Protein Kinase Cδ Activation by Interleukin-1β Stabilizes Inducible Nitric-oxide Synthase mRNA in Pancreatic β-Cells*

Exposure of pancreatic islets to cytokines such as interleukin (IL)-1β induces a variety of proinflammatory genes including type II nitric-oxide synthase (iNOS) which produces nitric oxide (NO). NO is thought to be a major cause of islet β-cell dysfunction and apoptotic β-cell death, which results in type I diabetes. Since protein kinase C (PKC) mediates some of the actions of cytokines in other cell types, our aim was to assess the role of PKC in IL-1β-induced iNOS expression in pancreatic β-cells. PKCδ, but not PKCα, was specifically activated in the rat INS-1 β-cell line by IL-1β as assessed by membrane translocation. Moreover, iNOS expression and NO production were significantly attenuated by the PKCδ specific inhibitor rottlerin and overexpression of a PKCδ kinase-dead mutant protein. Conversely, overexpression of PKCδ wild type protein significantly potentiated this response. These results were confirmed at the mRNA level by reverse transcriptase-polymerase chain reaction. However, a role at the level of transcriptional regulation appeared unlikely, since PKCδ was not required for the activation of NF-κB, activating protein 1, and activating transcription factor 2 signaling pathways in response to IL-1β. There was, however, a significant increase in iNOS mRNA stability mediated by PKCδ wild type, while PKCδ kinase-dead acted reciprocally, reducing iNOS mRNA stability. The results indicate that, in addition to transcriptional activation, mRNA stabilization is a key component of the mechanism by which IL-1β stimulates iNOS expression in β-cells and that PKCδ plays an essential role in this process. PKCδ activation may therefore have significant consequences with regard to cellular function and viability when β-cells are exposed to IL-1β and potentially other cytokines.

Type I diabetes is an inflammatory disease that is characterized by apoptotic destruction of pancreatic ␤-cells. It is thought to be mediated at least in part by cytokines such as interleukin (IL) 1 -1␤, tumor necrosis factor, and interferon-␥ released from infiltrating macrophages. Among other effects, these cytokines induce the expression of iNOS and the production of NO within ␤-cells, which is proposed to be a significant trigger for apoptosis (1)(2)(3)(4).
IL-1␤ induces the expression of iNOS protein in ␤-cells by activating the stress-activated protein kinase (SAPK)/c-Jun N-terminal kinase (JNK) and p38 mitogen-activated protein kinase as well as the NF-B pathway, which in turn activate ATF2, AP-1, and NF-B transcription factors, respectively (5)(6)(7)(8)(9). These translocate to the nucleus and bind to specific regions within the promoter to initiate iNOS message transcription (10). Gene regulation is also controlled post-transcriptionally at the level of message stability (11), and this is particularly true of iNOS mRNA, which is targeted for rapid degradation because of signals in the 3Ј-untranslated region (12,13).
PKC is a family of serine/threonine protein kinases consisting of 11 isoforms, comprising conventional, novel, and atypical subgroups, each with varying cofactor requirements and cellular distribution (14). Studies using pharmacological PKC activators or PKC overexpression have established that PKC activation is necessary for enhanced gene expression in response to proinflammatory stimuli (15)(16)(17) and in a limited number of cells is sufficient for iNOS induction (18,19). PKC is capable of activating AP-1 (20), ATF2 (21), and NF-B (22)(23)(24)(25)(26) and may also act upstream of the mitogen-activated protein kinase cascade to modulate activity of extracellular signal-regulated kinase 1/2, p38, and JNK/SAPK (27,28), so it could mediate its effects on cytokine signaling by acting at any of these sites.
