- Email: [email protected]
Aromatase activity in microsomal preparations of human genital skin fibroblasts: Influence of glucocorticoids
J. steroid Biochem. Vol. 33, No. 3, pp. 341-347, 1989 Printed in Great Britain. All rights reserved
0022-4731/89 $3.00 + 0.00 Press plc
Copyright Q 1989Pergamon
AROMATASE ACTIVITY IN MICROSOMAL PREPARATIONS OF HUMAN GENITAL SKIN FIBROBLASTS: INFLUENCE OF GLUCOCORTICOIDS GARY
D. BERKOVITZ,* TAREK BISAT and KATHRYN M. CARTER
Division of Pediatric Endocrinology, The Johns Hopkins Hospital, Baltimore, Maryland 21205, U.S.A. (Received 30 November 1988)
Summary-Skin is an important site of estrogen production in men. Although the aromatase complex in these cells appears to be similar to that of other human cells, the regulation of aromatase by glucocorticoids in cultured human skin fibroblasts is unique. We examined aromatase activity in microsomal-enriched fractions of cultured human skin fibroblasts in order to characterize better the factors that regulate the aromatase in these cells. The optimum pH for aromatase activity in microsomal preparations ranged between 7.0 and 7.5. When androstenedione was the substrate, the mean V,,,ar was 0.58 pmol/mg protein/h (range: 0.09-l .26 pmol/mg protein/h) and the mean K,,, was 27 nM (range: 9-50 nM). When aromatase activity was determined as a function of NADPH concentration, the mean V,,,,, was 0.39 pmol/mg protein/h (range 0.1 l-O.82 pmol/mg protein/h) and the mean K,,, was 180 PM (range: 86-300 FM). For skin fibroblasts exposed to DEX, aromatase activity in isolated microsomes and intact cells was stimulated demonstrating a typical time course with peak levels at 14 h and a decline toward baseline with prolonged (48-60 h) exposure. Cytosol from DEX-stimulated cells did not stimulate the aromatase activity in microsomal-enriched preparations from untreated cells. In addition,
cytosol from cells incubated with DEX for a prolonged period (60 h) did not inhibit the higher aromatase activity of microsomes from cells incubated with DEX for only 14 h. We previously demonstrated that skin fibroblasts incubated with DEX and CHX produced a superinduction phenomenon for aromatase activity. This superinduction of enzyme activity also occurred in the microsomal-enriched fraction and was unaffected by the cytosol of these cells. These studies exclude the possibility that the unique effects of DEX on the aromatase in human skin fibroblasts are due to the production of either inhibitory or stimulatory soluble factors within cytosol.
INTRODUCDON Extraglandular conversion of androgens to estrogens via the aromatase pathway provides 85% of the estrogen production in men [l, 21 approximately half of it occurring in muscle, skin and fat [3]. Consider-
able attention has been focused on the regulation of aromatase activity in cultured adipose tissue stromalvascular cells as they are considered to be an important site of aromatase activity [4-61. We demonstrated that cultured human skin fibroblasts are an important site of aromatase activity and that both the maximal level of enzyme activity, V,,, , and the affinity of the enzyme for substrate, K,,,, are similar to those of adipose stromal-vascular cells [7]. However, despite the similarities in the enzyme complex of the two tissues, the regulation of their aromatase activity differs in several important respects [8]. Notable among these differences is the observation that dexamethasone (DEX) stimulates aromatase activity in skin fibroblasts in a time-dependent *Address correspondence to: Gary D. Berkovitz, M.D., CMSC 3-l 10, Pediatric Endocrinolgy, The Johns Hopkins Hospital, 600 N. Wolfe Street. Baltimore, Maryland 21205 U.S.A.
fashion with peak levels at 12-24 h and subsequent decline to baseline levels. In addition, coincubation of cultured skin fibroblasts with DEX and cycloheximide (CHX) produces a superinduction of enzyme activity when aromatase activity is assayed in the absence of CHX. These observations underscore the unique aspects of the regulation of aromatase activity in skin fibroblasts. We performed studies of aromatase activity in microsomal-enriched fractions prepared from cultured human skin fibroblasts in order to characterize further the aromatase and to understand better its regulation by glucocorticoids. MATERIALS AND
METHODS
Cell culture
Genital skin specimens were obtained at the time of circumcision from ten normal newborn male subjects (NFS l-10). Informed written consent was obtained from mothers. Fibroblasts strains were propagated from explants of skin and the cells were cultured on Minimal Essential Medium with Earle’s salts and L-glutamine supplemented with nonessential amino acids, antibiotic-antimycotic, and 15% fetal bovine serum, as previously described [9]. 341
342
GARY
D.
BERKOVITZ
Preparation of crude subcellular fractions
Crude subcellular fractions of skin fibroblasts were prepared following a modification of previously described methodology [IO]. Confluent monolayers of genital skin fibroblasts were placed on ice, washed three times with Tris-buffered saline (0.01 M Tris-HC1 and 150 nM NaCl, pH 7.4) scraped from culture plates in Tris-buffered saline, collected by centrifugation (1000 g, 15 min) and lysed by sonication in 0.1 M Tris-HCl, pH 7.4. All subsequent steps were performed at 4°C. The cell sonicate was centrifuged (lOOOg, 15 min) to yield a crude nuclear pellet and the resultant supernatant centrifuged (lO,OOOg, 15 min) to produce a crude mitochondrial pellet. The supernatant was centrifuged (100,000 g, 1 h) to yield a microsomal pellet and cytosol. Following each step, the subcellular particles were resuspended by gentle homogenization in 0.1 M Tris-HCI, pH 7.4 using a Dounce homogenizer (pestle A). An aliquot of the homogenate was removed for protein assay [ 1l] and the remainder was reserved for aromatase assay. For routine assays of aromatase activity, genital skin fibroblast sonicates were prepared as described above, and centrifuged (lOOOg, 15 min). The resultant supernatant was subsequently centrifuged (lOO,OOOg,1 h), and the pellet containing mitochondria and microsomes was resuspended as described. Aromatase assay in whole cells
Aromatase activity in confluent monolayers of skin fibroblasts was determined from the generation of [‘H]H20 after incubation of cells with medium containing 150 nM [1-3H]androstenedione (A) for 3 h at 37°C as previously described [7, 121. Progesterone was added (5 PM) to all incubations to prevent loss of substrate via the Sa-reductase pathway [7].
et al.
Incubations were terminated by the addition of 10% trichloroacetic acid (1 ml) and precipitated protein removed by centrifugation. An aliquot of supernatant was extracted in ten vols of chloroform and incubated in a solution of 5% aqueous charcoal. The charcoal solution was centrifuged and an aliquot removed for liquid scintillation counting (efficiency = 37%). Aromatase activity was calculated from the amount of radiolabel in the sample, corrected for recovery and the amount of radiolabel in the l/l position of the substrate. Enzyme activity was normalized for the protein content of the assay aliquot [l I]. Aromatase kinetics using A as substrate were determined by incubating subcellular particles in 0.1 M Tris-HCl, pH 7.4 containing [l-‘H]A (6.25-250 nM), 5 mM NADPH and 5 PM progesterone for 2 h at 25°C. Maximal enzyme activity. I’,,,,,, and substrate affinity, K,,,, were calculated by Eadie-Hofstee analysis of enzyme kinetic data using linear regression analysis. Aromatase activity as a function of NADPH concentration was determined by incubating subcellular particles of 0.1 M Tris-HCl, pH 7.4 containing NADPH (6.25-SOOpM), 150nM [1-3H]A, and 5pM progesterone for 2 h at 25C. V,,,,, and K,,, were calculated as outlined above. Radioactive steroids and materials
[ 1-3H]A (28.2 Ci/mmol), was purchased from New England Nuclear and purified by paper chromatography before use. The percentage of radiolabel in the lp position of the [ 1-3H]A was 78, as determined by New England Nuclear. Radioinert dexamethasone (DEX) was purchased from Sigma (St Louis, MO.). The sources for tissue culture reagents were described previously [9]. solvents used for the various assays were reagent grade or better. RESULTS
Aromatase assay in subcellular particles
Aromatase activity in subcellular particles was determined from the generation of [3H]H,0 when (1-3H]A was converted to estrone according to a modification of methodology previously described [lo, 121. Subcellular particles were prepared as described above and incubated by shaking at 25°C in 0.1 M Tris-HCl, pH 7.4 containing 150 nM [ 1J H]A and 5 mM NADPH. Progesterone (5 p M) was added to all incubations to prevent loss of substrate via the 5x-reductase pathway (71. Incubations routinely contained 0.1-0.8 mg protein in 250~1 buffer. Unless specified otherwise, the incubation period was 2 h. Blank values (background) were established by incubating buffer alone with radiolabeled substrate, progesterone and NADPH, as initial experiments demonstrated that these values were similar to those obtained by incubating microsomes with radiolabeled substrate, progesterone, NADPH, and an excess of radioinert androstenedione (data not shown).
