|Accueil | Aperçu | Revues | Soumettre | Nous Contacter | English|
Apoptosis contributes to immune-mediated pancreatic β cell destruction in type I diabetes. Exposure of β cells to interleukin-1β (IL-1β) causes endoplasmic reticulum stress and activates proapoptotic networks. Here, we show that nuclear factor κB (NF-κB) and mitogen-activated protein kinase (MAPK) signaling pathways regulate the expression of CCAAT/enhancer-binding protein homologous protein (CHOP), which mediates endoplasmic reticulum stress-induced apoptosis. Both CHOP mRNA and protein increase in β cells treated with IL-1β. In addition, prolonged exposure to high glucose further increases IL-1β-triggered CHOP expression. IL-1β also causes increased expression of C/EBP-β and a reduction of MafA, NFATc2, and Pdx-1 expression in β cells. Inhibition of the NF-κB and MAPK signaling pathways differentially attenuates CHOP expression. Knocking down CHOP by RNA interference protects β cells from IL-1β-induced apoptosis. These studies provide direct mechanistic links between cytokine-induced signaling pathways and CHOP-mediated apoptosis of β cells.
Type I diabetes is an autoimmune disease characterized by the progressive depletion of pancreatic β cells leading to failed insulin production. During the course of insulitis, proinflammatory cytokines, such as interleukin (IL)3-1β, tumor necrosis factor α, and interferon γ, are released by invasive immune cells (1, 2). Pancreatic β cells are the target of autoimmune assault, and its persistence leads to increased apoptosis and the resultant loss of β cell mass (3, 4). Long term exposure of β cells to cytokines changes the expression of genes necessary for β cell function and promotes the activation of proapoptotic signals (5). Among the variety of proapoptotic networks induced by IL-1β, NF-κB and MAPK pathways play important roles in the pathogenesis of type I diabetes (6, 7).
NF-κB is formed by homo- or heterodimers of five NF-κB family members. When bound to a group of inhibitory κB (IκB) proteins, NF-κB dimers are inactive in the cytoplasm (8). Effects of NF-κB activation on β cell survival are controversial. Tumor necrosis factor α-induced NF-κB activation is reportedly antiapoptotic in pancreatic islet cells (9). However, other reports indicate that NF-κB activation induced by IL-1β either alone or with interferon γ promotes apoptosis in pancreatic β cells. Inhibition of IL-1β-induced NF-κB activation by adenovirus-mediated expression of a phosphorylation-defective mutant of IκBα prevents apoptosis in human and rat primary β cells and insulin-producing cell lines (10). Conditional and specific NF-κB blockade in vivo prevents diabetes caused by streptozotocin-induced β cell loss (11). Microarray and other analyses predicted that IL-1β-induced NF-κB activation increases expression of a variety of proapoptotic genes in β cells, including Fas, c-Myc, and CCAAT/enhancer-binding protein homologous protein (CHOP) (5, 12,–14). However, the molecular mechanisms by which NF-κB modulates transcription of these genes are largely unknown.
JNK is a MAPK activated by cytokines and other environmental stresses. IL-1β activates JNK signaling in β cells (15,–17). Cell-permeable peptide inhibitors of JNK prevent cytokine-induced apoptosis in insulin-producing cells (18). IL-1β also activates ERK1/2 and p38 MAPKs in islets and rat insulinoma cells (15, 19,–21). Beyond inhibiting insulin gene transcription, the actions of MAPKs in response to cytokines are not well characterized in β cells.
