Digging Of Whole Genome Copy Number Variation And Its Genetic Effects In Cattle And Goat

Genetic variation on the genome is an important basis for studying differences in phenotypic traits in animals.As one of the most important genetic variations,Copy Number Variation(CNV)can affect various complex traits and diseases,and its research on livestock has become a hot spot.At present,the methods of whole genome-wide CNV detection mainly include SNP chip,aCGH(array comparative genomic hybridization)and high-throughput sequencing.However,systematic studies on genome-wide CNV detection,diversity among different populations,and association analysis with phenotypic traits are still limited in cattle and goat.In this study,we firstly characterized the distribution for whole genome CNV detection in 47 US Holsteins using cattle CGH chip and 1023 goats from the worldwide goat populations using goat SNP chip,respectively.After that,we conducted the genome-wide association analysis(GWAS)between the identified CNVs and multiple complex traits in US Holstein bulls and African meat goats,respectively.Finally,three CNVs,which were overlapped with functional genes associated with meat and dairy production traits,were selected to explore for the genetic effects of CNVs on gene expression and/or phenotypes in Chinese native cattle and goats,respectively.The main results of this study are shown as follows:1.Detection of copy number variations in US Holstein bullsWith the CGH data,1758 CNV regions(CNVRs)were detected in 47 US Holstein bulls.After merging CNVRs,a total of 1043 CNVRs were identified,covering a total of 46,802,944 bp.It accounts for about 2.06% of the Bos Taurus genome sequence.Among them,702 CNVRs were copy number loss type(Loss),270 CNVRs were copy number gain type(Gain),and 71 CNVRs were mixed type of copy number gain and loss(Both).The length of Loss was about 2.5 times and 3 times that of Gain and Both,respectively.Among 29 autosomes,the number of CNVs on chromosomes 7 and 12 was the highest,while the average CNV length on chromosomes 17 and 27 were significantly longer than other chromosomes.Comparison of results for CNV detections in US Holsteins based on SNP and NGS technologies,22.13%(216/976)of CNVRs in this study were found to overlap with previous studies,accounting for 15,866,448 bp.Overlapping analysis of CNVRs with bovine quantitative trait loci(QTL)revealed that 91 CNVRs were overlapped with known QTL loci.Gene Ontology(GO)annotation for CNVRs observed a total of 761 functional genes,which were mainly enriched in biological processes related to chromosome assembly,olfactory stimulation,stress,and perception.2.Genome-wide association analysis of copy number variation and multiple complex traits in US Holstein bullsWe then performed GWAS of 297 CNVRs and 41 complex traits in US Holsteins using CNVtools.The results showed that 87 CNVRs were significantly associated with complex traits.Within them,39 CNVRs were simultaneously associated with at least 2 complex traits.Notably,24 CNVRs were related to daughter pregnancy rate(DPR).All 41 traits were divided into four categories and significant associated CNVRs were observed as following: 1)35 CNVRs were associated with eight reproduction traits;2)21 CNVRs were associated with 7 health traits;3)25 CNVRs were significantly associated with 6 production traits;4)39 CNVRs were markedly associated with 15 body conformation traits.Among them,35 CNVRs were overlapped with the coding regions and / or flanking sequences of 47 protein-coding genes.The expression profiles of CNVR-overlapping functional genes in 91 bovine tissues or cell types were analyzed,suggesting that the phenotype associated CNVRs may affect important traits by influencing on the CNV-overlapping genes.For example,the significant effects of CNVR661,CNVR213,and CNVR758 on phenotypes may be related to CYP4A11,CTR9,and CAPZA3,respectively.3.Detection and diversity of copy number variation in the worldwide goat populationsUsing the Caprine SNP50 genotyping data generated by the ADAPTmap Project,we investigated the diversity of CNV distribution in 1023 individuals from 50 goat breeds,which identified 6286 putative CNVs by PennCNV software.These CNVs were merged into 978 CNVRs,spanning ~262 Mb of total length and corresponding to ~8.96% of the goat genome.To construct a comparative CNV map,we then divided the samples into six subgroups,including Western Asia(WAS),Eastern Mediterranean(EME),Alpine & Northern Europe(ANE),Madagascar(MAD),Northwestern Africa(NWA),and Southeastern Africa(SEA)groups.The CNV frequency was the highest in NWA and the lowest in ANE.Among the 29 autosomes,the frequencies of CNV on chromosomes 1,5,6,12,14,and 17 were high,and the frequencies on chromosomes 21,22,and 28 were low.There were large differences for CNV distribution