We quantified the contribution of GLP-1 as a mediator of the therapeutic effects of dipeptidyl peptidase 4 (DPP-4) inhibition (vildagliptin) by using the GLP-1 receptor antagonist exendin [9-39] in patients with type 2 diabetes and in healthy subjects. Thirty-two patients with type 2 diabetes and 29 age- and weight-matched healthy control subjects were treated in randomized order with 100 mg once daily vildagliptin or placebo for 10 days. Meal tests were performed (days 9 and 10) without and with a high-dose intravenous infusion of exendin [9-39]. The main end point was the ratio of the areas under the curve (AUCs) of integrated insulin secretion rates (total AUCISR) and glucose (total AUCglucose) over 4 h after the meal. Vildagliptin treatment more than doubled responses of intact GLP-1 and glucose-dependent insulinotropic polypeptide and lowered glucose responses without changing AUCISR/AUCglucose in healthy subjects. Vildagliptin significantly increased this ratio by 10.5% in patients with type 2 diabetes, and exendin [9-39] reduced it (both P < 0.0001). The percentage reduction in the AUCISR/AUCglucose ratio achieved with exendin [9-39] was significantly smaller after vildagliptin treatment than after placebo treatment (P = 0.026) and was equivalent to 47 ± 5% of the increments due to vildagliptin. Thus, other mediators appear to contribute significantly to the therapeutic effects of DPP-4 inhibition.

Many studies have shown that human natural killer T (NKT) cells can promote immunity to pathogens, but their regulatory function is still being investigated. Invariant NKT (iNKT) cells have been shown to be effective in preventing type 1 diabetes in the NOD mouse model. Activation of plasmacytoid dendritic cells, modulation of B-cell responses, and immune deviation were proposed to be responsible for the suppressive effect of iNKT cells. We studied the regulatory capacity of human iNKT cells from control subjects and patients with type 1 diabetes (T1D) at disease clinical onset. We demonstrate that control iNKT cells suppress the proliferation of effector T cells (Teffs) through a cell contact-independent mechanism. Of note, suppression depended on the secretion of interleukin-13 (IL-13) by iNKT cells because an antibody blocking this cytokine resulted from the abrogation of Teff suppression; however, T1D-derived iNKT cells showed impaired regulation that could be attributed to the decrease in IL-13 secretion. Thus, alteration of the IL-13 pathway at disease onset may lead to the progression of the autoimmune response in T1D. Advances in the study of iNKT cells and the selection of agonists potentiating IL-13 secretion should permit new therapeutic strategies to prevent the development of T1D.

Protein S is an anticoagulant factor that also regulates inflammation and cell apoptosis. The effect of protein S on diabetes and its complications is unknown. This study compared the development of diabetes between wild-type and transgenic mice overexpressing human protein S and the development of diabetic glomerulosclerosis between mice treated with and without human protein S and between wild-type and protein S transgenic mice. Mice overexpressing protein S showed significant improvements in blood glucose level, glucose tolerance, insulin sensitivity, and insulin secretion compared with wild-type counterparts. Exogenous protein S improved insulin sensitivity in adipocytes, skeletal muscle, and liver cell lines in db/db mice compared with controls. Significant inhibition of apoptosis with increased expression of BIRC3 and Bcl-2 and enhanced activation of Akt/PKB was induced by protein S in islet β-cells compared with controls. Diabetic wild-type mice treated with protein S and diabetic protein S transgenic mice developed significantly less severe diabetic glomerulosclerosis than controls. Patients with type 2 diabetes had significantly lower circulating free protein S than healthy control subjects. This study shows that protein S attenuates diabetes by inhibiting apoptosis of β-cells and the development of diabetic nephropathy.

In humans, the glucagon response to moderate-to-marked insulin-induced hypoglycemia (IIH) is largely mediated by the autonomic nervous system. Because this glucagon response is impaired early in type 1 diabetes, we sought to determine if these patients, like animal models of autoimmune diabetes, have an early and severe loss of islet sympathetic nerves. We also tested whether this nerve loss is a permanent feature of type 1 diabetes, is islet-selective, and is not seen in type 2 diabetes. To do so, we quantified pancreatic islet and exocrine sympathetic nerve fiber area from autopsy samples of patients with type 1 or 2 diabetes and control subjects without diabetes. Our central finding is that patients with either very recent onset (<2 weeks) or long duration (>10 years) of type 1 diabetes have a severe loss of islet sympathetic nerves (Δ = -88% and Δ = -79%, respectively). In contrast, patients with type 2 diabetes lose no islet sympathetic nerves. There is no loss of exocrine sympathetic nerves in either type 1 or type 2 diabetes. We conclude that patients with type 1, but not type 2, diabetes have an early, marked, sustained, and islet-selective loss of sympathetic nerves, one that may impair their glucagon response to IIH.

