Denosumab stimulates spermatogenesis in infertile men with preserved Sertoli cell capacity

Abstract

Sperm production depends on proper Sertoli-germ cell interaction, and we hypothesized that receptor activator of nuclear factor κB ligand (RANKL) activity in Sertoli cells may influence spermatogenesis. Treatment with the RANKL inhibitor denosumab, normally used to treat osteoporosis, increased testicular weight, inhibin B, and germ cell proliferation in ex vivo testis cultures and in vivo in a humanized RANKL mouse. The effect on germ cell proliferation was positively associated with baseline serum concentrations of anti-müllerian hormone (AMH). In accordance, denosumab increased germ cell proliferation in ex vivo human testis cultures with low/moderate but not severe impairment of Sertoli cell function. In a placebo-controlled randomized clinical trial, denosumab had no effect on semen quality but increased sperm concentration in a subgroup of infertile men with serum AMH ≥38 pmol/L at baseline. In conclusion, high serum AMH may increase the probability of a beneficial response to denosumab treatment in infertile men, thus suggesting a possible venue for precision medicine in male infertility.

Keywords: AMH; RANKL activity; bone; denosumab; male infertility; personalized medicine; randomized clinical trial; spermatogenesis.

Graphical abstract

Figure 1

Denosumab increases testis weight, sperm count, and serum inhibin B in vivo and in an ex vivo testis culture model from humanized RANKL mice (A) Body and organ weights (g) and epididymal sperm count (10⁶) in humanized RANKL mice 6 weeks after subcutaneous (s.c.) injection with denosumab (5.0 mg/kg) or vehicle. (B) Serum inhibin B (1, pg/mL), AMH (2, ng/mL), testosterone (3, nmol/L), FSH (4, ng/mL), and LH (5, ng/mL) in humanized RANKL mice 6 weeks after s.c. injection with denosumab (5.0 mg/kg) or vehicle. (C) Serum inhibin B/FSH ratio (left) and serum testosterone/LH ratio (right) in humanized RANKL mice 6 weeks after s.c. injection with denosumab (5.0 mg/kg) or vehicle. (D) Schematic representation of the adult mouse testis tissue culture model. Testes from 8-week-old humanized RANKL mice were dissected into 1.5 mm³ pieces, placed on agarose gel cubes, and treated with FSH (100 ng/mL), OPG (100 ng/mL), or denosumab (100 ng/mL) via the culture media. (E) H&E staining showing representative images of testis tissue from humanized RANKL mice after 48 h of ex vivo culture (n = 13). Black insert indicates high-magnification area. Scale bars correspond to 250 (left) and 50 μm (right). (F) Representative images of immunohistochemical staining of proliferating (BrdU⁺, left) and apoptotic (cPARP⁺, right) germ cells in humanized RANKL mouse testis after 48 h of ex vivo culture (n = 13). Counterstaining with Mayer’s hematoxylin. Scale bars correspond to 50 μm. (G) Quantification of proliferating (BrdU⁺/mm², left) and apoptotic (cPARP⁺/mm², right) cells in testis cultures from humanized RANKL mice treated with FSH (100 ng/mL), OPG (100 ng/mL), or denosumab (100 ng/mL) for 48 h relative to vehicle treatment. (H) Correlation between serum AMH (ng/mL) and proliferative response (quantification of proliferating germ cells [BrdU⁺/mm²]) to FSH (100 ng/mL, left) or denosumab (100 ng/mL, right) normalized to vehicle treatment in humanized RANKL ex vivo mouse cultures. Data presented individually and as mean ± SEM with n = 9/10 (vehicle/Dmab) (A, B, and C) and with n = 13/10/12/11 (vehicle/FSH/OPG/Dmab) for BrdU and cPARP (G), and n = 13/10/11 (vehicle/FSH/Dmab) for AMH (H). Statistical tests: independent t test (A, B, and C) and mixed-effects analysis with Dunnett’s test to adjust for multiple comparisons (G) and simple linear regression analysis (H). Data were log transformed (B). Abbreviations: G, grams; SEM, standard error of the mean; s.c., subcutaneous. See also Figures S1–S3, and S8.

Figure 2

OPG is expressed in Sertoli cells from adult human testis with Sertoli cell-only but not in testis with full spermatogenesis or fetal Sertoli cells (A) Triple immunofluorescence showing representative images of the expression pattern of RANKL (sc-7628, red), RANK (HPA027728, blue), OPG (sc-21038, green), and DAPI (D3571, gray) in human testis with full spermatogenesis (n = 3). Scale bar corresponds to 25 μm. (B) Immunohistochemistry showing representative images of the expression pattern of RANKL (sc-7628), RANK (HPA027728), and OPG (sc-21038) and Sertoli cell markers GATA4 (nucleus) (sc-1237), SOX9 (nucleus) (AB5535), Vimentin (cytoplasm) (sc-373717), and MIC2 (membrane) (M3601) in human testis with full spermatogenesis and samples with Sertoli cell-only (n = 3). Counterstaining with Mayer’s hematoxylin. Scale bars correspond to 50 μm. (C) Immunohistochemistry showing representative images of the expression pattern of RANKL (sc-7628), RANK (HPA027728), and OPG (sc-21038) in human fetal testis from gestational week 9–11 (n = 8, age of sample shown are GW 11), including markers SOX9 (AB5535) and AMH (Sertoli cells) (sc-6886), OCT4 (gonocytes) (sc-5279), CYP11A1 (fetal Leydig cells) (HPA016436), and COUP-TFII (interstitial cells) (PP-H7147-00). Black inserts indicate high-magnification areas. Counterstaining with Mayer’s hematoxylin. Scale bars correspond to 25 μm. See also Figure S4 and Table S1 and S2.

