The evolution of genomic and epigenomic features in two Pleurotus fungi

Abstract

Pleurotus tuoliensis (Bailinggu, designated Pt) and P. eryngii var. eryngii (Xingbaogu, designated Pe) are highly valued edible mushrooms. We report de novo assemblies of high-quality genomes for both mushrooms based on PacBio RS II sequencing and annotation of all identified genes. A comparative genomics analysis between Pt and Pe with P. ostreatus as an outgroup taxon revealed extensive genomic divergence between the two mushroom genomes primarily due to the rapid gain of taxon-specific genes and disruption of synteny in either taxon. The re-appraised phylogenetic relationship between Pt and Pe at the genome-wide level validates earlier proposals to designate Pt as an independent species. Variation of the identified wood-decay-related gene content can largely explain the variable adaptation and host specificity of the two mushrooms. On the basis of the two assembled genome sequences, methylomes and the regulatory roles of DNA methylation in gene expression were characterized and compared. The genome, methylome and transcriptome data of these two important mushrooms will provide valuable information for advancing our understanding of the evolution of Pleurotus and related genera and for facilitating genome- and epigenome-based strategies for mushroom breeding.

Figure 1

Genomic features of P. tuoliensis (Pt) and P. eryngii var. eryngii (Pe) tabulated in circos plot. The circles from outside to inside represent length of contigs (colored and black stripes represent contigs derived from Pt and Pe, respectively), gene density, repeat elements content, coverage of RNA-seq reads in mycelium, DNA CG methylation level, siRNA reads coverage, respectively. Each feature was calculated based on 10 kb non-overlapping windows. Synteny gene pairs are linked by colored lines in the inner-most circle.

Figure 2

Phylogenetic relationship of conserved single-copy orthologs identified in fifteen fungal species. The maximum likelihood (ML) phylogenetic tree was constructed from 1385 conserved (existing in all analyzed fungal species) single-copy orthologs (aligned by MAFFT) using FastTree with the model of LG + CAT. In addition, those common conserved (including both conserved single-copy and multiple-copies orthologs), Pleurotus-specific (existing in all three Pleurotus taxa whereas absent in other fungi), taxa-specific (uniquely existing in one specific Pleurotus genome), and other types of orthologs are categorized and tabulated, respectively.

Figure 3

Clustered gene contents of CAZyme and oxidoreductase families in fifteen representative fungi genomes. Hierarchical clustering of wood-decay related genes were completed using the R package pvclust, in which the unbiased (AU) p-values (%) are computed by 1000 bootstrap re-samplings. Respective gene numbers of each gene family in corresponding genome is shown in each cell. Over-represented and under-represented gene families are depicted in red and blue backgrounds, respectively. Within the gene family on the right column, abbreviated names of CAZyme and lignolytic oxidoreductase families are denoted in black and red, respectively. Abbreviations of gene family names: GH, glycoside hydrolases; CE, carbohydrate esterases; POD, class II peroxidases; MCO, multicopper oxidases; CRO, copper-radical oxidases; CDH, cellobiose dehydrogenase; Cytb562, cytochrome b562; OXO, oxalate oxidase/decarboxylases; QRD, quinone reductases; DyP, dye-decolorizing peroxidases; HTP, heme-thiolate peroxidases; P450, cytochromes P450. Abbreviations of species names: Po, Pleurotus ostreatus; Pt, Pleutotus tuoliensis; Pe, Pleurotus eryngii var. eryngii; Fh, Fistulina hepatica; Ab, Agaricus bisporus; Lb, Laccaria bicolor; Cc, Coprinopsis cinerea; Cp, Coniophora puteana; Hp, Hydnomerulius pinastri; Pm, Pisolithus microcarpus; Ds, Dichomitus squalens; Pg, Phlebiopsis gigantean; Cn, Cryptococcus neoformans; Tm, Tremella mesenterica; Nc, Neurospora crassa.

Figure 4

Establishment of DNA methylation profiles in P. tuoliensis (Pt) and Pleutotus eryngii var. eryngii (Pe). (A) Averaged DNA methylation levels within (enclosed between TSS and TTS) and around (flanking +1kb) genic and TE regions in Pt; (B) Density curves of genes and TEs (upper) and distribution of DNA methylation levels (in the contexts of CG, CHG and CHH, bottom) along the contig 1 of Pt genome; (C) Numbers of identified DMRs (hyper- and hypo- methylated regions) within genic (defined as regions of gene body accompanied with flanking +1kb interval), inter-genic, and repeat regions of Pt and Pe; (D) Phylogenetic tree of gene homologs encoding DNA methyltransferases in fifteen fungal species. The Maximum Likelihood (ML) phylogenetic tree was constructed based on the DNA methylase domain (PF00145) of 56 DNA methyltransferases. Sequence alignment, tree construction, and further validation were completed using MAFFT and FastTree (with 1000 bootstrapping replicates), respectively.

Figure 5

Original and distribution of siRNAs and DNA methylation profiling of siRNA clusters in P. tuoliensis (Pt) and Pleutotus eryngii var. eryngii (Pe). (A) Length and genomic region distribution of siRNAs in Pt and Pe are tabulated in bar columns. Genic-regions defined as regions of gene body accompanied with flanking +1kb; (B) The Maximum Likelihood (ML) phylogenetic tree of identified Pleurotus DCL genes was constructed as described in Fig. 4 based on RNA dependent RNA polymerase domain (PF05183) and different colors denotes diverged clades of Pleurotus DCL genes; (C) DNA CG methylation level of siRNA clusters located in genic-, TE-, and intergenic-regions, respectively, in Pt; (D) DNA CG methylation level in siRNA cluster-related and siRNA cluster-non-related genic contexts in Pt.

Figure 6

DNA methylation around TE regions negatively regulate expression of adjacent genes in P. tuoliensis (Pt). (A) Negative impacts of DNA methylation in/around TEs (on the right) on the expression of their neighboring genes (evaluated as FPKM values, on the left) in Pt. “TE_TE”, “non_TE”, “TE_non”, and “non_non” denotes genes flanked with TEs on both upstream and downstream sides, only flanked with downstream TEs, only flanked with upstream TEs, and without any flanking TEs, respectively. Asterisks indicates both levels of gene expression and gene body DNA methylation were significantly different between gene set without any flanking TEs and other gene sets (p < 0.01; Mann-Whitney-Wilcoxon test); (B) Significant negative correlation between gene expression and DNA methylation level within corresponding genic regions. On the x axis, genes are categorized into bins of hierarchical DNA methylation levels. Asterisks indicates significantly negative correlation between DNA methylation level in different contexts and gene expression level (p < 0.01; Pearson’s product-moment correlation); (C) Exemplary overviewed IGV tracks illustrate the relative distributions of DNA methylation level, mRNA reads abundance, siRNA reads abundance along the ~35 kb region in contig 1: 1,825,947–1,860,003 of Pt.