《植物生物学》课程教学资源（文献资料）The Arabidopsis MERISTEM DISORGANIZATION 1 gene is

SEBthe plant journalSorietyThePlant Journal (2011)68,657-669doi:10.1111/j.1365-313X.2011.04718.xTheArabidopsisMERISTEMDISORGANIZATION1geneisrequired forthe maintenance of stem cells throughthereduction of DNA damageYuma Hashimura and Chiharu Ueguchi"Bioscienceand BiotechnologyCenter,NagoyaUniversity,Chikusa-ku,Nagoya 464-8601,JapanReceived 11 May 2011; revised 6 July 2011; accepted 19 July 2011; published online 31 August 2011.*For correspondence (fax +81 52 789 5214; e-mail cueguchi@agr.nagoya-u.ac.jp).SUMMARYInplants,stem cells reside in apical meristems,and providethe descendants required for post-embryonicgrowth anddevelopmentthroughoutthe lifeofaplant.Toidentifyanovelfactorrequiredforthemaintenanceofstemcells,weisolatedanArabidopsismutant,namedmeristemdisorganization1-1(mdo1-1),thatexhibitsseveral developmental defects,such asabnormal phyllotaxy and plastochron,stemfasciation and retardedroot growth.We found that themutant plants fail to maintain stem cells,resulting in the differentiation ordeath of stem cells.The mutant plants also showedseveral phenotypes related to DNA damage,suggestingthat themutant cells areexposed constitutively to DNA damage even without external genotoxic stress.Thegrowthdefectand thehypersensitivitytoDNA-damagingagents of mdo1-1were enhanced significantlywhencombinedwithalesionoftheATAXIA-TELANGIECTASIAMUTATED(ATM)gene,butnotoftheATM/RAD3-RELATED(ATR)gene,suggestingthatthefunction oftheMDO1gene iscloselyrelatedtothat ofATMkinaseTheMDO1geneencodes an unknown protein that is conserved ina widevariety of land plants.Theresultsthus suggestedthatthe MDO1geneproduct isrequiredforthe maintenance of stem cells throughareductionin DNA damage.Keywords: stem cell, apical meristem, DNA damage, cell death, cell differentiation.INTRODUCTIONin the SAM is dynamicallyregulated bytheWUS-CLAVATAIn plants, stem cells reside in apical meristems, and providethedescendantsrequiredforpost-embryonicgrowthand(CLVnegativefeedbackloop(Schoofetal.,2000).Asimilardevelopmentthroughout the lifeof a plant.Plantstemcellsbut distinct mechanism also operates in root tips.Thecan remain in an undifferentiated state by accepting signalsWUSCHEL-RELATEDHOMEOBOX5(WOX5)genefunctionsfrom organizing centers in specific microenvironmentsinQC cells to prevent the initialsfromundergoing differen-tiation,and overexpression of the CLV3-likegene CLE40calledstemcell niches.Intheshootapicalmeristem(SAM)reduces the meristematic activity of the RAM (Hobe et al.,stem cells are maintained in thecentral zone under thecontrol ofWUSCHEL(WUS)-expressing cellslocated just2003;Fiers etal.,2005;Sarkaretal.,2007).In additionbeneath the stem cell population (Mayer et al., 1998). In theseveral transcription factors specifically requiredfor themaintenance of apical meristems have been identified.Therootapicalmeristem(RAM),asetofspecializedstemcellsSHOOTMERISTEMLESS(STM)geneencodingaKnotted1calledinitialssurroundquiescentcenter(QC)cellsthatactasa root-organizing center (Dolan et al.,1993; van den Bergclass homeoboxprotein is necessaryfortheestablishmentetal.,1997).ThehighlyorganizedstructureofapicalmeroftheSAMduringembryogenesis,andforfurthermainte-istems is apparently indispensable for proper growth andnance in post-embryonic development (Barton and Poethig,development, e.g.the production of lateral organs in a1993;Endrizzi et al.,1996; Long etal.,1996).In the RAM,twospatially and temporally regulated manner (phyllotaxy andGRASfamilytranscriptionfactorgenes,SHORTROOT(SHR)plastochron,respectively).and SCARECROW(SCR),areinvolved inmaintainingtheExtensive genetic studies have revealed several factorsstem cell population (Di Laurenzio et al., 1996; Helariuttaet al.,2000; Sabatini et al.,2003).The PLETHORA1(PLT1)specifically involved in the maintenance of stem cells in eachtypeofapical meristem.The sizeofthe stemcellpopulationand PLT2 genes encoding AP2 transcription factors are2011TheAuthors657The Plant Journal 2011Blackwell Publishing Ltd

The Arabidopsis MERISTEM DISORGANIZATION 1 gene is required for the maintenance of stem cells through the reduction of DNA damage Yuma Hashimura and Chiharu Ueguchi* Bioscience and Biotechnology Center, Nagoya University, Chikusa-ku, Nagoya 464-8601, Japan Received 11 May 2011; revised 6 July 2011; accepted 19 July 2011; published online 31 August 2011. *For correspondence (fax +81 52 789 5214; e-mail cueguchi@agr.nagoya-u.ac.jp). SUMMARY In plants, stem cells reside in apical meristems, and provide the descendants required for post-embryonic growth and development throughout the life of a plant. To identify a novel factor required for the maintenance of stem cells, we isolated an Arabidopsis mutant, named meristem disorganization 1-1 (mdo1-1), that exhibits several developmental defects, such as abnormal phyllotaxy and plastochron, stem fasciation and retarded root growth. We found that the mutant plants fail to maintain stem cells, resulting in the differentiation or death of stem cells. The mutant plants also showed several phenotypes related to DNA damage, suggesting that the mutant cells are exposed constitutively to DNA damage even without external genotoxic stress. The growth defect and the hypersensitivity to DNA-damaging agents of mdo1-1 were enhanced significantly when combined with a lesion of the ATAXIA-TELANGIECTASIA MUTATED (ATM) gene, but not of the ATM/RAD3- RELATED (ATR) gene, suggesting that the function of the MDO1 gene is closely related to that of ATM kinase. The MDO1 gene encodes an unknown protein that is conserved in a wide variety of land plants. The results thus suggested that the MDO1 gene product is required for the maintenance of stem cells through a reduction in DNA damage. Keywords: stem cell, apical meristem, DNA damage, cell death, cell differentiation. INTRODUCTION In plants, stem cells reside in apical meristems, and provide the descendants required for post-embryonic growth and development throughout the life of a plant. Plant stem cells can remain in an undifferentiated state by accepting signals from organizing centers in specific microenvironments called stem cell niches. In the shoot apical meristem (SAM), stem cells are maintained in the central zone under the control of WUSCHEL (WUS)-expressing cells located just beneath the stem cell population (Mayer et al., 1998). In the root apical meristem (RAM), a set of specialized stem cells called initials surround quiescent center (QC) cells that act as a root-organizing center (Dolan et al., 1993; van den Berg et al., 1997). The highly organized structure of apical mer￾istems is apparently indispensable for proper growth and development, e.g. the production of lateral organs in a spatially and temporally regulated manner (phyllotaxy and plastochron, respectively). Extensive genetic studies have revealed several factors specifically involved in the maintenance of stem cells in each type of apical meristem. The size of the stem cell population in the SAM is dynamically regulated by the WUS-CLAVATA (CLV) negative feedback loop (Schoof et al., 2000). A similar but distinct mechanism also operates in root tips. The WUSCHEL-RELATED HOMEOBOX 5 (WOX5) gene functions in QC cells to prevent the initials from undergoing differen￾tiation, and overexpression of the CLV3-like gene CLE40 reduces the meristematic activity of the RAM (Hobe et al., 2003; Fiers et al., 2005; Sarkar et al., 2007). In addition, several transcription factors specifically required for the maintenance of apical meristems have been identified. The SHOOT MERISTEMLESS (STM) gene encoding a Knotted 1 class homeobox protein is necessary for the establishment of the SAM during embryogenesis, and for further mainte￾nance in post-embryonic development (Barton and Poethig, 1993; Endrizzi et al., 1996; Long et al., 1996). In the RAM, two GRAS family transcription factor genes, SHORT ROOT (SHR) and SCARECROW (SCR), are involved in maintaining the stem cell population (Di Laurenzio et al., 1996; Helariutta et al., 2000; Sabatini et al., 2003). The PLETHORA 1 (PLT1) and PLT2 genes encoding AP2 transcription factors are ª 2011 The Authors 657 The Plant Journal ª 2011 Blackwell Publishing Ltd The Plant Journal (2011) 68, 657–669 doi: 10.1111/j.1365-313X.2011.04718.x

658Yuma Hashimura and Chiharu Ueguchiredundantly required to both establish and maintain stemtoxictreatments(FulcherandSablowski,2009).Unlikethecell niches in root tips (Aida et al., 2004).The OBERON 1caseinanimals,becauseplantstemcellsproducegerm-line(OBE1)and OBE2genes encoding homeodomain fingercells in the final phase of the life cycle,there is a risk that theproteins are required for the maintenanceof boththe SAMmutations fixed in stem cells during the vegetativegrowthand the RAM (Saiga et al.,2008).phaseareinheritedbytheoffspring.Thus,theprotectionofIt has been reported in Arabidopsis that the structuralstem cells against DNA damage,including exclusion ofdamaged stem cells, is crucial not only for developmentalorganizationandproperfunctioningofapicalmeristemsarealso affected by lesions of the genes with general cellularnormalitybutalsoforreproductivefitness.functions,suchas chromatin organization,DNA replicationHere, we isolated a novel mutant, named meristemand cell cycle regulation.Mutations of the FASCIATA(FAS)disorganization 1-1 (mdo1-1), that exhibits abnormal shootgenes encoding subunits of chromatin assembly factor-1,morphogenesis as well as retarded growth of roots.Wewhich supports nucleosome assembly on replicating DNA,found that the developmental defects observed in mdo1-1result in thedisorganizationofapicalmeristems,leadingtoresult from the failure of the maintenance of stem cells inapical meristems,whichleadstothedisorganizationofthepleiotropicdevelopmentaldefects(LeyserandFurner,1992;Kayaetal.,2001).TheMGOUN1(MGO1)geneencodingameristem structures. In mdo1-1 cells, an elevated level oftype-lBtopoisomerase,which is requiredforthe regulationDSBs and increased expression of DNA damage-inducedof DNA coiling during replication and transcription, wasgenes,the expression of which is regulated by ATM,revealed to be important for the proper functioning of thesuggested that the mutant cells are exposed constitutivelySAMandRAM(Laufsetal.,1998;Takahashiet al.,2002;GraftoDNAdamage,leadingtoDSBs.Themdo1-1phenotypeset al., 2010).Altered expression levels of a key cell-cyclewere enhanced when combined with an atm mutation.Theregulatorgene,RETINOBLASTOMA-RELATED(RBR),affectMDO1gene encodes an unknown proteinthatis conservedthe maintenance of stem cells in apical meristems (Wild-inawidevarietyoflandplants.Theseresultsthus suggestedwater et al.,2005; Borghi et al.,2010).Lesions ofthe TEBICHIthat the MDO1gene product is requiredforthe maintenance(TEB)gene,theproductofwhich isproposedtobe involvedof stem cells through the protection of cells from DNAincell cycleregulation,result inthedisorganizationof apicaldamage.meristems(lnagakietal.,2006,2009).TheMGO3/TONSOKURESULTS(TSK)/BRUSHY1(BRU1)geneproduct,anuclearprotein,isneeded for meristem function as well as other cellularIsolationofthemdo1-1mutantfunctions,such as DNAdamageresponses and epigeneticWe obtained a novel mutant exhibiting aberrant shootgene silencing(Guyomarc'h et al.,2004; Suzuki et al.,2004;morphogenesisfromapool of EMS-mutagenizedM,seed-Takeda et al., 2004).Alithough the underlying mechanismlings ofArabidopsisthaliana,anddesignateditasmdo1-1remains unclear, the results suggest the requirement ofbased on themeristem phenotype(seebelow).AboutproperDNAmetabolism for themaintenance of stem cellthree-quarters and one-quarter of the self-progeny of theniches.heterozygousmdo1-1/+plants,which showed noapparentLivingorganisms are atriskfor exposureto endogenousdefective phenotype, exhibited the wild-type and mdo1-1and environmental hazards causing DNA damage and,phenotypes, respectively (576 wild type and 189 mutantconsequently,mutations.Fixedmutations,ifcausedinstemplants; =0.04, P>0.05), suggesting that mdo1-1 is acells, would be inherited after cell division, and not onlysingleand recessive Mendeliantrait.prevent cells from functioning normally but also causeserious problems,suchas cancer.ToavoidsuchdeleteriousShoot phenotype of mdo1-1situations,inanimals,stemcellsshowalowtolerancetoThe mdo1-1 mutation affected shoot development. In theDNA damage, and have a tendency for differentiation orseedling stage,the emergence ofthefirstpairofleaves (firstprogrammed cell death (apoptosis) in response to DNAand second leaves)orthe second leaf was occasionallydamage (Rich et al.,2000; Sherman et al.,2011).Cellulardelayed (Figure 1b,c). In some cases,three leaves (second,responses to DNAdamage are mediated bytwoproteinthird and fourth leaves) were generated nearly simulta-kinases:ATAXIA-TELANGIECTASIA MUTATED(ATM)andneously (Figure 1e,arrowheads). In the subsequent vege-ATM/RAD3-RELATED (ATR).ATM and ATR are activated bytative growth phase,leaves continued to emergein aDNA double-strand breaks (DSBs) and single-stranded DNA,respectively,and in turn positively regulate downstreamspatiallyand temporally unregulated manner (Figure1g).Themutant leaves showeda slightlyabnormal shapeandareactions,suchasthetranscriptionalupregulation ofseveralDNA repair genes (Shiloh, 2006; Flynn and Zou, 2010).reduced size, resulting in a compact rosette structure. In thereproductivegrowthphase,themutant inflorescences wereLikewise,itwasdemonstrated recentlythatplantstem cellsand their early descendants selectively undergo ATM/ATR-frequentlyfasciated,andsomeofthemceasedtogrow,withnew inflorescences developing from the arrested shootmediated programmed cell death upon certain mild geno-2011TheAuthorsThePlantJournal2011BlackwellPublishingLtd,ThePlantJournal,(2011),68,657-669

