Uncovering the Unusual Inhibition Mechanism of a Trypanosome Alternative Oxidase Inhibitor Displaying Broad-Spectrum Activity against African Animal Trypanosomes

Abstract

The glucose-dependent respiration of bloodstream forms of the parasite Trypanosoma brucei depends on an unusual and essential mitochondrial electron-transport system, consisting of glycerol-3-phosphate dehydrogenase and the trypanosome alternative oxidase (TAO). We report here the discovery of an allosteric inhibitor of TAO that displays highly potent activity (EC₅₀ values in the range 1-20 nM) against the important veterinary pathogens T. b. brucei, Trypanosoma evansi, Trypanosoma equiperdum, and Trypanosoma congolense, i.e., >5-fold greater potency than the standard drugs. The methylene-linked 2-methyl-4-hydroxybenzoate 2-pyridinyldiphenylphosphonium derivative (1) was the best inhibitor of recombinant TAO (IC₅₀ = 1.3 nM) via a noncompetitive/allosteric mechanism (K_i = 3.46 nM). Remarkably, X-ray crystallography showed that 1 was bound to a site of TAO ∼25 Å from the catalytic pocket. Although 1 demonstrated good safety toward mammalian cells in vitro (selectivity index >2300), it did not fully clear parasitemia in experimental animals, attributable to a high hepatic clearance.

1. (A) Historical Drugs Used for the Treatment of Gambiense and Rhodesiense HAT. (B) New Orally Active Drug Registered for Treating Both Stages of Gambiense HAT. (C) Structurally Related Early (Micromolar) TAO Inhibitors: Salicylhydroxamic Acid (SHAM) and 2,4-Dihydroxybenzoate (2,4-DHB) Derivatives. (D) Mitochondrion-Targeted (Nanomolar) Allosteric TAO Inhibitor, Compound 1

1

IC₅₀ values against rTAO and SAR of triarylphosphonium TAO inhibitors (PDPP = 2-pyridinyldiphenylphosphonium, TPP = triphenylphosphonium). Compound 1 is the most potent TAO inhibitor among all the phosphonium series we have reported so far. − ,, Reference compound: ascofuranone, IC₅₀ = 2 nM (Table ). All IC₅₀ values are the average and SEM of 4 independent determinations.

1. Synthesis of Compounds 1–12

2

Growth curve of T. brucei wild type s427 grown in either the continuous presence of 1 at indicated concentrations (filled symbols) or with the drug washed out with fresh media after 24 h (open symbols; the arrows indicate the time when compound 1 was withdrawn), or untreated (no drug control). The results presented are those of three biological repeats. Error bars represent SD of the biological replicates.

3

Assessment of DNA configuration. (A) Percentage of cells at various cell division stages in populations treated or not treated with 2 × EC₅₀ over a 24-h period. The percentages are the average of three independent determinations and SEM, obtained using flow cytometry after cell permeabilization and staining with propidium iodide. (B) DNA content of cells treated over a 24 h period with compound 1 at 2 × EC₅₀ as determined by fluorescence microscopy following DAPI staining. N, nuclear DNA; K, kinetoplast DNA. The percentages are the average of three independent determinations and SEM.

4

Inhibition mechanism and potency of compound 1 against TAO. (A) Lineweaver–Burk plots for TAO in the absence and presence of 5, 10, and 15 nM of the inhibitor. The curve revealed a noncompetitive inhibition pattern for 1 against TAO. (B) Dixon plot showing the K _i of 1 on TAO. The slopes of the Lineweaver–Burk primary plot were plotted against the varying concentrations of 1. K _i = −(intercept over the X-axis) of the Dixon secondary plot.

