Main

The destruction of pancreatic β-cells results in hypoinsulinaema and hyperglycaemia. Immunity and inflammation play an important role in β-cell destruction through the recruitment and activation of T cells and macrophages in pancreatic islets and the local production of inflammatory cytokines such as interleukin-1β (IL-1β), tumour necrosis factor-α (TNF-α) and interferon-γ (IFN-γ).1, 2 The pro-inflammatory cytokine IL-1β induces cell death in cultured islets.3 TNF-α can augment IL-1β-induced cytotoxicity of cultured β cells,4 whereas a combination of the cytokines IL-1β, TNF-α and IFN-γ has additive cytotoxicity for cultured islets.5 Evidence that this mechanism contributes to β-cell death in vivo comes from IL-1 and TNF-α blockade studies in the multiple low-dose streptozotocin (MLD-STZ) model of hyperglycaemia,6, 7 and blocking IL-1 action prolongs survival of mouse islets in an allograft model.8 In addition, a recent clinical trial showed that treatment with the IL-1 receptor antagonist improved blood glucose levels and β-cell secretory function and reduced markers of systemic inflammation in patients with type 2 diabetes.9

IL-1β and TNF-α are known to activate the three different mitogen-activated protein kinase (MAPK) pathways in pancreatic islets: extracellular signal-regulated kinase (ERK), p38 and c-Jun N-terminal kinase (JNK).10, 11, 12 Several studies have shown that JNK signalling is involved in IL-1β-induced β-cell death in vitro,13, 14 whereas most studies suggest that activation of ERK does not participate in cytokine-induced β-cell death.12, 15 However, the role of the p38 MAPK pathway is less clear with some studies finding that blockade of the p38 MAPK pathway can suppress β-cell death in the setting of cytokine or amylin-induced cytotoxicity or in islet grafts,15, 16, 17 whereas other studies have reported that p38 MAPK blockade can exacerbate cytokine-induced β-cell death or has no effect upon cytokine-induced β-cell death.13, 18

The p38 MAPK is activated by a wide variety of stresses such as pro-inflammatory cytokines (IL-1 and TNF-α), Toll-like receptor activation (innate immunity), reactive oxygen species, osmotic stress and UV irradiation.10 The p38 kinase is activated by dual phosphorylation of its activation loop through the action of the upstream kinases mitogen-activated protein kinase kinase 3 (MKK3) and MKK6, although other mechanisms of p38 activation can operate in response to specific stimuli.19 Once activated, the p38 kinase can phosphorylate a variety of transcription factors resulting in cellular responses such as apoptosis, inflammation and fibrosis.10 However, activation of this pathway can lead to different outcomes depending upon the individual stimulus and cell type involved. For example, p38 signalling promotes reactive oxygen species induced apoptosis of cultured tubular epithelial cells, whereas p38 signalling can suppress apoptosis in this cell type induced by the toxin 1,2-(dichlorovinyl)-L-cysteine.20, 21

Most of our knowledge on the contribution of p38 MAPK signalling to pancreatic islet damage comes from in vitro studies. However, one study has reported the use of systemic administration of a p38 inhibitor to prevent the onset of hyperglycaemia in non-obese diabetic mice.22 This is a very promising finding, and consistent with studies showing beneficial effects of systemic p38 inhibitor drugs in a range of disease models.23, 24, 25, 26, 27 However, there are concerns that systemic p38 blockade could potentially have deleterious effects.28 This raises the question of whether it may be possible to dissect the p38 pathway at the level of the upstream kinases, MKK3 and MKK6. A requirement for MKK3–p38 signalling has been demonstrated in mouse models of inflammatory arthritis.29 MKK3 is expressed in islets;30 however, the potential contribution of MKK3–p38 signalling in the development of pancreatic islet damage is unknown. To examine this question, we used mice deficient for the Mkk3 gene to examine the contribution of MKK3–p38 signalling in the MLD-STZ model in which both pro-inflammatory cytokines and the T-cell immune response play a pathogenic role in β-cell destruction leading to hyperglycaemia.6, 7, 31, 32, 33

