August 1983

7. Kline BE, R usch H P , Baumann CA, Lavik PS: The effect of pyridoxine on tumor growth . Cancer Res 3:825- 829, 1943

8. Miner DL, Miller JA, Baumann CA, R usch H P: The effect of pyridoxine and other B vitamins on t he production of liver cancer with p-dimethylaminoazobenzene. Cancer Res 3:296- 302, 1943

9. Tryfia tes GP , Morris HP: Vitamin B-6 and neoplasia, Vitamin B-6 Metabolism and Role in Growth. E dited by GP Tryfia tes. West­port, Connecticut, Food and Nutrition Press, 1980, pp 173-186

10. Holtz P , Palm D : Pharmacological aspects of vitamin B". P harmacal Rev 16:113-178, 1964

11. Weir DR, Morningstar W A: Effect of pyridoxine deficiency induced by desoxypy:ridoxine on acute lymphatic leukemia of adults. Blood 9:173- 182, 1954

12. P egg AE: Role of pyridoxal phosphate in mammalian polyamine biosynthesis. Biochem J 166:81-88, 1977

13. Murray A W, Froscio M: Induction of ornithine decarboxylase ac­t ivity in epidermis by a tumor promoter: decreased response in mice maintained on a vitamin BG-deficient diet. Biochem Biophys Res Commun 77:693- 699, 1977

14. Baird WM, Sedgewick JA, Boutwell RK: E ffects of phorbol and four diesters of phorbol on the incorporation of trit iated precur ­sors into DNA, RNA and protein in mouse epidermis. Cancer R es 31:1434-1439, 1971

15. Lowe NJ: Ultraviolet light and epidermal polyamines. J Invest Dermatol 77:147-153, 1981

16. Yamada M, Saito A, T amura Z: Fluorometric determination of pyridoxal-5-phosphate and pyridoxal in biological materials. I. P rocedure and application. Chern Pharm Bull (Tokyo) 14:482-484, 1966

17. Grigor KM, von Redich D, Glick D : Fluorometric determination of pyridoxal phosphate in microgram samples of tissue. Anal Bio­chem 50:28- 34, 1972

18. Connor MJ, Lowe NJ : Inhibit ion by retinoids of tape-stripping

0022-202X / 83/ 8101-0139$02.00/ 0 THE J OU RNAL OF I N VESTIGATIVE DERMATOLOGY, 81:139-144, 1983 Copyright © 1983 by The Williams & Wilkins Co.

1).5-3{3-HSD AND SEBACEOUS GLAND FUNCTION 139

provoked ornithine decarboxylase induction. Clin Res 31:145A, 1983

19. Coburn SP, Mahuren J D , Sallay S l: In vivo metabolism of 4'­deoxypyridoxine in rat and man. J Bioi Chern 251:1646- 1652, 1976

20. J ohnson FA, White SL, Halsey ML, Sevigny SWdeJ: Effect of deoxypyridoxine on the amount of pyridoxal phosphate in the livers and leukocytes of rats and on leukocyte number, size, and type. J Nutr 88:51-54, 1966

21. Coburn SP, Mahuren JD, Schaltenbrand WE, Wostmann BS, Madsen D: E ffects of vitamin B-6 deficiency and 4'-deoxy:pyri­doxine on pyridoxal phosphate concentrations, pyridoxine kinase and other aspects of metabolism in the rat. J Nutr 111:391-398, 1981

22. Tryfiates GP: Cofactor regulation of gene product expression. Vi­tamin B-6 effects of enzyme induction and expression of proteins in normal and hepatoma-bearing animals, Vitamin B-6 Metabo­lism and Role in Growth . Edited by GP T ryfiates. Westport, Connecticut , Food and Nutrit ion P ress, 1980, pp 309- 333

23. Hunter JE, Harper AE: Stability of some pyridoxal phosphate­dependent enzymes in vitamin B-6 deficient rats. J Nutr 106:653-664, 1976

24. Litwack G, Rosenfield S: Coenzyme dissociation, a possible deter­minant of short half-life of inducible enzymes in mammalian liver . Biochem Biophys R es Commun 52:181-188, 1973

25. Tryfiates GP: Vitamin B-6 effects on t he growth of Morris hepa­tomas and the development of enzymatic activity, Morris Hepa­tomas. Mechanism of Regulation. Edited by HP Morris, WE Criss. New York, P lenum Press, 1978, pp 607- 642

26. Disorbo DM, Litwack G: Changes in the intracellular level of pyridoxal 5' -phosphate affect the induction of tyrosine amino­transferase by glucocorticoids. Biochem Biophys Res Commun 99:1203-1208, 1981

Vol. 81, No. 2 Printed in U.S. A.