CHOP belongs to the CCAAT/enhancer-binding protein (C/EBP) family. It contains an N-terminal transcriptional activation domain and a C-terminal basic-leucine zipper domain. CHOP was identified as a transcription factor mediating endoplasmic reticulum (ER) stress-induced apoptosis. It is ubiquitously expressed at a low level and robustly expressed in many cells under stressful conditions, such as glucose deprivation and depletion of ER calcium stores. CHOP gene induction is modulated through the double-stranded RNA-activated protein kinase-like ER kinase/eukaryotic initiation factor 2α/activating transcription factor 4 (PERK/eIF2α/ATF4), inositol-requiring protein 1α/X-box-binding protein 1 (IRE1α/XBP1), and ATF6α pathways (22). The 5′-flanking region of the CHOP gene contains overlapping cis-acting CAAT/enhancer binding (CEB), ATF, cyclic AMP response element, and other DNA-binding elements. Many transcription factors have been reported to regulate CHOP expression, including C/EBP-β, ATF2, ATF3, XBP1, and ATF4 in different systems (23,–25). We previously observed that in pancreatic β cells, MafA can repress CHOP promoter activity through a MARE (Maf response element), which overlaps with the CEB/ATF/CRE site (26, 27). The ERK1/2 MAPKs control MARE activity on the CHOP gene. IL-1β induces AP-1 dimers composed of c-Fos, c-Jun, and/or JunB to bind to the AP-1 binding sites, which increase CHOP promoter activity (28).
CHOP has been implicated in the pathogenesis of type I diabetes by promoting β cell death (29). CHOP-deficient islets are resistant to the toxic effects of the nitric oxide donor, S-nitroso-N-acetyl-d,l-penicillamine. Akita mice, which bear a point mutant in the insulin II gene with a homozygous deletion of the CHOP gene, show delayed diabetes onset, indicating that CHOP exacerbates the diabetic phenotype (30, 31). CHOP deficiency improves β cell ultrastructure and promotes cell survival in high fat diet-fed eIF2αS/A mice. CHOP-null diabetic Leprdb/db mice display increased expression of the unfolded protein response and antioxidative stress response genes (22).
Because of the large demand on pancreatic β cells to synthesize insulin, they develop a highly functional ER. This makes them especially vulnerable to ER stress caused by hyperglycemia or inflammatory cytokines. IL-1β stimulation depletes ER calcium stores, increases ER stress, and eventually leads to β cell apoptosis (29, 32, 33). A putative NF-κB binding site resides in the rat CHOP gene promoter. Thus, we asked whether CHOP might be a direct target of NF-κB activated by IL-1β. As shown here, CHOP gene expression is directly modulated by NF-κB. Inhibition of the NF-κB pathway reduces CHOP expression and β cell apoptosis.
Rat INS-1 and mouse MIN6 cells were cultured in RPMI 1640 medium and Dulbecco's modified Eagle's medium (Invitrogen) respectively, supplemented with 10% heat-inactivated fetal bovine serum, 11 mm glucose, 2 mm glutamine, 1 mm sodium pyruvate, 50 μm β-mercaptoethanol, and 100 μg/ml each penicillin and streptomycin at 37 °C. Human islets were obtained from ICR Basic Science Islet Distribution Program and cultured in RPMI 1640 medium. Mouse islets were prepared by Donald Brock McNeal and kindly provided by Joyce J. Repa (Department of Physiology, University of Texas, Southwestern). The glucose concentrations were 25 mm for MIN6 and 11 mm for INS-1 cells. After the treatments cells were harvested in cold lysis buffer (50 mm HEPES, pH 7.5, 0.15 m NaCl, 1% Triton X-100, 0.2 mg/ml phenylmethylsulfonyl fluoride, 0.1 m NaF, 2 mm Na3VO4, 10 μg/ml aprotinin, 5 μg/ml pepstatin A, and 5 μg/ml leupeptin). Antibodies were as follows: phospho-IκBα (Ser32/Ser36), phospho-p38 (Thr180/Tyr182), cleaved poly(ADP-ribose) polymerase (PARP), and phospho-JNK (Thr183/Tyr185) from Cell Signaling; c-Jun, phospho-c-Jun (Ser63), c-Maf, c-Fos, NeuroD1, p50, p65, and CHOP from Santa Cruz Biotechnology; Pdx-1 from Upstate; and phospho-ERK1/2 (Thr202/Tyr204) from Sigma. Anti-ERK1/2 polyclonal antibody (Tyr691) was as described (34). Aliquots of cell lysates (20 μg of protein) were subjected to SDS-PAGE. Proteins were electrotransferred to nitrocellulose membranes (Millipore) and blocked with 5% nonfat powdered milk in 20 mm Tris, pH 7.5, 0.15 m NaCl containing 0.1% Tween 20 for 2 h at room temperature and incubated with primary antibody overnight at 4 °C, followed by incubation with horseradish peroxidase-conjugated secondary antibody diluted 1:3000 in the solution above for 30 min at room temperature. Proteins were detected by enhanced chemiluminescence, and signals were quantitated by densitometry.
