Glucocorticoids Regulate the Induction of Phosphoenolpyruvate ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY Q 1993 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 268, No. 17,Issue of June 15, pp. 12952-12957,1993 Printed in U.S.A.

Glucocorticoids Regulate the Induction of Phosphoenolpyruvate Carboxykinase (GTP) Gene Transcription during Diabetes* (Received for publication, January 27, 1993, and in revised form, March 19, 1993)

Jacob E. Friedman$, Jeung S. Yunj, YashomatiM. Patelll, Mary M. McGrane, and Richard W.Hanson From the Pew Center for Molecular Nutrition and Department of Biochemistry, Case Western Reserve University School of Medicine, Cleveland,Ohio 44106-4935 and the SEdison Animal Biotechnology Center, Ohio University, Athens,Ohio 45107

Thehormonalregulationoftranscriptionofthe several factors. First, the liver in diabetic animals is resistant phosphoenolpyruvate carboxykinase (GTP)(4.1.1.32) to insulin, which manifests itself as a failure of normal levels (PEPCK) gene during diabetes was studied using trans- of insulin to suppress gluconeogenesis (1).Second, the coungenic mice containing a chimeric gene consisting of ter-regulatory hormone glucagon is usually elevated in segments of the PEPCK promoter (-2000/+73, -4601 NIDDM (3) and has been shown to contribute to increased +73,-366/+73) linked to bovinegrowthhormone glucose production by the liver (4). A third factor, which has (bGH) reporter gene. The effect of diabetes and insulinreceived relatively little attention, is the role of circulating on transgenic mice containing a mutation in cAMP glucocorticoids in maintaining gluconeogenesis. The imporregulatory sequences at -901-82 and-2501-234 was also studied. In addition, we analyzed the transcrip- tance of glucocorticoids in the regulation of blood glucose in tional response of the PEPCK gene to adrenalectomy, diabetes was first reported in 1936 by Long and Lukens (5), the administration of glucocorticoids, and alterations who showed that adrenalectomy of diabetic animals reduced in dietary protein and carbohydrate. Our results indi- hyperglycemia. Exton and Park(6) laterdescribed the failure cate that deletion of the insulin regulatory sequence of of isolated perfused liver from an adrenalectomized rat to the PEPCK promoter did not affect dietary control of respond to a gluconeogenic stimulus of glucagon or CAMP, unless first pretreated with dexamethasone. Glucocorticoids PEPCK gene expression. However, glucocorticoids and the glucocorticoid regulatory unit appear to beessen- have generally been thought to play a “permissive” role by tial for induction ofPEPCK gene transcription by di- enhancing the availability of gluconeogenic substrates and abetes. By contrast, mutation of cAMP regulatory ele- increasing the sensitivity of the liver to theactions of glucagon ments of the PEPCK promoter did not limit induction and catecholamines (7,8). However, these steroids have been of PEPCK transcription by diabetes, nor did itaffect found to have substantial effects on transcription of a number negative regulation of transcription byinsulin. These of genes coding for hepatic enzymes that regulate gluconeoresults provide evidence for the interaction of insulin genesis, particularly PEPCK. and glucocorticoid regulatory elements in the control PEPCK catalyzes the conversion of oxaloacetate to phosof PEPCKgene transcription and suggest an important phoenolpyruvate and is considered a key regulatory step in role of glucocorticoids as a gluconeogenic activator gluconeogenesis (reviewed in Ref. 9). PEPCK activity (lo), during diabetes. protein (1l),and mRNA (12) are all elevated in animalmodels of NIDDM and in streptozotocin-induced diabetes (11).Unlike many other regulatory enzymes in metabolic pathways, the level of PEPCK in a tissue is primarily regulated by Increased hepatic gluconeogenesis is the primary factor hormonally induced changes in gene transcription (13, 14), responsible for fasting hyperglycemia in patients with diabe- and/or mRNA stability (15, 16). PEPCK mRNA has a halftes (1).In type I diabetes (insulin-dependent), insulin defi- life of approximately 30 min, so that changes in the transcripciency promotes unrestrained gluconeogenesis which is exag- tion rate of the gene have a marked effect on the rate of gerated by impaired glucose transport in peripheral tissues, enzyme synthesis (17). Hormonal signals which control most notably skeletal muscle (2). In non-insulin-dependent PEPCK gene transcription do so by altering a complex patdiabetes mellitus (NIDDM),‘ excessive glucose production tern of protein-protein and protein-DNA interactions. A numoccurs despite availability of insulin (1,2); thismay be due to ber of protein binding domains inthe PEPCKpromoter have * This work was supported in part by Grants DK-21859 and DK- been identified by DNase I footprinting analysis (18-21) and 2441 from the National Institutes of Health and by grants from the gel shift assay (22-25). The sequences necessary for many of Pew Charitable Trusts and the Thomas A. Edison Program of the the regulatory properties of the PEPCKpromoter are thought State of Ohio. The costs of publication of this article were defrayed to be contained within 500 base pairs of the start site of in part by the payment of page charges. This article must therefore transcription (24, 26). be hereby marked “adoertisement” in accordance with 18 U.S.C. Mutational analysis of the PEPCK promoter has demonSection 1734 solely to indicate this fact. $ Supported by Individual National Research Service Award DK- strated that the transcriptional regulation by CAMP occurs F3208477 from National Institutes of Health. TOwhom correspond- through DNA regions CRE-1 (-90/-82) and P3(1) (-2501 ence should addressed. Tel.: 216-368-6166; Fax: 216-368-4544. -234) (24), whereas glucocorticoids appear to activate tranII Trainee of the Metabolism Training Program supported by Na- scription through a complex glucocorticoid regulatory unit tional Institutes of Health Grant DK-07319. (GRU) located between -455 and -350 (27). A region of the The abbreviations used are: NIDDM, non-insulin-dependent diabetes mellitus; PEPCK, phosphoenolpyruvate carboxykinase; GRU, PEPCK promoter (between -407 and -416),which maps glucocorticoid regulatory unit; IRS, insulin response sequence; bGH, within the GRU, has also been identified by O’Brien et al. (28) as an insulin response sequence (IRS). A thyroid horbovine growth hormone.

