Evidence for Increased Tissue Androgen Sensitivity in Neurturin Knockout Mice

U SIMANAINEN and others Neurturin actions on androgen 218:2 151–163 Research sensitivity

Evidence for increased tissue androgen sensitivity in neurturin knockout mice

Ulla Simanainen, Yan Ru (Ellen) Gao, Reena Desai, Mark Jimenez, Jennifer Spaliviero, Janet R Keast1,† and David J Handelsman Correspondence ANZAC Research Institute, University of Sydney, Sydney, New South Wales 2139, Australia should be addressed 1Kolling Institute of Medical Research, University of Sydney, New South Wales 2065, Australia to D J Handelsman †J R Keast is now at Department of Anatomy and Neuroscience, University of Melbourne, Melbourne, Email Victoria 3010, Australia [email protected]

Abstract

Neurturin (NTN) is a member of the glial cell line-derived neurotrophic factor (GDNF) family Key Words and signals through GDNF family receptor alpha 2 (GFRa2). We hypothesised that epithelial " glial cell line-derived atrophy reported in the reproductive organs of Ntn (Nrtn)- and Gfra2 (Gfra2)-deficient mice neurotrophic factor (GDNF) family could be due to NTN affecting the hormonal environment. To investigate this, we compared " TGFb family the reproductive organs of Ntn- and Gfra2-deficient male mice in parallel with an analysis of " neurotrophic their circulating reproductive hormone levels. There were no significant structural changes " hypothalamic–pituitary– within the organs of the knockout mice; however, serum and intratesticular testosterone testicular axis and serum LH levels were very low. To reconcile these observations, we tested androgen " urogenital tract

Journal of Endocrinology sensitivity by creating a dihydrotestosterone (DHT) clamp (castration plus DHT implant) to create ﬁxed circulating levels of androgens, allowing the evaluation of androgen-sensitive endpoints. At the same serum DHT levels, serum LH levels were lower and prostate and seminal vesicle weights were higher in the Ntn knockout (NTNKO) mice than in the wild-type mice, suggesting an increased response to androgens in the accessory glands and hypothalamus and pituitary of the NTNKO mice. Testicular and pituitary responsiveness was unaffected in the NTNKO males, as determined by the response to the human chorionic gonadotrophin or GNRH analogue, leuprolide, respectively. In conclusion, our results suggest that NTN inactivation enhances androgen sensitivity in reproductive and neuroendocrine tissues, revealing a novel mechanism to inﬂuence reproductive function and

the activity of other androgen-dependent tissues. Journal of Endocrinology (2013) 218, 151–163

Introduction

Neurotrophic factors control the development of periph- 2007, Navarro et al. 2007). They also have important eral nerve circuits by promoting the survival of sensory effects on neurophysiological properties. For example, in and motor neurons and targeting their axons to form nociceptive sensory neurons, many neurotrophic factors functional connections (Huang & Reichardt 2003, da Silva promote sensitisation, which underlies inﬂammatory and & Wang 2011). Neurotrophic factors and their receptors neuropathic pain states (Pezet & McMahon 2006). continue to be expressed in adults and are important Neurturin (NTN) is a member of the glial cell mediators of axon regeneration after injury (Chen et al. line-derived neurotrophic factor (GDNF) family of

http://joe.endocrinology-journals.org Ñ 2013 Society for Endocrinology Published by Bioscientiﬁca Ltd. DOI: 10.1530/JOE-13-0056 Printed in Great Britain Downloaded from Bioscientifica.com at 09/25/2021 10:35:06AM via free access Research U SIMANAINEN and others Neurturin actions on androgen 218:2 152 sensitivity

neurotrophic factors (Buj-Bello et al. 1995, Kotzbauer et al. The initial goal of our study was to determine whether 1996, Airaksinen & Saarma 2002) that is produced by many NTN maintains epithelial integrity of male accessory sex peripheral tissues, including the reproductive organs organs by regulating their hormonal environment. To (Golden et al. 1999, Xian et al. 1999, Widenfalk et al. investigate this, we compared the reproductive organs of 2000, Meng et al. 2001). The GDNF family of factors Ntn knockout (NTNKO) and Gfra2 knockout (GFRa2KO) mediate their actions through a receptor complex consist- mice, in parallel with an analysis of their circulating sex ing of a ligand, a high-affinity binding component, a GPI- steroid hormone levels. The results from this study show linked GDNF family receptor alpha (GFRa) and a common an unexpected mismatch between androgen-dependent signalling component, a receptor tyrosine kinase, Ret accessory sex organs and androgen environment, prompt- (Airaksinen & Saarma 2002). From four GFRa subunits, ing us to investigate whether these mice had the ability to GFRa2 functions as a specific NTN receptor, and mice respond normally to hormonal stimulation. Our results deficient in Ntn (Nrtn)orGfra2 (Gfra2) bear similar also show that NTN deprivation leads to a heightened phenotypes (Heuckeroth et al. 1999, Rossi et al. 1999, response to androgens. This reveals a novel and powerful Wanigasekara et al. 2004). way for NTN to influence not only reproductive function The best-known action of NTN on reproduction is via but also the activity of other androgen-sensitive tissues. It its trophic effects on pelvic parasympathetic neurons also provides an additional mechanism by which NTN (Wanigasekara et al. 2004, Yan & Keast 2008). These may influence androgen-sensitive peripheral neurons that regulate the function of male reproductive organs. autonomic neurons are essential for initiation of penile erection and stimulation of glandular secretion. For example, penile erection is highly impaired in adult Materials and methods mice deficient in Ntn or its receptor, Gfra2, largely due to Mouse colonies and sample collection many fewer parasympathetic axons making connections with the cavernosal vessels during the early postnatal The generation of NTNKO and GFRa2KO mice has been period (Laurikainen et al. 2000, Nangle & Keast 2006). described elsewhere (Heuckeroth et al. 1999, Rossi et al. NTN also has powerful direct trophic effects on adult 1999). The mice were genotyped as described previously pelvic parasympathetic neurons (Laurikainen et al. 2000, (Rossi et al. 1999, Enomoto et al. 2000). The mice were Wanigasekara & Keast 2005, Bella et al. 2007), so reduced killed by cardiac exsanguination under anaesthesia

Journal of Endocrinology NTN signalling in adulthood would further impair (60 mg/kg ketamine and 10 mg/kg xylazine, i.p.), and signalling to their target organs. serum was stored frozen at K20 8C. Male reproductive Studies of Ntn-orGfra2-deficient mice have also organs analysed included prostate lobes, seminal vesicles revealed epithelial atrophy of male accessory sex organs (SVs) and testes. Prostate lobes (ventral prostate (VP), (Wanigasekara et al. 2004), and although it is generally dorsolateral prostate (DLP) and anterior prostate (AP)), SVs believed that the mechanism may be neurotrophic, the and testes were dissected free of fat and connective tissue specific mechanism of this defect has not been defined. In and weighed separately. Non-reproductive organs (kidney, these mice, the parasympathetic innervation of glandular spleen and heart) were also weighed. Prostate tissues tissue fails to develop properly, so it is possible that a loss were either snap–frozen with liquid nitrogen and stored at K of normal, regular neuronal activation leads to cellular 80 8C for RNA extraction or fixed in Bouin’s solution loss. However, NTN may have a more direct action on for 4 h at room temperature for studying histology. these and other non-neuronal tissues. This is supported by The NTNKO and GFRa2KO males were compared with the expression of NTN, GFRa2 and Ret in the testis, their respective wild-type (WT) littermates. All procedures prostate, pituitary, and hypothalamus (Widenfalk et al. were approved by the Animal Care and Ethics Committee of 1997, 2000, Golden et al. 1998, 1999, Xian et al. 1999, the Royal North Shore Hospital and University of Sydney. Meng et al. 2001). Moreover, there is a transient disruption of spermatogenesis in NTN-overexpressing mice (Meng Experimental design et al. 2001), NTN stimulates DNA synthesis in spermatogonia (Viglietto et al. 2000) and GFRa2-positive cells Experiment 1: intact males Sexually mature (12G4 are important for the differentiation of secretory (meanGS.D.) weeks of age) homozygous NTNKO and pituitary cells including the gonadotrophin-secreting GFRa2KO male mice were compared with their respective cells (Garcia-Lavandeira et al. 2009). WT littermates. Endpoints analysed were serum

