COPY NUMBER LOSSES IN SPECIFIC REGIONS OF COMMON VARIATION ARE SIGNIFICANTLY MORE PREVALENT IN INFERTILITY PATIENTS COMPARED WITH A POPULATION OF DEMONSTRATED FERTILITY AND LOW FETAL WASTAGE

Abstract

Chromosomal microarray analysis (CMA) is a valuable standard of care for assessment of individuals with developmental disabilities and other pathological states. However, interpretation of incident copy number variants (CNV) is highly dependent upon use of prevalence data from “normal” populations. Infertility which is present in up to 15% of any population is known to be genetically heterogeneous, however such patients would typically appear in CNV databases in the “normal” population and relevant findings may go unsuspected. We performed bioinformatic analysis of the variation observed in 539 consecutive women receiving services at a major US Infertility practice for the presence of potentially relevant CNV relating to their underlying condition or unsuspected genetic risks. We compared these to a population of 79 women with exceptional fertility (grand multiparas) as demonstrated by 5 or more uncomplicated term deliveries with no clinically apparent reproductive losses. CNV were assessed with an FDA approved CMA (Affymetrix, CytoscanDX). CNV association analysis of the entire databases of CNV segments in the “infertile” and “grand multipara” cohorts was undertaken with ParseCNV program (http://parsecnv.sourceforge.net/). A typically wide range of CNV and runs of homozygosity (ROH) was observed, with many overlapping known common regions of variation and the majority predicted to be likely benign. Potential impactful rare gene findings have been reported elsewhere, including two well described recurrent pathogenic CNV observed in 3 patients each, and an appreciable number of deletions encompassing all or part of genes for known over 40 autosomal recessive conditions suggesting CNV analysis would add positively to expanded carrier screening programs. A striking presence of genomic deletions in three regions in the “infertile” cohort (6-18%), none of which were observed in any of the “grand multiparas” (p value range: 0.02-≤0.000005) We used one-sided Fisher's exact test to test for enrichment of either deletion or duplication CNVs between the “infertile” case subjects and the comparison individuals. These differential CNVs all occurred in regions of common well-known CNV of gene families on chromosomes (chr) 7, 14, and 11 (typically 12-40% prevalence range for regions; Database of Genomic Variants [DGV]; http://dgv.tcag.ca/). There was no difference in the relative occurrence of any duplications or ROH. Deletions significantly differential in prevalence between these groups were observed in several regions of very common known CNV, which may have plausible link to key theories on reproductive success and evolution. The variants observed on the first two regions (chr7 & 14) involved variant transcripts of the T-cell receptor-g and -a respectively. On chromosome 11 the deletions involved the 52N1 and 52N5 families of olfactory receptor genes. Both the immune response and olfactory perceptions have been suggested to be involved with the evolution of eutherian mammalian reproduction and fertility in humans.