The Arabidopsis Locus RCB Mediates UpstreamRegulation of Mitotic Gene Expression1

Kristiina Himanen2, Christophe Reuzeau, Tom Beeckman, Siegbert Melzer, Olivier Grandjean,Liz Corben3, and Dirk Inze*

Department of Plant Systems Biology, Flanders Interuniversity Institute for Biotechnology, Ghent University,Technologiepark 927, B–9052 Gent, Belgium (K.H., T.B., L.C., D.I.); CropDesign N.V., B–9052 Gent, Belgium(C.R.); Universite de Liege, Sart-Tilman, B–4000 Liege, Belgium (S.M.); and Institut National de la RechercheAgronomique, Centre de Versailles, Route de Saint-Cyr, F–78026, France (O.G.)

Transcriptional regulation of cell cycle regulatory genes, such as B-type cyclins, is tightly linked with the mitotic activity inthe meristems. To study the regulation of a B-type cyclin gene, a targeted genetic approach was undertaken. An Arabidopsisline containing a fusion construct between the CYCB1;1 promoter and a bacterial �-glucuronidase marker gene (uidA) wasused in ethyl methanesulfonate mutagenesis. The mutants were screened for altered CYCB1;1::uidA expression patterns. Ina reduced CYCB1;1 expression mutant (rcb), the CYCB1;1::uidA expression was severely affected, being excluded from theshoot and root apical meristems and leaf primordia and shifted to cells associated with root cap and stomata. In additionto the overall reduction of the endogenous CYCB1;1 transcript levels, other G2-to-M phase-specific genes were alsodown-regulated by the mutation. In the mutant plants, the inflorescence stem growth was reduced, indicating low meristemactivity. Based on the altered CYCB1;1::uidA expression patterns in rcb root meristem, a model is proposed for RCB thatmediates the tissue specificity of CYCB1;1 promoter activity.

The eukaryotic cell cycle is controlled at two majorcheckpoints, one before DNA replication (G1 to S)and one before mitosis (G2 to M). In yeast (Saccharo-myces cerevisiae) and animals, progression throughthese checkpoints is driven by cyclin-dependent ki-nases (CDKs), the activity of which depends on in-teraction with their regulatory subunits, the cyclins.Different cyclins mediate the cell cycle phase-specificactivation of the CDKs and modify their substratespecificity in a spatial and temporal manner (Stalsand Inze, 2001). In Arabidopsis, five classes of CDKshave been identified (Vandepoele et al., 2002), butonly two of them have been extensively character-ized: CDKA;1 (formerly known as Cdc2a), which isactive at the G1-to-S and G2-to-M checkpoints, andCDKB1;1 (formerly known as Cdc2b), whose activityis highest at the G2-to-M-transition (Mironov et al.,1999; Joubes et al., 2000; Porceddu et al., 2001;Menges and Murray, 2002). CDKB1;1 represents a

plant-specific CDK that is thought to interact withthe mitotic cyclin CYCB1;1 at the G2-to-M transition(Criqui et al., 2000) and also both genes show acoordinated transcriptional up-regulation during theG2-to-M phase (Mironov et al., 1999). Three mainclasses of cyclins (A, B, and D type) have been de-scribed in plants, and in the Arabidopsis genome, atotal of 30 cyclins have been identified (Vandepoeleet al., 2002). D-type cyclins are preferentially inducedby mitogen stimuli at the G1 phase (Soni et al., 1995;Fuerst et al., 1996; Richard et al., 2002). Unlike theD-type cyclins, the levels of A- and B-type cyclins aretightly controlled by cell cycle progression. The pro-moter activities and transcript levels of A-type cyc-lins have been shown to increase at mid S phase(Fuerst et al., 1996; Shaul et al., 1996). For B-typecyclins, G2 and M phase-specific peaks in expressionhave been shown in many plant species, such asArabidopsis (Ferreira et al., 1994a), Catharanthus ro-seus (Ito et al., 1998), and Nicotiana sylvestris (Trehinet al., 1999).

As regulatory proteins, cyclins have a high turn-over rate, and their cyclic appearance results fromstringent regulation at the transcriptional level(Shaul et al., 1996; Criqui et al., 2001). A well-characterized example is the B-type cyclin, CYCB1;1.Transcriptional regulation of CYCB1;1 is tightly re-lated to cell cycle activity and can be used as amarker for detecting mitotically active tissues (Fer-reira et al., 1994b; Shaul et al., 1996; Colon-Carmonaet al., 1999). The expression of the CYCB1;1 promoter-driven bacterial gus (�-glucuronidase) gene (CYCB1;1;;uidA) has been shown to correlate well with the

1 This work was supported by the Interuniversity Poles of At-traction Program (Belgian State, Prime Minister’s Office, FederalOffice for Scientific, Technical, and Cultural Affairs, grant no.P5/13), by the Academy of Finland (fellowship to K.H.), and bythe Finnish Cultural Foundation for fellowships (fellowship toK.H.).

2 Present address: Laboratory of Plant Molecular Biology, Rock-efeller University, 1230 York Avenue, New York, NY 10021.

3 Present address: University of East Anglia, Norwich NR4 7TJ,UK.

* Corresponding author; e-mail dirk.inze@psb.ugent.be; fax32–9 –3313809.

Article, publication date, and citation information can be foundat http://www.plantphysiol.org/cgi/doi/10.1104/pp.103.027128.

1862 Plant Physiology, December 2003, Vol. 133, pp. 1862–1872, www.plantphysiol.org © 2003 American Society of Plant Biologists

mRNA localization of the endogenous gene (deAlmeida Engler et al., 1999). The cell cycle-specificgene activation of CYCB1;1 is mediated via cis-actingelements in the promoter (Ito et al., 1998, 2001; Trehinet al., 1999; Planchais et al., 2002).

Despite extensive analysis of the transcriptionalcontrol of the mitotic cyclin CYCB1;1, little is knownabout the regulatory pathways that affect its pro-moter activity. In an attempt to study the upstreamregulation of the CYCB1;1 promoter, a targeted ge-netic approach was undertaken. The CYCB1;1::uidAline was used in chemical mutagenesis, thereby al-lowing visualization of mutations that specificallyaffect the promoter activity. Mutants were screenedfor altered patterns of CYCB1;1::uidA expression. Wereport the identification and characterization of amutant line with reduced CYCB1;1::uidA expression(designated rcb). The rcb mutant lost CYCB1;1::uidAexpression in both shoot and root apical meristemsand leaf primordia, whereas the expression was ec-topically induced in cells associated with root cap andstomata. In addition, the overall endogenous levels ofthe CYCB1;1 transcripts and those of other mitoticgenes were reduced in the rcb mutant. In matureplants, the total length of the inflorescence stem andits diameter were significantly affected, indicating aneffect on the meristem activity. Based on the alteredCYCB1;1::uidA expression patterns in rcb root meris-tems, we propose a model in which RCB mediates thetissue specificity of the CYCB1;1 promoter activity.

