C

FIG. 3. CLV3 expression. In situ hybridization with digoxigenin-labelled CLV3 antisense probe of longitudinal sections through Arabidopsis shoot apical meristems of wild-type (A), clv1-4 (B), clv2-1 (C) or clv3-2 (D). In the wild-type, only a few cells at the tip of the meristem (the putative stem cells) express CLV3 RNA. In all clv mutants, stem cells accumulate in the meristems, and the CLV3 expression domain is enlarged.

membrane. CLV3 is the best candidate for a ligand that binds and activates the CLV1/CLV2 receptor so far (Fletcher et al 1999, Brand et al 2000). CLV3 encodes a relatively small protein with no conspicuous amino acid motifs. However, it carries an N-terminal signal peptide, making it likely that the CLV3 protein can be secreted and interact with the extracellular domains of the CLV1 and CLV2 receptors. There is no direct evidence yet that CLV3 acts as a secreted ligand. Periclinal chimeras derived from the unstable, transposon induced clv3-7loss-of-function mutant were used to investigate the cell-autonomy of CLV3 function (Fletcher et al 1999). Secondary shoots that showed a wild-type phenotype were isolated on clv3-7mutant plants, and seeds were collected from the revertant sectors after self-pollination. A wild-type (revertant) clv3 allele would segregate among the progeny of these somatic revertants only when the reversion occurred in the L2 layer of the meristem, which gives rise to the gametes. All reversions that occurred in the L1 or L3 layer should not be transmitted to the next generation. We found that many revertant sectors segregated only the mutant allele, indicating that CLV3 function was restored somatically in the L1 or L3 cell layer, but not in the L2. Restoration of CLV3 activity in a single cell layer is therefore sufficient to control the proliferation and differentiation of the meristem, and CLV3 can therefore function non-cell-autonomously.

CLV3 mRNA is found primarily in the L1 and L2 layer of the central zone of shoot and floral meristems (Fig. 3), while the RNA of CLV1 is found mostly in an underlying domain in the L3 (Clark et al 1997, Fletcher et al 1999). The expression pattern of CLV2 has not yet been analysed in great detail, but RNA can be detected in all shoot tissues of the plant (Jeong et al 1999). So the stage is set for CLV3 as a signal that is secreted from cells in the outer layers of the meristems and binds to a heterodimeric receptor, consisting of CLV1 and CLV2, in deeper layers. A combination of biochemistry and genetics provided further evidence that CLV1 function depends on CLV3 (Trotochaud et al 1999). In the wild-type, the CLV1 protein is found primarily in a 450 kDa protein complex, which is assumed to be the active complex (Fig. 2). Formation and stability of this complex depends on the presence of functional CLV3 and CLV2 protein. Two more proteins were identified as members of this complex: One of them is KAPP, a kinase associated protein type 2C phosphatase, which has been shown to down-regulate CLV signalling, possibly by inhibiting the auto- or transphosphorylating activity of CLV1 (Williams et al 1997). The second protein, ROP, is a small GTPase related to the Rho protein family. Other members of this family control cytoskeleton reorganization during pollen tube growth (Li et al 1998), thus raising the possibility that the cytoskeleton itself is a target for CLV signalling.

Since the central zone of the meristem expands in civ mutants, CLV signalling would act to restrict expansion of the central zone and promote the timely exit of cells into the surrounding peripheral zone. The studies on the tomato fasciated mutant have shown that expansion of the stem cell zone is promoted from L3. In Arabidopsis, the WUSCHEL (WUS) gene encoding a homeodomain transcription factor promotes stem cell fate, since wus mutants fail to initiate or maintain an active stem cell population in meristems, resulting in a premature differentiation of cells in the central zone (Laux et al 1996). WUS RNA is found only in the L3 layer of shoot and floral meristems, indicating that this gene may act non-cell autonomously (Fig. 4, Mayer et al 1998). However, wus mutants can still produce axillary inflorescences and occasionally abnormal flowers that lack most organs in the inner whorls, so WUS is not required for meristem initiation per se. Mutations in WUS or the CLV genes have essentially opposite phenotypes, but wus mutants are epistatic to civ mutants. Therefore, WUS could be a target gene for repression by the CLV proteins. Alternatively, WUS could be required to establish the cells upon which the CLV proteins can act. In civ mutants, WUS RNA is no longer confined to a few cells in deeper layers of the meristem (Fig. 4, Schoof et al 2000, Brand et al 2000), indicating that one consequence of CLV signalling is to down-regulate WUS, or to limit expansion of the WUS expression domain. The gradual enlargement of the shoot meristem in civ mutants could then be due to the deregulation of WUS. The increased WUS activity would specify more stem cells, resulting in an expanded central zone and eventually fasciation of the shoot meristem. Support for this model of CLV/WUS interaction came from two types of ectopic expression experiments:

