transformation of plants and soil microorganisms (biotechnology research (no. 3))

Over the last fifty years plant breeders have achieved impressive improvements inyield, quality and disease resistance. These gains suggest that many more modifi-cations might be introduced if appropriate genes can be identified. Current DNAtechniques allow the construction of transgenic plants and this important newbook reviews the current state of knowledge.

A team of leading researchers provide in-depth reviews at the cutting edge oftechnology for laboratory techniques for transformation of important soilmicroorganisms and recalcitrant plants of economic value. The book is dividedinto three sections: soil microorganisms; cereal crops; and industrially importantplants. The most effective methods used to date are compared and their meritsand limitations are discussed. Some chapters emphasize case studies and applica-tions. In cases where obstacles remain to be overcome, an overview of progress todate is given.

The book will serve as a general guide and reference tool for those working ontransformation in microbiology and plant science.

PLANT AND MICROBIAL BIOTECHNOLOGY RESEARCH SERIES: 3Series Editor: James Lynch

Transformation of Plantsand Soil Microorganisms

PLANT AND MICROBIAL BIOTECHNOLOGY RESEARCH SERIESSeries Editor: James Lynch

Titles in the series

1. Plant Protein EngineeringEdited by P. R. Shewry and S. Gutteridge

2. Release of Genetically Engineered and Other MicroorganismsEdited by J. C. Fry andM. Day

3. Transformation of Plants and Soil MicroorganismsEdited by K. Wang, A. Herrera-Estrella and M. Van Montagu

Transformation of Plants andSoil Microorganisms

Edited by

Kan WangICI Seeds, Iowa, USA

Alfredo Herrera-EstrellaCentro de Investigacion y Estudios Avanzados, Mexico

and

Marc Van MontaguUniversiteit Gent, Belgium

CAMBRIDGEUNIVERSITY PRESS

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGEThe Pitt Building, Trumpington Street, Cambridge, United Kingdom

CAMBRIDGE UNIVERSITY PRESSThe Edinburgh Building, Cambridge CB2 2RU, UK40 West 20th Street, New York NY 10011-4211, USA477 Williamstown Road, Port Melbourne, VIC 3207, AustraliaRuiz de Alarcon 13, 28014 Madrid, SpainDock House, The Waterfront, Cape Town 8001, South Africa

http ://www. Cambridge. org

© Cambridge University Press 1995

This book is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,no reproduction of any part may take place withoutthe written permission of Cambridge University Press.

First published 1995

First paperback edition 2004

A catalogue record for this book is available from the British Library

Library of Congress cataloguing in publication data

Transformation of plants and soil microorganisms / edited by Kan Wang.Alfredo Herrera-Estrella, and Marc Van Montagu.

p. cm. - (Plant and microbial biotechnology research series ; 3)Includes bibliographical references and index.ISBN 0 521 45089 6 hardback1. Plant genetic transformation. 2. Crops — genetic engineering.3. Genetic transformation. 4. Soil microbiology. I. Wang, Kan,1959-. II. Herrera-Estrella, Alfredo. III. Van Montagu, Marc.IV Series.SB123.57.T7 1995631.5'23-dc20 94-11609 CIP

ISBN 0 521 45089 6 hardbackISBN 0 521 54820 9 paperback

And he gave it for his opinion, that whoever could make two earsof corn or two blades of grass to grow upon a spot of groundwhere only one grew before, would deserve better of mankind,and do more essential service to his country than the whole raceof politicians put together.

Jonathan Swift (1667-1745)Gulliver's Travels (1726), Chapter 7, 'Voyage to Brobdingnag'

This book is sponsored by ICI Seeds,a Business Unit of ZENECA INC.

Contents

List of Contributors xiSeries Preface xiiiPreface xvAcknowledgements xviAbbreviations and Terms xvii

Part I Transformation of Soil Microorganisms

1 Pseudomonas 3Gareth Warren, Joyce Loper, Dallice Mills and Linda Thomashow

2 Nocardioform and Coryneform Bacteria 10Jan Desomer and Marc Van Montagu

3 Agrobacterium, Rhizobium, and Other Gram-Negative Soil Bacteria 23Alan G. Atherly

4 Filamentous Fungi 34Gustavo H. Goldman, Marc Van Montagu and Alfredo Herrera-Estrella

Part II Transformation of Cereal Crops

5 Rice Transformation: Methods and Applications 53Junko Kyozuka and Ko Shimamoto

6 Maize 65H. Martin Wilson, W. Paul Bullock, Jim M. Dunwell, J. Ray Ellis, Bronwyn Frame,James Register III, and John A. Thompson

7 Barley, Wheat, Oat and Other Small-Grain Cereal Crops 81RalfR. Mendel and Teemu H. Teeri

Part III Transformation of Industrially Important Crops

8 Leguminous Plants 101JackM. Widholm

9 Spring and Winter Rapeseed Varieties 125Philippe Guerche and Catherine Primard

X CONTENTS

10 Sunflower 137Gunther Hahne

11 Forest Trees 150Ronald R. Sederoff

Index 164

Contributors

Alan G. AtherlyDepartment of Zoology and GeneticsRoom 2216, Molecular Biology BuildingIowa State UniversityAmes, IA 50011USA

W. Paul BullockICI SeedsResearch Department2369 330th Street, Box 500Slater, IA 50244USA

Jan DesomerLaboratoire CentralSolvay, S.A.Rue de Ransbeek 310B-1120 BrusselsBelgium

Jim M. DunwellZENECA SeedsJealotfs Hill Research StationBracknell, Berkshire RG12 6EYEngland

J. Ray Ellis (deceased)ZENECA SeedsJealott's Hill Research StationBracknell, Berkshire RG12 6EYEngland

Bronwyn FrameICI SeedsResearch Department2369 330th Street, Box 500Slater, IA 50244USA

Gustavo GoldmanUniversidade de Sao PauloFaculdade de Ciencias Farmaceuticas de Ribeirdo

PretoVia do Cafe SIN14040-903 Ribeirdo Preto, SPBrazil

Philippe GuercheLaboratoire de Biologie CellulaireInstitut National de la Recherche Agronomique78026 Versailles CedexFrance

Giinther HahneInstitut de Biologie Moleculaire des PlantesCentre National de la Recherche ScientifiqueUniversite Louis Pasteur12, rue du General Zimmer67084 Strasbourg CedexFrance

Alfredo Herrera-EstrellaCentro de Investigacion y Estudios AvanzadosUnidad IrapuatoDepartamento de Ingenieria GeneticaKm. 9.6 del Libramiento None Carretera Irapuato-

LeonApartado Postal 62926500 IrapuatoGto., Mexico

Xll CONTRIBUTORS

Junko KyozukaCSIRO Division of Plant IndustryGPO Box 1600Canberra ACT 2601Australia

Joyce LoperUSDA-ARS Horticultural Crops Research

Laboratory3420 N. W. Orchard AvenueCorvallis, OR 97330USA

Ralf R. MendelInstitute of BotanyTechnical University of BraunschweigHumboldtstr. 13300 BraunschweigGermany

Dallice MillsDepartment of Botany and Plant PathologyOregon State UniversityCorvallis, OR 97331USA

Catherine PrimardLaboratoire de Biologie CellulaireInstitut National de la Recherche Agronomique78026 Versailles CedexFrance

James Register IIIICI SeedsResearch Department2369 330th Street, Box 500Slater, IA 50244USA

Ronald R. SederoffDepartment of Forestry, Genetics and BiochemistryNorth Carolina State UniversityRaleigh, NC 27695USA

Ko ShimamotoLaboratory of Plant Molecular GeneticsNara Institute of Science and Technology8916-5 TakayamaIkomaNara 630-01Japan

Teemu TeeriInstitute of BiotechnologyUniversity of HelsinkiKarvaamokuja 3FIN-00380 HelsinkiFinland

Linda ThomashowUSDA-ARS Root Disease and Biological Control

Research Unit367 Johnson HallWashington State UniversityPullman, WA 99164USA

John A. ThompsonZENECA SeedsJealott's Hill Research StationBracknell, Berkshire RG12 6EYEngland

Marc Van MontaguLaboratorium voor GeneticaUniversiteit GentLedeganckstraat 35B-9000 GentBelgium

Kan WangICI SeedsResearch Department2369 330th Street, Box 500Slater, IA 50244USA

Gareth WarrenDNA Plant Technology Corporation6701 San Pablo AvenueOakland, CA 94608-1239USA

JackM. WidholmDepartment of AgronomyUniversity of Illinois at Urbana-ChampaignW-203 Turner Hall1102 South Goodwin AvenueUrbana, IL 61801-4798USA

H. Martin WilsonICI SeedsResearch Department2369 330th Street, Box 500Slater, IA 50244USA

Series PrefacePlant and Microbial Biotechnology

The primary concept of this series of books is toproduce volumes covering the integration of plantand microbial biology in modern biotechnologicalscience. Illustrations abound, for example thedevelopment of plant molecular biology has beenheavily dependent on the use of microbial vectors,and the growth of plant cells in culture has drawnlargely on microbial fermentation technology. Inboth of these cases the understanding of microbialprocesses is now benefiting from the enormousinvestments made in plant biotechnology. It isinteresting to note that many educational institu-tions are also beginning to see things this way andintegrating departments previously separated byartificial boundaries.

Having set the scope of the series, the nextobjective was to produce books on subjects thathad not already been covered in the existing litera-ture and, it was hoped, to set some new trends.

One of the most commonly used techniques togenetically engineer both plants and micro-organisms is transformation. However, it seemedto me that, whereas transformation was of coursecovered in all molecular biology textbooks, a sub-stantive research monograph that would cover thisexciting and expanding field was not available.

The Genetics Laboratory at the University ofGent, under the direction of Marc Van Montagu,has been an earlier 'player' and is now a world

leader in transformation. Particularly, theirinvolvement in the characterization of the Ti(tumor inducing) plasmid from a soil bacterium(Agrobacterium tumefaciens) and its use in thetransformation of plants set one of the first scenesfor plant molecular biology to emerge. Marc VanMontagu was, therefore, an obvious choice as avolume editor. One of the great strengths of theGent Laboratory has been its international flavor.As I attempted to delve into the transformation ofTrichoderma, I found that the Gent team wasalready there and I was delighted to meet AlfredoHerrera-Estrella, who was active in the field inGent. He was an obvious choice to join the editor-ial team on his return to Mexico. Then to com-plete the team, Marc and Alfredo suggested thatKan Wang, who had completed her Ph.D. in Gentand now leads a crop transformation team inZENECA/ICI SEEDS in the United States ofAmerica, should take the editorial lead.

This editorial team has persuaded an outstand-ing international group to produce an excellentcollection of accounts of transformation in plantsand soil microorganisms. It should provide a goodstimulus to accelerate the pace of development ofagricultural and environmental biotechnology.

Jim Lynch

Preface

Advances in biology continue to be made at astriking and ever increasing rate, especially sincethe powerful technique of gene manipulationbecame available. For plant biologists, however,the possibility of engineering plants was not con-sidered until the mid 1970s, when gene transfermediated by the soil bacterium Agrobacteriumtumefaciens was discovered. From this moment on,enormous developments have occurred concern-ing not only practical aspects such as cropimprovement but also fundamental aspects relat-ing to the understanding of the biology of plants intheir interactions with the environment. One ofthe most important subjects is perhaps that ofplant-microbe interactions.

For either applied or academic research, it isessential to establish efficient transformation sys-tems. Although the basic transformation tech-niques have been well established for modelsystems such as Escherichia colt, yeast and tobacco,many important species have proven more diffi-cult to manipulate. During the past several years,many research groups took up the challenge oftransformation and have made significant pro-gress. To keep abreast with this rapid develop-ment, we have invited in this volume severalleading groups to share their experiences withprospective and practicing researchers in themicrobe/plant field. Together, we have endeav-ored to give readers details of one particular prob-lem of major importance - transformation of soilmicroorganisms and recalcitrant crops.

Part I consists of four chapters on the transfor-mation of soil microorganisms. Each chapterdescribes most, if not all, techniques used fortransformation in the laboratory of a specificmicrobe. In some cases conjugation, a naturalDNA transfer process, is widely used for trans-

formation; in other cases either a physical treat-ment such as electroporation or a physical-chemi-cal treatment such as polyethylene glycol-CaCl2 isused as an effective method. A comparison ofthese techniques and key points of each proce-dures are provided.

Many plant tissue culture systems, unlike theiranimal counterparts, allow for the regeneration ofwhole organisms. However, a number of plants,especially most of the economically valuable cropsand trees cannot be easily manipulated in culture.Part II covers transformation of cereal crops; threechapters deal with transformation of the mostimportant monocotyledonous plants - rice, maize,barley, wheat, etc. Great progress has been madeover the past several years in these areas and trans-formation of these crops is now becoming routinein a number of laboratories. Progress in the trans-formation of recalcitrant dicotyledonous plantsand woody species has also been significant. Fourchapters in Part III provide up to date informationon some industrially important plants, such assoybean, rapeseed, sunflower, and forest trees. Allthe chapters review current transformation tech-niques. The most effective methods are empha-sized, technical problems are highlighted andpotential applications are discussed.

We hope that this book encourages all of us tokeep putting forward our best efforts and that itinspires new people to enter this field of biotech-nology. We should not forget that the great impactof gene manipulation in more recent develop-ments in agriculture has yet to be seen by thegeneral public.

K. Wang, A. Herrera-Estrella and M. Van Montagu

xv

Acknowledgements

We thank G. Angenon, A. Caplan, M. De Block, are due to A. Uytterhaegen for her efficiency andR. Deblaere, W. Dillen, J. Desomer, K. patience during the preparation of this book.D'Halluin, G. Gheysen, E. Gobel, M. Holsters Finally, we are grateful to R. Harington ofand D. Inze for critical reading of the manuscripts, Cambridge University Press for his constant sup-and M. De Cock for typing tables. Special thanks port in the editing process.

xvi

Abbreviations and Terms

AABAABR

Ac

actactlachenesADHadhl

ADPagroinfection

ALS

alsamdS

Apaph

argB

aroA

ARS

attPAVGA188A188 X B73

haploid genome of cabbageabscisic acidantibiotic resistancemaize autonomous transposableelement activatoractin genea rice gene for actinsunflower kernelsalcohol dehydrogenasea gene for alcoholdehydrogenaseadenosine diphosphatetransfer of viral DNA viaAgrobacteriumpromoter from the mutatedacetolactate synthase gene fromArabidopsis thalianaacetolactate synthase geneacetamidase gene fromAspergillus nidulansampicillinaminoglycosidephosphotransferase geneornithine carbamoyl transferasegene from Aspergillus nidulans5-enolpyruvylshikimate synthasegene; enzyme is active in thesynthesis of aromatic aminoacidsautonomously replicatingsequencesphage attachment siteaminoethoxyvinyl glycinemaize inbred lineFl hybrid made between inbredlines Al88 and B73

B5BAbar

BC1BCGbenA3

bgll

ble

bml

bmlR

bmlR3

BMS

bpbzlB73CCcab

CaMVCaMV 35S

CaMV 19S

CAT

B5 medium6-benzyladeninephosphinothricin acetyltransferase gene fromStreptomyces hygroscopicusresult of the first backcrossbacille Calmette-Guerinp-tubulin A3 benomyl resistancegene from Neurospora crassaP-glucosidase gene fromTrichoderma reeseitransposon 5 bleomycinresistance gene(3-tubulin benomyl resistancegene from Aspergillus nidulansP-tubulin benomyl resistancegene(3-tubulin R3 benomyl resistancegene from ColletotrichumgraminicolaBlack Mexican Sweet (maizecultivar)base-pair(s)bronze 1 genemaize inbred linehaploid genome of turnipcapacitancechlorophyll alb binding proteingenecauliflower mosaic viruscauliflower mosaic virus 35SRNAcauliflower mosaic virus 19SRNAchloramphenicol acetyltransferase

XVll

XV111 ABBREVIATIONS AND TERMS

cat

cat!Cbcbhl

cDNAcf.u.CmcmsColElcoscosduction

CPCPMVCScv.CVS.

2,4-DDGTdhfrdicot(s)DNADs

EDTAelectroduction

EmEPEPSP

EPT

FiG418

GA3

glaAghAgoxAGUSgusAHfrHmR

H-NMR

hph

HPT

chloramphenicol acetyltransferase genecatalase 1 genecarbenicillincellobiohydrase I gene fromTrichoderma reeseicomplementary DNAcolony-forming unitschloramphenicolcytoplasmic male sterilecolicin Elbacteriophage X cohesive endstransduction of plasmids whichcarry cos sequencescoat proteincowpea mosaic viruschlorsulfuroncultivarcultivars2,4-dichlorophenoxyacetic aciddirect gene transferdihydrofolate reductase genedicotyledonous plant (s)deoxyribonucleic acidnon-autonomous dissociationelementethylenediaminetetra-acetic acidelectric pulse-mediated transferof plasmids directly betweenbacterial strainsearly methionine geneelectroporationenolpyruvylshikimate-3-phosphateefficient plasmidtransformationfasciation-inducingsynthetic aminoglycosideantibiotic, geneticingibberellic acidglucoamylase genecitrate reductase geneglucose oxidase gene(3-glucuronidasealternative name for uidAhigh frequency of recombinationhygromycin resistancehigh resolution nuclear magneticresonancehygromycin phosphotransferasegene from Escherichia colihveromvcin nhosphotransferase

hsp60

Hup

HygIBAIgnite

inaIna+

Ina"IncP, Q, W

INHintKanKan 15kbKmKmR

lacZ

LBIeu2

luclysA

MIJmobmobv.TnS

monocot(s)mRNAMSMtMtxR

4-MUMu\N6NAAraaDnifnodNOSnosNPTIInptll

NT1

60 kilodalton heat shock proteingenehydrogen uptake-relatedhydrogenasehygromycinindolebutyric acidherbicide containingphosphinothricin as activeingredientice nucleation geneice nucleation activeice nucleation inactiveincompatibility group P, Q andWisonicotinic acid hydrizideintegrase genekanamycinkanamycin (15 mg/1)1000 base-pairskanamycinkanamycin resistanceEscherichia coli gene forP-galactosidaseLuria-Bertani mediumP-isopropyl malatedehydrogenase gene fromSaccharomyces cerevisiaeluciferase gene from fireflyw-diaminopimelatedecarboxylase genemicroinjectionmobilization genestransposon 5 carrying plasmidmobilization functionsmonocotyledonous plant(s)messenger ribonucleic acidMurashige and Skoog mediummillions of tonsmethotrexate resistance4-methylumbelliferonea maize transposable elementplant mediumnaphthaleneacetic acidnitrate reductase genenitrogen fixation genenodulation genenopaline synthase promoternopaline synthase geneneomycin phosphotransferase IItransposon 5 neomycinphosphotransferase genecell line of Nicotiana tabacum

ABBREVIATIONS AND TERMS XIX

ocs

oliCR

orioriT

oscp-1pl5AparParoPAT

pat

pcbC

PCRPCVPDApda

PEGPenGp.f.u.pgallPhleopin2

polAPPT, Bastaproj.proto.psbAp.s.i.PTR

pyr\

pyrF

pyrG

Q4qa-2

RRO

Agrobacterium gene for octopinesynthase on a Ti-plasmidoligomycin resistance gene fromAspergillus nigerreplication origintransfer originosmotically sensitive cellincompatibility groupincompatibility grouppartitioning function genesparomomycinphosphinothricinacetyltransferasephosphinothricinacetyltransferase geneisopenicillin synthetase genefrom Cephalosporium acremoniumpolymerase chain reactionpacked cell volumepisatin-demethylating abilitypisatin demethylase gene fromNectria haematococcapolyethylene glycolpenicillin Gplaque-forming unitspre-progalacturonidase genephleomycinproteinase inhibitor II gene frompotatoDNA polymerase Aphosphinothricinmicroprojectile bombardmentprotoplast direct DNA uptakesoybean atrazine resistance genepounds per square inchT-DNAl 'and2 'genepromoters from Agrobacteriumtumefaciensorotidine 5'-monophosphatedecarboxylase gene fromTrichoderma reesei and Neurosporacrassaorotidine 5'-monophosphatedecarboxylase geneorotidine 5'-monophosphatedecarboxylase gene fromAspergillustransducing bacteriophage Q4catabolic dehydroquinase genefrom Neurospora crassaresistanceprimary transformants

Rl

rab 16A

RAPD

rbcS

RFLP

RiRIProlC

rRNARSVRubisco

Ssh\shlSmSpspoCl

spoCICSV40sull

T

TcT-DNATiTLTMVTn5TnSseql

Tn903TRtratrpC

tub2R

ubiuidAura3

progeny of primarytransformantsa rice abscisic acid responsivegenerapid amplified polymorphicDNAribulose-1,5-bisphosphatecarboxylase oxygenase smallsubunit generestriction fragment lengthpolymorphismroot inducingrepeat-induced point mutationa pathogenesis-related gene oftheTL-DNAoftheAgrobacterium rhizogenes Riplasmidribosomal ribonucleic acidrice stripe virusribulose-135-bisphosphatecarboxylase oxygenasespring varietiesshrunken 1 genea maize sucrose synthase genestreptomycinspectinomycinsporulation-specific geneclustersporulation-specific C genesimian virus 40enterobacteria sulfonamideresistance genetime constanttetracyclinetransferred DNAtumor inducingleft region of the T-DNAtobacco mosaic virusbacterial transposon 5deletion in transposon 5streptomycin resistance genetransposon 903right region of the T-DNADNA transfer genesanthranilate synthase gene fromAspergillus nidulansbenomyl-resistant p-tubulingene 2 from Trichoderma viridea maize gene for ubiquitingusA or p-glucuronidase geneorotidine 5'-monophosphatedecarboxylase gene

XX ABBREVIATIONS AND TERMS

ura5

virvir

y B, C3 E3

F, GvirD4

W

orotate phosphoribosyltransferase genevirulence functionvirulence-related operon on theTi plasmid of Agrobacteriumtumefaciensvirulence operons A, B, C, E, F,G from Agrobacterium tumefaciensvirulence gene D4 fromAgrobacterium tumefacienswinter variety

WTX-Gluc

yA

2 \i circle

70

wild-type5-bromo-4-chloro-3-indolylglucuronide, substrate for(3-glucuronidaseAspergillus nidulans yellow sporemutationSaccharomyces cerevisiaeminichromosomeCaMV 35S RNA promoter withduplicated enhancer

PART I TRANSFORMATION OFSOIL MICROORGANISMS

IPseudomonas

Gareth Warren, Joyce Loper, Dallice Mills and Linda Thomashow

Introduction

Some pseudomonads exist in close associationwith plants. Best characterized are the phyto-pathogenic strains that are epiphytic colonists.Some of these are specific in their ability to infectonly one plant species or are limited to particularcultivars within a species, as is true for somePseudomonas syringae pathovars. Other pseudo-monads colonize the rhizosphere. Best studiedamong the rhizosphere colonizers are strains thatbenefit plant health by their antagonism of patho-genic potential colonists (Defago & Haas, 1990).However, not all rhizosphere pseudomonads areof this sort: certain strains are neutral or even dele-terious to the host plant.

Genetic manipulation is indispensible forthe study of the interactions between plantand pseudomonad (Lindow, Panopoulos &McFarland, 1989). Introducing DNA into thepseudomonad is sometimes problematic: it is thesubject of this chapter. The techniques developedfor plant-related pseudomonads, based on priorexpertise, have been those of conjugal transfer anddirect transformation.

Transformation of Pseudomonas

Direct transformation of Pseudomonas syringae andP. fluorescens has been employed successfully(Mukhopadhyay, Mukhopadhyay & Mills, 1990),but optimal protocols may differ considerablyfrom strain to strain. The effects of divalentcations used in inducing competence for transfor-mation are highly strain and species dependent.

Electroporation is now an alternative to treat-ment with divalent cations for inducing DNAuptake; optimal protocols are probably less strainspecific and the technique is potentially more effi-cient. In the future it may become the preferredway for introducing DNA into Pseudomonas. Itspracticability has been demonstrated with Pseudo-monas aeruginosa (Smith & Iglewski, 1989; Smithet al., 1990), P. chlororaphis and P. oxalaticus(Wirth, Friesenegger & Fiedler, 1989), and P.putida (Fiedler & Wirth, 1988; Trevors &Starodub, 1990).

Conjugal transfer to Pseudomonas

Conjugation has usually been the method ofchoice for introducing DNA into pseudomonads.Conjugal receptor function appears to be univer-sal among Gram-negative bacteria, and conjuga-tion techniques can often be adapted with relativeease to a new strain. It is thought that the single-stranded mode by which DNA enters the conjugalrecipient reduces the probability of recognition bythe recipient's restriction system, in comparisonto the double-stranded mode of DNA entry that isusual in the transduction or transformation ofGram-negative bacteria.

The donor in conjugations with Pseudomonas isalmost always Escherichia coli, because the latter isamenable to efficient direct transformation and isa reliable host for the plasmid vectors commonlyutilized. When DNA appears in the conjugalrecipient cell and causes a heritable change in itsphenotype, that recipient is termed a transcon-jugant.

4 WARREN et al.

Some plasmids are able to replicate in a varietyof bacterial hosts, including E. coli and Pseudo-monas. They are said to possess a broad hostrange. When conjugation involves a broad hostrange plasmid, most transconjugants are recipi-ents in which the plasmid is now established as anew replicon.

Other plasmids have a narrow host range; forexample, ColEl and pBR322 are limited toenteric bacteria (e.g. E. coli) by their inability toreplicate in other types. The ability to be trans-ferred during conjugation, however, does notdepend on the ability to replicate in the recipient.Thus, narrow host range plasmids can be utilizedto carry DNA into Pseudomonas, and since they donot persist as independent replicons, selection forthe transferred DNA selects transconjugants inwhich the transferred DNA has recombined withthe recipient genome. In this use, narrow hostrange plasmids are termed suicide vectors. Twonarrow host range plasmids (one limited to E. coli,the other to Pseudomonas) can be joined to form aplasmid able to replicate in both (Van den Eede etal, 1992): vectors constructed on this principleare sometimes known as shuttle vectors.

Conjugative ('self-transmissible') plasmidssuch as RK2 (also known as RP1, RP4, and R68;Palombo et al, 1989) cause their host bacteriumto conjugate. They have evolved this ability to per-mit transfer of their own DNA; however, it shouldbe remembered that conjugation and DNA trans-fer (mobilization) are separate functions thatrequire distinct sets of genes. The products of themobilization genes recognize the transfer origin(oriT) from which the physical transfer of theplasmid DNA begins. Conjugative plasmids arenecessary for conjugation but are inconvenient asvectors, because the conjugation genes occupyrelatively large regions of DNA. Transmissible('mobilizable') plasmids do not cause conjugationbut can be transferred when conjugation occurs.They are generally used as vectors in combinationwith a coresident conjugative plasmid. In nature,transmissible plasmids such as RSF1010 (Derby-shire, Hatfull & Willetts, 1987) carry their owndistinct oriT together with a set of mobilizationgenes whose products recognize that origin specif-ically. However, transmissible vectors withoutmobilization genes can be constructed by makinguse of the oriT from a conjugative plasmid. Thislimits them to mobilization by just one type ofconjugative plasmid (a circumstance not favored

by evolution, but adequate for experimental pur-poses). DNA mobilization and specificity wereaddressed in an excellent review by Willetts &Wilkins(1984).

Many of the vectors in common use are trans-missible, broad host range plasmids that aredeleted derivatives of the incompatibility group P(IncP) plasmid RK2: they retain oriT and thetetracycline resistance gene of RK2. PlasmidpRK290 is a prototype of such IncP vectors (Dittaet al, 1985). They are most commonly mobilizedby pRK2013, a conjugative plasmid that is alsoderived from RK2 but lacks tetracycline resistanceand possesses a narrow host range (Figurski &Helinski, 1979). Other broad host range vectorshave been derived from the transmissible IncQplasmid RSF1010 (Bagdasarian et al, 1981) orfrom conjugative IncW plasmids (Leemans et al,1982).

Many suicide vectors, including the pSUPseries (Simon, Priefer & Puhler, 1983), employthe oriT of RK2 and utilize replicons of the ColElor pl5A incompatibility groups. These are mobil-ized by pRK2013 or a 'mobilized strain in whichpart of a conjugative plasmid is integrated into thebacterial chromosome (Simon et al., 1986).Plasmid pBR322, which contains an oriT, hasbeen used as a suicide vector by supplying ColEl -type mobilization functions in trans (Van Haute etal, 1983).

When a suicide plasmid is transferred, selectionrequires recombination in order to form the trans-conjugant. Since such recombination is a rareevent, success requires a higher efficiency of con-jugal DNA transfer than is necessary for selectionof recipients of broad host range plasmids.

Efficiency of transfer varies considerably, de-pending on the identity of the recipient strain, theages of the donor and recipient cultures, and howlong the cultures are incubated together. In gen-eral, stationary-phase recipient cultures haveproven more satisfactory than log-phase culturesbut log-phase cultures are preferred for donorstrains (including the helper if the mating is tri-parental). In a typical procedure, cells are washedby centrifugation, mixed on a nitrocellulose filterat a donorrrecipient ratio of 1:1 or 1:5 andincubated for intervals ranging from 5 h(Hamdan, Weller & Thomashow, 1991) to 48 h(Thomashow & Weller, 1988) on Luria Bertani(LB) agar at 28 °C. Cultures are then washedfrom the filters into sterile liquid and plated

Pseudomonas 5

directly on to selective media. Alternatively, cellsfrom individual donor colonies have been trans-ferred with a metal replicator to plates of LB agarseeded with approximately 108 recipient cells (andif the cross is triparental, seeded also with 107 cellsof the helper); after 48 h, transconjugants arereplicated to selective media.

Case studies

Pseudomonas syringae pv. syringae J900Strain J900 is an epiphytic pathogen of beans.DNA introduction into J900 has been used toinvestigate its pathogenic interaction with theplant. A novel strategy for bringing in DNA wasneeded because IncP vectors are unstable duringgrowth of J900 in planta, so that after a period of 1week, 75% of the cells may not contain the plas-mid. Moreover, strains containing IncP plasmidssometimes exhibit retarded growth and attentu-ated virulence.

The first element of the strategy was the con-struction of a shuttle vector capable of stable repli-cation in the target strain, utilizing an origin ofreplication from a narrow host range plasmidindigenous to P. syringae. The replication region(ori) from the cryptic plasmid pOSU900 (Mukho-padhyay et al., 1990), harbored by J900, wasselected because it was expected to function invarious pathovars of P. syringae. ori was recog-nized, among various fragments of pOSU900cloned in pBR322, by its ability to replicate in,and thereby confer pBR322-derived antibioticresistance on, a plasmidless Pseudomonas strain.

Direct transformation of Pseudomonas wasattempted using a CaCl2/heat-shock protocoldeveloped for E. coli (Maniatis, Fritsch & Sam-brook, 1982). Vectors containing the BarriHL frag-ment on which ori was originally cloned yieldedapproximately 2000 transformants per microgramof DNA. (An IncP vector also gave transformantsat approximately the same frequency: D. Mills &Y. Zhao, unpublished data). However, deletionsmade within the on-containing fragment resultedin plasmids that gave increased transformationfrequencies, with maximum levels approaching106 transformants per microgram of DNA. Thedeletion derivatives had copy numbers inPseudomonas up to six-fold higher than the originalplasmid, and the derivatives most efficient intransformation were those with the higher copy

numbers. It is not obvious why this should be so.The original pBR322-on plasmid was stably

maintained in P. syringae, with 98% of cells retain-ing the plasmid after a 2-week period of growth inplanta. However, the stability of the deletionderivatives in P. syringae was extremely variable,and ranged from approximately 1% to 100% dur-ing a similar growth period in planta. Those deriv-atives that were stable in P. syringae could alsotransform and be maintained without antibioticselection in P. fluorescens B10 (Kloepper, Schroth& Miller, 1980) and Pf-5 (Howell & Stipanovic,1979) at frequencies ranging from 103 to 105 permicrogram of DNA.

The inability to recover any transconjugants ofAgrobacterium tumefaciens, Rhizobium meliloti or ofa polA mutant of E. coli indicates that the oricloned from pOSU900 has a narrow host range.This may be useful for limiting unwanted dissemi-nation of the vector during studies in the eco-system.

Pseudomonas fluorescens strain MS 1650

Strain MS 1650 is an ice nucleation active (Ina+)epiphytic colonist (Warren et al> 1987). It wasconsidered desirable to introduce a defective inagene into MS 1650 for the purpose of generatingIna~ derivatives for experiments in the biologicalcontrol of frost damage by competitive exclusion.

Experiments with broad host range plasmidsshowed that MS 1650 was a relatively poor recipi-ent for conjugal transfer from E. coli donors.Suicide vectors utilizing the transfer origin ofpBR322 did not yield detectable numbers oftransconjugants. Therefore the defective ina genewas cloned into a vector that had been observed togive more efficient transfer than pBR322 into afew other strains of Pseudomonas (Warren,Corotto & Green, 1985). The new vector,pLVC18, contained the mobilization system ofthe broad host range IncQ plasmid RSF1010 butlacked its replicative origin so that suicide muta-genesis was still possible. Conjugative functionswere provided by an IncP helper plasmid. Itbecame possible to isolate very small numbers oftransconjugants with this system - enough to pro-ceed with the marker exchange strategy.

The availability of an alternative mobilizationsystem with greater efficiency allowed selection ofthe desired recombination events. One conjuga-tion/mobilization system may not be universally

6 WARREN et al

more efficient than all others, so it is necessary totest various systems empirically.

Pseudomonas fluorescens strain Pf-5

Strain Pf-5 is a biological control agent of plantdiseases caused by the soil-borne fungi Pythiumultimum (Howell & Stipanovic, 1980) andRhizoctonia solani (Howell & Stipanovic, 1979).Strain Pf-5 produces the antibiotics pyoluteorin,pyrrolnitrin, and 2,4-diacetylphloroglucinol. Ithas been desirable to introduce DNA into Pf-5 toanalyze the role of antifungal compounds in thesuppression of plant disease by this pseudo-monad.

Initial attempts to introduce plasmids via con-jugation with E. coli donors were unsuccessful.Although E. coli is relatively resistant to pyrrol-nitrin and 2,4-diacetylphloroglucinol, it is sensi-tive to pyoluteorin (Takeda, 1958). It could bedemonstrated that the E. coli donors were killedon the mating plates. Since Pf-5 is viable at 37 °Cbut does not produce pyoluteorin at this tempera-ture, coculture of Pf-5 with E. coli was attemptedat 37 °C. The E. coli donors survived and ade-quate numbers of transconjugants were recovered.Using this procedure, the suicide transposon vec-tor pLG221 (Boulnois et al> 1985) was used togenerate Tn5 mutants of Pf-5 that were deficientin pyoluteorin production (Kraus & Loper, 1992).Matings at 37 °C have also been used to increasethe recovery of transconjugants of P. aeruginosaLEC1, a soil isolate that suppresses Septoria triticiblotch of wheat (Flaishman et al., 1990).

The crucial step in the development of a work-ing protocol for conjugal transfer of DNA intoPf-5 was the recognition of a potential cause offailure with matings conducted under the usualconditions - in this case, the toxicological incom-patibility of donor and recipient during coculture.In other cases, cultural incompatibility has beenameliorated by other means, for example manipu-lation of the media formulation or selection of aresistant donor strain.

Pseudomonas fluorescens strain HV37a

HV37a is a root-colonizing pseudomonad thatinhibits the growth of several fungal pathogens(Gutterson et al.y 1988). It was desirable to intro-duce DNA for marker exchange to manipulateproduction of antifungal antibiotics, and see how

this influenced the interaction of the bacteria withthe plant and its fungal pathogens.

Transconjugants arose when broad host rangeIncP plasmids were mobilized into HV37a by stan-dard conjugation protocols. However, the low fre-quencies with which they were obtained made itunsurprising that suicide plasmids did not yielddetectable numbers of transconjugants. An alterna-tive method of obtaining marker exchange makesuse of the instability of some broad host range IncPvectors in HV37a (Jones & Gutterson, 1987).When selection for these vectors is relaxed, plas-midless strains arise in a few generations. If theunstable vector carries a chromosomal gene with adistinguishable allele, then after plasmid loss adetectable proportion of the resulting strains willhave substituted the plasmid-borne allele for theoriginal chromosomal allele. In some trans-conjugant cells, homologous recombinationbetween plasmid and chromosome will occur andcause their cointegration. Subsequent excisiverecombination can now leave an originallyplasmid-borne allele in the chromosome. Why doesan unstable vector favor the recovery of markerexchanges? Presumably, the instability provides aninternal selection for the cointegration event.

This strategy illustrates an alternative means ofintroducing specific alterations into the Pseudo-monas genome. It is probably more time consum-ing than the use of suicide plasmids, where thelatter is feasible. The plasmids that are unstable inHV37a may not be sufficiently unstable in otherpseudomonads, but it is likely that equivalent onescould be constructed. Instability or conditionalstability can be generated in most plasmids bymaking appropriate deletions or by selectingtemperature-sensitive mutations.

Pseudomonas fluorescens strain CHAO

Strain CHAO is a rhizobacterium that suppressestobacco black root rot caused by Thielaviopsis basi-cola. It excretes several metabolites with antifungalproperties, including hydrogen cyanide (Voisardet al, 1989). Introduction of DNA into this strainhas been desirable for genetic analysis, in particu-lar for transposon mutagenesis to identify thegenes for antifungal traits and test their impor-tance in disease suppression.

Strain CHAO has proved refractory to conjugalreception of the broad host range IncP plasmidRK2, and also to reception of other plasmids,

Pseudomonas 7

both broad host range and suicide, when mobi-lized by RK2 or pRK2013. A large deletion inRK2, including its primase gene, increased theefficiency of recovering CHAO transconjugants bytwo to three orders of magnitude (Voisard, Rella& Haas, 1988). The deletion derivative also mobi-lizes other oriTxju vectors at concomitantly higherefficiencies, enabling transposon mutagenesiswith suicide vectors.

The enhancement effect has not been defini-tively ascribed to the removal of the RK2 primasegene (and it should be noted that the primase geneis helpful in some types of interspecies con-jugation). However, some alternative explanationscan be eliminated. The effect cannot be due torestriction, since it extends to the mobilization ofother plasmids, and also to transfer of RK2between strains of CHAO. Likewise, the effectcannot be due to changes in plasmid maintenance,because RK2 replicates stably in CHAO onceintroduced. The effect is likely to be exerted onthe conjugal donor, either by affecting the struc-ture of mating aggregates or by changing someattribute of the DNA to be transferred (forexample, altering the make-up of the leader pro-teins).

The RK2 deletion that enhances transfer toCHAO also provides a lesser enhancement fortransfer into P. fluorescens S9, and would certainlybear testing with other recalcitrant pseudomon-ads. Other types of plasmid mutation couldenhance transfer to other strains. Mutagenesis ofthe vectors used for transfer, or of the conjugativeplasmids used to mobilize them, would appear tobe a reasonable strategy for other cases. The dis-covery of this means of improving transfer effi-ciency to CHAO adds to the diversity of empiricalapproaches brought to bear on the problem.

Pseudomonas fluorescens strain 2-79 and P.aureofaciens strain 30-84

Strains 2-79 and 30-84 colonize the roots of wheatand provide protection against take-all, an impor-tant soil-borne disease caused by the fungalpathogen Gaeumannomyces graminis var. tritici.Both strains produce the antibiotic phenazine-1-carboxylic acid, and strain 30-84 also producestwo hydroxylated phenazine derivatives. Mani-pulation of the genes involved in antibiotic bio-synthesis has been desirable to evaluate the role ofthe antibiotics both in disease suppression and as

a factor that may contribute to competitiveness inrhizosphere colonization.

The phenazine-producing strains have provenrecalcitrant as genetic recipients in conjugationlargely because phenazines are toxic to E. coli.Viability was reduced in donor populations by up tothree orders of magnitude within the first hour aftermixing with a culture of Pseudomonas fluorescens2-79 (L. Thomashow, unpublished data). Conjugalefficiency was increased by selecting spontaneousmutants of a donor strain that could grow inthe presence of phenazine-1-carboxylic acid(Thomashow & Weller, 1988) and by briefly expos-ing recipient strains to 10 mM Tris (pH 8) buffercontaining 1 mM ethylenediaminetetra-acetic acid(EDTA). Both Tris and EDTA, but especially thelatter, are disruptive to the outer membrane ofPseudomonas^ and prolonged exposure causes sub-stantial cell lysis, as indicated by increased viscosityof the culture. It is not known whether theTris/EDTA treatment facilitates conjugation byremoving a physical barrier associated with therecipient cell envelope or if metabolic processes,including phenazine synthesis, are affected.

It was not possible to select donor strains resis-tant to the mixture of three phenazines producedby P. aureofaciens 30-84. However, supplementa-tion of the media used for conjugation with either100 \xbA ferric ammonium citrate or 10 mM p-aminobenzoic acid suppressed phenazine biosyn-thesis and enabled the recovery of transconjugants(Pierson & Thomashow, 1992).

Conclusions

The majority of the cases considered aboveinvolve conjugal transfer as the means of introduc-ing DNA into a plant-associated pseudomonad.The efficiency of conjugal transfer depends onmany factors: restriction systems in the recipient,anti-ii. coli toxins produced by the recipient, andconjugal receptivity (which may vary considerablyfor different types of conjugative plasmid in thedonor). It is not currently possible to predict con-jugal receptivity; experimentation with variousconjugation and mobilization systems is recom-mended. However, the presence of restriction sys-tems and the production of toxins are amenable toexperimental investigation and manipulation;knowledge of these characteristics can be used toimprove transconjugant recovery frequencies.

8 WARREN et al.

Direct transformation has an appealing sim-plicity as a means of introducing DNA intoPseudomonas. When successful, it is also fasterthan the use of conjugal transfer. Electroporationis likely to be used increasingly as a means ofdirect transformation because its applicability isprobably quite general.

Acknowledgement

We thank Jennifer Kraus for excellent editorialsuggestions.

References

Bagdasarian, M., Lurz, R., Rukert, B., Franklin,F. C. H., Bagdasarian, M. M., Frey, J. & Timmis,K. N. (1981). Specific-purpose plasmid cloningvectors. II. Broad host range, high copy number,RSFlOlO-derived vectors, and a host-vector systemfor gene cloning in Pseudomonas. Gene, 16, 237-247.

Boulnois, G. J., Varley, J. M., Sharpe, G. S. &Franklin, F. C. H. (1985). Transposon donorplasmids, based on Collb-P9, for use in Pseudomonasputida and a variety of other Gram negative bacteria.Molecular and General Genetics, 200, 65-67.

Defago, G. & Haas, D. (1990). Pseudomonads asantagonists of soilborne plant pathogens: modes ofaction and genetic analysis. In Soil Biochemistry, Vol.6, ed. J. M. Bollag & G. Stotzky, pp. 249-291.Marcel Dekker: New York & Basel.

Derbyshire, K. M., Hatfull, G. & Willetts, N. S.(1987). Mobilisation of the non-conjugative plasmidRSF1010: a genetic and DNA sequence analysis ofthe mobilisation region. Molecular and GeneralGenetics, 206, 161-168.

Ditta, G., Schmidhauser, T., Yakobson, E., Lu, P.,Liang, X. W., Finlay, D., Guiney, D. & Helinski,D. R. (1985). Plasmids related to the broad hostrange vector, pRK290, useful for gene cloning andfor monitoring gene expression. Plasmid, 13, 149-153.

Fiedler, S. & Wirth, R. (1988). Transformation ofbacteria with plasmid DNA by electroporation.Analytical Biochemistry, 170, 38-44.

Figurski, D. H. & Helinski, D. R. (1979). Replicationof an origin-containing derivative of plasmid RK2dependent on a plasmid function provided in trans.Proceedings of the National Academy of Sciences, USA,76, 1648-1652.

Flaishman, M., Eyal, Z., Voisard, C. & Haas, D.(1990). Suppression of Septoria tritici by phenazine-or siderophore-deficient mutants of Pseudomonas.Current Microbiology, 20, 121-124.

Gutterson, N., Ziegle, J. S., Warren, G. J. & Layton,T. J. (1988). Genetic determinants for cataboliteinduction of antibiotic biosynthesis in Pseudomonasfluorescens HV37a. Journal of Bacteriology, 170,380-385.

Hamdan, H., Weller, D. M. & Thomashow, L. S.(1991). Relative importance of fluorescentsiderophores and other factors in biological controlof Gaeumannomyces graminis var. tritici byPseudomonas fluorescens strains 2-79 and M4-80R.Applied and Environmental Microbiology, 57,3270-3277.

Howell, C. R. & Stipanovic, R. D. (1979). Control ofRhizoctonia solani on cotton seedlings withPseudomonas fluorescens and an antibiotic producedby the bacterium. Phytopathology, 69, 480-482.

Howell, C. R. & Stipanovic, R. D. (1980).Suppression of Pythium ultimum-m&uczd damping-off of cotton seedlings by Pseudomonas fluorescens andits antibiotic, pyoluteorin. Phytopathology, 70,712-715.

Jones, J. D. G. & Gutterson, N. (1987). An efficientmobilizable cosmid vector, pRK2013, and its use ina rapid method for marker exchange in Pseudomonasfluorescens strain HV37a. Gene, 61, 299-306.

Kloepper, J. W., Schroth, M. N. & Miller, T. D.(1980). Effects of rhizosphere colonization by plantgrowth-promoting rhizobacteria on potato plantdevelopment and yield. Phytopathology, 70,1078-1082.

Kraus, J. & Loper, J. E. (1992). Lack of evidence for arole of antifungal metabolite production byPseudomonas fluorescens strain Pf-5 in biologicalcontrol of Pythium damping-off of cucumber.Phytopathology, 82, 264-271.

Leemans, J., Langenakens, J., De Greve, H., Deblaere,R., Van Montagu, M. & Schell, J. (1982). Broadhost range cloning vectors derived from theW-plasmid Sa. Gene, 19, 361-364.

Lindow, S. E., Panopoulos, N. J. & McFarland, B. L.(1989). Genetic engineering of bacteria frommanaged and natural habitats. Science, 244,1300-1307.

Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982).Molecular Cloning: A Laboratory Manual ColdSpring Harbor Laboratory Press, Cold SpringHarbor, NY.

Mukhopadhyay, P., Mukhopadhyay, M. & Mills, D.(1990). Construction of a stable shuttle vector forhigh-frequency transformation in Pseudomonassyringae pv. syringae. Journal of Bacteriology, 172,477-480.

Palombo, E. A., Yusoff, K., Stanisich, V. A.,Krishnapillai, V. & Willetts, N. S. (1989). Cloningand genetic analysis of tra cistrons of the Tra2/Tra3region of plasmid RP1. Plasmid, 22, 59-69.

Pierson, L. S. & Thomashow, L. S. (1992). Cloning

Pseudomonas 9

and heterologous expression of the phenazinebiosynthetic locus from Pseudomonas aureofaciens30-84. Molecular Plant-Microbe Interactions, 5,330-339.

Simon, R., O'Connell, M., Labes, M. & Puhler, A.(1986). Plasmid vectors for the genetic analysis andmanipulation of rhizobia and other gram-negativebacteria. Methods in Enzymology, 118, 640-659.

Simon, R., Priefer, U. & Puhler, A. (1983). A broadhost range mobilization system for in vivo geneticengineering: transposon mutagenesis in gram-negative bacteria. Bio/Technology, 1, 784-790.

Smith, A. W. & Iglewski, B. H. (1989).Transformation of Pseudomonas aeruginosa byelectroporation. Nucleic Acids Research, 17, 105-109.

Smith, M., Jessee, J., Landers, T. & Jordan, J. (1990).High efficiency bacterial electroporation: 1 X 1010Rcoli transformants/pg. Focus, 12, 38-40.

Takeda, R. (1958). Pseudomonas pigments. I.Pyoluteorin, a new chlorine-containing pigmentproduced by Pseudomonas aeruginosa. Hako KogakuZasshi, 36, 281-290.

Thomashow, L. S. & Weller, D. M. (1988). Role of aphenazine antibiotic from Pseudomonas fluorescens inbiological control of Gaeumannomyces graminis var.tritici. Journal of Bacteriology, 170, 3499-3508.

Trevors, J. T. & Starodub, M. E. (1990).Electroporation of pKKl silver-resistance plasmidfrom P. stutzeri AG259 into P. putida CYM318.Current Microbiology, 21, 103-108.

Van den Eede, G., Deblaere, R., Goethals, K., VanMontagu, M. & Holsters, M. (1992). Broad hostrange and promoter selection vectors for bacteriathat interact with plants. Molecular Plant-MicrobeInteractions, 5, 228-234.

Van Haute, E., Joos, H., Maes, M., Warren, G., Van

Montagu, M. & Schell, J. (1983). Intergenerictransfer and exchange recombination of restrictionfragments cloned in pBR322: a novel strategy for thereversed genetics of the Ti plasmids of Agrobacteriumtumefaciens. EMBO Journal, 2, 411^17.

Voisard, C , Keel, C , Haas, D. & Defago, G. (1989).Cyanide production by Pseudomonas fluorescens helpssuppress black root rot of tobacco under gnotobioticconditions. EMBO Journal, 8, 351-358.

Voisard, C , Rella, M. & Haas, D. (1988). Conjugativetransfer of plasmid RP1 to soil isolates ofPseudomonas fluorescens is facilitated by certainlarge RP1 deletions. FEMS Microbiology Letters, 55,9-14.

Warren, G. J., Corotto, L. V. & Green, R. L. (1985).Conjugal transmission of integrating plasmids intoPseudomonas syringae and Pseudomonas fluorescens. InAdvances in the Molecular Genetics of theBacteria-Plant Interaction, ed. A. A. Szalay & R. P.Legocki, pp. 212-214. Cornell UniversityPublishers: Ithaca, NY.

Warren, G. J., Lindemann, J., Suslow, T. V. & Green,R. L. (1987). Ice nucleation-deficient bacteria asfrost protection agents. In Biotechnology inAgricultural Chemistry, ACS Symposium Series no.334, ed. H. M. Le Baron, R. O. Mumma, R. C.Honeycutt & J. H. Duesing, pp. 215-227. AmericanChemical Society.

Willetts, N. & Wilkins, B. (1984). Processing ofplasmid DNA during bacterial conjugation.Microbiological Reviews, 48, 24-41.

Wirth, R., Friesenegger, A. & Fiedler, S. (1989).Transformation of various species of Gram-negativebacteria belonging to 11 different genera byelectroporation. Molecular and General Genetics, 216,175-177.

2Nocardioform and CoryneformBacteriaJan Desomer and Marc Van Montagu

Introduction

Nocardioform and coryneform bacteria are Gram-positive, soil-dwelling microorganisms with a highG + C content genome and include a large num-ber of species of medical, agricultural, or indus-trial interest. Many mycobacterial diseases such astuberculosis (caused by Mycobacterium tuber-culosis) and leprosy (M. leprae) have been underinvestigation for a long time, whereas, morerecently, opportunistic infections with mycobacte-ria (M. avium) have been described in patientstreated with immunosuppressive drugs and inindividuals with acquired immunodeficiency syn-drome. Mycobacterium bovis bacille Calmette-Guerin (BCG), an avirulent strain of M. bovis, hasbeen used as a vaccine for the prevention of tuber-culosis and has nonspecific immuno-stimulatingproperties. Mycobacterial components are fre-quently added as adjuvants to stimulate theimmune response to foreign antigens.

Nocardioform bacteria of the genus Rhodococcusare mostly saprophytic soil organisms but includehuman and animal pathogens (e.g. Rhodococcusbronchialis isolated from sputum of patients withpulmonary diseases; R. equi, which causes a puru-lent bronchopneumonia in foals, cattle, swine andoccasionally humans) (Goodfellow & Minnikin,1981; Goodfellow, 1986). Rhodococcus fascians isa pathogen on a large range of plants, causingfasciation, a disease characterized by the loss ofapical dominance and the development of adven-titious shoots. In severe infections, these stuntedshoots have the appearance of a 'leafy gall'(Tilford, 1936; Lacey, 1939). In addition, rhodo-cocci exhibit a wide range of notable metabolic

activities including biotransformation of steroids(Ferreira et ah, 1984), lignin degradation (Rast etaL, 1980), degradation of xenobiotic compounds(Appel, Raabe & Lingens, 1984; Hasegawa et al.,1985; Cook & Hiitter, 1986), production of bio-surfactants and flocculants (Cooper, Akit &Kosaric, 1982; Kurane et al> 1986), and degrada-tion of alkanes (Murai et al> 1980).

The coryneform bacteria are the most impor-tant bacterial group involved in commercial pro-duction of L-amino acids, which are used as tasteenhancers in human food and in supplementinganimal feed with essential amino acids. At thesame time, this group includes important humanand animal pathogens (e.g. Corynebacterium diph-theriae, the etiologic agent of diphtheria, andCorynebacterium xerosis, a skin pathogen), whereasphytopathogenic species cause devastating wiltingdiseases of economically important crops such astomato, potato, maize and ornamentals (Keddie &Jones, 1981).

Despite the medical, agricultural and industrialinterest in coryneform and nocardioform bacteria,their study has been hampered by the lack ofmethods for the introduction of genetic material.During the last decade, cloning systems for thesebacteria have been accumulating and are increas-ingly applied to complement the strain improve-ment of amino acid producers by 'classical'mutagenesis (Batt et at., 1985), to study patho-biology by creating specific mutants (Crespi et al,1992), and for the development of live vaccines(Stover er a/., 1991).

10

Nocardioform and coryneform bacteria 11

Introduction of DNA into coryneform andnocardioform bacteria

Introduction of genetic material into bacteria cangenerally be achieved in any of the following ways.

First, transformation is the most convenientmethod to introduce exogenously added DNAinto the bacteria of interest. Bacterial species thatdo not have an identified competent phase in theirlife cycle, or in which such a phase cannot beinduced by starvation or addition of metals, canbe transformed by polyethylene glycol (PEG)-mediated DNA uptake in protoplasts/spheroplastsor made transiently porous for DNA entry by highvoltage electroporation of intact cells. Second,transduction of bacterial DNA by phage particles(including 'cosduction' of plasmids containing thecohesive ends of the phage genome) requires theexistence of a generalized transducing phage thatcan at least recognize the receptor required forattachment to the bacterial surface. This latterprerequisite is not always available or easily identi-fied, but, in cases where generalized transducingphages are described, they constitute a convenienttool for fine mapping of genes by cotransduction.Third, exchange of genetic material between bac-terial strains can also be achieved by conjugationof autonomous, conjugative elements or by mobi-lization of nonautonomously transmittable plas-mids. As the recent publications on plasmidexchange between distantly related genera pointout, exciting new ways for introduction of DNAinto recalcitrant bacteria are appearing as more islearned about the properties of transmissible plas-mids. Fourth, protoplast fusion has been appliedto a lesser extent to exchange genetic materialbetween coryneform or nocardioform bacteria(Furukawa et al> 1988).

Transformation

Polyethylene glycol-mediated DNA uptake inprotoplasts/spheroplastsPEG-mediated DNA uptake in protoplasts orspheroplasts is a process that proceeds in threedistinct steps: (i) partial or complete removal ofthe cell wall, (ii) PEG-assisted introduction of thenaked DNA, and (iii) regeneration of an intactcell wall followed by screening or selecting fortransformants. Regeneration seems to be thebottleneck, given the strain specificity of the com-

position of regeneration media. Transfection bynaked phage DNA (where no regeneration isrequired because viable phage particles can beassembled in protoplasts) has been used success-fully to optimize the first two steps of the protocols(Brownell et al, 1982; Ozaki et ah, 1984; Yeh,Oreglia & Sicard, 1985; Sanchez et al., 1986),allowing a quicker development of a protocol forthe third, apparently limiting, step of protoplasttransformation.

The exact conditions to generate protoplasts orspheroplasts competent for transformation/trans-fection are as diverse as the transformed species,or the research groups working on these problems(Table 2.1), but some underlying common princi-ples can be distilled from the different reports.Most nocardioform and coryneform bacteria areless sensitive to lysozyme action than are otherGram-positive bacteria, although lysozyme hasbeen demonstrated to hydrolyze the glycosidicbonds between iV-acetylglucosamine andiV-acetylmuramic acid in these species. This recal-citrance has been attributed to the presence of acharacteristic layer of mycolic acids (Yoshihama etal> 1985). Most commonly used ways to increasethe sensitivity to lysozyme action are the additionof cell wall synthesis inhibitors such as glycine(0.5% -2.5%), or antibiotics (0.3-0.5 U penicillinG/ml; 0.3 -200 \yg ampicillin/ml ) to the culturedcells prior to lysozyme treatment (Table 2.1).Penicillin G inhibits cross-linking of the glycans,whereas glycine is incorporated into cell wallsinstead of D-alanine; however, due to its less effi-cient incorporation and cross-linking abilities, thisoften results in growth inhibition (up to 40%; Best& Britz, 1986) and in the formation of a looser cellwall. Excessively high glycine concentrationsresult in exaggeratedly misshapen cells that areinsensitive to lysozyme action (Serwold-Davis,Groman & Rabin, 1987). Periods required for thepretreatment vary from 1 to 14 h (Table 2.1). Pre-treatment of cells with isonicotinic acid hydrazide(INH) or cerulin (inhibitors of mycolic acid syn-thesis (concentrations not mentioned)) did notresult in significantly higher spheroplast produc-tion (Yoshihama et al., 1985). Less commonmethods to increase protoplasts/spheroplasts pro-duction are the limited addition of surface-activecompounds such as sodium dodecylsulfate (SDS)(Martin et al., 1987), the less generally applicableuse of lysozyme-sensitive mutants (Smith et al.,1986), or morphologically distinct auxotrophic

Table 2.1. Polyethylene glycol-mediated DNA uptake in protoplasts I spheroplasts in nocardioform and coryneform bacteria

Organism

Coryneform bacteriaCorynebacterium glutamicumCorynebacterium herculisBrevibacterium flavumMicrobacteriumammoniaphilum

Corynebacterium lilium

Corynebacterium glutamicum

Corynebacterium glutamicumBrevibacterium

lactofermentum

Corynbacterium glutamicum

Corynebacterium diphtheriaCorynebacterium ulceransCorynebacterium glutamicum

Brevibacteriumlactofermentum

Clavibacter michiganense

Nocardioform bacteriaRhodococcus erythropolis

Rhodococcus erythropolis

Rhodococcus sp (H13-A)Rhodococcus globerulusRhodococcus equiMycobacterium smegmatis

Buffer (osmoticstabilizer)3

13.5% Na-succinate

10% Na-succinate

0.41 M sucrose

10% Na-succinate

0.5 M sorbitol

13.5% Na-succinate

0.41 M sucrose

10.3% sucrose

0.5 M sucrose

7.33% D-mannitol

10.3% sucrose

10.3% sucrose

Enzyme

1 mglysozyme/ml,37°Cfor3h

1 mglysozyme/ml,0.5 jig Ap/ml for20 h

1 mglysozyme/ml,30 °C overnight

2mglysozyme/ml,2.5% glycine for20 h

2.5 mglysozyme/ml for1.5h

0.4 mglysozyme/ml for1 h (static)

300 nglysozyme/ml at35°Cfor4h

2 mg lysozyme/mlfor2h

2mglysozyme/ml,37 °C for 45 min

1 mglysozyme/ml,30°Cfor10-15 min

10mglysozyme/ml,35°Cfor2h

2mglysozyme/ml,37°Cfor2h

Pretreatment

0.5 U PenG/ml

4 jig Ap/ml inhypertonicmedium at 34 °Cf o r i h

0.3 U PenG/ml

2.5% glycine,0.3 |xg Ap/ml

2% glycine

2% glycine

0.3 U PenG/ml

0.5% glycine

3% glycine

200 |xg Ap/ml for2h

1 % glycine

PEG addition

20% PEG 6000

30% PEG 6000

20% PEG 6000

30% PEG

50% PEG 1000

20% PEG 8000

25% PEG 6000

20% PEG 6000

165-210JJLI20% PEG 4000

20% PEG 1000

25% PEG 8000

25% PEG 1000

Regeneration(osmoticstabilizer)

13.5% Na-succinate

10% Na-succinate

0.41 M sucrose

10% Na-succinate

0.5 M sorbitol

13.5% Na-succinate

0.25 M sucrose,0.25 M Na-succinate

0.5 M sorbitol

0.5 M sucrose

7.33% D-mannitol

10.3% sucrose

0.5 M sucrose

Transformationfrequency

2.4X1O2-3.5X1O5

4.1 X103

3.2 X102

2.6 X102

3.4 X104

>101 0

(transfection)

1.5X105

104

3.8 X103

(transfection)0.2-150

(transformation)

ioVg

3X103

(transfection)2X10

(transformation)

2-9 X102

105 (transfection)

6.4 X105

2.4 X105

3.4 X106

103-104

(transfection)

Reference

Katsumata ef al.(1984)

Yeh eta/. (1985)

Ozakiefa/. (1984)

Yeh ef al. (1986)

Yoshihama ef al.(1985)

Serwold-Davies efal. (1987)

Santamaria ef al.(1984, 1985)

Meletzus &Eichenlaub(1991)

Dabbs & Sole(1988)

Brownell ef al.(1982)

Vogt Singer &Finnerty (1988)

Jacobs ef al.(1987)

Notes:PEG, polyethylene glycol; Ap, ampicillin; PenG, penicillin G.aAII percentages are w/v.


mutants (particularly isoleucine), that result infacilitated protoplast formation but are not moresensitive to lysozyme action than is the parentstrain (Best & Britz, 1986).

Enzymatic treatments with lysozyme (1 mg/mlto 10 mg/ml) are performed in osmotically stabi-lized buffers (Table 2.1). No uniformity can befound in the required periods for enzymatic treat-ment that range from very short (10 min) toextremely long (24 h), although the tendency is toreduce exposure times to lysozyme, as prolongedincubation results in reduced regeneration andtransformation efficiencies. Most protocolsrecommend shaking during lysozyme digestion,except for those described by Yoshihama et ah(1985) and Serwold-Davis etal (1987).

Similarly, no consensus can be found in theeffectiveness of enzyme treatments. AlthoughYoshihama et al (1985) reported that the lyso-zyme treatment results only in osmotically sensi-tive but morphologically unaltered cells, severalother groups report that up to 99% of the cells aretransformed to round protoplasts as observed bylight microscopy. The use of other hydrolyticenzymes (mutanolysin or lysostaphin) in combi-nation with lysozyme did not result in increasedspheroplast production (Yoshihama et al.> 1985).

Uptake of DNA by the spheroplasts/protoplastsis facilitated by the addition of PEG. Generally,PEG with an average molecular weight of 6000 isused, in concentrations that vary between 20%and 50% (Table 2.1). Coryneform and nocardio-form bacteria are less sensitive to the presence ofPEG than are other Gram-positive bacteria suchas Bacillus subtilis (Chang & Cohen, 1979) anddilution with hypertonic media is sufficient toallow regeneration on osmotically stabilizedmedia. These latter also differ greatly in theirexact composition. Transformation efficienciesdepend on several variables, e.g. the nature of theDNA used (phage versus plasmid), the hostorganism used to prepare the transforming DNAand the recipient organism. Under optimal condi-tions the values can vary from as low as less thanone transformant/ng DNA to up to 106 transfor-mants/^g DNA. Transfection efficiencies arehigher than transformation efficiencies of identicalorganisms, reflecting possibly the regenerationfrequency of the cell wall required for transforma-tion.

Transformation efficiencies are linearly corre-lated with the concentration of transforming

DNA, but saturation thresholds have beenreported (Santamaria, Gil & Martin, 1985). Anadditional problem that might arise during PEG-mediated transformation of protoplasts/sphero-plasts is the loss of endogenous plasmids duringthe protoplasting/regeneration process. Evidencefor 'curing' following formation and regenerationof protoplasts/spheroplasts was provided forStaphylococcus aureus (Novick et ah, 1980), Strep-tomyces coelicolor (Hopwood, 1981), and Rhodo-coccus fascians (Desomer, Dhaese & VanMontagu, 1988).

High voltage electroporation of intact cellsIntroduction of DNA into bacterial cells via briefhigh voltage electric discharges has become thepreferred method for transformation of recalci-trant species during the last few years, because it issimple, fast and reproducible. The exact mecha-nism by which electrotransformation occurs, how-ever, remains obscure. When cell membranes areplaced in an electric field of a critical strength, areversible local disorganization and transientbreakdown occurs, allowing both molecular influxand efflux (Chassy, Mercenier & Flickinger,1988). However, when the imposed electric fieldbecomes too high, irreversible damage to themembranes may result in cell death. During 'pore'formation, transforming DNA can diffuse into thebacterial cells. Several protocols for electropora-tion and transformation efficiencies are summa-rized in Table 2.2. In general, no pretreatment ofthe cells is required, except for extensive washes ofthe harvested cells with the electroporationmedium. Nevertheless, Hermans, Boschloo & deBont (1990) and Haynes & Britz (1990) recom-mend the addition of 4 to 8 mg of INH, or INH incombination with 2.5% (w/v) of glycine to thegrowing cell cultures in order to obtain highertransformation efficiencies for M. aureum and Cglutamicum, respectively. However, Dunican &Shivnan (1989) found no significant improvementby pretreating the cultures with glycine, althoughthe transformation efficiencies obtained by theseauthors were already higher than those reportedby others. These high transformation efficienciescould be correlated with the low conductivity ofthe electroporation medium they used. Electro-poration can be achieved in media as simple aswater (Meletzus & Eichenlaub, 1991), supple-mented with PEG (Desomer et at., 1990), orglycerol (Dunican & Shivnan, 1989), which both

Table 2.2. High voltage electroporation of intact bacterial cells

Organism

Nocardioform bacteriaMycobacterium aureum

Mycobacterium smegmatis

Mycobacterium bovis BCGMycobacterium smegmatis

Rhodococcus fascians

Mycobacterium smegmatis(EPT)

Coryneform bacteriaClavibacter michiganense


Corynebacterium calluvaeBrevibacter iactofermentum

Brevibacter ammoniagenes


Electroporation

medium

10% sucrose, 7mMHepes, pH7.0, 1 mMMgCI2

7 mM Na-phosphate,pH7.2

272 mM sucrose10mM Hepes, pH7.0,10% glycerol

30% PEG 1000

10% glycerol

H2O

10% glycerol

10% glycerol10% glycerol

10% glycerol

0.5 M sucrose, 1 mMMgCI2

Electric settings

field(kV/cm)

12.5

6.25

6.25

12.5

12.5

12.5

12.5

12.5

R(n)

200

—

—

400

—

—

200

200

c

25

25

25

25

25

25

25

25

Time

constant(ms)

3.5-4.5

—

—

3.5-4.5

—

13.5

4.5-5

4.5-5

Transformation

efficiency

2.34 X104

5X103

10-500

105-107

104

105

2X103

1.5X1O3-1X1O7

102-103

103-105

103

5.2 X105

Remarks

Pretreatment with4|jig INH/ml

Phasmids

Homologousintegration

TransfectionShuttle plasmids

Pretreatment 2.5%glycine, 4-8 mgINH/ml

Reference

Hermans et al.(1990)

Snapper et al.(1988)

Husson et al.(1990)

Desomer et al.(1990)

Snapper et al.(1990)



Dunican & Shivnan(1989)

Dunican & Shivnan(1989)

Haynes & Britz(1989, 1990)

Notes:INH, isonicotinic acid hydrizide.


enhance efficiency and allow freezing, in portionsof the 'electrocompetent' cells.

The electric parameters are very similar in allcases. An exponentially decreasing pulse with aninitial field strength of 12.5 kV/cm is deliveredfrom a 25 ^F capacitance. The total resistance (R)is a combination of internal resistance (deter-mined by the conductivity of the electroporationmedium, the volume of the cells to be electropo-rated, etc., all parameters that are less amenable tomanipulation) and external resistances, which canbe changed simply by switching the resistors of theelectroporation apparatus. Both resistances deter-mine the time constant (T = RC) that defines thetime needed for the field strength to reach aboutone-third of the initial field strength, and for prac-tical reasons can be regarded as the effective dura-tion of the pulse. With the exception of Clavibactermichiganense, most time constants reported forefficient electroporation of nocardioform andcoryneform bacteria are in the range 3.5-5 ms(Table 2.2). A lower time constant (< 3 ms)results, at least for R. fascians, in a drastic decreasein transformation efficiency (J. Desomer, unpub-lished data). Considering the time needed fortransformants to form colonies, T has proven to bea good time-saving evaluation parameter.

Other critical values affect transformation effi-ciency by electroporation. A uniform physiologi-cal state of the cells at the time of harvest isgenerally ensured by collecting cells from lateexponential or early stationary phase cultures(A600=0.5 to 0.6). The density of electroporatedsamples fluctuates around 109 colony formingunits (c.f.u.)/ml.

A saturation at higher DNA concentrations,which could be explained by the existence of anelectrocompetent subpopulation, was observedfor several coryneform bacteria (Dunican &Shivnan, 1989) and for R. fascians (Desomer et al.,1990). For the latter species, electrotransforma-tion of a 160 kDa plasmid, albeit with low trans-formation efficiency, has been reported (Desomeret al, 1990), indicating that plasmid size need notlimit electrotransformation. Interestingly, M.smegmatis mutants could be obtained that exhib-ited an enhanced (four to five orders of magni-tude) plasmid transformation efficiency (EPT)phenotype, but were unaffected in efficiency oftransfection by phage D29 DNA, or in the effi-ciency of transformation by integrating vectors(Snapper et al> 1990). As it was shown that no

inactivation of a restriction/modification systemwas involved, it was assumed that EPT mutantsare affected in some aspects of plasmid replicationand maintenance (Snapper et at., 1990).

Conjugation

Self-transmissible extrachromosomal elementshave been described both in nocardioform andcoryneform bacteria. In Clavibacter flaccumfaciens,a 69 kb circular plasmid that encodes arsenite,arsenate and antimony resistance genes(pDGlOl) could be conjugated between severalsubspecies with transfer frequencies ranging from5 X 10~9 to 1.3 X 10"4. Crosses between C. flac-cumfaciens and C. michiganense were unsuccessful(Hendrick, Haskins & Vidaver, 1984). Also in thenocardioform bacteria, conjugative endogenousmetal resistance plasmids were found. In Rhodo-coccus fascians^ transfer of cadmium resistanceplasmids (120-160 kb) was observed betweenwild-type and cured strains with frequencies of10"4 to 10"2 (Desomer et al, 1988). Lithoauto-trophic, thallium-resistant, Nocardia opaca strainscan transfer both the thallium resistance (fre-quency 10"1 to 10~2 per donor cell) and the abilityto grow lithoautotrophically (frequency 10"4 to10") to cured N. opaca or Rhodococcus erythropolisstrains (Sensfuss, Reh & Schlegel, 1986). Whereasthe correlation between transfer of thallium resis-tance and conjugal transfer of 110 kb plasmidswas obvious (Sensfuss et al.> 1986), it has onlyrecently been discovered that the genetic informa-tion for hydrogen-autotrophic growth was locatedon large, conjugative, linear extrachromosomalelements (270-280 kb) (Kalkus, Reh & Schlegel,1990). Similarly sized linear plasmids werediscovered in the phytopathogen Rhodococcusfascians. In this case too, conjugal transfer(10^/acceptor) could be detected after appropri-ate tagging of these plasmids, but here the linearplasmids encode essential fasciation-inducinggenes (Crespi et al.y 1992). No thorough geneticanalysis of the conjugation mechanism in thesespecies has been published.

Although this intraspecies conjugation of self-transmissible extrachromosomal elements hasproven useful in the allocation of plasmid-residinggenes, the technique has been of limited use ingenetically engineering this group of bacteria. Therecent description of interspecies mobilization ofappropriate shuttle plasmids from Escherichia coli

16 DESOMER AND VAN MONTAGU

to both coryneform and nocardioform bacteria,therefore, opens up exciting new possibilities(Lazraq et al, 1990; Schafer et al, 1990; Gormley& Davies, 1991; Mazodier & Davies, 1991).

In the first applications of this system, shuttlevectors were constructed that contained an originof replication for Mycobacterium species (Lazraq etaLy 1990) or coryneform bacteria (Schafer et al.,1990), an origin of replication for E. coli, an IncP-type origin of transfer and a kanamycin resistancegene known to be expressed in both E. coli andcoryneform bacteria or mycobacteria. Escherichiacoli helper strains that provided IncP transfer func-tion in trans could mobilize the resulting plasmidsto M. smegmatis (pMYlO: transfer frequencybetween 2.2 X 10~5 and 1.2 X 10"7; Lazraq et al.,1990) or different coryneform bacteria (transferfrequencies ranging from 10~7to 10~2, dependingon the species used as recipient; Schafer et at.,1990). Interestingly, transfer frequencies to Cglutamicum were high enough to allow rescue ofnonreplicating vectors by homologous recombina-tion (Schwarzer & Puhler, 1991). Using the IncQplasmid pRSFlOlO or derivatives known to repli-cate in both E. coli and mycobacteria (Hermans etat., 1991), Gormley & Davies (1991) were able toestablish plasmid transfer by mobilization from anE. coli helper strain to M. smegmatis (frequency10"2 to 10~3). Due to its relative independence ofhost-encoded functions, pRSFlOlO can replicatestably in mycobacteria and overcome the necessityfor multiple replicons to suit the different hosts.This property could turn pRSFlOlO into a power-ful tool for genetic manipulation of these industri-ally and medically important bacteria.

Transduction

Generalized transduction has not frequently beenemployed for the introduction of genetic materialin nocardioform or coryneform bacteria. In 1976,Momose, Miyashiro & Oba reported isolating abacteriophage that mediated generalized trans-duction in Brevibacterium flavum at a frequency of10~6/plaque-forming unit (p.f.u.). More recently, ageneralized transducing bacteriophage for Rhodo-coccus erythropolis has been described (Q4; Dabbs,1987). Under optimal conditions, transduction toprototrophy of auxotrophic markers or transduc-tion of antibiotic resistance markers was found ata frequency of 10~8/p.f.u.

More sophisticated are the phasmids for use in

mycobacteria that consist of E. coli plasmid andthe (pseudo)temperate phages of mycobacteria.These phasmids can be propagated in E. coli asplasmids, transfected to fast-growing mycobacte-ria, and packaged into phage particles that canthen efficiently infect slow-growing mycobacteria(e.g. BCG; Jacobs, Tuckman & Bloom, 1987;Snappers a/., 1990).

A cosmid-type vector (containing the cohesiveends of Brevibacterium phage fla) for coryneformbacteria has been described that was packaged invivo upon infection of the harboring bacteria withfla and transduced with a frequency of 10^/p.f.u.

/., 1985).

Fate of transforming DNA

Plasmids and marker genes

Like other microorganisms, coryneform andnocardioform bacteria contain a multitude of plas-mids. An early survey of phytopathogenic Coryne-bacterium species resulted in the detection ofplasmids, ranging from 35 to 78 kb (Gross,Vidaver & Keralis, 1979), some of which are con-jugative (Hendrick et aL, 1984). Because of theirlarger sizes, they had little value for developmentas cloning vectors. Therefore, several researchgroups sought and found smaller plasmids incoryneforms, most of which were cryptic (forreviews, see Sandoval et al., 1985; Martin et al.,1987). For the nocardioform genera Rhodococcusand Mycobacterium, fewer plasmids have beendescribed. Large native plasmids that code forresistance to chloramphenicol and heavy metals(cadmium, antimony, and arsenate) have beendescribed in R. fascians (Desomer et al> 1988) andan unidentified Rhodococcus isolate (Dabbs &Sole, 1988). The former plasmids are also con-jugative (Desomer et al, 1988). In addition, large,linear plasmids (>200 kb) conferring the abilityfor hydrogen-auto trophic growth (Kalkus et al,1990) or phytopathogenicity (Crespi et al, 1992)were detected in rhodococci. An intermediate-sized plasmid (80 kb) is probably correlatedwith the production of a virulence-associatedprotein in Rhodococcus equi (Tkachuk-Saad &Prescott, 1991), whereas smaller cryptic plasmids(2.6-19.5 kb) were found in Rhodococcus species(Vogt Singer & Finnerty, 1988; Hashimoto et al>1992). In mycobacteria, the best characterized,


but cryptic, plasmid is pAL5000 (Labidi et al,1984; Rauzier, Moniz-Pereira & Gicquel-Sanzey,1988; Ranes era/., 1990).

Cloning vectors have been developed basedupon these small cryptic plasmids (Hashimoto etal.) 1992) or by deletion of the larger plasmids(Desomer et al, 1990). As marker genes, eitherheterologous genes from E. coli (aminoglycoside-phosphotransferase {aph) or chloramphenicolacetyltransferase (cat); Martin et al, 1987) orStreptomyces species have frequently been used(thiostrepton resistance; Vogt Singer & Finnerty,1988) as well as endogenous marker genes(spectinomycin/streptomycin resistance (Katsumataet al, 1984); chloramphenicol resistance (Desomeret al, 1990, 1992); arsenate resistance (Dabbs,Gowan & Andersen, 1990)). Examples of suchcloning vectors for corynebacteria, rhodococci,and mycobacteria are listed in Table 2.3.

Integration of nonreplicative vectors

Vectors that cannot replicate in coryneform ornocardioform bacteria, but can be selected for,due to the presence of a marker gene, can be res-cued upon introduction into these bacteria byintegration with an existing replicon. This integra-tion can occur via either homologous or illegiti-mate recombination.

Integration by homologous recombinationIntegration of transformed DNA by homologousrecombination has been studied in coryneformand nocardioform bacteria, mainly for appliedreasons (stable integration, removal of undesiredgenes) and is confined to the analysis of recombi-nation between vector-borne DNA fragments andtheir chromosomal homologous counterparts. Inmycobacteria, introduction of a pUC19-basedvector, unable to replicate in M. smegmatis> andcontaining the aph gene of Tn903 inserted into thepyrF gene of M. smegmatis, resulted in 10-500transformants/ng of electroporated plasmid DNA.Southern analysis revealed that 60% of these wereintegrated by single homologous recombinationthrough the pyrF DNA sequences flanking the aphgene. The remaining 40% of the transformantshad undergone a double homologous recombina-tion, resulting in the replacement of the intactchromosomal pyrF gene copy by the interruptedpyrF gene, and consequently resulting in uracilauxotrophy. Looping out of the vector sequences

by a second homologous recombination event inisolates where the plasmid had already been inte-grated by a single homologous recombinationoccurred at frequencies lower than 10~3 (Husson,James & Young, 1990).

Similar observations were obtained for integra-tion by homologous recombination of nonreplica-tive vectors containing an interrupted lysA gene inCorynebacterium glutamicum. Since the used vec-tors contain the mob functions of RP4 and, there-fore, are mobilizable in conjugations between E.coli SI7-1 and C. glutamicum, single homologousrecombinants were obtained with higher efficien-cies (5 X 102 to 5 X 103 integrants/mating). Only2% of these isolates had undergone a doublerecombination event, resulting in lysine auxo-trophy. Loss of vector sequences by a secondrecombination event in mutant strains with anintegrated vector was obtained only after 200 gen-erations, at low frequency (Schwarzer & Puhler,1991).

Reyes et al. (1991) constructed plasmids fromwhich a repliconless cartridge, called 'integron',can be isolated (it contains a selectable markerand the git A region of C. melassecola). These plas-mids can be obtained from the coryneform host tobe transformed, and thus provide a source of host-modified 'integron' DNA, that in turn can be usedfor ligation to the gene of interest and transforma-tion of the host. In this way, stable integrants ofcorynebacteria were obtained without restrictionmodification barriers (Reyes et al., 1991).

Integration after electrotransformation of non-replicating vectors by single recombination throughhomologous vector-borne DNA sequences andchromosomal counterparts has also been reportedfor rhodococci, with an efficiency of 102/pg of elec-troporated plasmid DNA (Desomer et al> 1990).Single- and double-homologous recombinationbetween a DNA fragment interrupted by thephleomycin resistance (PhleoR) gene, and the wild-type counterpart located on a conjugative plasmidwas forced by conjugation to a plasmidless acceptorstrain. Cmrhleo isolates (double recombinants)were obtained with a frequency of only 5 X 10"3 ofthat of the PhleoR mutants (M. Crespi & J.Desomer, unpublished data).

Integration by illegitimate recombination

In several nocardioform bacteria, integration ofnonreplicative plasmid DNA in the genome by


Table 2.3. Examples of cloning vectors

Coryneformbacteria

pULRS6

pULRS8

pULMJ51

MycobacteriumpMGV261pMGV361

RhodococcuspRK4pRF30

Size(kb)

6.1

5.8

11.3

4.54.5

5.313.2

ABR

markers3

HygKanCatKanCat

KanKan

KanCm

Comments

Contains promoterless kanamycin gene forshotgun cloning of promoters

Expression vector {hsp60) promoterIntegrative expression vector

Based on cryptic pRC3 plasmidDeletion derivative of pRF2 (160 kb)

Reference

Martfn etal. (1990)

Cadenas, Martin &Gil (1991)

Stover ef al. (1991)Stover et al. (1991)

Hashimoto ef al. (1992)Desomer etal. (1990)

Notes:aABR, antibiotic resistance; Kan, kanamycin; Cat, chloramphenicol (acetyltransferase); Cm, chloramphenicol; Hyg,hygromycin.

illegitimate recombination has been reported. InMycobacterium tuberculosis and BCG, electropora-tion of a pBR322-derived vector (containing theaph gene of Tn5seql as selectable marker) thatcould not replicate in these hosts still resulted intransformants (at a frequency 104 to 105 timeslower than when autonomously replicating DNAwas used), provided that DNA was linearizedprior to electroporation. Independent transfor-mants revealed different Southern hybridizationpatterns, indicating random integrating events(Kalpana, Bloom & Jacobs, 1991).

Similar observations were made in R. fasciansupon introduction of replication-deficient pUC13-or pUC18-based vectors containing a selectablemarker gene for this species. However, no lin-earization prior to electrotransformation wasrequired to obtain frequencies of integration ashigh as those in Mycobacterium. The presence ofintact ColEl replicon and ampicillin resistancegene on the inserted sequences allowed straight-forward cloning of the interrupted target DNAsequences. Furthermore, integration seemed tooccur via specific positions on the plasmids, whichwere determined by sequence analysis to overlapthe restriction enzyme Narl site in pUC13 orpUC18. No homologies could be found amongthe sequences of different target sites nor betweenthe recombining target and plasmid sequences,

indicating that recombination was illegitimate andessentially random along the genome. This latterproperty had already been suggested by the multi-tude of different phenotypes obtained (auxotro-phies, pigmentation mutants, phytopathogenicitymutants). Although the mechanism by which inte-gration of the plasmids occurs in R. fascians isunknown, available evidence seems to point in thedirection of a linear recombination intermediate,generated from the incoming circular plasmid(Desomer, Crespi & Van Montagu, 1991). Whiletransposition of movable genetic elements hasbeen described in mycobacteria (Leskiw et al,1990; Martin et al, 1990; McAdam et al, 1990),none of these mutagenesis systems achieves theefficiencies reported above.

Examples of application

Pathobiology of Rhodococcus fasciansinfection in plants

Rhodococcus fascians is a nocardioform phyto-pathogen that infects a large range of plants(Tilford, 1936; Lacey, 1939) and causes thedevelopment of numerous adventitious shoots.Although in earlier publications (Murai et al,1980; Murai, 1981) phytopathogenic propertieshad been associated with the presence of large


covalently closed circular cadmium resistance-encoding plasmids (a thesis contradicted byLawson, Gantotti & Starr (1982)), recently, largelinear extrachromosomal elements unique tovirulent strains have been described (Crespi et al.,1992). Transformation technology (Desomer etah, 1990) has allowed insertion of a marker geneinto the linear plasmid. By conjugation of theselinear fasciation-inducing (Fi) plasmids fromvirulent to avirulent cured strains, it was demon-strated unequivocally that they carried essentialfasciation-inducing genes (Crespi et al., 1992).The random insertion of nonreplicating plasmidsalong the Fi plasmid has identified at least threeloci, involved in fasciation induction, one encod-ing an isopentenyltransferase, the enzyme thatcatalyzes the first step in cytokinin biosynthesis.

The use of BCG for recombinant vaccines

The transformation procedures developed formycobacteria have allowed the construction ofrecombinant 'bacille Calmette-Guerin' (BCG),expressing antigens of different pathogens (Stoveret al., 1991). BCG has a long history of use toimmunize against tuberculosis and exhibits someunique features (oral administration, single inocu-lum requirement, innocuous persistence in vivo)that could allow development of a inexpensivemultivaccine. Therefore, an integrative expressionvector was developed based on hsp60 expressionsignals and the phage attachment site (a#P) andintegrase (int) gene of a temperate mycobacterio-phage. Integrated vectors are stably maintainedeven without antibiotic selection (Lee et al., 1991)and Zz5p60-driven expression of several antigensfrom a variety of pathogens was obtained (Stoveret al., 1991). In mice, inoculated by low titers ofBCG, long-lasting humoral and cellular immuneresponses were elicited and persistent BCGs wererecovered from host spleens. Heat-killed BCGswere not effective for immunization, supportingthe view that the persisting immune responseswith BCG are due to growth and persistence ofthe vaccine (Stover et al., 1991).

Concluding remarks

Although significant progress in the developmentof molecular-genetic tools for coryneform andnocardioform bacteria has been achieved in the

last decade, several improvements are stillrequired for effective genetic engineering. A firstdrawback is the lack of incompatibility studiesamong the different replicons, making simultane-ous introduction and stable inheritance of morethan one plasmid (e.g. in genetic repressor/opera-tor studies) unpredictable. Also, very little isknown about stability of the used replicons with-out selective pressure. Available information isscant on the structure of expression signals inthese bacteria, let alone regulation signals, andvirtually nothing is known about secretion signals.Studies of this kind could increase the 'tool box'for successful manipulation of the specific proper-ties of these bacterial genera.

References

Appel, M., Raabe, T. & Lingens, F. (1984).Degradation of o-toluidine by Rhodococcusrhodochrous. FEMS Microbiology Letters, 24,123-126.

Batt, C. A., Follettie, M. T., Shin, H. K., Yeh, P. &Sinskey, A. J. (1985). Genetic engineering ofcoryneform bacteria. Trends in Biotechnology, 3,305-310.

Best, G. R. & Britz, M. L. (1986). Facilitatedprotoplasting in certain auxotrophic mutants ofCorynebacterium glutamicum. Applied Microbiology andTechnology, 23, 288-293.

Brownell, G. H., Saba, J. A., Denniston, K. &Enquist, L. W. (1982). The development of aRhodococcus-actinophage gene cloning system.Developments in Industrial Microbiology, 23, 287-298.

Cadenas, R. F., Martin, J. F. & Gil, J. A. (1991).Construction and characterization of promoter-probe vectors for Corynebacteria using thekanamycin-resistance reporter gene. Gene, 98,117-121.

Chang, A. C. & Cohen, S. (1979). High frequencytransformation of Bacillus subtilis protoplasts byplasmid DNA. Molecular and General Genetics, 168,111-115.

Chassy, B. M., Mercenier, A. & Flickinger, J. (1988).Transformation of bacteria by electroporation.Trends in Biotechnology, 6, 303-309.

Cook, A. M. & Hiitter, R. (1986). Ring dechlorinationof deethylsimazine by hydrolases from Rhodococcuscorallinus. FEMS Microbiology Letters, 34, 335-338.

Cooper, D. G., Akit, J. & Kosaric, N. (1982). Surfaceactivity of the cells and extracellular lipids inCorynebacterium fascians CF15. Journal ofFermentation Technology, 60, 19-24.

Crespi, M., Messens, E., Caplan, A. B., Van


Montagu, M. & Desomer, J. (1992). Fasciationinduction by the phytopathogen Rhodococcus fasciansdepends upon a linear plasmid encoding a cytokininsynthase gene. EMBO Journal, 11, 795-804.

Dabbs, E. R. (1987). A generalised transducingbacteriophage for Rhodococcus erythropolis. Molecularand General Genetics, 206, 116-120.

Dabbs, E. R. & Sole, G. J. (1988). Plasmid-borneresistance to arsenate, arsenite, cadmium, andchloramphenicol in a Rhodococcus species. Molecularand General Genetics, 211, 148-154.

Dabbs, E. R., Gowan, B. & Andersen, S. J. (1990).Nocardioform arsenic resistance plasmids andconstruction of Rhodococcus cloning vectors. Plasmid,23, 242-247.

Desomer, J., Crespi, M. & Van Montagu, M. (1991).Illegitimate integration of non-replicative vectors inthe genome of Rhodococcus fascians uponelectrotransformation as an insertional mutagenesissystem. Molecular Microbiology, 5, 2115-2124.

Desomer, J., Dhaese, P. & Van Montagu, M. (1988).Conjugative transfer of cadmium resistance plasmidsin Rhodococcus fascians strains. Journal of Bacteriology,170, 2401-2405.

Desomer, J., Dhaese, P. & Van Montagu, M. (1990).Transformation of Rhodococcus fascians by high-voltage electroporation and development of R.fascians cloning vectors. Applied and EnvironmentalMicrobiology, 56, 2818-2825.

Desomer, J., Vereecke, D., Crespi, M. & VanMontagu, M. (1992). The plasmid-encodedchloramphenicol resistance protein of Rhodococcusfascians is homologous to the transmembranetetracycline efflux proteins. Molecular Microbiology,6, 2377-2385.

Dunican, L. K. & Shivnan, E. (1989). High frequencytransformation of whole cells of amino acidproducing coryneform bacteria using high voltageelectroporation. Bio I Technology, 7, 1067-1070.

Ferreira, N. P., Robson, P. M., Bull, J. R. & van derWalt, W. H. (1984). The microbial production of3a-H-4a-(3 '-propionic acid)-5a-hydroxy-7ap-methylhexahydro-indan-1 -one-8-lactone fromcholesterol. Biotechnology Letters, 6, 517-522.

Furukawa, S., Azuma, T., Nakanishi, T. & Sugimoto,M. (1988). Breeding an L-isoleucine producer byprotoplast fusion of Corynebacterium glutamicum.Applied Microbiology and Biotechnology, 29,248-252.

