Application of

Molecular Markers to

Rice Breeding in

Australia

Molecular markers for the sd-1 and fgrgenes

A report for the Rural Industries Research

and Development Corporation

by Stephen Garland &Robert Henry

May 2001

RIRDC Publication No 01/38

RIRDC Project No. USC-2A

© 2001 Rural Industries Research and Development Corporation.

All rights reserved.

ISBN 0 642 58260 2

ISSN 1440-6845

Application of molecular markers to rice breeding in Australia

Publication No. 01/38

Project No. USC-2A

The views expressed and the conclusions reached in this publication are those of the author

and not necessarily those of persons consulted. RIRDC shall not be responsible in any way

whatsoever to any person who relies in whole or in part on the contents of this report.

This publication is copyright. However, RIRDC encourages wide dissemination of its

research, providing the Corporation is clearly acknowledged. For any other enquiries

concerning reproduction, contact the Publications Manager on phone 02 6272 3186.

Researcher Contact Details

Professor Robert Henry

Centre for Plant Conservation Genetics

Southern Cross University PO Box 157

Lismore NSW 2480

Phone: 02 66203010

Fax: 02 66222080

Email: rhenry@scu.edu.au

RIRDC Contact Details

Rural Industries Research and Development Corporation

Level 1, AMA House

42 Macquarie Street

BARTON ACT 2600

PO Box 4776

KINGSTON ACT 2604

Phone: 02 6272 4539

Fax: 02 6272 5877

Email: rirdc@rirdc.gov.au.

Website: http://www.rirdc.gov.au

Published in May 2001

Printed on environmentally friendly paper by Canprint

ii

Foreword

The major objectives of the project were to identify, adapt and evaluate molecular markers for routine

use in the Australian rice-breeding program. The major component of the work involved the

development of markers for the semi-dwarfing gene sd-1 and the major fragrance gene fgr.

This publication provides a description of the development and assessment of four useful markers for

sd-1 and three useful markers for fgr. The utilisation of these markers within the Australian rice-

breeding program is discussed.

This project was funded from industry revenue which is matched by funds provided by the Federal

Government and is an addition to RIRDC’s diverse range of over 600 research publications. It forms

part of our Rice R&D program, which aims to improve the profitability and sustainability of the

Australian rice industry.

Most of our publications are available for viewing, downloading or purchasing online through our

website:

• downloads at www.rirdc.gov.au/reports/Index.htm

• purchases at www.rirdc.gov.au/eshop

Peter Core

Managing Director

Rural Industries Research and Development Corporation

iii

Acknowledgments

Thanks to all the staff from Yanco Agricultural Institute and Southern Cross University who

provided invaluable assistance. Thanks to Dr Susan McCouch and the Cornell Research

Foundation Inc, New York, their contribution was critical to the success of this project.

iv

Contents

Foreword ................................................................................................................................ii

Acknowledgments ................................................................................................................. iii

Executive Summary............................................................................................................... v

1. Introduction .................................................................................................................... 1

1.1 Semi-dwarfism ............................................................................................................ 1

1.2 Fragrance ................................................................................................................... 1

2. Objectives ...................................................................................................................... 3

3. Methodology................................................................................................................... 3

3.1 Mapping Population .................................................................................................... 3

3.2 DNA extraction............................................................................................................ 4

3.3 Supply and investigation of the RFLP probes............................................................. 4

3.4 Evaluation of marker utility.......................................................................................... 5

4. Results ........................................................................................................................... 6

4.1 The development of markers for sd-1 ......................................................................... 6

Number of alleles............................................................................................................. 10

4.2 The development of markers for fgr.......................................................................... 10

5. Discussion.................................................................................................................... 13

5.1 Markers for sd-1........................................................................................................ 13

5.2 Markers for fgr .......................................................................................................... 14

6. Implications .................................................................................................................. 15

7. Recommendations ....................................................................................................... 17

8. References................................................................................................................... 19

v

Executive Summary

The major objectives of the project were to identify, adapt and evaluate molecular markers for routine

use in the Australian rice-breeding program. The major component of the work involved the

development of markers for the semi-dwarfing gene sd-1 and the major fragrance gene fgr.

The semi-dwarfing gene (sd-1) is responsible for producing sturdy and moderately short plants with

higher harvest index compared to similar tall plants. sd-1 is the source of semi-dwarfism in the

Australian rice varieties. The genetics of fragrance in rice is more complicated and involves the

production of basmati or jasmine style aroma. Several chemicals are involved with aroma, however a

single constituent (2-acetyl-1-pyrroline) is believed to be a major component and is controlled by a

single gene.

The application of marker assisted selection for these genes, has potential to greatly improve the

efficiency of the Australian rice-breeding program. Markers have the capacity to assist in the selection

of traits that are expensive or laborious to assess. Molecular markers can be evaluated from a single

seedling leaf or seed samples, allowing selection to occur before the trait is expressed. The semi-dwarf

character, for example, could be detected before maturity or even as a heterozygote thus eliminating the

need for progeny testing in a backcrossing program. Grain quality characters such as aroma are very

difficult to assess accurately. Markers used in preliminary screening would reduce labour costs and

assist in selection.

Fragments of DNA linked to sd-1 and fgr were assessed for differences to develop markers for these

genes. The techniques utilised, ensured that the markers would be easily produced and identified.

Four useful markers have been produced for sd-1. Three useful markers have also been produced or

identified for fgr. The markers for fgr and sd-1 will be useful in the Australian rice-breeding program

and in rice breeding programs in general. This project has demonstrated that molecular markers are

suitable for variety identification and marker assisted selection within the Australian Industry.

It will be necessary to develop more markers and to assess those markers in a wider selection of rice

lines. The presence of a database containing the alleleic characterisations of general markers or those

markers linked to genes or quantitative traits, would allow for the selection of a suite of markers which

are polymorphic between Australian varieties, breeding lines and breeding samples. The markers could

be selected for use in germplasm identification, genetic mapping, and marker assisted selection

programs.

The challenge is to maximise the potential benefits of markers through the appropriate integration of

marker assisted selection into the breeding program. It may be necessary to restructure the breeding

program to obtain maximum benefits.

This project was the subject of a PhD project for Stephen Garland providing important training in rice

molecular biology.

1

1. Introduction

1.1 Semi-dwarfism

The agronomic benefits of shorter rice plants can include increased lodging resistance and increased

harvest index (reviewed by Mackill and Rutger 1979), however most dwarfing genes are associated with

unsuitable agronomic characteristics (reviewed by Cho et al., 1994, Maluszynski et al., 1986, Mackill

and Rutger, 1979).