12952

Regulation of PEPCK Gene Transcription duringDiabetes

12953

of 400 mg/dl 5-7 days before the study and were assigned to either mone responsive sequence is also present in the PEPCK or insulin replacement group. While the half-life of PEPCK promoter between -322 and -308 (29). In addition, liver- placebo mRNA is 30 min (8), bGH mRNA half-life is 48 h (35); it was specific expression of the PEPCKgene requires only proximal therefore necessary to administer insulin for an extended time period DNA regulatory elements located between base pairs -460 to observe the negative effect of insulin on the concentration of bGH and +69 after the start site of transcription (26, 30-32). A in diabetic mice. Animals receiving insulin were implanted with a preliminary analysis indicated that a segment of the PEPCK subcutaneous pellet that released insulin at 1.2 units/day for 5 days. promoter (region from base pair -460 to +69) provided the This treatmentrestored blood glucose to normal levels (100-150 mg/ dl). To observe the acute effects of insulin on transcription of the necessary cis-acting elements for dietary and hormonal re- PEPCK gene and the PEPCK-bGH transgene, diabetic mice not sponsiveness to CAMPand glucocorticoids in transgenic mice receiving chronic insulin were given an intraperitoneal insulin injec(30-32). tion of 1 unit of insulin at two consecutive 30-min intervals. After 1 The present studyfocuses on the complex metabolic control h the micewere killed and liver nuclei were prepared and stored of transcription of the PEPCK promoter in transgenic mice frozen a t -70 "C for nuclear runoff assay. A bilateraladrenalectomy was performed in some animals 4-6 days during diabetes. We analyzed the transcriptionalresponse of segments of the PEPCKpromoter to diabetes, insulin replace- before killing. After surgery, mice received 1%(w/v) NaCl to drink instead of water. To determine the effect of glucocorticoids, adrenalment, adrenalectomy, administration of glucocorticoids, and ectomized animals were injected intraperitoneally with dexamethaalterations in dietary protein and carbohydrate. Our results sone (1.25 mg/kg) at 18 and 6 h before killing. For dietary studies, indicate that during diabetes PEPCK gene transcription is transgenic mice were fed either normal laboratory chow ad libitum or regulated by a complex interaction between insulin and glu- were fed a high carbohydrate diet containing 81.5% sucrose, 12.2% casein, 0.3% DL-methionine, 4% cottonseed oil, 2% brewers' yeast, cocorticoids, which is focused on the GRU of the PEPCK promoter. Glucocorticoids, and not CAMP, play a dominant and a 1%mineral mix with vitamins for 1 week. For high protein feeding, animals were fed a diet containing 64% casein, 22% nutritive role in determining the extent of hepatic PEPCK gene tran- fiber, 11%vegetable oil, 2% brewers' yeast, and 1%mineral mix plus scription during diabetes. vitamins (26). Anesthetized animals were killed by cervical dislocation, and the liver removed and frozen a t -70 "C for subsequent analysis. RNA Extraction and Analysis-Total RNA was extracted from rat Materials-ATP, CTP, GTP, yeast tRNA, proteinase K, and re- liver by the modified acid-phenol guanidine thiocyanate procedure striction enzymes were purchased from Boehringer Mannheim. DNA (36). RNA samples (10 pg) were prepared with 37% deionized formwas labeled using a random primer labeling kit obtained from Boeh- aldehyde, 6% formaldehyde, and diethylpyrocarbonate-treatedwater ringer Mannheim. [ ( u - ~ * P ] ~ C(3000 T P Ci/mmol), [a-32P]UTP(3000 to a volume of 88 ml. The samples were heated at 50 "C for 1h, cooled mCi/mmol), and GeneScreen Plus were purchased from Du Pont- for 5 min, and blotted onto nitrocellulose filters using a 96-well dot New England Nuclear. RNase-free DNase I (1000 units) was obtained blot apparatus (Bio-Rad). After hybridization, the filters were washed from Promega. All other reagentswere of the highest purity available. extensively at 55 "C, dried, and the image intensity was scanned by The cDNA probes for c-fos and bovine growth hormone (bGH) were Phosphor Imager analysis (Molecular Dynamics, CA). The abundance the generous gifts of Dr. Thomas Curran (Roche Institute of Molec- of PEPCK mRNA and bGH mRNA was integrated and expressed ular Biology, Nutley, NJ) and Dr. Fritz Rottman (Case Western relative to actin mRNA to control for small differences