Chronic Exposure to High Glucose Concentrations Increases ...

0013-7227/99/$03.00/0 Endocrinology Copyright © 1999 by The Endocrine Society

Vol. 140, No. 10 Printed in U.S.A.

Chronic Exposure to High Glucose Concentrations Increases Proglucagon Messenger Ribonucleic Acid Levels and Glucagon Release from InR1G9 Cells* ERIC DUMONTEIL, BEATE RITZ-LASER, CHISTOPHE MAGNAN, ´ ANA GRIGORESCU, ALAIN KTORZA, AND JACQUES PHILIPPE ILE Diabetes Unit (E.D., B.R.-L., I.G., J.P.), Centre Me´dical Universtaire, 1211 Gene`ve 4, Switzerland CH-1211; Laboratoire de Physiopathologie de la Nutrition (C.M., A.K.), Groupe Endocrinologie me´tabolique, Universite´ Paris 7, Tour 23–33, F-75251 Paris Cedex 05, France ABSTRACT a cell function is impaired in diabetes. In diabetics, plasma levels of glucagon are high despite persistently elevated glucose levels and may even rise paradoxically in response to a glucose load; high plasma glucagon levels are accompanied by increased proglucagon gene expression. We have investigated the effects of high glucose concentrations on InR1G9 cells, a glucagon-producing cell line. We show here that chronically elevated glucose concentrations increase glucagon release by 2.5- to 4-fold, glucagon cell content by 2.5- to 3-fold, and

D

IABETES results from an inadequate response of the b cells of the endocrine pancreas to match insulin release to the prevailing glucose concentrations, resulting in hyperglycemia. It is now well established that sustained hyperglycemia leads to the so-called glucose toxicity. In support of the glucose toxicity concept, chronic exposure of primary islets or insulinoma cells to elevated glucose concentrations impairs glucose-responsive insulin secretion and biosynthesis (1, 2). The mechanisms by which chronic hyperglycemia reduces insulin production have been recently analyzed; high glucose leads to a paradoxical decrease in insulin messenger RNA (mRNA) levels through an inhibition of insulin gene transcription (3–9). The b cell is not the only cell type of the islet that does not function properly in diabetes. a cell function has been reported be abnormal in all forms of diabetes. The normal reciprocal relationship between glucose and glucagon concentrations is thus lost. In contrast to nondiabetics, in whom hyperglycemia suppresses and hypoglycemia stimulates glucagon secretion, in diabetics, plasma levels of glucagon are constantly high despite varying degrees of hyperglycemia; in fact, glucagon levels may rise paradoxically during hyperglycemia produced by a carbohydrate meal (10). In addition to these in vivo observations, it has been shown that prolonged culture of isolated mouse or rat islets at high Received January 26, 1999. Address all correspondence and requests for reprints to: Jacques Philippe, M.D., Diabetes Unit, Centre Me´dical Universitaire, 1211 Gene`ve 4, Switzerland Ch-1211. E-mail: [email protected]. * This work was supported by the Swiss National Science Foundation (grant 3200 – 046816), by the Institute for Human Genetics and Biochemistry and by the Berger Foundation.

proglucagon messenger RNA levels by 4- to 8-fold, whereas changes for 24 h have no effect on proglucagon messenger RNA levels. Persistently elevated glucose affects proglucagon gene expression at the level of transcription and insulin is capable of preventing this effect. We conclude that chronically elevated glucose may be an important factor in the a cell dysfunction that occurs in diabetes and thus that glucose may not only affect the b cell but also the a cell. (Endocrinology 140: 4644 – 4650, 1999)

