A Novel Potent Indazole Carboxylic Acid Derivative Blocks Spermatogenesis and Is Contraceptive in Rats after a Single Oral Dose 1

Transcription

1 BIOLOGY OF REPRODUCTION 78, (2008) Published online before print 23 January DOI /biolreprod A Novel Potent Indazole Carboxylic Acid Derivative Blocks Spermatogenesis and Is Contraceptive in Rats after a Single Oral Dose 1 Joseph S. Tash, 2,4,5 Barbara Attardi, Sheri A. Hild, Ramappa Chakrasali, 7 Sudhakar R. Jakkaraj, 7 and Gunda I. Georg 3,5,7 Department of Molecular and Integrative Physiology 4 and Interdisciplinary Center for Male Contraceptive Research and Drug Development, 5 University of Kansas Medical Center, Kansas City, Kansas 10 BIOQUAL, Inc., Rockville, Maryland Department of Medicinal Chemistry, 7 University of Kansas, Lawrence, Kansas ABSTRACT Women have historically been the focus for development of new contraceptive methods. The National Institutes of Health, World Health Organization, and Institute of Medicine have stressed the need to develop nonhormonal, nonsteroidal male contraceptive agents. We report results from initial dose-ranging studies of a new indazole carboxylic acid analogue, gamendazole. An infertility rate of 100% was achieved in seven out of seven proven-fertile male rats 3 wk after a single oral dose of mg/kg of gamendazole. Fertility returned by 9 wk in four of seven animals, with typical numbers of normal-appearing conceptuses. A fertility rate of 100% returned in four of six animals that became infertile at a single oral dose of 3 mg/kg of gamendazole. No differences in mating behavior were observed in either of the gamendazole-treated groups versus the control (vehicle-only) group. In the animals that showed reversible infertility, a transient increase in circulating FSH levels coincided with an initial decline in inhibin B levels after administration of gamendazole, but no other significant changes in circulating reproductive hormones were observed. Gamendazole inhibited production of inhibin B by primary Sertoli cells in vitro with a median inhibitory concentration of.8 þ 3.0 (SEM) ¾ M, suggesting that Sertoli cells are a primary target. A biotinylated gamendazole analogue revealed cytoplasmic and perinuclear binding of gamendazole in primary Sertoli cells. Gamendazole represents the most potent new oral antispermatogenic indazole carboxylic acid to date. Our results, however, demonstrate that additional dose-finding studies are required to improve reversibility and widen the therapeutic window before more detailed drug development of this potential nonhormonal male contraceptive agent can occur. fertilization, male contraception, Sertoli cells, spermatogenesis, testis 1 Supported by National Institutes of Health (NIH) U54 HD (to J.S.T.), NIH N01 HD (to G.I.G.), and U54 HD33994 Center for Reproductive Sciences, a National Institute of Child Health and Human Development (NICHD) Grant for Specialized Cooperative Centers Program in Reproductive Research (SCCPRR), at the University of Kansas Medical Center. Also supported by NICHD contract N01-HD awarded to BIOQUAL, Inc. Dedicated to the memory of Geoffrey M.H. Waites, Sc.D., Ph.D. and Thaddeus R.R. Mann, M.D., Ph.D., Sc.D., F.R.S. 2 Correspondence: FAX: ; jtash@kumc.edu 3 Current address: Department of Medicinal Chemistry, 717 Delaware Street SE, University of Minnesota, Minneapolis, MN Received: 3 October 200. First decision: 8 January Accepted: 10 December Ó 2008 by the Society for the Study of Reproduction, Inc. ISSN: INTRODUCTION World Health Organization statistics show that 122 million planned pregnancies occur worldwide per year [1]. Despite of the availability of many different female contraceptive methods and of condoms, an additional 87 million pregnancies were unintended (representing 42% of all pregnancies), and 4 million pregnancies were terminated by abortion. In the United States, the unintended pregnancy rate is 49% of all births, and about half of these are terminated by abortion [2]. Surprisingly, in 50% of unintended pregnancies, the women reported having used a contraceptive [2]. Thus, the development of novel reversible oral male contraceptive agents has been identified by the National Institutes of Health, Institute of Medicine, and World Health Organization as being a major advance needed to address this worldwide reproductive health issue [3 5]. The need for novel approaches to develop a male contraceptive is highlighted in these reports, because traditional approaches to male contraception have yielded mixed results, problems with reversibility, and/or unacceptable side effects. In a recent immunocontraception study, an infertility rate of 78% was achieved in monkeys that showed a high-titer immune response to a testis/epididymis-specific protein antigen [].A third of the animals, however, did not generate titers sufficient to warrant inclusion in fertility trials, a problem common to immunocontraception approaches [7, 8]. Nonetheless, these studies provide important proof-of-concept outcomes for immunocontraceptive approaches and help to identify potential targets for drug development. Alternatively, small molecules may be amenable to use as contraceptives, because they may pass through the blood-testis barrier more readily. One small compound, N-butyldeoxynojirimycin, an iminosugar used in the treatment of Gaucher disease and infection with human immunodeficiency virus, is a focus for development of reversible nonhormonal male contraceptive agents that target spermiogenesis [9]. We have focused on indazole carboxylic acids, which severely inhibit spermatogenesis and are reversible in rats, mice, rabbits, dogs, and nonhuman primates [10, 11]. In particular, lonidamine (LND; and its analogue gamendazole, trans-3-(1-benzyl--(trifluoromethyl)-1h-indazol-3-yl)acrylic acid) was shown to be an efficacious and reversible antispermatogenic agent in nonhuman primates [10], but its use has been limited by undesirable side effects [12, 13]. The reported side effects for LND include asthenia, muscular pain, testicular pain, nausea, vomiting, somnolence, and elevation of liver enzymes. Nonetheless, LND is currently in use in humans as an anticancer therapeutic agent [14 1]. To be acceptable as a male contraceptive agent, however, none of these side effects would be acceptable. 1127