testosterone and gonadotrophins as well as weights of male Signal was visualised using a Vectastain Elite anti-rabbit reproductive organs and selected non-reproductive organs ABC Kit (Vector Laboratories, Burlingame, CA, USA) (see above). according to the manufacturer’s instructions. Colour development was done using the 3,30-diaminobenzidine Experiment 2: androgen sensitivity Androgen tetrahydrochloride chromogenic substrate (Dako Australia sensitivity in the male mice was determined following Pty Ltd, Campbellfield, VIC, Australia). Negative control castration to remove the endogenous source of androgens sections were incubated with non-immune rabbit IgG and treatment with fixed androgen levels produced (Santa Cruz Biotechnology). The sections were counter- by dihydrotestosterone (DHT) implants (DHT clamp; stained with Harris haematoxylin. Simanainen et al. (2011)). DHT was chosen as a Cell proliferation was determined using a proliferating non-aromatisable androgen. The WT and NTNKO male cell nuclear antigen (PCNA) kit (Zymed, San Francisco, CA, mice were orchidectomised via scrotal sac under USA). This kit provides ready-to-use biotinylated anti-PCNA ketamine/xylazine anaesthesia (anaesthetic doses as primary antibodies, which were used according to the given above) and treated by subdermal implantation of manufacturer’s instructions. Microwave antigen retrieval 0.5 cm silastic tubing filled with w5 mg crystalline DHT was done using 10 mM citric acid buffer (pH 6) for 10 min at (Rossi et al. 1999, Simanainen et al. 2009) for 7 days. high power. Negative control sections were incubated with Treatment for 7 days was chosen based on previous non-immune mouse IgG (Santa Cruz Biotechnology). experience in analysing androgen sensitivity using the Mousesmallintestinesectionswereusedaspositive same type of DHT implants (Simanainen et al. 2011). controls. Prostate epithelial proliferating cell index and Endpoints analysed were serum DHT, 3a-diol and 3b-diol prostate epithelial height were quantified using the CAST- and gonadotrophins as well as male reproductive organ GRID V1.10 (Olympus Corp., Albertslund, Denmark) weights (organs listed above). software as described previously (Simanainen et al.2007). Mean tubular diameter per testis was determined from Experiment 3: testicular and pituitary fixed testes embedded in methacrylate resin, sectioned responsiveness The WT and NTNKO male mice and stained with 0.5% toluidine blue as described were administered an i.p. injection of 10 IU human previously. Mean tubular diameter per testis was calcu- chorionic gonadotrophin (hCG; Pregnyl, Organon Pty lated using the Cast software (Olympus) to trace the Ltd, Lane Cove, NSW, Australia) in saline to analyse the Journal of Endocrinology perimeter of 80–100 tubular cross sections, where each testicular responsiveness to LH/hCG (Ingman & Robertson tubular diameter was derived from the formula PZpd. 2007)orwith1mg of the GNRH analogue, leuprolide (Lucrin, Abbott), to analyse the pituitary and testicular responsiveness to GNRH stimulation (Handelsman et al. Hormone assays 1985a, Bergh & Damber 1988). Serum and tissues were collected 1 h after injection. Endpoints analysed Serum Mouse serum LH levels were measured by an were serum LH (leuprolide) and testosterone (hCG immunofluorometric assay as described previously and leuprolide). (Jimenez et al. 2005), using specific antibodies for mLH as validated previously (Simanainen et al. 2011). Mouse serum FSH levels were determined using a specific immunofluoro- Histology and immunohistochemistry metric assay as described previously (Jimenez et al. 2005). Testes were embedded in resin, and representative All assays were performed in a single batch. sections were analysed after 0.5% toluidine blue staining Serum levels of testosterone, DHT, 3a-diol and 3b-diol of 5 mm sections. Prostates were embedded in paraffin, cut were measured in extracts of 100 mlofmouseserumbyliquid into 5 mm sections and stained with haematoxylin and chromatography–tandem mass spectrometry (LC–MS/MS; eosin (H&E). Harwood & Handelsman 2009) as adapted for mouse serum Immunohistochemistry for AR was performed on (McNamara et al.2010). The quantitation limits for 5 mm dewaxed paraffin sections. Rabbit anti-AR (N-20; testosterone, DHT, 3a-diol and 3b-diol were 20, 100, 400 1:100 dilution; Santa Cruz Biotechnology, Inc.) antibody and 400 pg/ml respectively. The diols were measured was used to detect AR (Mulholland et al. 2011). separately, but presented as a sum of 3a-diol and 3b-diol. Microwave-induced (high-power) antigen retrieval was The androgen sensitivity index (ASI) was calculated as performed with 0.01 M citrate buffer, pH 6, for 12 min. a product of serum LH and testosterone (Hiort et al. 2000)

with the increased androgen sensitivity demonstrated by a expressed as median and 95% CIs for serum hormone lower ASI value. levels due to high variability of serum hormones in adult male mice and as mean and S.E.M. for others unless Testis Intratesticular testosterone was analysed by otherwise specified. P values !0.05 were considered stable isotope dilution LC–MS/MS (Harwood & Handels- statistically significant. man 2009) as adapted for mouse serum and reproductive tissue (McNamara et al. 2010). Briefly, whole frozen testis was homogenised in 500 ml PBS buffer (containing Results 0.5% BSA (w/v) and 5 mM EDTA, pH 7.4) and centrifuged Experiment 1: intact males (2000 g 10 min, 4 8C) to separate insoluble debris, and then 200 ml of supernatant were extracted with Body weights and organ weights Adult NTNKO 1 ml hexane:ethyl acetate (3:2 ratio) fortified with and GFRa2KO males were slightly (w10%) but signi- testosterone-1,2,3-d3 (d3-T) as the internal standard. The ficantly (PZ0.041) smaller than their respective WT organic layer, separated by freezing the aqueous layer, was littermates (WT vs KO (meanGS.D.): NTN, 27.1G2.7 vs dried and reconstituted in 1.2 ml of 20% methanol in PBS 25.6G2.6 g; GFRa2, 30.3G2.4 vs 27.1G2.5 g), so organ prior to injection onto the C8 column for analysis (1 ml). weights are reported relative to body weights. The relative In a separate experiment using radiolabelled steroid as a testis weight was slightly increased (PZ0.044) in both the tracer, no detectable steroid losses or metabolism was NTNKO and GFRa2KO males compared with the respect- detected when the testis homogenate was processed for ive WT controls (Fig. 1A), while the weights of epididymis analysis (data not shown), eliminating the need for a and SVs were not statistically different between the correction factor so that data are expressed as ng/testis. genotypes (Fig. 1B and C). The weights of the separated prostate lobes were not significantly different in the RNA extraction and real-time RT-PCR NTNKO and GFRa2KO males when compared with their For RT-PCR and real-time RT-PCR, total RNA was extracted respective WT littermates (Fig. 1D, E and F). The weights using the RNeasy Plus Mini Kit (Qiagen) and cDNA was of the non-reproductive organs (kidney, spleen and heart) synthesised with Omniscript reverse transcriptase in the NTNKO or GFRa2KO males did not differ from that (Qiagen) from 250 ng of total RNA using oligo dT of the non-reproductive organs in their respective WT

