Characterization of the mating-type loci in P. ostreatus strains PC9 and PC15

High-quality genome assemblies of P. ostreatus strains PC9 and PC15 are available [18–21], yet the mating-type loci remain incompletely characterized. To improve gene model accuracy within the mating-type regions and to ensure consistent locus characterization between strains, we reannotated the PC15 genome available from JGI (assembly accession number: GCA_000697685.1). In basidiomycetes, the matA locus is typically flanked by the highly conserved mitochondrial intermediate peptidase (MIP) and β-flanking (βFG) genes [22]. Using these conserved flanking genes, we identified and characterized the boundaries of the P. ostreatus matA locus. The locus spans ~5kb in PC9 and ~10kb in PC15; both are located on chromosome 2 (Fig 1A). The matA locus encodes two homeodomain (HD) genes, HD1 and HD2, which can form heterodimers with compatible HD proteins from the opposite mating partner, enabling intracellular recognition of sexual compatibility after mating [23]. Notably, PC9 carries two HD genes (HD1.1 and HD2.1) transcribed in opposite directions, whereas, PC15 harbors four HD genes (HD1.2, HD1.3, HD2.2, and HD2.3) in the HD locus (Fig 1A), suggesting a gene duplication event in PC15. A previous study reported only a single HD2 gene in the PC15 matA locus [13]; however, examination of the PC15 JGI genome assembly (GCA_000697685.1) confirmed the presence of two HD2 genes [19,21,20], consistent with the complete gene content observed in our analysis. To assess allelic divergence, we aligned the protein sequences of HD1 and HD2 from both strains. While all HD proteins share conserved homeodomain motifs, we still observed distinct sequence variations among HD1 and HD2 protein (S1A and S1B Fig), supporting the conclusion that PC9 and PC15 possess genetically distinct and mating-compatible matA alleles.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Characterization of mating-type loci and associated genes in P. ostreatus.

(A and B) MAT locus alignments (amino acid-based) of strains PC9 and PC15 (clinker v0.0.31). Genes encoding proteins having more than 95% identical amino acid residues are linked. The color code for links and genes highlights sequence homology and grey genes have no similarity to other genes in the alignment. (A) In the matA locus on chromosome 2 (Chr2), homeodomain genes HD1 and HD2 are shown in green and pink, respectively; the conserved mitochondrial intermediate peptidase (MIP) and beta flanking gene (βFG) are shown in teal and purple, respectively. (B) In the matB locus on chromosome 9 (Chr9), mating-type specific pheromone receptor genes (STE3) are shown in orange (within matB) and blue (outside matB); pheromone genes (PHB) are in green; the conserved major facilitator superfamily (MFS) transporter and UbiA family prenyltransferase (UbiA) are shown in teal and purple, respectively; other genes are shown in grey. Scale bars indicate 2.5 kb. (C and D) Phylogenetic trees of HD proteins (C) and Ste3 proteins (D) from P. ostreatus (PC9 in red underline; PC15 in yellow underline) and S. commune by the maximum-likelihood method using IQ-TREE.

https://doi.org/10.1371/journal.pgen.1011966.g001

In both PC9 and PC15, the matB locus is located on chromosome 9, spanning ~54 kb and ~40 kb, respectively (Fig 1B). The locus contains mating pheromone precursor (PHB) and pheromone receptor (STE3) genes, which are involved in premating recognition [6]. In the model basidiomycetes, the matB locus is flanked by various genes. For example, in C. cinereus, the B locus is bordered by a gene encoding major facilitator superfamily (MFS) transporter [24,25], whereas in S. commune, the matB locus is flanked by genes encoding a zinc finger transcription factor and hypothetical proteins [26]. Similarly, in P. ostreatus strain PC9 and PC15, the matB is flanked by an MFS transporter on one side and a prenyltransferase (UbiA) gene on the other, providing genomic boundaries for matB locus. To identify the mating-type genes within matB, we searched for conserved Pfam domains corresponding to STE3 (Pfam: PF02076) and PHB (Pfam: PF08015) domains. In total, four STE3 genes (STE3.1 to STE3.4) and five PHB genes (PHB1 to PHB5) were identified in PC9, and five STE3 genes (STE3.7 to STE3.11) and six PHB genes (PHB6 to PHB11) were identified in PC15 (S1 Table). Four additional genes encoding Ste3 proteins, STE3.5, STE3.6, STE3.12, and STE3.13, that share high sequence identity (> 96%) were found flanking the matB locus, suggesting that they are not mating-type-specific (S2 Table). Alignment of all nine STE3 receptors encoded in the matB locus revealed notable variation in the C-terminal regions (S1C Fig), consistent with their roles in mating signal recognition and allelic specificity. We further identified genes that encode mating pheromone precursors (PHB) within the matB locus. Across both alleles, we identified 11 PHB genes encoding peptides of 51–72 amino acids in length with a canonical C-terminal CAAX motif (C = cysteine, A = aliphatic residue, X = any residue) characteristic of mating pheromones [27]. Protein sequence alignment confirmed conservation of this motif among all P. ostreatus Phb peptides (S1E Fig).

