Post-embryonic plant growth is dependent on a functional shoot apical meristem (SAM) that provides cells for continuous development of new aerial organs. However, how the SAM is dynamically maintained during vegetative development remains largely unclear. We report here the characterization of a new SAM maintenance mutant, sha1-1 (shoot apical meristem arrest 1-1), that shows a primary SAM-deficient phenotype at the adult stage. The SHA1 gene encodes a novel RING finger protein, and is expressed most intensely in the shoot apex. We show that, in the sha1-1 mutant, the primary SAM develops normally during the juvenile vegetative stage, but cell layer structure becomes disorganized after entering the adult vegetative stage, resulting in a dysfunctional SAM that cannot initiate floral primordia. The sha1-1 SAM terminates completely at the stage when the wild-type begins to bolt, producing adult plants with a primary inflorescence-deficient phenotype. These observations indicate that SHA1, a putative E3 ligase, is required for post-embryonic SAM maintenance by controlling proper cellular organization.
The primary shoot apical meristem (SAM) in higher plants is established during embryogenesis, and contributes to the formation of all aerial organs including leaves, stem and flowers in an indeterminate style throughout the lifespan of the plant. In the SAM of Arabidopsis thaliana and other plants, a few long-term stem cells are specified into the three outermost cell layers of the central region of the apex. These self-renewing stem cells are undifferentiated and divide slowly to give rise to daughter cells, some of which retain their undifferentiated state, while others undergo differentiation and give rise to the above-ground plant body.
The functional maintenance of the SAM in Arabidopsis is achieved primarily by the homeodomain transcription factors SHOOTMERISTEMLESS (STM) and WUSCHEL (WUS) through two independent pathways (Clark et al., 1996; Lenhard et al., 2002). STM is a KNOTTED-like homeobox (KNOX) gene that is expressed throughout the SAM, and is required for preventing premature differentiation of meristem cells (Endrizzi et al., 1996). Three other KNOX genes, namely KNAT1, KNAT2 and KNAT6, function together with STM to restrict expression of the genes ASYMMETRIC LEAVES 1 (AS1) and AS2, which are required for organ initiation (Byrne et al., 2000; Semiarti et al., 2001). AS1 and AS2, in turn, repress the expression of the KNOX genes in organ primordia (Carles and Fletcher, 2003).
One of the most frequently asked questions in meristem biology is how the stem cell niche is correctly maintained to coordinate proper SAM function. In Arabidopsis, the answer has largely come from mutational analysis of WUS and CLAVATA3 (CLV3). WUS encodes a member of the WOX family of homeodomain transcription factors and is expressed in a domain underneath the stem cells termed the organizing center (Mayer et al., 1998). In the wus loss-of-function mutant, no self-maintaining stem cells are produced, and the SAM is prematurely terminated (Laux et al., 1996). By contrast, ectopic expression of WUS induces an ectopic stem cell fate (Lenhard et al., 2002), indicating that WUS is required and sufficient to maintain the correct position and number of the overlaying stem cells. CLV3 is a small, secreted peptide that is produced exclusively by the stem cells (Fletcher et al., 1999), and loss of CLV3 function results in ectopic expression of WUS, which shifts upwards by one or two cell layers (Schoof et al., 2000). By contrast, overexpression of CLV3 results in severely reduced transcription of WUS (Brand et al., 2000). These results, combined with functional analysis of CLV1 and CLV2 genes (Clark et al., 1997; Jeong et al., 1999; Kayes and Clark, 1998; Lenhard and Laux, 2003), suggest a dynamic negative feedback loop in which WUS induces the expression of CLV3 in stem cells, and CLV3 in turn presumably interacts with the CLV1–CLV2 receptor kinase complex to downregulate ultimately the expression of WUS in recipient cells (Schoof et al., 2000). Therefore, to understand how the stem cell niche is dynamically maintained, it is important to elucidate how the expression level and domain of WUS are tightly controlled.
So far, several genes have been shown to be involved in SAM maintenance by affecting the WUS/CLV signaling pathway. These include HANABA TARANU (HAN) (Zhao et al., 2004), STIMPY (STIP) (Wu et al., 2005), ULTRAPETALA1 (UTL1) (Carles et al., 2005), SPLAYED (SYD) (Kwon et al., 2005) and APETALA2 (AP2) (Wurschum et al., 2006). However, the mechanisms for dynamic maintenance of the SAM during vegetative development remain largely unclear. In an attempt to gain more information on SAM maintenance, we performed an extensive genetic screen for mutants defective in SAM development. We show here that SHA1, a putative RING finger E3 ligase, is involved in the regulation of post-embryonic SAM maintenance by controlling proper cellular organization.
