Deletion of protein kinase Cδ in mice modulates stability of inflammatory genes and protects against cytokine-stimulated beta cell death in vitro and in vivo
- First Online:
- Cite this article as:
- Cantley, J., Boslem, E., Laybutt, D.R. et al. Diabetologia (2011) 54: 380. doi:10.1007/s00125-010-1962-y
- 689 Downloads
Proinflammatory cytokines contribute to beta cell destruction in type 1 diabetes, but the mechanisms are incompletely understood. The aim of the current study was to address the role of the protein kinase C (PKC) isoform PKCδ, a diverse regulator of cell death, in cytokine-stimulated apoptosis in primary beta cells.
Islets isolated from wild-type or Prkcd−/− mice were treated with IL-1β, TNF-α and IFNγ and assayed for apoptosis, nitric oxide (NO) generation and insulin secretion. Activation of signalling pathways, apoptosis and endoplasmic reticulum (ER) stress were determined by immunoblotting. Stabilisation of mRNA transcripts was measured by RT-PCR following transcriptional arrest. Mice were injected with multiple low doses of streptozotocin (MLD-STZ) and fasting blood glucose monitored.
Deletion of Prkcd inhibited apoptosis and NO generation in islets stimulated ex vivo with cytokines. It also delayed the onset of hyperglycaemia in MLD-STZ-treated mice. Activation of ERK, p38, JNK, AKT1, the ER stress markers DDIT3 and phospho-EIF2α and the intrinsic apoptotic markers BCL2 and MCL1 was not different between genotypes. However, deletion of Prkcd destabilised mRNA transcripts for Nos2, and for multiple components of the toll-like receptor 2 (TLR2) signalling complex, which resulted in disrupted TLR2 signalling.
Loss of PKCδ partially protects against hyperglycaemia in the MLD-STZ model in vivo, and against cytokine-mediated apoptosis in vitro. This is accompanied by reduced NO generation and destabilisation of Nos2 and components of the TLR2 signalling pathway. The results highlight a mechanism for regulating proinflammatory gene expression in beta cells independently of transcription.
KeywordsBeta cell Cytokines Islets of Langerhans mRNA stabilisation NOS2 Protein kinase C Streptozotocin Toll-like receptor Type 1 diabetes
Thymoma viral proto-oncogene 1
B-cell leukemia/lymphoma 2
DNA-damage-inducible transcript 3
Eukaryotic translation initiation factor 2 alpha kinase 3
Extracellular signal regulated kinase
Glucose-stimulated insulin secretion
c-Jun N-terminal kinase
Protein kinase C
Mitogen-activated protein kinase
Myeloid cell leukemia sequence 1
Multiple low dose streptozotocin
Nuclear factor kappa B
Inducible nitric oxide synthase
Apoptotic destruction of pancreatic beta cells as part of an auto-inflammatory response is the hallmark of type 1 diabetes [1, 2, 3, 4]. The release of proinflammatory cytokines, such as IL-1β, TNF-α and IFNγ from the monocytic infiltrate is a key feature of the progression of the disease [1, 2, 3, 4]. IL-1β signalling in beta cells includes activation of various mitogen-activated protein kinase (MAPK) cascades. Another important apoptotic pathway is mediated by nuclear factor kappa B (NFκB), which leads to increased expression of inducible nitric oxide synthase (NOS2), thereby enhancing generation of the free radical nitric oxide (NO). The latter is a key, but not sole, mediator of beta cell apoptosis following exposure to proinflammatory cytokines, and in experimental models of type 1 diabetes [1, 2, 3, 4].
Another signalling pathway initiated by IL-1β involves activation of protein kinase C (PKC) . This comprises a family of serine/threonine protein kinases consisting of 11 isoforms with differing tissue distributions and functions, as well as varying co-factor requirements and substrates . Work from our laboratory established that the PKCδ isoform is activated following stimulation of INS-1 insulinoma cells with IL-1β, and that this activation contributed to cytokine-mediated apoptosis [7, 8]. In this model, PKCδ regulated Nos2 at post-transcriptional level, by actively stabilising Nos2 mRNA, thereby increasing its half-life . However, PKCδ is now widely recognised as a component of apoptotic signalling pathways triggered by diverse cytotoxic agents, including those such as ultraviolet radiation and DNA-damaging agents, in which NOS2 induction does not play a role. In these instances, PKCδ can impinge upon substrates in the distal steps of the apoptotic cascade [9, 10, 11]. Furthermore, a recent study has demonstrated a requirement for PKCδ in mediating beta cell apoptosis in vivo in response to high-fat feeding, a situation not thought to involve NO generation . To date, however, a role for PKCδ in animal models of type 1 diabetes has not been tested.
