BARBARA J. ATTARDI, JEAN A. ENGBRING, DAVID GROPP, AND SHERI ANN HILD*

Transcription

1 Journal of Andrology, Vol. 32, No. 5, September/October 2011 Copyright E American Society of Andrology Development of Dimethandrolone 17b-Undecanoate (DMAU) as an Oral Male Hormonal Contraceptive: Induction of Infertility and Recovery of Fertility in Adult Male Rabbits BARBARA J. ATTARDI, JEAN A. ENGBRING, DAVID GROPP, AND SHERI ANN HILD* From the Division of Reproductive Endocrinology and Toxicology, BIOQUAL Inc, Rockville, Maryland. ABSTRACT: Dimethandrolone undecanoate (DMAU: 7a,11b-dimethyl-19-nortestosterone 17b-undecanoate) is a potent orally active androgen with progestational activity that is in development for therapeutic uses in men. We hypothesized that because of its dual activity, DMAU might have potential as a single-agent oral hormonal contraceptive. To test this possibility, adult male rabbits (5/ group) of proven fertility were treated orally with vehicle or DMAU at 1.0, 2.5, 5.0, or 10.0 mg/kg/d for 12 or 13 weeks. Semen and blood samples were collected every other week through week 30. Sperm were decreased (P,.05) in semen samples from DMAU-treated rabbits at 2.5 and 5.0 mg/kg/d at weeks 12, 14, 16, 18, and 20 compared to week 0 (prior to treatment). The percentage of forward progressive motile sperm in those rabbits that still had measurable sperm was also reduced by DMAU treatment at 2.5 mg/kg/d at weeks 14, 16, 18, and 20 and at 5.0 mg/kg/d at week 18 (P,.05). At 1.0 mg/kg/d only 1 rabbit had reduced sperm numbers and motility. A mating trial was performed at week 15. The number of bred males that were fertile was 4 of 4 in the vehicle-treated group and 4 of 5, 0 of 4, and 2 of 5 in the 1.0, 2.5, and 5.0 mg/kg/d DMAU treatment groups. By week 22, sperm numbers and forward progressive motility increased, and they returned to pretreatment levels in all DMAU-treated rabbits by week 30. All bred males were fertile at week 31. Serum levels of testosterone, follicle-stimulating hormone (FSH), and luteinizing hormone (LH) were significantly suppressed in DMAU (1.0, 2.5, or 5.0 mg/kg/d)-treated rabbits during the 12-week dosing interval, but were comparable to pretreatment levels after cessation of dosing. These data indicate that DMAU suppressed the hypothalamic-pituitary-gonadal axis, resulting in severe oligospermia in the majority of rabbits in the 2.5 and 5.0 mg/kg/d dosing groups. Infertility was observed when sperm numbers decreased to about 10% of pretreatment levels. In rabbits dosed with DMAU at 10.0 mg/ kg/d, no effect on sperm numbers or motility was observed by week 12. Dosing continued for another week, and the rabbits underwent a gross necropsy on week 13 with removal of testes and epididymides for histology and preparation of testicular cytosol. Serum testosterone, FSH, and LH levels were considerably suppressed in these rabbits as in the lower-dose groups. The lack of oligospermia in the 10.0 mg/kg group as well as in the 2 fertile males in the 5.0 mg/kg group may have been due to high intratesticular levels of 7a,11bdimethyl-19-nortestosterone, the active metabolite of DMAU. Hence, as observed previously for testosterone, DMAU has a biphasic effect on spermatogenesis. Collectively, these data indicate that DMAU has the potential to be an orally active single-agent male hormonal contraceptive at an appropriate dose level and should be tested for contraceptive efficacy in nonhuman primates. Key words: Spermatogenesis, oligospermia, mating trial, gonadotropins, testosterone, liver enzymes. J Androl 2011;32: Hormonal contraception in men is aimed at inhibiting spermatogenesis to at least severe oligospermia (#10 6 sperm/ml). As both luteinizing hormone (LH; via the Leydig cell) and follicle-stimulating hormone (FSH; via the Sertoli cell) are required to Supported by National Institute of Child Health and Human Development contract NO1-HD awarded to BIOQUAL Inc. Correspondence to: Dr Barbara J. Attardi, Director, Molecular Endocrinology Laboratory, BIOQUAL Inc, 9600 Medical Center Dr, Rockville, MD ( bjattardi@bioqual.com). * Present address: National Center for Research Resources, National Institutes of Health, DHHS, Bethesda, MD Received for publication September 2, 2010; accepted for publication December 16, Previous Presentation: A part of this research was presented at the 41st annual meeting of the Society for the Study of Reproduction in Kailua-Kona, Hawaii, 2008 and the International Congress of Andrology in Barcelona, Spain, DOI: /jandrol achieve normal semen parameters, spermatogenic suppression is based on potent suppression of gonadotropin secretion (Nieschlag et al, 2003; Liu et al, 2008; Page et al, 2008). Achieving antigonadotropic activity by hormonal means results in simultaneously inhibiting testosterone production, which primarily results from LH stimulation of the Leydig cell. Androgen receptor (AR) signaling within both Leydig and Sertoli cells is also essential for fertility (Page et al, 2008). Thus, an effective hormonal contraceptive regimen must lead to strong negative feedback inhibition of both gonadotropins but also maintain circulating androgen levels to avoid symptoms of androgen deficiency. In men from east Asia, testosterone alone is effective, but one-third of Caucasian men do not respond to testosterone alone with sufficient suppression of spermatogenesis to result in effective contraception (Nieschlag et al, 2003). Therefore, in Caucasian men, progestins or gonadotro- 530