Goodfellow, M. (1986). Genus Rhodococcus Zopf1891, 28al. In Bergey's Manuaf of SystematicBacteriology, Vol. 2, ed. P. H. A. Sneath, N. S. Mair,M. E. Sharpe & J. G. Holt, pp. 1472-1481.Williams and Wilkins, Baltimore.

Goodfellow, M. & Minnikin, D. E. (1981). Thegenera Nocardia and Rhodococcus. In The Prokaryotes,Vol. II, ed. M. P. Starr, H. Stolp, H. G. Triiper,

A. Balows & H. G. Schlegel, pp. 2016-2027.Springer-Verlag, Berlin.

Gormley, E. P. & Davies, J. (1991). Transfer ofplasmid RSF1010 by conjugation from Escherichiacoli to Streptomyces lividans and Mycobacteriumsmegmatis. Journal of Bacteriology, 173, 6705-6708.

Gross, D. C , Vidaver, A. K. & Keralis, M. B. (1979).Indigenous plasmids from phytopathogenicCorynebacterium species. Journal of GeneralMicrobiology, 115, 479-489.

Hasegawa, S., Vandercook, C. E., Choi, G. Y.,Herman, Z. & Ou, P. (1985). limonoid debitteringof citrus juice sera by immobilized cells ofCorynebacterium fascians. Journal of Food Science, 50,330-332.

Hashimoto, Y., Nishiyama, M., Yu, F., Watanabe, I.,Horinouchi, S. & Beppu, T. (1992). Developmentof a host-vector system in a Rhodococcus strain andits use for expression of the cloned nitrile hydratasegene cluster. Journal of General Microbiology, 138,1003-1010.

Haynes, J. A. & Britz, M. L. (1989).Electrotransformation of Brevibacteriumlactofermentatum and Corynebacterium glutamicum:growth in Tween 80 increases transformationfrequencies. FEMS Microbiology Letters, 61,329-334.

Haynes, J. A. & Britz, M. L. (1990). The effect ofgrowth conditions of Corynebacterium glutamicum onthe transformation frequency obtained byelectroporation. Journal of General Microbiology, 136,255-263.

Hendrick, C. A., Haskins, W. P. & Vidaver, A. K.(1984). Conjugative plasmid in Corynebacteriumflaccumfaciens subsp. oortii that confers resistance toarsenite, arsenate, and antimony (III). Applied andEnvironmental Microbiology, 48, 56-60.

Hermans, J., Boschloo, J. G. & de Bont, J. A. M.(1990). Transformation of Mycobacterium aurum byelectroporation: the use of glycine, lysozyme andisonicotinic acid hydrazide in enhancingtransformation efficiency. FEMS MicrobiologyLetters, 72, 221-224.

Hermans, J., Martin, C , Huijberts, G. N. M.,Goosen, T. & de Bont, J. A. M. (1991).Transformation of Mycobacterium aurum andMycobacterium smegmatis with the broad host-rangeGram-negative cosmid vector pJRD215. MolecularMicrobiology, 5, 1561-1566.

Hopwood, D. A. (1981). Genetic studies withbacterial protoplasts. Annual Review of Microbiology,35, 237-272.

Husson, R. N., James, B. E. & Young, R. A. (1990).Gene replacement and expression of foreign DNA inmycobacteria. Journal of Bacteriology, 172, 519-524.

Jacobs, W. R. Jr, Tuckman, M. & Bloom, B. R.(1987). Introduction of foreign DNA into


mycobacteria using a shuttle plasmid. Nature{London), 327, 532-535.

Kalkus, J., Reh, M. & Schlegel, H. G. (1990).Hydrogen autotrophy of Nocardia opaca strains isencoded by linear megaplasmids. Journal of GeneralMicrobiology, 136, 1145-1151.

Kalpana, G. V., Bloom, B. R. & Jacobs, W. R. Jr(1991). Insertional mutagenesis and illegitimaterecombination in mycobacteria. Proceedings of theNational Academy of Sciences, USA, 88, 5433-5437.

Katsumata, R., Ozaki, A., Oka, T. & Furuya, A.(1984). Protoplast transformation of glutamate-producing bacteria with plasmid DNA. Journal ofBacteriology, 159, 306-311.

Keddie, R. M. & Jones, D. (1981). Saprophytic,aerobic coryneform bacteria. In The Prokaryotes, Vol.II, ed. M. P. Starr, H. Stolp, H. G. Triiper, A.Balows & H. G. Schlegel, pp. 1838-1878. Springer-Verlag, Berlin.

Kurane, R., Toeda, K., Takeda, K. & Suzuki, T.(1986). Culture conditions for production ofmicrobial flocculant by Rhodococcus erythropolis.Agricultural and Biological Chemistry, 50, 2309-2313.

Labidi, A., Dauget, C , Goh, K. S. & David, H. L.(1984). Plasmid profiles of Mycobacterium fortuitumcomplex isolates. Current Microbiology, 11, 235-240.

Lacey, M. S. (1939). Studies on a bacterium associatedwith leafy galls, fasciations and 'cauliflower' diseaseof various plants. Part III. Further isolation,inoculation experiments and cultural studies. Annalsof Applied Biology, 26, 262-278.

Lawson, E. N., Gantotti, B. V. & Starr, M. P. (1982).A 78-megadalton plasmid occurs in avirulent strainsas well as virulent strains of Corynebacterium fascians.Current Microbiology, 7, 327-332.

Lazraq, R., Clavel-Seres, S., David, H. L. & Roulland-Dussoix, D. (1990). Conjugative transfer of a shuttleplasmid from Escherichia coli to Mycobacteriumsmegmatis. FEMS Microbiology Letters, 69, 135-138.

Lee, M. H., Pascopella, L., Jacobs, W. R. Jr & Hatfull,G. F. (1991). Site-specific integration ofmycobacteriophage L5: integration-proficientvectors for Mycobacterium smegmatis, Mycobacteriumtuberculosis, and bacille Calmette-Guerin. Proceedingsof the National Academy of'Sciences,USA, 88,3111-3115.

Leskiw, B. K., Mevarech, M., Barritt, L. S., Jensen,S. E., Henderson, D. J., Hopwood, D. A., Bruton,C. J. & Chater, K. F. (1990). Discovery of aninsertion sequence, IS 116, from Streptomycesclavuligerus and its relatedness to other transposableelements from actinomycetes. Journal of GeneralMicrobiology, 136, 1251-1258.

Martin, C , Timm, J., Rauzier, J., Gomez-Lus, R.,Davies, J. & Gicquel, B. (1990). Transposition of anantibiotic resistance element in mycobacteria. Nature{London), 345, 739-746.

Martin, J. F., Cadenas, R. F., Malumbres, M.,Mateos, L. M., Guerrero, C. & Gil, J. A. (1990).Construction and utilization of promoter-probe andexpression vectors in corynebacteria.Characterization of corynebacterial promoters. InProceedings of the 6th International Symposium onGenetics of Industrial Microorganisms, ed. H. Heslot, J.Davis, J. Florent, L. Bobichon, G. Durand & L.Penasse, pp. 283-292. Societe Francaise deMicrobiologie, Paris.

Martin, J. F., Santamaria, R., Sandoval, H., del Real,G., Mateos, L. M., Gil, J. A. & Aguilar, A. (1987).Cloning systems in amino acid-producingcorynebacteria. Bio/Technology, 5, 137-146.

Mazodier, P. & Davies, J. (1991). Gene transferbetween distantly related bacteria. Annual Review ofGenetics, 25, 147-171.

McAdam, R. A., Hermans, P. W. M., van Soolingen,D., Zainuddin, Z. F., Catty, D., van Embden,J. D. A. & Dale, J. W. (1990). Characterization of aMycobacterium tuberculosis insertion sequencebelonging to the IS5 family. Molecular Microbiology,4, 1607-1613.

Meletzus, D. & Eichenlaub, R. (1991).Transformation of the phytopathogenic bacteriumClavibacter michiganense subsp. michiganense byelectroporation and development of a cloning vector.Journal of Bacteriology, 173, 184-190.

Miwa, K., Matsui, K., Terabe, M., Ito, K., Ishida, M.,Takagi, H., Nakamori, S. & Sano, K. (1985).Construction of novel shuttle vectors and a cosmidvector for the glutamic acid-producing bacteriaBrevibacterium lactofermentum and Corynebacteriumglutamicum. Gene, 39, 281-286.

Momose, H., Miyashiro, S. & Oba, M. (1976). On thetransducing phages in glutamic acid-producingbacteria. Journal of General Applied Microbiology, 22,119-129.

Murai, N. (1981). Cytokinin biosynthesis and itsrelationship to the presence of plasmids in strains ofCorynebacterium fascians. In Metabolism and MolecularActivities of Cytokinins, ed. J. Guern & C. Peaud-Lenoel, pp. 17-26. Springer-Verlag, Berlin.

Murai, N., Skoog, F., Doyle, M. E. & Hanson, R. S.(1980). Relationship between cytokinin production,presence of plasmids, and fasciation caused bystrains of Corynebacterium fascians. Proceedings of theNational Academy of Sciences, USA, 11, 619-623.

Novick, R., Sanchez-Rivas, C , Gruss, A. & Edelman,I. (1980). Involvement of the cell envelope inplasmid maintenance: plasmid curing during theregeneration of protoplasts. Plasmid, 3, 348-358.

Ozaki, A., Katsumata, R., Oka, T. & Furuya, A.(1984). Transfection of Corynebacterium glutamicumwith temperate phage <pCGl. Agricultural andBiological Chemistry, 48, 2597-2601.

Ranes, M. G., Rauzier, J., Lagranderie, M.,


Gheorghiu, M. & Gicquel, B. (1990). Functionalanalysis of pAL5000, a plasmid from Mycobacteriumfortuitum: construction of a 'mini'mycobacterium-Esc/zmc/Ma coli shuttle vector.Journal of Bacteriology, 172, 2793-2797.

Rast, H. G., Engelhardt, G., Ziegler, W. & Wallnofer,P. R. (1980). Bacterial degradation of modelcompounds for lignin and chlorophenol derivedlignin bound residues. FEMS Microbiology Letters, 8,259-263.

Rauzier, J., Moniz-Pereira, J. & Gicquel-Sanzey, B.(1988). Complete nucleotide sequence of pAL5000,a plasmid from Mycobacterium fortuitum. Gene, 71,315-321.

Reyes, O., Guyonvarch, A., Bonamy, C , Salti, V.,David, F. & Leblon, G. (1991). 'Integron'-bearingvectors: a method suitable for stable chromosomalintegration in highly restrictive Corynebacteria.Gene, 107, 61-68.

Sanchez, F., Pefialva, M. A., Patino, C. & Rubio, V.(1986). An efficient method for the introduction ofviral DNA into Brevibacterium lactofermentumprotoplasts. Journal of General Microbiology, 132,1767-1770.

Sandoval, H., del Real, G., Mateos, L. M., Aguilar, A.& Martin, J. F. (1985). Screening of plasmids innon-pathogenic corynebacteria. FEMS MicrobiologyLetters, 27, 93-98.

Santamaria, R., Gil, J. A., Mesas, J. M. & Martin, J. F.(1984). Characterization of an endogenous plasmidand development of cloning vectors and atransformation system in Brevibacteriumlactofermentum. Journal of General Microbiology, 130,2237-2246.

Santamaria, R. I., Gil, J. A. & Martin, J. F. (1985).High-frequency transformation of Brevibacteriumlactofermentum protoplasts by plasmid DNA. Journalof Bacteriology, 162, 463-467.

Schafer, A., Kalinowski, J., Simon, R., Seep-Feldhaus,A.-H. & Puhler, A. (1990). High-frequency conjugalplasmid transfer from Gram-negative Escherichia colito various Gram-positive coryneform bacteria.Journal of Bacteriology, 172, 1663-1666.

Schwarzer, A. & Puhler, A. (1991). Manipulation ofCorynebacterium glutamicum by gene disruption andreplacement. Bio/Technology, 9, 84-87.

Sensfuss, C , Reh, M. & Schlegel, H. G. (1986). Nocorrelation exists between the conjugative transfer ofthe autotrophic character and that of plasmids inNocardia opaca strains. Journal of GeneralMicrobiology, 132, 997-1007.

Serwold-Davis, T. M., Groman, N. & Rabin, M.(1987). Transformation of Corynebacteriumdiphtheriae, Corynebacterium ulcerans, Corynebacteriumglutamicum, and Escherichia coli with the C.diphtheriae plasmid pNG2. Proceedings of the NationalAcademy of Sciences, USA, 84, 4964-4968.

Smith, M. D., Flickinger, J. L., Lineberger, D. W. &Schmidt, B. (1986). Protoplast transformation incoryneform bacteria and introduction of an a-amylase gene from Bacillus amyloliquefaciens intoBrevibacterium lactofermentum. Applied andEnvironmental Microbiology, 51, 634-639.

Snapper, S. B., Lugosi, L., Jekkel, A., Melton, R. E.,Kieser, T., Bloom, B. R. & Jacobs, W. R. Jr (1988).Lysogeny and transformation in mycobacteria:stable expression of foreign genes. Proceedings of theNational Academy of Sciences, USA, 85, 6987-6991.

Snapper, S. B., Melton, R. E., Mustafa, S., Kieser, T.& Jacobs, W. R. Jr (1990). Isolation andcharacterization of efficient plasmid transformationmutants of Mycobacterium smegmatis. MolecularMicrobiology, 4, 1911-1919.

Stover, C. K., de la Cruz, V. F., Fuerst, T. R.,Burlein, J. E., Benson, L. A., Bennett, L. T., Bansal,G. P., Young, J. F., Lee, M. H., Hatfull, G. F.,Snapper, S. B., Barletta, R. G., Jacobs, W. R. Jr &Bloom, B. R. (1991). New use of BCG forrecombinant vaccines. Nature {London), 351,456-460.

Tilford, P. E. (1936). Fasciation of sweet peas causedby Phytomonas fascians n.sp. Journal of AgriculturalResearch, 53, 383-394.

Tkachuk-Saad, O. & Prescott, J. (1991). Rhodococcusequi plasmids: isolation and partial characterization.Journal of Clinical Microbiology, 29, 2696-2700.

Vogt Singer, M. E. & Finnerty, W. R. (1988).Construction of an Escherichia coli - Rhodococcusshuttle vector and plasmid transformation inRhodococcus spp. Journal of Bacteriology, 170,638-645.

Yeh, P., Oreglia, J., Prevots, F. & Sicard, A. M.(1986). A shuttle vector system for Brevibacteriumlactofermentum. Gene, 47, 301-306.

Yeh, P., Oreglia, J. & Sicard, M. (1985). Transfectionof Corynebacterium lilium protoplasts. Journal ofGeneral Microbiology, 131, 3179-3183.

Yoshihama, M., Higashiro, K., Roa, E., Akedo, M.,Shanabruch, W. G., Folletie, M., Walker, G. C. &Sinskey, A. J. (1985). Cloning vector system forCorynebacterium glutamicum. Journal of Bacteriology,162, 591-597.

3Agrobacterium, Rhizobium, and OtherGram-negative Soil BacteriaAlan G. Atherly

Introduction

The introduction of DNA into bacterial cells isessential for rapid genetic analysis, i.e., mapping,complementation analysis, and as a tool for intro-duction of plasmids that contain cloned frag-ments. Several different approaches are used withGram-negative soil bacteria, including transfor-mation (natural and electroporation assisted),transduction and conjugation. Protoplast fusionhas not been successful in Gram-negative bacte-ria. Of all the procedures for DNA introduction,transformation is the most frequently used owingto the advent of gene cloning and the need tointroduce plasmids (with cloned inserts) into bac-teria. In this respect, introduction of plasmids isimportant for complementation analysis, genera-tion of clone banks, integration of DNAsequences into the genome by recombination, andintroduction of a desired gene into a cell on a sta-bly maintained plasmid. Only a limited number ofbacterial species permit natural transformation(e.g. Haemophilus influenzae, Streptococcus pneumo-nia, Bacillus subtilis and Acinetobacter calcoaceticus),but several bacteria have become amenable totransformation after treatment with special chemi-cals and/or heat cycles (Escherichia coli, Rhizobium,Agrobacterium, and Bacillus cereus). More recently,a wide variety of bacteria have become trans-formable by the application of high electric cur-rents, which causes membrane permeability,allowing uptake of large DNA molecules. Thisprocess, termed electroporation, has had a dra-matic impact on the ability to insert any desiredDNA sequence into any bacterial species.

The genera Rhizobium, Bradyrhizobium and

Agrobacterium are economically important soilbacteria; Rhizobium and Bradyrhizobium fix nitro-gen in symbiosis with the family Leguminosae,and Agrobacterium, besides causing plant diseases,is used as a vector system for introduction of for-eign DNA into dicotyledonous plants. Rhizobiumand Agrobacterium are Gram-negative soil bacte-ria, possess considerable DNA sequence similarity(Gibbins & Gregory, 1972), and have similargrowth characteristics and requirements. Theslow-growing genus Bradyrhizobium is more dis-tantly related (Elkan, 1981). Other importantGram-negative bacteria considered in this revieware Azospirillum, Erwinia and Xanthomonas.Bacteria of the genus Azospirillum are diazotrophsassociated with the roots of grasses and causeincreased crop yields, and the genera Erwinia andXanthomonas cause many different plant diseases.

Vectors and selectable markers

A considerable number of DNA cloning vectorshave been developed over the years; however, onlya limited number are useful in Agrobacterium,Rhizobium and other Gram-negative soil bacteria.Three groups of broad host range vectors havebeen developed and have been used in a largenumber of Gram-negative bacteria includingKlebsiella, Serratia, Pseudomonas, Acinetobacter,Xanthomonas, Erwinia, Azorhizobium, as well asBradyrhizobium, Rhizobium and Agrobacteriumspecies. The first of these vectors (pRK290) wasdeveloped by Ditta et al (1980) from RK2, a large(56 kb) P-l incompatibility group plasmid.pRK290 is a relatively large cloning vector (20 kb)

23

24 ATHERLY

with a tetracycline resistance marker and lacks thetra genes for mobilization. pRK290 contains a sin-gle recognition site for restriction enzymes EcoBIand Bglli that is suitable for cloning. This vectorcan be easily transformed into Gram-negativebacteria using electroporation or conjugated fromEscherichia coli in conjunction with the helperplasmid pRK2013, which contains the RK2 trans-fer genes cloned onto a ColEl replicon (Figurski& Helinski, 1979). A cosmid of pRK290 was cre-ated by inserting a cos sequence into the uniqueBgRl site (pLAFRl, pVKlOO; Friedman et al.,1982; Knauf & Nester, 1982). These plasmids/cosmids have been widely used to support replica-tion of recombinant molecules in Agrobacteriumand Rhizobium. Other IncP plasmids, R772 andpTJS75, have also been used but have technicallimitations similar to those of pRK290, i.e. fewunique restriction sites, and scarcity of antibioticresistance markers. pTJS75 (Schmidhauser &Helinski, 1985) has a reduced size, but largeinserts tend to be unstable and rearrange (R. K.Prakash & A. G. Atherly, unpublished data).Another large IncP plasmid, pVSl, was isolatedfrom Pseudomonas aeruginosa and is capable ofreplication in a wide variety of Gram-negativebacteria, but not in E. coli (Itoh et al., 1984).However, when an 8 kb segment of pVSl contain-ing the replication, stability and mobilizationregions was ligated into pBR325 the resultingplasmid, pGV910, was capable of replication in alltested Gram-negative bacteria including E. coliand simultaneously acquired compatability withother IncP plasmids such as pRK290 (Van denEede et al, 1992). Thus, pGV910 can be used incombination with other P-type plasmids to form anovel binary vector system.

A second group of broad host range cloning vec-tors has been developed from the smaller 8.9 kb7wcQ/P4 plasmid RSF1010 (Guerry, van Embden& Falkow, 1974; Bagdasarian et al, 1981).Chloramphenicol/streptomycin resistant deriva-tives of RSF1010 have been prepared (pJRD215,10.2 kb; Davison et al, 1987) and have consider-able advantage over pRK290 as a wide range ofunique cloning sites is available (EcoBI, Sstl,Hindlll, Xmal, Xhol, Sail, BamUl, and Clal) andthe size of the vectors is greatly reduced, allowinglarger inserts for transformation.

A third group of cloning vectors comes from theIncW plasmid pSa, which contains a small (1.5 kb)well-characterized region of DNA that supports

replication in a wide range of bacterial hosts. Taitet al (1983), Leemans et al. (1982), Shaw et al.(1983) and Close, Zaitlin & Kado (1984) devel-oped a set of relatively low copy number plasmids,ranging from 5.8 to 15 kb, which have wide appli-cation in Gram-negative soil bacteria. These pSavectors have been modified by increasing the num-ber of convenient cloning sites and phenotypicmarkers. In addition to the endogenous spectino-mycin and streptomycin resistance genes, newlyconstructed plasmids include kanamycin(pGV1106, pSal51 and pSa4), sulfonamide(pGV1113), tetracycline (pGV1122) and chloram-phenicol (pGVl 124, pSa4 and pSal52) resistancegenes. pSa-derived vectors have been shown toreplicate in E. coli, KJebsiella, Serratia, Erwinia,Rhizobium, Agrobacterium, Pseudomonas andAlcaligenes species (Tait et al, 1983).

Other potential sources of cloning vectors forsoil bacteria are the replication origins of stableendogenous plasmids, with the addition of selec-table markers and cloning sites. Mozo, Cabrera &Ruiz-Argueso (1990) cloned the origin of DNAreplication (about 5 kb) from a small cryptic plas-mid from Rhizobium species (Hedysarum) andfound it to be more stable than RK2 derivatives ina variety of Rhizobium and Agrobacterium strains,in the absence of selective pressure. In additionAgrobacterium plasmid replication regions havebeen cloned (Gallie et al, 1984; Gallie, Hagiya &Kado, 1985a; Nishiguchi, Takanami & Oda,1987; Tabata, Hooykaas & Oda, 1989) and somecloning vectors have been constructed using thesereplicator regions (Gallie, Novak & Kado, 19856;Tabata et al, 1989). Tabata et al. (1989) insertedthe 6.8 kb replicator region of the Ti plasmidpTiB6S3 into a ColEl replicon and found it to bestably maintained in Agrobacterium. Likewise,Gallie et al. (19856) constructed a very stable lowcopy number plasmid/cosmid from the replicatorregion of pTAR of A. tumefaciens (pUCDIOOland pUCD2001). Some useful cloning vectors aresummarized in Table 3.1.

Methods for introduction of DNA intobacteria

Transformation

Early reports on the transformation of Agro-bacterium and Rhizobium occurred in the1960s and have been extensively reviewed

Table 3.1. Cloning vectors for genetic analysis of Gram-negative soil bacteriaGram-negative soil bacteria 25

Plasmids Size(kb)

Selectable markers and other characteristics Reference

pRK290pVK102pLAFRIpTJS75pJRD215pUCD2pUCD5pUCS2001pC22

202321.67.5

10.2131310.417.5

Tc, RK2 oh, IncPTc, Km, cos, RK2 oh, IncPTc, cos, RK2 oh, IncPTc, RK2 oh, IncPKm, Sp, cos, RFS1010 oh, IncOKm, Sp, Cb, Tc, ColE1 oh, pSa oh, lnc\NpUCD with cosKm, Tc, Ap, cos, ColE1 oh, pTAR ohKm, Cb, Sp, cos, pTAR oh, binary vector forAgrobactehum

D\Ua etal. (1980)Knauf&Nester(1982)Friedman etal. (1982)Schmidhauser & Helinski (1985)Davison etal. (1987)Close etal. (1984)Close et al. (1984)GalWe etal. (19856)Simoens etal. (1986)

Notes:Tc, tetracycline; Sp, spectinomycin; Km, kanamycin; Cb, carbenicillin; Sm, streptomycin; oh, origin of replication; Ap,ampicillin; cos, phage X cohesive ends.

(Schwinghamer, 1977; Kondorosi & Johnston,1981). Many of the data from these early reportswere not reproducible and gave widely varyingresults with respect to DNA concentration andfrequency of transformants.

Gram-negative bacteria can take up and stablyestablish exogenous DNA. Uptake is dependentupon a transitory state of competence for transfor-mation, which is generally related to both the con-ditions of growth and die circumstances underwhich the cells and DNA are combined. The pres-ence of divalent cations often plays an essentialrole. Escherichia coli was the first nonnaturallycompetent Gram-negative bacteria where condi-tions were found that allowed uptake of DNA(Mandel & Higa, 1970) and Ca2+ is an absoluterequirement. Subsequently, Holsters etal. (1978)developed a procedure for transformation ofAgrobacterium that has been modified (Ebert, Ha& An, 1987) to improve the transformation fre-quency. This procedure has also been adapted fortransformation of Rhizobium meliloti (Selvaraj &Iyer, 1981; Courtois, Courtois & Guillaume,1988) and its close relatives (Bullerjahn &Benzinger, 1982), but not Bradyrhizobium species.Transformation frequencies for both Rhizobiumand Agrobacterium are as high as 1CF5 with largeplasmids such as pRK290 and both requiregrowth in special medium, a temperature shock at0 °C and 37 °C, followed by a recovery period of3-4 h at 28 °C, also in a special medium.Competent agrobacteria can be stored at -70 °Cfor use at a later time (Holgen & Willmitzer,1988).

Transformation of B. japonicum poses a special

problem, and high frequencies of transformationhave been obtained only by electroporation (seeElectroporation, below). Berry & Atherly (1981)reported the transformation of B. japonicumstrains with RP1 after spheroplast formation andtreatment with polyethylene glycol; however, thefrequencies of transformation were very low(10-7).

The soil bacterium Acinetobacter calcoaceticus(strain BD4), in contrast, has been shown toacquire high natural competence during normalgrowth (Juni & Janik, 1969) and is transformed byseveral plasmids (Singer, Van Tuijl & Finnerty,1986). In fact, A. calcoaceticus is seemingly com-petent in natural environments as it is able to takeup DNA in groundwater in the presence of di-valent cations (Lorenz, Reipschlager & Wacker-nagel, 1992).

Electroporation

In 1973 it was discovered that high current pulsesin a narrow range of intensity and duration resultin a transient and reversible increase in plasmamembrane permeability (Zimmerman, Schultz &Pilwat, 1973; Zimmerman, 1983). After a micro-second pulse of electric current the lifespan of thefield-induced pores is strongly dependent on thetemperature, very likely due to the fluidity of theproteins and lipids in the membrane. As a conse-quence, time, intensity of the current, and thetemperature are very tightly regulated for each celltype. In practice, a highly concentrated suspen-sion of cells in a nonconductive medium isexposed to a high voltage source for a few

26 ATHERLY

microseconds by the discharge of a capacitorthrough the system, the time of exposure to thefield being determined by the time required forcapacitor discharge. The cell membrane acts as anelectrical insulator that is unable to pass current.The high voltage surge results in the formation ofpores that are large enough to allow macromole-cules such as double-stranded DNA molecules topass into the cell. The closure of the pores is a nat-ural decay process that is delayed by holding thecells at a low temperature, usually 0 °C.

Electroporation is simple and rapid and wasfirst applied to many different eukaryotic celltypes, some of which were refractory to transfor-mation. These included many mammalian, fungaland plant cell types (Potter, Weir & Leder, 1984;Fromm, Taylor & Walbot, 1985; Shigekawa &Dower, 1988). Only recently has electroporationbeen applied to both Gram-positive and Gram-negative bacteria and has generally met with uni-form success (Chassy & Flickinger, 1987; Dower,Miller & Ragsdale, 1988; Luchansky, Muriana &Klaenhammer, 1988). Correspondingly, the fol-lowing Gram-negative soil bacteria are easilytransformed using electroporation: Rhizobium,Bradyrhizobium, Agrobacterium, Erwinia, Pseudo-monaSy Xanthomonas and Azospirillum (Broek, vanGool & Vanderleyden, 1989; Jun & Forde, 1989;Mattanovich et al., 1989; Wirth, Friesenegger &Fiedler, 1989; Guerinot, Morisseau & Kapatch,1990; Hatterman & Stacey, 1990; Nagel et al>1990; White & Gonzolas, 1991). Electroporation-assisted transformation of DNA directly into soilbacteria has many advantages over conjugationfrom E. coli. For example, with Agrobacterium tri-parental matings from E. coli are routinely used toinsert the binary Ti plasmids for eventual infectionof plant cells, and this sometimes leads to plasmidrearrangements (An et at., 1988) as well as neces-sitating that the recipient Agrobacterium strainpossess an antibiotic resistance marker. Electro-poration-assisted transformation also allows theintroduction of DNA directly into B. japonicumstrains, where plasmid introduction was previ-ously restricted to conjugation (Guerinot et al.,1990; Hatterman & Stacey, 1990). Azospirillumcan now be transformed by means of electro-poration, where previously no method was avail-able for introducing DNA (White & Gonzolas,1991).

The efficiency of electroporation-assisted trans-formation was initially low (Wirth et al.> 1989) but

with growth media changes and optimization ofelectroporation conditions and equipment, theefficiency has increased to dramatic levels(106-10%ig DNA; Nagel et al.9 1990). Some fac-tors that affect the efficiency of electroporation-assisted transformation include excretion ofnucleases that degrade the added DNA, effectiverestriction endonuclease systems for digesting theunmodified DNA (Wirth et al., 1989), and thepresence of large quantities of extracellular poly-saccharide (Regue etal, 1992).

Electroporation can also assist in the transfer ofplasmids directly between bacterial strains, with-out the necessity of purifying the DNA from thedonor strain. This has been dubbed 'electro-duction' (Pfau & Youderian, 1990). The strainpossessing the donor plasmid is mixed with therecipient strain and the mixture is subjected to atypical electrical discharge for Gram-negative bac-teria (15-20 kV/cm, 200 Q, 25 ^F in a 0.10 cmwide cuvette). The frequency of plasmid transferis as high as 4 X 10~5 per recipient cell (Pfau &Youderian, 1990). This phenomenon is likely tobe due to the diffusion of the plasmid out of thedonor cells, via the generated pores, and subse-quent uptake by the recipient cells, and not due tofusion of the two cell types.

Transformation frequencies dramatically de-crease with increasing size of the double-strandedDNA molecule. In E. coli the transformation fre-quency is lower with plasmids larger than 50 kb.This restricts the preparation of genome librarieswith large insert sizes. In contrast, the plasmamembrane pores generated during electroporationof bacteria are seemingly quite big, as extremelylarge molecules of DNA can be transformed usingelectroporation. Mozo & Hooykaas (1991) wereable to introduce DNA molecules as large as 250kb (the pTiB6::R772 plasmid of A. tumefaciens)into A. tumefaciens, using electroporation, withlow but reproducible efficiencies: 2.7 X 10"9

transformants/survivor versus 1.4 X 10~6 for a 20kb plasmid with crude DNA preparations. Highelectroporation efficiencies were also obtainedwith E. coli cells using plasmids up to 136 kb insize (Leonardo & Sidivy, 1990).

Conjugation

Cloning vectors usually have mob or tra functionsremoved to prevent unwanted plasmid transfers

Gram-negative soil bacteria 27

into undesirable bacteria. For example, RK2derivative plasmids lack the tra genes, thus con-jugative transfer from a donor to a recipient strainis accomplished by supplying the tra genes in transon a ColEl plasmid (pRK600 or similar plas-mids). The ColEl plasmid replicates only in E.coli, not in Agrobacterium or Rhizobium, and therecipient strain is selected by growth on mediumwith an appropriate antibiotic.

Plasmids coding for symbiotic functions insome strains of R. leguminosarum, R. trifolii and R.phaseoli are self-transmissible (Johnson, Bibb &Beringer, 1978; Beynon, Beringer & Johnston,1980; Hooykaas et al, 1981; Rolfe et al, 1981;Hooykaas, Snijdewint & Schilperoort, 19826;Lamb, Hombrecher & Johnston, 1982; Scott &Ronson, 1982), but this is not the case in R.meliloti, R. fredii and B. japonicum (Kondorosi etal, 1982; Appelbaum et al, 1985; Atherly et al,1985). Several approaches have been devised totransfer entire plasmids from one strain toanother. For example, Hooykaas, Den Dulk-Ras& Schilperoort (1982a) developed a helper plas-mid (pRL180) that assists in the transfer of non-transmissible plasmids. Kondorosi et al (1982)inserted the mob region of the P-l type plasmidRP4 into the large plasmid of R. meliloti to obtainplasmid transfer. Mobilization of a nontransmissi-ble plasmid of R. trifolii was achieved by cointegra-tion with the transfer-proficient plasmid R68.45(Scott & Ronson, 1982). A more generally usedsystem utilizes a mob::Tn5 element present in aColEl-based vector (Simon, 1984). Conjugationof this plasmid into Gram-negative soil bacteria,with selection for kanamycin, results in the ran-dom insertion of the mob::Tn5 element into thegenome. This approach has been used for plasmidand chromosome transfer, utilizing tra genes intrans, with excellent success (Glazebrook &Walker, 1991), permitting high frequency ofrecombination (Hfr)-like mapping. The mob::Tn5system has also been used to transfer plasmidsfrom Azospirillum lipoferum into A. tumefaciens(Bally & Givaudan, 1988).

The Ti plasmids of A. tumefaciens are trans-missible between strains (Kerr, Manigault &Tempe, 1977; Genetello et al, 1977) as well asinto other Gram-negative bacteria, such as R. tri-folii (Hooykaas et al., 1977). Transfer of plasmidsbetween strains is seemingly dependent upon spe-cific virulence iyir) gene functions as mutations invirA, virG, virB, and virE operons (Steck & Kado,

1990), and, in some cases, virC (Gelvin &Habeck, 1990) greatly decreases plasmid trans-fers. The presence or absence of acetosyringonehas no effect. In contrast, Beijersbergen et al.(1992) found that the nonconjugative plasmid,pKT230 (a derivative of IncQ plasmid RSF1010),could be mobilized between Agrobacterium strainsand mobilization was dependent upon the pres-ence of acetosyringone, the virA, virG and virBoperons and virD4. These data suggest thatT-DNA (transferred-DNA) transfer from Agro-bacterium to plants may occur in a manner similarto bacterial conjugation.

Transduction

Although little used, transduction is a valuableadjunct to plasmid cloning vectors and transfor-mation in the genetic analysis of Rhizobium.Generalized transducing phages are available forR. meliloti (Casadesus & Olivaries, 1979; Sik,Hovath & Chatterjee, 1980; Finan et al, 1984;Martin & Long, 1984; Williams, Klein & Signer,1989; Novikova et al, 1990; Glazebrook &Walker, 1991), R. leguminosarum and R. trifolii(Buchanan-Wollaston, 1979) and B. japonicum(Shah, Sousa & Modi, 1981). Phage can be usedfor fine-structure mapping, strain constructionand for the transfer of large plasmids. The molec-ular size of most Rhizobium phage genomes is150-200 kb, thus allowing transfer of encapsu-lated intact plasmids (Martin & Long, 1984;Glazebrook & Walker, 1991). Unfortunately,most Rhizobium phages are very strain specific.

Vector and DNA stability

Most cloning vectors now in use in Gram-negativesoil bacteria are derived from broad host rangereplicons (Table 3.1) and many of these vectorsare unstable during growth under nonselectiveconditions. For example, Agrobacterium andRhizobium cells lose pRK290 and its derivatives ata rate ranging between 0.03% and 0.25%/genera-tion (Ditta et al, 1980; Close et al, 1984).Plasmids derived from the IncW origin from plas-mid pSa are lost at an even more rapid rate, rang-ing between 2% and 13%/generation dependingupon the construction (Close et al, 1984). Thisrapid loss likely reflects the nontransmissibility ofthese plasmids as pSa, which is self-transmissible,

28 ATHERLY

shows no loss even after 40 generations (Close etal, 1984).

A highly stable cloning vector is desirable forsymbiotically expressed genes. In Rhizobium cells,RK2-derived vectors are relatively unstable duringsymbiosis with the host plant (Long, Buikema &Ausubel, 1982; Lambert et al, 1987), althoughWeinstein, Roberts & Helinski (1992) stabilizedRK2-derived plasmids by introducing a 3.2 kbDNA fragment (pTR102 and pMW708) or a0.8 kb DNA fragment (pTRlOl and pMW707)derived from the RK2 stabilization region. Noplasmid loss was observed after 100 generations ofnonselective growth, nor during nodule passage.Gallie et al (19856) constructed a series of cloningvectors using the origin of replication from the sta-ble endogenous plasmid (pTAR) of AgrobacteriumLBA4301. These vectors (pUCD2001 andpUCDIOOl) are extremely stable in AgrobacteriumLBA4301 and Rhizobium meliloti RM102Z1 undernonselective conditions, showing no loss after 50generations. The copy number was low, asexpected from the presence of par gene functions,which are required for plasmid replication andpartitioning. Similarly, Simoens et al (1986) pre-pared a binary vector (pC22) utilizing the origin ofreplication from A. rhizogenes plasmid pRiHRl,which is very stable under nonselective condi-tions. Mozo et al (1990) cloned the origin ofDNA replication (about 5 kb) from a small crypticplasmid from Rhizobium species (Hedysarum) andfound it to be more stable than RK2 derivatives ina variety of Rhizobium and Agrobacterium strains inthe absence of selective pressure. This plasmidsequence seemingly possesses par genes similar toAgrobacterium plasmids. The Rhizobium species(Hedysarum) origin of replication is more stablethan RK2-derived plasmids in a variety ofRhizobium strains under symbiotic conditions,although foreign DNA inserts have not beentested in this plasmid or in pUCD2001 orpUCDIOOl.

Another approach for stably inserting a desiredDNA sequence into Rhizobium is via homologousrecombination, either into a stable endogenousplasmid or into the chromosome. Care must betaken that the insertion does not inactivate a genewith an unknown but desirable function, thus site-specific recombination is the safest approach.Legocki, Yun & Szalay (1984) inserted DNAsequences into Rhizobium species by homologousrecombination, but relied upon randomly cloned

fragments to find a 'nonessentiaP region of thegenome. In a related study (Yun, Noti & Szalay,1986), nif genes were part of a genomic target, butthis had the risk of creating undesired, symbioti-cally altered strains after integration. Acuna et al(1987) constructed a vector (pRJ1035) that wasshown to integrate unselected, cloned DNA into aspecific and nonessential region of the B. japon-icum genome. Unfortunately, this vector isrestricted to use with B. japonicum strains, sincerecombination depends upon the presence ofrepeated sequences not found in other species.

Another important consideration is genomicrearrangements that result from insertion of aplasmid into a bacterial cell. Plasmid transfer intoAgrobacterium and Rhizobium by conjugation fre-quently results in plasmid rearrangements (Berry& Atherly, 1984; An et al, 1988; Shantharam,Engwall & Atherly, 1988). Although no conclu-sive studies have been done, introduction of DNAby transformation, either natural or electropora-tion assisted, seemingly produces fewer rearrange-ments of cloned fragments.

Applications

Some obvious uses of introduced DNA intoGram-negative soil bacterial cells include comple-mentation analysis, introduction of a new, desir-able gene or genes, and gene replacement byhomologous recombination.

An example of complementation analysis is theexperiments of Long et al (1982), who were thefirst to clone a DNA fragment in R. meliloti thatconferred nodulation ability. They mated a R.meliloti clone bank from E. coli en masse into a nod~R. meliloti strain and selected for nod+ derivativesby direct selection on alfalfa plants (Medicagosativa). Similarly, Innes, Hirose & Kuempel(1988) constructed a clone from R. trifolii in aRK2-type plasmid and, after conjugation into apSym plasmid-cured background, found a 32 kbclone that was able to confer nodulation andnitrogen fixation. Hahn & Hennecke (1988),using a deletion mutant of B. japonicum defectivein nodulation, selected clones that complementedthe inability to form nodules. Many examplesexist, as complementation is a commonly usedapproach for cloning genes in mutant strainswhere the clone restores the original phenotype.

Introduction of new, desirable genes into


Rhizobium or other Gram-negative bacteria is lessfrequent, largely due to the lack of identified desir-able genes. Lambert et al. (1987) introduced thegene for hydrogenase, originally cloned from B.japonicum^ into Hup B. japonicum, R. meliloti andR. leguminosarum strains. It was believed that theadded hydrogenase would increase the hydrogenrecycling capability of these strains and improvethe efficiency of nitrogen fixation; however, theeffect is minimal. Even when and if desirablegenes are found and stably introduced intoRhizobium strains, it is unlikely that the newstrains could be utilized in a natural symbiosis inthe soil, as endogenous strains out-compete intro-duced strains unless massive numbers of bacteriaare added.

With the discovery of electroporation-assistedtransformation, Gram-negative soil bacteria canbe used for direct preparation of clone banks,bypassing the need to prepare the clone bank in E.coli and then conjugate the clones into the desiredstrain. Conjugation requires triparental mating,selectable markers on both the helper plasmid andthe mated plasmid, and frequently causesrearrangements of inserted fragments (see above).Preparation of a clone bank directly in a desirablestrain, or direct transformation for complementa-tion analysis, bypasses these obstacles. In thisrespect, Simoens et al. (1986) utilized the stablyreplicating binary cloning vector pC22, whichcontains T-DNA border sequences, a kanamycinresistance gene and BamHl and Xbal cloning sitesbetween the border sequences, and a strepto-mycin resistance gene for plasmid selection to pre-pare a clone bank of Arabidopsis thaliana in E. coli.This clone bank was conjugated into an appropri-ate Agrobacterium strain that was used to introduceDNA fragments into plant cells. Unfortunatelythey found that a large percentage (60%) ofthe clones were unstable in the absence of selec-tion, both in E. coli and in Agrobacterium.Rearrangements were likely due to the presence ofrepeated sequences, as Simoens et al. found alarge number of rearrangements in clones contain-ing ribosomal RNA sequences. In addition, theyfound that conjugation between E. coli andAgrobacterium preferentially selected against themore unstable DNA clones. Direct preparation ofthe library in the Agrobacterium host using electro-poration-assisted transformation, however, wouldbypass this problem.

Another use of electroporation-assisted cloning

into Gram-negative soil bacteria is the directcloning of DNA replicator regions of Gram-negative replicons. Replicator regions of Gram-negative soil bacteria have made very stablecloning vectors and it is desirable to find others. Adigest of total DNA from a desired strain wouldbe cloned into a ColEl-based vector, which is in-capable of replication in Gram-negative soil bacte-ria. The clone bank is transformed into aplasmidless strain, for example a cured Agro-bacterium strain, where only clones possessing thereplicator region replicate.

In summary, many stable cloning vectors arenow available for a wide range of Gram-negativesoil bacteria and with the advent of electropora-tion, it is relatively easy to introduce DNAsequences into any bacterial strain. It is very likelythat we will see stable genetically engineered soilbacteria in general use in the USA in the nearfuture, assuming that Federal regulatory agenciesallow their introduction into the soil.

ReferencesAcuna, G., Alverez-Morales, A., Hahn, M. &

Hennecke, H. (1987). A vector for the site-directed,genomic integration of foreign DNA into soybeanroot-nodule bacteria. Plant Molecular Biology, 9,41-50.

An, G., Ebert, P. R. Mitra, A. & Ha, S. B. (1988).Binary vectors. In Plant Molecular Biology ManualA3 Kluwer Academic Publisher, Dordrecht,pp. 1-19.

Appelbaum, E. R., McLoughlin, T. J., O'Connell, M.& Chartrain, N. (1985). Expression of symbioticgenes of Rhizobium japonicum USDA191 in otherrhizobia. Journal of Bacteriology, 163, 385-388.

Atherly, A. G., Prakash, R. K., Masterson, R. V., DuTeau, N. B. & Engwall, K. S. (1985). Theorganization of genes involved in symbiotic nitrogenfixation on indigenous plasmids of Rhizobiumjaponicum. Proceedings of the World SoybeanConference ill, ed. R. Shibles, pp. 229-300. WestviewPress, Boulder, Co.

Bagdasarian, M., Lurz, R., Riichert, B., Franklin,F. C. H., Bagdasarian, M. M., Frey, J. & TimmisK. N. (1981). Specific-purpose plasmid cloningvectors. II. Broad host range, high copy number,RSFlOlO-derived vectors, and a host-vector systemfor gene cloning in Pseudomonas. Gene, 16, 237-247.

Bally, R. & Givaudan, A. (1988). Mobilization andtransfer of Azospirillum lipoferum plasmid by theTn5-mob transposon into a plasmid-freeAgrobacterium tumefaciens strain. Canadian Journal ofMicrobiology, 34, 1354-1357.

30 ATHERLY

Beijersbergen, A., Dulk-Ras, A. D., Schilperoort, R. A.& Hooykaas, P. J. J. (1992). Conjugative transfer bythe virulence system of Agrobacterium tumefaciens.Science, 256, 1324-1327.

Berry, J. O. & Atherly, A. G. (1981). Introduction ofplasmids into Rhizobium japonicum by conjugationand spheroplast transformation. In Proceedings of the8th North American Rhizobium conference, ed. R. W.Clark & J. H. W. Stephens, pp. 115-128. Universityof Manitoba Press, Winnipeg, Manitoba.

Berry, J. O. & Atherly, A. G. (1984). Induced plasmidgenome rearrangements in Rhizobium japonicum.Journal of Bacteriology, 157, 218-224.

Beynon, J. L., Beringer, J. E. & Johnston, A. W. B.(1980). Plasmids and host range in Rhizobiumleguminosarum and Rhizobium phaseoli. Journal ofGenetic Microbiology, 120, 421-429

Broek, A. V., van Gool, A. & Vanderleyden, J. (1989).Electroporation of Azospirillum brasilense withplasmid DNA. FEMS Microbiology Letters, 61,177-182.

Buchanan-Wollaston, V. (1979). Generalizedtransduction in Rhizobium leguminosarum. Journal ofGenetic Microbiology, 112, 135-142.

Bullerjahn, G. S. & Benzinger, R. H. (1982). Genetictransformation of Rhizobium leguminosarum byplasmid DNA. Journal of Bacteriology, 150, 421-424.

Casadesus, J. & Olivaries, J. (1979). Generaltransduction in Rhizobium meliloti by athermosensitive mutant of bacteriophage DF2.Journal of Bacteriology, 139, 316-317.

Chassy, B. M. & Flickinger, J. L. (1987).Transformation of Lactobacillus casei byelectroporation. FEMS Microbiology Letters, 44,173-177.

Close, T. J., Zaitlin, D. & Kado, C. I. (1984). Designand development of amplifiable broad-host-rangecloning vectors: analysis of the vir region ofAgrobacterium tumefaciens plasmid pTiC58. Plasmid,12, 111-118.

Courtois, J., Courtois, B. & Guillaume, J. (1988).High-frequency transformation of Rhizobium meliloti.Journal of Bacteriology, 170, 5925-5927.

Davison, J., Heusterspeute, M., Chevalier, N., Ha-Thi, V. & Brunei, F. (1987). Vectors with restrictionsite banks, v. pJRD215, a wide-host-range cosmidvector with multiple cloning sites. Gene, 51,275-280.

Ditta, G., Stanfield, S., Corbin, D. & Helinski, D. R.(1980). Broad host range DNA cloning system forGram-negative bacteria, construction of a gene bankof Rhizobium meliloti. Proceedings of the NationalAcademy of Sciences, USA, 11, 7347-7351.

Dower, W. J., Miller, J. F. & Ragsdale, C. W. (1988).High efficiency transformation of E. coli by highvoltage electroporation. Nucleic Acid Research, 16,6127-6145.

Ebert, P. R., Ha, S. B. & An, G. (1987). Identificationof an essential upstream element in the nopalinesynthase promoter by stable and transient assays.Proceedings of the National Academy of Sciences, USA,84, 5745-5749.

Elkan, G. H. (1981). The taxonomy of theRhizobiaceae. In The Biology of the Rhizobiaceae,International Review of Cytology Suppl. 13, ed. K. L.Giles & A. G. Atherly, pp. 1-14. Academic Press,New York.

Figurski, D. & Helinski, D. R. (1979). Replication ofan origin containing derivative of plasmid RK2dependent on a plasmid function provided in trans.Proceedings of the National Academy of Sciences, USA,76, 1648-1652.

Finan, T. M., Hartwieg, E., Lemieux, K., Bergman,K., Walker, G. C. & Signer, E. R. (1984). Generaltransduction in Rhizobium meliloti. Journal ofBacteriology, 159, 120-124.

Friedman, A. M., Long, S. R., Brown, S. E., Buikema,W. J. & Ausubel, F. M. (1982). Construction of abroad host range cosmid cloning vector and its usein the genetic analysis of Rhizobium mutants. Gene,18, 289-296.

Fromm, M., Taylor, L. P. & Walbot, V. (1985).Expression of genes transferred into monocot anddicot plant cells by electroporation. Proceedings of theNational Academy of Sciences, USA, 82, 5824-5828.

Gallie, D. R., Hagiya, M. & Kado, C. I. (1985a).Analysis of Agrobacterium tumefaciens plasmidpTiC58 replication region with a novel high copynumber derivative. Journal of Bacteriology, 161,1034-1041.

Gallie, D. R, Novak, S. & Kado, C. I. (19856). Novelhigh and low copy number stable cosmids for use inAgrobacterium and Rhizobium. Plasmid, 14, 171-175.

Gallie, D. R., Zaitlin, D., Perry, K. L. & Kado, C. I.(1984). Characterization of the replication andstability regions of Agrobacterium tumefaciens plasmidpTAR. Journal of Bacteriology, 157, 739-745.

Gelvin, S. B. & Habeck, L. L. (1990). vir genesinfluence conjugal transfer of the Ti plasmid ofAgrobacterium tumefaciens. Journal of Bacteriology,172, 1600-1608.

Genetello, C , Van Larebeke, N., Holsters, M., DePicker, A., Van Montagu, M. & Schell, J. (1977). Tiplasmids of Agrobacterium as conjugative plasmids.Nature (London), 265, 561-563.

Gibbins, A. M. & Gregory, K. F. (1972). Relatednessamong Rhizobium and Agrobacterium speciesdetermined by three methods of nucleic acidhybridization. Journal of Bacteriology, 111, 129-141.

Glazebrook, J. & Walker, G. C. (1991). Genetictechniques in Rhizobium meliloti. Methods inEnzymology, 204, 398-418.

Guerinot, M. L., Morisseau, B. A. & Kapatch, T.(1990). Electroporation of Bradyrhizobium japonicum.


Molecular and General Genetics, 221, 287-290.Guerry, P., Van Embden, J. & Falkow, S. (1974).

Molecular nature of two non-conjugative plasmidscarrying drug resistance genes. Journal ofBacteriology, 117, 619-630.

Hahn, M. & Hennecke, H. (1988). Cloning andmapping of a novel nodulation region fromBradyrhizobium japonicum by geneticcomplementation of a deletion mutant. Journal ofBacteriology, 54, 55-61.

Hatterman, D. R. & Stacey, G. (1990). Efficienttransformation of Bradyrhizobium japonicum byelectroporation. Journal of Bacteriology and AppliedEnvironmental Microbiology, 56, 833-836.

Holgen, R. & Willmitzer, L. (1988). Storage ofcompetent cells for Agrobacterium transformation.Nucleic Acids Research, 16, 9877.

Holsters, M., De Waele, D., Depicker, A., Messens, E.,Van Montagu, M. & Schell, J. (1978). Transfectionand transformation of. Agrobacterium tumefaciens.Molecular and General Genetics, 163, 181-187.

Hooykaas, P. J. J., Den Dulk-Ras, H. & Schilperoort,R. A. (1982a). Method for the transfer of largecryptic, non-self-transmissible plasmids: ex plantatransfer of the virulence plasmid of Agrobacteriumrhizogenes. Plasmid, 8, 94-96.

Hooykaas, P. J. J., Klapwijk, P. M., Nuti, M. P.,Schilperoort, R. A. & Rorsch, A. (1977). Transfer ofthe Agrobacterium tumefaciens Ti plasmid to avirulentagrobacteria and rhizobia ex planta. Journal ofGeneral Microbiology, 98, 477-484.

Hooykaas, P. J. J., Snijdewint, F. G. M. &Schilperoort, R. A. (19826). Identification of theSym plasmid of Rhizobium leguminosarum strain 1001and its transfer to and expression in other rhizobiaand Agrobacterium tumefaciens. Plasmid, 8, 73-82.

Hooykaas, P. J. J., Van Brussel, A. A. N., Den Dulk-Ras,H., Van Slogteren, G. M. S. & Schilperoort, R. A.(1981). Sym-plasmids of Rhizobium trifolii expressedin different rhizobial species and Agrobacteriumtumefaciens. Nature {London), 291, 351-353.

Innes, R. W., Hirose, M. A. & Kuempel, P. L. (1988).Induction of nitrogen fixing nodules on cloverrequires only 32 kilobase pairs of DNA from theRhizobium trifolii symbiosis plasmid. Journal ofBacteriology, 170, 3793-3800.

Itoh, Y., Watson, J. M., Haas, D. & Leisinger, T.(1984). Genetic and molecular characterizationof Pseudomonas plasmid pVSl. Plasmid, 11,206-220.

Johnson, A. W. B., Bibb, M. J. & Beringer, J. E.(1978). Tryptophan genes in Rhizobium - theirorganization and transfer to other bacterial genera.Molecular and General Genetics, 165, 323-330.

Jun, S. W. & Forde, B. G. (1989). Efficienttransformation of Agrobacterium spp. by high voltageelectroporation. Nucleic Acids Research, 17, 8385.

Juni, E. & Janik, A. (1969). Transformation ofAcinetobacter calcoaceticus. Journal of Bacteriology, 98,281-288.

Kerr, A., Manigault, P. & Tempe, J. (1977). Transferof virulence in vivo and in vitro in Agrobacterium.Nature (London), 265, 560-561.

Knauf, V. & Nester, E. W. (1982). Wide host rangecloning vectors, a cosmid clone bank of anAgrobacterium Ti plasmid. Plasmid, 8, 45-54.

Kondorosi, A. & Johnston, A. W. B. (1981). Thegenetics of Rhizobium. In The Biology of theRhizobiaceae, International Review of Cytology Suppl.13, ed. K. Giles & A. G. Atherly, pp. 191-224.Academic Press, New York.

Kondorosi, A., Kondorosi, E., Pankhurst, C. E.,Broughton, W. J. & Banfalvi, Z. (1982).Mobilization of a Rhizobium meliloti megaplasmidcarrying nodulation and nitrogen fixation genes intoother rhizobia and Agrobacterium. Molecular andGeneral Genetics, 188, 433-440.

Lamb, J. W., Hombrecher, G. & Johnston, A. W. B.(1982). Plasmid-determined nodulation andnitrogen fixation abilities in Rhizobium phaseoli.Molecular and General Genetics, 186, 449-452.

Lambert, G. R., Harker, A. R., Cantrell, M. A.,Hanus, F. J., Russell, S. A., Haugland, R. A. &Evans H. J. (1987). Symbiotic expression of cosmid-borne B. japonicum hydrogenase genes. Applied andEnvironmental Microbiology, 53, 422-428.

Leemans, J., Langenakens, J., De Greve, H., Deblaere,R., Van Montagu, M. & Schell, J. (1982). Broad-host-range cloning vectors derived from the W-plasmid Sa. Gene, 19, 361-364.

Legocki, R. P., Yun, A. C. & Szalay, A. A. (1984).Expression of 15-galactosidase controlled by anitrogenase promoter in stem nodules ofAeschynomene scabra. Proceedings of the NationalAcademy Sciences, USA, 81, 5806-5810.

Leonardo, E. D. & Sidivy, J. M. (1990). A new vectorfor cloning large eukaryotic DNA segments inEscherichia coli. Bio/Technology, 8, 841-844.

Long, S. R., Buikema, W. & Ausubel, F. M. (1982).Cloning of Rhizobium meliloti nodulation genes bydirect complementation of Nod" mutants. Nature(London), 198, 495-498.

Lorenz, M. G., Reipschlager, K. & Wackernagel, W.(1992). Plasmid transformation of naturallycompetent Acinetobacter calcoaceticus in non-sterilesoil extract and ground water. Archives ofMicrobiology, 157, 355-360.

Luchansky, J. B., Muriana, P. M. & Klaenhammer,T. R. (1988). Application of electroporation fortransfer of plasmid DNA to Lactobacillus,Lactococcus, Leuconostoc, Listeria, Pediococcus,Bacillus, Staphylococcus, Enterococcus, andPropionibacterium. Molecular Microbiology, 2,637-646.

32 ATHERLY

Mandel, M. & Higa, A. (1970). Calcium-dependentbacteriophage DNA infection. Journal of MolecularBiology, 53, 159-162.

Martin, M. O. & Long, S. R. (1984). Generalizedtransduction in Rhizobium meliloti. Journal ofBacteriology, 159, 125-129.

Mattanovich, D., Rucker, F., Machado, A. C , Lamer,M., Regner, F., Steinkelner, H., Himmler, G. &Katinger, H. (1989). Efficient transformation ofAgrobacterium spp. by electroporation. Nucleic AcidsResearch, 17, 6747.

Mozo, T., Cabrera, E. & Ruiz-Argueso, T. (1990).Isolation of the replication DNA region from aRhizobium plasmid and examination of its potentialas a replicon for Rhizobiaceae cloning vectors.Plasmid, 23, 201-215.

Mozo, T. & Hooykaas, P. J. J. (1991) Electroporationof megaplasmids into Agrobacterium. Plant MolecularBiology, 16,917-918.

Nagel, R., Elliott, A., Mase., A., Birch, R. G. &Manners, J. M. (1990). Electroporation of binary Tiplasmid vector into Agrobacterium tumefaciens andAgrobacterium rhizogenes. FEMS Microbiology Letters,67, 325-328.

Nishiguchi, R., Takanami, M. & Oda, A. (1987).Characterization and sequence determination of thereplicator region in the hairy-root inducing plasmidpRiA4b. Molecular and General Genetics, 206, 1-8.

Novikova, N. I., Safronova, V. I., Lyudvikova, E. K.& Simarov, B. V. (1990). Production ofrecombinant transducing phage of Rhizobiummeliloti with a broad spectrum of lytic activity.Genetika, 26, 37-42.

Pfau, J. & Youderian, P. (1990). Transferringplasmid DNA between different bacterial specieswith electroporation. Nucleic Acids Research, 18,6165.

Potter, H., Weir, L. & Leder, P. (1984). Enhancer-dependent expression of human immunoglobingenes introduced into mouse pre-fl-lymphocytes byelectroporation. Proceedings of the National Academyof Sciences, USA, 81, 7161-7165.

Regue, M., Enfedaque, J., Camprubi, S. & Tomas,J. M. (1992). The O-antigen lipopolysaccharide inthe major barrier to plasmid DNA uptake byKlebsiella pneumoniae during transformation byelectroporation and osmotic shock. Journal ofMicrobiology Methods, 15, 129-134.

Rolfe, B. G., Djordjevic, M., Scott, K. F., Hughes,J. E., Badennoch-Jones, J., Gresshoff, P. M., Chen,Y., Dudman, W. G., Zurkowski, W. & Shine, J.(1981). Analysis of nodule formation in fast growingRhizobium strains. In Current Perspectives in NitrogenFixation, ed. A. Gibson & A. Newton, pp. 142-145.Australian Academic Press, Canberra.

Schmidhauser, T. J. & Helinski, D. R. (1985). Regionsof broad host range plasmid RK2 involved in

replication and stable maintenance in nine species ofGram-negative bacteria. Journal of Bacteriology, 164,446-451.

Schwinghamer, E. A. (1977). Genetic aspects ofnodulation and dinitrogen fixation by legumes, themicrosymbiont. In A Treatise on Dinitrogen Fixation,ed. R. W. F. Hardy & W. S. Silver, Section III, pp.577-622. Wiley (Interscience), New York.

Scott, D. B. & Ronson, C. W. (1982). Identificationand mobilization by cointegrate formation of anodulation plasmid in Rhizobium trifolii. Journal ofBacteriology, 151, 36-43.

Selvaraj, G. & Iyer, I. V. (1981). Genetictransformation of Rhizobium meliloti by plasmidDNA. Gene, 15, 279-283.

Shah, K., Sousa, S. & Modi, V. V. (1981). Studies ontransducing phage M-l for Rhizobium japonicum.Archives of Microbiology, 130, 262-266.

Shantharam, S., Engwall, K. S. & Atherly, A. G.(1988). Symbiotic phenotypes of soybean rootnodules associated with deletions and rearrangementsin the symbiotic plasmid of R. fredii USDA 191.Journal of Plant Physiology, 132, 431-438.

Shaw, C. H., Leemans, J., Shaw, C. H., VanMontagu, M. & Schell, J. (1983). A general methodfor the transfer of cloned genes to plant cells. Gene,23,315-330.

Shigekawa, K. & Dower, W. J. (1988). Electroporationof eukaryotes and prokaryotes, a general approach tothe introduction of macromolecules into cells.BioTechniques, 6, 742-751.

Sik, T., Hovath, J. & Chatterjee, S. (1980).Generalized transduction in Rhizobium meliloti.Molecular and General Genetics, 178, 511-516.

Simoens, C , Alliote, T., Mendel, R., Miiller, A.,Schiemann, J., Van Lijsebettens, M., Schell, J., VanMontagu, M. & Inze, D. (1986). A binary vector fortransferring genomic libraries to plants. Nucleic AcidsResearch, 14, 8073-8090.

Simon, R. (1984). High frequency mobilization ofGram-negative bacterial replicons by the in vitroconstructed Tn5-Mob transposon. Molecular andGeneral Genetics, 196, 413-420.

Singer, J. T., Van Tuijl, J. J. & Finnerty, W. (1986).Transformation and mobilization of cloning vectorsin Acinetobacter spp. Journal of Bacteriology, 165,301-303.

Steck, T. R. & Kado, C. I. (1990). Virulence genespromote conjugative transfer of the Ti plasmidbetween Agrobacterium strains. Journal ofBacteriology, 172, 2191-2193.

Tabata, S., Hooykaas, P. & Oda, A. (1989). Sequencedetermination and characterization of the replicatorregion in the tumor-inducing plasmid pTiB6S3.Journal of Bacteriology, 171, 1665-1672.

Tait, R. C , Close, T. J., Lundquist, R. C , Haga, M.,Rodriguez, R. L. & C. I. Kado. (1983).


Construction and characterization of a versatilebroad host range DNA cloning system for Gram-negative bacteria. Bio/Technology, 1, 269-275.

Van den Eede, G., Deblaere, R., Goethals, K., VanMontagu, M. & Holsters, M. (1992). Broad hostrange and promoter selection vectors for bacteriathat interact with plants. Molecular Plant-MicrobeInteractions, 5, 228-234.

Weinstein, M., Roberts, R. C. & Helinski, D. R.(1992). A region of the broad-host-range plasmidRK2 causes stable in planta inheritance of plasmidsin Rhizobium meliloti cells isolated from alfalfa rootnodules. Journal of Bacteriology, 174, 7486-7489.

White, T. J. & Gonzolas, C. F. (1991). Application ofelectroporation for efficient transformation ofXanthomonas campestris pv. oryzae. Phytopathology,81, 521-524.

Williams, M. V., Klein, S. & Signer, E. R. (1989).

Host restriction and transduction in R. meliloti. Appliedand Environmental Microbiology, 55, 3229-3230.

Wirth, R., Friesenegger, A. & Fiedler, S. (1989).Transformation of various species of Gram-negativebacteria belonging to 11 different genera byelectroporation. Molecular and General Genetics, 216,175-177.

Yun, A. C , Noti, J. D. & Szalay, A. A. (1986).Nitrogenase promoter-/acZ fusion studies ofessential nitrogen fixation genes in Bradyrhizobium1110. Journal of Bacteriology, 167, 784-791.

Zimmermann, U. (1983). Electrofusion of cells,principles and industrial potential. Trends inBiotechnology, 1, 149-155.

Zimmermann, U., Schultz, J. & Pilwat, G. (1973).Transcellular ion flow in Escherichia coli B andelectrical sizing of bacteria. Biophysics Journal, 13,1005-1013.

4Filamentous FungiGustavo H. Goldman, Marc Van Montagu and Alfredo Herrera-Estrella

Introduction

Fungi are eukaryotic, heterotrophic organisms withan absorptive mode of nutrition. Most fungi areboth unicellular and multinucleate, with rigid chi-tinous cell walls, and usually exhibit mycelial oryeast-like growth habit. Probably most of thebiotechnologically important soil fungi have no sta-ble recombination cycles in the laboratory. Thiscreates problems for physiological and geneticstudies in these species. Genetic manipulationusing transformation and gene cloning provides themost logically directed approach to dissect and,eventually, alter the physiology of these filamen-tous fungi. Towards this end, there has been strongpressure to develop techniques of basic molecularbiology suitable for these organisms. The mostintensively studied fungi are the unicellular yeastSaccharomyces cerevisiae and the filamentous fungiNeurospora crassa and Aspergillus nidulans. The mol-ecular genetic systems of these organisms haveserved as the basis for development of similar sys-tems in less tractable but economically importantfungal species (Timberlake & Marshall, 1989).

The first report of transformation of a fungalspecies mediated by DNA was published byMishra & Tatum (1973). Growing cultures of aninositol-requiring mutant of N. crassa were trans-formed with total DNA of the wild-type togetherwith calcium; from the conidia formed on suchcultures it was possible to select prototrophicstrains. This pioneering experiment was receivedwith skepticism and only some years later, Hinnen,Hicks & Fink (1978) reported transformation of S.cerevisiae using protoplasts from a Ieu2 mutant bytreatment with wild-type DNA in the presence of

calcium chloride. The utilization of protoplastswas immediately applied to the filamentous fungiN. crassa and A. nidulans (Case et al.> 1979;Tilburn et al.> 1983). With the passing years, thesetechniques of transformation have been extendedto other species of soil fungi (Table 4.1).

This chapter recapitulates the main technologi-cal developments, and the possibilities offered tobiotechnology by the application of genetic trans-formation techniques to soil fungi. It is our aim toreview the approaches being adopted and, whenpossible, to use examples with several soil fungi toillustrate these techniques. Transformation of soilyeasts will not be discussed in this chapter.

Procedures

Polyethylene glycol-mediatedtransformation

Until now, this technique has been the mostexploited for fungal cells and is considered as thestandard transformation system. However, itsapplication involves delicate handling of cellsbecause of the requirement for the use of proto-plasts or osmotically sensitive cells (OSCs). Asmentioned before, protoplast technology has beenan important achievement for the elaboration oftransformation protocols. Protoplasts or OSCs areproduced by removing the cell wall of either ger-minating spores or hyphae with cell-wall-degrad-ing enzymes. The most commonly used andsuccessful product for this purpose is Novozym-234, a commercially available hydrolytic enzymemixture secreted by the filamentous fungusTrichoderma harzianum, which is used alone or in

34

Filamentous fungi 35

Table 4.1. Examples of soil fungal species in which transformation has been achieved

Species

A. NONPATHOGENIC

Antibiotic producersCephalosporium acremoniumPenicillium chrysogenum

Enzyme producersAspergillus oryzaeAspergillus nigerTrichoderma reeseiTrichoderma reeseiTrichoderma reeseiTrichoderma reeseiTrichoderma viride

Biocontrol agentsGliocladium roseumGliocladium viridensMetarhizium anisopliaeTrichoderma harzianum

Trichoderma viride

Food processing (starter)Penicillium nalgiovense

MycorrhizaLaccaria laccata

B. PHYTOPATHOGENIC

Cochliobolus heterostrophusColletotrichum graminicolaColletotrichum lindemuthianumColletotrichum trifolii

Fusarium oxysporiumFusarium sambucinumFusarium sporotrichiodesFulvia fulvaGaeumannomyces graminisLeptosphaeriaMagnaporthe grisea (Pyricularia oryzae)Nectria haematococca (Fusarium solonii.sp. pisi)

Phytophthora capsiciPhytophthora parasiticaSeptoria nodorumUstilago hordei

Ustilago maydisUstilago nigraUstilago violacea

Markers

hph, niaDsul\

pyrGargBamdS, argBura3, urabpyrAhphpyrA

hphhphbenA3hph, bml

hph, tub!

amdS

hph

amdS, hphbmIR, bmlR3hph, amdShph, benomylresistance

hphhphhphhphbenomyl resistancehphargBhph, argB

hphhphhphhph

hphhphhph

Referene

Skatrud etal. (1987); Whitehead etal. (1990)Carramolino etal. (1989)

Mattern etal. (1987)Buxton, Gwynne & Davis (1985)Penttila etal. (1987)Berges & Barreau (1991)Smith etal. (1991)A. Herrera-Estrella (unpublished data)Cheng, Tsukagoshi & Udaka (1990)

Thomas & Kenerly (1989)Thomas & Kenerly (1989)Goettel etal. (1990)Goldman etal. (1992); Herrera-Estrella etal.(1990); Uhloa, Vainstein & Pederby (1992)

Herrera-Estrella etal. (1990); Goldman etal. (1993)

Geisen & Leistner (1989)

Barrett, Dixon & Lemke (1990)

Turgeon, Garber & Yoder (1985, 1987)Panaccione, McKiernan & Hanau (1988)Rodriguez & Yoder (1987)Dickman (1988)

Kistler& Benny (1988)Salch & Beremand (1988)Turgeon etal. (1987)Oliver et al. (1987)Henson, Blake & Pilgerarn (1988)Farman & Oliver (1987); Turgeon etal. (1987)Parsons, Chumley & Valent (1987)Dickman & Kolattukudy (1987); Turgeon st al.(1987); Rambosek & Leach (1987)

Bailey, Mena & Herrera-Estrella (1991)Bailey et al. (1991)Cooley etal. (1988)Holden, Wang & Leong (1988); Yoder & Turgeon(1985)

Wang etal. (1988); Yoder & Turgeon, (1985)Holden etal. (1988)Bej & Perlin (1988)

combination with other lytic enzymes such as(3-glucuronidase and chitosanase. All protoplastpreparations have to be protected by the presenceof an osmotic stabilizer at all times. Although sor-bitol, at concentrations between 0.8 and 1.2M,

has been commonly used and seems to be satisfac-tory for all fungi, mannitol, sodium chloride, andmagnesium sulfate are used for some fungi(Fincham, 1989). Mycelia or germinative tubesare usually selected as the starting material for the

36 GOLDMAN et at.

production of protoplasts. When cell walls havebeen partially or totally removed, cells are treatedwith a mixture of calcium chloride, polyethyleneglycol (PEG), and transforming DNA. Theexogenous DNA molecules are apparently inter-nalized while a PEG-induced protoplast fusiontakes place (no transformation occurs when PEGis omitted; Timberlake & Marshall, 1989). Afterthis treatment, the protoplasts are plated onappropriate regeneration medium that will allowselection for the expression of the phenotypeencoded by the transforming DNA (see selectionof transformants, below). While this phenomenonis always forced to occur by means of the combi-nation of calcium chloride and PEG, the concen-tration of these chemicals, the type of PEG andthe time during which the cells are exposed varyaccording to the cells to be transformed (Ballance,Buxton & Turner, 1983; Tilburn et al, 1983).

The two major problems faced when one is try-ing to apply this technique are the efficiency withwhich protoplasts are formed and their regenera-tion, although special attention should also bepaid to the time of recovery allowed to the cellsbefore selection.

Alternative methods: lithium acetate,electroporation and particle bombardment

Because PEG-mediated transformation is timeconsuming and some fungi are not amenable toprotoplast preparation and/or regeneration, alter-native transformation methods have been devel-oped. One of the alternative methods that avoidsprotoplast preparation is the use of high concen-trations of lithium acetate in combination withPEG to induce permeability of intact cells toDNA. Initially developed for S. cerevisiae (Ito etal., 1983), this method has been successfully usedfor N. crassa (Dhawale, Paietta & Marzluf, 1984)and Coprinus lagopus (Binninger et al., 1987). Inthe filamentous fungi, germinating spores wereexposed to the transforming DNA in the presenceof 0.1 M lithium acetate. It is not known howlithium acetate assists the passage of DNA intointact cells, and this method is not frequently usedfor filamentous fungi.

Recently, electroporation has become a valuabletechnique for the introduction of nucleic acids intoboth eukaryotic and prokaryotic cells (Miller,Dower & Tompkins, 1988; Forster & Neumann,1989). Intact cells, as well as cells treated with cell-

wall-degrading enzymes, are amenable to electro-poration (Fromm, Taylor & Walbot, 1985;Dower, Miller & Ragsdale, 1988; Shigekawa &Dower, 1988). When a cell is exposed to an elec-tric field, the membrane components becomepolarized and a potential difference developsacross it. If the voltage exceeds a threshold level,the membrane breaks down in localized areas andthe cell becomes permeable to exogenous mole-cules (Shigekawa & Dower, 1988). The inducedpermeability is reversible provided the magnitudeand/or duration of the electric field does notexceed a critical limit, otherwise the cell is dam-aged irreversibly (Shigekawa & Dower, 1988).There are also some reports about electroporationin filamentous fungi and the method is likely tobecome more popular, mainly due to its simplicityand reproducibility. Ward, Kodama & Wilson(1989) reported the transformation of protoplastsof A. awamori and A. niger by electroporation,obtaining transformation frequencies similar tothose obtained with PEG. Richey et al. (1989)transformed protoplasts of Fusarium solani and A.nidulans. These authors reported transformationfrequencies lower than those obtained by standardtransformation methods. Goldman, Van Montagu& Herrera-Estrella (1990) reported electropora-tion of T. harzianum using a combination of OSCsand PEG. An important factor identified in thisstudy was that the incubation time of the germina-tive tubes with the cell-wall-degrading enzymesgreatly influenced the final outcome of the trans-formation. In all of these experiments, transfor-mants could be obtained without the addition ofPEG. However, Goldman et al. (1990) obtained afour-fold higher transformation frequency when1 % (w/v) PEG was present in the medium duringthe delivery of the electric shock. Intact cells havealso been used for electroporation of filamentousfungi. Chakraborty, Patterson & Kapoor (1991)reported electroporation of germinative tubes ofN. crassa and conidia of Penicillium urticae andEdman & Kwon-Chung (1990) have obtainedsimilarly successful transformation of intact cells ofCryptococcus neoformans.

In most of these protocols, potentials between2000 and 3000 V/cm with a time constant of 10to 15 ms have been set up. In contrast with thecalcium chloride-PEG-mediated transformation,there is no standard medium established for elec-troporation (Richey et al., 1989; Ward et al., 1989;Goldman et al., 1990; Chakraborty et al., 1991).

Filamentous fungi 3 7

Thus, it is essential to take into account the fol-lowing parameters when establishing the electro-poration conditions for a new system: cell size, cellviability after shock delivery, and conductivity ofthe electroporation media.

Cells can also be transformed by literally shoot-ing them with microprojectiles carrying nucleicacids, with the consequent expression of the intro-duced genes (Klein et al, 1987; Sanford, Smith &Russell, 1993). The projectiles used can be metalparticles made of colloidal gold or tungsten coatedwith nucleic acids, or other types of hardened par-ticles such as bacteria and phages. This method iscalled particle bombardment or 'biolistic' (for bio-logical ballistics) transformation, and has alreadybeen applied to plants (Klein et al, 19SSa,b,c;Wang, Holden & Leong, 1988), mammals(Williams et al, 1991), bacteria (Shark et al,1991), and fungi (Armaleo et al, 1990; A. Bailey,personal communication). According to Armaleoet al. (1990), this technique of delivering nucleicacids into cells is still in its infancy but has alreadydemonstrated several advantages over preexistingtransformation methods: (i) its simplicity of appli-cation allows the simultaneous targeting of hun-dreds of millions of cells; (ii) it is not limited toone particular species and has been employed suc-cessfully both in systems already transformable byother means as well as in less tractable systems;and (iii) it is the only technique so far described bywhich mitochondria and chloroplasts (Fox,Sanford & McMullin, 1988; Johnston et al, 1988;Daniell et al, 1990) can be transformed. In fungi,transformation by particle bombardment has beenachieved for »S. cerevisiae, Schizosaccharomycespombe, N. crassa (Armaleo et al., 1990), Phyto-phthora capsici, P. citricola, P. cinnamomi, P. cit-rophthora (A. Bailey, personal communication),and Mucor circinelloides (F. Gutierrez, personalcommunication). This technique has been moresuccessful when self-replicating integrative vectorsare used and is specially interesting when assayingfor transient expression or one is titrating trans-acting factors. The future of this technique ispromising in filamentous fungi, but, as in the caseof electroporation, many biological and physicalparameters underlying this process need to be bet-ter defined.

Selection of transformantsMarker genes

Selection of transformants from the backgroundof nontransformed cells depends on the expres-sion of genes conferring adequate dominant selec-table phenotypes (Timberlake & Marshall, 1989).In the beginning, auxotrophic mutants weretransformed to prototrophy; in these cases, theselection of the transformants was usuallystraightforward (Fincham, 1989). As mentionedabove, the first report of transformation of a fila-mentous fungus was based on the complementa-tion of an inositol-requiring N. crassa mutantstrain with DNA from the inositol-independentparent (Mishra & Tatum, 1973). Later, Case et al.(1979) extended the types of selection for N.crassa by using the qa-2 gene encoding thecatabolic dehydroquinase from N. crassa. Sub-sequently, many different genes used as proto-trophic markers in recipient auxotrophic mutantswere used as selectable markers (for a review, seeFincham, 1989). An inconvenient consequence ofrelying on selection against auxotrophy in isolat-ing transformants is the need for the appropriateauxotrophic mutation in the recipient strain. Theisolation of certain auxotrophic mutants is some-times facilitated by a positive selection procedure.For example, mutations resulting in the loss oforotidine 5'-monophosphate carboxylase (encodedby ura3 or #yr4), which is required for uridinebiosynthesis, confer resistance to the normallyinhibitory analog 5-fluoroorotic acid (Alani, Cao& Kleckner, 1987; Diez et al., 1987; Smith,Bayliss & Ward, 1991).

A more versatile system is to use a dominantgene as the selectable marker. There are manygenes already available for filamentous fungi fromdifferent origins, such as those conferring resis-tance to hygromycin B, G418, phleomycin,oligomycin, copper, the fungicide benomyl, andan A. nidulans gene, amdS, that enables the organ-ism to grow on acetamide as sole carbon andnitrogen source (Austin, Hall & Tyler, 1990;Table 4.2). The existence of these markers meansthat most fungal species, even those not prev-iously subjected to laboratory investigations, cannow be genetically modified by transformation(Timberlake & Marshall, 1989). However, theproblem still exists that many fungal species arenaturally resistant to many of the antibiotics forwhich genes conferring resistance are available.

38 GOLDMAN et al.

Table 4.2. Some dominant selectable markers used for transformation in filamentous fungi

Marker gene Origin Phenotype

hphnpt\\benA3, bmlbmlR3o//CR

amdSblesul\

E. coliBacterial Tn5N. crassa, A. nidulansC. graminicolaA. nigerA. nidulansBacterial Tn5Enterobacteria

Hygromycin B resistanceKanamycin, G418 resistanceBenomyl resistanceBenomyl resistanceOligomycin resistanceAcetamide utilizationBleomycin, phleomycin resistanceSulfonamide resistance

Cotransformation

In cases where a transforming gene cannot directlybe selected for, an option is to look for its assimila-tion along with a more readily selectable marker.Theoretically, there is a high probability that a cellthat takes up one kind of DNA will also take upanother, specially if the ratio of cotransformingDNA : transforming DNA is kept high. Accordingto Fincham (1989), this phenomenon of cotrans-formation can be rationalized by supposing thatnot all protoplasts are equally prone to take upDNA and that those more competent to do so willtend to take up several molecules simultaneously.The frequencies of reported cotransformation arehighly variable and probably dependent upon theorganism and the transformation conditions.

Genetic purification of transformants

Generally, when a protoplast is transformed, thetransformed nuclei will exist within a populationof wild-type nuclei, since most of the protoplastsare multinucleate. Selection cannot be applied toindividual nuclei because the fungal thallus iscoencytial. The initially transformed colonies arelikely to be heterokaryons, with some nuclei trans-formed and some not, or with a heterologous mix-ture of independently transformed nuclei. If afungus forms uninucleate conidia, a heterokaryoncan be resolved very simply into its componentsby plating conidia and by isolating single colonies.If the conidia are multinucleate, genetic purifica-tion can be achieved more laboriously by severalsuccessive rounds of re-isolation of the trans-formed phenotype from single conidial colonies.After three rounds the probability of stochasticloss of one or other nuclear component is high(Fincham, 1989).

Fate of transforming DNA

Autonomously replicating vectors

Although most of the plasmids used for transfor-mation in filamentous fungi have no origin ofreplication, a few autonomously replicating vec-tors have been used. These vectors are derivedfrom naturally occurring mitochondrial plasmidsor vectors with 'autonomously replicatingsequences' (ARS). Autonomously replicating vec-tors in S. cerevisiae have a higher frequency oftransformation than do integrative vectors. Thisenhanced efficiency can be exploited to simplifythe isolation and cloning of genes. The origin ofreplication from the *S. cerevisiae nuclear plasmid,the 2 \i circle, is used extensively as a componentin yeast episomal vectors (Futcher, 1988).Unfortunately, the 2 \i replicon and yeast chromo-somal ARS do not promote autonomous replica-tion of vectors in any filamentous fungus so fartested. Therefore, sequences from chromosomalDNA and from endogenous fungal plasmids havebeen screened in attempts to identify sequencesthat promote autonomous replication of vectors.

Most fungal plasmids are found within mito-chondria, but, in some yeast species, plasmids arefound within nuclei. Circular plasmids appear tobe less common in fungi than linear ones and havebeen reported only in N. crassa, Cochliobolusheterostrophus and Cephalosporum acremonium(Taylor, Smolich & May, 1985; Samac & Leong,1989). Plasmids with diverse characteristics havebeen described in a wide range of fungal species.Unfortunately, there are only few examples of par-tial success of incorporation of mitochondrialreplicons into transformation vectors (Stahl et al.,1982; Esser et al., 1983; Stohl & Lambowitz,1983; Kuiper & de Vries, 1985). The replicons of


mitochondrial plasmids could be useful in con-structing autonomously replicating vectors for fil-amentous fungi because mitochondrial plasmidsare maintained at high copy number (Samac &Leong, 1989).

A more traditional approach to vector buildinghas been the incorporation of ARS. Thesesequences allow autonomous replication of thehybrid plasmid in transformants and, thus, enhancethe efficiency of transformation (Mishra, 1985).Integrated sequences are stable but replicatingsequences can be lost because they lack the cen-tromeres and telomere size needed to behave liketrue chromosomes. Buxton & Radford (1984)screened a library of 700 cloned N. crassa sequencesin the size range 1 to 7 kb for the ability to improvethe N. crassa transformation frequency wheninserted into a plasmid that also included pyr4+ orqa-2+ as a selective marker. Only four were found tohave significant effect. In an attempt to isolate anAspergillus origin of replication, Ballance & Turner(1985) utilized the fact that the N. crassa pyr4 geneis weakly expressed in yeast. Fragments of A. nidu-lans genomic DNA were cloned into a plasmid con-taining the N. crassa pyr4 gene and this library wasused to transform a yeast strain. Transformants thatwere mitotically unstable, i.e. lost the plasmidunder nonselective conditions, were subsequentlyidentified on the assumption that these containedautonomously replicating plasmids. These replicat-ing plasmids, complete with the A. nidulans DNAfragment acting as ARS, were re-isolated from yeastand used to transform A. nidulans. One plasmidderived in this manner was capable of transformingA. nidulans at frequencies 50- to 100-fold higherthan the original plasmid, yielding up to 5 X 103

transformants/Mg DNA (Ward, 1991). The A. nidu-lans sequence that conferred this high frequency oftransformation was termed ARS1. Further investi-gation of the function of ARS 1 has led to the con-clusion that it does not confer autonomousreplication in A. nidulans, although the unprovenpossibility exists that autonomous replicationoccurs for a brief period immediately after transfor-mation and is followed by stabilization due to inte-gration. Mapping the chromosomal location ofARS1 has shown it to be tightly linked to the cen-tromere of linkage group I (Cullen et al, 1987). Ithas since been confirmed that many sequencesfrom heterologous species that act as ARS in yeastare not necessarily ARS or origin of replication intheir native host (Maundrell et al., 1985).

Van Heeswijck (1986) and Roncero et al.(1989) presented evidence for the autonomousreplication of recombinant plasmids in Mucorcircinelloides leu transformants. Plasmids consist-ing of a unique fragment of Mucor DNA insertedinto YRpl7 or pBR322 give a high frequency oftransformation (up to 7800 leu transformants/jigDNA), are mitotically unstable, can be re-isolatedin an unmodified form from uncut transformantDNA, and are present as discrete extrachromo-somal DNA molecules. Subcloning of the recom-binant plasmids and analysis of subsequent leutransformants show that autonomous replicationis independent of the vector sequences and locatesa Mucor ARS within a 4.4 kb PstI fragment of theinsert DNA. Recently, it has been shown that highfrequency transformation and mitotically unstabletransformants are obtained when Ustilago maydisis transformed with a vector incorporating a puta-tive ARS from nuclear DNA of U. maydis. Thevector is present at high copy number, approxi-mately 25 molecules/cell, of vector DNA(Tsukuda et al., 1988). The same vector has beensuccessfully used for transformation of Phytoph-thora parasitica and P. capsici (M. Bailey, personalcommunication).

Integration of DNA into the chromosomes

Genetic modification of most filamentous fungidepends on the genomic incorporation of exoge-nously added DNA. Therefore, most of the plas-mids used for transformation in filamentous fungido not need a fungal origin of replication.Plasmids readily integrate into either homologousor heterologous sites. Genomic integration of cir-cular plasmids occurs in several ways. Accordingto the most commonly used classification, threetypes of integration event can be defined (Figure4.1). Type I involves integration of the plasmid ata region of homology within the genome (Figure4.1 (a)) and it is usually called homologous recom-bination. In general, the relative frequency ofhomologous integration is highly variable fromone organism to another: 100% in S. cerevisiae(Struhl, 1983), about 80% in A. nidulans (Yelton,Hamer & Timberlake, 1984) and only 1% to 5%in N. crassa (Case, 1986) or Coprinus cinereus(Binninger et al., 1987). Type II transformantshave the plasmid integrated into sites within thegenome where no known homology exists (Figure4.1(6)). The mechanism by which the DNA

40 GOLDMAN et al

(a)

TYPE ILEU 2

leu 2leu 2

Single copy integration at leu 2 locus• * OR

LEU2 LEU2 Ieu2

Tandem integration at LEU 2 locus

LEU 2

Ieu2Native LEU 2 locus unaffected

PLUSLEU2

Single copy integration not associatedwith leu 2 locus

ORLEU 2 LEU 2

Tandem integration not associated withleu 2 locus

LEU 2

Gene conversion or doublecross-over to givegene replacement

Figure 4.1. Patterns of plasmid integration in fungaltransformants. (a) Type I involves integration of the plasmidat a region of homology within the genome and is usuallycalled homologous recombination, (b) Type II transformantshave the plasmid integrated into sites within the genomewhere no known homology exists (heterologous or ectopicrecombination), (c) Type III, or a gene conversion event, noplasmid sequence would be detected within the genomebut the gene (for example, leu) would apparently have beenreplaced by the introduced copy.

recombines in type II (heterologous recombina-tion) has not been determined but there is pre-sumably a requirement for short stretches offortuitous sequence identity at the sites of appar-ently nonhomologous integration (Ward, 1991).In type III or a gene conversion event, no plasmidsequence would be detected within the genome,but the gene (leu in the example) would appar-ently have been replaced by the introduced copy(Figure 4.1(c)). Multiple plasmid copies inte-grated in tandem are a common feature of trans-formants. There are two obvious mechanismsthrough which this might come about: either

extrachromosomal plasmids first undergo homol-ogous recombination with each other to form cir-cular oligomers, which could then integrate byhomology with the chromosomal, single copy; or amonomeric plasmid recombines with its chromo-somal homolog and then tandem repeats arisethrough secondary integration (Fincham, 1989).

The use of linearized DNA has been shown togive higher efficiencies of integrative transforma-tion in 5. cerevisiae when compared to circularDNA (Suzuki et al, 1983). In contrast, lineariza-tion of vector DNA does not seem to have anyeffect when transforming N. crassa (Huiet & Case,1985) and it results in lower efficiencies whenused for transformation in Trichoderma species(Herrera-Estrella, Goldman & Van Montagu,1990). However, the use of linear moleculesappears to increase significantly homologousrecombination (Boylan et al., 1987; Aramayo,Adams & Timberlake, 1989). Genomic integra-tion of linear molecules often occurs by a processequivalent to a double cross-over event (Kinsey &Rambosek, 1984; Miller, Miller & Timberlake,1985). However, circularization of linear mole-cules prior to integration can produce tandemduplications (Miller et al., 1985).

Although integrated plasmids are mitoticallystable, introduced DNA sequences are oftenmeiotically unstable. In A. nidulans, tandemlyrepeated sequences are lost at variable, but readilydetectable, frequencies after self-fertilization orout-crossing (Tilburn et al., 1983). In N. crassa,duplicated sequences are eliminated at high fre-quency during the second phase by a processcalled 'repeat-induced point mutation' (RIP)(Selker et al., 1987; Selker & Garret, 1988).Considering that most of the soil nonpathogenicfungi lack a sexual phase, the behavior of insertedDNA sequences in meiosis is irrelevant.