One semi-dwarfing gene, sd-1 (d-47), has been used predominantly in the development of highly

productive rice cultivars. sd-1 increases lodging resistance, harvest index, responsiveness to nitrogen

fertilisation and improves general plant type (reviewed by Cho et al., 1994). The semi-dwarfing gene sd-

1 is incompletely recessive (Murai et al., 1995), and as a homozygote reduces total culm height by

about 25% (Rutger et al., 1986). The absolute reduction of internode length decreases from the first to

the sixth internode (Murai et al., 1995). There is currently no hard evidence to implicate sd-1 with the

biosynthesis of, or response to, any particular phytohormone involved with stem elongation.

Most of the Australian semi-dwarf cultivars have M7 (Carnahan et al., 1978) in their pedigree. M7 is a

cultivar that possesses the sd-1 locus. sd-1 is therefore the likely candidate dwarfing gene predominating

in the Australian industry.

sd-1 is partially recessive and the identification of the correct genotype is not completely accurate based

on phenotype (height class). The degree of expression is dependent on environmental conditions such as

fertiliser levels (Murai et al., 1995a) and problems with genotyping would be exasperated under

variable field conditions. A further complicating factor, that would also tend to widen genotype height

classes and increase class overlap in breeding populations, is the presence of numerous QTL loci

associated with height (see Li et al., 1995, Lu et al., 1996, Zhuang et al., 1997, Yan et al., 1998). The

correct identification of heterozygotes, that could be produced after a tall recurrent parent is

backcrossed to a semi-dwarf individual, would not be definite under controlled conditions.

Given the benefits of incorporating sd-1 into new varieties, the utilisation of molecular markers for the

identification of the genotype for the sd-1 locus would increase the efficiency of a breeding program.

Heterozygotes developed from a backcrossing exercise could be identified from small samples of

seedling leaves, alleviating the need for maturation of plants and genotyping based on the identification

of segregation in the progeny.

1.2 Fragrance

Several chemical constituents are important to the aroma of cooked rice (Grosch and Schieberle, 1997).

However, 2-acetyl-1-pyrroline (AP) is regarded as the most important component of aroma in the

basmati and jasmine style fragrant rices (Lorieux et al., 1996). AP is found in all parts of the rice plant,

except for the roots (Lorieux et al., 1996) and is also found, at concentrations up to 100 times lower, in

non-fragrant varieties (reviewed by Grosch and Schieberle, 1997).

Lorieux et al. (1996) reviewed the genetics of aromatic fragrance and concluded that a single recessive

gene was responsible for the production of fragrant rice plants.

2

Fragrance can be detected by tasting the associated flavour in individual seeds or assessing the aroma of

leaf tissue or grains after either heating in water or reacting with solutions of KOH or I2-KI (reviewed

by Tragoonrung et al., 1996).

Tasting individual grains has been the preferred method for quality selection of aromatic rice varieties

within the Australian breeding program (Reinke et al., 1991). However, there are problems with the

sensory detection of fragrance. There is considerable variation between analysts in their ability to detect

fragrance or the associated flavour. Some individuals have difficulty in detecting the aroma or taste. A

rapid decline of an individuals ability to distinguish between fragrant and non-fragrant samples with

each analysis performed is also experienced. An analyst's ability declines as the senses become saturated

or actual physical damage occurs. Abrasions to the tongue, causing bleeding, often result from chewing

numerous seeds. Caustic substances such as KOH would also cause damage to the nasal passages.

Sensory methods are therefore not suitable for processing large numbers of samples.

The chemical detection of AP is also possible but is time consuming and requires large samples (eg.

hundreds of grams, Widjaja et al., 1996, Lorieux et al., 1996). The development of a PCR based

molecular marker for the major component of fragrance (AP), for use within breeding programs, would

have many advantages over sensory or chemical detection methods. Many more plants could be

processed and sample sizes of 0.1g or less could be analysed.

Plants from a breeding program for fragrant lines, in the early stages of cultivar development, could be

assessed before maturity from small amounts of leaf tissue to reduce the number of seed samples tested

by sensory methods. Non-fragrant or heterozygous individuals would be avoided for the sensory or

chemical assessment of aroma or flavour quality or tests for other grain quality characteristics. As the

recessive gene could be detected in a heterozygous state, the marker would be useful for the

identification of offspring possessing fgr after backcrossing with homozygous non-fragrant plants.

3

2. Objectives

Several authors have reported polymorphisms within rice RFLP probes and sequence-tagged-sites

(Williams et al., 1991, Fukuoka et al., 1994, Ghareyazie et al., 1995, Xu et al., 1998a). Based on this

success and the many benefits associated with a PCR based marker, the objectives of this study were to

convert RFLP probes and other loci genetically mapped near sd-1 and fgr into PCR based markers.

Polymorphic sites could then be identified by sequencing homologous regions in parent cultivars of a

mapping population or identified as size differences between the PCR products. The polymorphic

markers could then be assessed for linkage to semi-dwarfism and fragrance and ultimately used to

produce PCR based markers for assessment and utilisation within the Australian rice breeding program.

The Australian cultivar Doongara, has an unknown source of semi-dwarfism, as no short stature plants

are present in its pedigree (Bluebelle / Calrose // Jojutla, Ko et al., 1994). A secondary aim of this study

was to determine if the semi-dwarfism in Doongara was also due to sd-1.

3. Methodology

Cho et al. (1994) identified the close linkage of the RFLP markers RG220 and RG109 (co-segregated),

to sd-1 (0.8 cM) on chromosome 1. Maeda et al. (1997) also identified linkage between these two

markers and sd-1. RG220 and RG109 were found to be at a genetic distance of 0.3 and 0.9 cM. The

single recessive fragrance gene (fgr) has been linked to the RFLP clone RG28 on chromosome 8, at a

genetic distance of 4.5cM (Ahn et al., 1992). Lorieux et al. (1996) confirmed the close linkage between

RG28 and fgr (5.8cM).

We screened for polymorphism in PCR amplified regions homologous to RFLP clones linked to sd-1

and fgr to develop PCR based markers. In addition, microsatellite markers previously mapped in the

vicinity of fgr were assessed for polymorphism information content to provide alternative markers for

fragrance.

3.1 Mapping Population

A population of 215 F2 seedlings, derived from a cross between Kyeema (tall, fragrant, long-grain,

Australian cultivar) and Doongara (semi-dwarf, non-fragrant, long grain, Australian cultivar) and 100

seedlings of each parent cultivar were transplanted to the field at 30cm intervals at Yanco Agricultural

Institute, NSW Agriculture, Yanco, in November 1997. Leaf material was collected in January 1998

and frozen for DNA extraction at a later date. Height measurements for phenotype determination were

taken in early March 1998 and were determined for the tallest panicle as the distance from the lower

node of the third internode to the top of the panicle. Height genotypes were determined for 50 F2 plants

by assessing height and height segregation for 8 F3 individuals from each F2 plant. The F3 plants were

grown in 20cm diameter pots from November 1998 to February 1999 at Southern Cross University.