in RNA Reserve University), respectively. The synthetic diets used in this loading. study were purchased from Nutritional Biochemical Corp. (Cleveland, Determination of Gene Transcription in Vitro-To observe the OH), and their composition has been described previously (26). Re- acute effects of insulin on PEPCK gene transcription, isolated nuclei combinant insulin (Humulin R, 100 units/ml; Eli Lilly Co., Indian- were prepared from mouse liver as described previously (13). Nuclear apolis, IN)and dexamethasone (American Reagent Laboratories, run-off transcription was initiated with nuclei (approximately 4 X Shirley, NY) were obtained from commercial sources. Custom-syn- lo'), 100 p1 of 2 X reaction buffer (35% glycerol, 10 mM MgC12,0.2 M thesized, time-release insulin pellets for implantation studies inmice KC1, 12 mM ATP, 6 mM GTP, 6 mM CTP, 10 mM creatinine were obtained from Innovative Research of America (Toledo, OH) phosphate, 0.04 mg/ml creatinine kinase, and 100 units/ml RNase and contained between 2.4 and 9.6 units of bovine insulin. inhibitor) for 45 min a t 26 "C with 100 pCi of [a-"P]UTP. Nuclei cDNA Probes-The cDNA probe used for hybridization with bGH were digested following the procedure of Schibler et al. (37) as mRNA was a 814-kilobasepair cDNA containing the entire structural modified by Lowell et al. (38). Within each experiment, equal amounts gene for bGH (33). The PEPCK probe was a 1.1-kilobase PstI-PstI of labeled RNA (between 10 and 20 x lo6 cpm) was hybridized for 72 fragment from the 3' end of the PEPCK cDNA as described previ- h at 45 "C with slot blot filters bound with 5 pg of linearized cDNA ously (26). Actin mRNA was hybridized with the mouse a-actin cDNA plasmid. Filters were prepared with cDNA probes for bGH, 3' (34). PEPCKIO, c-{os, a-actin, PTZ plasmid (vector, no insert), and geAnimals-The transgenic mice used in this study have been de- nomic DNA (0.2 pg). After hybridization, filters were washed extenscribed previously (26, 30) and contain the bGH gene ligated to the sively at 65 'C and treated with RNase Afor 30 min at 37 'C. Specific PEPCK promoter: -2000 to +73, PEPCK(-2000)-bGH -460 to hybridization was quantified by phosphoimager analysis after 1-3 +73, PEPCK(-460)-bGH: and -355 to +73, PEPCK(-355)-bGH. days of exposure. The abundance of each mRNA transcript was Transgenic mice containing -460 to +73 with a block mutation in integrated, background radioactivity subtracted, and expressed rela-go/-82 and -250/-234, PEPCK(CRE-l/P3I)-bGH were also stud- tive to hybridization to genomic DNA. ied.' After initial screening and characterization of several lines of mice, one line of each mutation or deletion was chosen for in-depth RESULTS study based on the typical overall expression patterns of the respective transgene: -2000(19); -460(36); -355(111); and CRE-l/PBI(lS). The objective of these studies was to determine the horTransgenic mice were identified by dot blot analysis of DNA from a monal mechanisms involved in theregulation of PEPCK gene segment of the tail, using a 32P-labeledbGH cDNA probe (26). The animals were 2-4 months of age at the time of the study and were transcriptionduring diabetes. Streptozotocin diabetes has given free access to water and were fed standard laboratory chow been shown to increase PEPCK synthesis by removing the unless otherwise specified. All animals were killed between 1O:OO a.m. inhibitory effect of insulin on gene transcripton and by inand 12:Oo p.m. creasing the concentration of glucagon and glucocorticoids Dietary and Hormonal Treatments-Animals were made diabetic (12). The levels of mRNA for PEPCK (from the endogenous by a single intraperitoneal injection of streptozotocin (200 mg/kg). PEPCK gene) and bGH mRNA are shown in Fig. 1. The Mice were fasted overnight prior to injection and re-fed 2 h later. Diabetes was confirmed by measuring blood glucose (tail vein) using levels of bGH mRNA increased up to 5-fold in the livers of and glucose dipsticks and Glucometer 3 (Ames Products, Indianapolis, diabetic mice with the PEPCK(-2000)-bGH IN). All diabetic animals had blood glucose concentrations in excess PEPCK(-460)-bGH genes. This level of induction of gene EXPERIMENTAL PROCEDURES

* Y.Patel, J. Yun, and R. W. Hanson,

manuscript in preparation.

expression was very similar in all transgenic mice and in the livers of non-transgenic mice (data not shown). In mice with