among the six subgroups on chromosomes 5,6,14,17,and 20.Based on the variance for each CNVR among six subgroups,obvious differences were observed for some CNVRs,such as CNVR1,CNVR4,CNVR8,CNVR9 on chromosomes 17,13,3,14,respectively.Clustering analysis of six subgroups was performed based on the top 5% CNVRs with CNV frequencies of each group.The results showed that SEA and MAD were clustered together within the three subgroups of African goats.Among the two subgroups in Europe,ANE was grouped together with WAS and then gather with EME.Based on GO annotations,there were 526 CNVRs with at least 2 CNV events in the population,which were overlapped with 1437 genes.Among them,154 genes were mapped to CNVRs with large differences across six subgroups(Variance > 0.01).GO analysis showed that these CNV-overlapping genes were mainly enriched in binding,metabolic processes and cell membrane composition.We found several important CNV-overlapping genes(e.g.EDNRA,ADAMTS20,ASIP,KDM5 B,ADAM8,SHH,DGAT1),which were involved in local adaptations such as coat color,muscle development,metabolic processes and milk productions.4.Genome-wide association analysis between copy number variation and growth traits in African meat goatsUsing Golden Helix SNP & Variation Suite(SVS)software,we detected common CNVs in 126 samples from four African indigenous goat breeds.After filtering,a total of 30 CNVs ranging from 30,237 bp to 4,910,757 bp were identified.These CNVs were then used for association analysis with six growth traits by a linear regression analysis.The results showed that CNV4 was significantly associated with chest width while CNV27 was markedly associated with both chest width and width of pin bones.GO annotations for these two genes showed that CNV4 and CNV27 were overlapped with ADCY1 and SNX29 gene,respectively.It has been widely demonstrated that SNX29 SNPs were markedly associated with meat production traits in livestock.Also,CNV27 was located in the regions involving in the difference for two predict transcripts of the SNX29 gene.Thus,we speculated that the CNV27 may affect the SNX29 gene functions,resulting in the dosage effects on growth and development of meat goats.5.Analysis for genetics effects of economic traits-related CNVs in Chinese cattle and goatBased on the comparison analysis results for CNV detection,33 genes were observed to overlapped with CNVs both in cattle and goats.To explore for the genetics effects on gene expressions and phenotypes,three CNVs,that were overlapped with functional genes related to meat or milk production traits(SHH,MAPK10,and DGAT1),were selected for subsequent analysis.Of them,bovine SHH CNV was located in the CDS,bovine MAPK10 CNV was located in the intron,and goat DGAT1 CNV was observed to overlapped with the whole DGAT1 gene.For CNV distributions,bovine SHH CNV and MAPK10 CNV and goat DGAT1 CNV revealed significant difference in Chinese cattle and goats,respectively.The association analysis for CNV and phenotypes showed significant associations: 1)SHH CNV was associated with the growth traits of three yellow cattle breeds.The chest depth of Qinchuan cattle,and the weight,body length and chest depth of the 18-month-old Nanyang cattle were significantly better in the individuals with SHH CNV gain type.The normal type was the advantageous type for chest circumference and body weight in Jinnan cattle.2)MAPK10 CNV was significantly associated with the birth weight,18-month body height,30-month body weight and average daily gain of Nanyang cattle.Individuals with normal type performed significantly better than those with other two copy number types(P < 0.05 or 0.01).2)DGAT1 CNV was found to associate with the hip bone index in Guanzhong dairy goats,where the hip index with one copy number was higher than 0.In Saanen dairy goats,DGAT1 CNV was associated with FPD and SNF in late lactation,and the loss type was the advantageous type.Therefore,bovine SHH CNV and MAPK10 CNV and goat DGAT1 CNV can be respectively used as molecular markers for improving beef production and milk production.The analysis for the dosage effects on gene expressions indicated that: 1)SHH CNV was negatively correlated with the expression of SHH gene in adipose tissue of adult Qinchuan cattle;2)In fetal stage,the MAPK10 CNV was negatively correlated with gene expressions of MAPK10 in muscle and brain and MYOG in muscle.In adult adipose tissue,we observed a weak negative correlation between MAPK10 and FOXC2,which plays a role in inhibiting fat differentiation.However,MAPK10 was positively correlated with FABP4,which could promote fat differentiation.These results suggest that the increased copy number of SHH CNV and MAPK10 CNV can inhibit the expression of functional genes in muscle or adipose tissue,resulting in dosage effects on the cattle growth traits.