THE JOURNAL OF B I O ~ I C A L CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 4, Issue of January 28, pp. 2516-2520, 1994 Printed in U.S.A.

Dexamethasone Down-regulation of Insulin Receptor Substrate-1 in 3T3-Ll Adipocytes”

(Received for publication, March 29, 1993, and in revised form, August 13, 1993)

Marilyn A. TurnbowSP, Susanne R. Kellern, K. Michael Rice$, and Charles W. Garner$ From the Wepartment of Biochemistry and Molecular Biology, %as Tech University Health Sciences Center, Lubbock. ‘Ikzas 79430 and the WeDartment of Biochemistry, Dartmouth Medical School, Hanovei New Hampshire 03755-3844

Insulin resistance resulting from prolonged exposure of intact animals or cultured cells to glucocorticoids is often attributed to postreceptor signaling defects. To better understand the specific effects of glucocorticoids on insulin signaling, we have characterized the effect of dexamethasone on the expression of an insulin signaling intermediate, the insulin receptor substrate-1 (IRS-1) in 3T3-Ll adipocytes.

Addition of dexamethasone resulted in a 40”7wo de- cline in steady-state IRS-1 protein over 2448 h of treat- ment. Dexamethasone did not significantly change the degradation rate of IRS-1 protein but decreased the net rate of amino acid incorporation into IRS-1 by 87%. Be- tween l and 2.5 h of treatment with dexamethasone, ac- tinomycin D, or both drugs given simultaneously, the concentration of IRS-1 mRNA declined with a half-life of 0.7-1.0 h. However, after 4 h of dexamethasone treat- ment, IRS-1 mRNA concentrations stabilized at -35% of the control level. The dexamethasone-induced decline in IRS-1 protein could be prevented by simultaneous ad- ministration of the glucocorticoid antagonist mifepris- tone, RU38486. These results suggest that in 3T3-Ll adi- pocytes the loss of IRS-1 protein after dexamethasone treatment can be accounted for chiefly by inhibition of the synthesis of IRS-1 mRNA.

In many mammalian tissues, activation of the insulin recep- tor is followed within seconds by the tyrosine phosphorylation of a M, 160,000-180,000 cytosolic protein, the insulin receptor substrate-1 (IRS-1)l (White et al., 1985; Sun et al., 1991). Fol- lowing tyrosine phosphorylation, one or more of the YMXW YXXM motifs of IRS-1 are bound by the src-homology 2 do- mains of phosphatidylinositol 3’-kinase, resulting in activation of the enzyme (Backer et al., 1992).

In cultured adipocytes, intact animals, and human subjects, glucocorticoid excess is associated with insulin resistance, a condition in which one or more of the normal cellular responses to insulin is reduced (Garvey et al., 1989; Grunfeld et al., 1981;

* This research was supported by National Institutes of Health Grant DK 42816 and by a grant from the South Plains Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertise- rnent” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

5 To whom correspondence should be addressed: Dept. of Biochemis- try and Molecular Biology, Texas Tech University Health Sciences Cen- ter, 3601 Fourth St., Lubbock, TX 79430. Tel.: 806-743-2512; Fax: 806- 743-2990.

1 The abbreviations used are: IRS-1, insulin receptor substrate-1; DMEM, high glucose Dulbecco’s modified Eagle’s medium; E64, trans- epoxysuccinyl-~-leucylamido(4-guanidino)-butane; FBS, fetal bovine se- rum; GR, glucocorticoid receptor; PAGE, polyacrylamide gel electropho- resis; RU 486, mifepristone, RU38486; TBS, Tris-buffered saline.

Guillaume-Gentil et al., 1990; van Putten et al., 1985; Wajngot et al., 1992). We have observed that dexamethasone treatment of 3T3-Ll adipocytes causes down-regulation of IRS-1 (Rice et al., 19921, suggesting that defective signal transduction through IRS-1 could contribute to the development of glucocor- ticoid-induced insulin resistance. We have begun an investiga- tion of the role of IRS-1 in this type of insulin resistance by studying IRS-1 regulation in 3T3-Ll adipocytes following treat- ment with dexamethasone.

