Exosomes from β-cells alleviated hyperglycemia and enhanced angiogenesis in islets of streptozotocin-induced diabetic mice

Introduction

Diabetes is a major chronic disease and brings a heavy burden to individuals, families and societies. According to IDF reports, 1 in 11 adults has diabetes globally, with a total number of up to 425 million.¹ The chronic complications jeopardize the quality of life and eventually lead to high morbidity and mortality of diabetic patients.^2,3 The failure and dysfunction of pancreatic islet β-cells is the core pathogenesis of both type 1 diabetes and type 2 diabetes.^4,5 Secretion of enough insulin from β-cells is essential for glucose homeostasis in body. Based on the finding from UKPDS, defects in function β-cell mass may have already occurred during normal glucose tolerance before the onset of type 2 diabetes,⁶ by apoptosis and de-differentiation.^7,8 Therefore, how to preserve function β-cell mass and delay the onset of the disease more efficiently through early intervention, especially activation of its internal repairing mechanism before the complete loss of its functions, has become a new target in prevention and treatment of diabetes.

The application of pancreas and islet transplantation has provided an opportunity to cure diabetes.^9,10 However, the lack of sources and immune rejection limit the widespread of this approach.^11,12 Exosomes with diameters ranging 60–200 nm approximately are small nanoscale vesicles secreted by cells.^13,14 It exhibits a characteristic “discoid” ellipse or disc-like vesicle structure under electron microscopy, and its bilayer lipid membrane is rich in cholesterol, lecithin and sphingomyelin. It contains proteins, mRNAs and miRNAs, which can transmit messages through a variety of ways to the recipient cells.¹⁵ The formation and secretion of exosome’s cystic structure occur through stimulation of internal environment and factors, and hold great potential as novel diagnostic and prognostic molecular biomarkers of certain diseases. In addition, studies have shown that exosomes secreted in the extracellular environment in physiological or disease states are capable of various functions including reducing oxidative stress reactions, inhibiting apoptosis, inducing cell proliferation and promoting angiogenesis.¹⁶ They also play an important role in the regulation of inflammatory responses, the repairing of tissue damages and other pathophysiological progresses.^16,17 Exosome-based therapeutic approaches have been applied in treating neurological disorders,^18,19 skeletal muscle regeneration²⁰ and heart remodeling in myocardial infarction.²¹ Skeletal muscle is the major site for insulin action. Sarcopenia was associated with insulin resistance in both obese and non-obese individuals. Mesenchymal stem cells (MSCs)-derived exosome treatment promoted myogenesis and angiogenesis in C2C12, and muscle regeneration in muscle injury mice.²⁰ Moreover, MSCs-derived exosomes were reported to participate in myocardial repair and ameliorate cardiac damage after myocardial infarction via activating S1P/SK1/S1PR1 signaling.²² However, less effectiveness in lowering blood glucose is observed in diabetic model.²³ Thus, it is conceivable to expect that exosomes from β-cells containing elements needed for the cells might be able to protect pancreatic islet β-cells. The results of this study are expected to lay a foundation for new ways of treatment of diabetes.

Materials and methods

Antibodies

Antibodies used in histological study were obtained from Abcam (Cambridge, UK) (guinea pig anti-mouse insulin antibody, goat anti-guinea pig antibody conjugated with CY3, goat anti-rabbit antibody conjugated with Alexa Fluor 488, donkey anti-goat antibody conjugated with CY3, and rabbit anti-F4/80 antibody), and Santa Cruz (California, USA) (rabbit anti-mouse glucagon antibody, goat anti-mouse CD31 antibody). Antibody used in the flow cytometry was purchased from BD Pharmingen (San Diego, USA) (anti-mouse F4/80-PE). Antibodies used in Western blotting include rabbit polyclonal anti-mouse CD9, rabbit anti-mouse CD63 and goat anti-rabbit IgG-HRP (Abcam, Cambridge, UK). Mouse anti-GAPDH monoclonal antibody was purchased from KangChen Bio-tech Inc. (Shanghai, China).

