Upregulation of insulin receptor substrate-2 in pancreatic β cells prevents diabetes

The insulin receptor substrate-2 (Irs2) branch of the insulin/IGF signaling system coordinates peripheral insulin action and pancreatic β cell function, so mice lacking Irs2 display similarities to humans with type 2 diabetes. Here we show that β cell–specific expression of Irs2 at a low or a high level delivered a graded physiologic response that promoted β cell growth, survival, and insulin secretion that prevented diabetes in Irs2^–/– mice, obese mice, and streptozotocin-treated mice; and that upon transplantation, the transgenic islets cured diabetes more effectively than WT islets. Thus, pharmacological approaches that promote Irs2 expression in β cells, especially specific cAMP agonists, could be rational treatments for β cell failure and diabetes.

Diabetes mellitus arises from dysregulated glucose sensing or insulin secretion (mature onset diabetes of youth, or MODY), autoimmune-mediated β cell destruction (type 1), or insufficient compensation for peripheral insulin resistance (type 2) (1). As insulin resistance develops, type 2 diabetes is avoided by expanding functional β cells that secrete sufficient insulin quickly enough to maintain glucose homeostasis. Diet, acute or chronic stress, and obesity are major determinants of peripheral insulin sensitivity, and age exacerbates these environmental effects. While it is clear that insulin resistance and β cell dysfunction contribute to type 2 diabetes, there is considerable debate regarding the relative importance of these two abnormalities (2, 3). Studies with humans and rodents reveal a tightly controlled inverse relationship between insulin sensitivity and insulin secretion, suggesting that a feedback mechanism balances β cell function with the prevailing level of peripheral insulin sensitivity (3). When β cell compensation fails, glucose intolerance and type 2 diabetes develop, but the mechanisms involved and whether they are linked at the molecular level are unknown (4).

How insulin sensitivity influences insulin secretion is poorly understood. Recent work with transgenic mice suggests that β cell function is regulated by insulin and IGF signaling (5). The receptors for insulin and IGF1 promote phosphorylation of the insulin receptor substrates, Irs1 and Irs2, which engage various downstream pathways, including the Grb2/Sos → ras and the PI3K → Akt cascades (6–8). Although these pathways were originally described in heterologous cell lines and peripheral tissues, recent work suggests that they play a central role in β cell function. Insulin action and secretion are closely linked at the molecular level through the Irs2 branch of the insulin/IGF signaling cascade (9, 10). Global disruption of Irs2 causes peripheral insulin resistance that is compensated for by relative hyperinsulinemia early in life; however, Irs2^–/– mice progress toward diabetes as β cell mass decreases and insulin secretion fails (9, 11–13). By contrast, peripheral insulin resistance caused by obesity or by disruption of Irs1 is compensated for, at least in part, by expansion of β cell mass (9, 14). Disruption of the receptors for insulin or IGF1 in β cells impairs first-phase insulin secretion, which contributes to glucose intolerance (11–13). Together these results suggest that the Irs2 branch of the insulin/IGF signaling pathway is essential for β cell function throughout life.

Identifying a single genetic defect that causes insulin resistance and β cell failure has been difficult to accomplish, leading many investigators to conclude that ordinary type 2 diabetes is a polygenic disorder (3). By contrast, disruption of Irs2 in mice reveals a single defect that causes both insulin resistance and β cell dysfunction that culminates in diabetes (9, 15). Mutations in IRS2 are unusual and probably not associated with common type 2 diabetes in people (16). However, genetic polymorphism or environmental stress that compromises Irs2 function or diminishes its downstream signals might explain the heterogeneity of common type 2 diabetes in terms of a defined signaling cascade. Since Irs2 plays an important role in β cell function, we reasoned that its upregulation might prevent diabetes under conditions that cause insulin resistance or physiological stress. To test this hypothesis, we generated transgenic mice expressing Irs2 at a low or a high level in β cells under the control of the rat insulin II promoter (rip). Our results show that Irs2 expression prevents diabetes that results from a variety of causes.

Upregulation of Irs2 in β cells. To determine whether β cell dysfunction is a pivotal element in the development of diabetes in Irs2–/– mice, we generated transgenic C57BL/6 mice expressing a low (rip9→Irs2) or a high (rip13→Irs2) level of Irs2 in pancreatic β cells under the control of rip. Irs2 was measured by specific immunoblotting of islet extracts containing equal amounts of total protein (Figure 1a). Irs2 was detected in WT islet extracts but was undetectable in Irs2–/– islets. Concurrent immunoblots revealed twofold and 12-fold increased Irs2 in rip9→Irs2 and rip13→Irs2 extracts (Figure 1a), which was similar to that obtained physiologically by activation of the endogenous Irs2 promoter with functional cAMP response element binding protein (21). Moreover, the relative increase of transgenic Irs2 in islets was comparable to the upregulation of endogenous Irs2 in Min6 β cells treated for 10 hours with exendin-4 or dibutyl-cAMP (Figure 1b).

Figure 1

Irs2 protein expression and β cell function. (a) Irs2 protein levels measured by specific immunoblotting in Irs2 immunoprecipitates of 200 μg total protein from islet extracts of 4-week-old WT mice, Irs2–/– mice, or two lines of transgenic mice (rip9→Irs2 and rip13→Irs2). (b) Irs2 protein levels measured by specific immunoblotting in Irs2 immunoprecipitates of 200 μg total protein from Min6 cells treated for 8 hours with the indicated concentrations of exendin-4 or dibutyl-cAMP (diB-cAMP) (c) Representative WT, rip9→Irs2, or rip13→Irs2 pancreas sections immunostained with anti-insulin antibodies (left panel) or anti-FLAG antibodies (right panel) that were detected by cyanine (Irs2) or fluorescein (Insulin) conjugated secondary antibodies, respectively. Original magnification, ×100. (d) Pancreatic insulin content per mg total pancreas was measured in acid/ethanol extracts of total pancreas from 6-month-old WT, rip9→Irs2, and rip13→Irs2 mice. Data are the mean values ± SEM of five mice per genotype. *P < 0.05, **P < 0.01. (e) Glucose-stimulated insulin release measured in 6-week-old WT and rip13→Irs2 mice. D-glucose (3 g/kg body wt) was injected intraperitoneally into 16-hour-fasted mice and blood samples were collected at the indicated timepoints. Results are expressed as mean values ± SEM of six WT and six rip→Irs2 mice (*P < 0.05). (f) Irs2 protein levels measured by specific immunoblotting in Irs2 immunoprecipitates from hypothalamic extracts of 4-week-old WT, rip9→Irs2, rip13→Irs2, Irs2–/–, or Irs2–/–:rip13→Irs2 mice. rip9, rip9→Irs2; rip13, rip13→Irs2.

