The S-domain serine/threonine receptor-like kinases (SRLKs) comprise one of the largest and most rapidly expanding subfamilies in the plant receptor-like/Pelle kinase (RLKs) family. The founding member of this subfamily, the S-locus receptor kinase (SRK), functions as the female determinant of specificity in the self-incompatibility (SI) responses of crucifers. Two classes of proteins resembling the extracellular S domain (designated S-domain receptor-like proteins, SRLPs) or the intracellular kinase domain (designated S-domain receptor-like cytoplasmic kinases, SRLCKs) of SRK are also ubiquitous in land plants, indicating that the SRLKs are composite molecules that originated by domain fusion of the two component proteins. Here, we explored the origin and diversification of SRLKs by phylogenomic methods.
Based on the distribution patterns of SRLKs and SRLCKs in a reconciled species-domain tree, a maximum parsimony model was then established for simultaneously inferring and dating gene duplication/loss and fusion /fission events in SRLK evolution. Various SRK alleles from crucifer species were then included in our phylogenetic analyses to infer the origination of SRKs by identifying the proper outgroups.
Two gene fusion events were inferred and the major gene fusion event occurred in the common ancestor of land plants generated almost all of extant SRLKs. The functional diversification of duplicated SRLKs was illustrated by molecular evolution analyses of SRKs. Our findings support that SRKs originated as two ancient haplotypes derived from a pair of tandem duplicate genes through random regulatory neo-/sub- functionalization in the common ancestor of the Brassicaceae.
The “S domain” (SD) was initially defined by the S-locus glycoprotein (SLG) and the S-locus receptor kinase (SRK), which are encoded by two closely-linked genes in the Brassica self-incompatibility (SI) determining locus, the S locus. SLG, which was the first S locus-derived gene identified, encodes a secreted glycoprotein, whereas SRK encodes a transmembrane receptor kinase with an extracellular domain that shares extensive sequence similarity with SLG . SRK is the female determinant of specificity in “self-pollen” recognition, and in self-incompatible species of the Brassicaceae (crucifers), the S haplotype-specific binding of SRK to its cognate pollen-borne ligand S locus cystein rich protein/S locus protein 11 (SCR/SP11) activates the SRK and triggers a signaling cascade that culminates in “self-pollen” rejection [2, 3]. The extracellular S domain of SRK is responsible for ligand binding [4, 5], whereas the intracellular kinase domain (KD) is thought to translate this signal into a cellular response by phosphorylating Arm Repeat Containing (ARC1) protein, an E3 ligase involved in protein ubiquitination [6, 7]. With the increasing availability of sequenced plant genomes, it has been realized that proteins having a structure resembling SRKs, designated S-domain receptor-like kinases (hereafter SRLKs), form one of the largest and most-rapidly expanding subfamilies within the plant receptor-like/Pelle kinase superfamily [8–11]. In addition, a large group of receptor-like cytoplasmic kinases (RLCKs) resembling the intracellular kinase domains of SRLKs but lacking the extracellular S-domain (designated S-domain receptor-like cytoplasmic kinase, SRLCKs) were also defined by their close phylogenetic relationship to the kinase domains of SRLKs [8, 12]. Interestingly, another class of proteins resembling SLGs, designated S-domain receptor-like proteins, SRLPs, is also ubiquitous in plants [13–15], suggesting that the composite SRLKs likely originated by fusion of the two split component proteins. Phylogenetic analyses of the kinase domains in Arabidopsis thaliana also suggested that SRLKs are not monophyletic and probably arose via independent recruitment of S domains .
Gene fusion is considered to be an important evolutionary path to create novelty in protein architectures (the linear arrangement of protein domains) and functions by forming composite proteins and linking components of extant signaling/biochemical pathways . In agreement with this notion, a number of chimeric genes generated by gene fusions have been reported to have important functions [17–20]. Based on the assumption that selection favors fusions of functionally-related proteins, identification of fusion-link (split proteins in some genomes and fused proteins in other genomes) has initially been used to predict genome-wide protein interactions and functions in the extremely compact genomes of prokaryotes and yeasts [16, 21], and more recently in the more complex eukaryotic genomes [22, 23]. Based on the distribution patterns of the composite proteins and the split proteins in either the species trees  or the domain trees , gene fusion events were inferred in a large number of sequenced genomes. Furthermore, a maximum parsimony algorithm has been established to analyze the evolution of protein architectures, in particular domain fusion and fission, based on the inferred ancestral architecture at each node in the species trees  or domain trees [25, 26]. In plants, because only the Arabidopsis and rice genomes have been included in such studies, very little is known about the evolution of domain architecture in other plant genomes. Despite the fact that multiple-domain proteins in super-protein families are normally composed of abundant and versatile domains and tend to undergo more independent gene fusion/fission events , analysis of gene fusion/fission events in a large gene subfamily such as SRLK subfamily is still lacking.
