Design & synthesis of a novel EPSP synthase inhibitor based on its ternary complex with shikimate-3-phosphate and glyphosate

Lyme disease enolpyruvyl-UDP-GlcNAc synthase: fosfomycin-resistant MurA from Borrelia burgdorferi, a fosfomycin-sensitive mutant, and the catalytic role of the active site Asp

MurAs (enolpyruvyl-UDP-GlcNAc synthases) from pathogenic bacteria such as Borrelia burgdorferi (Lyme disease) and tuberculosis are fosfomycin resistant because an Asp-for-Cys substitution prevents them from being alkylated by this epoxide antibiotic. Previous attempts to characterize naturally Asp-containing MurAs have resulted in no protein or no activity. We have expressed and characterized His-tagged Lyme disease MurA (Bb_MurA(H6)). The protein was most soluble at high salt concentrations but maximally active around physiological ionic strength. The steady-state kinetic parameters at pH 7 were k(cat) = 1.07 ± 0.03 s(-1), K(M,PEP) = 89 ± 12 μM, and K(M,UDP-GlcNAc) = 45 ± 7 μM. Mutating the active site Asp to Cys, D116C, caused a 21-fold decrease in k(cat) and rendered the enzyme fosfomycin sensitive. The pH profile of k(cat) was bell-shaped and centered around pH 5.3 for Bb_MurA(H6), with pK(a1) = 3.8 ± 0.2 and pK(a2) = 7.4 ± 0.2. There was little change in pK(a1) with the D116C mutant, 3.5 ± 0.3, but pK(a2) shifted to >11. This demonstrated that the pK(a2) of 7.4 was due to D116, almost 3 pH units above an unperturbed carboxylate, and that it must be protonated for activity. This supports D116's proposed role as a general acid/base catalyst. As fosfomycin does not react with simple thiols, nor most protein thiols, the reactivity of D116C with fosfomycin, combined with the strongly perturbed pK(a2) for D116, strongly implies an unusual active site environment and a chemical role in catalysis for Asp/Cys. There is also good evidence for C115 having a role in product release. Both roles may be operative for both Asp- and Cys-containing MurAs.

Synergistic inhibitor binding to Streptococcus pneumoniae 5-enolpyruvylshikimate-3-phosphate synthase with both monovalent cations and substrate

The inhibitor binding synergy mechanism of the bi-substrate enzyme Streptococcus pneumoniae 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) has been investigated with a linkage thermodynamics strategy, involving direct binding experiments of one ligand conducted over a range of concentration of the other. The results demonstrate that binding of the inhibitor glyphosate (GLP) is highly synergistic with both a natural substrate shikimate-3-phosphate (S3P) and activating monovalent cations. The synergy between GLP and S3P binding was determined to be 1600-fold and is in qualitative agreement with previous work on Escherichia coli EPSPS. The binding molar ratios of S3P and GLP were measured as 1.0 and 0.7 per EPSPS, respectively. Monovalent cations that have been shown previously to stimulate S. pneumoniae EPSPS catalytic activity and its inhibition by GLP were found here to exhibit a similar rank-order with respect to their measured GLP binding synergies (ranging from 0 to > or =3000-fold increase in GLP affinity). The cation specificity and the sub-millimolar concentrations where these effects occur strongly suggest the presence of a specific cation binding site. Analytical ultracentrifugation data ruled out GLP-binding synergy mechanisms that derive from, or are influenced by, changes in oligomerization of S. pneumoniae EPSPS. Rather, the data are most consistent with an allosteric mechanism involving changes in tertiary structure. The results provide a quantitative framework for understanding the inhibitor binding synergies in S. pneumoniae EPSPS and implicate the presence of a specific cation binding regulatory site. The findings will help to guide rational design of novel antibiotics targeting bacterial EPSPS enzymes.

New aromatic inhibitors of EPSP synthase incorporating hydroxymalonates as novel 3-phosphate replacements

A new, aromatic analogue of the EPSP synthase enzyme reaction intermediate 1 has been identified, which contains a 3-hydroxymalonate moiety in place of the usual 3-phosphate group. This simplified inhibitor was readily prepared in five steps from ethyl 3,4-dihydroxybenzoate. The resulting tetrahedral intermediate mimic 9 is an effective, competitive inhibitor versus S3P with an apparent Ki of 0.57 +/- 0.06 microM. This result demonstrates that 3-hydroxymalonates exhibit potencies comparable to aromatic inhibitors containing the previously identified 3-malonate ether replacements and can thus function as suitable 3-phosphate mimics in this system. These new compounds provide another example in which a simple benzene ring can be used effectively in place of the more complex shikimate ring in the design of EPSP synthase inhibitors. Furthermore, the greater potency of 9 versus the glycolate derivative 10 and the 5-deoxy-analog 11, again confirms the requirement for multiple anionic charges at the dihydroxybenzoate 5-position in order to attain effective inhibition of this enzyme.