Aromatase activity in subcellular fractions
Subcellular fractions of genital skin fibroblasts (NFS-1) were prepared as described and their aromatase activity determined as shown in Table 1. Aromatase activity in the various subcellular fracTable
1.Aromatase activity in subcellular fractions of human genital skin fibroblasts’
Fraction Sonicate l.OOOg lO,OOOg 100,000 g lOO,ooOg supernatant
Aromatase activity (pmol/h/culture plate)
Protein content (mg/culture plate)
0.26 0.062 0.057 0.050
1.00 0.29 0.17 0.13
0.002
0.42
*Data shown represent the mean of three determinations for aromatase activity and duplicate determinations for protein content. In order to compare the groups, aromatase activity and protein content have been normalized for the amount derived from one 60 mm culture plate.
Skin fibroblast microsomal aromatase
343
0.9 0.8 0.7 0.6 0.5 0.4 -
0.5
1.0
1.5
20
2.5
30
0.1
0.2
Time (h)
0.3
0.4 Protein
0.5
0.6
0.7
0.8
(mg)
Fig. 1. (Left) Time course of aromatase activity. Microsomal-enriched fractions prepared from confluent monolayers of genital skin fibroblasts (NFS-3) were incubated with 150 nM [I-‘HI-A, 5 PM progesterone, and 5 mM NADPH for 0.5-3 h. Aromatase activity was determined from the amount of [3H]H,0 generated. Data shown are values obtained from triplicate determinations. (Right) Protein concentration dependance of aromatase activity. Microsomal-enriched fraction prepared from confluent monolayers of genital skin fibroblasts (NFS-5) were diluted in buffer over a range of protein concentrations and incubated with 150 nM [I-‘HIA, 5 PM progesterone, and 5 mM NADPH for 2 h. Aromatase activity was determined from the amount of [3H]H,0 generated. Data shown are values obtained from duplicate determinations. tions accounted for 65% of the enzyme activity in the cell sonicate. Enzyme activity was similar in the crude nuclear (1000 g), mitochondrial (lO,OOOg), and microsomal (100,000 g) fractions suggesting that the nuclear and mitochondrial preparations contained some residual endoplasmic reticulum. Subsequent assays of aromatase activity were performed using a preparation that included particles from the 10,OOOg and 100,000 g fractions (see Methods). These subcellular particles are referred to as the microsomal-enriched fraction. Optimal conditions for determination of aromatase activity in microsomes
Microsomal-enriched fractions were prepared from genital skin fibroblasts (NFS-2) and resuspended in either buffer alone or in cytosol prepared from the cells. Aromatase activity in the microsomal fraction suspended in cytosol (0.024 pmol/mg protein/h) was similar to that of microsomal fractions suspended in buffer alone (0.018 pmol/mg protein/h). Subsequent
Table
?. Aromatase
kinetics
in microsomal-enriched
using androstenedione V mar Strain
NFS-I NFS-2 NFS-3 NFS-5 NFS-7 NFS-8 NFS-9 NFS-IO Mean + SEM
famolimn
preparations
as substrate
oroteinih)
0.1 I 0.09 0.17 0.84 1.16 0.16 0.84 I .26 0.58 t 0.18
K, (nM)
were performed using microsomal preparations suspended in buffer. In time course experiments using NFS-3 (Fig. 1), aromatase activity in microsomal-enriched preparations increased in a linear fashion between 0.5 and 3 h. Similar results were obtained using another strain of genital skin fibroblasts (NFS-4) aromatase activity being 0.26,0.47,0.67,0.84 and 1.2 pmol/mg protein/h and 0.5, 1, 1S, 2, 2.5 and 3 h, respectively. In subsequent assays, microsomes were incubated for 2 h. In order to validate the aromatase assay over a concentrations, range of protein microsomalenriched fractions were prepared from NFS-5 were diluted in buffer so that incubations contained from 0.10 to 0.73 mg protein. Under these conditions, aromatase activity increased in a linear fashion with respect to protein concentration (Fig. 1). The pH optimum for aromatase activity was determined by incubating microsomal preparations from NFS-6 in 0.1 M T&citrate buffer over a range of pH values from 4.5 to 9. Figure 2 shows a broad peak of aromatase activity between pH 7.0 and 7.5. Similar results were obtained using another strain of genital skin fibroblasts (NFS-7), aromatase activity being 0.14, 0.20, 0.20, 0.18 and 0.14 pmol/mg protein/h at pH 6.8, 7.2, 7.4, 7.6 and 8.0, respectively. All subseexperiments
43 47 50
quent experiments pH 7.4.
II
Determination of aromatase kinetics in microsomal preparations
14 28 I2 9 27 + 6.2
were carried
out at physiologic
Values for V,,, and K, in eight strains of genital skin fibroblasts using A as substrate are shown in
344
GARY
D.
BERKOVITZ et al.
Fig. 2. pH Optimum for aromatase activity. Microsomal-enriched fractions were prepared from confluent monolayers of genital skin fibroblasts (NFS-6) and aromatase activity was determined in 100mM T&citrate buffer @H 4.5-9.0) containing 150nM [3H]A, 5pM progesterone and 5mM NADPH. Data shown were calculated from duplicate determinations.
Table 2. The mean V,,,+ SEM was 0.58 f 0.18 pmol/mg protein/h and the mean K,,,k SEM was 27 k 6.2 nM. Data shown were calculated from duplicate dete~inations of enzyme activity at 8 different concentrations of A. Correlation coefficients for linear regression analysis of Eadie-Hofstee transformed data were 0.90 or better. Values for Y,, and K,,, in six strains of genital skin fibroblasts as a function of NADPH concentration are shown in Table 3. The mean I’,,,., + SEM was 0.39 + 0.14 pmol/mg protein/h and the mean K,,,& SEM was 180 & 38 PM. Data shown were calculated from duplicate determinations of enzyme activity using 8 different concentrations of NADPH. Correlation coefficients for linear regression analysis of Eadie-Hofstee transformed data were 0.73 or better (mean = 0.90). The influence of glucocorticoids matase aciivitj
on microsomal
aro-
Initial experiments were performed to compare the influence of DEX on the aromatase activity in whole cells and in microsomal preparations. Confluent monolayers of genital skin fibroblasts (NFS-8) were incubated for 14 h at 37°C in medium with or without 250 nM DEX. Aromatase activity was determined in triplicate samples of whole cells and of microsomalenriched fractions. Whole cell aromatase activity Table 3. Aromatase kinetics in microsomal-enriched
prepnrations as
(mean f SEM) was 2.6 + 0.09 pmohmg protein/h in untreated cells and 10.4 + 0.28 pmol/mg protein/h in cells incubated with DEX. Aromatase activity in microsomal preparation was also stimulated by incubation with DEX, levels (mean f SEM) being 0.82 k 0.03 pmol/mg protein/h and 4.20 k 0.03 pmol/mg protein/h in microsomal preparations from untreated and treated cells, respectively. These experiments were carried out using two other strains with similar results, the increase in aromatase activity in whole cells and the microsomal-enriched fractions being 18- and 21-fold, respectively in NFS-1, and 2 and 3 fold, respectively in NFS-2. In time course experiments, aromatase activity was determined in triplicate samples of whole cells and microsomal-enriched fractions after incubation of genital skin fibroblasts (NFS-1) with 250nM DEX for 0, 14 or 48 h at 37°C. In whole cells, aromatase activity (mean k SEM) increased from a basal level of 0.29 t 0.05 pmol/mg protein/h to 3.25 + 0.39 pmol/mg protein/h at 14 h and then declined to 1.21 10.35 pmol/mg protein/h at 48 h. Microsomal aromatase activity followed a similar time course, levels being 0.03, 0.34 and 0.19 pmol/mg protein/h at 0, 14 and 48 h. respectively. Whole cell and microsomal-enriched aromatase activity were compared after “superinduction” in the following experiment. Confluent monolayers of genital skin fibroblasts were incubated in medium alone or in medium containing 250nM DEX or 250 nM DEX plus 50 fig/ml CHX for 14 h at 37°C and were then washed three times in Hanks Balanced Salt Sofution (HBSS). For whole cell assays, triplicate plates were incubated with 150 nM rH]A plus 5 FM P for 4 h and aromatase activity was determined. For microsomal-enriched assays, monolayers were incubated with MEM alone for 3 h at 37C, prior to preparation and assay of triplicate microsomal-enriched samples. Table 4 shows the results of such an experiment using NFS-8. Aromatase activity in whole cells and microsomal-enriched preparations was induced by DEX and “superinduced” by DEX plus CHX. Subsequent studies were performed to determine whether cytosolic factors are involved in the stimulation of aromatase activity by DEX. Confluent monolayers of genital skin fibroblasts were incubated for 14 h at 37°C in the presence or absence of 250 nM DEX, and microsomai-enriched fractions and cytosol were prepared. An aliquot of the cytosolic fractions was removed for protein assay and microsomes were
a function of NADPH concentration Vnl””
Strain NFS-I NFS-3 NFS-7 NFS-8 NFS-9 NFS-10 Mean f SEM
Table 4. Influence of DEX and CHX on whole cell and microsomal
(pmol/mg protein~h) 0.1 I 0.12 1.15 0.8 I 0.32 0.82 0.39 + 0.14
aromatase activity 89 150 300 280 200 180 + 38
Sample
Aromatase activity (pmoljmg protein/h) Baseline DEX
DEX plus CHX
Whole cells
1.31 fO.14
4.74 * 0.09
6.44 + 0.41
Microsomes
0.33 + 0.003
I .79 F 0.09
3.49 1_0.09
Data
shown represent the mean of triplicate determinations
*SEM.
