Underlying mechanisms of glucocorticoid-induced β-cell death and dysfunction: a new role for glycogen synthase kinase 3

Abstract

Glucocorticoids (GCs) are widely prescribed for their anti-inflammatory and immunosuppressive properties as a treatment for a variety of diseases. The use of GCs is associated with important side effects, including diabetogenic effects. However, the underlying mechanisms of GC-mediated diabetogenic effects in β-cells are not well understood. In this study we investigated the role of glycogen synthase kinase 3 (GSK3) in the mediation of β-cell death and dysfunction induced by GCs. Using genetic and pharmacological approaches we showed that GSK3 is involved in GC-induced β-cell death and impaired insulin secretion. Further, we unraveled the underlying mechanisms of GC-GSK3 crosstalk. We showed that GSK3 is marginally implicated in the nuclear localization of GC receptor (GR) upon ligand binding. Furthermore, we showed that GSK3 regulates the expression of GR at mRNA and protein levels. Finally, we dissected the proper contribution of each GSK3 isoform and showed that GSK3β isoform is sufficient to mediate the pro-apoptotic effects of GCs in β-cells. Collectively, in this work we identified GSK3 as a viable target to mitigate GC deleterious effects in pancreatic β-cells.

Fig. 1. Treatment with LiCl or SB216763 decreases Dexa-induced cell death in INS-1 832/13 cell line and in rat islets.

A Cell death was assessed by 7-AAD incorporation measured by flow cytometry after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without SB216763 (SB) in INS-1 832/13 cells. B Quantification of 7-AAD positives INS-1 832/13 cells (n = 6). C Cell apoptosis was evaluated by Annexin V-FITC incorporation assay. Annexin V-FITC positive cells were measured by flow cytometry after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without SB216763 (SB) in INS-1 832/13 cells (n = 4). D Representative images of TUNEL-positive cells in islets, after 24 h incubation with or without Dexamethasone (Dexa), and 27 h incubation with or without LiCl. The TUNEL-positive nuclei are stained in brown. White arrows show positive cells. Scale bar: 200 µm. E Quantification of TUNEL-positive cells in islets (n = 4) after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without LiCl or SB216763 (SB). More than 20 islets per condition were counted. F Relative mRNA levels of Bim and Bcl-2, measured by qPCR (n = 4) in INS-1 832/13, after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without SB216763 (SB). Results are expressed as percentage of NT condition for each independent experiment. G Relative mRNA levels of Bim and Bcl-2, measured by qPCR (Bim: n = 3–4; Bcl-2 = 3–5) in islets, after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without LiCl or SB216763 (SB). Results are expressed as percentage of NT condition for each independent experiment. Cyclophilin A is used as housekeeping gene in figure F and G. H Representative Western Blot (top) and quantification (bottom) of Bim and Bcl-2 protein levels in INS-1 832/13 cells after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without SB216763 (SB) (n = 5). β-actin is used as loading control. Results are expressed as percentage of NT condition for each independent experiment. I Representative Western Blot (top) and quantification (bottom) of Bim and Bcl-2 protein levels in islets after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without LiCl (Bim n = 4; Bcl-2 n = 5). β-actin is used as loading control. Results are expressed as percentage of NT condition for each independent experiment. Results are shown as means ± S.E.M. n represents the number of independent experiments. *p < 0.05; **p < 0.01; ****p < 0.0001.

Fig. 2. Treatment with LiCl or SB216763 improves insulin secretion in Dexa condition in rat islets.

A Measurement of insulin secretion (n = 5) in rat islets after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without LiCl. Insulin secretion is assessed in medium containing 2.8 mM glucose or 16.7 mM glucose. Insulin in the medium was measured by ELISA and normalized to the DNA content. Results are expressed as percentage of NT 2.8 mM condition. B Measurement of insulin secretion (n = 5) in rat islets after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without SB216763 (SB). Insulin secretion is assessed in medium containing 2.8 mM glucose or 16.7 mM glucose. Insulin in the medium was measured by ELISA and normalized to the DNA content. Results are expressed as percentage of NT 2.8 mM condition. C Relative mRNA levels of Sgk1 and Kv1.5, measured by qPCR (n = 3) in islets, after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without LiCl or SB216763 (SB). Cyclophilin A is used as housekeeping gene. Results are expressed as percentage of NT condition for each independent experiment. Results are shown as means ± S.E.M. n represents the number of independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Fig. 3. Dexa does not change GSK3α and GSK3β expression in INS-1 832/13 and in rat islets neither their activity.

A Relative mRNA levels of Gsk3α and Gsk3β, measured by qPCR (n = 4) in INS-1 832/13, after 24 h incubation with or without Dexamethasone (Dexa). Results are expressed as percentage of NT condition for each independent experiment. B Relative mRNA levels of GSK3α and GSK3β, measured by qPCR (n = 7), in islets, after 24 h incubation with or without Dexamethasone (Dexa). Results are expressed as percentage of NT condition for each independent experiment. Cyclophilin A is used as housekeeping gene in A and B. C Representative Western Blot (left) and quantification (right) of P-GSK3β (S9) and GSK3β protein levels in INS-1 832/13 cells after 24 h incubation with or without Dexamethasone (Dexa) (n = 5). β-actin is used as loading control. Results are expressed as percentage of NT for each independent experiment. Black arrow shows the band corresponding to P-GSK3β (S9). D Representative Western Blot (left) and quantification (right) of P-GSK3α (S21) and GSK3α protein levels in INS-1 832/13 cells after 24 h incubation with or without Dexamethasone (Dexa) (n = 5). β-actin is used as loading control. Results are expressed as percentage of NT for each independent experiment. E Representative Western Blot (left) and quantification (right) of P-GSK3β (S9) and GSK3β protein levels in rat islets after 24 h incubation with or without Dexamethasone (Dexa) (n = 5). β-actin is used as loading control. Results are expressed as percentage of NT for each independent experiment. Black arrow shows the band corresponding to P-GSK3β (S9). F Representative Western Blot (left) and quantification (right) of P-GSK3α (S21) and GSK3α levels in rat islets after 24 h incubation with or without Dexamethasone (Dexa) (n = 5). β-actin is used as loading control. Results are expressed as percentage of NT for each independent experiment. Results are shown as means ± S.E.M. n represents the number of independent experiments. **p < 0.01.

