mutations in the reticulata gene dramatically alter internal architecture but have little effect on...

RESEARCH PAPER

Mutations in the RETICULATA gene dramatically alterinternal architecture but have little effect on overall organshape in Arabidopsis leaves

Rebeca Gonzalez-Bayon1,*, Elizabeth A. Kinsman3,*, Vıctor Quesada1, Antonio Vera4, Pedro Robles1,

Marıa Rosa Ponce1, Kevin A. Pyke2 and Jose Luis Micol1,†

1 Division de Genetica and Instituto de Bioingenierıa, Universidad Miguel Hernandez, Campus de Elche,E-03202 Elche, Spain2 Plant Science Division, School of Biosciences, University of Nottingham, Sutton Bonington Campus,Loughborough, Leicestershire LE12 5RD, UK3 School of Human and Life Sciences, Roehampton University, Holybourne Avenue, London SW15 4J, UK4 Division de Genetica, Universidad Miguel Hernandez, Campus de San Juan, E-03550 Alicante, Spain

Received 29 March 2006; Accepted 24 May 2006

Abstract

A number of mutants have been described in Arabi-

dopsis, whose leaf vascular network can be clearly

distinguished as a green reticulation on a paler lamina.

One of these reticulate mutants was named reticulata

(re) by Redei in 1964 and has been used for years as

a classical genetic marker for linkage analysis. Seven

recessive alleles of the RE gene were studied, at least

four of which seem to be null. Contrary to many other

leaf mutants studied in Arabidopsis, very little pleio-

tropy was observed in the external morphology of the

re mutants, whose only aberration obvious at first sight

is the reticulation exhibited by cotyledons and leaves.

The re alleles caused a marked reduction in the density

of mesophyll cells in interveinal regions of the leaf,

which does not result from perturbed plastid develop-

ment in specific cells, but rather from a dramatic

change in internal leaf architecture. Loss of function

of the RE gene seems to specifically perturb mesophyll

cell division in the early stages of leaf organogenesis.

The leaves of re mutants were nearly normal in shape

in spite of their extremely reduced mesophyll cell

density, suggesting that the epidermis plays a major

role in regulating leaf shape in Arabidopsis. The RE

gene was positionally cloned and found to be ex-

pressed in all the major organs studied. RE encodes

a protein of unknown function and is identical to the

LCD1 gene, which was identified based on the in-

creased sensitivity to ozone caused by its mutant allele

lcd1-1. Double mutant analyses suggest that RE acts

in a developmental pathway that involves CUE1 but

does not include DOV1.

Key words: Arabidopsis, leaf organogenesis, mesophyll

development, reticulata.

Introduction

The correct development of plant leaf internal tissues iscrucial to leaf function in order to optimize gaseous exchangeand light capture. However, the nature of the developmentalmechanisms that underlie the construction of the internalleaf architecture from different leaf cell types is unclear(reviewed in Fleming, 2002; Tsukaya, 2002, 2003, 2005;Tsiantis and Hay, 2003; Byrne, 2005; Canales et al., 2005).A detailed genetic dissection of the development of the inter-nal anatomy of leaves is required, and although specificepidermal cell types (Larkin et al., 1997) and vascular celldifferentiation (Carland and McHale, 1996; Nelson andDengler, 1997; Candela et al., 1999; Ye et al., 2002) havebeen studied in detail, an understanding of the interplay ofepidermal, palisade mesophyll, spongy mesophyll, andvascular development in constructing internal leaf structureandwhole leaf shape is lacking (Pyke and Lopez-Juez, 1999).

* These authors contributed equally to this work.y To whom correspondence should be addressed. E-mail: [email protected]

Journal of Experimental Botany, of 13

doi:10.1093/jxb/erl063This paper is available online free of all access charges (see http://jxb.oxfordjournals.org/open_access.html for further details)

ª 2006 The Author(s).This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) whichpermits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Journal of Experimental Botany Advance Access published July 26, 2006 by guest on January 27, 2016

http://jxb.oxfordjournals.org/D

ownloaded from

http://jxb.oxfordjournals.org/

A problem with screening approaches that require theisolation of mutants specifically perturbed in leaf internalarchitecture is the prediction of a clear morphologicalleaf phenotype which will reveal an underlying change inthe anatomical structure of the organ. Multiple sectioningas a screening method is likely to be both arduous andproblematic, especially in comparing tissues of equivalentdevelopmental status. In addition, pleiotropic effects onleaf morphogenesis may result from altered cell physiologyand consequently blur the distinction between leaf-specificmutants and general developmental mutants (Poethig, 1997;Bohmert et al., 1998; Berna et al., 1999).

Morphological mutants of Arabidopsis thaliana havebeen examined, in which the vascular network stands outas a colour difference on the lamina. Three broad phenotypicclasses exist: a dark green vascular pattern on a green lamina,a green vascular pattern on a pale lamina, and a pale vascularpattern on a green lamina (EA Kinsman, KA Pyke, unpub-lished data). Although many reticulate mutants are presentin stock centres and have been reported in different mutantcollections (Li et al., 1995; Berna et al., 1999; Serrano-Cartagena et al., 1999; Streatfield et al., 1999), the relation-ship of this phenotype to the underlying internal leaf structurehas been characterized in only a very few cases.

reticulata (re) was the first mutant with reticulate leavesisolated in Arabidopsis (Redei and Hirono, 1964). It hasbeen used for decades as a classical morphological map-ping line because of the ease of recognition and low degreeof pleiotropy of its phenotype. The characterization of reand another six mutant alleles of the RE gene is reportedhere. The results indicate that re mutations dramaticallyaffect mesophyll cell proliferation, but only slightly changeoverall leaf shape, which suggests that the correct de-velopment of the mesophyll tissues contributes little toleaf shape.

Materials and methods

Plant material and growth conditions

Seeds of the Arabidopsis thaliana (L.) Heynh. wild-type accessionsLandsberg erecta (Ler), Columbia-0 (Col-0), Wassilewskija-2 (Ws-2),and Enkheim-2 (En-2), as well as the N129 (which carries the re-1 andglabrous1mutations) andN734 (re-2) mutant lines,were obtained fromthe Nottingham Arabidopsis Stock Centre (NASC). They were grownfor the morphological and anatomical analyses illustrated in Figs 2, 5,6, and Table 2, as described in Kinsman and Pyke (1998). All theremaining cultures were performed as described in Ponce et al. (1998).The re-3 and re-4 mutants were isolated as described in Bernaet al. (1999), and re-5 was identified in a collection of alleles kindlyprovided by Javier Paz-Ares. Lines carrying the re-6 and re-7 alleleswere supplied by the NASC and are described at http://signal.salk.edu.Crosses were performed as described by Berna et al. (1999).The genetic backgrounds of the mutants are shown in Table 1.

Although the genetic background of re-1 was assumed to be Col,based on the information available at the NASC, it was found to bea hybrid, as determined by genotyping microsatellite markers, asdescribed in Ponce et al. (1999). Among the 20 microsatellitesgenotyped in re-1, 11 were homozygous for the Col-0 allele, 5 forthe Ler allele, and 4 for alleles unequivocally different from those ofLer and Col-0. These four alleles could not be assigned to any of the27 wild-type accessions that were genotyped in previous works(Quesada et al., 2002; Perez-Perez et al., 2002).

