Making ß(-like)-cells from exocrine pancreas.

Creating an abundant source of ß(-like)-cells has been a major goal in diabetes research for many decades. The concept of cell plasticity has inspired many strategies towards regenerative medicine, but its successes have been limited until very recently. Today, most cell types in the pancreas are considered candidates for the generation of ß(-like)-cells through transdifferentiation. While ß(-like)-cells that are in vitro differentiated from human embryonic stem cells are already being grafted in patients, ß(-like)-cells generated by transdifferentiation are not yet ready for clinical application. These cells would however offer several advantages over the current ß(-like)-cells generated by directed differentiation, especially concerning safety issues. In addition, perfect control of the transdifferentiation efficiency would through targeted drug delivery support a non-invasive cell therapy for diabetes. Lastly, focusing on the exocrine pancreas as prime candidate makes sense in view of their abundance and high plasticity. Keeping these hopeful perspectives in mind, it is worth to continue focused research on the mechanisms that control transdifferentiation from pancreas exocrine to ß-cells.

Accessory cells for ß-cell transplantation.

Despite recent advances, insulin therapy remains a treatment, not a cure, for diabetes mellitus with persistent risk of glycaemic alterations and life-threatening complications. Restoration of the endogenous ß-cell mass through regeneration or transplantation offers an attractive alternative. Unfortunately, signals that drive ß-cell regeneration remain enigmatic and ß-cell replacement therapy still faces major hurdles that prevent its widespread application. Co-transplantation of accessory non-islet cells with islet cells has been shown to improve the outcome of experimental islet transplantation. This review will highlight current travails in ß-cell therapy and focuses on the potential benefits of accessory cells for islet transplantation in diabetes.

Human blood outgrowth endothelial cells improve islet survival and function when co-transplanted in a mouse model of diabetes.

AIMS/HYPOTHESIS: As current islet-transplantation protocols suffer from significant graft loss and dysfunction, strategies to sustain the long-term benefits of this therapy are required. Rapid and adequate oxygen and nutrient delivery by blood vessels improves islet engraftment and function. The present report evaluated a potentially beneficial effect of adult human blood outgrowth endothelial cells (BOEC) on islet graft vascularisation and function. METHODS: Human BOEC, 5 × 10(5), were co-transplanted with a rat marginal-islet graft under the kidney capsule of hyperglycaemic NOD severe combined immunodeficiency (SCID) mice, and the effect on metabolic outcome was evaluated. RESULTS: Although vessel density remained unaffected, co-transplantation of islets with BOEC resulted in a significant and specific improvement of glycaemia and increased plasma C-peptide. Moreover, in contrast to control mice, BOEC recipients displayed reduced beta cell death and increases in body weight, beta cell proliferation and graft-vessel and beta cell volume. In vivo cell tracing demonstrated that BOEC remain at the site of transplantation and do not expand. The potential clinical applicability was underscored by the observed metabolic benefit of co-transplanting islets with BOEC derived from a type 1 diabetes patient. CONCLUSIONS/INTERPRETATION: The present data support the use of autologous BOEC in translational studies that aim to improve current islet-transplantation protocols for the treatment of brittle type 1 diabetes.

Differentiating neural crest stem cells induce proliferation of cultured rodent islet beta cells.

AIMS/HYPOTHESIS: Efficient stimulation of cycling activity in cultured beta cells would allow the design of new strategies for cell therapy in diabetes. Neural crest stem cells (NCSCs) play a role in beta cell development and maturation and increase the beta cell number in co-transplants. The mechanism behind NCSC-induced beta cell proliferation and the functional capacity of the new beta cells is not known. METHODS: We developed a new in vitro co-culture system that enables the dissection of the elements that control the cellular interactions that lead to NCSC-dependent increase in islet beta cells. RESULTS: Mouse NCSCs were cultured in vitro, first in medium that stimulated their proliferation, then under conditions that supported their differentiation. When mouse islet cells were cultured together with the NCSCs, more than 35% of the beta cells showed cycle activity. This labelling index is more than tenfold higher than control islets cultured without NCSCs. Beta cells that proliferated under these culture conditions were fully glucose responsive in terms of insulin secretion. NCSCs also induced beta cell proliferation in islets isolated from 1-year-old mice, but not in dissociated islet cells isolated from human donor pancreas tissue. To stimulate beta cell proliferation, NCSCs need to be in intimate contact with the beta cells. CONCLUSIONS/INTERPRETATION: Culture of islet cells in contact with NCSCs induces highly efficient beta cell proliferation. The reported culture system is an excellent platform for further dissection of the minimal set of factors needed to drive this process and explore its potential for translation to diabetes therapy.