redundantly required to both establish and maintain stem cell niches in root tips (Aida et al., 2004). The OBERON 1 (OBE1) and OBE2 genes encoding homeodomain finger proteins are required for the maintenance of both the SAM and the RAM (Saiga et al., 2008). It has been reported in Arabidopsis that the structural organization and proper functioning of apical meristems are also affected by lesions of the genes with general cellular functions, such as chromatin organization, DNA replication and cell cycle regulation. Mutations of the FASCIATA (FAS) genes encoding subunits of chromatin assembly factor-1, which supports nucleosome assembly on replicating DNA, result in the disorganization of apical meristems, leading to pleiotropic developmental defects (Leyser and Furner, 1992; Kaya et al., 2001). The MGOUN1 (MGO1) gene encoding a type-IB topoisomerase, which is required for the regulation of DNA coiling during replication and transcription, was revealed to be important for the proper functioning of the SAM and RAM (Laufs et al., 1998; Takahashi et al., 2002; Graf et al., 2010). Altered expression levels of a key cell-cycle regulator gene, RETINOBLASTOMA-RELATED (RBR), affect the maintenance of stem cells in apical meristems (Wild￾water et al., 2005; Borghi et al., 2010). Lesions of the TEBICHI (TEB) gene, the product of which is proposed to be involved in cell cycle regulation, result in the disorganization of apical meristems (Inagaki et al., 2006, 2009). The MGO3/TONSOKU (TSK)/BRUSHY 1 (BRU1) gene product, a nuclear protein, is needed for meristem function as well as other cellular functions, such as DNA damage responses and epigenetic gene silencing (Guyomarc’h et al., 2004; Suzuki et al., 2004; Takeda et al., 2004). Although the underlying mechanism remains unclear, the results suggest the requirement of proper DNA metabolism for the maintenance of stem cell niches. Living organisms are at risk for exposure to endogenous and environmental hazards causing DNA damage and, consequently, mutations. Fixed mutations, if caused in stem cells, would be inherited after cell division, and not only prevent cells from functioning normally but also cause serious problems, such as cancer. To avoid such deleterious situations, in animals, stem cells show a low tolerance to DNA damage, and have a tendency for differentiation or programmed cell death (apoptosis) in response to DNA damage (Rich et al., 2000; Sherman et al., 2011). Cellular responses to DNA damage are mediated by two protein kinases: ATAXIA-TELANGIECTASIA MUTATED (ATM) and ATM/RAD3-RELATED (ATR). ATM and ATR are activated by DNA double-strand breaks (DSBs) and single-stranded DNA, respectively, and in turn positively regulate downstream reactions, such as the transcriptional upregulation of several DNA repair genes (Shiloh, 2006; Flynn and Zou, 2010). Likewise, it was demonstrated recently that plant stem cells and their early descendants selectively undergo ATM/ATR￾mediated programmed cell death upon certain mild geno￾toxic treatments (Fulcher and Sablowski, 2009). Unlike the case in animals, because plant stem cells produce germ-line cells in the final phase of the life cycle, there is a risk that the mutations fixed in stem cells during the vegetative growth phase are inherited by the offspring. Thus, the protection of stem cells against DNA damage, including exclusion of damaged stem cells, is crucial not only for developmental normality but also for reproductive fitness. Here, we isolated a novel mutant, named meristem disorganization 1-1 (mdo1-1), that exhibits abnormal shoot morphogenesis as well as retarded growth of roots. We found that the developmental defects observed in mdo1-1 result from the failure of the maintenance of stem cells in apical meristems, which leads to the disorganization of the meristem structures. In mdo1-1 cells, an elevated level of DSBs and increased expression of DNA damage-induced genes, the expression of which is regulated by ATM, suggested that the mutant cells are exposed constitutively to DNA damage, leading to DSBs. The mdo1-1 phenotypes were enhanced when combined with an atm mutation. The MDO1 gene encodes an unknown protein that is conserved in a wide variety of land plants. These results thus suggested that the MDO1 gene product is required for the maintenance of stem cells through the protection of cells from DNA damage. RESULTS Isolation of the mdo1-1 mutant We obtained a novel mutant exhibiting aberrant shoot morphogenesis from a pool of EMS-mutagenized M2 seed￾lings of Arabidopsis thaliana, and designated it as mdo1-1 based on the meristem phenotype (see below). About three-quarters and one-quarter of the self-progeny of the heterozygous mdo1-1/+ plants, which showed no apparent defective phenotype, exhibited the wild-type and mdo1-1 phenotypes, respectively (576 wild type and 189 mutant plants; v2 = 0.04, P > 0.05), suggesting that mdo1-1 is a single and recessive Mendelian trait. Shoot phenotype of mdo1-1 The mdo1-1 mutation affected shoot development. In the seedling stage, the emergence of the first pair of leaves (first and second leaves) or the second leaf was occasionally delayed (Figure 1b,c). In some cases, three leaves (second, third and fourth leaves) were generated nearly simulta￾neously (Figure 1e, arrowheads). In the subsequent vege￾tative growth phase, leaves continued to emerge in a spatially and temporally unregulated manner (Figure 1g). The mutant leaves showed a slightly abnormal shape and a reduced size, resulting in a compact rosette structure. In the reproductive growth phase, the mutant inflorescences were frequently fasciated, and some of them ceased to grow, with new inflorescences developing from the arrested shoot 658 Yuma Hashimura and Chiharu Ueguchi ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 68, 657–669

DNA damage and stem cell maintenance 659inflorescence meristems in the early reproductive growthphase were used for the following analysis. Scanning electron microscopic analysis indicated thatthe wild-type SAMshowed adome-shapedstructurewithasmoothsurface(Figure 2a).We did not observe sucha regular structure in themdo1-1 shoot apex: the apical region of mdo1-1, a region米米米surrounded byfloral buds,was expandedand coveredwithalot of small convex structures (Figure2b).Longitudinal sections revealedthatthe mdo1-1shoot apex wasflat,enlargedlaterally and was not covered by a canonical surface layerstructure,i.e.theL1and L2 layers,as is usually observed inthe wild-type SAM (Figure 2c,d). In the mutant apical region,especiallyinthesurfacearea,weobservedalotofenlargedcells,suggesting thatthecellshad startedto differentiate.Toclarify this further, we carried out in situ hybridization anal-ysis ofthe STMand AINTEGUMENTA(ANT)genes,thetranscripts of whichcan mark undifferentiated and differen-tiating cells, respectively (Elliott et al., 1996; Long et al.,1996).STMexpression wasdetectedinthecentral regionofthe wild-type SAM and downregulated in the incipient floralbuds (Figure2e),ashasbeen reportedpreviously (Long etal.,1996).Inthemdo1-1shootapex,STMexpressionwasreducedinthesurfacecells(Figure2f),or,insomesamplesFigure1.Shoot phenotypes of the mdo1-1mutant.(a-c)Morphology ofmultipleSTM expression domainscould be observed(Fig7-day-old seedlingsPlants were grown on MS gellan gum plates with 16-h light/8-h darkure 2g).Although ANTexpression was downregulated in thefluorescent illumination at 22C.Col (a) and mdo1-1 mutants showingwild-type central region (Figure 2h),it was detected in a widedelayed emergence of the first pair of leaves (b) and the second leaf (c) arerange of cells in the mdo1-1 shoot apex (Figure 2i). Theseshown-Scalehars:5mn(d, e)Col (d) and mdo1-1(e)seedlings (10 days old) are shown.Thearrowandresults thus suggest that the structural organization of thearrowheads indicate the first leaf, and second,third and fourth leavesSAM in mdo1-1isdisrupted bythe failure of maintaining therespectively.Scalebars:5mm(f-h) Morphology of 26-day-old rosettes. Plants were grown on MS gellanundifferentiated stateof stem cells.gum plates for 10 days and then on rockfiber with 16-h light/8-h darkRAMphenotypeofmdo1-1fluorescent illumination at22°C.Inflorescencestems wereremoved.Col (f)mdo1-1 (g) and the mdo1-1 carrying genomic MDO1 allele (h) are shown.Themdo1-1 mutation also affected root growth.The rootScalebars: 1 cm.(i-l) Inflorescence phenotypes of 35-day-old plants. Plants were grown on MSelongation rate was substantially decreased in the mdo1-1gellan gum plates for 10 days and then on rock fiber with 16-h light/8-h darkmutant (Figure 3a), suggesting a functional defect of thefluorescent illuminationat22°C.Col (i),mdo1-1(j,.k)and mdo1-1carrying thegenomic MDO1allele (l) are shown.ThearrowmutantRAM.To examinethisfurther,weinvestigatedthein panel (j) indicates an inflorescence showing the stop-and-go phenotypeexpressionofmarkergenesspecifyingseveraltypesofstemScale bars: 1 cm.cellnichecells.QC25expression wasobservedinQCcells inthewildtype(Figure3b;Sabatinietal.,2003),whereasQC25expressionwasreducedinsomeQCcellsorwastotallyapices (the stop-and-go phenotype; Figure 1j,arrow).Theabsent in mdo1-1 (Figure 3c,d). This suggested that thephyllotaxy of the floral buds was also distorted (Figure 1k)maintenanceofQCfunction is,atleastpartially,impaired byTheabnormaldevelopmentalpatternofthelateral organsasthe mdo1-1mutation.Inthe wild type, columella initialswell as the frequently observed stem fasciation in the mdo1-1locateddirectlybelowtheQCweremaintainedinanundifplants strongly suggested that the mutant SAM is impairedferentiatedstate sothattheydidnotaccumulate starchstructurallyand/orfunctionally.Thus,wefocused uponthegranules that were visualized by Lugol staining (Willemsenmutational effect on apical meristemfunctioning.et al.,1998;Figure3b).In contrast,cells justbeneaththeQCcells accumulated starch granules in mdo1-1 (Figure 3c,d),SAMphenotypeofmdo1-1indicating a failure in themaintenanceof the undifferentiTo determine whether or not the structure of the SAM is af-ated state of columella initials,presumably because ofafected by the mdo1-1mutation, we inspected the mutantmalfunction of the QC cells and/orcolumella initialsperse.shoot apiceshistologically.Because severe phenotypes,WeobtainedessentiallythesameresultswhenQC46andstem fasciation and growth arrest, in addition to abnormalQC184(Sabatini etal.,2003),other QC markers,were usedphyllotaxy,wereseenduringthereproductivegrowthphase(Figure S1).2011TheAuthorsThePlantJournal2011Blackwell PublishingLtd,ThePlantJournal,(2011),68,657-669