Extraction and identification of exosomes

MIN6 cells, a kind gift from Dr. F. Liu (Department of Cellular and Structural Biology, University of Texas Health Science Center at San Antonio, TX, USA), were purchased originally from American Type Culture Collection²⁴ and cultured in high-glucose Dulbecco’s Modified Eagle Medium (Gibco, USA) supplemented with 15% exosome-free fetal bovine serum (FBS, Gibco, USA), 100μg/mL streptomycin, 100μg/mL penicillin, and 50μM β-mercaptoethanol. Exosomes were isolated from supernatant of MIN6 cells collected during a 48hrs period according to the previous reports.^25,26 Cellular debris was firstly removed through centrifugation at 4°C (at 300g for 10 mins, at 2000g for 10 mins, and at 10,000g for 30 mins). The supernatant was then centrifuged at 100,000g for 70 mins at 4°C. The pellets primarily contained exosomes were collected and resuspended in PBS and stored at −80°C. Western blotting was used to identify the exosomes by the marker protein CD9 and CD63.²⁵ The size distribution and concentration of the exosomes were measured by a ZetaSizer Nano ZS (Malvern Instruments, Worcestershire, UK). Accordingly, the samples were diluted in PBS, and tertiary measurements were performed at 25°C. Transmission electron microscopy (TEM) was used to observe the morphology of the exosomes.

Cell viability analysis

INS-1 (832/13) cells, a gift from Dr. Y. Liu (the Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences), were derived from parental INS-1 cells via a transfection-selection strategy by Dr. Christopher B. Newgard and his colleagues,²⁷ and were maintained in RPMI 1640 medium (Gibco, USA) supplemented with 10% FBS, 10mmol/L HEPES and 6.1 mmol/L glucose at 37°C and 5% CO₂ in a humid atmosphere. The cells were seeded in 96-well plates at a density of 2.5×10⁴/well. The cells were exposed to 0.4 mM palmitate in combination without or with different concentrations of exosomes (1.5×10⁹ partials/mL, 4.5×10⁹ partials/mL or 9×10⁹ partials/mL). After 48hrs, the cells were tested for the viability using Methylthiazolyldiphenyl-tetrazolium Bromide (Beyotime, China) assay.

Animal studies

Male C57BL/6J mice were purchased from Shanghai Slaccas Company (Shanghai, China). All experimental procedures involving animals were approved by the Animal Ethics Committee of Shanghai Jiao Tong University Affiliated Six People’s Hospital in accordance with the Guidelines for the Care and Use of Laboratory Animals (DWSY2014-068/DWLL2019-0275). All animals were housed under a standard 12hrs/12hrs light-dark cycle at constant temperature of 23±1°C with free access to standard food and water. At 6 weeks of age, the mice were administered with either high-dose of streptozotocin (STZ) (H-STZ, 200mg/kg) or low-dose of STZ (L-STZ, 140mg/kg) by intraperitoneal injection once on day 0 (Figure 1). Transplantation of exosomes was performed according to the previous report.²⁸ For H-STZ treated mice, 6×10⁹ particles of exosomes in 0.2 mL PBS or only PBS as control were transplanted in situ pancreas after their fasting blood glucose levels reached at least 11.1 mmol/L. Non-fasting blood glucose levels and body weight were measured every week since the transplantation (Figure 1A). As to L-STZ treated mice, 6×10⁹ particles of exosomes or PBS were transplanted in situ pancreas at day 10 after STZ injection, and 3×10⁹ particles of exosomes 11 days after first transplantation through tail vein injection. An intraperitoneal glucose tolerance test (IPGTT) was performed at day 7, day 17 and day 35 with 2mg/kg glucose and 6hrs fasting (Figure 1B). Serum samples were collected before and 30mins after 2mg/kg glucose load for insulin measurement at day 17 and 28. At the end of the experiment, both H-STZ- and L-STZ-treated mice were sacrificed, blood samples were collected and pancreatic tissues were preserved for insulin extraction and histology study.

Figure 1 Timeline for streptozotocin (STZ) injection, exosome transplantation and metabolic analyses. Notes: (A) High-dose of STZ (200mg/kg, once) was administered in mice to build a diabetic model. Exosomes derived from MIN6 cells or PBS were transplanted in situ pancreas after non-fasting glucose levels reached 11.1mmol/l. Body weight and non-fasted blood glucose concentrations were monitored once every week. (B) Low dose of STZ (140mg/kg, once) was intraperitoneally administered in mice. An intraperitoneal glucose tolerance test (IPGTT) was performed on day 7, day 17 and day 35. Exosomes or PBS were transplanted in situ pancreas on day 10 and were given via tail vein injection on day 21. Abbreviations: IPGTT, intraperitoneal glucose tolerance test; STZ, streptozotocin.

