The enediyne antitumor antibiotic C-1027 contains an unusual (S)-3-chloro-4,5-dihydroxy-beta-phenylalanine moiety, which requires an aminomutase for its biosynthesis. Previously, we established that SgcC4 is an aminomutase that catalyzes the conversion of L-tyrosine to (S)-beta-tyrosine and employs 4-methylideneimidazole-5-one (MIO) at its active site [Christenson, S. D., Liu, W., Toney, M. D., and Shen, B. (2003) J. Am. Chem. Soc. 125, 6062-6063]. Here, we present a thorough analysis of the properties of SgcC4. L-Tyrosine is the best substrate among those tested and most likely serves as the in vivo precursor for the (S)-3-chloro-4,5-dihydroxy-beta-phenylalanine moiety. The presence of MIO in the active site is supported by several lines of evidence. (1) Addition of ATP or divalent metal ions has no effect on its aminomutase activity. (2) SgcC4 has optimal activity at pH approximately 8.8, similar to the pH optima of MIO-dependent ammonia lyases. (3) SgcC4 is strongly inhibited by sodium borohydride and potassium cyanide, but preincubation with L-tyrosine or 4-hydroxycinnamate largely prevents this inhibition. (4) The difference spectrum between SgcC4 and its S153A mutant shows a positive peak at approximately 310 nm, indicative of MIO. (5) The S153A mutation lowers k(cat)/K(M) 640-fold. The SgcC4-catalyzed conversion of L-tyrosine to (S)-beta-tyrosine proceeds via 4-hydroxycinnamate as an intermediate. The latter also acts as a competitive inhibitor with respect to L-tyrosine and serves as an alternative substrate for the production of beta-tyrosine in the presence of an amino source. A full time course for the SgcC4-catalyzed interconversion between L-tyrosine, beta-tyrosine, and 4-hydroxycinnamate was measured and analyzed to provide estimates for the rate constants in a minimal mechanism. SgcC4 also exhibits a beta-tyrosine racemase activity, but alpha-tyrosine racemase activity was not detected.

Alanine racemase catalyzes the pyridoxal phosphate-dependent interconversion of the D-and L-isomers of alanine. Previous studies have shown that the enzyme employs a two-base mechanism in which Lys39 and Tyr265 are the acid/base catalysts. It is thus possible that stereoisomerization of the external aldimine intermediates occurs through a concerted double proton transfer without the existence of a distinct carbanionic intermediate. This possibility was tested by the application of multiple kinetic isotope effect (KIE) methodology to alanine racemase. The mutual dependence of primary substrate and solvent deuterium KIEs has been measured using equilibrium perturbation-type experiments. The conceptually straightforward measurement of the substrate KIE in H 2 O is complemented with a less intuitive protium washout perturbation-type measurement in D 2 O. The primary substrate KIE in the D f L direction at 25°C is reduced from 1.297 in H 2 O to 1.176 in D 2 O, while in the L f D direction it is reduced from 1.877 in H 2 O to 1.824 in D 2 O. Similar reductions are also observed at 65°C, the temperature to which the Bacillus stearothermophilus enzyme is adapted. These data strongly support a stepwise racemization of stereoisomeric aldimine intermediates in which a substrate-based carbanion is an obligatory intermediate. The ionizations observed in k cat /K M pH profiles have been definitively assigned based on the ∆H ion values of the observed pK a 's with alanine and on the pH dependence of k cat /K M for the alternative substrate serine. The acidic pK a in the bell-shaped curve is due to the phenolic hydroxyl of Tyr265, which must be unprotonated for reaction with either isomer of alanine. The basic pK a is due to the substrate amino group, which must be protonated to react with Tyr265-unprotonated enzyme. A detailed reaction mechanism incorporating these results is proposed.

Free energy profiles for alanine racemase from Bacillus stearothermophilus have been determined at pH 6.9 and 8.9 from global analysis of racemization progress curves. This required a careful statistical design due to the problems in finding the global minimum in mean square for a system with eight adjustable parameters (i.e., the eight rate constants that describe the stepwise chemical mechanism). The free energy profiles obtained through these procedures are supported by independent experimental evidence: (1). steady-state kinetic constants, (2). solvent viscosity dependence, (3). spectral analysis of reaction intermediates, (4). equilibrium overshoots for progress curves measured in D(2)O, and (5). the magnitudes of calculated intrinsic kinetic isotope effects. The free energy profiles for the enzyme are compared to those of the uncatalyzed and the PLP catalyzed reactions. At pH 6.9, PLP lowers the free energy of activation for deprotonation by 8.4 kcal/mol, while the inclusion of apoenzyme along with PLP additionally lowers it by 11 kcal/mol.