As the only members of the SRLK subfamily whose function is known, SRKs are well suited to investigate the functional diversification of SRLKs. In Brassica species but not Arabidopsis species, SRKs may be clearly divided into two classes, class I and class II . Moreover, phylogenetic analyses of SRK kinase domains showed that SRKs from Brassica species are not monophyletic, having descended from only two of the lineages that were presumably present in the Arabidopsis-Brassica ancestor, and that diversification of the Brassica S haplotypes took place after the separation of the two genera . Furthermore, theoretical analyses have long predicted that SI could have been first expressed in a two S-haplotype system causing incomplete suppression of selfing, with further differentiation among S haplotypes and enhancement of SI expression having evolved subsequently [30, 31]. This long-held hypothesis awaits further elaboration by molecular evolution studies.
In this study, we first retrieved SRLKs from five sequenced genomes representing the major lineages of land plants. SRLCKs were then delineated by their close phylogenetic relationship to SRLKs based on a maximum likelihood (ML) kinase domain tree. On the basis of a reconciled species-domain tree including both SRLKs and SRLCKs, gene duplication/loss and fusion events in SRLK evolution were inferred and dated by integrating the maximum parsimony ancestral architecture inference algorithm [25, 26] into the widely applied gene duplication/loss model . In addition, the origination of SRKs in the Brassicaceae was explored by reconstruction of SRK kinase domain phylogeny in the context of SRLK evolution.
SRLKs emerged in early land plants and expanded greatly in Angiosperms
SRLKs are composed principally of three modular domains in a configuration of S domain (SD)-transmembrane domain (TM)-kinase domain (KD). SDs are further divided into three subdomains in a configuration of B_lectin-SLG-PAN_APPLE (Figure 1). The B_lectin and PAN_APPLE domains have been proposed to be important for protein structure and dimerization of Brassica SRKs, while the hypervarible SLG domain plays a key role in SCR binding [4, 5]. Transmembrane domain (TM) prediction is not always accurate, thus TM will not be considered in our protein architecture analyses. Proteins with stand-alone SLG and PAN_APPLE domains are rare, whereas B_lectin domain proteins are more abundant, particularly in the spikemoss genome (Figure 1).
A total of 253 SRLK and 19 SRLP sequences with the domain architecture typical of SRKs (B_lectin-SLG-PAN_APPLE-TM-KD) or SLGs (B_lectin-SLG-PAN_APPLE) were retrieved and are hereafter referred to as typical SRLKs and SRLPs respectively (Figure 1 and Additional file 1: Table S1). Other combinations of B_lectin, SLG, PAN_APPLE, and KD domains are also ubiquitous in different plant species (Figure 1). These combinations may represent either the precursors of typical SRLKs/SRLPs or the degenerated products of the typical SRLKs/SRLPs. Our search also retrieved stand-alone B_Lectin and PAN_APPLE sequences in the genomes of green algae Chlamydomonas reinhardtii and Ostreococcus tauri, whereas sequences of stand-alone SLG domains or any combination of B_lectin, PAN_APPLE, SLG, and KD domains were not detected. In view of the fact that plant RLKs were likely generated after the divergence of land plants from green algae , we tentatively included sequences containing combinations of at least two of the four modular domains, adding 139 SRLK and 30 SRLP sequences with atypical domain architectures to our dataset. Nine proteins with stand-alone SLG, which characterize S-domains, were also included as SRLPs. In total, our dataset included 7, 9, 129, 38 and 209 SRLKs and 4, 12, 10, 5 and 27 SRLPs from the genomes of Physcomitrella patens (moss), Selaginella moellendorffii (spikemoss), Oryza sativa (rice), Arabidopsis thaliana (Arabidopsis), and Populus trichocarpa (poplar) respectively (Additional file 1: Table S1). In moss and spikemoss, a relatively small number SRLKs (7 and 9, respectively) are found, and SRLKs continued to expand immediately after the divergence of angiosperms, since there are 23.8, 19.3 and 33.9 times (normalized by genome size) as many members in rice, Arabidopsis and poplar, respectively, compared with moss.
Kinase domain tree and delineation of SRLCKs
To delineate SRLCKs by the kinase-domain tree, KD sequences from 392 SRLKs in our dataset and total 96 RLCKs of the five species from Phytozome V9.0 that were the best hits to the members of the gene family HOM000017 in PLAZA 2.5 database were then used to construct a maximum likelihood (ML) phylogenetic tree (Figure 2 and Additional file 2: Figure S1). The basal nodes of the phylogenetic tree are composed of 17 RLCKs, while two well-supported monophylectic groups (with a 100 bootstrap) in land plants are evident. The first group consists of 390 SRLKs, while the second group consists all 79 RLCKs as well as 2 moss SRLKs. Therefore, the 79 RLCKs are hereafter designated as SRLCKs (Additional file 1: Table S1 and Additional file 2: Figure S1). Considering protein architectures, SRLKs are not monophylectic but form two groups, the major group containing 390 SRLKs and the minor group containing 2 moss SRLKs. In the major group, two large subgroups correspond to the previously-identified SD-1 and SD-2 S-domain RLKs [8, 9], which are composed of 202 SRLKs, and 188 SRLKs, respectively.
