Evidence exists to suggest that androgens stimulate bone formation in the estrogen-deficient state, however the mechanism of action is unclear. The following study investigates the effect of dihydrotestosterone (DHT) on biochemical markers of bone turnover and calcium homeostasis in sham and oophorectomized (oophx) rats when either vehicle, 40, 80, or 160 mg/kg body weight (bw) DHT were administered at the time of operation or at 15 weeks postoperation. Serum alkaline phosphatase (ALP) increased following DHT administration in sham and oophx rats in all groups (mean ALP ± SEM [U/l] week 8; sham vehicle, 40 ± 7; sham 160 mg DHT/kg bw, 72 ± 5; oophx vehicle, 60 ± 6; oophx 160 mg DHT/kg bw, 88 ± 11) (p < 0.001). In contrast, serum osteocalcin was significantly suppressed in oophx rats administered DHT 15 weeks following operation (mean osteocalcin ± SEM [μg/l] week 8; oophx vehicle, 17.6 ± 3.5; oophx 160 mg DHT/kg bw, 10.5 ± 1) (p < 0.01). Urine deoxypyridinoline was significantly decreased when DHT was administered 15 weeks postoophorectomy (p < 0.001); however, urine hydroxyproline was not affected by DHT treatment in any group. Urine calcium was decreased by DHT treatment (mean Ca/Cr ± SEM week 8; sham vehicle, 0.87 ± 0.13; sham 160 mg DHT/kg bw, 0.24 ± 0.08; oophx vehicle, 0.68 ± 0.16; oophx 160 mg DHT/kg bw, 0.45 ± 0.1) (p < 0.005) which was associated with an increase in the renal tubular reabsorption of calcium (p < 0.05). This study demonstrates the direct effects of DHT on both bone cell activities and the renal handling of calcium.
ANDROGENS PARTIALLY RESTORE bone in both postmenopausal women and gonadectomized male and female rats.1–8 While androgen receptors have been identified on human osteoblast-like cells,9 it is unclear whether these responses are an effect of androgens on bone cells to stimulate bone formation or inhibit bone resorption. Alternatively, they may exert an indirect effect on bone by increasing muscle mass and thereby stimulating mechanical forces on the skeleton. Some androgens may be converted within the target cell to estrogen and thus effects may also be exerted via activation of the estrogen receptor.
In postmenopausal osteoporotic women, the synthetic androgen nandrolone decanoate increases forearm mineral density and calcium absorption with no effect observed on the biochemical marker of bone resorption, urine hydroxyproline excretion.4 Such data suggest that androgens act to increase bone mineral density by increasing bone formation. In contrast, Johanssen and coworkers found nandrolone combined with oral calcium supplementation, also in postmenopausal osteoporotic women, increased bone mineral content with no concominant rise in bone formation. They suggested that androgens have a positive effect on bone by inhibiting bone resorption in addition to increasing muscle mass.1
The oophorectomized (oophx) rat has been used extensively as a model of postmenopausal bone loss.10 Adrenalectomized female rats suffer bone loss similar to the oophx rat which is reversed by treatment with nandrolone decanoate.11 5α-dihydrotestosterone (DHT) partially restores cancellous bone volume in osteopenic oophx rats, reflecting a net gain of bone tissue rather than the prevention of further bone loss.5,8 Furthermore in female rats, the administration of the androgen antagonist flutamide induces bone loss independent of estrogen status.12
Such data demonstrate the importance of androgens for skeletal integrity in females as well as males. The mechanism of androgens on bone, however, remains controversial and to differentiate between an antiresorptive action and stimulation of osteoblast activity we have investigated the effects of DHT on biochemical indices of bone turnover and calcium homeostasis in sham and oophx rats. Unlike testosterone, DHT is not metabolized to estrogen and therefore the effects observed following such treatment are due to activation of the androgen receptor.
MATERIALS AND METHODS
Seventy-two 8-month-old female Sprague-Dawley rats (265–357 g) were obtained from the Gilles Plains Animal Resource Centre (Gilles Plains, South Australia). Each rat was meal fed 20 g/day commercial rat chow containing 0.7% calcium and 0.6% phosphorus and 200 U/kg vitamin D (Milling Industries Pty. Ltd., South Australia) and tap water was supplied ad libitum. The animals were housed at 26°C with a 12-h light–dark cycle. All procedures were approved by The Institute of Medical and Veterinary Science and The University of Adelaide animal ethic committees.
Experimental procedure and design
Forty-eight animals were randomly divided into eight groups, and preoperative fasting blood and urine samples were collected to provide baseline biochemistry. The rats were randomly allocated to either a sham or oophorectomy operation performed under halothane anaesthesia and were administered either vehicle (silastic tubing), 40, 80, or 160 mg/kg body weight (bw) DHT (Sigma Chemical Co., Milwaukee, WI, U.S.A.) by silastic implants (Dow-Corning Medical Silastic tubing, Dow-Corning, Midland, MI, U.S.A.) of lengths 1.5, 2.8, and 5.6 cm, respectively, at the time of operation. Twenty-four hour fasting blood and urine samples were collected weekly for 8 weeks.
