| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Endocrine Service and Diabetes and Heart Center of The Heart Institute, Hospital das Clinicas of The University of São Paulo Medical School, São Paulo, SP 05403-000, Brazil
Correspondence: Address all correspondence to: Bernardo L. Wajchenberg, Diabetes and Heart Center of The Heart Institute, Hospital das Clinicas of The University of São Paulo Medical School, Av. Eneas C. Aguiar 44, Bloco II, AB São Paulo, SP 05403-000, Brazil. E-mail: bernarwaj{at}globo.com.
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. Decline of ß-Cell...III. Factors for Progressive...IV. Clinical Intervention to...References |
|---|
Among the interventions to preserve or "rejuvenate" ß-cells, short-term intensive insulin therapy of newly diagnosed type 2 diabetes will improve ß-cell function, usually leading to a temporary remission time. Another intervention is the induction of ß-cell "rest" by selective activation of ATP-sensitive K+ (KATP) channels, using drugs such as diazoxide.
A third type of intervention is the use of antiapoptotic drugs, such as the thiazolidinediones (TZDs), and incretin mimetics and enhancers, which have demonstrated significant clinical evidence of effects on human ß-cell function.
The TZDs improve insulin secretory capacity, decrease ß-cell apoptosis, and reduce islet cell amyloid with maintenance of neogenesis. The TZDs have indirect effects on ß-cells by being insulin sensitizers. The direct effects are via peroxisome proliferator-activated receptor 
activation in pancreatic islets, with TZDs consistently improving basal ß-cell function. These beneficial effects are sustained in some individuals with time. There are several trials on prevention of diabetes with TZDs.
Incretin hormones, which are released from the gastrointestinal tract in response to nutrient ingestion to enhance glucose-dependent insulin secretion from the pancreas, aid the overall maintenance of glucose homeostasis through slowing of gastric emptying, inhibition of glucagon secretion, and control of body weight. From the two major incretins, glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP), only the first one or its mimetics or enhancers can be used for treatment because the diabetic ß-cell is resistant to GIP action. Because of the rapid inactivation of GLP-1 by dipeptidyl peptidase (DPP)-IV, several incretin analogs were developed: GLP-1 receptor agonists (incretin mimetics) exenatide (synthetic exendin-4) and liraglutide, by conjugation of GLP-1 to circulating albumin. The acute effect of GLP-1 and GLP-1 receptor agonists on ß-cells is stimulation of glucose-dependent insulin release, followed by enhancement of insulin biosynthesis and stimulation of insulin gene transcription. The chronic action is stimulating ß-cell proliferation, induction of islet neogenesis, and inhibition of ß-cell apoptosis, thus promoting expansion of ß-cell mass, as observed in rodent diabetes and in cultured ß-cells. Exenatide and liraglutide enhanced postprandial ß-cell function.
The inhibition of the activity of the DPP-IV enzyme enhances endogenous GLP-1 action in vivo, mediated not only by GLP-1 but also by other mediators. In preclinical studies, oral active DPP-IV inhibitors (sitagliptin and vildagliptin) also promoted ß-cell proliferation, neogenesis, and inhibition of apoptosis in rodents. Meal tolerance tests showed improvement in postprandial ß-cell function.
Obviously, it is difficult to estimate the protective effects of incretin mimetics and enhancers on ß-cells in humans, and there is no clinical evidence that these drugs really have protective effects on ß-cells.
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. Decline of ß-Cell...III. Factors for Progressive...IV. Clinical Intervention to...References |
|---|
|
II. Decline of ß-Cell Function |
|---|
|
TopAbstractI. IntroductionII. Decline of ß-Cell...III. Factors for Progressive...IV. Clinical Intervention to...References |
|---|
B. Pathophysiology
Insulin resistance and impaired insulin secretion are usually present in patients with classical DM2 as well as in most individuals with IGT. The relative contributions of impaired ß-cell function and insulin sensitivity in DM2 have been controversial, especially over what constitutes the primary genetic defect. Current evidence points to ß-cell dysfunction as the first demonstrable defect with limited capacity to compensate for the presence of insulin resistance. However, the modulating effect of insulin sensitivity on ß-cell function has to be considered for the assessment of insulin release in individuals at risk of developing DM2. The nature of this relationship is such that insulin sensitivity and ß-cell function are inversely and proportionally related, whereby the product of these two parameters is constant, being referred to as the disposition index (10), and in turn can be interpreted as a measure of the ability of the ß-cells to compensate for insulin resistance. Mathematically, this relationship is described by the hyperbolic relationship between the acute insulin response (AIR) and the metabolic action of insulin to stimulate glucose disposal (M) and is referred to as glucose homeostasis, with glucose concentration assumed to remain constant along the hyperbola. According to Stumvoll et al. (11), because glucose is one of the signals stimulating AIR in response to decreasing M, hypothetically, as with any normally functioning feed-forward system, AIR should not fully compensate for worsening M, since this would remove the stimulus for compensation. Evidence was provided from cross-sectional, longitudinal, and prospective data on Pima Indians (n = 413) and Caucasians (n = 60), demonstrating that fasting and postprandial glucose concentrations increase with decreasing M despite normal compensation of AIR. To denote this physiological adaptation to chronic stress (insulin resistance), Stumvoll et al. (11) have proposed the term "glucose allostasis." Allostasis (stability through change) would ensure continued homeostatic response (stability through remaining the same) to acute stress at some cumulative costs to the system. With increasing severity and over time, the allostatic load (increase in glycemia), while increasing the burden that persistent increases in insulin secretion place on the ß-cell and perhaps other parts of the system, may have pathological consequences, such as the development of DM2. However, not all insulin-resistant subjects develop this disease where, in approximately 80% of cases, increased insulin secretion is sufficient to offset overt hyperglycemia (12). Nevertheless, in some individuals, failure of the ß-cell to maintain adequate supplies of insulin triggers the development of DM2. Prospective studies (average follow-up of 7 yr) in Pima Indians, a very high-risk population for developing DM2, have shown that impaired ß-cell function, as assessed by AIR to glucose (AIRg), was a predictor of developing diabetes, independent of obesity and insulin resistance (prevalent in this population), evaluated by the hyperinsulinemic glucose clamp (13). In a longitudinal study of 48 Pima Indians with normal glucose tolerance followed for an average of 5 yr with multiple measurements performed during this period, 17 of these individuals progressed from normal glucose tolerance through IGT to diabetes and insulin secretion declining progressively by 78%, whereas insulin sensitivity declined by 14%, associated with an increase in mean body weight from 93.7 to 106.9 kg. In the remaining 31 individuals who did not develop diabetes, a similar 11% decrease in insulin sensitivity was associated with a 30% increase in insulin secretion. At baseline, however, when all subjects had normal glucose tolerance, those individuals who subsequently progressed to diabetes had ß-cell function markedly decreased for their degree of insulin resistance compared with those subjects who did not progress over time (14). From this and other studies, some of them previously reviewed, substantial evidence has accumulated showing that deterioration in ß-cell function typically precedes hyperglycemia and has an important influence on its course. Furthermore, most of the available evidence supports the view that DM2 is a heterogeneous disorder in which the major genetic factor is impaired ß-cell function, whereas insulin resistance is the major acquired factor at least in Caucasian populations. However, this may differ in a high-risk non-Caucasian population with severe genetic predisposition to insulin resistance, e.g., Indians (15). Compelling evidence that impaired insulin secretion is the major genetic trait has been provided via the assessment of the degree of ß-cell compensation to reduced insulin sensitivity among normal glucose-tolerant offspring or siblings of Caucasian familial DM2 kindreds (i.e., families with at least two siblings diagnosed with DM2 before age 65 yr) (16, 17). In effect, glucose-tolerant members of Caucasian familial DM2 kindreds (but not similar obese control subjects) exhibit impairment in ß-cell compensation for obesity-related insulin resistance. Furthermore, the heritability of insulin secretion is 67% (16).
