Iii

Figure 2.1 The sequences of systemins and the structures of their precursors. Prolines (P) and hydroxyprolines (O) are in boldface. In the precursor diagram, the systemin peptides are shown as black boxes. The gray boxes represent the leader sequences. After Ryan and Pearce (2003).

As a result, the transgenic plant is compromised in defense against Manduca sexta larvae.

Systemin homologues have been identified in many other species of the Solaneae subtribe in the Solanaceae family but not in other subtribes of this family, such as tobacco. In addition, the tomato systemin does not induce the wound response in tobacco, which suggests that it is highly diverged from the tobacco wound signal and not perceived by tobacco cells. Tobacco appears to be missing a strong leaf-to-leaf long-distance wound signaling system, although strong leaf-to-root and local wound signaling systems are present (Pearce et al., 1993; Zhang & Baldwin, 1997; Constabel etal., 1998).

Exogenous application of systemin to tomato suspension cultures causes a rapid increase in the pH of the medium. This alkalinization is a convenient assay for the biological activity of systemin. Although the tomato systemin does not generate an alkalinization response when applied to tobacco suspension cells, extracts from tobacco leaves can do so, suggesting that a systemin-like activity exists in tobacco. Facilitated by the alkalinization assay, the Ryan group isolated two peptides from tobacco, Tobacco Systemin I and Tobacco Systemin II (which were later renamed TobHypSys I and II), that function to induce the localized wound response (Pearce et al., 2001a). Both tobacco systemins are hydroxyproline-rich 18-amino acid peptides generated from proteolytic processing of a single 165-amino acid precursor (Fig. 2.1). TobHypSys I and II do not share significant sequence similarity to each other or to the tomato systemin. In addition, unlike the tomato systemins, the tobacco systemins are glycosylated. However, they are all proline-rich (or hydroxyproline-rich in the case of tobacco systemins) and have a -PPS- (or -OOS-) motif. Prolines and hydroxyprolines may allow some limited conformations in their secondary structures, which could play important roles in interaction with their receptors (Pearce & Ryan, 2003; Narvaez-Vasquez & Ryan, 2004).

Although the suppression of the tomato systemin gene blocks systemic wounding signaling, it does not have a significant affect on local wound signaling (McGurl et al., 1992), which suggests that local wound signaling in tomato is independent of systemin. The search for additional wound signaling molecules in tomato has led to the identification of three new peptide signals, tomato hydroxyproline-rich systemin (TomHypSys) I, II, and III (Pearce & Ryan, 2003). These three peptides are 20,18, and 15 amino acids in length respectively (Fig. 2.1). Like the TobHypSys systemins, they are hydroxyproline-rich and glycosylated. All three TomHypSys systemins are derived from a single wound-inducible 146-residues precursor that shares sequence similarity to the precursor of the tobacco systemins, which suggests they evolved from a common ancestor gene.

The systemin precursor, prosystemin, is synthesized in phloem tissues and localized in the cytosol (Narvaez-Vasquez & Ryan, 2004). In contrast, the precursors of the TomHypSys and TobHypSys signals are likely synthesized on the rough ER, processed and modified in ER and the Golgi apparatus, and then enter the secretory pathway. Therefore, it is plausible that the TomHypSys and TobHypSys systemins move from cell to cell through intercellular fluid and act as short-range signals to induce the local wound response, whereas the localization of systemin provides the advantage of rapid transportation through phloem to activate the systemic wound response.

Jasmonate is another mobile signal molecule that interplays with systemin to amplify wound signaling (Ryan & Pearce, 2003). Wounding causes the generation and release of systemin from damaged cells. It moves through the apoplast and is perceived by neighboring cells, resulting in the induction of jasmonate synthesis. As jasmonate moves through the plant, it induces the production of more prosystemin. This systemin-jasmonate cycle amplifies the wound signaling process to achieve a strong systemic wound response and effectively defend the plant from its predator. However, the question as to how prosystemin is released from the cytosol of un-wounded distal cells to the apoplast remains unanswered. Since polypeptides usually do not diffuse across the plasma membrane, a specific transporter would be required to transport prosystemin to the intercellular space where it is activated by proteolytic processing. In species such as tobacco that lacks systemin, jasmonate is mainly responsible for long-range wound signaling. The peptide signals such as TobHypSys I and II help to amplify wound signaling and activate the local wound response.