Twenty-four animals were randomly divided into four groups, and all animals were oophorectomized under halothane anaesthesia. At 15 weeks postoperation, the rats were administered either vehicle (silastic tubing), 40, 80, or 160 mg/kg (bw) DHT as for experiment 1. Twenty-four hour fasting blood and urine samples were collected every 2 weeks for 14 weeks.
Urine volumes were recorded. Urine creatinine, acidified urine calcium, serum alkaline phosphatase, serum alanine aminotransferase, serum calcium, albumin, total protein, and creatinine were analyzed on automated chemical analysers (Kone Progress Plus, Kone Corp., Ruukintie, Finland and Cobas Bio, Roche Ltd., Basel, Switzerland) using manufacturer recommended methods. Serum sodium, potassium, chloride, and bicarbonate were measured on a Fast 4 electrolyte system (Ciba-Corning 664, Medfield, MA, U.S.A.). Whole blood ionized calcium was measured on a calcium-pH analyzer (Ciba-Corning 634, Halstead Essex, U.K.). Urine hydroxyproline was measured by the method of Bergman and Loxley.13 Free urine deoxypyridinoline was measured by competitive enzyme immunoassay using a commercially available kit (Pyrilinks-D, Metra Biosystems Inc., CA, U.S.A.). Serum osteocalcin was measured by radioimmunoassay as described by Morris et al.14 DHT was measured by radioimmunoassay using a commercially available kit (Diagnostic System Laboratories Inc., Webster, TX, U.S.A.).
Calculations and statistical analyses
Urine calcium, hydroxyproline, and free deoxypyridinoline were expressed as a ratio to urine creatinine. Creatinine excretion (mmol/day) was calculated by multiplying urine creatinine (mmol/l) by 24-h urine volume (l/day). Ultrafiltrable calcium was calculated by the formula described by Morris et al.14 Anion gap was calculated from the difference between the sum of the serum sodium and potassium and the sum of the serum chloride and bicarbonate. The maximum tubular reabsorption of calcium (TmCa) was calculated by the formulas described by Marshall15; however, the calculated serum ultrafiltrable calcium was used as the measure of the filtered load of plasma calcium. Statistical analyses for hydroxyproline/creatinine ratio, alkaline phosphatase, and osteocalcin were performed on the difference between the experimental values and the baseline value to correct for individual variation.
One way analysis of variance (ANOVA) was used to analyze the effect of operation in rats receiving vehicle alone. Multiple comparisons of mean values were made by repeated measures ANOVA. When the assumption of equal correlations between biochemical analyses at each time point was violated, a conservative adjustment to the degrees of freedom was made using the following formula16:
Tukey's post hoc test was used to identify significant differences between mean values.17 All data were analyzed by SAS version 6.10 (SAS Institute, Cary, NC, U.S.A.) on a personal computer. A value of p < 0.05 was considered significant.
Serum DHT levels, which were unaffected by operation, correlated with dose administered in both sham and oophx rats (p < 0.001), and remained elevated throughout the duration of both experiments (Table 1). DHT administration did not significantly affect urine volume or 24-h urine creatinine excretion in either experiment (data not presented), and urine biochemical variables were expressed as a ratio to creatinine to correct for dilutional and bladder emptying errors.
Table Table 1. SERUM DIHYDROTESTOSTERONE LEVELS (pmol/l) MEASURED IN EXPERIMENT 1 AND 2
Serum alkaline phosphatase and osteocalcin were raised as a result of oophorectomy in rats receiving vehicle (p < 0.001) (Figs. 1a, 1b, 2a, and 2b). Alkaline phosphatase was elevated following DHT administration independent of dose (p < 0.001) in both sham and oophx rats. This effect was time dependent (p < 0.001) with maximal stimulation occurring between weeks 2 and 8 after commencing treatment (Figs. 1a and 1b). Osteocalcin was unaffected by DHT treatment in either operation group (Figs. 2a and 2b). There was no difference between the sham and oophx rats following DHT treatment due to an insignificant rise in the sham rats and an insignificant fall in the oophx rats.