The sha1-1 mutant shows arrested SAM development during the vegetative stage
From T-DNA insertion lines, we isolated a mutant that showed a SAM-deficient phenotype at the stage when the wild-type began to flower (Figure 1). Because detailed characterization revealed that development of the primary SAM of the mutant was arrested during seedling development, the mutant was designated sha1-1 (for shoot-apical-meristem arrest 1-1).
The sha1-1 mutant showed normal seedling morphology before 5 days after germination (DAG, Figure 2d), although its emerged leaves were somewhat smaller than those of the wild-type (Figure 2a). After 7 DAG, the leaf initiation in sha1-1 (Figure 2d–f) lagged behind the wild-type (Figure 2a–c), and at 13 DAG only about five leaves were visible in the sha1-1 seedlings (Figure 2f) compared with about eight leaves in the wild-type (Figure 2c). However, the sha1-1 mutant showed normal phyllotaxy during vegetative development (Figure 2e,f,h). The wild-type began to bolt at 20 DAG (Figure 2g), whereas no primary inflorescence formed in the sha1-1 plants (Figure 2 h,i). In contrast to the aerial parts, root development was not affected in the sha1-1 mutant (data not shown).
Observations by scanning electron microscope (Figure 3) and longitudinal sections (Figure 4) revealed details of the deficiencies in SAM structure in the mutant during phase changes. Before 5 DAG, the sha1-1 SAM was the same size and dome shape as the wild-type SAM, and the size of initiated leaf primordia in sha1-1 was comparable to that in the wild-type (Figure 3a–d). Significant differences between sha1-1 (Figure 3f) and wild-type (Figure 3e) in terms of SAM morphology were first observed at 7 DAG, indicating that development of the primary SAM of sha1-1 was arrested (Figure 3f), in contrast to the rapid increase seen in the size of the dome-shaped SAM in the wild-type (Figure 3e). The number and phyllotaxy of leaf primordia that formed in sha1-1 after this stage were almost normal, but their sizes were obviously smaller than those of the wild-type (Figure 3e–h), consistent with the slower rate of leaf initiation in sha1-1 seedlings (Figure 2e,f).
Furthermore, the wild-type SAM showed characteristic L1, L2 and L3 cell layers, and anticlinal cell division occurred in L1 and L2 cells (Figure 4a,c,e,g,i). The sha1-1 mutant showed the same SAM cellular structure as the wild-type at 5 DAG (Figure 4b). However, the SAM structure of sha1-1 became disorganized after 6 DAG (Figure 4d, f, h, j, l). In contrast to the wild-type, the plane of cell division in the L2 cells of sha1-1 SAM began to adopt an abnormal orientation at 6 DAG, although the L1 layer appeared normal at this stage (Figure 4d,f). At 7 DAG, cells in the L1 layer of sha1-1 SAM also became irregularly orientated (Figure 4h), and the cell layer structure was completely disturbed at 10 DAG (Figure 4j).
The wild-type SAM began to initiate the floral meristem (FM) at 9 DAG (arrowhead in Figure 3g), and multiple floral primordia were observed surrounding the SAM after 11 DAG (Figure 3i,k). Although the primary SAM of sha1-1 still maintained a dome-shape at 15 DAG (Figure 3l), the SAM was obviously dysfunctional because no FM or stem were formed during all the observed stages (Figures 3 h,j,l,m and 4j,l,m). At 20 DAG, the primary SAM of sha1-1 completely terminated (Figure 3m) or showed only a disorganized structure with small meristematic cells remaining in the uppermost region (Figure 4m). Some sha1-1 mutants (about 40%) developed no ectopic meristems at this stage, while the others (60%) formed a few ectopic meristems around the arrested primary SAM (data not shown).
Under our growth conditions, the sixth leaf of the wild-type that appeared at 11 DAG was the last one that was characteristic of a juvenile vegetative (JV) leaf based on their trichome distribution (data not shown, Telfer et al., 1997). We therefore infer that SHA1 functions in SAM maintenance from the adult vegetative (AV) stage onwards.