The Relationship Between the In Vitro Activity of 3/J-Hydroxysteroid Dehydrogenase 4 4

-5-Isomerase in Human Sebaceous Glands and Their

Secretory Activity In Vivo*

NICHOLAS B. SIMPSON, M.D. , M.R.C.P ., WILLIAM J . CUNLIFFE, M.D., F.R.C.P., AND

MALCOLM B. HoDGINS, B .Sc. , PH.D.

Departments of Dermatology, the R oyal Infirmary (NBS), Glasgow, Scotland, The General Infirmary (WJ C), Leeds, E ng land, and University of Glasgow (MBH), Glasgow, S cotland

3/J-Hydroxysteroid dehydrogenase .14-5-isomerase (.15-

3/J-HSD) catalyzes an early step in the synthesis of t es­tosterone from dehydroepiandrosterone (DHA). We compared enzyme activity in back skin biopsies with sebum excretion rate (SER) in 14 individuals. The rate of conversion of [7a-3H]DHA into [3H]-4-androstene-3,17-dione was measured in cryostat sections of skin and compared with the sebaceous gland content of the same

Manuscript received August 4, 1982; accepted for publication Decem­ber 23, 1982.

• This work forms part of a thesis submitted by NBS to the Univer­sity of Wales, Cardiff, for the degree .of Doctor in Medicina (M.D .), 1981.

Reprint requests to: Nicholas B . Simpson, M.D ., Department of Dermatology, The Royal Infirmary, Glasgow G4 OSF, Scotland.

Abbreviations: DHA: dehydroepiandrosterone NAD+: nicotinamide adenine dinucleotide SER: sebum excretion rate TLC: thin-layer chromatography

biopsies. Reaction rate was proportional t o the volume of sebaceous gland tissue in the sections . Enzyme activ ­ity was absent from sections without histologically iden­tifiable sebaceous gland tissue. This suggests that the ~5-3/J-HSD is localized in sebaceous glands. SER, m eas­ured by a modified photometric technique at the biopsy site, correlated highly with sebaceous gland volume and with the rate of conversion of DHA into androstenedione in the biopsy. For each biopsy, specific activity of .15-3/J­HSD in sebaceous glands was calculated by dividing the rate of formation of PH]-4-androstene-3,17 -dione by se­baceous gland volume. Specific activity of ~5-3/J-HSD did not correlate significantly with SER, suggesting that variations in concentration of ~5-3/J-HSD in sebaceous glands probably do not underlie variations in sebaceous gland activity.

The weak androgen dehydroepiandrosterone (DHA) stimu­lates sebaceous gland secretion in humans [1,2]. The increase

140. SIMPSON, CUNLIFFE, AND HODGINS

TABLE I. Details of patients No. Sex Age Grade of acne SER (~/cm'/3 h)

1 F 20 Mild 14.2 2 F 21 Mild 36.5 3 F 21 Mild 26.0 4 F 22 Mild 18.3 5 F 28 Moderate 15.5 6 F 28 Moderate 12.25 7 F 29 Moderate 27.5 8 M 18 Severe 33.0 9 M 18 Severe Unobtainable

10 M 18 Moderate 11.8 11 M 19 Mild 17.2 12 M 21 Moderate 35.2 13 M 21 Moderate 24.7 14 M 31 Moderate 16.4

in its plasma concentration in early puberty [3,4] is associated with an increase in sebaceous gland size and secretory activity [5,6] and could be a cause of acne [7]. DHA is converted into the potent androgens, testosterone and 5a-dihydrotestosterone, in human skin (8]. The activity of the 3,8-hydroxysteroid de­hydrogenase .14

-5-isomerase (.15-3/3-HSD) enzyme complex,

which catalyzes an early step in this pathway, might influence the level of "active androgen" in sebaceous glands, thereby controlling sebum secretion [9-13]. In order to test this hypoth­esis, we compared the activity of .15-3,8-HSD in human seba­ceous glands with the in vivo secretory activity of the same glands.

MATERIALS AND METHODS Patients and Clinical Assessment

Patient data are shown in Table I. Fourteen patients took part in the study-7 males and 7 females. Acne was graded according to the involvement of back skin, using a 3-point scale-mild, moderate, or severe (modified from [14]). Informed and signed consent was obtained from all patients and the study was conducted with local ethical committee approval.

Chemicals

[7a-3H]-Dehydroepiandrosterone (13.9 Ci/rnmol) was obtained from the Radiochemical Centre, Amersham, England. Purity was checked by thin-layer chromatography (TLC) on Kieselgel HF60 254 ± 336 (E. Merck and Co., Darmstadt, West Germany) using chloroform/acetone 185/15 (v/v) as solvent. Nicotinamide adenine dinucleotide (NAD+) and androstenedione were obtained from Sigma Chemical Company Limited, Kingston upon Thames, England. Organic solvents of analyt­ical grade were used as supplied.

