This Supplement includes articles based on presentations at the Eighth Servier - IGIS Symposium entitled ‘Animal Models of Islet Dysfunction’. These yearly Symposia are organized by the authors of this editorial, and held in an attractive venue in St-Jean Cap Ferrat in France. Their aim was to review the state-of-the-art in topics related to the biology of the islet, with emphasis on the pathogenesis and the treatment of type 2 diabetes. The objective of the Eighth Symposium was to review what we have learnt from animal models of islet dysfunction. This was a challenging task because models of spontaneous diabetes have to a major extent advanced our knowledge of the genetics and pathophysiology of the disease. Moreover, specific gene manipulations in animals have contributed to our understanding of the molecular mechanisms underlying the development of glucose intolerance. But, things are never simple in biology, and attention has recently been drawn to the multiple pitfalls that await the naïve researcher ready to manipulate genes in an animal. It is with the intention to warn that the two first papers of this volume are presented. In the first, it is reported that conditional gene targeting using the Cre/lox P strategy, while very useful for studies of glucose homeostasis and tissue function or dysfunction in diabetes and pancreas development, necessitates a variety of experimental caveats, most often in relation to the procedures used to generate Cre-driver lines. It is discussed that using bacterial artificial chromosome-derived transgenes or performing a Cre knock-in may be advantageous for improving the specificity of expression, as are systems for regulating Cre activity. The need for careful characterization of genetically manipulated models is illustrated in the second paper by the unexpected alterations in the NOD/Lt mouse model of type 1 diabetes following forced β-cell expression of non-mammalian genes ligated to an insulin promoter sequence. Part 1 considers the role of nuclear and other regulatory factors in type 2 diabetes. In the first paper, the role of sirtuins (SIRT) in the regulation of insulin secretion and β-cell survival is summarized. SIRT1 is an NAD-dependent protein deacetylase that deacetylates important transcription factors and co-factors [p53, forkhead transcription factor (FoxO1), nuclear factor-κB (NF-κB), peroxisome proliferator-activated receptor gamma co-activator 1 alpha (PGC-1α), etc.] regulating adipogenesis, myogenesis, insulin secretion and gluconeogenesis. Importantly, pharmacological activation of SIRT mimics the effects of caloric restriction and promotes longevity. SIRT1 acts as a positive regulator of insulin secretion in β cells, enhancing glucose-induced insulin release. The underlying mechanisms are complex. First, SIRT1 reduces the expression of the mitochondrial uncoupling protein, UCP2. Second, SIRT1 deacetylates FoxO1 and thereby represses its activation of pro-apoptotic target genes. In the context of nuclear factors, the pathogenesis of maturity-onset diabetes of the young (MODY), the monogenic form of type 2 diabetes mellitus, is of special interest. The disease is characterized by impairment of glucose-stimulated insulin secretion. MODY develops because of mutations in six different genes, among which mutation of the hepatic nuclear factor (HNF-1α) gene is the most common (MODY 3). Recently it was demonstrated that collectrin, a kidney-specific gene of unknown function, is a novel target of HNF-1α in pancreatic β cells. As collectrin promotes the formation of soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complexes and controls insulin exocytosis, alterations in its regulation may be associated with impaired insulin secretion, and collectrin may thus be a target for the treatment of type 2 diabetes. In addition to MODY, diabetes diagnosed during the first 6 months of life is likely to have a genetic aetiology. Permanent neonatal diabetes mellitus (PNDM) develops most often due to heterozygous activating mutations in KCNJ11, encoding the KIR 6.2 subunit of the ATP-sensitive potassium channel. Most of these patients respond to treatment with sulphonylureas. PNDM in infants diagnosed under 6 months of age also develops because of heterozygous activating mutations in the ABCC8 gene that encodes SUR1, the regulatory subunit of the KATP channel complex. Specifically, the functional effects of two missense mutations in the Kir 6.2 subunit of the KATP channel, K170T and E322K, were investigated. They both cause PNDM. The K170T mutation affects ATP-binding/transduction, whereas the E322K mutation influences the interaction with SUR1. Considering these important breakthroughs in the understanding of diabetes pathogenesis, an important issue is the identification of promising novel research directions. An interesting hypothesis is that an imbalance between proinsulin biosynthesis and insulin secretion may lead to progressive β-cell death. When a disconnection between insulin production and secretion develops, such as in type 2 diabetes, autophagy increases to maintain stable β-granule numbers. When increased autophagy becomes chronic, autophagy-mediated cell death occurs. Our understanding of the complexity of gene regulation has markedly advanced by the discovery of miRNAs. These are short, ∼22 nucleotide-long non-coding RNAs, which are believed to participate in the regulation of evolutionarily conserved regulatory pathways. Recently, it has been shown that miRNAs can play a role in the regulation of insulin secretion and glucose homeostasis. They are involved in the control of energy metabolism, pancreatic β-cell development and function, muscle cell proliferation and differentiation, and cholesterol and fat metabolism. Methods are presently being developed, which allow the testing of miRNAs as viable therapeutic targets. In Part 2, the modifications of β-cell glucose sensing and β-cell apoptosis as causes of diabetes are considered. Insulin secretion is dependent on mitochondrial metabolism in β cells. Mitochondrial diabetes occurs when a threshold level of mitochondrial DNA (mtDNA) mutation is reached. Maternally inherited diabetes and deafness (MIDD) is clinically manifested when 10–30% of the mtDNA genome, including the region coding for leucine tRNA, develops point mutations. The disease develops mainly because of increased β-cell apoptosis. Similarly, transgenic mouse models in which β-cell mtDNA mutations accumulate, develop diabetes. The lack of an inhibitory effect on glucose-stimulated insulin release suggests that the production of reactive oxygen species or defective oxidative metabolism are not the mechanisms by which diabetes develops in these mice. Hence, it is important to elucidate how relatively low levels of random mtDNA mutations provoke β-cell apoptosis. Glucose metabolism generates a rise in the [ATP] : [ADP] ratio, which promotes inhibition of KATP channels, resulting in membrane depolarization, activation of voltage-dependent Ca2+ channels, Ca2+ entry and insulin secretion. It is suggested that in rodent and human β cells, there is an inverse U-shaped response to hyperexcitability. Hence, any mechanism provoking hyperexcitability of islets, for example, decreased KATP channel density or activity, causes an initial hypersecreting phenotype, whereas further increases in excitability cause progression to an undersecretory diabetic phenotype. Such a pattern resembles the progression in type 2 diabetes – from initially compensatory insulin hypersecretion to β-cell failure. In this context, it is useful to remember that β cells are equipped with many other ion channels. Among these, the Kv channels are discussed here. These are responsible for the falling phase of action potentials, and their regulation affects insulin secretion. Another intriguing aspect of the regulation of insulin secretion is the functional heterogeneity of insulin granules in the β cell. Only granules that have been ‘primed’ are released in the exocytotic process. The protein Munc13 plays an important role in this priming process, which renders the vesicles fusion competent with the plasma membrane of the β cell. Munc13 interacts with syntaxin and other factors such as ATP, N-ethylmaleimide-sensitive factor (NSF), Cl−, etc. to regulate insulin granule priming. The precise mechanisms of the interactions between Munc13 with these factors and the SNARE complex remain to be clarified. Among the many other factors, which are involved in the docking and eventually fusing of granules with the plasma membrane is granuphilin that mediates the docking by linking Rab27a on the granule membrane and syntaxin-1a on the plasma membrane. Nevertheless, evidence is presented here suggesting that both undocked granules and newly docked granules are releasable by glucose stimulation, although fusion from previously docked granules apparently occurs more rapidly in the stimulation. A better understanding of the priming process may lead to novel approaches to correct the deficient insulin secretion of the type 2 diabetic patient. As insulin secretion is a multicellular process, a signalling system is present, which coordinates the activity of β cells within an islet as well as activity between islets. Individual β cells sense the state of activity of their neighbours and regulate their own activity accordingly. The integrating mechanisms include both indirect networks (e.g. neurotransmitters, hormones and ions) and direct cell-to-cell communication. The latter process takes place via gap junctions; these are small domains of the cell membrane where adjacent cells face each other across the extracellular space, which is reduced to a narrow ‘gap’ of only 2–3 nm wide. Recent work has demonstrated that connexin 36 (Cx36) is a component protein of gap junctions in pancreatic β cells. In addition, there are studies suggesting that connexin-dependent signalling is required for the fine regulation of insulin biosynthesis, storage and release under basal conditions and after glucose stimulation. Loss of Cx36 resulted in pancreatic dysfunctions that are similar to those observed in diabetes. Thus, impaired Cx36-dependent signalling may participate in the pathophysiology of diabetes. Part 3 discusses β-cell adaptation and maladaptation to hyperfunction. An increased influx of fatty acids into β cells in obesity and overnutrition provokes β-cell failure by virtue of so-called lipotoxicity. Investigations of the mechanisms underlying this finding have demonstrated that long-term exposure to saturated fatty acids such as palmitic acid (PA) decreases insulin secretion through sterol regulatory element-binding protein (SREBP)-1c activation. In contrast, polyunsaturated fatty acids, including eicosapentanoic acid (EPA), antagonize PA-suppressed insulin release by suppressing the expression of SREBP-1c. Adaptation to increased function in, for example, obesity requires augmentation in β-cell mass. Overall β-cell mass is controlled by a dynamic balance between neogenesis, proliferation, cell size and apoptosis. It is unclear whether compensatory β-cell hyperplasia caused by insulin resistance is achieved by proliferation of existing β cells or by neogenesis from progenitor cells embedded in duct epithelia. Furthermore, the mechanism/s regulating this adaptation are unknown. Recent studies suggest that β-cell compensation for insulin resistance is a proliferative response of existing β cells to growth factor signalling, which requires FoxO1 participation. Hence, FoxO1 plays an important role in β-cell proliferation. In most cell types, FoxO1 shuttles between the nucleus and cytoplasm, but in β cells, FoxO1 is constitutively localized to the cytoplasm because of continuous stimulation by endogenously produced insulin. However, hyperglycaemia and oxidative stress lead to nuclear redistribution of FoxO1 in β cells. This is associated with increased expression of the transcription factors, NeuroD and MafA. FoxO1 nuclear translocation is probably part of a protective response against β-cell dysfunction induced by hyperglycaemia. During oxidative stress, FoxO1 activity is regulated by a balance between acetylation and deacetylation to prevent excessive FoxO1-dependent transcription, which would otherwise lead to apoptosis and cellular atrophy. Importantly, constitutive nuclear expression of FoxO1 prevents the proliferative and antiapoptotic actions of glucagon-like peptide-1 in cultured β cells. RNA expression profiling of cultured β cells with FoxO1 gain-of-function reveals repression of genes involved in glycolysis, nitric oxide synthesis, ion transport and G-protein-receptor signalling. Thus, FoxO1 seems to provoke a state of metabolic diapause in which growth, development and physiological activities are blocked to protect β cells against oxidative and environmental stresses. With regard to the regulation of β-cell mass by growth factors, insulin, incretins and nutrients, it seems that phosphoinositide 3 kinase/Akt signalling constitutes a converging pathway. Overexpression of a constitutively active form of Akt1 in β cells provokes an increase in both the size and the number of cells. Akt has also been implicated in β-cell regeneration. Akt kinase could regulate β-cell mass by activation of the cyclin-dependent kinase 4 complex. As a serine/threonine kinase, Akt1 is an interesting target for designing drugs, which induce proliferation and survival of β cells, provided that oncogenic effects could be avoided. Also, the neural network in the endocrine pancreas participates in the coordination of secretory functions and growth of islet cells. Among neurotransmitters, an important role is played by acetylcholine (ACh), which is released from intrapancreatic parasympathetic nerve endings during the preabsorptive and absorptive phases of feeding. ACh markedly enhances glucose-induced insulin release. This effect is exclusively mediated by the M3 subtype muscarinic receptor. Mutant mice that lack M3 receptors only in β cells present decreased insulin release and glucose intolerance. In contrast, mice overexpressing M3 receptors in β cells demonstrate enhanced insulin release and improved glucose tolerance. In patients with type 2 diabetes, the insulin response to glucose or a mixed meal is decreased. This defect is accounted for by both impaired stimulus-secretion coupling in β cells and a decreased β-cell mass. However, reduced β-cell mass should be considered as an imbalance between formation and destruction rates, rather than as a static condition. Presently, the determination of β-cell mass in humans is not possible in vivo. However, studies in primates and minipigs have revealed correlations between in vivo functional tests and actual β-cell mass. It is suggested that β-cell mass may also be predicted in humans by studying insulin secretion induced by intravenous glucose and/or arginine. Part 4 opens with two reviews of a spontaneous non-obese model of type 2 diabetes, the Goto Kakisaki (GK) rat. The GK rat strain was developed by selective breeding of Wistar rats with the highest normal blood glucose levels in an oral glucose tolerance test. Glucose intolerance and defective glucose-induced insulin secretion are constant features of the GK rat. In the Stockholm colony of GK rats, β-cell density and relative volume of islet endocrine cells were comparable in 2 to 3-month-old GK and control Wistar rats. Similar results were reported in a study in Dallas with GK rats obtained from Sendai. Accordingly, in the Stockholm colony, the impaired insulin response was accounted for by defective stimulus-secretion coupling because of both primary genetic defects and glucotoxicity. In this context, recent studies have demonstrated that the expression of exocytotic SNARE proteins is significantly reduced in islets of GK rats. In contrast, GK rats from the Paris colony show poor proliferation and/or survival of islet endocrine precursor cells during foetal life, with inhibited islet neogenesis. As a result, total pancreatic β-cell mass and insulin stores are decreased by 60% in adult animals. The primary defect is thought to be impaired Igf2/Igf1-R pathway in the pancreatic rudiment during early foetal life. It is thus interesting that even in a single animal model for type 2 diabetes, β-cell biology may show such discordant pictures. The role of epigenetic factors, probably modulated by the subtle differences to be found between animal facilities, clearly needs to be investigated in this model, as probably also in other models. In this context, the known relationship between poor foetal growth and subsequent development of features of the metabolic syndrome is an important chapter in diabetes pathophysiology. Animal models have been established to clarify the mechanisms behind these observations, which stem from epidemiological studies. It has been demonstrated that reduction in nutrients during foetal development programmes the endocrine pancreas as well as insulin-sensitive peripheral tissues. An important finding that whether calories or proteins are restricted, malnourished pups are born with defective β cells and insulin resistance, these defects persist throughout life. Despite the similar endpoint, different cellular and physiological mechanisms are proposed. Hormones operative during foetal life like insulin, insulin-like growth factors (IGFs) and glucocorticoids, as well as specific molecules like taurine, or islet vascularization are implicated as possible factors amplifying the defect. The molecular mechanisms responsible for intrauterine programming of the β cells are not known, although programming of mitochondria and epigenetic regulation have been proposed. Much further work is needed in this novel area, which may be of major importance for the prevention of at least some proportion of the present world diabetes epidemics. In summary, this volume offers the readership of Diabetes, Obesity and Metabolism a ‘tasting tour’ in the vast and rich territory, which is the world of experimental models of type 2 diabetes. No two animal models yield identical information, and no single model will faithfully represent the human disease with all its variety. It is the painstaking combination of selected aspects of each model that will advance our knowledge of normal and abnormal β-cell function in man. Needless to say, it is our conviction that novel, improved modes of treatment for type 2 diabetes cannot emerge without such studies. First Servier – IGIS Symposium: Birth, Life, and Death of a β-Cell in Type 2 Diabetes. Diabetes 2001; 50 (Suppl. 1). Second Servier – IGIS Symposium: Kinetics of Insulin Release in Health and Type 2 Diabetes. Diabetes 2002; 51 (Suppl. 1). Third Servier – IGIS Symposium: Regulation of Insulin Production: in Search of Therapeutic Targets. Diabetes 2002; 51 (Suppl. 3). Fourth Servier – IGIS Symposium: Novel Factors in the Regulation of β-Cell Function. Diabetes 2004; 53 (Suppl. 1). Fifth Servier – IGIS Symposium: Impact of Treatment on Islet Function in Type 2 Diabetes: A Critical Appraisal. Diabetes 2004; 53 (Suppl. 3). Sixth Servier – IGIS Symposium: Type 1 and Type 2 Diabetes: Less Apart than Apparent? Diabetes 2005; 54 (Suppl. 2). Seventh Servier – IGIS Symposium: The Islet – Brain – Peripheral Tissue Network and Type 2 Diabetes. Diabetes 2006; 55 (Suppl. 2). The outstanding logistic support offered by the team of Laurence Alliot, Servier, France makes the Servier-IGIS Symposia an unforgettable experience. Our sincere thanks are due to the Secretary of the IGIS Board, Prof. Alain Ktorza, for his support of its scientific activities and to Catriona Donagh for her superb assistance in the preparation of this volume.