Our transgenic production strategy incorporated a FLAG tag at the C terminus of Irs2 to facilitate immunostaining. Pancreas sections immunostained with anti-FLAG antibodies revealed that recombinant Irs2 protein was restricted to the insulin-positive β cells in rip9→Irs2 and rip13→Irs2 islets (Figure 1c). Moreover, the pancreatic insulin content increased 1.3-fold in rip9→Irs2 mice and 2.7-fold in rip13→Irs2 mice, revealing a dose effect of Irs2 expression on β cell insulin content (Figure 1d). During an intraperitoneal glucose challenge, insulin secretion during the first 30-minute interval was threefold higher in rip13→Irs2 mice than in WT mice (AUC WT: 1,880 ± 50 nmol × min/l; AUC rip13→Irs2: 5,440 ± 50 nmol × min/l) (Figure 1e). However, glucose homeostasis was normal in rip13→Irs2 mice from birth until the experiment was terminated after 24 weeks, as fasting glucose never fell below 79 ± 3 mg/dl and fed glucose never rose above 159 ± 6 mg/dl. Both transgenic mouse lines displayed normal fertility, growth, and adiposity, and had normal life spans (data not shown). Irs2 protein was never detected in hypothalamus (Figure 1f), nor was it detected in liver, muscle, or adipose tissues (data not shown).

Irs2 in β cells prevents diabetes in Irs2–/– mice. Male C57BL/6 Irs2–/– mice developed hyperglycemia between 4 and 6 weeks of age, which progressed to overt diabetes during the next 5–6 weeks until they died (Figure 2a). To determine whether the Irs2 transgene prevented diabetes in the Irs2–/– mice, we crossed rip9→Irs2 or rip13→Irs2 mice with Irs2–/– mice. As expected, transgenic Irs2 was expressed at a low level in Irs2–/–:rip9→Irs2 islets and at a high level in Irs2–/–:rip13→Irs2 islets (data not shown). Irs2–/–:rip9→Irs2 mice survived for 24 weeks because hyperglycemia progressed slowly toward diabetes between 10 and 24 weeks of age (Figure 2a). By contrast, glucose levels of Irs2–/–:rip13→Irs2 mice were normal during the 24-week experiment, revealing a dose effect for β cell Irs2 expression on glucose homeostasis (Figure 2a).

Figure 2

Metabolic effects of Irs2 overexpression in β cells. (a) Blood glucose concentrations were measured from tail bleeds of fed male mice. Values are mean ± SEM obtained from WT, rip13→Irs2, Irs2–/–, and rip→Irs2 mice crossed into a C57BL/6 Irs2–/– background (Irs2–/–:rip9→Irs2 and Irs2–/–:rip13→Irs2). Data are the mean values ± SEM of four (Irs2–/–:rip9→Irs2) to ten (WT, Irs2–/–, rip13→Irs2, and Irs2–/–:rip13→Irs2) mice of the indicated ages. (b) Glucose-tolerance tests were performed on 8-week-old fasted mice following intraperitoneal loading with 2 g D-glucose per kg body wt. Blood samples were taken at the timepoints indicated and glucose was determined as described. Results are mean ± SEM of four to ten mice (***P < 0.001, Irs2–/– vs. Irs2–/–:rip13→Irs2). (c) Random-fed serum insulin levels (mean values ± SEM, ***P < 0.001) measured by ELISA.

While Irs2–/– mice at 8 weeks of age displayed fasting hyperglycemia (Irs2–/–: 189 ± 10 mg/dl; WT: 109 ± 6 mg/dl), Irs2–/– mice expressing the transgenic Irs2 were nearly normal (Irs2–/–:rip9→Irs2: 124 ± 9 mg/dl; Irs2–/–:rip13→Irs2: 121 ± 8 mg/dl). As previously shown, the Irs2–/– mice were severely hyperglycemic during an intraperitoneal glucose challenge (Figure 2b). However, at 8 weeks of age, glucose intolerance was slightly improved in the Irs2–/–:rip9→Irs2 mice and completely normalized in the Irs2–/–:rip13→Irs2 mice (Figure 2b). Moreover, the Irs2–/–:rip13→Irs2 mice survived more than a year after the Irs2–/– mice died, displaying low random fed blood glucose levels at 15 months (WT: 121 ± 7 mg/dl; Irs2–/–:rip13→Irs2: 94 ± 6 mg/dl).

Irs2–/– mice are insulin resistant as a result of reduced insulin action in liver, fat, and muscle (9, 22). Before diabetes developed at 6 weeks, serum insulin levels were equally elevated in Irs2–/–, Irs2–/–:rip9→Irs2, and Irs2–/–:rip13→Irs2 mice to compensate for insulin resistance (Figure 2c). Insulin levels declined dramatically in Irs2–/– mice at 8 weeks, coinciding exactly with the onset of severe diabetes (Figure 2c). Compensatory insulin secretion in Irs2–/–:rip9→Irs2 mice gradually declined between 12 and 25 weeks until they died with severe diabetes (Figure 2c). Irs2–/–:rip13→Irs2 mice never developed diabetes due to persistent compensatory hyperinsulinemia, revealing the graded physiological response to Irs2 expression in transgenic β cells functioning in a genetically insulin resistant mouse (Figure 2c and data not shown).

Irs2 promotes β cell development and growth. Our previous work suggests that Irs2 signaling regulates survival of pancreatic β cells (18, 19). Compared with WT mice at 8 weeks of age, islet area in Irs2–/– pancreas sections was reduced about threefold, owing to a slightly reduced density of small islets containing 50% fewer β cells (Figure 3a and Table 1). By contrast, islet area in the rip13→Irs2 sections increased twofold, mainly due to increased density of normal-sized islets (Table 1). Islet density and the β cell content increased in Irs2–/–:rip13→Irs2 mice, a change that was also revealed by the increased ratio of β cells to α cells (Table 1).

