Luteinizing Hormone Receptor-Mediated Effects on Initiation of Spermatogenesis in Gonadotropin-Deficient (hpg) Mice Are Replicated by Testosterone 1

Transcription

1 BIOLOGY OF REPRODUCTION 70, (2004) Published online before print 3 September DOI /biolreprod Luteinizing Hormone Receptor-Mediated Effects on Initiation of Spermatogenesis in Gonadotropin-Deficient (hpg) Mice Are Replicated by Testosterone 1 Jennifer A. Spaliviero, Mark Jimenez, Charles M. Allan, and David J. Handelsman 2 Andrology Laboratory, ANZAC Research Institute, Concord Hospital, University of Sydney, Sydney, NSW 2139, Australia ABSTRACT Testosterone (T) is an absolute requirement for spermatogenesis and is supplied by mature Leydig cells stimulated by LH. We previously showed in gonadotropin-deficient hpg mice that T alone initiates qualitatively complete spermatogenesis bypassing LH-dependent Leydig cell maturation and steroidogenesis. However, because maximal T effects do not restore testis weight or germ cell number to wild-type control levels, additional Leydig cell factors may be involved. We therefore examined 1) whether chronic hcg administration to restore Leydig cell maturation and steroidogenesis can restore quantitatively normal spermatogenesis and testis development and 2) whether nonandrogenic Leydig cell products are required to initiate spermatogenesis. Weanling hpg mice were administered hcg ( IU i.p. injection three times weekly) or T (1-cm subdermal Silastic implant) for 6 weeks, after which stereological estimates of germinal cell populations, serum and testicular T content, and testis weight were evaluated. Human CG stimulated Leydig cell maturation and normalized testicular T content compared with T treatment where Leydig cells remained immature and inactive. The maximal hcg-induced increases in testis weight and serum T concentrations were similar to those for T treatment and produced complete spermatogenesis characterized by mature, basally located Sertoli cells (SCs) with tripartite nucleoli, condensed haploid sperm, and lumen development. Compared with T treatment, hcg increased spermatogonial numbers, but both hcg and T had similar effects on numbers of spermatocytes and round and elongated spermatids per testis as well as per SC. Nevertheless, testis weight and germ cell numbers per testis and per SC remained well below phenotypically normal controls, confirming the involvement of non-leydig cell factors such as FSH for quantitative normalization of spermatogenesis. We conclude that hcg stimulation of Leydig cell maturation and steroidogenesis is not required, and that T alone mostly replicates the effects of hcg, to initiate spermatogenesis. Because T is both necessary and sufficient for initiation of spermatogenesis, it is likely that T is the main Leydig cell secretory product involved and that additional LH-dependent Leydig cell factors are not essential for induction of murine spermatogenesis. leydig cells, luteinizing hormone, spermatogenesis, testis, testosterone INTRODUCTION Spermatogenesis is a highly complex and precisely regulated process of cellular division and differentiation gov- 1 This work was supported by the NHMRC. 2 Correspondence: FAX: ; djh@anzac.edu.au Received: 15 May First decision: 7 June Accepted: 18 August by the Society for the Study of Reproduction, Inc. ISSN: erned by two hormones, testosterone (T) and FSH. Whereas FSH acts directly on its Sertoli cell (SC) membrane receptor [1], T is produced within the testicular interstitial space by mature Leydig cells, which are stimulated and tightly regulated by LH [2, 3]. Leydig cells also secrete nonandrogenic products, including growth factors, cytokines, and vasoactive peptides, which contribute to the intercellular communication between Leydig, Sertoli, and peritubular cells [3, 4]. Although it has long been known that T is essential for the initiation and maintenance of spermatogenesis [5], recent genetic FSH-deficient models have shown that qualitatively normal mouse spermatogenesis can be initiated and maintained in the absence of FSH [6 9], although in each model spermatogenesis remains quantitatively subnormal. Any specific physiological role for such nonandrogenic Leydig cell products, however, remains to be established. Previous studies based upon inhibiting endogenous T secretion in mature rodents by gonadotropin withdrawal either by using testosterone and estradiol implants [10] or hypophysectomy [11] or by eliminating Leydig cells with ethane dimethanesulphonate (EDS) treatment [12] suggest that Leydig cell factors other than T are not required to restore spermatogenesis. These models, however, examined maintenance or reinitiation of spermatogenesis because germ cell development had been completed before experimental treatments. Furthermore, the uncertain influence of incomplete gonadotropin suppression, other pituitary hormones, and nonspecific effects of the cytotoxin EDS limits interpretation of these experiments. Whether Leydig cell factors other than T are essential for initiation of spermatogenesis has not been reported. Hypogonadal (hpg) mice have a congenital deficiency of hypothalamic gonadotropin-releasing