Morphological and ultrastructural analyses

Methods for leaf clearing and fixation, embedding, quantification ofleaf and cellular anatomical features, and observation of leaf tissue byconfocal microscopy are as described previously (Kinsman and Pyke,1998; Candela et al., 1999). Venation pattern diagrams were obtainedby drawing leaf veins on the screen of a Wacom Cintiq 18SXInteractive Pen Display and using the Adobe Photoshop 5.0 program.Leaf cell fixation and separation were performed as described pre-

viously (Pyke and Leech, 1992; Kinsman and Pyke, 1998). Suspen-sions of separated cells were tapped out from pieces of processed leaftissue mounted in 0.1 M Na2EDTA on microscope slides. Numbers ofchloroplasts per cell were counted by eye by viewing with Nomarskimicroscope optics, and areas of individual cells and plastids weremeasured using the Lucia image analysis system.

Table 1. The re mutants studied in this work

aa: amino acids. ,: T-DNA insertion. NC: not confirmed.

Allele Alternative name Geneticbackground

Mutagen DNA change Predictedprotein changed

Origine

Type of mutation Affected regionc

re-1 re, N129 Hybridb Unknown Deletion Whole RE gene No protein 1re-2 N734 En-2 EMS Transition G1225!A 189 aa missing 2re-3 ven2-1 Ler EMS Transition G1658!A Gly328!Arg 3re-4 ven2-2 Ler EMS Transition C1378!T Pro295!Leu 3re-5 ven2-3 Ws-2 T-DNA Rearrangement Nucleotide 6508 of At2g41790 is

joined to nucleotide –45 of RENo protein 4

re-6 N584529 Col-0 T-DNA Insertion ,660 395 aa missing 5re-7 N537745a Col-0 T-DNA Insertion NC; annotated as ,2045 – 5

a This line was obtained at the end of the research described here and has not yet been characterized in detail.b See Materials and methods.c Numbers refer to nucleotides in genomic DNA, starting from position +1 of the transcription unit.d The wild-type RE protein includes 432 amino acids.e 1: NASC (donated by M Koornneef); 2: NASC (donated by AR Kranz); 3: Berna et al. (1999); 4: This work; 5: SIGnAL collection (http://

signal.salk.edu).

2 of 13 Gonzalez-Bayon et al.

by guest on January 27, 2016http://jxb.oxfordjournals.org/

Dow

nloaded from


For scanning electron microscopy, plant material was preparedas described in Serrano-Cartagena et al. (2000). Micrographs weretaken in a JSM-840 JEOL scanning electron microscope.

Positional cloning and molecular characterization of

the re mutations

High-resolution mapping of the re-3 mutation was performed usingSSLP markers (Bell and Ecker, 1994) as described in Ponceet al. (1999). For allele sequencing, PCR products covering theAt2g37860 transcription unit were obtained using as a templateDNA extracted from Col-0, Ler, En-2, and re-2, re-3, re-4, and re-5homozygous plants. Sequencing reactions were carried out withABI PRISM BigDye Terminator Cycle Sequencing kits in 5 ll re-action volumes. Sequencing electrophoreses were performed on anABI PRISM 3100 Genetic Analyser.To confirm the presence and position of T-DNA inserts, DNA was

extracted from the N584529 and N537745 lines and PCR amplifiedusing the primers shown in supplementary Table 1 at JXB online, andboth strands of the amplification products obtained were sequenced.Southern blot analyses were performed as previously described(Perez-Perez et al., 2004), after ScaI or BamHI restriction of geno-mic DNA extracted from Col-0, Ler and re-1 and re-3 homo-zygous plants. Sequences of the primers used to PCR synthesizethe probe are shown in supplementary Table 1 at JXB online.

Gene expression analyses

For semi-quantitative reverse-transcriptase PCR (RT-PCR) analysis,RNA was extracted from plant material (50–100 mg), reverse trans-cribed using random hexanucleotide primers, and PCR amplified asdescribed in Quesada et al. (1999). Sequences of the gene-specificprimers used are detailed in supplementary Table 1 at JXB online.Control reactions using primers for the ORNITHINE TRANSCAR-BAMILASE (OTC) gene (Quesada et al., 1999) were performed onthe same cDNA samples.To produce the PRE:GUS construct, the complete genomic region

(0.5 kb) located upstream of the RE transcription unit and down-stream of the preceding gene (At2g37840) was PCR amplified usingCol-0 genomic DNA as a template and the pGusAt2g37860F andpGusAt2g37860R primers, cloned into the pGEM-3Zf(+) vector andtransferred to the pDW137 binary vector, which contains the

b-Glucuronidase (GUS) reporter gene (Blazquez et al., 1997). ThePRE:GUS construct was mobilized into Agrobacterium tumefaciensLBA4404 cells, and fully sequenced to confirm its structural integ-rity prior to being transferred into Col-0 plants by the floral dipmethod (Clough and Bent, 1998). Transgenic plants were selected on50 lg ml�1 kanamycin-supplemented medium. GUS staining wasperformed as described in Blazquez et al. (1997).

Results

Leaf morphology of the re mutants

The mutants studied here are listed in Table 1. Crosses ofthe N129 line (carrying the re recessive mutation, herenamed re-1) to several other reticulate mutants from stockcentres revealed a second recessive allele of RE, which wasnamed re-2 (N734). Another three reticulate allelic mutantswere obtained (Berna et al., 1999; this work) and initiallynamed ven2-1 (venosa2-1), ven2-2, and ven2-3. Comple-mentation tests demonstrated that these lines carried re-cessive alleles of the RE gene, for which reason they wererenamed re-3, re-4, and re-5. After the positional cloningof the RE gene, two additional recessive alleles, re-6 andre-7, were obtained from a collection of knockouts.

The only abnormality obvious in the re mutants at thelevel of the overall plant form is the fact that their coty-ledons, vegetative leaves, and cauline leaves display aconspicuous vein pattern, which is normally seen as a dark-green venation on a paler green background (Fig. 1). Noobvious mutant phenotypic traits were found in the roots,the inflorescence, the floral organs, or the fruits of the remutants (data not shown).

All the re mutants displayed very similar leaf pheno-types, irrespective of their genetic background. A fewextensions of the vasculature to the leaf margins in the

Fig. 1. Cotyledons (left) and third node vegetative leaves (right) of the re mutants and their wild types. All plants were homozygous for the mutationsindicated. The wild-type genetic background of each mutant is indicated in brackets. The Ws-2 wild type is not represented since it is similar to Col-0 andEn-2. Pictures were taken 20 d after sowing. Bar=1 mm.

The Arabidopsis RETICULATA gene 3 of 13


Dow

nloaded from


re/re leaves resulted in teeth that correspond to the presenceof the apical and lateral hydathodes. Lateral teeth are visibleonly during the early stages of leaf expansion in the wildtypes (Candela et al., 1999), but remain evident in re/refully expanded leaves.

Overall leaf shape is not very different between there mutants and their wild types, the only exceptions beingthe above-mentioned marginal irregularities and organ size,which is slightly, but clearly, reduced (Fig. 1). Measure-ments of the 1st to 9th node leaf laminae showed noconsistent differences between the wild type and the re-1mutant in the width/length and perimeter/area ratios duringmost of the stages of leaf development (Fig. 2A, B).Consistent with the small reduction observed for the thirdleaf (Fig. 1), the width, length, perimeter, and area wereslightly reduced in all the leaves of the re-1 mutantcompared with the wild type (Fig. 2A, B).

Measurement of the vascular length in relation to leafexpansion revealed that re-1/re-1 and wild-type first leavesare very similar in the early stages of leaf development(Fig. 2C). As the leaf reaches full expansion, however, vas-

cular development is decreased in the mutant, resulting in areduction in the vascular length per unit area in re-1/re-1ma-ture first leaves. The leaf venation pattern of the re mutantswas also analysed and it was found in all cases that it issimpler than that of their corresponding wild types (Fig. 3).