2 1128 TASH ET AL. TABLE 1. Criteria for assessment of spermatogenic index based on testicular morphology from Whitsett et al. [21]. Stage The successful development of a male contraceptive also will have a significant effect on world health and associated health care costs and risks. A potent reversible male contraceptive might help to decrease the risks and costs associated with 0% of the 4 million pregnancies terminated by an abortion performed under medically unsafe conditions [1]. It also might have a significant impact on reducing the bulk of unintended pregnancies that result in premature and lowbirth-weight babies that account for nearly half of the $29.3 billion spent during 2001 on obstetrics hospital charges [2]. In this regard, in the United States, just the costs associated with an unplanned birth range between $147 for a normal birth to more than $ for a premature baby with problems [2]. A male contraceptive might help to decrease the 13% of the global disease burden that is attributable to maternal conditions, such as hemorrhage, infection, or unsafe abortion among women aged years [17]. Thus, the successful development of a male contraceptive will offer real alternatives for family planning and contribute significantly to reducing these health and economic burdens worldwide. To become an acceptable male contraceptive drug, a candidate agent is required to meet high efficacy standards with minimal to no toxicity. A male contraceptive drug will be taken by otherwise healthy individuals; thus, the number and level of acceptable side effects are extremely low to none. A second criterion is that the drug have a high reversibility rate ideally, comparable to or better than those of existing female contraceptive methods. The third main criterion is the potential effects of the drug on pregnancies. The health of fetuses/infants/ children from fathers who have taken the drug previously should not be affected; there must be no increase in the level of prenatal loss, birth defects, or infant mortality in the children of men who elect to stop taking the drug and to have children again. To meet these challenges, our goal was to generate new analogues of LND with increased antispermatogenic potency and decreased side effects. Here, we report the results of drug discovery and initial dose-ranging and mating experiments for a novel LND analogue with potent antispermatogenic activity. MATERIALS AND METHODS Morphologic hallmarks 1 Only spermatogonia present. 2 Spermatogonia and spermatocytes present. 3 Spermatogonia, spermatocytes, and round (early) spermatids present with less than 5 late spermatids per tubule. 4 Spermatogonia, spermatocytes, and round spermatids present with up to 25 late spermatids per tubule. 5 All cell types present with late spermatids per tubule. All cell types present with more than 75 late spermatids per tubule. Animals and Compound Administration All the experiments using animals were conducted under protocols approved by the University of Kansas Medical Center and BIOQUAL, Inc., institutional animal care and use committees as specified by federal guidelines. Before testing in animals, all new analogues were subjected to mutagenicity screening at a final compound concentration of 100 lm with and without S9 (rat liver microsomes; Moltox) using an Ames test kit (Environmental Bio- Detection Products, Inc.). Compounds proving to be mutagenic (either with or without S9 or at any concentration) were not further tested in animals. For initial screening and characterization of antispermatogenic activity, control and test groups each contained five male Long-Evans rats (age, 5 70 days; weight, g). Similar numbers of animals per group have been used in previous studies [18, 19]. Initial screening was performed in three groups of animals receiving a single i.p. injection of 25 mg/kg, 200 mg/kg, or an equivalent volume of vehicle control (at 2.5 ml/kg). Compounds that were more potent than LND by i.p. administration at 25 mg/kg were then tested by oral administration (single oral gavage) and included in mating trials. For oral administration, gamendazole was formulated in 10% ethanol/sesame oil. For the mating trials, one group of rats (n ¼ 7 per group) was treated with vehicle (5 ml/kg), and the remaining groups were treated with gamendazole at 0.75, 1.5, 3.0, or.0 mg/kg as a single dose or at.0 mg kg 1 day 1 for 7 days. Processing of Tissues for Histology In preliminary experiments using LND as a standard, disruption of the germinal epithelium was noted 48 h after a single i.p. injection. The lumens of the seminiferous tubules, however, were filled with released cells, making it impossible to assess the spermatogenic index. Between 5 and 7 days after a single administration, clearance of shed cells and debris from the lumens had stabilized, making it possible to quantify the spermatogenic index. During early experiments, animals were killed on Day 7, but later, Day 5 was chosen to reduce costs with no effect on the resulting data. After animals were killed, the testes were removed and weighed. Testes were processed for histology as described previously [20] except that the tissue was stained with hematoxylinand-eosin. In the postmating trial, testes sections were stained with periodic acid-schiff and hematoxylin. In the initial screening experiments described above, at death, two animals from each group were randomly selected for tissue necropsy (including brain, lung, stomach, heart, liver, spleen, kidneys, and gut). In cases when animals died or were killed before the 5-day posttreatment schedule (for AF2785, two of six animals at 200 mg/kg i.p.; for gamendazole, three of five animals at 200 mg/kg i.p.), these rats also were subjected to necropsy. The tissues from this screening experiment were evaluated by a veterinary pathologist to provide initial information regarding potential toxicity. More detailed and extensive toxicity studies are needed to further evaluate the toxicity of gamendazole. Spermatogenic Index The method described by Cook et al. [18], as modified by Whitsett et al. [21], was used to score the effects of test compounds on spermatogenesis. The assessment includes not only testis weight but also a spermatogenic index. In one cross-section through the middle of the testis, all tubules are examined for the presence of spermatogonia, spermatocytes, and spermatids. Based on the number of each cell type per tubule, a score is assigned as detailed in Table 1. Mating Trials Two mating trial assays were performed to evaluate the dose-dependent antifertility effects of gamendazole in rats. Adult male Crl:CD (SD) outbred rats (weight, 300 g; age, 9 wks; Charles River Laboratories) were used for these mating trial assays. Fertility was established as described previously [22]. In the first assay, proven-fertile male rats (n ¼ per group) received a single oral dose of vehicle (5 ml/kg of 10% ethanol in sesame oil) or 0.75, 1.5, or 3.0 mg/kg of gamendazole (day of dosing considered to be Day 0 of Week 0). Mating trials were performed as described previously [23] at Weeks 1 9, and male rats were killed and a gross necropsy performed on Week 10 [22]. For the second study, proven-fertile male rats (n ¼ 7 per group) were dosed orally as follows: vehicle control at 5 ml kg 1 day 1 for 7 days (Days 0 ), gamendazole at mg/kg on Day 0 (single dose), or gamendazole at mg kg 1 day 1 each day for 7 days (Days 0 ). Mating trials were performed at Weeks 1 10 (each week) and Weeks 12, 14, 18, and 2. At Week 27, the male rats were killed, and gross necropsy was performed. Blood samples were collected from the tail vein of the male rats throughout both studies to obtain serum samples for hormone assays. The number of conceptuses was determined after necropsy of the female rats, and the conceptuses were grossly evaluated and classified as either normalappearing or resorbing. The number of normal conceptuses per pregnant female rat between the two groups was compared statistically by one-way ANOVA. At necropsy, testes, ventral prostate, and seminal vesicles were excised and weighed. Both epididymides and the right testis were excised, weighed, and preserved in Bouin solution [23]. The left testis was homogenized and used to determine the number of mature spermatid heads [23]. These testicular endpoints were compared to their respective control groups using a Student t- test (see Tables 4 and 5).