hormone [13] because of a naturally occurring mutation causing a major deletion of the GnRH gene [14]. Consequently, hpg mice undergo normal male sexual differentiation (because of gonadotropin-independent fetal Leydig cell function) but become androgen deficient from birth [15] while remaining fully androgen responsive. This contrasts with tfm mice or rats, which are functionally androgen deficient from conception and exhibit a similar spermatogenic block at pachytene spermatocyte stage [16]. However, because of their androgen resistance, the effects of androgen treatment cannot be evaluated. In hpg mice, Leydig cells remain sparse ( 10% of wild type) and morphologically immature [17], resulting in gonads that do not undergo initiation of spermatogenesis [13]. Although initiation of qualitatively complete spermatogenesis in hpg mice can be achieved with exogenous T alone, testis weight and germ cell numbers remain 30% 40% of wild-type controls [6]. In contrast, after GnRH gene transfer [18] or transplantation of hypothalami containing GnRH neurons [19], hpg mice show restoration of full quantitative spermatogenesis. The present

2 hcg AND T INITIATION OF SPERMATOGENESIS 33 FIG. 1. Depicted are testis weights (upper left panel), serum T concentration (upper right panel), testicular T content (lower left panel), and testicular T concentration (lower right panel) for groups of hpg mice that were untreated (n 6 19) or treated with hcg at doses of 0 (saline, n 5 6), 0.1 (n 8), 0.3 (n 5), 1 (n 6), 10 (n 3 7), or 100 (n 7) IU i.p. injection daily or with a subdermal 1-cm T implant (n 7 15) in place for 6 weeks and for phenotypically normal littermates (n 5 12). The hcg treatment effect is maximal at the 1-IU dose for all four variables, producing similar effects to T treatment, although both effects of hcg and T treatments remained lower than phenotypically normal (nonhpg) littermates. Note logarithmic scale on y-axis for serum T concentration and testicular T content. Asterisk indicates significant difference from untreated hpg, and hash indicates significant difference from T treatment (P 0.05). Solid filled bars indicate hpg mice, and grayfilled histogram indicates phenotypically normal littermates. study was designed to determine whether restoration of LH activity (with hcg as a long-acting homolog of LH) leading to Leydig cell maturation and secretory activity could rectify the remaining quantitative deficiencies in testis development and germ cell numbers. This could also determine whether Leydig cell factors other than T are important in pubertal testis development and quantitative elaboration of germinal cell populations. MATERIALS AND METHODS Mice The hpg colony was housed under standard conditions at the University of Sydney and ANZAC Research Institute. Genotypes (N/N, N/hpg, hpg/ hpg) were assayed from tail DNA by polymerase chain reaction as described previously [6]. The Animal Ethics Committee of University of Sydney or CSAHS Animal Welfare Committee approved all experiments. Experimental Design Weanling (21-day old) hpg male mice (n 5 15 per group) were either injected (i.p.) with hcg (Pregnyl, Organon Australia Pty Ltd, Sydney, Australia) at a range of doses (0.1U 100 IU/mouse) three times weekly for 6 weeks or given a subdermal implant of T (Sigma, Sydney, Australia) in a 1-cm silastic implant (id 1.47 mm, od 1.95 mm, ends sealed with Silastic adhesive, Dow-Corning, Sydney, Australia) for 6 weeks. This T treatment produces maximal stimulation of spermatogenesis [6]. In addition, controls included groups of untreated and saline (vehicle)-injected mice and untreated phenotypically normal littermates. All experiments were terminated when mice reached 9 weeks of age. Tissue and Blood Collection Terminal blood was collected under anesthesia via cardiac puncture and serum stored frozen at 20 C. Body weight and the wet weight of fresh (right) and fixed (left) testes and fresh (right) and fixed (left) epididymis and spleen were recorded. The right testis was quickly removed, snap frozen in liquid nitrogen, and stored at 80 C. The mouse was then perfused via the left ventricle with warm (37 C) heparin and saline solution followed by 30 ml of Bouins fixative. The left testis was immersed in fixative overnight and subsequently processed and embedded in methacrylate resin by using the Technovit 7100 embedding kit (Heraeus, Germany). Sections were cut with a Reichter Jung microtome. Thin (7 m) sections stained with toluidine blue were used for photomicrographs and macroscopic histology. All procedures have been fully described previously [20]. Stereology Thick (25 m) sections were prepared for stereology from upper, lower, and middle sections of the testis and stained with periodic acid-schiff. Sertoli and spermatogenic cell types were classified according to Russell et al [21]. Sections were evaluated by using the optical disector technique and CAST software (Olympus, Copenhagen, Denmark) as described previously [20]. Hormone Assays Frozen testis was homogenized in phosphosaline and TritonX-100 buffer and extracted with hexane:ethyl acetate (3:2). Serum and testicular T were assayed by