A high-throughput whole cell screen to identify inhibitors of Mycobacterium tuberculosis

Abstract

Tuberculosis is a disease of global importance for which novel drugs are urgently required. We developed a whole-cell phenotypic screen which can be used to identify inhibitors of Mycobacterium tuberculosis growth. We used recombinant strains of virulent M. tuberculosis which express far-red fluorescent reporters and used fluorescence to monitor growth in vitro. We optimized our high throughput assays using both 96-well and 384-well plates; both formats gave assays which met stringent reproducibility and robustness tests. We screened a compound set of 1105 chemically diverse compounds previously shown to be active against M. tuberculosis and identified primary hits which showed ≥ 90% growth inhibition. We ranked hits and identified three chemical classes of interest-the phenoxyalkylbenzamidazoles, the benzothiophene 1-1 dioxides, and the piperidinamines. These new compound classes may serve as starting points for the development of new series of inhibitors that prevent the growth of M. tuberculosis. This assay can be used for further screening, or could easily be adapted to other strains of M. tuberculosis.

Fig 1. Optimization of fluorescence measurements.

DREAM8 was dispensed into 384-well plates and fluorescence was measured at varying excitation wavelengths when the emission wavelength was fixed at 590nm (A) or at varying emission wavelengths when a fixed excitation of 558nm was used (B). Data are the average ± SD from four wells.

Fig 2. Optimization of growth conditions.

(A) Serial dilutions of CHEAM3 were plated in triplicate in a 384 well plate and measured for fluorescence before and after five days of incubation. The signal to background (S:B) for each inoculum was calculated (average signal D0÷average signal D5) to identify the greatest amplitude to measure growth of cells over the course of 5 days. (B) 320 experimental wells of a 384 well plate were inoculated with CHEAM3 diluted to a starting OD₅₉₀ of 0.01, 0.02 or 0.03. Plates were read on D4 and D5. The average fluorescence for each condition (n = 320) was plotted, with error bars indicating the standard deviation. The calculated signal to background (S:B) is shown above each bar.

Fig 3. High throughput screen validation scatter plots: Two plates each containing minimum, midpoint, and maximum signal controls were run in 384-well plates on three separate days using the final assay conditions.

Recombinant M. tuberculosis expressing (A) DsRed or (B) mCherry was grown for 5 days. Relative fluorescence units (RFU) were measured in each well.

Fig 4

Small molecule compound library screen: A selected library of 1105 small molecules from NIH-SRI/TAACF was screened in replicate experiments against (A) CHEAM3 or (B) DREAM8. The % inhibition for each compound was calculated. The results from the first and second runs are plotted on the x- and y-axis, respectively, for both strains. For each strain the Pearson coefficient of linear correlation between the two replicate data sets was calculated in Graphpad Prism and is shown in boxed text in the upper right corner of each plot. (C) The average % inhibition of CHEAM3 and DREAM8 growth was calculated and plotted on the x- and y- axis, respectively. The calculated Pearson co-efficient comparing the data generated from the two different strains is shown in boxed text in the upper right corner of the plot.

Fig 5. Selected hit compounds from screen.

Three hit chemotypes identified in our screen were noted as being of interest for further development. Their structures and MICs are shown.