Figure 3

Islet morphometry. (a) Insulin immunostaining of representative pancreas sections from 8-week-old C57BL/6 mice and from rip13→Irs2, Irs2–/–/rip13→Irs2, and Irs2–/– mice (original magnification, ×100). (b) Western blot of Pdx1 protein expression in isolated islets from 4-week-old male mice of the indicated genotypes. Each lane was loaded with 100 μg of total islet protein from one animal and is representative of at least three independent experiments.

Table 1

Islet characteristics

To determine whether β cells in Irs2–/–:rip13→Irs2 islets were dividing at a higher rate than those in WT or rip13→Irs2 islets, we injected 8-week-old mice with the thymidine analogue BrdU to measure mitogenesis. BrdU incorporation into β cells during a 6-hour interval increased three- to fourfold in Irs2–/–:rip13→Irs2 mice compared with WT or rip13→Irs2 mice; no BrdU-positive cells were detected in Irs2–/– mice (Table 1). The average β cell size never changed upon expression of the Irs2 transgene. Thus, compensatory islet expansion during insulin resistance required Irs2 signaling to increase the number of normal-sized β cells.

The effect of Irs2 upregulation on islet gene expression. A possible mechanism by which Irs2 promotes expansion of β cells is through upregulation of the homeodomain transcription factor Pdx1 (also called Idx1 and Ipf1). Pdx1 is critical for development of the pancreas in mice and humans, and its complete disruption blocks pancreas development; in adult β cells, Pdx1 promotes normal glucose sensing and insulin secretion, and suppresses apoptosis (18, 23–25). Previous results show that Pdx1 expression is nearly lost in Irs2–/– islets, but can be restored by downregulation of Foxo1 (10). Pdx1 was detected equally by immunoblotting in WT and rip9→Irs2 islets (data not shown), whereas it was barely detected in our Irs2–/– islet extracts of equal protein concentration (Figure 3b). Consistent with a specific role for Irs2 in Pdx1 expression, Pdx1 protein was strongly upregulated rip13→Irs2 in WT and Irs2–/– mice containing the rip13→Irs2 transgene. (Figure 3b). Since upregulation of Pdx1 increases mitogenesis in Irs2–/– islets, upregulation of Pdx1 might contribute to the increased number of β cells in Irs2–/–:rip13→Irs2 islets (18, 25).

Pdx1 is reported to upregulate the expression of many genes that promote glucose-stimulated insulin secretion, including Glut2 (26). We used Affymetrix MG-U74v2 arrays to estimate the change in Glut2 mRNA between WT and rip13→Irs2 islets. The specificity of the Glut2 probe set was validated against the Glut1, Glut2, Glut3, and Glut4 probes using samples from rip13→Irs2 islets, brain, fat, liver, and muscle (Figure 4a). As expected, Glut2 was detected in liver and in rip13→Irs2 islets, but absent in the other test tissues; Glut1 and Glut3 were restricted to brain; and Glut4 was expressed exclusively in adipose and muscle tissue (Figure 4a). Compared with WT islets, Glut2 mRNA increased threefold (P < 0.005) in rip13→Irs2 islets (Figure 4b). Immunostaining revealed Glut2 in the plasma membrane of rip13→Irs2 and Irs2–/–:rip13→Irs2 β cells, whereas it was barely detected in the plasma membrane of WT islets (Figure 4c). The mRNA for other glycolytic enzymes also increased in rip13→Irs2 islets, including glucokinase (1.5-fold, P = 0.04), aldolase-1 (1.5-fold, P = 0.03), GAPDH (threefold, P < 0.005), and Pgk-1 (1.9-fold, P = 0.01); the β subunit of pyruvate dehydrogenase was also increased threefold (P < 0.005) (Figure 4b). Identical results were found for Irs2–/–:rip13→Irs2 islets. Since the secretion of insulin is tightly coupled to the rate of glucose metabolism, increased activity of glycolytic enzymes is consistent with improved β cell function (27).

Figure 4

Gene expression in WT tissues and rip13→Irs2 islets. (a) Affymetrix MG-U74v2 arrays (A, B, and C arrays) were used to estimate the relative mRNA levels for Glut1 (93738_at), Glut2 (103357_at), Glut3 (93804_at), and Glut4 (102314_at) in 6-week-old WT brain (B), adipose (F), rip13→Irs2 islets (I), liver (L), and muscle (M); the Affymetrix probe set is indicated in parentheses. Genes called present (Pr) or absent (A) by GeneChip (version 5) are indicated on the figure. The bars show averages of normalized expression measurements obtained from samples of two mice. (b) The fold change for mRNA in rip13→Irs2 islets normalized against WT islets is shown for Glut2, enzymes of the glycolytic pathway (glucokinase, Gck; phosphofructokinase-1, Pfk1; aldolase-1, Aldo1; GAPDH, Gapd; phosphoglycerokinase, Pgk1; phosphoglyceromutase-1, Pgam1; enolase, Eno1; pyruvate kinase, Pk3; pyruvate), and the pyruvate dehydrogenase α subunit (Pdha1) and β subunit (Pdhb). The expression of each gene in rip13→Irs2 islets is reported relative to the normalized level in WT islets; the P value for this comparison is indicated. The accession number for each enzyme is indicated above the P value. (c) Glut2 immunostaining of pancreas sections from 8-week-old WT, rip13→Irs2, and Irs2–/–:rip13→Irs2 mice (original magnification, ×400).

Irs2 in β cells prevents diabetes in obese and old mice. Peripheral insulin resistance develops during obesity and aging in mice and people, and progresses to diabetes when β cells fail to compensate with increased insulin secretion (28). To determine whether expression of Irs2 in β cells promotes compensation for obesity-induced diabetes, WT C57BL/6 mice or rip13Irs2 mice were weaned and maintained for 60 days on a low- or a high-fat diet. Mice fed the high-fat diet were obese at 12 weeks of age (C57BL/6: 36.5 ± 2.7 g, n = 6; rip13→Irs2: 37.3 ± 2.5 g, n = 6) compared with those eating a low-fat diet (C57BL/6: 24.6 ± 2.8 g, n = 6; rip13→Irs2: 25.4 ± 2.6 g, n = 6). Obese mice displayed fasting hyperglycemia (obese C57BL/6: 206 ± 16 mg/dl; lean C57BL/6: 101 ± 18 mg/dl) and glucose intolerance (Figure 4a). By contrast, obese rip13Irs2 mice had statistically normal fasting glucose levels (129 ± 15 mg/dl) and normal glucose tolerance (Figure 5a). Consistent with these results, fasting hyperinsulinemia was greater in obese rip13Irs2 mice than in obese C57BL/6 WT mice, and during the intraperitoneal glucose challenge insulin levels increased significantly in obese rip13→Irs2 mice but not in obese C57BL/6 WT mice (Figure 5b).

