NF-kappa B prevents beta cell death and autoimmune diabetes in NOD mice

Abstract

Whereas NF-kappaB has potent antiapoptotic function in most cell types, it was reported that in pancreatic beta cells it serves a proapoptotic function and may contribute to the pathogenesis of autoimmune type 1 diabetes. To investigate the role of beta cell NF-kappaB in autoimmune diabetes, we produced transgenic mice expressing a nondegradable form of IkappaBalpha in pancreatic beta cells (RIP-mIkappaBalpha mice). beta cells of these mice were more susceptible to killing by TNF-alpha plus IFN-gamma but more resistant to IL-1beta plus IFN-gamma than normal beta cells. Similar results were obtained with beta cells lacking IkappaB kinase beta, a protein kinase required for NF-kappaB activation. Inhibition of beta cell NF-kappaB accelerated the development of autoimmune diabetes in nonobese diabetic mice but had no effect on glucose tolerance or serum insulin in C57BL/6 mice, precluding a nonphysiological effect of transgene expression. Development of diabetes after transfer of diabetogenic CD4(+) T cells was accelerated in RIP-mIkappaBalpha/nonobese diabetic mice and was abrogated by anti-TNF therapy. These results suggest that under conditions that resemble autoimmune type 1 diabetes, the dominant effect of NF-kappaB is prevention of TNF-induced apoptosis. This differs from the proapoptotic function of NF-kappaB in IL-1beta-stimulated beta cells.

Fig. 1.

Inhibition of NF-κB activation by mIκBα in β cells. (A) Primary islet cells from transgenic mice (Tg+) and nontransgenic mice (Tg−) were treated with 10–40 ng/ml TNF-α for 30 min, and cell lysates were analyzed by immunoblotting for IκBα. (B) Primary islet cells from Tg+ and Tg− mice were treated with TNF-α for 30 min and then stained with insulin Ab to identify β cells (green), RelA/p65 Ab (red) to determine its subcellular distribution, and Hoechst33342 (blue) to localize nuclei for confocal microscopy. Nuclear translocation of RelA/p65 resulted in magenta nuclei after merging of the red and blue channels (arrowheads, blue nuclei indicating impaired nuclear translocation of RelA/p65; arrows, magenta nuclei indicating RelA/p65 nuclear translocation). Images are representative of three independent experiments.

Fig. 2.

Altered sensitivity of mIκBα-islet cells to cytokine-induced death. (A) Primary islet cells from Tg+ and Tg− mice were incubated with different concentrations of TNF-α and IFN-γ. After 9 days, cell viability was assessed by the MTT assay (∗, P = 0.02; means ± SE from four independent experiments performed in triplicate). (B) Tg+ and Tg− islet cells were treated with TNF-α and IFN-γ for 9 days, and the extent of apoptosis was evaluated by Hoechst33342 staining (∗, P = 0.02; means ± SE from two independent experiments). (C and D) mIκBα-islet cells were treated as in B. They were stained with insulin Ab together with Hoechst33342 and examined by fluorescent microscopy (∗, P = 0.00002; total cell numbers observed were 247 and 199, respectively). (E) mIκBα-islet cells were incubated with 5 ng/ml IL-1β plus 100 units/ml IFN-γ (I5F) for 5 days, and cell viability was assessed by the MTT assay (∗, P = 0.02; means ± SE from six independent experiments performed in triplicate). (F) Tg+ and Tg− islet cells were subjected to I5F treatment for 48 h. Cell lysates were immunoblotted for iNOS expression.

Fig. 3.

Altered sensitivity of IKKβ-deficient β cells to cytokine-induced death. (A) Genomic DNA isolated from pancreatic islets of three Ikkβ^{Δβ cell} mice, liver and lung of one such mouse, and pancreatic islets of a control Ikkβ^F/F mouse were analyzed by PCR to detect the deleted and nondeleted Ikkβ alleles. (B) Islet cells from Ikkβ^{Δβ cell} or Ikkβ^F/F mice were treated with TNF-α and IFN-γ for 9 days, and cell viability was assessed by the MTT assay (∗, P = 0.02; means ± SE from three independent experiments performed in triplicate). (C) Islet cells from Ikkβ^{Δβ cell} or Ikkβ^F/F mice were subjected to I5F treatment for 5 days, and cell viability was assessed by the MTT assay (∗, P = 0.0003; means ± SE from three independent experiments performed in triplicate).

Fig. 4.

Accelerated diabetes in RIP-mIκBα/NOD mice. (A) Female RIP-mIκBα/NOD mice were analyzed for urine and/or blood glucose weekly (●, transgenic, n = 23; ○, nontransgenic, n = 28). (B) Male RIP-mIκBα/NOD mice were followed as above (●, transgenic, n = 30; ○, nontransgenic, n = 30). (C) Development of diabetes was monitored in female mice of another RIP-mIκBα/NOD mouse line (●, transgenic, n = 28; ○, nontransgenic, n = 26). (D) TUNEL staining combined with insulin immunohistochemistry was conducted on pancreatic sections from 8-wk-old female RIP-mIκBα/NOD mice and control littermates. The number of insulin-positive β cells that were TUNEL-positive was determined (∗, P = 0.02; n = 4 each).

Fig. 5.

NF-κB inhibition in β cells increases susceptibility to immune-mediated destruction after adoptive transfer (AT). (A) Effect of TNF-α sequestration on development of diabetes in sublethally irradiated RIP-mIκBα/NOD mice that received lymphocytes from diabetic NOD mice. Mice were treated with 0.5 mg of neutralizing TNF-α Ab or control IgG three times per week for the observation period, and development of diabetes was monitored (●, Ab-treated, n = 21; ○, control IgG, n = 12). (B) Incidence of diabetes after transfer of lymphocytes from diabetic NOD mice was monitored in RIP-mIκBα/NOD mice and nontransgenic littermates (●, transgenic, n = 8; ○, nontransgenic, n = 11). (C) Effect of TNF-α sequestration on development of diabetes in irradiated RIP-mIκBα/NOD mice that received CD4⁺ T cells from diabetic BDC2.5-SCID mice. Mice were treated as in A (●, Ab-treated, n = 8; ○, control IgG, n = 8). The numbers of TUNEL-positive β cells in treated (n = 4) and control mice (n = 7) were determined as in Fig. 4D (means ± SE) and are shown in Inset. (D) Incidence of diabetes after transfer of CD4⁺ T cells from diabetic BDC2.5-SCID mice was monitored in RIP-mIκBα/NOD mice and nontransgenic littermates (●, transgenic, n = 25; ○, nontransgenic, n = 25).

Fig. 6.

Decreased expression of NF-κB-dependent antiapoptotic molecules in β cells from RIP-mIκBα mice. (A) Tg+ and Tg− islet cells maintained in 0.2% serum for 24 h were treated with 10 ng/ml TNF-α for the indicated times. Total cytoplasmic RNA was analyzed by RT-PCR for expression of the indicated mRNAs. (B) Islet cells maintained in 0.2% serum for 24 h were treated with 2 ng/ml TNF-α or 2 ng/ml TNF-α plus 100 units/ml IFN-γ. Total RNA was analyzed by RT-PCR. (C) After pretreatment with 50 μM zVAD-fmk, islet cells were treated in the same way as in B for 48 h. Expression of XIAP and c-FLIP isoforms was analyzed by immunoblotting.