Type 2 diabetes develops as a consequence of

-Cell Function and Viability in the Spontaneously Diabetic GK Rat Information From the GK/Par Colony B. Portha, M.-H. Giroix, P. Serradas, M.-N. Gangnerau, J. Movassat, F. Rajas, D. Bailbe, C. Plachot, G. Mithieux, and J.-C. Marie The GK rat model of type 2 diabetes is especially convenient to dissect the pathogenic mechanism necessary for the emergence of overt diabetes because all adult rats obtained in our department (GK/Par colony) to date have stable basal mild hyperglycemia and because overt diabetes is preceded by a period of normoglycemia, ranging from birth to weaning. The purpose of this article is to sum up the information so far available related to the biology of the -cell in the GK/Par rat. In terms of -cell function, there is no major intrinsic secretory defect in the prediabetic GK/Par -cell, and the lack of -cell reactivity to glucose (which reflects multiple intracellular abnormalities), as seen during the adult period when the GK/Par rats are overtly diabetic, represents an acquired defect (perhaps glucotoxicity). In terms of -cell population, the earliest alteration so far detected in the GK/Par rat targets the size of the -cell population. Several convergent data suggest that the permanently reduced -cell mass in the GK/Par rat reflects a limitation of -cell neogenesis during early fetal life, and it is conceivable that some genes among the set involved in GK diabetes belong to the subset of genes controlling early -cell development. Diabetes 50 (Suppl. 1):S89 S93, 2001 From the Laboratoire de Physiopathologie de la Nutrition (B.P., M.-H.G., P.S., M.-N.G., J.M., D.B., C.P., J.-C.M.), Université D. Diderot, Paris; and the Institut National de la Santé et de la Recherche Médicale (F.R., G.M.), Faculté Laënnec, Lyon, France. Address correspondence and reprint requests to Prof. B. Portha, Laboratoire de Physiopathologie de la Nutrition, CNRS ESA 7059, Université D. Diderot/Paris 7, 2 Place Jussieu, Tour 33, 75251 Paris Cedex 05, France. E-mail: portha@paris7.jussieu.fr. Received for publication 21 May 2000 and accepted in revised form 28 August 2000. This article is based on a presentation at a symposium. The symposium and the publication of this article were made possible by an unrestricted educational grant from Les Laboratoires Servier. GLP, glucagon-like protein; SERCA, calcium-atpase. Type 2 diabetes develops as a consequence of interplay among -cell dysfunction, peripheral insulin resistance, and elevated hepatic glucose production. However, it is not known which is the primary abnormality and which are abnormalities secondary to elevated plasma glucose, so-called glucose toxicity. To delineate the primary abnormalities, it is desirable to analyze individuals destined to become diabetic before the development of the disease. The advantage of using an animal model is that the development of diabetes can be predicted and thus it is possible to dissect the pathogenic mechanism necessary for the emergence of overt diabetes. The Goto-Kakizaki Wistar rat (GK rat) is especially useful because all adult animals of both sexes exhibit type 2 diabetes. This spontaneous diabetes model was produced by selective breeding (with glucose intolerance as a selection index) repeated over many generations, starting from a nondiabetic Wistar rat colony. The characteristics of GK animals bred in our colony in Paris (GK/Par) for more than 10 years (1) are very stable and remain close to those of the animals in the original Japanese colony (2): all of the rats have a basal mild hyperglycemia and impaired glucose tolerance. Males and females are similarly affected, and their diabetic state is stable over 72 weeks of follow-up (3). In adult GK rats, plasma insulin release in vivo in response to intravenous glucose is abolished (1,3). In vitro studies of insulin release with the isolated perfused pancreas (1) or with perifused islets (4) indicate that both early and late phases of glucoseinduced insulin release are markedly affected in the adult GK rat. Concerning insulin action in adult GK rats, we have reported decreased insulin sensitivity in the liver, in parallel with moderate insulin resistance in extrahepatic tissues, i.e., muscle and adipose tissue (5,6). In our colony (GK/Par), hyperglycemia is preceded by a period of normoglycemia, ranging from birth to weaning (7). Therefore, during this