Molecular and Cellular Endocrinology, 8 (1977) 283-289 0 Elsevier/North-Holland Scientific Publishers, Ltd.

ANDROGEN METABOLISM IN RELATION TO GROWTH STIMULATION BY A

UTERINE CELL LINE *

Per STENSTAD, Kjetill @STGAABD and Kristen B. EIK-NES

Institute of Biophysics, N. T.H., University of Trondheim, N 7034 Trondheim-NTH, Norway

Received 28 March 1977; accepted 17 June 1977

The metabolism of radioactive testosterone, 5cr-dihydrotestosterone, 4-androstene-3@,17fl- diol or 4-androstene-3cu,lIlp-diol by the human cell line NHIK 3025, derived from a carcinoma of the uterine cervix, was studied. The cells were grown in Eagle’s minimal essential medium (MEM) with a steroid concentration of 10e7 M for 4 days. Androgen metabolism by this cell line is essentially the same as for other androgen-responsive cells. The most interesting testoster- one metabolite in this system is 4-androstene-3&17pdiol, and the separation of this compound from 4-androstene-3o,l7pdiol and the two corresponding So-reduced diols is described.

Since 4-androstene-3p,l7pdiol is a more potent growth factor for these cells than testoster- one, the small conversion of testosterone to 4-androstene-3p,l7p-diol observed could be respon- sible for the growth stimulation by testosterone.

Keywords: androgen metabolism; androgen responsiveness.

The work of Wibe et al. (1976) shows that the human cell line NHIK 3025, stemming from a carcinoma of the uterine cervix, is no longer responsive to estra-

diol-17& while T ** stimulates growth of these cells. It was also found that under the conditions of that study, DHT had no measurable effect on cell proliferation. This T-DHT discrepancy has also been observed in growth studies of primary myo-

blasts (Powers and Florini, 1975). However, 38,170 stimulated the NHIK 3025 cells significantly better than T. Androgen metabolism in these cells has therefore been

investigated in order to clarify the observed growth stimulation effects. It was of special interest to see whether any metabolic interconversion took

place between the two active compounds in this system, T and 3/3,17/I. Very little is known about metabolism leading to production of 3&17& but such metabolism has previously been noted in cerebral cortex, hypophysis, hypothalamus, prostate

* This work was in part supported by a research grant from the Norwegian Council for Science and the Humanities and by research contract No. l-ND-42812 from the National Institutes of Health, Bethesda, Md., U.S.A. ** The following abbreviations are used: T = 4androstene-3-one,l7p-ol; DHT = So-androstane- 3-one,17p-ol; A4 = 4-androstene-3,17-dione; k-o1,17-one = 5o-androstane-3a-o1,17-one; 3a,17p = 4-androstene-3a,l7@diol; 3p,17p = 4-androstene-3&17Pdiol; 3oadiol = 5o-androstane- 3o,17pdiol; 3&adiol= Sa-androstane-3@,17@diol.

283

284 P. Stenstad et al.

(Genot et al., 1975) and in rabbit skeletal muscle in vitro (Thomas and Dorfman, 1964).

In the current investigation we have studied the metabolism of radioactive T, DHT, 3a,17fl or 3@,17/3 by cells from the cell line NHIK 3025 while growing in a culture medium for 4 days.

MATERIAL AND METHODS

The solvents used for extraction and chromatography were purchased from Merck and Co.

Nonradioactive steroids were supplied by Steraloids Inc., Pawling, N.Y., except for 4-androstene-3o,l7&diol which was a gift from Merck and Co.

The following radioactive steroids were employed: [l ,2-3H] testosterone (43.7 Ci/mmol) and [ 1 ,2-3H]4-androstene-3,1 7-dione (46.2 Ci/mmol) from New England Nuclear; ]I ,2-3H] So-dihydrotestosterone (48 Ci/mmol), [ 1 ,2-3H]5a-androstane- 3&17&diol (41 Cijmmol), [ 1 ,2-3H] So-androstane-3&17&diol (42 Ci/mmol) and [l ,2-3~]4-androstene-3~,17~-diol (39.7 Ci/mmol) from Amersham. Radioactive 4-androstene-3~,17~-diol was obtained by incubating ~l,2-3H]testosterone (43.7 Cifmmol) with cells from the cell line NHIK 3025 under the conditions stated (see below). Radioactive 4-androstene-3Lu,l ‘I@-diol was isolated from these incubations as discussed in the Methods section. All radioactive steroids were purified by thin- layer chromatography before being used as substrates in incubation studies.