PKC␦ and PKC␣ are the predominant PKC isozymes expressed in islets and insulinoma cells (29), but their role in IL-1␤-induced gene expression has not been defined. Determination of that role was the aim of the current study. Using adenoviral constructs for overexpression of wild-type and kinase-dead PKC isoforms in the INS-1 cell line, we demonstrate that PKC␦ is necessary for IL-1␤-stimulated iNOS protein expression and NO production and that the underlying mechanism is control of iNOS mRNA stability.
Generation and Expression of PKC␣ and PKC␦ Recombinant Adenovirus-cDNA for mouse PKC␣ and rat PKC␦ (generous gifts from Fredrick Mushinski and Peter Parker, respectively) were first subcloned into pALTER (Promega, Annandale, New South Wales, Australia) using conventional molecular biological techniques (30). Mutagenesis was performed according to the protocol using the mutagenic primers 5Ј-TACGCCATCAGGATCCTGAAG-3Ј and 5Ј-CTTTGCAAT-CAGGTACCTGAAGAAG-3Ј (for PKC␣ and PKC␦, respectively) (Beckman Instruments, Gladesville, Australia) targeted to conserved lysine residues in the ATP binding domain. Mutants were designated PKC␣KD (K368R substitution) and PKC␦KD, (K376R substitution). Overexpression of kinase-dead forms have previously been shown to act in an isoenzyme-specific dominant negative fashion (31,32), while our data (not shown) and others have confirmed that the PKC␦KD mutant is catalytically inactive (20,33). Wild type and mutant PKC cDNAs were then subcloned from pALTER into pXCMV, an adenoviral shuttle vector constructed in this laboratory by subcloning the pRcCMV (Invitrogen) expression cassette into pXCX3, derived from pXCX2 (34). Recombinant adenovirus was then prepared by recombination with the adenovirus plasmid pJM17, essentially as described by Graham and Prevec (35). The MX17 control is an adenovirus that contains just the virus backbone (pJM17) and the expression cassette from pXCX3, but without cDNA. Adenovirus-mediated expression of PKC␣ and PKC␦ proteins was determined by infecting INS-1 cells at 10 -20 plaqueforming units/cell, in 200 l of medium and then incubating for 1 h at 37°C, mixing every 15 min. Virus was then removed, and fresh medium was applied. At 48 h postinfection, proteins were separated by 10% SDS-PAGE, and PKC␣ or PKC␦ expression was determined by immunoblotting as described below.
Analysis of IL-1␤-induced PKC␦ Translocation to the Plasma Membrane-At 48 h of culture, INS-1 cells were exposed to 300 pg/ml IL-1␤ (R & D Systems, Minneapolis, MN) for the times indicated. Cells were then washed in ice-cold PBS and cytosol, and membrane fractions were prepared, essentially as described (36). Proteins in individual fractions were separated by SDS-PAGE on 10% gels (all reagents from Bio-Rad) and immunoblotted as described below.
Immunoblot Analysis-Whole cell lysate fractions were prepared from INS-1 cells by washing monolayers in ice-cold PBS and resuspending in modified Laemmli sample buffer (37) containing 1% SDS and 10% mercaptoethanol. Membrane and cytosolic fractions prepared above were also resuspended in this buffer. Proteins were separated by SDS-PAGE on 10% gels, transferred to nitrocellulose membranes (Transblot; Bio-Rad) by electroblotting, blocked with 5% milk powder, and probed with appropriate antibodies, i.e. mouse anti-PKC␣ antibody from Transduction Laboratories (San Diego, CA), rat anti-PKC␦ and iNOS antibodies from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and cRel, IB␣, IB␤, phospho-p38, JNK, and extracellular signalregulated kinase 1/2 antibodies from New England Biolabs (Beverly, MA). Donkey anti-rabbit horseradish peroxidase-conjugated secondary antibody was from Jackson ImmunoResearch Laboratories Inc. (West Grove, PA), and goat anti-mouse horseradish peroxidase-conjugated secondary antibody was from Caltag Laboratories (Burlingame, CA). The antigen-antibody complexes were visualized by ECL detection (Amersham Pharmacia Biotech) and exposure to x-ray film (Fuji, Tokyo, Japan). Signal intensities were determined by densitometric analysis (on a Molecular Dynamics Personal Densitometer SI and software by IPLabGel, Signal Analytics, VA).