SD-1 SRLKs are specific for angiosperms, indicating that this group is approximately 140 million years old . In contrast to SD-1, SD-2 is a very diverse group of SRLKs from all species with well-dissolved clades. Given that the oldest evidence for the existence of vascular plants is found in Upper Ordovician, SD-2 SRLKs are inferred to be more than 450 million years old (Figure 2 and Additional file 2: Figure S1). Based principally on the topology of the trees, clade support values and branch length, we tentatively defined 6 (SD-1a, SD-1b, SD-1c, SD-1d, SD-1e, SD-1f) and 7(SD-2a, SD-2b, SD-2c, SD-2d, SD-2e, SD-2f and SD-2g) subgroups in each of the SD-1 and SD-2 groups. Interestingly, the A. thaliana SRK falls in the SD-1b subgroup, sharing a most recent common ancestor with 3 Arabidopsis, 5 rice, and 2 poplar SRLKs (Figure 2 and Additional file 2: Figure S1). More importantly, except for the two moss SRLKs, all SRLKs or SRLCKs cluster together, which strongly suggests that extant SRLKs are likely derived from one major S-domain recruitment event in land plant evolution.
Inference and dating of gene duplication/loss and fusion/fission events in the evolutionary history of SRLKs
The gene duplication-loss model embedded in Notung has been used to infer and date gene duplication/loss events in RLK evolution . After a reconciled species-domain tree was generated by Notung, we inferred the ancestral architectures of all nodes along the reconciled tree according to our model (Figure 3 and Additional file 3: Figure S2). Gene duplication/loss and fusion/fission events were inferred and dated. SRLKs were shown to have a rapid gene expansion, whereas SRLCKs tended to have undergone a decay comparing with the inferred number of ancestors in early land plants (Figure 4). A leap of common ancestor number from 10 in the ancestor of land plants to 34 of angiosperms represents a critical stage in SRLK expansion, which coincides with the emergence of angiosperms and the rapid increase in gene duplication events. In the past 450 million years after the divergence of moss and the embryophyta ancestor, two duplications of SRLKs was inferred, while as many as 172 duplications were inferred to have occurred in the past 100 million years after the divergence of poplar and its rosid ancestor. In Arabidopsis, the results suggest rapid birth and death of SRLKs, and the loss of most SRLCKs (Figure 4).
Based on the distribution patterns of SRLKs and SRLCKs in the reconciled tree, gene fusion/fission events in the evolutionary history of SRLKs were also inferred (Figure 2 and Additional file 3: Figure S2). In total, 2 gene fusion events were inferred, while no fission event was detected. The major ancient gene fusion event occurred in the common ancestor of land plants and generated the ancestral gene of 390 extant SRLKs. The other minor gene fusion event occurred after moss have diverged from the common ancestor of land plants, generating two extant moss SRLKs. Notably, scarcity of gene fusion and the lack of fission events in the evolution of SRLKs subfamily suggests that functional diversification of SRLKs is principally driven by sub-or neo-funtionalization of duplicated genes.
Origination of brassicaceae SRKs in the context of SRLK evolution
The A. thaliana SRK falls within the SD-1b group, a small monophyletic group in angiosperms that includes 10 non-SRK members, which are good candidates to investigate the functional diversification of SRLKs (Figure 2). We thus performed phylogenetic analysis of SRKs by including KD sequences of 47 SRKs from the Brassica/Raphanus and Arabidopsis/Capsella lineages as well as SD-1b members (Additional file 4: Table S2 and Figure 5A). A functional SRK from self-fertile A. thaliana accession Wei was included to replace the ΨSRK (At4g21370), because its function in SI has been demonstrated . In addition, a likely functional SRK sequence (CruSRK) from the newly-evolved self-fertile species Capsella rubella is also included .
The topology of our rooted ML kinase domain tree is similar to that of an unrooted ML kinase domain tree based on nucleotide sequences . Three separate well-supported clades, Brassica class I, Arabidopsis/Capsella, and Brassica class II, are evident (Figure 5A). The A. thalina non-SRK SRLKs, Arabidopsis Receptor Kinase (ARK1/ARK2/ARK3), do not fall within the outgroups in the kinase domain tree (Figure 5A), but form a well-supported orthologous group in the Brassicaceae with Brassica class-II SRKs. In contrast, they do fall within outgroups in S-domain trees in our (Additional file 5: Figure S3) and other studies [37–39]. A. thaliana SRK (At4g21370) and most SRKs from A. lyrata, A. halleri, and Capsella grandiflora also form a distinct orthologous group in the Arabidopsis/Capsella lineage. In addition, Brassica class-I SRKs and Arabidopsis/Capsella SRKs (except for AlSRK20) appear to form a large orthologous group in Brassicaceae, albeit with a relatively low bootstrap value (Figure 5A). A. thaliana SRK (At4g21370) and ARK3 (At4g21380) are located in different orthologous groups and are arranged in tandem (Figure 5B). The same arrangement of SRK and ARK3 orthologs was also retained in all characterized S haplotypes of A. lyrata (Figure 5B).
Discussion and conclusions
The architecture of the SRLKs was likely established after the divergence of land plants from green algae approximately 1000 million years ago, but before the divergence of vascular plant lineage from the moss lineage. Consistent with their predicted function in perceiving various external signals, the SDs of SRLKs are very variable in both sequence and architecture. Because of the highly variable nature of the SDs and the resulting poor sequence alignments, it was difficult to use these domains for investigating the trajectory of SRLK evolution. In contrast, kinase-domain sequences are more conservative likely due to constrains imposed by the requisite interactions with other signaling partners, and were thus used in our study to simplify the interpretation of SRLK evolution.