Histology study, Western blotting and insulin measurement

For immunofluorescence analysis, the pancreas was fixed in 4% formaldehyde and embedded in paraffin. Pancreatic sections were incubated at 4°C with guinea pig anti-insulin antibody (1:100 dilution), rabbit anti-glucagon antibody (1:100 dilution), or goat anti-CD31 antibody (1:50 dilution) overnight. Thereafter, the sections were stained with goat anti-guinea pig antibody conjugated with CY3 (1:300 dilution) for insulin, goat anti-rabbit antibody conjugated with Alexa Fluor 488 (1:400 dilution) for glucagon, and donkey anti-goat antibody conjugated to CY3 for CD31 (1:500 dilution). Pancreatic sections were also incubated with rabbit anti-F4/80 antibody (1:500 dilution) overnight for F4/80 staining. The digital images were acquired by a Zeiss Axio-Imager Standard Microscope (Carl Zeiss, Germany). Quantification of CD31 in islets was conducted by ZEN software (ZEN Version 2.0, Carl Zeiss Inc., Germany) from three nonconsecutive sections. Image was acquired using Zeiss 40× Objective. Data were expressed as the average ratio of CD31-positive density area in each islet over insulin-positive area of islets, as described.²⁹

To determine CD9 and CD63 expression, MIN6 cells and exosomes extracted from the cells were loaded to 12% SDS-PAGE gels and followed by transferring onto nitrocellulose membranes. The signals were detected by enhanced chemiluminescence (Thermo Fisher, USA) and imaged by Image Quant LAS 4000 mini bio-molecular imager (GE Healthcare, Uppsala, Sweden). Insulin content in plasma and in acetic acid extracts from pancreatic tissues was measured by mouse insulin ELISA kits (Hong Kong University Antibody and Immunoassay Services, China).

Flow cytometry

Splenocytes were isolated from freshly obtained spleen from each L-STZ animals at day 35 after STZ. In brief, the spleens were taken out and the spleen cells were separated with a 100μm cell strainer. After removing red cells using Lysing Buffer (BD Pharmingen), single splenic cells were washed with PBS, incubated with the fluorescent-labeled F4/80 antibody (F4/80-PE) at 4°C in dark for 30mins, and washed with PBS twice before acquisition. 20–50×10³ events were acquired by Aria II (Becton Dickinson, CA, USA) and analyzed by FlowJoV10.4.2 software (a Becton Dickinson, CA, USA).

Statistical analysis

Statistical analysis was performed with GraphPad Prism 5 for windows (GraphPad Software, San Diego, USA). The data in the study were showed as mean ± SEM, and the differences between/among groups were tested for significance with Student’s test or ANOVA and followed by Bonferroni’s multiple comparison test. Log-rank (Mantel-Cox) test was exploited to the survival curves comparison. P-values were two-tailed, and P<0.05 considered to be a significant difference.

Results

Characterization of exosomes

The size distribution of the vesicles in the supernatant derived from MIN6 cells was mainly in the range of 60–200 nm with an average size of 118.75±38.74nm (Figure 2A), which was consistent with the exosome size distribution from previous reports.³⁰ Vesicles with size of about 100 nm could be observed clearly from the TEM image (Figure 2B) and the subcellular structures exhibited the characteristic spherical morphology of exosomes. Moreover, CD9 and CD63, typical markers of exosomes, were highly expressed in the vesicles (Figure 2C).

Figure 2 Characterization of exosomes derived from MIN6 cells. Notes: (A) Size distribution of exosomes evaluated by a ZetaSizer Nano ZS. (B) Transmission electron microscopy image of the exosomes, scale bar: 100nm. (C) Western blot analysis of exosome marker CD9 and CD63. (D) Cell viability in INS-1(832/13) (INS-1) cells. INS-1 cells were exposed to palmitate (PA, 0.4 mmol/L) and different concentrations of exosomes (Exo, 0–1.5-4.5–9.0X109/mL) for 48 hrs. Data were presented as fold-change of INS-1 cell viability over cells treated without palmitate and exosomes. * P<0.05, ** P<0.01, between indicated groups (n=4). Abbreviations: PA, palmitate; Exo, exosomes.