The sha1-1 mutant shows reduced expression of WUS and CLV3, but not STM
The SAM of Arabidopsis is maintained primarily by WUS and STM through two independent pathways. To investigate which pathway is affected in the sha1-1 mutant, we used RT-PCR to analyze the transcription levels of genes known to be related to SAM maintenance. No obvious difference was found in the transcription levels of any of the meristem regulators tested in sha1-1 and wild-type plants at 5 DAG (Figure 5a), which is in agreement with the above observations that the sha1-1 mutant developed a normal SAM before this stage. In plants at 15 DAG, messenger levels of CLV1, CLV2, STM and KNAT1 were not significantly altered in the mutant (Figure 5a). By contrast, transcripts of WUS and CLV3 were not detected in the sha1-1 mutant (Figure 5a), indicating that the WUS/CLV3 pathway is affected in this mutant.
The implication that SHA1 affects the WUS pathway was further confirmed by GUS reporter gene analysis. In 10 DAG seedlings carrying the STM::GUS reporter gene, there was no obvious difference in expression between the sha1-1 mutant and the wild-type (Figure 6a,b). By contrast, whereas intense staining was detected in the shoot apex of the wild-type seedlings carrying the WUS::GUS reporter (Figure 6c), staining was absent or only very faint in the sha1-1 seedlings under the same conditions (Figure 6d).
The LEAFY (LFY) gene is expressed in newly emerging leaf primordia and is rapidly up-regulated after initiation of floral primordia (Blazquez et al., 1997). Thus, LFY represents an important marker of the transition from the vegetative to the reproductive phase. As described above, the SAM of sha1-1 had failed to initiate floral primordia even by the time that the wild-type began to bolt (Figures 2 g,h and 3g–m), a feature that should be reflected in the expression pattern and intensity of LFY. Indeed, intense GUS signal was readily detected in the presumptive flower primordia in LFY::GUS-expressing wild-type plants at 15 DAG (Figure 6e), the stage at which multiple floral primordia have been initiated (Figure 3k). In the sha1-1 mutant, however, GUS staining was detected only in the younger leaf primordia and was completely absent from the central region of the shoot apex (arrow in Figure 6f), in agreement with the morphological observations that the sha1-1 mutant failed to enter the reproductive phase at 15 DAG (Figure 3l).
SYD has been reported to directly target WUS for transcriptional regulation (Kwon et al., 2005). In the syd mutant, there is a reduction in WUS expression at the stage when phenotypic defects in the SAM are first observed. To verify whether this reduction also occurs in the sha1-1 mutant, we analyzed WUS expression at various developmental stages. Unlike syd, the sha1-1 mutant showed a transcription level of WUS similar to that of the wild-type at 7 DAG (Figure 5b), the stage when SAM development was arrested in the mutant (Figure 3f). By contrast, transcription of WUS was greatly reduced in the mutant after 8 DAG (Figure 5b).
WUS is epistatic to SHA1 in SAM maintenance
To evaluate further the relationship between SHA1 and WUS, we generated a sha1-1 wus double mutant. In the single wus mutant, one to four leaf primordia have initiated by about 14 DAG, and further leaf primordia initiation occurs after about one week’s interruption (Laux et al., 1996). The phenotype of all 12 sha1-1 wus double mutant plants isolated was indistinguishable from that of the wus single mutant (Figure 7a–f) up to 20 DAG, the stage at which about nine leaves were initiated in sha1-1 (Figure 2h). Observation of SAM morphology in plants at 10 DAG revealed that the SAM of the sha1-1 wus mutant was terminated completely at this stage (Figure 7j), which is similar to the wus single mutant (Figure 7i). By contrast, a dome-shaped SAM structure was still found in sha1-1 plants (Figure 7h). This observation indicates that WUS is epistatic to SHA1 in SAM maintenance. At 30 DAG, the wus mutant has initiated numerous leaf primordia. By contrast, few leaf primordia were initiated in the double mutant (data not shown). This may be explained by the observation that, although both sha1-1 wus and wus showed terminated SAM morphology, the central region of the shoot apex in sha1-1 wus plants was much more severely disturbed compared with that in wus (Figure 7i,j). This result suggests that SHA1 has an enhancing effect on WUS during vegetative SAM maintenance.
