Only in the last decade did modeling studies predict, and knockout experiments confirm, that type 2 diabetes is a "2-hit" disease in which insulin resistance is necessarily accompanied by beta-cell defect(s) preventing the compensatory upregulation of insulin secretion. This long- delayed insight was associated with the development of a constant, the "disposition index," describing the beta-cell sensitivity-secretion relationship as a rectangular hyperbola. Shifts in insulin sensitivity are accompanied by compensatory alterations in beta-cell sensitivity to glucose. Insulin-sensitive subjects do not require a massive insulin response to exogenous glucose to maintain a normal blood glucose. But if their insulin sensitivity decreases by 80%, as in late pregnancy, they need a fivefold greater insulin response to achieve an identical disposition index. Women with gestational diabetes have an insulin response similar to that of normal volunteers; at first glance, this suggests similar islet function, but the utility of the disposition index is to normalize this response to the amplitude of third trimester insulin resistance, revealing severe beta-cell deficiency. The index is a quantitative, convenient, and accurate tool in analyzing epidemiologic data and identifying incipient impaired glucose tolerance. Separate major issues remain, however: the causes of insulin resistance, the causes of the failure of adequate beta-cell compensation in type 2 diabetes, and the nature of the signal(s) from insulin-resistant tissues that fail to elicit the appropriate beta-cell increment in sensitivity to glucose and other stimuli. The disposition index is likely to remain the most accurate physiologic measure for testing hypotheses and solutions to these challenges in whole organisms.

A model for the relationship between ionic and metabolic oscillations and plasma insulin oscillations is presented. It is argued that the pancreatic beta-cell in vivo displays two intrinsic frequencies that are important for the regulation of plasma insulin oscillations. The rapid oscillatory activity (2--7 oscillations [osc] per minute), which is evident in both ionic and metabolic events, causes the required elevation in cytoplasmic Ca(2+) concentration ([Ca(2+)](i)) for the exocytosis of insulin granules. This activity is important for regulation of the amplitude of plasma insulin oscillations. The frequency of the rapid oscillatory ionic activities is regulated by glucose and allows the beta-cell to respond in an analogous way, with gradual changes in [Ca(2+)](i) and insulin release in response to the alterations in glucose concentration. The slower oscillatory activity (0.2--0.4 osc/min), which is evident in the metabolism of the beta-cell, has a frequency corresponding to the frequency observed in plasma insulin oscillations. The frequency is not affected by changes in the glucose concentration. This activity is suggested to generate energy in a pulsatile fashion, which sets the frequency of the plasma insulin oscillations. It is proposed that the slow oscillations in [Ca(2+)](i) observed in vitro are a manifestation of the metabolic oscillations and do not represent an in vivo phenomenon.

Whereas the mechanisms underlying oscillatory insulin secretion remain unknown, several models have been advanced to explain if they involve generation of metabolic oscillations in beta-cells. Evidence, including measurements of oxygen consumption, glucose consumption, NADH, and ATP/ADP ratio, has accumulated to support the hypothesis that energy metabolism in beta-cells can oscillate. Where simultaneous measurements have been made, these oscillations are well correlated with oscillations in intracellular [Ca(2+)] and insulin secretion. Considerable evidence has been accumulated to suggest that entry of Ca(2+) into cells can modulate metabolism both positively and negatively. The main positive effect of Ca(2+) is an increase in oxygen consumption, believed to involve activation of mitochondrial dehydrogenases. Negative feedback by Ca(2+) includes decreases in glucose consumption and decreases in the mitochondrial membrane potential. Ca(2+) also provides negative feedback by increasing consumption of ATP. The negative feedback provided by Ca(2+) provides a mechanism for generating oscillations based on a model in which glucose stimulates a rise in ATP/ADP ratio that closes ATP-sensitive K(+) (K(ATP)) channels, thus depolarizing the cell membrane and allowing Ca(2+) entry through voltage-sensitive channels. Ca(2+) entry reduces the ATP/ADP ratio and allows reopening of the K(ATP) channel.

