LETTERS Conserved factors regulate signalling in Arabidopsis thaliana shoot and root stem cell organizers Ananda K. Sarkar1 *{, Marijn Luijten2 *, Shunsuke Miyashima3 , Michael Lenhard1 {, Takashi Hashimoto3 , Keiji Nakajima3 , Ben Scheres2 , Renze Heidstra2 & Thomas Laux1 Throughout the lifespan of a plant, which in some cases can last more than one thousand years, the stem cell niches in the root and shoot apical meristems provide cells for the formation of complete root and shoot systems, respectively. Both niches are superficially different and it has remained unclear whether common regulatory mechanisms exist. Here we address whether root and shoot meristems use related factors for stem cell maintenance. In the root niche the quiescent centre cells, surrounded by the stem cells, express the homeobox gene WOX5 (WUSCHEL-RELATED HOMEOBOX 5)1 , a homologue of the WUSCHEL (WUS) gene that non-cell-autonomously maintains stem cells in the shoot meri- stem2 . Loss of WOX5 function in the root meristem stem cell niche causes terminal differentiation in distal stem cells and, redundantly with other regulators, also provokes differentiation of the proximal meristem. Conversely, gain of WOX5 function blocks differentiation of distal stem cell descendents that normally differentiate. Importantly, both WOX5 and WUS maintain stem cells in either a root or shoot context. Together, our data indicate that stem cell maintenance signalling in both meristems employs related regulators. Higher organisms evolved the ability to keep founder cells undifferentiated and pluripotent by signals provided in specialized stem cell niches3 . In the model plant Arabidopsis thaliana, root and shoot meristem stem cell niches are organized differently compared with each other4 . In the root meristem, the stem cells surround a small group of rarely dividing cells, termed the quiescent centre (Fig. 1a), and give rise to distal (columella), lateral (lateral root cap and epidermis) and proximal (cortex, endodermis and stele) cell types. Ablation studies show that short range signals from the quiescent centre keep only the directly abutting stem cells undifferentiated5 , similar to most animal stem cell niches studied so far3 . In contrast, in the shoot meristem, a zone of three stem cell layers is maintained by an underlying organizing centre2 (Supplementary Fig. 1a), which regulates the stem cell pool as a whole. In addition, the regulatory genes described so far for root and shoot stem cell niches are different6 . In the shoot meristem, a feedback mechanism between the organizing centre and stem cells dynamically regulates maintenance of the stem cell pool: WUS activity in the organizing centre keeps stem cells undifferentiated and induces expression of the signal peptide CLAVATA3 (CLV3), which in turn restricts the size of the WUS expression domain (Supplementary Fig. 1a)7,8 . We asked whether a related stem cell maintenance mechanism might operate in the root by analysing the WOX family for functions in the root meristem1 . Expression of WOX5 messenger RNA1 and a WOX5-GUS reporter gene initiates in the embryonic cell lineage that gives rise to the quiescent centre (Supplementary Fig. 1b), and persists in the quiescent centre (Fig. 1b; and Supplementary Fig. 1c) during postembryonic root growth. This expression pattern is strikingly similar to that of WUS in the shoot meristem2 (Supplementary Fig. 2), raising the question of whether both genes might have similar functions in the respective stem cell niches. In seedlings of the putative null allele wox5-1 (Supplementary Fig. 1d, e), the cells at the quiescent centre position have abnormal shape and are enlarged in comparison to wild type (Fig. 1c, d, arrows, and Supplementary Table 1). Notably, adjacent columella stem cells are even larger, suggesting that they have undergone differentiation (Fig. 1d, bottom arrow). No other abnormality was observed in wox5-1 mutants. Complementation of the mutant defects by a WOX5 complementary DNA construct confirmed that the wox5-1 mutation causes the observed phenotype (Supplementary Fig. 3a). *These authors contributed equally to this work. 1 Institute of Biology III, University of Freiburg, Scha¨nzlestraße 1, 79104 Freiburg, Germany. 