Another potential role of PKCδ is in linking endoplasmic reticulum (ER) stress to c-Jun N-terminal kinase (JNK) signalling and thus apoptosis . This stress response is activated in beta cells by cytotoxic cytokines [14, 15, 16, 17, 18], potentially as a result of depletion of Ca2+ from the lumen of the ER, such that the ability of the latter to fold and export secretory protein is compromised . Signalling molecules in the ER membrane sense the accumulation of unfolded proteins and initiate an adaptive response. However, if this fails to rebalance the folding capacity of the ER, apoptosis is activated [20, 21]. In beta cells, the best characterised of the transmembrane ER stress sensors is the protein kinase, eukaryotic translation initiation factor 2 alpha kinase 3 (EIF2AK3 also known as PERK), which (among other effects) is mainly responsible for transcriptional upregulation of the pro-apoptotic transcription factor, DNA-damage-inducible transcript 3 (DDIT3, also known as CHOP).
The goal of the current study was to determine whether activation of PKCδ by cytokines contributes to cytokine-mediated apoptosis in primary beta cells and to investigate the underlying mechanisms, with particular reference to ER stress and mRNA stabilisation of inflammatory genes such as Nos2. We also sought to test the potential involvement of PKCδ in a mouse model of type 1 diabetes.
Prkcd null mice were generated by insertion of a LacZ/neo cassette into the first transcribed exon of the gene, as previously described [22, 23]. Routine genotyping was carried out by PCR analysis of tail-tip DNA, using a forward primer corresponding to a 5′ untranslated region of the Prkcd locus and reverse primers corresponding to either exon 1 (wild type)—or the Lac-Neo insert (see Electronic supplementary material [ESM] Table 1). Ethics approval for mouse studies was granted by the Garvan Institute/St Vincent’s Hospital Animal Ethics Committee. Mice were maintained on a hybrid 129/SV C57BL/6 background, using Prkcd heterozygous breeding pairs. They were fed a standard chow diet and had free access to drinking water. Age-matched wild-type and Prkcd−/− littermates (8–12 weeks old) were used for experiments.
A stock solution of 4 mg/ml streptozotocin (STZ) was prepared in 0.1 mol/l sodium citrate (pH 4.5) immediately before usage. After fasting for 4–6 h, mice were injected i.p. with 40 mg/kg STZ once a day for 5 days. On days 2, 5, 9, 11, 12, 16 and 24, food was withdrawn in the morning and blood withdrawn from the tail vein 6 h later for analysis of blood glucose. In some instances, mice were killed at day 11, and pancreases were removed and fixed in 4% paraformaldehyde for immunohistochemistry.
Islet isolation, cytokine treatment and insulin secretion assays
Islets were isolated by pancreatic digestion, and purified using a Ficoll-paque gradient (GE Healthcare, Chalfont St Giles, UK) before overnight culture in RPMI 1640 with 11 mmol/l glucose and 10% FCS (Invitrogen, Mulgrave, Vic, Australia). A cytokine mixture comprising IL-1β, TNF-α and IFNγ was added at concentrations and for times indicated in the text. In some studies the selective TLR2 agonist Pam2CSK4 (Invivogen, San Diego, CA, USA) was added at 1 μg/ml for 30 min. For insulin secretion assays, islets were preincubated for 1 h in HEPES-buffered KRB containing 0.1% BSA and 2 mmol/l glucose. Batches of five islets were incubated at 37°C for 1 h in 130 μl KRB containing 0.1% BSA and 2 mmol/l glucose (basal), supplemented with glucose (20 mmol/l), or KCl (25 mmol/l) as indicated in the text. Insulin release was determined by RIA (Linco/Millipore, Billerica, MA, USA).
Apoptosis and NO assays
Apoptosis was measured using an ELISA kit (Roche Applied Science, Castle Hill, NSW, Australia) which quantifies the apoptotic mono- and oligonucleosomes in a sample. Islets (40–80) were lysed in 0.2 ml of the supplied lysis buffer, incubated for 30 min at room temperature, and the lysate was spun at 200×g for 10 min . The assay was performed using 20 μl of the supernatant fraction in the ELISA according to the manufacturer’s instructions. NO was assayed using the Greiss reaction as previously described  using a 0.1 ml aliquot of the islet culture medium.
