Flutamide-Induced Testicular Undescent in the Rat is Associated with Alterations in Genitofemoral Nerve Morphology

()022-534 7/94/1512-0509$03,00 /0 Tm: JOURNAL OF UROLOGY Copyright© 1994 by AMERICAN URO:,,OGlCAL ASSOC!AT!ON, Il-,c

FLUTAMIDE-INDUCED TESTICULAR UNDESCENT IN THE RAT IS ASSOCIATED WITH ALTERATIONS IN GENITOFEMORAL NERVE MORPHOLOGY D. A. HUSMANN*, T. B. BOONE AND M. J. McPHAUL From the Department of Urology, Mayo Clinic, Rochester, Minnesota and the Departments of Internal Medicine and Surgery, University of Texas Southwestern Medical Center, Dallas, Texas

ABSTRACT

Androgen regulation of testicular descent is an established fact. However, the mechanisms by which androgens assert their influence is unknown. One of the leading hypotheses regarding androgenic control of testicular descent is based upon the assumption that testosterone regulates the development of the sexually dimorphic genitofemoral nucleus and nerve (GFN). To investigate whether or not alterations in the genitofemoral nerve are indeed androgen dependent and associated with testicular undescent, we administered flutamide in a time-specific manner to pregnant SpragueDawley rats. The GFN of ten animals with unilateral intra-abdominal testicular undescent was compared with the contralateral GFN associated with the descended testicle. Digital microscopic evaluations found that the GFN associated with testicular undescent did express altered morphologic abnormalities. Specifically, the GFN associated with the cryptorchid testicle had significant reductions in the neural diameter (.0169 ± .0024 mm. 2 versus .0275 ± .0079 mm. 2 ; p <.01), the number of large myelinated fibers per mm, 2 (8A ± 3.3 versus 18.8 ± 5; p <.001) and the number of small myelinated nerve fibers per mm, 2 (632 ± 87.4 versus 1090,7 ± 104.3; p <.001). These data suggest that testicular undescent in the rat is associated with morphologic alterations in the ipsilateral genitofemoral nerve, The exact mechanism of how these alterations are related to cryptorchidism remains to be elucidated. KEY WORDS:

cryptorchism, flutamide, rats, testis

The hypothesis that the control of testicular descent is through the interplay between three different components, androgens, the genitofemoral nerve (GFN) and the cremasteric muscle, was first suggested by L. G, Lewis in 1948. 1 Since that time, support for the role of the GFN in testicular descent has been published by Beasley and Hutson and associates, who found that incision of the GFN or division of the spinal cord above the origin of the GFN will result in testicular undescent.2-4 These later studies led to the hypothesis that androgens act upon the sexually dimorphic genitofemoral nuclei to induce descent of the testes. 2-4 There is also evidence that the cremasteric muscle has an important role in both testicular descent and genitofemoral nerve and nuclei development. In particular, the finding that an incision through the gubernacular bulb results in obliteration of the cremasteric muscular sac, altered genitofemoral nuclei morphology and testicular undescent 5 • 6 reconfirms two of the three components of L, G, Lewis's hypothesis. With the recent discovery that time-specific flutamide can result in undescent of the testes, 708 it became possible to test the hypothesis that alterations in the genitofemoral nerve and cremasteric muscle coexist in the presence of antiandrogen-induced cryptorchidism (failure of the testis to descend). MATERIALS AND METHODS

Flutamide is a nonsteroidal antiandrogen that inhibits androgen uptake and nuclear binding in target tissues. 90 10 Administration of this agent on gestational days 16 and 17 in the Sprague-Dawley rat resulted in the production of a novel animal model to study cryptorchidism. 7 •8 As noted in previous publications, 7• 8 the time-specific antiandrogen administration of flutamide can produce animals with either unilateral or bilateral cryptorchidism. In addition, the location of the undescended Accepted for publication September 14, 1993. * Requests for reprints: Department of Urology, E 17A, Mayo Clinic, 200 First St. S.W., Rochester, Minnesota 55905. 509

