Upregulation of caveolin-1 and its colocalization with cytokine receptors contributes to beta cell apoptosis

Caveolin-1 (cav-1), the principal structural and signalling protein of caveolae, is implicated in various signalling events, including apoptotic cell death in type 2 diabetes. However, the precise role of beta cells in apoptosis has not been clearly defined. In this study, we investigated the involvement of cav-1 in cytokine-induced beta cell apoptosis and its underlying mechanisms in the rat beta cell line, INS-1 and isolated islets. Treatment of cytokine mixture (CM, TNFα + IL-1β) significantly increased the mRNA and protein expression of cav-1, and resulting in increased formation of caveolae. We found that IL-1 receptor 1 and TNF receptor localized to plasma membrane lipid rafts in the control cells and CM treatment recruited these receptors to the caveolae domain. After cav-1 siRNA transfection, CM-dependent NF-κB activation was reduced and consequently downregulated the mRNA expression of iNOS and IL-1β. Finally, decreased cell viability by CM treatment was ameliorated in both INS-1 cells and isolated islets treated with cav-1 siRNA. These results suggest that increased cav-1 expression and recruitment of cytokine receptors into caveolae contribute to CM-induced beta cell apoptosis.


Cytokine mixture treatment increases recruitment of cytokine receptors into caveolae in INS-1 cells.
To investigate whether cav-1 affected the activation of NF-κB signalling, we analysed the localization of IL-1R1 and TNFR1 by CM stimulation. Control and CM treated cells were fractionated by sucrose gradient density centrifugation in the absence of detergent, a procedure widely used to isolated caveolae-enriched membrane domain 20 . Western blot analysis of the fractions showed that the majority of the cav-1 was present at the low density 5~30% sucrose interface of the gradient (Fig. 2b, . Protein amounts in each fraction showed that the CM treated cells contained higher levels of protein in the caveoale domain. (Fig. 2a). A significant proportion of IL-1R1 and TNFR1 was recruited to the caveolae fractions in CM treated cells, whereas flotillin-1, a membrane lipid raft marker, was similar in control and CM treated INS-1 cells. When control and CM-treated cells were co-immunostained with anti-IL1R1, anti-TNFR1 and anti-cav-1 antibodies, the receptors were found to be colocalized with cav-1 in the membrane of CM treated cells compared with control cells (Fig. 2c). Mean fluorescent intensity was calculated and showed a significant increase in IL-1R1 and TNFR1 localization with cav-1 protein in CM treated INS-1 cells compared with control (Fig. 2d).
Down-regulation of cav-1 expression protects against cytokine mixture-induced beta cell apoptosis. Given that recruitment of cytokine receptors into caveolae was increased by CM treatment, we investigated whether reduction of cav-1 expression could protect against beta cell apoptosis induced by CM. First, we downregulated cav-1 expression using 21-nucleotide siRNA at concentrations of 5, 10, 25, and 50 μM and found a marked reduction in cav-1 mRNA after 24 h of 10 μM siRNA transfection (Fig. 3a). We confirmed that protein expression of cav-1 was also reduced in cells treated with 10 μM of cav-1 siRNA, compared with control siRNA treated cells (Fig. 3b,c). To investigate whether decreased cav-1 expression would affect the levels of beta cell apoptosis, we transfected INS-1 cells with cav-1 siRNA in the presence or absence CM, and measured the level of apoptotic cells via annexin-V staining followed by flow cytometry analysis. As shown in Fig. 3d,e, the percentage of annexin-V-FITC positive cells increased during CM treatment, but was significantly reduced in cav-1 siRNA transfected cells (Fig. 3d,e). Moreover, cav-1 siRNA transfected cells had significantly higher cell viability upon CM treatment compared to control siRNA transfected cells (Fig. 4a). To confirm that downregulation of cav-1 affected CM-induced cell viability in islets, we transfected cav-1 siRNA into isolated islet cells from mice www.nature.com/scientificreports www.nature.com/scientificreports/ and rats. As shown in Fig. 4b,c, CM treatment significantly reduced cell viability compared to control cells treated with control siRNA and the reduction was attenuated by cav-1 siRNA in CM treated cells (Fig. 4b,c). In contrast, cells overexpressing cav-1 were more sensitive to CM treatment, and cell viability was dramatically decreased in cells transfected with pcDNA-cav-1 compared with cells transfected with control vector (Supplementary Fig. 3).

Down-regulation of cav-1 expression inhibits cytokine mixture-induced NF-κB activation.
