Comprehensive analysis of methods used for the evaluation of compounds against Mycobacterium tuberculosis

In vivo efficacy of apramycin in murine infection models

Apramycin is a unique aminoglycoside with a dissociation of antibacterial activity and ototoxicity. We assessed the antibacterial efficacy of apramycin in two murine models of infection, Mycobacterium tuberculosis aerosol infection and Staphylococcus aureus septicemia. In both infection models, the efficacy of apramycin was comparable to that of amikacin. These results suggest that apramycin has the potential to become a clinically useful agent against drug-resistant pathogens and support further development of this promising unique aminoglycoside.

In vitro and in vivo activities of the nitroimidazole TBA-354 against Mycobacterium tuberculosis

Nitroimidazoles are a promising new class of antitubercular agents. The nitroimidazo-oxazole delamanid (OPC-67683, Deltyba) is in phase III trials for the treatment of multidrug-resistant tuberculosis, while the nitroimidazo-oxazine PA-824 is entering phase III for drug-sensitive and drug-resistant tuberculosis. TBA-354 (SN31354[(S)-2-nitro-6-((6-(4-trifluoromethoxy)phenyl)pyridine-3-yl)methoxy)-6,7-dihydro-5H-imidazo[2,1-b][1,3]oxazine]) is a pyridine-containing biaryl compound with exceptional efficacy against chronic murine tuberculosis and favorable bioavailability in preliminary rodent studies. It was selected as a potential next-generation antituberculosis nitroimidazole following an extensive medicinal chemistry effort. Here, we further evaluate the pharmacokinetic properties and activity of TBA-354 against Mycobacterium tuberculosis. TBA-354 is narrow spectrum and bactericidal in vitro against replicating and nonreplicating Mycobacterium tuberculosis, with potency similar to that of delamanid and greater than that of PA-824. The addition of serum protein or albumin does not significantly alter this activity. TBA-354 maintains activity against Mycobacterium tuberculosis H37Rv isogenic monoresistant strains and clinical drug-sensitive and drug-resistant isolates. Spontaneous resistant mutants appear at a frequency of 3 × 10(-7). In vitro studies and in vivo studies in mice confirm that TBA-354 has high bioavailability and a long elimination half-life. In vitro studies suggest a low risk of drug-drug interactions. Low-dose aerosol infection models of acute and chronic murine tuberculosis reveal time- and dose-dependent in vivo bactericidal activity that is at least as potent as that of delamanid and more potent than that of PA-824. Its superior potency and pharmacokinetic profile that predicts suitability for once-daily oral dosing suggest that TBA-354 be studied further for its potential as a next-generation nitroimidazole.

Some Nigerian anti-tuberculosis ethnomedicines: a preliminary efficacy assessment

ETHNOPHARMACOLOGICAL SIGNIFICANCE

Nigerian herbalists possess indigenous ethnomedicinal recipes for the management of tuberculosis and related ailments. A collaborative preliminary modern scientific evaluation of the efficacy of some Nigerian ethnomedicines used by traditional medicine practitioners (TMPs) in the management of tuberculosis and related ailments has been carried out.

MATERIALS AND METHODS

Ethnomedicinal recipes (ETMs) were collected from TMPs from locations in various ecological zones of Nigeria under a collaborative understanding. The aqueous methanolic extracts of the ETMs were screened against Mycobacterium bovis, BCG and Mycobacterium tuberculosis strain H37Rv using the broth microdilution method.

RESULTS

Extracts of ETMs screened against BCG showed 69% activity against the organism. The activities varied from weak, ≤2500 µg/mL to highly active, 33 µg/mL 64% of the extracts were active against Mycobacterium tuberculosis The activities of the extracts against Mycobacterium tuberculosis varied from weak, ≤2500 µg/mL to highly active, 128 µg/mL. There was 77% agreement in results obtained using BCG or Mycobacterium tuberculosis as test organisms.

CONCLUSION

The results show clear evidence for the efficacy of the majority of indigenous Nigerian herbal recipes in the ethnomedicinal management of tuberculosis and related ailments. BCG may be effectively used, to a great extent, as the organism for screening for potential anti-Mycobacterium tuberculosis agents. A set of prioritization criteria for the selection of plants for initial further studies for the purpose of antituberculosis drug discovery research is proposed.