The mechanisms driving the pulsatility of insulin secretion in vivo and in vitro are still unclear. Because glucose metabolism and changes in cytosolic free Ca(2+) ([Ca(2+)](c)) in beta-cells play a key role in the control of insulin secretion, and because oscillations of these two factors have been observed in single isolated islets and beta-cells, pulsatile insulin secretion could theoretically result from [Ca(2+)](c) or metabolism oscillations. We could not detect metabolic oscillations independent from [Ca(2+)](c) changes in beta-cells, and imposed metabolic oscillations were poorly effective in inducing oscillations of secretion when [Ca(2+)](c) was kept stable, which suggests that metabolic oscillations are not the direct regulator of the oscillations of secretion. By contrast, tight temporal and quantitative correlations between the changes in [Ca(2+)](c) and insulin release strongly suggest that [Ca(2+)](c) oscillations are the direct drivers of insulin secretion oscillations. Metabolism may play a dual role, inducing [Ca(2+)](c) oscillations (via changes in ATP-sensitive K(+) channel activity and membrane potential) and amplifying the secretory response by increasing the efficiency of Ca(2+) on exocytosis. The mechanisms underlying the oscillations of insulin secretion by the isolated pancreas and those observed in vivo remain elusive. It is not known how the functioning of distinct islets is synchronized, and the possible role of intrapancreatic ganglia in this synchronization requires confirmation. That pulsatile insulin secretion is beneficial in vivo, by preventing insulin resistance, is suggested by the greater hypoglycemic effect of exogenous insulin when it is infused in a pulsatile rather than continuous manner. The observation that type 2 diabetic patients have impaired pulsatile insulin secretion has prompted the suggestion that such dysregulation contributes to the disease and justifies the efforts toward understanding of the mechanism underlying the pulsatility of insulin secretion both in vitro and in vivo.

A strong genetic component of the beta-cell defect of type 2 diabetes is undisputed. We recently developed a modification of the classic hyperglycemic clamp to assess beta-cell function in response to various stimuli (10 mmol/l glucose, additional glucagon-like peptide [GLP]-1, and arginine). Subjects at risk for developing type 2 diabetes (impaired glucose-tolerant individuals, women with gestational diabetes, and individuals with a family history of type 2 diabetes) clearly showed a significantly decreased mean secretory response to all secretagogues compared with controls. We also showed that normal glucose-tolerant carriers of the Gly972Arg polymorphism in the insulin receptor substrate 1 have significantly reduced insulin secretion in response to glucose and arginine but not to GLP-1. More remarkably, however, the relative impairment of the different secretory phases varied greatly in the same individual, indicating a substantial heterogeneity of beta-cell dysfunction. Specific prominence of this heterogeneity may reflect a specific cellular defect of the beta-cell. In subjects sharing this pattern of heterogeneity, any underlying genetic variant may be enriched and thus more likely not only to be identified but also to be related to a pathophysiological mechanism. In conclusion, we believe that careful clinical characterization of beta-cell function (and dysfunction) is one way of identifying and understanding the genetic factors leading to the insulin secretory failure of type 2 diabetes.

Insulin is released from the pancreas in a biphasic manner in response to a square-wave increase in arterial glucose concentration. The first phase consists of a brief spike lasting approximately 10 min followed by the second phase, which reaches a plateau at 2-3 h. It is widely thought that diminution of first-phase insulin release is the earliest detectable defect of beta-cell function in individuals destined to develop type 2 diabetes and that this defect largely represents beta-cell exhaustion after years of compensation for antecedent insulin resistance. In this article, the origins of these concepts are reviewed and recent evidence is presented suggesting that reductions in both phases of insulin release are equally early, that they precede insulin resistance other than that simply due to obesity, and that they therefore may represent the primary genetic risk factor predisposing individuals to type 2 diabetes.

Type 2 diabetes is a heterogeneous disorder due to prevalent insulin resistance associated with deficient insulin secretion or to a prevalent defect of insulin secretion associated with impaired insulin action. The definition is supported by the high frequency at which insulin resistance can be demonstrated in type 2 diabetic patients. Nonetheless, insulin resistance is not a sufficient mechanism to cause diabetes. Impaired beta-cell function is a necessary defect in all conditions of impaired glucose regulation; however, it manifests itself in a different manner in fasting and glucose-stimulated conditions. In the fasting state, the basal insulin secretory rate increases as a function of the progressive decline in insulin action. As such, the fasting plasma insulin concentration is often taken as a marker for insulin sensitivity. After glucose challenge, a specific alteration of acute insulin release is an early and progressive defect. The latter might represent an intrinsic defect, but its continuous decline is affected by glucotoxicity and lipotoxicity. To understand the impact of beta-cell dysfunction in type 2 diabetes on metabolic homeostasis, it is useful to consider the different phases of insulin secretion separately. Insulin secretion can be divided into basal (postabsorptive) and stimulated (postprandial) states. The former prevails during the interprandial phases and plays a major role during the overnight fast; the latter regulates glucose metabolism when carbohydrate is abundant and must be disposed of. Data in animals and humans support a crucial physiological role of first-phase insulin secretion in postprandial glucose homeostasis. This effect is primarily achieved in the liver, allowing prompt inhibition of endogenous glucose production and limiting the postprandial rise in plasma glucose level. In type 2 diabetes, loss of the early surge of insulin release is an early and quite common defect that may have a pathogenetic role in the development of postprandial hyperglycemia, possibly requiring specific therapeutic intervention.

