    |
  ![[help]](/images/buttons/help.gif) |
 |
|
Fall 1998, Vol.5, No.2
THE RECEPTOR FOR AGE (RAGE): Ann Marie Schmidt, M.D.
Introduction The incidence of diabetes is on the rise throughout the world. In the United States alone, it is estimated that more than 12 million people suffer from this debilitating disorder.1 Diabetes represents a range of disorders typified by abnormal regulation of glycemia; largely mediated either by relative insulin deficiency or insulin resistance. Regardless of etiology, however, it is well-established that in diabetes, long-term complications ensue from abnormal regulation of glucose metabolism. Despite aggressive attempts at maintenance of strict euglycemia, however, diabetes often eventuates in accelerated dysfunction of vascular and inflammatory cells in persons with the disorder, resulting in a wide range of complications that increase morbidity and mortality. In the macrovasculature, for example, accelerated and aggressive atherosclerosis leads to the development of premature heart attacks and strokes.2-3 Microvascular disease may lead to the devastating complications of blindness and renal failure.4-5 In certain complications, such as impaired wound healing, impotence, and neuropathy, dysfunction of vascular, inflammatory and neural components is believed to contribute to progressive impairment of cellular structure and function.6-8 Furthermore, it is established that the prevalence and severity of periodontal disease is enhanced in patients with diabetes.9-10 Much attention has been focussed on dissecting the individual factors that lead to the development of complications in diabetes. Elevated levels of blood glucose have been linked to cellular perturbation. One consequence of hyperglycemia is the increased generation of sorbitol and fructose by the enzyme aldose reductase. Markedly increased production of sorbitol has been associated with the development of such complications in diabetes as retinopathy, neuropathy and nephropathy.11 Furthermore, the increased production of diacylglycerol in hyperglycemia has been linked to increased activation of the ß-isoform of protein kinase C, and subsequent alterations in cellular phenotype associated with vascular dysfunction.12-13 Our studies have focussed on the indirect sequelae of hyperglycemia; one consequence of which is the irreversible formation of Advanced Glycation Endproducts, or AGEs, which accumulate in diabetes. Interactions of these AGEs with their cellular receptor, Receptor for AGE (RAGE), may provide a novel and directed target for the prevention of the vascular and inflammatory cell complications that characterize diabetes. AGEs Form and Accumulate in Diabetes Upon exposure to aldose or reducing sugars, proteins and lipids undergo nonenzymatic glycation and oxidation; the early, reversible products of these interactions are known as Schiff bases/Amadori products. Measurement of such adducts of hemoglobin provides a measure of glycemic control over weeks to months, and is a helpful gauge for the clinician attempting to optimally manage a diabetic patient's glycemia.14 Our laboratory has focussed on the series of reactions downstream of these events; a heterogeneous set of reactions may eventuate in which the irreversible AGEs form. These include such species as pentosidine, carboxymethyllysine, pyralline and methylglyoxal.15-18 AGEs accumulate in the plasma and tissues in patients with diabetes and their presence has been linked to the pathogenesis of complications.19 It should be noted that AGEs also accumulate in a range of other settings including renal failure, amyloidoses, inflammation and aging. Alteration of cellular properties by AGEs. An important feature of AGEs is their ability to form cross-links, a process which may lead to changes in cellular structure and integrity. Such alterations may have important impact on cellular barrier function. Furthermore, nonenzymatic glycoxidation of basement membrane-associated structures may prevent their facilitation of cell attachment, and modification of growth factors may suppress their mitogenic activity.20-21 Such receptor-independent factors likely play important roles in the development of cellular perturbation, especially over long periods of time. However, it is likely that more temporally proximal events occur that initiate a series of cellular alteration steps which, we hypothesize, may trigger events leading to the development of vascular and inflammatory cell complications. Identification of Receptor for AGE (RAGE). It had been recognized that AGEs interact with cultured cells such as endothelial cells and monocytes in a manner consistent with interaction with specific cell surface binding sites. Using a strategy of AGE albumin radiolabelled with 125I, we purified to homogeneity two cell surface binding sites for AGEs from detergent extract of bovine lung. The first polypeptide was identical to lactoferrin, and the second was a newly-described protein we called Receptor for AGE (RAGE), a member of the immunoglobulin superfamily of cell surface molecules.22-23 The putative hydropathy plot of RAGE indicated that there was an 332-amino acid extracellular region composed of one "V"-type immunoglobulin domain followed by two "C"-type immunoglobulin domains. This was followed by a hydrophobic transmembrane spanning domain and lastly, by a highly-charged cytosolic domain (Fig. 1). The extracellular two-thirds of RAGE, soluble or sRAGE, was capable of binding AGEs and preventing their interaction with, and activation of, cellular receptors. While other cell surface interaction sites for AGEs have been identified, the best-characterized of these is RAGE.24-26
Figure 1. The predicted hydropathy plot of RAGE. Based on the cloning and identification of cDNA of RAGE, a hydropathy plot was generated using computer analysis. This plot predicts that there is a 332 amino acid extracellular domain composed of one "V"-type immunoglobulin domain, followed by two "C"-type immunoglobulin domains. This portion of the molecule is followed by a hydrophobic transmembrane spanning domain, and lastly, by a highly-charged cytosolic domain.
