INTRO

Cell exlular euthe estexpansorgan sthe conthe pla

The tion anexpanswall (Gnormaland havorientation of the cellulose has been attributed to the corticalmicrotubules of the cytoskeleton. This is based on two typesof evidence. First, cortical microtubules are normally foundaligned in the same orientation as the cellulose microfibrils.Second, disruption of the microtubules has led to abnormalcellulose microfibril patterns and abnormal cell expansion(Giddings and Staehelin, 1991). A model has been proposed inwhich microtubules act like tracks between which the cellulosesynthesis complex moves (Giddings and Staehelin, 1991). Inthis way the orientation of the microtubules dictates the orien-tation of the cellulose microfibrils without any direct contactbetween the two.

Regulation of the extent of expansion is less well under-stood. The driving force for expansion is turgor pressuregenerated by the presence of a high osmotic potential in thevacuoles within the expanding cell. Water movement into thecell is thought to be passive, driven only by the difference inosmotic potential (Cosgrove, 1993). Expansion occurs by

along amer thees cs tof thll coor p

ressu

he ugge butvidual cells is subject to a

higher regulation that has predetermined the final shape of theorgan (Kaplan and Hagemann, 1991). In animal cells, there isalso little known about the regulation of cell expansion.Although most animal cells do not have cell walls, currentstudies of the primary plant cell wall indicate that it resemblesthe extracellular matrix surrounding animal cells (Roberts,1989). Therefore, there are likely to be aspects of the regula-tion of cell expansion that are common to both animals andplants as well as some that are unique to each.

The

Arabidopsis root is particularly well suited to studies ofcell expansion. Cell expansion in the body of the root isregulated so that the primary axis is normally parallel to thedirection of growth. This makes for long cylindrical cells.Polarized expansion in the longitudinal orientation (which wewill refer to as “elongation”) occurs primarily in a region atthe tip of the root known as the elongation zone (Steeves andSussex, 1989).

By screening mutagenized plants for roots that have an

DevelopmPrinted in

Regulaplant osentingroot o

mutanculatioresulteexpansthe mu

t onble phe

athw

pansisis

SUMM

Cond

Marie-

Departm

*Present e-mail: be

g the length of the wall.nd reforming of bondsnetworks, one made of other made of pectinonstant synthesis and

the cell wall. At somee wall so that no moremponents are added. Atressure. When the wallre, expansion ceases

genetic control of cellsted that plant form is

at the level of the organ.

1237

a change in the orien- mutant combinationsnotype suggesting thatay or encode partially

on, cell polarity,

ent 121, 1237-1252 (1995)Great Britain © The Company of Biologists Limited 1995

tion of cell expansion is essential to the formation ofrgans. We have characterized 21 mutations, repre-

expansion may not be dependentation of the microtubules. Dou

ARY

itional root expansion mutants of Arabidopsis

Theres Hauser*, Atsushi Morikami and Philip N. Benfey

ent of Biology, New York University, 1009 Main Building, New York, N.Y. 10003, USA

address: Centre of applied Genetics, Universitaet fuer Bodenkultur, Gregor Mendel Str. 33 A-1180 Vienna, Austrianfeyp@ACFcluster.nyu.edu

DUCTION

pansion is a key parameter of development in multicel-caryotes. Regulation of cell expansion is required for

ablishment of cell shape and polarity. In plants, cellion is of paramount importance in the determination ofhape. To the extent that organ shape affects function,trol of cell expansion can be critical to the survival ofnt. two principal parameters of cell expansion are orienta-d extent. There is evidence that the orientation of cellion is controlled by the cellulose microfibrils in the cell

iddings and Staehelin, 1991). These molecules arely found arranged transverse to the direction of growthe been likened to hoops around a barrel. Control of the

addition of cell wall components This requires a constant breakinwithin the two co-extensive polycellulose and hemicellulose and(Roberts, 1989). It also requirtransport of structural componentpoint there is a “rigidification” obonds are broken and no more wathis point the wall resists the turgpressure equals the turgor p(Cosgrove, 1993).

Very little is known about texpansion. In fact, it has been sdetermined not at the cellular levelIn this theory, expansion of indi

six loci, that cause abnormal cell expansion in thef Arabidopsis thaliana. The phenotype of thesets is conditional upon the rate of root growth. Cal-n of cell volumes indicated that the mutationsd in defects in either the orientation or the extent ofion or in both. Analysis of cortical microtubules intants suggested that a shift in the orientation of cell

resulted in loss of the conditionalthe genes may act in a similar predundant functions.

Key words: root development, cell exmicrotubules, core mutants, Arabidop

1238

increashave deand Begeneticing sixreporteupon thwere nDetermCORE CorticamutantsCORE genes mfunctio

MATE

Plant sStrains (selewskArabido(cob-1),previousGenetic bein andfied COwere isoincludin1991), (EMS), mutagenroots wiprogenyphenotyBenfey,CORE mof rootsto the re

MappinThe chrmeasurimicrosamorphicand visi(Col), p(Col), qu(Ler). pWS. Theregatingindepenmodifiedmicrosaacrylam

Map disThe reclethal/st(both ec

Recomposition

For liO.8% w(microsa

ed diameter we and others have identified mutants thatregulated cell expansion (Benfey et al., 1993; Hausernfey, 1993; Baskin et al., 1992). Here, we present a and cellular characterization of 21 mutants represent- loci. Three of these loci have not been previouslyd. The phenotype of all 21 mutants was conditionale rate of root growth. For this reason these mutants

amed Conditional Root Expansion (CORE) mutants.ination of cellular dimensions suggested that differentgenes control the orientation and extent of expansion.l microtubule orientation was determined for all. The loss of the conditional phenotype when themutations were combined indicated that the COREay act in a similar pathway or provide redundant

ns.

(microsatellite). For pom2, 4% with GPA1 (CAPS), 12% with m429(CAPS). For cud, 16% with DHSI (CAPS), 26% with AG (CAPS).

Confocal microscopy, immunolocalization ofmicrotubules, nuclear staining and image analysisSeedlings were stained with acriflavine as described (Aeschbacher etal., 1995). Hypocotyls were stained with two additional 30 minutesteps in 95% ethanol after fixation, to reduce the chlorophyll content.Microtubule immunolocalization was a modification of the method ofColasanti et al. (1993); Doonan et al. (1993) and performed asfollows. Seedlings were fixed in a solution of 4% paraformaldehyde,5% DMSO, 0.1% glutaraldehyde, 50 mM PIPES, 5 mM EGTA, 5 mMMgSO4, pH 7.0 by vacuum infiltration at 4˚C for 1 hour. Seedlingswere then washed three times in wash buffer (50 mM PIPES, 5 mMEGTA, 5 mM MgSO4, pH 7.0) for 30 minutes. Seedlings were thentransferred to a solution of 2% driselase (Sigma), which was vacuuminfiltrated for 1 hour at room temperature. The seedlings were washedtwice followed by a 10 minute incubation with 100% methanol at −20˚C. Seedlings were washed twice then infiltrated for 15 minutes