Islets were washed, lysed and protein content determined by bicinchoninic acid assay (Pierce/Thermo Scientific, Rockford, IL, USA). Fifteen micrograms of protein was resolved on a 7% SDS-PAGE gel (Invitrogen) before transfer to polyvinylidene fluoride membranes. Membranes were blocked with milk and probed with the following antibodies for 2 h at room temperature: anti-PKCα (610108) and anti-PKCε (610086; BD biosciences, San Jose, CA, USA); anti-PKCβ (SC209), anti-PKCδ (SC213), anti-PKCζ (SC216), anti-DDIT3/CHOP (SC575), anti-IκBβ (SC945), anti-Mcl1 (SC819), anti-NOS2 (SC651) and anti-14-3-3β (SC1657; Santa Cruz Biotechnology, Santa Cruz, CA, USA); Anti-BCL2 (2876), anti-phospho-JNK (Thr183/Tyr185, 9251), anti-phospho-p44/42 MAPK (Thr202/Tyr204, 4377), anti-phospho-AKT1 (S473, 9271), anti-phospho-p38 MAPK (Thr180/Tyr182, 9211) and anti-phospho-EIF2AK3/PERK (Thr980, 3179; Cell Signaling Technology, Danvers, MA, USA); anti-α-tubulin (T90267) and anti-β-actin (Sigma, St Louis, MI, USA) and anti-rabbit or anti-mouse horseradish peroxidase-conjugated secondary antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA, USA). Following chemiluminescent detection, densitometry was performed using ImageJ1.38q (NIH, Bethesda, MD, USA).
Quantitative PCR and mRNA stabilisation
Islets were exposed to 400 U/ml IL-1β for 6 h, and then treated with 1 μmol/l actinomycin D for times as indicated. RNA was extracted and RT-PCR was carried out as previously described using primers as listed in ESM Table 1 for Nos2, β-actin (Actb), and Ddit3. Alternatively, cDNA derived from islets before and 3 h after actinomycin D was applied to a PCR array containing 84 inflammation genes (SABiosciences, Frederick, MD, USA), RT-PCR was performed using an ABI PRISM7900 HT instrument (Applied Biosystems, CA, USA) and data were analysed according to the manufacturer’s instructions. Gene expression was normalised to that of the mean of four housekeeping genes: Gusb, Hprt1, Hsp90ab1 and Gapdh.
Statistics were performed using GraphPad Prism5 (Graphpad Software, La Jolla, CA, USA) or Excel (Microsoft, Redmond, WA, USA) software. Paired and unpaired t tests and two-way ANOVA were performed as appropriate. A p value of <0.05 was regarded as significant.
Pancreatic islets from Prkcd−/− mice are deleted in PKCδ without compensation from other PKC isoforms
Deletion of PKCδ partially protects against apoptosis due to cytokines, but does not alter glucose-stimulated insulin secretion
Proximal cytokine signalling pathways in pancreatic islets are unaltered by deletion of Prkcd
Cytokine-stimulated ER stress is not modulated by Prkcd deletion
PKCδ contributes to Nos2 mRNA stabilisation
PKCδ post-transcriptionally regulates multiple mRNA transcripts in islets
TLR2 signalling is inhibited by deletion of Prkcd
A role for PKCδ in regulating expression of components of the TLR2 signalling complex might be expected to impact on signalling downstream of that receptor. We tested this directly using the specific TLR2 agonist Pam2CSK4. As shown in Fig. 6b, this compound resulted in large fold increases in the phosphorylation of JNK, ERK and p38 in wild-type islets. In all instances these responses were less pronounced in islets deleted in Prkcd, and this difference attained statistical significance in the case of JNK and ERK.
Mice deleted in Prkcd are partially protected in an autoimmune model of beta cell destruction
PKCδ is a major mediator of diverse forms of apoptosis in many cell types [9, 10, 11]. Its role in beta cells, however, has chiefly been addressed in the context of regulation of GSIS, which has proved controversial [40, 41, 42]. In our hands, functional inhibition of PKCδ by overexpression of a kinase-dead mutant of the enzyme in isolated rat islets using adenovirus was without effect on GSIS . Moreover, overexpression of a similar construct specifically in beta cells of transgenic mice also failed to demonstrate a major role in regulating secretion . By contrast, a partial requirement for PKCδ in GSIS was reported using islets from Prkcd knockout mice . The results shown here, using a different knockout model, are in keeping with the results of the kinase-dead approach, suggesting that PKCδ plays little role in GSIS. We also demonstrate for the first time that activation of PKCδ does not appear to be involved in the inhibition of insulin secretion that occurs following exposure to cytokines.
As a member of the novel subgroup of PKC isoforms, PKCδ is traditionally activated in receptor signalling cascades by diacylglycerol, which is generated in turn from the breakdown of phosphoinositides by phospholipase C . The best characterised function of PKCδ is as a positive regulator of apoptosis, but the underlying mechanisms are complex and diverse [9, 10, 11]. We now show that loss of PKCδ protects against beta cell death in response to cytokines but not STZ. These results point to a role for PKCδ specifically in cytokine-stimulated apoptosis, as STZ acts via chemical disruption of DNA and non-enzymatic generation of NO in vitro . Note that this is different from its mode of action in vivo, where multiple low doses of STZ are thought to trigger an autoimmune attack on beta cells [38, 39]. Our current results are therefore consistent with previous data showing that cytokines, especially IL-1, activate the phospholipase C pathway in beta cells , and we previously provided evidence that PKCδ serves as a component of the downstream signalling cascade . The present study, however, strongly supports the argument against a major involvement of PKCδ as an upstream component in the ERK, JNK and p38 pathways that are activated by cytokines and implicated in beta cell apoptosis.