The observed segregation ratios between height genotypes, for the 50 genotyped F2 individuals, were

tested by χ2 analysis against the expected ratio for a single gene (1 : 2 : 1).

124 F2 plants (including the 50 F2 genotyped for height) were classified as fragrant, segregating or non-

fragrant by tasting dehulled F3 seed (ground between front teeth before tasting). If five F

3 seeds were

tasted and all were fragrant, the F2 individual was considered fragrant, otherwise at least 12 seeds were

tasted to separate segregating from non-fragrant F2 individuals. The observed segregation ratio of non-

4

fragrant:segregating:fragrant, was tested by χ2 analysis against the expected ratio for a single gene (1 :

2 : 1).

3.2 DNA extraction

DNA was extracted from approximately 0.6 g of frozen leaf tissue as generally described by Weining

and Henry (1995) for 50 F2 individuals genotyped for height and parent cultivars (Kyeema and

Doongara). The following modifications were included; 2 ml of 2.5% sarkosyl extraction buffer and 2

ml of phenol/chloroform/isoamyl alcohol (25:24:1) were used; DNA was precipitated with 1.3 volumes

of isopropanol in addition to 0.13 volumes of 3M sodium acetate (pH 4.8) and then washed twice with

70% ethanol. The DNA was resuspended in 200 µl of Tris-EDTA (TE) buffer. DNA preparations were

diluted with TE buffer to a final concentration of approximately 50 to 100 ng per µl.

3.3 Supply and investigation of the RFLP probes

The probes RG109, RG220 and RG28 (rice-etiolated-leaf genomic library, PstI restricted, pUC9

vector, cultivar IR36) were supplied by Dr. Susan McCouch and the Cornell Research Foundation Inc,

New York. The clones were bi-directionally sequenced using commercial pUC DNA sequencing primers

and the ABI PrismTM, BigDyeTM Terminator Cycle Sequencing Ready Reaction Kit. Gel separation and

electropherogram production were performed by the Australian Genome Research Facility, University

of Queensland.

Probe-specific PCR primers were designed with the aid of MACVECTORTM 6.0, Sequence Analysis

Software, Oxford Molecular Group Inc. Primers were synthesised by Pacific Oligos Pty Ltd, SCU,

Lismore, NSW, Australia. PCR reactions using template DNA from Kyeema and Doongara were

carried out on a Perkin Elmer, Gene Amp PCR System 9700. The reaction volume was 20µl containing

1X Boehringer Mannheim PCR Buffer (1.5 mM MgCl2), 50 to 100 ng of genomic DNA and 200µM

dNTPs. The PCR reaction mix also included 150 nM of each primer, 2 mM total MgCl2, and 0.5 units

of Taq DNA Polymerase (Boehringer Mannheim). The temperature cycling conditions were 1 minute at

94° C; followed by 35 cycles of 94° C for 30 seconds, 54° C for 30 seconds and 72° C for 1 minute;

with a final hold at 72° C for 5 minutes.

PCR products were purified with a QIAquick PCR Purification Kit (QIAGEN) and sequenced as

discussed above. Sequence polymorphism between Kyeema and Doongara, the parents of the mapping

population were identified. Primers flanking the polymorphism, if required, were designed and

synthesised as previously indicated, except the forward primer was 5´end labeled with the fluorescent

phosphoramidite dye, HEX or FAM (Perkin-Elmer). Sizing of the PCR products were performed on an

ABI Prism 310 Genetic Analyser with the following run and analysis conditions (Module GS STR

POP4 C, 20 minute run time, size standard GENESCAN 500 - TAMRA, Local Southern Sizing

method, Capillary-length-to-detector 30 cm). The sizing of DNA fragments was relative and not

necessarily an absolute measure of size.

Sequence tagged sites from the genetic map of Harushima et al. (1998) were also selected for

assessment. These STS were developed from RFLP probes produced from the cultivar Nipponbare and

included C86, C1439, G1133, L543, L819, R2625, S2523, S2596, S13312, and S13471. These STS

were expected to be near sd-1, based on the relative position of integrated markers from the genetic

maps of Harushima et al. (1998) and Causse et al. (1994) (the latter map includes sd-1) as presented in

the genetic map of Maeda et al. (1997). The STS and clone RZ538 (EcoR1, rice etiolated leaf cDNA

library, variety IR36, Causse et al., 1994) was also selected for assessment as it had been genetically

5

mapped close to sd-1 and end sequence was available. DNA sequences for the STS were obtained from

the Ricegenes database.

Fifty F2 individuals from the mapping population were genotyped for the polymorphic markers. The

genetic distance between fgr or sd-1 and the linked markers were determined as the percentage of

recombinant chromosomes (cM).

3.4 Evaluation of marker utilityTo assess the usefulness of the PCR based markers within the Australian rice industry, the marker

alleles were characterised for 24 rice samples. The rice samples analysed included 16 Australian

breeding lines or Australian commercial cultivars. Half of the 24 samples were fragrant lines and also

included 13 tall and 11 semi-dwarf lines. Rice seed samples were supplied by Yanco Agricultural

Institute (YAI), New South Wales Agriculture, Yanco, Australia. (See Ko et al. (1994) for some of the

Australian pedigrees, others are available through the YAI if required). Five plants for each variety

were grown in a glasshouse at Southern Cross University, Lismore, NSW, Australia. Approximately 1

to 3 cm of mature leaf was sampled from each of the five plants for each cultivar. The leaf material

from the five plants for each cultivar was bulked producing a total weight of about 0.3 g. The DNA was

then extracted as described above.

Microsatellite markers RM42 and RM223 (Chen et al., 1997) were selected in addition to the markers

developed from the RFLP clone RG28 for assessment of polymorphism and allele characterisation.

These markers are linked to fgr and will provide alternate markers for this gene. RG28 and RG1 are

linked at a map distance ranging from 10 cM (Causse et al., 1994) and 12 cM (Lorieux et al., 1996) to

25.5 cM (Cho et al., 1998). RG28 and RG1 flank the genome region that contains fgr. RG28 was not

mapped by Chen et al. (1997) and Xu et al. (1998b). However, the location of RG28 and fgr, in relation

to RM42 and RM223, can be estimated based on the relative position of RG1 and other nearby

markers, in these genetic maps, and in the genetic maps of Cho et al. (1998), Lu et al. (1996) and

Causse et al. (1994).

RG28 probably lies between RM42 and RM223. RM42 may therefore be a few cM further from fgr

than RG28 and RM223 is possibly closer. In most cases these microsatellite markers are likely to be

within 10 cM of fgr and therefore useful for gene tagging within a breeding program.

