| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Stárka
Institute of Endocrinology, CZ 116 94 Prague, Czech Republic
Address all correspondence and requests for reprints to: Martin Hill, Ph.D., Institute of Endocrinology, Národní 8, 116 94 Praha 1, Czech Republic. E-mail: . mhill{at}endo.cz
Abstract
Epitestosterone has been demonstrated to act at various levels as a weak antiandrogen. So far, its serum levels have been followed up only in males. Epitestosterone and its major circulating precursor pregnenolone sulfate and T were measured in serum from 211 healthy women and 386 men to find out whether serum concentrations of epitestosterone are sufficient to exert its antiandrogenic actions. In women, epitestosterone exhibited a maximum around 20 yr of age, followed by a continuous decline up to menopause and by a further increase in the postmenopause. In men, maximum epitestosterone levels were detected at around 35 yr of age, followed by a continuous decrease. Pregnenolone sulfate levels in women reached their maximum at about age 32 yr and then declined continuously, and in males the maximum was reached about 5 yr earlier and then remained nearly constant. Epitestosterone correlated with pregnenolone sulfate only in males. In both sexes a sharp decrease of the epitestosterone/T ratio around puberty occurred. In conclusion, concentrations of epitestosterone and pregnenolone sulfate are age dependent and, at least in prepubertal boys and girls, epitestosterone reaches or even exceeds the concentrations of T, thus supporting its role as an endogenous antiandrogen. The dissimilarities in the course of epitestosterone levels through the lifespan of men and women and its relation to pregnenolone sulfate concentrations raise the question of the contribution of the adrenals and gonads to the production of both steroids and even to the uniformity of the mechanism of epitestosterone formation.
THE 17
-EPIMER OF T (epitestosterone, EpiTe) has previously been considered to be a by-product of the 
5-steroid pathway, without biological significance (1). In 1987, Nuck and Lucky (2) reported the effect of EpiTe on the flank organ of a golden hamster, in which it inhibited the T effect on the pilosebaceous unit; they hypothesized that EpiTe could act as an antiandrogen. This finding encouraged the authors of this study, among others, to investigate EpiTe antiandrogenic effects in vitro as well as in vivo, using various rodent and human models.
Besides the ability of EpiTe to inhibit 5-reductase in rats, reported already previously by others (3), the authors have demonstrated that EpiTe can reduce the weight of androgen-dependent organs in rats and/or mice and compete with synthetic androgen methyltrienolone for ARs in rat prostate cytosol with Ki about half of that of dihydrotestosterone (4). Furthermore, EpiTe appears to act as a competitive inhibitor of the testicular P450C17 enzyme in rats and humans (5, 6) and, at least in rats, influence the secretion and production of LH and FSH (7, 8). Recently, significantly lower EpiTe concentrations were found in hair samples from balding men (9); it was concluded that EpiTe may act as a weak antiandrogen, the efficiency of which could be potentiated by the complexity of its actions at various levels.
However, it is unclear whether circulating EpiTe levels are sufficient to exert antiandrogenic actions. In previous studies the authors have focused on the role of EpiTe and androgens in males with respect to their plausible involvement in the pathogenesis of typical androgen-dependent diseases of older men, such as benign prostate hyperplasia (BPH). It has been shown that in childhood the EpiTe/T ratio is close to or even higher than one but that it decreases sharply during the prepubertal period and puberty, remaining nearly constant in adulthood (10). Similar results have been obtained by others (11). This points to some importance, at least, of EpiTe before puberty in males, when it may attenuate the effects of T.
In contrast to the very low serum EpiTe levels in older men, relatively high concentrations of this steroid have been found in human prostatic tissue from men operated on for BPH. EpiTe concentrations in this tissue were about half those of dihydrotestosterone but twice as high as those of T (12).
In humans, the biosynthesis of EpiTe arises from pregnenolone, a certain portion of which is converted to 5- androstene-3ß,17-diol, which in turn serves as a substrate for 3ß-hydroxysteroid dehydrogenase/4,5-isomerase giving rise to EpiTe, thus avoiding the DHEA and androstenedione involved in the T pathway (13). As has been demonstrated by Dehennin (14), the main sources of EpiTe are testicular Leydig cells, but a certain proportion of the circulating steroid may come from the adrenals.
