Hormonally mediated plasticity of motoneuron morphology in the adult rat spinal cord: A cholera toxin-HRP study

The dorsolateral nucleus (DLN) and the spinal nucleus of the bulbocavernosus (SNB) of the rat lumbar spinal cord are sexually dimorphic groups of motoneurons that innervate striated perineal muscles involved in male copulatory behavior. Androgens control the development of these motoneurons and their target muscles, and continue to influence the system in adulthood. Given that several features of SNB motoneuron morphology have been shown to be androgen sensitive in adult male rats, we examined the effects of androgen manipulations on the morphology of motoneurons in the DLN in adult rats. Adult male rats were castrated and implanted with testosterone-filled or blank implants, or were subjected to a sham-castration procedure. Six weeks after treatment, motoneurons in the DLN were retrogradely labeled with cholera toxin-horseradish peroxidase (HRP) after injection into the ischiocavernosus (IC) muscle and their morphology assessed. Measures of the radial extent and coverage of the dendritic arbor of DLN motoneurons projecting to the IC (DLN-IC motoneurons) were similar across the groups, indicating comparable degrees of HRP transport. However, DLN-IC motoneurons in castrates with blank implants possessed both shorter dendritic lengths and smaller somas than those of castrates treated with testosterone. Castrates with testosterone implants had DLN-IC motoneurons that were significantly larger than those of sham castrates in dendritic length and soma area. These results suggest that motoneurons in the DLN, like those in the SNB, possess a significant degree of structural plasticity in adulthood which is influenced by androgens.     

Hormonally mediated plasticity of motoneuron morphology in the adult rat spinal cord: a cholera toxin-HRP study.

The dorsolateral nucleus (DLN) and the spinal nucleus of the bulbocavernosus (SNB) of the rat lumbar spinal cord are sexually dimorphic groups of motoneurons that innervate striated perineal muscles involved in male copulatory behavior. Androgens control the development of these motoneurons and their target muscles, and continue to influence the system in adulthood. Given that several features of SNB motoneuron morphology have been shown to be androgen sensitive in adult male rats, we examined the effects of androgen manipulations on the morphology of motoneurons in the DLN in adult rats. Adult male rats were castrated and implanted with testosterone-filled or blank implants, or were subjected to a sham-castration procedure. Six weeks after treatment, motoneurons in the DLN were retrogradely labeled with cholera toxin-horseradish peroxidase (HRP) after injection into the ischiocavernosus (IC) muscle and their morphology assessed. Measures of the radial extent and coverage of the dendritic arbor of DLN motoneurons projecting to the IC (DLN-IC motoneurons) were similar across the groups, indicating comparable degrees of HRP transport. However, DLN-IC motoneurons in castrates with blank implants possessed both shorter dendritic lengths and smaller somas than those of castrates treated with testosterone. Castrates with testosterone implants had DLN-IC motoneurons that were significantly larger than those of sham castrates in dendritic length and soma area. These results suggest that motoneurons in the DLN, like those in the SNB, possess a significant degree of structural plasticity in adulthood which is influenced by androgens.     

A 'window of time' during which testosterone determines the opiatergic control of LH release in the adult male rat

Male rats castrated before puberty (when 26 days of age) showed a progressively decreasing susceptibility to the inhibitory effects of morphine (5 mg/kg) upon LH secretion for up to 28 days after gonadectomy (approximately 100%, 40% and 10% inhibition at 5, 12 and 28 days after castration), but thereafter morphine again caused approximately 50% reduction in serum LH values; the minimum inhibition found at 28 days after castration (age 54 days) occurred at the time at which male rats normally reach puberty. When rats were castrated at 59 days of age, morphine maximally suppressed serum LH concentrations (to less than 70%) 2 and 5 days after castration, but had no effect thereafter. In prepubertal castrates, testosterone replacement between Days 26 and 50 of life resulted in responses to morphine similar to those found in rats castrated after puberty, i.e. serum LH levels were not reduced. Morphine significantly reduced LH levels in prepubertal castrates given testosterone after 60 days of age. Treatment with morphine consistently elevated serum prolactin concentrations (greater than 100%) in castrated rats of all ages, regardless of the time elapsed after gonadectomy. These results indicate a transient fall in the inhibitory opioidergic tone upon LH secretion as the normal age of puberty approaches, that the ability of opiates to alter LH release in adulthood may depend upon testicular steroids secreted during the peripubertal period, and that the LH responses do not reflect general changes in the neuroendocrine response to opiates after castration since the prolactin response to morphine remains intact in rats castrated before and after puberty.     