Are bigger data sets better for machine learning? Fusing single-point and dual-event dose response data for Mycobacterium tuberculosis

Tuberculosis is a major, neglected disease for which the quest to find new treatments continues. There is an abundance of data from large phenotypic screens in the public domain against Mycobacterium tuberculosis (Mtb). Since machine learning methods can learn from past data, we were interested in addressing whether more data builds better models. We now describe using Bayesian machine learning to assess whether we can improve our models by combining the large quantities of single-point data with the much smaller (higher quality) dual-event data sets, which use both dose-response data for both whole-cell antitubercular activity and Vero cell cytotoxicity. We have evaluated 12 models ranging from different single-point, dual-event dose-response, single-point and dual-event dose-response as well as combined data sets for three distinct data sets from the same laboratory. We used a fourth data set of active and inactive compounds from the same group as well as a smaller set of 177 active compounds from GlaxoSmithKline as test sets. Our data suggest combining single-point with dual-event dose-response data does not diminish the internal or external predictive ability of the models based on the receiver operator curve (ROC) for these models (internal ROC range 0.83-0.91, external ROC range 0.62-0.83) compared to the orders of magnitude smaller dual-event models (internal ROC range 0.6-0.83 and external ROC 0.54-0.83). In conclusion, models developed with 1200-5000 compounds appear to be as predictive as those generated with 25 000-350 000 molecules. Our results have implications for justifying further high-throughput screening versus focused testing based on model predictions.

Pharmacologically active metabolites, combination screening and target identification-driven drug repositioning in antituberculosis drug discovery

There has been renewed interest in alternative strategies to address bottlenecks in antibiotic development. These include the repurposing of approved drugs for use as novel anti-infective agents, or their exploitation as leads in drug repositioning. Such approaches are especially attractive for tuberculosis (TB), a disease which remains a leading cause of morbidity and mortality globally and, increasingly, is associated with the emergence of drug-resistance. In this review article, we introduce a refinement of traditional drug repositioning and repurposing strategies involving the development of drugs that are based on the active metabolite(s) of parental compounds with demonstrated efficacy. In addition, we describe an approach to repositioning the natural product antibiotic, fusidic acid, for use against Mycobacterium tuberculosis. Finally, we consider the potential to exploit the chemical matter arising from these activities in combination screens and permeation assays which are designed to confirm mechanism of action (MoA), elucidate potential synergies in polypharmacy, and to develop rules for drug permeability in an organism that poses a special challenge to new drug development.

Looking back to the future: predicting in vivo efficacy of small molecules versus Mycobacterium tuberculosis

Selecting and translating in vitro leads for a disease into molecules with in vivo activity in an animal model of the disease is a challenge that takes considerable time and money. As an example, recent years have seen whole-cell phenotypic screens of millions of compounds yielding over 1500 inhibitors of Mycobacterium tuberculosis (Mtb). These must be prioritized for testing in the mouse in vivo assay for Mtb infection, a validated model utilized to select compounds for further testing. We demonstrate learning from in vivo active and inactive compounds using machine learning classification models (Bayesian, support vector machines, and recursive partitioning) consisting of 773 compounds. The Bayesian model predicted 8 out of 11 additional in vivo actives not included in the model as an external test set. Curation of 70 years of Mtb data can therefore provide statistically robust computational models to focus resources on in vivo active small molecule antituberculars. This highlights a cost-effective predictor for in vivo testing elsewhere in other diseases.

Expression of antimicrobial drug tolerance by attached communities of Mycobacterium tuberculosis

There is an urgent need to improve methods used to screen antituberculosis drugs. An in vitro assay was developed to test drug treatment strategies that specifically target drug-tolerant Mycobacterium tuberculosis. The H37Rv strain of M. tuberculosis survived antimicrobial treatment as attached microbial communities when maintained in tissue culture media (RPMI-1640) with or without lysed human peripheral blood leukocytes. When cultured planktonically in the presence of Tween-80, bacilli failed to form microbial communities or reach logarithmic phase growth yet remained highly susceptible to antimicrobial drugs. In the absence of Tween, bacilli tolerated drug therapy by forming complex microbial communities attached to untreated well surfaces or to the extracellular matrix derived from lysed human leukocytes. Treatment of microbial communities with DNase I or Tween effectively dispersed bacilli and restored drug susceptibility. These data demonstrate that in vitro expression of drug tolerance by M. tuberculosis is linked to the establishment of attached microbial communities and that dispersion of bacilli targeting the extracellular matrix including DNA restores drug susceptibility. Modifications of this in vitro assay may prove beneficial in a high-throughput platform to screen new antituberculosis drugs especially those that target drug-tolerant bacilli.

