Rapid synergistic interaction between thyroid hormone and

Communication

THEJOURNAL OF BIOLOGICAL CHEMISTRY Vol. 261 No. 21 Issue of July 25 pp. 9583-9586 1986 The kmericin Society of Bioiogicd Chernisd, Inc. Printed in U.S.A.

Q 1986 by

Rapid Synergistic Interaction between Thyroid Hormone and Carbohydrate on rnRNAsl4 Induction*

several specific mRNAs (1-7). Moreover, we and others (812) demonstrated that carbohydrate and triiodothyronine (T3’) interact synergistically to regulate the level of these lipogenic enzymes and theirmRNAs. Nevertheless, the mechanism by which this interaction occurs remains unclear. The development of a cDNA clone to hepatic mRNAslr has (Received for publication, April 8, 1986) provided a useful model for the study of thyroid hormone action. Because this mRNA responds to T3 within 20 min, it Cary N. Mariash, Steven SeeligS, is likely that it is a primary target gene (13-15). In addition, Harold L. Schwartz, and Jack H. Oppenheimer its tissue-specific expression suggests that itis associated with From the Division of Endocrinology and Metabolism, lipogenesis (16). Previously we demonstrated that in the Department of Medicine, University of Minnesota, steady state this mRNA responds both to TS and carbohyMinneapolis, Minnesota 55455 drates. However, neither the site of interaction of T3 and Recent studies have shown that hepatic rnRNASl4 carbohydrates, northe temporal relationshipswith other thyresponds rapidly to thyroid hormone administration. roid related events, such as enhanced carbohydrateutilization Moreover, this mRNA is known to increase in mass (17, 18), have been defined. The availability of the cDNA to with the administration of a high carbohydrate fat- this rapidly responsive mRNA has now allowed us to address free diet. Therefore, it appears to share many of the these questions.

same properties of the known hepatic lipogenic enzymes. Because the lipogenic enzymes display a synergistic interactionbetween thyroid hormones and carbohydrates, we investigated the kinetics of response of mRNAslr to carbohydratefeeding, as well as its inter(T3). W e found that actionwithtriiodothyronine mRNASI4responds rapidly to the dietary administraa 2-fold increase tion of sucrose in euthyroid rats, with within 30 min, and a 25-fold increase by 4 h. On the other hand, when given to hypothyroid rats, sucrose ultimately lead to only a 2-%fold increase in thelevel of mRNAS14,attaining a level less than that found in starved euthyroid rats. The diminished response of mRNASI4to sucrose in hypothyroidism could not be enhanced by insulin administration. However, administration of replacement doses of T3(400 ng/100 g of body weight) immediately restored the rapid response to sucrosefeeding. The response of sucrose andT3 was synergistic. Dose-response studies with Ts indicated that the rapid interaction between T3 and sucrose was limited by the occupancy of the T3nuclear receptor. A similar synergistic response to T3 and glucose was noted in primary hepatocyte cultures, thus indicating is not due that thesynergism between these two stimuli to changes in extrahepatic hormones or metabolites. Our data aremost consistent with the hypothesis that the T3-nuclear receptor complex multiplies a signal generated by carbohydrate metabolism to induce hepatic mRNAS14. The interaction does not appear to require the preliminary induction of carbohydratemetabolizing enzymes and theirmRNAs.

MATERIALS AND METHODS

Animal Treatment-Male Sprague-Dawley rats, 150-175g,were rendered hypothyroid by addition of 0.025% methimazole to the drinking water for 3 weeks. During the 3rd week, all rats showed the expected arrest in weight gain. Food was removed from the cages at 400 pm the day prior to killing. The morning of the experiment rats were giveneither 400 ng of T3/100g of body weight intravenously, or 1 cc/IOO g of body weight of 60% sucrose by oral gavage, or both. Control animalsreceived saline intravenously or by gavage as appropriate. To some animals, 10 units of regular pork insulin (U-100, Eli Lilly & Co.) was administered intraperitoneally. Hepatocyte cultures were prepared as described (19),except that cells were maintained on Falcon tissue culture Petri plates (BectonDickinson and Co., Cockeysville, MD) without collagen coating. After an initial 4 h in modified Williams Medium E containing 10% calf serum, the medium was changed to the testmedium. The serum-free test medium was supplemented with sodium selenite (2 X IO-’ M), human transferrin (5 ng/ml), bovine serum albumin (500 mg/liter), M),adenine (8.7 aprotinin (2000 units/liter), orotic acid (6.4 X X 1O“j M), and thymidine (6.2 X 10” M), in addition to insulin and dexamethasone as previously noted. Transferrin, albumin, aprotinin, and dexamethasone were obtained from Sigma. Spot 14 mRNA Qunntitation-Total hepatic RNA was isolated from liver in guanidine hydrochloride, or from hepatocyte cultures in guanidine thiocyanate, as described elsewhere (20). Preparation of the (32P]cDNAprobe, and dot-blot hybridization has been reported (14). For these studies, an internal standard containing mRNAsl, was applied to all dot-blots to allow comparison between experiments. The internalstandard was derived from one large pool of liver total RNA. One unit is equivalent to 1%of the internal standard. Dotblots were quantitated by video densitometry (21) or counting of the individual dots by liquid scintillation spectrometry. Comparisons were performed by one-way analysis of variance, Statistical differences between groups were calculated using the Newman-Kuells procedure. All values are expressed as themean standard error of at least four determinations per group.