Pancreatic islet and progenitor cell surface markers with cell sorting potential.

AIMS/HYPOTHESIS: The aim of the study was to identify surface bio-markers and corresponding antibody tools that can be used for the imaging and immunoisolation of the pancreatic beta cell and its progenitors. This may prove essential to obtain therapeutic grade human beta cells via stem cell differentiation. METHODS: Using bioinformatics-driven data mining, we generated a gene list encoding putative plasma membrane proteins specifically expressed at distinct stages of the developing pancreas and islet beta cells. In situ hybridisation and immunohistochemistry were used to further prioritise and identify candidates. RESULTS: In the developing pancreas seizure related 6 homologue like (SEZ6L2), low density lipoprotein receptor-related protein 11 (LRP11), dispatched homologue 2 (Drosophila) (DISP2) and solute carrier family 30 (zinc transporter), member 8 (SLC30A8) were found to be expressed in early islet cells, whereas discoidin domain receptor tyrosine kinase 1 (DDR1) and delta/notch-like EGF repeat containing (DNER) were expressed in early pancreatic progenitors. The expression pattern of DDR1 overlaps with the early pancreatic and duodenal homeobox 1 (PDX1)⁺/NK6 homeobox 1 (NKX6-1)⁺ multipotent progenitor cells from embryonic day 11, whereas DNER expression in part overlaps with neurogenin 3 (NEUROG3)⁺ cells. In the adult pancreas SEZ6L2, LRP11, DISP2 and SLC30A8, but also FXYD domain containing ion transport regulator 2 (FXYD2), tetraspanin 7 (TSPAN7) and transmembrane protein 27 (TMEM27), retain an islet-specific expression, whereas DDR1 is undetectable. In contrast, DNER is expressed at low levels in peripheral mouse and human islet cells. Re-expression of DDR1 and upregulation of DNER is observed in duct-ligated pancreas. Antibodies to DNER and DISP2 have been successfully used in cell sorting. CONCLUSIONS/INTERPRETATION: Extracellular epitopes of SEZ6L2, LRP11, DISP2, DDR1 and DNER have been identified as useful tags by applying specific antibodies to visualise pancreatic cell types at specific stages of development. Furthermore, antibodies recognising DISP2 and DNER are suitable for FACS-mediated cell purification.

Ngn3 expression during postnatal in vitro beta cell neogenesis induced by the JAK/STAT pathway.

The basic helix-loop-helix protein Neurogenin3 specifies precursor cells of the endocrine pancreas during embryonic development, and is thought to be absent postnatally. We have studied Ngn3 expression during in vitro generation of beta-cells from adult rat exocrine pancreas tissue treated with epidermal growth factor and leukaemia inhibitory factor. This treatment induced a transient expression of both Ngn3 and its upstream activator hepatocyte nuclear factor 6. Inhibition of EGF and LIF signalling by pharmacological antagonists of the JAK2/STAT3 pathway, or knockdown of Ngn3 by RNA interference prevented the generation of new insulin-positive cells. This study demonstrates that in vitro growth factor stimulation can induce recapitulation of an embryonic endocrine differentiation pathway in adult dedifferentiated exocrine cells. This could prove to be important for understanding the mechanism of beta-cell regeneration and for therapeutic ex vivo neogenesis of beta cells.

Human and rat beta cells differ in glucose transporter but not in glucokinase gene expression.

Glucose homeostasis is controlled by a glucose sensor in pancreatic beta-cells. Studies on rodent beta-cells have suggested a role for GLUT2 and glucokinase in this control function and in mechanisms leading to diabetes. Little direct evidence exists so far to implicate these two proteins in glucose recognition by human beta-cells. The present in vitro study investigates the role of glucose transport and phosphorylation in beta-cell preparations from nondiabetic human pancreata. Human beta-cells differ from rodent beta-cells in glucose transporter gene expression (predominantly GLUT1 instead of GLUT2), explaining their low Km (3 mmol/liter) and low VMAX (3 mmol/min per liter) for 3-O-methyl glucose transport. The 100-fold lower GLUT2 abundance in human versus rat beta-cells is associated with a 10-fold slower uptake of alloxan, explaining their resistance to this rodent diabetogenic agent. Human and rat beta-cells exhibit comparable glucokinase expression with similar flux-generating influence on total glucose utilization. These data underline the importance of glucokinase but not of GLUT2 in the glucose sensor of human beta-cells.

Methyl succinate antagonises biguanide-induced AMPK-activation and death of pancreatic beta-cells through restoration of mitochondrial electron transfer.

BACKGROUND AND PURPOSE: Two mechanisms have been proposed to explain the insulin-sensitising properties of metformin in peripheral tissues: (a) inhibition of electron transport chain complex I, and (b) activation of the AMP activated protein kinase (AMPK). However the relationship between these mechanisms and their contribution to beta-cell death and dysfunction in vitro, are currently unclear. EXPERIMENTAL APPROACH: The effects of biguanides (metformin and phenformin) were tested on MIN6 beta-cells and primary FACS-purified rat beta-cells. Cell metabolism was assessed biochemically and by FACS analysis, and correlated with AMPK phosphorylation state and cell viability, with or without fuel substrates. KEY RESULTS: In MIN6 cells, metformin reduced mitochondrial complex I activity by up to 44% and a 25% net reduction in mitochondrial reducing potential. In rat beta-cells, metformin caused NAD(P)H accumulation above maximal glucose-inducible levels, mimicking the effect of rotenone. Drug exposure caused phosphorylation of AMPK on Thr(172) in MIN6 cell extracts, indicative of kinase activation. Methyl succinate, a complex II substrate, appeared to bypass metformin blockade of complex I. This resulted in reduced phosphorylation of AMPK, establishing a link between biguanide-induced mitochondrial inhibition and AMPK activation. Corresponding assessment of cell death indicated that methyl succinate decreased biguanide toxicity to beta-cells in vitro. CONCLUSIONS AND IMPLICATIONS: AMPK activation can partly be attributed to metformin's inhibitory action on mitochondrial complex I. Anaplerotic fuel metabolism via complex II rescued beta-cells from metformin-associated toxicity. We propose that utilisation of anaplerotic nutrients may reconcile in vitro and in vivo effects of metformin on the pancreatic beta-cell.

STAT3 modulates ß-cell cycling in injured mouse pancreas and protects against DNA damage.

Partial pancreatic duct ligation (PDL) of mouse pancreas induces a doubling of the ß-cell mass mainly through proliferation of pre-existing and newly formed ß-cells. The molecular mechanism governing this process is still largely unknown. Given the inflammatory nature of PDL and inflammation-induced signaling via the signal transducer and activator of transcription 3 (STAT3), the activation and the role of STAT3 in PDL-induced ß-cell proliferation were investigated. Duct ligation stimulates the expression of several cytokines that can act as ligands inducing STAT3 signaling and phosphorylation in ß-cells. ß-Cell cycling increased by conditional ß-cell-specific Stat3 knockout and decreased by STAT3 activation through administration of interleukin-6. In addition, the level of DNA damage in ß-cells of PDL pancreas increased after deletion of Stat3. These data indicate a role for STAT3 in maintaining a steady state in the ß-cell, by modulating its cell cycle and protection from DNA damage.

Glucose sensing in pancreatic beta-cells: a model for the study of other glucose-regulated cells in gut, pancreas, and hypothalamus.

Nutrient homeostasis is known to be regulated by pancreatic islet tissue. The function of islet beta-cells is controlled by a glucose sensor that operates at physiological glucose concentrations and acts in synergy with signals that integrate messages originating from hypothalamic neurons and endocrine cells in gut and pancreas. Evidence exists that the extrapancreatic cells producing and secreting these (neuro)endocrine signals also exhibit a glucose sensor and an ability to integrate nutrient and (neuro)hormonal messages. Similarities in these cellular and molecular pathways provide a basis for a network of coordinated functions between distant cell groups, which is necessary for an appropriate control of nutrient homeostasis. The glucose sensor seems to be a fundamental component of these control mechanisms. Its molecular characterization is most advanced in pancreatic beta-cells, with important roles for glucokinase and mitochondrial oxidative fluxes in the regulation of ATP-sensitive K+ channels. Other glucose-sensitive cells in the endocrine pancreas, hypothalamus, and gut were found to share some of these molecular characteristics. We propose that similar metabolic signaling pathways influence the function of pancreatic alpha-cells, hypothalamic neurons, and gastrointestinal endocrine and neural cells.

DRB1*0403 protects against IDDM in Caucasians with the high-risk heterozygous DQA1*0301-DQB1*0302/DQA1*0501-DQB1*0201 genotype. Belgian Diabetes Registry.

The human leukocyte antigen (HLA) class II genotype DQA1*0301-DQB1*0302/DQA1*0501-DQB1*0201 has been identified as a marker strongly predisposing to insulin-dependent diabetes mellitus (IDDM) in Caucasian populations. Its frequency in control populations (1-3%) is still, however, 1 order of magnitude higher than the prevalence of IDDM, suggesting that its penetrance can be modified by protective factors. In this study we searched for such a factor in the DRB1 locus by studying DRB1*04 polymorphism in 174 European Caucasian IDDM patients and 73 nondiabetic control subjects, all sharing the HLA-DR3/DR4 phenotype. Significant protection was encoded by the DRB1*0403 allele, which was observed in 5 of 49 control subjects (10%) and none of 171 IDDM patients (0%) with the DQA1*0301-DQB1*0302/DQA1*0501-DQB1*0201 genotype (RR = 0.02 [0.01-0.18], P < 0.0005). These data support the concept that protective HLA class II genes can overrule the risk caused by HLA-DQ susceptibility dimers. They also contribute to a possible strategy to screen for nondiabetic individuals with increased genetic risk of developing IDDM.

Adult human pancreatic duct and islet cells exhibit similarities in expression and differences in phosphorylation and complex formation of the homeodomain protein Ipf-1.

The homeodomain transcription factor encoded by the pancreatic and duodenal homeobox gene-1 (Ipf-1) is essential for pancreatic ontogenesis. Whether Ipf-1 is also involved in the neogenesis of beta-cells in the adult pancreas is unknown. We examined whether Ipf-1 is expressed in adult human pancreatic ducts, which are thought to generate new beta-cells. In tissue sections, virtually all duct cells were immunopositive for Ipf-1, as were the islet beta-cells but not the acinar cells. After isolation and culture, both duct and islet cell preparations contained the Ipf-1 immunoreactive proteins p42 and p45 (42 and 45 kDa, respectively) in similar proportions, but the expression levels were twofold lower in duct cells. After 4 h of labeling, the endocrine cells exhibited a sevenfold higher phosphorylation of p42 than the duct cells, whereas p45 was phosphorylated only in endocrine cells. Homeobox binding transcription factor complexes with Ipf-1 in duct cells differed from those in endocrine cells in terms of gel mobility, sequence specificity, and affinity. The observed similarities in Ipf-1 expression by adult human pancreatic duct cells and endocrine cells may reflect their common ontogenic origin, whereas the differences in Ipf-1 phosphorylation and complex formation may correlate with their divergent differentiation.

Inhibition of cytokine-induced NF-kappaB activation by adenovirus-mediated expression of a NF-kappaB super-repressor prevents beta-cell apoptosis.

Cytokine-induced beta-cell death is an important event in the pathogenesis of type 1 diabetes. The transcription factor nuclear factor-kappaB (NF-kappaB) is activated by interleukin-1beta (IL-1beta), and its activity promotes the expression of several beta-cell genes, including pro- and anti-apoptotic genes. To elucidate the role of cytokine (IL-1beta + gamma-interferon [IFN-gamma])-induced expression of NF-kappaB in beta-cell apoptosis, rat beta-cells were infected with the recombinant adenovirus AdIkappaB((SA)2), which contained a nondegradable mutant form of inhibitory kappaB (IkappaB((SA)2), with S32A and S36A) that locks NF-kappaB in a cytosolic protein complex, preventing its nuclear action. Expression of IkappaB((SA)2) inhibited cytokine-stimulated nuclear translocation and DNA-binding of NF-kappaB. Cytokine-induced gene expression of several NF-kappaB targets, namely inducible nitric oxide synthase, Fas, and manganese superoxide dismutase, was prevented by AdIkappaB((SA)2), as established by reverse transcriptase-polymerase chain reaction, protein blot, and measurement of nitrite in the medium. Finally, beta-cell survival after IL-1beta + IFN-gamma treatment was significantly improved by IkappaB((SA)2) expression, mostly through inhibition of the apoptotic pathway. Based on these findings, we conclude that NF-kappaB activation, under in vitro conditions, has primarily a pro-apoptotic function in beta-cells.

5' insulin gene polymorphism confers risk to IDDM independently of HLA class II susceptibility.

The polymorphic variable number of tandem repeats in the 5' upstream region of the human insulin gene is a well-known non-human leukocyte antigen locus contributing to genetic susceptibility to IDDM. Controversy exists about the question as to whether INS susceptibility haplotypes are or are not preferentially inherited together with HLA-DR4 haplotypes. We investigated whether genetic interaction between INS and the HLA complex can be better defined using DQ genotypic and phenotypic markers in addition to DR serology. The 5' INS 1/1 genotype was positively associated with IDDM both in non-DR4 subjects (relative risk = 4.3; 95% confidence interval, 1.6-11.5) and DR4 subjects (relative risk = 4.2; 95% confidence interval, 1.9-9.0). Further subdivision of IDDM patients and matched control subjects according to HLA-DQA1 and HLA-DQB1 genotype or phenotype also failed to show any association between 5' INS and HLA class II genes in diabetic patients. The 5' INS and HLA class II polymorphisms therefore provide independent risk markers, which may both contribute to the genetic screening of a high-risk population among nondiabetic individuals.

Expression and functional activity of glucagon, glucagon-like peptide I, and glucose-dependent insulinotropic peptide receptors in rat pancreatic islet cells.

Rat pancreatic alpha- and beta-cells are critically dependent on hormonal signals generating cyclic AMP (cAMP) as a synergistic messenger for nutrient-induced hormone release. Several peptides of the glucagon-secretin family have been proposed as physiological ligands for cAMP production in beta-cells, but their relative importance for islet function is still unknown. The present study shows expression at the RNA level in beta-cells of receptors for glucagon, glucose-dependent insulinotropic polypeptide (GIP), and glucagon-like peptide I(7-36) amide (GLP-I), while RNA from islet alpha-cells hybridized only with GIP receptor cDNA. Western blots confirmed that GLP-I receptors were expressed in beta-cells and not in alpha-cells. Receptor activity, measured as cellular cAMP production after exposing islet beta-cells for 15 min to a range of peptide concentrations, was already detected using 10 pmol/l GLP-I and 50 pmol/l GIP but required 1 nmol/l glucagon. EC50 values of GLP-I- and GIP-induced cAMP formation were comparable (0.2 nmol/l) and 45-fold lower than the EC50 of glucagon (9 nmol/l). Maximal stimulation of cAMP production was comparable for the three peptides. In purified alpha-cells, 1 nmol/l GLP-I failed to increase cAMP levels, while 10 pmol/l to 10 nmol/l GIP exerted similar stimulatory effects as in beta-cells. In conclusion, these data show that stimulation of glucagon, GLP-I, and GIP receptors in rat beta-cells causes cAMP production required for insulin release, while adenylate cyclase in alpha-cells is positively regulated by GIP.

Reprogramming of human pancreatic exocrine cells to ß-like cells.

Rodent acinar cells exhibit a remarkable plasticity as they can transdifferentiate to duct-, hepatocyte- and islet ß-like cells. We evaluated whether exocrine cells from adult human pancreas can similarly respond to proendocrine stimuli. Exocrine cells from adult human pancreas were transduced directly with lentiviruses expressing activated MAPK (mitogen-activated protein kinase) and STAT3 (signal transducer and activator of transcription 3) and cultured as monolayers or as 3D structures. Expression of STAT3 and MAPK in human exocrine cells activated expression of the proendocrine factor neurogenin 3 in 50% to 80% of transduced exocrine cells. However, the number of insulin-positive cells increased only in the exocrine cells grown initially in suspension before 3D culture. Lineage tracing identified human acinar cells as the source of Ngn3- and insulin-expressing cells. Long-term engraftment into immunocompromised mice increased the efficiency of reprogramming to insulin-positive cells. Our data demonstrate that exocrine cells from human pancreas can be reprogrammed to transplantable insulin-producing cells that acquire functionality. Given the large number of exocrine cells in a donor pancreas, this approach presents a novel strategy to expand cell therapy in type 1 diabetes.

Inducible VEGF expression by human embryonic stem cell-derived mesenchymal stromal cells reduces the minimal islet mass required to reverse diabetes.