testicle can range from ectopic to perirenal in nature. 7• 8 In an effort to determine if alterations in the genitofemoral nerve were related to antiandrogen-induced testicular undescent the following protocol was performed: Protocol One. Ten timed pregnant Sprague-Dawley rats received injections of 10 mg. of flutamide (dissolved in 10% ethanol in triolein) per 100 gm. body weight on gestational days 16 and 17. Injection vehicle (10% ethanol in triolein) alone was administered to 3 additional timed pregnant animals for controls. All neonatal animals were killed on the 30th postpartum day. Necropsy was performed under a 5X to 15X operating diploscope (Applied Fiberoptics, Southbridge, Massachusetts). In an effort to correlate alterations in the genitofemoral nerve to flutamide-induced testicular undescent, we chose ten pups with unilateral cryptorchidism. All pups included in the study had the undescended testicle in the perinephric area, the most proximal location found in this cryptorchid animal model. 708 By using unilaterally cryptorchid animals with the most proximal location for testicular undescent, we believed that any alteration in the GFN related to testicular undescent would become apparent. The contralateral GFN associated with the descended testes would serve as a flutamide-exposed control population. Alterations in the GFN morphology in this latter group should demonstrate neuronal changes caused by the antiandrogen, but not necessarily related to cryptorchidism. To establish normal values in animals not exposed to flutamide, controls were also obtained from ten male and female pups whose mothers had been injected with the vehicle alone on gestational days 16 and 17. At the time of sacrifice the cremasteric muscles of the selected pups were excised and the weight was recorded. Due to variations in weight of the pups from multiple litters, the muscle weight was recorded in mg./100 gm. body weight. Weights are given as mean values ± the standard error of the mean.

510

FLUTAMIDE-INDUCED TESTICULAR UNDESCENT IN THE RAT

Following removal of the cremasteric muscle, the genitofemoral nerves were excised proximal to the take-off of the femoralis and cremasteric neural branches. 11 It should be noted that no abnormal branchings of the nerve were found following treatment with flutamide. After excision the nerve was fixed in a 0.1 M. phosphate buffered solution (pH 7.4) containing 1% formaldehyde and 2% gluteraldehyde overnight at 4C. In the morning the specimens were rinsed in phosphate buffer (three changes of solution over 1 hour). They underwent secondary fixation with 1 % Os0 4 at 4C for 2 hours and were subsequently embedded in epon. Microscopic sections were taken at 5 µm. and analyzed by digital microscopy for quantitative morphometric analysis. Specifically, an IBAS image analysis system was used (Krontron Image Analysis Division, Carl Zeiss, Inc., Thornwood, New York). Video images were captured from a Zeiss Axophot microscope using a 40X plan neofluar objective (Carl Zeiss Inc.). The microscope was fitted with a Newvicon tube videocamera (Hamamutsu, Japan), and the image was displayed on a Sony RGB 12-inch monitor for analysis. The number of motorneurons per genitofemoral nerve was counted and expressed per area (number per .01 mm. 2 ). The myelinated nerve fibers were broken into their classically described bimodal pattern of small (25 to 30 µm.) versus large fibers (45 to 60 µm.) 12 •13 and counted directly. These values are also expressed as number per .01 mm. 2 ± the standard error of the mean. It should be noted that the cremasteric muscle in the rat everts to form the cremasteric sac into which the testis descends. Therefore, one of our major concerns in this study was whether or not the absence of the testicle in the scrotum could affect GFN morphology or cremasteric weight. Protocol two was therefore established to verify that the absence of the testes from the cremasteric sac, that is, the scrotum,8 did not result in alterations in nerve morphology or wet weight of the muscle. Protocol Two. Ten male rat pups 14 days of age underwent a unilateral inguinal orchiectomy; the animals were killed on the 30th day of life. We deliberately chose the fourteenth day of life to perform unilateral castration since the cremasteric muscle is fully formed by that day, 8 • 14 and the testicle is still intraabdominal (not descending until the 21st or 22nd postpartal day). 8 •14 It should be noted that in this species the cremasteric muscle lines the scrotum with the testicle descending into its sac-like structure. Orchiectomy prior to testicular descent therefore prevents the maximal mechanical stretch of the cremasteric sac that occurs with descensus of the testis. Any alterations of the genitofemoral nerve that would become apparent in animals from Protocol Two would be the result of ineffective development of the genitofemoral nerve due to cremasteric muscle disuse. At the time of the anatomic dissections, the cremasteric muscles were removed and weighed and weights were recorded relative to the animals' total body weight (mg./ 100 gm. body weight). The genitofemoral nerve was excised on both the side of unilateral orchiectomy and on the side of the descended testicle. Both nerves were fixed and analyzed as in Protocol One. Statistical evaluations. Statistical analysis performed within this paper used either a paired t test or, where indicated, two sample t tests with unilateral variance. P values <.01 were considered significant. RESULTS