Internal architecture of re/re leaves

In the re mutants, the leaf lamina was ridged rather thansmooth, with the contours of the vascular pattern evident onthe leaf surface (Fig. 4A, B). This suggested that there maybe changes in the internal leaf anatomy, which correlatewith the venation pattern. Examination of intact first leavesby confocal microscopy showed that their interveinal re-gions were extremely reduced in cell density and appearedto have holes within the matrix of mesophyll cells (Fig.4C–K). This was confirmed by transverse sections of re/releaves, which revealed an internal leaf structure dramati-cally different from the wild type and which directly reflectsthe reticulation pattern observed at the whole leaf level(Fig. 5). Areas which lie between the vascular strands hada greatly reduced number of both palisade and spongy

A

Leaf length (mm)0

Leaf

wid

th (m

m)

0

5

10

15

20

25

10 20 30 40 50Leaf area (mm2)

B

0

Leaf

per

imet

er (m

m)

0

20

40

60

80

100

120

100 200 300 400 500 600Leaf area (mm2)

C

0

Vasc

ular

leng

th (m

m)

0

20

40

60

80

5 10 15 20 25 30 35

Col-0re-1

Fig. 2. Analysis of some leaf shape parameters for Col-0 and re-1/re-1 leaves. (A, B) Relationship between leaf length and width (A), and leaf area andperimeter (B). Data were obtained for 1st to 9th node leaves from four different Col-0 and re-1/re-1 plants collected 40 d after sowing. (C) Relationshipbetween leaf area and total vascular length during the development of first leaves in Col-0 and re-1/re-1 plants. Data were collected from the first leavesof 4–9 plants and daily between the 11th and 19th day after sowing.

Fig. 3. Vascular networks of re/re and wild-type (A–I) first and (J–R) third node leaves. All plants were homozygous for the mutations indicated. Thegenetic background of each re mutant is indicated in brackets. The leaves were collected 21 d after sowing and cleared as described in the Materials andmethods. Bar=1 mm.



Dow

nloaded from


mesophyll cells compared with the wild type (Fig. 5B, C).By contrast, the vascular bundles consisting of vascularcells and bundle sheath cells were normal and complete.Extensive sectioning indicated that a cellular bridgebetween the adaxial and abaxial epidermis often occurs inthe proximity of vascular strands while in the interveinalregions, mesophyll cell density is much lower and contig-uous cells linking the two epidermises in the vertical planeare much rarer (Fig. 5). Consequently, the density ofchloroplast-containing cells per unit leaf area is greatlyreduced in the interveinal regions, compared with thevascular regions of the lamina (Table 2).

Quantification of the internal leaf architecture of the re-1mutant (Table 2) showed that for transverse sections acrosshalf leaves, i.e. from the midvein to leaf margin, the pro-portion of transect occupied by mesophyll cells is reduced

and the fraction occupied by airspace is concomitantlygreater than in wild-type leaves. This difference is alsomanifested as a reduction in the number of palisade andspongy mesophyll cells as a proportion of the cell transects(data not shown). There appears to be a preferential lossof palisade mesophyll cells over spongy mesophyll cellssince the ratio of palisade:spongy cell transects is only 0.37in re-1/re-1 leaves compared with 0.48 in the wild type.It is clear from the transverse sections of these leaves (Fig.5) that, in the interveinal regions, the loss of mesophyllcells and the resultant increase in airspace is even greaterthan the average data reveal for entire half-leaf transects.The palisade and spongy mesophyll cells which are presentin re-1/re-1 leaves appeared normal (data not shown) andcontained chloroplasts similar in number and morphologyto those of the wild type (Table 2, and data not shown).

Fig. 4. (A, B) Scanning electron micrographs of the adaxial epidermis of third leaves from (A) Col-0 and (B) re-1/re-1 plants, the latter showing itscontoured vasculature on the surface of the lamina. (C–K) Reduced internal cell density in re/re leaves, as viewed by confocal microscopy by showingfluorescing native chlorophyll within mesophyll cells. All plants were homozygous for the mutations indicated. The pictures show details of themesophyll of first leaves collected 21 d after sowing and represent projections of eight optical sections. Values of thickness of the leaf sections rangedfrom 48.2 to 128 lm. Bar=1 mm (A, B) and 50 lm (C–K).

Fig. 5. Leaf sections of re/re and wild-type leaves. Transverse sections are shown from first leaves of (A) Col-0, (B) re-1/re-1, and (C) re-2/re-2 plants.Arrows indicate abaxial epidermal lobes. Similar results were obtained with the remaining re mutants and their corresponding wild types. Bar=50 lm.



Dow

nloaded from


The content of chloroplasts per bundle sheath cell, on theother hand, was significantly higher in the re-1 mutant.Thus, its reticulate phenotype does not result from per-turbed plastid development, but rather from a dramaticchange in internal leaf architecture. The paleness of theinterveinal regions seems to be mainly due to reducedmesophyll cell density and increased air space, which,furthermore, makes the vasculature more apparent. Never-theless, the increased number of chloroplasts shown in thebundle sheath cells would also contribute to the dark greencolour of the vasculature.

Generally, re-1/re-1 leaves were thinner than wild-typeleaves (Table 2), but increased leaf thickness was oftenobserved close to a vascular strand, which is consistent withthe observation that the vascular pattern has a raised profileon the adaxial surface of the mutant leaves (Fig. 4B).Transverse sections revealed lobes of epidermal tissue,particularly on the abaxial leaf surface, which are notobserved in wild-type leaves (Fig. 5). Scanning electronmicroscopy indicated that epidermal cell size and morphol-ogy are not modified in the leaves of re mutants (seesupplementary Fig. 1 at JXB online).

Changes in the development of the internal mesophylltissue in the re-1 mutant were evident at the early stagesof leaf primordial development (Fig. 6), although the

shoot apical meristem was unchanged (data not shown). Itappears that reduced cell division in the re-1/re-1 palisadecell layer results in looser cell packing even at this earlystage. The mutant palisade cells also appeared to be morevacuolated than the wild type in sections of equivalentdevelopmental stage (Fig. 6).

Positional cloning of the RE gene and characterizationof re alleles

The re-1 mutation has been known for a long time to mapat chromosome 2 (Redei and Hirono, 1964). An F2 map-ping population derived from a Col-03re-3/re-3 crosswas genotyped for molecular markers (see Materials andmethods), which allowed us to find linkage to the nga361and nga168 markers. For one of the genes within thecandidate region, At2g37860, recently named LCD1 (LOWCELL DENSITY1), a mutant allele had been described,lcd1-1, which causes a reticulate leaf phenotype (Barthand Conklin, 2003). The At2g37860 gene was sequencedin all the remutants (Fig. 7), finding the mutations shown in

Fig. 6. Transverse sections from midway along the length of the firstleaf primordium from (A) Col-0 and (B) re-1/re-1 plants. At the time offixation, the primordium of the fourth leaf was evident on the apicalmeristem. Both primordia were approximately 250 lm in length. ad:adaxial. ab: abaxial. Bar=20 lm.

Table 2. Leaf thickness, tissue proportions, and mean values forcell size, chloroplast number, and chloroplast size and cell indexfrom first leaves of the wild-type Col-0 and the re-1 mutant

Standard errors are given in parentheses.

