Novel insights of whole exome sequencing in non-obstructive azoospermia patients

Objectives: Male infertility is a multifactorial pathology suggested as the underlying factors in majority of cases of severe male infertility, especially non-obstructive azoospermia. The recent emergence of next-generation sequencing (NGS) offers an opportunity to analyze many genes at once and to develop various bioinformatic approaches, including variant burden investigation. Gene-based burden tests identify a set of rare variants in a given gene by comparing case and control cohorts. Another method – Gene Set Enrichment Analysis (GSEA) – analyzes gene networks, that share common regulation, biological function or chromosomal location. There is still a lack of reports on WES implementation with burden tests, gene network analysis and routine clinical diagnostics in male infertility research. The purpose of this study was to investigate the potential of exome sequencing in idiopathic azoospermia cases.

Methods: Whole exome sequencing was performed on 21 non-obstructive azoospermia patients. A gene set of previously described and novel candidate genes of azoospermia was compiled.

Samples were sequenced using the Twist Comprehensive Exome Panel. The resulting sequences were mapped against the human genome GRCh38 reference sequence using BWAMEM. Copy number variations (CNVs) were annotated using AnnotSV. Samples were analyzed and filtered with the Illumina's BaseSpace Variant Interpreter. Variants considered as pathogenic and likely pathogenic were confirmed by Sanger sequencing. Genetic burden test was performed with TRAPD. P value < 0.05 was considered significant. Protein interactions were investigated with ConsensusPathDB, STRING and CytoScape.

Results: SNV analysis detected two genetic variants of unknown significance (VUS) in genes, affecting the hypothalamic–pituitary–gonadal axis: NR5A1 and FGFR1. Clinical investigation did not demonstrate hypogonadism in the subject group. NM_004959(NR5A1):c.763C>T was interpreted at first as likely pathogenic (LP), according to ACMG guidelines. The patient's phenotype did not reflect the expected phenotype. In CNV investigation, VUS deletion in AD genes TUBG1 and TUBG2 was found (seq[GRCh38] 17q21.2(42613632-42659961)x1). No previously described known pathogenic genetic variants were found.

Genetic variant burden was elevated in 1473 genes. 302 genes with increased loss-of-function (LoF) variant set were present in more than one sample. Variant burden of genes TKFC, DPM1, UBE2J2, MTCH2, GCLC, NPIPB11, OR2T33, POTEG was elevated in > 50% of samples. Over-representation analysis with pathway based set of genes with high variant burden demonstrated 26 pathways, half of the pathways (13) being involved in sperm development, especially sumoylation (4). Over-representation analysis with protein complex-based sets obtained 14 protein sets, all involved in DNA repair and genomic integrity. STRING interaction analysis between genes with high variant burden showed two genome instability networks.

Conclusions: Based on a preselected diagnostic gene panel, we identified four VUS in exomes, involved in hypothalamic–pituitary–gonadal axis. In patients with azoospermia, an increased burden of genetic variants is observed in interrelated genes involved in genome instability and spermatogenesis. These findings can add supporting information to the knowledge base of infertility diagnostics. WES as a routine diagnostics method in azoospermic patients calls for the further investigation.