The demonstration of ß-cell dysfunction before the onset of DM2 and the recognition of the confounding effect of obesity on insulin sensitivity cast doubt on the theory that insulin resistance is the primary cause of DM2 (15). Several studies of the insulin-secretory capacity of glucose-tolerant individuals with a predisposing ethnicity or a family history of DM2 have indicated that ß-cell dysfunction, like insulin resistance, occurs in genetically predisposed individuals with normal glucose tolerance, well before the emergence of overt diabetes (see references cited in Ref. 15). Of note in these studies is the fact that there were no significant differences with respect to insulin sensitivity between the predisposed subjects and the control groups (glucose-tolerant subjects with no family history of DM2), suggesting that alterations in insulin secretion precede insulin resistance in patients at risk for developing DM2.
Considering that the vast majority (85–90%) of DM2 patients are obese and that intraabdominal obesity, being the major determinant of insulin resistance (not only in obese individuals but also in apparently nonobese individuals who have evidence of increased abdominal fat) represents a significant genetic component (18), then insulin resistance occurring as a result of this could be considered genetic. Nevertheless, most obese individuals who are insulin resistant are not diabetic and what distinguishes them from those who are diabetic is, as indicated previously, the ability of their pancreatic ß-cells to compensate for insulin resistance. Although insulin resistance may be critical for the development of diabetes, irreversibly impaired insulin secretion plays, as expected, an essential role in diabetes persistence after weight loss and restoration to normal or near-normal insulin sensitivity (19).
Regarding the 10–15% of patients with DM2 who are not obese, whether they become diabetic depends on the balance between the severity of insulin resistance and the ability of the ß-cell to compensate for the insulin resistance, as in the case of obese individuals. A spectrum could exist; at one extreme, the nonobese individual may be insulin resistant before or after the development of diabetes. High-fat diet, decreased physical fitness, increase in visceral fat, smoking, pregnancy, certain medications, and hyperglycemia itself (glucose toxicity) can all cause insulin resistance. At the other extreme of the spectrum, insulin resistance is absent, and the immediate cause of diabetes is then impaired insulin secretion. Only a minority of DM2 patients belong to this group, such as those with maturity-onset diabetes of the young and nonobese blacks and Swedes. There are no well-documented cases in which impaired insulin secretion is absent and insulin resistance is the immediate cause of DM2 (20).
The defect of insulin secretion in DM2 is related to two confounding components: insulin deficiency and disturbed kinetics of secretion combined with impaired glucose stimulus-secretion coupling (21).
1. Insulin deficiency.
Insulin deficiency in DM2 is relative to the prevailing hyperglycemia and its measurement (immunoreactive or "true" insulin), at least in the early stages of the disease where these could show normal or increased values. However, because this level of insulin is insufficient to control the prevailing hyperglycemia, it would indicate that "insulin deficiency" is more evident when "true" insulin is measured. Besides, DM2 is characterized by an increased fasting ratio of proinsulin [long-chain precursor of insulin which is processed in the ß-cell to produce equimolar quantities of mature ("true") insulin and C-peptide] to total immunoreactive insulin (reflecting all immunoreactive species including proinsulin and conversion intermediates) (PI/IRI) indicating a reduced processing of proinsulin, which correlates with ß-cell dysfunction and is predictive of disease development (22). From the observed significant negative correlation between fasting PI/IRI ratio and the acute maximal insulin responses to arginine and glucose in DM2, it has been proposed that the PI/IRI ratio, a relatively easy parameter to obtain, indicates a reduced ß-cell capacity in patients with DM2 (23).
The determination of insulin in fasting plasma samples has the caveat that in this situation the demand for insulin is at its lowest. Only after stimulation does the magnitude of the insulin deficiency become fully manifested.
2. ß-Cell secretory defect.
It is well recognized that the relatively selective loss of glucose stimulation manifested by the lack of the first phase of secretion and decreased second phase (24) in DM2 patients as well in subjects with IGT, demonstrates multiple abnormalities in both qualitative and quantitative measures of insulin secretion. Prandial insulin release after an oral glucose load or mixed meal is also delayed (25), permitting the elevated postprandial glucose concentrations characteristic of these individuals, despite relatively normal fasting glucose levels in IGT (26). Sensitivity to nonmetabolizable stimuli such as arginine remains normal, although the magnitude of glucose-dependence of the insulin response may be attenuated. Baseline insulin secretion is normally pulsatile, with a periodicity of 5 to 10 min. The defective release of insulin in DM2 involves reduced diurnal oscillations (27), impaired ultradian oscillations (28), and their reduced entrainment (29) as well as impaired rapid pulsatile secretion (30). The hypothesis of impaired insulin oscillations as a primary ß-cell defect in DM2 was supported by the observation of defective oscillatory insulin release in first-degree relatives of such patients (31). However, Ritzel et al. (32) observed that the parameters of pulsatile insulin secretion were similar in normal subjects and DM2 and IGT subjects by deconvolution, spectral, and autocorrelation analysis and approximate entropy. This could also partly account for some of the insulin resistance in a number of studies showing that equal amounts of insulin presented to target organs have improved action when delivered in a pulsatile manner (see references cited in Ref. 33). These defects of secretion, particularly the lack of response to glucose, acting in concert with the insulin deficiency along with disproportionate amounts of proinsulin, compromise the ability of residual ß-cells to provide sufficient insulin to achieve euglycemia in the presence of insulin resistance or otherwise.
C. Decreased ß-cell mass
Insights into changes that occur in the ß-cells have been gained mainly from diabetic animal models that support reduced ß-cell mass as a significant contributory factor to diminished insulin secretion in DM2. Many of the signals that regulate the balance of replication of the existing cells/neogenesis from stem cells and cell death through necrosis or apoptosis and thus determine net ß-cell mass have been identified, but it is less clear which of these factors contribute to the failure of ß-cell mass augmentation. Obviously, diminished proliferation or increased apoptosis or both will result in lower ß-cell mass. There is evidence that increased ß-cell apoptosis in the Zucker diabetic fatty rat (ZDF), an animal model of DM2, underlies the decreased ß-cell mass seen in these animals (34). Autopsy data examining ß-cell mass in obese and diabetic humans have been sparse and sometimes contradictory. It is difficult to distinguish between the two mechanisms, cell formation and cell death, in human tissue sections mainly because dead cells are removed rapidly from the islet by macrophages and neighboring cells, making it hard to quantify cell death. Although cell proliferation can be quantified in tissue sections using markers, this only provides a single snapshot in time that probably does not reflect accurately the complex dynamic of the process (21). Recently, the study by Butler et al. (4), as mentioned in Section I, showed relative ß-cell volume to be increased by 50% in obese compared with lean nondiabetic pancreas and was attributed to increased neogenesis of islets from exocrine ductal tissue. Relative ß-cell volume was decreased in obese IGT and more in obese DM2 compared with nondiabetic controls, being attributable to accelerated cell apoptosis. Other studies (35, 36) have also shown a reduction in ß-cell mass in DM2, leading to a rapid emphasis of this etiological factor.
