Inducible Nitric-oxide Synthase and Nitric Oxide Donor Decrease Insulin Receptor Substrate-2 Protein Expression by Promoting Proteasome-dependent Degradation in Pancreatic β-Cells

Insulin receptor substrate-2 (IRS-2) plays a critical role in the survival and function of pancreatic β-cells. Gene disruption of IRS-2 results in failure of the β-cell compensatory mechanism and diabetes. Nonetheless, the regulation of IRS-2 protein expression in β-cells remains largely unknown. Inducible nitric-oxide synthase (iNOS), a major mediator of inflammation, has been implicated in β-cell damage in type 1 and type 2 diabetes. The effects of iNOS on IRS-2 expression have not yet been investigated in β-cells. Here, we show that iNOS and NO donor decreased IRS-2 protein expression in INS-1/832 insulinoma cells and mouse islets, whereas IRS-2 mRNA levels were not altered. Interleukin-1β (IL-1β), alone or in combination with interferon-γ (IFN-γ), reduced IRS-2 protein expression in an iNOS-dependent manner without altering IRS-2 mRNA levels. Proteasome inhibitors, MG132 and lactacystin, blocked the NO donor-induced reduction in IRS-2 protein expression. Treatment with NO donor led to activation of glycogen synthase kinase-3β (GSK-3β) and c-Jun N-terminal kinase (JNK/SAPK) in β-cells. Inhibition of GSK-3β by pharmacological inhibitors or siRNA-mediated knockdown significantly prevented NO donor-induced reduction in IRS-2 expression in β-cells. In contrast, a JNK inhibitor, SP600125, did not effectively block reduced IRS-2 expression in NO donor-treated β-cells. These data indicate that iNOS-derived NO reduces IRS-2 expression by promoting protein degradation, at least in part, through a GSK-3β-dependent mechanism. Our findings suggest that iNOS-mediated decreased IRS-2 expresssion may contribute to the progression and/or exacerbation of β-cell failure in diabetes.

␤-cells are, therefore, major contributors to the progression of diabetes.
Of interest, high fat diet reduced IRS-2 protein expression in islets of 90% pancreatectomized rats without alteration in IRS-2 mRNA expression (13). These data raise the possibility that high fat diet feeding might reduce IRS-2 protein expression by enhancing protein degradation.
Interleukin-1␤ (IL-1␤) has been proposed as a common important player in the pathogenesis of type 1 and type 2 diabetes (14). The blockade of IL-1␤ improves glycemic control and ␤-cell function in patients with type 2 diabetes (15). A previous study has shown that circulating levels of IL-1 receptor antagonist, an endogenous antagonist of IL-1␤, is associated with ␤-cell capacity in patients with type 1 diabetes (16). In nonobese diabetic mice, gene disruption of IL-1␤ receptor delays the development of diabetes (17).
Nitric oxide (NO) produced by inducible nitric-oxide synthase (iNOS, 2 also known as NOS2) has been implicated in ␤-cell damage and death in both type 1 and type 2 diabetes (18 -21) and during islet transplantation (22)(23)(24). Previous studies have shown that iNOS expression plays an important role in IL-1␤-induced dysfunction and death of cultured ␤-cells (20, 25, 26 -28), although other studies have concluded that cytokine induces ␤-cell damage through both NO-dependent and -independent pathways (29,30). The expression of iNOS is increased in ␤-cells and intraislet macrophages of rodent models of type 1 and type 2 diabetes (18,(31)(32)(33)(34)(35). Aminoguanidine, an iNOS inhibitor, prevents the development of obesity-induced diabetes in Zucker fa/fa rats (36). iNOS depletion and iNOS inhibitor have been shown to block or ameliorate diabetes development in multiple low dose streptozotocin-treated mice and nonobese diabetic mice, murine models of type 1 diabetes (18,27,28,37), although controversial results have been also reported (38,39). Moreover, ␤-cell-specific iNOS expression per se leads to insulin-dependent diabetes and loss of ␤-cells without insulitis in mice (41). However, it is not fully understood how NO and iNOS induce and/or exacerbate ␤-cell damage and loss of functional ␤-cell mass in diabetes. Here, we show that iNOS and NO donor reduce the protein expression of IRS-2 by promoting proteasome-dependent degradation of IRS-2 in cultured insulinoma cells and mouse islets.