Primer sequences for RM42 and RM223 are available from Chen et al. (1997). Primers were

synthesised by Pacific Oligos Pty Ltd, SCU, Lismore, NSW, Australia. PCR reactions were carried out

on a Perkin Elmer, Gene Amp PCR System 9600 or 9700. The reaction volume was 20µl containing 1X

Boehringer Mannheim PCR Buffer (1.5 mM MgCl2), approximately 50 ng of genomic DNA, 200µM

dNTPs and 200 nM [R110] dUTP (Perkin Elmer). The PCR reaction mix also included 100 nM of each

primer, 3mM total MgCl2, and 0.5 units of Taq DNA Polymerase (Boehringer Mannheim). The

temperature cycling conditions for the microsatellite markers were 1 minute at 94° C followed by 30

cycles of 94° C for 15 seconds, 54° C for 15 seconds and 72° C for 1 minute. Sizing of microsatellite

alleles was performed on an ABI Prism 310 Genetic Analyser

Polymorphism Information Content values (PIC) were calculated for each marker using Nei’s Gene

Diversity measure Hs (Nei 1973)

n

Hs = 1 - Σ xk

2

k = 1

Where xk is the frequency of the kth allele.

6

4. Results4.1 The development of markers for sd-1

The observed segregation ratios between height genotypes, for 50 F2 individuals, are presented in Table

1. The results are consistent with a single gene controlling semi-dwarf stature χ2 (0.05, df=2), p=0.5)].

Table 1 Observed segregation ratio between height genotypes for 50 F2 individuals from the mapping

population (Kyeema/Doongara) and the expected segregation ratio for a single gene. χ2 (0.05, df=2)

analysis indicated no significant difference between observed and expected ratios, p=0.5.

Homozygous

Semi-dwarf

Heterozygote Homozygous

Tall

Significnce

Observed

frequency11 23 16

not significant

Expected

frequency12.5 25 12.5

Four DNA polymorphisms were identified in homologous regions of RFLP clones and STS, between

Kyeema and Doongara, that could be detected as a size difference or the loss or gain of restriction site.

The markers, primer details and alleleic detection methods are supplied in Table 2. The PCR conditions

for the sequence tagged sites (STS) listed in Table 2 were described in the methods section as indicated

for the conversion of the RFLP and STS to PCR based markers. The PCR conditions for SCU-Rice-

SSR-2 were the same as for the RM prefixed microsatellite markers without the addition of flourescent

dUTPs.

Table 2. Name and primer details for the PCR based markers for sd-1. The RFLP clones used to

develop the PCR based markers are included in the marker name. The exception is the microsatellite

marker SCU-Rice-SSR-2 that was developed from RG109.

Marker Name Primer pairDetection

Method

SCU-Rice-SSR-2

5'-AGCTACTATCAGACAACAGAAAACG-

3' Size Separation

(Source RG109) 5'-ATACGCCAAGATTTCCTAAAC-3'

SCU-Rice-STS-13312 5'-GGAACTGCTCTGCTACTAACCCTG-3' Size Separation

5'-CGTCGGGAAACGAATCCAAC-3'

SCU-Rice-STS-S13471 5'- GCACCACATACTCACTTCTGCTTG-3' Restriction

5'-GCAGGTGTAGCTCTTGTCACAGTG-3' Digest (Pst I)

SCU-Rice-STS-

RG22O 5'-ATTGACCTGTGGATTGCTGTCTG-3' Restriction

7

5'-TGGCGTCTGCTGCTCATTTG-3' Digest (Hae III)

Alleles for SCU-Rice-STS-S13312 are presented if Figure 1. Alleles for SCU-Rice-SSR-2 are

presented in Figure 2. Alleles for SCU-Rice-STS-S13471 and SCU-Rice-STS-RG220 after agarose gel

electrophoresis and ethidium bromide staining of restriction-enzyme digested, PCR-products, are

presented in Figures 3 and 4.

Figure 1 Capillary electrophoresis output, demonstrating alleles for the marker SCU-Rice-STS-13312,

for the varieties of Kyeema and Doongara and one heterozygote F2 individual from the mapping

population (Kyeema / Doongara). Alleles differ by 1 bp. Horizontal scale in bp. Size is relative and not

an accurate measure. Expected allele sizes based on sequence information were 206 and 207 bp.

Doongara allele - 305

bp

Heterozygote

Kyeema allele - 297

bp

209 bp allele - Kyeema

Heterozygote

208 bp allele-Doongara

8

Figure 2 Capillary electrophoresis output, demonstrating alleles for the marker SCU-Rice-SSR-2, for

the varieties of Kyeema and Doongara and one heterozygote F2 individual from the mapping population

(Kyeema / Doongara). Alleles differ by 8 bp. Horizontal scale in bp. Size is relative and not an accurate

measure. Expected allele sizes based on sequence information were 303 and 311 bp.

Figure 3 Hae III digestion of PCR products of marker SCU-Rice-STS-RG220 after agarose gel

electrophoresis and ethidium bromide staining. Lanes 1 to 3 are individuals homozygous for the allele

found in Doongara. Note the presence of 2 bands approximately positioned at the expected size of 312

and 365 bp. A band is also faintly visible at the expected size of 110 bp. Lanes 7 to 9 are individuals

homozygous for the allele found in Kyeema or Amaroo. Note the presence of 1 distinct band

approximately positioned at the expected size of 677 bp. The product at 110 bp was very faint but could

be detected on the original gel. Lanes 4 to 6 are heterozygous individuals from the mapping population

(Kyeema / Doongara).

Figure 4 PstI digested PCR products of the marker SCU-Rice-STS-S13471 after agarose gel

electrophoresis and ethidium bromide staining. Lanes 1 and 4 are individuals homozygous for the allele

1 2 3 4 5

100 bp ladder

..500

300

..200

9

found in Kyeema. Note the presence of 2 bands approximately positioned at the expected size of 100

and 185 bp. Lanes 5 to 7 are individuals homozygous for the allele found in Doongara. Note the

presence of 1 distinct band approximately positioned at the expected size of 283 bp. Lanes 2 and 3 are

heterozygous individuals from the mapping population (Kyeema / Doongara).

No recombinants in 50 F2 individuals were found for sd-1, SCU-Rice-SSR-2 (RG109), and SCU-Rice-

STS-S13471. One recombination (map distance of 1 cM) was identified between sd-1 and both, SCU-

Rice-STS-RG220 and SCU-Rice-STS-S13312. Two recombination events were recorded between

SCU-Rice-STS-RG220 and SCU-Rice-STS-S13312 (2 cM).

Markers SCU-Rice-SSR-2, SCU-Rice-STS-RG220, SCU-Rice-STS-S13312 were assessed for levels

for polymorphism-information-content (PIC) across 24 rice samples. The marker SCU-Rice-STS-

S13471 was assessed for 5 cultivars as low levels of polymorphism and similar polymorphism patterns

suggested further screening would be unproductive.

