The role of interferon regulatory factor-1 in cytokine-induced mRNA expression and cell death in murine pancreatic beta-cells.

ARTICLE

INTRODUCTION

Several mechanisms have been implicated in the pathological processes leading to ß-cell dysfunction and death in type 1 diabetes, including ß-cell interaction with cytotoxic T cells through MHC class I molecules [1], the induction of iNOS expression and NO production [2], the production of oxygen free radicals [3], and the activation of an apoptotic program in ß-cells [4]. Most of these mechanisms require the presence of cytokines, such as IL-1ß and IFN-gamma, secreted by infiltrating immunocytes [5]. These cytokines can up-regulate MHC class I [6] and induce iNOS expression in ß-cells [2] and some of the ß-cell neighbors [7-9], and they can also lead to ß-cell apoptosis [4]. On the other hand, cytokines can also induce or upregulate expression of proteins which protect ß-cell from injury, such as heat shock protein 70, manganese superoxide dismutase, heme oxygenase (reviewed in [10, 11]) and serine protease inhibitor-3 (SPI-3) [12]. SPI-3 is an acute phase protein which may play a role in cellular protection against injury [13]. The observation that chemical serine protease inhibitors prevent IL-1ß-induced ß-cell dysfunction [14, 15], and the recent report on IL-1ß-induced SPI-3 gene expression in rat pancreatic ß-cells [12], suggest that this protease inhibitor may be part of the ß-cell defense mechanisms triggered in response to cell damage.

Most of the effects of cytokines are tightly regulated at the transcriptional level. By binding to its cell surface receptor, IFN-gamma activates the tyrosine kinases Jak1 and Jak2 [16]. These kinases phosphorylate the signal transducer and activator of transcription-1alpha (STAT1alpha), which in turn binds to gamma-activated sites of various genes [16]. One of the genes activated by STAT1alpha is interferon regulatory factor-1 (IRF-1) [17]. IRF-1 is a transcription factor which regulates expression of MHC class I in different tissues [18] and iNOS expression in macrophages [19]. IRF-1 is also required for induction of apoptosis in T cells [20] and triggering autoimmune diseases such as collagen-induced arthritis and experimental allergic encephalomyelitis [21]. IRF-1 expression is upregulated by IFN-gamma and IL-1ß in human and rat islets and in the rat insulinoma cell line RINm5F [22], and IRF-1^­/­ mice backcrossed with diabetes-prone NOD mice do not develop hyperglycemia [23]. These observations suggest that IRF-1-mediated gene expression, either in ß-cells or immune-effector cells, may play an important role for ß-cell destruction during insulitis.

In the present study, we used pancreatic islets and FACS-purified ß-cells from IRF-1^­/­ mice to evaluate the role for IRF-1 in cytokine-induced MHC class I, SPI-3 and iNOS mRNA expression, and ß-cell death.

MATERIALS AND METHODS

Animals

The IRF-1^­/­ mice were a generous gift from Dr. Tak Mak of the Ontario Cancer Institute (Ontario, Canada). The IRF-1^­/­ mice were generated by using the construct pMIRF1neoB containing 4.9 kb homologues IRF-1 genomic DNA with a 1.2 kb deletion which was replaced by the neomycin resistance gene [24]. The deletion removed the exons required for the DNA-binding domain, resulting in an inactive IRF-1 product. The IRF-1^­/­ mice were then backcrossed 3-times into a C57Bl/6 background. The animals were bred and maintained, under filter hoods and fed with sterilized mouse diet and water, at the experimental animal facility of the Catholic University of Leuven. Wild type C57Bl/6 (wt) mice (purchased from Harlan Nederland, Horst, The Netherlands, and maintained under similar conditions as the IRF-1^­/­ mice) were used as controls for the IRF-1^­/­ mice. The homozygosity of IRF-1^­/­ mice was confirmed in pancreatic (see Figure 2) and spleen tissue (data not shown) by the absence of IRF-1 mRNA expression.

Islet and ß-cell isolation and culture

Islets were obtained from 10 week old male and female IRF-1^­/­ and wt mice. Pancreases were digested with collagenase and filtered over 500 µm pore mesh nylon screens. From the filtered material, islets were hand-picked using a siliconized glass pipette, placed into the culture medium and cultured for 16 hours at 37° C. The islets were then dissociated with trypsin (20 µg/ml) and sorted by FACStar flow cytometer (Becton Dickinson, Sunnyvale, CA), as previously described for rat ß-cells [25]. Evaluation by electron microscopy showed that these preparations contained more than 95% viable ß-cell (data not shown).

