help button home button Endocrine Society Molecular Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Molecular Endocrinology, doi:10.1210/me.2005-0268
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
20/8/1867    most recent
Author Manuscript (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ortis, F.
Right arrow Articles by Eizirik, D. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ortis, F.
Right arrow Articles by Eizirik, D. L.
Molecular Endocrinology 20 (8): 1867-1879
Copyright © 2006 by The Endocrine Society

Cytokine-Induced Proapoptotic Gene Expression in Insulin-Producing Cells Is Related to Rapid, Sustained, and Nonoscillatory Nuclear Factor-{kappa}B Activation

Fernanda Ortis, Alessandra K. Cardozo, Daisy Crispim, Joachim Störling, Thomas Mandrup-Poulsen and Décio L. Eizirik

Laboratory of Experimental Medicine (F.O., A.K.C., D.C., D.L.E.), Université Libre de Bruxelles, 1070 Brussels, Belgium; Laboratory for ß-Cell Biology (J.S., T.M.-P.), Steno Diabetes Center, DK-2820 Gentofte, Denmark; and Department of Molecular Medicine (T.M.-P.), Karolinska Institute, SE-17177 Stockholm, Sweden

Address all correspondence and requests for reprints to: Professor Décio L. Eizirik, Laboratory of Experimental Medicine, Université Libre de Bruxelles, Route de Lennik, 808-CP-618, 1070 Brussels, Belgium. E-mail: deizirik{at}ulb.ac.be.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Cytokines, such as IL-1ß and TNF-{alpha}, contribute to pancreatic ß-cell death in type 1 diabetes mellitus. The transcription factor nuclear factor-{kappa}B (NF-{kappa}B) mediates cytokine-induced ß-cell apoptosis. Paradoxically, NF-{kappa}B has mostly antiapoptotic effects in other cell types. The cellular actions of NF-{kappa}B depend on the cell type, the nature and duration of the stimulus, the periodicity, and the degree of activity of the particular dimers involved. To clarify the reasons behind the proapoptotic effects of NF-{kappa}B in pancreatic ß-cells, we compared the pattern of cytokine-induced NF-{kappa}B activation between rat insulin-producing cells (INS-1E cells) and fibroblasts (208F cells). NF-{kappa}B activation was induced in INS-1E cells and in 208F cells after exposure to cytokines, but apoptosis was induced only in INS-1E cells, with a more pronounced proapoptotic effect of IL-1ß than of TNF-{alpha}. NF-{kappa}B activation in IL-1ß-exposed INS-1E cells was earlier and more marked as compared with TNF-{alpha}-exposed INS-1E cells or IL-1ß-exposed 208F cells. Both cytokines induced a prolonged (up to 48 h) and stable NF-{kappa}B activation in INS-1E cells, whereas IL-1ß induced an oscillatory NF-{kappa}B activation in 208F cells. p65/p65 and p65/p50 were the predominant NF-{kappa}B dimers in IL-1ß-exposed INS-1E cells and 208F cells, respectively. IL-1ß induced a differential usage of cis-elements in the inducible nitric oxide synthase promoter region in the two cell-lines and an increase in ERK1/2 activity in INS-1E cells but not in 208F cells. Cytokine-induced expression of I{kappa}B isoforms and other NF-{kappa}B target genes (Fas, MCP-1, and inducible nitric oxide synthase) was severalfold higher in INS-1E cells than in 208F cells. These results suggest that cytokine-induced NF-{kappa}B activation in insulin-producing cells is more rapid, marked, and sustained than in fibroblasts, which correlates with a more pronounced activation of downstream genes and a proapoptotic outcome.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
PANCREATIC ß-CELLS ARE the target of an autoimmune assault in type 1 diabetes mellitus (T1D) (1). During the ensuing inflammatory reaction, named "insulitis," proinflammatory cytokines such as IL-1ß, TNF-{alpha}, and interferon (IFN)-{gamma} are released in the vicinity of ß-cells by activated macrophages and T cells, contributing to ß-cell dysfunction and death (reviewed in Ref. 2). Under in vitro conditions, IL-1ß induces functional ß-cell impairment and, in combination with IFN-{gamma} and/or TNF-{alpha}, causes ß-cell death, mostly by apoptosis (2). Apoptosis is probably the main cause of ß-cell death at the onset of T1D (2, 3, 4, 5) and following islet transplantation (2, 6, 7).

The transcription factor nuclear factor-{kappa}B (NF-{kappa}B) (8, 9) has an important role in cytokine-induced ß-cell apoptosis, because apoptosis is prevented in human and rat primary ß-cells, and in insulin-producing cell lines, by inhibition of NF-{kappa}B activation (10, 11, 12). NF-{kappa}B blocking by an I{kappa}B{alpha}(S/A)2 superrepressor also prevents rat ß-cell apoptosis induced by IFN-{gamma} + double-stranded RNA (a byproduct of viral infection) (13). Furthermore, conditional and specific NF-{kappa}B blockade in vivo protects pancreatic ß-cells against toxic/immuno-mediated diabetes after multiple low doses of streptozotocin (14). Paradoxically, NF-{kappa}B-regulated genes have been shown to inhibit apoptosis in diverse cell types (15, 16, 17, 18, 19, 20).

NF-{kappa}B is formed by homo- or heterodimers of five NF-{kappa}B family members (8). These dimers are usually present in an inactive form in the cytoplasm, where they remain bound to a group of related inhibitory {kappa}B (I{kappa}B) proteins (8, 21). Stimulation by proinflammatory cytokines, such as IL-1ß and TNF-{alpha}, results in NF-{kappa}B activation through the classical pathway (22), leading to degradation of the I{kappa}B proteins and consequent translocation of NF-{kappa}B to the nucleus (21, 22). The set of genes and the cellular responses induced by NF-{kappa}B are determined by both the activation of specific dimers and by the duration, periodicity, and level of activity of the particular dimer involved (23, 24, 25). The different isoforms of I{kappa}B play distinct and complementary roles in the regulation of specific NF-{kappa}B dimers (15, 16, 22, 26), resulting in both quantitative and qualitative changes in gene expression (23, 27). These characteristics are specific for different cell types and might vary in a particular cell type exposed to different stimuli (9, 15). Because no information is available regarding these parameters in ß-cells, it remains to be determined whether NF-{kappa}B activation in these cells has specific activation characteristics that cause it to exert a proapoptotic effect, whereas it exerts an antiapoptotic effect in most other cell types.

