Kinetics of the inhibition and recovery of spermatogenesis induced by treatment with WIN 18,446, a male contraceptive, in mice

Abstract

Background:

Retinoic acid is essential for spermatogenesis. Genetic deletion of the retinoic acid synthesizing enzymes, Aldh1a1/1a2, in the testes causes infertility in mice. An inhibitor of the ALDH1A1/1A2 enzymes, WIN 18,446 reversibly inhibit spermatogenesis and is a promising approach to male contraception. Previously we reported that a four-week treatment of WIN 18,446 inhibits spermatogenesis and 9-week recovery from treatment normalized fertility of treated mice. However, the precise kinetics of this process has not been studied.

Objectives:

To extend our knowledge of kinetics of ALDH1A inhibition, we studied the changes in the seminiferous epithelium and retinoic acid synthesis capacity of the testes during four weeks of WIN 18,446 treatment and during nine weeks of recovery.

Materials and methods:

Male mice were fed a diet containing WIN 18,446 for 4 weeks followed by a normal diet for up to 8 weeks. Frequently, during the treatment and recovery period, five mice were euthanized and testes were analyzed for testicular histology and retinoic acid synthesis capacity and compared with controls.

Results:

Testes weights progressively decreased and the seminiferous epithelium deteriorated over time with WIN 18,446 treatment and returned to normal after 8 weeks. Retinoic acid synthesis capacity was significantly inhibited 3 days after the WIN 18,446 treatment and recovered after 7 days of no treatment.

After 4 weeks of treatment, complete blockage of spermatogenesis with only spermatogonia and Sertoli cells was observed. The RA biosynthetic capacity of the testes was significantly reduced before the disruption of spermatogenesis was observed and recovered prior to the reinitiation of spermatogonial differentiation.

Conclusions:

Effects of ALDH1A inhibition on spermatogenesis are reversible. Our observation that strong inhibition of ALDH1A disrupts the seminiferous epithelium prior to the completion of a full cycle of spermatogenesis suggests that episodic inhibition of ALDH1A may function as a male contraceptive.

Introduction

Retinoic acid (RA), a metabolite of vitamin A, is critical for spermatogenesis in mammals. It stimulates the A_undiff to A₁ transition of spermatogonia and is also likely involved in spermiogenesis and spermiation^1,2. Cells in the testes express all three major enzymes involved in RA biosynthesis, ALDH1A1, 1A2 and 1A3³. The roles of these enzymes in spermatogenesis have been studied extensively in mice using both genetic⁴ and pharmacological^5–8 approaches, which suggested that ALDH1A1 and 1A2 but not 1A3 are critical in spermatogenesis. Although RA’s role in male reproduction is undisputed, mechanisms of RA action and cells that produce RA during spermatogenesis are not fully elucidated^1,2.

Our group has been examining feasibility of developing male contraceptives that target the enzymes that are responsible for RA biosynthesis, ALDH1As. WIN 18,446 is a strong and irreversible inhibitor of ALDH1A1, 1A2 and 1A3, and oral administration of this compound abrogates spermatogenesis in many mammalian species, including humans^8–11. Importantly, cessation of the WIN 18,446 treatment fully restores spermatogenesis and fertility⁷, providing evidence that inhibition of ALDH1A may be a feasible method of reversible contraception for men. WIN 18,446 is not being developed as a male contraceptive due to side effects, including hepatic lipidosis and disulfiram reactions with alcohol. However, it is an excellent prototype compound that allows for examination of the roles of ALDH1A enzymes in spermatogenesis in adults. In particular, the kinetics of suppression of spermatogenesis and how that relates to inhibition of RA biosynthetic capacity via ALDH1A activity have not been ascertained. Thus, we used a well-established mouse model to determine the changes in testicular cells and RA biosynthetic capacity during WIN 18,446 treatment and after the cessation of the treatment to extend our knowledge of the pharmacodynamics of this class of compounds, which may have utility as male contraceptives.

