Abstract
Self-incompatibility (SI) in plants has evolved independently multiple times and S-RNase-based gametophytic self-incompatibility (GSI) is most common. The Rosaceae family possesses both self-recognition (Prunus) and nonself-recognition (Malus) GSI systems, and the latter is widespread in flowering plants. Gillenia trifoliata is a Rosaceae species related to Prunus and Malus, providing utility for understanding SI evolution. Gillenia is sister taxon to Malus, but unlike Malus, has not undergone polyploidisation. In addition, the common ancestor of Gillenia and Prunus is close to the origin of the subfamily. Using a highly contiguous Gillenia genome, orthologous regions to both Malus and Prunus S-loci were identified. Only the Prunus-like S-locus was highly polymorphic and had signatures of a functional S-locus including positive selection of the S-RNase. This suggests a self-recognition system controls SI in Gillenia, and the common ancestors of Gillenia and Prunus, and Gillenia and the apple tribe, likely had a self-recognition SI system. Comparative genomics between Gillenia and Malus suggest apple lost the self-recognition mechanism, and a nonself-recognition mechanism evolved independently from a rudimentary locus with at least one male S-determinant. Repetitive sequences in the Malus-like S-locus in Gillenia may facilitate illegitimate recombination, suggesting putative mechanisms of evolution of nonself-recognition S-loci.
Similar content being viewed by others
Introduction
Self-incompatibility (SI) is a pre-fertilisation genetic mechanism to promote outcrossing. SI systems have evolved independently at least 35 times in flowering plants1 and the gametophytic SI (GSI) system (in which SI is determined by the genotype of the haploid pollen gamete) has the broadest taxonomic range2,3,4,5. Among GSI systems, the S-RNase-based GSI system is the putative ancestral state for core eudicots6. In this system, a single highly polymorphic S-locus contains a pistil (female) S-determinant (S-RNase) and pollen (male) S-determinant(s) (S-F-box) genetically linked within an S-haplotype3. When active, the pistil S-RNase degrades cellular RNA in the pollen tube thus inhibiting growth7. The conditions under which the S-RNase is active, depends on whether the system utilises a self- or nonself-recognition mechanism (‘self’ referring to the S-haplotype). For S-RNase-based GSI self-recognition systems, an inactive pistil S-RNase becomes cytotoxic upon recognition by a single self-pollen S-F-Box, whereas, in the case of nonself-recognition, an active pistil S-RNase is detoxified upon recognition by multiple possible nonself-pollen S-locus F-Box (SLF) (or S-locus F-Box Brothers (SFBB) in Rosaceae)8,9. The different genetic architectures of the two mechanisms have major evolutionary consequences: self-recognition systems co-evolve within an S-haplotype and rare S-haplotypes benefit from negative-frequency dependence (i.e., absence of recognition of rare S-alleles increases fitness)10; nonself-recognition systems co-evolve within a population, and the effective fitness of negative-frequency dependence also depends on the ‘completeness’ of the multiple nonself SLFs to recognise a wide-range of population S-RNases8,11,12; mechanisms which contribute to this ‘completeness’ are therefore integral to understanding SI function and evolution13.
Phylogenetic analyses of S-RNase GSI S-genes show that S-RNase (T2 RNase) and SLF genes from species spanning two major clades (superrosids and superasterids) cluster within monophyletic clades suggesting a single origin of each S-gene component4,6,14. Given nonself-recognition GSI systems are well-characterised in Petunia (Solanaceae), Antirrhinum (Plantaginaceae), Citrus (Rutaceae), and Malus, Pyrus and Rosa (Rosaceae)9,15,16,17,18,19,20,21,22, it is often concluded that nonself-recognition with multiple SLF factors was the likely ancestral state of the S-RNase GSI system in eudicots4,8,20. Within Rosaceae, Prunus bears a self-recognition S-RNase GSI system17,18, which is designated on the basis of a distinct Prunus-specific S-locus F-box (SFB) clade within SLF phylogenies23 and loss-of-function mutations in SFB causing self-compatibility24,25, among other lines of evidence. It is postulated the Prunus self-recognition system evolved independently, and the Rosaceae ancestor had a nonself-recognition system4,8, like that of Rosa and Malus. The Prunus S-locus is orthologous to the S-loci of Rosa and Fragaria, and it was recently shown that Rosa putatively bears a multiple SLF nonself-recognition GSI system while the Fragaria recognition mechanism is unknown20,23,26. The Malus multiple SFBB nonself-recognition S-locus is paralogous to Prunus, and putatively arose through convergent de novo evolution23,27. Lack of highly contiguous genome assemblies has hampered studies of S-locus evolution.
The Gillenia genus occupies a phylogenetically informative position within Rosaceae to aid exploration of SI evolution. Gillenia is sister taxon to the Maleae tribe comprising approximately 1000 species, among which apple, Japanese pear, and Sorbus bear nonself-recognition GSI systems at the same genomic locus16,21,28. Studies suggest the Maleae tribe (base chromosome, 17, and rarely 15) evolved from an ancestor similar to Gillenia (base chromosome, 9), through autopolyploidisation, aneuploidy, and diploidisation29,30, and recently, high synteny was shown between G. trifoliata and apple (Malus domestica) genomes where 12 (of 17) homoeologous apple chromosomes shared end-to-end co-linearity with six Gillenia chromosomes31. Gillenia therefore provides utility to deconvolute the ancient polyploidisation of the Maleae tribe. Also, the last common ancestor of Prunus and Gillenia is thought to be close to the origin of the second largest Rosaceae subfamily (Amygdaloideae)32,33, providing insight into putative ancestral states. This study aimed to identify and characterise the S-locus in Gillenia to understand evolution of SI systems, through comparison of S-loci between the highly contiguous genomes of G. trifoliata and apple root-stock cultivar Malus domestica ‘M9’ and informative Rosaceae species.
Results
Pollination assays demonstrate self- and cross-incompatibility in Gillenia
In a previous study, facilitated self-pollination of Gillenia flowers failed to set fruit or seed over successive seasons suggesting Gillenia was self-incompatible31. Pollen tube growth assays were undertaken to explore self-incompatibility phenotype using hand pollination with self-pollen. It was found that all self-pollination assays resulted in no pollen tube growth confirming previous observations and presence of a self-incompatibility phenotype (Fig. 1a,e,i).
In the previous study, differential cross-compatibility phenotypes were observed between the three Gillenia individuals31. To determine reciprocal cross-incompatibility phenotype, pollen tube growth assays were undertaken after hand pollination with cross-pollen from different individuals. Complete pollen tube growth was observed in reciprocal crosses between individual ‘G2’ and individuals ‘G1’ or ‘G3’ (Fig. 1b,d,f,h), demonstrating cross-compatibility. However, limited or no pollen tube growth was observed in reciprocal crosses between individuals ‘G1’ and ‘G3’ (Fig. 1c,g) confirming previous observations and suggesting cross-incompatibility.
Self- and cross-(in)compatibility tested with pollen tube growth assays. Compound fluorescence images showing development of pollen tubes at 24 h after self- or cross-pollination of or between Gillenia individuals (‘G1’, ‘G2’, and ‘G3’). Pollen tube growth following self- (a,e,i) or reciprocal cross- pollination (b-d,f-h). Compatible reciprocal crosses were observed in a previous study31 between ‘G2’ and ‘G1’ (b,d) and ‘G2’ and ‘G3’ (f,h), while incompatible crosses were observed in a previous study in reciprocal crosses between ‘G1’ and ‘G3’ (c,g). Germinated pollen and pollen tubes are observed as green fluorescence. Scale bar, 500 μm. Analysis also performed at 48 h and showed reduced signal in all crosses (data not shown). Table shows summary of analysis from three replicates each with five styles; +, compatible; - incompatible.