Urine hydroxyproline/creatinine and deoxypyridinoline/creatinine were increased as a result of oophorectomy in rats receiving vehicle (p < 0.001, p < 0.01, respectively) (Table 2); however, they were unaffected by DHT administration in either group. Although not statistically significant, DHT treatment resulted in an increase in both urine hydroxyproline and deoxypyridinoline in sham rats such that no statistical difference was observed between DHT-treated sham and oophx rats (Table 2). Urine calcium/creatinine was unaffected by operation in rats receiving vehicle (Figs. 3a and 3b). DHT administration decreased urine calcium independent of dose (p < 0.005) in both sham and oophx rats. This effect was time dependent and occurred at 2 weeks following commencement of DHT treatment (p < 0.005) (Figs. 3a and 3b). TmCa was unaffected by oophorectomy but was increased following DHT administration (p < 0.05) (Table 2) and this effect was not dose dependent but was time dependent with maximal responses occurring at weeks 6–8, following DHT administration (data not presented). Ionized calcium was not affected by oophorectomy but was significantly decreased at all doses of DHT treatment (p < 0.05), (Table 2) with maximal suppression occurring at weeks 6–8 following commencement of DHT administration (data not presented).
Table Table 2. EXPERIMENT 1: BIOCHEMICAL VARIABLES FOLLOWING DHT ADMINISTRATION FROM 0 TO 8 WEEKS POSTOPERATION
Serum creatinine decreased following oophorectomy (p < 0.05) and was decreased by DHT in both sham and oophx rats (p < 0.05) (Table 2). Body weight increased over time (p < 0.005); however, it was unaffected by DHT (data not presented). No relationship was observed between weight and serum creatinine (data not presented). Serum alanine aminotransferase (ALT) and albumin were unchanged throughout the duration of the experiment (mean ALT ± SEM [U/l]; sham vehicle, 17.2 ± 4.7; sham DHT, 14.5 ± 2.6; oophx vehicle, 19.1 ± 4.9; oophx DHT, 20.2 ± 3.0; mean albumin SEM (g/l); sham vehicle, 34.6 ± 0.6; sham DHT, 33.9 ± 0.2; oophx vehicle, 33.3 ± 0.8; oophx DHT 33.8 ± 0.2).
Serum alkaline phosphatase was increased as a result of DHT administration independent of dose (p < 0.05), and this effect was time dependent with maximal serum levels not occurring until 14 weeks after commencing treatment (29 weeks postoperation) (p < 0.001) (Fig. 1c). Serum osteocalcin was decreased following DHT treatment (p < 0.01). This effect was also independent of dose but was time dependent with maximal suppression occurring between weeks 10 and 14 after commencing treatment (25 and 29 weeks postoophorectomy; p < 0.001) (Fig. 2c).
Urine deoxypyridinoline/creatinine was decreased by DHT treatment (p < 0.001), although urine hydroxyproline/creatinine was unaffected (Table 3). The fall in urine calcium/creatinine following DHT treatment did not reach statistical significance (p = 0.13) (Fig. 3c). TmCa was elevated by DHT treatment (p < 0.05) (Table 3) and this effect was time dependent (p < 0.025) (data not presented), but again independent of dose. Serum ionized calcium (Table 3), alanine ALT, albumin, and body weight were unaffected by DHT administration and were unchanged throughout the experiment (mean ALT ± SEM [U/l]; oophx vehicle, 15.7 ± 1.6; oophx DHT, 20.0 ± 1.5; mean albumin SEM (mmol/l); oophx vehicle, 33.8 ± 0.4; oophx DHT, 34.7 ± 0.2). Serum creatinine decreased upon DHT treatment (p < 0.005) (Table 3). No relationship was identified between serum creatinine and body weight (data not presented).
Table Table 3. EXPERIMENT 2: BIOCHEMICAL VARIABLES FOLLOWING DHT ADMINISTRATION FROM 15 TO 29 WEEKS POSTOOPHORECTOMY
DHT treatment in osteopenic oophx rats results in a net gain in bone volume rather than the prevention of further bone loss.5 The acrual of bone is due to an increase in the bone formation rate, with increased surface extent of flurochrome labels resulting in both increased trabecular thickness and number in the tibae5 and stimulation of periosteal bone formation in the femur.8 At high doses of DHT, comparable to those used in the present study, the extent of osteoclast surface and number is reduced.5 It is of interest that at these high doses of DHT the bone formation rate does not differ from oophx control rats although, at least over the short term, bone mineral density is significantly increased.5,8 In the present study, we have demonstrated significant biochemical changes with DHT treatment providing further information on the activities of androgens on bone and calcium metabolism.
An increase of serum alkaline phosphatase was observed in both sham and oophx rats following DHT administration either at operation or 15 weeks postoperation. The response was rapid, with stimulation evident after 2 weeks in the groups receiving DHT at operation but slower in the osteopenic oophx rats with 14 weeks of treatment required to detect stimulation. No changes were detected in serum albumin levels, a protein synthesized by the liver, or alanine aminotrasferase, a liver enzyme, suggesting that the effect of DHT is on bone, rather than liver alkaline phosphatase. This effect is consistent with the induction of alkaline phosphatase by DHT in isolated osteoblast cells in vitro which was blocked by hydroxyflutamide, confirming this activity is mediated by the androgen receptor.18 The resultant levels of alkaline phosphatase in both sham and oophx rats were comparable, suggesting that these dosages provided maximal stimulation of osteoblasts independent of ovarian status.