Mouse Islet Isolation-Islets were isolated from male wildtype C57BL/6 mice, iNOS knock-out (Ϫ/Ϫ) mice on C57BL/6 background, wild-type BKS mice, and obese, diabetic (db/db) mice on BKS background (Jackson Laboratory, Bar Harbor, ME) at 9 -12 weeks of age by collagenase digestion followed by centrifugation over a Histopaque (Sigma) gradient, as described previously (44). Upper laparotomy was performed by midline incision, and after clamping the common bile duct at its entrance to the duodenum, 3 ml of M199 medium containing collagenase P (1 mg/ml; Roche Applied Science) was injected into the duct. The swollen pancreas was surgically removed.
Human Islets-Human islets from two non-diabetic subjects (male aged 39 years; female aged 45 years) and two type 2 diabetes patients (male aged 63 years on anti-diabetic medication for 5 years; male age 42 years on anti-diabetic medication for 4 years) were provided by the National Disease Research Interchange (Philadelphia, PA).
Western Blot Analysis-Cells or islets were lysed on ice for 30 min in lysis buffer (10 mM Tris-HCl, pH 7.4, 5 mM EDTA, 1% Nonidet P-40, 0.4% sodium deoxycholate, 10 mM sodium pyrophosphate, 10 mM sodium fluoride, 2 mM sodium vanadate, 1 mM phenylmethylsulfonyl fluoride, protease inhibitor mixture (Sigma)), as described previously (45). The protein samples were denatured by boiling for 5 min and separated in a 7.5 or 10% polyacrylamide gel and electrophoretically transferred to nitrocellulose membranes (Bio-Rad). The membranes were blocked in 2% ECL advance blocking agent (GE Healthcare) for 1 h at room temperature and incubated with primary antibody for 2 h at room temperature or overnight at 4°C. This was followed by incubation with secondary antibody conjugated with horseradish peroxidase for 1 h at room temperature. The blots were visualized by an enhanced chemiluminescence method using an ECL advance Western blotting detection kit (GE Healthcare). Bands of interest were scanned using an HP Scanjet 4850 and were quantified by NIH ImageJ 1.410 software (National Institutes of Health, Bethesda, MD).
Immunoprecipitation-Lysates were preincubated with 20 l of protein A/G-agarose beads (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) for 2 h at 4°C to minimize nonspecific absorption. The supernatants were then incubated with 2 g of anti-IRS-2 antibody (Santa Cruz Biotechnology, Inc.) and 20 l of protein A/G-agarose beads at 4°C overnight. After centrifugation at 1,000 ϫ g for 5 min, the pellets were washed five times with Tris-buffered saline (10 mM Tris-HCl, pH 7.4, 150 mM NaCl) and dissolved in 30 l of SDS-sample buffer.
Evaluation of mRNA Expression Levels-Total RNA was purified using TRIzol reagent (Invitrogen). The first-strand cDNA was synthesized from 1 g of total RNA using a high capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA). Real-time PCR reactions were performed using 10 ng of cDNA and TaqMan probes (Applied Biosystems) for IRS-2 (Rn01482270_s1 or Hs0065185_m1) and 18 S ribosomal RNA (Hs99999901_s1), conducted with Mastercycler ep realplex (Eppendorf, Westbury, NY). Results were normalized to 18 S ribosomal RNA as an endogenous reference gene, and the relative amount of each mRNA was calculated by the comparative Ct (threshold cycle) method. iNOS mRNA content in the islets was evaluated by RT-PCR, as described previously (46,47), using specific primers for mouse and human iNOS (mouse, 5Ј-ACAGCCTCAGAGT-CCTTCAT-3Ј and 5ЈTTGTCACCACCAGCAGTAGT-3Ј; human, 5Ј-CAGTACGTTTGGCAATGGAGACTGC-3Ј and 5Ј-GGTCACATTGGAGGTGTAGA GCTTG-3Ј). RT-PCR products were quantified using a densitometer and image analyzer (Bio-Rad) (46). 36B4 gene expression was used as an internal control (48).
Evaluation of Cell Viability-Cell viability of INS-1/832 cells and islet cells was assessed using Sytox Green (Molecular Probes, Inc., Eugene, OR) and TOX-8 (Sigma) according to the manufacturers' instructions. For Sytox staining, cells were incubated with Sytox Green (1 M) for 20 min in the dark and observed under a Nikon Eclipse TE2000-5 inverted fluorescence microscope.