M.-T. Hauser, A. Morikami and P. N. Benfey

RIALS AND METHODS

i

R

M

p

ont

o

o

d

ti

oeo

st

trains, growth conditions and genetic analysisor ecotypes) Landsberg erecta (Ler), Columbia (Col), Was-ja (Ws) of Arabidopsis thaliana (L) were obtained from thepsis Stock Center (Columbus, OH). The mutants cobra-1lion’s tail-1 (lit-1) and pom-pom1-1 (pom1-1) were grown asly described (Benfey et al., 1993; Hauser and Benfey, 1993).crosses were performed essentially as described by Schiefel- Somerville (1990); Benfey et al. (1993). The newly identi-E mutants, quill (qui), cudgel (cud), and pom-pom2 (pom2)

lated among 145,000 plants grown from different populationsg: pools of T4 T-DNA mutagenized WS lines (Feldmann,

2 Col seeds mutagenized with ethyl methanesulfonateM2 Col seeds mutagenized with X-rays and M2 Col seedsized with fast neutrons (Lehle Seeds). Plants that exhibitedth a diameter larger than wild-type were isolated and their were retested in the M3 generation for the conditionale on nutrient agar plates as previously described (Hauser and1993). Pairwise crosses were made between the differentutants to determine complementation groups. Regeneration

from calli that had been induced from shoots was accordinggeneration regime described by Valvekens et al. (1988).

gmosomal location of the mutant genes was determined byg the recombination frequency between the mutant gene and

with 0.5% Nonidet P-40 in wash buffer. Seedlings were washed twicethen labelled for at least 1 hour at room temperature with 300 µlwashing buffer containing 3% BSA (Sigma) and a 1:75 dilution ofthe rat monoclonal antibody YOL1/34 anti-yeast-α-tubulin subunitantibody (Sera Labs). The antibody recognizes most eucaryotic α-tubulins including yeast, mammal, and plant. The antibody has alower affinity to native microtubules and a higher affinity to fixedmicrotubules. Seedlings were washed three times then incubated for1 hour at room temperature in 300 µl wash buffer containing 3% BSA(Sigma) and a 1:100 dilution of a tetramethylrhodamine-5-isothio-cyanate-(TRITC) labelled secondary antibody (anti-rat IgG, Sigma).Seedlings were rinsed in a wash buffer/glycerin series (30%, 50%,70%) and the roots were mounted on a microscope slide with Citifluor(Ted Pella). Confocal laser scanning microscopy (CLSM) wasperformed with a Zeiss confocal microscope with a helium line at 543nm.

For root tip staining, young seedlings were fixed in 3.7% formalde-hyde in PBT buffer (130 mM NaCl, 10 mM NaHPO4 buffer, pH 7.0,0.1% Tween 20) for at least 4 hours at room temperature, washedthree times with PBT, and stained with 2 µM 4,6-diamindino-2-phenylindole (DAPI, Sigma) in PBT buffer for 10 minutes. Excessstain was removed by washing with PBT, then seedlings were washedin 50% and 70% glycerin in PBT and mounted with Citifluor (TedPella). The root tips were analyzed with a Zeiss confocal microscopewith the Argon/Krypton laser line of 488 nm. Images were stored asTIFF files, manipulated in Adobe Photoshop and printed on a KodakXL7700 dye sublimation printer.

ellite markers (Bell and Ecker, 1994), cleaved amplified poly- sequence markers (CAPS) (Konieczny and Ausubel, 1993)ble markers. The homozygous mutants lit-1 (Col), cob-1m1-1 (Ws), pom1-2 (Col), cud-1 (Col), pom2-1 (Col), qui-1i-2 (Ws) were mapped by crossing them to Landsberg erectam1-1 was also crossed to Col and lit-1 was also crossed to resulting F1 plants were selfed to generate F2 seedlings seg-

for the mutant phenotype. DNA was isolated from 15-20ent axenically grown mutant seedlings using a slightly method of Konieczny and Ausubel (1993). We modified theellite mapping protocol to separate the fragments on 5% poly-de gels (19:1) in 1× Tris Borate EDTA (TBE), pH 8.3 buffer.tances were calculated from recombination frequencies (RF).mbination frequency of cud-1 which is male gametophyticrile was calculated as the number of homozygous plantstypes) of the total F2 plants exhibiting the mutant phenotype. bination frequencies were as follows, approximate map are listed in Table 1: , 18.8% with DFR (CAPS), 20% with LFY (CAPS). For cob,ith LFY. For qui, 8% with PVV4 (CAPS), 10% with nga59tellite). For pom1, 4.7% with nga59, 1.2% with nga63

Cell dimensions, cell volume calculations and growthcondition shift experimentsCellular dimensions were determined from CLSM pictures taken inthe differentiation zone of wild-type and CORE mutant roots.Between 150 and 200 cells were measured from four to 14 differentroots for wild type and each mutant. For hypocotyls, approximately80 cells were measured from five different plants. Cell volumes werecalculated assuming that epidermis and cortex cells were approxi-mately cylindrical for both wild type and mutants. This was based onobservations of numerous longitudinal and fresh transverse sections.In transverse sections, the cross-sectional areas of endodermis cellshad a rectangular shape. For these cells, the circumferential wall wastaken as the length and the radial wall was used for the width so thatvolume = circumferential length × radial width × cell length. Ascontrols for these measurements, cross-sectional areas were obtaineddirectly from digitized images of fresh sections as described previ-ously (Benfey et al., 1993). Areas obtained in this way were approx-imately equivalent to areas calculated from diameters measured fromoptical sections.

For the growth condition shift experiments, seeds were planted on

C

nutrient agar medium supplemented with 0.1% sucrose. After 8 daysgrowth, root epidermal cells were marked with carbon grains asdescribed by Okada and Shimura (1990). Seedlings were then trans-ferred to either nutrient agar medium supplemented with 4.5% sucroseor 0.1% sucrose. At intervals of 4, 8, 12, 24 and 36 hours root tipswere photographed under a Nikon SMZ-U stereomicroscope.Tracings of root tips and the location of carbon grains were made fromprojections of 35 mm slides.

RESULTS

Phenotypic and genetic analysis of root expansionmutants with conditional phenotypesTo investigate the genetic basis of root morphogenesis wescreened progeny of mutagenized seeds for abnormal rootexpansion. We isolated 21 mutants that had reduced root lengthand exhibited a large variability in root diameter along thelength revealelion’s tviouslywe desthese m(pom2)cob annant (Bmale-spollinapollen 1 mutamal locmarker

The phenotyperoots growing atnamed this clas(CORE) mutantsrate is dependencentration of sucrtemperature conBenfey, 1993). Wgrown under congrowth they exhitype growth ratesmedium with a lotemperature and lhad relatively nrevealed that the relatively normalIn particular, the containing 0.5%

This flower pheextremely low fe

To determine wfor the developmroots induced frolit-1, cob-1, cud-mutant phenotyp(data not shown)stimulus in a maof the direction o

Root morpholoThe structure of analyzed by cosections through in cell shape amdetermine how tmeasured cellula

Tabl

Mutant/ allele

lit-1cob-1cob-2cob-3qui-1 I 0 Col x-ray recessivequi-2 Ws T-DNA recessivequi-3 Col EMS recessivepom1-1 I 5 Ws T-DNA recessivepom1-2 Col EMS recessivepom1-3 Ws T-DNA recessivepom1-4 Ws T-DNA recessivepom1-5 Ws T-DNA recessivepom1-6 Ws T-DNA recessivepom1-7 Ws T-DNA recessivepom1-8 Ws T-DNA recessivepom1-9 Col EMS recessivepom1-10 Col EMS recessivepom1-11 Col EMS recessivecud-1 IV 76 Col x-ray dominant†pom2-1 II 60 Col fast neutron recessivepom2-2 Col fast neutron recessive

*Approximate map positions determined by mapping with CAPS andmicrosatellite markers (see Materials and Methods).