To know if the exosomes derived from β-cells have any protective effect, we, then, co-cultured INS-1(832/13) cells exposed to palmitate (0.4mmol/L) together with the exosomes at indicated concentrations for 48hrs. As expected, exposure of palmitate significantly induced cell death (P<0.05) (Figure 2D). Addition of exosomes dose dependently improved the cell viability induced by palmitate (P<0.01, Figure 2D).

Exosome treatment prolonged the survival time of H-STZ-treated diabetic mice

Three to four days after STZ injection, non-fasting blood glucose concentrations in H-STZ-treated mice increased significantly (before STZ injection: 6.01±0.82 mmol/L vs after STZ injection: 21.01±5.11 mmol/L, P<0.001). Transplantation of the exosomes significantly decreased the non-fasting glucose levels despite its failure to reach normal levels (P<0.05, Figure 3A), and prolonged the median survival time (15 vs 30 days) (P<0.01, Figure 3B). No significant differences in the body weight, serum and pancreatic insulin content could be viewed between exosome treated and non-treated diabetic mice (Figure 3C–E).

Figure 3 Effect of exosomes (Exo) on streptozotocin (STZ)-induced diabetic mice. Notes: Male C57 mice were administrated with high-dose of STZ (200mg/kg), and Exo was transplanted afterwards. (A) Non-fasting blood glucose levels, (B) survival time, (C) body weight, (D) serum insulin and (E) insulin content extract from pancreas were compared. Data were showed as mean ± SEM. *P<0.05, **P<0.01, ***P<0.001, between indicated groups (Control, n=3; STZ, n=5; STZ+Exo, n=11). Abbreviations: Exo, exosomes; STZ, streptozotocin.

Exosome treatment improved glucose metabolism, islet morphology and increased insulin content in L-STZ-treated mice

L-STZ-treated mice exhibited progressively impaired glucose tolerance at day 7, 17 and 35, respectively (Figure 4A–F) (P<0.05). At day 17 and 28, glucose-induced insulin secretion and pancreatic insulin content at the end of experiments decreased dramatically compared with normal controls (P<0.001, Figure 4g-i). In consistency, there was a disruption in islet architecture and decrease in insulin staining in these mice (Figure 5). However, compared to mice with only STZ injection, the exosome-treated mice exhibited an improvement of the glucose tolerance (P<0.05, Figure 4B and C) and a decrease in the area under curve (AUC) of IPGTT (P<0.01, Figure 4E and F). Consistently, the exosome treatment significantly elevated glucose-induced insulin secretion and pancreas insulin content (P<0.05, Figure 4G–I), and maintained the insulin staining and islet architectures (Figure 5). These results together suggested that the exosomes had a protective effect on β-cells through promoting the cell survival, and preserving the insulin storage and islet morphology.

Figure 4 Effect of exosomes (Exo) on glucose metabolism in low-dose streptozotocin (STZ)-treated mice. Notes: Male C57 mice were treated with 140mg/kg STZ, and Exo derived from MIN β-cells was transplanted at 10 days after STZ injection. (A-F) Intraperitoneal glucose tolerance tests (IPGTT) were performed on day 7 (A), day 17 (B) and day 35 (C). Area under curve (AUC) of IPGTT was calculated (D-F). Serum insulin levels before and after glucose load were detected at day 17 (G) and day 28 (H). (I) Insulin content extracted from pancreas were detected. Data were showed as mean ± SEM, *P<0.05, **P<0.01, ***P<0.001 between indicated groups, or between STZ + Exo and STZ-treated mice (Control, n=4; STZ, n=5; STZ+Exo, n=5). Abbreviations: AUC, Area under curve; Exo, exosomes; IPGTT, intraperitoneal glucose tolerance test; STZ, streptozotocin.

