Molecular and Cellular Endocrinology, 86 (1992) 11l- 118 0 1992 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/92/$05.00

MOLCEL

111

02783

Androgens increase insulin receptor mRNA levels, insulin binding, and insulin responsiveness in HEp-2 larynx carcinoma cells Giorgio Sesti a, Maria Adelaide Marini a, Paola Briata b, Antonella Nadia Tullio ‘, Antonio Montemurro a, Patrizia Borboni ‘, Roberto De Pirro d, Roberto Gherzi ’ and Renato Lauro a ” Dipartimento di Medicina Interna, Cattedra di Endocrinologia, II Unic,ersitridi Roma, Rome, Italy, h Laboratorio di Immunobiologia, Istituto Nazionale per la Ricerca sul Cancro (IST), Genol,a, Italy, ’ Laboratorio Cellife/Istituto Nazionale per la Ricerca sul Cancro (IST), Genoc,a, Italy, and ’ Cattedra di Endocrinologia, lhkersitci di Ancona, Ancona, Italy (Received

Key words: Insulin

action;

Insulin

receptor

5 February

gene; Androgen;

1992; accepted

HEp-2

27 March

1992)

larynx carcinoma

Summary Androgen receptors have been found in human larynx and androgens have been supposed to play an important role in promoting the growth of laryngeal carcinomas. The molecular mechanism underlaying this phenomenon is not at all understood. Aim of this work was to investigate the effects of two androgens (testosterone and dihydrotestosterone) on insulin receptor mRNA levels and insulin binding activity as well as on either metabolic or growth-promoting actions of insulin in a human larynx carcinoma cell line (HEp-2). We found that HEp-2 cells express a high affinity insulin receptor. Both androgens significantly increase insulin receptor mRNA levels and insulin receptor number in HEp-2 cells. Insulin action, evaluated either as total glucose utilization or as [“Hlthymidine incorporation into DNA, significantly increased in HEp-2 treated with androgens in comparison to control cultures. Altogether, our data allow us to speculate that the increased insulin effectiveness we observed in the larynx carcinoma cell line HEp-2 after androgen treatment might be involved in the regulation of larynx cancer cells growth.

Introduction The process of cell growth and differentiation depends on the intricate interplay of many effec-

Correspondence to: Dr. Roberto Gherzi, Cellife Laboratory, National Cancer Institute (IST), Viale Benedetto XV, 10, 16132 Genova, Italy. Tel. 10-355504; Fax 10-352999. Supported, in part, by grants from Consiglio Nazionale delle Ricerche and Minister0 Universita e Ricerca Scientifica e Tecnologica 40% and 60% (to R.L.) and from Minister0 Universita e Ricerca Scientifica e Tecnologica 60% (to R.G.).

tor molecules involved in the regulation of gene transcription. Some of these cellular effecters, like the insulin receptor (IR), belong to the family of protein tyrosine kinases which affect gene expression through a not yet understood cascade of protein phosphorylation and/or de-phosphorylation which, ultimately, leads to the activation of some transcription factor (Morley and Traugh, 1990; Manzella et al., 1991). Another class of effecters is constituted by the members of the recently identified steroid-/ thyroid hormone-/ retinoic acid-receptor superfamily, which function