†cudgel is also male-gametophytic lethal. This may be due to a linkedmutation resulting perhaps from a deletion induced by the X-ray mutagenesis.

1239onditional root expansion mutants

of all 21 mutants was conditional upon the a maximum rate. As noted above, we haves of mutants, Conditional Root Expansion. We have shown that wild-type root growtht on various factors. These include the con-ose in nutrient agar media as well as light andditions (Benfey et al., 1993; Hauser and

hen the 21 root expansion mutants wereditions that provide maximal wild-type rootbited the expanded phenotype (Fig. 1). Wild- can be diminished by growing the plants onwer concentration of sucrose or by reducingight. Under these conditions the mutant rootsormal diameters (Fig. 2). Optical sectionscellular morphology of the mutants was also under these permissive conditions (Fig. 3).cellular morphology of cob-1 grown on media sucrose was difficult to distinguish from

of the root. Reciprocal crosses among the mutantsd six complementation groups. Three loci, cobra (cob),ail (lit) and pom-pom1 (pom1) have been described pre- (Benfey et al., 1993; Hauser and Benfey, 1993). Herecribe the isolation and characterization of new alleles ofutants as well as three new loci, quill (qui), pom-pom2 and cudgel (cud) (Table 1). All mutations except ford cud-1 were found to be recessive. cob is semidomi-enfey et al., 1993). cud-1 was found to be dominant and

terile as evidenced by 100% wild-type progeny whenting wild type and 50% mutant progeny when receivingfrom wild type. The segregation of the progeny of cud-nt plants was 1:1 mutant to wild type. The chromoso-ations of the loci were determined by use of moleculars (Table 1).

wildtype (Fig. 3B) (Benfey et al., 1993). This was also true ofpom2 (data not shown). Under these conditions cob-1, pom1-8 (Fig. 2B), pom1-7 (Fig. 3D), qui-2 (Fig. 2C) and cud-1 hadreduced root length but no apparent radial cell expansion. Theonly mutant with any detectable radial cell expansion whengrown under these conditions was lit-1 (Figs 2D, 3E and 3F).However, this was dramatically reduced from the extent ofexpansion when grown on higher sucrose levels (compare Figs2D and 1B).

To test whether the expansion at higher sucrose levels wasinfluenced by the increased osmotic potential, we grew theCORE mutants on two different concentrations of mannitol, anon-metabolized carbohydrate. We found that there was no dif-ference in root expansion in any of the CORE mutants whengrown on low (0.25%) or high (2.25%) mannitol-containingmedium. The root phenotype of the CORE mutants was similarto the phenotype on normal growth medium supplementedwith 0.5% sucrose.

There was no apparent abnormality in the aerial phenotypeof CORE mutant seedlings when grown under restrictive orpermissive conditions (Figs 1 and 2). When grown on soil, themutants lit and pom1 exhibited reduced growth and smallerleaves. The flower phenotypes of the CORE mutants werenormal with the exception of pom2 which had reduced filamentelongation as well as thickening of the style (data not shown).

e 1. Name and origin of conditional root expansionmutants

Chromosome Map Geneticnumber* position* Ecotype Mutagen status

V 70 Col EMS recessiveV 90 Col EMS semidominant

Col x-ray semidominantWs T-DNA semidominant

notype was probably the reason for thertility of the pom2 mutants.

hether the mutant phenotypes were specificental context of the seedling, we examinedm callus culture. Callus-induced roots from1, pom1-1, qui-1 and pom2-1 displayed thee when they grew under restrictive conditions. All the mutants responded to a gravitropicnner similar to wildtype although the changef growth was slower than wildtype.

gy and cell volume of CORE mutantsthe expanded root of the CORE mutants wasnfocal laser scanning microscopy. Opticalthe differentiation zone revealed differencesong the different mutants (Fig. 4B-G). To

he shape changes affected cell volumes wer dimensions for the epidermis, cortex and

1240

endodermdistingui(Fig. 4) determin

The cedramaticvaried gcell lengof cell exsignificacob-1, qthere wepidermalength aequivalecells the (Table 2approxim

Cell v(See Macud-1 thedescribed

M

Fig. 1. Wlion’s tail

.-T. Hauser, A. Morikami and P. N. Benfey

r

nua

nn

)

t

ice versa. From the volume calculations, the mutants couldvided into three groups. The first class had volumes ofrmis, cortex and endodermis that were significantly

ose, 11 days after germination (DAG). (A) Wild-type WS, (B)om2-1. Bar, 1 mm.

is (Table 2). The cell walls in the stele were hard tosh in the optical sections due to high autofluorescenceand therefore the cellular dimensions could not be

and vbe diepide

ildtype and CORE seedlings on medium supplemented with 4.5% sucr-1, (C) cobra-1, (D) quill-1, (E) pom-pom1-1, (F) cudgel-1, (G) pom-p

ed for these tissues.ll length measurements indicated that all mutants hadally reduced cell elongation (Table 2). The mutantseatly in the extent of radial expansion. The ratio ofth to diameter provided an indication of the polaritypansion. In wild-type epidermal cells the length wastly greater than the diameter (Table 2). However, ini-1, and cud-1 the reverse was true, suggesting thats a change in polarity of cell expansion in thel cells (Table 2). For lit-1, pom1-1 and pom2-1 thed diameter of epidermal cells were approximatelyt (Table 2). For wild-type cortex and endodermal

length was also significantly greater than the diameter. For all mutants the cortex and endodermis cells hadately equivalent cell lengths and diameters (Table 2).

olumes were calculated from the cellular dimensionserials and Methods). For the mutants cob-1, qui-1 and volume calculation was based on the shift in polarity above so that the cell length was now the diameter

smaller than wildtype. The mutant lit-1 was the only memberof this class. Cell volumes in lit-1 roots were approximately onequarter those of wildtype (Table 3). This indicates that mutationof the lit gene results in defective control of the extent of cellexpansion so that normal cell volumes are never obtained. Inthe second class, represented by cob-1, the epidermal cellvolume was approximately equal to wildtype (Table 3). Thisindicates that the cob mutation does not have a dramatic effecton the extent of expansion since volume (at least of theepidermis) is conserved. It suggests that the primary defect incob could be a shift in the polarity of cell expansion. The thirdclass of mutants had cell volumes that were significantly greaterthan wildtype for at least one tissue. In qui-1 and cud-1 roots,epidermal tissue had a greater volume than wildtype (Table 3).In both pom1-1 and pom2-1 there was increased volume inepidermis as well as cortex. This suggests that these mutationscause a loss of regulation of the extent of expansion.