Varietal screening and Polymorphism Information Content (PIC) values for markers SCU-Rice-STS-

RG220, SCU-Rice-STS-S13312, SCU-Rice-STS-S13471 and SCU-Rice-SSR-2 are presented in Table

3. SCU-Rice-SSR-2 was the most polymorphic marker with a PIC value of 0.50 and three alleles

identified.

All of the semi-dwarf varieties analysed were Australian breeding lines or cultivars. The tall varieties of

Kyeema, Pelde, Goolarah, Inga and YRF203 are Australian. The rest of the tall samples were more

distantly related foreign varieties, mostly indica subspecies. (see Garland et al., 1999, for an assessment

of genetic similarity between these varieties).

No polymorphisms were identified, for the markers linked to sd-1, between alleles possessed by the tall

Australian cultivars and most of the Australian semi-dwarf lines. Alleleic differences were only present

between both Doongara and YRF204, and the tall Australian varieties of Kyeema, Pelde, Goolarah,

Inga and YRF203. Most of the tall foreign varieties possessed distinct alleleic differences for the

markers SCU-Rice-STS-S13312 and/or SCU-Rice-SSR-2, to the semi-dwarf specimens. Exceptions

were for the comparison of Doongara and YRF204 with Khao Dawk Mali 105, and Della with all semi-

dwarf cultivars except for Doongara and YRF204. Low levels of polymorphism were detected for the

marker SCU-Rice-STS-RG220.

Table 3 Varietal screening and Polymorphism Information Content values for the markers SCU-Rice-

SSR-2, SCU-Rice-STS-RG220, SCU-Rice-STS-S13312, and SCU-Rice-STS-S13471. Alleles for

SCU-Rice-SSR-2 are in base pairs as determined by capillary electrophoresis. D = allele found in

Doongara. K = allele found in Kyeema. nd = no data.

Marker SCU-Rice-Semi-Dwarf SSR- 2 STS - RG220 STS -S13312 STS - S13471

Doongara 305 D D D

YRF204 305 D K D nd

Amaroo 297 K K K

Namaga 297 K K K

Millin 297 K K nd

Echuca 297 K K nd

Jarrah 297 K K nd

Illabong 297 K K nd

10

YRW4 297 K K nd

YRK4 297 K K nd

Langi 297 K K nd

TallKyeema 297 K K K

Pelde 297 K K K

Goolarah 297 K K nd

Inga 297 K K nd

YRF203 297 K K nd

Della 297 K K nd

Azucena 297 K D nd

Millagrossa 295 K D nd

Dumsiah 295 K D nd

Dumsorkh 295 K D nd

Basmati 370 295 K D nd

Moosa Taroom 110 295 K D D ndKhao Dawk Mali 105 305 D D nd

Number of alleles 3 2 2 2PIC value 0.5 0.2 0.4 -

4.2 The development of markers for fgr

Segregation ratios for the fragrance genotypes were found to be consistent with that of a single

fragrance gene (Table 4).

Table 4 Observed segregation ratio for 124 F2 individuals (Kyeema / Doongara) and the expected

segregation ratio for a single gene. χ2(0.05, df = 2) analysis indicated no significant difference between

observed and expected ratios, p = 0.12.

Fragrant Segregating Non fragrant Significan

Observed

number 28 55 41

not

significant

Expected

number 31 62 31

A small microsatellite mono T repeat in homologous fragments of RG28 was identifed that differed in

repeat number between Kyeema and Doongara. Doongara had 9 tandem T repeats, RG28 (Variety

IR36) had 8 and Kyeema 7. The primers SCU-Rice-SSR-1.F (HEX labelled) (5´-

GATCTCACTCCAAGTAAACTCTGAC-3´) and SCU-Rice-SSR-1.R

(5´-ACTGCCATTGCTTCTGTTCTC-3´) were designed to flank the microsatellite region producing an

expected product of 130 bp for Doongara (marker SCU-Rice-SSR-1). The primers amplified a single

product that was determined as 125 bp in Kyeema and as expected, 2 bp larger in Doongara. Size

determined by the 310 Genetic Analyser is not necessarily an absolute measure when using a general

size standard. PCR conditions for SCU-Rice-SSR-1 were the same as those for the RM prefixed

markers without the addition of fluorescent dUTPs. Alleles for SCU-Rice-SSR-1 are presented if Figure

5.

11

Figure 5 Capillary electrophoresis output, demonstrating alleles for the marker SCU-Rice-SSR-1 for

the varieties Kyeema, Dumsiah and Doongara. Horizontal scale in bp. The peak to the right of the

Kyeema allele is an artefact. Results are also presented for 2 homozygous and one heterozygote F2

individual from the mapping population (Kyeema / Doongara).

The microsatellite marker SCU-Rice-SSR-1 was closely linked to the fragrance gene in the mapping

population. Four recombinants were found in the 50 F2 plants (4cM). The four recombinants included 1

individual homozygous for fgr and heterozygous for SCU-Rice-SSR-1 and 3 homozygous non-fragrant

individuals that were heterozygous for SCU-Rice-SSR-1.

Alleles for the markers SCU-Rice-SSR-1, RM223, and RM42, for a range of fragrant and non-fragrant

rice samples are presented in Table 5. The three alleles found for SCU-Rice-SSR-1 in the fragrant

cultivars represented 7, 8 and 9 T repeats (PIC = 0.48). Two alleles were found in the non-fragrant

cultivars, the 9 bp allele predominated and only the breeding line YRK4 possessed the second allele of 8

bp. The markers RM223 (PIC = 0.66) and RM42 (PIC = 0.64) were more polymorphic. Forty-eight

percent of the possible 144 pair-wise comparisons between the fragrant and non-fragrant cultivars for

SCU-Rice-SSR-1 did not produce a distinguishing allelic difference. For this same assessment, 12.5%

for RM42, 33.3% for RM223 and 8% for all three markers combined were determined. The rice

samples that could not be comprehensively distinguished by allelic differences were Della from

Doongara, Amaroo, Echuca, Inga, Jarrah, Langi, Millin, Namaga, Pelde and YRW4; and YRF203 and

YRF204 from Doongara.