The dose-response relationship between the hepatic sinusoidal insulin level and glucose production by the liver is such that a half-maximally effective concentration is at or slightly below the hormone levels seen basally after an overnight fast. In the normal individual, the direct effect of the hormone on the hepatocyte is far more important in restraining glucose production than its indirect effect mediated via a suppression of lipolysis. Because insulin regulates the liver in a direct fashion, its effect occurs within several minutes. Thus, the speed with which insulin works and the sensitivity of the liver to it predict that first-phase insulin release should have a significant effect in quickly suppressing hepatic glucose production. On the other hand, nonhepatic tissues are much less sensitive to insulin and respond slowly as a result of the need for insulin to cross the endothelial barrier. As a result, first-phase insulin is unlikely to significantly alter peripheral glucose disposal. Simulation studies in humans and dogs in which the effects of first-phase insulin were simulated confirmed the aforementioned predictions. In addition, they confirmed the ability of second-phase insulin release to have significant effects on both glucose production and utilization.

Glucose induces biphasic insulin secretion by the islet beta-cell. Based on recent knowledge on glucose signaling in the beta-cell, the underlying mechanisms for this biphasicity could be envisaged as follows. Glucose-induced elevation of cytosolic free Ca(2+) concentration, which is due to the electrophysiological events that originate in closure of the ATP-sensitive K(+) (K(ATP)) channel, most likely triggers the first phase. The second phase is produced by gradual augmentation and potentiation of Ca(2+)-triggered insulin release by the K(ATP) channel-independent, nonionic signals. Protein acylation may be involved in the nonionic signaling. In patients lacking functional K(ATP) channels, however, the first phase of glucose-induced insulin secretion is clearly retained, casting doubt on the simplistic view outlined above. In this pathological condition, the K(ATP) channel-independent, most likely nonionic, glucose action alone is sufficient for the first-phase response.

The insulin secretory response by pancreatic beta-cells to an acute "square wave" stimulation by glucose is characterized by a first phase that occurs promptly after exposure to glucose, followed by a decrease to a nadir, and a prolonged second phase. The first phase of release is due to the ATP-sensitive K(+) (K(ATP)) channel-dependent (triggering) pathway that increases [Ca(2+)](i) and has been thought to discharge the granules from a "readily releasable pool." It follows that the second phase entails the preparation of granules for release, perhaps including translocation and priming for fusion competency before exocytosis. The pathways responsible for the second phase include the K(ATP) channel-dependent pathway because of the need for elevated [Ca(2+)](i) and additional signals from K(ATP) channel-independent pathways. The mechanisms underlying these additional signals are unknown. Current hypotheses include increased cytosolic long-chain acyl-CoA, the pyruvate-malate shuttle, glutamate export from mitochondria, and an increased ATP/ADP ratio. In mouse islets, the beta-cell contains some 13,000 granules, of which approximately 100 are in a "readily releasable" pool. Rates of granule release are slow, e.g., one every 3 s, even at the peak of the first phase of glucose-stimulated release. As both phases of glucose-stimulated insulin secretion can be enhanced by agents such as glucagon-like peptide 1, which increases cyclic AMP levels and protein kinase A activity, or acetylcholine, which increases diacylglycerol levels and protein kinase C activity, a single "readily releasable pool" hypothesis is an inadequate explanation for insulin secretion. Multiple pools available for rapid release or rapid conversion of granules to a rapidly releasable state are required.