Col-0 re-1/re-1

Leaf thickness (lm) 253 (24) 143 (6)Proportion of tissue type (%)a

Mesophyll 51.4 38.2Airspace 24.7 35.5Vascular 1.1 1.4Epidermis 22.7 24.8

Mesophyll cellsMesophyll cell plan area (lm2;n >100)

5088 (147) 3780 (157)

Chloroplasts per mesophyll cell(n >50)

99 (4) 100 (4)

Chloroplast plan area (lm2;n >300)

45.3 (0.5) 39.6 (0.5)

Cell indexb 0.88 1.04Bundle sheath cellsBundle sheath cell plan area(lm2; n >100)

1251 (46) 1381 (60)

Chloroplasts per bundle sheathcell (n >50)

26.5 (1.0) 39 (1.5)

Chloroplast plan area(lm2; n >300)

23.3 (0.3) 21.9 (0.3)

Cell indexb 0.49 0.62

a Mean proportion of tissue types and leaf thickness are derived fromhalf leaf transects from three positions along the leaf length from fivedifferent first leaves of mutant and wild-type plants. In each transect, leafthickness was measured at approximately 20 points across the leaf width.

b Cell index is calculated as (chloroplast number per cell3meanchloroplast plan area)/cell plan area.

F3

R3 R4 R1 R5 R2

F1 F4 F5 F2 F6

TGAre-6

re-4 re-3

(G A) re-7

1 2289

re-2100 bp

∆re-1(C T) (G A)

ATG

Fig. 7. Schematic representation of the structure of the RE gene, withindication of the re mutations. Exons and introns are indicated by boxesand lines, respectively. White boxes correspond to the 59 and 39;untranslated regions. The n symbol is used to indicate a deletion. Thepositions and directions of the oligonucleotide primers used for thestructural characterization of re alleles are represented by horizontalarrows below the RE gene map. Oligonucleotide sequences are detailed inthe supplementary Table 1 at JXB online.



Dow

nloaded from


Table 1. This demonstrated that lcd1-1 and the remutationsare alleles of the same gene, which is named RE in thiswork. The RE gene symbol has been used for four decadesin significant publications and is widely known to theArabidopsis community, criteria that have been in manycases followed to decide which gene symbol should begiven precedence, as indicated in the guidelines of Meinkeand Koornneef (1997).

A search of the Salk collection of T-DNA insertion linesrevealed two with the At2g37860 gene disrupted, whichwere confirmed by complementation tests to carry re allelesand named re-6 and re-7. The sequence of the RE gene inthe wild-type accessions studied was identical, the onlyexception being nucleotide position 2060 of its transcrip-tion unit, which was C in Col-0, but G in Ler and En-2. Thismeans that the RE protein has a phenylalanine residue inCol-0 and a leucine in Ler and En-2 at position 399.

As regards the mutants, the re-3 point mutation convertsa glycine into arginine at position 328, and re-4 causesthe substitution of proline by leucine at position 295. Thesetwo last changes affect a conserved region of the RE protein(Fig. 7; see supplementary Fig. 2 at JXB online). The re-2mutation removes 189 amino acids from the C-terminalregion of the RE protein, eliminating most of their con-served domains (see supplementary Fig. 2 at JXB online).The re-5 mutant carries a rearrangement joining nucleotide-45 of RE (At2g37860) to nucleotide 6508 of At2g41790.This mutant carries, in addition, a T-DNA insertion be-tween the At2g41770 and At2g41780 genes, which, to-gether with At2g41790, are located at a distance of 1.6 Mbfrom RE in the wild type. In the re-5 mutant, the promoterof At2g41790 drives the expression of an At2g41790-REchimeric transcript, which is predicted to be translatedonly into the protein product of At2g41790, which in turn ispartially aberrant because its C-terminus includes aminoacids translated from the promoter of RE. The distancebetween the 59 cap of the chimeric At2g41790-RE mRNAand the initiation codon of its RE portion is about 3 kb,which would not allow ribosome assembly at the RE tran-slation initiation signal (Kawaguchi and Bailey-Serres,2002). As regards re-1, it carries a deletion of 24 kb, en-compassing RE and another 10 genes (At2g37790–At2g37880). Consistent with this, no signal was detectedin Southern blots of restricted genomic DNA of the re-1mutant probed with a RE-specific probe (see Materials andmethods and supplementary Table 1 at JXB online).

The expression of the RE gene in the third leaves ofthe re mutants and their wild types was analysed usingseveral primer sets (Fig. 7; see supplementary Table 1 atJXB online). Transcripts of the expected size were detectedin the re-2, re-3, and re-4 mutants, which carry pointmutations, but not in re-1, as a consequence of its deletionaffecting the whole RE gene. Amplification and sequencingof the full-length re-5 cDNA confirmed its predictedchimeric nature. Two RE truncated mRNAs were identified

in the re-6 mutant, one of them transcribed upstreamand the other downstream of the T-DNA insert, as occursin other insertional mutants. The upstream mRNA is pre-dicted to be translated into only 37 amino acids of the REwild-type protein. The downstream mRNA lacks 217nucleotides in its 59 region, includes some T-DNA nucleo-tides, and cannot be translated because it does not includea translation initiation signal.

The RE gene encodes a predicted protein of 432 aminoacids with a molecular weight of 46.6 kDa (see supple-mentary Fig. 2A at JXB online). Searches in the databasesmade it possible to identify two full length cDNAs,At2g37860.1 and At2g37860.2, the first one 96 nucleotideslonger than the other. Translation of At2g37860.1 wouldproduce a protein of 348 amino acids, 84 residues shorterthan the predicted full length protein. The attempts toidentify the At2g37860.1 transcript were not successful,suggesting that it is either artefactual, unstable, producedin developmental stages other than those studied here, orstrictly restricted to a reduced number of cells duringdevelopment.

In silico analyses (http://crombec.botanik.uni-koeln.de/seq_view.ep) detected a putative transit peptide to chloro-plasts in the N-terminal part of the protein (amino acids1–47) and two probable transmembrane regions (aminoacids 247–269 and 321–343). RE is closely related to otherplant proteins of unknown function in Arabidopsis (theclosest homologue to RE in the genome of Arabidopsisbeing the At5g22790 gene, which maps at chromosome 5),Euphorbia esula, and Oryza sativa. The C-terminal regionsof all these proteins are highly homologous, whereas theN-terminal region of all of them, except CAE05717 of rice,is rich in glycine residues. No homology was found withproteins of biological systems other than plants, suggestingthat RE is a plant-specific protein.

RE expression analyses

To determine the spatial pattern of expression of RE inthe wild types, total RNA was isolated from roots, rosetteleaves, stems, flower buds, and open flowers of Col-0, En-2and Ler. RT-PCR amplifications indicated that expres-sion of RE is not restricted to leaves, as shown by thesingle band of 288 bp (the expected size for the amplifi-cation product corresponding to the last three exons ofAt2g37860.2) detected in all the organs studied (see sup-plementary Fig. 2B at JXB online). Similar amounts ofRE transcripts were obtained from the different organs stu-died, except in stems, where fewer transcripts were detected.The same conclusions can also be reached by examiningthe expression profiles of RE obtained from Genevestigator(http://www.genevestigator.ethz.ch; Zimmermann et al.,2004).

RE expression was further analysed by transformingCol-0 plants with a RE promoter: b-glucuronidase (GUS)



Dow

nloaded from


construct (see Materials and methods). GUS activity washigh in the vasculature of developing leaf primordia anddecreased as leaves aged, becoming restricted in fullyexpanded leaves to the margins, hydathodes, and to thebasal region of the lamina (Fig. 8). GUS activity was alsodetected in stipules, root tips, hydathodes of cauline leaves,stamens, and in the abscission zone of the funiculus. Thispattern of GUS activity is consistent with that of REexpression detected by RT-PCR.