RESULTS

Isolation of the rcb Mutant and Genetic Analysis

To study the regulation of the CYCB1;1 promoteractivity, 2,000 seeds of the transgenic Arabidopsisline containing a CYCB1;1::uidA promoter fusion(Ferreira et al., 1994b) were mutagenized with ethylmethanesulfonate. The M1 plants were self-fertilized,and 7,850 of the M2 plants (314 M2 families) werescreened for altered patterns of CYCB1;1::uidA pro-moter activity in the root tips. From the mutantscreening, one line with rcb expression was chosenfor further analysis. In this line, the CYCB1;1::uidAexpression appeared to be ectopically induced in theroot cap cells, whereas it was absent from the rootapical meristem (Fig. 1, A and B). To confirm that thealtered CYCB1;1::uidA expression in rcb was notcaused by a mutation in the CYCB1;1::uidA promoter,the wild-type and rcb promoters (1.27 kb) from theT-DNA constructs were isolated by PCR and se-quenced. The sequences did not differ, indicatingthat the altered expression was caused by an inde-pendent mutation.

Before further phenotypic analysis, the mutant linewas subjected to three successive backcrosses withthe CYCB1;1::uidA starter line. In the F1 population,gain of CYCB1;1::uidA expression in the root capwas detected in all of the progeny, indicating that the

rcb mutation was semidominant. In the F2 popula-tion, three classes of expression patterns were ob-served. When compared with the homozygous wild-type (RCB/RCB; Fig. 1A) and homozygous mutantplants (rcb/rcb; Fig. 1B), the heterozygous mutants(rcb/RCB; Fig. 1C) had intermediate expression pat-terns. As with the homozygous mutants, the het-erozygous mutants (F1, 100%; and F2, 50%) showed again of CYCB1;1::uidA expression in the lateral rootcap. In the root apical meristem, however, interme-diate intensities of the CYCB1;1::uidA expressionwere observed, whereas in the homozygous mutants,no expression was detected in that tissue. In the F2population, the segregation pattern of ectopicCYCB1;1::uidA expression in the root cap again sug-gested the semidominant nature of the rcb mutation.

To further confirm that the rcb mutation was inde-pendent from the CYCB1;1::uidA transgene, testcrosses were performed between the rcb mutants andthe C24 wild-type plants. The CYCB1;1::uidA expres-sion pattern in the F1 population resembled that ob-tained from the backcrosses, although the GUS stain-ing was often much weaker because of the reducedamount of CYCB1;1::uidA T-DNA in the progeny. Inthe F2 population, CYCB1;1::uidA-driven GUS stain-ing patterns typical for wild-type (RCB/RCB, 24%),heterozygous (rcb/RCB, 47%), rcb mutant (rcb/rcb,24%), and GUS-negative (5%) genotypes were en-countered. This result confirmed that the mutationwas independent from the uidA promoter fusion it-self because the wild-type (RCB/RCB) CYCB1;1::uidAexpression patterns could be recovered without in-troducing wild-type alleles of the uidA construct.However, the segregation pattern of the GUS-negative plants was only 5% in the F2 population,indicating that rcb contained two alleles of the uidAtransgene. The segregation pattern of GUS-negativelines (5%) was also confirmed by kanamycin selec-tion: Only 10 of 200 F2 plants were sensitive to theselection. This result again underlines the fact thatthe rcb mutation resides outside the CYCB1;1::uidAfusion because the mutation was able to relocalizethe expression of the two CYCB1;1::uidA transgenes.

To determine the genomic locus of the rcb muta-tion, genetic mapping was initiated with amplifiedfragment length polymorphism (AFLP) markers (Pe-ters et al., 2001). To introduce mapped markers intothe rcb mutant line, the M3 line of rcb (in ArabidopsisC24 ecotype) was crossed with Columbia-0 (Col-0)ecotype. From a segregating F2 population, 40 mu-tants, four wild-type plants, and the parental lineswere selected by GUS assay of 2-week-old root seg-ments. The selected lines were processed for bulkedsegregant AFLP analysis according to Peters et al.(2001). From each of the 16 primer combinations used,on average five polymorphic markers for each wereidentified for C24 and Col-0 ecotypes. The bulkedsegregant analysis showed linkage with eight markerslocated in the lower arm of chromosome 2 for the rcb

Arabidopsis Locus Affecting Mitotic Gene Expression

Plant Physiol. Vol. 133, 2003 1863

mutation (data not shown). Also, the non-linkage wasanalyzed, and the rcb mutation was not linked withthe polymorphic markers on chromosomes 1, 3, 4, or 5.The CYCB1;1 gene itself is located in chromosome 4(At4g37490). In addition, single simple length poly-morphism markers (Bell and Ecker, 1994) for each ofthe Arabidopsis chromosomes were tested to confirmthe putative loci. Analysis of 40 mutants with chromo-some 2-specific markers (nga1126 [chromosome 2, 51cM], nga168 [chromosome 2, 74 cM], and an INDELmarker in chromosome 2 [79 cM]) gave segregation

patterns of 28:23:28 individuals for the C24 genotype,11:19:13 for the heterozygous mutant genotype, and1:0:1 for the Col-0 genotype, respectively. These datasuggest that the rcb mutation is located on the lowerarm of chromosome 2.

The CYCB1;1::uidA Expression and CYCB1;1 mRNAPatterns in the rcb Mutant

The rcb mutant was selected on the basis of aremarkable change in the CYCB1;1 promoter-driven

Figure 1. Bright- and dark-field images of whole-mount GUS-stained plants. A, CYCB1;1::uidA inwild type. B, CYCB1;1::uidA in rcb mutant. C,CYCB1;1::uidA in a heterozygous (rcb/RCB) plantwith weak expression in the meristem combinedwith root cap expression. Anatomical analysis ofhistochemically stained Arabidopsis seedlingroots. D, Longitudinal section (5 �m) of 9-d-oldwild-type seedling root in which CYCB1;1::uidAstaining is restricted to epidermal and corticalcell files. E, Longitudinal section of rcb mutantwith CYCB1;1::uidA expression shifted to cells ofthe lateral root cap and absence of expression inthe root apical meristem. Differential interferencecontrast microscopy of histochemically stainedleaves at different developmental stages in 4-d-old seedlings (F–K). F to H, CYCB1;1::uidA inwild type. I to K, CYCB1;1::uidA in rcb. F and I,Shoot apices showing the emerging first pair ofleaf primordia (SAM). G and J, Young developingleaves. H and K, Detailed view of adaxial epider-mis of developing leaves showing the stomatalcomplexes. H, CYCB1;1::uidA expression re-stricted to developing stomata at the meristemoidstage in wild type. K, Localization ofCYCB1;1::uidA in palisade cells underneath ma-ture guard cells; arrows indicate unstained sto-mata. Expression of CYCB1;1::uidA during inflo-rescence development (L and M). L, GUS stainingin wild-type inflorescences at different develop-mental stages. CYCB1;1::uidA is expressed in theentire pistil. M, Presence of GUS staining in wild-type ovaries and absence from the nectaries. N,CYCB1;1::uidA expression in rcb inflorescences.CYCB1;1::uidA is absent from pistils but stronglyexpressed in the nectaries (O). C, cortex; CRC,columella root cap; E, epidermis; En, endoder-mis; LRC, lateral root cap; S, stele. Bars � 200�m (A–C), 100 �m (D, E, F, and I), 1 mm (G andJ), and 50 �m (H and K).

Himanen et al.