3 NEW REVERSIBLE NONHORMONAL ORAL MALE CONTRACEPTIVE 1129 FIG. 1. Sequential testing strategy for novel indazole carboxylic acids. New analogues of LND were subjected to sequential go-no go tests as part of an overall structure-activity relationship approach. Compounds that failed a test were not carried into the next phase and dropped from further study. The criteria for pass/fail are summarized above each go-no go decision point in the strategy. The tests, in sequence, were 1) Ames test (go ¼ no mutagenicity), 2) single i.p. administration at 25 and 200 mg/kg (go ¼ greater than 50% of the animals respond, with spermatogenic index,3 at 25 mg/ kg), 3) single oral administration at 25 mg/kg and lower (go ¼ efficacy is same or better than equivalent i.p. dose), 4) single dose-ranging efficacy mating (go ¼ minimum of 0% infertility for further dose finding, 100% infertility priority), and 5) single dose-mating/longitudinal recovery (go ¼ greater than 0% recovery, 100% recovery priority). Details of each test are presented in Materials and Methods. Following the mating/recovery study in the figure is a summary of the next key components of dose-ranging studies that will be required to assess the compounds in preparation for more detailed proof-ofconcept and suitability for further drug development. IND, Investigational new drug. Serum Hormone Assays Serum samples collected throughout the present study were analyzed for inhibin B by an ELISA kit (Oxford Bio-Innovation Ltd.). The RIAs for measuring testosterone and FSH in the rat have been described previously [22]. Briefly, testosterone was measured in unextracted serum samples using a kit (Coat-A-Count) from Diagnostic Products Corp. Rat FSH was measured using National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) reagents supplied by Dr. A.F. Parlow (National Hormone and Peptide Program, Torrance, CA) and according to the procedures received with the reagents. The standard was NIDDK-rFSH-RP-2. Preparation of Sertoli Cell-Enriched Cultures from Rats and Mice Testes were collected from ten 1-day-old rats, washed, and decapsulated. Sertoli cells were obtained using the method described by Delfino and Walker [24]. The final preparations contained more than 95% Sertoli cells as determined by Oil Red O staining. Mouse Sertoli cells were isolated from 27-day-old animals using the method described by Heckert et al. [25]. Sertoli Cell ATP assay Isolated Sertoli cells were plated in black, 9-well dishes and incubated for 72 h at 338C. Culture medium (serum-free, supplemented Dulbecco modified Eagle medium/f-12) was removed and replaced with fresh medium containing various concentrations of gamendazole (10 10 to M) dissolved in ethanol or dimethyl sulfoxide (final concentration in medium, 0.1%). Cells were returned to 338C and incubated for 48 h. The ATP content of the cells was measured at 48 h using the PerkinElmer ATP-Lite M kit, containing mammalian lysis buffer, lyophilized substrate, and reconstitution buffer, according to the manufacturer s instructions. Luminescence was measured in a scintillation counter (PerkinElmer TopCount). The data were plotted and the median inhibitory concentrations (IC 50 ) estimated using GraphPad PRISM software. Sertoli Cell Inhibin B Immunoassay Isolated Sertoli cells were plated, incubated, and then treated with various concentrations of gamendazole (10 11 to 10 M) and incubated for 72 h as described above. Medium was collected for measurement of inhibin B using a kit from Sertotec according to the manufacturer s directions except that the boiling step was omitted, and the standard curve (recombinant human inhibin B) was prepared in Sertoli cell culture medium. The limit of detection (lowest standard) was 5. pg/ml. Binding of Gamendazole to Primary Sertoli Cells In Vitro Primary mouse Sertoli cells were isolated as described above and plated on pretreated glass coverslips. The cells were washed with PBS to remove excess media and then fixed in 2% paraformaldehyde, followed by permeabilization with 1% NP-40. The cells were blocked with 5% BSA in buffer for 1 h at room temperature, washed with buffer, and then incubated with a ultraviolet (UV) cross-linking biotinylated analogue of gamendazole (GMZ-BT-UV) for 1 h at 378C. The cells were washed with buffer to remove unbound drug, followed by incubation in the dark for 20 min at room temperature with fluorescein-avidin D (Vector Laboratories). Cross-linking was accomplished with a UV mineral lamp at long wavelength for 20 min at 48C. Coverslips were mounted with VectaShield mounting medium (Vector Laboratories) suitable for fluorescence. Statistical Analyses Data were checked for normality and for homogeneity of variance before performing parametric tests. The appropriate statistical method for data analyses are described within the text, or each table, as necessary. RESULTS Initial Ranking of Compounds Based on Efficacy Using Testis Weight and Blockage of Spermatogenesis as Parameters More than 150 new analogues of LND were synthesized (patent pending) by modifying positions in the molecule that had not been reported previously [10, 11, 2, 27]. A sequential go-no go testing strategy was developed to remove compounds showing mutagenicity, poor efficacy, and readily identifiable toxicity before proceeding to initial dose-ranging mating trials (Fig. 1). Of the 150 new analogues, 17 were mutagenic (either with or without S9 added to the buffer mixture, or both) and, thus, were not carried into animal studies. Next, compounds that were not mutagenic were screened in male rats at a single dose of 25 mg/kg i.p. as described in Materials and Methods. Of the remaining new analogues, 15 showed antispermatogenic activity in the initial animal screening. Gamendazole, however, was the first compound subsequently carried into mating trials and reversibility studies, thus becoming a new lead compound for further study. Both AF2785 and AF234 (Fig. 2) were synthesized and tested for comparison with LND analogues already reported in the literature [2, 27]. Of previously published compounds, only AF2785 and AF234 were chosen as reference standards to LND, because all other previous indazole carboxylic acid analogues were either less potent, more toxic, and/or poorly