The receptor of systemin was recently identified by its binding to systemin. The receptor is called SR160, a 160-kDa transmembrane protein with an extracellular leucine-rich repeat (LRR) domain and an intracellular serine/threonine kinase domain (Scheer & Ryan, 2002). It turns out that SR160 shares a very high sequence similarity to BRI1 in Arabidopsis. BRI1 was previously identified as the cell surface receptor of brassinolide, a steroid hormone that regulates plant growth and development (Li & Chory, 1997). Further evidence that SR160 is indeed the tomato brassinolide receptor came from another independent study aimed to isolate the tomato gene CU-3 that functions like BRI1 (Montoya et al., 2002). The study found that the cloned CU-3 gene, mutation of which results in insensitivity to brassinolide, encodes SR160. It was further revealed that the cu-3 mutant is not only defective in brassinolide signaling but also compromised in systemin signaling (Scheer et al., 2003). The above studies demonstrate that BRI1/SR160 has a dual function.

It is possible that SR160/BRI1 initially functions in the defense response but was co-opted as a component in brassinolide signaling during the course of evolution, although the reverse may be the case (Ryan & Pearce, 2003). Nevertheless, it raises an interesting question as to how perception of systemin and brassinolide by the same receptor activates two distinctive intracellular biochemical pathways.

It is yet to be determined if SB160 is also the receptor for the other systemins. However, even if different receptors are involved, it is clear that the signaling pathways mediated by different systemins converge at an early step and, therefore, they activate a similar physiological response. As in many signaling pathways mediated by peptide hormones in animals, a kinase cascade may be involved in activation of the plant wound response. Both the tomato systemin and the TobHypSys systemins activate a 48-kDa MAPK (Schaller & Ryan, 1996; Stratmann et al., 2000; Ryan & Pearce, 2003). Other intracellular events involved in the systemin signaling pathway include the release of Ca2+ from vacuoles to the cytosol as well as activation of phospholipase A2. These signaling events lead to generation of jasmonate from linolenic acid, which, together with ethylene, activates defense genes.

2.2.2 RALF regulates plant growth and development

During the purification of the tobacco systemins, another peptide from tobacco leaves was found to cause rapid alkalinization of the medium of the tobacco suspension cells. The peptide was named rapid alkalinization factor (RALF) (Pearce et al., 2001b). RALF is a 49-amino acid peptide derived from a 115-amino acid pre-proprotein. RALF induces a stronger and more rapid alkalinization response than the two tobacco systemins. However, unlike the systemins, it does not induce the tobacco PIN genes and, therefore, is not a signal of the wound response.

To determine its possible biological function, RALF was applied to excised plants and germinating seeds. It was found that RALF caused an arrest of seedlings' root growth by affecting the elongation zone and the root meristem (Pearce et al., 2001b). It also blocked root hair formation. RALF appears to be expressed in a variety of tissues and its homologues are present in plants species across the plant kingdom. The mechanism by which RALF regulates root growth remains elusive. Neither is it clear whether RALF affects other growth and developmental process.

2.2.3 ENOD40 regulates nodulation and cell proliferation

Some nitrogen fixation bacteria (rhizobia) in soil establish an endosymbiotic association with legumes. Biological nitrogen fixation is carried out by rhizobia in a specific type of organs called root nodules, which form following rhizobia infection. The nodule formation involves communication between the bacteria and the host as well as between different types of root cells. Flavonoids secreted by legume roots activate the rhizobial genes involved in production of Nod factors. Nod factors share the similar basic structure comprising lipo-chitin oligomers with a four or five b-1,4-linked N-acetyl glucosamine backbone in which the N-acetyl group of the

Pericycle cells Inner cortex cells

Figure 2.2 Schematic drawing of a part of a root cross section undergoing nodulation. Bacterial infection induces root hair curling and formation of an infection thread that carries the bacteria into inner cell layers. EN0D40 expression is first detected in the pericycle cells, which undergo limited cell division following the infection. The inner cortex cells dedifferentiate to enter the cell cycle, leading to formation of a nodule primordium. After Geurts and Bisseling (2002).