*

There is general agreement that thyroid hormone action is initiated at the nuclear level and involves the regulation of * This work was supported by National Institutes of Health Grants 1-R01-AM32885, 1-R01-AM19812, and T-32-AM07203. Part of this work was presented at the 9thInternational Thyroid Congress, September, 1985, Sao Paolo, Brazil. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Present address: Dept of Pediatrics, University of Minnesota, Minneapolis, MN 55455.

RESULTS

We have previously demonstrated that Ta administration to hypothyroid rats leads to a rapid induction of mRNASI4(9, 10). On the other hand, no information is available regarding the response kinetics of mRNASI4to carbohydrates. Therefore, prior to studying the interaction of T3and carbohydrates, it became necessary to determine the rapidity of response to

9583

The abbreviation used is: TB,triiodothyronine.

9584

Rapid T3 and Carbohydrate Interaction on mRNASII

a carbohydrate stimulus. To this end, we starved euthyroid rats overnight to minimize the effects of nocturnal eating and to assure a uniform base line. The following morning (8:30 am) 1 cc of 60% sucrose was administered to all rats by oral gavage. Fig. 1 demonstrates the response of mRNAS14during the first 4 h after the sucrose stimulus. A significant ( p < 0.0,5)increase in the level of mRNAs14could be measured 30 min after the carbohydrate load. This rapid and linear response is quite similar to that previously noted for mRNASI4 induction by thyroid hormone (14). These data indicate that in the euthyroid rat, aproduct related to carbohydrate metabolism is capable of rapidly and directly inducing the formation of mRNAs14.Furthermore, the magnitude of this response is nearly as great as the response to T3. Previously we had shown a synergism between carbohydrate feeding and thyroidal status on the level of several of the lipogenic enzymes (5). Since mRNAs14also responds to both thyroid hormone and carbohydrate feeding, Euthyroid we questioned whether this mRNA also displayed an interaction between these two stimuli. As anticipated, Fig. 2 shows that the acute administration of 60% sucrose to overnight-starved hypothyroid rats leads to only a 6.7 unit increase in the level of mRNAs14. Indeed, the level attained4hafter sucrose administration was less, although not significantly, than that observed in the starved euthyroid controls (8.7 & 6.0 units uersus 13.1 f 13.0 units, respectively). Thus, the mRNAs14 response of the hypothyroid starved rat is markedly resistant to carbohydrate feeding. In earlier studies we had shown that the liver of hypothyroid, starved, and diabetic rats demonstrate qualitatively and quantitatively similar patterns of mRNAs, as determined by two-dimensional gel electrophoresis of in vitro translated products (11). Since starvationand diabetes mellitus are

t Hypothyroid

FIG. 2. Effect of sucrose on mRNAsll in 4 h. A representative experiment in overnight-starved euthyroidand hypothyroid rats. The animals were treated as in Fig. 1. All rats were killed 4 h after sucrose or saline. The dotted bars represent animals given saline; the hatched bars animals given sucrose. Each bar represents themean and standard error of 4 rats. In some experiments hypothyroid rats failed to respond significantly tosucrose within 4 h.