112

in response to binding of the corresponding ligand as transcription factors of specific target genes (for review see: Evans, 1988; Beato, 1989). The human larynx is a target tissue for androgens (Saez and Sakai, 1976). Androgen receptors have been demonstrated in either the normal human larynx or in the laryngeal epithelial carcinomas (Tuohimaa et al., 1981; Beckford et al., 1985). Furthermore, it has been reported that androgens might play an important role in promoting the growth of laryngeal carcinomas (Mattox et al., 1984; Vecerina-Volic et al., 1987). Insulin, acting through its membrane receptor, regulates either cellular metabolism or cell growth and differentiation programs (for review see: Rosen, 1987; Kahn and White, 1988). After initial binding to its receptors, insulin exerts a variety of biologic effects: (i) the activation of transport systems (for hexoses, amino acids and ions); (ii) the stimulation of the activity of several enzymes; (iii) the activation of the DNA synthesis; and (iv) the modulation of the number of its own receptors (Rosen, 1987; Kahn and White, 1988). Insulin receptor gene expression regulation has not been yet completely clarified even if - 1800 base pairs of the IR gene promoter region have been cloned and sequenced (Araki et al., 1987; Seino et al., 1989; Tewari et al., 1989). Since androgen receptors (like glucocorticoid and progesterone receptors) are DNA binding proteins which, upon activation by the specific ligand, truns-activate the expression of target genes (Evans, 1988; Beato, 1989), we hypothesized: (i> that the insulin receptor gene expression is regulated by androgens as it is regulated by glucocorticoids (McDonald and Goldfine, 1988; Rouiller et al., 1988; Briata and Gherzi, 1990) and progestins (Papa et al., 1990); (ii> that androgen action leads to a modulation of insulin responsiveness in androgen-treated cells. In the present study we demonstrate, for the first time, that androgens increase insulin receptor mRNA levels and insulin binding activity in the larynx carcinoma HEp-2 cells. This, in turn, is associated with an increase of insulin responsiveness, in terms of either metabolic or growth-promoting actions, in this transformed human laryngeal cell line.

Materials

and methods

Reagents and plasmids D-[U-‘“ClGlucose (280 mCi/mmol), [methyl‘Hlthymidine (20 Ci/mmol), Multiprime DNA labelling system and Hybond-N nylon membranes were purchased from Amersham; [a- j2 P]dCTP (N 3000 Ci/mmol) was from New England Nuclear; DEAE 52 (Na45) membranes were from Schleicher and Schuell. [‘2’I]A14 monoiodo human insulin (300-350 mC/mM) was a gift of Prof. R. Navalesi (Pisa, Italy). Insulin-like growth factor I (IGF-I) was from Bachem. IGF-II was from Boehringer-Mannheim. Growth hormone (GH) was from Lilly Research Center. Porcine and human insulin were kindly provided by Novo Industries Anti-insulin receptor monoclonal antibodies MA-10 and MA-20, raised in mice against highly purified human placental receptor (Forsayeth et al., 1987), were purified to the homogeneity as previously described (Sesti et al., 1988). Cell culture media and fetal calf serum (FCS) were from ICN Laboratories. All materials for agarose gel electrophoresis were purchased from USB. Testosterone, dihydrotestosterone (DHT), and bovine insulin were from Sigma Chemical Co. All other reagents, of molecular biology grade, were as described in Gherzi et al. (1989) and Briata et al. (1990). A 4.1 kb EcoRI fragment of the IR cDNA (pHIR 12.1), obtained from the American Type Culture Collection, and a 1.2 kb PstI fragment of human p-actin cDNA (pBAC) were prepared and used as described (Briata et al., 1990); rat brain GLUT1 cDNA probe (pGTH-14) (Birnbaum et al., 1986) was the kind gift of Dr. Morris Birnbaum (Harvard University Medical School). Cells HEp-2 epidermoid carcinoma cells, derived from human larynx (Moore et al., 1955), were originally obtained from Flow/ICN. Cells were grown in monolayers in Eagle’s modified minimal essential medium (MEM), supplemented with 10% fetal calf serum. Confluent cells were subcultured in 6-well or 24-well plates and used for experiments as described in the Results section and in figure legends. The medium was changed

113

to HEp-2 ments.

