Large-scale analyses of the X chromosome in 2,354 infertile men discover recurrently affected genes associated with spermatogenic failure

Abstract

Although the evolutionary history of the X chromosome indicates its specialization in male fitness, its role in spermatogenesis has largely been unexplored. Currently only three X chromosome genes are considered of moderate-definitive diagnostic value. We aimed to provide a comprehensive analysis of all X chromosome-linked protein-coding genes in 2,354 azoospermic/cryptozoospermic men from four independent cohorts. Genomic data were analyzed and compared with data in normozoospermic control individuals and gnomAD. While updating the clinical significance of known genes, we propose 21 recurrently mutated genes strongly associated with and 34 moderately associated with azoospermia/cryptozoospermia not previously linked to male infertility (novel). The most frequently affected prioritized gene, RBBP7, was found mutated in ten men across all cohorts, and our functional studies in Drosophila support its role in germ stem cell maintenance. Collectively, our study represents a significant step towards the definition of the missing genetic etiology in idiopathic severe spermatogenic failure and significantly reduces the knowledge gap of X-linked genetic causes of azoospermia/cryptozoospermia contributing to the development of future diagnostic gene panels.

Keywords: RBBP7; X chromosome; azoospermia; cryptozoospermia; genes; genetics; genomics; male infertility; next-generation sequencing; spermatogenesis.

Figure 1

Flowchart representing the classification of genes mutated in more than one man Genes have been classified in two main categories on the basis of whether they have been previously associated with human male infertility. Genes not previously associated with male infertility were further prioritized and classified according to the potential relationship with the human phenotype. (A) Filtered variants as described in the material and methods; (B) from the 100 genes, 55 were further selected on the basis of expression profile and published data

Figure 2

Variants in genes not previously linked to male infertility, ranked highest on the basis of testis expression, MGI data, and literature (A) Variants in genes for which various model organisms support a role in spermatogenesis; (B) model organisms are currently not available or are newly acquired human genes. B, Barcelona cohort; B, Barcelona Replication cohort; M, MERGE cohort; M, MERGE replication cohort; G, GEMINI cohort; N, Nijmegen/Newcastle cohort; Azoo, azoospermic; Crypto, cryptozoospermic; SCO, Sertoli cell-only; iSCO, incomplete SCO; SGA, spermatogonial arrest; iSGA, incomplete SGA; iSCA, incomplete spermatocytic arrest; MeiA, meiotic arrest; MatA, maturation arrest; Hypo, hypospermatogenesis; NA, not available; Mmu, Mus muscuslus; Dme, Drosophila melanogaster; Cel, Caenorhabditis elegans; LoF, loss of function; Chr, chromosome; Chrt, chromatin.

Figure 3

Drosophila Caf1-p55 depletion affects GSC maintenance and differentiation (A) DAPI staining of testes from control nos-Gal4/UAS-GFP and knockdown nos-Gal4/UAS-GFP; UAS-Caf1-p55-RNAi. GFP was visualized by direct fluorescence (in green). Scale bars correspond to 100 μm. (B) Percentage of testes showing an underdeveloped phenotype in control nos-Gal4/UAS-GFP (n = 42) and knockdown nos-Gal4/UAS-GFP; UAS-Caf1-p55-RNAi (n = 36) males. (C) Immunostaining of testes from control white (left panels) and knockdown nos-Gal4/UAS-GFP; UAS-Caf1-p55-RNAi males (right panels) with αdBigH1 antibodies (in red), Vasa (in green) antibodies. DNA was stained with DAPI. White dotted circles indicate the undeveloped testes. Scale bars correspond to 100 μm. (D) Photomicrograph of the tip regions of the control nos-Gal4/UAS-GFP testis and the knockdown nos-Gal4/UAS-GFP; UAS-Caf1-p55-RNAi testis in (C). Note the complete lack of early germ cells (dBigH1 and Vasa positive cells inside the white dotted circle) in the nos-Gal4/UAS-GFP; UAS-Caf1-p55-RNAi testis. Scale bars correspond to 25 μm.

Figure 4

Protein-protein interactions predicted for the 41 genes with a prioritization score ≥ 5 Significantly more interactions were observed in these proteins than expected for a similarly sized dataset of randomly selected proteins (PPI enrichment p value < 7.44e−14)