The majority of circulating pregnenolone is sulfated; though mainly of adrenal origin, it may serve as a supply for EpiTe precursors. Changes of pregnenolone sulfate levels during life in both sexes were reported as early as 1983 by de Peretti and Mappus (15). More recently, Morley et al. (16) followed up pregnenolone sulfate and six other hormonal variables, along with cognitive and physical tests, in a cohort of exceptionally healthy men in search of suitable parameters of predicative value for the development of age-related dysregulations. Taking into consideration that pregnenolone and its more abundant sulfate are indeed the major precursors in EpiTe biosynthesis, it may be of interest whether some relationship between circulating EpiTe and pregnenolone sulfate does exist and, if so, whether it is confined to males and what changes there are in the ratio of the two circulating steroids through life.
The present study provides data on serum EpiTe, pregnenolone sulfate, and T in a large group of male and female subjects of normal population during their lives. The following questions were addressed: 1) What are the levels of circulating EpiTe during life; 2) do they differ according to sex; 3) are serum concentrations of EpiTe sufficient to exert its antiandrogenic actions with respect to actual T levels; and 4) is there any relationship between circulating EpiTe and pregnenolone sulfate?
Materials and Methods
Subjects
The study included 211 females (10–77 yr) and 386 males (1–91 yr), randomly selected in the framework of iodine deficiency screening in the Czech Republic. The subjects with major apparent medical problems such as endocrinopathies, oligo- or amenorrhea in women of reproductive age, or those receiving medications known to affect endocrine status with particular respect to steroid metabolism were excluded. All participants signed informed consent to the use of their blood samples for research purposes.
Steroid determination
Blood was withdrawn from the cubital vein between 0800 h and 1000 h. Not more than 2 h later, the serum was separated and stored in a freezer at -20 C until processed. The serum EpiTe was determined by the method of Bílek et al. (17). Specific antiserum raised against 17-hydroxy-4-androstene-3-one-3-(O-carboxymethyl)-oxime-BSA and radioiodinated epitestosterone-3-(O-carboxymethyl)-oxime tyrosine methyl ester derivative as a tracer were used as reagents. The sensitivity of the analysis in serum was 0.033 nmol/liter, and inter- and intraassay coefficients of variation were 10.9% and 7.7%, respectively. The sample volume was 350 µl. The identity of measured EpiTe was validated by analyzing 15 samples of pooled sera from males and females as follows: serum (1 ml), spiked with 10,000 cpm of [1,2,6,7(n)3H]testosterone (Radiochemical Center, Amersham Pharmacia Biotech, Uppsala, Sweden) was extracted twice with diethyl ether (3 ml); the extract after evaporation was partitioned between 80% methanol with water (1 ml) and 1 ml light petroleum (boiling point 60–80 C). The water-methanol phase after evaporation was subjected HPLC. A binary gradient was used at a constant flow rate of 1 ml/min. Mobile phase A was 15% acetonitrile in water containing 100 mg/liter ammonium bicarbonate. Mobile phase B was methanol. The gradient was as follows: 1-min delay at 40% B, switch to 65% B, and delay up to the 8th minute followed by switch to 100% B and delay up to the 12th minute when the mobile phase was adjusted to 40% B; rinsing proceeded up to the 18th minute. The temperature was kept at 40 C. The column was an ET 250/4, NUCLEOSIL 100-5 C18 from Macherey-N\|[auml ]\|gel (Düren, Germany) The fractions containing EpiTe and T (reverse transcription = 10.4 and 11.6 min, respectively) were collected and following evaporation of the solvent, the fraction containing EpiTe was analyzed by gas chromatography-mass spectroscopy (GC-MS) after derivatization with O-(2,3,4,5,6-pentafluorobenzyl)-hydroxylamine hydrochloride at 60 C for 60 min and subsequent derivatization with a mixture of bis-(trimethysilyl)trifluoroacetamide and trimethychlorosilane (99:1) at 60 C for 45 min. GC separation was carried out with a ZEBRON ZB-50 (15m x 0.25 mm) middle polar capillary column with 0.15-µm film thickness, catalog no. 7e.g.-G004-05, (PHENOMENEX, St. Torrance, CA). The temperature of the injection port was 300 C.
The following protocol was used: temperature gradient: plateau at 120 C (1 min), linear gradient 40 C/min from 120 C to 240 C (3 min), linear gradient 10 C/min from 240 C to 300 C (6 min), and plateau at 300 C (1 min); pressure gradient: high-pressure (pulsed splitless) injection at 60 kPa (1 min), linear gradient 10.5 kPa/min from 30 to 62 kPa (3.05 min), linear gradient 2.6 kPa/min from 62 to 77.5 kPa (5.95 min), and plateau at 300 C (1 min). The duration of the analysis was 11 min.