2 Department of Molecular Cell Biology, Utrecht University, Padualaan 8, 3584 CH Utrecht, The Netherlands. 3 Graduate School of Biological Sciences, Nara Institute of Science and Technology, 8916-5 Takayama, Ikoma, Nara 630-0192, Japan. {Present addresses: Cold Spring Harbor Laboratory, One Bungtown Road, Cold Spring Harbor, New York 11724, USA (A.K.S.); John Innes Centre, Colney Lane, Norwich, N4 7UH, UK (M.K.). col vas lrc epc en qc cscCSC QC Stem cells Wild type Wild type QC CSC dc wox5-1 ba Differentiation Differentiation Figure 1 | WOX5 expression in the quiescent centre is required for root meristem development. a, Schematic of A. thaliana root meristem with columella (col), columella stem cells (CSC), quiescent centre (QC), lateral root cap (lrc), epidermis (ep), cortex (c), endodermis (en) and vascular bundle (vas), redrawn with permission from ref. 4. b, WOX5 mRNA expression in the quiescent centre. c–d, wox5-1 roots display enlarged cells at the quiescent centre and columella stem cell (CSC) positions. Scale bars, 50 mm. Vol 446|12 April 2007|doi:10.1038/nature05703 811 Nature©2007 Publishing Group Of several quiescent-centre-specific reporter genes9 , only QC184 is not expressed in wox5-1 and expression is lacking as early as the heart embryo stage (Fig. 2a, b, d, e; and Supplementary Table 2), indicating that WOX5 functions in root development from early embryogenesis. In contrast, QC25, QC46 and WOX5-GUS are expressed in slightly expanded domains compared to wild type (Fig. 2c, f; Supplementary Fig. 3c–h; and Supplementary Table 2). These data indicate that WOX5 is required for some aspects of quiescentcentre-specific gene expression, but seems not to be a major factor in quiescent centre specification. We asked whether in wox5-1 roots the distal stem cells might have undergone premature differentiation, which in the columella can be visualized by the accumulation of starch-granule-containing organelles involved in gravitropic sensing. Indeed, in wox5-1 roots, the cells at the stem cell position, but not the quiescent centre, accumulate starch granules (Fig. 2e, f; and Supplementary Table 2). In contrast, the proximal meristem appears normal. These findings indicate that WOX5 gene expression in the quiescent centre is required to non-cell-autonomously maintain the distal stem cells as undifferentiated, similar to the role of WUS in the shoot2 . We then investigated whether WOX5 activity is also sufficient to repress columella cell differentiation, by ubiquitously expressing a dexamethasone (DEX)-inducible WOX5 gene. DEX induction results in small cells that lack starch granules in the place of differentiated columella and lateral root cap cells (Fig. 2h, i; Supplementary Fig. 4a; and Supplementary Table 1), indicating that they have not undergone normal differentiation. Unlike quiescent centre cells, the small cells divide, as shown by the continuously increasing cell number and expression of the cell cycle marker CYCB1-GUS (Supplementary Fig. 4b) and quiescent centre markers are not induced (Fig. 2h). Consistent with a loss of differentiated columella cells, 35S[GVG]-WOX5 (35S-WOX5; Supplementary Methods) roots display reduced gravitropism (Supplementary Fig. 4c). After withdrawal of DEX, however, the small cells readily accumulate starch grains typical for differentiated columella cells (Fig. 2l). Normally, laser ablation of the quiescent centre causes differentiation of stem cells9,10 , but this did not have any effect in 35S-WOX5 (Fig. 2j, k). Thus, ectopic WOX5 expression is sufficient to block differentiation of columella stem cell daughters without the requirement of any further quiescent centre signalling. Previous studies highlighted auxin-dependent transcription of the PLETHORA (PLT) genes and SHORTROOT/SCARECROW (SHR/ SCR) functions as essential components in root meristem maintenance and we asked how WOX5 interacts with these pathways9,11–13 . In mutants for the auxin response regulators BODENLOS (BDL) or MONOPTEROS (MP; Fig. 3d–f; and Supplementary Table 3), WOX5 expression is rarely detected, consistent with the notion that BDL/MP-mediated auxin signalling is required for the embryonic initiation of the quiescent centre13,14 . Auxin accumulation in the quiescent centre region provides patterning information to the root meristem mediated through PLT genes9,12,15 . In wox5-1 mutants, auxin accumulation seems normal based on the expression of the auxin response reporter gene DR5-GUS (Supplementary Fig. 5a, b). In plt1 plt2 double mutants, WOX5 expression is occasionally slightly expanded (Fig. 3g–i; and Supplementary Table 3), whereas PLT1 expression is normal in wox5-1 (Supplementary Fig. 5c, d and Supplementary Table 2). The SHR protein moves from provascular cells into the quiescent centre and activates transcription of the SCR gene16 , which is required to specify quiescent centre identity and to maintain stem cells11 . WOX5 expression is reduced or undetectable in shr and scr mutants (Fig. 3j–l; Supplementary Fig. 5g–i; and Supplementary Table 3), whereas an SCR reporter gene is correctly expressed in wox5-1 roots (Supplementary Fig. 5e, f; and Supplementary Table 2). Collectively, our data indicate that WOX5 expression depends on the induction in the root pole by MP-mediated auxin signalling and on SHR/SCR activity, whereas PLT1 and PLT2 have only a minor role in confining WOX5 expression to the quiescent centre. The requirement of WOX5 for only columella stem cell maintenance suggests unidirectionality of this signalling pathway in stem cell control, similar to WUS-mediated stem cell control in the shoot stem cell niche. We investigated whether effects of the wox5-1 mutation on the proximal meristem might be masked by redundant functions. Indeed, in scr-4, shr-1 single mutants and plt1 plt2 double mutants, the wox5-1 mutation promotes differentiation in the proximal region of the root meristem in addition to its effects on distal stem cells (Fig. 3m; and Supplementary Fig. 5j, k). This suggests that WOX5 gene expression in the quiescent centre redundantly contributes to proximal stem cell activity or alternatively that WOX5 has a stemcell-independent function in the proximal meristem, as has been reported for its upstream regulator SCR11 (Supplementary Fig. 7). The striking similarities of WOX5 and WUS expression patterns and functions in root and shoot stem cell niche organizers raised the ih b e a j c f k g d l QC46/+DEX QC184/WT QC CSC COL QC184/wox5-1 QC (CSC) COL QC (CSC) COL QC184/WT } WOX5-GUS/WT QC CSC COL WOX5-GUS/wox5-1 } QC46/–DEX QC184/wox5-1 } +DEX –DEX +DEX } QC QC QC QC Figure 2 | WOX5 represses differentiation in the columella. a–f, wox5-1 heart-torpedo stage embryos (d) and root tips (e) lack QC184 expression, express WOX5-GUS (f) in an enlarged domain and accumulate starch at the CSC position (e, f). g, h, 35S-WOX5 expression gives small cells in the columella, which lack QC46 expression and starch grains (h). j–l, Quiescent centre ablation (arrow) does not induce differentiation (j, k), whereas transfer to non-inducing medium does (l; control, i). Nomarski (a–i, k, l) and confocal (j) median longitudinal images are shown. GUS staining, blue; starch, violet; waved brackets, extra small cells; COL, columella; CSC, columella stem cells; QC, quiescent centre; DEX, dexamethasone. Scale bars, 20 mm (a, d), 50 mm (b, c, e, f, i–l) and 10 mm (g, h). LETTERS NATURE|Vol 446|12 April 2007 812 Nature©2007 Publishing Group question of whether both genes are functionally equivalent. Indeed, expression of the WUS cDNA from the WOX5 promoter completely restores quiescent centre and stem cells in the wox5-1 root meristem (Fig. 4a, b; and Supplementary Fig. 6a, b). Interestingly, ectopic shoot tissue was never observed when WUS was expressed in the quiescent centre. This contrasts to previous studies expressing WUS in a broader range of cell types17 and favours the interpretation that organizer signalling only maintains stem cells as undifferentiated, whereas the fate of differentiating daughters is determined by the tissue context. Conversely, expression of a WUS-WOX5 transgene rescues premature termination of inflorescence meristems and occasionally of floral meristems in the putative null allele wus-1 (Fig. 4c–f; and Supplementary Fig. 6d–g). The rescued wus-1 mutants resemble the weak wus-6/jam allele18 , suggesting that the WUSWOX5 transgene complements WUS function but provides a reduced level of organizing centre signalling. Defective ovule development in wus mutants is not rescued by WUS-WOX5 (Supplementary Fig. 6c), indicating that both regulators are interchangeable in only stem cell maintenance. Significantly, the early embryo genes WOX8 and WOX9 are not able to rescue wus-1 defects (H. Breuninger and T.L., unpublished observations), showing that the ability to regulate stem cell maintenance is not a general property of WOX proteins. 0 5 10 15 20 25 30 j k ih l g m RootHeart embryoGlobular embryo * scr-4 * bdl * plt1 plt2 * Wild type bdl mp Columbia wox5-1 scr-4 wox5-1 scr-4 shr-1 wox5-1 shr-1 plt1 plt2 wox5-1 plt1 plt2 Meristematiccortexcells ca b d e f scr-4 wox5-1 scr-4 3 4 6 3 4 6 3 4 6 3 4 6 days Figure 3 | WOX5 interaction with SHR/SCR and auxin pathways. a–l, Hybridization with WOX5 antisense probe. WOX5 expression was detected in wild type (a–c), absent in bdl and mp (d–f, inset shows sporadic exceptions), deregulated in plt1 plt2 (g–i), and undetectable in scr-4 (j–l, insets show sporadic exceptions). m, wox5-1 enhances differentiation in the proximal meristem in scr-4 (inset), shr-1 and plt1 plt2 mutants. Error bars, s.d. Longitudinal median histological (a, b, d, e, g, h, j, k) and optical (c, f, i, l, m) sections are shown. Four-day-old roots (c, i, l) or middle stage embryos (f) were used. Hybridization signal is red–blue; asterisk, reddish colour is staining independent of hybridization reaction. Arrows, quiescent centre position. Scale bars, 25 mm (a, b, d–f, g, h, j, k) and 50 mm (c, i, l). ba wus-1 WUS-WOX5/wus-1 wus-1 WUS-WOX5/wus-1 st pe pe st WOX5-WUS/wox5-1wox5-1 c d e f Figure 4 | WOX5 and WUS are interchangeable in stem cell control. a, b, WOX5-WUS expression restores quiescent centre and columella stem cells (white arrows) in wox5-1. Violet, starch grains. c–f, Rescue of the shoot meristem stem cell niche. wus-1 shoot (c) and floral meristem (e) fail to maintain stem cells and terminate prematurely. WUS-WOX5 expression restores an indeterminate inflorescence meristem (d) and complete flowers (f). Twenty-six-day-old plants (c, d) and flowers (e, f) are shown. White arrows indicate shoot apex (c, d). pe, petal; st, stamen. Scale bars, 1 cm (c, d), 1 mm (e, f) and 50 mm (a, b). NATURE|Vol 446|12 April 2007 LETTERS 813 Nature©2007 Publishing Group The extent of mechanistic similarity and evolutionary relationship between root and shoot stem cell niches has remained unclear thus far. Here we show that the organizers of both niches employ related regulators to provide stem cell maintaining signals. However, there also seem to be differences between both niches. For example, CLV3 homologues promote meristem differentiation but do not influence stem cell maintenance in the root on overexpression and it is thus