Potential mechanisms of ß-cell adaptation can be summarized as follows (37): 1) functional up- and down-regulation of secretory machinery; 2) ß-cell recruitment; 3) ß-cell turnover and islet mass changes [evidence mainly from rodent work: increasing/decreasing cell size; ß-cell turnover; de novo differentiation (neogenesis) of ß-cells within islets and new islets budding from exocrine acinar cell pancreas; and ß-cell death by necrosis vs. apoptosis (programmed ß-cell death)].
Opinions diverge regarding the relative contribution of a decrease in cell mass vs. an intrinsic defect in the secretory machinery. A decrease in ß-cell mass is likely to play a role in the pathogenesis of human DM2 as it does in rodent models of the disease, as indicated above. However, in contrast with type 1 diabetes, which has ß-cell mass reduction of 70–80% at the time of diagnosis, DM2 initial pathological studies suggested ß-cell loss of 25–50% (38, 39), although this has been disputed by others (40). Because ß-cells cannot be measured in vivo, it remains unclear whether DM2 has a lower ß-cell mass early in life, has failed to increase ß-cell mass in the face of a given insulin resistance, or has a progressive ß-cell loss as suggested by epidemiological evidence (1, 2, 3). Moreover, the secretion defect is probably more severe than could be accounted for solely by the reduction in ß-cell mass in DM2 (41). Based on results obtained in rodent models of the disease and cultured rodent and human islets, it seems reasonable to assume that dyslipidemia and hyperglycemia negatively affect ß-cell mass by increasing ß-cell apoptosis in human DM2 (21, 42). Section III discusses recent hypotheses on the mechanisms of glucotoxicity and lipotoxicity besides other factors for progressive loss of ß-cell function and mass.
Summary/conclusions.
There is a progressive deterioration in ß-cell function over time in DM2, independent of the type of treatment, where pancreatic islet function has been found to be about 50% of normal at the time of diagnosis, regardless of the degree of insulin resistance, as indicated by the UKPDS. The decline of ß-cell function, the first demonstrable defect, is the limited capacity to compensate for the presence of insulin resistance, both usually present in classical DM2 as well as in most individuals with IGT. Most of the evidence supports the view that DM2 is a heterogeneous disorder in which the major genetic factor is impaired ß-cell function where insulin resistance would be the major acquired factor at least in Caucasians. The defect of insulin secretion in DM2 is related to two confounding components: insulin deficiency and ß-cell secretory defect. On the other hand, there is an impaired glucose sensing in the 
-cells leading to hyperglucagonemia.
The reduction of ß-cell mass is attributable to accelerated apoptosis. Regarding the relative contribution of a decrease in cell mass vs. an intrinsic defect in the secretory machinery, the latter is probably more severe than could be accounted for by the reduction in ß-cell mass in DM2.
|
III. Factors for Progressive Loss of ß-Cell Function and Mass |
|---|
|
TopAbstractI. IntroductionII. Decline of ß-Cell...III. Factors for Progressive...IV. Clinical Intervention to...References |
|---|
The fact that glucose induces apoptosis in ß-cells is probably linked to the relative specificity of this toxicity toward the ß-cell but not to other islet or most nonislet cell types. The ß-cell is extremely sensitive to small changes in ambient glucose. When these changes are of short duration and lie within the physiological range, such as after a meal, they lead to insulin secretion. When of longer duration and more pronounced in magnitude, perhaps they are translated by the ß-cell glucose-sensing pathways into proapoptotic signals (21). One such signal might be endoplasmic reticulum (ER) stress triggered by an increase in insulin biosynthesis (48) leading to increased insulin secretion as well as increased proinsulin biosynthesis to replenish ß-cell insulin stores. The increase in proinsulin biosynthesis in turn causes an increased flux of protein through the ER of the ß-cell, which is quite high compared with other cell types even under physiological conditions, and any further increase is expected to be conducive to ER stress. Furthermore, chronic hyperglycemia could also lead to long-term increases in cytosolic Ca2+ [as opposed to the normal short-term increases arising from glucose-induced closure of ATP-dependent potassium channels (KATP)] that could in turn be proapoptotic (49).
Long-term hyperglycemia also induces the generation of reactive oxygen species (ROS) leading to chronic oxidative stress because the islets express very low levels of antioxidant enzymes and activity. ROS, particularly hydroxyl radicals, interfere with normal processing PDX-1 (pancreas duodenum homeobox-1) mRNA, a necessary transcription factor for insulin gene expression and glucose-induced insulin secretion besides being a critical regulator of ß-cell survival (47, 46). The generation of ROS (and reactive nitrogen species) will ultimately activate stress-induced pathways [nuclear factor 
B (NF-B), stress kinases, and hexosamines] to manipulate cell fate (50, 51, 52). Del Guerra et al. (53), studying islets isolated from the pancreata of 13 DM2 patients (age ranging from 49 to 75 yr, and diabetes duration from 2 to 8 yr, except one patient with 23-yr known duration of the disease) and 13 matched nondiabetic cadaveric organ donors, demonstrated several functional and molecular defects, confirming that DM2 islets release less insulin than control islets accompanied by altered expression of glucotransporters and of glucokinase, reduced activation of AMP-activated protein kinase and alterations in some transcription factors regulating ß-cell differentiation and function. Markers of oxidative stress, such as nitrotyrosine and 8-hydroxy-2'-deoxyguanosine concentrations, were significantly higher in DM2 than control islets and correlated with the degree of glucose-stimulated insulin release impairment. The addition of glutathione in the incubation medium determined reduction of oxidative stress (as suggested by diminished levels of nitrotyrosine), improved glucose-stimulated insulin secretion, and increased insulin mRNA expression. Thus, Del Guerra et al. (53) concluded that the functional impairment of DM2 islets could be, at least in part, reversible by reducing islet cell oxidative stress. It is important to emphasize that in this study the percentage of ß-cells was only slightly (
10%), although significantly, reduced in diabetic islets compared with control islets. As reasoned by Robertson et al. (47), if the steady decline in ß-cell function in DM2 is attributable to any significant extent to ongoing apoptosis via chronic oxidative stress with no deterioration in ß-cell replication, then interference with apoptosis by antioxidants or any other therapy, might provide a much needed new approach to conventional treatment that could stabilize ß-cell mass.
B. Lipotoxicity
Diabetes is associated with dyslipidemia characterized by an increase in circulating free fatty acids (FFAs) and changes in lipoprotein profile. In healthy humans, besides the hyperinsulinemia induced by an acute elevation of FFAs, there is also an increase in glucose-stimulated insulin secretion after prolonged FFA infusion (48 and 96 h) (54, 55) but not in nondiabetic individuals genetically predisposed to developing DM2 (55). In healthy control subjects, the FFA-induced insulin resistance was compensated by the enhanced insulin secretion, whereas persistently elevated FFAs may contribute to progressive ß-cell failure (ß-cell lipotoxicity) in individuals genetically predisposed to DM2. Santomauro et al. (56) have demonstrated that overnight administration of the nicotinic acid analog acipimox lowered plasma FFA as well as fasting insulin and glucose levels, reduced insulin resistance, and improved oral glucose tolerance with decreased insulin levels in lean and obese nondiabetic subjects and in subjects with IGT and DM2. The significant decrease in insulin levels suggested that plasma FFA support between 30 and 50% of basal insulin levels. A sustained (7-d) reduction in plasma FFA concentrations in DM2, with acipimox, was also associated with enhanced insulin-stimulated glucose disposal (reduced insulin resistance) associated with a decreased content of intramyocellular long-chain fatty acids (FAs) and improvement in oral glucose tolerance test with a slight decrease in mean plasma insulin levels (57). These data suggest that physiological increases in plasma FFA concentrations in humans potentiate glucose-stimulated insulin secretion and are unlikely to be "lipotoxic" to ß-cells (58) but may contribute to progressive ß-cell failure in at least some individuals who are genetically predisposed to developing DM2 (55). For all studies reported in relation to the FFA-ß-cell interaction it is important to emphasize that the stimulatory effects on glucose-stimulated insulin secretion are physiological in nature, particularly during the fasted-to-fed transition. Circulating FFAs help maintain a basal rate of insulin secretion, keeping adipose tissue lipolysis in check.