1864 Plant Physiol. Vol. 133, 2003

GUS staining patterns. To further characterize thecell specificity of the CYCB1;1::uidA expression in rcbroots, the wild-type expression pattern was analyzedin anatomical sections. In the primary root apicalmeristem of the wild-type plant, the CYCB1;1::uidAexpression was detected in a cell type-specific man-ner in the epidermal and cortical cell files (Fig. 1D),whereas in the newly developed meristems of lateralroot primordia, the expression was uniform (Ferreiraet al., 1994b; Beeckman et al., 2001). In rcb, this ex-pression pattern was severely altered. The expressionwas absent from the epidermal and cortical cell filesin the root apex but was ectopically induced in thelateral root cap initials and their immediate daughtercells (Fig. 1E), where no expression in the wild typecould be detected. Other meristematic tissues ofwild-type and rcb plants were analyzed to seewhether the effect of rcb mutation was root specific.

Wild-type shoot apical meristem and emerging leafprimordia expressed CYCB1;1::uidA strongly (Fig.1F). In young wild-type leaf primordia, the basipetalgradient of cell division activity could be observed(Fig. 1G; see also Donnelly et al., 1999). During thewild-type stomata development, the expression wasrestricted to the meristemoids (Fig. 1H; see also Sernaand Fenoll, 1997). In rcb, the meristematic cells in theshoot apex, leaf primordia, and young leaves had noCYCB1;1::uidA promoter activity (Fig. 1I). Instead, inrcb, strong expression was present in the area ofhydathodes (Fig. 1J) and ectopically induced in thepalisade parenchyma cells beneath each developingstomatum (Fig. 1K).

Also, in rcb flowers, an altered CYCB1;1::uidA ex-pression was observed. In the flowers of wild-typeplants, CYCB1;1::uidA expression was strong in de-veloping pistil, excluding the stigma and nectaries(Fig. 1, L and M). In contrast, in the rcb mutant line,no GUS staining was detected in the pistil, whereasthe nectaries were strongly stained (Fig. 1, N and O).

Thus, in different tissues and organs of the rcbmutant, a shift in CYCB1;1::uidA localization takesplace in comparison with the wild type. Both in theroot and shoot apical meristems, the strong meris-tematic CYCB1;1::uidA expression was lost in rcb,whereas an ectopic expression was induced in tissuesthat usually have no mitotic activity. In addition, thesemidominant nature of the rcb mutant indicates thatRCB could encode a dose-dependent activator of theCYCB1;1 promoter in meristematic tissues and per-haps a repressor outside the meristems.

Root Cap Maturation in rcb

In the rcb roots, the CYCB1;1::uidA expression wasaffected in a cell type-specific manner, namely ec-topically induced in the lateral root cap initial cells.We analyzed the CYCB1;1::uidA expression in moredetail in wild-type and rcb roots at different stages ofroot cap development: In rcb, it characteristically

changed during root cap development, whereas inwild-type plants, it was high in the developing lateralroot primordia (Fig. 2A) until starch accumulated inthe newly forming root cap cells (Fig. 2B). Upon dif-ferentiation of statocyte layers in the columella, theCYCB1;1::uidA expression diminished (Fig. 2C), and in

Figure 2. Comparative analysis of CYCB1;1::uidA expression duringlateral root development and de novo root cap formation under dark-field stereoscopy. A to D, Young lateral root primordia, emerginglateral roots, outgrowing lateral roots, and lateral root with fully de-veloped root caps in wild type, respectively. E to H, Young lateral rootprimordia, emerging lateral roots, outgrowing lateral roots, and lateralroot with fully developed root caps in rcb mutant, respectively. F,Statoliths (arrow) become refractive under dark-field optics. G,CYCB1;1::uidA is only expressed (arrow) in rcb at the moment stato-cyte organization starts to appear. I and J, CYCB1;1::uidA expression in2,4-dichlorophenoxyacetic acid-treated seedling roots of wild typeand rcb, respectively. Bars � 1 mm (A–H), 200 �m (I and J).

Arabidopsis Locus Affecting Mitotic Gene Expression

Plant Physiol. Vol. 133, 2003 1865

mature wild-type root caps, it was undetectable (Fig.2D). In contrast, in rcb, the CYCB1;1::uidA expressionfollowed an opposite pattern and appeared to betightly linked with the development of the statocytetissues. In young lateral root primordia (before devel-opment of the statocytes), CYCB1;1::uidA was not ex-pressed (Fig. 2, E and F). The CYCB1;1::uidA expres-sion in lateral root caps of rcb appeared with thematuration of the statocytes in the columella (Fig. 2G).At the time the statocyte layers in the columella werefully developed, the lateral root cap cells showedstrong GUS staining, whereas the columella remainedunstained (Fig. 2H).

In rcb mutants, the CYCB1;1::uidA expression wasabsent from the root apical meristem and was exclu-sively associated with the root cap-specified zone.We tested two conditions under which the size andcell patterning of root meristems were severely al-tered. Treatment with the auxin transport inhibitornaphthylphthalamic acid is known to cause spanningof tissue with root cap identity upwards from theroot tip (Sabatini et al., 1999). In the rcb mutant, thetreatment with 10�5 m naphthylphthalamic acid ledto expansion of the root cap with the characteristicCYCB1;1::uidA expression for rcb (data not shown). Fur-thermore, synthetic auxin 2,4-dichlorophenoxyaceticacid was used to induce ectopic cell division in themeristem. In both wild-type and rcb plants, treatmentwith 2,4-dichlorophenoxyacetic acid increased the mer-istematic tissue in both plant types and induced astrong expression of CYCB1;1::uidA in the expandedmeristem of wild-type plants (Fig. 2I) but not in the rcbroot. In the mutant, the typically restricted pattern ofCYCB1;1::uidA expression was observed (Fig. 2J).

Shoot Phenotype

To analyze the shoot-specific growth phenotypes ofrcb, the mutant and CYCB1;1::uidA wild-type plants(C24 ecotype) and the CYCB1;1 knockout line fromthe Versailles T-DNA collection and the correspond-ing wild-type plants (Wassilewskija [Ws] ecotype)

were grown under short- (8 h/16 h) and long- (16 h/8h) day light conditions. Growth was analyzed by com-paring the number and size of rosette leaves, numberof rosette and lateral branches, plant height, floweringtime, and flower development. Observations weremade on 40 plants in at least three replicates for thegrowth conditions. Although under constant lightconditions no consistent growth phenotypes were de-tected, under short- and long-day conditions, thegrowth of rcb plants was considerably altered whencompared with that of the wild-type controls (Table I).The growth of rcb plants was repressed at the stage ofinflorescence stem development, which were approx-imately 25% shorter in height and more rigid than thewild-type plants (Fig. 3A). Flowering time in rcb wasdelayed by approximately 6 d. Rosette leaf growthwas similar to that of wild-type controls, without anysignificant difference in leaf development and leafnumber. Rosette branch development was reduced,whereas lateral branches developed as those of controlplants. The growth of the CYCB1;1 knockout plantswas comparable with that of the wild-type plants,with only a significantly reduced number of leavesand rosette branches in the mutant plants (Table I).

In rcb, the silique length and diameter were signif-icantly shorter (40%) and larger than those of wild-type plants, respectively (Table I). Furthermore, un-der both long- and short-day conditions, the stemdiameter increased significantly in rcb plants fromthe basis of the rosette when the flowering stem was2 d old (1-cm total height) until the end of the devel-opment. An increase in diameter of the lateralbranches of rcb was similarly observed.