By integrating the ancestral architecture inference algorism  into the widely applied gene duplication-loss parsimony model [8, 32], we established a maximum parsimony model suitable to infer and date gene duplication/loss and fusion/fission events in SRLK evolution (Figure 3). Our results suggest that almost all (except for 2 moss SRLKs) SRLKs of land plants are derived from a single ancient domain fusion event. Continuous expansion of SRLKs by gene duplication has played pivotal roles in shaping the phylogenies of extant SRLKs. In contrast with previous interpretations based merely on topology of the phylogenetic trees [8–10], we show that SD-1 and SD-2 group SRLKs were generated by the same ancient gene-fusion event that likely occurred in the common ancestor of land plants. Mis-annotation of genomic sequences, however, may be accounted for the inconsistence between our results and the published papers [8–10]. When the poplar genome of Phytozome v6.0, in which only 10% annotated gene models are supported by full length cDNAs, were used for detecting fusion/fission events, we could detect 5 fusion events and 23 fission events occurred in poplar. However, no fusion or fission event was found using the updated poplar genome of Phytozome v9.0, in which 218 out of 228 SRLKs/RLCKs were supported by assembly ESTs (Additional file 1: Table S1). As in all other such studies, we assigned an equal cost for gene fusion, fission, duplication, loss, or speciation in order to avoid prior bias stemming from uncertainties relating to the relative frequency of these events  and their dependence on the particular genomes investigated [22, 23]. Since the distribution patterns of the composite SRLKs and the split SRLCKs on the reconciled tree are critical for our analysis, expansion and high-rate retention of both the composite SRLK and split SRLCK genes is essential. We might have underestimated the actual number of gene fusion/fission events in our analyses because the domain architectures (composite or split) of the most recent common ancestors (MRCAs) at the leaf nodes with lost genes can only be inferred by the parsimony principle (Additional file 3: Figure S2). Similar requirements can be largely fulfilled in most RLK subfamilies such as Lysin motif-type RLKs . We thus propose that our method could be extrapolated to analyze gene fusion/fission events in other multiple-domain super-protein families.
Functional diversification following the origination of multiple-domain proteins is a common theme in studies of protein family evolution [41–43]. Although great efforts have been devoted to investigate the diversification of SRKs by reconstructing their phylogeny, such studies are limited by the lack of suitable outgroups . Utilizing SD-1b members as outgroups, we show that in the Brassicaeae, SRKs do not form a monophyletic clade. More intriguingly, ARK1/ARK2/ARK3 are confidently clustered with Brassica class-II SRKs, forming a distinct orthologous group in the Brassicaceae (Figure 5B). This orthologous relationship between ARK1/ARK2/ARK3 and class-II SRKs has not been previously revealed because S-domain trees were used in most studies [28, 38, 39]. Even in studies that used the kinase-domain trees, neither ARK1/ARK2/ARK3  nor Brassica SRKs were included [37, 38]. Furthermore, A. thaliana SRK confidently clustered with most SRKs of Arabidopsis/Capsella species, forming another distinct orthologous group in the Arabidposis/Capsella lineage (Figure 5B). Since SRK and ARK3 are arranged in tandem in the S haplotypes of Arabidopsis species, we propose the following model on the origins of SRKs. An ancient duplication in the common ancestor of Brassicaceae produced two tandemly-arranged paralogous genes, designated as ancestor of SRKI (A_SRKI) and ancestor of SRKII (A_SRKII). Subsequently, random neo-fuctionalzation and/or sub-funtionalization of A_SRKI and A_SRKII produced two ancient S haplotypes, from which SRKs might have been derived by further diversification (Figure 6). This model provides a mechanism for the establishment of the long-proposed ancient two-S haplotype SI system in the common ancestor of the Brassicaceae [30, 31]. However, the clustering of Arabidopsis/Capsella SRKs with Brassica/Raphanus class-I SRKs in the same orthologous group remains ambiguous because of the relatively low support value. The different gene orders of SLGs and SRKs in known class-I and class-II S haplotypes (Figure 7) and the inference that SLGs were derived from an ancient duplicated copy of SRK [1, 36, 44], support the notion that class-I and class-II SRKs might also be derived from two tandemly duplicated paralogous genes. Similar examples of the evolution of functional orthologous genes from ancient paralogs have been reported for AGAMOUS and PLENA as well as SRnases in diploid strawberry . Paralogs produced by tandem duplications such as A_SRKI and A_SRKII might be prone to this random evolution, because they are free from constrains of different genomic contexts. Regulatory neo-functionalization is the most likely evolutionary scenario for paralogs produced by tandem duplications . However, no overlap in the expression domains of ARK3 and SRKa in self-incompatible A. lyrata has been observed in either vegetative or floral tissues (Figure 8), suggesting that regulatory sub-functionalization played a significant role in SRK evolution. Nevertheless, certain SRK variants are expressed in leaf tissue , indicating that regulatory neo-functionalization also played a role. Together with the fact that Arabidopsis ARK3 is orthologous to Brassica class-II SRKs, our findings indicate that the biochemical functions of SRKs and ARK3, and probably other SRLKs in the SD-1b group, are likely the same or very similar. This conclusion is in line with the existence of a conserved DUF 3660 motif between SDs and KDs in ARK1, ARK2, ARK3, and all SRKs, and particularly with the findings that ARK1, ARK2, and ARK3 interact with and phosphorylate a number of the Arabidopsis plant U-box/ARM-repeat (AtPUB-ARM), ARC1 homologs of A. thaliana. In view of that ARKs and other SRLKs may be involved in plant innate immunity responses [50–53], the overlap of SRK signaling with that of plant immunity mediated by ARK3 and other SD-1b group SRLKs suggested by our molecular evolution study is intriguing and should be pursued by further experimental investigations.
