Development and application of a cellular, gain-of-signal, bioluminescent reporter screen for inhibitors of type II secretion in Pseudomonas aeruginosa and Burkholderia pseudomallei

Abstract

The type II secretion (T2S) system in gram-negative bacteria comprises the Sec and Tat pathways for translocating proteins into the periplasm and an outer membrane secretin for transporting proteins into the extracellular space. To discover Sec/Tat/T2S pathway inhibitors as potential new therapeutics, the authors used a Pseudomonas aeruginosa bioluminescent reporter strain responsive to SecA depletion and inhibition to screen compound libraries and characterize the hits. The reporter strain placed a luxCDABE operon under regulation of a SecA depletion-responsive upregulated promoter in a secA deletion background complemented with an ectopic lac-regulated secA copy. Bioluminescence was indirectly proportional to the isopropyl-β-D-thiogalactopyranoside concentration and stimulated by azide, a known SecA ATPase inhibitor. A total of 96 compounds (0.1% of 73,000) were detected as primary hits due to stimulation of luminescence with a z score ≥5. Direct secretion assays of the nine most potent hits, representing five chemical scaffolds, revealed that they do not inhibit SecA-mediated secretion of β-lactamase into the periplasm but do inhibit T2S-mediated extracellular secretion of elastase with IC(50) values from 5 to 25 µM. In addition, seven of the nine compounds also inhibited the T2S-mediated extracellular secretion of phospholipase C by P. aeruginosa and protease activity by Burkholderia pseudomallei.

Figure 1

Kinetics and intensity of the luminescent response of four bioluminescent reporter strains to the SecA ATPase inhibitor sodium azide. The P. aeruginosa reporter strains listed below with lac-promoted complemented secA deletions and SecA-depletion-responsive promoter-luxCDABE transcriptional fusions were grown in IPTG-containing medium overnight, subcultured into medium lacking IPTG, and added to microplate wells containing a final concentration of sodium azide of 0.5 mM. The z-scores (number of standard deviations above or below the DMSO-only control well values, calculated from the average of 16 replicates) for each strain are plotted vs. time. Samples are indicated as follows as the strain number with the P. aeruginosa gene used for transcriptional fusion of luxCDABE in parentheses: MDM1145 (PA1365) (○), MDM1144 (PA2408) (□), MDM1143 (PA2403) (◇), and MDM1167 (PA2404) (△).

Figure 2

Comparison of growth and bioluminescence of ΔglmU and ΔsecA P. aerguinosa strains carrying the SecA-depletion responsive promoter from PA1365 fused to P. luminescens luxCDABE. Bioluminescent reporter strains MDM1145 (ΔsecA) and MDM1170 (ΔglmU), both containing the PA1365-luxCDABE element, were grown overnight in LBGS containing 0.05mM IPTG (■) (control culture), 0.025mM IPTG (△), or 0.0125mM IPTG (◇), then sub-cultured into LBGS without IPTG, except for the control, which was maintained at 0.05mM IPTG. Growth (A₆₀₀) was followed for 7.5h for (A) MDM1145 (solid lines) and (B) MDM1170 (dashed lines). (C) Cells (200 μl) from each culture were withdrawn at indicated times, added to 96-well opaque white microplates, and relative light units (RLU) were measured. The averages of triplicate readings are plotted; error bars represent the standard deviation.

Figure 3

Kinetics and intensity of the luminescent response of MDM1145 to compounds from a pilot screen. Three significant hits, (□) 5,7-dichloro-8-hydroxyquinoline, (◇) 8-hydroxyquinoline, (○) demeclocycline, and two compounds that failed to stimulate luminescence significantly, (■) abietic acid, and (◆) 1-(3-chlorophenylamino)-4-phenylphthalazine, were added to microplate wells at 25 μM, and the luminescence of reporter strain MDM1145 was measured for 4.5h. The z-scores (number of standard deviations above the luminescence detected in the DMSO negative control wells, calculated as the average of triplicate readings) are shown vs. time.

Figure 4

Luminescence (RLU) from eleven 384-well microplate wells containing reporter strain MDM1145 in a high throughput screen for T2S inhibitors. RLU values are shown at 180 min. for 352 positive controls (◆, induced by 25 μM DHQ; average of 13,000 +/−1,375) in positions 1–352, for 3,520 samples (○) in positions 353–3,872, and for 352 negative controls (■, no induction, DMSO only; average of 4,550 +/−350) in positions 3,873–4,224. Eighteen samples were designated as hits in this particular set of plates (0.5% primary hit rate) because their RLU values exhibited z-scores >5 (i.e., >5 standard deviations above the average negative control value, denoted as a horizontal line at 6,320 RLU).

Figure 5

Profiling of hit compounds against a panel of luminescent reporter strains. Nine confirmed hits were profiled for their ability to stimulate luminescence of efflux-deficient strains carrying the same SecA-depletion responsive promoter as used in the screening strain (PA1365) as well as a variety of promoters designed to respond to inhibition of related (solid shaded and unshaded bars) and unrelated (cross-hatched and stippled bars) targets. The strain number, the number of the P. aeruginosa PAO1 gene whose promoter was fused to the P. luminescens luxCDABE operon, and the target to which the promoter-lux fusion was designed to respond are noted in the graph legend. Cells were grown as described in Materials and Methods and incubated with 25 μM compound. Luminescence was measured at 2 h and RLU values are presented as the number of standard deviations above or below the average negative (DMSO only) control (z-score).