The cell culture The human cell line employed - NHIK 3025 (Nordbye and Oftebro, 1969;

Oftebro and Nordbye, 1969) - is derived from an early stage of carcinoma of the cervix uteri. These cells were incubated in a culture medium containing the follow- ing solutions (obtained from Gibco/Biocult): 86% (v~v) MEM (Eagle) with Earle’s salt and without L-~utamine, 1% (v/v) MEM nonessenti~ amino acid solution, 1% (v/v) L-gl~itamine solution and 2% (v/v) penicillin-streptomycin solution, To this mixture was added 10% (v/v) fetal calf serum also purchased from Gibco/BiocuIt.

The stock culture of cells was handled as published (Pettersen et al., I973), and the incubation experiments were conducted with cells from stock cultures in the log phase, 2-4 days after replating (Wibe et al., 1976). The cells were plated on a plas- tic Petri dish (Nunc) containing 5 ml culture medium. For studies on metabolism of radioactive steroids, 5 dishes, each containing about 50,000 cells, were used. Petri dishes containing only medium served as control. The dishes were incubated at 37°C in an atmosphere of 5% COa using a COa incubator (National Appliance CO.). After 4 h incubation the medium was removed and replaced with 5 ml fresh medi- um containing 10m7 M radioactive steroid. The amount of radioactivity (3H) was 400,000 cpm/ml culture medium. All samples were then incubated for 92 h.

Androgen metabolism by a uterine cell line 285

In order to measure celI growth in these experiments, 3 dishes containing 250 cells/dish were plated at the same time, and treated identically. Dishes containing cells and culture medium alone served as control. Cell growth in these dishes was determined as published by Wibe et al. (1976) and was in all cases in accordance with the steroid response found in previous experiments (Wibe et al., 1976). Finally, additional cell growth experiments were conducted as stated above with 3@,17& T or 3cu,I7@ added to the culture medium.

Separation of the steroids Upon termination of incubation the medium (5 ml) was transferred to centrifuge

tubes, the cells scraped off the Petri dish and washed out of this container with 0.9% sodium chloride solution (2 X 2 ml) which was added to the incubation medi- um. This mixture was extracted twice with 10 ml ethylacetate each time. In order to achieve maximal recovery of radioactive steroids, the extracting solvent con- tained the following nonradioactive steroids (1 pg/ml): T, A4, DHT, 3o-o1,17-one, 3a,l7@, 3/3,17& 3a-adiol and 3/3-adiol. The residue from extraction was exposed to thin-layer chromatography (TLC) on glass plates covered with silica gel (Merck silica gel 60 F-254, batch no. 5715) in the solvent system toluene : methanol (9 : 1). The plates were chromatographed twice. Radioactivity on the plate was located using a PW 4007 TLC scanner. TLC in this system gave the following frac- tions: fraction I containing polar and unknown radioactivity (Rr 0.07-o. 13); frac- tion II which could contain 3a,17@, 3j3,17& 3cr-adiol and 3&adiol (Rr 0.22); frac- tion III which could contain T (Rr 0.33); fraction IV which could contain DHT and 3a-01,17-one (Rr 0.46); fraction V, which could contain A4 (Rf 0.67).

Fractions I, II and IV were scraped off, the silica gel was extracted with diethyl- ether and the extracts were evaporated to dryness. The residue from fraction I was rechromatographed on silica gel (batch no. 5715) in the solvent system ethyl- acetate : n-hexane : acetic acid (7.5 : 20 : 5). Four distinct areas of radioactivity with Rr’s of 0.23,0.34,0.43 and 0.51 could be located. The nature of radioactivity in these areas still remains to be identified.