A new highly contiguous genome for Gillenia reveals two putative S-loci
Due to their highly polymorphic nature, characterisation of S-loci requires highly contiguous genome assemblies. Gillenia version 1 (Gtr.v1) genome assembly had high synteny with selected Rosaceae genomes31 and comparison of Gtr.v1 with apple or peach (Prunus persica) genomes revealed syntenic regions in Gtr.v1 with their respective characterised S-loci. The apple S-locus, located on the distal end of apple linkage group 1734, was syntenic with the distal end of Gtr.v1 Chr03, while the Prunus S-locus, located on the distal end of peach linkage group 635, was syntenic with the proximal end of Gtr.v1 Chr05 (Supplemental Fig. S1). Nucleotide sequence alignment between published Maleae S-RNase and S-locus F-Box Brother genes (see Methods) identified annotated gene models on Gtr.v1 Chr03 within the region syntenic with the apple S-locus, while alignment with published Prunus S-RNase genes identified regions only on unanchored contigs. This suggested Gtr.v1 lacked necessary contiguity for S-locus characterisation. To resolve this, a new Gillenia genome (Gtr.v2) was assembled using PacBio® Hi-Fi long-read sequencing (see Methods). High co-linearity was observed between Gtr.v1 and v2 genomes (Supplemental Fig. S2a,b) and BUSCO scores were comparable between versions. Gtr.v2 was haplotype-resolved with less unanchored scaffolds and almost no gaps (Table 1), illustrating higher contiguity.
Nucleotide sequence alignment between Gtr.v2 and published Maleae and Prunus S-RNases, as described above, identified homology (Supplemental Fig. S2c), suggesting Gillenia possessed two putative S-loci: one on Chr03 orthologous to the Maleae S-locus (‘Mal S-locus’), and a second on Chr05 orthologous to the Prunus S-locus (‘Pru S-locus’). Automated and manual annotation identified putative S-genes within the S-loci on each haplotype. For simplicity haplotype 1 (gene model ID prefix ‘Hap1’) was used as reference for further analysis.
Both Maleae and Prunus-like S-loci contain putative S-RNase genes
S-RNase-based GSI requires the activity of S-RNases belonging to Class III RNase T2 subfamily within the ribonuclease superfamily4. Within Class III, S-RNase with SI function can be distinguished from S-lineage RNases with no SI function in having a maximum of two introns36, presence of amino acid patterns 1 and 2 and absence of pattern 414, and a basic isoelectric point (8 ≤ pI ≤ 10)23. Nucleotide sequence alignment between Gtr.v2 and published Maleae and Prunus S-RNases and further Conserved Domain Database search of putative loci (see Methods) identified seven candidate S-RNase genes: three on Chr05 in the ‘Pru S-locus’ and four on Chr03 in the ‘Mal S-locus’, however, one ‘Mal S-locus’ S-RNase (Hap1g27798) was putatively non-coding. The six protein-coding candidates were assessed for molecular features characteristic of S-RNases with SI function (Fig. 2). Only ‘Mal S-locus’ Hap1g12044 and Hap1g12045 and ‘Pru S-locus’ Hap1g16438 encoded full-length proteins, had a maximum of two introns, encoded amino acid patterns 1 and 2, and were absent of S-lineage-specific amino acid pattern 4, and Hap1g12044 and Hap1g16438 had a predicted basic isoelectric point (Fig. 2a, Supplemental Fig. S3). S-RNase proteins have a characteristic secondary structure comprised of five conserved domains and hypervariable regions25,37. Analysis of secondary structure of the three full-length S-RNases showed, in addition to patterns 1 and 2 previously described, all three possessed the five conserved domains and the Rosaceae hypervariable domain (Supplemental Fig. S3). It was observed that the active site Lys residue in the C3 domain (K118 in Pruavi-S1, BAA83479), which is conserved in all S-RNases, was predicted to be mutated (K→T) in proteins of ‘Mal S-locus’ Hap1g12044 and ‘Pru S-locus’ Hap1g16438, while the haplotype-2 allele of ‘Pru S-locus’ gene, Hap2g17313, did not have this mutation (Supplemental Fig. S3); a K→T mutation in the active site of C3 domain may alter S-RNase activity.
Molecular features of candidate Gillenia S-RNases. (a) Analysis of molecular signature of S-RNase genes in Gillenia candidates with two references (Pyrus pyrifolia Pypyr-S1 BAA32412.1 and Prunus avium Pruavi-S1 BAA83479.1). Protein sequence length (res., amino acid residues) shows truncation of some predicted proteins. Functional S-RNase genes have a maximum of two introns36, presence of amino acid patterns 1 and 2 and absence of amino acid pattern 4 (six residues ‘[CG]P[QLRSTIK][DGIKNPSTVY][ADEIMNPSTV][DGKNQST]’ occurring between pattern-2 and C3)14, and a basic isoelectric point (8 ≤ pI ≤ 10)23. pI calculated in Geneious Prime. (b) Phylogenetic analysis of S-RNase proteins from Gillenia, Prunus, and Maleae species. Clade nomenclature23 shows clades of ‘Prunus S-RNases’ (orange), ‘Maleae S-RNases’ (blue), a cluster containing ‘S-lineage 1’ and ‘Malus S-lineage 2’ clades with no known SI function (green) and outgroups of ‘RNase 2/3/LE-like’ genes (grey). Maximum likelihood bootstrap values from 100 datasets shown at branches. Scale is substitutions per site. (c) Expression of reference haplotype-1 Gillenia S-RNase genes in style/stigma, stamen (anther), and germinated pollen tissues, shown as transcripts per million (TPM), coloured banner of model ID indicates lineage corresponding to the phylogenetic tree in (b). Letters represent statistical significance groups from Tukey’s HSD test, all unlabelled data belong to the same group.
Phylogenetic analysis with published Maleae and Prunus S-RNase and S-lineage proteins showed that ‘Pru S-locus’ Hap1g16438 clustered in the ‘Prunus S-RNase’ clade, while the other two ‘Pru S-locus’ proteins clustered in the S-lineage clade which is not associated with SI function, and the three ‘Mal S-locus’ proteins clustered within the ‘Maleae S-RNases’ clade (Fig. 2b, Supplemental Fig. S4). S-RNase genes are characterised by high expression in the style and stigma23,25. Expression using mRNA-seq on style/stigma, stamen, and germinated pollen tissues was assessed for all seven putative S-RNase genes (Fig. 2c, Supplemental Fig. S5). The full-length S-RNases ‘Mal S-locus’ Hap1g12044 and Hap1g12045, and ‘Pru S-locus’ Hap1g16438, were expressed in the style/stigma and not male reproductive tissues, but Hap1g16438 expression was 30-fold higher in the style/stigma compared to the ‘Mal S-locus’ RNases. Of the two ‘Mal S-locus’ S-RNase genes, Hap1g12045 had higher expression compared to Hap1g12044. Taken together, only the ‘Pru S-locus’ Hap1g16438 had all molecular features consistent with a functional S-RNase, whereas the two full-length ‘Mal S-locus’ candidates had either insufficiently basic isoelectric point (Hap1g12045) or very low expression in style/stigma tissues (Hap1g12044).
The ‘Pru S-locus’ S-RNase is highly polymorphic, has evolved under positive diversifying selection, and its alleles correlate with incompatibility phenotypes
S-genes are highly polymorphic through a combination of negative frequency-dependence, providing long-term maintenance of diverse alleles, and positive diversifying selection, facilitating molecular evolution of different recognition specificities10,37. To assess the degree of polymorphism amongst Gillenia S-RNases putative alleles were deduced from PacBio® long-read IsoSeq and Illumina short-read transcriptomic datasets from three Gillenia individuals. For analysis of allelic diversity, ‘Pru S-locus’ Hap1g16438 and ‘Mal S-locus’ Hap1g12045 were selected based on RNA-seq expression levels (Fig. 2c). Amongst Gillenia individuals, four alleles were identified for ‘Pru S-locus’ Hap1g16438 with divergent nucleotide (64.3–71.3%) and predicted protein (71.0–82.4%) identity. Four alleles were also identified for ‘Mal S-locus’ Hap1g12045 and had greater conservation of nucleotide (99.3–99.9%) and predicted protein (98.3–99.6%) identity (Supplemental Fig. S6a-d). Additionally, single nucleotide polymorphism (SNP) density was calculated in a genome-wide context in the style/stigma transcriptomes of the three Gillenia individuals (Supplemental Fig. S6e) and showed increased SNP density at the ‘Pru S-locus’ compared to the ‘Mal S-locus’. Together, the data suggest only the ‘Pru S-locus’ Hap1g16438 had sufficient polymorphic characteristics to control SI recognition specificity.