Double mutant analysis

To gain insight into the genetic interactions between mu-tations causing reticulation, re-3/re-3 plants were crossedto homozygotes for the loss-of-function mutations cue1-5(chlorophyll a/b binding protein underexpressed; Liet al., 1995) and dov1 (differential development of vascularassociated cells; Kinsman and Pyke, 1998), which exhibita dark-green leaf vasculature on a pale-green lamina(Fig. 9). The cue1-5 mutant is smaller than the re mutantsand its reticulation does not become less apparent withageing. These two traits were shared by the RE/-;cue1-5/cue1-5 and re-3/re-3;cue1-5/cue1-5 siblings found in theF2 progeny of a re-3/re-33cue1-5/cue1-5 cross, whosegenotypes were confirmed by sequencing the CUE1 and

RE genes. This indicates that cue1-5 is epistatic to re-3 andsuggests that RE and CUE1 act in the same developmentalpathway affecting leaf development. The phenotype ofthe dov1/dov1;re-3/re-3 double mutants was merely addi-tive, as expected if the DOV1 and RE genes act in anindependent manner.

Discussion

The re mutants seem to be leaf-specific

The anatomical development of the wild-type leaf hasbeen characterized in some detail (Pyke et al., 1991) anda plethora of mutants with altered leaf shape is available(Tsuge et al., 1996; Berna et al., 1999; Serrano-Cartagenaet al., 1999) for the model species Arabidopsis thaliana.However, many leaf mutants are not exclusively affectedin leaf organogenesis. For example, in a large-scale screen-ing for mutants affected in their leaf development 153 mu-tants were isolated, 78 of which were also visibly affectedin their flower development (Berna et al., 1999; Roblesand Micol, 2001). Since the lateral organs of plants are as-sumed to have evolved from leaves, it is not surprising thata number of leaf mutants display, in addition, defects in

Fig. 8. Histochemical assay for GUS activity in transgenic plants expressing PRE:GUS in a wild-type Col-0 background. (A) Expanded fifth leaf, and(B) third and fourth expanding leaves and fifth leaf primordia of a 19-d-old plant. (C) Cauline leaf, (D) inflorescences, (E) main root tip, and (F) seeds ofa 34-d-old plant. Bar=50 lm (A, B, E, F) and 1 mm (C, D).

Fig. 9. Genetic interactions between the re-3, dov1, and cue1-5 mutations. Rosettes are shown from single mutants (A–C) and double mutants (D, E).Pictures were taken 18 d after sowing. Bar=1 mm.



Dow

nloaded from


floral organs and fruits. On the other hand, pleiotropyappears in some cases because the leaf phenotype is due tothe ectopic derepression in the leaves of the shoot apicalmeristem (Lincoln et al., 1994) or floral organ identity genes(Goodrich et al., 1997; Serrano-Cartagena et al., 1999).

The final shape and size of all lateral organs is achievedthrough the genetic control of cell expansion, proliferation,and identity along the dorsoventral, mediolateral, andproximodistal axes. Mutants affected in leaf dorsoventral-ity, such as kanadi (kan, Kerstetter et al., 2000), phabulosa(phb; McConnell and Barton, 1998; McConnell et al.,2001), and phavoluta (phv; McConnell et al., 2001), displayradialization in both their leaves and floral organs, ashappens in the phantastica (phan) mutants of Antirrhinummajus (Waites and Hudson, 1995; Waites et al., 1998) andlam1 of Nicotiana tabacum (McHale and Marcotrigiano,1998). Furthermore, the angustifolia mutants of Arabidop-sis are affected in cell expansion (an; Tsuge et al., 1996)and cell proliferation (an3; Horiguchi et al., 2005) along themediolateral axis, and display narrow leaves and flowerorgans. The same can be said with regard to the proximo-distal axis, as shown by the rotundifolia mutants ofArabidopsis, which display short leaves and floral organsbecause of defective cell expansion (rot3; Tsuge et al.,1996) and cell proliferation (rot4; Narita et al., 2004).The CINCINNATA (CIN) gene of Antirrhinum majus(Nath et al., 2003; Crawford et al., 2004) controls thegrowth of both leaves and petals. The asymmetric leaves 1(as1; Byrne et al., 2000) and as2 (Semiarti et al., 2001)Arabidopsis mutants are affected in both the proximodistaland mediolateral patterning of leaves, and their flowersare also abnormal.

Although many mutants isolated based on their leafphenotypes are pleiotropic, this is not the case for the remutants, whose only abnormality as regards externalmorphology is the reticulation displayed by the cotyledons,vegetative leaves, and cauline leaves. Only a few other leafmutants are known to be leaf-specific, such as liguleless1and liguleless2, which are uniquely necessary for ligule andauricle development in the leaves of maize (Harper andFreeling, 1996).

The re alleles strongly disrupt internal leaf architecturebut not plastid development

Mutation at the RE locus causes reticulation, due toa marked reduction in the density of mesophyll cells ininterveinal regions of the leaf, which does not result fromperturbed plastid development in specific cells but ratherfrom a dramatic change in internal leaf architecture.Contrary to what has been concluded from previous studieson other mutants, this characterization of the RE generevealed that the cellular basis for the reticulate phenotypeof leaves does not necessarily result from aberrations inplastid development. It has been shown, in addition, that

this phenotypic class may be usefully exploited as a basefor mutant screens which may reveal altered internal leafarchitecture.

It has been suggested that altered internal leaf archi-tecture results from the absence of a putative chloroplast-derived signal which controls palisade mesophyll celldifferentiation and expansion (Pyke et al., 2000). In fact, anumber of mutants with reticulate, pale or albino pheno-types show aberrant mesophyll development. Unlike re,all these mutants show defective chloroplast development,which suggests a role for the RE gene in a mesophyll-specific cell proliferation regulatory system that is notdependent on plastid development.

Mutations that disrupt internal leaf architecture andcause pale or albino phenotypes include pac (pale cress;Reiter et al., 1994; Grevelding et al., 1996), dal1 (DAG-like1; Babiychuk et al., 1997), var1 and var2 (yellowvariegated1 and 2; Chen et al., 2000; Takechi et al., 2000;Sakamoto et al., 2002), im (immutans; Wetzel et al., 1994;Carol et al., 1999; Wu et al., 1999) and sca3 (scabra3;Hricova et al., 2006), all of them in Arabidopsis, and thedcl (defective chloroplast and leaves-mutable; Keddieet al., 1996) and dag (Dof affecting germination; Sommeret al., 1985; Chatterjee et al., 1996) mutants in tomato andAntirrhinum majus, respectively.

The reticulate mutants reported in Arabidopsis includecue1 and dov1. In the leaves of cue1, palisade parenchymacells show a spheric rather than a columellar shape withlarger air-spaces and smaller chloroplasts than the wild type(Li et al., 1995). CUE1 encodes the plastid phosphoenol-pyruvate (PEP)/phosphate translocator (PPT), which islocalized on the inner envelope and imports PEP, the firstsubstrate of the shikimate pathway which produces aro-matic amino acids and a variety of secondary metabolites(Streatfield et al., 1999). The normal-looking bundle sheathcells and plastids found in paraveinal regions of cue1/cue1leaves, but abnormal mesophyll cells and plastids, indicatethat mesophyll cells specifically require PPT or that it isnot required by bundle sheath cells. The epistasis of cue1-5to re-3 that was observed indicates a functional relation-ship between CUE1 and RE and suggests a role for REin the shikimate pathway.

The dov1 mutant has normal bundle sheath cells andchloroplasts, whereas its mesophyll cells are normal in sizeand number, but contain plastids that are reduced in sizeand number, are vacuolated, and lack grana. The DOV1gene remains to be cloned. Our double mutant analysessuggest that RE and DOV1 are involved in differentprocesses affecting leaf development.

