CLIN. CHEM. 39/1, 93-96 (1993) CLINICAL CHEMISTRY, Vol. 39, No. 1, 1993 93 Quantification of p2-Microglobulinand Albumin in Plasma and Peritoneal Dialysis Fluid by a Noncompetitive Immunoenzymometric Assay A. Kandoussi, C. Cachera, D. Pagniez, R. Saile, and A. Tacquet An enzyme-linked immunosorbent assay (ELISA) was developed for measuring j32-microglobulin (/32m) and albumin in continuous ambulatory peritoneal dialysis (CAPD) fluid, Plasma concentrations of f32m were twofold greater in hemodialysis patients (41.3 ± 13.3 mg/L) than in CAPD patients (23.6 ± 5.5 mg/L) matched for duration of treatment. Measurement of p2m in CAPD fluid showed a substantial loss of this protein, -31% of total body 132m, compared with a 5% loss of a protein of middle molecular mass (albumin). Because of the molecular sieving effects of the peritoneal membrane, peritoneal clearance of $2m was sixfold greater than that of albumin. Whether /32m losses prevent or delay the incidence of dialysis-induced amyloidosis in these patients remains to be established. AddItional Keyphrases: enzyme-linked immunosorbent assay hemodialysis compared . proteins amyloidosis Albumin has been established as a sensitive indicator of renal glomerular injury (1) because its relative mo- lecular mass (Mr 68 000) only slightly exceeds the ex- clusion limit of glomerular basement membrane perme- ability. On the other hand, /32-microglobulin (/32m) is a low-Mr (11 800) protein associated with histocompatibil- ity antigens and is freely ifitered by the kidney (2,3).’ A normal kidney is able to reabsorb almost all of the 2m, which is then further catabolized by the proximal tubule (4). f32m in serum is increased in renal diseases that have reduced glomerular filtration (5). The interest in /32m has recently increased as the protein constituent of amyloid deposition in long-term hemodialysis patients (6, 7). Because 132m is not removed by conventional hemodialysis, it may accumulate in blood and tissues and induce the formation of amyloidosis. Such a risk has been encountered in long-term hemodialysis, but is only anecdotal in patients undergoing continuous ambula- tory pentoneal dialysis (CAPD) (8, 9). This study was designed to develop a sensitive enzyme-linked immuno- sorbent assay (ELISA) for measuring f32m and albumin in CAPD fluid without prior concentration and to com- pare the peritoneal clearance of these two proteins. Materialsand Methods Materials Microtiter plates were purchased from Costar, Cam- bridge, MA. Horseradish peroxidase (EC 1.11.1.7) was Laboratoire de Recherches Nephrologiques (Pr. A. Tacquet), H#{243}pital Calmette, Blvd. du Pr. Leclercq, 59037 Lille Cedex, France. ‘Nonstandard abbreviations: 2m, p2-microglobulin; CAPD, continuous ambulatory peritoneal dialysis; and ELISA, enzyme- linked immunosorbent assay. Received March 6, 1992; accepted August 28, 1992. from Boebringer Mannheim, Mannheim, Germany. Peroxidase substrate (o-phenylenediamine dihydrochlo- ride), gelatin, and pure human albumin (standard) were from Sigma Chemical Co., St. Louis, MO. Rabbit anti- human antibodies against fl2m and against albumin were obtained from Dako Ltd., Glostrup, Denmark. The protein content of the CAPD fluid was deter- mined by the method of Lowry et al. (10). A pool of plasma from healthy donorswas used as secondary standard for (32m. It was calibrated by ELISA against a purified 132m preparation from Calbiochem France-Bio- chem (Meudon, France). Procedures Preparation of the antibody-enzyme conjugate. Horse- radish peroxidase-labeled antibodies were prepared as described by Nakane and Kawaoi (11). We mixed 5 mg of peroxidase (Boehringer Mannheim; grade I, 250 kU/g of lyophilisate), dissolved in 1 mL of phosphate buffer (0.1 moliL, pH 6.8), with 1 mL of 0.3 mol/L NaHCO3 solution and 0.1 mL of 2,4-dinitrofluorobenzene solution (0.1 gIL in ethanol). After 2 h incubation, we added 1 mL of sodium periodate (80 mmol/L). The reaction was stopped after 1 h by adding 0.2 mL of redistilled glyc- erol; the solution was then dialyzed against 1L of sodium carbonate buffer (0.01 mollL, pH 9.5) and was mixed with antibody fractions (10 mg in 1 mL) in the same buffer. The mixture was incubated for 3 h at room temperature and dialyzed against 1L of phosphate buffer (0.1 molIL, pH 7.4). The conjugate solution was stored at -20 #{176}C in small portions after the addition of an equal volume of glycerol to each portion. Peroxidase substrate solution. We used o-phenylene- diamine dihydrochloride, 3 g/L, in 0.1 mol/L phosphate- citrate buffer (pH 5.5, and containing hydrogen perox- ide, 3.5 mmol/L) within 30 mm of preparation. Immunoassay. Coating; washing; addition of conju- gate, substrate, and HC1; and finally spectrophotometnc reading were performed by an automated ELISA Pro- cessor (Behring Institute, Marburg, Germany). To min- imize nonspecific binding to microtiter wells, we in- cluded in the assay buffer (for dilution of antigen or conjugate) 5 g of gelatin per liter. Microtiter plates were washed with 0.1 mol/L phosphate-buffered saline before they were coated with 100 L of antibody in each well by incubation overnight at room temperature. After four washes of the coated plate with phosphate-buffered saline, 100 L of each dilution of the samples and of the standards was added to the wells and incubated for 2h at 37#{176}C. The plates were washed and incubated again for 2 h at 37#{176}C with 100 ML of peroxidase-antibody conjugate per well. After washing, 100 pL of fresh