Use of transformation for analysis ofgene function

Self-cloning in filamentous fungi

Although there are several reports on cloning offilamentous fungal genes in Escherichia coli and*S. cerevisiae (Ballance, 1986; for a short review,see Schrank et al, 1991; Goldman et al, 1992),many fungal genes will not be expressed in thesespecies and so cannot be isolated by interspecific


complementation in these hosts. Additionally, anycloning method that involves isolation of the spe-cific mRNA species depends on its abundance inthe total mRNA population. This is not the casefor many genes of interest, especially regulatorygenes. Theoretically, the establishment of a trans-formation system could allow that fungal genescould be cloned by complementation of muta-tions, i.e. self-cloning. So, a genomic library of aparticular fungal species would be constructedand used to transform the strain bearing a particu-lar mutation. Transformants would be selected inwhich the mutation had been complemented andthe plasmid bearing the sequence responsible forthe complementation would be isolated again.According to Ward (1991), this type of self-cloning requires two features: (i) transformationfrequencies must be high enough so that a genebank can be screened using realistic quantities ofDNA and protoplasts, and (ii) it must be possibleto re-isolate or rescue any complementing plas-mid. The lack of satisfactory shuttle vectors makesit somewhat more difficult to clone genes by com-plementation in filamentous fungi, but in somecases intact bacterial plasmids can be recoveredfrom transformants. In cases in which there is nodetectable free plasmid remaining, a transformingsequence can sometimes still be recovered bycleaving the transformant DNA with a restrictionenzyme that cuts once, but no more than once,within the sequence duplicated through type Iintegration (see p. 39; Fincham, 1989). The frag-ments generated in this way are circularized withligase and the reconstituted plasmid is selected bytransformation of E. coli (Yelton et ah, 1984).Yelton, Timberlake & van den Hondel (1985)developed a more efficient way of cloning comple-menting A. nidulans genes by constructing a cos-mid library - a plasmid with bacteriophage X cossequences - including trpC+ for selection in A.nidulans, ampicillin and chloramphenicol resis-tance genes for selection in E. coli, and a BarriHIcloning site that would accept fragments cleavedwith the 'tetra-cutting' endonuclease Mbol orSau3Al. A cosmid library carrying A. nidulansgenomic fragments of 35 to 40 kb, was used totransform a trpCyA (yellow spored) double-mutant strain. DNA isolated from the rrp+, green-spored selected transformants was packaged invitro and the cosmids recovered by infection of asuitable E. coli strain, where they could be selectedby ampicillin resistance. The identity of these

clones was then confirmed by a new transforma-tion experiment with the rescued E. coli plasmids.

Cosmid rescue has also been used to clonegenes from one species by detecting their expres-sion in a different species. For example, a disease-determinant gene, pisatin-demethylating ability(PDA) from the phytopathogen Nectria haemato-cocca was cloned by detecting its expression in A.nidulans (Weltring et ah, 1988). It should be possi-ble to use identical or related strategies to isolategenes of interest from any other fungal species. Arecent example of this technology has been pro-vided by the cellulolytic fungus T. reesei. Barreauet al. (1991) cloned an A. niger gene for invertaseby transformation of the QM9A14 Ura5" strain ofthis species. First, the authors developed a veryefficient transformation system for this speciesusing pyrimidine auxotrophic ura3 (pyrty andura5 mutants (Berges & Barreau, 1991). Then, asibling (sib) selection (see below) procedure wascarried out in order to clone the structural gene ofan A. niger invertase by direct expression in ura5+

transformants. Two cosmid clones were obtainedand, using oligonucleotides derived from the N-terminus and internal tryptic peptides from thepurified protein, one of the clones was identifiedas containing the gene for the secreted form of A.niger invertase.

A widely used method for gene identification inN. crassa is named sib selection (Akins &Lambowitz, 1985). Sib selection refers to the con-struction of an ordered gene bank in which E. coliclones, each containing a plasmid with a differentN. crassa DNA fragment, are maintained sepa-rately. Transformation is originally by DNAextracted from pools of large numbers of theseclones and subsequently by smaller and smallersubpools until the individual clone containingDNA capable of complementation of the N. crassamutation of interest is identified.

An important point about cloning by comple-mentation is the need to confirm that the clonedgene obtained is not really some other sequencethat acts as a suppressor of the mutation beingcomplemented. Initial tests to rule out this possi-bility are to show that the cloned gene will com-plement several different allelic mutations and,if available, deletions of that gene, or to showby several crossings that the cloned gene mapsat the expected locus (provided the gene hasbeen mapped in the first place) (Ward, 1991). Inorganisms that do not have a sexual cycle, these

42 GOLDMAN et ah

tests are not available and then, of course, con-firmation of whether or not the cloned gene is asuppressor becomes more difficult.

Gene disruption and replacement

Homologous integration of transforming plasmidsopens up the possibility of gene disruption andgene replacement techniques. It often happensthat a cloned DNA sequence looks like a func-tional gene in that it is transcribed, contains anopen reading frame, and perhaps has some inter-esting similarities to known genes in other organ-isms, but it cannot be assigned a function becauseno mutations have been identified in it. In suchcases, the first step in understanding the gene is touse the clone to disrupt the equivalent sequence inthe genome, thus creating a null mutant. In genereplacement, as opposed to gene disruption, thepurpose is to retain gene activity but modify itsproduct or its mode of transcription.

The most common methodology for gene dis-ruption in filamentous fungi is the procedurecalled one-step disruption, originally described byRothstein (1983) for 5. cerevisiae. This procedureconsists of inserting a copy of a selectable marker(e.g. hph> encoding resistance to the antibiotichygromycin) into the cloned gene under investiga-tion. This construction is then used to transformand convert a hygromycin-sensitive strain into aresistant one. There are many examples of suc-cessful gene disruption in N. crassa and A. nidu-lans (Miller et al., 1985; Paietta & Marzluf, 1985).Hoskins et al. (1990) have carried out a gene dis-ruption analysis in the antibiotic producer C. acre-monium. In this case, the one-step disruptionprocedure using hph as a marker was inserted intothe genomic region immediately upstream frompcbC (the gene coding for isopenicillin syn-thetase). Approximately 4% of the C. acremoniumtransformants obtained were unable to produceP-lactam antibiotics.

The first procedure for gene replacement infungi was described for 5. cerevisiae (Scherer &Davies, 1979). In this method, transformationwas carried out with a plasmid containing both amodified form of the target gene and a separateselectable marker. Transformants with the plas-mid integrated by homology into the target genehad tandemly arranged copies of both the targetgene and the modified version that was to replaceit, with the rest of the plasmid including the selec-

tive marker between them. After screening forsubclones (after ten cycles of budding) that hadlost the marker by crossing over between tandemgene copies, there was either restoration of thenatural gene or its replacement, depending onwhether the recombinant event occurred to theright or to the left of the sequence distinguishingthe natural and modified gene copies (Scherer &Davies, 1979; Fincham, 1989). Miller et al(1985), working with A. nidulans, replaced thespoCl C gene of the spoCl (sporulation-specific)gene cluster with a partly deleted derivative fol-lowing essentially the same method describedabove. A plasmid carrying the modified spoClsequence together with trpC+ as a selectablemarker was used to transform a trpC mutantstrain. A transformant with trpC+ integrated bycrossing-over within spo tended to lose trpC+ dur-ing vegetative growth by further crossing-overbetween the flanking spo sequences.

Titration of trans-acting gene products

The introduction by transformation of multiplecopies of a as-acting sequence that binds to theprotein product of a trans-acting regulatory genecan give valuable information about the functionof that gene (Fincham, 1989). Andrianopoulos &Hynes (1988) showed that multiple copies ofsome as-acting regulatory sequences titrate awaytheir corresponding rrans-acting transcriptionalregulators, leading to an inability to induce genesin the regulon. Titrations are not expected to beuseful in those instances where there is a largeexcess of trans-acting factor or where the trans-active regulatory gene is autogenously controlled(Timberlake & Marshall, 1989).

Applications to biotechnology

One of the most obvious applications of transfor-mation is the development of gene expression sys-tems for filamentous fungi. These organisms arepotentially attractive as host systems for heterolo-gous gene expression because of their high secre-tory capacity. Many species of filamentous fungi,including Aspergillus, Trichoderma, Achlya, Mucor,Penicillium and Cephalosporium have been in com-mercial use for decades. Transformation is nowbeing used to improve existing fungal strains byproviding them with the genes to produce


enzymes or antibodies. The introduction of 'newphenotypes' or overexpression of existing ones in agiven species could theoretically improve the bio-logical process carried out by it. Cross-speciesexpression can also be exploited to study a num-ber of biological and biochemical problemsincluding structure/function relationships of aparticular gene product in an isozyme-free back-ground, evolutionary relationships between func-tionally equivalent genes, elements that controltranscription, translation and post-translationalmodification of a gene product, and pathway engi-neering in fungi (Fowler & Berka, 1991).

Cross-species gene expression can be used to'engineer' biosynthetic pathways by introducingcloned genes from one organism into another. Forexample, the successful cloning of a penicillinbiosynthetic gene cluster from Penicillium chryso-genum has allowed the introduction of at leastthree linked genes that are necessary for penicillinbiosynthesis into fungi that do not normally makepenicillin (Smith et al.> 1990; Fowler & Berka,1991). Increasing the copy number of a fungalgene through transformation usually leads toincreased gene expression (Fowler & Berka,1991). Recent examples of this concept includethe expression of the genes for glucose oxidase(goxA) (Whittington et al, 1990), glucoamylase(glaA) (Fowler, Berka & Ward, 1990), andprepro-polygalacturonidase II (pgall) (Bussink,Kester & Visser, 1990) in A. niger as well as(3-glucosidase (bglL) gene expression in T. reesei(Barnett, Berka & Fowler, 1991).

An intensive effort for increasing protein secre-tion is being carried out in the cellulolytic pro-ducer Trichoderma reesei (for a review, seeNevalainen et aL, 1991). This species is wellknown for its ability to produce large quantities ofcellulase. Under the appropriate conditions, pro-duction strains secrete into the culture mediumwell over 50% of all the protein produced, whichrepresents over 40 g/1 (Knowles et al., 1989).Another very attractive feature present in T. reeseiis a protein glycosylation system that appears tomodify extracellular proteins in a manner verysimilar to that found in mammalian cells(Salovouri et al., 1987). Two interesting projectsemploying this species are increased cellulase pro-duction using genetic engineering and the use ofstrong promoters of the cellulase gene for heterol-ogous gene expression. The former is being doneby using cloned cellulase genes and strains with

completely different cellulase profiles (Nevalainenet al.> 1991). In this sense, transformation withindividual cellulase genes would be useful and asimple tool to alter the quantitative pattern of cel-lulolytic enzymes produced by T. reesei. Kubicek-Pranz, Gruber & Kubicek (1991) have shownthat, on increasing the copy number of one of theenzymes of the complex, cellobiohydrolase II,transformants exhibited an increased specificactivity against crystalline cellulose in vitro.

To date, there is just one report of heterologousgene expression in T. reesei. Harkki et al. (1989)studied the ability of T. reesei to express andsecrete bovine chymosin. They inserted chymosincomplementary DNA (cDNA) sequences intoexpression units that included the promoter andterminator regions of the highly expressed cello-biohydrolase I (cbhl) gene. Several expressionunits were constructed that employed differentconfigurations of cbhl and chymosin signal pep-tides and they were introduced into T. reesei. Inthe resulting transformants, more than 90% of thechymosin that was synthesized was extracellularand yields of 40 mg/1 were produced by some iso-lates. The chymosin produced by T. reesei wasprocessed to an active form (Berka & Barnett,1989).

Aspergillus niger and A. awamori also have theability to secrete copious amounts of proteins insubmerged culture. Since these strains are widelyregarded as safe for the production of food-gradeenzymes, harnessing even a portion of this capa-city for the production of high value enzymes orpharmaceutical proteins could provide an eco-nomically significant advantage over more expen-sive approaches such as mammalian cell cultures.Extracellular yields of glucoamylase from A. nigercould be improved by increasing the gene copynumber (Berka & Barnett, 1989). Multiple copiesof the A. niger glucoamylase (glaA) gene wereintroduced into both A. niger and A. awamori.Multiple, integrated glaA gene copies were foundto be arranged in tandem repeats that were stablein the absence of selective pressure. Transfor-mants that contained multiple copies of the glaAgene overproduced ^/aA-specific mRNA, result-ing in increased enzyme synthesis. In a similarapproach, A. niger was used for the production ofhen egg-white lysozyme, which was correctly pro-cessed and folded, as shown by two-dimensionalH-NMR. In this case, lysozyme was routinely pro-duced at 10 mg/ml (Archer et al., 1990). Also in

44 GOLDMAN et al.

1990, Ward et al. selected a strain of A. awamorifor the production of a glucoamylase-chymosin fusion protein. In their work, theyreported the enzyme generated as being autocat-alytically released from the fusion and active.

Fungi, however, are the most important class ofplant pathogens and also affect plant productivityin positive ways. For example, symbiotic mycor-rhizae increase the ability of plant roots to obtainlimiting nutrients. Genetic engineering provides anew opportunity to study mechanisms controllingsymbiosis and pathogenesis by allowing the identi-fication of symbiosis- and pathogenicity-relatedgenes.

Kolattukudy and colleagues provided an exam-ple of an enzyme as a pathogenicity determinant.In their work, they introduced the Fusarium solanipisi cutinase gene into Mycosphaerella species, aparasitic fungus that affects papaya fruits only ifthe fruit skin is mechanically breached beforeinoculation. The transformants obtained of thewound-requiring fungus were then capable ofinfecting intact papaya fruits (Dickman, Podilia &Kolattukudy, 1989).

Nectria haematococca, a fungal pathogen of pea,carries genes that encode pisatin demethylase(PDA), a cytochrome P-450 monooxygenasethat detoxifies the phytoalexin pisatin. BecausePDA is required by N. haematococca for patho-genicity on pea, pisatin helps to defend peaagainst N. haematococca. The possibility thatpisatin is a general defense factor (i.e. PDA canconfer pathogenicity to fungi not normally path-ogenic on pea) was investigated by Schafer et al.(1989). Genes encoding PDA were transformedinto and highly expressed in Cochliobolus hetero-strophus (a fungal pathogen of maize but not ofpea) and in A. nidulans (a saprophytic fungus),neither of which produces a significant amountof PDA. Recombinant C. heterostrophus was nor-mally virulent on maize, but it also caused symp-toms on pea, whereas recombinant A. nidulansdid not affect pea.

In spite of all recent efforts to develop new tech-niques and approaches to study soil fungi, it isclear that, compared to the sophisticated molecu-lar biological tools available for well-characterizedorganisms such as E. coli and S. cerevisiae, thedevelopment of gene expression systems for fila-mentous fungi is still at an early stage. Neverthe-less, the potential advantages and commercialapplications for expression of both homologous

and heterologous genes continues to provide theimpetus for intense research efforts.

Acknowledgements

The authors are grateful to Drs Allan Caplan,Dominique Van Der Straeten and June Simpsonfor critical reading of the manuscript, MartineDe Cock for typing it, Karel Spruyt, VeraVermaercke, and Stefaan Van Gijsegem for draw-ings and photographs.

References

Akins, R. A. & Lambowitz, A. M. (1985). Generalmethod for cloning Neurospora crassa nuclear genesby complementation of mutants. Molecular andCellular Biology, 5, 2272-2278.

Alani, E., Cao, L. & Kleckner, N. (1987). A methodfor gene disruption that allows repeated use ofURA3 selection in the construction of multiplydisrupted yeast strains. Genetics, 116, 541-545.

Andrianopoulos, A. & Hynes, M. J. (1988). Cloningand analysis of the positively acting regulatory geneamdR from Aspergillus nidulans. Molecular andCellular Biology, 8, 3532-3541.

Aramayo, R., Adams, T. H. & Timberlake, W. E.(1989). A large cluster of highly expressed genes isdispensable for growth and development inAspergillus nidulans. Genetics, 122, 65-71.

Archer, D. B., Jeense, D. J., MacKenzie, D. A.,Brightwell, G , Lambert, N., Lowe, G., Radford,S. E. & Dobson, C. M. (1990). Hen egg whitelysozyme expressed in, and secreted from, Aspergillusniger is correctly processed and folded.Bio/Technology, 8, 741-745.

Armaleo, D., Ye, G.-N., Klein, T. M., Shark, K. B.,Sanford, J. C. & Johnston, S. A. (1990). Biolisticnuclear transformation of Saccharomyces cerevisiaeand other fungi. Current Genetics, 17, 97-103.

Austin, B., Hall, R. M. & Tyler, B. M. (1990).Optimized vectors and selection for transformationof Neurospora crassa and Aspergillus nidulans tobleomycin and phleomycin resistance. Gene, 93,157-162.

Bailey, A. M., Mena, G. L. & Herrera-Estrella, L.(1991). Genetic transformation of the plantpathogens Phytophthora capsici and Phytophthoraparasitica. Nucleic Acids Research, 19, 4273-4278.

Ballance, D. J. (1986). Sequences important for geneexpression in filamentous fungi. Yeast, 2, 229-236.

Ballance, J., Buxton, F. P. & Turner, G (1983).Transformation of Aspergillus nidulans by theorotidine-5'-phosphate decarboxylase gene of


Neurospora crass a. Biochemical and BiophysicalResearch Communications, 112, 284-289.

Ballance, D. J. & Turner, G. (1985). Development ofa high-frequency transforming vector for Aspergillusnidulans. Gene, 36, 321-331.

Barnett, C. C , Berka, R. M. & Fowler, T. (1991).Cloning and amplification of the gene encoding anextracellular p-glucosidase from Trichoderma reesei:evidence for improved rates of saccharinification ofcellulosic substrates. Bio/Technology, 9, 562-567.

Barreau, C , Boddy, L. M., Peberdy, J. F. & Berges,T. (1991). Direct cloning of an Aspergillus nigerinvertase gene by transformation of the QM9414ura5" strain of the cellulase producer Trichodermareesei. EMBO Workshop on Molecular Biology ofFilamentous Fungi, Berlin (FRG), 24-29 August,1991, Abstract A3.

Barrett, V., Dixon, R. K. & Lemke, P. A. (1990).Genetic transformation of a mycorrhizal fungus.Applied Microbiology and Biotechnology, 33, 313-316.

Bej, A. K. & Perlin, M. (1988). Apparenttransformation and maintenance in basidiomycetemitochondria of a plasmid bearing the hygromycin Bgene. Genome, 19 (Suppl. 1), 300.

Berges, T. & Barreau, C. (1991). Isolation of uridineauxotrophs from Trichoderma reesei and efficienttransformation with the cloned ura3 and ura5 genes.Current Genetics, 19, 359-365.

Berka, R. M. & Barnett, C. C. (1989). Thedevelopment of gene expression systems forfilamentous fungi. Biotechology Advances, 7, 127-154.

Binninger, D. M., Skrzynia, C , Pukkila, P. J. &Casselton, L. A. (1987). DNA-mediatedtransformation of the basidiomycete Coprinuscinereus. EMBO Journal, 6, 835-840.

Boylan, M. T., Mirabito, P. M., Willett, C. E.,Zimmerman, C. R. & Timberlake, W. E. (1987).Isolation and physical characterization of threeessential conidiation genes from Aspergillus nidulans.Molecular and Cellular Biology, 7, 3113-3118.

Bussink, H. J. D., Kester, H. C. M. & Visser, J.(1990). Molecular cloning, nucleotide sequence andexpression of the gene encoding prepro-polygalacturonidase II of Aspergillus niger. FEBSUtters, 273, 127-130.

Buxton, F. P., Gwynne, D. I. & Davies, R. W. (1985).Transformation of Aspergillus niger using the argBgene of Aspergillus nidulans. Gene, 37, 207-214.

Buxton, F. P. & Radford, A. (1984). Thetransformation of mycelial spheroplasts ofNeurospora crassa and the attempted isolation of anautonomous replicator. Molecular and GeneralGenetics, 196, 339-344.

Carramolino, L., Lozano, M., Perez-Aranda, A.,Rubio, V. & Sanchez, F. (1989). Transformation ofPenicillium chrysogenum to sulfonamide resistance.Gene, 11, 31-38.

Case, M. E. (1986). Genetical and molecular analysesof qa-2 transformants in Neurospora crassa. Genetics,113, 569-587.

Case, M. E., Schweizer, M., Kushner, S. R. & Giles,N. H. (1979). Efficient transformation of Neurosporacrassa by utilizing hybrid plasmid DNA. Proceedingsof the National Academy of Sciences, USA, 76,5259-5263.

Chakraborty, B. N., Patterson, N. A. & Kapoor, M.(1991). An electroporation-based system for highefficiency transformation of germinated conidia offilamentous fungi. Canadian Journal of Microbiology,37, 858-863.

Cheng, C , Tsukagoshi, N. & Udaka, S. (1990).Transformation of Trichoderma viride using theNeurospora crassa pyr\ gene and its use in theexpression of a Taka-amylase A gene fromAspergillus oryzae. Current Genetics, 18, 453-456.

Cooley, R. N., Shaw, R. K., Franklin, F. C. H. &Caten, C. E. (1988). Transformation of thephytopathogenic fungus Septoria nodorum tohygromycin B resistance. Current Genetics, 13,383-389.

Cullen, D., Wilson, L. J., Grey, G. L., Henner, D. J.,Turner, G. & Ballance, D. J. (1987). Sequence andcentromere proximal location of a transformationenhancing fragment ansl from Aspergillus nidulans.Nucleic Acids Research, 15, 9163-9175.

Daniell, H., Vivekananda, J., Nielsen, B. L., Ye, G. N.& Sanford, J. C. (1990). Transient foreign geneexpression in chloroplasts of cultured tobacco cellsafter biolistic delivery of chloroplast vectors.Proceedings of the National Academy of Sciences, USA,87, 88-92.

Dhawale, S. S., Paietta, J. V. & Marzluf, G. A. (1984).A new, rapid and efficient transformation procedurefor Neurospora. Current Genetics, 8, 77-79.

Dickman, M. B. (1988). Whole cell transformation ofthe alfalfa fungal pathogen Colletotrichum trifolii.Current Genetics, 14, 241-246.

Dickman, M. B. & Kolattukudy, P. E. (1987).Transformation oiFusarium solani f. sp. pisi usingthe cutinase promoter. Phytopathology, 11, 1740.

Dickman, M. B., Podilia, G. K. & Kolattukudy, P. E.(1989). Insertion of cutinase gene into a woundpathogen enables it to infect intact host. Nature{London), 342, 446-448.

Diez, B., Alvarez, E., Cantoral, J. M., Barredo, J. L. &Martin, J. F. (1987). Selection and characterizationof pyrG mutants of Penicillium chrysogenum lackingorotidine-5'-phosphate decarboxylase andcomplementation of the#yr4 gene of Neurosporacrassa. Current Genetics, 12, 277-282.

Dower, W. J., Miller, J. F. & Ragsdale, C. W. (1988).High efficiency transformation of E. coli by highvoltage electroporation. Nucleic Acids Research, 16,6127-6145.

46 GOLDMAN et al.

Edman, J. C. & Kwon-Chung, K. J. (1990). Isolationof the URA5 gene from Cryptococcus neoformans var.neoformans and its use as a selective marker fortransformation. Molecular and Cellular Biology, 10,4538-4544.

Esser, K., Kuck, U., Stahl, U. & Tudzynski, P.(1983). Cloning vectors of mitochondrial origin foreukaryotes: a new concept in genetic engineering.Current Genetics, 13, 327-330.

Farman, M. L. & Oliver, R. P. (1988). Thetransformation of protoplasts of Leptosphaeriamaculans to hygromycin B resistance. CurrentGenetics, 13, 327-330.

Fincham, J. R. S. (1989). Transformation in fungi.Microbiology Reviews, 53, 148-170.

Forster, W. & Neumann, E. (1989). Gene transfer byelectroporation. A practical guide. In Electroporationand Electrofusion in Cell Biology, ed. E. Neumann,A. E. Sowers & C. A. Jordan, pp. 299-318. PlenumPress, New York.

Fowler, T. & Berka, R. M. (1991). Gene expressionsystems for filamentous fungi. Current Opinion inBiotechnology, 2, 691-697.

Fowler, T., Berka, R. M. & Ward, M. (1990).Regulation of the glaA gene of Aspergillus niger.Current Genetics, 18, 537-545.

Fox, T. D., Sanford, J. C. & McMullin, T. W. (1988).Plasmids can stably transform yeast mitochondrialacking endogenous mtDNA. Proceedings of theNational Academy of Sciences, USA, 85, 7288-7292.

Fromm, M., Taylor, L. P. & Walbot, V. (1985).Expression of genes transferred into monocot anddicot plant cells by electroporation. Proceedings of theNational Academy of Sciences, USA, 82, 5824-5828.

Futcher, A. B. (1988). The 2|x circle plasmid ofSaccharomyces cerevisiae. Yeast, 4, 27-40.

Geisen, R. & Leistner, L. (1989). Transformation ofPenicillium nalgiovense with the amdS gene ofAspergillus nidulans. Current Genetics, 15, 307-309.

Goettel, M. S., Leger, R. J. S., Bhairi, S., Jung, M. K.,Oakley, B. R., Roberts, D. W. & Staples, R. C.(1990). Pathogenicity and growth of Metarhiziumanisopliae stably transformed to benomyl resistance.Current Genetics, 17, 129-132.

Goldman, G. H., Demolder, J., Dewaele, S., Herrera-Estrella, A., Geremia, R. A., Van Montagu, M. &Contreras, R. (1992). Molecular cloning of theimidazole-glycerolphosphate dehydratase gene ofTrichoderma harzianum by genetic complementationin Saccharomyces cerevisiae using a direct expressionvector. Molecular and General Genetics, 234,481-488.

Goldman, G. H., Temmerman, W., Jacobs, D.,Contreras, R., Van Montagu, M. & Herrera-Estrella, A. (1993). A nucleotide substitution in oneof the fl-tubulin genes of Trichoderma viride confersresistance to the antimitotic drug methyl

benzimidazole-2-yl-carbamate. Molecular andGeneral Genetics, 240, 73-80.

Goldman, G. H., Van Montagu, M. & Herrera-Estrella, A. (1990). Transformation of Trichodermaharzianum by high-voltage electric pulse. CurrentGenetics, 17, 169-174.

Harkki, A., Uusitalo, J., Bailey, M., Penttila, M. &Knowles, J. K. C. (1989). A novel fungal expressionsystem: secretion of active calf chymosin from thefilamentous fungus Trichoderma reesei.Bio/Technology, 7, 596-603.

Henson, J. M., Blake, N. K. & Pilgeram, A. L. (1988).Transformation of Gaeumannomyces graminis tobenomyl resistance. Current Genetics, 14, 113-117.

Herrera-Estrella, A., Goldman, G. H. & VanMontagu, M. (1990). High-efficiencytransformation system for the biocontrol agents,Trichoderma spp. Molecular Microbiology, 4, 839-843.

Hinnen, A., Hicks, J. B. & Fink, G. R. (1978).Transformation of yeast. Proceedings of the NationalAcademy of Sciences, USA, 75, 1929-1933.

Holden, D. W., Wang, J. & Leong, S. A. (1988).DNA-mediated transformation of Ustilago hordei andUstilago nigra. Physiological and Molecular PlantPathology, 33, 235-239.

Hoskins, J. A., O'Callaghan, N., Queener, S. W.,Cantwell, C. A., Wood, J. S., Chen, C. J. & Skatrud,P. L. (1990). Gene disruption of xhepcbAB geneencoding ACV synthetase in Cephalosporiumacremonium. Current Genetics, 18, 523-530.

Huiet, L. & Case, M. (1985). Molecular biology of theqa gene cluster of Neurospora. In Gene Manipulationsin Fungi, ed. J. W. Bennett & L. L. Lasure, pp.229-244. Academic Press, Orlando, FL.

Ito, H., Fukuda, Y., Murata, K. & Kimura, A. (1983).Transformation of intact yeast cells treated withalkali cations. Journal of Bacteriology, 153, 163-168.

Johnston, S. A., Anziano, P. Q., Shark, K., Sanford,J. C. & Butow, R. A. (1988). Mitochondrialtransformation in yeast by bombardment withmicroprojectiles. Science, 240, 1538-1541.

Kinsey, J. A. & Rambosek, J. A. (1984).Transformation of Neurospora crassa with the clonedam (glutamate dehydrogenase) gene. Molecular andCellular Biology, 4, 117-122.

Kistler, H. C. & Benny, U. K. (1988). Genetictransformation of the fungal plant wilt pathogen,Fusarium oxysporum. Current Genetics, 13, 145-149.

Klein, T. M., Fromm, M., Weissinger, A., Tomes, D.,Schaaf, S., Sletten, M. & Sanford, J. C. (1988a).Transfer of foreign genes into intact maize cells withhigh-velocity microprojectiles. Proceedings of theNational Academy of Sciences, USA, 85, 4305^1309.

Klein, T. M., Gradziel, T., Fromm, M. E. & Sanford,J. C. (19886). Factors influencing gene delivery intoZea mays cells by high-velocity microprojectiles.Bio/Technology, 6, 559-563.


Klein, T. M., Harper, E. C , Svab, Z., Sanford, J. C ,Fromm, M. E. & Maliga, P. (1988c). Stable genetictransformation of intact Nicotiana cells by theparticle bombardment process. Proceedings of theNational Academy of Sciences, USA, 85, 8502-8505.

Klein, T. M., Wolf, E. D., Wu, R. & Sanford, J. C.(1987). High-velocity microprojectiles for deliveringnucleic acids into living cells. Nature (London), 327,70-73.

Knowles, J., Penttila, M., Harkki, A., Nevalainen, H.,Teeri, T., Saloheimo, M. & Uusitalo, J. (1989).Applications of the molecular biology of Trichodermareesei. In Molecular Biology of Filamentous Fungi,Foundation for Biotechnical and IndustrialFermentation Research, Vol. 6, ed. H. Nevalainen &M. Penttila, pp. 113-118. Foundation forBiotechnical and Industrial Fermentation Research,Helsinki.

Kubicek-Pranz, E. M., Gruber, F. & Kubicek, C. P.(1991). Transformation of Trichoderma reesei withthe cellobiohydrolase II gene as a means forobtaining strains with increased cellulase productionand specific activity. Journal of Biotechnology, 20,83-94.

Kuiper, M. T. R. & de Vries, H. (1985). Arecombinant plasmid carrying the mitochondrialplasmid sequence of Neurospora intermedia LaBelleyields new plasmid derivatives in Neurospora crassatransformants. Current Genetics, 9, 471-477.

Mattern, I. E., Unkles, S., Kinghorn, J. R., Pouwels,P. H. & van den Hondel, C. A. M. J. J. (1987).Transformation of Aspergillus oryzae using the A.nigerpyrG gene. Molecular and General Genetics, 210,460-461.

Maundrell, K., Wright, A. P. H., Piper, M. & Shall, S.(1985). Evaluation of heterologous ARS activity in5. cerevisiae using cloned DNA from S. pombe.Nucleic Acids Research, 13, 3711-3722.

Miller, B. L., Miller, K. Y. & Timberlake, W. E.(1985). Direct and indirect gene replacements inAspergillus nidulans. Molecular and Cellular Biology, 5,1714-1721.

Miller, J. F., Dower, W. J. & Tompkins, L. S. (1988).High-voltage electroporation of bacteria: genetictransformation of Campylobacter jejuni with plasmidDNA. Proceedings of the National Academy of Sciences,USA, 85, 856-860.

Mishra, N. C. (1985). Gene transfer in fungi.Advances in Genetics, 23, 73-178.

Mishra, N. C. & Tatum, E. L. (1973). Non-Mendelian inheritance of DNA-induced inositolindependence in Neurospora. Proceedings of theNational Academy of Sciences, USA, 70, 3875-3879.

Nevalainen, H., Penttila, M., Harkki, A., Teeri, T. &Knowles, J. (1991). The molecular biology ofTrichoderma and its application to the expression ofboth homologous and heterologous genes. In

Molecular Industrial Mycology, ed. S. A. Leong &R. M. Berka, pp. 129-148. Marcel Dekker, NewYork.

Oliver, R. P., Roberts, I. N., Harling, R., Kenyon, L.,Punt, P. J., Dingemanse, M. A. & van den Hondel,C. A. M. J. J. (1987). Transformation of Fulviafulva, a fungal pathogen of tomato, to hygromycin Bresistance. Current Genetics, 12, 231-233.

Paietta, J. V. & Marzluf, G. A. (1985). Genedisruption by transformation of Neurospora crassa.Molecular and Cellular Biology, 5, 1554-1559.

Panaccione, D. G., McKiernan, M. & Hanau, R. M.(1988). Colletotrichum graminicola transformed withhomologous and heterologous benomyl-resistancegenes retains expected pathogenicity to corn.Molecular Plant-Microbe Interactions, 1, 113-120.

Parsons, K. A., Chumley, F. G. & Valent, B. (1987).Genetic transformation of the fungal pathogenresponsible for rice blast disease. Proceedings of theNational Academy of Sciences, USA, 84, 4161-4165.

Penttila, M., Nevalainen, H., Ratto, M., Salminen, E.& Knowles, J. (1987). A versatile transformationsystem for the cellulolytic filamentous fungusTrichoderma reesei. Gene, 61, 155-164.

Rambosek, J. & Leach, J. (1987). Recombinant DNAin filamentous fungi: progress and prospects. CRCCritical Reviews in Biotechnology, 6, 357-393.

Richey, M. G., Marek, E. T., Schardl, C. L. & Smith,D. A. (1989). Transformation of filamentous fungiwith plasmid DNA by electroporation.Phytopathology, 79, 844-847.

Rodriguez, R. J. & Yoder, O. C. (1987). Selectablegenes for transformation of the fungal plantpathogen Glomerella cingulata f. sp. phaseoli(Colletotrichum lindemuthianum). Gene, 54, 73-81.

Roncero, M. I. G., Jepsen, L. P., Stroman, P. & vanHeeswijck, R. (1989). Characterization of a leuAgene and an ARS element from Mucor circinelloides.Gene, 84, 335-343.

Rothstein, R. J. (1983). One-step gene disruption inyeast. In Recombinant DNA, part C, Methods inEnzymology, Vol. 101, ed. R. Wu, L. Grossman &K. Moldave, pp. 202-211. Academic Press, NewYork.

Salch, Y. P. & Beremand, M. N. (1988). Developmentof a transformation system for Fusarium sambucinum.Journal of Cellular Biochemistry, Suppl. 12C, 290(No. Y 341).

Salovuori, I., Makarow, M., Rauvala, H., Knowles, J.& Kaariainen, L. (1987). Low molecular weighthigh-mannose type glycans in a secreted protein ofthe filamentous fungus Trichoderma reesei.Bio/Technology, 5, 152-156.

Samac, D. A. & Leong, S. A. (1989). Mitochondrialplasmids of filamentous fungi: characteristics anduse in transformation vectors. MolecularPlant-Microbe Interactions, 2, 155-159.

48 GOLDMAN et al.

Sanford, J. C , Smith, F. D. & Russell, J. A. (1993).Optimizing the biolistic process for differentbiological applications. In Recombinant DNA, partH, Methods in Enzymology, Vol. 217, ed. R. Wu3pp 483-508. Academic Press, San Diego.

Schafer, W., Straney, D.5 Ciufetti, L., Van Etten,H. D. & Yoder, O. C. (1989). One enzyme makes afungal pathogen, but not a saprophyte, virulent on anew host plant. Science, 246, 247-249.

Scherer, S. & Davis, R. W. (1979). Replacement ofchromosome segments with altered DNA sequencesconstructed in vitro. Proceedings of the NationalAcademy of Sciences, USA, 76, 4951-4955.

Schrank, A., Tempelaars, C , Sims, P. F. G., Oliver,S. G. & Broda, P. (1991). The trpC gene ofPhanerochaete chrysosporium is unique in containingan intron but nevertheless maintains the order offunctional domains seen in other fungi. MolecularMicrobiology, 5, 467-476.

Selker, E. U., Cambareri, E. B., Jensen, B. C. &Haack, K. R. (1987). Rearrangement of duplicatedDNA in specialized cells of Neurospora. Cell, 51,741-752.

Selker, E. U. & Garrett, P. W. (1988). DNA sequenceduplications trigger gene inactivation in Neurosporacrassa. Proceedings of the National Academy of Sciences,USA, 85, 6870-6874.

Shark, K. B., Smith, F. D., Harpending, P. R.,Rasmussen, J. L. & Sanford, J. C. (1991). Biolistictransformation of a procaryote, Bacillus megaterium.Applied and Environmental Microbiology, 57, 480-485.

Shigekawa, K. & Dower, W. J. (1988). Electroporationof eukaryotes and prokaryotes: a general approach tothe introduction of macromolecules into cells.BioTechniques, 6, 742-751.

Skatrud, P. L., Queener, S. W., Carr, L. G. & Fisher,D. L. (1987). Efficient integrative transformation ofCephalosporium acremonium. Current Genetics, 12,337-348.

Smith, D. J., Burnham, M. K. R., Edwards, J., Earl,A. J. & Turner, G. (1990). Cloning andheterologous expression of the penicillin biosyntheticgene cluster from Penicillium chrysogenum.Bio/Technology, 8, 39-41.

Smith, J. L., Bayliss, F. T. & Ward, M. (1991).Sequence of the cloned pyr4 gene of Trichodermareesei and its use as a homologous selectable markerfor transformation. Current Genetics, 19, 27-33.

Stahl, U., Tudzynski, P., Kiick, U. & Esser, K.(1982). Replication and expression of abacterial-mitochondrial hybrid plasmid in thefungus Podospora anserina. Proceedings of the NationalAcademy of Sciences, USA, 79, 3641-3645.

Stohl, L. L. & Lambowitz, A. M. (1983).Construction of a shuttle vector for the filamentousfungus Neurospora crassa. Proceedings of the NationalAcademy of Sciences, USA, 80, 1058-1062.

Struhl, K. (1983). The new yeast genetics. Nature{London), 305, 391-397.

Suzuki, K., Imai, Y., Yamashita, I. & Fukui, S.(1983). In vivo ligation of linear DNA molecules tocircular forms in the yeast Saccharomyces cerevisiae.Journal of Bacteriology, 155, 747-754.

Taylor, J. W., Smolich, B. D. & May, G. (1985). Anevolutionary comparison of homologousmitochondrial plasmid DNAs from three Neurosporaspecies. Molecular and General Genetics, 201,161-167.

Thomas, M. D. & Kenerley, C. M. (1989).Transformation of the mycoparasite Gliocladium.Current Genetics, 15, 415-420.

Tilburn, J., Scazzocchio, C , Taylor, G. G., Zabicky-Zissman, J. H., Lockington, R. A. & Davies, R. W.(1983). Transformation by integration in Aspergillusnidulans. Gene, 26, 205-221.

Timberlake, W. E. & Marshall, M. A. (1989). Geneticengineering of filamentous fungi. Science, 244,1313-1317.

Tsukuda, T., Carleton, S., Fotheringham, S. &Holloman, W. K. (1988). Isolation andcharacterization of an autonomously replicatingsequence from Ustilago maydis. Molecular andCellular Biology, 8, 3703-3709.

Turgeon, B. G., Garber, R. C. & Yoder, O. C. (1985).Transformation of the fungal maize pathogenCochliobolus heterostrophus using the Aspergillusnidulans amdS gene. Molecular and General Genetics,201, 450-453.

Turgeon, B. G., Garber, R. C. & Yoder, O. C. (1987).Development of a fungal transformation systembased on selection of sequences with promoteractivity. Molecular and Cellular Biology, 1, 3297-3305.

Ulhoa, C. J., Vainstein, M. H. & Pederby, J. F.(1992). Transformation of Trichoderma species withdominant selectable markers. Current Genetics, 21,23-26.

Van Heeswijck, R. (1986). Autonomous replication ofplasmids in Mucor transformants. Carlsberg ResearchCommunications, 51, 433-443.

Wang, J., Holden, D. W. & Leong, S. A. (1988). Genetransfer system for the phytopathogenic fungusUstilago maydis. Proceedings of the National Academyof Sciences, USA, 85, 865-869.

Ward, M. (1991). Aspergillus nidulans and otherfilamentous fungi as genetic systems. In ModernMicrobial Genetics, ed. U. N. Streips & R. E. Yasbin,pp. 455-496. Wiley-Liss, New York.

Ward, M., Kodama, K. H. & Wilson, L. J. (1989).Transformation of Aspergillus awamori and A. nigerby electroporation. Experimental Mycology, 13,289-293.

Ward, M., Wilson, L. J., Kodama, K. H., Rey, M. W.& Berka, R. M. (1990). Improved production ofchymosin in Aspergillus by expression as a


glucoamylase-chymosin fusion. Bio/Technology, 8,435-440.

Weltring, K.-M., Turgeon, B. G., Yoder, O. C. &VanEtten, H. D. (1988). Isolation of a phytoalexin-detoxification gene from the plant pathogenic fungusNectria haematococca by detecting its expression inAspergillus nidulans. Gene, 68, 335-344.

Whitehead, M. P., Gurr, S. J., Grieve, C , Unkles,S. E., Spence, D., Ramsden, M. & Kinghorn, J. R.(1990). Homologous transformation ofCephalosporium acremonium with the nitratereductase-encoding gene (niaD). Gene, 90, 193-198.

Whittington, H., Kerry-Williams, So Bidgood, K.,Dodsworth, N., Peberdy, J., Dobson, M.,Hinchliffe, E. & Ballance, D. J. (1990). Expressionof the Aspergillus niger glucose oxidase gene in A.niger, A. nidulans and Saccharomyces cerevisiae.Current Genetics, 18, 531-536.

Williams, R. S., Johnston, S. A., Riedy, M., DeVit,

M. J., McElligott, S. G. & Sanford, J. C. (1991).Introduction of foreign genes into tissues of livingmice by DNA-coated microprojectiles. Proceedings ofthe National Academy of Sciences, USA, 88,2726-2730.

Yelton, M. M., Hamer, J. E. & Timberlake, W. E.(1984). Transformation of Aspergillus nidulans byusing a trpC plasmid. Proceedings of the NationalAcademy of Sciences, USA, 81, 1470-1474.

Yelton, M. M., Timberlake, W. E. & van den Hondel,C. A. M. J. J. (1985). A cosmid for selecting genesby complementation in Aspergillus nidulans: selectionof the developmentally regulated^ locus.Proceedings of the National Academy of Sciences, USA,82, 834-838.

Yoder, O. C. & Turgeon, B. G. (1985). Molecularbases of fungal pathogenicity to plants. In GeneManipulations in Fungi, ed. J. W. Bennett & L. L.Lasure, pp. 417-448. Academic Press, Orlando, FL.

PART II TRANSFORMATION OFCEREAL CROPS

5Rice Transformation:Methods and ApplicationsJunko Kyozuka and Ko Shimamoto

Introduction

Genetic transformation offers new approaches tomany fundamental problems in plant biology. Inparticular, transgenic plants are essential tools inunderstanding in vivo functions of plant genes andmolecular mechanisms of their regulation (Schell,1987). In addition, genetic transformation pro-vides novel approaches to crop improvement.Early examples are herbicide- (Comai et al.> 1985;De Block et ah, 1987), virus- (Powell-Abel et al,1986) and insect- (Vaeck et al, 1987; Perlak et al.,1990) resistant dicotyledonous crops such astobacco and potato. Recently, to improve agro-nomically important traits, more sophisticatedstrategies have been designed and applied to anumber of crop species (Fraley, 1992).

In contrast to dicotyledonous species, produc-tion of transgenic monocotyledonous (monocot)plants has been difficult despite extensive effortsover many years. Because of this, expression ofgenes derived from monocot species is oftenexamined in transgenic dicot species (Lamppa,Nagy & Chua, 1985; Keith & Chua, 1986; Colotet al, 1987; Ellis et al, 1987). However, increas-ing evidence obtained from transgenic experi-ments suggests that there are differences inregulation of gene expression between monocotand dicot species. Monocot genes are not alwaysexpressed correctly or at all in transgenic dicotplants (Keith & Chua, 1986; Colot et al, 1987;Ellis et al.y 1987). Moreover, anatomical differ-ences between monocot and dicot species oftenmake it difficult to evaluate accurately tissue/cellspecific expression of monocot-derived genes intransgenic dicot plants (Schernthaner, Matzke &

Matzke, 1988; Lloyd et ai3 1991; Matsuoka &Sanada, 1991).

Transgenic monocot plants are thereforerequired for elucidation of monocot gene expres-sion and of differences in the system of geneexpression between monocot and dicot plants; amonocot species amenable to genetic engineeringis needed as a model plant for studies in gene reg-ulation of monocot genes. In addition, the devel-opment of techniques for routine generation oftransgenic monocots is of critical importance inplant breeding, because they include major cropspecies such as rice, maize, wheat and barley.

Recently, routine generation of fertile trans-genic rice plants (Oryza sativa) by direct DNAtransfer to protoplasts has been achieved(Shimamoto et al, 1989; Shimamoto, 1991). Inaddition to this, rice has a number of favorablefeatures as a model monocot plant that are sharedwith Arabidopsis thaliana, a model dicot plant forplant molecular genetics (Meyerowitz, 1989).These features include well-developed linkage andrestriction fragment length polymorphism(RFLP) maps (Kinoshita, 1984; McCouch et al.,1988; Wang & Tanksley, 1989), and a smallgenome three to five times as big as that of A.thaliana that contains less repetitive DNA com-pared to other plant species (Zhao et al, 1989;Nishibayashi, 1992)

In this chapter, we review the present status ofrice transformation (Table 5.1), and discuss keytechniques and unsolved problems in the methodsfor generation of transgenic rice plants. Further-more, we describe results of recent research activi-ties making use of transgenic rice and discuss thepotential of rice as a model monocot species for

53

54 KYOZUKA AND SHIMAMOTO

Table 5.1. Transgenic rice plants

Cultivar Method

Protoplast transformationYamahoushiPi4, Taipei 309Taipei 309YamahoushiNipponbareTaipei 309Chinsurah Boro IINipponbare, Taipei309

IR54Norin 8, SasanishikiNipponbareNipponbareTaipei 309ToridelNipponbareTaipei 309

NipponbareNipponbareTaipei 309

EPPEGEPEPEPEPPEGPEG

PEGEPEPEPEPEPEPEP, PEG

PEGEPPEG

Microprojectile bombardmentGulfmont

IR 54, IR26IR36, IR72

Selection

G418—KmR

HmR

HmR

KmR, G418HmR

HmR

KmR

HmR

HmR

HmR

KmR

HmR

HmR

HmR, MtxR

HmR

HmR

—

Hm,Bialaphos

Introduced genes

35S-npfll, 35S-uidAmaize adM-uidA35S-npfll3bS-hph, roiC-uidA3bS-hph, 3bS-uidA35S-npfll, 3bS-uidA3bS-hph3bS-hph

35S-npfll, 3bS-uidA3bS-hph, 3bS-uidA3bS-hph, 3bS-uidA3bS-hph, maize adh}-uidA3bS-npt\\3bS-hph, maize3bS-hph, maize adh\-uidA3bS-hph,p1',2'-uidA,3bS-dhfr

3bS-hph, maize Ac3bS-hph, rice cab-uidArice actin-uidA

3bS-hph, 3bS-bar 35S-uidA

Fertility

NTNTNTNT+NT+NT

NT++NT+NT+NT

NT++

+

Reference

Toriyama etal. (1988)Zhang &Wu (1988)Zhang etal. (1988)Matsuki etal. (1989)Shimamoto etal. (1989)Bauraw&Hall (1990)Datta etal. (1990)Hayashimoto & Murai (1990)

Peng etal. (1990)Tada etal. (1990)Terada & Shimamoto (1990)Kyozuka etal. (1990)Davey etal. (1991)Izawa etal. (1991)Kyozuka etal. (1991)Meijer ef al. (1991)

Murali etal. (1991)Tada etal. (1991)Zhang etal. (1991)

Christou etal. (1991)

Notes:EP, electroporation; PEG, polyethylene glycol; G418, synthetic aminoglycoside antibiotic, geneticin; Km, kanamycin; Hm,hygromycin; Mtx, methotrexate; 35S, CaMV 35S RNA promoter; adrW, maize alcohol dehydrogenase gene 1 promoter; rolC,promoter from a pathogenesis-related gene of the TL-DNA of the Agrobacterium rhizogenes Ri plasmid; P1', 2', T-DNA 1' and21 gene promoters from Agrobacterium tumefaciens; npt\\, neomycin phosphotransferase II gene; uidA, (3-glucuronidase gene;hph, hygromycin phosphotransferase gene; dhfr, dihydrofolate reductase gene; bar, phosphinothrin acetyltransferase gene;Ac, maize autonomous transposable element activator; NT, not tested.

studies in regulation of gene expression that isunique to monocot plants.

Methods for generation of transgenicrice plants

Protoplast transformation

Agrobacterium-mediated transformation is routinelyused for production of transgenic dicot plants.However, monocot species, especially cereals, areat best weakly susceptible to Agrobacterium tumefa-ciens. Thus, many laboratories have sought alterna-tive transformation methods for cereals. Among anumber of methods studied, protoplast transforma-tion and microprojectile bombardment have beensuccessfully used to produce fertile transgenicplants in rice and maize {Zea mays) (Table 5.1).

In cereals, rice is exceptional in that plantregeneration from protoplasts has been well estab-lished (Abdullah, Cocking & Thompson, 1986;Toriyama, Hinata & Sasaki, 1986; Yamada, Yang& Tang, 1986; Kyozuka, Hayashi & Shimamoto,1987). Therefore, rice transformation can be rou-tinely performed with direct DNA transfer intoprotoplasts (Table 5.1) and subsequent plantregeneration. Production of transgenic maize(Rhodes et al> 1988) and orchardgrass (Dactylisglomerate Horn et al> 1988) by protoplast trans-formation has also been achieved; however, theirfertilities have yet to be demonstrated.

The 'quality' of embryogenic suspension cul-tures from which protoplasts are isolated, is one ofthe most critical factors for successful productionof transgenic plants. Normally we use protoplastsisolated from embryogenic suspension cultures

Rice 55

derived from calli originating from mature seeds.Such protoplasts show high plating efficiency (ca10%). A detailed protocol for initiation and main-tenance of embryogenic suspension cultures wasdescribed by Kyozuka, Shimamoto & Ogura(1989) and Kyozuka & Shimamoto (1991).Suspension cultures of most japonica cultivarsproduce protoplasts capable of high frequencycolony formation 1-2 months after initiation ofthe suspension culture. Suspension cultures arerenewed every 6 months to avoid loss of theirmorphogenic capacity during prolonged culture.

The efficiency of generating transgenic riceplants varies depending on cultivars, presumablydue to differences in their adaptability to cultureconditions in vitro and in their competence foraccepting foreign DNA. Although transgenicplants have been obtained in some indica rice(Datta et al> 1990), protoplast culture of indicavarieties is generally more difficult than that ofjaponica varieties (Kyozuka, Otoo & Shimamoto,1988). For the recalcitrant varieties, includingsome japonica varieties and most of indica vari-eties, further improvement of culture conditions isnecessary to generate transgenic plants repro-ducibly.

Both electroporation and PEG treatment areused to introduce foreign DNA into rice proto-plasts (Table 5.1). When conditions are opti-mized, the efficiency of DNA uptake byprotoplasts does not seem to be much differentbetween these two methods. When protoplasttransformation was compared with the micro-projectile bombardment, the former is more fre-quently used for production of transgenic calli andplants in a wide variety of cereal species, includingmaize (Rhodes et al., 1988), barley (Hordeum vul-gare; Lazzeri et al., 1991) and orchardgrass (HornetaL, 1988).

Recently, production of fertile transgenic indicaand japonica rice plants by microprojectile bom-bardment has been reported; the scutellar tissue ofimmature embryo was bombarded and plantswere regenerated from calli derived from thetransformed tissue (Christou, Ford & Kofron,1991). The main advantage of this technique isthe minimum requirement of tissue culture, whichmakes the method relatively genotype indepen-dent. However, two steps need to be furtherexamined: the frequency of the plant regenerationfrom transformed calli and the degree of chimeraformation in primary transgenic plants. Further-

more, collection of a large number of immatureembryos is laborious and requires a constantsupply of mature plants. Regeneration of fertiletransgenic maize plants using microprojectilebombardment has also been reported by twogroups (Fromm et al., 1990; Gordon-Kamm et al.,1990). Both groups adopted embryogenic suspen-sion cells as target materials. At present, the suc-cess is limited to a few genotypes and thefrequency of transformation is relatively low.Therefore, the frequency of plant regenerationfrom transformed cells should be improved toestablish the method for the routine production oftransgenic maize plants.

Selectable markers

The effective selection of transformed cellsdepends greatly on selectable markers and theselection procedures used. Hygromycin (hph),kanamycin (nptll), and bialaphos (bar) resistancegenes have been used as selectable marker genesfor rice transformation (Table 5.1). However,effective selection of transformed cells is notalways easy with the KmR marker because ricecells generally have background resistance tokanamycin, and furthermore albino or sterileplants were often obtained from rice callus afterselection with G418, another aminoglycosideinactivated by the NPT II enzyme (K.Shimamoto, unpublished data). AlthoughDekeyser et al. (1989) suggested that the bar geneencoding phosphinothricin acetyltransferase(PAT), which confers resistance to herbicidebialaphos, is an efficient selection marker gene inrice, its usefulness in rice transformation needs tobe further examined.

In our procedure, selection with hygromycin Bis started after 10 to 14 days from the initiation ofprotoplast culture and selection is repeated twicefor 7-10 days each (Figure 5.1). Our preliminaryexperiments showed that transformation with lin-earized plasmid carrying the marker gene pro-duced more HmR calli than that with circularplasmid. HmR calli thus selected are transferred tothe regeneration medium and then shoots arisefrom these transformed calli within 2-6 weeks.According to our procedure, transformed riceplantlets can be obtained within 8-10 weeks afterelectroporation. The frequency of plant regenera-tion from HmR calli is approximately 60%-80%(for a detailed protocol, see Kyozuka &


0Protoplasts

oo o10 |

<*y

Smallcolony

O

100

N®

>

/ ; /

^

Smallcallus

10 mm 1 mm

-i 1 , 1 18 1 2 week

Shoot

10 mm10 mm

Figure 5.1. Steps in protoplast transformation in rice. Electroporated protoplasts are embedded in agarose blocks andcultured by mixed nurse culture method (Kyozuka etai, 1987). After 10 days of the culture, the nurse cells are removed andagarose blocks are transferred to fresh medium containing 30 ^g hygromycin B/ml (first selection). Hm-resistant coloniesbecome visible (0.5-1.0 mm in diameter) at the end of the first selection; then agarose blocks are transferred on to solidmedium with 30 [xg hygromycin B/ml (second selection). At the end of the second selection (ca 5 weeks afterelectroporation), Hm-resistant calli become about 2 mm in diameter. Then the Hm-resistant (transformed) calli aretransferred to regeneration medium without hygromycin. Shoots are obtained from 60%-80% of Hm-resistant calli within1-2 months from the transfer to regeneration medium.

Shimamoto, 1991). Albino plants rarely appearedafter selection with hygromycin, however,somaclonal variations seem to occur more often intransgenic plants than in plants regenerated fromnontransformed protoplasts. Also, low fertilitywas observed in some transgenic plants probablydue to somaclonal variations and ploidy changes.

Intron enhancement of gene expression

In cereal species, steady-state levels of mRNA andproteins are increased by the presence of intronsin vector constructs. So far, the enhancing effectof introns has been found not only with intronsfrom cereal species such as maize adhl introns 1,2, 3, 6, 8, and 9 (Kyozuka et al, 1990; Callis,Fromm & Walbot, 1987; Oard, Paige & Dvorak,1989; Mascarenhas et al, 1990; Luehrsen &Walbot, 1991), maize bz intron 1 (Callis et al.,1987), maize shl intron 1 (Vasil et al, 1989),maize actin gene (act) intron 3 (Luehrsen &Walbot, 1991) and rice actin gene {act) intron 1(McElroy et al, 1990), but also with intron 1 ofthe catalase gene (cat) derived from the dicotspecies castor bean (Tanaka et al, 1990).Stimulating effects of the first intron of maizeadhl gene (Kyozuka et al, 1990), rice act geneintron 1 (McElroy et ah, 1990) and castor bean car

intron 1 (Tanaka et al, 1990) on the expression ofthe (3-glucouronidase (uidA) reporter gene havebeen described in rice.

The degree of enhancement is dependent onthe origin of the intron, the sequences flanking theintron, the reporter gene and the host cells usedfor experiments. The first intron of maize adhlgene increased the (3-glucuronidase (GUS) activ-ity in rice protoplasts four- to six-fold when the35S promoter of cauliflower mosaic virus (CaMV)was used, while 11- to 18-fold increases wereobserved with the maize adhl promoter (Kyozukaet al, 1990). These results are in agreement withthe finding by Callis et al (1987) that the firstintron of maize adhl gene enhanced chloram-phenicol acetyltransferase (CAT) activity 16- to112-fold in maize protoplasts when driven by theadhl promoter, but 5- to 22-fold when the CaMV35S promoter was used (Callis et al, 1987). Thestrong enhancement of gene expression obtainedby insertion of the first intron of castor bean catgene in the N-terminal region of the codingsequence of the uidA gene driven by the 35S pro-moter was demonstrated in transiently, as well asstably, transformed rice cells (Tanaka et al,1990). Ten- to 90-fold increases in GUS activitywere observed. Northern blot analysis showedthat the increase in GUS activity was correlated

Rice 57

Table 5.2. Frequency of cotransformation

No. of HmR

calliNo. of Gus+ calli

Experiment 1Experiment 2

5838

3015

51.739.5

Notes:Protoplasts were cotransformed with 25 |xg of plasmid DNAcarrying 35S-hph gene and 25 |xg of plasmid DNA carryinguidA gene.

with the increased level of mature mRNA and effi-cient splicing. When the same construct was intro-duced into tobacco cells, little increase of geneexpression was observed (Ohta et al.9 1990) andthe intron was not spliced efficiently (Tanaka etaL, 1990).

Cotransformation

As a means to introduce nonselectable genes intorice, cotransformation has been used successfully.In cotransformation, a plasmid carrying the non-selectable gene and another carrying the selectablemarker gene are mixed and introduced into proto-plasts. In our experiments, when a GUS plasmidis mixed with the HmR plasmid in the ratio of 1:1(50 \\g each/ml), the efficiency of cotransforma-tion was 30%-50% at the level of expression(Table 5.2). The cotransformation frequencydoes not seem to vary, regardless of which nonse-lectable genes are used for each experiment. Oneclear advantage of this method is that constructionof a composite plasmid carrying both the selec-table marker gene and the nonselectable gene isnot required. By using the cotransformationmethod, more than three different genes can beintroduced into a protoplast by single treatment.Cotransformed genes can be segregated in theprogeny when they are not linked, giving rise totransgenic plants carrying no selection markergene.

Integration patterns of foreign DNA

Integration patterns of foreign DNA in thegenome of transgenic rice were analyzed bySouthern blot analysis using various fragments ofHmR plasmids as hybridization probes (H.Fujimoto et al> unpublished data). The resultsindicated that five out of six HmR transformants

contained one or two functional hph sequencesand no other fragmented pieces of the hphsequence were detected. With respect to the copynumber of the integrated hph gene, the results aregenerally in good accordance with those fromtransgenic cereals obtained by other groups, andthe majority of the transgenic plants containedone or two copies of the transgene. Our study alsorevealed that there are three different patterns ofintegration of the HmR plasmid into rice genome.First, almost the entire unit of the plasmid DNA isintegrated into a single site. Second, several frag-mented pieces of the plasmid DNA are integratedinto multiple sites of rice genome. Third, a tan-dem repeat of the entire plasmid is integrated(four- to five-copy) into one site. Further analysisof integrated foreign DNA in transgenic riceshould be important for understanding the mech-anism of DNA integration and for using trans-genic plants effectively in the introduction ofeconomically important genes.

Unsolved problems

Despite rapid progress in rice transformationbased on protoplast culture, further improve-ments will be required to fully realize the potentialof transgenic approaches in rice research. First,establishment of transformation protocols forsome of the major indica varieties of rice areurgently required because this variety is grown inthe majority of rice-growing countries in theworld. Recent developments in indica rice trans-formation and plant regeneration from protoplastsindicates that careful identification of stepsinvolved in generation of transgenic plants willeventually solve the problems and it will not belong before protoplast-mediated transformationbecomes possible in most of the major indica vari-eties of rice. Second, controlled integration of for-eign DNA will be desired to utilize transgenic riceplants for some research areas. Deleted or re-arranged copies of the plasmid DNA are oftendetected in the chromosomes of transgenic riceplants and they may cause undesired mutations inaddition to the somaclonal variations that occurinevitably through protoplast culture. At present,our understanding of factors influencing patternsand efficiency of integration of transgenes is lim-ited. One possible approach is to establish amethod for gene targeting using homologousrecombination (Paszkowski et al., 1988). The


gene-targeting technique should allow the integra-tion of foreign genes exclusively into specific siteson the chromosomes. Gene targeting may also beused to disrupt a gene of interest in order to revealin vivo function where the actual role of the gene isunknown.

Regulated expression of genes revealedby in vivo GUS histochemical analysis

Analysis of regulation of gene expression inhomologous transgenic system has many advan-tages because introduced genes exhibit their nat-ural properties in the homologous host plants.However, for most cereals, such a system is notavailable, and none is as well developed as thosefor rice. Therefore, rice can be used as a hostspecies to analyze in vivo functions not only ofpromoters of rice genes but also of promoters iso-lated from other cereal species.

The first promoter examined in transgenic riceplants is the 35S promoter of CaMV (Terada &Shimamoto, 1990) because it has been usedextensively for expression of agronomically usefulgenes in dicot species. The quantitative and histo-chemical analysis of the 35S promoter expressionin transgenic rice plants and their progeny demon-strated that it directs GUS expression in a numberof tissues including root, leaf, flower and seeds,and that the level of GUS expression in leaf androot was comparable to that in transgenic tobaccoplants. A general conclusion on the expressionpattern of the 35S promoter is that it is a 'constitu-tive' promoter in rice and no major differences inits expression pattern between rice and tobaccoare detected. Thus, the 35S promoter should beuseful for introduction of agriculturally importantgenes such as coat protein genes of viruses (seep. 60).

Some studies demonstrated that the 35S pro-moter was less effective in cereal cells than in dicotcells (Hauptmann etal, 1987). Non-optimal con-ditions of transformation and/or lack of compe-tence of cells in DNA uptake used might havecontributed to the low activity of the 35S pro-moter.

We examined the regulated expression of theuidA reporter gene fused with maize alcohol de-hydrogenase 1 gene (adhl) promoter and a ricerbcS promoter (small subunits of Rubisco (ribu-lose-l,5-bisphosphate carboxylase oxygenase)) in

transgenic rice plants (Plate 5.1). Maize adhl isone of the best studied genes in higher plants. Inmaize plants, alcohol dehydrogenase (ADH) ispresent in pollen, embryo, endosperm andseedling roots. In transgenic rice plants, the maizeadhl promoter directs constitutive GUS expres-sion in shoots, root caps of seedling roots andmature roots, anthers, anther filaments, pollen,scutellum and endosperm (Plate 5.l(a)-(d)).Although the histochemical GUS analysis withtransgenic rice plants made more detailed obser-vation of the expression possible, the overallspatial expression pattern of maize adhlpromoter-tt/dA gene in rice is similar to the distri-bution of ADH protein in maize plants except forthe expression in shoots of transgenic rice. Themost likely and the most attractive explanation forthis difference in expression is that GUS expres-sion conferred by the adhl promoter in transgenicrice may be regulated by cellular factor(s) presentor active only in rice shoots but not in maizeshoots. This difference in the expression patternbetween the two species may provide a suggestionto understanding the mechanism underlyingspecies differentiation in expression of homolo-gous genes.

Anaerobic induction of ADH proteins has beenwell documented in several plant species. Thus,anaerobic induction was carefully examined using5- to 7-day-old seedlings derived from selfed prog-enies of primary transgenic plants. The expressionwas strongly (up to 81-fold) induced in roots ofseedlings in response to anaerobic treatment for24 h, concomitant with an increase in the level ofuidA mRNA. Our results also indicate that induc-tion in the expression by maize adhl promotertakes place in specific regions of the root in trans-genic rice plants and the spatial pattern of expres-sion changes distinctly after induction.

These results indicate that the maize adhl-uidAfusion gene is expressed in a regulated mannerthat reflects the natural property of the promoter.As the maize adhl promoter did not confer suffi-cient expression in transgenic tobacco plants, it isevident that transgenic rice plants are more appro-priate hosts for studying the expression of mono-cot (cereal) genes.

Expression of genes encoding small subunits ofRubisco (rbcS) is leaf specific and regulated bylight. To investigate the regulation of rice rbcSexpression more precisely, the rbcS-uidA fusiongene was introduced into transgenic rice plants.

Rice 59

The histochemical study detected GUS activity inmesophyll cells of both leaf sheath and leaf bladebut not in epidermis or vascular tissues (Plate5.1(e)). No activity was detected in roots, flowersor seeds (Plate 5(/)-(*)). The levels of GUSexpression examined in five independent trans-genic plants ranged from 10 to 150 nmol 4-MU/min per mg protein, which is one order of magni-tude higher than that obtained with the 35S pro-moter. Little difference in the GUS activitybetween leaf sheath and leaf blade was detected.The expression of the rbcS-uidA gene was inducedby light in primary transformants as well as Rlprogeny plants and the induction took place at thelevel of transcription.

Other examples of promoters whose functionshave been studied in transgenic rice plants includethe rolC of the Ri plasmid (Matsuki et al, 1989),rice cab (Tada et aL, 1991) and rice actin gene(Zhang, McElroy & Wu, 1991). In leaves androots of transgenic rice plants, the rolC promoterdirected GUS activity only in vascular tissues.The activity of the rice cab gene promoter wasdetected in leaves, stems and floral organs but notin roots, and its expression was induced by light.The promoter of the rice actin gene is a usefulconstitutive promoter because it confers the highlevel of expression in all the tissues examined.

Examples examined so far of the regulatedexpression of promoters derived from rice as wellas from other cereals clearly show that transgenicrice is a valuable host for studying the expressionof monocot genes. Many interesting and impor-tant features of monocot gene expression will berevealed using transgenic rice plants in the nearfuture. Among others, as-elements responsible fortissue specific or developmentally regulatedexpression of monocot genes can be determinedby using transgenic rice plants. Transgenic riceplants should become useful tools to elucidate invivo functions of trans-acting factors that havebeen isolated from various plant species includingmonocots (Katagiri & Chua, 1992).

Introduction and expression of usefulgenes in transgenic rice plants

Genetic transformation provides a powerful toolfor crop improvement. In some dicot species,agronomically useful traits such as resistance toherbicides (Comai et al> 1985; De Block et al.,

1987), insects (Vaeck et al, 1987; Perlak et al,1990) and viruses (Powell-Abel et al, 1986) andalso male sterility (Mariani et al., 1990) havebeen conferred by genetic engineering. Novelapproaches are being taken to alter other agro-nomically important traits of crops. However, thistechnique has not been applied to cereals, mainlydue to the lack of reliable transformation strate-gies until recently. Establishment of an efficienttransformation system in rice has made it possibleto apply gene transfer methods to improve rice,one of the most important crops in the world.

Transposable elements

Transposon tagging has important uses in the iso-lation of new genes and a number of genes havebeen isolated by this method in maize andAntirrhinum majus (Wienand & Saedler, 1988;Carpenter & Coen, 1990). However, becauseactive transposable elements have not been identi-fied in the rice genome so far, the maize autono-mous transposable element activator (Ac) andnonautonomous element dissociation (Ds) wereintroduced into rice to develop a transposon-tagging system.

In order to introduce Ac, a phenotypic assay forexcision of the Ac element was employed (Izawa etal, 1991; Mural, Kawagoe & Hayashimoto1991). In this assay, excision of the Ac element isrecognized by an HmR (hygromycin resistant)phenotype, since excision of the Ac element fromthe untranslated leader sequence of the hph genereconstitutes a functional HmR gene. Excision andreintegration of the Ac element in the rice genomewas examined by Southern blot analysis, demon-strating that the introduced Ac element was activein transformed calli and transgenic plants.Sequence analysis of excision sites indicated thatthe Ac element was excised in rice in a mannersimilar to that in maize.

As an alternative to the use of an autonomousAc element, transposition of a nonautonomous Dselement was examined by cotransformation withthe Ac transposase gene fused with the 35S pro-moter. This Ac transposase gene is not able totranspose because it lacks the ends of Ac necessaryfor transposition. The Ds element was inserted inthe chimeric HmR gene in place of Ac and it wasfound that the Ds element can be excised fromand reintegrated into the rice chromosome by theaction of the transposase produced in trans by the


Ac transposase gene. Excision of the Ds elementfrom the hph gene was monitored as the appear-ance of HmR cells. Transposition of the Ds ele-ment was examined by Southern blot analysis andby sequencing its excision sites and rice DNAflanking integrated Ds elements. The analysisindicated that the Ds element actively transposesin rice in the presence of the Ac transposase gene.Sequences of Ds excision sites are similar to thoseof Ac excision sites in transgenic rice and 8 bpduplication of target sequences observed in maizewas also found in rice. These results suggestedthat a two-element (AclDs) system can be alsoused for tagging genes in rice. Interestingly, mostof the Hm calli did not contain integrated copiesof the Ac transposase gene, suggesting that tran-siently expressed Ac transposase acted on the Dselement and caused its transposition into the ricegenome. This gives us an interesting possibility ofgenerating transgenic plants carrying only nonau-tonomous Ds elements scattered throughout thegenome. Then, these Ds plants can be used forscreening for possible mutations caused by Dsinsertion in the next generation.

One characteristic of the Ac transposition is thatAc often transposes to sites linked to its originallocation. If it is also the case in rice, generation oftransgenic rice plants carrying transposable ele-ments on known chromosomes should greatlyfacilitate effective gene tagging by Ds. To mapinsertion sites of the elements on rice chromo-somes, DNA sequences flanking integrated Dselements can be isolated by the inverse poly-merase chain reaction method (H. Hashimoto &K. Shimamoto, unpublished data).

waxy gene for controlling starchcomposition of the grain

Starch composition in endosperm is one of theeconomically important traits in rice. Starch ofwild-type grain consists of 15%-30% amylose and70%-85% amylopectin, whereas the starch inendosperm of the waxy mutant completely lacksamylose. Amylose content varies greatly amongrice cultivars and affects grain qualities, graintastes and cooking properties. Although severalother factors contribute to the amylose contents ofrice seeds, they are determined primarily byexpression levels of the waxy gene.

A genomic region containing the rice waxy genehas been cloned and sequenced (Wang et al.3

1990). Therefore opportunities exist for manipu-lation of the amylose content in rice grains bygenetic engineering techniques. In our laboratory,several modified waxy genes were introduced intorice. Expression of antisense waxy genes fusedwith two promoters that are known to express indeveloping endosperm indeed reduces amylosecontent of the seed. Also it was found that trans-genic plants produced pollens that are not stainedby iodine. Our preliminary analysis suggests thatamylose content of the seed can be altered byintroduction of modified waxy genes. Becauseother genes such as those for ADP-glucosepyrophosphorylase or a branching enzyme knownto be involved in starch biosynthesis in cerealgrains have been isolated, it should be feasible toalter a wide variety of starchs in rice in the future.

Coat protein gene of rice stripe virus

Coat protein (CP)-mediated protection againstvirus diseases has been applied with a number ofdicot species since the first report showing that theCP of tobacco mosaic virus (TMV) expressed intransgenic plants conferred resistance to TMV(Powell-Abel et al., 1986). Transgenic rice plantsexpressing the CP of rice stripe virus (RSV) wasgenerated in our laboratory (Hayakawa et al.,1992). RSV is a member of the Tehui virus groupand is transmitted by small brown planthoppers.In Japan, Korea, China, Taiwan and the formerUSSR, RSV causes serious damage to rice. TheCP expression vector used in the experiment con-sisted of the 35S promoter, the first intron of cas-tor bean catalase gene, the coding sequence of theCP gene and polyadenylation site from the nopa-line synthase gene. Western blot analysis usingprimary transgenic plants revealed that, out of 33independent clones each of which gave rise to sev-eral plants, 19 expressed detectable levels of theCP. The amount of the CP produced in the riceleaves was estimated to be up to 0.5% of the totalsoluble protein. In the assay for viral resistance,transgenic plants expressing CP did not exhibitdisease symptoms, whereas the nontransformedcontrol plants as well as transformed plants notexpressing the CP showed clear disease symp-toms, indicating that the resistance to RSVdepended on expression of the introduced CPgene. The CP gene was stably transmitted to theprogeny of primary transgenic plants and CP

Rice 61

expression and the viral resistance were observedin the progeny plants.

This study indicated that introduction of CPgenes is a promising approach to introduce viralresistance in cereals. This strategy is applicable toother viruses such as Tungro virus, which is caus-ing severe damage to indica rice in many Asiancountries.

8-Endotoxin gene of Bacillus thuringiensis

Genes encoding insect control proteins fromBacillus thuringiensis (B.t.) have been introducedinto several crops for protection against insects. Inrice, insect damage is one of the major agriculturalproblems in many Asian countries. Two lepi-dopterans, rice stem borer and rice leaf folder, arelargely responsible for the insect damage to rice.Therefore, introduction of B.t. toxin genes shouldbe a useful approach for protection against the pests.

Preliminary attempts to express a B.t. toxingene in rice have been described (Yang et al.,1989; Xie, Fan & Ni, 1990). In these studies theB.t. coding sequence was translationally fusedwith the uidA gene and introduced into rice byprotoplast transformation. Southern blot analysisof resultant transgenic plants showed integrationof the B.t.-uidA fusion gene in the rice genome.Furthermore, GUS expression has been detectedin roots of the primary transgenic plants. Whetherthe transgenic plants exhibit resistance to any ofrice pests has not been reported.

In order to use the B.t. gene effectively againstmajor rice pests, a number of factors influencingprotein expression needs to be considered. Forinstance, it has been known that the level of B.t.toxin genes in plants is very low and it is generallynot sufficient to control insects in field conditions.One of the reasons for low level of expression is itscodon usage which is substantially different fromthat of plant genes. Extensive modification of thesequence has dramatically improved expressionlevel of the B.t. gene in plants (Perlak et aL, 1990).Thus, a similar modification may also be neededin order to apply the B. t. toxin for insect control ofrice. Another consideration is the choice of pro-moters. Stemborers, for instance, enter the stemsof plants and grow inside them. This should betaken into account when a B. t. expression vector isconstructed. In the future, however, efficientexpression of the B.t. gene in rice will be achievedand it should improve resistance of rice to insects.

Conclusions

Rice transformation is now well established andbeing used for genetic engineering and for studiesin regulation of genes derived from monocotyle-donous plants. Establishment of routine transfor-mation in rice has depended on well-definedprotoplast culture and subsequent plant regenera-tion. Nevertheless, further improvement isrequired in protoplast culture of major indica vari-eties and reduction of undesirable mutations gen-erated during the process of production oftransgenic plants. Understanding the mechanismsof nonhomologous recombination and compe-tence of cells for DNA uptake should help to fiir-ther improve techniques of rice transformation inthe future.

Development of a nearly saturated RFLP mapand use of transposable elements will providevaluable tools for identifying agriculturally usefulgenes of rice. These should expand our list ofgenes that can be introduced into rice after modi-fication for better productivity.

In gene regulation, roles of specific sequences inpromoter regions of various genes during develop-ment and differentiation will come from studiesusing transgenic plants. Furthermore, factorsdetermining the spatial and temporal expressionof genes should be better understood by makinguse of transgenic plants. As our understanding ofthese cis and trans factors involved in gene expres-sion advances, chances of generating novel riceplants with various improved traits will consider-ably increase.

References

Abdullah, R., Cocking, E. C. & Thompson, J. A.(1986). Efficient plant regeneration from riceprotoplasts through somatic embryogenesis.Bio/Technology, 4, 1087-1090.

Battraw, M. J. & Hall, T. C. (1990). Histochemicalanalysis of CaMV 35S promoter-P-glucuronidasegene expression in transgenic rice plants. PlantMolecular Biology, 15, 527-538.

Callis, Jo Fromm, M. & Walbot, V. (1987). Intronsincrease gene expression in cultured maize cells.Genes and Development, 1, 1183-1200.

Carpenter, R. & Coen, E. S. (1990). Floral homeoticmutations produced by transposon-mutagenesis inAntirrhinum majus. Genes and Development, 4,1483-1493.


Christou, P., Ford, T. L. & Kofron, M. (1991).Production of transgenic rice (Oryza sativa L.)plants from agronomically important indica andjaponica varieties via electric discharge by particleacceleration of exogenous DNA into immaturezygotic embryos. Bio/Technology, 9, 957-962.

Colot, V., Robert, L. S., Kavanagh, T. A., Bevan,M. W. & Thompson, R. D. (1987). Localization ofsequences in wheat endosperm protein genes whichconfer tissue-specific expression in tobacco. EMBOJournal, 6, 3559-3564.

Comai, L., Gacciotti, D., Hiatt, W. R., Thompson,G., Rose, R. E. & Stalker, D. (1985). Expression inplants of a mutant aroA gene from Salmonellatyphimurium confers tolerance to glyphosate. Nature{London), 317, 741-744.

Datta, S. K., Peterhans, A., Datta, K. & Potrykug, I.(1990). Genetically engineered fertile indica-ricerecovered from protoplasts. Bio I Technology, 8,736-740.

Davey, M. R., Kothari, S. L., Zhang, H., Rech, E. L.,Cocking, E. C. & Lynch, P. T. (1991). Transgenicrice: characterization of protoplast-derived plantsand their seed progeny. Journal of ExperimentalBotany, 42, 1159-1169.

De Block, M., Botterman, J., Vandewiele, M., Dockx,J., Thoen, C , Gossele, V., Mowa, N. R.,Thompson, C , Van Montagu, M. & Leemens, J.(1987). Engineering herbicide resistance in plants byexpression of a detoxifying enzyme. EMBO Journal,6,2513-2518.

Dekeyser, R., Claes, B., Marichal, M., Montague,M. C. & Caplan, A. (1989). Evaluation of selectablemarkers for rice transformation. Plant Physiology, 90,217-223.

Ellis, J. G., Llewellyn, D. J., Dennis, E. S. & Peacock,W. J. (1987). Maize adh-1 promoter sequencescontrol anaerobic regulation: addition of upstreampromoter elements from constitutive genes isnecessary for expression in tobacco. EMBO Journal,6, 11-16.

Fraley, R. (1992). Sustaining the food supply.Bio/Technology, 10, 40-43.

Fromm, M., Morrish, F., Armstrong, C , Williams, R.,Thomas, J. & Klein, T. M. (1990). Inheritance andexpression of chimeric genes in the progeny oftransgenic maize plants. Bio/Technology, 8, 833-844.

Gordon-Kamm, W. J., Spencer, T. M., Mangano,M. L., Adams, T. R., Daines, R. J., Start, W. G.,O'Brien, J. V., Chambers, S. A., Adams, W. R.,Willetts, N. G., Rice, T. B., Mackey, C. J., Krueger,R. W., Kausch, A. P. & Lemaux, P. G. (1990).Transformation of maize cells and regeneration offertile transgenic plants. Plant Cell, 2, 603-618.

Hauptmann, R. M., Ozias-Akins, P., Vasil V.,Tabaeizadeh, Z., Rogers, S. G., Horsch, R. B.,Vasil, I. K. & Fraley, R. T. (1987). Transient

expression of electroporated DNA inmonocotyledonous and dicotyledonous species.Plant Cell Reports, 6, 265-270.

Hayakawa, T., Zhu, Y., Itoh, X., Kimura, Y., Izawa,T., Shimamoto, K. & Toriyama, S. (1992).Genetically engineered rice resistant to rice stripevirus, an insect transmitted virus. Proceedings of theNational Academy of Science, USA, 89, 9865-9869.

Hayashimoto, A. & Murai, N. (1990). Apolyethyleneglycol-mediated protoplasttransformation system for production of fertiletransgenic rice plants. Plant Physiology, 93, 857-863.

Horn, M. E., Shillito, R. D., Conger, B. V. & Harms,C. T. (1988). Transgenic plants of orchardgrass(Dactylis glomerata L.) from protoplasts. Plant CellReports, 7, 469-472.

Izawa, T., Miyazaki, C , Yamamoto, M., Terada, R.,Iida, S. & Shimamoto, K. (1991). Introduction andtransposition of the maize transposable element Acin rice (Oryza sativa L.). Molecular and GeneralGenetics, 227, 391-396.

Katagiri, F. & Chua, N.-H. (1992). Plant transcriptionfactors: present knowledge and future challenges.Trends in Genetics, 8,. 22-27.

Keith, B. & Chua, N.-H. (1986). Monocot and dicotpre-mRNAs are processed with different efficienciesin transgenic tobacco. EMBO Journal, 5, 2419-2425.

Kinoshita, T. (1984). Gene analysis and linkage map.In Biology of Rice, ed. S. Tsunoda & M. Takahashi,pp. 187-274. Japan Scientific Society Press, Tokyo.

Kyozuka, J., Fujimoto, H., Izawa, T. & Shimamoto,K. (1991). Anaerobic induction and tissue-specificexpression of maize Adhl promoter in transgenic riceplants and their progeny. Molecular and GeneralGenetics, 228, 40-48.

Kyozuka, J., Hayashi, Y. & Shimamoto, K. (1987).High frequency plant regeneration from riceprotoplasts by novel nurse culture methods.Molecular and General Genetics, 206, 408-413.

Kyozuka, J., Izawa, T., Nakajima, M. & Shimamoto,K. (1990). Effect of the promoter and the firstintron of maize Adhl on foreign gene expression inrice. Maydica, 35, 353-357.

Kyozuka, J., Otoo, E. & Shimamoto, K. (1988). Plantregeneration from protoplasts of indica rice:genotypic differences in culture response. Theoreticaland Applied Genetics, 76, 887-890.

Kyozuka, J. & Shimamoto, K. (1991). Transformationand regeneration of rice protoplasts. Plant TissueCulture Manual, ed K. Lindsey, pp. 1-17. KluwerAcademic Publishers, Amsterdam.

Kyozuka, J., Shimamoto, K. & Ogura, H. (1989).Regeneration of plants from rice protoplasts.Biotechnology in Agriculture and Forestry, Vol. 8, PlantProtoplasts and Genetic Engineering I, ed. Y. P. S.Bajaj, pp. 109-203. Springer-Verlag, BerlinHeidelberg, New York.

Rice 63

Lamppa, G., Nagy, F. & Chua, N.-H. (1985). Light-regulated and organ-specific expression of wheatCab gene in transgenic tobacco. Nature (London),316, 750-752.

Lazzeri, P. A., Brettschneider, R., Luhrs, R. & Lorz,H. (1991). Stable transformation of barley via PEG-induced direct DNA uptake into protoplasts.Theoretical and Applied Genetics, 81, 437-444.

Lloyd, J. C , Raines, C. A., John, U. P. & Dyer, T. A.(1991). The chloroplast FBPase gene of wheat:structure and expression of the promoter inphotosynthetic and meristematic cells of transgenictobacco plants. Molecular and General Genetics, 225,209-216.

Luehrsen, K. R. & Walbot, V. (1991). Intronenhancement of gene expression and the splicingefficiency of introns in maize cells. Molecular andGeneral Genetics, 225, 81-93.

Mariani, C , De Beuckeleer, M., Truettner, J.,Leemans, J. & Goldberg, R. B. (1990). Induction ofmale sterility in plants by a chimaeric ribonucleasegene. Nature (London), 347, 737-741.

Mascarenhas, D., Mettler, I. J., Pierce, D. A. & Lowe,H. W. (1990). Intron-mediated enhancement ofheterologous gene expression in maize. PlantMolecular Biology, 15, 913-920.

Matsuki, R., Onodera, H., Yamauchi, T. & Uchimlya,H. (1989). Tissue-specific expression of the rolCpromoter of the Ri plasmid in transgenic rice plants,Molecular and General Genetics, 220, 12-16.

Matsuoka, M. & Sanada, Y. (1991). Expression ofphotosynthetic genes from the C4 plant, maize, intobacco. Molecular and General Genetics, 225,411-419.

McCouch, R., Kochert, G., Yu, Z. H., Wang, Z. Y.,Khush, G. S., Coffinan, W. R. & Tanksley, S. D.(1988). Molecular mapping of rice chromosomes.Theoretical and Applied Genetics, 76, 815-829.

McElroy, D., Zhang, W., Cao, J. & Wu, R. (1990).Isolation of an efficient actin promoter for use in ricetransformation. Plant Cell, 2. 163-171.

Meijer, E. G. M., Schilperoort, R. A., Rueb, S., Os-Ruygrok, P. E. & Hensgens, L. A. M. (1991).Transgenic rice cell lines and plants: expression oftransferred chimeric genes. Plant Molecular Biology,16, 807-820.

Meyerowitz, E. M. (1989). Arabidopsis, a useful weed.Cell, 56, 263-269.

Murai, N., Li, Z., Kawagoe, Y. & Hayashimoto, A.(1991). Transposition of the maize Activator elementin transgenic rice plants. Nucleic Acids Research, 19,617-622.

Nishibayashi, S. (1992). Is genome of rice small? RiceGenetics Newsletters, 8, 152-154.

Oard, J. H., Paige, D. & Dvorak, J. (1989). Chimericgene expression using maize intron in cultured cellsofbreadwheat. Plant Cell Reports, 8, 156-160.

Ohta, S., Mita, S., Hattori, T. & Nakamura, K.(1990). Construction and expression in tobacco ofP-glucuronidase (GUS) reporter gene containing theintron within the coding sequence. Plant CellPhysiology, 31, 805-815.

Paszkowski, J., Baur, M., Bogucki, A. & Potrykus, I.(1988). Gene targeting in plants. EMBO Journal, 1,4021-4026.

Peng, J., Lyznik, L. A., Lee, L. & Hodges, T. (1990).Transformation of indica rice protoplasts with gusAand neo genes. Plant Cell Reports, 9, 168-172.

Perlak, F. J., Deaton, R. W., Armstrong, T. A., Fuchs,R. L., Sims, S., Greenplate, J. T. & Fischhoff, D. A.(1990). Insect resistant cotton plants.Bio/Technology, 8, 939-943.

Powell-Abel, P., Nelson, R. S., De, B., Hoffinann, N.,Rogers, S. G., Fraley, R. T. & Beachy, R. N.(1986). Delay of disease development in transgenicplants that express the tobacco mosaic virus coatprotein gene. Science, 232, 738-743.

Rhodes, C. A., Pierce, D. A., Mettler, I. J.,Mascarenhas, D. & Detmer, J. J. (1988). Geneticallytransformed maize plants from protoplasts. Science,240, 204-207.

Schell, J. (1987). Transgenic plants as tools to studythe molecular organization of plant genes. Science,237, 1176-1183.

Schernthaner, J. P., Matzke, M. A. & Matzke, A. J. M.(1988). Endosperm-specific activity of a zein genepromoter in transgenic tobacco plants. EMBOJournal, 7, 1249-1255.

Shimamoto, K. (1991). Transgenic rice plants. InMolecular Approaches to Crop Improvement, ed. E. S.Dennis & D. J. Llewellyn, pp. 1-15. Springer, Wien.

Shimamoto, K., Terada, R., Izawa, T. & Fujimoto, H.(1989). Fertile transgenic rice plants regeneratedfrom transformed protoplasts. Nature (London), 338,274-276.

Tada, Y., Sakamoto, M. & Fujimura, T. (1990).Efficient gene introduction into rice byelectroporation and analysis of transgenic plants: useof electroporation buffer lacking chloride ions.Theoretical and Applied Genetics, 80, 475-480.

Tada, Y., Sakamoto, M., Matsuoka, M. & Fujimura,T. (1991). Expression of a monocot LHCPpromoter in transgenic rice. EMBO Journal, 10,1803-1808.

Tanaka, A., Mita, S., Ohta, S., Kyozuka, J.,Shimamoto, K. & Nakamura, K. (1990).Enhancement of foreign gene expression by a dicotintron in rice but not in tobacco is correlated with anincreased level of mRNA and an efficient splicing ofthe intron. Nucleic Acids Research, 18, 6767-6770.

Terada, R. & Shimamoto, K. (1990). Expression ofCaMV 35S-GUS gene in transgenic rice plants.Molecular and General Genetics, 220, 389-392.

Toriyama, K., Arimoto, Y., Uchimiya, H. & Hinata,


K. (1988). Transgenic rice plants after direct genetransfer into protoplasts. Bio/Technology, 6,1072-1074.

Toriyama, K., Hinata, K. & Sasaki, T. (1986).Haploid and diploid plant regeneration fromprotoplasts of anther callus in rice. Theoretical andApplied Genetics, 73, 6-19.

Vaeck, M., Reynaerts, A., Hofte, H., Jensens, S., DeBeuckeleer, M., Dean, C , Zabeau, M., VanMontagu, M. & Leemans, J. (1987). Transgenicplants protected from insect attack. Nature {London),328, 33-37.

Vasil, V., Clancy, M., Ferl, R. J., Vasil, I. K. &Hannah, L. C. (1989). Increased gene expression bythe first intron of maize Shrunken-1 locus in grassspecies. Plant Physiology, 91, 1575-1579.

Wang, Z. Y. & Tanksley, S. D. (1989). Restrictionfragment length polymorphism in Oryza sativa L.Genome, 32, 1113-1118.

Wang, Z. Y., Wu, Z. L., Xing, Y. Y., Zheng, F. G.,Guo, X. I., Zhang, W. G. & Hong, M. M. (1990).Nucleotide sequence of rice waxy gene. Nucleic AcidsResearch, 18, 5898-5899.

Wienand, U. & Saedler, H. (1988). Plant transposableelements: unique structures for gene tagging andgene cloning. In Plant DNA Infectious Agents, ed. T.Hohn & J. Schell, pp. 205-228. Springer, Wien.

Xie, D. X., Fan, Y. L. & Ni, P. C. (1990). Transgenic

rice plant obtained by transferring the Bacillusthuringiensis toxin gene into a Chinese rice cultivarZhonghua 11. Rice Genetics Newsletters, 1, 147-148.

Yamada, Y., Yang, Z. Q. & Tang, D. T. (1986). Plantregeneration from protoplast-derived callus of rice{Oryza sativa L.). Plant Cell Reports, 5, 85-88.

Yang, H., Guo, S. D., Li, J. X., Chen, X. J. & Fan,Y. L. (1989). Transgenic rice plants produced fromprotoplasts following direct uptake of Bacillusthuringiensis 8-endotoxin protein gene. Rice GeneticsNewsletters, 6, 159-160.