Figure 2. RAGE and AGEs co-localize in human diabetic vasculature. A and B demonstrate immunostaining of adjacent sections of a renal arterial vessel from a patient with diabetes for RAGE (A) and AGE (B) epitopes. C and D show immunostaining of adjacent sections of a renal arterial vessel from an age-matched normal control for RAGE (C) and AGE (D) epitopes. Compared with age-matched control individuals, the vasculature of patients with diabetes demonstrates increased deposition of AGEs and expression of RAGE. RAGE and Target Cell Dysfunction in Diabetes RAGE is present on cells targeted for dysfunction in diabetes. RAGE is present at low, homeostatic levels in a range of cell types such as endothelium, monocytes, vascular smooth muscle, mesangium and neurons. In diabetes, however, the expression of RAGE is enhanced, and strikingly co-localizes with that of AGE. For example, in diabetic vasculature, increased formation and deposition of AGEs in the endothelium and vascular smooth muscle co-localize with increased expression of RAGE27-30 (Fig. 2). These data strongly suggested the possibility that upregulation of RAGE in diabetic tissues might be linked to the development of vascular and inflammatory cell complications. Interestingly, enhanced expression of RAGE is also noted in the developing neurons of the central nervous system, and co-localizes with increased expression of the polypeptide amphoterin, with which RAGE interacts.31 Previous studies had indicated that amphoterin, at least in in vitro studies, mediates neurite outgrowth. When embryonic neurons were cultured on plastic dishes coated with amphoterin, incubation with anti-RAGE F(ab')2 or soluble RAGE inhibited neurite outgrowth.31 Taken together, these data suggested the possibility that neuronal RAGE may have an integral role in development of the nervous system. The interaction of AGEs with RAGE perturbs cellular function. RAGE interacts with a range of cell types important in the development of complications that typify diabetes. Endothelial cells. In homeostasis, the expression of endothelial RAGE is quite low; in states of perturbation, however, its expression is enhanced. AGEs, both those prepared in vitro and those derived from patients with diabetes, such as those found on the surface of red blood cells32 bound to cultured endothelial cells with Kd 50 nM, in a RAGE-dependent manner.22, 32 In all cases, the binding of these AGEs to endothelial cells was blocked in the presence of either anti-RAGE IgG, or in the presence of excess soluble RAGE (sRAGE; the extracellular two-thirds of RAGE).22 Suggestive of a role for AGE-endothelial RAGE interaction in the development of diabetic vascular complications were the findings that AGE-RAGE interaction resulted in enhanced endothelial permeability33 and increased production of Vascular Cell Adhesion Molecule-1 (VCAM-1), an adhesion molecule which mediates adherence of mononuclear inflammatory cells to stimulated endothelium via its monocyte counterligand VLA-4. Both processes were blocked in the presence of either anti-RAGE IgG or sRAGE.34 Mononuclear phagocytes. Inflammatory cells such as mononuclear phagocytes are important contributing factors to the development of vascular lesions, especially in the presence of increased cholesterol35 as well as important contributors in the inflammatory response. Interestingly, beyond diabetic atherosclerosis, monocyte binding to modified adducts such as AGE-ß2-microglobulin via RAGE,29, 36-37 may be important in the pathogenesis of bone and joint destruction observed in patients with dialysis-related amyloidosis.