In the ten animals with perinephric unilateral undescended testes exposed to time-specific flutamide, the weight of the cremasteric muscle on the side of the intra-abdominal testes was significantly lower (37 ± 10 mg./100 gm. body weight) compared with the cremasteric muscle on the side of the descended testes, (56 ± 6 mg./100 gm. body weight, p <.001 by paired t test analysis). No significant difference was found between the cremasteric muscles exposed to flutamide with descended testes and the normal male controls (60 ± 8 mg./ 100 gm. body weight).

To investigate whether there was a coexisting alteration in the genitofemoral nerve associated with testicular undescent, morphometric analysis was used to compare the GFN on the side of descended versus the side of the undescended testis in unilateral cryptorchid animals exposed to flutamide. These comparisons revealed a significant alteration of the GFN morphology related to testicular undescent. Specifically, a decrease in both GFN diameter (.0274 ± .0075 mm. 2 versus .0170 ± .0020 mm. 2 ; p <.01), the number of large myelinated nerves (18.75 ± 4.62/.01 mm. 2 versus 8.42 ± 3.12/.01 mm. 2 ; p <.001) and the number of small myelinated nerves (1090.6 ± 98.96/ .01 mm. 2 versus 632.05 ± 82.90/.01 mm. 2 ; p <.001) were found in the GFN associated with perinephric cryptorchidism, compared with the contralateral GFN associated with the descended testicle. Examples of representative sections of the GFN are shown in figure 1, and the results of morphometric analysis are summarized in figures 2-4. As is apparent from these figures, the genitofemoral nerves exposed to flutamide and associated with undescended testes were almost identical in morphology to normal female controls. As noted in figures 2 and 3, the GFN of the flutamide-exposed animals associated with testicular descent did have a reduction in the genitofemoral area and the number of small and large myelinated nerves compared with normal male controls with descended testicles; however, the reductions were not statistically significant, p >.02. In all parameters evaluated, the GFN of the flutamide-exposed animals associated with testicular descent and the normal male control animals were significantly different from normal female controls, p <.001. After reviewing these data, we became concerned that the absence of the testis from the scrotum may have been responsible for the observed reduction in cremasteric weight and the alteration in morphology of the genitofemoral nerve associated with the undescended testicle. Protocol Two was therefore developed to investigate whether the absence of the testicle from the cremasteric sac was responsible for either of the alterations. To investigate this possibility, unilateral castration was performed on ten neonatal rats at postpartum day 14 (note: the testicle descends into the cremasteric sac on postpartal day 21 or 22). At 30 days of life when the animals were killed, no significant differences were dectected between the cremasteric muscle on the side with the descended testes and the side with the unilateral castration: 61 ± 9 mg./100 gm. body weight versus 54 ± 10 mg./100 gm. body weight, respectively (p = .52 by paired t test analysis). Similarly no significant differences could be found in the GFN diameter (.0280 ± .0075 versus .0270 ± .035), the number of large myelinated nerve fibers per .01 mm. 2 (31 ± 3.6 versus 28.3 ± 5.2), or the number of small myelinated nerve fibers per .01 mm. 2 (1054 ± 92.6 versus 1023 ± 105.2). All reported numbers equal the mean and standard error of the mean for ten separate specimens. These studies suggest that the absence of testicular mass within the cremasteric sac does not affect cremasteric muscle weight or genitofemoral nerve development at 30 days of age. DISCUSSION