Sulfonylureas are ATP-sensitive potassium (KATP) channel blockers commonly used in the treatment of type 2 diabetes mellitus (T2DM). Activation of KATP channels plays a neuroprotective role in ischemia; thus, whether sulfonylureas affect the outcomes of stroke in patients with T2DM needs to be further studied. In our study, streptozotocin (STZ)-induced diabetic mice subjected to transient middle cerebral artery occlusion (MCAO) showed larger areas of brain damage and poorer behavioral outcomes. Blocking the KATP channel by tolbutamide increased neuronal injury induced by oxygen-glucose deprivation (OGD) in vitro and permanent MCAO (pMCAO) in vivo. Activating the KATP channel by diazoxide reduced the effects of both the OGD and pMCAO. Western blot analysis in STZ mouse brains indicated an early increase in protein levels of N-methyl-d-aspartate receptor 2B and postsynaptic density protein-95, followed by a decrease in phosphorylation of glycogen synthase kinase 3β. Our systematic meta-analysis indicated that patients with T2DM treated with sulfonylureas had a higher odds ratio for stroke morbidity than those who received comparator drugs. Taken together, these results suggest that sulfonylurea treatment in patients with T2DM may inhibit the neuroprotective effects of KATP channels and increase the risk of stroke.

Fibrosis is emerging as a hallmark of metabolically dysregulated white adipose tissue (WAT) in obesity. Although adipose tissue fibrosis impairs adipocyte plasticity, little is known about how aberrant extracellular matrix (ECM) remodeling of WAT is initiated during the development of obesity. Here we show that treatment with the antidiabetic drug metformin inhibits excessive ECM deposition in WAT of ob/ob mice and mice with diet-induced obesity, as evidenced by decreased collagen deposition surrounding adipocytes and expression of fibrotic genes including the collagen cross-linking regulator LOX Inhibition of interstitial fibrosis by metformin is likely attributable to the activation of AMPK and the suppression of transforming growth factor-β1 (TGF-β1)/Smad3 signaling, leading to enhanced systemic insulin sensitivity. The ability of metformin to repress TGF-β1-induced fibrogenesis is abolished by the dominant negative AMPK in primary cells from the stromal vascular fraction. TGF-β1-induced insulin resistance is suppressed by AMPK agonists and the constitutively active AMPK in 3T3L1 adipocytes. In omental fat depots of obese humans, interstitial fibrosis is also associated with AMPK inactivation, TGF-β1/Smad3 induction, aberrant ECM production, myofibroblast activation, and adipocyte apoptosis. Collectively, integrated AMPK activation and TGF-β1/Smad3 inhibition may provide a potential therapeutic approach to maintain ECM flexibility and combat chronically uncontrolled adipose tissue expansion in obesity.

To determine the roles of insulin and insulin-like growth factor 1 (IGF-1) action in adipose tissue, we created mice lacking the insulin receptor (IR), IGF-1 receptor (IGF1R), or both using Cre-recombinase driven by the adiponectin promoter. Mice lacking IGF1R only (F-IGFRKO) had a ∼25% reduction in white adipose tissue (WAT) and brown adipose tissue (BAT), whereas mice lacking both IR and IGF1R (F-IR/IGFRKO) showed an almost complete absence of WAT and BAT. Interestingly, mice lacking only the IR (F-IRKO) had a 95% reduction in WAT, but a paradoxical 50% increase in BAT with accumulation of large unilocular lipid droplets. Both F-IRKO and F-IR/IGFRKO mice were unable to maintain body temperature in the cold and developed severe diabetes, ectopic lipid accumulation in liver and muscle, and pancreatic islet hyperplasia. Leptin treatment normalized blood glucose levels in both groups. Glucose levels also improved spontaneously by 1 year of age, despite sustained lipodystrophy and insulin resistance. Thus, loss of IR is sufficient to disrupt white fat formation, but not brown fat formation and/or maintenance, although it is required for normal BAT function and temperature homeostasis. IGF1R has only a modest contribution to both WAT and BAT formation and function.

