Date of Award:


Document Type:


Degree Name:

Doctor of Philosophy (PhD)


Nutrition, Dietetics, and Food Sciences

Committee Chair(s)

Deloy G. Hendricks


Deloy G. Hendricks


Deborah R. Campbell


Nedra K. Christensen


LeGrand Ellis


Ronald Munger


Early introduction of cow's-milk-based infant formulas, in particular the ABBOS epitope of bovine serum albumin (BSA), has been implicated as an autoimmune trigger in the pathogenesis of insulin dependent diabetes mellitus (IDDM). A direct enzyme-linked immunosorbant assay (ELISA), using polyclonal anti-BSA antibodies, was developed to determine the BSA content of cow's milk and 15 infant formulas. Powdered high-whey (60%) formulas averaged 41 mg BSA/100 ml; 2% milk contained 52 mg BSA/100 ml; and the high-casein formulas averaged 13 mg/100 ml. BSA content of powdered polymeric formulas and cow's milk varied directly with the whey protein concentration (correlation coefficient = 0.8445, Q = 0.008). BSA was not detected in any hydrolyzed powdered formula or commercially sterile liquid preparation regardless of protein composition. The absence of BSA was confirmed by polyacrylamide gel electrophoresis. It is unlikely that the ABBOS epitope is present in the formulas testing negative for BSA due to enzymatic hydrolysis and heat denaturation of these formula preparations.

A laboratory technology was developed that could be upgraded to produce BSA free protein bases used in the manufacture of infant formula. Affinity chromatography, using paramagnetic beads with an immobilized antibody against BSA, was applied to extract BSA from cow's milk and whey isolates. Monoclonal and polyclonal antibody-activated beads were used to capture BSA from samples. The capture efficiency in milk was 11% and 19% for polyclonal beads, and 59% for monoclonal beads. Capture efficiency of monoclonal beads of 91% was significantly greater in both acid and sweet whey compared to the polyclonal beads exhibiting a capture efficiency 31% and 24% in acid and sweet whey, respectively. Capture efficiency of monoclonal and polyclonal beads did not differ significantly in milk, acid whey, or sweet whey. Removal of BSA from a known sample of 25ng of BSA treated with polyclonal beads was 70% effective with a capture efficiency of 35%. A net reduction of 99.9% of the BSA could be expected by coupling immunocapture with molecular sieving. Immunocapture was most effective in removing BSA when only small amounts were present in the sample.