2 Attardi et al N DMAU as an Oral Male Hormonal Contraceptive 531 pin-releasing hormone (GnRH) antagonists have been combined with testosterone to augment inhibition of LH and FSH secretion and, consequently, spermatogenesis. A variety of progestins together with different testosterone formulations have been tested for male hormonal contraception in clinical trials for more than 30 years with varying degrees of success (Nieschlag et al, 2003; Sitruk-Ware, 2006; Page et al, 2008). Some progestins interact not only with progestin receptors (PR) but also with other classes of steroid hormone receptors ARs, estrogen receptors, glucocorticoid receptors, and/or mineralocorticoid receptors and show antagonist as well as agonist activities. In this regard, we have recently demonstrated that progestins with greater inherent androgenic activity are more effective in suppressing LH secretion in a castrate rat model than more pure progestins (Attardi et al, 2010). Dimethandrolone undecanoate (DMAU: 7a,11b-dimethyl-19-nortestosterone 17b-undecanoate) is a potent orally active androgen in development for therapeutic uses in men. Cleavage of the 17b-ester bond by esterases in vivo releases the biologically active androgen dimethandrolone (DMA). Like other 19-norandrogens, DMA also binds to PRs and demonstrates progestational activity both in vitro and in vivo (stimulation of endometrial gland arborization in the estrogen-primed immature female rabbit; Attardi et al, 2006). The combined androgenic and progestational activity of DMAU suggests it might have potential as a single-agent oral male hormonal contraceptive. To test this possibility, we treated proven fertile male rabbits with daily oral doses of DMAU for 12 or 13 weeks and examined effects on sperm parameters, serum levels of gonadotropins and testosterone, and fertility. Our results indicated that sperm concentration and motility were significantly suppressed by DMAU treatment, with 2.5 mg/kg/d being the most effective dose. Both a lower dose (1.0 mg/kg/d) and a higher dose (5.0 mg/kg/d) did not render all treated rabbits infertile. Furthermore, none of the rabbits treated with the highest dose, 10.0 mg/kg/d, showed any suppression of sperm concentration or motility. Thus, like testosterone in the rat and rabbit (Desjardins et al, 1973; Ewing et al, 1973; Walsh and Swerdloff, 1973), DMAU has a biphasic effect on spermatogenesis in the rabbit, but at an appropriate dose level, DMAU is an effective orally active hormonal antifertility agent. Materials and Methods Chemicals and Materials DMAU was synthesized at Southwest Foundation for Biomedical Research, San Antonio, Texas, under NICHD contract NO1-HD , and was 99% pure based on highperformance liquid chromatography analysis. Needles, syringes, Vacutainer SST tubes, and related animal supplies were purchased from NLS Inc (Baltimore, Maryland). Porcine semen extender (Enduraguard) and Leja 4-chamber 20-mm slides for use with the computer assisted semen analysis (CASA) system were purchased from Minitube of America Inc (Verona, Wisconsin). Latex probe covers (condom size) for use with the artificial vagina were purchased from ProSupply (Manchester, Connecticut). Neutral-buffered 10% formalin, semen collection tubes, and reagent grade chemicals were purchased from Sigma-Aldrich Inc (St Louis, Missouri) or VWR Inc (West Chester, Pennsylvania). Food-grade sesame oil (Hain Pure Foods, Boulder, Colorado) was purchased from a local grocery store. Reagents for alanine aminotransferase (ALT), aspartate aminotransferase (AST), and gamma glutamyl transpeptidase (GGT) assays were purchased from Randox Laboratories Ltd (Oceanside, California), and reagents for sorbitol dehydrogenase (SDH) assays were purchased from Sigma-Aldrich. Animals New Zealand White male rabbits of proven fertility (fertility was established by the supplier) were purchased from Harlan Laboratories (Oxford, Michigan). The adult males were individually housed in stainless steel cages and received 125 g Teklad 2031 rabbit diet (Indianapolis, Indiana) daily and tap water ad libitum. The rabbits diet was supplemented with kale as an extra source of fiber, and rabbits received enrichment toys in compliance with BIOQUAL s standard operating procedures. The photoperiod was 12 hours light/12 hours dark. The environmental conditions of the animal rooms were maintained as recommended in the National Research Council Guide for the Care and Use of Animals (1996). All study protocols were approved by BIOQUAL s institutional animal care and use committee. Experimental Design Rabbits (5/group) were dosed orally with vehicle, 10% ethanol in sesame oil, at 0.5 ml/kg/d, or with DMAU at 1.0, 2.5, 5.0, or 10.0 mg/kg/d for 12 weeks. This dosing interval covered the time required for spermatogenesis and epididymal transport in the rabbit. Not all treatment groups were evaluated during the same time frame. Body weights were obtained weekly, and adjustments were made to daily doses based on the most recent body weight. Prior to dosing, rabbits were trained to use an artificial vagina to collect ejaculates using a teaser female doe (Williams, 1993; Foote and Carney, 2000; Naughton et al, 2003). Semen samples were collected and analyzed every 2 weeks both during (weeks 0, 2, 4, 6, 8, 10, 12) and after (weeks 14, 16, 18, 20, 22, 24, 26, 28, 30) the dosing interval. Sperm concentration, motility, and forward progressive motility were determined using the Sperm Vision CASA system (Minitube) according to the manufacturer s recommendations. Semen samples with high numbers of sperm were diluted with porcine semen extender, and the dilution was entered into the Sperm Vision program. At week 15, when sperm concentration and motility had reached a nadir, each male rabbit was mated with 2 females to assess fertility.