Targeting dormant bacilli to fight tuberculosis

Tuberculosis (TB) is an infectious disease caused by Mycobacterium tuberculosis (Mtb), which kills about 2 million people annually. Furthermore, 2 billion people worldwide are latently infected with this organism, with 10% of them reactivating to active TB due to re-growth of nonreplicating (dormant) Mtb residing in their tissues. Because of the huge reservoir of latent TB it is important to find novel drugs/drug combinations killing dormant bacilli (microaerophiles, anaerobes and drug-tolerant persisters) surviving for decades in a wide spectrum of granulomatous lesions in the lungs of TB patients. Antibiotic treatment of drug-susceptible TB requires administration of isoniazid, rifampin, pyrazinamide, ethambutol for 2 months, followed by isoniazid and rifampin for 4 months. To avoid reactivation of dormant Mtb to active pulmonary TB, up to 9 months of treatment with isoniazid is required. Therefore, a strategy to eliminate dormant bacilli needs to be developed to shorten therapy of active and latent TB and reduce the reservoir of people with latent TB. Finding drugs with high rate of penetration into the caseous granulomas and understanding the biology of dormant bacilli and in particular of persister cells, phenotypically resistant to antibiotics, will be essential to eradicate Mtb from humans. In recent years unprecedented efforts have been done in TB drug discovery, aimed at identifying novel drugs and drug combinations killing both actively replicating and nonreplicating Mtb in vitro, in animal models and in clinical trials in humans.

Fusing dual-event data sets for Mycobacterium tuberculosis machine learning models and their evaluation

The search for new tuberculosis treatments continues as we need to find molecules that can act more quickly, be accommodated in multidrug regimens, and overcome ever increasing levels of drug resistance. Multiple large scale phenotypic high-throughput screens against Mycobacterium tuberculosis (Mtb) have generated dose response data, enabling the generation of machine learning models. These models also incorporated cytotoxicity data and were recently validated with a large external data set. A cheminformatics data-fusion approach followed by Bayesian machine learning, Support Vector Machine, or Recursive Partitioning model development (based on publicly available Mtb screening data) was used to compare individual data sets and subsequent combined models. A set of 1924 commercially available molecules with promising antitubercular activity (and lack of relative cytotoxicity to Vero cells) were used to evaluate the predictive nature of the models. We demonstrate that combining three data sets incorporating antitubercular and cytotoxicity data in Vero cells from our previous screens results in external validation receiver operator curve (ROC) of 0.83 (Bayesian or RP Forest). Models that do not have the highest 5-fold cross-validation ROC scores can outperform other models in a test set dependent manner. We demonstrate with predictions for a recently published set of Mtb leads from GlaxoSmithKline that no single machine learning model may be enough to identify compounds of interest. Data set fusion represents a further useful strategy for machine learning construction as illustrated with Mtb. Coverage of chemistry and Mtb target spaces may also be limiting factors for the whole-cell screening data generated to date.

Identification of novel inhibitors of nonreplicating Mycobacterium tuberculosis using a carbon starvation model

During Mycobacterium tuberculosis infection, a population of bacteria is thought to exist in a nonreplicating state, refractory to antibiotics, which may contribute to the need for prolonged antibiotic therapy. The identification of inhibitors of the nonreplicating state provides tools that can be used to probe this hypothesis and the physiology of this state. The development of such inhibitors also has the potential to shorten the duration of antibiotic therapy required. Here we describe the development of a novel nonreplicating assay amenable to high-throughput chemical screening coupled with secondary assays that use carbon starvation as the in vitro model. Together these assays identify compounds with activity against replicating and nonreplicating M. tuberculosis as well as compounds that inhibit the transition from nonreplicating to replicating stages of growth. Using these assays we successfully screened over 300,000 compounds and identified 786 inhibitors of nonreplicating M. tuberculosis In order to understand the relationship among different nonreplicating models, we tested 52 of these molecules in a hypoxia model, and four different chemical scaffolds in a stochastic persister model, and a streptomycin-dependent model. We found that compounds display varying levels of activity in different models for the nonreplicating state, suggesting important differences in bacterial physiology between models. Therefore, chemical tools identified in this assay may be useful for determining the relevance of different nonreplicating in vitro models to in vivo M. tuberculosis infection. Given our current limited understanding, molecules that are active across multiple models may represent more promising candidates for further development.