extracellular glucose concentrations results in paradoxically increased glucagon secretion (11, 12). The loss of a-cell response to changes in glucose concentrations appears to be selective inasmuch as glucagon secretion in diabetes is normally suppressed by increased FFA and is hyperresponsive to stimulation by arginine (13). While the existence of a-cell dysfunction in diabetes is generally accepted, the explanation for the apparent a-cell insensitivity or even paradoxical response to glucose remains unclear. It has been suggested that a-cell dysfunction is an abnormality separate from the b-cell disorder or alternatively that it is the passive consequence of insulin deficiency (13). Accumulating evidence suggests that neither view is strictly correct. Indeed, while insulin treatment leads to reductions of plasma glucagon levels, it may not restore the normal a-cell ability to secrete glucagon appropriately in response to changes in glucose concentration. The concept of glucose toxicity could thus apply to the observed a cell dysfunction. We recently showed that, in vivo, insulinopenic diabetes is characterized by both elevated plasma glucagon and proglucagon mRNA levels; these levels rise despite marked hyperglycemia suggesting that the a cell becomes progressively insensitive to glucose (14). We have investigated the possibility that chronically elevated glucose levels might affect the function of glucagonproducing cells. We show here that exposure of InR1G9 cells to high glucose leads to an increase in glucagon release, glucagon cell content and proglucagon mRNA levels; this increase does not occur within 24 h but, over a period of days, is reversible, and prevented by insulin. The mechanism by which high glucose increases proglucagon gene expression is through the stimulation of transcription and the first 138

4644

PROGLUCAGON GENE EXPRESSION

bp of the promoter are necessary for this effect. We conclude that hyperglycemia may not only affect the b cells by decreasing insulin bioynthesis but also the a-cells by stimulating glucagon production. Materials and Methods Cell culture and transfection studies InR1G9 cells are glucagon-producing cells and grow in monolayers (15, 16). As most islet cell lines InR1G9 cells are usually cultured in 11 mm glucose; this glucose concentration may not be optimal for a cell function, as previously demonstrated for b cells (3). We cultured InR1G9 cells in RPMI 1640 at a glucose concentration of either 5 or 11 mm, supplemented with 5% FCS and 5% new-born calf serum, 100 U/ml of penicillin, and 100 mg/ml of streptomycin. Cells were passaged once weekly after trypsinization and fed every 48 h with medium except that new medium without serum was added 24 h before glucagon measurements. For counting, cells were plated at a density of 105 cells in 12-well plates; cells were detached with trypsin-EDTA after 7 days and counted in a hematocytometer chamber for four successive periods. The number of nonviable cells was substracted from the total number of cells. Cell viability was assessed by trypan blue exclusion. Cells were transfected in suspension by the DEAE dextran method (17) with 3 mg of a CAT reporter plasmid and 1 mg of the plasmid pSV2A pap to monitor transfection efficiency. PSV2Apap contains the human placental alkaline phosphatase gene driven by the simian virus 40 long terminal repeat (18). 59-deleted mutants of the rat glucagon gene upstream sequence were subcloned into poCAT as previously described (17). Transfections were performed between weeks 13 and 20. Insulin was added to the culture medium (in the absence of serum) at a concentration of 1028 m in specific experiments. Of note, the insulin concentrations in the media (from FCS and newborn calf serum) averaged 2.9 to 5.5 mU/ml. Cell extract were prepared 48 h after transfection and analyzed for CAT and alkaline phosphatase activity as described (18 –19). Protein concentrations were determined with a Bio-Rad Laboratories, Inc. protein assay kit.

Northern blot analysis Cells from a 10 cm diameter culture plate per experimental condition (; 107 cells) were lysed in guanidine thiocyanate and RNA was extracted through a cesium chloride gradient as previously described (16). Twenty micrograms of total RNA, representing each experimental condition were analyzed by Northern blot (16) using [32P]-labeled rat proglucagon and mouse b-actin complementary DNA probes; quantification of the signals was done by scanning of the membranes using a Phosphor imager (Molecular Dynamics, Inc.).

