Association of ESX1 gene variants with non-obstructive azoospermia in Chinese males

Abstract

Genetic factors are one of the most important causes of non-obstructive azoospermia (NOA). ESX1 is an X-linked testis-biased expressed gene, and a potential biomarker for testicular sperm retrieval in NOA patients, yet few systematic studies have investigated its association with NOA. Here, we performed selected exonic sequencing in a large cohort of Chinese males, and four novel missense mutations (including one compound mutation), one novel synonymous mutation of ESX1 unique to NOA patients were identified. We analyzed the effects of ESX1 mutations on cyclin A degradation and cell cycle progression by immunoprecipitation assay and flow cytometry, and found that the compound mutant p.[P365R; L366V] ESX1 compromised the stabilizing effect of ESX1 on polyubiquitinated cyclin A, thereby causing the failure of M phase arrest in cells. Further studies showed that the deleterious effect of the compound mutations on ESX1 protein function was attributed to p.P365R but not p.L366V alteration. The novel ESX1 mutation p.P365R might confer high risk for NOA in Han Chinese population, probably via affecting cell cycle control.

Figure 1

Missense mutations of ESX1 identified in patients with NOA. (a) Four missense mutations were validated by PCR-Sanger sequencing. Mutation sites are marked with a black box. (b) The evolutionary conservation of each amino acid altered by these mutations was analyzed via multiple protein sequence alignments. (c) Schematic representation of ESX1 protein and its missense mutations.

Figure 2

The p.[P365R; L366V] mutation compromised the stabilizing effect of ESX1 on polyubiquitinated cyclin A. Cyclin A and Flag-Ubiquitin were transfected into 293 T (a) or Hela (b) cells together with wild type or mutant ESX1-HA expression vectors, as indicated. Cells transfected with empty vector of pcDNA 3.1-HA and further treated with MG132 were used as positive control. Cell lysates were then prepared and immunoprecipitated with anti-Cyclin A antibodies. Cyclin A polyubiquitination was detected with anti-Flag antibody. The grouping of blots cropped from different parts of the same gel, or from different gels, fields, or exposures were divided by black lines. Compared with wild type ESX1, polyubiquitination of cyclin A was significantly decreased by the p.[P365R; L366V] mutation.

Figure 3

p.[P365R; L366V] ESX1 disrupted the inhibitory effect of ESX1 on cell cycle of 293 T and Hela cells. Wild type or mutant ESX1-HA expression vectors were transfected into cells as indicated. Empty vector was used as negative control (NC). Cells were synchronized at S phase with thymidine blocking for 36–48 h, then washed with PBS and treated with cytosine for another 6 h. (a) Cells were harvested and cell cycle profiles were analyzed by flow cytometry and the FlowJo software. The G0/G1, S, or G2 phase was marked with blue, green and pink colour, respectively. Relative cell ratio of different phases in 293 T (b) and Hela (c) generated by FlowJo from at least three independent experiments were analyzed statistically using GraphPad Prism 5 software. Compared with wild type ESX1, p.[P365R; L366V] ESX1 lost the ability to arrest cells at M phase. * P < 0.05.

Figure 4

The p.P365R but not p.L366V mutation attributes to the deleterious effect on ESX1 function. p.P365R or p.L366V ESX1 overexpression plasmids were transfected into 293 T and Hela cells as indicated. Wild type ESX1 was used as control. After synchronization as described above, cells were either lysed for IP to detect the polyubiquitination of cyclin A (a and b), or subjected to flow cytometry (c). Relative cell ratio of different phases in 293 T (d) and Hela (e) were analyzed statistically using GraphPad Prism 5 software. The grouping of blots cropped from different parts of the same gel, or from different gels, fields, or exposures were divided by black lines. Compared with wild type ESX1, p.P365R but not p.L366V failed to stabilize polyubiquitinated cyclin A, and further lost the ability to arrest cells at M phase. * P < 0.05.

Figure 5

ESX1 mRNA levels in NOA patients harboring ESX1 mutations. RNA was extracted from FFPE testicular samples and then reverse-transcribed into cDNA. Compared with patient with OA, ESX1 mRNA level was substantially decreased in patients harboring the p.R281H (W245) or p.[P365R; L366V] (W075) mutation. ** P < 0.01.

Figure 6

Hematoxylin and eosin staining of testis sections from (a) a patient with obstructive azoospermia (OA), and (b) the NOA patient carrying the ESX1 p.[P365R; L366V] mutation. Only Sertoli cells could be detected within the testis tissue of this NOA patient (Sertoli-cell-only-syndrome, SCO), whereas the tissue of the OA patient displayed full spermatogenesis. Scale bar: 50 μm.