Shotgun Kinetic Target-Guided Synthesis Approach Enables the Discovery of Small-Molecule Inhibitors against Pathogenic Free-Living Amoeba Glucokinases

Abstract

Pathogenic free-living amoebae (pFLA) can cause life-threatening central nervous system (CNS) infections and warrant the investigation of new chemical agents to combat the rise of infection from these pathogens. Naegleria fowleri glucokinase (NfGlck), a key metabolic enzyme involved in generating glucose-6-phosphate, was previously identified as a potential target due to its limited sequence similarity with human Glck (HsGlck). Herein, we used our previously demonstrated multifragment kinetic target-guided synthesis (KTGS) screening strategy to identify inhibitors against pFLA glucokinases. Unlike the majority of previous KTGS reports, our current study implements a "shotgun" approach, where fragments were not biased by predetermined binding potentials. The study resulted in the identification of 12 inhibitors against 3 pFLA glucokinase enzymes─NfGlck, Balamuthia mandrillaris Glck (BmGlck), and Acanthamoeba castellanii Glck (AcGlck). This work demonstrates the utility of KTGS to identify small-molecule binders for biological targets where resolved X-ray crystal structures are not readily accessible.

Keywords: glucokinases; kinetic target-guided synthesis; multifragment screening; small molecule inhibitors; sulfo-click reaction.

Figure 1

Schematic representation of the multifragment screening using KTGS. First, the complementary reactive fragments are incubated in a 96-well plate in the absence (red) and presence (blue) of the target. Fragments that bind to the target and lie in close proximity to one another irreversibly react to form a multidentate ligand. All incubations are then subjected to selected reaction monitoring (SRM) LC–MS/MS analysis, where ligated products are deemed as hits if their concentrations are amplified in the presence of the target when compared to the absence of the target.

Figure 2

Cluster map of N-acylsulfonamide (SZTA) hits from the KTGS screen against NfGlck. Compounds were clustered using Murcko frameworks and ECFP4 fingerprints with a Tanimoto similarity cutoff of 0.9. The white numbers in dark purple boxes correspond to the cluster number, while black numbers in light-purple boxes correspond to the number of cluster members. Compounds are colored according to CNS MPO scores, which range from 0 to 6, with red compounds indicating a lower CNS MPO score and blue compounds possessing a higher score.

Scheme 1. Synthesis of the N-Acylsulfonamides by (a) EDCI-Promoted Coupling; (b) Reaction of Selenocarboxylates, Generated from Carboxylic Acids, with Sulfonyl Azides

Figure 3

Structures of the synthesized methylated analogues. SZ and TA methylated analogues retain sulfonyl azide and thio acid functionalization, respectively, while the opposite portion of the N-acylsulfonamide core is substituted with a methyl group.

Figure 4

Histogram charts depicting the (a) central nervous system multiparameter optimization (CNS MPO) scores, (b) molecular weight (MW), (c) calculated octanol/water partition coefficient at pH 7 (c Log D), and (d) calculated octanol/water partition coefficient (c Log P) profile of the enumerated set of 1710 N-acylsulfonamides (SZTAs). All physicochemical values were calculated using Jchem for Excel (Chemaxon), and histograms were generated with StarDrop (Optibrium).

Figure 5

Cluster maps of sulfonyl azide (SZ) and thio acid (TA) fragment libraries. The libraries were clustered using Murcko frameworks and ECFP4 fingerprints with a Tanimoto similarity threshold of 0.8.