To investigate whether CM-induced activation of NF-κB was affected by cav-1 expression, we analysed the expression levels of phosphorylated IKKα, IKKβ, and IκBα in cav-1 siRNA treated cells. As shown in Fig. 5a,b, the levels of phosphorylated IKKα, IKKβ, and IκBα by CM treatment were significantly reduced in cav-1 downregulated cells compared with control siRNA treated cells. Moreover, cav-1 downregulation significantly attenuated CM-induced NF-κB p65 phosphorylation in the nucleus (Fig. 5a,b). To investigate whether downstream target of NF-κB activation was affected by cav-1 depletion, mRNA expression level of IL-1β and inducible nitric oxide synthase (iNOS) 21 was examined in cav-1 siRNA treated cells with or without CM. We found that mRNA expression of these genes were significantly increased after 6 h treatment of CM and the levels were reduced by cav-1 downregulated cells (Fig. 5c).

Down-regulation of cav-1 expression increases glucose-stimulated insulin secretion.
Since previous reports have shown that inhibition of cav-1 expression enhances insulin secretion 13 , we measured insulin secretion in the presence of 3 mM and 17 mM glucose after transfection with control and cav-1 siRNA in INS-1 cells. Insulin secretion in cells transfected with control siRNA was increased approximately 2.5-fold upon treatment with 17 mM glucose compared to treatment with 3 mM glucose. Moreover, transfection with cav-1 siRNA significantly increased insulin secretion upon treatment with both 3 mM and 17 mM glucose compared with control cells (Supplementary Fig. 4).

Discussion
In this study, we elucidated the role of cav-1 in cytokine-induced beta cell apoptosis and also attempted to determine its underlying mechanisms in the INS-1 cell line. Prolonged exposure to pro-inflammatory cytokines, particularly the combination of IL-1β and TNFα, induced pancreatic beta cell apoptosis, which is one of the major causes of islet inflammation during the development of diabetes. Activation of NF-κB via phosphorylated IKKα/IKKβ and NF-κB p65 translocation into the nucleus is essential for the progressive loss of beta cells in cytokine-induced diabetes, and the inhibition of this process could be an effective strategy for beta cell protection 16,17,22,23 . As previously reported, there are many potential mechanisms of cytokine-induced beta cell death 1 , and receptor-mediated signalling strongly contributes to the activation of intracellular death signals, including NF-κB signalling 24 .
We found that expression levels of IL-1R1 and TNFR1 were increased by cytokine treatment. A study by Boni-Schnetzler et al. demonstrated that IL-1R1 expression is induced by palmitate and blocking IL-1R strongly inhibited proinflammatory factors stimulated by free fatty acids in human and mouse islets 25 . Moreover, IL-1R antagonist improves glycemia and beta cell secretory function in patients with type 2 diabetes 26 . TNF-induced cell death was also mediated by TNFR1 27 , and anti-TNF therapeutics are currently used to treat inflammatory diseases such as rheumatoid arthritis and Crohn's disease 28 . These results suggest that increased expression and activation of these receptors may play an important role in beta cell inflammation during the development of diabetes.
Cav-1 plays an important role in metabolic signalling and is involved in glucose and/or cholesterol homeostasis, vesicle transport, proliferation, and apoptosis 12 . Catalan et al. demonstrated that expression of cav-1 mRNA is elevated in both visceral and subcutaneous adipose tissue of obese type 2 diabetic patients compared to lean controls 29 and Wehinger et al. reported that overexpression of cav-1 in beta cells promoted apoptosis 14 . In the present study, CM treatment increased the number of caveolae structures, as well as the mRNA and protein expression levels of cav-1. These results suggest that upregulation of cav-1 plays an important role in transmitting signals from the cell surface via intracellular signalling pathways that regulate inflammation and type 2 diabetes.
Cav-1 is known to interact with many caveolae-localized signalling molecules via caveolin scaffolding domain (CSD) and an aromatic-rich caveolin binding motif (CBM) on the associated proteins 30 . Members of the TNF receptor subfamily contain a CBM motif and IL-1R1 effectors (IRAK, TRAF6, and IκB) are recruited to cav-1-containing lipid rafts 31,32 . The finding that cytokine signalling resulted in the recruitment of two receptors in caveolae on the plasma membrane is consistent with previous studies, and suggests that cav-1 dependent lipid rafts are critical for cytokine activation.
