Small Molecules Targeting Novel Mycobacterium tuberculosis Drug Targets

Abstract

Tuberculosis is the leading cause of fatality among all bacterial infections worldwide. The emergence of Mycobacterium tuberculosis (Mtb, the causative agent of tuberculosis) strains resistant to first and second line drugs is rapidly reducing the number of antibiotics available for clinical use, resulting in a desperate need for the discovery of new therapeutic agents. One important strategy is to target novel pathways for which there are no pre-existing resistance mechanisms. This thesis highlights strategies to selectively inhibit two novel Mtb drug targets: hydrolase important for pathogenicity 1 (Hip1), a serine protease critical for the virulence of the bacterium, and Mycobacterium tuberculosis dethiobiotin synthase (MtDTBS), a protein involved in the biosynthesis of biotin, a vitamin essential for the growth and pathogenicity of Mtb. Chapter One describes the structure and specificity of the target enzymes Hip1 and MtDTBS. Preliminary studies on the use of peptidomimetic α-keto esters as inhibitors of Hip1 and small molecule cyclopentyl fragments as binders of MtDTBS are also discussed. Chapter Two reports a library of compounds derived from the potent inhibitor 1.04 (Ki = 309 ± 15 pM) with reduced MW and or increased aqueous solubility. The reported library includes inhibitors with truncated sequences that contain either an N-terminal Cbz or pyrazine group. Additionally, a tripeptide derivative of 1.04 containing a 5-memebered lactam (a cyclic derivative of Gln) at P1 is also reported. The truncated inhibitors exhibit reduced inhibitory activity against Hip1 relative to 1.04, indicating that a tripeptide sequence is essential for potent inhibition, as evidenced by the tripeptide containing the Gln-lactam at P1 which showed comparable inhibitory potency (Ki = 126 ± 7 pM) relative to the parent compound 1.04. Chapter Three describes the structural characterisation of the lead tripeptide inhibitors containing either Leu (1.04) or Gln-Lac at P1, bound to Hip1, as determined by X-ray crystallography. Chapter Four focuses on the structure-guided chemical optimisation of the hit compound 2 (identified by in silico screening; KD = 3.4 ± 0.4 mM) targeting the DAPA binding site of MtDTBS. Attaching an acidic group to the pposition of the aromatic ring of the scaffold, as well as an additional carboxy group on the cyclopentyl ring, significantly improved binding affinity. Further optimisation led to tetrazole 7a as a particularly tight binder (KD = 57 ± 5 nM) and moderately potent inhibitor (Ki = 5 ± 1 μM) of MtDTBS. Chapter Five reports a new series of small molecule binders of MtDTBS derived from the lead compound 7a. Various tetrazoles were designed to investigate the role of charge delocalisation on binding to MtDTBS, while analogues containing related carboxylic acid bioisosteres were prepared to determine what other functionalities (aside from a tetrazole) at the p-position of the aromatic ring can improve binding. Methylated tetrazoles and the analogous benzyl tetrazole (of 7a) showed diminished binding affinity (KD = >100 μM) relative to 7a (KD = 57 ± 5 nM), while the deoxygenated derivative showed comparable activity (KD = 69 ± 4 nM), suggesting that delocalisation of a negative charge over the tetrazole and aryl groups of 7a is critical for tight binding to MtDTBS. Analogues containing bioisosteres displayed weak binding affinities (KD = ≥ 150 μM) to MtDTBS and this is attributed to the bioisosteres being too large to bind to the narrow terminus of the DAPA-pocket, as elucidated by X-ray crystallography. Analogues containing a phosphonate moiety designed to bind to the P-loop of MtDTBS were found to have weaker binding affinity relative to 7a (KD = 1.4 ± 0.03 and 1.4 ± 0.08), and this was deemed to be due to the bulky phosphonate ethyl esters binding away from the P-loop, as revealed by the co-crystal structure of one of these compounds bound to MtDTBS.