Journal of Endocrinology (Invitrogen Australia Pty). The ﬁnal RT reaction mixtures littermates (Fig. 1G, H and I). The results for absolute were diluted 1:5 for storage at K20 8C. Quantitative real- organ weights (data not shown) were directionally similar time RT-PCR analyses for androgen receptor (Ar), MP25 to the body weight relative organ weights. (androgen-dependent, VP marker (Lin et al. 2002)), Histology Testis morphology was not affected by NTN cyclophilin and b-actin were performed on cDNA using or GFRa2 inactivation and was similar among the NTNKO, the QuantiTect SYBR Green PCR kit (Qiagen) and Rotor- GFRa2KO and WT males, with each testis demonstrating Gene 2000 System (Corbett Research, Mortlake, NSW, complete spermatogenesis and tubules containing Australia) as described previously (Simanainen et al. 2007). elongated spermatids (Fig. 2A) and qualitatively normal Primer sequences, product size and annealing tempera- Leydig cells (Supplementary Figure 1A and B, see section tures were as described previously (Lin et al. 2002, on supplementary data given at the end of this article). Simanainen et al. 2007). b-Actin primers were provided in Tubular diameter per testis was further analysed in the WT the SABiosciences RT2 Real-Time PCR kit (Jomar and NTNKO males and was found to be similar between Bioscience Pty, Kensington, SA, Australia). the genotypes (Supplementary Figure 1C). Prostate morphology under light microscopy with H&E staining Statistical analysis appeared to be normal in the NTNKO and GFRa2KO males Statistical analysis was performed using two-way ANOVA when compared with the respective WT males (Fig. 2B, with treatment groups and genotype as the ﬁxed main shown for ventral lobe only), and the epithelial height was effects. In case of non-homogenous variances (according similar between the genotypes (Supplementary Figure 1D). to Levene’s test, P!0.01), the non-parametric Kruskal– Prostate epithelial cell proliferation as analysed by Wallis ANOVA was used, followed by the Mann–Whitney PCNA-positive epithelial cells (as percentage of epithelial U test. The statistical analysis was performed using SPSS, cells counted) was similar in the NTNKO or GFRa2KO Inc. and NCSS (Kaysville, UT, USA) software. Data are males when compared with the WT males (Fig. 3A).

A Testis D Ventral prostate G Kidney

6 0.4 12 WT GFRα2 WT NTN

GFRα2KO * * 0.3 9 NTNKO 4 0.2 6 2 0.1 3 mg/g body weight 0 0 0

B Epididymis E Dorsolateral prostate H Spleen 1.5 0.8 4

0.6 3 1.0 0.4 2 0.5 0.2 1

0 0 0

C Seminal vesicle F Anterior prostate I Heart 10 1.5 6 8 1.0 4 6 4 0.5 2 2 mg/g body weight mg/g body weight 0 0 0

Journal of Endocrinology Figure 1 Relative (mg/g body weight) weights of the (A) testis, (B) epididymis, WT (WT GFRa2 and WT NTN) littermates. Data are shown as meanGS.E.M. (C) seminal vesicles, (D) ventral prostate, (E) dorsolateral prostate and Numbers of mice were 4 for the GFRa2KO males and 4 for the respective WT (F) anterior prostate as well as the non-reproductive organs (G) kidney, males and 14 for the NTNKO males and 20 for the respective WT males. (H) spleen and (I) heart in the GFRa2KO and NTNKO males and their respective *Signiﬁcantly different from that of the respective WT males (P!0.05).

Molecular analysis AR protein expression in prostate observed for serum testosterone levels, intratesticular epithelia appeared to be similar in the KO and WT males testosterone levels were significantly reduced in both the (Supplementary Figure 1E), and the Ar mRNA levels in the NTNKO (PZ0.012) and GFRa2KO (PZ0.05) males prostate (VP) were not significantly modified by GFRa2or compared with their respective WT littermates (Fig. 4B). NTN inactivation (Fig. 3B). Similarly, the mRNA levels of Similarly, intratesticular levels of DHT and DHT metab- androgen-dependent MP25 were not significantly olites 3a-diol and 3b-diol were significantly lower in the modified by GFRa2 or NTN inactivation (Fig. 3C). NTNKO and GRFa2KO males when compared with their The expression of Ar and MP25 was quantified relative to respective WT littermates (Supplementary Figure 2, see that of cyclophilin and b-actin as housekeeping genes. section on supplementary data given at the end of this The results were not affected by the housekeeping genes, article). Serum DHT levels were low but detectable in all and the data are shown relative to cyclophilin. the males and not significantly affected by the genotype

(Fig. 4C). Serum and intratesticular E2 was undetectable in Steroid hormone and gonadotrophin levels all the males. Serum testosterone was detectable in all the males, but Serum LH levels were significantly lower than the its levels were significantly reduced in both the NTNKO respective WT levels in both the GFRa2KO (PZ0.012) and (PZ0.013) and GFRa2KO (PZ0.029) males compared with NTNKO (PZ0.001) males (Fig. 4D), while serum FSH levels their respective WT littermates (Fig. 4A). As has been were not significantly affected by either NTN or GFRa2

A Testis histology WT NTNKO GFRα2KO

B Prostate histology

WT NTNKO GFRα2KO

Figure 2 Representative images of (A) testis and (B) prostate (ventral lobe) histology in the WT, NTNKO and GFRa2KO males. No major differences were observed in the cellular structure of the testis or prostate lobes.

inactivation (Fig. 4E). The ASI was markedly reduced in both whereas the NTNKO males demonstrated much greater the GFRa2KO (PZ0.029) and NTNKO (PZ0.024) males serum LH suppression, to less than half that of the intact when compared with their respective WT littermates WT males (Fig. 6A). The LH response to castration with or (Fig. 4F), indicating increased androgen sensitivity. without DHT treatment was signiﬁcantly (PZ0.043) depen- Journal of Endocrinology dent on the genotype (WT vs NTNKO), with greater suppression being observed in the NTNKO males (signiﬁcant Experiment 2: DHT clamp (androgen sensitivity) interaction between genotype and treatment; PZ0.005).