4 1130 TASH ET AL. FIG. 2. Chemical structures of LND, AF234, AF2785, and gamendazole (GMZ; trans-3-(1-benzyl--(trifluoromethyl)-1h-indazol-3-yl)acrylic acid; patent pending) and UV cross-linking biotinylated gamendazole (patent pending). reversible. To rank AF2785, AF234, and the newly designed compounds based on efficacy relative to LND, dose-response experiments were first conducted to identify a dose of LND that would give a response in approximately 50% of the animals tested as follows: A positive response in an animal was based on a reduction in testis weight of greater than 30% at five days after a single compound administration as well as a loss of spermatids (number of spermatids per tubule, 5) in more than 50% of the tubules. For all compounds that elicited a response, reductions in testis weight always matched loss of spermatids. Using these criteria, a dose of 25 mg/kg of LND was found to give an average 50% response (i.e., n. 2 animals/group) in replicate groups of five animals. Using the same criteria for AF2785, AF234, and most of the new analogues that had an antispermatogenic effect, the response was usually all or none. In other words, animals that did not respond were nearly identical in testis weight to controls, and testicular histology appeared to be near normal. In the responding animals, the testes were routinely 30 to 50% lower in weight compared with controls, and histology revealed at least 75% of the tubules with five or fewer spermatids. Similar findings were observed in rats evaluated 7 days after oral administration of LND or gamendazole at 25 mg/kg. Testis weights were decreased by 32% (P, 0.02) and 54% (P, ) in LNDand gamendazole-treated rats, respectively. In addition, the percentage of tubules exhibiting normal morphology were significantly decreased (P, 0.05) by oral treatment with LND (32% 10%) and gamendazole (1% 1%) as compared to vehicle-treated rats (99% 1%; mean SEM, n ¼, based on Kruskal-Wallis ANOVA on ranks followed by Student- Newman-Keuls test). In the responding animals, the disrupted germinal epithelium appeared to be identical to what has been published previously [10, 2, 28 32] (Fig. 3). A dose of 400 mg/kg of LND is near the reported median lethal dose (LD 50 ) and caused 40% mortality, whereas no animal losses occurred at all with the lower doses of LND. Thus, for routine screening, we used two doses to test new analogues. A low dose of 25 mg/ kg was used so that potency relative to LND could be determined; and a high dose of 200 mg/kg was given to screen for toxicity and to determine whether increased mortality resulted relative to that with LND and two other analogues, AF2785 and AF234. Table 2 summarizes the results for gamendazole, LND, AF2785, and AF234. No mortality was observed with any LND analogue or gamendazole at 25 mg/kg; however, gamendazole was the only analogue with 100% efficacy at this dose. In contrast, mortality was observed at 200 mg/kg of AF2785 or gamendazole. Because a minimum dose of 25 mg/kg produced an antispermatogenic effect and was 8-fold lower than the lethal dose (see below), detailed LD 50 determinations were deferred to future, more detailed dose-finding and toxicology studies. Additional dose-ranging studies of compounds emerging from the first phase of testing subsequently were conducted. Experiments by oral administration were thus carried out (Fig. 1), and these experiments determined that gamendazole also was orally bioavailable and efficacious as a contraceptive at lower doses (see below and Fig. 4). Gamendazole Causes Loss of Late-Stage Spermatogenic Cells Gamendazole was identified from the design and testing of more than 150 new LND analogues. The structure of gamendazole is presented in Figure 2. Gamendazole became the focus for detailed investigation because it was the first compound to show an antispermatogenic effect in 100% of animals at the screening dose of 25 mg/kg (Fig. 3 and Table 2), whereas LND inhibited spermatogenesis in 50% of animals at this dose. The antispermatogenic effect of gamendazole at 25 mg/kg (Fig. 3, C and D) was histologically similar to LND (Fig. 3, E and F). Compared with the control (Fig. 3, A and B), both LND and gamendazole cause a loss of the structured spermatogenic cell-type layering, with loss of spermatozoa and TABLE 2. Summary of antispermatogenic response and mortality data five days posttreatment in rats after single IP administration of LND and a selected group of analogues. Compound Response a (25 mg/kg) Mortality b (25 mg/kg) Response a (200 mg/kg) Mortality b (200 mg/kg) LND 2/5 (40%) 0/5 5/5 (100%) 0/5 c AF2785 3/5 (0%) 0/5 3/4 (75%) 2/ AF234 3/5 (0%) 0/5 4/5 (80%) 0/5 Gamendazole 5/5 (100%) 0/5 2/2 (100%) 3/5 a The fraction represents the number of animals responding by a decline in testis weight.30%/the total number of animals tested in the group. b The fraction represents number of animals deceased/number of animals tested. c Mortality in 2 out of 5 animals was observed at a single oral dose of 400 mg/kg.