Fig. 4. Treatment with SB216763 reduces the nuclear GR localization following Dexa treatment.

A Representative Western Blot (left) and quantification (right) of nuclear and cytoplasmic GR protein levels in INS-1 832/13 after 3 h incubation with or without Dexamethasone (Dexa) and with or without SB216763 (SB) (n = 5). Tubulin is used as the marker of the cytoplasmic fraction. Histone H3 is used as the marker of the nuclear fraction. Results are expressed as the percentage of Dexa condition for each independent experiment. B Representative images showing DAPI (upper panel) and GR staining (lower panel) in INS-1 832/13 after 3 h incubation with or without Dexamethasone (Dexa) and 3 h with or without SB216763 (SB). White arrows show examples of nuclear staining of the GR. Scale bar: 20 μm. C Representative Western Blot (left) and quantification (right) of nuclear and cytoplasmic GSK3β protein level in INS-1 832/13 after 3 h incubation with or without Dexamethasone (Dexa) and with or without SB216763 (SB) (n = 5). Tubulin is used as the marker of the cytoplasmic fraction. Histone H3 is used as the marker of the nuclear fraction. Results are expressed as the percentage of NT condition for each independent experiment. D Representative Western Blot (left) and quantification (right) of nuclear and cytoplasmic GSK3α levels in INS-1 832/13 after 3 h incubation with or without Dexamethasone (Dexa) and with or without SB216763 (SB) (n = 5). Tubulin is used as the marker of the cytoplasmic fraction. Histone H3 is used as the marker of the nuclear fraction. Results are expressed as the percentage of NT condition for each independent experiment. Results are shown as means ± S.E.M. n represents the number of independent experiments. **p < 0.01.

Fig. 5. Treatment with LiCl or SB216763 reduces the GR expression in INS-1 832/13 cells as well as in rat and human islets.

A Relative mRNA level of GR, measured by qPCR (n = 4), in INS-1 832/13, after 24 h treatment with or without Dexamethasone (Dexa) and 27 h with or without SB216763 (SB). Results are expressed as percentage of NT condition for each independent experiment. B Representative Western Blot (right) and quantification (left) of GR protein level in INS-1 832/13 after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without SB216763 (SB) (n = 5). β-actin is used as loading control. Results are expressed as percentage of NT condition for each independent experiment. C Relative mRNA level of GR, measured by qPCR (n = 3) in rat islets, after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without LiCl or SB216763. Results are expressed as percentage of NT condition for each independent experiment. D Representative Western Blot (right) and quantification (left) of GR protein level in rat islets after 24 h incubation with or without Dexamethasone (Dexa) and 27 h incubation with or without LiCl (n = 4). β-actin is used as loading control. Results are expressed as percentage of NT condition for each independent experiment. E Relative mRNA level of human GR, measured by qPCR (n = 4), in human islets, after 24 h treatment with or without LiCl. Results are expressed as percentage of NT condition for each independent experiment. Cyclophilin A is used as housekeeping gene in figure A and C. TBP is used as housekeeping gene in figure E.

Fig. 6. GSK3β knock-down is sufficient to decrease GR expression and GC-induced β-cell death.

A Representative Western Blot (left) and quantification (right) of GSK3β and GSK3α levels in INS-1 832/13 after 24 h incubation with GSK3β siRNA (upper panel, n = 6) or with GSK3α siRNA (lower panel, n = 6). β-actin is used as loading control for GSK3α and tubulin is used as loading control for GSK3β. Results are expressed as percentage of siCTL condition, represented by dashed line, for each independent experiment. B Representative Western Blot (left) and quantification (right) of GR protein level in INS-1 832/13 after 24 h incubation with GSK3β siRNA (upper panel, n = 6), or with GSK3α siRNA (lower panel, n = 6). Tubulin is used as loading control. Results are expressed as percentage of siCTL condition for each independent experiment. C Relative mRNA level of GSK3β (upper panel) and GSK3α (lower panel), measured by qPCR, in INS-1 832/13, after infection with lentivirus coding for scrambled shRNA (shScr), GSK3β.1 shRNA (shGSK3β.1) or GSK3β.2 shRNA (shGSK3β.2) (n = 3). Results are expressed as percentage of scrambled shRNA condition, represented by dashed line, for each independent experiment. D Representative Western-blot (upper panel) and quantification (lower panel) of GR protein level in GFP sorted INS-1 832/13 after infection with lentivirus coding for scrambled shRNA or GSK3β.2 shRNA (n = 2). β-actin is used as loading control. Results are expressed as percentage of scrambled shRNA condition for each independent experiment. E Cell death reported by 7-AAD incorporation measured by flow cytometry, and quantification (n = 5), after infection of INS-1 832/13 cells with lentivirus coding for scrambled shRNA, GSK3β.1 shRNA or GSK3β.2 shRNA, and 24 h after incubation with (right panel) or without Dexamethasone (Dexa) (left panel). Cyclophilin A is used as housekeeping gene in figure C. Results are shown as means ± S.E.M. n represents the number of independent experiments. *p < 0.05; **p < 0.01.