Serum steroid hormone and gonadotrophin Organ weights Prostate and SV weights were signi- levels Androgen sensitivity was further examined in ficantly influenced by the treatments (P!0.001) and the NTNKO males using the DHT clamp, where the mice genotype (PZ0.03), with a significant interaction being were castrated to remove endogenous androgens and then observed between genotype and treatment (P!0.01). provided with DHT implants to deliver comparable non- Castration reduced prostate and SV weights compared with aromatisable androgen levels. DHT treatment following those of the intact mice, whereas DHT treatment following castration resulted in similar circulating DHT levels as well castration increased the weights in both the WT and NTNKO as levels of two main DHT metabolites, 3a-diol and 3b-diol males (Fig. 6B and C). In both the castrated and DHT-treated (combined as diols), in the NTNKO and WT males (Fig. 5A castrated males, the prostate and SVs were significantly and B), demonstrating the validity of the DHT clamp. heavier in the NTNKO males than in the WT males (Fig. 6B Following castration with or without DHT implant, serum and C). All prostate lobes responded similarly to castration or testosterone was non-detectable (Fig. 5A and B). to castration and DHT treatment (data not shown). Serum LH levels were significantly influenced by the treatments (P!0.001). Castration increased serum LH levels almost twofold from intact WT levels in the WT and NTNKO Experiment 3: pituitary and testicular responsiveness males (Fig. 6A). DHT treatment following castration suppressed the elevated post-castration LH levels to levels Serum steroid hormone and gonadotrophin comparable to those of the intact males in the WT mice, levels To determine the responsiveness of the pituitary

A Epithelial Discussion proliferation index 6 WT GFRα2 The GDNF family of neurotrophic factors has well-known GFRα2KO roles in the differentiation and maintenance of the 4 WT NTN nervous system (Airaksinen & Saarma 2002). On the NTNKO other hand, while GDNF signalling is important for kidney 2 development and spermatogenesis (Sariola & Saarma PCNA positive % Epithelial cells 2003), the role of other GDNF families of neurotrophic 0 factors outside the nervous system is not well understood. In humans and rodents, the prostate gland has a rich B Ar expression 6 sympathetic and parasympathetic innervation (Vaalasti & Hervonen 1979, 1980), and previous rodent studies of

4 autonomic denervation have suggested signiﬁcant (Martinez-Pineiro et al. 1993, Lujan et al. 1998, Diaz et al. 2010) but complex and ill-deﬁned roles of autonomic 2 expression nerves in the regulation of glandular structure and Relative mRNA Relative function. Therefore, the original aim of this study was 0 to utilise transgenic mouse models lacking either a C MP25 expression 80 A Serum testosterone D Serum LH

20 WT GFRα2 1.5 60 GFRα2KO 15 WT NTN 1.0 NTNKO 40 10 * ng/ml ng/ml 0.5 * 5 expression 20 * * Relative mRNA Relative 0 0 0 B Testicular testosterone E Serum FSH 0.6 8 6 0.4 Journal of Endocrinology Figure 3 (A) Prostate epithelial proliferation index (percentage of epithelial cells that are 4 0.2 ng/ml PCNA positive) in the NTNKO, GFRa2KO and respective WT (WT GFRa2andWT 2 NTN) males as analysed by stereology. (B and C) Real time RT-PCR analysis of ng/mg testis * * 0 Ar (B) and androgen-dependent MP25 (C) mRNA expression relative to 0 cyclophilin mRNA expression in the ventral prostate of the GFRa2KO and C Serum DHT F ASI (LH × testosterone) NTNKO males and their respective WT littermates. Data are shown as 3 20 meanGS.E.M. Numbers of mice were four each for the GFRa2KO, NTNKO and 15 respective WT males. 2 10 ng/ml 1 and testis to GNRH analogue stimulation as well as the 5 testicular steroidogenic response to LH, the WT and 0 0 * * NTNKO males were administered the GNRH analogue, leuprolide, or hCG respectively. Serum testosterone levels ! Figure 4 were significantly influenced by the treatments (P 0.001) (A) Serum testosterone (ng/ml), (B) intratesticular testosterone (ng/mg and genotype (P!0.01), while serum LH levels were testis) and (C) serum DHT (ng/ml) concentrations as well as serum (D) LH significantly influenced by the treatments only (ng/ml) and (E) FSH (ng/ml) concentrations for the GFRa2KO and NTNKO a ! males and their respective WT (WT GFR 2 and WT NTN) littermates. Data (P 0.001). The administration of hCG reduced (31 and are shown as medianG95% CI. (F) The ASI was determined by the product 11% compared with intact WT levels in the WT and of absolute values for serum testosterone and LH multiplied with each G NTNKO males respectively) serum LH levels and increased other. Data are shown as mean S.E.M. For testosterone and DHT, numbers of mice were 4 for the GFRa 2KO and WT GFRa2 males, 12 for the NTNKO (790 and 1125% compared with intact WT levels in the males and 15 for the WT NTN males. For intratesticular testosterone, WT and NTNKO males respectively) serum testosterone numbers of mice were four for the GFRa 2KO and WT GFRa2 males, seven levels in both the NTNKO and WT males, while that of for the NTNKO males and seven for the WT NTN males. For serum LH and FSH as well as for ASI, numbers of mice were four for the GFRa2KO and WT leuprolide caused a marked increase in serum LH and GFRa2 males, eight for the NTNKO males and seven for the WT NTN males. testosterone levels (Fig. 7). *Significantly different from that of the respective WT males (P!0.05).

A Serum DHT B Serum diols rats following preganglionic parasympathectomy, but a

1.0 WT NTN 8 greater reduction following preganglionic sympath- NTNKO 0.8 6 ectomy (McVary et al. 1994). These differential effects 0.6 may also be related to the targeting of each group of axons, 4 ng/ml 0.4 ng/ml with parasympathetic nerves mainly innervating the 2 0.2 ND ND ND ND epithelium and sympathetic axons innervating the 0 0 smooth muscle (Bruschini et al. 1978, Wang et al. 1991). Cast Cast+DHT Cast Cast+DHT NTN and GFRa inactivation has previously been confirmed to significantly modify the parasympathetic Figure 5 Serum levels (ng/ml) of (A) DHT and (B) main DHT metabolites 3a-diol and innervations of the prostate with the NTNKO and GFRaKO 3b-diol (combined as diols) in the NTNKO and respective WT (WT NTN) male mice having an almost complete loss of VIP C males following castration (Cast) or castration with DHT (Cast DHT) terminals innervating the epithelium of the reproductive treatment. Data are shown as medianG95% CI. NDZlevels under the LOQ. Numbers of mice were five each for the NTNKO and WT males for CastC organs (Wanigasekara et al. 2004). DHT and six and eight for the NTNKO and WT males for Cast respectively. In contrast to the grossly normal development of androgen-dependent accessory sex organs, the circulating ligand (NTNKO) or a receptor (GFRa2KO) for the NTN and intratesticular testosterone and DHT levels were signalling pathway to characterise the role and markedly reduced (by 83–97%) in both the NTNKO and signalling mechanisms of parasympathetic nerves in GFRaKO mice. The significantly reduced circulating prostate development. androgen levels together with reduced circulating LH While only a minor role for NTN signalling was levels and grossly normal weights of androgen-dependent revealed in prostate development, we found a novel role accessory sex organs suggested that the inactivation of for NTN via GFRa2 in the regulation of tissue androgen NTN signalling led to increased responsiveness to andro- sensitivity. The inactivation of NTN signalling had only a gens. This was subsequently confirmed directly using the modest effect on prostate growth with statistically normal DHT clamp. prostate weights, although the VP and DLP weights were While the striking reduction in circulating and consistently lower than those of the respective WT mice in intratesticular testosterone levels is suggested to be due both the NTNKO and GFRaKO mice. A previous study to reduced Leydig cell synthesis and secretion of testoster-