The concept that the genitofemoral nerve (GFN) could play an important role in testicular descent is not novel. Indeed, L. G. Lewis in 1948 suggested that the interplay between androgens, the cremasteric muscle and the GFN regulated the descent of the testes. 1 This concept is supported by the finding that division of the GFN or spinal cord transection above the Ll-2 GF nucleus (also known as the cremasteric nucleus) results in cryptorchidism. 2- 4 Further fortifying this hypothesis was the knowledge that the GF nucleus and nerve are sexually dimorphic and that their morphology is influenced by androgens. 6· 13· 15 In particular the genitofemoral nucleus and nerve of the male animal has more neuronal fibers than those of the female. The sexual dimorphism of the male's GF nucleus and nerve is due to an increase in the substance P neuronal path-

511

PLUTAivHDE-KNDUCED TESTICULAR UNDESCENT KN THE RAT

FIG. 1. Genitofemoral nerves at 30 days of age associated with testicular descent and undescent. This animal was exposed to flutamide on gestational days 16 and 17 only and was killed 30 days postpartum. Necropsy revealed unilateral cryptorchidism with undescended testicle located in perirenal area. A, genitofemoral nerve on side of testicular undescent, X 1500. B, genitofemoral nerve associated with testicular descent, X 1500. 0.04 . - - ~ - - - - - - - - - - - - - - - - - - - , 0.035 0.03 0.025 0.02 0.015 0.01 0.005 0 Descended

Undescended

Female

Male

Fm. 2. Comparison of genitofemoml nerve area in animals with unilateral cryptorchidism and controls. Mean plus and minus standard error of mean for GFN area in animals with unilateral undescended testes. Ten nerves were evaluated from each study group. Statistically significant differences were found between GFN of normal control males and flutamide-exposed GFN on ipsilateral side of descended testicle when compared with normal control females (p <.001) or flutamide-exposed GFN on ipsilateral side of undescended testicle (p <.001 for normal males and p <.01 for flutamide exposed GFN on ipsilateral side of descended testicle).

ways (regulatory fibers for nociceptive and thermal afferents from the testes), serotonin fibers (thermal regulatory fibers) and the myelinated motoneurons (responsible for transmission of the neural impulses that result in cremasteric muscle contractions). In essence, these neuronal pathways are responsible for regulating testicular position in response to an environmen tal stimulus, predominantly temperature or pain. The findings of significant ipsilateral alterations in genitofemoral nerve morphology and cremasteric muscle wet weight revealed in this paper suggest that both of these alterations develop in response to antiandrogens and that they are associated with the presence of cryptorchidism. Our data imply that androgens could regulate testicular descent in the rat by one of three mechanisms: A) by direct androgenic stimulation of the developing genitofemoral nucleus; B) by direct androgenic stimulation of the developing cremasteric muscle; or C) both the neuronal and muscular systems could be affected by the localized production of androgen-induced paracrine factors. 6 • 16- 22 The first hypothesis, that androgens act directly on the genitofemoral nucleus or nerve to produce its altered morphology, cannot be supported by current scientific data. One of the basic tenets of androgen physiology is that this hormone can only function when its receptor is present. Current investigations into the ontologic development of androgen receptors in the lumbar spinal cord have revealed that the receptor does not develop in the lumbar spinal cord until the 10th postpartum

Descended

Undescended

Female

Male

FIG. 3. Comparison of number of large myelinated nerves in animals with unilateral cryptorchidism and control population. Mean± standard error of mean of ten animals in each study group. Number of large myelinated nerves within genitofemoral nerve is expressed per nerve area (0.01 mm.2). Statistically significant differences were found between normal control male population compared with either normal control female populations (p <.001) or flutamide exposed GFN ipsilateral to undescended testicle (p <.001). Genitofemoral nerve on ipsilateral side of descended testicle was significantly different from normal control female population (p <.001) and flutamide exposed GFN on ipsilateral side of descended testicle (p <.001). Comparison between normal male controls and flutamide-exposed GFN on ipsilateral side of descended testicle approached but did not reach statistical significance (p >.02).