Measurement of Sebum Excretion Rate (SER)

SER was measured in 13 patients by a modification of the technique of Cunliffe, Kearney, and Simpson [15]. An area of clinically normal scapular skin 1.5 x 1.2 em (1.8 cm2

) with an identifiable pilosebaceous orifice at the center was delineated by adhesive tape. Residual surface lipid was removed by the application of absorbent papers that were changed every 10 min until a pattern of lipid spots corresponding to individual sebaceous orifices (follicular pattern) could be seen. The area was protected from the rubbing of clothing and the patients rested for 3 h during which time only cold drinks were allowed. After 3 h the accumulated lipid was sampled by the application of a finely ground glass plate and assayed by the use of a "Lipometre" (L'Oreal Labora­tories, France). The .reading in millivolts was converted to Jig of lipid/ cm2/3 h by 1·eference to a calibration curve. ·

Finely etched glass plates were used for this study. The sapphire sampling plates of Cunliffe et a\ [15] were found to be inaccurate on back skin as the surface lipid level of back skin is 10-fold less than that on the forehead. A calibration curve (Fig 1) was constructed using skin surface lipid eluted from the foreheads of 6 adult volunteers by diethyl ether; epithelial debris was removed by filtration and the lipid made up in serial dilutions in acetone. Ten microliter aliquots were dispensed onto the 1.8 cm2 plates, giving 10-80 11-g. em - 2 (6 replicates at each dilution). The acetone was allowed to evaporate, the sampling plates placed in the Lipometre, and the appropriate millivolt value recorded.

Vol. 81, No. 2

Biopsy and Sectioning of Tissues

After sampling the skin surface lipid, the delineated area was anes­thetized by subcutaneous injection oflignocaine (1% aqueous). A 4-mm punch biopsy, centered on the central pilosebaceous orifice, was taken and immediately frozen on solid C02• Specimens were stored overnight at -25°C, after which the subcutaneous fat was removed by naked eye dissection, the tissue embedded in ice, and sectioned on a cryostat at -25°C. Serial sections 16-Jiffi thick were cut and examined for the presence of sebaceous gland tissue. When sebaceous gland acini became visible, 10 consecutive sections were cut and alternate sections taken-5 onto a microscope slide and 5 into an incubation tube at - 25°C. The procedure was repeated, until no sebaceous tissue could be identified, using fresh microscope slides and incubation tubes. This yielded 3-7 incubations per biopsy.

Histometry

Tissue sections were fixed in formol saline and stained with hema­toxylin and eosin. The thickness of the tissue section was measured at 3 points by focusing on the extreme cell layers with a calibrated microscope stage, and the surface area measured by drawing using a Leitz Orthoplan microscope with drawing tube attachment and planim­etry at 20x magnification for total section area and 125X magnification for sebaceous gland acini and pilosebaceous duct and sebaceous duct epithelium. One observer was used throughout.

6."-3{1-HSD Assay

The conversion of [7a -3H]DHA into [3H]-4-androstene-3,17-clione (also called [3H]-androstenedione) was measured by a modification of the methods of Shirley and Cook [16] and Chakraborty et al [13]. In the standard enzyme assay, 5 tissue sections were disintegrated by vortexing in 0.2 ml of a solution containing 0.1 JiM [7 a-3H]DHA, 0.6 mM

NAD+, and 0.05 M Tris-HCl buffer, pH 7.6. The mixture was incubated for 60 min at 37°C, then the reaction was stopped by the addition of 2.0 ml chloroform/methanol, 1/1 (v/v) containing nomadioactive DHA and androstenedione (400 Jig of each) as carriers.

Two time courses were performed. In each, 65 tissue sections con­taining sebaceous gland tissue were collected in a Teflon-glass homog­enizer at -25°C and homogenized by hand in 1.0 ml of 0.05 Tris-HCl buffer, pH 7.6, for 10 min at 0°C on ice. Aliquots (0.1 ml) of the tissue homogenate were divided among 10 incubation tubes at 0°C. Reactions were started by adding 0.1 ml of 0.05 M Tris-HCl buffer, pH 7.6, containing 0.2 JiM [7a -3H]DHA and 1.2 mM NAD+. Duplicate incuba­tions were performed in air at 37°C for 0, 15, 30, 60, and 120 min and were stopped as above.

400

300

Ol 200 - ~ 1J ro ~

'w a: ~ 100 2 0 [L

'_J

0

0 20 40

Sebum (>Jg .cm-2 )

60 80

FIG 1. Calibration curve for the "Lipometre." This curve is of the form y = A[1-e-B<• - •C> ], where A (the asymptotic value) = 400, B (the slope parameter) = 0.033, and C (the intercept on the ordinate) = 3.8.

