J. Endocrinol.lnvest. 14.·31-35, 1991

Androgen receptors in normal and pathological thyroids M. Marugo*, G. Torre**, D. Bernasconi*, L. Fazzuoli*, S. Cassulo*, and G. Giordano* *Cattedra di Fisiopatologia Endocrina e Cattedra di Endocrinologia e Patologia Costituzionale, and **Cattedra di Fisiopatologia Chrurgica, Universita di Genova, 16132 Genova, Italy ABSTRACT. The purpose of this study was to determine the presence of cytosolic and nuclear androgen receptors (AR) in both normal and nodular thyroid tissues from twelve women and six men. Samples of benign thyroid nodules and corresponding surrounding normal tissue were processed by the single saturation point assay, using [3H] R1881 ([3H] Methyltrienolone) at final concentration of 5 nM. The results show the presence of AR (cytosolic and/or nuclear) in all examined tissue samples. The nuclear AR content was higher (p < 0.01) in normal rather than in nodular thyroid tissues. The same pattern was observed when nuclear AR were

analyzed according to the sex. In addition, nuclear AR content was significantly (p < 0.05) higher in normal thyroid tissue from men than from women. Our data suggest an androgen influence on thyroid tissue. If androgens are supposed to exert an antagonist role on estrogen actions also in thyroid tissue, the presence of higher nuclear AR concentration in the male rather than in the female normal thyroid may justify the lower incidence of thyroid diseases in men. Moreover, the lower AR levels found in male as well as in female nodular and goitrous tissues support the hypothesis that androgens may act with an antagonist mechanism on thyroid growth.

INTRODUCTION

sexes (6, 7) supports the hypothesis that androgens may have a physiological role on the thyroid. Basic prerequisite for the androgen action on thyroid was the presence of androgen receptors (AR) in thyroid tissue . Several authors have demonstrated (8-10) the presence of androgen cytosolic receptors in both normal and pathological human thyroids (i .e. papillary carcinomas, follicular adenomas and colloid goiters). In this study, we assayed cytosolic (ARc) and nuclear (ARn) androgen receptors in normal thyroid and benign nodules (such as adenomas and nontoxic goiters). The purpose was to determine whether the nodules and the adjacent histologically normal tissues have AR and whether there is any difference on the receptor conteht between female and male thyroids.

The relationships between sexual steroids and the thyroid function are well known. In women rather than in men, the incidence of thyroid disorders is greater and also the occurrence of goiters during puberty, pregnancy, and menopause give proof of growth modulation exerted by gonadal steroids. Estrogens and androgens can indirectly act through hypothalamic-pituitary-thyroid axis (1, 2) and on plasmatic concentrations of thyroid hormones by alterating hormone-binding proteins (3). Moreover, estrogens may have a role in thyroid growth modulation also through the action of local growth-factors. Preliminary experiments show that the IGF-I production by primary cultured human thyrocytes was stimulated by incubation with estradiol (unpublished data). The presence of estrogen and progesterone receptors in human normal and pathological thyroid tissues has been demonstrated in previous studies (4-7). The finding of similar estradiol receptor quantities in nodular and normal thyroids in both

MATERIALS AND METHODS Thyroid tissues was obtained from patients (12 women and 6 men, age range 30 - 65 yr) who underwent partial thyroidectomy for benign thyroid nodules . No patient was treated with exogenous hormones and all subjects showed normal thyroid function (T3 1.02 ± 0.07 ng/ml, T4 89.3 ± 5.1 ng/ml, TSH 1.79 ± 0.52IlU/ml, M±SE). Pathological and adjacent, histologically normal, thyroid tissues were immediately frozen after

Key-words: Androgen receptors, normal and nodular thyroid. Correspondence: Prof. Mario Marugo. Catted ra di Fis iopato log ia Endocrina, ISMI (Istituto Scientifico di Medicina Interna). Universita di Genova. Viale Benedetto XV. 6.16132 Genova. Italy.

Received May 10, 1990; accepted October 8, 1990.

31

M Marugo, G. Torre, D. Bernasconi, et ai,

surgery, and stored in liquid nitrogen until processing, The histological examination of the nodular tissue showed an heterogeneous structure formed by follicles of different dimensions bordered by hyperplastic, flat or normal epithelium, The assay of androgen receptors, both cytosolic and nuclear, was carried out by an exchange assay previously described (11), All processing was done at 4 C, if not indicated, The specimens were pulverized with a tissue pulverizer (Grindomat MM/Retsch, KG, Haan, Germany) at -70 C, resuspended in 10 vol of Buffer A (10 mM Tris pH 7.4 at 20 C, 1.5 mM EDTA, 10% glycerol v/v) and pelletted by centrifugation at 800xg for 10 min, The supernatant was adjusted to contain 10 mM sodium molybdate, and then centrifuged at 1OO,OOOxg for 60 min to yield the soluble cytosol fraction, The nuclear pellet was treated with pancreatic DNase 1 (20 flg/ml final concentration, Sigma Chemical Company, St. Louis, Mo) for 30 min and then centrifuged at 800xg for 15 min, Samples for DNA assay (200 fll) were drawn before adding DNase. The pellet was incubated in KCI Buffer (10 mM Tris pH 7.4 at 20 C, 1.5 mM EDTA, 10 mM thioglycerol, 0.6 M KCI, and 10% v/v glycerol), and centrifuged at 1OO,OOOxg for 30 min, The supernatant fluid (KCI-extract) was diluted with Buffer A to decrease the ionic strength, The residual pellets, after washing, were resuspended in Buffer A at the starting volume, This