3 532 Journal of Andrology N September ÙOctober 2011 Female rabbits were euthanized approximately 10 days after mating and their uteri examined for implantation sites. Females were not proven fertile by the supplier, but the presence and number of corpora lutea on the ovaries were determined to confirm mating and establish that the females had ovulated. An additional mating trial was undertaken at week 31 to ascertain recovery of fertility. In rabbits dosed with 10.0 mg/kg/d of DMAU, no effect on sperm numbers or motility was observed by week 12. Therefore, dosing continued for another week, and the rabbits underwent a gross necropsy 24 hours after the last oral dose (week 13). At necropsy, the testes and epididymides were excised for histological evaluation and weighed. One testis was used to prepare cytosol for determination of testosterone and DMA concentrations (Hild et al, 2004), and for comparison, testicular cytosols were also prepared from untreated rabbits (n 5 3). The other testis and both epididymides were preserved in Bouin solution. Preserved testes were embedded in glycol methacrylate medium, and cross-sections (2 mm) were stained with periodic acid Schiff (PAS) and hematoxylin. Epididymides were embedded in paraffin, and longitudinal sections (5 mm) were stained with PAS and hematoxylin. Blood Collection and Analysis Blood was obtained from the male rabbits every 2 weeks on the same schedule as the semen collection. During the dosing interval, these samples were collected prior to dosing, 24 hours after the previous day s dose. In addition, blood was collected on odd weeks during the treatment interval (weeks 1 11) at approximately 4 hours after oral dosing. Serum was harvested from the clotted blood and assayed for levels of testosterone, LH, FSH, and DMA. Serum levels of DMA obtained 24 hours after the last oral dose represented trough levels of drug, whereas those obtained at 4 hours after dosing approximated peak levels of DMA based on previous pharmacokinetic studies. Serum levels of testosterone were determined using DPC Coat- A-Count radioimmunoassay (RIA) (Diagnostic Products Corp, Los Angeles, California). Serum samples and testicular cytosols were extracted with 1-chlorobutane, reconstituted in zero calibrator, and assayed as described in the kit insert. The EC 90 (effective concentration: the concentration at which 90% of the radioligand is bound) value from the standard curve for the RIA was set as the limit of detection and varied from 0.11 to 0.33 ng/ml. Serum levels of FSH and LH were measured in rabbit serum using reagents from the National Hormone and Peptide Program (supplied by Dr A. F. Parlow) following the procedures received with the reagents. The standards were National Institute of Diabetes and Digestive and Kidney Disease rabbit FSH reference preparation AFP-10996C and rabbit LH reference preparation AFP-7818C. The limits of detection (EC 90 ) ranged from 0.79 to 1.44 ng/ml for FSH and from 0.22 to 0.33 ng/ml for LH, both based on 200 ml ofserum. Levels of DMA and potential immunoreactive metabolites in serum and testicular cytosols were determined using a specific RIA developed at BIOQUAL Inc (Attardi et al, 2006). Samples were extracted by methanol precipitation prior to incubation with the primary antiserum from rabbit No , bleed No. 5, at a final dilution of 1: The limit of detection, from 24 to 55 pg/ml in different assays, was calculated as the mean + 3 SD of the values obtained for the pretreatment samples (17 ml extracted serum per tube). Serum liver enzymes (AST, ALT, GGT, and SDH) were analyzed by Ani Lytics Inc (Gaithersburg, Maryland) as described previously (Hild et al, 2010). Statistical Analyses All statistical analyses were performed using SigmaStat for Windows, version 3.5 (SPSS Inc, Chicago, Illinois). All tests were 2-tailed with the significance set as a Sperm concentration, progressive motility, and total motility were compared across time for each treatment group using 1-way analysis of variance for repeated measures (ANOVA-RM), as the data did not meet the criteria for a 2-way ANOVA-RM (normality and homogeneity of variances). The specific tests used were Friedman s ANOVA-RM on ranks followed by Dunn s comparison to control for all data from rabbits receiving 2.5 mg/kg/d, and ANOVA-RM followed by the Holm-Sidak comparison to control for all data from rabbits receiving 5.0 mg/kg/d. Data from day 0 served as the control (baseline) values for these analyses. The ratio of fertile males: males bred for each treatment group was compared to vehicle-treated rabbits using a z test of proportions. Graphs were prepared using SigmaPlot version 10.0 (SPSS Inc) or GraphPad PRISM (GraphPad Software, San Diego, California). SigmaStat was used to calculate the area under the curve (AUC) for serum levels of testosterone, FSH, and LH over the treatment interval, 0 12 weeks, and recovery interval, weeks. This program was also used to determine the AUC 0 13wk for serum levels of DMA. Significant treatment effects on the AUC 0 12wk for serum testosterone and FSH were determined by Kruskal-Wallis analysis of variance (ANOVA) on ranks followed by the Student-Newman-Keuls multiple comparison test. Significant treatment effects on AUC 0 12wk for serum LH and AUC 14 30wk for serum testosterone, FSH, and LH were determined by ANOVA, and the Holm-Sidak multiple comparison test was used for a significant F value. Likewise, significant treatment effects on AUC 0 13wk for serum DMA were determined by ANOVA followed by the Holm-Sidak test. Intratesticular testosterone and DMA levels in untreated and DMAU-treated rabbits were compared using Student s t test. Data on intratesticular testosterone were log 10 transformed prior to analysis to meet the assumptions of the parametric test. Serum liver enzyme levels, ALT, AST, GGT, and SDH in rabbits receiving 10.0 mg/kg/d of DMAU were compared across time by 1-way ANOVA-RM. Data were log 10 transformed prior to analysis to meet the assumptions of the test. For a significant F value (P,.05), the Holm-Sidak test was used to determine differences. Values on day 0 served as the baseline for each rabbit. Results Effect of DMAU on Spermatogenesis and Fertility In vehicle-treated rabbits, sperm concentration varied throughout the study interval, ranging from 100 to 600