Physcomitrella patens (moss), Selaginella moellendorffii (spikemoss), Oryza sativa (rice), Arabidopsis thaliana (Arabidopsis), and Populus trichocarpa (poplar), which represent the major lineages in land plant evolution, were used in this study. The annotated protein sequences of the 5 sequenced genomes were downloaded (http://www.phytozome.net). To identify SRLK and SRLP sequences, an HMMer search was performed by the standard profiles of the modular domains of S-domains , the B_lectin, SLG, and PAN_APPLE domains. After searching the sequences of primary screening against the Pfam database with the established “trusted cut off”, we detected in the moss, spikemoss, rice, Arabidopsis, and poplar genomes, respectively, 13, 136, 134, 47, and 244 B_lectin containing proteins; 9, 13, 111, 36, and 194 SLG-containing proteins; as well as 2, 10, 102, 37 and 167 PAN_APPLE-containing proteins (Additional file 1: Table S1). Because proteins belonging to the same homologous group are readily and confidently retrieved from PLAZA 2.0 database (http://bioinformatics.psb.ugent.be/plaza/), we thus retrieved RLCK sequences from the homologous group (HOM000017, which contain all SRLKs identified from Phytozome 9.0 ), and then updated each RLCK sequence with the best hit from Phytozome v9.0 using BLASTP. At last, these hits filtered for only one kinase domain using Pfam, were used as candidates for SRLCKs (Additional file 1: Table S1).
Sequence alignment and phylogenetic analysis
The composite SRLKs and their split component proteins, SRLCKs and SRLPs (Figure 3A), are ubiquitous in plants suggesting the occurence of fusion and/or fission events. To explore the trajectory of SRLK evolution, a kinase domain tree including both SRLKs and SRLCKs was constructed. The amino-acid sequences of the kinase domains of 392 SRLKs and 96 RLCKs were aligned using ClustalX (Version 2.0) with Gonnet 250 protein weight matrix and the pairwise parameters of gap opening 10 and gap extension 0.2  (Additional file 6: Sporting dataset 1) Arabidopsis homologs of Right Open Reading 1 (RIO1) family kinase (At 5 g37350 and At 2 g24990) were used as outgroups . ProtTest v2.4  was used to select the best-fit model of protein evolution for the alignment. Then, according to the best-fit model predicted by ProtTest v2.4, a rooted maximum likelihood (ML) tree was constructed with the JTT substitution model using the PhyML v3.0 online program, and the support of interior branches was assessed with the aLRT bootstrap method . Finally, the phylogenetic tree was displayed and edited using MEGA v5.0 .
According to this phylogenetic tree, the A. thaliana SRK (At4g21370), ARK1 (At1g65790), ARK2 (At1g65800), and ARK3 (At4g21380) sequences as well as 7 SRLKs from rice and poplar form a well-supported subgroup. Kinase domain sequences from these 11 proteins and Brassicaceae 47 SRK sequences (Additional file 4: Table S2 and Additional file 7: Supporting dataset 2) were retrieved from Uniprot and were included to construct another ML tree using PhyML v3.0 online program. We included SRKs from as many Brassicacea species as possible, including Brassica oleracea, B. rapa, B. napus, and Raphanus sativus in the Brassica/Raphanus lineage and A. thaliana, A. lyrata, A. halleri, Capsella grandiflora, and C. rubella in the Arabidopsis/Capsella lineage.
Inference and dating of gene duplication/loss and fusion/fission events
A reconciled species-domain tree was generated using the Notung program, which offers a unified framework for incorporating duplication-loss parsimony into phylogenetic analysis . After reconciling the ML domain tree with the species tree of the five land plant species constructed using the NCBI Taxonomy Browser (http://www.ncbi.nlm.nih.gov/Taxonomy/CommonTree/wwwcmt.cgi), gene duplicate- on/loss and fusion events were inferred and dated. Based on the reconciled tree, the ancestral protein architectures were inferred by a modified maximum parsimonious protein ancestral architecture inference algorithm [25, 26]. The extant protein architectures at the leaves are used to initialize the tree. Instead of traversing the whole tree, we infer the ancestral protein architecture of MRCA at each node sequentially from the leaves to the root. To avoid prior bias, gene duplication/loss, gene fusion/fission, and speciation were assigned an equal cost of 1. However, when a gene fusion/fission event occurred at a node, a gene duplication/speciation event must also have occurred, thus the node is assigned a cost of 2. In the bifurcating terminal branches of the reconciled species-domain tree, three configurations could be detected (Figure 3B and 3C). When both leaf nodes are of the same domain architecture (SRLK or SRLCK), we inferred that the internal node has a MRCA of the same architecture. The two proteins could be paralogs derived from a duplication event or orthologs generated by speciation events (Figure 3B). When the two leaf nodes are of different domain architectures (one SRLK and one SRLCK), the architectures of the MRCAs were inferred by those of the deeper branches. The trees were traversed twice and the ancestral architectures yielding the lowest cost were selected (Figure 3C). After the architectures of these outer nodes were inferred, they were treated as leaves to initiate another round of ancestral architecture-inferring process until the ancestral architectures at all inner nodes were inferred, after which gene duplication/loss and fusion/loss events were inferred and dated.