Figure 3. Implantation of polytetrafluorethylene (PTFE) tubes with adsorbed AGE albumin into rats incites monocyte migration and activation. PTFE tubes impregnated with either AGE albumin (A) or native albumin (B) were implanted into the subcutaneous tissue of rats. The tubes were removed after 4 days and examined for the presence of infiltrating cells by hematoxylin and eosin staining. Our studies have indicated that AGEs bound to human peripheral blood-derived mononuclear phagocytes with Kd 50 nM. This binding was largely inhibited in the presence of RAGE blockade. The interaction of AGEs with monocyte RAGE resulted in enhanced chemotaxis and haptotaxis of monocytes; a process blocked in the presence of either anti-RAGE F(ab')2 or sRAGE.29, 38 In vivo, when polytetrafluoroethylene tubes were incubated with AGE rat serum albumin and implanted into the backs of rats, a florid attraction and deposition of mononuclear inflammatory cells ensued. In contrast, incubation of PTFE tubes with native rat serum albumin incited minimal inflammatory response, which was localized to the tissue-graft interface38 (Fig. 3). AGE-monocyte RAGE interaction also results in cellular activation, with implications for the development of inflammatory and vascular sequelae. Incubation of AGE-ß2-microglobulin with human monocytes resulted in increased elaboration of Tumor Necrosis Factor-a into cellular supernatants, a process inhibited in the presence of sRAGE.29 These data strongly suggested that the interaction of AGEs with monocyte RAGE might have implications for a range of circumstances in which AGEs form and accumulate, such as diabetes and renal failure. Smooth Muscle Cells. RAGE is present on the surface of vascular smooth muscle cells. The interaction of AGEs with cultured smooth muscle cells resulted in their increased migration and activation, as evidenced by enhanced production of factors such as Monocyte Chemoattractant Protein-1.39 In this context, smooth muscle cell RAGE has multiple implications in the context of vascular perturbation and injury. The interaction of AGEs with RAGE activates cell-signal
Figure 4. Ligation of RAGE by AGEs initiates a series of cell-signalling events. The interaction of AGEs with cellular RAGE results in activation of signalling events including p21ras, MAP kinase and NF-kB. Intermediate events in these cascades are under active investigation. We hypothesize that such events alter cellular properties in a manner predisposing to the development of vascular and inflammatory cell complications of diabetes and other disorders in which AGEs excessively form and accumulate. domain23 that, we hypothesize, initiates a series of cell signalling events that eventuate in altered gene expression. Previous studies indicated that an initial event in AGE-RAGE interaction was the generation of enhanced cellular oxidant stress,40 as AGE-RAGE interaction resulted in increased generation of thiobarbituric acid reactive substances, increased mRNA for heme oxygenase-1 and activation of the transcription factor NF-kB.40 Recently, we identified that intracellular events proximal to AGE-RAGE-mediated activation of NF-kB likely included activation of p21ras and MAP kinases41 (Fig. 4). Current studies are focused on delineating precise effector molecules with which the cytosolic domain of RAGE interacts in order to mediate these events. Taken together, these studies indicated that the ligation of RAGE by AGEs had the capacity to alter cellular properties in a manner predisposing to critical perturbation events favoring vascular and inflammatory cell dysfunction characteristic of the diabetic state. Is Antagonism of AGE-RAGE Interaction a Novel Therapeutic Target of Diabetes? The enhanced expression of RAGE in diabetic tissue, as well as the pathological sequelae of AGE-RAGE interaction in cells targeted for dysfunction in diabetes, such as endothelial cells, smooth muscle cells and monocytes, point to this interaction as a possible novel therapeutic target for the complications of diabetes. Other approaches to treatment and prevention of the complications of diabetes have included efforts such as rigorous euglycemic control, as exemplified by the results of the DCCT trials, in which strict control of hyperglycemia resulted in significantly diminished complications