Materials and Methods

Ethical statement

All procedures in animal studies were approved by the UW Institutional Animal Care and Use Committee (Protocol no 4038–01).

Study design

Breeding age male mice (C57BL/6 background, n=60) were fed AIN93M diet containing WIN 18,446 (2mg/g diet)⁸ for up to 4 weeks. Five of the sixty mice per time point were euthanized after 3, 7, 14, 21 and 28 of treatment. After euthanasia, testes were collected and weighed. One testis was snap-frozen on dry ice, and the other testis was bisected, half was fixed in modified Davidson’s fixative and the other half snap-frozen on dry ice. The remaining 35 mice were switched to AIN93M (control) diet (without WIN 18,446) for up to 8 weeks to determine the kinetics of recovery from WIN 18,446 treatment. Of these 35 mice, five mice per time point were euthanized at days 3, 7, 14, 21, 28, 42 and 56 after returning to the control diet. No treatment controls (n=15) were fed AIN93M diet for the entire study period and their testes were examined at weeks 4, 8 and 12 after diet initiation (Supplementary Fig 1). Mice were between 8–25 weeks old at the start of the study and between 15–25 weeks old at the end of the study.

Testicular Histology

Testes specimens were fixed in modified Davidson’s fixative for 24 – 48 hrs and transferred to 70% ethanol prior to being embedded in paraffin blocks, and tissue sectioned and stained with Periodic Acid-Schiff /hematoxylin (PAS-h). Stained testes sections were reviewed by a board-certified veterinary pathologist (JS) who scored the amount of ongoing spermatogenesis using a modified Johnsen scoring system^12,13 (Supplementary table 1). Additionally, sections of testes from day 56 mice were stained with von Kossa stain to evaluate for mineralization.

Retinoic Acid Biosynthetic Capacity.

The RA biosynthetic capacity of the testes was determined as previously reported¹⁴ with some modifications. Briefly, one testis from each mouse was homogenized in a 25 mM Tris buffer (pH 7.4, Sigma) containing 0.25M sucrose (Sigma) and 1mM DTT(Sigma) using a bead beater. The homogenate was centrifuged at 10,000 g for 10 minutes (allegra 25R, Beckman Coulter) and the supernatant was collected. Protein concentrations were determined with a Bradford assay (Biorad, cat # 500–0006). RA biosynthetic capacity (ALDH1A enzyme activity) was determined with 15 μg protein in an enzyme activity buffer (10 mM HEPES, 150 mM KCl, 2 mM EDTA, pH 7.5; all chemicals from Sigma) with 1 μM retinal (Sigma) and 2 mM NAD (Sigma) after 15 min incubation at 37 °C. Enzyme assays were quenched with acetonitrile: methanol (1:1, v/v, both from Supelco) containing an internal standard, all-transRA-d5 (Toronto Research Chemicals). Samples were then centrifuged at 3,500 g for 10 min, and the RA in the supernatant was measured using UPLC-MS (Agilent 1290 UPLC and AB Sciex 5500 qTrap Q-LIT mass spectrometer) as previously described¹⁵.

Statistical analysis

Statistical analyses were carried out using Prism (v. 10, Graph Pad, La Jolla CA). Comparison among multiple groups were carried out using one-way ANOVA followed by a Dunnett’s post-hoc testing for between group comparisons. P-values less than 0.05 were considered significant.

Results

Testes weights decreased as WIN 18,446 treatment duration increased (Fig.1). Testes weights were significantly reduced compared with control mice after 14 days of WIN 18,446 treatment and remained smaller than the untreated controls during the entire drug treatment and recovery (cessation) period.

Figure 1.