Measurement of IL-1␤-induced iNOS Expression and NO Production-INS-1 and islets cells were either infected with recombinant adenovirus as described above or exposed to PKC inhibitors (Calbiochem) 30 min prior to the application of 300 pg/ml IL-1␤. After 24-h exposure to IL-1␤, 100 l of medium was analyzed for nitrite via the Griess diazo reaction (38). Analysis of iNOS expression in INS-1 cells was performed by washing the remaining cells in PBS and immunoblotting cell lysates as described above, using a rat anti-iNOS antibody.
Isolation of Nuclear Extracts for Immunoblot Analysis-INS-1 cells were treated appropriately, harvested by washing with ice-cold PBS, and resuspended in 1 ml. Cells were pelleted by centrifugation at 13,000 ϫ g for 20 s. The PBS was removed, and cells were resuspended in 175 l of ice-cold buffer A (10 mM Hepes, pH 7.9, 10 mM KCl, 1.5 mM MgCl 2 , 0.1 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 100 g/ml aprotonin, 100 g/ml leupeptin). Cells were incubated for 10 min on ice before the addition of 9 l of 10% (v/v) Nonidet P-40 (Sigma) and vortexed for 20 s. The nuclear fraction was pelleted at 13,000 ϫ g for 20 s and washed once in buffer A, without resuspending the pellet. 40 l of buffer C (10 mM Hepes, pH 7.9, 420 mM KCl, 1.5 mM MgCl 2 , 0.1 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 100 g/ml aprotonin, 100 g/ml leupeptin) was then applied, and the pellet was resuspended by vortexing and vigorous agitation for 30 min at 4°C. This nuclear fraction was then pelleted for 15 min at 13,000 ϫ g, and supernatant was removed for analysis by SDS-PAGE and immunoblotting.
Luciferase Reporter Assays-INS-1 cells were transfected with pCIneo (Promega) and B, API (kind gifts from Malcom Handel, Garvan Institute) and ATF2 (Promega) luciferase reporter DNA using Superfectamine (Life Technologies, Inc.) according to the manufacturer's instructions. The B luciferase plasmid has six NF-B sites and a ␥-fibrinogen basal promoter, while the AP-1 plasmid has three AP-1 sites also driven by the ␥-fibrinogen basal promoter. Stably transfected populations were selected using neomycin (400 g/ml) over 6 weeks. Stable cell lines were then seeded in 96-well plates at 5 ϫ 10 4 cells/well and exposed to adenovirus after a 24-h culture. 48 h postinfection, cells were exposed to 300 pg/ml IL-1␤ for 4 h, and luciferase activity was measured using the LucLite luciferase assay kit (Amersham Pharmacia Biotech) according to the manufacturer's instructions. Luminescence was measured using a Canberra Packard (Sydney, Australia) TopCount microplate scintillation counter.
RNA Isolation and Semiquantitative RT-PCR of iNOS Message-INS-1 cells were cultured in six-well plates (at 1 ϫ 10 6 cells/well) and infected with adenovirus as described above. At 48 h postinfection, cells were exposed to 300 pg/ml IL-1␤ for either 6 or 12 h. To measure message stabilization, cells were exposed to actinomycin D (1 M) for the indicated times. Cells were then washed in PBS and resuspended in 1 ml of Trizol/well (Life Technologies, Inc.). RNA was extracted according to the manufacturer's instructions and quantified using absorbance at 260 nm. cDNA synthesis was performed using a preamplification kit (Life Technologies, Inc.) according to the manufacturer's instructions, using 5 g of RNA in each reaction. PCR was then performed using 2 l of cDNA, oligonucleotides specific for either ␤-actin or iNOS (39), and AmpliTaq Gold (PerkinElmer Life Sciences). The PCR amplification protocol was designed to allow quantification of iNOS and ␤-actin PCR product while in the exponential phase of amplification. Products were amplified using a 5-min hot start at 95°C and 18 (␤-actin) or 24 -28 cycles (iNOS) of 30 s at 95°C, 30 s at 55°C, and 60 s at 74°C. Products were analyzed by gel electrophoresis on 1.5% agarose gels, stained with ethidium bromide, visualized on a GelDoc 1000 illuminator (Bio-Rad), and analyzed using IPLabGel software.