Kyeema - 7 T repeats

Dumsiah - 8 T repeats

F2 Homozygote - 7 T repeats

F2 Heterozygote

F2 Homozygote - 9 T repeats

Doongara - 9 T repeats

12

Table 5 Microsatellite alleles in bp (based on 310 Genetic Analyser results) for the markers SCU-Rice-SSR-1,

RM42 and RM223 for a range of fragrant and non-fragrant cultivars. (PIC–Polymorphism Information Content)

(*more alleles may be present)

Fragrant

CultivarsSCU-Rice-SSR-1 RM42 RM223

Kyeema 125 167 151

Goolarah 125 167 145

Khao Dawk

Mali 105125 167 145

Azucena 125 167 147

Millagrossa 126 160 169 145 153*

Dumsiah 126 160 147

Moosa

Tarom 110126 127 160 149

Dumsorkh 127 160 149

Basmati 370 127 160 149

YRF203 127 167 151

YRF204 127 167 151

Della 127 165 167 149 151

Non-

Fragrant

Cultivars

SCU-Rice-SSR-

1RM42 RM223

Doongara 127 167 151

Amaroo 127 165 149

Echuca 127 165 149

Illabong 127 165 147

Inga 127 165 149

Jarrah 127 165 149

Langi 127 165 149

Millin 127 165 147 149

Namaga 127 165 149

Pelde 127 165 149

YRW4 127 165 149

YRK4 126 165 140 149

Number of Alleles 3 4 6

PIC value 0.48 0.64 0.66

13

5. Discussion

5.1 Markers for sd-1

The Australian semi-dwarf cultivars have a pedigree that indicates the presence of the sd-1 gene. The

exception is Doongara, however this study demonstrates that the source of semi-dwarfism in Doongara

is also sd-1. Four useful PCR based markers, closely linked to sd-1, have been produced. The marker

SCU-Rice-SSR-2 has the greatest immediate utility. Alleleic detection for this marker simply involves

size separation and the marker was able to distinguish between a reasonable proportion of the tall and

semi-dwarf samples tested in this investigation.

A suitable polymorphic marker was detected for most of the pair-wise comparisons between the semi-

dwarf and tall varieties. Most of the semi-dwarf or tall varieties had comprehensively different alleles

for the markers SCU-Rice-SSR-2 or SCU-Rice-STS-S13312, for 9 to 12 varieties from the opposing

height category. However, the marker alleles from the tall varieties of Pelde, Goolarah, Della and

YRF203 were only different to two semi-dwarf cultivars.

The lack of polymorphisms between some of the tall and semi-dwarf cultivars is not unexpected due to

the narrow genetic base of the Australian industry (reviewed by Ko et al., 1994) and the high levels of

genetic similarity between Australian varieties (Garland et al., 1999). Conversely, because the

Australian cultivars are highly related, it could be expected that markers linked to sd-1 would be

different between the tall and semi-dwarf lines. Possible explanations for the lack of polymorphism

between some of the tall and semi-dwarf Australian lines include:

1) Recombination between the linked markers present in the tall parent lines and sd-1

during cultivar development;

2) Lack of polymorphism between the tall and semi-dwarf parent lines originally used to

produce the cultivars analysed; or

3) Through mutation.

The lack of alleleic differences between the genetically divergent cultivars of Doongara (semi-dwarf)

and Khao Dawk Mali 105 (tall), certainly supports the possibility that a lack of polymorphism existed

between parent lines. Recombination between loci within 1 cM is also quite possible given the number

of meiotic events that would occur over the several years needed for cultivar development. Selection for

certain traits, either naturally or as part of the breeding program, could also greatly increase the chance

of producing cultivars where recombinations have occurred between sd-1and linked loci. It is of interest

to note that agronomic traits and QTLs that would be under selective pressure have been identified in

the same genome region as sd-1. These traits include, shattering resistance (Oba and Kikuchi,1991);

number of panicles per plant, number of filled grains per panicle, total number of spikelets per panicle,

% spiklet fertility, 1000 grain weight, grain weight per plant (Zhuang et al., 1997); and heading date

(see Ricegenes consensus map).

In order to confirm the cause for the lack of universality of different alleles to distinguish between tall and semi-

dwarf cultivars, we would need to:

1) Know what alleles were present in the parent cultivars,

2) Observe nucleotide instability to demonstrate mutational causes, and

14

3) Know the associations between allele types and the degree of expression of the

quantitative characteristics, in order to assess the effect of selective pressures on

recombination in the vicinity of sd-1. Obviously more work is necessary to identify the

mechanism involved. Undertaking such an investigation would be of interest, especially

for the assessment of marker stability.

Further markers for sd-1 may be needed for some crosses between semi-dwarf and tall varieties. As

indicated by the success of this investigation, markers could be developed from the numerous other

clones and STS linked to sd-1. Clones genetically mapped between G370 and C86 (Harushima et al.,

1998) and the clones RZ739 and RZ161 from the Ricegenes consensus map

(http://genome.cornell.edu/rice/quickqueries/) could be included. These genetic maps are also suitable

resources for the development of further markers for fgr.

5.2 Markers for fgr

Given that SCU-Rice-SSR-1, RM223 and RM42 are suitable markers for all sources of fragrance (fgr)

and the large proportion of fragrant and non-fragrant lines distinguished, these markers will be highly

useful in the Australian rice breeding program and in rice breeding programs in general. Della was the

only fragrant variety tested with low levels of separation from the non-fragrant lines. This was due to

the identification of two common alleles for the markers RM223 and RM42. Markers for fgr in Della

individuals will be distinguishable from alleles in non-fragrant lines in some cases, depending on which

allele(s) for the marker is present for the individual specimen. The heterogeneity in the Della sample was

not unexpected as it is susceptible to out-crossing (Garland et al. 1999). As heterogeneity was detected

in several samples, it is recommended that further screening be performed for different accessions or

seed samples to make the allelic characterisation comprehensive. In addition, it would be wise to verify

polymorphisms between parents of a cross before a program of marker aided selection is undertaken.

To summarise, four useful PCR based, co-dominant markers, have been produced for sd-1, the semi-

dwarfing gene utilised in the Australian industry. Three useful PCR based co-dominant markers, have

also been produced or identified for fgr, the gene responsible for the major component of fragrance in

rice. The markers for fgr and sd-1 will be useful in the Australian rice-breeding program and in rice

breeding programs in general.

15

6. Implications

It is envisaged that the markers developed in this study will be used for accelerated backcrossing and for

selection during inbreeding. Marker assisted backcrossing can halve the number of backcrosses

necessary to incorporate a gene of interest into a preferred genetic background (reviewed by Dudley,

1993). Marker assisted backcrossing is especially attractive for recessive genes, like sd-1 and fgr. There

is no need to identify heterozygous individuals in the backcross generations by traditional methods. The

heterozygous individuals carry the gene of interest and are needed for the production of the next

backcross. The traditional genotyping procedure is time consuming and involves the detection of

segregation in progeny produced by selfing individuals from each backcross generation.

A marker assisted exercise to incorporate fgr or sd-1 into a desirable genetic background would involve

a cross between a superior cultivar and one possessing the gene of interest, followed by repeated

backcrossing of selected progeny to the superior cultivar. Progeny would be selected, before maturity

and without genotype assessment, that had the lowest genome composition from the varietal source of

fgr or sd-1. Genome composition would be assessed and progeny selected based on the presence of the

lowest proportion of alleles from the varietal source of fgr or sd-1, for markers unlinked or distantly

linked to the gene of interest, and the presence of alleles for the markers closely linked to those genes.

