Role of the ZWILLE gene in the regulation of central shoot meristem cell fate during Arabidopsis embryogenesis
1998; Springer Nature; Volume: 17; Issue: 6 Linguagem: Inglês
10.1093/emboj/17.6.1799
ISSN1460-2075
Autores Tópico(s)Plant tissue culture and regeneration
ResumoArticle16 March 1998free access Role of the ZWILLE gene in the regulation of central shoot meristem cell fate during Arabidopsis embryogenesis Bernard Moussian Bernard Moussian Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany Search for more papers by this author Heiko Schoof Heiko Schoof Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany Search for more papers by this author Achim Haecker Achim Haecker Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany Search for more papers by this author Gerd Jürgens Gerd Jürgens Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany Search for more papers by this author Thomas Laux Corresponding Author Thomas Laux Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany Search for more papers by this author Bernard Moussian Bernard Moussian Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany Search for more papers by this author Heiko Schoof Heiko Schoof Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany Search for more papers by this author Achim Haecker Achim Haecker Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany Search for more papers by this author Gerd Jürgens Gerd Jürgens Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany Search for more papers by this author Thomas Laux Corresponding Author Thomas Laux Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany Search for more papers by this author Author Information Bernard Moussian1, Heiko Schoof1, Achim Haecker1, Gerd Jürgens1 and Thomas Laux 1 1Lehrstuhl für Entwicklungsgenetik, Auf der Morgenstelle 1, D–72076 Tübingen, Germany *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:1799-1809https://doi.org/10.1093/emboj/17.6.1799 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Postembryonic development in higher plants is marked by repetitive organ formation via a self-perpetuating stem cell system, the shoot meristem. Organs are initiated at the shoot meristem periphery, while a central zone harbors the stem cells. Here we show by genetic and molecular analyses that the ZWILLE (ZLL) gene is specifically required to establish the central–peripheral organization of the embryo apex and that this step is critical for shoot meristem self-perpetuation. zll mutants correctly initiate expression of the shoot meristem-specific gene SHOOT MERISTEMLESS in early embryos, but fail to regulate its spatial expression pattern at later embryo stages and initiate differentiated structures in place of stem cells. We isolated the ZLL gene by map-based cloning. It encodes a novel protein, and related sequences are highly conserved in multicellular plants and animals but are absent from bacteria and yeast. We propose that ZLL relays positional information required to maintain stem cells of the developing shoot meristem in an undifferentiated state during the transition from embryonic development to repetitive postembryonic organ formation. Introduction Stem cells in the center of the shoot meristem are the ultimate source from which all tissues of the growing shoot are derived (Barlow, 1978; Clark, 1997; Meyerowitz, 1997; Laux and Mayer, 1998). They are considered undifferentiated in the sense that they lack morphological features associated with cells in mature tissue, such as a large central vacuole. Clonal analyses suggest that the pluripotent shoot meristem stem cells are specified in response to positional cues (Ruth et al., 1985). Differentiating daughter cells enter specific developmental pathways according to their positions and are incorporated into organ primordia in the peripheral zone. Several genes have been shown to regulate the maintenance of the stem cell population and thus the indeterminate nature of the shoot meristem itself. The first specific regulator identified was the maize KNOTTED1 (KN1) gene, which encodes a homeodomain protein that promotes meristem cell fate in a non-cell-autonomous manner (Vollbrecht et al., 1991; Smith et al., 1992; Sinha et al., 1993). The related SHOOT MERISTEMLESS (STM) gene from Arabidopsis is required to prevent stem cells from being incorporated into organ primordia, and thus from differentiating (Clark et al., 1996; Endrizzi et al., 1996; Long et al., 1996). The WUSCHEL (WUS) gene is necessary for cell identity in the meristem center and wus mutations lead to termination of meristem activity (Laux et al., 1996). While WUS and STM are required for shoot and floral meristem activity throughout development, mutations in the TERMINAL FLOWER gene specifically result in the conversion of the inflorescence meristem center into determinate flowers (Alvarez et al., 1992). The CLAVATA (CLV) genes promote organ formation and/or regulate meristem cell proliferation and appear to have antagonistic roles to the genes mentioned above (Clark, 1997). STM and CLV1 seem to competitively regulate the balance between undifferentiated cells and organ formation in response to positional information (Clark et al., 1996; Laux and Schoof, 1997). CLV1 encodes a putative membrane-bound receptor kinase, suggesting a function in a signaling pathway (Clark et al., 1997). Genetic analysis suggests that WUS is a putative target for STM and CLV1 regulation (Endrizzi et al., 1996; Laux et al., 1996). Development of the primary shoot meristem is initiated during the establishment of the basic body organization of the embryo (Laux and Jürgens, 1997) as visualized by transcriptional activation of the STM gene in central apical cells of the globular embryo (Long et al., 1996). Histological analyses of organ initiation (Spurr, 1949; Kaplan, 1969) and the analogous defects of cotyledon and leaf formation in the stm mutant (Endrizzi et al., 1996) suggest that at least some similar mechanisms are involved during the initiation of cotyledons at the globular embryo stage, and leaf initiation in postembryonic shoot meristems. However, there are also differences suggesting that the apical domain of the globular embryo is not fully equivalent to a shoot meristem (Laux and Mayer, 1998). For example, the initiation of leaves but not cotyledons is affected by mutations in the WUS (Laux et al., 1996) or ZLL genes (Jürgens et al., 1994; Endrizzi et al., 1996). While recent genetic and molecular analyses have provided models for the way in which the shoot meristem is maintained (Clark, 1997; Meyerowitz, 1997; Laux and Mayer, 1998), little is known about how the organization and regulatory networks that govern shoot meristem self-perpetuation are established during embryogenesis. Here we report the isolation of the ZLL gene and an analysis of its role. Functional analyses of zll mutants and gene expression data both indicate that ZLL is required for stem cell fate within the developing embryonic shoot meristem. Results In order to study early events in shoot meristem development, we analyzed 16 EMS (ethyl methanesulfonate)-induced mutant alleles of the ZLL gene (Jürgens et al., 1994). All zll mutants showed indistinguishable seedling phenotypes that suggested a defect in embryonic shoot meristem establishment. Complementation analyses between zll-3 and the similar mutant pinhead (pnh) (McConnell and Barton, 1995) demonstrated that the two mutations are allelic (data not shown). Unless stated otherwise, we used zll-3 for our analyses. zll embryos form differentiated structures instead of a shoot meristem In mature wild-type embryos, the shoot meristem can be recognized as a ridge of cells, five to six cells long and two cells wide (Figure 1A), flanked by the first two true leaf primordia which have arisen from it. In contrast, mature zll embryos displayed a single bulge of cells, three to nine cells across, spanning the whole apex (Figure 1B). In some cases, the cell number was reduced and the shape of the apex was flat (Figure 1C). Figure 1.Embryonic and seedling phenotype of zll mutants. (A–C) Scanning electron microscopy of mature embryo apices where one cotyledon has been removed (*). (A) In wild-type, the shoot meristem (arrow) has initiated the first two leaf primordia, p. In zll embryos, either a single bulge (B, arrow) or a flat apex (C, arrow) has formed instead of a shoot meristem. (D–G) 10-day-old seedlings. (D) In wild-type, the shoot meristem has initiated a rosette of leaves, l. zll seedlings either display an empty apex (E, arrow), a filamentous structure (F, arrow) or a solitary leaf (G, l) instead of a shoot meristem. c, cotyledon. Each bar represents 50 μm in A–C, 1 mm in D–G. Download figure Download PowerPoint After germination, the shoot meristem in wild-type seedlings continually initiates new organs and gives rise to a rosette of leaves (Figure 1D). zll seedlings, in contrast, showed an 'empty', flat apex (Figures 1E and 2A), solitary filamentous structures (Figure 1F) or leaves (Figure 1G) instead of a shoot meristem. A similar range of seedling phenotypes has been described for the pnh mutant (McConnell and Barton, 1995). In zll seedlings, the orientation of solitary leaves relative to the cotyledons was variable, in contrast with wild-type. Thus, no regular wild-type leaf was formed, indicating that zll shoot meristem development was disrupted before the initiation of the first true leaves. Infrequently, intermediate forms between filamentous structures and leaves, radially symmetric or partially fused leaves, were observed (data not shown). The number of seedlings with central structures correlated with the number of embryos with bulged apices, suggesting that an apical bulge represents a central structure primordium. All