Tuberculosis (TB) remains a global health problem throughout the developing world, causing 2–3 million deaths annually. In the past 25 years, the problem has worsened, primarily as a result of the growing epidemic of human immunodeficiency virus (HIV). TB is caused by Mycobacterium tuberculosis, a bacterium that is distinguished from other pathogens by its unique and complex lipid surface and its slow replication time. M. tuberculosis has evolved numerous ways of evading both the innate and adaptive immune responses.
Ominously, M. tuberculosis continues to evolve to evade drug killing mechanisms. In 2005, 42 HIV-positive people entered the hospital in Tugela Ferry in KwaZulu-Natal, South Africa, and were diagnosed with TB, which is generally treatable even when the patient is HIV positive. All of these patients died, however, with an average time to death of 16 days. They shared a common infection, XDR-TB, caused by an extensively drug-resistant (XDR) strain that was resistant to at least four and possibly all seven TB drugs. Unfortunately, since that time, the incidence of XDR-TB has been increasing at alarming rates throughout the province of KwaZulu-Natal.
Evolution has caused the emergence of mutations that confer a survival advantage in the face of killing drugs; furthermore, it is unclear whether these strains have also acquired increased virulence. M. tuberculosis that causes XDR-TB is emerging as a more significant pathogen in the 21st century than other M. tuberculosis strains. Unfortunately, there exist no FDA-approved drugs for treating XDR-TB, and the existing TB vaccine is inadequate. Clearly, novel interventions are needed to control both drug-sensitive and drug-resistant TB.
Advanced Genetic Tools: Novel Reporter Phages, Transposon Delivery, and Specialized Transduction
The mycobacteriophage TM4 has proved to be a highly useful phage for genetically manipulating M. tuberculosis and M. bovis because it has been engineered into a highly efficient delivery system. Genetically manipulable versions can be generated by introducing an Escherichia coli cosmid into a nonessential region of this phage. The resulting chimeric molecule is a shuttle phasmid that replicates in an E. coli plasmid, can be packaged into bacteriophage lambda heads, and replicates in mycobacteria as a phage. To prevent killing of the recipient cell, we isolated temperature-sensitive mutations within the mycobacteriophage genome. This allows for replication of the phage at a permissive temperature, but not at a nonpermissive temperature, which enables the delivery of the cargo genes without leading to death of the recipient cell.
The cargo genes to be delivered include reporter genes such as those that encode luciferase and green fluorescent protein; transposons; and allelic exchange substrates. By delivering a reporter gene, such as firefly luciferase, we are able to determine more rapidly than before whether an M. tuberculosis cell is alive or dead after being treated with drugs. While the luciferase reporter test was able to accurately assess drug susceptibility of M. tuberculosis in South Africa, it still required 7–10 days to grow the cultures. In the past year, in collaboration with Graham Hatfull (HHMI Professor, University of Pittsburgh), we have engineered phages containing green fluorescent protein and demonstrated that we can detect viable M. tuberculosis cells within hours of infection. Because the XDR-TB strains can be fatal within a matter of a few weeks, our goal with this improved reporting system is to be able to use sputum samples from patients suspected of harboring drug-resistant mutations to assess drug susceptibility rapidly. The value of such an assessment lies in its ability to determine quickly if the patient has XDR-TB. If so, an appropriate treatment regimen or isolation can be initiated. Efforts are currently under way to use this system directly on patients in South Africa to make rapid assessments of which drugs would be effective in curing specific M. tuberculosis infections.
The phage delivery system has been a workhorse to generate mutants in M. tuberculosis and M. bovis. This is achieved by delivering either a transposon or an allelic exchange substrate. Transposon libraries are readily generated following infection of M. tuberculosis or M. bovis cells with the delivery phage, and the resulting drug-resistant mutants are cells in which the transposon is randomly inserted into the genome of tubercle bacilli. Alternatively, we can precisely delete specific genes in the genome, using a methodology called specialized transduction, to ascertain the function of the deleted gene. The robustness of this system allows us to make a set of precise deletions (knockouts) of M. tuberculosis.
Control of Tuberculosis
Control of TB can be accomplished either by drugs or through an effective immune response. XDR-TB has evolved in populations that are immunocompromised, primarily HIV-infected individuals, to be resistant to all current TB drugs. Basic knowledge about the biology of M. tuberculosis and its interaction with immune systems is essential to the development of both chemotherapies (new drugs) and immunotherapies (new prophylactic and possibly therapeutic vaccines). Such knowledge can be achieved by genetic manipulation of both the tubercle bacillus and the mammalian host.
Control of TB by Drugs: Lessons from Isoniazid
Antibiotics revolutionized the treatment of bacterial infections in the 1940s, starting with penicillin. In 1947, Selman Waksman was the first to discover streptomycin as an antibiotic that was active against M. tuberculosis. However, it was soon discovered that monotherapy is not effective, since M. tuberculosis quickly evolved resistant mutants. In 1952, three groups discovered that isoniazid (INH) had excellent activity against M. tuberculosis, and it became a second effective drug.