RIAs Glucagon was determined in both culture media and acid-ethanol extracted cells by radioimmunoassays (Bio data, Rome, Italy). The protease inhibitor aprotinin was added at 250 KIU/ml of fresh culture media 24 h before measurement of immunoreactive glucagon. For glucagon cell content determination, InR1G9 cells were washed with PBS buffer and immunoreactive glucagon was measured after acid/ethanol (1.5% HCL, 75% ethanol) extraction (20). The sensitivity of the assay was 14.5 ng/liter for glucagon and the coefficient of variation within and between assays was 6%.

Presentation of the results and data analysis Results are presented as mean 6 sem. Statistical analysis of differences between groups were carried out using ANOVA and the paired, two-tailed Student’s t tests.

Results High glucose exposure increases glucagon release, glucagon cell content and proglucagon mRNA levels

To investigate whether changes in glucose concentrations could affect glucagon release, glucagon cell content and pro-

4645

glucagon mRNA levels, we incubated InR1G9 cells in either 5 or 11 mm glucose for 5 weeks; at the end of each period of 1 week, glucagon release (for the last 24 h), glucagon cell content and proglucagon mRNA levels were measured. As shown in Fig. 1, glucagon released in the medium during 24 h was 2- to 3-fold higher at 11 mm compared with 5 mm; similar results were obtained for glucagon cell content and proglucagon mRNA levels, although the difference in proglucagon mRNA levels between 5 and 11 mm was higher, in the range of 4- to 8-fold, compared with both release and cell content. To study whether the changes observed in proglucagon mRNA levels between 5 and 11 mm glucose were stable over a more prolonged period and reversible, we incubated InR1G9 cells in 5 and 11 mm glucose for 12 weeks (Fig. 2). We also tested for the reversibility of these changes by switching at weeks 10 and 11 subsets of the cells incubated at 5 mm to 11 mm glucose. Proglucagon mRNA levels were measured weekly as before. As shown in Fig. 2, a 3- to 4-fold higher level of proglucagon mRNA was still observed after 12 weeks of culture in 11 mm glucose. In addition, changing glucose from 5 to 11 mm at weeks 10 and 11 resulted in a return to high mRNA levels values at weeks 11 and 12; reversibility of the increase was also verified by switching cells from 11 mm to 5 mm glucose. We conclude from these data that chronically high glucose levels result in a reversible increase in glucagon release, glucagon cell content and proglucagon mRNA levels over prolonged periods in InR1G9 cells. To ascertain that cell number and viability were not affected by the different glucose concentrations used, viable cells were determined by trypan blue exclusion and cells were counted at the end of each week for a 4-week period (Fig. 3). We did not observe any difference in cell number whether the cells were cultured at 5 or 11 mm glucose; furthermore, cell viability was similar at both glucose concentrations, between 90 and 95%. Chronic changes in glucose concentrations affect glucagon release, glucagon cell content and proglucagon mRNA levels

To determine whether glucagon release, glucagon cell content and proglucagon mRNA levels are affected by less prolonged changes in glucose concentrations, we switched InR1G9 cells incubated in 5 mm glucose to either 2 or 11 mm and cells incubated in 11 mm to either 2 or 5 mm for 24 h and measured glucagon release, glucagon cell content and proglucagon mRNA levels. As shown in Fig. 4, changes from 5 to either 2 or 11 mm or from 11 to either 2 or 5 mm glucose did not significantly affect either glucagon release, glucagon cell content or proglucagon mRNA levels. Of note similar results were obtained from isolated primary rat islets incubated under the same conditions (Dumonteil, E., manuscript in preparation). To investigate the time course of glucose effects on proglucagon mRNA levels, we switched InR1G9 cells chronically incubated at 11 mm glucose to 5 mm and measured proglucagon mRNA levels daily. Significant decreases in mRNA levels occurred between day 4 and 5; stable values were then observed from day 6 (Fig. 5A). Similarly, cells chronically incubated at 5 mm glucose were switched to 11 mm and proglucagon mRNA levels were assessed daily. As shown in Fig. 5B, mRNA levels started to increase after 4 days