Statistics-All results are presented as mean Ϯ S.E. Statistical significance was determined using unpaired Student's t test.

Role of PKC␦ in IL-1␤ Signaling in INS-1 Cells-Pancreatic
islets and ␤-cell lines have been well documented to contain PKC, of which the predominant isoenzymes expressed are PKC␣ and PKC␦ (29,40). To determine whether these isozymes are involved in IL-1␤ signaling in the clonal insulin secreting INS-1 cell line, we first measured PKC translocation to the plasma membrane as an indicator of activation. In response to IL-1␤, PKC␦ translocates from the cytosolic to the membrane fraction of INS-1 cells as early as 2 min after stimulation with IL-1␤ (Fig. 1A). No such translocation with PKC␣ was observed. Densitometric analysis (Fig. 1B) revealed that accumulation of PKC␦ in the membrane fraction was more than doubled by 2 min stimulation and significantly elevated for at least 7.5 min.
Role of PKC␦ in IL-1␤-induced iNOS Expression and NO Production-To determine whether PKC activation participated in iNOS induction by IL-1␤ in ␤-cells, we measured iNOS expression and NO production in the presence of specific PKC inhibitors (41). IL-1␤ stimulated the induction of NO production and iNOS expression at 24 h in INS-1 cells (Fig. 2, A and  B). When incubated with 30 nM Go6976, a PKC inhibitor selective for PKC␣ (41,42), iNOS expression and NO production were not inhibited (but instead were enhanced). However, rottlerin (10 M), which selectively inhibits PKC␦, almost completely inhibited production of NO. Most importantly, rottlerin also completely inhibited IL-1␤-stimulated NO accumulation in isolated rat pancreatic islets (Fig. 2C), confirming that this effect is not limited to the INS-1 cells.
These results therefore implicate PKC␦ in IL-1␤ signaling and prompted us to investigate its role more directly. Recombinant adenovirus capable of overexpressing WT and KD mutant forms of PKC␦ were generated and used to infect INS-1 cells. Recombinant PKC adenovirus induced a high degree of overexpression by 72 h relative to MX17 control virus (Fig. 3A). A direct role for PKC␦ in IL-1␤-induced NO production and iNOS expression was confirmed using the adenoviral constructs as shown in Fig. 3, B and C, respectively. IL-1␤ stimulated a robust induction of iNOS and a 6-fold increase in NO production in INS-1 cells infected with control virus. Overexpression of PKC␦WT, however, significantly potentiated the NO response to IL-1␤ without affecting basal levels. Similar effects were seen on iNOS expression. Effects of PKC␦KD overexpression were reciprocal to those of PKC␦WT, with IL-1␤stimulated iNOS expression markedly reduced and NO production levels almost completely abolished.
PKC␦ Modulates IL-1␤-induced iNOS mRNA Levels-The results above suggest that PKC␦ activation is required for IL-1␤-stimulated iNOS protein expression. To assess the underlying mechanism, we first measured iNOS mRNA levels after a 12-h exposure to IL-1␤. Semiquantitative RT-PCR analysis (Fig. 4A) showed that IL-1␤ robustly induces iNOS message from virtually undetectable levels. Overexpression of PKC␦WT significantly potentiated message abundance, over 2-fold compared with the MX17 control virus. Conversely, overexpression of PKC␦KD reduced iNOS message levels to 40% of control. The results confirm that PKC␦ plays a major role in controlling IL-1␤Ϫinduced iNOS expression by effects mediated at the level of mRNA.
