DMRT1 repression using a novel approach to genetic manipulation induces testicular dysgenesis in human fetal gonads

Abstract

Study question: Does loss of DMRT1 in human fetal testis alter testicular development and result in testicular dysgenesis?

Summary answer: DMRT1 repression in human fetal testis alters the expression of key testicular and ovarian determining genes, and leads to focal testicular dysgenesis.

What is known already: Testicular dysgenesis syndrome (TDS) is associated with common testicular disorders in young men, but its etiology is unknown. DMRT1 has been shown to play a role in the regulation of sex differentiation in the vertebrate gonad. Downregulation of DMRT1 in male mice results in trans-differentiation of Sertoli cells into granulosa (FOXL2+) cells resulting in an ovarian gonadal phenotype.

Study design, size, duration: To determine the effect of DMRT1 repression on human fetal testes, we developed a novel system for genetic manipulation, which utilizes a Lentivral delivered miRNA during short-term in vitro culture (2 weeks). A long-term (4-6 weeks) ex vivo xenograft model was used to determine the subsequent effects of DMRT1 repression on testicular development and maintenance. We included first and second-trimester testis tissue (8-20 weeks gestation; n = 12) in the study.

Participants/materials, setting, methods: Human fetal testes were cultured in vitro and exposed to either of two DMRT1 miRNAs (miR536, miR641), or to scrambled control miRNA, for 24 h. This was followed by a further 14 days of culture (n = 3-4), or xenografting (n = 5) into immunocompromised mice for 4-6 weeks. Tissues were analyzed by histology, immunohistochemistry, immunofluorescence and quantitative RT-PCR. Endpoints included histological evaluation of seminiferous cord integrity, mRNA expression of testicular, ovarian and germ cell genes, and assessment of cell number and protein expression for proliferation, apoptosis and pluripotency factors. Statistical analysis was performed using a linear mixed effect model.

Main results and the role of chance: DMRT1 repression (miR536/miR641) resulted in a loss of DMRT1 protein expression in a sub-population of Sertoli cells of first trimester (8-11 weeks gestation) human fetal testis; however, this did not affect the completion of seminiferous cord formation or morphological appearance. In second-trimester testis (12-20 weeks gestation), DMRT1 repression (miR536/miR641) resulted in disruption of seminiferous cords with absence of DMRT1 protein expression in Sertoli (SOX9+) cells. No differences in proliferation (Ki67+) were observed and apoptotic cells (CC3+) were rare. Expression of the Sertoli cell associated gene, SOX8, was significantly reduced (miR536, 34% reduction, P = 0.031; miR641 36% reduction, P = 0.026), whilst SOX9 expression was unaffected. Changes in expression of AMH (miR536, 100% increase, P = 0.033), CYP26B1 (miR641, 38% reduction, P = 0.05) and PTGDS (miR642, 30% reduction, P = 0.0076) were also observed. Amongst granulosa cell associated genes, there was a significant downregulation in R-spondin 1 expression (miR536, 76% reduction, P < 0.0001; miR641, 49% reduction, P = 0.046); however, there were no changes in expression of the granulosa cell marker, FOXL2. Analysis of germ cell associated genes demonstrated a significant increase in the expression of the pluripotency gene OCT4 (miR536, 233%, P < 0.001). We used the xenograft system to investigate the longer-term effects of seminiferous cord disruption via DMRT1 repression. As was evident in vitro for second-trimester samples, DMRT1 repression resulted in focal testicular dysgenesis similar to that described in adults with TDS. These dysgenetic areas were devoid of germ cells, whilst expression of FOXL2 within the dysgenetic areas, indicated trans-differentiation from a male (Sertoli cell) to female (granulosa cell) phenotype.

Limitations, reasons for caution: Human fetal testis tissue is a limited resource; however, we were able to demonstrate significant effects of DMRT1 repression on the expression of germ and somatic cell genes, in addition to the induction of focal testicular dysgenesis, using these limited samples. In vitro culture may not reflect all aspects of human fetal testis development and function; however, the concurrent use of the xenograft model which represents a more physiological system supports the validity of the in vitro findings.

Wider implications of the findings: Our findings have important implications for understanding the role of DMRT1 in human testis development and in the origin of testicular dysgenesis. In addition, we provide validation of a novel system that can be used to determine the effects of repression of genes that have been implicated in gonadal development and associated human reproductive disorders.

Study funding/competing interest(s): This project was funded by a Wellcome Trust Intermediate Clinical Fellowship (Grant No. 098522) awarded to RTM. LBS was supported by MRC Programme Grant MR/N002970/1. RAA was supported by MRC Programme Grant G1100357/1. RMS was supported by MRC Programme Grant G33253. This work was undertaken in the MRC Centre for Reproductive Health which is funded by the MRC Centre grant MR/N022556/1. The funding bodies had no input into the conduct of the research or the production of this manuscript. The authors have declared no conflicts of interest.