Figure 3

The effect of denosumab in human testis cultures depends on OPG expression and Sertoli cell function (A) Flowchart of selection of patient material for the study of denosumab treatment in human testis cultured ex vivo. All patient materials were initially stained with D2-40 (marker of germ cell neoplasia in situ), and patients with D2-40-positive cells in the tissue were excluded. Five patients were excluded due to poor technical quality, morphology, SCO in all tubules, severe dysgenesis, or no possibility of comparison between treatment and vehicle. (B) Representative images of negative immunohistochemical staining with D2-40 in human testis tissue cultures with full spermatogenesis (left), spermatogenic arrest (middle), and SCO (right) (n = 11). Counterstaining with Mayer’s hematoxylin. Scale bars correspond to 50 μm. (C) Flowchart showing selected patients included in quantification of apoptosis (cPARP) or proliferation (BrdU). All seminiferous tubule in cultured tissue fragments from the 15 included patients was examined, and each stratified based on amount of OPG-positive staining (<20% positive staining, 20%–50% positive staining, and <50% positive staining) in Sertoli cell cytoplasm. (D) Stratification of human testis culture based on low (<20%), medium (20–50%), or intense OPG staining in Sertoli cell cytoplasm with representative images shown (n = 15). Specimens were immunohistochemically stained for OPG (sc-21038). Counterstaining with Mayer’s hematoxylin. Scale bars correspond to 50 μm. (E) Immunohistochemical staining showing representative images of proliferating (BrdU⁺) and apoptotic (cPARP⁺) germ cells (n = 11). Counterstaining with Mayer’s hematoxylin. Scale bars correspond to 50 μm. (F) Quantification of proliferating (BrdU⁺/mm², left) and apoptotic (cPARP⁺/mm², right) cells in human testis cultures treated with denosumab (100 ng/mL) for 48 h relative to vehicle control treatment (each patient is its own control) and stratified based on amount of OPG staining in Sertoli cells (<20%, 20%–50%, or >50%). Data presented individually and as mean ± SEM (F) with n = 9/10/8 (<20%/20%–50%/<50%) for BrdU in vehicle and Dmab and n = 9/11/11 (<20%/20%–50%/>50%) for cPARP in vehicle and Dmab. Scale bars correspond to 50 μm. Statistical tests: paired (one-sided) sample t test (F). Abbreviations: BrdU, bromodeoxyuridine; cPARP, cleaved PARP; OPG, osteoprotegerin; SC, Sertoli cell; SCO, Sertoli cell-only; SEM, standard error of the mean. See also Figure S9 and Table S10.

Figure 4

Changes in seminal sRANKL and OPG and serum levels of reproductive hormones from baseline to day 80 (A) Sperm concentration, total sperm count, total motile sperm, seminal fluid sRANKL and OPG at baseline and day 80. Blue line with p values indicate significant changes from baseline to day 80 in the denosumab group. (B) Reproductive hormones in serum at baseline and day 80. Blue lines with p values indicate significant changes from baseline to day 80 in the denosumab group, while red lines with p values indicate significant changes from baseline to day 80 in the placebo group. Vertical black line indicates significant difference between the treatment groups. Data are presented as changes for each individual from baseline to day 80 (blue dotted lines, denosumab; red dotted lines, placebo). The bold dots and lines in (A) represent the median and in (B) the mean for each group. Statistical tests: changes within each group from baseline to day 80 in sperm parameters and seminal fluid sRANKL and OPG were evaluated by Wilcoxon test, and differences between groups were evaluated by Mann-Whitney U test. Changes within each group from baseline to day 80 in reproductive hormones were evaluated by paired sample t test, and differences between groups were evaluated by independent t test. Abbreviations: AMH, anti-müllerian hormone; Dmab, denosumab; FSH, follicle-stimulating hormone; LH, luteinizing hormone; sRANKL, soluble RANKL. See also Tables S4 and S7.

Figure 5

Changes in semen variables and serum levels of reproductive hormones from baseline to day 80 in two subgroups of infertile men (A) Upper: sperm parameters at baseline and day 80 in men with serum AMH ≥38 pmol/L. Lower: changes in sperm parameters at day 80 normalized to baseline for each individual. (B) Reproductive hormones in serum at baseline and day 80 in men with serum AMH ≥38 pmol/L. (C) Upper: sperm parameters at baseline and day 80 in men with serum AMH ≥38 pmol/L and serum sRANKL ≥0.24 pmol/L. Lower: changes in sperm parameters at day 80 normalized to baseline for each individual. (D) Reproductive hormones in serum at baseline and day 80 in men with serum AMH ≥38 pmol/L and serum sRANKL ≥0.24 pmol/L. Data are presented as changes for each individual from baseline to day 80 (B and D) (blue dotted lines, denosumab; red dotted lines, placebo). The bold dots and lines represent the mean for each group (B and D). Blue lines with p values indicate significant changes from baseline to day 80 in the denosumab group, while vertical black line indicates significant difference between the treatment groups (B and D). In (A) and (C) upper panels, data are presented as back-transformed mean with 95% confidence interval. The PRR was defined as an increase at day 80 compared to baseline. Statistical tests: changes within each group from baseline to day 80 in semen parameters and reproductive hormones were evaluated by paired sample t test, and differences between groups were evaluated by analysis of covariance (A and C, upper) and chi-squared test (A and C, lower) for sperm parameters and an independent t test for reproductive hormones (B and D). Abbreviations: AMH, anti-müllerian hormone; Dmab, denosumab; FSH, follicle-stimulating hormone; LH, luteinizing hormone; PRR, positive response rate. See also Figure S6 and S7 and Table S8.