Effect of Exendin-4 on Autophagy Clearance in Beta Cell of Rats with Tacrolimus-induced Diabetes Mellitus

Growing evidence suggests that GLP-1 protects beta cells against various cellular injuries by modulating autophagy. In this study, we examined whether exendin-4 (Ex-4), a GLP-1 analog, had preventive effects on tacrolimus (Tac)-induced beta cell injury by improving autophagy clearance. Rats with Tac-induced diabetes mellitus exhibited increased autophagy-associated protein expression, light chain 3B levels, and autophagic vacuole numbers in pancreatic beta cells. Additionally, Tac increased autophagy in a dose- and time-dependent manner in vitro, and inhibition of autophagosome using 3-methyladenine reduced Tac-induced islet cell injury by decreasing reactive oxygen species production and apoptosis. Ex-4 treatment decreased Tac-induced hyperglycaemia, oxidative stress, and apoptosis, accompanied by decreased autophagy-associated protein expression and autophagosome numbers. In vivo and in vitro studies showed that Tac treatment impaired lysosomal function and autophagosome-lysosome fusion; these processes were improve by Ex-4 treatment. Moreover, addition of bafilomycin A1, an inhibitor of lysosomal function, abolished the protective effects of Ex-4. Our findings reveal that Tac-induced diabetes mellitus was a state of excessive burden of autophagosomes and impairment of autophagy clearance and that Ex-4 protected against Tac-induced pancreatic islet injury by reducing the burden of autophagosomes via activation of autophagosome clearance. Thus, Ex-4 had therapeutic effects on Tac-induced pancreatic beta cell injury.

by modulation of autophagy. Using experimental rats and INS-1 cells, we studied the status of autophagy in Tac-induced pancreatic islet cell injury and the effects of Ex-4 on modulating Tac-induced autophagy. The results of our study clearly demonstrated the rationale for use of Ex-4 in the management of Tac-induced diabetes mellitus based on autophagy regulation.

Results
Tac-induced diabetes mellitus, oxidative stress, and apoptosis were suppressed by Ex-4 treatment. Rats treated with Tac exhibited symptoms of diabetes mellitus, as demonstrated by increased area under the curve of glucose (AUCg) values and lower levels of serum insulin as compared with those in rats treated with vehicle (VH). Concomitant treatment with Ex-4 reversed these changes (Fig. 1a-c and Table 1). Morphologically, islets treated with Tac showed lower immunoreactivity for insulin and decreased numbers of beta cells compared with those in rats treated with VH; these effects were reversed by Ex-4 (Fig. 1d,e). The number of alpha cells identified by the ratio of glucagon-positive cells to insulin-positive cells was markedly increased in the Tac group but not in the Tac plus Ex-4 group (Fig. 1f). Oxidative stress, as measured using 8-hydroxy-2′ -deoxyguanosine , was higher in pancreatic islet tissue sections and serum from rats in the Tac group than that from rats in the VH group; however, Ex-4 treatment attenuated these changes ( Fig. 1g-i). Apoptosis, as measured using terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) assays and enumeration of active caspase-3-positive cells, was markedly increased in the Tac group as compared with those in the VH and VH plus Ex-4 groups; notably, however, Ex-4 treatment reduced both parameters (Fig. 1g,j,k).
Ex-4 reduced Tac-induced autophagy substrate levels and autophagosome formation. Next, we examined alterations in autophagy substrate (ubiquitin protein aggregate and p62) levels and autophagosome formation (as measured by the expression of microtubule-associated protein 1 light chain 3 beta [LC3B]) in tissue sections from experimental rats. In the Tac group, confocal microscopy revealed that increased numbers of ubiquitin-, p62-, and LC3B-positive cells coincided with insulin positivity, indicating the presence of these markers in beta cells (Fig. 2a-c). Moreover, the numbers of ubiquitin-, p62-, and LC3B-positive cells were significantly higher in the Tac group than in the VH group. However, Ex-4 attenuated these changes ( Fig. 2d-f). High-magnification images revealed that ubiquitin was colocalized with LC3B in a punctate manner in beta cells from rats in the Tac group (Fig. 2g). Thus, to determine whether this effect was dependent on time or Tac concentration, we observed the expression of ubiquitin, p62, and LC3B in cultured INS-1 cells. Interestingly, the expression levels of these proteins were increased with Tac treatment in a time-and concentration-dependent manner (Fig. 3a,b). Addition of Ex-4 attenuated these increases in a concentration-dependent manner during Tac treatment (Fig. 3c).