Measurement of Nitrite-Nitrite accumulation in culture medium was determined by Griess reagent (Sigma). 50 l of culture medium was mixed and incubated with 50 l of Griess reagent for 15 min at room temperature, and absorbance at 540 nm was measured in a microplate reader. Serial dilutions of sodium nitrite were used as standards.
Statistical Analysis-The data were compared using one-way analysis of variance followed by Tukey's least significant difference test or unpaired Student's t test. A value of p Ͻ 0.05 was considered statistically significant. All data are expressed as mean Ϯ S.E. In accord with decreased IRS-2 expression by IL-1␤, basal and insulin-stimulated phosphorylation of IRS-2 and Akt was attenuated by IL-1␤, which was partially prevented by iNOS inhibitor, L-NIL, in INS-1/832 cells (Fig. 4). In the absence of IL-1␤, L-NIL did not affect the phosphorylation status of IRS-2  iNOS-dependent Decrease in IRS-2 Protein Expression in Mouse Islet Cells-Similar to INS-1/832 cells, treatment with IL-1␤ alone or IL-1␤ plus IFN-␥ decreased IRS-2 protein expression in cultured islets isolated from wild-type mice along with induction of iNOS expression (Fig. 5A), whereas mRNA content of IRS-2 was not decreased (mouse IRS-2 mRNA level: control, 100 Ϯ 9%; IL-1␤, 95 Ϯ 14%) IL-1␤ plus IFN-␥ induced more profound decrease in IRS-2 protein expression in association with greater increases in iNOS expression and nitrite accumulation in wild-type islets (Fig. 5B). In islets from iNOSdeficient mice, however, neither IL-1␤ alone nor IL-1␤ plus IFN-␥ affected IRS-2 expression or increased nitrite concentration in the culture medium (Fig. 5). iNOS inhibitor, L-NIL (200 M), blocked IL-1␤ plus IFN-␥-induced decrease in IRS-2 protein expression and nitrite accumulation in wild-type mouse islets (data not shown).

IL-1␤ Reduces IRS-2 Protein Expression in an iNOS-dependent Manner in Pancreatic
NO Donor Reduces IRS-2 Protein Expression in ␤-Cells-NO donor, GSNO, decreased the protein expression of IRS-2 in a time-dependent manner in INS-1/832 cells (Fig. 6A). GSNO (300 M) significantly decreased IRS-2 expression starting at 2 h after the addition of the NO donor. Treatment with GSNO for 5 h elicited the inhibitory effects on IRS-2 expression in a dose-dependent manner. A significant reduction in IRS-2 protein expression was observed starting at a dose of 100 M (Fig.  6B). To assess cell viability, we evaluated cell membrane integrity and metabolic cell viability by Sytox staining and TOX-8, respectively. Treatment with GSNO (300 M) for 5 h did not significantly affect cell viability (data not shown). Another NO donor, SNAP (300 M), also decreased IRS-2 protein expression in INS-1/832 cells (Fig. 6C). GSNO treatment for 5 h dosedependently decreased IRS-2 protein expression in ␤TC-6 cells as well (Fig. 6D). NO donors did not alter GAPDH protein expression in INS-1/832 and ␤TC-6 cells.
To investigate an involvement of proteasome-dependent protein degradation, we examined the effects of inhibitors for proteasome, MG132 and lactacystin. MG132 (1 M) and lactacystin (1 M) inhibited GSNO-induced reduction in IRS-2 protein expression in INS-1/832 cells and mouse islets (Figs. 8 (B-D) and 11C). GSNO treatment led to increased ubiquitination of IRS-2 protein in INS-1/832 cells (Fig. 8E). These findings support a role of ubiquitination-involved proteasome-dependent degradation in NO donor-induced suppression of IRS-2 protein expression in ␤-cells. Consistent with the role of protein degradation, treatment with GSNO reduced IRS-2 expression in the presence of cycloheximide, an inhibitor of protein synthesis, in INS-1/832 cells (Fig. 8F).    for GSK-3␤ and JNK/SAPK, respectively, in GSNO-treated INS-1/832 cells (Fig. 9). To the contrary, mTOR activity was decreased by GSNO, as reflected by decreased phosphorylation of p70S6K, a substrate of mTOR. The protein expression of glycogen synthase, c-Jun, and p70S6K was not altered by GSNO treatment or the inhibitors.