In rodent islets, increased FFAs have been shown to be proapoptotic in ß-cells (59). Exposure of cultured human islets to saturated FAs such as palmitate are highly toxic to the ß-cell, inducing ß-cell apoptosis, decreased ß-cell proliferation, and impaired ß-cell function. In contrast, the monounsaturated FAs such as oleate are protective against both palmitate and glucose-induced apoptosis and increase in proliferation. The deleterious effect of palmitic acid was mediated via ceramide-mitochondrial apoptotic pathways, whereas induction of the mitochondrial protein Bcl-2 by oleic acid may contribute to the protective effect of monounsaturated FAs, such as palmitoleic or oleic acids (60).
A similar balance between pro- and antiapoptotic effects is found for lipoprotein action on insulin-secreting cells from mouse pancreatic islets. The sole available study testing the hypothesis that lipoproteins modulate the function and survival of the cells has demonstrated that purified human very low-density lipoprotein and low-density lipoprotein particles reduced insulin mRNA levels and ß-cell proliferation and were proapoptotic, whereas high-density lipoprotein protected ß-cells against these proapoptotic effects. The protective effects of high-density lipoprotein were mediated, partially at least, by inhibition of caspase-3 cleavage and activation of Akt/protein kinase B, whereas proapoptotic lipoproteins seem to act via c-Jun N-terminal kinase (61). These results are highly suggestive that the changes in lipoprotein profile observed in DM2 could contribute to the pathogenesis and progression of ß-cell failure.
Regarding the mechanism(s) by which lipotoxicity can impair ß-cell function, the emerging evidence has suggested that long-chain fatty acyl-coenzyme A (CoA) might be involved in the ß-cell dysfunction that occurs after prolonged exposure to FFAs, mediating its deleterious effects, at least in rodents, as initially proposed by Prentki and Corkey (62) and recently reviewed by Yaney and Corkey (63). According to these authors, the simultaneous presence of elevated glucose and FA results from glucose as an oxidative fuel, in accumulation of cytosolic citrate, the precursor of malonyl-CoA, which in turn inhibits carnitine-palmitoyl-transferase-1, the enzyme responsible for transport of FA into the mitochondria, blocking their oxidation and energy production, resulting in cytosolic accumulation of long-chain fatty acyl-CoAs. Thus, this model proposes that glucose concentration plays a critical role in the effect of FAs. Whether long-chain acyl-CoA accumulation directly affects ß-cell function or whether it serves as a precursor for other active molecules such as diacylglycerols or phospholipids directly activating protein kinase C isoforms somehow synergizing with glucose to enhance insulin secretion was not characterized, whereas the nature of the effectors downstream of lipid metabolite accumulation remains unknown (62, 63, 64).
As reported by Poitout and Robertson (64), the "malonyl-CoA/long chain-acyl-CoA" proposed as a biochemical basis for lipotoxicity implies that the effects of FA are greatly influenced by concomitant glucose concentration. Therefore, in the presence of physiological glucose concentrations, elevated FA should be readily oxidized in the mitochondrion and should not harm the ß-cell. Under circumstances in which both FA and glucose are elevated, accumulation of metabolites derived from FA esterification would inhibit glucose-induction insulin secretion and insulin gene expression. Thus, glucotoxicity and lipotoxicity are closely interrelated, in the sense that lipotoxicity does not exist without chronic hyperglycemia. Furthermore, the effects of glucose on lipid metabolism are so intense that according to Poitout and Robertson (64), lipotoxicity can be viewed as one mechanism of glucotoxicity. Hence, generation of ROS may be an alternative mechanism of both gluco- and lipotoxicity. In effect, exposure of islets to palmitate induces generation of ROS (65) and treatment of islets with metformin, which has antioxidant properties, protecting them from deleterious effects of FAs (66).
C. Proinflammatory cytokines and leptin
The chronic increase in inflammatory mediators observed in DM2 might not only affect insulin-sensitive tissues and blood vessel walls but could also affect pancreatic ß-cells. In addition to genetic factors, development of DM2, with a central role for the functional ß-cell mass, is strongly influenced by environmental factors, including decreased physical activity, nutrition, and obesity. This promotes the following factors, which are possible mediators of an inflammatory process (67).
1. Adipocyte-secreted factors.
Locally produced hormones and cytokines possess important auto/paracrine properties. Some are also released into circulation and have endocrine effects. In particular, leptin, TNF-, IL-6, and IL-1 receptor antagonist are produced and secreted by fat tissue, being increased in obesity, and have been causally linked to insulin resistance (18, 68).
Leptin is expressed primarily in adipose tissue, representing the most obvious exponent of the adipocyte. Recently, leptin has also been considered as a proinflammatory cytokine because of its structural similarity with other cytokines and its receptor-induced signaling pathways (69). In rodent islets, leptin induces ß-cell proliferation and protects from FFA-induced ß-cell apoptosis. In contrast, chronic exposure of human islets to leptin leads to ß-cell apoptosis via increasing release of IL-1ß and decreasing release of IL-1 receptor antagonist in the islets (70). Other cytokines released by adipocytes, including TNF- and IL-6, may also modulate ß-cell survival, although it is unclear whether the amount released into the circulation is sufficient to affect ß-cells (70). Furthermore, it may well be that these cytokines are only effective in the presence of other cytokines.
2. Increased cell nutrients.
Elevated glucose concentrations and FFAs, in addition to their role of cell nutrients, also have a dual effect on ß-cell turnover. Depending on duration of exposure to glucose or FFA and on the genetic background of the islets, glucose and FFAs may induce ß-cell proliferation and have pro- or antiapoptotic effects. According to Maedler et al. (50), elevated glucose concentrations induce ß-cell production of IL-1ß leading to ß-cell apoptosis. Furthermore, chronic hyperglycemia increases production of ROS, which may cause oxidative damage in ß-cells, as highlighted previously. Both IL-1ß and ROS activate the transcription of the NF-B, which plays a critical role in mediating inflammatory responses. On the other hand, increased concentrations of FFAs, particularly saturated, may affect the viability of the ß-cells directly, as discussed earlier, or via obesity, i.e., adipocyte-secreted cytokines (TNF-, IL-6, and leptin) may act directly on the ß-cells or activate the innate immune system.
3. Innate immune system and autoimmunity.
The innate immune system is considered to provide rapid host defenses until the slower adaptive immune response develops. In addition to the endocrine activity of the adipocytes, described above, macrophages and endothelium may contribute to increasing serum levels of IL-1ß, IL-6, and TNF- in DM2 (71) and may act on the pancreatic islets and impair ß-cell secretory function or activate the innate immune system (67). Similarly, these cytokines induce the liver to produce acute-phase proteins such as C-reactive protein, haptoglobin, fibrinogen, plasminogen activator inhibitor, and serum amyloid A.