5 1131 NEW REVERSIBLE NONHORMONAL ORAL MALE CONTRACEPTIVE FIG. 4. Dose-dependent effect on testis weight 5 days after single oral administration of gamendazole or LND. The high variability observed in the LND-treated rats results from the fact that two of five rats did not show a response to LND. doses of gamendazole (3 and mg/kg), which were used for the mating trial, similar loss of spermatids and spermatozoa were observed (Fig. 3, G J). A higher retention of seminiferous epithelial organization (9% normal tubules for 3 mg/kg and 85% normal tubules for mg/kg), however, was observed, with less separation of cells and no vacuolation. Much less sloughing of the larger primary spermatocyte cells and smaller secondary spermatocytes was seen. This result was confirmed by the quantitative spermatogenic index (Table 3). The enhanced potency of gamendazole versus LND was revealed by a dose-dependent decline in testis weight 5 days after single oral administration (Fig. 4): An ED50 of approximately 25 mg/kg was estimated for LND based on a 50% decline in testis weight from the control to the minimum resulting testis weight in the group treated with higher drug concentrations. An ED50 of 0.8 mg/kg was calculated for gamendazole based on the dosing study with 0 to 3 mg/kg. Partially Reversible Infertility after a Single Oral Dose FIG. 3. Comparison of single doses of gamendazole or LND on testicular histology. A and B) Testis from a control animal administered vehicle alone. C and D) Testis from an animal that responded to 25 mg/kg i.p. of LND. E and F) Testis from an animal treated with 25 mg/kg i.p. of gamendazole. G and H) Testis from an animal treated with mg/kg of oral gamendazole. I and J) Testis from an animal treated with 3 mg/kg of oral gamendazole. Bar ¼ 200 lm. The potent inhibition of spermatogenesis by gamendazole suggested that it may have antifertility activity. Dose-ranging efficacy mating trials in proven-fertile male rats demonstrated that single oral doses at 1.5, 3.0, and.0 mg/kg were effective at inducing infertility in rats (Tables 4 and 5). At a single oral dose of 1.5 mg/kg, gamendazole induced infertility in two of six rats for only 1 wk (Week 3 or 4) of the 9-wk mating trial (Table 4). At a single oral dose of 3.0 mg/kg, four of six males were rendered infertile by Week 4, and all four of these males recovered fertility by Week 5 (Table 4). A single oral dose of.0 mg/kg of gamendazole produced 100% infertility 4 wk TABLE 3. Spermatogeneic index in seminiferous tubules five days after a single dose of LND analogues. spermatids from the germinal epithelium. The histologic patterns suggest that gamendazole produced more extensive loss of spermatids compared with LND. Spermatogenic index data support this observation (Table 3). Sloughing of the larger cells with dense chromatin pattern and high nuclear:cytoplasmic ratio (primary spermatocytes) and smaller, less densely staining cells with less prominent chromatin pattern (secondary spermatocytes) also was seen with LND and gamendazole at this higher dose. Vacuolation of the different spermatogenic cell types and disruption of the spermatogonial layer on the basement membrane layer was seen as well. At the lower oral Oral dose Compound b 3 mg/kg Controla LND Not tested AF234 Not tested AF2785 Not tested Gamendazole a IP dose b mg/kg Not tested Not tested Not tested mg/kgb mg/kgb Control animals were administered vehicle alone (see Materials and Methods for details). b Values are mean SD.

6 1132 TASH ET AL. TABLE 4. Gamendazole-induced infertility in adult male rats: Response to a single oral dose. Treatment group No. males infertile/ no. males treated No. males recovering fertility/no. males rendered infertile No. normal implants/ pregnant rat a Vehicle control 0/ 1 1 Gamendazole 0.75 mg/kg 0/ mg/kg 2/ 2/ mg/kg 4/ 4/ a Mean SEM for Week 9 of the study. See Materials and Methods for details on the mating protocols and analysis of results. There were no significant differences (P ¼ 0.8) in the number of normal implantation sites per pregnant female among the treatment groups (ANOVA). after a single oral administration (Table 5). The mg/kg dose was the lowest dose tested that produced a 100% infertility rate after a single oral administration. The period of complete infertility lasted for 2 wk, followed by complete recovery of fertility in four of seven animals by wk. The remaining three animals in the group did not recover fertility. When administered at.0 mg/kg for seven consecutive days, a similar onset of infertility was observed; however, only two of seven animals recovered fertility during the 2-wk mating trial. In all treatment groups, the number of conceptuses and the proportion of abnormal conceptuses were not significantly different for the animals that recovered fertility compared with those of the control group. Pathology Results Five days after the single i.p. dose, animals were killed and gross necropsy performed. Pathology was examined in at least one randomly chosen animal per gamendazole or LND analogue treatment group (Table 2). Any animals that showed abnormal health or behavior effects also were examined in detail. All tissues were examined for histopathology; only affected tissues are reported. Following LND treatment at 200 mg/kg, some areas of the liver contained arteries and veins that were engorged with blood, whereas other areas did not appear to be congested. For the kidney, some engorgement of arteries and swollen arterioles was observed, with occasional large amounts of blood apparent between the proximal tubules. Engorged arteries and veins also were observed in the pancreas. These side effects usually were noted in at least one of the five animals at 200 mg/kg of LND. These side effects were not observed in animals dosed at 25 mg/kg; however, it should be noted that at 25 mg/kg, only approximately 50% of animals exhibited an inhibition of spermatogenesis. Following treatment with AF2785 at 25 mg/ kg, scattered areas near the periphery of the liver and pancreas contained distended, blood-filled sinuses. Evidence of necrosis, pyknotic nuclei, and swelling of the remaining cells in the area were observed. Following treatment with 25 mg/kg of AF234, small, isolated, necrotic nodules were observed near the periphery of the liver and pancreas. At 200 mg/kg, congestion and distension of most veins and some arteries in both the liver and pancreas were observed. After treatment with 25 mg/kg of gamendazole, histopathological findings in all organs of treated animals were nonremarkable, with no evidence of inflammation, necrosis, hemorrhage, or tumors. It should be noted that 100% of the animals treated with 25 mg/kg of gamendazole exhibited reduced spermatogenesis. At the higher dose of 200 mg/kg, gamendazole resulted in mortality in three of the five animals. In the two animals that survived, however, all of