EXPERIMENTAL. PROCEDURES Materials-Unless otherwise noted, all reagents were obtained from

Sigma. RU 486 was a gift from Roussel-Uclaf, Romainville, France. Cell Culture-3T3-Ll fibroblasts (CL 173, American Type Culture

Collection, Rockville, MD) were cultured in high glucose Dulbecco’s modified Eagle’s medium (DMEM, Sigma D5648) + 10% calf serum in 6- or 10-cm dishes or in T-25 flasks (Sarstedt, Arlington, TX). One to two days after reaching confluence, cells were placed in DMEM + 10% fetal bovine serum (FBS) and subjected to a standard differentiation protocol (Frost and Lane, 1985). Adipocytes were used between 8 and 11 days after the initiation of differentiation. The day before an experiment, cells were fed fresh DMEM plus 10% FBS. Drug dissolved in vehicle or vehicle alone was added 24 h later.

RNA Isolation and RNase Protection Assay--Total RNA was isolated by the method of Chomczynski and Sacchi (1987) using 1.5 ml of TRI ReagentTM (Molecular Research Center, Cincinnati, OH) according to the protocol supplied by the Molecular Research Center. Ribonuclease protection assays were performed on 15-20 pg of total cellular RNA using the RPA IIm kit (Ambion, Austin, TX) as previously described (Turnbow and Gamer, 1993). Riboprobes were synthesized from the T3 promoter of a BluescriptTM SK1+ plasmid (Promega, Madison, WI) con- taining a 1006-base pair fragment of the cDNA coding region for mouse IRS-1 (Keller et al., 1993). Probes were made using the Ambion T3 MAXIscriptTM in vitro transcription kit with reactions -6 PM in [a-32P]CTP (DuPont NEN). After digestion with 1000 unitdm1 RNase TI (Ambion), processed samples were run on 8 M urea (Bio-Rad), 5% Long RangerTM acrylamide (AT Biochem, Malvern PA) minigels. Bands were detected by exposing gels to Kodak XRP-5 x-ray film. The integrated optical intensity of bands was quantitated using a Biolmage Visage 2000 (BioImage, Ann Arbor, MI) computer-assisted image analyzing system. Total RNA was measured by absorption at 260 nm. Because 24 h of dexamethasone treatment does not change total RNA when com- pared with control ( p > 0.1 by Student’s t test; data not shown), inte- grated optical intensities were compared on a per dish basis with the control values indicated within each experiment. Positions of bands were estimated by ethidium bromide staining of a 0.16-1.77-kilobase RNA ladder (Life Technologies, Inc.) run in the same gel. Undigested probe and protected fragments migrated at 410 and 365 nucleotides, respectively. The concentration of IRS-1 mRNA was determined to be approximately 8 copies/control adipocyte by excision and scintillation counting of protected bands as described in the RPA IITM protocol.

Preparation of Cellular Extracts-Extracts for immunoprecipitation and immunoblotting were prepared by placing dishes or flasks of cells on ice, removing medium, and washing cells twice with 20 m Tris, pH 7.4, plus 150 m NaCl (TBS) at 4 OC. Cells were lysed using 1.0 ml of pH 7.6 lysis buffer: 20 mM Tris, 50 mM NaC1, 5 m EDTA, 30 mM sodium pyrophosphate, 50 m NaF, 1 mM sodium vanadate, 1% Triton X-100 (Pierce Chemical Co.) with 1 mM phenylmethanesulfonyl fluoride, 10 y E64, and 1 mM diisopropyl fluorophosphate added to the buffer imme-

2516

Dexamethasone Down-regulation of IRS-1 2517 diately before use. Cells were broken by passage through a 25-gauge needle, and lysates were spun at 13,000 x g for 30 min. Supernatants were either used immediately or were stored a t -70 "C.

Antibodies-Polyclonal a-p3 antibodies were produced in rabbits against the mouse IRS-1 peptide sequence CGANLGTSPALPGDEAAG (Keller et al., 1991). Polyclonal a-CT antibodies were produced in rab- bits against the rat IRS-1 carboxyl-terminal peptide CYASINFQKQPE- DRQ (Sun et al., 1991). Antibodies were affinity-purified by adsorption of the antiserum to the appropriate peptide-agarose column and elution at pH 2.5, followed by neutralization and dialysis (Rice et al., 1993).