The cell volume of the endodermis was reduced in allmutants. qui-1 showed the most dramatic reduction. This may

1241Conditional root expansion mutants

be duecortex calculain wildof the the rooof voluall the

To droot wCOREof the standargraduadiamet1 and was a s

Fig. 2.8, (C) q(H) pom

at mtmmee medleqm

W n medium supplemented with 0.5% sucrose. (A) Wild-type Col, (B) pom-pom1-u -pom2-1 and cobra-1, (F) cudgel-1 and pom-pom1-1, (G) cudgel-1 and quill-1,

ildtype and CORE single and double mutants grown oill-1, (D) lion’s tail-1. Double mutants between (E) pom-pom2-1 and lion’s tail-1. Bar, 1 mm.

to positional constraints of these cells between thend the stele. It is important to note that the range ofed cell volumes was much larger in the mutants thantype (Table 3). This may due to the conditional natureutations which causes a variability in the diameter of

. An alternative explanation is that the tight regulatione and shape has been lost in the different tissues ofutants.

termine if the mutant phenotypes were specific for the examined the dimensions of hypocotyl cells. Theutants differed little from wildtype in the cell lengths

pidermis, cortex and endodermis (Table 4). The large deviation of the cell length data results from the

increase of cell length along the hypocotyl. Ther of the hypocotyl was similar to wildtype for lit-1, cob-ui-1. However, for pom1-1, cud-1 and pom2-1 thereall increase in hypocotyl diameter. This increase was

due primarily to a slight increase in the radial expansion of thetwo cortex layers. The radial expansion of the hypocotyl inthese three mutants (approximately 40%) was still significantlysmaller than the radial expansion of the roots (200-260% forthese three mutants). When grown in the dark to stimulate morerapid elongation of hypocotyl cells there was still no majorincrease in the hypocotyl diameter (data not shown). Weconclude that the radial expansion phenotype of the COREmutants affects primarily root cells.

Apparent cell division rate of mutant rootsUnder conditions that favor higher growth rates for wild-typeroots, the CORE mutant roots are expanded and very short.We wished to determine whether the reduction in cellelongation was sufficient to account for the reduced growthrate of the mutants. Root growth was determined bymeasuring the root length differences between days 4 and 6

1242

after gersevere dlength imutants.cell divicell lengFor the growth a

M mi and P. N. Benfey

Fig. 3. Opthrough thlion’s tailquill-2, (I

.-T. Hauser, A. Morika

mination (Table 5). As would be expected from theefects in cell elongation, root growth (as defined byncrease) was severely reduced in all the CORE Assuming constant cell cycle time, we calculated thesion rate by dividing the root growth per day by theth (plus or minus the standard deviation) (Table 5).mutants qui-1 and pom2-1 the reduction in rootppeared to be due primarily to the reduction of cell

elongation. For the other mutants the shorter root length andthe slower root growth appeared to be a combination ofreduced cell elongation and an apparent reduction in thedivision rate. The most dramatic apparent reduction in celldivision rate was observed in lit-1 and cud-1.

From optical sections of root tips (Fig. 4H-L) we calcu-lated the cell number in the cortex tissue file between theinitial cells and the first cell that exhibited polar or increased

tical sections through the roots of seedlings grown on medium supplemented with 0.5% sucrose. Median longitudinal sectionse differentiation zone of (A) wild-type Ws, (B) cobra-3, (C) quill-2, (D) pom-pom1-7, (E) half of the expanded part of the root of

-1, (F) unexpanded part of same lion’s tail-1 root, (G) double mutant of cobra-1 and pom-pom2-1, (H) double mutant of cudgel-1 and) double mutant of cudgel-1 and pom-pom1-1, (J) double mutant of lion’s tail-1 and pom-pom2-1. Bar, 50 µm.

1243Conditional root expansion mutants

cell expansion. This is approximately equthat has been termed the post-mitotic i(PIG) zone (Baluska et al., 1993). In wconsists of about 20 cells. For the mutan

Table 2. Cellular dim ly expanded roots of wild type and CORE mutantsEndodermis

Epidermis tex length/radial width Genotype length/diameter iameter circumferential length Stele diameter Root diameter

wild type 174±38.5 ±3.5 112.8±2913.2±1.8 ±2.8 6.2±1.4 38.0±4.7 120.5±29

13.5±2.6

lit-1 20.0±9.0 ±7.6 18.4±6.623.6±7.6 ±4.6 13.9±1.9 82.9±13.8 229.6±17

20.0±2.5

cob-1 21.7±8.8 ±6.6 15.4±5.068.9±11.2 ±10.0 15.2±3.2 52.5±8.9 289.7±39.4

21.8±8.8

qui-1 33.7±11.7 ±8.9 21.9±7.065.4±27.8 ±9.7 8.4±1.1 44.8±5.1 259.6±56.5

12.7±2.4

pom1-1 51.7±21.5 ±16.9 27.5±13.044.1±9.7 ±10.7 10.1±3.1 58.2±22 259.8±55.8

17.9±4.9

cud-1 30.2±11.564.9±18.2

pom2-1 59.0±21.534.9±8.0

All measurements are in micrometers.

Table 3. Mean and range* of calculateexpanded root cells

Epidermis Cortex Genotype mean (max-min) mean (max-

wildtype 24 (34-14) 40 (57-22lit-1 6 (10-2) 17 (26-8)cob-1 26 (44-8) 15 (25-5)qui-1 58 (102-14) 15 (15-4)pom1-1 79 (130-28) 63 (104-2cud-1 47 (81-13) 25 (37-13pom2-1 56 (92-21) 64 (96-32

*Calculation of mean used mean values for cell diTable 2. Lower values for range calculated using medeviation for both diameter and length, upper values deviation for both diameter and length.

All measurements are in picoliters.

Length (mm)

wt 1.3±0.4lit-1 0.9±0.2cob-1 1.5±0.4qui-1 0.9±0.2pom1-1 1.3±0.4cud-1 0.8±0.3pom2-1 n.d.

Measurements were made 4 days after germination

ensions of ful

Corlength/d

15618.0

20.432.7

20.130.8

26.526.3

38.845.2

parent reduction in cell cells in this region. Theave a cortex cell number

y in this region. These expansion and differen- specific number of cell

restrictive

re sensitive to the effectsd experiments in which

to restrictive conditions. nutrient medium supple-l cells at the tip of the root

3 319.8±79.2

5 238.2±33.5

cud-1 which have the greatest apdivision rate there were only about 4mutants qui-1, pom1-1 and pom2-1 hof about 9, 7 and 10 respectiveldata suggest that the timing of celltiation is not necessarily linked to adivisions.

Cell expansion upon shifting toconditions To determine if all growing cells weof the cobra mutation we performeplants were shifted from permissiveSeedlings were grown for 8 days onmented with 0.1% sucrose. Epiderma

25.1±6.5 19.4±6.535.5±6.2 14.9±5.9 70.3±15.

19.9±2.6

47.1±16.1 32.8±10.541.5±5.3 11.5±1.9 54.7±13.