Cell culture was performed in HAM's F-10 medium supplemented with 10 mM glucose, 50 µM IBMX (3-isobutyl-1-methylxanthin), 1% BSA (Boehringer Mannheim, Mannheim, Germany), 0.1 mg/ml streptomycin (Continental Pharma, Puteaux, Belgium), 0.075 mg/ml penicillin (Laboratoires Diamant, Brussels, Belgium) and 2 mM L-glutamine (GIBCO, Paisley, Scotland) [26]. Islets were cultured for 24 hours in suspension culture dishes of 3 cm diameter (Nunc, Naperville, IL USA) at a concentration of 50-100 islets in 1 ml of medium. Purified single ß-cells (3 x 10³ cells per cup) were cultured in Falcon 96-well microtiter plates (Becton Dickinson, New Jersey, NJ) containing 200 µL of medium. In a group of experiments with purified ß-cells, performed to determine medium nitrite accumulation, we used 3-6 x 10⁴ cells per cup in 200 µl of the above described medium. For long term culture (6 and 9 days), culture medium was changed every 3 days and fresh cytokines added. Since the culture conditions described above were originally developed for maintenance of rat pancreatic ß-cells [26], we initially tested whether they were adequate for mouse pancreatic ß-cells (obtained from C57Bl/6/DBA mice). Using FACS-purified ß-cells, we observed 82 ± 2% viable cells after 10 days of culture (n = 3), which was similar to previous data with rat ß-cells [27].

Cytokine treatment and nitrite determination

The effect of cytokines was examined after 2 hours, 24 hours, 3, 6 and 9 days of culture in the presence of recombinant murine IFN-gamma (1,000 U/ml, 10 U/ng, Holland Biotechnology, Leiden, The Netherlands) and/or recombinant human IL-1ß (50 U/ml, 38 U/ng, a kind gift of Dr. C.W. Reynolds from the National Cancer Institute, Bethesda, USA). In some experiments to evaluate ß-cell viability, we also added recombinant murine tumor necrosis factor-alpha (TNF-alpha; 1,000 U/ml, 220 U/ng, Innogenetics, Gent, Belgium). These concentrations of cytokines were selected based on previous studies with pancreatic islets [22, 28, 29]. In the experiments aimed to determine mRNA expression, islets or ß-cells were exposed to 50 U/ml IL-1ß or 50 U/ml IL-1ß + 1,000 U/ml IFN-gamma, but not to IFN-gamma alone. This was because of the limited amount of mouse islets and ß-cells available, and that IFN-gamma alone does not affect mouse ß-cell function or survival [28] (Pavlovic and Eizirik, unpublished data). Culture media were collected after 24 hours for nitrite determination (nitrite is a stable product of NO oxidation), which was performed spectrophotometrically at 546 nm wavelength after colored reaction with the Griess reagent [30].

mRNA isolation and RT-PCR

Poly(A)⁺ RNA was isolated from islets or cell aggregates (0.5 to 1 x 10⁵ cells) using oligo(dT)₂₅-coated polystyrene Dynabeads (Dynal, Oslo, Norway). The reverse transcription reaction was performed at 42° C for 1 hour, and contained (per 10 µl) mRNA equivalent to 6 x 10³ cells, 1x reverse transcription buffer, 5 mM MgCl₂, 1 mM of each dNTP, 2.5 µM random hexamer primers and 100 units of Moloney murine leukemia virus reverse transcriptase (Perkin Elmer, Norwalk, CT). The subsequent PCR reaction contained (in 25 µl reaction solution): 5 µl cDNA, 0.4 µM of forward and reverse primers, 200 µM of each dNTP, 1x PCR buffer, 2 mM MgCl₂, and 0.625 U AmpliTaq Gold DNA polymerase (Perkin Elmer). PCR specificity and efficiency was improved by using hot start PCR with a 12 min predenaturation at 95° C and then 27 (glyceraldehyde-3-phosphate dehydrogenase ­ GAPDH), 28 (MHC-I) and 31 (iNOS, IRF-1, SPI-3) cycles at 94° C for 45 s, 58° C for 45s and 72° C for 80 s. For the insulin gene, cDNA was diluted 40-fold and PCR performed for 22 cycles using the thermo-profile described above. The number of cycles was selected to allow linear amplification of the cDNA under study. To enable comparisons between the relative amounts of target cDNA among the different samples, external standards for each target cDNA and for the house keeping gene GAPDH cDNA were used in the PCR reactions. In order to prepare these standards, PCR products of specific bands amplified from cDNA were detected on the ethidium bromide-stained agarose gel and their concentrations estimated by comparison against known amounts of molecular weight markers run in the same gel. Dilution series containing decreasing amounts of target cDNA templates were then prepared and amplified simultaneously with islet or ß-cell cDNA samples. The PCR amplification efficiencies for cDNA samples and their respective standards were then compared by densitometry (see below). The primer sequences used were IRF-1F 5'-TCTGAGTGGCATATGCAGATGGAC-3'; IRF-1R 5'-GGTCAGAGACCCAAACTATGGTGC-3'; MHC-IF 5'-TCGCTGAGGTATTTCGTCAC-3'; MHC-IR 5'-TCTTCGTTCAGGGCGATGTA-3'; iNOS-F 5'-GACAGCACAGAATGTTCCAG-3'; iNOS-R 5'-TGGCCAGATGTTCCTCTATT-3'; GAPDH-F 5'-TCACTCAAGATTGTCAGCAA-3', GAPDH-R 5'-AGATCCACGACGGACACATT-3'; insulin-F 5'-CCCACCCAGGCTTTTGTCAA-3'; insulin-R 5'-GTAGAGGGAGCAGATGCTGG-3'; SPI3-F: 5'-CAAATTTGTCCCAATGTCTGC-3'; SPI3-R: 5'-AAGGTTGCACAGTCCCATCAA-3'. The ethidium bromide-stained agarose gels were photographed under UV-transillumination using Kodak Digital Science DC40 camera (Kodak, Rochester, NY, USA) and the PCR band intensities on the image were quantified by Biomax 1D Image analysis software (Kodak) and expressed in pixel intensities (O.D.). The target cDNA present in each sample was corrected for the respective GAPDH value.