In the present study, we performed time course studies to compare the pattern of cytokine-induced NF-{kappa}B activation between INS-1E cells [a well-differentiated ß-cell line; NF-{kappa}B is proapoptotic in ß-cells (10, 11, 12, 13, 28)] exposed to either IL-1ß or TNF-{alpha}, and in rat fibroblasts, in which NF-{kappa}B is antiapoptotic (19, 20). Cytokines induced NF-{kappa}B activation in both INS-1E and rat fibroblast 208F cells, but apoptosis was induced only in INS-1E cells, with IL-1ß showing a higher proapoptotic effect than TNF-{alpha}. NF-{kappa}B activation in INS-1E cells exposed to IL-1ß has distinct characteristics as compared with 208F cells. These include an earlier, stronger, stable and prolonged NF-{kappa}B activation, differential composition of NF-{kappa}B members in the activated complex, differential expression of NF-{kappa}B-regulated genes, differential usage of cis-elements in the promoter region of target genes, and induction of ERK1/2 activity, known to enhance the transcriptional activity of p65 (29). These observations suggest that IL-1ß-induced NF-{kappa}B activation has a specific pattern in INS-1E cells, which may explain the surprising proapoptotic effect of this transcription factor in ß-cells.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
INS-1E Cells, But Not 208F Cells, Are Sensitive to Cytokine-Induced Cell Death
We first examined the effects of IL-1ß or TNF-{alpha} on INS-1E cells viability. IL-1ß alone increased cell death by apoptosis in INS-1E cells after 24 h and 48 h of exposure (Fig. 1AGo), with no significant changes in the percentage of necrotic cells (data not shown). On the other hand, exposure of INS-1E cells to a 20-fold higher concentration of TNF-{alpha} alone failed to increase the percentage of apoptotic cells (Fig. 1AGo). In subsequent experiments we tested whether a combination of IL-1ß with TNF-{alpha} potentiated induction of apoptosis in INS-1E cells. After a 48-h exposure to IL-1ß (100 U/ml) + TNF-{alpha} (100 U/ml) or IL-1ß (100 U/ml) + TNF-{alpha} (1000 U/ml), respectively, the observed apoptotic index was 1.3 ± 0.2 and 3.3 ± 0.7 (n = 4) as compared with 2.4 ± 0.5 (n = 4) in cells exposed to IL-1ß (100 U/ml) alone. Thus, TNF-{alpha} did not potentiate IL-1ß-induced apoptosis in INS-1E cells under the present experimental conditions. IFN-{gamma} (100 U/ml) alone induced cell death, with an apoptotic index of 4.0 ± 1.3 and 9.0 ± 1.0 after 24 h and 48 h, respectively. This effect was amplified severalfold by combination with either IL-1ß or TNF-{alpha} (Fig. 1BGo). The combination of IL-1ß (100 U/ml) with IFN-{gamma} induced a higher apoptotic index than the combination of TNF-{alpha} (100 U/ml or 500 U/ml) + IFN-{gamma} (Fig. 1BGo), confirming that IL-1ß has a more potent proapoptotic activity in INS-1E cells than TNF-{alpha}. On the other hand, neither IL-1ß alone nor combinations of IL-1ß (100 U/ml) + IFN-{gamma} or IL-1ß (100 U/ml) + IFN-{gamma} + TNF-{alpha} (1000 U/ml) affected 208F cells viability. Thus, the apoptotic index in 208F cells was lower than 0.6 at 48 h under all the different conditions tested (n = 3–5). After transfection with an adenoviral construct encoding an I{kappa}B{alpha}(S/A)2 superrepressor protein [same viral construct used in our previous experiments with ß-cells (10, 13, 30)]; however, apoptosis was induced in 208F cells by a combination of IL-1ß + TNF-{alpha} + IFN-{gamma}, with an apoptotic index 6.4 ± 0.8 (n = 6; P < 0.001 vs. control, noninfected cells). AdI{kappa}B{alpha}(S/A)2 alone did not modify apoptosis (apoptotic index, 0.6 ± 0.3; n = 6), and a control adenovirus, encoding the enhanced green fluorescent protein neither induced apoptosis alone nor potentiated cytokine-induced apoptosis in 208F cells (data not shown). This proapoptotic effect of NF-{kappa}B blocking in the presence of cytokines in 208F cells is in marked contrast with our data, and data from other groups, showing that a similar NF-{kappa}B blocker protects primary ß-cells (rodent and human) and insulin-producing cells against apoptosis induced by IL-1ß (11), IL-1ß + IFN-{gamma} (10, 14), double-stranded RNA + IFN-{gamma} (13), and IL-1ß + TNF-{alpha} + IFN-{gamma} (12).


Figure 1
View larger version (17K):
[in this window]
[in a new window]
  		Fig. 1. Cytokine-Induced Apoptosis in INS-1E Cells

A, INS-1E cells exposed to IL-1ß (50 or 100 U/ml) or TNF-{alpha} (1000 U/ml) for 24 h (white bars) or 48 h (black bars). B, INS-1E cells exposed to IL-1ß (100 U/ml) or TNF-{alpha} (100, 500, or 1000 U/ml) in combination with IFN-{gamma} (100 U/ml) for 48 h. Cell viability was determined with the DNA-binding dyes HO 342 and propidium iodide (PI), and the apoptotic index was calculated as described in Materials and Methods (note the different scales in panels A and B). Data are the means ± SEM of three to six experiments. *, P ≤ 0.05; **, P ≤ 0.01 vs. control, paired t test; #, P < 0.01 vs. IL-1ß + IFN-{gamma}, ANOVA.

 
Cytokines Induce a Differential Kinetics of NF-{kappa}B Activation in INS-1E and 208F Cells
To determine whether IL-1ß induces a different NF-{kappa}B activation kinetics in pancreatic ß-cells, we compared INS-1E cells exposed to IL-1ß or TNF-{alpha} with 208F cells exposed to IL-1ß during time course experiments (Fig. 2Go). Both cytokines induced NF-{kappa}B activation in INS-1E cells (Fig. 2Go), but IL-1ß induced an earlier and more pronounced peak of NF-{kappa}B activation after 10 min than TNF-{alpha} (P < 0.01). After 30 min, however, IL-1ß and TNF-{alpha} induced a similar and sustained increase in NF-{kappa}B activation, lasting for at least 48 h. In contrast, IL-1ß-induced NF-{kappa}B activation in 208F cells followed an oscillatory pattern, with a first peak of activation at 30 min and a second peak at 1 h, followed by stabilization of NF-{kappa}B activation between 90 min and 12 h, and then a decrease at 24 h. This is in good agreement with recent observations that cytokines induce an oscillatory pattern of NF-{kappa}B activation in fibroblasts (23). The above-described intensity and kinetics of NF-{kappa}B activation in INS-1E cells (tested at 10 min, 30 min, and 24 h) were not altered by the following combination of cytokines: IL-1ß + IFN-{gamma} or IL-1ß + TNF-{alpha}, as compared with IL-1ß alone; and IFN-{gamma} + TNF-{alpha} as compared with TNF-{alpha} alone (data not shown). IFN-{gamma} alone did not induce NF-{kappa}B activation (data not shown).


Figure 2
View larger version (15K):
[in this window]
[in a new window]
  		Fig. 2. NF-{kappa}B Activation Kinetics in INS-1E and 208F Cells during Persistent Exposure to Cytokines

INS-1E cells were exposed to 100 U/ml IL-1ß (black diamond) or 1000 U/ml TNF-{alpha} (open square), whereas 208F cells were exposed to 100 U/ml IL-1ß (open triangle) at the indicated time points. OD values (arbitrary units) of specific NF-{kappa}B mobility shift were quantified, normalized (as described in Materials and Methods), and graphed at the indicated nonlinear time scale. The results shown represent means ± SE for three to four independent experiments.