To explore the evolutionary divergence of mating-type genes across basidiomycetes, we constructed maximum-likelihood phylogenetic trees of HD and Ste3 protein from P. ostreatus and the model mushroom Schizophyllum commune (Fig 1C and 1D). Phylogenetic analysis of HD proteins revealed two distinct, well-supported clades corresponding to HD1 and HD2 (bootstrap support = 100%). Within each clade, sequences clustered by species, forming subgroups indicative of taxon-specific diversification (Fig 1C). The clear separation between HD1 and HD2, along with their internal species-level groupings, suggests long-term evolutionary maintenance and potential functional specialization of these mating-type genes. In contrast, Ste3 proteins grouped first by allele type and then further separated by species (Fig 1D). Within these, Ste3.2 and Ste3.8 shared 99.5% amino acid identity, and Ste3.3 and Ste3.9 shared 96.7% identity (S1D Fig and S2 Table). By contrast, the protein sequences of Ste3.1 versus Ste3.7 (75.7% identity) and Ste3.4 versus Ste3.11 (55.4% identity) are considerably more divergent, further supporting their roles as distinct mating-type pheromone receptors (S1D Fig).

Multiallelic mating-type loci in P. ostreatus

In tetrapolar fungi, mating-type loci are often multiallelic rather than biallelic, resulting in the presence of hundreds or even thousands of mating-types, as seen in C. cinereus and S. commune, which have over 12,000 and 20,000 mating types, respectively [28]. P. ostreatus also follows a multiallelic tetrapolar mating system, though the exact structure and diversity of its mating-type alleles remain poorly characterized. To investigate its multiallelism, we compared the matA and matB loci across several P. ostreatus genomes (Table 1). Our dataset comprised two monokaryotic strains, SS5 and SS14, derived from strain TWF713, as well as additional publicly available P. ostreatus genomes retrieved from NCBI and JGI databases. These include monokaryotic strains CCMSSC00389 (389) [29], CCMSSC03989 (3989) [30], two haplotypes of the dikaryotic strain gfPleOstr1.1 (gf and gfalt) [31], and the dikaryotic strains DSM11191 (DSM) and 595 [32]. We then examined the matA and matB regions across all strains, focusing on their organization, gene content, and allelic variation to better understand the molecular basis of multiallelism in P. ostreatus. In this study, matA and matB alleles are defined based on complete nucleotide and protein identity. Alleles are distinct if they differ in nucleotide and protein sequence.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. P. ostreatus genomes used in the present study.

https://doi.org/10.1371/journal.pgen.1011966.t001

Comparison of HD genes across ten P. ostreatus genomes revealed 17 HD1 and 12 HD2 alleles within these matA loci (Fig 2A and S3 Table). While at least 12 distinct matA alleles have been reported previously [11], their structural organization and sequence-level variation were not described. To avoid potential overlap with existing allele designations, we named the characterized alleles in this study from A21-A29. However, we cannot rule out the possibility that some of these may be identical in sequence or in mating type to previously reported alleles. SS14 harbors a matA allele identical to that of PC9 (allele A2), exhibiting complete sequence identity at both nucleotide and amino acid levels. Nine novel matA alleles were identified among the remaining haplotypes (Fig 2A). Six haplotypes contained a single HD1/HD2 gene pair (Fig 2A), although duplication of HD1 was observed in SS5 (A21), 389 (A22), and 595 (A26). Notably, SS5 (A21) carries three HD1 genes, a configuration similar to that recently reported for P. ostreatus strain PC80, derived from an artificially cultivated strain HeiKang 650 [33]. The gfalt haplotype (A29) exhibited duplication of both HD1 and HD2, consistent with our previous observations in PC15. Moreover, the matA locus of gfalt (A29) expanded to 73 kb—substantially larger than the 6 kb locus in PC9—and contained an insertion of five genes, including four hypothetical proteins and one transposase (Fig 2A). Given the sequence variation and gene duplications we had observed, we performed phylogenetic analysis to assess the divergence of HD1 and HD2 proteins across strains. This analysis revealed two distinct clusters for HD1 (HD1.1 to HD1.17) and HD2 (HD2.1 to HD2.13), confirming the divergence of these gene families (Fig 2C).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Comparative analysis of matA and matB loci across P. ostreatus strains.

