Compartmentalization and quantitation of protein in human milk.

Human milk protein was determined by three colorimetric methods and by Kjeldahl analysis. The distribution of nitrogen (N) and protein was determined within various milk compartments. Total N, whey, casein, nonprotein nitrogen (NPN), cell N and N in the fat fraction were analyzed by micro-Kjeldahl analysis after a series of centrifugation and ultracentrifugation separations. Fresh milk samples (colostrum, transitional milk and mature milk) were centrifuged at 500 X g to separate milk cells and at 5000 X g to skim the milk. Decelled milk and skimmed milk were ultracentrifuged at 189,000 X g to separate fat and casein micelles from whey. NPN was analyzed after trichloroacetic acid precipitation. Whole milk, decelled milk, skimmed milk and whey were analyzed for protein with the Lowry method, modified for fat-containing samples, the Bradford dye-binding assay (Bio-Rad) and the Pierce bicinchoninic acid (BCA) assay. Cell nitrogen had a tendency to be lower in mature milk than in colostrum. Colostrum contained only 6% casein protein, whereas mature milk contained 13%. Fat from skimming was lower in N than fat from ultracentrifugation. Average NPN levels were similar for milk from all three lactation periods, and constituted 10% of colostrum N and 25% of mature milk N. Protein determined by the Bio-Rad method on whole milk samples had the lowest variability (square root MSE) when correlated to Kjeldahl values. All three assays had lower variability when analyzing whey and skimmed milk than when analyzing whole milk. The Lowry method and the Bio-Rad method had low variability for whey and skimmed milk samples, but the Lowry method yielded analytical values closest to Kjeldahl protein values. The BCA method consistently overestimated Kjeldahl protein by 30%.