Regulation of PEPCK Gene Transcription duringDiabetes Control

W

Diabetic + insulin HlghCHO diet

PEPCK

PEPCK(-ZOOO)-bGH PEPCK(-460)-bGH PEPCK(-355)-bGH

FIG. 1. Effect of diabetes chronic insulin replacement, or dietary carbohydrate on PEPCK and bGH mRNA in the livers of transgenic mice. Micewere made diabetic by streptozotocin injection and insulin replaced for 5 days using pellets implanted subcutaneously. A separate group of nondiabetic mice with the PEPCK(-355)-bGH transgene were fed a high carbohydrate (80%) diet for 1 week. Mice were killed, and the levels of PEPCK, bGH, and actin mRNA was determined by dot blot analysis (see “Experimental Procedures”).Control values were arbitrarily set toa value of 1,and theeffect of diabetes, insulin treatment, or high carbohydrate diet was expressed as -fold change compared to the control value. The control group means were: PEPCK = 0.97 f 0.32; PEPCK(-2000)-bGH = 1.51 f 0.72; PEPCK(-460)-bGH = 3.49 f 1.84; PEPCK(-355)-bGH = 5.58 ? 1.51. Results are the mean from 2-3 animals per treatment; symbols defined a t right.

the PEPCK(-355)-bGH gene, endogenous PEPCK mRNA inthe liver increased 4.5-fold over nondiabetic controls, whereas the level of bGH mRNA increased by only 2.5-fold. Chronic insulin replacement in diabetic mice reduced hyperglycemia and decreased PEPCK mRNA to control levels in the livers of all mice studied. However, hepatic bGH mRNA decreased about 50% in mice with the PEPCK(-2000)-bGH and PEPCK(-460)-bGH genes, and by 21% in the livers of diabetic mice containing the PEPCK(-355)-bGH transgene. The inability of insulin to fully reduce the concentration of bGH mRNA to control levelsmaybe due in part to the relatively long half-life of bGH mRNA (48 h) and possibly to the short length of insulin treatment. Feeding nondiabetic mice containing the PEPCK(-355)-bGH transgene a high carbohydrate diet for 1 week lowered the hepatic concentration of both PEPCK and bGH mRNA by 80% (Fig. 1). Insulin rapidly decreases PEPCK gene transcriptionin hepatoma cells, even in the presence of cAMP (14). To determine whether acute treatment with insulin also decreased PEPCK-bGH gene transcription in transgenic mice, nuclear runoff experiments were performed using nuclei isolated from the livers of normal, diabetic, or diabetic mice after acute insulin injection (Figs. 2 and 3). Inthe livers of diabetic mice, transcription of the PEPCK gene and the PEPCK(-2000)bGH and PEPCK(-460)-bGH transgenes was increased 3-4fold abovenondiabetic controls anddecreased to control levels 60 min after acute insulin injection (Fig. 3). In diabetic mice with the PEPCK(-355)-bGHtransgene however, PEPCK gene transcription in the liver was increased 4-fold above nondiabetic controls, whereas transcription of the PEPCKbGH transgene increased only 1.5-fold. Insulin reduced PEPCK gene transcription in the liver to control levels but

did not significantly affect PEPCK(-355)-bGH gene transcription. Diabetes increased c-fos genetranscription 2-%fold above nondiabetic controls, and acute insulin administration in diabetic animals did not significantly affect c-fosgene transcription (data not shown). Because diabetes failed to induce transcription of the PEPCK(-355)-bGH transgene, perhaps due to the deletion of the GRU, we hypothesized that glucocorticoids and the missing glucocorticoid regulatory element were essential for increased PEPCK gene transcription in diabetes. To determine the role of glucocorticoids in the induction of PEPCK gene expression by diabetes, we measured PEPCK mRNA in response to diabetes in the absence of adrenal steroid hormones. Adrenalectomy had no effect on basal levels of PEPCK mRNA, whereas diabetes alone increased PEPCK mRNA 4-fold (Fig. 4). However, in diabetic adrenalectomized mice there was no effect of diabetes on PEPCK mRNA; replacement with dexamethasone amplified PEPCK mRNA 3.5-fold, to levels similar to those noted in the livers of diabetic mice. These findings indicate that glucocorticoids and theglucocorticoid regulatory element are essentialfor the induction of PEPCK gene transcription during diabetes. However, previous studies have shown that whereas glucocorticoids alone increase PEPCK gene expression in liver cells, glucocorticoids in combination with cAMP has anadditive effect in increasing PEPCK gene transcription (8). To determine if control of PEPCK gene expression in diabetes requires the previously characterized cAMP regulatory elements within the PEPCK promoter(24), we studied transgenic animals with the PEPCK promoter (-460) containing block mutations in key regulatory elements necessary for induction of transcription by CAMP. In transgenic animals with mutations in CRE-1/ P3(I), diabetes increased bothPEPCKandPEPCK-bGH gene transcription in liver 3.1- and 5.5-fold abovebasal levels, respectively (Fig. 5). Insulin decreased PEPCK-bGH gene transcription to below control levels. These results suggest that mutation of downstream elements involved in cAMP control did not limit the induction of PEPCK gene expression by diabetes, and,furthermore, that negative regulation of PEPCK gene transcription by insulin was not mediated through sites previously identified as cAMP regulatory elements. We next examined the transcriptional response of the PEPCK promoter to diet in mice with the PEPCK(-355)bGH transgene. Mice were placed on a high carbohydrate (80%) diet for 1 week to decrease PEPCK or a high protein (65%)carbohydrate-free diet to increase PEPCK gene expression. A high carbohydrate diet decreased transcription of both the endogenous PEPCK gene and the PEPCK(-355)-bGH transgene in the liver by almost 90% compared to chow-fed controls. A diet high in proteinand free of carbohydrate induced the level of transcription of the PEPCKgene 6.2-fold and PEPCK/bGH gene 2.8-fold in comparison to chow-fed mice (Fig. 6). This level of gene transcription is similar to that observed in the nuclei from the livers of mice with the PEPCK(-460)-bGH transgene.’ Acute insulin injection of micefed a high protein diet decreased transcription of both the PEPCK and PEPCK(-355)-bGH genes to levels similar to those noted with chow-fed controls. Thus, in contrast to the loss of hormonal induction by diabetes observed in mice with the PEPCK(-355)-bGH transgene, dietary control over PEPCK gene expression was maintained, suggesting that there is an alternative site on the PEPCK promoter which is involved in dietary regulation of PEPCK gene transcription.