RNA extraction and reverse transcription polymerase chain reaction (RT-PCR)
RT-PCR was used to examine the spatial expression of AlSRK and AlARKs in A. lyrata organs. Total RNA from root, stem, leaf, stamen, style, and stigma tissues of A. lyrata was isolated using the Trizol reagent (Invitrogen, USA) according to the manufacturer’s instructions. The residual genomic DNA in the total RNA was removed by treatment with RNase-free DNaseI and the total RNA was further purified with phenol chloroform-isoamyl alcohol. RT-PCR was performed using SuperScript™II RNase HI Reverse Transcriptase (Invitrogen, USA) using the following primer pairs: 5′-GACAACGCGTGTGAGACCTAT-3′ and 5′-CATTAGG AGCTGCAGTTGCTC-3′ for AlSRKa, and 5′-GACCAATGCGATGATTACAAAG -3′ and 5′-CAGTAGCTGTTTCAATTAGT-3′ for AlARK3. The PCR conditions of AlSRKa/AlARK3 were 94°C for 3 min followed by 40 cycles of the following: 94°C for 30 s, 58°C/52°C for 30 s, and 72°C for 40 s. The amplification products were then analyzed with agarose gel electrophoresis.
Availability of supporting data
The data sets supporting the results of this article are included within the article (and its additional files)
S-domain serine/threonine receptor-like kinase
plant receptor-like/Pelle kinase
S-locus receptor kinase
S-domain receptor-like protein
S-domain receptor-like cytoplasmic kinases
S locus cystein rich protein/S locus protein 11
Arm Repeat Containing
Arabidopsis Receptor Kinase
Most recent common ancestor
Common ancestor of Brassicaceae SRKs
Ancestor of SRKI
Ancestor of SRKII
Arabidopsis plant U-box/ARM-repeat
Right Open Reading 1.
We thank Professor June Nasrallah and Professor Mikhail Nasrallah (Cornell University) for critical reading of the manuscript and providing A. lyrata seeds. This work was supported by the China Science Foundation (NSFC) (31070207) and the Major Research Plan from the Ministry of Science and Technology of China (No. 2013CB945100) to PL.
Department of Ecology, College of Resources and Environmental Sciences, China Agricultural University
Stein JC, Howlett B, Boyes DC, Nasrallah ME, Nasrallah JB: Molecular cloning of a putative receptor protein kinase gene encoded at the self-incompatibility locus ofBrassica oleracea.Proc Natl Acad Sci USA 1991,88(19):8816–8820.PubMedView Article
Kachroo A, Schopfer CR, Nasrallah ME, Nasrallah JB: Allele-specific receptor-ligand interactions in Brassica self-incompatibility.Science 2001,293(5536):1824–1826.PubMedView Article
Takayama S, Shimosato H, Shiba H, Funato M, Che FS, Watanabe M, Iwano M, Isogai A: Direct ligand-receptor complex interaction controls Brassica self-incompatibility.Nature 2001,413(6855):534–538.PubMedView Article
Kemp BP, Doughty J: S cysteine-rich (SCR) binding domain analysis of the Brassica self-incompatibility S-locus receptor kinase.New Phytol 2007,175(4):619–629.PubMedView Article
Naithani S, Chookajorn T, Ripoll DR, Nasrallah JB: Structural modules for receptor dimerization in the S-locus receptor kinase extracellular domain.Proc Natl Acad Sci USA 2007,104(29):12211–12216.PubMedView Article
Stone SL, Anderson EM, Mullen RT, Goring DR: ARC1 is an E3 ubiquitin ligase and promotes the ubiquitination of proteins during the rejection of self-incompatible Brassica pollen.Plant Cell 2003,15(4):885–898.PubMedView Article
Stone SL, Arnoldo M, Goring DR: A breakdown of Brassica self-incompatibility in ARC1 antisense transgenic plants.Science 1999,286(5445):1729–1731.PubMedView Article
Lehti-Shiu MD, Zou C, Hanada K, Shiu SH: Evolutionary history and stress regulation of plant receptor-like kinase/pelle genes.Plant Physiol 2009,150(1):12–26.PubMedView Article
Shiu SH, Karlowski WM, Pan R, Tzeng YH, Mayer KF, Li WH: Comparative analysis of the receptor-like kinase family in Arabidopsis and rice.Plant Cell 2004,16(5):1220–1234.PubMedView Article
Shiu SH, Bleecker AB: Expansion of the receptor-like kinase/Pelle gene family and receptor-like proteins in Arabidopsis.Plant Physiol 2003,132(2):530–543.PubMedView Article
Shiu SH, Bleecker AB: Receptor-like kinases from Arabidopsis form a monophyletic gene family related to animal receptor kinases.Proc Natl Acad Sci USA 2001,98(19):10763–10768.PubMedView Article
Lalonde BA, Nasrallah ME, Dwyer KG, Chen CH, Barlow B, Nasrallah JB: A highly conserved Brassica gene with homology to the S-locus-specific glycoprotein structural gene.Plant Cell 1989,1(2):249–258.PubMed