Assessment of ß-cell viability

The viability of islet cell preparations was assessed after 5 days exposure to cytokines. The islets were incubated for 15 min with propidium iodide (PI, 10 µg/ml) and Hoechst (HO) 342 (20 µg/ml) [27]. PI is a highly polar dye, which penetrates only cells with damaged membranes, staining their nuclei red; HO 342 freely crosses the plasma membrane, entering cells with both damaged and intact membranes and leading to DNA being stained blue [27, 29]. The approximate percentage of dead cells was estimated by two individual observers who were unaware of the sample identity. Note that evaluation of cell death in whole islets is difficult due to superposition of cells. Thus, these observations should be considered as semi-quantitative. The percentages of apoptotic and necrotic cells in the single ß-cell preparations were assessed after 3, 6 and 9 days exposure to cytokines [29]. The cells were examined under an inverted fluorescence microscope with UV excitation at 340-380 nm. Viable cells were identified by their intact nuclei with blue fluorescence (HO 342), necrotic cells by their intact nuclei with yellow-red fluorescence (HO 342 + PI), apoptotic cells by their fragmented nuclei, exhibiting either a blue (HO 342; early apoptosis) or yellow-red fluorescence (HO 343 + PI; late apoptosis). This fluorescence assay for single ß-cells has been validated by electron microscopy, and offers the advantage of being quantitative [27]. Under each experimental condition, a minimum of 500 cells were counted. The necrosis and apoptosis indices were calculated as ((% necrotic or apoptotic cells under experimental condition ­ % necrotic or apoptotic cells in control)/(100 ­ % dead cells in control)) x 100 [31].

Statistical analysis

Results are presented as means ± SEM of data obtained from islets and ß-cells isolated from male and female animals. No significant differences were observed between male and female mice regarding the different parameters studied (data not shown), allowing pooling of the data. The statistical differences between the groups were determined by paired Student's t-test or, when indicated, by analysis of variance (ANOVA) followed by the multiple t-tests with the Bonferroni correction. In all experiments, islets or pure ß-cell preparations obtained from one set of animals (3 to 5 mice) were considered as one individual observation. Thus, each independent experiment was performed with a separate set of animals and on a separate day.

RESULTS

Cytokine-induced IRF-1 mRNA expression in islets isolated from wt mice

To determine whether IFN-gamma and/or IL-1ß induce IRF-1 expression in mouse pancreatic islets, as previously described for rat and human islets [22], islets obtained from wt mice were exposed to these cytokines for 2 hours (Figure 1). There was a low basal expression of IRF-1 mRNA in wt islets, which was strongly upregulated by both IL-1ß (5-fold) and IFN-gamma (20-fold), or by a combination of these two cytokines (50-fold) (Figure 1). This increased IRF-1 expression was maintained even after 24 hours exposure to cytokines (Figure 2). On the other hand, islets obtained from IRF-1^­/­ mice did not express IRF-1 mRNA under either basal condition or after exposure to IL-1ß or IL-1ß + IFN-gamma (Figure 2; data not shown), confirming the IRF-1-deficient state of these mice.