Figure 1

DMRT1 knockdown does not affect seminiferous cord formation in the first trimester human fetal testis.(A) H + E staining of pre-culture control (left panel) and after culture for 13 days following an initial 24 h lentiviral exposure to scrambled miRNA (middle panel) or miR536 (right panel). Seminiferous cords are delineated by broken white lines. DMRT1 mRNA expression in (B) miR536- and (C) miR641-exposed tissues compared to scrambled miRNA controls (n = 4 for each DMRT1-miRNA exposure group). (D) Immunofluorescence analysis of SOX9 (green) and DMRT1 (red) expression in Sertoli cells in fetal testis tissue following transduction with either scrambled miRNA (left panel), miR536 (middle panel) or miR641 (right panel) lentiviral constructs. Scale bars: (A) 20 μM and (D) 50 μM. Graphs show mean ± SEM. Data analyzed using a linear mixed effects model in R.

Figure 2

DMRT1 expression is lost in Sertoli cells of the second-trimester human fetal testis after in vitro exposure to miR536 and miR641. H + E staining of pre-culture control (left panel) and after culture for 13 days following an initial 24 h lentiviral exposure to scrambled miRNA (middle panel) or miR641 (right panel). Normal histological appearance of the testicular tissue is demonstrated. DMRT1 mRNA expression in (B) miRNA536- and (C) miRNA641-exposed tissues compared to scrambled miRNA controls (n = 3 for each DMRT1-miRNA treatment group, *P < 0.05). (D) Immunofluorescence analysis of SOX9 (green) and DMRT1 (red) expression in Sertoli cells in fetal testis tissue following transduction with either scrambled miRNA (left panels), miR536 (middle panels) or miR641 (right panels) lentiviral constructs. Scale bar (A) 20 μM and (D) 50 μM. Graphs show mean ± SEM. Data analyzed using a linear mixed effects model in R.

Figure 3

DMRT1 knockdown in second-trimester human fetal testis alters expression of genes involved in testicular and ovarian development. RT-qPCR analysis of genes associated with development of (A) Sertoli (B) granulosa and (C) germ cells in second-trimester human fetal testes tissue culture for 13 days after an initial 24 h exposure to scrambled miRNA, miR536 or miR641. Expression was normalized to RPS29 (n = 2–3 per treatment group, *P < 0.05, **P < 0.01, ***P < 0.001). Graphs show mean ± SEM. Data analyzed using a linear mixed effects model in R.

Figure 4

DMRT1 knockdown in vitro results in disruption of seminiferous cords in second-trimester human fetal testis tisse. Immunofluorescence analysis of the expression of SOX9 (Sertoli cells; green) and DMRT1 (Sertoli and germ cells; red) in fetal second-trimester testis tissue cultured for 13 days following an initial 24 h transduction with either scrambled miRNA, miR536 or miR641 lentiviral constructs. Regions of disruption of seminiferous cords can be identified (broken white line) where DMRT1 expression has been reduced but SOX9 expression has been retained. Scale bar: 100 μM.

Figure 5

DMRT1 knockdown results in ‘focal testicular dysgenesis’ in second-trimester human fetal testis xenografts. Double immunohistochemistry for DMRT1 (brown; Sertoli and germ cells) and SOX9 (blue; Sertoli cells) in (A) pre-graft control, or after xenografting for 6 weeks following an initial 24 h transduction period in culture with (B) scrambled miRNA, (C) miR536 or (D) miR641. Red asterisk = normal seminiferous cord structure, black arrowheads show areas of cord breakdown and the absence of DMRT1 protein; black arrows indicate germ cells and DYS identifies focal dysgenetic areas in which there has been a complete breakdown of seminiferous cords. Scale bar: 50 μM.

Figure 6

Germ and somatic cell composition and function is altered in ‘focal dysgenetic’ regions of second-trimester human fetal testis resulting from DMRT1 knockdown. Double immunofluorescence for (A) Ki67 (green; proliferation marker) and Cleaved Caspase 3 (red; apoptotic marker), (B) DMRT1 (green; Sertoli and germ cells) and CYP11A1 (red; Leydig cells), and (C) AP2γ (green; gonocytes) and MAGE-A4 (red; pre-spermatogonia). Nuclear counterstain (gray; DAPI/Hoescht). Focal dysgenetic areas are indicated by dashed white lines. White asterisk highlights a single AP2γ+ germ cell within a dysgenetic area (bottom right panel), White arrow indicates a surviving MAGE-A4+ germ cell within a dysgenetic area. Scale bar: 50 μM.

Figure 7

DMRT1 knockdown results in FOXL2 expression in a region of ‘focal dysgenesis’ in a second-trimester human fetal testis xenograft. FOXL2 protein expression in xenografts of control and DMRT1-miRNA transduced second-trimester human fetal testis tissue. (A) Double immunohistochemistry for SOX9 (Blue) and DMRT1 (brown) identifies regions of loss of DMRT1 with ‘Focal Dysgenesis’. (B) FOXL2 immunoexpression (green) in scattered cells in one fetus in an area devoid of DMRT1 (red) expression following initial transduction with miR536. Scale bar: 20 μM.