Drug Discovery

Overcoming the difficulties in treating TB and the emerging problem of drug resistance will require a greatly expanded collection of drugs directed against new targets and new physiologic states of the bacteria. We are leveraging the Broad's chemical screening capabilities to systematically identify novel chemicals that can kill Mtb in a variety of states. We will also use newly developed genomic and proteomic techniques to rapidly identify the protein targets of these chemicals to allow for lead optimization.

We have already developed and executed three small-molecule screens targeting M. tuberculosis (Mtb) in a variety of physiological states, yielding a number of new small molecule leads, new insights into the activities of known bioactive compounds, and/or new understanding of the cellular processes required for Mtb growth and survival.

Beyond the initial screening efforts, we are continuing to develop new methods for identifying the targets of the small molecule inhibitors of Mtb. These include:

  • Affinity-based approaches that capitalize on the proteomic expertise available here at the Broad
  • Genetic library-based strategies described previously
  • Novel strategies for the enrichment and isolation of resistant mutants from non-replicating cultures

Additionally, we are working in collaboration with the Broad Institute Chemical Biology and Therapeutics group to discover and develop promising leads that could shorten the duration of TB therapy in patients. This chemical effort is in collaboration with GlaxoSmithKline (GSK), who will perform downstream in vivo testing of candidates.

Non-Replicating Persistence

TB is able to survive within the infected individual despite treatment with current antibiotics. This is likely due to the ability of the bacteria to adopt physiologic states that elude the activity of current drugs.  What is needed are new drugs that are able to target these alternative in vivo states and expedite eradication of the pathogen.

We believe strongly that screening for inhibitors of exponentially growing Mtb in rich medium does a poor job of recapitulating the state of the bacterium during human disease, particularly the state of persistence. It has been shown across a range of animal models that bacteria within the host are exposed to conditions that restrict their growth, including hypoxia, acidic pH, and nutrient limitation, among others. The vulnerable biological pathways under these conditions likely differ drastically from those during exponential growth, which is why current TB antibiotics are ineffective against this state and necessitate long treatment periods for cure.

We developed a screening assay designed to identify compounds active against non-replicating bacteria, with the goal of identifying compounds that target the non-replicating bacteria in vivo that are believed to be responsible for the protracted six-to-nine month drug regimens that are required to cure TB infection. We have now screened over 400,000 molecules in this assay and have identified several promising scaffolds for which we are working on identifying their mechanisms of action. Studies include using these as in vivo probes to determine the ability of such compounds to shorten the necessary duration of therapy. These compounds are also part of the Broad-GSK collaboration.

Genetic Library Construction

TB genetics is slow and difficult. It can take four to six months to construct a single mutant to evaluate its role in infection. This poses a considerable barrier to TB research and to the development of a systematic, comprehensive understanding of the requirements for infection and disease. At the Broad, we are taking a systematic approach to TB genetics, creating libraries en masse that will enable us to evaluate the contribution of every gene to the growth and survival of Mtb in conditions relevant to human tuberculosis, including exposure to antibiotics and growth and survival in host cells. Once completed, these tools will be made readily available to the TB research community at large, through mechanisms (to be determined) that maximize access both in the US and abroad.

We are constructing the following three unique, arrayed libraries of Mtb strains that will enable us to functionally characterize the complete Mtb genome.

  • Loss-of-function mutants in genes non-essential for in vitro growth
  • Knockdown strains in which the levels of an essential protein can be modulated
  • Overexpression strains in which bacterial proteins can be overproduced in a controlled manner

Host-Pathogen Interactions

One of the unique features of TB infection is that patients' have extremely heterogeneous responses to infection, ranging from clearance of infection, development of latent (asymptomatic) infection, to development of acute, life-threatening disease. Further, about 10% of those that develop only latent infection will develop reactivation disease at some point in their life, moving them to the category of life-threatening illness. The determinants of these different outcomes are currently unknown but represent a potentially critical avenue of investigation if we can identify these determinants and modulate them in favor of clearance of infection.

Researchers at the Broad are poised to capitalize on recent advances in genomic technologies developed at the Institute, specifically the Klarman Cell Observatory, to study the factors that determine infection outcome. Lines of inquiry include development of:

  • Methods to characterize the expression profile of individual host cells in response to infection and to correlate it with infection outcome. The research aims to shed light on the biology underlying the vast heterogeneity seen in TB patient outcomes. This work was performed in a model system of another pathogen, Salmonella, and we now intend to apply these methods in order to study the regulatory networks and responses of host cells infected with TB, to understand the host determinants of infection outcome using single cell host expression analysis.
  • Methods to simultaneously characterize Mtb expression states during infection and correlate them with infection outcome, as the pathogen likely also plays an important role in determining infection outcome.
  • Increasingly complex models of TB-macrophage interaction that recapitulate in vivo complexity, including macrophage activation and the inclusion of additional cell types (i.e. T cells and adaptive immunity).