SRK (S-locus receptor kinase) is the receptor that allows stigma epidermal cells to discriminate between genetically related (‘self’) and genetically unrelated (‘non-self’) pollen in the self-incompatibility response of the Brassicaceae. SRK and its ligand, the pollen coat-localized SCR (S-locus cysteine-rich protein), are highly polymorphic, and their allele-specific interaction explains specificity in the self-incompatibility response. The present article reviews current knowledge of the role of SRK in the recognition and response phases of self-incompatibility, and highlights the new insights provided by analysis of a transgenic self-incompatible Arabidopsis thaliana model.
Ligand-activated transmembrane receptor protein kinases mediate a wide range of biological processes in plants, including regulation of vegetative and reproductive development, pollination and fertilization, hormone perception, and disease resistance. Understanding how receptor protein kinases fulfil their diverse functions requires knowledge of the ligand that binds to their extracellular domain and activates their cytoplasmic kinase domain as well as the downstream signalling cascades that are triggered by receptor activation to produce a cellular response. Although much effort has been devoted in recent years to the characterization of plant receptor protein kinases, collectively known as RLKs (receptor-like kinases), the majority of the large number of RLKs encoded by plant genomes have no known biological function, and only in a few cases have ligands been identified or signalling cascades been described.
SRK (S-locus receptor kinase) is one receptor protein kinase for which both a biological function and activating ligand are known. SRK is the receptor that allows the stigma to discriminate between genetically related (‘self’) and genetically unrelated (‘non-self’) pollen in the self-incompatibility response of the Brassicaceae . Genetic self-incompatibility systems are physiological out-crossing mechanisms found in more than half of angiosperm species . In the Brassicaceae family, self-incompatibility is controlled by variants of a single highly polymorphic genetic locus, called the S-locus, for which up to 100 variants can occur in any self-incompatible species . The SRK gene was identified in Brassica as a stigma-expressed polymorphic gene that is genetically linked to the S-locus  and was later demonstrated to be the stigma determinant of self-incompatibility specificity . Subsequently, a second S-locus-linked polymorphic gene that was expressed in anthers was identified and shown to encode the pollen determinant of self-incompatibility specificity  and to be the ligand for SRK [6,7]. This gene was designated the SCR (S-locus cysteine-rich protein) gene because it encodes a small secreted protein of ~50 amino acids containing eight cysteine residues . Subsequent studies in Brassica and self-incompatible members of other Brassicaceae genera confirmed that each functional S-locus variant contains an SRK and an SCR gene, which together comprise the ‘S haplotype’ .
A dramatic demonstration of the fact that SRK and SCR are not only the sole determinants of self-incompatibility specificity, but also the primary determinants of the out-crossing mode of mating in the Brassicaceae, was the transfer of the self-incompatibility trait to Arabidopsis thaliana, the close self-fertile relative of self-incompatible Arabidopsis lyrata and Arabidopsis halleri [9,10]. A. thaliana harbours non-functional S haplotypes that contain defective SRK or SCR genes [11–16]. However, transformation of A. thaliana with functional SRK and SCR gene pairs isolated from self-incompatible A. lyrata or Capsella grandiflora conferred a self-incompatibility phenotype on several accessions of A. thaliana [9,10,17,18]. These transgenic complementation experiments demonstrate that the SRK-mediated signalling pathway was retained in A. thaliana. The availability of these self-incompatible A. thaliana SRK–SCR transformants has enabled the use of experimental approaches that are difficult or impossible to implement in Brassica species. The present brief review outlines the unique features that distinguish SRK from other plant RLKs, describes what is known about the role of SRK in the recognition and response phases of the self-incompatibility response, and highlights some of the often unexpected insights into the regulation and biological functions of SRK that have emerged from use of the transgenic self-incompatible A. thaliana model.
Unique features of SRK
Like other RLKs, the SRK protein has a single-pass trans-membrane domain with an extracellular domain at its N-terminus followed by a transmembrane domain and a serine/threonine protein kinase domain towards its C-terminus (Figure 1A). However, unlike other functionally characterized RLKs, which belong either to the LRR-RLK (leucine-rich repeat RLK) subfamily or to the CrRLK1L (Catharanthus roseus RLK1-like) subfamily, SRK is the prototypic and only functionally characterized member of the SD-RLK (S-domain RLK) subfamily . SRK and other SD-RLKs are characterized by the presence of a distinctive extracellular domain, called the S domain. As shown in Figure 1(A), starting at its N-terminus, this domain, which contains several N-glycosylation sites, consists of two contiguous B-lectin domains followed by a region containing 12 conserved cysteine residues, the C-terminal six of which are contained within a PAN/APPLE domain that is responsible at least in part for the ability of SRK to self-interact .