August 1983

TABLE II. Recrystallization of [ 3 H]-4-androstene-3,17-dione to constant specific activity"

Specific activity Solvents

Crystals Supernatant

Acetone/n-hexane Acetone/n-hexane Chloroform/n-hexane Chloroform/n-hexane Acetone/n-hexane

29,465 28,659 28,158 28,745 28,032

" Theoretical sp act: 30,960 dpm/mg.

Extraction and Isolation of Steroids

38,274 31,174 30,765 32,312 28,780

Reaction mixtures were evaporated to dryness at 40°C and parti­tioned between diethyl ether and distilled water 8/1 (v/v). The ether fraction was taken to dryness under an air stream, the residue dissolved in a few drops of ethyl acetate/methanol 2/1 (v/v), and the steroids separated by TLC on silica gel HF60 254-366 using the solvent system, chloroform/acetone 185/15 (v/v). Carrier steroid bands were identified by fluorescence at 254 nm (androstenedione) and 366 nm (DHA), scraped from the plate into small glass columns, and the steroids eluted with 8 ml ethyl acetate/methanol, 2/1 (v/v). The eluates were taken to dryness under an air stream and dissolved in 2.0 ml methanol. Aliquots (0.2 ml) were added to 6 ml of a toluene scintillant containing 0.3% (w I v) 2,5-diphenyloxazole (PPO) and 0.01% (w /v) 1,4-di-2(4-methyl-5-phenyloxazoyl}-benzene (POPOP) , in duplicate. Counting was per­formed in a Packard Tri-carb liquid scintillation spectrometer model 3214 with a counting efficiency of 42% for tritium.

Statistical Analysis

Linear regression has been calculated by the method of least squares. Significance testing of the correlation coefficients and of the difference between the intercept on the ordinate and zero were performed by the t-test. Partial correlation was calculated from the appropriate simple correlation coefficients.

RESULTS

Histometry

There was no histologic evidence of pilosebaceous duct ob­struction in any biopsy. Sebaceous gland area and section t hickness were measured on each section. The average section thickness was 16 J.lm ± 1.2 SD. The volume of sebaceous gland in a set of sections was the product of sebaceous gland area and average section thickness. Because alternate sections were taken for histometry and enzyme assay, it was assumed that t he volume of sebaceous gland in each enzyme assay was very similar to that in the corresponding sections taken for histo­metry. The total volume of the sections from each biopsy ranged from 0.464- 0.949 mm3 (mean 0.633 ± 0.151 SD). There was no significant positive correlation between sebaceous gland volume and section volume, either in any single biopsy or when measurements from all sets of sections were pooled (r = 0.164, 0.1 <p < 0.2, n = 85).

Identification of Reaction Products

After TLC about 3% of the applied radioactivity remained at t he origin of the chromatogram and 96% ± 1% SD was present in the DHA and androstenedione carrier bands. A pooled sample of eluted [3H)-4-androstene-3,17-dione was recrystal­lized to constant specific activity from several solvents [8]. Table II shows that [3H)-4-androstene-3, 17 -dione was the major metabolite recovered from the incubations. The identity of e HJ-4-androstene-3,17-dione formed by incubation of [7a-3H] DHA with cryostat sections of human skin and NAD+ has been previously established [13].

Reproducibility of the Enzyme Assay

The average recovery of radioactivity after the ether/water partition stage was 86.5% ± 5.0% SD. Radioactivity recovered in DHA and androstenedione fractions after TLC averaged

6.5 -3(3-HSD AND SEBACEOUS GLAND FUNCTION 141

E' 6 a. 0 .._, w z 0 0 w 4 z w f-(f)

0 a: 0 z <! 2

I M ~

X

i 9 0

•

control

..... ::: g:::::::::.a:::::::::::::::::::-g:::::::::::::::::::·.::::::::·. :::::::::::~~

0 15 30 60 120

Time (min)

FIG 2. Time courses for conversion of [7a-3H]DHA into [3H]andro­stenedione by pooled skin sections for 2 different individuals (solid lines) with appropriate no-tissue controls (brol1en lines).

25

2

15

1()

5

a

25

b.

• • • •• - . ' ...... ·::~-'f.l. . . . ~ ·--- .­.,. .

• •

• •• •

Sebaceous gland volume mm3

12

Q.t

Sebaceous gland volume mm3

FIG 3. The rate of conversion of [7a-"H]DHA into ["H]androstene­dione as a function of sebaceous gland volume: (a) in 85 individual incubations (r = 0.69; p < 0.001); (b) as regression lines for 14 biopsies.