Journal of Endocrinology using these mice has suggested atrophy of the prostate one, it may be attributable to either intrinsic defects in epithelial layer. However, following a more detailed non- Leydig cell steroidogenesis or reduction in Leydig cell biased stereological analysis of the epithelial height, we stimulation drive by circulating LH. An intact hypo- did not ﬁnd major differences in the histological appear- thalamic–pituitary–testicular axis (HPT axis) would ance or height of prostate epithelium between the WT and normally respond to low circulating testosterone levels KO prostates. The minor reduction observed in the present with an increase in pituitary gonadotrophin secretion due study supports a relatively minor role of parasympathetic to a lack of androgenic negative feedback. Yet, in the activity in prostate growth, suggested previously by a NTNKO and GFRa2KO males, both of which lacked NTN small decrease in prostate weight observed in pre-pubertal signalling, the very low circulating androgen levels were

A Serum LH B Prostate C Seminal vesicle 3 3 12 WT NTN NTNKO 2 2 8 mg mg ng/ml 1 1 4

0 0 0 Intact Cast Cast+ Intact Cast Cast+ Intact Cast Cast+ DHT DHT DHT

Figure 6 Changes in (A) serum LH levels (ng/ml) and (B) prostate (combined prostate intact WT males. By two-way ANOVA, serum LH levels as well as prostate lobes) or (C) seminal vesicle weights for the intact NTNKO and respective and SV weights were signiﬁcantly inﬂuenced by the treatments (P!0.001) WT (WT NTN) males or following castration (Cast) or castration and DHT and genotype (P%0.043). Data are shown as medianG95% CI for LH levels (CastCDHT) treatment. Dotted line compares the values with those of the and as meanGS.E.M. for organ weights, nR5.

A Serum LH GFRaKO males (Widenfalk et al. 2000). The preservation of 4 spermatogenesis despite markedly reduced circulating and WT NTN intratesticular androgen levels is consistent with the fact NTNKO 3 that intratesticular testosterone levels required for induction (Singh et al. 1995)ormaintenance 2 (Handelsman et al. 1999) of murine spermatogenesis are ng/ml much lower than ambient levels prevailing in the testis or 1 bloodstream with analogous findings reported in rats despite a significantly reduced number of seminiferous 0 tubules and interstitial fluid testosterone levels (Zirkin Intact hCG Leuprolide et al. 1989). Hence, increased androgen sensitivity in the B Serum testosterone testis of the NTNKO mice could support the maintenance 60 of normal spermatogenesis despite markedly lowered

circulating and intratesticular levels. Although E2 can induce murine spermatogenesis (Ebling et al. 2000)in 40 gonadotrophin-deﬁcient hpg mice by a mechanism requir- ing a functional AR (Lim et al. 2008) and stimulation of ng/ml 20 pituitary FSH secretion (Allan et al. 2010), serum E2 was not detectable in either the WT or KO mice by LC–MS/MS consistent with our previous ﬁndings (McNamara et al. 0 Intact hCG Leuprolide 2010). Hence, the remarkable maintenance of spermatogenesis in this study is unlikely to involve E2 due to the

absence of sufficiently increased circulating FSH and E2 Figure 7 Changes in serum (A) LH levels (ng/ml) and (B) testosterone levels (ng/ml) in levels involved in E2 induction of spermatogenesis. the intact NTNKO and respective WT (WT NTN) males or following injection Similarly, it is unlikely that another unknown testicular with hCG or GNRH analogue leuprolide. Dotted line compares the values androgen could explain the presence of normal sperma- with those of the intact WT males. By two-way ANOVA, serum testosterone levels were significantly influenced by the treatments (P!0.001) and togenesis despite serum and testicular testosterone genotype (P!0.01), while serum LH levels were significantly influenced by and DHT concentrations that, together with reduced Journal of Endocrinology ! G the treatments only (P 0.001). Data are shown as median 95% CI. serum LH levels, indicate inactive Leydig cell steroido- Numbers of mice were 10 and 12 for intact, 6 and 5 for hCG-treated and 8 and 7 for leuprolide-treated NTNKO and WT males respectively. genesis, the only plausible source of such a hypothetical unknown androgen. Analogous findings of histologically K K normal testis in Tgfb1 (Tgfb1)KO(Tgfb1 / ) mice despite accompanied by reduced serum LH levels but unchanged 95% reduction in intratesticular testosterone levels serum FSH levels. This implies impaired hypothalamic– compared with the WT mice (Ingman & Robertson 2007) pituitary LH secretion and/or an enhanced negative have been reported. testosterone feedback set point (Aiman et al.1979, To directly test androgen responsiveness, we used an Handelsman et al. 1985a, Hiort et al. 2000, Simanainen open-loop DHT clamp, whereby in castrated mice, steady- et al. 2011). The preserved HPT responses to stimulation state DHT delivery stabilises circulating androgen levels, with hCG (testicular responsiveness) and the GNRH ana- allowing for the direct examination of the suppression logue leuprolide (pituitary and testicular responsiveness) of pituitary gonadotrophin secretion and androgen- exclude any significant functional defect in the testis dependent organ weights. While castration reduced the or pituitary. These findings together with largely circulating androgen (testosterone and DHT) levels to non- unchanged prostate and SV weights and AR expression in detectable levels (compared with detectable androgen levels the NTNKO and GFRa2KO males, despite marked lowering inintactmales),DHTtreatmentproducedsimilarcirculating of circulating serum testosterone (and LH) levels, DHT levels in both the WT and NTNKO males, thus supported increased responsiveness to low levels of validating the DHT clamp. DHT-induced suppression of circulating androgens. serumLHlevelswassignificantlygreaterintheNTNKOmales The testis appeared to be structurally normal with all when compared with the WT males. Similarly, the response stages of spermatogenesis being present. This is consistent of the prostate and SVs to DHT was significantly greater in with the retained fertility reported in mature NTNKO and the NTNKO males, despite them having similar prostate