Western Blotting-Cell lysates were diluted with 0.33 volume of 4 x sample buffer (380 mM Tris, pH 6.8,8% SDS, 40% glycerol, 4 m~ EDTA, 0.016% bromphenol blue, 81 mM dithiothreitol, and 0.4 mM sodium vanadate), heated to 55 "C for 10 min, and separated by SDS-PAGE on 6% gels. Proteins were transblotted overnight at 23 V to Immobilon P (Millipore, Bedford, MA) in 96 mM glycine, 12.5 mM Tris, 0.01% SDS, and 10% methanol a t 4 "C. Filters were blocked 1 h in 3% instant nonfat dry milk (Carnation) in TBS, washed in TBS, exposed to 2 pg/ml affin- ity-purified a-CT in 1% milk for 2 h, washed, and exposed to a 15000 dilution of horseradish peroxidase-conjugated donkey anti-rabbit IgG (Amersham Corp.) in 1% milk for 30 min. After several more washes, bound antibodies were visualized by chemiluminescence using the ECL system (Amersham). Filters were exposed to Kodak XRP-5 film for 30 s to 5 min. Signals were quantitated using the Visage 2000 image-ana- lyzing system. Total protein was measured using the reagents and protocol in the Pierce Chemical Co. BCA protein assay kit. Because 24 h of dexamethasone treatment did not change total protein when com- pared with control ( p > 0.2 by Student's t test; data not shown), inte- grated optical intensities were compared on a per dish basis with the control values indicated within each experiment.

3sS Metabolic Labeling, Immunoprecipitation, and Visualization- For pulse-chase labeling, adipocytes in T-25 flasks (Sarstedt) were fed with 2.5 ml of DMEM containing 10% FBS and -200 pCi of EXPRE35S3sS 3sS-protein labeling mix (NEG-072, DuPont NEN). After 16 h, cells were washed three times with DMEM and were then fed with 5 ml of unlabeled medium containing 10% FBS plus or minus 100 nM dexamethasone. Cells were washed and lysed as above after the indi- cated incubation times. For pulse labeling studies, cells in T-25 flasks were fed with 2.5 ml of DMEM containing 10% FBS and -200 pCi of 35S-protein labeling mix and were harvested at the indicated times. Immunoprecipitations were performed as follows: for each ml of lysate, 20 pg of affinity-purified a-p3 antibodies was adsorbed to 30 pl (settled volume) of protein A-agarose beads. Tubes were inverted 12 timedmin for 1 h a t 4 "C; beads were washed with TBS and resuspended to 60 pl in TBS containing 1 mM phenylmethanesulfonyl fluoride, 10 p~ E64, and 1 mM diisopropyl fluorophosphate. Each milliliter of cell lysate was precleared by mixing a t 12 rpm for 1 h a t 4 "C with 20 pl (settled volume) of protein A-agarose beads. One milliliter of precleared super- natant was added to 60 p1 of antibody-conjugated beads and was rotated at 12 rpm for 2 h a t 4 "C. Beads were then washed with 1 ml of TBS containing 1 mM phenylmethanesulfonyl fluoride, 10 p~ E64, and 1 mM diisopropyl fluorophosphate. Liquid was removed, and 60 p1 of elution buffer (500 m~ Tris, pH 6.8, 2 mM EDTA, 8 M urea, 100 mM dithio- threitol, 4% SDS, and 0.016% bromphenol blue) was added to each tube. Tubes were heated to 55 "C for 10 min. Eluted protein was separated from the agarose beads by piercing the bottom of the tube with a 25- gauge needle and centrifuging the liquid into another tube. Proteins were separated by SDS-PAGE in 6% gels. Gels were prepared for fluo- rography using ResolutionTM (E. M. Corporation, Chestnut Hill, MA), dried at 55 "C under vacuum, and exposed to Kodak XRP-5 film a t -70 "C. Relative integrated optical intensities were quantitated using the Visage 2000 image-analyzing system. 14C-Methylated proteins ( A m - ersham Corp.) were included on each gel as molecular weight markers.