Signal Transduction Pathways Controlling iNOS Gene Transcription-The transcription factor AP-1 plays a role in regu-lating iNOS gene expression in pancreatic ␤-cells (2,5). Because PKCs are important regulators of AP-1 activity in many cells, acting upstream of JNK/SAPK, we first examined the potential for PKC␦ to act in this pathway. Fig. 5A shows the results of Western blot analysis to measure JNK/SAPK activation using phosphospecific antibodies that recognize only the activated form of these kinases. IL-1␤ clearly activated this kinase but in a manner that was unaffected by the presence of PKC␦WT or PKC␦KD. In Fig. 5B, we show activation of AP-1mediated luciferase expression in INS-1 stably expressing the AP-1 reporter plasmid and exposed to IL-1␤ for 4 h. Clearly, overexpression of either PKC␦WT or PKC␦KD had no effect on AP-1 activity stimulated by IL-1␤.
We next analyzed the ATF2 transcription factor and the role

FIG. 2. Inhibition of PKC␦ by rottlerin inhibits IL-1␤ induced NO production and iNOS expression in INS-1 cells and rat islets.
A, INS-1 cells were treated with Go6976 (30 nM) or Rottlerin (10 M) for 30 min prior to the addition of IL-1␤ (300 pg/ml). After 24 h, medium was taken to determine levels of NO production by the Griess reaction. Data are the mean of four separate experiments (**, p Ͻ 0.01 when compared with control INS-1 cells with IL-1␤). B, whole cell lysates were prepared in parallel, and proteins were separated by SDS-PAGE on 10% gels and transferred to nitrocellulose membranes, which were blotted with anti-iNOS antibody. The immunoblot shown is representative of four separate experiments. C, whole rat islets were also exposed to IL-1␤ for 24 h with and without rottlerin, and NO production was measured as above.
of PKC upon this pathway, since ATF2 has been shown to be regulated by PKC in other systems (21). We again used phosphospecific antibodies to measure the effect of PKC␦WT and PKC␦KD protein on IL-1␤-induced activation of p38 (Fig. 6A) and a specific luciferase reporter to measure ATF2 activity in INS-1 cells exposed to IL-1␤. While the IL-1␤-induced phosphorylation of p38 was unaltered by PKC␦ overexpression, there were effects at the level of ATF2-mediated luciferase expression. In fact, PKC␦WT clearly and significantly augmented both the basal and IL-1␤ response. This suggests that PKC␦ might be sufficient for up-regulating ATF2 reporter activity. However, this effect is independent of IL-1␤, since the response to PKC␦WT and IL-1␤ appeared additive, and the PKC␦KD mutant did not affect stimulated reporter activity.
We next investigated the effect of PKC␦ on various aspects of the NF-B signaling cascade. NF-B exists as a cytosolic heterodimer, where activation is mediated by degradation of an inhibitory (IB) subunit, which thereby facilitates translocation of the active subunit (p65) to the nucleus. The time course of degradation and resynthesis of IB␣ and IB␤ in response to IL-1␤ are shown in Fig. 7A. We also show the appearance of p65 in the nucleus over this time. Both IB␣ and IB␤ are degraded in response to IL-1␤ with the concomitant appearance of p65 in the nucleus indicating activation of the NF-B pathway. There is, however, no effect of either PKC␦WT or PKC␦KD on the kinetics of these events. Transcriptional activity of NF-B in INS-1 cells stably transfected with the B luciferase reporter construct was also investigated (Fig. 7B). IL-1␤ induced a 20fold increase in luciferase expression over 4 h, which was significantly enhanced by overexpression of the PKC␦WT construct, as was basal luciferase expression. There was, however, no inhibition of IL-1␤-induced B-mediated luciferase expression by PKC␦KD. As with the ATF2 data presented above, these results suggest that the PKC␦ is sufficient to increase basal NF-B reporter activity but acts independently of IL-1␤.
