Inhibitors of pathogen intercellular signals as selective anti-infective compounds

Abstract

Long-term antibiotic use generates pan-resistant super pathogens. Anti-infective compounds that selectively disrupt virulence pathways without affecting cell viability may be used to efficiently combat infections caused by these pathogens. A candidate target pathway is quorum sensing (QS), which many bacterial pathogens use to coordinately regulate virulence determinants. The Pseudomonas aeruginosa MvfR-dependent QS regulatory pathway controls the expression of key virulence genes; and is activated via the extracellular signals 4-hydroxy-2-heptylquinoline (HHQ) and 3,4-dihydroxy-2-heptylquinoline (PQS), whose syntheses depend on anthranilic acid (AA), the primary precursor of 4-hydroxy-2-alkylquinolines (HAQs). Here, we identified halogenated AA analogs that specifically inhibited HAQ biosynthesis and disrupted MvfR-dependent gene expression. These compounds restricted P. aeruginosa systemic dissemination and mortality in mice, without perturbing bacterial viability, and inhibited osmoprotection, a widespread bacterial function. These compounds provide a starting point for the design and development of selective anti-infectives that restrict human P. aeruginosa pathogenesis, and possibly other clinically significant pathogens.

Figure 1. Anthranilic Acid Analogs and HAQ Biosynthetic Pathway

(A) Chemical structures of AA, MA, and five halogenated AA analogs. (B) The HAQ biosynthetic pathway and hypothetical steps mediated by the pqsA-D gene products.

Figure 2. Halogenated AA Analogs Inhibit HAQ Biosynthesis

Culture supernatants of PA14 cells grown in LB media minus or plus 1.5 mM or 6 mM 6FABA, 1.5 mM or 6 mM 6CABA, 1.5 mM 4CABA, or 1.5 mM MA, were collected at 5, 10, 15, and 24 h; and HHQ, PQS, and HQNO concentrations (mg/l) were quantified by LC/MS. Data points are the average of triplicate experiments, ± standard deviation. To facilitate presentation, below-detection levels of HAQs are indicated as 0.01.

Figure 3. AA Analogs Compete with AA for the PqsA Active Site

(A) The AA analogs do not perturb AA accumulation. Intracellular AA concentrations (mM) were determined for PA14 cells grown minus or plus 1.5 mM 6FABA, 6CABA, or 4CABA, and for pqsA ⁻ mutant cells. (B) Exogenous AA partially overcomes AA analog inhibition of HHQ synthesis in PA14 cells. Increasing concentrations of AA were added to cultures grown in 0.125 mM 6FABA, 0.75 mM 6CABA, or 0.125 mM 4CABA, which reduced HHQ production by 75% in supernatants of OD₆₀₀ = 4 cell cultures. The competing AA concentrations were 0, 3, 6, and 10 times that of each AA analog.

Figure 4. The AA Analogs Promote Survival of B+I Mice to PA14 Pathogenesis, and Restrict In Vivo HHQ Production

(A) PA14-infected mice were injected 6 h post-B+I with 6FABA, 6CABA, or 4CABA. Mice infected with pqsA⁻ cells served as an additional control group. Data were averaged for a minimum of two independent experiments, with n(PA14) = 40; n(PA14/6FABA) = 36; n(PA14/6CABA) = 30; n(PA14/4CABA) = 20 and n(pqsA⁻) = 20. Standard deviations were <20% for the time points between 0 and 4 d, and <12% thereafter. Cox proportional hazards regression showed that mouse mortality due to PA14 infection was significantly reduced by injection with 6FABA (hazard ratio = 0.398, p = 5.6e-04), 6CABA (hazard ratio = 0.397, p = 1.2e-03), and 4CABA (hazard ratio = 0.325, P = 1.8e-03), compared with control saline; pqsA⁻ virulence was highly attenuated (hazard ratio = 0.118, p = 9.1e-06). (B) HHQ levels (μg/mg) at 12 h post-B+I for rectus adbominus muscle directly underlying the infection site in untreated B+I mice, and for mice treated 6 h post-B+I with 6FABA, 6CABA, or 4CABA. n = 5 for each experimental condition. The t-test (p = 0.011) and Wilcoxon rank sum test (p = 0.03) showed that the difference in HHQ levels between control and 4CABA-treated mice was statistically significant.

Figure 5. The AA Analogs Restrict Systemic Bacterial Proliferation, but Not Local Proliferation at the Infection Site

PA14-infected mice were left untreated (controls) or injected 6 h post-B+I with 6FABA, 6CABA, or 4CABA. The animals were subsequently sacrificed at 22 h post-B+I, and bacteria CFU/mg were determined for (A) muscle tissue that directly underlies the infection site; (B) muscle tissue adjacent to the infection site; and (C) blood. Numbers above dots correspond to the number of mice for each condition. The statistical significance of the CFU/mg differences between the treated and untreated control mice were determined using a Wilcoxon rank sum test. p-Values for the difference in adjacent muscle in response to 6FABA, 6CABA, and 4CABA treatment versus no treatment were 0.00001, 0.00066, and 0.00267, respectively; p-values for the difference in blood with 6FABA, 6CABA, and 4CABA, treatment versus no treatment were 0.00147, 0.00331, and 0.00331, respectively.

Figure 6. The AA Analogs Inhibit HAQ Production in B. thailandensis Cell Supernatants

Mass spectrometry (MS) determination of unsaturated HAQs in OD₆₀₀ = 3 culture supernatants of B. thailandensis cells grown in the absence or presence of 3 mM 6FABA, 6CABA, or 4CABA. Note that while saturated HAQs, including HHQ, occur in B. thailandensis cells, their levels were below the quantification threshold.

Figure 7. The AA Analogs Increase the Osmotic Sensitivity of Several Bacterial Pathogens

Ten times (10×) serial dilutions of the indicated bacterial pathogens were spotted onto high-salt tester plates containing no AA analog, 6 mM 6FABA, 6 mM 6CABA, or 1.5 mM 4CABA. Note that at these concentrations, bacterial growth on regular low-salt LB plates was unaffected.