(A and B) MAT locus alignments (amino acid-based) of 12 haplotypes PC9, PC15, SS5, SS14, CCMSSC00389 (389), CCMSSC03989 (3989), gfPleOstr1.1 (haplotypes gf and gfalt), and two dikaryotic strains: DSM11191 (DSM) and 595 (clinker v0.0.31). Genes encoding proteins having more than 95% identical amino acid residues are linked. The color code for links and genes highlights sequence homology and grey genes have no similarity to other genes in the alignment. (A) Homeodomain genes HD1 and HD2 are shown in green and pink, respectively; the conserved MIP and βFG are shown in teal and purple, respectively. (B) Pheromone receptor genes (STE3) are shown in orange (within matB) and blue (outside matB); pheromone genes (PHB) are in green; the conserved MFS and UbiA are shown in teal and purple, respectively; other genes are shown in grey. (C and D) Phylogenetic trees of 30 HD proteins (C) and 48 Ste3 proteins (D) from 12 P. ostreatus haplotypes by the maximum-likelihood method using IQ-TREE (PC9 in red underline; PC15 in yellow underline).

https://doi.org/10.1371/journal.pgen.1011966.g002

At the matB locus, we identified ten new alleles (B21-B30) that are absent in either PC9 or PC15, with locus sizes ranging from 45 kb to 150 kb (Fig 2B). The locus is flanked by MFS and UbiA genes, and most alleles contain three to five STE3 and four to six PHB genes. One to two STE3-like genes were also found immediately outside the UbiA boundary in different strains; these receptors share high similarity and therefore were considered non-mating-type-specific. The matB loci of SS5 (B21), SS14 (B22), and 595 (B28) were notably expanded—to 105 kb, 77 kb and 150 kb, respectively—compared to the 54 kb locus in PC9. Several additional genes were detected within these expanded regions, including an integrase, a retrotransposon Gag protein, and hypothetical proteins, which are unlikely to be directly involved in mating but may have contributed to the locus expansion. Phylogenetic analysis of Ste3 protein sequences illustrates their diversity across P. ostreatus strains, aligning with the multiallelic nature of the matB locus and with the locus’s role in facilitating diverse mating interactions (Fig 2D). In total, we identified distinct 12 matB distinct alleles across strains containing 48 STE3-like pheromone receptor genes, confirming the multiallelic configuration of both loci in P. ostreatus.

To test whether allelic variation affects mating compatibility, we performed reciprocal crosses among SS5 (A21B21), SS14 (A2B22), PC9 (A2B1) and PC15 (A1B2). SS5 is compatible with both PC9 and PC15, as evidenced by the presence of clamp connections—hallmark structures of dikaryotic hyphae in basidiomycetes—and successful fruiting body development (Fig 3). The SS14 x PC9 cross produced neither clamp connections nor fruiting bodies, as expected from their both harbor the A2 allele of the matA locus (Fig 3). In contrast, the SS14 x PC15 cross yielded clamp connections and fruiting bodies (Fig 3). These results support that P. ostreatus follows a tetrapolar mating system, in which successful mating requires compatibility at both matA and matB loci.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Clamp connection and fruiting body formation in interstrain crosses of P. ostreatus.

Representative crosses of SS5 (A21B21) x PC9 (A2B1), SS5 (A21B21) x PC15 (A1B2), SS14 (A2B22) x PC9 (A2B1), and SS14 (A2B22) x PC15 (A1B2) display hallmark features of successful mating in a tetrapolar system, including clamp connections (arrows) and the fruiting body formation. Scale bar = and 1 cm (upper panels) and 10 µm (lower panels).

https://doi.org/10.1371/journal.pgen.1011966.g003

Construction of the congenic strain PC9.15 via genetic backcrossing

PC9 and PC15 are monokaryotic strains derived from the dikaryotic strain N001, but only ~94% of their genomes are aligned with high similarity (~97%). This high genetic diversity likely contributes to the significant growth differences observed between the two strains. Given the robust growth and well-annotated genome of PC9, we selected the PC9 genetic background for introgression of the PC15 mating-type loci to generate a congenic strain (Fig 4A). To construct a congenic strain with PC9 background, we first crossed PC9 and PC15 and isolated meiotic progeny by dissecting individual basidiospores. Randomly selected progeny carrying the A1B2 mating type as determined by PCR was then backcrossed with PC9 (A2B1). After nine additional rounds of backcrossing, we obtained a congenic strain that we designated PC9.15, which carries the A1B2 mating-type loci from PC15 (Figs 4B and S2A). We found that PC9.15 exhibited growth behavior comparable to PC9 on nutrient-rich medium (YMG) and readily formed fruiting bodies when crossed with PC9 (Fig 4A and 4C). Overall, phenotypic characterization of PC9.15 showed that it retains growth and reproductive traits of PC9 while carrying the A1B2 mating-type loci from PC15 (S2 Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Construction of the congenic strain PC9.15 with a PC9 genetic background and PC15 mating-type loci.