Androgen action at the target musculature regulates brain‐derived neurotrophic factor protein in the spinal nucleus of the bulbocavernosus

We have previously demonstrated that brain‐derived neurotrophic factor (BDNF) interacts with testosterone to regulate dendritic morphology of motoneurons in the highly androgen‐sensitive spinal nucleus of the bulbocavernosus (SNB). Additionally, in adult male rats testosterone regulates BDNF in SNB motoneurons and its target muscle, the bulbocavernosus (BC). Because BDNF is retrogradely transported from skeletal muscles to spinal motoneurons, we hypothesized that testosterone could regulate BDNF in SNB motoneurons by acting locally at the BC muscle. To test this hypothesis, we restricted androgen manipulation to the SNB target musculature. After castration, BDNF immunolabeling in SNB motoneurons was maintained at levels similar to those of gonadally intact males by delivering testosterone treatment directly to the BC muscle. When the same implant was placed interscapularly in castrated males it was ineffective in supporting BDNF immunolabeling in SNB motoneurons. Furthermore, BDNF immunolabeling in gonadally intact adult males given the androgen receptor blocker hydroxyflutamide delivered directly to the BC muscle was decreased compared with that of gonadally intact animals that had the same hydroxyflutamide implant placed interscapularly, or when compared with castrated animals that had testosterone implants at the muscle. These results demonstrate that the BC musculature is a critical site of action for the androgenic regulation of BDNF in SNB motoneurons and that it is both necessary and sufficient for this action. Furthermore, the local action of androgens at the BC muscle in regulating BDNF provides a possible mechanism underlying the interactive effects of testosterone and BDNF on motoneuron morphology. © 2013 Wiley Periodicals, Inc. Develop Neurobiol 73: 587–598, 2013     

Exogenous testosterone prevents motoneuron atrophy induced by contralateral motoneuron depletion

Gonadal steroids exhibit neuroprotective and neurotherapeutic effects. The lumbar spinal cord of male rats contains a highly androgen-sensitive population of motoneurons, the spinal nucleus of the bulbocavernosus (SNB), whose morphology and function are dependent on testosterone in adulthood. Unilateral SNB motoneuron depletion induces dendritic atrophy in contralateral SNB motoneurons, but this atrophy is reversed in previously castrated males treated with testosterone. In the present experiment we test the hypothesis that the morphology of SNB motoneurons is protected from atrophy after contralateral motoneuron depletion by exogenous testosterone alone (i.e., with no delay between castration and testosterone replacement). We unilaterally depleted SNB motoneurons by intramuscular injection of cholera toxin conjugated saporin. Simultaneously, some saporin-injected rats were castrated and immediately given replacement testosterone. Four weeks later, contralateral SNB motoneurons were labeled with cholera toxin conjugated HRP, soma sizes were measured, and dendritic arbors were reconstructed. Contralateral SNB motoneuron depletion induced somal atrophy and dendritic retraction, but testosterone treatment prevented both of these effects. Thus, the presence of high-normal levels of testosterone prevents motoneuron atrophy induced by contralateral motoneuron depletion. These data support a therapeutic role for testosterone in preventing atrophy induced by motoneuron injury.     

Androgen-sensitivity of somata and dendrites of spinal nucleus of the bulbocavernosus (SNB) motoneurons in male C57BL6J mice

In rats, androgens in adulthood regulate the morphology of motoneurons in the spinal nucleus of the bulbocavernosus (SNB), including the size of their somata and the length of their dendrites. There are conflicting reports about whether androgens exert similar influences on SNB motoneurons in mice. We castrated or sham-operated C57BL6J mice at 90 days of age and, thirty days later, injected cholera toxin conjugated horseradish peroxidase into the bulbocavernosus muscle (to label SNB motoneurons) on one side, and into intrinsic foot muscles contralaterally (to label motoneurons of the retrodorsolateral nucleus (RDLN)). Castrated mice had significantly smaller SNB somas compared to sham-operated mice while there were no differences in soma size of RDLN motoneurons. Dendritic length in C57BL6J mice, estimated in 3-dimensions, also decreased significantly after adult castration. In rats, androgens act directly through androgen receptors (AR) in SNB motoneurons to control soma size and nearly all SNB motoneurons contain AR. Since SNB somata in C57BL6J mice shrank after adult castration, we used immunocytochemistry to characterize AR expression in SNB cells as well as motoneurons in the RDLN and dorsolateral nucleus (DLN). A pattern of labeling matched that seen previously in rats: the highest percentage of AR-immunoreactive motoneurons are in the SNB (98%), the lowest in the RDLN (25%) and an intermediate number in the DLN (78%). This pattern of AR labeling is consistent with the possibility that androgens also act directly on SNB motoneurons in mice to regulate soma size in mice.     