Pericycle cells Inner cortex cells

Figure 2.2 Schematic drawing of a part of a root cross section undergoing nodulation. Bacterial infection induces root hair curling and formation of an infection thread that carries the bacteria into inner cell layers. EN0D40 expression is first detected in the pericycle cells, which undergo limited cell division following the infection. The inner cortex cells dedifferentiate to enter the cell cycle, leading to formation of a nodule primordium. After Geurts and Bisseling (2002).

terminal sugar is replaced by an acyl chain (Geurts & Bisseling, 2002). Nod factors made by different rhizobial species are modified in a variety of ways and are a major determinant of host specificity.

Nod factors are the signaling molecules secreted by rhizobia to the root surface where they are recognized by specific receptors of host cells. They induce many host genes termed nodulin or nodule-specific genes and initiate a series of coordinated events, leading to reprogramming of differentiated root cells and formation of a nodule primordium (Fig. 2.2) (Geurts & Bisseling, 2002).

Rhizobia enter plants through root hairs. Rhizobia infection causes root hair to deform by reorienting tip growth (root hair curling). An infection thread containing the bacteria is then formed within the curled root hair. The infection thread grows toward the base of the root hair cell and eventually to the nodule primordium. The response of the epidermal cells to rhizobia is induced within minutes after Nod factor application. The response of inner cell layers to Nod factors can be detected within an hour. Following the rhizobial infection, pericycle cells undergo a limited number of cell divisions. The inner cortex cells then dedifferentiate to enter the cell division cycle, leading to formation of the nodule primordium. The nodule primordium differentiates to form a nodule after bacteria are released from the infection thread.

ENOD40 is one of the early nodulin genes (ENOD). The first ENOD4G gene, GmENOD4D, was identified from soybean (Yang et al., 1993). GmENOD40 is strongly induced during nodule development prior to the start of N2 fixation. This suggests that the gene is involved in the nodulation process but not N2-fixation process. During early stages of nodule development, the GmENOD40 gene is expressed first in the pericycle of root vascular bundle and then in dividing root cortical cells and the nodule primordium.

ENOD40 produces a 0.7-kb transcript that was initially thought to function as a nontranslatable RNA because it lacked a significant open reading frame (ORF) (Crespi et al., 1994). Further insights into the possible function of ENOD40 come from the study on the ENOD40 gene from Medicago truncatula (MtENOD40) (Sousa et al., 2001). When MtENOD40 was delivered into epidermal cells and the outermost cortical layer by ballistic microtargeting, it induced division of inner cortical cells to initiate nodule formation. This phenomenon was used as an assay for the MtENOD40 activity. The alfalfa ENOD40 contains two small ORFs (sORF I and sORF II) that encode 13 and 27 amino acids respectively. The sORF I region is highly conserved among the ENOD40 genes from different species. It was found that sORF I or sORF II alone is sufficient for ENOD40 activity. In addition, their activity requires translation of the sORFs. The other regions of the transcript may play a role in controlling stability or regulating the translation process. Therefore, the study indicates that the short peptides encoded by the sORFs are responsible for ENOD40 activity.

A study using in vitro translation of soybean ENOD40 in wheat germ extracts revealed that two peptides (Peptides A and B) of 12 and 24 residues, respectively, were translated from the single mRNA (Rohrig et al., 2002). Both peptides are translated from the region conserved among the ENOD40 genes. The studies on MtENOD40 and Soybean ENOD40 demonstrate that multiple peptides are produced from the polycistronic transcripts by translation of multiple small ORFs. Interestingly, both Peptides A and B bind to nodulin 100, which is a subunit of sucrose synthase (Rohrig et al., 2002). The discovery suggests that ENOD40 peptides may control photosyn-thate use, but it does no provide an obvious answer as to how the regulation of sucrose synthase activity by ENOD40 leads to cortical cell dedifferentiation and proliferation.

The ENOD40 gene is expressed not only during nodulation but also in phloem cells of stems. In addition, ENOD40 genes have been isolated from non-legume dicot species and monocot species such as tomato, rice, and maize that do not form nodules (van de Sande et al., 1996; Kouchi et al., 1999). Therefore, ENOD40 may have a general role as a regulator of cell proliferation in plants.