Figure 5 Effect of exosomes (Exo) on islet morphology of low-dose streptozotocin (STZ)-treated mice. Notes: Mice were injected with 140mg/kg of STZ and transplanted with Exo after 10 days of STZ. The pancreatic sections from the mice were dual-stained for insulin in red and glucagon in green. DAPI staining blue indicated nuclei. Scale bar =20μm. Abbreviations: Exo, exosomes; STZ, streptozotocin.

Exosome treatment enhanced CD31 expression in islets from L-STZ mice

Disruption of inner vascularization is associated with STZ-induced β-cell death. To further understand the mechanism of diabetic alleviation using the exosome treatment, a histological analysis was carried out in these animals. As shown in Figure 6, the expression of CD31, a marker for endothelial cells, was weaker after L-STZ injection compared with normal control. However, its expression was greatly enhanced after the exosomes, indicating an association of the exosomes and vascular density (Figure 6).

Figure 6 Effect of exosomes (Exo) on endothelial cell proliferation in islets. Notes: Mice were injected with 140mg/kg of streptozotocin (STZ) and transplanted with Exo after 10 days of STZ. (A) The pancreas sections were dual-stained for insulin in red and CD31 in green. DAPI staining blue indicated nuclei. Scale bar =50 μm. Dashed line indicates the area of islets. (B) Quantification of CD31 density in islets, data were expression as ratio of CD31 density over islet area (12–16 islets/group). *P<0.05, ***P<0.001 between indicated groups. Abbreviations: Exo, exosomes; STZ, streptozotocin.

Exosome treatment tended to reduce infiltration of macrophages in islets from L-STZ-treated mice

To decipher if exosomes play any role on inflammatory regulation, pancreas sections from mice at 35 days after L-STZ were stained with F4/80, a major macrophage marker. As shown in Figure 7A, F4/80-positive cells were increased in islets from STZ-treated mice. Exosome treatment tended to lower the expression of F4/80 in islet cells (Figure 7A). However, no changes could be observed in the relative frequency of macrophages in splenocytes isolated from STZ-treated alone or together with exosome-treated mice when compared with that of control mice (Figure 7B).

Figure 7 Effect of exosomes (Exo) on macrophage infiltration in mice treated with streptozotocin (STZ). Notes: Mice were injected with 140mg/kg of STZ and transplanted with Exo after 10 days of STZ. (A) The pancreas sections were stained with F4/80, a marker for macrophages. Scale bar =50μm. Quantifications of F4/80-positive cells within islets were shown below. (B) Splenocytes were isolated from freshly obtained spleen, F4/80-positive cells were analyzed with flow cytometry. Data were showed as mean ± SEM, n=4. *P<0.05, between indicated groups. Abbreviations: Exo, exosomes; STZ, streptozotocin.

Discussion

In the present study, we provide evidences demonstrating for the first time that the exosomes derived from β-cell lines exert a protective effect on STZ-induced glucose disturbance, in which the exosomes prolong the survival time in H-STZ-induced diabetic mice and alleviate glucose intolerance in L-STZ-treated mice. Most importantly, the exosome treatments confer resistance to insulin decrease in pancreas and serum, and maintain islet morphology in L-STZ mice. Furthermore, we find that these protective effects are associated with the increase of CD31 expression in pancreas and decrease of macrophage infiltration in islets suggesting a role of the exosomes on neovascularization and modulation of immune response within pancreatic islets, which might contribute to the protective effects of the exosomes on STZ-treated mice.

STZ, a highly selective pancreatic β-cells cytotoxic agent, has been widely applied in animal experiments to build diabetic model. In this study, we built two types of diabetic mouse model with different dosages of STZ injection. The H-STZ treatment destroyed pancreatic islet β-cells within a short time, resulting in rapid β-cells necrosis, significant hyperglycemia and diabetes in the mice. The L-STZ could not induce a rapid β-cell death within a short time, but induce abnormal glucose metabolism in the mice.³¹ Transplantation of the exosomes remarkably prolonged the survival time of H-STZ-treated mice, as well as improved glucose tolerance of the L-STZ-treated mice. Moreover, there were significant improvements in the serum insulin levels and pancreatic insulin content in L-STZ-treated mice. The islet morphology and insulin storage were preserved after exosome transplantation in these mouse models observed by the immunofluorescence analysis. Despite the limitation of exosomes failure to bring dysfunctional β-cells back to normal level, the above data indicated that the exosomes derived from β-cells played an essential role in the functional recovery of β-cell and internal glucose homeostasis maintenance.