Cytokine-induced MHC class I, iNOS, SPI-3 and insulin mRNA expression and medium nitrite accumulation in islets isolated from wt and IRF-1^­/­ mice

MHC class I mRNA was expressed under basal condition both in wt and IRF-1^­/­ islets (Figure 2). These levels were higher in wt than in the IRF-1^­/­ islets (Table 1). Upon addition of the cytokines IL-1ß or IL-1ß + IFN-gamma, MHC class I mRNA levels increased in both groups, reaching higher levels in islets isolated from wt mice (Table 1). However, expressing the data as fold-increase above the basal level, indicated similar values for wt (IL-1ß, 1.4 ± 0.4, n = 7; IL-1ß + IFN-gamma, 2.0 ± 0.02, n = 7) and IRF-1^­/­ islets (IL-1ß, 1.4 ± 0.1, n = 6; IL-1ß + IFN-gamma, 2.5 ± 0.3, n = 7).

SPI-3 mRNA expression was not detectable in the wt islets under basal condition (Figure 2, Table 1), but was induced by IL-1ß. This induction was significantly decreased after exposure to IL-1ß + IFN-gamma, suggesting an inhibitory role for IFN-gamma. Islets from IRF-1^­/­ mice also expressed SPI-3 mRNA under basal condition. This expression was upregulated by IL-1ß, but, contrary to the observations recorded for wt islets this effect was strongly potentiated by the addition of IFN-gamma (Figure 2, Table 1).

There was no iNOS expression under basal condition in wt or IRF-1^­/­ islets, but in both strains iNOS mRNA was induced by IL-1ß, and further increased by IFN-gamma. Following cytokine exposure, iNOS mRNA expression was lower in IRF-1^­/­ than in wt islets (Table 1).

Insulin mRNA was similar in both wt and IRF-1^­/­ islets under basal conditions, with optical densities (O.D. corrected by GAPDH O.D.; islet treatment as described in Table 1) of respectively 1.4 ± 0.1 (n = 5) and 1.4 ± 0.1 (n = 4). After exposure to IL-1ß, this expression was significantly reduced in wt islets (O.D. 0.9 ± 0.1; p < 0.01 versus basal condition, i.e. no cytokine added, n = 5), but it was unaffected in IRF-1^­/­ islets (O.D. 1.4 ± 0.1; n = 4). Simultaneous exposure to IL-1ß + IFN-gamma led to a further decrease in insulin mRNA expression in wt islets (O.D. 0.6 ± 0.1; p < 0.002 versus basal condition, n = 5), and had also an inhibitory effect, albeit less severe than in wt islets, on insulin mRNA expression in the IRF-1^­/­ islets (O.D. 0.9 ± 0.1; p < 0.002 versus basal condition, n = 4).

Nitrite production by wt islets doubled upon addition of IL-1ß and was increased 5-fold upon stimulation with the cytokine combination IL-1ß + IFN-gamma (Table 1). There was no increase in the nitrite accumulation in IRF-1^­/­ islets after exposure to IL-1ß alone, but these islets responded with a 3-fold increase in nitrite production following stimulation with IL-1ß + IFN-gamma (Table 1).

Viability of islet cells isolated from wt and IRF-1^­/­ mice after 5 days exposure to IL-1ß + IFN-gamma

Under basal conditions both wt and IRF-1^­/­ islets were well preserved after 5 days culture (Figure 3), showing mostly viable cells (< 10% dead cells). The combination of IL-1ß + IFN-gamma induced islet cell death (Figure 3), but the degree of cell destruction was clearly lower in the IRF-1^­/­ islets, where death occurred in approximately 31 ± 4% of the cells versus 85 ± 3% in the wt islets (n = 4, p < 0.0001).

Cytokine-induced MHC class I, iNOS and SPI-3 mRNA expression and medium nitrite accumulation in FACS-purified pancreatic ß-cells isolated from wt and IRF-1^­/­ mice

MHC class I mRNA was expressed under basal condition both in wt and IRF-1^­/­ ß-cells (Table 2). Upon addition of the cytokines IL-1ß or IL-1ß + IFN-gamma, MHC class I mRNA levels increased in both groups (Table 2). When expressing the data as fold-increase above the basal level, there were similar values for wt (IL-1ß, 1.4 ± 0.1; IL-1ß + IFN-gamma, 3.6 ± 0.3; n = 3) and IRF-1^­/­ ß-cells (IL-1ß, 2.0 ± 0.3; IL-1ß + IFN-gamma, 4.2 ± 0.6; n = 3).