Autophagy was observed in pancreatic beta cells in rats with Tac-induced diabetes mellitus.
To visualise the induction of autophagy in experimental animals, we performed transmission electron microscopy (TEM) on islets (Fig. 4). Beta cells from Tac-treated rats contained markedly reduced numbers of insulin granules compared with those in the VH group (Fig. 4a,b). High-magnification TEM images (Fig. 4c) showed double membrane-bound vacuoles containing intact cytoplasmic organelles (suggestive of autophagosomes) and single membrane-bound vacuoles containing degraded cytoplasmic organelles (suggestive of autolysosomes). Counting and measuring the areas of autophagic vacuoles/vesicles in micrographs also indicated that there was a significant increase in autophagy in cells from the Tac group (Fig. 4d,e). Combined treatment with Ex-4 and Tac attenuated the formation of autophagic vacuoles and induced recovery of the number of insulin granules.

Tac-induced pancreatic beta cell injury was associated with autophagy. To determine whether
Tac-induced beta cell injury was mediated through autophagy, we treated INS-1 cells with the autophagy inhibitor 3-methyladenine (3-MA). The effects of autophagy inhibition were evaluated by measuring changes in fluorescence levels of 2′ ,7′ -dichlorodihydrofluorescein diacetate (H 2 DCF-DA, for detecting reactive oxygen species [ROS] production) and annexin V (for detecting apoptosis). 3-MA, which inhibits class III phosphatidylinositol 3-kinase to block autophagosome formation at the early stage of autophagy, significantly decreased the percentages of H 2 DCF-DA-and annexin V-positive Tac-treated INS-1 cells (Fig. 5). These data suggested that Tac-induced pancreatic beta cell injury resulted in a state of excessive autophagosome formation and that autophagy was responsible for Tac-induced beta cell injury.

Ex-4 reversed Tac-induced impaired autophagy clearance by improving lysosomal function.
Next, we evaluated lysosomal function to determine whether Tac-induced pancreatic islet injury was associated with autophagy clearance. First, we examined lysosomal protein expression in experimental animals. Tac treatment significantly decreased lysosomal-associated membrane protein 2A (LAMP-2A) expression within beta cells, as shown in Fig. 6a; however, treatment with Ex-4 attenuated these effects. Next, we measured the activation of lysosomal function using INS-1 cells under combined Tac plus Ex-4 treatment. Cells were stained with LysoTracker and analyzed using flow cytometry and confocal imaging (Fig. 6b). Tac-treatment significantly reduced the fluorescence intensity, whereas Ex-4 treatment recovered the fluorescence intensity to that of control cells. Similar patterns were observed for cathepsin B and LAMP-2A (Fig. 6c,d). Bafilomycin A1 (BA) was used as a negative control of lysosomal function and was shown to abolish lysosomal acidification.

Ex-4 enhanced autophagosome-lysosome fusion in Tac-induced pancreatic beta cell injury.
To examine autophagosome-lysosome fusion in pancreatic beta cells, we examined the colocalization of LC3B (an autophagosome marker), LAMP-2A (a lysosome marker), and insulin (a pancreatic beta cell marker) in tissue sections from experimental rats. Punctate LC3B expression in beta cells was obviously observed in the Tac group, and the combined treatment with Ex-4 attenuated this change. In contrast, LAMP-2A expression was decreased The data are presented as means ± standard errors (SEs). n = 8 per group. # P < 0.05 versus the VH or VH plus Ex-4 groups; $ P < 0.05 versus the Tac group. Scale bar = 50 μ m.
in the Tac group compared with that in the VH group, but was recovered in the Tac plus Ex-4 group (Fig. 7a). Moreover, combined treatment with Ex-4 markedly enhanced LC3B localization to lysosomes (Fig. 7b). Next, we treated INS-1 cells with Tac and/or Ex-4 after transfection with DDK-tagged LC3B and then performed double immunofluorescence to detect DDK and LAMP-2A. Similar to the results in experimental rats, Ex-4 reduced LC3B punctate expression and enhanced colocalization of LC3B and LAMP-2A as compared with that in the Tac group ( Fig. 8a-c). We used BA as a negative control for autophagosome-lysosome fusion. Addition of BA markedly induced accumulation of LC3B and suppressed the expression of LAMP2A.
The protective effects of Ex-4 were causally related to lysosomal function. To determine whether the protective effects of Ex-4 on Tac-induced injury were mediated through improved lysosomal function, INS-1 cells were treated with BA, and the effects of BA were evaluated by measuring changes in H 2 DCF-DA and annexin V. Inhibition of lysosomes by BA in Tac plus Ex-4 in INS-1 cells significantly increased the percentages of H 2 DCF-DA-and annexin V-positive cells as compared with that in Tac plus Ex-4 cells not treated with BA ( Fig. 9).