Role of PKC␦ in iNOS mRNA Stability-Since we did not observe a requirement for PKC␦ activity in the regulation in the three major signaling pathways known to control IL-1␤induced iNOS gene expression in pancreatic ␤-cells, we next assessed whether PKC␦ might exert effects at the level of iNOS mRNA stability. This was assessed by stimulating INS-1 cells for 6 h with IL-1␤ and then halting transcription with actinomycin D. RNA was harvested at 2, 4, and 6 h after this point. The results demonstrate that in control cells iNOS mRNA is degraded following actinomycin D treatment, with iNOS mRNA having a half-life (t1 ⁄2 ) of ϳ6 h (Fig. 8). PKC␦WT clearly reduces the rate of degradation such that at 6 h postapplication of actinomycin D, iNOS message levels are less than 30% below the peak levels prior to actinomycin D treatment. Conversely, PKC␦KD overexpression increased the rate of iNOS mRNA degradation such that more then 80% of the message is degraded after 6 h in the presence of actinomycin D, and its t1 ⁄2 is reduced to less than 3 h. These observations strongly implicate a role for PKC␦ in iNOS mRNA stabilization and provide a

FIG. 4. Reciprocal modulation of IL-1␤-induced iNOS mRNA levels by PKC␦KD and PKC␦WT. INS-1 cells were infected with
MX17 control virus and PKC␦WT and PKC␦KD recombinant adenoviruses as described under "Experimental Procedures," and at 48 h postinfection they were exposed to 300 pg/ml IL-1␤ for 12 h. RNA was extracted using Trizol reagent, and RT-PCR was performed using oligonucleotides specific for rat iNOS. ␤-Actin PCR products were isolated at 18 cycles, and iNOS PCR products were isolated at 24 cycles (known to be the exponential phase for iNOS cDNA amplification). These were separated on a 1.5% agarose gel, stained with ethidium bromide, and analyzed using Bio-Rad and IPLabGel software. The top shows a plot of mean data Ϯ S.E. from four individual experiments (where *, p Ͻ 0.05; **, p Ͻ 0.01), while the bottom shows a representative gel. mechanism by which PKC␦ can regulate IL-1␤ induced iNOS mRNA levels and protein expression. DISCUSSION We have described a novel role for PKC␦ in regulating NO generation in IL-1␤-stimulated pancreatic ␤-cells. This was elucidated using specific PKC inhibitors and recombinant adenoviruses for overexpression of PKC␦WT or -KD. Our results indicate that PKC␦ activation was absolutely required for upregulation of iNOS mRNA and protein levels, as well as generation of NO, in cells stimulated for 12-24 h with IL-1␤. However, overexpression of PKC␦KD did not inhibit any of the three major transcriptional pathways (AP-1, ATF-2, and NF-B) known to be important for iNOS induction in IL-1␤-stimulated ␤-cells (1-6). Moreover, when transcription was inhibited using actinomycin D, iNOS mRNA was shown to degrade from stimulated levels with a t1 ⁄2 of ϳ6 h. This was shortened to less than 3 h in cells overexpressing PKC␦KD, whereas iNOS mRNA degradation was markedly inhibited in the presence of PKC␦WT.
IL-1␤ has previously been demonstrated to generate a rapid (Ͻ10-min) increase in the PKC activator diacylglycerol in pancreatic islets (43) and ␤-cell lines (5) and to induce the phosphorylation of known PKC substrates (43). Our results now suggest that it is PKC␦ that is specifically activated by IL-1␤ in INS-1 cells under these circumstances. Importantly, it is well documented that even short term exposure to IL-1␤ is sufficient to trigger much longer acting proinflammatory and proapoptotic events in pancreatic ␤-cells (2, 3). However, regulation of PKC␦ is complex. For example, it is known to be targeted by upstream kinases, both on a region known as the activation loop and on certain tyrosine residues, and phosphorylation of these sites augments PKC␦ activity (44). Therefore, we do not exclude the possibility that PKC␦ is activated in response to IL-1␤, and thereby stabilizes iNOS mRNA, by mechanisms other than a rapid rise in diacylglycerol.