Skin fibroblast microsomal aromatase
345
Table 5. Influence of cytosolic factors on DEX-stimulated aromatase activity in microsomal-enriched tions Source of microsomes: Source of cytosol:
Strain
Baseline Baseline 0.12 (0.11-0.13) 0.02 (0.014-0.026)
NFS-I NFS-5
prepara-
Aromatase activity (pmol/mg protein/h) Baseline DEX DEX DEX Baseline DEX 0.16 (0.160.16) 0.03 (0.034.03)
1.40 (1.4-1.4) 0.18 (0.17xl. 19)
I .70 (1.61.8) 0.18 (O.l&o.l8)
Data shown represent the mean and range of duplicate determinations.
in cytosol as shown in Table 5. Aromatase activity was determined in duplicate samples and was normalized for the microsomal-enriched fraction protein content of the sample by subtracting the protein contribution of the cytosol from the total protein content in the sample. In this and all other experiments using cytosol, the microsomes and cytosol in each sample were derived from the same number of cells. Aromatase activity in microsomes from DEX-treated cells was higher than that of untreated cells, and cytosol from DEX treated cells did not enhance the aromatase activity of microsomes. In time course experiments, genital skin fibroblasts (NFS-7) were incubated with medium containing 250 nM DEX for 12 or 60 h at 37°C and microsomalenriched fractions and cytosol were prepared. Aromatase activity was determined in microsomal preparations resuspended in cytosol as shown in Table 6. Activity was determined in triplicate samples and normalized for the protein content of microsomal-enriched preparations. Cytosol from cells incubated for 60 h did not diminish the activity of the microsomal-enriched fraction incubated for only 12 h. The influence of cytosolic factors on the superinduction of aromatase activity by DEX and CHX was examined. Confluent monolayers of genital skin fibroblasts (NFS-3) were incubated with 250nM DEX in the presence or absence of 50 pg/ml CHX for 14 h at 37°C. Cells were washed three times in HBSS and incubated in MEM for 3 h at 37°C and microsomal-enriched fractions and cytosol were prepared. Aromatase activity was determined in microsomal samples resuspended in cytosol as shown in Table 7. resuspended
Activity was determined in triplicate samples and normalized for the content of microsomal-enriched fraction protein. Aromatase activity in microsomal preparations derived from cells incubated with DEX plus CHX was three fold higher than in the microsomal-enriched preparations derived from cells incubated with DEX alone. Furthermore, cytosol from cells incubated with DEX plus CHX did not enhance the aromatase activity in microsomal preparations from cells incubated with DEX alone. DISCUSSION
Human skin fibroblasts grown in cell culture provide a valuable model for studying the factors that regulate aromatase activity in extraglandular tissue [7,8, 131. A variety of similarities exist between the aromatase complex in skin and that of other cultured human cells. In particular, the K, values of the aromatase in human skin fibroblasts [7] are similar to those in intact adipose tissue stromal-vascular cells [4], human placental cells [14], human choriocarcinema trophoblast cells [15], and human fetal hepatocytes [ 161,mean values ranging between 14 and 50nM. Our current study using microsomal preparations permitted us to characterize further the aromatase in human skin fibroblasts and to explore the unique aspects of its regulation. We studied aromatase activity in subcellular preparations containing both microsomes and mitochondria as enzyme activity was similar in subcellular particles contained in the 10,OOOg and 100,OOOg pellets. We validated our assay by demonstrating that aromatase activity increased in a linear fashion with time and with increasing protein concentration. The
Table 6. Influence of cytosolic factors on time course of DEX-stimulated aromatase activity in microsomal-enriched preparations Source of Source of Aromatase (omolimn
microsomes cytosol activity oroteinihl
l2h 12h 0.47 * 0.01
l2h 60 h 0.45 f 0.003
Data shown represent the mean of triplicate determinations
60 h 12h 0.19 * 0.003
60 h 60h 0.22 f 0.006
+ SEM.
Table 7. Influence of cytosolic factors on the superinduction of aromatase activity in microsomalenriched preparations Source of Source of Aromatase (pmol/mg
microsomes cytosol activity protein/h)
DEX DEX I.0 f 0.01
DEX DEX + CHX I .o f 0.02
Data shown represent the mean of triplicate determinations
DEX + CHX DEX 2.7 i 0.02 + SEM.
DEX + CHX DEX + CHX 3.5 f 0.14
346
GARY D. BERKOVITZ et al.