The peptides derived from ENOD40 are apparently localized in the cytosol. However, expression of ENOD40 gene in outer cell layers of alfalfa roots induces cell division of inner layers, suggesting that the EN0D40 peptide signals produced in the epidermal cells may move to the inner cell layers to induce cell proliferation (Sousa et al., 2001). Such a short-range movement can be achieved by diffusing through plasmodesmata. Alternatively, the EN0D40 action may generate a secondary signal that transmits the information to other cells to induce cell division. More work is needed to understand the mode of action underlying the EN0D40-mediated nodu-lation and cell proliferation processes.

2.2.4 PSK (phytosulfokine) is a mitogenic factor

Synthetic hormones such as auxin and cytokinin are usually added to cultured cells to stimulate cell proliferation. In addition, dispersed culture cells must reach a required initial cell density in order to proliferate. Low-density suspension cell cultures display strikingly low mitotic activity that could not be improved by supplementation with known plant hormones. However, adding conditioned medium derived from rapidly growing cell culture to a low-density culture can stimulate proliferation, suggesting that a mitogenic factor is secreted into the medium.

A suspension culture system of mesophyll cells prepared from Asparagus officinalis was used to purify the mitogenic factor (Matsubayashi & Sakagami, 1996). Two mitogenic factors were identified to be a sulfated pentapeptide (H-Tyr (S03H)-Ile-Tyr(S03H)-Thr-Gln-0H) and a sulfated tetrapeptide (H-Tyr(S03H)-Ile-Tyr(S03H)-Thr-0H). The two peptides are derived from a common precursor. The peptides were named phytosulfokine-a and -b (PSK-a and -b). Truncated PSK which is missing the two C-terminal residues still retained up to 20% activity, whereas truncation of any of the three N-terminal residues abolished its activity, indicating that the N-terminal tripeptide fragment is the active core (Matsubayashi et al., 1996). In addition, the sulfate group was found to be essential for its activity.

Incubation of low-density cells with either of the PSKs at nanomolar concentrations in combination with other hormones stimulates cell proliferation. The failure in proliferation of low-density cell culture is likely caused by the dilution of PSK below its active concentration in initial cultures. It has been proposed that PSKs are synthesized in culture cells in response to auxin and cytokinin and secreted into medium to regulate cell dedifferentiation and activate the cell cycle (Matsubayashi etal., 1999).

PSK-a was also identified in the conditioned medium of rice culture cells (Matsubayashi et al., 1997). The cloned rice PSK gene encodes a 89-amino acid product, preprophytosulfokine, which contains a 22-amino acid cleavable leader sequence (Yang et al., 1999). The PSK sequence is located near the C-terminus of the precursor. The PSK gene is expressed at a relatively high level in rice culture cells. It is also expressed at a low level in rice seedlings, indicating that the hormone play an important role in cell proliferation in culture and intact plants. Rice culture cells overexpressing the rice PSK gene were found to divide approximately twofold faster than wild-type cells. In contrast, antisense-mediated suppression of the rice gene caused a decrease in cell division of the rice cultures.

A homology search for Arabidopsis genes encoding the PSK sequence led to the identification of at least four putative AtPSKs, named AtPSK1, AtPSK2, AtPSK3, and AtPSK5, whose designated numbers reflect their chromosomal location (Yang et al., 2001). These four genes encode 67-87-amino acid polypeptides. The deduced Arabidopsis PSK precursors and the rice PSK precursor share little overall sequence similarity except that they are identical in an 8-amino acid region (termed the PSK domain), which includes the PSK sequence. In Arabidopsis cell cultures, both PSK-a and PSK-b are secreted to the medium. The Arabidopsis PSK genes are expressed in apical meristem, leaves, hypocotyls, and roots, again indicating that the PSK signals function not only in culture cells but also in intact plants.

Specific high-affinity binding activities for PSK-a have been detected in the outer surface of rice's plasma membrane. The specific binding was altered by the pH condition and ionic interaction, suggesting that the ligand-receptor binding is controlled by ionic interaction (Matsubayashi & Sakagami, 2000). The photoaffin-ity cross-linking study revealed the existence of two putative PSK receptors in rice. They are glycosylated proteins with molecular weights of 120 and 160 kDa respectively. The carrot cell line NC contains a relatively high concentration of high-affinity PSK-binding protein and therefore was used to purify the PSK receptor. Photoaffinity labeling of membrane proteins from the carrot cell line with a photoactivatable PSK analog revealed that a 120-kDa protein and a minor 150-kDa protein, both of which are glycosylated proteins, bind to PSK. The two binding proteins were purified and identified as an LRR-containing receptor kinase that shares sequence similarity to those of the known hormone receptors such as BRI1 and CLV1 (Matsubayashi et al., 2002). The 150-kDa protein is likely a modified form of the 120-kDa receptor.