Structures of the SRK and SCR proteins
Another unique feature is that SRK and its ligand SCR are encoded by genes that are physically linked in the genome, and that alleles of these genes are inherited as a single genetic unit. Moreover, both proteins exhibit an extraordinarily high degree of intraspecific sequence polymorphism, consistent with their role as receptors and ligands involved in recognition. A self-incompatible species typically harbours a large number of S haplotypes and thus contains a large series of receptor variants and a corresponding series of ligand variants, and statistical analysis of sequence alignments suggests that SRK and SCR co-evolve. SRK variants can exhibit more than 30% amino acid sequence divergence [21,22]. SCR variants are even more polymorphic and typically share less than 50% amino acid sequence similarity [5,22–24], although they all appear to assume the same 3D structure, a CSαβ (cysteine-stabilized αβ) structure similar to defensins (Figure 1B). Comparisons of SCRs show conservation of only eight cysteine residues, a glycine residue between the first and second cysteine residues, and an aromatic amino acid residue between the third and fourth cysteine residues [5,23–25].
SRK and the recognition phase of self-incompatibility
The self-incompatibility response of the Brassicaceae is triggered within minutes of contact between ‘self’ pollen and the stigma epidermal cell, causing inhibition of pollen hydration, germination and pollen tube growth. Molecular and biochemical characterization of SRK and SCR has provided an explanation for how the stigma epidermal cell is able to recognize specifically self pollen from among the various pollen grains that bombard its surface. First, the SRK gene is expressed predominantly in stigma epidermal cells  and the SRK protein is an integral plasma membrane protein [27,28]. For its part, the SCR gene is expressed in the anther tapetum, and the SCR protein is incorporated into the pollen coat during pollen maturation [7,23]. Secondly, biochemical studies demonstrated that SCR is the ligand that binds to the extracellular S domain of SRK, and that SRK and SCR exhibit allele-specific interactions, whereby an SRK variant will only bind, and be activated by, the SCR variant encoded by the same S haplotype [6,7]. On the basis of these observations, the recognition of self pollen occurs as follows. In a self-pollination, the pollen-borne self SCR is delivered to the surface of the stigma epidermal cell and diffuses, or is transported, across the epidermal cell wall to gain access to SRK. The binding of this self SCR to the extracellular domain of SRK activates the SRK kinase, and is believed to trigger a cellular response in stigma epidermal cells that causes inhibition of pollen germination and tube penetration into the stigma epidermal cell wall. By contrast, in a cross-pollination, the SCR variant delivered to the stigma can neither bind nor activate SRK, and the pollen inhibitory signalling cascade is not triggered, thus allowing pollen tube development to proceed.
The high specificity that characterizes the SRK–SCR interaction raises the question of which residues in receptor and ligand are responsible for this specificity. Statistical analysis of the sequences of the ligand-binding extracellular S domain of SRKs identified highly variable residues having a high probability of being subject to positive selection, and are thus inferred to be important for self-incompatibility specificity [22,29]. However, the validity of these inferences has so far only been tested for a small number of variants by structure–function analysis studies involving domain swapping between variants and site-directed mutagenesis of individual residues or groups of residues. Although it was possible to perform structure–function analysis studies of the small SCR protein in Brassica because of the availability of a simple pollination bioassay for this protein , a similar analysis of the ~400-amino-acid-long SRK extracellular domain had to await the introduction of several different self-incompatibility specificities into A. thaliana [18,31]. The ease of transformation in this species was instrumental in allowing the functional analysis in planta of the large number of engineered mutant variants required for structure–function analysis of SRK.
Together, the structure–function analysis studies of SCR and the SRK extracellular domain have revealed, surprisingly, that only a few of the many polymorphic amino acid residues found in these proteins are essential for the ability of SRK to recognize its cognate SCR and trigger the self-incompatibility response in stigmas [30,31]. These results suggest that a change in the specificity of receptor and ligand involves only a few amino acid changes. However, because the number of variants that have been analysed so far is small, there are at present no general rules that allow identification of specificity residues in SRK or SCR sequences. Moreover, determining whether any specificity-determining residues that are identified represent contact points between SRK and SCR will require an as yet unavailable high-resolution 3D structure of SRK in its ligand-bound and ligand-unbound forms. Pinpointing the specificity residues in SRK and SCR is critical for understanding how the two proteins co-evolve to maintain their interaction and for addressing the intriguing and difficult question of how new specificities arise in the two-component SRK–SCR system .
