X chromosome-linked CNVs in male infertility: discovery of overall duplication load and recurrent, patient-specific gains with potential clinical relevance

Abstract

Introduction: Spermatogenesis is a highly complex process involving several thousand genes, only a minority of which have been studied in infertile men. In a previous study, we identified a number of Copy Number Variants (CNVs) by high-resolution array-Comparative Genomic Hybridization (a-CGH) analysis of the X chromosome, including 16 patient-specific X chromosome-linked gains. Of these, five gains (DUP1A, DUP5, DUP20, DUP26 and DUP40) were selected for further analysis to evaluate their clinical significance.

Materials and methods: The copy number state of the five selected loci was analyzed by quantitative-PCR on a total of 276 idiopathic infertile patients and 327 controls in a conventional case-control setting (199 subjects belonged to the previous a-CGH study). For one interesting locus (intersecting DUP1A) additional 338 subjects were analyzed.

Results and discussion: All gains were confirmed as patient-specific and the difference in duplication load between patients and controls is significant (p = 1.65 × 10(-4)). Two of the CNVs are private variants, whereas 3 are found recurrently in patients and none of the controls. These CNVs include, or are in close proximity to, genes with testis-specific expression. DUP1A, mapping to the PAR1, is found at the highest frequency (1.4%) that was significantly different from controls (0%) (p = 0.047 after Bonferroni correction). Two mechanisms are proposed by which DUP1A may cause spermatogenic failure: i) by affecting the correct regulation of a gene with potential role in spermatogenesis; ii) by disturbing recombination between PAR1 regions during meiosis. This study allowed the identification of novel spermatogenesis candidate genes linked to the 5 CNVs and the discovery of the first recurrent, X-linked gain with potential clinical relevance.

Figure 1. Position of DUP1A relative to LIN00685 and PPP2R3B on the X chromosome (PAR1).

Diagram of the Xp22.33, showing the presence of all known protein-coding genes (Red) and DUP1A (Blue). Enlarged is a 100 Kb region showing the location of PPP2R3B and LINC00685 in relation to the DUP1A minimum and maximum. This gain will certainly duplicate the antisense element LINC00685, but does not fully duplicate PPP2R3B – skewing the ration between the gene and its negative regulator.

Figure 2. Relative expression of PPP2R3B and LINC00685 in testes biopsies in patients with different phenotypes (Johnsen Score) (Chalmel, et al., 2012).

Taken from EBI Expression Atlas. (http://www.ebi.ac.uk/gxa/home).