Chemiluminescent Protease Probe for Rapid, Sensitive, and Inexpensive Detection of Live Mycobacterium tuberculosis

Abstract

Tuberculosis (TB) is a top-ten cause of death worldwide. Successful treatment is often limited by insufficient diagnostic capabilities, especially at the point of care in low-resource settings. The ideal diagnostic must be fast, be cheap, and require minimal clinical resources while providing high sensitivity, selectivity, and the ability to differentiate live from dead bacteria. We describe here the development of a fast, luminescent, and affordable sensor of Hip1 (FLASH) for detecting and monitoring drug susceptibility of Mycobacterium tuberculosis (Mtb). FLASH is a selective chemiluminescent substrate for the Mtb protease Hip1 that, when processed, produces visible light that can be measured with a high signal-to-noise ratio using inexpensive sensors. FLASH is sensitive to fmol of recombinant Hip1 enzyme in vitro and can detect as few as thousands of Mtb cells in culture or in human sputum samples within minutes. The probe is highly selective for Mtb compared to other nontuberculous mycobacteria and can distinguish live from dead cells. Importantly, FLASH can be used to measure antibiotic killing of Mtb in culture with greatly accelerated timelines compared to traditional protocols. Overall, FLASH has the potential to enhance both TB diagnostics and drug resistance monitoring in resource-limited settings.

Figure 1

Fast luminescent affordable sensor of Hip1 (FLASH). (A) Following proteolytic cleavage of the FLASH probe by Mtb Hip1, self-elimination and chemiexcitation steps ultimately lead to light emission. (B) Mtb produces Hip1 protease which cleaves the FLASH probe, producing light. (C) Light produced by probe cleavage is measured over time. Total light output in a given time period (dashed area) is summed to yield integrated luminescence.

Figure 2

FLASH is a sensitive probe for Mtb Hip1 activity. (A) Light emitted by the FLASH probe upon incubation with various concentrations of Mtb Hip1. (B) Time course of integrated luminescence from part A. (C) Total integrated luminescence after 1 h of incubation with Mtb Hip1 (mean ± SD, n = 3). Horizontal lines show the mean (solid) ± 3 SD (dashed) of control samples lacking enzyme. Each enzyme concentration was compared to the control samples via one-way ANOVA with Dunnett’s test (***, p < 0.001). The best fit line shows linear regression to the log-transformed data. (D) Kinetic analysis of the FLASH probe. Data were fitted to the Michaelis–Menten equation to yield kinetic parameters. (E) Inhibition of Mtb Hip1 by CSL157 (mean ± SD, n = 3). Mtb Hip1 was preincubated with inhibitor for 30 min at 37 °C before the addition of the FLASH probe. Data were fitted to a four-parameter logistic equation to yield IC₅₀. All parameters are reported as the 95% confidence interval.

Figure 3

FLASH is a quantitative probe for Mtb cells. (A) To determine the limit of detection, cultures of Mtb were serially diluted into medium or into processed human sputum and then incubated with the FLASH probe. Total integrated luminescence after 1 h of incubation of the FLASH probe with (B) Mtb H37Rv in 7H9 medium, (C) mc²6020 in 7H9 medium, and (D) mc²6020 in processed human sputum (mean ± SD, n = 3). Horizontal lines show the mean (solid) ± 3 SD (dashed) of control samples lacking cells. For all experiments, each sample was compared to the no-bacteria control via one-way ANOVA with Dunnett’s test (***, p < 0.001; for part C, all comparisons yielded p < 0.001). Best fit lines show linear regressions to the log-transformed data (excluding cell concentrations that are not significantly different from the control). Limits of detection were calculated by determining the cell number for which the best fit line intercepts the mean + 3 SD of the control samples.

Figure 4

FLASH is selective for Mtb. (A) Phylogenetic tree of potential Mtb Hip1 homologues found in NTMs, other lung pathogens, and commensal members of the human airway and oral microbiomes. Also included is Rv2223c, an uncharacterized peptidase encoded by Mtb with sequence similarity to Hip1. Bacteria are colored by their ability to process FLASH at high cell densities (orange, active; blue, inactive; black, not tested). (B) FLASH signal for 6 × 10⁴ cells of each NTM in 7H9 medium (mean ± SD, n = 3). Each sample was compared to the no-bacteria control via one-way ANOVA with Dunnett’s test (***, p < 0.001). (C) FLASH signal for M. tuberculosis (Mtb), M. gordonae (Mgo), M. intracellulare (Min), M. scrofulaceum (Msc), and M. avium (Mav) at the indicated cell number (mean ± SD, n = 3). (D) Sequence similarity of potential homologues to Mtb Hip1 and limits of detection calculated from part C for each NTM.

Figure 5

FLASH provides a quantitative measure of Mtb viability. (A) Mtb cultures were treated with RIF for up to 9 days. Samples were removed throughout the treatment period and incubated with FLASH for 1 h, or with CellTiter-Blue (CTB) for 24 h. (B) FLASH and CTB measurements for cultures treated for 7 days with RIF (mean ± SD, n = 3). Marker colors correspond to RIF concentrations shown in part A. (C) FLASH signal dependence on RIF concentration for each day (mean ± SD, n = 3). Dose response for killing by RIF as measured by the FLASH probe (D) or CTB (E) (mean ± SD, n = 3). Data were normalized to DMSO (100% viability) and 10 μM RIF (0% viability) and fitted to a two-parameter logistic function. IC₅₀ values are reported as 95% confidence intervals. (F) Time course of mc²6020 treated with the critical concentrations of rifampicin (RIF, 1 μg/mL), ethambutol (EMB, 5 μg/mL), isoniazid (INH, 0.1 μg/mL), pyrazinamide (PZA, 100 μg/mL), or streptomycin (STR, 1 μg/mL). For all days, the signal from untreated cultures was compared to each of the treated cultures via a two-way ANOVA with Dunnett’s test (n = 3; ***, p < 0.001 for the comparison between untreated cultures and each of the antibiotic conditions). (G) Time course of H37Rv (WT) Mtb and RpoB H526D mutant Mtb (H526D) treated with DMSO (black) or the critical concentration of RIF (red). For each day, the RIF- and DMSO-treated conditions were compared via an independent t test (n = 3; ***, p < 0.001). (H) Luminescent signal from H37Rv (WT) or H526D after 6 days of culture in the presence or absence of RIF. Samples are compared to the WT Mtb strain treated with RIF via one-way ANOVA with Dunnett’s test (n = 3; ***, p < 0.001).