RANKL regulates male reproductive function

Abstract

Infertile men have few treatment options. Here, we demonstrate that the transmembrane receptor activator of NF-kB ligand (RANKL) signaling system is active in mouse and human testis. RANKL is highly expressed in Sertoli cells and signals through RANK, expressed in most germ cells, whereas the RANKL-inhibitor osteoprotegerin (OPG) is expressed in germ and peritubular cells. OPG treatment increases wild-type mouse sperm counts, and mice with global or Sertoli-specific genetic suppression of Rankl have increased male fertility and sperm counts. Moreover, RANKL levels in seminal fluid are high and distinguishes normal from infertile men with higher specificity than total sperm count. In infertile men, one dose of Denosumab decreases RANKL seminal fluid concentration and increases serum Inhibin-B and anti-Müllerian-hormone levels, but semen quality only in a subgroup. This translational study suggests that RANKL is a regulator of male reproductive function, however, predictive biomarkers for treatment-outcome requires further investigation in placebo-controlled studies.

Conflict of interest statement

All authors declare no competing interests except for M.B.J. who holds two patents on the use of RANKL inhibitors to treat male infertility. R.B. has been on the advisory Board of Amgen and B.L. is now employed by Radius Health. All other authors state no conflicts of interest.

Figures

Fig. 1. Gonadal RANKL signaling and influence of OPG in mice.

a Triple immunofluorescence with RANKL (sc-9073, red), SOX9 (Sertoli cell marker, green), and VASA (germ cell marker, blue). b Double immunofluorescence with RANK (HPA0277728, green), VASA (red), and DAPI (blue), or c OPG (sc-21038, blue), SOX9 (green), and DAPI (red). d Triple immunofluorescence with RANKL (sc-9073, red), RANK (HPA0277728, green), OPG (sc-21038, blue), and DAPI (gray) (all IF in 16-week-old WT mice). White insert marks high magnification area. e, f Phenotypic effects of OPG (1 mg/kg, red) or vehicle (blue) injected twice weekly for 2 or 5 weeks on testicular and epididymal weight and total epididymal sperm count in 10-week-old mice, (e, left panel; p = 0.03, right panel; p = 0.04). Note, y-axes do not intersect x-axes at 0 (e, f). Scale bars correspond to 25 μm (a–d). Data are presented individually and as mean ± SEM, with n (vehicle/OPG) = 15/10 (left and middle panels, e) n (vehicle/OPG) = 9/10 (right panel, e), and n (vehicle/OPG) = 5/5 (all panels, f). Statistical test: two-sided Student’s t test (e, left and right panels) with *p < 0.05.

Fig. 2. Phenotypic effects of systemic versus Sertoli cell-specific suppression of RANKL.

a Rosa reporter expression in Rankl fl/fl, Sertoli specific (AMHCre;Ranklfl/fl, SC-Rankl), and global (VasaCre;Ranklfl/fl, Rankl −/−) RANKL-deficient mice. b, c Expression of RANKL at mRNA and protein level in Rankl fl/fl (blue), SC-Rankl (red), and Rankl −/− (green) mice (b, fl/fl vs. SC: p = 0.923, fl/fl vs. −/−: p = 0.004). d Testicular and epididymal weight of Rankl fl/fl, SC-Rankl, and Rankl −/− mice, (left panel; p = 0.003, p = 0.159, right panel; p = 0.008, p = 0.033). e Sperm count and ABC sperm motility of Rankl fl/fl, SC-Rankl, and Rankl −/− mice, (left panel; p = 0.013, p = 0.107). f Morphology and expression of Sertoli cell marker GATA-4 in testis of Rankl fl/fl, SC-Rankl, and Rankl −/− mice. Counterstaining with Mayer’s hematoxylin. Scale bars correspond to 50 μm. g Gene expression of Inhibin B, Amh, Rank, and Opg in Rankl fl/fl, SC-Rankl, and Rankl −/− mice, (top right panel; p = 0.01, p = 0.02, bottom right panel; p = 0.07, p = 0.04). h Fertility study of Rankl fl/fl, SC-Rankl, and Rankl −/− male mice exposed to three healthy females for 5 days, (left panel; p = 0.02, middle panel; p = 0.01). i Serum testosterone (nmol/L) in Rankl fl/fl, SC-Rankl, and Rankl −/− mice. j Fertility study of Rankl fl/fl, SC-Rankl, and Rankl −/− mice after OPG treatment (1 mg/kg x 2 weekly for 2 weeks), (left panel; p = 0.02, middle panel; p = 0.02, p = 0.01, right panel; p = 0.005). The OPG-treated Rankl −/− mice did not give birth to pups. Note, y-axes do not intersect x-axes at 0 (d). Data are presented individually and as mean ± SEM, with n (Rankl fl/fl/SC-Rankl/Rankl −/−) = 12/8/12 (b), n (= 14/6/12 (d), n = 6/9/9 (left panel, e), n = 6/5/6 (right panel, e), n = 4–6 (f), n = 12/8/8 (top left panel, g), n = 7/6/8 (top right, bottom left and right panels, g), n (male/female) = 6/18 (Rankl fl/fl) and n = 4/12 (SC-Rankl) and n = 4/12 (Rankl −/−) for all panels (h), n = 8 (i), n (male/female) = 3/9 (Rankl fl/fl + OPG) and n = 4/12 (SC-Rankl + OPG) and n = 2/6 (Rankl −/− + OPG) (j). Statistical tests: one-way analysis of variance (ANOVA) with Dunnett’s test to adjust for multiple comparisons (b, d, e, g, h (middle panel), j (middle and right panels)) or Pearson chi square test (h (left panel), j (left panel)) with *p < 0.05, **p < 0.01.