To identify the origin of the stem diameter in-crease, the stems were sectioned at different posi-tions, starting from 1 cm above the rosette (below anylateral branches) and at approximately 10 cm fromthe rosette. The rcb stem sections were larger but hada normal radial symmetry consisting of an epidermallayer, four to five layers of chlorenchyme, a layer ofalternating vascular bundles and intervascular fibers,and a central zone. In control plants, the average

Table I. Phenotypic measurements of wild-type C24 and rcb mutants under short-day conditions 11 weeks after sowing

Values were averaged from five sections for each sample. Observations were made on 40 plants in at least three differentreplicates.

Phenotypic trait CYCB1;1::uidA (C24) rcb (C24) Wild Type (Ws) CYCB1;1 Mutant (Ws)

Longest rosette leaf (cm) 81.6 80.5 79.3 80.4No. of rosette leaves 18 16 59.6 28.5Plant height (cm) 22.2 16.5 35.5 30.6No. of rosette branches 6.4 4.1 11.1 2.6No. of lateral branches 7.1 7.3 7.2 10.1Time to anthesis (d) 78 83 73 72Silique length (mm) 14.3 8.6 14.5 14.4Cells in inflorescence stem

No. of cells in pith 24.1 29.8 ND a NDNo of cells in cortex 5.0 5.2 ND NDNo. of interfascicular cells 5.2 9.1 ND ND

aND, Not determined.

Himanen et al.

1866 Plant Physiol. Vol. 133, 2003

number of vascular strands in the stem sections wasnine, instead of an average of 11 in rcb plants (Fig. 3,B and C). The structure of the vascular strands in rcbdiffered from that of control plants by larger vascularbundles at the base of the stem and at one-third of theoverall stem height. The difference in shape of thevascular bundles appears to be due to an increase incell size both in the phloem and in the xylem regions(Fig. 3, D and E). Also, the cells between the vascularbundles were enlarged, contributing to the overallexpansion of the stem diameter (Fig. 3, F and G). Thenumber of cells in the inflorescence stem was esti-mated by calculating the number of cells from 10images taken of sections cut 1 cm above the rosette.Although the cell numbers did not change in thecortex, in the central stele, and especially in the in-terfascicular zone, considerably more cells werepresent in rcb than in the wild type (Table I).

Kinematic analysis of root growth from germina-tion until 2 weeks of age, quantification of the num-ber of lateral roots, and flow cytometric analysis ofroot and leaf tissues did not reveal any differencesbetween rcb and wild-type plants (data not shown).

rcb Down-Regulates a Set of MitoticGenes in Arabidopsis

To test whether the rcb mutation affected other cellcycle genes, transcript levels of CYCA2;1, CYCB1;1,CYCB2;1, CDKA;1, and CDKB1;1 were analyzed bysemiquantitative RT-PCR as described previously by

Himanen et al. (2002). One week after germination,seedlings were used for RNA extraction. The mutanthad reduced CYCB1;1 transcript levels when com-pared with those of wild-type plants, in agreementwith the reduced GUS activity. In addition to down-regulation of CYCB1;1, other G2-to-M and M phase-specific genes, such as CYCB2;1 and CDKB1;1, alsoshowed similarly reduced transcript levels (Fig. 4A).Markers of the G1-to-S cell cycle phase, such asCYCA2;1 and CDKA;1, were not affected by the mu-tation. To further confirm the reduced endogenoustranscript levels of CYCB1;1, in situ hybridization wasperformed on shoot apices from wild-type and mutant

Figure 3. Shoot phenotype. A, Wild-type and rcb plants grown under long-day conditions. B and C, Hand sections of 5- to6-week-old wild-type C24 and rcb plants, respectively. D and E, Transverse section of the basal part of the stem showinga vascular bundle of wild type and rcb, respectively. F and G, Transverse section of the basal part of the stem showing theinterfascicular region of wild type and rcb, respectively. C, cortex; E, epidermis; iV, interfascicular region; P, phloem; Pt,pith; V, vascular bundle; X, xylem. Bars � 0.5 mm (B), 1 mm (C), and 10 �m (D–G).

Figure 4. Reverse transcription (RT)-PCR gel blot. A, Transcript levelof the cell cycle regulatory genes (CYCB1;1, CYCB2;1, CYCA2;1,CDKA;1, CDKB1;1, and ACTIN-2) in 1-week-old wild-type and rcbseedlings. B to E, In situ hybridization of CYCB1;1 mRNA on shootapical meristem tissue of rcb mutant and wild-type plants. B, rcbCYCB1;1 sense probe. C, rcb hybridized with CYCB1;1 antisenseprobe. D, Wild type hybridized with CYCB1;1 sense probe. D, Wildtype hybridized with CYCB1;1 antisense probe.

Arabidopsis Locus Affecting Mitotic Gene Expression

Plant Physiol. Vol. 133, 2003 1867

plants grown under conditions described by Corbesieret al. (1996). In the rcb shoot apical meristem, no signalof CYCB1;1 mRNA could be detected with antisense orsense probes (Fig. 4, B and C). In wild-type plants, thelocalization of CYCB1;1 mRNA was patchy when hy-bridized with antisense probes, whereas the senseprobes gave no signal (Fig. 4, D and E).

DISCUSSION

rcb Shows an “Inverse” Development of theCYCB1;1::uidA Expression Pattern during Root CapMaturation Compared with the Wild Type

The most striking phenotype of the rcb mutantwas the ectopic expression of CYCB1;1::uidA ob-served in the root cap where wild-type plants do notshow any expression. Our results indicate that theCYCB1;1::uidA expression pattern is strictly localizedin specific cell types, depending on the developmen-tal stage of the particular organ. In the wild-type rootapical meristem, the CYCB1;1::uidA is expressed inthe dividing cortical and epidermal cell files, whereasin the rcb, no expression is detected. However, dur-ing the lateral root development in rcb, theCYCB1;1::uidA expression is only induced upon rootcap maturation and becomes restricted to the lateralroot cap cells, whereas it is not expressed in themature root cap of the wild type. In pea (Pisumsativum), the root cap meristem is regulated indepen-dently from the primary root apical meristem and isprogrammed to produce a species-specific amount ofroot cap cells (Hawes et al., 1998). When the organreaches a certain size, cell production ceases. Thecells differentiate progressively through a series ofdevelopmental stages until the cells at the peripheryof the root cap separate as border cells. The separa-tion of these metabolically active cells depends on theenvironmental conditions, such as water potential.When incubated in water with gentle agitation, theborder cells respond immediately by expansion andrelease from the root cap, whereas under dry condi-tions, they remain attached to the root. When theborder cells are still attached to a mature root cap, theroot cap meristem arrests in the G1 phase of the cellcycle (Brigham et al., 1998). As a sign of the G1 arrest,the mature root cap cells fail to express the histoneH2 gene (Tanimoto et al., 1993). However, upon re-moval of the border cells, a cell division marker geneis induced within 15 min (Woo and Hawes, 1997).