radioimmunoassay as described previously [22]. Undetectable T levels were assigned the detection limit of the assay for analysis. Human FSH was measured with a two-site immunoradiometric assay (DELFIA, Wallac, Turku, Finland) as described previously [22]. This FSH assay detects epitopoes on both subunits to measures intact, dimeric human FSH but does not cross-react with hcg (10 6 IU hcg is measured as 0.3 IU/l FSH). Antibodies to hcg were detected by specific binding of iodinated hcg. Briefly, hcg (Pregnyl, Organon Australia Pty Ltd) was iodinated with chloramine T (Sigma) and separated from free iodine with Con A Sepharose (Pharmacia, Sydney, Australia). Serum (100 l, diluted 1:100 in assay buffer) was incubated with either buffer or excess hcg (20 g per tube) and CPM I 125 hcg overnight (18 h) at room temperature with bound and free tracer separated by 20% polyethylene glycol precipitation. Statistics For analysis, testis and epididymis weight were corrected for body weight to adjust for differences in body size caused by litter size or treatment. Organ weights and T levels were log transformed prior to analysis to maintain equality of variance. Quantitative data from stereological analysis are presented as cells per testis or per SC. In some analyses, organ weights and hormone data that did not differ between untreated and salinetreated hpg mice were combined for subsequent data analysis. Data were analyzed by one-way ANOVA with suitable linear contrasts and Fishers LSD test for posthoc comparisons by using NCSS 2001 software (NCSS, Kaysville, UT). All data were expressed as mean SEM. RESULTS Testis and Organ Weight Testis weight (per body weight) was increased by all hcg doses in a dose-dependent fashion compared with un-

3 34 SPALIVIERO ET AL. FIG. 2. Testicular histology. Low-power cross-sections ( 10) of representative toluidine stained testis show phenotypically normal (nonhpg) littermate (A) with complete spermatogenesis, including mature seminiferous tubules with lumen formation (L) as compared with untreated hpg mice (B) which have immature undeveloped tubules with no lumen and incomplete spermatogenesis. Bar 100 m. Both hcg (C) and T (D) induced a similar degree of tubular maturation, with increased tubular diameter and lumen formation. Bar 50 m. Higher magnification ( 100) shows Leydig cell hyperplasia and hypertrophy in the interstitial space of hcg-treated testes (E) as compared with sparse immature Leydig cells in T-treated testes (F). Arrows indicate Leydig cell clusters in the interstitium. In contrast, both hcg-treated (G) and T-treated (H) testes show similar induction of tubular epithelial maturation with completion of spermatogenesis. Basal SCs with characteristic irregular nuclei are indicated with an asterisk. All stages of germ cell development are present, and arrowheads indicate mature haploid spermatozoa. Bar in (E H) 10 m. treated or saline-treated hpg mice ( [n 25] and [n 11], respectively, for pooled data) (Fig. 1). An hcg dose of 1 IU produced plateau stimulation of testis weight. Testis weight after hcg doses between 1 and 100 IU did not differ significantly from the effects of T treatment. After treatment with 1U hcg or T, however, testes remained at 31% and 24%, respectively, of phenotypically normal littermate controls. Neither body nor spleen weight were significantly different between groups treated with hcg or T (data not shown). Testosterone Human CG produced dose-dependent increases in serum and testicular T compared with untreated or saline-treated hpg mice (Fig. 1). For testicular T, plateau effects were achieved with 1 IU of hcg, and maximal hcg effects at doses of 1 IU or higher were similar to phenotypically normal littermate controls but significantly higher than the effects of T treatment. Serum T was stimulated by 1 IU hcg and higher doses with plateau effects produced by a dose of 1 IU hcg. All hcg doses of 1 IU or higher produced serum T similar to those after T treatment. Human CG Antibodies Antibodies to hcg were detected in serum of all hpg mice treated with 10 IU (median 8.9% specific tracer binding, range 1.2% 19%, n 7) and 100 IU (median 5.8% specific tracer binding, range 4.7% 7.6%, n 7) of hcg but not in any untreated hpg mice (n 6) or hpg mice treated with saline (n 6), 0.1 (n 8), 0.3 (n 5), or 1 (n 6) IU or in phenotypically normal littermate control (n 5) mice. There was no significant difference between