SRK signalling and the response phase of self-incompatibility
By contrast with the recognition phase of the self-incompatibility response, which is relatively well characterized, the signalling cascade that is triggered within the stigma epidermal cell by the SRK–SCR interaction is poorly understood. Thus the question of how self pollen is inhibited has not been answered. However, three proteins have been proposed to function in SRK-mediated signalling by genetic analysis of Brassica self-compatible mutants and interaction studies in yeast: MLPK (M-locus protein kinase), which is a membrane-tethered kinase that interacts with the SRK kinase domain , ARC1 (arm repeat-containing protein 1), an E3 ubiquitin ligase that interacts with, and is phosphorylated by, SRK , and the Exo70A1 component of the exocyst complex, which interacts with ARC1 . On the basis of analysis of the latter two proteins, and the fact that the exocyst complex has been shown to function in polarized secretion in yeast and animal systems, it has been proposed that the stigma epidermal cell secretes factors required for proper hydration or germination of pollen grains via the exocyst complex. Activation of SRK would cause activation of ARC1, which would ubiquitinate Exo70A1 and target it for degradation. As a consequence, ‘compatibility’ factors would not be secreted and pollen grains will be unable to hydrate and germinate.
Interestingly, analysis of the role of A. thaliana orthologues of MLPK, ARC1 and Exo70A1 using the transgenic A. thaliana SRK–SCR self-incompatible model in conjunction with T-DNA (transferred DNA) knockout mutations does not support a role for these proteins in the self-incompatibility response of A. thaliana SRK–SCR transformants . Although the reason for this discrepancy is not known, one possibility is that self-incompatibility signalling is based on multiple signalling pathways rather than a single linear pathway . Each of these pathways might make a partial contribution to the overall self-incompatibility response, with different branches of the pathway being utilized preferentially in different Brassicaceae species. In any case, whatever the SRK signalling pathway proves to be, any model of self-incompatibility signalling must account for the fact that a single stigma epidermal cell can discriminate between self pollen and non-self pollen. Thus, rather than being a global response of the stigma epidermal cell, the pollen-inhibitory signalling cascade triggered by the SRK–SCR interaction must be restricted to cytoplasmic microdomains beneath the point of contact with an incompatible pollen grain.
SRK and pistil development
For almost two decades after its discovery, SRK was thought to function exclusively in the self-incompatibility response. It was not until the transgenic A. thaliana self-incompatible model was used for mutagenesis of the self-incompatibility response that an additional role for SRK was uncovered [37,38]. This mutant screen was performed in SRK–SCR transformants of the Col-0 accession. Col-0[SRK–SCR] transformants differ from SRK–SCR transformants of some other accessions, such as C24, Cvi, Sha, Kas and Hodja, which do not set seed because their stigmas express an intense self-incompatibility response that persists throughout stigma development [10,15]. Instead, Col-0[SRK–SCR] transformants express self-incompatibility in mature flower buds and very young flowers, but not in older stigmas [9,10]. Because they set seed, it is possible to use these plants for standard chemical mutagenesis. A screen for Col-0[SRK–SCR] plants exhibiting a modified self-incompatibility response identified a mutation that enhanced the self-incompatibility phenotype of these plants, resulting in lack of seed set . The mutation also caused stigma exsertion (i.e. positioning of the stigma above the anthers) due to increased pistil elongation resulting from increased cell proliferation. Unexpectedly, the stigma exsertion phenotype was dependent on an active SRK kinase, revealing a previously unsuspected role for SRK and SRK-mediated signalling in the promotion of cell proliferation in the pistil. How SRK signalling produces different outcomes in the stigma epidermal cell and other cells of the pistil remains to be determined.
Map-based cloning of the mutation identified RDR6 (RNA-dependent RNA polymerase 6), which functions in the production of tasiRNAs (trans-acting siRNAs). Subsequent analysis demonstrated that ARF3 (auxin-response factor 3), a known target of RDR6, acts non-cell-autonomously from its site of expression in cells below the stigma to enhance the self-incompatibility response and simultaneously down-regulate auxin responses in stigma epidermal cells . These unexpected results reveal the importance of communication between epidermal and subepidermal cells for stigma function, and suggest the tantalizing possibility that auxin may be involved in the regulation of self-incompatibility.
Other roles for SRK?
Receptor protein kinases often function in multiple processes. For example, the LRR-RLK ERECTA functions in stomatal patterning, anther development, cell proliferation in vegetative tissues and biotic interactions , whereas the CrRLK1L receptor kinase FERONIA functions in fertilization, cell elongation, polarized growth and response to fungal pathogens .