In Arabidopsis, the border cells appear to be tightlyassociated with the root cap and are not releasedduring water agitation treatment (Hawes et al., 1998).The slow growth rate of the root cap indicates thatthe cell division activity is generally low. This con-clusion is supported by lack of CYCB1;1::uidA (ourobservations) and CYCA2;1 expression, as well as bythymidine labeling in wild-type root caps (Burssenset al., 2000a). In rcb, an opposite development isobserved, because the marker gene for mitotic activ-

ity, CYCB1;1:uidA, is ectopically induced in the lat-eral root cap cells. However, no differences in theroot cap size or structure are observed in the mu-tants, indicating that the ectopic CYCB1;1::uidA ex-pression alone is not adequate to drive additional celldivisions in these cells.

In rcb Inflorescence, Meristem Activity Is Affected ButNot Organ Initiation

Based on the phenotypic analysis of the rcb mutant,we conclude that in addition to the regulation ofCYCB1;1 tissue specificity in roots, RCB is necessaryfor inflorescence meristem activity. Plant develop-ment is generally due to a balance between cell divi-sion in meristematic tissues to produce organs andcell growth and differentiation to allow organ growthand maturation. The stem phenotype appearedrather late, during rosette and inflorescence growth,namely at the onset of inflorescence stem growth.Because the number of lateral organs remained un-altered, the size and the early functioning of theshoot apical meristem were probably not stronglyaffected. However, the meristem function may bemodified at a later stage, perhaps by affecting the cellproduction rate in the meristem responsible for in-florescence stem growth. Furthermore, increasednumber of cells in the transverse sections of the in-florescence stems was detected, suggesting that thesegrowth activities are regulated independently fromeach other.

In the flower organs of wild-type plants,CYCB1;1::uidA expression was found in the young tipof the pistil where actively dividing cells are present(Liu et al., 1997; Broadhvest et al., 2000). In rcb, theCYCB1;1::uidA expression pattern is different becauseit is shifted from the pistil to the stamen. The siliquesof rcb were also significantly shorter and wider thanthose of the wild-type lines, indicating that RCB wasnecessary to maintain CYCB1;1 gene expression inpistils of the wild-type plants, and lack of CYCB1;1expression in rcb pistils repressed the mature siliquelength.

Redundancy in Plant Cell Cycle

The rcb mutation affected various organs. In rootsand leaves, where changes in CYCB1;1::uidA expres-sion were detected, no growth phenotypes were ob-served. On the other hand, the inflorescence stemhad a strong growth phenotype probably because ofthe effect of the rcb mutation on the expression ofCYCB1;1 in the shoot apex. In addition, a similarresponse was observed in the siliques, which re-mained shorter but broader than in wild type. Inter-estingly, the phenotype of a CYCB1;1 knockout mu-tant differed from that observed in rcb, showing adecrease in number of rosette leaves and rosettebranches. These observations indicate that the tissue

Himanen et al.

1868 Plant Physiol. Vol. 133, 2003

specificity of CYCB1;1 expression and CYCB1;1knockout had a different effect on the shoot apicalmeristem activity.

In addition to the down-regulation of CYCB1;1, thercb mutation also down-regulated the M phase-specific genes (CYCB2;1, and CDKB1;1) at the tran-scriptional level. The fact that the rcb mutant was stillable to grow indicated that the functions of thesegenes have probably been taken over by other genes.These data suggest a high level of redundancy for cellcycle genes in plants. Similar situations are found inyeast with three G1 cyclins (Cln1, Cln2, and Cln3).Only loss-of-function of the Cln3 delays slightly en-try into S phase; however, cells that are double mu-tants for certain combinations of these genes have amore severe cell cycle delay than the single mutants(Thomas, 1993). Also, the four mitotic B-type cyclinsof yeast (Clb1, Clb2, Clb3, and Clb4) are highly re-dundant (Tjandra et al., 1998). Only deletion of Clb2retards mitosis, whereas any of the other three mi-totic cyclins can be deleted without noticeable phe-notypes (Amon et al., 1993). Nevertheless, only dele-tion of the three other genes can create a Clb2-dependent genetic background. In addition, the yeastcyclins appear to be capable of complementing eachother’s functions, despite the cell cycle phase inwhich they normally function. In yeast, two addi-tional B-type cyclins, Clb5 and Clb6, are produced atthe onset of the S phase, but they still display func-tional redundancy with the other four B-type cyclins(Segal et al., 1998).

In mammals, such flexibility does not seem to exist.In mice (Mus musculus), deletion of one of the twoB-type cyclins (CycB1) is lethal, although the functionof the other, CycB2, can be compensated (Brandeis etal., 1998). In animal systems, transcriptional up-regulation of the B-type cyclins have been shown toplay a central role in regulating the entry into mitosis(Murray and Kirschner, 1989; Nurse, 1990). In plants,CYCB1;1 has been suggested to be the main mitoticcyclin (Doerner et al., 1996). Ectopic expression ofCYCB1;1 stimulates cell division activity in root api-cal meristems, indicating that the level of CYCB1;1 isa limiting factor for the entry into mitosis. However,the absence of severe growth phenotypes in rcb ar-gues against this conclusion. Clearly, reduced levelsof CYCB1;1, CYCB2;1, and CDKB1;1 had no dramaticeffects on cell cycle progression. Other B-type, oreven A-type, cyclins probably take over the role ofcyclins with defective function.

Not much is known about the redundancy betweenplant cyclins. Immunolocalization studies in maize(Zea mays) have revealed different subcellular local-izations for several closely related plant cyclins(Mews et al., 1997, 2000), and strictly localized pat-terns also have been reported for the D-type cyclinsin the floral meristems of snapdragon (Antirrhinummajus; Gaudin et al., 2000). In addition, we haveshown strictly tissue-specific expression patterns for

CYCB1;1::uidA in the cortical and epidermal cells inwild-type root apical meristems. However, based ondatabase information of Arabidopsis, at least 30 cyc-lins are predicted (Vandepoele et al., 2002). Such ahigh number of cyclins is more than that reported forany other organism to date. The actual function of allthe cyclins still remains to be studied, but plant-specific features of growth (no cell migration) anddevelopment (postembryonic organ development)may require special regulation of the cell cycle. Towhat extent these numerous plant cyclins can conferfunctional redundancy between each other still re-mains to be elucidated.

RCB May Have a Dual Function in Regulating theTissue Specificity of the CYCB1;1 Expression

We aimed at identifying direct regulators ofCYCB1;1 promoter activities. In the rcb mutant, iden-tified from the screen of ethyl methanesulfonate-mutagenized CYCB1;1::uidA plants, a shift in tissuespecificity of GUS activity was encountered, indicat-ing that RCB itself may be situated upstream ofthe CYCB1;1 promoter regulation. Based on theCYCB1;1::uidA expression patterns in wild-type andrcb mutants, we propose a model for the function ofRCB. In this model, RCB plays a dual role both as apositive regulator for CYCB1;1 promoter activation inthe outer layers of the root apical meristem and as arepressor in lateral root cap cells (Fig. 5A). Thisconclusion derives from the observed loss ofCYCB1;1::uidA expression in the root apical meristemand, on the other hand, on the gain-of-expression inthe lateral root cap in rcb (Fig. 5B). Further supportfor the model comes from the analysis of the het-erozygous mutants in which the meristematic expres-sion has been reduced to an intermediate level incombination with a gain-of-expression in the root cap(Fig. 5C). It is interesting that a similar change inCYCB1;1::uidA expression patterns has been ob-served in the shoot of rcb. The typically strong ex-pression pattern of wild-type plants in the shootapical meristem is absent in rcb; instead, a shift toparenchyma cells beneath the developing stomata isobserved. It is tempting to speculate that RCB en-codes a trans-acting factor that would act both as anactivator and a repressor, depending on the interac-tion with other proteins. The existence of trans-actingfactors, which act both positively and negatively onthe expression of cell cycle genes, is not unprece-dented. In animals, the heterotrimeric transcriptionfactor NFY has been shown to transcriptionally in-hibit CycB1, CycB2, and Cdc25 promoters upon DNAdamage-induced G2 arrest (Manni et al., 2001). Inaddition to negative regulation, CCAAT-bindingtranscription factors also mediate positive regulationof cell cycle phase-specific CycB1 expression (Katulaet al., 1997; Sciortino et al., 2001). Together with othertranscription factors, NFYs appear to mediate balanc-

Arabidopsis Locus Affecting Mitotic Gene Expression

Plant Physiol. Vol. 133, 2003 1869

ing between activation and repression of genes in atissue-specific manner (Gilthorpe et al., 2002).