The pcDNA-IκBαΔN construct was kindly provided by Zhijian Chen (University of Texas Southwestern). Recombinant human IL-1β was purchased from BIOSOURCE (Carlsbad CA). The inhibitors U0126, SP600125, and BAY11-7082 were purchased from Sigma and Biomol (Plymouth Meeting, PA).
SYBR Green-based PCR was performed on the ABI 7500 DNA Sequence Detection System with standard fluorescent chemistries and thermal cycling conditions: 50 °C for 2 min, 95 °C for 10 min for one cycle, an additional 40 cycles at 95 °C for 15 s, 60 °C for 1 min, and 62 °C for 2 min for one cycle. SYBR Green supermix with ROX was purchased from Bio-Rad. 18 S rRNA was used as an internal expression control. The primers are: CHOP forward, 5′-CCTAGCTTGGCTGACAGAGG and reverse, 5′-GGGCACTGACCACTCTGTTT; c-Jun forward, 5′-CTGATCATCCAGTCCAGCAA and reverse, 5′-GACACTGGGAAGCGTGTTCT; c-Fos, forward 5′-CCAGTCAAGAGCATCAGCAA and reverse, 5′-GTACAGGTGACCACGGGAGT; Pdx-1 forward, 5′-TGGATGAAATCCACCAAAGC and reverse, 5′-TTCAACATCACTGCCAGCTC; and 18 S rRNA forward, 5′-TTGACGGAAGGGCACCACCAG and reverse, 5′-GCACCACCACCCACGGAATCG.
INS-1 cells (10-cm dishes) were cross-linked with formaldehyde at a final concentration of 1% for 8 min at room temperature. Reactions were terminated by the addition of 0.125 m glycine for 5 min. Cells were washed twice with cold phosphate-buffered saline and scraped in phosphate-buffered saline followed by centrifugation at 3000 × g for 5 min at 4 °C. Pellets were resuspended in 0.3 ml of 1% SDS, 10 mm EDTA, 50 mm Tris-HCl, pH 8.1, 10 μg/ml aprotinin, 5 μg/ml pepstatin A, 5 μg/ml leupeptin, 0.2 mg/ml phenylmethylsulfonyl fluoride and incubated on ice for 10 min, and 0.7 ml of 0.01% SDS, 1.2 mm EDTA, 16.7 mm Tris-HCl, pH 8.1, 1.1% Triton X-100, 167 mm NaCl, 10 μg/ml aprotinin, 5 μg/ml pepstatin A, 5 μg/ml leupeptin, and 0.2 mg/ml phenylmethylsulfonyl fluoride was added. Lysates were then sonicated using a Sonic Dismembrator 500 (Fisher Scientific) to shear cross-linked chromatin into 500–1000-bp pieces followed by centrifugation at 16,000 × g for 10 min. 2% of the reaction was used as input. Each immunoprecipitation reaction consisted of 0.06 mg of chromatin and 5 μg of antibody with rotation overnight at 4 °C, followed by incubation with 40 μl of protein A-agarose beads and 10 μg of salmon sperm DNA for 2 h at 4 °C. Extensive washing was performed as described previously (35). Precipitates were rotated for 20 min at room temperature in 1% SDS and 0.1 m NaHCO3. To reverse cross-links, 20 μl of 5 m NaCl, 20 μl of 0.5 m EDTA, and 20 μl of Tris-HCl, pH 6.5, were added and incubated at 65 °C overnight. Proteinase K (20 μg) was added for 1 h at 55 °C. DNA was extracted with phenol/chloroform/isoamyl alcohol, precipitated with EtOH, and analyzed by real time PCR. The following primers were used to amplify the rat CHOP gene promoter: forward, 5′-GGGCATAAGAGCATCACACA and reverse, 5′-AACGCTTCCCAAAGACTGAG. The following primers, forward and reverse, respectively, were used to amplify the insulin gene promoter: mInsII, 5′-AACTGGTTCATCAGGCCATC and 5′-ACTGGGTCCCCACTACCTTT; hIns, 5′-GAGGAAGAGGTGCTGACGAC and 5′-CCATCTCCCCTACCTGTCAA (gel analysis); 5′-GTCCTGAGGAAGAGGTGCTG and 5′-CCATCTCCCCTACCTGTCAA (real time PCR analysis).