Capacitance measurements were applied to mouse pancreatic beta-cells to elucidate the cellular mechanisms underlying biphasic insulin secretion. We report here that only <50 of the beta-cell's >10,000 granules are immediately available for release. The releasable granules tightly associate with the voltage-gated alpha(1C) Ca(2+) channels, and it is proposed that the release of these granules accounts for first-phase insulin secretion. Subsequent replenishment of the releasable pool by priming of previously nonreleasable granules is required for second-phase insulin secretion. The latter reaction depends on intragranular acidification due to the concerted action of granular bafilomycin-sensitive v-type H(+)-ATPase and 4,4-diisothiocyanostilbene-2,2-disulfonate--blockable ClC-3 Cl(-) channels. Lowering the cytoplasmic ATP/ADP ratio prevents granule acidification, granule priming, and refilling of the releasable pool. The latter finding provides an explanation to the transient nature of insulin secretion elicited by, for example, high extracellular K(+) in the absence of metabolizable fuels.

A full biphasic insulin response is the most sensitive index for well-coupled beta-cell signal transduction. While first-phase insulin response is extremely sensitive to potentiating and inhibiting modulations, full expression of second-phase response requires near maximally activated beta-cell fuel metabolism. In the isolated rat pancreas, accelerated calcium entry or activation of protein kinase (PK)-A or PKC result in no insulin response in the absence of fuel metabolism. At submaximal levels of beta-cell fuel secretagogue, arginine (which promotes calcium entry) or glucagon (which activates PKA) produces a small first-phase insulin response but minimal or no second-phase response; carbachol (which activates PKC and promotes calcium entry) generates biphasic insulin response in the presence of minimal fuel (3.3 mmol/l glucose). Glucagon produces full biphasic response in the presence of 10.0 mmol/l glucose, whereas arginine requires near-maximal stimulatory glucose (16.7 mmol) to produce full biphasic insulin response. Thus, PKA and PKC signal pathways potentiate primary signals generated by fuel secretagogues to induce full biphasic insulin response, while calcium recruitment alone is insufficient to potentiate primary signals generated at low levels of fuel secretagogue. We suggest that three families of PKs (calmodulin-dependent PK [CaMK], PKA, and PKC) function as distal amplifiers for stimulus-secretion coupling signals originating from fuel metabolism, as well as from incretins acting through membrane receptors, adenylate cyclase, and phospholipase C. Several isoenzymes of PKA and PKC are present in pancreatic beta-cells, but the specific function of most is still undefined. Each PK isoenzyme is activated and subsequently phosphorylates its specific effector protein by binding to a highly specific anchoring protein. Some diabetes-related beta-cell derangements may be linked to abnormal function of one or more PK isoenzymes. Identification and characterization of the specific function of the individual PK isoenzymes may provide the tool to improve the insulin response of the diabetic patient.

Rapid and sustained stimulation of beta-cells with glucose induces biphasic insulin secretion. The two phases appear to reflect a characteristic of stimulus-secretion coupling in each beta-cell rather than heterogeneity in the time-course of the response between beta-cells or islets. There is no evidence indicating that biphasic secretion can be attributed to an intrinsically biphasic metabolic signal. In contrast, the biphasic rise in cytoplasmic Ca(2+) concentration ([Ca(2+)](i)) induced by glucose is important to shape the two phases of secretion. The first phase requires a rapid and marked elevation of [Ca(2+)](i) and corresponds to the release of insulin granules from a limited pool. The magnitude of the second phase is determined by the elevation of [Ca(2+)](i), but its development requires production of another signal. This signal corresponds to the amplifying action of glucose and may serve to replenish the pool of granules that are releasable at the prevailing [Ca(2+)](i). The species characteristics of biphasic insulin secretion and its perturbations in pathological situations are discussed.