apices (the stop-and-go phenotype; Figure 1j, arrow). The phyllotaxy of the floral buds was also distorted (Figure 1k). The abnormal developmental pattern of the lateral organs as well as the frequently observed stem fasciation in the mdo1-1 plants strongly suggested that the mutant SAM is impaired structurally and/or functionally. Thus, we focused upon the mutational effect on apical meristem functioning. SAM phenotype of mdo1-1 To determine whether or not the structure of the SAM is af￾fected by the mdo1-1 mutation, we inspected the mutant shoot apices histologically. Because severe phenotypes, stem fasciation and growth arrest, in addition to abnormal phyllotaxy, were seen during the reproductive growth phase, inflorescence meristems in the early reproductive growth phase were used for the following analysis. Scanning elec￾tron microscopic analysis indicated that the wild-type SAM showed a dome-shaped structure with a smooth surface (Figure 2a).We did not observe such a regular structure in the mdo1-1 shoot apex: the apical region of mdo1-1, a region surrounded by floral buds, was expanded and covered with a lot of small convex structures (Figure 2b). Longitudinal sec￾tions revealed that the mdo1-1 shoot apex was flat, enlarged laterally and was not covered by a canonical surface layer structure, i.e. the L1 and L2 layers, as is usually observed in the wild-type SAM (Figure 2c,d). In the mutant apical region, especially in the surface area, we observed a lot of enlarged cells, suggesting that the cells had started to differentiate. To clarify this further, we carried out in situ hybridization anal￾ysis of the STM and AINTEGUMENTA (ANT) genes, the transcripts of which can mark undifferentiated and differen￾tiating cells, respectively (Elliott et al., 1996; Long et al., 1996). STM expression was detected in the central region of the wild-type SAM and downregulated in the incipient floral buds (Figure 2e), as has been reported previously (Long et al., 1996). In the mdo1-1 shoot apex, STM expression was reduced in the surface cells (Figure 2f), or, in some samples, multiple STM expression domains could be observed (Fig￾ure 2g). Although ANT expression was downregulated in the wild-type central region (Figure 2h), it was detected in a wide range of cells in the mdo1-1 shoot apex (Figure 2i). These results thus suggest that the structural organization of the SAM in mdo1-1 is disrupted by the failure of maintaining the undifferentiated state of stem cells. RAM phenotype of mdo1-1 The mdo1-1 mutation also affected root growth. The root elongation rate was substantially decreased in the mdo1-1 mutant (Figure 3a), suggesting a functional defect of the mutant RAM. To examine this further, we investigated the expression of marker genes specifying several types of stem cell niche cells. QC25 expression was observed in QC cells in the wild type (Figure 3b; Sabatini et al., 2003), whereas QC25 expression was reduced in some QC cells or was totally absent in mdo1-1 (Figure 3c,d). This suggested that the maintenance of QC function is, at least partially, impaired by the mdo1-1 mutation. In the wild type, columella initials located directly below the QC were maintained in an undif￾ferentiated state so that they did not accumulate starch granules that were visualized by Lugol staining (Willemsen et al., 1998; Figure 3b). In contrast, cells just beneath the QC cells accumulated starch granules in mdo1-1 (Figure 3c,d), indicating a failure in the maintenance of the undifferenti￾ated state of columella initials, presumably because of a malfunction of the QC cells and/or columella initials per se. We obtained essentially the same results when QC46 and QC184 (Sabatini et al., 2003), other QC markers, were used (Figure S1). (a) (f) (i) (j) (k) (l) (g) (h) (b) (c) (d) (e) Figure 1. Shoot phenotypes of the mdo1-1 mutant. (a–c) Morphology of 7-day-old seedlings. Plants were grown on MS gellan gum plates with 16-h light/8-h dark fluorescent illumination at 22C. Col (a) and mdo1-1 mutants showing delayed emergence of the first pair of leaves (b) and the second leaf (c) are shown. Scale bars: 5 mm. (d, e) Col (d) and mdo1-1 (e) seedlings (10 days old) are shown. The arrow and arrowheads indicate the first leaf, and second, third and fourth leaves, respectively. Scale bars: 5 mm. (f–h) Morphology of 26-day-old rosettes. Plants were grown on MS gellan gum plates for 10 days and then on rockfiber with 16-h light/8-h dark fluorescent illumination at 22C. Inflorescence stems were removed. Col (f), mdo1-1 (g) and the mdo1-1 carrying genomic MDO1 allele (h) are shown. Scale bars: 1 cm. (i–l) Inflorescence phenotypes of 35-day-old plants. Plants were grown on MS gellan gum plates for 10 days and then on rock fiber with 16-h light/8-h dark fluorescent illumination at 22C. Col (i), mdo1-1 (j, k) and mdo1-1 carrying the genomic MDO1 allele (l) are shown. The arrow in panel (j) indicates an inflorescence showing the stop-and-go phenotype. Scale bars: 1 cm. DNA damage and stem cell maintenance 659 ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 68, 657–669

660 Yuma Hashimura andChiharu Ueguchi(a)(a)cF(c)(d)OL910732678Daysaftergermination(b)Cd(hFigure3.Rootphenotypes of themdo1-1mutant(a)Thekinetics ofrootgrowth.Plantsweregerminated and grown onthesurface of MS gellan gum plates with 16-h light/8-h dark fluorescentillumination at 22°C. The length of main roots was determined every day.Each value represents an average, with standard error, for 20 plants. Thestrains used were: Col (red circles); mdo1-1 (blue triangles); and mdo1-1carrying genomic MDO1 allele (green squares).Figure 2. Shoot apical meristem (SAM) phenotypes of the mdo1-1 mutants.(bd)ExpressionofQC25and accumulationof starchgranules inroottips(a, b) SEM images of 28-day-old Col (a) and mdo1-1 (b) shoot apices. ScaleBlue and purple represent QC25-expressing cells and accumulated starchbars:100um.granules, respectively. QC25 (b) and QC25 mdo1-1(c, d) Longitudinal plastic sections of 28-day-old Col (c) and mdo1-1 (d) shoot(c,d) seedlings(7-daysold) were subjected to GUS stainingfollowed by Lugolapices.Scale bars: 50 μm.staining. The arrowheads indicate the position of columella initials. Scale(e-g) Expression of the STM gene in 14-day-old Col (e) and mdo1-1bars: 50 μm.(f,g)shootapices.AccumulationofSTMmRNAwasanalyzed byin situ(e-g) Expression of SCRpro:GFP-TIP. Col (e) and mdo1-1hybridization.The arrowhead in panel (e) indicates an incipient leaf primor-(f, g) seedlings (7-days old) were counterstained with propidium iodide (PI).dium. Scale bars: 50 μm.Thearrowhead inpanel (f) indicates acellfile showingreduced expression of(h, i) Expression of the ANT gene in 28-day-old Col (h) and mdo1-1 (i) shootSCRpro:GFP-TIP.The arrowhead and arrow in panel g) indicate reduced andapices. The accumulation of ANT mRNA was analyzed by in situ hybridization.discontinuous expression,and ectopic expression of SCRpro:GFP-TiP,Scale bars: 50 μm.respectively. Scale bars: 100 μm.AsdemonstrateduponexpressionofaSCARECROW(Figure 3f,g). It should be noted that the emergence of(SCR)promoter-GFP fusion gene(SCR:GFP-TIP;Saitoectopicexpressionof SCR:GFP-TIPinthemutantsteleswasetal.,2005),SCRisexpressed inQCcells,cortical/endoder-accompanied by the disappearance of expression of themal initials and endodermis cell layers in the wild typefusion gene in the endodermal cell files (Figure 3g).This(Figure 3e). In the mdo1-1 root tips, the SCR:GFP-,TIPfinding suggested that the function of cortical/endodermalexpression was significantlyreduced in someendodermisinitialsand/orthecell fateofendodermalcellfilesarenotcell layers, or was abolished in a discontinuous mannermaintained properly in the mdo1-1roots.2011TheAuthorsThePlantJournal2011BlackwellPublishingLtd,ThePlantJournal,(2011),68,657-669

As demonstrated upon expression of a SCARECROW (SCR) promoter-GFP fusion gene (SCR:GFP-cTIP; Saito et al., 2005), SCR is expressed in QC cells, cortical/endoder￾mal initials and endodermis cell layers in the wild type (Figure 3e). In the mdo1-1 root tips, the SCR:GFP-cTIP expression was significantly reduced in some endodermis cell layers, or was abolished in a discontinuous manner (Figure 3f,g). It should be noted that the emergence of ectopic expression of SCR:GFP-cTIP in the mutant steles was accompanied by the disappearance of expression of the fusion gene in the endodermal cell files (Figure 3g). This finding suggested that the function of cortical/endodermal initials and/or the cell fate of endodermal cell files are not maintained properly in the mdo1-1 roots. 2 3 4 5 6 0 1 1 2 3 4 5 6 7 8 9 10 Days after germination Root length (cm) (a) (b) (c) (d) (e) (f) (g) Figure 3. Root phenotypes of the mdo1-1 mutant. (a) The kinetics of root growth. Plants were germinated and grown on the surface of MS gellan gum plates with 16-h light/8-h dark fluorescent illumination at 22C. The length of main roots was determined every day. Each value represents an average, with standard error, for 20 plants. The strains used were: Col (red circles); mdo1-1 (blue triangles); and mdo1-1 carrying genomic MDO1 allele (green squares). (b–d) Expression of QC25 and accumulation of starch granules in root tips. Blue and purple represent QC25-expressing cells and accumulated starch granules, respectively. QC25 (b) and QC25 mdo1-1 (c, d) seedlings (7-days old) were subjected to GUS staining followed by Lugol staining. The arrowheads indicate the position of columella initials. Scale bars: 50 lm. (e–g) Expression of SCRpro:GFP-cTIP. Col (e) and mdo1-1 (f, g) seedlings (7-days old) were counterstained with propidium iodide (PI). The arrowhead in panel (f) indicates a cell file showing reduced expression of SCRpro:GFP-cTIP. The arrowhead and arrow in panel (g) indicate reduced and discontinuous expression, and ectopic expression of SCRpro:GFP-cTIP, respectively. Scale bars: 100 lm. (a) (b) (c) (d) (e) (f) (g) (h) (i) Figure 2. Shoot apical meristem (SAM) phenotypes of the mdo1-1 mutants. (a, b) SEM images of 28-day-old Col (a) and mdo1-1 (b) shoot apices. Scale bars: 100 lm. (c, d) Longitudinal plastic sections of 28-day-old Col (c) and mdo1-1 (d) shoot apices. Scale bars: 50 lm. (e–g) Expression of the STM gene in 14-day-old Col (e) and mdo1-1 (f, g) shoot apices. Accumulation of STM mRNA was analyzed by in situ hybridization. The arrowhead in panel (e) indicates an incipient leaf primor￾dium. Scale bars: 50 lm. (h, i) Expression of the ANT gene in 28-day-old Col (h) and mdo1-1 (i) shoot apices. The accumulation of ANT mRNA was analyzed by in situ hybridization. Scale bars: 50 lm. 660 Yuma Hashimura and Chiharu Ueguchi ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 68, 657–669