monolayers

the day before

the experi-

Binding studies Insulin binding to confluent monolayers was performed in 100 mM Hepes buffer (pH 7.9) containing 120 mM NaCl, 1.2 mM MgSO,, 2.5 mM KCl, 15 mM sodium acetate, 10 mM glucose and 10 mg/ml bovine serum albumin (BSA) (IBB). Cells were incubated with [‘*“I]human insulin (50 pM) in the absence or in the presence of various amounts of native human insulin or other ligands and the specific binding was measured as previously described (Gherzi et al., 1989). To perform experiments aimed to investigate the effect of androgens on insulin receptor binding, cells were incubated in MEM medium containing 0.5% dialyzed BSA in the presence of either 0.1% (v/v> ethanol (the solvent of androgens: control cultures) or different amounts of testosterone or dihydrotestosterone for various periods of time at 37°C and, thereafter, [‘*‘I]insulin binding was assayed as described above. Cells were counted (two wells of each culture condition) and the binding was normalized to 10h cells. Insulin receptor internalization and down-regulation were evaluated as previously reported in detail (Gherzi et al., 1989; and references cited therein). RNA preparation and Northern blot analysis Total RNA was prepared from HEp-2 cells by the guanidine/cesium chloride method (Chirgwin et al., 1979). Polyadenylated RNA (poly(A)+ RNA) was prepared according to the procedure of Jakobsen et al. (1990) using Dynabeads oligo(dT),, (Dynal). pHIR 12.1, pGTH-14, and pBAC were labelled by random priming with hexanucleotide fragments using [ w3* P]dCTP (Feinberg and Vogelstein, 1984). Polyadenylated RNA (4 pg per lane) was denatured, electrophoresed and blotted to Hybond-N nylon membranes as reported (Briata et al., 1990). Filters were hybridized for 16 h at 65°C using radioactive probes, washed as previously described (Briata et al., 1990) and exposed to Hyperfilms at - 70°C. Autoradiograms were densitometrically scanned (using a Ultroscan LKB laser densitome-

ter) and the relative arbitrary units.

absorbances

expressed

in

Total glucose incorporation Confluent monolayers of HEp-2 cells were treated either with ethanol (O.l%, v/v, final concentration: control cells) or with androgens (dissolved in 0.1% ethanol). Then, monolayers were incubated for 30 min at 37°C in Krebs-RingerHepes buffer (pH 7.81, containing 5.5 mM glucose and 10 mg/ml BSA, in the absence (basal activity) or in the presence of various amounts of insulin. Then, D-[U-‘4C]glucose (0.5 PCi) was added for 45 min at 37°C. Assays were stopped by washing the cells 3 times with ice-cold phosphate-buffered saline (PBS). Cells were solubilized in 0.03% (w/v> sodium dodecyl sulfate (SDS) and the incorporated radioactivity was counted. An aliquot of the solubilized material was used to measure the protein content (see below). Nonspecific trapping of the labelled glucose was measured by stopping the reaction immediately after the addition of the D-[U-‘4C]glucase and it was subtracted from the counts of each experimental point. Thymidine incorporation into the DNA and biochemical assays HEp-2 monolayers were treated with androgens as described above. [“HlThymidine incorporation into DNA of HEp-2 cells was measured as previously described (Gherzi et al., 1989) with the slight modifications detailed in the legend to Fig. 6. The protein content of solubilized monolayers was assessed by the method of Bradford (1976). The results of insulin binding and total glucose incorporation assays were normalized taking into account the protein content of each culture. Results Insulin binding to HEp-2 cells Since, to our knowledge, HEp-2 cells have never been used to investigate insulin-receptor interactions, we performed preliminary experiments in order to characterize the IR in these cells and to determine optimal insulin binding conditions. As expected for an housekeeping gene