Irrespective of the actual mechanism, our results clearly suggest that IL-1␤ initiates a bifurcating signaling cascade in the pancreatic ␤-cells; one arm controls iNOS gene transcription through a relatively well characterized signaling pathway, whereas a second arm, poorly understood but possessing a requirement for PKC␦, regulates iNOS mRNA stabilization. Both arms act in concert to ensure efficient regulation of NO generation. This is consistent with the observation that overexpression of PKC␦WT was not in itself sufficient to up-regulate iNOS expression in the absence of IL-1␤ stimulation. This also explains previous findings that activation of PKC with phorbol esters was insufficient to stimulate NO generation in clonal ␤-cells but that the phorbol esters could potentiate the response to IL-1␤ (45).
A similar dual control mechanism has been recently observed in other cell types for IL-1␤-induced genes such as cyclooxygenase-2 (46) and IL-8 (47) (48). However, in the published studies, mRNA stabilization appeared to involve activation of p38 rather than PKC, although the latter was not specifically addressed. Our data are the first to describe a role for PKC␦ in this bifurcating cascade, by which iNOS mRNA is stabilized. In our studies, p38 did not lie downstream of PKC␦ in a linear signaling pathway, since p38 activation was not affected by overexpression of either of the PKC␦ constructs. However, it is not excluded that both kinases might phosphorylate, and thereby regulate, a common substrate that acts as trans-acting factor in the control of mRNA stabilization.
The full importance that message stabilization has in regulating iNOS gene expression has only recently become apparent. Two studies (12,13) have now shown that iNOS mRNA is highly transcribed, but the 3Ј-untranslated region severely destabilizes the RNA structure such that it is efficiently degraded in an unstimulated cell. ATTTA motifs are thought to contribute to structural instability that targets the RNA for degradation, but these are specifically silenced in the presence of certain proinflammatory stimuli, the message is stabilized, and iNOS mRNA levels increase. In this context, a role for the RNA binding protein HuR has recently been invoked (13), but the upstream pathways linking this to receptor occupation are not defined. On the other hand, there is a limited literature to suggest that PKC does promote stabilization of some gene transcripts such as those of lactate dehydrogenase. An AT-rich region in the 3Ј-untranslated region of the lactate dehydrogenase mRNA sequence has been identified as the PKC-stabilizing region (49). The trans-acting factors that target this region are unknown. However, since iNOS and lactate dehydrogenase mRNA share similar AT-rich sequence motifs, it is clearly possible that PKC might regulate stability of both transcripts through a common mechanism.
It is well documented that PKC activation is sufficient for induction of NF-B-responsive genes in many cell types, but a demonstration that PKC is necessary for up-regulation of iNOS expression by proinflammatory stimuli has been limited to studies using macrophages and astrocytes (15)(16)(17). In only one instance was a specific requirement of PKC␦ reported, and even this was not selective, since PKC␣ and PKC␤I were also implicated (15). A role for PKC in stabilizing iNOS mRNA has only been proposed once previously, but this was based on the use of phorbol esters in macrophages and so could not discriminate between PKC isoforms (50). The PKC requirement in the earlier studies (15)(16)(17) was generally explained at the level of NF-B activation. In contrast, our results suggest that PKC␦ is not necessary for IL-1␤-stimulated NF-B activation in pancreatic islets. We did also observe an IL-1␤-independent up-regulation of NF-B (and ATF2) reporter activity in cells overexpressing PKC␦WT, but in both instances PKC␦ did not appear to be acting in the upstream signaling cascades. This suggests a site of action at the level of phosphorylation of p65 (or ATF-2) or of other proteins with which they interact to regulate transcription. However, since the effects of PKC␦ on NF-B and ATF2-mediated transcription were independent of IL-1␤, we have not further pursued the underlying mechanisms.