4 hcg AND T INITIATION OF SPERMATOGENESIS 35 FIG. 3. Total SC numbers in hpg testis. All hcg doses (0.1 U [n 5], 0.3 U [n 5], or 1 U [n 6]) and T-treated (n 15) mice showed similar SC numbers per testis with only a nonsignificant increase compared with untreated hpg mice (n 13). Controls were phenotypically normal littermates (n 7). Asterisk indicates significant difference from untreated hpg, and hash indicates significant difference from T treatment (P 0.05). Solid-filled bars indicate hpg mice, and gray-filled histogram indicates phenotypically normal littermates. hcg antibody titres in the 10- and 100-IU treatment groups. To avoid possible confounding effects of hcg immunoneutralization, stereological analyses of hcg effects were restricted to the 1 IU and lower doses. Follicle-Stimulating Hormone Pregnyl from the same batches used in the experiments contained no detectable FSH, with 1 IU hcg having less than IU human FSH immunoreactivity. Histology Control testis (Fig. 2A) demonstrated complete spermatogenesis with all tubules showing well-developed lumina compared with hpg testis (Fig. 2B), where spermatogenesis is arrested at the pachytene spermatocyte stage and seminiferous tubules have no lumen. Both 1 U hcg (Fig. 2C) and T (Fig. 2D) induced complete spermatogenesis. Tubules with well-formed lumen and similar tubular diameters were observed with both treatments, but they remained smaller than phenotypically normal controls. Human CG treatment produced hypertrophy and hyperplasia of Leydig cells in the interstitium (Fig. 2E), whereas with T treatment, Leydig cells remained immature and spindle shaped as indicated by arrows (Fig. 2F). Both hcg (Fig. 2G) and T (Fig. 2H) treatment led to maturation of the germinal epithelium, which was spatially organized into identifiable stages with mature basal SC nuclei (stars) and mature haploid sperm (arrows). Stereology SC. SC numbers (Fig. 3) were significantly increased by maximal hcg or T treatment compared with untreated hpg mice ( million per testis). There was no significant difference between maximal effects of hcg and T treatments ( million per testis vs million per testis), whereas SC numbers remained lower than in phenotypically normal littermate controls (55% and 52%, respectively). Germ cells. Spermatogonial (SG) numbers (Fig. 4) were increased by 1 IU hcg treatment, but lower doses (0.1, 0.3 IU) and T treatment had no effect compared with untreated controls. Maximal hcg effects produced significantly fewer SG than phenotypically normal littermate controls ( vs million per testis). For spermatocyte (Sc), round (RS) and elongated (ES), maximal hcg effect (1 IU) increased germ cell numbers to the same extent as T treatment (Fig. 4). Nevertheless, for both hcg and T treatment, each germ cell population remained lower than phenotypically normal littermate controls. Lower hcg doses produced small, nonsignificant increases in haploid cells (RS, ES), which were much lower than the 1 IU hcg dose. Germ cell:sc ratio. The mean number of germ cells supported by each SC was evaluated by the germ cell:sc ratio FIG. 4. Germ cell populations. Total number of spermatogonia (upper left panel), spermatocytes (upper right panel), round spermatids (lower left panel), and elongated spermatids (lower right panel) in hpg mice treated with 0.1U (n 5), 0.3 U (n 5),or1U(n 6) hcg or 1-cm T implant (n 15) for 6 weeks. Controls were untreated hpg (n 13) and phenotypically normal littermates (n 7). Treatment of hpg mice with hcg (1 IU) stimulated more spermatogonia but produced similar numbers of spermatocytes, round and elongated spermatids, compared with T treatment. Asterisk indicates significant difference from untreated hpg, and hash indicates significant difference from T treatment (P 0.05). Solid-filled bars indicate hpg mice, and gray-filled histogram indicates phenotypically normal littermates.

5 36 SPALIVIERO ET AL. FIG. 5. Germ cell to SC ratios. The number of spermatogonia per SC (upper left panel) was increased with hcg treatment compared with T treatment, but both hcg and T treatments had similar effects on the spermatocytes (upper right panel), round (lower left panel) and elongated (lower right panel) spermatids, to SC ratios. Neither hcg nor T treatment restored the germ cell to SC ratios to normal (N). Asterisk indicates significant difference from untreated hpg, and hash indicates significant difference from T treatment (P 0.05). Solid-filled bars indicate hpg mice, and gray-filled histogram indicates phenotypically normal littermates. for each germ cell type. For SG, the SG:SC ratio was significantly increased ( 2-fold) by 1 IU hcg treatment to levels significantly greater than those for T treatment (Fig. 5). For Sc, the ratio was increased by both 1 IU hcg and T treatment to a similar extent ( 7.5-fold). For the haploid germ cells, both RS (Fig. 5) and ES (Fig. 5) were undetectable in the untreated hpg mice but were increased to similar germ cell ratios by 1 IU hcg and T treatment. Lower hcg doses ( 1 IU) had lesser effects on all germ cell ratios. All ratios remained significantly lower than in phenotypically normal littermate controls. DISCUSSION The primary physiological role of mature Leydig cells is secretion of T, which is required for both testicular development and extra-gonadal androgenic effects. It has long been known that the normal development of mammalian spermatogenesis requires Leydig cell T production [5] and unhindered T action [16, 23]. It is also now well established that, in mice functionally deficient in FSH, T alone is sufficient to initiate [6] and maintain [9] mouse spermatogenesis. Similar effects of endogenous T exposure presumably explain the presence of qualitatively normal spermatogenesis in mice genetically deficient in either the FSH subunit [7] or the FSH receptor [8, 24]. Whether Leydig cells contribute anything other than T, however, to induction