the organs were nonremarkable, with no evidence of inflammation, necrosis, hemorrhage, or tumor. In subsequent testing at lower doses down to mg/kg (oral), which gave a 100% infertility rate, all organs were nonremarkable, with no evidence of inflammation, necrosis, hemorrhage, or tumors. Body Weight and Reproductive Organ Changes after Treatment The animals used for antispermatogenic activity were 0 days of age, and vehicle-treated rats exhibited normal weight gain during the 5-day experiment (Fig. 5). At 25 mg/kg of LND, AF2785, or AF234, either a delay of 2 days was observed in body weight gain, followed by an increase that paralleled that of the control group or a slight decline for 2 days, followed by an increase that paralleled that of the control group. At 200 mg/kg of LND, a larger decline was observed in body weight for 2 days, followed by maintenance of the same weight for the remaining 3 days. Animals dosed with AF2785 or AF234 exhibited a delay in weight gain similar to that observed with 25 mg/kg. No delay was seen in body weight gain, and the rate of weight gain in gamendazoletreated animals was parallel to that of the controls for the entire 5-day period. Although the 25 mg/kg of gamendazole group started out at a slightly higher body weight, this weight was not significantly different from controls at any point. At 200 mg/kg, the surviving animals had a 2-day delay in weight gain, followed by an increase that paralleled the control animals. In contrast to the i.p. administration experiments described above, with parallel single oral administration experiments at 25 mg/kg, the body weights of the rats were not different among treatment groups (P. 0.05; data not shown). The rats from the two mating trial studies (Tables 4 and 5) were subjected to necropsy at study termination (Week 10 or 27), and the data on weights of reproductive organs and testicular endpoints were obtained. Data for only the 3 and mg/kg single-dose groups and their respective vehicle-control TABLE 5. Gamendazole induced infertility in adult male rats: Comparison of single dose versus seven daily doses. Treatment group No. males infertile/ no. males treated No. males recovering fertility/ no. males rendered infertile No. normal implants/ pregnant rat a Vehicle control 5 ml/kg per day for 7 days (Days 0 ) 0/ Gamendazole mg/kg on Day 0 (single dose) 7/7 4/ mg/kg per day for 7 days (Days 0 ) 7/7 2/ a Mean SEM for Week 18 of the study. See Materials and Methods for details on the mating protocols and analysis of results. There were no significant differences (P ¼ 0.80) in the number of normal implantation sites per pregnant female among the treatment groups (ANOVA).

7 NEW REVERSIBLE NONHORMONAL ORAL MALE CONTRACEPTIVE 1133 FIG. 5. Body weight changes following a single i.p. administration of LND, AF2785, AF234, and gamendazole. Animals were 0 days of age at the start of the experiment and were weighed on the day before compound administration, on the day of administration (Day 0, arrow), and both 2 and 5 days later. A) Compounds administered at 25 mg/kg. B) Compounds administered at 200 mg/kg. Controls received 5 ml/kg of carrier (buffered dimethyl sulfoxide). group are presented in Table. In the 3 mg/kg group, all the animals that were rendered infertile recovered fertility, and no significant difference was observed in testicular or epididymal weight compared with those in control animals. A 10% reduction was seen in spermatid counts and a 18% reduction in the proportion of normal seminiferous tubules 10 wk after the treatment even though all of these animals were fertile. In the mg/kg of gamendazole group, data are presented for all the animals as well as grouped by animals that recovered fertility versus those that remained infertile 27 wk after treatment. For all treated animals, all the male reproductive parameters were significantly lower in comparison to the control animals. When subgrouped by recovered-fertile versus infertile animals, however, all these parameters were still lower than those in vehicle-treated rats, but they were closer to control values. Histology of the testes from the animals used in the mating trials (Fig. ) revealed that the recovery of fertility in all of the animals treated with 3 mg/kg was matched with restoration of TABLE. Body, testis, and epididymal parameters after single oral dose of gamendazole. Mean paired epididymal weight (g) b Tubules with mature spermatids (%) b Paired testes weight Spermatid head count (g) b (no. cells x 10 per testis) b Treatment group Final body weight (Necropsy Week) N a (g) Vehicle control Week þ Week Gamendazole 3.0 mg/kg on Day 0 (Week 10) * 82 þ 2* * * ( vs ) 27 11* (45 13 vs 2 1) * ( vs ).0 mg/kg on Day 0 (Week 27) * ( vs ) a N ¼ Number of males. b For the Week 27 study ( mg/kg dose), the values in parenthesis represent mean SEM for the recovered fertile vs. infertile males, respectively (n ¼ 4 fertile; 3 infertile). * Significantly different (P, 0.05) from respective vehicle control group based on Student t-test.

8 1134 TASH ET AL. FIG.. Testicular histology in gamendazole-treated animals following mating trials. A and B) Testis from a control animal administered vehicle alone. C and D) Testis from an animal that was rendered infertile by 3 mg/kg of gamendazole but recovered fertility. E and F) Testis from an animal that was rendered infertile by mg/kg of gamendazole but recovered fertility. G and H) Testis from an animal that was rendered infertile by mg/kg of gamendazole and remained infertile. Bar ¼ 200 lm. normal spermatogenesis (Fig., C and D) similar to paired control (vehicle) animals (Fig., A and B). In the animals treated with mg/kg, those that recovered fertility also showed recovery of normal seminiferous tubule histology (Fig., E and F). The animals that failed to recover fertility showed not only a failure to repopulate the germinal epithelium, but also a loss of spermatogonia resulting in tubules with a Sertoli cell only phenotype (Fig., G and H). Behavior after Administration For 2 3 h after injection of LND, AF2785, or AF234 (25 or 200 mg/kg), the animals became lethargic. For AF2785, two of six animals died from seizures within 40 min. For gamendazole at 200 mg/kg only, three animals were killed after initial spasms that ceased by 20 min but were followed by labored breathing and a semiconscious state. In all of the 25 mg/kg and the remainder of the 200 mg/kg of gamendazole groups, as well as at all the oral doses tested, no evidence of lethargy (reduced alertness and awareness, lack of interest at cage opening, and prolonged periods of inactivity relative to controls) was found at either 25 or 200 mg/kg. In the mating trials described in Tables 4 and 5, in which gamendazole was administered orally, no differences in mating behavior were observed between control and gamendazole-treated animals at any of the doses tested.

