Sweet and sour: unraveling diabetic vascular disease.

Diabetes is the epidemic of the 21st century.1 Rare in the past, diabetes has grown into an increasingly common disease both in developed countries and most recently also in the third world. The most important factor for this unforeseen trend appears to be the increase in body weight around the world attributable to the changes in lifestyle over the last decades.2 It is likely that the increasing prevalence of diabetes will greatly affect the cardiovascular disease burden in the future. Although the morbidity and mortality of cardiovascular disease has fallen over the last 3 decades, this trend may flatten or even reverse. Thus, a better understanding of the consequences of diabetes in the vasculature and the heart is of great importance. Indeed, diabetes markedly affects the function of the cardiovascular system, both in the microcirculation as well as in large conduit arteries supplying vital organs such as the heart, brain and kidney.3 As a consequence, diabetes surpasses other conditions such as dyslipidemia and hypertension as a risk predictor for myocardial infarction, stroke, and renal failure4; in fact, because of its severe prognosis, diabetes may be considered the cancer of the vasculature. For vascular homeostasis, endothelial cells are of utmost importance.5 Indeed, these cells produce a variety of mediators, surface proteins, and autocoids involved in vasomotion, coagulation, and inflammation. In humans and animal models of diabetes in vivo as well as in endothelial cells in culture exposed to high glucose concentrations, marked functional changes are observed.3 A major mediator of endothelial function is nitric oxide (NO), which is synthesized from L-arginine via endothelial nitric oxide synthase (eNOS).5 In diabetes, endothelium-dependent relaxations are reduced both in humans and experimental animals.5 Surprisingly, though, early studies found an increased rather than a reduced expression of eNOS in human aortic endothelial cells in culture treated with high glucose and, later, also in arteries obtained from diabetic animals.6 Although this initially seemed paradoxical, it soon became clear that biologically active NO is chemically inactivated under these condition because of the production of reactive oxygen species (ROS). ROS and, in particular, superoxide (O2 ) rapidly interact with NO to form peroxynitrite, a highly reactive molecule that leads to nitrosylation of vital cellular proteins, among others prostacyclin synthase and superoxide dismutase, as well as DNA damage (Figure). Thus, endothelial dysfunction in diabetes is most likely related to these biochemical alterations: sweet and sour seems to underlie early diabetic vascular disease. The molecular and cellular mechanisms involved in glucose-induced endothelial dysfunction have been partially clarified over the last decade. It appears that high glucose concentrations activate protein kinase C and, specifically, its 2 isoform.7 Because glucose, but not mannitol, exerts the above mentioned effects on NO and ROS and also specifically activates protein kinase C 2 in endothelial cells, these effects appear to be unrelated to changes in osmolality, but rather reflect activation of distinct cellular pathways (Figure). Protein kinase C 2 has been implicated in the regulation and activation of membrane-associated NAD(P)H-dependent oxidases and subsequent production of superoxide anions and hence may initiate oxidative stress,8 which is a hallmark of diabetic vascular dysfunction. More recently, another pathway possibly involved in this context has been characterized. Indeed, p66shc is a mitochondrial adaptor protein involved in ROS metabolism and apoptosis.9 Interestingly, genetic deletion of p66shc in mice prolongs life span by approximately one-third, blunts cellular ROS production, reduces the cellular susceptibility to ROS, and prevents age-dependent endothelial dysfunction.10 Most strikingly, experiments in p66 knockout mice with streptozotocin-induced diabetes have revealed that endothelium-dependent relaxations in the aorta are preserved in spite of markedly elevated glucose levels in these animals. Concomitantly, in contrast to their wild-type littermates, ROS formation is not upregulated in diabetic p66 / mice, nor is ONOO formation increased. As a consequence, the amount of nitrosylated proteins remains at a normal level.11 Thus, these results suggest that glucose may not only activate protein kinase C 2 but also (possibly via the latter enzyme) other pathways such as p66shc to increase ROS production under diabetic conditions. ROS are known to activate transcription factors, such as nuclear factor B, that regulate inflammatory and procoagulant genes (Figure). These molecular events may well explain vascular dysfunction, inflammation, and disease associated with diabetes in patients. In this issue of Circulation Research, Romero et al expand on these concepts by their study on the effects of diabetes on arginase activity.12 L-Arginine is not only a substrate for eNOS but also for arginase. Arginase exists in 2 isoforms: arginase I is located in the cytoplasm and expressed abunThe opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association. From Cardiology, CardioVascular Center, University Hospital, Zurich; and Cardiovascular Research, Institute of Physiology, University of Zurich, Switzerland. Correspondence to Thomas F. Lüscher, MD, FRCP, FESC, Professor and Chairman of Cardiology and Head of Cardiovascular Research, University Hospital Zürich, Rämistrasse 100, CH-8091 Zürich, Switzerland. E-mail [email protected] (Circ Res. 2008;102:9-11.) © 2008 American Heart Association, Inc.