Zhang, W., McElroy, D. & Wu, R. (1991). Analysis ofrice Actl 5' region activity in transgenic rice plants.Plant Cell, 3, 1155-1165.

Zhang, W. & Wu, R. (1988). Efficient regeneration oftransgenic rice plants from protoplasts and correctlyregulated expression of foreign genes in plants.Theoretical and Applied Genetics, 76, 835-840.

Zhang, H. M., Yang, H., Reach, E. L., Golds, T. J.,Davis, A. S., Mulligan, B. J., Cocking, E. C. &Davey, M. R. (1988). Transgenic rice plantsproduced by electroporation-mediated plasmiduptake into protoplasts. Plant Cell Reports, 1,379-383.

Zhao, X., Wu, T., Xie, Y. & Wu, R. (1989). Genome-specific repetitive sequences in the genus Oryza.Theoretical and Applied Genetics, 78, 201-209.

6Maize

H. Martin Wilson, W. Paul Bullock, Jim M. Dunwell, J. Ray Ellis*,Bronwyn Frame, James Register III and John A. Thompson

Introduction

Genes advantageous for pathogen or insect resis-tance, or for enhancing yield, are continuouslybeing sought by plant breeders. Many of thesegenes exist outside the species targeted forimprovement. This fact has led researchers tocontemplate ways of overcoming the biologicalbarriers to gene transfer. More than a quarter of acentury has passed since the first attempt wasmade to introduce DNA into maize (Zea mays)through physical intervention (Coe & Sarkar,1966). Although this attempt, which involveddirect injection of DNA into apical meristems ofseeds, was not successful, it did serve to identify aprincipal problem - namely that 'The cell wall, amassive barrier to large structures, may have to bedisrupted mechanically or chemically, or other-wise circumvented'. Another observation made in1966 by Coe & Sarkar was that 'the numberof meristem cells actually penetrated andobserved may be too small to have included afortuitous, observable transformation'. Overtwo decades later the efficient combination ofDNA delivery and transformed cell selectionhas proved to be the key to production of trans-genic plants of all species and particularly ofmaize, which as a cereal of major commercialimportance has attracted substantial attention inrecent years.

Maize transformation, defined as the produc-tion of transgenic plants that are fertile and pro-duce transgenic seed, was achieved for the first

* We dedicate this review to the memory of our colleagueRay Ellis, who died in April, 1993.

time in 1990 (Fromm et al., 1990; Gordon-Kammet aly 1990). Since then a number of commercialresearch groups have conducted field trials oftransgenic maize, demonstrating the attainment ofmore widespread success.

In this review we briefly consider some of theapproaches that have produced interesting prelim-inary results - Agrobacterium-mediated DNA deliv-ery, whisker-mediated DNA delivery, germ-linetransformation targets (pollen, ovules andembryos), and microinjection - and then concen-trate on the three currently most successfulapproaches, namely direct gene transfer to proto-plasts, particle bombardment of regenerable cellcultures, and, most recently, electroporation ofimmature embryos.

Interesting preliminary results

Agrobacterium-mediated delivery

Agrobacterium tumefaciens is the most widely usedvector for transformation of dicotyledonous(dicot) species but its utility in maize transforma-tion remains to be demonstrated. Monocotyle-dons (monocots) are rarely natural hosts forAgrobacterium (De Cleene & De Ley, 1976; DeCleene, 1985) and therefore are not expected tobe susceptible to gene transfer mediated byAgrobacterium. This expectation has been chal-lenged, however, by the demonstration thatAsparagus officinalis (Liliaceae) could be trans-formed with DNA delivered by Agrobacterium toyield fertile transgenic plants (Hooykaas-VanSlogteren & Hooykaas, 1984; Bytebier et al,1987).

65

66 WILSON et al

Experimental inoculation of maize seedlingswith wild-type Agrobacterium strains led to claimsthat nopaline or octopine synthase activities,encoded by transferred DNA (T-DNA) from theTi plasmid, could be detected in seedling tissuesand in progeny obtained from the inoculatedplants (Graves & Goldman, 1986). However, thesynthesis of opines cannot be unequivocallyascribed to T-DNA transfer, and these results onmaize remain to be substantiated. More recently,a potent inhibitor of Agrobacterium virulenceinduction has been found in maize seedlings(Sahi, Chilton & Chilton, 1990).

A breakthrough in demonstrating the potentialof Agrobacterium as a gene vector for maize was theunambiguous transfer of maize streak virus DNAfrom Agrobacterium to inoculated plants by a pro-cedure commonly termed 'agroinfection'(Grimsley et al, 1987). To effect the transfer,tandemly repeated or partially repeated copies ofthe viral genome were inserted between the T-DNA borders of the vector plasmid and the shootmeristem of seedlings inoculated with the engi-neered Agrobacterium strain (Grimsley et al,1988). Transfer of the T-DNA was detected bythe presence of replicative forms of the viralgenome and the development of viral diseasesymptoms in the host plant. The success ofagroinfection in demonstrating T-DNA transferto maize results from the sensitive detection sys-tem, since the marker DNA becomes highlyamplified by viral replication and systemic spread.

The agroinfection technique has not producedstably transformed maize plants because the repli-cating viral genome is not transmitted to germ-line cells and does not become integrated into thenuclear genome. Attempts to detect T-DNAtransfer using sensitive but nonreplicating markergenes, e.g. uidA, have not been successful (N.Grimsley, personal communication), indicatingthat Agrobacterium delivers T-DNA to very fewinitial target cells. Despite these limitations, theagroinfection procedure has permitted analysis ofthe physical and biological factors necessary forAgrobacterium-mediated gene transfer to maizeand other cereals. For example, the nonrequire-ment of virulence inducers, the utility of binaryvectors, the relative efficiency of C58 strains andthe developmental competence of shoot meris-tems have been demonstrated (Grimsley et al.,1987; Boulton et al, 1989; Grimsley, 1990;Schlappi & Hohn, 1992). The inefficacy of

octopine strains to infect maize has been ascribedto the inhibitory action of the virF gene on the Tiplasmid (Jarchow, Grimsley & Hohn, 1991).

Integration of T-DNA into the genome of ricehas been reported (Ranieri et al, 1990). In addi-tion, the production of transgenic maize plantsfrom cultured shoot apices inoculated withAgrobacterium was reported by Gould et al. in1991. However, molecular and genetic evidencefor integration of T-DNA into the progenyremains inconclusive.

Agrobacterium clearly has potential as a genevector for maize, but the efficiency of thebacterium-plant interaction must be improved forthis approach to compete with direct gene trans-fer.

Germline transformation targets

Introduction of cloned genes directly into thegermline is an attractive goal for maize transfor-mation technology, since this would obviate theneed to regenerate from totipotent cells via tissueculture. However, germline transformation hasnot been convincingly demonstrated in maize, norin any other plant species.

From a practical point of view, pollen is themost amenable germline cell because of its abun-dance and accessibility. In the mature maizepollen grain (the male gametophyte) the germlineis represented by the two sperm cells that liewithin the vegetative cell and are surrounded bythe complex pollen wall. Although numerousattempts to transform maize pollen have beenpublished, and success has been claimed, in nocase is convincing molecular and genetic evidenceavailable. Most of these studies relied on poorlydefined morphological traits, encoded by genomicDNA, as genetic markers. However, use of well-defined molecular markers cloned on plasmidshas failed to produce transformants (Roeckel etal, 1988; Booy, Krens & Huizing, 1989) or failedto provide the crucial molecular and genetic evi-dence (De Wet et al, 1988).

Experimental procedures have generally beendesigned to permit DNA uptake into hydratingpollen grains or into the pollen tubes of germinat-ing grains by imbibition (De Wet et al, 1985;Sanford, Skubik & Reisch, 1985; De Wet et al,1986; Ohta, 1986; Waldron, 1987). Proven obs-tacles to DNA uptake include potent nucleaseactivity (Roeckel et al, 1988; Booy et al, 1989)

Maize 67

and pollen wall impermeability (Kranz & Lorz,1990). As an alternative, attempts have beenmade to introduce DNA into immature spikeletsof maize tassels in the hope of inducing geneuptake by developing microspores (Bennetzen etaly 1988). Efforts have also been made to targetthe growing pollen tube as a means of carryingmarker DNA along the silks to the embryo sac (G.Feix, personal communication; Langridge et al>1992). The demonstration of endophytic bacteriain maize plants has emphasized how readily arte-factual transformation results can be obtainedwhen nonaxenic target materials are used(Konstantinov, Mladenovic & Denic, 1991).

Delivery of marker genes directly to the egg cellhas rarely been attempted because of its inaccessi-bility within the ovary. Two approaches underinvestigation are injection of DNA into the prede-termined position of the egg cell in the ovule(Wagner, Dumas & Mogensen, 1990) or the enzy-matic isolation of the embryo sac and egg cell fol-lowed by gene delivery and fertilization in vitro(Kranz, Bautor & Lorz, 1991).

Embryos also have been investigated as recipi-ents for gene transfer by imbibition. Here, the tar-get cells are those in the shoot meristem that willultimately give rise to gametocytes. Both zygoticand somatic embryos have been used as DNArecipients, but no transgenic plants have been pro-duced (Topfer et al, 1989; Lupotto & Lusardi,1990). Nevertheless, transient expression ofmarker genes taken up by this approach has beenobtained using embryos isolated from a range ofcereal species (Topfer et al, 1989, 1990).

Despite the lack of success in targeting germlinecells for transformation to date, these cells arelikely to remain an option for experimentationbecause of their inherent potential benefits. Theavailability of marker genes that have been provento function in transgenic maize should now allowearlier claims of germline transformation to be rig-orously evaluated.

Microinjection

As an approach to maize transformation, micro-injection has received limited attention comparedto other delivery systems. The considerable tech-nical skill and expensive, sophisticated equipmentrequired have contributed to the slow progress inexperimentation. The main advantage of micro-injection over other delivery methods is the preci-

sion with which DNA can be delivered intodenned target cells (Neuhaus & Spangenberg,1990). While microinjection leads to high trans-formation frequencies in suitable single cell sys-tems (up to 20% in tobacco protoplasts; Schnorfet al> 1991), the rate of injection is slow, and thereare currently no single cell targets in maize thatcan be both efficiently injected and cultured at lowdensity.

Attention has therefore turned with the cerealsto multicellular systems as potential DNA recipi-ents via microinjection (Potrykus, 1990). Successin the production of transgenic Brassica napusplants following microinjection of microspore-derived embryos (Neuhaus et ah, 1987) promptedthe use of zygotic cereal embryos in transforma-tion studies. Zygotic embryos can be consideredan attractive target system for transformation,given the reduced effect of genotype, somaclonalvariation and culture-related infertility comparedto long-term cell cultures. However, it is clear thatan inevitable consequence of transforming struc-tures with many cells, such as zygotic embryos, isthe production of chimeras. Since the fate of indi-vidual cells in an embryo cannot currently bedescribed, it is necessary to analyze sexual off-spring for the presence of introduced genes.

In collaboration with Gunter Neuhaus at ETH,Zurich, we have performed an extensive evalua-tion of the microinjection of zygotic proembryosas a route to maize transformation. Culture condi-tions were established for proembryos of an elite,inbred maize genotype. Microinjection was per-formed into the meristematic region of theembryos using a DNA construct carrying the nptllgene. The embryos were then grown in a micro-culture system and the resulting plants transferredto soil. Analysis of these plants was restricted topolymerase chain reaction (PCR) analysis of tasseltissue. Three individual tassels out of over 100analyzed were identified and independently veri-fied as being PCR positive for the presence of thenptll gene. The progeny of a further 340 plantsderived from microinjected embryos were thenscreened for the presence and expression of thenptll marker gene, using PCR analysis andkanamycin selection. Despite the large number ofplants analyzed (>40 000), no evidence wasfound for transgene transmission to the progeny ofmicroinjected individuals.

In light of these data and the lack of any othercontradictory data in plants it has become clear

68 WILSON et al.

that the possibility exists that cells within zygoticembryos lack the competence for transformation,as discussed by Potrykus (1990). The difficulty oftargeting injections to germline progenitor cells inmaize embryos large enough to be cultured mayalso dramatically reduce the efficiency of transfor-mation. However, in view of the extensive screen-ing of injected material described above, it seemslikely there is a fundamental problem in effectingtransformation of meristematic cells. This issuerequires more basic research if further progress isto be made.

In vitro fertilized gametes offer an additionalopportunity for microinjection (Kranz et al,1991). However, the development of a transfor-mation system based on these target cells willrequire considerable further effort, and the degreeof technical sophistication needed will preventwidespread experimentation.

Whisker-mediated transformation

A simple, rapid and inexpensive means of del-ivering naked DNA to nonregenerable maizesuspension culture cells (cv. Black Mexican Sweet- BMS) without the need for sophisticated equip-ment has been demonstrated by Kaeppler et al.(1990). The method involves vortex treatment ofsuspension cells with silicon carbide whiskers andplasmid DNA. The rapid mixing resulting fromthis treatment appears to lead to whisker penetra-tion of cells and DNA delivery. BMS cells werefound to produce transient (3-glucuronidase(GUS) activity at a frequency of 140 blue color-forming units per 250 JULI (packed cell volume) ofcells.

More recently, stable transformation of BMSusing this approach has been reported by the samegroup (Kaeppler et al., 1992). Cells were treatedwith a plasmid carrying the bar and uidA genes(pBARGUS) and selected on an herbicide (activeingredient phosphinothricin). An average of 40GUS-expressing units and 3.4 stably transformedcallus lines were recovered from a 300 |UL1 packedcell volume of BMS cells.

Approaches proven to generatetransformed maize plants

Three approaches have proved successful to-dateand will be considered in detail below. These are

(i) direct gene transfer to protoplasts, (ii) particlebombardment and (iii) electroporation of imma-ture zygotic embryos and type 1 callus. In com-mon to all three are the use of target cells fromwhich whole plants can be regenerated and thedirect delivery of DNA to these cells. To offer util-ity in a transformation system, transformed cul-tured cells must be capable of periods of sustainedgrowth and division (for selection), and fullexpression of totipotency (for regeneration of fer-tile plants). This has proved an elusive formula inmaize (and other cereals) and has delayed successfor several years.

Direct gene transfer to protoplasts

Isolated protoplasts offer advantages over otherrecipient cell systems for DNA delivery since,using protoplasts, it is possible to define preciseconditions for efficient DNA uptake into a largepopulation of cells (Negrutiu et al., 1987).Procedures for direct gene transfer involving poly-ethylene glycol (PEG) treatment or electropora-tion are well established for many species andhave been used with a range of cereals, includingmaize. In addition, there is speculation that theprocess of enzymatic protoplast isolation mightconstitute the trigger in cereals for something akinto a 'wound response', thus increasing cell compe-tence for transformation and regeneration(Potrykus, 1990). Multicellular structures result-ing from protoplast culture are generally of single-cell origin (Thompson, Abdullah & Cocking,1986), so the generation of chimeras is avoided.

Although protoplasts are attractive as a trans-formation target, maize protoplast culture remainstechnically difficult. As a result, slow progress hasbeen made in the production of transgenic maizeplants from protoplasts.

Protoplasts isolated directly from maize planttissue are, as with other cereal species, extremelyrecalcitrant in culture. Numerous attempts toinduce sustained cell division in such protoplastshave been made without success (Potrykus,Harms & Lorz, 1976). At best, cereal leaf proto-plasts can be induced to undergo only a few celldivsions (Hahne, Lorz & Hahne, 1990). However,protoplasts of maize leaf or root origin are of valuein transient expression experiments where infor-mation on regulation of promoter function may beobtained (Junker et al., 1987).

Protoplasts from nonregenerable BMS suspen-

Maize 69

sion cultures can be induced to undergo sustainedcell division (Chourey & Zurawski, 1981).However, the plating efficiency of such culturesremains low (at best 10%) compared to dicotspecies, despite improvements in division fre-quency obtained through, for example, the use ofnurse cultures (Ludwig et al., 1985) or condi-tioned medium (Somers et al., 1987).

Stable transformation of BMS protoplasts hasbeen achieved following either electroporation(Fromm, Taylor & Walbot, 1986) or PEG-mediated DNA uptake (Armstrong et ah, 1990).Information on the cotransformation of un-linked genes has been obtained in BMS usingnptll/uidA (Lyznik et al, 1989), nptll/aphTV(Armstrong et al., 1990) and nptll/cat (ICISeeds; J. A. Thompson et al., unpublished data).Cotransformation occurs at frequencies of40% - 80%, as described for other species (Saul &Potrykus, 1990). Cationic methods have also beenused to transform maize protoplasts. Kanamycin-resistant BMS transformants were recovered afterprotoplast transfection with genomic DNA from aline previously transformed with the nptll gene(Antonelli & Stadler, 1990).

Protoplasts isolated from cell lines such asBMS, and from those of endosperm origin, havealso been used widely in studies of transient geneexpression and provide useful information on con-struct function, for example intron and enhancereffects (Callis, Fromm & Walbot, 1987; Last etal., 1991; Quayle & Feix, 1992).

Embryogenic maize cultures offer the best pos-sibility for the recovery of plants from protoplasts.The development of embryogenic suspension cul-tures capable of yielding division-competent cellsis, however, a time-consuming and poorly under-stood process. Rhodes, Lowe & Ruby (1988a)succeeded in regenerating plants of A188 and B73from embryogenic suspension-derived protoplastsof immature embryo origin, using feeder layers tostimulate protoplast division. Unfortunately allthe regenerated plants were sterile, probably as aconsequence of the age of the cultures used forprotoplast isolation (from 18 months to 4 years).Transgenic plants of A188 were produced by thesame group (Rhodes et al., 19886) following elec-troporation and kanamycin selection. Thirty-eightplants were regenerated from ten different celllines. However, these plants were also sterile.

Several attempts have been made to develop anefficient system for the production of fertile plants

from immature embryo-derived embryogenicmaize suspension protoplasts that would be usefulfor transformation studies. Shillito et al. (1989)succeeded in regenerating fertile plants fromembryogenic suspension protoplasts of a B73-derived inbred line despite the occurrence ofmany morphological and reproductive abnormali-ties. Again, feeder layers of nurse suspension cellswere used to enhance protoplast division fre-quency in this work and in that of Prioli &Sondahl (1989), where fertile plants were regener-ated from protoplasts of a Cateto inbred lineadapted to tropical conditions. A wide range ofmorphological abnormalities were also observedin the regenerants described in this report.

Complex interactions between protoplast andfeeder cell lines were noted by Petersen, Sulc &Armstrong (1992), where plants with only tassels,or plants with only ears, were regenerated fromprotoplasts of crosses involving A188, B73 andBMS. Protoplast-derived regenerants were usedas embryo donors to develop lines with improvedprotoplast response.

Microspore-derived cultures have also beenused to develop embryogenic suspension culturesfrom which regenerable protoplasts were isolated(Sun, Prioli & Sondahl, 1989; Mitchell & Petolino,1991). No fertile dihaploid plants were describedin these reports and again a range of phenotypicvariants were noted in the regenerated plants.

In contrast, Morocz et al. (1990) found rela-tively low frequencies of developmental abnor-malities in the plants regenerated fromembryogenic suspension protoplasts of the com-plex synthetic maize genotype He/89. An auxin-autotrophic culture (patent applied for, EPapplication 0465875A1) derived from this geno-type was shown to yield division-competent proto-plasts as early as 10 weeks after culture initiation.More than 500 plants were obtained from theresulting callus and of these 60%-70% set seedwhen selfed or sibbed. The use of young culturesfrom a tissue culture-adapted genotype undoubt-edly contributes to the high plant regeneration fre-quency and fertility seen in this work.

This efficient regeneration system was success-fully used to achieve stable transformation follow-ing PEG-mediated DNA uptake and subsequentselection on phosphinothricin (Donn, Nilges &Morocz, 1990). Plants were regenerated from theresulting transformed callus, with the majoritybeing fertile. This therefore represents the first

70 WILSON et al.

report of fertile transgenic plant production frommaize protoplasts.

It remains to be seen whether elements of theserecent reports (Donn et al, 1990; Morocz et al,1990) can be used to bring more general successin the production of transgenic maize plants fromprotoplasts. At present, genotype constraints andthe reduced vigor and fertility of plants regener-ated from protoplasts probably outweigh thebenefits of protoplasts as recipients for the inte-gration of foreign DNA. Transient expression inprotoplasts will continue to provide valuableinformation on the functional analysis of genes.

Particle bombardment

The use of high velocity microprojectiles to carryDNA into living cells is now routine in many labo-ratories following the pioneering work of Sanford,Wolf, Klein and Allen at Cornell University(Klein et al, 1987; Sanford et al, 1987).Transient expression of an introduced gene wasfirst reported in maize in 1988 by Klein et alStable transformation of cultured maize cells(Klein et al, 1989) and regeneration of fertiletransgenic maize plants from transformed cul-tured cells (Fromm et al, 1990; Gordon-Kamm etal, 1990) were subsequently reported. Particlebombardment is currently used in a number ofcommercial laboratories, including our own, togenerate transformed maize plants. Plates 6.1-6.9illustrate various stages in particle bombardment-mediated maize transformation.

Early successes with maize (e.g. see Frommet al, 1990; Gordon-Kamm et al, 1990) wereachieved using DNA coated on to tungsten parti-cles that were accelerated by means of gunpowder-driven macroprojectiles or macrocarriers. Morerecent particle bombardment work (Russell, Roy& Sanford, \992a,b; Sanford, Smith & Russell,1993) has shown that acceleration of DNA-coatedgold particles driven by helium gas provides amore efficient means for generation of transientand stable transformants in a number of biologicalsystems. The use of gold particles and a helium-driven device (the Dupont PDS-1000) is now rou-tine in many laboratories, including most of thoseworking in the area of maize transformation.

A number of different protocols detailing meth-ods of introducing genes into regenerable corncells through particle bombardment have beendescribed (Fromm et al, 1990; Gordon-Kamm et

al, 1990). The protocol used at ICI Seeds is asfollows. Gold particles (3 mg of 1 |mm-sized parti-cles) are washed and coated with plasmid DNA(5 ^g) as described in the Dupont PDS-1000manual. Ten microliters of gold particle/DNAsuspension are then loaded on to each macro-carrier and thoroughly dried in a desiccator priorto bombardment. A fine wire mesh (150 |xm)screen (after Gordon-Kamm et al (1990), whoused a 100 |xm screen), placed between themacrocarrier holder and the target tissue, hasbeen found to increase gold particle dispersionand reduce tissue death. Suspension cells (2 dayspost-subculture), or unorganized callus tissue, arevacuum filtered on to filter paper (Whatman no.4) discs (0.25 ml PCV/filter), cultured overnighton the surface of N6 medium (suspension cells)or MS medium (callus), and then bombarded(900 p.s.i., 1/16 gap distance) the next day.Transformed clone selection is accomplished byincluding 1 mg bialaphos/1 in the culture medium.An osmotic pretreatment of unorganized embryo-genie callus (culture on 0.5 M osmoticum for 30min) just prior to bombardment can enhance bothtransient expression and stably transformed clonerecovery (Russell etal, 19926).

DNA-coated particles can be introduced intovirtually any plant cell accessible to bombard-ment. This fact has led to consideration of whatconstitutes a suitable target cell for the generationof fertile transformed plants. To date, only cellscapable of giving rise to cultures from whichplants can be regenerated have proved useful forthis purpose. Claims purporting to represent sta-ble transformation of differentiated cells with sub-sequently normal development have not beensubstantiated (in any plant species). Target cellsmost widely used in corn are embryogenic cellsuspension cultures (e.g. see Gordon-Kamm etal, 1990) and callus (e.g. see Fromm etal, 1990).

The great majority of cells in target tissue donot receive DNA during bombardment. Sanford(1990) has estimated a rate of one transient trans-formant per 1000 to 10 000 bombarded targetcells, and rates of stable transformation at 2%-5%of the rate of transient transformation. These esti-mates suggest that stable transformation, on aver-age, occurs at a rate of one event per hundreds ofthousands, if not millions, of bombarded cells.The need is therefore plain for a means to selectout the transformed cells.

Kanamycin selection has been used to select out

Maize 71

maize transformants following particle bombard-ment with the nptll gene (J. Register, unpublisheddata) .There does, however, appear to be endoge-nous resistance to the antibiotic, and growth ofnontransformed tissues ('escapes') is common.Similarly, attempts to use the selective agentChlorsulfuron, and a mutant form of the ah genehave been successful (Fromm et aL, 1990) butagain escapes frequently occur (ICI Seeds,unpublished data). Hygromycin has also beenused as a selective agent following bombardmentwith the gene for hygromycin phosphotransferase(Walters et al, 1992). Here, growth on mediacontaining hygromycin first at 15 mg/1 and then at60 mg/1 was required for more than 9 weeksbefore sectors unequivocally resistant to theaminoglycoside could be identified. Recently,glyphosate has been used successfully as a selec-tive agent in conjunction with a mutant enolpyru-vyl shikimate-3-phosphate (EPSP) synthase gene(report from Monsanto, 1992 World Congress onCell and Tissue culture, Washington, DC). Todate, however, the selection system most widelyused following particle bombardment-mediatedtransformation is expression of the bar gene (iso-lated from Streptomyces hygroscopicus; Thompsonet al.y 1987) conferring resistance to phos-phinothricin (PPT). PPT is an analog of gluta-mine that inhibits the amino acid biosyntheticenzyme glutamine synthase of bacteria and plants.Bialaphos is a tripeptide precursor of PPT pro-duced by some strains of Streptomyces, in whichtwo alanine residues are linked to the PPT moiety;the active PPT moiety is released intracellularly bypeptidase activity (see Mazur & Falco, 1989).Although PPT has been used successfully as theselective agent in corn protoplast systems (Donnet al., 1990; and see above), attempts to select onPPT following particle bombardment-mediatedtransformation with the bar gene have generallyfailed (ICI Seeds, unpublished data). However,for reasons that are not understood, if bialaphos,in the range 1-5 mg/1, is used instead of PPT, veryeffective selection for embryogenic suspensioncells transformed with the bar gene can beachieved (Gordon-Kamm et al., 1990; Spencer etal., 1990). Escapes with the bialaphos/6<zr systemfollowing particle bombardment of embryogenicsuspension cells are rare.

There are two areas where limitations in effi-ciency are encountered with bombardment ofembryogenic cell cultures. First, the number of

independent transgenic events from which plantscan be regenerated is relatively low, and, second,transgenic regenerants often show poor fertility.

It is possible to produce large numbers (600 to800) of stably transformed cell lines from singlebombardments of certain plant cell cultures (e.g.NT1 of tobacco; Russell et aL, 19926). However,with embryogenic corn cultures the average fre-quency of stable transgenic clone generation iscloser to one or two per bombardment in most ofthe successful experiments where numbers havebeen reported (e.g. Gordon-Kamm et al., 1991).This difference may relate to the nature of growthin embryogenic cell cultures compared to that inthe populations of relatively homogeneous cellsfound in friable tobacco cultures such as NT1.Friable embryogenic maize cultures can be char-acterized as aggregates of cells organized intoproembryonic structures at various stages ofdevelopment. Division to generate new structuresoccurs from a subset of these cells; other cellswithin structures divide in the course of organizedgrowth. The net result is that the number of cellsfrom which stably transformed, selectablecolonies can arise post-bombardment is relativelysmall. In the case of a culture such as NT1tobacco, nearly all cells are capable of giving riseto stably transformed colonies.

Regeneration is not possible from all stablytransformed clones and when plants can be pro-duced they are not always fertile. Where numbersare given, it is clear that seed production on trans-formed regenerants is very low, with Gordon-Kamm et al (1991) reporting a total of 176 seedsfrom 76 plants of one cell line and a total of 268seeds from 219 plants of another cell line. Cornplants grown from seed in the greenhouse typi-cally set between 300 and 400 seeds per plant(depending, of course, on genotype). The cause oflow fertility in transgenic regenerants does notappear to be related in any general way to thepresence of transgenes because in subsequent gen-erations transgenic plants usually show normalfertility (Gordon-Kamm et al, 1991; ICI Seeds,unpublished data). In addition, transgenic regen-erants often appear 'stressed', i.e. they display pre-mature senescence, poor tassel and eardevelopment and stunted growth, whereas prog-eny of these plants display none of these symp-toms and appear normal in all respects.

The causes of poor fertility and phenotypicalabnormality in transgenic regenerants are

72 WILSON et al.

unknown. These problems are encountered tovarying degrees with regenerated corn plants ingeneral and are thought to be related to the cultureprocess. The phenomenon appears more markedin plants recovered from selection in vitro. Keepingthe periods spent in culture and under selection toa minimum may be the most practical short-termsolution to overcoming these problems.

We have shown (ICI Seeds, unpublished data)that improvements in the efficiency with whichtransgenic clones can be produced are oftenaccompanied by a loss of regenerability, or by aloss in fertility of the plants that are produced.Deterioration in culture productivity in terms offertile transgenic plant generation has also beensuggested to occur as a consequence of culture age(Fromm et al, 1990). One practical approach toovercome this problem is to maintain a bank ofcryopreserved, productive cultures from whichtarget tissue can be withdrawn when required.

Substantial progress will be required in under-standing the control and dynamics of embryo-genie culture development and plant regenerationbefore these culture systems can be fully exploitedin the generation of fertile transgenic plantsthrough particle bombardment. There has, unfor-tunately, been little fundamental work aimed ataddressing the inefficiencies of culture initiationand fertile plant regeneration in cereal species todate.

Wounding and electroporation of immaturezygotic embryos and type 1 callus

Recently, D'Halluin et al (1992) described anelectroporation method for the production oftransgenic maize plants. This method utilizes astarget tissue immature zygotic embryos of a sizefrom which callus can be readily initiated(1.0-1.5 mm, typically about 10-14 days' post-pollination) or finely chopped type 1 callus. Thefreshly isolated embryos are treated for 1-2 minwith an enzyme solution (0.3% (w/v) macero-zyme, salts, buffer and 10% (w/v) mannitol) andthen electroporated (after Dekeyser et al, 1990) inthe presence of linearized DNA. Following selec-tion on kanamycin-containing medium, plants areregenerated from actively growing embryogenictissue.

The authors of this method claim that wound-ing and/or degrading of intact tissue is an impor-tant prerequisite for success with DNA uptake.

This wounding may be achieved by physicalmeans, e.g. cutting, or by chemical means such asan enzyme treatment for 1-2 min. Earlier workshowed that electroporation could be used tointroduce DNA into cells of intact rice leaf bases(Dekeyser et al, 1990). Successful uptake wasconfirmed by transient gene expression assays. Inthis latter study, wounding - other than the dis-secting out of the leaf bases - was not implicated.

In one example of the method, corn of thegenotype H99 was transformed with a plasmidcontaining an nptll cassette and the barnase gene(Hartley, 1988) under the control of the tapetal-specific promoter TA29 from tobacco. It has beenshown that expression of the barnase gene intapetal tissue of tobacco (Nicotiana tabacum) andoilseed rape (Brassica napus) results in failure ofnormal anther development and, as a conse-quence, male sterility. The transgenic corn plantsproduced by the method of D'Halluin et al.(1992) were male sterile, but otherwise phenotyp-ically normal. Progeny from these plants showed a1:1 segregation for the nptll gene. The nptll nega-tive progeny were male fertile, while the nptll pos-itive progeny were male sterile. The barnase genewas detected by Southern analysis in all of themale sterile plants.

Further work to understand more fully themechanism for DNA uptake into maize cellsmediated by electroporation is clearly required.Nevertheless, the approach has yielded fertile,transgenic maize plants and has the promise ofproviding a means for testing transgenes with onlythe briefest exposure to growth in vitro.

Applicability of successful approaches tocommercial germplasm

Reliance on a culture phase limits application ofthe above approaches to those genotypes fromwhich the requisite response can be obtained. Inthe case of PEG-mediated DNA uptake intoprotoplasts, success has been restricted to onecomplex tissue culture-adapted genotype.Bombardment efforts to date have focused on theuse of friable, embryogenic target tissue devel-oped, for the most part, from immature Fl hybridembryos of the genotype A188 X B73 (immatureembryos dissected out of A188 plants pollinatedwith B73 pollen). Claims of success with elitemaize inbreds have been made (stiff stalks of the

Maize 73

B73 type) but until very recently there had beenno published reports. Koziel et al (1993), how-ever, have reported the introduction of a syntheticgene coding for an insecticidal protein into an eliteinbred of Lancaster parentage. This was achievedby bombardment of immature zygotic embryosfollowed by culture initiation, selection ofembryogenic callus and plant regeneration.Success following electroporation of immaturezygotic embryos or type 1 callus has been achievedonly with H99 and Pa91, two maize inbred linesknown to be highly responsive in culture initiationexperiments.

Transgenes can be introduced into any maizeline from transgenic culture-responsive lines bybackcrossing but clearly, from a commercial pointof view, direct transformation of elite inbreds hasmajor advantages. Transgene expression andagronomic performance can be assessed immedi-ately, backcrossing is not required, and productdevelopment time is reduced. Several approachesare being employed in attempts to overcomegenotype restrictions. Media improvements(Duncan et al, 1985), and improved culture sys-tems (selection of a certain cell type, culture main-tenance protocols; Shillito et al, 1989), have beentested with some success. An advance more read-ily reproducible and effective appears to be that ofVain, Flament & Soudain (1989a) and Vain, Yean& Flament (19896), who showed that silver nitrateand/or aminoethoxyvinyl-glycine (AVG) couldenhance friable embryogenic response from cul-tured immature embryos of A188. Songstad,Armstrong & Petersen (1991) subsequentlyshowed that enhanced rates of response could beobtained with B73 and derivatives. This result hasalso been observed at ICI Seeds (W. P. Bullock etah, unpublished data). Further exploitation of theencouraging lead identified by Vain and col-leagues may result in success with other elitemaize genotypes.

A different approach has been taken byArmstrong, Romero-Sevenson & Hodges (1992),who have attempted to identify the genetic com-ponents in A188 for culture responsiveness andintrogress it into elite germplasm. However, froma practical point of view it would appear to bequicker and easier to introgress a transgene fromA188 X B73 germplasm into elite lines than tointrogress a complex trait such as culture respon-siveness into elite lines as they are continuallybeing developed by breeders.

Molecular characterization of transgenicmaizeTransgene integration

The methods that have been used successfully totransform maize all employ direct DNA transfer.In general, transgene integration in maize(Spencer et al., 1992) follows the trends observedwith direct DNA delivery in other plant species(Riggs & Bates, 1986; Schocher et al, 1986;Klein et al, 1988; Bellini et al, 1989; Tomes et al,1990). These trends are illustrated in Table 6.1,which shows a summary of results obtained in ourlaboratory from Southern analysis of over 100independently transformed callus lines producedby particle bombardment of embryogenicA188 X B73 suspension cells (J. Register, un-published data). Here, the selectable and nonse-lectable genes were on the same introducedplasmid.

The data in Table 6.1 illustrate several points.First, the absence of selection for gene expressioncorrelates with a greater incidence of rearrange-ments (indicated by the presence of fragments ofunexpected length) and deletions (nonselectablegene versus selectable gene). Second, integrationof multiple gene copies (more than five) occurswith a frequency approaching 40%, although inaround 90% of these cases the transgenes appearto be integrated at a single site (data not shown).This is a frequently observed phenomenon withintroduction of transgenes through direct DNAdelivery. Third, similar copy numbers of intactselectable and nonselectable genes were present inindividual transformants.

At the whole plant level, transgene transmissionhas been studied through meiosis (Fromm et al,1990; Gordon-Kamm etal, 1990, 1991; Spenceretal, 1992; Walters etal, 1992). Generally, trans-genes are transmitted with fidelity from the pri-mary transformed cell through to progeny oftransgenic plants. There are, however, exceptions.Some transgene instability through meiosis wasreported by Potrykus et al (1985), followingdirect DNA delivery in tobacco. Similarly, incorn, transformants have been noted where trans-gene transmission has apparently occurred onlythrough female gametes (Register et al, 1992;Spencer etal, 1992; Walters et al, 1992).

As shown in Table 6.1, direct gene transfertends to introduce multiple gene copies into cellsand these gene copies often display complex inte-

74 WILSON et al

Table 6.1. Characterization ofDNA integration after particle bombardment-mediated direct DNA delivery to embryogenic maize cells

Frequency of integration (%)

DNA rearrangementsUnrearranged gene copiesRearranged gene copiesUnrearranged and rearranged gene copiesDeleted genes

Copy number estimates1-45-10>10

Selectable gene(npt\\)

547

39—

592219

Nonselectable gene(uidA)

27194212

622216

Notes:Copy numbers were estimated from reconstructions and are expressed per haploidgenome.

gration patterns (e.g. see Bellini et al., 1989;Tomes et al, 1990; Spencer et al, 1992).Although direct gene transfer and Agrobacterium-mediated transformation have been comparedhead-to-head only once (Czernilofsky et al,19S6a,b), it is clear that complex integration pat-terns occur more frequently following direct genetransfer. It remains to be shown definitivelywhether these integration characteristics have anyfunctional significance but indirect evidenceindicates that transgene instability may also begreater following direct gene transfer (for exam-ple, compare Chyi et al, 1986; Muller et al, 1987;and Heberle-Bors et al, 1988; with Bellini et al,1989; Tomes et al, 1990; and Spencer et al,1992).

Transgene expression

Study of transgene expression in primary maizetransformants indicates that transgenes (driven bythe 35S promoter) that confer resistance to aselective agent at the callus stage usually express atthe whole plant level (e.g. see Gordon-Kamm etal, 1991). Expression of nonselectable transgenes(e.g. uidA) introduced either on the same plasmidas the selectable marker gene (Fromm et al, 1990;Register et al, 1992; ICI Seeds, unpublished data)or on a separate plasmid (Gordon-Kamm et al,1990) is typically more variable. We haveobserved many cases in which expression of an

apparently intact nonselectable gene occurs infewer plants (regenerated from a single callus)than the selectable gene (Register et al, 1992; ICISeeds, unpublished data).

Inheritance of transgene expression has beenstudied for the most part in transgenic individualsproduced by Agrobacterium-mediated transforma-tion. The first study of transgene expression inplants transformed via direct DNA delivery wasreported in 1985 by Potrykus et al These authorsnoted that kanamycin-resistant tobacco transfor-mants gave progeny that showed - with few excep-tions - the expected segregation ratios for a singledominant trait. Later studies confirmed this resultand also documented occasional ratios consistentwith segregation of multiple loci and loss of geneexpression (Morota & Uchimiya, 1988; Bellini etal, 1989; Tomes et al, 1990; Scheid, Paskowski& Potrykus, 1991). In transgenic maize, bialaphosor PPT resistance conferred by the bar gene hasbehaved as a single dominant trait in most of theprogenies studied (Gordon-Kamm et al, 1991;Register et al 1992; Spencer et al, 1992; Waltersetal, 1992).

There are few studies of the inheritance ofexpression of nonselectable genes in plants follow-ing direct DNA delivery. Results to date suggestthat both Mendelian and non-Mendelian segrega-tion can occur (Morota & Uchimiya, 1988;Tomes et al, 1990). There are only two suchreports in maize (Fromm et al, 1990; Koziel et al,

(0

Plate 5.1. In vivo histochemical GUS analysis by using transgenic rice plants, (a)-(d) Expression of maize ad^-uid/K gene.{e)-(h) Expression of rice rbcS-uidA gene. (a),(e) Expression in a leaf. (b),(f) Expression in a root. (c),(g) Expression in a flower.(d),(h) Expression in a seed.

V

Plates 6.1-6.9. 1: Transient expression of the uidA gene in friable embryogenic callus tissue. 2: Transformed embryogeniccallus clone selected on 1 mg bialaphos/l. 3: Transformed somatic embryo expressing ft-glucuronidase (GUS) activity.4: Transformed plantlet during regeneration phase. 5: GUS activity in a leaf of a transgenic plant. Note vascular-specificexpression. 6: GUS activity in a root tip of a transgenic plant. 7: Control plant reaction 7 days after leaf painting with Ignite(herbicide containing phosphinothricin as active ingredient). 8: Leaf from a transgenic plant carrying the bar gene andexpressing resistance to Ignite. 9: Seed production on a regenerated transgenic plant.

Maize 75

1993), and in these instances expression of thetransgenes segregated as single dominant traits.

A number of reports on plant transformationhave shown stable inheritance of transgenes, withloss of expression. This has been observed follow-ing introduction of a native gene (e.g. see Napoli,Lemieux & Jorgenson, 1990; Van der Krol et al,1990) or multiple transgenes (e.g. Matzke et al>1989; Goring, Thomson & Rothstein, 1990), orfollowing self-pollination of transformants (Scheidet aL, 1991; de Carvalho et aL, 1992). We havemade similar observations in progeny of trans-genic corn following self-pollinations (Register etal., 1992). A number of mechanisms for transgenesilencing (also referred to as cosuppression ortrans inactivation) have been proposed (Matzke etal, 1989; Napoli et al, 1990; Grierson et aL,1991; Mol, VanBlokland & Kooter, 1991; Scheidet al, 1991). The only factor linking reports ofgene silencing in transgenic plants is the presenceof multiple copies of the silenced gene (or parts ofthe gene). It will be interesting to see if transgeniccorn produced via direct DNA delivery (with itsattendant increase in copy number and frequencyof rearrangement) is particularly susceptible tothis kind of epigenetic modification.

Although an in-depth consideration of factorsaffecting transgene expression is outside the scopeof this review, it is worth briefly considering two ofthe most important. These are: regulatorysequences of the introduced gene, and transgeneinsertion site. Evaluation of regulatory sequencescan be made in transient expression systems usingparticle bombardment or protoplasts (e.g. seeBodeau & Walbot, 1992). Presently, the onlypractical approach to addressing the insertion siteis to produce a number of independent transfor-mants and to select those in which the desiredtype and level of transgene expression is achieved.Site-specific recombination or creation of chro-mosome 'domains' (e.g. Hall et al.y 1991; Breyneet aL> 1992) to achieve expression independent ofthe transgene insertion site are approaches thatcould allow construct testing with the minimumnumber of transgenic individuals.

Concluding remarks

Progress in maize transformation is being made atwhat seems to be an ever-increasing pace. An effi-ciency sufficient for selection and recovery of

transformed plants can be achieved by threeapproaches: (i) direct gene transfer to protoplasts,(ii) particle bombardment, and (iii) electropora-tion of zygotic embryos. There is little doubt thatover the coming years these transformation tech-niques will become increasingly simple, more effi-cient and applicable to a wider range of genotypes.

The commercial goal with maize transforma-tion is, in the first instance, to introduce perfor-mance-enhancing transgenes without disruptingthe highly selected elite germplasm used forhybrid production. Small-scale field trials of trans-genic maize plants carrying various single genetraits (e.g. for insect, virus and herbicide resis-tance) have already been conducted in the USAand Europe. In the longer term, transformationmay become a tool that will allow the directedmanipulation and exchange of genes between elitelines, as well as the introduction of complex multi-gene traits such as those involved in yield or thosecontrolling recombination.

Recent developments

Recently, the use of silicon carbide whiskers (seewhisker-mediated transformation, above) togetherwith embryogenic maize suspension culturesresulted in the production of fertile, transgenicmaize plants (Frame et al.> 1994). Transformationefficiency was estimated as one fifth to one tenththat achieved with microprojectile bombardment.In addition, stably transformed cell lines wereobtained from whisker treatment of embryogeniccallus derived from an elite stiff stalk maize inbredvariety, indicating that the method can be extendedto target tissues other than suspension cells.

References

Antonelli, N. M. & Stadler, J. (1990). Genomic DNAcan be used with cationic methods for highlyefficient transformation of maize protoplasts.Theoretical and Applied Genetics, 80, 395-401.

Armstrong, C. L., Peterson, W. L., Buchholz, W. G.,Bowen, B. A. & Sulc, S. L. (1990). Factors affectingPEG-mediated stable transformation of maizeprotoplasts. Plant Cell Reports, 9, 335-339.

Armstrong, C. L., Romero-Sevenson, J. & Hodges,T. K. (1992). Improved tissue culture response ofan elite maize inbred through backcross breedingand identification of chromosomal regions important

76 WILSON et al.

for regeneration by RFLP analysis. Theoretical andApplied Genetics, 84, 755-762.

Bellini, C , Guerche, P., Spielmann, A., Goujard, J.,Lesaint, C. & Caboche, M. (1989). Genetic analysisof transgenic tobacco plants obtained by liposome-mediated transformation: absence of evidence forthe mutagenic effect of inserted sequences in sixtycharacterised transformants. Journal of Heredity, 80,361-367.

Bennetzen, J. L., Lin, C , McCormick, S. &Staskawicz, B. J. (1988). Transformation of Adh nullpollen to Adh+ by 'macroinjection'. Maize GeneticsCooperative Newsletter, 62, 113-114.

Bodeau, J. P. & Walbot, V. (1992). Regulatedtranscription of the maize Bronze-2 promoter inelectroporated protoplasts requires the Cl and Rgene products. Molecular and General Genetics, 233,379-387.

Booy, G., Krens, F. A. & Huizing, H. J. (1989).Attempted pollen-mediated transformation of maize.Plant Physiology, 135, 319-324.

Boulton, M. I., Buchholz, W. G., Marks, M. S.,Markham, P. G. & Davies, J. W. (1989). Specificityof Agrobacterium-mediated delivery of maize streakvirus DNA to members of the Gramineae. PlantMolecular Biology, 12, 31-40.

Breyne, P., Van Montagu, M., Depicker, A. &Gheysen, G. (1992). Characterisation of a plantscaffold attachment region in a DNA fragment thatnormalises transgene expression in tobacco. PlantCell, 4, 463-471.

Bytebier, B., Deboeck, F., Greve, H., Van Montagu,M. & Van Hernalsteens, J. P. (1987). T-DNAorganisation in tumor cultures and transgenic plantsof the monocotyledon Asparagus officinalis.Proceedings of the National Academy of Sciences, USA,84, 5345-5349.

Callis, J., Fromm, M. & Walbot, V. (1987). Intronsincrease gene expression in cultured maize cells.Genes and Development, 1, 1183-1200.

Chourey, P. S. & Zurawski, D. B. (1981). Callusformation from protoplasts of a maize cell culture.Theoretical and Applied Genetics, 59, 341-344.

Chyi, Y.-S., Jorgenson, R. A., Goldstein, D.,Tanksley, S. D. & Figueroa, F. (1986). Locationsand stability of Agrobacterium-mediated T-DNAinsertions in the Lycopersicon genome. Molecular andGeneral Genetics, 204, 64-69.

Coe, E. H. & Sarkar, K. R. (1966). Preparation ofnucleic acids and a genetic transformation attemptin maize. Crop Science, 6, 432-435.

Czernilofsky, A. P., Hain, R., Herrera-Estrella, L.,Lorz, H., Goyvaerts, E., Baker, B. & Schell, J.(1986a). Fate of selectable marker DNA integratedinto the genome of Nicotiana tabacum. DNA, 5,101-113.

Czernilofsky, A. P., Hain, R., Baker, B. & Wirtz, U.

(19866). Studies of the structure and functionalorganisation of foreign DNA integrated into thegenome of Nicotiana tabacum. DNA, 5, 473^182.

D'Halluin, K., Bonne, E., Bossut, M., De Beuckeleer,M. & Leemans, J. (1992). Transgenic maize plantsby tissue electroporation. Plant Cell, 4, 1495-1505.

de Carvalho, F., Gheysen, G., Kushnir, S., VanMontagu, M. C , Inze, D. & Castreana, C. (1992).Suppression of (3-1,3-glucanase transgeneexpression in homozygous plants. EMBO Journal,11, 2595-2602.

De Cleene, M. (1985). The susceptibility ofmonocotyledons to Agrobacterium tumefaciens.Phytopathologie Zeitschrift, 113, 81-89.

De Cleene, M. & De Ley, J. (1976). The host range ofcrown gall. Botanical Reviews, 42, 389-466.

De Wet, J. M. J., Bergquist, R. R., Harlan, J. R.,Brink, D. E., Cohen, C. E., Newell, C. A. & DeWet, A. E. (1985). Exogenous gene transfer inmaize (Zea mays) using DNA-treated pollen. In TheExperimental Manipulation of Ovule Tissues, ed. G. P.Chapman, S. H. Mantell & R. W. Daniels, pp.197-209. Longman, New York.

De Wet, J. M. J., Berthaud, J., Cubero, J. I. &Hepburn, A. G. (1988). Genetic transformation ofcereals. In Biotechnology in Tropical CropImprovement: Proceedings of the InternationalBiotechnology Workshop 12-15 January 1987, pp.27-32. ICRISAT Center, Patancheru, A. P. 502324, India.

De Wet, J. M. J., De Wet, A. E., Brink, D. E.,Hepburn, A. G. & Woods, J. A. (1986).Gametophyte transformation in maize (Zea mays,Gramineae). In Biotechnology and Ecology of Pollen,International Conference on Biotechnology andEcology of Pollen, ed. D. L. Mulcahy, G. B.Mulcahy & E. Ottaviano, pp. 59-64. Springer-Verlag, New York.

Dekeyser, R. A., Claes, B., De Ryke, R. M. U.,Habets, M. E., Van Montagu, M. C. & Caplan,A. B. (1990). Transient gene expression in intactand organised rice tissues. Plant Cell, 2, 591-602.

Donn, G., Nilges, M. & Morocz, S. (1990). Stabletransformation of maize with a chimeric, modifiedphosphinothricin acetyl transferase gene fromStreptomyces viridochromogenes. In Abstracts of VllthInternational Congress on Plant Cell and Tissue Culture,IAPTC, A2-38, p. 53.

Duncan, D. R., Williams, M. E., Zehr, B. E. &Widholm, J. M. (1985). The production of calluscapable of plant regeneration from immatureembryos of numerous Zea mays genotypes. Planta,165, 322-332.

Frame, B., Drayton, P., Bagnall, S., Lewnau, C ,Bullock, P., Wilson, M., Dunwell, J., Thompson, J.& Wang, K. (1994). Production of fertile transgenicmaize plants by silicon carbide whisker-mediated

Maize 77

transformation. In Vitro Cellular and DevelopmentalBiology, 30A, 35.

Fromm, M. E., Taylor, L. P. & Walbot, V. (1986).Stable transformation of maize after gene transfer byelectroporation. Nature (London), 319, 791-793.

Fromm, M. E., Morrish, F., Armstrong, C. L.,Williams, R., Thomas, J. & Klein, T. M. (1990).Inheritance and expression of chimeric genes in theprogeny of transgenic maize plants. Bio I Technology,8, 833-839.

Gordon-Kamm, W. J., Spencer, T. M., Mangano,M. L., Adams, T. R., Daines, R. J., Start, W. G.,O'Brien, J. V., Chambers, S. A., Adams, W. R.,Willetts, N. G., Rice, T. B., Mackey, C. J., Krueger,R. W., Kausch, A. P. & Lemaux, P. G. (1990).Transformation of maize cells and regeneration offertile transgenic plants. Plant Cell, 2, 603-618.

Gordon-Kamm, W. J., Spencer, T. M., O'Brien, J. V.,Start, W. G., Daines, R. J., Adams, T. R.,Mangano, M. L., Chambers, S. A., Zachweija,S. J., Willetts, N. G., Adams, W. R., Mackey, C. J.,Krueger, R. W., Kausch, A. P. & Lemaux, P. G.(1991). Transformation of maize using particlebombardment: an update and perspective. In VitroCellular and Developmental Biology, 27P, pp. 21-27,Tissue Culture Association.

Goring, D. R., Thomson, L. & Rothstein, S. J. (1990).Transformation of a partial nopaline synthase geneinto tobacco suppresses the expression of a residentwild-type gene. Proceedings of the National Academyof Sciences, USA, 88, 1770-1774.

Gould, J., Devey, M., Hasegawa, O., Ulian, E. C ,Peterson, G. & Smith R. H. (1991). Transformationof Zea mays L. using Agrobacterium tumefaciens andthe shoot apex. Plant Physiology, 95, 426-434.

Graves, A. C. F. & Goldman, S. L. (1986). Thetransformation of Zea mays seedlings withAgrobacterium tumefaciens. Plant Molecular Biology, 1,43-50.

Grierson, D., Fray, R. G., Hamilton, A. J., Smith, C. J.S. & Watson, C. F. (1991). Does co-suppression ofsense genes in transgenic plants involve antisenseRNA? Trends in Biotechnology, 9, 122-123.

Grimsley, N. (1990). Agroinfection. PhysiologiaPlantarum, 79, 147-153.

Grimsley, N., Hohn, T., Davis, J. W. & Hohn, B.(1987). Agrobacterium-mediated delivery ofinfectious maize streak virus into maize plants.Nature (London), 325, 177-179.

Grimsley, N., Ramos, C , Hein, T. & Hohn, B.(1988). Meristematic tissues of maize plants aremost susceptible to agroinfection with maize streakvirus. Bio/Technology, 6, 185-189.

Hahne, B., Lorz, H. & Hahne, G. (1990). Oatmesophyll protoplasts - their response to variousfeeder cultures. Plant Cell Reports, 8, 590-593.

Hall, G., Allen, G. C , Loer, D. S., Thompson, W. F.

& Spiker, S. (1991). Nuclear scaffolds and scaffold-attachment regions in higher plants. Proceedings ofthe National Academy of Sciences, USA, 88,9320-9324.

Hartley, R. W. (1988). Barnase and Barstar expressionof its cloned inhibitor permits expression of a clonedribonuclease. Journal of Molecular Biology, 202,913-915.

Heberle-Bors, E., Charvat, B., Thompson, D.,Schernthaner, J. P., Barta, A., Matzke, A. J. M. &Matzke, M. A. (1988). Genetic analysis of T-DNAinsertions into the tobacco genome. Plant CellReports, 1, 571-57r4.

Hooykaas-Van Slogteren, G. M. S. & Hooykaas,P. J. J. (1984). Expression of Ti plasmid genes inmonocotyledonous plants infected withAgrobacterium tumefaciens. Nature (London), 311,763-764.

Jarchow, E., Grimsley, N. H. & Hohn, B. (1991). virF,the host-range-determining virulence gene ofAgrobacterium tumefaciens, affects T-DNA transfer toZea mays. Proceedings of the National Academy ofSciences, USA, 88, 10426-10430.

Junker, B., Zimny, J., Luehrs, R. & Lorz, H. (1987).Transient expression of chimeric genes in dividingand non-dividing cereal protoplasts after PEG-induced DNA uptake. Plant Cell Reports, 6,329-332.

Kaeppler, H. F., Gu, W., Somers, D. A., Rines, H. W.& Cockburn, A. F. (1990). Silicon carbide fiber-mediated DNA delivery into plant cells. Plant CellReports, 9, 415-418.

Kaeppler, H. F., Somers, D. A., Rines, H. W. &Cockburn, A. F. (1992). Silicon carbide fiber-mediated stable transformation of plant cells.Theoretical and Applied Genetics, 84, 560-566.

Klein, T. M., Fromm, M., Weissinger, A., Tomes, D.,Schaaf, S., Sletten, M. & Sanford, J. C. (1988).Transfer of foreign genes into intact maize cellsusing high velocity microprojectiles. Proceedings of theNational Academy of Sciences, USA, 85, 4305-4309.

Klein, T. M., Kornstein, L., Fromm, M. E. &Sanford, J. C. (1989). Genetic transformation ofmaize cells by particle bombardment. PlantPhysiology, 91, 440-444.

Klein, T. M., Wolf, E. D., Wu, R. & Sanford, J. C.(1987). High velocity microprojectiles for delivery ofnucleic acids into living cells. Nature (London), 327,70-73.

Konstantinov, K., Mladenovic, S. & Denic, M.(1991). Recombinant DNA technology in maizebreeding, v. Plant and indigenous bacterial strains.Transformation by the gene controlling kanamycinresistance. Genetika (Beograd), 23, 121-135.

Koziel, M. G., Beland, G. L., Bowman, C , Carozzi,N. B., Crenshaw, R., Crossland, L., Dawson, J.,Desai, N., Hill, M., Kadwell, S., Launis, K., Lewis,

78 WILSON et ah

K., Maddox, D., McPherson, K., Meghji, M. R.,Merlin, E., Rhodes, R., Warren, G. W., Wright, M.& Evola, S. V. (1993). Field performance of elitetransgenic maize plants expressing an insecticidalprotein derived from Bacillus thuringiensis.Bio/Technology, 11, 194-200.

Kranz, E., Bautor, J. & Lorz, H. (1991). In vitrofertilisation of single, isolated gametes of maizemediated by electrofusion. Sexual Plant Reproduction,4, 12-16.

Kranz, E. & Lorz, H. (1990). Micromanipulation andin vitro fertilization with single pollen grains ofmaize. Sexual Plant Reproduction, 3, 160-169.

Langridge, P., Brettschneider, R., Lazzeri, P. & Lorz,H. (1992). Transformation of cereals viaAgrobacterium and the pollen pathway: a criticalassessment. Plant Journal, 2, 631-638.

Last, D. I., Brettell, R. I. S., Chamberlain, D. A.,Chaudhury, A. M., Larkin, P. J., Marsh, E. L.,Peacock, W. J. & Dennis, E. S. (1991). pEMU: animproved promoter for gene expression in cerealcells. Theoretical and Applied Genetics, 81, 581-588.

Ludwig, S. R., Somers, D. A., Peterson, W. L.,Pohlman, R. F., Zarowitz, M. Z., Gengenbach,B. G. & Messing, J. (1985). High frequency callusformation from maize protoplasts. Theoretical andApplied Genetics, 71, 344-351.

Lupotto, E. & Lusardi, M. C. (1990). A potentialapproach for transformation: DUT (desiccation-uptake technique). Evidence of transient expressionin embryogenic structures and regenerated plants.Maize Genetics Cooperative Newsletter, 64, 22-23.

Lyznik, L. A., Ryan, R., Ritchie, S. W. & Hodges,T. K. (1989). Stable co-transformation of maizeprotoplasts with gush and neo genes. Plant MolecularBiology, 13, 151-161.

Matzke, M. A., Primig, M., Trnovsky, J. & Matzke,A. J. M. (1989). Reversible methylation andinactivation of marker genes in sequentiallytransformed tobacco plants. EMBO Journal, 8,643-649.

Mazur, B. J. & Falco, S. C. (1989). The developmentof herbicide resistant crops. Annual Review of PlantPhysiology, 40, 441-470.

Mitchell, J. C. & Petolino, J. F. (1991). Plantregeneration from haploid suspension and protoplastcultures from isolated microspores of maize. Journalof Plant Physiology, 137, 530-536.

Mol, J., VanBlokland, R. & Kooter, J. (1991). Moreabout co-suppression. Trends in Biotechnology, 9,182-183.

Morocz, S., Donn, G., Nemeth, J. & Dudits, D.(1990). An improved system to obtain fertileregenerants via maize protoplasts isolated from ahighly embryogenic suspension culture. Theoreticaland Applied Genetics, 80, 721-726.

Morota, H. & Uchimiya, H. (1988). Inheritance and

structure of foreign DNA in progenies of transgenictobacco obtained by direct gene transfer. Theoreticaland Applied Genetics, 76, 161-164.

Muller, A. J., Mendel, R. R., Schiemann, J., Simoens,C. & Inze, D. (1987). High meiotic stability of aforeign gene introduced into tobacco byAgrobacterium-mediated transformation. Molecularand General Genetics, 207, 171-175.

Napoli, C , Lemieux, C. & Jorgenson, R. (1990).Introduction of a chimeric chalcon synthase gene inpetunia results in reversible co-suppression ofhomologous gene in trans. Plant Cell, 2, 279-289.

Negrutiu, I., Shillito, R., Potrykus, I., Biasini, G. &Sala, F. (1987). Hybrid genes in the analysis oftransformation conditions. Plant Molecular Biology,8, 363-373.

Neuhaus, G. & Spangenberg, G. (1990). Planttransformation by microinjection techniques.Physiologia Plantarum, 79, 213-217.

Neuhaus, G., Spangenburg, G., Mittelsten Scheid, O.& Sweiger, H.-G. (1987). Transgenic rapeseedplants obtained by the microinjection of DNA intomicrospore-derived embryos. Theoretical and AppliedGenetics, 75, 30-36.

Ohta, Y. (1986). High-efficiency genetictransformation of maize by a mixture of pollen andexogenous DNA. Proceedings of the National Academyof Sciences, USA, 83, 715-719.

Petersen, W. L., Sulc, S. & Armstrong, C. L. (1992).Effect of nurse cultures on the production of macro-calli and fertile plants from maize embryogenicsuspension culture protoplasts. Plant Cell Reports, 10,591-594.

Potrykus, I. (1990). Gene transfer to plants: assessmentand perspectives. Physiologia Plantarum, 79, 125-134.

Potrykus, I., Harms, C. T. & Lorz, H. (1976).Problems in culturing cereal protoplasts. In CellGenetics in Higher Plants, ed. D. Dudits, G. L.Farkas & G. Lazar, pp. 129-140. Akademiai Kiado,Budapest.

Potrykus, L, Paszkowski, J., Saul, M. W., Petruske, J.& Shillito, R. D. (1985). Molecular and generalgenetics of a hybrid foreign gene introduced intotobacco by direct gene transfer. Molecular andGeneral Genetics, 199, 169-177.

Prioli, L. M. & Sondahl, M. R. (1989). Plantregeneration and recovery of fertile plants fromprotoplasts of maize (Zea mays L.). Bio/Technology,7, 589-594.

Quayle, T. J. A. & Feix, G. (1992). Functional analysisof the —300 region of maize zein genes. Molecularand General Genetics, 231, 369-374.

Ranieri, D. M., Bottino, P., Gordon, M. P. & Nester,E. W. (1990). Agrobacterium-mediatedtransformation of rice (Oryza sativa L.).Bio/Technology, 8, 33-38.

Register III, J. C , Sillick, J. M., Peterson, D. J.,

Maize 79

Bullock, W. P. & Wilson, H. M. (1992). Molecularanalysis of transgenic maize callus, plants andprogeny. Journal of Cellular Biochemistry, 16F, 210.

Rhodes, C. A., Lowe, K. S. & Ruby, K. L. (1988a).Plant regeneration from protoplasts isolated fromembryogenic maize cell cultures. Bio/Technology, 6,56-60.

Rhodes, C. A., Pierce, D. A., Mettler, I. J.,Mascarenhas, D. & Detmer, J. J. (19886).Genetically transformed maize plants fromprotoplasts. Science, 240, 204-207.

Riggs, C. D. & Bates, G. W. (1986). Stabletransformation of tobacco by electroporation:evidence for plasmid concatenation. Proceedings ofthe National Academy of Sciences, USA, 83,5602-5606.

Roeckel, P., Heizmann, P., Dubois, M. & Dumas, C.(1988). Attempts to transform Zea mays via pollengrains. Effects of pollen and stigma nucleaseactivities. Sexual Plant Reproduction, 1, 156-163.

Russell, J. A., Roy, M. K. & Sanford, J. C. (1992a).Physical trauma and tungsten toxicity reduce theefficiency of biolistic transformation. PlantPhysiology, 98, 1050-1056.

Russell, J. A., Roy, M. K. & Sanford, J. C. (19926).Major improvements in biolistic transformation ofsuspension-cultured tobacco cells. In Vitro Cellularand Developmental Biology, 28P, 97-105.

Sahi, S. V., Chilton, M. D. & Chilton, W. S. (1990).Corn metabolites affect growth and virulence ofAgrobacterium tumefaciens. Proceedings of the NationalAcademy of Sciences, USA, 87, 3879-3883.

Sanford, J. C. (1990). Biolistic plant transformation.Physiologia Plantarum, 79, 206-209.

Sanford, J. C , Klein, T. M., Wolf, E. D. & Allen, N.(1987). Delivery of substances into cells and tissuesusing a particle bombardment process. Journal ofParticle Science Technology, 5, 27-37.

Sanford, J. C , Skubik, K. A. & Reisch, B. I. (1985).Attempted pollen-mediated plant transformationemploying genomic donor DNA. Theoretical andApplied Genetics, 69, 571-574.

Sanford, J. C , Smith, F. D. & Russell, J. A. (1993).Optimising the biolistic process for differentbiological applications. Methods in Enzymology, 217,483-509

Saul, M. W. & Potrykus, I. (1990). Direct genetransfer to protoplasts: fate of the transferred genes.Developmental Genetics, 11, 176-181.

Scheid, O. M., Paskowski, J. & Potrykus, I. (1991).Reversible inactivation of a transgene in Arabidopsisthaliana. Molecular and General Genetics, 228,104-112.

Schlappi, M. & Hohn, B. (1992). Competence ofimmature maize embryos for Agrobacterium-mediated gene transfer. Plant Cell, 4, 7-16.

Schnorf, M., Neuhaus-Url, G., Galli, A., Lida, S.,

Potrykus, I. & Neuhaus, G. (1991). An improvedapproach for transformation of plant cells bymicroinjection: molecular and genetic analysis.Transgenic Research, 1, 23-30.

Schocher, R. J., Shillito, R. J., Saul, M. W.,Paszkowski, J. & Potrykus, I. (1986). Co-transformation of unlinked foreign genes into plantsby direct gene transfer. Bio/Technology, 4,1093-1096.

Shillito, R. D., Carswell, G. K., Johnson, C. M.,Dimaio, J. J. & Harms, C. T. (1989). Regenerationof fertile plants from protoplasts of elite inbredmaize. Bio I Technology, 1, 581-587.

Somers, D. A., Birnberg, P. R., Petersen, W. L. &Brenner, M. L. (1987). The effect of conditionedmedium on colony formation from 'Black MexicanSweet' corn protoplasts. Plant Science, 53, 249-256.

Songstad, D. D., Armstrong, C. L. & Petersen, W. L.(1991). AgNO3 increases type III callus productionfrom immature embryos of maize inbred B73 and itsderivatives. Plant Cell Reports, 9, 699-702.

Spencer, T. M., Gordon-Kamm, W. J., Daines, R. J.,Start, W. G. & Lemaux, P. G. (1990). Bialaphosselection of stable transformants from maize cellculture. Theoretical and Applied Genetics, 79,625-631.

Spencer, T. M., O'Brien, J. V., Start, W. G., Adams,T. R., Gordon-Kamm, W. J. & Lemaux, P. G.(1992). Segregation of transgenes in maize. PlantMolecular Biology, 18, 201-210.

Sun, C. S., Prioli, L. M. & Sondahl, M. R. (1989).Regeneration of haploid and dihaploid plants fromprotoplasts of sweet and supersweet (sh2sh2) corn.Plant Cell Reports, 8, 313-316.

Thompson, C. J., Mowa, N. R., Tizard, R., Crameri,R., Davies, J. E., Lauwereys, M. & Botterman, J.(1987). Characterisation of the herbicide-resistancegene bar from Streptomyces hygroscopicus. EMBOJournal, 6, 2519-2523.

Thompson, J. A., Abdullah, R. & Cocking, E. C.(1986). Protoplast culture of rice using mediasolidified with agarose. Plant Science, 47, 123-133.

Tomes, D. T., Weissinger, A. K., Ross, M., Higgins,R., Drummond, B. J., Schaaf, S., Malone-Schoneberg, J., Staebel, M., Flynn, P., Anderson, J.& Howard, J. (1990). Transgenic tobacco plants andtheir progeny derived by microprojectilebombardment of tobacco leaves. Plant MolecularBiology, 14, 261-268.

Topfer, R., Gronenborn, B., Schaefer, S., Schell, J. &Steinbiss, H. (1990). Expression of engineeredwheat dwarf virus in seed-derived embryos.Physiologia Plantarum, 79, 158-162.

Topfer, R., Gronenborn, B., Schell, J. & Steinbiss, H.(1989). Uptake and transient expression of chimericgenes in seed-derived embryos. Plant Cell, 1, 133-139.

Vain, P., Flament, P. & Soudain, P. (1989a). Role of

80 WILSON et al.

ethylene in embryogenic callus initiation andregeneration in Zea mays L. Journal of PlantPhysiology, 135, 537-540.

Vain, P., Yean, H. & Flament, P. (19896).Enhancement of production and regeneration ofembryogenic type III callus in Zea mays L. byAgNO3. Plant Cell, Tissue and Organ Culture, 18,143-151.

Van der Krol, A. R., Mur, L. A., Beld, M., Mol,J. N. M. & Stuitje, A. R. (1990). Flavenoid genes inpetunia: addition of a limited number of gene copiesmay lead to a suppression of gene expression. PlantCell, 2, 291-299.

Wagner, V. T., Dumas, C. & Mogensen, H. L.(1990). Quantitative three-dimensional study on theposition of the female gametophyte and itsconstituent cells as a prerequisite for corn (Zeamays) transformation. Theoretical and AppliedGenetics, 79, 72-76.

Waldron, J. C. (1987). Pollen transformation. MaizeGenetics Cooperative Newsletter, 61, 36-37.

Walters, D. A., Vetsch, C. S., Potts, D. E. &Lundquist, R. C. (1992). Transformation andinheritance of a hygromycin phosphotransferasegene in maize plants. Plant Molecular Biology, 18,189-200.

7Barley, Wheat, Oat and Other Smallgrain Cereal Crops

Ralf R. Mendel and Teemu H. Teeri

Introduction

The most important crop plants of the worldbelong to the monocotyledonous grasses. Fromthe perspective of genetically transforming thesesuccessful outdoors plants with new genes, theyhave properties in common. The tissues of thesegrasses lack what is called the wound response -the characteristic of the cells at a site of physicaldamage to proliferate and form calloid protectivetissue - the cereal tissue rather seals the woundsite by lignification and sclerification of the cells.

This alternative strategy in reaction to wound-ing seems to be the basis of the recalcitrance ofcereal plants to the manipulations of the experi-mentalist. The two most commonly used genetransfer techniques utilize either the capability ofAgrobacterium to mate with plants or the capacityof a naked plant cell (protoplast), where DNA canbe introduced by physical means, to regenerate itscell wall, start to divide and finally to differentiateinto a mature plant. Neither of these methodsfunctions easily for the cereal grasses. Agro-bacteria deliver their transferred DNA into cells ata wound site - destined to deteriorate. Protoplastsof cereal plants have been very hard to regenerateand especially to retain their capacity to differenti-ate after regeneration. This situation has led to thedevelopment of alternative, often very imagina-tive, gene transfer techniques. The literature ofcereal transformation is littered with reports ofexotic methods that mostly have given ambiguousresults and are very hard to repeat.

Still, the characteristics of the cereal life-styleare not a fundamental obstacle to either plantregeneration from protoplasts or gene transfer

itself. Investment of efforts has made it possible togenerate stably transformed plants from both riceand maize (Rhodes et al.> 1988; Toriyama et al.,1988). However, compared to the success withthose plants, gene transfer to the small grain cere-als of the moderate climatic regions is still in itsbeginning. Only a few papers are published andmost information comes from congress abstractsand via personal communications. Still, remark-ably, the first reports of obtaining transgenic, fer-tile wheat (Vasil et al., 1992) and oat (Somers etal.y 1992) have recently been published.

In the context of this chapter, stable transfor-mation means the introduction of one or moreintact copies of one or more foreign genes into thegenome of the plant, where the foreign genes arestably maintained through cell divisions and areexpressed. Two forms of stably transformed planttissue are possible to generate: (i) transgenic cellcultures, and (ii) transgenic mature plants. In thelatter case, regular Mendelian inheritance of theforeign genes also is required.

For successful stable transformation, three con-ditions have to be met: (i) the foreign DNA mustpenetrate the cell, (ii) it has to stabilize by integra-tion into the chromosomes, and (iii) the trans-formed cell must regenerate into either aproliferating tissue culture or a mature plant.

It is possible to monitor the first step of trans-formation, the penetration, independently fromthe steps that follow. Free DNA, once it hasentered the plant cell, will persist there for severaldays, find its way to the nucleus, and have itsgenes expressed by the cell's machinery.Eventually the DNA is lost, but the phenomenonof transient expression is important in assessing

81

82 MENDEL AND TEERI

penetration by using an established marker genesystem. The most sensitive markers for penetra-tion are complete genomes of plant viruses, whichstart to replicate and spread as a result of penetra-tion of a single molecule. In this way it has beenestablished that agrobacteria are capable of intro-ducing DNA also to cereal cells (Grimsley et al.,1987). The histochemically detectable reportergenes are, however, most useful as penetrationmarkers. By staining the tissue after gene delivery,it is possible to count the number and evaluate thedistribution of the cells that have been penetrated.Also, detection of the activity of histochemicallydetectable reporter genes gives most informationfor optimization, since the number of transformedfoci is more important than the quantified averageability of the cells to express the particular con-struct used.

The next step of transformation, stabilization ofthe foreign genes, can be measured by followingpersisting marker gene expression. This is, how-ever, more complicated because both of the histo-chemically detectable reporters that have beenused for plants (uidA and lacZ) code for extremelypersistent enzymes. Usually the assessment of sta-bilization is done together with regeneration.Active stabilization of foreign genes has not beenused in direct gene transfer and the experimental-ists have relied on natural stabilization through'illegitimate recombination', a phenomenon witha fortunately high frequency of up to several percent (Klein et al, 1988).

The step from the penetration and subsequentstabilization to regeneration is not trivial. Thelogic is simple: DNA has to enter cells with fullregeneration capacity. Penetration of foreignDNA to cereal cells has been easy to demonstratewith several techniques. The difficulty in the abil-ity to penetrate regenerating cells is equal to thedifficulty in obtaining transgenic cereal plants.

Stabilization of foreign genes in the cells of adifferentiated plant is only the first step.Consistent inheritance of the transgenes in thenext generation according to the laws of Mendel(1865) is necessary not only for the demonstrationof true integration into chromosomes but also forthe practical maintenance of the transgenic plants.The first transgenic dicotyledonous plantsobtained with Agrobacterium vectors were shownto fulfil this condition by rule (Budar et al., 1986).It is important to realize that this is not necessarilytrue for new plant species transformed with new

techniques (e.g. see Rogers & Rogers, 1992). It isclear that the assessment of gene transfer andtransmission of the transgenes to further genera-tions in these plants must be done by using thestrongest and most clear-cut methods of molecu-lar analysis (Potrykus, 1990).

Methods of transformation

The principal difficulty in transformation of cerealplants is not in the methods of DNA delivery, butin those of culturing the cells. It is, in fact, specialthat the Solanaceous species (tobacco, tomato,petunia, potato, etc.) react so well in vitro andenabled the whole concept of transgenic matureplants to be established ten years ago.

The development of cereal transformationmethods has followed two lines of thought. First,much effort has been put into overcoming thetechnical problems in protoplast culture. The abil-ity to regenerate protoplasts to mature plantspractically means ability to transform. Second,attempts have been made to bypass the problemsin the protoplast culture by developing alternativemethods of DNA delivery into intact cells withmore potential to regenerate into mature plants.

In the following, we discuss first the protoplasttechniques and the one successful case of thealternative delivery method, particle bombard-ment. The required techniques of in vitro tissueculture are included. After a section for themarker and reporter gene constructions used, acollection of the other methods of DNA deliveryare presented. At the moment, they have at thebest a status of being 'promising', although someof them may prove to be 'successful' as well. Sincethe alternative methods represent a large amountof the work done towards transformation of cerealplants, we feel that they should be presented to thereader.

Protoplasts

The isolation of protoplasts from different tissuesof the Gramineae is relatively straightforward(Morrish, Vasil & Vasil, 1987). Vital protoplastscan be obtained in large amounts, e.g. from youngleaves. However, protoplasts isolated from differ-entiated cereal tissues have one thing in common:they are unable to divide. Nevertheless, mesophyllprotoplasts, for example, are excellent material for

Small-grain cereal crops 83

physiological studies as well as for experiments oftransient gene expression.

In order to test a given gene construct in tran-sient expression, protoplasts have to be isolated inlarge amounts and should be uniform and vital.These requirements are met by protoplasts iso-lated either from leaves or from suspension cellcultures. The transformation scheme is principallythe same as for other monocots and dicots, i.e.DNA-uptake is mediated by polyethylene glycol(PEG), electroporation, or a combination of both(see e.g. Lazzeri et al., 1991). One to 3 days afterDNA transfer, the cells are harvested, and expres-sion of the reporter gene is measured usually bydetermining the enzymatic activity it codes for.Also immunodetection of the gene product ormeasurement of the reporter mRNA is possible.There are reports of use in cereals for most of thecommon reporter genes such as raprll, uidA, cat,bar and luc.

When using protoplasts to assay transient geneexpression, it is important to realize that proto-plasts are highly stressed cells and do not neces-sarily reflect all of the properties they had in thetissue from which they were derived. Still, ascheme for correct regulation of gene expressionin cereal protoplasts by plant hormones has beenpublished (see below).

The crucial point for stable transformation viathe protoplast approach is the embryogenicity ofthe suspension culture as protoplast source. Twomain sources of cells are successfully used for thispurpose: immature embryos and anthers.Suspension cultures derived from these tissueshave been shown to be embryogenic (Liihrs &Lorz, 1987a,6; Jahne et al, 1991a; Vasil, Redway& Vasil, 1990) and will frequently give protoplaststhat are able to divide, depending on the skill ofthe experimenter, on the quality of the plant mate-rial, and on the genotype used. Also mesocotyl-derived suspension cultures have been shown tofulfill these criteria (Miiller, Schulze & Wegner,1989). Approximately 3-4 months after taking theimmature embryos or the anthers into culture, thesuspensions are fine enough to give large amountsof vital and dividing protoplasts that are able toform protoplast-derived colonies.

In principle, it should be possible to use thesecell lines as starting material also for obtaining sta-bly transformed protoplast-derived plants. How-ever, the time window where the cell lines willretain their embryogenicity is rather narrow (1-3

months), and protoplasts isolated from later culti-vation stages of the cultures will develop into callithat are no longer able to regenerate into plants.Thus, in order continuously to have embryogenicsuspension cell cultures available, it is necessary tore-establish the embryogenic cell culture repeat-edly, or to take the existing cell culture into cryo-preservation. However, suspension cell culturesthat have lost their embryogenic potential are notuseless. They are often good sources of proto-plasts for gene transfer experiments that aim togenerate transgenic calli.

There are numerous basic and applied prob-lems that can be studied at the level of transgeniccell cultures without the urgent necessity of havingregenerated plants available. The study of mecha-nisms for the recognition and integration of for-eign genes into cereal genomes, the physiologicalquestions connected with housekeeping genesinvolved in primary metabolic steps, the recogni-tion of targeting and processing signals of a for-eign or endogenous gene by the cereal cell, andthe ability of the cells to produce foreign protein inactive form and its effects on the basic physiologyof the cell can all be investigated in tissue culture.For all these areas of research, the protoplastapproach is an excellent transformation systemthat gives stably transformed cell colonies in ashort time.

Particle gun

Originally the particle gun was seen as one of thefantastic (but improbable) new ways to solve theproblems in transformation of recalcitrant species.Remarkably, and unlike many of the other alterna-tive methods, it could quickly be adopted by newlaboratories, it functions very reproducibly, and ithas proven to be a truly new approach to directgene transfer. Thus, not only does it solve manyexisting problems, but also it penetrates intototally new areas such as transformation oforganelle genomes (Svab, Hajdukiewicz &Maliga, 1990).

The principle of the particle gun is simple:DNA is precipitated on inert microcarriers of goldor tungsten, which can be accelerated to a speedwhere the particles have sufficient impact to pene-trate intact plant cells. Once inside the cell, theDNA dissolves and what follows is common todirect transfer of DNA into wall-less protoplasts:transient expression of the genes in the transferred

84 MENDEL AND TEERI

DNA can be measured, and the sequences have achance to integrate into the chromosomes by 'ille-gitimate recombination'.

The fundamental novelty of the method is thatyou end up with an intact cell with foreign DNA.Thus, this cell has not been subjected to thetrauma of protoplasting and should have retainedits capability of cell division and finally of differen-tiation to a whole new plant - if it ever had suchpotential.

The delivery of foreign DNA into cereal tissuesby particle gun bombardment can be assessed bymeasuring transient expression of the transferredgenes. Histochemically detectable reporter geneactivities give most information for optimization,since they allow counting of the transformed foci.Transient assay of particle-bombarded tissue hasalso valuable applications of its own. It makes pos-sible tissue-specific assays of gene constructionsdirectly in intact tissue and without the need ofstable transformation (see below).

In order to obtain transgenic plants, a naturaltarget for bombardment would be a meristematiccell. However, these cells are small, compact andhard to reach. The in vitro induction of the regen-erative potential in nonmeristematic cells is morepractical. Tissue cultures initiated from immatureembryos or microspores, and forming embryo-likestructures regenerating to normal, fertile plantscan be obtained from cultivars of barley (Hordeumvulgare; Ahloowalia, 1987; Jahne et al., 1991a;Jahne, Lazzeri & Lorz, 19916), wheat (Triticumaestivum; Vasil et al., 1990) and oat (Avena sativa;Bregitzer et al., 1991). It is also possible that thehighly focussing bombardment device of Sautteret al. (1991) will prove useful in the transforma-tion of cells that naturally have high regenerationcapacity.

Promoters and marker genes

The 35S promoter of cauliflower mosaic virus isthe most widely used promoter for plant transfor-mation. Experiments of transient and stable trans-formation with barley and wheat protoplasts showthat it is suitable also for driving gene expressionin these small grain cereals. However, there seemsto be a difference in the strength of this promoterbetween transient and stable transformation: intransient assays it is a rather weak promoterwhereas its performance in stable transformants iscomparable to those levels known from dicots

such as tobacco (Mendel et al., 1990; J. Schulze,A. Nerlich & R. Mendel, unpublished data),allowing the selection of large numbers of trans-genic colonies.

During the last years, numerous groups havetried to improve the performance of the 35S pro-moter in monocots by introducing monocot-specific sequences into the 5'-untranslated leaderof reporter gene constructs, e.g. the exon 1 andintron 1 of the maize shl gene (Maas et al., 1991),the intron 1 of the maize adhl gene (Callis,Fromm & Walbot, 1987) and the intron 1 of therice actl gene (McElroy et al., 1990, 1991). Theseexperiments were performed mainly in maize. Itwas shown that also in barley protoplasts theinsertion of exon 1/intron 1 of the shl gene intothe nontranslated leader could increase the tran-sient expression levels of cat up to 1000-fold andthose of luc up to 250-fold (Maas et al, 1991;Topfer et al., 1992). However, when exchangingcat for uidA as reporter gene, these effects were notany longer so pronounced (Lazzeri & Lorz, 1991).

Another way to boost transient expression lev-els in monocots is the systematic testing in tran-sient assays of constitutive monocot promotersinserted in front of reporter genes. When the riceactin gene promoter was tested (McElroy et al.,1990), fused in front ofuidA, its performance was1000-fold better than that of the 35S promoter inbarley protoplasts (Spickernagel et al., 1991) andabout 50-100-fold better than that of 35S whenusing the particle bombardment approach withbarley suspension cells (R. R. Mendel, unpub-lished data).

Last et al. (1991) found yet another possibilityfor obtaining high transient gene expression inmonocots. They constructed an 'artificial' pro-moter consisting of six repeats of an anaerobicresponsive element of maize plus four repeats ofthe ocs enhancer in front of a modified adhl pro-moter with the adhl intron 1. This promoter(pEmu), fused in front of uidA, showed a perfor-mance similar to that of the rice actin promoterwhen it was tested in the same transient assays(protoplasts and particle bombardment) that wereused for the actin promoter (Spickernagel et al.,1991; R. R. Mendel, unpublished data). Withdicots, however, there are indications that toostrong promoters are not optimal for selection ofstable transformants: Hinchee et al. (1991)observed with petunia that the percentage ofescapes increased with promoter strength due to


cross-feeding protection so that a promoter ofmoderate strength is more suitable for driving theselected marker.

There seems to be a general problem in usinguidA as an effective reporter gene in barley andwheat, and this is particularly obvious in stablytransformed cells where in many cases the expres-sion of uidA is completely repressed. This, how-ever, is not due to the 35S promoter per se, or tothe plasmid carrying the uidA gene as such, or tothe lack of transcription factors etc., for the fol-lowing reasons, (i) The 35S promoter is veryactive in combination with the nptll gene that islocalized adjacent to 35S-uidA on the same pieceof DNA that had been introduced into barleygenome without rearrangements (Schulze,Nerlich & Mendel, 19916). (ii) 35S-uidA can beeffectively expressed in transient assays in barleycells when the microprojectile-bombardmentapproach is used (Mendel et al., 1989, 1990;Schulze et al, 1991c). (iii) Also the muchstronger, monocot-specific promoter pEmu (Lastet al., 1991) driving the uidA gene is not able toincrease P-glucuronidase (GUS) activity after sta-ble integration into barley cells, although in tran-sient assays after microprojectile bombardmentthere is a dramatic increase in uidA expressioncompared to the 35S-wzdA construct (J. Schulze,A. Nerlich and R. Mendel, unpublished data).The integration of the uidA coding sequence intothe genome seems to be the crucial point inrepressing its activity. The reason for this phe-nomenon is not known. One could speculate thatthe base composition determining the methylationpattern of this bacterial gene is not compatiblewith the cereal gene expression machinery oncethe foreign gene has been integrated, whereas withthe nptll gene sequence this would not be thecase. Also in stably transformed wheat callus, uidAexpression was observed to vary strongly (Mejza etal, 1991; Vasil era/., 1991).

When the performance of 35S-w/JA or pEmu-uidA in barley and wheat on the one hand and inmaize, rice and dicots on the other is comparedwith transient assays (bombardment of suspen-sion culture cells), it is obvious that small-graincereals are generally less effective expressing theintroduced constructs (J. Schulze, A. Nerlich andR. Mendel, unpublished data; Brettschneider,Becker & Lorz, 1991). Taking into account theabove-mentioned effects after stable integrationone has to conclude that the uidA gene is a prob-

lematic reporter gene for small-grain cereals. Thegene luc coding for luciferase might be a bettercandidate, but information about its performancein these recipients is scarce. Maas et al. (1991) andLanahan et al. (1992) have shown that it is princi-pally possible to use luc as a reporter for transientgene expression in barley protoplasts and in ourown experiments we confirmed these data(R. R. Mendel, unpublished data; R. Hannus,unpublished data).

For selecting stable transformants, the mostuseful marker genes are nptll and bar. For nptll,selection on kanamycin gives a considerable back-ground growth of nontransformed colonies. Itsanalog, the compound G418 (also known asgeneticin), is much more effective in selectingtransgenic colonies in barley for example (Schulzeet al.> 19916). However, the selective agent mayalso have an effect on regeneration. We have expe-rienced that G418 increases albinism in barleywhereas kanamycin does not (A. Ritala, unpub-lished data). For rice, Battraw & Hall (1992)report a reverse effect of these chemicals.

The nptll gene under the control of the 35Spromoter is completely sufficient for selectinglarge numbers of stably transformed cell coloniesin barley (Lazzeri et al, 1991; Schulze et al.,19916; Ritala et al., 1993). The bar gene undercontrol of the 35S promoter has also been success-fully used to select stable transformants in barley(Louwerse etal, 1991; Schulze & Mendel, 1991),wheat (Vasil et al., 1992) and oat (Somers et al.,1992). Genes mediating resistance to the antibi-otics hygromycin (Kaneko et al., 1991) andmetothrexate (Chen, Dale & Mullineaux, 1991;Mejza et al., 1991) can be used as well; however,there is not so much information available.

Other methods of DNA delivery

Macroinjection into floral tillersCircumvention of in vitro regeneration would behighly advantageous and several approaches havebeen suggested in this direction. One of them ismacroinjection of DNA directly into floral tillers.De la Pefia, Lorz & Schell (1987) injected DNAcontaining the marker gene nptll into the tillercavity of immature rye tillers (Secale cereale),thereby flooding the immature inflorescence, theflorets and very likely also the enclosed sexualorgans with DNA. It was thought that, once in theanther or ovule, the foreign DNA would penetrate

86 MENDEL AND TEERI

the spore mother cell and would be incorporatedinto the micro- or megaspores. After fertilization,seed setting, harvesting and germination, theseedlings were screened for kanamycin resistance.From 3023 grains, three plants were obtained thatgrew in the presence of kanamycin and expressednptll in their leaves. In genomic DNA, Southernblots revealed appropriately hybridizing restric-tion fragments. However, no transmission to theprogeny was demonstrated.

Rogers & Rogers (1992) used the sameapproach with the genes for GUS and thaumatin,both under control of an aleurone-specific a-amy-lase promoter. Without any selection, the har-vested barley grains were screened for theexpression of the uidA gene by using a piece ofaleurone tissue from each seed. From seeds posi-tive in this test, the embryos were germinated andSouthern blot analysis of DNA from the seedlingswas performed. With a frequency of about 1%,positive plants were obtained that were fertile andcould be selfed so that the next generation alsocould be analyzed. It turned out that somehybridizing bands were lost when different tillersfrom the same plant were compared. In addition,reduction of the intensity of hybridization signals,as well as the appearance of new hybridizingbands in different tissue samples of the sameplants, were taken as an indication of ongoingrearrangements of the introduced foreign DNAwithin the recipient plant. Different stabilities ofthe two reporter genes coding for GUS and thau-matin were observed and conclusions about theimportance of the methylation pattern for genestability were drawn. It was observed that the veryunstable uidA marker maintained the methylationpattern characteristic of prokaryotic DNA, whilethe monocot gene for thaumatin showed someloss of A-methylation and apparent acquisition ofC-methylation in CpG dinucleotides. In furtherexperiments with nonmethylated and with in vitromethylated plasmid DNA, it was shown that thepresence of A^-methyladenine in the transformingplasmid DNA was associated with rapid loss of theDNA from new tillers on the plants, while theabsence of A^-methyladenine coupled withmethylation of C residues at CpG dinucleotidesgreatly increased the stability of the same plasmidand allowed transmission into the second genera-tion. Rogers & Rogers speculated about the pres-ence of a previously undefined system in barleythat discriminates between foreign and genomic

DNA by identifying DNA that lacks the propermethylation pattern followed by removal of thelatter in actively dividing cells. Integration into thegenome was not shown, rather there were indica-tions of an extrachromosomal location of theintroduced DNA that persisted through manymultiples of cell division and was inheritable bythe F2 generation.