Responses were recorded in selected ion monitoring mode monitoring the molecular ion (m/z = 555.5). In addition, three different GC gradients were used to check the correct identification of the substances. The detector voltage was set at 2 kV and the sampling rate was 0.25 sec. The temperatures of the interface and the ion source were 310 C and 240 C, respectively.
The results were corrected to losses during extraction and chromatography using recovered [3H]testosterone. In parallel, the samples were analyzed by RIA. The correlation of RIA (y) with GC-MS (x) was expressed by a linear equation y = 0.160 + 0.983 • x with a correlation coefficient of R = 0.924.
T was determined using standard RIA (18), the only modification being that radioiodinated testosterone-3-O(carboxymethyl)oxime tyrosine methyl ester derivative was used as a tracer. Pregnenolone sulfate was determined as has been described elsewhere (19).
Statistical analyses
The dependencies of EpiTe, pregnenolone sulfate, and the steroid ratios on age were evaluated using stepwise polynomial regression. Because of the non-Gaussian distribution in the concentrations of all three measured steroids, the original data were subjected to power transformation (20, 21) to attain the minimum skewness of the studentized residuals. The retransformed 95% confidence intervals of prediction were considered to be the age-dependent limits of the reference range of the steroid. The degree of polynomial was determined using the correlation coefficient of multiple regression adjusted to degrees of freedom, SE of estimation (the square root of the mean squared error), mean absolute error (the average of the absolute values of the residuals), and Akaike information criterion (20, 21). The statistical significance of the model was determined using Fisher’s test. Regression diagnostic plots were used for the detection of outliers and high leverage points (20, 21). However, all outliers and high leverage points excluded from the computation of predictions and confidence intervals were retained in the figures for completeness.
Two-way ANOVA with sex as the first and age as the second factors was used for evaluation of age and sex relationships in steroids and their ratios. The original data were transformed to minimum skewness of residuals to stabilize the group variances and to approximate a Gaussian distribution of the data. For the detection of outliers, an analysis of residuals was used.
The individual differences between the two groups were tested using the method of least significant differences (LSD). The group means with 95% confidence intervals obtained from ANOVA treatment of the transformed data were retransformed to the original scale. All computations described above were performed using Statgraphics Plus version 3.3 software (Manugistics Inc., Rockville, MA).
Pearson’s method was used for the evaluation of mutual correlations between the measured steroids. To avoid nonconstant variance, a non-Gaussian distribution of the data, and to straighten the simple monotonic curvilinear relationships between the variables, power transformation to minimum skewness in each of the two dimensions was applied (20). The principal axis and 95% confidence ellipsoids were computed in Excel 97 (Microsoft Corp., Redmont, WA) using a method described elsewhere (21). The results obtained were retransformed to the original scale.
Results
Serum EpiTe levels in women and men are shown in Fig. 1
. The peak of EpiTe levels in women was found around the 20th year of age (Fig. 1A), and in men it was shifted to around the 35th year (Fig. 1B). In contrast to men, an increasing and accelerating trend starting in menopause and proceeding in the postmenopause is apparent in women. The differences among the mean levels of EpiTe in various age groups are shown in Fig. 2. The confidence intervals of the group means, which are not overlapped, imply statistically significant differences in intergroup mean values.
|
|
. In both sexes a sharp decrease with age occurred, remaining nearly constant in adulthood. As shown in Fig. 4, there was no overlap between confidence intervals in males and females, which means that the differences in each age group were significant.
|
|
. In females (A) only one distinct maximum was found, at around 30 yr, and in males an indistinct maximum was found at around 25 yr; subsequently, it did not change much until approximately 52 yr, after which it decreased continuously again. The differences in individual age groups according to sex are shown in Fig. 6.
|
|
. Although in men (B) the ratio displayed a slight but significant (P < 0.01) increasing trend, in women (A) a U-shaped age dependence was found, with a decreasing trend up to the third decade, from which time a continuous increase followed. The differences were highly significant (ANOVA, P < 0.0001). The differences in individual age groups according to sex are shown in Fig. 8. In men, the increasing overall trend of the EpiTe/pregnenolone sulfate ratio in senescence was still significant but was much less pronounced than in women.
|
|
, EpiTe significantly correlated with pregnenolone sulfate in men, but no correlation was found in women. No significant correlations were found between the levels of EpiTe and T in the two sexes.
|
Hitherto, EpiTe levels in serum have been systematically investigated only in men, with respect to the possible role of this metabolite as an endogenous antiandrogen (9), particularly in relation to its potential role in the pathogenesis of BPH (10, 12). Here, circulating EpiTe and pregnenolone sulfate were measured in a population sample of both men and women from all age groups.