The residue from fraction II was rechromatographed on Schleicher and Schiill F-1500 silica gel plates (batch no. 1979 or 1981) in the solvent system dichloro- methane : ethylacetate (9 : 1). The plates were chromatographed 5 times and TLC was conducted at 4°C. Authentic 3cu,l7& 3cr-adiol, 3Sadiol and 3&17fi exhibited R, values of 0.24,0.38,0.49 and 0.55 respectively in this chromatography system.

The residue from fraction IV was also rechromatographed on Schleicher and Schtill silica gel plates in the solvent system dichloromethane : ethylacetate (9 : l), but the plates were chromatographed twice and TLC was conducted at 18-20°C. Authentic 3cu,17-one showed Rf of 0.48 and DHT had Rf of 0.61 in this TLC sys- tem.

Sample radioactivity corresponding to these Rr values was scraped off and counted in a liquid scint~ation counter (Isocap/300). A part of this radioacti~ty was mixed with authentic nonradioactive steroids (T, A4, DHT, 3cu-o&17-one,

286 P. Stenstad et al.

3a,17& 3&17fi, 3o-adiol or 3&adiol) and crystallized to constant specific activity (cpm/mg) from three different solvent combinations. The variation in specific

radioactivity during these crystallizations was within ?5% of average specific radio- activity (data not shown).

In all experiments known amounts of radioactive T, A4, 3a,l7& 3/3,170,3a-adiol and 3&diol were dissolved in 5 ml medium and processed like sample steroids. The

recovery of steroids from these blank samples was used for correcting losses during

extraction and chromatography of steroids in the incubation samples.

RESULTS AND DISCUSSION

The results from study of T metabolism by cells from the cell line NHIK 3025

are presented in table 1, the values for metabolites and unused substrate being given as percentages of the substrate added to the medium at the start of incubation. It is evident that these cells can form 3&l 7/I from T. Moreover, the isomer 3a,l7/3 is also produced in this system, and in addition T is converted to A4 and to the So-reduced compounds DHT, 3a,-adiol, 3/I-adiol and 3a-01,17-one. Finally, several polar T- metabolites are formed (table 2), but these conversion products have up to now not been identified. Attempts to isolate estradiol-170 from these incubations failed.

When growing these cells with radioactive T under the conditions given, we also looked at T-metabolites in the medium and the cell fraction separately. At the end of incubation the cell fraction was washed with 0.9% sodium chloride solution (3 X 2 ml) before being extracted for steroids. Table 2 gives the relative amounts of metabolites in the cells and in the incubation medium. However, no conspicuous

partition of steroids between these two fractions could be determined. Table 3 contains data on bioconversion of radioactive 3&17/I by these cells. The

Table 1 Metabolism of [3HJtestosterone by the cell line NHIK 3025. 50,000 cells were incubated in 5 ml medium containing lo-’ M testosterone for 92 h. The metabolites are given as percent- ages of the substrate added to the medium.

Metabolites

4-Androstene-3a!,l’Ip_diol 4-Androstene-3P,17p_diol Se-Androstane-3or,l7p-diol So-Androstane-3pJ7Pdiol 4-Androstene-3,17dione So-Dihydrotestosterone Sex-Androstane-3or-o1,17one

Mean t SD (n = 5)

5.90 ? 0.78 0.51 f 0.06

14.38 f 1.68 0.30 f 0.06 0.84 f 0.07 1.80 zt 0.12 0.27 f 0.02

Testosterone unmetabolized 63.26 * 3.65

Androgen metabolism by a uterine cell line 281

Table 2 Metabolism of [3H]testosterone by the cell line NHlK 3025. 50,000 cells were incubated in 5 ml medium containing lo-’ M testosterone for 92 h. The metabolites are presented as per- centages of the total amount of radioactive steroids isolated in the medium and in the cell frac- tion respectively.