Cross-compatibility phenotypes were assessed for correlation with putative S-RNase alleles. Alleles of ‘Pru S-locus’ Hap1g16438 were identified within each individual: G1 (alleles pS1 (Hap1g16438), pS2 (Hap2g17313)), G2 (alleles pS3, pS4), and G3 (allele pS2) (Supplemental Fig. S6f). Compatible reciprocal crosses were observed in the pollen tube growth assays between G2xG1 and G2xG3 (Fig. 1), suggesting the ‘Pru S-locus’ S-alleles correlated with these cross-compatibility phenotypes. Only one haplotype could be identified in G3 (pS2), and two alleles were identified in G1 (pS1, pS2), suggesting G1♂xG3♀ should yield a cross-compatibility phenotype while G3♂xG1♀ should have a cross-incompatibility. Pollen tube growth was not observed in either cross, suggesting additional male factors may be contributing to the cross-incompatibility phenotype of G1♂xG3♀. In contrast to the ‘Pru S-locus’ S-alleles, the ‘Mal S-locus’ alleles, in addition to the high sequence conservation (> 98.3% protein identity), did not correlate with incompatibility phenotypes (G1 (mS1, mS2), G2 (mS1, mS3), G3 (mS1, mS4)).
S-RNase genes display signatures of positive diversifying selection as nonsynonymous mutations change protein sequence and enable evolution of diverse recognition specificities10,37. Candidate S-RNase genes in Gillenia were therefore analysed for selective pressure using a branch-based phylogenetic approach (using CODEML) with S-RNases from eight species each from Maleae and Prunus (see Methods). Branch models assuming heterogeneous nonsynonymous/synonymous rate ratio (ω = dN/dS) for different branches of Maleae and Prunus-like S-RNase phylogenies were analysed (Table 2). In the Prunus S-RNase lineage, the alternate hypothesis (H1), which assumes the Gillenia branch has a different ω ratio to the Prunus branch, is rejected (2Δℓ< χ21,5%) in favour of the null hypothesis (H0) of homogeneous ω ratio on all branches, suggesting the ‘Pru S-locus’ S-RNase in Gillenia is under similar selective pressure as known functional Prunus S-RNases. In comparison, in the Maleae S-RNase lineage, the null hypothesis of homogeneous ω ratio is rejected (2Δℓ> χ21,5%) in favour of the alternate hypothesis that assumes the Gillenia branch has a different ω ratio to that in the Maleae branch, and correspondingly, a near 2-fold increase towards positive selection was evident in the Maleae branch comprised of known functional S-RNases, compared with the Gillenia branch. Branch-site models assuming heterogeneous ω across branches and amino acid residues (sites) were also analysed (Table 2). Compared to the Prunus-lineage S-RNases, the Maleae-lineage S-RNases observed a greater increase in proportions of 2a and 2b site classes (in which foreground branches (Maleae) are under positive selection while background branches (Gillenia) are under purifying selection or neutral evolution, respectively) between null and alternate hypotheses, and this change derived predominantly from a decrease in sites under neutral evolution (site class 1), suggesting the Gillenia ‘Mal S-locus’ S-RNase was under greater neutral selection compared to the Gillenia ‘Pru S-locus’ S-RNase, and supports branch model findings that the Gillenia ‘Pru S-locus’ S-RNase is under similar pressures to functional Prunus S-RNases.
Both Maleae and Prunus-like S-loci contain putative S-locus F-box genes
Nucleotide sequence alignment between Gtr.v2 and published Maleae SFBB and Prunus SFB genes followed by Conserved Domain Database search (see Methods) identified eight S-locus F-box genes on Chr03 ‘Mal S-locus’ and 16 on the Chr05 ‘Pru S-locus’. Phylogenetic analysis resolved the lineage of the predicted proteins (Fig. 3a; Supplemental Fig. S7). Of the eight F-boxes on the Chr03 ‘Mal S-locus’, three clustered with SFBB proteins, while four clustered with or near the SFB-like clade. Of the 16 F-boxes on Chr05 ‘Pru S-locus’, one clustered with SFB proteins, nine clustered with S-Locus F-box Like (SLFL) proteins, and the remaining six clustered in outgroups. S-locus F-box genes are typically highly expressed in pollen and stamen tissues16,18. mRNA-seq analysis of style/stigma, stamen, and germinated pollen tissues revealed that while all three SFBB genes and five SLFL genes have high expression in stamen and pollen tissues, low expression also occurred in style/stigma tissues. In contrast, the single SFB gene (Hap1g16439) was highly expressed in germinated pollen and stamen tissues and no expression was detected in the style/stigma (Fig. 3b; Supplemental Fig. S8). Analysis of protein structure of the pollen/stamen expressed Gillenia S-F-box candidates showed that each had the conserved F-box and variable domains specific to each lineage (Supplemental Fig. S9). Owing to a putatively recent segmental duplication at the ‘Mal S-locus’, two SFBB genes (Hap1g27799 and Hap1g27800) were identical and the third SFBB gene (Hap1g12043) shared 93.7% nucleotide identity with Hap1g27799 and Hap1g27800. Using mRNA-seq data, sequence diversity was found to be extremely low among the Maleae-lineage Gillenia SFBB genes (93.5–100% nucleotide identity), compared with Prunus-like SFB gene Hap1g16439 (82.4–99.9% nucleotide identity) (Supplemental Fig. S10), supporting ‘Pru S-locus’ as a more highly polymorphic locus.
Gillenia S-locus F-boxes. (a) Phylogenetic analysis of S-locus F-box proteins from Gillenia, Prunus, and Maleae species. Clade nomenclature23 shows lineages of ‘S-locus F-Box Brothers’ (SFBB, blue), ‘S-Locus F-box Like’ (SLFL, green), ‘S-locus F-Box’ (SFB, orange), and ‘SFB-like’ (grey). Scale is substitutions per site. (b) Expression of reference haplotype-1 Gillenia S-locus F-box genes in style/stigma, stamen (anthers), and germinated pollen tissues, shown as transcripts per million (TPM), coloured banner of model ID indicates lineage corresponding to the phylogenetic tree in (a). Letters represent statistical significance groups from Tukey’s HSD test, unlabelled data belong to the ‘f’ group.
Evolution of ‘Mal S-locus’ and ‘Pru S-locus’ in Rosaceae genomes
Genomic architecture of the two putative S-loci in Gtr.v2 was characterised (Fig. 4a). The Chr05 ‘Pru S-locus’ in Gillenia contains the core components of SI regulation (S-RNase and SFB) and multiple SLFL genes, like the self-recognition S-loci observed in Prunus species38. In contrast, the Chr03 ‘Mal S-locus’ has more than one S-RNase, and few, near-identical SFBB-like genes, which is unlike the architecture of nonself-recognition S-loci observed in multiple species and comprised of a single highly polymorphic S-RNase gene with multiple SFBB genes16,21. As the Chr05 S-locus has the conserved architecture of functional S-loci, this suggests a Prunus-like self-recognition mechanism is regulating SI in Gillenia.
Comparative genomics was performed to understand putative origins and evolution of the ‘Mal S-locus’ and ‘Pru S-locus’ within apple. To assist this, a highly contiguous genome was constructed for the semi-vigorous apple (M. domestica Borkh) rootstock ‘M9’ using PacBio® HiFi long-read sequencing. This assembly achieved haplotype-level resolution, yielding 17 pseudomolecules per haplotype. Through scaffolding against the apple ‘GDDH13’ genome39, the pseudomolecules were built by 37 and 36 scaffolds for haplotypes 1 and 2, respectively. The order and orientation of scaffolds within each pseudomolecule per haplotype are shown in the AGP (A Golden Path) files (Supplemental Table S1). The resulting assembly showed an improved N50 of 38.4 Mb and 33.4 Mb for the two haplotypes (Supplemental Table S2). Moreover, 21 (constituting 95.04% of haplotype 1) and 29 (constituting 94.25% of haplotype 2) scaffolds were successfully anchored to the ‘M9’ genetic map (Supplemental Fig. S11), demonstrating synteny and anchoring. BUSCO scores of the ‘M9’ genome demonstrated high assembly quality (Supplemental Fig. S12).