Fig. 3. Expression of RANKL, RANK, and OPG in human testis and effects of RANKL inhibition.

a Gene expression of RANKL, RANK, and OPG in normal human testis (blue) and Sertoli cell only (red), (p = 0.002, p = 0.003, p = 0.002). b Triple immunofluorescence with RANKL (sc-7628, red), SOX9 (Sertoli cell marker, green), and MAGE-A4 (germ cell marker, blue) in normal testis. c Triple immunofluorescence RANK (HPA0277728, blue), SOX9 (green), and MAGE-A4 (red) in normal testis. d Triple immunofluorescence with OPG (sc-21038, green), SOX9 (blue), and MAGE-A4 (red) in normal testis. e Triple immunofluorescence with RANKL (sc-7628, red), RANK (HPA0277728, blue), OPG (sc-21038, geen), and DAPI (gray) in normal testis. f Apoptotic germ cells in ex vivo cultures treated with vehicle, Denosumab (100 ng/ml), or OPG (50 ng/ml) for 48 h investigated by IHC staining for cleaved PARP (cPARP, marker of apoptosis). Counterstaining with Mayer’s hematoxylin. g Number of cPARP positive cells per area after 48 h of treatment with vehicle (blue), Denosumab (red), or OPG (green), (p = 0.002, p = 0.041). h Paired measurement of sRANKL in the media of ex vivo testis cultures (indicated by different colors) following 48 h of vehicle, Denosumab, or OPG treatment, (p = 0.027, p = 0.047). Scale bars correspond to 25 µm (b–e) or 50 μm (f). Data presented individually and as mean ± SEM, with n (normal testis/Sertoli cell only) = 5/3 (a), n (vehicle/Dmab/OPG) = 8/8/8 (g), and n (vehicle/Dmab/OPG) = 4/4/4 (h). Statistical test: unpaired two-sided Student’s t test (a) or paired two-sided Student’s t test (g, h) with *p < 0.05, **p < 0.01. Abbreviations: Dmab Denosumab.

Fig. 4. Expression of RANKL, RANK, and OPG in male reproductive tract and effects of denosumab on human spermatozoa.

a Expression of RANKL (sc-9073), RANK (sc-9073), and OPG (sc-8468) in human caput, corpus, and cauda epididymis. Counterstaining with Mayer’s hematoxylin. Scale bars correspond to 125 µm for low magnification (upper panels) and 50 µm for high magnification (lower panels). Black inserts mark high magnification areas. b Expression of RANKL (sc-9073), RANK (sc-9073), and OPG (sc-8468) in human prostate gland and seminal vesicle. Counterstaining with Mayer’s hematoxylin. Scale bars correspond to 100 µm. c Immunostaining of RANKL (sc-9073) in human spermatozoa. Counterstaining with Mayer’s hematoxylin. Scale bar corresponds to 10 µm. d Intracellular Ca2+ signals induced by Denosumab in concentrations from 200 µg/mL to 0.195 µg/mL (indicated by different colors). ΔF/F0 (%) indicates the percentage change in fluorescence that translates into cytoplasmic calcium changes. Progesterone (100 nM) was used afterwards to test for an augmented effect, while human tubular fluid (HTF) was used as a negative control (repeated twice in three donors). e Acrosome reacted spermatozoa (% of number of spermatozoa) from healthy donors after treatment with Denosumab (6 µg/mL, red), the positive control ionomycin (10 µM, green), or vehicle treatment (HTF with 0.2% DMSO, blue), (p < 0.0001, p < 0.0001). Data presented as mean (D) or mean ± SEM (e) with n = 3 (d, e). Statistical test: one-way analysis of variance (ANOVA) with Dunnett’s test to adjust for multiple comparisons (e) with ***p < 0.001. Abbreviations: Dmab Denosumab, Prog. Progesterone.