The re alleles have little effect on leaf shape

Whilst many studies have focused on roles for differentialplanes of cell division and leaf expansion in controlling leafshape (Poethig, 1997), the relative roles of the epidermis



Dow

nloaded from


and the mesophyll are far from clear (Dale, 1988). The anand rot3 mutations in Arabidopsis result in changes in leafallometry, to which a major contributory factor is abnor-mality in the polarity of epidermal cell expansion (Tsukayaet al., 1994; Tsuge et al., 1996). In leaves of cotton, theOkra mutation dramatically affects leaf shape, yet canproduce an altered leaf phenotype when present solely inthe L1-derived epidermis, as found by studies of chimeras(Dolan and Poethig, 1998).

Impaired mesophyll cell development resulting frommutation at the RE locus dramatically alters internal leafarchitecture, but has little effect on leaf shape. Only a slightdecrease in the final leaf area of the re mutants was found,suggesting that RE might also contribute to some extentto leaf expansion. Leaf thickness was reduced in re/reinterveinal regions, resulting in ridges of venation on theleaf surface. In addition, it was found that the size andmorphology of the epidermal cells was not different in there mutants and their wild types. Taken together, theseobservations suggest that the epidermis plays a major rolein controlling leaf shape in Arabidopsis, whereas thedivision and expansion of the internal mesophyll paren-chyma tissue would appear to be secondary in this process.This is not totally surprising, since Avery (1933) alreadyclaimed that the epidermis controls leaf expansion, basedon observations performed in tobacco. Since the epidermis,which is genetically distinct and lineage-related, is the onlycontiguous cellular structure which covers the entire leaf, itwould seem logical that the epidermal layer may containpre-determined information concerning leaf shape, whichcan be modulated to varying extents by the underlyingmesophyll tissue. In an analogous developmental system,the Arabidopsis petal, the epidermis is the primary tissuesince the mesophyll is rudimentary and poorly developed(Pyke and Page, 1998). As a result, the expansion of epi-dermal cells is a major factor in controlling overall petalexpansion and petal shape.

The evidence derived from this characterization of theleaf phenotype of re mutants is also compatible with thehypothesis that the control of leaf shape is non-cellautonomous, as suggested by the tangled mutant of maize(Smith et al., 1996). If leaf shape were controlled at tissueor organ level rather than at cell level, reduced mesophyllcell packing would not presumably change leaf shape at all,but merely produce a leaf with more internal airspace, aswas shown in the re mutants.

The fact that the leaves of re mutants are thinner ininterveinal areas because the mesophyll tissue is incompleteand the observation that the abaxial epidermis appears to belobed in places suggest that the development of themesophyll parenchyma plays a more important role in thecontrol of leaf thickness. Indeed, the extent of palisademesophyll cell elongation in a plane perpendicular to thelamina surface is closely correlated with leaf thicknessthroughout the process of Arabidopsis leaf development

and in mature leaves (Pyke et al., 1991). Despite thereduction in mesophyll cell number in the re lines, thepalisade and spongy mesophyll cells which are presentmaintain their position-specific characteristics, with similarsize and shape and chloroplast complement as the wildtype. Nevertheless, an increased number of chloroplasts perbundle sheath cell was observed in the re-1 mutant, sug-gesting that RE might also restrict the number of chloro-plasts in these cells. It is interesting that reduced internalmesophyll proliferation affects vascular development,which might indicate that recruitment of cells along thevascular cell differentiation pathway from mesophyll par-enchyma is limiting in this mutant.

Nature and action of the RE gene

The RE gene was positionally cloned and it was found tocorrespond to LCD1, which was identified based on theincreased sensitivity to ozone caused by its mutant allelelcd1-1 (Barth and Conklin, 2003). The lcd1-1 mutationdoes not affect the photosynthetic apparatus or chloroplast-protective mechanisms (Barth and Conklin, 2003), whichis consistent with the observation here that the reticulatephenotype was due to defects in mesophyll cells but notin chloroplasts.

Mutants that display both developmental and defencedefects have been reported, in which some defence com-ponents affect cell proliferation or vice versa (see, amongothers, Vanacker et al., 2000; Holt et al., 2002; Jin et al.,2002; Song et al., 2004). This was also the case for lcd1-1,whose impaired defence responses have been interpretedas secondary effects of the reduction in the number ofleaf parenchyma cells, since O3 degradation would yielda higher number of radical oxygen species per cell, andpronounced tissue damage, in a mesophyll extremelyreduced in cell density (Barth and Conklin, 2003). Furtherresearch will be required, however, to exclude completelythat the perturbations in defence associated with mutationsat RE (LCD) cause their developmental aberrations.

The lcd1-1 mutant carries a nonsense mutation thateliminates only the last C-terminal amino acid of the REprotein. No other allele of a plant gene has been describedin which the absence of a single, C-terminal amino acidresults in a mutant phenotype (Barth and Conklin, 2003).The results presented here demonstrate for re-1 and predictfor re-2, re-5, and re-6 that they are null mutations, allof which damage the RE gene much more severely thanlcd1-1. Therefore, the isolation and characterization of alarge series of loss-of-function re alleles allowed the in-volvement of the RE gene in the control of internal leafarchitecture in Arabidopsis to be confirmed.

RE encodes a plant-specific protein, to which no func-tion could be assigned on the basis of its similarity to otherproteins found in the databases. In silico analysis detectedthe presence of two transmembrane regions, and a putative



Dow

nloaded from


transit peptide to the chloroplast in the N-terminal part ofthe RE protein. Therefore, the involvement of RE in someplastid function not required for plastid development can-not be ruled out, given that its absence has no visible effectson plastid biogenesis. On the other hand, RE was notidentified as being included in the chloroplast proteome ina recent attempt to isolate all chloroplast proteins, by meansof tandem mass spectrometry (Kleffman et al., 2004).Therefore, a function for RE outside this organelle cannotbe excluded.

Using a reporter transgene, RE activity was detectednot only in leaves but also in roots and floral organs.Interestingly, RE activity was especially high in leaf pri-mordia, as expected from the role assumed for this gene inthe early stages of leaf development. In fact, the leafphenotype of the re mutants suggests that RE functions inearly primordia development to control the division ofmesophyll parenchyma cells as the lamina expands. Sub-sequent differentiation pathways of those cells are normalin all features other than the final ratio of the major celltypes of the leaf.

Supplementary data

Supplementary data can be found at JXB online.

Acknowledgements

We wish to thank the NASC and the ABRC for seeds, JavierPaz-Ares for providing the T-DNA seed collection from which re-5was isolated, and Anton Page, Sarah Case, Jose Manuel Serrano, andVeronica Garcıa-Sempere for excellent technical assistance. Thiswork was funded by the Biotechnology and Biological SciencesResearch Council (grant no. GO4843 to KAP), and the Ministerio deEducacion y Ciencia of Spain (grant nos BMC2002-02840 andBFU2005-01031 to JLM). RGB was fellow of the Ministerio deEducacion y Ciencia of Spain.

References

Avery GS. 1933. Structure and development of the tobacco leaf.American Journal of Botany 20, 565–592.

Babiychuk E, Fuangthong M, Van Montagu M, Inze D,Kushnir S. 1997. Efficient gene tagging in Arabidopsis thalianausing a gene trap approach. Proceedings of the National Academyof Sciences, USA 94, 12722–12727.

Barth C, Conklin PL. 2003. The lower cell density of leafparenchyma in the Arabidopsis thaliana mutant lcd1-1 is associ-ated with increased sensitivity to ozone and virulent Pseudomonassyringae. The Plant Journal 35, 206–218.

Bell DJ, Ecker JR. 1994. Assignment of 30 microsatellite loci tothe linkage map of Arabidopsis. Genomics 19, 137–144.

Berna G, Robles P, Micol JL. 1999. A mutational analysis of leafmorphogenesis in Arabidopsis thaliana. Genetics 152, 729–742.