(A) PC9, PC15, and PC9.15 were grown on YMG agar media in 5.5-cm diameter Petri plates for 10 days. Scale bar = 1 cm. (B) Crossing strategy used to generate the PC9.15 congenic strain: Monokaryotic “Bc1#” progeny were obtained from a PC9 x PC15 cross. Progeny carrying the PC15 (A1B2) mating type were backcrossed to PC9 (A2B1) to generate monokaryotic “Bc2” progeny. This process was repeated for ten crosses to produce PC9.15. (C) Fruiting body development from a PC9.15 x PC9 cross at days 1, 3, and 5. Scale bars = 1 cm.

https://doi.org/10.1371/journal.pgen.1011966.g004

High-quality genome assembly and analysis of the congenic strain PC9.15

To confirm the genome similarity of PC9.15 to PC9 and to characterize its chromosomal architecture, we generated a high-quality, near-chromosome-level genome assembly of the P. ostreatus congenic strain PC9.15. Using Nanopore sequencing, we achieved 128x coverage and assembled a contiguous genome of ~35.3 Mb (Table 2), comparable to the genome sizes of PC9 (35.0 Mb) and PC15 (34.3 Mb) [18–21]. The PC9.15 genome is distributed across 14 scaffolds, ranging from 38.7 kb to 5.36 Mb, including one scaffold representing the mitochondrial genome. Among the 11 main scaffolds analyzed, nine exhibit telomere-to-telomere assembly, one possesses telomeric repeats at a single end, and one lacks telomeric sequences. Although P. ostreatus is known to contain 11 chromosomes [14], our assembly of PC9.15 includes 13 scaffolds. The two additional scaffolds, representing only 1.4% of the total genome size, likely representing unresolved gaps or unplaced regions, such as centromeric or telomeric sequences. For example, scaffold 12 (468,928 bp) contains telomeric repeats and may correspond to a terminal chromosome region, while scaffold 13 (38,780 bp) shows high coverage depth but without gene density, a characteristic associated with centromeres [34], suggesting it may represent a misplaced centromeric fragment. Collectively, these results indicate that the PC9.15 genome assembly reaches near-chromosome-level resolution, with 11 chromosome-like scaffolds.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Genomic features of three P. ostreatus genome assemblies: PC9, PC9.15, and PC15. nt, nucleotides.

https://doi.org/10.1371/journal.pgen.1011966.t002

To assess genome completeness and structural features of the PC9.15 assembly, we analyzed gene density, sequence coverage, and BUSCO scores. Gene density is consistent across 11 scaffolds, averaging approximately 41 genes per 100 kb, with an average sequencing depth of 71.8 reads per 10 kb (Fig 5A). Coverage is generally uniform across chromosomes, with localized increases on chromosomes 2 and 6 (Fig 5A), possibly due to repetitive sequences, such as transposable elements (TEs). Assembly completeness, assessed using the basidiomycota_odb10 database [35], yielded a BUSCO score of 99.4%, comparable to that of PC9 (98.6%) and PC15 (99.3%) [18], confirming the high quality of the PC9.15 assembly. Overall, these results confirm that the PC9.15 genome is well-assembled, complete, and suitable for further genomics analysis.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Genome architecture of P. ostreatus congenic strain PC9.15 and comparative analysis of assembled genomes of PC9.15, PC9, and PC15.

(A) Tracks (outer to inner) represent the distribution of genomic features in our PC9.15 assembly: (1) size (in Mb) of PC9.15 scaffolds, with numbers prefixed by the letter “C” indicating the chromosome and “sc” indicating the scaffold; (2) gene density (along 100-kb sliding windows, ranging between 0 and 50 protein-coding genes); (3) distribution of TEs along the genome; (4) telomere repeat frequency (along 1-kb sliding windows, ranging between 10 and 30 repeats); and (5) depth of sequencing coverage (along 10-kb sliding windows, ranging between 0 and 150 depth). (B) Whole-genome dot-plot alignment of PC9.15 (query; y-axis) against PC9 (reference; x-axis) generated with D-Genies. Each dot represents an aligned segment, colored by nucleotide identity as indicated in the scale. The summary values report the fraction of total query (PC9.15) sequence length that is unaligned (“No match”) or aligns to the reference within each identity bin, including the fraction aligning at >75% identity (99.47%). (C) Circos plot showing regions of similarity shared by the PC9.15 (C1-C11) and PC9 (C1-C11) (identity > 95%, length > 10kb). (D) Circos plot showing regions of similarity shared by the PC9.15 (C1-C12) and PC15 (C1-C12) strains (identity > 95%, length > 10kb). The grey box represents segment A. Tracks (outer to inner) represent the distribution of genomic features in each assembly: (1) chromosome sizes (in Mb) of chromosomes, with numbers prefixed by the letter “C” indicating the chromosome and “sc” indicating the scaffold; (2) distribution of telomere repeats with 1-kb sliding windows, ranging between 10-30 repeats.

