This chapter contains sections titled: Introduction Case 1: Human Protein Tyrosine Phosphatase Designing and Synthesizing Dual-site Inhibitors The Target Initial Leads Extension of the Initial Fragment Discovery and Incorporation of the Second Fragment The Search for Potency and Selectivity Finding More “Drug-like” Molecules Decreasing Polar Surface Area on Site 2 Monoacid Replacements on Site 1 Core Replacement Case 2: MurF Pre-filtering by Solution-phase NMR for Rapid Co-crystal Structure Determinations The Target Triage of Initial Leads Solution-phase NMR as a Pre-filter for Co-crystallization Trials Conclusion Acknowledgments References

The requirement for D-alanine in the peptidoglycan layer of bacterial cell walls is fulfilled in part by alanine racemase (EC 5.1.1.1), a pyridoxal 5'-phosphate (PLP)-assisted enzyme. The enzyme utilizes two antiparallel bases focused at the C, position and oriented perpendicular to the PLP ring to facilitate the equilibration of alanine enantiomers. Understanding how this two-base system is utilized and controlled to yield reaction specificity is therefore a potential means for designing antibiotics. Cycloserine is a known alanine racemase suicide substrate, although its mechanism of inactivation is based on transaminase chemistry. Here we characterize the effects of a Y265F mutant (Tyr265 acts as the catalytic base in the L-isomer case) of Bacillus stearothermophilus alanine racemase on cycloserine inactivation. The Y265F mutant reduces racemization activity 1600-fold [Watanabe, A., Yoshimura, T., Mikami, B., and Esaki, N. (1999) J. Biochem. 126, 781-786] and only leads to formation of the isoxazole end product (the result of the transaminase pathway) in the case Of D-cycloserine, in contrast to results obtained using the wild-type enzyme. L-Cycloserine, on the other hand, utilizes a number of alternative pathways in the absence of Y265, emphasizing the importance of Y265 in both the inactivation and racemization pathway. In combination with the kinetics of inactivation, these results suggest roles for each of the two catalytic bases in racemization and inactivation, as well as the importance of Y265 in "steering" the chemistry to favor one pathway over another.

The D-Ala-D-Ala adding enzyme (MurF) from Streptococcus pneumoniae catalyzes the ATP-dependent formation of the UDP-MurNAc-pentapeptide, a critical component of the bacterial cell wall. MurF is a potential target for antibacterial design because it is unique to bacteria and performs an essential non-redundant function in the bacterial cell. The recent discovery and subsequent cocrystal structure determination of MurF in complex with a new class of inhibitors served as a catalyst to begin a medicinal chemistry program aimed at improving their potency. We report here a multidisciplinary approach to this effort that allowed for rapid generation of cocrystal structures, thereby providing the crystallographic information critical for driving the inhibitor optimization process. This effort resulted in the discovery of low-nanomolar inhibitors of this bacterial enzyme.

In a broad genomics analysis to find novel protein targets for antibiotic discovery, MurF was identified as an essential gene product for Streptococcus pneumonia that catalyzes a critical reaction in the biosynthesis of the peptidoglycan in the formation of the cell wall. Lacking close relatives in mammalian biology, MurF presents attractive characteristics as a potential drug target. Initial screening of the Abbott small-molecule compound collection identified several compounds for further validation as pharmaceutical leads. Here we report the integrated efforts of NMR and X-ray crystallography, which reveal the multidomain structure of a MurF-inhibitor complex in a compact conformation that differs dramatically from related structures. The lead molecule is bound in the substrate-binding region and induces domain closure, suggestive of the domain arrangement for the as yet unobserved transition state conformation for MurF enzymes. The results form a basis for directed optimization of the compound lead by structure-based design to explore the suitability of MurF as a pharmaceutical target.

Alanine racemase (EC 5.1.1.1) catalyzes the interconversion of alanine enantiomers, and thus represents the first committed step involved in bacterial cell wall biosynthesis. Cycloserine acts as a suicide inhibitor of alanine racemase and as such, serves as an antimicrobial agent. The chemical means by which cycloserine inhibits alanine racemase is unknown. Through spectroscopic assays, we show here evidence of a pyridoxal derivative (arising from either isomer of cycloserine) saturated at the C4' carbon position. We additionally report the L- and D-cycloserine inactivated crystal structures of Bacillus stearothermophilus alanine racemase, which corroborates the spectroscopy via evidence of a 3-hydroxyisoxazole pyridoxamine derivative. Upon the basis of the kinetic and structural properties of both the L- and D-isomers of the inhibitor, we propose a mechanism of alanine racemase inactivation by cycloserine. This pathway involves an initial transamination step followed by tautomerization to form a stable aromatic adduct, a scheme similar to that seen in cycloserine inactivation of aminotransferases.

The stability of insulinotropin was evaluated as a function of pH and temperature in the present study so as to predict stability at ambient and subambient temperatures. For traditional Arrhenius analysis, there is generally a loss of statistical information for the estimates of the energy of activation (E(a)) and the frequency factor (A) due to the use of two separate regression steps. Therefore, a comparison was made between traditional Arrhenius analysis and a non-linear fitting approach utilizing log (C/C-0) = -0.434 t A exp(-E(a)/RT) where C is the concentration, 0 denotes the initial value, t is time, T is temperature, and R is the gas constant. A good agreement was found in the estimates by the two methods; however, the nonlinear method did provide an improvement in confidence limits under some conditions. Arrhenius plots for the degradation of the peptide appear to obey the Arrhenius equation from 5 to 50 degrees C under alkaline and acidic conditions; however, some deviation was observed at neutral pH. The analysis predicts that at 25 degrees C, maximal stability for the peptide is at approximately pH 8. Both methods predict 10% loss of potency in 4-5 months at pH 7 and 25 degrees C, which appears to be reasonably consistent with ongoing real time studies.

Transforming growth factor β3 (TGF-β3) is a dimeric protein (Mr 25 444) which regulates the proliferation of epithelial cells. A reversed-phase high-performance liquid chromatography (RP-HPLC) method was developed for TGF-β3 in order to investigate the solution stability of TGF-β3. Peak width and retention time were evaluated as a function of (1) the initial percent of acetonitrile, (2) the gradient slope, (3) the percentage of trifluoracetic acid, and (4) column temperature. By varying the above conditions, particularly temperature, it was possible to reduce retention times and peak widths by a factor of 2 and 6, respectively. The observed temperature effects on peak width were consistent with the concept that conformational changes can be induced in the protein during the chromatographic migration. A balance was found between on-column degradation and improved peak width at elevated temperatures. The final assay was found to be linear over a 10–150 μg/ml drug range with a correlation coefficient of 0.998 and acceptable reproducibility (R.S.D. 2.09%, n = 6). The assay has been useful for monitoring drug loss in ongoing stability studies; however, it was found to have poor resolving power when certain stressed samples of TGF-β3 were applied, suggesting that complimentary assay methodologies must be developed.