Sato K, Nishio T, Kimura R, Kusaba M, Suzuki T, Hatakeyama K, Ockendon DJ, Satta Y: Coevolution of the S-locus genes SRK, SLG and SP11/SCR inBrassica oleraceaandB. rapa.Genetics 2002,162(2):931–940.PubMed
Suzuki G, Watanabe M, Toriyama K, Isogai A, Hinata K: Molecular cloning of members of the S-multigene family in self-incompatibleBrassica campestris L.Plant Cell Physiol 1995,36(7):1273–1280.PubMed
Marcotte EM, Pellegrini M, Ng HL, Rice DW, Yeates TO, Eisenberg D: Detecting protein function and protein-protein interactions from genome sequences.Science 1999,285(5428):751–753.PubMedView Article
Kusaba M, Nishio T, Satta Y, Hinata K, Ockendon D: Striking sequence similarity in inter- and intra-specific comparisons of class I SLG alleles fromBrassica oleraceaandBrassica campestris: implications for the evolution and recognition mechanism.Proc Natl Acad Sci USA 1997,94(14):7673–7678.PubMedView Article
Long M, Langley CH: Natural selection and the origin of jingwei, a chimeric processed functional gene in Drosophila.Science 1993,260(5104):91–95.PubMedView Article
Rogers RL, Bedford T, Lyons AM, Hartl DL: Adaptive impact of the chimeric gene Quetzalcoatl inDrosophila melanogaster.Proc Natl Acad Sci USA 2010,107(24):10943–10948.PubMedView Article
Suetsugu N, Mittmann F, Wagner G, Hughes J, Wada M: A chimeric photoreceptor gene, NEOCHROME, has arisen twice during plant evolution.Proc Natl Acad Sci USA 2005,102(38):13705–13709.PubMedView Article
Snel B, Bork P, Huynen M: Genome evolution: gene fusion versus gene fission.Trends Genet 2000,16(1):9–11.PubMedView Article
Nakamura Y, Itoh T, Martin W: Rate and polarity of gene fusion and fission inOryza sativaandArabidopsis thaliana.Mol Biol Evol 2007,24(1):110–121.PubMedView Article
Kummerfeld SK, Teichmann SA: Relative rates of gene fusion and fission in multi-domain proteins.Trends Genet 2005,21(1):25–30.PubMedView Article
Yanai I, Wolf YI, Koonin EV: Evolution of gene fusions: horizontal transfer versus independent events.Genome Biol 2002.,3(5): research0024
Fong JH, Geer LY, Panchenko AR, Bryant SH: Modeling the evolution of protein domain architectures using maximum parsimony.J Mol Biol 2007,366(1):307–315.PubMedView Article
Forslund K, Henricson A, Hollich V, Sonnhammer EL: Domain tree-based analysis of protein architecture evolution.Mol Biol Evol 2008,25(2):254–264.PubMedView Article
Vogel C, Teichmann SA, Pereira-Leal J: The relationship between domain duplication and recombination.J Mol Biol 2005,346(1):355–365.PubMedView Article
Fobis-Loisy I, Miege C, Gaude T: Molecular evolution of the S-locus controlling mating in the brassicaceae.Plant Biol (Stuttg) 2004,6(2):109–118.View Article
Edh K, Widen B, Ceplitis A: The evolution and diversification of S-locus haplotypes in the Brassicaceae family.Genetics 2009,181(3):977–984.PubMedView Article
Uyenoyama MK: On the evolution of genetic incompatibility systems. IV. Modification of response to an existing antigen polymorphism under partial selfing.Theor Popul Biol 1988,34(3):347–377.PubMedView Article
Bateman AJ: Self-incompatibility systems in agiosperm.1. Theory.Heredity 1952,6(3):285–310.View Article
Chen K, Durand D, Farach-Colton M: NOTUNG: a program for dating gene duplications and optimizing gene family trees.J Comput Biol 2000,7(3–4):429–447.PubMedView Article
Moore MJ, Bell CD, Soltis PS, Soltis DE: Using plastid genome-scale data to resolve enigmatic relationships among basal angiosperms.Proc Natl Acad Sci USA 2007,104(49):19363–19368.PubMedView Article
Tsuchimatsu T, Suwabe K, Shimizu-Inatsugi R, Isokawa S, Pavlidis P, Stadler T, Suzuki G, Takayama S, Watanabe M, Shimizu KK: Evolution of self-compatibility in Arabidopsis by a mutation in the male specificity gene.Nature 2010,464(7293):1342–1346.PubMedView Article
Guo YL, Bechsgaard JS, Slotte T, Neuffer B, Lascoux M, Weigel D, Schierup MH: Recent speciation ofCapsella rubellafromCapsella grandiflora, associated with loss of self-incompatibility and an extreme bottleneck.Proc Natl Acad Sci USA 2009,106(13):5246–5251.PubMedView Article
Kusaba M, Dwyer K, Hendershot J, Vrebalov J, Nasrallah JB, Nasrallah ME: Self-incompatibility in the genus Arabidopsis: characterization of the S locus in the outcrossing A. lyrata and its autogamous relative A. thaliana.Plant Cell 2001,13(3):627–643.PubMed