Testosterone metabolites differentially maintain adult morphology in a sexually dimorphic neuromuscular system

The lumbar spinal cord of rats contains the sexually dimorphic, steroid‐sensitive spinal nucleus of the bulbocavernosus (SNB). Androgens are necessary for the development of the SNB neuromuscular system, and in adulthood, continue to influence the morphology and function of the motoneurons and their target musculature. However, estrogens are also involved in the development of the SNB system, and are capable of maintaining function in adulthood. In this experiment, we assessed the ability of testosterone metabolites, estrogens and nonaromatizable androgens, to maintain neuromuscular morphology in adulthood. Motoneuron and muscle morphology was assessed in adult normal males, sham‐castrated males, castrated males treated with testosterone, dihydrotestosterone, estradiol, or left untreated, and gonadally intact males treated with the 5α‐reductase inhibitor finasteride or the aromatase inhibitor fadrozole. After 6 weeks of treatment, SNB motoneurons were retrogradely labeled with cholera toxin‐HRP and reconstructed in three dimensions. Castration resulted in reductions in SNB target muscle size, soma size, and dendritic morphology. Testosterone treatment after castration maintained SNB soma size, dendritic morphology, and elevated target muscle size; dihydrotestosterone treatment also maintained SNB dendritic length, but was less effective than testosterone in maintaining both SNB soma size and target muscle weight. Treatment of intact males with finasteride or fadrozole did not alter the morphology of SNB motoneurons or their target muscles. In contrast, estradiol treatment was completely ineffective in preventing castration‐induced atrophy of the SNB neuromuscular system. Together, these results suggest that the maintenance of adult motoneuron or muscle morphology is strictly mediated by androgens. © 2009 Wiley Periodicals, Inc. Develop Neurobiol 70: 206–221, 2010.     

Testosterone increases acetylcholine receptor number in the “levator ani” muscle of the rat

The “levator ani” muscle of male rats provides a neuromuscular system in which both the muscle and its motoneurons have high levels of androgen receptors. Two weeks of castration caused a 48% loss of acetylcholine receptors in this muscle. One week of testosterone propionate injections initiated one week after castration increased receptor number by 27% over untreated castrate levels. These changes paralleled changes in muscle protein content. In contrast, castration and testosterone treatments of castrates had no effect on total, Triton X-100-extractable acetylcholinesterase activity. This system may provide a useful model of synaptic plasticity.     

Gonadal hormones during puberty organize environment-related social interaction in the male rat

This study examined the role of gonadal androgens during puberty on the development of environment-related social interaction (SI) in male rats. SI in an unfamiliar environment versus SI in a familiar environment was evaluated in young adult rats as a function of sex and gonadal status. Intact male rats at 60 days of age exhibited a differential response to the two environments, whereas SI in intact female rats at 60 days was equivalent in the two environments. Furthermore, male rats castrated as juveniles and tested for SI at 60 days displayed a pattern of environment-related SI similar to SI in intact adult female rats. This effect of juvenile castration on SI in male rats was prevented by chronic exposure to testosterone propionate (TP) over Days 30 through 60. SI in male rats castrated in adulthood, on the other hand, was not altered either 2 or 4 weeks postcastration. The results from this study indicate that pubertal secretions of gonadal androgen(s) are necessary for the development of environment-related SI in male rats. In contrast, secretions of gonadal androgens in adulthood do not appear to be critical for the continued expression of environment-related SI, as suggested by the observation that environment-related SI in male rats remains unchanged by castration in adulthood.     

