Despite numerous attempts over 16 years, the results of aldose reductase inhibitor (ARI) trials for the treatment of diabetic neuropathy have not proven efficacy. This paper reviews each of the ARI trials, examines confounding factors, and proposes a future course. The confounding factors considered are pharmacokinetics (ARI penetration of human nerve), length of trial (in terms of the natural history of diabetic neuropathy), trial endpoints (reversibility or slowing of progression), reproducibility of clinical measurements (in terms of power calculations), standardization and quality control of endpoints, and clinically meaningful differences in endpoints. We conclude that ARIs are most likely to have a beneficial effect in the management of diabetic distal symmetrical polyneuropathy and autonomic neuropathy but that the clinical role of ARIs is to slow the progression of diabetic neuropathy rather than to reverse it. Future trials should be designed with adequate statistical power, with consideration of the variability of the endpoint measurements for long enough duration, and with rigorous quality control to definitively confirm the utility of ARIs in the treatment of diabetic distal symmetrical polyneuropathy and autonomic neuropathy.

Antioxidant treatment has been shown to prevent nerve dysfunction in experimental diabetes, providing a rationale for a potential therapeutic value in diabetic patients. The effects of the antioxidant alpha-lipoic acid (thioctic acid) were studied in two multicenter, randomized, double-blind placebo-controlled trials. In the Alpha-Lipoic Acid in Diabetic Neuropathy Study, 328 patients with NIDDM and symptomatic peripheral neuropathy were randomly assigned to treatment with intravenous infusion of alpha-lipoic acid using three doses (ALA 1,200 mg; 600 mg; 100 mg) or placebo (PLAC) over 3 weeks. The total symptom score (TSS) (pain, burning, paresthesia, and numbness) in the feet decreased significantly from baseline to day 19 in ALA 1,200 and ALA 600 vs. PLAC. Each of the four individual symptom scores was significantly lower in ALA 600 than in PLAC after 19 days (all P < 0.05). The total scale of the Hamburg Pain Adjective List (HPAL) was significantly reduced in ALA 1,200 and ALA 600 compared with PLAC after 19 days (both P < 0.05). In the Deutsche Kardiale Autonome Neuropathie Studie, patients with NIDDM and cardiac autonomic neuropathy diagnosed by reduced heart rate variability were randomly assigned to treatment with a daily oral dose of 800 mg alpha-lipoic acid (ALA) (n = 39) or placebo (n = 34) for 4 months. Two out of four parameters of heart rate variability at rest were significantly improved in ALA compared with placebo. A trend toward a favorable effect of ALA was noted for the remaining two indexes. In both studies, no significant adverse events were observed. In conclusion, intravenous treatment with alpha-lipoic acid (600 mg/day) over 3 weeks is safe and effective in reducing symptoms of diabetic peripheral neuropathy, and oral treatment with 800 mg/day for 4 months may improve cardiac autonomic dysfunction in NIDDM.

Foot ulceration and lower limb amputation are still common complications of diabetes. Diabetic peripheral neuropathy and peripheral vascular disease are the most important etiologic factors, but there is a complex interplay between these abnormalities and a number of other contributory factors, such as altered foot pressures, limited joint mobility, glycemic control, ethnic background, and cardiovascular parameters. Identification of patients at high risk of ulceration is nevertheless simple, and education of such patients can achieve a major reduction in amputation and ulceration rates.

The peripheral nerve disorders associated with diabetes are complex and probably involve a variety of causative mechanisms. This may give rise to difficulty in the classification of individual cases. A broad separation into rapidly reversible or more persistent phenomena is helpful. The former, which can be categorized as "hyperglycemic neuropathy," include minor sensory symptoms, reduced nerve conduction velocity, and resistance to ischemic conduction failure. From analogy with experimental studies in animals, nerve hypoxia is likely to play a significant role in their origin. Of the more persistent phenomena, a distal symmetric polyneuropathy that predominantly affects sensory and autonomic function is the most common manifestation. A distal axonopathy of dying-back type may represent the underlying pathogenetic basis. Other more persistent phenomena consist of focal and multifocal lesions giving rise to cranial, thoraco-abdominal, and limb neuropathies, including proximal lower limb motor neuropathy (diabetic amyotrophy). Some of these may have an ischemic basis. Multifocal proximal lesions can summate to produce an approximately symmetric diffuse distal neuropathy. Focal lesions at sites of entrapment or external compression may reflect an abnormal susceptibility of diabetic nerve to compressive damage. There is also evidence that focal inflammatory, including vasculitic, lesions may be involved in proximal lower limb neuropathies. Finally, superimposed chronic inflammatory demyelinating polyneuropathy may occur. For the evaluation of possible treatment regimens, it is essential that cases should be correctly classified as to type. Thus, the features falling into the category of hyperglycemic neuropathy should not contaminate the assessment of distal symmetric polyneuropathy. For this type, a widely accepted scheme for staging devised by P. J. Dyck is available. Other schemes are also available for the assessment of such cases, with differing degrees of complexity. Evaluation by serial nerve biopsies has also been proposed.