The cellular and molecular mechanisms of insulin secretion are being intensively investigated, yet most researchers are seemingly unaware of the complexity of the dynamic regulation of the secretion. In this article, we summarize studies of the physiology of insulin secretion performed over several decades. The insulin response of perifused islets of rats, perfused rat pancreas, or that of a human, to a square-wave glucose stimulus is biphasic, a transient first-phase response of 4- to 10-min duration followed by a gradual rise in secretion rates (second-phase response). Several hypotheses have been proposed to account for the phasic nature of insulin secretion; they are briefly discussed in this review. We have favored the hypothesis that nutrient stimulators such as glucose, in addition to a primary and almost immediate secretory signal, with time induce both stimulatory and inhibitory messages in the beta-cell, and those messages modulate the primary insulinogenic signal. Indeed, studies in the rat pancreas and in humans have demonstrated that short stimulations with glucose generate a state of refractoriness of the insulin secretion, which we have termed time-dependent inhibition (TDI). Nonnutrient secretagogues such as arginine induce strong TDI independent of the duration of stimulation. Once the agent is removed, TDI persists for a considerable period. In contrast, prolonged stimulations with glucose (and other nutrients) lead to the amplification of the insulin response to subsequent stimuli; this can be demonstrated in the perfused rat pancreas, in perifused islets from several rodents, and in humans. We have termed this stimulatory signal time-dependent potentiation (TDP). The generation of TDP requires higher glucose concentrations and prolonged stimulation; the effect is retained for some time after cessation of the stimulus. Of major interest is the observation that, while the acute insulin response to glucose is severely reduced in glucose-intolerant animals and humans, TDP seems to be intact. The cellular mechanisms of TDI and TDP are poorly understood, but data reviewed here suggest that they are distinct from those that lead to the acute insulin response to stimuli. A model is proposed whereby the magnitude and kinetics of the insulin response to a given stimulus reflect the balance between TDP and TDI. Researchers studying the cellular and molecular mechanisms of insulin release are urged to take into consideration these complex and opposing factors which regulate insulin secretion.

The beta-cell mitochondria are known to generate metabolic coupling factors, or messengers, that mediate plasma membrane depolarization and the increase in cytosolic Ca(2+), the triggering event in glucose-stimulated insulin secretion. Accordingly, ATP closes nucleotide-sensitive K(+) channels necessary for the opening of voltage-gated Ca(2+) channels. ATP also exerts a permissive action on insulin exocytosis. In contrast, GTP directly stimulates the exocytotic process. cAMP is considered to have a dual function: on the one hand, it renders the beta-cell more responsive to glucose; on the other, it mediates the effect of glucagon and other hormones that potentiate insulin secretion. Mitochondrial shuttles contribute to the formation of pyridine nucleotides, which may also participate in insulin exocytosis. Among the metabolic factors generated by glucose, citrate-derived malonyl-CoA has been endorsed, but recent results have questioned its role. We have proposed that glutamate, which is also formed by mitochondrial metabolism, stimulates insulin exocytosis in conditions of permissive, clamped cytosolic Ca(2+) concentrations. The evidence for the implication of these and other putative messengers in metabolism-secretion coupling is discussed in this review.

Several reports indicate that hypoglycemic sulfonylureas augment Ca(2+)-dependent insulin secretion via mechanisms other than inhibition of the ATP-sensitive K(+) channel. The effect involves a 65-kd protein in the granule membrane and culminates in intragranular acidification. Lowering of granule pH is necessary for the insulin granule to gain release competence. Proton pumping into the granule is driven by a v-type H(+)-ATPase, but requires simultaneous Cl(-) uptake into the granule via metabolically regulated ClC-3 Cl(-) channels to maintain electroneutrality. Here we discuss the possibility that modulation of granule ClC-3 channels represents the mechanism whereby sulfonylureas directly potentiate the beta-cell exocytotic machinery.

Regulated exocytosis forms the basis for many intercellular signaling processes, for example, in hormone secretion or neurotransmitter release. During regulated exocytosis, the membrane of a secretory vesicle fuses with the plasma membrane in a tightly controlled reaction that is most often triggered by calcium. Recent advances have allowed major insights into the molecular mechanisms that mediate regulated exocytosis. In the present review, we will briefly discuss two key features of regulated exocytosis that have been particularly well studied recently. First, we will examine the current understanding of the membrane fusion reaction that underlies regulated exocytosis and that is effected by SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) and munc18-like proteins similar to other membrane fusion reactions. Second, we will describe the role of the major candidates for the calcium sensors that trigger exocytosis, a protein family called synaptotagmins. Although our understanding of regulated exocytosis is as yet incomplete, the results from the studies of SNAREs, munc18s, and synaptotagmins have provided a molecular anchor for a more complete future description.