Discussion
In this study, we investigated the effects of Ex-4 on autophagy in the context of Tac-induced pancreatic beta cell injury. The results of our study clearly demonstrated that Tac treatment increased autophagy substrate expression and autophagosome formation. However, addition of Ex-4 reduced these alterations by improving lysosomal dysfunction and the frequency of autolysosome-lysosome fusion. These findings provide a rationale for use of Ex-4 in the management of Tac-induced diabetes mellitus in clinical practice.
To date, few studies have examined autophagy in dysfunctional beta cells of rats with Tac-induced diabetes mellitus. Therefore, in this study, we examined markers of autophagy substrates and autophagosomes in rats with Tac-induced diabetes mellitus. In this study, we used p62 and ubiquitinated protein aggregates as markers of autophagy substrates based on previous reports showing that such toxic aggregates are increased in neurodegenerative diseases, such Huntington's disease, and in pancreatic beta cell-specific autophagy-deficient mice 10,11 . Pancreatic beta cells are known to bear a high burden of protein synthesis and folding. Therefore, the disposal of misfolded or denatured proteins is important, and the autophagosome/lysosome system plays a key role in this process 12 . Here, we found that Tac treatment increased the levels of p62 and ubiquitin-positive beta cells and LC3B. Furthermore, TEM observations revealed abundant autophagic vacuoles accompanying reduced numbers of insulin-bearing vesicles. These findings suggested that Tac-induced diabetes mellitus was characterised by overloaded protein aggregates and the accumulation of large amounts of autophagic vacuoles.
In addition, to clarify the casual relationship between Tac-induced beta cell injury and autophagy, we evaluated the effects of an autophagy inhibitor in cultured INS-1 cells (Fig. 5). Our findings showed that blocking autophagosome formation by 3-MA significantly decreased oxidative stress and apoptosis caused by Tac, suggesting that excessive autophagosome formation was responsible for Tac-induced oxidative stress and apoptotic cell death.
Next, we evaluated autophagy clearance by measuring lysosomal function because lysosomes are the main organelles involved in processing autophagosomes. From our study, we found that Tac treatment impaired lysosomal function, as demonstrated by the increased lysosomal pH and reduced hydrolase enzyme activity. Subsequently, damaged lysosomes could not fuse with autophagosomes, as demonstrated by reduced colocalization of markers for lysosomes and autophagosomes. Combined treatment with Ex-4 significantly reversed Tac-induced lysosomal defects, e.g., increased LAMP-2A and cathepsin B activity, reduced lysosomal pH, and decreased the autophagosome-lysosome fusion rate. Thus, we suggest that lysosomal impairment may be involved in Tac-induced pancreatic beta cell injury and that Ex-4 may improve the clearance of Tac-induced autophagosomes by enhancing lysosomal activity.
In this study, we also found that BA treatment increased oxidative stress and apoptotic cell death as compared with that in the Tac plus Ex-4 group without BA treatment, suggesting that the protective effects of Ex-4 on Tac-induced pancreatic beta cells were closely associated with lysosomal function. Thus, we cautiously concluded that Tac-induced pancreatic beta cell injury may be causally associated with lysosomal dysfunction. However, we found that Ex-4 had protective effects, even in the presence of BA. Ex-4 was still able to "rescue" the phenotype of Tac treatment, including maintaining activity of cathepsin B and LAMP-2A. The reason for this protective effect of Ex-4 in the state of lysosomal inhibition was unclear. However, it is possible that Ex-4 may act on a step of the autophagy process well before lysosomal acidification and may also affect lysosomal acidification. Alternatively, Ex-4 may exert protective effects via direct antioxidative and anti-apoptotic effects or via an AMP-dependent protein kinase (AMPK)-dependent mechanism, which could affect beta cell survival 13,14 .
The role of GLP-1 analogs in autophagy is unclear. Liraglutide (LRG), another GLP-1 analog, inhibits free fatty acid (FFA)-induced cell injury in hepatocytes or pancreatic beta cells by activating autophagy 15,16 . LRG has also been shown to improve cardiac function in mice fed a high-fat diet (HFD) by increasing autophagy 17 . On the other hand, consistent with our results, LRG has been shown to reduce LC3 levels, autophagosome formation, and apoptosis following chronic exposure of renal tubular epithelial cells to high glucose (HG) 18 . These contradictory findings suggest that GLP-1 analogs may act differently on autophagy according to the intensity and duration of stress rather than different mechanisms of action. FFA and HFD are relatively weak stresses compared with Tac or HG. Therefore, activation of autophagy is required to prevent FFA-or HFD-induced cell injury. However, Tac and HG induce a state of excessive autophagosome formation, and reduced autophagy is needed to prevent the cell injury. In this situation, GLP-1 analogs may exert their protective effects by activating autophagy or clearing autophagosomes according to the intensity and duration of stress.