Mendel et ah (1990) used macroinjection intofloral tillers for transferring the genes nptll anduidA. With a frequency of 1.6%, kanamycin-resistant plants were obtained that exhibited verylow NPTII activities in leaves. No NPTII activitywas detectable in F2 plants. DNA isolated fromFl and F2 plants was further analyzed. Poly-merase chain reaction (PCR) primers specific foran internal 412 bp fragment of the nptll codingsequence allowed the amplification of the targetsequence in selected Fl and F2 plants in thosecases where NPT II activity was no longerdetectable. Southern hybridization of genomicDNA demonstrated that a hybridizing 1.7 kbwpdl-specific fragment was inherited from Fl toF2 plants; however, surprisingly, different plantsshowed the same fragment size even when Pstlwas used to cut inside the nptll coding sequence inorder to visualize the flanking DNA. It appearedthat independent transformants showed an identi-cal hybridization pattern that was inherited fromFl to F2. Further, the fragment size did not fitinto the restriction map of the introduced plasmidDNA and, most unexpectedly, nondigestedgenomic DNA showed discrete wpfll-hybridizingbands. Obviously the introduced plasmid DNAoccurred in a nonintegrated, i.e. extrachromoso-mal, location. This phenomenon was also dis-cussed by Rogers & Rogers (1992). It remains tobe seen whether these plasmid molecules occurfreely in the cytoplasm or are localized inorganelles.

Du et ah (1989) utilized macroinjection intofloral tillers of Triticum monococcum with nptll asthe selected marker and the gene for soybean legu-min as a cargo gene, and reported a frequency of0.6% for the occurrence of kanamycin-resistantseedlings. Dot hybridizations of genomic DNAwith legumin as probe were positive in 7 out of 20tested green plants. No Southern blot data werepresented.

In a critical review of cereal transformation,Potrykus (1990) argued that with in vivo tech-niques there is the chance that foreign DNA does


not exist in barley cells but rather in contaminat-ing endophytes that are intimately connected withthe plant cell and very difficult to culture outsidethe plant. Rogers & Rogers (1992) exclude thispossibility by showing that the a-amylase pro-moter-mdA construct was correctly transcribed inaleurone tissue from the barley promoter (asassessed by both Sj nuclease protection assays andreverse transcriptase-PCR assays). Mendel et al.(1990, 1991) could not exclude this possibilityand announced further investigations into thisdirection. Konstantinov et al. (1990), however,were able to recover kanamycin-resistant bacteriafrom the F5 generation of maize plants after theyhad macroinjected DNA with nptll as a markerinto the plants before sporogenesis. In the F5 gen-eration, wprll-specific hybridizing fragments wereobserved that were of uniform size in independenttransformants, so that the endophyte theory wastaken into consideration. The recovered bacteriacontained plasmids that hybridized with the nptllprobe, and were identified as Pseudomonas andAcinetobacter species.

The macroinjection approach is an interestingmethod for studying the fate of foreign DNAintroduced into a full-size organism (i.e. an intactplant) rather than into a single cell. Presently,however, this procedure is of purely academicvalue and it is rather unlikely that it will developinto a reliable transformation method of generalapplicability for barley or other monocot or dicotplants.

DNA transfer via the pollen tube pathwayA long time before macroinjection was developed,other approaches were suggested to circumventregeneration in vitro. Most of them proposedintroduction of DNA into gametes, followed byfertilization and zygotic embryogenesis. This kindof approach would be simpler, faster and cheaperthan the in vitro methods, and would also avoidthe problem of somaclonal variation. It was a logi-cal consequence to favor the use of pollen as a vec-tor for DNA because ovules are difficult to isolateand the injection into the embryo sac in situseemed to be too tedious and unpredictable. Itwas hoped that pollen was easily accessible forDNA transfer and that the pollen tube woulddeliver the DNA to the egg cell (for a review, seeHess, 1987).

De Wet et al (1985) and Ohta (1986) sug-gested the use of DNA-treated pollen as a DNA

vector for pollinating fertile plants of maize, andHess (1987) described a similar transformationsystem for Nicotiana glauca, but without molecu-lar-genetic evidence. Ahokas (1987) showed thatDNA uptake into pea pollen was facilitated byliposomes, and Abdul-Baki et al. (1990) demon-strated the introduction of labeled DNA intopollen grains of Nicotiana gossei by electropora-tion. Twell et al. (1989) bombarded pollen withDNA using the particle gun approach and demon-strated transient gene expression of the markergene uidA. However, although DNA could betaken up into pollen of plants of diverse phylo-genetic origin, there is no case reported yet thatdecisively demonstrates a successful gene transferusing this kind of approach.

Another pollen tube pathway approach wasdescribed by Zhong-xun & Wu (1988) for ricewhere, after pollination, the stigmas were cut offand DNA solution was directly applied to the styleusing the pollen tube as a microcapillary. In DNAhybridizations of seed-derived marker-selectedplants, positive signals were obtained. Our dupli-cation of the experiment with barley was not suc-cessful (G. H. Patel & T. H. Teeri, unpublisheddata).

A third approach was proposed by Picard et al.(1988). They applied DNA with the nptll markergene on to just-pollinated stigmas of wheat plantsand found seedlings with kanamycin resistanceand expression of the nptll gene among the prog-eny of the DNA-treated flowers. However, nomolecular-genetic evidence was presented. Later,several groups tried to apply this method to othercereals, and failed. Only recently, by using thesame approach, the group of Jacquemin reportedthe transfer of the maize zein gene into wheatplants (Delporte et al.> 1991). nptll was used asthe selectable marker for the nonselected cargogene. DNA of kanamycin-resistant plants wasshown to contain both nptll and zein-gene-specific hybridization signals, and after selfing, thetissue-specific expression of a-zein was immuno-logically detected in the endosperm. Furthermolecular-genetic work is in progress.

The pollen tube approach of Picard was alsoused by us in similar experiments with barley(Mendel et al., 1990). DNA with 35S-nptll asmarker was directly applied to the stigma of barleyplants that were pollinated 5-20 min before DNAapplication. DNA was applied in a buffer thatmaintained it stable for more than 1 h. In total

88 MENDEL AND TEERI

1058 plants were treated, 11 200 seeds wereobtained and germinated on 150 mg kanamycin/1,and 305 green plants selected. Some of theselected plants showed very weak NPT II signalsthat, after selfing the plants, were lost in the prog-eny. Hybridization of genomic DNA (Fl and F2)with a nprll-specific probe revealed in some casespositive signals, the size of which, however, didnot fit into the restriction map of the utilized geneconstruct. Moreover, in some selected plants anidentical hybridization pattern occurred, althoughthese plants were derived from independentmother plants. These observations are very similarto those of the macroinjection approach and canbe interpreted in the same way (see above).

Kartel et al. (1990) applied the pollen tubeapproach for introducing the nptli gene into bar-ley plants followed by selection for kanamycinresistance. In Southern blots, the selected plantsshowed «/>rII-specific fragments; however, nohybridizing fragments were detectable in the self-ing progeny of kanamycin-resistant plants.NPT II assays were not performed.

Hess, Dressier & Nimmrichter (1990) pre-sented a system named 'pollen-mediated indirectgene transfer' with Agrobacterium as the infectiveagent for the pollen. The authors pipetted Agro-bacterium tumefaciens containing the marker genenptll on to wheat spikelets. After selection of theprogeny for kanamycin resistance, NPT II assaysof the selected plants showed extremely lowNPT II activities. The transmission of the trans-genes was followed for two generations.Although transmission was observed, the ratiosfailed to follow Mendelian laws. Also, rearrange-ment and loss of signals in Southern analysis wasdetected, suggesting instability. The authorsreport that probing with Ti plasmid-specific virprobes was negative, thus demonstrating that theplants were not contaminated by Agrobacterium.A rigid molecular analysis of these plants withrespect to the integrative state of the DNA willbe necessary.

Recently, Heberle-Bors critically reviewed allkinds of pollen-mediated gene transfer approaches,and came to the conclusion that 'the elegant ideaof using mature pollen as a super vector for genetransfer fell short of experimental reality'(Heberle-Bors et al, 1990; Heberle-Bors, 1991).We do not exclude that some kind of DNA trans-fer is being brought about by these approaches.However, these experiments are far from being

reproducible and predictable. From the presentpoint of view, it is unlikely that the pollen tubeapproach will develop into a reliable transforma-tion method of general applicability for monocotor dicot plants.

Electroporation of intact explantsWith the most sensitive methods of detection (i.e.transformation with a virus genome), it was shownmany years ago that nucleic acid can enter intactplant cells in electroporation conditions(Morikawa et al., 1986). Lindsey & Jones (1987)were able to show low levels of transient expres-sion from foreign DNA after electroporation ofintact sugar beet tissue culture cells. Finally,Dekeyser et al. (1990) could develop the system toa level of practical use and to analyze tissue-specific expression of recombinant genes in cellsof intact plant tissue. Significantly, this work wasdone with a monocot plant, rice, and an immedi-ate application, as suggested by the authors, is sta-ble transformation of cereal plants. They alsotested for transient expression in leaf bases of sev-eral other plants, including barley. With a closelyrelated method, stable transformation of maizewas recently achieved (D'Halluin et al., 1992). Itremains to be experienced whether this methodwill be generally useful.

Imbibition of seedsThe establishment of genetic transformation tech-niques of plants during the last decade has alsodefined the most useful molecular-genetic tools inthe assessment of gene transfer. This has made itpossible to reevaluate claims of gene transfer toplants, presented at the time when only ambigu-ous markers for the event were available (e.g. Doy,Gresshoff & Rolfe, 1973; Johnson, Grierson &Smith, 1973). Topfer et al. (1989) have testedwhether in fact it is possible to enter plant cells byexogenous DNA during the imbibition of the dryembryo. They soaked dry embryos in DNA solu-tion containing a specific reporter gene constructand, indeed, could show transient expression inboth cereal and legume tissues. The natural con-tinuation of these experiments is to select for sta-bilization of transferred resistance genes. No datafrom these kinds of experiment have so far beenpublished, and the usefulness of dry seed imbibi-tion as a transformation method is still open todebate.


MicroinjectionAlthough plant cells have strong cell walls, it ispossible to penetrate cells by techniques ofmicroinjection (Toyoda et al., 1990). As withother methods of direct gene transfer, the targetcell must have the potential to regenerate into awhole plant. The advantage of microinjection isthat the investigator can target the DNA into cellsthat he expects to regenerate. In spite of many tri-als, however, transgenic cereal plants have not yetbeen obtained (Potrykus, 1991).

Transformation with AgrobacteriumBy far the most successful method of gene transferfor very many species of dicotyledonous plants isthe use of Agrobacterium-based vectors. Agro-bacteria have a unique ability to penetrate cells at awound site and actively integrate the transferredDNA stably into plants chromosomes.

For penetration, Agrobacterium have utilized avariation of interbacterial mating, or conjugation.With a specific system, the bridge between thecells is generated and the transferred DNA entersthe plant cell as a single-stranded molecule(Stachel & Zambryski, 1986). The integration ofthe transferred DNA is probably also specific,although the mechanism is as yet not understood.

The next step, regeneration and subsequentproliferation, is also controlled by the pathogenicAgrobacterium. The transferred DNA containsgenes that are expressed in the transformed plantcell and lead to the synthesis of plant hormones,promoting further cell division and ensuring thatthe transformed cell, starting to produce uniquenutrients to the benefit of the bacterium, willflourish and divide. When the Agrobacterium sys-tem is used for gene transfer by the investigator,the last step of control is removed and replaced byaddition of exogenous growth factors, leadingfinally to the achieving of normally differentiatedtransgenic plants.

The first demonstration that Agrobacterium vec-tors can deliver DNA also into cereal (maize) cellswas obtained by assaying the activity of infectedviral sequences (Grimsley et al., 1987). Sincethen, Agrobacterium tumefaciens has been shown tointeract - very inefficiently - with the cerealsmaize (Gould et al, 1991), wheat (Gould et al,1992) and rice (Raineri et al, 1990; Li et al,1991). The main obstacle seems to be the growthhabits of cereal cells, especially the fact thatwounding is not followed by proliferation of the

cells at the wound site. Agrobacterium-mtdi&tzdgene transfer to monocot plants that, opp.osing tothe rule, do have a wound response, was demon-strated early (Hernalsteens et al, 1984; Hooyakaas-Van Slogteren, Hooyakaas & Schilperoort, 1984).

Although no breakthrough has yet beenreported, there is certain promise in Agro-bacterium-medisLted gene transfer to cereals. Inprinciple, agrobacteria could be used to transformany plant cells, provided that the transformed cellwould have the capacity to regenerate. It hasbecome clear that somehow bypassing the lack ofwound response of cereals is critical.

Miscellaneous methodsElectrophoretic migration of DNA (35S-uidA)into barley caryopses, followed by transientexpression of uidA, was described by Ahokas(1989). Using a laser microbeam, Kaneko et al.(1991) introduced the gene for hygromycin phos-photransferase (HPT) into barley microspore cellsand selected for hygromycin-resistant callus.

In dicots, further ways for DNA transfer havebeen reported that are applicable also to cerealcells. For introducing DNA into tobacco proto-plasts the use of PEG or electroporation could bereplaced by a short pulse of mild sonication(Joersbo & Brunstedt, 1990). Sonication was alsouseful for stably introducing DNA into cells ofwhole leaf segments of tobacco (Zhang et al.,1991). Another way was shown by Kaeppler et al.(1990): suspension culture cells of maize andtobacco showed transient gene expression aftervortexing of the cells in the presence of silicon car-bide fibers. Weber et al. (1990) used a lasermicrobeam for cutting holes of defined dimen-sions into cell walls of various dicot cells and tis-sues followed by DNA uptake and transient geneexpression as well as stable transformation.

Transformation of small-grain cereals

Barley

Transient gene expressionBarley protoplasts are most easily isolated fromyoung leaves or from an established tissue culture.For transient assays, the mesophyll protoplasts areequally suitable as suspension culture protoplasts,although the former are usually not able to divide

90 MENDEL AND TEERI

Qunker et al, 1987; Mendel et al, 1990). Thisalso means that transient gene expression is notbound to cell division. In electroporated barleymesophyll protoplasts, it has been possible todemonstrate the chloroplast targeting of achimeric nptW gene with a pea transit peptide(Teeri et al, 1989). Also protoplasts from thestarchy endosperm (Lee et al, 1991) and from thealeurone layer (Lee etal, 1989) of developing bar-ley grains were able to show transient expressionof the reporter genes uidA and cat. In the case ofendosperm protoplasts, two endosperm-specificpromoters (of genes coding for wheat glutenin andbarley chymotrypsin inhibitor 2), fused to theuidA coding sequence, could be shown to be func-tional.

While many dicot protoplasts quickly dediffer-entiate and will not represent the tissue from whichthey were isolated, this does not always take placewith cereal protoplasts. Several reports of correctgene regulation in barley protoplasts have beenpublished. Jacobsen & Close (1991) could demon-strate the correct regulation of a barley high-pi a-amylase promoter, fused to uidA as a reportergene, by gibberellic acid (GA3) and abscisic acid(ABA) as tested by transient expression assays inprotoplasts prepared from mature barley aleuronelayers. Similar experiments were performed bySkriver et al. (1991) with barley aleurone proto-plasts in order to delineate hormone-responsiveelements of the ABA-responsive rice rabl6A geneand of the barley a-amylase gene. Salmenkallio etal. (1990) compared the regulation of a low-pi a-amylase promoter of barley fused to nptll as areporter. Correct regulation by GA3 and ABA wasobserved in protoplasts isolated from aleurone lay-ers. In protoplasts of the scutellar epithelium, GA3had no effect on the low level of expression, and inmesophyll protoplasts the gene construct was notexpressed at all, both cases also reflecting the prop-erties of the intact tissues.

By using particle bombardment, it is possible toassay transiently the expression of introducedgenes in cells that still are integral to the plant tis-sue. Lanahan et al. (1992) mapped the GA3response complex in a barley a-amylase gene byperforming transient assays in intact aleuronelayers of a barley grain. Variation inherent inparticle bombardment was reduced by coexpress-ing a second reporter (ubi-luc) and by comparingthe expression of the tested uidA fusion with thisstandard control.

Stable transformationUntil now, there is no known case of genetic-ally proven stably transformed, mature barleyplants. However, transgenic protoplast derivedor bombarded cell cultures containing andexpressing one or more foreign genes have beenobtained (Lazzeri & Lorz, 1990; Lazzeri et al.,1991; Mendel et al., 1990, 1991; Ritala et al.,1993).

Although protoplasts are easily prepared fromvarious barley tissues, only those isolated fromsuspension cultures have been able to divide. Ifthe suspension culture will form embryoids, thereis a chance that the protoplasts will also regenerateinto mature plants. Regeneration of protoplast-derived colonies has been reported by Yan et al.(1990) and Jahne et al. (19916). In principle,transgenic mature barley should emerge fromthese experiments in the near future.

Two problems remain in stable transformationof protoplasts. First, the embryogenic suspensioncultures are not stable, and must be reinitiated fre-quently. Second, they can all be initiated fromonly particular varieties. Therefore, the chances todevelop the protoplast approach into a gene trans-fer system routinely applicable for a wide varietyof genotypes are small.

Embryogenic tissue culture that does not gointo suspension, and from which protoplast isola-tion is not possible, can be initiated from a muchwider selection of barley cultivars (Jahne et al.,I991a,b). These tissue cultures are ideal for parti-cle bombardment, and although no reports oftransgenic barley by this method are published, itprobably will be the most general method of bar-ley transformation.

With barley tissue cultures that do not possessthe capacity for differentiation, it has been easy todemonstrate, by selecting for the stabilization ofan antibiotic resistance marker gene, that foreignDNA delivered by particle bombardment can bestably integrated into barley chromosomes(Mendel et al., 1991; Ritala et al., 1993). Similarresults have been obtained by direct gene transferto protoplasts (Lazzeri et al., 1991; Mendel et al.,1990; Schulze et al., 19916). Stable integration ofone or more copies of the foreign DNA into thehost genome was demonstrated by hybridizationof digested genomic DNA with probes derivedfrom the coding regions of the transferred genes.It is important to realize that this is a major step inbarley transformation and, as mentioned, scien-


tists can already study many problems in trans-genic barley tissue culture cells.

In order to study the ability of barley to produceforeign proteins, we have stably introduced thegene for the seed storage protein vicilin of Viciafaba (genomic sequence and complementaryDNA (cDNA)) under control of the barleyhordein promoter into barley cells in order tostudy its integration, stability and its possibleexpression (K. Greger et al., unpublished data), aswell as a fungal heat stable p-glucanase secreted inan active form from stably transgenic malting bar-ley cells (Aspegren et al., 1992).

Wheat

Transient gene expressionBack in 1986, transient expression of the cat gene(Ou-Lee, Turgeon & Wu, 1986) driven by the35S promoter and of the nptll gene (Werr & Lorz,1986) driven by the maize shl promoter werereported for protoplasts of Einkorn wheat,Triticum monococcum. Oard, Paige & Dvorak(1989) showed for protoplasts of bread wheat,Triticum aestivum, transient expression of a 35S-cat construct. Inclusion of intron 6 of maize adhlgene between the 35S and cat sequence increasedcat expression more than 100-fold. Also proto-plasts from the aleurone layer of developing wheatgrains were able to show transient expression ofthe reporter gene cat (Lee et al., 1989).

Stable transformationThere is already vast experience of cell and tissueculture of wheat, and protoplast isolation fromdifferent explants is routine (e.g. see Morrish etal, 1987). As early as 1985, Lorz, Baker & Schellreported the transfer of the nptll gene into proto-plasts of Triticum monococcum suspension culturecells and the selection of stably transformed calluscolonies that were not morphogenic. As men-tioned above for barley, and for bread wheat, themain bottleneck for obtaining transgenic plants isthe rapid loss of embryogenicity of morphogenicsuspension cultures and the protoplasts derivedfrom them. Again the regeneration is also stronglygenotype dependent. There have been severalattempts to improve the situation, e.g. by usingdifferent protocols for initiating the suspensioncultures (Wang & Nguyen, 1990). Langridge,Lazzeri & Lorz (1991) used immature embryos ofa wheat translocation line carrying the short arm

of rye chromosome IRS instead of its own shortarm of chromosome IB and could define two dis-tinct regions on IRS that enhanced the embryo-genicity of callus cultures derived from it.

Hayashi & Shimamoto (1988), isolating proto-plasts directly from immature embryos, reportedthe formation of roots and albino shoots fromprotoplast-derived callus colonies (cv. ChineseSpring). Harris et al. (1988), using an anther-derived suspension culture of the wheat cultivarChris, obtained rare green plantlets from proto-plasts. In 1990, Vasil's group reported the firstsuccessful case of regenerating green plants fromprotoplasts isolated from immature-embryo-derived suspension cultures (Vasil et al., 1990).After flowering, however, these otherwise healthyplants showed the phenomenon of precocioussenescence so that no progeny could be obtained.Also Chang et al. (1991) were able to regenerate afew green plants from suspension-culture-derivedprotoplasts (cv. Mustang). So, in essence all pre-requisites are given for generating the first trans-genic protoplast-derived wheat plants. However,as yet there is no such case known.

As for barley, it is rather unlikely that the proto-plast approach will develop into a gene transfersystem routinely applicable for a wide variety ofwheat genotypes. The particle bombardmentmethod remains the most promising alternative.Tissue culture cells of wheat have been trans-formed to express foreign marker genes (Vasil etal., 1991). Also, embryogenic tissue cultures,capable of regeneration into mature plants, havebeen available (Vasil et al., 1990). These twomethods were recently combined. Vasil et al.(1992) were able to obtain transgenic, fertile and(for the transgenes) true breeding wheat plantsfrom two varieties of Triticum aestivum. They usedas selectable marker the bar gene, coding for phos-phinothricin resistance, in the control of the 35Spromoter and enhanced with the maize adhlintron 1. Although the nonselected marker uidAcontrolled by the adhl promoter was expressedonly weakly, this work, together with that donewith oat described below, represent a major break-through in small-grain cereal transformation.

Oat

Using the particle gun approach, Kakuta et al.(1992) introduced 35S-uidA into embryo cells ofimbibed oat and wheat seeds. Two to 7 days after

92 MENDEL AND TEERI

bombardment, transient uidA expression wasmonitored in cells of the embryos, stems leavesand roots of the seedlings.

Most remarkably, fertile transgenic oat plantswere reported recently by Somers et al. (1992).They used in their experiments an oat line thathad been developed for the ability to produceregenerable, embryogenic tissue cultures. Thesecultures were bombarded with 35S-bar and adhl-uidA constructs, and selected for herbicide resis-tance. Thirty-eight transgenic plants could beregenerated and grown to maturity. One of themwas fully fertile, and transferred both markers toits progeny, where they were expressed. Co-expression of the nonselected uidA marker (GUSactivity) was observed in 75% of the transgenictissue culture lines.

Rye

Not very much work towards transformation ofrye has been published, except for that by De laPena et al. (1987), where the macroinjectionapproach was first tried for a small-grain cereal.De la Pena et al. described injection of DNA (withthe marker gene nptIT) into the tiller cavity ofimmature rye tillers. After screening the progenyfor kanamycin resistance, among 3023 resultinggrains three plants were obtained that grew in thepresence of kanamycin and expressed NPT IIactivity in the leaves. In genomic DNA, Southernblots revealed appropriately hybridizing restric-tion fragments. However, no transmission to theprogeny was demonstrated. Later, preliminaryexperiments showed that the progeny plants con-tained wpdl-specific hybridizing fragments; how-ever, no correct segregation was observed(H. Lorz, personal communication).

Orchardgrass (Dactylis glomerata)

Horn et al. (1988) transferred the hygromycinresistance gene under control of the 35S promoterto protoplasts of orchardgrass and regeneratedgreen, hygromycin-resistant plants that exhibitedthe correct hybridization pattern in a Southernanalysis of genomic DNA.

Guinea grass (Panicum maximum)

Vasil et al. (1988) optimized DNA transfer intoprotoplasts of Guinea grass using a 35S-cat con-

struct and transient gene expression, andHauptmann et al. (1988) obtained stably trans-formed cell lines with 35S-hph as a selectablemarker.

Sorghum (Sorghum vulgare)

Hagio, Blowers & Earle (1991) transformed sus-pension cell cultures of sorghum using the particlegun approach and selected cell colonies stablyexpressing hph and nptll (driven by the 35S pro-moter). The cotransferred 35S-uidA wasexpressed at low levels. Also Blowers, Hagio &Earle (1991) successfully used the particle gun toobtain stably transformed cell cultures (wprll, hph>uidA). Battraw & Hall (1991) transformed proto-plasts of Sorghum bicolor with the genes for NPT IIand GUS and obtained stable transformants.

Turfgrass (Festuca arundinacea)

Ha, Wu & Thorne (1992) transformed tall fescueprotoplasts with the gene for HPT and selectedhygromycin-resistant cell lines that could beregenerated into plants. Using a similar approachand the marker genes hph and bar, Takamizo et al.(1992) obtained transgenic tall fescue plantsexpressing the transgenes.

Discussion

There are suitable promoters and effective selec-table markers available for transforming cells ofsmall-grain cereals and generating stably trans-formed cell cultures. Also the physical introduc-tion of foreign DNA into a recipient cell does notseem to be problematic. However, the followingsteps - stable integration of the introduced DNAinto the host genome and regeneration of thetransformed cells to mature plants - seem to causeproblems. Most importantly this is a problem oftissue culture technology. A successful transfor-mation method requires that a large number ofrecipient cells can be induced to regenerate intomature plants. It might also be speculated thatsmall-grain cereals could be more efficient thanother species in recognizing the introduced DNAto be 'foreign' (e.g. by screening for the methyla-tion pattern of the incoming gene) and activatingmechanisms that ultimately lead either to the exci-sion of the foreign DNA or to its irreversible


repression. However, being successfully intro-duced into dividing cells, the foreign DNA is atleast in some cases able to give high and stableexpression (e.g. nptll driven by 35S in barley tis-sue culture). Yet we do not know whether or notthe expression will change during the process ofplant regeneration.

Furthermore, we do not know whether thereare special requirements for efficient gene expres-sion in cereal cells. It is well established that manyhighly expressed nuclear genes of cereals andother monocots are characterized by the occur-rence of so called CpG islands within their pro-moter region and within the 5'-part of the codingsequence, whereas their intron regions do notexhibit these characteristics (Martinez, Martin &Cerff, 1989). However, all the marker andreporter genes commonly used are devoid of CpGislands. It remains to be established whether thisstructural characteristic will be of importance forthe genetic stability of a transferred gene duringsexual transmission and for the stability in theprogeny. Other factors should also be taken intoconsideration when one is planning to transfer agiven gene into a cereal species: splicing mecha-nisms of pre-mRNAs might be slightly differentbetween cereals and dicots. This was indicated bya report of Keith & Chua (1986), who demon-strated that the wheat rbcS gene was not correctlyspliced in tobacco. In another case, however, thegene for sorghum phosphoenolpyruvate carboxy-lase was correctly spliced in the dicot.

In summary, a large number of approaches hasbeen tried to stably transform small-grain cereals,with the aim of obtaining green fertile plants. Thisgoal has been reached for wheat and oat for thefirst time recently. Besides the protoplastapproach and the particle bombardment method,there are numerous techniques belonging to thegroup of nonorthodox methods; however, theresults are far from being reproducible and pre-dictable. Most importantly, although there is evi-dence indicative of stabilization and correctexpression of the foreign DNA transferred bythese approaches, the DNA is rarely transmittedto the progeny. Meiotic divisions seem to elimi-nate the foreign genetic material. Thus, only thetransfer of DNA into protoplasts and the micro-projectile-mediated transfer of genes throughintact cell walls can at the moment be recom-mended for approaching the problem of stablytransforming small-grain cereals. But these two

methods have bottlenecks that have to be taken intoaccount: with the protoplast approach it is the lossof the regenerative potential of the selected trans-genic colonies, and with the particle bombardmentit is the question of which is the appropriate targettissue and which is the most suitable system ofselection. Nevertheless, both of these methods havea chance to develop into reliable methods of generalapplicability to small-grain cereals.

Recent developments

This chapter is based on literature that was avail-able in the beginning of 1993. The field is devel-oping rapidly and now, one year later, morebreakthroughs have taken place. Barley, regener-ating to fertile plants, has been transformed byone of us (Ritala et al., 1994) and others (Wan &Lemaux, 1994), and sorghum by Casas et al.(1993). After the first reported success in a partic-ular plant, it is necessary that transformation isrepeated by others and made more efficient. Thishas taken place for wheat (Vasil et ah> 1993;Weeks, Anderson & Blechl, 1993; Becker,Brettschneider & Lorz, 1994; Nehra et a/., 1994).Common to the recent accomplishments has beenthe use of particle bombardment as method ofDNA delivery, the bar gene as a selectable marker,and of immature embryos as targets. This gives aclear guideline for development of transformationsystems for the remaining small-grain cerealspecies. A long and laborious phase of methoddevelopment in cereal transformation has appar-ently come to an end. The exciting part of thework may begin.

References

Abdul-Baki, A. A., Saunders, J. A., Matthews, B. F. &Pittarelli, G. W. (1990). DNA uptake duringelectroporation of germinating pollen grains. PlantScience, 70, 181-190.

Ahloowalia, B. S. (1987). Plant regeneration fromembryo-callus culture in barley. Euphytica, 36,659-665.

Ahokas, H. (1987). Transfection by DNA-associatedliposomes evidenced at pea pollination. Hereditas,106, 129-138.

Ahokas, H. (1989). Transfection of germinating barleyseed electrophoretically with exogenous DNA.Theoretical and Applied Genetics, 11, 469-472.

94 MENDEL AND TEERI

Aspegren, K., Ritala, A., Mannonen, L., Hannus, R.,Kurten, U., Salmenkallio, M., Kauppinen, V. &Teeri, T. H. (1992). Secretion of a fungal heatstable (3-glucanase from tobacco and barley cells.Progress in Molecular Biology (NOMBA Symposium),abstract 2, Helsinki.

Battraw, M. & Hall, T. C. (1991). Stabletransformation of Sorghum bicolor protoplasts withchimeric neomycin phosphotransferase II and (3-glucuronidase genes. Theoretical and Applied Genetics,82, 161-168.

Battraw, M. & Hall, T. C. (1992). Expression of achimeric neomycin phosphotransferase II gene infirst and second generation transgenic rice plants.Plant Science, 86, 191-202.

Becker, D., Brettschneider, R. & Lorz, H. (1994).Fertile transgenic wheat from microprojectilebombardment of scutellar tissue. Plant Journal, 5,299-307.

Blowers, A. D., Hagio, T. & Earle, E. D. (1991).Isolation of autonomously-replicating plasmids fromtransformed Sorghum cells. In Molecular Biology ofPlant Growth and Development, Third InternationalCongress of Plant Molecular Biology, ed. R. B. Hallick,abstract 921. University of Arizona, Tucson.

Bregitzer, P., Bushnell, W. R., Rines, H. W. &Somers, D. A. (1991). Callus formation and plantregeneration from somatic embryos of oat (Avenasativa L.). Plant Cell Reports, 10, 243-246.

Brettschneider, R., Becker, D. & Lorz, H. (1991).Transient gene expression in cereal cells aftermicroprojectile bombardment. In InternationalAssociation of Plant Cell and Tissue Culture, Meeting ofthe German Section, abstract vl5. University ofHamburg, Hamburg.

Budar, F., Thia-Toong, L., Van Montagu, M. &Hernalsteens, J.-P. (1986). Agrobacterium-mediatedgene transfer results mainly in transgenic plantstransmitting T-DNA as a single mendelian factor.Genetics, 114, 303-313.

Callis, J., Fromm, M. & Walbot, V. (1987). Intronsincrease gene expression in cultured maize cells.Genes and Development, 1, 1183-1200.

Casas, A. M., Kononowicz, A. K., Zehr, U. B.,Tomes, D. T., Axtell, J. D., Butler, L. G., Bressan,R. A. & Hasegawa, P. M. (1993). Transgenicsorghum plants via microprojectile bombardment.Proceedings of the National Academy of Sciences, USA,90, 11212-11216.

Chang, Y.-F., Wang, W. C , Warfield, C. Y., Nguyen,H. T. & Wong, J. R. (1991). Plant regenerationfrom protoplasts isolated from long-term cellcultures of wheat (Triticum aestivum L.). Plant CellReports, 9, 611-614.

Chen, D. F., Dale, P. J. & Mullineaux, P. M. (1991).Transformation of Triticum monococcum L. cells andanalysis of long term stability of foreign genes. In

Molecular Biology of Plant Growth and Development,Third International Congress of Plant MolecularBiology, ed. R. B. Hallick, abstract 418. University ofArizona, Tucson.

De la Pena, A., Lorz, H. & Schell, J. (1987).Transgenic rye plants obtained by injecting DNAinto young floral tillers. Nature, 325, 274-276.

De Wet, J. M. J., Bergquist, R. R., Harlan, J. R.,Brink, D. E., Cohen, C. E., Newell, C. A. & DeWet, A. E. (1985). Exogenous gene transfer inmaize (Zea mays) using DNA-treated pollen. InExperimental Manipulation of Ovule Tissue, ed. G. P.Chapman, S. H. Mantell & W. Daniels, pp.197-209. Longman, London.

Dekeyser, R. A., Claes, B., De Rycke, R. M. U.,Habets, M. E., Van Montagu, M. C. & Caplan,A. B. (1990). Transient gene expression in intactand organized rice tissues. Plant Cell, 2, 591-602.

Delporte, F., Mingeo, C , Salomon, F. & Jacquemin,J.-M. (1991). Expression of zein family protein inwheat endosperm (Triticum aestivum L.) aftertransformation by copollination with plasmidicDNA. In Eucarpia Symposium on GeneticManipulation in Plant Breeding, abstract I/P4.Tarragona.

D'Halluin, K., Bonne, E., Bossut, M., De Beuckeleer,M. & Leemans, J. (1992). Production of fertiletransgenic maize plants by electroporation ofimmature embryos and type I callus. Plant Cell, 4,1495-1505.

Doy, C. H., Gresshoff, P. M. & Rolfe, B. G. (1973).Biological and molecular evidence for thetransgenesis of genes from bacteria to plant cells.Proceedings of the National Academy of Sciences, USA,70, 723-726.

Du, J., Liu, H., Xie, D. X., Hua, X. J. & Fan, Y. L.(1989). Injection of exogenous DNA into youngfloral tillers of wheat. Genetic Manipulation in Plants,5, 8-12.

Gould, J., Devey, M., Hasegawa, O., Ulian, E. C ,Peterson, G. & Smith, R. H. (1991).Transformation of Zea mays L. using Agrobacteriumtumefaciens and the shoot apex. Plant Physiology, 95,426-434.

Gould, J. H., Devey, M. E., Ko, T. S., Peterson, G.,Hasegawa, O. & Smith, R. H. (1992).Transformation of Gramineae using Agrobacteriumtumefaciens. Journal of Cellular Biochemistry, 16F,207.

Grimsley, N., Hohn, T., Davies, J. W. & Hohn, B.(1987). Agrobacterium-mediated delivery ofinfectious maize streak virus into maize plants.Nature (London), 325, 177-179.

Ha, S. B., Wu, F.-S. & Thome, T. K. (1992).Transgenic tall fescue (Festuca arundinacea Schreb.)plants regenerated from protoplasts. Journal ofCellular Biochemistry, 16F, 207.


Hagio, T., Blowers, A. D. & Earle, E. D. (1991).Stable transformation of sorghum cell cultures afterbombardment with DNA-coated microprojectiles.Plant Cell Reports, 10, 260-264.

Harris, R., Wright, M., Byrne, M., Varnum, J.,Brightwell, B. & Schubert, K. (1988). Callusformation and plantlet regeneration from protoplastsderived from suspension cultures of wheat {Triticumaestivum L.). Plant Cell Reports, 7, 337-340.

Hauptmann, R. M., Vasil, V., Ozias-Akins, P.,Tabaeizadeh, Z., Rogers, S. G., Fraley, R. T.,Horsch, R. B. & Vasil, I. K. (1988). Evaluation ofselectable markers for obtaining stable transformantsin the Gramineae. Plant Physiology, 86, 602-606.

Hayashi, Y. & Shimamoto, K. (1988). Wheatprotoplast culture: embryogenic colony formationfrom protoplasts. Plant Cell Reports, 7, 414-417.

Heberle-Bors, E. (1991). Germ line transformation inhigher plants. Newsletter of the InternationalAssociation of Plant Cell and Tissue Culture, 64, 2-10.

Heberle-Bors, E., Benito Moreno, R. M., Alwen, A.,Stoger, E. & Vincente, O. (1990). Transformationof pollen. In Progress in Plant Cellular and MolecularBiology, ed. H. J. J. Nijkamp, L. H. W. van der Plas& J. van Aartrijk, pp. 244-251. Kluwer AcademicPublishers, Dordrecht.

Hernalsteens, J.-P., Thia-Toong, L., Schell, J. & VanMontagu, M. (1984). An Agrobacterium-transformedcell culture from the monocot Asparagus offtcinalis.EMBO Journal, 3, 3039-3041.

Hess, D. (1987). Pollen-based techniques in geneticmanipulation. International Review of Cytology, 107,367-395.

Hess, D., Dressier, K. & Nimmrichter, R. (1990).Transformation experiments by pipettingAgrobacterium into the spikelets of wheat {Triticumaestivum L.). Plant Science, 72, 233-244.

Hinchee, M. A. W., Fry, J. E., Sanders, P. R.,Armstrong, C. L., Petersen, W. L., Songstad, D. D.,Mathis, N. L., Muskopf, Y. M., Barnason, A. R.,Laytom, J. G., Corbin, D. R., Connor-Ward, D. V.& Horsch, R. B. (1991). Experimental gene transfersystems for plants. In Molecular Biology of PlantGrowth and Development, Third International Congressof Plant Molecular Biology, ed. R. B. Hallick, abstract63. University of Arizona, Tucson.

Hooyakaas-Van Slogteren, G. M. S., Hooyakaas,P. J. J. & Schilperoort, R. A. (1984). Expression ofTi plasmid genes in monocotyledonous plantsinfected with Agrobacterium tumefaciens. Nature{London), 311, 763-764.

Horn, M. E., Shillito, R. D., Conger, B. V. & Harms,C. T. (1988). Transgenic plants of orchardgrass{Dactylis glomerata L.) from protoplasts. Plant CellReports, 7, 469-472.

Jacobsen, J. V. & Close, T. J. (1991). Control oftransient expression of chimaeric genes by

gibberellic acid and abscisic acid in protoplastsprepared from mature barley aleurone layers. PlantMolecular Biology, 16, 713-724.

Jahne, A., Lazzeri, P. A., Jager-Gussen, M. & Lorz, H.(1991a). Plant regeneration from embryogenic cellsuspensions derived from anther cultures of barley{Hordeum vulgare L.). Theoretical and AppliedGenetics, 82, 74-80.

Jahne, A., Lazzeri, P. A. & Lorz, H. (19916).Regeneration of fertile plants from protoplastsderived from embryogenic suspensions of barley{Hordeum vulgare L.). Plant Cell Reports, 10, 1-6.

Joersbo, M. & Brunstedt, J. (1990). Direct genetransfer to plant protoplasts by mild sonication.Plant Cell Reports, 9, 207-210.

Johnson, C. B., Grierson, D. & Smith, H. (1973).Expression of \plac5 DNA in cultured cells of ahigher plant. Nature New Biology, 244, 105-107.

Junker, B., Zimny, J., Luhrs, R. & Lorz, H. (1987)Transient expression of chimaeric genes in dividingand non-dividing cereal protoplasts after PEG-induced DNA uptake. Plant Cell Reports, 6,329-332.

Kaeppler, H. F., Gu, W., Somers, D. A., Rines, H. W.& Cockburn, A. F. (1990). Silicon carbide fiber-mediated DNA delivery into plant cells. Plant CellReports, 9, 415-418.

Kakuta, H., Hashidoko, Y., Seki, T., Matsui, M.,Anai, T., Hasegawa, K. & Mizutani, J. (1992). Insitu foreign gene expression in seedlings of somemonocotyledonous crops using the improvedparticle gun driven by compressed N2 gas. Journal ofCellular Biochemistry, 16F, 237.

Kaneko, T., Kuroda, H., Hirota, N., Kishinami, I.,Kimura, N. & Ito, K. (1991). Two transformationsystems of barley - electroporation and laserperforation. In Barley Genetics vi Vol. 1, ed. L.Munck, K. Kirkegaard & B. Jensen, pp. 231-233.Munksgaard International Publishers, Copenhagen.

Kartel, N. A., Zabenkova, K. I., Maneshina, T. V. &Ablov, S. E. (1990). Barley plants with introducedgene for kanamycin resistance. Doklady AkademiiNauk BSSR, 34 (N3), 261-263.

Keith, B. & Chua, N.-H. (1986). Monocot and dicotpre-mRNAs are processed with different efficienciesin transgenic tobacco cells. EMBO Journal, 5,2419-2425.

Klein, T. M., Harper, E. C , Svab, Z., Sanford, J. C ,Fromm, M. E. & Maliga, P. (1988). Stable genetictransformation of intact Nicotiana cells by theparticle bombardment process. Proceedings of theNational Academy of Sciences, USA, 85, 8502-8505.

Konstantinov, K. L., Mladenovic, S. D., Denic, M. P.& Plecas, M. (1990). Maize genome and genome ofbacteria living in maize - transformation with abacterial gene. In VI NATO Advanced Studies - PlantMolecular Biology, p. 99. Schloss Elmau, Bavaria.

96 MENDEL AND TEERI

Lanahan, M. B., Ho, T.-H. D., Rogers, S. W. &Rogers, J. C. (1992) A gibberellin response complexin cereal a-amylase gene promoters. Plant Cell, 4,203-211.

Langridge, P., Lazzeri, P. & L6rz, H. (1991). Asegment of rye chromosome 1 enhances growth andembryogenesis of calli derived from immatureembryos of wheat. Plant Cell Reports, 10, 148-151.

Last, D. I., Brettell, R. I. S., Chamberlain, D. A.,Chaudhury, A. M., Larkin, P. J., Marsh, E. L.,Peacock, W. J. & Dennis, E. S. (1991). pEmu: animproved promoter for gene expression in cerealcells. Theoretical and Applied Genetics, 81, 581-588.

Lazzeri, P. A., Brettschneider, R., Luhrs & Lorz, H.(1991). Stable transformation of barley via PEG-induced direct DNA uptake into protoplasts.Theoretical and Applied Genetics, 81, 437-444.

Lazzeri, P. A. & Lorz, H. (1990). Regenerablesuspension and protoplast cultures of barley andstable transformation via DNA uptake intoprotoplasts. In Genetic Engineering in Crop Plants, ed.G. W. Lycett & D. Grierson, pp. 231-238.Butterworths, London.

Lazzeri, P. A. & Lorz, H. (1991). Protoplast cultureand transformation in barley. In Barley Genetics vi,Vol. 1, ed. L. Munck, K. Kirkegaard & B. Jensen,pp. 240-244. Munksgaard International Publishers,Copenhagen.

Lee, B. T., Murdoch, K., Topping, J., Jones, M. G. K.& Kreis, M. (1991). Transient expression of foreigngenes introduced into barley endosperm protoplastsby PEG-mediated transfer or into intact endospermtissue by microprojectile bombardment. PlantScience, 78, 237-246.

Lee, B., Murdoch, K., Topping, J., Kreis, M. & Jones,M. G. K. (1989). Transient gene expression inaleurone protoplasts isolated from developingcaryopses of barley and wheat. Plant MolecularBiology, 13, 21-39.

Li, X.-Q., Liu, C , Peng, J., Gelvin, S. B. & Hodges,T. K. (1991). GUS expression in rice tissues usingAgrobacterium-mediated transformation. In MolecularBiology of Plant Growth and Development, ThirdInternational Congress of Plant Molecular Biology, ed.R. B. Hallick, abstract 385. University of Arizona,Tucson.

Lindsey, K. & Jones, M. G. K. (1987). Transient geneexpression in electroporated protoplasts and intactcells of sugar beet. Plant Molecular Biology, 10,43-52.

Lorz, H., Baker, B. & Schell, J. (1985). Gene transferto cereal cells mediated by protoplasttransformation. Molecular and General Genetics, 199,178-182.

Louwerse, J. D., Gersman, M., van Dam, K. J.,Hensgens, L. A. M. & van der Mark, F. (1991).Towards the stable transformation of barley using

particle bombardment. In Molecular Biology of PlantGrowth and Development, Third International Congressof Plant Molecular Biology, ed. R. B. Hallick, abstract910. University of Arizona, Tucson.

Luhrs, R. & Lorz, H. (1987a). Initiation ofmorphogenic cell-suspension and protoplast culturesof barley. Planta, 175, 71-81.

Luhrs, R. & Lorz, H. (1987b). Plant regeneration invitro from embryogenic cultures of spring- andwinter-type barley (Hordeum vulgare L.) varieties.Theoretical and Applied Genetics, 75, 16-25.

Maas, C , Laufs, J., Grant, S., Korfhage, C. & Werr,W. (1991). The combination of a novel stimulatoryelement in the first exon of the maize Shrunken-1gene with the following intron 1 enhances reportergene expression up to 1000-fold. Plant MolecularBiology, 16, 199-207.

Martinez, P., Martin, W. & Cerff, R. (1989).Structure, evolution and anaerobic regulation of anuclear gene encoding cytosolic glyceraldehyde-3-phosphate dehydrogenase from maize. Journal ofMolecular Biology, 208, 551-565.

McElroy, D., Blowers, A. D., Jenes, B. & Wu, R.(1991). Construction of expression vectors based onthe rice actin 1 (Actl) 5' region for use in monocottransformation. Molecular and General Genetics, 231,150-160.

McElroy, D., Zhang, W., Cao, J. & Wu, R. (1990).Isolation of an efficient actin promoter for use in ricetransformation. Plant Cell, 2, 163-171.

Mejza, S. J., Nguyen, T., Driller, J. P., Walker, L. S.,Ulrich, T. & Wong, J. R. (1991). Stable co-transformation of genes in Triticum aestivum cellsafter particle bombardment. In Molecular Biology ofPlant Growth and Development, Third InternationalCongress of Plant Molecular Biology, ed. R. B. Hallick,abstract 411. University of Arizona, Tucson.

Mendel, G. (1865). Versuche iiber Pflanzenhybriden.Verhandlungen des naturforschenden Vereines in Bru'nn,4, 3-47.

Mendel, R. R., Clauss, E., Hellmund, R., Schulze, J.,Steinbiss, H. H. & Tewes, A. (1990). Gene transferto barley. In Progress in Plant Cellular and MolecularBiology, ed. H. J. J. Nijkamp, L. H. W. van der Plas& J. van Aartrijk, pp. 73-78. Kluwer AcademicPublishers, Dordrecht.

Mendel, R. R., Miiller, B., Schulze, J., Kolesnikov, V.& Zelenin, A. (1989). Delivery of foreign DNA tointact barley cells by high-velocity microprojectiles.Theoretical and Applied Genetics, 78, 31-34.

Mendel, R. R., Schulze, J., Clauss, E., Nerlich A.,Cerff, R. & Steinbiss, H. H. (1991). Gene transferto barley. In Barley Genetics vi, Vol. 1, ed. L.Munck, K. Kirkegaard & B. Jensen, pp. 240-244.Munksgaard International Publishers, Copenhagen.

Morikawa, H., Iida, A., Matsui, C , Ikegami, M. &Yamada, Y. (1986). Gene transfer into intact plant


cells by electroinjection through cell walls andmembranes. Gene, 41, 121-124.

Morrish, F. M., Vasil, V. & Vasil, I. K. (1987).Developmental morphogenesis and geneticmanipulation in tissue and cell cultures of theGramineae. Advances in Genetics, 24, 431-499.

Miiller, B., Schulze, J. & Wegner, U. (1989).Establishment of barley cell suspension cultures ofmesocotyl origin suitable for isolation of dividingprotoplasts. Biochemie und Physiologie der Pflanzen,185, 123-130.

Nehra, N. S., Chibbar, R. N., Leung, N., Caswell, K.,Mallard, C , Steinhauer, L., Baga, M. & Kartha,K. K. (1994). Self-fertile transgenic wheat plantsregenerated from isolated scutellar tissues followingmicroprojectile bombardment with two distinct geneconstructs Plant Journal, 5, 284-297.

Oard, J. H., Paige, D. & Dvorak, J. (1989). Chimericgene expression using maize intron in cultured cellsofbreadwheat. Plant Cell Reports, 8, 156-160.

Ohta, U. (1986). High efficiency genetictransformation of maize by a mixture of pollen andexogenous DNA. Proceedings of the National Academyof Sciences, USA, 83, 715-719.

Ou-lee, T. M., Turgeon, R. & Wu, R. (1986).Expression of a foreign gene linked to either a plant-virus or a Drosophila promoter, after electroporationof protoplasts of rice, wheat and sorghum.Proceedings of the National Academy of Sciences, USA,83, 6815-6819.

Picard, E., Jacquemin, J. M., Granier, F., Bobin, M. &Forgeois, P. (1988). Genetic transformation ofwheat (Triticum aestivum) by plasmid DNA uptakeduring pollentube germination. In vmh InternationalWheat Genetics Symposium, pp. 779-781. CambridgeUniversity Press, Cambridge.

Potrykus, I. (1990). Gene transfer to cereals: anassessment. Bio/Technology, 8, 535-542.

Potrykus, I. (1991). Gene transfer to plants:assessment of published approaches and results.Annual Review of Plant Physiology and PlantMolecular Biology, 42, 205-225.

Raineri, D. M., Bottino, P., Gordon, M. P. & Nester,E. W. (1990) Agrobacterium-mediatedtransformation of rice (Oryza sativa L.).Bio/Technology, 8, 33-38.

Rhodes, C. A., Pierce, D. A., Mettler, I. J., Mascarenhas,D. & Detmer, J. J. (1988). Genetically transformedmaize plants from protoplasts. Science, 240, 204-207.

Ritala A., Aspegren K., Kurten U., Salmenkallio-Marttila M., Mannonen L., Hannus, R.,Kauppinen, V., Teeri, T. H. & Enari, T.-M. (1994).Fertile transgenic barley by particle bombardment ofimmature embryos. Plant Molecular Biology, 24,317-325.

Ritala, A., Mannonen, L., Salmenkallio, M., Kurten,U., Hannus, R., Aspegren, K., Mendez-Lozano, J.,

Teeri, T. H. & Kauppinen, V. (1993). Stabletransformation of barley tissue culture by particlebombardment. Plant Cell Reports, 12, 435-440.

Rogers, S. W. & Rogers, J. C. (1992). The importanceof DNA methylation for stability of foreign DNA inbarley. Plant Molecular Biology, 18, 945-962.

Salmenkallio, M., Hannus, R., Teeri, T. H. &Kauppinen, V. (1990). Regulation of a-amylasepromoter by gibberellic acid and abscisic acid inbarley protoplasts transformed by electroporation.Plant Cell Reports, 9, 352-355.

Sautter, C , Waldner, H., Neuhaus-Url, G., Galli, A.,Neuhaus, G. & Potrykus, I. (1991). Micro-targeting:high efficiency gene transfer using a novel approachfor the acceleration of micro-projectiles.Bio/Technology, 9, 1080-1085.

Schulze, J. & Mendel, R. R. (1991). Effect of 5-azacytidine on the expression of foreign genes instably transformed barley cell lines. In InternationalAssociation of Plant Cell and Tissue Culture, Meeting ofthe German Section, abstract vl2. University ofHamburg, Hamburg.

Schulze, J., Nerlich, A. & Mendel, R. R. (1991a).Effect of 5-azacytidine on the expression of foreigngenes in stably transformed barley. In MolecularBiology of Plant Growth and Development, ThirdInternational Congress of Plant Molecular Biology, ed.R. B. Hallick, abstract 360. University of Arizona,Tucson.

Schulze, J., Nerlich, A., Ryschka, S., Steinbiss, H. H.& Mendel, R. R. (19916). Stable transformation ofbarley using a highly-efficient protoplast system.Physiologia Plantarum, 82, A3 2.

Schulze, J., Nerlich, A., Ryschka, S., Steinbiss, H. H.& Mendel, R. R. (1991c). Expression of foreigngenes in stably transformed barley. In EucarpiaSymposium on Genetic Manipulation in Plant Breeding,abstract I/P16. Tarragona.

Skriver, K., Olsen, F. L., Rogers, J. C. & Mundy, J.(1991). Czs-acting DNA elements responsive togibberellin and its antagonist abscisic acid.Proceedings of the National Academy of Sciences, USA,88, 7266-7270.

Somers, D. A., Rines, H. W., Gu, W., Kaeppler, H. F.& Bushnell, W. R. (1992). Fertile, transgenic oatplants. Bio/Technology, 10, 1589-1594.

Spickernagel, R., Lazzeri, P. A., Liitticke, S. & Lorz,H. (1991). Transient marker gene expression fromdifferent promoter constructs after PEG mediatedDNA uptake into barley protoplasts. In InternationalAssociation of Plant Cell and Tissue Culture, Meeting ofthe German Section, abstract 39. University ofHamburg, Hamburg.

Stachel, S. E. & Zambryski, P. C. (1986).Agrobacterium tumefaciens and the susceptible plantcell: a novel adaptation of extracellular recognitionand DNA conjugation. Cell, 47, 155-157.

98 MENDEL AND TEERI

Svab, Z., Hajdukiewicz, P. & Maliga, P. (1990). Stabletransformation of plastids in higher plants.Proceedings of the National Academy of Sciences, USA,87, 8526-8530.

Takamizo, T., Wang, Z., Suginobu, K., Perez-Vincente, R., Potrykus, I. & Spangenberg, G.(1992). Biotechnological approaches to forage grassimprovement: Festuca and Lolium. Journal of CellularBiochemistry, 16F, 212.

Teeri, T. H., Patel, G. H., Aspegren, A. & Kauppinen,V. (1989). Chloroplast targeting of neomycinphosphotransferase II with a pea transit peptide inelectroporated barley mesophyll protoplasts. PlantCell Reports, 8, 187-190.

Topfer, R., Gronenborn, B., Schell, J. & Steinbiss,H.-H. (1989). Uptake and transient expression ofchimeric genes in seed-derived embryos. Plant Cell,1, 133-139.

Topfer, R., Maas, C , Horicke-Grandpierre, C ,Schell, J. & Steinbiss, H. H. (1992). Expressionvectors for high-level gene expression indicotyledonous and monocotyledonous plants. InMethods in Enzymology, vol. 217, ed. R. Wu, pp.66-78. Academic Press, New York.

Toriyama, K., Arimoto, Y., Uchimiya, H. & Hinata,K. (1988). Transgenic rice plants after direct genetransfer into protoplasts. Bio/Technology, 6, 1072-1074.

Toyoda, H., Yamaga, T., Matsuda, Y. & Ouchi, S.(1990). Transient expression of the (3-glucuronidasegene introduced into barley coleoptile cells bymicroinjection. Plant Cell Reports, 9, 299-302.

Twell, D., Klein, T. M., Fromm, M. E. &McCormick, S. (1989). Transient expression ofchimeric genes delivered into pollen bymicroprojectile bombardment. Plant Physiology, 91,1270-1274.

Vasil, V., Brown, S. M., Re, D., Fromm, M. E. &Vasil, I. K. (1991). Stably transformed callus linesfrom microprojectile bombardment of cellsuspension cultures of wheat. Bio/Technology, 9,743-747.

Vasil, V., Castillo, A. M., Fromm, M. E. & Vasil, I. K.(1992). Herbicide resistant fertile transgenic wheatplants obtained by microprojectile bombardment ofregenerable embryogenic callus. Bio/Technology, 10,667-674.

Vasil, V., Hauptmann, R., Morrish, F. M. & Vasil,I. K. (1988). Comparative analysis of free DNAdelivery and expression into protoplasts of Panicummaximum Jacq. (Guinea grass) by electroporationand polyethylene glycol. Plant Cell Reports, 1,499-503.

Vasil, V., Redway, F. & Vasil, I. K. (1990).Regeneration of plants from embryogenicsuspension culture protoplasts of wheat (Triticumaestivum L.). Bio I Technology, 8, 429-434.

Vasil, V., Srivastava, V., Castillo, A. M., Fromm,M. E. & Vasil, I. K. (1993). Rapid production oftransgenic wheat plants by direct bombardment ofcultured immature embryos. Bio I Technology, 11,1553-1558.

Wan, Y. & Lemaux, P. G. (1994). Generation of largenumbers of independently transformed fertile barleyplants. Plant Physiology, 104, 37-48.

Wang, W. C. & Nguyen, H. T. (1990). A novelapproach for efficient regeneration from long-termsuspension culture of wheat. Plant Cell Reports, 8,639-642.

Weber, G., Monejambashi, S., Wolfrum, J. &Greulich, K.-O. (1990). Genetic changes induced inhigher plant cells by a laser microbeam. PhysiologiaPlantarum, 79, 190-193.

Weeks, J. T., Anderson, O. D. & Blechl, A. E. (1993).Rapid production of multiple independent lines offertile transgenic wheat (Triticum aestivum). PlantPhysiology, 102, 1077-1084.

Werr, W. & Lorz, H. (1986). Transient geneexpression in a Gramineae cell line: a rapidprocedure to study plant promoters. Molecular andGeneral Genetics, 202, 368-373.

Yan, Q., Zhang, X., Shi, J. & Li, J. (1990). Long-termcallus cultures of diploid barley (Hordeum vulgareL.). Kexue tonqbao, 35, 1581-1583.

Zhang, L.-J., Cheng, L.-M., Xu, N., Zhao, N.-M., Li,C.-G., Yuan, J. & Jia S.-R. (1991). Efficienttransformation of tobacco by ultrasonication.Bio/Technology, 9, 996-997.

Zhong-xun, L. & Wu, R. (1988). A simple method forthe transformation of rice via the pollen-tubepathway. Plant Molecular Biology Reporter, 6,165-174.

PART III TRANSFORMATION OFNDUSTRIALLYMPORTANT CROPS

8Leguminous PlantsJack M. Widholm

Introduction

Many of the most important grain and foragecrops of the world are legumes so there has beenan interest in their tissue culture and transforma-tion for many years. Since legumes are dicotyle-dons (dicots), they are in general very susceptibleto infection by both Agrobacterium tumefaciens andAgrobacterium rhizogenes, so these systems havebeen used extensively.

Legumes in general and grain legumes in partic-ular have been recalcitrant as far as plant regener-ation from tissue culture is concerned. This hasmade many of the transformation techniques diffi-cult, and so has delayed the overall progress.However, it is possible to transform the two mostimportant crop legumes, soybean and alfalfa, andregenerate transformed fertile plants. Thus, weare now ready to attempt to improve leguminouscrops using transformation. A review of legumetransformation has been presented by Nisbet &Webb (1990).

In the following sections, we summarize manyof the transformation studies that have been car-ried out with leguminous crops species and drawconclusions about the shortcomings and prospectsfor each. The species are divided into sections onGrain and Forage, based upon primary crop use.

Grain legumesSoybean (Glycine max)Soybean is a very important oil- and protein-producing crop, so many studies have been car-ried out with the goal of obtaining transformedplants.

Transformation by Agrobacterium tumefaciensWhile most dicots, including legumes as describedin other sections of this review, are very suscepti-ble to A. tumefaciens, soybean plants do not formtumors very well when inoculated with virulentstrains and there are large differences in responsesfound with different genotypes (e.g. see Owens &Cress, 1985; Byrne etal, 1987). Similar genotypeeffects were also found in the formation ofkanamycin-resistant callus on aseptic hypocotylsegments following inoculation with A. tumefa-ciens carrying the nptll gene (Hinchee et ai,1988). These reports identified Peking as one ofthe most responsive genotypes, and this black-seeded noncommercial line has been used inmany subsequent studies.

In early soybean transformation work, cotyle-dons, nodes and internodes of 2-3-week-oldaseptic Forrest soybean seedlings were injectedwith A. tumefaciens strain A722 by Facciotti et al.(1985). This strain harbored a cointegrate,virulent plasmid containing the nptll gene drivenby a small subunit promoter from soybean ribu-lose-bisphosphate carboxylase, with an octopinesynthase terminator. Hormone-independent,octopine-positive and kanamycin-resistant callusand suspension cultures could be recovered fromtissues near the wound sites after growth on MSbasal medium with 500 mg carbenicillin/1. Whenthe transformed callus was grown in the light thenptll poly(A+)RNA was 5-10 times higher thanwhen in the dark. Light also increased the NPT IIenzyme activity and the kanamycin resistance ofthe callus. Thus, the light-inducible promoteroperates properly in transformed callus even whendriving an nptll gene.

101

102 WIDHOLM

A shoot regeneration system from detachedcotyledons of 4-10-day-old, aseptically germi-nated soybean seedlings was used to obtain trans-genic plants with the cultivars Peking and MaplePresto, which were identified as being the mostresponsive to A. tumefaciens out of 100 genotypes(Hinchee et al, 1988). The cotyledons weredipped in A. tumefaciens strain A208 carrying nos>nptll, nos and cauliflower mosaic virus (CaMV)35S, uidA, nos (pMON 9749) and were placedadaxial side down on B5 medium (Gamborg,Miller & Ojima, 1968) salts with 1.15 mg6-benzyladenine (BA)/1, 500 mg carbenicillin/1and 100 mg cefotaxime/1 to inhibit bacterialgrowth, and 200-300 mg kanamycin/1 to select fortransformed cells. After 3 weeks some of the tis-sues expressed (3-glucuronidase (GUS) enzymeactivity in tissue sections when measured histo-chemically with X-Gluc as substrate.

This procedure was used with 1400 cotyledons,mostly from Peking plants, using cointegrate plas-mids pMON9749 or pMON894, which carriesnptll and a modified petunia EPSPS (enolpyruvyl-shikimate-3-phosphate synthase) gene which con-fers glyphosate resistance, driven by an enhancedCaMV 35S promoter. A total of 128 plants devel-oped via adventitious shoot formation in the pres-ence of kanamycin. Eight of these plants wereactually transformed when NPT II or GUS activ-ity was measured in leaves or resistance tokanamycin or glyphosate was measured in leaf cal-lus. No transgenic plants were recovered withoutkanamycin selection when 100 shoots were ana-lyzed.

Two transgenic plants produced seed thatshowed a 3:1 segregation of the transformingDNA and the encoded traits (NPT II and GUSenzyme activity in one case and kanamycin andglyphosate resistance in the other). The geneticand molecular biological data are consistent with asingle site of integration of one or a few genecopies.

Zhou & Atherly (1990) used a modification ofthe Hinchee et al (1988) method with cotyledonsfrom 5-day-old Peking seedlings. The cotyledonswere removed, air-dried for 30 min and inocu-lated by wounding the proximal end with a scalpeldipped in A. tumefaciens strain A281 with the viru-lent plasmid pTiBo542 and a binary with nptlland uidA. The cotyledons were incubated on amedium similar to that used before for 48 h in thedark and for 24 h in the light. The cotyledons

were then transferred to the same medium with500 mg carbenicillin/1, 100-200 mg cefotaxime/1and 250 mg kanamycin/1 where callus developedon up to 94% of the cotyledons and shoot primor-dia on 8% within 2-3 weeks. Fourteen shoots thatformed were maintained on the same mediumwith the antibiotics but without BA. Six plantswere recovered and three out of five plants thatcontained the maize Ac active transposable ele-ment inserted inside the uidA gene had GUS-pos-itive sectors when assayed with X-Gluc. Theseresults show that this cotyledon system can pro-duce transformed plants and that the Ac elementis active in soybean. This activity was studied fur-ther with transformed callus where kanamycin-selected callus with the uninterrupted uidA geneshowed GUS activity in 100% of the cases, while45% of the calli with Ac in the gene showed GUSactivity. There was no GUS activity detected inselected callus containing the uidA gene with theinactive Ac element Ds inserted into it.

Transgenic soybean plants were also producedby injecting about 30 |xl of A. tumefaciensC58Z707 carrying a binary plasmid with nptll intothe plumule, cotyledonary node and adjacentregions of 18-24 h aseptically germinated A0949seedlings with one of the cotyledons removed(Chee, Fober & Slightom, 1989). The seedlingswere incubated for 4 h on moistened paper towelsand were then transplanted to soil for normalplant growth. A total of about 2200 plants wererecovered from about 4000 inoculated seedlingsand leaves of these were analyzed for NPT IIactivity by the in situ gel assay. Sixteen plants con-tained NPT II activity and this was not due tocontaminating bacteria since no A, tumefacienswere recovered when the leaf extracts were platedon a selective medium. Ten of the 16 NPT II pos-itive plants clearly contained the nptll gene andthis was confirmed by the polymerase chain reac-tion (PCR). However, progeny from only one ofthese plants carried nptll sequences and then onlythree out of 36 were positive. This low inheritanceratio may be caused by chimerism of the R0plant.

This method of injecting germinating seedlingsdid produce some transformed progeny but themethod is labor intensive and the frequency verylow (0.07% of the originally injected seedlingsproduced transgenic plants). Also the methodmay not show genotype specificity.

Another transformation system utilized imma-

Leguminous plants 103

ture cotyledons that were wounded by beingpushed into 500 |xm stainless steel or nylon meshwith a spatula and then were incubated overnightwith A. tumefaciens strains LBA4404 or EHA101with a binary plasmid containing nptll and themaize zein gene (Parrott et al> 1989). The tissuewas then incubated in an embryogenesis induc-tion medium with high naphthaleneacetic acid(NAA) (10 mg/1) and 500 mg cefotaxime/1 and insome cases with 10 mg G418/1 to select trans-formed cells. Embryos were removed at 30 daysand then placed on a maturation and germinationmedium with the cefoxitin. There were great dif-ferences in the numbers of plants produced by thedifferent genotypes used and LBA4404 seemed todecrease the regeneration frequency. From a totalof 9800 cotyledons, 14 plants were recovered,three of which were established to have beentransformed by Southern hybridization. Two ofthese three plants came from embryos formedwithout G418 selection. These plants producedseed but none of the progeny were transformed,indicating that the R0 plants were apparentlychimeric and none of the seed was produced fromtransformed cells.

Transformation by Agrobacterium rhizogenesHairy roots have been produced by inoculatingwounds on 2-3-week-old greenhouse grownplants made with a multiple needle devicebetween nodes 2 and 3 with A. rhizogenes strainR1000 (pRiA46) (Owens & Cress, 1985). When23 soybean cultivars and three Glycine soja plantintroductions were inoculated, hairy roots formedon five of the soybean and two of the G. soja geno-types. In this case, Peking did respond but Biloxiproduced a larger number of roots. Inoculation ofaseptic cotyledons from 1- or 3-day-old seedlingsin cuts on the adaxial side also resulted in hairyroot formation with Peking and Biloxi.

Savka et al (1990) tested four A. rhizogenesstrains with aseptic 6-15-day-old seedlings of 10soybean cultivars by inoculating with a scalpeldipped in bacteria by cutting the cotyledon abaxialsurface several times and by making 2 cm longlongitudinal cuts on the hypocotyls. The seedlingswere incubated in large culture tubes where rootsformed at both inoculation sites. The A. rhizogenesstrain K599 (cucumopine) produced roots oncotyledons of all cultivars, whereas the other threestrains (8196, mannopine; 1855, agropine; A4,agropine) were much less effective. The K599-

induced roots were all opine positive, but onlyabout a third of those produced by the otherstrains were. None of the roots produced onhypocotyls contained opines, so these were notconsidered to be transformed.

The hairy roots grew rapidly in liquid or onsolidified media and were used in soybean cystnematode (Heterodera glycines) propagation stud-ies. The nematodes formed mature cysts on thecultured roots, followed by formation of second-generation cysts, which indicates that an entire lifecycle can be completed with this system.

So far there has been no indication that the soy-bean hairy root cultures can produce shoots bydirect shoot formation or by embryogenesis ororganogenesis following callus induction.

Hairy root cultures have also been producedwith the wild perennial Glycine species canescens,clandestina and argyrea by inoculating hypocotylsof 6-12-day-old aseptic seedlings with their rootsremoved by puncturing with a needle on a syringecontaining A. rhizogenes (Rech et at., 1988, 1989;Kumar, Jones & Davey, 1991). Transgenic plantswere produced when the hairy roots from G.canescens and G. argyrea were cultured on amedium with 1.1-10 mg BA/1 and 0.05 mgindolebutyric acid (IBA)/1. The roots formed cal-lus and then shoots formed on the callus.

Transformation by microprojectile bombardmentIn the first work with a particle acceleration devicewith soybean, Christou, McCabe & Swain (1988)bombarded immature embryos with 1-5 (xm goldspheres coated with plasmid DNA containing thenptll gene with the CaMV 35S promoter and nospolyadenylation sequences. The accelerationdevice utilizes a force generated by the volatiliza-tion of a water droplet by an electrical discharge toaccelerate an aluminum foil sheet that hits a stop-ping screen, thus allowing the gold particles tocontinue on into the target tissue. The system isoperated under a partial vacuum. Protoplasts pre-pared from the bombarded tissue were culturedand 3 weeks after the bombardment were selectedwith 50 or 100 mg kanamycin/1. Selected colonies,which were recovered at a frequency of about 10~5,were grown on 100 mg kanamycin/1. These cellclones all contained NPTII activity and nptllgene sequences. However, plants could not beregenerated from these transformed cultures.

Meristems from immature soybean seed embry-onic axes have also been bombarded with gold

104 WIDHOLM

particles coated with DNA containing theCaMV 35S, nptll, nos or CaMV 35S, uidA, nosgene constructs (McCabe et al, 1988). The tis-sues were then cultured on a high cytokinin(13.3 jxM BA) medium to induce multiple shootformation in the dark for 1-2 weeks. Shootsformed on a medium with 1.7 JJLM BA and thesecould be rooted on hormone-free medium or begrafted onto soybean seedlings to obtain wholeplants. Three to eight shoots could be recoveredfrom each axis. Histochemical assays for GUSactivity showed that expression could be observedin sectors of the tissues 2 days after bombardmentand NPT II activity could also be measured in tis-sue homogenates. Overall about 2% of the shootscontained some transformed tissue. In one case,one plant resulting from 389 grafted shootsshowed NPT II enzyme activity and three out often progeny were also positive. One of these prog-eny plants contained transforming DNA.

Two of the transformed plants, obtained fol-lowing bombardment of meristems with DNA-coated gold particles with both uidA and nptll(plant 4615) or with uidA (plant 3993) genes asdescribed above, were analyzed for gene expres-sion and inheritance (Christou et ah, 1989). Thesetwo plants expressed the genes in all leaves sowere not sectorial chimeras. Plant 4615 producedtransformed progeny in the ratio 2.3:1 and the Rltransformed plants produced progeny that eitherwere all transformed or in a 3:1 ratio, which fits ahomozygous or heterozygous Rl state, respec-tively. Plant 3993 produced transformed Rl pro-geny at a 1:1 ratio and this was also 1:1 in the R2indicating some problem with transmission of thetransformed trait through the pollen. This wasconfirmed by staining pollen for GUS activity. Inall cases the enzyme activities and gene sequenceswere inherited together. In plant 4615 progeny,the nptll and uidA genes were inherited as a block,indicating integration of both genes at a single site.

Since the soybean, particle bombardment,transformation system does not employ a resis-tance marker for selection, the transformed plantsare identified by histochemical assay of the expres-sion of the uidA gene (Christou, 1990; Christou &McCabe, 1992). The first step is to find expres-sion in 2-4 mm thick sections taken from thebasal end of elongating shoots. GUS-positiveshoots are grown further, and primary leaves andpetioles and midribs of the second and third trifo-liate leaves are assayed. A wide variety of GUS

expression patterns are found, which clearly sup-ports the chimeric nature of the transformationevents. Plants with patterns expected for transfor-mation of the L2 or germ layer were found to pro-duce transformed progeny. In about 450experiments where about 180 000 shoots wereregenerated, 21 600 (12%) showed GUS expres-sion but apparently only about 2% of those thatexpressed GUS were germline transformants andproduced transformed progeny.

Transgenic Rl plants obtained by particle bom-bardment with the CaMV 35S, uidA, nos genewere assayed for GUS activity histochemicallyusing X-Gluc (Yang & Christou, 1990). Theactivity staining of thin sections showed that cer-tain tissues expressed GUS activity at high levels(root pericycle, xylem parenchyma and stem andleaf phloem), others at intermediate levels (leafmesophyll, certain ground tissues of stem and leafmidrib and trichome and guard cells) and some atlow or zero levels (root procambium, phloem andcortex, stem vascular cambium and the majorityof leaf midrib cortex cells).

An embryogenic suspension culture of the culti-var Fayette was initiated by Finer & Nagasawa(1988); it grew relatively slowly but consisted ofsmall clumps of globular stage embryos. Thesecultures continuously increased in mass by adven-titious embryo formation from cells at or near thesurface of the preexisting embryos. The cultureswere initiated and maintained on a high 2,4-dichlorophenoxyacetic acid (2,4-D) medium bymanually selecting the embryogenic-appearingclumps at each transfer. Embryo formation can beinduced within a month on a regenerationmedium without hormones and these embryoscan be matured within another month and, fol-lowing desiccation, can germinate and form com-plete plants.

These embryogenic suspension cultures wereused in particle bombardment studies with theBiolistics Particle Delivery System from DuPontwith 1.1 jim tungsten particles coated with DNAcarrying hph and/or uidA genes driven by theCaMV 35S promoter (Finer & McMullen, 1991).Selection for hygromycin (50 |xg/ml) resistancewas imposed 1-2 weeks after bombardment andwithin 4-6 weeks live growing regions were visibleon the cell clumps, which were selectively trans-ferred to fresh medium for another 2-3 months atwhich point plant regeneration procedures wereinitiated. When GUS activity was measured histo-


chemically, 3 days after bombardment, there werean average of 709 positive foci, and each bom-bardment produced about three stably trans-formed clones. These results give a stabletransformation versus transient expression fre-quency of about 0.4%. An average of 25 embryoswere produced per clump and 20% of these ger-minated to form plants from a total of 12 differenttransgenic clones. The regenerated plantsexpressed GUS activity and contained transform-ing DNA. Some transformed progeny were alsoproduced.

Transformation of protoplasts by direct uptake ofDNAThe first report of the regeneration of soybeanplants from protoplasts was that of Wei & Xu(1988), who obtained plants from four out of sixChinese cultivars used. The protoplasts were iso-lated from immature cotyledons and nodular cal-lus formed on a medium with 2,4-D and BA andshoots subsequently formed on a medium withNAA, BA, kinetin, zeatin and casein hydrolysate.Up to 35% of the calli formed shoots with one cul-tivar (Heilong 26) and the regenerated plants werefertile. Plant regeneration from soybean immatureembryo protoplasts has also been reported byLuo, Zhao & Tian (1990).

We have been able to repeat this work usingseveral more cultivars (7 out of 15 cultivars pro-duced plants) including some current commer-cially important ones (Dhir, Dhir & Widholm1991c, 19926). So far the regenerated plants havebeen fertile and usually produce more than 20seeds. To date little somaclonal variation has beennoted in hundreds of R2 Clark 63 progeny.

Transient expression studies were carried outwith Clark 63 protoplasts using the cat and uidAgenes to optimize the DNA uptake methods (Dhiret al.y 19916). The optimal conditions included asingle electroporation pulse (500 V/cm fieldstrength and 1000 |JLF capacitance) with about 4%polyethylene glycol (PEG) 6000. To obtain stabletransformation, Clark 63 protoplasts were electro-porated with two different plasmids, one with hphand uidA genes controlled by CaMV 35S promot-ers and nos terminators and another with nptllwith a CaMV 35S promoter and nos terminatorand the mannityl opine biosynthesis region (Dhiret aly 19916, 1992a), using similar electroporationmethods, selection for the antibiotic resistancemarkers carried by the plasmids and regeneration

of shoots and plants. However, further study ofthe progeny does not confirm the presence of thetransformed genes or opine synthesis so one can-not say that transformed progeny can be obtainedusing this system at present. Because of theseinconsistencies these papers have been with-drawn. Other studies in this laboratory haveshown that transformed calli can be clearly pro-duced but the regeneration of plants is difficult.

Transformation by microinjectionAn ovary microinjection method was used by Liuet al (1990) to transfer the Solanum nigrum psbAatrazine resistance gene into soybean. Twomicroliters of 100 |xg plasmid DNA/ml, carryingthe psbA gene, was injected into 507 ovaries 1 dayafter pollination and 1220 seeds were harvestedfrom these. Seed set was only slightly decreased incomparison with uninjected controls. Plantsgrown from the treated seeds in the field weresprayed with the herbicide atrazine and seven ofthe 866 plants that grew from the seed showedresistance. Fluorescence induction curves con-firmed the atrazine tolerance of some plants. Dot-blot hybridization with isolated chloroplast DNAfrom two of the resistant plants showed that therewere pBR322 plasmid sequences present. Theplasmid sequences, rather than the psbA gene,were used for probing, since the resistance iscaused by a single base change and the genesequences are highly conserved between species.The authors feel that the psbA gene was in thechloroplast because the pBR322 sequences werefound in purified chloroplast DNA and the psbAgene used had only the native promoter, whichshould be expressed only in the chloroplast.

This microinjection method looks promisingbut it must require careful and precise injectiontechniques to be successful so may not be possiblein many laboratories. Likewise additional geneticand molecular analyses are needed to clearly sub-stantiate the claims.

Conclusions about soybean transformationSoybean has been transformed by A. tumefaciensand microprojectile bombardment to obtaintransformed progeny. None of the methods usedis ideal at present so the one chosen would dependupon the laboratory skills, funding levels and thegenotypes desired. The A. tumefaciens method islow frequency, is very genotype specific andproduces chimeric plants. The microprojectile

106 WIDHOLM

method is also low frequency, produces chimericplants, requires much labor and costly equipment(including royalty payments), but apparently isnot genotype specific. This method, when appliedto embryogenic suspensions, needs first of all thesuspension cultures, which require time and skillfor initiation, but the selection is straightforwardand nonchimeral plants can be obtained in about8 months.

Pea (JHsum sativum)

There have been a number of studies carried outwith P. sativum (pea) because this is an importantfood crop. In one study, seven wild-type A. tume-faciens strains were used to induce tumors ongreenhouse-grown plants, shoot cultures andaseptically germinated seedlings of five cultivars(Puonti-Kaerlas, Stabel & Eriksson, 1989). Threeof the strains were very effective on all cultivars,with A281 being most effective, followed by C58and GV3101 (pGV3304). These results show thatthe bacterial strain was more important than peagenotype as far as tumor induction is concerned.

Kanamycin-resistant (75 |xg/ml) callus wasobtained in cocultivation experiments with epi-cotyls, stem and leaf pieces, with a frequency usu-ally near 30%. Cocultivation with microcoloniesformed from protoplasts also produced kanamycin-resistant colonies with two cultivars (frequenciesof 5 X 10~3 for Filby and 10"4 for Bello). Nopalinesynthase activity was expressed in 91% of thekanamycin-resistant calli and all 12 lines analyzedcontained nptll sequences.

Pieces of pea shoot cultures and epicotyls fromaxenic pea seedlings were cocultivated for 2-3days with A. tumefaciens carrying hph or nptllgenes, in liquid medium containing 0.5 mg BA/1and 2,4-D (Puonti-Kaerlas, Eriksson &Engstrom, 1990). The explants were washed andwere transferred to callus initiation medium with500 |xg claforan/ml to inhibit bacterial growth and15 |xg hygromycin B/ml or 75 jig kanamycin/ml toselect transformed cells. The best medium fortransformed callus formation contained 0.5 mgBA/1 and 2,4-D, where from 135-392 hygromycin-resistant calli formed on 200 explants from threecultivars within 1 month. Other trials with all fivecultivars with both kanamycin or hygromycin pro-duced 30-240 resistant calli per 200 explants.Shoots could be regenerated from hygromycin-selected callus of one cultivar, whereas none could

be obtained from any of the kanamycin-selectedcalli. The transformed shoots contained hphsequences. The regenerated plants had producedflowers but no seed at the time of publication.

Recently the pea cvs. Puget and Stivo have beentransformed using A. tumefaciens strain GV3101(pGV2260::pGV1503) carrying hph using themethods of Puonti-Kaerlas et al (1990) asdescribed above (Puonti-Kaerlas, Eriksson &Engstrom, 1992a). Shoot induction occurred with0.2%-15% of the hygromycin selected calli in 4-9months and about half of these could be rooted.Plants regenerated from different Stivo-transformed calli were fertile and the Rl and R2progeny inherited the transformed gene as aMendelian, dominant trait. Progeny were ana-lyzed by Southern and Northern blots and a leafspot test with concentrated hygromycin solutions(1%) where lesions formed on untransformedleaves. Hygromycin resistance was confirmed bycallus formation from leaf pieces on a mediumwith 15 mg hygromycm/ml. The transformedplants and their progeny were tetraploids.Whether the production of tetraploids is associ-ated with the transformation or the tissue cultureprocess is not known. However, ploidy increasesare common in plants regenerated from tissue cul-tures.

Tumor induction frequencies induced by fivedifferent A. tumefaciens strains were measuredwith epicotyl segments from 7-day-old axenic, eti-olated, Madria pea seedlings (Kathen & Jacobsen,1990). The segments were soaked for 1 h with thebacterial suspensions, were blotted dry and thenwere incubated on MS medium with 0.1 mg BA/1and 0.1 mg picloram/1. Tumor induction frequen-cies of from 26% to 36% were obtained with threestrains, C58C1 being most effective. No tumorscould be induced on immature zygotic embryos orapical domes, both of which are known to producesomatic embryos (Kysely et al, 1987). Low fre-quencies were found with immature leaflet (5%)and leaf discs (11%), while epicotyl segments,nodal explants and whole apices gave frequenciesfrom 47% to 68%. Cocultivation with 50 |xM ace-tosyringone was inhibitory and increasing thecocultivation time to 5 days stimulated tumor for-mation. Transformation frequency was alsoaffected by the pea genotype.

Following transformation of nodus explantswith A. tumefaciens carrying either hph or nptllselectable marker genes, selection either during


shoot formation or after shoot formation at therooting stage produced from 2.5% to 4.9% and0.9% to 1.1% transformed shoots from 1075nodus explants. Analysis of eight kanamycin-resistant plants detected NPT II activity in five.Twelve of these plants flowered but no seed wasset.

It is of interest to note that the frequency ofrecovery of transformed shoots was similar for thenptll and hph containing strains even though thelatter was seven-fold more efficient with hypocotylsegments at producing transformed callus. Theauthors feel that, owing to the greater toxicity ofhygromycin, the selection of resistant shoots wastoo stringent with this antibiotic in comparisonwith kanamycin.

A plant regeneration system from nodal thincell layer segments obtained from 10-12 day-oldaxenic pea seedlings was developed by Nauerby etal (1991). After 14 days in liquid medium, up toeight 1-2 cm long shoots could be removed asthey formed over a 2-3 month period. The origi-nal explant contained one preformed bud. Thefour pea cultivars tested produced an average of3.3-4.1 shoots per explant with a shoot-formingfrequency of between 88% and 97%. The rootingfrequency varied from 14% to 49% and the pod-forming frequencies of the regenerated plants var-ied from 12% to 57%. This method can producerooted plants within 7 weeks, but the number ofshoots, the rooting frequency and fertility are allrelatively low.

Preliminary transformation experiments withthe nodal, thin cell layer segments have shownGUS expression in shoots regenerated from eightout of 148 explants following a 3-day cocultiva-tion with A. tumefaciens carrying the uidA gene ona binary vector. Following cocultivation theexplants were incubated in a medium containing750 |xg vancomycin/ml to stop bacterial growthbut apparently not with any antibiotic or otheragent to select for transformants.

There have been several studies, in addition tothose described above, that have optimized A.tumefaciens transformation of pea tissues but havenot yet produced transformed plants. Hobbs,Jackson & Mahon (1989) inoculated 17-day-oldplants grown in growth chambers of 13 differentgenotypes with wild-type A. tumefaciens strainsC58, A281 and ACH5. The A281 strain inducedmore and larger tumors. Similar results werefound with immature leaves that were cocultivated

for 2 days and then incubated on medium with nogrowth regulators but with 500 |xg carbenicillin/ml.In all cases there were positive responses but therewere differences between pea genotypes. Thetumor tissue grew without growth regulators, con-tained the specific opines for each strain and hadboth right (TR) and left (TL) transferred DNA(T-DNA) sequences.

This same laboratory has also transformed5-day-old seedling stem segments with several dis-armed A. tumefaciens strains (Lulsdorf et al.,1991). Following cocultivation, transformed cal-lus was induced from 2-3 mm end-pieces onmedium with 1 mg BA/1 and 1 mg 2,4-D/l and500 mg carbenicillin/1 and 100 mg cefotaxime/1 toinhibit bacterial growth, and 50 mg kanamycin/1or 25 mg hygromycin/1 to select the transformedcallus. Optimal conditions included a 4-daycocultivation period with a tobacco, nurse, sus-pension culture and the A. tumefaciens strainEHA101 (pBI1042), where about 75% of the sec-tions formed transformed callus using kanamycin(35S, 35S, alfalfa mosaic virus enhancer, nptll) orhygromycin (nos, hph) selection. The cointegratevector was more efficient than the binary vectorsystem used.

Hussey, Johnson & Warren (1989) studied theinduction of tumors or hairy roots when varioustissues of cultured pea shoots were inoculatedwith wild-type strains. Hairy roots could beinduced by the A. rhizogenes strain 9402 andtumors by A. tumefaciens strain C58 but not byACH5. Inoculation of 0.5 mm slices of the shoot,beginning at the tip by a 48 h cocultivation andthen incubation on a medium with 0.5 mg BA/1,0.25 mg IBA/1 and 500 mg ampicillin/1, producedtumors and hairy roots first on the youngest tissueand then progressively down the stem. Tumorsand hairy roots could also be induced by stabbingthe apical dome, following dissection of the leavesand primordia, with a needle dipped in A. tumefa-ciens or A. rhizogenes^ respectively. These studiesindicate that meristematic cells can be trans-formed by the Agrobacterium systems.

Hairy roots had also been induced on axenic,3-day-old, pea epicotyls from five cultivars eitherfollowing decapitation or on wounded regionsalong the epicotyl (Bercetche et al., 1987). Theinoculated plants were incubated in medium intubes in the dark. Transformed roots were obtainedwith A. rhizogenes strain 1855 but not with strains2659 and 8196. The cortical cells from which the

108 WIDHOLM

hairy roots were ultimately derived were partlypolyploid but the resulting roots were diploid.

Pea protoplasts should also provide a system fortransformation by direct DNA uptake if plantregeneration can be obtained. Puonti-Kaerlas &Eriksson (1988) were able to obtain callus forma-tion from protoplasts isolated from epicotyls of12-day-old aseptic seedlings of 10-pea cultivars.Plating in agarose slabs 2 days after preparationwas beneficial for increasing the plating efficien-cies. High plating efficiencies were obtained withprotoplasts from leaves, shoot cultures and epi-cotyls of the best cultivar, Filby. Shoot-like struc-tures developed in about 1 % of the calli from twoof the genotypes but normal plants with roots didnot form.

Protoplasts prepared from shoot apices ofyoung, in vitro-grown Filby and Belman peaseedlings were electroporated with plasmids con-taining the nptll or hph genes (Puonti-Kaerlas,Ottoson & Eriksson, 19926). After two days theprotoplasts were embedded in 0.6% (w/v) SeaPlaque agarose and selection began immediatelyor after 12 days with 75 mg kanamycin/1 or 15 mghygromycin/1. Resistant transformed colonieswere selected with hygromycin and none wasrecovered with kanamycin. No plants could beregenerated from the transformed colonies.

Protoplasts were also prepared from shootapices or lateral shoot buds from aseptically grownembryonic axes (Lehminger-Mertens & Jacobsen,1989). When the protoplast derived callus wasgrown on media with 5-10 |xM picloram or 2,4-D, somatic embryos formed on 10% -30% of thecalli from two cultivars, whereas 1 % or less of thecalli from three other cultivars formed embryos.Embryo maturation was stimulated by 1.5 |xMgibberellic acid (GA3) and 0.12 JJLM NAA andgermination occurred with 2.9 |ULM GA3.Complete plants formed and could be potted insoil. These plants had normal morphology andproduced seed.

Leaf mesophyll protoplasts were used in tran-sient expression experiments following electropo-ration with DNA carrying the uidA gene (Hobbs etaU 1990). The nos and CaMV 35S promotersgave similar expression levels but the duplicated35S promoter had higher activity. One of thegenotypes gave 5-20-fold higher levels of GUSactivity than did the other genotype.

These studies indicate that it should be possibleto transform protoplasts from certain genotypes

using direct DNA uptake methods and followingthe selection of the transformed colonies plantregeneration can be accomplished. However, thesuccessful use of such a system has not beenreported. Likewise the A. tumefaciens transforma-tion system with pea explants that can regenerateplants seems to be relatively efficient and the pas-sage of the transformed genes to progeny has beendemonstrated by Puonti-Kaerlas et al. (1992a).

Moth bean (Vigna aconitifolia)

Protoplasts isolated from primary leaves of 9-10-day-old axenic moth bean seedlings were incu-bated with plasmids carrying the nptll gene drivenby a pair of tandemly arranged nos promoters orby the CaMV 35S promoter (Kohler et al.,19876). DNA uptake was induced by PEG and insome cases by electroporation. Selection wasapplied after 7 days, when the protoplasts wereembedded in agarose slabs and 75 |xg kanamycin/mlwas added. Transformation frequencies of about10~5 were obtained with cultivar 560, but the fre-quency was about 10-fold lower with 909.Electroporation did not increase the frequencywhen applied in addition to PEG.

