Effects of low and high relative molecular protein mass on four methods for total protein determination in urine.

Four commonly used methods for the determination of total protein in urine were compared. These were two biuret methods using different precipitants, a Ponceau S method and a Coomassie Brilliant Blue method. The protein content of the urines was also evaluated by sodium dodecylsulphate polyacrylamide gel electrophoresis. The biuret method with ethanolic phosphotungstic acid as precipitant correlated best with the Coomassie Brilliant Blue method (r = 0.944; p less than 0.001) but less well with the Ponceau S (r = 0.895; p less than 0.001) or biuret-trichloroacetic acid (r = 0.874; p less than 0.001) methods. For urines with normal electrophoretic protein patterns, the imprecise biuret-trichloroacetic acid method (cv = 18.5%) gave the greatest number of false high results (23 in 36 urines) as assessed by electrophoresis. False low results were common in low relative molecular mass (Mr) proteinuria, especially with the biuret-tricholoroacetic acid and Ponceau S methods. High Mr proteinuria rarely caused false low results. Discrepancies between methods appear to have resulted from incomplete precipitation of low Mr protein by trichloroacetic acid.