WIN18,446 treatment reduces testicular weights of mice and cessation of the treatment increases testicular weights. Male mice were orally treated with WIN 18,446 (provided in diet) for up to 28 days and then fed a control diet for up to 56 days. At predetermined times, a group of mice were euthanized and testes were collected and weighed to determine effects of ALDH1A inhibition on spermatogenesis. Control group was fed a control diet without WIN 18,446 during the entire study period. One way ANOVA was performed followed by multiple comparisons using Dunnett test. All groups were compared to the control group. *, p<0.0001

Testicular histology demonstrated marked disruption of the seminiferous epithelium with continued WIN 18,446 treatment (Fig. 2). Some tubular degeneration and vacuolation, as well as occasional retained spermatids, was observed as early as 7 days after the onset of treatment with WIN 18,446. After 14 days of WIN 18,446 treatment, more pronounced disruption of the seminiferous epithelium with germ cell degeneration and disorganization, apoptotic spermatids (presumed), and multinucleated giant cells (symplastic spermatids) were noted. After 21 days of the treatment, most testicular tubules showed severe epithelial disruption with loss of germ cells and frequent multinucleated (symplastic) cells, and after 28 days, the majority of tubules contained only Sertoli cells and spermatogonia, similar to previous studies^8,10,16. On the recovery side, after 3 days of the drug cessation, most tubules still contained only Sertoli cells and spermatogonia. At day 7 following drug cessation, more spermatogonia were seen and some tubules showed rare spermatocytes. At recovery day 14, variable numbers of spermatocytes were observed in most tubules and PAS positive material was present in some tubules from testes at recovery day 14 and later. By day 21, many spermatocytes were seen in most tubules, with round or elongating spermatids present in a few tubules. Round spermatids were most often the most differentiated germ cell present in tubules by 28-day recovery, as previously reported⁸. More tubules with elongating spermatids and some with spermatozoa, although substantially reduced in number, were present by 42-day recovery. By 56-day recovery, many tubules had normal spermatogenesis, but approximately 10% of tubules were atrophic or had severe disruption of the seminiferous epithelium. These included tubules with PAS positive material (approximately 1–2 per testes); tubules lined only with Sertoli cells and spermatogonia; and tubules with vacuolation of the seminiferous epithelium. Some tubules had partial depletion or disorganization of germ cells, and rarely multinucleated cells (symplastic spermatids). There was no mineralization noted on von Kossa stain of the 56-day testes.

Figure 2.

WIN 18,446 treatment causes deterioration of testicular cells and cessation of the treatment restores spermatogenesis. Testis of mice during the treatment of WIN 18,446 and recovery from WIN 18,446 treatment were evaluated by a modified Johnsen score after paraffin embedded tissue sections were stained with PAS-H. (A) Testes were evaluated using a modified Johnsen score. One way ANOVA was performed followed by multiple comparisons using Dunnett test. All groups were compared to the control group. **, p<0.0001; *, p=0.0001 (B) Representative testicular tissue sections at different time points. Arrow denotes WIN 18,446 treatment, thicker portion being longer dosing. At day 14 of treatment, there is pronounced disruption of the seminiferous epithelium with germ cell degeneration and disorganization (arrow; inset shows higher magnification with asterisk noting area of cell loss and eosinophilic material). This worsens at day 21 of treatment, and frequent multinucleated cells are seen (arrow; inset shows higher magnification). By day 21 of treatment, and at days 3 and 7 of recovery, the majority of tubules contain only Sertoli cells and spermatogonia (arrow; inset at day 3 of recovery shows higher magnification). Increasing numbers of spermatocytes are observed at days 14 and 21 of recovery, and PAS positive material in some tubules is present (black asterisk). Occasional tubules with PAS positive material were seen at later timepoint as well, although not shown here. At day 28 of recovery, round spermatids are frequently the most differentiated germ cell present in tubules (arrowhead). Spermatogenesis is restored at day 56, although some disorganized and atrophic tubules are present (asterisk). Bar = 50 microns.

Testes became smaller with increased duration of WIN 18,446 treatment and directly associated with worsening modified Johnsen score (Supplementary Fig 2). In contrast, during the recovery period, the modified Johnsen score improved steadily without significant changes in testes weights, likely due to the asynchronous nature of spermatogenesis recovery.