Electroporation or PEG were used with the pro-toplast system described above, with the same twocultivars to stimulate the uptake of a plasmid car-rying the nptll gene alone and one with nptll andcat (Kohler et al, 1987a). The transformation ratewas higher when kanamycin selection began 7days after DNA uptake rather than after 28 days.Again the transformation frequency was muchhigher with cultivar 560 than with 909. The cul-ture medium was also very important in deter-mining transformation frequencies. Transientexpression of the cat gene could be observed 48 hafter electroporation. No stable transformation ortransient expression was seen with suspensioncultured cells without cell wall removal. Southernhybridization analysis showed that nptll DNAsequences were present in the selected calluslines. When the selected callus lines were placedon regeneration medium without growthregulators, about 10% of the calli regeneratedplants.

Moth bean protoplasts isolated and cultured asdescribed above were cocultivated 3 days afterpreparation for 48 h with A. tumefaciens contain-ing a cointegrate plasmid containing nptll drivenby the nos promoter (Eapen et al, 1987). After


being washed, the cells were cultured in a mediumwith 500 fxg cefotaxime/ml to inhibit bacterialgrowth. After 21 days the cells were plated inagarose and 75 |xg kanamycin/ml added to selecttransformants. Colonies growing after 3-4 weekswere grown with 100 (xg kanamycin/ml for anadditional 2-4 weeks. Again the transformationfrequency of the cultivar 560 (4 X 10~5) was muchhigher than with 909 (5 X 10~7). The selected callicontained nptll DNA sequences and 23% of themcontained nopaline synthase activity. Shoot budsand plants were then regenerated.

The same moth bean protoplast system wasused in another study of PEG-mediated trans-formation with a plasmid carrying the nptll genedriven by the CaMV 35S promoter (Kohler et ah,1989). Again the cultivar 560 was transformed ata much higher rate than another cultivar, inthis case 88. The transformation frequency wasincreased by 1-2-fold when the protoplasts wereX-irradiated (10 Gy) 1 h after incubation withDNA.

These studies with moth bean protoplasts showthat transformation by direct DNA uptake can berelatively efficient and shoots can be regeneratedwithin about 3 months from about 10% of theselected calli (Kohler et ah, 1987a; Eapen et ah,1987). Even though transformed plants wereregenerated, no information concerning the char-acteristics of the plants including fertility and pas-sage of the transforming DNA to progeny havebeen reported.

Broad bean (Viciafaba)

Hairy roots were induced on axenic 5-6-day-oldV. faba seedling epicotyls by wounding with asyringe filled with A. rhizogenes strains A4and 15834 containing the binary plasmidpGSGLUCl, which carries nptll and uidA genesdriven by the 1' and 2' promoters, respectively, ofthe TR-DNA of A. tumefaciens (Schiemann &Eisenreich, 1989). Hairy roots formed within14-21 days and 45% expressed GUS activity.These roots could be cultured on hormone-freesolid, but not on liquid medium. No antibioticswere needed to suppress bacterial growth.

Hairy roots were also induced on axenic V. fabacotyledons and stems of 10-day-old seedlingsusing A. rhizogenes strain LBA9402 with the Riplasmid and the binary vector pBIN19, which car-ries the nptll gene (Ramsey & Kumar, 1990). The

stems were inoculated by placing bacteria in a1 cm long cut, half-way through the stem abovethe cotyledonary node. The excised cotyledonswere cut with a scalpel to make five slits on theadaxial surface and bacteria were spread over thesurface. After 2 days of cocultivation the tissueswere transferred to a medium with 250 |jLg cefo-taxime/ml to inhibit bacterial growth. Rootsformed after 14 days on all eight genotypes butthere were large differences in the frequencies.Sixteen of 25 roots analyzed contained NPT IIactivity. Chromosome counts showed the roots tobe 51% diploid, 45% tetraploid, with the rest hav-ing a higher ploidy level. Four of 65 root clonesexamined had chromosome rearrangements.Stems produced a higher proportion of diploidsthan did cotyledons. Apparently plants could notbe regenerated from the transformed roots.

These two studies show that transformed hairyroot cultures can be easily initiated from manyfaba bean genotypes but with no apparent plantregeneration.

Vicia narbonensis

A system for regenerating plants has been devel-oped for V. narbonensis, a close relative of V. faba,via embryogenesis from callus obtained fromexcised young shoot tips (Pickardt, HuancarunaPerales & Schieder, 1989). This regeneration pro-tocol was coupled with A. tumefaciens trans-formation to produce transformed plants byPickardt et al (1991). Shoot tips and epicotylsfrom 2-3-day-old aseptic seedlings were precul-tured for 1 day in liquid MS medium containing4 mg picloram/1 with shaking. A small amount ofA. tumefaciens strain C58C1 containing the co-integrate vector pGV3850HPT (CaMV 35S, hphand nos, nos) was added to the liquid medium andcocultivation was carried out for 48 h with shak-ing. The segments were rinsed and incubated for48 h on the same medium, solidified with 0.25%(w/v) Gelrite, with 500 |xg claforan/ml to inhibitbacterial growth and then were selected on thismedium with 30 |xg hygromycin/ml. After 28 daysthe cultures were transferred to the same mediumexcept that 60 |xg hygromycin/ml and 200 |xgclaforan/ml were used. After 3 months on thismedium with transfers every 28 days, the cultureswere placed on regeneration medium with low-ered auxin content and transferred every 28 daysuntil somatic embryos formed. These were

110 WIDHOLM

matured and germinated to form complete plants.Hygromycin concentrations sufficient to preventgrowth of controls were present in each mediumused.

A total of 438 of the original 2500 treatedexplants had callus still growing on the regenera-tion medium and 41 of the 44 tested containednopaline. From a total of 142 embryos formedfrom these calli, 26 showed hygromycin resistanceand 7 of 12 shoots formed from these were nopa-line positive. Southern hybridization of five nopa-line positive shoots showed the presence ofT-DNA. The rooted shoots failed to develop fur-ther.

Recently, Pickardt et al (1992) have reportedthe regeneration of two fertile V. narbonensisplants transformed with the Brazil nut 2S albumingene using methods similar to those describedabove. Some of the progeny expressed this highsulfur amino acid protein in leaves but the levels inseed were very low, perhaps because theCAMV 35S promoter was used. This workdemonstrates that V. narbonensis can be trans-formed to produce transformed progeny.

Lentil (Lens culinaris)

Two-week-old greenhouse-grown lentil seedlingswere inoculated with four wild-type A. tumefaciensstrains (C58, ACH5, GV3111, A281) by stabbingwith a needle at the first and second internodes(Warkentin & McHughen, 1991). Shoot apicesexcised from 4-day-old seedlings were also inocu-lated by immersing in a bacterial suspension. Theexplants were blotted on filter paper and werecocultivated for 3 days on MS basal medium fol-lowed by rinsing with 500 (xg carbenicillin/ml andincubation on the same medium with 100 |xgcefotaxime/ml. All four A. tumefaciens strainsformed tumors at all of the wound sites on thelentil stems. The tumors were larger wheninduced by strains GV3111 and ACH5. From81% to 100% of the shoot apices developedtumors, the largest being formed by strain A281.The tumors produced the appropriate opine oropines and the one line analyzed contained T-DNA sequences. These results indicate that lentiltransformation can be carried out efficiently. Ifthis can be coupled with the apex regenerationsystem described by Williams & McHughen(1986) then transformed plants should be obtain-able.

Cowpea (Vigna unguiculata)

Tumors could be induced on stems of 7-day-oldcowpea seedlings when virulent A. tumefaciensstrains LBA1010 (with the octopine type plasmidpTiB6) and LBA958 (with the nopaline type plas-mid pTiC58) were inoculated by puncturing thestem with a toothpick soaked in the bacteria(Garcia, Hille & Goldbach, 1986).

Leaf discs from 6-day-old cowpea plants werealso submerged for 30 s with diluted A. tumefa-ciens C58C1 carrying an nptll (wos, nptll, ocs) geneon the nononcogenic Ti plasmid pGV3850:: 1103neo (dim) (Garcia et al, 1986). The discs werewashed briefly, blotted dry and incubated oncallus-inducing medium with a petunia cell feederlayer and a filter paper for separation. After 2 daysthe discs were transferred to medium with 200 (xgcefotaxime/ml and 200 |xg vancomycin/ml toinhibit bacterial growth and with 50 |xg G418/mlto select transformed cells. After six 20-day trans-fers, on a medium with the cefotaxime, van-comycin and 100 |xg kanamycin/ml, the selectedcalli were grown on a medium without any anti-biotics. The selected calli contained nopaline andtransforming DNA sequences (Southern hybridi-zation). Plants could not be regenerated from thetransformed calli.

The same cowpea leaf disc method, with some-what altered selection protocols, was used toinsert a full-length copy of the cowpea mosaicvirus (CPMV) mRNA carried on a binary vectorwith nptll (Garcia et al., 1987). The selectedtransformed callus was used to study the expres-sion of the CPMV cDNA and showed that theCaMV 35S promoter was 10-fold more activethan the nos promoter.

Cowpea transformation was also carried outwith surface-sterilized mature seeds of five geno-types, which were incubated in water for 12 h andthen the embryos with the cotyledons removedwere sliced longitudinally (Penza, Lurquin &Fillippone, 1991). These pieces were soaked for2 min with A. tumefaciens virulent strain A281 car-rying the nptll gene and were then transferred to aculture medium containing 5 mg BA/1 and 1 mgNAA/1 for 2 days. The explants were rinsed in amedium containing 500 mg claforan/1 and afterblotting were grown on a medium containing50 mg G418/1 to select transformed cells.Antibiotic-resistant callus formed on 10% orfewer of the treated embryos but no plants could


be regenerated. The growing cells did containnptll sequences, as shown by dot blots.

The embryo sections were also cocultivatedwith a disarmed A. tumefaciens strain C58(pGV2260) with p35SGUSINT as the binarycontaining both uidA (containing an intron) andnptll genes. In this case the cocultivation andregeneration media both lacked growth regulatorsbut the latter contained 500 mg claforan/1 toinhibit bacterial growth. The shoots that formedafter 21 days from axillary buds were sectionedand stained for GUS activity. About 50% of theshoots contained GUS-expressing cells in thestem but not in leaf tissue. While some cells wereapparently transformed, none of the shoots sur-vived when placed on a medium containing 50 mgG418/1 indicating that too few of the cells wereexpressing NPT II activity.

This cowpea system appears to have promisefor producing transformed shoots, since chimeraltissues were readily obtained. The use of selectionfrom the beginning to eliminate escapes and tochannel only the transformed cells into shoot for-mation would seem to be one possible way toimprove the results.

Dry bean (Phaseolus vulgaris)

Hairy root cultures have been initiated by Aird,Hamill & Rhodes (1988) by infecting aseptic drybean plants at puncture sites on the stems with A.rhizogenes LBA9402. The hairy roots were excisedand were cultured in B5 liquid medium with500 mg ampicillin/1. The cultures contained TLand TR T-DNA sequences, as shown by Southernhybridization. The chromosome number was theeuploid number of 22 after 7 months in culture.No additional information was presented.

Hairy root cultures were also initiated byMcClean et al. (1991) from 1 cm long hypocotylsegments from aseptic seedlings of one genotypeby inoculating A. rhizogenes strain A4RS (pRiB278b) carrying the binary plasmid pGA482 onthe cut end of the segments set upright. After 2days of cocultivation, the segments were trans-ferred two times at 2-day intervals on MS mediumwith 800 fig cefotaxime/ml and 500 |xg carbeni-cillin/ml to inhibit bacteria and then to the samemedium with 75 |mg kanamycin/ml to select fortransformed roots. The roots could be removedand be cultured in liquid medium with 75 |xg

kanamycin/ml. The roots contained NPT IIenzyme activity and nptll gene sequences.

Neither of the reports of hairy root induction ondry bean plants described plant regeneration sothis must not have occurred.

Tumor induction assays were performed with1-day germinated, surface-sterilized dry beanseedlings of 19 diverse genotypes that were inocu-lated with the virulent A. tumefaciens strains A208(nopaline), A281 (agropine) and LBA 4001(octopine) by puncturing the cotyledonary noderegion three times after removal of the cotyledonsand then placing bacteria on the wound (McCleanet aly 1991). Following growth for 7 days onmoistened paper towels in a culture room, theseedlings were transplanted in soil and gall forma-tion determined after 14 days. All genotypesshowed a high susceptibility to A. tumefaciens,with little difference between strains.

One genotype was also inoculated with the avir-ulent strain C58Z707(C58Z707/pGA482) with abinary plasmid containing nos, nptll, nos.Cotyledonary nodal regions from aseptic 1-7-day-old seedlings were inoculated with bacteria afterpuncturing with a needle. Following a 3-daycocultivation on MS medium, the explants weretransferred three times at 3-day intervals on MSmedium with 500 fig carbenicillin/ml and 800 |xgcefotaxime/ml to eliminate bacteria. The explantswere then cultured on the same medium with thecarbenicillin and 200 juig kanamycin/ml to selectfor transformed cells. The kanamycin concentra-tion was gradually increased over a 2-monthperiod to 500 |xg/ml. The selected growing callusdid contain NPT II enzyme activity.

While it is possible to induce hairy roots andtumors readily with dry bean, no transformedplants have apparently been regenerated usingthese systems.

The navy bean cv. Seafarer has been trans-formed using electric-discharge particle accelera-tion to deliver gold particles coated with plasmidDNA into apical meristems of germinating seeds(Russell et al., 1993). Multiple shoots that formedon a high cytokinin medium were screened forGUS activity histochemically or for bialaphosresistance by placing in medium with the herbi-cide. About two shoots formed on each explantand about 0.5% of these expressed GUS activityin at least a small portion. About 0.03% of theshoots produced seed carrying the GUS activity.Some of the GUS-negative shoots showed

112 WIDHOLM

bialaphos resistance and these produced progenythat were not damaged when sprayed with Basta,showing that the bar gene driven by the CAMV35S promoter can impart clearcut herbicide resis-tance. Thus, stably transformed, fertile, dry beanplants can be produced by the electric-dischargeparticle acceleration technique.

Peanut (Arachis hypogaea)

Peanut, or groundnut, is an important food cropthat, like other large-seeded legumes, has beendifficult to manipulate in culture to obtain regen-erated plants. There has been one report describ-ing tumor induction with peanut tissues (Lacorteet al., 1991). Four-week-old, greenhouse-grownplants of five cultivars were inoculated with fourdifferent virulent A. tumefaciens strains (A281,Bo542, A208, T37) by applying the bacteria toneedle wounds on stem internodes. All of thestrains except T37 induced tumors on all geno-types, with A281 producing tumors that appearedearlier and were larger. Tumors could also be pro-duced by all four strains on wounded stems of invitro-grown aseptic plants. These tumors wouldgrow as friable callus on MS medium with nogrowth regulators. The tumor tissues containedthe correct opine.

The A. tumefaciens strain A281 carryingpTDO2 with the nptU and uidA genes was used toinfect cotyledon segments and embryonic axesfrom surface-sterilized seed and leaf and petioleexplants from 7-day-old aseptic, in vitro-grownplantlets. The tissues were inoculated with bacte-ria, blotted and cultured on MS medium for 24 h.After being rinsed and blotted, the tissues werecultured on MS medium with 500 |xg cefo-taxime/ml to inhibit bacterial growth. Tumorsformed with all five cultivars on all explant sourcesexcept cotyledons which formed roots. Bacteria-free tumor tissues would grow without growthregulators and contained agropine. About half ofthe tissues were kanamycin resistant and con-tained NPTII enzyme activity.

This study shows that peanut can be readilytransformed by A. tumefaciens, but there is a needfor plant regeneration protocols so that trans-formed plants can be recovered.

Forage legumesAlfalfa (Medicago sativa), lucerne

Alfalfa is the most widely grown forage legumeand has been one of the easiest legumes to regen-erate from callus (Bingham et al, 1975), suspen-sion cultures (Atanassov & Brown, 1984) andprotoplasts (Kao & Michayluk, 1980). Genotypespecificity for plant regeneration is very evidentand there has even been a recurrent selection pro-gram for plant regeneration ability that hasresulted in the release of genotypes for generaluse (Bingham, 1989, 1991). The first linereleased, Regen-S, was selected mostly from thecultivar Saranac, which showed 12% regenera-tion in cycle 0, 50% in cycle 1 and 67% in cycle 2(Bingham et a/., 1975). Regen-S produces about90% as much herbage as does Saranac and soperforms in a manner close to that of a commer-cial cultivar. The line Regen-SY is especially sus-ceptible to A. tumefaciens transformation. Ideally,however, transformation should be done with thebest current commercial varieties to producerapidly acceptable materials that do not need tobe backcrossed. However, Bingham & McCoy(1986) believed that alfalfa plants regeneratedfrom tissue culture contain enough somaclonalvariation that backcrossing would always beneeded anyway.

There are several examples of alfalfa transfor-mation using A. tumefaciens, including a report byShahin et al. (1986). Surface-sterilized 5 cm longstem segments from mature plants (cultivarCUF101) grown in the field were cut into2-3 mm slices, submerged in a suspension of A.tumefaciens (LBA4404 containing a binary vectorwith nos, nptll), blotted and then incubated on afilter paper on medium containing 2 mg 2,4-D/land 1 mg BA/1. After 48 h the slices were placed ona similar medium with lower BA (0.25 mg/1) andwith 500 mg cefatoxime/1 to eliminate bacteriaand 50 mg kanamycin/1 to select transformed calli.Kanamycin-resistant calli formed on 12%—15% ofthe slices and many plants were regenerated fol-lowing somatic embryo induction on media with2 mg (or higher) 2,4-D/l. All of the selected callitested contained NPT II enzyme activity, but notall of the regenerated plants did. The presence oftransforming DNA in the regenerated plants wasconfirmed by Southern hybridization. The regen-erated plants were morphologically normal andwere fertile.


An autotetraploid Medicago varia Hungariangenotype, which was selected by A. Atanassovfrom a M. sativa X Medicago falcata hybrid, wasused in transformation experiments by Chabaudet al. (1988). Wounded leaflet and petiole seg-ments from aseptic seedlings were dipped into anA. tumefaciens (two strains both with pVW130 car-rying nptll) suspension, blotted with filter papersand cocultivated for 2-4 days on a complexmedium (B5h), which is B5 with 1 mg 2,4-D/l,0.2 mg kinetin/1, 800 mg glutamine/1, 100 mg ser-ine/1, 10 mg glutathione/1 and 1 mg adenine/1using a feeder layer of alfalfa suspension culturedcells or 50 |ULM acetosyringone under a filterpaper. The explants were rinsed with water andwere blotted and then transferred to the samemedium with 300 (xg carbenicillin/ml to inhibitbacterial growth and 50 or 100 |xg kanamycin/mlto select for transformed calli and subsequentlysomatic embryos that begin to form after 6-8weeks. Embryos were matured on a medium lack-ing growth regulators but containing the originalantibiotics. The mature embryos were rooted on amedium that also contained the antibiotics.

In these experiments the leaflets produced moretransformed callus (36%) than petioles (9%) andA. tumefaciens strain A281 was more efficient(34%) than strain LBA4404 (14%). A 4-daycocultivation was better (43%) than 2 (10%) or 3days (19%). There was not much difference intransformation frequency between 50 and 100 |xgkanamycin/ml for selection. The regeneratedplants produced callus with NPTII enzyme activ-ity and nptll sequences. Maximum transformationfrequencies near 70% were obtained when allparameters were optimized.

D'Halluin, Botterman & De Greef (1990) havetransformed M. sativa line RA-3, which is a highlyregenerable genotype selected from Regen-S,using A. tumefaciens strain CSSClRif* carryingthe disarmed T-DNA vector pMP90. The twobinary vectors used contained nptll driven by thenos promoter and the bar gene driven by theCaMV 35S promoter or the nptll and bar genesdriven by the 1' and 2' promoters, respectively, ofthe divergent TR T-DNA promoter as well as hphdriven by CaMV 35S. Stem and petiole segments1 cm long from aseptically grown shoot cultureswere dipped in a late log-phase A, tumefaciens cul-ture and then were cocultivated on top of a filterpaper covering a tobacco feeder layer in a mediumcontaining 25 fiM NAA and 10 |xM kinetin. After

2-3 days the sections were rinsed and were thengrown on the same medium with 500 or 1000 (xgcarbenicillin/ml to eliminate bacteria, andkanamycin, hygromycin or phosphinothricin toselect transformed cells. After about two 3-4-week subcultures on the selective medium,somatic embryogenesis was induced with a 3-4day pulse of 50 |xM 2,4-D in liquid selectivemedium. The embryos were allowed to germinateto form complete plants on selective media over aperiod of one to several months.

In this system petioles produced transformedcallus more readily than they did stems. Selectionimmediately after cocultivation with 50 or100 |ULg/ml kanamycin, 75 |xg hygromycin/ml or5 or 10 |xg phosphinothricin/ml produced trans-formed plants, with kanamycin and hygromycingiving a higher frequency. A large number ofplants were regenerated and most of these wereresistant to the herbicide Basta when sprayed.Plants with the bar gene driven by the CaMV 35Spromoter tended to be more resistant than thosewith the TR2'-bar gene and the levels of phos-phinothricin acetyltransferase activity generallycorrelated with the herbicide resistance. Most ofthe transformed regenerated plants also expressedNPT II.

This study shows that alfalfa can be trans-formed with a useful gene to produce plants thatexpress the gene at varying levels. In this case theherbicide resistance gene is expressed at a highenough level in some plants to impart resistance tothe usual field application rates.

In another attempt to place a useful gene intoalfalfa, Schroeder et al. (1991) used A. tumefaciensstrain LBA4404 with a binary plasmid containinga chicken ovalbumin cDNA clone driven by theCaMV 35S promoter with a nos terminator. Theplasmid also contained nptll. The ovalbumin genewas used because the ovalbumin protein containsa relatively high level of cysteine plus methionine(6.5%). Leaves from greenhouse-grown plants ofthree cultivars, two of which are of commercialimportance, were surface sterilized; segmentswere submerged in a bacterial suspension for10 min, blotted and cocultivated for 3 days on thecomplex B5h medium described above. The seg-ments were then washed and placed on B5hmedium with 50 |xg kanamycin/ml to select trans-formed callus and 100 jig cefotaxime/ml to elimi-nate the bacteria. Callus was transferred after 4weeks to the same medium or to the same

114 WIDHOLM

medium with higher 2,4-D (11 mg/1) and kinetin(1.1 mg/1), where somatic embryos formed inthree more weeks. The embryos were allowed tomature and form plants on media withoutkanamycin.

The noncommercial cultivar Rangelander pro-duced transformed calli on 10%-15% of theexplants, whereas the success rate with the twocommercially important ones was only0.1%-0.5%, even with an altered protocol. Theproduction of somatic embryos from each calluswas also much lower with the latter two cultivars.However, transformed plants were obtained fromall three genotypes; they did express the ovalbu-min cDNA at varying levels to accumulate fromabout 0.001% to 0.01% of the total leaf solubleprotein as measured by immunoblots. The proteinappeared to be stable, as shown by pulse-chaselabeling experiments and by levels found in leavesof different ages. Of 33 Rangelander plants ana-lyzed, 31 contained the ovalbumin cDNA butonly 23 contained the protein.

These results show that it is possible to obtainexpression of a useful gene in commercially impor-tant alfalfa varieties, but the levels of the proteinaccumulated seems to be too low to be significantfor increasing the leaf sulfur amino acid content.Whether another promoter or altered gene canincrease the production is open to debate.

There are also a number of reports of the pro-duction of hairy roots using A. rhizogenes^ fol-lowed by regeneration of transformed plants. Anexample is the work of Sukhapinda, Spivey &Shahin (1987), who used stem segments frommature plants of two cultivars grown in the green-house. The sections were surface sterilized, rinsedthree times, cut into 1-2 cm sections and placedin an inverted position on a medium containing250 |xg cefatoxime/ml. A drop of a bacterial sus-pension was spread on the top of each section.Hairy roots, which formed profusely on most sec-tions in 2-3 weeks, were excised and placed on agrowth-regulator-free medium with 100 |xg cefa-toxime/ml to inhibit bacterial growth. From 43%to 60% of the roots produced nopaline and theyalso expressed the typical hairy root phenotype,lack of geotropism and extensive branching.Embryogenesis was induced by transfer of theroots to a medium with 50 jxM 2,4-D to inducecallus formation and then to a complex mediumcontaining asparagine, glutamine, adenine sulfate,5 jxM 2,4-D and 0.5 jxM BA. After a week the

medium was modified by substituting 2% (w/v)sucrose for 2% glucose. Plants were regenerated,from the embryos formed, by continuous subcul-turing on a medium lacking growth regulators.

A total of 382 transformed plants were pro-duced using A. rhizogenes strain A4 with binaryvectors containing two soybean genes or the maizetransposable element Mu 1 as well as nos. Therewere multiple copies of the TL- and TR-DNA aswell as the Ri plasmid. The plants containednopaline as did their progeny. The transformedplants were phenotypically normal except forextensive and shallow roots.

In a report by Spano et al. (1987), stem seg-ments from 1-2-month-old aseptically grownplants of three cultivars were inoculated with A.rhizogenes strain NCPPB1855 harboring thepRil855 (agropine type) plasmid by injection witha syringe or by placement on the cut ends. Hairyroots were removed after 1 month and wereplaced in liquid MS medium with 1000 mg car-benicillin/1 to inhibit bacterial growth. About 50%of the hairy roots contained agropine. Bacteria-free callus was formed within 2 months on solidmedium with 2-5 mg 2,4-D/l and 0.25 mgkinetin/1 with decreasing levels of the antibiotic.Somatic embryos then formed on solid mediumwithout any growth regulators and these thenformed complete plants on the same medium.Only one of the cultivars, which had previouslybeen shown to be highly regenerative, formedplants. One of the plants that contained agropinealso contained three to five copies of the TL-DNAand two copies of the TR-DNA in truncated form.The regenerated transformed plants had larger,shortened roots with more highly developedlateral roots, unlike the usual tap root system ofnormal alfalfa plants. The transformed plants hadshortened internodes and more and smallerleaves. The plants were fertile.

These two examples of hairy root transforma-tion did produce transformed, fertile plants aftercallus formation and embryogenesis and notdirectly by shoot formation from the roots. Similarresults were obtained by Golds et al (1991),where transformed hairy roots were produced bythree cultivars, but plants were obtained fromroots of only one of the cultivars and only througha callus phase. The hairy root production was effi-cient from stem sections with cotransformationfrequencies of about 50% using both cointegrateand binary vector systems. The callus formation


and embryogenesis system appears to require sev-eral months and is genotype specific. The plantsproduced do have abnormal, compact root sys-tems that are likely to be detrimental under mostfield conditions.

Alfalfa mesophyll protoplasts have been trans-formed via electroporation or PEG-mediateddirect DNA uptake using plasmids with nos, hph,nos or CaMV35S, nptll (Larkin et al, 1990).Transformed colonies were selected withhygromycin or kanamycin, respectively, at fre-quencies near 1QT4 to 10~5. Plants were regener-ated from the kanamycin-selected colonies viasomatic embryogenesis. Samples of selected calliand regenerated plants contained NPT II enzymeactivity.

Protoplasts from leaves of M. borealis line 94shoot cultures were electroporated with pGA472,which carries the nptll gene (Kuchuk et al., 1990).The protoplast viability was about 50%-60% afterelectroporation with 10 |xg plasmid DNA and40 |xg of calf thymus DNA using 5 X 105 proto-plasts at 560 V/cm and 1 |xF. Colonies whichformed after 2-3 weeks were transferred to a solid-ified medium with 50 |xg kanamycin sulfate/mland 1 mg BA/1. A total of 56 green coloniesformed, of which 15 expressed stable kanamycinresistance. Shoot regeneration was obtained fromfive of these clones and roots did form on theseshoots in the presence of kanamycin. The twoplants tested contained NPT II enzyme activity.

Another possible method for transformingalfalfa was investigated by Senaratna et al. (1991),who prepared desiccated somatic embryos (about10%—15% water) and then these were incubatedfor 2 h with pBI221.2 with the uidA gene drivenby the CaMV 35S promoter. GUS activity wasseen after histochemical staining on the surface ofgerminating embryos 3 days after imbibition andon up to 80% of the plantlets formed 14 days afterimbibition. PCR analysis also showed the pres-ence of a 250 bp segment spanning the promoterand uid A gene in embryos imbibed with the plas-mid DNA.

These results indicate that some cells can takeup plasmid DNA upon imbibition but few of theapical meristematic cells that form the plant weretransformed. This method has been used previ-ously with imbibing seeds (Ledoux & Huart,1968) but in general the overall results obtained inthe past work did not clearly demonstrate trans-formation of the resulting plants. Thus, the use-

fulness of this general method needs further con-firmation.

Certain alfalfa genotypes can be readily trans-formed using A. tumefaciens or by direct DNAuptake with protoplasts to produce transgenic fer-tile plants that express the integrated genes. A lim-itation on the process is the genotype specificity ofplant regeneration. Agrobacterium rhizogenes alsocan transform efficiently, but here again there isplant regeneration genotype specificity and thetransformed plants have the generally undesirablehairy root phenotype of shortened and morehighly branched roots.

The genotype specificity of transformation ofalfalfa apparently may not be an important limita-tion because there is enough variation inducedduring tissue culture so that several backcrosseswill be needed to eliminate this and produce com-mercially acceptable lines, according to Bingham& McCoy (1986).

Birdsfoot trefoil (Lotus corniculatus)

A number of forage legumes other than alfalfahave been transformed using A. rhizogenes and A,tumefaciens. There have been several reports withL. corniculatus, including that by Jensen et al.(1986), who infected wound sites of asepticseedlings with A. rhizogenes strain 15834 carryinga cat gene driven by a soybean leghemoglobin pro-moter inserted into the T-region of the Ri plas-mid. The roots that formed at the wound siteswere cultured in the presence of 500 |xgclaforan/ml to eliminate the bacteria. Plants wereregenerated apparently after callus induction with2,4-D through somatic embryogenesis or organo-genesis. The methods used to accomplish thiswere not explained. The regenerated plants werestudied at the molecular level as far as integratedDNA structure and expression of the cat geneunder the control of the soybean leghemoglobinpromoter. The gene control was root nodule spe-cific just as in the soybean plant.

Birdsfoot trefoil was also transformed by Petit etal. (1987): hypocotyls of aseptic seedlings werewounded with a needle or scalpel dipped in A. rhi-zogenes strains 15834 (agropine-type C58) or8196 (mannopine-type). After 14 days, stempieces with developing roots (up to 90% formedroots) were transferred to medium with 500 |xgclaforan/ml and then the roots were excised andcultured further on the same medium for 20-30

116 WIDHOLM

days in the dark. Shoots formed within 20-30days in the light and could be excised and rootedto form complete plants. The process from inocu-lation to potting of plants took about 132 days.About 90% of the 30 transformed plants exam-ined were transformed, since they synthesizedopines and contained Ri DNA sequences. Rootnodules appeared to have normal nitrogenaseactivities and transcript levels for nodule-specificleghemoglobin and constitutive ubiquitin. Theseplants apparently had the typical hairy rootphenotype and have been used in basic studiesof this phenotype (Shen et al, 1988). These stud-ies showed that, unlike A. tumefaciens where auxinand cytokinin synthesis carried by the Ti DNAcaused the tumorous phenotypes, the Ri genessomehow make the transformed cells more sensi-tive to auxin.

In another study, seedling hypocotyl and stemsegments from 1-2-month-old L. corniculatusplants were incubated with A. rhizogenes for1-2 min, and after 2-3 days were then incubatedwith 300 |xg cefotaxime/ml to eliminate the bacte-ria (Tabaeizadeh, 1989). The wild-type A. rhizo-genes contained the binary pBinl9 with nos, nptliand CaMV 35S, cat, nos gene constructs. Rootsformed only on inoculated segments (15% and35% of the stem and hypocotyl segments,respectively). Plant regeneration was obtained onregeneration medium containing up to 30 |xgkanamycin/ml from about 50% of the hairy roots.Shoots could be regenerated from wild-type rootsin up to 10 |xg kanamycin/ml. Enzyme activities(CAT, NPT II) could be detected in transformedshoots and roots, and cat gene sequences were alsopresent in the transformed plants. The processfrom inoculation to transformed plant productionrequired only 2 months. These regenerated trans-formed L. corniculatus plants 'did not show anymorphological differences with respect to leafshape, plant height, etc. compared to seed grownplants' according to the author Tabaeizadeh(1989).

Hairy root cultures were also initiated by Webbet al (1990) from stolon segments or seedlings ofL. corniculatus, T. repens and T. pratense, followingincubation with a wild-type A. rhizogenes strainC58C1 using a needle. Following decontamina-tion by growth with several antibiotics during sev-eral subcultures, the cultures were maintained inliquid or on solid medium. The integrated copynumber of the TL-DNA varied from one to eight,

while the chromosome number was near normal.Plants could not be regenerated from the twoTrifolium species but shoots regenerated sponta-neously from the L. corniculatus hairy roots after 8weeks in culture. Many of these plants had normalcharacteristics (height, stem thickness, tannincontent, flower color, plant dry weight) but rootmorphology, chlorophyll content, leaf size andflower and seed production were altered. Nitrogenfixation was not affected.

Transgenic L. corniculatus plants containing afull-length cDNA clone of the soybean cytosolicglutamine synthetase, which is expressed in rootsand root nodules, were obtained, using the meth-ods of Petit et al. (1987), by Miao et aL (1991). Byusing a glutamine synthetase-wzdA gene fusionconstruct the expression was followed by measur-ing GUS activity histochemically. The gene wasexpressed with the correct tissue specificity in L.corniculatus and tobacco, which was also trans-formed.

Rapidly growing birdsfoot trefoil hairy root cul-tures were obtained following transformation of8-10-day-old aseptic seedling hypocotyls withwild-type A. rhizogenes (C58Cl-pRi 15834) (Morris& Robbins, 1992). Shoots formed readily uponillumination in liquid medium. These shootscould form fertile regenerated plants or could bemaintained in liquid medium as shoot-organcultures. Condensed tannin formation was stud-ied in these cultures and in callus initiated fromthem.

Transformation of L. corniculatus was alsoaccomplished using A. tumefaciens by Yu & Shao(1991). Cotyledons excised from axenic 5-10-day-old seedlings were cut in half, incubated for10 min with A. tumefaciens strain C58C1 contain-ing a selectable marker gene (nos, nptll) and areporter gene (nos), blotted dry with filter paperand then cocultivated for 2 days on a mediumwith 0.1 mg BA/1. Selection was then carried outon the same medium with 100 mg kanamycin/1and bacterial growth inhibited by 300 (xg cefo-taxime/ml. Within 21 days, buds had formed ongrowing calli on 80 of the 200 treated cotyledonexplants. Shoots longer than 3 cm were rooted ingrowth-regulator-free medium and were thentransplanted to pots where seed was produced. Allselected calli, shoots and plants contained nopa-line. The one plant analyzed contained NPT IIactivity and T-DNA sequences as determined bydot blot. The progeny contained nopaline, indi-


eating that the transforming DNA was passed tothe next generation.

Birdsfoot trefoil thus can be readily transformedwith both A. rhizogenes or A. tumefaciens to obtainfertile transformed plants. The success is due tothe rapid regeneration of shoots directly fromhairy roots and from callus. The plants regener-ated from hairy roots generally do have the typicalaltered root phenotype, which is likely to affectfield performance.

White clover (Trifolium repens)

A highly regenerable white clover genotype wastransformed by placing A. tumefaciens on the cutends of stolon internode segments followed byincubation for 1-3 days in a moist Petri dish(White & Greenwood, 1987). The ends were cutoff and placed on callusing medium with 500 |xgcefotaxime/ml to stop bacterial growth and100 |xg kanamycin/ml to select transformed plantcells. Kanamycin-resistant growth was obtained,with the four different binary vectors used, on24%-44% of the treated segments with a 3-daycocultivation period. The callus was cultured onthe antibiotic medium with 21-day subcultureintervals for about 3 months, when plants wereregenerated on regeneration medium with orwithout 100 |xg kanamycin/ml. The regeneratedshoots produced nopaline with certain constructs,had NPT II enzyme activity and contained one totwo copies of the nptll gene sequences. One rea-son for the success of this system is the high regen-eration capacity of the callus of the genotype used.

Hairy roots were also produced by Webb et al.(1990) from stolon segments or seedlings of T.repens and T. pratense, as described in the birdsfoottrefoil section, above. Whole plants could not beregenerated from these roots.

Stylosanthes species

Tumors were induced on leaf sections of theStylosanthes species humilis^ hamata, guianensis andscabra by A. tumefaciens strains A281 and C58 atrelatively low frequencies, with no tumors beingformed on some cultivars of the latter two species(Manners, 1987).

Stems and leaf sections from 4-5-week-old 5.humilis seedlings grown under aseptic conditionswere treated with an A. rhizogenes suspension for2 min, blotted dry and incubated for up to 4 days

before washing and transfer to medium with500 |xg cefotaxime/ml to inhibit bacterial growthand 50 |xg kanamycin/ml to select transformedcells (Manners & Way, 1989). The A. rhizogenescontained the binary plasmid pGA492 (wos,nos-nptll gene fusion). The highest frequency oftransformed root production was 86% with stems,following a 3-day coincubation period in the dark.After 6 weeks the bacteria-free kanamycin-resistant transformed roots were placed on shootregeneration medium with different BA concen-trations, where shoots formed on about 23% ofthe roots with 2 mg BA/1. Transformed plantscould be produced within about 18-24 weeksafter inoculation.

Opines were found in some of the transformedroots but not in any of the regenerated plants,whereas NPT II activity was found in all roots andregenerated plants. Most of the plants had mor-phological abnormalities, with about half beingdwarf and multistemmed. All plants were fertileand the dwarfhess could be separated fromkanamycin resistance in the next generation. Onlyone of the 25 regenerated plants had completelynormal morphology and this plant was kanamycinresistant but lacked the TL-DNA sequences, indi-cating that the Ri virulence plasmid was not trans-ferred in this case.

Lotononis bainesii

Lotononis bainesii is a subtropical forage legumethat can be easily regenerated from tissue cultures(Bovo, Mroginski & Rey, 1986). Leaf discs weretransformed (Wier et al.> 1988) by A. tumefacienscarrying a gene fusion of the cat gene and a syn-thetic DNA sequence that codes for a peptidecontaining relatively high concentrations of theessential amino acids lysine (16.7%, w/v), trypto-phan (11.5%) and methionine (8.3%); methodsused were similar to those developed for tobaccoby Horsch et al (1985). The chimeric fused genewas driven by the CaMV 19S promoter with thenos terminator. Transformed callus was selectedwhen the discs were incubated after 48 h ontobacco nurse cultures on medium with 30 |xgkanamycin sulfate/ml (the plasmid also carried thenptll gene) and 500 jxg carbenicillin/ml to elimi-nate the bacteria. The callus was placed on freshselective medium, where shoots grew and formedroots on 300 fig kanamycin/ml but untransformedshoots did not grow or form roots and had low-

118 WIDHOLM

ered chlorophyll levels. Seeds obtained from threeof the transformed plants showed kanamycinresistance with segregation ratios fitting a domi-nant, single gene model. Amino acid analysis ofwhole plants showed some increases in certain ofthe expected amino acids in some plants.

Sainfoin {Onobrychis vidifolia)

Hypocotyls and cotyledons of aseptic 10-14-day-old sainfoin plants were inoculated by injecting Arhizogenes into wounds and then were incubated inan upright position for 21-28 days in solidmedium (Golds et al, 1991). From 50% to 78%of the hypocotyls or cotyledons, respectively,formed hairy roots with strain A4T, which carriedpRiA46. Inoculated leaves were unresponsive.Most (70% to 80%) of the roots would growrapidly on a medium with 500 |xg cefotaxime/ml.Shoots regenerated spontaneously from some ofthe root cultures to form plants that were shorterand had smaller leaves and prolific, irregularlybranched masses of fine negatively geotropicroots. The transformed roots contained agropineand mannopine but the regenerated plants didnot. Southern hybridization confirmed the pres-ence of TL- and TR-DNA sequences in theregenerated plants.

Transformed plants can be readily obtained withbirdsfoot trefoil, white clover, 5. humilus, L. baine-sii and sainfoin as described in this section usingAgrobacteriwn systems and either shoot regenera-tion from hairy roots or plant regeneration fromcallus. The transformation success dependsgreatly on the relative ease of obtaining regenera-tion from these systems.

Summary and conclusions

A summary of the published results for eachspecies discussed here is given in Table 8.1. Anumber of the spaces are blank because it is diffi-cult to determine the category to use on the basisof the results presented. This summary is given toshow what is reported at this time and manychanges would be expected in the future as moreprogress is made.

The grain legumes as a group have producedtransformed progeny in the cases of soybean, drybean, pea and V. narbonensis (Table 8.1). Trans-

formed plants have also been produced by mothbean but there are no reports of progeny. Theother grain legumes listed, broad bean, lentil,cowpea and peanut, have not produced trans-formed plants. The deficiency with the grainlegumes lies predominantly with the recalcitranceof these species in plant regeneration. One wouldanticipate that further effort will surmount thisproblem, just as it has with soybean where about10 years ago some workers felt that this speciescould not regenerate plants from tissue culture.Since then, callus, suspension and protoplastregeneration systems have been developed.

The soybean transformation is not easy and allof the successful techniques, A. tumefaciens infec-tion, microprojectile bombardment and proto-plast electroporation, have shortcomings. The A.tumefaciens system is low frequency, produceschimeric plants with many escapes and is verygenotype specific. The microprojectile system isnot genotype specific but requires large numbersof immature embryos, is low frequency, produceschimeric plants and uses expensive equipmentand supplies, and royalties also apply if a commer-cial product is produced. If microprojectiles areused to transform suspension cultures then theembryogenic culture must first be initiated, whichis difficult. The use of suspensions improves theselection so that nonchimeric plants can be recov-ered, but again the equipment and royalty costscan be high. The protoplast system should beapplicable to about half the genotypes, is slow (8months or more) but should produce non-chimeric, fertile plants, although this has not beenconfirmed as yet. Certainly improvements may bemade in all of the systems to make them moreideal.

There has been better success with the foragelegumes where alfalfa and birdsfoot trefoil can bereadily transformed to produce regenerated, fer-tile plants. The one problem with alfalfa is thatplant regeneration shows genotype specificity sothat the best commercial genotypes cannot betransformed easily.

The other four forage legumes listed, whiteclover, Stylosanthes, L. bainesii and sainfoin, havebeen successfully transformed to produce plantsand those of Stylosanthes and L. bainesii have pro-duced seed. The most apparent reason for the bet-ter success with the forage, in contrast to that withthe grain legumes, is the superior plant regenera-tion capabilities of the former.


Table 8.1. Summary of transformation results with leguminous plants

Species

Soybean

Pea

Moth beanFaba beanVicia narbonensisLentilCowpeaDry bean

PeanutAlfalfa

Birdsfoot trefoil

White clover

Stylosanthes

Lotononis bainesiiSainfoin

Method"

Af.A.r.Proj.Proj.

Proto.At .A.r.Proto.Proto.A.r.A.t.A.t.A.t.A.t.A.r.Proj.A.t.A.t.A.r.Proto.A.tA.r.

A.t.A.r.

A.t.A.r.A.t.A.t.

Startingmaterial

SeedlingSeedlingImmature seedSuspensionculture

Immature seedPlantPlantSeedlingSeedlingSeedlingSeedlingSeedlingSeedlingSeedlingSeedlingMature seedPlantSeedlingPlantsLeavesSeedlingSeedling

StolonStolonSeedlingLeavesPlantLeavesSeedling

Genotypespecific?

YesIntermed.NoNo

Intermed.Intermed.YesYesYes

No

No

No

YesYes

Yes

Yes

Yes

Regenerateplants?

YesNoYesYes

YesYesNoNoYesNoYes

NoNoNoYesNoYesYesYesYesYes

YesNo

NoYesYes

Plantsfertile?

Yes

YesYes

YesNo

No

Yes

Yes

YesYes

YesYes

YesYes

Remarks

Plants not transformed

RapidRapid, shoots directly fromroot

Notes:aA.t, Agrobacterium tumefaciens; A.r., Agrobacterium rhizogenes; Proj., microprojectile bombardment; Proto., protoplastdirect DNA uptake.

While there have been no reports of the produc-tion of transformed forage legume plants usingprotoplast systems, there have been many reportsdescribing the regeneration of plants from proto-plasts, so the technique should be possible. Plantshave been regenerated from protoplasts of S. guia-nensis, L. corniculatus and T. repens (Webb, Wood-cock & Chamberlain, 1987) among others, asreviewed by Vieira et al (1990).

As with many other species, some legumes canform shoots from hairy roots, although only birds-foot trefoil seems to form shoots directly withoutan intervening callus phase. In a review by Tepfer(1990), a total of 116 dicotyledonous species arelisted that have been transformed with A rhizo-

genes; transformed plants have been regeneratedfrom 37 of them. However, not all of the plantregeneration was direct formation from the root.

In the transformation systems that utilize A. rhi-zogenes for gene insertion, almost all of the regen-erated plants carry the Ri T-DNA, which impartscertain abnormal characteristics to the plants,including shortened and enlarged roots. Whilethis type of root, which permeates a smaller soilmass than normal, might be advantageous undercertain conditions, in general this phenotypewould be detrimental.

It should be possible with binary Ri plasmid sys-tems to separate the integrated Ri plasmid fromthe integrated binary vector DNA by genetic seg-

120 WIDHOLM

regation if these sequences are not closely linkedon the chromosome. This has been accomplishedwith one Ti system (de Framond et al> 1986).This separation could eliminate the detrimentalhairy root phenotype but retain the useful gene.

Most legumes, being dicots, are sensitive toboth A. tumefaciens and A. rhizogenes, althoughthere can be some host genotype specificity, aswell as bacterial strain specificity. Overall, how-ever, relatively good infection rates can beattained with the right combinations of plantgenotype and Agrobacterium strain.

In several places in this chapter the productionof chimeric plants was described. This conclusionwas reached because only certain sectors of thetransformed plants expressed the transformedgene and in a few documented cases only seedsfrom these sectors produced plants that alsoexpressed the transformed genes. This informa-tion does not allow a determination of whether ornot the nonexpressing tissue actually contains thetransformed gene in an inactive state. Christou(1990) showed, however, that, when some of theGUS nonexpressing tissues in chimeric soybeanplants were analyzed, the uidA gene was absent.Thus, only in rare cases has the absence of thetransformed gene been documented in chimericplants. The most correct description of theseplants would be that they are chimeric for expres-sion of the transformed genes unless further mole-cular studies have been carried out.

While somaclonal variation does occur withmost tissue culture systems, the results presentedhere and our results with soybean protoplasts andorganogenic and embryogenic cultures (Barwale& Widholm, 1987) indicate that somaclonal varia-tion occurs at a low enough frequency that the fewvariants can be discarded with little harm. Theremaining normal-appearing soybean families donot show any detrimental agronomic characteris-tics in field trials (Stephens, Nickell & Widholm,1991). One should therefore always regenerate anumber of plants (10-20?) so that the undesirableones can be discarded. This is especially impor-tant from the standpoint of gene expression, sincethis is known to often be variable in different indi-viduals. Enough plants need to be available to givea choice in order to find the desired expressionlevel. In the case of alfalfa, however, Bingham &McCoy (1986) felt that the somaclonal variationis so great in any regenerated plants that back-crossing is required in all cases.

There can be some debate about the effect ofhow the genes are integrated because the Agro-bacterium vectors integrate a more or less set pieceof DNA, in contrast to the direct DNA uptake andmicroprojectile systems where random pieces andconcatemers can be integrated. This discussioncan best be carried out with other systems, sincethe data with legumes are rather limited at themoment, although successful gene expression inprogeny has certainly been observed with bothsystems.

There are many general traits that could beimproved in any crop plant including disease,drought, heat, cold, salt and insect resistance,depending upon the growing area. Some of thesecan already be targeted since there are antifungal(chitinase, p-glucanase) and insecticidal (Bacillusthuringiensis endotoxins) genes available. Most ofthe other traits cannot be approached presently,since the genes controlling these traits are notavailable.

Nutritional quality is also very important toboth grain and forage legumes. Legume seeds aregenerally deficient in sulfur amino acids so theaddition of proteins high in methionine is beingattempted (Parrott et al, 1989).

There have already been a number of transgenicplants produced with genes that could improvethe crop. These would include soybean withglyphosate (Hinchee et al.y 1988) or atrazineresistance (Liu et al> 1990), alfalfa with Bastaresistance (D'Halluin et al, 1990) or chickenovalbumin expression (Schroeder et at., 1991),dry bean with Basta resistance (Russell et al.>1993) and L. bainesiiwixh a synthetic gene encod-ing a protein with a high lysine, tryptophan andmethionine content (Bovo et aly 1986). Many ofthe transgenic plants also synthesize opines somight be useful in testing the 'opine concept'(Petit et aly 1983), which postulates that opinesproduced by plants act as chemical mediators ofparasitism between Agrobacterium and the plants.A symbiosis might be encouraged if beneficialbacteria contained the opine utilization genes.

Basic studies include the integration of trans-posable elements into soybean (Zhou & Atherly,1990) and alfalfa (Sukhapinda et al, 1987) andstudies of tissue-specific expression of the soybeanleghemoglobin gene (Jensen et al, 1986) and theglutamine synthetase gene (Miao et al, 1991),both in birdsfoot trefoil. Basic studies on the auxinsensitivity of hairy roots were also done with birds-


foot trefoil (Shen et al, 1988). Facciotti et al(1985) studied light induction of a chimeric soy-bean small subunit gene in soybean callus trans-formed by A. tumefaciens.

While the transformation progress with legumeshas been slow, we now have some systems that aresuccessful and more progress is expected. Thus,there should be clear evidence of crop improve-ment in the next few years. Then questions aboutwhich new genes to use, degree of stability of geneexpression and the effect of these genes on plantperformance will have to be addressed.

Acknowledgements

I thank E. T. Bingham and Guangbin Luo forreading the manuscript and for making useful sug-gestions, and T. Eriksson, J. Puonti-Kaerlas, O.Schieder and D. R. Russell for supplying manu-scripts in press or unpublished information. Theunpublished data were obtained with the supportof the Illinois Agricultural Experiment Station andthe Illinois Soybean Program Operating Board.

References

Aird, E. L. H., Hamill, J. D. & Rhodes, M. J. C.(1988). Cytogenetic analysis of hairy root culturesfrom a number of plant species transformed byAgrobacterium rhizogenes. Plant Cell, Tissue and OrganCulture, 15, 47-57.

Atanassov, A. & Brown, D. C. W. (1984). Plantregeneration from suspension culture and mesophyllprotoplasts ofMedicago sativa L. Plant Cell, Tissueand Organ Culture, 3, 149-162.

Barwale, U. B. & Widholm, J. M. (1987). Somaclonalvariation in plants regenerated from cultures ofsoybean. Plant Cell Reports, 6, 365-368.

Bercetche, J., Chriqui, D., Adam, S. & David, C.(1987). Morphogenesis and cellular reorientationsinduced by Agrobacterium rhizogenes (strains 1855,2659 and 8196) on carrot, pea and tobacco. PlantScience, 52, 195-210.

Bingham, E. T. (1989). Registration of Regen-S alfalfagermplasm useful in tissue culture andtransformation research. Crop Science, 29,1095-1096.

Bingham, E. T. (1991). Registration of alfalfa hybridRegen-SY germplasm for tissue culture andtransformation research. Crop Science, 31, 1098.

Bingham, E. T., Hurley, L. V., Kaatz, D. M. &Saunders, J. W. (1975). Breeding alfalfa which

regenerates from callus tissue in culture. CropScience, 15, 719-721.

Bingham, E. T. & McCoy, T. J. (1986). Somaclonalvariation in Alfalfa. Plant Breeding Reviews, 4,123-152.

Bovo, O. A., Mroginski, L. A. & Rey, H. Y. (1986).Regeneration of plants from callus tissue of thepasture legume Lotononis bainesii. Plant Cell Reports,5, 295-297.

Byrne, M. C , McDonnell, R. E., Wright, M. S. &Carnes, M. G. (1987). Strain and cultivar specificityin the Agrobacterium - soybean interaction. PlantCell, Tissue and Organ Culture, 8, 3-15.

Chabaud, M., Passiatore, J. E., Cannon, F. &Buchanan-Wollaston, V. (1988). Parametersaffecting the frequency of kanamycin resistant alfalfaobtained by Agrobacterium tumefaciens mediatedtransformation. Plant Cell Reports, 7, 512-516.

Chee, P. O., Fober, K. A. & Slightom, J. L. (1989).Transformation of soybean {Glycine max) byinfecting germinating seeds with Agrobacteriumtumefaciens. Plant Physiology, 91, 1212-1218.

Christou, P. (1990). Morphological description oftransgenic soybean chimeras created by the delivery,integration and expression of foreign DNA usingelectric discharge particle acceleration. Annals ofBotany, 66, 379-386.

Christou, P. & McCabe, D. E. (1992). Prediction ofgerm-line transformation events in chimeric R0transgenic soybean plantlets using tissue-specificexpression patterns. Plant Journal, 2, 283-290.

Christou, P., McCabe, D. E. & Swain, W. F. (1988).Stable transformation of soybean callus by DNA-coated gold particles. Plant Physiology, 87, 671-674.

Christou, P., Swain, W. F., Yang, N. S. & McCabe,D. E. (1989). Inheritance and expression of foreigngenes in transgenic soybean plants. Proceedings of theNational Academy of Sciences, USA, 86, 7500-7504.

de Framond, A. J., Back, E. W., Chilton, W. S.,Kayes, L. & Chilton, M. D. (1986). Two unlinkedT-DNAs can transform the same tobacco plant celland segregate in the Fl generation. Molecular andGeneral Genetics, 202, 125-131.

D'Halluin, K., Botterman, J. & De Greef, W. (1990).Engineering of herbicide-resistant alfalfa andevaluation under field conditions. Crop Science, 30,866-871.

Dhir, S. K., Dhir, S., Hepburn, A. & Widholm, J. M.(1991a). Factors affecting transient gene expressionin electroporated Glycine max protoplasts. Plant CellReports, 10, 106-110.

Dhir, S. K., Dhir, S., Savka, M. A., Belanger, F., Kriz,A. L., Farrand, S. K. & Widholm, J. M. (1992a).Regeneration of transgenic soybean (Glycine max)plants from electroporated protoplasts. PlantPhysiology, 99, 81-88.

Dhir, S. K., Dhir, S., Sturtevant, A. P. & Widholm,

122 WIDHOLM

J. M. (19916). Regeneration of transformed shootsfrom electroporated soybean {Glycine max (L.)Merr.) protoplasts. Plant Cell Reports, 10, 97-101.

Dhir, S. K., Dhir, S. & Widholm, J. M. (1991c). Plantletregeneration from cotyledonary protoplasts of soybean{Glycine max (L.) Merr.). Plant Cell Reports, 10, 39-43.

Dhir, S. K., Dhir, S. & Widholm, J. M. (19926).Regeneration of fertile plants from protoplasts ofsoybean {Glycine max L.): genotypic differences inculture response. Plant Cell Reports, 11, 285-289.

Eapen, S., Kohler, F., Gerdemann, M. & Schieder, O.(1987). Cultivar dependence of transformation ratesin moth bean after co-cultivation of protoplasts withAgrobacterium tumefaciens. Theoretical and AppliedGenetics, IS, 207-210.

Facciotti, D., O'Neal, J. K., Lee, S. & Shewmaker,C. K. (1985). Light-inducible expression of achimeric gene in soybean tissue transformed withAgrobacterium. Bio/Technology, 3, 241-246.

Finer, J. J. & McMullen, M. D. (1991).Transformation of soybean via particlebombardment of embryogenic suspension culturetissue. In Vitro Cellular and Developmental Biology,27P, 175-182.

Finer, J. J. & Nagasawa. A. (1988). Development of anembryogenic suspension culture of soybean [Glycinemax (L.) Merrill]. Plant Cell, Tissue and OrganCulture, 15, 125-136.

Gamborg, O. L., Miller, R. A. & Ojima, K. (1968).Nutrient requirements of suspension cultures ofsoybean root cells. Experimental Cell Research, 50,148-151.

Garcia, J. A., Hille J. & Goldbach, R. (1986).Transformation of cowpea {Vigna unguiculata) cellswith an antibiotic resistance gene using a Tiplasmid-derived vector. Plant Science, 44, 37-46.

Garcia, J. A., Hille, J., Vos, P. & Goldbach, R. (1987).Transformation of cowpea Vigna unguiculata with afull-length DNA copy of cowpea mosaic virus m-RNA. Plant Science, 48, 89-98.

Golds, T. J., Lee, J. Y., Husnain, T., Ghose, T. K. &Davey, M. R. (1991). Agrobacterium rhizogenesmediated transformation of the forage legumesMedicago sativa and Onobrychis viciifolia. Journal ofExperimental Botany, 42, 1147-1157.

Hinchee, M. A. W., Conner-Ward, D. V., Newell,C. A., McDonnell, R. E., Sato, S. J., Gasser, C. S.,Fischhoff, D. A., Re, D. B., Fraley, R. T. & Horsch,R. B. (1988). Production of transgenic soybeanplants using Agrobacterium-mediated DNA transfer.Bio/Technology, 6, 915-922.

Hobbs, S. L. A., Jackson, J. A., Baliski, D. S.,DeLong, C. M. O. & Mahon, J. D. (1990).Genotype- and promoter-induced variability intransient p-glucuronidase expression in peaprotoplasts. Plant Cell Reports, 9, 17-20.

Hobbs, S. L. A., Jackson, J. A. & Mahon, J. D. (1989).

Specificity of strain and genotype in thesusceptibility of pea to Agrobacterium tumefaciens.Plant Cell Reports, 8, 274-277.

Horsch, R. B., Fry, J. E., Hoffmann, N. L., Eichholtz,D., Rogers, S. G. & Fraley, R. J. (1985). A simpleand general method for transferring genes intoplants. Science, 227, 1229-1231.

Hussey, G., Johnson, R. D. & Warren, S. (1989).Transformation of meristematic cells on the shootapex of cultured pea shoots by Agrobacteriumtumefaciens and A. rhizogenes. Protoplasma, 148,101-105.

Jensen, J. S., Marcker, K. A., Otten, L. & Schell, J.(1986). Nodule-specific expression of a chimaericsoybean leghaemoglobin gene in transgenic Lotuscorniculatus. Nature {London), 321, 669-674.

Kao, K. N. & Michayluk, M. R. (1980). Plantregeneration from mesophyll protoplasts of alfalfa.Zeitschriftfur Pflanzenphysiologie, 96, 135-141.

Kathen, A. D. & Jacobsen, H. J. (1990). Agrobacteriumtumefaciens-mediated transformation of Pisumsativum L. using binary and cointegrate vectors.Plant Cell Reports, 9, 276-279.

Kohler, F., Cardon, G., Pohlman, M., Gill, R. &Schieder, O. (1989). Enhancement oftransformation rates in higher plants by low-doseirradiation: are DNA repair systems involved in theincorporation of exogenous DNA into the plantgenome? Plant Molecular Biology, 12, 189-199.

Kohler, F., Golz, C , Eapen, S., Kohn, H. & Schieder,O. (1987a). Stable transformation of moth beanVigna aconitifolia via direct gene transfer. Plant CellReports, 6, 313-317.

Kohler, F., Golz, C , Eapen, S. & Schieder O.(19876). Influence of plant cultivar and plasmid-DNA on transformation rates in tobacco and mothbean. Plant Science, 53, 87-91.

Kuchuk, N., Komarnitski, I., Shakhovsky, A. & Gleba,Y. (1990). Genetic transformation of Medicagospecies by Agrobacterium tumefaciens andelectroporation of protoplasts. Plant Cell Reports, 8,660-663.

Kumar, V., Jones, B. & Davey, M. R. (1991).Transformation by Agrobacterium rhizogenes andregeneration of transgenic shoots of the wild soybeanGlycine argyrea. Plant Cell Reports, 10, 135-138.

Kysely, W., Myers, J. R., Lazzeri, P. A., Collins, G. B.& Jacobsen, H. J. (1987). Plant regeneration viasomatic embryogenesis in pea {Pisum sativum L.).Plant Cell Reports, 6, 305-308.

Lacorte, C , Mansur, E., Timmerman, B. & Cordeiro,A. R. (1991). Gene transfer into peanut {Arachishypogaea L.) by Agrobacterium tumefaciens. Plant CellReports, 10, 354-357.

Larkin, P. J., Taylor, B. H., Gersmann, M. & Brettell,R. I. S. (1990). Direct gene transfer to protoplasts.Australian Journal of Plant Physiology, 17, 291-302.


Ledoux, L. & Huart, R. (1968). Integration andreplication of DNA of M. lysodeikticus in DNA ofgerminating barley. Nature (London), 218,1256-1259.

Lehminger-Mertens, R. & Jacobsen, H. J. (1989).Plant regeneration from pea protoplasts via somaticembryogenesis. Plant Cell Reports, 8, 379-382.

Liu, B., Yue, S., Hu, N., Li, X., Zhai, W., Li, N.,Zhu, R., Zhu, L., Mao, D. & Zhou, P. (1990).Transfer of the atrazine-resistance gene of blacknightshade to soybean chloroplast genome and itsexpression in transgenic plants. Science in China B,33, 444-452.

Lulsdorf, M. M., Rempel, H., Jackson, J. A., Baliski,D. S. & Hobbs, S. L. A. (1991). Optimizing theproduction of transformed pea (Pisum sativum L.)callus using disarmed Agrobacterium tumefaciens.Plant Cell Reports, 9, 479-483.

Luo, X., Zhao, G. & Tian, Y. (1990) Plantregeneration from protoplasts of soybean (Glycinemax L.). Acta Botanica Sinica, 32, 616-621.

Manners, J. M. (1987). Transformation of Sty fosanthesspp. using Agrobacterium tumefaciens. Plant CellReports, 6, 204-207.

Manners, J. M. & Way, H. (1989). Efficienttransformation with regeneration of the tropicalpasture legume Stylosanthes humilis usingAgrobacterium rhizogenes and a Ti plasmid-binaryvector system. Plant Cell Reports, 8, 341-345.

McCabe, D. E., Swain, W. F., Martinell, B. J. &Christou, P. (1988). Stable transformation ofsoybean (Glycine max) by particle acceleration.Bio/Technology, 6, 923-926.

McClean, P., Chee, P., Held, B., Simental, J., DrongR. F. & Slightom, J. (1991). Susceptibility of drybean (Phaseolus vulgaris L.) to Agrobacteriuminfection: transformation of cotyledonary andhypocotyl tissues. Plant Cell, Tissue and OrganCulture, 24, 131-138.

Miao, G. H., Hirel, B., Marsolier, M. C , Ridge,R. W. & Verma, D. P. S. (1991). Ammonia-regulated expression of a soybean gene encodingcytosolic glutamine synthetase in transgenic Lotuscorniculatus. Plant Cell, 3, 11-22.

Morris, P. & Robbins, M. P. (1992). Condensedtannin formation by Agrobacterium rhizogenestransformed root and shoot organ cultures of Lotuscorniculatus. Journal of Experimental Botany, 43,221-231.

Murashige, T. & Skoog, F. (1962). A revised mediumfor rapid growth and bioassays with tobacco tissuecultures. Physiologia Plantarum, 15, 473-497.

Nauerby, B., Madsen, M., Christiansen, J. &Wyndaele R. (1991). A rapid and efficientregeneration system for pea (Pisum sativum), suitablefor transformation. Plant Cell Reports, 9, 676-679.

Nisbet, G. S. & Webb, K. J. (1990). Transformation

in legumes. In Biotechnology in Agriculture andForestry, Vol. 10, Legumes and Oilseed Crops I, ed.Y. P. S. Bajaj, pp. 38-48. Springer-Verlag, Berlin.

Owens, L. D. & Cress, D. E. (1985). Genotypicvariability of soybean response to Agrobacteriumstrains harboring the Ti or Ri plasmids. PlantPhysiology, 11, 87-94.

Parrott, W. A., Hoffinan, L. M., Hildebrand, D. F.,Williams, E. G. & Collins, G. B. (1989). Recoveryof primary transformants of soybean. Plant CellReports, 7, 615-617.

Penza, R., Lurquin P. F. & Fillippone, E. (1991).Gene transfer by cocultivation of mature embryoswith Agrobacterium tumefaciens: applications tocowpea (Vigna unguiculata Walp). Journal of PlantPhysiology, 138, 39-43.

Petit, A., David, C , Dahl, G., Ellis, J., Guyon, P.,Casse Delbart, F. & Tempe, J. (1983). Furtherextension of the opine concept: plasmids inAgrobacterium rhizogenes cooperate for opinedegradation. Molecular and General Genetics, 190,204-214.

Petit, A., Stougaard, J., Kiihle, A., Marcker, K. A. &Tempe, J. (1987). Transformation and regenerationof the legume Lotus corniculatus'. a system formolecular studies of symbiotic nitrogen fixation.Molecular and General Genetics, 207, 245-250.

Pickardt, T., Huancaruna Perales, E. & Schieder, O.(1989). Plant regeneration via somaticembryogenesis in Vicia narbonensis. Protoplasma, 149,5-10.

Pickardt, T., Meixner, M., Schade, V. & Schieder, O.(1991). Transformation of Vicia narbonensis viaAgrobacterium-mediated gene transfer. Plant CellReports, 9, 535-538.

Pickardt, T., Saalbach, I., Machemehl, F., Saalbach,G. Schieder, O. & Muntz, K. (1992). Expression ofthe sulfur-rich Brazil nut 2S albumin in transgenicVicia narbonensis plants. In Vitro CellularDevelopmental Biology, Suppl. 28, 92A.

Puonti-Kaerlas, J. & Eriksson, T. (1988). Improvedprotoplast culture and regeneration of shoots in pea(Pisum sativum L.). Plant Cell Reports, 1, 242-245.

Puonti-Kaerlas, J., Eriksson, T. & Engstrom, P.(1990). Production of transgenic pea (Pisum sativumL.) plants by Agrobacterium tumefaciens-mediated genetransfer. Theoretical and Applied Genetics, 80, 246-252.

Puonti-Kaerlas, J., Eriksson, T. & Engstrom, P.(1992a). Inheritance of a bacterial hygromycinphosphotransferase gene in the progeny of primarytransgenic pea plants. Theoretical and AppliedGenetics, 84, 443-450.

Puonti-Kaerlas, J., Ottosson, A. & Eriksson, T.(1992b). Survival and growth of pea protoplastsafter transformation by electroporation. Plant Cell,Tissue and Organ Culture, 30, 141-148.

Puonti-Kaerlas, J., Stabel, P. & Eriksson, T. (1989).

124 WIDHOLM

Transformation of pea {Pisum sativum L.) byAgrobacterium tumefaciens. Plant Cell Reports, 8,321-324.

Ramsey, G. & Kumar, A. (1990). Transformation ofViciafaba cotyledon and stem tissues byAgrobacterium rhizogenes: infectivity and cytologicalstudies. Journal of Experimental Botany, 41,841-847.

Rech, E. L., Golds, T. J., Hammatt, N., Mulligan,B. J. & Davey, M. R. (1988). Agrobacteriumrhizogenes mediated transformation of the wildsoybeans Glycine canescens and G. clandestina:production of transgenic plants of G. canescens.Journal of Experimental Botany, 39, 1225-1285.

Rech, E. L., Golds, T. J., Husnain, T., Vainstein,M. H., Jones, B., Hammatt, N., Mulligan, B. J. &Davey, M. R. (1989). Expression of a chimaerickanamycin resistance gene introduced into the wildsoybean Glycine canescens using a cointegrate Riplasmid vector. Plant Cell Reports, 8, 33-36.

Russell, D. R., Wallace, K. M., Bathe, J. H., Martinell,B. J. & McCabe, D. E. (1993). Stabletransformation oiPhaseolus vulgaris via electric-discharge mediated particle acceleration. Plant CellReports, 12, 165-169.

Savka, M. A., Ravillion, B., Noel, G. R. & Farrand,S. K. (1990). Induction of hairy roots on cultivatedsoybean genotypes and their use to propagate thesoybean cyst nematode. Phytopathology, 80,503-508.

Schiemann, J. & Eisenreich, G. (1989).Transformation of field bean {Viciafaba L.) cells:expression of a chimaeric gene in cultured hairyroots and root-derived callus. Biochemie undPhysiologie der Pflanzen, 185, 135-140.

Schroeder, H. E., Khan, M. R. I., Knibb, W. R.,Spencer, D. & Higgins, T. J. V. (1991). Expressionof a chicken ovalbumin gene in three lucernecultivars. Australian Journal of Plant Physiology, 18,495-505.

Senaratna, T., MacKensie, B. D., Kasha, K. J. &Procunier, J. D. (1991). Direct DNA uptake duringthe imbibition of dry cells. Plant Science, 79,223-228.

Shahin, E. A., Spielmann, A., Sukhapinda, K.,Simpson, R. B. & Yashar, M. (1986).Transformation of cultivated alfalfa using disarmedAgrobacterium tumefaciens. Crop Science, 26,1235-1239.

Shen, W. H., Petit, A., Guern, J. & Tempe, J. (1988).Hairy roots are more sensitive to auxin than normalroots. Proceedings of the National Academy of Sciences,USA, 85, 3417-3421.

Spano, L., Mariotti, D., Pezzotti, M., Damiani, F. &Arcioni, S. (1987). Hairy root transformation inalfalfa (Medicago sativa L.). Theoretical and AppliedGenetics, 73, 523-530.

Stephens, P. A., Nickell, C. D. & Widholm, J. M.

(1991). Agronomic evaluation of tissue-cultured-derived soybean plants. Theoretical and AppliedGenetics, 82, 633-635.

Sukhapinda, K., Spivey R. & Shahin, E. A. (1987). Ri-plasmid as a helper for introducing vector DNA intoalfalfa plants. Plant Molecular Biology, 8, 209-216.

Tabaeizadeh, Z. (1989). Genetic transformation of apasture legume, Lotus corniculatus, L. (Birdsfoottrefoil). Biotechnology Letters, 11, 411-416.

Tepfer, D. (1990). Genetic transformation usingAgrobacterium rhizogenes. Physiologia Plantarum, 79,140-146.

Vieira, M. L. C , Jones, B., Cocking, E. C. & Davey,M. R. (1990). Plant regeneration from protoplastsisolated from seedling cotyledons of Stylosanthesguianensis, S. macrocephala and 5. scabra. Plant CellReports, 9, 289-292.

Warkentin, T. D. & McHughen, A. (1991). Crowngall transformation of lentil {Lens culinaris Medik.)with virulent strains of Agrobacterium tumefaciens.Plant Cell Reports, 10, 489-493.

Webb, K. J., Jones, S., Robbins M. P. & Minchin,F. R. (1990). Characterization of transgenic rootcultures of Trifolium repens, Trifolium pratense andLotus corniculatus and transgenic plants of Lotuscorniculatus. Plant Science, 70, 243-254.

Webb, K. J., Woodcock, S. & Chamberlain, D. A.(1987). Plant regeneration from protoplasts ofTrifolium repens and Lotus corniculatus. Plant Breeding,98, 111-118.

Wei, Z. & Xu, Z. (1988). Plant regeneration fromprotoplasts of soybean {Glycine max L.). Plant CellReports, 7,348-351.

White, D. W. R. & Greenwood, D. (1987).Transformation of the forage legume Trifolium repensL. using binary Agrobacterium vectors. PlantMolecular Biology, 8, 461-469.

Wier, A. T., Thro, A. J., Flores, H. E. & Janyes, J. M.(1988). Transformation of Lotononis bainesii Bakerwith a synthetic protein gene using the leaf disktransformation-regeneration method. Phyton, 48,123-131.

Williams, D. J. & McHughen, A. (1986). Plantregeneration of the legume Lens culinaris Medik.(lentil) in vitro. Plant Cell, Tissue and Organ Culture,7, 149-153.

Yang, N. S. & Christou P. (1990). Cell type specificexpression of a CaMV 35S-gus gene in transgenicsoybean plants. Developmental Genetics, 11, 289-293.

Yu, J. P. & Shao, Q. Q. (1991). Transformation ofLotus corniculatus L. mediated by Agrobacteriumtumefaciens. Science in China B, 34, 932-937.

Zhou, J. H. & Atherly, A. G. (1990). In situ detectionof transposition of the maize controlling element{Ac) in transgenic soybean tissues. Plant Cell Reports,8, 542-545.

9Spring and Winter Rapeseed Varieties

Philippe Guerche and Catherine Primard

Introduction

Rapeseed is one of the most important oilseedcrops after soybeans and cottonseed, representing10% of world oilseed production in 1990(McCalla & Carter, 1991). It has been domesti-cated for more than 3000 years, along with theother Brassica species having noteworthy eco-nomic value, in Asia and in Europe (where rape-seed was first mentioned as a crop in thethirteenth century). It has been introduced intoCanada only very recently (1942). The produc-tion of oilseeds, meal and oil has been increasingcontinuously for the last 30 years for food andfeed grains, mainly by expansion of the area undercultivation (Robbelen, 1991). By the year 2000,China should be the leading producer with 9.2 Mt(26%), followed by India with 7.8 Mt (22%),European Community (12 countries), 7.6 Mt(21%), Canada 3.8 Mt (11%) and eastern Europe2.6 Mt (7%) (Carr & MacDonald, 1991).

Oilseed rape is a cruciferous species, belongingto the genus Brassica (Tribe Brassiceae, FamilyBrassicaceae), resulting from the natural hybridi-zation between a cabbage (B. oleracea L., CC,2n= 18, Western Europe and Northwest Africa)and a turnip rape (B. campestris L., AA, 2n = 20,Europe and Asia). Rape is an amphidiploid (B.napus, AACC, 2n = 38), whose center of diversityis the intersection of its parental areas. Artificialcrosses (Chevre et ai, 1991) or somatic fusionshave been used to introgress genes from relatedspecies (for a review, see Trail, Richards & Wu,1989). The development of a detailed geneticlinkage map (isozymes, restriction fragmentlength polymorphism (RFLP), random amplified

polymorphic DNA (RAPD)) is of interest foranalysis of genetic relationships between species.

Brassica napus is a herbaceous plant that canreach 2 m in height. The flowers and pods areproduced in a mass concentrated near the top ofthe plant, ensuring a high harvest efficiency usingmodern management and harvest practices. As aresult of the long history of breeding effortsfocused on this crop, rapeseed is the only poly-ploid Brassica offering spring as well as winter cul-tivars. In western Europe, winter rapeseedcompletes its cycle in 250-310 days and springrapeseed in only 120-150 days, the former havingthe higher yield (around 4 tons/ha) (Renard et al>1992). Rapeseed is a semi-autogamous species.Pollination is by wind and insects (bees), drawnby the nectar of its bright yellow flowers. Out-crossing is estimated to be 10%-30% in theabsence of incompatibility genes. Thus, breederscould aim to create a number of varietal types,such as populations or pure lines, but syntheticvarieties or hybrids (provided an allogamizationsystem is available) are better able to exploit het-erosis.

Oil and proteins are the important rapeseedproducts, accounting for 40%-46% and 20%-28%,respectively, of the rapeseed dry matter, followedquantitatively by carbohydrates (20%), of whichstarch accounts for only 2%-3%.

Rapeseed edible oil is of high quality with a highunsaturated fatty acid content (mainly oleic acid(60%) and linoleic acid (22%)). The oil also hassome qualities for nonfood industrial uses (com-bustible, lubricating oil, paint). Rapeseed proteinshave a well-balanced amino acid compositionresulting in possibilities for products of high nutri-

125

126 GUERCHE AND PRIMARD

tive value if other antinutritional compounds pre-sent in the seeds can be avoided.

Considerable effort has been made to improverapeseed agronomic qualities by selective breedingtechniques. In this way the so-called 'double zero'cultivars were developed containing no erucic acidin the oil and a very low content of glucosinolatecompounds in the meal. These types of cultivarare named canola in Canada. The future goals forimproving this crop include increased yield anddisease resistance, in addition to altering the oilcontent and composition. The protein content ofthe meal is negatively correlated with oil content.Conventional breeding coupled with emergingbiotechnologies will continue to play a major rolein this improvement.

The main objectives may be summarised as fol-lows. To increase seed oil content without increas-ing erucic acid and to lower a-linolenic acid(responsible for 'room odour' at high temperature)from 9% to 5-7%. For industry, specific varietieswith high erucic acid content (66%), linoleic acid(30%), oleic acid (80%), or a reduction of a-linolenic acid to 0% are desired. Biochemical andmolecular analyses of biosynthesis of fatty acidsfrom rapeseed or other species (Arabidopsisthaliana, Borago ojficinalis, Cuphea ellipsis) aredeveloping with the aim of cloning genes impli-cated in fatty acid desaruration, chain elongationand transacylation reactions. The eventual goal isto modify the composition of fatty acids by genetransfer. Other important objectives are to lowerthe content of glucosinolate compounds, cholineesters and cellulose. Modification of the proteincomposition to enhance the balance of essentialamino acids (methionine, lysine and tryptophan) isyet another goal. All of these objectives may beachieved by gene transfer (Vandekerckhove et at.,1989; De Clercq et al, 1990).

More productive and better-adapted rapeseedvarieties with improved tolerance to biotic or abi-otic stress can be developed by breeders lookingfor earliness, drought and cold tolerance, lodgingresistance, apetalous flowers, and insect andherbicide resistance. Some objectives may beachieved through gene transfer:

1. The introduction of a cloned foreign gene intothe rapeseed genome in order to add a new char-acter, such as a gene supplying resistance to aherbicide, a pest or a type of stress or to modifythe metabolite composition of an organ, such asfatty acid or protein components.

2. The inactivation of a gene by the insertion of amarker gene in order to find mutants.

3. The study of the expression of a given gene in ahomologous or heterologous system, in order tostudy promoters and regulation of gene expres-sion.

Transformation techniques

A number of gene transfer methods have beenused to mediate delivery of foreign genes intorapeseed. Agrobacterium-medizted transformationhas been most commonly used but some directgene transfer techniques have also been tried withsuccess. Table 9.1 summarizes the reports whereefforts were made to improve the efficiency oftransformation and where transgenic rapeseedplants were obtained. Most of these experimentswere done with genes that allowed positive selec-tion of the transformed tissue (hairy root pheno-type induced by expression of the Agrobacteriumrhizogenes transferred DNA (T-DNA) genes orantibiotic or herbicide resistance genes). The fol-lowing conclusions may be drawn from the refer-ences listed in Table 9.1.

Transformation by Agrobacterium

Transformation by Agrobacterium rhizogenesWinter and spring varieties of rapeseed are verysusceptible to infection by A. rhizogenes and mostorgans when inoculated with this bacteria give riseto hairy root proliferation with a very high effi-ciency (50%-100% of inoculated organ seg-ments). These roots have a very high growth rateon hormone-free medium. This allows the roots torapidly outgrow the bacteria on a medium con-taining standard antibiotics and thus it is easy todecontaminate the transformed tissue.

The first transgenic rapeseed plant wasobtained by Ooms et al> in 1985 after regenera-tion of a clonal hairy root explant induced by A.rhizogenes infection. This plant was unfortunatelysterile and showed the classical phenotype ofmany plants from other species regenerated fromhairy root (mainly wrinkled leaves, short inter-nodes and abundant secondary root system)(Tepfer, 1984). But in later experiments, fertileplants were regenerated, although exhibiting thehairy root phenotype and with varying degreesof reduced fertility. They produced enoughviable pollen to fertilize an emasculated wild-type

Spring and winter rapeseed varieties 127

rapeseed; in this way, backcrossed progeny couldbe obtained.

The 'hairy root' phenotype prevents direct useof these plants in breeding programs, but someresearchers have taken advantage of this easy pro-cedure to produce transgenic plants. Two strate-gies may be used: cotransformation experiments(with a wild-type (WT) A. rhizogenes plus a dis-armed A. tumefaciens carrying a plasmid with aplant-selectable marker gene and/or a gene ofinterest) or binary vector experiments (with an A.rhizogenes carrying two plasmids, a wild type Riplasmid plus a disarmed T-DNA plasmid, carry-ing the genes of interest). Cotransformed regener-ated plants bear both Ri TL-TR DNA and thedesired construct. Segregation analysis of thecotransferred A. rhizogenes T-DNA and thegene(s) carried by the disarmed T-DNA is per-formed by backcrossing progeny to the wild-typeparent. When the two markers are integratedindependently or at a distance that allows recom-binations, the BC1 plants, which carry only thedisarmed T-DNA (with the desired gene), can beselected in the progeny.

Generally, plant tissues used for transformationare segments of axis tissues, such as hypocotyl,vegetative stem or floral axis. These fragments areprepared and cultivated in vitro. In the case of flo-ral or even vegetative axis tissues, the segments areset up in gelified medium upside down to avoidnatural rooting. Inoculation is performed withfreshly subcultured bacteria, either in liquid or onsolid medium. In case of cotransformation, A. rhi-zogenes and A. tumefaciens are mixed just beforeinoculation generally in a 1:1 ratio. A smallamount of bacteria is laid on the upper woundedsurface of the segments. The delay before hairyroot development is varied from 10 days to 3weeks. Young roots are harvested and indepen-dently cultured on a decontamination medium.When the plasmid used allows it, cotransformedroot clones can be immediately screened on theproper selective medium. The regeneration effi-ciency of the transformed roots has beenimproved by preconditioning the roots beforeregeneration in a 2,4-dichlorophenoxyacetic acid(2,4-D)-containing medium (Guerche et al,1987a). All tested cultivars are able to regenerateplants, but with varying regeneration frequency,depending on whether the cultivar used was aspring (very efficient) or winter (less efficient)variety. Each of the resulting root clones may then

be regenerated into several plants. Cotrans-formation rate (percentage of cotransformed hairyroot clones/total number of hairy root clones)seems to be higher when the binary vector strategyis used (80%-90% versus 15%-20%) as was firstshown by Hamill et al (1987).

It is noteworthy that not only hairy roots butshoots can be obtained after A. rhizogenes inocula-tion. Damgaard & Rasmussen (1991) observedthat when decapitated hypocotyls with an intactroot system were inoculated by A. rhizogenes car-rying either a binary vector (WT Ri plasmid and adisarmed T-DNA carrying a kanamycin resistancegene) or a cis vector (only the WT Ri T-DNAplasmid in which the kanamycin resistance genewas introduced), up to 58% of these hypocotylsdeveloped shoots directly at the point of infectionand that 4%-15% of these shoots show a hairyroot phenotype and NPT II activity. It will beinteresting to discover whether it is possible toobtain directly, by a binary vector strategy, trans-formed shoots containing only the kanamycinresistance gene, without the WT Ri T-DNA ashas been observed with tomato (Shahin et al,1986). It is still too early to know whether thecotransformation with two bacteria leads moreoften to the integration of the two exogenousDNAs at independent loci than the binary vectortransformation strategy.

According to the different studies and the vari-ous transformation strategies used to obtain trans-genic rapeseed, the copy number of the insertedgene can vary between one and ten copies. Often,if several copies are present, they behave as aunique locus, due to tandem or inverted integra-tions. The copy number of the TL-DNA of A. rhi-zogenes is higher than that of the TR-DNA in thefew plants studied as the TL part may be inte-grated alone but the TR part is always associatedwith the TL.

Polymerase chain reaction (PCR) techniquesallow rapid assays on a large number of progeny tocounter-select plants with integrated Ri T-DNAand to select those bearing the gene of interest.The ease of the procedure and the high frequencyof transformation allow the possible use of thisstrategy for the transfer of genes without anyselective means.

Transformation by Agrobacterium tumefaciensBecause of the unfavorable phenotype induced bythe Ri T-DNA, transformation procedures with a

Table 9.1. Synopsis of the major transformation experiments carried out on rapeseed

Variety

Jet 9 (W)

Brutor(S)

HM 81 (S)

Bienvenue (W)Brutor (S)

Line (S)1046 (S)

Samourai (W)

Samourai (W)

Samourai (W)

Westar (S)

Westar (S)

Westar (S)

Westar (S)

Westar (S)R8494(W)

Westar (S)

Topaz (S)

Topaz (S)

Westar (S)Cobra (S)

Westar (S)Westar (S)Drakkar (S)R8494

Brutor (S)

Wintervarieties

Spring varieties

Brutor (S)

At.

Dis nop

Dis nop

Dis nopDis oct

Dis nop

Dis nop

WT nopDis nopDis nop

Dis nop

Dis nop

Dis oct

Dis octDis nop

Dis octDis nop

As. DGT

WT

WT

WT

WT

WT

WT

WT

WT

WT

EP

MIJ

PEG

EP

Organ

Stem

Stem

Stem

Cotyledonarynode

Decapitatedhypocotyl

Floral axis

Floral axis

Floral axis

Stem disc

Floral stemepiderm

Floral stemepiderm

Hypocotyl

Hypocotyl

Cotyledonarypetiole

Microspore-derived

embryo

Microspore-derivedembryo

Floral stem

StemHypocotyl

Leafprotoplasts

Microspore-derivedembryos

Leafprotoplasts

Leafprotoplasts

Promoter

NOS

NOS

19S70ALS

ALS

NOS35S

35S

NOSNOS35S

35S35S35S

35S

NOS

NOS35SPTR

19S

35S,SV40

35S

19SNOS35S

Selection

Hairy-root

Hairy-root

Hairy-root

Hairy-root-Kan50

Kan (noselection onKan-medium)

CS100Kan 60CS100

CS100

Kan 100

Mtx 0.01

Kan 15CAM7CAM7

Kan 25Kan 25Kan 50PPT20

Kan 15

Kan 50Hyg 10

Kan 50

Kan 150GUS

Kan 100PPT20Kan 50

Paro 20Kan 70

Kan

Hyg 30

Par 20Hyg 25PPT5

Putativetransgenictissue ratea

50% root

70% root

19%-58%shoots

30% root

30% root

100% root

7% shoot

32% shoot

6.6% shoot5% shoot5% shoot

1.5% shoot3.2% shoot40% (S,W)30% (S,W)

shoot

55% shoot

14.7%29% embryo

0.6% embryo

30% shoot

2% shoot34% (S)18% (W) shoot

5.2 X10"6

colony

28% (35S)49% (SV40)

embryo

1-4X10"3

colony

2.6x10~4col2X10 "col3 .8X10" col

Number of Ftransgenicplants

1p -

173p +

1p

17p(90% +co-transformed)

9%-30% op +4%-15%

NPTII +

3p(15% +co-transf root)

3p (21 % +co-transf root)

13p(80% +co-transf root)

200p +

24p(10%) +

12p(2%) +10p (1.6%) +5p (2.5%) +

3p (0.7%) +9p(1.9%)748p (S,W) +550p (S,W)

22p +

40% NPTII +25% NPTII +(of resistantembryos)

3p (on 600 +cocultivatedembryos)

13p(7%ofthe +shoots)

6p +228p +

2p +

+

8p

30p

Copynumber

>5 (1p)

1-4 (5p)

2-3 dp)

5-10 dp)

2-3 (19p)

2-3 (5p)

1-2 (6p)

1-3 (12p)

1-8 (26p)

1-4 (6p)

5-10 dp)

s*1 <3p)2-5 (18p)

1 (2p)

>2 (8p)

1,3*2(12col)

Remarks

Sterility of theregenerated plant. Hairy

High efficiency of rootregeneration (up to90%). Hairy root

pnenoiypeHairy root phenotype

Binary vector. Highfrequencyco-transformation.Possible segregation ofthe two markers in the

progenyCis and binary vectors.Direct regeneration ofshoots at the infectionpoint. Selection with

NPTII and opine assaysBinary vectors.

Co-transformation.Attempts at plantregeneration wereperformed on only alimited number of rootclones. Possiblesegregation of the twomarkers in the progeny

Cis and binary vectors.More than 50%escapes on Kan 50 and10% on Kan 100. Lowefficiency of octopinestrain

Cis vector. Unsuccessfulselection on Kan.One-third escapes

Cis and binary vectors.Low efficiency selectionon Kan and CAM.One-third escapes

Binary vectors

Binary vector. 25% of thetransgenic plantsshowed no detectablePAT or NPTn activity

Binary vector. Fewescapes. 55% is theresult of 1 experimentwith 40 infected

expiantsBinary vector. 1000

plantlets regenerated.NPTII test on 75. Morethan 60% escapes.Doubling needed. Riskof chimeras

Binary vector. Lowkanamycin resistancelevel. Doubling needed.Risk of chimeras

Binary vectors. Highfrequency of escapes.Nop strain of A.t.allowed transformationof Cobra (W)

Cis vectors. No escapesBinary vectors.Cotransformation with 2different strains carrying2 different markergenes. 57% of thecotransformed plantshave the 2 copies linked

Paromomycine allowedselection of colonies athigher density than Kan

Necessity of a secondround of embryogenesisto avoid chimeras, notissues selection on Kan

Transformationexperiment using totalDNA of a Hyg resistantB. nigra line

Improvement ofelectroporationefficiency by GUStransient expressionassays

References

Ooms etal.(1985)

Guerche etal.(1987a)

Hrouda etal.(1988)

Boulter etai.(1990)

Damgaard &Rasmussen(1991)

Primard etal.(unpublisheddata)

Fry et al.(1987)

Pua ef al.(1987)

Charest ef al.(1988)

Radke ef al.(1988)

De Block &Debrouwer(1989)

Moloney et al.(1989)

Pechan (1989)

Swanson etal.(1989)

Boulter ef al.(1990)

Misra(1990)De Block &

Debrouwer(1991)

Guerche etal.d987i>)

Neuhaus etal.(1987)

Golz ef al.(1990)

Rouan &Guerche(1991)

Notes:Abbreviations: (S), Spring; (W), Winter; At, Agrobacterium tumefaciens; A.r., A. rhizogenes; WT, wild-type; Dis nop, disarmed T-DNA (nopaline strain); Dis oct, disarmed T-DNA (octopinestrain); DGT, direct gene transfer; EP, electroporation; MU, microinjection; PEG, polyethylene glycol treatment; NOS, nopaline synthase promoter; 19S, CaMV 19S RNA promotor; 35S,CaMV 35S RNA promotor; 70, CaMV 35S RNA promotor with duplicated enhancer; PTR, 1' and 2' gene promotors from A. tumefaciens T-DNA; SV40, simian virus 40 promotor; ALS,promotor from the mutated acetolactate synthase gene from Arabidopsis thaliana; Kan 15, kanamycin (15 mg/l); Hyg, hygromycine; GUS, p-glucuronidase; PPT, phosphinothricin; Paro,paromomycin; CS, chlorsulfuron; Mtx, methotrexate; CAM, chloramphenicol; F, fertility of the plant; p, plant; col, colony.s Putative transgenic tissue rate is the fraction of presumed transgenic shoots or roots divided by the total number of treated expiants X 100 for Agrobacterium transformation and the %ratio of the number of resistant colonies tn the total number of colonies submitted to selection for EP and PEG experiments."Copy number: the minimal and maximal copy number found in the plants studied (in parens).


disarmed A. tumefaciens that have proven highlysuccessful in tobacco have been applied by severalgroups to rapeseed. For large-scale transformationprograms it is necessary to have an efficientmethod of transformation. One of the main pre-requisites for this is a high level of regenerationfrom the chosen explant. Leaf disc, the tissue themost commonly used for tobacco transformation,does not seem to have the highest regenerationpotential in rapeseed. Stem sections, thin stemlayers, cotyledon petioles, hypocotyls ormicrospore-derived embryos appear to have moreregeneration potential. The case of microspores oryoung-derived embryos is of interest because ofthe high potential number of embryos obtainedfrom microspore culture of B. napus, and becausethey are haploid, giving the opportunity to pro-duce (after chromosome doubling (spontaneousor with colchicine)) directly homozygous trans-genic plants. The risk of chimeras obtained whentransformations take place in an already pluri-cellular embryo may be avoided by the ability ofB. napus microspore-derived embryos to undergoan embryogenic process. These embryogenicstrains give rise to secondary embryos issued fromone cell, so cloning a transformation event.However, to date, diploid tissues have been morecommonly used.