4 Attardi et al N DMAU as an Oral Male Hormonal Contraceptive 533 Figure 1. Sperm concentration (A) and progressive sperm motility (B) in semen samples collected from vehicle- and dimethandrolone undecanoate (DMAU)-treated (2.5 or 5.0 mg/kg/d) rabbits over the 30-week study interval. * indicates significantly different from week 0 (P,.05). Bars and brackets represent mean 6 SE, n 5 5, except for sperm motility in semen samples from those rabbits that still had measurable sperm; the n is indicated by the number above the bar in parentheses sperm/ml (Figure 1A). The percentage of forward progressive motile sperm tended to average between 40% and 60% (Figure 1B), and total sperm motility averaged 75% 85% (data not shown) in vehicletreated rabbits. The mean number of sperm in semen samples from rabbits treated with 2.5 or 5.0 mg/kg/d of DMAU was significantly decreased at weeks 10, 12, 14, 16, 18, and 20 compared to week 0 (prior to treatment; Figure 1A). By week 22, sperm numbers started to increase again, and these returned to pretreatment levels in all DMAU-treated rabbits by week 30, suggesting recovery of spermatogenesis. DMAU treatment at 2.5 and 5.0 mg/kg/d also reduced forward progressive sperm motility in semen samples from those rabbits that still had measurable sperm at weeks 14, 16, 18, and 20 (Figure 1B). By week 22, an increase in forward progressive sperm motility was observed, and motility returned to pretreatment levels in all DMAU-treated rabbits by week 30. Similar effects were observed on total sperm motility, which was significantly decreased compared to week 0 at weeks 14, 16, and 18 in the 2.5 mg/kg/d group and at weeks 16 and 18 in the 5.0 mg/ kg/d group. At 1.0 mg/kg/d, DMAU resulted in suppression of sperm concentration, forward progressive motility, and total sperm motility in only 1 of 5 rabbits, and at 10.0 mg/kg/d, DMAU did not affect sperm concentration or motility at any time over the 13- week dosing interval.

5 534 Journal of Andrology N September ÙOctober 2011 Table 1. Results of mating trials Wk 15 Trial Wk 31 Trial Males Females Males Females No. Fertile/ No. Corpora Lutea No. Conceptuses in No. Fertile/ No. Corpora Lutea No. Conceptuses in Treatment No. Bred in Bred Females a Pregnant Females a No. Bred in Bred Females a Pregnant Females a Vehicle 4/ / DMAU at 1.0 mg/kg/d 4/ /1 8 8 DMAU at 2.5 mg/kg/d 0/4 b / DMAU at 5.0 mg/kg/d 2/ / Abbreviation: DMAU, dimethandrolone undecanoate. a Values represent the mean 6 SE, n b Significantly different (P 5.034) from vehicle-treated males based on z test of proportions. At week 15, when sperm concentrations and motility were significantly suppressed in DMAU-treated rabbits, a mating trial was performed. All 4 vehicle-treated males that mated were fertile (1 male in this group failed to mate; Table 1), whereas all 4 males that mated in the 2.5 mg DMAU/kg/d group were infertile (1 male in this group failed to mate; Table 1); this correlated with low sperm numbers in their semen samples (, sperm/ml). Three males in the 5.0 mg DMAU/kg/d group were also infertile; 2 others, which had higher numbers of sperm in their semen samples ( sperm/ml), were fertile. The latter males exhibited normal fertility, as the number of conceptuses corresponded to the number of corpora lutea, an indirect indicator of ova ovulated. Likewise, the sole male rabbit that had low sperm numbers (, sperm/ml) in the 1 mg/kg/d DMAU group was infertile. Based on these data, infertility was observed when sperm numbers decreased to about 10% of pretreatment numbers ( sperm/ml at week 0). A second mating trial was performed at week 31. All males that mated were fertile, indicating that the males that were rendered infertile at week 15 had recovered fertility (Table 1). Serum Levels of DMA, Testosterone, FSH, and LH Trough levels of DMA (24-hour time point) in vehicle- and DMAU-treated rabbits are illustrated in Figure 2A. Serum DMA was at the limit of detection in vehicletreated rabbits. Trough levels of DMA remained elevated over the 12-week dosing interval, but were undetectable in the majority of rabbits by week 14, following discontinuation of dosing at week 12. AUC 0 13wk (pg/ml 6 wk) increased linearly with dose (r , Figure 2B): (vehicle); (1.0 mg/kg); (2.5 mg/kg); (5.0 mg/kg); (10.0 mg/kg). Overall, serum DMA levels (AUC 0 13wk ) were significantly higher (P,.05) in rabbits that received 10.0 mg/kg of DMAU than in all other groups and in the 5.0 mg/kg group compared to the 1.0 or 2.5 mg/kg groups. There was no significant (P..05) difference in the AUC 0 13wk between rabbits treated with DMAU at 1.0 and 2.5 mg/kg/ d. At 4 hours after dosing in rabbits treated with 1.0, 2.5, or 5.0 mg/kg/d of DMAU, serum DMA levels were roughly Figure 2. Serum dimethandrolone (DMA) levels in rabbits dosed orally with dimethandrolone undecanoate (DMAU) at 1.0, 2.5, and 5.0 mg/kg/d over the first 14 weeks of the study (includes 12-week treatment interval) (A). The high-dose group (10.0 mg/kg/d) was dosed orally for 13 weeks, and treatment was terminated because of the lack of effect on sperm parameters. Points represent mean 6 SE, n 5 5. Limit of detection was 42.0 pg/ml for the groups that received 1.0, 2.5, or 5.0 mg/kg/d and 54.7 for the 10.0 mg/kg/d group. The mean area under the curve for weeks 0 13(AUC 0 13wk ; pg/ml 6wk) 6 SE, n 5 5, was calculated for all groups and plotted vs dose (B). Means with different letters are significantly different.