In ß-cells from wt mice SPI-3 mRNA was not detectable under basal condition (Table 2), but it was induced by IL-1ß. This induction was somewhat decreased after exposure to IL-1ß + IFN-gamma. ß-cells obtained from IRF-1^­/­ mice expressed SPI-3 mRNA under basal condition and this expression was upregulated by IL-1ß, an effect further potentiated by IFN-gamma (Table 2).

There was no iNOS mRNA expression under basal conditions in wt or IRF-1^­/­ ß-cells, but in both strains there was a minor iNOS expression following IL-1ß exposure, which was further increased by IFN-gamma (Table 2). Neither ß-cells isolated from wt nor those isolated from IRF-1^­/­ mice increased medium nitrite accumulation in response to IL-1ß treatment. Exposure to IL-1ß + IFN-gamma induced a nearly 2-fold increase in medium nitrite in both wt and IRF-1^­/­ ß-cells. Contrary to the observation in whole islets (Table 1), ß-cells isolated from IRF-1^­/­ mice produced similar amounts of nitrite as wt ß-cells.

Cell death in FACS-purified ß-cells isolated from wt and IRF-1^­/­ mice following 3, 6 and 9 days exposure to combinations of cytokines.

Exposure of wt or IRF-1^­/­ ß-cells to IL-1ß + IFN-gamma had only a minor impact on the necrosis and apoptosis indices in purified ß-cells (data not shown). For this reason, cells were exposed to the cytokine combination IL-1ß + IFN-gamma + TNF-alpha, which had been previously shown to induce apoptosis in human ß-cells [29]. Combinations of cytokines induced a significant and similar increase in necrosis and apoptosis indices in both wt and IRF-1^­/­ ß-cells after 6 days of culture. Following 9 days of cytokine exposures most of the wt or IRF-1^­/­ ß-cells were dead (Figure 4).

DISCUSSION

We have previously suggested that IRF-1 activation is required for cytokine-induced iNOS expression in the rat insulinoma cell line RINmF5 and in rat and human pancreatic islets [22]. This suggestion was mostly based on indirect evidence, and at this point it was not possible to block IRF-1 expression and/or its nuclear binding. An alternative approach to study this question is the use of mice with targeted disruption of the IRF-1 gene [24]. Since our previous studies were performed in human or in rat islets (or rat insulin-producing RINmF5 cells) [22], we began by testing whether wild type mouse islets also express IRF-1 in response to cytokines. We observed that IFN-gamma, alone or in combination with IL-1ß, induces an early (2 hours) and marked increase in IRF-1 expression in mouse islets, thus validating the use of this experimental model to answer our question.

Following exposure to IL-1ß or IL-1ß + IFN-gamma, islets isolated from IRF-1^­/­ mice presented a 30-50% reduction in iNOS mRNA expression and nitrite production, as compared to wt islets. This partial decrease is different from results obtained in IRF-1^­/­ macrophages, where a nearly complete absence of iNOS activity in response to cytokines and LPS was observed [19]. Interestingly, FACS-purified ß-cells from both wt and IRF-1^­/­ islets failed to produce NO in response to IL-1ß alone. In agreement with these data, we observed that the mouse insulinoma cell line MIN6 also fails to produce NO in response to IL-1ß (Pavlovic and Eizirik, unpublished data). This differs from results obtained in rat ß-cells and insulinoma cell lines, where IL-1ß alone induces NO production [2, 32-34], but is similar to our previous observations with human islets [35] or human ductal cells [9], where IL-1ß + IFN-gamma are required for iNOS expression. Exposure of wt or IRF-1^­/­ ß-cells to a combination of IL-1ß + IFN-gamma resulted in a clear and similar induction of iNOS expression and NO production, indicating that the potentiating effect of IFN-gamma on IL-1ß-induced iNOS expression in ß-cell does not require IRF-1. We have recently described that STAT1 has a more important role than IRF-1 in iNOS expression in insulin producing cells [33]. Thus, it is conceivable that IFN-gamma-induced STAT1 activation in the IRF-1^­/­ may compensate for the lack of IRF-1. The fact that whole mouse islets produced nitrite in response to IL-1ß alone, and that IRF-1^­/­ islets have a lower nitrite synthesis than wt islets, suggests that part of the nitrite production is derived from islet non-ß-cells. Indeed, freshly isolated islets contain around 30-40% non-ß-cells, of both endocrine and non-endocrine origin. Among these non-ß-cells, macrophages, endothelial and ductal cells are potential sites of NO production [7-9]. Macrophages can contribute to islet NO production by both direct NO synthesis [7] and by releasing cytokines which in turn further up-regulate iNOS expression in ß-cells [32, 36]. As mentioned above, IRF-1 is required for iNOS expression and NO production by macrophages [19]. Thus, it is conceivable that a functional inhibition of islet macrophages in IRF-1^­/­ mice contributed, at least in part, for the observed decrease in total islet NO production by these mice.