The number of backcrosses are greatly reduced when markers for the recurrent parent are incorporated

into a marker assisted backcrossing exercise (reviewed by Dudley, 1993).

Markers for other traits of interest could also be incorporated into such an exercise. The Australian rice

industry is currently trying to develop a basmati style variety suitable for Australian conditions.

Basmati style cultivars are fragrant, experience grain elongation with little increase in breadth after

cooking, have low to moderate amylose content (<25%), and a soft texture (Ahn et al., 1993). A topical

example of accelerated backcrossing can be provided using the development of a basmati style variety

and utilizing markers for sd-1, fgr, markers for other traits of interest and randomly distributed markers.

Marker assisted selection may provide the power necessary for the development of productive semi-

dwarf Basmati style cultivars given the low levels of success achieved by traditional breeding.

Difficulties experienced, include a lack of desirable recombinants and reversion to the recurrent parent

type in backcrossing exercises (reviewed by Ahn et al., 1993). This may reflect problems with

segregation distortion biased towards the genotype of the recurrent parent. Markers will be able to

identify the possibly lower than expected numbers of individuals possessing important genomic regions

from the basmati cultivar.

The initial cross, in our example of marker assisted backcrossing, could involve a highly productive

semi-dwarf Australian long grain cultivar, with grain quality characteristics of greatest similarity to

basmati rice, crossed with a traditional basmati cultivar. F2 progeny would be selected based on the

presence of marker alleles for sd-1, fgr, and the least proportion of basmati derived alleles for markers

in other genome regions. The selection of alleles for markers linked to other specific genes that could be

incorporated into this process, include 2 that are important to cooking quality. Developed microsatellite

markers for the Waxy gene, a gene that controls amylose content (Ayres et al., 1997, Tan et al., 1999),

should definitely be utilized. Markers linked to the genome region containing the starch branching

enzyme III (Harrington et al., 1997), a gene that controls the relative proportions of amylose and

amylopectin in rice grains, should also be incorporated into the exercise.

Progeny could also be selected based on the presence of alleles for QTL loci from the parent line

possessing the superior phenotype for the trait of interest. A major QTL for cooked-kernel elongation

16

has been identified on chromosome 8 (Ahn et al., 1993). It would probably be appropriate to select

progeny that contain alleles for markers in this genome region that originated from the basmati parent.

Markers for other QTL, such as those traits with numerous QTL of low effect (see review by Yano and

Sasaki, 1997) could also be considered. However, as indicated by Redona and Mackill (1996) alleles for

markers linked to QTLs that are associated with superior measurement for the traits, do not necessarily

come from the superior parent. The appropriate alleles for selection will be cross specific and possibly

environmentally influenced. These markers will need to be identified through empirical studies.

17

7. Recommendations

1) It is recommended that the markers for sd-1 be applied to marker assisted backcrossing. The marker

assisted conversion of tall varieties, for example Kyeema or Pelde, to semi-dwarfs is an example.

The markers for sd-1 are probably not of great benefit to the selection of semi-dwarfs during

inbreeding. Visual selection in the standard breeding program is probably sufficient.

2) Markers for fgr should be applied to selection in both inbreeding and backcrossing exercises due to

the time consuming and inaccurate nature of assessing fragrance.

3) Markers for both traits should be assessed on the parents of the initial cross to confirm marker

polymorphism, on the F1 plants to confirm the success of the cross and in subsequent generations.

Marker assessment of the F2 generation is particularly important during inbreeding to ensure that

recombination does not occur between the gene of interest and the marker and thus reduce the

accuracy of detection of the gene in later generations.

4) Heterogeneity was detected in several samples, for some of the markers linked to genes of interest

identified in this study, and it is recommended that further screening be performed. Alleleic

screening of different accessions or seed samples, that may be involved in marker assisted selection

programs, should be undertaken in order to make the allelic characterisation comprehensive. In

addition, it would be wise to verify polymorphism between parents of a cross before a program of

marker aided selection is undertaken.

5) There were some potential crosses identified that did not possess a suitable marker for either sd-1 or

fgr. Given the success of this investigation further screening of linked RFLP probes would probably

be the most efficient method for the development of further markers for these two genes.

6) A comprehensive fingerprinting exercise should now be undertaken to extend the work of Garland et

al., 1999). There are hundreds of microsatellite markers now available for rice. Alleles for 100 or

more microsatellite markers, distributed through the genome, should be determined for numerous

individuals, from several accessions, for each cultivar or rice sample of importance to the

Australian rice-breeding program. The presence of a database containing the alleleic

characterisations would allow for the selection of a suite of markers which are polymorphic between

Australian varieties, breeding lines and breeding samples, for use in germplasm identification,

genetic mapping, and marker assisted selection programs.

7) The challenge is to maximise the potential benefits of markers through the appropriate integration of

marker assisted selection into the breeding program. It may be necessary to restructure the breeding

program to obtain maximum benefits. For example, it will be necessary to screen F2 individuals

during inbreeding exercises. In general, individual plants will need to be identified. It may be

necessary to transplant by hand or reduce the density of planting to enable tagging of individuals or

groups of individuals for accurate re-sampling in the selection process.

8) In order to facilitate the integration of markers into the breeding program it may be appropriate to

create a new position of Molecular Plant Breeder. This position would work in close collaboration

with breeders and cereal chemistry. The initial appointment could involve a specific task. For

example, marker assisted development of a suitable Basmati style variety through accelerated

backcrossing. Suitable procedures would be developed and refined for possible extension of marker

assisted selection into the wider breeding program.

18

19

8. References

Ahn SN, Bollich CN, McClung AM, Tanksley SD (1993) RFLP analysis of genomic regions

associated withcooked-kernel elongation in rice. Theor Appl Genet 87: 27-32

Ahn SN, Bollich CN, Tanksley SD (1992) RFLP tagging of a gene for aroma in rice. Theor Appl

Genet 84: 825-828

Ayres NM, McClung AM, Larkin PD, Bligh HFJ, Jones CA, Park WD (1997) Microsatellites and a

single-nucleotide polymorphism differentiate apparent amylose classes in an extended pedigree

of US rice germ plasm. Theor Appl Genet 94: 773-781

Carnahan HL, Johnson CW, Tseng ST (1978) Registration of 'M7' Rice. Crop Sci. 18:356-357

Causse MA, Fulton TM, Cho YG, Ahn SN, Chunwongse J, Wu K, Xiao J, Yu Z, Ronald PC,

Harrington SE,Second G, McCouch SR, Tanksley SD (1994) Saturated molecular map of the

rice genome based on a interspecific backcross population. Genetics 138: 1251-1274

Chen X, Temnykh S, Xu Y, Cho YG, McCouch SR (1997) Development of a microsatellite

framework map providing genome-wide coverage in rice (Oryza sativa L.). Theor Appl Genet