unknown whether roots, similar to shoots, employ a WUS–CLV3 related feedback mechanism19,20 . Our results suggest that WOX5 in the root stem cell niche has a more direct function in stem cell signalling, rather than in specifying quiescent centre identity (Supplementary Fig. 7). The nature of the signals that are induced by WUS/WOX5 expression in the stem cell organizers is presently unresolved. WOX5 protein itself or a downstream factor might move to stem cells as the long-postulated shortrange factor5 . To date, WOX5 protein has not been robustly detected, presumably owing to very low expression levels. However, on the basis of localization of a functional WUS–GUS fusion protein, movement of WUS protein out of the organizing centre seems not to be necessary for its function in shoot meristem stem cell maintenance, indicating that factors downstream of WUS act as signals (Supplementary Fig. 6h, i). Palaeobotanical evidence for early root structures is controversial, but the current view is that roots have evolved independently at least twice in Lycophytina and Euphyllophytina21,22 . It remains to be shown whether WUS and WOX5 shared an ancestral function in stem cell control since root and shoot separation, or whether they have been recruited for this function after the diversification of stem cell niches. METHODS Plant work. wox5-1 (SALK038262) and wox5-3 (SALK147644) are transfer (T)DNA insertion alleles in the Columbia (Col) accession and were obtained from the Arabidopsis Biological Resource Center (ABRC, USA) and the Nottingham Arabidopsis Stock Center (NASC, UK). 73DR5-GUS23 was kindly provided by J. Marfett, and mpU51 (ref. 14) by G. Ju¨rgens (Tu¨bingen). Root length and meristem size were measured as described10 . The number of meristematic cells was obtained by counting cortex cells showing no signs of rapid elongation. Quiescent centre laser ablations were performed in 4-day-old roots using a Leica SP2 inverted confocal laser scanning microscope5 . Roots were stained for amyloplasts using Lugol (Sigma) 24-h after ablation24 . For confocal microscopy, root tips were mounted in 10 mg ml21 propidium iodide solution. Microscope settings were as described5 . Expression analysis. Starch granules24 and b-glucuronidase activity7 were visualized as described. 35S-WOX5 seedlings were cultured on standard 0.53 MS medium supplemented with or without 1 mM dexamethasone. In situ hybridization of sections, whole-mount embryos, and 4-day-old seedlings were performed as described1,24 . The PLT1 riboprobe has been described previously12 . The WOX5 riboprobe was prepared from a cDNA without the homeobox to avoid crosshybridization with related mRNAs. Further experimental details are provided as Supplementary Information. Received 9 January; accepted 23 February 2007. 1. Haecker, A. et al. Expression dynamics of WOX genes mark cell fate decisions during early embryonic patterning in Arabidopsis thaliana. Development 131, 657–668 (2004). 2. Mayer, K. F. X. et al. Role of WUSCHEL in regulating stem cell fate in the Arabidopsis shoot meristem. Cell 95, 805–815 (1998). 3. Spradling, A., Drummond-Barbosa, D. & Kai, T. Stem cells find their niche. Nature 414, 98–104 (2001). 4. Laux, T. The stem cell concept in plants: a matter of debate. Cell 113, 281–283 (2003). 5. van den Berg, C., Willemsen, V., Hendriks, G., Weisbeek, P. & Scheres, B. Shortrange control of cell differentiation in the Arabidopsis root meristem. Nature 390, 287–289 (1997). 6. Vernoux, T. & Benfey, P. N. Signals that regulate stem cell activity during plant development. Curr. Opin. Genet. Dev. 15, 388–394 (2005). 7. Schoof, H. et al. The stem cell population of Arabidopsis shoot meristems is maintained by a regulatory loop between the CLAVATA and WUSCHEL genes. Cell 100, 635–644 (2000). 