Another mRNA induced by cytokines in mouse islets and ß-cells was SPI-3. SPI-3 is a natural protease inhibitor, induced by cytokines and other forms of cell stress in hepatocytes [37], brain cells [13] and rat pancreatic ß-cells [12]. SPI-3 may have a role in cellular protection against injury [13, 37], and it has been previously described that chemical protease inhibitors prevent IL-1ß-induced ß-cell dysfunction [14, 15]. Interestingly, while IFN-gamma induced a clear decrease in IL-1ß-induced SPI-3 expression in wt islets or ß-cells, the same cytokine potentiated by 5-fold the stimulatory effect of IL-1ß in IRF-1^­/­ islets or ß-cells. This indicates that IRF-1 mediates the inhibitory effect of IFN-gamma on SPI-3 expression and that, in the absence of this transcription factor, a stimulatory component of the cytokine action (possibly mediated via STAT1 activation) predominates and further increases the cellular content of SPI-3 mRNA. Blockers of NO production induce a partial protection against the deleterious effects of cytokines in rat [2, 38] and mouse islets [28]. Thus, it is conceivable that a partial decrease in NO production (see above), paralleled by a several-fold increase in SPI-3 expression, contributed to the observed protection against cytokine-induced cell death in IRF-1^­/­ pancreatic islets, as compared to the wt islet cells. This improved viability was accompanied by a better preserved insulin mRNA expression, reinforcing the idea that ß-cells located in the IRF-1^­/­ islets are partially protected from IL-1ß + IFN-gamma effects. On the other hand, FACS purified ß-cells from IRF-1^­/­ mice were as sensitive as ß-cells from wt mice to cytokine-induced cell death. The modes of ß-cell death were both necrosis and apoptosis, which is somewhat different from our previous observations in human islet cells, where combination of cytokines induced mostly apoptosis [29]. Considering that IRF-1^­/­ ß-cells exposed to cytokines also overexpress SPI-3, it seems that this serine protease inhibitor, in the absence of a decreased NO (or other radicals) production, is not sufficient to prevent cytokine-induced cell death. As discussed above, other islet cell types may contribute to cytokine-induced ß-cell death in whole islets. In the absence of these non-ß-cells, IRF-1 is not a major mediator of the direct, toxic effects of combinations of cytokines in ß-cells.

Another potential mechanism by which cytokines contribute to ß-cell destruction in type 1 diabetes mellitus is via upregulation of ß-cell MHC class I expression [6]. The requirement of IRF-1 for upregulation of MHC class I expression varies in different tissues [18], although there is no information regarding the ß-cells. We have observed that islets and ß-cells isolated from IRF-1^­/­ mice present a partial (30%) decrease in basal MHC class I expression, but are able to respond to cytokines with an increased expression of this mRNA. This suggests that, as discussed above for iNOS, other pathways of cytokine signal transduction besides IRF-1 are important for this phenomenon in pancreatic ß-cells.

As a whole, the present and previous data obtained in other tissues from IRF-1^­/­ mice, indicate that IFN-gamma contributes to the expression of iNOS and MHC class I mRNA by different signal transduction pathways in diverse cell types [19, 20, 23, 39]. In this respect, IRF-1 is probably more important for cytokine signaling in the immune system [21] than in pancreatic ß-cells (present data). If this is the case, the recent observation that NOD mice deficient in IRF-1 do not develop diabetes [23], is probably explained by the inhibition of immune effector cells, rather than by the blocking of IFN-gamma effects at the ß-cell level.

CONCLUSION

Acknowledgments.

The technical assistance of E. Verheugen, R. Leemans and J. Laureys is gratefully acknowledged. This work was supported by grants from the Juvenile Diabetes Foundation International (JDFI 1-1998-4); Flemish Scientific Research Foundation (FWO, G.0216.99 and a postdoctoral fellowship (C.A. Gysemans)); from the Belgium National Ministry of Science (IUAP P4/21); and by a Shared Cost Action in Medical and Health Research of the European Community (BMH4-CT98-3448).

REFERENCES

1. Foulis A K. 1987. The pathogenesis of beta cell destruction in type I (insulin-dependent) diabetes mellitus. (Review) J. Pathol. 152: 141.