other parts of the seedling, such as the cotyledons, shoot axis and the root meristem, were not visibly affected in zll mutants, suggesting that ZLL function is required specifically in the embryo for shoot meristem development. Figure 2.Postembryonic zll development. (A–C, F–K) Scanning electron microscopy images, (A–G and I) zll, (H and K) wild-type. zll seedlings lacking a primary shoot meristem (A) initiate adventitious shoot meristems at the base of the cotyledons [(B) and (C), arrows; range corresponds to rectangle in (A)]. (D) In longitudinal median sections the differentiated central apical cells are still present (arrow) after initiation of ASMs (asm). (E) DAPI-stained seedlings suggest that ASMs (arrow) were initiated at the petiole of the cotyledons. (F) The elongated ASM (arrow) has initiated supernumerary leaf primordia at its periphery. (G) The zll inflorescence meristem, im, is fasciated and gives rise to supernumerary floral meristems, fl. (H) Wild-type comparison with (G). (I) Occasionally, zll adventitious fasciated shoots fail to initiate axillary shoot meristems (arrow). (K) Wild–type comparison with (I). c, cotyledon; f, filamentous structure; l, leaf; s, stem; v, vasculature. Each bar represents 50 μm. Download figure Download PowerPoint zll mutations affect developmental decisions in the embryo apex Occasionally, individual zll embryos initiated two leaf primordia at the periphery of the apex as in wild-type embryos. In such 'escapes', a functional primary shoot meristem was formed, in contrast with zll seedlings displaying the range of phenotypes described above (Table I). Additional observations suggested that different zll alleles, although they produced qualitatively indistinguishable seedling defects, differed in the frequency of escape seedlings. The reason for this quantitative difference is unclear, but the parental genotype did not play a role and reciprocal crosses between pairs of zll alleles resulted in transheterozygous progenies that displayed an escape frequency similar to or intermediate between their parents, suggesting that different alleles reflect different doses of ZLL activity (data not shown). Table 1. Primary shoot meristem formation in zll embryos zll seedling phenotypes E F L 2L Seedlings (n) 126 180 68 63 PSM (%) 0 0 0 98 In zll seedlings that displayed an empty apex (E), a central filamentous structure (F), a solitary leaf (L) or at least two leaves (2L), the percentage of plants that showed a primary shoot meristem (PSM) is given. zll seedlings form adventitious shoot meristems Although zll mutants were defective in embryonic shoot meristem development, they could initiate de novo adventitious shoot meristems (ASM) postembryonically in the axils of the cotyledons (Figure 2B–F). In seedlings with either a flat apex or with a radially symmetric central structure, ASMs were formed on average 10 days after germination in both cotyledonary axils (Figure 2C–E). In contrast, in seedlings with a bilaterally symmetric central leaf, ASMs were initiated 5 days later on average and were restricted to the cotyledonary axil facing the upper side of the leaf (data not shown). Histological analyses and staining of nuclei showed that ASMs were positioned at the base of cotyledonary petioles (Figure 2D–E), indicating that the ASMs were not simply lagging primary shoot meristems. zll ASMs correctly initiated leaves at the periphery, but compared with wild-type shoot meristems they were larger and elongated (Figure 2F and G). This resulted in fasciated shoots giving rise to leaves and floral meristems in an abnormal arrangement (Figure 2G). Occasionally, zll adventitious shoots did not initiate meristems in the axils of cauline leaves (Figure 2I), similar to the pnh mutant (McConnell and Barton, 1995). However, we did not observe this defect in the primary shoots of escape zll plants, suggesting that it is a secondary effect of the fasciated shoot architecture. Thus, ZLL seems not to be necessary for shoot meristem formation or function per se, but to be specifically required for establishing a functional primary shoot meristem in the embryo. ZLL is required to prevent cell differentiation in the center of the embryonic apex In order to study the cellular basis of the zll defect, we analyzed embryo and seedling development at the histological level. During early stages of embryogenesis (globular stage to torpedo stage) no differences were found between wild-type and zll (data not shown). In median sections the shoot meristem can be recognized as a group of small, densely staining cells with large nuclei, which are located between the cotyledons in the mature embryo (Figure 3A) and between the leaf primordia in the seedling (Figure 3B). The surrounding differentiated cells are larger and stain less densely, and their central vacuole occupies most of the cell lumen. Figure 3.Cell types are altered in zll apices. 