Syntenic analysis performed between Chr05 ‘Pru S-locus’ and homoeologous Malus ‘M9’ Chr04/12 was conducted to explore putative S-locus evolution (Fig. 4b). The region containing the core SI components (S-RNase and SFB) was absent in both apple chromosomes and M9.Chr12 harboured additional non-syntenic sequence with M9.Chr04 suggesting an insertion. The putative M9.Chr12 insertion had PIF/Harbinger transposons at its borders, and within the insertion, hAT transposons bordered a repeating unit of ribosomal/photosynthesis-related and glycosyl hydrolase domains, suggesting a range of insertion mechanisms. One SLFL-like gene was identified on each Malus chromosome, sharing synteny with two different Gillenia SLFL genes, consistent with shared ancestry.
Syntenic analysis was conducted between Chr03 ‘Mal S-locus’ and homoeologous Malus ‘M9’ Chr09/17 to understand evolution of this putative S-locus (Fig. 4c). Consistent with evolutionary dynamics of S-loci, few syntenic links were identified between the ‘Mal S-locus’ and the apple orthologous chromosomes. However, microsynteny was detected between SFBB-like Gillenia Hap1g12043 and Malus SFBB3a allele on M9.Chr17 (Fig. 4c; dark blue connector line), and between a transposable element in Gillenia and apple chromosomes (described below), suggesting independent evolution of the ‘Mal S-locus’ in the two species, but also putative shared ancestry of one SFBB gene.
To explore putative origins of the Chr03 ‘Mal S-locus’, an analysis of synteny between this region and identified syntenic regions from four genomes from species representing all three Rosaceae subfamilies (Rosoideae (Rosa), Amygdaloideae (Prunus), and Dryadoideae (Dryas)) was conducted. Despite syntenic regions either side of the Gillenia Chr03 'Mal S-locus’, no synteny was detected between the ‘Mal S-locus’ and syntenic regions from the four species (Fig. 4d), suggesting the Chr03 ‘Mal S-locus’ may have arisen independently in the Malodae ancestor.
Evolution of the Gillenia ‘Mal and Pru S-loci’ in Rosaceae. (a) Schematic of the ‘Pru S-locus’ and ‘Mal S-locus’ in Gillenia. The ‘Pru S-locus’ on Gtr.Chr05 bears a single S-RNase (hap1g16438, blue triangle), single SFB (hap1g16439, turquoise triangle), and multiple SLFL genes (hap1g27801-g27808, green triangles), as well as genes with other functions (grey triangles). The ‘Mal S-locus’ on Gtr.Chr03 bears multiple S-RNase genes both full-length (hap1g12044/hap1g12045, blue triangles) and partial S-RNase genes (hap1g27798/hap1g12046, yellow triangles) and three SFBB genes (hap1g12043/hap1g27799/hap1g27800, green triangles). (b,c) Syntenic analysis between Gillenia S-loci with orthologous syntenic regions on Malus domestica ‘M9’ homoeologous chromosomes. Gillenia S-loci from (a) superimposed to scale onto Gtr chromosome segments. Yellow region denotes putative S-loci (i.e., core components (S-RNase and SFB) of the ‘Pru S-locus’, the ‘Mal S-locus’ on Gtr.Chr03, or the Malus Chr17 S-locus on ‘M9’ including S-RNase and SFBB genes and excluding a single outgroup F-box gene). ‘M9’ S-gene colour coding follows (a). M9.Chr12 insertion denoted by a dashed border, annotations below M9.Chr12 are (top-bottom), PIF/Harbinger transposons (dark red triangles), hAT transposons (light red triangles), and high-frequency domains identified within the insertion forming a repeating unit of photosynthesis (green) or ribosomal (brown) related domains with glycosyl hydrolase (dark blue) domains. Connecting lines denote synteny of minimum length 300 bp with greater than 78% identity, dark blue connecting line denotes microsynteny detected between SFBB genes Gtr.Chr03 hap1g12043 and M9.Chr17. SFBB3a allele (MinLength = 1126 bp; 84.78%id). (d) Syntenic analysis between Gtr.Chr03 ‘Mal S-locus’ with orthologous syntenic regions from four species representing three Rosaceae subfamilies, Prunus (Amygdaloideae), Rosa (Rosoideae), Dryas (Dryadoideae). Gillenia S-loci from (a) superimposed to scale on Gtr.Chr03 segment and highlighted yellow. Connecting lines denote synteny of minimum length 120 bp with greater than 78% identity. Pavi, Prunus avium Tieton Genome v2.040; Pdul, Prunus dulcis Nonpareil Genome v1.041; RcHm, Rosa chinensis Homozygous Genome v2.042; Dryas octopetala genome43. Prunus and Rosa genomes sourced from Genome Database for Rosaceae44, Dryas genome sourced from NCBI (www.ncbi.nlm.nih.gov).
Features of Gillenia ‘Mal S-locus’ point to putative gene conversion mechanisms
One hypothesis for the origin of the Maleae S-locus is through de novo evolution23 putatively arising through gene conversion13. Gene conversion can result from illegitimate recombination between non-homologous sequences or between duplicate sequences with high sequence similarity at ectopic and proximal locations, causing non-reciprocal (unidirectional) transfer of genetic information45. The Chr03 ‘Mal S-locus’ in Gillenia was assessed for features which may facilitate gene conversion (Fig. 5). Three segmental duplications, varying in size from 18.283 to 21.617 kb, were identified within the Chr03 ‘Mal S-locus’ which were bordered by highly similar sequences identified by the EDTA annotation pipeline46 as belonging to the same annotation family (Fig. 5a; Supplemental Fig. S13). Each segment included an SFBB gene positioned in tandem with an RNase_T2 domain-containing gene of either a full-length S-RNase (hap1g12045, block-3), non-coding partial S-RNase (hap1g27798, block-1), or remnant RNase_T2 domain only (block 2). Proximal positioning of highly similar sequence may permit illegitimate recombination and unidirectional transfer of genetic information.
Transposable elements (TEs) can accumulate in regions of low recombination such as S-loci47,48 and their presence does not imply causative effect in genome evolution; however, TE movements may propagate highly similar sequence to enable rare illegitimate recombination49,50. An intact Ty1/Copia LTR (identity 98.5%) positioned adjacent to Gillenia SFBB hap1g12043 and encompassing non-coding partial S-RNase hap1g27798 was found to be syntenic with three locations across Malus M9.Chr09 and M9.Chr17 (Fig. 5b, maroon lines) dispersed proximal to Malus S-locus SFBB genes (Fig. 5b, green line), suggesting both conservation and active replication proximal to other SFBB genes respectively before and after whole genome duplication in Malus. An annotated Ty3/Gypsy LTR also had synteny with TEs determined to be from the same family in multiple locations on both M9.Chr09 and M9.Chr17 (Fig. 5b, light orange lines), but with conservation of numbers suggesting no further replication after whole genome duplication. The intact Ty1/Copia LTR was also found to be actively transcribed in Gillenia with transcripts spanning the non-coding partial RNase_T2 hap1g27798 gene (Supplemental Fig. S14). Presence of the intact Ty1/Copia LTR within the Malus S-locus may suggest a mechanism to propagate highly similar sequence throughout the locus.