 
Differential Molecular Composition of Activated NF-{kappa}B in INS-1E and 208F Cells after Cytokine Exposure
To analyze the molecular composition of cytokine-activated NF-{kappa}B dimers in our experimental models, we selected five time points (10 min, 45 min, 2 h, 12 h, and 24 h) of the time course experiments. Two complexes were induced in INS-1E cells exposed to cytokines, with the first complex ("a"; indicated by an arrow in Fig. 3Go) already present after 10 min of exposure. In IL-1ß-treated INS-1E cells (INS-IL), an antibody directed against the NF-{kappa}B member p65 ({alpha}-p65) completely supershifted this complex at 10 min (lane 3 in Fig. 3Go, A and B), 45 min (Fig. 3AGo, lane 6, and Fig. 3BGo, lane 8), 2 h (data not shown), 12 h (Fig. 3AGo, lane 9) and 24 h (lane 13 in Fig. 3Go, A and B). Interestingly, complex a was only partially supershifted by an antibody against the NF-{kappa}B member p50 ({alpha}-p50) (Fig. 3AGo, lanes 4, 7, 10, and 14; and Fig. 3BGo: lanes 4, 9, and 14), mostly after 24 h of IL-1ß treatment. In INS-1E cells treated with TNF-{alpha} (INS-TNF), complex a was totally supershifted by anti-p65 (Fig. 3AGo, lanes 3, 6, 9, and 13; and Fig. 3BGo, lanes 3, 8, and 13) but not by anti-p50 (Fig. 3AGo, lanes 4, 7, 10, and 14; and Fig. 3BGo, lanes 4, 9, and 14), at all time points analyzed. The complex "b" in INS-1E cells (indicated by an arrow in Fig. 3AGo and 3BGo), which is stronger after 45 min of exposure to either IL-1ß or TNF-{alpha}, was not supershifted by anti-p65 at the different time points studied (Fig. 3AGo, lanes 6, 9, 13; and Fig, 3BGo, lanes 3, 8, and 13). On the other hand, anti-p50 completely supershifted this band after 45 min (Fig. 3AGo, lane 7; and Fig. 3BGo, lane 9), 2 h (data not shown), 12 h (Fig. 3AGo, lane 10), and 24 h (lane 14 in Fig. 3Go, A and B). Neither complex a nor complex b were supershifted by antibodies directed against the NF-{kappa}B family members p52 (Fig. 3BGo, lanes 5, 10, and 15) and c-Rel (Fig. 3BGo, lanes 6, 11, and 16). The NF-{kappa}B complex a was also present in IL-1ß-treated 208F cells (208F-IL) at 10 min and remained detectable up to 24 h (Fig. 3AGo, indicated by arrow). This complex was totally supershifted by both anti-p65 (Fig. 3AGo, lanes 3, 6, 9, and 13) and anti-p50 (Fig. 3AGo, lanes 4, 7, 10, and 14) antibodies, at all time points analyzed. Of note, complex b was totally supershifted in 208F cells only by the anti-p50 antibody (Fig. 3AGo, lanes 7, 10, and 14). The specificity of these complexes was confirmed by competition with excess of an unlabeled {kappa}B-binding site probe (Fig. 3AGo, lane 15; and Fig. 3BGo, lane 17). Moreover, an unrelated antibody against signal transducer and activator of transcription (STAT)1 (Fig. 3AGo, lane 11) or interferon regulatory factor (IRF)-1{alpha} (Fig. 3BGo, lane 18) had no effect on the migration of the two complexes.


Figure 3
View larger version (82K):
[in this window]
[in a new window]
  		Fig. 3. Molecular Composition of Cytokine-Activated NF-{kappa}B in INS-1E and 208F Cells

Experiments were performed with INS-1E cells exposed to IL-1ß (100 U/ml) (INS-IL) or TNF-{alpha} (1000 U/ml) (INS-TNF) and 208F cells exposed to IL-1ß (100 U/ml) (208F-IL) at the indicated time points. Control cells were left untreated. The specificity of NF-{kappa}B mobility shift (indicated by arrows) was confirmed by competition with the nonradioactive (cold) {kappa}B-binding site oligonucleotide at 24 h (panel A, lane 15; and panel B, lane 17) and by use of antibodies directed against the NF-{kappa}B proteins p65 (panel A, lanes 3, 6, 9, and 13; and panel B, lanes 3, 8, and 13), p50 (panel A, lanes 4, 7, 10, and 14; and panel B, lanes 4, 9, and 14), p52 (panel B, lanes 5, 10, and 15) and c-Rel (panel B, lanes 6, 11, and 16). The specificity of the supershifted bands was tested by the use of an unrelated antibody against STAT-1 (panel A, lane 11) and IRF1 (panel B, lane 18). The figure is representative of three (panel A) or two (panel B) similar experiments.

 
These results suggest that in INS-1E cells IL-1ß or TNF-{alpha} induce a complex preferentially constituted by p65 homodimers, with a minor contribution by the p65/p50 heterodimer. Complex b, also induced by both cytokines in INS-1E cells, appears to be preferentially composed by p50 homodimers. 208F cells exposed to IL-1ß have a different pattern of NF-{kappa}B activation, with p65/p50 as the predominant NF-{kappa}B dimer detected in complex a at all time points studied.

Characterization of the Kinetics of Cytokine-Induced Expression/Degradation of the I{kappa}B{alpha}, -ß, and -{epsilon} Isoforms
In mammalian cells the major regulatory I{kappa}B proteins are {alpha}, ß, and {epsilon} (8), and it has been shown that the functional characteristics of these isoforms are primarily the result of differences in the kinetics of their degradation and resynthesis (23). We therefore analyzed the kinetics of expression of I{kappa}B{alpha}, I{kappa}Bß, and I{kappa}B{epsilon} at the mRNA and protein levels in INS-1E and 208F cells after cytokine exposure. The three I{kappa}B isoforms ({alpha}, ß, and {epsilon}) are expressed in both INS-1E and 208F cells, as evidenced by real-time RT-PCR (Fig. 4Go, A–C). IL-1ß-treated INS-1E cells had a higher I{kappa}B{alpha} mRNA expression (peak of fold induction at 45 min of 32 ± 5.3) than TNF-{alpha}-treated INS-1E cells (fold induction at 45 min of 10.7 ± 4.5; P < 0.05 vs. IL-1ß); this difference was more pronounced when comparing INS-1E and 208F cells treated with IL-1ß (fold induction at 45 min of 1.6 ± 0.3 for 208F cells; P < 0.02 vs. INS-1E cells) (Fig. 4AGo). IL-1ß and TNF-{alpha} induced similar kinetics of I{kappa}Bß and I{kappa}B{epsilon} mRNA expression in INS-1E cells (Fig. 4Go, B and C, respectively), but IL-1ß tended to induce a more pronounced expression of these mRNAs. Cytokine-induced I{kappa}Bß and I{kappa}B{epsilon} mRNA expression in INS-1E cells (Fig. 4Go, B and C, respectively) was delayed and of lower magnitude when compared with I{kappa}B{alpha} mRNA expression (Fig. 4AGo; note the different scales in Fig. 4Go, A–C). In 208F cells, induction of I{kappa}B{alpha}, I{kappa}Bß, and I{kappa}B{epsilon} mRNA by IL-1ß was lower than in INS-1E cells at most time points studied (Fig. 4Go, A–C). In summary, intensity of induction of the three I{kappa}B isoforms was the highest in INS-1E cells treated with IL-1ß, followed by INS-1E cells exposed to TNF-{alpha} and 208F cells treated with IL-1ß, respectively.