As indicated by Donath et al. (67), when apoptotic cells are present in high enough numbers or when apoptosis is the consequence of exposure to cytokines such as IL-1ß and TNF-, they can provoke an immune response. Moreover, pronounced activation of the acute-phase response is associated with islet cell autoantibodies in patients with DM2 (72). After glucose and FFA-induced ß-cell apoptosis, it is conceivable that depending on age and on genetic and/or environmental factors, some DM2 patients may show mobilization of T cells reactive to ß-cell antigens, culminating in autoimmune destruction of ß-cells similar to that observed during earlier stages in "classical" type 1 diabetes (21). This response may be so discrete and desynchronized in time and space that it has evaded detection in earlier autopsy studies. The precise role of the innate and acquired immune system in the ongoing process of ß-cell demise in DM2 remains to be investigated.
D. Islet cell amyloid
The relevance of amyloid deposition in the deterioration of ß-cell function has been the subject of debate for many years. As indicated in Section I, deposits composed mainly of islet amyloid polypeptide (IAPP), also known as amylin, have been reported in up to 90% of DM2 individuals, compared with 10–13% of nondiabetic counterparts (4). IAPP is a 37-amino acid, ß-cell peptide that is costored and coreleased with insulin in response to ß-cell secretagogues. The normally soluble peptide is found in the circulation at 5–15 pmol/liter concentration in man, and like C-peptide (but not insulin) it is excreted by the kidney; the relationship of circulating insulin-like molecules to IAPP would therefore be more accurate with C-peptide than insulin (73).
Several bodies of evidence support a potential role of IAPP in the pathophysiology of ß-cell loss in DM2. In effect, spontaneous DM2 occurring in humans, monkeys, and cats shares close homology in IAPP sequence, which spontaneously forms amyloid fibrils in an aqueous environment. Nevertheless, proline substitutions in the region IAPP 24–29 that does not form fibrils in an aqueous solution, as found in rats and mice, do not spontaneously develop DM2 but instead require selective genetic manipulation to develop diabetes (see reference citations in Ref. 73). Transgenic mice expressing human IAPP (hIAPP) in ß-cells when obese, spontaneously develop diabetes characterized by islet amyloid and decreased ß-cell mass (74). Prospective studies in these mice support the hypothesis that the mechanism of the decreased ß-cell mass is increased apoptosis (75). It has been suggested that fibril size of IAPP may determine the capacity of amyloid to cause cell death and that intermediate-sized oligomeric material is chiefly responsible for ß-cell toxicity. Exposure of mouse islets to intermediate-sized aggregates of hIAPP caused disruption and vesiculation of cell membranes after 24 h, and both necrotic and apoptotic cells were evident after 48 h. These particles were more cytotoxic to mouse islet cells than matured particles containing large aggregates (76). The findings from the Mayo Clinic that only a minority (about 10%) of the cases with IFG had islet amyloid (evaluated in slides stained by Congo red and examined under polarized light for birefringence) present, while already having a deficit in presumptive ß-cell mass of approximately 40%, is consistent with the suggestion that small IAPP oligomers (not detectable by light microscopy) are the cause of ß-cell loss, whereas the large extracellular amyloid deposits visible by light microscopy are inert (4). Alternatively, it is possible that IAPP formation is secondary to the onset of hyperglycemia and not of primary importance in the pathophysiology of DM2 (4).
In the recently published review of islet amyloid (73), the authors concluded that in human DM2, islet amyloidosis largely results from diabetes-related pathologies (such as diabetes-associated abnormal proinsulin processing, which could contribute to destabilization of granular IAPP) and is not an etiological factor for hyperglycemia. However, the associated progressive ß-cell destruction leads to severe islet dysfunction and insulin requirement (Fig. 1
).
|
Two possible explanations account for the impaired ß-cell function consequent to decreased ß-cell mass (21): 1) increased insulin demand on residual ß-cells per se leading to changes in function (by ER stress or other mechanisms); and 2) hyperglycemia consequent to decreased ß-cell mass driving the impairment in ß-cell function. However, the decreasing ß-cell mass experimentally more often than not leads to a more or less prolonged period of hyperglycemia (79, 80). Moreover, in many instances there is compensatory regeneration of ß-cells that might not be as well-differentiated as older, fully mature cells. Regardless, there are in vitro studies on islet and in vivo studies by glucose infusion showing that high glucose for prolonged periods of time leads to impaired ß-cell mass and/or function. Therefore, accepting that glucose plays a central role among those factors contributing to ß-cell demise whereas transient postprandial hyperglycemic excursions may predominantly induce ß-cell proliferation in insulin-resistant individuals, this adaptive mechanism may fail in the long term and be overridden by ß-cell apoptosis. However, it is unlikely that glucotoxicity acts alone, and the negative contribution of saturated FAs, lipoproteins, leptin, and circulating and locally produced cytokines will burn out the ß-cells further. These factors will induce apoptosis and/or necrosis, which in the presence of proinflammatory cytokines may activate specific immunological phenomena, ultimately resulting in autoimmunity as indicated above (46).
Notwithstanding, DM2 in humans progresses over time, whereas impaired ß-cell function appears to be reversible at least to a certain degree in the early stages of the disease. In effect, the overnight infusion of glucagon-like peptide 1 (GLP-1) in DM2 improved first- and second-phase insulin responses to a 2-h hyperglycemic clamp. All responses on GLP-1 were not significantly different from nondiabetic control subjects (81). Besides, a recovery of the first-phase insulin secretion was observed after the addition of rosiglitazone given for 6 months to failing sulfonylurea and metformin regimen in DM2 with a duration of diabetes of 7.6 ± 2.1 yr (82). Similarly, the iv administration of the incretin mimetic, exenatide, maintained for 30 min, to DM2 patients treated with diet/exercise, with a duration of the disease of 4 ± 2 yr, restored the first and second phases of insulin secretion, after glucose challenge, with a pattern similar to that found in healthy paired controls (83). These results are compatible with the view that the deficient pattern of insulin secretion is mainly functional in nature and that the reduction in pancreatic islet mass is moderate in patients with DM2 (4, 83). However, the recovery of insulin secretion should be tested in further studies focusing on patients with more advanced stages of DM2. Furthermore, if an individual with DM2, even with severe hyperglycemia, is rendered euglycemic by either pharmacological means or changes in lifestyle, ß-cell function, in particular glucose responsiveness, can be restored, and this will be discussed later. This is particularly true in newly diagnosed DM2 (84, 85) where the limiting threshold for reversibility of decreased ß-cell mass has probably not been passed. Rendering an individual with DM2 euglycemic interrupts the vicious circle linking decreased ß-cell function with hyperglycemia. This might not only restore insulin secretory patterns but also allow for some restoration of ß-cell mass. Although there are many reasons to believe that this may occur in humans as it does in animal models of reduced ß-cell mass, this has never been documented by virtue of the fact that ß-cell mass cannot be measured noninvasively at the time of writing.
Understanding the mechanisms of ß-cell death and thus the decreased ß-cell mass and impaired function has provided the basis for a new therapeutic target, ß-cell preservation, particularly when one considers the reversibility of impaired ß-cell function and possibly ß-cell mass as discussed above. In this context, it should be noted that long-acting incretin mimetics, which besides enhancing glucose-sensing and insulin secretory capacity of the endocrine pancreas have also been shown in preclinical studies in rodents, have additional effects on ß-cell mass by stimulating proliferation and inhibiting apoptosis (86).
Other examples include the thiazolidinediones (TZDs), which have been linked to not only improving insulin secretory capacity but also preventing the loss of ß-cell mass by reducing apoptosis with maintenance of ß-cell neogenesis in ZDF rats (87) as well as in other murine models of DM2 (88). In human beings, TZDs were able to preserve and recover ß-cell function (82, 89) besides improving insulin sensitivity.