2. Eizirik D L, Pavlovic D. 1997. Is there a role for nitric oxide in ß-cell dysfunction and damage in IDDM? Diabetes Metab. Rev. 13: 293.

3. Rabinovitch A. 1993. Roles of cytokines in IDDM pathogenesis and islet ß-cell destruction. Diabetes Rev. 1: 215.

4. Mauricio D, Mandrup-Poulsen T. 1998. Apoptosis and the pathogenesis of IDDM: a question of life and death. (Review) Diabetes 47: 1537.

5. Mandrup-Poulsen T. 1996. The role of interleukin-1 in the pathogenesis of IDDM. (Review) Diabetologia 39: 1005.

6. Pavlovic D, van de Winkel M, van der Auwera B, Chen M C, Schuit F, Bouwens L, Pipeleers D. 1997. Effect of IFN-gamma and glucose on major histocompatibility complex class I and class II expression by pancreatic ß- and non-ß-cells. J. Clin. Endocrinol. Metab. 82: 2329.

7. Wiegand F, Kroncke K D, Kolb-Bachofen V. 1993. Macrophage-generated nitric oxide as cytotoxic factor in destruction of alginate-encapsulated islets. Protection by arginine analogs and/or coencapsulated erythrocytes. Transplantation 56: 1206.

8. Steiner L, Kroncke K D, Fehsel K, Kolb-Bachofen V. 1997. Endothelial cells as cytotoxic effector cells: cytokine-activated rat islet endothelial cells lyse syngeneic islet cells via nitric oxide. Diabetologia 40: 150.

9. Pavlovic D, Chen M C, Bouwens L, Eizirik D L, Pipeleers D. 1999. Contribution of ductal cells to cytokine responses by human pancreatic islets. Diabetes 48: 29.

10. Eizirik D L, Sandler S, Palmer J P. 1993. Repair of pancreatic beta-cells. A relevant phenomenon in early IDDM? (Review) Diabetes 42: 1383.

11. Eizirik D L. 1996. Beta-cells defence and repair mechanisms in human pancreatic islets. Horm. Metab. Res. 28: 302.

12. Chen M C, Schuit F, Pipeleers D G, Eizirik D L. 1999. IL-1ß induces serin protease inhibitor 3 (SPI-3) gene expression in rat pancreatic ß cells detection by differential display of mRNA. Cytokine (in press).

13. Tsuda M, Kitagawa K, Imaizumi K, Wanaka A, Tohyama M, Takagi T. 1996. Induction of SPI-3 mRNA, encoding a serine protease inhibitor, in gerbil hippocampus after transient forebrain ischemia. Brain Res. Mol. Brain Res. 35: 314.

14. Welsh N, Bendtzen K, Sandler S. 1990, Influence of proteases on inhibitory and stimulatory effects of interleukin-1ß on ß-cell function. Diabetes 40: 290.

15. Eizirik D L, Bendtzen K, Sandler S. 1991. Short exposure of rat pancreatic islets to interleukin-1ß induces a sustained but reversible impairment in ß-cell function: influence of protease activation, gene transcription, and protein synthesis. Endocrinology 128: 1611.

16. Schindler C, Darnell J E Jr. 1995. Transcriptional responses to polypeptide ligands: the JAK-STAT pathway. (Review) Annu. Rev. Biochem. 64: 621.

17. Li X, Leung S, Qureshi S, Darnell J E Jr, Stark G R. 1996. Formation of STAT1-STAT2 heterodimers and their role in the activation of IRF-1 gene transcription by interferon-gamma. J. Biol. Chem. 271: 5790.

18. Hobart M, Ramassar V, Goes N, Urmson J, Halloran P F. 1996. The induction of class I and class II major histocompatibility complex by allogeneic stimulation is dependent on the transcription factor interferon regulatory factor-1 (IRF-1): observations in IRF-1 knockout mice. Transplantation 62: 1895.

19. Kamijo R, Harada H, Matsuyama T, Bosland M, Gerecitano J, Shapiro D, Le J, Koh S I, Kimura T, Green S J, et al. 1994. Requirements for transcription factor IRF-1 in NO synthase induction in macrophages. Science 263: 1612.

20. Tamura T, Ishihara M, Lamphier M S, Tanaka N, Oishi I, Aizawa S, Mitsuyama T, Mak T W, Tahi S, Taniguchi T. 1995. An IRF-1-dependent pathway of DNA damage-induced apoptosis in mitogen-activated T lymphocytes. Nature 376: 596.

21. Tada Y, Ho A, Matsuyama T, Mak T W. 1997. Reduced incidence and severity of antigen-induced autoimmune disease in mice lacking interferon regulatory factor-1. J. Exp. Med. 185: 231.