:c E' a. s Q) c 0

~ c Q)

~ "E <t

J. )( ..,

IQ

142 SIMPSON, CUNLIFFE, AND HODGINS

66.5% ± 6.9% SD of starting radioactivity. No correction for procedural losses was applied in calculating the quantity of CHJ-4-androstene-3,17-dione.

Time Course of Conversion of {7a- 3H]DHA into ( JH]-4-androstene-3, 17-dione

Fig 2 shows 2 time courses and their respective controls over 120 min. Reaction velocity is constant for up to 120 min and 20% conversion of [7 a-3H]DHA into CHJ-4-androstene-3,17-dione.

Rate of Conversion of [7a-3H]DHA into fH]-4-androstene-3,17-dione Compared with Sebaceous Gland Content of Skin

Fig 3a shows the comparison between rate of [3H]-4-andros­tene-3,17-dione formation and sebaceous gland volume for the 85 individual incubations from 14 biopsies. Fig 3b shows the regression lines for comparison of rate of CHJ-4-androstene-3,17-dione formation with sebaceous gland volume in the sets of 4-7 incubations from each biopsy; the appropriate correlation coefficients are shown in Table III. In all except 3 incubations the quantity of [3H]-4-androstene-3,17 -dione formed in 1 h fell within the linear part of the time course. It is clear that the rate of conversion of [7a-3H]DHA into androstenedione is di­rectly proportional to sebaceous gland volume. In only 1 biopsy was the extrapolated intercept of the regression line at the ordinate significantly different from zero. In several incubations

TABLE III. Significance levels for the correlations between rate of conversion of [7a-3H} DHA into {"H]-4-androstene-3,17-dione and

sebaceous gland volume in each of 14 biopsies"

Patient No. of Correlation incubations coefficient (r)

1 7 0.83 (p < 0.05) 2 7 0.95 (p < 0.001) 3 5 0.64 (0.1 < p < 0.5) 4 4 . 0.38 (0.6 <p < 0.7) 5 7 0.70 (0.05 < p < 0.1) 6 7 0.58 (0.1 < p < 0.5) 7 7 0.72 (0.05 < p < 0.1) 8 5 0.48 (0.1 <p < 0.5) 9 7 0.92 (p < 0.01)

10 7 0.93 (p < 0.01) 11 6 0.99 (p < 0.001) 12 7 0.90 (p < 0.01) 13 6 0.84 (p < 0.05) 14 4 0.99 (p < 0.01)

" The linear regression lines are shown in Fig 3b.

8·

6 •

4

2

0

0 0·5 lo() 1·5 2 ·0 2 ·5 3 -Q 3-5

10 x Sebaceous g land vo lume cmm3J

FIG 4. The rate of conversion of [7a-~H]DHA into [3H]androstene­dione in a biopsy as a function of sebaceous gland volume in 14 biopsies from 14 donors (r = 0.79; p < 0.001).

Vol. 81, No.2

..c ~40 E • • (.) • rn- 30 2 Q) • +-' • ro ~

20 c Oo • 0 • ·~ •

~ • • (.) 10 --X Q)

E :J 0 .0

~ 0 0·5 10 1·5 2{) 2·5 3·0 3·5

lOx Sebaceous g land volume cmm3J

FIG 5. Sebum excretion rate as a function of sebaceous gland volume in 13 biopsies from 13 donors (r = 0.85; p < 0.001). Biopsies containing a single gland (filled circles); biopsies containing several separate glands (open circles). Correlation for single glands only, r = 0.88 (p < 0.001) .

.J:

.£2. E

4

u • • -Ol ::J.. 30

•

• QJ +-' ro 20 ~ r:J> c 0 • • 0 • +-' 10 ~ u X QJ 0

E :J .0 0 2 4 6 8 Q)

(f) 10-5 x [3HlAndrostenedione ( cprn / h)

FIG 6. Sebum excretion rate as a function of the rate of conversion of [7a-~H]DHA into [3H]androstenedione in a biopsy (r = 0.78; p < 0.01). Biopsies containing a single gland (filled circles); biopsies con­taining several separate glands (open circles). Correlation for single glands only, r = 0.66 (p < 0.05).

without histologically identifiable sebaceous gland in the sec­tions there was no detectable formation of [3H]-4-androstene-3,17 -dione. Similar results were obtained when sebaceous gland volume was calculated as a fraction of total section volume. There was no significant positive correlation betwe~n the rate of CHJ-4-androstene-3,17-dione formation and total section vol­ume in any single biopsy or when results from all incubations were taken together (data not shown) . The results support the view that A5-3,8-HSD of human skin is localized mainly in sebaceous glands.