Aging is associated with the development of chronic diseases such as insulin resistance and type 2 diabetes. A reduction in mitochondrial fat oxidation is postulated to be a key factor contributing to the progression of these diseases. Our aim was to investigate the contribution of impaired mitochondrial fat oxidation toward age-related disease. Mice deficient for malonyl CoA decarboxylase (MCD(-/-)), a mouse model of reduced fat oxidation, were allowed to age while life span and a number of physiological parameters (glucose tolerance, insulin tolerance, indirect calorimetry) were assessed. Decreased fat oxidation in MCD(-/-) mice resulted in the accumulation of lipid intermediates in peripheral tissues, but this was not associated with a worsening of age-associated insulin resistance and, conversely, improved longevity. This improvement was associated with reduced oxidative stress and reduced acetylation of the antioxidant enzyme superoxide dismutase 2 in muscle but not the liver of MCD(-/-) mice. These findings were recapitulated in aged mice treated with an MCD inhibitor (CBM-3001106), and these mice also demonstrated improvements in glucose and insulin tolerance. Therefore, our results demonstrate that in addition to decreasing fat oxidation, MCD inhibition also has novel effects on protein acetylation. These combined effects protect against age-related metabolic dysfunction, demonstrating that MCD inhibitors may have utility in the battle against chronic disease in the elderly.

Sensory neurons are particularly susceptible to neuronal damage in diabetes, and silent mating type information regulation 2 homolog 1 (Sirt1) has been recently identified as a key gene in neuroprotection and wound healing. We found that the expression of Sirt1 was downregulated in trigeminal sensory neurons of diabetic mice. A microRNA microarray analysis identified microRNA-182 (miR-182) as a Sirt1 downstream effector, and the expression level of miR-182 was increased by Sirt1 overexpression in trigeminal neurons; Sirt1 bound to the promoter of miR-182 and regulated its transcription. We also revealed that miR-182 enhanced neurite outgrowth in isolated trigeminal sensory neurons and overcame the detrimental effects of hyperglycemia by stimulating corneal nerve regeneration by decreasing the expression of one of its target genes, NOX4. Furthermore, the effects of miR-182 on corneal nerve regeneration are associated with a functional recovery of corneal sensation in hyperglycemic conditions. These data demonstrate that miR-182 is a key regulator in diabetic corneal nerve regeneration through targeting NOX4, suggesting that miR-182 might be a potential target for the treatment of diabetic sensory nerve regeneration and diabetic keratopathy.

The brain influences glucose homeostasis, partly by supplemental control over insulin and glucagon secretion. Without this central regulation, diabetes and its complications can ensue. Yet, the neuronal network linking to pancreatic islets has never been fully mapped. Here, we refine this map using pseudorabies virus (PRV) retrograde tracing, indicating that the pancreatic islets are innervated by efferent circuits that emanate from the hypothalamus. We found that the hypothalamic arcuate nucleus (ARC), ventromedial nucleus (VMN), and lateral hypothalamic area (LHA) significantly overlap PRV and the physiological glucose-sensing enzyme glucokinase. Then, experimentally lowering glucose sensing, specifically in the ARC, resulted in glucose intolerance due to deficient insulin secretion and no significant effect in the VMN, but in the LHA it resulted in a lowering of the glucose threshold that improved glucose tolerance and/or improved insulin sensitivity, with an exaggerated counter-regulatory response for glucagon secretion. No significant effect on insulin sensitivity or metabolic homeostasis was noted. Thus, these data reveal novel direct neuronal effects on pancreatic islets and also render a functional validation of the brain-to-islet neuronal map. They also demonstrate that distinct regions of the hypothalamus differentially control insulin and glucagon secretion, potentially in partnership to help maintain glucose homeostasis and guard against hypoglycemia.