period, young GK rats can be considered to be prediabetic. DECREASED -CELL NUMBER AND MULTIPLE -CELL FUNCTIONAL DEFECTS IN THE ADULT GK/Par RAT WITH OVERT DIABETES In the adult GK rat, total pancreatic -cell mass is decreased (by 60%) in the range of the decrease in pancreatic insulin stores (1,7,8) (Fig. 1). This alteration of the -cell population cannot be ascribed to increased -cell apoptosis but is related, at least partly, to significantly decreased -cell replication (Fig. 1). Moreover, the adult GK pancreas exhibits two different populations of islets: large islets, which are disrupted by connective tissue (7) and display heterogeneity in the staining of the -cells, and small islets, with heavily stained -cells and normal architecture (Fig. 1). The islets of adult GK/Par rats, at least after collagenase digestion, show decreased -cells and low insulin content compared with control islets. The islet DNA content was decreased to a similar extent; this is consistent with our morphometric data (Fig. 2), which indicate that there is no major change in the relative contribution of -cells to total endocrine cells in the GK islets. In addition, in GK islets, S89

-CELL FUNCTION AND VIABILITY IN GK RAT -cell apoptotic -cell BrdU labeling % -cells index (%) index (%) in the pancreas FIG. 1. Left panel: Percentage of -cells in total pancreas, -cell mitotic index (BrdU labeling index), and -cell apoptotic index in 4-monthold male GK/Par rats and control Wistar rats. Values are expressed as means ± SE. *P < 0.001 compared with Wistar rats. Right panel: Immunoperoxidase staining for insulin in pancreas of 4-month-old male GK/Par rats (original magnification 125). Adult GK/Par pancreas exhibits two different populations of islets: large islets, which are disrupted by connective tissue and display heterogeneity in the staining of -cells (bottom), and small islets with heavily stained -cells and normal architecture (top). insulin content, when expressed relative to DNA, remains lower than in control islets, which supports degranulation in the -cells of diabetic animals. The notion that in the GK -cell the lesion responsible for loss of glucose-induced insulin secretion is mostly upstream of the effector system is supported by data indicating that GK islets are duly responsive to nonnutrient secretagogues, such as sulfonylureas or a combination of barium and theophylline (9). We have reported that glucose transport and glucose phosphorylating activity are not modified in the GK -cells (4,10). Consistent with these conclusions is the present observation that the expression of glucose transporter GLUT2 (by reverse transcriptase polymerase chain reaction) is normal in GK islets (Fig. 3). It is also noteworthy that in GK (as well as Wistar) rat islets, we were unable to detect expression of glucose-6-phosphatase (whereas in GK rat liver, glucose-6-phosphatase was easily detected) (Fig. 3). This is an interesting point because in another GK colony (11), it has been reported that islet glucose-6-phosphatase activity was increased, as was cycling between glucose and glucose-6-phosphate (11). We have shown that impaired glucose-induced insulin release in GK islets is associated with perturbation of multiple mitochondrial functions. More specifically, we reported that aerobic, but not anaerobic, glycolysis is impaired in GK islets (4,9,10), and we showed that mitochondria of GK islets exhibit a specific decrease in the activities of flavin adenine dinucleotide dependent glycerophosphate dehydrogenase (4,10) and branched-chain ketoacid dehydrogenase (12). Although this certainly may contribute to lower oxidation rates, it does not exclude other mechanisms. Indeed, we found that the -cells of adult GK rats had a significantly smaller mitochondrial volume than control -cells (13). No major deletion or restriction fragment polymorphism could be detected in mtdna from adult GK islets (13); however, they contained markedly less mtdna than did control islets. The lower islet mtdna was paralleled by decreased content of some islet mt mrnas, such as cytochrome b (13). In accordance with this, insufficient increase of ATP generation in response to high glucose was shown by our group (4) (Fig. 4). Finally, the impaired insulin response to glucose may be attributed to impaired elevation of intracellular Ca 2+ concentration, which seems to be a consequence of the failure by glucose to augment L-type Ca 2+ channel activity because of insufficient plasma membrane depolarization, reflecting impaired closure of ATP-sensitive K + channels (Fig. 4). We have recently obtained data supporting the view that abnormal Ca 2+ handling by the endoplasmic reticulum may also participate in defective Ca 2+ signaling; we investigated the cytosolic calcium response to high glucose in single perifused GK islets, as measured by dual-wave spec- S90