RESULTS Down-regulation of IRS-1 Protein by Dexamethasone and Ac-

tinomycin &Treatment of 3T3-Ll adipocytes with 100 nM dexamethasone resulted in a decrease of IRS-1 protein to 47% of its original level after 24 h and a further decline to 27% after 48 h, as shown in Fig. 1. In comparison, cells treated with 500 nM actinomycin D, an irreversible inhibitor of RNA synthesis, had 32% of IRS-1 remaining after 16 h. The rates of down- regulation were similar for the first 8 h. After this time, the rate of down-regulation in the presence of dexamethasone was much slower than in the presence of actinomycin D. The

0 4D ED 16D 24C 24D48C48D

Time, h

0 8 1 6 2 4 3 2 4 0 4E

Time, h FIG. 1. Down-regulation of steady-state IRS-1 protein. Adipo-

cytes were treated with vehicle (A), 100 nM dexamethasone (O), or 500 nM actinomycin D (0). At the times indicated, cells were lysed, and IRS-1 protein was detected by immunoblotting as described under "Ex- perimental Procedures." n ? 3; error = S.E. When error bars are not visible, they are smaller than the plot symbol. The inset shows a typical immunoblot of IRS-1 from control ( C ) and dexamethasone-treated (Dl adipocytes.

steady-state concentration of IRS-1 protein in control cells did not change.

Effect of Dexamethasone on Loss and Accumulation of 35S Label in IRS-1-We next examined the rate of degradation of IRS-1 protein by pulse-chase labeling using 35S-labeled amino acids. Upon removal of 35S-labeled medium, the cells were treated with or without 100 nM dexamethasone, and IRS-1 was immunoprecipitated from cell lysates at the times indicated in Fig. 2. The loss of 35S-labeled IRS-1 protein was monitored by fluorescence-enhanced autoradiography of protein. The calcu- lated half-lives of IRS-1 were 12.1 h in control and 10.2 h in dexamethasone-treated cells. By analysis of covariance, the ob- served decrease in half-life with dexamethasone treatment was not significant ( p > 0.2). The value we obtained for the half-life of IRS-1 protein in control adipocytes is less than that reported earlier (Rice et al., 1993). Depending on the passage number and developmental age of the cells, we have observed control half-lives ranging from 12 to 25 h for IRS-1. Consequently, controls and treatments were performed on cells plated at the same time.

We assessed net accumulation of label into IRS-1 protein using medium containing 35S-labeled amino acids plus or mi- nus 100 nM dexamethasone. After an initial lag period, the amount of labeled IRS-1 increased linearly from 2 to 8 h (Fig. 3). After 8 h of treatment, many of the nonspecific protein bands in the dexamethasone-treated sample (Fig. 3, inset) were labeled as intensely as those in the control sample, suggesting that dexamethasone did not decrease accumulation of label into IRS-1 protein by decreasing the uptake of 35S-labeled amino acids. The net rate of incorporation of 35S-labeled amino acids into IRS-1 in dexamethasone-treated cells was 13% of that found in control cells. By analysis of covariance, the difference in the two rates was significant ( p < 0.0001).

Down-regulation of IRS-1 mRNA by Dexamethasone and Ac- tinomycin &During treatment with 500 nM actinomycin D and with 100 nM dexamethasone, IRS-1 mRNA levels declined rapidly to 30% of their initial value (Fig. 4, A and B ) . However, between 4 and 48 h after the initiation of dexamethasone treat- ment, the steady-state concentration of the mRNA stabilized a t -35% of the amount present in control cells.

The rate of loss of steady-state IRS-1 mRNA was compared using treatments of 100 nM dexamethasone alone, 500 nM ac-

2518 Dexamethasone Down-regulation of IRS-1

.- C a .I-

I! a

U

100

5 0

30

cu - a Time, h 2 10 - 1 0 4 8 1 2 1 6 2 0 2 4

Time, h FIG. 2. Degradation of aS-labeled IRS-1 protein. Adipocytes

were pulse-chase labeled with 35S-amino acids as described under “Ex- perimental Procedures.” After removal of labeled medium, cells were incubated with unlabeled medium without dexamethasone (0) or with 100 nM dexamethasone (0) for the times indicated. IRS-I was isolated by immunoprecipitation, separated by SDS-PAGE, and was visualized by fluorescence-enhanced autoradiography of the dried gels. By analy- sis of covariance, the slopes of the two lines were not different ( p > 0.2). n t 3; error = S.E. When error burs are not visible, they are smaller than the plot symbol. The inset shows a typical autoradiogram of 3sS- labeled IRS-1 protein immunoprecipitated from control (C) and dexa- methasone-treated (D) adipocytes.