Chronic hyperglycemia may cause growth factor alterations that are likely to participate in tissue remodeling typical for diabetic late complications. However, few details of such events are known. The ocular vitreous fluid allows studies of growth factor levels in human eyes (after vitrectomy). The vitreous is highly inert and protected by the blood-retina barrier and thus probably reflects growth factor production by the normal retina. Vitreous from patients with proliferative diabetic retinopathy (PDR) was compared with vitreous obtained from patients with nonproliferative eye disease and with vitreous from patients without diabetes but with marked neovascular proliferations due to ischemia. This design permits us to distinguish diabetes-related from non-diabetes-related alterations. Insulin-like growth factor I (IGF-I), IGF-II, IGF binding protein 2 (IGFBP-2), and IGFBP-3 were elevated 3- to 13-fold in nondiabetic retinal ischemia and 1.5- to 3-fold in PDR, indicating that the changes were not restricted to diabetes. These changes may partially be explained by leakage of serum into the vitreous, since IGFs and IGFBPs are 20- to 50-fold higher in serum than in vitreous, and vitreous protein content was 1.5-fold elevated in PDR subjects and 5-fold in ischemia patients compared with control subjects. TGF-beta is a proposed antiangiogenic factor in the eye. TGF-beta2 was the predominant subtype in vitreous, and its total amount was not altered in PDR patients. More importantly, the active fraction of TGF-beta was decreased by 30 and 70% in PDR and nondiabetic retinal ischemia patients, respectively. Since plasmin may control TGF-beta activation, the serum protein alpha2-antiplasmin was measured and found to be significantly elevated to 150 and 250% of control values in PDR and ischemia patients, respectively. Thus, influx of serum proteins due to microvascular disturbances and hypoxia is proposed as a possible cause for vitreous alterations of IGF-I and of active TGF-beta. These changes seem to occur late in the sequence of events leading to PDR and are not specific for diabetes, but they were also observed in other diseases characterized by retinal hypoxia.

Islet transplantation is a treatment for diabetes that has the potential to normalize glucose levels and prevent the development of complications. In spite of the simplicity of the concept and the urgent need to provide such a treatment to patients, there has been a frustrating lack of progress. This perspective delves into the scientific and political impediments to success. The scientific barriers are the need to find a satisfactory source of insulin-producing tissue and the requirement to prevent this tissue from being destroyed by immune rejection and autoimmunity. The problems and potential of allografts, xenografts, and the development of cell lines are discussed. Multiple approaches to the prevention of immune destruction are considered, including immunobarrier devices, immunosuppression, development of tolerance, and genetic manipulation. The political barriers discussed include the problems of high expectations, the controversy surrounding targeted research, the balance between basic and applied research, the roles of industry and academia, the concerns about xenotransplantation, and the difficulties in developing a planned approach to the problem.

The metabolism of the storage polysaccharide glycogen is intimately linked with insulin action and blood glucose homeostasis. Insulin activates both glucose transport and glycogen synthase in skeletal muscle. The central issue of a long-standing debate is which of these two effects determines the rate of glycogen synthesis in response to insulin. Recent studies with transgenic animals indicate that, under appropriate conditions, each process can contribute to determining the extent of glycogen accumulation. Insulin causes stable activation of glycogen synthase by promoting dephosphorylation of multiple sites in the enzyme. A model linking this action to the mitogen-activated protein kinase signaling pathway via the phosphorylation of the regulatory subunit of glycogen synthase phosphatase gained widespread acceptance. However, the most recent evidence argues strongly against this mechanism. A newer model, in which insulin inactivates the enzyme glycogen synthase kinase-3 via the protein kinase B pathway, has emerged. Though promising, this model still does not completely explain the molecular basis for the insulin-mediated activation of glycogen synthase, which remains one of the many unknowns of insulin action.