Figure 5

Metabolic characterization of WT and rip13→Irs2 mice maintained on a low-fat diet (LFD) or high-fat diet (HFD). (a) An intraperitoneal glucose tolerance test (2 g D-glucose per kg body wt) was performed on fasted C57BL/6 WT or rip13→Irs2 mice maintained on a high-fat or low-fat diet for 60 days after weaning. Blood samples were taken at the timepoints indicated and glucose levels were determined as described. Results are expressed as mean ± SEM of six WT and six rip13→Irs2 male mice (**P < 0.01, ***P < 0.001). (b) Serum insulin levels for the indicated timepoints during a glucose tolerance test. Data are the mean values ± SEM (*P < 0.05, **P < 0.01, ***P < 0.001). (c) Glucose tolerance test in 6-month-old C57BL/6 WT and rip13→Irs2 mice. Data are the mean values ± SEM of eight mice per genotype (*P < 0.05).

Irs2 expression falls as mice age (10). At 6 months, C57BL/6 mice on a normal diet developed mild glucose intolerance, whereas the rip13→Irs2 mice were normal (Figure 5c). At 40 weeks, islet area was 3.1% ± 0.6% in WT mice, 6% ± 1% in rip9→Irs2 mice, and 10% ± 2% in rip13→Irs2 mice, revealing a graded response to a low and high level of Irs2 expression. Thus, transgenic Irs2–mediated β cell expansion compensates for insulin resistance that develops during aging.

Irs2 promotes β cell survival signaling. Acute or chronic stress and autoimmune responses upregulate proinflammatory cytokines (including TNF-α, IL-1β, and IFN-γ) that promote destruction of β cells at least in part by contributing to apoptosis (29–32). By contrast, Irs protein signaling promotes cell growth and inhibits apoptosis in various cellular backgrounds (19). In isolated WT murine islets, IGF1 stimulated phosphorylation of Erk1/2, Akt, and the Akt target Foxo1 (Figure 6a). By contrast, in Irs2–/– islets the phosphorylation of these targets was reduced, and cleaved/activated caspase-3 accumulated and was insensitive to IGF1 stimulation, which is consistent with decreased growth and survival of Irs2–/– β cells (Figure 6a). To determine whether upregulation of Irs2 restores these signals, islets were isolated from WT mice, or Irs2–/–:rip13→Irs2 and rip13→Irs2 mice, cultured for 12 hours, and stimulated with IGF1. IGF1 stimulated phosphorylation of Erk1/2, Akt, and Foxo1 in WT islets; however, the basal and IGF1-stimulated phosphorylation of these proteins was significantly increased in Irs2–/–:rip13→Irs2 and rip13→Irs2 islets (Figure 6b). Moreover, upregulation of Irs2 eliminated the accumulation of cleaved cascapse-3, even before IGF1 stimulation, suggesting that Irs2 signaling plays a critical role in β cell survival (Figure 6b).

Figure 6

IGF1 and insulin signaling in isolated islets. (a) Islets were isolated from WT or Irs2–/– C57BL/6 mice and incubated overnight before stimulation with insulin or IGF1 for 20 minutes as described in Methods. Each lane was loaded with 100 μg of total islet protein. The results are representative of at least three independent experiments. (b) Islets were isolated from WT, Irs2–/–:rip13→Irs2, and rip13→Irs2 mice and incubated overnight before stimulation with IGF1 for 20 minutes as described in Methods. The same amount of total protein (300 μg) was loaded onto the gel and transferred to a nitrocellulose membrane. Membranes were analyzed for the presence of Erk1/2 and phosphorylated Erk (pErk1/2); Akt1/2 and phosphorylated Akt (pAkt1/2); Foxo1 and phosphorylated Foxo1 (pFoxo1); and cleaved caspase-3 (Casp3).

Apoptosis and function of rip13→Irs2 islets transplanted in mice. Since recombinant Irs2 dramatically reduced the level of cleaved caspase-3, we reasoned that β cell apoptosis might be reduced in rip13→Irs2 islets. We induced β cell apoptosis in situ by injecting a low dose of streptozotocin for 5 consecutive days into 8-week-old WT or rip13→Irs2 mice with comparable body weight (21.7 ± 3 g) and comparable levels of random-fed serum glucose (158 ± 11 mg/dl) and plasma insulin (255 ± 60 pmol/l). Blood glucose and serum insulin levels were measured before streptozotocin injection and 6, 12, and 15 days after the first streptozotocin injection. At 15 days, T cell infiltration was comparable in C57BL/6 WT and rip13→Irs2 mice (Figure 7a). However, low-dose streptozotocin caused insulinopenic diabetes in C57BL/6 WT mice, whereas the rip13→Irs2 mice were normal (Figure 7, b and c). Apoptotic β cells were readily detected in WT mice, but apoptosis was reduced more than 50% in rip13→Irs2 mice (Figure 7d). Thus, transgenic Irs2 prevented the onset of diabetes 15 days after streptozotocin injection, at least in part by suppressing β cell apoptosis.