DNAdamageandstemcell maintenance661BILITY1(ATBRCA1)genes.The products of these genesCell death observed in the mdo1-1stem cell nicheshave been demonstrated to be involved in DNA repair, andAs shown byFigure 3(f.g),we observed that some cells weretheir expression is induced by DSBs (Klimyuk and Jones,stained with propidium iodide (Pl). PI stains the cell walls of1997;Deveaux etal.,2000; Lafarge and Montane,2003)living cells but is also usedto detectdead cellsthat havelostTotal RNAwaspreparedfromshootapextissuesandthentheir membrane integrity (Truernit and Haseloff, 2008)subjected to quantitative real-time RT-PCR(qRT-PCR)anal-WhenroottipswerestainedwithPlalone,cells instemcellysis. It was revealed that, in mdo1-1, the transcript levels ofniches,QC cells,several initials and the early descendantsthesegenes wereincreased2-2.5-foldcomparedwith thosein the wild type (Figure 5a).Furthermore, we checked thewere stained preferentially in mdo1-1, but not in the wildtype (Figure 4a,b). To confirm cell death in the mdo1-1expressionofCYCB1;1becauseits expression isalsoknownmutant, root tips were stained with Sytox Orange,anotherto be upregulated by irradiation, a typical treatmentcell death marker (Truernit and Haseloff,2008).As shown inresulting in DSBs (Culligan et al., 2006). Plants carrying aFigure 4(d), we observed a similar staining pattern in theCYCB1;1:GUS fusion containinga mitotic destruction boxmdo1-1 roots:that is, cells in the stem cell niches were(Colon-Carmona etal.,1999)were subjected to GUS staining.TheGUSactivitywas stronglyenhanced bythemdo1-1preferentially stained with Sytox Orange.mutation in both shoot apices and root tips (Figure 5c,e).DNA damage in the mdo1-1 mutantThis enhancement neither results from enhanced cell divi-sion activity norfrom cellcycle arrest at the G2/M boundaryIt was reported recently that the cells in plant stem cellnichesarehypersensitivetoDNAdamage,leadingtoDSBsbecause,comparedwiththewild-typebackground,expres-sion of CYCB1;2:GUS containing a destruction box, the(Fulcherand Sablowski,2009).Themdo1-1rootphenotypedescribed above led us to theidea that the mdo1-1mutationexpression of which is induced by the duration of the G2/Mcauses DNA damage in the cells.To examine this idea, wetransition,but not by DNA damage responses (Culliganfirstexaminedtheexpressionof theRAD51,GAMMAetal.,2006),wasnotdrasticallyenhanced in themutantRESPONSE1(ATGR1)andBREASTCANCERSUSCEPTIshootapex,and was instead slightly reduced in themutantroot tips (Figure 5g,i). These results thus strongly suggestthat the mdo1-1 cells are exposed to DNA damage consti-tutively,even without externalgenotoxic stress.(a)(b)It is known thatATMactivatedin responseto DSBsmediates the transcriptional activation of several genes,including RAD51,ATGR1,ATBRCA1and CYCB1;1(Garciaetal.,2003;Culligan etal.,2006).Therefore, it would beexpected that the level of DNA damage including DSBs isenhanced in mdo1-1 cells. To examine this directly, DSBswereanalyzedusingtheterminaldeoxynucleotidetransfer-ase-mediated dUTP nick-end labeling (TUNEL)method(Gavrieli et al., 1992), which can detect 3'-OH break ends.Labeled samples were further stained with 4',6-diamidino-2-phenylindole (DAPl) to detect nuclei. In the wild-type shootapex,apparent TUNELstaining was not detected (Figure5j).(cdIncontrast,weobserved a lotof TUNEL-stained nuclei inawide range of cells in the mutant apex (Figure 5k). Essen-tiallythe same results were obtained forthemutant roottips(Figure 5m). It should be noted that the TUNEL-stainednuclei weredetectednot onlyin stemcellnichesbut also indifferentiating cells, such as cells in growing leaves and inthe differentiation zone of the root tip (Figure 5k,m).Theseresults strongly suggest that a lesion of the MDO1functionresultsinincreasedlevelsofDSBs.Wethen examined whether or not mdo1-1plants aresensitive toDNA-damaging agents.Seedlings grown in MSmediumfor8dayswerefurthergrownfor2weeksinthesame medium containing various concentrations of bleo-Figure 4.Dead cells detected in mdo1-1root tipsmycin,a chemical agent leading to DSBs.As shown inCol (a, c) and mdo1-1 (b, d) seedlings (7-days old) were stained withpropidium iodide (a, b) and Sytox Orange (c, d). Scale bars: 100 μm.Figure6,thegrowthofthemdo1-1plants wasaffectedmore2011TheAuthorsThePlantJournal2011Blackwell PublishingLtd,ThePlantJournal,(2011),68,657-669

Cell death observed in the mdo1-1 stem cell niches As shown by Figure 3(f,g), we observed that some cells were stained with propidium iodide (PI). PI stains the cell walls of living cells but is also used to detect dead cells that have lost their membrane integrity (Truernit and Haseloff, 2008). When root tips were stained with PI alone, cells in stem cell niches, QC cells, several initials and the early descendants were stained preferentially in mdo1-1, but not in the wild type (Figure 4a,b). To confirm cell death in the mdo1-1 mutant, root tips were stained with Sytox Orange, another cell death marker (Truernit and Haseloff, 2008). As shown in Figure 4(d), we observed a similar staining pattern in the mdo1-1 roots: that is, cells in the stem cell niches were preferentially stained with Sytox Orange. DNA damage in the mdo1-1 mutant It was reported recently that the cells in plant stem cell niches are hypersensitive to DNA damage, leading to DSBs (Fulcher and Sablowski, 2009). The mdo1-1 root phenotype described above led us to the idea that the mdo1-1 mutation causes DNA damage in the cells. To examine this idea, we first examined the expression of the RAD51, GAMMA RESPONSE 1 (ATGR1) and BREAST CANCER SUSCEPTI￾BILITY 1 (ATBRCA1) genes. The products of these genes have been demonstrated to be involved in DNA repair, and their expression is induced by DSBs (Klimyuk and Jones, 1997; Deveaux et al., 2000; Lafarge and Montane´, 2003). Total RNA was prepared from shoot apex tissues and then subjected to quantitative real-time RT-PCR (qRT-PCR) anal￾ysis. It was revealed that, in mdo1-1, the transcript levels of these genes were increased 2–2.5-fold compared with those in the wild type (Figure 5a). Furthermore, we checked the expression of CYCB1;1 because its expression is also known to be upregulated by c irradiation, a typical treatment resulting in DSBs (Culligan et al., 2006). Plants carrying a CYCB1;1:GUS fusion containing a mitotic destruction box (Colo´ n-Carmona et al., 1999) were subjected to GUS stain￾ing. The GUS activity was strongly enhanced by the mdo1-1 mutation in both shoot apices and root tips (Figure 5c,e). This enhancement neither results from enhanced cell divi￾sion activity nor from cell cycle arrest at the G2/M boundary because, compared with the wild-type background, expres￾sion of CYCB1;2:GUS containing a destruction box, the expression of which is induced by the duration of the G2/M transition, but not by DNA damage responses (Culligan et al., 2006), was not drastically enhanced in the mutant shoot apex, and was instead slightly reduced in the mutant root tips (Figure 5g,i). These results thus strongly suggest that the mdo1-1 cells are exposed to DNA damage consti￾tutively, even without external genotoxic stress. It is known that ATM activated in response to DSBs mediates the transcriptional activation of several genes, including RAD51, ATGR1, ATBRCA1 and CYCB1;1 (Garcia et al., 2003; Culligan et al., 2006). Therefore, it would be expected that the level of DNA damage including DSBs is enhanced in mdo1-1 cells. To examine this directly, DSBs were analyzed using the terminal deoxynucleotide transfer￾ase-mediated dUTP nick-end labeling (TUNEL) method (Gavrieli et al., 1992), which can detect 3¢-OH break ends. Labeled samples were further stained with 4¢,6-diamidino- 2-phenylindole (DAPI) to detect nuclei. In the wild-type shoot apex, apparent TUNEL staining was not detected (Figure 5j). In contrast, we observed a lot of TUNEL-stained nuclei in a wide range of cells in the mutant apex (Figure 5k). Essen￾tially the same results were obtained for the mutant root tips (Figure 5m). It should be noted that the TUNEL-stained nuclei were detected not only in stem cell niches but also in differentiating cells, such as cells in growing leaves and in the differentiation zone of the root tip (Figure 5k,m). These results strongly suggest that a lesion of the MDO1 function results in increased levels of DSBs. We then examined whether or not mdo1-1 plants are sensitive to DNA-damaging agents. Seedlings grown in MS medium for 8 days were further grown for 2 weeks in the same medium containing various concentrations of bleo￾mycin, a chemical agent leading to DSBs. As shown in Figure 6, the growth of the mdo1-1 plants was affected more (a) (b) (c) (d) Figure 4. Dead cells detected in mdo1-1 root tips. Col (a, c) and mdo1-1 (b, d) seedlings (7-days old) were stained with propidium iodide (a, b) and Sytox Orange (c, d). Scale bars: 100 lm. DNA damage and stem cell maintenance 661 ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 68, 657–669

662 Yuma Hashimuraand ChiharuUeguchi(b)O(a)43.5M(d)(h(e)1.50.5RAD51/TUB2ATGR1/TUB2ATBRCA1/TUB2Figure5.ElevatedlevelofDNAdamageinthemdo1-1mutant(a) Expression of DNA damage-induciblegenes.Total RNA was prepared from tissues including shoot apices of 14-day-old seedlings and then subjected toreal-timeRT-PCR analysis.The level of TUB2mRNA was used as an internal control, and the values in mdo1-1(blue bars) are expressed as ratios to the values in Col(red bars)Values represent averages of three biological replicates with standard deviation.(b-e) Expression of CYCB1;1:GUS.Seedlings of Col (b, d) and mdo1-1 (c,e) were subjected to GUS staining.Longitudinal sections of shoot apices (b,c; 14-day-oldseedlings) and whole-mount images of root tips (d, e; 7-day-old seedlings) are shown. Scale bars: 50 μm for (b, c); 100 μm for (d, e).(f-i) Expression of CYCB1;2:GUS. Seedlings of Col (f, h) and mdo1-1 (g,i) were subjected to GUS staining. Longitudinal sections of shoot apices (f, g; 14-day-oldseedlings)andwhole-mount images of roottips (h,i;7-day-oldseedlings)are shown.Scalebars:50μmfor (f,g):100μmfor(h,i)(i-m) Detection ofDSBs in themdo1-1mutant byTUNELassay.Col (, I and mdo1-1 (k,m)seedlings (14-days old) were subjected to theTUNELassay. Longitudinalsections of shootapices (j,k) and root tips (l, m)are shown.Nomarski images (left),TUNEL-stained images (middle) and DAPl-stained nuclei (right) are shown.Scalebars: 50 μm for (j, k); 100 μm for (l, m).severely than that of the wild-type ones. We obtained similartrast, in combination with the mdo1-1allele, the majority ofresults when mitomycin C (MMC), a DNA cross-linkingthe mdo1-1atm-4 double mutants (16 outof 22plants)agent leading toDSBs,was used (FigureS2).These resultsexhibited a severe growth defect. In addition to the abnor-thus indicated that mdo1-1 plants are hypersensitive tomal phyllotaxy observed in the mdo1-1 single mutant, itsDNA-damagingagents.rosette structures were markedly smaller than those of theparentalsinglemutants,mdo1-1andatm-4(Figure7a).AfterGenetic relationship betweenMDO1andATMprolonged cultivation,there was no furthergrowth and theTheATMandATRproteinkinasesfunctionas key regulatorsplants died.The rest of the mdo1-1atm-4double mutantsin DNA damage responses (Shiloh, 2006; Flynn and Zou,(six out of 22 plants)grew like the mdo1-1 single mutant2010).Toelucidatethe linkbetweenDNAdamageresponsesduring the vegetative growth phase, and then developedand themeristem phenotypeof mdo1-1, we constructedinflorescencestemsthatalsoshowedabnormalfloralbudphyllotaxy fromaxial buds,but notfrom primary meristems,mdo1-1atmandmdo1-1atrdoublemutantsandanalyzedtheir growth phenotypes.Themutations,atm-4and atr-4,which hadpresumablyceasedtogrow(FigureS3a).Even inareT-DNA insertional allelesand arefrom a Col backgroundthis double mutant population, fruit development was(lnagaki et al.,2009).Theatm-4 singlemutant showed nodrastically impaired (Figure S3b).Todetermine whether orapparent growth defect in both the vegetative and repro-notthe rootphenotype of mdo1-1isenhancedby atm-4,ductive growth phases (Figure 7a and Figure S3a). In con-7-day-old seedlings were stained with Pl. Although no2011TheAuthorsThePlantJournal2011BlackwellPublishingLtd,ThePlantJournal,(2011),68,657-669

