miR-320a induces pancreatic β cells dysfunction in diabetes by inhibiting MafF

Abstract

A variety of studies indicate that microRNAs (miRNAs) are involved in diabetes. However, the direct role of miR-320a in the pathophysiology of pancreatic β cells under diabetes mellitus remains unclear. In the current study, islet transplantation and hyperglycemic clamp assays were performed in miR-320a transgenic mice to explore the effects of miR-320a on pancreatic β cells in vivo. Meanwhile, β cell-specific overexpression or inhibition of miR-320a was delivered by adeno-associated virus (AAV8). In vitro, overexpression or downregulation of miR-320a was introduced in cultured rat islet tumor cells (INS1). RNA immunoprecipitation sequencing (RIP-Seq), luciferase reporter assay, and western blotting were performed to identify the target genes. Results showed that miR-320a was increased in the pancreatic β cells from high-fat-diet (HFD)-treated mice. Overexpression of miR-320a could not only deteriorate the HFD-induced pancreatic islet dysfunction, but also initiate pancreatic islet dysfunction spontaneously in vivo. Meanwhile, miR-320a increased the ROS level, inhibited proliferation, and induced apoptosis of cultured β cells in vitro. Finally, we identified that MafF was the target of miR-320a that responsible for the dysfunction of pancreatic β cells. Our data suggested that miR-320a could damage the pancreatic β cells directly and might be a potential therapeutic target of diabetes.

Keywords: MafF; diabetes; miRNA; pancreatic islet; β cells.

Graphical abstract

Figure 1

miR-320a was increased in HFD damaged pancreatic β cells (A) Body weight was measured every three weeks since the age of 8 weeks and results for the 12 weeks; data are expressed as mean ± SEM, n = 6. ∗p < 0.05, calculated by two-way ANOVA analysis. (B) Blood glucose was measured every three weeks since the age of 8 weeks and results for the 12 weeks; data are expressed as mean ± SEM, n = 6. ∗∗p < 0.01, calculated by two-way ANOVA analysis. (C) Insulin level was measured every three weeks since the age of 8 weeks and results for the 12 weeks; data are expressed as mean ± SEM, n = 5. ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis. (D) Islets were isolated from each time point mice and cultured for 48 h. Then insulin secretion from islets was determined in response to 2.8- or 22-mM glucose; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, ∗∗p < 0.01, calculated by two-way ANOVA analysis. (E) Representative images of immunofluorescence staining for miR-320a (red), insulin (green), and Hoechst (blue) in islets from wild-type mice. Scale bar, 25 μm. (F) Relative miR-320a expression in isolated islets measured by real-time PCR; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, calculated by one-way ANOVA analysis. (G) and (H) Apoptosis of INS1 were determined by Annexin V/PI flow cytometric analysis; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis. (I) Relative miR-320a expression in INS1 measured by real-time PCR; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis.

Figure 2

Overexpression of miR-320a aggravated the HFD-induced damages in pancreatic β cells in vivo (A) Pancreatic islets from miR-320a transgenic (TG) or wild-type (WT) mice were isolated and cultured. Insulin secretion was determined in response to 2.8 or 22mM glucose; data are expressed as mean ± SEM, n = 5. ∗p < 0.05, calculated by two-way ANOVA analysis. (B) Model of islet transplantation. (C) Streptozocin (STZ)-induced diabetic WT mice were transplanted with an appropriate amount (100) of WT islets or miR-320a transgenic (TG) mice islets. Blood glucose levels were monitored for 6 weeks, and results expressed as area under the curve; data are expressed as mean ± SEM, n = 5. ∗p < 0.05, calculated by Student’s t test. Diabetic mice receiving WT or TG mice islets were fasted overnight and glucose tolerance tests (1.0 g/kg body weight) were performed at 6 weeks post-transplant. Diabetic mice receiving WT or TG mice islets were fasted overnight and hyperglycemic clamp experiments were performed at 6 weeks post-transplant. Plasma glucose (D) and - GINF (E) were measured during hyperglycemic clamps. Insulin (F) and C-peptide levels (G) before and during the last 20 min of the hyperglycemic clamp and (H) disposition indexes (DI = GINF/insulin x C-peptide) during the last 20 min of hyperglycemic clamp in mice; data are expressed as mean ± SEM, n = 5. ∗p < 0.05, ∗∗p < 0.01, calculated by two-way ANOVA analysis. (I) Representative images of immunofluorescence staining for TUNEL (green), insulin (red), and Hoechst (blue) in transplanted islets from WT-WT or TG-WT mice. Scale bar, 10 μm. Data are expressed as mean ± SEM, n = 3, ∗∗p < 0.01, calculated by Student’s t test. (J) Diabetic mice receiving WT or TG mice islets were fed with high-fat diet for another 6 weeks and blood glucose levels were monitored, and results expressed as area under the curve, data are expressed as mean ± SEM, n = 5. ∗∗p < 0.01, calculated by two-way ANOVA analysis. Diabetic mice receiving WT or TG mice islets were fasted overnight and glucose tolerance tests (1.0 g/kg body weight) were performed at 6 weeks post-transplant. Diabetic mice receiving WT or TG mice islets were fasted overnight and hyperglycemic clamp experiments were performed at 6 weeks post-transplant. Plasma glucose (K) and GINF (L) were measured during hyperglycemic clamps. Insulin (M) and C-peptide levels (N) before and during the last 20 min of the hyperglycemic clamp, data are expressed as mean ± SEM, n = 5. ∗p < 0.05, ∗∗p < 0.01, calculated by two-way ANOVA analysis. (O) Disposition indexes during the last 20 min of hyperglycemic clamp in mice; data are expressed as mean ± SEM, n = 5. ∗p < 0.05, calculated by one-way ANOVA analysis. (P) Representative images of immunofluorescence staining for TUNEL (green), insulin (red), and Hoechst (blue) in transplanted islets from WT-WT+HFD or TG-WT+HFD mice. Scale bar, 10 μm. Data are expressed as mean ± SEM, n = 3, ∗p < 0.05, calculated by Student’s t test.