optimum pH for aromatase activity of microsomes from human skin fibroblasts was between pH 7.0 and 7.5. This is in the same range as the optimum pH (7.0) determined for the aromatase of human placental microsomes [17]. Leshin et al., studied estrogen production in cultured skin fibroblasts from the Sebright Bantam chicken as these ceils have extremely high levels of aromatase activity. The optimum pH of microsomal aromatase in these cells ranged from 6.5 to 7.0 when testosterone was used as substrate [lo]. The aromatase complex consists of at least two major microsomal components: an NADPH cytochrome P-450 reductase and a cytochrome P-450 mono-oxygenase [12]. The mean K, of the aromatase in microsomal-enriched fractions derived from cultured human skin fibroblasts as 27 nM when androstenedione was substrate. This value is similar to that obtained by others [l&20] for human placental microsomes (range 2345 nM). By contrast, enzyme kinetic studies of microsomal aromatase in cultured adipose tissue stromal vascular cells demonstrates a low affinity (K,,, = 1030 nM) and a high affinity (K,, = 100 nM) component [21]. This complex pattern was attributed to the presence of lipid in the microsomal membranes. None the less, the Km of the high affinity site is in the same range as that described by us for microsomal preparations of cultured skin fibroblasts. The K,,, values in human skin fibroblast microsomal-enriched preparations are also similar to the Km (2 nM) determined for aromatase in microsomes of Sebright Bantam chicken skin fibroblasts
DOI. The V,,, of the microsomal aromatase in human skin fibroblasts ranged from 0.09 to 1.26 pmol/mg protein/h (mean = 0.58 pmol/mg protein/h). The wide range of values most likely reflects differences in the number of enzyme complexes suggesting that the variation in intact fibroblasts [7] is also due to differences in enzyme content rather than differences in intracellular regulators, cofactor pools, or substrate availability. The mean K,,, for NADPH in human skin fibroblasts was 180 PM. Routine assays contained 5 mM NADPH, insuring saturation of NADPH binding sites. It is of interest that the mean K,,, for NADPH in Sebright Bantam skin fibroblasts was reported to be 0.13 PM, the value being approximately one thousand times lower than the K, in human skin fibroblasts [lo]. Investigation of microsomal aromatase activity also permitted a better understanding of the mechanisms involved in the glucocorticoid stimulation of aromatase activity. We determined aromatase activity in whole cells and in microsomal fractions at a substrate concentration (150 nM A) which is approximately six times the K,. Hence, aromatase activity approximated V,,,,,. Experiments using cytosol from DEX-stimulated and unstimulated cells excluded the involvement of a cytosolic factor in the DEX stimulation Aromatase activity increased in parallel in
whole cells and in microsomal-enriched preparations following DEX stimulation, suggesting that the effect of DEX results from an increase in the number of enzyme complexes. As mentioned earlier, the DEX-stimulation of aromatase activity in skin fibroblasts is unique in several aspects: namely, in the decline in aromatase activity from peak levels after prolonged exposure to DEX, and in the superinduction of aromatase activity by DEX and CHX. In time course experiments, microsoma1 aromatase activity mirrored that of whole cells, demonstrating a peak activity at 14 h and a decline by 48 h. We previously proposed that the decline in aromatase activity with prolonged exposure to DEX might be due to a loss of DEX receptor function, the induction of an inhibitory protein which interferred with aromatase activity or the induction of a repressor protein which interferred with transcription of the aromatase gene [8]. We subsequently demonstrated that DEX did indeed down-regulate the levels of glucocorticoid receptors in genital skin fibroblasts. However, experiments in which we demonstrated a dissociation between changes in DEX receptor levels and changes in aromatase activity suggested that the down-regulation of DEX receptor binding did not account entirely for the decline in aromatase activity [ 131.Our current studies in which microsomal preparations from cells incubated with DEX for either 14 h (short-term) or 60 h (long-term) were mixed with cytosol from these cells demonstrated that the decline in aromatase activity after 14 h was not due to the generation of a soluble inhibitory cytosolic factor. Coincubation of cells with DEX and CHX led to a superinduction of aromatase in whole cells when enzyme activity was assayed in the absence of CHX. In microsomal assays, we incubated cells in medium alone for 3 h prior to aromatase assay. Under these conditions, the aromatase activity of microsomal preparations was also superinduced. Experiments in which we mixed microsomal-enriched preparations and cytosol from DEX-induced and from superinduced cells demonstrated that the superinduction was not due to the presence of a soluble stimulatory cytosolic factor in the superinduced cells. We previously proposed that the aromatase in human skin fibroblasts might be under positive and negative control [8, 131.To explain superinduction we suggested that DEX might induce both aromatase mRNA and a putative aromatase repressor mRNA. However, the presence of CHX would prevent the repressor from being expressed and would permit a greater accumulation of aromatase message. In the absence of CHX, the aromatase mRNA would be translated leading to a greater than usual accumulation of aromatase. Alternatively, the amount of mRNA encoding aromatase in the presence of DEX and CHX might not be different from the amount transcribed in the presence of DEX alone. However, the effect produced by catabolism of the enzyme would be modified since the mRNA is transcribed
Skin fibroblast microsomal aromatase
over 34 h rather than 14 h. Our current work does not permit us to distinguish between these two possibilities. However, experiments using microsomal preparations lend credence to the concept that superinduction of aromatase activity in skin fibroblasts is due to a specific and dramatic increase in aromatase content. Acknowledgements-We thank MS Kathy Klein for her excellent secretarial assistance and Mrs Shirley Ho for her technical assistance. This work was supported in part by USPHS Research Grants ROI-DK-35339 and ROI-DK-00180.
347
Migeon C. J.: Time-dependent biphasic response of aromatase to dexamethasone in cultured human skin fibroblasts. J. c/in. Endocr. Metab. 63 (1986) 468474. 9. Brown T. R. and Migeon C. J.: Cultured human skin
fibroblasts: a model for the study of androgen action. Molec. Cell. Biochem. 36 (1981) 3-22. IO. Leshin M., Baron J.. George F. W. and Wilson J. D.:
Increased estrogen formation and aromatase activity in fibroblasts cultured from the skin of chickens with the henny feathering trait. J. biol. Chem. 256 (1981) 43414344.
Il. Lowry 0. H., Rosebrough N. J., Farr A. L. and Randall
Endocr. Metab. 49 (1979) 905-916. 3. Longcope C., Pratt J. H.,’ Schneider S. H. and Fineberg
R. J.: Protein measurement with the Folin phenol reagent. J. biol. Chem. 193 (1951) 265-275. 12. Thompson Jr. E. A. and Siierti P. K.: Utilization of oxygen and reduced dinucleotide phosphate by human placental microsomes during aromatization of androstenedione. J. biol. Chem. 249 (1974) 5367-5372. 11 Berkovitz, G. D., Carter K. M.,. Migeon C. J. and _-’ Brown T. R.: Down-regulation of the ghtcocorticoid receptor by dexamethasone in cultured human skin fibroblasts: implications for the regulation of aromatase activity. J. c/in. Endocr. Metab. 66 (1988) 1029-1036. 14. Lobo J. O., Bellino F. L. and Bankert L.: Estrogen synthetase activity in human term placental cells in monolayer culture. Endocrinology 116 (1985) 889-895. 15. Johnston J. O., Wright C. L. and Metcalf G. W.: Time-dependent inhibition of aromatase in trophoblastic tumor cells in tissue culture. J. steroid Biochem. 20
E. S.: Aromatization of androgens by muscle and adipose tissue in LVoo.J. clin. Endocr. Metab. 46 (1978)
(1984) 1221-1226. 16. Lanoux M. J., Cleland W. H., Mendelson C. R., Carr
REFERENCES 1. MacDonald P. C., Rombaut R. P. and Siiteri P. K.: Plasma precursors of estrogen. Extent of conversion of plasma androstenedione to estrone in normal males and non-pregnant normal, castrate and adrenalectomized females. J. clin. Endocr. Metab. 27 (1967) 1103-I 1Il. 2. MacDonald P. C.. Madden J. D., Brenner P. E., Wilson J. D. and Siiteri. P. K.: Origin of estrogen in normal men and in women with testicular feminization. J. clin.
146-152. 4. Ackerman
G. E.. Smith M. E., Mendelson C. R., MacDonald P. C. and Simpson E. R.: Aromatization of androstenedione by human adipose tissue stromal cells in monolayer culture. J. clin. Endocr. Metab. 53 (1981)
412417. 5. Simpson E. R.. Ackerman
G. E., Smith M. E. and Mendelson C. R.: Estrogen formation in stromal cells of adipose tissue of women: induction by glucocorticoids. Proc. natn. Acad. Sri., U.S.A. 78 (1981) 5690-5694.
6. Mendelson
C. R., Cleland W. H., Smith M. E. and Simpson E. R.: Regulation of aromatase activity of stromal cells derived from human adipose tissue. Endocrinology 11 (1982) 1077-1085.
7. Berkovitz G. D., Fujimoto M., Brown T. R., Brodie A.
M. and Migeon C.-J.: Aromatase activity in cultured human aenital skin tibroblasts. J. clin. Endocr. Mefab. 59 (198+ 665-67 1. 8. Fujimoto M., Berkovitz
G. D., Brown T. R. and
B. R. and Simpson E. R.: Factors affecting the conversion of androstenedione to estrogens by human fetal hepatocytes in monolayer culture. Endocrinology 117 (1985) 361-368. 17. Ryan K. J.: Biological aromatization of steroids, J. biol. Chem. 234 (1959) 268-272. 18. Covey D. F., Hood W. F. and Parikh V. D.: IO/JPropynyl-substituted steroids. J. biol. Chem. 256 (1981)
1096-1079. 19. Brueggemeier R. W., Snider C. E. and Kimball J. G.: A photoaffinity inhibitor of aromatase. Steroids 40 (1982) 679-689.
20. Burkhart J. P., Weintraub P. M., Wright C. L. and Johnston J. 0.: Novel silyated steroids-as aromatase inhibitors. Steroids 45 (1985) 357-374. 21. Cleland W. H., Mendelson C. R., Simpson E. R.: Aromatase activity of membrane fractions of human adipose tissue stromal cells and adipocytes. Endocrinology 113 (1983) 2155-2160.