severely than that of the wild-type ones. We obtained similar results when mitomycin C (MMC), a DNA cross-linking agent leading to DSBs, was used (Figure S2). These results thus indicated that mdo1-1 plants are hypersensitive to DNA-damaging agents. Genetic relationship between MDO1 and ATM The ATM and ATR protein kinases function as key regulators in DNA damage responses (Shiloh, 2006; Flynn and Zou, 2010). To elucidate the link between DNA damage responses and the meristem phenotype of mdo1-1, we constructed mdo1-1 atm and mdo1-1 atr double mutants and analyzed their growth phenotypes. The mutations, atm-4 and atr-4, are T-DNA insertional alleles and are from a Col background (Inagaki et al., 2009). The atm-4 single mutant showed no apparent growth defect in both the vegetative and repro￾ductive growth phases (Figure 7a and Figure S3a). In con￾trast, in combination with the mdo1-1 allele, the majority of the mdo1-1 atm-4 double mutants (16 out of 22 plants) exhibited a severe growth defect. In addition to the abnor￾mal phyllotaxy observed in the mdo1-1 single mutant, its rosette structures were markedly smaller than those of the parental single mutants, mdo1-1 and atm-4 (Figure 7a). After prolonged cultivation, there was no further growth and the plants died. The rest of the mdo1-1 atm-4 double mutants (six out of 22 plants) grew like the mdo1-1 single mutant during the vegetative growth phase, and then developed inflorescence stems that also showed abnormal floral bud phyllotaxy from axial buds, but not from primary meristems, which had presumably ceased to grow (Figure S3a). Even in this double mutant population, fruit development was drastically impaired (Figure S3b). To determine whether or not the root phenotype of mdo1-1 is enhanced by atm-4, 7-day-old seedlings were stained with PI. Although no 2 5 3 3.5 4 0 0.5 1 1.5 2 2. 5 Relative level of mRNA RAD51/TUB2 ATGR1/TUB2 ATBRCA1/TUB2 (a) (j) (l) (m) (k) (b) (d) (e) (h) (i) (c) (f) (g) Figure 5. Elevated level of DNA damage in the mdo1-1 mutant. (a) Expression of DNA damage-inducible genes. Total RNA was prepared from tissues including shoot apices of 14-day-old seedlings and then subjected to real-time RT-PCR analysis. The level of TUB2 mRNA was used as an internal control, and the values in mdo1-1 (blue bars) are expressed as ratios to the values in Col (red bars). Values represent averages of three biological replicates with standard deviation. (b–e) Expression of CYCB1;1:GUS. Seedlings of Col (b, d) and mdo1-1 (c, e) were subjected to GUS staining. Longitudinal sections of shoot apices (b, c; 14-day-old seedlings) and whole-mount images of root tips (d, e; 7-day-old seedlings) are shown. Scale bars: 50 lm for (b, c); 100 lm for (d, e). (f–i) Expression of CYCB1;2:GUS. Seedlings of Col (f, h) and mdo1-1 (g, i) were subjected to GUS staining. Longitudinal sections of shoot apices (f, g; 14-day-old seedlings) and whole-mount images of root tips (h, i; 7-day-old seedlings) are shown. Scale bars: 50 lm for (f, g); 100 lm for (h, i). (j–m) Detection of DSBs in the mdo1-1 mutant by TUNEL assay. Col (j, l) and mdo1-1 (k, m) seedlings (14-days old) were subjected to the TUNEL assay. Longitudinal sections of shoot apices (j, k) and root tips (l, m) are shown. Nomarski images (left), TUNEL-stained images (middle) and DAPI-stained nuclei (right) are shown. Scale bars: 50 lm for (j, k); 100 lm for (l, m). 662 Yuma Hashimura and Chiharu Ueguchi ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 68, 657–669

DNAdamageandstemcell maintenance663(a)Bleomycin (μg ml-1)00.250.51.01.52.0米米米水Col米咪水mdo1-1mdo1-1mdo1-1?Satr-4mdo1-1atm*兴atm-4大来福(b)mdo1-1atm-4Figure 6. Hypersensitivity of mdo1-1 to bleomycin.Plants were grown on MS agar plates for 8days and thenfurther grown onMS agarplates containingthe indicated concentrations of bleomycin for14days.Rosette structures aftertheremoval of thebolted inflorescences, ifnecessary,are shown.Scale bar:1 cm.CoPl-stained cells were observedin the atm-4 roottips,alargernumberofPl-stainedcellswasdetectedinthemdo1-1atm-4doublemutantthaninthemdo1-1singlemutantroottips(Figure 7b).These results thus suggested that the atm-4mutation has additive effects on the mdo1-1shoot and rootphenotypes.In contrastto themdo1-1 atm-4doublemutant,themdo1-1atr-4doublemutantexhibitedthemdo1-1singlemdo1-1mdo1-mutant phenotype in both shoots and roots (Figure 7 andmdo1-1atm-4atr-aFigure S3b), indicating that the atr-4 mutation does notenhancethemdo1-1phenotype.Figure 7. The atm-4 mutation enhances the mdo1-1 phenotypes.We further examined whetherornot the atm-4mutation(a) Rosette morphology of 32-day-old plants. Inflorescence stems of theindicated plantsexceptformdo1-1atm-4were removed.Scalebar:1cmenhances the sensitivity of mdo1-1 to DNA-damaging(b) Phenotypes of root tips. Seedlings (7-days old) were stained withagents.As shown in Figure6,themdo1-1atm-4doublepropidium iodide.Scale bar:100 μmmutantshowedstrongersensitivitytobleomvcinthaneitherthemdo1-1or atm-4singlemutant:thedoublemutant plantsdied with just 0.25 mg m/-1 bleomycin. We obtained essen-tiallythe sameresultwhenMMCwasused(FigureS2).TakenTo determine whether or not the mis-sense mutation intogether,theseresults stronglysuggest thatthefunction ofAt1g56260isindeedresponsibleforthemdo1-1phenotype,MDO1iscloselyrelatedtothatofATM,butnottothatofATRwe carried out complementation analysis.We isolated aapproximately4-kb Spel-Spel genomicfragmentencom-IdentificationoftheMDO1genepassing the At1g56260 coding region as well asapproxi-To identifytherelevantgeneforthemdo1-1mutation,wemately2.2kbputativepromoterregion from BACclonecarried out map-based cloning.The mutation was roughlyF14G9,andcloned itintoanappropriateT-DNAvector.Themapped to themiddle region ofchromosomel.Subsequentresultantconstructwasthenintroduced intothemdo1-1fine mapping revealed that the mutation is located within aplants.Thetransgenicplants showednormalgrowth inbothapproximately 40-kb region,in which 19 putativegenesshootsandroots (Figures 1h,land3a).Theresultsindicated(At1g56250-At1g56420)areannotated.Genomic sequenc-that the wild-type At1g56260 allele can fully rescue theing of all putative coding regions in the mdo1-1genomemdo1-1phenotype,and thus weconcluded that thispartic-allowed us to find a unique base substitution, a G-Aular gene is MDO1.The MD01 gene encodes a 127 amino acid protein,transition,in thethird exonof At1g56260.The mutationresultsinaGly→Gluaminoacidsubstitutionatthe77ththe primary sequence of which contains no known funccodon (Figure 8a).tional domain, including a subcellular localization signal.@2011TheAuthorsThePlantJournal2011Blackwell PublishingLtd,ThePlantJournal,(2011),68,657-669

PI-stained cells were observed in the atm-4 root tips, a larger number of PI-stained cells was detected in the mdo1-1 atm-4 double mutant than in the mdo1-1 single mutant root tips (Figure 7b). These results thus suggested that the atm-4 mutation has additive effects on the mdo1-1 shoot and root phenotypes. In contrast to the mdo1-1 atm-4 double mutant, the mdo1-1 atr-4 double mutant exhibited the mdo1-1 single mutant phenotype in both shoots and roots (Figure 7 and Figure S3b), indicating that the atr-4 mutation does not enhance the mdo1-1 phenotype. We further examined whether or not the atm-4 mutation enhances the sensitivity of mdo1-1 to DNA-damaging agents. As shown in Figure 6, the mdo1-1 atm-4 double mutant showed stronger sensitivity to bleomycin than either the mdo1-1 or atm-4 single mutant: the double mutant plants died with just 0.25 mg ml)1 bleomycin. We obtained essen￾tially the same result when MMC was used (Figure S2). Taken together, these results strongly suggest that the function of MDO1 is closely related to that of ATM, but not to that of ATR. Identification of the MDO1 gene To identify the relevant gene for the mdo1-1 mutation, we carried out map-based cloning. The mutation was roughly mapped to the middle region of chromosome I. Subsequent fine mapping revealed that the mutation is located within a approximately 40-kb region, in which 19 putative genes (At1g56250–At1g56420) are annotated. Genomic sequenc￾ing of all putative coding regions in the mdo1-1 genome allowed us to find a unique base substitution, a G fi A transition, in the third exon of At1g56260. The mutation results in a Gly fi Glu amino acid substitution at the 77th codon (Figure 8a). To determine whether or not the mis-sense mutation in At1g56260 is indeed responsible for the mdo1-1 phenotype, we carried out complementation analysis. We isolated a approximately 4-kb SpeI–SpeI genomic fragment encom￾passing the At1g56260 coding region as well as approxi￾mately 2.2 kb putative promoter region from BAC clone F14G9, and cloned it into an appropriate T-DNA vector. The resultant construct was then introduced into the mdo1-1 plants. The transgenic plants showed normal growth in both shoots and roots (Figures 1h,l and 3a). The results indicated that the wild-type At1g56260 allele can fully rescue the mdo1-1 phenotype, and thus we concluded that this partic￾ular gene is MDO1. The MDO1 gene encodes a 127 amino acid protein, the primary sequence of which contains no known func￾tional domain, including a subcellular localization signal. Bleomycin (μg ml–1) 0 0.25 1.0 0.5 2.0 1.5 atm-4 mdo1-1 atm-4 Col mdo1-1 Figure 6. Hypersensitivity of mdo1-1 to bleomycin. Plants were grown on MS agar plates for 8 days and then further grown on MS agar plates containing the indicated concentrations of bleomycin for 14 days. Rosette structures after the removal of the bolted inflorescences, if necessary, are shown. Scale bar: 1 cm. Col atm-4 atr-4 mdo1-1 atm-4 mdo1-1 atr-4 mdo1-1 Col atm-4 atr-4 mdo1-1 atm-4 mdo1-1 mdo1-1 atr-4 (a) (b) Figure 7. The atm-4 mutation enhances the mdo1-1 phenotypes. (a) Rosette morphology of 32-day-old plants. Inflorescence stems of the indicated plants except for mdo1-1 atm-4 were removed. Scale bar: 1 cm. (b) Phenotypes of root tips. Seedlings (7-days old) were stained with propidium iodide. Scale bar: 100 lm. DNA damage and stem cell maintenance 663 ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 68, 657–669