Islet transplantation has been hampered by loss of function due to poor revascularization. We hypothesize that co-transplantation of islets with human embryonic stem cell-derived mesenchymal stromal cells that conditionally overexpress VEGF (hESC-MSC:VEGF) may augment islet revascularization and reduce the minimal islet mass required to reverse diabetes in mice. HESC-MSCs were transduced by recombinant lentiviruses that allowed conditional (Dox-regulated) overexpression of VEGF. HESC-MSC: VEGF were characterized by tube formation assay. After co-transplantation of hESC-MSC:VEGF with murine islets in collagen-fibrin hydrogel in the omental pouch of diabetic nude mice, we measured blood glucose, body weight, glucose tolerance and serum C-peptide. As control, islets were transplanted alone or with non-transduced hESC-MSCs. Next, we compared functional parameters of 400 islets alone versus 200 islets co-transplanted with hESC-MSC:VEGF. As control, 200 islets were transplanted alone. Metabolic function of islets transplanted with hESC-MSC:VEGF significantly improved, accompanied by superior graft revascularization, compared with control groups. Transplantation of 200 islets with hESC-MSC:VEGF showed superior function over 400 islets alone. We conclude that co-transplantation of islets with VEGF-expressing hESC-MSCs allowed for at least a 50% reduction in minimal islet mass required to reverse diabetes in mice. This approach may contribute to alleviate the need for multiple donor organs per patient.

Differential expression of rat insulin I and II messenger ribonucleic acid after prolonged exposure of islet beta-cells to elevated glucose levels.

Prolonged exposure of rat islet beta-cells to 10 mmol/liter glucose has been previously shown to activate more cells into a glucose-responsive state (>90%) than has exposure to 6 mmol/liter glucose (50%). The present study demonstrates that this recruitment of more activated cells results in 4- to 6-fold higher levels of proinsulin I and proinsulin II messenger RNA (mRNA). However, only the rate of proinsulin I synthesis is increased. Failure to increase the rate of proinsulin II synthesis in the glucose-activated cells results in cellular depletion of the insulin II isoform, which can be responsible for degranulation of beta-cells cultured at 10 mmol/liter glucose. Higher glucose levels (20 mmol/liter) during culture did not correct this dissociation between the stimulated insulin I formation and the nonstimulated insulin II formation. On the contrary, the rise from 10 to 20 mmol/liter glucose resulted in a 2-fold reduction in the levels of proinsulin II mRNA, but not of proinsulin I mRNA; this process further increased the ratio of insulin I over insulin II to 5-fold higher values than those in freshly isolated beta-cells. The present data suggest that an elevated insulin I over insulin II ratio in pancreatic tissue is a marker for a prolonged exposure to elevated glucose levels. The increased ratio in this condition results from a transcriptional and/or a posttranscriptional failure in elevating insulin II formation while insulin I production is stimulated in the glucose-activated beta-cells.

Escherichia coli and Saccharomyces cerevisiae acetylornithine aminotransferase: evolutionary relationship with ornithine aminotransferase.

Genes argD and ARG8, encoding the acetylornithine aminotransferase (ACOAT) subunit in Escherichia coli and Saccharomyces cerevisiae, respectively, have been cloned and sequenced. The deduced amino acid sequences show substantial similarity. Moreover, they resemble ornithine aminotransferase (OAT) sequences (i.e., those from yeast, rat and man); the observed similarities are statistically significant, indicating that the enzymes are homologous. However, in contrast to OATs, which appear to be substrate (i.e., ornithine)-specific, S. cerevisiae ACOAT transaminates ornithine about as efficiently as E. coli does. The evolutionary relationship between ACOATs and OATs is discussed in terms of substrate ambiguity.

AMP-activated protein kinase can induce apoptosis of insulin-producing MIN6 cells through stimulation of c-Jun-N-terminal kinase.

We have recently shown that conditions known to activate AMP-activated protein kinase (AMPK) in primary beta-cells can trigger their apoptosis. The present study demonstrates that this is also the case in the MIN6 beta-cell line, which was used to investigate the underlying mechanism. Sustained activation of AMPK was induced by culture with the adenosine analogue AICA-riboside or at low glucose concentrations. Both conditions induced a sequential activation of AMPK, c-Jun-N-terminal kinase (JNK) and caspase-3. The effects of AMPK on JNK activation and apoptosis were demonstrated by adenoviral expression of constitutively active AMPK, a condition which reproduced the earlier-described AMPK-dependent effects on pyruvate kinase and acetyl-coA-carboxylase. The effects of JNK activation on apoptosis were demonstrated by the observations that (i). its inhibition by dicumarol prevented caspase-3 activation and apoptosis, (ii). adenoviral expression of the JNK-interacting scaffold protein JIP-1/IB-1 increased AICA-riboside-induced JNK activation and apoptosis. In primary beta-cells, AMPK activation was also found to activate JNK, involving primarily the JNK 2 (p54) isoform. It is concluded that prolonged stimulation of AMPK can induce apoptosis of insulin-producing cells through an activation pathway that involves JNK, and subsequently, caspase-3.