Fig 4 shows the comparison between the rate of [:JH]-4-androstene-3,17-dione formation in each biopsy (the sum of all incubations from the biopsy) and total sebaceous gland volume . Given that A5-3,8-HSD activity is localized in sebaceous glands, the positive linear correlation without significant intercept in­dicates that, in the group of individuals studied, the specific activity of A5-3,8-HSD (rate of conversion of [7a-3H]DHA into [

3H]-4-androstene-3,17 -dione per unit sebaceous gland volume) is relatively constant, i.e., there is no difference in specific activity between small and large glands.

August 1983

Sebum Excretion Rate

In the limited number of subjects studied there was no correlation between SER and the age, sex, or grade of acne of the patients. There was a highly significant positive correlation between SER and sebaceous gland volume (Fig 5) . Acini be­longing to several separate sebaceous glands were identified in 2 of the biopsies (open circles in Fig 5). When these were excluded a similar correlation was obtained.

Sebum Excretion Rate and .6.5-3/J-HSD Activity

Fig 6 shows that the rate of conversion of [7o:-3H]DHA into [

3 H]-4-androstene-3,17-dione in a biopsy is positively correlated with SER measured in vivo over the biopsy site. The correlation remained if the 2 biopsies with several sebaceous glands were omitted. For each biopsy the specific activity of .6.5-3/J-HSD in the sebaceous glands was calculated by dividing the rate of formation of [3H]-4-androstene-3,17-dione by sebaceous gland volume. Fig 7 shows that there is no significant correlation between the specific activity of .6.5-3/J-HSD and SER. Again, omission of the 2 biopsies with several sebaceous glands did not ch ange the result. The same conclusion was reached by calcu­lation of the partial correlation coefficients for SER and the rate of [3H]-4-androstene-3,17-dione formation from the simple correlation coefficients in Figs 4, 5, and 6, which gave a value T1.2.3 = 0.23.

DISCUSSION

Studies of body site variation in [13,17] and embryonic de­velopment of [18] DHA metabolism in human skin suggest a sebaceous gland localization of .6.5-3/J-HSD-a conclusion drawn from histochemical studies of skin hydroxysteroid dehydrogen­ases [9- 12,19]. However, the preferred 3/J-hydroxysteroid sub­strate for the histochemical detection of hydroxysteroid dehy­drogenase activity in human sebaceous glands is 3fJ-hydroxy-5o:-androstan-3-one [9,12]; D. 4-5 -isomerase cannot be detected histochemically. Thus, even though biochemical analysis has demonstrated that DHA is converted into .6.4 -steroids in whole skin this does not necessarily confirm a sebaceous gland local­izati~n of the complete 3/J-hydroxysteroid dehydrogenase .6.4-

5-

isomerase complex. Nevertheless, it has now been shown that microdissected human sebaceous glands can convert DHA into .6.4-steroids and that in forehead skin the sebaceous gland is the main site of this reaction [20]. The present finding of direct

.r: i2... NE 40 u Ol

:::l.

~ 3 0 ~ c .Q 20 Q) u ~ 10

E ::J .0 Ql

(/) 0

• • • • • •

• oo • • • •

2:0

I0-6x1 3

H) Androstenedione/sebaceous gland v o lume

(cpm / h / mm3 )

FIG 7. Sebum excretion rate as a function of the specific activity of t."-3,8-HSD in sebaceous glands (r = 0.06; 0.8 < p < 0.9). Biopsies containing a single gland (filled circles); biopsies containing several glands (open circles). Correlation for single glands only, r = 0.12; 0.7 <p < 0.8) .

~5-3,8-HSD AND SEBACEOUS GLAND FUNCTION 143

proportionality between the rate of conversion of DHA in_to androstenedione and sebaceous gland content of sets of senal sections within individual biopsies further strengthens the case that .6.5-3/J-HSD of back skin is a sebaceous gland enzyme. While this may turn out to be true for other body sites, conversion of DHA into .6.4-steroids has been demonstrated in plucked terminal hair follicles [21,22] and in axillary apocrine glands [20].

It appears that the concentration of enzyme in sebaceous glands (sp act = rate of androstenedione f?r~a~ion + s~baceous gland volume) did not differ among those ~divtduals wtth small and those with large sebaceous glands (Ftg 4). Yet there was a significant conelation between the rate of androstenedione formation in a biopsy and SER in male and female subjects. This results from the fact that .6.5-3/J-HSD is localized in seba­ceous glands the size of which correlates with SER (Fig 5). When the specific activity of .6.5-3/J-HSD is compared with SER there is no significant correlation. Thus, the present study does not support the view that variation of .6.5-3,8-HSD concentration in sebaceous glands underlies interindividual variations in SER.