664 Yuma Hashimuraand ChiharuUeguchi(a)Gly-→GluAt1g56260VGGA→GAA→ATGTGA(b)A.thaliana88888881888888z.officinaleTEG.maxETM.truncatulinifer>nat.eaiiis.T.aestivuz.mayso.sativapinastenate80100120A.thalianaKTIEI127DWEE1HIEQ--PNNEAFLEEEDNSNMVE27B.naDnznt进培话培场培officinaleG:epuncatulaUUneTr.officinal4-INmajus-eduli:estivnt330O.sative-pinaster118P.patensFigure 8. Molecular characterization of the MDO1 gene.(a)Genomic structureof theMDO1gene.The structureofthe chromosomal region encompassingtheMDO1geneis shownschematically.Theclosed boxesrepresent exons including the MDO1open reading frame, with presumed initiation and termination codons.The open boxes represent 5'and 3'untranslatedregions.The position of the mdo1-1 mutation is indicated by an arrow. The relevant nucleotide and amino acid substitutions in mdo1-1 are also indicated.(b) Alignment of the predicted sequence of the MDO1 protein and putative homologs. Sequence alignment was conducted with cLusTAL x (Tompson et al., 1997)and visualized using GENEDoc (Nicholas et al., 1997). Amino acids highlighted in black and gray are those conserved completely and conserved in more than 75% ofthe presented homologs,respectively.The substituted amino acid in mdo1-1is indicated by an arrow. Accession numbers used to construct the alignment arefllows:BT011713(Arabidopsis thaliana,Atg56260),V025118(Brassica napus),DY351937(Zingiberofficinale),BU546344(Glycine max,AW257229(Medicagtruncatula),EC941125(Vitisvinifera),DY817411(Taraxacumofficinale),AJ807040(Antirrhinummajus),FP094644(Phyllostachysedulis),BJ244238(Triticumaestivum),BT083651(Zeamays),CB670331(Oryza sativa),CT575900 (Pinuspinaster)andXM_001782167(Physcomitrellapatens).Interestingly, the MDO1 sequence is highly conserved in aDISCUSSIONwide variety of land plants, including a moss,but not inotherorganisms,suchasanimals.Themdo1-1mutationisWedemonstratedthatthestemcellsinmdo1-1apicalmer-located within a five amino acid tract (FIGEL)that isistems have a tendency to lose their undifferentiated state,completely conserved inallknown homologs (Figure8b).resultingincell differentiationorcell death.InthemutantThis fact suggests that this region plays an important role,shoot apex,the surface cells in stem cell niches began toand thatthemdo1-1 mutation substantially affects thebecome enlarged,and ANT expression increased but,function of the MDO1protein.inversely,STMexpressiondecreased (Figure2).Similarly,in the mutant root tips, QC cells failed to maintain theirExpressionoftheMDO1geneidentity,columella initials accumulated starchgranules andTo examine the expression of the MDO1 gene, we con-cortical/endodermal initials lost theirfunction (Figure 3).WestructedaMDO1promoter:GUSfusion(MDO1pro:GUS)asalso observedthe preferential death of stem cells and theiraprobe.Asshown inFigure9,the MDO1geneis ex-earlydescendants intheroottips(Figure4).Itisapparentpressed nearly ubiquitously throughout the whole plant,thatthese abnormalities leadto the disorganization of apicalbut at a low level.In seedlings,strong staining was mainlymeristems.Inmdo1-1shoots,several lateral organs,leavesobserved inmeristematictissues,suchas intheSAM,rootand floralbuds did notdevelop in a spatiallyand temporallytips, growing leaves and lateral root primordia. The vas-regulatedmanner(Figure1).Thisirregularphyllotaxyandculatureswerealsostronglystainedthroughoutthewholeplastochronprobablyresultfromthedisorganizationoftheplant.mutant SAM. It was demonstrated previously that the2011TheAuthorsThePlantJournal2011BlackwellPublishingLtd,ThePlantJournal,(2011),68,657-669

Interestingly, the MDO1 sequence is highly conserved in a wide variety of land plants, including a moss, but not in other organisms, such as animals. The mdo1-1 mutation is located within a five amino acid tract (FIGEL) that is completely conserved in all known homologs (Figure 8b). This fact suggests that this region plays an important role, and that the mdo1-1 mutation substantially affects the function of the MDO1 protein. Expression of the MDO1 gene To examine the expression of the MDO1 gene, we con￾structed a MDO1 promoter:GUS fusion (MDO1pro:GUS) as a probe. As shown in Figure 9, the MDO1 gene is ex￾pressed nearly ubiquitously throughout the whole plant, but at a low level. In seedlings, strong staining was mainly observed in meristematic tissues, such as in the SAM, root tips, growing leaves and lateral root primordia. The vas￾culatures were also strongly stained throughout the whole plant. DISCUSSION We demonstrated that the stem cells in mdo1-1 apical mer￾istems have a tendency to lose their undifferentiated state, resulting in cell differentiation or cell death. In the mutant shoot apex, the surface cells in stem cell niches began to become enlarged, and ANT expression increased but, inversely, STM expression decreased (Figure 2). Similarly, in the mutant root tips, QC cells failed to maintain their identity, columella initials accumulated starch granules and cortical/endodermal initials lost their function (Figure 3). We also observed the preferential death of stem cells and their early descendants in the root tips (Figure 4). It is apparent that these abnormalities lead to the disorganization of apical meristems. In mdo1-1 shoots, several lateral organs, leaves and floral buds did not develop in a spatially and temporally regulated manner (Figure 1). This irregular phyllotaxy and plastochron probably result from the disorganization of the mutant SAM. It was demonstrated previously that the At1g56260 GGA→GAA Gly→Glu ATG TGA * 20 * 40 * 60 A.thaliana : B.napus : Z.officinale : G.max : M.truncatula : V.vinifera : T.officinale : A.majus : P.edulis : T.aestivum : Z.mays : O.sativa : P.pinaster : P.patens : MAKSQIEP-GVPITLQELYPSSLFYKEGVSLRVTAMLRGYSVETAIGVIEDGGRSLKINTQNIR MANSQIES-GAPITLQELYPSSPFFMEARSLRVTGLLKGYSVETAIGVIEDGEKSLKINTQHLR MATSIIKP-AVPVLLKEIESGSQFFKQGASLRVTGMLHAYSEDTAVAVIADANISFKINTVHLR MASLEIKS-GALVSLQDLRPSSPFFKQGASLRITGKLHEYSIETGLATIIDGDDILKVSTKHLR MAFFEGKS-GALVSLQDMRPSSPFFKQGTSVRIIGKLHEYSSETGLATVIDGNDILKVSTEHLK MMSSAVKS-GALVSLQDLQPSSPFFKQGASLRVTGKLQEYSVETAIAIVIDGSANLKINTQHLR MASASIPS-GARVLLQELDSSSSHFKQGASLRVTGKLQEYSVETAIAIVADGGATLAVDTQHLR MGTTSALSA-GALVNLKEVNPSSPIFNQGVSLRVTGKLQDYNLETAVVVIVDEESSLKVDTQYLK MASPALKP-GVPITLRELVPSSEMFKQGTSLRVTGNLQSYDVDPAIAVIQDGSMSLKVDTQHLR MAPSTLKP-GVPITLQELEPSSEMFKQGASLRVTGILQSYDVDSAVVVIQDGSARLKIDTQNLR MASAGLEP-GVPVILRELEPSSEMFKQGASLRVTGSLQSFDVESATATIQDGSVSLKVDTQHLR MASSVLQP-GVPVTLQELEPSSESFRQGASLRVTGVLQSYDLNSAIAVIQDGGASLKVDTQNLR MASASFSASTVKAGVLVMLDELNPSSPFFTNGASLRLTGRLQEFSVETAIAVIVDGGATFQIDTQNLR MAVVTLNP-GLVVRLQELQPGSSFVLPNQSLRVVGSLQSFDSTSGIAVLVDAGASLRLDLEHLR : 63 : 63 : 63 : 63 : 63 : 63 : 63 : 64 : 63 : 63 : 63 : 63 : 68 : 63 * 80 * 100 * 120 * A.thaliana : B.napus : Z.officinale : G.max : M.truncatula : V.vinifera : T.officinale : A.majus : P.edulis : T.aestivum : Z.mays : O.sativa : P.pinaster : P.patens : 80 100 120 DVSFRVGSIYQFIGELHIEQ-PNNEAILQARTGRNVDGIDMNLYRKTIELLRQFLK-EEDNSNMVE DVSFRVGSVYQFIGELHIE-PNNEPILQARTGRNVDGIDINLYRKTIELLRRFLEVEEDNRNMVE DLSFRFGSYYQFIGELHIL-PEDVAILQARVGRIVDGLDPNLYNQSLQLRRQFES-ELVNS- DLTFQVGSVYQFIGELLIQ-PDNEGVLQARVGRNVDGIDLNLYHQSLLLLRQFQTNHLNNPATM￾DLKFQVGSVYQFIGELLIR-TDNEGVLQAHVGRNVDGIDLNLYHQSLLLLKQFQANHLNNSAT- DLTFRAGSIYQFIGELLIQ-PDNEAILQARVGRNVDGIDLNLYHQSLQLVRQFQADHMNDQAA- -LNLRVGSLYQFIGELSIQ-PNNEGILKARVGRNVDGMDLNLYQQSLKLLRQFQSDQISHLENLQ -INIRPGSLYQFIGELDIK-PDNEAILKARVGRNVDGMDLNLYRQSLQLLKQFQSEQTNIQKD- DISFRTNSMYQFIGELLIQ-PDNDAILQARVGRNVDGLDLNLYQQSLLIRRQYEAKLRSSRRA- DISFRSNSTFQFIGELLIQ-PNNDAILQARVGRNVDGLDLNLYQQSLIIRRQHEAKLLSSRRP- DVSFRTNSVYQFIGELQIR-EVDDAILLARIGRNVDGLDMNLYQQALLIRRQHEAKLLSSRRA- EISFRTNSTYQFIGELLIK-PDNDAVLQARVGRNVDGIDLNLYQQSLLIRRQYEAQLRS-RRA- DIHFRIDSLYQFIGELLIQPGLQPHQAILQARVGRNVDGMDMHLYRKSLQLRRQFEAKYMGMQST- ELPLRVGSLFEFIGELEVD-TTQQARVGRNVDGMDMKLFDKVLQLRRRFEKDYKHN- : 127 : 127 : 122 : 126 : 125 : 125 : 126 : 125 : 125 : 125 : 125 : 124 : 133 : 118 (a) (b) Figure 8. Molecular characterization of the MDO1 gene. (a) Genomic structure of the MDO1 gene. The structure of the chromosomal region encompassing the MDO1 gene is shown schematically. The closed boxes represent exons including the MDO1 open reading frame, with presumed initiation and termination codons. The open boxes represent 5¢ and 3¢ untranslated regions. The position of the mdo1-1 mutation is indicated by an arrow. The relevant nucleotide and amino acid substitutions in mdo1-1 are also indicated. (b) Alignment of the predicted sequence of the MDO1 protein and putative homologs. Sequence alignment was conducted with CLUSTAL X (Tompson et al., 1997) and visualized using GENEDOC (Nicholas et al., 1997). Amino acids highlighted in black and gray are those conserved completely and conserved in more than 75% of the presented homologs, respectively. The substituted amino acid in mdo1-1 is indicated by an arrow. Accession numbers used to construct the alignment are follows: BT011713 (Arabidopsis thaliana, At1g56260), EV025118 (Brassica napus), DY351937 (Zingiber officinale), BU546344 (Glycine max), AW257229 (Medicago truncatula), EC941125 (Vitis vinifera), DY817411 (Taraxacum officinale), AJ807040 (Antirrhinum majus), FP094644 (Phyllostachys edulis), BJ244238 (Triticum aestivum), BT083651 (Zea mays), CB670331 (Oryza sativa), CT575900 (Pinus pinaster) and XM_001782167 (Physcomitrella patens). 664 Yuma Hashimura and Chiharu Ueguchi ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 68, 657–669