4646


Endo • 1999 Vol 140 • No 10

FIG. 1. Effect of chronically high glucose concentrations on glucagon release, glucagon cell content and proglucagon mRNA levels. InR1G9 cells were cultured in RPMI-1640 medium containing either 5 or 11 mM glucose for 5 weeks. Glucagon release (A), glucagon cell content (B) and proglucagon mRNA levels (C and D) were measured at the end of each weekly period. Glucagon release and content are expressed in ng/5009000 cells, whereas proglucagon mRNA levels were quantified by Northern blot (C), corrected for the respective b-actin signals and are expressed as percentage of the value obtained at week 1 at 11 mM glucose (D). Values are derived from three independent experiments; * and “ indicate significant differences (P , 0.01 and P , 0.05, respectively)

to reach 95% of the values observed in cells chronically incubated in 11 mm glucose. These data indicate that in InR1G9 cells glucose does not affect proglucagon mRNA levels within 24 h; it takes indeed several days and up to a week to reach a new steady state. Insulin prevents the increase in proglucagon mRNA levels induced by high glucose

We and others previously reported that insulin was capable to decrease glucagon biosynthesis through a decrease in proglucagon gene transcription (14, 16, 19, 21, 22). To study whether insulin might affect the increase in proglu-

cagon gene expression induced by chronically elevated glucose concentrations, we switched InR1G9 cells incubated in 5 mm glucose to 11 mm in the presence or absence of 1028 m insulin for 1 week; proglucagon mRNA levels were measured at the end of the experimental period. Figure 6 shows that insulin was able to prevent the increase in proglucagon mRNA levels induced by high glucose. High glucose affects proglucagon gene expression at the transcription level

To determine the mechanisms by which chronically high glucose increases proglucagon mRNA levels, we transfected


4647

FIG. 3. Different glucose concentrations do not affect cell number. InR1G9 cells were cultured in RPMI-1640 medium containing either 5 or 11 mM glucose and plated at a density of 105 cells; they were counted after 7 days for four successive periods. Values are derived from four different experiments.

proglucagon gene expression at the level of transcription and that the effect is mediated within 138 bp of the proglucagon gene promoter. Discussion

FIG. 2. Reversibility of the glucose effects on proglucagon mRNA levels. InR1G9 cells were cultured in RPMI-1640 medium containing either 5 or 11 mM glucose for 12 weeks. At weeks 10 and 11, aliquots of InR1G9 cells cultured at 5 mM glucose were switched to 11 mM and cells cultured at 11 mM were switched to 5 mM. Proglucagon mRNA levels were measured at the end of weeks 1, 10, 11, and 12 by Northern blot as in Fig. 1. A, A representative Northern blot; B, proglucagon mRNA levels are corrected for the respective b-actin signals and are expressed as percentage of the values obtained at week 1 at 11 mM glucose. Values are derived from three independent experiments; * indicate significant differences (P , 0.01).

292 base pairs (bp) of the rat proglucagon gene 59-flanking sequence linked to the CAT reporter gene (2292 CAT) into InR1G9 cells chronically cultured in either 5 or 11 mm glucose. Transcriptional activity of 2292 CAT was found to be 4-fold higher in cells incubated chronically at 11 compared with 5 mm glucose (Fig. 7). In contrast, changes in glucose concentrations from 5 to either 2 or 11 mm or from 11 to either 2 or 5 mm 24 h after transfection did not affect CAT activity. To localize further the DNA sequences, which mediate the glucose effect on proglucagon gene expression, we assessed transcriptional activity of shorter fragments of the rat proglucagon gene (2213 CAT, 2168 CAT, and 2138 CAT) into InR1G9 cells cultured in either 5 or 11 mm glucose. As shown on Fig. 8, the first 138 bp of the promoter are still capable of mediating the glucose effect. Our data thus indicate that chronically high glucose affects