22. Flodström M, Eizirik D L. 1997. interferon-gamma-induced interferon regulatory factor-1 (IRF-1) expression in rodent and human islet cells precedes nitric oxide production. Endocrinology 138: 2747.

23. Nakazawa T, Satoh J, Takakashi K, Sakata Y, Masuda T, Taniguchi T, Kato I, Okamoto H, Toyota T. 1998. IRF-1-deficient nonobese diabetic mice do not develop autoimmune diabetes. Diabetes 45 (Suppl. 3): 0806 (Abstract).

24. Matsuyama T, Kimura T, Kitagawa M, Pfeffer K, Kawakami T, Watanabe N, Kündig T M, Amakawa R, Kishikara K, Wakeham A, et al. 1993. Targeted disruption of IRF-1 or IRF-2 results in abnormal type I IFN gene induction and aberrant lymphocyte development. Cell 75: 83.

25. Pipeleers D G, In't Veld P A, van de Winkel M, Maes E, Schuit F, Gepts W. 1985 A new in vitro model for the study of pancreatic A and B cells. Endocrinology 117: 806.

26. Ling Z, Hannaert J C, Pipeleers D. 1994. Effects of nutrients, hormones and serum on survival of rat islet beta cells in culture. Diabetologia 37: 15.

27. Hoorens A, van de Casteele M, Klöppel G, Pipeleers D. 1996. Glucose promotes survival of rat pancreatic ß cells by activating synthesis of proteins which suppress a constitutive apoptotic program. J. Clin. Invest. 98: 1568.

28. Cetkovic-Cvrlje M, Eizirik D L. 1994. TNF-alpha and IFN-gamma potentiate the deleterious effects of IL-1ß on mouse pancreatic islets mainly via generation of nitric oxide. Cytokine 6: 399.

29. Delaney C A, Pavlovic D, Hoorens A, Pipeleers D G, Eizirik D L. 1997. Cytokines induce deoxyribonucleic acid strand breaks and apoptosis in human pancreatic islet cells. Endocrinology 138: 2610.

30. Green L, Wagner D, Glogowski J, Skipper P L, Wishnok J S, Tannenbaum S R. 1982. Analysis of nitrate, nitrite and (¹⁵N)nitrate in biological fluids. Anal. Biochem. 126: 131.

31. Pipeleers D, van de Winkel M. 1986. Pancreatic ß cells possess defense mechanisms against cell-specific toxicity. Proc. Natl. Acad. Sci. USA 83: 5267.

32. Heitmeier M R, Scarim A L, Corbett J A. 1997. Interferon-gamma increases the sensitivity of islets of Langerhans for inducible nitric-oxide synthase expression induced by interleukin-1. J. Biol. Chem. 272: 13697.

33. Darville M I, Eizirik D L. 1998. Regulation by cytokines of the inducible nitric oxide synthase promoter in insulin-producing cells. Diabetologia 41: 1101.

34. Ling Z, Chen M C, Smismans A, Pavlovic D, Schuit F, Eizirik D L, Pipeleers D G. 1998. Intercellular differences in interleukin-1ß-induced suppression of insulin synthesis and stimulation of non-insulin protein synthesis by rat pancreatic ß-cells. Endocrinology 139: 1540.

35. Eizirik D L, Sandler S, Welsh N, Cetkovic-Cvrlje M, Nieman A, Geller D A, Pipeleers D G, Bendtzen K, Hellerström C. 1994. Cytokines suppress human islet function irrespective of their effects on nitric oxide generation. J. Clin. Invest. 93: 1968.

36. Corbett J A, McDaniel M L. 1995. Intraislet release of interleukin-1 inhibits beta-cell function by inducing beta-cell expression of inducible nitric oxide synthase. J. Exp. Med. 181: 559.

37. Pages G, Rouayrenc J F, Le Cam G, Mariller M, Le Cam A. 1990. Molecular characterization of three rat liver serine-protease inhibitors affected by inflammation and hypophysectomy. Protein and mRNA analysis and cDNA cloning. Eur. J. Biochem. 190: 385.

38. Southern C, Schulster D, Green I C. 1990. Inhibition of insulin secretion by interleukin-1ß and tumour necrosis factor-alpha via L-arginine-dependent nitric oxide generating mechanism. FEBS Lett. 276: 42.

39. Khan I A, Matsuura T, Fonseka S, Kasper L H. 1996. Production of nitric oxide (NO) is not essential for protection against acute Toxoplasma gondii infection in IRF-1^­/­ mice. J. Immunol. 156: 636.