Gillenia ‘Mal S-locus’ has repeated blocks and repetitive sequences. (a) ‘Mal S-locus’ structure showing regions of putative segmental duplications with dashed lines indicating borders at repeat regions belonging to the same annotation family determined by the EDTA pipeline (teal triangles), found at the left and right borders of segmental blocks 2 and 3, and right border of segmental block 1. Top, grey boxes indicate regions of segmental duplications. Gene annotations are SFBB (green), full-length S-RNase genes (blue, hap1g12044 and hap1g12045), partial S-RNase gene fragments (blue, hap1g27798 and hap1g12046), and gene with no known S-locus function (grey, hap1g27797). RNase_T2 domains (yellow). Repeat regions (teal from the same family, orange from the same TE family, grey from additional and different TE families). LTR annotations are Ty1/Copia and Ty3/Gypsy (green), LTR retrotransposons (dark red). TIR annotations are hAT (light red), PIF/Harbinger (purple), Mutator (cream), CACTA (light blue). LINE elements (blue). (b) Syntenic analysis between Gillenia ‘Mal S-locus’ with orthologous syntenic regions on Malus domestica ‘M9’ homoeologous chromosomes, annotations follow Fig. 4c. Connecting lines denote synteny of minimum length 290 bp with greater than 78% identity. Connecting lines highlight syntenic relationships of interest: green, Gillenia hap1g12043 and M9.Chr17 SFBB3a allele; maroon, Gillenia intact Ty1/Copia LTR with locations of the same TE family on both Malus chromosomes; light-orange, Gillenia annotated Ty3/Gypsy LTR with locations of the same family on both Malus chromosomes.
Discussion
In this study, we provide data which suggest a Prunus-like self-recognition mechanism governs SI in Gillenia. We have shown that Gillenia is a self-incompatible species and bears orthologous regions to Maleae nonself-recognition and Prunus self-recognition SI systems. However, only the ‘Pru S-locus’ had molecular features of a functional S-RNase including highest expression in the style/stigma tissues and evidence of positive diversifying selection, had highly polymorphic S-alleles, had S-RNase allele combinations which explain some SI phenotypes, and had conserved genome architecture relative to Prunus S-loci. The ‘Mal S-locus’ had more than one S-RNase, had fewer S-locus F-box genes necessary for a nonself-recognition mechanisms (typically greater than eight)16,21, S-genes lacked necessary degrees of polymorphism, evidence for positive selection of the S-RNase was weak, and S-RNase alleles did not correlate with cross-incompatibility phenotypes.
Some limitations of the data suggest the mechanism controlling SI regulation in Gillenia needs further analysis. A cross-incompatibility phenotype shown in the G1♂xG3♀ which may theoretically be cross-compatible, suggest other factors may be contributing to the incompatibility mechanism—this may suggest involvement of the Chr03 ‘Mal S-locus’. The function of the Chr03 ‘Mal S-locus’ remains to be elucidated, as genes present at the locus show sequence conservation and expression in reproductive organs. Two independent S-loci control self-incompatibility in Fragaria, but both loci are highly polymorphic51, with the Chr03 ‘Mal S-locus’ lacking this characteristic, a two-loci system does not seem feasible in Gillenia. Further work with a larger population and/or through segregation analyses would provide further clarity on the self-incompatibility mechanism in Gillenia.
Nonself-recognition systems are found in a wide range of species including species belonging to Solanaceae, Plantaginaceae, and Rosaceae families9 and are hypothesised to be the ancestral state4. A functional Prunus-like S-locus controlling self-incompatibility in Gillenia suggests the common ancestor of the Malodae supertribe (Gillenia + Maleae) also used a Prunus-like self-recognition mechanism. This may suggest evolution of widespread nonself-recognition systems is relatively labile. The data here suggest the ‘Pru S-locus’ may have been lost in the ancestor to the Maleae tribe but this requires further exploration of Maleae species beyond Malus. After putative loss, the Maleae nonself-recognition S-locus may have evolved independently from a rudimentary locus structure consisting of at least one SFBB, present in the Malodae ancestor, similar to the Gillenia Chr03 ‘Mal S-locus’. Evolution of the Maleae nonself-recognition system may have resulted from selective pressure to escape self-compatibility8. Self-compatibility may arise either, prior to polyploidisation, through loss of the self-recognition system1, or, after polyploidisation, through competitive interaction of diploid pollen3,52. Either hypothesis is possible, given the self-recognition system is absent on both apple Chr04/12 (supports deletion prior to polyploidisation), and apple Chr12 contains TEs bordering the insertion (supports TE movement induced by polyploidisation50. The Maleae S-locus putatively evolved de novo via gene conversion13. The data here show abundance of highly similar sequences throughout the Chr03 ‘Mal S-locus’. These sequences may facilitate illegitimate recombination and unidirectional transfer of genetic information, contributing to gene conversion.
Evidence of a Prunus-like S-locus controlling self-incompatibility in Gillenia suggests the common ancestor of the Amygdaloideae subfamily32,33 utilised a self-recognition system. In terms of the ancestral state of Rosaceae, only low support was found for an SFB gene on the Dryas (Dryadoideae subfamily) genome scaffold with high synteny to the ‘Pru S-locus’ (data not shown), but the low-quality assembly of the Dryas genome may contribute to this. However, a putative SFB gene was identified as an S-locus F-box gene in Fragaria23 (Rosoideae subfamily). Combined with the putatively dynamic nature of evolution of the Maleae nonself-recognition mechanism, it may be parsimonious to postulate that the Rosaceae ancestor bore an SFB-based self-recognition system, rather than a nonself-recognition system, as found in Maleae and Rosa20.
Materials and methods
Plant materials and growth conditions
Mature Gillenia trifoliata plants, purchased from Wake Robin Nursery, Balclutha, New Zealand (www.wakerobin.co.nz), were grown in glasshouse conditions as previously described31. Mature Malus domestica ‘M9’ (clone NZ9) plants, purchased from Waimea Nursery, Nelson, New Zealand (www.waimeanurseries.co.nz), were grown in research orchard conditions in Palmerston North, New Zealand.
Genetic compatibility and pollen tube growth assays
Genetic self- and cross-compatibility studies in Gillenia were undertaken by hand-pollinating ten flowers per genotype using separate brushes with fresh pollen collected from flowers of the same (self-pollen) or different (cross-pollen) plant and performed in insect-free glasshouse conditions as previously described31. In vivo pollen tube growth assay53 was performed with modifications. Briefly, pistils were collected at 24 h and 48 h after pollination, fixed in FAA (37% formaldehyde/glacial acetic acid/50% ethanol, 5:5:90) and stored at 4 °C until use. Pistils were incubated in 4 N NaOH for 2 h to soften tissues, then washed thoroughly with water. Softened pistils were washed three times with 0.1% K3PO4, then incubated in 0.1% aniline blue in 0.1% K3PO4 for 24 h at 20 °C in darkness. Stained pistils were washed twice using 0.1% K3PO4 and gently squashed between a glass slide and cover slip in a drop of 0.1% aniline blue solution. Pollen tube growth in styles was observed with a compound fluorescence microscope (Nikon Eclipse Ni-E, Japan) with UV light at a wavelength of 365 nm. Three replicates per cross/self-pollination test were observed, with each replicate showing five styles.
In vitro pollen tube cultivation54 was performed with modifications. Mature pollen (BBCH stage 6531) was cultivated in 600 µL growth media (550 µM Ca(NO3)2, 1.6 mM H3BO3, 1.6 mM MgSO4, 1 mM KNO3, 440 mM sucrose, 5 mM 2-(N-morpholino) ethanesulfonic acid (MES), pH 6.5), and incubated in dark at 120 rpm at 25 °C for 5 h. Pollen tube growth was confirmed under the microscope. Germinated pollen grains were collected with centrifugation at 13,000 rpm, and the pelleted pollen was immediately processed for RNA extraction.
Nucleic acid isolation, library construction and sequencing
For Gillenia, high molecular weight nuclear genomic DNA was extracted from isolated nuclei collected from young leaves as described previously31 and submitted to Leiden Genome Technology Center (Netherlands) for PacBio® HiFi library construction and sequencing.
For Malus domestica ‘M9’, young, fully expanded leaves (20 g) were collected and 2 µg of nuclear genomic DNA was extracted from isolated nuclei as previously described55, and used for PacBio® HiFi library construction and sequencing. A Single Molecule, Real-Time (SMRT) sequencing library (PacBio®) to obtain High Fidelity reads (HiFi) was generated by the service provider (Neogen® Genomics, USA) and sequenced on the PacBio® Sequel II platform to generate subreads. Consensus reads (HiFi reads) were produced using the SMRT Link software of the PacBio analysis tool (https://github.com/caseywdunn/docker-smrtlink).