As described above, only two roles have been demonstrated for SRK, in self-incompatibility and pistil development. However, it had been proposed that SRK also functions in interspecific incompatibility, i.e. inhibition of pollen derived from other species (heterospecific pollen) . This interspecific incompatibility has been observed in pollinations between Brassica species and in intergeneric pollinations between Brassica and Arabidopsis, but its molecular basis is not known. Nevertheless, cytological observations and reciprocal pollinations have suggested that the S-locus itself, presumably SRK, is responsible for the recognition and inhibition of heterospecific pollen by the stigma , and that interspecific incompatibility and self-incompatibility share a stigma-based pollen-inhibition pathway. Indeed, the inhibition of heterospecific pollen is cytologically identical with that of conspecific self pollen, i.e. it is typically manifested at the stigma surface by the failure of hydration and germination of pollen grains. Moreover, in pollinations between a self-incompatible and a self-compatible species, interspecific incompatibility is often unilateral, such that heterospecific pollen is inhibited on the stigmas of the self-incompatible species, but not in the reciprocal pollination.
The possible role for SRK in the inhibition of heterospecific pollen was tested in the case of intergeneric unilateral incompatibility between Brassica oleracea and A. thaliana . The A. thaliana stigma does not discriminate against pollen from other genera of the Brassicaceae, and it inhibits the hydration and germination only of pollen from species outside the Brassicaceae family [43,44]. Thus Brassica pollen grains hydrate and germinate efficiently on A. thaliana stigmas and produce pollen tubes that grow into the pistil  (Figure 2). Interestingly, these Brassica pollen tubes were not directed to the A. thaliana ovules (Figure 2D), as occurs in A. thaliana pistils pollinated with conspecific pollen (Figure 2E), indicating that interspecific incompatibility operates in the ovary of A. thaliana, because of the inability of Brassica pollen tubes to perceive attractants emanating from the A. thaliana ovule . Indeed, more recent in vitro pollen-tube-guidance assays demonstrated that A. thaliana ovules produce diffusible species-specific chemoattractants . In contrast, A. thaliana pollen grains are inhibited at the surface of mature Brassica stigmas (Figure 2B). However, immature Brassica stigmas are fully receptive to A. thaliana (Figure 2C), and the ability of these stigmas to inhibit A. thaliana pollen is first observed 1 day before anthesis, coincident with the acquisition of the ability to reject self pollen by these stigmas. However, despite these similarities to self-incompatibility, the stigmas of a self-compatible B. oleracea strain carrying a null mutation in SRK were still able to inhibit A. thaliana pollen. This result demonstrates that a functional SRK is not required for pollen inhibition in this particular case of unilateral intergeneric incompatibility .
Intergeneric pollinations between Brassica and A. thaliana
Thus no evidence has been found so far for the convergence of self-incompatibility and interspecific incompatibility. It should be noted that analysis of A. thaliana pollination revealed that the very early events of the pollination process, namely adhesion and hydration of pollen grains at the stigma surface, are mediated by lipophilic molecules located on the surface of pollen grains . Consequently, it is possible that in wide pollinations, i.e. between plants belonging to different genera and families, the inability of pollen grains to hydrate does not require receptor-based signalling in the stigma epidermis.
Much progress has been made on understanding how SRK allows the stigma to discriminate between self and non-self pollen. However, much still needs to be learned. Regarding the recognition phase of the self-incompatibility response, it is important to determine the crystal structure of the receptor extracellular domain and its ligand-binding pocket, analyse a large number of receptor and ligand variants to derive general rules for identification of the specific amino acids that determine specificity, and possibly gain insight into the diversification and co-evolution of SRK and SCR variants and the mechanism underlying the generation of new self-incompatibility specificities. Also critical is deciphering the signalling cascade that is triggered upon binding of the receptor to its SCR ligand and unravelling the role of auxin in self-incompatibility. In addition to elucidating mechanistic aspects of the self-incompatibility response, these studies will reveal any overlaps that might exist between the self-incompatibility signalling pathway and other plant signalling pathways and possibly open new avenues of research into the biological roles of the orphan receptors in the SD-RLK family.
Regulation of Fertilization and Early Seed Development: A Biochemical Society Focused Meeting held at the University of Bath, U.K., 11–13 September 2013. Organized and Edited by James Doughty (University of Bath, U.K.) and Thomas Dresselhaus (University of Regensburg, Germany).
The present article is based on work supported by the National Science Foundation [grant number IOS-1146725]. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the article.