SHA1 encodes a novel RING finger protein
sha1-1 was isolated as a T-DNA insertion mutant, and the backcrossed BC2F2 population segregated in a 3:1 ratio (wild-type:sha1-1 mutant), in which the sha1-1 phenotype co-segregated with the T-DNA insert. This result indicates that the recessive trait of sha1-1 is caused by a single T-DNA insertion. To isolate the sequence flanking the T-DNA insertion site in the mutant, we performed TAIL-PCR as previously described (Liu et al., 1995). Sequencing the resulting PCR fragments for the sha1-1 mutant revealed that the two T-DNA fragments were inserted in the promoter of the gene At5g63780 (Figure 8a,b). The SHA1 gene encodes a putative RING domain-containing protein of 368 amino acids (Figure 8d), with a predicted size of 40.5 kDa. No other characterized domains are present in SHA1. A comparison of the predicted SHA1 amino acid sequence with sequences in the current protein databases (NCBI; BLAST network server) showed that SHA1 exhibits 71% similarity to another Arabidopsis gene At5g08750 (Figure 8d), which we named SHA2. The predicted SHA1 gene product also shows 45% identity to the rice gene OsSHA (accession number XP_474481) across the whole length of the protein (Figure 8d). Because we found no ortholog of SHA1 in other organisms such as human or rat, we infer that the SHA (SHA1/SHA2) gene is plant-specific.
RT-PCR analysis using SHA1-specific primers showed that transcription of SHA1-1 was extremely reduced in the mutant plants (Figure 8e). To verify that reduction of the SHA1 transcript (Figure 8e) was responsible for the mutant phenotype in sha1-1 plants, we examined an additional allele, sha1-2, that showed T-DNA insertions in the promoter of the gene (Figure 8a,c). The sha1-2 mutant also showed the same SAM arrest phenotype as the sha1-1 mutant (data not shown), indicating that insertion of T-DNA into the promoter of At5g63780 is responsible for the sha1 mutant phenotype.
Investigation of tissue-specific expression revealed that SHA1 is transcribed ubiquitously, with the highest level in inflorescences (data not shown). Further analysis using promoter::GUS fusions confirmed that native promoter for SHA1 is expressed most intensely in the shoot apex (Figure 9a,c). Observation by longitudinal section revealed that SHA1 is expressed in the base of stem and young leaf primordia, and also showed circular-strip expression around the rib zone (Figure 9e). Deletion of the promoter in the sha1-1 mutant led to complete disappearance of the GUS signal from the apex, and GUS activity could be detected only in the embryonic-derived cotyledons (Figure 9b,d). These observations agree well with the above finding that SHA1 functions in post-embryonic SAM maintenance, and also explained the residual expression of SHA1 in the mutant when cotyledon-contained whole plants were used (Figure 8c).
We introduced the antisense SHA1 CDS into wild-type plants to verify whether SHA1 is the causal gene of the sha1-1 phenotype, antisense SHA1 transgenic lines were constructed. Out of about 120 recovered T1 plants, we identified 15 independent transformants that, although varying in severity from line to line, had a phenotype of altered phyllotaxy at the seedling stage (Figure 10c,d), indicating a defect in SAM function. Analysis of SHA1 expression revealed that the SHA1 mRNA level decreased with the phenotypic severity in the antisense transgenic lines (Figure 10e). At later growth stages, the transgenic lines with weak phenotype could form a primary inflorescence, although this occurred about two weeks later (data not shown). By contrast, the lines showing a severe phenotype at the seedling stage developed an adult plant closely resembling that of the sha1-1 mutant (Figure 10f–i). Therefore, we conclude that SHA1 is the causal gene for the SAM arrest phenotype in sha1-1.
SHA1 is required for vegetative SAM maintenance
We present here the role of a RING domain-containing protein, SHA1, in post-embryonic SAM maintenance in Arabidopsis. We have shown that the primary SAM in the sha1-1 mutant develops normally during the JV stage, but the cellular structure of the SAM becomes disorganized from the AV stage onwards, resulting in plants with a completely dysfunctional SAM that cannot initiate primary inflorescences and floral primordia. These results indicate that SHA1 is required for vegetative SAM maintenance. Although ectopic meristems are able to form around the terminated SAM at later growth stages, these meristems cannot function normally, as assessed by the abnormal phenotypes observed in adventitious shoots and flowers (data not shown), suggesting that, in addition to its role in development of the SAM, SHA1 also functions in FM development.
SHA1 affects the WUS pathway in SAM maintenance
The post-embryonic development of higher plants is dependent on the maintenance of a functional SAM. In Arabidopsis, this maintenance is achieved primarily by the homeodomain transcription factors WUS and STM through two independent pathways (Carles and Fletcher, 2003). In this study, we have provided several lines of evidence that SHA1 specifically affects the WUS pathway in post-embryonic SAM maintenance. First, transcription of WUS, but not of STM, was extremely reduced in the mutant (Figure 5a). Second, STM::GUS was expressed normally in the SAM of sha1-1, whereas WUS::GUS expression was absent from sha1-1 seedlings in comparison to wild-type seedlings (Figure 6). Last, the sha1-1 wus double mutant showed a phenotype similar to that of the wus single mutant (Figure 7), suggesting that both genes act in one pathway.
