Hydroxamates as substrates and inhibitors for FMN-dependent 2-hydroxy acid dehydrogenases

Long-chain hydroxy acid oxydase (HAO) is a member of a flavoenzyme family with significant amino acid sequence similarity and strongly conserved three-dimensional structure; in particular, active-site amino acids involved in catalysis are invariant, with one exception, and numerous enzymatic studies suggest an identical chemical mechanism involving an intermediate carbanion for all family members. Known physiological substrates are a variety of L-2-hydroxy acids. Peroxisomal HAO differs from the other family members in that its actual physiological substrate is not known; it was first described as an L-amino acid oxidase, and recently was identified as an enzyme that converts creatol (hydroxycreatinine) to methylguanidine (a metabolite involved in a variety of uremic syndromes). Creatol (2-amino-5-hydroxy-1-methyl-4(5H)imidazolone) is not a 2-hydroxy acid. We show in this work that 2-hydroxyphenyl acetohydroxamate (HYPAH, the hydroxamate of mandelic acid), a compound that bears similarity both to mandelate (one of the best substrates known) and to creatol, is turned over by HAO, but between 10- and 100-fold less efficiently than mandelate itself. The compound also binds to the active site of homologous flavocytochrome b(2) (L-lactate dehydrogenase). Comparative pH-rate studies for mandelate and its hydroxamate suggest that HYPAH may bind in its ionized form. Both pH-rate profiles are bell-shaped curves, as are those determined for two other family members, flavocytochrome b(2) and mandelate dehydrogenase; while the group with an acid pK(a) between 5 and 6 is most likely the active-site histidine (the residue which abstracts the substrate C2 proton), the identity of the basic group is less clear. It has been proposed to be one of the active site arginines (Lehoux, I., and Mitra, B. (1999) Biochemistry38, 5836-5848); we suggest as an alternative that it could be the lysine residue that interacts with the flavin N1 and O2 positions and stabilizes the negative charge of reduced flavin. In addition to these studies, we have found that HAO is competitively inhibited by benzohydroxamate, which is one atom shorter than HYPAH; its affinity is nearly 100-fold lower than that of the substrate, in contrast to the strong inhibition it exerts on mandelate racemase (Maurice, St. M., and Bearne, S. L. (2000) Biochemistry39, 13324-13335). In the latter case, the 100-fold higher affinity compared to mandelate was proposed to arise from the fact that the hydroxamate can mimic the enolic intermediate which lies on the reaction pathway after C2 proton abstraction. Thus our results do not support the existence of a similar enolic intermediate for HAO (and probably its homologues), although they do not disprove it.