Figure 5. Soluble RAGE (sRAGE) acts as a decoy to interrupt AGE-cellular target interactions. Soluble RAGE (sRAGE), the extracellular two-thirds of RAGE, binds AGEs and prevents their interaction with, and activation of, cellular surface targets such as RAGE. SRAGE may represent a novel, directed target for the prevention and treatment of complications in diabetes. in the microvasculature. However, merely a trend toward decreased complications of the macrovasculature was observed.42 Furthermore, it is clear that not all patients may be effectively or safely managed with rigorous maintenance of euglycemia. Other strategies, such as aldose reductase inhibitors11 and inhibitors of the ß-isoform of protein kinase C 12-13 have been suggested as means to inhibit the pathologic sequelae of the direct effects of elevated levels of blood glucose. Other strategies designed to diminish the formation of AGEs, such as use of aminoguanidine43-45 or agents designed to destroy AGE-crosslinks46 have further suggested the importance of AGEs as an important target in diabetes. A complementary approach, however, was clearly needed since use of agents such as aminoguanidine would have no impact on pre-formed/irreversible AGEs in diabetic tissues. Furthermore, there is no direct evidence to indicate that AGE-crosslinks are the only pathogenic features of these modified adducts. In that context, we postulated that soluble RAGE (sRAGE), the extracellular two-thirds of RAGE, may represent a novel target for the therapy of diabetic complications based on initial studies in in vitro binding and cell culture assays. Multiple studies suggested that the binding of AGEs (either those prepared in vitro or those derived from in vivo sources such as the urine of patients with renal failure) could be inhibited in the presence of excess soluble (s) RAGE.29, 33, 41, 47 Functionally, for example, excess sRAGE inhibited AGE-mediated mononuclear cell chemotaxis,38 increased expression of VCAM-1 in endothelial cells,34 hyperpermeability of endothelial cell monolayers,33 and increased elaboration of TNF-a into mononuclear cell supernatants.29 These data suggested that sRAGE bound AGEs and prevented their interaction with, and activation of, cell surface RAGE (Fig. 5). In in vivo studies, intravenous administration of AGE albumin into normal mice resulted in its rapid clearance from the circulation; this was in contrast to the markedly delayed clearance of radiolabelled albumin. These data suggested that AGE albumin was able to interact specifically with targets on the vessel wall, such as RAGE. Consistent with these observations and suggestive of a role for RAGE as a target in diabetes, the clearance of radiolabelled AGE albumin was significantly attenuated in the presence of sRAGE; however, consistent with its specificity for AGE-modified adducts, administration of sRAGE had no effect on the clearance of native albumin. In normal mice, administration of sRAGE inhibited AGE-mediated increased expression of IL-6 in liver tissue47 and suppressed vascular permeability mediated by infusion of syngeneic diabetic red blood cells.33 Taken together, these data suggested that sRAGE could interfere with the ability of AGEs to interact with and activate cellular targets. The critical test of this hypothesis, however, was whether sRAGE might exert protective effects in diabetic animals. To further explore these issues, we first characterized sRAGE prepared in a baculovirus expression system. After rat sRAGE was purified to homogeneity and rendered free of detectable levels of endotoxin, it was tested in diabetic rats upon radiolabelling with 125I. 125I-sRAGE was infused into the jugular veins of rats rendered diabetic with streptozotocin.33 In rats diabetic for 9-11 weeks, the t1/2 for elimination of sRAGE was 21.7 + 0.43 hours, compared with 13.6 + 0.79 hours in nondiabetic, age-matched controls.33 In both cases, the major organ in which sRAGE appeared to deposit was the kidney. An important marker of vascular dysfunction in diabetes is the presence of vascular hyperpermeability. The onset of microalbuminuria has been demonstrated epidemiologically to predict the development of clinically-relevant vascular disease and its related increased morbidity and mortality within years of the first demonstration of abnormal urinary excretion of albumin.48-49 To test this concept in diabetic rodents,we measured the tissue-blood isotope ratio. 125I-albumin and 51Cr-labelled red blood cells were utilized as a means to test vascular permeability, previously demonstrated to be elevated in spontaneously diabetic or streptozotocin-induced diabetic rats.50 In this model, as expected, vascular hyperpermeability was increased in a range of organs after 9-11 weeks of diabetes; particularly in the skin, intestine and kidney, compared with age-matched control rats. 33 At a lower dose of sRAGE (2.25 mg/kg; intravenously), hyperpermeability was blocked in diabetic intestine and skin, and largely blocked (approximately 60%) in diabetic kidney. However, at a higher dose of sRAGE, calculated to achieve expected plasma concentrations of 40-60 µg/ml, hyperpermeability in diabetic kidney was reversed by 90% (Fig. 6). These beneficial results occurred within one hour of administration of sRAGE. Together, these data provided our first experimental evidence that administration of sRAGE might impact on established diabetic complications; indeed, in this model, reversal of associated hyperpermeability in diabetic rats by sRAGE has broad implications, especially for microvascular kidney and retinal disease that may cause significant consequences in patients with diabetes. |