associated with low insulin concentrations, we wished to test the possibility thatthe diminished increase in hepatic mRNAs14elicited by carbohydrate in thehypothyroid animal was due to an inadequate insulin response. Because preliminary experiments demonstrated that 10 units of regular insulin/100 g of body weightproduced significant hypoglycemia, all rats received a large amount of sucrose ( 5 m1/100 g of body weight of 60% sucrose) in addition to the insulin. Although this is a relatively large dose of insulin, and despite a significant diminution in plasma glucose induced by insulin admin'OOT istration (676 -+ 82 mg/dl uersus 266 2 104 mg/dl, hypothyroid versus hypothyroid insulin p = 0.0145),no significant increase in the level of mRNAs14 above that attained with sucrose alone was found (Fig. 3). Thus, the diminished response to carbohydrate feeding in the hypothyroid animal cannot be attributed to insufficient insulin secretion. Because thyroidal status influences the metabolism of glucose (17, 18), we asked whether the lack of response to carbohydrate in the hypothyroid rat could be due to diminished activity of the enzymes responsible for the hepatic glucose metabolism. Furthermore, we had previously found that a product derived from the mitochondrial metabolism of pyruvate is responsible for the carbohydrate induction of malic enzyme mRNA as well as for mRNAS14(19, 20). If the diminished response of mRNAs14to carbohydrate in hypothyroidism was due to a reduction of the glucose-metabolizing enzymes, one would anticipate alag of several hours to several days after administration of TSbefore the response to carboI I I I I hydrate feeding is restored. However, when we administered 1 2 3 4 a replacement dose of T3 (400 ng/100 g of body weight) along with sucrose gavage to hypothyroid rats, we found that the4h response to sucrose was not significantly different from that Time (hours) obtained after feeding sucrose to euthyroid rats (74.5 st 14.6 FIG. 1. Time course of response to sucrose in euthyroid rats. versus 77.9 & 14.4 units, respectively). Thus, T3immeunits After overnight starvation, euthyroid rats were administered 60% sucrose (1 cc/100 g of body weight) by gavage. At the times indicated, diately restored the ability to respond to carbohydrate feeding. 4 rats from each group were killed. Total RNA was extracted from Moreover, the level of mRNAs14 was significantly greater than the liver, and mRNAsll level was quantitated by dot-blot hybridiza- the sum of the individual responses to T3 and sucrose. The tion. The ordinate gives the relative level of mRNAs14corrected for rapid synergism between T3and sucrose is summarized in Fig. an internal standard applied to all dot-blots. Each point depicts the mean and standard error. The 30-min time point is significantly ( p 4. This finding suggests that theTs-nuclear receptor complex interacts with the carbohydrate signal and that this process < 0.05) greater than the0 time control.

+

Rapid T3 and Carbohydrate Interaction on

*O

T

Starved

Sucrose Sucrose

+ Insulin

Euthyroid

+

Sucrose

FIG. 3. Effect of insulin in hypothyroid rats. The experimental protocol was identical to thatin Fig. 1.Insulin (10units of regular pork insulin, subcutaneously) was administered simultaneously with the sucrose gavage ( 5 cc/lOO g of body weight). Each bar is the mean and standard error of 4 rats. The euthyroid control rats did not receive insulin. Quantitation of mRNAslowas assayed by cutting each dot from the nitrocellulose filter, and counting was done by liquid scintillation spectrophotometry.

loo

7

1

2

3

9585

posed by Rodbard (22). The response to sucrose alone was 5.3 f 1.1 units, whereas the maximum response (at 200 pg of T3 + sucrose) was 118.1 f 10.8 units. However, 90% of the maximum response occurred with 1.5 pg of Ts, a dose which leads to near saturation of the Ts-nuclear receptor complex during the 4 h of the study (23). Therefore, the ability of T3 to interact withsucrose in the inductionof mRNAS14appears to be limited by the occupancy of the Ts-nuclear receptor complex. Lastly, we wished to determine whether the interaction between TSand carbohydrates required alterations in extrahepatic hormones or metabolites. Therefore, we cultured isolated hepatocytesfor 3days in serum-freemedium containing either low glucose (5.5 mM glucose, no T3),maximally effective concentrations of glucose (27.5 mM glucose, no T3),maximally effective concentrations of T3 (5.5 mM glucose, 5 x lo-? M T3),or both (27.5 mM glucose, 5 X M TJ. The results are shown in Fig. 5. Whereas glucose or T3 enhancement alone led to a 2-3-fold increase in the level of mRNAsI4, thecombined treatment led to a 25-fold increase inthe level of mRNASI4.These data indicate that synergistic the interaction between T3and carbohydrates occurs within the hepatic cell withoutrequiringalterationsinextrahepatichormones or metabolites. DISCUSSION

These studies were undertaken to characterize further the response of mRNAsI4tocarbohydratesandtodetermine whether T3 and carbohydrates interacted synergistically to regulate the level of this mRNA. Our finding in euthyroid 150