The age dependence curves of the two steroids differed in men and in women. For EpiTe in men, a similar age curve (Fig. 1
) as that reported previously (10) was obtained, with a maximum at around 35 yr. In women the maximum was found about 15 yr earlier, and an increasing trend appeared after the 60th year. Stepwise age-related changes (Fig. 2) revealed significantly higher levels in males with the exception of the prepubertal period and senescence (after the 60th year). With the exception of the prepubertal levels, circulating T levels in adult males are 5 to 10 times higher than those of EpiTe, but in females, from the end of puberty until senescence, the levels of both steroids are close to each other, and in prepubertal girls and postmenopausal women EpiTe is even prevalent (Fig. 3). In both sexes a sharp decline of the EpiTe/T ratio before puberty was found. The sex differences between corresponding age groups were significant (Fig. 4).
Concerning pregnenolone sulfate, the prepubertal and adult levels of this steroid in women were close to those reported by de Peretti and Mappus (15), with a maximum near the 30th year, but in males only an indistinct maximum was found about the 30th year, followed by a slight decline until 55 yr, after which the decline becomes distinct, as reported by others (16).
This study aimed to contribute to the question of whether EpiTe might play the role of an endogenous antiandrogen. The relationship of circulating EpiTe to pregnenolone sulfate may add some information as to whether in men and women the contribution of the adrenals and gonads to the production of EpiTe is similar or whether these sources undergo some changes during the lifespan. Taking into account the fact that EpiTe is bound to AR with one-tenth to one-third affinity as T (5, 22), it may be concluded that in women in reproductive or postmenopausal age, EpiTe could only marginally counteract the effect of circulating androgens. However, in prepubertal children of both sexes, it might function as a factor of hormonal homeostasis. Questions remain as to the role of EpiTe formation and accumulation in target tissues (e.g. in the prostate in which it may be involved in the regulation of intracellular hormone levels and may also provide a contribution to the cellular response).
Another question is whether the mechanism of EpiTe formation proposed by Weusten et al. (13) for EpiTe production in testicular tissue is the only one prevailing in humans or whether other biosynthetic mechanisms are operating in other tissues. Weusten et al. (13) have suggested a mechanism of EpiTe biosynthesis not involving 17-hydroxylase/C17–20-lyase. According to these authors, pregnenolone is directly metabolized to 5-androstene-3ß,17-diol, which in turn serves as a substrate for 3ß-hydroxysteroid dehydrogenase/4,5-isomerase, thus yielding EpiTe. This hypothesis has been indirectly supported by the discovery of a tight correlation between EpiTe and 5-androstene-3ß,17-diol levels in both peripheral and spermatic blood in men (13).
A different situation is seen when the concentrations of EpiTe and its precursor pregnenolone sulfate are correlated. When the EpiTe/pregnenolone sulfate ratio was plotted as a function of age (Fig. 7) or stepwise age-related changes were calculated (Fig. 8), quite different patterns were obtained for men and women: In men, the ratio displayed a slight but significant increasing tendency, but in women a U-shape dependence was found.
Circulating pregnenolone sulfate is mostly if not entirely of adrenal origin (23). The major proportion of EpiTe in men is believed to be formed in the testis (13), even though a recent report has demonstrated that in hypogonadal (but not in eugonadal) men, it responded to ACTH, indicating adrenal participation (11). These facts were a further reason for considering pregnenolone sulfate as well as EpiTe and T to find out whether some relationship between the levels of pregnenolone sulfate and EpiTe does exist.
Although in men a positive correlation between EpiTe and pregnenolone sulfate was found, no such a correlation could be demonstrated in women (Fig. 9). No definite conclusion may be drawn from these results. However, it may be speculated that EpiTe in women is produced by a different mechanism than it is in men or that it is derived by the same mechanism but is metabolized differently. The absence of a correlation between pregnenolone sulfate and EpiTe levels in women excludes a higher contribution by the adrenals. The different mechanism of EpiTe formation means that EpiTe is, at least partially, a by-product of the classical route of androgen biosynthesis, including P450C17 action. The interconversion of T to EpiTe via androstenedione and consequent 17-reduction, which can be seen in some species, must also be taken into account. However, Thijssen et al. (24) have demonstrated that in men the peripheral conversion of labeled androstenedione and labeled T to EpiTe can account for less than 5% of the total urinary excretion of EpiTe.