0022-4731/89 $3.00 + 0.00 Press plc
Copyright Q 1989Pergamon
AROMATASE ACTIVITY IN MICROSOMAL PREPARATIONS OF HUMAN GENITAL SKIN FIBROBLASTS: INFLUENCE OF GLUCOCORTICOIDS GARY
D. BERKOVITZ,* TAREK BISAT and KATHRYN M. CARTER
Division of Pediatric Endocrinology, The Johns Hopkins Hospital, Baltimore, Maryland 21205, U.S.A. (Received 30 November 1988)
Summary-Skin is an important site of estrogen production in men. Although the aromatase complex in these cells appears to be similar to that of other human cells, the regulation of aromatase by glucocorticoids in cultured human skin fibroblasts is unique. We examined aromatase activity in microsomal-enriched fractions of cultured human skin fibroblasts in order to characterize better the factors that regulate the aromatase in these cells. The optimum pH for aromatase activity in microsomal preparations ranged between 7.0 and 7.5. When androstenedione was the substrate, the mean V,,,ar was 0.58 pmol/mg protein/h (range: 0.09-l .26 pmol/mg protein/h) and the mean K,,, was 27 nM (range: 9-50 nM). When aromatase activity was determined as a function of NADPH concentration, the mean V,,,,, was 0.39 pmol/mg protein/h (range 0.1 l-O.82 pmol/mg protein/h) and the mean K,,, was 180 PM (range: 86-300 FM). For skin fibroblasts exposed to DEX, aromatase activity in isolated microsomes and intact cells was stimulated demonstrating a typical time course with peak levels at 14 h and a decline toward baseline with prolonged (48-60 h) exposure. Cytosol from DEX-stimulated cells did not stimulate the aromatase activity in microsomal-enriched preparations from untreated cells. In addition,
cytosol from cells incubated with DEX for a prolonged period (60 h) did not inhibit the higher aromatase activity of microsomes from cells incubated with DEX for only 14 h. We previously demonstrated that skin fibroblasts incubated with DEX and CHX produced a superinduction phenomenon for aromatase activity. This superinduction of enzyme activity also occurred in the microsomal-enriched fraction and was unaffected by the cytosol of these cells. These studies exclude the possibility that the unique effects of DEX on the aromatase in human skin fibroblasts are due to the production of either inhibitory or stimulatory soluble factors within cytosol.
INTRODUCDON Extraglandular conversion of androgens to estrogens via the aromatase pathway provides 85% of the estrogen production in men [l, 21 approximately half of it occurring in muscle, skin and fat [3]. Consider-
able attention has been focused on the regulation of aromatase activity in cultured adipose tissue stromalvascular cells as they are considered to be an important site of aromatase activity [4-61. We demonstrated that cultured human skin fibroblasts are an important site of aromatase activity and that both the maximal level of enzyme activity, V,,, , and the affinity of the enzyme for substrate, K,,,, are similar to those of adipose stromal-vascular cells [7]. However, despite the similarities in the enzyme complex of the two tissues, the regulation of their aromatase activity differs in several important respects [8]. Notable among these differences is the observation that dexamethasone (DEX) stimulates aromatase activity in skin fibroblasts in a time-dependent *Address correspondence to: Gary D. Berkovitz, M.D., CMSC 3-l 10, Pediatric Endocrinolgy, The Johns Hopkins Hospital, 600 N. Wolfe Street. Baltimore, Maryland 21205 U.S.A.
fashion with peak levels at 12-24 h and subsequent decline to baseline levels. In addition, coincubation of cultured skin fibroblasts with DEX and cycloheximide (CHX) produces a superinduction of enzyme activity when aromatase activity is assayed in the absence of CHX. These observations underscore the unique aspects of the regulation of aromatase activity in skin fibroblasts. We performed studies of aromatase activity in microsomal-enriched fractions prepared from cultured human skin fibroblasts in order to characterize further the aromatase and to understand better its regulation by glucocorticoids. MATERIALS AND
METHODS
Cell culture
Genital skin specimens were obtained at the time of circumcision from ten normal newborn male subjects (NFS l-10). Informed written consent was obtained from mothers. Fibroblasts strains were propagated from explants of skin and the cells were cultured on Minimal Essential Medium with Earle’s salts and L-glutamine supplemented with nonessential amino acids, antibiotic-antimycotic, and 15% fetal bovine serum, as previously described [9]. 341
342
GARY
D.
BERKOVITZ
Preparation of crude subcellular fractions
Crude subcellular fractions of skin fibroblasts were prepared following a modification of previously described methodology [IO]. Confluent monolayers of genital skin fibroblasts were placed on ice, washed three times with Tris-buffered saline (0.01 M Tris-HC1 and 150 nM NaCl, pH 7.4) scraped from culture plates in Tris-buffered saline, collected by centrifugation (1000 g, 15 min) and lysed by sonication in 0.1 M Tris-HCl, pH 7.4. All subsequent steps were performed at 4°C. The cell sonicate was centrifuged (lOOOg, 15 min) to yield a crude nuclear pellet and the resultant supernatant centrifuged (lO,OOOg, 15 min) to produce a crude mitochondrial pellet. The supernatant was centrifuged (100,000 g, 1 h) to yield a microsomal pellet and cytosol. Following each step, the subcellular particles were resuspended by gentle homogenization in 0.1 M Tris-HCI, pH 7.4 using a Dounce homogenizer (pestle A). An aliquot of the homogenate was removed for protein assay [ 1l] and the remainder was reserved for aromatase assay. For routine assays of aromatase activity, genital skin fibroblast sonicates were prepared as described above, and centrifuged (lOOOg, 15 min). The resultant supernatant was subsequently centrifuged (lOO,OOOg,1 h), and the pellet containing mitochondria and microsomes was resuspended as described. Aromatase assay in whole cells
Aromatase activity in confluent monolayers of skin fibroblasts was determined from the generation of [‘H]H20 after incubation of cells with medium containing 150 nM [1-3H]androstenedione (A) for 3 h at 37°C as previously described [7, 121. Progesterone was added (5 PM) to all incubations to prevent loss of substrate via the Sa-reductase pathway [7].
et al.
Incubations were terminated by the addition of 10% trichloroacetic acid (1 ml) and precipitated protein removed by centrifugation. An aliquot of supernatant was extracted in ten vols of chloroform and incubated in a solution of 5% aqueous charcoal. The charcoal solution was centrifuged and an aliquot removed for liquid scintillation counting (efficiency = 37%). Aromatase activity was calculated from the amount of radiolabel in the sample, corrected for recovery and the amount of radiolabel in the l/l position of the substrate. Enzyme activity was normalized for the protein content of the assay aliquot [l I]. Aromatase kinetics using A as substrate were determined by incubating subcellular particles in 0.1 M Tris-HCl, pH 7.4 containing [l-‘H]A (6.25-250 nM), 5 mM NADPH and 5 PM progesterone for 2 h at 25°C. Maximal enzyme activity. I’,,,,,, and substrate affinity, K,,,, were calculated by Eadie-Hofstee analysis of enzyme kinetic data using linear regression analysis. Aromatase activity as a function of NADPH concentration was determined by incubating subcellular particles of 0.1 M Tris-HCl, pH 7.4 containing NADPH (6.25-SOOpM), 150nM [1-3H]A, and 5pM progesterone for 2 h at 25C. V,,,,, and K,,, were calculated as outlined above. Radioactive steroids and materials
[ 1-3H]A (28.2 Ci/mmol), was purchased from New England Nuclear and purified by paper chromatography before use. The percentage of radiolabel in the lp position of the [ 1-3H]A was 78, as determined by New England Nuclear. Radioinert dexamethasone (DEX) was purchased from Sigma (St Louis, MO.). The sources for tissue culture reagents were described previously [9]. solvents used for the various assays were reagent grade or better. RESULTS
Aromatase assay in subcellular particles
Aromatase activity in subcellular particles was determined from the generation of [3H]H,0 when (1-3H]A was converted to estrone according to a modification of methodology previously described [lo, 121. Subcellular particles were prepared as described above and incubated by shaking at 25°C in 0.1 M Tris-HCl, pH 7.4 containing 150 nM [ 1J H]A and 5 mM NADPH. Progesterone (5 p M) was added to all incubations to prevent loss of substrate via the 5x-reductase pathway (71. Incubations routinely contained 0.1-0.8 mg protein in 250~1 buffer. Unless specified otherwise, the incubation period was 2 h. Blank values (background) were established by incubating buffer alone with radiolabeled substrate, progesterone and NADPH, as initial experiments demonstrated that these values were similar to those obtained by incubating microsomes with radiolabeled substrate, progesterone, NADPH, and an excess of radioinert androstenedione (data not shown).