Blazquez MA, Soowal LN, Lee I, Weigel D. 1997. LEAFYexpression and flower initiation in Arabidopsis. Development124, 3835–3844.

Bohmert K, Camus I, Bellini C, Bouchez D, Caboche M,Benning C. 1998. AGO1 defines a novel locus of Arabidopsiscontrolling leaf development. EMBO Journal 17, 170–180.

Byrne ME. 2005. Networks in leaf development. Current Opinionin Plant Biology 8, 59–66.

Byrne ME, Barley R, Curtis M, Arroyo JM, Dunham M,Hudson A, Martienssen RA. 2000. Asymmetric leaves1mediatesleaf patterning and stem cell function in Arabidopsis. Nature408, 967–971.

Canales C, Grigg S, Tsiantis M. 2005. The formation andpatterning of leaves: recent advances. Planta 221, 752–756.

Candela H, Martınez-Laborda A, Micol JL. 1999. Venationpattern formation in Arabidopsis thaliana vegetative leaves.Developmental Biology 205, 205–216.

Carland FM, McHale NA. 1996. LOP1: a gene involved in auxintransport and vascular patterning in Arabidopsis. Development122, 1811–1819.

Carol P, Stevenson D, Bisanz C, Breitenbach J, Sandmann G,Mache R, Coupland G, Kuntz M. 1999. Mutations in theArabidopsis gene IMMUTANS cause a variegated phenotype byinactivating a chloroplast terminal oxidase associated with phy-toene desaturation. The Plant Cell 11, 57–68.

Chatterjee M, Sparvoli S, Edmunds C, Garosi P, Findlay K,Martin C. 1996. DAG, a gene required for chloroplast differen-tiation and palisade development in Antirrhinum majus. EMBOJournal 15, 4194–4207.

Chen M, Choi Y, Voytas DF, Rodermel S. 2000. Mutations in theArabidopsis VAR2 locus cause leaf variegation due to the loss ofa chloroplast FtsH protease. The Plant Journal 22, 303–313.

Clough SJ, Bent AF. 1998. Floral dip: a simplified method forAgrobacterium-mediated transformation of Arabidopsis. ThePlant Journal 16, 735–743.

Crawford BC, Nath U, Carpenter R, Coen ES. 2004. CINCIN-NATA controls both cell differentiation and growth in petal lobesand leaves of Antirrhinum. Plant Physiology 135, 244–253.

Dale JE. 1988. The control of leaf expansion. Annual Review ofPlant Physiology and Plant Molecular Biology 39, 267–295.

Dolan L, Poethig RS. 1998. The Okra leaf shape mutation in cottonis active in all cell layers of the leaf. American Journal of Botany85, 322–327.

Fleming AJ. 2002. The mechanism of leaf morphogenesis. Planta216, 17–22.

Goodrich J, Puangsomlee P, Martin M, Long D, Meyerowitz EM,Coupland G. 1997. A Polycomb-group gene regulates homeoticgene expression in Arabidopsis. Nature 386, 44–51.

Grevelding C, Suter-Crazzolara C, von Menges A, Kemper E,Masterson R, Schell J, Reiss B. 1996. Characterisation of a newallele of pale cress and its role in greening in Arabidopsis thaliana.Molecular and General Genetics 251, 532–541.

Harper L, Freeling M. 1996. Interactions of liguleless1 andliguleless2 function during ligule induction in maize. Genetics144, 1871–1882.

Holt III BF, Boyes DC, Ellerstrom M, Siefers N, Wiig A,Kauffman S, Grant MR, Dangl J. 2002. An evolutionarilyconserved mediator of plant disease resistance gene function isrequired for normal Arabidopsis development. DevelopmentalCell 2, 807–817.

Horiguchi G, Kim GT, Tsukaya H. 2005. The transcription factorAtGRF5 and the transcription coactivator AN3 regulate cellproliferation in leaf primordia of Arabidopsis. The Plant Journal43, 68–78.

Hricova A, Quesada V, Micol JL. 2006. The SCABRA3 nucleargene encodes the plastid Rpo Tp RNA polymerase, which isrequired for chloroplast biogenesis and mesophyll cell proliferationin Arabidopsis. Plant Physiology 141, 942–958.



Dow

nloaded from


Jin H, Axtell MJ, Dahlbeck D, Ekwenna O, Zhang S,Staskawicz B, Baker B. 2002. NPK1, an MEKK1-like mitogen-activated protein kinase, regulates innate immunity and develop-ment in plants. Developmental Cell 3, 291–297.

Kawaguchi R, Bailey-Serres J. 2002. Regulation of trans-lational initiation in plants. Current Opinion in Plant Biology5, 460–465.

Keddie J, Carroll B, Jones J, Gruissem W. 1996. The DCL geneof tomato is required for chloroplast development and palisadecell morphogenesis in leaves. EMBO Journal 15, 4208–4217.

Kerstetter RA, Bollman K, Taylor RA, Bomblies K, Poethig RS.2001. KANADI regulates organ polarity in Arabidopsis. Nature411, 706–709.

Kinsman EA, Pyke KA. 1998. Bundle sheath cells and cell-specificplastid development in Arabidopsis leaves. Development 125,1815–1822.

Kleffmann T, Russenberger D, Von Zychlinski A, Christopher W,Sjolander K, Gruissem W, Baginsky S. 2004. The Arabidopsisthaliana chloroplast proteome reveals pathway abundance andnovel protein functions. Current Biology 14, 354–362.

Larkin JC, Marks MD, Nadeau J, Sack F. 1997. Epidermal cellfate and patterning in leaves. The Plant Cell 9, 1109–1120.

Li H, Culligan K, Dixon RA, Chory J. 1995. CUE1: a mesophyllcell-specific positive regulator of light-controlled gene expressionin Arabidopsis. The Plant Cell 7, 1599–1610.

Lincoln C, Long J, Yamaguchi J, Serikawa K, Hake S. 1994. Aknotted 1-like homeobox gene in Arabidopsis is expressed in thevegetative meristem and dramatically alters leaf morphology whenoverexpressed in transgenic plants. The Plant Cell 6, 1859–1876.

McConnell JR, Barton MK. 1998. Leaf polarity and meristemformation in Arabidopsis thaliana. Development 125, 2935–2942.

McConnell JR, Emery J, Eshed Y, Bao N, Bowman J,Barton MK. 2001. Role of PHABULOSA and PHAVOLUTA inradial patterning in shoots. Nature 411, 709–713.

McHale NA, Marcotrigiano M. 1998. LAM1 is required fordorsoventrality and lateral growth of the leaf blade in Nicotiana.Development 125, 4235–4243.

Meinke DW, Koornneef M. 1997. Community standards forArabidopsis genetics. The Plant Journal 12, 247–253.

Narita NN, Moore S, Horiguchi G, Kubo M, Demura T, FukudaH, Goodrich J, Tsukaya H. 2004. Overexpression of a novelsmall peptide ROTUNDIFOLIA4 decreases cell proliferation andalters leaf shape in Arabidopsis. The Plant Journal 38, 699–713.

Nath U, Crawford BC, Carpenter R, Coen E. 2003. Geneticcontrol of surface curvature. Science 299, 1404–1407.

Nelson T, Dengler N. 1997. Leaf vascular pattern formation. ThePlant Cell 9, 1121–1135.

Perez-Perez JM, Ponce MR, Micol JL. 2004. The ULTRACUR-VATA2 gene of Arabidopsis thaliana encodes an FK506-bindingprotein involved in auxin and brassinosteroid signaling. PlantPhysiology 134, 101–117.