https://doi.org/10.1371/journal.pgen.1011966.g005

To evaluate genomic similarity and structural variation, we compared the PC9.15 genome to its parental strains, PC9 and PC15. Annotation of the PC9.15 genome using funannotate (v1.8.14) [36] identified 14,118 genes (S5 Table), higher than in PC9 (11,875 genes) [18] or PC15 (12,330 genes) [19–21]. To control for annotation inconsistencies arising from version differences, we re-annotated the PC9 genome using the same version of funannotate (v1.8.14), which yielded 12,003 genes. Therefore, the higher gene count observed in PC9.15 likely reflects a combination of improved assembly contiguity. To assess genome-wide relatedness, we used D-Genies to generate whole-genome dot-plot alignments between PC9.15 and its parental strains, PC9 and PC15 [37]. D-Genies summary statistics indicated that ~99.47% of the PC9 assembly aligned to PC9.15 at >75% nucleotide identity (Fig 5B), whereas only ~56.31% of the PC15 assembly aligned to PC9.15 under the same threshold. This demonstrates that PC9.15 is genetically closer to PC9, consistent with its congenic background. Comparative analysis revealed several chromosomal rearrangements in PC9.15 relative to PC9. Circos plot alignment showed that a segment from PC9 chromosome 4 had been translocated to PC9.15 chromosome 2, and a region from PC9 chromosome 7 translocated to and became inverted on PC9.15 chromosome 4 (Fig 5C). Additionally, we observed a unique region (segment A) on chromosome 4 of PC9.15 that did not align with the PC9 genome (Fig 5C), but partially matched chromosome 4 of PC15 (Fig 5D), indicating that this region retains PC15-derived sequence. The persistence of this segment may reflect suppressed recombination, potentially caused by repetitive sequences, structural rearrangements, or local chromosomal features that reduce crossover frequency, as observed in analogous recombination-suppressed regions [38].

Moreover, we annotated the PC9 mitochondrial genome, revealing a 65 kb circular molecule (S3 Fig). Given that PC9 and PC15 are descended from the strain N001, we expected their mitochondrial genomes to be identical. Indeed, our comparison showed 99.9% similarity between the mitochondrial genomes of PC9, PC15, and PC9.15 (S4 Fig and S6 Table).

Analysis of mating-type loci in the congenic strain PC9.15

To confirm the mating-type configuration of the congenic strain PC9.15, we analyzed the structure of its matA and matB loci and compared them to those of PC9 and PC15. As in parental strains, the matA and matB loci of PC9.15 are located on chromosome 2 and 9, respectively. The matA locus contains four HD genes (HD1.2, HD1.3, HD2.2, and HD2.3), and the matB locus includes five STE3 genes (STE3.7 to STE3.11) and six pheromone precursor genes (PHB6 to PHB11), identical to PC15 (Fig 6A). These results confirm that PC9.15 carries the A1B2 mating-type loci inherited from PC15.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Characterization of mating-type loci and associated genes in PC9.15.

Structures of the matA and matB loci (A) and the large-scale chromosomal (B) matA and matB loci compared across the PC9, PC9.15, and PC15 strains. (A) MAT locus alignments (amino acid-based) of strains PC9, PC9.15, and PC15 (clinker v0.0.31). Genes encoding proteins having more than 95% identical amino acid residues are linked. The color code for links and genes highlights sequence homology and grey genes have no similarity to other genes in the alignment. In the matA locus, homeodomain genes HD1 and HD2 are shown in green and pink, respectively; the conserved MIP and βFG are shown in teal and purple, respectively. In the matB locus, mating-type specific pheromone receptor genes (STE3) are shown in orange (within matB) and blue (outside matB); pheromone genes (PHB) are in green; the conserved MFS and UbiA are shown in teal and purple, respectively; other genes are shown in grey. (B) Large-scale alignment of the matA and matB loci between PC9, PC9.15, and PC15. The matA and matB loci of PC9 is shown as red line; those of PC9.15 and PC15 are shown as pink lines. The shades of grey show nucleotide sequence identity. The figures were generated by Easyfig v2.2.2.

https://doi.org/10.1371/journal.pgen.1011966.g006

To further identify the recombination sites around the mating-type loci, we expanded the sequence alignment to ~3 Mb for the matA locus and ~600 kb for the matB locus (Fig 6B). PC9.15 retained segments of the PC15 genomic background near both loci, with approximately 1.13 Mb around matA and 300 kb around matB. These retained regions raise the possibility of locally reduced recombination around the mat loci. Despite these regions of difference, the remainder of the PC9.15 genome is largely derived from PC9. Together, these data confirm the successful construction of a congenic strain that harbors the PC15 mating-type loci (A1B2) within the PC9 genetic background.