Most patients with diabetes die from macrovascular complications. Little is known about the pathogenesis of diabetic vascular disease, but recent advances in molecular genetics and oxidation chemistry provide clues to the mystery of diabetes and atherosclerosis. Genetic variants of well-known proteins such as lipoprotein lipase and apolipoprotein E are common. These proteins are suitable candidates for mediating diabetic vascular risk because their variants can produce hypertriglyceridemia, a risk factor for atherosclerosis in diabetes. However, mutations could have different effects on lipoprotein flux across arteries depending on whether expression is dominant in the vascular space or the vascular wall. Lipoproteins retained in the arterial wall are subject to oxidative modification, which could be dependent on glycoxidation, the enzyme myeloperoxidase, or reactive nitrogen species derived from nitric oxide. Accelerated vascular disease in diabetes is likely the result of complex interactions between metabolic derangements such as hyperglycemia, mutations in genes controlling lipid metabolism, and antioxidant defense mechanisms.

Enteroviruses have been examined for their possible role in the etiology of IDDM for nearly 40 years, yet the evidence remains inconclusive. The mechanism of acute cytolytic infection of beta-cells, proposed by earlier studies, appears to be incompatible with the long preclinical period of autoimmunity preceding IDDM. Advances in molecular biology have improved our understanding of enteroviral biology and of potential alternative pathogenic mechanisms through which enteroviruses may cause diabetes. The focus of future human studies will likely shift from people with IDDM to those with prediabetic autoimmunity to determine whether acute enteroviral infections can promote progression from autoimmunity to overt diabetes. We propose that such studies use assays to detect enteroviral RNA, in addition to IgM serology. RNA assays can overcome sensitivity and type-specificity limitations of IgM assays as well as identify diabetogenic strains of enteroviruses, if such exist. Evaluation of the role of enteroviruses in triggering beta-cell autoimmunity in humans will require large prospective studies of young children. The Diabetes Autoimmunity Study in the Young--one of very few such studies currently underway--is focusing on potential interactions between HLA class II genes and enteroviral infections. Future studies will likely examine interactions between viral infections and non-HLA IDDM candidate genes, including those that may determine beta-cell tropism of candidate viruses.

Evidence is reviewed that free fatty acids (FFAs) are one important link between obesity and insulin resistance and NIDDM. First, plasma FFA levels are elevated in most obese subjects. Second, physiological elevations in plasma FFA concentrations inhibit insulin stimulated peripheral glucose uptake in a dose-dependent manner in normal controls and in patients with NIDDM. Two possible mechanisms are identified: 1) a fat-related inhibition of glucose transport or phosphorylation, which appears after 3-4 h of fat infusion, and 2) a decrease in muscle glycogen synthase activity, which appears after 4-6 h of fat infusion. Third, FFAs stimulate insulin secretion in nondiabetic individuals. Some of this insulin is transmitted in the peripheral circulation and is able to compensate for FFA-mediated peripheral insulin resistance. FFA-mediated portal hyperinsulinemia counteracts the stimulation of FFAs on hepatic glucose production (HGP) and thus prevents hepatic glucose overproduction. We speculate that, in obese individuals who are genetically predisposed to develop NIDDM, FFAs will eventually fail to promote insulin secretion. The stimulatory effect of FFAs on HGP would then become unchecked, resulting in hyperglycemia. Hence, continuously elevated levels of plasma FFAs may play a key role in the pathogenesis of NIDDM in predisposed individuals by impairing peripheral glucose utilization and by promoting hepatic glucose overproduction.

