Formation of Immunochemical Advanced Glycosylation End Products Precedes and Correlates With Early Manifestations of Renal and Retinal Disease in Diabetes Paul J. Beisswenger, Zenji Makita, Thomas J. Curphey, Lynn L. Moore, Smith Jean, Truls Brinck-Johnsen, Richard Bucala, and Helen Vlassara Elevated levels of advanced glycosylation end products (AGEs) have been found in multiple tissues in associa- tion with diabetic vascular complications and during the microalbuminuric phase of diabetic nephropathy. In this study, we have used an AGE-specific enzyme-linked im- munosorbent assay (ELISA) to measure skin AGEs to determine whether elevated levels can be detected before the onset of overt microangiopathy. Subjects with type I diabetes (n = 48) were graded for the degree of nephrop- athy (normal [23], microalbuminuria [12], or macroalbu- minuria [12]) and retinopathy (none [13], background [20], or proliferative [15]). Subgroups with a premi- croalbuminuric phase of albumin excretion (^28 mg/24 h, n = 27) or with the earliest stages of retinopathy (ji = 27) were identified. A significant increase in tissue AGEs was found as urinary albumin increased during the premi- croalbuminuric phase of nephropathy even when the data were adjusted for age and duration of diabetes (P = 0.005). Immunoreactive AGEs also increased as normal renal status advanced to microalbuminuria and mac- roalbuminuria (P = 0.0001 across groups). Significant elevation of AGEs was also found in association with the earliest stages of clinically evident retinopathy (early background versus minimal grades). In addition, higher AGE levels were found in subjects with proliferative retinopathy when compared with those with less severe retinopathy (f* < 0.004 across groups). In contrast, no significant differences were found in tissue AGE levels between groups with or without early retinopathy based on pentosidine or fluorescent AGE measurements, al- though fluorescent AGEs correlated with albumin excre- tion. In conclusion, levels of collagen-linked AGEs, when measured by an AGE-specific ELISA, reveal a correlation with preclinical stages of diabetic nephropathy and early retinopathy not indicated by other methods and may prove useful as early markers of microangiopathy in type I diabetes. Diabetes 44:824-829, 1995 From the Section of Diabetes, Endocrinology and Metabolism (P.J.B.), Department of Medicine, Dartmouth Medical School, the Dartmouth-Hitchcock Medical Center, Hanover and Lebanon, New Hampshire; and the Picower Institute for Medical Research (R.B.,H.V.), Manhasset, New York. Address correspondence and reprint requests to Dr. Paul J. Beisswenger, Associate Professor of Medicine, Section of Endocrinology, Diabetes, and Metab- olism, Dartmouth-Hitchcock Medical Center, 1 Medical Center Drive, Lebanon, NH 03756. Received for publication 13 October 1994 and accepted in revised form 23 February 1995. AGE, advanced glycosylation end product; ANCOVA, analysis of covariance; ANOVA, one-way analysis of variance; ELISA, enzyme-linked immunosorbent assay; UAE, urinary albumin excretion. T he presence of advanced glycosylation end prod- ucts (AGEs) on vascular structures is believed to be important in the pathogenesis of diabetic mi- crovascular and macrovascular sequelae (1), par- ticularly in the progression of tissue damage leading to diabetic renal and retinal disease. A host of AGE-induced biochemical and biological responses have also been linked to the accelerated atherosclerotic damage found in patients with longstanding diabetes (2-5). Although a substantial body of basic experimental data supports the concept that AGEs play an important role in the development of diabetic sequelae, most studies conducted to date have primarily focused on subjects without detectable microangiopathy or with advanced diabetic sequelae, with the exception of one recent study that demonstrated in- creased levels of collagen-linked pentosidine and fluorescent AGEs in association with microalbuminuria (2). Studies on the relationship between tissue AGEs and diabetic vascular sequelae have also been impeded by the limited number of sensitive and specific assays. Several recent studies have shown that levels of a tissue-derived fluorescent AGE, pentosidine, and the glycoxidation product carboxymethyllysine are elevated in association with aging and diabetic vasculopathy (2,3,5-7). However, AGEs are known for their heterogeneity, and many of them do not exhibit fluorescence (8). Because of this, assays that mea- sure specific AGEs or that depend on fluorescence may not reflect total AGE accumulation in diabetes. Recently, using an AGE-specific immunoassay (9), tissue and serum AGEs, as well as those on hemoglobin and plasma lipoproteins (11,12), were quantified in diabetic patients with and without chronic renal failure (10). Further clinical studies are needed to investigate the value of this assay in evaluating the role of AGEs in the development of diabetic microvascular se- quelae. The goal of the current study was to examine the indepen- dent association between tissue levels of AGEs as deter- mined by this enzyme-linked immunosorbent assay (ELISA) and the earliest renal and retinal sequelae of diabetes by performing cross-sectional studies of a population with type I diabetes. We also have used this assay to determine the relationship between long-term glycemic control and AGEs to validate earlier studies that showed a positive relationship 824 DIABETES, VOL. 44, JULY 1995 Downloaded from http://diabetesjournals.org/diabetes/article-pdf/44/7/824/337784/44-7-824.pdf by guest on 04 November 2022