Cell Biology and Toxicology, Vol. 8, No. 4, 1992 267

TESTOSTERONE METABOLISM IN AN IN VITRO SKIN MODEL S A N D R A R. S L I V K A Advanced Tissue Sciences La Jolla, California

The metabolic activity of skin is important in penetration of topically applied compounds. Currently, animal or cadaver skin is used to evaluate the relationship between metabolism and penetration. In the present study, testosterone metabolism and penetration in a three-dimensional human skin model consisting of keratinocytes and fibroblasts derived from neonatal foreskins was characterized. Pieces of the model were placed on tissue culture inserts with HEPES-buffered medium on the dermal side. Penetration of [3H]testosterone was faster at 32 °C than 4 °C suggesting that metabolism affected penetration. To evaluate this metabolism, [3H]testosterone was applied to the stratum corneum side of the skin model. Radiolabeled metabolites released into the medium after incubation were separated by HPTLC and analyzed by autoradiography. This skin model metabolized [3H]testosterone to both more polar and non-polar compounds which were similar to metabolites of neonatal foreskins. The appearance of non-polar compounds was earlier than the appearance of polar compounds. Both dermal fibroblasts and differentiated epidermal keratinocytes contributed to the metabolism of testosterone. Two testosterone metabolites, dihydrotestosterone and androstane-3, 17 diol, were reduced by addition of the cytochrome P-450 inhibitor metyrapone and were only produced by the keratinocytes. In conclusion, this model is a reproducible source of metabolically active skin and therefore a good alternative to animal or cadaver skin for evaluation of the contribution of metabolism to penetration. INTRODUCTION

Metabolism by skin, the largest organ in the body, inactivates many endogenous or exogenous compounds or alternatively produces more biologically active compounds (Price, 1975; Bickers et al., 1982). Also, metabolism by skin can alter the percutaneous absorption of topically applied compounds (Kao et al., 1985). The degree to which metabolic activity

1. Address all correspondence to: Sandra R. Slivka, c/o Elizabeth Whalen, Advanced Tissue Sciences, 10933 N. Torrey Pines Rd., La Jolla, CA 92037. Tel: (619)-450-5825. 2. Key Words: animal alternative, metabolism, skin model, testosterone. 3. Abbreviations: CHC13, chloroform; HHBSS, HEPES buffered Hank's Balanced Salt Solution; HPTLC, high performance thin layer chromatography; MeOH, methanol. Cell Biology and Copyright © 1992

Toxicology, Vol. 8, No, 4, pp. Princeton Scientific Publishing ISSN: 0742-2091

267-276

Co.,

Inc.

268 Slivka

plays a role in penetration and metabolic fate of topically applied compounds is dependent on the viability of the tissue substrate. An in vitro skin model with reproducible metabolic activity can be useful for studying the metabolism of various compounds by skin. Our laboratory has developed a skin model as an alternative to animal or cadaver skin for toxicology testing and penetration studies (Slivka et al., 1993). This three-dimensional skin model consists of keratinocytes and fibroblasts derived from human neonatal foreskins. To produce a dermal model, fibroblasts were seeded onto nylon mesh and grown for 33 days until a physiological dermal-like matrix consisting of collagen, proteoglycans, and glycosaminoglycans was formed. No exogenous matrix proteins are added to these cultures. Keratinocytes were seeded on to the dermal model and grown at the air/liquid interface until a morphologically differentiated epidermis consisting of basal, spinous, granular and stratum corneum layers were formed. The epidermis of this model was shown to express appropriate differentiation markers including K1 (67 Kd) keratin, filaggrin, inv01ucrin, and ceramide types III and IV. The epidermis of this model provides a barrier which discriminates the polarity of various compounds. This model is currently used as an alternative to animals for evaluation of the toxicity and irritancy potential of topically applied compounds (Slivka and Zeigler, 1993). In this study, this model was used to test the effect of metabolic activity on penetration of testosterone. The testosterone metabolites were evaluated by HPTLC. Using the P-450 inhibitor metyrapone, metyrapone sensitive and insensitive metabolites were found. Additionally, using this model dermal and epidermal contributions to this metabolism were evaluated. MATERIALS AND METHODS