Fig. 5. Serum and seminal fluid concentrations of sRANKL in relation to semen quality.

a (left) Serum sRANKL (pmol/L), (middle) seminal sRANKL (pmol/L), and (right) seminal/serum sRANKL ratio in healthy (blue) and infertile men (red), (left panel; p < 0.0001, middle; p < 0.0001, right; p < 0.0001). b Longitudinal measurements of seminal sRANKL levels in 31 infertile men with an average of 14 days between sample collections. c Semen quality variables and seminal sRANKL levels in a pooled linear regression model of both healthy and infertile men. d Seminal sRANKL levels and semen quality stratified in groups according to WHO references for normal (blue) vs. low (red) semen quality, (left panel; p = 0.096, left middle; p = 0.041, right middle; p = 0.0009, right; p = 0.0034). Pooled analyses of all healthy and infertile men. e Seminal/serum sRANKL ratios and semen quality stratified in groups according to WHO classification of normal vs. low semen quality, (left panel; p < 0.0001, left middle; p = 0.005, right middle; p < 0.0001, right; p < 0.0001). Pooled analyses of all healthy and infertile men. f Receiver operating characteristic (ROC) curve analysis and table showing sensitivity and 1-specificity for semen quality variables, seminal sRANKL, and sRANKL seminal/serum ratio as well as area under the curve (AUC) with 95% confidence intervals (CI). g Sex steroid levels in serum and seminal sRANKL levels in a linear regression model of both healthy and infertile men. All beta and p-values are adjusted for duration of abstinence. Statistical tests: two-sided Student’s t test (a, d, e) and linear regression model (c, g) with *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 6. Reproductive hormones and semen quality before and after treatment with Denosumab (60 mg) in 10 infertile men.

a Serum AMH (pmol/L), Inhibin B (pg/ml), and Inhibin B/FSH ratio, (left panel; p = 0.004, middle; p = 0.001, right; p = 0.03). b Total sperm count (106). c Sperm concentration (106/ml) and total number of progressive motile sperm (106). d Individual changes in serum inhibin B at day 80, (p = 0.001). e Individual changes in total progressive motile sperm at day 80. f Baseline serum OPG and treatment response evaluated by percental change in serum Inhibin B at day 80. g Baseline serum OPG and treatment response evaluated by percental change in total progressive motile sperm at day 80 normalized to baseline, (right panel; p = 0.015). h Baseline serum Inhibin B and treatment response evaluated by relative change in total progressive motile sperm at day 80 normalized to baseline, (right panel; p = 0.042). i Baseline serum AMH and treatment response evaluated by relative change in total progressive motile sperm at day 80 normalized to baseline. j DNA fragmentation at baseline, day 20, and 120. k Serum sRANKL and OPG normalized to baseline values, (left to right; p < 0.0005, p < 0.0005, p < 0.0005, p < 0.0005, p = 0.002). l Seminal fluid levels of sRANKL (pmol/L) at day 80 to 120 compared with baseline, (p = 0.002). Dotted line indicates upper reference limit for the ELISA. All hormones and semen quality variables are presented after normalization to mean of the two samples delivered prior to treatment start except in Fig. B. Note, y-axes and x-axes do not intersect at 0 (a,b, d-f, k). Data presented as mean ± SEM (b), individually (d-l), or combined mean ± SEM (blue) and individually (dotted) (a, c, g, h, i, j, k) with n = 10 (a-k), n = 8 (l) Statistical test: paired two-sided Student’s t test (a, d, k, l) and unpaired two-sided Student’s t test (g, h). *p < 0.05, **p < 0.01, ***p < 0.001.