WIN 18,446 treatment significantly reduced RA biosynthetic capacity as early as 3 days after the start of the treatment. RA biosynthetic capacity remained reduced for the rest of the treatment period (Fig. 3). RA biosynthetic capacity showed a small degree of recovery after 3 days of drug cessation and variable recovery after 7 and 14 days of drug cessation. RA synthesis capacity was not different from control testes (from no treatment control) in testes of mice that had recovery period of greater than 21 days.

Figure 3.

WIN 8,446 treatment strongly inhibits retinoic acid (RA) synthesis capacity of testes. Proteins of testes were extracted and used to determine testicular RA synthesis capacity. RA levels in enzyme reaction were adjusted by protein concentrations and incubation time. One way ANOVA was performed followed by multiple comparisons using Dunnett test. All groups were compared to the control group. *, p=0.0001. RA synthesis was almost completely inhibited by WIN 18,446 as early as 3 days after the treatment and recovered after 3 days of drug cessation. Almost a full recovery was observed in some of testes of mice that were recovered for 7 days after the WIN 18,446 treatment.

Discussion

RA, synthesized by ALDH1A enzymes, is essential in spermatogenesis at two different stages, including spermatogonial differentiation and spermiogenesis/spermiation^1,2. Our group has been examining effects of ALDH1A inhibitors on spermatogenesis and fertility^5–8 to develop a new class of drugs for male contraceptives. Because RA signaling is involved in many physiological functions in animals, concerns were raised regarding such approaches. However, genetic approaches that cause deletion of ALDH1A enzymes only in testes during development (using germ cell specific or Sertoli cell specific Cre-recombinase) or in the whole body postnatally (using tamoxifen inducible Cre-recombinase)^3,17 did not appear to affect the health of the animals significantly. These data support the feasibility of the approach of targeting ALDH1A enzymes for male contraception.

WIN 18,446 is a strong, irreversible inhibitor of ALDH1A enzymes and causes spermatogenesis blockage in diverse species, including human^8–11. In our previous studies, we showed complete recovery of fertility in mice treated with WIN 18,446 for 4 weeks followed by 9 weeks of drug cessation⁷. In addition, the offspring of the treated males showed normal fertility, suggesting that ALDH1A enzymes may be effective, reversible contraceptive targets with no adverse transgenerational impact on spermatogenesis. However, we also observed rare mineralization of tubules in testes of WIN 18,446-treated and recovered males. Thus, we wanted to further characterize the effects of ALDH1A inhibition on spermatogenesis during WIN 18,446 treatment and in the recovery period using histological and biochemical analyses to further optimize drug administration of this class (ALDH1A inhibitors) in the future. Our study showed that RA biosynthetic capacity is almost completely blocked by WIN 18,446 as early as 3 days after the treatment, a time when the testes show essentially normal spermiogenesis. A previous study has shown early histologic changes at 2 and 4 days of WIN treatment, characterized by an increased percentage of tubules (including stage VIII and stage IX tubules) with aligned spermatozoa¹⁶. In the current study, we did not calculate the percentage of tubules with aligned spermatozoa. Continued inhibition of ALDH1A activity resulted in tubular degeneration and vacuolation of the seminiferous tubules at day 7 of WIN treatment and progressive deterioration thereafter. It is interesting to observe that testes weights closely reflect worsening modified Johnsen score (Supplementary Figure 2). We interpret this to suggest that ALDH1A inhibition in both Sertoli and germ cells induce more significant defects than blockage of spermatogonial differentiation. In agreement with this interpretation, it was reported that loss of ALDH1A enzymes in Sertoli cells influence spermatogonial differentiation and spermiation³. In addition, Chung et al. reported spermiation abnormalities to be the earliest abnormalities in spermatogenesis in mice treated with pharmacological inhibitor of RAR signaling¹⁸. Endo et al. also demonstrated that ALDH1A inhibition influences spermatogonial differentiation, meiosis initiation, elongation of spermatozoa and spermiation using subcutaneous injections of WIN 18,446¹⁶.