For RNA extraction, floral tissues at BBCH stages31 were immediately collected into lysis buffer, processed with a micro-homogeniser, and extracted with Spectrum Plant Total RNA Kit (Sigma-Aldrich, NZ). For PacBio® long-read IsoSeq RNA-seq, a single composite library of pooled style/stigma and stamen tissues of open flowers (BBCH-65) from three individuals was prepared and sequenced on Sequel II by the Australian Genome Research Facility (AGRF, Australia). For short-read RNA-seq, RNA was extracted from three individuals, ten flowers each, from pre-opened (BBCH-60) style/stigma tissues, mature (BBCH-65) stamens, and germinated pollen grains prepared from mature (BBCH-65) flowers. mRNA-seq library construction and sequencing was performed by Novogene (Hong Kong) using the NovaSeq PE150 platform.
Genome assembly
For the Gillenia genome, PacBio® Hi-Fi long-read DNA sequencing data were de novo assembled using hifiasm/v0.18.5-r50056. The resulting haplotype 1 and 2 contigs were processed in parallel and scaffolded into two sets of chromosome-scale assemblies using Gillenia genome v131 as reference with homology-based assembler RagTag/v1.0.257. Assembly statistics were calculated with Assemblathon58 and MUMmer/v3.2359. Telomeres were checked using telomeric-identifier (https://github.com/tolkit/telomeric-identifier). BUSCO/v5.2.2 analysis was used to verify assembly completeness60,61.
For the Malus domestica ‘M9’ genome, the haplotype-resolved genome was assembled using hifiasm/v0.18.5-r50056 in conjunction with PacBio® HiFi reads. Statistical evaluation of the resulting paired haplotypes was executed using Assemblathon58. The scaffolding was done with RagTag software57 using the whole genome alignment to the apple ‘Golden Delicious’ double haploid GDDH13v1.1 genome39 performed by nucmer software59 to produce an ‘agp’ file, which was then used by agptools (https://github.com/WarrenLab/agptools) to produce the two final haplotype assembly fasta files. The syntenic alignments (minimum alignment length of 18,000) between each haplotype and the reference genome were visualized through mummerplot R dot plots (https://jmonlong.github.io/Hippocamplus/2017/09/19/mummerplots-with-ggplot2/).
Additionally, the dual sets of haplotype-resolved scaffolds were anchored by aligning the 17 linkage groups established with a genetic map constructed for ‘M9’. This map was based on 646 and 626 SNPs markers per haplotype, respectively, which were extracted from the Illumina Infinium® 20 K apple SNP array62 acquired from DNA of an ‘M9’ x ‘Robusta5’ crossing population63. The arrangement and alignment of scaffolds were performed in accordance with marker order from the linkage map using ALLMAPS/v0.7.759. The final 17 pseudomolecules were merged into a single FASTA file per haplotype. Metrics of each haplotype were collected using Assemblathon58 and BUSCO values were calculated based on the eudicots_odb10 (2,326 genes) and embryophyte_odb10 (1,614 genes) datasets.
Genome annotation
De novo repeats were deleted on Gtr.v2 using EDTA/v246,64. The Gtr.v2 haplotyped genomes were soft masked using RepeatMasker/v4.0.565,66. High quality PacBio long-read RNA IsoSeq data from pooled tissues (see ‘Nucleic acid isolation’ methods) was processed using pbmm2 (https://github.com/PacificBiosciences/pbmm2), with the haplotyped genomes indexed using ‘pbmm2 index’, followed by aligning IsoSeq reads with ‘pbmm2 align’. Illumina mRNA-Seq data were aligned to the soft-masked haplotyped genomes using STAR/v2.7.10a67. Alignments were fed into BRAKER/v2.1.568 as evidence to predict gene models. The gene functional annotation was performed on the Mercator4 web server69,70 (https://www.plabipd.de/mercator_main.html).
Manual annotation was undertaken using alignment with short- and long-read RNA-seq data, Conserved Domain Database search using pfam00445 for S-RNase genes and pfam00646 and pfam01344 for S-locus F-box genes, homology identification using nucleotide BLAST (e-value < 1e-15) with published S-genes (Maleae S-RNases: U19793.1, AB002139.1, EF689008.1; Prunus S-RNases: AB028153.1, AB252415.1, AJ298314.1; Maleae SFBBs: AB270793.1, AB270796.1, AB270798.1; Prunus SFB/SLFLs: AB111518.1, AB092624.1, AB092627.1, AB092626.1), and whole chromosome alignment with characterised S-loci from Pyrus BAC clones AB545981 and AB54598271.
Genome synteny analyses
Microsyntenic analysis used Synteny Viewer tool (Genome Database for Rosaceae44 to locate syntenic anchors followed by nucleotide alignment of regions containing syntenic anchors with progressiveMauve72 plugin in Geneious Prime/v2023.0.1 (Biomatters Ltd, Auckland, NZ) to explore genome annotation detail. Microsyntenic regions were analysed with dot plots generated using MUMmer/v3.2359 and visualised with NGenomeSyn73 with parameters described in text.
De Novo transcriptome assembly
To analyse polymorphisms of S-gene candidates in Gillenia, nine short-read mRNA sequencing libraries from three tissue types (style/stigma, stamen, and germinated pollen) from three individuals (see Nucleic acid isolation methods above) were used for de novo transcriptome assembly using trinityrnaseq/v2.14.074. Cleaned data were aligned to the haplotyped genomes and quantified using the nf-core/RNASeq pipeline75.
Phylogenetic analysis
Putative S-RNase and S-F-box genes were identified via nucleotide BLAST search (e-value < 1e-15) with published S-genes for Maleae and Prunus (accession numbers same as used for manual annotation, see ‘Genome annotation’ methods section) and by Conserved Domain Database search of a 4 MB region encompassing each putative S-loci using pfam00445 for S-RNase genes and pfam00646 and pfam01344 for S-locus F-box genes. Phylogenetic analyses were conducted in Geneious Prime/v2023.0.1 (Biomatters Ltd, Auckland, NZ). Protein sequences for all putative S-genes were aligned using Clustal Omega/v1.2.276. Regions of alignment were used to generate phylogenetic trees with PhyML/v3.3.2018062177 with JTT substitution model and bootstrap analysis of 100 datasets.
Positive selection analyses
Protein and nucleotide sequences used for branch-wise comparisons were: Prunus persica-S1 AB252415.1, P. dulcis-Sj AAZ57491.1, P. armeniaca-S2 AAT69245.1, P. mume-S11 ABV71999.1, P. salicina-Se BAF91848.1, P. cerasus-S36a ABW17266.1, and P. avium-S22 ABR19609.1 for comparison with ‘Pru S-locus’ Hap1g16438 (pS1 allele), and M. domestica-S27 AAB70516.1, M. spectabilis-S4 ADB85482.1, Pyrus communis-Sk BAE92269.1, P. bretschneideri-S34 ABD72921.1, Sorbus aucuparia-S18 ABP01658.1, Cydonia oblonga-S1 ATL75302.1, Eriobotrya japonica-S6, and Crataegus pinnatifida-S2 for comparison with ‘Mal S-locus’ Hap1g12045 (mSa1 allele). Codon alignments were prepared from nucleotide sequences using Clustal Omega/v1.2.276 and manually checked. Phylogenetic trees were prepared with RAxML78 using protein alignments prepared with Clustal Omega/v1.2.276, and manually checked for known species relationships32. Ratios of non-synonymous substitutions to synonymous substitutions were calculated with codeml in PAML/v4.9i79, using default parameters.
Data availability
The datasets generated and/or analysed are available in the NCBI repository. For Gillenia genome, BioProjects PRJNA1196517 (haplotype 1, genome JBJYYC000000000) and PRJNA1197676 (haplotype 2, genome JBJYYD000000000) using BioSample SAMN45544832. For Malus genome, BioProjects PRJNA1196521 (haplotype 1, genome JBLJWW000000000) and PRJNA1197688 (haplotype 2, genome JBLJWX000000000) using BioSample SAMN45552643. For Gillenia IsoSeq, SRA SRR32173724 and SRR32173725 using BioSample SAMN45544833. For Gillenia mRNA-seq, SRA SRR32161566-SRR32161574 using BioSamples SAMN45544834-SAMN45544842. Also, the Gillenia genome and gene and TE annotations are available at the Genome Database for Rosaceae (www.rosaceae.org).