Castration reduces motoneuron soma size but not dendritic length in the spinal nucleus of the bulbocavernosus of wild‐type and BCL‐2 overexpressing mice

Motoneurons in the spinal nucleus of the bulbocavernosus (SNB) and their target muscles, bulbocavernosus and levator ani (BC/LA), constitute an androgen‐sensitive neuromuscular system. Testosterone regulates SNB soma size, SNB dendritic length, and BC/LA muscle mass in adult male rats. Recent evidence indicates that the cell death‐regulatory protein, Bcl‐2, may also play a role in adult neural plasticity. The present study examined whether gonadal hormones and/or the Bcl‐2 protein influence the morphology of the SNB neuromuscular system in adult B6D2F1 mice. In Experiment 1, adult wild‐type and Bcl‐2 overexpressing males were castrated and implanted with silastic capsules containing testosterone or left blank. Six weeks after castration, cholera toxin‐horseradish peroxidase was injected into the BC muscle to label SNB dendrites. Animals were killed 48 h later, and BC/LA muscle mass, SNB soma size, and SNB dendritic arbors were examined. In Experiment 2, wild‐type and Bcl‐2 overexpressing males were castrated or sham castrated, implanted with testosterone‐filled or blank capsules, and examined 12 weeks later. In both experiments, BC/LA muscle mass and SNB soma size were significantly reduced in castrates receiving blank capsules. Surprisingly, however, there was no effect of hormone manipulation on any of several measures of dendritic length. Thus, the dendritic morphology of SNB motoneurons appears to be relatively insensitive to circulating androgen levels in B6D2F1 mice. Bcl‐2 overexpression did not influence BC/LA muscle mass, SNB soma size, or SNB dendritic length, indicating that the morphology of this neuromuscular system and the response to castration are not altered by forced expression of the Bcl‐2 protein. © 2002 Wiley Periodicals, Inc. J Neurobiol 53: 403–412, 2002     

Castration reduces motoneuron soma size but not dendritic length in the spinal nucleus of the bulbocavernosus of wild-type and BCL-2 overexpressing mice

Motoneurons in the spinal nucleus of the bulbocavernosus (SNB) and their target muscles, bulbocavernosus and levator ani (BC/LA), constitute an androgen-sensitive neuromuscular system. Testosterone regulates SNB soma size, SNB dendritic length, and BC/LA muscle mass in adult male rats. Recent evidence indicates that the cell death-regulatory protein, Bcl-2, may also play a role in adult neural plasticity. The present study examined whether gonadal hormones and/or the Bcl-2 protein influence the morphology of the SNB neuromuscular system in adult B6D2F1 mice. In Experiment 1, adult wild-type and Bcl-2 overexpressing males were castrated and implanted with silastic capsules containing testosterone or left blank. Six weeks after castration, cholera toxin-horseradish peroxidase was injected into the BC muscle to label SNB dendrites. Animals were killed 48 h later, and BC/LA muscle mass, SNB soma size, and SNB dendritic arbors were examined. In Experiment 2, wild-type and Bcl-2 overexpressing males were castrated or sham castrated, implanted with testosterone-filled or blank capsules, and examined 12 weeks later. In both experiments, BC/LA muscle mass and SNB soma size were significantly reduced in castrates receiving blank capsules. Surprisingly, however, there was no effect of hormone manipulation on any of several measures of dendritic length. Thus, the dendritic morphology of SNB motoneurons appears to be relatively insensitive to circulating androgen levels in B6D2F1 mice. Bcl-2 overexpression did not influence BC/LA muscle mass, SNB soma size, or SNB dendritic length, indicating that the morphology of this neuromuscular system and the response to castration are not altered by forced expression of the Bcl-2 protein.     

Influence of testosterone on the development of the ischiocavernosus muscle of the rat

The mean area of the neuromuscular endplates and the acetylcholinesterase (AChE) activity at the myotendinous junction of the ischiocavernosus muscle were studied in normal, castrated and testosterone-treated castrated Wistar rats by the Koelle method. The mean endplate area was found to be smaller in castrated rats, compared to normal ones (p less than 0.001), while testosterone treatment restored its normal size (0.8 greater than p greater than 0.7). The terminal AChE activity in castrated rats was as strong and spread as in juvenile ones, while it was almost absent in normal and in testosterone-treated castrated rats. The same parameters were examined in the tibialis anterior muscle of the same rats, chosen as a specimen of 'nonhormone-dependent' muscle, without finding any difference among the single groups.     