SHA1 affects WUS expression indirectly by controlling SAM cellular organization
The sha1-1 mutant shows down-regulated WUS expression during vegetative SAM development. Because the sha1-1 mutant showed WUS expression comparable to that of the wild-type at the stage of SAM arrest (Figures 3f and 5b), the SAM arrest phenotype of sha1-1 is not caused by the reduced WUS transcription level. We propose that the down-regulated WUS transcription in the sha1-1 mutant is an indirect consequence of meristem degeneration.
The SAM of Arabidopsis is partitioned into radial domains comprising discrete cell layers and concentric zones, and characterization of the WUS–CLV feedback interaction loop indicates that integration of meristem development requires short-range intercellular signals between these different layers. In addition, analysis of mutants with a disorganized SAM structure further shows that the organized SAM cellular arrangement is critical for the proper cell-to-cell communication that is required for tight control of WUS expression (Kaya et al., 2001; Suzuki et al., 2004). In this study, we showed that loss of SHA1 function caused improper cellular organization in the SAM after 6 DAG (Figure 4). Since correct cellular organization is needed for proper cell-to-cell communication and cell fate decision, disruption of the SAM cellular structure in sha1 would result in failure of the cells at OC to perceive the signalling from other parts of the SAM to stably maintain transcription of WUS. As a result, cells in the WUS expression domain are unable to maintain their identity at late SAM developmental stage, resulting in a reduction of WUS transcription that lags behind the phenotypic defects of the SAM (Figures 3f, 4 h and 5b). In contrast to WUS, transcription of STM is regulated via a different pathway by regulators such as CUC and AS1/AS2 (Aida et al., 1999; Byrne et al., 2000; Semiarti et al., 2001). These regulators remain unaltered in the sha1-1 mutant (data not shown). The reduction the signalling in WUS expression level results in fewer stem cells being specified at the very apex, and thus fewer daughter cells being available for amplification by STM, leading to the slower formation of leaf primordia observed in the mutant (Figures 2 and 3). Formation of the last two leaves at 15 DAG (Figure 3l) consumed all of the remaining SAM cells, resulting in sha1-1 plants with the SAM-deficient phenotype (Figures 1c and 3m).
Further analysis of the temporal relation between meristem termination and changes in WUS expression is needed to clarify the relationship between SHA1 and WUS in vegetative SAM maintenance.
Possible mechanisms of SHA1 function in SAM maintainance
The SHA1 gene encodes a putative RING domain-containing ubiquitin–protein E3 ligase. In Arabidopsis, about 90% of the components of the ubiquitin/26S proteasome pathway have been revealed to encode subunits of the E3 ligases, which confer substrate specificity and have been shown to be involved in many plant developmental processes such as embryogenesis, photomorphogenesis, hormone signaling and disease resistance (Vierstra, 2003). Some ligase-encoding genes such as RHA2b (Lechner et al., 2002) or BIG BROTHER (Disch et al., 2006) have been found to be intensely expressed in meristematic regions, suggesting that they have a role in meristem function. To our knowledge, however, there is no evidence so far of the involvement of E3 ligases in the regulation of SAM maintenance. The RING variant domain contained in SHA1 is a C4HC3-type RING finger motif (C-x·2-C-x·10-45-C-x·1-C-x·7-H-x·2-C-x·11-25-C-x·2-C) that is found in a number of cellular and viral proteins, some of which have been shown to have ubiquitin E3 ligase activity both in vivo and in vitro (Coscoy et al., 2001; Dodd et al., 2004; Hassink et al., 2005; Swanson et al., 2001). In fact, most of the RING and modified RING proteins in Arabidopsis have been revealed to have ligase activity (Stone et al., 2005). From this point of view, we propose that SHA1 acts as an E3 ligase to specifically target an as yet unknown negative regulator of SAM maintenance for breakdown through the 26S proteasome pathway. On contrary, the RING domain could also function in mediating protein–protein or protein–DNA interactions, and RING domain-containing proteins without E3 ligase activity are known to be involved in a wide range of physiological and cellular processes such as development (Borden, 2000). Interestingly, the SHA1 gene is predicted to encode a membrane protein with three transmembrane domains, suggesting a role in cell-to-cell communication. Therefore, we do not exclude the possibility that SHA1 may, either acting independently or by forming large complex through the RING domain, directly regulate SAM development.