8. Brand, U., Fletcher, J. C., Hobe, M., Meyerowitz, E. M. & Simon, R. Dependence of stem cell fate in Arabidopsis on a feedback loop regulated by CLV3 activity. Science 289, 617–619 (2000). 9. Sabatini, S. et al. An auxin-dependent distal organizer of pattern and polarity in the Arabidopsis root. Cell 99, 463–472 (1999). 10. Wildwater, M. et al. The RETINOBLASTOMA-RELATED gene regulates stem cell maintenance in Arabidopsis roots. Cell 123, 1337–1349 (2005). 11. Sabatini, S., Heidstra, R., Wildwater, M. & Scheres, B. SCARECROW is involved in positioning the stem cell niche in the Arabidopsis root meristem. Genes Dev. 17, 354–358 (2003). 12. Aida, M. et al. The PLETHORA genes mediate patterning of the Arabidopsis root stem cell niche. Cell 119, 109–120 (2004). 13. Weijers, D. et al. Auxin triggers transient local signaling for cell specification in Arabidopsis embryogenesis. Dev. Cell 10, 265–270 (2006). 14. Berleth, T. & Ju¨rgens, G. The role of the Monopteros gene in organising the basal body region of the Arabidopsis embryo. Development 118, 575–587 (1993). 15. Blilou, I. et al. The PIN auxin efflux facilitator network controls growth and patterning in Arabidopsis roots. Nature 433, 39–44 (2005). 16. Nakajima, K., Sena, G., Nawy, T. & Benfey, P. Intercellular movement of the putative transcription factor SHR in root patterning. Nature 413, 307–311 (2001). 17. Gallois, J. L., Nora, F. R., Mizukami, Y. & Sablowski, R. WUSCHEL induces shoot stem cell activity and developmental plasticity in the root meristem. Genes Dev. 18, 375–380 (2004). 18. Hamada, S. et al. Mutations in the WUSCHEL gene of Arabidopsis thaliana result in the development of shoots without juvenile leaves. Plant J. 24, 91–101 (2000). 19. Casamitjana-Martinez, E. et al. Root-specific CLE19 overexpression and the sol1/2 suppressors implicate a CLV-like pathway in the control of Arabidopsis root meristem maintenance. Curr. Biol. 13, 1435–1441 (2003). 20. Hobe, M., Muller, R., Grunewald, M., Brand, U. & Simon, R. Loss of CLE40, a protein functionally equivalent to the stem cell restricting signal CLV3, enhances root waving in Arabidopsis. Dev. Genes Evol. 213, 371–381 (2003). 21. Kenrick, P. & Crane, P. R. The origin and early evolution of plants on land. Nature 389, 33–39 (1997). 22. Kenrick, P. & Crane, P. The Origin and Early Diversification of Land Plants: a Cladistic study (Smithsonian Institution Press, Washington, 1997). 23. Ulmasov, T., Hagen, G. & Guilfoyle, T. J. ARF1, a transcription factor that binds to auxin response elements. Science 276, 1865–1868 (1997). 24. Willemsen, V., Wolkenfelt, H., de Vrieze, G., Weisbeek, P. & Scheres, B. The HOBBIT gene is required for formation of the root meristem in the Arabidopsis embryo. Development 125, 521–531 (1998). Supplementary Information is linked to the online version of the paper at www.nature.com/nature. Acknowledgements We are grateful to E. van der Graaff for the data shown in Supplementary Fig. 4c, and E. Tucker, T. Demlow and B. Geiges for experimental help. We thank G. Ju¨rgens for mp seeds, E. Kiegle and J. Haseloff for mGFP5ER and UAS vectors, M. Terlou for software development for root and meristem measurements, P. Graf for photography and E. Tucker and M. Tucker for comments on the manuscript. This work was sponsored by grants from the European Union (REGIA) and the Deutsche Forschungsgemeinschaft (to T.L.), a Netherlands Genomics Initiative grant (to R.H.), and grants from the Japan Society for the Promotion of Science and the Novartis Foundation (to K.N.). Author Information Sequence of WOX5 mRNA is available in GenBank under accession numbers NM_111961 and AY251398. Reprints and permissions information is available at www.nature.com/reprints. The authors declare no competing financial interests. Correspondence and requests for materials should be addressed to T.L. (laux@biologie.uni-freiburg.de). LETTERS NATURE|Vol 446|12 April 2007 814 Nature©2007 Publishing Group