T+

mRNASI4

T

T

4

Time (Hours) FIG. 4. Rapid synergistic response to Ta and sucrose in hypothyroid rats. The animal manipulationand RNA extraction were identical to thatdescribed in Fig. 1. Where indicated, T3(400 ng/100 g of body weight) was administered by tail vein. Each point represents in the the mean and standarderror of 4-24 rats. The standard errors 0 time controls, 1 and 2 h Ts. 1 and 2 h sucrose, are obscured by their respective points. Animals given sucrose alone are depicted by open circles, T, alone by closed squares, and T3 + sucrose by open squares. The shaded area indicates the 95% confidence interval for an additive response at 4 h based on the responses to T3 or sucrose alone.

does not depend upon a preliminary induction of other proteins. In orderto localize further the siteof interaction of T3with sucrose on the induction of mRNAs14, we examined the 4-h response to sucrose with varying concentrations of T3 from 0.4 ng/100 g of body weight t o 200 pg/lOO g of body weight. The resultswere fit (correlation coefficient = 0.89) by nonlinear least squares analysis to the four-parameter model pro-

Control Glucose T3 Both FIG.5. Levels of mRNASl4 in hepatocyte cultures. Hepatocytes obtained from fed, euthyroid rats were cultured as indicated under “Materials and Methods”. When indicated, maximal amounts (determined from previous experiments) of T3or glucose were added to thecultures 4 h after plating. Total RNA was extracted from four plates ineach group. The datarepresent the mean and standard error for a representative experiment. Two other experiments show similar results.

9586

Rapid T3and Carbohydrate Interaction mRNAs14 on

ratsthat mRNAs14 responded as rapidly to sucrose asit responds to T3suggests that thecarbohydrate stimuluswhich regulates this mRNA, andperhapsothercarbohydrate-responsive mRNAs, also acts at a nuclear site. Since other sites of regulation cannot be excluded, further studies arerequired to support this hypothesis. Evidence to support the hypothesis that the T3-nuclear receptor complex regulates the level of this mRNA in the nucleus, comes from studies from the laboratory of Towle and co-workers (13), who have shown that theheteronuclear RNA precursor of mRNAS14rises following T3administration prior to a measurable increase in the mature mRNA. On the other hand, transcriptional “run-on” assays have indicated that the increase in nuclear and cytosolic mRNAs14 followingboth T3 and carbohydrate feeding is associated with only a transient and minimal increase in transcription rate (24). If the run-on assays correctly reflect the in uiuo transcription rate, this mRNA is regulated by alterations in the stability of the precursor heteronuclear RNA. The precise mechanism by which hormones and metabolites lead to precursor stability within the nucleus remains to be determined. A surprising finding in our study was the poor response to sucrose feeding in the hypothyroid rat. Previously we had found that feeding a high carbohydrate fat-free diet for several weeks to hypothyroid rats led to large increases in mRNAS14 (9).Because hypothyroid rats may have diminished pancreatic responses to glucose (25), we felt that perhaps inadequate insulin secretion accounted for the slow response to sucrose feeding. However,the results following insulin administration in the current study eliminated this possibility. Large doses of insulin, sufficient to cause a significant decrement in the plasma glucose content, did not lead to enhanced production of mRNAs14. On the other hand, the rapid enhancement of the diminished response in hypothyroidism by administration of a relatively small dose of T3 indicates that acarbohydratederived signal can be immediately produced in the hypothyroid rat. Such a resultis based on the assumption that ittakes several hours to several days to restore all the carbohydratemetabolizing enzymes to normal (26). A rapid response to carbohydrate immediately after T3 replacement would suggest that the carbohydrate signal interacts with the T3 signal in the nucleus of the hepatic cell. This conclusion is supported by the finding that theinteraction of sucrose with T3is limited by the occupancy of the T3-nuclear receptor complex. Furthermore, the data suggest that T3 acts by multiplying this carbohydrate signal at the nuclear level to produce the synergistic response. On the basis of previous studies, we believe that the immediate carbohydrate signal is derived from the mitochondrial oxidation of pyruvate (19, 20). Nevertheless, when a high carbohydrate fat-free dietis administered over a long time, other factors must also be generated which lead to the amplification of the carbohydrate signal and allow large increases in the amount of mRNAs14 to be produced. The synergistic interaction between T3 and carbohydrates and the demonstration that this synergism occurs directly in the hepatic cell, indicate that mRNAsI4 is a useful model for the study of the molecular regulation of hepatic lipogenic enzymes by both hormones andintracellular metabolites.