day. 23 - 25 The inability to detect androgen receptors present in the fetal or neonatal spinal cord, 25 the time of maximal androgenic activity in descent of the testis, 5 • 6 • 26 suggests that this is not the location for androgenic activity in testicular descent. In contrast to the spinal cmd, both the paratesticular areas and the cremasteric muscle demonstrate androgen receptors as early as gestational day 15, the earliest time point investigated to date. 14 (Also personal observation, D. A. Husmann). These latter findings have led to the speculation13 that androgen receptors develop in the spinal cord after being induced by an as yet unidentified factor or factors produced by the muscle. Regarding the second hypothesis, it is possible that androgens may be having a direct effect on the cremasteric muscle, independent of neural innervation. This hypothesis is based on the finding that androgen can stabilize sexually dimorphic muscle development in the absence of innervation. 18• 25 This hypothesis would suggest that androgens act locally, either on the gubemacular mesenchyme, the progenitor of the cremasteric muscle in this species, or on the cremasteric muscle itself. According to this hypothesis it would be the direct action of androgen on the developing cremasteric muscle and not the indirect action of androgen on its neural innervation that would be responsible for both the full development of the cremasteric

512

FLUTAMIDE-INDUCED TESTICULAR UNDESCENT IN THE RAT 1,200

-I ll

C:

e::::, Q)

1,000

ci

C: C: "'O

Q) N

800

•

cii E .5 E

600

>-o

400

Q) ...

Eo

1a

E

200

rJ")

0 Descended

Undescended

Female

Male

FIG. 4. Comparison of number of small myelinated nerve fibers in genitofemoral nerve in animals with unilateral cryptorchidism and control population. Values expressed represent mean± standard error of mean for ten specimens in each study group. Number of small myelinated nerves is expressed per genitofemoral nerve area (0.01 mm. 2). Statistically significant differences were found between genitofemoral nerve of control males and ipsilateral flutamide exposed GFN associated with testicular descent when compared with either control females (p <.001) or flutamide-exposed ipsilateral GFN associated with undescended testicle (p <.001).

muscle and the genitofemoraL nerve. This hypothesis is supported by the known facts that incision of the gubernacular bulb can prevent testicular descent and can also permanently alter genitofemoral nuclei development. 5 • 6 In essence, the available published data imply that the direct effect of androgens on the genitofemoral nuclei in the Ll-2 spinal cord is not the etiology of the GFN abnormality. Rather the findings are most consistent with the current hypothesis of neuronal development, which states that the nerve's target organ produces paracrine factors that modulate neuronal survival by retrograde transport for these factors back to the neuronal cell. 6· 17· 22 According to this hypothesis, the target organ, either the gubernaculum or the cremasteric muscle arising from the gubernaculum, is the site of androgenic activity at a critical time during embryological development. 7· 8· 23· 26· 27 Androgens would act either directly or indirectly (through the production of androgen-mediated paracrine factors) on the cremasteric muscle to prevent its fatty degeneration. 18· 28 Full development of the GFN would subsequently depend either upon the complete development of the cremasteric muscle or upon the action of the androgenic-induced paracrine factors produced by the muscle. At present, either of the latter two hypotheses are in complete agreement with Jost's theory of sexual development. 29 In essence, each testicle would produce high local levels of testosterone at a crucial point in development. The hormone will act on a local target organ, possibly the paratesticular mesenchyme or the gubernaculum, the latter of which is the site of the developing cremasteric muscle in this species. The production of paracrine or growth factors from these structures would be responsible for full development of the genitofemoral nerve and possibly testicular descent. The production of unilateral cryptorchidism would occur after one of two events: Either the localized levels of androgens fail to produce a direct effect on the target organ or, alternatively, inadequate levels of functional paracrine factors are produced by the target organ. The experiments revealed in this paper cast the biological phenomena of testicular undescent in a new light and suggest a number of immediate experiments relating to the expression of the androgen receptor in dimorphic muscle groups and in their associated motor nuclei. In addition, they open the way to investigation of possible androgen-regulated paracrine factors as the mediators of testicular descent. Indeed a multitude of paracrine factors is already known to be under androgenic control and could theoretically alter the morphologies of the genitofemoral nerve, the paratesticular appendages and the anatomy of the inguinal and scrotal regions. In particular, the