Metabohtes

Unknown compounds 4-Androstene-3cx,l7p-diol 4-Androstene-3~,17~-diol 5-Androstane-3o,l7p-diol 5-Androstane-3p,l7p-diol 4-Androstene-3 ,17dione 5~Dihydrotestosterone Sa-Androstane-3a-ol,l7-one Testosterone

Medium Cell fraction

3.0 4.7 6.7 4.6 0.6 1.0

16.0 10.2 0.3 1.0 0.9 5.2 2.0 5.7 0.3 0.2

70.2 67.6

metabolism of this substrate is reduced compared to that of T, but small amounts of T, 3a,l7@ and 3a-adiol were formed.

When the cells were grown with radioactive 3a,l7/3 metabolism to T was greater than when 3/3,17/l was the substrate (table 3). Thus, these cells contain A4-, 3o- and 3&hydroxysteroid dehydrogenases, and these enzymatic activities are reversible.

A direct reduction of 3a,17P or of 3@,17/3 to the correspond~g 3.adiol in our system is a most unlikely metabolic event. When the cells were incubated with 30,170 no 3/3-adiol could be determined (table 3) and with 3a,17P as substrate, 3~ adiol formation was low (table 3), though minute metabolism to DHT was detected

Table 3 Metabolism of [3H]4-androstene-30r,17@diol and of [3H]4-androstene-3P,17fldiol by the cell line NHIK 3025. 50,000 cells were incubated in 5 ml medium containing lo-’ M steroids for 92 h. The metabolites are given as percentages of the substrate added to the medium.

Metabolites Substrate used

30,170 (mean f SD, n = 5) 3p,17p (mean * SD, n = 5)

Testosterone 20.68 f 1.20 4.06 it 0.33 4-Androstene-3,17dione 0.16 t 0.08 a

4-Androstene-3a!,lIlpdiol a 1.24 b. 1.09 So-Androstane-3-oneJ7p-ol 0.59 + 0.05 a

5cu-Androstane-3cu,l7p_diol 0.75 t 0.12 1.79 f 0.16

Unmetabol~ed substrate 74.12 f 8.10 82.05 t 5.01

a Less than 500 cpm in tlnal sample.

288 P. Stenstad et al.

in the latter experiment. Production of 3ol-adiol or 3P-adiol from either 3a,17P or 3/3,17fl probably proceeds via T as intermediate in our incubations, as observed in other biosystems (Ungar et al., 1957; Gross0 and Ungar, 1964; Kundu et al., 1965).

Incubations of 50,000 cells in 5 ml medium with radioactive DHT resulted in almost complete metabolism of the substrate to 3cr-adiol (table 4). The activity of the 5&androstane3cr-hydroxysteroid dehydrogenase in the cells from line NHIK 3025 is much higher than that of the enzyme(s) converting T to 3rr,17/3 or to 3&17fl (tables 1 and 4). The rate-limiting step in metabolism of T to 3o-adiol by these cells is the 5~-reductase (tables 1 and 4).

En previous work (Wibe et al., 1976) we were unable to demonstrate significant increase in growth of these cells when incubated (250 cells plated/5 ml medium) with concentrations of DHT ranging from 10m9 to low6 M. The data of table 4 could, however, point to the possibility that lack of growth promotion with DHT is due to rapid conversion of DHT to 3a-adiol, a steroid which is inert as a growth inducer in this system in vitro (Wibe et al., 1976). When, however, 250 cells were grown in 5 ml medium with 10V7 M DHT, most of the substrate was left unmetab- olized at the end of 4 days of incubation (table 4). It could, however, be that the intracellular DHT concentration is too low in these growth experiments. Following separation of cells and medium after 4 days incubation with 10m7 M DHT under conditions similar to those of Wibe et al. (1976) (1000 cells plated /S ml medium), we found that the relative amounts of DHT and 3or-adiol were the same in the cells and in the medium (99% DHT and 0.8% 3ol-adiol). Thus, lack of DHT response in the cell growth experiment of Wibe et al. (1976) is probably not due to the absence of DHT as such in the system.