Diversity and genomic organization of mating-type loci in P. ostreatus

In this study, we investigated the structure and diversity of matA and matB alleles in P. ostreatus. Given the diversity of mating-type loci across basidiomycetes, we compared our findings to other well-characterized systems. In C. cinereus, the matB locus is organized into three distinct gene groups, each containing a single pheromone receptor gene (RCB) and multiple pheromone genes (PHB) [25]. Similarly, phylogenetic analysis of Ste3 proteins in P. ostreatus revealed five distinct groups (Fig 2D). In most P. ostreatus strains, PHB genes are located near STE3 genes, and the number of PHB genes is typically equal to or greater than the number of STE3 genes. The frequent physical proximity of PHB and STE3 gene, together with the higher copy number of PHB relative to STE3, indicates a characteristic genomic organization of the mating-type locus in P. ostreatus. This organization may allow P. ostreatus to fine-tune mating recognition through multiple pheromone-receptor combinations. Unlike the tightly clustered arrangement seen in C. cinereus, this dispersed structure likely reflects a different evolutionary path for generating mating-type diversity. In P. ostreatus, genes encoding the MAP kinase signaling cascade (e.g., STE20, STE11, STE12) are dispersed across multiple chromosomes rather than residing within a MAT locus as in C. neoformans [10]. Instead, the matB includes several reductases, transporters, and hypothetical proteins whose functions remain unclear but may still play a role in mating.

The tetrapolar mating-type system, with multiallelic matA and matB loci, promotes outcrossing in basidiomycetes [39]. For example, in C. cinereus, over 200 alleles per MAT locus yield a 99.5% outcrossing probability, whereas Ustilago maydis, with over 25 b alleles and only two a mating types, exhibits a lower outcrossing probability of 50% [40]. The diversity of mating-type alleles in P. ostreatus likely enhances outcrossing in natural populations. Our analysis identified 11 alleles at the matA locus and 12 alleles at the matB locus across 12 P. ostreatus haplotypes, suggesting the existence of at least 132 distinct mating types. These findings not only validate the multiallelic structure of P. ostreatus mating-type loci but also provide a detailed molecular characterization of allele diversity. Given this extensive allelic variation, additional matA and matB alleles likely remain unidentified in natural populations. Future studies on wild-isolated P. ostreatus strains will further enhance our understanding of the diversity and complexity of these loci and provide deeper insights into the evolution and regulation of the mating-type system in P. ostreatus.

Expansion and allelic diversity of MAT loci across strains

Across twelve P. ostreatus haplotypes, we identified 11 matA alleles and 12 matB alleles, confirming extensive multiallelism at both MAT loci. We also observed significant expansion in the physical size of these loci, including one expanded matA locus (73 kb in the gfalt haplotype versus 6 kb in PC9) and three expanded matB loci (77–150 kb versus 50 kb in PC9). The expanded regions contain additional genes, notably a transposase, an intergrase/recombinase-like protein, and a retrotransposon Gag protein, suggesting that transposable elements (TEs) and retrotransposon activity may has contributed to the structural expansion. Similar TE-associated expansion of the MAT locus has been reported in Cryptococcus, where accumulation of transposons within the MAT region has contributed to its expansion [41]. Moreover, the transition from an ancestral tetrapolar to a bipolar mating system in Cryptococcus involved fusion of the previously unlinked P/R and HD loci via chromosomal translocations, resulting in an expanded contiguous MAT locus [42]. Together, these findings support the hypothesis that TE insertions, together with integrase/recombinase activity, create a genomic environment that promotes structural rearrangements while preserving mating-type diversity in P. ostreatus.