We present evidence in our studies that chronically elevated glucose concentrations increase glucagon release, glucagon cell content, and proglucagon gene expression in InR1G9 cells. The effects of glucose are rather slow taking more than 1 day to be fully manifest. The changes induced by glucose are fully reversible inasmuch as switching from 5 to 11 mm glucose or from 11 to 5 mm glucose for 6 to 7 days results in a 3- to 5-fold increase or decrease, respectively, in glucagon release, glucagon cell content and proglucagon mRNA levels. The mechanisms by which chronically high glucose affects glucagon gene expression are to be fully defined but glucose effects are mediated through an increase in gene transcription and DNA sequences within 138 bp of the proglucagon gene promoter are necessary for this increase. The lag time between the change in glucose concentration and the observed consequences on proglucagon gene expression suggests that new protein synthesis is required to affect gene transcription; we hypothesize that high glucose may either decrease the abundance or function of a transcription factor that inhibits glucagon gene expression or inversely increase or induce the synthesis of a positively acting factor. Precise definition of the DNA sequences mediating the glucose effect should help resolving these possibilities. An important aspect of these studies is that insulin is capable of preventing the effects of a prolonged exposure to high glucose concentrations. Because InR1G9 cells were incubated in media containing 10% serum, thus in the presence

4648


Endo • 1999 Vol 140 • No 10

FIG. 5. Time-course of the changes in proglucagon mRNA levels induced by glucose. InR1G9 cells were incubated in RPMI-1640 medium containing either 11 mM (A) or 5 mM glucose (B). After 12 weeks, cells were then switched to 5 mM (A) or 11 mM glucose (B), respectively, and proglucagon mRNA levels were measured every day for 7 days and quantified by Northern blot as in Fig. 1. Values are derived from three different experiments and are expressed as the percentage of the values obtained at day 0 at 11 mM glucose. * and “ indicate significant differences (P , 0.01 and P , 0.05, repectively).

FIG. 4. Changes in glucose concentrations during 24 h do not affect glucagon release, glucagon cell content, or proglucagon mRNA levels. InR1G9 cells were cultured in RPMI-1640 medium containing either 5 or 11 mM glucose; cells incubated with 5 mM glucose were then switched to medium containing either 2 or 11 mM, whereas cells incubated in 11 mM glucose were switched to either 2 or 5 mM glucose for 24 h. At the end of the incubation period, glucagon was measured both in the medium (A) or cells (B) and proglucagon mRNA levels by

of 0.3 to 0.5 mU/ml insulin, it is possible that the chronic glucose effects on glucagon release, glucagon cell content and proglucagon gene expression might have been underestimated due to the potential inhibitory effects of the low concentrations of insulin. Insulin has been reported to inhibit proglucagon gene expression both in vivo and in vitro (14, 16, 19, 21–23); insulin is able to acutely decrease proglucagon gene transcription through the enhancer element G3, which Northern blot (C). Values are derived from three different experiments and are expressed as the percentage of the values obtained at 11 mM glucose. * and “ indicate significant differences (P , 0.01 and P , 0.05, respectively)


FIG. 6. Insulin prevents the glucose-induced increase in proglucagon mRNA levels. InR1G9 cells were cultured in RPMI-1640 medium containing either 5 or 11 mM glucose. Cells incubated with 5 mM glucose were switched to 11 mM with (1i) or without 1028 M insulin for 7 days. At the end of the incubation period, total RNA was analyzed by Northern blot. Values are derived from three different experiments and are expressed as the percentage of the values obtained at 11 mM glucose. “ indicate significant difference (P , 0.05).