Insulin resistance, characterized by reduced responsiveness to normal circulating concentrations of insulin, is a common feature of almost all patients with type II diabetes. The presumed central roles of both peripheral and hepatic insulin resistance suggest that the enhancement of insulin action might be an effective pharmacological approach to diabetes. Thiazolidinediones are a new class of orally active drugs that are designed to enhance the actions of insulin. These agents reduce insulin resistance by increasing insulin-dependent glucose disposal and reducing hepatic glucose output. Clinical studies in patients with type II diabetes, as well as other syndromes characterized by insulin resistance, have demonstrated that thiazolidinediones may represent a safe and effective new treatment. Although the precise mechanism of action of these drugs remains unknown, transcriptional changes are observed in tissue culture cells that produce enhanced insulin action. This regulation of gene expression appears to be mediated by the interactions of thiazolidinediones with a family of nuclear receptors known as the peroxisome proliferator-activated receptors (PPARs). The further elucidation of the molecular actions of these drugs may reveal much about the underlying mechanisms of insulin resistance.

The effect of insulin to acutely stimulate glucose uptake into muscle and adipose tissue is essential for normal glucose homeostasis. The GLUT4 glucose transporter is a major mediator of this action, and insulin recruits GLUT4 from an intracellular pool to the plasma membrane. An important pathologic feature of obesity, NIDDM, and to a lesser extent IDDM is resistance to insulin-stimulated glucose uptake. Investigations of the mechanisms have revealed tissue-specific regulation of GLUT4 with decreased gene expression in adipose cells but not in skeletal muscle. This has led to the hypothesis that alterations in the trafficking of the GLUT4 vesicle or in the exposure or activation of the GLUT4 transporter may cause insulin resistance in skeletal muscle in obesity and diabetes. Exercise training increases GLUT4 expression in muscle in association with enhanced glucose tolerance in vivo. Transgenic mice have been created to investigate other approaches to improve insulin action on glucose transport. Overexpression of GLUT4 in adipocytes of transgenic mice increases the proportion of GLUT4 on the plasma membrane and enhances insulin sensitivity in vivo. Altering insulin signaling by overexpressing p21ras in adipocytes of transgenic mice results in increased GLUT4 on the plasma membrane in the absence of insulin and increases insulin sensitivity in vitro and in vivo. Thus, glucose transport is a pivotal step in whole-body insulin action. Strategies to increase the number of GLUT4 transporters that are functionally inserted in the plasma membrane in muscle and adipocytes may lead to new therapies to treat or prevent NIDDM.

The role of T-cells in the pathogenesis of IDDM has been an area of much interest, and investigators have recently acquired new tools for studies on T-cells with the advent of T-cell clones that are reactive with islet antigens. Derived from NOD mice, diabetogenic T-cell lines and clones have for the most part been CD4+ and T-helper 1 (Th1)-like in their cytokine production. Some CD8+ cytotoxic clones have also been reported, although these have generally not transferred diabetes in the absence of CD4+ T-cells. The T-cell clones that have been described can also be separated on the basis of their antigen reactivity. While many of the T-cell lines and clones described react with islets, isolated islet cells, or islet membrane preparations, others have known antigen specificities, reacting with defined islet cell proteins such as insulin, GAD, and heat shock proteins. Particularly in the case of insulin-reactive clones, diabetogenicity has also been demonstrated. In light of the many possible T-cell reactivities that may arise from the islet lesion, the question of whether there is a dominant initiating antigen is a particularly intriguing one.

IDDM (type I diabetes) is generally believed to result from T-cell-mediated autoimmune destruction of the insulin-producing beta-cells in the pancreatic islets of Langerhans. In the last few years, considerable progress has been made with regard to the identification and characterization of candidate autoantigens recognized by autoantibodies; several of these candidate autoantigens are recognized by T-cells, including insulin, GAD65 and GAD67, heat-shock protein 65 (hsp65), and islet-cell antigen 69 (ICA69). In addition to these, a number of unidentified beta-cell antigens, including insulin-secretory granule membrane proteins and a 38-kDa protein, have been shown to stimulate T-cells of IDDM patients. However, T-cell autoreactivity to islet antigens is not specific for IDDM, and the T-cell target antigens are not specific for beta-cells. Moreover, the autoantigens involved in the initiation of the insulitis must be defined, and the mechanism of the T-cell-dependent beta-cell destruction remains to be unraveled. This review focuses on T-cell autoreactivity in IDDM in humans and the implications of the present knowledge for immunointervention and monitoring of immunotherapeutic trials.