Regulation of PEPCK Gene Transcription during PEPCK(-2000)-bGH

-

--

Cuntrcll I)iat,ctic+l Diahelic rEr(.Y h(;H

c.FOS

rTz u.rctin

-

1,rnurnic n\ 4

a

. .

-

0

12955

PEPCK(-460)-bGH

-

r

Diabetes

s

-

L

u

I ,

Control PEPCK b(;H

= -

c-FOS

-

Diabetic

ag

PEPCK(-355)-bGH

,"'

PEPCK b(;H

CI

c.FOS

VTZ

FTZ

a-actin titnomir DNA

n.aclin I-

-- e

1)iabctic

Control

I)iubctic+l

"

(;rnomlc

DSA

.

I)iahclic*l

-

-

-

2.

0

0

FIG.2. Effect of diabetes and insulininjection on transcription of genes b y nuclei from liver of transgenic mice. Nuclei were isolated from control, diabetic, and diabetic mice 60 min after insulin injection. Nascent transcripts were labeled with [ "PIUTP and hybridized to cDNA sequences that were immobilized to nitrocellulose (see "Experimental Procedures"). Shown are autoradiographic signals from a representative experiment.

T

Control

T

ADX Diabetic Diabetlc-ADX Diabetic-ADX+ Dex

0

1

2

3

4

5

mRNA bva (Fold U u n g o )

FIG.4. Effect of glucocorticoida and diabetes on hepatic P E P C K mRNA. Mice were made diabetic by streptozotocin injection and a bilateral adrenalectomy (ADX) was performed in some animals 4-6 days before killing (see "Experimental Procedures"). To determine the effect of glucocorticoids, adrenalectomized-diabetic PEPCK PEPCKI-1000I-bCH PEPCK(-46Ol-bCH PEPCK(.35S)-bGH animals were injected intraperitoneally with dexamethasone (1.25 FIG.3. Effect of diabetesand ineulin injection on transcrip- mg/kg) a t 18 and 6 h before killing. Hepatic PEPCK mRNAwas P E P C K gene and PEPCK-bGH transgenes in nuclei normalized to the value for actin mRNA and expressed as the mean tion of the from mouse liver. Mice were made diabetic by streptozotocin injec- f standard error of the mean (2.80 k 0.47). The control group mean tion, and insulin was injected 60 min before killing (see "Experimental was arbitrarily set to a value of 1, and experimental groups were Procedures"). Nuclei were isolated and transcription measured as expressed as -fold change compared to the control animals. Data are described in the legend to Fig. 2. The resultant autoradiograms were for 3-5 animals per group. scanned and the values expressed as the mean f standard error of the mean for 3-4 animals from each group; symbols defined a t right. The control group means were: PEPCK = 3.23 f 1.47; mutated forms of the PEPCK promoter transiently transPEPCK(-2000) = 5.22 f 3.62; PEPCK(-460) = 4.74 f 2.65; fected into cells has failed to give reproducible and definitive PEPCK(-355) = 5.16 f 3.23. The control mean was set to a value of regulation by insulin,R despite normal regulation of the en1. and experimental groups were expressed as -fold change compared dogenous PEPCK gene. to the respective control animals. Using cells that were isolated by selection for expression of