is localized between 2290 to 2260 bp relative to the transcriptional start site (19 –21); however, as the glucoseinduced changes on proglucagon gene transcription are mediated by the proximal 138 bp of the promoter, the mechanisms by which insulin prevents the glucose effects will thus need to be further analyzed. Prevention of the glucose effect is likely to be indirect; however, in view of the lag time between the changes in glucose concentrations and increased gene transcription, one possibility would be that insulin counteracts the effects of glucose on the synthesis of a specific protein involved in the regulation of proglucagon gene expression. In type 2 diabetes, b cell function deteriorates progressively over time leading to more pronounced hyperglycemia (1–2). The concept of glucose toxicity implicates that a sustained elevation in glucose levels exerts by itself deleterious effects on pancreatic b cells. Impaired b cell function in the presence of high glucose has been demonstrated in several experimental systems (1–9); proinsulin gene expression as well as insulin biosynthesis and secretion are impaired. The effect of glucose on proinsulin mRNA levels has been proposed to decrease the activity of positively acting transcription factors such as IDX-1, IEF-1, and the RIPE3b-binding proteins interacting with the gene promoter and possibly up-regulating transcriptional repressors of insulin gene transcription, such as C/EBP b (6 –9). a cell function is also impaired in diabetes; there is both a loss of glycemic control of glucagon secretion and hyperresponsiveness of glucagon secretion to stimuli such as amino acids (13). The respective roles of insulin and glucose in these abnormalies have been abundantly debated; it appears likely, however, that they are both influential in the control

4649

FIG. 7. Chronically elevated glucose regulates proglucagon gene expression through an increase in transcription. InR1G9 cells were cultured in RPMI-1640 medium containing either 5 or 11 mM glucose. Cells were transfected with 3 mg of 2292 CAT containing 292 bp of the rat proglucagon gene promoter linked to the CAT reporter gene. Cells incubated in 5 mM glucose were then switched to either 2 or 11 mM glucose and cells incubated in 11 mM glucose were switched to either 2 or 5 mM for 24 h. Cells were then harvested and CAT activity was measured. Values are derived from three different experiments and are expressed as the percentage of the values obtained at 11 mM glucose. * indicate significant difference (P , 0.01 and P , 0.05, respectively).

of a cell function. Insulin has been clearly shown to affect both glucagon release and proglucagon gene expression, whereas glucose is the major regulator of glucagon secretion. In diabetes, plasma glucagon and proglucagon mRNA levels are chronically elevated despite hyperglycemia (14); insulin treatment is capable of correcting the elevated glucagon levels to normal (13). Elevated glucose by itself may impair a cell function. Prolonged culture of isolated mouse and rat islets at high glucose concentrations results indeed in paradoxically increased glucagon secretion (11, 12); these data are reminiscent of the paradoxical rise in glucagon levels during hyperglycemia produced by a carbohydrate meals in human diabetics (10). It has also been proposed that a cells have a glucose-sensing system that is reversibly attenuated by hyperglycemia (24) and that glucagon responsiveness to hypoglycemia may be improved by insulin-independent correction of hyperglycemia (25). Our results extend these finding by suggesting that sustained elevations of glucose levels may increase glucagon release and proglucagon gene expression and thus participate in abnormal a cell function observed in diabetes. There is no reason to believe that chronically elevated glucose affects only b cells; a cells display indeed similar glucosesensing mechanisms as b cells, such as a glucose transporter, GLUT-1, and the rate-limiting enzyme in glucose metabolism, glucokinase (26, 27). Glucose is the major regulator of both glucagon and insulin secretion, whereas acutely raised glucose levels stimulate and inhibit insulin and glucagon release, respectively, sustained elevations of glucose inhibit

4650


5. 6.

7.

8. 9.

10. 11.

