The purification of A/B transferase has been attempted since 1970. Sources rich in A/B transferase activity such as human milk and plasma were used. Although the attempts had some success in identifying the fractions that were enriched with enzyme activity, they were not pure. In 1990, Henrik Clausen, a colleague of mine at the now defunct Biomembrane Institute, obtained two fractions of pure proteins from human lungs that seemed to be the soluble form of A transferase. The reverse-phase chromatography used at the last step of the purification procedures denatured the isolated proteins, and no enzymatic activity was observed. The partial amino acid sequences of one of the proteins revealed that the protein was the cystic fibrosis antigen protein. Apparently, one or several of the patients whose lungs were used for the purification seemed to have suffered from the disease. The partial amino acid sequences for the other proteins were not found in the DNA and protein sequence databases. If the activity was not lost during the last step, the specific activity was calculated to be high (5.7 units/mg).