When different factors influencing the shootregeneration of cotyledon petioles from variousBrassica species and cultivars of B. napus are com-pared, substantial differences in the regenerationrate between genotypes are displayed (Dale &Ball, 1991). A precise study of the regenerationability of sections and thin epidermal layers fromthe floral axis points to the importance of theauxin content in the culture medium. On auxin-containing medium, two winter genotypes (cvs.Bienvenu and Darmor) show very few regeneratedshoots from thin layers, whereas the spring-typecv. Brutor shows a high regeneration rate (>50%).Conversely, axis sections from all three genotypesproduce shoots on a medium without auxin.Moreover the winter cv. Darmor showed the high-est regeneration rate (>90%). According to thesource of the explant and the composition of plantgrowth regulators in the medium, the cultivarstested may have regeneration potential indepen-dent of their spring or winter type (Julliard et ah,1992). Some reports in Table 9.1 underline thegreater difficulty in transforming the winter culti-vars, at least those presently tested. Experiments

were done mainly with the spring variety Westarwith variable success, but are generally low com-pared to tobacco (Horsch et al, 1983).

The other prerequisite is the ability of A. tume-faciens to transform the rapeseed tissues. Two fac-tors are combined here: the virulence of thebacteria strain towards B. napus, and the cellaccessibility to the bacterium. Both octopine andnopaline strains have been used to obtain trans-genic rapeseed plants, but nopaline strains havebeen more widely used, partly because undis-armed nopaline strains show greater virulence,even for winter varieties (Holbrook & Miki,1985). However, Charest et al (1989) showedthat adding acetosyringone to the bacteria culturemedium could enhance the virulence of octopinestrains towards B. napus. As regards accessibility,in published protocols, the bacterium will haveaccess mainly to the surface cells. Cytologicalstudies show that if shoots originate from cells inthe inner cell layers of the treated tissue, the fre-quency of transformation is much lower than ifthey are on or near the surface (Moloney et al.>1989).

The conditions of the plant culture, the type oforgan and age of the plant are chosen primarilywith regard to the regeneration capacity, andexplants are prepared to be cultured in vitro. Theagrobacteria are prepared so as to be as virulent aspossible: culture conditions and subculturerhythm should provide cells in their fast growthphase and, finally, the density of the last resuspen-sion of the bacteria in the inoculation medium isimportant. For inoculation, explants are generallydipped entirely into the bacterial suspension.Time of inoculation is variable (from a few min-utes to several days) according to the organ type,as too long a delay may affect the survival of thetissues or their regeneration ability (Radke et al,1988).

The duration of the coculture (fragments areblotted dry and transferred on to a solid mediumwith or without filter paper) is also to be assessed.Little is known concerning the T-DNA integra-tion event in the cell genome except that dividingcells are said to be the best targets. Thus, thecoculture time may be a crucial phase with regardto the physiological state of the tissue being trans-formed.

Most of the authors noticed that cocultivationand further decontamination of the tissues beforeor during regeneration have dramatic effects on


the plant regeneration rate, lowering by at leastone order of magnitude its efficiency. There is noprecise study of the effect on this point of theAgrobacterium strain used, but in vitro culture con-ditions and the antibiotic used to eliminate theAgrobacterium have been more thoroughly investi-gated. For the former, two main areas were inves-tigated in order to reduce hypersensitivity tobacteria-induced damage: preconditioning of thetissue before inoculation, and complementing thecoculture medium with a feeder layer of healthyfast growing cells (references cited in Table 9.1).Carbenicillin seems to be the most commonlyused for tissue decontamination, but this antibi-otic also has an auxin-like effect that has to betaken into account in the medium composition fortissue regeneration.

The other limiting step is the selection of trans-genic shoots. Again three main factors are to beconsidered: the construct characteristics, the tim-ing and severity of the selection of transformedcells. First, the promoter strength controlling theexpression level of the resistance of the trans-formed cells surrounded by dying tissues is animportant factor. Different promotors and markercoding sequences (all of them widely used fortobacco transformation (for a review, see Klee &Rogers, 1989)) have been used with success toselect transgenic rapeseed plants. However, thestudy of the relative strengths of some promoters(cauliflower mosaic virus 35S, 2' and nopalinesynthase (nos)) demonstrated that the level of geneexpression in transformed calli may be less in rapecalli than in tobacco calli, although the ranking ofthese promoters is conserved (35S constructs givethe the highest level, nos the lowest) (Harpster etah, 1988). Kanamycin has been used as selectiveagent in many systems. For rapeseed transforma-tion, except for one report (Pua et al., 1987),kanamycin is the most efficient selective agent.Hygromycin also works well.

Most publications show that the number ofputative transgenic shoots (or embryos) increaseswhen the selective agent concentration is loweredin the medium but in return the number ofescapes increases dramatically when the concen-tration approaches the minimum lethal dose foruntransformed tissue. The same problem isobserved if application of selection is delayed fortoo long after cocultivation.

At the last in vitro culture step, some authorsnoticed that the transgenic shoots were often vitri-

fied or difficult to root (Fry, Barnason & Horsch,1987; De Block, Debrouwer & Terming, 1989).Ethylene production, which inhibits shoot forma-tion and causes senescence of the tissue, can belimited by the addition of AgNO3 and good venti-lation of the in vitro cultivated tissues. Vitrificationcan be avoided by decreasing the water potentialof the medium, reducing the relative humidity inthe culture vessels and lowering the cytokinin con-centration. Rooting can be improved by adding alow concentration of auxin in a well-aeratedgelling agent medium (using perlite for example).

Lastly, it seems that the variability in the differ-ent transformation experiments is extremely highand that the repeatability of the same transforma-tion procedure between different laboratories isextremely low, so few general rules can be drawnfrom these investigations for the moment.

The copy number of the transgene(s) in the fewtransgenic plants that have been analyzed at theDNA level varies from one to more than tencopies of the foreign gene. Varying degrees ofcomplexity in their integration patterns can beobserved but it seems that multiple copies of theT-DNA integrate mainly as inverted or tandemrepeats. These integrations behave as singlegenetic loci rather than independent copies. Insome studies, it seems that up to 25% of the trans-genic rapeseed shoots that were obtained onmedium containing a selective agent gave rise toplants in which the transgene was present but itsexpression was not detectable (De Block et al>1989).

Direct gene transfer

Few transgenic plants have been obtained bydirect gene transfer experiments. One paperreports the microinjection of eight-cell micro-spore-derived embryoids with a very high effi-ciency (around 40%), but no transgenic plantswere shown (Neuhaus et aL, 1987).

To date, only direct gene transfer of protoplastshas allowed the recovery of transgenic rapeseedplants. Generally, mesophyll or hypocotyl proto-plasts derived from in wro-cultured plants weresubjected to electroporation or polyethylene gly-col (PEG) treatment. Experimental conditionsused for these two techniques are approximatelythe same as those developed for tobacco proto-plast transformation.

The frequency of transformation is always low

Table 9.2. Synopsis of the major transformation experiments with agronomic applications carried out on rapeseed

Variety Trans Promotor andselective agent

Gene of interest Expression

RNA Protein Results Reference

Westar

Westar

Westar

WestarDrakkar

Brutor

Drakkar

WestarProfit

Westar

A t

A t

A t

A t

EP

A t

A t

A t

35S Kan

NOS Kan

PTR Kan Hyg

PTR Kan Hyg

NOS Kan

35S PPT

35S CS, Kan

35S Kan

2S albumin gene (napin)from Brassica napus

35S pro-Metallothionein I3'NOS

Enkephalin substitution inArabidopsis 2S albumingene

Arabidopsis/BrazW nut 2Sgene

Methionine richsubstitution in 2SArabidopsis gene

Pro Iectin-2S Brazil nutgene

Anther specific promotorTA29-barnase codingsequence

Mutated acetolactatesynthase gene from A.thaliana

Maize oleosin codingsequence with napinpromotor

+ Seed specific expression

+ Constitutive expression. Thetransgenic plants are 10xmore CdCI? tolerant than WT

+ + Seed specific expression.10-50 nmol enkephalin/g ofseed

+ + Seed specific expression0.1 % of salt soluble seed

+ + protein1-2% of salt soluble protein

+ + Lectin-like expression.0.02-0.06% of salt solubleseed protein

+ Anther specific expression.Male sterility of transgenicplants

+ + Transgenic plants are 30xmore resistant to CS thancontrols

+ + Seed specific expression. 1 %of total seed proteins. Welltargeted in protein bodies

Radke etal. (1988)

Misra(1989)

Vandekerckhoveetal. (1990)

De Clercq et al.(1990)

Guerche etal.(1990)

Mariani etal.(1990)

Miki etal. (1990)

Lee etal. (1991)

Notes:Abbreviations: Trans, transformation technique; 35S, CaMV 35S RNA promotor; NOS,tumefaciens T-DNA; Kan, kanamycin; CS, chlorsulfuron; PPT, phosphinothricin.

nopaline synthase promotor; PTR, 1' and 2' gene promotors from A.


in comparison with that obtained from tobaccoprotoplasts (l%-5% by both methods (Shillito etah, 1985)) mainly because the plating efficiency islower for rapeseed protoplasts (30%) than fortobacco protoplasts (90%-100%) and because ofthe difficulties of selecting the transgenic colonies(Rouan & Guerche, 1991). Indeed, low densitycultivation of rapeseed protoplast-derived micro-colonies is difficult and, for this reason, theauthors use agarose-embedded culture (Shillito etah, 1983) to select and regenerate the few candi-date colonies into plants. The selection of thetransformed colonies on phosphinothricin-con-taining medium seems to give better results thanselection on kanamycin or hygromycin becauseresistant colonies grow faster, allowing a shorterselection time and a higher regeneration efficiencyof the colonies. Here, too, the efficiency of trans-formation is largely dependent on the genotypeand the protoplast culture conditions.

The electroporation efficiency can be improvedby the use of GUS transient expression assays(Jefferson et ah, 1987) to optimize the parametersand Chapel & Glimelius (1990) noticed that acombination of PEG and electroporation treat-ment led to a better transformation efficiency.

All these experiments were performed on springvarieties for which protoplast culture is better con-trolled than for winter varieties.

Conclusion

Rapeseed is'one of the species for which all the tis-sue culture techniques required for transforma-tion experiments exist but all the necessary stepsto achieve transformation are less efficient than formodel transformation species (tobacco, etc.).Except for experiments by De Block et ah (1989;De Block & Debrouwer, 1991) transformationefficiency is reduced by one or two orders of mag-nitude in comparison with tobacco (Horsch et ah,1983). It seems that the constructs with selectablemarker genes having strong promotors (35S, 70)are recommended for lowering the escape / trans-genic plants ratio. Considering all the literature, itis always possible to select transformants with anyselective agent, except with chloramphenicol, pro-vided selective conditions are well established(Table 9.1).

By using one of these different techniques, it isalso always possible, with time and effort, toobtain transgenic rapeseed plants in laboratories

where some experience with rape tissue cultureexists. To ensure reliable transformation rates,several modifications of already published proto-cols have been necessary, either to adapt them tonew cultivars or simply to experimental conditionsin each laboratory. It is clear that if thesedescribed techniques are to be used, much workneeds to be done to improve regeneration of tissuefrom cultivated varieties (mainly winter types). Itwill also be necessary to find tricks to decontami-nate tissue from Agrobacterium without injuring itsregeneration potential, and to select the transgenicshoots whilst minimizing escapes in the process.At present, spring cultivars, especially Westar,have higher transformation rates than wintertypes. Two strategies are then possible: (i) tochoose to introduce the gene of interest into amodel spring cultivar, easily transformed and lesstime-consuming for the analysis of the genetictransmission, followed by at least four backcrossesto a more recalcitrant but agronomically desiredcultivar; or (ii) to choose a different transforma-tion technique, such as cotransformation with A.rhizogenes/A. tumefaciens. The latter is more time-consuming because one more step of segregationis required to eliminate the Ri T-DNA, but is effi-cient enough to obtain the required number oftransformed plants. Little is known on the stabilityof the transferred genes because the number oftransgenic plants studied is low and no systematicor statistical studies have been applied to this field,but it seems that extinction of some transgenesmay occur as noted in other species (Matzke et ah,1989; linn et al, 1990).

Agronomic applications of rapeseedtransformation

Some transformation experiments have alreadybeen carried out with different genes of agronomicor commercial interest. These experiments aredetailed in Table 9.2. Most of them refer to seedprotein expression, either the production of 'phar-maceutical' plants or the enrichment of the mealin some specific amino acids by increasing thequality or the quantity of a specific storage pro-tein. In these preliminary experiments, the level ofexpression of these transgenes was too low to leadto a notable alteration of the amino acid composi-tion of the meal. The future aims of the authorsare now to enhance the level of gene expression.


Efforts are now being made to produce herbi-cide-resistant plants to allow treatment of rapefields with herbicide, both to preferentially killweeds and also to kill the offspring of old rapeseedvarieties (the seed can survive for ten years in thesoil). Phosphinothricin (De Block et al, 1989),glyphosate (C. Primard et al, unpublished data)and sulfonylurea (Miki et al, 1990; C. Primard etal, unpublished data) resistant rapeseed plantshave been obtained by transformation techniques.At present, there are no commercial varieties, butsome of the transgenic plants are being used inrisk assessment programs in Canada and Francein order to measure the possible diffusion of thetransgenes into wild species (Kerlan et al, 1991).

The nuclear male sterility created by Mariani etal (1990) has also been transferred into rapeseedplants and is currently being tested to measure theefficiency of this hybrid production strategy.Rapeseed hybrids will soon be commercially veryimportant and this artificial means to control polli-nation could compete with natural genetic controls(nuclear or cytoplasmic male sterility, incompati-bility genes).

Some experiments are also in progress to modifythe oil quality of the seeds. Recently a Calgenegroup obtained high stearate transgenic rapeseedplants by an antisense strategy (Kridl et al, 1991).The expression of the antisense gene blocks the syn-thesis of the 89 desaturase, which is the first enzymeof the desaturation pathways of C18 fatty acid inrapeseed. These kinds of preliminary experimenthave proved that it is certainly possible to modifythe fatty acid metabolic pathway in rapeseed.

On one hand, rapeseed is a crop of considerableeconomic importance; on the other hand, it showsgood aptitude for in vitro culture, and transgenicrapeseed plants can be obtained by several of thepresent-day techniques. Therefore, it will cer-tainly be in the near future one of the first candi-dates for the commercialization of transformedcultivars. However, unless new transformationtechniques are developed much work has to bedone before rapeseed could be considered as amodel species for transformation.

Acknowledgements

The authors thank Kirk Schnorr, Ian Small andGeorges Pelletier for critical reading of the manu-script and helpful discussions.

References

Boulter, M., Croy, E., Simpson, P., Shields, R., Croy,R. & Shirsat, A. (1990). Transformation of Brassicanapus L. (oilseed rape) using Agrobacteriumtumefaciens and Agrobacterium rhizogenes: acomparaison. Plant Science, 70, 91-99.

Carr, R & McDonald, B. (1991). Rapeseed in achanging world: processing and utilisation. GroupeConsultatif International de Recherche sur le Colza,Rapeseed Congress, Saskatoon, 9-11 July, 1, 39-56.

Chapel, M. & Glimelius, K. (1990). Temporaryinhibition of cell wall synthesis improves thetransient expression of the GUS gene in Brassicanapus mesophyll protoplasts. Plant Cell Reports, 9,105-108.

Charest, P. J., Holbrook, L. A., Gabard, J. Iyer, V. N.& Miki, B. L. (1988). Agrobacterium -mediatedtransformation of thin cell layer explants fromBrassica napus L. Theoretical and Applied Genetics, 75,438-445.

Charest, P., Iyer, V. N. & Miki, B. L. (1989). Factorsaffecting the use of chloramphenicolacetyltransferase as a marker for Brassica genetictransformation. Plant Cell Reports, 7, 628-631.

Chevre, A. M., This, P., Eber, F., Deschamps, M.,Renard, M., Delseny, M. & Quiros, F. (1991).Characterization of disomic addition lines Brassicanapus-Brassica nigra by isozyme, fatty acid andRFLP markers. Theoretical and Applied Genetics, 81,43-49.

Dale, P. J. & Ball, L. F. (1991). Plant regenerationfrom cotyledonary explants in a range of Brassicaspecies and genotypes. Groupe ConsultatifInternational de Recherche sur le Colza, RapeseedCongress, Saskatoon, 9-11 July, 4, 1122-1127.

Damgaard, O. & Rasmussen, O. (1991). Directregeneration of transformed shoots in Brassica napusfrom hypocotyl infections with Agrobacteriumrhizogenes. Plant Molecular Biology, 17, 1-8.

De Block, M. & Debrouwer, D. (1991). Two T-DNA's co-transformed into Brassica napus by adouble Agrobacterium tumefaciens infection aremainly integrated at the same locus. Theoretical andApplied Genetics, 82, 157-263.

De Block, M., Debrouwer, D. & Tenmng, P. (1989).Transformation of Brassica napus and Brassicaoleracea using Agrobacterium tumefaciens and theexpression of the bar and neo genes in the transgenicplants. Plant Physiology, 91, 694-701.

De Clercq, A., Vandewiele, M., Van Damme, J.,Guerche, P., Van Montagu, M., Vandekerckhove, J.& Krebbers, E. (1990). Stable accumulation ofmodified 2S albumin seed storage proteins withhigher methionine contents in transgenic plants.Plant Physiology, 94, 970-979.

Fry, J., Barnason, A. & Horsch, R B. (1987).


Transformation of Brassica napus with Agrobacteriumtumefaciens based vectors. Plant Cell Reports, 6,321-325.

Golz, C , Kohler, F. & Schieder, O. (1990). Transferof hygromycin resistance into Brassica napus usingtotal DNA of a transgenic B. nigra line. PlantMolecular Biology, 15, 475-483.

Guerche, P., Charbonnier, M., Jouanin, L., Tourneur,C , Paszkowski, J. & Pelletier, G. (19876). Directgene transfer by electroporation in Brassica napus.Plant Science, 52, 111-116.

Guerche, P., De Almeida, E. R. P., Schwarzstein,M. A., Gander, E., Krebbers, E. & Pelletier, G.(1990). Expression of the 2S albumin fromBertholletia excelsa in Brassica napus. Molecular andGeneral Genetics, 221, 306-314.

Guerche, P., Jouanin, L., Tepfer, D. & Pelletier, G.(1987a). Genetic transformation of oilseed rape{Brassica napus) by the Ri T-DNA of Agrobacteriumrhizogenes and analysis of inheritance of thetransformed phenotype. Molecular and GeneralGenetics, 206, 382-386.

Hamill, J. D., Prescott, A. & Martin, C. (1987).Assessment of the efficiency of cotransformation ofthe T-DNA of disarmed binary vectors derived fromAgrobacterium tumefaciens and the T-DNA of A.rhizogenes. Plant Molecular Biology, 9, 573-584.

Harpster, M. H., Townsend, J. A., Jones, J. D. G.,Bedbrook, J. & Dunsmuir, P. (1988). Relativestrengths of the 35S cauliflower mosaic virus, 1', 2'and nopaline synthase promoters in transformedtobacco, sugarbeet and oilseed rape callus tissue.Molecular and General Genetics, 212, 182-190.

Holbrook, L. & Miki, B. (1985). Brassica grown galltumourigenesis and in vitro of transformed tissue.Plant Cell Reports, 4, 329-333.

Horsch, R. B., Fraley, R. T., Rogers, S. G., Sanders,P. R., Lloyd, A. & Hoffinann, N. (1983).Inheritance of functional foreign genes in plants.Science, 223, 496-498.

Hrouda, M., Dusbabkova, J. & Necasek, J. (1988)Detection of Ri T DNA in transformed oilseed raperegenerated from hairy roots. Biologia Plantarum, 30,234-236.

Jefferson, R. A., Kavanagh, T. A. & Bevan, M. W.(1987). GUS fusions: fi Glucuronidase as a sensitiveand versatile gene marker in higher plants. EMBOJournal, 6, 3901-3907.

Julliard, J., Sossountzov, L., Habricot, Y. & Pelletier,G. (1992). Hormonal requirements and timecompetency for shoot organogenesis in two cultivarsof Brassica napus. Physiologia Plantarum, 84,521-530.

Kerlan, M. C , Chevre, A. M., Eber, F., Botterman, J.& De Greef, W. (1991). Risk assessment of genetransfer from transgenic rapeseed to wild species inoptimal conditions. Groupe Consultatif International

de Recherche sur le Colza, Rapeseed Congress,Saskatoon, 9-11 July, 4, 1028-1033.

Klee, H. J. & Rogers, S. G. (1989). Plant gene vectorsand genetic transformation: plant transformationsystems based on the use of Agrobacteriumtumefaciens. Cell Culture and Somatic Cell Genetics ofPlants, Vol. 6: Molecular Biology of Plant NuclearGenes, ed. J. Schell & I. K. Vasil, pp. 1-23.Academic Press Inc, New York.

Kridl, J. C , Knutzon, D. S., Johnson, W. B.,Thompson, G. A., Radke, S. E., Turner, J. C. &Knauf, V. C. (1991). Modulation of stearoyl-ACPdesaturase levels in transgenic rapeseed. InternationalSociety for Plant Molecular Biology, ThirdInternational Congress, Tucson, 6-11 October,Congress Abstracts, 723.

Lee, W. S., Tzen, J. T. C , Kridl, J. C , Radke, S. E. &Huang, A. H. C. (1991). Maize oleosin is correctlytargeted to seed oil bodies in Brassica napustransformed with the maize oleosin gene. Proceedingsof the National Academy of Sciences, USA, 88,6181-6185.

Linn, F., Heidmann, I., Saedler, H. & Meyer, P.(1990). Epigenetic changes in the expression of themaize Al gene in Petunia hybrida: role of numbers ofintegrated gene copies and state of methylation.Molecular and General Genetics, 222, 329-336.

MacCalla, A. & Carter, C. (1991). Rapeseed in achanging world: the impacts of global issues. GroupeConsultatif International de Recherche sur le Colza,Rapeseed Congress, Saskatoon, 9-11 July, 1, 1-14.

Mariani, C , De Beuckeleer, M., Truettner, J.,Leemans, J. & Goldberg, R. (1990). Induction ofmale sterility in plants by a chimaeric ribonucleasegene. Nature {London), 347, 737-741.

Matzke, M., Primig, M., Trnovsky, J. & Matzke, A.(1989). Reversible methylation and inactivation ofmarker genes in sequentially transformed tobaccoplants. EMBO Journal, 8, 643-649.

Miki, B. L., Labbe, H., Hattori, J., Ouellet, T., Gabard,J., Sunohara, G., Charest, P. J. & Iyer, V. N. (1990).Transformation of Brassica napus canola cultivars withArabidopsis thaliana acetohydroxide synthase genes:analysis of herbicide resistance. Theoretical and AppliedGenetics, 80, 449^58.

Misra, S. (1989). Heavy metal tolerant transgenicBrassica napus L. and Nicotiana tabacum L. plants.Theoretical and Applied Genetics, 78, 161-168.

Misra, S. (1990). Transformation of Brassica napus L.with a 'disarmed' octopine plasmid of Agrobacteriumtumefaciens: molecular analysis and inheritance of thetransformed phenotype. Journal of ExperimentalBotany, 41, 224, 269-275.

Moloney, M. M., Walker, J. M. & Sharma, K. K.(1989). High efficiency transformation of Brassicanapus using Agrobacterium vectors. Plant Cell Reports,8, 238-242.


Neuhaus, G., Spangenberg, G., Mittelsten Scheid, O.& Schweiger, H. G. (1987). Transgenic rapeseedplants obtained by the microinjection of DNA intomicrospore derived embryoids. Theoretical andApplied Genetics, 75, 30-36.

Ooms, G., Bains, A., Burrell, M., Karp, A., Twell, D.& Wilcox, E. (1985). Genetic manipulation incultivars of oilseed rape (Brassica napus) usingAgrobacterium. Theoretical and Applied Genetics, 71,325-329.

Pechan, P. (1989). Successful cocultivation of Brassicanapus microspores and proembryos withAgrobacterium. Plant Cell Reports, 8, 387-390.

Pua, E. C , Mehra-Palta, A., Nagy, F. & Chua, N. H.(1987). Transgenic plants of Brassica napus L.Bio/Technology, 5,815-817.

Radke, S. E., Andrews, B. M., Moloney, M. M.,Crouch, M. L., Kridl, J. C. & Knauf, V. C. (1988).Transformation of Brassica napus L. usingAgrobacterium tumefaciens: developmentally regulatedexpression of the reintroduced napin gene.Theoretical and Applied Genetics, 75, 685-694.

Renard, M., Brun, H., Chevre, A., Delourme, R.,Guerche, P., Mesquida, J., Morice, J., Pelletier, G.& Primard, C. (1992). Colza oleagineux. InAmelioration des Especes Vegetales Cultivees, ed. A.Gallais & H. Bannerot, pp. 135-145. INRA Ed.,Paris.

Robbelen, G. (1991). Rapeseed in a changing world:plant production potential. Groupe ConsultatifInternational de Recherche sur le Colza, RapeseedCongress, Saskatoon, 9-11 July, 1, 29-38.

Rouan, D. & Guerche, P. (1991). Transformation andregeneration of oilseed rape protoplasts. In PlantTissue Culture Manual, ed. K. Lindsey, Vol. B4,pp. 1-24. Kluwer Academic Publishers, Dordrecht.

Shahin, E., Sukhapinda, K., Simpson, R. & Spivey, R.(1986). Transformation of cultivated tomato by abinary vector in Agrobacterium rhizogenes: transgenicplants with normal phenotypes harbor binary vectorT-DNA, but no Ri-plasmid T-DNA. Theoretical andApplied Genetics, 72, 770-777.

Shillito, R. D., Paszkowski, J. & Potrykus, I. (1983).Agarose plating and a bead type culture techniqueenable and stimulate development of protoplast-derived colonies in a number of plant species. PlantCell Reports, 2, 244-247.

Shillito, R. D., Saul, M. W., Paszkowski, J., Miiller,M. & Potrykus, I. (1985). High efficiency directgene transfer to plants. Bio/Technology, 3,1099-1103.

Swanson, E. & Erickson, L. (1989). Haploidtransformation in Brassica napus using an octopine-producing strain of Agrobacterium tumefaciens.Theoretical and Applied Genetics, 78, 831-835.

Tepfer, D. (1984). Genetic transformation of severalspecies of higher plants by Agrobacterium rhizogenes:phenotypic consequences and sexual transmission ofthe transformed genotype and phenotype. Cell, 37,959-967.

Trail, F., Richards, C. & Wu, F. S. (1989). Geneticmanipulation in Brassica. In Biotechnology inAgriculture and Forestry, Vol. 9, Plant Protoplasts andGenetic Engineering II, ed. Y. P. S. Bajaj, pp.197-216. Springer-Verlag, Berlin.

Vandekerckhove, J., Van Damme, J., Van Lijsebettens,M., Botterman, J., De Block, M., Vandewiele, M.,De Clercq, A., Leemans, J., Van Montagu, M. &Krebbers, E. (1989). Enkephalins produced intransgenic plants using modified 2S seed storageproteins. Bio/Technology, 7, 929-932.

10Sunflower

Gunther Hahne

Introduction

The cultivated sunflower {Helianthus annuus L.) isone of the four most important species grownworldwide as oil crops. Sunflower oil is appreci-ated for its high nutritional value, due to a bal-anced content of fatty acids. This position as animportant, and in some countries almost exclu-sive, source of one of the basic food components,has been made possible by a very successful appli-cation of conventional breeding techniques.The economic importance of sunflower hasrecently stimulated interest in biotechnologicalapproaches, which are expected to further extendthe possibilities for improvement of this species. Amajor issue in this context is, of course, the trans-fer of isolated genes that will confer novel charac-ters on this important crop.

Although first reports of sunflower tissues cul-tured in vitro date back to the early times of planttissue culture (e.g. De Ropp, 1946; Hildebrandt,Riker & Duggar, 1946; Henderson, Durrell &Bonner, 1952; Kandler, 1952), serious interest inthe in vitro culture of sunflower tissues and cells isquite recent, a situation that is reflected in thecomparatively small number of publications deal-ing with this issue. Moreover, this species hasproven quite recalcitrant to regeneration, andtechnological advances as well as the comprehen-sion of the underlying problems still flow at a slowrate. In the absence of a significant number ofpublications on transgenic sunflowers, the pur-pose of this review cannot be a critical discussionof the stability of foreign genes in such plants, noran evaluation of the methods by which these genescould have been introduced. This review rather

attempts to compile the existing tissue culture sys-tems mat might be, or have already been, used toapproach the problem of sunflower transforma-tion, and to highlight the difficulties that havebeen encountered at various levels.

Biological background information

The genus Helianthus belongs to the familyAsteraceae (formerly Compositae) and is com-posed of 67 species (Heiser, 1976) which areannual or persistent herbaceous plants, growing inclimates ranging from arid to more temperate con-ditions (Heiser et at., 1969). While some of thesespecies are known as noxious weeds in certainareas, and others are grown as ornamental plants,the only species with some economic value areHelianthus tuberosus (Jerusalem artichoke) andHelianthus annuus (the cultivated sunflower).Most sunflower lines cultivated today are F2hybrids, the production of which necessitates anintricate system of cytoplasmic male sterility com-prising sterile and fertile inbred lines as well asrestorer lines. This fact has important conse-quences for the strategies to be chosen for a practi-cally useful transformation protocol. A trulyuseful protocol must be applicable to inbred linesand must not be limited to a small range of geno-types.

Any approach to transformation has to take theparticular biological properties of the studiedplant into account. A brief description of thegrowth cycle of sunflower will facilitate the under-standing of the specific problems encounteredduring transformation and regeneration, and willmake the choice of certain explants more obvious.

137

138 HAHNE

Sunflower seeds are quite large (8-10 mm) andenclosed in a hard shell (pericarp) composed oftwo halves containing not only the embryo, butalso an air cavity that generally harbors fungalspores often difficult to eliminate. Elimination ofbacterial and fungal contaminants from sunflowertissues is perceived as a major obstacle in estab-lishing tissue cultures. Upon germination, sun-flower kernels (achenes) quickly produce a longand sturdy hypocotyl bearing two large cotyle-dons. Adult sunflowers are tall plants with largeleaves. The size of cultivated lines ranges from1.5 m to more than 2.5 m. Their size, theirrequirement for high light intensities, and theirattractiveness to a number of insect and fungalpests can present problems for the greenhouseculture of sunflower (Jeannin & Hahne, 1991).Male sterile lines used as female parents, and F1hybrids used for seed production, usually haveone single stem terminating in one head ofimpressive size. Pollinator lines are in generalbranched and bear multiple, smaller inflores-cences. The transition from the vegetative to thereproductive phase is in most cultivars relativelyindependent of environmental factors such as daylength, and appears to be developmentallyregulated (Steeves et al> 1969). Depending ongenotype and growth conditions, anthesis isapproximately 8-12 weeks after sowing. As ischaracteristic for Asteraceae, the inflorescencesare radially symmetrical and composed of a multi-tude of small florets. The sterile ray flowers withtheir fused petals are responsible for the attractiveyellow outer whorl, whereas the center containsthe inconspicuous but fertile disc flowers. Theopening of these flowers is progressive and pro-ceeds from the circumference to the center of theinflorescence. Depending on the size of the flowerhead, this process may take 5-10 days. Embryodevelopment is consequently asynchronous withinone inflorescence. The resulting embryos developquickly and reach their final length approximately12-15 days after fertilization, but full maturityrequires that the seeds remain some two addi-tional months on the plant. For experimental pur-poses, however, this time can be shortened andthe dehusked immature seeds can be germinatedin vitro when harvested only 10-20 days after pol-lination. More detailed information on sunfloweranatomy and development can be found in theauthoritative publication edited by Carter (1978).

Approaches to sunflower transformation

Because of the convenient experimental proper-ties of sunflower seedlings (size of hypocotyl andcotyledons), sunflower has been a favorite subjectfor physiological studies. Its susceptibility toAgrobacterium tumefaciens has been recognizedalready quite early, and crown gall tumors havebeen the subject of numerous studies (e.g. DeRopp, 1946; Matzke et al.y 1984; Ursic, 1985;Yao, Jingfen & Kuochang, 1988). An adaptationof the early protocols for the production of trans-genic callus rather than tumors, however, has onlybeen taken quite recently (Everett, Robinson &Mascarenhas, 1987; Nutter er a/., 1987;Escand6n& Hahne, 1991). These experiments made use ofdisarmed strains of A. tumefaciens. The leaf discapproach, which has been so successful fortobacco (Horsch et ah, 1985), is not applicable tosunflower because no reliable regeneration systemfrom leaves is available. Long-term clonal propa-gation of sunflower shoot cultures as a convenientand axenic source of leaves for such experiments isdifficult or even impossible, due to the tendency ofprecocious flowering in vitro that puts an end tothe vegetative multiplication phase after only a fewweeks (e.g. Henrickson, 1954; Paterson, 1984).Regeneration from cultured tissues, in particularfrom transgenic cultures such as callus or cell sus-pensions, remains difficult. The cases whereregeneration from (nontransformed) callus hasbeen successful either made use of immatureembryos as donor material (Wilcox McCann,Cooley & van Dreser, 1988; Espinasse & Lay,1989; Espinasse, Lay & Volin, 1989) or musttoday be considered as an exceptional event,restricted to specific conditions (Greco et al>1984; Paterson & Everett, 1985; Lupi et al,1987).

Although gene transfer to sunflower tissues hasbeen routine practice for a long while, andalthough transformed plants have been described,the production of transgenic sunflowers is far frombeing a routine procedure. The approaches pub-lished to date suffer from very low efficiencies andpoor reliability (Everett et aL, 1987; Schrammeijeret al.y 1990). One procedure that is reported not tosuffer from these shortcomings has recently beencommunicated by Malone-Schoneberg et al.(1991; Bidney, Malone-Schoneberg & Scelonge,1992a). Their protocol consists of a combinationof A. tumefaciens and the particle gun. Details of

Sunflower 139

this approach have been published for tobacco(Nicotiana species; Bidney et al> 19926).Information concerning sunflower is availableonly in conference abstracts (Malone-Schoneberget al., 1991; Bidney et aL, 1992a), and details onthe production of transgenic sunflowers are notavailable.

Regeneration: a fundamental necessity

The production of transgenic plants involves twodistinct processes: (i) the transfer of the foreigngene into a cell, and (ii) the regeneration of awhole plant from the progeny of this cell. Sinceregeneration is the limiting step for sunflowertransformation, I will devote a rather large sectionto the review of the different regeneration systemsavailable.

Adventitious regeneration from tissues

Adventitious regeneration from an explant canoccur either directly, i.e. some cells of the tissueare stimulated to develop into a somatic embryoor a shoot without a prolonged dedifferentiatedphase, or indirectly, where cells of the explant firstform a callus, which can later be induced to regen-erate plants. Both pathways are compatible withtransformation. Direct regeneration systems arefaster and usually show less somaclonal variationthan indirect ones, but selection for transformedtissues is much easier on the callus stage.

Direct regenerationBoth direct somatic embryogenesis and directshoot morphogenesis have been described for sun-flower. Immature zygotic embryos have a highmorphogenic capacity for somatic embryogenesis(Finer, 1987; Freyssinet & Freyssinet, 1988) aswell as shoot morphogenesis (Jeannin & Hahne,1991). To induce direct regeneration from imma-ture embryos, these are generally isolated from theinflorescence when they have reached a length of1-5 mm. They are then cultured on a mediumsupplemented with either auxin (Finer, 1987) orcytokinin (Freyssinet & Freyssinet, 1988; Jeannin& Hahne, 1991). A high osmotic pressure appearsto favor the embryogenic response. Fertile plantshave been obtained from a range of genotypes,although with different efficiencies. For regenera-tion, this approach is fast and relatively reliable,

but overall efficiency is limited (0.1-0.5 plantsregenerated/explant). The constant availability ofthe explant material requires a considerable effortin the greenhouse, and the long exposure of thedonor plant to varying environmental conditionsresults in an elevated experimental variability(Jeannin & Hahne, 1991). Direct morphogenesisfrom immature embryos has not yet been used fortransgenic plant production.

The cotyledons of mature seeds or germinatingseedlings give rise to shoots directly formed ontheir surface (Power, 1987; Nataraja &Ganapathi, 1989; Chraibi et al., 1991, 1992;Knittel, Escandon & Hahne, 1991). Experi-mentally, this system is convenient because theperiod required for the production of the donormaterial is less than a week, and this process caneasily be performed in a strictly controlled envi-ronment. The number of initially induced shootscan be rather high, although a large proportion ofthem may be abortive and, of the remainder, notall shoots may develop to full plants owing tocompetition. The morphogenic response is criti-cally dependent on the developmental stage of theexplant (Knittel et al., 1991) and shows a stronggenotype dependence. This latter limitation, how-ever, can be overcome by culturing the explants inliquid rather than solid medium for 2 weeks(Chraibi et al., 1992), or in the presence of in-hibitors of ethylene production or action (Chraibiet at., 1991). Fertile, regenerated plants areobtained quickly: the time between sowing of thedonor plant and harvesting seeds from the regen-erated plant may be as short as 4 months. Thisregeneration system should be sufficiently effi-cient to be used for transformation, but the cotyle-don is not easily infected with A. tumefaciens.

Indirect regenerationCallus induction is possible on virtually any tissueof young or adult sunflower plants. Because of itsconvenience, hypocotyl is often used for callusinduction (Greco et al, 1984; Paterson & Everett,1985; Piubello & Caso, 1986; Robinson &Adams, 1987). These calli may be vigorouslygrowing and can be maintained for long periods.They also represent a convenient source for cellsuspension cultures. Mature tissues that are suit-able as explants for plant regeneration from callus,however, are limited to the hypocotyl of certaingenotypes (Greco et aL, 1984; Paterson & Everett,1985; Lupi et al, 1987). Unfortunately, such a

140 HAHNE

genotype with high regeneration capacity is notavailable to the public. Indirect embryogenesis hasbeen demonstrated on epidermal thin layerexplants from hypocotyl of an inbred line(Pelissier et al., 1990). In this system, a high num-ber of embryos could be induced, but failed togerminate. Secondary embryogenesis has permit-ted the regeneration of fertile plants, but frequen-cies are not indicated for this step.

As is true for direct regeneration, immatureembryos are a good source also for morphogeniccallus (Wilcox McCann et al, 1988; Witrzens etal., 1988; Espinasse & Lay, 1989; Espinasse et al.,1989). This material shows a pronounced geno-type X medium interaction (Espinasse & Lay,1989). For callus induction, embryos must be iso-lated early, at approximately heart-to-torpedostage (0.2-1.5 mm), corresponding to approxi-mately 4 days after pollination. In most cases thecalli are not cultured for long periods, becausetheir regeneration potential is highest shortly aftercallus induction and then decreases rapidly.Although induction of embryogenesis in suspen-sion cultures of sunflower has been described(Prado & Berville, 1990), I am not aware of anycallus or cell suspension culture that has beenmaintained for a long period and still shows a rea-sonably high morphogenic activity.

Regeneration from callus may occur via shootmorphogenesis (Cavallini & Lupi, 1987;Espinasse et al., 1989) or somatic embryogenesis(Paterson & Everett, 1985), irrespective of thetype of donor material used for callus induction.The efficiency of the infection of sunflowerhypocotyl with A. tumefaciens is well documented(Matzke et al, 1984; Everett et aL, 1987), and theproduction of transformed callus presents noproblem (Escandon & Hahne, 1991). The feasi-bility of this approach for plant transformation hasbeen demonstrated, albeit with low efficiency andreliability (Everett et al, 1987). No informationis available concerning the use of immatureembryos for the production of transgenic callusand the possibility of subsequent plant regenera-tion.

Regeneration from existing meristems

Because of the difficulties encountered withregeneration from adventitious shoots andsomatic embryos, alternative strategies have beenexplored. The apical meristem of young plantlets

can be induced to proliferate and allows the regen-eration of a relatively high number of shoots thatcan be transferred to the greenhouse (Paterson,1984). However, the meristem is not easily acces-sible for A. tumefaciens, and thus the use of thisapproach for the production of transgenic plants islimited by its low efficiency (Schrammeijer et aL>1990). This limitation seems to be eliminatedby the combination of A. tumefaciens and the par-ticle gun already mentioned above (Malone-Schoneberg et a/., 1991; Bidney et al, 1992fc), butfull experimental details concerning sunflowerexplants are not yet published.

Regeneration from single cells

ProtoplastsProtoplasts have been isolated and cultured froma variety of sunflower tissues, including hypocotyl(Lenee & Chupeau, 1986; Moyne et aL, 1988;Chanabe, Burrus & Alibert, 1989; Schmitz &Schnabl, 1989; Dupuis, Pean & Chagvardieff,1990), cotyledons (Bohorova, Cocking & Power,1986), petioles (Schmitz & Schnabl, 1989), leaves(Guilley & Hahne, 1989; Schmitz & Schnabl,1989), and suspension cultures (G. Hahne,unpublished data). The isolation of protoplastspresents no particular difficulties, and theobtained yields are comparatively high. Yieldsaverage 2-5 million protoplasts/g tissue for mostdonor materials. Media have been defined that aresuitable for the regeneration of callus from allthese materials. The genotype appears to be ofsubordinate importance for the ability of proto-plasts to develop into callus, and most of thedescribed media can be used for all protoplasttypes. Sunflower protoplasts are in general quitestable, with the exception of mesophyll proto-plasts, which tend to be fragile under some condi-tions. The plating efficiencies for the first divisionsare high, a property that makes them useful fortransient expression studies. In order to obtainreasonable plating efficiencies for callus produc-tion, it is advisable to embed the protoplasts in asemi-solid matrix. Both agarose (Moyne et al>1988; Dupuis et aL, 1990) and alginate (Fischer &Hahne, 1992) are convenient and have been used.While it is not difficult to produce vigorouslygrowing callus from protoplasts, successful shootinduction has rarely been reported for H. annuus>and fertile plants have been obtained in only twocases. Binding et al (1981) reported shoot regen-

Sunflower 141

eration from shoot-tip-derived protoplasts, butunfortunately the experimental details and thegenotype were insufficiently described for theresults to be reproduced. Burrus et al (1991)obtained a small number of fertile plants fromhypocotyl protoplasts of a genotype that had beenspecially selected for its regeneration capacity.The feasibility of plant regeneration from sun-flower protoplasts has thus been demonstrated.The regeneration protocol itself is quite classic; itcan therefore be assumed that the genotypic effectis the decisive factor. This genotype, derived froman interspecific hybrid of H. annuus and H. petio-laris, is unfortunately privately owned and notavailable to the public. We have recently estab-lished a protocol for the regeneration of fertileplants from protoplasts of an inbred sunflower line(Fischer, Klethi & Hahne, 1992). This protocoldiffers from that described by Burrus et al (1991)in many aspects, and both appear to be specific fortheir respective genotypes. Given the availabilityof a suitable genotype, there is no fundamentalobstacle to the use of protoplasts for the produc-tion of transgenic plants other than the overallefficiency, which is still rather low even in the spe-cially selected genotype (Burrus etal, 1991).

Pollen and oocytesAlthough pollen and oocytes are not usually con-sidered as objects for plant transformation, theydo have some potential for genetic manipulation(e.g. Neuhaus et al, 1987). Since none of themore traditional approaches forms the basis of asound transformation system for sunflower, Iinclude these reproductive cells in this briefoverview of the state of the in vitro techniques forsunflower.

Plants have been regenerated from culturedanthers of interspecific hybrids in the genusHelianthus (Bohorova, Atanassov & Georgieva-Todorova, 1985; Giirel, Nichterlein & Friedt,1991), and haploid plants have been obtainedthrough androgenesis in sunflower (Mezzaroba &Jonard, 1986). However, frequencies and repro-ducibility are still too low to permit routine appli-cation of this technique.

Although the isolation of oocytes is more diffi-cult and yields are much lower than is the case formicrospores, gynogenesis is an alternativeapproach for haploid plant production in sun-flower (Hongyuan et al, 1986; Gelebart & San,1987). This technique, although successful in

some cases, is not widely applied because of tech-nical difficulties and low efficiency.

Possible strategies to use pollen, microspores,or oocytes for the production of transgenic plantsinclude microinjection and the use of the particlegun (see below).

Properties of regenerated plants

While shoots obtained from preexisting meristemspresent no major difficulties for the transfer intothe greenhouse, except premature flowering invitro, most adventitious shoots and even somaticembryos are vitrified to a more or less seriousextent. Vitrification is a phenomenon encounteredin many tissue culture systems (e.g. see Bottcher,Zoglauer & Goring, 1988), but its causes arepoorly understood. It must, at present, beresolved empirically for each case. No generalrecipe exists to prevent vitrification in the case ofsunflower, but factors that might be consideredinclude the gas phase, relative humidity, waterpotential of the medium, and hormonal composi-tion of the media used throughout the regenera-tion procedure. Trivial factors such as the type ofcontainer and its position in the culture room maybe decisive for the production of shoots with aquality sufficient for a transfer into the green-house. With few exceptions, adventitious shootsand even somatic embryos present enormous diffi-culties in root induction, and this step is oftenresponsible for a very low efficiency of the transferof shoots to the greenhouse (Freyssinet &Freyssinet, 1988; Burrus et aly 1991). Once rootshave been obtained, plants can usually be accli-mated without major difficulties, and most ofthem will proceed to the flowering stage.Alternatively, shoots that fail to produce roots canstill be transferred to the greenhouse when graftedon to an established rootstock. We have obtainedvigorously growing, protoplast-derived shoots bythis technique, and seed set was unproblematic(Fischer et al, 1992).

Due to the generally very low regenerationefficiencies, not many studies are available con-cerning tissue-culture induced variability. The re-generated plants themselves (RO) are generallystunted (e.g. Finer, 1987; Wilcox McCann et al,1988; Jeannin & Hahne, 1991) and are often nottaller than 20-50 cm, although they may reach thesize of normal sunflower in exceptional cases.Even in lines that are characterized by only one

142 HAHNE

stem bearing a single head, it is not unusual for ROplants to be highly branched and to produce mul-tiple inflorescences. In short, the overall appear-ance of regenerated sunflowers hardly resemblesthat of normal plants. However, the reasons forthis are physiological rather than genetic, and thisphenotype is not transmitted to ensuing genera-tions (Freyssinet & Freyssinet, 1988; Jeannin &Hahne, 1991).

From the few published studies on genetic sta-bility of tissue culture-derived sunflower plants itemerges that plants derived from immatureembryos by direct organogenesis or somaticembryogenesis appear to be genetically quite sta-ble. Significant deviations from the parent pheno-type have not been detected in the offspring ofregenerated plants during several generations(Freyssinet & Freyssinet, 1988; Jeannin & Hahne,1991). Karyotypic variations have not beendetected in three generations after plant regenera-tion from immature embryos (Jeannin, Poirot &Hahne, 1990; Jeannin & Hahne, 1991).

Somaclonal variants with altered coumarin con-tent have been obtained after mercuric chloride-treatment of immature embryo-derived callus(Roseland, Espinasse & Grosz, 1991). No detailedanalysis is available concerning plants derived byany of the other regeneration routes. OnlyWitrzens et al (1988) have indicated the occur-rence of possible somaclonal variants after regen-eration from immature embryo-derived callus ofone genotype.

The basic problems in culture of sunflowerin vitro

We can conclude that a variety of regenerationsystems is available for sunflower. Regenerationhas been demonstrated to be possible from almostany conceivable multi- or unicellular explant, butonly a few of them can be considered as actuallyestablished, universally applicable protocols(these are based mainly on the use of immatureembryos). The recurring problems are (i) low effi-ciency, (ii) low reliability, and (iii) low universal-ity, i.e. the frequently encountered limitation toone or few genotypes. The approaches to regener-ation that are more generally applicable and havea reasonable efficiency for plant regeneration, i.e.direct regeneration systems and callus induced onimmature embryos, cannot easily be used in con-junction with A. tumefaciens. In contrast, those

systems that are quite amenable to Agrobacterium-mediated transformation have only low regenera-tion efficiencies or very limited applicability. Theefficiency of virtually any regeneration system iscritically dependent on the genotype used, withthe more efficient lines often not being available tothe public. Of the public lines, the inbred line HA300 in both its male sterile (A) and male fertile (B)forms presents perhaps the highest regenerationpotential, although its performance is in manyrespects not satisfactory. A good regenerationpotential from immature embryos is found in HA89B. Moreover, the performance of a public linesuch as HA 300 may vary in function because ofthe supplier from whom the seeds have beenobtained.

Transformation

Genetic markers

Any method used for the transformation of planttissues or cells occurs with an efficiency far below100%. If long and tedious screening on the level ofregenerated plants is to be avoided, a powerfuland reliable selection for transformed cells isindispensable. A number of selectable markershave been used in transformation experiments ofvarious sunflower explants. The nptll gene codingfor the enzyme, neomycin phosphotransferase II(NPT II), is perhaps the most universally utilizedgenetic marker for transformed plant tissues. Thisenzyme inactivates by phosphorylation a numberof compounds belonging to the class of amino-glycoside antibiotics, including kanamycin, paro-momycin, gentamycin, etc. Kanamycin hasproven its efficiency for tobacco and many otherspecies (e.g. see Fraley, Rogers & Horsch, 1987;Potrykus, 1991), and has been employed for theselection of transformed callus obtained from sun-flower hypocotyl (Everett et al> 1987) or proto-plasts (Moyne et al, 1989). Selection has beenpossible in these cases, but the regenerationcapacity of the callus was severely diminished(Everett et al.y 1987). Other authors observed ahigh spontaneous resistance of sunflower tissuestowards kanamycin and were unable to use thiscompound for selection, while other aminoglyco-sides such as paromomycin or G418 could beused successfully (Escandon & Hahne, 1991).Phosphinothricin (PPT; 'Basta'), inactivated bythe product of the bar gene (phosphinothricin

Sunflower 143

acetyl transferase (PAT; De Block et aL, 1987))also has proven to be a useful selective marker(Escandon & Hahne, 1991). The reasons for theconflicting observations concerning kanamycinare not clear, but may be due to a genotypic effect.Furthermore, factors such as the auxin : cytokininratio and N-source in the culture medium may beimportant for the efficiency of the selection(Escandon & Hahne, 1991).

Quantitative and semi-quantitative markersthat have been used successfully with sunflowertissues include (3-glucuronidase (GUS), NPT II,and chloramphenicol acetyl transferase (CAT).The availability of several quantifiable markersallows the use of an internal standard for transientexpression assays. The experimental variabilitycan thus be decreased, and individual experimentsrendered more comparable (as shown for tobaccoby Lepetit et al. (1991)). Quantitative assays areusually employed for transient expression studies.The histological GUS reaction, which is semi-quantitative when interpreted with caution, hasalso been utilized for the detection of stable trans-formation events in sunflower. No backgroundproblems are encountered with the GUS stainwhen bacterial expression is excluded (e.g. byusing the GUS-intron construction (Vancanneytet al., 1990)), but sunflower tissue has a tendencyto turn brown during the fixation and stainingprocess, rendering the blue color difficult todetect.

Agrobacterium-mediated gene transfer

All experiments resulting in transformed sun-flower tissues have made use of A. tumefaciens,with one exception (microinjection; Espinasse-Gellner, 1992). All transformed sunflower plantspublished to date with a detailed experimentaldescription have been obtained using this vector(Everett et al., 1987; Schrammeijer era/., 1990).

Crown gall tumors of sunflower were used earlyon to study the integration (Ursic, 1985) andexpression (Matzke et al., 1984) of foreign genes.The Agrobacterium strains used in these studieswere of wild-type, and although the resultingtumors could be established in vitro (Ursic, 1985),no plants have been regenerated from these cul-tures for obvious reasons. This aspect will not beconsidered in more detail here.

Two publications exist to date describing trans-formed sunflower plants, and two additional

approaches have been described in conferenceabstracts. First, Everett et al. (1987) used a dis-armed strain of A. tumefaciens to transformhypocotyl explants of a genotype that regeneratedefficiently from callus induced on this explant.Upon selection on kanamycin, however, theregeneration potential was lost to a great extent. Itwas necessary to subculture the selected callus onnonselective medium for several months beforeshoots could be regenerated. Seventeen plantswere obtained and were shown to be transgenic bykanamycin resistance and Southern analysis. Nofurther publication followed this initial report, buta short note by Hartmann (1991) indicated thatsome difficulties were encountered and that theregeneration and selection system needed furtherwork. Second, Schrammeijer et al (1990) usedlongitudinally cut meristems of mature embryosfor their transformation experiments; they treatedthe specimens with a suspension of A. tumefaciens.The bacteria were indeed able to penetrate intothe meristematic tissue in a few cases, which thengave rise to plants containing transformed celllines. This approach has the benefit of being verystraightforward, and at least theoretically beingindependent of genotypic restrictions. However,the efficiency is extremely low. Of the 1500 meri-stems treated, two transgenic shoots wereobtained with chimeric transgene expression.Progeny could be obtained from one shoot, but noexpression of the uidA gene was detectable in theoffspring.

The approach of Schrammeijer et al. (1990) hasbeen used in a modified version by Malone-Schoneberg et al. (1991; Bidney et aL, 1992a) toproduce transgenic sunflower plants. Their origi-nal technique combines the particle gun approachand Agrobacterium as a very efficient vector forDNA delivery. Some experimental details aregiven in the related paper (Bidney et al., 19926)using tobacco as a model, and also describingtransformed sectors on sunflower shoots. Thesmall, dense projectiles of the particle gun areused in this approach not for DNA delivery (as inthe technique that has become standard today; fora review, see Christou, 1992), but to createmicrolesions that allow access of Agrobacterium tothe tissue in a very efficient way (Bidney et al.,19926). This combination increases the transfor-mation efficiency for tobacco leaves by 100-fold,and it can be expected that it is equally efficientfor sunflower meristems. A number of transgenic

144 HAHNE

sunflowers have been produced using thisapproach (Bidney et al, 1992a). The descriptionof technical details and characterization of theobtained plants (Bidney et al 1992c) is not verydetailed, but it appears that most of Bidney et a/.'sprimary transformants produced with this tech-nique are chimeric for the expression of the intro-duced gene(s). Our own experiments withembryonic axes indicate that an appropriate selec-tion scheme can limit the occurrence of chimeras(N. Knittel et al, unpublished data).

Direct gene transfer

Particle gunThe particle gun in its many versions has beenused successfully for the transformation of a num-ber of plant species that had hitherto defied allefforts towards transformation (e.g. see Klein etal, 1988; McCabe, Martinell & Christou, 1988),and for transient expression studies in tissues thatare inaccessible for protoplast isolation (e.g. seeHamilton et al, 1992; van der Leede-Plegt et al,1992). It would therefore appear logical that thispowerful tool had been applied to sunflower.However, apart from the novel approach used byBidney et al (\992a,b), no successful applicationof the particle gun to any sunflower tissues ororgans has been reported so far.

MicroinjectionTransgenic plants of Brassica napus have been pro-duced by microinjection in young, microspore-derived somatic embryos at the four to eight cellstage (Neuhaus etal, 1987). Homogeneous trans-genic plants could be obtained from the primarychimeric regenerants by secondary embryo-genesis. Microinjection thus appears to be a feasi-ble approach to the transformation of arecalcitrant species, provided the regeneration hasbeen established for an organ that can be micro-injected.

A similar approach has been taken for sun-flower, where embryos cultured in ovulo(Espinasse et al, 1991) are injected with DNA.The technique has been presented only as a con-ference abstract (Espinasse-Gellner, 1992); exper-imental details and a characterization of thetransgenic plants are thus not available. Theauthors have, however, reported on their successto regenerate plants from 3-day-old embryos ofinterspecific hybrids cultured in ovulo (Espinasse

et al, 1991). An evaluation of efficiency and uni-versality of this approach must await furtherdetails.

ProtoplastsDirect gene transfer into protoplasts is a well-established technique that is used extensively inmany species in one of two ways, namely electro-poration and polyethylene glycol (PEG)-mediatedDNA uptake. Analysis of the treated protoplastsafter only a short time (24-48 h) allows the detec-tion of a fairly strong signal in a significant numberof protoplasts (Negrutiu et al, 1988; Jung et al,1992), a phenomenon that has been termed 'tran-sient expression'. When protoplasts are cultured forlonger periods, the burst of transgene expressiondue to nonintegrated DNA disappears, and theexpression of DNA stably integrated in the genomecan be detected in a considerably lower percentageof protoplasts. This technique has yielded trans-genic plants in a steadily increasing number ofspecies (for a review, see Potrykus, 1991).

The feasibility of transient expression has beendemonstrated for sunflower using both electro-poration and PEG to stimulate DNA uptake(Kirches, Frey & Schnabl, 1991). Kirches et alhave defined the experimental parameters formesophyll protoplasts of the hybrid line TrimasoP.The optimal conditions are comparable to thoseoptimized for protoplasts from other species.

Stably transformed transgenic callus has beenobtained after PEG-mediated gene transfer intosunflower protoplasts (Moyne et al, 1989). Thefrequency of callus resisting kanamycin selectionand expressing the nptll gene was relatively low(4 X 1CT6) in this study. Since no regeneration pro-tocol is available for the genotype used, transgenicplants have not been obtained from this callus.

From these experiments it is clear that sunflowerprotoplasts do not have properties that preventtheir use in DNA transfer experiments. However,this approach is likely to develop into a viable alter-native for the production of transgenic sunflowerplants only if the frequencies can be considerablyimproved for both stable transformation eventsand, even more important, for plant regeneration.

Conclusion and perspectives

A certain number of gene transfer techniques havebeen established for sunflower, and, likewise, a

Sunflower 145

diverse range of regeneration systems is available.It is the combination of these two componentsinto one efficient protocol that is the source ofproblems for the production of transgenic sun-flower plants. The tissues for which relatively reli-able and universal regeneration protocols areavailable, i.e. immature embryos and cotyledons,are not easily transformed by A. tumefaciens. Incontrast, hypocotyl is relatively efficiently trans-formed, but its regeneration is difficult and seemspossible with a very limited choice of genotypesand in only few laboratories. The cases are rarewhere both gene transfer and subsequent plantregeneration have been successful. They aremostly characterized by extremely low efficienciesand a poor reproducibility. A possible exception isthe transformation of meristems by a combinationof A. tumefaciens and the particle gun (Bidney etal, I992a,b).

The most important obstacle in the waytowards the transgenic sunflower is beyond doubtthe still very poorly developed in vitro culture ofsunflower cells and tissues. With few notableexceptions, sunflower cell and tissue culture ischaracterized by a very pronounced genotypedependence, poor reproducibility in differentexperimental environments, and insufficient effi-ciency. It is at present unclear what the reasonsare for this recalcitrance. Wild relatives of sun-flower (Bohorova et al, 1986; Chanabe et al,1991; Krasnyanski et al, 1992) and certain inter-specific hybrids (Chandler & Beard, 1983; Burruset al, 1991) are in many respects less problematic,although less effort has been devoted to them.Perhaps the major handicap for workers in sun-flower in vitro culture is the strong genotypedependence of the regeneration process. Althoughgenotypes suitable for different aspects have beenidentified, lines that regenerate from the undiffer-entiated state have in general been private prop-erty (e.g. see Paterson & Everett, 1985; Everett etal, 1987; Burrus et al, 1991) that is accessible forother laboratories only under restrictive condi-tions, if at all. One or several model genotypesthat may have only a mediocre regenerationpotential or agronomic value, but could serve as apositive control, have not yet been identifiedamong the public lines. The establishment ofsuch a reference line could considerably facilitateand accelerate research towards the transformedsunflower.

In view of the low frequencies obtained in the

few published experiments that resulted in trans-genic sunflower plants, the absence of informationon the stability and inheritance of the foreigngenes is hardly surprising.

Interest in sunflower tissue culture has been ris-ing during the last years, and this is reflected bythe increasing number of publications on this sub-ject. Original approaches have been developedand considerable advances have been made con-cerning both gene transfer and plant regeneration,and there is reason to hope that it is only a ques-tion of time until both aspects are combined into aprotocol for routine sunflower transformation.

The list of some potentially useful genes to betransferred into sunflower, once a reliable trans-formation system has been established, will not bemuch different from a list applicable to any otherspecies. Apart from the study of genes and geneconstructions interesting for fundamental studies,but without immediate interest for application, anumber of biological and technological problemsmight be addressed by using transgenic sun-flowers. Sunflowers are grown virtually all over theworld, and the needs are evidently different foreach region. Immediate application could improveresistance against herbicides and insect attack,and in the more distant future against fungalpathogens. The other domain where it is hopedthat biotechnology will contribute to a new devel-opment, in due term, is the change of productquality. Novel utilizations of vegetable oil lead to achanging demand for oils of different quality, con-cerning not only the degree of desaturation butalso characters such as chain length. Furthermore,sunflower is used in some areas as a fodder crop,and improved amino acid composition of its stor-age proteins could render this application moreinteresting.

Today, all commercially successful hybrids, i.e.the overwhelming majority of sunflower seedsgrown worldwide, are based on the same cycto-plasmic male sterile (cms) cytoplasm. This poten-tially hazardous situation could be overcome byproducing novel cms lines and their restorer sys-tems by interspecific hybridization. The utilizationof genetic engineering approaches that are verypromising in other crops might provide for moreflexibility and less complicated breeding.

Other breeding goals identified today concernmore complex characters, such as drought resis-tance, early flowering for culture in more northernlatitudes, and height reduction. In the absence of

146 HAHNE

a clear understanding of the underlying physiol-ogy, these characters may at present be moresuccessfully addressed by conventional breedingand in vitro culture techniques, such as exploitingsomaclonal variation or somatic hybridization.

References

Bidney, D. L., Malone-Schoneberg, J. & Scelonge,C. J. (1992a). Stable transformation of sunflowerand field evaluation of transgenic progeny. InSunflower Research Workshop, pp. 57-58. NationalSunflower Association, Fargo.

Bidney, D. L., Scelonge, C. J. & Malone-Schoneberg,J. B. (1992c). Transformed progeny can berecovered from chimeric plants regenerated fromAgrobacterium tumefaciens treated embryonic axes ofsunflower. In Proceedings of the 13th InternationalSunflower Conference, pp. 1408-1412. InternationalSunflower Association, Pisa.

Bidney, D., Scelonge, C , Martich, J., Burrus, M.,Sims, L. & Huffinan, G. (19926). Microprojectilebombardment of plant tissues increasestransformation frequency by Agrobacteriumtumefaciens. Plant Molecular Biology, 18, 301-313.

Binding, H., Nehls, R., Kock, R., Finger, J. &Mordhorst, G. (1981). Comparative study onprotoplast regeneration in herbaceous species of theDicotyledoneae class. Zeitschrift fiirPflanzenphysiologie, 101, 119-130.

Bohorova, N., Atanassov, A. & Georgieva-Todorova,J. (1985). In vitro organogenesis, endrogenesis andembryo culture in the genus Helianthus L. Zeitschriftfiir Pflanzenziichtung, 95, 35-44.

Bohorova, N. E., Cocking, E. C. & Power, J. B.(1986). Isolation, culture and callus regeneration ofprotoplasts of wild and cultivated Helianthus species.Plant Cell Reports, 5, 256-258.

Bottcher, I., Zoglauer, K. & Goring, H. (1988).Induction and reversion of vitrification of plantscultured in vitro. Physiologia Plantarum, 72,560-564.

Burrus, M., Chanabe, C , Alibert, G. & Bidney, D.(1991). Regeneration of fertile plants fromprotoplasts of sunflower (Helianthus annuus L.).Plant Cell Reports, 10, 161-166.

Carter, J. F. (1978). Sunflower Science and Technology.American Society of Agronomy, Crop ScienceSociety of America, Soil Science Society of America,Inc., Publishers, Madison.

Cavallini, A. & Lupi, M. C. (1987). Cytological studyof callus and regenerated plants of sunflower(Helianthus annuus L.). Plant Breeding, 99,203-208.

Chanabe, C , Burrus, M. & Alibert, G. (1989).

Factors affecting the improvement of colonyformation from sunflower protoplasts. Plant Science,64, 125-132.

Chanabe, C , Burrus, M., Bidney, D. & Alibert, G.(1991). Studies on plant regeneration fromprotoplasts in the genus Helianthus. Plant CellReports, 9, 635-638.

Chandler, J. M. & Beard, B. H. (1983). Embryoculture of Helianthus hybrids. Crop Science, 23,1004-1007.

Chraibi, K., Castelle, J. C , Latche, A., Roustan, J. P.& Fallot, J. (1992). Enhancement of shootregeneration potential by liquid medium culturefrom mature cotyledons of sunflower (Helianthusannuus L.). Plant Cell Reports, 10, 617-620.

Chraibi, K. M., Latche, A., Roustan, J. P. & Fallot, J.(1991). Stimulation of shoot regeneration fromcotyledons of Helianthus annuus by the ethyleneinhibitors, silver and cobalt. Plant Cell Reports, 10,204-207.

Christou, P. (1992). Genetic transformation of cropplants using microprojectile bombardment. PlantJournal, 2, 275-281.

De Block, M., Botterman, J., Vanderwiele, M., Dockx,J., Thoen, C , Goessele, V., Mowa, N. R.,Thompson, C , Van Montagu, M. & Leemans, J.(1987). Engineering herbicide resistance in plants byexpression of a detoxifying enzyme. EMBO Journal,6,2513-2518.

De Ropp, R. S. (1946). The isolation and behavior ofbacteria-free crown-gall tissue from primary galls ofHelianthus annuus. Phytopathology, 37, 201-206.

Dupuis, J. M., Pean, M. & Chagvardieff, P. (1990).Plant donor tissue and isolation procedure effect onearly formation of embryoids from protoplasts ofHelianthus annuus L. Plant Cell, Tissue and OrganCulture, 22, 183-189.

Escandon, A. J. S. & Hahne, G. (1991). Genotype andcomposition of culture medium: factors important inthe selection for transformed sunflower (Helianthusannuus) callus. Physiologia Plantarum, 91, 367-376.

Espinasse, A. & Lay, C. (1989). Shoot regeneration ofcallus derived from globular to torpedo embryosfrom 59 sunflower genotypes. Crop Science, 29,201-205.

Espinasse, A., Lay, C. & Volin, J. (1989). Effects ofgrowth regulator concentrations and explant size onshoot organogenesis from callus derived fromzygotic embryos of sunflower (Helianthus annuus L.).Plant Cell, Tissue and Organ Culture, 17, 171-181.

Espinasse, A., Volin, J., Dybing, C. D. & Lay, C.(1991). Embryo rescue through in ovulo culture inHelianthus. Crop Science, 31, 102-108.

Espinasse-Gellner, A. (1992). A simple and directtechnique of transformation in sunflower. InSunflower Research Workshop, p. 50. NationalSunflower Association, Fargo.

Sunflower 147

Everett, N. P., Robinson, K. E. P. & Mascarenhas, D.(1987). Genetic engineering of sunflower(Helianthus annuus L.). Bio I Technology > 5,1201-1204.

Finer, J. (1987). Direct somatic embryogenesis andplant regeneration from immature embryos of hybridsunflower {Helianthus annuus L.) on a high sucrose-containing medium. Plant Cell Reports, 6, 372-374.

Fischer, C. & Hahne, G. (1992). Structural analysis ofcolonies derived from sunflower (Helianthus annuusL.) protoplasts cultured in liquid and semisolidmedia. Protoplasma, 169, 130-138.

Fischer, C , Klethi, P. & Hahne, G. (1992).Protoplasts from cotyledon and hypocotyl ofsunflower {Helianthus annuus L.): shoot regenerationand seed production. Plant Cell Reports, 11,632-636.

Fraley, R. T., Rogers, S. G. & Horsch, R. B. (1987).Genetic transformation in higher plants. CRCCritical Reviews in Plant Sciences, 4, 1-46.

Freyssinet, M. & Freyssinet, G. (1988). Fertile plantregeneration from sunflower (Helianthus annuus L.)immature embryos. Plant Science, 56, 177-181.

Gelebart, P. & San, L. H. (1987). Obtention deplantes haploides par culture in vitro d'ovaires etd'ovules non fecondes de tournesol (Helianthusannuus L.). Agronomie, 7, 81-86.

Greco, B., Tanzarella, O. A., Carrozo, G. & Blanco,A. (1984). Callus induction and shoot regenerationin sunflower (Helianthus annuus L.). Plant ScienceLetters, 36, 73-77.

Guilley, E. & Hahne, G. (1989). Callus formationfrom isolated sunflower (Helianthus annuus)mesophyll protoplasts. Plant Cell Reports, 8,226-229.

Giirel, A., Nichterlein, K. & Friedt, W. (1991). Shootregeneration from anther culture of sunflower(Helianthus annuus) and some interspecific hybridsas affected by genotype and culture procedure. PlantBreeding, 106, 68-76.

Hamilton, D. A., Roy, M., Rueda, J., Sindhu, R. K.,Sanford, J. & Mascarenhas, J. P. (1992). Dissectionof a pollen-specific promoter from maize bytransient transformation assays. Plant MolecularBiology, 18,211-218.

Hartmann, C. L. (1991). Agrobacterium transformationin sunflower. In Sunflower Research Workshop, pp.95-99. National Sunflower Association, Fargo.

Heiser, C. B. (1976). The Sunflower. University ofOklahoma Press, Norman.

Heiser, C. B., Smith, D. M., Clevenger, S. B. &Martin, W. C. J. (1969). The North AmericanSunflowers. In T. Delevoryas (Ed.), Memoirs of TheTorrey Botanical Club. The Seeman Printery,Durham, NC.

Henderson, J. H. M., Durrell, M. E. & Bonner, J.(1952). The culture of normal sunflower stem

callus. American Journal of Botany, 39, 467-473.Henrickson, C. E. (1954). The flowering of sunflower

explants in aseptic culture. Plant Physiology, 29,536-538.

Hildebrandt, A. C , Riker, A. J. & Duggar, B. M.(1946). The influence of the composition of themedium on growth in vitro of excised tobacco andsunflower tissue cultures. American Journal ofBotany, 33, 591-597.

Hongyuan, Y., Chang, Z., C , D., Hua, Y., Yan, W. &Xiaming, C. (1986). In vitro culture of unfertilizedovules in Helianthus annuus L. In Haploids of HigherPlants in Vitro ed. H. Han & Y. Hongyuan, pp.182-190. Springer Verlag, Berlin.

Horsch, R. B., Fry, J. E., Hoffman, N. L., Eichholz,D., Rogers, S. G. & Fraley, R. T. (1985). A simpleand general method for transferring genes intoplants. Science, 227, 1229-1231.

Jeannin, G. & Hahne, G. (1991). Donor plant growthconditions and regeneration of fertile plants fromsomatic embryos induced on immature zygoticembryos of sunflower (Helianthus annuus L.). PlantBreeding, 107, 280-287.

Jeannin, G., Poirot, M. & Hahne, G. (1990).Regeneration de plantes fertiles a partir d'embryonszygotiques immatures de tournesol. InCinquantenaire de la culture in vitro Versailles(France), 24-25 October 1989, ed. C. Dore, pp.275-276. Paris: Ed. INRA.

Jung, J. L., Bouzoubaa, S., Gilmer, D. & Hahne, G.(1992). Visualization of transgene expression on thesingle protoplast level. Plant Cell Reports, 11,346-350.

Kandler, O. (1952). Uber eine physiologischeUmstimmung von Sonnenblumenstengelgewebedurch Dauereinwirkung von fi-Indolylessigsaure.Planta, 40, 346-349.

Kirches, E., Frey, N. & Schnabl, H. (1991). Transientgene expression in sunflower mesophyll protoplasts.Botanica Acta, 104, 212-216.

Klein, T. M., Fromm, M. E., Weissinger, A., Tomes,D., Schaaf, S., Sletten, M. & Sanford, J. C. (1988).Transfer of foreign genes into intact maize cellsusing high velocity microprojectiles. Proceedings of theNational Academy of Sciences, USA, 85, 4304-4309.

Knittel, N., Escandon, A. S. & Hahne, G. (1991).Plant regeneration at high frequency from maturesunflower cotyledons. Plant Science, 73, 219-226.

Krasnyanski, S., Polgar, Z., Nemeth, G. & Menczel,L. (1992). Plant regeneration from callus andprotoplast cultures of Helianthus giganteus L. PlantCell Reports, 11,7-10.

Lenee, P. & Chupeau, Y. (1986). Isolation and cultureof sunflower protoplasts (Helianthus annuus L.):Factors influencing the viability of cell coloniesderived from protoplasts. Plant Science, 43, 69-75.

Lepetit, M., Ehling, M., Gigot, C. & Hahne, G.

148 HAHNE

(1991). An internal standard improves the reliabilityof transient expression studies in plant protoplasts.Plant Cell Reports, 10, 401-405.

Lupi, M. C , Bennici, A., Locci, F. & Gennai, D.(1987). Plantlet formation from callus and shoot-tipculture of Helianthus annuus (L.). Plant Cell, Tissueand Organ Culture, 11, 47-55.

Malone-Schoneberg, J. B., Bidney, D., Scelonge, C ,Burrus, M. & Martich, J. (1991). Recovery of stabletransformants from Agrobacterium tumefaciens treatedsplit shoot axes of Helianthus annuus. In VitroCellular and Developmental Biology, 27, 152A.

Matzke, M. A., Susani, M., Binns, A. N., Lewis,E. D., Rubenstein, I. & Matzke, A. J. M. (1984).Transcription of a zein gene introduced intosunflower using a Ti plasmid vector. EMBO Journal,3, 1525-1531.

McCabe, D. E., Martinell, B. J. & Christou, P. (1988).Stable transformation of soybean (Glycine max)plants. Bio/Technology, 87, 923-926.

Mezzaroba, A. & Jonard, R. (1986). Effets du stade deprelevement et des pretraitements sur ledeveloppement in vitro d'antheres prelevees sur letournesol cultive (Helianthus annuus L.). ComptesRendus de VAcademie des Sciences, Paris, Serie III,tome 303, 181-186.

Moyne, A. L., Tagu, D., Thor, V., Bergounioux, C ,Freyssinet, G. & Gadal, P. (1989). Transformedcalli obtained by direct gene transfer into sunflowerprotoplasts. Plant Cell Reports, 8, 97-100.

Moyne, A. L., Thor, V., Pelissier, B., Bergounioux,C , Freyssinet, G. & Gadal, P. (1988). Callus andembryoid formation from protoplasts of Helianthusannuus. Plant Cell Reports, 1, 437-440.

Nataraja, K. & Ganapathi, T. R. (1989). In vitroplantlet regeneration from cotyledons of Helianthusannuus cv. Morden (sunflower). Indian Journal ofExperimental Botany, 27, 777-779.

Negrutiu, I., Shillito, R., Potrykus, J., Biasini, G. &Sala, F. (1988). Hybrid genes in the analysis oftransformation conditions. Plant Molecular Biology,8, 363-373.

Neuhaus, G., Spangenberg, G., Mittelsten-Scheid, O.& Schweiger, H. G. (1987). Transgenic rapeseedplants obtained by the microinjection of DNA intomicrospore-derived embryoids. Theoretical andApplied Genetics, 75, 30-36.

Nutter, R., Everett, N., Pierce, D., Panganiban, L.,Okubara, P., Lachmansingh, R., Mascarenhas, D.,Welch, H., Mettler, I., Pomeroy, L., Johnson, J. &Howard, J. (1987). Factors affecting the level ofkanamycin resistance in transformed sunflower cells.Plant Physiology, 84, 1185-1192.

Paterson, K. E. (1984). Shoot tip culture of Helianthusannuus - flowering and development of adventitiousand multiple shoots. American Journal of Botany, 71,925-931.

Paterson, K. E. & Everett, N. P. (1985). Regenerationof Helianthus annuus inbred plants from callus. PlantScience, 42, 125-132.

Pelissier, B., Bouchefra, O., Pepin, R. & Freyssinet, G.(1990). Production of isolated somatic embryosfrom sunflower thin cell layers. Plant Cell Reports, 9,47-50.

Piubello, S. M. & Caso, O. H. (1986). In vitro cultureof sunflower (Helianthus annuus L.) tissues. Phyton,46, 131-137.

Potrykus, I. (1991). Gene transfer to plants:assessment of published approaches and results.Annual Reviews of Plant Physiology and PlantMolecular Biology, 42, 205-225.

Power, C. J. (1987). Organogenesis from Helianthusannuus inbreds and hybrids from the cotyledons ofzygotic embryos. American Journal of Botany, 74,497-503.

Prado, E. & Berville, A. (1990). Induction of somaticembryo development by liquid culture in sunflower(Helianthus annuus L.). Plant Science, 67, 73-82.

Robinson, K. E. P. & Adams, D. O. (1987). The roleof ethylene in the regeneration of Helianthus annuus(sunflower) plants from callus. PhysiologiaPlantarum, 71, 151-156.

Roseland, C , Espinasse, A. & Grosz, T. J. (1991).Somaclonal variants of sunflower with modifiedcoumarin expression under stress. Euphytica, 54,183-190.

Schmitz, P. & Schnabl, H. (1989). Regeneration andevacuolation of protoplasts from mesophyll,hypocotyl, and petioles from Helianthus annuus L.Journal of Plant Physiology, 135, 223-227.

Schrammeijer, B., Sijmons, P. C , van den Elzen,P. J. M. & Hoekema, A. (1990). Meristemtransformation of sunflower via Agrobacterium. PlantCell Reports, 9, 55-60.

Steeves, T. A., Hicks, M. A., Naylor, J. M. & Rennie,P. (1969). Analytical studies on the shoot apex ofHelianthus annuus. Canadian Journal of Botany, 47,1367-1375.

Ursic, D. (1985). Eight DNA insertion events ofAgrobacterium tumefaciens Ti-plasmids in isogenicsunflower genomes are all distinct. Biochemical andBiophysical Research Communications, 131, 152-159.

van der Leede-Plegt, L. M., van der Ven, B. C. E.,Bino, R. J., van der Salm, T. P. M. & van Tunen,A. J. (1992). Introduction and differential use ofvarious promoters in pollen grains of Nicotianaglutinosa and Lilium longiflorum. Plant Cell Reports,11, 20-24.

Vancanneyt, G., Schmidt, R., O'Connor-Sanchez, A.,Willmitzer, L. & Rocha-Sosa, M. (1990).Construction of an intron-containing marker gene:splicing of the intron in transgenic plants and its usein monitoring early events in Agrobacterium-mediated plant transformation. Molecular and

Sunflower 149

General Genetics, 220, 245-250.Wilcox McCann, A., Cooley, G. & van Dreser, J.

(1988). A system for routine plantlet regeneration ofsunflower {Helianthus annuus L.) from immatureembryo-derived callus. Plant Cell, Tissue and OrganCulture, 14, 103-110.

Witrzens, B., Scowcroft, W. R., Downes, R. W. &Larkin, P. J. (1988). Tissue culture and plant

regeneration from sunflower {Helianthus annuus) andinterspecific hybrids {H. tuberosus X H. annuus).Plant Cell, Tissue and Organ Culture, 13, 61-76.

Yao, X., Jingfen, J. & Kuochang, C. (1988). Transferand expression of the T-DNA harboured byAgrobacterium tumefaciens in cultured explants ofHelianthus annuus. Acta Botanica Yunnanica, 10,159-166.

11Forest Trees

Ronald R. Sederoff

Introduction

Forest trees dominate the temperate and tropicalecosystems and the wood produced by trees is themost abundant biological material on the earth'ssurface (Gammie, 1981). Wood provides fuel formuch of the population of the world, and wood isa leading industrial raw material. In the UnitedStates, wood accounts for 25% of the value of allindustrial materials and exceeds them in amount(National Research Council, USA, 1990).Increased environmental concern requires thatmore forest is preserved and that less cuttingbe carried out in natural forests. To meetthe increased demand for wood, more woodneeds to be produced on less land throughgenetic improvement and intensive management(Buonjiorno & Grosenick, 1977; USDA ForestService, 1982).

Genetic improvement of forest trees is a slowprocess. The most advanced tree-breeding pro-grams have carried out only a few generations ofselection. Forest trees are largely undomesticatedplants and differ from many agronomic crops thathave been cultivated and selected for thousands ofyears. Furthermore, trees are not annual crops,and one generation for many species can take 20years or more.

Genetic engineering has great promise for agri-culture because it can accelerate traditional breed-ing, bypass long generation times and crossreproductive barriers through gene transfer tech-nology. If this is true for agricultural crops, thengenetic engineering should be a greater advantageto forestry because traditional methods have beenslower and more difficult. The application of

genetic engineering technology to tree improve-ment has been slow also because the necessarytechniques and information are not available.Woody plants have been neglected as objects ofinterest in molecular biology and little is knownabout the molecular basis of biological processesthat are unique to perennial woody plants, such asdormancy, wood formation, or the transition fromjuvenility to maturity (Hutchison & Greenwood,1991). There is a need to acquire more basic bio-logical information about forest trees to addressmany practical goals through genetic engineering.

DNA transfer and tree improvement

There are three general types of genetic modifica-tions for accelerated tree improvement throughgenetic engineering. These are (i) improved resis-tance to pests and pathogens (biotic stress),(ii) modified metabolism and development, and(iii) resistance to different kinds of abiotic stress.

Woody plants are characterized by high levels ofgenetic diversity. High levels of variation aremaintained within species and within populations.High diversity is associated with large geographicranges, outcrossing breeding systems, and broadseed dispersal, but much of the variation remainsunexplained (Hamrick, Godt & Sherman-Broyles,1992). As perennial woody plants, forest trees aresubjected to long-term stress from pests andpathogens. Woody perennials differ from annualcrops because resistance must be maintained dur-ing different seasons over many years. Such stressis most extreme for long-lived forest trees, whereselection occurs over tens or even hundreds ofyears. Transformation may aid in solving major

150

Forest trees 151

problems of epidemic diseases in trees, such asthose caused by white pine blister rust(Cronartium ribicola Fischer), Dutch elm disease(Ceratocystis ulmi (Buism.) C. Mor.), or chestnutblight (Endothia parasitica). Insect pests, whethergenerally endemic, such as Southern pine barkbeetle (Dendroctonus frontalis Zimmermann), orepidemic, such as gypsy moth (Lymantria disparL.), are a serious concern in forest ecology andmanagement. Strategies for genetic engineeringusing Bacillus thuringiensis or proteinase inhibitorscould lead to reduced damage or to reduced use ofchemicals in the natural environment (Strauss,Howe & Goldfarb, 1991). Damage from pests andpathogens are major factors reducing forest pro-ductivity around the world.

The second main category of traits forimproved trees affect the development or metabo-lism of forest trees. Those of most immediateinterest affect growth and yield. Growth and mor-phology are under moderate to strong geneticcontrol, and accelerated improvement for yieldand quality would be valuable (Zobel & Talbert,1984). Control of fertility is an important traitbecause it could lead to controlled breeding, bio-logical containment, and increased vegetativegrowth. Control of growth by photoperiod and thetransition from juvenile to mature growth areinherited traits that have major effects on growthand quality of forest trees.

The third general category of traits for geneticengineering of forest trees are those affecting abi-otic stress. Trees must resist a variety of physicaland chemical stresses over long periods of time.The ability of trees to respond to temperature,flooding, drought, atmospheric pollution andglobal climate change are all factors related totraits that may be targets for genetic engineering.Recently, trees have been considered in combina-tion with directed genetic modifications for bio-remediation (Stomp etal., 1992).

Forest trees are different from domesticatedcrops

The transformation systems for forest trees havesome special requirements, compared to domesti-cated crops. Due to high levels of genetic diver-sity, different plants within a species vary greatlyin their ability to regenerate in culture and in theirresponse to pathogens such as Agrobacterium(Clapham et a/., 1990; Riemenschneider, 1990;

Bergmann, 1992; A.-M. Stomp, personal com-munication). Considerable effort is oftenexpended in screening for the specific genotypethat will perform appropriately, particularlybecause there are no cultivars or inbred lines. Insome species, clones are available, or genotype iscontrolled in part, using open pollinated progenyfrom a known seed tree. An ideal transformationsystem would be relatively insensitive to variationin genotype.

Transformation studies in most forest trees aremade difficult by the long reproductive cycles. It isnot possible to go through several reproductivegenerations of segregation to sort out mosaics orgenetic instabilities after transformation. Veri-fication of transformation depends on cell culturesystems and vegetative propagation. Thisapproach has been used successfully in walnut(Juglans regia L.) transformation, where trans-formed plants were produced after repetitivesomatic embryogenesis (McGranahan et ah, 1988,1990). Embryos, cocultivated with Agrobacteriumtumefaciens, were multiplied by repetitive embryo-genesis, and selected for kanamycin resistance insubsequent embryonic generations. Selectedembryo subclones were germinated and grown asplants or micropropagated.

A third limitation is the problem of juvenilemature correlation. Many traits cannot be readilypredicted for a mature tree based on the perfor-mance or phenotype of the juvenile tree. Woodproperties, for example, are significantly differentin juvenile and mature trees (Zobel & vanBuijtenen, 1989). Genetic engineering of maturetraits would be greatly facilitated by good juvenilemature correlations.

There are two major barriers to the applicationof genetic engineering to forest trees. The first isthe lack of information about the biological mech-anisms and the genetic basis of processes that aretargets for genetic engineering. The second is theabsence of adequate methods for DNA transfer invirtually any forest tree species, although notableprogress has been made recently with a fewspecies.

The purpose of this chapter is to review the cur-rent status of DNA transfer technology in foresttrees. We also address DNA transfer in the con-text of specific problems of interest in woodyplants and some specific objectives related togenetic engineering of forest trees. Some aspectsof genetic engineering of forest trees have recently

152 SEDEROFF

been reviewed (Neale & Kinlaw, 1992). Manyaspects of DNA transfer and genetic engineeringof trees have also been discussed in other articles(see von Arnold, Clapham & Ekberg, 1990;Charest & Michel, 1991; Ahuja & Libby, 1993;Sederoff& Stomp, 1993).

Transformation in conifers

The most widely planted forest trees throughoutthe temperate latitudes are conifers. Among thepredominant species are loblolly pine (Pinus taedaL.) in the southeastern United States, Douglas fir(Pseudotsuga menziesii (Mirb) Franco) in thenorthwestern United States, and Monterey pine(Pinus radiata D. Don) in the southern hemi-sphere, grown as an exotic. In the more northernlatitudes, spruces (such as Norway spruce; Piceaabies (L.) Karst), are important. Most conifers aregrown for structural wood or for pulp and paper.In spite of their significant commercial value,conifers have only recently become material forexperiments in DNA transformation. In the last 7years, transfer and expression of foreign genes incells and tissues of coniferous species have beenclearly established. The majority of the evidencehas come from two different technical approaches,one using Agrobacterium as a biological vector forDNA transfer and the other using microprojectilebombardment for physical transfer of DNA.

Agrobacterium-mediated DNA transfer inconifers

Before 1985, the only evidence for DNA transferin conifers were morphological criteria for infec-tion of conifers by the crown gall bacterium (DeCleene & De Ley, 1976, 1981). It is widelybelieved that Agrobacterium-induced crown galldisease is limited to angiosperms, particularlydicotyledons (dicots). Reports in the literature ofinfection of conifers by Agrobacterium are foundfrom the 1930s (Smith, 1935a,6). A wide surveyshows 56% of tested species of gymnosperms aresusceptible, whereas 58% of angiosperms are sus-ceptible (De Cleene & De Ley, 1976). At least 66different conifers are susceptible to Agrobacterium(for a review, see Sederoff & Stomp, 1993).