MATERIALS AND METHODS

Multiple Low-Dose Streptozotocin (MLD-STZ) Model

Mkk3−/− gene deficient mice on the C57BL/6J background were bred in house at Monash Animal Services, Clayton, Australia.34 Male mice (22–26 g) were given daily intraperitoneal injections of 40 mg/kg streptozotocin (Sigma-Aldrich, St Louis, MO, USA) dissolved in 0.1 M sodium citrate pH 4.5 for 5 consecutive days. Blood glucose was measured once weekly following a 16 h fast using the glucose oxidase method. Hyperglycaemia was defined as >7 mmol/l fasting blood glucose. Animals were killed at different times (5 days, 2 or 4 weeks) after the start of STZ injections for analysis of the following parameters (group sizes are provided in the figure legends). Day 5 is the peak of STZ-induced islet cell apoptosis, which occurs before leukocyte infiltration. Week 2 is the peak of islet leukocyte infiltration and the peak of leukocyte-induced islet cell apoptosis. Week 4 is used for analysis of pancreatic insulin content and to allow for serial measurements of blood glucose. All animal experimentation was approved by the Monash Medical Centre Animal Ethics Committee.

Histochemistry

Gomori's aldehyde fuchsin staining was performed on 4 μm paraffin sections of formalin-fixed pancreatic tissue as described previously.35 The percent islet area stained was determined by image analysis using Image Pro Plus 4.0 software (Media Cybernetics, CA, USA). The perimeter of the islet was traced to measure the islet area and then the area of aldehyde fuchsin staining within the islet was assessed using a pre-set threshold and expressed as a percentage.

PAS staining of 4 μm formalin-fixed sections was used to score the degree of insulitis using a semiquantitative scoring system as follows: 0, normal; 1+, minor peri-islet mononuclear cell infiltration; 2+, moderate intra-islet mononuclear cell infiltration (<50% of islet area); 3+, severe intra-islet cell infiltration (>50% of islet area) with damage to islet architecture.

Pancreas Insulin Content

Groups of 4 wild-type (WT) and Mkk3−/− mice were killed 4 weeks after MLD-STZ. The pancreas was removed and weighed, and then insulin was extracted by homogenization in 5 ml acid ethanol (150 mM HCl in 75% ethanol). After overnight incubation at 4°C, insulin was quantified in the sample supernatant by ELISA (Linco Research, St Charles, MO, USA).

Immunohistochemistry Staining

Immunoperoxidase staining using an avidin–biotin complex (ABC) based system was performed on 4 μm paraffin sections. The following primary antibodies were used: rabbit antibody against cleaved caspase-3 (Cell Signaling Technology, Beverly, MA, USA); guinea-pig anti-insulin antibody (Dako); F4/80, which detects mouse macrophages (Serotec, Oxford, UK); GK1.5, which detects mouse CD4 and; YTS169.4, which detects mouse CD8.

To quantify apoptotic islet cells, serial sections were stained for cleaved caspase-3 and insulin. For each islet cross-section, the number of cleaved caspase-3 positive cells was counted within the islet. In the serial section, the area of insulin staining in each islet was analysed by image analysis to normalize the number of apoptotic cells relative to the area of insulin staining and record the number of apoptotic cells per mm2. This method was also used to confirm that apoptotic cells were insulin positive and to determine the proportion of apoptotic β-cells in the islets.

Isolation of Mouse Pancreatic Islets

Normal mice, or mice after 2 weeks of MLD-STZ, were killed and the common bile duct ligated at the distal end with nylon suture and then cannulated with a 30-gauge needle and syringe to infuse 3 ml of 1 mg/ml ice-cold collagenase P (Roche Biochemicals, Mannheim, Germany) in Hank's balanced salt solution (HBSS). The inflated pancreas was excised and incubated in 2 ml of collagenase P solution at 37°C for 12 min. After a brief vigorous shake to disperse the digested pancreas, islets were washed three times in HBSS with centrifugation for 30 s at 200 × g. The islet pellet was either resuspended in RNAlater (Ambion) for RNA extraction, or in culture media for apoptosis studies. The material was then put in petri dishes under a dissecting microscope with reflected lighting and a dark base. Individual islets were recovered using sterile forceps and then used for RNA extraction or cell culture studies.

Glucose-Induced Insulin Secretion by Isolated Islets

Islets were isolated from groups of four normal WT and Mkk3−/− mice. Triplicate samples of 8–15 hand-picked islets were incubated overnight at 37°C in RPMI-1640 media with 10% fetal calf serum. The islets were washed and incubated in 0.5 ml Krebs-Ringer Hepes-buffered saline (KRHS) in 3 mmol/l glucose at 37°C for 30 min and then incubated in 0.5 ml KRHS in 20 mmol/l glucose for another 30 min. Supernatants were assessed for insulin content by ELISA. Islets were extracted and analysed for total insulin. The percentage of insulin secreted by islets within the 30 min period of high glucose stimulation was calculated.