The PSK receptor gene encodes a 1021-amino acid polypeptide with a predicted cleavable signal peptide. It contains an extracellular N-terminus with 21 LRR repeats, a transmembrane domain, and a cytoplasmic serine/threonine kinase domain. Suppression of the receptor gene by an antisense construct caused inhibition of PSK-mediated cell proliferation, further demonstrating its function as the PSK receptor.

Apparently, multiple hormones including auxin, cytokinin, and PSK interact to regulate differentiation and proliferation of cultured cells. It is intriguing to speculate that many other auxin- and/or cytokinin-mediated biological pathways in plants may involve peptide signals. Identification of downstream components in the PSK signaling pathway should provide novel insights into molecular mechanisms that control cell differentiation and proliferation.

2.2.5 CLAVATA 3 (CLV3) regulates stem cell homeostasis

Embryogenesis in animals generates a miniature version of an adult organism. In contrast, a plant embryo has a much simpler structure that contains two stem-cell populations: the shoot apical meristem and the root apical meristem (Clark, 2001). The body patterns of an adult plant are largely established by postembryonic pattern formation at these meristems. The shoot apical meristem (SAM) acts as a

Leaf primordium

Leaf primordium

Figure 2.3 Schematic drawing of organization of an apical meristem. The meristem is organized into three layers: L1, L2, and L3. The stem cells are indicated as the gray area. The arrows indicate the flow of cells as a result of stem cell proliferation. Two leaf primordia are formed on the flanks of the meristem. After Clark (2001).

Figure 2.3 Schematic drawing of organization of an apical meristem. The meristem is organized into three layers: L1, L2, and L3. The stem cells are indicated as the gray area. The arrows indicate the flow of cells as a result of stem cell proliferation. Two leaf primordia are formed on the flanks of the meristem. After Clark (2001).

self-renewing source of undifferentiated stem cells whose descendents become incorporated into organ and tissue primordia and acquire different fates, leading to formation of the above-ground organs such as leaves, flowers, vasculature and other tissues of the stem, whereas the root meristem is responsible for development of the root system (Sharma et al, 2003).

In order for the shoot meristem to form organs continuously, different regions of the shoot meristem have to establish constant communication so that a balance between stem cell proliferation through cell division and cell departure from the meristem to form lateral organs is maintained (see also Chapter 6). Cells of SAM consist of three clonally distinct cell layers (Fig. 2.3) (Clark, 2001). Cells of the L1 layer (the epidermal layer) and the L2 layer (the subepidermal layer) divide in an anticlinal fashion, and therefore they are single cell thick. Cells of the underlying L3 layer divide in various orientations. Since all the differentiated cell types of the adult plant are derived from a small number of stem cells, cell lineage patterns do not play a critical role in regulating cell fate. Instead, cell fate is determined by a positional effect that requires communication within and between the different layers of cells.

Genetic screens for Arabidopsis mutants with an altered SAM morphology have identified three CLAVATA genes (CLV1, CLV 2, and CLV3) whose loss-of-function mutations result in an enlarged meristem (Clark et al., 1993, 1995; Kayes & Clark, 1998). Some of the clv mutants accumulate over 1000-fold more undif-ferentiated cells in SAM than do wild-type plants. Genetic analysis revealed that these three genes function in the same pathway by preventing unrestricted stem cell proliferation. For instance, the phenotypes of clvl and clv3 null mutants and clvl clv3 double mutants are nearly identical.