The Luciferase Reporter Assay System and passive lysis buffer were purchased from Promega (Madison, WI). INS-1 cells were transfected with a luciferase reporter plasmid containing three repeats of wild type (GGAAATGCCAG) or mutant NF-κB binding sites (CCAAATGCCAG or CCAAATGTAAG) from the rat CHOP promoter (beginning with −693 bp) with FuGENE6 HD reagent (Roche Applied Science). After 48 h, cells were harvested with passive lysis buffer supplemented with 100 mm β-glycerophosphate, 2 mm Na3VO4, 100 mm NaF, 10 μg/ml aprotinin, 5 μg/ml pepstatin A, 5 μg/ml leupeptin, 0.2 mg/ml phenylmethylsulfonyl fluoride. Reporter activity in the supernatants was assayed with the Luciferase Assay System using a TD-20/20 bioluminometer (Turner Designs).
INS-1 cells were transfected with 25 nm small interfering RNA oligonucleotides by Lipofectamine 2000 RNAiMax following the manufacturer's protocols (Invitrogen). CHOP small interfering RNA oligonucleotide sequences (Ambion) were: 1, AAACCUUCACUACUCUUGATT (sense) and UCAAGAGUAGUGAAGGUUUTT (antisense); 2, GGCUCAACGAGGAAAUCGATT (sense) and UCGAUUUCCUCGUUGAGUUGC (antisense).
Results were expressed as means ± S.E. determined from three independent experiments. Statistical significance was calculated by ANOVA with the post hoc analysis as in the figure legends.
Recent reports of microarray studies have indicated that IL-1β affects expression of a variety of genes, many of which are predicted to be NF-κB downstream targets in primary rat β cells and β cell lines (5, 12). To evaluate the impact of NF-κB and other signaling pathways, we first examined the effects of IL-1β on the expression of a set of genes involved in β cell function and survival using quantitative real time PCR. MIN6 cells were cultured under standard conditions and treated with 20 ng/ml IL-1β up to 6 h. CHOP mRNA increased after 30 min of IL-1β exposure, reached a peak around 60 min, and gradually returned to the initial level after several hours. The maximum enhancement was about 2.5-fold compared with the control (Fig. 1A). A similar effect was noted in INS-1 cells. Transcription of c-Jun started to increase at 10 min and dramatically increased to 16-fold over the control 30 min after the addition of IL-1β (Fig. 1B). IL-1β treatment augmented c-Fos transcription about 3-fold with similar kinetics (Fig. 1C). Pdx-1, NeuroD1 (BETA2), and MafA have been shown to activate the insulin gene promoter synergistically, whereas C/EBP-β and c-Jun inhibit the insulin promoter (36,–39). As one test of whether IL-1β impairs β cell function, the expression of insulin gene transcription factors was monitored. Pdx-1 mRNA was decreased to approximately 30% of the control at 2 h (Fig. 1D). However, mRNAs encoding neither MafA nor NeuroD1 were significantly affected. There was a slight increase in both insulin I and II mRNA observed within 6 h of IL-1β treatment (supplemental Fig. 1, C and D).