paracrine function of neurotrophic growth factor (NGF) is known to be controlled by androgens. 19 Diminished secretion of NGF by a nerve's target organ is known to result in neuronal death, a fact similar to the findings of reduced small and large myelinated neurons recorded in this paper. An alternate paracrine factor that could play a major role is epidermal growth factor (EGF). Epidermal growth factor is also under androgenic control2° and is of significance since it serves to stabilize the wolffian ducts by preventing their dissolution. 21 Abnormal EGF secretion in the peritesticular areas by androgens during embryologic development could explain the increased incidence of epididymal abnormalities seen in testicular undescent. 30 In addition, EGF is also known to increase hyaluronic acid production.31 This finding is significant since inadequate production of hyaluronic acid would prevent swelling of the gubernacular bulb, 8 which serves to dilate the inguinal canal, thereby allowing the testis to descend from the abdomen into the scrotum. Either inadequate stimulation or abnormal production of EGF by androgens could therefore explain both the pathogenesis of cryptorchidism and the increased epididymal abnormalities seen in this entity. Another paracrine factor that could possibly play a role is calcitonin gene-related peptide (CGRP). This peptide is known to be responsive to both androgens and an as yet unidentified soluble factor from the sexually dimorphic muscles. 31 Calcitonin gene-related peptide causes increased rhythmic contractions of the gubernaculum. The rhythmic contractions of the gubernaculum are believed to aid the transinguinal migration of the testicle. 32· 33 Since CGRP is released through the GFN, any alteration in its neural structure (as demonstrated in this paper) could theoretically result in impairment of the peptide's release and its subsequent physiologic function. In summary, time-specific antiandrogen administration results in unilateral cryptorchidism with associated ipsilateral morphologic alterations of the genitofemoral nerve and reductions in cremasteric muscle weight. The exact mechanisms of how these alterations are related to cryptorchidism remains to be elucidated. REFERENCES

1. Lewis, L. G.: Cryptorchidism. J. Urol., 60: 345, 1948. 2. Beasley, S. W. and Hutson, J. M.: Effect of division of genitofemoral nerve on testicular descent in the rat. Aust. N. Z. J. Surg., 77: 49, 1987. 3. Beasley, S. W. and Hutson, J. M.: The role of the gubernaculum in testicular descent. J. Urol., 140: 1191, 1988. 4. Hutson, J.M., Beasley, S. W. and Bryan, A. D.: Cryptorchidism in spina bifida and spinal cord transection: a clue to the mechanism of transinguinal descent of the testis. J. Pediatr. Surg., 23: 275, 1988. 5. Frey, H. L. and Rajfer, J.: Role of the gubernaculum and intraabdominal pressure in the process of testicular descent. J. Urol., 131: 574, 1984. 6. Newton, B. W. and Hamill, R. W.: Target regulation of the seritonin and substance P innervation of the sexually dimorphic cremasteric nucleus. Brain Res., 485: 149, 1989. 7. Spencer, J. R., Torrado, T., Sanches, R. S., Vaughn, E. D. and Imperato-McGinley, J.: Effects of flutamide and finasteride on rat testicular descent. Endocrinology, 129: 741, 1991. 8. Husmann, D. A. and McPhaul, M. J.: Time specific androgen blockade with flutamide inhibits testicular descent in the rat. Endocrinology, 129: 1409, 1991. 9. Sufrin, G. and Coffey, D. S.: Flutamide: mechanism of action of a new nonsteroidal antiandrogen. Invest. Urol., 13: 429, 1976. 10. Kemppainen, J. A., Lane, M. V., Sar, M. and Wilson, E. M.: Androgen receptor phosphorylation, turnover, nuclear transport and transcriptional activity. J. Biol. Chem., 267: 968, 1992. 11. Hebel, R. and Stromberg, M. W.: Anatomy and Embryology of the Rat. Worthsee, Germany: BioMed Verlag, pp. 182-184, 1986. 12. Jones, E. G. and Cowan, W. M.: Nervous Tissue in Histology, 4th ed. Edited by L. Weiss and R. 0. Greiss. New York: McGraw Hill Book Company, pp. 333-337, 1977.