The three CLV genes have been cloned. The CLV3 gene encodes a 96-amino acid protein that shares no significant sequence similarity to other proteins with known functions (Fletcher et al., 1999). It contains a predicted 18-amino acid N-terminal signal peptide that directs the protein into the secretory pathway. CLVl encodes a receptor kinase and, like BRI1 and the PSK receptor, it has an extracellular LRR domain and an intracellular kinase domain (Clark et al., 1997). CLV2 is similar to CLV1 but lacks a cytoplasmic kinase domain. The identification of the CLV proteins suggests that CLV1 and CLV2 function as the receptors, and CLV3 acts as the extracellular ligand. CLV1 and CLV2 likely form a heterodimer (Jeong et al., 1999; Trotochaud et al., 1999). Binding of CLV3 with CLV1/CLV2 activates the receptor, which then recruits additional signaling components to form a 450-kDa signaling complex, leading to activation of downstream effectors (Trotochaud et al., 2000; Clark, 2001).

Studies on expression patterns of the CLV genes have generated important insights into their functions. CLV3 is expressed predominantly in the L1 and L2 layers at the apex of the shoot apical meristem. However, CLVl is predominantly expressed in the L3 cells, largely beneath the CLV3 expression domain (Fletcher et al., 1999). The results indicate that CLV3 acts as a short-distance signal that is secreted from the L1 and L2 layers and diffused to the inner cell layers where it is perceived by CLV1/CLV2 to restrict stem cell proliferation.

If the CLV pathway functions to restrict stem cell proliferation, then how is stem cell homeostasis maintained? Sharma et al. (2003) has proposed a feedback loop model in which the CLV pathway and a positive, stem-cell-promoting pathway interact to maintain the balance between cell loss and cell division. The homeodomain transcription factor WUSCHEL (WUS) is a key component of the cell-promoting pathway. WUS is required for maintaining the stem cell reservoir, and its loss-of-function mutant is unable to maintain a pool of pluripotent stem cells, resulting in meristem termination (Laux et al., 1996; Mayer et al., 1998). The WUS-mediated positive pathway acts to promote stem cell proliferation. The resulting enlargement of the stem cell population leads to increased production of the CLV3 ligand. As a consequence, it activates the CLV-mediated negative pathway. The enhanced CLV-signaling then causes reduction in the level of WUS transcription, which in turn reduces production of CLV3. The resulting reduction in the CLV-mediated negative pathway then activates the positive pathway. Through this mutual regulation equilibrium is attained.

The identification of the components in the CLV1 450-kDa complex would lead to the identification of additional players involved in the CLV pathway. At least two proteins have been identified from the complex (Trotochaud et al., 1999). The first one is the kinase-associated protein phosphatase (KAPP). In contrast to protein kinases, phosphatases often switch off a response through dephosphorylation. KAPP likely negatively regulates CLV1 action by dephosphorylating CLV1. Another component of the CLV1 complex is a Rho/Rac-GTPase-related protein. Members of this family such as Ras are important components in relaying signals through a series of intracellular biochemical reactions including the kinase cascade.

2.2.6 S-locus cysteine-rich proteins determine specificity of self-incompatibility in the Brassicaceae

During a compatible pollen-pistil interaction, a desiccated pollen grain adheres tightly to the surface of stigma, where it hydrates and germinates. The pollen tube then elongates into the female tissues. The pollen tube elongation is guided toward individual ovules where sperm cells are released to fertilize the egg cell and central cells, leading to development of embryo and endosperm (Pruitt, 1999).

Self-incompatibility (SI) is a phenomenon in which self-pollen is recognized and rejected by pistil whereas non-self pollen is accepted (see also Chapter 10). SI is used by many plant species to prevent inbreeding and maintain genetic variability. The SI system is genetically controlled by a single locus called the S locus, which contains multiple genes that could be multi-allelic. Different mechanisms govern pollen recognition and rejection. The SI system of the Solanaceae species is called gametophytic self-incompatibility. In this system, if the single S allele carried by the haploid pollen matches either of the two S alleles present in the diploid tissues of the pistil, the pollen will be rejected (Matton et al., 1994). Usually self-pollen can germinate but pollen tube elongation is inhibited after they enter the style.

In contrast, in the sporophytic self-incompatibility system used by the Bras-sicaceae species, the genotype of the male parent, not that of the pollen itself, determines the outcome. If either of the two S alleles present in the male parent matches either of the two S alleles carried by the pistil, pollen hydration and germination is arrested. The self-incompatible response in this SI system occurs on the stigma surface by blocking the very early events of pollen-pistil interactions, suggesting that cell surface molecules are involved in pollen recognition. The molecular mechanism that determines sporophytic self-incompatibility in the Brassicaceae is described here.