Figure 4
View larger version (28K):
[in this window]
[in a new window]
  		Fig. 4. Time Course Analysis of IL-1ß and TNF-{alpha}-Induced Expression/Degradation of the I{kappa}B{alpha}, I{kappa}Bß, and I{kappa}B{epsilon} Isoforms in INS-1E and 208F Cells

INS-1E cells were treated with IL-1ß (100 U/ml) (black diamonds) or TNF-{alpha} (1000 U/ml) (open squares), and 208F cells were treated with IL-1ß (100 U/ml) (open triangles) at the indicated time points. A–C, Real-time RT-PCR was performed with specific primers for I{kappa}B{alpha} (A), I{kappa}Bß (B), and I{kappa}B{epsilon} (C), corrected by GAPDH expression. The data are expressed as fold variation of the respective controls (considered as 1). D–F, OD values of I{kappa}B isoforms recognized by antibodies specific against I{kappa}B{alpha} (D), I{kappa}Bß (E), and I{kappa}B{epsilon} (F) in Western blot experiments performed using cytoplasmic fractions obtained from EMSA experiments (Fig. 2Go). The results were normalized for actin and graphed at the indicated nonlinear time scale. Results are means ± SE for three independent experiments.

 
We next analyzed cytokine-induced I{kappa}B{alpha}, I{kappa}Bß, and I{kappa}B{gamma} degradation by Western blot assays. I{kappa}B{alpha} degradation was earlier (after 10 min) and more intense in IL-1ß-treated INS-1E cells, as compared with TNF-{alpha}-treated INS-1E cells or with IL-1ß-treated 208F cells (Fig. 4DGo; P < 0.05 vs. IL1ß-treated INS-1E cells), where degradation of I{kappa}B{alpha} was detectable only after 30 min of treatment (Fig. 4DGo). In 208F cells, cytokine-induced I{kappa}B{alpha} degradation was less intense even when compared with TNF-{alpha}-treated INS-1E cells (Fig. 4DGo). After 2 h of cytokine exposure, I{kappa}B{alpha} protein returned to basal levels in both INS-1E and 208F cells, but after 8 h there was a second and less intense wave of degradation (Fig. 4DGo). Cytokine-induced I{kappa}Bß and I{kappa}B{epsilon} degradation in INS-1E cells (Fig. 4Go, E and F, respectively) was slower and less marked than that observed for I{kappa}B{alpha}. IL-1ß again tended to have a more marked effect than TNF-{alpha}. On the other hand, IL-1ß failed to induce I{kappa}Bß and I{kappa}B{epsilon} degradation in 208F cells (Fig. 4Go, E and F). The absolute levels of the quantified I{kappa}B proteins (OD quantification of specific bands recognized by I{kappa}B antibodies) were similar between INS-1E and 208F control cells (data not shown). These results indicate that IL-1ß induces a more intense profile of degradation of I{kappa}B isoforms in INS-1E cells as compared with 208F cells, mirrored by a compensatory NF-{kappa}B-dependent later increase in I{kappa}B mRNA expression.

Cytokines Induce Differential Expression Profile of NF-{kappa}B Target Genes in INS-1E and 208F Cells
IL-1ß and TNF-{alpha} induced a progressive increase in MCP-1 mRNA expression in INS-1E cells, with a peak around 2–4 h and decrease after 12 h, but it did not return to control levels even after 48 h (Fig. 5AGo). Thus, at this time point, MCP-1 expression is still 40.2 ± 19- and 6.5 ± 2.2-fold higher than control for IL-1ß and TNF-{alpha} treatment, respectively. TNF-{alpha}-induced MCP-1 expression, however, was lower than the induction by IL-1ß; for instance, at 4 h TNF-{alpha}-induced MCP-1 expression was 10.8 ± 2.4-fold lower than the induction by IL-1ß (P ≤ 0.05, vs. IL-1ß) (Fig. 5AGo). Basal expression of MCP-1 mRNA was higher in 208F cells than in INS-1E cells (data not shown). IL-1ß induced a progressive increase in MCP-1 mRNA expression in 208F cells, with a decrease after 6–8 h of treatment, but this induction was much less marked than in INS-1E cells (Fig. 5AGo). Expression of Fas mRNA was induced in INS-1E cells by both IL-1ß and TNF-{alpha}, with a similar temporal profile and a peak around 2–4 h (Fig. 5BGo). However, as observed for MCP-1, IL-1ß induced a more marked increase in Fas mRNA expression than TNF-{alpha} (at 4 h, P ≤ 0.02 vs. IL-1ß) (Fig. 5BGo). Fas mRNA expression decreased after 12 h of exposure to both cytokines, but it did not return to control levels even after 48 h. IL-1ß induced a similar profile of Fas mRNA in 208F cells, with a peak around 4 h, but this induction was less marked than in INS-1E cells (Fig. 5BGo).


Figure 5
View larger version (23K):
[in this window]
[in a new window]
  		Fig. 5. Cytokines Induce Differential Expression Kinetics of NF-{kappa}B Target Genes in INS-1E and 208F Cells

INS-1E cells were exposed to IL-1ß (100 U/ml) (black diamonds) and TNF-{alpha} (1000 U/ml) (white square), and 208F cells were exposed to IL-1ß (100 U/ml) (open triangles). mRNA was extracted and real-time RT-PCR performed for MCP-1 (A), FAS (B), and iNOS (C). The results were corrected by GAPDH expression and shown as fold induction compared with control values (considered as 1). Results are means ± SE of three independent experiments.

 
Expression of inducible nitric oxide synthase (iNOS) mRNA was induced in INS-1E cells by both IL-1ß and TNF-{alpha}, with similar kinetics characterized by a peak after 4 h of exposure (Fig. 5CGo). IL-1ß, however, induced a 30.5 ± 3.6-fold higher induction of iNOS than TNF-{alpha} (P ≤ 0.02). iNOS expression decreased after 12 h of exposure to both cytokines (Fig. 5CGo). On the other hand, IL-1ß alone did not induce iNOS expression in 208F cells (data not shown); in these cells iNOS induction was only observed when using a combination of IL-1ß + IFN-{gamma} + TNF-{alpha} (data not shown). Moreover, INS-1E cells were sensitive to cell death induced by the nitric oxide (NO) donor 1,2,3,4-oxatriazolium, 5-amino-3(3,4-dichlorophenyl)-chloride (GEA), with only 36 ± 19% (P ≤ 0.05 vs. control; n = 4) and 1 ± 0.1% (P ≤ 0.001 vs. control; n = 4) cells remaining viable after a 24-h exposure to 25 µM or 50 µM GEA, respectively. On the other hand, 208F cells were relatively resistant to exogenous NO, with an observed viability of 94 ± 1% after exposure to 25 µM GEA (P > 0.5 vs. control; n = 4) and of 66.1 ± 11.0% after 50 µM GEA (P ≤ 0.05 vs. control; n = 4).

To further reveal differences in the intensity of cytokine-induced NF-{kappa}B activation in INS-1E and 208F cells, we transfected INS-1E and 208F cells with a NF-{kappa}B reporter plasmid containing six copies of a NF-{kappa}B consensus-binding site (pNF-{kappa}B-Luc). This NF-{kappa}B luciferase construct was activated by cytokines in both cell lines, but whereas in INS-1E cells IL-1ß induced a 10-fold increase in the luciferase activity as compared with control (Fig. 6Go, INS-IL), TNF-{alpha} induced only a 4-fold increase (P ≤ 0.001 vs. IL-1ß-treated INS-1E cells) (Fig. 6Go, INS-TNF). NF-{kappa}B activation was even lower in 208F cells treated with IL-1ß (Fig. 6Go, 208F-IL) or with a combination of IL-1ß + TNF-{alpha} + IFN-{gamma}, in the range of 1.2- to 1.3-fold induction (Fig. 6Go, 208F-3cyt; P ≤ 0.001 vs. IL-1ß-treated INS-1E cells). Similar results were obtained after 18 h treatment of the different cell lines with cytokines (data not shown).