Subsequently, transformation in conifers wasestablished at a cellular or subcellular level.Evidence for transformation was obtained from

expression of genes transferred by Agrobacterium,including those coding for opine synthesis, phyto-hormone synthesis, and expression of kanamycinresistance. By 1990, such evidence had beenobtained in nine pine species and Douglas fir(Sederoff et al, 1986; Dandekar et al, 1987;Gupta, Dandekar & Durzan, 1988; Morris, Castle& Morris, 1988, 1989; Stomp et al, 1988, 1990;Ellis et al, 1989; Loopstra, Stomp & Sederoff,1990). Similarly, transformation had beenreported for four spruce species (Clapham &Ekberg, 1988; Ellis et al, 1989; Morris et al,1989), two true firs (Clapham & Ekberg, 1986,1988; Morris et al, 1988) and western hemlock(Tsuga heterophylla (Raf.) Sarg.) (Morris et al,1988). Studies showing expression of markergenes for p-glucuronidase, chloramphenicolacetyltransferase, luciferase, and neomycin phos-photransferase have been successful for loblollypine, Monterey pine, Scots pine (Pinus sylvestrisL.), Douglas fir, jack pine (Pinus banksianaLamb.), white spruce (Picea glauca (Moench)Voss), black spruce (Picea mariana (Mill.) B.S.P.)and hybrid larch (Bekkaoui et al, 1988; Gupta etal, 1988; Dunstan, 1989; Tautorus et al, 1989;Wilson, Thorpe & Moloney, 1989; Duchesne &Charest, 1991; Stomp, Weissinger & Sederoff,1991).

In sugar pine (P. lambertiana Dougl.), stabletransformation has been obtained in culturedcells. Sugar pine shoots derived from cytokinin-treated cotyledons were inoculated with an hyper-virulent strain of Agrobacterium tumefacienscontaining a binary plasmid system (Stomp et al,1988; Loopstra et al, 1990). Galls produced onshoots proliferated as callus in culture in theabsence of phytohormone, and in the presence ofkanamycin. Stable transformation was verifiedafter more than a year in culture by Southernblots and by NPT II activity. Similar results havebeen obtained for Douglas fir. Phytohormone-independent callus has been grown from tumorsinduced in culture on shoots established in vitro(Dandekar et al, 1987). Ellis et al (1989) andMorris et al (1989) have confirmed and extendedevidence for physical transfer and expression ofintroduced genes in this species.

In the genus Picea several species of sprucehave also been transformed by Agrobacteriumincluding Norway spruce (P. abies), Sitka spruce(P. sitchensis (Bong.) Carr.), white spruce,Englemann spruce (P. englemanni Parry ex

Forest trees 153

Engelm.) (Clapham & Ekberg, 1986, 1988; Elliset aL, 1989, 1991a; Hood et al, 1989, 1990;Clapham et aL, 1990). Stably transformed calluscultures established following Agrobacterium infec-tion have been obtained and characterized forNorway spruce (Hood et al., 1990) and for whitespruce (Ellis etal., 1989). Hood etal. (1990) inoc-ulated shoots of Norway spruce and then culturedthe resulting tumors in vitro. Transformation wasverified by Southern blotting and by opine synthe-sis.

Induction of the Agrobacterium virulencecascade in Douglas fir

Many strains of Agrobacterium that have highlevels of virulence in dicotyledonous plants do notwork well in conifers (Morris et aL, 1989; Stomp etaL, 1990). The native inducers of the Agro-bacterium virulence cascade induced by dicots areacetosyringone and hydroxyacetosyringone,although many other phenolic and hydroxy-phenylpropanoid compounds can act as inducers.In Douglas fir, Morris & Morris (1990) identifiedthe phenylpropanoid glucoside coniferin as themajor native phenolic virulence gene inducer.Coniferin is known to be a major soluble phenoliccomponent of conifer xylem. It is considered to bea storage form of coniferyl alcohol, the major pre-cursor of guaiacyl lignin (Sederoff & Chang,1991). In developing xylem of white spruce,scraped from wood after peeling the cambium andphloem, coniferin was found at levels of more than4% of the wet weight (Savidge, 1989). It is pre-sumed that coniferin is located in the vacuolesbecause it is present at such high levels.

Transformation by bombardment

Microprojectile bombardment provides a methodfor studying transient expression in forest trees,particularly in systems that are either recalcitrantor have not been explored. Bombardments resultin localized staining usually in single cells, and canpenetrate tissue below the epidermis (Stomp et al.,1991). Many forest tree species have been studiedwith this technique, e.g. Douglas fir (Goldfarb etaL, 1991); Populus (McCown et aL, 1991); blackspruce (Duchesne & Charest, 1991).

An abscisic acid (ABA)-inducible promoterfrom wheat has been found to be particularlyactive in conifer cells and tissues. The promoter

from the early methionine (Em) gene fused togusA (Marcotte, Russell & Quatrano, 1989) hasbeen used in bombardment experiments on a vari-ety of conifer cells and tissues. These includeNorway spruce somatic embryos (Robertson et aL,1992), differentiating xylem (Loopstra,Weissinger & Sederoff, 1992), loblolly pinecotyledons (Stomp et aL, 1990) and black sprucecallus (Duchesne & Charest, 1991). ABA is knownto be important in the regulation of embryonicdevelopment of conifers and has been shown toaffect development of somatic embryos of Norwayspruce and other spruce species (Hakman & vonArnold, 1985; Hakman & Fowke, 1987; vonArnold & Woodward, 1988). Several promotersfrom dicots have been tested in conifers, but noneis as active in conifers as the ABA-inducible Empromoter. Ellis etal. (1991a, 1992) observed activ-ity for cauliflower mosaic virus (CaMV) 35S,nopaline synthase (nos), Em from wheat, a ubiqui-tin promoter from Arabidopsis, Rubisco (ribulose-1,5-bisphosphate carboxylase oxygenase) (rbc)promoters from Arabidopsis, soybean and larch(Larix), a phosphoenol pyruvate (PEP)-carboxy-lase promoter, an alcohol dehydrogenase promoterfrom maize, and an auxin-inducible promoterfrom soybean. All showed activity after bombard-ment, which decreased within days. A heat shockpromoter from soybean could be induced andexpressed for up to 4 weeks after bombardment.

Transient expression of a promoter from theDcS gene from carrot (Daucus) has been reportedto occur following bombardment of Norwayspruce cell and embryo cultures (Newton et aL,1992). DcS is an ABA-responsive gene expressedabundantly and late during carrot embryogenesis(Hatzopoulos, Fong & Sung, 1990). Expressionwas observed in suspension cells, but less expres-sion was observed in embryogenic callus orsomatic embryos. Expression in embryos or sus-pension cells was enhanced if cultures were pre-treated with ABA. In black spruce, gusA wasexpressed by both the CaMV 35S promoter andthe Em promoter from wheat. The Em promoterappeared far more active (Duchesne & Charest,1991). In a subsequent study, Duchesne &Charest (1992) found a similar result withembryogenic lines of hybrid larch (Larix sp.) andshowed that expression increased following addi-tion of ABA to the culture medium.

Microprojectile bombardment has been used inNorway spruce somatic embryos to generate

154 SEDEROFF

stable transformation in cell lines (Robertson etal, 1992). Mature somatic embryos were bom-barded with pRT99gus, a plasmid that containsthe nptll and gusA genes both fused to theCaMV 35S promoter. Assays of cell lines selectedon antibiotic show that all of the eight lines testedshow coexpression of both marker genes, but atvariable levels. None of the transformed linesremained embryogenic.

Of particular interest in forest trees is the use ofmicroprojectile bombardment to observe expres-sion of promoters in differentiating xylem. Treesare a good system in which to study the differenti-ation of wood, and wood formation is an impor-tant target for genetic engineering (Whetten &Sederoff, 1991) because of the potential for modi-fication of wood properties. Loopstra et al (1992)have studied the expression of three promoters inloblolly pine stem sections (Em from wheat,CaMV 35S, and nos). Expression of the gusAreporter gene could be detected in different celltypes, including tracheids, ray parenchyma, andaxial parenchyma associated with resin canals.Again, the Em promoter was the most efficient.

Transformation in hardwoods

Hardwoods, the general term for dicot treespecies, include a much larger and more diversenumber of species than is found among theconifers (softwoods). Wood properties are morevariable in hardwoods, and different hardwoodsare used for a larger diversity of products (Zobel &van Buijtenen, 1989). Hardwoods are used forpulp, paper and structural material. In addition,hardwoods are a major source of fuel around theworld, and are used for a great variety of specialtyproducts. The major cultivated hardwood foresttrees are in the genus Eucalyptus. Many other treesare cultivated to a lesser extent but have featuresthat make them promising candidates for geneticengineering. Our discussion here is focused ontrees grown for wood and wood products. Manyimportant tree species are being investigated forother purposes, such as trees being geneticallymodified for improved rubber production(Kitayama et al, 1990) or for horticultural cropssuch as apple (James et al, 1989) or walnut(McGranahan^ra/., 1990).

Transgenic conifers

Perhaps the most significant recent advance intransformation of forest trees has been the pro-duction of the first transgenic conifers. Transgenicplants have been produced in white spruce andlarch. Ellis et al (19916, 1992, 1993) were able toregenerate stably transformed plants of whitespruce following microprojectile bombardment.Late stages of somatic embryos were bombardedand selected using a sublethal level of kanamycin.Cell lines were obtained that were able to formdifferentiated embryos. Transformed seedlingsexpressed genes for gusA and Bacillus thuringiensisendotoxin (Ellis et al, 1993). European larch(Larix decidua) was transformed by Agrobacteriumrhizogenes (Huang, Diner & Karnosky, 1991;Huang & Karnosky, 1991). Agrobacterium rhizo-genes induced hairy roots and/or adventitiousshoot buds from transformed cells. Adventitiousbuds were excised, cultured, rooted and trans-ferred to the greenhouse as plantlets. Trans-formation was verified by opine synthesis, andSouthern blot hybridization. The next steps willbe the development of routine protocols for trans-formation in these species and other commerciallyimportant conifers.

Populus as a model system

The genus Populusy which includes the aspens andthe poplars, has been increasing in importance as acommercial species. Poplars, particularly hybridpoplars, have fast growth, good form, short rota-tion times, and are easily propagated vegetatively.They are grown in Europe and several northernregions of the USA. Poplars have become a modelforest tree for molecular genetics owing to thesmall size of their genome (Arumuganathan &Earle, 1991), efficient regeneration in culture(Ahuja, 1986; McCown et al, 1988) and ability tobe transformed by Agrobacterium (Parsons et al,1986; Fillatti et al, 1987; Pythoud et al, 1987; DeBlock, 1990). Poplars are well suited for funda-mental studies of tree physiology, and as a modelsystem for genetic engineering of woody plants.Wood from poplars and aspens is used for a varietyof processed wood products such as pulp, paper,plywood, hardboard, and packing materials, inaddition to furniture (Burns & Honkala, 1990).

A major problem in Populus transformation isthe variation within the genus for transformationand regeneration. Some sections of the genus arereadily transformed and regenerated, whereasothers remain recalcitrant. Often the species of

Forest trees 155

interest for practical applications are poorly trans-formed, particularly some of the fast growingcommercial hybrids. There are many advantagesto the Populus species for basic research and itremains one of the most useful species for molecu-lar genetics of forest trees.

Populus trichocarpa X P. deltoides (clone H l l ) aclone with significant economic potential, hasbeen transformed with a genetically modifiedstrain of A. rhizogenes (Parsons et al, 1986;Pythoud et al, 1987). A plasmid containing thevir region of the supervirulent plasmid pTiBo542dramatically increased the infectivity. Trans-formation was verified by Southern blots andregeneration of shoots containing the gene forNPTII was also reported. Poplar cells trans-formed with these strains grow in the absence ofadded growth regulators, and synthesize strain-specific opines. The highest frequency of transfor-mation was obtained with small greenwood stemsections that were cocultivated with bacteria andtransferred to medium lacking hormones. Trans-formed cells gave rise to rapidly growing tumorsthat proliferated indefinitely.

Wang et al (1990) reported transformation andregeneration from leaf explants of Populus tomen-tosa Carr. DNA transfer was indicated by expres-sion of the transferred DNA (T-DNA) gene 4 andexpression of chloramphenicol acetyltransferase.An alternative method for transformation ofpoplar (P. tremula L. X P. alba L.) by has beenreported where stem internodes were inoculatedwith a suspension of two strains of Agrobacterium.A wild-type strain that induces shoot differentia-tion in tumors and a disarmed strain carrying abinary vector were mixed together and used toinoculate stems. Shoots were obtained fromtumors that expressed both nptll andgusA derivedfrom the binary strain (Brasiliero et al., 1991).Subsequent efforts have produced transgenicplants with additional constructs in as little as 3months (Leple etal, 1992).

Herbicide resistant Populus

The first transgenic forest tree was a poplar carry-ing glyphosate tolerance. Fillatti et al (1987) usedan oncogenic binary system of A. tumefaciens(C58/587/85) containing a wild-type C58 plas-mid, and an engineered plasmid (pPMG587/85),to infect leaf discs of a hybrid aspen cloneNC5339 (P. alba L. X P. grandidentata Michx.).

Transformation was obtained by cocultivating leafsegments with Agrobacterium and a feeder layer oftobacco suspension cells (Horsch et al, 1985).The regenerated transformed plants carried anactive kanamycin resistance gene and an alteredaroA gene for resistance to the herbicideglyphosate (Comai et al, 1985). The aroA genecodes for 5-enolpyruvylshikimate synthase, anenzyme active in the synthesis of aromatic aminoacids. Transformation was also confirmed bySouthern and Western blotting.

The transformed poplar plants were resistant toglyphosate at levels of 0.07 kg/ha, providing fur-ther evidence that the gene for resistance isexpressed and functional (Fillatti et al, 1988).The level of tolerance, although clearly differentfrom that of untransformed controls, is not at thelevel that would be needed for commercial appli-cation (Riemenschneider et al, 1988). It may beexpected that higher levels of activity will beobtained with new constructs. These experimentsare a landmark in forest biotechnology becausethey represent the first example of a tree modifiedby genetic engineering.

De Block (1990) examined factors affectingtransformation in a hybrid aspen (clone 357;Populus alba X P. tremula) and a hybrid poplar(clone 064; Populus trichocarpa Torr. & Gray X P.deltoides Bartr. ex Marsh). Shoot-tip necrosiseffects produced in culture were overcome bymodification of the nitrate/ammonia ratio of themedium (De Block, 1990). Efficient transforma-tion was obtained using a disarmed Agrobacteriumtumefaciens strain carrying chimeric genes forresistance to phosphotricine (glyphosinate) resis-tance (De Block et al, 1987). The enzyme phos-photricin acetyltransferase (PAT) acetylatesphosphotricine and inactivates it. The gene bar,codes for PAT. Glyphosinate inhibits glutaminesynthetase resulting in lethal accumulation ofammonium (Murakami et al, 1986). Trees trans-formed with the bar gene were normal in appear-ance, and were completely resistant to field levelpreparations of the herbicide (Botterman et al,1991). A similar study has been carried out toconfer resistance to Basta where A. rhizogenes wasused to introduce the gene into a hybrid clone (P.tremula X P. alba). Regenerated plants were toler-ant to the herbicide (Devillard, 1992).

156 SEDEROFF

Transformation of Populus for insectresistance

Populus is a preferred target of the gypsy moth, themajor insect pest of hardwoods in North America(Montgomery & Wallner, 1988). Insect resistanceis, therefore, an important trait for geneticimprovement in Populus. Most studies of transfor-mation in Populus use Agrobacterium to transferDNA to plant cells; however, McGown et al.(1991) used direct gene transfer with a particlegun (electric discharge). In their system, three dif-ferent target tissues (protoplast-derived cells,organized cultured nodules, and stem explantswere bombarded. Two unrelated hybrid Populusgenotypes were tested (P. alba X P. grandidentata'Crandon' and P. nigra 'Beautifolia' X P. tricho-carpa). Four plants were recovered from the moreeasily manipulated genotype (P. alba X P. grandi-dentata) that were transformed for wos-NPT II,CaMV 35S-GUS, and 35S-BT (modified endo-toxin gene from Bacillus thuringiensis). One of theplants was highly resistant to feeding of two lepi-dopteran pests, the forest tent caterpillar (Mala-cosoma disstria Hubner) and the gypsy moth(Lymantria dispar L.). Transformation by bom-bardment has an advantage over Agrobacterium^where transformation is limited by host pathogenincompatibility. Recovery of transformed plantsdepended upon pretreatment of target tissues,optimization of bombardment parameters, andthe use of a selection technique employing flood-ing of the target tissues.

The gene coding for potato proteinase inhibitorII (pin2) is one of the best-characterized plantdefense genes (Ryan, 1990), and acts throughproduction of a small proteinaceous inhibitor ofanimal digestive enzymes. The inhibitors accumu-late in foliage after insect attack or mechanicalwounding. The pin2 gene is inducible in potatoand retains inducibility when transferred totobacco (Thornburg et al> 1987). The wound-inducible pin2 promoter from potato, in a genefusion with cat (gene for chloramphenicol amino-transferase), was transferred into hybrid poplarwhere it was still wound inducible (Klopfenstein eta/., 1991). The poplar hybrid was P. alba L. X P.grandidentata Michx. clone Hansen, which is anatural hybrid originating in southeast Iowa.Agrobacterium tumefaciens transformation was car-ried out using a binary miniplasmid pRT45, inarmed A281 containing the pin2-cat gene fusion.

Northern and Southern blot analysis verified thattransformation of single gene copies had occurredand that expression of pin2 was wound inducible.

Transformation in Eucalyptus

One of the world's most important genera of for-est trees is Eucalyptus. Eucalypts are grown forpulp, paper and wood on several continents andare the most important tree for wood products intropical forestry. Some hybrids of Eucalyptus (e.g.E. grandis Hile ex Maiden X E. camaldulensisDehn.) are used in Brazil for charcoal (Zobel &van Buijtenen, 1989). In intensively managedplantations, hybrids of Eucalyptus (e.g. grandis Xurophylla S.T. Blake) are grown on a 6 year rota-tion (Zobel & Talbert, 1984). Transformation inthis group of forest trees could have a significanteffect on the practice of plantation forestry in thetropical and subtropical regions of the world.

At least one species of Eucalyptus (E. tereticornisSm.) has been reported to be susceptible toAgrobacterium (Jindal & Bhardwaj, 1986). Anotherreport of transformation of E. gunniiH. by A. rhi-zogenes, was based on phenotypic effects in culture(Adam, 1986). Transient gene expression hasbeen reported in E. citrodora Hook, using electro-poration of protoplasts (Manders etal, 1992) andin E. gunnii (Teulieres et al., 1991; Teulieres,Leborgne & Boudet, 1991). In E. gunnii, transientexpression of gusA was observed in protoplastsderived from callus or suspension culture afterelectroporation or treatment with PEG. Transientexpression was also observed after electroporationof intact cells.

Transformation in energy and biomassspecies

Several forest tree species have been a focus ofattention for biomass production for fuel. A pro-gram in the US has focused on five plants for shortrotation woody crops (SRWC) for energy. InPopulus and sweetgum (Liquidambar styracifluaL.), efficient transformation and regeneration oftransformed plants have been achieved. In theother species of interest for SRWC, black locust(Robinia pseudoacacia L.), sycamore (Platanusoccidentalis L.), and silver maple (Acer saccariniumL.), less work has been done (Harry & Sederoff,1989).

In the southeastern USA, Liquidambar styraci-

Forest trees 157

flua (American sweetgum) is a rapidly growinghigh quality hardwood that is used as an ornamen-tal species and grown for high quality paper(Kormanik, 1990). It has the potential to beimproved by genetic engineering for insect resis-tance, lignin content or cold hardiness (Sullivan &Lagrimini, 1992). Two groups have indepen-dently used Agrobacterium tumefaciens to transformand regenerate sweetgum. Chen (1991) has devel-oped a root nodule system (Z.-Z. Chen & A.-M.Stomp, personal communication) that has a highcapacity for shoot regeneration. Cocultivation ofleaf pieces with Agrobacterium gives rise to trans-formed cell lines from which plants can be regen-erated (Chen, 1991). Sullivan & Lagrimini (1992)introduced genes for insect resistance, enhancedperoxidase activity, and reduced peroxidase activ-ity. In both laboratories, transformation was veri-fied by molecular hybridization or expression ofreporter genes.

Black locust is susceptible to A. tumefaciens andA. rhizogenes (Davis & Keathley, 1989).Inoculation in vitro produced tumors that werephytohormone independent. All tumors had seg-ments of T-DNA, and callus obtained from infec-tion of A281 carrying pGA472 as a binary waskanamycin resistant and showed integration ofnptll DNA by Southern analysis.

In Sweden, willows, Salix (species) have been afocus of attention for fuel production. Salix growsrapidly in high latitudes and is readily propagated.Salix clones are being tested for use as an energy-generating tree species as one component of anational alternative energy program to reducedependence on nuclear power. Willows readilycoppice and are useful for high yield cropping sys-tems. Vahala, Stabel & Eriksson (1989) trans-formed Salix stem explants with A. tumefaciensstrains C58 and GV3101 and observed growth onhormone-free medium containing kanamycin.Transformation was confirmed by molecularhybridization and by opine synthesis. In addition,transformation has been observed in leaf discs ofwillow that produced callus and roots (Rocha &Maynard, 1990). Hauth & Beiderbeck (1992) usedA, rhizogenes to infect roots of Salix alba L. andestablished cultures of hairy roots after infection.

Transformation in yellow poplar

Yellow poplar (Liriodendron tulipifera L.) is nativeto the central eastern USA and Canada. The tree

is characterized by good form, high quality woodand rapid growth. It is sometimes considered tobe one of the most important hardwood speciesin the USA (Merkle & Sommer, 1991). Somaticembryogenesis of yellow poplar was firstreported in 1986 (Merkle & Sommer, 1986).Low frequencies of conversion of embryos toplantlets limited the usefulness of this system fortransformation studies. Methods for conversionof embryos at higher frequency were developed(Merkle et al, 1990). Populations of synchro-nized mature somatic embryos were obtainedfrom proembryonic masses cultured in liquidmedium and selected for size on stainless steelsieves. Mature embryos transferred to basalmedium produced plantlets.

Embryogenic suspension cultures of yellowpoplar have been transformed for the gusA geneusing either direct DNA uptake into protoplasts ormicroprojectile bombardment of intact cells(Wilde, Meagher, & Merkle, 1991). In subse-quent studies, transgenic plants were producedfrom transformed suspension cells (Wilde et al>1992). Plasmid DNA containing genes for GUSand NPT II was introduced by microprojectilebombardment and mature plants were regener-ated. Although, work on transformation of yellowpoplar is recent, it has significant promise as amodel tree species because of the excellent systemfor regeneration. It will be difficult to apply thesystem to different genotypes because somaticembryonic cultures must be produced from differ-entiating zygotic embryos.

Nitrogen fixation and genetic engineering inforest trees

Actinorhizal plants can establish symbiotic rela-tionships with the actinomycete Frankia, andthereby fix nitrogen in root nodules (Normand &Lalonde, 1986). Alnus (alder) is an actinorhizalforest tree found in high latitudes that could beuseful for land reclamation and reforestation. Thecapacity of alder (Alnus incana; Alnus glutinosa L.Gaertn.) for transformation by A. tumefaciens hasbeen demonstrated (Mackay, Seguin & Lalonde,1988). Tumor formation, integration of T-DNAand expression of opine synthesis were found,using standard strains of ACH5 or C58. Electro-poration of gusA driven by the 35S promoter ofCaMV into protoplasts of Alnus incana showedtransient expression. Expression was maximal at 1

158 SEDEROFF

to 2 days and declined by day 4 (Seguin &Lalonde, 1988).

Betula (birch) is closely related to Alnus butdoes not have a symbiotic system for nitrogen fixa-tion. Genes involved in symbiotic nitrogen fixa-tion in Alnus, if transferred to birch, might conferthe ability to establish nitrogen fixation (Normand& Lalonde, 1986). Agrobacterium tumefaciens wasused to produce tumors in four clones of paperbirch (Betula papyrifera Marsh.). Transformationwas verified by Southern blotting, and opine syn-thesis (Mackay et al> 1988).

Reforestation on nitrogen-deficient soils is animportant problem in tropical forestry and sub-tropical forestry where soil degradation hasoccurred. Trees that fix nitrogen are also impor-tant for agroforestry. Species of Casuarinaceaethat are nodulated by the actinomycete Frankiacan also fix atmospheric nitrogen and can grow onnitrogen-deficient soils. The family includes somehigh yielding, fast growing trees with a high valuefor firewood. Allocasuarina verticillata Lam., amember of this family of nitrogen-fixing trees, hasbeen transformed using Agrobacterium rhizogenes(Phelep et al.) 1991). Hairy root formation wasinduced, followed by induced or spontaneousshoot formation. Shoots were rooted to producetransgenic plants with extensive ageotropic rootsystems. Transformation has been verified bySouthern blot analysis and by the detection of spe-cific opines. Phelep et al envisioned introducinggenes that could contribute to pest or herbicideresistance for trees that can fix nitrogen as agro-nomic crops.

Some other trees of interest

Bolyard, Hajela & Sticklen (1991) have initiatedtransformation experiments in the hybrid pioneerelm (JJlmus 'Pioneer'). The purpose of theseexperiments is the eventual introduction of genesthat could enhance resistance to the fungus thatcauses Dutch elm disease. Both microprojectilebombardment and Agrobacterium infection of stemsections resulted in transformed callus, on thebasis of Southern blotting and fluorogenic detec-tion of reporter gene activity.

Oaks are a large and diverse group of trees thathave not been intensively studied at the molecularlevel. Initial attempts to transform Quercus roburL. are promising. Nodal stem explants appear tobe transformed by A. rhizogenes 9402 or A. tume-

faciens A281, both containing gusA (Roest et al>1991). Sterile root clones were established andshowed GUS activity.

Neem (Azadirachta indica) is an extremely ver-satile tropical tree that is used for veneer, firewoodand as a shade tree. The leaves of neem are alsoused for their insecticidal properties. Neem hasbeen transformed and regenerated (Naina, Gupta& Mascarenhas, 1989). Agrobacterium tumefaciensstrains containing a recombinant derivative ofpTiA6 were used to infect seedlings grown in vitro.Induced tumors gave rise to transformed callus.Plantlets were obtained from the callus and frominduced shoots. Transformation was indicated byautotrophic growth of callus, synthesis of octopineand resistance to kanamycin.

Barriers to progress

Although several species will provide good modelsystems for the production of transgenic plants,many of the most commercially important species,for example pines, remain recalcitrant to transfor-mation. The barriers to transformation and regen-eration rest, in significant part, in limitations ofthe tissue culture systems. The central problem isthe introduction of DNA into cells that will becapable of regeneration. In some cases, the cellsthat can be transformed are not the cells that canregenerate. It is essential that cells can survive thestress of the transformation protocols, whether thestress is biological, resulting from Agrobacteriuminfection, or physical, resulting from bombard-ment. In either case, cells that are transformedmust proliferate, survive selection, and regenerateto form plantlets through organogenesis orembryogenesis. Culture systems can be very slowfor many forest tree species (von Arnold &Woodward, 1988; Tautorus, Fowke & Dunstan,1991).

Many difficulties are simply problems of effi-ciency, where several steps in the process eachoccur at low frequency. It may be that successrequires only increases in frequency of rare events.Although much discussion has taken place regard-ing the relative merits of Agrobacterium transfor-mation and bombardment, both methods havebeen successful in specific systems and the choiceof method depends upon the biology of the sys-tem. Improvements continue in marker systems,vectors and transformation efficiency. Screening

Forest trees 159

and definition of genetic variation for cultureparameters could have significant effects.Development of somatic embryogenesis forconifer species, for example, has often requiredscreening for rare genotypes that could proliferateand differentiate in culture.

Societal implications

Finally, the ecological and social implications ofgenetic engineering of forest trees are significant.Special interest groups in the public and privatesectors are concerned about the potential threat tothe value and stability of our forests posed bygenetic engineering. Their concerns result fromconflicting interests and the economic pressures toharvest trees from ecologically sensitive stands orfrom recreation areas. Opposition to genetic engi-neering of trees exists in part because of a pre-sumed threat to the natural forests. However, theincreasing economic pressure for wood and woodproducts may require use of intensive manage-ment of plantations with genetically engineeredtrees to produce enough wood efficiently onsmaller areas of land. Another area of concern liesin the potential introduction of foreign genes intothe natural populations. Methods to survey geneflow into natural populations need to be devel-oped as well as methods to restrict potential geneflow. It may be necessary to engineer reproductivesterility into such clones to contain foreign geneswithin managed programs.

References

Adam, S. (1986). Obtention de racines transformeeschez Eucalyptus gunnii H. par Agrobacteriumrhizogenes. Annales de Recherches Sylvicoles, pp. 7-21.

Ahuja, M. R. (1986). Aspen. In Handbook of Plant CellCulture, vol. 4, ed. D. A. Evans, W. R. Sharp & P. V.Ammirato, pp. 626-651. Macmillan, New York.

Ahuja, M. R. & Libby, W. J. (1993). ClonalForestry.Vol. I Genetics and Biotechnology. Springer-Verlag,Berlin, Heidelberg.

Arumuganathan, K. & Earle, E. (1991). Nuclear DNAcontent of some important plant species. PlantMolecular Biology Reporter, 9, 208-218.

Bekkaoui, F., Pilon, M., Laine, E., Raju, D. S. S.,Crosby, W. L. & Dunstan, D. I. (1988). Transientgene expression in electroporated Picea glaucaprotoplasts. Plant Cell Reports, 7, 481-484.

Bergmann, B. (1992). Genotype influence onadventitious organogenesis and susceptibility toAgrobacterium tumefaciens in Pinus. Ph.D.dissertation, North Carolina State University.

Bolyard, M. G., Hajela, R. K. & Sticklen, M. B.(1991). Microprojectile and Agrobacterium-mediztedtransformation of pioneer elm. Journal ofArboriculture, 17, 34-37.

Botterman, J., D'Halluin, K., De Block, M., De Greef,W. & Leemans, J. (1991). Engineering of glufosinateresistance and evaluation under field conditions:phosphinothricin-acetyltransferase gene expressionin transgenic tobacco, tomato, potato, poplar,alfalfa, rape and sugarbeet to confer herbicideresistance. In Herbicide Resistance in Weeds and Crops,11th LongAshton International Symposium,pp. 355-363. Butterworth-Heinemann, New York.

Brasileiro, A. C. M., Leple, J. C , Muzzin, J.,Ounnoughi, D., Michel, M. F. & Jouanin, L.(1991). An alternative approach for gene transfer intrees using wild-type Agrobacterium strains. PlantMolecular Biology, 3, 441-452.

Buonjiorno, J. & Grosenick, G. L. (1977). Impact ofworld economic and demographic growth on forestproducts consumption and wood requirements.Canadian Journal of Forest Research, 1, 392-399.

Burns, R. M. & Honkala, B. H. (1990). Silvics of NorthAmerica, Vol. 2 Hardwoods. Forest Service, U. S.Department of Agriculture Handbook 654.

Charest, P. J. & Michel, M. F. (1991). Basics of plantgenetic engineering and its potential applications totree species. Petawawa National Forestry Institute,Chalk River, Ontario. Information Report No. PI-X-104.

Chen, Z.-Z. (1991). Nodular culture andAgrobacterium mediated transformation fortransgenic plant production in Liquidambarstyraciflua L.(sweetgum). Ph.D. dissertation, NorthCarolina State University.

Clapham, D. & Ekberg, 1. (1986). Induction oftumors by various strains of Agrobacteriumtumefaciens on Abies nordmanniana and Picea abies.Scandinavian Journal of Forest Research, 1, 435-437.

Clapham, D. H. & Ekberg, I. (1988). Induction oftumors by various strains of Agrobacteriumtumefaciens on Abies nordmanniana and Picea abies. InGenetic Manipulation of Woody Plants, ed. H. W.Hanover & D. E. Keathley, p. 463. Plenum Press,New York.

Clapham, D. H., Ekberg, I., Eriksson, G., Hood, E. E.& Norell, L. (1990). Within population variation insusceptibility to Agrobacterium tumefaciens A281 inPicea abies (L.) Karst. Theoretical and AppliedGenetics, 79, 654-656.

Comai, L., Facciotti, D., Hiatt, W. R., Thompson, G.,Rose, R. & Stalker, D. (1985). Expression in plantsof a mutant aroA gene from Salmonella typhimurium

160 SEDEROFF

confers tolerance to glyphosate. Nature (London),317, 741-744.

Dandekar, A. M., Gupta, P. K., Durzan, D. J. &Knauf, V. (1987). Transformation and foreign geneexpression in micropropagated Douglas-fir(Pseudotsuga menziesii). Bio/Technology, 5, 587-590.

Davis, J. M. & Keathley, D. E. (1989). Detection andanalysis of T-DNA in crown gall tumors andkanamycin resistant callus of Robinia pseudoacacia.Canadian Journal of Forest Research, 19, 1118-1123.

De Block, M. (1990). Factors influencing the tissueculture and the Agrobacterium-mediatedtransformation of hybrid aspen and poplar clones.Plant Physiology, 93, 1110-1116.

De Block, M., Botterman, J., Vandewiele, M., Dockx,J., Thoen, C , Gossele, V., Rao Mowa, N.,Thompson, C , Van Montagu, M. & Leemans, J.(1987). Engineering herbicide resistance in plants byexpression of a detoxifying enzyme. EMBO Journal,6,2513-2518.

De Cleene, M. & De Ley, J. (1976). The host range ofcrown gall. Botany Review, 42, 389-466.

De Cleene, M. & De Ley, J. (1981). The host range ofinfectious hairy-root. Botany Review, 47, 147-194.

Devillard, C. (1992). Transformation in vitro dutremble (Populus tremula X Populus alba) parAgrobacterium rhizogenes et regeneration de plantestolerant au basta. Compte Rendu de VAcademie desSciences. Serie 3, Sciences de la Vie, 314, 291-298.

Duchesne, L. C. & Charest, P. J. (1991). Transientexpression of the beta-glucuronidase gene inembryogenic callus of Picea mariana followingmicroprojection-tungsten biolistic microprojectile-mediated DNA transfer. Plant Cell Reports, 10,191-194.

Duchesne, L. C. & Charest, P. J. (1992). Effect ofpromoter sequence on transient expression of thebeta-glucuronidase gene in embryogenic calli ofLarix X eurolepis and Picea mariana followingmicroprojection-hybrid larch and black spruce callusculture transformation by biolistic microprojectilemethod. Canadian Journal of Botany, 70, 175-180.

Dunstan, D. I. (1989). Transient gene expression inelectroporated conifer protoplasts: electroporation ofwhite spruce, black spruce and jack pine protoplasts.In Vitro, 25, Pt.2, 63A.

Ellis, D., Roberts, D., Sutton, B., Lazaroff, W., Webb,D. & Flinn, B. (1989). Transformation of whitespruce and other conifer species by Agrobacteriumtumefaciens. Plant Cell Reports, 8, 16-20.

Ellis, D. D., McCabe, D., Russell, D., Martinell, B. &McCown, B. H. (1991a). Expression of inducibleangiosperm promoters in a gymnosperm, Piceaglauca (white spruce). Plant Molecular Biology, 17,19-27.

Ellis, D. D., McCabe, D., Mclnnis, S., Martinell, B. &McCown, B. H. (19916). Transformation of white

spruce by electrical discharge particle acceleration.In Application of Biotechnology to Tree Culture,Protection and Utilization, ed. B. E. Hassig, T. K.Kirk, W. L. Olsen, K. F. Raffa & J. M. Slavicek.USDA Forest Service, Columbus, OH.

Ellis, D. D., McCabe, D., Mclnnis, S.,Ramachandran, R. & McCown, B. (1992).Transformation of Picea glauca-gene expression andthe regeneration of stably transformed embryogeniccallus and plants. International Conifer BiotechnologyWorking Group, abstract no. 29.

Ellis, D. D., McCabe, D. E., Mclnnis, S.,Ramachandran, R., Russell, D. R., Wallace, K. M.,Martinell, B. J., Roberts, D. R., Raffa, K. F. &McCown, B. H. (1993). Stable transformation ofPicea glauca by particle acceleration. Bio/Technology,11, 84-89.

Fillatti, J. J., Hassig, B., McCown, B., Comai, L. &Riemenschneider, D. (1988). Development ofglyphosate tolerant Populus plants throughexpression of a mutant aroA gene from Salmonellatyphimurium. In Genetic Manipulation of WoodyPlants, ed. J. Hanover & D. Keathley, pp. 243-250.Plenum Press, New York.

Fillatti, J. J., Selmer, J., McCown, B., Hassig, B. &Comai, L. (1987). Agrobacterium mediatedtransformation and regeneration of Populus.Molecular and General Genetics, 206, 192-199.

Gammie, J. (1981). World timber to the year 2000.The Economist Intelligence Unit Special Report, No. 98,Economist Intelligence Unit Ltd, London.

Goldfarb, B., Strauss, S. H., Howe, G. T. & Zaerr,J. B. (1991). Transient gene expression ofmicroprojectile-introduced DNA in Douglas fircotyledons. Plant Cell Reports, 10, 517-521.

Gupta, P. K., Dandekar, A. M. & Durzan, D. J.(1988). Somatic preembryo formation andtransient expression of a luciferase gene in Douglasfir and loblolly pine protoplasts. Plant Science, 58,85-92.

Hakman, I. & Fowke, L. C. (1987). Somaticembryogenesis in Picea glauca (white spruce) andPicea mariana (black spruce). Canadian Journal ofBotany, 65, 656-659.

Hakman, I. & von Arnold, S. (1985). Plantletregeneration through somatic embryogenesis inPicea abies. Journal of Plant Physiology, 121, 149-158.

Hamrick, J. L., Godt, M. J. W. & Sherman-Broyles,S. L. (1992). Factors influencing levels of geneticdiversity in woody plant species. New Forests, 6,95-124.

Harry, D. E. & Sederoff, R. R. (1989). Biotechnology inBiomass Crop Production: The Relationship ofBiomassProduction and Genetic Engineering. Oak RidgeNational Laboratory, Environment SciencesDivision of Publications, No. 3411.

Hatzopoulos, P., Fong, F. & Sung, R. (1990). Abscisic

Forest trees 161

acid regulation of Dc8, a carrot embryonic gene.Plant Physiology, 94, 690-695.

Hauth, S. & Beiderbeck, R. (1992) In vitro culture ofAgrobacterium induced hairy roots of Salix alba L.Silvae Genetica, 41, 46-48.

Hood, E. E., Clapham, D. H., Ekberg, I. &Johannson, T. (1990). T-DNA presence and opineproduction in tumors of Picea abies (L.) Karstinduced by Agrobacterium tumefaciens A281. PlantMolecular Biology, 14, 111-117.

Hood, E. E., Ekberg, I., Johannson, T. & Clapham,D. H. (1989). T-DNA presence and opineproduction in Agrobacterium tumefaciens A281induced tumors on Norway spruce. Journal ofCellular Biochemistry, Suppl. 13 Part D, 260.

Horsch, R. B., Fry, J. E., Hofrman, N. L., Eichholz,D., Rogers, S. G. & Fraley, R. T. (1985). A simpleand general method for transferring genes intoplants. Science, 227, 1229-1231.

Huang. Y., Diner, A. M. & Karnosky, D. F. (1991).Agrobacterium rhizogenes-mediated genetictransformation and regeneration of a conifer: Larixdecidua. In Vitro Cellular and Developmental Biology,4, 201-207.

Huang, Y. & Kamosky, D. F. (1991). A system forgymnosperm transformation and plant regeneration:Agrobacterium rhizogenes and Larix deadi/a-Europeanlarch hairy root culture and propagation. In Vitro,27, 153A.

Hutchison, K. W. & Greenwood, M. S. (1991).Molecular approaches to gene expression duringconifer development and maturation. Forest Ecologyand Management, 43, 273-286.

James, D. J., Passey, A. J., Barbara, D. J. & Bevan, M.(1989). Genetic transformation of apple (Maluspumila Mill.) using a disarmed Ti-binary vector.Plant Cell Reports, 7, 658-661.

Jindal, K. K. & Bhardwaj, L. N. (1986). Occurrence ofcrown gall on Eucalyptus tereticornis Sm. in India.Indian Forester, 112, 1121.

Kitayama, M., Takahashi, M., Surzycki, S. &Togasaki, R. (1990). Transformation of callus tissuefrom Hevea brasillensis and Jasminium officinale. PlantPhysiology 93, 1, Suppl.46.

Klopfenstein, N. B., Shi, N. Q., Kernan, A., McNabb,H. S., Jr, Hall, R. B., Hart, E. R. & Thornburg,R. W. (1991). Transgenic Populus hybrid expresses awound inducible potato proteinase inhibitor II-CATgene fusion. Canadian Journal of Forest Research, 21,1321-1328.

Kormanik, P. P. (1990). Liquidambar styraciflua,Sweetgum. In Silvics of North America, ed. R. M.Burns & B. H. Honkala, Technical Coordinators.USDA Forest Service Publication No. 654, pp.400-405.

Leple, J. C , Brasileiro, A. C. M., Michel, M. F.,Delmotte, F. & Jouanin, L. (1992). Transgenic

poplars: expression of chimeric genes using fourdifferent constructs. Plant Cell Reports, 11, 137-141.

Loopstra, C. A., Stomp, A.-M. & Sederoff, R. R.(1990). Agrobacterium mediated DNA transfer insugar pine. Plant Molecular Biology, 15, 1-9.

Loopstra, C. A., Weissinger, A. K. & Sederoff, R. R.(1992). Transient expression in differentiatingwood. Canadian Journal of Forest Research. 22,993-996.

Mackay, J., Seguin, A. & Lalonde, M. (1988). Genetictransformation of 9 in vitro clones oiAlnus andBetula by Agrobacterium tumefaciens. Plant CellReports, 7, 229-232.

Manders, G., Dossantos, A. V. P., Vaz, F. B. D.,Davey, M. R. & Power, J. B. (1992). Transient geneexpression in electroporated protoplasts ofEucalyptus citriodora Hook. Plant Cell Tissue andOrgan Culture, 30, 69-75.

Marcotte, W. R. Jr, Russell, S. H. & Quatrano, R. S.(1989). Abscisic acid-responsive sequences from theEm gene of wheat. Plant Cell, 1, 969-976.

McCown, B. H., McCabe, D. E., Russell, D. R.,Robison, D. J., Barton, K. A. & Raffa, K. F. (1991).Stable transformation of Populus and incorporationof pest resistance by electric discharge particleacceleration. Plant Cell Reports, 9, 590-594.

McCown, B. H., Zeldin, E. L., Pinkalla, H. A. &Dedolph, R. R. (1988). Nodule culture: adevelopmental pathway with high potential forregeneration, automated micropropagation, andplant metabolite production from woody plants. InGenetic Manipulation of Woody Plants, ed. J. W.Hanover & D. E. Keathley, pp. 149-166. PlenumPress, New York.

McGranahan, G. H., Leslie, C. A., Uratsu, S. L. &Dandekar, A. (1990). Improved efficiency of thewalnut somatic embryo gene transfer system. PlantCell Reports, 8, 512-516.

McGranahan, G. H., Leslie, C. A., Uratsu, S. L.,Martin, L. A. & Dandekar, A. M. (1988).Agrobacterium-mediated transformation of walnutsomatic embryos and regeneration of transgenicplants. Bio/Technology, 6, 800-804.

Merkle, S. A. & Sommer, H. E. (1986). Somaticembryogenesis in tissue cultures of Liriodendrontulipifera. Canadian Journal of Forest Research, 16,420-422.

Merkle, S. A. & Sommer, H. E. (1991). Yellow-poplar(Liriodendron spp.). In Biotechnology in Agricultureand Forestry, Vol 16 Trees III, ed. Y. P. S. Bajaj,pp. 94-110. Springer-Verlag, Berlin, Heidelberg.

Merkle, S. A., Wiecko, A. T., Sotak, R. J. & Sommer,H. E. (1990). Maturation and conversion ofLiriodendron tulipefera somatic embryos. In VitroCellular and Developmental Biology, 26, 1086-1093.

Montgomery, M. E. & Wallner, W. E. (1988). Thegypsy moth: a westward migrant. In Dynamics of

162 SEDEROFF

Forest Insect Populations, ed. A. A. Berryman, pp.353-375. Plenum Press, New York.

Morris, J. W., Castle, L. A. & Morris, R. O. (1988).Transformation of pinaceous gymnosperms byAgrobacterium. In Genetic Manipulation of WoodyPlants, ed. J. W. Hanover & D. E. Keathley, p. 481.Plenum Press, New York.

Morris, J. W., Castle, L. A. & Morris, R. O. (1989).Efficacy of different Agrobacterium tumefaciens strainsin transformation of pinaceous gymnosperms.Physiological and Molecular Plant Pathology, 34,451-462.

Morris, J. W. & Morris, R. O. (1990). Identification ofan Agrobacterium tumefaciens virulence gene inducerfrom the pinaceous gymnosperm Pseudotsugamenziesii. Proceedings of the National Academy ofSciences, USA, 87, 3614-3618.

Murakami, T., Anzai, H., Imai, S., Satoh, A.,Nagaoka, K. & Thompson, C. J. (1986). Thebialiphos biosynthetic genes of Streptomyceshygroscopicus: molecular cloning and characterizationof the gene cluster. Molecular and General Genetics,205, 42-50.

Naina, N. S., Gupta, P. K. & Mascarenhas, A. F.(1989). Genetic transformation and regeneration oftransgenic neem (Azadirachta indica) plants usingAgrobacterium tumefaciens. Current Science, 58,184-187.

National Research Council, USA (1990). ForestryResearch: A Mandate for Change. A report from thecommittee on Forestry Research, Commission onLife Sciences, Board on Biology & Board onAgriculture.

Neale, D. B. & Kinlaw, C. S. (1992). Forestbiotechnology. Forest Ecology and Management, 43,179-324.

Newton, R. J., Yibrah, H. S., Dong, N., Clapham,D. H. & von Arnold, S. (1992). Expression of anabscisic acid responsive promoter in Picea abies (L.)Karst. following bombardment from an electricaldischarge particle accelerator-microprojectileparticle acceleration application to Norway sprucetissue culture and cell culture transformation. PlantCell Reports, 11, 188-191.

Normand, P. & Lalonde, M. (1986). The genetics ofactinorhizal Frankia, a review. In Third InternationalSymposium on Nitrogen Fixation with Nonlegumes,Plant Soil, 90, 429-454.

Parsons, T. J., Sinkar, V. P., Stettler, R. F., Nester,E. W. & Gordon, M. P. (1986). Transformation ofpoplar by Agrobacterium tumefaciens. Bio I Technology,4, 533-536.

Phelep, M., Petit, A., Martin, L., Douhoux, E. &Tempe, J. (1991). Transformation and regenerationof a nitrogen fixing tree, Allocasuarina verticillataLam. Bio/Technology, 9, 461-466.

Pythoud, F., Sinkar, V. P., Nester, E. W. & Gordon,

M. P. (1987). Increased virulence of Agrobacteriumrhizogenes conferred by the vir region of pTiBo542:application to genetic engineering of poplar.Bio/Technology, 5, 1323-1327.

Riemenschneider, D. E. (1990). Susceptibility of intraand interspecific hybrid poplars to Agrobacteriumtumefaciens C58. Phytopathology, 80, 1099-1102.

Riemenschneider, D. E., Haissig, B. E., Sellmer, J. &Fillatti, J. J. (1988). Expression of an herbicidetolerance gene in young plants of a transgenic hybridpoplar clone. In Somatic Cell Genetics of WoodyPlants, ed. M. R. Ahuja, pp. 73-80. KluwerAcademic Publishers, Dordrecht.

Robertson, D., Weissinger, A. K., Ackley, R., Glover,S. & Sederoff, R. R. (1992). Genetic transformationof Norway spruce (Picea abies (L.) Karst) usingsomatic embryo explants by microprojectilebombardment. Plant Molecular Biology, 19, 925-935.

Rocha, S. P. & Maynard, C. A. (1990). In vitromultiplication and Agrobacterium transformation ofwillow (Salix lucida). In Vitro Cellular andDevelopmental Biology, 26, 43A (abstract 111).

Roest, S., Brueren, H. G. M. J., Evers, P. W. &Vermeer, E. (1991). Agrobacterium mediatedtransformation of oak (Quercus roburL.)-propagation. Acta Horticulture, 289, 259-260.

Ryan, C. A. (1990). Protease inhibitors in plants:genes for improving defenses against insects andpathogens. Annual Reviews of Phytopathology, 28,425-449.

Savidge, R. A. (1989). Coniferin, a biochemicalindicator of commitment to tracheid differentiationin conifers. Canadian Journal of Botany, 67,2663-2668.

Sederoff, R. & Chang, H.-M. (1991) Ligninbiosynthesis. In Wood Structure and Composition, ed.M. Lewin & I. Goldstein, pp. 263-285. MarcelDekker, Inc., New York.

Sederoff, R. & Stomp, A.-M. (1993). DNA transfer inconifers. In Clonal Forestry I: Genetics andBiotechnology ed. M. R. Ahuja & W. J. Libby,pp. 241-254. Springer-Verlag, Berlin, Heidelberg.

Sederoff, R. R., Stomp, A.-M., Chilton, W. S. &Moore, L. W. (1986). Gene transfer into loblollypine by Agrobacterium tumefaciens. Bio/Technology, 4,647-649.

Seguin, A. & Lalonde, M. (1988). Gene transfer byelectroporation in Betulaceae protoplasts: Alnusincana. Plant Cell Reports, 7, 367-370.

Smith, C. O. (1935a). Crown gall on conifers.Phytopathology, 25, 894.

Smith, C. O. (19356). Crown gall on the Sequoia.Phytopathology, 25, 439-440.

Stomp, A.-M., Han, K.-H., Perkins, E. J. & Gordon,M. P. (1992). The use of genetically engineeredtrees and plants for environmental restoration andwaste management. In Proceedings of the Waste

Forest trees 163

Management and Environmental Sciences Conference,San Juan, Puerto Rico, pp. 498-503. Department ofEnergy Publication.

Stomp, A.-M., Loopstra, C , Chilton, W. S., Sederoff,R. R. & Moore, L. W. (1990). Extended host rangeof Agrobacterium tumefaciens in the genus Pinus. PlantPhysiology, 92, 1226-1232.

Stomp, A.-M., Loopstra, C , Sederoff, R. R., Chilton,S., Fillatti, J., Dupper, G., Tedeschi, P. & Kinlaw,C. (1988). Development of a DNA transfer systemfor pines. In The Genetic Manipulation of WoodyPlants, Basic Life Sciences, vol. 44, ed. J. W.Hanover & D. E. Keathley, pp. 231-241. PlenumPress, New York.

Stomp, A.-M., Weissinger, A. K. & Sederoff, R. R.(1991). Transient expression from microprojectile-mediated DNA transfer in Pinus taeda. Plant CellReports, 10, 187-190.

Strauss, S. H., Howe, G. T. & Goldfarb, B. (1991).Prospects for genetic engineering of insect resistancein forest trees. Forest Ecology and Management, 43,181-209.

Sullivan, J. & Lagrimini, L. M. (1992).Transformation of Liquidambar styraciflua. PlantPhysiology, 99, 1, Suppl. 46.

Tautorus, T. E., Bekkaoui, F., Pilon, M., Dalta,R. S. S., Crosby, W. L., Fowke, L. C. & Dunstan,D. (1989). Factors affecting transient geneexpression in electroporated black spruce {Piceamariana) and jack pine (Pinus banksiana) protoplasts.Theoretical and Applied Genetics, 78, 531-536.

Tautorus, T. E., Fowke, L. C. & Dunstan, D. I.(1991). Somatic embryogenesis in coniferpropagation; a review. Canadian Journal of Botany,69, 1873-1899.

Teulieres, C , Grima-Pettenati, J., Curie, C , Teissie,J. & Boudet, A. M. (1991). Transient foreign geneexpression in polyethylene glycol treated orelectropulsated Eucalyptus gunnii protoplasts. PlantCell Tissue and Organ Culture, 25, 125-132.

Teulieres, C , Leborgne, N. & Boudet, A. M. (1991).Direct gene transfer procedures for Eucalyptusgenetic transformation: Eucalyptus protoplasttransformation using polyethylene glycol orelectroporation for beta-glucuronidase reporter geneexpression in culture. In Vitro, 28, 121 A.

Thornburg, R. W., An, G., Cleveland, T. E., Johnson,R. & Ryan, C. A. (1987). Wound-inducibleexpression of a potato inhibitor II-chloramphenicolacetyltransferase gene fusion in transgenic tobaccoplants. Proceedings of the National Academy ofSciences, USA, 84, 744-748.

USDA Forest Service (1982). An analysis of thetimber situation in the United States, 1952-2030.Forest Resources Report No. 23. US GovernmentPrinting Office, Washington, DC.

Vahala, T., Stabel, P. & Eriksson, T. (1989). Genetictransformation of willows (Salix spp.) byAgrobacterium tumefaciens. Plant Cell Reports, 8,55-58.

von Arnold, S., Clapham, D. & Ekberg, I. (1990). Hasbiotechnology a future in forest tree breeding? ForestTree Improvement, 23, 31-48.

von Arnold, S. & Woodward, T. (1988).Organogenesis and embryogenesis in mature zygoticembryos of Picea sitchensis. Tree Physiology, 4,291-300.

Wang, S. P., Xu, Z. H. & Wei, Z. M. (1990). Genetictransformation of leaf explants of Populus tomentosa.Acta Botanica Sinica, 3, 172-177.

Whetten, R. & Sederoff, R. (1991). Geneticengineering of wood. Forest Ecology and Management,43,301-316.

Wilde, H. D., Meagher, R. B. & Merkle, S. A. (1991).Transfer of foreign genes into yellow-poplar(Liriodendron tulipifera). In Woody PlantBiotechnology, ed. M. R. Ahuja, pp. 227-232.Plenum Press, New York.

Wilde, H. D., Meagher, R. B. & Merkle, S. A. (1992).Expression of foreign genes in transgenic yellow-poplar plants. Plant Physiology, 98, 114-120.

Wilson, S. M., Thorpe, T. A. & Moloney, M. M.(1989). PEG-mediated expression of GUS andCAT genes in protoplasts from embryogenicsuspension cultures of Picea glauca. Plant CellReports, 7, 704-707.

Zobel, B. J. & Talbert, J. (1984). Applied Forest TreeImprovement. J. Wiley & Sons, New York.

Zobel, B. J. & van Buijtenen, J. P. (1989). WoodVariation: Its Causes and Control. Springer-Verlag,New York.

Index

Page numbers in italics refer to figures and tables

abscisic acid (ABA), 90abscisic acid (ABA)-inducible promoter, 153Ac transposase gene, 59-60acetosyringone, 106, 153Acinetobacter calcoaceticus transformation, 25acquired immunodeficiency syndrome (AIDS), 10adhl promoter, 58Agrobacterium, 23

clone bank use, 29electroporation, 26Eucalyptus transformation, 156gene vector for maize, 66pKT230 mobilization, 27plasmid rearrangement, 28plasmid replication region cloning, 24Populus insect resistance, 156replication of recombinant molecules, 24transferred DNA, 27transformation, 24-5vector stability, 28vector system, 23vector transgene inheritance, 82virulence cascade in Douglas fir, 153virulence inhibition, 66

Agrobacterium rhizogenesalfalfa transformation, 114, 115birdsfoot trefoil transformation, 115, 116, 117black locust transformation, 157broad bean transformation, 109dry bean transformation, 111-12legume transformation, 119, 120nitrogen-fixing tree transformation, 158pea transformation, 107Populus transformation, 155rapeseed transformation, 126-7sanfoin transformation, 118soybean transformation, 103

strain K599, 103TL-DNA, 127

Agrobacterium tumefaciensalfalfa transformation, 112, 113, 115birdsfoot trefoil transformation, 115-16, 117black locust transformation, 157broad bean transformation, 109cowpea transformation, 110, 111dry bean transformation, 111legume transformation, 118, 120lentil transformation, 110Lotononis bainesii transformation, 117nitrogen-fixing tree transformation, 157-8nopaline strains, 130octopine strains, 130pea transformation, 106, 107peanut transformation, 112Populus transformation, 155, 156rapeseed transformation, 127, 130-1soybean transformation, 101-3, 105-6sunflower infection efficiency, 140sunflower transformation, 138, 143-4, 145sweetgum transformation, 157transformation vector, 65

Agrobacteriuw-based vectors, small-grain cereals, 89Agrobacterium-mediated transformation

conifers, 152-3transgenic maize, 65-6, 74

agroinfection, 66agropine, 114alcohol dehydrogenase (ADH), 58

anerobic induction of proteins, 58alder, 157alfalfa, 112-15

Agrobacterium rhizogenes transformation, 115Agrobacterium tumefaciens transformation, 112, 113,

115

164

Index 165

hairy root production, 114ovalbumin gene insertion, 114protoplast transformation technique, 115regeneration, 112somaclonal variation, 120transformation, 118, 119

alfalfa, transgenicagropine, 114Basta resistance, 120GUS activity, 115nopaline content, 114NPT II activity, 112, 113, 115ovalbumin expression, 120

Alnus, 157aminoethoxyvinyl-glycine (AVG), 73ampicillin, 11

resistance gene, 18antibiotic resistance marker transduction, 16antifungal genes, 120antisense gene expression, 134Arabidopsis thaliana, 53Arachis hypogaea, 112aroA gene, 155Asparagus officinale, 65Aspergillus

electroporation transformation, 36invertase gene cloning, 41protein secretion, 43-4

Aspergillus nidulans> 34autonomously replicating sequence incorporation,

39cloning complementing genes, 41cosmid library, 41gene replacement, 42homologous integration, 39introduced DNA sequence stability, 40

atrazine resistance, 120gene, 105

autonomously replicating sequences (ARS), 38, 39auxin, 130

regeneration induction, 139auxin : cytokinin ratio for sunflower transformation,

143auxotrophic markers, transduction, 16auxotrophic mutants, 37Azadirachta indica, 158Azospirillum, 23

electroporation, 26Ti plasmids, 27

Bacillus subtilis, PEG sensitivity, 13Bacillus thuringiensis

endotoxin, 120, 1548-endotoxin gene, 61forest trees, 151insect control proteins, 61

bacteriophage, transducing, 16

fcargene, 85, 91alfalfa transformation, 113herbicide resistance in navy bean, 112marker for small-grain cereal transformation, 93Populus herbicide resistance, 155selective agent for transgenic maize, 71see also bialaphos

barleya-amylase gene, 90electroporation of mesophyll protoplasts, 90foreign DNA stable integration, 90foreign protein production, 91protoplasts, 89-90reporter gene expression, 90stable transformation, 90-1transformation, 93transient gene expression, 89-90

barnase gene, 72Basta

resistance, 120see also phosphinothricin (PPT)

BCG, 10persisting immune responses, 19recombinant vaccine use, 19

Betula, 158bialaphos, 71

see also bar genebialaphos resistance

gene, 55navy bean, 111

biolistic transformation, 37bioremediation, forest trees, 151birch, 158birdsfoot trefoil, 115-17

Agrobacterium rhizogenes transformation, 115, 116,117

Agrobacterium tumefaciens transformation, 115-16,117

CAT activity, 116glutamine synthase gene, 120glutamine synthetase-w*dA gene fusion construct,

116hairy root production, 116, 121NPT II activity, 116opines, 117transformation, 118, 119

black locust transformation, 157Black Mexican Sweet (BMS) protoplast

transformation, 69Bradyrhizobium, 23

nitrogen fixation, 23transformation, 25

Brassica napus, 125microinjection technique, 67

broad bean, 109hairy root induction, 109NPT II activity, 109

166 Index

cab gene, 59CaCl2/heat-shock protocol, 5cadmium resistance plasmids, 15CaMV 19S promoter, Lotononis bainesii

transformation, 117CaMV 35S promoter, 58, 84, 85, 102

alfalfa transformation, 113birdsfoot trefoil transformation, 116conifers, 153, 154cowpea transformation, 110DNA containing, 104moth bean transformation, 109navy bean transformation, 112rapeseed, 131soybean transformation, 105Vicia narbonensis transformation, 109, 110

cat genebirdsfoot trefoil transformation, 115, 116Lotononis bainesii transformation, 117moth bean transformation, 108soybean transformation, 105

cell wall synthesis inhibitors, 11cellobiohydrolase II, 43cellulase

gene transformation, 43production, 43

cellulose, 126Ceratocystis ulmi> 151cereal plants

gene transfer techniques, 81protoplasts, 81

cereal plants, small-grain, 81artificial promoter, 84C residues at CpG dinucleotides, 86constitutive monocot promoters, 84CpG islands, 93DNA delivery methods, 85-9DNA macroinjection into floral tillers, 85-7DNA transfer via pollen tube pathway, 87-8electrophoretic migration of DNA, 89electroporation, 88, 89embryonic suspension cell cultures, 83foreign DNA, 86, 87gene expression requirements, 93illegitimate recombination, 84laser microbeam technique, 89marker genes, 82, 84-5mesocotyl-derived suspension cultures, 83microinjection, 89microprojectile-mediated gene transfer, 93particle gun technique, 83-4pEmu promoter, 85penetration markers, 82pollen-mediated indirect gene transfer, 88polyethylene glycol (PEG), 89promoters, 84-5protoplasts, 82-3, 93

regeneration of transformed cells to mature plants,92

reporter genes, 83, 86seed inbibition, 88silicon carbide fibers, 89stabilization of foreign genes, 82stable integration of DNA, 92stable transformant selection, 85stable transformation, 81suspension culture embryogenicity, 83transformation, 89-92transformation with Agrobacterium, 89transformation methods, 82-9, 93transgene inheritance, 82transient expression boosting, 84

cerulin, 11chestnut blight, 151chitinase gene, 120Chlorsulfuron, 71choline esters, 126chromosome domains, 75chymosin complementary DNA, 43Clavibacter flaccumfaciens, 15clone bank, 29cloning vectors, 17, 18

function removal, 26-7coat protein (CP) gene, 60-1ColEl plasmid, 27

host range, 4ColEl replicon, 18complementation analysis, 28coniferin, 153conifers

abscisic acid-inducible promoter, 153adventitious shoot buds, 154Agrobacterium-mediated transfer, 152-3callus growth, 152crown gall bacteria, 152microprojectile bombardment transformation,

153-4NPT II activity, 152pine family transformation, 152spruce transformation, 152, 153transformation, 152^4transgenic, 154xylem, 154

conjugal receptor function, 3conjugal transfer, DNA, 7conjugation

coryneform bacteria, 15-16genes, 4nocardioform bacteria, 15-16

conjugative plasmids, mutagenesis, 7Coprinus cinereus homologous integration, 39Coprinus lagopus lithium acetate transformation, 36Corynebacterium diphtheriae, 10Corynebacterium glutamicum transformation efficiency, 13

Index 167

Corynebacterium integration by homologousrecombination, 17

Corynebacterium xerosis, 10coryneform bacteria, 10-19

cloning systems, 10cloning vectors, 17, 18conjugation, 15-16cosmid-type vector, 16DNA introduction, 11, 12, 13, 14, 15-16electroporation, 13, 14, 15illegitimate recombination integration, 17-18integration of nonreplicative, 17lysozyme action sensitivity, 11PEG-mediated DNA uptake in

protoplasts/spheroplasts, 11, 12, 13plasmids, 16-17self-transmissible extrachromosomal elements,

15transduction, 16

cosmid library for Aspergillus nidulans, 41cosmid rescue, 41cotransformation

filamentous fungi, 38gene introduction into rice plants, 57

cowpea, 110-11GUS activity, 111transformation, 118, 119

cowpea mosaic virus (CPMV) mRNA, 110CpG islands, 93Cronartium ribicola, 151crop improvement

genetic engineering, 59genetic transformation, 53, 59

crop species, major, 53crown gall

bacteria, 152tumors, 138, 143

Cryptococcus neoformans, 36cytokinin, regeneration induction, 139

2,4-D, 115, 127Dactylis glomerata, 92Dc8 promoter, 153Dendroctonus frontalis, 15159 desaturase, 1342,4-diacetylphloroglucinol, 6dicotyledenous plants

gene expression regulation, 53transgene inheritance, 82

diphtheria, 10divalent cations, 3DNA

cloning vectors, 23-4, 27-8conjugal recipient entry, 3conjugal transfer, 7conjugation introduction into

nocardioform/coryneform bacteria, 15-16

direct transformation, 8electroporation, 13, 14, 15fate of transforming, 16-18imbibition into dry embryo, 88integron, 17linearized, 40microinjection techniques, 67PEG-mediated uptake in protoplasts/spheroplasts,

11, 12, 13stability of introduced sequences, 40transduction introduction into

nocardioform/coryneform bacteria, 16transformation efficiency, 13transformation introduction into

nocardioform/coryneform bacteria, 11, 12, 13, 14,15

uptake by protoplasts, 13uptake by spheroplasts, 13see also transforming DNA

DNA introduction into bacteria, 24-7applications, 28-9conjugation, 26-7electroporation, 25-6, 29nocardioform/coryneform, 11, 12, 13, 14, 15, 16plasmid rearrangement, 28transduction, 27transformation, 24-5

T-DNA see transferred DNADouglas fir

Agrobacterium virulence cascade, 153transformation, 152

dry bean, 111-12hairy root culture, 111transformation, 118, 119tumor induction. 111

Ds element with Ac transposase gene, 59-60Dutch elm disease, 151, 158

early methionine gene, 153EDTA, Pseudomonas exposure effects, 7electroduction, 26electroporation, 3, 8, 29

barley protoplasts, 90clone bank, 29DNA introduction into rice protoplasts, 55electric parameters, 15filamentous fungi, 26-7Gram-negative soil bacteria, 25-6high voltage of intact cells, 13, 14, 15moth bean transformation, 108pea transformation, 108rapeseed transformation, 132small-grain cereals, 88, 89sunflower protoplasts, 144transformation efficiency, 15

elm, hybrid pioneer, 158Em gene, 153, 154

168 Index

embryonic suspension cell culturesprotoplast transformation in rice, 54-5small-grain cereals, 83

endosperm, starch composition, 60Endothia parasitica, 151enhanced plasmid transformation efficiency (EPT)

phenotype, 155-enolpyruvylshikimate synthase, 155erucic acid, 126Erwinia, 23

electroporation, 26Escherichia coli, 3

DNA uptake, 25shuttle plasmids, 15-16transformation frequency, 26

ethylenediaminetetra-acetic acid see EDTAEucalyptus, 154

transformation, 156

fasciation, 10induction, 19

fasciation-inducing genes, 19Festuca arundinacea> 92filamentous fungi, 34-44

autonomously replicating vectors, 38-9cotransformation, 38dominant genes, 37, 38electroporation transformation, 26-7gene disruption, 42genetic manipulation, 34genetic purification of transformants, 38genomic integration of circular plasmids, 39homologous recombination, 39lithium acetate transformation, 36marker genes, 37one-step disruption, 42particle bombardment in transformation, 27plasmid integration patterns, 40polyethylene glycol-mediated transformation, 34-6protoplasts in transformation, 34self-cloning, 40-2transformant selection, 37-8transformation, 39

floral tillers, DNA macroinjection, 85-7forage legumes

alfalfa, 112-15birdsfoot trefoil, 115-17Lotononis bainesii) 117-18sanfoin, 118transformation success, 118-19

forest trees, 150-9abiotic stress, 151actinorhizal, 157Bacillus thuringiensis, 151biomass production, 156-7bioremediation, 151conifer transformation, 152-4

DNA transfer, 150-1, 152ecological impact of genetic engineering, 159energy production, 156epidemic disease, 151fertility control, 151fuel production, 156-7genetic engineering, 150, 151-2, 157-8genetic improvement, 150growth control, 151hardwoods, 154-8improvement, 150-1insect pests, 151juvenile mature correlation, 151long-term stresses, 150nitrogen fixation, 157-8photoperiod, 151plantlet formation, 158proteinase inhibitors, 151regeneration cells, 158reproductive cycle length, 151societal impact of genetic engineering, 159transformation efficiency, 158-9transformation systems, 151

fungi, transformation in soil species, 35

Gaeumannomyces graminis var. tritici, 1gene

constitutive monocot promoters, 84disruption, 42introduction of new desirable, 28-9regulatory sequences of introduced, 75

gene cloningcomplementation, 41-2cosmid rescue, 41

gene expressioncross-species, 43transient in wheat, 91

gene function analysis, 40-4gene disruption, 42gene replacement, 42titration of rraws-acting gene products, 42

gene replacement, 42homologous recombination, 28

genetic engineering, forest trees, 157-8germline transformation targets, transgenic maize,

66-7gibberellic acid (GA3), 90glaA gene, 43P-glucanase gene, 120glucoamylase, 43-4glucosinolate, 126P-glucuronidase (GUS)

activity in fiber-mediated transformation of maize,68

expression in transgenic rice plants, 58, 59gene, 86rice protoplast activity, 56-7

Index 169

sunflower transformation, 143glutamine synthase gene, 120glutamine synthetase-wzdA gene fusion construct, 116glycerol, 13glycine, 11, 13Glycine max, 101-8Glycine soya, 103glyphosinate, 155glyphosphate

resistance, 120selective agent for transgenic maize, 71

grain legumes, 101-12broad bean, 109dry bean, 111-12lentil, 110moth bean, 108-9pea, 106-8peanut, 112

Gram-negative soil bacteriabroad host range vectors, 23cloning vectors, 25DNA introduction, 23DNA replicator region direct cloning, 29exogenous DNA uptake, 25introduction of new desirable genes, 28-9pGV910 replication, 24pSA vectors, 24

grasses, wound response absence, 81guinea grass transformation, 92^wsAgene, 153, 154

Populus expression, 155yellow poplar transformation, 157

gypsy moth, 151Populus target, 156

hardwoodscultivated, 154transformation, 154-8

Helianthus annuus, 137Helianthus tuberosus, 137herbicide resistance

Populus, 155sunflower, 145

heterokaryons, 38HmR

calli, 55, 57phenotype, 59see also hygromycin

homologous recombination, integration by, 17host range

broad, 4narrow, 4, 5

hph gene, 55-6, 57, 92Agrobacterium tumefaciens transformation of pea,

106-7see also hygromycin

hsp60 expression signals, 19

hybridization, interspecific, 145hydrogen-autotrophic growth, 15hydrogen-auxotrophic growth, 16hydrogenase gene, 29hydroxyacetosyringone, 153hygromycin

alfalfa transformation, 113pea transformation, 106, 107phosphotransferase (HPT) gene, 89rapeseed transformation, 131resistance gene see hph geneselective agent for transgenic maize, 71Vicia narbonensis transformation, 109-10

ice nucleation active (Ina+) epiphytic colonist, 5immunosuppression, 10ina gene, 5IncP (incompatibility group P), 4

transfer, 16IncW plasmids, 4insect resistance

Populus, 156sunflower, 145

insecticidal genes, 120integrative vectors, 19integron, 17introns, gene expression enhancement, 56-7isonicotinic acid hydrazide (INH), 11isopentyltransferase, 19

Jerusalem artichoke, 137

kanamycinalfalfa transformation, 112, 113, 115dry bean transformation, 111Lotononis bainesii transformation, 117-18moth bean transformation, 108, 109pea transformation, 106, 107rapeseed transformation, 131resistance gene see nptll genesunflower regeneration, 142transgenic maize, 70-1

laser microbeam technique, 89leafy gall, 10leghemaglobin gene expression, 120legumes, 101

forage, 112-18grain, 101-12, 118hairy root formation, 119somaclonal variation, 120transformation success, 118-21

Lens culinaris, 110lentil, 110

transformation, 118, 119tumor induction, 110

leprosy, 10

170 Index

a-linoleic acid, 126Lotononis bainesii, 117-18

transformation, 118, 119Lotus corniculatus, 115-17luc gene, 85lucerne see alfalfaLymantria dispar, 151lysostaphin, 13lysozyme, 13

maizeAgrobacterium-based vectors, 89agronomic performance, 73alcohol dehydrogenase (ADH), 58DNA introduction, 65embryo reception of gene transfer, 67embryogenic cultures, 69endophytic bacteria, 67inbred elite, 72-3insecticidal protein synthetic gene, 73marker gene delivery of egg cell, 67microspore-derived cultures for protoplasts, 69pollen grain, 66protoplast culture, 68T-DNA, 66

maize, transgenic, 65Agrobacterium-mediated DNA delivery, 65-6Agrobacterium-mediated transformation, 74agroinfection, 66applicability to commercial germplasm, 72-3Black Mexican Sweet (BMS) cells, 68culture productivity degeneration, 72direct gene transfer, 68-70, 73-4electroporation, 72fertile, 75fertile plant production, 69fertility, 71-2friable embryogenic response, 73germline transformation targets, 66-7GUS activity, 68inheritance of transgene expression, 74-5kanamycin selection, 70-1microinjection, 67-8microprojectile bombardment, 55molecular characterization, 73-5nptll marker, 87particle bombardment, 70-2performance enhancement, 75phenotypic abnormality, 71-2protoplast transformation, 54, 68-70proven approaches, 68-72seed production, 71stresses, 71techniques, 75transformation efficiency, 75transgene expression, 74-5transgene integration, 73-4

transgene introduction, 73whisker (fiber)-mediated transformation, 68, 75wounding for DNA uptake, 72zygotic proembryo microinjection, 67-8

marker genes, filamentous fungi, 37Medicago sativa, 112-18Medicago varia> 113meisosis, transgene transmission, 73microinjection techniques

Brassica napus, 67rapeseed, 131small-grain cereals, 89soybean, 105sunflower, 144zygotic proembryo in maize, 67-8

microprojectile bombardmentconifers, 153-4DNA containing CaMV 35S constructs, 104gold particles, 104wpdlgene, 104soybean, 103-5sunflower, 143-4transgenic rice plant generation, 55uidA gene, 104see also particle bombardment

mitochondrial plasmid replicons, 38-9mob::Tn5 system, 27mobilization genes, 4mobilizer strains, 4monocotyledenous plants

gene expression regulation, 53transgenic, 53

moth beanprotoplast transformation technique, 108-9transformation, 108-9, 118, 119

Mucor, autonomously replicating sequenceincorporation, 39

mutanolysin, 13mycobacteria

cloning vectors, 17, 18phasmids, 16plasmids, 16-17

mycobacterial diseases, 10Mycobacterium aureum transformation efficiency, 13Mycobacterium avium, 10Mycobacterium bovis bacille Calmette-Guerin (BCG),

10Mycobacterium leprae> 10Mycobacterium smegmatis, integration by homologous

recombination, 17Mycobacterium tuberculosis, 10

illegitimate recombination, 18mycolic acid, 11

navy bean, 111-12GUS activity, 111

Nectria haematococca, 44

Index 171

gene cloning, 41neenij 158Neurospora crassa, 34

autonomously replicating sequence incorporation,39

electroporation transformation, 36gene identification method, 41homologous integration, 39inositol-requiring mutant, 34, 37lithium acetate transformation, 36particle bombardment, 37repeat-induced point mutation (RIP), 40

nitrogen fixation, forest trees, 157-8nocardioform bacteria, 10-19

cloning systems, 10cloning vectors, 17, 18conjugation, 15-16DNA introduction, 11, 12, 13, 14, 15-16electroporation, 13, 14, 15illegitimate recombination integration, 17-18integration of nonreplicative, 17lysozyme action sensitivity, 11pathobiology of infection in plants, 18-19PEG-mediated DNA uptake in

protoplasts/spheroplasts, 11, 12, 13plasmids, 16-17self-transmissible extrachromosomal elements, 15transduction, 16

nonreplicative vector integration, 17nopaline

rapeseed transformation, 130synthase activity, 66

nos genebirdsfoot trefoil transformation, 116conifers, 154soybean transformation, 104

Novozym-234, 34npdl gene, 55, 85

Agrobacterium tumefaciens transformation of pea,106-7

alfalfa transformation, 112, 113birdsfoot trefoil transformation, 116conifer transformation, 154cowpea transformation, 110macroinjection into floral tillers, 86marker for DNA transfer to pollen tube, 87, 88microprojectile bombardment, 104moth bean transformation, 108peanut transformation, 112Populus expression, 155reporter gene in barley, 90soybean transformation, 101, 102, 103sunflower regeneration, 142

nucleic acids, microprojectiles, 37

oaks, 158oat gene transfer, 91-2

octopinerapeseed transformation, 130synthase activity, 66

oilseed production, 125see also rapeseed

oleic acid, 126one-step disruption, 42Onobrychis viciifolia, 118oocyte regeneration, 141opines

birdsfoot trefoil transformation, 117synthesis, 120utilization genes, 120

orchardgrass transformation, 92protoplast, 54

oriT, 4^^TRK2 vector mobilization, 7Oryza sativa, 53osmotically sensitive cells (OSCs), 34ovalbumin gene

alfalfa transformation, 113-14expression in alfalfa, 120

Panicum maximum, 92paromomycin, 142particle bombardment, 83-4

barley transformation, 93cell transformation rate, 70DNA coated, 70efficiency, 71illegitimate recombination, 84microcarriers, 83oat gene transfer, 91-2sorghum transformation, 93transgenic maize, 70-2wheat gene transfer, 91see also microprojectile bombardment

pathogenicity determinants, 44pBR322, 4

host range, 4pBR322-0n plasmid, 5pea, 106-8

Agrobacterium rhizogenes transformation, 107Agrobacterium tumefaciens transformation, 106, 107electroporation, 108GUS expression, 107NPT II activity, 107protoplast transformation technique, 108transformation, 118, 119tumor induction frequency, 106, 107

peanut, 112transformation, 118, 119

pEmu promoter, 85penicillin biosynthetic gene cluster cloning, 43penicillin G, 11pGV910 replication, 24Phaseolus vulgaris, 111-12

172 Index

phasmids, mycobacterial, 16phenazine derivatives, 7phenazine-1-carboxylic acid, 7phleomycin resistance gene, 17phosphinothricin (PPT), 142

rapeseed transformation, 132resistance, 71

phosphotricine resistance, 155phytopathogenicity, 3, 16Phytophthora

particle bombardment, 37vector for transformation, 39

pm2gene, 156pisatin demethylase, 44Pisum sativum, 106-8PJRD215, 24plant defence genes, 156plasma membrane permeability, high current-pulse

effects, 25plasma membrane pores, electroporation, 26plasmids

bacterial hosts, 4conjugative, 4extrachromosomal linear, 15fasciation-inducing linear, 19fungal, 38homologous integration, 42host range, 4instability, 6linear, 16self-transmissible, 27size and transformation frequency, 26transfer between bacterial strains, 26transfer of entire, 27transmissible, 4yeast, 38

pMON894, 102pMON9749, 102point mutation, repeat-induced (RIP), 40pollen, 66

regeneration, 141pollen grain, DNA uptake, 66-7pollen tube pathway, DNA transfer in small-grain

cereals, 87-8pollen-mediated indirect gene transfer, 88polyethylene glycol (PEG)

direct gene transfer to maize protoplasts, 68-9DNA introduction into rice protoplasts, 55electroporation, 13moth bean transformation, 108, 109small-grain cereals, 89sunflower transformation, 144

polyethylene glycol (PEG)-mediated DNA uptake inprotoplasts/spheroplasts, 11, 12> 13

Populusy 154-7DNA transfer, 155genome size, 154

gusA expression, 155herbicide resistant, 155insect resistance, 156nptll expression, 155transformation, 154-7

potato proteinase inhibitor gene, 156pRJ1035 vector, 28pRK290, 4, 23-4

Agrobacterium transformation, 25cosmid, 24Rhizobium transformation, 25

pRK2013,4, 24protein secretion, 43proteinase inhibitors, forest trees, 151protoplast transformation

embryogenic suspension cultures, 54-5transgenic rice plant generation, 54-5

protoplastsalfalfa transformation, 115barley, 89-90Black Mexican Sweet (BMS) origin, 69cereal plants, 81culture, 61direct gene transfer for maize transformation, 68-70DNA uptake, 13endogenous plasmid loss, 13feeder cell line interactions, 69filamentous fungi transformation, 34green plant regeneration, 91legumes, 119maize embryogenic cultures, 69maize endosperm origin, 69microspore-derived cultures, 69moth bean transformation, 108-9osmotic stabilizers, 35pea transformation, 108PEG-mediated DNA uptake, 11, 12, 13production increase, 11, 13rapeseed transformation, 131-2small-grain cereals, 82-3, 93soybean transformation, 105stable transformation in barley, 90sunflower, 140-1, 144wheat, 91

pRSFlOlO, 16pRT99gus plasmid, 154pSA vectors, 24ps&Agene, 105pseudomonads

DNA introduction, 3genetic manipulation, 3phytopathogenic, 3rhizosphere colonizers, 3transformation, 5-7

Pseudomonasconjugal transfer, 3-5electroporation, 26

Index 173

transformation, 3Pseudomonas aeruginosa LEC1, 6Pseudomonas aureofaciens strain 30-84, 7Pseudomonas fluorescens

strain 2-79, 7strain CHAO, 6-7strain HV37a, 6strain MS 1650, 5-6strain Pf-5, 6transformation, 3

Pseudomonas syringae, 3pv. syringae J900, 5transformation, 3

pSUP suicide vectors, 4pyoluteorin, 6pyrF gene, 17pyrrolnitrin, 6Pythium ultimum, 6

Quercus robur transformation, 158

rapeseed, 125Agrobacterium rhizogenes transformation, 126-7Agrobacterium tumefaciens transformation, 127,

130-1agronomic applications of transformation, 133, 132,

134auxin in regeneration, 130coculture time, 130-3cotransformation with A. rhizogeneslA. tumefaciens,

127direct gene transfer, 131-3double zero cultivars, 126electroporation technique, 133gene transfer, 126hairy root production, 127herbicide-resistance, 134microinjection techniques, 131microspore-derived embryos, 130nuclear male sterility, 134oil quality, 125-6, 134polymerase chain reaction (PCR) techniques, 127protoplasts, 131-3regeneration tissue, 130spring cultivars, 125, 133transformation efficiency, 133transformation experiments, 128-9transformation techniques, 126-7, 128-9, 130-3transgenic shoot selection, 131winter cultivars, 125, 133

rapeseed, transgenicdisarmed T-DNA, 127fertile, 127transgene copy number, 131

rbcS promoter, 58rbcS-uidA fusion gene, 58-9recombination

excisive, 6

homologous, 6illegitimate, 82, 84site-specific, 75

reforestation, 158regeneration

adventitious from tissues, 139-40direct, 139efficiency, 141, 142from existing meristems, 140indirect, 139-40morphogenic callus, 140oocytes, 141plant form, 141-2plant properties, 141pollen, 141root induction, 141single cell, 140-1speed, 139sunflower, 145systems for sunflower, 142transgenic plants, 139-42

regulatory sequences, as-acting, 42repeat-induced point mutation (RIP), 40replicons, Gram-negative, 29reporter genes

construct insertion, 84small-grain cereals, 83

RH2,4Rhizobium, 23

complementation analysis, 28electroporation, 26homologous recombination, 28introduction of new desirable genes, 28-9nitrogen fixation, 23phage genomes, 27plasmid rearrangement, 28plasmid transfer, 27replication of recombinant molecules, 24site-specific recombination, 28transduction, 27transformation, 24-5vector stability, 28

Rhizoctonia solani, 6rhizosphere colonizers, 3rhodococci

cloning vectors, 17, 18electrotransformation, 17metabolic activity, 10plasmids, 16

Rhodococcus bronchialis, 10Rhodococcus fascians, 10, 15

illegitimate recombination, 18pathobiology of infection in plants, 18-19

Ri plasmid systems, legumes, 119-20rice

Ac element introduction, 59Agrobacterium-based vectors, 89

174 Index

rice (com.)genome HmR plasmid integration patterns, 57pests and Bacillus thuringiensis gene, 61pollen tube pathway DNA transfer, 87transformation, 53-4, 57transposon tagging system, 59-60

rice actingene, 59promoter, 84

rice plants, transgenic, 53CaMV 35S promoter, 58cotransformation, 57direct DNA transfer, 54efficiency, 55gene expression regulation, 58-9gene regulation, 61generation methods, 54-8genome size, 53GUS expression, 58, 59integration patterns of foreign DNA, 57intron enhancement of gene expression, 56-7microprojectile bombardment, 55plasmid DNA rearrangement, 57protoplast transformation, 54-5rbcS promoter, 58rbcS-uidA fusion gene, 58-9restriction fragment length polymorphism (RFLP)

maps, 53, 61rice stripe virus coat protein, 60ro/C of Ri plasmid, 59selectable markers, 55-6trans-acting factors, 59transgene integration, 57transposable elements, 61undesirable mutations, 61useful gene introduction/expression, 59-61

rice protoplasts, 54-5culture, 61electroporation for DNA introduction, 55PEG for DNA introduction, 55transformation, 56

rice stripe virus, coat protein (CP), 60-1RK2

deletion, 7derivative plasmids, 27primase gene, 7tetracycline resistance gene, 4tra gene absence, 27

rolCpromoter, 59of Ri plasmid, 59

root induction, 141RSF1010, 4rye

DNA macroinjection into floral tillers, 85-6gene transfer, 92

Saccharomyces cerevisiae, 34gene disruption, 42gene replacement, 42homologous integration, 39linearized DNA, 40lithium acetate transformation, 36particle bombardment, 37

Salix, 157sanfoin, 118

transformation, 118, 119seed inbibition, small-grain cereals, 88self-cloning

filamentous fungi, 40-2transformant selection, 41

shoot morphogenesis, sunflower, 139shuttle vectors, 4, 16

for pseudomonads, 5sibling selection, 41silicon carbide whiskers (fibers), 75

small-grain cereals, 89silver nitrate, 73sodium dodecylsulfate (SDS), 11somaclonal variation, legumes, 120somatic embryogenesis

sunflower, 139yellow poplar, 157

sorghum transformation, 92, 93Southern pine bark beetle, 151soybean, 101-8

Agrobacterium rhizogenes transformation, 103Agrobacterium tumefaciens transformation, 101-3,

105-6cyst nematode propagation studies, 103hairy root cultures, 103leghemaglobin gene expression, 120microinjection transformation technique, 105microprojectile bombardment, 103-5protoplast transformation by DNA direct uptake,

105somaclonal variation, 120transformation, 118, 119transposable element integration, 120uidA gene expression, 104

soybean, transgenic, 102atrazine resistance, 120embryogenic suspension culture, 104glyphosphate resistance, 120GUS activity, 102, 104, 105hygromycin resistance, 104NPT II activity, 102, 103RI plants, 104

spheroplastsDNA uptake, 13endogenous plasmid loss, 13PEG-mediated DNA uptake, 11, 12, 13production increase, 11, 13

spoCIC gene, 42

Index 175

starch composition of endosperm, 60sterility, nuclear male, 134Stylosanthes transformation, 118, 119suicide vectors, 4, 6sunflower, 137-46

Agrobacterium tumefaciens transformation, 138, 145Agrobacterium-mediated gene transfer, 143-4auxin : cytokinin ratio for transformation, 143callus induction, 139, 140crown gall tumors, 138, 143culture-derived, 142direct gene transfer, 144direct somatic embryogenesis, 139electroporation technique, 144embryo microinjection, 144Fj hybrid cultivars, 137, 138flower structure, 138gene transfer, 138growth cycle, 137-8GUS activity, 143HA 89B line regeneration, 142HA 300 line regeneration, 142herbicide resistance, 145hypocotyl explant transformation, 143in vitro culture, 142, 145indirect embryogenesis, 140insect attack resistance, 145interspecific hybridization, 145male sterile lines, 138meristem regeneration, 140microinjection, 144morphogenic callus, 140NPT II activity, 143particle gun transformation, 143-4pollinator lines, 138polyethylene glycol technique, 144protoplasts, 140-1, 144regenerated plant properties, 141-2regeneration, 139-42, 145shoot formation, 139, 140-1shoot morphogenesis, 139single cell regeneration, 140-1somaclonal variants, 142tissue culture, 145transformation, 138-9, 142-4

sweetgum transformation, 157

take-all, 7thallium resistance plasmids, 15thaumatin gene, 86Thielaviopsis basicola, 6Ti plasmids, 27tobacco

barnase gene expression, 72black root rot, 6kanamycin-resitant transformants, 74mosaic virus (TMV) coat protein, 60

NT1 cell line, 71rraws-acting gene product titration, 42transconjugant, 3

recovery frequency, 7transcriptional regulators, zraws-acting, 42transduction, 27

coryneform bacteria, 16nocardioform bacteria, 16

transfer efficiency, 4transfer vector mutagenesis, 7transferred DNA (T-DNA), 66transformants

genetic transformation, 38selection for filamentous fungi, 37-8

transformationbiotechnology applications, 42-4conifers, 152-4direct, 8efficiency, 13electroporation efficiency, 13, 15forest trees, 151-2Gram-negative soil bacteria, 24-5hardwoods, 154-8stable, 81-2sunflower, 142-4

transforming DNAfate, 38-40integration into chromosomes, 39-40

transgeneinsertion site, 75integration in maize, 73-4silencing, 75

transgene expressioninheritance, 74-5inheritance in maize, 74transgenic maize, 73, 74-5

transposable elementsactivator (Ac), 59dissociation (Ds), 59introduction into rice, 59

transposonmutagenesis, 6tagging, 59-60

Trichoderma reeseiheterologous gene expression, 43protein glycosylation system, 43protein secretion, 43

Tris, Pseudomonas exposure effects, 7trpC\ 42tuberculosis, 10turfgrass transformation, 92

uidA gene, 58alfalfa transformation, 115barley, 90macroinjection into floral tillers, 86marker for oat gene transfer, 92

176 Index

w/dAgene (cont.)microprojectile bombardment, 104pea transformation, 108peanut transformation, 112small-grain cereals, 84, 85soybean transformation, 102, 105

Ulmus, 158

vectorsautonomously replicating, 38-9building, 39cosmid-type, 16integrative expression, 19

Viciafaba, 109Vicia narbonensis, 109-10

transformation, 118, 119vicilin, 91Vigna aconitifolia, 108-9Vigna unguiculata, 110-11vir gene, 27viral resistance, cereals, 61

walnut transformation, 151waxy gene, 60

wheatembryogenic tissue cultures, 91protoplasts, 91stable transformation, 91transient gene expression, 91

white clover transformation, 118, 119white pine blister rust, 151willow, 157wilting diseases, 10wood, 150

processed products, 154woody plants

genetic diversity, 150molecular basis of biological processes, 150short rotation crops, 156see also forest trees

Xanthomonas, 23electroporation, 26

xylem, genetic engineering, 154

yellow poplar transformation, 157

Zea mays, 65

transformation of plants and soil microorganisms (biotechnology research (no. 3))

Documents