The use of (E)- and (Z)-phosphoenol-3-fluoropyruvate as mechanistic probes reveals significant differences between the active sites of KDO8P and DAHP synthases

The enzymes 3-deoxy-d-manno-2-octulosonate-8-phosphate (KDO8P) synthase and 3-deoxy-d-arabino-2-heptulosonate-7-phosphate (DAHP) synthase catalyze a similar aldol-type condensation between phosphoenolpyruvate (PEP) and the corresponding aldose: arabinose 5-phosphate (A5P) and erythrose 4-phosphate (E4P), respectively. While KDO8P synthase is metal-dependent in one class of organisms and metal-independent in another, only a metal-dependent class of DAHP synthases has thus far been identified in nature. We have used catalytically active E and Z isomers of phosphoenol-3-fluoropyruvate (E)- and (Z)-FPEP, respectively] as mechanistic probes to characterize the differences and/or the similarities between the metal-dependent and metal-independent KDO8P synthases as well as between the metal-dependent KDO8P synthase and DAHP synthase. The direct evidence of the overall stereochemistry of the metal-dependent Aquifex pyrophilus KDO8P synthase (ApKDO8PS) reaction was obtained by using (E)- and (Z)-FPEPs as alternative substrates and by subsequent (19)F NMR analysis of the products. The results reveal the si face addition of the PEP to the re face of the carbonyl of A5P, and establish that the stereochemistry of ApKDO8PS is identical to that of the metal-independent Escherichia coli KDO8P synthase enzyme (EcKDO8PS). In addition, both ApKDO8PS and EcKDO8PS enzymes exhibit high selectivity for (E)-FPEP versus (Z)-FPEP, the relative k(cat)/K(m) ratios being 100 and 33, respectively. In contrast, DAHP synthase does not discriminate between (E)- and (Z)-FPEP (the k(cat)/K(m) being approximately 7 x 10(-)(3) microM(-)(1) s(-)(1) for both compounds). The pre-steady-state burst experiments for EcKDO8PS showed that product release is rate-limiting for the reactions performed with either PEP, (E)-FPEP, or (Z)-FPEP, although the rate constants, for both product formation and product release, were lower for the fluorinated analogues than for PEP 125 and 2.3 s(-)(1) for PEP, 2.5 and 0.2 s(-)(1) for (E)-FPEP, and 9 and 0.1 s(-)(1) for (Z)-FPEP, respectively]. The observed data indicate substantial differences in the PEP subsites and open the opportunity for the design of selective inhibitors against these two families of enzymes.