Figure 6
View larger version (9K):
[in this window]
[in a new window]
  		Fig. 6. Differential Intensity of Cytokine-Induced NF-{kappa}B Reporter Gene Expression in INS-1E and 208F Cells

INS-1E and 208F cells were cotransfected with an NF-{kappa}B luciferase reporter gene and the internal control pRL-CMV, encoding Renilla luciferase. After overnight transfection, the cells were exposed to IL-1ß (100 U/ml) (IL) or TNF-{alpha} (1000 U/ml) (TNF), or to a combination of IL-1ß+TNF-{alpha}+IFN-{gamma} (3cytk) for 4 h and assayed for luciferase activities. The results are normalized for Renilla luciferase activity and are expressed as fold change against control (considered as 1). Results are means ± SE of three to six experiments. *, P ≤ 0.001 vs. INS-1E cells treated with IL-1ß, paired t test.

 
Role of the NF-{kappa}B-Binding Sites in the Induction of iNOS Promoter by Cytokines in INS-1E and 208F Cells
We have shown previously that a proximal NF-{kappa}B-binding site (–106 to –97), an octamer (Oct)-binding site (located 20 bp downstream from the proximal NF-{kappa}B site), and a distal NF-{kappa}B-binding site (–965 and –918) in the iNOS rat promoter (up to –1002 bp) are required for IL-1ß-induced iNOS promoter activity in insulin-producing cells (31, 32). In the present study we used constructs of the iNOS promoter containing mutations in these transcription factor-binding sites, to further investigate whether differences in cytokine-induced iNOS gene expression in INS-1E and 208F cells (Fig. 5Go) are related to usage of different cis-elements in the promoter region.

Disruption of the distal NF-{kappa}B-binding site in INS-1E cells, either by deletion of the 5'-region (construct 918luc) or by mutation of the distal NF-{kappa}B-binding site (dNFmut), reduced IL-1ß-induced iNOS promoter luciferase activity by 80% as compared with the wild-type promoter 1002luc (Fig. 7Go, INS-IL). Disruption of either the NF-{kappa}B proximal binding site (pNFmut) or the Oct-binding site [Octmut, which interacts with NF-{kappa}B to promote iNOS promoter activity (32)], reduced IL-1ß-induced iNOS promoter luciferase activity by 50% compared with the wild-type vector 1002luc (Fig. 7Go, INS-IL). On the other hand, these binding sites were of less relevance for TNF-{alpha}-induced iNOS promoter activity. Thus, only the 918luc construct conferred a significant decrease in luciferase activity as compared with 1002luc (Fig. 7Go, INS-TNF). For 208F cells only a disruption of the proximal NF-{kappa}B binding induced a decrease in the luciferase activity induced by IL-1ß treatment as compared with 1002luc (Fig. 7Go, 208F-IL), indicating differential cofactor usage in the distal and Oct-binding sites between ß- and non-ß-cells.


Figure 7
View larger version (22K):
[in this window]
[in a new window]
  		Fig. 7. Differential Usage of the iNOS Promoter cis-Acting Elements by Cytokines in INS-1E and 208F Cells

Cells were transfected with the wild-type iNOS promoter construct piNOS-1002luc (1002), or with the same construct mutated either in the distal (dNFmut) or proximal (pNFmut) NF-{kappa}B-binding sites, or in the Oct-binding site (Oct mut), or with a 5'-deletion piNOS-918luc (918) (31 32 ). Luciferase activity of the constructs after 16 h of cytokine exposure was calculated as fold change against control (no cytokines added) considered as 1. Fold changes of each construct after cytokine exposure are expressed as percentage of the luciferase activity of piNOS-1002luc (considered as 100%), for the same treatment. IL-1ß induced higher luciferase activity in piNOS1002luc vector (1002luc) in INS-1E cells (7.1 ± 1.4-fold induction; P ≤ 0.01 vs. control untreated cells), as compared with TNF-{alpha} treatment of INS-1E cells (1.7 ± 0.1-fold induction; P ≤ 0.01 vs. control untreated cells) or IL-1ß treatment of 208F cells (1.9 ± 0.9; P ≤ 0.001 vs. control untreated cells). Results are means ± SE of five to eight experiments. **, P ≤ 0.001; *, P ≤ 0.01; and #, P ≤ 0.02 vs. piNOS1002luc in the same cell type and treatment, paired t test.

 
Cytokines Induce a Differential Activation of ERK in INS-1E and 208F Cells
Activation of ERK1/2 contributes to cytokine-induced apoptosis and enhanced NF-{kappa}B transcriptional activity in ß-cells (29, 33). Therefore, we next analyzed cytokine-induced ERK1/2 activation in both INS-1E and 208F cells. In INS-1E cells IL-1ß induced an early (starting as early as 10 min) and transitory ERK1/2 phosphorylation, followed by a second peak of activation between 2 and 4 h (Fig. 8Go). In contrast, there was no induction of ERK1/2 phosphorylation in 208F cells treated with IL-1ß or in INS-1E cells after TNF-{alpha} treatment (Fig. 8Go). IL-1ß-induced ERK1/2 phosphorylation in INS-1E cells was significantly higher (P ≤ 0.05) as compared with TNF-{alpha}-treated INS-1E cells at 10, 30, and 90 min, and 2 and 4 h; it was also higher (P ≤ 0.05) than the phosphorylation observed in IL-1ß-treated 208F cells at 30, 60, and 90 min and 2 h. None of these treatments modified total ERK1/2 expression (data not shown).


Figure 8
View larger version (13K):
[in this window]
[in a new window]
  		Fig. 8. IL-1ß Induces ERK Activation in INS-1E Cells but Not in 208F Cells

INS-1E cells were exposed to IL-1ß (100 U/ml) (black diamonds) or TNF-{alpha} (1000 U/ml) (white square) whereas 208F cells (open triangles) were exposed to IL-1ß (100 U/ml) for the indicated time periods. Western blot experiments were performed using antibodies against P-ERK1/2 and total ERK1/2. The graphic represents OD values of P-ERK1/2 corrected by total ERK1/2 and normalized for the respective control values (considered as 1). Results are means ± SE for three to four independent experiments.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
Activation of the transcription factor NF-{kappa}B plays a crucial role in cytokine-induced ß-cell apoptosis in vitro (10, 11, 12) and in vivo (14), which raises the intriguing question of whether NF-{kappa}B activation possesses specific proapoptotic characteristics in ß-cells. To address this question, we compared the pattern of NF-{kappa}B activation induced by IL-1ß or TNF-{alpha} in insulin-producing INS-1E cells and by IL-1ß in rat fibroblast 208F cells. The rationale for selecting these treatments and cell types is that both IL-1ß and TNF-{alpha} activate NF-{kappa}B induction in ß-cells (Refs. 2 , 34 , and 35 and present data) and, especially in combination with IFN-{gamma}, trigger ß-cell apoptosis (Ref. 2 and present data); of note, IL-1ß has a 2-fold more intense proapoptotic effect than TNF-{alpha}. On the other hand, IL-1ß, alone or in combination with TNF-{alpha} + IFN-{gamma}, activates NF-{kappa}B but does not induce cell death in 208F fibroblast cells (present data). In contrast, NF-{kappa}B activation has mostly an antiapoptotic effect in fibroblasts (19, 20), a finding we confirmed by blocking NF-{kappa}B with an adenoviral construct expressing an I{kappa}B{alpha} superrepressor protein and thus sensitizing 208F cells to cytokine-induced cell death. The present findings indicate a distinct pattern of cytokine-induced NF-{kappa}B activation in INS-1E and 208F cells. These differences include: 1) an earlier and more potent NF-{kappa}B activation in IL-1ß-treated INS-1E cells; 2) an IL-1ß-induced oscillatory pattern of NF-{kappa}B activation in 208F cells but not in INS-1E cells; 3) a differential composition of NF-{kappa}B members in the activated complex; 4) differential kinetics of I{kappa}B expression and degradation; 5) a quantitatively and qualitatively differential expression of key NF-{kappa}B target genes; 6) differential usage of cis-elements in the promoter region of the iNOS target gene; and 7) ERK1/2 activation by IL-1ß in INS-1E cells, but not by TNF-{alpha} in INS-1E cells or by IL-1ß in 208F cells.