Summary/conclusions.
Numbered among the factors for progressive loss of ß-cell function and mass are glucotoxicity and lipotoxicity, which are closely interrelated in the sense that lipotoxicity does not exist without chronic hyperglycemia and, depending on duration of exposure to glucose or FFAs and on genetic background of the islets, may induce ß-cell proliferation and have pro- or antiapoptotic effects.
Proinflammatory cytokines and leptin, produced by fat tissue, could also affect ß-cells. In particular, leptin, TNF-, IL-6, and IL-1 receptor antagonist are a linkage of obesity to DM2. Leptin in islets leads to ß-cell apoptosis, whereas TNF- and IL-6 may also modulate ß-cell survival. Apoptotic cells can provoke mobilization of T cells reactive to ß-cell antigens, culminating in autoimmune destruction of ß-cells similar to that observed at earlier stages of type l diabetes.
Regarding the IAPP, several factors support its role in the pathophysiology of ß-cell loss in DM2. Alternatively, it is possible that IAPP formation is secondary to the onset of hyperglycemia and not of primary importance in the pathophysiology of DM2.
The link of reduced ß-cell mass to impaired function may be due to an increased demand on residual ß-cells per se leading to changes in function (ER stress or other mechanisms) or related to the hyperglycemia consequent to decreased ß-cell mass, driving the impairment in ß-cell function. In vitro and in vivo studies suggested that persistently elevated glucose levels play a central role among those factors (FFAs, lipoproteins, leptin, and cytokines) contributing to ß-cell demise.
Understanding the mechanisms of ß-cell death and thus decreased ß-cell mass and impaired function has provided the basis to ß-cell preservation, particularly when one considers that the impaired ß-cell function and possibly ß-cell mass appear to be reversible, particularly at early stages of the disease where the limiting threshold for reversibility of decreased ß-cell mass has probably not been passed.
|
IV. Clinical Intervention to Preserve or "Rejuvenate" ß-Cells |
|---|
|
TopAbstractI. IntroductionII. Decline of ß-Cell...III. Factors for Progressive...IV. Clinical Intervention to...References |
|---|
In a similar study (85), 138 newly diagnosed DM2 patients with fasting glucose greater than 200 mg/dl were hospitalized and treated with CSII for 2 wk. After CSII, the patients were followed longitudinally on diet alone. Optimal glucose control was achieved within 6.3 ± 3.9 d in 126 patients. The remission rates (percentage of individuals maintaining near euglycemia) at the third, sixth, 12th, and 24th months were 72.6, 67.0, 47.1, and 42.3%, respectively. Patients who maintained glycemic control for more than 12 months (remission group) had greater recovery of ß-cell function than those who did not (nonremission group) when assessed immediately after CSII. HOMA of ß-cell function (HOMA-ß) and AUC insulin during iv glucose tolerance test, as well as change in AIR, were significantly higher in the remission group. Furthermore, proinsulin decreased highly significantly; thus the proinsulin/insulin ratio was also reduced highly significantly (indication of an improvement in the quality of insulin secretion) as well as the HOMA of insulin resistance (HOMA-IR; surrogate for evaluation of the degree of insulin resistance) in the remission group. Li et al. (85) concluded that the improvement of ß-cell function, particularly the restoration of the first-phase insulin secretion, could be responsible for the remission.
Summary/conclusions.
Among the clinical interventions to preserve or "rejuvenate" ß-cells, short-term intensive insulin therapy of newly diagnosed DM2 has been proposed: 2- to 3-wk intensive therapy with multiple daily insulin injections of NPH plus regular or CSII. The improvement of ß-cell function, especially the restoration of the first-phase insulin secretion, would lead to remission at least for a period of time. Furthermore, proinsulin decreased highly significantly as did PI/IRI ratio, indicating an improvement in the quality of insulin secretion.
B. Modulation of the ß-cell ATP-sensitive K+ (KATP) channel
It has been well demonstrated that insulin release by the ß-cell is initiated by an increase in the intracellular Ca2+ concentration that is mediated by Ca2+ influx through voltage-gated Ca2+ channels in the plasma membrane. The opening and closing (gating) of these Ca2+ channels is determined by ß-cell membrane potential, which is in turn regulated by the activity of the KATP channel. In the unstimulated ß-cell, KATP channels are open and the outward movement of K+ ions through these channels holds the membrane potential at a hyperpolarized level at which voltage-gated Ca2+ channels are closed. When plasma glucose concentration rises, glucose uptake and metabolism by the pancreatic ß-cell are enhanced and, possibly through changes in intracellular adenine nucleotide concentrations, induce KATP channel closure. This leads to membrane depolarization, opening voltage-gated Ca2+ channels, and an increase in cytosolic Ca2+ that triggers the exocytosis of insulin. Drugs like the sulfonylureas act by inhibiting the KATP channel directly, and thereby through depolarizing the ß-cell and stimulating Ca2+ influx they enhance insulin secretion. Another class of drugs is the K-channel openers. These comprise a structurally unrelated group of compounds that have the common property of opening KATP channels, which hyperpolarizes the ß-cells and prevents insulin release, even in the presence of glucose. The most potent KATP channel opener in the ß-cell used in clinical medicine is diazoxide (95).
The concept of deficient insulin stores as a contributing factor to ß-cell dysfunction in DM2 was recognized 30 yr ago when diazoxide was used to inhibit insulin secretion (and induce ß-cell "rest") in patients with DM2 receiving insulin, leading to restoration of the insulin response to a combined stimulation of ß-cells with glucagon plus tolbutamide, not observed in control diabetics receiving placebo with insulin (96). Diazoxide has successfully been used to improve ß-cell function in animal studies, such as the recovery of ß-cell responsiveness in 90% pancreatectomized diabetic rats (97), and to prevent the progress of deranged ß-cell function in rats with streptozotocin-induced diabetes (98). Besides, in vitro studies of human islets have demonstrated that chronic hyperglycemia desensitizes ß-cells to glucose being accompanied by three major Ca2+ abnormalities: elevated basal Ca2+, loss of glucose-induced rise in Ca2+, and oscillatory activity disturbance with a decrease in glucose-induced slow oscillations. Relieving overstimulation with coculture with diazoxide (and 486 mg/dl glucose) significantly restored postculture insulin responses to glucose, lowered basal Ca2+, and normalized glucose-induced oscillatory activity, but failed to restore Ca2+ during postculture glucose stimulation (99). Therefore, the induction of ß-cell rest by selective activation of ß-cell KATP channels preserving insulin stores and pulsatile insulin secretion may provide a strategy to protect ß-cells from chronic overstimulation and to improve islet function. In effect, short-term diazoxide treatment of obese DM2 patients with poor metabolic control exerted moderate but beneficial effects on important parameters of insulin secretion: absent insulin response to an iv glucose challenge and glucose potentiation of arginine-induced insulin secretion after insulin treatment, whereas a response to both tests, albeit moderate, could be demonstrated after insulin plus diazoxide treatment. These effects could be dissociated from confounding effects of changes of glycemia because insulin infusion was used to achieve close to identical degrees of glycemia during the two treatment periods. Consequently, the results of this study indicate the beneficial effects of ß-cell rest. Besides, a significant reduction of the molar ratio of proinsulin to insulin was observed only after diazoxide plus insulin (100). The response to glucose and glucose potentiation of arginine-induced insulin secretion has been shown to correlate to the functioning islet mass as demonstrated after islet autotransplantation in humans (101).