the co-transfected neomycin resistance (neo) gene, Forest et al. (40) found that a PEPCK-CAT fusion gene from -437 to +69 of the PEPCKpromoter did not respond to insulin alone The interaction of glucocorticoids, CAMP,and insulin plays but gave a reproducible insulin-induced inhibition in the an important physiological role in modulating expression of presence of glucocorticoids. When shorter fragments of the the gene for PEPCK. In rat liver and in H4IIE hepatoma PEPCK promoter (between -402 and -271) were used, the cells, insulin inhibits basal PEPCK gene transcription and inhibition of transcription from the PEPCK promoter caused prevents the cAMP and/or glucocorticoid-induced transcrip- by insulin wasreducedfrom 90% to a mean of 57%; this tion (15). The negative effect of insulin on PEPCK gene suggests the deletion of an insulin response element, and transcription occurs rapidly (80% decrease in 30 min) (39), perhaps the presence of another element 3' from -271 in the does not require protein synthesis (15), and does not appear PEPCK promoter. O'Brien et al. (28) later confirmed the to involve an alteration in intracellular cAMP concentration presence of a 10-basepair sequence, located between positions (41). Although it is clear that insulin can strongly inhibit -416 and -407, relative to the transcription start site of the PEPCK gene transcription, efforts at defining the molecular PEPCK gene, which mediates an insulin response in H4IIE mechanism(s) responsible for insulin's regulation of PEPCK cells when inserted into a heterologous promoter, termed the gene have proven quite difficult. One reason for this difficulty insulin regulatory sequence (IRS). The glucocorticoid regurelates to the complex pattern involved in hormonal regula- latory region of the PEPCK promoter was identified by cotion of the PEPCK gene (i.e. multiple transcription factors transfection of various PEPCK-CAT plasmids with 5' delebinding at several sites on the promoter). A second difficulty tion end points together with a glucocorticoid receptor expresinvolves the failure of insulin to reproducibly affect transiently transfected DNA. For reasons that areunclear, analyA. Wynshaw-Boris, E. A. Park, and R. W.Hanmn. unpublished sis by more than one laboratory of fusion genes containing results. DISCUSSION

Regulation of PEPCK Gene 1’ranscription during Diabetes

12956

T Conlrol

5-

0

LC

Dlabetlc Dlabafc + insulln

E

4T

3-

2-

1-

n-

d .

PEPCK(CRE-l/P3)-bGH PEPCK

FIG.5. Effect of diabetes and insulin injection on transcription of PEPCK gene and PEPCK(CRE-l/P3)-bGH transgenes from mouse liver nuclei. Nuclei were isolated from mouse liver

and transcription measured as describedin legend to Fig. 2. The and the results of 3-4 animals resultant autoradiograms were scanned from each group quantitated; symbols defined at right. The control group was arbitrarily set to a value of 1, and experimental groups were expressed as -fold change compared to the control animals. The values are expressed as the mean +- standard error of the mean. The control group means were: PEPCK = 0.66 -+ 0.15; PEPCK(CRE-1/ P3)-bGH= 6.35 f 2.34.

B

Regular Diet

0 HCCHO IHCPAO

-

HCPRO+INSUUN

U

0

LC

PEPCK

FIG.6. Effect of dietary carbohydrate, protein, and insulin injection on transcription of the PEPCK gene and PEPCK(-366)-bGH transgenes in nuclei from mouse liver. Mice were placed on custom designed diets for 1 week as described under “ExperimentalProcedures.” Nuclei were isolated from the pooled liver of three mice of each group and transcription assay carried out as described in the legend of Fig. 2. Data are the means from two separate determinations.Thecontrolgroup mean was arbitrarily setto a value of 1,and experimental groups were expressed as -fold change comparedto the control animals. The control group means were: PEPCK = 3.46; PEPCK(-355)-bGH = 5.16; symbols defined at right.

sion vector into H41IE cells. DNase I footprint analysis revealed the presence of accessory binding factors and glucocorticoid receptor binding sites located well 3‘ to the 5’ boundary of this element (27), and this 110-base pair region was designated a glucocorticoid regulatory unit (GRU). The ability to introduce chimeric genes into the germline of transgenic animals provides a powerful tool for characterizing the in vivo function of hormone response elements within a promoter underphysiological conditions. Our results demonstrate that in transgenic mice with a deletion of the