Charlesworth D, Bartolome C, Schierup MH, Mable BK: Haplotype structure of the stigmatic self-incompatibility gene in natural populations ofArabidopsis lyrata.Mol Biol Evol 2003,20(11):1741–1753.PubMedView Article
Charlesworth D, Mable BK, Schierup MH, Bartolome C, Awadalla P: Diversity and linkage of genes in the self-incompatibility gene family inArabidopsis lyrata.Genetics 2003,164(4):1519–1535.PubMed
Paetsch M, Mayland-Quellhorst S, Neuffer B: Evolution of the self-incompatibility system in the Brassicaceae: identification of S-locus receptor kinase (SRK) in self-incompatibleCapsella grandiflora.Heredity 2006,97(4):283–290.PubMedView Article
Zhang XC, Wu X, Findley S, Wan J, Libault M, Nguyen HT, Cannon SB, Stacey G: Molecular evolution of lysin motif-type receptor-like kinases in plants.Plant Physiol 2007,144(2):623–636.PubMedView Article
Orengo CA, Thornton JM: Protein families and their evolution-a structural perspective.Annu Rev Biochem 2005, 74:867–900.PubMedView Article
Todd AE, Orengo CA, Thornton JM: Evolution of function in protein superfamilies, from a structural perspective.J Mol Biol 2001,307(4):1113–1143.PubMedView Article
Todd AE, Orengo CA, Thornton JM: Evolution of protein function, from a structural perspective.Curr Opin Chem Biol 1999,3(5):548–556.PubMedView Article
Tantikanjana T, Nasrallah ME, Stein JC, Chen CH, Nasrallah JB: An alternative transcript of the S locus glycoprotein gene in a class II pollen-recessive self-incompatibility haplotype ofBrassica oleraceaencodes a membrane-anchored protein.Plant Cell 1993,5(6):657–666.PubMed
Kramer EM, Jaramillo MA, Di Stilio VS: Patterns of gene duplication and functional evolution during the diversification of the AGAMOUS subfamily of MADS box genes in angiosperms.Genetics 2004,166(2):1011–1023.PubMedView Article
Boskovic RI, Sargent DJ, Tobutt KR: Genetic evidence that two independent S-loci control RNase-based self-incompatibility in diploid strawberry.J Exp Bot 2010,61(3):755–763.PubMedView Article
Moore RC, Purugganan MD: The evolutionary dynamics of plant duplicate genes.Curr Opin Plant Biol 2005,8(2):122–128.PubMedView Article
Prigoda NL, Nassuth A, Mable BK: Phenotypic and genotypic expression of self-incompatibility haplotypes inArabidopsis lyratasuggests unique origin of alleles in different dominance classes.Mol Biol Evol 2005,22(7):1609–1620.PubMedView Article
Samuel MA, Mudgil Y, Salt JN, Delmas F, Ramachandran S, Chilelli A, Goring DR: Interactions between the S-domain receptor kinases and AtPUB-ARM E3 ubiquitin ligases suggest a conserved signaling pathway in Arabidopsis.Plant Physiol 2008,147(4):2084–2095.PubMedView Article
Chen X, Shang J, Chen D, Lei C, Zou Y, Zhai W, Liu G, Xu J, Ling Z, Cao G: A B-lectin receptor kinase gene conferring rice blast resistance.Plant J 2006,46(5):794–804.PubMedView Article
Pastuglia M, Roby D, Dumas C, Cock JM: Rapid induction by wounding and bacterial infection of an S gene family receptor-like kinase gene inBrassica oleracea.Plant Cell 1997,9(1):49–60.PubMed
Pastuglia M, Swarup R, Rocher A, Saindrenan P, Roby D, Dumas C, Cock JM: Comparison of the expression patterns of two small gene families of S gene family receptor kinase genes during the defence response inBrassica oleraceaandArabidopsis thaliana.Gene 2002,282(1–2):215–225.PubMedView Article
Sanabria N, Goring D, Nurnberger T, Dubery I: Self/nonself perception and recognition mechanisms in plants: a comparison of self-incompatibility and innate immunity.New Phytol 2008,178(3):503–513.PubMedView Article
Fukai E, Fujimoto R, Nishio T: Genomic organization of the S core region and the S flanking regions of a class-II S haplotype inBrassica rapa.Mol Genet Genomics 2003,269(3):361–369.PubMedView Article
Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez R: Clustal W and Clustal X version 2.0.Bioinformatics 2007,23(21):2947–2948.PubMedView Article
Abascal F, Zardoya R, Posada D: ProtTest: selection of best-fit models of protein evolution.Bioinformatics 2005,21(9):2104–2105.PubMedView Article
Guindon S, Dufayard JF, Lefort V, Anisimova M, Hordijk W, Gascuel O: New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0.Syst Biol 2010,59(3):307–321.PubMedView Article
Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S: MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods.Mol Biol Evol 2011,28(10):2731–2739.PubMedView Article
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.