In a study of the regulation of the P450C17 enzyme in the direction toward 17-hydroxylation (typical of adrenals) or C17–20-cleavage operating predominantly in the testis, Miller et al. (25) pointed to the importance of the availability of reducing equivalents, which depend on electron-donating partners formed by other cooperating enzymes occurring in the respective tissue. Our previous finding that EpiTe acts as a competitive inhibitor of both activities in the human testis (6) indicates that EpiTe is, at least partially, a by-product of T biosynthesis in this tissue.
In conclusion, the authors’ measurements have demonstrated that the concentrations of EpiTe and pregnenolone sulfate are age dependent and that at least in prepubertal boys and girls, EpiTe concentration approaches, or even surpasses, those of T, thus leaving it free to seek its role in hormonal homeostasis. In both sexes, the dissimilarities in the course of EpiTe levels during life, and its relation to pregnenolone sulfate concentrations, raise the question of the contribution of the adrenals and gonads to the production of both steroids and even of the uniformity of the mechanism of EpiTe formation.
Acknowledgments
We express our sincere thanks to Karel Pacák, M.D., D.Sc. (NIH, Bethesda, MD), for his valuable help in preparation of the manuscript.
Footnotes
This work was supported by Grant 5398-3 from the Internal Grant Agency of the Czech Ministry of Health.
Abbreviations: BPH, Benign prostate hyperplasia; EpiTe, epitestosterone; LSD, least significant differences.
Received January 29, 2001.
Accepted January 24, 2002.
References
íková M, Hampl R 1989 Epitestosterone—an endogenous antiandrogen? J Steroid Biochem 33:1019–1021[CrossRef][Medline]
íková M, Hampl R, Stárka L 1992 Epitestosterone—a potent competitive inhibitor of C21-steroid side chain cleavage in the testis. J Steroid Biochem Mol Biol 43:721–724[CrossRef][Medline]
íková M, Hampl R, Hill M, Stárka L 1993 Inhibition of steroid 17 alpha-hydroxylase and C17,20-lyase in the human testis by epitestosterone. J Steroid Biochem Mol Biol 46:515–518[CrossRef][Medline]
ík O, Perheentupa A, Bièíková M, Huhtaniemi I, Hampl R, Stárka L 1994 The effect of epitestosterone on gonadotrophin synthesis and secretion. J Endocrinol 143:353–358íková M, Kanceva R, Lap
ík O, Hill M, Stárka L 1993 The effect of epitestosterone on the plasma levels of LH and FSH in ovariectomized immature rats. J Steroid Biochem Mol Biol 45:321–324[CrossRef][Medline]
ík O, Hampl R, Hill M, Stárka L 1995 Plasma levels of epitestosterone from prepuberty to adult life. J Steroid Biochem Mol Biol 55:405–408[CrossRef][Medline]
ík R, Stárka L 1996 Concentration of the endogenous antiandrogen epitestosterone and androgenic C19-steroids in hyperplastic prostatic tissue. Prostate 28:347–351[CrossRef][Medline]
ák P, Lukeová S, Kozák I, Chrpová M, Stárka L 1978 The use of iodinated steroid as radioligand for testosterone radioimmunoassay. J Steroid Biochem 9:771–773[CrossRef][Medline]
íková M, Számel I, Hill M, Tallová J, Stárka L 2001 Allopregnanolone, pregnenolone sulfate, and epitestosterone in breast cyst fluid. Steroids 66:55–57[CrossRef][Medline]
J, Forina M 1994 Chemometrics for analytical chemistry. vol. 2. PC-aided regression and related methods. Chichester: Horwood
This article has been cited by other articles:
 |
|
 |
|
  P. Liere, A. Pianos, B. Eychenne, A. Cambourg, S. Liu, W. Griffiths, M. Schumacher, J. Sjovall, and E.-E. Baulieu Novel lipoidal derivatives of pregnenolone and dehydroepiandrosterone and absence of their sulfated counterparts in rodent brain J. Lipid Res., December 1, 2004; 45(12): 2287 - 2302. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
, power of the transformation (