glucocorticoid regulatory unit (GRU), no significant increase in PEPCK-bGH gene transcription was observed in diabetic mice. Likewise, adrenalectomy in non-transgenic mice prevented the increase in PEPCK mRNA normally observed in diabetic mice, whereas dexamethasone injection increased PEPCK gene expression 4-fold. This effect of glucocorticoids has been reported previously for the synthesis of PEPCK in the livers of rats (42). Our results clearly support the importance of the GRU (27) and glucocorticoids for positive induction of PEPCK gene transcription. Although previous studies have shown that glucocorticoids injected into animals do not substantially increase PEPCK gene transcription compared with glucagon and/or cAMP (13); this effect is probably due to thefact that the injection of glucocorticoid hormones will stimulate insulin release and preventthe induction of PEPCK gene expression (43). Transcription from the PEPCKpromoter lacking the GRU did not respond to insulin, when this hormone was administered to diabetic animals. This could be due to the lack of interaction of insulin or its agents with glucocorticoid binding factors, since mutations that disrupt the IRS concomitantly diminish the response to glucocorticoids (28,40). Insulin may thus work through a mechanism involving interference with glucocorticoid receptor binding or binding of accessory factors to the PEPCK promoter. These results provide additional evidence for interaction of insulin and glucocorticoids as a key regulatory mechanism of PEPCK gene transcription. Deletion of the GRU did not disrupt the transcriptional response of the PEPCK promoter to a diet high in carbohydrate, or to insulin, when it was administered to mice fed a high protein, carbohydrate-free diet. If the metabolic effect of feeding the mice a high carbohydratediet is mediated by insulin alone, then the IRS at -416 to -407 is clearly not required for the response of the PEPCK promoter to this hormone. It is likely, however, that theeffect of carbohydrate feeding to an intact animal involves multiple hormonal signals, mediated in part by transcriptional regulatory elements in the PEPCKpromoter that are more 3‘ to the deleted IRS. For example, carbohydrate feeding could reduce the concentration of cAMP or other intracellular mediators. With high protein feeding, nutritional and hormonal conditions are somewhat similar to diabetes; the plasma insulin to glucagon ratio decreases, and glucocorticoids increase (44). Since the increased transcription of the PEPCK(-355)-bGH gene in the livers of mice fed a high protein carbohydrate-free diet was less than the endogenous PEPCK gene, the response to a high protein diet could involve glucocorticoid regulation. However, since this induction of gene transcription was similar to that observed in nuclei from the livers of transgenic mice with PEPCK(-460)-bGH gene,’ factors other than glucocorticoid-mediated changes maybeinvolved. It is worth noting that DNA segments farupstreamin thePEPCK promoter (-4800 base pairs) have been found to contain enhancer elements that amplify transcription 10-20-fold relative to theminimal PEPCK promoter (45). The involvement of cAMP regulatory elements in insulin control of PEPCK gene transcription was studied in mice with mutations in two promoter regions that have been demonstrated to prevent acuteinduction of the PEPCKpromoter by acute BtgAMP or glucagon injection? In transgenic animals with mutations in both the CRE and the P3(I) regions of the PEPCKpromoter, diabetesincreased hepatic PEPCKbGH gene transcription 5.5-fold, whereas insulin treatment decreased transcription below the levels noted with nuclei isolated from the livers of control mice. These results generally agree with Forest et al. (40), who reported that in stably

Regulation of PEPCK Gene Transcription during Diabetes transfected HUIE cells, deletion of the CREfrom the PEPCK promoter reduced the cAMP induction of PEPCK-CAT activity from 6.2- to 2.6-fold, whereas the induction of transcription from the PEPCK promoter by cAMP was totally inhibited by insulin. These results suggest that mutation of downstream elements involved in cAMP control do not limit the induction of PEPCK gene expression by diabetes and, furthermore, that the effect of insulin is not mediated through the sites identified as cAMP regulatory elements. It is also possible that diabetes and insulin involve changes in transcription factors that do not bind to these distinct regions on the promoter, or that themechanism of insulin actioninvolves more global changes involving multiple interactions of regulatory proteins on the promoter. The outcome of these studies suggests glucocorticoids may be more important in the regulation of PEPCK gene expression than previously believed, especially with regard to the metabolic complications found in diabetes. PEPCK gene expression and activity is inappropriately elevated in animal models of NIDDM (11,12), aswell as insulin-deficient type I diabetic models as shown here. Although the molecular mechanism(s) responsible for increased hepatic glucose production in NIDDM are currently unknown, it is possible that failure of insulin to modulate glucocorticoid-dependent gene transcription may play a critical role in sustaining the increased rates of gluconeogenesis characteristic of NIDDM. Future studies willbe necessary to address the important role of glucocorticoids as a key gluconeogenic activator during diabetes. Acknowledgment-We thank FrankMularofor assistance in the execution of these studies.

expert technical

REFERENCES 1. Consoli, A,, and Jujhan, N. (1990) Ann. Med. 22, 191-195 2. DeFronzo, R., Simonson, D., and Ferrannini, E. (1982) Diabetobgia 2 3 , 313-319 3. Unger, R. H., Aguilar-Parada, E., Meuller, W. A., and Esentraut, A. M. (1970) J. Clin. Invest. 49,837-845 4. Baron, A. D., Schaeffer, L., Shragg, P., andKolterman, 0. G. (1987) Diabetes 36,274-283 5. Long, C. N. H., and Lukens, F.D. W. (1936) J. Exp. Med. 63,465-469 6. Friedmann, N., Exton, J. H., and Park,C. R. (1967) Biochem. Biophys. Res. Cornmun. 29,113-119 7. Exton, J. H. (1979) in Glucocorticoid Hormone Action (Baxter, J. D., and Rousseau, G. G., e&) pp. 517-521, Springer-Verlag, New York 8. Salavert, A., and Iynedjian, B. P. (1982) J. Biol. C k m . 267,13404-13412 9. Hanson, R.W., and Garber, A. J. (1972) Am. J. Clin. Nutr. 26,1010-1021 10. Chang, A. Y., and Schneider,D. I. (1970) Diabetologia 6,274-278 11. Veneziale, C. M., Donofrio, J. C., and Nishimura, H. (1983) J. Biol. Chem. 268,14257-14262

