Abstract

We have compared in vivo pyridoxine responsiveness with in vitro cystathionine beta-synthase activity in extracts of confluent fibroblasts from 14 synthase-deficient patients. Enzyme activity was measured with and without addition of its cofactor, pyridoxal-5'-phosphate, using a radioisotopic assay which detects as little as 0.25% of control activity. Six of seven lines from responsive patients had measurable activity without the added cofactor (0.6-15% of mean control). Two of these lines showed a five- and sevenfold stimulation of cystathionine beta-synthase activity with added pyridoxal-5'-phosphate; in the other four, the cofactor addition increased activity only modestly, as in controls. Two of seven lines from nonresponsive patients had measurable activity (each 3% of mean control) which increased two- and fivefold with the added cofactor. Cystathionine beta-synthase activity was undetectable in one line from a responsive patient and in five lines from nonresponsive ones. To characterize control and mutant synthase further, dissociation constants for pyridoxal-5'-phosphate were estimated and thermostability (54 degrees C) was studied in two control and five mutant lines. In one mutant, both parameters were normal; in the others, the affinity for the cofactor was reduced 3-to 11-fold and thermostability was much impaired. We conclude that at least three general classes of cystathionine beta-synthase mutants exist: those with no residual activity; those with reduced activity and normal affinity for pyridoxal-5' phosphate; and those with reduced activity and a reduced affinity for the cofactor. Pyridoxine responsiveness in vivo cannot be correlated simply with the presence or absence of residual synthase activity in vitro or with stimulation of in vitro enzyme activity by cofactor.