Figure 7

Survival and function of rip13→Irs2 mice during streptozotocin-induced diabetes. (a) Representative H&E-stained pancreas sections from untreated WT or rip13Irs2 islets compared with pancreas sections obtained 15 days after one injection per day for 5 days of low-dose (40 mg/kg) streptozotocin (strep; original magnification, ×200). (b) Blood glucose levels in WT or rip13→Irs2 mice were measured before and after five (daily) injections of low-dose (40 mg/kg) streptozotocin. Results are expressed as mean ± SEM of eight WT and eight rip13→Irs2 mice (**P < 0.01); (c) Glucose levels (mg/dl) divided by serum insulin levels (ng/dl) are shown for experimental days 0, 6, and 15 (***P < 0.001). (d) Apoptotic cells were detected in deparaffinized sections using a rhodamine DNA fragmentation detection assay. The number of apoptotic nuclei per β cell is shown for day 15. Values are expressed as mean ± SEM of eight WT and eight transgenic mice (***P < 0.001). (e) Transplantation of WT islets into streptozotocin-diabetic mice. C57BL/6 mice were treated with 100 mg/kg streptozotocin for 3 consecutive days. Blood glucose levels were measured in samples obtained through tail bleeds of fed mice before transplantation (shaded area) and at the indicated ages after transplantation of 300 (filled circles) or 150 (open circles) WT islets. (f) Transplantation of rip13→Irs2 islets into streptozotocin-diabetic mice was conducted in identical fashion with 250 (filled circles), 120 (open circles), or 50 (inverted triangles) islets. Values are mean ± SEM of at least three mice per experiment.

To test whether rip13→Irs2 islets are resistant to the effects of stress induced by transplantation, WT or rip13→Irs2 islets were transplanted under the kidney capsule of streptozotocin-diabetic mice. In this experiment, 8-week-old C57BL/6 mice were treated with 100 mg/kg streptozotocin for 3 consecutive days to cause severe diabetes (Figure 7, e and f). Without islet transplantation, the diabetic mice died about 10 days after the first injection. By contrast, transplantation of 300 WT islets (>10,000 islets/kg) normalized random-fed glucose levels by 50% within 3–5 days, and glucose levels were nearly normal 60 days later when the experiment was terminated (Figure 7e). By contrast, transplantation of 150 WT islets prevented death but failed to normalize glucose levels (Figure 7e), whereas 80 islets were completely ineffective as the diabetic mice died (not shown). Importantly, 50 rip13→Irs2 islets (β cell mass equivalent to 80 WT islets) normalized random-fed glucose levels by 50% after 20 days, and quicker treatment was obtained with 120 and 250 rip13→Irs2 islets (Figure 7f). BrdU incorporation measured in the islet grafts 60 days after transplantation revealed that DNA synthesis was twofold greater in rip13→Irs2 islets (2.12% ± 0.32% BrdU positive β cells, n = 3) than in WT islets (1.2% ± 0.2% BrdU positive β cells, n = 3). Thus, upregulation of Irs2 in transplanted β cells significantly reduced the number of islets needed to cure diabetes in mice, at least in part by promoting proliferation.

The primary cause of type 2 diabetes is unknown; however, there is general agreement that insulin resistance is an early event in the onset of the disease (2, 33). Many insulin-resistant patients secrete sufficient insulin for many years before β cell function fails to compensate and type 2 diabetes emerges (2, 34, 35). Systemic failure of Irs2 in mice causes peripheral insulin resistance followed by β cell failure and diabetes, although the time interval is compressed relative to human type 2 diabetes (36). Here we show that upregulation of Irs2 exclusively in β cells promotes glucose tolerance in old mice and prevents diabetes in Irs2–/– and obese mice. Moreover, IRS2 protects β cells from destruction by streptozotocin and improves the function of isolated β cells used for transplantation. Thus, upregulation of Irs2 in β cells might be a common mechanism to prevent or treat many forms of diabetes.

Massive destruction of β cells is well known in type 1 diabetics, and human autopsy studies reveal that β cell mass is reduced at least 50% in obese patients with type 2 diabetes (37, 38). Thus, insufficient β cell mass relative to peripheral requirements might be a key pathogenic factor of both types of diabetes (38). One way to preserve β cell mass is to reduce apoptosis (25, 39). Streptozotocin in multiple low doses induces β cell apoptosis and diabetes in mice and other mammals, at least in part by promoting T cell infiltration that promotes the production of TNF-α and IFN-γ in the pancreas (40, 41). Upregulation of Irs2 in mouse β cells does not reduce T cell infiltration during streptozotocin treatment; however, it reduces apoptosis of β cells by 50%, and prevents the onset of diabetes for at least 15 days. Several pathways promote survival of β cells, and many of these mechanism are mediated by insulin/IGF1 → Irs2 signaling (19). Moreover, Irs2 signaling mediates the phosphorylation of Bad and the forkhead transcription factor Foxo1, which control important survival pathways (42, 43). The activation of Akt is an important mechanism used by insulin and IGF1 to inhibit apoptosis in a variety of cellular environments (44). Recent results suggest that Akt2 might promote β cell survival by linking Irs2 to Foxo1, especially during cytokine-induced stress (8).

Phosphorylated Bad dissociates from the antiapoptotic protein Bcl2, which inhibits activation of the caspase cascade, including the accumulation of the cleaved/activated effector caspase-3, which promotes apoptosis (42). IGF1 or insulin stimulation ordinarily inhibits the accumulation of cleaved/activated caspase-3 through activation of Akt; however, insulin- and IGF1-stimulated Akt phosphorylation is reduced in Irs2–/– islets, increasing the accumulation of cleaved/activated caspase-3 that promotes β cell apoptosis (19). IGF1 stimulation weakly reduces the levels of cleaved/activated caspase-3 in our WT isolated islets, consistent with low or undetectable Irs2 levels following overnight incubations. However, transgenic upregulation of Irs2 almost completely blocks the accumulation of activated caspase-3, even before addition of exogenous IGF1. Thus, Irs2 is an important inhibitor of caspase-mediated apoptosis.

Foxo1 is an important Akt substrate that links insulin/IGF signaling to gene regulation (43). Phosphorylated Foxo1 is ordinarily excluded from the nucleus, which changes the expression of various genes including the upregulation of Pdx1 (10). Pdx1 plays an important role in gut and β cell development, and promotes glucose sensing and insulin secretion in adult β cells (45–47). Partial Pdx1 deficiency leads to an organ-level defect in insulin secretion and diabetes (25). Transgenic upregulation of Foxo1 constitutively increases the level of nuclear Foxo1, which downregulates Pdx1 and impairs β cell function (10). By contrast, reduced expression of Pdx1 in Irs2–/– islets is consistent with nuclear accumulation of dephosphorylated Foxo1, whereas increased Pdx1 levels in Irs2–/–:Foxo1+/– islets reflect the reduced content of cellular Foxo1 (10). Similarly, upregulation of Irs2 in rip13→Irs2 or Irs2–/–:rip13→Irs2 β cells increases Foxo1 phosphorylation to promote its exclusion from the nucleus, which is consistent with elevated Pdx1. Thus, Pdx1 might mediate many of the effects of Irs2 in β cells, as transgenic upregulation of Pdx1 restores β cell function and promotes normal glycemia in Irs2–/– mice (18). Interestingly, Pdx1 expression persists in some lines of C57BL/6 mice lacking Irs2, which protects them from diabetes (48).