16.1±4.2

d volumes of fully

Endodermismin) mean (max-min)

) 9 (14-4)6 (8-3)6 (9-2)3 (4-1)

1) 5 (9-1)) 6 (10-2)) 6 (9-3)

ameters and lengths froman minus standardused mean plus standard

ivalent to the regionsodiametric growthildtype this regionts cob-1, lit-1, and

were marked with carbon grains (Okada and Shimura, 1990).Plants were then transferred to nutrient medium supplementedwith either 4.5% sucrose or with 0.1% sucrose as a control.Images of roots taken at intervals were then analyzed for therelative movement of the carbon grains (Fig. 5). Estimates ofelongation rates were determined from distance increases

Table 4. Hypocotyl cell dimensionsDiameter Epidermis length Cortex length Endodermis length

(µm) (µm) (µm) (µm)

252±32 142±60 106±30 97.0±31297±39 116±38 75.9±20 62.1±17292±27 152±47 84.4±28 63.6±19263±33 109±65 82.8±36 70.4±24319±57 93.4±52 71.5±47 61.7±30311±20 96.5±27 76.5±17 65.5±16363±49 113.6±38 80.2±20 68.0±23

.

1244 M.-T. Hauser, A. Morikami and P. N. Benfey

Condit

between adjacent carbon grains during a particular timeinterval. Increases in root diameter were used as estimates ofradial expansion for the same time interval.

Abnormal radial expansion was already observable 4 hoursafter the shift to restrictive conditions (Fig. 5B). Abnormalradial expansion appeared to be restricted to a subset ofgrowing cells (Fig. 5B,E). Above this region there were cellsin which elongation occurred without any apparent abnormalradial expansion. In particular, the relative movement ofcarbon grains located approximately 350 µm from the root tipindicated continued elongation of these cells while there waslittle increase in the root diameter at this point (Fig. 5E). Themeristematic zone also exhibited little or no change in shapeduring the first 8-12 hours after the shift to restrictive con-ditions (Fig. 5B-D). Estimates of relative expansion ratessuggested that in the region of abnormal radial expansion, therate of radial expansion exceeded that of elongation. Thisprovides additional evidence that the cobra mutation results ina shift in the orientation of polar expansion.

Orientation of cortical microtubules All the CORE mutants have abnormal polarity as evidenced bythe relative dimensions of the wall lengths of their epidermal

cells. Particularly striepidermal cells in cob. ulation predicts that thof the microtubules whexpansion (Giddings anexamine whether suchmutants, we analyzed mization with an anti-tub

In wild-type roots, zone were aligned trpredicted by the modeldifferentiation zone thrandom (Fig. 6D). Indramatic shift in the From the staining patte

Table 6. Analysis of CORE double mutantsSeedling pheno

Parental genotype Wild-type Mut1 Mumut1 × mut2 Total Obs./Exp. Obs./Exp. Obs./E

cob-1 × lit-1 (Benfey et al., 1993)qui × lit n.d.pom1-1 × lit-1 228 133/128 44/42.8 38/4cud-1 × lit-1 170 59/63.8 64/63.8 21/2pom2-1 × lit-1 218 121/122.6 46/40.9 39/4qui-2 × cob-1 347 64/65.1 283/281.9pom1-10 × cob-1 750 430/421.9 256/281.3

430/421.9 320/328.2cud-1 × cob-1 213 77/79.9 106/106.5pom2-1 × cob-1 148 82/83.3 18/27.8 32/2

82/83.3 66/64.9pom1-1 × qui-1 564 334/328 212/219cud-1 × qui-2 129 61/48.4 38/48.4 15/1pom2-1 × qui-1 155 76/87.2 69/58.1cud-1 × pom1-1 249 105/93.4 116/124.5pom2-1 × pom1-1 292 149/164.3 122/109.5cud × pom2 n.d.

*cud is dominant and male gametophytic lethal and therefore the dihybrid cross segregates 3 : 3 : 1 : 1 = wt : c†cob is semidominant and therefore we observed a gradient of slightly different phenotypes which we scored

+/+, cob/cob mut2/+, cob/+ mut2/mut2, +/+ mut2/mut2 and in case of the cross of cob with qui probably also coheterozygote we observed enhancement of the cob semidominant phenotype.

‡Due to very similar phenotype it was not possible to distinguish pom1, pom2 and qui as single mutants. The §Some of the plants scored as double mutants survived to give F3 seeds. The analysis of these F3 as well as tes

double mutants had the genotype cob/cob.qui/+. Therefore we placed this number in parentheses.**The χ2 value for these gave a P value smaller than 0.05. This is probably due to the semidominant cob phen

double mutant homozygotes from the double mutant combination with a heterozygous cobra allele. Abbreviationn.d., not determined.

Fig. 4. Optical sections of wildtype and CORE mutant roots of plantsgrown on medium supplemented with 4.5% sucrose. Medianlongitudinal sections through the differentiation zone of (A)wild-type Col, (B) lion’s tail-1, (C) cobra-1, (D) quill-1, (E) pom-pom1-1, (F) cudgel-1, (G) pom-pom2-1. Median longitudinalsections through the root tips of (H) wild-type Col, (I) lion’s tail-1,(J) cobra-1, (K) pom-pom1-1 and (L) pom-pom2-1. Bar, 50 µm.

Table 5. Comparisodivision rate unde

Root growth (mm/day)

WT 6.0lit-1 0.15cob-1 0.3qui-1 1.0pom1-1 0.65cud-1 0.1pom2-1 1.3

Root growth was determinbetween day 4 and 6 after gecalculated by dividing the rostandard deviation; see Table

1245ional root expansion mutants

king is the shift in polarity of theThe prevailing model of expansion reg-

n of root growth and apparent cellr maximal growth rate conditions

Division rate (cell number increase/day)

epidermis cortex endodermis

28-44 32-49 43-725-14 5-13 6-14

10-23 11-23 15-3022-46 27-55 34-669-21 12-30 16-432-5 3-6 4-8

16-35 21-42 30-57

ed by measuring the root length differencesrmination. The apparent cell division rate wasot growth per day by the cell length (± the 2).

ere should be a shift in the orientationen there is a change in the polarity ofd Staehelin, 1991) (see Discussion). To a shift had occurred in the COREicrotubule orientation by in situ local-

ulin antibody. cortical microtubules in the elongationansversely to the axis of growth as (Fig. 6D). In cells that had entered thee microtubule orientation was more

cob-1 there did not appear to be aorientation of microtubules (Fig. 6B).rn it appeared that the orientation of the

types

t2 Double xp. Obs./Exp. χ2

2.8 13/14.3 0.901.3 26/21.3* 1.40.9 12/13.6 0.93

,

(39/21.7)† § 0.0264/46.9† 8.7**

0.3630/26.6*,† 0.8

7.8 16/9.3† 8.9**0.04

38/36.5‡ 0.46.2 15/16.1* 5.67

10/9.69‡ 3.4828/31.1 2.3321/18.3‡ 3.25

ud : mut2 : dbl together. They include the genotypes cob/cobb/+ qui/+ and cob/+ +/+. In a cob/+, qui/+ trans-

double mutant phenotype was easily scored. tcross results revealed that the surviving putative

otype which made it difficult to distinguishs: mut., mutant; obs, observed; exp, expected,