While drug resistance compromises drug therapy, it also provides clues to the mechanisms of resistance to the drug, as well as clues to the drug's own mechanism of action. This knowledge provides us with strategies for developing more effective therapies. Acquisition of that knowledge, however, awaited the development of genetic techniques for investigating M. tuberculosis, which our lab achieved for the first time in the late 1980s.
In the early 1990s, our isolation of genetically transformable M. smegmatis strains led to the discovery that INH is a prodrug that is activated by catalase peroxidase. Our early genetic studies had demonstrated that INH and another TB drug, ethionamide (ETH), share a previously unknown common target encoded by a gene we named inhA. Subsequent studies revealed that inhA encodes an enzyme that is an enoyl reductase, a critical enzyme in mycolic acid biosynthesis. In collaboration with James Sacchettini (Texas A&M; University), we used x-ray crystallographic analysis to show that activated INH forms an adduct with the cofactor NAD (nicotinamide adenine dinucleotide), a novel paradigm for drug action. Additional genetic studies have identified novel mutations in previously unstudied genes that confer resistance to INH or ETH. Subsequent biochemical analyses have revealed that these mutations either alter NADH/NAD ratios inside the cell or are required for ETH activation, in either case fully supporting the novel model of action we have proposed.
By isolating temperature-sensitive mutations in inhA, we have demonstrated that inhA inactivation is sufficient to induce cell lysis and death in mycobacterial cells, proving that InhA is an excellent drug target. Knowledge of this entire mechanism has allowed us to identify novel compounds from combinatorial chemical libraries that inhibit inhA-encoded enoyl reductase, yet do not require activation, and that kill INH-resistant cells.
The inhA model can thus serve as an attractive template for developing new drugs that attack strains of XDR-TB.
Control of TB by Immune Effectors: Lessons from BCG and Immune Evasion Mechanisms
Vaccines have proved to be the most effective interventions in controlling infectious diseases since Edward Jenner's discovery of the smallpox vaccine in the late 18th century. BCG (bacille Calmette-Guérin), which was developed as an attenuated strain of M. bovis in 1904, has been administered to half the world's population since the 1940s. Unfortunately, BCG has not proved to be effective in controlling the global TB epidemic. Since BCG had been shown to be effective in early clinical trials, it was postulated that BCG had lost its efficacy following years of propagation in the laboratory. To test this hypothesis, we and others reconstructed the BCG mutants in both M. tuberculosis and M. bovis by specialized transduction, and found these mutants to be limited in their ability to control M. tuberculosis infections. Therefore, we hypothesized that M. tuberculosis has evolved numerous mechanisms to evade killing by both innate and adaptive immunity. An understanding of these mechanisms at the molecular level will lead to the development of novel immunotherapies.
M. tuberculosis is a highly successful pathogen because it has learned how to evade host killing mechanisms. Genomic analysis has revealed that the M. tuberculosis genome is composed of more than 4,000 genes, less than a fourth of which are essential for growth of the M. tuberculosis cells in artificial media. Many of the remaining 3,000 genes are likely involved in the evasion of host immune response. Studies have suggested that M. tuberculosis evades immunity in macrophages by inhibiting MHC (major histocompatibility complex) class I and class II presentations and by preventing phagosome-lysosome fusion, intracellular signaling, and cytokine expression and apoptosis. Our group has been developing methods to screen libraries of mutants of M. tuberculosis and M. bovis to identify the genes that encode the diverse strategies for evading immunity. For example, the isolation of many of these genes has been shown to encode enzymes that modify the lipids that cover the surface of tubercle bacilli. One such set of lipids, mycolic acids (the signature lipid molecule of mycobacteria), which are each 80 carbons in length, are decorated with unique chemical structures. Each of these structures, which are all essential for virulence and immune evasion, are encoded by a specific gene, and inactivation of any of these genes leads to avirulence and defects in immune evasion.
Surface molecules are not the only mechanism that is critical for virulence; in addition, the tubercle bacillus actively secretes immune evasion effectors. For example, by identifying the first virulence-specific secretion system of M. tuberculosis, namely secA2, we discovered that this system is required for the secretion of superoxide dismutase (SOD), an immune evasion effector. In collaboration with Steven Porcelli and John Chan (both at the Albert Einstein College of Medicine), we demonstrated that the secreted SOD is required to prevent apoptosis of M. tuberculosis–infected macrophages. Macrophages that now undergo apoptosis with the secA2 mutant present MHC class I antigens to CD8 cells much more efficiently than the virulent parental M. tuberculosis strains, and are thus more immunogenic. Studies in both mice and guinea pigs have shown that this mutant is a more effective vaccine against M. tuberculosis infection than is BCG.
Our current studies are now focused on how mycobacteria evade killing by innate immune mechanisms. By combining mutations that are devoid of specific immune evasion effectors, it should be possible to generate more efficacious vaccines and immunotherapies.