FIG. 8. The 138 first base pairs of the proglucagon gene promoter are sufficient to mediate the effects of chronically elevated glucose. InR1G9 cells were cultured in RPMI-1640 containing either 5 or 11 mM glucose for 10 weeks. Cells were transfected with 3 mg 2292 CAT, 2200 CAT, 2168 CAT and 2138 CAT containing 292 bp, 200 bp, 168 bp, and 138 bp, respectively, of the rat proglucagon gene promoter linked to the CAT gene. CAT activities are derived from three different experiments and are expressed as the percentage of the values obtained at 11 mM glucose. * and “ indicate significant difference (P , 0.01 and P , 0.05, respectively).

insulin release and biosynthesis as well as proinsulin gene expression; the effects of sustained elevations of glucose on glucagon release and proglucagon gene expression would be expected to be the opposite. Our recently published in vivo data as well as the present data suggest indeed that chronically high glucose concentrations stimulate glucagon release and proglucagon gene expression (14). It must be stressed, however, that these results have been obtained from experiments done with a cultured cell line producing glucagon; the response of these cells to glucose may be different compared with those of primary a cells. There is, however, no satisfactory primary a cell system at present to validate our results, but more detailed knowledge of the glucose sensing mechanisms of the a cell should allow a better understanding of the physiological and pathophysiological role of glucose in a cell function.

12. 13. 14.

15. 16. 17. 18. 19. 20.

21. 22.

23.

Acknowledgments We thank C. Robert-Tissot for typing the manuscript and I. Constant and N. Klages for expert technical assistance.

References 1. Ja¨rvinen HY 1992 Glucose toxicity. Endocr Rev 13:415– 431 2. Leahy JL, Bonner-Weir S, Weir GC 1992 b-cell dysfunction induced by chronic hyperglycemia. Diabetes Care 15:442– 451 3. Robertson RP, Zhang HJ, Pyzdrowski KL, Walseth TF 1992 Preservation of insulin mRNA levels and insulin secretion in HIT cells by avoidance of chronic exposure to high glucose concentrations. J Clin Invest 90:320 –325 4. Olson LK, Redmon JB, Towle HC, Robertson RP 1993 Chronic exposure of

24. 25. 26.

27.