or maintenance of spermatogenesis remains debatable. Over the past decade, an increasingly large number of nonandrogenic products of the Leydig cells have been identified as possibly contributing to the local regulation of spermatogenesis. An incomplete list includes the hormones estradiol, inhibin, activin, oxytocin, GnRH, CRH, TRH, POMC, and -endorphin; growth factors IGF-1, bfgf, PDGF, and TGF and - ; cytokines IL-1 and IL-6; and vasoactive factors endothelin-1 and angiotensin-ii [25, 26]. Such nonandrogenic Leydig cell products might act upon peritubular, Sertoli, or interstitial vascular or neural cells, but no clear physiological mechanisms or specific intratesticular role of such local nonandrogenic Leydig cell products has been identified [3]. Although paracrine modulators of Leydig cell function are readily detected by the sensitive LH/hCG bioassay of Leydig cell testosterone production, functional bioassays for Leydig cell products influencing spermatogenesis are not available because of the complexity of cellular functions and interactions. A useful paradigm to evaluate putative nonandrogenic Leydig cell factors is whether spermatogenesis can be induced or maintained in the absence of functional Leydig cells. Hence, previous studies inactivating Leydig cell function by either LH deprivation via hypophysectomy [11], GnRH immunoneutralisation [27], or EDS-induced Leydig cell destruction [12] have suggested that T accounts for most or all LH-mediated actions. However, each of these model systems involved only the mature testis, so the evaluation was restricted to the maintenance or reinitiation of previously established spermatogenesis [10, 28]. In contrast, this study has found that hcg can induce qualitatively complete spermatogenesis in hpg mice, which are gonadotropin and androgen deficient from birth [15]. The maximal hcg effects on testis growth as well as absolute and relative germinal cell populations were very similar to those produced by T alone. Although there was a significantly greater effect of hcg than T on spermatogonial populations, all more advanced germ cell populations were affected similarly by hcg and T treatment. This implies that the main, or the only important, Leydig cell product for initiation of murine spermatogenesis is T. Caveats of this conclusion are that differences in germ cell dynamics may be overlooked, and that this pharmacological dosage cannot fully exclude other local effects having secondary modulatory effects. The differences in spermatogonial numbers may reflect similar observations of hcg effects on spermatogonial proliferation in the absence of FSH and not replicated fully by testosterone previously reported as due to Leydig cell secretions acting directly on spermatogonia [29] or via paracrine effects on SC secretions [30 32]. Conversely, in the present experiments the failure of hcg to fully restore quantitatively normal spermatogenesis any more than T can achieve indicates that either non-leydig cell factors or Leydig cell factors at times before weaning might be important in explaining the per-

6 hcg AND T INITIATION OF SPERMATOGENESIS 37 sisting deficit in testis growth and germinal development when either hcg or T are supplied from weaning onwards. The immunogenicity of higher hcg doses prevented an examination of their efficacy. This study utilized hcg as a long-acting, structurally homologous analog of LH. It binds to and activates the same LH receptor and exerts similar effects on Leydig cell maturation and steroidogenesis [2]. Human CG stimulates increased serum T concentrations within 2 h after administration and remains elevated for several days [33], allowing administration at longer intervals than daily. Prolonged hcg treatment can induce hyperplasia and hypertrophy of mature [34] and immature [35] Leydig cells, as well as those reappearing after EDS treatment [36]. In hpg mice, LH stimulates steroidogenesis and proliferation [17, 37 39]. One drawback of hcg is its immunogenicity as a foreign protein. In this study we observed that lower doses ( 1 IU) of hcg did not produce anti-hcg antibodies, whereas doses of 10 or 100 IU injected i.p. three times weekly for 6 weeks (total 18 doses) cause development of anti-hcg antibodies. This is consistent with a previous study that used doses of 10 or 100 IU and found that neutralizing anti-hcg antibodies developed consistently after 14 or more days but not after 7 daily injections [40]. Although the reason for the dose-dependent immunogenicity of hcg in mice is unclear, it has permitted an evaluation of hcg effects at doses up to 1 IU three times weekly for 6 weeks without interference by anti-hcg antibodies. Although Pregnyl is an extract of human urine, it is unlikely that contamination with FSH [41] or any intrinsic FSH-like bioactivity of hcg [42] would have influenced the experimental findings, because the FSH content of the highest dose (100 IU) of Pregnyl used contained IU FSH. By contrast, daily treatment with 1 IU of pure recombinant human FSH has no effect on testis weight [43]. The present study observed that, despite hcg-induced maturation of Leydig cell function as indicated