1 μm longitudinal median sections of mature embryo (A and C) or seedling (B and D) apices. In wild–type (A and B), the shoot meristem displays small, densely staining cells (arrows). Apical cells in zll mutants (C and D) appear to have initiated differentiation (arrows); they are larger and stain less densely than the corresponding wild-type cells. c, cotyledon; l, leaf; p, primordium; v, vasculature. Each bar represents 20 μm. Download figure Download PowerPoint In contrast, we did not detect meristem cells in mature zll embryos or in zll seedlings. The apices of zll embryos were variably shaped, and the cells had neither prominent nuclei, nor did they stain more intensively than surrounding differentiated cells (Figure 3C). In zll seedlings, the cells in the apex were larger and more vacuolated than wild-type meristem cells, suggesting that these cells had initiated differentiation (Figure 3D). In contrast with the shoot meristem, root meristem organization appeared to be normal in zll mutants (data not shown), suggesting that the two meristems are regulated by different processes. ZLL regulates the spatial pattern of STM expression late in embryogenesis In order to determine the stage at which ZLL is required, we analyzed the expression of STM as a molecular marker for shoot meristem cell identity in zll embryos compared with wild-type. STM expression is confined to central apical cells of wild-type embryos from the globular stage onwards and is downregulated in the organ anlagen at the periphery of the shoot meristem, but maintained in the center, during postembryonic development (Long et al., 1996). STM expression appeared not to be affected in zll mutants during early stages of embryogenesis (Figure 4A, B, E, and F), suggesting that the shoot meristem program was initiated correctly. However, at later embryo stages the spatial expression pattern of STM was markedly altered: expression was downregulated in the center of the apex (Figure 4G), in contrast with wild-type (Figure 4C), and became confined to a small group of lateral cells. This downregulation of central STM expression is consistent with the differentiation of the zll embryo apex and may explain why zll seedlings occasionally mimic a weak stm phenotype (Endrizzi et al., 1996) and produce fused leaves at the position of the shoot meristem. STM expression appeared to be normal in zll mutants during postembryonic development (Figure 4D and H), as expected from the mutant phenotype. Figure 4.Expression patterns of STM in wild-type and zll mutants. In situ hybridization of longitudinal median sections of wild-type (A–D) and zll (E–H) with STM antisense RNA. (A and E) early heart stage embryo; (B and F) torpedo stage embryo. STM expression (arrows) is confined to apical cells between the cotyledonary primordia, c, both in wild-type (A and B) and in zll (E and F) embryos. (C and G) Bent cotyledon stage embryo. In wild-type (C), STM is expressed in a single group in the center of the shoot meristem primordium (arrow). In zll embryos (G) the expression is reduced in the central part of the apex (arrow), but instead is present in two patches close to the axils of the cotyledons, c. (D and H) Inflorescences. STM is expressed in the center of both the inflorescence meristem, im, and floral primordia, fl, in wild-type (D) and in zll (H). Tissue was collected, fixed and processed under the same conditions. Each bar represents 10 μm. Download figure Download PowerPoint Differentiation of central apical cells in zll embryos is not affected by clv mutations CLV and STM antagonistically regulate the balance between undifferentiated cells and cell differentiation in shoot and floral meristems, with CLV possibly promoting organ formation at the periphery and STM promoting meristem cell fate in the center (Clark et al., 1993, 1996). Our observation that STM expression in zll embryos was downregulated at the apex center raised the possibility that an increased CLV/STM ratio caused ectopic differentiation in zll. We therefore analyzed zll clv double mutants to determine whether CLV activity was required for ectopic cell differentiation in zll apices. zll clv1 embryos displayed the zll single mutant defects, such as flat apices (Figure 5B) or solitary central organs, suggesting that CLV1 is not required for differentiation of central apical cells in zll embryos. This observation is consistent with findings that CLV mutations cannot rescue embryonic shoot meristem development in stm (Clark et al., 1996). Adventitious shoot meristems in zll clv1 double mutants were drastically increased in size by the clv1 mutation (Figure 5C), as expected from the enlarged meristem in clv1 single mutants. Double mutant flowers had similar but more variable organ numbers compared with clv1 (Table II). The same phenotypes as for zll clv1 were observed for zll clv3, consistent with the notion that CLV1 and CLV3 act in the same process (Clark et al., 1995). Figure 5.zll clv double mutants. (A and B) Scanning electron microscopy of mature embryo apices where one cotyledon has been removed (*).