FIG. 4. CTV3 and WUS expression in wild-type and elvi mutant embryos. (A and C). In situ hybridization with CTV3 digoxigenin-labelled antisense probe (A) Longitudinal section through a wild-type embryo, bent-cotyledon stage. (C) In a elvi mutant embryo, the CTV3 expression domain expands, indicating that stem cells accumulate. (B and D) In situ hybridization with WUS digoxigenin-labelled antisense probe.(B) Wild-type embryo. WUS is expressed in deeper cell layers underlying the stem cells. (D) elvi mutant embryo. WUS is now expressed in a larger domain in the L2 and L3 layers. Bars, 20 m. Abbreviations: L1, L2, L3, meristem layers; H, hypocotyl; Co, cotyledon.

FIG. 4. CTV3 and WUS expression in wild-type and elvi mutant embryos. (A and C). In situ hybridization with CTV3 digoxigenin-labelled antisense probe (A) Longitudinal section through a wild-type embryo, bent-cotyledon stage. (C) In a elvi mutant embryo, the CTV3 expression domain expands, indicating that stem cells accumulate. (B and D) In situ hybridization with WUS digoxigenin-labelled antisense probe.(B) Wild-type embryo. WUS is expressed in deeper cell layers underlying the stem cells. (D) elvi mutant embryo. WUS is now expressed in a larger domain in the L2 and L3 layers. Bars, 20 m. Abbreviations: L1, L2, L3, meristem layers; H, hypocotyl; Co, cotyledon.

Transgenic Arabidopsis plants that express C~LV3 constitutively resemble wildtype at early stages of development, but soon after emergence of the first leaves, the meristem arrests and ceases to produce further organs (Brand et al 2000). In some transgenic plants, the meristem resumes activity and will even produce an inflorescence with flowers that lack the inner organs. The opposite phenotype is observed in clv3 loss-of-function mutants, where stem cells accumulate in the centre of shoot and floral meristems, and additional organs, whorls of organs and/or undifferentiated tissue are formed. Thus, C~LV3 loss-of-function causes excessive accumulation of stem cells, while CLV3 gain-of-function causes premature differentiation and loss of stem cells, possibly by causing a down-regulation of WUS.

Transgenic plants expressing WUS in a broader domain of the meristem resembled clv-mutants with large and fasciated meristems (Schoof et al 2000). When WUS was expressed from the ANT promoter in organ primordia, these primordia were converted into shoot meristems that expressed CLV3,

FIG. 5. Model for the feedback regulation of stem cell fate. The surface of the shoot apical meristem is indicated by the black line, gene names are roughly positioned where the genes are expressed. CLV3 acts from the tip of the meristem and activates the CLV1/CLV2 receptor complex. Activity of the CLV pathway represses transcription and/or function of WUS, and possibly other genes. WUS promotes non-cell autonomously stem cell fate in the overlying cells, and also promotes expression of the CLV genes in their respective domains. Together, these gene functions establish a feedback loop with negative and positive interactions for the control of stem cell fate in meristems.

suggesting that WUS is indeed sufficient to specify stem cell identity. However, WUS function is not required for C~LV3 expression (Brand et al 2000). It is likely that there additional factors that function together with WUS in a stem cell-promoting pathway. Mutations in the POLTERGEIST (POL) gene have been identified as partial suppressors of mutations in the CLV genes (Yu et al 2000). When the CLV genes are functional, pol mutants are nearly indistinguishable from wild-type plants. POL appears to act downstream of the CLV signalling pathway and redundantly with WUS, indicating that this gene could be a target for repression by the CLV signal transduction pathway.

A possible model consistent with these observations is that WUS, together with other genes like POL, acts as an organizer and promotes stem cell fate in the overlying cells (Fig. 5). These stem cells express the CLV genes, which in turn act to restrict the expression or function of WUS, so that these two antagonistic activities can constantly readjust the size of the central zone. During organ formation, cells in the peripheral zone need to be replaced, which are ultimately derived from the central zone. One attractive hypothesis is that the central zone is influenced by signals emanating from organ primordia, which could act via the CLV pathway to control cell division rates or fate in the central zone.

A cknowledgement

Work in R. S.'s laboratory was supported by the Deutsche Forschungsgemeinschaft through SFB 243.

0 0

Post a comment