Inhibition of GSK-3␤ Significantly Ameliorated NO Donorinduced Suppression of IRS-2 Expression in
GSK-3␤ inhibitor, SB216763 (10 M), significantly prevented GSNO-induced reduction in IRS-2 protein expression in INS-1/832 cells (Fig. 9). In contrast, JNK inhibitor, SP 600125 (10 M), failed to significantly increase IRS-2 protein expression in GSNO-treated INS-1/832 cells, although it up-regulated slightly but significantly IRS-2 protein expression in NO donoruntreated cells. There was little effect if any of mTOR inhibitor, rapamycin (1 M), on IRS-2 expression in NO donor-treated ␤-cells. SP600125 and rapamycin effectively reduced activities of JNK/SAPK and mTOR, as judged by the phosphorylation status of c-Jun and p70S6K.

DISCUSSION
Here, we show that iNOS and NO donor decrease IRS-2 expression in insulinoma cell lines and mouse islet cells by increasing proteasome-dependent protein degradation of IRS-2. Our data clearly indicate that iNOS is required for IL-1␤and IL-1␤ plus IFN-␥-induced reduction in IRS-2 expression in cultured pancreatic ␤-cells.
IRS-2 has a critical role in functional ␤-cell mass. These findings, therefore, suggest that iNOS may mediate proinflammatory cytokine (i.e. IL-1␤)-induced damage and reduced mass of ␤-cells, at least in part, by suppressing IRS-2 expression in ␤-cells. iNOS expression is observed in ␤-cells and macrophages in the islets of diabetic rodents (18,(31)(32)(33)(34)(35). It is reasonable to speculate that NO derived from iNOS expressed in both intraislet macrophages and ␤-cells might contribute to decreased IRS-2 protein expression and ␤-cell dysfunction in vivo in diabetes.
Previous studies have revealed the transcriptional regulation of IRS-2 in ␤-cells. Glucagon-like peptide-1 (GLP-1) and GLP-1 receptor agonists promote the transcription and mRNA level of IRS-2 through cAMP-response element-binding protein-mediated mechanisms (49,50). Acute exposure to glucose results in a rapid increase in IRS-2 mRNA levels in ␤-cells (51), although underlying mechanisms remain to be determined. In contrast, neither cytokine nor NO donor altered IRS-2 mRNA levels in INS-1/832 cells and mouse islets (Figs. 4 and 8), whereas the protein expression of IRS-2 was markedly reduced. Our findings highlight a role of ubiquitination-involved proteasome-dependent degradation of IRS-2 in the detrimental effects of proinflammatory cytokines and NO donor in ␤-cells. In contrast to IRS-2, iNOS and NO donor did not alter the protein expression of other components of the insulin signaling pathway, including insulin receptor, Akt, GSK-3␤, mTOR, and p70S6K (Figs. 5, 9, 10, 11) (data not shown). It is conceivable, therefore, that NOmediated promotion of proteasomal degradation may be specific to IRS-2 rather than a consequence of general activation of the ubiquitin-proteasome pathway.
iNOS and NO donor can induce ␤-cell death. Previous studies have shown that iNOS mediates ␤-cell death induced by IL-1␤ (21,52). In our experiments, however, treatment with IL-1␤ (5 units/ml) or IL-1␤ (10 units/ml) plus IFN-␥ (100 ng/ml) for up to 24 h and exposure to GSNO (300 or 400 M) for 5 h did not affect cell viability of INS-1/832 cells and cultured mouse islets, although marked suppression of IRS-2 expression was observed. These results indicate that under these experimental conditions, the NO-mediated decrease in IRS-2 protein In contrast, lactocystin did not alter the phosphorylation status of glycogen synthase in mouse islets treated with and without GSNO. D, IRS-2 protein expression was decreased in islets isolated from obese, diabetic (db/db) mice compared with wild-type mice. The protein expression of GAPDH did not differ between the two groups. iNOS mRNA content was increased in islets from db/db mice relative to wild-type mice, whereas 36B4 expression did not differ between the two groups. *, p Ͻ 0.05; ***, p Ͻ 0.001 versus WT. n ϭ 4/group. Error bars, S.E.

iNOS Decreases IRS-2 Expression in Pancreatic ␤-Cells
is not attributable to reduced cell viability of the ␤-cells treated with cytokine or NO donor. Considering a critical role of IRS-2 in ␤-cell survival, however, it is possible that reduced IRS-2 expression might facilitate NO-involved ␤-cell death in diabetes.
Moreover, our results indicate that activation of GSK-3␤ plays a role in NO donor-induced reduction in IRS-2 protein expression in ␤-cells. Consistent with a previous study (52), iNOS and NO donor decreased basal and insulin-stimulated phosphorylation of Akt in ␤-cells (Fig. 5). This was associated with activation of GSK-3␤, as indicated by reduction in inhibitory phosphorylation of GSK-3␤ at serine 9 and elevation in phosphorylation of glycogen synthase, an endogenous substrate of GSK-3␤ (Figs. 9 -11). (Akt inhibits GSK-3␤ activity by phosphorylating serine 9 in GSK-3␤.) The inhibition of GSK-3␤ by pharmacological inhibitors or siRNA-mediated knockdown significantly prevented NO donor-induced reduction in IRS-2 expression in INS-1/832 cells and cultured mouse islets. NO donor treatment also resulted in increased phosphorylation (activation) of JNK/SAPK in ␤-cells, as shown previously (52). The inhibition of JNK/SAPK, however, did not block the effects of NO donor on IRS-2 expression in ␤-cells.
A previous study has shown that mTOR mediates high glucose-and insulin-like growth factor-I (IGF-I)-induced proteasome-dependent degradation of IRS-2 in ␤-cells (53). However, NO donor treatment led to a decrease in mTOR activity, and rapamycin failed to inhibit NO donor-induced reduction in IRS-2 protein expression in INS-1/832 cells (Fig. 9). These findings indicate that high glucose/IGF-I and NO donor promote IRS-2 protein degradation through distinct signaling pathways.
␤-Cell-specific overexpression of constitutively active GSK-3␤ causes reductions in ␤-cell mass and proliferation of ␤-cells (54). Of interest, these authors observed a trend of reduced IRS-2 protein expression in islets of the ␤-cell-specific GSK-3␤ transgenic mice compared with wild-type mice. However, the decreased IRS-2 protein expression was not statistically analyzed in their study. Nor was IRS-2 mRNA expression examined (54). Conversely, IRS-2 protein expression is up-regulated in the islets of ␤-cell-specific GSK-3␤ knock-out mice relative to wild-type mice (55). Importantly, GSK-3␤ inhibitors, SB216763 and LiCl, increased IRS-2 protein expression in cultured isolated islets and mouse MIN 6 insulinoma cells under serum deprivation (55), a condition that leads to activation of GSK-3␤. These observations seem consistent with our findings, although these authors did not evaluate IRS-2 expression at the mRNA level. Together, our results support the notion that activation of GSK-3␤ results in decreased protein expression of IRS-2 in ␤-cells.
␤-Cell-specific gene disruption of GSK-3␤ rescues the development of diabetes and reduced ␤-cell mass in IRS-2 knock-out mice (40). These results argue that GSK-3␤ can function as a downstream effector in mediation of the deleterious effects of IRS-2 deficiency in ␤-cells. Our results support an important role of GSK-3␤ in ␤-cell damage. Of note, our data indicate that GSK-3␤ may function as an upstream negative regulator of IRS-2 as well. Reduced IRS-2 expression leads to attenuated Akt activity in ␤-cells, which, in turn, results in activation of GSK-3␤ via decreased phosphorylation of serine 9 in GSK-3␤.
One can speculate, therefore, that activation of GSK-3␤ and reduction in IRS-2 expression might comprise a vicious cycle through a positive feedback mechanism, which contributes to progression of ␤-cell damage and diabetes. Further studies are required to clarify this point.
In conclusion, IL-1␤ alone and in combination with IFN-␥ reduces IRS-2 protein expression by promoting protein degradation in pancreatic ␤-cells. iNOS is required for the inhibitory effect of IL-1␤ and IL-1␤ plus IFN-␥ on IRS-2 protein expression. Our results indicate that GSK-3␤ plays an important role in NO-mediated suppression of IRS-2 protein expression in ␤-cells.