Inhibition of Testosterone Conversion to Dihydrotestosterone in Men Treated Percutaneously by Progesterone

Comments

PIERRE MAUVAIS-JARVIS, FREDERIQUE KUTTENN, AND NICOLE BAUDOT Biochemical Department, Faculty of Medicine Pitie-Salpetriere, 75634 Paris, France

ABSTRACT. Unlabeled progesterone was added topical treatment and incubated with the same at different doses to human skin homogenates procedure as during in vitro experiments. An inhi- originating from the pubic area of normal men. bition of testosterone 5a-reduction to dihydrotes- The transformation of radioactive testosterone to tosterone was also observed after this percuta- dihydrotestosterone was inhibited by this in vitro neous administration of progesterone. Thus pro- addition of progesterone. In other experiments gesterone, when percutaneously applied upon the 100 mg of progesterone dissolved in alcohol was skin of human beings, might be useful as an applied upon the skin of 2 normal men. A skin antiandrogen. (/ Clin Endocrinol Metab 38: 142, specimen was obtained before and after this 1974)

TJECENT studies upon metabolism in human ically, the inhibition of testosterone 5a-reduction •*•*• skin have demonstrated the transformation by progesterone might be considered as an anti- of several steroid hormones in this organ and androgenic effect of endocrinological and derma- there is increasing evidence that these transfor- tological significance, since progesterone admin- mations are of endocrinological importance. In istered percutaneously is largely metabolized in particular, the 5a-reduction of testosterone to situ (8) before reaching the peripheral circula- 17(3-hydroxy-5a-androstan-3-one (dihydrotestos- tion. terone, DHT) which seems to be an essential In this study, the results of experiments relat- prerequisite for the action of testosterone in ing the inhibition by progesterone of testoster- male accessory organs (1,2) has been also ob- one 5a-reduction by human skin are presented. served in vitro in human skin (3,4). In this lab- These studies have been carried out both in oratory (5,6), the reduction'of testosterone to vitro by addition of unlabeled progesterone to DHT and 5a-androstane-3q,17(3-diol (andro- skin homogenates incubated with radioactive stanediol) by human skin wjas also suggested, testosterone and "in vivo plus in vitro" by incu- since in normal man, testosterone was a better bation of skin from normal men previously sub- precursor of androstanediol when the labeled tes- mitted to a percutaneous treatment with proges- tosterone was administered .through the skin terone. than when the radioactive precursor was injected Both experiments resulted in an inhibition of in the peripheral circulation. i the 5a-reduction of testosterone to DHT. . On the other hand, the Sa-reduction of progesterone to 5a-pregnane-3,20-dione (dihydropro- Materials and Methods gesterone) was observed both in vitro (7) and Chemicals in vivo (8) by comparing, as for testosterone, Testosterone-1,2-:{H (SA 42.75 C/mmole) was the fate of radioactive progesterone simulta- obtained from New England Nuclear Corp. and neously administered by percutaneous and intra- purified in the laboratory. Its purity was found to venous routes. Moreover, Voigt, Fernandez and be at least 95% after chromatography and isotope Hsia (9) have found that a number of steroids dilution. Reference steroids were purchased from inhibited the 5a-reduction of testosterone by Roussel Uclaf and Ikafarm laboratories; glucose-6- microsomal preparation of human,skin; among phosphate, glucose-6-phosphate dehydrogenases, these, the most potent inhibitor was progester- NADF and NADPH were obtained from Boehringer one which competed effectively with testosterone chemical corp. for the active site of the enzyme. Thus theoret- Specimens Six normal men were studied. They were 30-40 Received June 18, 1973. yr old. In 4 cases pubic skin samples were obtained 142