It was also observed that, in adults with autoimmune diabetes, diazoxide treatment at onset preserved some residual insulin secretion for up to 18 months, when compared with a placebo group, when both groups were also receiving multiple insulin injection therapy (102). A similar study was performed in childhood type 1 diabetes, demonstrating that partial inhibition of insulin secretion for 3 months with diazoxide at onset of the disease, in addition to multiple daily insulin injections, extended the period of remission and temporarily preserved residual insulin production compared with placebo (maximum at 6 months, followed by a progressive decline) (103).
Although the beneficial effect of diazoxide may be due to induction of ß-cell "rest," it could also reflect in part the antiapoptotic effect of this type of drug because it was demonstrated that incubation of murine pancreatic islets and ß-cells with glucose induced apoptosis of ß-cells that is Ca2+ dependent, because introduction to the culture medium of diazoxide, which blocked glucose-induced Ca2+ increase, inhibited apoptosis (45). It is interesting to point out that the closure of KATP channels by the sulfonylureas tolbutamide and glibenclamide may also induce Ca2+-dependent ß-cell apoptosis in rodent and human islets (45, 104, 105), but this effect has only been observed in vitro and not consistently (106). The induction of ß-cell rest by a new selective KATP-channel opener (NN414) preserved insulin stores and pulsatile insulin secretion without restoring the orderliness of insulin secretion in human islets, isolated from cadaveric organ donors, cultured at high glucose concentrations (198 mg/dl) (107). According to Ritzel et al. (107), the concept of ß-cell rest may provide a strategy to protect ß-cells from chronic overstimulation and improve islet function. The authors also provide evidence that impaired glucose-regulated insulin secretion in DM2 partially involves mechanisms distinct. Afrom insulin stores and insulin secretion rates (ISRs).
In line with findings indicated in Section IV.A related to insulin therapy of newly diagnosed DM2, a clinical study by Alvarsson et al. (108), comparing insulin and sulfonylurea (gIibenclamide) treatment of recently (<2 yr) diagnosed DM2, showed that treatment with insulin given twice daily as premixed insulin (30% soluble and 70% NPH) preserved ß-cell function (glucagon-stimulated C-peptide response) more effectively than glibenclamide for the duration of the study (2 yr). On the other hand, the nonsulfonylurea secretagogues (meglitinides) repaglinide and natiglinide, which work similarly to sulfonylurea agents but have a more rapid onset and shorter duration of action, when applied for their respective circulating half-lives (less than 2 h) in vitro, appear not to have any apoptotic effect on cultured human islets. Thus, short-acting insulin secretagogues may be preferred to long-acting ones, such as glibenclamide, due to the fact that patients will be less exposed to proapoptotic stimuli during prolonged treatment with short-acting agents. This was according to in vitro studies of human islets, but in vivo confirmation is necessary because several additional factors including ß-cell proliferation and regeneration may compensate for glibenclamide-induced apoptosis (105).
Summary/conclusions.
The induction of ß-cell "rest" by selective activation of ß-cell KATP channels, using drugs such as diazoxide and the KATP channel opener NN414, that hyperpolarize ß-cells, preserves insulin stores and pulsatile insulin secretion, thus protecting ß-cells from chronic overstimulation and improving islet function besides reducing the PI/IRI ratio. Although the beneficial effect of those selective KATP channel openers may be due to inducing ß-cell "rest," it could also reflect in part the antiapoptotic effect of these drugs by blocking the glucose-induced cytosolic Ca2+ increase, secondary to the ß-cell KATP channels closure associated with the rise in plasma glucose, as shown in in vitro studies which have demonstrated that glucose-induced apoptosis of ß-cells is Ca2+ dependent. The closure of KATP channels by sulfonylureas, tolbutamide and glibenclamide, may also induce Ca2+-dependent ß-cell apoptosis also in in vitro studies in rodent and human islets. If these findings translate to patients with DM2, at least some long-acting sulfonylureas may have adverse effects on ß-cells, whereas short-acting ones will be preferred due to the fact that patients will be less exposed to proapoptotic stimuli during protracted treatment with these agents.
C. Antiapoptotic drugs
1. Leptin.
As indicated earlier, leptin in ZDF rat islet protects the ß-cell from FFA-induced apoptosis (109). However, it induces ß-cell apoptosis in human islets by the mechanism previously discussed (70).
2. Aminoguanidine.
Aminoguanidine acts as an antioxidant in vivo in diabetic animal models, preventing ROS formation and lipid peroxidation in cells and tissues, while also preventing oxidant-induced apoptosis (110). Moreover, at higher doses this inhibits inducible nitric oxide synthase with reduction in nitric oxide, which has been shown to mediate IL-1ß-induced impairment of ß-cell function and ultimately cause ß-cell apoptosis as reported in cultured prediabetic ZDF islets (59, 111). Because ROS are known to increase intracellular advanced glycation end-products (AGEs) and antioxidants are known to inhibit AGE formation (110), presumably the time required for formation of immunodetectable intracellular AGEs is much longer than the time-course for oxidant-induced apoptosis, suggesting that intracellular AGE formation does not play a causal role in this process of oxidant-induced apoptosis. Furthermore, diabetic models using aminoguanidine have also shown a reduction in diabetes-related complications such as retinopathy, neuropathy, and nephropathy. Specifically, reduction in retinal microvessel formation, albuminuria, and the prevention or decrease in motor and sensory nerve conduction velocity have been reported (see references cited in Ref. 112). In line with previously indicated considerations (47), the prescription of antioxidants as adjunct therapy in DM2 is warranted to determine whether this approach would prevent continued deterioration in ß-cell function, particularly when glycemic control is not satisfactory (47). Unfortunately, studies with antioxidant vitamins to date have not improved glycemic control. Additionally, aminoguanidine has not been used in humans because of its side effect profile in many organs.
Summary/conclusions.
Aminoguanidine acts as an antioxidant in vivo, in diabetic animal models, preventing ROS formation and lipid peroxidation and oxidant-induced apoptosis and at higher doses inhibits inducible nitric oxide synthase with reduction in nitric oxide, which was shown to ultimately cause ß-cell apoptosis in cultured prediabetic ZDF rat islets. Aminoguanidine has not been used in humans because of its side effect profile in many organs.
3. Effect of TZDs on ß-cells.
The TZDs are agonists of peroxisome proliferator-activated receptor (PPAR), a nuclear receptor that regulates transcription of genes involved in lipid and glucose metabolism. Although predominantly expressed in adipose tissue, PPAR is present in other insulin-sensitive tissues, including the pancreatic islet cells (113). Fully 10% of genes transcribed in adipose tissue are differentially expressed with the use of a PPAR agonist, as well as 2% of all genes expressed in liver and 1% of those expressed in skeletal muscle (114). In adipose tissue, stimulation of PPAR increases adipocyte differentiation, resulting in an increased number of small, insulin-sensitive adipocytes (115). The development of these insulin-sensitive cells enhances glucose uptake, improving glycemic control. Improved insulin sensitivity, reported for TZDs, both in monotherapy and combination in DM2, has the potential to protect ß-cells because they may reduce the demand on the pancreas (see references cited in Ref. 116). Lowering plasma glucose may reduce the risk of glucotoxicity. In addition, by increasing adipose-tissue mass, TZDs promote FA uptake and storage in adipose tissue, thus reducing levels of circulating FFAs where this has the potential to alleviate lipotoxicity. Hence, TZDs retain fat where it belongs, according to the "FA steal" hypothesis (115). Reductions in FFAs may be mediated by suppression of TNF- expression, which has been demonstrated by in vitro studies such as those of Arner (117), who has reported that rosiglitazone suppresses TNF- to almost undetectable levels in human preadipocytes. Thus, according to Arner (117), TZDs reduce the release of FFAs from the adipocytes by two mechanisms: 1) increasing the insulin sensitivity of the cell, and thereby improving the antilipolytic effect of insulin; and 2) reducing levels of TNF-, known to suppress the antilipolytic properties of insulin by inhibiting the insulin-signaling cascade.