95: 553-567

Cho YG, Eun MY, McCouch SR, Chae YA (1994) The semidwarf gene, sd-1, of rice (Oryza sativa

L.). II.Molecular mapping and marker-assisted selection. Theor Appl Genet 89: 54-59

Cho YG, McCouch SR, Kuiper M, Kang M-R, Pot J, Groenen JTM, Eun MY (1998) Integrated map

of AFLP,SSLP and RFLP markers using a recombinant inbred population of rice (Oryza sativa

L.). Theor Appl Genet 97: 370-380

Dudley JW (1993) Molecular markers in plant improvement: Manipulation of genes affecting

Quantitative traits. Crop Sci. 33:660-668

Fukuoka S, Inoue T, Miyao A, Monna L (1994) Mapping of sequence-tagged sites in rice by single

conformation polymorphism. DNA Research 1:271-277

Garland SH, Lewin L, Abedinia M, Henry R, Blakeney A (1999) The use of microsatellite

polymorphisms for the identification of Australian breeding lines of rice (Oryza sativa L.).

Euphytica 108: 53-63

Ghareyazie B, Huang N, Second G, Bennett J (1995) Classification of rice germplasm I. Analysis

using ALPand PCR-based RFLP. Theor Appl Genet 91: 218-227

Grosch W, Schieberle P (1997) Flavor of cereal products – A review. Cereal. Chem 74: 91-97

Harrington SE, Bligh HFJ, Park WD, Jones CA, McCouch SR (1997) Linkage mapping of starch

branching enzyme III in rice (Oryza sativa L.) and prediction of location of orthologous genes

in other grasses. Theor Appl Genet 94: 564-568

20

Harushima Y, Yano M, Shomura A, Sato M, Shimano T, Kuboki Y, Yamamoto T, Lin SY, Antonio

BA, Parco A, Kajiya H, Huang N, Yamamoto K, Nagamura Y, Kurata N, Khush GS, Sasaki T

(1998) A high-density genetic linkage map with 2275 markers using a single F2 population.

Genetics 148: 479-494

Ko HL, Cowan DC, Henry RJ, Graham GC, Blakeney AB, Lewin LG (1994) Random amplified

polymorphic DNA analysis of Australian rice (Oryza sativa L.) varieties. Euphytica 80: 179-

189

Li Z, Pinson SRM, Stansel JW, Park WD (1995) Identification of quantitative trait loci (QTLs) for

heading date and plant height in cultivated rice (Oryza sativa L.). Theor Appl Genet 91: 374-

381

Lorieux M, Petrov M, Huang N, Guiderdoni E, Ghesquière A (1996) Aroma in rice: genetic analysis

of quantitative trait. Theor Appl Genet 93: 1145-1151

Lu C, Shen L, Tan Z, Xu Y, He P, Chen Y (1996) Comparative mapping of QTLs for agronomic traits

of rice across environments using a double haploid population. Thoer Appl Genet 93: 1211-

1217

Mackill DJ, Rutger JN (1979) The inheritance of induced-mutant semidwarfing genes in rice. The

Journal of Heredity 70: 335-341

Maeda H, Ishii T, Mori H, Kuroda J, Horimoto M, Takamure I, Kinoshita T, Kamijima O (1997) High

density molecular map of semidwarfing gene, sd-1, in rice (Oryza sativa L.). Breeding Science

47: 317-320

Murai M, Shinbashi N, Sato S, Sato K, Araki H, Ehara M (1995) Effect of the dwarfing gene from

Dee-geo-woo-gen on culm and internode lengths, and its response to fertilizer in rice. Breeding

Science 45: 7-14

Nei M (1973) Analysis of gene diversity in subdivided populations. Proc Natl Acad Sci USA 70:

3321-3323

Oba S, Kikuchi F (1991) Grain-shattering gene linked with semidwarfing gene sd-1.International Rice

Research Institute (1991) Rice genetics II. P.O. Box 933, Manila, Philippines.

Reinke RF, Welsh LA, Reece JE, Lewin LG, Blakeney AB (1991) Procedures for the quality selection

of aromatic rice varieties. International Rice Research Newsletter 16(5): 10-11

Redona ED, Mackill DJ (1996) Mapping quantitative trait loci for seedling vigor in rice using RFLPs.

Theor Appl Genet 92: 395-402

Rutger JN, Azzini LE, Brookhouzen PJ (1986) Inheritance of semidwarf and other useful mutant

genes in rice. International Rice Research Institute (1986) Rice genetics. P.O. Box 933 Manila,

Philippines: 261-271

Tan YF, Li JX, Yu SB, Xing YZ, Xu CG (1999) The three important traits for cooking and eating

quality of rice grains are controlled by a single locus in an elite rice hybrid, Shanyou 63. Theor

Appl Genet 99: 642-648

21

Tragoonrung S, Sheng JQ, Vanavichit A (1996) Tagging an aromatic gene in lowland rice using bulk

segregant analysis. International Rice Research Institute. 1996. Rice genetics III. Proceedings

of the third international rice genetics symposium. 16-20 Oct 1995. Manila: 613-618

Weining S, Henry RJ (1995) Molecular analysis of the DNA polymorphism of wild barley (Hordeum

spontaneum) germplasm using the polymerase chain reaction. Genetic Resources and Crop

Evolution 42: 273-281

Widjaja R, Craske JD, Wootton M (1996) Comparative studies on volatile components of non-

fragrant and fragrant rices. J Sci Food Agric 70: 151-161

Williams MNV, Pande N, Nair S, Mohan M, Bennett J (1991) Restriction fragment length

polymorphism analysis of polymerase chain reaction products amplified from mapped loci of

rice (Oryza sativa L.) genomic DNA. Theor Appl Genet 82: 489-498

Xu J, Constantino SV, Magpantay G, Bennett J, Sarkarung S, Huang N (1998a) Classification of rice

germplasm II. Discrimination of indica from japonica via analysis of amplicon length

polymorphisms. Plant Cell Reports 17: 640-645

Xu Y, Shen L, McCouch SR, Zhu L (1998b) Extension of the rice DH population genetic map with

microsatellite markers. Chinese Science Bulletin 43: 149-153

Yan J, Zhu J, He C, Benmoussa M, Wu P (1998) Molecular dissection of development behavior of

plant height in rice (Oryza sativa L.). Genetics 150: 1257-1265

Yano M, Sasaki T (1997) Genetic and molecular dissection of quantitative traits in rice. Plant

Molecular Biology 35: 145-153

Zhuang J-Y, Lin H-X, Lu J, Qian H-R, Hittalmani S, Huang N, Zheng K-L (1997) Analysis of QTL ×

environment interaction for yield components and plant height in rice. Theor Appl Genet 95:

799-808