It has been previously described that the pattern of oscillation and periodicity plays an important role in the set of genes induced by NF-{kappa}B (23, 36). Against this background, we presently evaluated cytokine-induced expression of I{kappa}B{alpha}, MCP-1, iNOS, and Fas mRNAs, which we have previously shown to be modified by cytokines via NF-{kappa}B activation in ß-cells (30, 31, 37, 38). IL-1ß led to a more pronounced expression of NF-{kappa}B target genes in INS-1E cells, as compared with TNF-{alpha}, which correlates well with the more proapoptotic effect of IL-1ß in these cells. In 208F cells, IL-1ß-induced expression of NF-{kappa}B target genes was consistently lower, as compared with INS-1E cells, or even absent as in the case of iNOS expression. iNOS contributes to cytokine-induced ß-cell death via production of high amounts of the radical nitric oxide (NO) (39, 40). NO is involved in cytokine-induced endoplasmic reticulum stress, a key event for triggering ß-cell apoptosis (41), and microarray analysis of cytokine-treated insulin-producing cells has shown that nearly 50% of the genes modulated by cytokines are NO dependent (42).

Fas and MCP-1 proteins are both expressed in pancreatic ß-cells in response to IL-1ß and play a role in in vivo destruction of ß-cells during the insulitis process (2, 43, 44). NF-{kappa}B is a key regulatory factor in the induction of Fas (38) and MCP-1 (37) by cytokines in ß-cells and, in good agreement with the present findings, we have previously observed that both p65 and p50 NF-{kappa}B family members bind to Fas and MCP-1 promoters in ß-cells after IL-1ß exposure (37, 38).

The specificity of NF-{kappa}B dimer complexes for endogenous promoters depends not only on the {kappa}B site sequences itself, but also on protein-protein interactions in the context of the chromatin (24). In fibroblasts the MCP-1, I{kappa}B-{alpha}, and Fas genes are induced by the p50:p65, p52:p65, and p65:p65 dimers; MCP-1 can also be induced by the p50:c-Rel heterodimer (24). Taking this into account, the higher expression of these genes in INS-1E cells, compared with fibroblasts, could be due to a different composition of NF-{kappa}B dimers induced by cytokines. In line with this possibility, we observed that the composition of the activated NF-{kappa}B dimers is different in INS-1E and 208F cells. Thus, whereas in 208F cells the dominant NF-{kappa}B dimer induced by IL-1ß is constituted of p65:p50 [which is in good agreement with previous observations (24)], the main dimer observed in INS-1E cells during the first 12 h of IL-1ß exposure contained p65 in a homodimer composition, with a minor contribution by the p65/50 dimer after exposure to IL-1ß or TNF-{alpha}. Of note, dimer exchange during prolonged NF-{kappa}B activation, as presently observed, is another regulatory mechanism involved in duration and generation of differential transcriptional effects by NF-{kappa}B (25).

I{kappa}B{alpha} is known to participate in a feedback loop where newly synthesized I{kappa}B{alpha}, strongly induced by NF-{kappa}B, enters the nucleus and sequesters NF-{kappa}B dimers to the cytoplasm, inhibiting their activation (45). This generates an oscillatory pattern of NF-{kappa}B activation that is damped by the I{kappa}Bß and I{kappa}B{epsilon} isoforms, which have slower induction kinetics (23). In this study we observed that IL-1ß induces a stronger I{kappa}B{alpha} expression in INS-1E cells as compared with 208F cells, and a later expression of I{kappa}Bß and I{kappa}B{epsilon} as compared with I{kappa}B{alpha}. Interestingly, NF-{kappa}B activation is oscillatory in 208F cells but not in INS-1E cells, suggesting that additional elements, which remain to be identified, contribute to a stable or oscillatory pattern of NF-{kappa}B activation in different cell types. I{kappa}B{alpha} has affinity for p65- and c-Rel-containing dimers, specially for the p65:p50 heterodimers, whereas I{kappa}B{epsilon} has a selective role in the regulation of p65 homodimers and c-Rel/p65 heterodimers (46) and is probably involved in the regulation of later phases of NF-{kappa}B gene activation by p65:c-rel complexes (46). It is thus conceivable that the observed dimer exchange, in combination with the differential kinetics in expression/degradation of I{kappa}B isoforms, explains why NF-{kappa}B nuclear binding oscillates in 208F cells whereas it is stable in INS-1E cells.

IL-1ß induced a more pronounced and earlier I{kappa}B{alpha} degradation in INS-1E cells as compared with TNF-{alpha}-treated INS-1E cells or IL-1ß-treated 208F cells (Fig. 4Go). The earlier degradation of I{kappa}B{alpha} in INS-1E cells exposed to IL-1ß is in line with the earlier NF-{kappa}B activation observed in these cells (Fig. 2Go). TNF-{alpha} signaling via TNFR1 in other cell types results in the rapid activation of I{kappa}B kinase and nearly complete degradation of I{kappa}B{alpha} within 10 min (22). This is not the case in INS-1E cells, which have a slower and less marked I{kappa}B{alpha} degradation in the presence of TNF-{alpha}. IL-1ß and TNF-{alpha} induced a similar temporal profile of I{kappa}Bß and I{kappa}B{epsilon} protein degradation and resynthesis in INS-1E cells, with IL-1ß inducing a more pronounced response than TNF-{alpha} (Fig. 4Go, E and F). On the other hand, IL-1ß failed to induce I{kappa}Bß and I{kappa}B{epsilon} degradation in 208F cells.

Chromatin reorganization is an important element for the regulation of NF-{kappa}B signaling (36), dividing NF-{kappa}B target genes in two categories, namely early response genes that do not require chromatin changes for its expression, and late response genes that require chromatin changes. This explains why I{kappa}B{alpha}, previously shown to be an early-response gene (23, 36), is expressed before Fas in both INS-1E and 208F cells. We have previously shown, in the context of chromatin, that the long-lasting stimulatory effect of IL-1ß on MCP-1 expression in ß-cells is related to an earlier and more sustained binding of NF-{kappa}B to the MCP-1 promoter (37), as compared with macrophages exposed to lipopolysaccharide (36). A similar mechanism may contribute to the observed pattern of prolonged expression of other NF-{kappa}B target genes in INS-1E cells, as compared with fibroblasts. On the other hand, the present observation that IL-1ß induces a higher activation of an NF-{kappa}B reporter and an iNOS reporter construct in INS-1E cells, as compared with INS-1E cells stimulated with TNF-{alpha} or 208F cells exposed to IL-1ß, suggests that neither chromatin changes nor a differential composition in NF-{kappa}B members provides the whole explanation for the differences observed between the diverse stimuli/cell types. Indeed, transiently transfected promoters are both accessible to transcription factors independently of chromatin configuration and promiscuous regarding binding to NF-{kappa}B family members (24). This suggests that posttranslational modifications such as phosphorylation, acetylation, methylation, or sumoylation may contribute to enhance/modify NF-{kappa}B responses in IL-1ß-treated INS-1E cells. Posttranslational changes modulate both strength and duration of NF-{kappa}B responses in other cell types (47); for example, the I{kappa}B kinase complex can directly regulate phosphorylation of the p65 subunit, which is associated with a cell type- and stimulus-dependent enhancement of transcriptional activity of NF-{kappa}B (48). Moreover, recent findings indicate that the MAPK ERK, which is differentially regulated by IL-1ß and TNF-{alpha} (Ref. 49 and present data), increases the transactivating capacity of NF-{kappa}B in insulin-producing cells and thereby NF-{kappa}B-mediated iNOS activation (29). In line with these data, we observed that IL-1ß, but not TNF-{alpha}, induced ERK phosphorylation in INS-1E cells but failed to do so in 208F cells. This may contribute to the observed differences between TNF-{alpha}- and IL-1ß-induced NF-{kappa}B activation and subsequent biological effects in INS-1E cells.