by their morphology and intratesticular and serum T concentrations, there was no further increase in testis weight or haploid germ cell number compared with effects of T alone. Hence, hcg-stimulated Leydig cell maturation and secretory activity commencing at weaning still failed to restore testis weight or spermatogenesis to levels comparable with phenotypically normal control mice. The restoration of normal testis development in hpg mice after GnRH gene transfer [18] and fetal hypothalamic transplants [19] proves the defect in testis growth in hpg mice is not caused by any genetic limitation intrinsic to the testis. There are a number of possible explanations for the failure to rectify this deficit. First, other non-leydig cell factors such as FSH or paracrine factors influencing SCs may be necessary. The most important role of FSH for initiation of spermatogenesis may be the control of SC proliferation in the perinatal period [44 46]. Because each SC provides functional and structural support for a finite number of developing germ cells, the size of the SC population determines the final germ cell carrying capacity of the adult testis [46]. Transgenic mouse models expressing human FSH [20, 22] or an activating mutation of the human FSH receptor [47] in mouse SCs on an hpg background support the possibility that the functional deficiency of FSH is partly responsible. In these models, the addition of FSH effects on SCs, in conjunction with T, further rectifies the deficit in testis growth and spermatogenesis, though the net effect still falls short of wildtype mice. Despite a profusion of paracrine factors possibly involved in regulation of spermatogenesis, none have an established physiological role [25, 26]. A second caveat is that the timing or pattern of LH replacement may have important effects, which our experiment did not replicate. For example, postnatal LH effects before the time of weaning [48], when our experiments commenced, or effects that depend on the pulsatility of endogenous LH secretion [49, 50] were not replicated in our experiments. Finally, the hcg dose, limited by its immunogenicity, may have been insufficient to fully replicate LH effects. Both the early onset and the dose limitation of chronic LH stimulation could be addressed by constructing a transgenic model from the transgenic hcg mouse model as described elsewhere [51, 52] transferred onto the hpg background, similar to those we have used to characterize FSH effects [20, 22, 47]. Future studies of hcg effects in our FSH transgenic models may determine the magnitude of this interaction in the initiation of spermatogenesis. We conclude that hcg treatment can initiate qualitatively complete spermatogenesis in the gonadotropin-deficient hpg mouse. With the exception of spermatogonia, the testicular effects of the maximal hcg doses were identical to the maximal effects of T treatment on testis weight, SC numbers, and absolute and relative (to SC) germ cell populations, whereas hcg alone also stimulated Leydig cell maturation and normalized intratesticular T content. Yet, the functional maturation of Leydig cells or the contribution of any nonandrogenic Leydig cell factor(s) caused by hcg treatment cannot rectify the remaining quantitative deficit of spermatogenesis in the androgen or hcg-treated hpg mouse. This remaining deficit may be because of a lack of other regulators of spermatogenesis including FSH, paracrine intratesticular factors, or failure to fully replicate physiological patterns of LH secretion by hcg treatment commencing at weaning. Because T is both necessary and sufficient to induce spermatogenesis, and because it replicates all hcg effects, it is likely that T accounts for all hcg effects in the induction of spermatogenesis. ACKNOWLEDGMENTS The authors gratefully acknowledge the valued technical assistance of Julie Simpson and Fiona Thorn. REFERENCES 1. Simoni M, Gromoll J, Nieschlag E. The follicle-stimulating hormone receptor: biochemistry, molecular biology, physiology, and pathophysiology. Endocr Rev 1997; 18: Saez JM. Leydig cells: endocrine, paracrine, and autocrine regulation. Endocr Rev 1994; 15: Habert R, Lejeune H, Saez JM. Origin, differentiation and regulation of fetal and adult Leydig cells. Mol Cell Endocrinol 2001; 179: Gnessi L, Fabbri A, Spera G. Gonadal peptides as mediators of development and functional control of the testis: an integrated system with hormones and local environment. Endocr Rev 1997; 18: Steinberger E. Hormonal control of mammalian spermatogenesis. Physiol Rev 1971; 51: Singh J, O Neill C, Handelsman DJ. Induction of spermatogenesis by androgens in gonadotropin-deficient (hpg) mice. Endocrinology 1995; 136: Kumar TR, Wang Y, Lu N, Matzuk MM. FSH is required for ovarian follicle maturation but not for male fertility. Nat Genet 1997; 15: Dierich A, Sairam MR, Monaco L, Fimia GM, Gansmuller A, LeMeur M, Sassone-Corsi P. Impairing follicle-stimulating hormone (FSH) signalling in-vivo: targeted disruption of the FSH receptor leads to aberrant gametogenesis and hormonal imbalance. Proc Natl Acad Sci U S A 1998; 95: Handelsman DJ, Spaliviero JA, Simpson JM, Allan CM, Singh J.