MATERIALS AND METHODS

Plant Material

The transgenic Arabidopsis line CYCB1;1::uidA (Ferreira et al., 1994b) wasin the genetic background of the C24 ecotype. For in vitro culture, seedswere sterilized and sown on K1 germination media (Valvekens et al., 1988)without vitamins. Plants were grown in a growth chamber with continuouslight (110 �E m�1 s�1 photosynthetically active radiation supplied bycool-white fluorescent tubes) held at 22°C. Plants were grown on squareplates (Greiner Labortechnik, Frickenhausen, Germany) in a vertical posi-tion to facilitate the accessibility to the root system. Material was collectedat various time points depending on the requirements of each experiment.For phenotypic characterization, seedlings were used 36 h, 4 d, 1 week, or 2weeks after germination. After 3 weeks of sterile culture, plants weretransferred to a mixture of soil and vermiculite (1:3 [w/v]) for self-fertilization under the same growth conditions.

Mutagenesis

The Arabidopsis CYCB1;1::uidA promoter fusion line was chemicallymutagenized with ethyl methanesulfonate. Treatment with 0.3% (w/v)ethyl methanesulfonate was performed for 12 h, after which the seeds wereextensively washed with water. M1 plants were germinated on K1 mediumand transferred to soil after 2 weeks of growth for self-fertilization. M2progenies were analyzed in a GUS assay-based mutant screening. From eachline, 25 seeds were germinated together with wild-type controls. Rootcuttings were collected for the GUS assay after 1 week of growth. The lineswere screened for alterations in the pattern or intensity of the CYCB1;1::uidAexpression. The selected putative mutant lines were self-fertilized, and theM3 plants were subjected to three successive backcrosses with the starterline before phenotypic analysis. Intactness of the CYCB1;1::uidA promoterfusion in the mutants was controlled by sequencing. The promoters fromboth wild type and rcb were isolated by standard PCR amplification with Pfu(Promega, Madison, WI). Primers were designed to hybridize with the flank-ing sequences on each side of the actual CYCB1;1 promoter in theCYCB1;1::uidA construct and inside the promoter (5�CGCGATCCAGACT-GAATGCCCACAGGCCG with 5�CGTGCCACGCGCTACAGACCACGCCCand 5�GGGCGTGGTCTGTAGCGCGTGGCACG with 5�GCCTGGGGTGC-CTAATGAGTGAGAATTGACGG). PCR products were run on 0.9% (w/v)agarose gels, purified with Qiaquick Gel Extraction Kit (Qiagen, Hilden,Germany), and sequenced. Sequences from three replicate samples were

compared with the promoter from the mutant starter line in multiple align-ments (GCG Wisconsin Package; Accelrys, San Diego, CA).

Genetic Analysis

For mapping of the rcb mutant, an AFLP mapping system was used (Voset al., 1995). Homozygous mutant plants from the M3 population werecrossed with the parental line of the wild-type Col-0 ecotype, for whichAFLP markers were available. The F1 plants were self-fertilized, and thesegregating F2 progenies were screened by GUS assay to select mutantindividuals for mapping. AFLP reactions were performed on total genomicDNA isolated from 40 mutant plants, four wild-type individuals, and thetwo parental lines with the DNeasy protocol (Qiagen). The mutant DNAsamples were pooled in bulks of eight and were subjected to digestion withSacI and MseI enzymes, together with the wild-type and parental samples.Pre-amplification reactions were done using SacI and MseI primers withoutselective nucleotides, and the AFLP fingerprints were generated with twoselective nucleotides added to each primer. Adaptors and primers wereobtained from Genset (Paris). The SacI primers were 33P-labeled for visual-ization of the fragments by autoradiography. The AFLP banding patternsfrom 16 primer combinations were scored for presence or absence of bandsrepresenting polymorphic markers between Col-0 and C24 ecotypes. Themap position was deduced based on linkage and non-linkage of the markersfor the mutation. The chromosomal location of the uidA T-DNA was deter-mined from markers that were present in the mutant pools but absent fromthe four wild-type individuals. The marker analysis was performed bystandard PCR as described by Bell and Ecker (1994).

In Situ Hybridization

For in situ hybridization, the plant material was grown according to themethod of Corbesier et al. (1996). Longitudinal sections of shoot apices from2-month-old plants were hybridized as described by Segers et al. (1996). TheCYCB1;1 [35S]UTP-labeled antisense and sense probes were prepared fromthe linearized pGEMCYC1At vector using SP6 and T7 RNA polymerase,respectively.

Histochemical GUS Assays

For histochemical GUS assays, complete seedlings or root cuttings werestained in multiwell plates (Falcon, Becton-Dickinson, Bedford, MA). GUSassays were performed as described by Beeckman and Engler (1994). Sam-ples mounted in Tris-saline buffer or lactic acid were observed and photo-graphed with a stereoscope (Stemi SV11, Zeiss, Jena, Germany) or with

Figure 5. Model for the putative role of thewild-type RCB and the effects of homozygousand heterozygous rcb mutations. A, RCB activa-tor for CYCB1;1 genes in the root apical meris-tem and repressor for the same gene in thelateral root cap tissues in wild type. B, In thehomozygous rcb mutant, lack of CYCB1;1 ex-pression in the root apical meristem and ectopi-cal induction in the lateral root cap. C, In theheterozygous mutant, induction of intermediateCYCB1;1 expression in the root apical meristemand in the root cap cells.

Himanen et al.

1870 Plant Physiol. Vol. 133, 2003

differential contrast optics on a standard light microscope (Leica, Wien,Austria).

Microscopy

For anatomical sections, samples from the GUS assays were fixed over-night in 1% (v/v) glutaraldehyde and 4% (w/v) paraformaldehyde in 50mm phosphate buffer. Samples were dehydrated and embedded in Techno-vit 7100 resin (Heraeus Kulzer, Wehrheim, Germany) according to themanufacturer’s instructions. For proper orientation of the samples, trans-parent strips were used to facilitate alignment of the tissues (Beeckman andViane, 2000). Sections of 5-�m were cut with a microtome (Minot 1212, Leitz,Wetzlar, Germany), dried on object glasses, counterstained for cell wallswith 0.05% (w/v) ruthenium red (Fluka Chemical, Buchs, Switzerland) intap water for 30 s, mounted in DePex, and covered with coverslips foranalysis and photography.