The endoplasmic reticulum quality control protein activating transcription factor 6 (ATF6) has emerged as a novel metabolic regulator. Here, we show that adenovirus-mediated overexpression of the dominant-negative form of ATF6 (dnATF6) increases susceptibility to develop hepatic steatosis in diet-induced insulin-resistant mice and fasted mice. Overexpression of dnATF6 or small interfering RNA-mediated knockdown of ATF6 decreases the transcriptional activity of peroxisome proliferator-activated receptor α (PPARα)/retinoid X receptor complex, and inhibits oxygen consumption rates in hepatocytes, possibly through inhibition of the binding of PPARα to the promoter of its target gene. Intriguingly, ATF6 physically interacts with PPARα, enhances the transcriptional activity of PPARα, and triggers activation of PPARα downstream targets, such as CPT1α and MCAD, in hepatocytes. Furthermore, hepatic overexpression of the active form of ATF6 promotes hepatic fatty acid oxidation and protects against hepatic steatosis in diet-induced insulin-resistant mice. These data delineate the mechanism by which ATF6 controls the activity of PPARα and hepatic mitochondria fatty acid oxidation. Therefore, strategies to activate ATF6 could be used as an alternative avenue to improve liver function and treat hepatic steatosis in obesity.

Skin fluorescence (SF) noninvasively measures advanced glycation end products (AGEs) in the skin and is a risk indicator for diabetes complications. N-acetyltransferase 2 (NAT2) is the only known locus influencing SF. We aimed to identify additional genetic loci influencing SF in type 1 diabetes (T1D) through a meta-analysis of genome-wide association studies (N = 1,359) including Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) and Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR). A locus on chromosome 1, rs7533564 (P = 1.9 × 10(-9)), was associated with skin intrinsic fluorescence measured by SCOUT DS (excitation 375 nm, emission 435-655 nm), which remained significant after adjustment for time-weighted HbA1c (P = 1.7 × 10(-8)). rs7533564 was associated with mean HbA1c in meta-analysis (P = 0.0225), mean glycated albumin (P = 0.0029), and glyoxal hydroimidazolones (P = 0.049), an AGE measured in skin biopsy collagen, in DCCT. rs7533564 was not associated with diabetes complications in DCCT/EDIC or with SF in subjects without diabetes (nondiabetic [ND]) (N = 8,721). In conclusion, we identified a new locus associated with SF in T1D subjects that did not show similar effect in ND subjects, suggesting a diabetes-specific effect. This association needs to be investigated in type 2 diabetes.

Insulin controls glucose uptake into adipose and muscle cells by regulating the amount of GLUT4 in the plasma membrane. The effect of insulin is to promote the translocation of intracellular GLUT4 to the plasma membrane. The small Rab GTPase, Rab10, is required for insulin-stimulated GLUT4 translocation in cultured 3T3-L1 adipocytes. Here we demonstrate that both insulin-stimulated glucose uptake and GLUT4 translocation to the plasma membrane are reduced by about half in adipocytes from adipose-specific Rab10 knockout (KO) mice. These data demonstrate that the full effect of insulin on adipose glucose uptake is the integrated effect of Rab10-dependent and Rab10-independent pathways, establishing a divergence in insulin signal transduction to the regulation of GLUT4 trafficking. In adipose-specific Rab10 KO female mice, the partial inhibition of stimulated glucose uptake in adipocytes induces insulin resistance independent of diet challenge. During euglycemic-hyperinsulinemic clamp, there is no suppression of hepatic glucose production despite normal insulin suppression of plasma free fatty acids. The impact of incomplete disruption of stimulated adipocyte GLUT4 translocation on whole-body glucose homeostasis is driven by a near complete failure of insulin to suppress hepatic glucose production rather than a significant inhibition in muscle glucose uptake. These data underscore the physiological significance of the precise control of insulin-regulated trafficking in adipocytes.

Atrasentan, a selective endothelin A receptor antagonist, has been shown to reduce albuminuria in type 2 diabetes. We previously showed that the structural integrity of a glomerular endothelial glycocalyx is required to prevent albuminuria. Therefore we tested the potential of atrasentan to stabilize the endothelial glycocalyx in diabetic apolipoprotein E (apoE)-deficient mice in relation to its antialbuminuric effects. Treatment with atrasentan (7.5 mg/kg/day) for 4 weeks reduced urinary albumin-to-creatinine ratios by 26.0 ± 6.5% (P < 0.01) in apoE knockout (KO) mice with streptozotocin-induced diabetes consuming an atherogenic diet, without changes in gross glomerular morphology, systemic blood pressure, and blood glucose concentration. Endothelial cationic ferritin surface coverage, investigated using large-scale digital transmission electron microscopy, revealed that atrasentan treatment increases glycocalyx coverage in diabetic apoE KO mice from 40.7 ± 3.2% to 81.0 ± 12.5% (P < 0.05). This restoration is accompanied by increased renal nitric oxide concentrations, reduced expression of glomerular heparanase, and a marked shift in the balance of M1 and M2 glomerular macrophages. In vitro experiments with endothelial cells exposed to laminar flow and cocultured with pericytes confirmed that atrasentan reduced endothelial heparanase expression and increased glycocalyx thickness in the presence of a diabetic milieu. Together these data point toward a role for the restoration of endothelial function and tissue homeostasis through the antialbuminuric effects of atrasentan, and they provide a mechanistic explanation for the clinical observations of reduced albuminuria with atrasentan in diabetic nephropathy.