Genetic and molecular studies carried out a decade ago identified two multi-allelic genes associated with the S locus in Brassica (Nasrallah et al., 1987; Stein et al., 1991). 0ne of them encodes S-locus glycoprotein (SLG), and the other encodes S-locus receptor kinase (SRK). SLG is a glycoprotein that is localized to the cell wall of papillae, the epidermal cells of the pistil where pollen-pistil interaction occurs (Kandasamy et al., 1989). SRK is a transmembrane protein which contains an extracellular domain (S-domain) with high sequence similarity to SLG and an intracellular serine/threonine kinase domain. SLG and SRK are tightly linked and highly polymorphic. Like SLG, SLK is predominantly expressed in the papilla cells. Suppression of LSG or LRK was found to convert an otherwise incompatible interaction into self-compatible (Takayama & Isogai, 2003), demonstrating that both proteins function in pollen recognition. It is believed that SRK is the female determinant of SI specificity in the stigma, and that SLG enhances the strength of the SI response by facilitating SRK maturation and stability (Dixit et al., 2000; Dixit & Nasrallah, 2001).

After the identification of the female determinant, many efforts were directed toward identifying the male determinant of self-incompatibility. The following criteria are applied to define a male determinant (Takayama & Isogai, 2003): (1) it is encoded by a gene located at the S locus; (2) it exhibits allelic diversity reflecting different variants of the S locus (designated S haplotypes); (3) it is expressed in sporophytic cells because the SI is determined by sporophytic genotypes; (4) it is localized on the surface of pollen grains; and (5) it physically interacts with SRK.

In a search for the male determinant, a gene located between SLG8 and SRK8 of the Brassica S8 haplotype was identified as a good candidate (Schopfer et al., 1999). This gene exhibits polymorphism associated with different S haplotypes and is specifically expressed in anthers. It encodes a small cysteine-rich protein with a molecular weight of approximately 9 kDa. The protein was named S-locus cysteine-rich protein (SCR). A definitive proof that SCR functions as the male determinant of SI came from transgenic studies. Pollen of an S2 haplotype plant is normally compatible with stigma of an S6 plant. However, when SCR6 was transformed into an S2 haplotype, pollen from the transgenic plants was rejected by S6 stigma.

The SCR genes are expressed in the microspores and tapetal cells, the diploid maternal tissues surrounding the developing male gametophytes (pollen) (Takayama et al., 2000). The tapetal cells undergo programmed cell death during microgameto-genesis and provide their constituents to pollen coating. The SCR expression pattern explains why the Brassica SI system is sporophytically determined. Further studies revealed that SCR is localized mainly in the pollen coat (Takayama & Isogai, 2003).

SCR has a cleavable leader sequence and is secreted. SCRs encoded by different alleles generally share approximately 40% sequence identity. This indicates an extensive divergence of this protein, consistent with their role as the SI specificity determinants. SCRs generally contain eight cysteine residues that form four disulphide bonds (Takayama et al., 2001). It is speculated that specificity of SI might be determined by other residues. Following pollination, SCR is diffused from pollen coats through the papilla wall and binds to the cognate SRK (Kachroo et al., 2001; Takayama et al., 2001). The ligand binding induces a conformational change of SRK, which activates the kinase through autophosphorylation. Activated SRK phosphorylates downstream targets to initiate a signaling cascade and activate specific biochemical response, leading to the rejection of self-pollen.

The extensive studies in the last decade have generated a fairly clear picture of the molecular mechanism involved in pollen recognition of the SI system. However, the way SCR-SRK interaction is translated into self-pollen arrest remains largely unknown. One possible scenario (Kachroo et al., 2002) is that the SCR-SRK signaling pathway results in release of calcium from the papillar cells to the pollen-papillae interface. Pollen maintains a calcium gradient. Uptake of calcium by the pollen grain changes the calcium gradient, leading to blockage of pollen germination. Another possibility is that the SCR-SRK interaction leads to activation of a physiological response in the papillae cells, which is analogous to the disease resistance response. Such a response would cause the papillae to secret components that could be detrimental to pollen.

Was this article helpful?

0 0

Post a comment