fraction contained KCI-nonextractable nuclear receptors. In order to decrease the nonspecific binding and to remove endogenous steroids, the cytosolic and KCI-extractable fractions were mixed for 1 hour with buffer containing dextran-coated charcoal (DCC) 0,1% and then pelleted, Samples for the protein assays (Bradford's method using bovine serum albumin as standard) (12) were drawn from the supernatant after ultracentrifugation. The DNA contents were determined according to the method of Burton (using calf thymus DNA as standard) (13), before adding the DNase, The recovery of nuclei from the homogenate, based on DNA analysis, averaged about 75%. Incubation of cytosol fraction, KCI extractable fraction and KCI nonextractable fraction was carried out for 1 h at 20 C in quadruplicate, using [3H] R1881 (17-Methly [3H] methyltrienolone, 89 Ci/mmol, New England Nuclear, Boston, Ma) at final concentration of 5 nM and 2OO-fold molar excess triamcinolone acetonide (added to prevent binding of R1881 to the progesterone receptor), The nonsaturable binding, measured by parallel incubation in the presence of 200fold molar excess corresponding unlabelled R1881 (New England Nuclear, Boston, Ma), was subtracted from the total binding, The unbound hormone was removed by 0.25% DCC incubation for 30 min and centrifuged for 15 min. Aliquots of 400 fll of supernatant were placed in vials containing 8

Table 1 - Androgen receptors concentration (cytosolic ARc fmol/mg protein; nuclear ARn fmol/mg DNA) in normal and nodular thyroid, Case (no,)

Sex

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

F F

Mean

SE

M

F M

F F F F M

F F M M

F F F

M

Age (yr)

Normal Thyroid ARc

30 38 40 62 65 49 58 39 62 51 54 41 45 59 33 46 64 61

7,5 3,1 3,5 4,0 6,2 54 1,6 5,1 7.2 2,1 16,0 4,5 17,3 8,0 23,6 21,5 0 1,5

49,8 2,66

7,6 1,66

32

Nodular Thyroid ARn

ARc

ARn

26,5 30,8 350,0 53,3 373.2 23,7 257,6 265,8 85,7 1314 168,0 103,2 412,9 125,2 91,9 159,1 281,0 160,2

1,0 2,5 5,2 2,5 3,0 0 1,6 3,5 9,8 3,6 1,5 10,7 27,3 2,8 2,3 84 22,0 2,1

36,5 11,7 113,7 38,3 198,8 52,8 56,5 64,8 49.5 38.2 46,8 27,0 26,2 34,0 36,3 63,0 62,8 22,0

172.2 29,04

6,1

175

544 10,01

Thyroid and androgen receptors

ml of scintillation liquid (Atomlight NEN Chemicals) and then counted in a Beckman LS 7000 B-counter (Beckman Instruments Inc. Ca, 55% efficiency). By this method the binding sites, both occupied and unoccupied with the endogenous hormone, were measured. Results are expressed as fmol!mg protein for cytosolic receptors or fmol/mg DNA for nuclear reptors. The nuclear AR level is the sum of the level in KCI extractable and KCI non-extractable fraction. One pool of normal and one of pathological tissue specimens was also analyzed according to the Scatchard method (14), because in most cases a limited amount of tissue was available. Extraction and incubation techniques were similar but the assay was performed by different concentrations of [3H] R1881 ranging 0.625-10nM. Scatchard analysis of the binding data yielded a receptor concentration of 115 fmol/mg DNA for ARn KCI-extractable and 42 fmol/mg DNA for ARn KCI-nonextractable, in normal specimens. In nodular thyroids, the ARn concentration was 28 fmol/mg DNA in the former fraction and 15 fmol/mg DNA in the latter. The Kd did not differ in normal (0.62-0.69 nM) and in nodular (0.5-0.63 nM) thyroid tissue. The results obtained using the single saturation point assay correlated optimally with those using Scatchard's analysis; the concentration of ligand adopted (5nM) provided a reliable, although underestimated, measurement of AR content. The correlation of the sin-

gle saturation point assay with the Scatchard analysis was r = 0.92. Statistical analysis was performed by the nonparametric Mann Whytney test; p values