Ar expression. While AR expression appeared to be normal (Widenfalk et al. 1997, Golden et al. 1998, Japon et al. 2002), in target tissues such as the prostate, we did not look at the while the expression of GFRa2 can be observed in the AR co-factors that could be modified by NTN. Interestingly, developing pituitary gland and in the hypothalamus the differences between the WT and NTNKO males appeared (Widenfalk et al. 1997, Golden et al. 1998, Xian et al. to remain even in the androgen-deprived environment of 1999). In the hypothalamus, it is suggested that NTN in the castrated males. Comparable observations of changes supraoptic and paraventricular nuclei may be a source of in androgen responsiveness have been reported previously NTN for a number of neuronal populations that express in uraemic rats (Handelsman et al.1985b)aswellasin NTN receptors and have efferent projections to these humanised mice with variable-length CAG repeats in the hypothalamic nuclei, including neurons in the medial AR (Simanainen et al.2011). It is speculated that the latter preoptic nucleus (Golden et al. 1998). NTN, GFRa2 and Ret might be an adaptive phenotype termed ontogenic are present in the adult murine testis (Golden et al. 1999, regression (Handelsman & Dong 1992, Handelsman 2004), Viglietto et al. 2000, Wong et al. 2002), where they are for example, by epigenetic modifications, with the promoter expressed in the testicular Sertoli cells as well as in germ activity of androgen-sensitive genes being enhanced to cells at spermatogenic stages IX–XII and I–II of the cycle maintain the same activity as ARs operating at higher (Golden et al. 1999, Widenfalk et al. 2000, Meng et al. 2001). sensitivity (Zhang & Ho 2011). GFRa2 and Ret are also expressed in the testis (Cao et al. NTN, GDNF, artemin and persephin comprise a group 1996).ThepresenceofNTNsignallingcomponents of structurally related neurotrophic factors belonging to throughout the HPT axis, together with our data demon- the transforming growth factor-b (TGFb) superfamily strating the effects of NTN inactivation in intact males (Airaksinen & Saarma 2002). Therefore, it is interesting (testicular responsiveness) and following DHT clamp to consider our results in relation to a recent study (pituitary and reproductive organ responsiveness), suggests K K characterising Tgfb1 / mice (Ingman & Robertson that the effect of NTN may be related to AR signalling rather 2007). Comparable to the NTNKO and GFRaKO males in than being tissue specific. This could be at the level of gene the present study, these mice had significantly reduced transcription or signal transduction. serum LH and testosterone levels but normal weight and The loss of circulating androgens with an increase in morphology of androgen-dependent accessory sex organs. tissue androgen sensitivity in Ntn-deficient animals also This suggests that the tissue response to androgens may has important implications for the peripheral nervous K K also be increased in Tgfb1 / males. The similar pheno- system. Many of the sensory and autonomic neurons Journal of Endocrinology K K types of the NTNKO and Tgfb1 / males suggest either innervating the reproductive organs express ARs, which disruption of a common signalling pathway or possible mediate diverse effects on the growth and physiological cooperation between NTN and TGFb signal transduction. properties of these neurons in developing, postnatal and The latter is supported by the finding that TGFb and GDNF mature animals (Melvin & Hamill 1987, 1989, Keast 2006, cooperate in various steps of GDNF signalling, including Brock et al. 2007, Purves-Tyson et al. 2007). Previous GFRa membrane localisation (Peterziel et al. 2002). To our studies in NTN and GFRa2 gene KO mice have revealed a knowledge, this potential interaction of TGFb and NTN deficit in the innervation of reproductive organs by signalling has not been studied. In addition, while GDNF cholinergic nitrergic neurons, closely matching the regulates spermatogenesis as demonstrated by both expression patterns of GFRa2 within the pelvic ganglia GDNF-KO and -overexpressing mice (Meng et al. 2000), (Laurikainen et al. 2000, Wanigasekara et al. 2004). While biologically significant crosstalk between NTN and GDNF it is reasonable to ascribe at least some of this deficit to the is unlikely to be relevant due to significantly different absence of NTN signalling, it is also possible that some testicular phenotypes between the GDNF- and NRTN- aspects of the phenotype are obscured by concomitant overexpressing mice as well as between the KO mice actions on androgen signalling in these neurons. To our (Meng et al. 2001). knowledge, lumbosacral sensory neurons and pelvic NTN signalling is mediated by the binding of NTN to autonomic neurons have not been investigated as a site GFRa2 and their subsequent interaction with Ret, the of signalling crosstalk between androgens and the GDNF extracellular domain. One or all members of NTN family of neurotrophic factors. The temporal patterns of signalling are expressed in the HPT axis, allowing direct secretion of these factors by different tissues and organs biological effects. NTN and Ret can be detected in adult within the urogenital tract have not been determined, but pituitary gland and in the large secretory neurons of they could exert a significant influence on androgen supraoptic and paraventricular nuclei of the hypothalamus actions in this system.

In conclusion, we have reported a novel role of Bruschini H, Schmidt RA & Tanagho EA 1978 Neurologic control of neurotrophic factors in the regulation of male hormonal prostatic secretion in the dog. Investigative Urology 15 288–290. Buj-Bello A, Buchman VL, Horton A, Rosenthal A & Davies AM 1995 GDNF environment by modifying the tissue responsiveness to is an age-specific survival factor for sensory and autonomic neurons. low levels of circulating androgens. We have demon- Neuron 15 821–828. (doi:10.1016/0896-6273(95)90173-6) strated that Ntn deficiency leads to a significantly Cao T, Shannon M, Handel MA & Etkin LD 1996 Mouse ret finger protein (rfp) proto-oncogene is expressed at specific stages of mouse increased response to DHT treatment following castration, spermatogenesis. Developmental Genetics 19 309–320. (doi:10.1002/ suggesting increased tissue androgen sensitivity. This (SICI)1520-6408(1996)19:4!309::AID-DVG4O3.0.CO;2-D) increased responsiveness to circulating androgen levels Chen ZL, Yu WM & Strickland S 2007 Peripheral regeneration. Annual Review of Neuroscience 30 209–233. (doi:10.1146/annurev.neuro.30. could, therefore, impact not only reproductive function 051606.094337) but also hormone-sensitive diseases. Diaz R, Garcia LI, Locia J, Silva M, Rodriguez S, Perez CA, Aranda-Abreu GE, Manzo J, Toledo R & Hernandez ME 2010 Histological modifications of the rat prostate following transection of somatic and autonomic nerves. Anais da Academia Brasileira de Cieˆncias 82 397–404. Supplementary data (doi:10.1590/S0001-37652010000200015) This is linked to the online version of the paper at http://dx.doi.org/10.1530/ Ebling FJ, Brooks AN, Cronin AS, Ford H & Kerr JB 2000 Estrogenic JOE-13-0056. induction of spermatogenesis in the hypogonadal mouse. Endocrinology 141 2861–2869. (doi:10.1210/en.141.8.2861) Enomoto H, Heuckeroth RO, Golden JP, Johnson EM & Milbrandt J 2000 Development of cranial parasympathetic ganglia requires sequential Declaration of interest actions of GDNF and neurturin. Development 127 4877–4889. The authors declare that there is no conflict of interest that could be Garcia-Lavandeira M, Quereda V, Flores I, Saez C, Diaz-Rodriguez E, perceived as prejudicing the impartiality of the research reported. Japon MA, Ryan AK, Blasco MA, Dieguez C, Malumbres M et al. 2009 A GRFa2/Prop1/stem (GPS) cell niche in the pituitary. PLoS ONE 4 e4815. (doi:10.1371/journal.pone.0004815) Golden JP, Baloh RH, Kotzbauer PT, Lampe PA, Osborne PA, Milbrandt J & Funding Johnson EM Jr 1998 Expression of neurturin, GDNF, and their receptors This study was supported by the University of Sydney Bridging Grant (U S), in the adult mouse CNS. Journal of Comparative Neurology 398 139–150. NHMRC Project Grants 1006582 (U S), 402627 (J R K) and 570877 (J R K), and (doi:10.1002/(SICI)1096-9861(19980817)398:1!139::AID-CNE9O3.0. Senior Research Fellowship 632903 (J R K). CO;2-2) Golden JP, DeMaro JA, Osborne PA, Milbrandt J & Johnson EM Jr 1999 Expression of neurturin, GDNF, and GDNF family-receptor mRNA in the developing and mature mouse. Experimental Neurology 158 Acknowledgements 504–528. (doi:10.1006/exnr.1999.7127) Ms Keely McNamara and Mr Shai Joseph are thanked for their technical Handelsman DJ 2004 Reversible hypogonadism: ontogenic reversion in Journal of Endocrinology help. physiology, pathology and ageing. In: Hormone Interventions in Adult Life, Proceedings of the 8th International HypoCCS Symposium, 93–110. Handelsman DJ & Dong Q 1991 Ontogenic regression: a model of stress References and reproduction. In Stress & Reproduction, vol 86, pp. 333–345. Aiman J, Griffin JE, Gazak JM, Wilson JD & MacDonald PC 1979 Eds Sheppard KE, Boublik JH, Funder JW. Serono Symposia Androgen insensitivity as a cause of infertility in otherwise normal Publications, Raven Press. men. New England Journal of Medicine 300 223–227. (doi:10.1056/ Handelsman DJ, Spaliviero JA & Turtle JR 1985a Hypothalamic–pituitary NEJM197902013000503) function in experimental uremic hypogonadism. Endocrinology 117 Airaksinen MS & Saarma M 2002 The GDNF family: signalling, biological 1984–1995. (doi:10.1210/endo-117-5-1984) functions and therapeutic value. Nature Reviews. Neuroscience 3 Handelsman DJ, Spaliviero JA & Turtle JR 1985b Testicular function in 383–394. (doi:10.1038/nrn812) experimental uremia. Endocrinology 117 1974–1983. (doi:10.1210/ Allan CM, Couse JF, Simanainen U, Spaliviero J, Jimenez M, Rodriguez K, endo-117-5-1974) Korach KS & Handelsman DJ 2010 Estradiol induction of spermato- Handelsman DJ, Spaliviero JA, Simpson JM, Allan CM & Singh J 1999 genesis is mediated via an estrogen receptor-a mechanism involving Spermatogenesis without gonadotropins: maintenance has a lower neuroendocrine activation of follicle-stimulating hormone secretion. testosterone threshold than initiation. Endocrinology 140 3938–3946. Endocrinology 151 2800–2810. (doi:10.1210/en.2009-1477) (doi:10.1210/en.140.9.3938) Bella AJ, Fandel TM, Tantiwongse K, Brant WO, Klein RD, Garcia CA & Harwood DT & Handelsman DJ 2009 Development and validation of a Lue TF 2007 Neurturin enhances the recovery of erectile function sensitive liquid chromatography–tandem mass spectrometry assay to following bilateral cavernous nerve crush injury in the rat. Journal of simultaneously measure androgens and estrogens in serum without BrachialPlexus and Peripheral NerveInjury2 5.(doi:10.1186/1749-7221-2-5) derivatization. Clinica Chimica Acta 409 78–84. (doi:10.1016/j.cca. Bergh A & Damber JE 1988 Treatment with an LHRH agonist or hCG 2009.09.003) increases interstitial fluid volume and permeability to Evans blue in the Heuckeroth RO, Enomoto H, Grider JR, Golden JP, Hanke JA, Jackman A, mouse testis. International Journal of Andrology 11 449–456. Molliver DC, Bardgett ME, Snider WD, Johnson EM Jr et al. 1999 Gene (doi:10.1111/j.1365-2605.1988.tb01017.x) targeting reveals a critical role for neurturin in the development and Brock JA, Handelsman DJ & Keast JR 2007 Postnatal androgen deprivation maintenance of enteric, sensory, and parasympathetic neurons. Neuron dissociates the development of smooth muscle innervation from 22 253–263. (doi:10.1016/S0896-6273(00)81087-9) functional neurotransmission in mouse vas deferens. Journal of Hiort O, Holterhus PM, Horter T, Schulze W, Kremke B, Bals-Pratsch M, Physiology 581 665–678. (doi:10.1113/jphysiol.2007.128728) Sinnecker GH & Kruse K 2000 Significance of mutations in the