114

product, IR is expressed by HEp-2 cells. The HEp-2 insulin receptor specifically binds insulin and maximal insulin binding is obtained incubating cells with [“51]insulin at 4°C (data not shown). At this temperature, the steady state of insulin binding was reached in 5 h and maintained unchanged up to 20 h of incubation (data not shown). We also measured insulin binding at 24°C and we found that, at this temperature, the equilibrium binding (85% of the maximal binding at 4°C) was reached in 90 min and maintained up to 5 h (data not shown). However, since some experiments, aimed to investigate insulin biological action (total glucose incorporation assay), have to be performed at 37°C we measured insulin receptor binding also at this temperature. At 37°C binding equilibrium was achieved in 30 min and maintained up to 3 h (data not shown). Unless otherwise stated, we performed the majority of binding experiments at 4°C. Next, we investigated the potency of various ligands to compete with [““Ilhuman insulin for binding to IR. As extensively reported in other cell lines, our insulin binding studies in HEp-2 cells demonstrated that: (i) human, porcine and bovine insulin bind to IR with a similar potency (Fig. 1); (ii) Scatchard plot of insulin binding data is curvilinear (Fig. 2); (iii> the anti-insulin receptor monoclonal antibody MA-10 is more potent than MA-20 antibody in inhibiting insulin binding to HEp-2 cells (Fig. 1). It is noteworthy that the monoclonal antibody MA-IO, which recognizes the IR and not the IGF-I receptor (Forsayeth et al., 1987), completely inhibits insulin binding to HEp-2 cells (Fig. l), further supporting the evidence that, in these cells, insulin specifically binds its own receptor; (iv) lo-’ M IGF-I, 10Ph M IGF-II and 10Ph M GH do not affect insulin binding to HEp-2 cells (Fig. 1); (v) upon binding to its receptor, insulin is internalized in a time-, and energy-dependent manner temperature-, (data not shown); (vi) insulin (lOPh M) causes a down-regulation of its own receptors (data not shown). Androgens increase insulin binding to HEp-2 cells In order to investigate the effect of androgens on insulin receptor binding, HEp-2 cells were incubated for 24 h at 37°C in MEM (containing

Ligand Concentration (M) Fig. I. [““I]Insulin

binding

to HEp-2

cells is differently

placed by various ligands. Cells were incubated

man insulin (SO pM) in the presence of the indicated trations of these ligands: human

CO), bovine MA-20 GH

insulin

monoclonal

(01,

insulin

MA-10

antibody

( * ), porcine

monoclonal

(0). for YO min at 24°C.

and methods.

The

insulin bound

All the experiments are expressed

conccninsulin

antibody

(ml.

( + ), IGF-I (W). IGF-II (x ).and amount

insulin bound was then assayed as described cate. Results

dis-

with [“iI]hu-

were carried

as percentage

in the absence

of [“‘I]human under Materials out in tripli-

of the [“51]human

of any unlabelled

competitor

ligand.

0.5% dialyzed BSA) in the absence or in the presence of different concentrations (lo- “’ M to 5 X lo-” M) of either testosterone or DHT. Then, cells were washed twice with IBB and insulin binding was assayed at 4°C. Both testosterone and DHT significantly increase insulin binding in a concentration-dependent manner (Fig. 3). DHT in prois - 20% more potent than testosterone ducing this effect. This finding is in agreement with the relative biological potency of the two androgens on target tissues (Dorfman and Shipley, 1956). Time-course studies demonstrate that both androgens produce their maximal effect on insulin binding in 24 h. Androgens’ action on insulin binding is maintained for up to 48 h while the half-maximal effect occurs in 12 h (data not shown). Scatchard analyses of binding curves (Fig. 2) demonstrate that both androgens increase the number of insulin binding sites per cell without affecting insulin receptor affinity.

115

1.2 I

7

0

Control

n Testosterone t DHT

0

2

4

6

8

I0

12

Bound (fmol/106 cells) increase [“sI]insulin binding to HEp-2 Fig. 2. Androgens cells. Cells were incubated for 24 h at 37°C in the presence of 0.1% ethanol (the solvent of androgens; control cells: CO). 10~ ’ M testosterone (m) or lO_” M DHT) ( * ). Then, [“‘Ilhuman insulin binding was measured for 5 h at 4°C in the absence or in the presence of increasing amounts of unlabelled human insulin. Nonspecific binding was measured in the presence of lO_” M unlabelled human insulin and subtracted. Cells were counted (two wells of each cell culture condition) and the binding was normalized to 10” cells. Results of three independent experiments, performed in duplicate, have been averaged and are presented. The abscissa could also be expressed as O-100 pM. The ordinate is a unit-less ratio.