9 NEW REVERSIBLE NONHORMONAL ORAL MALE CONTRACEPTIVE 1135 FIG. 7. Comparison of circulating inhibin B, FSH, and testosterone levels in proven-fertile male rats treated with vehicle or gamendazole at mg/kg. Gamendazole-treated rats were separated into animals that recovered fertility versus animals that remained infertile over a 27-wk period. A) Inhibin B: The limit of detection was 25 pg/ml. The human inhibin B standard was reconstituted in castrate rat serum; therefore, the results are relative to human inhibin B. B) FSH: The limit of detection was 1.5 ng/ml. C) Testosterone: The limit of detection was 0.04 ng/ml. Sertoli Cells Are a Target for Gamendazole It is important to note that no significant difference in circulating testosterone levels was observed between control and gamendazole-treated animals (Fig. 7C). In addition, gamendazole produced no significant differences in weights of the seminal vesicles and ventral prostate, two androgendependent organs (data not shown). Previous studies of LND analogues demonstrated that Sertoli cells are the primary target of the drug, leading to the loss of spermatids [33]. Analysis of additional serum hormone levels in the animals treated with mg/kg of gamendazole (Fig. 7A) demonstrated a temporary but significant decline in circulating inhibin B levels (P, 0.05, ANOVA-repeated measures followed by Holm-Sidak multiple comparison) that was matched with a transient but not significant increase in FSH levels (Fig. 7B). It also should be noted that inhibin B levels remained low in the animals that remained infertile after gamendazole treatment but rebounded after the initial drop in animals that recovered fertility. Serum FSH levels were inversely related to inhibin B levels, but increases were not significant. The effect of gamendazole on serum inhibin B levels suggested that gamendazole also may act in the testis via Sertoli cells [33], and this was confirmed by measuring inhibin B secretion in primary cultures of rat Sertoli cells incubated with varying concentrations of gamendazole (Fig. 8A). Indeed, inhibin B production in Sertoli cells was remarkably sensitive to gamendazole, with an IC 50 of only (SEM) M. Figure 8A presents the data from one of three replicate experiments used to determine the IC 50.By contrast, Sertoli cell ATP levels, a marker of cell viability, were

10 113 TASH ET AL. FIG. 8. Gamendazole responses and binding in primary Sertoli cells in vitro. Sertoli cells were incubated in the presence or absence of gamendazole at the concentrations indicated. The media were collected and assayed for inhibin B (A) after 72 h of treatment, and the cells for ATP content (B) after 48 hr of treatment. not reduced until gamendazole concentrations were 10 5 Mor greater (Fig. 8B). Thus, a five-log difference was observed between a physiologically significant biological marker of gamendazole action on Sertoli cells versus a toxic effect. Because gamendazole was derived from the cancer therapeutic agent LND, gamendazole also was examined, as an additional assessment of in vitro toxicity, for an effect of growth of a variety of ovarian cancer cell lines using the (3-(4,5- Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay. Growth of ID8-T1, ID8, SKOV3, and CAOV3 was inhibited by gamendazole, with IC 50 values of , , , and M, respectively. These IC 50 values were comparable to the gamendazole concentrations that caused ATP leakage from the primary Sertoli cell. This difference in dose for a biological versus toxic effect on cells in vitro is promising for further development of gamendazole. The spatial binding of gamendazole was demonstrated in cultured primary mouse Sertoli cells using GMZ-BT-UV (Fig. FIG. 9. Binding of gamendazole (GMZ-BT-UV) to mouse Sertoli cells in vitro. A) GMZ-BT-UV binding in formalin-fixed, detergent-lysed Sertoli cells. B) Binding in the presence of 10-fold excess gamendazole. C) Binding in the presence of 10-fold excess LND. D) Control incubation with fluorescein isothiocyanate-avidin alone. Bar ¼ 50 lm. 2). Staining was predominantly perinuclear (punctate staining that could be nucleolar), with less pronounced cytoplasmic staining (Fig. 9A). Control incubations with GMZ-BT-UV in the presence of a 10-fold molar excess gamendazole (Fig. 9B) or LND (Fig. 9C) blocked the localization, demonstrating that the pattern was compound specific. Nonspecific binding with fluorescein isothiocyanate-avidin D alone also was very low (Fig. 9D). These results suggest that Sertoli cells are a target of gamendazole action. DISCUSSION The present study presents initial dose-ranging results that suggest gamendazole should continue on the preclinical pathway as a small-molecule, male contraceptive drug candidate. In support of this goal, we report that gamendazole is the most potent analogue of LND synthesized thus far, producing a 100% infertility rate in rats after a single oral dose of only mg/kg and a 7% infertility rate after a single oral dose of 3 mg/kg. Furthermore, the recovery of fertility in the animals rendered infertile was 100% at 3 mg/kg and 57% at mg/kg. It also is important to note that the number and proportion of normal-appearing conceptuses in the animals that regained fertility were identical to those in control animals. By comparison, previous analogues of LND required multiple doses at mg/kg to induce complete infertility in the rat [2, 27]. Thus, gamendazole appears to be an improvement over other LND analogues; however, the single oral dose results reported here, although promising, also suggest that lower multiple dose-ranging experiments are necessary to determine if a wider therapeutic window exists that can produce the ideal 100% infertility rate but yield 100% recovery of fertility once the dosing regimen is terminated. It also must be emphasized that a near-100% return to fertility likely will be required for a male contraceptive drug intended for marketing to men who have not yet had their families. To approach these goals properly, additional dose-ranging experiments are planned as well as formulation, pharmacokinetics, pharmacodynamics, and other studies of the compounds. In general, an oral nonhormonal male contraceptive offers advantages over current hormone-based contraceptives, in that the latter generally require combinations of hormones and multiple routes of administration, including repeated oral doses plus injection of additional agents, a long period of administration

