European flat oyster (Ostrea edulis) is an ecologically and economically important marine bivalve, that has been severely affected by the intracellular parasite Bonamia ostreae. In this study, a flat oyster SNP array (~14,000 SNPs) was used to validate previously reported outlier loci for divergent selection associated with B. ostreae exposure in the Northeast Atlantic Area. A total of 134 wild and hatchery individuals from the North Sea, collected in naïve (NV) and long-term affected (LTA) areas, were analysed. Genetic diversity and differentiation were related to the sampling origin (wild vs hatchery) when using neutral markers, and to bonamiosis status (NV vs LTA) when using outlier loci for divergent selection. Two genetic clusters appeared intermingled in all sampling locations when using outlier loci and their frequency was associated with their bonamiosis status. When both clusters were compared, outlier datasets showed high genetic divergence (FST > 0.25) unlike neutral loci (FST not ≠ 0). Moreover, the cluster associated with LTA samples showed much higher genetic diversity and significant heterozygote excess with outlier loci, but not with neutral data. Most outliers mapped on chromosome 8 (OE-C8) of the flat oyster genome, supporting a main genomic region underlying resilience to bonamiosis. Furthermore, differentially expressed genes previously reported between NV and LTA strains showed higher mapping density on OE-C8. A range of relevant immune functions were specifically enriched among genes annotated on OE-C8, providing hypotheses for resilience mechanisms to an intracellular parasite. The results suggest that marker-assisted selection could be applied to breed resilient strains of O. edulis to bonamiosis, if lower parasite load and/or higher viability of the LTA genetic cluster following B. ostreae infection is demonstrated.

Sampling

The analysis was performed on 134 flat oyster individuals collected in 2019 from locations with different bonamiosis prevalence: i) three oyster beds from the North Sea (OS, GBR, NO); ii) two hatchery batches (OSH, WZH) from the same area (FAO subarea 27.4 https://www.fao.org/fishery/en/area/27/en; Fig. 1) (Table 1). Hatchery spat was produced from wild individuals of the same area that were conditioned in the hatchery until spawning. Hatchery samples were included in the study considering their interest in restocking bonamiosis-affected areas by using Bonamia-resilient strains from marker-assisted selection programs. These samples provided additional information for the association analysis and validation of markers, and specifically in one case, the Wadden Sea, the only representative sample was from the hatchery. Furthermore, hatchery data allowed an estimation of the contribution of parents for checking the reproduction protocols in batches at hatcheries to maintain genetic diversity in future restoration programs. The sampling locations were sorted according to bonamiosis status from previous information as: i) naïve (NV), supposedly to have never been in contact with the parasite; and ii) long-term affected (LTA), where parasite presence was first reported over 30 years ago and monitored over time to the present (Engelsma et al. 2010).

DNA extraction and SNP genotyping

Genomic DNA was extracted from gill tissue using the E.Z.N.A.® Mollusc DNA kit (OMEGA Bio – Tek) following the manufacturer’s recommendations. Briefly, between 30 – 50 mg of fresh gill was treated with a solution containing 20 µL of proteinase K and 700 µL of lysis buffer (CSPL) overnight while gently shaking to mix thoroughly at 65 ºC. Samples were subjected to RNAse A treatment and to several buffer washes by centrifugation. DNA quantity and purity were measured with a Nanodrop spectrophotometer and DNA concentrations were normalized for genotyping.

Samples were shipped to Identigen (Dublin, Ireland) for genotyping on a medium-density Affymetrix Axiom SNP array (14,950 SNPs; 1 SNP per ~ 63 kb of the flat oyster genome; 935.6 Mb; see below). After filtering SNPs with < 90% call rate, 11,641 were retained, which supports the robustness of the SNP chip platform (Gutierrez et al., 2017). A total of 611 SNPs were monomorphic across all sampling locations and discarded for structure analyses, since they would not influence the results. Three different SNP datasets were considered according to the analyses performed (see Results section) and the information previously reported by Vera et al. (2019): i) the whole dataset, ii) the two outlier datasets related to bonamiosis resistance according to more or less stringent statistical criteria (consistent and suggestive, respectively), and iii) the neutral dataset.

Files provided can be opened with Notepad. Therefore, no specific software is required.