Whole-Plant Analysis, AnatomicalSections, and Microscopy

Plants were grown at 20°C in small pots on soil in the greenhouse underlong-day conditions (16 h of light/8 h of darkness) with sodium lamps andat 15°C under short-day conditions (8 h of light/16 h of darkness) conditionswith cool-white lights (75–100 �E m�2, Philips, Eindhoven, The Nether-lands). Stem samples taken at 1 and 10 cm above the rosette were sectionedmanually. Tissues samples were treated 20 min in 50% (v/v) sodium hypo-chlorite, washed in water for 30 min, fixed for 5 min in acetic acid, stainedfor 5 min in 0.1% (w/v) toluidine blue, and washed in tap water. Sampleswere examined through a Wild M3 Microscope (Leica) and photographedunder bright field. Data were collected and analyzed with Optimas 6.0(MediaCybernetics, http://www.optimas.com/optimas.htm).

RT-PCR

Endogenous transcript levels of a set of cell cycle regulatory genes wereanalyzed by semiquantitative RT-PCR as described by Burssens et al.(2000b) and Himanen et al. (2002). For the cDNA synthesis, total RNA wasprepared with RNeasy reagents (Qiagen) from 1-week-old wild-type and rcbseedlings. The cDNA was prepared from three independent RNA samples,and 15, 20, and 25 cycles of PCR were tested to verify the exponential phaseof the amplification.

ACKNOWLEDGMENTS

The authors acknowledge Christiane Genetello for the effort in maintain-ing the mutant lines during the project, Janny Peters and Tom Gerats forhelp in determining the map position of the RCB locus, Katia Belcram for hertechnical contribution in the identification and analysis of shoot phenotypes,Nathalie Detry for in situ hybridizations, Rebecca Verbanck for artwork, JanZethof for INDEL primers, Meeta Mistry for critical reading of the manu-script, and Martine De Cock for help in preparing it.

Received May 21, 2003; returned for revision June 25, 2003; accepted Sep-tember 9, 2003.

LITERATURE CITED

Amon A, Tyers M, Futcher B, Nasmyth K (1993) Mechanisms that help theyeast cell cycle clock tick: G2 cyclins transcriptionally activate G2 cyclinsand repress G1 cyclins. Cell 74: 993–1007

Beeckman T, Burssens S, Inze D (2001) The peri-cell-cycle in Arabidopsis. JExp Bot 52: 403–411

Beeckman T, Engler G (1994) An easy technique for the clearing of histo-chemically stained plant tissue. Plant Mol Biol Rep 12: 37–42

Beeckman T, Viane R (2000) Embedding of thin plant specimens for ori-ented sectioning. Biotechnol Histochem 75: 23–26

Bell CJ, Ecker JR (1994) Assignment of 30 microsatellite loci to the linkagemap of Arabidopsis. Genomics 19: 137–144

Brandeis M, Rosewell I, Carrington M, Crompton T, Jacobs MA, Kirk J,Gannon J, Hunt T (1998) Cyclin B2-null mice develop normally and are

fertile whereas cyclin B1-null mice die in utero. Proc Natl Acad Sci USA95: 4344–4349

Brigham LA, Woo H-H, Wen F, Hawes MC (1998) Meristem-specific sup-pression of mitosis and a global switch in gene expression in the root capof pea by endogenous signals. Plant Physiol 118: 1223–1231

Broadhvest J, Baker SC, Gasser CS (2000) SHORT INTEGUMENTS 2 pro-motes growth during Arabidopsis reproductive development. Genetics155: 899–907

Burssens S, de Almeida Engler J, Beeckman T, Richard C, Shaul O,Ferreira P, Van Montagu M, Inze D (2000a) Developmental expression ofthe Arabidopsis thaliana CycA2;1 gene. Planta 211: 623–631

Burssens S, Himanen K, van de Cotte B, Beeckman T, Van Montagu M,Inze D, Verbruggen N (2000b) Expression of cell cycle regulatory genesand morphological alterations in response to salt stress in Arabidopsisthaliana. Planta 211: 632–640

Colon-Carmona A, You R, Haimovitch-Gal T, Doerner P (1999) Spatio-temporal analysis of mitotic activity with a labile cyclin-GUS fusionprotein. Plant J 20: 503–508

Corbesier L, Gadisseur I, Silvestre G, Jacqmard A, Bernier G (1996) Designin Arabidopsis thaliana of a synchronous system of floral induction by onelong day. Plant J 9: 947–952

Criqui MC, Parmentier Y, Derevier A, Shen W-H, Dong A, Genschik P(2000) Cell cycle-dependent proteolysis and ectopic overexpression ofcyclin B1 in tobacco BY2 cells. Plant J 24: 763–773

Criqui MC, Weingartner M, Capron A, Parmentier Y, Shen W-H, Heberle-Bors E, Bogre L, Genschik P (2001) Sub-cellular localisation of GFP-tagged tobacco mitotic cyclins during the cell cycle and after spindlecheckpoint activation. Plant J 28: 569–581

de Almeida Engler J, De Vleesschauwer V, Burssens S, Celenza JL Jr, InzeD, Van Montagu M, Engler G, Gheysen G (1999) Molecular markers andcell cycle inhibitors show the importance of cell cycle progression innematode-induced galls and syncytia. Plant Cell 11: 793–808

Doerner P, Jørgensen J-E, You R, Steppuhn J, Lamb C (1996) Control ofroot growth and development by cyclin expression. Nature 380: 520–523

Donnelly PM, Bonetta D, Tsukaya H, Dengler RE, Dengler NG (1999) Cellcycling and cell enlargement in developing leaves of Arabidopsis. Dev Biol215: 407–419

Ferreira P, Hemerly A, de Almeida Engler J, Bergounioux C, Burssens S,Van Montagu M, Engler G, Inze D (1994a) Three discrete classes ofArabidopsis cyclins are expressed during different intervals of the cellcycle. Proc Natl Acad Sci USA 91: 11313–11317

Ferreira PCG, Hemerly AS, de Almeida Engler J, Van Montagu M, EnglerG, Inze D (1994b) Developmental expression of the Arabidopsis cyclingene cyc1At. Plant Cell 6: 1763–1774

Fuerst R, Soni R, Murray J, Lindsey K (1996) Modulation of cyclin tran-script levels in cultured cells of Arabidopsis thaliana. Plant Physiol 112:1023–1233

Gaudin V, Lunness PA, Fobert PR, Towers M, Riou-Khamlichi C, MurrayJAH, Coen E, Doonan J (2000) The expression of D-cyclin genes definesdistinct developmental zones in snapdragon apical meristems and islocally regulated by the cycloidea gene. Plant Physiol 122: 1137–1148

Gilthorpe J, Vandromme M, Brend T, Gutman A, Summerbell D, Totty N,Rigby PWJ (2002) Spatially specific expression of Hoxb4 is dependent onthe ubiquitous transcription factor NFY. Development 129: 3887–3899

Hawes M, Brigham L, Wen F, Woo H, Zhu Y (1998) Function of root bordercells in plant health: pioneers in the rhizosphere. Annu Rev Phytopathol36: 311–327

Himanen K, Boucheron E, Vanneste S, de Almeida Engler J, Inze D,Beeckman T (2002) Auxin-mediated cell cycle activation during earlylateral root initiation. Plant Cell 14: 2339–2351