References
Igic, B., Lande, R. & Kohn, J. R. Loss of self-incompatibility and its evolutionary consequences. Int. J. Plant Sci. 169, 93–104 (2008).
Charlesworth, D., Vekemans, X., Castric, V. & Glémin, S. Plant self-incompatibility systems: a molecular evolutionary perspective. New Phytol. 168, 61–69 (2005).
Zhang, Y. & Xue, Y. in In Self-Incompatibility in Flowering Plants: Evolution, Diversity, and Mechanisms, 193–215 (eds Franklin-Tong, V. E.) (Springer, 2008).
Zhao, H. et al. Origin, loss, and regain of self-incompatibility in angiosperms. Plant. Cell. 34, 579–596 (2022).
Allen, A. M. & Hiscock, S. J. In Self-Incompatibility in Flowering Plants: Evolution, Diversity, and Mechanisms, 73–101 (ed Vernonica E. Franklin-Tong) (Springer Berlin Heidelberg, 2008).
Steinbachs, J. E. & Holsinger, K. E. S-RNase-mediated gametophytic self-incompatibility is ancestral in eudicots. Mol. Biol. Evol. 19, 825–829 (2002).
McClure, B. In Self-Incompatibility in Flowering Plants: Evolution, Diversity, and Mechanisms, 217–236 (ed Vernonica E. Franklin-Tong) (Springer Berlin Heidelberg, 2008).
Fujii, S., Kubo, K. & Takayama, S. Non-self- and self-recognition models in plant self-incompatibility. Nat. Plants. 2, 16130 (2016).
Kubo, K. et al. Collaborative non-self recognition system in s-rnase–based self-incompatibility. Science 330, 796–799 (2010).
Castric, V. & Vekemans, X. Plant self-incompatibility in natural populations: a critical assessment of recent theoretical and empirical advances. Mol. Ecol. 13, 2873–2889 (2004).
Harkness, A. & Brandvain, Y. Non-self recognition-based self-incompatibility can alternatively promote or prevent introgression. New Phytol. 231, 1630–1643 (2021).
Vekemans, X. & Castric, V. When the genetic architecture matters: evolutionary and ecological implications of self versus nonself recognition in plant self-incompatibility. New Phytol. 231, 1304–1307 (2021).
Kubo, K. et al. Gene duplication and genetic exchange drive the evolution of S-RNase-based self-incompatibility in Petunia. Nat. Plants. 1, 14005 (2015).
Vieira, J., Fonseca, N. A., Vieira, C. P., An & S-RNase-based gametophytic self-incompatibility system evolved only once in eudicots. J. Mol. Evol. 67, 179–190 (2008).
Zhou, J. et al. Structural and transcriptional analysis of S-locus F-box genes in Antirrhinum. Sex. Plant Reprod. 16, 165–177 (2003).
Sassa, H. et al. S locus F-box brothers: multiple and pollen-specific F-box genes with S haplotype-specific polymorphisms in Apple and Japanese Pear. Genetics 175, 1869–1881 (2007).
Entani, T. et al. Comparative analysis of the self-incompatibility (S-) locus region of Prunus mume: identification of a pollen-expressed F-box gene with allelic diversity. Genes Cells. 8, 203–213 (2003).
Ushijima, K. et al. Structural and transcriptional analysis of the self-incompatibility locus of almond: identification of a pollen-expressed F-box gene with haplotype-specific polymorphism. Plant. Cell. 15, 771–781 (2003).
Tsukamoto, T., Hauck, N. R., Tao, R., Jiang, N. & Iezzoni, A. F. Molecular and genetic analyses of four nonfunctional S haplotype variants derived from a common ancestral S haplotype identified in sour Cherry (Prunus cerasus L). Genetics 184, 411–427 (2010).
Vieira, J. et al. The identification of the Rosa S-locus and implications on the evolution of the Rosaceae gametophytic self-incompatibility systems. Sci. Rep. 11, 3710 (2021).
Aguiar, B. et al. Patterns of evolution at the gametophytic self-incompatibility Sorbus aucuparia (Pyrinae) S pollen genes support the non-self recognition by multiple factors model. J. Exp. Bot. 64, 2423–2434 (2013).
Liang, M. et al. Evolution of self-compatibility by a mutant Sm-RNase in citrus. Nat. Plants. 6, 131–142 (2020).
Aguiar, B. et al. Convergent evolution at the gametophytic self-incompatibility system in Malus and Prunus. PLoS One 10, e0126138 (2015).
Ushijima, K. et al. The S haplotype-specific F-box protein gene, SFB, is defective in self-compatible haplotypes of Prunus avium and P. mume. Plant J. 39, 573–586 (2004).
Tao, R. & Iezzoni, A. F. The S-RNase-based gametophytic self-incompatibility system in Prunus exhibits distinct genetic and molecular features. Sci. Hort. 124, 423–433 (2010).
Du, J. et al. Molecular characteristics of S-RNase alleles as the determinant of self-incompatibility in the style of Fragaria viridis. Hortic. Res. 8, 185 (2021).
Lv, S. et al. The origin and evolution of RNase T2 family and gametophytic self-incompatibility system in plants. Genome Biol. Evol. 14, evac093 (2022).
Minamikawa, M. et al. Apple S locus region represents a large cluster of related, polymorphic and pollen-specific F-box genes. Plant Mol. Biol. 74, 143–154 (2010).
Evans, R. C. & Campbell, C. S. The origin of the Apple subfamily (Maloideae; Rosaceae) is clarified by DNA sequence data from duplicated GBSSI genes. Am. J. Bot. 89, 1478–1484 (2002).
Velasco, R. et al. The genome of the domesticated Apple (Malus X domestica Borkh). Nat. Genet. 42, 833–839 (2010).
Ireland, H. S. et al. The Gillenia trifoliata genome reveals dynamics correlated with growth and reproduction in Rosaceae. Hortic. Res. 8, 233 (2021).
Xiang, Y. et al. Evolution of Rosaceae fruit types based on nuclear phylogeny in the context of geological times and genome duplication. Mol. Biol. Evol. 34, 262–281 (2017).
Potter, D. et al. Phylogeny and classification of Rosaceae. Plant Syst. Evol. 266, 5–43 (2007).
Maliepaard, C. et al. Aligning male and female linkage maps of Apple (Malus pumila Mill.) using multi-allelic markers. Theor. Appl. Genet. 97, 60–73 (1998).
Ballester, J. et al. Location of the self-incompatibility gene on the almond linkage map. Plant. Breed. 117, 69–72 (1998).
Igic, B. & R Kohn, J. Evolutionary relationships among self-incompatibility RNases. Proc. Natl. Acad. Sci. USA. 98, 13167–13171 (2001).
Vieira, J., Morales-Hojas, R., Santos, R. A. & Vieira, C. P. Different positively selected sites at the gametophytic self-incompatibility pistil S-RNase gene in the Solanaceae and Rosaceae (Prunus, Pyrus, and Malus). J. Mol. Evol. 65, 175–185 (2007).
Matsumoto, D. & Tao, R. Distinct self-recognition in the Prunus S-RNase-based gametophytic self-incompatibility system. Hortic. J. 85, 289–305 (2016).
Daccord, N. et al. High-quality de Novo assembly of the Apple genome and methylome dynamics of early fruit development. Nat. Genet. 49, 1099–1106 (2017).
Wang, J. et al. Chromosome-scale genome assembly of sweet Cherry (Prunus avium L.) Cv. Tieton obtained using long-read and Hi-C sequencing. Hortic. Res. 7, 122 (2020).
D’Amico-Willman, K. M. et al. Whole-genome sequence and methylome profiling of the almond [Prunus dulcis (Mill.) D.A. Webb] cultivar ‘Nonpareil’. G3 (Bethesda) 12 (2022).
Raymond, O. et al. The Rosa genome provides new insights into the domestication of modern roses. Nat. Genet. 50, 772–777 (2018).