Consistent with the changes in amounts of mRNAs, immunoblotting showed that CHOP expression was low in untreated β cells, dramatically enhanced after 24 h of IL-1β stimulation, and sustained for at least 48 h (Figs. 1E and and22G). It has been reported that low IL-1β concentrations (0.01–0.02 ng/ml) stimulate proliferation of human islets and decrease apoptosis, whereas increasing amounts of IL-1β have the opposite effect (40). We found that as little as 1 ng/ml IL-1β was sufficient to promote CHOP expression and that the induction of CHOP was concentration-dependent (Fig. 1E). Expression of c-Jun was detected in untreated β cells. An increase was observed around 10 min followed by a peak at 60 min and then a return to untreated amounts around 24 h. A doublet of c-Jun, MafA, and some of the other factors appears due to phosphorylation of the induced proteins by MAPKs and other kinases. The induction of c-Jun expression by IL-1β was also concentration-dependent (Fig. 1F and data not shown). In a manner similar to c-Jun, c-Fos expression increased to a small extent at 60 min of IL-1β treatment (Figs. 1G and and22E). Given the well established link between insulin transcription factors and β cell function, we also examined their expression. Pdx-1, NFATc2, and MafA were down-regulated by 24 h. NeuroD1 was also reduced but to a lesser extent (Fig. 1, H and I). In contrast, expression of C/EBP-β, an inhibitory insulin gene transcription factor, was enhanced after 2 h of stimulation (Fig. 1J). The association of C/EBP-β on the insulin gene promoter was detected after as little as 2 h of IL-1β treatment by ChIP assay in both human islets and MIN6 cells (Fig. 1, K and L). These data suggest that IL-1β impairs β cell function through both down-regulation of insulin gene transcription activators and up-regulation of inhibitory factors.
To explore the signaling pathways involved in the regulation of CHOP expression triggered by IL-1β, we examined pathways known to be activated by cytokines. Phosphorylation of IκBα was utilized to monitor the activation of NF-κB. Phospho-IκBα was barely detectable in untreated β cells, indicating that NF-κB activation was low under normal conditions (Fig. 2A). The phosphorylated protein was easily detected within 5 min of treatment with 1 ng/ml IL-1β. Activation of NF-κB was concentration-dependent, biphasic, and returned to nearly undetectable levels after 24 h (Fig. 2A). Immunoblotting revealed that ERK1/2, JNK, and p38 were all activated by IL-1β in β cells but with distinct kinetics (Fig. 2, B–D). Phospho-p38 was detected as early as 5 min after IL-1β addition and persisted for up to 24 h, whereas phospho-JNK reached a peak around 30 min and rapidly returned to basal levels. IL-1β induced biphasic activation of ERK1/2, with activity detected after 24 h. We also observed that the expression of the ERK1/2 downstream target c-Fos reached a peak around 3 h and that the maximum amount of the JNK substrate c-Jun was noted around 60 min (Fig. 2, E and F). SP600125 (20 μm), a JNK inhibitor, reduced JNK activity and also attenuated total c-Jun, phospho-c-Jun, and modestly reduced CHOP expression; U0126 (10 μm), a MEK1/2 inhibitor, blocked ERK1/2 activity and reduced c-Fos and to a small extent CHOP expression (Fig. 2, E–G, and supplemental Fig. 3A). None of the MAPK inhibitors had any effect on NF-κB activation (supplemental Fig. 3A). To block NF-κB activation, a pharmacological inhibitor BAY11-7082 was utilized. It partially reduced IL-1β-induced CHOP expression, indicating that CHOP might be a downstream target of the NF-κB pathway (Fig. 2G). BAY11-7082 also prevented the IL-1β-induced reduction in the concentration of NFATc2 and appeared to attenuate the effect on MafA (Fig. 2G). These results are consistent with earlier reports that the NF-κB pathway is involved in controlling CHOP expression and β cell function.