Induction of Apoptosis in Cultured Mouse Islets

Isolated pancreatic islets from normal WT or Mkk3−/− mice were isolated as described above, cultured for 72 h in DMEM with 10% FCS, and then incubated for 24 h with either 0.5 mM STZ or with a recombinant cytokine mix (10 ng/ml IL-1β, 10 ng/ml TNF-α, 10 ng/ml IFN-γ) (R&D Systems). Apoptosis was quantified by Cell Death Detection ELISA kit (Roche) and normalized against cellular DNA content.

Real Time RT-PCR

Total cellular RNA was extracted from isolated pancreatic islets using the RNAeasy Micro kit (Qiagen, Doncaster, Victoria, Australia) and reverse transcribed using the Superscript First-Strand Synthesis kit (Invitrogen) with oligo-dT primers. Real-time PCR was performed with the primers and probes listed in Table 1 using Rotor-Gene 3000 system (Corbett Research, Sydney, Australia) using the QuantiTect Probe PCR kit (Qiagen) with thermal cycling conditions of 37°C for 10 min to activate uracil–DNA glycosylase, 95°C for 15 min, followed by 45 cycles of 95°C for 15 s and 60°C for 60 s. Primers and probes used are listed in Table 1. Standard curves were established for target PCR products including cyclophilin, using serial dilutions of each purified PCR product. Samples were run and the relative abundance of each mRNA was calculated using the ΔΔCt method and normalized against the cyclophilin mRNA level. All samples were analysed twice and the average result taken.

Table 1 Sequences of primers and probes for real-time RT-PCR
Full size table

Statistical Analysis

Data are presented as mean±s.e.m. Data were analysed by one-way ANOVA with Newman–Keuls multiple comparison post-test or by Mann–Whitney U-test.

RESULTS

Characterization of Pancreatic Islet Function in Mkk3−/− Mice

Adult Mkk3−/− and WT mice had equivalent levels of fasting blood glucose (5.3±0.6 and 4.8±0.5 mmol/l in Mkk3−/− vs WT, respectively) and fasting plasma insulin (213±23 and 218±39 nmol/l in Mkk3−/− vs WT). In addition, glucose-induced insulin released by isolated pancreatic islets was equivalent in Mkk3−/− and WT mice (Figure 1a). Furthermore, basolateral glucose transporter (Glut2) mRNA levels were equivalent in isolated islets from Mkk3−/− and WT mice, which is important since STZ is taken up into β-cells via this transporter.36, 37 We attempted to analyse MKK3 and MKK6 expression and phosphorylation in islets from WT and Mkk3−/− mice; however, immunostaining of tissue sections was unsuccessful despite trying several commercial antibodies and a wide range of antigen retrieval conditions and western blotting was not sufficiently sensitive to detect MKK3 or MKK6 in lysates of isolated islets.

Figure 1
figure 1

Pancreatic islet function in Mkk3−/− mice. (a) Islets isolated from normal WT (open bars) or Mkk3−/− (closed bars) mice showed a very similar ability to secrete insulin in response to stimulation with high glucose. (b) RT-PCR analysis showed no difference in Glut2 mRNA levels between islets isolated from normal WT (open bars) and Mkk3−/− (closed bars) mice.

Full size image

MLD-STZ-Induced Hyperglycaemia

Control WT mice developed hyperglycaemia within 2 weeks of MLD-STZ injections (Figure 2a), with an accumulative 80% incidence (33/41 mice with hyperglycaemia at week 4). Consistent with the development of hyperglycaemia, STZ-WT mice showed a marked reduction in pancreatic insulin levels (Figure 2b), and a substantial loss of Fuchsin stained insulin granules in islet β-cells at the 4-week time point (Figure 2c–e). In contrast, Mkk3−/− mice were completely protected from MLD-STZ-induced hyperglycaemia (0/17 mice with hyperglycaemia at week 4) (Figure 2a). Indeed, Mkk3−/− mice were still protected from hyperglycaemia at 8 weeks after MLD-STZ administration (data not shown). There was a partial reduction in pancreatic insulin levels and in the area of islet Fuchsin staining in STZ-Mkk3−/− mice, but this was mild compared to the changes seen in STZ-WT mice, and did not affect blood glucose levels (Figure 2b–f).