The effects of Ex-4 on autophagy in Tac-induced pancreatic beta cell injury are summarized in Fig. 10. Tac treatment augments autophagosome formation in pancreatic beta cells to adapt to stress. However, autophagosomes are not effectively degraded due to impaired autophagy clearance (lysosomal dysfunction and autophagosome fusion with lysosomes). Therefore, excessive autophagosome formation leads to apoptotic cell death. In this process, Ex-4 improves autophagic clearance rate by enhancing lysosomal function and autophagosome fusion with lysosomes. Through this mechanism, Ex-4 may protect against Tac-induced pancreatic beta cell dysfunction.
In conclusion, impaired lysosome-associated autophagic degradation or clearance is an important element in the pathogenesis of Tac-induced pancreatic beta cell injury, and Ex-4 may be useful for management of Tac-induced diabetes mellitus by modulating autophagy. were housed in a controlled-light environment at the animal facility of the Catholic University of Korea. The Animal Care and Use Committee of the Catholic University of Korea approved the experimental protocol (CUMC-2012-0117-04), and all procedures performed in this study were in accordance with ethical guidelines for animal studies. Animals were maintained on a 12/12-h light/dark cycle, with a 0.05% salt diet (Research Diets, New Brunswick, NJ, USA) and water ad libitum. After acclimatization for 1 week, weight-matched rats were randomised to four groups (n = 8 each) and were treated daily with a subcutaneous (s.c.) injection of 1.5 mg/kg Tac (Prograft; Astellas Pharma, Ibaraki, Japan) or 1 mL/kg VH (olive oil; Sigma-Aldrich, St. Louis, MO, USA) with or without Ex-4 (1 μ g/kg, intraperitoneal [i.p.] injection; Sigma-Aldrich) for 4 weeks. Routes of administration and doses of drugs were chosen based on previous studies 19,20 . Rats were pair-fed, and their body weights were monitored daily. After the 4-week treatment, animals were housed individually in metabolic cages (Tecniplast, Buguggiate, Italy) for the measurement of urine volume over 24 h. The following day, animals were anaesthetised, and blood samples and tissue specimens were obtained for further analysis. Serum Scr levels, blood urea nitrogen (BUN) levels, and trough levels of Tac in the blood were measured as described previously 20 . Pancreatic function and islet size. An intraperitoneal glucose tolerance test (IPGTT) was performed at the end of the 4 weeks of treatment. Briefly, after 1 day of fasting, 50% dextrose (1.5 g/kg) was injected, and the blood glucose concentration was measured just before and at 30, 60, 90, and 120 min after the injection using a glucose analyser (Accu-Check; Roche Diagnostics, Basel, Switzerland). The AUCg was calculated by trapezoidal estimation from the values obtained in the IPGTT. The fasted serum insulin level was measured using a sandwich enzyme-linked immunosorbent assay (ELISA; Millipore Corp., St. Charles, MO, USA). For the quantification of islet size, the numbers of beta cells or alpha cells were assessed in each using quantification of captured images from immunohistochemistry with insulin or glucagon (TDI Scope Eye Version 3.6 for Windows; Seoul, Korea). Insulin-or glucagon-positive cells were evaluated by counting approximately 20 randomly selected nonoverlapping islets for the eight animals in each group.       Tac treatment augments autophagosome formation in pancreatic beta cells to adapt to stress and lysosomal dysfunction. However, autophagosomes are not effectively degraded due to impaired autophagy clearance (lysosomal dysfunction and autophagosome fusion with lysosomes). Therefore, excessive autophagosomes lead to apoptotic cell death. In this process, Ex-4 improves the autophagic clearance rate by enhancing lysosomal function and autophagosome fusion with lysosomes.
in each group. Quantitative analysis was performed by calculating the percent positive area showing the same intensity per islet using histogram equalization (TDI Scope Eye). Immunoblot analysis. INS-1 cells were plated in 96-well plates at a density of 25,000 cells/well in complete RPMI medium. After 24 h, the cells were treated with serum-free RPMI medium containing Tac (5-20 μ g/mL) and Ex-4 (0.1-100 nM) with or without BA. At the indicated times, INS-1 cells were lysed in 10 mM Tris (pH 7.5) containing 1% sodium dodecyl sulphate (SDS) and 1 mM NaVO 4 . Equal amounts of proteins were resolved using SDS-polyacrylamide gel electrophoresis and electroblotted onto polyvinylidene fluoride membranes (Millipore Corp.). The membranes were incubated for 1 h at room temperature in blocking solution containing 5% nonfat milk (BD Difco, Sparks, MD, USA) in PBS, followed by overnight incubation at 4 °C with the primary antibodies diluted in Signal Boost immunoreaction enhancer (Millipore Corp.). The next day, the blots were incubated for 1 h with peroxidase-conjugated secondary antibodies (Cell Signaling Technologies). Antibody binding was detected with commercial chemiluminescence kits (ATTO Corp., Tokyo, Japan). Quantification was performed using relative densities, with the control group set at 100%; densities were normalised to that of the beta-actin bands from the same gel (Quantity One version 4.4.0; Bio-Rad, Hercules, CA, USA).

