Quinoline ATP Synthase Inhibitors with Activity Against Multidrug Resistant Acinetobacter baumannii and Pseudomonas aeruginosa

Abstract

The Gram-negative, pathogenic bacteria Acinetobacter baumannii (AB) and Pseudomonas aeruginosa (PA) have been identified as a particular threat due to rising multidrug resistance, and antibiotics with novel mechanisms of action are needed. Bacterial bioenergetics is a promising but underdeveloped drug target since the complexes of oxidative phosphorylation are critical to cell survival in these organisms. Building from our previous work using quinoline derivatives to inhibit the ATP synthase of PA, we report a new set of 14 quinoline derivatives that demonstrates potent inhibition of the AB ATP synthase, with the best inhibitor having an IC50 of 230 ng/mL in vitro, expands the quinoline structure-activity relationship against the PA enzyme, and establishes molecular strategies for achieving selectivity between PA and AB. Furthermore, several compounds demonstrated potent antibacterial activity against multidrug resistant strains of AB and PA indicating ATP synthase as a promising new area for broad spectrum antibiotic development in AB.

Keywords: ATP synthase; Acinetobacter baumannii; Pseudomonas aeruginosa; antibiotics; inhibitors.

Figure 1

A) The structure of AB F₁F_o ATP synthase determined by cryo‐EM (PDB ID: 7P2Y). B) Detail of the putative quinoline binding site centered around Asp60 (brown) on subunit c with contributions from other residues from subunit a (pale yellow) and subunit c (pale green). C) Structures of previously reported quinoline‐derived inhibitors of PA ATP synthase. Key positions on the quinoline backbone are numbered. D) Segments of the amino acid sequences of subunit a (left) and subunit c (right) that contribute to the putative quinoline binding site are aligned to compare AB with PA, MT, and Homo sapiens (HS). The critically conserved Asp60 is highlighted in red. Other potential interactions include aliphatic (yellow), aromatic (blue), and polar (green) residues.

Scheme 1

Synthesis of quinoline compounds.

Figure 2

Inhibition of AB ATP synthesis activity in inner membrane vesicles by the most potent compounds: WSA238 (A), WSA255 (B), and WSA257 (C). Individual data points (n≥3) were fit with a variable slope dose response curve (solid line). Dashed lines indicate the 95 % confidence bands for the fit.

Figure 3

Comparison of potency against AB and PA ATP synthase. A) IC₅₀ values from dose response fits are plotted for AB (filled bars) and PA (open bars) ATP synthase inhibition. Compounds are grouped by western modification and colors correspond to the C2 amine (black, N‐ethyl piperidine; red, N‐cyclopentyl piperidine; green, benzyl pyrrolidine; blue, N‐methyl piperizine). Error bars represent standard error of the IC₅₀ parameter of the fit, and statistically significant differences between AB and PA are marked (ns, p>0.05; **, p≤0.01; ***, p≤0.001). B) Hill slopes from dose response fits are plotted for AB (filled bars) and PA (open bars) ATP synthase inhibition and colored as in panel A. Error bars represent standard error of the Hill parameter of the fit. WSA248 is omitted from these plots due to its low potency.

Figure 4

Trends across western changes compared to unmodified within each C2. IC₅₀ or MIC for each unmodified quinoline (R¹‐R³=H) are listed in μg/mL. For each modification, a red arrow indicates less potent activity (higher IC₅₀ or MIC), a green arrow indicates more potent activity (lower IC₅₀ or MIC), and a black bar indicates no change.

Figure 5

Representative high‐scoring docking poses, calculated in MOE and depicted using PyMol, are shown in the putative binding site (PDB ID: 7P2Y) at the interface of subunit a (pale yellow) and subunit c (pale green). Inhibitors WSA236 (A), WSA249 (B−C), WSA269 (D), WSA252 (E), and WSA257 (F) are shown in cyan. Specific interactions determined by MOE are shown as dashed lines with hydrogen bonds colored green and ionic interactions colored purple.