DNAdamageandstemcell maintenance665mdo1-1stemcellsmightperturbthepositional information,(a)(Cresulting in the disorganization of meristem structures.Indeed,weobservedtheectopicformationof SCR-express-ing cell files in the steles of themutant root tips (Figure 3).ThemultipleSTMexpressiondomains inthemdo1-1SAM(Figure 2) may also be the consequence of the ectopicregeneration of stem cells. Or,alternatively,the single STM-expression domain may be simply divided into multipledomains through cell differentiation of medial stem cells inthe stem cell niches. In either case, such multiple STM-expressingdomainswillresultinstemfasciationAnimalstemcellsgenerallyshowalowtolerancetogenotoxic stress, and tend to undergo differentiation orbprogrammed cell death in response to DNA damage (Richetal.,2000;Sherman etal.,2011).Likewise,it wasdemonstratedrecentlythatplantstemcellsandtheirearlydescendants selectivelyundergoATM/ATR-dependent andnon-apoptotic programmed cell death under certain mildgenotoxic treatments(Fulcher andSablowski,2009).Wealso observed that the cells in stemcell niches diedpreferentially in mdo1-1 roottips, even though an elevatedDSB level wasdetectednotonlyin stemcell niches but alsoindifferentiatingcells(Figures4and5).TheseobservationsFigure 9.Expression of MDO1pro:GUS: 10-day-old seedlings of Col harbor-ing MDO1pro:GUSwere subjected toGUS stainingthus suggest that the elevated level of DSBs observed inAwholeplant (a),alongitudinal sectionofa shootapex (b)andaroot tip (c)mdo1-1 cells induces cell death preferentially in the rootare shown. Scale bars: 2 μm for (a); 50 μm for (b); and 100 μm for (c)stem cellniches.Cell differentiation in stem cell niches maybecausedbythemalfunctionoforganizingcentercells,suchas in the QC,and/or byadirect effect of DNA damage,as inpositions oflateralorgan primordiaaredeterminedbyauxinthecaseofanimalstemcells.ltremainsobscurewhetheroraccumulation that occurs through polar auxin transportnot the preferential cell death is ATM/ATR-dependent pro-through the surface cell layerof the SAM (Reinhardt et al.,grammed cell death.It is apparent that the ATM kinase is2000,2003a).In disorganizedmdo1-1shootapices,in whichactivated inmdo1-1cellsbecausethe expressionof severalsurface cells start to differentiate, the spatial and temporalATM-regulated genes was enhanced and the level of DSBspattern of auxin accumulation will be perturbed,resulting inwaselevated(Figure5),supportingtheideathatcelldeathinthemdo1-1rootstemcellnichesisATM/ATR-dependentabnormalphyllotaxyandplastochron.Weobservedareducednumberofstemcellsinmdo1-1programmed cell death. On the other hand, this idea seemsapical meristems (Figures 2and 3).Nevertheless,exceptforto be inconsistent withthe finding thatthenumber of deadthe transient arrest of shoot growth (stop-and-go pheno-cells increased with the introduction of the atm-4 mutationtype),the growth of the mdo1-1 shoots and roots was notintothemdo1-1background(Figure7).Onepossiblereasoncompletely arrested (Figures 1 and 3). This fact stronglyis that the atm-4 allele used is hypomorphic.Or, alterna-suggests that new stem cells are supplied continuouslytotively,in the mdo1-1 atm-4 double mutant, ATR instead ofthe mutant apical meristems.The new stem cells maybeATMmightplaya regulatoryrole inprogrammedcell death.providedthroughthedivisionoftheresidualstemcellsand/Inanyevent,furtherexperimentswill resolvetheissueintheor,in an alternative way,through theregenerationofstemfuture.cellsfromdifferentiatingones.Inplants,ithasbeenreportedFrom the results described in this study,taken together,thatregenerationofstemcellnichescantakeplaceinboththe following simple scenario explaining thegrowth pheno-shoot and root apical meristems, even after the surgicaltype of mdo1-1has emerged.An elevated level of DNAremoval of authentic stem cell niches (Reinhardt et al.damage includingDSBs leadsto deathand the differentia-2003b;Senaetal.,2009).Similarly,inmdo1-1apicalmeristion of stem cells in apical meristems (maybe through theactivation of the ATM/ATR pathway),finally resulting intemsasupplyofregeneratedstemcellsmaycompensatefothe loss of stemcells.Althoughthe newmeristems regen-severalvisiblephenotypes,suchasstemfasciation,througheratedthrough surgicalexperimentsresembledtheauthen-disorganization of the meristem structure. This view isticonesstructurally,themdo1-1shootapicesdidnotshowconsistent with the previous reports that mutations inregular structures.Continuous, not transient, lesions of theDNArepairgenes occasionally cause stem fasciation.For2011TheAuthorsThePlantJournal2011Blackwell PublishingLtd,ThePlantJournal,(2011),68,657-669

positions of lateral organ primordia are determined by auxin accumulation that occurs through polar auxin transport through the surface cell layer of the SAM (Reinhardt et al., 2000, 2003a). In disorganized mdo1-1 shoot apices, in which surface cells start to differentiate, the spatial and temporal pattern of auxin accumulation will be perturbed, resulting in abnormal phyllotaxy and plastochron. We observed a reduced number of stem cells in mdo1-1 apical meristems (Figures 2 and 3). Nevertheless, except for the transient arrest of shoot growth (stop-and-go pheno￾type), the growth of the mdo1-1 shoots and roots was not completely arrested (Figures 1 and 3). This fact strongly suggests that new stem cells are supplied continuously to the mutant apical meristems. The new stem cells may be provided through the division of the residual stem cells and/ or, in an alternative way, through the regeneration of stem cells from differentiating ones. In plants, it has been reported that regeneration of stem cell niches can take place in both shoot and root apical meristems, even after the surgical removal of authentic stem cell niches (Reinhardt et al., 2003b; Sena et al., 2009). Similarly, in mdo1-1 apical meris￾tems a supply of regenerated stem cells may compensate for the loss of stem cells. Although the new meristems regen￾erated through surgical experiments resembled the authen￾tic ones structurally, the mdo1-1 shoot apices did not show regular structures. Continuous, not transient, lesions of the mdo1-1 stem cells might perturb the positional information, resulting in the disorganization of meristem structures. Indeed, we observed the ectopic formation of SCR-express￾ing cell files in the steles of the mutant root tips (Figure 3). The multiple STM expression domains in the mdo1-1 SAM (Figure 2) may also be the consequence of the ectopic regeneration of stem cells. Or, alternatively, the single STM￾expression domain may be simply divided into multiple domains through cell differentiation of medial stem cells in the stem cell niches. In either case, such multiple STM￾expressing domains will result in stem fasciation. Animal stem cells generally show a low tolerance to genotoxic stress, and tend to undergo differentiation or programmed cell death in response to DNA damage (Rich et al., 2000; Sherman et al., 2011). Likewise, it was demonstrated recently that plant stem cells and their early descendants selectively undergo ATM/ATR-dependent and non-apoptotic programmed cell death under certain mild genotoxic treatments (Fulcher and Sablowski, 2009). We also observed that the cells in stem cell niches died preferentially in mdo1-1 root tips, even though an elevated DSB level was detected not only in stem cell niches but also in differentiating cells (Figures 4 and 5). These observations thus suggest that the elevated level of DSBs observed in mdo1-1 cells induces cell death preferentially in the root stem cell niches. Cell differentiation in stem cell niches may be caused by the malfunction of organizing center cells, such as in the QC, and/or by a direct effect of DNA damage, as in the case of animal stem cells. It remains obscure whether or not the preferential cell death is ATM/ATR-dependent pro￾grammed cell death. It is apparent that the ATM kinase is activated in mdo1-1 cells because the expression of several ATM-regulated genes was enhanced and the level of DSBs was elevated (Figure 5), supporting the idea that cell death in the mdo1-1 root stem cell niches is ATM/ATR-dependent programmed cell death. On the other hand, this idea seems to be inconsistent with the finding that the number of dead cells increased with the introduction of the atm-4 mutation into the mdo1-1 background (Figure 7). One possible reason is that the atm-4 allele used is hypomorphic. Or, alterna￾tively, in the mdo1-1 atm-4 double mutant, ATR instead of ATM might play a regulatory role in programmed cell death. In any event, further experiments will resolve the issue in the future. From the results described in this study, taken together, the following simple scenario explaining the growth pheno￾type of mdo1-1 has emerged. An elevated level of DNA damage including DSBs leads to death and the differentia￾tion of stem cells in apical meristems (maybe through the activation of the ATM/ATR pathway), finally resulting in several visible phenotypes, such as stem fasciation, through disorganization of the meristem structure. This view is consistent with the previous reports that mutations in DNA repair genes occasionally cause stem fasciation. For (a) (b) (c) Figure 9. Expression of MDO1pro:GUS: 10-day-old seedlings of Col harbor￾ing MDO1pro:GUS were subjected to GUS staining. A whole plant (a), a longitudinal section of a shoot apex (b) and a root tip (c) are shown. Scale bars: 2 lm for (a); 50 lm for (b); and 100 lm for (c). DNA damage and stem cell maintenance 665 ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 68, 657–669