Endo • 1999 Vol 140 • No 10

HIT cells to high glucose concentrations paradoxically decreases insulin gene transcription and alters binding of insulin gene regulatory protein. J Clin Invest 92:514 –519 Eizirik DL, Korbutt GS, Hellerstro¨m C 1992 Prolonged exposure of human pancreatic islet to high glucose concentrations in vitro impairs the b-cell function. J Clin Invest 90:1263–1268 Sharma A, Olson LK, Robertson RP, Stein R 1995 The reduction of insulin gene transcription in HIT-T15 b cells chronically exposed to high glucose concentration is associated with the loss of RIPE3b1 and STF-1 transcription factor expression. Mol Endocrinol 9:1127–1134 Olson LK, Sharma A, Peshavaria M, Wright CVE, Towle HC, Robertson RP, Stein R 1995 Reduction of insulin gene transcription in HIT-T15 b cells chronically exposed to a supraphysiologic glucose concentration is associated with loss of STF-1 transcription factor expression. Proc Natl Acad Sci USA 92:9127–9131 Poitout V, Olson LK, Robertson RP 1996 Chronic exposure of bTC-6 cells to supraphysiologic concentrations of glucose decreases binding of the RIPE3b1 insulin gene transcription activator. J Clin Invest 97:1041–1046 Seufert J, Weir GC, Habener JF 1998 Differential expression of the insulin gene transcriptional repressor CCAAT/enhancer-binding protein b and transactivator islet duodenum homeobox-1 in rat pancreatic b cells during the development of diabetes mellitus. J Clin Invest 101:2528 –2539 Buchanan KD, McCarroll AM 1972 Abnomalies of glucagon metabolism in untreated diabetes mellitus. Lancet 2:1394 –1395 Segerstro¨m K, Anderson A, Lundquist G, Peterson B, Hellerstro¨m C 1976 Regulation of the glucagon release from mouse pancreatic islets maintained in tissue culture at widely different glucose concentrations. Diabetes 2:45– 48 Ziegler M, Hahn HJ, Ziegler B, Kohler E, Fiedler H 1975 Paradoxical glucagon response after stimulation with glucose and arginine in isolated pancreatic and rat islets. Diabetologia 11:63– 69 Unger RH, Orci L 1983 Glucagon II. In: Lefe`bvre PJ (ed) Handb Exp Pharm 66/II. Springer-Verlag, Berlin, Heidelberg, pp 431– 450 Dumonteil E, Magnan C, Ritz-Laser B, Meda P, Dussoix P, Gilbert M, Ktorza A, Philippe J 1998 Insulin, but not glucose lowering corrects the hyperglucagonemia and increased proglucagon mRNA levels observed in insulinopenic diabetes. Endocrinology 139:4540 – 4546 Takaki R, Ono J, Nakamura M, Yokogawa Y, Kumae S, Hiraoka T, Yamaguchi K, Uchida S 1986 Isolation of glucagon-secreting cell lines by cloning insulinoma cells. In Vitro Cell Dev Biol 22:120 –126 Philippe J 1989 Glucagon gene transcription is negatively regulated by insulin in a hamster islet cell line. J Clin Invest 84:672– 677 Philippe J, Drucker DJ, Knepel W, Japeal L, Habener JF 1988 a-cell-specific expression of the glucagon gene is conferred to the glucagon promoter element by the interactions of DNA-binding proteins. Mol Cell Biol 8:4877– 4888 Henthorn P, Zervos P, Raducha M, Harris H, Kadesch T 1988 Expression of a human placental alkaline phosphatase gene in transfected cells: use as a reporter for studies of gene expression. Proc Natl Acad Sci USA 85:6342– 6346 Philippe J 1991 Insulin regulation of the glucagon gene is mediated by an insulin-responsive DNA element. Proc Natl Acad Sci USA 88:7224 –7227 Bernard C, Thibault C, Berthault MF, Magnan C, Saulnier C, Portha B, Pralong W, Penicaud L, Ktorza A 1998 Pancreatic B cell regeneration after 48h glucose infusion in mildly diabetic rats is not correlated with functional improvement. Diabetes 47:1058 –1065 Philippe J, Morel C, Cordier-Bussat M 1995 Islet-specific proteins interact with the insulin-response element of the glucagon gene. J Biol Chem 270:3039 –3045 Chen L, Komiya I, Inman L, McCorkle K, Alam T, Unger RH 1989 Molecular and cellular responses of islets during perturbations of glucose homeostasis determined by in situ hybridization histochemistry. Proc Natl Acad USA 86:1367–1371 Chen L, Komiya I, Inman L, O’Neil J, Appel M, Alam T, Unger RH 1989 Effect of hypoglycemia and prolonged fasting on insulin and glucagon gene expression. J Clin Invest 84:711–714 Gerich JE 1983 Glucose in the control of glucagon secretion. In: Lefe`bvre PJ (ed) Handb Exp Pharm 66/II. Springer-Verlag, Berlin, Heidelberg, pp 3–18 Shi ZQ, Rastogi KS, Lekas M, Efendis S, Drucker DJ, Vranic M 1996 Glucagon response to hypoglycemia is improved by insulin-independent restoration of normoglycemia in diabetic rats. Endocrinology 137:3193–3199 Heimberg H, De Vost A, Pipeleers D, Thorens B, Schuit F 1995 Differences in glucose transporter gene expression between rat pancreatic a- and b-cells are correlated to differences in glucose transport but not in glucose utilization. J Biol Chem 270:8971– 8975 Heimberg H, De Vos A, Moens K, Quartier E, Bouwens L, Pipeleers D, Van Schaftingen E, Madsen O, Schuit F 1996 The glucose sensor protein glucokinase is expressed in glucagon-producing a-cells. Proc Natl Acad Sci USA 93:7036 –7041