7 38 SPALIVIERO ET AL. Spermatogenesis without gonadotropins: maintenance has a lower testosterone threshold than initiation. Endocrinology 1999; 140: Meachem SJ, Wreford NG, Robertson DM, McLachlan RI. Androgen action on the restoration of spermatogenesis in adult rats: effects of human chorionic gonadotrophin, testosterone and flutamide administration on germ cell number. Int J Androl 1997; 20: Sun YT, Wreford NG, Robertson DM, de Kretser DM. Quantitative cytological studies of spermatogenesis in intact and hypophysectomised rats: identification of androgen-dependent stages. Endocrinology 1990; 127: Sharpe RM, Fraser HM, Ratnasooriya WD. Assessment of the role of Leydig cell products other than testosterone in spermatogenesis and fertility in adult rats. Int J Androl 1988; 11: Cattanach BM, Iddon CA, Charlton HM, Chiappa SA, Fink G. Gonadotrophin-releasing hormone deficiency in a mutant mouse with hypogonadism. Nature 1977; 269: Mason AJ, Hayflick JS, Zoeller RT, Young WS, Phillips HS, Nikolics K, Seeburg PH. A deletion truncating the gonadotropin-releasing hormone gene is responsible for hypogonadism in the hpg mouse. Science 1986; 234: O Shaughnessy PJ, Baker P, Sohnius U, Haavisto AM, Charlton HM, Huhtaniemi I. Fetal development of Leydig cell activity in the mouse is independent of pituitary gonadotroph function. Endocrinology 1998; 139: Vanha-Perttula T, Bardin CW, Allison JE, Gumbreck LC, Stanley AJ. Testicular feminization in the rat: morphology of the testis. Endocrinology 1970; 87: Baker PJ, O Shaughnessy PJ. Role of gonadotrophins in regulating numbers of Leydig and Sertoli cells during fetal and postnatal development in mice. Reproduction 2001; 122: Mason AJ, Pitts S, Nikolics K, Szonyi E, Wilcox JN, Seeburg PH, Stewart TH. The hypogonadal mouse: reproductive functions restored by gene therapy. Science 1986; 234: Krieger DT, Perlow MJ, Gibson MJ, Davies TF, Zimmerman EA, Ferin M, Charlton HM. Brain grafts reverse hypogonadism of gonadotropin releasing hormone deficiency. Nature 1982; 298: Haywood M, Spaliviero J, Jimemez M, King NJ, Handelsman DJ, Allan CM. Sertoli and germ cell development in hypogonadal (hpg) mice expressing transgenic follicle-stimulating hormone alone or in combination with testosterone. Endocrinology 2003; 144: Russell LD, Ettlin RA, Hikim APS, Clegg ED. Histological and Histopathological Evaluation of the Testis. Clearwater, Florida: Cache River Press; Allan CM, Haywood M, Swaraj S, Spaliviero J, Koch A, Jimenez M, Poutanen M, Levallet J, Huhtaniemi I, Illingworth P, Handelsman DJ. A novel transgenic model to characterize the specific effects of follicle-stimulating hormone on gonadal physiology in the absence of luteinizing hormone actions. Endocrinology 2001; 142: Quigley CA, DeBellis A, Marschke KB, El-Awady MK, Wilson EM, French FF. Androgen receptor defects: historical, clinical and molecular perspectives. Endocr Rev 1995; 16: Abel MH, Wootton AN, Wilkins V, Huhtaniemi I, Knight PG, Charlton HM. The effect of a null mutation in the follicle-stimulating hormone receptor gene on mouse reproduction. Endocrinology 2000; 141: Spiteri-Grech J, Nieschlag E. Paracrine control of testicular function: a review. J Reprod Fertil 1993; 98: de Kretser DM, Loveland KL, Meinhardt A, Simorangkir D, Wreford N. Spermatogenesis. Hum Reprod 1998; 13: McLachlan RI, Wreford NG, Tsonis C, dekretser DM, Robertson DM. Testosterone effects on spermatogenesis in the gonadotropin-releasing hormone-immunized rat. Biol Reprod 1994; 50: McLachlan RI, Wreford NG, Meachem SJ, de Kretser DM, Robertson DM. Effects of testosterone on spermatogenic cell populations in the adult rat. Biol Reprod 1994; 51: Mather JP, Attie KM, Woodruff TK, Rice GC, Phillips DM. Activin stimulates spermatogonial proliferation in germ-sertoli cell cocultures from immature rat testis. Endocrinology 1990; 127: Huang HFS, Nieschlag E. Suppression of the intratesticular testosterone is associated with quantitative changes in spermatogonial populations in intact adult rats. Endocrinology 1986; 118: Soder O, Bang P, Wahab A, Parvinen M. Insulin-like growth factors selectively stimulate spermatogonial, but not meiotic, deoxyribonucleic acid synthesis during rat spermatogenesis. Endocrinology 1992; 131: Allard