Accumulating evidence suggests that ß-cell fate after exposure to cytokines depends on the time course and severity of perturbation of key ß-cell gene networks (2, 50, 51). These networks are regulated, at least in part, by the transcription factors NF-{kappa}B (30) and STAT-1 (52). As discussed above, there is considerable complexity in what was previously thought as a single NF-{kappa}B pathway. The present observations, i.e. IL-1ß-induced NF-{kappa}B activation in insulin-producing cells is more intense and sustained than in fibroblasts, involves different members of the NF-{kappa}B family, and leads to a more marked activation of downstream genes potentially involved in insulitis and ß-cell death, provide the first indication of why NF-{kappa}B is proapoptotic in ß-cells. Additional clarification of the regulation of NF-{kappa}B and other stress-related transcription factors in pancreatic ß-cells will be crucial to allow us to understand and, hopefully, to prevent, ß-cell death in T1D.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Cell Culture and Time Course Experiments
Rat insulin-producing INS-1E cells, a kind gift from Professor C. Wolheim (Center Medical Universitaire, Genenva, Switzerland), were cultured in RPMI medium containing 5% fetal calf serum (53). INS-1E cells were used between passages 66–80; these cells have a well-preserved function, as judged by glucose-induced insulin release and glucose oxidation (data not shown). Rat fibroblast 208F cells [European Collection Cell Cultures (ECACC), Salisbury, UK] were cultured in DMEM supplemented with 10% fetal calf serum. For time-course experiments cells were exposed for 10, 30, 45, 60, and 90 min, and 2, 4, 6, 8, 12, 24, and 48 h to recombinant human IL-1ß (100 U/ml; a kind gift from Dr. C. W. Reinolds, National Cancer Institute, Bethesda, MD) or to recombinant murine TNF-{alpha} (1000 U/ml; Innogenetics, Gent, Belgium). The rationale for the selection of the time points was that NF-{kappa}B is activated by IL-1ß after 10–30 min (Ref. 54 and present data), whereas an increase in cell death is detected only after a 24-h exposure to the cytokine (present data). Cytokine concentrations were selected based on our previous dose-response experiments (Refs. 2 , 33 , and 42 and present data), using cell survival as endpoint.

Assessment of Cell Viability
Cells were cultured in 96-well dishes (6000 cells per condition) and exposed to IL-1ß (10, 50, or 100 U/ml) or TNF-{alpha} (100, 500, and 1000 U/ml) alone, or in combination with 0.036 µg/ml recombinant rat IFN-{gamma} (R&D systems, Oxon, UK) (corresponding to ~100 U/ml of IFN-{gamma}) for 24 h and 48 h. In some experiments an NO donor, GEA 3162 (Alexis, San Diego, CA) was also used. The percentage of viable, necrotic and apoptotic cells was determined in at least 600 cells in each experimental condition, by inverted fluorescence microscopy after addition of the DNA dyes Hoechst 342 (20 µg/ml) and propidium iodide (10 µg/ml) (42, 55, 56, 57). Viability was evaluated by at least two observers, one of whom was unaware of sample identity. The agreement between findings obtained by the two observers was at least 90%. The apoptosis index was calculated as [(% apoptotic cells in experimental conditions – % apoptotic cells in control)/(100-dead cells in control)] x 100 (55, 57). In the present series of experiments, the percentage of apoptosis in control INS-1E cells was 6 ± 1% (n = 10).

Transfection with Recombinant Adenovirus
The recombinant replicative-deficient adenovirus, encoding a mutated nondegradable I{kappa}B{alpha} protein with serines 32 and 36 mutated to alanines (AdI{kappa}B(S/A)2) (58) or a control virus encoding the enhanced green fluorescent protein were prepared and used as previously described (10). The 208F cells were infected overnight at 37 C, at a multiplicity of infection of 10.

EMSAs
Nuclear extracts were prepared from INS-1E and 208F cells (59) and DNA binding by nuclear proteins (4 µg) and supershift analysis with anti-p50, anti-p65, anti-c-Rel, and anti-p52 antibodies (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) were performed as previously described (37). The sequence of NF-{kappa}B consensus oligonucleotide was 5'-AGCTTCAGAGGGGACTTTCCGAGA-3'. The specificity of the protein-DNA complex was assessed by the use of a non radioactive NF-{kappa}B consensus oligonucleotide, whereas the specificity of the supershift complex was tested using a nonrelated antibody ({alpha}-STAT or {alpha}-IRF-1). The specific NF-{kappa}B mobility shift bands detected in the x-ray film were quantified by Biomax 1D image analysis software (Eastman Kodak, Rochester, NY) and expressed as arbitrary units of OD. The results of each individual experiment were normalized by the peak value, considered as 100. The results are expressed as mean ± SE for three to four independent experiments.

Real-Time Quantitative RT-PCR Analysis
After exposure to cytokines during time course experiments, cells were harvested, and reverse transcriptase reaction was performed using poly (A)+RNA (41). Expression of I{kappa}B isoforms ({alpha}, ß, and {epsilon}), and target genes for NF-{kappa}B (MCP-1, iNOS, and Fas) was determined by the use of real-time RT-PCR using SYBR Green fluorescence on a LightCycler instrument (Roche Diagnostics, Mannheim, Germany) by the standard curve method (60, 61). The PCR amplification reactions and preparation of standards were performed as previously described (60). Expression values of the gene of interest were corrected by the expression values of the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and shown as fold-induction compared against control values (no cytokine added), considered always as 1. We have shown previously that cytokines do not modify GAPDH expression in insulin-producing cells (41, 42, 62), and these observations were confirmed in the present series of experiments (data not shown). Primer sequences and their respective PCR fragment lengths were as follows (first are shown the primers used for the standard curve, and then the primers used for the real-time PCR): I{kappa}B{alpha}1: forward (F), 5'-AAGGACGAGGATTACGAGCA-3'; reverse (R), 5'-CAAAGTCACCAAGTGCTCCA-3' (504 bp); I{kappa}B{alpha}2: F, 5'-TTGGTCAGGTGAAGGGAGAC-3'; R, 5'-GTCTCGGAGCTCAGGATCAC-3' (143 bp). I{kappa}Bß1: F, 5'-TGGCTACGTCACTGAGGATG-3'; R, 5'-AGGCTCCGGTTTATTGAGGT-3' (562 bp); I{kappa}Bß2: F, 5'-ACTCAGAGCCAGGACCACAC-3'; R, 5'-GGGTGTGGCCATAGTTT-3' (142 bp). I{kappa}B{epsilon}1: F, 5'-TTCTGTTGCTTGGCTTTCCT-3'; R, 5'-CTCATCTTCCACGTTCAGCA-3' (588 bp); I{kappa}B{epsilon}2: F, 5'-CCTGGACCTCCAACTGAAGA-3'; R, 5'-ATGTCGGCTCCATTCTGAAG-3' (108 bp). Fas1: F, 5'-ATGTTCGAATGCAAGGGACT-3'; R, 5'-GCAAGGCTCAAGGATGTCTT-3' (410 bp); Fas2: F, 5'-GGAGGAGTACACGG-ACAGGA-3'; R, 5'-TTTCTTTGCACCTGCACTTG-3' (131 bp). Primers used for GAPDH, MCP-1, and iNOS were described elsewhere (63).