.6.5-3,8-HSD could still play an important role in the andro­genic control of sebum secretion. The temporal correlation between secretion of DHA [3,4] and the onset of sebaceous gland activity [5,6] suggests a role for this steroid in sebaceous gland physiology. A related steroid, .6.5 -androstene-3fJ,17 fJ-diol, is also an extremely potent stimulator of sebum secretion in the rat [23]. The sebotrophic effects of DHA and .6.5-androstene-3fJ,17fJ-diol most probably require their conversion into testos­terone or 5a-dihydrotestosterone; the first step in this transfor­mation requires .6.5-3/J-HSD. However, it is likely that the flux through the pathway is controlled by supply of substrate rather than enzyme activity per se. Alternatively, it is possible that induction of .6.5-3,8-HSD in sebaceous glands takes place early in puberty (similarly to late fetal life [18,19]), reaching relatively constant levels in the adult. Thus it may be of interest to extend the present study to a younger age group.

This study also emphasizes the importance (often over­looked) of taking into account the cellular distribution of an enzyme when attempting to correlate its activity in an organ or tissue with some physiologic parameter or with a disease proc­ess. Elevation of skin .6.5-3/J-HSD activity [24] or testosterone Sa-reductase [25,26] in hirsutism and acne could simply reflect the fact that both enzymes in skin are concentrated largely in sebaceous glands [20,27] which may be increased in size in the above disorders [28,29].

The assistance of Mrs. A. Dunlop, Mr. D. Forster, Dr. J. Kearney, and Miss C. Staiton is acknowledged. Dr. D. St. Leger of L'Oreal Laboratories, France provided the Lipometre. The authors are indebted to the late Professor J. A. Milne for his advice and encouragement.

REFERENCES 1. Pochi PE, Strauss JS, Mescon H: The role of adrenocortical steroids

in the control of human sebaceous gland activity. J Invest Der­matol 41:391- 399, 1963

2. Pochi PE, Strauss JS: Sebaceous gland response in man to the administration of testosterone, t.' -androstenedione and dehydro­isoandrosterone. J Invest Dermatol 52:32-36, 1969

3. Sizonenko PC, Paunier L: Hormonal changes in puberty. III. Cor­relation of plasma dehydroepiandrosterone, testosterone, FSH, and LH with stages of puberty and bone age in normal bo:ys and girls and in patients with Addison's disease or hypogonadism or prematw-e or late adrenarche. J Clin Endocrinol Metab 41:894-904, 1975

4. de Peretti E, Forest M: Unconjugated dehydroepiandrosterol!-e plasma levels in normal subjects from b1rth to adolescence m human: the use of ·a sensitive radioimmunoassay. J Clin Endo­crinol Metab 43:982-991, 1976

5. Ramasastry P, Downing DT, Pochi PE, Strauss JS: Chemical composition of human skin surface lipids from birth to puberty. J Invest Dermatol 54:139-144, 1970

6. Pochi PE, Strauss JS, Downing DT: Skin surface lipid composition, acne, pubertal development and urinary excretion of testosterone and 17 -ketosteroids in children. J Invest Dermatol 69:485-489, 1977

144 ALVAREZ ET AL

7. Milne JA; The metabolism of androgens by sebaceous glands. Br J Dermatol 81 (suppl 2):23-28, 1969

8. Hay JB, Hodgins MB: Metabolism of androgens in vitro by human facial and axillary skin. J Endocrinol59:475-486, 1973

9. Baillie AH, Caiman KC, Milne JA: Histochemical distribution of hydroxysteroid dehydrogenase in human sebaceous glands. Br J Dermatol 77:610-616, 1965

10. Baillie AH, Thomson J, Milne JA: The distribution of hydroxyste­roid dehydrogenase in human sebaceous glands. Br J Dermatol 78:451-457, 1966

11. Caiman KC, Muir A V, Milne JA, Young H: Survey of the distri­bution of steroid dehydrogenases in sebaceous gland of human skin. Br J Dermatol 82:567- 571, 1970

12. Muir AV: Studies on the t..5-3,8-hydroxysteroid dehydrogenases in the sebaceous glands of human skin. PhD Thesis, University of Glasgow, 1970

13. Chakraborty J, Thomson J, MacSween MP, Muir A V, Caiman KC, Grant JK, Milne JA: The in vitro metabolism of [7a-3H]dehy­droepiandrosterone by human skin. Br J Dermatol 83:477-482, 1970 .

14. Burton JL, Cunliffe WJ, Stafford L, Shuster S: The prevalence of acne vulgaris in adolescence. Br J Dermatol85:119-126, 1971

15. Cunliffe WJ, Kearney JN, Simpson NB: A modified photometric technique for measuring sebum excretion rate. J Invest Dermatol 75:394-398, 1980

16. Shirley IM, Cooke BA: Metabolism of dehydroepiandrosterone by the separated zones of the human foetal and newborn adrenal cortex. J Endocrinol 44:411-419, 1969

17. Hay JB, Hodgins MB: Metabolism of androgens by human skin in acne. Br J Dermatol 94:123-133, 1974

18. Sharp F, Hay JB, Hodgins MB: Metabolism of androgens by human

0022-202X/ 83/8102·0144$02.00/0 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY, 81:144-148, 1983 Copyright © 1983 by The Williams & Wilkins Co.