Perez-Perez JM, Serrano-Cartagena J, Micol JL. 2002. Geneticanalysis of natural variations in the architecture of vegetativeleaves in Arabidopsis thaliana. Genetics 162, 893–915.

Poethig RS. 1997. Leaf morphogenesis in flowering plants. ThePlant Cell 9, 1077–1087.

Ponce MR, Quesada V, Micol JL. 1998. Rapid discriminationof sequences flanking and within T-DNA insertions in theArabidopsis genome. The Plant Journal 14, 497–501.

Ponce MR, Robles P, Micol JL. 1999. High-throughput geneticmapping in Arabidopsis. Molecular and General Genetics 261,408–415.

Pyke KA, Leech RM. 1992. Nuclear mutations radically alterchloroplast division and expansion in Arabidopsis thaliana.Plant Physiology 99, 1005–1008.

Pyke KA, Lopez-Juez E. 1999. Cell differentiation and leafmorphogenesis in Arabidopsis. Critical Reviews of Plant Science18, 527–546.

Pyke KA, Marrison JL, Leech RM. 1991. Temporal and spatialdevelopment of the cells of the expanding first leaf of Arabid-opsis thaliana (L.) Heynh. Journal of Experimental Botany 42,1407–1416.

Pyke KA, Page AM. 1998. Plastid ontogeny during petal de-velopment in Arabidopsis. Plant Physiology 116, 797–803.

Pyke KA, Zwobka M, Day A. 2000. Marking cell layers inBrassica napus with spectinomycin provides a new tool forstudying cell fate and demonstrates a requirement for chloroplastsin palisade cell differentiation. Journal of Experimental Botany51, 1713–1720.

Quesada V, Garcıa-Martınez S, Piqueras P, Ponce MR,Micol JL. 2002. Genetic architecture of NaCl tolerance inArabidopsis thaliana. Plant Physiology 130, 951–963.

Quesada V, Ponce MR, Micol JL. 1999. OTC and AUL1, twoconvergent and overlapping genes in the nuclear genome ofArabidopsis thaliana. FEBS Letters 461, 101–106.

Redei GP, Hirono Y. 1964. Linkage studies. Arabidopsis Inform-ation Service 1, 9–10.

Reiter RS, Coomber SA, Bouret TM, Bartley GE, Scolnik PA.1994. Control of leaf and chloroplast development by theArabidopsis gene pale cress. The Plant Cell 6, 1253–1264.

Robles P, Micol JL. 2001. Genome-wide linkage analysis ofArabidopsis genes required for leaf development. Molecular andGeneral Genetics 266, 12–19.

Sakamoto W, Tamura T, Hanba-Tomita Y, Sodmergen,Murata M. 2002. The VAR1 locus of Arabidopsis encodes achloroplastic FtsH and is responsible for leaf variegation in themutant alleles. Genes and Cells 7, 769–780.

Semiarti E, Ueno Y, Tsukaya H, Iwakawa H, Machida C,Machida Y. 2001. The ASYMMETRIC LEAVES2 gene ofArabidopsis thaliana regulates formation of a symmetric lamina,establishment of venation and repression of meristem-relatedhomeobox genes in leaves. Development 128, 1771–1783.

Serrano-Cartagena J, Candela H, Robles P, Ponce MR,Perez-Perez JM, Piqueras P, Micol JL. 2000. Genetic analysis ofincurvata mutants reveals three independent genetic operations atwork in Arabidopsis leaf morphogenesis.Genetics 156, 1363–1377.

Serrano-Cartagena J, Robles P, Ponce MR, Micol JL. 1999.Genetic analysis of leaf form mutants from the ArabidopsisInformation Service collection. Molecular and General Genetics261, 725–739.

Smith LG, Hake S, Sylvester AW. 1996. The tangled-1 mutationalters cell division orientations throughout maize leaf developmentwithout altering leaf shape. Development 122, 481–489.

Sommer H, Carpenter R, Harrison BJ, Saedler H. 1985. Thetransposable element Tam3 of Antirrhinum majus generates anovel type of sequence alteration upon excision. Molecular andGeneral Genetics 199, 225–231.

Song JT, Lu H, Greenberg JT. 2004. Divergent roles in Arabi-dopsis thaliana development and defense of two homologousgenes, ABERRANT GROWTH AND DEATH2 and AGD2-LIKEDEFENSE RESPONSE PROTEIN1, encoding novel aminotrans-ferases. The Plant Cell 16, 353–366.

Streatfield SJ, Weber A, Kinsman EA, Hausler RE. Li J,Post-Beittenmiller D, Kaiser WM, Pyke KA, Flugge UI,Chory J. 1999. The phosphoenolpyruvate/phosphate translocatoris requiredfor phenolic metabolism, palisade cell development, andplastid-dependent nuclear gene expression. The Plant Cell 11,1609–1622.

Takechi K, Sodmergen, Murata M, Motoyoshi F, Sakamoto W.2000. The YELLOW VARIEGATED (VAR2) locus encodes a



Dow

nloaded from


homologue of FtsH, an ATP-dependent protease in Arabidopsis.Plant Cell Physiology 41, 1334–1346.

Tsiantis M, Hay A. 2003. Comparative plant development: the timeof the leaf? Nature Reviews Genetics 4, 169–180.

Tsuge T, Tsukaya H, Uchimiya H. 1996. Two independent andpolarized processes of cell elongation regulate leaf blade expan-sion in Arabidopsis thaliana (L.) Heynh. Development 122,1589–1600.

Tsukaya H. 2002. Leaf development. In: Somerville CR,Meyerowitz EM, eds. The Arabidopsis Book. Rockville, FL:American Society of Plant Biologists, 1–23.

Tsukaya H. 2003. Organ shape and size: a lesson from studies ofleaf morphogenesis. Current Opinion in Plant Biology 6, 57–62.

Tsukaya H. 2005. Leaf shape: genetic controls and environmentalfactors. International Journal of Developmenta Biology 49,547–555.

Tsukaya H, Tsuge T, Uchimiya H. 1994. The cotyledon: a superiorsystem for studies of leaf development. Planta 195, 309–312.

Vanacker H, Lu H, Rate DN, Greenberg JT. 2001. A role forsalicylic acid and NPR1 in regulating cell growth in Arabidopsis.The Plant Journal 28, 209–216.

Waites R, Hudson A. 1995. phantastica: a gene required fordorsoventrality of leaves in Antirrhinum majus. Development121, 2143–2154.

Waites R, Selvadurai HR, Oliver IR, Hudson A. 1998. ThePHANTASTICA gene encodes a MYB transcription factor in-volved in growth and dorsoventrality of lateral organs inAntirrhinum. Cell 93, 779–789.

Wetzel CM, Jiang CZ, Meehan LJ, Voytas DF, Rodermel SR.1994. Nuclear-organelle interactions: the immutans variegationmutant of Arabidopsis is plastid autonomous and impaired incarotenoid biosynthesis. The Plant Journal 6, 161–175.

Wu D, Wright DA, Wetzel C, Voytas DF, Rodermel S. 1999.The IMMUTANS variegation locus of Arabidopsis defines a mito-chondrial alternative oxidase homolog that functions duringearly chloroplast biogenesis. The Plant Cell 11, 43–55.

Ye ZH, Freshour G, Hahn MG, Burk DH, Zhong R. 2002.Vascular development in Arabidopsis. International Review ofCytology 220, 225–256.

Zimmermann P, Hirsch-Hoffmann M, Hennig L, Gruissem W.2004. Genevestigator. Arabidopsis microarray database and analy-sis toolbox. Plant Physiology 136, 2621–2632.



Dow

nloaded from


mutations in the reticulata gene dramatically alter internal architecture but have little effect on...

Documents