1246 M.-T. Hauser, A. Morikami and P. N. Benfey

Fig. 5. The effect on cell expansion of a shift from permissive to restrictive conditions. cobra was grown for 8days on medium supplemented with 0.1% sucrose and root epidermal cells were marked with carbon grains.Seedlings were then transferred to media supplemented with 4.5% sucrose or 0.1% sucrose. (A) cob-1 root tip at 0time after transfer to 4.5% sucrose. (B) cob-1 at 4 hours after transfer to 4.5% sucrose. (C) cob-1 at 8 hours aftertransfer to 4.5% sucrose. (D) cob-1 at 12 hours after transfer to 4.5% sucrose. (E) Tracings of root tips shown inA-D with location of carbon grains indicated. Lines connect equivalent regions of the root as evidenced by thepresence of distinct carbon grains. Dotted line indicates point above which there is very little abnormal radialexpansion. Bar, 100 µm. (F) Tracings of root tips of cob-1 seedlings transferred from 0.1% sucrose to 0.1%sucrose. From left to right, times are 0, 4 and 8 hours after transfer. Bar, 300 µm. Note the carbon grains at thevery tip of both sets of roots are on root cap cells and therefore show little movement relative to the tip.

1247Conditional root expansion mutants

microtutubule outsideappearefor qushiftedpom2-1entatio

AnalyTo exa

Fig. 6. Omicrotu

b

i- (

n

sm

b

bules may be somewhat more random and the micro-undles appeared somewhat thicker. As with wildtype,the elongation zone the cob-1 microtubule orientationd much more random (Fig. 6B). The same was found1 and cud-1 in which epidermal polarity appearedFig. 6C,G). Cortical microtubules of lit-1, pom1-1 andwere also found to be predominantly in the same ori- as wildtype (Fig. 6A,E,F).

is of CORE mutant interactionsine the genetic interactions of the CORE mutants we

generated double mutant combinations (Table 6). Combina-tions of lit-1 and the other CORE mutants tested had pheno-types similar to lit-1 for the root expansion character. Thedouble mutants had reduced cell volume of the epidermis,cortex and endodermis (Fig. 7A-D). However, the root lengthof these double mutants was dramatically shorter than eitherparent. The double mutants also had an aerial phenotype thatwas severely stunted (Fig. 8A-D). This was particularly evidentin the combinations of lit-1 with cud-1 and with pom1-1 (Fig.8A,D).

All cud-1 double mutant combinations had root expansion

ptical sections of the root epidermis of seedlings grown on medium supplemented with 4.5% sucrose after immunolocalization ofules (A) lion’s tail-1, (B) cobra-1, (C) quill-1, (D) wild-type Col, (E) pom-pom1-1, (F) pom-pom2-1, (G) cudgel-1. Bar, 10 µm.

1248 M.-T. Hauser, A. Morikami and P. N. Benfey

Fig. 7. Optical sections of CORE double mutants grown on medium supplemented with 4.5% sucrose. Median longitudinal sections through thedifferentiation zone of double mutants of (A) cudgel-1 and lion’s tail-1, (B) cobra-1 and lion’s tail-1, (C) pom-pom1-1 and lion’s tail-1, (D)pom-pom2-1 and lion’s tail-1, (E) cudgel-1 and quill-2, (F) cudgel-1 and pom-pom1-1. Median longitudinal sections through root tips of doublemutants of (G) cudgel-1 and cobra-1, (H) quill-1 and pom-pom1-1, (I) cobra-1 and quill-2, (J) cobra-1 and pom-pom1-10, (K) pom-pom1-1 andpom-pom2-1, (L) cobra-1 and pom-pom2-1. Bar, 50 µm.

1249Conditional root expansion mutants

Fig. 8. Seedlings of CORE double mutants grown on medium supplemented with 4.5% sucrose approximately 17 DAG. Double mutant of (A)cudgel-1 and lion’s tail-1, (B) quill-3 and lion’s tail-1, (C) pom-pom1-1 and lion’s tail-1, (D) pom-pom2-1 and lion’s tail-1, (E) cudgel-1 andquill-2, (F) cudgel-1 and pom-pom1-1, (G) cudgel-1 and cobra-1, (H) quill-1 and pom-pom1-1, (I) cobra-1 and quill-1, (J) cobra-1 and pom-pom1-10, (K) pom-pom1-1 and pom-pom2-1, (L) cobra-1 and pom-pom2-1. Bar, 1 mm.

H

1250

characteristics reminiscent of cud-1 single mutants. Thephenotype of the double mutants included dramaticallyexpanded stele tissue as well as radially expanded epidermaland cortex tissue (Fig. 7E-G). In addition, double mutants ofboth lit-1 and cud-1 had extremely reduced elongation zonesthat resulted in roots with cells that were expanded in the zonewhere cell division normally takes place (Fig. 7C,D,F,G).Severely stunted aerial phenotypes similar to those of the lit-1double mutants were observed in the double mutants of cud-1with the other CORE mutants (Fig. 8E-G).

Double mutant combinations of cob-1 with pom1-10 orpom2-1 appeared to have an additive phenotype for expansionin the epidermis and cortex (Fig. 7J,L). The dramatic radialexpansion of the epidermis which is characteristic of cob-1 wasapparent in these double mutant combinations. However, incombination with cud-1 and qui-2, cob-1 did not appear toprovide any additional radial expansion of the epidermisbeyond that which is characteristic of these two mutants (Fig.7G,I). In a similar fashion, double mutants with qui-2 did nothave increased epidermal cell expansion (Fig. 7E,I). We notethough, that double mutants with any of the alleles of qui weremore affected in the aerial part of the plant (compare Fig. 8Eand 8G, or 8H and 8J). The pom2-1 mutation seemed toenhance slightly all other mutant phenotypes when in a doublemutant combination (Fig. 7K,L). Double mutant combinationswith pom2-1 exhibited a nearly wild-type aerial part, but dueto the extremely reduced fertility it was not possible to get F3seeds. Double mutant combinations with either pom1-1 orpom1-10 were characterized by shorter roots that had differ-entiated and expanded cells in what would normally be themeristematic zone (Fig 7C,F,H,J,K).

