ADVANCED GLYCOSYLATED ENDPRODUCTS (AGES) IN NONDIABETIC PATIENTS UNDERGOING DIALYSIS

Renal failure is associated with the accumulation of advanced glycosyl ated endproducts as a consequence of an impaired renal clearance. We c onfirmed the finding that in sera of ESRD patients the AGE-beta(2)-mic roglobulin fraction is significantly elevated in contrast to healthy c ontrols but also to diabetic patients without dialysis. Furthermore, w e investigated the question, as to whether the continuously high gluco se concentration in the abdominal cavity of CAPD patients leads to a l ocal production of AGE. We found increased levels of AGE in the dialys ate of CAPD patients after long dwell times as compared to the serum c oncentrations. In immunohistochemical studies we proved the presence o f high amounts of AGE in the mesothelial and submesothelial tissue of patients on perennial CAPD. In in vitro experiments mesothelial cells, cultured on glyosylated collagen, showed a disturbed, net-like growth and an impaired matrix production. Further studies demonstrated that the disturbed matrix production was paralleled by an impaired producti on of the metalloproteinase gelatinase and its natural inhibitors TIMP -1 and TIMP-2. These alterations may (at least in part) be responsible for the long-term damage of the peritoneal membrane observed in CAPD patients. The extent of the non-enzymatic and irreversible glycosylati on of proteins (AGE synthesis) depends on the time and intensity of gl ucose exposure. II is well established that AGE accumulate with age [C erami et al. 1987] and in diabetic patients. AGE are supposed to cause tissue damage, like e.g. arteriosclerosis by their stimulation of gro wth factors and their influence on collagen synthesis [Vlassara et al. 1992]. In 1991 Makita described elevated serum AGE levels in patients undergoing dialysis and found a significant correlation between serum creatinine and AGE levels [Makita et al. 1991]. Hemodialysis lowered the AGE levels only transiently (for approximately three hours) even w hen high flux dialysis filters were applied. Normalization of renal fu nction by successful kidney transplantation, however, leads to a rapid fall of circulating AGE levels [Makita et al. 1994]. We measured seru m AGE concentrations with a competitive ELISA (polyclonal antibody aga inst AGE-bovine serum albumin) in four groups of patients: 1) Patients undergoing dialysis (HD and PD, n=7) with diabetes mellitus, 2) Patie nts undergoing dialysis (HD and PD, n=32) without diabetes mellitus, 3 ) Non-uremic diabetic patients (n=24), 4) Healthy people serving as a control (n=7). In order to determine glycosylated proteins of medium s ize molecular weight, especially the fraction of the AGE beta(2)-micro globulin, the fractions from 5,000 up to 20,000 kDa were purified by f iltration and in addition, a beta(2)-microglobulin-affinity chromatogr aphy was carried out. It became evident, that the AGE-beta(2)-microglo bulin fraction was significantly elevated only in patients undergoing dialysis (groups 1 and 2) in contrast to diabetic patients without dia lysis and the healthy control group (groups 3 and 4). These results co nfirm the findings of Miyata et al. and other investigators [Dolhofer- Bliesener et al. 1995] and lead to the hypothesis that accumulation of glycosylated beta(2)-microglobulin might be involved in the pathogene sis of dialysis-associated amyloidosis by cytokine stimulation. This w as further supported by Miyata et al. who found AGE beta(2)-microglobu lin in the amyloid of patients on dialysis [1994].In a further clinica l investigation we were interested in a special problem in patients tr eated with peritoneal dialysis: During CAPD there is a continuously hi gh glucose concentration in the abdominal cavity which might cause an AGE formation in the peritoneal cavity or the peritoneal membrane, res pectively AGE levels were measured in the serum and the dialysate (aft er 2 and 12 hours of incubation of the dialysate) of 15 non-diabetic P D patients. Tn two of these patients we prepared samples of the omentu m for immunohistochemical staining, gained during abdominal surgery. T he samples were screened for AGE products by incubating them with an A GE antibody and application of the avidin-biotin complex method. The r esults are means of two measurements in each of the 15 patients and ar e given in U per mg of total protein (1U equals 50% reduction of the A GE antibody binding in the ELISA). The serum showed a mean of 0.26+/-0 .12 U, the dialysate 0.04+/-0.01 U after 2 hours of incubation and 0.4 6+/-0.21 U after 10 hours. The difference between serum and dialysate levels (after 10 hours of incubation) was significant (p <0.02, Studen t's t-test). Immunhistochemical staining of the samples of the omentum showed AGE accumulation in the mesothelial and submesothelial interst itium of the peritoneal membrane. Our results prove an accumulation of AGE in the dialysate and the peritoneal membrane of CAPD patients. Si nce intraperitoneal AGE concentration (after long dwells) exceed serum AGE concentration, we postulate an increased AGE synthesis in the mes othelial and submesothelial tissue during CAPD and a ''washing out eff ect'' after long incubation periods. These high concentrations of loca l AGE could possibly cause damage to the peritoneal membrane, especial ly in long-term treatment. Further experiments were carried out to inv estigate possible influences of local AGE-accumulation on growth rate and matrix synthesis of peritoneal mesothelial cells. Human mesothelia l cells were grown on culture wells coated with glycosylated collagen (AGE collagen). Mesothelial cells risen on wells coated only with coll agen (without AGE) served as control. After 20 hours of culture signif icant differences in cell growth rate and matrix synthesis were observ ed under the microscope: Mesothelial cells grown on collagen showed th eir typical polygonal structure (''cobblestone appearance''), whereas cells grown on AGE collagen lost this structure and showed a longitudi nal, net-like growth. Similar influences of AGE on cell growth were fo und in cultures of synovial cells [Miyata 1994]. To get an idea of the reasons for this pathological growth rate and matrix synthesis, we me asured the gelatinase activity of the supernatants in the culture well s with the help of a zymogram assay. In addition we determined gelatin ase m-RNA with the PCR method. We found, that cells grown on AGE colla gen produced gelatinase, whereas cells cultivated on collagen without AGE did not. We got similar results for the gelatinase inhib