As noted above, CHOP was predicted to be a downstream target of IL-1β-activated NF-κB. Because there is a putative NF-κB binding site in the 5′-flanking region of the CHOP gene, we tested whether NF-κB could bind directly to the CHOP promoter. ChIP assays showed that IL-1β stimulation caused a statistically significant increase in binding of the NF-κB subunit p50 to the CHOP promoter (Fig. 3A and supplemental Fig. 2A) which was reduced by BAY11-7082. Although p65 binding appeared to be increased, the difference was not statistically significant, and no differences were found in binding of other subunits. Consistent with these results, the increase in CHOP gene expression caused by a 1-h exposure to IL-1β was prevented by treatment with BAY11-7082 (Fig. 3B). As further support, we used reporter assays, transfecting INS-1 cells with plasmids containing either wild type or mutant NF-κB binding sites from the rat CHOP promoter. IL-1β stimulated luciferase activity in β cells transfected with a plasmid harboring the wild type binding site. Promoter activity was completely abolished after IL-1β treatment of cells expressing the reporter linked to a mutant binding site (Fig. 3D). Cotransfection of the NF-κB super suppressor, which lacks the N-terminal domain of IκBα and is not readily degraded, inhibited IL-1β-induced reporter gene expression. BAY11-7082 also partially blocked the effect (Fig. 3C). These observations are in agreement with the idea that CHOP gene expression in β cells is directly regulated by NF-κB.
Because β cells are glucose sensors, glucose is a key factor in maintaining β cell function and survival. To examine whether glucose affects IL-1β-induced CHOP expression, INS-1 cells were preincubated in medium containing different glucose concentrations for 24 h. Immunoblotting showed that 11 or 16 mm glucose strongly enhanced CHOP protein expression (Fig. 4A). However, β cells in high glucose produced little or no more CHOP mRNA (Fig. 4D). IL-1β had similar effects on CHOP transcription independent of glucose preincubation conditions. To explore possible mechanisms of protein induction, we considered pathways sensitive to glucose. NF-κB activation was similar regardless of glucose concentration (supplemental Fig. 3B). Previous studies showed that glucose has little effect on JNK or p38 (15). On the other hand, ERK1/2 activities are much higher in 11 or 16 mm glucose than in low glucose (Fig. 4B). The MEK1/2 inhibitor U0126 attenuated IL-1β-induced CHOP expression in INS-1 cells cultured in high glucose overnight (Fig. 4C). These results suggest that sustained activation of ERK1/2 caused by high glucose may alter the translation of CHOP mRNA or CHOP protein stability.
Apoptosis causes β cell loss in type I diabetes. Activation of the NF-κB pathway is antiapoptotic in most cell types, but has been suggested to be both pro- and antiapoptotic in pancreatic β cells (6, 9). To examine whether IL-1β-activated NF-κB leads to β cell apoptosis, we utilized cleaved PARP as a marker of apoptosis. Few INS-1 cells underwent apoptosis before or after IL-1β treatment for 24 h. A greater increase in PARP cleavage was observed following a 48-h incubation. In mouse islets, a basal level of PARP cleavage was observed without IL-1β, and the cytokine markedly increased cleavage within 24 h (supplemental Fig. 3D). To explore further whether CHOP expression is involved in β cell apoptosis, we knocked down CHOP expression by RNA interference. Both real time PCR and immunoblotting confirmed reduced CHOP expression by small interfering RNA oligonucleotides (Fig. 5A and data not shown). β cells with less CHOP expression displayed reduced IL-1β-induced β cell apoptosis. Addition of the NF-κB inhibitor BAY11-7082 partially attenuated PARP cleavage in INS-1 cells and in mouse islets (Fig. 5B and supplemental Fig. 3D). These results suggest that IL-1β-activated NF-κB is proapoptotic in insulin-producing cells. Furthermore, the findings are consistent with the idea that CHOP plays an important role in β cell apoptosis.
Pancreatic β cells are selectively destroyed by the immune system in type I diabetes through multiple signaling pathways (2, 4). Infiltrating immune cells and also β cells themselves are able to express IL-1β (41). CHOP expression is increased in β cells from both diabetic mice and human (42). Because CHOP mediates ER stress-induced apoptosis, the regulation of the CHOP gene appears to be critical in determining β cell fate. Although CHOP-deficient mice do not display any detectable phenotype under basal conditions, β cells isolated from CHOP-deficient mice are protected from apoptosis caused by either nitric oxide or the accumulation of a folding-defective mutant of proinsulin (30, 31). Indirect effects of NF-κB on CHOP expression have been identified. For example, ATF4, a downstream target of NF-κB, is a transcription factor that increases expression of the CHOP gene (28). Our studies support the conclusion that NF-κB also has a direct effect on CHOP expression. NF-κB binds to the region of the CHOP gene promoter containing the NF-κB consensus site; mutation of that element reduces sensitivity of a reporter to IL-1β, and inhibition of NF-κB partially blocked CHOP expression. Thus, the NF-κB signaling pathway has both indirect and direct inputs to CHOP expression in β cells.