White, D. & Pirro, S. The complete genome sequences of three species of Mountain Avens (Dryas, Rosaceae). Biodiversity Genomes 2022 (2022).
Jung, S. et al. 15 Years of GDR: new data and functionality in the genome database for Rosaceae. Nucleic Acids Res. 47, D1137–D1145 (2019).
Wang, X. Y. & Paterson, A. H. Gene conversion in angiosperm genomes with an emphasis on genes duplicated by polyploidization. Genes (Basel). 2, 1–20 (2011).
Ou, S. et al. Benchmarking transposable element annotation methods for creation of a streamlined, comprehensive pipeline. Genome Biol. 20, 275 (2019).
Wang, S., Kakui, H., Kikuchi, S., Koba, T. & Sassa, H. Interhaplotypic heterogeneity and heterochromatic features May contribute to recombination suppression at the S-locus in Apple (Malus×domestica). J. Exp. Bot. 63, 4983–4990 (2012).
Langley, C. H., Montgomery, E., Hudson, R., Kaplan, N. & Charlesworth, B. On the role of unequal exchange in the containment of transposable element copy number. Genet. Res. 52, 223–235 (1988).
Talbert, P. B. & Henikoff, S. Centromeres convert but don’t cross. PLoS Biol. 8, e1000326 (2010).
Vicient, C. M. & Casacuberta, J. M. Impact of transposable elements on polyploid plant genomes. Ann. Botany. 120, 195–207 (2017).
Boskovic, R. I., Sargent, D. J. & Tobutt, K. R. Genetic evidence that two independent S-loci control RNase-based self-incompatibility in diploid strawberry. J. Exp. Bot. 61, 755–763 (2010).
Huang, S. X. et al. Competitive interaction between two functional S-haplotypes confer self-compatibility on tetraploid Chinese Cherry (Prunus Pseudocerasus Lindl. CV. Nanjing Chuisi). Plant Cell Rep. 27, 1075–1085 (2008).
Gu, C. et al. Identification of S-genotypes in Chinese Cherry cultivars (Prunus Pseudocerasus Lindl). Tree. Genet. Genomes. 6, 579–590 (2010).
Jahnen, W., Lush, W. M. & Clarke, A. E. Inhibition of in vitro pollen tube growth by isolated S-Glycoproteins of Nicotiana alata. Plant. Cell. 1, 501–510 (1989).
López-Girona, E. et al. CRISPR-Cas9 enrichment and long read sequencing for fine mapping in plants. Plant. Methods 16 (2020).
Cheng, H., Concepcion, G. T., Feng, X., Zhang, H. & Li, H. Haplotype-resolved de Novo assembly using phased assembly graphs with hifiasm. Nat. Methods. 18, 170–175 (2021).
Alonge, M. et al. Automated assembly scaffolding using ragtag elevates a new tomato system for high-throughput genome editing. Genome Biol. 23, 258 (2022).
Bradnam, K. R. et al. Assemblathon 2: evaluating de Novo methods of genome assembly in three vertebrate species. Gigascience 2, 10 (2013).
Kurtz, S. et al. Versatile and open software for comparing large genomes. Genome Biol. 5, R12 (2004).
Simao, F. A., Waterhouse, R. M., Ioannidis, P., Kriventseva, E. V. & Zdobnov, E. M. BUSCO: assessing genome assembly and annotation completeness with single-copy orthologs. Bioinformatics 31, 3210–3212 (2015).
Seppey, M., Manni, M. & Zdobnov, E. M. BUSCO: Assessing Genome Assembly and Annotation Completeness. Methods in Molecular Biology 1962, 227–245 (2019).
Bianco, L. et al. Development and validation of a 20K single nucleotide polymorphism (SNP) whole genome genotyping array for Apple (Malus X domestica Borkh). PLoS One. 9, e110377 (2014).
Celton, J. M., Tustin, D. S., Chagné, D. & Gardiner, S. E. Construction of a dense genetic linkage map for Apple rootstocks using SSRs developed from Malus ESTs and Pyrus genomic sequences. Tree. Genet. Genomes. 5, 93–107 (2009).
Su, W., Ou, S., Hufford, M. B. & Peterson, T. A tutorial of EDTA: extensive de Novo TE annotator. Methods Mol. Biol. 2250, 55–67 (2021).
Chen, N. Using RepeatMasker to identify repetitive elements in genomic sequences. Current Protocols in Bioinformatics Chap. 4, Unit 4 10 (2004).
Tempel, S. Using and Understanding repeatmasker. Methods Mol. Biol. 859, 29–51 (2012).
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Bruna, T., Hoff, K. J., Lomsadze, A., Stanke, M. & Borodovsky, M. BRAKER2: automatic eukaryotic genome annotation with GeneMark-EP + and AUGUSTUS supported by a protein database. NAR Genomics Bioinf. 3, lqaa108 (2021).
Lohse, M. et al. Mercator: a fast and simple web server for genome scale functional annotation of plant sequence data. Plant, Cell Environ. 37, 1250–1258 (2014).
Schwacke, R. et al. MapMan4: A refined protein classification and annotation framework applicable to multi-omics data analysis. Mol. Plant. 12, 879–892 (2019).
Okada, K. et al. Related polymorphic F-box protein genes between haplotypes clustering in the BAC contig sequences around the S-RNase of Japanese Pear. J. Exp. Bot. 62, 1887–1902 (2011).
Darling, A. C., Mau, B., Blattner, F. R. & Perna, N. T. Mauve: multiple alignment of conserved genomic sequence with rearrangements. Genome Res. 14, 1394–1403 (2004).
He, W. et al. NGenomeSyn: an easy-to-use and flexible tool for publication-ready visualization of syntenic relationships across multiple genomes. Bioinformatics 39, btad121 (2023).
Grabherr, M. G. et al. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat. Biotechnol. 29, 644–652 (2011).
Ewels, P. A. et al. The nf-core framework for community-curated bioinformatics pipelines. Nat. Biotechnol. 38, 276–278 (2020).
Sievers, F. & Higgins, D. G. Clustal Omega for making accurate alignments of many protein sequences. Protein Sci. 27, 135–145 (2018).
Guindon, S. et al. New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0. Syst. Biol. 59, 307–321 (2010).
Stamatakis, A. RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22, 2688–2690 (2006).
Yang, Z. PAML 4: phylogenetic analysis by maximum likelihood. Mol. Biol. Evol. 24, 1586–1591 (2007).
Acknowledgements
The authors thank Monica Dragulescu and G Wadasinghe for maintenance of the Gillenia plants, Richard Newcomb and Kevin Davies for internal review of the manuscript, and Chao Gu and Ying Xu from Nanjing Agricultural University for pollen tube assay protocols.
Funding
This work was funded by Strategic Science Investment Fund (SSIF) grant C11X1702 from the New Zealand Ministry of Business, Innovation and Employment, and by Project-ANR-21-CE20-0005 On the PopuLation gEnomics of trAnsposable elementS in frUit tReEs – PLEASURE from the Agence Nationale de la Recherche (ANR).
Author information
Authors and Affiliations
Contributions
Molecular biology experimentation by RW for Gillenia, and DB and TF for Malus. Pollen tube germination and growth assays by RW. De novo transcriptome assemblies, mRNA-seq and IsoSeq computational analyses by CHD. De novo assembly of contigs by WM, AC and DC for Gillenia, scaffolding to chromosome-level and telomere checks by CHD. Malus ‘M9’ assembly by ELG. Gtr.v2 repeats detection and masking by THC, gene prediction and annotation by CHD and THC, and by HSI for manual annotation. Genome synteny analyses by CHD, HSI, and RS for Gillenia, and ELG for Malus. Phylogenetic analyses by RW and HSI. Positive selection analyses by HSI. Manuscript written by HSI and RS.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Wang, R., Deng, C.H., Cornille, A. et al. Characterisation of the Gillenia S-locus provides insight into evolution of the nonself-recognition self-incompatibility system in apple. Sci Rep 15, 14630 (2025). https://doi.org/10.1038/s41598-025-99335-8
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41598-025-99335-8