I I

0 2 6 8

Time, h FIG. 3. Accumulation of aS-labeled amino acids into IRS-1 pro-

tein. Adipocytes were placed in medium containing 35S-labeled amino acids, with no dexamethasone (0) or containing 100 nM dexamethasone (O), as described under “Experimental Procedures.” Cells were har- vested at the times indicated. IRS-1 was isolated and detected as de- scribed in Fig. 2. By analysis of covariance, the slopes of the two lines between 2 and 8 h were significantly different ( p < 0.0001). n 2 3; error = S.E. When error burs are not visible, they are smaller than the plot symbol. The inset shows a typical autoradiogram of 3sS-labeled IRS-1 protein immunoprecipitated from control ( C ) and dexamethasone- treated (D) adipocytes a t 8 h after addition of unlabeled medium.

tinomycin D alone, and both drugs given simultaneously. The reason for the initial lag period is not known, but it was con- sistently observed after treatment with actinomycin D and, to a lesser extent, after treatment with dexamethasone. Because of this lag period, half-lives were calculated from data obtained between 1.0 and 2.5 h of treatment. tm was 0.74 h with dexa- methasone and 0.71 h with actinomycin D. When both drugs were given simultaneously, tu2 was 1.01 h. By analysis of co- variance, the slopes of the three lines were not different ( p > 0.82).

Reversal of the Dexamethasone Effect on IRS-1-While the effect of actinomycin D on RNA synthesis is irreversible, Fig. 5, A and B, shows that dexamethasone down-regulation of both

A 10 I

‘I‘

0 0 1 6 2 4 3 2 4 0 4 0

T h e , h

10 ’ 0.0 1 .0 2.0 3 . 0

T h e , h FIG. 4. Long term (A) and short term ( B ) down-regulation of

IRS-1 mRNA. Adipocytes were treated with vehicle (A), 100 nM dexa- methasone (O), 500 nM actinomycin D (O), or 100 nM dexamethasone plus 500 nM actinomycin D (0). At the times indicated, total RNA was isolated as described under “Experimental Procedures.” IRS-1 mRNA was detected by ribonuclease protection assay performed with -20 pg of total RNAand a 365-nucleotide 32P-labeled riboprobe. Bands in undried gels were visualized by autoradiography. n t 3; error = S.E. When error burs are not visible, they are smaller than the plot symbol. A, the inset shows the 365-nucleotide protected IRS-1 mRNA bands from ribonucle- ase protection assays performed on total RNA from dexamethasone- treated (D) and control (C) adipocytes. B, by analysis of covariance, the slopes of the lines between 1 and 2.5 h were not different ( p > 0.82).

IRS-1 mRNA and protein could be reversed by removal of dexa- methasone. Cells were treated with 100 nM dexamethasone for 24 h, followed by removal of the drug and refeeding with fresh medium. Within 1 day, both IRS-1 mRNA and protein returned to levels found in cells treated with vehicle alone.

As shown in Fig. 6, the ED50 for down-regulation of IRS-1 expression over 24 h was 3 nM dexamethasone. The effect of dexamethasone was maximal at 10 rm and above. Down-regu- lation in the presence of 100 nM dexamethasone was prevented by simultaneous treatment with the glucocorticoid antagonist RU 486 (ED50 = 160 nM RU 486.

DISCUSSION Dexamethasone causes insulin resistance as measured by

several markers, including a reduction in insulin-stimulated glucose uptake (Grunfeld et al., 1981) and a decrease in glucose oxidation (Olefsky, 1975). While the precise cause or causes are not yet known, it is known that dexamethasone affects insulin signaling at several levels. Dexamethasone decreases insulin binding in 3T3-Ll adipocytes (Grunfeld et al., 1981; van Putten et al., 19851, decreases glucose transporter GLUT 1 and GLUT 4 steady-state mRNA (Chu et al., 19901, and depletes cell-sur- face glucose transporters in rat adipocytes (Garvey et al., 1989). In view of these effects on signaling, we investigated the effect of dexamethasone on the expression of another element of in- sulin signaling in 3T3-Ll adipocytes, the insulin receptor sub- strate protein IRS-1.