B. PORTHA AND ASSOCIATES FIG. 2. Characteristics of collagenase-isolated islets from 4-month-old male GK/Par rats and control Wistar rats. Data are means ± SE. The number of islet preparations is between 3 (morphometric studies) and 18 (biochemical determinations). In each experiment, DNA and insulin content values were obtained from two to five groups of 20 islets each, and the percentages of -cells and non- endocrine cells (,, and PP) were estimated in four to five islets. **P < 0.05, ***P < 0.001 compared with related value in control Wistar group. trophotometry using fura-2 (14). The most prominent difference is detected in the first 5 min following high glucose, because the GK islet lacks the initial reduction of cytosolic calcium. This initial reduction is thapsigargin-sensitive in the normal islet, suggesting that the sequestration of calcium by endoplasmic reticulum, attributed to activation of calcium- ATPases (SERCAs), is impaired in the GK -cell (Fig. 4). Such a conclusion is consistent with the report that SERCA3 gene expression is downregulated in GK islets (15). Alternatively, impaired calcium sequestration in the GK -cell can also be accounted for by insufficient cytosolic ATP generation in response to high glucose (Fig. 4). FIG. 3. Detection of glucose-6-phosphatase (Glc6Pase) and GLUT2 mrnas in islet and liver from 4-month-old male GK/Par rats and control Wistar (W) rats. Total RNAs were purified from freshly isolated islets and a frozen liver sample from GK and Wistar rats, using the Qiagen purification procedure. Glucose-6-phosphatase cdna (1,095 bp, exons 1 5) was amplified from 1 µg total RNA in 50 µl amplification medium as previously described (25), and 20 µl was analyzed on 1.5% agarose gel. GLUT2 cdna (808 bp, exons 6 11) was amplified from the same RNA samples and analyzed accordingly. Digestion products of phage- by HindIII/EcoRI were analyzed as size markers (DNA fragments appearing were 2,027, 1,904, 1,584, 1,330, 983, 832, and 564 bp long, as shown on the left of the figure from top to bottom). FIG. 4. Model for defective glucose-induced insulin release in the GK/Par -cell. Impaired insulin response to glucose may be attributed to impaired elevation of intracellular Ca 2+ concentration, which seems to be a consequence of the failure by glucose to augment L-type Ca 2+ channel activity, in its turn due to insufficient plasma membrane depolarization, reflecting impaired closure of the ATP-sensitive K + channels; this is the result of insufficient cytosolic ATP generation by glucose. Abnormal Ca 2+ handling by the endoplasmic reticulum (ER) in response to high glucose may also participate in the defective Ca 2+ signaling: the sequestration of calcium by ER during high-glucose exposure (attributed to activation of the SERCAs) is impaired in the GK rat -cell. Impaired calcium sequestration can also be accounted for by insufficient cytosolic ATP generation in response to high glucose: in GK islets, glucose fails to increase inositol triphosphate (IP3) accumulation. This is linked to an anomaly in targeting the phosphorylation of phosphoinositides: the activity of phosphatidylinositol kinase, the first of the two phosphorylating activities responsible for generating phosphatidylinositol biphosphate, is reduced. Moreover, deficient calcium handling and ATP supply in response to glucose probably also contribute to abnormal activation of phosphatidylinositol kinases and phospholipase C. S91