androgen receptor gene in males with idiopathic infertility. Journal of expression of neurturin in testis. Molecular and Cellular Endocrinology Clinical Endocrinology and Metabolism 85 2810–2815. (doi:10.1210/jc. 184 33–39. (doi:10.1016/S0303-7207(01)00649-9) 85.8.2810) Mulholland DJ, Tran LM, Li Y, Cai H, Morim A, Wang S, Plaisier S, Huang EJ & Reichardt LF 2003 Trk receptors: roles in neuronal signal Garraway IP, Huang J, Graeber TG et al. 2011 Cell autonomous role of transduction. Annual Review of Biochemistry 72 609–642. (doi:10.1146/ PTEN in regulating castration-resistant prostate cancer growth. Cancer annurev.biochem.72.121801.161629) Cell 19 792–804. (doi:10.1016/j.ccr.2011.05.006) Ingman WV & Robertson SA 2007 Transforming growth factor-b1 null Nangle MR & Keast JR 2006 Loss of nitrergic neurotransmission to mouse mutation causes infertility in male mice associated with testosterone corpus cavernosum in the absence of neurturin is accompanied by deficiency and sexual dysfunction. Endocrinology 148 4032–4043. increased response to acetylcholine. British Journal of Pharmacology 148 (doi:10.1210/en.2006-1759) 423–433. (doi:10.1038/sj.bjp.0706760) Japon MA, Urbano AG, Saez C, Segura DI, Cerro AL, Dieguez C & Navarro X, Vivo M & Valero-Cabre A 2007 Neural plasticity after peripheral Alvarez CV 2002 Glial-derived neurotropic factor and RET gene nerve injury and regeneration. Progress in Neurobiology 82 163–201. expression in normal human anterior pituitary cell types and in (doi:10.1016/j.pneurobio.2007.06.005) pituitary tumors. Journal of Clinical Endocrinology and Metabolism 87 Peterziel H, Unsicker K & Krieglstein K 2002 TGFb induces GDNF 1879–1884. (doi:10.1210/jc.87.4.1879) responsiveness in neurons by recruitment of GFRa1 to the plasma Jimenez M, Spaliviero JA, Grootenhuis AJ, Verhagen J, Allan CM & membrane. Journal of Cell Biology 159 157–167. (doi:10.1083/jcb. Handelsman DJ 2005 Validation of an ultrasensitive and specific 200203115) immunofluorometric assay for mouse follicle-stimulating hormone. Pezet S & McMahon SB 2006 Neurotrophins: mediators and modulators of Biology of Reproduction 72 78–85. (doi:10.1095/biolreprod.104.033654) pain. Annual Review of Neuroscience 29 507–538. (doi:10.1146/annurev. Keast JR 2006 Plasticity of pelvic autonomic ganglia and urogenital neuro.29.051605.112929) innervation. International Review of Cytology 248 141–208. Purves-Tyson TD, Arshi MS, Handelsman DJ, Cheng Y & Keast JR 2007 Kotzbauer PT, Lampe PA, Heuckeroth RO, Golden JP, Creedon DJ, Androgen and estrogen receptor-mediated mechanisms of testosterone Johnson EM Jr & Milbrandt J 1996 Neurturin, a relative of action in male rat pelvic autonomic ganglia. Neuroscience 148 92–104. glial-cell-line-derived neurotrophic factor. Nature 384 467–470. (doi:10.1016/j.neuroscience.2007.05.043) (doi:10.1038/384467a0) Rossi J, Luukko K, Poteryaev D, Laurikainen A, Sun YF, Laakso T, Laurikainen A, Hiltunen JO, Thomas-Crusells J, Vanhatalo S, Arumae U, Eerikainen S, Tuominen R, Lakso M, Rauvala H et al. 1999 Retarded Airaksinen MS, Klinge E & Saarma M 2000 Neurturin is a neurotrophic growth and deficits in the enteric and parasympathetic nervous factor for penile parasympathetic neurons in adult rat. Journal system in mice lacking GFRa2, a functional neurturin receptor. of Neurobiology 43 198–205. (doi:10.1002/(SICI)1097- Neuron 22 243–252. (doi:10.1016/S0896-6273(00)81086-7) 4695(200005)43:2!198::AID-NEU9O3.0.CO;2-D) Sariola H & Saarma M 2003 Novel functions and signalling pathways for Lim P, Allan CM, Notini AJ, Axell AM, Spaliviero J, Jimenez M, Davey R, GDNF. Journal of Cell Science 116 3855–3862. (doi:10.1242/jcs.00786) McManus J, MacLean HE, Zajac JD et al. 2008 Oestradiol-induced da Silva S & Wang F 2011 Retrograde neural circuit specification by target- spermatogenesis requires a functional androgen receptor. Reproduction, derived neurotrophins and growth factors. Current Opinion in Fertility, and Development 20 861–870. (doi:10.1071/RD08144) Neurobiology 21 61–67. (doi:10.1016/j.conb.2010.07.007) Lin TM, Ko K, Moore RW, Simanainen U, Oberley TD & Peterson RE 2002 Simanainen U, Allan CM, Lim P, McPherson S, Jimenez M, Zajac JD, Effects of aryl hydrocarbon receptor null mutation and in utero and Davey RA & Handelsman DJ 2007 Disruption of prostate epithelial lactational 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure on prostate androgen receptor impedes prostate lobe-specific growth and function. Journal of Endocrinology and seminal vesicle development in C57BL/6 mice. Toxicological Endocrinology 148 2264–2272. (doi:10.1210/en.2006-1223) Sciences 68 479–487. (doi:10.1093/toxsci/68.2.479) Simanainen U, McNamara K, Gao YR & Handelsman DJ 2009 Androgen Lujan M, Paez A, Llanes L, Angulo J & Berenguer A 1998 Role of autonomic sensitivity of prostate epithelium is enhanced by postnatal androgen innervation in rat prostatic structure maintenance: a morphometric receptor inactivation. American Journal of Physiology. Endocrinology and analysis. Journal of Urology 160 1919–1923. (doi:10.1016/S0022- Metabolism 296 E1335–E1343. (doi:10.1152/ajpendo.00017.2009) 5347(01)62444-3) Simanainen U, Brogley M, Gao YR, Jimenez M, Harwood DT, Martinez-Pineiro L, Dahiya R, Nunes LL, Tanagho EA & Schmidt RA 1993 Handelsman DJ & Robins DM 2011 Length of the human androgen Pelvic plexus denervation in rats causes morphologic and functional receptor glutamine tract determines androgen sensitivity in vivo. changes of the prostate. Journal of Urology 150 215–218. Molecular and Cellular Endocrinology 342 81–86. McNamara KM, Harwood DT, Simanainen U, Walters KA, Jimenez M & Singh J, O’Neill C & Handelsman DJ 1995 Induction of spermatogenesis by Handelsman DJ 2010 Measurement of sex steroids in murine blood androgens in gonadotropin-deficient (hpg) mice. Endocrinology 136 and reproductive tissues by liquid chromatography–tandem mass 5311–5321. (doi:10.1210/en.136.12.5311) spectrometry. Journal of Steroid Biochemistry and Molecular Biology 121 Vaalasti A & Hervonen A 1979 Innervation of the ventral prostate of the rat. 611–618. (doi:10.1016/j.jsbmb.2010.02.001) American Journal of Anatomy 154 231–243. (doi:10.1002/aja. McVary KT, Razzaq A, Lee C, Venegas MF, Rademaker A & McKenna KE 1001540208) 1994 Growth of the rat prostate gland is facilitated by the autonomic Vaalasti A & Hervonen A 1980 Autonomic innervation of the human nervous system. Biology of Reproduction 51 99–107. (doi:10.1095/ prostate. Investigative Urology 17 293–297. biolreprod51.1.99) Viglietto G, Dolci S, Bruni P, Baldassarre G, Chiariotti L, Melillo RM, Melvin JE & Hamill RW 1987 The major pelvic ganglion: androgen control Salvatore G, Chiappetta G, Sferratore F, Fusco A et al. 2000 Glial cell of postnatal development. Journal of Neuroscience 7 1607–1612. line-derived neutrotrophic factor and neurturin can act as paracrine Melvin JE & Hamill RW 1989 Androgen-specific critical periods for the growth factors stimulating DNA synthesis of Ret-expressing spermato- organization of the major pelvic ganglion. Journal of Neuroscience 9 gonia. International Journal of Oncology 16 689–694. 736–742. Wang JM, McKenna KE & Lee C 1991 Determination of prostatic secretion Meng X, Lindahl M, Hyvonen ME, Parvinen M, de Rooij DG, Hess MW, in rats: effect of neurotransmitters and testosterone. Prostate 18 Raatikainen-Ahokas A, Sainio K, Rauvala H, Lakso M et al. 2000 289–301. (doi:10.1002/pros.2990180403) Regulation of cell fate decision of undifferentiated spermatogonia by Wanigasekara Y & Keast JR 2005 Neurturin has multiple neurotrophic GDNF. Science 287 1489–1493. (doi:10.1126/science.287.5457.1489) effects on adult rat sacral parasympathetic ganglion neurons. European Meng X, Pata I, Pedrono E, Popsueva A, de Rooij DG, Janne M, Rauvala H & Journal of Neuroscience 22 595–604. (doi:10.1111/j.1460-9568.2005. Sariola H 2001 Transient disruption of spermatogenesis by deregulated 04260.x)