To test if the conditional phenotype was maintained in thedouble mutants we analyzed root growth under normally per-missive growth conditions. All of the double mutant combina-

tions tested had in which the sing(Figs 2E-H, 3G-Jleast partially red(see Discussion)low fertility prec

DISCUSSION

The regulation oto the determinaprincipal paramextent. We have result in deregulanewly isolated aloci have been chmutations includroot, the variabilof the root and From calculationthe mutants intoequal to, and gmutations result entation of expa

The volume clit may be an inacompensating rathis gene regulatFor cobra we hathe mutant is apsuggests that theorientation of cein these mutants

M.-T. Hauser, A. Morikami and P. N. Benfey

Fig. 9. Schematic representation of the three classes of CORE mutants. (A) In lion’s tail cell compared to wildtype. (B) In cobra there is approximate conservation of cell volume. Therefreduction in elongation. This leads to a shift in the polarity of expansion. (C) In the other foudefect in volume regulation so that the cells attain a greater volume than wildtype. This resultcircumferential expansion that overcompensates for the reduction in elongation.

expanded roots when grown under conditionsle mutants do not have detectable expansion). This suggests that these genes may have atundant functions or act in a similar pathway

. Most of these double mutants had extremelyluding further genetic testing.

f cell expansion is of paramount importancetion of organ shape in higher plants. The twoeters of cell expansion are orientation andidentified six genetic loci that when mutated,tion of cell expansion. Three of the loci werend additional alleles of previously identifiedaracterized. Common features of these CORE

e the dramatically reduced elongation of the ity in the degree of expansion along the lengththe root-specificity of the expansion defects.s of cell volumes we have been able to place three classes: those with volumes less than,reater than wildtype. This indicates that thein abnormal regulation of the extent or the ori-nsion or of both. alculations indicate that the primary defect inbility to achieve correct elongation with littledial expansion (Fig. 9A). This suggests thates the cell’s ability to attain its proper volume.ve determined that epidermal cell volume of

proximately equal to wildtype (Fig. 9B). This primary defect may be in the control of thell expansion. The principal axis of expansion may be in the radial rather than the longitu-

volume of the outer three layers is reduced asore, radial expansion compensates for ther mutants, represented here by pom-pom1 there is as from a major increase in radial and

dinal dpom-pcantly play a

The roorientTo anaexpansconditmutatitransitments anisotrgrowthof exptheoriethe conmicrotdramaIt is poin micobservarrestethe mexpansAnothchangetrue thnot besuggesanisotranisotrcan onexpansmicrotbundleorganigationsame shifted

Themay nexpanscomestreated(Baski

Sevdivisiobased rates. Bstate scautionticularamounthat a lfunctiodo noteffect severaceases

1251Conditional root expansion mutants

irection. The third class, which includes quill, cudgel,om1 and pom-pom2, has cell volumes that are signifi-greater than wildtype. This indicates that these genesrole in controlling the extent of expansion (Fig. 9C).

le of microtubules in determining theation of expansionlyze further the effect of the cobra mutation on cellion we shifted plants from permissive to restrictive

ions. These experiments suggested that the cobraon affected cells only for a limited period during their through the elongation zone. Preliminary measure-also suggested that the mutation caused a change fromopic growth in the longitudinal direction to anisotropic in the radial orientation. If a true shift in the polarityansion occurs in this mutant it could be used to tests concerning the role of different macromolecules introl of cell expansion. It was expected that the cortical

these mutants there is no abnormal cell expansion equivalentto that seen in the CORE mutants.

Because plant growth regulators can have a dramatic effecton plant cell expansion we grew the CORE mutants ondifferent concentrations of auxin, cytokinin, gibberellic acid,abscissic acid, the ethylene precursor 1-aminocyclopropane-1-carboxylic acid (ACC), silver ions (which inhibit ethyleneaction), the auxin transport inhibitor 2,3,5-triiodobenzoic acid(TIBA), and the gibberellic acid biosynthesis inhibitor,ancymidol. The root expansion phenotype was not noticeablyaffected by any of these treatments (data not shown).

Other mutations have been identified that affect root cellexpansion. In the sabre mutant there is a shift in orientation ofexpansion primarily in root cortex cells (Benfey et al., 1993;Aeschbacher et al., 1995). It was found that reduction ineffective ethylene levels resulted in partial rescue of thephenotype. This suggested that the SABRE gene and ethyleneregulated two counteracting processes that determined the

t

r

u

e

so

so

e

l

l

ubules in the epidermal cells of cobra would exhibit aic shift in polarity. This was found not to be the case.ssible that this is a situation in which a dramatic shiftotubule orientation would not be required to obtain theed cell shapes. One possibility is that elongation isd while “normal” radial expansion is unimpaired bytation. In this case, the mutation must allow radialion to continue beyond the limit found in wildtype.r possibility is that the CORE mutations result in a from anisotropic to isotropic expansion. If this wereen a dramatic shift in microtubule orientation wouldobserved. Rough estimates of relative expansion ratest that at least in cobra there has been a shift fromopic expansion in the longitudinal direction toopic expansion in the radial direction. However, thisly be resolved by accurate measurements of cellularion rates. It should also be noted that although theubules do not shift their orientation, the microtubules appear to be more aggregated and slightly more dis-zed than wildtype. A key question for future investi- is whether the cellulose microfibrils are still in therientation as the microtubules or whether they have their orientation. e observations raise the possibility that microtubulest play a critical role in regulating the orientation of cell

extent of radial expansion of certain cells (Aeschbacher et al.,1995). A temperature sensitive screen identified three mutantswith abnormal root expansion (Baskin et al., 1992). At therestrictive temperature these mutants have abnormal aerialgrowth as well. Complementation analysis revealed that thesemutants are not allelic to the CORE mutants (R.Williamson,M.-T.Hauser and P.Benfey unpublished results).

Double mutant analysis suggests CORE genes mayhave similar functionsThe mutant phenotype of all of the CORE mutants is condi-tional on the root growing at a maximum rate. Unlike a simpletemperature-sensitive mutation in which a protein foldingdefect would be implicated, the CORE mutant conditionalphenotype suggests that there may be conditional redundancyof function. In this hypothesis, under submaximal growth con-ditions, if any one of the CORE gene activities is removedanother gene activity can substitute for it. However, undermaximum growth conditions all of the CORE genes would benecessary for regulated root cell expansion. A prediction of thishypothesis is that combinations of the mutations at redundantloci should result in a loss of the conditional phenotype. Thisis what we found for all tested double mutant combinations ofthe CORE mutants. This is evidence for the CORE mutantsproviding partially redundant functions which might be accom-

ion in Arabidopsis root cells. Further evidence for thisfrom studies of ethylene treated roots and of roots

with microtubule stabilizing and destabilizing agentsn and Williamson, 1992; Baskin et al., 1994). ral of the CORE mutants appeared to have reduced celln rates. The calculations of cell division rates wereon the assumption that cells were dividing at constantecause we do not have direct evidence for this steady

ituation these observations should be interpreted with. We note, however, apparent division rates were par-y low for the mutants lit, cud, and cob that have a larget of expansion in the stele (see Table 2). It is possibleower cell division rate is caused by defects in the normalning of nutrient transport of the root vascular tissue. We believe that abnormal cell expansion is a secondaryof reduced cell division because we have analyzed mutants in which cell division slows down and thenentirely (Benfey et al., 1993; Scheres et al., 1995). In

plished if they acted in a multimeric complex or in a similardevelopmental process.

In addition to loss of the conditional phenotype, novel phe-notypes appeared in some double mutant combinations. Theseincluded complete cessation of root growth and severelyaffected aerial organs. Cessation of root growth was particu-larly evident in double mutant combinations in which steletissue was affected in both single mutants. Thus, this may bea secondary effect of impairment of vascular tissue function.The aerial phenotypes may also be attributable to impaired rootfunction. Alternatively, they may indicate that the CORE genesplay important functions in shoot cell expansion but that underour growth conditions a single gene mutation is always maskedby functional redundancy.