-CELL FUNCTION AND VIABILITY IN GK RAT BrdU labeling index of -cells (%) apoptotic index of -cells (%) pancreas weight (mg) total -cell mass (mg/pancreas) -cell mass (µg/mg pancreas) FIG. 5. Pancreas weight, total -cell mass, -cell BrdU labeling index, and -cell apoptotic index of 7-day-old GK/Par and Wistar (W) rat neonates. Values are expressed as means ± SE; the number of observations is four in each group. ***P < 0.001 compared with Wistar rats. Finally, we investigated phosphoinositides (16) and camp metabolism (17) in GK islets. Whereas carbachol was able to promote normal inositol generation in GK islets, high glucose failed to increase inositol phosphate accumulation (16). The inability of glucose to stimulate inositol phosphate production is not related to defective phospholipase C activity per se (total activity in islet homogenates is normal). Rather, it is linked to abnormal targeting of the phosphorylation of phosphoinositides; the activity of phosphatidylinositol kinase, which is the first of the two phosphorylating activities responsible for the generation of phosphatidylinositol biphosphate, is clearly reduced (16). Moreover, deficient calcium handling and ATP supply in response to glucose probably also contribute to abnormal activation of phosphatidylinositol kinases and phospholipase C. Concerning camp, it is remarkable that its intracellular content is already very high in GK -cells at low glucose (17). This is related to increased expression of the cyclase isoforms 2, 3, and 7, and of the G olf form of Gs proteins (18). Furthermore, camp is not further enhanced at increasing glucose concentrations (at variance with the situation in normal -cells) (17). This suggests that there exists a block in the steps linking glucose metabolism to activation of adenylate cyclase in the GK -cell. This contrasts strikingly with the capacity of the GK -cell to respond to glucagon-like protein (GLP)-1 such that it is able to restore the secretory competence to glucose with a clear biphasic response (17). This proves that the glucose-incompetence of the GK -cell is not irreversible and emphasizes the usefulness of GLP-1 as a therapeutic agent in type 2 diabetes. -CELL POPULATION AND -CELL FUNCTION IN THE GK/Par RAT DURING THE PREDIABETIC PERIOD Extensive follow-up of the animals after delivery has revealed that GK/Par pups become overtly hyperglycemic between 3 and 4 weeks of age. In GK neonates, total -cell mass is clearly decreased compared with that in Wistar rats FIG. 6. In vitro insulin release by freshly isolated islets from 1-weekold GK/Par rats in response to nutrient secretagogues. Data are expressed as absolute values (top) or as percentage of basal value determined at 2.8 mmol/l glucose in each group of islets (bottom). Each bar represents mean ± SE for 32 42 batches of islets obtained from six distinct islet preparations. ***P < 0.001 compared with the respective values obtained at 2.8 mmol/l glucose or 10 mmol/l leucine. Experiments with leucine or glutamine were carried out in the absence of glucose. S92

B. PORTHA AND ASSOCIATES (Fig. 5). Such -cell growth retardation cannot be ascribed to decreased -cell replication or to increased apoptosis (Fig. 5). We therefore postulate that the recruitment of new -cells from the precursor pool is defective in the young GK rat; other data from our group (19 23) suggest that the permanently reduced -cell mass in the GK model reflects a limitation of -cell neogenesis during early fetal life. At the same age, under in vitro static incubation conditions, GK rat islets release less insulin at basal glucose; however, they amplify their secretory response to high glucose, leucine, or leucine plus glutamine to the same extent as Wistar rat islets (Fig. 6). Therefore, there does not seem to exist a major intrinsic secretory defect in the prediabetic GK rat -cell. Consequently, the lack of -cell reactivity to glucose, as seen during the adult period, when the GK rats are overtly diabetic, represents an acquired defect (glucotoxicity?). In conclusion, the earliest alteration so far detected in the GK/Par rat targets the size of the -cell population. It is conceivable that some genes among those involved in GK/Par diabetes (24) belong to the subset of genes controlling early -cell development. REFERENCES 1. Portha B, Serradas P, Bailbé D, Suzuki K, Goto Y, Giroix M-H: -cell insensitivity to glucose in the GK rat, a spontaneous non-obese model for type II diabetes. Diabetes 41:486 491, 1991 2. Goto Y, Kakisaki M, Masaki N: Spontaneous diabetes produced by selective breeding of normal Wistar rats. Proc Jpn Acad Ser B 51:80 85, 1975 3. Berthelier C, Kergoat M, Portha B: Lack of deterioration of insulin action with aging in the GK rat: a contrasted pattern as compared to non-diabetic rat. Metabolism 46:890 896, 1997 4. Giroix MH, Sener A, Bailbé D, Leclercq-Meyer V, Portha B, Malaisse