BACKGROUND: There has been much recent interest in accumulation of advanced glycation end-products (AGE) in uraemic patients. Analysis of AGE has been difficult, because commonly used methodologies, i.e. immunodetection assays or fluorescence measurements, reflect group reactivity and are not specific for chemically defined substances. Some investigators measured individual AGE compounds, e.g. pentosidine, carboxymethyllysine, pyrraline or imidazolone, but a systematic assessment of known compounds using specific HPLC methods in diabetic and non-diabetic end-stage renal disease (ESRD) patients during treatment has not been performed. METHODS: For the present study, the concentrations of early and late products of the Maillard reaction in plasma and ultrafiltrate were monitored during high-flux dialysis sessions in diabetic and non-diabetic patients. AGE were analysed by fluorescence spectroscopy and size exclusion chromatography with fluorescence detection. Specific HPLC methods were used to quantify the Amadori product fructoselysine and the AGE compounds pentosidine and pyrraline in acid or enzymatic hydrolysates. RESULTS: Using size exclusion chromatography, we confirmed a similar fluorescent peak distribution for diabetic and non-diabetic ESRD patients. Main fractions were found at approximately 70, approximately 14 and <2 kDa, confirming results obtained by other authors. In diabetic patients, the fluorescence intensity of the low molecular weight fraction was higher. Uraemic patients differed from controls mainly by the fluorescence of the low molecular weight fraction. The peak spectrum in ultrafiltrates was similar to that in plasma regarding low molecular weight fractions and the 14 kDa peak, but no protein-bound fluorescence was found at 70 kDa. HPLC analysis revealed a significant reduction of plasma pentosidine during high-flux dialysis in non-diabetic (from 9.1+/-5.1 to 8.5+/-4.7 pmol/mg protein; P<0.05) and diabetic patients (from 10.0+/-9.1 to 6.8+/-4.0 pmol/mg protein; P<0.05). In contrast, plasma fructoselysine showed only a non-significant trend to decrease in diabetic (from 3.24+/-0.88 to 3.05+/-0.77 nmol/mg protein) and non-diabetic patients (from 2.69+/-0.52 to 2.56+/-0.50 nmol/mg protein). Pyrraline, a nonfluorescent late AGE product derived from reaction of 3-deoxyglucosone with lysine, could not be detected (detection limit approximately 40 pmol/mg protein). Comparing HPLC and size exclusion analysis, it was found that pentosidine accumulated in the range of low molecular weight substances and was removed by high-flux dialysis. CONCLUSIONS: High-flux dialysis reduces the plasma concentration of fluorescent AGE compounds, i.e. pentosidine, but the Amadori product fructoselysine is not removed, indicating that this compound is protein associated.
BACKGROUND: There has been much recent interest in accumulation of advanced glycation end-products (AGE) in uraemic patients. Analysis of AGE has been difficult, because commonly used methodologies, i.e. immunodetection assays or fluorescence measurements, reflect group reactivity and are not specific for chemically defined substances. Some investigators measured individual AGE compounds, e.g. pentosidine, carboxymethyllysine, pyrraline or imidazolone, but a systematic assessment of known compounds using specific HPLC methods in diabetic and non-diabetic end-stage renal disease (ESRD) patients during treatment has not been performed. METHODS: For the present study, the concentrations of early and late products of the Maillard reaction in plasma and ultrafiltrate were monitored during high-flux dialysis sessions in diabetic and non-diabeticpatients. AGE were analysed by fluorescence spectroscopy and size exclusion chromatography with fluorescence detection. Specific HPLC methods were used to quantify the Amadori product fructoselysine and the AGE compounds pentosidine and pyrraline in acid or enzymatic hydrolysates. RESULTS: Using size exclusion chromatography, we confirmed a similar fluorescent peak distribution for diabetic and non-diabetic ESRDpatients. Main fractions were found at approximately 70, approximately 14 and <2 kDa, confirming results obtained by other authors. In diabeticpatients, the fluorescence intensity of the low molecular weight fraction was higher. Uraemic patients differed from controls mainly by the fluorescence of the low molecular weight fraction. The peak spectrum in ultrafiltrates was similar to that in plasma regarding low molecular weight fractions and the 14 kDa peak, but no protein-bound fluorescence was found at 70 kDa. HPLC analysis revealed a significant reduction of plasma pentosidine during high-flux dialysis in non-diabetic (from 9.1+/-5.1 to 8.5+/-4.7 pmol/mg protein; P<0.05) and diabeticpatients (from 10.0+/-9.1 to 6.8+/-4.0 pmol/mg protein; P<0.05). In contrast, plasma fructoselysine showed only a non-significant trend to decrease in diabetic (from 3.24+/-0.88 to 3.05+/-0.77 nmol/mg protein) and non-diabeticpatients (from 2.69+/-0.52 to 2.56+/-0.50 nmol/mg protein). Pyrraline, a nonfluorescent late AGE product derived from reaction of 3-deoxyglucosone with lysine, could not be detected (detection limit approximately 40 pmol/mg protein). Comparing HPLC and size exclusion analysis, it was found that pentosidine accumulated in the range of low molecular weight substances and was removed by high-flux dialysis. CONCLUSIONS: High-flux dialysis reduces the plasma concentration of fluorescent AGE compounds, i.e. pentosidine, but the Amadori product fructoselysine is not removed, indicating that this compound is protein associated.