A high-density SNP genetic linkage map for the silver-lipped pearl oyster, Pinctada maxima: a valuable resource for gene localisation and marker-assisted selection

Abstract

Background: The silver-lipped pearl oyster, Pinctada maxima, is an important tropical aquaculture species extensively farmed for the highly sought "South Sea" pearls. Traditional breeding programs have been initiated for this species in order to select for improved pearl quality, but many economic traits under selection are complex, polygenic and confounded with environmental factors, limiting the accuracy of selection. The incorporation of a marker-assisted selection (MAS) breeding approach would greatly benefit pearl breeding programs by allowing the direct selection of genes responsible for pearl quality. However, before MAS can be incorporated, substantial genomic resources such as genetic linkage maps need to be generated. The construction of a high-density genetic linkage map for P. maxima is not only essential for unravelling the genomic architecture of complex pearl quality traits, but also provides indispensable information on the genome structure of pearl oysters.

Results: A total of 1,189 informative genome-wide single nucleotide polymorphisms (SNPs) were incorporated into linkage map construction. The final linkage map consisted of 887 SNPs in 14 linkage groups, spans a total genetic distance of 831.7 centimorgans (cM), and covers an estimated 96% of the P. maxima genome. Assessment of sex-specific recombination across all linkage groups revealed limited overall heterochiasmy between the sexes (i.e. 1.15:1 F/M map length ratio). However, there were pronounced localised differences throughout the linkage groups, whereby male recombination was suppressed near the centromeres compared to female recombination, but inflated towards telomeric regions. Mean values of LD for adjacent SNP pairs suggest that a higher density of markers will be required for powerful genome-wide association studies. Finally, numerous nacre biomineralization genes were localised providing novel positional information for these genes.

Conclusions: This high-density SNP genetic map is the first comprehensive linkage map for any pearl oyster species. It provides an essential genomic tool facilitating studies investigating the genomic architecture of complex trait variation and identifying quantitative trait loci for economically important traits useful in genetic selection programs within the P. maxima pearling industry. Furthermore, this map provides a foundation for further research aiming to improve our understanding of the dynamic process of biomineralization, and pearl oyster evolution and synteny.

Figure 1

Schematic representation of reference mapping families. Ovals represent females, squares represent males and diamonds represent families consisting of N offspring of unknown sex. Pink lines show the maternal contribution to the subsequent generation and blue lines show the paternal contribution. The population of origin for F₀ oysters is indicated by the letter in the sample ID: B for Bali, A for Aru and W for West Papua. The two unknown sires with no genotypes, U01 and U02, are indicated in red text.

Figure 2

The sex-average maps for linkage groups 1–7. SNP IDs in bold indicate framework SNPs placed at a LOD > 3 and remaining SNPs have been placed in their most likely position at a LOD < 3. SNPs located within known biomineralization genes are indicated in bold italics.

Figure 3

The sex-average maps for linkage groups 8–14. SNP IDs in bold indicate framework SNPs placed at a LOD > 3 and remaining SNPs have been placed in their most likely position at a LOD < 3. SNPs located within known biomineralization genes are indicated in bold italics.

Figure 4

Frequency of the sex-average inter-marker distances (cM) across the fourteen P. maxima linkage groups. Only intervals > 0 cM were included. Over 49% of all intervals are below 1 cM, demonstrating an even spread of markers throughout the genome.

Figure 5

The cumulative Kosambi cM for the sex-average, female and male maps. The extent and patterns of localised regional sex-specific recombination rates are illustrated for each linkage group. The overall female-to-male ratio (R) for each linkage group is also reported.

Figure 6

Comparison of standardised female and male interval distances of LG1 and LG2 revealing highly variable sex-specific recombination along both linkage groups. Regression analysis was performed by visually determining breakpoints (dashed lines) and grouping data into three slopes, left, middle and right. The male map is compressed near the centromeres and expanded near the telomeres, and the opposite was observed for the female map. The average slope of the lines in the two middle sections (centromeric) is 0.07 (±0.02) and is significantly different from 1 (P < 0.05). The average male-to-female recombination ratio for the slopes near the centromere is 1:5.98, indicating a male "cold-spot" for recombination. The average slope of the lines near the telomeres are 4.29 (±0.56) for the left group and 5.20 (±3.06) for the right, and again are significantly different from 1 (P < 0.05).

Figure 7

A plot of the female vs male inter-marker distances (cM) for all pairs of adjacent markers. The dashed line represents a 1:1 sex ratio whereby recombination is the same in both sexes. The majority of the points fall close to either 0 on the x-axis, or 0 on the y-axis indicating both strong female biased and strong male biased recombination throughout all intervals.

Figure 8

Mean linkage disequilibrium (LD) estimates at different linkage map distances throughout the P. maxima genome for r ² and D’.