WJ: Metabolic, ionic and secretory response to D-glucose in islets from rats with acquired or inherited non-insulin-dependent diabetes. Biochem Med Metab Biol 50:301 321, 1993 5. Bisbis S, Bailbé D, Tormo M-A, Picarel-Blanchot F, Derouet M, Simon J, Portha B: Insulin resistance in the GK rat: decreased number but normal kinase of insulin receptor in the liver. Am J Physiol 265:E807 E813, 1993 6. Picarel-Blanchot F, Berthelier C, Bailbé D, Portha B: Impaired insulin secretion and excessive hepatic glucose production are both early events in the diabetic GK rat. Am J Physiol 271: E755 E762, 1996 7. Movassat J, Saulnier C, Portha B: -Cell mass depletion precedes the onset of hyperglycaemia in the GK rat, a genetic model of NIDDM. Diabet Metab 21:365 370, 1995 8. Movassat J, Saulnier C, Serradas P, Portha B: Impaired development of pancreatic -cell mass is a primary event during the progression to diabetes in the GK rat. Diabetologia 40:916 925, 1997 9. Giroix MH, Vesco L, Portha B: Functional and metabolic perturbations in isolated pancreatic islets from the GK rat, a genetic model of non-insulin dependent diabetes. Endocrinology 132:815 822, 1993 10. Giroix MH, Sener A, Portha B, Malaisse WJ: Preferential alteration of oxydative relative to total glycolysis in islets of rats with inherited or acquired noninsulin dependent diabetes. Diabetologia 36:305 309, 1993 11. Östenson CG, Abdel-Halim SM, Andersson A, Efendic S: Studies on the pathogenesis of NIDDM in the GK (Goto-Kakisaki) rat. In Lessons from Animal Diabetes VI. Shafrir E, Ed. Boston, Birkhäuser, 1996, p. 301 317 12. Giroix MH, Saulnier C, Portha B: Decreased pancreatic islet response to L-leucine in the spontaneously diabetic GK rat: enzymatic, metabolic and secretory data. Diabetologia 42:965 977, 1999 13. Serradas P, Giroix MH, Saulnier C, Gangnerau MN, Borg H, Welsh M, Portha B, Welsh N: Mitochondrial DNA content is specifically decreased in adult, but not fetal, pancreatic islets of the GK rat, a genetic model of non-insulin dependent diabetes. Endocrinology 136:5623 5631, 1995 14. Marie JC, Bailbé D, Portha B: Defective glucose-dependent cytosolic Ca 2+ handling in islets of GK and nstz rat models of NIDDM (Abstract). Diabetes 48:A334, 1999 15. Varadi A, Molnar E, Ostenson CG, Ashcroft SJH: Isoforms of endoplasmic reticulum Ca 2+ -ATPase are differentially expressed in normal and diabetic islets of Langerhans. Biochem J 319:521 527, 1996 16. Giroix MH, Morin L, Wolf B, Portha B: Defective islet phosphoinositide metabolism and insulin secretion in the GK rat model of type 2 diabetes (Abstract). Diabetologia. In press 17. Dachicourt N, Serradas D, Bailbé D, Portha B: Abnormal camp content and insulin release in islets of spontaneously diabetic GK rats (Abstract). Diabetologia 40:A114, 1997 18. Frayon S, Pessah M, Giroix MH, Mercan D, Boissard C, Malaisse WJ, Portha B, Garel JM: G olf identification by RT-PCR in purified normal pancreatic B cells and in islets from rat models of non-insulin-dependent diabetes. Biochem Biophys Res Commun 254:269 272, 1999 19. Serradas P, Gangnerau MN, Giroix MH, Saulnier C, Portha B: Impaired pancreatic -cell function in the fetal GK rat: impact of diabetic inheritance. J Clin Invest 101:899 904, 1998 20. Movassat J, Portha B: Beta cell growth in the neonatal Goto-Kakisaki rat and regeneration after treatment with streptozotocin at birth. Diabetologia 42:1098 1106, 1999 21. Plachot C, Movassat J, Saulnier C, Portha B: -Cell regeneration after pancreatectomy in the adult GK rat, a genetic model of MIDDM (Abstract). Diabetes 48:A244, 1999 22. Portha B, Movassat J, Plachot C, Saulnier C: Anomalies de la croissance et du potentiel régénératif des cellules beta pancréatiques chez le rat GK, modèle de DNID. Flammarion Mèdecine-Sciences, Journées de Diabétologie, 25 38, 1999 23. Miralles F, Portha B: Early development of -cells is impaired in the GK rat model of NIDDM. Diabetes 50 (Suppl. 1):S84 S88, 2001 24. Gauguier D, Froguel P, Parent V, Bernard C, Bihoreau MT, Portha B, James MR, Pénicaud L, Lathrop M, Ktorza A: Chromosomal mapping of genetic loci associated with non-insulin dependent diabetes in the GK rat. Nat Genet 12:38 43, 1996 25. Rajas F, Bruni N, Montano S, Zitoun N, Mithieux G: The glucose-6-phosphatase gene is expressed in human and rat small intestine: regulation of expression in fasted and diabetic rats. Gastroenterology 117:132 139, 1999 S93