Cues that Elicit Ultrasounds from Adult Male Mice

Adult male mice (Mus musculus) which have a prior history ofexperience with other adult male and adult female mice readilyproduce 70 kHz ultrasonic vocalizations in the presence of urinefrom adult females but not in the presence of urine from adultmales. Urine from immatures of either sex does not elicit ultrasoundsfrom socially experienced adult males. The ultrasound elicitingpotency of adult male urine was not improved substantially followingcastration of adult males, injection of testosterone propionateto castrated adult males, administration of estradiol benzoateto castrated adult males, or neonatal castration. Ovarian hormonesdo not appear to be necessary for either the appearance at puberty,or the maintenance during adulthood, of the ultrasound elicitingcues of female urine. Stage of estrus did not have a major modulatingeffect on urinary cues eliciling male ultrasounds. Treatmentsthat did not substantially reduce the signal value of adultfemale urine include ovariectomy before or after puberty, ovariectomywith adrenalectomy, and neonatal administration of testosterone.The administration of testosterone to ovariectomized adult females,and hypophyseclomy, virtually eliminated the ability of urinefrom adult females to elicit ultrasounds from socially experiencedadult males. The implication of pituitary hormones in the modulationof female urinary cues thai elicit ultrasounds is particularlyinteresting since pituitary factors are also implicated in theproximal causation of postparturient maternal aggression, whichadult male ultrasounds may function to moderate.     

Some hormonal factors (hypophysectomy, castration and testosterone administration) modifying the course of "compensatory" muscle hypertrophy in the rat.

1. Maximum compensatory hypertrophy of the soleus and plantaris muscle in male rats is attained seven days after tenotomy of the gastrocnemius muscle (39% and 9% respectively). When tenotomy of the gastrocnemius was performed seven days ater hypophysectomy, hypertrophy in these two muscles was aproximately half that found in control animals. 2. After 81-day castration of young male rats the weight of the saleus and plantaris was reduced and hypertrophy following tenotomy of the gastrocneumius muscle did not develop. 3. Chronically castrated rats received testosterone two weeks prior to tenotomy of the gastrocnemius and a week during the muscle hypertrophy phase. Hypertrophy of the soleus in castrated rats which had received testosterone seven days after tenotomy of the gastrocnemius was 25% as compared with muscles of castrated animals. The corresponding value in the plantaris muscle was 10%. 4. These results indicate that even calf muscles of the rat, namely the soleus and plantaris muscles, are significantly affected by testosterone under these conditions, although it is not, as yet, clear whether its action is direct or indirect.     

Androgens negatively regulate myostatin expression in an androgen-dependent skeletal muscle

Myostatin is an important negative regulator of skeletal muscle growth, while androgens are strong positive effectors. In order to investigate the possible interaction between myostatin and androgen pathways, we followed myostatin expression in the androgen-dependent levator ani (LA) muscle of the rat as a function of androgen status. By testosterone deprivation (castration), we induced LA growth arrest in young male rats, whilst atrophy in adult ones, however, both processes could be reversed by testosterone supplementation. After castration, a significant up-regulation of active myostatin protein (and its propeptide) was found, whereas the subsequent testosterone treatment reduced myostatin protein levels to normal values in both young and adult rats. Similarly, a testosterone-induced suppression of myostatin mRNA levels was observed in castrated adult but not in young animals. Altogether, androgens seem to have strong negative impact on myostatin expression, which might be a key factor in the weight regulation of LA muscle.     

Effect of castration on hypothalamic testosterone metabolism in the male rat

In vitro studies were performed of hypothalamic testosterone (T) metabolism 30 days after castration of adult male rats. No changes were seen in T conversion into dihydrotesterone and estrogens in the castrated rats. Plasma T levels were decreased while plasma estradiol concentrations did not differ from those of intact controls. It was suggested that the hypothalamic T metabolism probably is not androgen dependent.     