Genomic rearrangement and recombination at the MAT loci

We observed regions surrounding both MAT loci, spanning ~1.13 Mb around matA and ~300 kb around matB lack recombination (Fig 6B). These regions are larger than the MAT loci themselves (matA ~ 10 kb; matB ~ 50 kb). Based on the P. ostreatus linkage map, the genome-wide average physical-to-genetic distance is estimated to be ~ 35 kb/cM [43], corresponding to ~ 32.1 cM for the 1.13 Mb matA-linked region and ~8.5 cM for the 300 kb matB-linked region. Thus, over 10 backcross generations, the probability of at least one recombinant arising within the ~ 1.13 Mb matA flanking region is relatively high (~96%), suggesting possibility of recombination suppression. On the other hand, the probability of at least one recombinant arising within the ~ 300 kb matB flanking region is 59%, leaving an ~ 41% chance that no recombinant occurs across this interval. Notably, recombination suppression around mating-type loci has been documented in other fungi; for example, Schizothecium tetrasporum exhibits suppression across a ~ 1.47 Mb region surrounding the mating-type locus [44].

Comparative analysis revealed several chromosomal rearrangements in the PC9.15 genome relative to PC9 and PC15. Chromosome translocations have been previously documented between PC9 and PC15 [18], so it is not surprising that the congenic strain also underwent large-scale rearrangements. For example, a portion of PC9 chromosome 4 (C4) has translocated to chromosome 11 (C11) in PC15 [18], and the same sequence (segment B) has translocated to chromosome 2 (C2) in PC9.15 (S5A and S5C Fig). Initially, we hypothesized that segment B originated from the centromeric region, which is often enriched in transposons and repetitive elements prone to recombination errors, as observed in C. neoformans [10]. However, further analysis revealed that segment B is present on both chromosome 2 (C2) and chromosome 4 (C4) in the PC9 genome, indicating that duplication events had occurred before the translocation was observed (S5B and S5D Fig). These findings suggest that the translocation of segment B likely resulted from recombination errors due to sequence similarity between duplicated segments.

Collectively, this study provides a detailed characterization of the P. ostreatus MAT loci in strains PC9 and PC15 and established PC9.15 as a congenic strain and a new genetic resource for precise genetic studies. We identified a multiallelic mating-type system with 11 matA and 12 matB alleles across twelve haplotypes. The PC9.15 congenic strain established here offers a versatile tool for probing the genetic and evolutionary dynamics of mating-type loci and for conducting future functional studies in this species.

Strains and culture conditions

Pleurotus ostreatus strains PC9 and PC15 differentiated from dikaryotic strain N001 [11], were used as the starting parental strains for genetic crosses. Spores were collected from a fruiting body of TWF713. Two monokaryon strains, SS5 and SS14, each originating from a single basidiospore, were established in the laboratory for analysis. Strains were maintained on yeast and malt extract with glucose (YMG) medium solidified with 1.5% (w/v) agar and incubated at 25°C.

Crossing, fruiting conditions and isolation of meiotic progeny

To induce dikaryon formation, the mating-compatible monokaryotic strains PC9 and PC15 were co-cultured on YMG agar medium for two weeks. Agar blocks from the mating interface were then transferred into sterile bottles containing sawdust medium (100 g of sawdust, 30 g wheat bran, tap water to 150 mL). These bottles were then incubated at 25°C in darkness for ten days. To induce fruiting body formation and obtain F1 progeny, cultures were shifted to 15°C under a 12 hr light/12 hr dark cycle. Fruiting bodies appeared after approximately seven days and matured within 3–4 days. Spores (F1 progeny) were harvested in Millipore water and micromanipulated using a dissecting microscope, with each plate containing 18 basidiospores. To assess reproductive performance, we recorded the number of germinated progeny per plate. Individual progeny were cultured, and their mating types were determined by PCR using mating-type-specific primers (S4 Table).

Congenic strain construction

PC9 was crossed with PC15, and single basidiospores were isolated. Three progeny carrying the PC15 mating type were selected and backcrossed to PC9. This process was repeated for a total of ten backcrosses to produce PC9.15, a congenic strain with mating loci derived from PC15 in the PC9 genetic background.

Genome sequencing and assembly comparison

The congenic strain PC9.15 and the monokaryotic strains SS5, and SS14 were cultured on YMG agar at 25°C for 7 days. Mycelia were then transferred to 50 mL of YMG liquid medium and incubated at 25°C and shaken at 200 rpm for 4 days. Genomic DNA was extracted using the cetyltrimethylammonium bromide (CTAB) method and purified by extraction with chloroform and phenol-chloroform and precipitation with isopropanol and ethanol, followed by an AMPure XP (Beckman Coulter) purification step. Libraries were prepared using the genomic DNA by Ligation kit (SQK-LSK109), and Nanopore long-read sequencing was performed using a FLO-MIN106 R9 flow cell (Oxford Nanopore) for PC9.15 and a R10 version flow cell (Oxford Nanopore) for SS5 and SS14. This approach yielded ~0.6 M reads for PC9.15, with a mean length of 7,400 bp, resulting in ~128X genome coverage. For SS5 and SS14, sequencing produced ~0.76 M and ~0.65M reads, with mean read lengths of 11,981 bp and 10,348 bp, corresponding to ~202X and ~142X coverage, respectively. De novo assembly of Nanopore long reads was performed using NECAT [45].