Work from several laboratories has demonstrated that cytokines also induce ER stress in beta cells [14, 15, 16, 17, 18]. However, the extent to which this contributes to apoptosis under these conditions remains to be resolved. We favour the view that, at least in vitro, ER stress makes less of a contribution to beta cell apoptosis in response to cytokines than it does with other cytotoxic stimuli such as saturated fatty acids . Here we demonstrate that induction of the ER stress markers phospho-EIF2A3 and DDIT3 by cytokines is not altered by deletion of Prkcd, even though there was a protection against apoptosis. This would suggest that PKCδ acts either independently of ER stress, or downstream of it. Because we observed modulation of NO generation, but not ER stress, our results would tend to support the argument against the view that NO serves as an upstream trigger of ER stress in cytokine-stimulated apoptosis [19, 45]. However, the 25% inhibition of NO demonstrated here may have been insufficient to affect ER stress.
Both our current results and previous studies [36, 37] suggest that a major protective role afforded by deletion of Prkcd might involve the destabilisation of multiple mRNA transcripts. Preeminent among these would be Nos2 such that the corresponding mRNA would be degraded more rapidly in the absence of PKCδ, thereby limiting NO generation. In general, however, the role of PKCδ in Nos2 expression and NO generation shown here for primary islets is more modest than that previously elaborated using INS-1 cells . Despite this, we did observe a protective effect of Prkcd deletion using the MLD-STZ model. This suggests that mechanisms in addition to reduced NO generation might be active in vivo. In support of this, we provided data suggesting that PKCδ potentially regulates stabilisation of multiple beta cell gene transcripts, and to a greater extent than Nos2. Although these results will need to be confirmed and extended in future studies, it seems more than coincidental that many of the candidate genes cluster to the IL-R1 and/or TLR2 signalling pathways. We believe the latter would provide the more productive avenue for further investigation. First, we were unable to provide direct evidence for a downregulation of IL-1R signalling (although we only examined this in conjunction with TNF-α and IFNγ, which may have confounded a modest effect). Second, TLR2 signalling to MAPK pathways was disrupted by deletion of Prkcd. This is noteworthy, as TLR2 has been implicated in beta cell dysfunction in both type 1 and type 2 diabetes [46, 47, 48]. Interestingly, the role of TLR2 in the setting of type 1 diabetes appears to involve recruitment of autoimmune cells that mediate secondary necrosis, rather than a beta cell autonomous effect on apoptosis . Whether PKCδ also specifically modulates this process is a topic for future studies. However, Prkcd−/− mice subjected to the MLD-STZ protocol showed a significantly lower increase in fasting blood glucose compared with wild-type mice, although the time of onset of hyperglycaemia was similar between the two groups. This would be consistent with activation of PKCδ playing a modulatory role in progression of the disease. The MLD-STZ model, however, generates only a mild insulitis [38, 39], and so we were unable to quantify the extent to which loss of PKCδ protects via effects in the beta cell directly, versus effects within immune cells potentially regulating their recruitment during insulitis. Such analyses would be better undertaken in future by backcrossing Prkcd null mice onto the NOD background.
In conclusion, there is a growing appreciation that cytokines can regulate expression of (especially proinflammatory) genes via mRNA stabilisation, as well as via transcription [49, 50]. Various PKC isoforms, including PKCδ, are implicated in these stabilisation pathways [36, 37]. Although much effort has been devoted to elucidating the transcriptional networks regulated by cytokines in beta cells, our results suggest that mRNA stabilisation pathways might also be a topic of great relevance to type 1 diabetes, particularly in the context of a role for PKCδ.
This work was supported in part by a grant from the Juvenile Diabetes Research Foundation (JDRF), a joint JDRF/National Health and Medical Research Council Special Program Grant, a postdoctoral research fellowship from the Faculty of Medicine, University of New South Wales, and a generous donation from S. Gross (Sydney, NSW, Australia). J. Cantley holds the GlaxoSmithKline–Don Chisholm Research Fellowship. We are grateful to D. Robinson and A. Davenport for technical support with the mouse studies, and to C. King for provision of reagents and helpful advice (all at the Garvan Institute of Medical Research, Sydney, NSW, Australia).
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.