Mechanisms involving the effects of the exosomes on diabetic alleviation might be related to the following aspects. Firstly, the exosomes exerted their above-mentioned effects through modulation of immune response. It is accepted that STZ exerts the cytotoxic effect in β-cells through excessive production of reactive oxygen species, activation of inflammation, and cell apoptosis.³² In the present study, the exosome treatment of L-STZ mice tended to decrease the infiltration of macrophages induced by STZ in islets. Macrophages play a crucial role in diabetes development through the induction of pancreatic inflammation³³ and are featured by functional diversity and plasticity.

The ability of exosomes on immune regulation has been reported. However, the reported role of exosomes in immunity appears to be complex with both anti-inflammatory and pro-inflammatory properties probably depending on the parent cells they are derived from along with the contexts and environments.^34–36 It has been demonstrated that exosomes from red blood cells are pro-inflammatory and are responsible for induction of TNF-α production by monocytes.³⁷ Additionally, exosome-like vesicles released from adipose tissue differentiate monocytes into activated macrophages with increased secretion of TNF-α and IL-6. Injection of these exosomes led to insulin resistance in mice through TLR4 pathway.³⁸ Similarly, exosomes from lipid-modified muscles are rich in lipid and can be incorporated into pancreas.³⁹ It also regulated the adaptations of β-cell mass,⁴⁰ suggesting a role of exosomes on inflammation and metabolism regulation. Yet, there are several studies supporting an immune suppressive role of the exosomes. MSCs-derived exosomes have been reported to suppress the secretion of pro-inflammatory factor TNF-α and IL-1β, but increase the concentration of anti-inflammatory factor TGF-β.⁴¹ Most interestingly, contrary to the findings, exosomes derived from adipose-derived stem cells attenuate high-fat diet-induced obesity and improve glucose tolerance and insulin sensitivity. Our data showing that exosomes protected islets against inflammation in islets through lowering the infiltration of macrophages are in line with the latter reports. The deep mechanisms involved in the regulatory effect of exosomes on macrophage need to be further elucidated.

Furthermore, we demonstrated in the present study that exosomes exert their ability on β-cell functional recovery through increasing endothelial cell proliferation and re-establishing vasculature in the islets. Islets are highly vascularized and its vasculature is essential for β-cell proliferation. Revascularization is one of the key factors for the improvement of islet function after transplantation. Endothelial cell CD31 is an important angiogenic and vascularizing factor,⁴² and has been commonly used as the marker for endothelial cell differentiation and angiogenesis.^43,44 The finding in the present study that the expression of CD31 increased significantly in the pancreas from mice treated with the exosome compared with those treated with STZ only indicates a role of the exosomes on revascularization regulation in the pancreas. This notion is in agreement with the previous findings, in which exosomes derived from stem cells or cell lines (eg cardiomyocytes H9c2) promoted angiogenesis on would healing, ischemic stroke and myocardial ischemic injury through induction of expression of a number of growth factors.^45–47 However, caution needs to be exercised in applying such angiogenesis-prone agents whenever potential risks in tumorigenicity exist.

Since MIN6 cell line has been shown to be not a pure, but a mixed population of cells expressing pancreatic endocrine hormones and bioactive molecules such as insulin, glucagon, somatostatin and GPR109A,^48,49 the exosome proteomics was conducted. A total of 1126 proteins were identified. Among them, 1029 proteins with gene ontology annotations were analyzed and shown in Figure S1. However, apart from insulin, the above-mentioned hormones and GPR109A were not found in the exosome, indicating that other factors might be involved in the effect of exosome on STZ-induced models. Further studies are needed to explore the mechanisms underlying the roles of exosomes on β-cell protection.

Conclusion

Our study demonstrated for the first time that exosomes derived from β-cells have the ability to maintain the homeostasis of glucose metabolism in STZ-induced diabetic mice, preserve islet architectures, and promote angiogenesis. These findings indicate a role of the exosomes on neovascularization regulation, which may provide a potential therapeutic strategy for diabetic treatment.