12957

12. Shafrir, E. (1988) in Frontiers in Diubetes Research (Shafrir, E., ed) pp. 304-315, John Libbey & Co., Baltimore, MD 13. Lamers, W. H., Hanson, R. W., and Meisner, H. (1982) Proc. Natl. Acad. Sci. U. S. A. 79,5137-5141 14. Granner, D.K., Andreone, T., Sasaki, K., and Beale, H. (1983) Nature 306,549-551 15. Sasaki, K., Cripe, T. P., Koch, S. R., Andreone, T. L., Petersen, D. D., Beale, E. G., and Granner,D. K. (1984) J. Biol. Chem. 269,15242-15251 16. Hod, Y., and Hanson, R. W. (1988) J. Bwl. Chem. 263,7747-7752 17. Nelson, K., Cimbala, M. A., and Hanson, R. W. (1980) J. Biol. Chem. 2 6 6 , 8509-8515 18. Roesler, W. J., Vandenbark, G. R., and Hanson,R. W. (1989) J. Biol. C k m . 264.9657-9664 ,---. 19. Sh%;J., Wynshaw-Boris, A., Short, H. P., and Hanson, R. W. (1986) J. Biol. Chem. 261,9721-9726 20. Quinn. P. G.. Wone. T. W.. Maenuson. M. A,. Shabb. J. B.. and Granner. D. K. (198i)) MorCell. B h . 8: 3467-3475 21. Magnuson, M. A,, Quinn, P. G., and Granner, D.K. (1987) J. Biol. Chem. 262,14917-14920 22. Park, E. A., Roesler, W. J., Liu, J., Klemm, D. J., Gurney, A. L., Thatcher, J. D., Shuman, J., Friedman, A., and Hanson, R. W. (1990) Mol. Cell. Biol. 10,6264-6272 23. Klemm, D. J., Roesler, W. J., Liu, J., Park, E.A., and Hanson,R. W. (1990) Mol. Cell. Biol. 10, 480-485 24. Liu, J., Park, E. A,, Gurney, A. L., Roesler, W. J., and Hanson, R. W. (1991) J. Biol. Chem. 266,19095;19102 25. Park, E. A., Gurney, A. L., Nizlelskl, S. E., Hakimi, P., Cao, Z., Moorman, A., and Hanson, R. W. (1993) J. Biol. Chem. 268,613-619 26. McGrane, M. M., DeVente, J., Yun, J., Bloom, J., Park, E., WynshawBoris, A., Wagner T., Rottman, F. M., and Hanson,R. W. (1988) J. Biol. Chem. 263,11443-11451 27. Imai, E., Stromstedt, P., Quinn, P. G., Carlstedt-Duke, J.,Gunstafsson, J., and Granner, D. K. (1990) Mol. Cell. Biol. 10,4712-4719 28. O'Brien, R. M., Lucas, P. C., Forest, C. D., Magnuson, M. A., and Granner, D. K (1990) Science 249,533-537 29. Giralt, M., Park, E. A,, Gurney, A. L., Liu, J., Hakimi, P., and Hanson, R. W. (1991) J. Biol. Chem. 266,21991-21996 30. McGrane, M. M., Yun, J., Moorman, A. F. M., Lamers, W. H., Hendrick, G. K., Arafah, B. M., Park, E. A., Wagner, T. E., and Hanson, R. W. (1990) J. Biol. Chem. 266,22371-22379 31. Short, M. K., Clouthier, D. E., Schaefer, I. M., Hammer, R. E., Magnuson, M. A., and Beale, E. G. (1992) Mol. Cell. Biol. 1 2 , 1007-1020 32. Eisenberger, C. L., Nechushtan, H., Cohen, H., Shani, M., and Reshef, L. (1991) Mol. Cell. Biol. 12,1396-1403 33. Helms, S. R., and Rottman, F. M. (1990) Nucleic Acids Res. 18,255-259 34. Giebelhaus, D. H., Weitlauf, H. M., and Schultz, G. A. (1985) Deu. Biol. 1 0 7 , 407-413 35. Kumara-Siri, M. H., and Surks, M. I. (1985) J . Biol. Chem. 2 6 0 , 1452914537 36. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162,156-159 37. Schibler W. U. Hagenbuchle, O., Wellauer, P. K., and Pittet, A. C. (1983) Cell 33,501-'508 38. Lowell,C.A., Stearman, R. S., and Morrow, J. F. (1986) J. Biol. Chem. 261,8453-8461 39. Sasaki, K., and Granner, D. K. (1988) Proc. Natl. Acad. Sci. U. S. A. 86, ""

I

I

,

~

29.54-29.58 _ ~ -.__ _ .

40. Forest, C. D., OBrien, R. M., Lucas, P. C., Magnuson, M. A., and Granner, D. K (1990) Mol. Endocrinol. 4,1302-1310 41. Beale, E. G., Koch, S. R., Brotherton, A. F. A,, Sheorain,V. S., and Granner, D. K. (1986) Diabetes 36,546-549 42. Reshef, L., Hanson, R. W., and Ballard, F. J. (1970) J. Biol. Chem. 2 4 6 ,

_""

5974-59RA -_.1

43. Gunn, J. M., Hanson, R. W., Meyuhas, O., Reshef, L., and Ballard, F. J. (1975) Biochem. J. 160,195-203 44. Peret, J., Foustock, S., Chanez, M., Bois-Joyeux, B., and Assan, R. (1981) J. Nutr. 111,1173-1184 45. IP, Y. T., Poon, D., Stone, D., Granner, D. K., and Chalkley, R. (1990) Mol. Cell. Biol. 10,3770-3781