Figure 3

Overexpression of miR-320a damaged pancreatic β cells in vitro INS1 cells were transfected with miR-320a mimics/inhibitors and then subjected to palmitate (0.3 mmol/L) stimulation. (A) and (B) Effects of miR-320a mimics/inhibitors on proliferation were determined by EdU flow cytometric analysis in cultured INS1 cells; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, calculated by one-way ANOVA analysis. (C) and (D) Effects of miR-320a mimics/inhibitors on apoptosis were determined by Annexin V/PI flow cytometric analysis in cultured INS1 cells; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis. (E) and (F) Effects of miR-320a mimics/inhibitors on the ROS levels were determined by DCFH-DA flow cytometric analysis in cultured INS1 cells; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis.

Figure 4

miR-320a suppressed the function of pancreatic β cells via MafF (A) Heatmap and Volcano plot (B) of the genes in RIP-seq using anti-Ago2 antibody in INS1 cells following miR-320a mimics or miR-con transfection. n = 2. (C) Venn diagram showing the overlap number of candidate target genes of miR-320a identified by Ago2-RIP-seq, which contained the binding sites of miR-320a among rat, mouse, and human species predicted by RNAhybrid. (D) Regulation of miR-320a on 3′-UTR of Ift57, Krt8, Ddrgk1, Cald1, MafF, Mlec, Meltf and Klhl36 in INS1 cells by luciferase reporter assay; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, calculated by Student’s t test. (E) Sequence alignment of binding sites between miR-320a and the 3′-UTR of MafF among several species. (F and G) Relative expression of MafF detected by western blot. Data are expressed as mean ± SEM, n = 3, ∗p < 0.05, calculated by one-way ANOVA analysis.

Figure 5

MafF restoration attenuated miR-320a induced pancreatic β cells dysfunction in vivo (A)–(D) Body weight and blood glucose levels were monitored for 7 weeks and results were expressed as area under the curve; data are expressed as mean ± SEM, n = 10. ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis. (E) Pancreatic islets from mice treated with AAV8-insulin vectors under normal diet were isolated and cultured. Insulin secretion was determined in response to 2.8 or 22mM glucose; data are expressed as mean ± SEM, n = 5. ∗p < 0.05, calculated by two-way ANOVA analysis. Plasma glucose (F) and GINF (G) were measured during hyperglycemic clamps. (H) and (I) Insulin and C-peptide levels before and during the last 20 min of the hyperglycemic clamp during the last 20 min of hyperglycemic clamp in mice; data are expressed as mean ± SEM, n = 6. ∗p < 0.05, calculated by two-way ANOVA analysis. (J) Disposition indexes (DI = GINF/insulin x C-peptide) during the last 20 min of hyperglycemic clamp in mice; data are expressed as mean ± SEM, n = 4. ∗p < 0.05, calculated by two-way ANOVA analysis. (K) and (L) Representative images of immunofluorescence staining for TUNEL (green), insulin (red), and Hoechst (blue) in islets from HFD mice treated with pancreatic β cells specific insulin vectors (AAV8-insulin). Scale bar, 20 μm. Data are expressed as mean ± SEM, n = 5, ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis.

Figure 6

MiR-320a damaged pancreatic β cells via inhibiting MafF/Nrf2 complex (A) Relative MafF protein levels in HFD mice detected by western blot; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis. (B) Relative MafF protein levels detected INS1 cells treated with palmitate by western blot; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis. (C) Relative MafF protein levels detected by western blot; data are expressed as mean ± SEM, n = 4. ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis. (D) Proliferation was determined by immunohistochemical staining for EdU (green) and Hoechst (blue) in cultured INS1 cells; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, ∗∗p < 0.01, calculated by one-way ANOVA analysis. (E) Apoptosis was determined by Annexin V/PI flow cytometric analysis in cultured INS1 cells; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, calculated by one-way ANOVA analysis. (F) The ROS levels were determined by DCFH-DA flow cytometric analysis in cultured INS1 cells; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, calculated by one-way ANOVA analysis. (G) Relative Nrf2 protein level detected by western blot. (H) Relative mRNA levels of NQO1, HO-1 measured by real-time PCR under HFD conditions; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, calculated by one-way ANOVA analysis. (I) Relative mRNA levels of NQO1, HO-1 measured by real-time PCR under palmitate conditions; data are expressed as mean ± SEM, n = 3. ∗p < 0.05, calculated by one-way ANOVA analysis (J) Schematic representation of the role of miR-320a in β cells dysfunction. Overexpression of miR-320a enhanced the oxidative stress of β cells via inhibiting the MafF/Nrf2 signal pathway to increase the ROS level, inhibit the proliferation and induce the apoptosis, thus leading to the reduction of insulin release, which finally impaired the secretion of insulin.