In incubations of cells from the cell line NHIK 3025 with 10V7 M 3/3,17/3, increased growth rate could be determined (Wibe et al., 1976). One may argue whether this effect is obtained in our system in vitro via conversion to T (table 3). However, as shown by Wibe et al. (1976), T gave no detectable growth stimulation of these cells at concentrations from 10m9 M or lower, while 30,17fi gives a signifi-

Table 4 Metabolism of [3HJ5~~~ydrotestosterone by the cell line NHIK 3025. Cefls were incubated in medium containing IO-’ M Sodihydrotestosterone for 92 h. The metabolites are given as per- centages of the substrate added to the medium. -

Metabolites 50,000 cells/5 ml medium 250 cells/S ml medium (mean i: SD, n = 5) (mean ?: SD, n = 3)

So-Androstane-3a,l7pdiol 89.48 f 6.52 0.77 + 0.03 So-Androstane-3pJ7pdiol 3.28 t 0.33 5a-Androstane-3a-olJ7-one 0.40 t 0.15

So-Dihydrotestosterone 1.79 f 0.50 98.83 ?: 0.81 unmetabo~zed

Androgen metabolism by a uterine cell line 289

Table 5 Growth data relative to control values for cells from the cell line NHIK 3025 after incubation for 4 days in the presence of steroids at concentrations given. Deviations recorded are based on standard error of the mean of steroid treated and control cells (n = 5). The system used has been published by Wibe et al. (1976).

Steroid added to medium Colony size Plating efficiency a

lo-r0 M Testosterone 1.01 * 0.04 1.00 f 0.11

10-l* M 4-Androstene-3~,17~diol 1.10 f 0.03 b 1.01 + 0.07 1.13 f 0.06 b 0.98 f 0.05

lo-’ M 4-Androstene-3o,l7fl-diol 1.01 * 0.04 1.00 f 0.05 1.03 + 0.05 1.00 -t 0.11

a Ratios of mean number of colonies per dish with more than 4 cells. b Mean value of colony size higher than control, significant to P < 0.01 and P < 0.04 respec-

tively, according to Student’s t-test.

cant effect even at lo-” M (table 5). These results have also been verified using the Coulter Counter method for deterring cell growth described by Wibe et al. (1976). This steroid concentration is well below the level of endogenous testoster- one in the medium used, which was determined to be 0.7 X 10V9 M. Therefore, the effect of 3&17/3 on growth of these cells cannot simply be explained by bioconver- sion of added 3/$17@ to T. Furthermore, 3ar,l7/3 is a better substrate for this conver- sion than 3&17p (table 3) and 3o,17@ gives no significant growth stimulation of these cells (table 5) at a concentration when T and 3/3,17fl act as growth stimu- lators. The observed effect of 3/3,17/I appears thus to be dependent on the presence of unmetabolized 313,170.

Although the metabolism of T to 30,170 is low (table l), the growth-promoting effect of 3@,17p is found at concentrations so low (table 5) that this metabolism can possibly be responsible for the growth-stimulating effects of T (Wibe et al., 1976).

REFERENCES

Genot, A., Loras, B., Monbon, M. and Bertrand, J. (1975) J. Steroid Biochem. 6,1247-1251. Grosso, L.L. and Ungar, F. (1964) Steroids 3,67-75. Kundu, N., Sandberg, A.A. and Slaunwhite Jr., W.R. (1965) Steroids 6,543-551. Nordbye, K. and Oftebro, R. (1969) Exp. Cell Res. 58,458-459. Oftebro, R. and Nordbye, K. (1969) Exp. Cell Res. 58,459-460. Pettersen, E.O., Oftebro, R. and Brustad, T. (1973) Int. J. Radiat. Biol. 24, 285-296. Powers, M.L. and Florini, J.R. (1975) Endocrinology 97,1043-1047. Thomas, P.Z. and Dorfman, R.I. (1964) J. Biol. Chem. 239,766-772. Ungar, F., Gut, M. and Dorfman, R.I. (1957) J. Biol. Chem. 224,191-200. Wibe, E., Qstgaard, K. and Eik-Nes, K.B. (1976) Mol. Cell. Endocrinol. 5,359-364.