Combining computational methods for hit to lead optimization in Mycobacterium tuberculosis drug discovery

PURPOSE

Tuberculosis treatments need to be shorter and overcome drug resistance. Our previous large scale phenotypic high-throughput screening against Mycobacterium tuberculosis (Mtb) has identified 737 active compounds and thousands that are inactive. We have used this data for building computational models as an approach to minimize the number of compounds tested.

METHODS

A cheminformatics clustering approach followed by Bayesian machine learning models (based on publicly available Mtb screening data) was used to illustrate that application of these models for screening set selections can enrich the hit rate.

RESULTS

In order to explore chemical diversity around active cluster scaffolds of the dose-response hits obtained from our previous Mtb screens a set of 1924 commercially available molecules have been selected and evaluated for antitubercular activity and cytotoxicity using Vero, THP-1 and HepG2 cell lines with 4.3%, 4.2% and 2.7% hit rates, respectively. We demonstrate that models incorporating antitubercular and cytotoxicity data in Vero cells can significantly enrich the selection of non-toxic actives compared to random selection. Across all cell lines, the Molecular Libraries Small Molecule Repository (MLSMR) and cytotoxicity model identified ~10% of the hits in the top 1% screened (>10 fold enrichment). We also showed that seven out of nine Mtb active compounds from different academic published studies and eight out of eleven Mtb active compounds from a pharmaceutical screen (GSK) would have been identified by these Bayesian models.

CONCLUSION

Combining clustering and Bayesian models represents a useful strategy for compound prioritization and hit-to lead optimization of antitubercular agents.

A medicinal chemists' guide to the unique difficulties of lead optimization for tuberculosis

Tuberculosis is a bacterial disease that predominantly affects the lungs and results in extensive tissue pathology. This pathology contributes to the complexity of drug development as it presents discrete microenvironments within which the bacterium resides, often under conditions where replication is limited and intrinsic drug susceptibility is low. This consolidated pathology also results in impaired vascularization that limits access of potential lead molecules to the site of infection. Translating these considerations into a target-product profile to guide lead optimization programs involves implementing unique in vitro and in vivo assays to maximize the likelihood of developing clinically meaningful candidates.

Rapid in vivo assessment of drug efficacy against Mycobacterium tuberculosis using an improved firefly luciferase

Objectives

In vivo experimentation is costly and time-consuming, and presents a major bottleneck in anti-tuberculosis drug development. Conventional methods rely on the enumeration of bacterial colonies, and it can take up to 4 weeks for Mycobacterium tuberculosis to grow on agar plates. Light produced by recombinant bacteria expressing luciferase enzymes can be used as a marker of bacterial load, and disease progression can be easily followed non-invasively in live animals by using the appropriate imaging equipment. The objective of this work was to develop a bioluminescence-based mouse model of tuberculosis to assess antibiotic efficacy against M. tuberculosis in vivo.

Methods

We used an M. tuberculosis strain carrying a red-shifted derivative of the firefly luciferase gene (FFlucRT) to infect mice, and monitored disease progression in living animals by bioluminescence imaging before and after treatment with the frontline anti-tuberculosis drug isoniazid. The resulting images were analysed and the bioluminescence was correlated with bacterial counts.

Results

Using bioluminescence imaging we detected as few as 1.7 × 10³ and 7.5 × 10⁴ reporter bacteria ex vivo and in vivo, respectively, in the lungs of mice. A good correlation was found between bioluminescence and bacterial load in both cases. Furthermore, a marked reduction in luminescence was observed in living mice given isoniazid treatment.

Conclusions

We have shown that an improved bioluminescent strain of M. tuberculosis can be visualized by non-invasive imaging in live mice during an acute, progressive infection and that this technique can be used to rapidly visualize and quantify the effect of antibiotic treatment. We believe that the model presented here will be of great benefit in early drug discovery as an easy and rapid way to identify active compounds in vivo.

