Genome-Wide Association Study And Genomic Selection Of Growth Traits And Inner Shell Color Traits In Hyriopsis Cumingii

The triangle sail mussel（Hyriopsis cumingii,Lea 1852）is the most important pearl mussel in China,and it produced more than 70%of the world’s total pearl output.However,the improvement of pearl production and quality of freshwater pearl culture industry,characterized by H.cumingii,has been in an unbalanced stage.In order to further improve the ability of H.cumingii to produce high-quality pearls,genome-wide association study and genomic selection were carried out on growth traits and inner shell color traits that significantly correlated with pearl quality;Besides,a 30 K c GPS SNP array was developed.As expected,this study will provide important technical support for genetic improvement of H.cumingii.The main results are as follows:1.Genome-wide association study reveals the genetic architecture of growth and inner shell color traits in triangle sail mussel（Hyriopsis cumingii）Five hundred mussels were genotyped by GBS sequencing.After strict quality control,a dataset of 463 mussels and 80562 SNPs were retained.Pearson correlation analysis showed that the correlations between growth traits and inner shell color traits were weak（near to zero）,indicating that it was difficult to achieve co-breeding effects of growth and inner shell color traits.Based on these SNPs,the population genetic structure and the decay rate of linkage disequilibrium（LD）within 463 samples were analyzed,a slight population stratification was detected,the genetic relationship between the samples was low（-0.1-0.4）,and the LD decay rate was rapid,these results demonstrated that the genetic background between the 463 mussels was complex.GEMMA’s mixed linear model was used to perform the genome-wide association study（GWAS）of growth and inner shell color traits,the five growth traits were typical quantitative traits and 12 significant SNPs were identified,with SNP Hcu＿1736＿570064 was identified in shell length,shell height,shell weight and body weight,LG19＿50918535 was identified in both shell length and body weight traits.Also,a total of 29 SNPs were identified for four inner shell color traits,and a*showed an obvious aggregation peak in chromosome 17,indicating that this region may be closely related to the color formation of inner shell.Furthermore,five of the 12 SNPs associated with growth traits proved to be reliable growth-associated SNPs in a new validate population.Cyclin,SNF,PMP22,CPED1,F-box and APC/C genes were found to be significantly associated with growth traits,which may influence the growth and development of H.cumingii by regulating cell mitosis,mantle epithelial tissue activity,and Ga²⁺absorption and transport capacity.For inner shell color traits,three key genes,HSP78,MAS and PKS maybe the key to color formation of inner shell.These results provide some new insights into the genetic structure and physiological regulation mechanism of growth and inner shell color traits in H.cumingii.However,detailed functional identification and regulation mechanism analysis merits more investigation.2.Genomic selection for improvement of growth traits in triangle sail mussel（Hyriopsis cumingii）764 mussels（496 for reference population,268 for selection candidates）were sequenced by GBS technique and a total of 97090 SNPs were identified.Based on genome-wide SNPs,the heritability was estimated to be moderate to high for the following five growth traits:shell length,shell height,shell width,shell weight,and body weight（0.36±0.11 to 0.51±0.10）.Then,the effects of five GS models（i.e.,GBLUP,Bayes A,Bayes B,Bayes Cπ,and Bayes LASSO）and different SNP numbers（selected using three SNP screening strategies）on the predictive ability of the genomic estimated breeding values（GEBVs）were compared.The GBLUP model containing10000 SNPs selected by GWAS strategy was established and was used to calculate GEBVs in selection candidates.The individuals with high GEBVs were then selected as broodstocks for the fast-growth strain（FS）,and the rest for control strains（CS）.At last,encouraging results were detected that the FS performed significantly better than CS in each growth traits.When the mussels were five months old,four growth traits increased by 5.74–28.77%and 11.93%–54.58%at the WY and CM sites,respectively;When the mussels were fifteen months old,five growth traits increased by 6.91%–26.10%and 9.02%–29.32%at the WY and CM sites,respectively.These results may assist in the future large-scale commercial application of GS in H.cumingii and provide a cost-efficient GS strategy for other aquaculture species as well.3.Genomic selection improves inner shell purpleness in triangle sail mussel Hyriopsis cumingiiA genotype dataset of 764 mussels and 97090 high-quality SNPs was constructed and used for GS analysis.In reference population,the four inner shell color phenotypes,lightness（L*）,redness（a*）,yellowness（b*）,and color difference（d E*）were 59.97±7.27,2.66±2.57,1.79±4.77,and 40.65±7.46,respectively,and the genomic heritability of four traits were moderately high（0.39–0.67）.Five GS models（GBLUP,Bayes A,Bayes B,Bayes Cπ,and Bayes LASSO）were applied to assess the prediction accuracy,where Bayes B performed the best and it was chosen to develop a cost-effective GS method using low-density SNPs.Lastly,one GS method that consisted of Bayes B model and 5000 GWAS-informative SNPs was successfully established and the genomic estimated breeding value（GEBV）of each candidate individual was predicted.The mussels with the highest GEBVs（n=30）were selected as broodstocks for breeding the purple strain（PS）,while the rest for control strain（CS）.Significant differences in inner shell color a*and d E*were detected between PS and CS,as well as the proportion of deep purple individuals rose from 9.57%and 12.38%to 51.16%and 26.92%in Wuyi（WY）and Chongming（CM）sites,respectively.These results demonstrated that GS was an efficient approach for enhancing the purpleness of the inner shell in H.cumingii and paved the way for future large-scale commercial applications of GS on shell color.4.Development of the 30 K c GPS SNP array and preliminary application in genetic diversity and population structure for three population of Hyriopsis cumingiiBased on the deep resequencing population sample data,a 30 K c GPS SNP array was designed and constructed for H.cumingii,which was composed of 30621 high-quality SNPs evenly distributed across the genome.Evaluation results showed that the detection rate of the SNP array in random samples was up to 99.60%,the consistency with resequencing data was up to 97.55%,the technical repeatability was up to 98.76%,and the overall error was less than 2.5%.The designed 30 K c GPS SNP array was preliminary applied in genotyping three population mussels,results showed that the heterozygosity of all SNPs were above 20%,and the detection rate was as high as99.31%,indicating that the 30 K SNP array had high stability and accuracy.Based on25485 SNPs obtained after quality control,the population structure characteristics,genetic diversity parameters and genetic differentiation（Fst）between the three populations were evaluated,and selection signatures were further screened.Principal component analysis（PCA）results clearly distinguished the wild population from the two elite populations,but the two elite populations could not be distinguished from each other.In addition,genetic diversity of the two elite populations was almost the same as that of the wild population after multiple generations of artificial selection,even two elite populations performed better in Ho and PIC.The Fst results showed that there was a moderate degree of genetic differentiation between the two elite populations and the wild population（0.0526-0.0764）,but no differentiation within the two elite populations（0.0270）.Further,the SNPs with top 1%of Fst values among the three populations were extracted and mapped,however,no obvious cluster peak phenomenon appeared on the chromosomes of the genome.In order to further identified the causal SNPs that influence the inner shell color formation,the allele frequency of the two elite populations（purple inner shell color,white inner shell color）were compared.All SNPs had consistent main alleles,but the allele frequency varied greatly,186 SNPs were identified with absolute subtract value of allele frequency>0.2,and they were primarily mapped on chromosomes 2,3,8,17 and 19.Also,chromosome 17 was identified in another GWAS test about purple and white.These results implied that chromosome 17is probably crucial for the development of the inner shell color formation.