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 February 2015. at 10:44 For personal use only. No other uses without permission. . All rights reserved. COMMENTS .143 at laparotomy for urologic nonendocrine diseases. reduction of testosterone to DHT, the identification None of the patients studied had renal failure. In of the other metabolites of testosterone was not the 2 other cases, skin was obtained after a topical performed. The chromatographic systems used in application on the same pubic area of 100 mg of this study are described in Table 1. Dried extracts unlabeled progesterone dissolved in absolute ethanol were submitted to a first paper chromatography. (prepared by Besins-Iscovesco laboratories). The Reference steroids were visualized by the Zimmer- total application of progesterone was fractionated mann reaction. Labeled testosterone, DHT and into 4 percutaneous applications of progesterone for androstenedione were located by scanning with a 24 hr between 2 biopsies of pubic skin performed radiochromatogram scanner Packard model 7200 and under saline anesthesia. All skin specimens were im- eluted. An aliquot of each spot was counted in a mediately stored at —20 C, approximately for 2 liquid scintillation spectrometer Packard model weeks, because NADPH does not seem to penetrate 3376. The remaining steroid material was purified the fresh skin samples (10). by 2 thin-layer chromatographies and 1 paper chromatography, as described in Table 1; and identi- Homogenization of skin samples fied by addition of the unlabeled carrier steroid The general procedure employed was very simi- and recrystallization to constant specific activity lar to that described by Voigt, Fernandez and Hsia (Table 2). (9). Approximately 100 mg of skin was used The final specific activity of material from the after removal of subcutaneous fat. Skin samples area corresponding in the first paper chroma- were minced with fine scissors and homogenized in tography to testosterone and androstenedione was Krebs Ringer phosphate buffer with an ultra 95% of the calculated value. Turrax until there was no visible piece of intact In the ligroin-propanediol chromatographic sys- skin and then transferred into a Tenbroeck homog- tem, DHT and epiandrosterone have a very close enizer. All operations were carried out at —4 C mobility. So, the material of this peak was examined for less than IS min. by reverse isotopic dilution with DHT and epiandrosterone. The data indicated that 85% of the DHT Incubation peak was effectively DHT (Table 2) and 15% could be epiandrosterone. The same rate had been obtained The medium consisted of 10 ml of Krebs Ringer by Gomez and Hsia (3). Phosphate buffer (pH 7.4) with addition of Peni- r> cillin G, 50 U/ml; glucose anhydre, 10~ mole/ml Results and either NADPH 1 nmole/ml or NADP 1 n-mole/ ml with or without glucose-6-phosphate, 5 M-mole/ml Effect of cofactors (Table 3) and glucose-6-phosphate dehydrogenase, 1 U/ml. In one experiment only 0.79% and 0.13% of Immediately prior to incubation, radioactive tes- radioactive testosterone was recovered as DHT tosterone (15 X 10(i dpm = 158 pmoles) in 0.1 ml and androstenedione in the absence of cofactors. of ethanol was added to each flask and evaporated In the presence of NADP the transformation of to dryness. The incubation was carried out for 1 hr at 37 C in a Dubnoff shaking incubator. In one case testosterone into DHT and androstenedione was the incubation was performed for 1, 2 and 3 hr. not significantly greater. By contrast, in the • presence of a NADPH generating system, there Extraction was more androstenedione-(22.9%) than DHT (8:4%) formed from radioactive testosterone. All incubations were terminated by addition of acetic acid in order to have a pH close to 5.0. The optimal reduction of testosterone to DHT Mixtures were then kept in a freezer at —10 C was obtained when NADPH was added to the overnight. Steroids were extracted with ethyl acetate, medium (10.5 to 15.3% of the incubated radio- washed with sodium carbonate (10%) and water, active testosterone). In that case, only 1.8 to then evaporated to dryness under vacuum. Non- 6.1% of testosterone was oxidized to androstene- steroid lipids were separated from steroids with dione. 709? aqueous methanol/heptane system. The re- covery of tritium in the methanol phase was greater TABLE 1. Solvent systems used in chromatography than 90%. Paper partition chromatography: Chromatography and identification of steroids 1. Ligroin:Propanediol 1:1 2. Cyclohexane:Benzene 1:1 In the present experiment, only testosterone, DHT and androstenedione were identified and mea- Thin-layer chromatography: sured. The essential object of this study being the 1. Ethyl acetate:Benzene 1:1 inhibitory action of progesterone upon the 5a- 2. Chloroform:Methanol 99:1