Figure 2
figure 2

Mkk3−/− mice are resistant to hyperglycaemia in MLD-STZ. (a) Fasting blood glucose was measured on a weekly basis after MLD-STZ administration. Most WT mice developed hyperglycaemia (>7 mmol/l) by week 3. In contrast, fasting blood glucose remained normal in Mkk3−/− mice after MLD-STZ administration (***P<0.0001 vs WT). (b) The insulin content of whole pancreas tissue was measured in normal mice (open bars) and mice at week 4 of MLD-STZ (closed bar) (n=4). (c) The degree of β-cell granulation in pancreatic islets was assessed by Gomori's aldehyde fuchsin staining in normal mice (open bars) and at week 4 of MLD-STZ (closed bars) in WT and Mkk3−/− mice (n=7–8). Examples of Fuchsin staining in: (d) normal WT mice; (e) MLD-STZ in WT mice showing a dramatic reduction in β-cell granulation, and; (f) MLD-STZ in Mkk3−/− mice which are largely protected from loss of β-cell granulation.

Full size image

MLD-STZ-Induced Islet Cell Apoptosis

Loss of β-cells through apoptosis is an important mechanism in the induction of hyperglycaemia in the MLD-STZ model.38 There are two peaks of β-cell apoptosis—an early induction of apoptosis before islet inflammation on day 5 and a second peak during the development of islet inflammation.38 We examined islet cell apoptosis by immunostaining for cleaved caspase-3. Examination of MLD-STZ WT mice demonstrated significant islet cell apoptosis on day 5 (in the absence of detectable islet inflammation) and on day 14 (Figure 3). MLD-STZ in Mkk3−/− mice induced significant islet cell apoptosis on day 5; however, Mkk3−/− mice showed a marked reduction in the second phase of islet cell apoptosis on day 14 (Figure 3). To examine whether MKK3 signalling is directly involved in islet cell apoptosis, we performed in vitro studies using islets isolated from normal WT and Mkk3−/− mice. The addition of STZ or a cytokine mix induced significant apoptosis in WT islets, with no protection from STZ- or cytokine-induced apoptosis evident in Mkk3−/− islets (Figure 4).

Figure 3
figure 3

Islet cell apoptosis in MLD-STZ. Immunostaining for cleaved caspase-3 in pancreas tissue sections. (a) Normal WT mouse pancreas with no staining. (b) Day 14 of MLD-STZ in WT mice showing several apoptotic cells with nuclear staining for cleaved caspase-3 (arrows). (c) Fewer apoptotic cells (arrows) are seen in islets of Mkk3−/− mice on day 14 of MLD-STZ. (d) Quantification of apoptotic cells based upon immunostaining for cleaved caspase-3 in normal mice (n=3), on day 5 (n=4–5), or on day 14 (n=8–10) after STZ administration in WT (open bars) and Mkk3−/− (closed bars) mice. *P<0.05 vs WT by ANOVA with Newman–Keuls multiple comparison post-test.

Full size image
Figure 4
figure 4

STZ- and cytokine-induced apoptosis of cultured islets. Islets were isolated from normal WT or Mkk3−/− mice, cultured and then stimulated for 24 h with either 0.5 mM STZ or a cytokine mix (CTK=10 ng/ml IL-1β, 10 ng/ml TNF-α, 10 ng/ml IFN-γ). Apoptosis was measured using the cell death detection ELISA kit. Values are presented as an OD ratio of sample to untreated negative control after normalizing by DNA content. There are no significant differences between WT and Mkk3−/− islets.

Full size image

MLD-STZ-Induced Islet Inflammation

The second phase of MLD-STZ-induced pancreatic damage results in the development of hyperglycaemia at 2 weeks. Previous studies in this model have shown that insulitis leading to a loss of insulin producing β-cells and the development of hyperglycaemia is dependent on T cells and macrophages,31, 39, 40, 41, 42, 43 with roles for individual cytokines (IL-1, TNF-α and IFN-γ) in β-cell destruction identified.6, 7, 32, 44 In the current study, MLD-STZ WT mice developed significant insulitis at 2 weeks (Figure 5). Immunohistochemistry staining showed peri-islet and intra-islet infiltration of F4/80+ macrophages, CD4+ T cells and CD8+ T cells in islets at the 2-week time point (Figure 6). Furthermore, MLD-STZ WT mice showed a significant increase in islet mRNA levels for IFN-γ, IL-4, IL-1β, TNF-α, inducible nitric oxide synthase (iNOS) and monocyte chemoattractant protein-1 (MCP-1) as determined by real-time RT-PCR analysis of isolated islets (Figure 7). Western blotting was attempted to examine the islet content of these proteins, but was unsuccessful due to insufficient material in these samples.