Steady-state amounts of IRS-1 protein declined by 4670% compared with control values after 2 4 4 8 h of dexamethasone treatment. The rate of degradation of IRS-1 protein was not

~ e x a m ~ ~ ~ a s o n e own-regu~ation of IRS-1 2519

'"1 A

2 3

Time, days

Der Removed

0 2 3

Tlme, days FIG. 5. IRS-1 F&NA (A) and protein ( B ) following removal of

dexamethasone. Adipocytes were treated for 24 h with vehxcle alone

fresh medium, and harvested at the times indicated. Lysates were ana- (0) or with 100 nM dexamethasone (Der) (0). Cells were washed, refed

lyzed for IRS-1 mRNA by ribonuclease protection assay or for protein by immunoblotting as described under '%xperimental Procedures." n 2 3;

plot symbol. error = S.E. When error bum are not visible, they are smaller than the

I t

0 .1 1 1 0 1 0 0 1000 10000

treatment (nM) FIG. 6. Protection of IRS-1 protein in the presence of dexa-

methasone. Adipocytes were treated with the indicated concentrations of dexamethasone (0) or with the indicated concentrations of RU 486 in the presence of 100 it^ dexamethasone (0). Cells treated with dexa- methasone only were given fresh medium on the day treatment was initiated. ARer 24 h of incubation, ceHs were lysed, and IRS-1 protein was d e W by immunoblottin~ as described under '%xperimentai Pro- cedures." n = 3; error = S.E.

significantly different from the rate of degradation in control cells. This is in sharp contrast to the down-regulation caused by chronic insulin treatment, which produces a 5-10-fold increase in the rate of degradation of IRS-1 protein (Rice et ai., 1993). However, we did see a ?-fold decrease in the rate of accwnula- tion of label in IRS-1 protein in dexamethasone-treated cells, suggesting that i ~ b i t i o n of synthesis rather than acceleration of degradation is responsible for the dex~ethason~induced down-regulation of 1-1 protein. Two factors that influence the rate of ECS-1 protein synthesis are the rate of protein trans-

lation and the steady-state concen~ation of mRNA. We decided to ~nvestigate the rate of protein synthesis at a later time because we suspected that the primary effects of dexametha- sone were at the mRNA level.

Using ribonuclease protection to assay IRS-1 mRNA, we de- termined that the steady-state IRS-1 mRNA concentration de- creased to 2640% of its initial value after 4 h of dexametha- sone treatment and remained at this level over the course of the experiment. This decrease in mRNA i s s ~ c i e n t to account for most if not all of the decrease in IRS-1 protein. Increased deg- radation of mRNA could also contribute to a portion of the decline in steady-state protein in the presence of dexametha- sone. Because of the low concentration of IRS-1 mRNA, we chose to evaluate its rate of degradation using ribonuclease protection. The measured half-life in the presence of dexa- methasone or actinomycin D was 0.7 h. The half-life in the presence of both was 1.0 h, a value that is not significantly different from the haIf-life in the presence of either dexametha- sone or actinomycin D alone. The absence of an effect of dexa- methasone on the rate of mRNA degradat~on in the presence of actinomycin D suggests that dexamethasone inhibits the syn- thesis of IRS-I mRNA without s i ~ i ~ ~ a n t l y affecting mRNA s~bili ty. It is possible, however, that actinomycin D blocks the ability of d e x ~ e t h a s o n ~ to regulate mRNA stability.

As shown in Fig. 5, IRS-1 protein could be protected from dexamethasone-induced down-regulation by simultaneous treatment with the glucocorticoid antagonist RU 486. This in- dicates that the dexamethasone effect on IRS-1 is probably mediated through the glucocorticoid receptor. RU 486 has an affinity for the GR that is similar to that of dexamethasone (Lazer and Agarwal, 1986). Our observed ED60 for reversal of the effect of 100 m dexamethasone was 160 IIM RU 486, a nearly equimolar concentration. If RU 486 acts as an antago- nist in this system, it is probable that RU 486 exerts its effect by competition with dexamethasone for binding to the GR.