In contrast to genetic approaches that cause permanent loss of ALDH1A, pharmacological approach of ALDH1A inhibition allows us to observe the kinetics of recovery of spermatogenesis in relation to endogenous RA synthesis. We observed recovery of RA synthesis capacity as early as 3 days after the drug cessation when most of the tubules appear to contain only Sertoli cells and spermatogonia. Previous studies have shown round spermatids that appeared normal in number up to step 7 present after 14 days of WIN withdrawal, with elongating spermatids and spermatozoa rarely encountered¹⁰, while in our study, round and elongating spermatids were present in a few tubules at 21 days after the withdrawal. Differences may be due to more sustained delivery of WIN 18,446 in our study as it was delivered by diet. While testes weight did not recover completely, a majority of the tubules contained fully differentiated spermatozoa at 8 weeks after stopping the drug treatment. This observation suggests that mice are likely fertile at this time. Indeed, we previously observed that the mice treated with WIN 18,446 for one month were fertile after 9 weeks of recovery⁷.

We detected that RA synthesis capacity are more variable at early stages of drug cessation and sometimes appear to be even higher than untreated mice (Figure 3). This may reflect higher spermatogonial contribution to total testicular cells as enzyme source in the assay was extracted from whole testes.

Our studies demonstrate that targeting ALDH1A enzymes for male contraception is feasible and likely safe. Future studies will address if longer ALDH1A inhibition can cause permanent damage to spermatogonia, and whether recovery takes longer after a longer duration of treatment or if the kinetics of recovery are independent of treatment duration. In addition, it is notable that mice with germ cell and Sertoli cell specific knock out of Aldh1a1-1a3 contained spermatogonia and Sertoli cells³. A recent study showed that spermatogenesis recovered in ~78% of tubules in vitamin A deficient mice after 15 weeks of vitamin A sufficient diet and that there are some areas of tubules more refractory to vitamin A treatment¹⁹. The atrophic tubules seen in our study in 8-week recovery mice may reflect these more vulnerable areas. We did not perform epididymal sperm counts in this study to correlate spermatogenesis blockage and recovery with viable sperms in cauda. However, based on our finding that spermatids were not observed after 2 weeks recovery period, it may be possible to devise a cyclic drug administration protocol similar to women’s hormonal contraceptives to improve the safety profile of an ALDH1A inhibitor. For example, after administration of ALDH1A inhibitor for a full spermatogenic cycle, drug holidays could be given for a period of time followed by shorter period of ALDH1A inhibitor administration.

Supplementary Material

Suppl Fig 1

Supplementary Figure 1. Schematic representation of study design. Male mice (8–25wo, n=75) were fed a diet containing WIN 18,446 (n=60) or a control diet (n=15) for a predetermined period. A group of mice (n=5) were euthanized at different time points to collect testes during the WIN 18,446 treatment period (up to 28 days). Remaining mice that were treated with WIN 18,446 were switched to the control diet and a group of mice were euthanized to collect testes for up to 8 weeks. Control diet fed mice were also euthanized at different time points to serve as normal testes controls. At the start of the study, mice were at 8–25 weeks of age to evaluate testes of mice at the similar age at the end of the study (15–25 wo).

Suppl Fig 2

Supplementary Figure 2. Testes weights are associated with histological changes in testicular cell composition. Modified Johnsen score was associated with testes weight during (A) WIN 18,446 treatment and (B) recovery. Results from regression analyses are presented (C) during treatment and (D) during recovery. Linear regression analysis shows significant correlation between testes weight and Modified Johnsen score for both phases. However, linear regression fits the data for the treatment period while non-linear regression fits the data for the recovery period better (R² 0.7801 vs. 0.8778).