Aromatase activity in subcellular fractions
Subcellular fractions of genital skin fibroblasts (NFS-1) were prepared as described and their aromatase activity determined as shown in Table 1. Aromatase activity in the various subcellular fracTable
1.Aromatase activity in subcellular fractions of human genital skin fibroblasts’
Fraction Sonicate l.OOOg lO,OOOg 100,000 g lOO,ooOg supernatant
Aromatase activity (pmol/h/culture plate)
Protein content (mg/culture plate)
0.26 0.062 0.057 0.050
1.00 0.29 0.17 0.13
0.002
0.42
*Data shown represent the mean of three determinations for aromatase activity and duplicate determinations for protein content. In order to compare the groups, aromatase activity and protein content have been normalized for the amount derived from one 60 mm culture plate.
Skin fibroblast microsomal aromatase
343
0.9 0.8 0.7 0.6 0.5 0.4 -
0.5
1.0
1.5
20
2.5
30
0.1
0.2
Time (h)
0.3
0.4 Protein
0.5
0.6
0.7
0.8
(mg)
Fig. 1. (Left) Time course of aromatase activity. Microsomal-enriched fractions prepared from confluent monolayers of genital skin fibroblasts (NFS-3) were incubated with 150 nM [I-‘HI-A, 5 PM progesterone, and 5 mM NADPH for 0.5-3 h. Aromatase activity was determined from the amount of [3H]H,0 generated. Data shown are values obtained from triplicate determinations. (Right) Protein concentration dependance of aromatase activity. Microsomal-enriched fraction prepared from confluent monolayers of genital skin fibroblasts (NFS-5) were diluted in buffer over a range of protein concentrations and incubated with 150 nM [I-‘HIA, 5 PM progesterone, and 5 mM NADPH for 2 h. Aromatase activity was determined from the amount of [3H]H,0 generated. Data shown are values obtained from duplicate determinations. tions accounted for 65% of the enzyme activity in the cell sonicate. Enzyme activity was similar in the crude nuclear (1000 g), mitochondrial (lO,OOOg), and microsomal (100,000 g) fractions suggesting that the nuclear and mitochondrial preparations contained some residual endoplasmic reticulum. Subsequent assays of aromatase activity were performed using a preparation that included particles from the 10,OOOg and 100,000 g fractions (see Methods). These subcellular particles are referred to as the microsomal-enriched fraction. Optimal conditions for determination of aromatase activity in microsomes
Microsomal-enriched fractions were prepared from genital skin fibroblasts (NFS-2) and resuspended in either buffer alone or in cytosol prepared from the cells. Aromatase activity in the microsomal fraction suspended in cytosol (0.024 pmol/mg protein/h) was similar to that of microsomal fractions suspended in buffer alone (0.018 pmol/mg protein/h). Subsequent
Table
?. Aromatase
kinetics
in microsomal-enriched
using androstenedione V mar Strain
NFS-I NFS-2 NFS-3 NFS-5 NFS-7 NFS-8 NFS-9 NFS-IO Mean + SEM
famolimn
preparations
as substrate
oroteinih)
0.1 I 0.09 0.17 0.84 1.16 0.16 0.84 I .26 0.58 t 0.18
K, (nM)
were performed using microsomal preparations suspended in buffer. In time course experiments using NFS-3 (Fig. 1), aromatase activity in microsomal-enriched preparations increased in a linear fashion between 0.5 and 3 h. Similar results were obtained using another strain of genital skin fibroblasts (NFS-4) aromatase activity being 0.26,0.47,0.67,0.84 and 1.2 pmol/mg protein/h and 0.5, 1, 1S, 2, 2.5 and 3 h, respectively. In subsequent assays, microsomes were incubated for 2 h. In order to validate the aromatase assay over a concentrations, range of protein microsomalenriched fractions were prepared from NFS-5 were diluted in buffer so that incubations contained from 0.10 to 0.73 mg protein. Under these conditions, aromatase activity increased in a linear fashion with respect to protein concentration (Fig. 1). The pH optimum for aromatase activity was determined by incubating microsomal preparations from NFS-6 in 0.1 M T&citrate buffer over a range of pH values from 4.5 to 9. Figure 2 shows a broad peak of aromatase activity between pH 7.0 and 7.5. Similar results were obtained using another strain of genital skin fibroblasts (NFS-7), aromatase activity being 0.14, 0.20, 0.20, 0.18 and 0.14 pmol/mg protein/h at pH 6.8, 7.2, 7.4, 7.6 and 8.0, respectively. All subseexperiments
43 47 50
quent experiments pH 7.4.
II
Determination of aromatase kinetics in microsomal preparations
14 28 I2 9 27 + 6.2
were carried
out at physiologic
Values for V,,, and K, in eight strains of genital skin fibroblasts using A as substrate are shown in
344
GARY
D.
BERKOVITZ et al.
Fig. 2. pH Optimum for aromatase activity. Microsomal-enriched fractions were prepared from confluent monolayers of genital skin fibroblasts (NFS-6) and aromatase activity was determined in 100mM T&citrate buffer @H 4.5-9.0) containing 150nM [3H]A, 5pM progesterone and 5mM NADPH. Data shown were calculated from duplicate determinations.
Table 2. The mean V,,,+ SEM was 0.58 f 0.18 pmol/mg protein/h and the mean K,,,k SEM was 27 k 6.2 nM. Data shown were calculated from duplicate dete~inations of enzyme activity at 8 different concentrations of A. Correlation coefficients for linear regression analysis of Eadie-Hofstee transformed data were 0.90 or better. Values for Y,, and K,,, in six strains of genital skin fibroblasts as a function of NADPH concentration are shown in Table 3. The mean I’,,,., + SEM was 0.39 + 0.14 pmol/mg protein/h and the mean K,,,& SEM was 180 & 38 PM. Data shown were calculated from duplicate determinations of enzyme activity using 8 different concentrations of NADPH. Correlation coefficients for linear regression analysis of Eadie-Hofstee transformed data were 0.73 or better (mean = 0.90). The influence of glucocorticoids matase aciivitj
on microsomal
aro-
Initial experiments were performed to compare the influence of DEX on the aromatase activity in whole cells and in microsomal preparations. Confluent monolayers of genital skin fibroblasts (NFS-8) were incubated for 14 h at 37°C in medium with or without 250 nM DEX. Aromatase activity was determined in triplicate samples of whole cells and of microsomalenriched fractions. Whole cell aromatase activity Table 3. Aromatase kinetics in microsomal-enriched
prepnrations as
(mean f SEM) was 2.6 + 0.09 pmohmg protein/h in untreated cells and 10.4 + 0.28 pmol/mg protein/h in cells incubated with DEX. Aromatase activity in microsomal preparation was also stimulated by incubation with DEX, levels (mean f SEM) being 0.82 k 0.03 pmol/mg protein/h and 4.20 k 0.03 pmol/mg protein/h in microsomal preparations from untreated and treated cells, respectively. These experiments were carried out using two other strains with similar results, the increase in aromatase activity in whole cells and the microsomal-enriched fractions being 18- and 21-fold, respectively in NFS-1, and 2 and 3 fold, respectively in NFS-2. In time course experiments, aromatase activity was determined in triplicate samples of whole cells and microsomal-enriched fractions after incubation of genital skin fibroblasts (NFS-1) with 250nM DEX for 0, 14 or 48 h at 37°C. In whole cells, aromatase activity (mean k SEM) increased from a basal level of 0.29 t 0.05 pmol/mg protein/h to 3.25 + 0.39 pmol/mg protein/h at 14 h and then declined to 1.21 10.35 pmol/mg protein/h at 48 h. Microsomal aromatase activity followed a similar time course, levels being 0.03, 0.34 and 0.19 pmol/mg protein/h at 0, 14 and 48 h. respectively. Whole cell and microsomal-enriched aromatase activity were compared after “superinduction” in the following experiment. Confluent monolayers of genital skin fibroblasts were incubated in medium alone or in medium containing 250nM DEX or 250 nM DEX plus 50 fig/ml CHX for 14 h at 37°C and were then washed three times in Hanks Balanced Salt Sofution (HBSS). For whole cell assays, triplicate plates were incubated with 150 nM rH]A plus 5 FM P for 4 h and aromatase activity was determined. For microsomal-enriched assays, monolayers were incubated with MEM alone for 3 h at 37C, prior to preparation and assay of triplicate microsomal-enriched samples. Table 4 shows the results of such an experiment using NFS-8. Aromatase activity in whole cells and microsomal-enriched preparations was induced by DEX and “superinduced” by DEX plus CHX. Subsequent studies were performed to determine whether cytosolic factors are involved in the stimulation of aromatase activity by DEX. Confluent monolayers of genital skin fibroblasts were incubated for 14 h at 37°C in the presence or absence of 250 nM DEX, and microsomai-enriched fractions and cytosol were prepared. An aliquot of the cytosolic fractions was removed for protein assay and microsomes were
a function of NADPH concentration Vnl””
Strain NFS-I NFS-3 NFS-7 NFS-8 NFS-9 NFS-10 Mean f SEM
Table 4. Influence of DEX and CHX on whole cell and microsomal
(pmol/mg protein~h) 0.1 I 0.12 1.15 0.8 I 0.32 0.82 0.39 + 0.14
aromatase activity 89 150 300 280 200 180 + 38
Sample
Aromatase activity (pmoljmg protein/h) Baseline DEX
DEX plus CHX
Whole cells
1.31 fO.14
4.74 * 0.09
6.44 + 0.41
Microsomes
0.33 + 0.003
I .79 F 0.09
3.49 1_0.09
Data
shown represent the mean of triplicate determinations
*SEM.