Moreover, the rapid response of this mRNA to both T3 and carbohydrates suggests that theresponse of mRNAsI4to these signals is primary, and notmediated through the induction of other hepatic or extrahepatic mRNAs or proteins. Finally, the rapid restoration of the response to carbohydrates provides the best evidence tht the T3-nuclear receptor complex multiplies within the nucleus a carbohydrate-generated signal responsible for the stabilization of the heteronuclear RNA precursor of mRNAs14. Acknowledgments-We want to thankDr. Henry Pitot andNancy Turner for the serum-free hepatocyte media recipe. We also wish to thank Robert Gunville and Anna Martinez-Tapp, Mary Ellen Domeier, and Deborah Iden for their excellent technical help. The administrative assistance of Patrice Schaus and the secretarial support of Kate Steinmeyer are greatly appreciated. REFERENCES 1. Fitch, W. M., and Chaikoff, I. L. (1960) J. Biol. Chem. 235,554557 2. Gibson, D. M., Lyons, R. T., Scott, D. F., and Muto, Y. (1972) Adu. Enzyme Regul. 10,187-204 3. Diamant, S., Gorin, E., and Shafrir, E. (1972) Eur. J. Biochem. 26,553-559 4. Roncari, D. A. K., and Murthy, V. K. (1975) J. Biol. Chem. 250, 4134-4138 5. Morris, S. M., Nilson, J. H., Jenik, R. A., Winherry, L. K., McDevitt, M.A., and Goodridge, A. G. (1982) J . Biol. Chem. 257,3225-3229 6. Magnuson, M. A., and Nikodem, V. M. (1983) J. Biol. Chem. 258.12712-12717 7. Sul, H., Wise, L. S., Brown, M.L., and Rubin, C. S. (1984) J. Biol. Chem. 259,555-559 8. Mariash, C. N., Kaiser, F. E., Schwartz, H. L., Towle, H. C., and Oppenheimer, J. H. (1981). J. Clin. Inuest. 65, 1126-1134 9. Liaw,C.,Seelig, S., Mariash, C.N., Oppenheimer, J. H., and Towle, H. C. (1983) Biochemistry 22,213-221 10. Topliss, D. J., Mariash, C. N., Seelig, S., Carr, F. E., and Oppenheimer, J. H. (1983) Endocrinology 112,1868-1870 11. Carr, F. E., Bingham, C., Oppenheimer, J. H., Kistner, C., and Mariash, C. N. (1984) Proc. Nutl. Acad. Sci. U. S. A. 81, 974978 12. Miksicek, R. J., and Towle, H. C. (1982) J. Biol.ChPm. 257, 11829-11835 13. Narayan, P., Liaw, C.W., and Towle, H. C. (1984) Proc. Nutl. Acad. Sci. U. S. A. 81,4687-4691 14. Jump, D. B., Narayan, -P., Towle, H., and Oppenheimer, J. H. (1984) J. Biol. Chem. 259, 2789-2797 15. Seelig, S., Jump, D. B., Towle, H. C., Liaw, C., Mariash, C. N., Schwartz, H. L., and Oppenheimer, J. H. (1982) Endocrinology 110,671-673 16. Jump, D. B., and Oppenheimer, J. H. (1985) Endocrinology 117, 2259-2266 17. Okajima, F., and Ui, M. (1979) Biochem. J. 182,565-575 18. Muller, M., and Seitz, H. (1980) Pflugers Arch. 386, 47-52 19. Mariash, C. N., and Oppenheimer, J. H. (1984) Metabolism 33, 545-552 20. Mariash, C. N., and Schwartz, H. L. (1986) Metabolism 35,452456 21. Mariash, C. N., Seelig, S., and Oppenheimer, J. H. (1982) Anal. Biochem. 121,388-394 22. Rodbard, D. (1974) Endocrinology 94, 1427-1437 23. Oppenheimer, J. H., Coulombe, P., Schwartz, H. L., and Gutfeld, N. W. (1978) J. Clin. Inuest. 60,987-997 24. Narayan, P., and Towle, H. C. (1985) Mol. Cell. Biol. 5, 26422646 25. Lenzen, S., Joost, H. G., and Hasselblatt, A. (1976) Endocrinology 99,125-129 26. Muller, M., and Seitz, H. (1984) Klin. Wochenschr 62, 11-18