(A) In clv1 a shoot meristem (arrow) and two leaf primordia, p, are present (compare with wild-type, Figure 1A). (B) zll clv1 double mutant embryos lack a shoot meristem (arrow; compare with zll, Figure 1B and C). (C) Scanning electron microscopy of a zll clv1 seedling apex. The adventitious meristem (arrow) is enlarged compared with zll (Figure 2F). c, cotyledon. Each bar represents 50 μm. Download figure Download PowerPoint Table 2. Comparison of the floral organ numbers Genotype n Organs sepals petals stamens carpels wt 10 4.0±0.0 4.0±0.0 6.0±0.0 2.0±0.0 zll-3 25 4.2±0.4 4.2±0.5 5.8±0.3 2.2±0.3 clv1-4 34 5.0±0.9 5.5±0.8 8.0±1.2 5.5±1.3 zll-3 clv1 39 4.5±0.8 5.4±1.5 7.0±2.5 5.4±2.4 Mean numbers of floral organs and standard deviations are given. Map-based cloning of the ZLL gene The ZLL gene was initially mapped with respect to morphological markers and was found to be linked to the TT3 gene on chromosome 5 (data not shown). To isolate the ZLL gene, we used RFLP and PCR-based mapping of recombination breakpoints in about 2500 meiotic F1 events from the cross zll (Ler)×wild-type (Nd). We isolated ∼1 Mb of contiguous genomic DNA from yeast artificial chromosome (YAC) libraries (Creusot et al., 1995), spanning the region between the TT3 and the CRA1 genes. We initiated chromosome walks from both directions with bacterial artificial chromosome (BAC) and P1 clones (Liu et al., 1995) and eventually localized ZLL on a single BAC clone of ∼100 kb, CO59 (Figure 6A). From this BAC we constructed a cosmid library, established a cosmid contig and narrowed down the position of the gene to a region of ∼30 kb in which we did not find any recombination. Using the cosmid clone c3 (Figure 6A), we isolated eight cDNA clones that represented a single transcript. Two independent cDNA clones of ∼3.3 kb that were isolated from different libraries both displayed an open reading frame of 2964 nucleotides. The presence of stop codons upstream of this open reading frame indicates that these cDNAs contain the full coding region. We confirmed that this reading frame represented the ZLL gene by identifying mutations in the cognate genomic DNA of eight different ZLL alleles (Figure 6B), seven of which represented G to A changes, typical for mutations induced by EMS, while one (zll-15) was an insertion of a single A. Figure 6.Isolation of the ZLL gene and sequence comparison. (A) Map of genomic clones and location of the ZLL gene. The numbers of recombination breakpoints in 2500 meiotic events are given. (B) The ZLL protein sequence deduced from the longest reading frame of ZLL cDNA. The mutations in zll-13 and zll-9 result in stop codons (designated *). The mutation in zll-15 causes a frame-shift, and the mutation in zll-8 changes an exon–intron border; both mutations result in predicted translational stops after a few codons (also designated *). The mutations in zll-16, zll-7, zll-6 and zll-2 change the amino acid sequence. The most highly conserved region shown in (D) is underlined. (C) Schematic multiple alignment of ZLL, AGO1 and C.elegans sequences. Lines represent stretches of sequences with >37% similarity. Identical residues in all sequences compared are shown underneath. The missense mutations in zll-16, zll-7, zll-6 and zll-2 affect four of these residues. (D) Comparison of the ZLL protein region between residues 873 and 942 with sequences derived from the following sources: rice expressed sequence tag (EST; DDBJ/EMBL/GenBank accession no. C29031), C.elegans open reading frame (Wilson, 1994) (DDBJ/EMBL/GenBank accession no. Z69661), rat EST (DDBJ/EMBL/GenBank accession no. H31693) and human EST (DDBJ/EMBL/GenBank accession no. R91199). Residues identical to the ZLL sequence are shown in boxes. Download figure Download PowerPoint ZLL is a member of a novel gene family specific to multicellular organisms The ZLL gene encodes a protein of 988 amino acids (Figure 6B). The predicted protein is hydrophilic, suggesting that ZLL is soluble in the cytosol. Its N-terminal region (residues 1–123) is highly proline rich (14%) and does not show significant similarities with known sequences. In contrast, the protein sequence from residue 124 to 988 shows 75% identity with the Arabidopsis ARGONAUTE1 (AGO1) protein (DDBJ/EMBL/GenBank accession no. U91995) and 43% identity with hypothetical proteins from Caenorhabditis elegans (e.g. DDBJ/EMBL/GenBank accession no. Z69661; Figure 6C). In addition, very high similarities between ZLL and sequences from plants, vertebrates and invertebrates were found at the C-terminus (Figure 6D). The four stop-mutations identified suggest that this part of the protein is not present in the respective