6 Attardi et al N DMAU as an Oral Male Hormonal Contraceptive fold higher (data not shown) than the trough levels of DMA shown in Figure 2A. Serum levels of testosterone, FSH, and LH in vehicleand DMAU-treated rabbits are shown in Figure 3, and comparison of the AUCs for the treatment (weeks 0 12) and recovery (weeks 14 30) phases are presented in Table 2. In vehicle-treated rabbits, serum testosterone levels fluctuated throughout the study interval, averaging between 1 and 2 ng/ml (Figure 3A). In rabbits treated with 1.0, 2.5, or 5.0 mg/kg/d DMAU, circulating testosterone was undetectable during the 12-week treatment interval, but increased to detectable levels throughout the remainder of the study (Figure 3A; Table 2). As observed in vehicle-treated rabbits, testosterone levels varied in the DMAU-treated rabbits during weeks 14 to 30. Serum FSH levels ranged from 7 to 10 ng/ml in the rabbits prior to treatment (Figure 3B), and these levels were maintained in vehicle-treated rabbits. FSH was suppressed to about 4 ng/ml for the duration of the 12-week dosing interval in rabbits treated with DMAU at 1.0, 2.5, or 5.0 mg/kg/d (Figure 3B; Table 2). There was no difference in the degree of suppression of FSH among the 3 dose groups. After cessation of treatment, serum FSH returned to pretreatment levels in all rabbits and remained there for the duration of the study. Serum LH levels were low in the vehicle-treated males and varied throughout the study (Figure 3C; Table 2). During the treatment interval, serum LH was at or near the limit of detection in males that received DMAU regardless of the dose administered. Serum LH returned to pretreatment levels following cessation of DMAU treatment. Overall, serum levels (AUCs) of testosterone, FSH, and LH were significantly suppressed (P,.05) in DMAUtreated rabbits compared to vehicle-treated rabbits during the 12-week treatment interval. Following cessation of DMAU treatment, serum testosterone, FSH, and LH returned to pretreatment levels and were not different (P..05) from levels in vehicle-treated rabbits for weeks (Table 2). High-Dose DMAU Treatment No effect on sperm numbers or motility was observed by week 12 in rabbits dosed with DMAU at 10.0 mg/kg/d (Figure 4). Dosing was continued for another week, and the rabbits underwent a gross necropsy 24 hours after the last oral dose on week 13. High-dose DMAU treatment also resulted in suppression of serum testosterone to nondetectable levels and decreased serum FSH levels from about 6 ng/ml at week 0 to approximately 4 ng/ml for the duration of the 13-week dosing interval (Figure 5), comparable to the effects of lower doses of DMAU on these hormones. The normally low levels of serum LH were also decreased to near the limit of detection in these rabbits (Figure 5). Thus, the hypothalamic-pituitary-testicular axis was suppressed by 10.0 mg/kg/d DMAU in terms of gonadotropin and testosterone production, but sperm production was not affected. Furthermore, testicular sections from DMAUtreated rabbits showed complete spermatogenesis through formation of condensed spermatids, similar to vehicle-treated animals (Figure 6). However, the amount of interstitial tissue in the testis of rabbits treated with 10.0 mg/kg/d DMAU was reduced compared to that in vehicle-treated rabbits. In particular, the Leydig cells were shrunken in appearance and had reduced cytoplasm (Figure 6A). These observations are compatible with the presence of mature sperm in the epididymides of rabbits treated with 10.0 mg/kg DMAU and correlate with the normal numbers of sperm in the ejaculated semen. Measurement of testosterone levels in testicular cytosol indicated that testosterone production from the Leydig cells was significantly suppressed in rabbits treated with 10.0 mg/kg/d DMAU compared to untreated rabbits (Figure 7). In contrast, intratesticular DMA levels were elevated in the DMAU-treated rabbits (Figure 7), but were at the limit of assay detection in untreated rabbits, as expected. As DMA is at least 10 times more potent than testosterone (Attardi et al, 2006), intratesticular DMA was presumably sufficient to maintain spermatogenesis in these rabbits. As DMAU at 10.0 mg/kg/d had been shown previously to affect liver function in the rabbit within 14 days of treatment (Hild et al, 2010), serum samples obtained prior to dosing on days 0, 15, 30, 60, and 90 were analyzed for ALT, AST, GGT, and SDH using established methods. By day 15, serum levels of all 4 liver enzymes were significantly elevated (Table 3), confirming the results of our previous study (Hild et al, 2010). However, serum levels of AST, ALT, and SDH decreased by day 60, but only the level of serum ALT remained significantly higher than that observed prior to treatment (day 0). In contrast, serum GGT levels were significantly elevated for the duration of the treatment interval. Discussion The data presented here indicate that DMAU at 1.0, 2.5, and 5.0 mg/kg/d for 12 weeks suppressed the hypothalamic-pituitary-gonadal axis and resulted in severe oligospermia in 1 of 5, 5 of 5, and 3 of 5 rabbits, respectively. Sperm concentrations began to decline by about week 10 and remained significantly suppressed through week 20. Progressive sperm motility also declined significantly by week 14 and returned to

7 536 Journal of Andrology N September ÙOctober 2011 Figure 3. Serum testosterone (A), follicle-stimulating hormone (FSH) (B), and luteinizing hormone (LH) (C) levels in vehicle- and dimethandrolone undecanoate (DMAU)-treated rabbits (1.0, 2.5, and 5.0 mg/kg/d) over the 30-week study interval. Points represent mean 6 SE, n 5 5. Limits of detection were 0.33 ng/ml for testosterone, 1.44 ng/ml for FSH, and 0.33 ng/ml for LH.