TABLE 2. Characterization of testosterone, andros- 00 r^ tenedione and D H T I •—I CM \O 1) Testosterone Specific x No. activity Solvents crystallization dpm/nmole Isooctane-methanol 1540 o-o Dichloromethane-hexane 2 525 Ethyl acetate-benzene 3 530

80 mg of unlabeled testosterone were added to an aliquot containing 150,000 dpm from the area of the O O O O peak corresponding to testosterone. The calculated specific activity (SA) was 541 dpm/nmole. The data indicate that 95% of the peak was effectively unme- tabolized testosterone.

2) Androstenedione a Specific E No. activity

Solvents crystallization dprn/nmole f-4 — OC t/j O Isooctane-benzene 1651 O '~- Dichloromethane-hexane 2 633 p 0.8 Ethyl acetate-benzene 3 625

25 mg of unlabeled androstenedione were added to O io an aliquot containing 55,000 dpm of the peak of q o androstenedione. The calculated SA was 639 dpm/ d d M-mole. The data indicate that 95% of the peak was effectively androstenedione.

3) DHT Specific No. activity Solvents crystallization dpm/mnole Isooctane-benzene 1356 2 >^ Dichloromethane-hexane 2 302 o a Ethyl acetate-benzene 3 300 o

40 mg of unlabeled DHT were added to an aliquot containing 43,300 dpm from the area of the peak corresponding to DHT. The calculated SA was 350 CM ^H dpm/M-mole. Data indicate that 85% of the peak was effectively DHT. T3

Effect of temperature H The reduction of testosterone proceeds op- X rsi t^ >—I f> CM f*5 timally at 37 C. When skin was heated to 80 C, Hogg no reduction of testosterone to DHT occurred, o even in the presence of NADPH. Thus, it was % concluded that the enzymatic activity is sup- -2 pressed at high temperature. £

Time course of testosterone 5a-reduction ^ U h hr< ^ Q . w 1/3 t/5 C/l

TABLE 4. Inhibition of testosterone 5a-reduction in homogenates of human skin from pubic area of normal men incubated for 1 hr with l,2-:1H-testosterone + NADPH (other metabolites of testosterone than DHT were not included in this table) DHT recovered after incubation Type of Progesterone % of total radioactivity experiment added pmole/100 mg/hr recovered I. In vitro Case 1 0 19.6 13.6 1 mg 5.3 4.2 100 ng 2.3 2.0 Case 2 0 19.7 15.3 10 Hg 3.0 2.5 1 M« 3.9 3.0 II. In vivo (percutaneous) Case 1 9 12.4 10.0 100 mg 3.3 2.9 Case 2 0 15.1 12.8 100 mg 2.5 1.3

I. In vitro experiments: addition to incubates of decreasing doses of progesterone. II. In vivo experiments: percutaneous application of 100 mg of progesterone dissolved in ethanol to 2 normal men before collection of skin. tion. Indeed, a skin specimen was fractionated one on the pubic area of two normal men leads into three samples incubated respectively for 1, to similar results as during in vitro experiments. 2 and 3 hr with radioactive testosterone and The fall in the conversion of testosterone to NADPH. The percent conversion of radioactive DHT recorded after percutaneous administra- testosterone to DHT was approximately 12% tion of progesterone was respectively 70.3 and for 1 hr, 10% for 2 hr, and 8% for 3 hr. Also, all 86.5% of the percent conversion noted before the other experiments were carried out for only 1 hr. treatment.