666YumaHashimura and ChiharuUeguchi1mutation.The atm-4 (SALK_036940)and atr-4 (SALK_054383)example,mutations of theArabidopsis MRE11gene,themutants wereobtainedfromtheArabidopsis Biological Resourceproduct of which is required for the recognition of DSBs,Center(ABRC,http://abrc.osu.edu).QC25,QC46andQC184(ecotypeexhibit several developmental phenotypes,including stemWassilewskija) were kindly provided by Drs B. Scheres (Utrechtfasciation(BundockandHooykaas,2002).MutationallesionsUniversity,theNetheriands)and M.Aida(NAiST,Nara Instituteofof theArabidopsis BRCA2genes,required for efficientDSBScience and Technology,Japan).SCR:GFP-,TIP,CYCB1;1:GUS andrepair,result in stem fasciation and abnormal phyllotaxy,CYCB1;2:GUS were kind gifts from Drs M.Morita (NAIST),P.Doerner (Edinburgh University,UK) and M, Umeda (NAIST),and M.albeit at a low incidence (Abe et al., 2009).Ito (Nagoya University,Japan),respectivelyItisintriguingthatseveralfasciationmutants,suchasfasPlants were grown with 16-h light/8-h dark fluorescent illumina-andmgo3/bru1/tsk,sharenotallbutsomeofthephenotypestion at 22°C.Plants were germinated andgrown on rockfiberobservedinmdo1-1.TheFASgeneproductsareassumedto(Grodan, http://www.grodan.com) or MS medium. Transformationensure the stable propagation of epigenetic states of theofArabidopsisplantswascarriedoutaccordingtotheAgrobacterium-mediatedfloral-dip method (Cloughand Bent, 1998).chromosome, which is required for the maintenance ofseveral meristem genes,such as WUS (Kaya etal., 2001).Mapping of theMD01geneTherefore, one can envisage that, independently of DNAMapping was carried out by PCR genotyping using single nucleo-damage responses, the mdo1-1 mutation also affects mer-tide polymorphisms (SNPs) between Col and Ler. The primeristemgeneexpression,resulting in theobservedmeristeminformation on cleaved amplified polymorphic sequence (CAPS)defects. Although this possibility cannot be excluded atmarkers for rough mapping was kindly provided by Drs Miyo Moritapresent, sucha'direct'effect of mdo1-1on gene expression,and Masao Tasaka (NAIST).The primer pairs of CAPS markers forthe subsequent fine mapping were designed based on the SNPifanymightbesmallerthanthatofDNAdamageresponsesinformation from The Arabidopsis Information Resource (TAIR,in a stressful situation in which stem cells die preferentiallyhttp://www.arabidopsis.org)and Monsanto Co.In any event, the elucidation of the molecular relationshipPCR genotypingbetweenMDO1andthesefactorsmustawaitfurther studies.We found that the atm-4 mutation enhanced the mdo1-1The sequences of PCR primers used are summarized in Table S1.phenotypes:namely,severegrowth arrest,anincreasedPlants were genotyped as to the MDO1 locus by CAPS analysisincluding nested PCR and restriction analysis. The genomic DNAlevel of cell death and enhanced sensitivity to DNA-damag-fragment was first amplified using MDO1-GF and MDO1-GR, andingagents.This thus suggests thatthefunction of MDO1isthen the PCR product was subjected to second PCR using MDO1-1Fcloselyrelatedto that ofATM.BecauseATMplays acentraland MDO1-TR, followed by Mboll digestion. The wild-type androlein thecellularresponsestoDSBs(Shiloh,2006),MDO1atm-4alleles were detected byPCR using atm-4F and atm-4R for theislikelytobeinvolvedinthedamageresponsetoDSBs,suchwild-type ATM allele,and atm-4F and LBa1 for the atm-4 allele. Theas the reduction of internal genotoxic stress, the protectionwild-type and atr-4 alleles were detected by PCR using atr-4F andatr-4R for the wild-type ATR allele, and atr-4F and LBa1 for the atr-4of chromosomal DNAagainst DNAdamageorDNA repair.allele.Because of the lack of known functional domains in theMDO1 primary sequence, further genetic and molecularPlasmidconstructionstudies are needed to elucidate the precise molecularForthe complementation test,a approximately 4-kb Spelgenomicfunctionof MDO1fragment encompassing the entire MDO1 coding region and theTo our current knowledge, the MDO1 gene is foundapproximately 2.2-kb putative promoter region was first isolatedspecifically in a wide variety ofland plants.Thisfact stronglyfrom BAC cloneF14G9(GenBankaccession number,AC069159),suggests that the gene was acquired during the course ofand then inserted into the Xbal site of pUC119 (Vieira and Messing,evolution from an aquatic ancestor to land plants.Land1987) to yield pCUA301. The Sal-Kpnl fragment including theMDO1genomic region was inserted between the Sal and Kpnl sitesplants,ascomparedwithplantslivinginawaterenviron-ofpBIB-Hm(Becker,1990)toyield pCUA307.ment, cannot escape several environmental hazards thatToconstructa MDO1promoter-GUS fusion gene(MDO1pro:-lead to DNAdamage, such as radiation,drought and highGUS),the approximately 2.2-kbputativepromoter fragment wasconcentrations of oxygen. The acquisition of the MDO1genefirst amplified by PCR from pCUA301 using MDO1-GWF1 andmighthavebeenoneoftheimportanteventsrequiredforMDO1-GWR1,andthenthePCRproductwassubjectedtosecondPCR using attB1 adaptor and attB2 adaptor.The resultant DNAtheevolution of landplants.fragmentwasclonedintopDONR221(lnvitrogen,http://www.invi-trogen.com) using Gateway BP Clonase (lnvitrogen).After confir-EXPERIMENTALPROCEDURESmation of thecloned sequence,theMDO1promoter fragment wasmoved into thepGWB533T-DNA vector (Nakagawa etal.,2007)Plant materialsusing Gateway LR clonase (lnvitrogen)to create an MDO1pro:GUSThe Columbia (Col) ecotype of A. thaliana (L.) Heynh. was used asfusiontoyieldpCUA320.the wild-type strain.The mdo1-1 mutant was isolated by screeningFor in situ hybridization,a 600-bp DNA fragment encompassingEMS-mutagenized M2 seeds, which were purchased from Lehlethe ANT cDNA was amplified from cDNA derived from wild-typeSeeds (http:/www.arabidopsis.com), and was used for furthershoot apices using ANT-1F and ANT-1R. The resultant fragmentanalyses after being backcrossed three times.The Landsberg erectawas digested with Xhol and Pstl, and then inserted between the(Ler) ecotype was used formap-based cloning to identify themdo1-Xhol and Pstl sites ofpBluescriptlIl SK(+) (Stratagene,nowAgilent2011TheAuthorsThePlant Journal 2011Blackwell Publishing Ltd,ThePlant Journal,(2011),68,657-669

example, mutations of the Arabidopsis MRE11 gene, the product of which is required for the recognition of DSBs, exhibit several developmental phenotypes, including stem fasciation (Bundock and Hooykaas, 2002). Mutational lesions of the Arabidopsis BRCA2 genes, required for efficient DSB repair, result in stem fasciation and abnormal phyllotaxy, albeit at a low incidence (Abe et al., 2009). It is intriguing that several fasciation mutants, such as fas and mgo3/bru1/tsk, share not all but some of the phenotypes observed in mdo1-1. The FAS gene products are assumed to ensure the stable propagation of epigenetic states of the chromosome, which is required for the maintenance of several meristem genes, such as WUS (Kaya et al., 2001). Therefore, one can envisage that, independently of DNA damage responses, the mdo1-1 mutation also affects mer￾istem gene expression, resulting in the observed meristem defects. Although this possibility cannot be excluded at present, such a ‘direct’ effect of mdo1-1 on gene expression, if any, might be smaller than that of DNA damage responses in a stressful situation in which stem cells die preferentially. In any event, the elucidation of the molecular relationship between MDO1 and these factors must await further studies. We found that the atm-4 mutation enhanced the mdo1-1 phenotypes: namely, severe growth arrest, an increased level of cell death and enhanced sensitivity to DNA-damag￾ing agents. This thus suggests that the function of MDO1 is closely related to that of ATM. Because ATM plays a central role in the cellular responses to DSBs (Shiloh, 2006), MDO1 is likely to be involved in the damage response to DSBs, such as the reduction of internal genotoxic stress, the protection of chromosomal DNA against DNA damage or DNA repair. Because of the lack of known functional domains in the MDO1 primary sequence, further genetic and molecular studies are needed to elucidate the precise molecular function of MDO1. To our current knowledge, the MDO1 gene is found specifically in a wide variety of land plants. This fact strongly suggests that the gene was acquired during the course of evolution from an aquatic ancestor to land plants. Land plants, as compared with plants living in a water environ￾ment, cannot escape several environmental hazards that lead to DNA damage, such as radiation, drought and high concentrations of oxygen. The acquisition of the MDO1 gene might have been one of the important events required for the evolution of land plants. EXPERIMENTAL PROCEDURES Plant materials The Columbia (Col) ecotype of A. thaliana (L.) Heynh. was used as the wild-type strain. The mdo1-1 mutant was isolated by screening EMS-mutagenized M2 seeds, which were purchased from Lehle Seeds (http://www.arabidopsis.com), and was used for further analyses after being backcrossed three times. The Landsberg erecta (Ler) ecotype was used for map-based cloning to identify the mdo1- 1 mutation. The atm-4 (SALK_036940) and atr-4 (SALK_054383) mutants were obtained from the Arabidopsis Biological Resource Center (ABRC, http://abrc.osu.edu). QC25, QC46 and QC184 (ecotype Wassilewskija) were kindly provided by Drs B. Scheres (Utrecht University, the Netherlands) and M. Aida (NAIST, Nara Institute of Science and Technology, Japan). SCR:GFP-cTIP, CYCB1;1:GUS and CYCB1;2:GUS were kind gifts from Drs M. Morita (NAIST), P. Doerner (Edinburgh University, UK) and M. Umeda (NAIST), and M. Ito (Nagoya University, Japan), respectively. Plants were grown with 16-h light/8-h dark fluorescent illumina￾tion at 22C. Plants were germinated and grown on rockfiber (Grodan, http://www.grodan.com) or MS medium. Transformation of Arabidopsis plants was carried out according to the Agrobacte￾rium-mediated floral-dip method (Clough and Bent, 1998). Mapping of the MDO1 gene Mapping was carried out by PCR genotyping using single nucleo￾tide polymorphisms (SNPs) between Col and Ler. The primer information on cleaved amplified polymorphic sequence (CAPS) markers for rough mapping was kindly provided by Drs Miyo Morita and Masao Tasaka (NAIST). The primer pairs of CAPS markers for the subsequent fine mapping were designed based on the SNP information from The Arabidopsis Information Resource (TAIR, http://www.arabidopsis.org) and Monsanto Co. PCR genotyping The sequences of PCR primers used are summarized in Table S1. Plants were genotyped as to the MDO1 locus by CAPS analysis, including nested PCR and restriction analysis. The genomic DNA fragment was first amplified using MDO1-GF and MDO1-GR, and then the PCR product was subjected to second PCR using MDO1-1F and MDO1-1R, followed by MboII digestion. The wild-type and atm-4 alleles were detected by PCR using atm-4F and atm-4R for the wild-type ATM allele, and atm-4F and LBa1 for the atm-4 allele. The wild-type and atr-4 alleles were detected by PCR using atr-4F and atr-4R for the wild-type ATR allele, and atr-4F and LBa1 for the atr-4 allele. Plasmid construction For the complementation test, a approximately 4-kb SpeI genomic fragment encompassing the entire MDO1 coding region and the approximately 2.2-kb putative promoter region was first isolated from BAC clone F14G9 (GenBank accession number, AC069159), and then inserted into the XbaI site of pUC119 (Vieira and Messing, 1987) to yield pCUA301. The SalI–KpnI fragment including the MDO1 genomic region was inserted between the SalI and KpnI sites of pBIB-Hm (Becker, 1990) to yield pCUA307. To construct a MDO1 promoter-GUS fusion gene (MDO1pro:- GUS), the approximately 2.2-kb putative promoter fragment was first amplified by PCR from pCUA301 using MDO1-GWF1 and MDO1-GWR1, and then the PCR product was subjected to second PCR using attB1 adaptor and attB2 adaptor. The resultant DNA fragment was cloned into pDONR221 (Invitrogen, http://www.invi￾trogen.com) using Gateway BP Clonase (Invitrogen). After confir￾mation of the cloned sequence, the MDO1 promoter fragment was moved into the pGWB533 T-DNA vector (Nakagawa et al., 2007) using Gateway LR clonase (Invitrogen) to create an MDO1pro:GUS fusion to yield pCUA320. For in situ hybridization, a 600-bp DNA fragment encompassing the ANT cDNA was amplified from cDNA derived from wild-type shoot apices using ANT-1F and ANT-1R. The resultant fragment was digested with XhoI and PstI, and then inserted between the XhoI and PstI sites of pBluescript II SK(+) (Stratagene, now Agilent 666 Yuma Hashimura and Chiharu Ueguchi ª 2011 The Authors The Plant Journal ª 2011 Blackwell Publishing Ltd, The Plant Journal, (2011), 68, 657–669