8 Attardi et al N DMAU as an Oral Male Hormonal Contraceptive 537 Table 2. AUC for serum levels of T, FSH, and LH Serum T Levels, ng/ml 6 wk Serum FSH Levels, ng/ml 6 wk Serum LH Levels, ng/ml 6 wk Treatment Group AUC 0 12wk AUC 14 30wk AUC 0 12wk AUC 14 30wk AUC 0 12wk AUC 14 30wk Vehicle a b b b DMAU at 1.0 mg/kg/d c c c DMAU at 2.5 mg/kg/d c c c DMAU at 5.0 mg/kg/d c c c Abbreviations: AUC, area under the curve; DMAU, dimethandrolone undecanoate; FSH, follicle-stimulating hormone; LH, luteinizing hormone; T, testosterone. a Vehicle (10% ethanol in sesame oil) was dosed orally at 0.5 ml/kg/d. Figure 4. Sperm concentration (A) and motility (B) in semen samples collected from rabbits treated with vehicle or dimethandrolone undecanoate (DMAU) at 10.0 mg/kg/d for 13 weeks. Data are expressed as percentage of control. The control (100%) represents the mean values from 4 semen analyses obtained prior to study assignment for each rabbit. Bars and brackets represent mean 6 SE, n 5 5. pretreatment levels by week 22. Mating trials were performed at weeks 15 and 31 based on semen analyses at weeks 14 and 30. Infertility at week 15 was correlated with oligospermia, as 1 of 5, 4 of 4, and 2 of 5 males were infertile in the 1.0, 2.5, and 5.0 mg/kg/d DMAU dose groups. All males had regained fertility by week 31. Our results indicate that male rabbits were rendered infertile when sperm numbers decreased to #10% of pretreatment levels (, /ml). DMAU suppressed serum LH, FSH, and testosterone equally well at 1.0, 2.5, 5.0, and 10.0 mg/kg/d during the dosing interval. LH and testosterone were decreased too close to the limit of detection, whereas FSH was decreased to about 4 ng/ml at all dose levels. As FSH could be only partially suppressed following androgen treatment, it is likely that it is regulated by another factor, that is, inhibin. However, there is currently no assay available to measure serum inhibin levels in the rabbit. Serum testosterone, FSH, and LH returned to pretreatment levels following cessation of dosing. This was true in all rabbits, whether sperm concentrations were drastically reduced, resulting in infertility in the mating trial at week 15, or sperm concentrations were much less severely affected and the animals remained fertile. Thus, inhibition of gonadotropin secretion is necessary, but not sufficient, to bring about oligospermia and infertility, at least in the rabbit. The factors that account for contraceptive failure in the face of effective gonadotropin suppression are not well understood. Suppression of LH and FSH levels experimentally in men by treatment with GnRH agonists leads to a profound reduction in both testosterone production and spermatogenesis (see Page et al, 2008), but in the LH receptor deficient mouse, spermatogenesis is maintained despite a marked reduction in androgens (Zhang et al, 2004). In men treated with a combination of testosterone enanthate and cyproterone acetate, higher levels of testosterone resulted in a lesser degree of spermatogenic suppression (Meriggiola et al, 2002). Taken together, these results have led to the hypothesis that high levels of exogenous

9 538 Journal of Andrology N September ÙOctober 2011 Figure 5. Serum testosterone (A), follicle-stimulating hormone (FSH) (B), and luteinizing hormone (LH) (C) levels in vehicle- and highdose dimethandrolone undecanoate (DMAU)-treated rabbits (10.0 mg/kg/d) over the 13-week study interval. Points represent mean 6 SE, n 5 5. Limits of detection were 0.11 ng/ml for testosterone, 1.38 ng/ml for FSH, and 0.22 ng/ml for LH. androgens or residual androgen production after withdrawal of LH and FSH may allow spermatogenesis to proceed and bring about contraceptive failure (Page et al, 2008). In this model, serum DMA levels (AUC 0 13wk ) were proportional to the dose administered. Another possibility is that androgens at sufficient levels within the testes will support spermatogenesis. Intratesticular testosterone concentrations in men are approximately 100-fold higher than circulating concentrations, but the importance of these high levels in the testis is not clear (Page et al, 2008). No quantitative relationship between sperm concentration and intratesticular androgen levels has been demonstrated in men after prolonged gonadotropin inhibition (Matthiesson et al, 2005; Page et al, 2007). Clearly, a better understanding of the regulation and role of intratesticular androgens, testosterone or exogenous androgens, in men is needed to increase the efficacy of current methods of hormonal contraception. In the adult rat, there is also far more testosterone present in the seminiferous tubules than is required to maintain normal spermatogenesis, but there is a significant correlation between testosterone concentration within the testis and maintenance of advanced spermatogenesis (Zirkin et al, 1989). Whether there is a relationship between intratesticular