In contrast, DHT did not affect serum osteocalcin levels in sham-operated rats and decreased osteocalcin in oophx rats only when administered at 15 weeks following oophorectomy. Therefore, a differential effect of this androgen on osteoblast products has been observed. It is proposed that alkaline phosphatase and osteocalcin synthesis reflect different aspects of osteoblastic activity,19 and therefore this discordant effect of DHT on their serum levels may indicate that androgens act at a specific stage of osteoblast maturation. These data are consistent with a model that DHT stimulates osteoblasts at this matrix maturation stage of development when alkaline phosphatase is synthesized without stimulating osteocalcin synthesis.
The effects of DHT on biochemical markers of bone resorption were similar to those observed on serum osteocalcin levels. Urine deoxypyridinoline was significantly reduced in osteopenic rats when DHT was administered 15 weeks postoophorectomy, while urine hydroxyproline excretion was unaffected. Since deoxypyridinoline is considered to be a more specific marker of bone resorption20 it may better reflect such changes compared with hydroxyproline. Nandrolone decanoate, a synthetic androgen, decreases osteocalcin in young and old rats when administered immediately following operation,21 which was attributed to an overall antiresorptive effect.
The differences between the effects of DHT on the biochemical variables when DHT was administered either at the time of operation or 15 weeks postoophorectomy possibly reflects an interaction between DHT and the rate of bone resorption at the time of treatment. In osteopenic rats, the bone turnover rate as measured by osteoclast and osteoblast surface is not significantly different from that of sham rats in the cancellous bone of the tibae at 120 days postoperation.8,22 When DHT was administered at the time of operation the effects on bone resorption may have been masked due to the increased number and activity of bone cells associated with the elevated bone turnover rate following estrogen deficiency.22 In contrast, the osteopenic rats had a lower number of active bone cells at the time of treatment and thus the effect of DHT to decrease bone resorption was detectable. Serum osteocalcin was also decreased by DHT treatment in the osteopenic oophx rats, further supporting this theory.
Urine hydroxyproline, urine deoxypyridinoline, alkaline phosphatase, and osteocalcin were increased as a result of oophorectomy compared with ovary-intact rats receiving vehicle alone consistent with increased bone turnover. It is interesting to note that DHT abolished the effect of oophorectomy on bone turnover because the resultant levels of all bone biochemical variables measured following DHT administration were not significantly different between sham and oophx rats. The major factors contributing to this effect were the greater increase of alkaline phosphatase and the slight but not significant increase in serum osteocalcin and deoxypyridinoline and hydroxyproline excretion in sham rats compared with oophx rats. These results provide further evidence for an interaction between DHT and estrogen which has previously been reported in the oophx rat model8 and requires additional investigation.
Urine calcium excretion was significantly decreased, with DHT administered immediately following operation in both sham and oophx rats, and, although not reaching statistical significance in oophx rats when administered 15 weeks following oophorectomy, this trend was maintained. Nandrolone decanoate administration also decreases urine calcium excretion in postmenopausal women2,3 and in young oophx rats.21 The decrease in urine calcium observed in the present study was associated with an increase in the tubular reabsorption of calcium in the kidney, suggesting a direct action of DHT on the kidney in agreement with the findings of Need and colleagues in postmenopausal women.4 Ionized calcium was significantly decreased in sham and oophx rats administered DHT at the time of operation although no effect was observed when DHT was administered to the osteopenic oophx rats. This may reflect increased calcium incorporation into bone with increased bone formation. The ability of DHT to conserve calcium at the level of the kidney contributes to the availability of calcium for incorporation into bone and the subsequent increase in bone volume.
In summary, this study has demonstrated that even at high doses DHT exerts a stimulatory effect on alkaline phosphatase levels with a decrease in ionized calcium levels consistent with stimulation of bone formation. There was no increase in body weight or serum creatinine levels suggesting that this action was a direct effect on bone cell activity rather than by stimulation of muscle tissue with an indirect effect on bone. DHT also suppressed urine deoxypyridinoline excretion and serum osteocalcin levels in osteopenic oophx rats, indicating an antiresorptive action in female rats. This action appears to be relatively weak since it was not detectable when DHT was administered immediately following either sham or oophorectomy operations. In addition, DHT increased the renal tubular reabsorption of calcium, suggesting a direct effect on the kidney. Since DHT is not metabolized to estrogen, each of these actions must be mediated by the androgen receptor. However, the further understanding of these pleiotropic activities and the possible interactions between DHT and the rate of bone resorption and/or estrogen warrants further study, particularly with respect to the effects of DHT on bone cell gene expression.