The IRS-1 response to dexamethasone appears to be bipha- sic. We observed a time-dependent decline in steady-state con- centration of IRS-1 mRNA during the first 4 h of dexametha- sone treatment. Thereafter, IRS-1 mRNA stabilized at -35% of its initial concentration. This may be the results of dexametha- sone-induced d o ~ - r e ~ l a t ~ o n of GR itself, Hoeck et al. (1989) observed in NIH 3T3 cells that after 3 h of exposure to dexa- methasone, the GR mRNA was down-re~lated to a new steady state that was -25% of its previous concentration. Our own preliminary observations indicate that the GR is similarly regulated in 3T3-Ll adipocytes.2 Consequently, the desensiti- zation of the IRS-1 response after 4 h of treatment may be caused by the dexamethasone-induced down-regulation of the GR.

In summary, dexamethasone, unlike insulin, does not exert a significant effect on the rate of degradation of IRS-1 protein in 3T3-Ll adipocytes. Dexamethasone does effect a reversible de- crease in the mRNA for IRS-l. Most of the decline in IRS-1 protein observed after 24 h of dexamethasone treatment is attributab~e to a reduction in transcription of IRS-1 mWA.

R E F E R E ~ C ~ S Backer, J. M., Myers, M. G., Shoelaon, S. E., Chin, D. J., Sun, X. J., Miralpek, M.,

Hu, E, Margolis, B., Skolnik, E. X, Schlessinger, J., and White, M. F. (1992) EMBO J. 9,346943479

Chu, D. T,, Isaaeeon, C, M., and Bell, I? A. (1990) Bioctrem. Soc, Banns. 18, 1247-1248

Chomezynski, P., and Sacchi, N. (1987) A w l , Bioehpm. 162,156-159

kost, S. C., and Lane, M. n. (1985) J. Bwl. Chpm. Uro, 2646-2652 Garvey, W. T., Huecksteadt, T. €?, Monzon, R., and Marshall, S. (1989) E d m ~ i d -

M. A. Turnbow* S. R. Keller, XI M. Rice, and C. W. Gamer, unpub- lished data.

2520 Dexamethasone Down-regulation of IRS-1

Grunfeld, C., Baird, K., Van Obberghen, E., and Kahn, C. R. (1981) Endocrinology

Guillaume-Gentil, C., Rohnedeanrenaud, F., Abramo, F., Bestetti, G. E., Rossi, 0.

Hoeck, W., Rusconi, S., and Groner, B. (1989) J. Biol. Chem. 264,1439614402 Keller, S. R., Kitagawa, K.,Aebersold, R., Lienhard, G. E., and Garner, C. W. (1991)

Keller, S. R., Aebersold, R., Garner, C. W., and Lienhard, G. E. (1993) Biochirn.

Lazar, G., and A g a n v a l , M. K. (1986) Biochem. Biophys. Res. Commun. lS4.44-50 Olefsky, J. M. (1975) J. Clin. Inuest. 66, 149S1508

ogy 134,2063-2073

10s. 1723-1730

L., and Jeanrenaud, B. (1990) Endocrinology 128, 1873-1879

J. Bwl. Chem. 286,12817-12820

Biophys. Acta 117% 323-326

Rice, K. M., Lienhard, G. E., and Garner, C. W. (1992) J. Biol. Chem. 267, 10163- 10167

Rice, K. M., Turnbow, M. A., and Garner, C. W. (1993) Biochem. Biophys. Res. Commun. 190,961-967

Sun, X. J., Rothenberg, P., Kahn, C. R., Backer, J. M., Araki, E., Wilden, F! A. Cahill, D. A., Goldstein, B. J., and White, M. E (1991) Nature 362.73-77

‘hrnbow, M. A., and Garner, C. W. (1993) Bio&hnques 16,267-270 van Putten, J. I? M., Wieringa, T., and Krans. H. M. J. (1985) Diabetdogia 28

Wajngot, A,, Giacca, A., Grill. V., Vranic, M., and Efendic, S. (1992) Proc. Natt.

White, M. E , Maron, R., and Kahn, C. R. (1985) Nature 918,183-186

445-451

Acad. Sci. U. S A . 88,6035-6039