Fig. 3. The effect of androgens on insulin binding to HEp-2 cells is concentration dependent. Subconfluent cells were incubated for 24 h at 37°C in the presence of 0.1% ethanol (control), testosterone (dissolved in 0.1% ethanol) or dihydrotestosterone (dissolved in ().I%> ethanol; DHT) at the indicated concentrations. Then, insulin binding was performed for 5 h at 4°C as described under Materials and methods. Nonspecific binding (i.e. the [“‘l]insulin binding in the presence of 1 @M unlabelled insulin) was calculated and subtracted. Results are expressed as percent of insulin binding over control. Five sets of experiments (performed in triplicate) have been averaged and presented. The values+ standard error of the mean for [“‘I]insulin binding to control cultures (not treated with androgens) were 1012+29 cpm. Standard errors never exceeded 39~ of mean values.

TDC Arzdrogens increase insulin receptor mRNA lelsels in HEp-2 cells It has been recently demonstrated that glucocorticoids and progestins increase IR mRNA levels in different cell lines (McDonald and Goldfine, 1988; Rouiller et al., 1988; Briata and Gherzi, 1990; Papa et al., 1990). Taking into account that androgens significantly increase insulin-receptor binding in HEp-2 cells and that the cellular actions of these hormones are mediated by receptors which belong to the same superfamily as glucocorticoid and progesterone receptors, we investigated the effects of testosterone and DHT on IR mRNA levels. As shown in Fig. 4, both androgens increase IR mRNA levels by - 3-fold. DHT is slightly more active than testosterone in producing its effect on IR mRNA levels. As a control, we hybridized our Northern blots with pBAC. As shown in Fig. 4, the amount of p-actin mRNA is similar in control and in androgen-

TDC

-GLUT1

HIRL n”^.,

-PACT Fig. 4. Northern blot analysis of poly(A)+ RNA from control and androgen-treated HEp-2 cells. Subconfluent HEp-2 cells were treated with lo-’ M testosterone (T) or DHT (D) or with 0.1% ethanol (the solvent of androgens) (C) for I6 h at 37°C. Poly(A)+ RNA was prepared from the cells according to the procedures detailed under Materials and methods. 4 fig of poly(A)’ RNA were electrophoresed and transferred to Hybond-N membranes. Filters were hybridized with either insulin receptor (HIR), GLUTl, or p-actin radiolabelled cDNA probes (as indicated), washed and autoradiographed. Figure shows representative autoradiograms (exposed 3 days at -80°C for HIR and GLUTl, 16 h at -80°C for p-actin) of three performed. The two bands corresponding to the major transcripts of the IR gene have been densitometrically scanned. The relative absorbances (expressed in arbitrary units) were 975 (0, 3015 (D), 275.5 (TX

treated cells. To better investigate the effect of androgens on total glucose incorporation (see below), we hybridized Northern blots with a ‘rat brain/ erythroid/ HepG2’ glucose transporter gene (GLUT11 cDNA probe. As depicted in Fig. 4, neither testosterone nor DHT affects GLUT1 mRNA levels in HEp-2 cells.

140 _

Androgens enhance insulin responsil~enessin HEp-2 cells Taking into account the above reported results, we investigated whether androgen-treated HEp-2 cells exhibit enhanced responsiveness to insulin. One of the major biological actions of insulin is the stimulation of glucose metabolism in target cells. In HEp-2 cells, insulin rapidly stimulates total glucose incorporation in a dose-dependent manner, the half-maximal effect occurring at 5 X lo-” M (Fig. 5 and data not shown). Both the time course of glucose incorporation (data not shown) and the concentration dependency of insulin action (see above) are similar to those reported in other human cell types including fibroblasts and adipocytes (Ciaraldi et al., 1979; Howard et al., 1979). Both testosterone and DHT enhance maximal glucose incorporation into HEp-2 cells and reduce by approximately one order of magnitude the concentration of insulin producing the halfmaximal effect (Fig. 5). This finding agrees with previously reported observations that an increase in IR number per cell leads to an increase in cellular sensitivity to insulin, in terms of glucose transport (Ebina et al., 1985; Chou et al., 1987). The demonstration that androgens do not affect GLUT1 mRNA levels in HEp-2 cells (Fig. 4) rules out the possibility that the enhanced glucose incorporation, in response to insulin, observed in androgen-treated cells, depends on an increased GLUT1 expression induced by androgens. It is well known that insulin, acting through its specific receptors, stimulates [“Hlthymidine incorporation into DNA and the magnitude of this action depends on the number of IR present on the cell surface (Chou et al., 1987; and literature cited therein). On the other hand, androgens, per [‘Hlthymidine incorporation into se, increase