EK, Blanchard KT, Boekelheide K. Exogenous stem cell factor (SCF) compensates for altered endogenous SCF expression in 2,5- hexanedione-induced testicular atrophy in rats. Biol Reprod 1996; 55: Hodgson YM, de Kretser DM. The temporal response of rat testes to hcg stimulation during sexual maturation. Int J Androl 1984; 7: Christensen AK, Peacock KC. Increase in Leydig cell number in testes of adult rats treated chronically with an excess of human chorionic gonadotropin. Biol Reprod 1980; 22: Chemes HE, Rivarola MA, Bergada C. Effect of HCG on the interstitial cells and androgen production in the immature rat testis. J Reprod Fertil 1976; 46: Teerds KJ, de Rooij DG, Rommerts FF, van den Hurk R, Wensing CJ. Proliferation and differentiation of possible Leydig cell precursors after destruction of the existing Leydig cells with ethane dimethyl sulphonate: the role of LH/human chorionic gonadotrophin. J Endocrinol 1989; 122: Charlton HM, Halpin DMG, Iddon C, Rosie R, Levy G, McDowell IFW, Megson A, Morris JF, Bramwell A, Speight A, Ward BJ, Broadhead J, Davey-Smith G, Fink G. The effects of daily administration of single and multiple injections of gonadotropin-releasing hormone on pituitary and gonadal function in the hypogonadal (hpg) mouse. Endocrinology 1983; 113: Scott IS, Charlton HM, Cox BS, Grocock CA, Sheffield JW, O Shaughnessy PJ. Effect of LH injections on testicular steroidogenesis, cholesterol side-chain cleavage P450 mrna content and Leydig cell morphology in hypogonadal mice. J Endocrinol 1990; 125: O Shaughnessy PJ. Steroidogenic enzyme activity in the hypogonadal (hpg) mouse testis and effect of treatment with luteinizing hormone. J Steroid Biochem Mol Biol 1991; 39: Risbridger GP, Robertson DM, de Kretser DM. The effects of chronic human chorionic gonadotropin treatment on Leydig cell function. Endocrinology 1982; 110: Padmanabhan V, Chappel SC, Beitins IZ. An improved in vitro bioassay for follicle-stimulating hormone (FSH): suitable for measurement of FSH in unextracted human serum. Endocrinology 1987; 121: Chopineau M, Martinat N, Galet C, Guillou F, Combarnous Y. beta- Subunit residues are crucial to confer FSH activity to equine LH/CG but are not sufficient to confer FSH activity to human CG. J Endocrinol 2001; 169: Singh J, Handelsman DJ. The effects of recombinant FSH on testosterone-induced spermatogenesis in gonadotropin deficient (hpg) mice. J Androl 1996; 17: Orth JM. The role of follicle-stimulating hormone in controlling Sertoli cell proliferation in testes of fetal rats. Endocrinology 1984; 115: Orth JM. FSH-induced Sertoli cell proliferation in the developing rat is modified by beta-endorphin produced in the testis. Endocrinology 1986; 119: Orth JM, Gunsalus GL, Lamperti AA. Evidence from Sertoli-cell depleted rats indicates that spermatid number depends on numbers of Sertoli cells produced during perinatal development. Endocrinology 1988; 122: Haywood M, Tymchenko N, Spaliviero J, Koch A, Jimenez M, Gromoll J, Simoni M, Nordhoff V, Handelsman DJ, Allan CM. An activated human follicle-stimulating hormone (FSH) receptor stimulates FSHlike activity in gonadotropin-deficient transgenic mice. Mol Endocrinol 2002; 16: Mann DR, Gould KG, Collins DC, Wallen K. Blockade of neonatal activation of the pituitary-testicular axis: effect on peripubertal luteinizing hormone and testosterone secretion and on testicular development in male monkeys. J Clin Endocrinol Metab 1989; 68: Veldhuis JD, Iranmanesh A, Godschalk M, Mulligan T. Older men manifest multifold synchrony disruption of reproductive neurohormone outflow. J Clin Endocrinol Metab 2000; 85: Veldhuis JD, Pincus SM, Mitamura R, Yano K, Suzuki N, Ito Y, Makita Y, Okuno A. Developmentally delimited emergence of more orderly luteinizing hormone and testosterone secretion during late prepuberty in boys. J Clin Endocrinol Metab 2001; 86: Risma KA, Clay CM, Nett TM, Wagner T, Yun J, Nilson JH. Targeted overexpression of luteinizing hormone in transgenic mice leads to infertility, polycystic ovaries, and ovarian tumors. Proc Natl Acad Sci U S A 1995; 92: Matzuk MM, DeMayo FJ, Hadsell LA, Kumar TR. Overexpression of human chorionic gonadotropin causes multiple reproductive defects in transgenic mice. Biol Reprod 2003; 69: