Yasumoto, K., H. Tsuji, and H. Mitsuda 1977 Isolation from rice bran of a bound form of vitamin B6 and its identification as 5'-O-(ß-D-glucopyranosyl)-pyridoxine. Agric. Biol. Chem. 41:1061–1067.
Zhang, Z., and D.B. McCormick 1992 Uptake and metabolism of N-(4'-pyridoxal)amines by isolated rat liver cells. Arch. Biochem. Biophys. 294:394–397.
Zhang, Z., J.F. Gregory III, and D.B. McCormick 1993 Pyridoxine-5'-ß-D-glucoside competitively inhibits uptake of vitamin B6 into isolated rat liver cells. J. Nutr. 123:85–89.
JOHN VANDERVEEN: Do you know if the traditional procedures for measuring nutrients in foods measure the glycosides as well? They are microbiological, I think, in both cases.
DONALD McCORMICK: Right. the answer to your question is "not uniformly." The results vary because the ability to measure glycosides depends on the nonspecific glucosidases that may or may not release it, and their presence varies from one microorganism to another. Some assays specifically do not do well; for instance, the streptococcal assays for B6 apparently do not have glucosidases that can release the B6 . So no, it is no good for assessing that.
It varies, so what you are dealing with, with any biological assay, including microbiological assays, are variations on competence for releasing bound forms, meaning that there is an error in that technique, too.