Role of MKK3 and p38 MAPK in cytokine-induced death of insulin-producing cells

Abstract

The aim of the present investigation was to elucidate further the importance of p38 MAPK (mitogen-activated protein kinase) in nitric oxide- and cytokine-induced beta-cell death. For this purpose, isolated human islets were treated with d-siRNA (diced small interfering RNA) and then exposed to the nitric oxide donor DETA/NONOate [2,2'-(hydroxynitrosohydrazono)bis-ethanamine]. We observed that cells treated with p38alpha-specific d-siRNA, but not with d-siRNA targeting GL3 (a firefly luciferase siRNA plasmid) or PKCdelta (protein kinase Cdelta), were protected against nitric oxide-induced death. This was paralleled by an increased level of Bcl-XL (B-cell leukaemia/lymphoma-X long). For an in-depth study of the mechanisms of p38 activation, MKK3 (MAPK kinase 3), MKK6 and their dominant-negative mutants were overexpressed in insulin-producing RIN-5AH cells. In transient transfections, MKK3 overexpression resulted in increased p38 phosphorylation, whereas in stable MKK3-overexpressing RIN-5AH clones, the protein levels of p38 and JNK (c-Jun N-terminal kinase) were decreased, resulting in unaffected phospho-p38 levels. In addition, a long-term MKK3 overexpression did not affect cell death rates in response to the cytokines interleukin-1beta and interferon-gamma, whereas a short-term MKK3 expression resulted in increased cytokine-induced RIN-5AH cell death. The MKK3-potentiating effect on cytokine-induced cell death was abolished by a nitric oxide synthase inhibitor, and MKK3-stimulated p38 phosphorylation was enhanced by inhibitors of phosphatases. Finally, as the dominant-negative mutant of MKK3 did not affect cytokine-induced p38 phosphorylation, and as wild-type MKK3 did not influence p38 autophosphorylation, it may be that p38 is activated by MKK3/6-independent pathways in response to cytokines and nitric oxide. In addition, it is likely that a long-term increase in p38 activity is counteracted by both a decreased expression of the p38, JNK and p42 genes as well as an increased dephosphorylation of p38.

Figure 1. p38 isoform expression patterns in response to cytokine stimulation

(A) Total RNA was purified from RIN-5AH cells and human T-lymphocyte Jurkat cells and real-time PCR analysis of p38α, p38β, p38γ and p38δ was performed. Data are p38/GAPDH ratios expressed as percentage of p38α in RIN-5AH cells not exposed to cytokines. Results are means for two different experiments. (B) Real-time PCR analysis of p38α, p38β, p38γ and p38δ in isolated rat islets and mouse endothelial IBE cells.

Figure 2. p38α mediates NO-induced human islet cell death

(A) Human islet cells were dispersed and transfected with d-siRNA against GL3 (control), p38α or PKCδ. On the next day, 2 mM DETA/NONOate was added and the cells were incubated for another 24 h. Cells were then stained with propidium iodide and Hoechst and inspected with a fluorescence microscope. Cell nuclei with red staining (necrosis), white condensed or fragmented staining (apoptosis) and blue staining (viable cells) were counted. Results are combined rates of apoptosis and necrosis and are expressed as percentage of the total cell number. Results are means±S.E.M. for four separate observations. *P<0.05 versus the corresponding control using Student's paired t test. (B) Two days after transfection of dispersed human islet cells with GL3-d-siRNA or p38-d-siRNA, the cells were lysed in SDS sample buffer and analysed for p38α and JNK2 expression using the immunoblot technique. The immunoblot is representative of two separate experiments. (C) Two days after transfection of dispersed islet cells with GL3-d-siRNA or p38-d-siRNA and 1 day after addition of 2 mM DETA/NONOate to some of the groups, the cells were lysed in SDS sample buffer and analysed for Bcl-XL and Bax expression using the immunoblot technique. Two representative experiments are shown. Protein loading is visualized by Amido Black staining.

Figure 3. p38, JNK2 and p44/42 are not hyperphosphorylated in MKK3/6-overexpressing RIN-5AH clones

MKK3wt, MKK3m, MKK6wt or MKK6m clones (at least three different clones per group) were stimulated with IL-1β (50 units/ml) and IFN-γ (1000 units/ml) (cytokines) for 30 min, lysed and analysed by immunoblotting with phospho-specific antibodies recognizing phosphorylated p38-MAPK (A, D), JNK2 (B, E) and p44/42-ERK (C, F) and total p38, JNK2 and p44/42. The immunoblots shown in (A–C) are representative of five experiments and the means±S.E.M. for these experiments are shown in (D–F). *P<0.05 using ANOVA and Student–Newman–Keuls post-ANOVA test. Abbreviation: Neo, neomycin-resistance.