Figure 5
figure 5

Insulitis in MLD-STZ. The degree of insulitis was assessed on PAS-stained sections. (a) An untreated WT mouse shows normal islet structure. (b) An example of peri-islet and intra-islet mononuclear cell infiltration with a marked loss of architecture in a WT mouse 2 weeks after MLD-STZ (3+ score). (c) Graph showing the degree of insulitis as assessed using a semi-quantitative scoring system. *P<0.001 compared to WT by Mann–Whitney U-test.

Full size image
Figure 6
figure 6

Macrophage and T-cell infiltration in MLD-STZ. Immunostaining of normal WT mice shows a small number of F4/80+ macrophages in pancreatic islets and only occasional CD4+ or CD8+ T cells in normal islets. At 2 weeks after MLD-STZ in WT mice there is a marked peri-islet and intra-islet F4/80+ macrophage infiltrate and a marked islet infiltrate of CD4+ and CD8+ T cells. In contrast, no leukocytic infiltrate is seen in Mkk3−/− mice at 2 weeks after MLD-STZ. Photomicrographs of immunostaining are representative of leukocyte infiltration as analysed in groups of 4–6 mice.

Full size image
Figure 7
figure 7

Analysis of pro-inflammatory molecules in islets in MLD-STZ. Pancreatic islets were isolated from normal WT and Mkk3−/− mice (n=4–5), and from WT and Mkk3−/− mice 2 weeks after MLD-STZ (n=8–10). Islet mRNA levels were analysed by real-time RT-PCR for: (a) IFN-γ, (b) IL-4, (c) IL-1β, (d) TNF-α, (e) iNOS and (f) MCP-1. Normal mice (day 0, open bars) and MLD-STZ mice (day 14 MLD-STZ, closed bars). *P<0.05 compared to WT by ANOVA.

Full size image

Mkk3−/− mice were largely protected from the second phase of MLD-STZ-induced pancreatic damage at the 2-week time point. This was shown by a reduction in the severity of insulitis (Figure 5). Consistent with this analysis, immunostaining found no significant peri- or intra-islet infiltration by F4/80+ macrophages, CD4+ T cells or CD8+ T cells (Figure 6). This reduction in leukocyte infiltration in STZ Mkk3−/− mice was associated with a lack of upregulation of the islet mRNA levels for IFN-γ, IL-1β, TNF-α and MCP-1 (Figure 7). However, there was an increase in islet IL-4 and iNOS mRNA levels in MLD-STZ Mkk3−/− mice that is comparable to that seen in MLD-STZ in WT mice (Figure 7), which is likely due to expression of these molecules by intrinsic islet cells.

DISCUSSION

Mice deficient in the Mkk3 gene were found to have normal islet function in terms the ability to secrete insulin in response to glucose stimulation. The normal levels of blood glucose and plasma insulin in fasted animals also indicates normal islet function in Mkk3−/− mice. Thus, the remarkable resistance of Mkk3−/− mice to MLD-STZ induced hyperglycaemia is likely to be due to a modified immune response. The potential mechanisms by which Mkk3−/− mice are protected from hyperglycaemia in this model are considered below.

The MLD-STZ model has two distinct phases of pancreatic injury.38 STZ induces acute toxic effects upon islet β-cells which is followed by leukocytic infiltration and the production of pro-inflammatory cytokines which completes β-cell destruction leading to hyperglycaemia. Mkk3−/− mice were not protected from the acute toxic effects of MLD-STZ administration on the basis that levels of islet cell apoptosis were comparable to than seen in MLD-STZ WT mice on day 5. In addition, cultured Mkk3−/− islets were equally susceptible to STZ induced apoptosis as those from WT mice. This is consistent with the equivalent expression of Glut2, the transport molecule that takes STZ into β-cells,36, 37 in WT and Mkk3−/− islets.