Whether Irs2 signaling promotes β cell growth is difficult to determine. Before birth, Irs2 signaling is not required for β cell development, as Irs2–/– neonates have sufficient islet function to avoid diabetes for the first 8 weeks of life. However, β cell mass fails to expand in Irs2–/– mice even though peripheral insulin resistance is expected to promote expansion as it does in Irs1–/– mice (9). Alone, upregulation of Irs2 does not increase mitogenesis in β cells; however, islet density increases equally in both rip13→Irs2 and Irs2–/–:rip13→Irs2 mice. Alternatively, BrdU labeling is detectable in Irs2–/–:rip13→Irs2 mice, suggesting that insulin resistance mediates β cell mitogenesis through Irs2 signaling. Unfortunately, our experiments do not reveal the mechanism by which Irs2 increases islet density. However, Kitamura et al. suggest that nuclear exclusion of Foxo1 from ductal precursors might promote the formation of new duct-associated islets by upregulating Pdx1 (10). Thus, relative hyperinsulinemia during the initial response to insulin resistance might promote Irs2-mediated β cell growth. Unfortunately, this mechanism is expected to eventually fail, as Irs2 signaling diminishes during progressive insulin resistance.

Although our work focuses on murine β cells, preliminary experiments show that upregulation of IRS2 in isolated human islets by adenoviral-mediated infection increases glucose-stimulated insulin release and BrdU incorporation (data not shown). Thus, pharmacologic or genetic approaches that upregulate Irs2 in β cells or promote Irs2-mediated downstream signals, such as reduction of nuclear Foxo1, might also improve β cell function in people. Tools to upregulate Irs2 have not been available because the regulation of Irs2 gene expression is poorly studied. However, results with HeLa cells suggest that human IRS2 is moderately upregulated by cAMP and strongly upregulated by progesterone or the combination of cAMP and RU486 (49).

In our experiments, glucagen-like peptide 1, exendin-4 or dibutyl-cAMP strongly upregulated Irs2 protein in Min6 β cells, and similar results were observed with isolated human islets (data not shown). A mechanism for the regulation of Irs2 by cAMP was recently suggested by Jhala et al., who demonstrated a functional CRE in the murine Irs2 gene (21). Thus, the trophic effect of GLP1/exendin-4 in rodents and people might be explained, at least in part, by the upregulation of IRS2 (50, 51). GLP1 secretion decreases in people with type 2 diabetes, but subcutaneous administration of GLP1 improves glucose homeostasis; lowers body weight, hemoglobin A(1C), and free fatty acids; and increases insulin sensitivity and insulin action (52, 53). Moreover, GLP1/exendin-4 inhibits both cytokine and fatty acid–induced apoptosis in β cells, and delays the development of diabetes in Zucker diabetic fatty rats by reducing β cell loss (54). Since these changes are also expected upon increased Irs2 protein levels, GLP1/exendin-4 or other β cell–specific cAMP agonists might prevent β cell failure by promoting Irs2 signaling.

As absolute or relative insulin insufficiency is associated with all forms of diabetes, β cell replacement via transplantation represents an attractive alternative for treating human diabetics; however, the isolation of sufficient (at least 10,000 islets/kg) high-quality human islets that survive and function after transplantation is challenging (55). Given the capacity of the rip13→Irs2 transgene to preserve or enhance β cell function during transplantation in mice, upregulation of IRS2 by pharmacologic or genetic approaches just before transplantation might improve the function of human islets and reduce the number needed for successful transplantation.

In summary, upregulation of Irs2 in β cells disrupts disease progression in mice, and if expressed highly enough, can prevent the onset of diabetes caused by Irs2 disruption or diet-induced obesity. The Irs2 signaling system includes many elements that propagate the insulin/IGF signal to cytoplasmic regulatory enzymes and to nuclear regulatory elements. In β cells, this includes certain MODY genes, revealing an intriguing link between insulin signaling and autosomal early onset diabetes (10, 18). Further investigation of the Irs2 branch of the insulin/IGF signaling cascade might explain the relationship between β cell failure and peripheral insulin resistance that causes type 2 diabetes. If the IRS2 signaling cascade is a master regulator that integrates β cell function with peripheral insulin action, upregulation of IRS2 by nutritional, pharmacologic, or genetic intervention could be a general way to slow progression or prevent the onset of diabetes.

The work was supported by grants from the NIH (R01-DK55326-04 to M.F. White) and the Deutsche Forschungsgemeinschaft (HE-3321 to A.M. Hennige) as well as by the Juvenile Diabetes Foundation (J. Ye). M.F. White is an associate investigator at the Howard Hughes Medical Institute.

Conflict of interest: The authors have declared that no conflict of interest exists.

Nonstandard abbreviations used: mature onset diabetes of youth (MODY); insulin receptor substrate (Irs); rat insulin II promoter (rip); murine genome (MG); glucose transporter-2 (Glut2); area under the curve (AUC).