androgen levels in the rabbit testes and spermatogenic suppression is not known. The present results suggest that high levels of intratesticular DMA might be sufficient to maintain semen parameters and fertility despite inhibition of gonadotropin secretion. This contention is supported by the failure of DMAU at the highest dose (10.0 mg/kg/d) to suppress spermatogenesis in any of the rabbits. Despite profound inhibition of intratesticular testosterone concentrations by DMAU at this dose, DMA levels in the testes were 5 6-fold higher in DMAU-treated rabbits than in untreated rabbits. As DMA is at least 10 times more potent than testosterone (Attardi et al, 2006), the intratesticular DMA levels may have been sufficient to maintain spermatogenesis in these rabbits and in the 2 of 5 rabbits in the 5.0 mg/kg/d DMAU group that were also fertile. However, we have no data on intratesticular DMA levels in the groups dosed with DMAU at 1.0, 2.5, or 5.0 mg/kg/d, as these rabbits were reused in subsequent studies once normal fertility and sperm parameters were reestablished. Hence, as observed previously for testosterone in rats and rabbits (Desjardins et al, 1973; Ewing et al, 1973; Walsh and Swerdloff, 1973), DMAU has a biphasic effect on spermatogenesis in the rabbit. The male rabbit, which develops severe oligospermia in response to appropriate doses of DMAU, may provide a model for elucidation of the parameters that account for the success or failure of hormonal contraception in the male, such as intratesticular androgen levels. A further advantage of DMAU as a potential male hormonal contraceptive is its ease of administration. Opinion polls indicate that a pill is the most popular option for a male hormonal contraceptive (see Nieschlag et al, 2003). Administration of testosterone/progestin combinations by 2 different modalities (injections, implants, topical gels, or orally) is cumbersome and

10 Attardi et al N DMAU as an Oral Male Hormonal Contraceptive 539 Figure 6. Morphology of testes from rabbits treated with dimethandrolone undecanoate at 10.0 mg/kg/d for 13 weeks (A) or vehicle (B). Rabbits were euthanized 24 hours after the last dose. less attractive. Furthermore, the use of a single-agent oral contraceptive such as DMAU may alleviate some of the troublesome side effects associated with use of progestins in combination with testosterone in men, such as weight gain due to bloating and water retention, greater decrease in high-density lipoprotein and increase in hematocrit than what is produced by testosterone alone, acne, and increase in proinflammatory cytokines and nocturnal sweating (Nieschlag et al, 2003; Sitruk- Ware, 2006; Page et al, 2008); but this needs to be tested directly. For either hormonal therapy or contraception in men, DMAU would require chronic administration. As other synthetic alkylated androgens such as 17b-methyltestosterone have been shown to be hepatotoxic (delorimier et al, 1965; Heywood et al, 1977; Tennant et al, 1981; Taylor and Snowball, 1984; Ishak and Zimmerman, 1987), we investigated the effect of DMAU on liver function using the adult male rabbit as a model (Hild et al, 2010). Potential effects of DMAU on liver function (bromsulfonphthalein retention and serum liver enzymes) in that study could be detected as early as 14 days after daily treatment with 10.0 mg/kg/d. In the present work, we took advantage of the long-term treatment of male rabbits with DMAU to further evaluate its possible hepatotoxicity. Serum samples obtained on days 0, 15, 30, 60, and 90 from rabbits dosed with 10.0 mg/kg/d of DMAU for 13 weeks were analyzed for ALT, AST, GGT, and SDH. Transient increases (day 15) in all 4 liver enzymes were observed, but with the exception of GGT, which remained significantly elevated for the duration of the treatment interval, liver enzyme levels decreased by day 60, suggesting the animals accommodate to continued Table 3. Serum levels of liver enzymes in rabbits treated with DMAU at 10.0 mg/kg/d for 13 weeks a Figure 7. Intratesticular levels of testosterone (T) and dimethandrolone (DMA) in untreated rabbits or rabbits that received 10.0 mg/kg/d of dimethandrolone undecanoate (DMAU) for 13 weeks. Data represent mean 6 SE, n 5 3 for untreated and n 5 5 for DMAU-treated rabbits. * indicates significantly different (P,.05) from untreated animals. Study Day AST (U/L) ALT (U/L) GGT (U/L) SDH (U/L) b b b b c c c c c,d c,d c b,c b,d d c b,c b,d d c b,c Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; DMAU, dimethandrolone undecanoate; GGT, gamma glutamyl transpeptidase; SDH, sorbitol dehydrogenase. a Reference values for these enzymes in male rabbits are U/L for AST, U/L for ALT (Loeb and Quimby, 1989), and U/L for GGT (Prasad, 2008). Normal values for SDH in rabbits were unavailable. b,c,d Means with different letters within a column are significantly different from one another (P,.05).