The protective effect of Mkk3 gene deletion in the MLD-STZ model was attributed to suppression of islet inflammation on day 14 following the initial STZ-mediated pancreatic insult. It is known that both T cells and macrophages play an important role in islet destruction leading to hyperglycaemia in the MLD-STZ model,31, 39, 40, 41, 42, 43 and thus the lack of T cell and macrophage infiltration seen in Mkk3−/− mice is likely to be a major reason for their protection from MLD-STZ induced hyperglycaemia. Furthermore, infiltrating macrophages and T cells are a major source of the pro-inflammatory cytokines IL-1β, TNF-α and IFN-γ that promote islet destruction in this model.6, 7, 17, 32, 44 However, it should be noted that Mkk3−/− islets were not protected from cytokine-induced apoptosis in vitro, suggesting that protection of Mkk3−/− mice from islet destruction in the MLD-STZ model is due to a failure to upregulate islet mRNA levels for these pro-inflammatory cytokines rather than any intrinsic resistance of Mkk3−/− β-cells to cytokine-induced cytotoxicity. In addition, this finding is consistent with a study in which treatment with a p38 inhibitor drug prevented upregulation of IL-1β and TNF-α mRNA levels in human islets and improved their function after transplantation into diabetic athymic mice.17

The lack of islet monocyte/macrophage infiltration in Mkk3−/− mice in the MLD-STZ model may be due to a failure to upregulate islet MCP-1 mRNA levels as shown by real-time RT-PCR analysis of isolated islets. This, in turn, may be due to the failure to upregulate islet IL-1β mRNA levels since IL-1 has been shown to induce MCP-1 production by cultured human islets and rat β-cells.45, 46 Furthermore, IL-1 induced upregulation of MCP-1 mRNA levels in cultured rat β-cells is, in part, dependent upon p38 MAPK signalling.45

It is interesting that the lack of MKK3–p38 signalling did not prevent upregulation of IL-4 mRNA in islets in the MLD-STZ model. Previous studies have shown that IL-4 is produced by intrinsic islet cells,32 indicating that STZ-induced pancreatic injury was able to upregulate IL-4 mRNA levels in the absence of leukocyte infiltration and MKK3–p38 signalling. However, IL-4 is not protective in the MLD-STZ model.32

The role of nitric oxide in islet damage, including cytokine-induced islet cytotoxicity, is a controversial topic.47 A number of studies have argued against a role for nitric oxide in the MLD-STZ model,48, 49 whereas iNOS−/− mice are partially protected from MLD-STZ-induced hyperglycaemia with a reduction in islet leukocytic infiltration.50 Consistent with previous studies, we found a significant increase in islet iNOS mRNA in WT mice in the MLD-STZ model. Of interest, normal Mkk3−/− mice showed higher basal levels of iNOS mRNA in islets compared to normal WT mice, suggesting an inhibitory role for MKK3–p38 signalling in regulation of iNOS gene expression in intrinsic islet cells. MLD-STZ in Mkk3−/− mice induced an increase in islet iNOS mRNA levels even in the absence of an increase in the expression of well-known inducers of iNOS expression, such as IFN-γ, IL-1β and TNF-α. Indeed, Mkk3−/− mice were protected from apoptosis at day 14 of MLD-STZ despite augmented islet iNOS mRNA levels, suggesting that nitric oxide may only be a minor player in islet apoptosis in vivo, although further studies are needed to validate this conclusion.

The results of the current study are consistent with a recent report in which administration of a p38 inhibitor, FR167653, was shown to reduce insulitis and development of hyperglycaemia in non-obese diabetic mice.22 However, there are concerns that p38 blockade could have deleterious effects based upon the fetal lethal phenotype of p38α gene knockout mice,51 impaired clearance of pneumococcal pneumonia and tuberculosis infections in mice treated with a p38 inhibitor,52 exacerbation of renal injury in some models of kidney disease53, 54 and clinical trials in which p38 inhibitors have caused hepatotoxicity.28 Our study has identified that signalling via MKK3 is sufficient for leukocyte-mediated islet cell destruction leading to hyperglycaemia. This opens up the possibility for the use of selective MKK3 inhibitors to prevent the onset of hyperglycaemia, or to prevent its reoccurrence after transplantation.

In conclusion, MKK3 signalling plays an essential role in the development of hyperglycaemia in the MLD-STZ model. This is postulated to operate via MKK3 signalling in damaged islets leading to MCP-1 production that, in turn, induces macrophage infiltration with the leukocyte-derived cytokines causing destruction of the remaining islet cells resulting in insulin deficiency and hyperglycaemia.