Bivalve-specific gene expansion in the pearl oyster genome: implications of adaptation to a sessile lifestyle

Abstract

Introduction: Bivalve molluscs have flourished in marine environments, and many species constitute important aquatic resources. Recently, whole genome sequences from two bivalves, the pearl oyster, Pinctada fucata, and the Pacific oyster, Crassostrea gigas, have been decoded, making it possible to compare genomic sequences among molluscs, and to explore general and lineage-specific genetic features and trends in bivalves. In order to improve the quality of sequence data for these purposes, we have updated the entire P. fucata genome assembly.

Results: We present a new genome assembly of the pearl oyster, Pinctada fucata (version 2.0). To update the assembly, we conducted additional sequencing, obtaining accumulated sequence data amounting to 193× the P. fucata genome. Sequence redundancy in contigs that was caused by heterozygosity was removed in silico, which significantly improved subsequent scaffolding. Gene model version 2.0 was generated with the aid of manual gene annotations supplied by the P. fucata research community. Comparison of mollusc and other bilaterian genomes shows that gene arrangements of Hox, ParaHox, and Wnt clusters in the P. fucata genome are similar to those of other molluscs. Like the Pacific oyster, P. fucata possesses many genes involved in environmental responses and in immune defense. Phylogenetic analyses of heat shock protein70 and C1q domain-containing protein families indicate that extensive expansion of genes occurred independently in each lineage. Several gene duplication events prior to the split between the pearl oyster and the Pacific oyster are also evident. In addition, a number of tandem duplications of genes that encode shell matrix proteins are also well characterized in the P. fucata genome.

Conclusions: Both the Pinctada and Crassostrea lineages have expanded specific gene families in a lineage-specific manner. Frequent duplication of genes responsible for shell formation in the P. fucata genome explains the diversity of mollusc shell structures. These duplications reveal dynamic genome evolution to forge the complex physiology that enables bivalves to employ a sessile lifestyle in the intertidal zone.

Keywords: Biomineralization; C1q; Genome; Heat shock proteins; Hox; ParaHox; Pearl oyster; Pinctada fucata.

Fig. 1

Lineage-specific genes/gene families are more numerous than conserved gene families among molluscs. a The number of common gene families assigned using OrthoMCL DB. b The number of gene families not assigned using OrthoMCL DB, but detected among mollusc species. c Gene composition of the three mollusc genomes. Bars represent the number of genes that are either members of common gene families (assigned to an OrthoMCL DB gene family; blue), shared families (not assigned using OrthoMCL DB, but shared by at least two mollusc species; green), lineage-specific families (present only in a single mollusc genome; yellow), and orphan genes (with no putative homolog; orange)

Fig. 2

Enriched GO categories may reflect the sedentary lifestyles of bivalves in the intertidal zone. The percentage of all annotated genes is shown. Asterisks indicate GO categories that are significantly (p < 0.01) enriched in bivalve-specific gene models

Fig. 3

Expansion of heat shock protein 70 (HSP70) genes has occurred in bivalve genomes. a Genes that contain conserved Pfam domains related to heat shock chaperones. See also Additional file 1: Table S4. b Unrooted maximum likelihood molecular phylogeny of HSP70 domain sequences of selected animal genomes. The blue area highlights a group of predominantly bivalve genes. Protein sequences of Pinctada fucata and Crassostrea gigas are marked with purple and green lines, respectively. Thick purple and green lines indicate that the clade comprises only one bivalve species, suggesting lineage-specific gene expansion. Yellow lines designate gene pairs of P. fucata and C. gigas supported by high bootstrap values (≥80 %), indicating that the genes probably existed in the common ancestor of the two bivalves. Nine outer red arcs, except the largest one on the left, show groups composed of four or more protostome genes. Nodes of the tree supported with high bootstrap values (≥80 %) are marked with black dots with the number, while larger black dots without a number indicate 100 % bootstrap support. The scale bar represents expected substitutions per site in the aligned regions. The first three letters of the protein ID indicate the species name; pfu: P. fucata, Cgi: C. gigas, Lgi: Lottia gigantea, Hro: Helobdella robusta, Cte: Capitella teleta, and Dme: Drosophila melanogaster. HSP70 domain sequences with lengths of 200 amino acids or more were used for analysis with the WAG + GAMMA + F model

Fig. 4

Tandem duplication and expansion of genes related to innate immune recognition has occurred in all three bivalve lineages. a Genes that contain selected Pfam domain candidates related to recognition of non-self antigens. See also Additional file 1: Table S7. b Unrooted maximum likelihood molecular phylogeny of C1q domain sequences of three mollusc genomes (Pinctada fucata: purple, Crassostrea gigas: green, and Lottia gigantea: black), and MgC1q proteins of Mytilus galloprovincialis (red). Wedges indicate the merged gene group, which is composed of sequences only from bivalve species, and the number of constituent genes shown in the arc. Wedges without numbers indicate groups that include two genes. Merged groups that contain two or more tandemly arranged genes in the genome are marked with asterisks. Branches showing orthologous gene groups of three bivalves are highlighted in yellow. Nodes supported with high bootstrap values (≥80 %) are marked with black dots. Branches with slashes were collapsed in order to fit the tree onto the page. The scale bar represents expected substitutions per site in the aligned regions. C1q domain sequences with lengths of 100 amino acids or more were used for analysis with the WAG + GAMMA + F model

Fig. 5

SMP gene families are clustered in the P. fucata genome. Relative position and orientations of genes are indicated. Lengths of scaffolds, genes, and intergenic regions are not to scale. a Shematrin. b N19. c Nacrein and nacrein-like. d MSI60 and MSI60-related. e Alveolin-like and MP10. f Chiobiase. g Chitinase-like. h EGF-like (i) Tyrosinase. j Fibronectin domain-containing. k Serine protease inhibitor. l Peroxidase-like. See also Additional file 1: Table S5 for detail

Fig. 6

Hox, ParaHox, and Wnt gene clusters in the P. fucata genome resemble those of other protostomes. The relative position and orientation of the genes are indicated. a Hox gene cluster. P. fucata Hox genes are located on 3 scaffolds. b ParaHox gene cluster. P. fucata ParaHox genes are aligned on a single scaffold. The gray box indicates a non-ParaHox gene. c Wnt gene cluster. P. fucata Wnt1, 6, 9, and 10 genes are found on a single scaffold and the gene order is the same as that of L. gigantea. Lengths of scaffolds, genes, and intergenic regions are not to scale. See also Additional file 1: Tables S6-S8 for details