1. Zimmet, P, Alberti, KG, Shaw, J. Global and societal implications of the diabetes epidemic. Nature. 2001. 414:782-787.
   View this article via: PubMed CrossRef Google Scholar
2. Taylor, SI, Accili, D, Imai, Y. Insulin resistance or insulin deficiency. Which is the primary cause of NIDDM? Diabetes. 1994. 43:735-740.
   View this article via: PubMed Google Scholar
3. Kahn, SE. The relative contributions of insulin resistance and beta-cell dysfunction to the pathophysiology of type 2 diabetes. Diabetologia. 2003. 46:3-19.
   View this article via: PubMed CrossRef Google Scholar
4. Ehrmann, DA, et al. Insulin secretory responses to rising and falling glucose concentrations are delayed in subjects with impaired glucose tolerance. Diabetologia. 2002. 45:509-517.
   View this article via: PubMed CrossRef Google Scholar
5. Kitamura, T, Kahn, CR, Accili, D. Insulin receptor knockout mice. Annu. Rev. Physiol. 2003. 65:313-332.
   View this article via: PubMed CrossRef Google Scholar
6. Yenush, L, White, MF. The IRS-signalling system during insulin and cytokine action. Bioessays. 1997. 19:491-500.
   View this article via: PubMed CrossRef Google Scholar
7. Kido, Y, Nakae, J, Accili, D. Clinical review 125: the insulin receptor and its cellular targets. J. Clin. Endocrinol. Metab. 2001. 86:972-979.
   View this article via: PubMed CrossRef Google Scholar
8. Garofalo, RS, et al. Severe diabetes, age-dependent loss of adipose tissue, and mild growth deficiency in mice lacking Akt2/PKBβ. J. Clin. Invest. 2003. 112:197-208. doi:10.1172/JCI200316885.
   View this article via: JCI PubMed Google Scholar
9. Withers, DJ, et al. Disruption of IRS-2 causes type 2 diabetes in mice. Nature. 1998. 391:900-904.
   View this article via: PubMed CrossRef Google Scholar
10. Kitamura, T, et al. The forkhead transcription factor Foxo1 links insulin signaling to Pdx1 regulation of pancreatic β cell growth. J. Clin. Invest. 2002. 110:1839-1847. doi:10.1172/JCI200216857.
    View this article via: JCI PubMed Google Scholar
11. Kulkarni, RN, et al. beta-cell-specific deletion of the Igf1 receptor leads to hyperinsulinemia and glucose intolerance but does not alter beta-cell mass. Nat. Genet. 2002. 31:111-115.
    View this article via: PubMed Google Scholar
12. Kulkarni, RN, et al. Tissue-specific knockout of the insulin receptor in pancreatic β cells creates an insulin secretory defect similar to that in type 2 diabetes. Cell. 1999. 96:329-339.
    View this article via: PubMed CrossRef Google Scholar
13. Xuan, S, et al. Defective insulin secretion in pancreatic β cells lacking type 1 IGF receptor. J. Clin. Invest. 2002. 110:1011-1019. doi:10.1172/JCI200215276.
    View this article via: JCI PubMed Google Scholar
14. Bonner-Weir, S. Islet growth and development in the adult. J. Mol. Endocrinol. 2000. 24:297-302.
    View this article via: PubMed CrossRef Google Scholar
15. Avruch, J. A signal for b-cell failure. Nature. 1998. 391:846-847.
    View this article via: PubMed CrossRef Google Scholar
16. Ehrmann, DA, Tang, X, Yoshiuchi, I, Cox, NJ, Bell, GI. Relationship of insulin receptor substrate-1 and -2 genotypes to phenotypic features of polycystic ovary syndrome. J. Clin. Endocrinol. Metab. 2002. 87:4297-4300.
    View this article via: PubMed CrossRef Google Scholar
17. Sutton, R, Peters, M, McShane, P, Gray, DW, Morris, PJ. Isolation of rat pancreatic islets by ductal injection of collagenase. Transplantation. 1986. 42:689-691.
    View this article via: PubMed Google Scholar
18. Kushner, JA, et al. Pdx1 restores beta cell function in Irs2 knockout mice. J. Clin. Invest. 2002. 109:1193-1201. doi:10.1172/JCI200214439.
    View this article via: JCI PubMed Google Scholar
19. Withers, DJ, et al. Irs-2 coordinates Igf-1 receptor-mediated beta-cell development and peripheral insulin signalling. Nat. Genet. 1999. 23:32-40.
    View this article via: PubMed Google Scholar
20. Flier, SN, Kulkarni, RN, Kahn, CR. Evidence for a circulating islet cell growth factor in insulin-resistant states. Proc. Natl. Acad. Sci. U. S. A. 2001. 98:7475-7480.
    View this article via: PubMed CrossRef Google Scholar
21. Jhala, US, et al. cAMP promotes pancreatic beta-cell survival via CREB-mediated induction of IRS2. Genes Dev. 2003. 17:1575-1580.
    View this article via: PubMed CrossRef Google Scholar
22. Previs, SF, Withers, DJ, Ren, JM, White, MF, Shulman, GI. Contrasting effects of IRS-1 vs. IRS-2 gene disruption on carbohydrate and lipid metabolism in vivo. J. Biol. Chem. 2000. 275:38990-38994.
    View this article via: PubMed CrossRef Google Scholar
23. Jonsson, J, Carlsson, L, Edlund, T, Edlund, H. Insulin-promoter-factor 1 is required for pancreas development in mice. Nature. 1994. 371:606-609.
    View this article via: PubMed CrossRef Google Scholar
24. Stoffers, DA, Zinkin, NT, Stanojevic, V, Clarke, WL, Habener, JF. Pancreatic agenesis attributable to a single nucleotide deletion in the human IPF1 gene coding sequence. Nat. Genet. 1997. 15:106-110.
    View this article via: PubMed CrossRef Google Scholar
25. Johnson, JD, et al. Increased islet apoptosis in Pdx1+/– mice. J. Clin. Invest. 2003. 111:1147-1160. doi:10.1172/JCI200316537.
    View this article via: JCI PubMed Google Scholar
26. Ahlgren, U, Jonsson, J, Jonsson, L, Simu, K, Edlund, H. beta-cell-specific inactivation of the mouse Ipf1/Pdx1 gene results in loss of the beta-cell phenotype and maturity onset diabetes. Genes Dev. 1998. 12:1763-1768.
    View this article via: PubMed CrossRef Google Scholar
27. Yoon, JC, et al. Suppression of beta cell energy metabolism and insulin release by PGC-1alpha. Dev. Cell. 2003. 5:73-83.
    View this article via: PubMed CrossRef Google Scholar
28. Matthews, DR, et al. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985. 28:412-419.