Plant materials and growth conditions
The wild-type Columbia-0 and all other materials used in this study were grown under the conditions described by Morita-Yamamuro et al. (2005). The wus null allele was obtained from SAIL (accession number: SAIL_150_G06), http://signalsalk.edu/cgi-bim/tdnaexpress?GENE=&FUNCTION=&TDNA=SAIL_I50-G06. Double mutant plants were genotyped by the gene-specific primers listed in Table 1. Transgenic Arabidopsis seeds harboring STM::GUS and WUS::GUS were kindly provided by Dr M. Tasaka (Graduate School of Biological Sciences, Nara Institute of Science and Technology, Japan) and Dr T. Laux (Institute of Biology III, Freiburg University, Freiburg, Germany), and introgressed into the sha1-1 mutant background.
Table 1. Primers used in this study.
sha1;1, sha1;2 LB
SHA1;1, SHA1;2 LP
SHA1;1, SHA1;2 RP
For promoter::GUS construction
The sha1-1 mutant was isolated by screening T-DNA insertion lines provided by RIKEN Yokohama Institute (Yokohama, Japan). To determine the genomic sequence flanking the T-DNA insertion, thermal asymmetric interlace (TAIL)-PCR was performed as described previously using T-DNA-specific primers (Liu et al., 1995; Morita-Yamamuro et al., 2005). The PCR fragments obtained were directly sequenced or cloned into the pCR2.1 vector (http://www.invitrogen.co.jp/) and then sequenced. The sequences bordering the T-DNA were determined using the primer pairs listed in Table 1.
All above-ground plant tissues were used for RNA isolation. RNA isolation, reverse transcription and PCR conditions have been described previously (Morita-Yamamuro et al., 2005). The number of PCR cycles varied from 25 to 35 depending on the mRNA level. The PCR primers used are listed in Table 1. PCR products were electrophoresed on an agarose gel and visualized by ethidium bromide staining.
Histology, GUS staining and scanning electron microscopy
Antisense SHA1 transformation was performed essentially according to the protocol supplied with the GATEWAY system (http://www.invitrogen.co.jp/gateway). In brief, SHA1 CDS was first introduced into the TOPO cloning vector pENTRTM/D-TOPO in an antisense orientation, by using specifically designed primers as described by the manufacturer. After verifying the orientation of the inserted SHA1 sequence in the plasmid, the antisense SHA1 CDS was further transformed into the binary vector pGWB2 by an LR reaction (Morita-Yamamuro et al., 2005). The pGWB2 vector carries the CaMV 35S promoter and directs expression of the CDS in the antisense orientation. The fusion plasmid was transformed into Agrobacterium tumefaciens GV2260 by electroporation, and then introduced into wild-type Arabidopsis plants using the Agrobacterium-mediated transformation method described previously (Clough and Bent, 1998). To select for transgenic progeny, T1 seeds from primary transformants were planted on germination medium containing 50 µg l−1 kanamycin and 50 µg l−1 hygromycin.
Promoter::GUS fusion construct
Genomic fragments of the full-length promoter region of the SHA1 gene were amplified from genomic DNA of wild-type Columbia using the primer sets listed in Table 1. Construction of a promoter::GUS fusion protein was performed by the method described above, except that the destination vector used was pGWB3, a binary vector in which the GUS gene is fused at the C-terminus and there is no promoter.
We are grateful to Mr Wataru Kato and Akira Iwata for helpful comments and technical assistance. We thank Dr T. Laux (Freiburg University) and Dr M. Tasaka (Nara Institute of Science and Technology) for providing WUS::GUS and STM::GUS, respectively, and the Arabidopsis Biological Resource Center for providing the wus mutant (SAIL_150_G06) and sha1-2 (CS851103). This work was supported by Grants-in-Aid for Scientific Research (numbers 17370011, 17051001 and 17780001), CREST (Core Research for Evolutional Science and Technology) of the Japan Science and Technology Corporation, and partially by the 21st century COE (Center of Excellence) Hokkaido University (to J.Y. and to Y.S. as a postdoctoral fellowship). S.-G.Y. acknowledges a fellowship from the Japan Society for the Promotion of Science (16-04445: 2004–2006).