FLUTAMIDE-Ii\/DUCED TESTICULAR UNDESCENT IN THE RAT 13. Nagy, J. I. and Senba, E.. Neural relations of cremaster motoneurons, spinal cord systems and the genitofemoral nerve in the rat. Brain Res. Bull., Hi: 609, 1985. 14. Husmann, D. A. and McPhaul, M. J.: Localization of the androgen receptor in the developing rat gubernaculum. Endocrinology, 128: 383, 1991. 15. Kojima, M. and Sano, Y.: Sexual differences in the topographical distribution of serotonergic fibers in the anterior column of rat lumbar spinal cord. Anat. Embryol., 170: 117, 1984. 16. Hauser, K F. and Toran-Allerand, C. D.: Androgen increases the number of cells in fetal mouse spinal cord cultures: implications for motoneuron survival. Brain Res., 484: 157, 1989. 17. Oppenheim, R. W., Haverkamp, L. J., Prevette, D., McManaman, J. L. and Appel, S. H.: Reduction of naturally occurring motoneuron death in vivo by a target-derived neurotrophic factor. Science, 240: 919, 1988. 18. Fishman, R. B. and Breedlove, S. M.: Neonatal androgen maintains sexually dimorphic muscles in the absence of innervation. Muscle Nerve, 11: 553, 1988. 19. Ishii, D. N. and Shooter, E. M.: Regulation of nerve growth factor synthesis in mouse submaxillary glands by testosterone. J. Neurochem., 25: 843, 1975. 20. Noguchi, S., Ohba, Y. and Oka, T.: Involvement of epidermal growth factor deficiency in pathogenesis of oligospermia in streptozocin -induced diabetic mice. Endocrinology, 128: 2141, 1991. 21. Gupta, C., Siegel, S. and Ellis, D.: The role of EGF in testosteroneinduced reproductive tract differentiation. Dev. Biol., 146: 106, 1991. 22. Erikson, C. A. and Turley, E. A.: The effects of epidermal growth factor on neural crest cells in tissue culture. Exp. Cell Res., 169: 267, 1987. 23. Breedlove, S. M. and Arnold, A. P.: Hormonal control of a developing neuromuscular system. II. Sensitive periods for the androgen-induced masculinization of the rat spinal nucleus of the bulbocavernosus. J. Neurosci., 3: 424, 1983.

513

24. Breedlove, S. M. and Arnold, A. P.: Hormone accumulation in a sexuaily dimorphic motor nucleus of the rat spinal cord. Science, 210: 564, 1980. 25. Fishman, R. B., Chism, L., Firestone, G. L. and Breedlove, S. M.: Evidence for androgen receptors in sexually dimorphic perinea! muscles of neonatal male rats. Absence of androgen accumulation by the perinea! motoneurons. J. Neurobiol., 21: 694, 1990. 26. Husmann, D. A. and McPhaul, M. J.: Reversal offlutamide induced cryptorchidism by prenatal time specific androgens. Endocrinology, 131: 1711, 1992. 27. Goldstein, L. A. and Sengelaub, D.R.: Hormonal control of neuron number in sexually dimorphic spinal nuclei of the rat. IV. Masculinization of the spinal nucleus of the bulbocavemosus with testosterone metabolites. J. Neurobiol., 21: 719, 1990. 28. Radhakrishnan, J., Morikawa, Y., Donahoe, P. K. and Hendren, W. H.: Observations on the gubernaculum during descent of the testis. Invest. Urol., 16: 365, 1979. 29. Jost, A.: Problems of fetal endocrinology: the gonadal and hypophyseal hormones. Recent Prog. Horm. Res., 8: 379, 1953. 30. Hadziselmovic, F. and Kuslin, E.: The role of the epididymis in descensus testes and the topographical relationship between the testis and epididymis from the sixth month of pregnancy until immediately after birth. Anal. Embryo!., 155: 191, 1979. 31. Popper, P., Ulbarri, C. and Miceyvych, P. E.: The role of target muscles in the expression of calcitonin gene-related peptide mRNA in the spinal nucleus of the bulbocavernosus. Mo!. Brain Res., 13: 43, 1992. 32. Yamanaka, J., Metcalfe, S. A., Hutson, J.M. and Mendelsohn, F. A. 0.: Testicular descent. IL Ontogeny and response to denervation of calcitonin gene related peptide receptors in neonatal rat gubernaculum. Endocrinology, 132: 280, 1993. 33. Momose, Y., Griffiths, A. L. and Hutson, J. M.: Testicular descent. HI. The neonatal mouse gubernaculum shows rhythmic contraction in organ culture in response to calcitonin gene-related peptide. Endocrinology, 131: 2881, 1992.