Skin fibroblast microsomal aromatase
345
Table 5. Influence of cytosolic factors on DEX-stimulated aromatase activity in microsomal-enriched tions Source of microsomes: Source of cytosol:
Strain
Baseline Baseline 0.12 (0.11-0.13) 0.02 (0.014-0.026)
NFS-I NFS-5
prepara-
Aromatase activity (pmol/mg protein/h) Baseline DEX DEX DEX Baseline DEX 0.16 (0.160.16) 0.03 (0.034.03)
1.40 (1.4-1.4) 0.18 (0.17xl. 19)
I .70 (1.61.8) 0.18 (O.l&o.l8)
Data shown represent the mean and range of duplicate determinations.
in cytosol as shown in Table 5. Aromatase activity was determined in duplicate samples and was normalized for the microsomal-enriched fraction protein content of the sample by subtracting the protein contribution of the cytosol from the total protein content in the sample. In this and all other experiments using cytosol, the microsomes and cytosol in each sample were derived from the same number of cells. Aromatase activity in microsomes from DEX-treated cells was higher than that of untreated cells, and cytosol from DEX treated cells did not enhance the aromatase activity of microsomes. In time course experiments, genital skin fibroblasts (NFS-7) were incubated with medium containing 250 nM DEX for 12 or 60 h at 37°C and microsomalenriched fractions and cytosol were prepared. Aromatase activity was determined in microsomal preparations resuspended in cytosol as shown in Table 6. Activity was determined in triplicate samples and normalized for the protein content of microsomal-enriched preparations. Cytosol from cells incubated for 60 h did not diminish the activity of the microsomal-enriched fraction incubated for only 12 h. The influence of cytosolic factors on the superinduction of aromatase activity by DEX and CHX was examined. Confluent monolayers of genital skin fibroblasts (NFS-3) were incubated with 250nM DEX in the presence or absence of 50 pg/ml CHX for 14 h at 37°C. Cells were washed three times in HBSS and incubated in MEM for 3 h at 37°C and microsomal-enriched fractions and cytosol were prepared. Aromatase activity was determined in microsomal samples resuspended in cytosol as shown in Table 7. resuspended
Activity was determined in triplicate samples and normalized for the content of microsomal-enriched fraction protein. Aromatase activity in microsomal preparations derived from cells incubated with DEX plus CHX was three fold higher than in the microsomal-enriched preparations derived from cells incubated with DEX alone. Furthermore, cytosol from cells incubated with DEX plus CHX did not enhance the aromatase activity in microsomal preparations from cells incubated with DEX alone. DISCUSSION
Human skin fibroblasts grown in cell culture provide a valuable model for studying the factors that regulate aromatase activity in extraglandular tissue [7,8, 131. A variety of similarities exist between the aromatase complex in skin and that of other cultured human cells. In particular, the K, values of the aromatase in human skin fibroblasts [7] are similar to those in intact adipose tissue stromal-vascular cells [4], human placental cells [14], human choriocarcinema trophoblast cells [15], and human fetal hepatocytes [ 161,mean values ranging between 14 and 50nM. Our current study using microsomal preparations permitted us to characterize further the aromatase in human skin fibroblasts and to explore the unique aspects of its regulation. We studied aromatase activity in subcellular preparations containing both microsomes and mitochondria as enzyme activity was similar in subcellular particles contained in the 10,OOOg and 100,OOOg pellets. We validated our assay by demonstrating that aromatase activity increased in a linear fashion with time and with increasing protein concentration. The
Table 6. Influence of cytosolic factors on time course of DEX-stimulated aromatase activity in microsomal-enriched preparations Source of Source of Aromatase (omolimn
microsomes cytosol activity oroteinihl
l2h 12h 0.47 * 0.01
l2h 60 h 0.45 f 0.003
Data shown represent the mean of triplicate determinations
60 h 12h 0.19 * 0.003
60 h 60h 0.22 f 0.006
+ SEM.
Table 7. Influence of cytosolic factors on the superinduction of aromatase activity in microsomalenriched preparations Source of Source of Aromatase (pmol/mg
microsomes cytosol activity protein/h)
DEX DEX I.0 f 0.01
DEX DEX + CHX I .o f 0.02
Data shown represent the mean of triplicate determinations
DEX + CHX DEX 2.7 i 0.02 + SEM.
DEX + CHX DEX + CHX 3.5 f 0.14
346
GARY D. BERKOVITZ et al.