The double mutant combinations of the CORE mutants alsorevealed possible epistatic relationships. For the character ofroot expansion, lit was epistatic to all the other mutants tested,cud was epistatic to all mutants except lit while the double

1252

mutant phenotypes of the other four muand pom2, were essentially additive. Inresults should take into account whetherdouble mutant analysis represented the nabsence of molecular data or deletion tesible to make such designations. Thus, exexercised in interpreting these results. Wthat the apparent epistasis of lit to the othbe plausible, in that lit is impaired in its avolume and therefore to elongate propseems reasonable that this phenotype weither that of cob, in which the orientaprimarily affected, or the other four mutatoo much expansion. It is important to nepistasis is only for the root expansiondouble mutants also lose their conditionaabnormal aerial organs. An alternative double mutant phenotypes are the resulhypomorphic mutations in the same pat

In summary we have identified six lothe orientation, the extent or both aspeccells. The loss of the conditional mutations in these loci are combined inact in a similar pathway. Molecular analproducts should shed additional light upof cell expansion.

We thank K. Roberts, P. Linstead and C. Dand support. We received expert technical asV. Blanc, D. Tang, M. Yung, K. Schultheisand M. S. Hsieh. We thank M. Huelskamp alating discussions. This publication benefittcomments and criticisms of L. Di LaurenzMalamy, R. Aeschbacher, J. Schiefelbein andJ. Schiefelbein for providing the qui-3, pomK. Feldmann for providing the pom1-1 alleBlanc for the pom1-10 allele. We thank J. Ecdiated seed, C. Leonard for help with confocfor confocal staining protocols, R. White forM-T. H. was supported by a Schroedinger zur Foerderung der wissenschaftlichen FoThe work in P. N. B.’s laboratory was supGM43788) from the NIH.

REFERENCES

Aeschbacher, R.A., Hauser, M.-T., Feldmann(1995). The SABRE gene is required for nArabidopsis. Genes Dev. in press.

Baluska, F., Brailsford, R.W., Hauskrecht, M., JP.W. (1993). Cellular dimorphism in the maize

M.-T. Hauser, A. Morikami and

tants cob, qui, pom1terpretation of these

the allele used in theull phenotype. In theter lines it is imposs-

treme caution must bee note, nonetheless,

er six mutants wouldbility to increase cellerly and expand. Itould be epistatic totion of expansion isnts, in which there isote that this apparent character. These litl phenotype and havepossibility is that the

microtubules, ethylene and gibberellin in the differentiation of cellularbehaviour in postmitotic growth zones. Acta Bot. 106, 394-403.

Baskin, T.I., Betzner, A.S., Hoggart, R., Cork, A. and Williamson, R.E.(1992). Root morphology mutants in Arabidopsis thaliana. Aust. J. PlantPhysiol. 19, 427-438.

Baskin, T.I., Wilson, J.E., Cork, A. and Williamson, R.E. (1994).Morphology and microtubule organization in Arabidopsis roots exposed tooryzalin or taxol. Plant Cell Physiol. 35, 935-942.

Baskin, T.I. and Williamson, R.E. (1992). Ethylene, microtubules and rootmorphology in wild-type and mutant Arabidopsis seedlings. Curr. Top. Pl.Biochem. Physiol. 11, 118-130.

Bell, C.J. and Ecker, J.R. (1994). Assignment of 30 microsatellite loci to thelinkage map of Arabidopsis. Genomics 18, 137-144.

Benfey, P.N., Linstead, P.J., Roberts, K., Schiefelbein, J.W., Hauser, M.-T.and Aeschbacher, R.A. (1993). Root development in Arabidopsis: fourmutants with dramatically altered root morphogenesis. Development 119,57-70.

Colasanti, J., Cho, S.O., Wick, S. and Sundaresan, V. (1993). Localizationof the functional p34-cdc2 homolog of maize in root tip and stomatalcomplex cells: Association with predicted division sites. Pl. Cell 5,1011-1111.

P. N. Benfey

t of combinations of

hway. ci that regulate either

ts of expansion in rootphenotype when thedicates that they may

ysis of the CORE geneon the genetic control

uckett for their kind helpsistance from H. Bondar,s, K. Gallagher, J. Lim,nd C. Schultz for stimu-ed enormously from theio, J. Wysocka-Diller, J. the reviewers. We thank1-9 and pom1-11 alleles,le and C. Schultz and V.ker for providing X-irra-al microscopy, S. Ruzin use of his slide scanner.fellowship from “Fondsrschung” (J0676-MOB).ported by a grant (RO1-

Cosgrove, D.J. (1993). How do plant cell walls extend? Plant Physiol. 102,1-6.

Doonan, J., Zhang, H., and Traas, J. (1993). Indirect Immunofluorescence onArabidopsis seedlings. In Arabidopsis, The Complete Guide (ed. D. Flandersand C. Dean). Norwich, U.K.: AFRC Institute of Plant Science Research.

Feldmann, K.A. (1991). T-DNA insertion mutagenesis in Arabidopsis:Mutational spectrum. Plant J. 1, 71-82.

Giddings, T.H.Jr. and Staehelin, L.A. (1991). Microtubule-mediated controlof microfibril deposition: A re-examination of the hypothesis. In TheCytoskeletal Basis of Plant Growth and Form (ed. C.W. Lloyd), pp. 85-100.London: Academic Press.

Hauser, M.-T. and Benfey, P.N. (1993). Genetic Regulation of RootExpansion in Arabidopsis thaliana. In NATO-ASI Plant Molecular BiologySeries (ed. P. Pugdomenech and G. Coruzzi), pp. 31-40. New York:Springer-Verlag.

Kaplan, D.R. and Hagemann, W. (1991). The relationship of cell andorganism in vascular plants. BioScience 41, 693-703.

Konieczny, A. and Ausubel, F.M. (1993). A procedure for mappingArabidopsis mutations using co-dominant ecotype-specific PCR-basedmarkers. Plant J. 4, 403-410.

Okada, K. and Shimura, Y. (1990). Reversible root tip rotation in Arabidopsisseedlings induced by obstacle-touching stimulus. Science 250, 274-276.

Roberts, K. (1989). The plant extracellular matrix. Current Opinion in CellBiology 1, 1020-1027.

Scheres, B., Di Laurenzio, L., Willemsen, V., Hauser, M.-T., Janmaat, K.,Weisbeek, P. and Benfey, P.N. (1995). Mutations affecting the radialorganisation of the Arabidopsis root display specific defects throughout theradial axis. Development 121, 53-62.

Schiefelbein, J.W. and Somerville, C. (1990). Genetic control of root hairdevelopment in Arabidopsis thaliana. Pl. Cell 2, 235-243.

, K.A. and Benfey, P.N.ormal cell expansion in

ackson, M.B. and Barlow,root cortex: involvement of

Steeves, T.A. and Sussex, I.M. (1989). Patterns in Plant Development.Cambridge: Cambridge University Press.

Valvekens, D., Montagu, M.V. and Lusebettens, M.V. (1988).Agrobacterium tumefaciens-mediated transformation of Arabidopsisthaliana root explants by using kanamycin selection. Proc. Natl. Acad. Sci.USA 85, 5536-5540.

(Accepted 5 January 1995)