The growth of the skin model used here has been described previously (Slivka et al., 1993). Conventional histology was performed. This model can be purchased laser cut into 13 x 13 mm squares from Advanced Tissue Sciences (La Jolla, CA, Skin 2 Model ZK 2000). For studies on percutanous absorption and metabolism the skin model squares are placed epidermal side up on Millicells (Millipore, Bedford, MA) in 6-well culture dishes in a manner similar to the static permeation assay previously described (Slivka et al., 1993). One ml of HEPES Buffered Hanks Balanced Salt Solution (HHBSS) was placed beneath the millicell. When metyrapone, a P-450 inhibitor (Mukhtar et al., 1987) was used, 100 [xM metyrapone was added to the HHBSS 3 hr prior to the application of testosterone. When epidermis and dermis of the skin model were evaluated separately, the epidermis was removed from the dermis by thermolysin treatment. This treatment cleanly separates dermal fibroblasts from epidermal keratinocytes (Slivka and Zeigler, 1993). Thermolysin (Sigma, St. Louis, MO) (0.2 mg/ml) was incubated with the skin model square for 1 hr and the epidermis was removed with a cell scraper. The epidermis square or dermis was then placed on Minicells.

Cell Biology and Toxicology, Vol. 8, No. 4, 1992 269

To begin the permeation or metabolism assay, a 9.5 x 9.5 mm applicator pad (Whatman 3M) was placed centrally on the epidermal side of the skin model and 1-5 l.tCi of [3H]testosterone (Dupont NEN, Boston, MA ) in 20 ~tl of HHBSS was placed onto the applicator pad. The dishes were covered and placed in a 32°C incubator. For experiments at 4 °, the 6-well dishes were placed on ice 15 min prior to starting the experiments and the incubation was carded out on ice in a refrigerator. To study the permeation of topically applied compounds, 50 ~tl aliquots were removed from the underside of the Millicell at the indicated times and replaced with HHBSS (50 ktl). The aliquots were counted in a liquid scintillation counter and the data were expressed as percent dose permeated. To evaluate testosterone metabolism, the HHBSS (entire 1 ml) containing testosterone metabolites which have permeated the skin model following the incubation was removed and extracted into methylene chloride (3 ml). The extract was vortexed and the aqueous and organic phase was allowed to separate (15 min). The aqueous phase (HHBSS) was removed and the methylene chloride phase evaporated to dryness by heating under a stream of N2. The samples were the applied to HPTLC plates (Whatman #4806-421). The plates were then developed in standard TLC tanks containing chloroform:methanol 98:1 or 99:1. The HPTLC plate was removed when the solvent front reached the top and air dried. Standards for metabolites were the following radiolabeled compounds; androst-4-ene-3,17-dione, androstane-3,17-diol, androsterone, dihydrotestosterone (Dupont NEN). For autoradiography, the plate was sprayed with Enhance (Dupont NEN Nef 970) and exposed to Kodak XAR film (Sigma, St. Louis, MO). The films were developed in Kodak GBX developer and fixer (Sigma, St. Louis, MO) according to manufacturer's directions. Autoradiograms were scanned by densitometry in a Shimadzu Scanner (Shimadzu, Kyoto, Japan). RESULTS

In the three-dimensional skin model, topically applied [3H]testosterone was absorbed into the medium faster at 32°C than at 4°C (Fig. 1). To determine if the temperature dependence of absorption was due to metabolism, the percutaneously absorbed radioactivity was analyzed by thin layer chromatography (Fig. 2). The skin model metabolized testosterone into both non-polar testosterone and polar metabolites. The metabolites observed had identity to those produced when neonatal foreskin was incubated with testosterone. No metabolites were observed in the medium at 4°C (data not shown). The time and application dependence of this metabolism was evaluated (Fig. 3). When testosterone was applied to the epidermal side of the model, the non-polar metabolites were observed in the medium earlier than the polar metabolites. When the mesh was placed on the

270

SHvka

tissue culture insert epidermal side down and the testosterone applied to the dermal side, the polar metabolites were not well absorbed into the medium.

0

1

2

TIME (hr)

FIGURE 1. Comparison of Testosterone Permeation at 4°C and 32°C. [3H]Testosterone (1 IxCi) was applied to the epidermal side of the skin model. The effect of temperature on permeation was assessed by observed the percent dose in the medium as a function of time.