Although rosiglitazone seems to be a pure PPAR agonist, the other TZD approved for the medical therapy is pioglitazone, which seems to also act like a partial PPAR agonist (118). In general, PPAR regulates genes involved in FA uptake and oxidation (liver, kidney, heart, and skeletal muscle), inflammation, and vascular function, whereas PPAR, as shown, regulates genes involved in FA uptake and storage, inflammation, and glucose homeostasis (119).
a. Animal data.
In obese ZDF rats, the overaccumulation of fat in the pancreatic islets induces ß-cell dysfunction, reflecting a mutation of the leptin receptor that blocks the normal triglyceride-lowering action of leptin on islets, as discussed previously, leading to lipotoxicity through exaggerated production of nitric oxide (111). In isolated islets from obese ZDF rats, addition of troglitazone halved the triglyceride content, doubled insulin secretion stimulated by arginine, and produced a greater than 30-fold increase to that stimulated by glucose (120). This is consistent with the ability of TZDs to prevent TNF--induced inhibition of insulin signaling by islets. In the same animal model, rosiglitazone treatment was shown to maintain ß-cell proliferation and to produce a 5-fold attenuation in the net rise in ß-cell death, preventing the loss of ß-cell mass indicated previously (87). Because excessive ß-cell apoptosis is associated with excessive accumulation of intracellular triglycerides, as reported earlier (59), staining of islet cells for insulin in diabetic db/db mice has shown evidence of regranulation when the islet cell triglyceride content had been reduced with rosiglitazone (121). Furthermore, pioglitazone given to three murine models of DM2 improved insulin secretory capacity and prevented the loss of ß-cell mass by reduction of oxidative stress with 1.5- to 15-fold higher levels of pancreatic insulin (122). The decrease of TNF- expression by pioglitazone in adipose tissue in db/db mice (123) might be associated, as indicated above, with an improvement in ß-cell dysfunction and survival. Treatment with rosiglitazone has also been shown to prevent islet amyloidosis in nondiabetic hIAPP transgenic mice (124).
In conclusion, the most relevant data in rodents indicated that the TZDs decrease ß-cell apoptosis, maintaining its neogenesis and preventing islet amyloidosis.
b. In vitro data in human ß-cells.
PPAR mRNA is expressed in isolated human pancreatic islets to a greater extent than PPAR mRNA, with predominance of PPAR2. This could be interpreted as a condition favoring lipid accumulation and, hence, lipid-induced damage (125). The expression of PPAR2 was markedly and time-dependently reduced by exposure to progressively higher concentrations of FFAs. This effect was not affected by the concomitant presence of high glucose. The presence of FFAs also produced deleterious cytostatic effect, inhibiting glucose-stimulated insulin release by 80%, as well as reducing islet insulin content by 75% and reducing insulin mRNA expression (125). However, incubation with rosiglitazone, a well-known PPAR agonist, prevented FFA-induced down-regulation of PPAR and insulin mRNA expression along with glucose-stimulation insulin release (125). In the same model, high concentrations of FFAs produced an almost 3-fold increase in rates of islet cell death, in association with significant increases in activity of the protease enzymes caspase 3 (apopain) and caspase 9, key mediators of apoptosis (104). Incubation with rosiglitazone at a concentration of 15 mg/ml attenuated islet cell death and normalized caspase activity levels (126).
Zeender et al. (127) were able to protect cultured human islets using pioglitazone (and sodium salicylate) against apoptosis and impaired function induced by IL-1ß and high glucose by blocking NF-B activation.
Lin et al. (128), noting the association of DM2 with increased ß-cell apoptosis and the presence of islet amyloid derived from IAPP, showed that IAPP induces apoptosis in cultured human pancreatic islets and that the addition of rosiglitazone to the incubation prevented the IAPP-induced apoptosis.
c. Mechanism of action of TZDs on the ß-cell
1) Indirect effects by amelioration of insulin sensitivity.
TZDs consistently lower fasting and postprandial glucose concentrations in clinical studies as well as plasma FFAs in type 2 diabetics, with a reduction of gluco- and lipotoxicity. Furthermore, as mentioned above, TZDs may directly protect the ß-cell from lipotoxic insult (125). Besides, insulin concentrations also decrease in the majority of trials, mirroring a decreased secretory demand on ß-cells. Such changes indicate that TZDs act as insulin sensitizers, which has been confirmed by direct measurements in in vivo studies in humans. Although the insulin-sensitizing effect of TZDs is well established, less is known about their influence on insulin secretion. It now appears that TZDs normalize the asynchronous insulin secretion that characterizes ß-cell failure. In a placebo-controlled study in patients with DM2, 13-wk treatment with rosiglitazone was shown to increase the ability of an oscillatory glucose infusion to program high-frequency pulsatile insulin secretion, despite the absence of any direct action on ß-cell secretory capacity. It was postulated that the improvement in ß-cell function could be related to a reduction in glucotoxicity due to the improved glycemic control and/or improved insulin sensitivity seen with TZDs (129). This could suggest an increased ability of the ß-cell to sense and respond to glucose changes within the physiological range after TZD treatment.
2) Direct effects via PPAR activation in pancreatic islets.
As a class effect, TZDs consistently improve basal ß-cell function, as measured by the HOMA model and as observed during TZD monotherapy and combination therapy (89, 130, 131, 132, 133, 134). Further evidence that TZDs exert beneficial effects on ß-cells derive from a study in which 2-month treatment with pioglitazone restored the first-phase insulin response to an iv glucose tolerance test in patients with IGT and with frank DM2 (135). Similar results were reported in a randomized 6-month study comparing the addition of rosiglitazone vs. insulin in patients with DM2 who were inadequately controlled on glimepiride and metformin, using a frequently sampled iv glucose tolerance test (FSIVGTT) (82). At study end, the AIRg was significantly increased with rosiglitazone, with no effect of exogenous insulin apparent. ß-Cell function determined by the disposition index, calculated as the product of AIRg and the insulin sensitivity index (Si = 1/HOMA-IR) also increased greatly. It should be noted that the restoration of the first-phase insulin response to glucose was independent of the correction of glucose toxicity.
Furthermore, extension studies with TZDs indicate that improvements in ß-cell function are sustained in some individuals over time. Four 52-wk clinical trials involving more than 3,700 patients with DM2, reviewed by Campbell (136), showed that pioglitazone was an effective long-term treatment, both as monotherapy and in combination with metformin or sulfonylurea. As well as maintaining glycemic control over the long term, pioglitazone also yielded benefits in terms of improvements in fasting insulin (which was reduced as proinsulin and C-peptide), lipid parameters, and hypoglycemia compared with other monotherapies or combination treatments. The longest follow-up study was that by Bell and Ovalle (137, 138) of type 2 diabetics on sulfonylurea, metformin, and rosiglitazone for 60 months; 22 (68%) patients in the study who were put on the triple therapy remained relatively well controlled. The predictor of failure and the need for insulin after a mean duration of 30 months was the nonsignificant or lacking increase in the meal- or glucagon-stimulated C-peptide. At 60 months, a frequent finding was also a reduction in fasting<