We found that inhibition of cav-1 reduced CM-induced NF-κB activation and lead to a decrease in mRNA expression of target genes (IL-1β and iNOS). Oakley et al. reported that IL-1R mediated NF-κB signalling was dependent on cav-1, as evidenced by the fact that inhibiting lipid raft formation reduced IL-1β-mediated NF-κB activation 31 . Garrean et al. also observed a significant reduction in LPS-induced IκBα degradation and NF-κB activation in Cav-1 −/− lungs relative to wild type mice 33 . These findings, in addition to those reported here, support a role for cav-1 in NF-κB activation in beta cells.
Recently, it was demonstrated that inhibition of cav-1 expression protected beta cell death induced by palmitate. Wehinger et al. found that over expression of cav-1 in MIN-6 cells promoted palmitate-induced beta cell apoptosis and implicated Src family kinase-mediated tyrosine phosphorylation in this pathway 14 14 , and this phosphorylation is associated with enhanced sensitivity in response to various cytotoxic stimuli 18,35 . In this study, we confirmed that CM also phosphorylated tyrosine-14 of cav-1, and our results suggested that cav-1 phosphorylation on tyrosine-14 augmented sensitivity to CM in INS-1 cells, restoration of CM induced beta-cell apoptosis might be also caused by loss of cav-1 phosphorylation site tyrosine-14.
It was reported that cav-1 depletion in vitro and in vivo results in insulin secretion. When unstimulated condition (low glucose level), cav-1 bound to insulin granule proteins including cdc42, guanosine 5′-triphosphate and vesicle associated membrane protein 2, but stimulation with glucose induced the dissociation of cav-1 from insulin granules and promoted insulin secretion 13 . Additionally, cav-1-deficient mice had higher plasma insulin levels and postprandial hyperinsulinemia under fasting or high-fat diet conditions 11 . Moreover, Wen et al. also demonstrated that cav-1 silencing significantly enhanced insulin production and secretion 34 . Therefore, these results demonstrate that downregulation of cav-1 could protect beta cells from CM treatment via reduced apoptosis and induced insulin secretion.
Previous study demonstrated that cav-1 deficiency mice showed lean phenotype and developed insulin resistance 11 , suggesting that the plasma insulin levels are affected by insulin action in peripheral tissues as well as insulin secretion in beta-cells. Although cav-1 deficiency in INS-1 cells results in increase insulin secretion compared with control cells under physiological glucose concentration, but whether hypoglycemia occurs and physiological role of cav-1 in insulin secretion in in vivo will be investigated in beta cell specific cav-1 KO mice.
In summary, we proposed a schematic mechanism (Fig. 6) in which cav-1 is involved CM-mediated beta cell apoptosis. Increased expression of cav-1 and caveolae structure was observed in CM-treated cells and recruitment of cytokine receptors into caveolae contributed to CM-induced beta cell apoptosis. Moreover, silencing cav-1 Figure 5. Down-regulation of cav-1 expression inhibits cytokine mixture-induced NF-κB activation. (a) INS-1 cells were transfected with CON siRNA or Cav-1 siRNA for 24 h, followed by treatment with the cytokine mixture (CM; IL-1β 20 ng/ml, TNFα 20 ng/ml). After 24 h, cell lysates were subjected to western blot analysis with specific antibodies. Actin and Lamin B1 were used as cytosolic and nuclear loading controls, respectively. (b) The densities of western blot signals were measured, and the relative expression level was normalized to that of actin, Lamin B1, and the non-phosphorylated form (IKK, IκB, and NF-κB). (c) Cells were treated as described in (a) and mRNA levels of IL-1β and iNOS was determined after 6 h incubation with the CM. Values are means ± SEM from triplicate experiments, *p < 0.05 vs. CON siRNA treated with CON, # p < 0.05 vs. CON siRNA treated with CM.
Transmission electron microscopy (TEM) analysis. Cells (1 × 10 6 ) were fixed in 4% paraformaldehyde and then in 1% osmium tetroxide. Samples were dehydrated via ethanol grade series, infiltrated with propylene oxide, and embedded with Epoxy resin (Poly bed 812 kit; Polysciences, Inc., Warrington, PA, USA). Embedded samples were cut into 65 nm-thick sections and stained with uranyl acetate and lead citrate. Samples were imaged using transmission electron microscopy (TEM, Philips CM200; Field Emission Instruments, USA), and images were acquired using XR41B CCD camera (Advanced Microscopy Techniques, MA, USA) Sodium carbonate extraction and sucrose density gradient fractionation of caveolae. Experiments were carried out following the detergent-free protocol developed by Song KS et al. 20 . Briefly, INS-1 cells were washed with ice-cold phosphate-buffered saline, trypsinised and homogenised. Sonicated cell samples were mixed with equal volumes of 80% sucrose solution in MES buffered saline (25 mM MES, pH 6.5, 0.15 M NaCl), placed in an ultracentrifuge tube, and overlaid with 4 ml of 30% sucrose and 4 ml of 5% sucrose in MES-buffered saline containing 0.25 M Na 2 CO 3 . Gradients were generated by centrifugation at 200,000 × g for 18 h in a SW41 rotor (Beckman Coulter, INC., Atlanta, USA). Fractionations were collected from the top of the gradient and dissolved in 1 × Laemmli SDS sample buffer prior to western blot analysis.