5

Structure of TAO–compound 1 complexes (PDB ID: 9M2A). (A) Complex structure of TAO–1 in the crystal. Each crystal lattice was formed by two dimers of TAOformed by chains A-B and C-Drelated by a noncrystallographic 2-fold axis. Compound 1 was bound only to chain A. Compound 1 is shown in gray stick model with the 2Fo–Fc map contoured at 1.0 σ, showing the electron density for 1 bound far from the active site (the diiron center) of Chain A. Fe atoms and OH^– ions are respectively shown as purple and red spheres and present in the active site of all chains. (B) Stick model of X-ray crystallographic structure for rTAO–1 complex. Compound 1 was bound to TAO by varied interactions with different amino acid residues at the binding site. The binding was established by hydrogen bonds and stacking interactions and hydrophobic interactions. The 4-hydroxy-2-methyl benzoic acid head is marked as , while the triphenylphosphonium tail as ☆, both linked by the methylene linker. The electron density map is shown contoured to 1.0 σ. (C) The amino acids binding compound 1. The amino acid residues located within 4 Å of the 1 molecule are shown in stick models. The hydrogen bonds are shown in dash lines. The omit electron density map for molecule 1 is contoured 1.0 σ (blue) and 1.5 σ (purple), respectively. (D) Dimer structure of TAO–1 complex. Chain A is shown in the rainbow cartoon from blue (N terminus) to red (C terminus), and chain B is gray. Compound 1 is shown in the yellow stick model. Fe atoms and OH^– ions are shown as purple and red spheres, respectively. Compound 1 is bound only to chain A, while the Fe atoms and OH^– ion are in both chains. (E) 90° orientation of chain D, revealing that 1 was bound far away (∼25 Å) from the active site of TAO. Fe atoms and OH^– ions present in the active site are shown as purple and red spheres, respectively.

6

Active site structure of TAO in the absence (A) or presence (B) of compound 1. (A) The active site in the absence of 1 (PDB ID: 9KUN). The key amino acids are shown as sticks and labeled. Fe atoms and OH^– ions are shown as purple and red spheres, respectively. (B) The active site structure in the presence of 1 (PDB ID: 9M2A). Note an ∼70° inward conformational change of R96. (C) Superposed structures of the cartoon models for the ligand-free TAO and TAO–1 complex showing the active site amino acids (PDB ID: 9KUN and 9M2A). Amino acid residues in the catalytic site are shown in the ball and stick model; those for the ligand-free form are colored pink and the TAO–1 is colored green. (D) Structures of the ligand-free form in the surface model superposed with the cartoon model of the TAO–1 structure, with the amino acids shown as sticks (PDB ID: 9KUN and 9M2A). (E) Surface model of the structure of ligand-free TAO (PDB ID: 9KUN). The active site cavity is marked with a black-contoured white star ☆. (F) Surface models of structures of TAO–1 (PDB ID: 9M2A). The active site cavity, marked with a black-contoured white star ☆, was blocked.

7

Protein sequence alignment of AOXs. The alignment was created with ESPript 3.0 alignment editor. The protein sources and their NCBI accession numbers are indicated. The predicted secondary structure elements identified in T. b. brucei are shown above the alignment: arrows (β-strands), TT (β-turns), coils (α-helices), and η (3₁₀-helix). Amino acid residues shown in white with a red background shade are the residues conserved among AOXs. Amino acid residues implicated for binding compound 1 in T. b. brucei AOX are indicated by black triangles. All amino acids binding compound 1 are conserved in T. congolense except for the change of S140 and V199 in T. b. brucei to K140 and I199 in T. congolense.

8

Structures of compound 1 binding sites showing the coordinating residues. (A) Crystal structure of T. b. brucei AOX-compound 1 complex (PDB ID: 9M2A). (B) AlphaFold modeled structure of T. congolense AOX docked with compound 1. The structure revealed that the side chain of K140 (underlined) interferes with the binding position of the methylene linker of compound 1. (C) Superposed structures of T. b. brucei (crystal structure, green amino acids) and T. congolense (modeled structure, pink amino acids) AOX-compound 1 complexes.