Insulin Concentration (M)

Fig. 5. Effect of androgens on insulin-stimulated total glucose incorporation in HEp-2 cells. Subconfluent cells were incubated for 16 h at 37°C with IO-’ M testosterone co), lo-’ M DHT (0) or with 0.1%’ ethanol (the solvent of androgens) (0). Then, HEp-2 cells were incubated with the indicated concentrations of insulin for 30 min at 37°C. Thereafter, D-[U-“C]glucose (0.5 PCi) was added and incubation was continued for 45 min at 37°C as described under Materials and methods. All the experimental points were assayed in triplicate. The results, expressed as percentage of incorporation over the basal, are the average of four independent experiments that varied within 15% of each other. The values *standard deviation for glucose incorporation in cultures not treated with insulin were 849 + 98 cpm (control). 900 i 90 cpm (DHT) and 897 k 76 cpm (testosterone), respectively.

DNA in HEp-2 cells (Fig. 6). When we measured insulin-stimulated [“Hlthymidine incorporation into DNA in HEp-2 cells, we observed that cells treated with either testosterone or DHT display a 75-95% enhanced incorporation when compared to control cells. Also in this case, DHT is slightly more efficient than testosterone in enhancing insulin-stimulated [ “Hlthymidine incorporation into DNA (Fig. 6). Discussion In this study we demonstrate that androgens enhance both metabolic and growth-promoting insulin actions in HEp-2 laryngeal carcinoma cells by increasing insulin receptor number. This effect is associated with enhanced insulin receptor mRNA levels.

117

I

I

0

IO-"

10.9

1o-‘O huh

10'

10.'

r

IO4

[Ml

Fig. 6. Effect of androgens on insulin-stimulated thymidine incorporation into DNA of HEp-2 cells. Subconfluent HEp-2 cells, in 24.well plates, were incubated in MEM containing 0.5% dialyzed BSA for 30 h at 37°C. Then, 1OK’ M testosterone to), lO_’ M DHT (X) or ethanol (0.196, the solvent of androgens: control cultures (0) were added for 12 h at 37°C prior to the addition of insulin (different concentrations from 0 to lo-’ M). Cells were incubated for 2.5 h at 37°C with 0.5 pCi/well of [jH]thymidine. Cells were washed 3 times with ice-cold PBS and solubilized with 0.03% SDS. Trichloroacetic acid (final concentration 10%) was then added to the extracts for 1 h at 4°C and the precipitates were collected by filtration onto Whatman 3MM filters, washed with 20 ml of ice-cold 10% trichloroacetic acid and assayed for radioactivity. The results are the average of five independent experiments performed in quadruplicate and are presented as cpm (per 10” cells) of [‘Hlthymidine incorporated into DNA & SEM.

Steroid receptors belong to a family of trunsacting regulatory proteins which, upon activation by their respective steroidal ligand, interact with &-acting elements in target genes (so-called ‘steroid-responsive elements’), resulting in the control of gene expression (Evans, 1988; Beato, 1989). The recent cloning of the human estradiol, glucocorticoid, progesterone, mineralocorticoid and androgen receptors demonstrated a high degree of structural similarity within the steroid hormone receptor superfamily (see Beato, 1989 for a review). Insulin receptor mRNA levels are enhanced by glucocorticoids (McDonald and Goldfine, 1988; Rouiller et al., 1988; Briata and Gherzi, 1990) and progesterone (Papa et al., 1990). A notcanonical glucocorticoid-responsive element (GRE) has been localized in the sequence of the