    View this article via: PubMed CrossRef Google Scholar
29. Hammonds, P, et al. Insulin-secreting beta-cells possess specific receptors for interleukin-1 beta. Growth Regul. 1990. 261:97-100.
30. Zumsteg, U, et al. Differential interleukin-1 receptor antagonism on pancreatic beta and alpha cells. Studies in rodent and human islets and in normal rats. Diabetologia. 1993. 36:759-766.
    View this article via: PubMed CrossRef Google Scholar
31. Mandrup-Poulsen, T, Bendtzen, K, Dinarello, CA, Nerup, J. Human tumor necrosis factor potentiates human interleukin 1-mediated rat pancreatic beta-cell cytotoxicity. J. Immunol. 1987. 139:4077-4082.
    View this article via: PubMed Google Scholar
32. Suarez-Pinzon, WL, Strynadka, K, Schulz, R, Rabinovitch, A. Mechanisms of cytokine-induced destruction of rat insulinoma cells: the role of nitric oxide. Endocrinology. 1994. 134:1006-1010.
    View this article via: PubMed CrossRef Google Scholar
33. Shulman, GI. Cellular mechanisms of insulin resistance. J. Clin. Invest. 2000. 106:171-176.
    View this article via: JCI PubMed CrossRef Google Scholar
34. Reaven, GM. Multiple CHD risk factors in type 2 diabetes: beyond hyperglycaemia. Diabetes Obes. Metab. 2002. 4(Suppl. 1):S13-S18.
    View this article via: PubMed CrossRef Google Scholar
35. Reaven, GM. Insulin resistance: why is it important to treat? Diabetes Metab. 2001. 27:247-253.
    View this article via: PubMed Google Scholar
36. White, MF. IRS proteins and the common path to diabetes. Am. J. Physiol. Endocrinol. Metab. 2002. 283:E413-E422.
    View this article via: PubMed Google Scholar
37. Clark, A, et al. Islet amyloid, increased A-cells, reduced B-cells and exocrine fibrosis: quantitative changes in the pancreas in type 2 diabetes. Diabetes Res. 1988. 9:151-159.
    View this article via: PubMed Google Scholar
38. Butler, AE, et al. Beta-cell deficit and increased beta-cell apoptosis in humans with type 2 diabetes. Diabetes. 2003. 52:102-110.
    View this article via: PubMed CrossRef Google Scholar
39. Pick, A, et al. Role of apoptosis in failure of β-cell mass compensation for insulin resistance and β-cell defects in the male Zucker diabetic fatty rat. Diabetes. 1998. 47:358-364.
    View this article via: PubMed CrossRef Google Scholar
40. Herold, KC, et al. Regulation of cytokine production during development of autoimmune diabetes induced with multiple low doses of streptozotocin. J. Immunol. 1996. 156:3521-3527.
    View this article via: PubMed Google Scholar
41. Holstad, M, Sandler, S. A transcriptional inhibitor of TNF-alpha prevents diabetes induced by multiple low-dose streptozotocin injections in mice. J. Autoimmun. 2001. 16:441-447.
    View this article via: PubMed CrossRef Google Scholar
42. Brazil, DP, Hemmings, BA. Ten years of protein kinase B signalling: a hard Akt to follow. Trends Biochem. Sci. 2001. 26:657-664.
    View this article via: PubMed CrossRef Google Scholar
43. Tran, H, Brunet, A, Griffith, EC, Greenberg, ME. The many forks in FOXO’s road. Sci. STKE. 2003. 2003:RE5.
    View this article via: PubMed Google Scholar
44. Datta, SR, et al. Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell. 1997. 91:231-241.
    View this article via: PubMed CrossRef Google Scholar
45. Peers, B, Leonard, J, Sharma, S, Teitelman, G, Montminy, MR. Insulin expression in pancreatic islet cells relies on cooperative interactions between the helix loop helix factor E47 and the homeobox factor STF-1. Mol. Endocrinol. 1994. 8:1798-1806.
    View this article via: PubMed CrossRef Google Scholar
46. Watada, H, et al. The human glucokinase gene beta-cell-type promoter: an essential role of insulin promoter factor 1/PDX-1 in its activation in HIT-T15 cells. Diabetes. 1996. 45:1478-1488.
    View this article via: PubMed CrossRef Google Scholar
47. Waeber, G, Thompson, N, Nicod, P, Bonny, C. Transcriptional activation of the GLUT2 gene by the IPF-1/STF-1/IDX-1 homeobox factor. Mol. Endocrinol. 1996. 10:1327-1334.
    View this article via: PubMed CrossRef Google Scholar
48. Suzuki, R., et al. 2003. Pdx1 expression in Irs2 deficient mouse beta-cells is regulated in a strain-dependent manner. J. Biol. Chem. doi:10.1074/jbc.M307004200.
49. Vassen, L, Wegrzyn, W, Klein-Hitpass, L. Human insulin receptor substrate-2 (IRS-2) is a primary progesterone response gene. Mol. Endocrinol. 1999. 13:485-494.
    View this article via: PubMed CrossRef Google Scholar
50. Holst, JJ. Therapy of type 2 diabetes mellitus based on the actions of glucagon-like peptide-1. Diabetes Metab. Res. Rev. 2002. 18:430-441.
    View this article via: PubMed CrossRef Google Scholar
51. Li, Y, et al. Glucagon-like peptide-1 receptor signaling modulates beta cell apoptosis. J. Biol. Chem. 2003. 278:471-478.
    View this article via: PubMed CrossRef Google Scholar
52. Lugari, R, et al. Evidence for early impairment of glucagon-like peptide 1-induced insulin secretion in human type 2 (non insulin-dependent) diabetes. Horm. Metab. Res. 2002. 34:150-154.
    View this article via: PubMed Google Scholar
53. Zander, M, Madsbad, S, Madsen, JL, Holst, JJ. Effect of 6-week course of glucagon-like peptide 1 on glycaemic control, insulin sensitivity, and beta-cell function in type 2 diabetes: a parallel-group study. Lancet. 2002. 359:824-830.
    View this article via: PubMed CrossRef Google Scholar
54. Farilla, L, et al. Glucagon-like peptide-1 promotes islet cell growth and inhibits apoptosis in Zucker diabetic rats. Endocrinology. 2002. 143:4397-4408.
    View this article via: PubMed CrossRef Google Scholar
55. Goss, JA, et al. Achievement of insulin independence in three consecutive type-1 diabetic patients via pancreatic islet transplantation using islets isolated at a remote islet isolation center. Transplantation. 2002. 74:1761-1766.
    View this article via: PubMed Google Scholar