Ito M, Araki S, Matsunaga S, Itoh T, Nishihama R, Machida Y, DoonanJH, Watanabe A (2001) G2/M-phase-specific transcription during theplant cell cycle is mediated by c-Myb-like transcription factors. Plant Cell13: 1891–1905

Ito M, Iwase M, Kodama H, Lavisse P, Komamine A, Nishihama R,Machida Y, Watanabe A (1998) A novel cis-acting element in promotersof plant B-type cyclin genes activates M phase-specific transcription.Plant Cell 10: 331–341

Joubes J, Chevalier C, Dudits D, Heberle-Bors E, Inze D, Umeda M,Renaudin J-P (2000) CDK-related protein kinases in plants. Plant MolBiol 43: 607–620

Katula KS, Wright KL, Paul H, Surman DR, Nuckolls FJ, Smith JW, TingJP, Yates J, Cogswell JP (1997) Cyclin-dependent kinase activation and

Arabidopsis Locus Affecting Mitotic Gene Expression

Plant Physiol. Vol. 133, 2003 1871

S-phase induction of the B1 cyclin gene are linked through the CCAATelements. Cell Growth Diff 8: 811–820

Liu J-q, Seul U, Thompson R (1997) Cloning and characterization of apollen-specific cDNA encoding a glutamic-acid-rich protein (GARP)from potato Solanum berthaultii. Plant Mol Biol 33: 291–300

Manni I, Mazzaro G, Gurtner A, Mantovani R, Haugwitz U, Krause K,Engeland K, Sacchi A, Soddu S, Piaggio G (2001) NF-Y mediates thetranscriptional inhibition of the cyclin B1, cyclin B2 and cdc25C promotersupon induced G2 arrest. J Biol Chem 276: 5570–5576

Menges M, Murray JAH (2002) Synchronous Arabidopsis suspension cul-tures for analysis of cell-cycle gene activity. Plant J 20: 203–212

Mews M, Sek FJ, Moore R, Volkmann D, Gunning BES, John PCL (1997)Mitotic cyclin distribution during maize cell division: implications for thesequence diversity and function of cyclins in plants. Protoplasma 200:128–145

Mews M, Sek FJ, Volkmann D, John PCL (2000) Immunodetection of fourmitotic cyclins and the Cdc2a protein kinase in the maize root: theirdistribution in cell development and dedifferentiation. Protoplasma 212:236–249

Mironov V, De Veylder L, Van Montagu M, Inze D (1999) Cyclin-dependent kinases and cell division in higher plants: the nexus. Plant Cell11: 509–521

Murray AW, Kirschner MW (1989) Cyclin synthesis drives the early em-bryonic cell cycle. Nature 339: 275–280

Nurse P (1990) Universal control mechanism regulating onset of M-phase.Nature 344: 503–508

Peters JL, Constandt H, Neyt P, Cnops G, Zethof J, Zabeau M, Gerats T(2001) A physical amplified-fragment length polymorphism map of Ara-bidopsis. Plant Physiol 127: 1579–1589

Planchais S, Perennes C, Glab N, Mironov V, Inze D, Bergounioux C(2002) Characterisation of cis-acting element involved in cell cycle phase-independent activation of Arath;CycB1;1 transcription and identificationof putative regulatory proteins. Plant Mol Biol 50: 109–125

Porceddu A, Stals H, Reichheld J-P, Segers G, De Veylder L, De PinhoBarroco R, Casteels P, Van Montagu M, Inze D, Mironov V (2001) Aplant-specific cyclin-dependent kinase is involved in the control of G2/Mprogression in plants. J Biol Chem 276: 36354–36360

Richard C, Lescot M, Inze D, De Veylder L (2002) Effect of auxin, cytokinin,and sucrose on cell cycle gene expression in Arabidopsis thaliana cellsuspension cultures. Plant Tissue Organ Cult 69: 167–176

Sabatini S, Beis D, Wolkenfelt H, Murfett J, Guilfoyle T, Malamy J,Benfey P, Leyser O, Bechtold N, Weisbeek P et al. (1999) An auxin-dependent distal organizer of pattern and polarity in the Arabidopsis root.Cell 99: 463–472

Sciortino S, Gurtner A, Manni I, Fontemaggi G, Dey A, Sacchi A, Ozato K,Piaggio G (2001) The cyclin B1 gene is actively transcribed during mitosisin HeLa cells. EMBO Rep 2: 1018–1023

Segal M, Clarke DJ, Reed SI (1998) Clb5-associated kinase activity isrequired early in the spindle pathway for correct preanaphase nuclearpositioning in Saccharomyces cerevisiae. J Cell Biol 143: 135–145

Segers G, Gadisseur I, Bergounioux C, de Almeida Engler J, Jacqmard A,Van Montagu M, Inze D (1996) The Arabidopsis cyclin-dependent kinasegene cdc2bAt is preferentially expressed during S and G2 phases of thecell cycle. Plant J 10: 601–612

Serna L, Fenoll C (1997) Tracing the ontogeny of stomatal clusters inArabidopsis with molecular markers. Plant J 12: 747–755

Shaul O, Mironov V, Burssens S, Van Montagu M, Inze D (1996) TwoArabidopsis cyclin promoters mediate distinctive transcriptional oscilla-tion in synchronized tobacco BY-2 cells. Proc Natl Acad Sci USA 93:4868–4872

Soni R, Carmichael J, Shah Z, Murray J (1995) A family of cyclin Dhomologs from plants differentially controlled by growth regulators andcontaining the conserved retinoblastoma protein interaction motif. PlantCell 7: 85–103

Stals H, Inze D (2001) When plant cells decide to divide. Trends Plant Sci6: 359–364

Tanimoto EY, Rost TL, Comai L (1993) DNA replication-dependent histoneH2A mRNA expression in pea root tips. Plant Physiol 103: 1291–1297

Thomas JH (1993) Thinking about genetic redundancy. Trends Genet 9:395–398

Tjandra H, Compton J, Kellogg D (1998) Control of mitotic events by theCdc42 GTPase, the Clb2 cyclin and a member of the PAK kinase family.Curr Biol 8: 991–1000

Trehin C, Glab N, Perennes C, Planchais S, Bergounioux C (1999) Mphase-specific activation of the Nicotiana sylvestris Cyclin B1 promoterinvolves multiple regulatory elements. Plant J 17: 263–273

Valvekens D, Van Montagu M, Van Lijsebettens M (1988) Agrobacteriumtumefaciens-mediated transformation of Arabidopsis thaliana root explantsby using kanamycin selection. Proc Natl Acad Sci USA 85: 5536–5540

Vandepoele K, Raes J, De Veylder L, Rouze P, Rombauts S, Inze D (2002)Genome-wide analysis of core cell cycle genes in Arabidopsis. Plant Cell14: 903–916

Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Hornes M, FrijtersA, Pot J, Peleman J, Kuiper M et al. (1995) AFLP: a new technique forDNA fingerprinting. Nucleic Acids Res 23: 4407–4414

Woo H-H, Hawes MC (1997) Cloning of genes whose expression is corre-lated with mitosis and localized in dividing cells in root caps of Pisumsativum L. Plant Mol Biol 35: 1045–1051

Himanen et al.

1872 Plant Physiol. Vol. 133, 2003