The outputs of NF-κB, like many other essential signaling pathways, are complicated. For example, many types of human tumors have constitutively active NF-κB (43). Active NF-κB up-regulates the expression of genes that maintain cancer cell proliferation and protect them from conditions that would otherwise cause apoptosis. Defects in NF-κB result in increased susceptibility to apoptosis because NF-κB regulates antiapoptotic genes, such as TRAF1 and TRAF2 (tumor necrosis factor receptor-associated factor), thereby controlling the activities of caspases that are essential in most apoptotic processes (8, 44). In contrast, in pancreatic β cells, the gene expression profile caused by cytokine-induced NF-κB activation is predominantly proapoptotic (5, 45). The opposite effects of NF-κB in these different settings may due to the different activators and the involvement of different death effectors and may also relate to the sensitivity of β cells to ER stress. The balance between β cell survival and apoptosis mediated by a variety stimuli critical in the onset of type I diabetes may involve properties and mechanisms triggered by NF-κB activation.
The negative impact on insulin gene transcriptional activators and the rapid induction of inhibitory factors might lead to the expectation that IL-1β would also rapidly impair insulin gene transcription. However, insulin mRNA did not significantly change over 6 h of IL-1β stimulation. We found previously that IL-1β begins to reduce insulin gene transcriptional elongation after 6 h. Thus, it seems likely that a short exposure to IL-1β was not sufficient to deplete the huge preexisting insulin mRNA pool. The fact that NF-κB inhibition largely restored the amount of essential insulin gene transcriptional activators is further support that in response to IL-1β, NF-κB impairs β cell function.
MAPKs have also been implicated in β cell loss. Inhibition of JNK by chemical blockers or inhibitory peptides protects β cells from apoptosis, indicating that JNK signaling contributes to IL-1β-induced β cell apoptosis (46). We found that IL-1β activates JNK and its downstream target c-Jun in β cells. Surprisingly, JNK activation was the most transient of the pathways examined, lasting only about 30 min. Confirming earlier findings, a JNK inhibitor blocked JNK activation and attenuated c-Jun and CHOP expression. Gurzov et al. (47) showed that JunB protects β cells against IL-1β- and interferon γ-mediated apoptosis through the inhibition of inducible nitric-oxide synthase and CHOP (47). Further studies are needed to identify how active JNK regulates the components of the AP-1 complex on the CHOP gene promoter.
In contrast to JNK, activation of p38 and ERK1/2 was sustained for hours. p38 phosphorylates CHOP on Ser78/Ser81, consistent with its prolonged activity (48). ERK1/2 have a potential inhibitory effect on CHOP expression via MafA and a stimulatory effect through c-Fos binding to the AP-1 sites on the CHOP promoter (26). Interestingly, prolonged activation of ERK1/2 was associated with an increase in CHOP protein, suggesting potential actions on CHOP translation or stability. Finally, inhibitors of the three MAPK pathways had little effect on NF-κB activation, suggesting that they act parallel to NF-κB under these conditions.
In conclusion, our findings provide further support for the idea that CHOP plays a direct role in IL-1β-triggered pancreatic β cell apoptosis, which is accomplished through the regulation of NF-κB as well as MAPK pathways. Therefore, CHOP might be a potential target for preventing β cell death. Further studies will focus on the molecular mechanisms by which CHOP initiates β cell apoptosis.
We thank Zhijian (James) Chen for constructs, Joyce Repa for mouse islets, Eric Wauson and Jihan Osborne for comments on the manuscript, Kathy McGlynn for technical assistance, and Dionne Ware for administrative assistance.
3The abbreviations used are: