What ails the β ‐cell?

The 11th Servier-IGIS meeting was held in St. Jean Cap Ferrat from 25 to 27 March 2010. The year's theme was ‘The stressed β-cell’, itself a testimonial to the changing cell biological, biochemical and clinical landscape of diabetes in the 21st century. Frontline investigators presented novel and creative mechanisms of cellular stress as the basis for β-cell failure in type 2 diabetes, and many new ideas were debated to account for the protean manifestations of β-cell dysfunction. Nonetheless, it was realized that significant hurdles remain in matching the ingenuity of the ideas to the prosaic reality of the slow and progressive degenerative process that affects the endocrine pancreas in diabetic patients. The pathogenesis of β-cell dysfunction in type 2 diabetes poses an exceptional intellectual challenge 1. Whether by dint of genetic or of acquired abnormalities, the β-cells of diabetic patients respond poorly to glucose challenge: laggards and braggarts at once, they fail to mount an appropriately timed response but, when they finally bestir themselves, they do it with zestful vengeance, exposing the body to too much insulin at the wrong time 2. Thus, although insulin secretion at some stages in type 2 diabetes is abnormally high, it still remains insufficient to control glycemia due to the increased demand. The mechanism underlying this abnormality is not clear. In addition to cell-autonomous (intrinsic) defects that can be demonstrated long before disease onset 3, there are cell-nonautonomous contributors to β-cell dysfunction: first and foremost hyperglycemia, also lipids, other hormones, autonomic nerves and the ever elusive cytokine set, not to mention changes in endothelial function and vascular flow, that are unlikely to relieve the β-cell's predicament 4. Nor are changes limited to the insulin secretory aspects of β-cell function. Over time, the number of β-cells undergoes changes too, although their extent is unlikely to be large, and their modality is somewhat shrouded in mystery. The relationship between β-cell mass and type 2 diabetes is also clouded by the intrinsic difficulty of assessing endocrine islet mass, by the large individual variation even among normal subjects, by the fact that human studies are by virtue of necessity limited to dead people, and by the uncertainties of quantifying immunohistochemical determinations 5. To complicate matters, clinical outcome data show that ‘therapeutic’ interventions may affect β-cell function while doing little to reverse this progressive deterioration, beyond providing stopgap relief 6. Against this backdrop, the 11th Servier-IGIS meeting brought into focus the dynamics of β-cell stress, an intentionally vague definition that reflects the investigators' state of mind as much as the β-cell's predicament. On the good news front, there's material aplenty, and of great interest. Originally identified as an alternative energy source during nutrient deprivation, autophagy has emerged as a mechanism of tissue remodelling, and appears to be required to maintain islet structure and to promote healthy β-cell function while preserving β-cell mass 7. Evidence is beginning to accumulate that autophagy protects against endoplasmic reticulum (ER) stress and promotes turnover of subcellular organelles—such as mitochondria and ER—thus providing a checkpoint for their integrity and function. It remains unclear what the physiologic cues regulating this process are, and how they are affected by diabetes. Data were presented to support the notion that autophagy is regulated by free fatty acids (FFAs) and other nutrients 8. But it is unclear at what point a physiologic process of cellular turnover becomes a disease process, and whether sheer nutrient oversupply, or individual components thereof, tilt the balance from physiological organelle turnover to accelerated cell death. While the models of defective autophagy presented at the meeting do provide a convincing picture of β-cell dysfunction 9, less clear is how the same process may underlie cellular pathologies as distinct as hepatosteatosis, adipocyte hypertrophy 7 and β-cell dysfunction. More facts need to come to light before the relevance of this exciting work on type 2 diabetes can be firmly established. The complex of events known as unfolded protein response (UPR) has also attracted interest from β-cell investigators for nearly a decade. As a protein factory, the β-cell is uniquely susceptible to the untoward consequences of a translational process running amok when exposed to continuous nutrient oversupply. In cells that produce large amounts of peptide hormone, the ER might be exquisitely sensitive to variations in translational rates, or to the presence of misfolded or otherwise incompletely processed proteins. What triggers this process, and does it affect β-cell function, mass or both? And where does the boundary lie between a healthy degree of UPR and abnormal UPR? This was discussed at the meeting and evidence was presented which suggests that several nutritional cues can set the stage for UPR activation, including FFAs, phospholipase A2 (iPLA2), reactive oxygen species (ROS) and inflammation. It is intriguing that FFAs have been shown to have differential effects on distinct branches of the UPR, underlying their ability to predispose to, or to prevent, β-cell dysfunction. Activation of iPLA2 might represent a link between activation of intracellular signaling pathways by FFAs and ER stress. And ROS generation, a perennial favourite of β-cell failure aficionados, might bridge oxidative stress with ER stress 10. An inflammatory milieu can also provide a mechanism to subsume the various branches of nutrient-based signalling under the common denominator of activation of ER stress. Some of these arguments have an unmistakable chicken–and–egg quality, reflecting the complexity of understanding a chronic disease process in an—to the clinical researcher—anatomically inaccessible environment. It is envisioned that the early abnormalities brought about by UPR activation primarily affects insulin secretion. Over time, β-cell mass might also become depleted through apoptotic processes. The latter point is clearly illustrated by the different properties of two mediators of the UPR, the serine