A workshop on autoreactive T-cell responses in NOD mice was held to optimize autoreactive T-cell detection methodologies. Using different proliferation assay protocols, 1 of the 11 participating laboratories detected spontaneous T-cell responses to GAD(524-543) and insulin(9-23) in their NOD mice. Two other laboratories were able to detect autoreactive responses when using enzyme-linked immunospot assay (ELISPOT) and enzyme-linked immunosorbent assay (ELISA) analysis of cytokines in culture supernatants, suggesting that these assays provided greater sensitivity. To address the divergent findings, a follow-up mini-workshop tested NOD mice from four different colonies side-by-side for T-cell proliferative responses to an expanded panel of autoantigens, using the protocol that had enabled detection of responses in the 1st International NOD Mouse T-Cell Workshop. Under these assay conditions, 16 of 16 NOD mice displayed proliferative responses to whole GAD65, 13 of 16 to GAD(524-543), 9 of 16 to GAD(217-236), 7 of 16 to insulin(9-23), and 5 of 16 to HSP277. Thus, spontaneous proliferative T-cell responses can be consistently detected to some beta-cell autoantigens and peptides thereof. Overall, the results suggest that more sensitive assays (e.g., ELISPOT, ELISA analysis of cytokines in supernatants, or tetramer staining) may be preferred for the detection of autoreactive T-cells.

Several self-antigens have been reported as targets of the autoimmune response in nonobese diabetic (NOD) mice. The aim of this workshop was to identify autoantibody assays that could provide useful markers of autoimmunity in this animal model for type 1 diabetes. More than 400 serum samples from NOD (4, 8, and 12 weeks of age and at diabetes onset), BALB/c, and B6 mice were collected from six separate animal facilities, coded, and distributed to five laboratories for autoantibody measurement. Insulin autoantibodies (IAA) were measured by radiobinding assay (RBA) by four laboratories and by enzyme-linked immunosorbent assay (ELISA) in one laboratory. Using the 99th percentile of BALB/c and B6 control mice as the threshold definition of positivity, IAA by RBA were detected in NOD mice at frequencies ranging from 10 to 30% at age 4 weeks, from 26 to 56% at 8 weeks, from 42 to 56% at 12 weeks, and from 15 to 75% at diabetes onset. With ELISA, IAA signals differed significantly between control mouse strains and increased with age in both control and NOD mice, with frequencies in NOD animals being 0% at 4 weeks, 14% at 8 weeks, 19% at 12 weeks, and 42% at diabetes onset. For IAA, the ELISA results were relatively discordant with those of RBA. GAD autoantibody (GADA) and IA-2 autoantibody (IA-2A) signals obtained by RBA were low (maximum 2.5% of total) but were increased in NOD mice compared with control mice at diabetes onset (GADA 29-50%; IA-2A 36-47%). ELISA also detected GADA (42%) and IA-2A (50%) at diabetes onset, with results concordant with those of RBA. Remarkably, GADA and IA-2A frequencies varied significantly with respect to the source colony of NOD mice. Furthermore, whereas neither GADA nor IA-2A correlated with IAA, there was strong concordance between GADA and IA-2A in individual mice. Sera with increased binding to GAD and IA-2 also had increased binding to the unrelated antigen myelin oligodendrocyte glycoprotein, and binding to GAD could not be inhibited with excess unlabeled antigen, suggesting nonspecific interactions. In sum, this workshop demonstrated that IAA measured by sensitive RBA are a marker of autoimmunity in NOD mice and draw into question the true nature of GADA and IA-2A in this animal model.

Recent advances in molecular and cell biology may allow for the development of novel strategies for the treatment and cure of type 1 diabetes. In particular, it is now possible to envisage restoration of insulin secretion by gene or cell-replacement therapy. The beta-cell is, however, remarkably sophisticated, and many of the features of this highly differentiated secretory cell will have to be faithfully mimicked in surrogate cells. In particular, insulin is normally secreted in a well-regulated fashion in rapid response to the metabolic needs of the individual and most specifically (but not exclusively) to changes in circulating levels of glucose. Such regulated secretion will be indispensable in order to avoid both hyper- and hypoglycemic episodes and depends on the ability of cells to store insulin in secretory granules before exocytosis in response to physiological stimuli. Furthermore, any newly created insulin-secreting cell will have to be able to adapt to alterations in insulin requirements that accompany changes with exercise, body weight, and aging. Fine tuning of insulin secretion over the longer term will also be important to avoid "clinical shifting" that could be caused by over-insulinization, including increased adiposity and cardiovascular disease. Finally, it will be necessary to ensure that newly created or implanted (surrogate) beta-cells are protected in some way from recognition by the immune system and in particular from autoimmune destruction.