In addition to iNOS described here, several other genes have also been shown to be regulated by PKC␦, including manganese superoxide dismutase (51), a regulator of inflammation, and genes responsible for cell growth such as p21 and p27 (52,53) and differentiation, i.e. involucrin (54). The expression of many other regulatory genes can also be modulated in the presence of phorbol esters and/or PKC inhibitors, thus implicating various PKC isozymes (55,56). While in many instances regulation by PKC is likely to be transcriptional, our data and the recent finding that genes like iNOS are highly regulated at a posttranscriptional stage (12,13) and that some other gene transcripts may contain PKC recognition sequences (49) clearly suggest that post-transcriptional mRNA stabilization could provide a second mechanism by which PKC modulates gene expression. In the context of pancreatic ␤-cells, it will be of great interest to determine whether a PKC␦-mediated mRNA stabilization is also involved in regulation of other genes known FIG. 8. PKC␦ controls iNOS message stability. INS-1 cells were infected with MX17 control virus and PKC␦WT and PKC␦KD recombinant adenoviruses as described above, and at 48 h postinfection they were exposed to 300 pg/ml IL-1␤ for 6 h. Actinomycin D (1 M) was applied at 6 h, and RNA was extracted using Trizol reagent at 2, 4, and 6 h after. RT-PCR was performed over 28 cycles using oligonucleotides specific for rat iNOS. Products were separated on a 1% agarose gel and analyzed using Bio-Rad and IPLabGel software. Shown is a plot with mean data Ϯ S.E. from three individual experiments (*, p Ͻ 0.05; **, p Ͻ 0.001). f, MX17; ‚, PKC␦WT; E, PKC␦KD. The insert shows a representative gel. Mean mRNA levels were not significantly different for PKC␦WT at 6 h with IL-1␤ compared with MX17, while PKC␦KD reduces iNOS mRNA at 6 h to 77 Ϯ 6% that of MX17.

FIG. 7. NF-B activation and transcriptional activity is not dependent on PKC␦ activity. A, INS-1 cells were infected with MX17
control virus and PKC␦WT and PKC␦KD recombinant adenoviruses as described under "Experimental Procedures," and at 48 h postinfection they were exposed to 300 pg/ml IL-1␤ for the indicated time course. Whole cell lysates and nuclear extracts were separated by 10% SDS-PAGE, and proteins were transferred to nitrocellulose membrane. Whole cell lysates were probed using anti-IB␣ and anti-IB␤ antibodies. Nuclear extracts were probed using anti-p65 antibodies. Representative blots from four separate experiments are shown. B, INS-1 cells stably transfected with a -luciferase reporter were infected with adenovirus as described under "Experimental Procedures," and at 48 h postinfection they were exposed to 300 pg/ml IL-1␤ for a further 4 h. Luminescence was measured according to the manufacturer's instructions. Results are mean data of five separate experiments (**, p Ͻ 0.01; ***, p Ͻ 0.001 when compared with the relevant MX17 control). to be induced by IL-1␤.
In conclusion, we have demonstrated a novel requirement for PKC␦ in IL-1␤-induced iNOS induction in INS-1 and islet ␤-cells and shown this to be mediated by mRNA stabilization. IL-1␤-induced NO produced by iNOS has been implicated in ␤-cell dysfunction, secretory defects, and eventually cell death by apoptosis. Therefore, an obvious focus of future studies will be to determine whether inhibition of PKC␦ abrogates any aspect of the cellular dysfunction initiated by IL-1␤ in pancreatic ␤-cells.