Mycobacterium tuberculosis is resistant to streptolydigin

Drug resistant strains of Mycobacterium tuberculosis (Mtb) undermine tuberculosis (TB) control. Streptolydigin is a broadly effective antibiotic which inhibits RNA polymerase, similarly to rifampicin, a key drug in current TB chemotherapeutic regimens. Due to a vastly improved chemical synthesis streptolydigin and derivatives are being promoted as putative TB drugs. The microplate Alamar Blue assay revealed that Streptococcus salivarius and Mycobacterium smegmatis were susceptible to streptolydigin with minimum inhibitory concentrations (MICs) of 1.6 mg/L and 6.25 mg/L, respectively. By contrast, the MICs of streptolydigin and two derivatives, streptolydiginone and dihydrostreptolydigin, against Mtb were ≥ 100 mg/L demonstrating that Mtb is resistant to streptolydigin in contrast to previous reports. Further, a porin mutant of M. smegmatis is resistant to streptolydigin indicating that porins mediate uptake of streptolydigin across the outer membrane. Since the RNA polymerase is a validated drug target in Mtb and porins are required for susceptibility of M. smegmatis, the absence of MspA-like porins probably contributes to the resistance of Mtb to streptolydigin. This study shows that streptolydigin is not a suitable drug in TB treatment regimens.

A dual read-out assay to evaluate the potency of compounds active against Mycobacterium tuberculosis

Tuberculosis is a serious global health problem caused by the bacterium Mycobacterium tuberculosis. There is an urgent need for discovery and development of new treatments, but this can only be accomplished through rapid and reproducible M. tuberculosis assays designed to identify potent inhibitors. We developed an automated 96-well assay utilizing a recombinant strain of M. tuberculosis expressing a far-red fluorescent reporter to determine the activity of novel compounds; this allowed us to measure growth by monitoring both optical density and fluorescence. We determined that optical density and fluorescence were correlated with cell number during logarithmic phase growth. Fluorescence was stably maintained without antibiotic selection over 5 days, during which time cells remained actively growing. We optimized parameters for the assay, with the final format being 5 days' growth in 96-well plates in the presence of 2% w/v DMSO. We confirmed reproducibility using rifampicin and other antibiotics. The dual detection method allows for a reproducible calculation of the minimum inhibitory concentration (MIC), at the same time detecting artefacts such as fluorescence quenching or compound precipitation. We used our assay to confirm anti-tubercular activity and establish the structure activity relationship (SAR) around the imidazo[1,2-a]pyridine-3-carboxamides, a promising series of M. tuberculosis inhibitors.

Activities of drug combinations against Mycobacterium tuberculosis grown in aerobic and hypoxic acidic conditions

Mycobacterium tuberculosis is exposed to hypoxia and acidity within granulomatous lesions. In this study, an acidic culture model of M. tuberculosis was used to test drug activity against aerobic 5-day-old (A5) and hypoxic 5-, 12-, and 19-day-old (H5, H12, and H19, respectively) bacilli after 7, 14, and 21 days of exposure. In A cultures, CFU and pH rapidly increased, while in H cultures growth stopped and pH increased slightly. Ten drugs were tested: rifampin (R), isoniazid (I), pyrazinamide (Z), ethambutol (E), moxifloxacin (MX), amikacin (AK), metronidazole (MZ), nitazoxanide (NZ), niclosamide (NC), and PA-824 (PA). Rifampin was the most active against A5, H5, H12, and H19 bacilli. Moxifloxacin and AK efficiently killed A5 and H5 cells, I was active mostly against A5 cells, Z was most active against H12 and H19 cells, and E showed low activity. Among nitrocompounds, NZ, NC, and PA were effective against A5, H5, H12, and H19 cells, while MZ was active against H12 and H19 cells. To kill all A and H cells, A5- and H5-active agents R, MX, and AK were used in combination with MZ, NZ, NC, or PA, in comparison with R-I-Z-E, currently used for human therapy. Mycobacterial viability was determined by CFU and a sensitive test in broth (day to positivity, MGIT 960 system). As shown by lack of regrowth in MGIT, the most potent combination was R-MX-AK-PA, which killed all A5, H5, H12, and H19 cells in 14 days. These observations demonstrate the sterilizing effect of drug combinations against cells of different M. tuberculosis stages grown in aerobic and hypoxic acidic conditions.