11 540 Journal of Andrology N September ÙOctober 2011 DMAU treatment. Furthermore, we did not observe any increases in liver enzyme levels in either adult male rats or cynomolgus monkeys treated daily for 28 days with DMAU at several doses in toxicology studies. In conclusion, the present findings suggest that DMAU has potential as an orally active single-agent reversible male hormonal contraceptive, with minimal hepatotoxicity, and should be tested in a nonhuman primate model. Acknowledgments The authors would like to thank Dr Sailaja Koduri for her help with the RIAs and Robert Heikkila of Ani Lytics Inc, Gaithersburg, Maryland, for analysis of serum samples for ALT, AST, GGT, and SDH levels. The expertise of the following technicians is gratefully acknowledged: Huifen Gao, Bruce Till, Anne Semon, Trung Pham, Laurent Pessaint, Jessica Pray, and Yangqing Zhao. Care of the animals was provided by Martha Washington and Guideon Nkamdjeu. We thank also Drs Richard Blye of the Contraception and Reproductive Health Branch (CRHB), NICHD, and Jerry R. Reel for their input into the design of this study and useful discussions, and Dr June Lee of the CRHB for her continued support of this research. References Attardi BJ, Hild SA, Reel JR. Dimethandrolone undecanoate: a new potent orally active androgen with progestational activity. Endocrinology. 2006;147: Attardi BJ, Koduri S, Hild SA. Relative progestational and androgenic activity of four progestins used for male hormonal contraception assessed in vitro in relation to their ability to suppress LH secretion in the castrate rat. Mol Cell Endocrinol. 2010;328: delorimier AA, Gordan GS, Lowe RC, Carbone JV. Methyl testosterone, related steroids, and liver function. Arch Intern Med. 1965;116: Desjardins C, Ewing LL, Irby DC. Response of the rabbit seminiferous epithelium to testosterone administered via polydimethylsiloxane capsules. Endocrinology. 1973;93: Ewing LL, Stratton LG, Desjardins C. Effect of testosterone polydimethylsiloxane implants upon sperm production, libido and accessory sex organ function in rabbits. J Reprod Fertil. 1973;35: Foote RH, Carney EW. The rabbit as a model for reproductive and developmental toxicity studies. Reprod Toxicol. 2000;14: Heywood R, Chesterman H, Ball SA, Wadsworth PF. Toxicity of methyl testosterone in the beagle dog. Toxicology. 1977;7: Hild SA, Attardi BJ, Koduri SA, Till BA, Reel JR. Effects of synthetic androgens on liver function using the rabbit as a model. J Androl. 2010;31: Hild SA, Attardi BJ, Reel JR. The ability of a gonadotropin-releasing hormone antagonist, acyline, to prevent irreversible infertility induced by the indenopyridine, CDB-4022, in adult male rats: the role of testosterone. Biol Reprod. 2004;71: Ishak HK, Zimmerman HJ. Hepatotoxic effects of the anabolic androgenic steroids. Semin Liver Dis. 1987;7: Liu PY, Swerdloff RS, Anawalt BD, Anderson RA, Bremner WJ, Elliesen J, Gu YQ, Kersemaekers WM, McLachlan RI, Meriggiola MC, Nieschlag E, Sitruk-Ware R, Vogelsong K, Wang XH, Wu FC, Zitzmann M, Handelsman DJ, Wang C. Determinants of the rate and extent of spermatogenic suppression during hormonal male contraception: an integrated analysis. J Clin Endocrinol Metab. 2008;93: Loeb WF, Quimby FW, eds. The Clinical Chemistry of Laboratory Animals. 1st ed. New York, NY: McGraw-Hill; Matthiesson KL, Stanton PG, O Donnell L, Meachem SJ, Amory JK, Berger R, Bremner WJ, McLachlan RI. Effects of testosterone and levonorgestrel combined with a 5a-reductase inhibitor or gonadotropin-releasing hormone antagonist on spermatogenesis and intratesticular steroid levels in normal men. J Clin Endocrinol Metab. 2005;90: Meriggiola MC, Costantino A, Bremner WJ, Morselli-Labate AM. Higher testosterone dose impairs sperm suppression induced by a combined androgen-progestin regimen. J Androl. 2002;23: Naughton CK, Nelson DR, Thomas AJ Jr. Development of an inexpensive artificial vagina for semen collection from rabbits. J Androl. 2003;24: Nieschlag E, Zitzmann M, Kamischke A. Use of progestins in male contraception. Steroids. 2003;68: Page ST, Amory JK, Bremner WJ. Advances in male contraception. Endocr Rev. 2008;29: Page ST, Kalhorn TF, Bremner WJ, Anawalt BD, Matsumoto AM, Amory JK. Intratesticular androgens and spermatogenesis during severe gonadotropin suppression induced by male hormonal contraceptive treatment. J Androl. 2007;28: Prasad K. Serum biochemical changes in rabbits on a regular diet with and without flax lignan complex following a high cholesterol diet. Int J Angiol. 2008;17: Sitruk-Ware R. New progestagens for contraceptive use. Hum Reprod Update. 2006;12: Taylor W, Snowball S. The effects of long-term administration of methyl testosterone on the development of liver lesions in BALB/c mice. J Pathol. 1984;143: Tennant BC, Balazs T, Baldwin BH, Hornbuckle WE, Castleman WL, Boelsterli U, Kallfelz FA. Assessment of hepatic function in rabbits with steroid-induced cholestatic liver injury. Fundam Appl Toxicol. 1981;1: Walsh PC, Swerdloff RS. Biphasic effect of testosterone on spermatogenesis in the rat. Invest Urol. 1973;11: Williams J. Semen analysis and fertility assessment in the rabbit. In: Chapin RE, Heindel JJ, eds. Male Reproductive Toxicology. San Diego, CA: Academic Press Inc; 1993; Methods in Toxicology; Vol. 3, part A. Zhang FP, Pakarainen T, Zhu F, Poutanen M, Huhtaniemi I. Molecular characterization of postnatal development of testicular steroidogenesis in luteinizing hormone receptor knockout mice. Endocrinology. 2004;445: Zirkin BR, Santulli R, Awoniyi CA, Ewing LL. Maintenance of advanced spermatogenic cells in the adult rat testis: quantitative relationship to testosterone concentration within the testis. Endocrinology. 1989;124: