| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular article |
a Department of Physiology and Pharmacology, School of Medicine, Oregon Health Sciences University, Portland, Oregon 97201-3098
ABSTRACT
Testosterone at physiological levels cannot exert negative feedback action on LH secretion in long-term castrated male monkeys. The cellular basis of this refractoriness is unknown. To study it, we compared two groups of male rhesus macaques: one group (group 1, n = 4) was castrated and immediately treated with testosterone for 30 days; the second group (group 2, n = 4) was castrated and treated with testosterone for 9 days beginning 21 days after castration. Feedback control of LH by testosterone in group 1 was normal, whereas insensitivity to its action was found in group 2. Using the endpoints of concentrations of aromatase activity (P450AROM messenger RNA [mRNA]) and androgen receptor mRNA in the medial preoptic anterior hypothalamus and in the medial basal hypothalamus, we found that aromatase activity in both of these tissues was significantly lower, P < 0.01, in group 2 compared with group 1 males. P450AROM mRNA and androgen receptor mRNA did not differ, however. Our data suggest that the cellular basis of testosterone insensitivity after long-term castration may reside in the reduced capacity of specific brain areas to aromatize testosterone. Because P450AROM mRNA did not change in group 2 males, we hypothesize that an estrogen-dependent neural deficit, not involving the regulation of the P450AROM mRNA, occurs in long-term castrated monkeys.
androgen receptor, estradiol, hypothalamus, Leydig cells, LH, pituitary hormones, steroid hormones, testes, testosterone
INTRODUCTION
Feedback control of LH secretion in male rhesus macaques has been studied in many laboratories (reviewed in [1]). Androgens secreted by the testes are intimately involved in the regulation of LH secretion. After castration (Cx) LH concentrations in the systemic circulation rise many-fold above those found in the circulation of intact males [2, 3] and are maintained at precastration levels after treatment with exogenous testosterone [2, 3].
Many studies indicate that brain aromatization mediates, at least partially, androgen regulation of gonadotropin release (reviewed in [4]). Our studies in nonhuman primates suggest that estrogen acts in concert with testosterone in the negative feedback control of gonadotropin release [3]. These results differ from previous reports [5] that indicate that testosterone continues to maintain its suppressive effect on gonadotropin secretion even after removal of estradiol (E2). Reconciliation of the two sets of data is possible if the larger doses of testosterone, used in experiments mentioned earlier [5], maintain cellular levels of E2. In contrast, male monkeys castrated for a period of time (>21 days) become refractory to the negative feedback actions of testosterone [3, 5, 6]. Little information has been published that explains the cellular components of this refractoriness. We reasoned that brain aromatization is associated with refractoriness to testosterone because of the following observations between species: even though Cx in rats produces a decline in hypothalamic aromatase activity [7], aromatization does not seem to play an important role in feedback regulation of gonadotropins by androgen in male rodents [8]. Therefore, we predicted that Cx in rats does not produce a state of refractoriness to testosterone as it does in nonhuman primates and, indeed, this seems to be true [9]. On the other hand, if we assume that aromatization plays a role in feedback control of gonadotropins in primates, and there is some evidence to support this idea [10], then refractoriness to testosterone would be expected to develop after Cx, and it does [3, 5, 6].
In addition, the brains of male rhesus monkeys contain P450 aromatase activity (AA), which declines after Cx in many brain areas, including the medial preoptic/anterior hypothalamus (MPAH) [11], and can be increased by testosterone in many of the same areas when administered at the time of Cx [12]. Significant amounts of GnRH can also be found in these regions [11].
Because brain areas that contain GnRH also contain AA, we previously studied the effects of brain aromatization on gonadotropin and testosterone secretion in intact male rhesus monkeys treated with an aromatase inhibitor (1, 4, 6-androstatriene-3, 17-dione; ATD) [10]. Our results clearly indicate that after ATD treatment, systemic levels of LH, but not FSH, are elevated. These elevations occur despite increases in systemic testosterone concentrations. Thus, aromatization seems to play an important mediating role in the physiological actions of androgen in negative feedback control of gonadotropins in primates. In this paper, we ask the following question: Might brain aromatization be involved in the refractoriness to the negative feedback actions of testosterone on LH secretion, which occurs in long-term castrated male rhesus monkeys? To answer this question, we studied brain AA, P450AROM messenger RNA (mRNA), and androgen receptor (AR) mRNA in MPAH and in the medial basal hypothalamus (MBH) of two groups of male rhesus monkeys. One group was refractory to the actions of testosterone on LH secretion; the other was responsive to the actions of testosterone.
MATERIALS AND METHODS
Animals
Eight adult male rhesus monkeys (Macaca mulatta, 5–10 yr of age) were used in this study. The housing and care of the animals have been described previously [11]. Experiments and animal care were conducted in accordance with the principles and procedures outlined by the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Blood samples (3 ml) were obtained from each animal through a cardiac catheter at 0900 and 2100 h, each day for 37 days (7 days before and 30 days after Cx) before they were killed. The placement of the catheters and the procedures for obtaining blood without disturbing the males have been described [13]. Testosterone was administered at the time of Cx in group 1 (n = 4) by placing four (4-cm in length) Silastic elastomer capsules (0.335 cm inside diameter; 0.465 cm outside diameter) filled with crystalline steroid s.c. between the scapulae. In this group, the capsules remained in place for 30 days. In group 2 (n = 4), the testosterone-filled capsules were implanted 21 days after Cx and remained in place for 9 days.
The tissues that were used for measuring aromatase activity, P450AROM mRNA and AR mRNA, by the water assay and ribonuclease protection assays, respectively, were obtained from the two treatment groups mentioned earlier. On the day of autopsy (between 0900 and 1200 h), each animal was anesthetized with ketamine hydrochloride (15 mg/kg body weight, i.m.) for transport from its cage to the autopsy room and then killed by injecting a lethal dose of pentobarbitol (85 mg/kg body weight, i.m.).
The brain was rapidly removed from the cranium and rinsed in ice-cold 0.9% saline. It was placed in a chilled monkey brain matrix (Activational Systems Inc.; distributed by Ted Pella Inc., Redding, CA) and cut into 4 mm coronal sections through the length of the preoptic area, hypothalamus, and adjacent amygdala. Individual regions were dissected from these coronal slices using diagrams of the monkey brain from the atlases of Snider and Lee [14] and Martin and Bowden [15]. The dissected tissues were immediately frozen on dry ice and stored at -80°C until they were assayed. The regional dissections used in the current study were somewhat larger than those used in our previous punch dissection studies [16] and, in many cases, encompassed more than one discrete brain nucleus or subregion. This dissection was necessary to obtain sufficient RNA for the P450AROM and AR mRNA protection assays. Specifically, MPAH included medial preoptic nucleus (n.), preoptic periventricular n., suprachiasmatic n., rostral part of the supraoptic n., paraventricular n., and anterior hypothalamic region and anterior n. of the hypothalamus. MBH included arcuate n., ventromedial hypothalamic n., medial-caudal part of the supraoptic n., median eminence, intermediate hypothalamic region, ventral part of the hypothalamic periventricular n., and tuber cinerium.
RNA Isolation
Total RNA was isolated by homogenizing tissues in 4 M guanidium isothiocyanate, 10 mM EDTA, 2% sodium N-lauryl sarcosine, 1% (v/v) ß-mercaptoethanol, 50 mM Tris HCL pH 7.6, in the presence of 10 mM vanadyl ribonucleoside complexes. The guanidium isothiocyanate homogenate was centrifuged through 5.7 M cesium chloride [17], and the RNA pellet was dissolved in 5 mM EDTA, 1% dodecyl sulfate, and 10 mM Tris HCL pH 7.4, and extracted once with phenol, once with chloroform/isoamyl alcohol (49/1,v/v), and concentrated by ethanol precipitation. Ten micrograms total RNA from each tissue was used in subsequent assays.
Ribonuclease Protection Assay
Tissue concentrations of P450AROM mRNA were measured by an RNAase protection assay using the 455 nt 32P-labeled cRNA probe transcribed from the 5' coding region of the rhesus monkey P450AROM complementary DNA (cDNA) as described previously [16]. Cyclophilin mRNA, which was used as a control for RNA loading on the gels, was measured using a 185 nt [32P] cRNA probe, which was transcribed from a rhesus monkey p1B15 cyclophilin cDNA cloned into pGEM-3Z vector (provided by Dr. Sergio Ojeda at the Oregon Regional Primate Research Center). The length of the protected cyclophilin mRNA fragment in the ribonuclease protection assay was 158 nt.
Protected areas on the polyacrylamide gels were quantified using a Molecular Imaging System (BioRad GS-525, Hercules, CA). Exposure time was 40 h. The phosphoimage signal for each sample was compared with the signals obtained from a standard curve of known amounts of P450AROM sense RNA (31.25–2000 fg), which were run with each gel. The results were expressed as P450AROM mRNA/µg total RNA analyzed.
Tissue concentrations of AR mRNA in monkey brain were measured exactly as described previously [18].
Aromatase Activity Assay
Aromatase activity was assayed in brain and adipose tissue by a radiometric technique that quantifies the incorporation of tritium from [1ß-3H]androstenedione into 3H-labeled water. The technique was validated previously in our laboratory for tissue from the rhesus monkey [11].
Hormone Assays
Testosterone, dihydrotestosterone (DHT), androstenedione, and E2 were quantified in systemic serum by radioimmunoassay after chromatography on Sephadex LH-20 as described previously [19, 20]. The intra-assay and inter-assay coefficients of variation for testosterone were 8.6% and 10.4%, respectively; for DHT they were 8.3% and 16.4%; for androstenedione they were 9.3% and 13.7%; and for E2 they were 5.3% and 16.5%. The lower limit of detectability (i.e., the first point on the standard curve lying outside 2 standard deviations of cpms in tubes containing unlabeled hormone, was 11.1 fmol/tube for testosterone, 10.2 fmol/tube for DHT, 6.7 fmol/tube for androstenedione, and 10.3 fmol/tube for E2.
Biologically active LH was quantified in systemic serum by a dispersed mouse Leydig cell bioassay, which was validated in our laboratory [21]. The intra-assay and inter-assay coefficients of variation for LH were 8.7% and 16.9%, respectively. The lower limit of detectability was 1.56 ng LER-1909-2 (LH standard)/tube.
Statistical Analysis
Pretreatment samples Morning and evening hormone values in serum from each male (n = 8), bled for 7 days before Cx, were determined to provide a baseline for comparison with post-treatment concentrations. Because of sample loss, only 41 determinations per time period were analyzed by a one-way ANOVA for repeated measures. Before analysis, the data were tested for homogeneity of variances. The E2 data were homogeneous. All other data were log-transformed before being analyzed.
Post-treatment samples Because males in group 2 were treated with testosterone for only 9 days (the day of Cx was considered Day 0), the hormone concentrations in their serum were compared with serum obtained from only the first 9 days after Cx in group 1 males. Serum samples obtained from the day of Cx and the final day of the experiment were not used in the analyses because matched morning/evening serum samples could not be obtained from these days. Therefore, only data from 32 matched serum samples (4 males x 8 days) for each time period were used to compare hormone levels in group 1 with group 2 males. These data were analyzed by a two-way ANOVA for repeated measures. If a significant F value was obtained by the ANOVA, post hoc comparisons were determined by the Newman-Keuls multiple range test. The data in the various treatment groups were tested for homogeneity of variances before the ANOVA and if the variances differed, the data were log-transformed for the analysis. All the data in the post-Cx period were log-transformed.
Because of the small number of subjects, differences in concentrations of AA, P450AROM mRNA, and AR mRNA in MPAH and MBH were analyzed by Mann-Whitney U-tests. All the statistical tests were performed using the computer program, GB-STAT (Dynamic Microsystems Inc., Silver Spring, MD). Probability values of P < 0.01 were considered statistically significant.
RESULTS
Steroid concentrations in serum during the week preceding Cx in both group 1 and group 2 males showed that testosterone, DHT, androstenedione, E2, and LH were all higher in serum in evening samples compared with morning samples, P < 0.01 (Table 1). After Cx, the serum of males, which were treated with testosterone for the entire period after Cx (group 1), contained mean testosterone concentrations of 9.1 ± 0.6 ng/ml (n = 32) at 0900 h and 8.3 ± 0.5 ng/ml (n = 32) at 2100 h during the 8-day sampling period (Fig. 1A). These means did not differ significantly from one another, P > 0.01. In group 2 males, those not treated with testosterone until Day 21 post-Cx, serum testosterone concentrations fell to nondetectable amounts during the 21 days following Cx but rose sharply after treatment with testosterone, reaching a mean concentration of 10.7 ± 0.8 ng/ml of serum at 0900 h and 9.8 ± 0.7 ng/ml of serum at 2100 h (n = 32) during the 8-day sampling period (Fig. 1B). These means did not differ significantly from one another (P > 0.01). The mean concentrations of testosterone in group 1 males were 10% and 15% lower in the morning and evening hours, respectively, than those found in group 2 males.
|
|
Concentrations of E2 in the systemic circulation of groups 1 and 2 are presented in Figure 1, C and D. Concentrations of E2 in serum obtained in the morning and evening hours from group 1 males treated with exogenous testosterone differed significantly (P < 0.01; Table 1), but compared with the pretreatment period, the pattern was reversed (i.e., the morning levels [0.010 ± 0.001 ng/ml serum, n = 32] exceeded those found in the evening [0.006 ± 0.0008 ng/ml, n = 32]). In group 2 males (Table 1), serum E2 concentrations did not change with time of day before or after treatment with testosterone (P > 0.01). Levels of E2 in serum collected at 0900 h were 0.013 ± 0.002 ng/ml vs. 0.012 ± 0.002 ng/ml at 2100 h (n = 32).
Figure 2 presents the concentrations of DHT (A and B) and androstenedione (C and D) in serum from the same males mentioned in Figure 1. Serum content of these hormones was significantly reduced after Cx, (P < 0.01; group 2 males during the 8-day sampling period after Cx) and reached higher levels after treatment with exogenous testosterone (P < 0.01; group 1 males during the 8-day sampling period after Cx). Levels of DHT and androstenedione in serum collected at 0900 vs. 2100 h did not differ significantly (P > 0.01; Table 1) in either group 1 or group 2 males during the 8 days used for comparative purposes.
|
The concentrations of bioactive LH in serum of group 1 males are presented in Figure 3A and in Table 1. In this group, serum testosterone concentrations of approximately 9 ng/ml suppressed LH concentrations in the systemic circulation, whereas twice this amount of testosterone was associated with elevated amounts of LH in gonad-intact males (Table 1). The concentrations of bioactive LH in serum from group 2 males are presented in Figure 3, panel B and Table 1. These latter data indicate that refractoriness to the negative feedback actions of testosterone occurred in group 2 males despite the maintenance of systemic testosterone concentrations, which were effective in group 1 males.
|
The quantities of AA, P450AROM mRNA, and AR mRNA in MPAH and MBH of males from groups 1 and 2 are presented in Figure 4. Aromatase activity was significantly lower (P < 0.01) in both MPAH and MBH of group 2 males, which did not have continuous exposure to testosterone after Cx, compared with group 1 males, which received testosterone treatment during the entire 30-day period following Cx. We also measured AA and P450AROM mRNA and AR mRNA in the following brain tissues: lateral preoptic anterior hypothalamus; lateral hypothalamus; dorso medial hypothalamus; bed nucleus of the stria terminalis; septum, cortical, medial, and basal amygdala; hippocampus; parietal and cingulate cortex; cerebellum; medulla; and abdominal adipose tissue. Neither AA nor P450AROM mRNA and AR mRNA differed between group 1 and group 2 males in any of these tissues (data not shown). Despite the difference in AA between the two groups of males in MPAH and MBH, the quantities of P450AROM mRNA and AR mRNA in these tissues did not differ significantly (P > 0.01).
|
DISCUSSION
The present study suggests that the cellular basis of testosterone insensitivity after long-term Cx may reside in the reduced capacity of specific brain areas to aromatize androgens. After Cx, treatment of males with testosterone-filled Silastic elastomer capsules maintained serum levels of testosterone at a relatively constant level of approximately 9 ng/ml. This amount of testosterone, if administered at the time of Cx, exerted negative feedback control of LH secretion for the entire period of treatment. In a previous study, testosterone replacement, which achieved a serum level of approximately 6–7 ng/ml, was unable to maintain systemic LH serum concentrations in the suppressed condition in male monkeys unless supplemented with small amounts of E2 [3]. It appears that the level of testosterone replacement in the present study was near the threshold for testosterone to exert its negative feedback actions on LH secretion. During this 30-day period, the testosterone, which diffused from the implants, was metabolized by 5
-reductase to DHT and probably by 17ß-hydroxysteroid dehydrogenase to androstenedione, and aromatase to E2, because all three steroids were elevated in the systemic circulation of treated males. The cases for E2 and androstenedione, however, are less compelling than for DHT because of the variance that we found in our measurements of these two hormones, perhaps indicating some contribution of the adrenal gland to the peripheral pool of these hormones. When treatment with testosterone was delayed for 21 days, the testosterone replacement was no longer capable of exerting negative feedback action on LH secretion during the 9 days of treatment, an observation that has been reported previously by us [3] and others [5, 6].
A negative feedback relationship between testosterone and bioactive LH secretion, which was clear in castrated males treated with exogenous testosterone, was not clear in intact males. Rather, the quantities of testosterone in serum from intact males appeared to be dependent upon LH secretion because the quantities of both hormones were elevated in serum at 2100 h (data shown in Table 1). Furthermore, testosterone concentrations in serum from intact males during this time period were approximately twice as high as concentrations in serum of males that were treated with exogenous testosterone, but LH was elevated in serum of intact males compared with group 1 males. Previously, we found differences in feedback regulation of gonadotropin secretion between gonad-intact and castrated males (Macaca fasicularis) treated with exogenous testosterone. In these studies, aromatase inhibitors disrupt feedback regulation of LH secretion of intact but not castrated males treated with testosterone [22]. Also, intact males will not release LH in response to an estrogen challenge, whereas castrated males treated with testosterone are responsive [23]. It should be noted, however, that our treatment protocols provided relatively constant amounts of testosterone in the systemic circulation, whereas the central nervous system and pituitary gland are exposed to rhythmic patterns of testosterone in gonad-intact males. Nevertheless, these data seem to indicate that in order to fully understand the control of gonadotropin secretion in male monkeys, one must understand the function of testicular secretions, not just the actions of testosterone.
In the context of this paradigm, we found that the capacity of MPAH and MBH to aromatize androgens was significantly diminished in those males in which the refractoriness to testosterone was established. This effect was specific for MPAH and MBH because treatment effects were not found in other brain tissues that aromatize androgens, such as amygdala and abdominal adipose tissue. Thus, it appears that in MPAH and in MBH, continuous exposure to testosterone or one of its active metabolites, in the absence of the testes, is necessary to maintain aromatization at a high level. Even though group 2 males treated with testosterone had elevated levels of both DHT and E2 in peripheral circulation, these hormones do not make a difference in restoring the negative feedback loop that is characteristic of the group 1 male.
The present study confirms previous reports on the diurnal nature of testosterone [24–27], DHT [26], and E2 [26] secretion in male rhesus monkeys. To this list can now be added androstenedione and LH, the serum concentrations of which are significantly higher in evening compared with morning hours. It is interesting that E2 concentrations in serum of group 1 males but not group 2 males differed significantly as a function of time of day. Even though this rhythm was opposite from that found in gonad-intact males, it may indicate that a requirement for negative feedback regulation of LH release by steroids in this species is a conditioning of hypothalamic neurons by episodic exposure to estrogen during a 24-h period. The only steroid in this study that fulfills this hypothetical requirement is E2 in group 1 males, in which the negative feedback loop is intact. Although we cannot explain the presence of an E2 rhythm in group 1 and its absence in group 2 males, its existence seems to depend upon the uninterrupted presence of testosterone in the systemic circulation of the male. It is important to note that brain tissues in this study were collected between 0900 and 1200 h, the time when E2 concentrations in the systemic circulation were elevated compared with the 2100-h samples. Does the capacity of tissues to aromatize androgens vary with time of day? If so, the rhythm that we observed may be a consequence of in situ estrogen production and, if found in the brain, could be an important mechanism for negative feedback regulation of gonadotropin secretion.
In addition to serving as a substrate for aromatization, testosterone regulates transcription, stability, or both of P450AROM mRNA in MPAH and MBH of male rhesus monkeys [28]. However, we found no differences in the quantities of P450AROM mRNA in MPAH and MBH between our two experimental groups, which might indicate that testosterone was equally effective in both groups of males, but that the effect on aromatase activity occurs distal to the control of transcription. Many observations refute the assumption that steroid effects on AA are always at the level of the genome. It seems likely that intracellular messengers, such as cAMP and their interactions with neurotransmitters and neuropeptides, can also exert regulatory effects on AA (reviewed in [29]).
MPAH and MBH of male rhesus monkeys have high concentrations of AR mRNA [18]. However, a comparison of the amounts of AR mRNA between treatment groups revealed no differences between them. This observation suggests that differences in AR mRNA cannot explain the cellular basis of the androgen-resistant state observed in our experiments. These observations do not necessarily mean that androgen effects on P450AROM mRNA in the male monkey brain are not mediated through the AR or that AR has not changed between the treatment groups. In male rats, testosterone regulates brain AA through an AR-mediated mechanism [30]. In this species, AR mRNA does not change in response to androgen treatment in MBH and the preoptic area [31] even though cytosolic and nuclear AR, measured by binding techniques, vary predictably with serum androgen concentrations [31]. The failure of AR mRNA to correlate with AR measurements in the brain differs from similar observations reported for other androgen-target tissues such as the prostate. In the rat prostate, the amounts of AR mRNA correlate with the presence of cytosolic AR [31]. An understanding of this discrepancy between these two androgen target tissues may explain differences in androgen action in these tissues. In the brain, androgen affects neuronal activity, whereas in the prostate, it initiates and maintains growth.
At the cellular level, estrogen exerts a wide range of neurotropic effects. Estrogen facilitates neurite development in vitro [32], the plasticity of GABAergic synapses of the hypothalamus [33], and up-regulates proenkephalin mRNA of the ventral medial hypothalamic nucleus [34]. In rodents, testosterone and E2 act upon astroglia, which mediate their effects (reviewed in [35]). In nonhuman primates, estrogen induces glial and synaptic plasticity of hypothalamic neurons [36] and increases ensheathing of neuronal somas by glial processes of GnRH neurons [37]. Thus, a deficit in in situ estrogen formation by MPAH and MBH of male monkeys could be an underlying cause of the insensitivity to the negative feedback control of LH by testosterone, which is found in this species.
FOOTNOTES
First decision: 28 February 2000.
1 Supported by NIH grants HD18196 and D43 TW HD00669. 
2 Correspondence. FAX: 505 494 4352; reskoj{at}ohsu.edu
Accepted: April 28, 2000.
Received: January 26, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  D. H Abbott, D. K Barnett, J. E Levine, V. Padmanabhan, D. A Dumesic, S. Jacoris, and A. F Tarantal Endocrine Antecedents of Polycystic Ovary Syndrome in Fetal and Infant Prenatally Androgenized Female Rhesus Monkeys Biol Reprod, July 1, 2008; 79(1): 154 - 163. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) ± SEM (vertical line). Group 1 males (n = 4) in A and C were treated with testosterone (T) from the time of Cx (30-day T); group 2 males (n = 4) in B and D were treated with T for 9 days beginning on Day 21 after Cx (9-day T). See Results for other details.