Ghrelin signaling has major effects on energy and glucose homeostasis, but it is unknown whether ghrelin's functions are centrally and/or peripherally mediated. The ghrelin receptor, growth hormone secretagogue receptor (GHS-R), is highly expressed in the brain and detectable in some peripheral tissues. To understand the roles of neuronal GHS-R, we generated a mouse line where Ghsr gene is deleted in all neurons using synapsin 1 (Syn1)-Cre driver. Our data showed that neuronal Ghsr deletion abolishes ghrelin-induced spontaneous food intake but has no effect on total energy intake. Remarkably, neuronal Ghsr deletion almost completely prevented diet-induced obesity (DIO) and significantly improved insulin sensitivity. The neuronal Ghsr-deleted mice also showed improved metabolic flexibility, indicative of better adaption to different fuels. In addition, gene expression analysis suggested that hypothalamus and/or midbrain might be the sites that mediate the effects of GHS-R in thermogenesis and physical activity, respectively. Collectively, our results indicate that neuronal GHS-R is a crucial regulator of energy metabolism and a key mediator of DIO. Neuronal Ghsr deletion protects against DIO by regulating energy expenditure, not by energy intake. These novel findings suggest that suppressing central ghrelin signaling may serve as a unique antiobesity strategy.

The mechanisms of metabolic improvements after Roux-en-Y gastric bypass (RYGB) surgery are not entirely clear. Therefore, the aim of our study was to investigate the role of obesity and RYGB on the human skeletal muscle proteome. Basal muscle biopsies were obtained from seven obese (BMI >40 kg/m(2)) female subjects (45.1 ± 3.6 years) pre- and 3 months post-RYGB, and euglycemic-hyperinsulinemic clamps were used to assess insulin sensitivity. Four age-matched (48.5 ± 4.7 years) lean (BMI <25 kg/m(2)) females served as control subjects. We performed quantitative mass spectrometry and microarray analyses on protein and RNA isolated from the muscle biopsies. Significant improvements in fasting plasma glucose (104.2 ± 7.8 vs. 86.7 ± 3.1 mg/dL) and BMI (42.1 ± 2.2 vs. 35.3 ± 1.8 kg/m(2)) were demonstrated in the pre- versus post-RYGB, both P < 0.05. Proteomic analysis identified 2,877 quantifiable proteins. Of these, 395 proteins were significantly altered in obesity before surgery, and 280 proteins differed significantly post-RYGB. Post-RYGB, 49 proteins were returned to normal levels after surgery. KEGG pathway analysis revealed a decreased abundance in ribosomal and oxidative phosphorylation proteins in obesity, and a normalization of ribosomal proteins post-RYGB. The transcriptomic data confirmed the normalization of the ribosomal proteins. Our results provide evidence that obesity and RYGB have a dynamic effect on the skeletal muscle proteome.

In the current study, we used muscle-specific TRIB3 overexpressing (MOE) and knockout (MKO) mice to determine whether TRIB3 mediates glucose-induced insulin resistance in diabetes and whether alterations in TRIB3 expression as a function of nutrient availability have a regulatory role in metabolism. In streptozotocin diabetic mice, TRIB3 MOE exacerbated, whereas MKO prevented, glucose-induced insulin resistance and impaired glucose oxidation and defects in insulin signal transduction compared with wild-type (WT) mice, indicating that glucose-induced insulin resistance was dependent on TRIB3. In response to a high-fat diet, TRIB3 MOE mice exhibited greater weight gain and worse insulin resistance in vivo compared with WT mice, coupled with decreased AKT phosphorylation, increased inflammation and oxidative stress, and upregulation of lipid metabolic genes coupled with downregulation of glucose metabolic genes in skeletal muscle. These effects were prevented in the TRIB3 MKO mice relative to WT mice. In conclusion, TRIB3 has a pathophysiological role in diabetes and a physiological role in metabolism. Glucose-induced insulin resistance and insulin resistance due to diet-induced obesity both depend on muscle TRIB3. Under physiological conditions, muscle TRIB3 also influences energy expenditure and substrate metabolism, indicating that the decrease and increase in muscle TRIB3 under fasting and nutrient excess, respectively, are critical for metabolic homeostasis.

The challenges of achieving optimal glycemic control in type 2 diabetes highlight the need for new therapies. Inappropriately elevated endogenous glucose production (EGP) is the main source of hyperglycemia in type 2 diabetes. Because activation of central ATP-sensitive potassium (KATP) channels suppresses EGP in nondiabetic rodents and humans, this study examined whether type 2 diabetic humans and rodents retain central regulation of EGP. The KATP channel activator diazoxide was administered in a randomized, placebo-controlled crossover design to eight type 2 diabetic subjects and seven age- and BMI-matched healthy control subjects. Comprehensive measures of glucose turnover and insulin sensitivity were performed during euglycemic pancreatic clamp studies following diazoxide and placebo administration. Complementary rodent clamp studies were performed in Zucker Diabetic Fatty rats. In type 2 diabetic subjects, extrapancreatic KATP channel activation with diazoxide under fixed hormonal conditions failed to suppress EGP, whereas matched control subjects demonstrated a 27% reduction in EGP (P = 0.002) with diazoxide. Diazoxide also failed to suppress EGP in diabetic rats. These results suggest that suppression of EGP by central KATP channel activation may be lost in type 2 diabetes. Restoration of central regulation of glucose metabolism could be a promising therapeutic target to reduce hyperglycemia in type 2 diabetes.

Modest elevations in C-reactive protein (CRP) are associated with type 2 diabetes. We previously revealed in mice that increased CRP causes insulin resistance and mice globally deficient in the CRP receptor Fcγ receptor IIB (FcγRIIB) were protected from the disorder. FcγRIIB is expressed in numerous cell types including endothelium and B lymphocytes. Here we investigated how endothelial FcγRIIB influences glucose homeostasis, using mice with elevated CRP expressing or lacking endothelial FcγRIIB. Whereas increased CRP caused insulin resistance in mice expressing endothelial FcγRIIB, mice deficient in the endothelial receptor were protected. The insulin resistance with endothelial FcγRIIB activation was due to impaired skeletal muscle glucose uptake caused by attenuated insulin delivery, and it was associated with blunted endothelial nitric oxide synthase (eNOS) activation in skeletal muscle. In culture, CRP suppressed endothelial cell insulin transcytosis via FcγRIIB activation and eNOS antagonism. Furthermore, in knock-in mice harboring constitutively active eNOS, elevated CRP did not invoke insulin resistance. Collectively these findings reveal that by inhibiting eNOS, endothelial FcγRIIB activation by CRP blunts insulin delivery to skeletal muscle to cause insulin resistance. Thus, a series of mechanisms in endothelium that impairs insulin movement has been identified that may contribute to type 2 diabetes pathogenesis.

Diabetic retinopathy is predominantly caused by vascular endothelial growth factor (VEGF)-induced vascular leakage; however, the underlying mechanism is unclear. Here we designed an in vivo transglutaminase (TGase) activity assay in mouse retina and demonstrated that hyperglycemia induced vascular leakage by activating TGase2 in diabetic retina. VEGF elevated TGase2 activity through sequential elevation of intracellular Ca(2+) and reactive oxygen species (ROS) concentrations in endothelial cells. The TGase inhibitors cystamine and monodansylcadaverin or TGase2 small interfering RNA (siRNA) prevented VEGF-induced stress fiber formation and vascular endothelial (VE)-cadherin disruption, which play a critical role in modulating endothelial permeability. Intravitreal injection of two TGase inhibitors or TGase2 siRNA successfully inhibited hyperglycemia-induced TGase activation and microvascular leakage in the retinas of diabetic mice. C-peptide or ROS scavengers also inhibited TGase activation in diabetic mouse retinas. The role of TGase2 in VEGF-induced vascular leakage was further supported using diabetic TGase2(-/-) mice. Thus, our findings suggest that ROS-mediated activation of TGase2 plays a key role in VEGF-induced vascular leakage by stimulating stress fiber formation and VE-cadherin disruption.