Figure 4. Lack of effect of stable expression of MKK3/MKK6 on cytokine-induced apoptosis and nitric oxide production

(A) Stable RIN-5AH clones were exposed to cytokines IL-β (50 units/ml) and IFN-γ (1000 units/ml) at 37 °C for 24 h. Cells were then stained with propidium iodide and analysis of cell viability was performed using the FACSCalibur flow cytometer. Results are presented as percentage dead cells of total cell count. Results shown are means±S.E.M. for six independent experiments. (B) L-NMMA (2 mM) was added to all stable clones. At the same time, the cytokines IL-1β (50 units/ml), IFN-γ (1000 units/ml) and TNF-α (1000 units/ml) were added to some groups. Cells were incubated at 37 °C for 72 h and cell viability was analysed using flow cytometry. Results are presented as percentage dead cells of total cell count. Results shown are means±S.E.M. for five independent experiments.

Figure 5. Transient overexpression of MKK3 and MKK6 in sorted RIN-5AH cells

RIN-5AH cells were transfected with pEGFP-C1 and the indicated plasmids using Lipofectamine™. On the next day, the GFP-positive cells were FACS-sorted and re-plated. Cells were analysed, 24 h later, for MKK3 and MKK6 overexpression by immunoblotting. The immunoblot is representative of four experiments.

Figure 6. Effect of MKK3/6 overexpression on basal p38, JNK2 and p44/42 phosphorylation

RIN-5AH cells were transfected, sorted and lysed for immunoblotting as described in the legend to Figure 5. Samples were probed with phospho-specific antibodies recognizing phosphorylated p38-MAPK, JNK2 and p44/42-ERK. Non-phospho-specific ERK antibody was used to visualize total p44/p42 levels. The immunoblot is representative of four experiments.

Figure 7. Effects of transient MKK3/6 overexpression on cytokine-stimulated p38 phosphorylation

(A) Immunoblot showing p38 phosphorylation in response to IL-1β (50 units/ml) and IFN-γ (1000 units/ml) and MKK3/6 overexpression. Two days after transfection and 1 day after FACS, the cells were exposed to the cytokines for 30 min, lysed, separated by SDS gel electrophoresis and analysed by immunoblotting with indicated antibodies. Untreated cells were used as control. (B) Results from immunoblots as the one shown in (A) were quantified by densitometry. Values of phospho-protein bands were related to those of non-phospho-specific protein bands. Results shown are means±S.E.M. for eight independent experiments. *P<0.05 using ANOVA and Student–Newman–Keuls post-ANOVA test. (C) Effects of okadaic acid (OA) and sodium orthovanadate (OV) on cytokine- and MKK3-stimulated p38 phosphorylation. Two days after transfection and 1 day after FACS, cells were pretreated with OA (100 nM) and OV (100 μM) for 10 min. After pretreatment, some groups of cells were exposed to the cytokines IL-β (50 units/ml) and IFN-γ (1000 units/ml) for 30 min. Cells were then lysed and cell proteins were separated by SDS gel electrophoresis and analysed by immunoblotting with p38 and phospho-specific p38 antibodies. The Figure is representative of three observations.

Figure 8. Effect of MKK3/6 overexpression on cytokine-induced apoptosis, nitric oxide production and iNOS levels in transiently transfected cells

(A) Two days after transfection and 1 day after FACS, RIN-5AH cells were exposed to the cytokines IL-β (50 units/ml) and IFN-γ (1000 units/ml) for 24 h. Cells were then stained with propidium iodide and cell viability was assessed using the FACSCalibur flow cytometer. Results are presented as percentage dead cells of the total cell count. Results are means±S.E.M. for six independent experiments. *P<0.05 using ANOVA and Dunnet's post-ANOVA test. (B) Two days after transfection and 1 day after FACS, L-NMMA (2 mM) was added to all the groups of cells and, at the same time, IL-1β (50 units/ml), IFN-γ (1000 units/ml) and TNF-α (1000 units/ml) (cytokines) were added to some groups. Cells were incubated for 72 h and cell viability was assessed using flow cytometry. Results are presented as percentage dead cells of total cell count and are the means±S.E.M. for five independent experiments. (C, D) Two days after transfection and 1 day after FACS, cells were exposed to the cytokines IL-1β (50 units/ml) and IFN-γ (1000 units/ml). Samples were collected the day after addition of cytokines and the levels of nitrite were measured spectrophotometrically (C). Results shown are means±S.E.M. for six independent experiments. Cells were also exposed to the combination of IL-1β (50 units/ml) and IFN-γ (1000 units/ml) for 6 h, lysed, separated on the SDS gel and analysed for iNOS expression by immunoblotting (D).

Figure 9. MKK3 does not affect p38 autophosphorylation in response to IL-1β, DETA/NONOate and sorbitol

Two days after transfection and 1 day after FACS, cells were pretreated with SB203580 (10 μM) for 30 min and then stimulated with IL-1β (50 units/ml), DETA/NONOate (2.5 mM; DETA/NO) and sorbitol (0.4 M) for 30 min. Cells were lysed, separated by SDS gel electrophoresis and analysed by immunoblotting with p38 and phospho-specific p38 antibodies. The Figure is representative of three independent experiments.