optimum pH for aromatase activity of microsomes from human skin fibroblasts was between pH 7.0 and 7.5. This is in the same range as the optimum pH (7.0) determined for the aromatase of human placental microsomes [17]. Leshin et al., studied estrogen production in cultured skin fibroblasts from the Sebright Bantam chicken as these ceils have extremely high levels of aromatase activity. The optimum pH of microsomal aromatase in these cells ranged from 6.5 to 7.0 when testosterone was used as substrate [lo]. The aromatase complex consists of at least two major microsomal components: an NADPH cytochrome P-450 reductase and a cytochrome P-450 mono-oxygenase [12]. The mean K, of the aromatase in microsomal-enriched fractions derived from cultured human skin fibroblasts as 27 nM when androstenedione was substrate. This value is similar to that obtained by others [l&20] for human placental microsomes (range 2345 nM). By contrast, enzyme kinetic studies of microsomal aromatase in cultured adipose tissue stromal vascular cells demonstrates a low affinity (K,,, = 1030 nM) and a high affinity (K,, = 100 nM) component [21]. This complex pattern was attributed to the presence of lipid in the microsomal membranes. None the less, the Km of the high affinity site is in the same range as that described by us for microsomal preparations of cultured skin fibroblasts. The K,,, values in human skin fibroblast microsomal-enriched preparations are also similar to the Km (2 nM) determined for aromatase in microsomes of Sebright Bantam chicken skin fibroblasts
DOI. The V,,, of the microsomal aromatase in human skin fibroblasts ranged from 0.09 to 1.26 pmol/mg protein/h (mean = 0.58 pmol/mg protein/h). The wide range of values most likely reflects differences in the number of enzyme complexes suggesting that the variation in intact fibroblasts [7] is also due to differences in enzyme content rather than differences in intracellular regulators, cofactor pools, or substrate availability. The mean K,,, for NADPH in human skin fibroblasts was 180 PM. Routine assays contained 5 mM NADPH, insuring saturation of NADPH binding sites. It is of interest that the mean K,,, for NADPH in Sebright Bantam skin fibroblasts was reported to be 0.13 PM, the value being approximately one thousand times lower than the K, in human skin fibroblasts [lo]. Investigation of microsomal aromatase activity also permitted a better understanding of the mechanisms involved in the glucocorticoid stimulation of aromatase activity. We determined aromatase activity in whole cells and in microsomal fractions at a substrate concentration (150 nM A) which is approximately six times the K,. Hence, aromatase activity approximated V,,,,,. Experiments using cytosol from DEX-stimulated and unstimulated cells excluded the involvement of a cytosolic factor in the DEX stimulation Aromatase activity increased in parallel in
whole cells and in microsomal-enriched preparations following DEX stimulation, suggesting that the effect of DEX results from an increase in the number of enzyme complexes. As mentioned earlier, the DEX-stimulation of aromatase activity in skin fibroblasts is unique in several aspects: namely, in the decline in aromatase activity from peak levels after prolonged exposure to DEX, and in the superinduction of aromatase activity by DEX and CHX. In time course experiments, microsoma1 aromatase activity mirrored that of whole cells, demonstrating a peak activity at 14 h and a decline by 48 h. We previously proposed that the decline in aromatase activity with prolonged exposure to DEX might be due to a loss of DEX receptor function, the induction of an inhibitory protein which interferred with aromatase activity or the induction of a repressor protein which interferred with transcription of the aromatase gene [8]. We subsequently demonstrated that DEX did indeed down-regulate the levels of glucocorticoid receptors in genital skin fibroblasts. However, experiments in which we demonstrated a dissociation between changes in DEX receptor levels and changes in aromatase activity suggested that the down-regulation of DEX receptor binding did not account entirely for the decline in aromatase activity [ 131.Our current studies in which microsomal preparations from cells incubated with DEX for either 14 h (short-term) or 60 h (long-term) were mixed with cytosol from these cells demonstrated that the decline in aromatase activity after 14 h was not due to the generation of a soluble inhibitory cytosolic factor. Coincubation of cells with DEX and CHX led to a superinduction of aromatase in whole cells when enzyme activity was assayed in the absence of CHX. In microsomal assays, we incubated cells in medium alone for 3 h prior to aromatase assay. Under these conditions, the aromatase activity of microsomal preparations was also superinduced. Experiments in which we mixed microsomal-enriched preparations and cytosol from DEX-induced and from superinduced cells demonstrated that the superinduction was not due to the presence of a soluble stimulatory cytosolic factor in the superinduced cells. We previously proposed that the aromatase in human skin fibroblasts might be under positive and negative control [8, 131.To explain superinduction we suggested that DEX might induce both aromatase mRNA and a putative aromatase repressor mRNA. However, the presence of CHX would prevent the repressor from being expressed and would permit a greater accumulation of aromatase message. In the absence of CHX, the aromatase mRNA would be translated leading to a greater than usual accumulation of aromatase. Alternatively, the amount of mRNA encoding aromatase in the presence of DEX and CHX might not be different from the amount transcribed in the presence of DEX alone. However, the effect produced by catabolism of the enzyme would be modified since the mRNA is transcribed
Skin fibroblast microsomal aromatase
over 34 h rather than 14 h. Our current work does not permit us to distinguish between these two possibilities. However, experiments using microsomal preparations lend credence to the concept that superinduction of aromatase activity in skin fibroblasts is due to a specific and dramatic increase in aromatase content. Acknowledgements-We thank MS Kathy Klein for her excellent secretarial assistance and Mrs Shirley Ho for her technical assistance. This work was supported in part by USPHS Research Grants ROI-DK-35339 and ROI-DK-00180.
347
Migeon C. J.: Time-dependent biphasic response of aromatase to dexamethasone in cultured human skin fibroblasts. J. c/in. Endocr. Metab. 63 (1986) 468474. 9. Brown T. R. and Migeon C. J.: Cultured human skin
fibroblasts: a model for the study of androgen action. Molec. Cell. Biochem. 36 (1981) 3-22. IO. Leshin M., Baron J.. George F. W. and Wilson J. D.:
Increased estrogen formation and aromatase activity in fibroblasts cultured from the skin of chickens with the henny feathering trait. J. biol. Chem. 256 (1981) 43414344.
Il. Lowry 0. H., Rosebrough N. J., Farr A. L. and Randall
Endocr. Metab. 49 (1979) 905-916. 3. Longcope C., Pratt J. H.,’ Schneider S. H. and Fineberg
R. J.: Protein measurement with the Folin phenol reagent. J. biol. Chem. 193 (1951) 265-275. 12. Thompson Jr. E. A. and Siierti P. K.: Utilization of oxygen and reduced dinucleotide phosphate by human placental microsomes during aromatization of androstenedione. J. biol. Chem. 249 (1974) 5367-5372. 11 Berkovitz, G. D., Carter K. M.,. Migeon C. J. and _-’ Brown T. R.: Down-regulation of the ghtcocorticoid receptor by dexamethasone in cultured human skin fibroblasts: implications for the regulation of aromatase activity. J. c/in. Endocr. Metab. 66 (1988) 1029-1036. 14. Lobo J. O., Bellino F. L. and Bankert L.: Estrogen synthetase activity in human term placental cells in monolayer culture. Endocrinology 116 (1985) 889-895. 15. Johnston J. O., Wright C. L. and Metcalf G. W.: Time-dependent inhibition of aromatase in trophoblastic tumor cells in tissue culture. J. steroid Biochem. 20
E. S.: Aromatization of androgens by muscle and adipose tissue in LVoo.J. clin. Endocr. Metab. 46 (1978)
(1984) 1221-1226. 16. Lanoux M. J., Cleland W. H., Mendelson C. R., Carr
REFERENCES 1. MacDonald P. C., Rombaut R. P. and Siiteri P. K.: Plasma precursors of estrogen. Extent of conversion of plasma androstenedione to estrone in normal males and non-pregnant normal, castrate and adrenalectomized females. J. clin. Endocr. Metab. 27 (1967) 1103-I 1Il. 2. MacDonald P. C.. Madden J. D., Brenner P. E., Wilson J. D. and Siiteri. P. K.: Origin of estrogen in normal men and in women with testicular feminization. J. clin.
146-152. 4. Ackerman
G. E.. Smith M. E., Mendelson C. R., MacDonald P. C. and Simpson E. R.: Aromatization of androstenedione by human adipose tissue stromal cells in monolayer culture. J. clin. Endocr. Metab. 53 (1981)
412417. 5. Simpson E. R.. Ackerman
G. E., Smith M. E. and Mendelson C. R.: Estrogen formation in stromal cells of adipose tissue of women: induction by glucocorticoids. Proc. natn. Acad. Sri., U.S.A. 78 (1981) 5690-5694.
6. Mendelson
C. R., Cleland W. H., Smith M. E. and Simpson E. R.: Regulation of aromatase activity of stromal cells derived from human adipose tissue. Endocrinology 11 (1982) 1077-1085.
7. Berkovitz G. D., Fujimoto M., Brown T. R., Brodie A.
M. and Migeon C.-J.: Aromatase activity in cultured human aenital skin tibroblasts. J. clin. Endocr. Mefab. 59 (198+ 665-67 1. 8. Fujimoto M., Berkovitz
G. D., Brown T. R. and
B. R. and Simpson E. R.: Factors affecting the conversion of androstenedione to estrogens by human fetal hepatocytes in monolayer culture. Endocrinology 117 (1985) 361-368. 17. Ryan K. J.: Biological aromatization of steroids, J. biol. Chem. 234 (1959) 268-272. 18. Covey D. F., Hood W. F. and Parikh V. D.: IO/JPropynyl-substituted steroids. J. biol. Chem. 256 (1981)
1096-1079. 19. Brueggemeier R. W., Snider C. E. and Kimball J. G.: A photoaffinity inhibitor of aromatase. Steroids 40 (1982) 679-689.
20. Burkhart J. P., Weintraub P. M., Wright C. L. and Johnston J. 0.: Novel silyated steroids-as aromatase inhibitors. Steroids 45 (1985) 357-374. 21. Cleland W. H., Mendelson C. R., Simpson E. R.: Aromatase activity of membrane fractions of human adipose tissue stromal cells and adipocytes. Endocrinology 113 (1983) 2155-2160.