zll alleles (Figure 6B). Except for ZLL and AGO1, the latter of which is required for leaf development (Bohmert et al., 1998), all other members of this sequence group are derived from genome or random cDNA sequencing projects and their functions are unknown. We did not identify sequence motifs suggestive of known functional domains in ZLL. Missense mutations affect residues that are identical in all available sequences (Figure 6C), suggesting that these residues are essential for a general function of this class of putative proteins. Interestingly, we did not find any ZLL-related sequences from bacteria or budding yeast for which the complete genome sequences are known, indicating that ZLL and other members of the gene family play roles specific to a multicellular context. Expression of ZLL We used in situ hybridization to analyze the ZLL expression pattern. ZLL mRNA was found in provascular cells at all stages of development (Figure 7A–E). Postembryonically, ZLL expression in provascular cells seemed to presage organ initiation (Figure 7C). Cross sections of embryo axes showed that ZLL expression was restricted to small groups of cells, presumably at the phloem pole (Figure 7E). In addition to provascular cells, ZLL mRNA was detected in the apex late in embryogenesis (Figure 7D) at about the time when zll apical development deviates from wild-type. ZLL mRNA, as a molecular marker for ZLL expressing cells, was distributed in provascular cells of zll mutant embryos indistinguishably from wild-type (data not shown). This indicates that the ZLL-expressing cells were still present in the mutant and confirmed our histological observation that the morphology of the vascular system was not notably altered by mutations in the ZLL gene. In order to make sure that we only detected the ZLL transcript and not transcripts from related gene family members in Arabidopsis, we used three different regions of the ZLL gene as probes (see Materials and methods), all of which gave identical results in in situ hybridization experiments and did not result in cross-hybridization to genomic Arabidopsis DNA under low stringency hybridization conditions (data not shown). Figure 7.ZLL gene expression pattern. In situ hybridization of ZLL antisense (A–E) and ZLL sense control RNA (F). In globular stage (A) and torpedo stage (B) embryos, ZLL is expressed in provascular cells (black arrow) but not in the embryo apex (white arrow). (C) In inflorescence, im, and floral, fl, meristems ZLL is expressed in provascular cells, apparently presaging the initiation of organs (arrow). (D) In bent cotyledon stage embryos ZLL is additionally expressed in the embryo apex (arrow). (E) Cross section of the embryo axis. ZLL expression is confined to what appear to be the phloem poles (arrow). (F) Sense control in a mature embryo. c, cotyledon; v, provascular cells; h, hypocotyl; e, endodermis; p, pericycle; *, cells of the inner integument that show a brownish color independently of the in situ hybridization. Download figure Download PowerPoint Discussion Our data address a critical process in shoot meristem development: the transition from early position-dependent activation of shoot meristem-specific genes in the embryo to a self-perpetuating stem cell system required for repetitive postembryonic organ formation. Mutations in the ZLL gene specifically disrupt embryonic shoot meristem development and result in the differentiation of stem cells. Here we discuss the evidence that the primary defect in zll mutants is the failure to maintain central cells of the embryonic shoot meristem primordium in an undifferentiated state. We then address the critical events leading to the establishment of a self-perpetuating shoot meristem organization and the role of the ZLL gene in this process. ZLL is required for shoot meristem cell fate in the embryo The strongest defect observed in zll seedlings is a flat apex with cells that appear to have lost meristematic cell features and initiated differentiation. Therefore, we suggest that the primary defect in zll embryos is the failure to maintain apical cells in an undifferentiated state. In this view, ectopic cell differentiation and organ formation are secondary effects: being unspecified, the central cells are permitted to switch to an alternative developmental pathway. Differentiation of these cells may represent an intrinsic 'default' pathway, or be initiated in response to external cues. The shoot meristem program appears to be initiated correctly in zll globular embryos. Later in embryogenesis, however, the cells in the center of the apex, which in wild-type constitute the stem cells, behave like peripheral cells in zll: they downregulate STM expression and initiate differentiation. After the initiation of the first true leaf primordia, the shoot meristem self-perpetuates independently of ZLL. The fact t
Referência(s)