cloned IR promoter region (Araki et al., 1987). Until now, no clear consensus sequence has been defined for progesterone and androgen response elements (Beato, 1989; Govindan, 1990) in the insulin receptor gene promoter region. However, it has been demonstrated that a GRE 15mer is able to mediate the induction of gene transcription by either progesterone or androgens (Ham et al., 1988). Even if no formal evidence has been provided, it is highly believable that glucocorticoid and progesterone receptors, activated by their specific ligand, interact with the not-canonical GRE in the IR promoter (or with not yet identified GREs located 5’ upstream in the gene) increasing IR gene transcription rate. Androgens display a marked effect on growth and differentiation of target organs such as prostate, seminal vesicles, and epididymis (Wilson et al., 1981). Some experimental evidence also indicates that androgens could play a role in sustaining the growth of laryngeal carcinoma (Mattox et al., 1984; Vecerina-Volic et al., 1987). The molecular basis of this phenomenon has not yet been clarified. The possibility that androgens exert a growthpromoting action on target cells by affecting the expression and/or the activity of growth factor receptors has been supported by recent experimental evidence demonstrating that the synthetic androgen R1881 increases either the epidermal growth factor receptor phosphorylation or the inositol phosphate levels in a human prostatic cell line (Sehgal and Powis, 1991). We decided to investigate whether two androgens (testosterone and DHT) affect insulin receptor mRNA levels in a cultured epidermoid cell line, derived from human larynx, in order to investigate the possibility that an androgen-insulin interplay might regulate larynx cancer cell growth and metabolism. Our data demonstrate that androgens both enhance maximal response to insulin and produce a left shift in dose-response curves of about one order of magnitude. Thus, androgens, enhancing insulin action, can increase the proliferative potential of larynx cancer cells, at least in culture. In this respect, of particular interest is the recent observation from Papa and coworkers (1990) that progestins enhance insulin growth-promoting action in human breast cancer cell lines. Also in

11X

this case, the enhanced insulin effectiveness depends on a increment of insulin receptor number induced by the progestin treatment (Papa et al., 1990). On the basis of our data, it is impossible to exclude that androgens increase IR transcripts by enhancing mRNA stability. However, according to previous observations on the effect of glucocorticoids on IR gene expression (McDonald and Goldfine, 1988; Rouiller et al., 1988), it is likely that also androgens enhance IR mRNA levels affecting the transcription rate of the IR gene. Although androgen treatment of HEp-2 cells 3-fold increases IR mRNA levels over the control, we observed that insulin receptor binding as well as insulin-stimulated glucose incorporation and thymidine incorporation into DNA are SO100% higher in androgen-treated cells. Even if we believe that the discrepancy between the increment of IR mRNA levels and the increment of insulin receptor binding is not dramatic, our results cannot exclude that androgens regulate at different levels the biosynthesis of IR in HEp-2 cells. Acknowledgments We are grateful to Morris Birnbaum for the GLUT1 cDNA probe and to Ezio Caniglia for his excellent technical assistance. References Araki. E., Shimada, F., Uzawa. U., Mori, M. and Ebina. Y. (1987) J. Biol. Chem. 262. 16186-16191. Beam. M. (1989) Cell 56, 335-354. Beckford, N.S.. Rood, S.R. and Schaid, D. (1985) Ann. Otol. Y4. 634-640. Birnbaum. M.J.. Haspel, H. and Rosen, O.M. (1986) Proc. Nat]. Acad. Sci. USA 83. 578445788. Bradford, M.M. (1976) Anal. Biochem. 72, 24X-254. Briata, P. and Gherzi, R. (1990) Biochem. Biophys. Res. Commun. 170, I184- 1190. Briata. P., Briata. L. and Gherzi, R. (1990) Biochem. Biophys. Res. Commun. 169, 397-405. Chirgwin. J.M.. Przybyla, A.E., McDonald, R.J. and Rutter, W.J. (1979) Biochemistry 24. 5294-5299. Chou. C.-K.. Dull, T.J.. Russell, D.S., Gherzi, R., Lebwohl, D.. Ullrich, A. and Rosen, O.M. (1987) J. Biol. Chem. 262, 1x42- 1847. Ciaraldi. T.P.. Kolterman, O.G.. Siegel. J.A. and Olefsky, J.M. (197Yl Am. J. Physiol. 236, E621-E625.

