Diversity and regulatory impact of copy number variation in the primate Macaca fascicularis

Abstract

Background: Copy number variations (CNVs) are a significant source of genetic diversity and commonly found in mammalian genomes. We have generated a genome-wide CNV map for Cynomolgus monkeys (Macaca fascicularis). This crab-eating macaque is the closest animal model to humans that is used in biomedical research.

Results: We show that Cynomolgus monkey CNVs are in general much smaller in size than gene loci and are specific to the population of origin. Genome-wide expression data from five vitally important organs demonstrates that CNVs in close proximity to transcription start sites associate strongly with expression changes. Among these eQTL genes we find an overrepresentation of genes involved in metabolism, receptor activity, and transcription.

Conclusion: These results provide evidence that CNVs shape tissue transcriptomes in monkey populations, potentially offering an adaptive advantage. We suggest that this genetic diversity should be taken into account when using Cynomolgus macaques as models.

Keywords: CNV; Cynomolgus monkey; Gene expression; Olfactory receptors; eQTL.

Fig. 1

CNV genotypes. a Geographic origin of four natural populations, from where the tested Cynomolgus monkeys were caught. b Number of duplications and deletions detected per individual by combining three CNV calling approaches. c Size distribution of the inferred CNV regions across all 21 individuals (n = 15,183). d Number of CNV regions detected among and across the four populations. e Hierarchical clustering of the log₂ − ratio genotypes of all CNV regions (n = 15,183) in 21 individuals. f Loadings of the first and second principal component based on a PCA performed on the log₂- ratio genotypes of all CNV regions with deletions (n = 8495) in 21 individuals

Fig. 2

eQTLs. a Number of detected cis-eQTL per tissue under 10% false discovery rate (FDR). b Average expression levels of eQTL genes in each tissue versus the average expression level of all genes in the respective tissue. c Nominal p-value of all detected cis-eQTL as a function of the distance to the transcription start site (TSS) of the eQTL CNV to its associated gene. d Density of detected cis-eQTLs as a function of the distance to the transcription start site (TSS) of the eQTL CNV to its associated gene

Fig. 3

eQTL loci. a region on chromosome 7 containing olfactory receptor cis-eQTLs in both kidney (orange) and lung (blue). CNV loci are color coded according to the copy number signal (copy number gain in blue, copy number loss in red). The red box highlights the CNV locus, which shows duplication events associated with gene expression changes of proximal olfactory receptor genes. Triangles indicate an association reported from the genome-wide cis-eQTL mapping, while stars indicate additional associations revealed by the eQTL region analysis. b eQTL association plots for olfactory receptor cis-eQTLs on chromosome 7 in both kidney (orange) and lung (blue). CNV genotype refers to the quantitative genotype defined as the median log₂-ratio of CGH probes contained within the CNV

Fig. 4

eQTL loci. a Region on chromosome 3 containing the ABCB4 cis-eQTL in lung (blue triangle). CNV loci are color coded according to the copy number signal (copy number gain in blue, copy number loss in red). The red box highlights the CNV locus, which shows deletion events associated with gene expression changes of the ABCB4 gene. b eQTL association plots of ABCB4 expression levels with the deletion CNV found in individuals from the Philippines. CNV genotype refers to the quantitative genotype defined as the median log₂-ratio of CGH probes contained within the CNV