Wanigasekara Y, Airaksinen MS, Heuckeroth RO, Milbrandt J & Keast JR isoforms by real time detection PCR using SYBR Green I. Neuroscience 2004 Neurturin signalling via GFRa2 is essential for innervation of Letters 320 141–145. (doi:10.1016/S0304-3940(01)02282-0) glandular but not muscle targets of sacral parasympathetic ganglion Xian CJ, Huang BR & Zhou XF 1999 Distribution of neurturin neurons. Molecular and Cellular Neurosciences 25 288–300. (doi:10.1016/ mRNA and immunoreactivity in the peripheral tissues of adult j.mcn.2003.10.019) rats. Brain Research 835 247–258. (doi:10.1016/S0006-8993(99) Widenfalk J, Nosrat C, Tomac A, Westphal H, Hoffer B & Olson L 1997 01590-5) Neurturin and glial cell line-derived neurotrophic factor receptor-b Yan H & Keast JR 2008 Neurturin regulates postnatal differentiation (GDNFR-b), novel proteins related to GDNF and GDNFR-a with specific of parasympathetic pelvic ganglion neurons, initial axonal cellular patterns of expression suggesting roles in the developing and projections, and maintenance of terminal fields in male urogenital adult nervous system and in peripheral organs. Journal of Neuroscience organs. Journal of Comparative Neurology 507 1169–1183. (doi:10.1002/ 17 8506–8519. cne.21593) Widenfalk J, Parvinen M, Lindqvist E & Olson L 2000 Neurturin, RET, Zhang X & Ho SM 2011 Epigenetics meets endocrinology. Journal of GFRa-1 and GFRa-2, but not GFRa-3, mRNA are expressed in mice Molecular Endocrinology 46 R11–R32. (doi:10.1677/JME-10-0053) gonads. Cell and Tissue Research 299 409–415. (doi:10.1007/ Zirkin BR, Santulli R, Awoniyi CA & Ewing LL 1989 Maintenance of s004419900068) advanced spermatogenic cells in the adult rat testis: quantitative Wong YW, Sia GM & Too HP 2002 Quantification of mouse glial cell-line relationship to testosterone concentration within the testis. derived neurotrophic factor family receptor a 2 alternatively spliced Endocrinology 124 3043–3049. (doi:10.1210/endo-124-6-3043)

Received in ﬁnal form 12 May 2013 Accepted 15 May 2013 Accepted Preprint published online 15 May 2013 Journal of Endocrinology