TEM.
After fixation in 2.5% glutaraldehyde in 0.1 M phosphate buffer, pancreatic tissues were postfixed with 1% O S O 4 and embedded in Epon 812. Ultrathin sections were cut, stained with uranyl acetate/lead citrate, and photographed with a JEM-1200EX transmission electron microscope (JEOL Ltd., Tokyo, Japan). Sections were scanned randomly at 20 different spots per sample at 5000× magnification. The numbers and areas of autophagic vacuoles (autophagosomes and autolysosomes) per cell in the scanned areas were measured using imaging software (TDI Scope Eye).
LysoTracker staining. One day after cell seeding, the cells were treated with serum-free RPMI medium containing Tac (20 μ g/mL) and Ex-4 (10 nM) in the presence or absence of BA (5 nM Apoptosis. One day after cell seeding, the cells were treated with serum-free RPMI medium containing Tac (20 μ g/mL) and Ex-4 (10 nM) in the presence or absence of 3-MA (10 mM) or BA (5 nM) for 12 h. Trypsinised INS-1 cells were treated with 5 μ L of fluorescein isothiocyanate (FITC)-conjugated annexin V (BD Biosciences) in 1× binding buffer (BD Biosciences) for 15 min at room temperature according to the manufacturer's protocol. The stained cells were analyzed using flow cytometry on a FACSCalibur instrument (BD Biosciences). Values are expressed as the percentage of fluorescent cells relative to the total cell count. Intracellular ROS generation. One day after cell seeding, the cells were treated with serum-free RPMI medium containing Tac (20 μ g/mL) and Ex-4 (10 nM) in the presence or absence of 3-MA (10 mM) or BA (5 nM). After 12 h, cells were incubated with 10 μ M H 2 DCF-DA (Molecular Probes) for 1 h at 37 °C according the manufacturer's instructions and then analyzed using a FACSCalibur (BD Biosciences). Values are expressed as the percentage of fluorescent cells relative to the total cell count.
Statistical analysis. The data are expressed as the mean ± standard error (SE) of at least three independent experiments. Multiple comparisons between groups were performed by one-way analysis of variance (ANOVA) with Bonferroni post-hoc tests using IBM SPSS Statistics (Version 19.0; IBM Corp., Armonk, NY, USA). Results with P values of less than 0.05 were considered significant.