Glucose is an important regulator of cell growth and metabolism. Thus, it is likely that some of the adverse effects of hyperglycemia are reflections of normal regulation by abnormal concentrations of glucose. How the cell senses glucose, however, is still incompletely understood. Evidence has been presented that the hexosamine biosynthesis pathway serves this function for regulation of aspects of glucose uptake, glycogen synthesis, glycolysis, and synthesis of growth factors. Excess hexosamine flux causes insulin resistance in cultured cells, tissues, and intact animals. Further evidence for the possible role of this pathway in normal glucose homeostasis and disease is that the level of activity of the rate-limiting enzyme in hexosamine synthesis, glutamine:fructose-6-phosphate amidotransferase, is correlated with glucose disposal rates (GDRs) in normal humans and transgenic mice.

In the present study, we have compared and analyzed published data related to the pathogenesis of the large vessel disease in diabetes. The prevailing opinion appears to be that diabetes accelerates the mechanism that leads to the development of classical atherosclerosis. However, as an alternative, we have amassed data that point to the presence of a diabetic macroangiopathy. This phenomenon comprises a constellation of nonatherosclerotic large vessel abnormalities. Today, we know that accumulation of periodic acid-Schiff (PAS)-positive material, as laminin, fibronectin, and type IV collagen, occurs together with hyaluronic acid and various types of connective tissue and calcium deposition. All these changes occur independent of the presence of atherosclerosis in the large vessels of diabetic patients. It seems to us that these observations emphasize that the concept of a specific diabetic macroangiopathy is a more fruitful working hypothesis than the usual theory of a link between atherosclerosis and diabetes. It provides a causal relationship (although the mechanism is unknown) between such changes and the abnormal metabolism in diabetes and a background for research strategy and tactics, aiming finally at the possibility of prevention and/or treatment of this common and dangerous disease.

Exposure of proteins to reducing sugars results in nonenzymatic glycation with the ultimate formation of advanced glycation end products (AGEs). One means through which AGEs modulate cellular functions is through binding to specific cell surface acceptor molecules. The receptor for AGEs (RAGE) is such a receptor and is a newly identified member of the immunoglobulin superfamily expressed on endothelial cells (ECs), mononuclear phagocytes (MPs), and vascular smooth muscle cells (SMCs) in both vivo and in vitro. Binding of AGEs to RAGE results in induction of cellular oxidant stress, as exemplified by the generation of thiobarbituric acid-reactive substances, expression of heme oxygenase type I, and activation of the transcription factor NF-kB, with consequences for a range of cellular functions. AGEs on the surface of diabetic red cells enhance binding to endothelial RAGE and result in enhanced oxidant stress in the vessel wall. By using reagents to selectively block access to RAGE, the role of this receptor in AGE-mediated perturbation of cellular properties can be dissected in detail.

Considerable interest has been focused in recent years on the mechanism of collagen cross-linking by high glucose in vitro and in vivo. Experiments in both diabetic humans and in animals have shown that over time collagen becomes less soluble, less digestible by collagenase, more stable to heat-induced denaturation, and more glycated. In addition, collagen becomes more modified by advanced products of the Maillard reaction, i.e., immunoreactive advanced glycation end products and the glycoxidation markers carboxymethyllysine and pentosidine. Mechanistic studies have shown that collagen cross-linking in vitro can be uncoupled from glycation by the use of antioxidants and chelating agents. Experiments in the authors' laboratory revealed that approximately 50% of carboxymethyllysine formed in vitro originates from pathways other than oxidation of Amadori products, i.e., most likely the oxidation of Schiff base-linked glucose. In addition, the increase in thermal stability of rat tail tendons exposed to high glucose in vitro or in vivo was found to strongly depend on H2O2 formation. The final missing piece of the puzzle is that of the structure of the major cross-link. We speculate that it is a nonfluorescent nonultraviolet active cross-link between two lysine residues, which includes a fragmentation product of glucose linked in a nonreducible bond labile to both strong acids and bases.

Advanced glycosylation end products (AGEs) form principally from the rearrangement of early glycation products, i.e., Amadori products, which produce a class of stable moieties that possess distinctive chemical crosslinking and biological properties. It has been generally believed that proteins with half-lives of longer than a few weeks are most susceptible to advanced glycosylation and that the highest levels of AGEs occur on proteins that comprise the long-lived structural components of connective tissue matrix and basement membrane.