kinase Perk and the bHLH transcription factor Chop. Loss of Perk function might cause chronic β-cell stress or hamper its ability to properly respond to ER stress 11. Chronic UPR activation can also lead to activation of the proapoptotic transcription factor Chop; Chop deletion in insulin-resistant mice can restore β-cell mass and forestall diabetes progression. A similar phenotype has been reported for the Chop-interacting protein C/EBPβ, providing further evidence of the participation of this pathway in β-cell dysfunction 12. If UPR activation can have totally diverging outcomes depending on the nature of the activating process, one wonders whether it is desirable to tinker with it—especially in light of the fact that the process is not limited to β-cells, but appears to involve virtually every tissue in the body. A notable attempt to promote more efficient transit of peptides through the ER is being made with small molecules that act as chaperones to relieve the load on the ER, something akin to unclogging or widening the drain of an overflowing bathtub 13. These chaperones are chemical probes that provide proof of principle, but whether this approach has a sufficiently large therapeutic window, that is, whether the UPR can be rendered more efficient without wreaking havoc on other adaptive processes, remains unsettled at this time. The role of uncoupling proteins in regulating insulin secretion has also been explored in detail. UCP2, the main mitochondrial uncoupler in β-cells, is regulated through post-translational mechanisms by glucose, and activated in response to nutrients such as FFAs, and by ROS 14. The notion that UCP2 is primarily regulated post-translationally may help reconcile conflicting claims on the role of this important component of the cellular machinery. Although mitochondria get all the attention, it should be remembered that peroxisomal FFA oxidation also provides an important source of ROS, thus connecting disturbed lipid metabolism and β-cell dysfunction 15. The idea of tweaking mitochondrial respiration to modulate insulin secretion is not new, but these recent advances support the somewhat unconventional view that making mitochondria less efficient might help defuse nutrient-induced toxicity, and prevent glucotoxicity and lipotoxicity in β-cells 16. Many cellular biological mechanisms have been identified to play a role—central or supportive—in β-cell dysfunction. Are any of them specific to the β-cell, or to distinct stages of the disease process, and are any of them reversible? Is it even desirable to reverse them? The mechanisms reviewed here appear to affect both β-cell function (insulin secretion) and number. In trying to integrate these processes into the pathogenesis of β-cell dysfunction and to leverage this knowledge to therapeutic ends, one has to be careful what one wishes for. For while it is likely that reversing the functional deficit of insulin secretion can have beneficial effects, there is simply not enough data to conclude that saving β-cells teetering on the brink of apoptosis, or laden with misfolded peptides, is either possible or desirable. For all the efforts that have gone into demonstrating that there is increased β-cell loss in diabetes, there is still no evidence that such loss is disproportionate to cellular dysfunction, that is, that diabetic patients prematurely lose functioning or otherwise salvageable β-cells, as opposed to β-cells that have passed their expiration date. There is ample evidence in human and experimental animal studies that diabetic islets are studded with ‘empty’β-cells, which presumably reflect cells that have become functionally exhausted. Are these the cells where degenerative processes take place? Apoptosis, autophagy, UPR and mitochondrial uncoupling are probably meant from an evolutionary perspective to be protective mechanisms; for all we know, by the time β-cells get targeted for these outcomes, their functional abilities may be so severely and irreversibly compromised as to render them useless and potentially harmful as a source of cellular damage, inflammation, vascular stress or tissue reactions. Would prolonging the life of this subset of β-cells offer clinical benefits? And can these tightly regulated processes, hovering on the brink between cell life and death, be pharmacologically targeted with the precision that would be required to distinguish salvageable β-cells from unsalvageable ones? As we move into a new decade of islet research, buoyed by the successes of a decade that has witnessed the resurgence of interest in the role of the endocrine pancreas in type 2 diabetes, capped by the introduction of new β-cell-tropic agents in diabetes management, we should be mindful of some of its lessons: (i) β-cell loss is a slow process and it is exceedingly hard to quantify it; thus, therapies that target β-cell integrity and attempt to enhance islet mass will require long-term outcome studies and will be fraught with the difficulties described above; (ii) β-cells have evolved to be singularly sensitive nutrient sensors; treating the cell-autonomous abnormalities of β-cell function cannot be divorced from treating what lies upstream—the abnormal metabolic profile in the bloodstream together with abnormal hormonal and neural inputs; (iii) there remain significant gaps in our understanding of the specificity of the process of β-cell dysfunction, that is, what are the relevant roles of cell-autonomous versus cell-nonautonomous abnormalities in the process; why does the functional (secretory) abnormality beget the quantitative deficit; and is the latter ever a problem; namely, does the number of β-cells ever become the limiting factor in diabetes control? As these interesting fields move forward to address the issues raised here, the Servier-IGIS meetings will continue to provide a critical forum to showcase progress. We are indebted to Laurence Alliot and Alain Ktorza (Servier) for their leadership and support of the IGIS Symposia. We thank the indefatigable Martine Zeitoun and Catriona Donagh (Servier) for their logistical and editorial support. We are grateful to Laboratoires Servier, Paris, for their continuing commitment to the IGIS project. The authors do not declare any conflict of interest relevant to this manuscript.