F I G U R E 2. Testosterone Metabolism by the Skin Model and Neonatal Foreskin. [3H]Testosterone (5 ~Ci) was applied to the epidermal side of the skin model. Following a 24 hr incubation at 32°C, the medium was extracted and evaluated for testosterone and metabolites by HPTLC (CHCL3:MeOH 98:2) followed by autoradiography.

Cell Biology and Toxicology, Vol. 8, No. 4, 1992 271

FIGURE 3. Time Dependence of Testosterone Metabolism. [3H]Testosterone (5 ktCi) was applied to the epidermal or dermal side of the skin model (skin model was placed upside down on the millicell). Following a 1, 2, or 3 hr incubation at 32°C, the medium was collected from replicate samples, and testosterone and its metabolites were analyzed by HPTLC (CHCL3:MeOH 98:2) followed by autoradiography. The reproducibility of these cultures was examined. At 3 hr of incubation, four different cultures with different keratinocyte lines were assayed in duplicate and scanned by densitometry. The non-polar fraction (two bands near solvent front), the dihydrotestosterone/ androsterone bands, polar bands near the origin, and unmetabolized testosterone represented 21.5+1.0, 31.2+6.4, 27.5+5.6, and 2.9+0.9 percent of the total metabolites (±S.D.}, respectively. The in vitro skin model was incubated with 100 ~tM metyrapone, a cytochrome P-450 enzyme inhibitor (Mukhtar et al., 1987), for 3 hr. Following this incubation testosterone was applied and the mesh incubated for an additional 3 hr. In the presence of metyrapone the production of dihydrotestosterone and androstane-3,17-diol was markedly reduced. The production of other metabolites, androsterone and androstane-3,17-dione and the most polar metabolites, was not decreased by metyrapone. To evaluate the dermal and epidermal components of metabolism, the model was separated into epidermis and dermis with thermolysin. The epidermis and dermis were each incubated with [3H]testosterone and the commercially available radiolabeled metabolites of testosterone (Fig. 4). Each metabolite was then tested on the epidermis and dermis of the skin model independently.

272

Slivka

F I G U R E 4. Metabolism by the Epidermis and Dermis of the Skin Model. [3H]Testosterone or [3H]testosterone metabolites (5 I.tCi/ml) were incubated on the epidermis or dermis of the skin model. Following a 3 hr incubation at 32°C the metabolites in the medium were analyzed by t-IPTLC (CHCL3:MeOH 98:2) followed by autoradiography.

Cell Biology and Toxicology, Vol. 8, No. 4, 1992 273

The epidermis and the intact skin model (epidermis + dermis) profile was identical to epidermis alone for each metabolite tested and was omitted from Fig. 4 for clarity. As shown previously (Pinsky et al., 1974), dermal fibroblasts however do contribute to skin metabolism (Fig. 4). In general, the dermis alone produced only more non-polar metabolites while the epidermis produced both polar and non-polar metabolites. The dermis metabolized dihydrotestosterone, androstane-diol, androstene-dione, and testosterone to more non-polar compounds whereas there was no metabolism of androsterone. Only the epidermis produced androstane-dione and the most polar metabolites (observed at the origin). Fig. 5 shows that the metyrapone sensitive metabolites, androstane-dione and dihydrotestosterone, were most likely produced as follows: testosterone to dihydrotestosterone to androstanediol. This is apparently correct because the androstenedione was only produced from dihydrotestosterone or testosterone by the epidermis of the Skin model. The conversion of dihydrotestosterone occurred only in the epidermis while its conversion to more polar metabolites occurred in both epidermis and dermis.

FIGURE 5. Evaluation of Metyrapone Sensitive Metabolites. [3H]Testosterone (5 I.tCi) was applied to the epidermal side of the side of the skin model preincubated with (+) or without (-) metyrapone (100 I.tM). Following a 2 hr incubation at 32°C, [3H]testosterone and its metabolites in the medium were analyzed by HPTLC (CHCL3:MeOI-I99:1) and autoradiography. Co-migration with radiolabeled metabolites is as indicated. The proposed pathways for testosterone metabolism in the skin model is shown in Fig. 6 with the contribution by dermis and epidermis as indicated. The most striking, although not unexpected, result is that only the epidermis can produce the androstane-3,17-diol (metyrapone

274

Slivka

sensitive metabolites) and the most polar metabolites at the origin which are not P-450 dependent.

epidermis NADPH 5-a-reductase metyrapone inhibits ~

testosterone

"~

dihydrotestosterone

Active metabolite

Inactive metabolites

epidermis or dermis

epidermis metyrapone inhibits

rostane-,a, ,a-a

4-androsten-3, 17-dione

17b-diol

and 5a or 5b-androstanedione epidermis epidermis

epidermis

or

dermis

¢ 5a-~

rosterone

~

polar r~etabolites

epidermis FIGURE 6. Proposed Pathways Used by Epidermal Keratinocytes and Dermal Fibroblasts in the Metabolism of Testosterone. The data in Figures 4 and 5 were compiled to show the possible pathways of testosterone metabolism.

Cell Biology and Toxicology, Vol. 8, No. 4, 1992 275

DISCUSSION In these studies, the reproducible metabolic activity of an in vitro three-dimensional skin model derived from human tissue was demonstrated. As expected from the work of Kao et al. (1985), this metabolism affected the permeation of testosterone through the skin model. This model metabolized the steroid testosterone to biologically active (dihydrotestosterone) as well as biologically inactive compounds. The pattern of metabolites observed was similar to that observed with human neonatal foreskin. The low standard deviation obtained in metabolism assays performed with four independent cultures suggests that these cultures are a reproducible source of metabolically active skin. The skin model can be used to evaluate the production of P-450 metabolites. The 5-alpha reductase pathway which produces dihydrotestosterone and androstane-3,17-diol was shown here to be sensitive to the cytochrome P-450 inhibitor metyrapone. This technique is applicable to the study of P-450 dependent metabolism of other compounds as has been performed with cultured keratinocytes and mouse skin (Bickers et al., 1982) Using this skin model epidermal and dermal components of metabolism by skin can b e evaluated. The metyrapone sensitive pathways were found only in epidermis. Additionally, the most polar metabolites were produced only by the epidermis. Dermal fibroblasts also metabolized testosterone as has been shown by Pinsky et al. (1974) and are therefore important in the skin model. In conclusion, this three-dimensional skin model is useful for studying metabolism of other substances and the role of metabolism in percutaneous absorption. The major advantage of this model is that it is a reproducible source of metabolically active skin. The use of this model could reduce the need for cadaver or animal skin in toxicology testing. REFERENCES BICKERS, D.R., MARCELO, C.L., DUTTA-CHOUDHURY, T., and MUKHTAR, H. (1982). "Studies on Microsomal Cytochrome P-450, Monooxygenases and Epoxide Hydrolase in Cultured Keratinocytes and Intact Epidermis from BALB/C Mice." J. Pharm. and Expt. Ther. 223:163168. KAO, J., PATTERSON, F.K., and HALL, J. (1985). "Skin Penetration and Metabolism of Topically Applied Chemical in Six Mammalian Species, Including Man: An in Vitro Study of Benzo[a]pyrene and Testosterone." Toxicology and Applied Pharmacology 81:502-516. MUKHTAR, H., ATHAR, M., and BICKERS, D.R. (1987). "Cytochrome P-450 Dependent Metabolism of Testosterone in Rat Skin." Biochem. Biophys. Res. Comm. 145:749-753. PINSKY, L., KAUFMAN, M., STRAISFELD, C., and SHANFIELD, B. (1974). "Lack of difference in testosterone metabolism between cultured skin fibroblasts of human adult males and females." J. Clin. Endocrin. Metab. 39:395-398. PRICE, V.H. (1975). "Testosterone Metabolism in the Skin." Arch. Dermatol. 111:1496-1502. SLIVKA, S.R., LANDEEN, L., ZEIGLER F., ZIMBER, M., and BARTEL, R.L. (1993). "Characterization, Barrier Function and Metabolism in an In Vitro Skin Model." J of Invest. Derm. (in press).

276

S~vka

SLIVKA, S.R. and ZEIGLER, F.C. (1993). "Use of an In Vitro Skin Model for Determining Epidermal and Dermal Contributions to Irritant Responses." J. Toxicol.--Cut. and Ocular Toxicol. (in press).

received: 7/20/92 accepted: 10/5/92

Testosterone metabolism in an in vitro skin model.

The metabolic activity of skin is important in penetration of topically applied compounds. Currently, animal or cadaver skin is used to evaluate the r...
470KB Sizes 0 Downloads 0 Views