11 NEW REVERSIBLE NONHORMONAL ORAL MALE CONTRACEPTIVE 1137 before the onset of infertility, and a long lag period for the return of fertility [34]. Immunocontraceptive approaches also require injection and show significant variability, including a lack of immune response in a significant subset of individuals [, 8]. Another key aspect of gamendazole is the lack of side effects noted with previous LND analogues [12, 13]. In the studies reported here, LND and similar analogues produced lethargy, temporary declines in weight gain, and histopathology of liver and pancreas after administration in rats. In the present study, however, none of these observed side effects were noted after administration of gamendazole at doses sufficient to cause infertility. Mortality was observed at 200 mg/kg of gamendazole. Thus, the therapeutic window for a single dose of gamendazole, although better than that of LND, must be widened further. Because male contraceptives would be taken by otherwise healthy individuals, a lack of toxic side effects is essential and will require detailed and prolonged assessment before such compounds can be proposed for human use. Gamendazole may have a wider therapeutic window than LND, in that an approximately 1-fold difference exists between the i.p. IC 50 for an antispermatogenic effect versus the approximate LD 50 for LND and AF2785, as compared to a 7-fold difference for gamendazole (Table 2). As additional dose-ranging studies are carried out, however, our goal must be to obtain an acceptable therapeutic window that will allow the candidate compound to be carried into clinical trials. It should be noted that the level of reversibility following a single oral dose of gamendazole at mg/kg was 57%. A similar level of reversibility recently was reported using an immunocontraceptive approach []. Our finding that a single lower oral dose of gamendazole (3 mg/kg) produced a 7% reduction in fertility with a 100% recovery of fertility is promising. Efficacy combined with reversibility may be improved by using a lower dose at a repeated interval. The temporal pattern for loss of fertility in rats after a single challenge by gamendazole is consistent with the loss of spermatids as predicted by Russell et al. [35]. Given the duration of the spermatogenic cycle, the multiple low-dose option also is attractive, in that it is conceivable that a once-a-week or once-a-month oral male contraceptive pill could selectively block the spermatid stage of spermatogenesis, also allowing eventual repopulation of the testis after terminating compound administration [35]. Another important consideration for the nonhormonal approach to male contraception is a lack of significant longterm changes in circulating hormone levels. Serum testosterone levels were not affected by gamendazole treatment; however, a general decline in serum testosterone was observed over the study interval, a phenomenon that has been observed with increasing age [3, 37]. The lack of a treatment effect on circulating testosterone was consistent with maintenance of the androgen-dependent male accessory glands and the lack of treatment effects on mating behavior. In contrast, serum inhibin B levels dropped to nondetectable in gamendazole-treated rats and were associated with infertility. The suppression of circulating inhibin B, a Sertoli cell-specific product [38, 39], suggests that gamendazole acts directly on the Sertoli cell. This hypothesis is further supported by our in vitro work demonstrating direct binding of gamendazole to mouse Sertoli cells and suppression of inhibin B secretion from rat Sertoli cells in culture. Disruption of the Sertoli cell-spermatid junctional complexes occurs in response to LND analogues [33]. Indeed, the increased potency of gamendazole has allowed the confirmation of Sertoli cells as a target for the action of gamendazole in vitro and the identification of proteinbinding targets that are consistent with the observed premature loss of spermatids via disruption of these junctional complexes [40]. Circulating inhibin B levels recovered in rats that regained fertility, whereas inhibin B remained nondetectable in rats that remained infertile. The association between infertility and low serum inhibin B levels has been observed in men, and correlative data suggest that germ cells may mediate regulation of inhibin B production and release from the Sertoli cell [39, 41, 42]. Hence, the lack of recovery of inhibin B production in rats rendered irreversibly infertile is not surprising, but whether this phenomenon is the result or the cause of continued infertility is unknown. A slight increase in serum FSH levels was associated with suppression of inhibin B levels, consistent with the release from negative feedback by inhibin B [41]. Other factors, however, such as activins and other members of the transforming growth factor b superfamily (e.g., bone morphogenetic proteins), which have opposing actions to inhibin [43, 44], also may have impacted the release of FSH form the pituitary, because the increase in FSH levels was not statistically significant. A contraceptive approach that does not cause significant changes in endogenous hormones or rely on hormone or hormone analogue administration may offer advantages. For example, long-term administration of hormone-based contraceptives has raised concerns over undesirable cardiovascular, bone, and cancer risk side effects that have been observed with female hormone-based contraceptives [45]. Nonetheless, detailed preclinical testing and clinical trials will need to be undertaken before new nonhormonal contraceptive methods can become available. In conclusion, gamendazole has emerged from these initial studies showing improvements in efficacy and toxicity that suggest it or its future analogues should continue as male contraceptive candidates. Future research on gamendazole should include additional dose-finding experiments to widen the therapeutic window and increase reversibility as well as experiments to elucidate its precise mechanism of action [40]. In particular, continued development of gamendazole as a contraceptive agent will require more detailed toxicity studies. An efficacious male contraceptive that produces a 100% infertility rate after a single oral dose represents a new generation of potential male contraceptive agents, because gamendazole represents a new class of potent nonhormonal agents that act by inhibiting spermatogenesis. ACKNOWLEDGMENTS The authors wish to acknowledge the excellent technical support of Michael J. Wulser, Jennifer Hughes, Sotirios-E. Macheras, and summer medical students Adam Gregg, Melissa K. Emerson, and Brent Burroughs. Thanks to Stacy Wolfe and Vijayalaxmi Gupta, Ph.D., for photomicroscopy. Thanks to George Enders, Ph.D., and Ajay Nangia, M.D., for interpretation of the testicular histology. The authors also wish to thank of the following BIOQUAL technicians for their technical expertise: Janet Burgenson, David Gropp, Jessica Luke, Margaret Krol, Trung Pham, and Bruce Till. Thanks to E. Mitch Eddy, Ph.D. (National Institutes of Health/ National Institute of Environmental Health Sciences), and T. Raj Kumar, Ph.D. (University of Kansas Medical Center) for critical reading of the manuscript. Thanks to Dr. Hyun Kim (Contraception and Reproductive Health Branch, National Institute of Child Health and Human Development) for advice and support. REFERENCES 1. World Health Organization. The World Health Report 2005: Make every mother and child count. Geneva, Switzerland; 2005: James D. The costs of unintended pregnancy. Manag Care Interface 200; 19(suppl A): DePaolo LV, Hinton BT, Braun RE. Male contraception: Views to the 21st Century, Bethesda, MD, USA, 9 10 September Trends Endocrinol Metab 2000; 11: 8.