Western blotting. Cells were lysed in mammalian protein extraction buffer (GE Healthcare, Milwaukee, WI, USA). Nuclear and cytoplasmic proteins were extracted according to the NE-PER TM Nuclear and Cytoplasmic Extraction Reagents manufacturer's instructions (Thermo Fisher Scientific, Madison, WI, USA). Thirty micrograms of protein samples were separated by SDS-PAGE, transferred to nitrocellulose membranes, and incubated with specific antibodies. The following antibodies were used at the dilution indicated: anti-cav-1, anti-IL-1R1, anti-TNFR, anti-IKKα, anti-IKKβ, anti-p-IKKα/β, anti-IκB, anti-p-IκB, anti-NFκB p65, anti-p-NFκB p65 (1:1000; Cell Signalling Technology, Boston, MA, USA), and anti-β-actin (1:10,000; Santa Cruz Biotechnology, Santa Cruz, CA, USA). After washing, the membranes were incubated with a secondary antibody conjugated with Figure 6. Schematic of the mechanism by which involvement of cav-1 and caveolae in CM-induced beta cell apoptosis in pancreatic beta-cells. Cytokine mixture treatment into beta cells inhibited insulin secretion and induced apoptosis. Cytokine mixture treatment increased caveolae structure as well as cav-1 expression and cytokine receptors (TNFR1 and IL1-R1) were recruited into caveolae. Therefore, activation of NF-kB signaling pathway increased the expression level of inflammatory response genes, which leads to beta cell apoptosis.
Immunofluorescence microscopy. Cell cultured on glass coverslips were fixed with 4% formaldehyde and permeabilized with PBS containing 0.5% triton X-100 and blocked with 2% BSA in PBS for 1 h. They were then incubated with mouse monoclonal antibody against cav-1, rabbit polyclonal antibody against IL-1R1 and TNFR1 overnight. After washing the coverslips were incubated with Alexa Fluor 594 TM -conjugated anti-rabbit, DyLight TM 488-conjugated anti-mouse secondary antibodies. The labelled cells were observed under a confocal microscope.
Transfection. For  Animals and islet isolation. Four-week-old male C57BL/6 J mice and 6 week old male Sprague-Dawley (SD) rat were purchased from Korea Research Institute of Bioscience & Biotechnology (KRIBB, Daejeon, South Korea) and Daehan Biolink (Daehan Biolink Co. LTD., Eumsung, South Korea). All animals were kept in the animal facility for 7 d to allow for environmental adaptation. The mice were fed standard chow for 2 wk and were then sacrificed for islet isolation. The collagenase digestion technique was used as previously described 36 , and islets were separated by centrifugation on Histopaque-1077 (Sigma-Aldrich, St. Louis, MO, USA). Healthy islets of appropriate sizes were individually-picked under a stereomicroscope. The islets were dissociated into single cells by trypsinisation. All experiments were approved by the Institutional Animal Care and Use Committee of Eulji University (EUIACUC18-7) and confirmed that all experiments were performed in accordance with relevant guidelines and regulations.
Glucose-stimulated insulin secretion assay. INS-1 cells were plated in 24-well plates and transfected with either control siRNA and cav-1 siRNA for 24 h. Cells were starved in Krebs-Ringer bicarbonate buffer without glucose for 2 h and insulin secretion was stimulated by treatment of cells with 3 mM or 17 mM glucose for 1 h. At the end of the incubation, the amount of insulin released into the supernatant was quantified using a rat insulin EIA kit (Alpco Diagnostics, Windham, NH, USA), according to the manufacturer's instructions. The amount of insulin released was normalized to the total amount of protein.

Statistical analysis.
Results are expressed as mean ± SEM of three separate experiments. Shapiro-Wilk normality test was performed in SPSS (Statistical Analysis System Institute 2010, IBM corp., Chicago, IL, USA) and differences between the control group and the treated group were assessed via the Student's t test. Analysis of variance (ANOVA), followed by Scheffe's multiple comparison test, were used to determine the significance of any differences among more than two groups. P < 0.05 was considered significant.