Vol. 81, No.2

foetal skin. J Endocrinol 70:491-499, 1976 19. Sharp F: The evolution and distribution of hydroxysteroid dehy­

drogenase activity in human foetal skin throughout gestatwn. Histochem J 10:517-528, 1978

20. Hay JB, Hodgins MB: Distribution of androgen metabolising: en­zymes in isolated tissues of human forehead and axillary skin. J Endocrinol 79:29-39, 1978

21. Fazekas AG, Lanthier A: Metabolism of androgens by isolated human beard hair follicles. Steroids 18:367-379, 1971

22. Fazekas AG, Sandor T: The metabolism of dehydroepiandrosterone by human scalp hair follicles. J Clin Endocrinol Metab 36:582-586, 1973

23. Archibald A: A study of factors influencing sebaceous gland activity in the rat. PhD Thesis, University of Newcastle Upon Tyne, 1973

24. Thomas JP, Oake RJ: Androgen metabolism in the skin of hirsute women. J Clin Endocrinol Metab 38:19- 22, 1974

25. Kuttenn F, Mowszowicz I, Schaison G, Mauvais-Jarvis P: Androgen production and skin metabolism in hirsutism. J Endocrinol 75:83-91, 1977

26. Sansone G, Reisner RM: Differential rates of conversion of testos­terone to dihydrotestosterone in acne and in normal human skin- a possible pathogenic factor in acne. J Invest Dermatol 56:366-372, 1971

27. Takayasu S, Wakirnoto H, Harni S, Sano S: Activity of testosterone 5a-reductase in various tissues of human skin. J Invest Dermatol 74:187-191, 1980

28. Plewig G: Acne vulgaris: proliferative cells in sebaceous glands. Br J Dermatol 90:623-630, 1974

29. Burton JL, Johnson G, Libman L, Shuster S: Skin virilism in women with hirsutism. J Endocrinol 53:349-354, 1972

Vol. 81, No.2 Printed in U.S.A.

The Hea~ing of Superficial Skin Wounds Is Stimulated by External Electrical Current

OSCAR M. ALVAREZ, PH.D., PATRICIA M . MERTZ, B.A., RICHARD V. SMERBECK, B.S., AND

WILLIAM H. EAGLSTEIN, M.D.

Department of Dermatology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, U.S.A.

We studied the effects of direct electric current sup­plied by an energized silver-coated electrode on dermal and epidermal wound healing. Keratome-induced wounds (0.3 mm deep) on the skin of young domestic pigs were treated with either an energized (50- 300 p.A) electrode (DC), an unenergized electrode (placebo), or left untreated. Wounds were excised on days 1-7 after wounding and the epidermis was separated from the dermis. The epidermal sheet was evaluated for reepithe­lialization and the dermis was assayed for collagen bio­synthetic capacity. Dermal collagen production among treatments did not differ markedly on days 1-4 after

Manuscript receiv"ed September 21, 1982; accepted for publication February 2, 1983. ·

This work was supported by a grant from the Sybron Corporation and Pittsburgh Skin and Cancer Foundation.

Reprint requests to: Patricia M. Mertz, Department of Dermatology, University of Pittsburgh School of Medicine, Scaife Hall RC 513, Pittsburgh, Pennsylvania 15261.

Abbreviations: CFU: colony-forming units DC: direct electric current HTso: healing time for 50% of wounds TCA: trichloroacetic acid

wounding. However, a highly significant increase (p < 0.001) in the collagen synthetic capacity was observed on days 5, 6, and 7 in wounds treated with DC. There was no significant difference in collagen synthesis among treatments when collagen production was corrected for DNA content. The rate of wound epithelialization was also significantly accelerated (p < 0.05) in DC-treated wounds. These results suggest that the proliferative and/ or migratory capacity of epithelial and connective tissue cells involved in repair and regeneration can be affected by an electrical field.

Recently the effect of electrical stimuli on tissues has been the subject of considerable experimental attention. Cells have complex electrical systems which are sensitive to electrical field changes. Varied effects, both specific and general, have been reported following electrical field alterations. Metabolic, behav­ioral, physiologic, and immunologic changes, in a wide variety of systems including bacteria, insects, plants, and mammalian cells in culture have been observed after exposure to electric fields [ 1-7].

Electrically stimulated bone healing and regeneration has been reported by many investigators [8]. Direct current has