Effects of estradiol and testosterone on the activity of pyruvate kinase of the cardiac and skeletal muscles of rats as a function of age and sex

The specific activities of pyruvate kinase of cardiac and skeletal (gastrocnemius) muscles of adult rats of both sexes are lower than those of immature rats. The activity does not change after adulthood in the cardiac muscle, but decreases in the gastrocnemius. The activity of pyruvate kinase of the heart of immature and adult rats of both sexes decreases after castration, but is unaffected in old rats. Castration has no effect on the activity of pyrovate kinase of the gastrocnemius muscle of rats of both sexes at any age. In invo administration of estradiol (50 μg/100 g body weight) increases the activity of pyruvate kinase of the heart of castrated male and female rats of the three ages. For the skeletal muscle, the activity increases in castrated adult female and old male rats only. A higher dose (100 μg) of estradiol has variable effects on pyruvate kinase of the heart of male and female castrated rats of different ages. This dose increase pyruvate kinase significantly in the skeletal muscle of old castrated male and female rats. However, it decreases it in the skeletal muscle of adult castrated male rats. Testosterone (100 μm) increases the activity of pyruvate kinase of the heart of castrated male rats. This increase is lower in old age. It has no effect in the heart of castrated female rats of any age. Testosterone (50 μg) increases pyruvate kinase activity of the skeletal muscle of young ovariectomized rats only. A higher dose (100 μg) causes a significant increase in pyruvate kinase of the skeletal muscle of castrated adult and old male, and young and adult female rats, respectively. These data show that sex steroid hormones induce pyruvate kinase of striated muscles, and that the age- and sex-dependent variations may be due to changes in the levels of receptor proteins.     

Castration‐induced upregulation of muscle ERα supports estrogen sensitivity of motoneuron dendrites in a sexually dimorphic neuromuscular system

The spinal cord of rats contains the sexually dimorphic motoneurons of the spinal nucleus of the bulbocavernosus (SNB). In males, SNB dendrites fail to grow after castration, but androgen or estrogen treatment supports dendritic growth in castrated males. Estrogenic support of SNB dendrite growth is mediated by estrogen receptors (ER) in the target muscle. ERα expression in cells lacking a basal lamina (referred to as “extra‐muscle fiber cells”) of the SNB target musculature coincides with the period of estrogen‐dependent SNB dendrite growth. In the SNB target muscle, extra‐muscle fiber ERα expression declines with age and is typically absent after postnatal (P) day 21 (P21). Given that estradiol downregulates ERα in skeletal muscle, we tested the hypothesis that depleting gonadal hormones would prevent the postnatal decline in ERα expression in the SNB target musculature. We castrated male rats at P7 and assessed ERα immunolabeling at P21; ERα expression was significantly greater in castrated males compared with normal animals. Because ERα expression in SNB target muscles mediates estrogen‐dependent SNB dendrogenesis, we further hypothesized that the castration‐induced increase in muscle ERα would heighten the estrogen sensitivity of SNB dendrites. Male rats were castrated at P7 and treated with estradiol from P21 to P28; estradiol treatment in castrates resulted in dendritic hypertrophy in SNB motoneurons compared with normal males. We conclude that early castration results in an increase in ERα expression in the SNB target muscle, and this upregulation of ERα supports estrogen sensitivity of SNB dendrites, allowing for hypermasculinization of SNB dendritic arbors. © 2013 Wiley Periodicals, Inc. Develop Neurobiol 73: 921–935, 2013     

Effects of testosterone on the metabolism of folate coenzymes in the rat

1. The effects of castration and testosterone treatment on enzymic activities involved in folate coenzyme metabolism in the liver and in accessory sex organs of male adult rats were studied. 2. In the liver of castrated rats the concentration of 10-formyltetrahydrofolate (10-HCO-H(4)folate) synthetase and tetrahydrofolate (H(4)folate) dehydrogenase were significantly decreased whereas that of 5,10-methylenetetrahydrofolate dehydrogenase increased; the treatment with five doses of testosterone caused a return to normal values of these activities. 3. In the prostate of castrated rats a pronounced decrease in H(4)folate dehydrogenase, serine hydroxymethyltransferase and 10-HCO-H(4)folate synthetase activities was observed. The administration of testosterone restored the enzymic activities to normal values. 4. In the seminal vesicles of castrated rats only 10-HCO-H(4)folate synthetase was markedly depressed; testosterone treatment not only restored activity to normal values but raised it to higher than normal values. The slight changes observed in other enzymic activities also returned to normal values with the hormone treatment. 5. These results are discussed in relation to a possible control mechanism of folate metabolism by testosterone.