Effect of unlabeled progesterone incubated in Discussion vitro with radioactive testosterone (Tables 4 and 5) The work of Strauss and Pochi (11) on the stimulation by androgens of sebaceous gland de- All skin specimens were incubated with 15 X 3 velopment has demonstrated that the skin was 10° dpm of H-testosterone in absence and in a target organ for androgens. The transformation presence of respectively: 1 mg, 100 \ig, 10 u.g and of testosterone to DHT by the skin has been 1 |ig of unlabeled progesterone. By comparison with the control experiments, a very important fall in the conversion of testosterone to DHT TABLE 5. Percent of inhibition of testosterone was noted when progesterone was added at dif- 5a-reduction by progesterone either incubated with ferent levels to the skin incubate (65 to 83% of testosterone (in vitro) or percutaneously administered to normal men (in vivo) inhibition). However, in the present experimen- tal conditions, the inhibition of testosterone 5a- % inhibition of DHT reduction by progesterone does not appear as formation from dose dependent. In particular, the fact that the Addition of progesterone testosterone conversion of testosterone to DHT was higher In vitro after addition of 1 mg of progesterone than with 1 mg (case 1) 69.1 lower amounts remains unclear. 100 ng (case 1) 85.0 10 M£ (case 2) 83.7 Percutaneous administration of unlabeled pro- 1 |xg (case 2) 80.2 gesterone (Tables 4 and 5). In vivo The incubation of skin samples recovered be- 100 mg (case 1) 70.3 fore and after a 24-hr application of progester- 100 mg (case 2) 86.5

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 February 2015. at 10:44 For personal use only. No other uses without permission. . All rights reserved. JCT. & M 1974 146 COMMENTS Vol 38 • No 1 further localized in sebaceous glands and hair gards the enzyme-coenzyme substrate relation- follicles (10,12). In addition, in the skin as well ship, even if NADPH is the required coenzyme as in male accessory organs, this enzymatic reac- from the 5a-reduction of testosterone or proges- tion requires the presence of NADPH (3,4,9). terone. Moreover, a relationship between testosterone More recently, Voigt and Hsia (19) have re- 5- by Sansone and Reisner (13) showing that the carboxylic acid. This compound dissolved in skin of patients with acne vulgaris compared to acetone, and topically applied to the flank skin of normals was more effective in reducing organs of hamsters inhibited locally the enlarge- testosterone to DHT. ment of the sebaceous glands provoked by tes- Most of the antiandrogens used at the present tosterone, whereas the 5u-reduction of testoster- time are known to interfere with the cytoplasmic one to DHT and androstanediol was blocked. protein binding specifically DHT in male acces- This study confirms the possibility of the inhibi- sory cells (14,15), thus preventing the transfer tion of testosterone Sa-reduction by compounds of the DHT-cytoplasmic protein complex to the topically applied. The steroid used by this group nuclear chromatin (16). However, at least in (19) was interesting because it had no known male accessory organs, the cytoplasmic protein hormonal activity and was not metabolized in has a particularly high binding affinity for DHT situ by the Sa-reductase. However, Voigt and and not for testosterone (14). As a result, the Hsia did not mention if such a compound may 5androgen action. Also, the ministered. Indeed, no data are given upon the inhibition of this metabolic event may be consid- absorption rate of 4-androsten-3-one-17(3-car- ered on different points of view as antiandro- boxylic acid through the skin. By contrast, the genic. Indeed, Dorfman (17) has previously total absorption of progesterone through human showed that the inunction of the chick comb skin has been calculated both in vitro (20) and with progesterone inhibited its stimulation by in vivo (8) and averages 10% of the adminis- testosterone whereas progesterone diminished the tered dose. The clinical use of progesterone as growth of male accessory organs of mice when an antiandrogen was criticized by Voigt and administered subcutaneously with testosterone. Hsia (19) because this steroid might be a pre- In addition, Dorfman (17.) suggested that pro- cursor for active androgens in testis or adrenal gesterone might be a naturally occurring anti- after its passage in blood stream. Our results androgen which can prevent, during fetal life, showing that more than 70% of progesterone the deleterious effect of androgen on the hypo- topically applied was metabolized into the hu- thalamic anterior pituitary-gonadal axis of fe- man skin do not agree with this assumption. In males. The inhibition of testosterone Sa-reduc- addition, the metabolic products of progesterone tion by progesterone was first demonstrated by 5a-reduction into the skin (essentially dihydro- Voigt, Fernandez and Hsia (9) in human skin progesterone and allopregnanediol) are not microsomes and after by Massa and Martini known as having an androgenic effect. Moreover, (18) in homogenates of rat hypophysis and no trace of these compounds was recovered in prostate. This inhibition was interpreted by the blood after percutaneous administration of Voigt and Hsia as the result of a competition progesterone (unpublished data). between progesterone and testosterone for the In conclusion, whatever the exact site and the active site of the Sa-reductase (9). However, this modality by which progesterone exerts its inhib- competition has not been classified as fully com- itory activity upon the reduction of testosterone petitive. In the present study, the results ob- to DHT, the fact that this steroid hormone acts tained in vitro with homogenates of pubic skin identically when it is topically applied to human from normal men were essentially qualitative skin and when it is added in vitro to skin homog- and cannot give additional information on the enates provides evidence that the percutaneous nature of the competitive reaction between tes- administration of progesterone is effective on a tosterone and progesterone. One can only say biological point of view. Thus it may be of par- that reduced nicotinamide nucleotides being ticular interest in several skin diseases such as added in excess during skin incubations, the roie acne vulgaris and seborrhea which are related to of the coenzyme may be partly excluded as re- hyperresponsiveness of the skin to androgens.