Ligand geometry of the ternary complex of 5-enolpyruvylshikimate-3-phosphate synthase from rotational-echo double-resonance NMR

The 46-kDa enzyme 5-enolpyruvylshikimate-3-phosphate (EPSP) synthase catalyzes the condensation of shikimate 3-phosphate (S3P) and phosphoenolpyruvate (PEP) to form EPSP. The reaction is inhibited by N-(phosphonomethyl)glycine (Glp), which, in the presence of S3P, binds to EPSP synthase to form a stable ternary complex. As part of a solid-state NMR characterization of this structure, we have used dipolar recovery at the magic angle (DRAMA) and rotational-echo double resonance (REDOR) to determine intra- and interligand internuclear distances. DRAMA was used to determine the single 31P-31P distance, while REDOR was used to determine one 31P-15N distance and five 31P-13C distances. These experimental distances were used as restraints in molecular dynamics simulations of an S3P-Glp complex to examine the geometry of the two ligands relative to one another in the ternary complex. The simulations were compared to unrestrained simulations of the EPSP synthase tetrahedral intermediate and its phosphonate analog. The results suggest that Glp is unlikely to bind in the same fashion as PEP, a conclusion that is consistent with recent studies that have questioned the role of Glp as a transition-state or intermediate analog.

An EPSP synthase inhibitor joining shikimate 3-phosphate with glyphosate: synthesis and ligand binding studies

A novel EPSP synthase inhibitor 4 has been designed and synthesized to probe the configurational details of glyphosate recognition in its herbicidal ternary complex with enzyme and shikimate 3-phosphate (S3P). A kinetic evaluation of the new 3-dephospho analog 12, as well as calorimetric and (31)P NMR spectroscopic studies of enzyme-bound 4, now provides a more precise quantitative definition for the molecular interactions of 4 with this enzyme. The very poor binding, relative to 4, displayed by the 3-dephospho analog 12 is indicative that 4 has a specific interaction with the S3P site. A comparison of Ki(calc) for 12 versus the Ki(app) for 4 indicates that the 3-phosphate group in 4 contributes about 4.8 kcal/mol to binding. This compares well with the 5.2 kcal/mol which the 3-phosphate group in S3P contributes to binding. Isothermal titration calorimetry demonstrates that 4 binds to free enzyme with an observed Kd of 0.53 +/- 0.04 microM. As such, 4 binds only 3-fold weaker than glyphosate and about 150-fold better than N-methylglyphosate. Consequently, 4 represents the most potent N-alkylglyphosate derivative identified to date. However, the resulting thermodynamic binding parameters clearly demonstrate that the formation of EPSPS x 4 is entropy driven like S3P. The binding characteristics of 4 are fully consistent with a primary interaction localized at the S3P subsite. Furthermore, (31)P NMR studies of enzyme-bound 4 confirm the expected interaction at the shikimate 3-phosphate site. However, the chemical shift observed for the phosphonate signal of EPSPS x 4 is in the opposite direction than that observed previously when glyphosate binds with enzyme and S3P. Therefore, when 4 occupies the S3P binding site, there is incomplete overlap at the glyphosate phosphonate subsite. As a glyphosate analog inhibitor, the potency of 4 most likely arises from predominant interactions which occur outside the normal glyphosate binding site. Consequently, 4 is best described as an S3P-based substrate-analog inhibitor. These combined results corroborate the previous kinetic model [Gruys, K. J., Marzabadi, M. R., Pansegrau, P. D., & Sikorski, J. A. (1993) Arch. Biochem. Biophys. 304, 345-351], which suggested that 4 interacts well with the S3P subsite but has little, if any, interaction at the expected glyphosate phosphonate or phosphoenolpyruvate-Pi subsites.

New EPSP synthase inhibitors: synthesis and evaluation of an aromatic tetrahedral intermediate mimic containing a 3-malonate ether as a 3-phosphate surrogate

A new analog of the EPSP synthase enzyme reaction intermediate 1, containing a 3-malonate ether moiety in place of the usual 3-phosphate group, was synthesized from 3,5-dihydroxybenzoic acid. This simple, synthetically accessible aromatic compound (5) is an effective competitive inhibitor versus S3P with an apparent Ki of 1.3 +/- 0.22 microM. This result demonstrates that a simple benzene ring can be a suitable achiral substitute for the more complex shikimate ring in the design of EPSP synthase inhibitors. Furthermore, the greater potency of 5 versus the phenol 6, glycolate 7 and the gallic acid analog 8 demonstrates the requirement for multiple anionic charges at the dihydroxybenzoate 5-position in order to attain effective inhibition of this enzyme. However, this 3-malonate ether substituted compound was at least 10-fold less effective as a bisubstrate inhibitor than the corresponding 3-phosphate. This suggests that tetrahedral intermediate mimics possessing a 3-malonate ether moiety are less effective than their corresponding 3-phosphates in accessing the optimal enzyme conformation stabilizing 1.