Dorfman, R.I. and Shipley, R.A. (1956) Androgens, Biochemistry, Physiology and Clinical Significance. John Wiley, New York. Ebina, Y.. Edery, M.. Ellis, L., Standring. D., Beaudoin, J., Roth, R.A. and Rutter, W.J. (1985) Proc. Nat]. Acad. Sci. USA 82, XO14~8018. Evans. R.M. (1988) Science 240, Xx9-X95. Feinberg, A.P. and Vogelstein, B. (1983) Anal. Biochem. 132, h-13. Forsayeth, J.R., Montemurro, A., Maddux, B., De Pirro, R. and Goldfine, I.D. (1987) J. Biol. Chem. 262, 413444140. Gherzi, R., Sesti, G., Andraghetti, G., De Pirro. R., Lauro, R., Adezati, L. and Cordera, R. (1989) J. Biol. Chem. 264, X627-8635. Govindan, M.V. (lY90) Mol. Endocrinol. 4, 417-427. Ham, J.. Thompson, A., Needham, M., Webb, P. and Parker, M. (1988) Nucleic Acids Res. 16, 526335276. Howard, B.V., Mott, D.M., Fields, R.M. and Bennett, P.H. (1979) J. Cell. Physiol. 101, 129-138. Jakobsen. K.S., Breivold, E. and Hornes. E. (199Ol Nucleic Acids Res. 18. 3669. Kahn, C.R. and White, M.F. (19Xx) J. Clin. Invest. X2, IlSll 1156. Manzella, J.M., Rychlik, W., Rhoads, R.E., Hershey, J.W.B. and Blackshear. P.J. (1991) J. Biol. Chem. 266. 23X3-2389. Mattox, D.E., Von Hoff. D.D. and McGuire. W.L. (1984) Arch. Otolaryngol. 1 IO, 721-724. McDonald, A.R. and Goldfine. I.D. (1988) J. Clin. Invest. XI, 499-504. Moore, A.E., Sabachewsky. L. and Toolan. H.W. (lYS5) Cancer Res. 15. 59X-602. Morley. S.J. and Traugh, J.A. (1990) J. Biol. Chem. 265. 10611~10616. Papa, V., Reese, C.C.. Brunetti, A., Vigneri, R., Siiteri, P.K. and Goldfine. I.D. (1990) Cancer Res. 50, 7X58-7862. Rosen, O.M. (1987) Science 237. 1452~145X. Rouiller, D.G., McKeon, C., Taylor. S.I. and Garden, P. (1988) J. Biol. Chem. 263, 131X5-13190. Saez, S. and Sakai, F. (1976) J. Steroid Biochem. 7, 919-921. Sehgal, I. and Powis, G. (1991) Program of the Annual Meeting of the American Society for Cancer Research, Houston, 19Y1, p. 45 (abstract). Seino, S., Seino. M., Nishi, S. and Bell. G.I. (1989) Proc. Nat]. Acad. Sci. USA 86, 11441 IX. Sesti, G., Marini, M.A.. Montemurro, A., Di Daniele, N., Bertoli. A., Cordera, R.. Andraghetti, G., De Pirro. R., Lauro, R., Monaco, F. and Roche, J. (1988) C.R. Sot. Biol. 182, 167-176. Tewari, D.S., Cook, D.M. and Taub, R. (lY8Y) J. Biol. Chem. 264. 1623X-16245. Tuohimaa. P.T.. Kallio, S., Heinijoki, J.. Aitasalo, K.. Virolainen, E., Karma. P. and Tuohimaa. P.J. (1981) Acta Otolaryngol. 91, 149-154. Vecerina-Volic. VS., Stojkovic, R.R., Krajima, Z. and Gamulin, S. (1987) Arch. Otolaryngol. Head Neck Surg. 113.411-413. Wilson, J.D., George, F.W. and Griffin, J.E. (19X1) Science 211. 1278-1284.