Acknowledgments 9. Voigt. W., E. P. Fernandez, and S. L. Hsia, J Biol Chem 245: 5594, 1970. 'Phis work was supported in part by a grant of the Fondation pour la Recherche Medicale Franraise. 10. Takayasu, S., and K. Adachi, J Clin Endocrinol The authors are indebted to Drs. Jardin, Banzet Metab 29: 1580, 1969. and Cady for provision of skin samples. 11. Strauss, J. S., and P. E. Pochi, Recent I'rogr Harm Res 19: 385. 1963. References 12. Takayasu S., and K. Adachi, Endocrinology 90: 73, 1972. 1. Bruchovsky, N., and J. D. Wilson. J Biol Chein 13. Sansone, G., and R. M. Reisner, / Invest 243: 2012, 1968. Dennatol 56: 366, 1971. 2. Anderson, K. M., and S. Liao, Nature 219: 277, 14. Mainwaring, W. I. P., J Endocrinol 45: 531, 1968. 1969. o. Gomez, E. C, and S. L. Hsia, Biochemistry 7: 24, 1968. 15. Liao, S., and S. Fang, Vitam Harm 27: 18, 4. Wilson, J. D., and j. D. Walker, / Clin Invest 1969. 48: 371, 1969. 16. Mainwaring, W. I. P., and B. M. Peterken, 5. Mauvais-Jarvis, P., J. P. Bercovici, and F. / Biol Chem 125: 285, 1971. Gauthier, / Clin Endocrinol Metab 29: 417, 17. Dorfman, R. I., In James, H. T., and L. 1969. Martini (eds.). Proceed of the Third Intl. (>. , O. Crepy, J. P. Bercovici, and F. Congress on Hormonal Steroids, Excerpla Gauthier, / Clin Invest 49: 31, 1970. Medica Found. 1971, p. 995. 7. Frost. P., E. C. Gomez, G. D. Weinstein. J. 18. Massa, R., and L. Martini, Cyn Invest 2: 253, Lamas, and S. L. Hsia, Biochemistry 8: 948, 1972. 1969. 19. Yoigt, W., and S. L. Hsia, Endocrinology 92: S. Mauvais-Jarvis, P., N. Baudot, and J. P. 1216, 1973. Bercovici, / Clin Endocrinol Metab 29: 1580, 20. Feldman. R. J., and H. I. Maibach. / Invest. 1969. Dennatol 52: 89, 1969.