For comparison of the PC9, PC15, and PC9.15 genomes, we used the dnadiff program from the MUMmer package [46], which computed sequence identity and generates alignment statistics between them.

Genome annotation

Genome annotation for PC9.15, PC15, SS5, and SS14 was performed using funannotate (v1.8.1) [36] and the pipeline described at https://funannotate.readthedocs.io/en/latest/index.html. For PC9.15, we then used funannotate train with RNA-seq data and masked genome to generate preliminary gene models. RNA was extracted from the PC9.15 strain and sequenced on an Illumina NextSeq 500 Mid Output 300 cycles using a paired-end 150 bp configuration (PE150). For strains PC15, SS5, SS14, CCMSSC00389 [29], CCMSSC03989 [30], gfPleOstr1.1 (haplotypes gf and gfalt) [31], and 595 [32], gene prediction was performed directly with funannotate predict after masking, without RNA-seq-based training. Functional annotation was performed using the funannotate remote option, which incorporated InterProScan and antiSMASH to identify InterPro domains and secondary metabolite clusters, respectively. Finally, funannotate annotate was used to integrate functional annotations, including Pfam domains, clusters of orthologous groups (COGs), and EggNOG classifications. The full annotation of PC9.15 is provided in S5 Table.

The accession numbers of BioSample: SAMN05327826 (CCMSSC00389), SAMN09434509 (CCMSSC03989), SAMN23459969 (595), SAMEA8562045 (gfPleOstr1.1), and SAMEA8562045 (gfPleOstr1 alternate haplotype).

For STE3.2 in the PC9 genome, we manually curated the gene model by extending its 3’ end to include RNA-seq-supported 3’UTR sequence that was missing from the previous annotation.

Annotation of the PC9 mitochondrial genome

A putative large duplication in PC9 scaffold_14 was identified through self-alignment of the sequence and is suspected to be an artifact resulting from the circular nature of the mitochondrial DNA and overlapping regions in the assembled Nanopore reads. This hypothesis was confirmed by aligning the contig with the mitochondria genome of P. ostreatus P51 [47]. The duplicated region was manually removed, resulting in an assembled mitochondrial genome of PC9 that closely matched the size of the P51 mitochondria genome. Annotation was performed using MITOS2 [48].

Genome and mating-type loci comparison

General genome assembly statistics, including scaffold length and N50, were calculated using the count_fasta_residues.pl script (https://github.com/SchwarzEM/ems_perl/blob/master/fasta/count_fasta_residues.pl). Assembly completeness was assessed with BUSCO v6.0.0 using the basidiomycota_odb10 database [35,49]. Repetitive elements were identified using a custom repeat library generated using a previously described pipeline [50], and analyzed with RepeatMasker [51], TransposonPSI [52], and LTRharvest [53]. The results were further analyzed using the one_code_to_find_them_all script [54]. Telomeric repeats were identified by scanning for the motif (TTAGGG)n in scaffold sequences [55].

Genome comparison between PC9 and PC9.15 was performed using Circos (v0.69.0) [56] and D-genies (Minimap v2.24) [37] to visualize synteny and structural differences. Mating-type loci from both genomes were aligned using Easyfig [57] or clinker [58], and protein-level comparison of HD1, HD2, Ste3, and Phb were performed with needleall [59]. Multiple sequence alignments of HD, Ste3, and Phb proteins were generated using MEGA11 with the MUSCLE algorithm [60]. A phylogenetic tree was constructed by using IQ-TREE v3.0.1 [61] and further visualized using iTOL v7 [62]. The HD1/HD2 and receptor protein sequences used for phylogenetic tree construction were obtained from Schizophyllum commune. The corresponding sequences were retrieved from NCBI GenBank. The GenBank accession numbers can be found in S7 Table.

Multiple sequence alignments of HD1, HD2, Ste3, and Phb proteins were generated with the T-coffee program [63], and alignment results were visualized with Jalview v2.11.5.0 [64].

Cell wall staining and imaging

The fungal cultures were stained with SCRI Renaissance 2200 (SR2200), which labels fungal cell walls without affecting growth, and imaged using a LSM980 Airyscan 2 confocal microscope (Carl Zeiss).

Statistics

Two-tailed unpaired Student’s t-test was performed to determine the statistical difference between control and experimental samples in GraphPad Prism 9. P < 0.05 was considered significant; asterisks demonstrate statistical significance, as calculated by two-tailed unpaired Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001).