Western Blot Analysis
Cytoplasmic protein extracts were obtained from the cells used in EMSA experiments, as described above. An equal amount of protein (60 µg) was subjected to a 10–12% SDS-PAGE and transferred to a nitrocellulose membrane. Immunoblot analysis was performed with anti-I{kappa}B{alpha}, anti-I{kappa}Bß, anti-I{kappa}B{epsilon}, and antiactin antibodies (Santa Cruz Biotechnology), anti-P-ERK1/2, anti-ERK1/2 (Cell Signaling Technology, Beverly, MA) and then with a secondary antirabbit horseradish peroxidase-labeled anti-IgG (Santa Cruz Biotechnology or Cell Signaling Technology). The specific bands recognized by the antibodies were quantified by Biomax 1D image analysis software (Eastman Kodak) and expressed as OD. The intensity values for the proteins analyzed were corrected by the values of the housekeeping protein actin. Results of ERK experiments were normalized by the peak value considered as 1, and expressed as mean ± SE for three to four independent experiments.

Promoter Reporter Assays
INS-1E and 208F cells were cotransfected by lipofection using lipofectAMINE 2000 (Invitrogen, Paisley, Scotland, UK) with the pRL-CMV encoding Renilla luciferase (Promega Corp. Madison, WI) as an internal control and with either the pNF-{kappa}B-Luciferase (BD Biosciences, Palo Alto, CA) test plasmid, or the control vector containing only the TATA-like promoter (pTAL) (38). For the studies with the iNOS promoter, the following constructs were used: wild-type iNOS promoter region (piNOS-1002luc); 5'-deleted sequence (piNOS-918luc); mutants for, respectively, the proximal NF-{kappa}B binding site (piNOS-pNFmut), the distal NF-{kappa}B binding site (piNOS-dNFmut), and the Oct-binding site (piNOS-Octmut) (31, 32). After transfection, the cells were exposed to IL-1ß (100 U/ml), TNF-{alpha} (1000 U/ml), or IFN-{gamma} (0.036 µg/ml), alone or in combination, for 4 h, 16 h, or 18 h. Luciferase activities were assayed with the Dual-Luciferase Reporter Assay System (Promega Corp.) (37). The test values were corrected for the luciferase values of the internal control plasmid pRL-CMV and expressed as fold-induction of luciferase activity of cytokine-treated cells compared with control (no cytokine added), considered as 1 (37).

Statistical Analysis
Data are shown as means ± SEM, and comparisons between groups were carried out either by t test (paired or unpaired) or by ANOVA followed by Student’s t test with the Bonferroni correction, as indicated. A P value of ≤ 0.05 was considered as statistically significant.


   ACKNOWLEDGMENTS
 
We thank M. A. Neef, G. Vandenbroeck, M. Urbain, J. Schoonheydt, N. El Amrite, and R. Leeman from Laboratory of Experimental Medicine, Université Libre de Bruxelles (Brussels, Belgium), and A. Hillesöe, from the Steno Diabetes Center (Gentofte, Denmark), for excellent technical support; and Dr. Nathan Goodman, Institute for Systems Biology (Seattle, WA) and Dr. Hamid Bolouri, BioComputation Group at the Science and Technology Research Institute (University of Hertfordshire, UK), for helpful discussions.


   FOOTNOTES
 
This work was supported by the Communauté Française de Belgique—Actions de Recherche Concertées, Fonds National de la Recherche Scientifique (Belgium), European Foundation for the Study of Diabetes, and Novo Nordisk Programme in Diabetes Research, the European Union 6th frame Program—Project EuroDia, The Danish Diabetes Association, and The Sehested-Hansen Foundation. A.K.C. is the recipient of a postdoctoral fellowship from the Juvenile Diabetes Research Foundation International and D.C. received a postdoctoral fellowship from the Coordenaçao de Aperfeiçoamento de Pessoal de Nivel Superior (Brazil).

First Published Online March 23, 2006

Abbreviations: 208F-IL, IL-1ß-treated 208F cells; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; IFN, interferon; iNOS, inducible nitric oxide synthase; INS-IL; IL-1ß-treated INS-1E cells; INS-TNF, INS-1E cells treated with TNF-{alpha}; IRF-1, interferon regulatory factor-1; NF-{kappa}B, nuclear factor-{kappa}B; Oct, octamer; STAT, signal transducer and activator of transcription; T1D, type 1 diabetes mellitus.

Received for publication July 5, 2005. Accepted for publication March 17, 2006.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Gepts W 1965 Pathologic anatomy of the pancreas in juvenile diabetes mellitus. Diabetes 14:619–633[Medline]
  2. Eizirik DL, Mandrup-Poulsen T 2001 A choice of death—the signal-transduction of immune-mediated ß-cell apoptosis. Diabetologia 44:2115–2133[CrossRef][Medline]
  3. Kurrer MO, Pakala SV, Hanson HL, Katz JD 1997 ß-Cell apoptosis in T cell-mediated autoimmune diabetes. Proc Natl Acad Sci USA 94:213–218[Abstract/Free Full Text]
  4. O’Brien BA, Harmon BV, Cameron DP, Allan DJ 1997 Apoptosis is the mode of ß-cell death responsible for the development of IDDM in the nonobese diabetic (NOD) mouse. Diabetes 46:750–757[Abstract]
  5. Moriwaki M, Itoh N, Miyagawa J, Yamamoto K, Imagawa A, Yamagata K, Iwahashi H, Nakajima H, Namba M, Nagata S, Hanafusa T, Matsuzawa Y 1999 Fas and Fas ligand expression in inflamed islets in pancreas sections of patients with recent-onset type I diabetes mellitus. Diabetologia 42:1332–13340[CrossRef][Medline]
  6. Davalli AM, Scaglia L, Zangen DH, Hollister J, Bonner-Weir S, Weir GC 1996 Vulnerability of islets in the immediate posttransplantation period. Dynamic changes in structure and function. Diabetes 45:1161–1167[Abstract]
  7. Biarnes M, Montolio M, Nacher V, Raurell M, Soler J, Montanya E 2002 ß-Cell death and mass in syngeneically transplanted islets exposed to short- and long-term hyperglycemia. Diabetes 51:66–72[Abstract/Free Full Text]
  8. May MJ, Ghosh S 1997 Rel/NF-{kappa}B and I{kappa}B proteins: an overview. Semin Cancer Biol 8:63–73[CrossRef][Medline]
  9. Mercurio F, Manning AM 1999 NF-{kappa}B as a primary regulator of the stress response. Oncogene 18:6163–6171