Attenuated mutants of Mycobacterium tuberculosis hold promise as potential vaccine candidates for the prevention of tuberculosis. To pursue this goal, two groups of researchers have created auxotrophs of M. tuberculosis and tested their infectivity and ability to protect animals from challenge with virulent tuberculosis. By allelic exchange, Jackson et al. created two purine auxotrophs, and Hondalus et al. created a leucine auxotroph of M. tuberculosis. These strains were unable to replicate in macrophages in vitro and, when used to infect mice, virulence was substantially attenuated. When tested for protective immunity to a challenge of aerosolized virulent M. tuberculosis in the guinea pig model, the purine auxotrophs conferred protection comparable to that of conventional bacillus Calmette-Guerin (BCG) in limiting growth of M. tuberculosis in the lungs, but were less efficacious than BCG in limiting growth of M. tuberculosis in the spleen. In a mouse model, the leucine auxotroph conferred protection after vaccination that was equivalent to that of BCG in prolonging survival after intravenous infection of virulent M. tuberculosis, but the leucine auxotroph vaccine was less effective than BCG in reducing the organ burden of infection after intravenous or aerosolized challenge. These studies emphasize the challenge of achieving an optimum balance of attenuation and immunogenicity for tuberculosis vaccine development. Future studies are needed to determine whether larger immunizing doses or booster immunizations can improve the immunogenicity of these attenuated strains.

Leila Alland

A recent study has shed light on a long-established relationship between the morphology and virulence of tuberculosis. In his original description of Mycobacterium tuberculosis, Koch noted that the bacteria formed braided microscopic bundles, a morphology that was later called �cording�. Further studies demonstrated an association between the virulence of an individual mycobacterial strain and the strength of its cording, indicating that components of the cell envelope are responsible for both cording and virulence, but these factors had not yet been identified. To elucidate the genetic basis for cording and virulence, Glickman et al. undertook a genetic screen for mutants that failed to form cords. They identified a previously unknown mycobacterial gene, pcaA, which they showed was required for synthesis of the proximal cyclopropane ring of the main cell wall mycolic acid of both bacillus Calmette-Guerin and M. tuberculosis. When mice were infected with the pcaA mutant of M. tuberculosis, the pcaA mutant replicated normally in vivo but failed to persist within and kill infected mice. These results establish the involvement of cyclopropane synthetases in the virulence and persistence of M. tuberculosis infection.

Leila Alland

Not only are cholesterol and TACOs bad for your arteries, they may also promote entry of pathogenic mycobacteria into your cells. Mycobacteria such as Mycobacterium tuberculosis are internalized by host macrophages and are able to survive intracellularly because of their ability to recruit and retain a protein called TACO (tryptophane aspartate-containing coat protein) at the microbacterial phagosome. TACO prevents vesicle fusion with the lysosomes that would normally destroy intracellular bacteria. The mechanism by which TACO binds the phagosomal membrane was a mystery, as TACO lacks a transmembrane domain. In a recent issue of Science, Gatfield and Pieters reported that TACO associated with the membrane in a cholesterol-dependent manner, and that cells depleted of cholesterol were unable to internalize mycobacteria. The authors proposed that mycobacteria preferentially enter cells at cholesterol-rich areas of the cell�s plasma membrane and are subsequently sequestered in TACO-coated phagosomes, preventing lysosomal delivery and promoting bacterial survival. They suggested that agents designed to target the cholesterol-mediated entry of M. tuberculosis may be developed as tuberculosis therapeutics.

Kris Novak

Identification of a toxin produced by Mycobacterium ulcerans may provide clues to the identity of M. tuberculosis virulence factors. Although most pathogenic bacteria produce toxins, which destroy tissues and modulate the immune response, an M. tuberculosis toxin has not yet been identified. Recently, George et al. reported the isolation of a cytotoxin from a related bacterium, M. ulcerans, the causative agent of Buruli ulcer. This is a rare West African disease characterized by progressive necrotic skin lesions. The authors isolated a polyketide-derived macrolide, which they called �mycolactone�, from a mycobacterial extract and showed that it was cytotoxic and caused cell cycle arrest in cultured fibroblasts. Intradermal injection of guinea pigs with the purified toxin produced a lesion similar to that of Buruli ulcer in humans. The authors suggested that this toxin may be one member of a family of virulence factors associated with mycobacterial diseases. The M. tuberculosis genome sequencing project has demonstrated many polyketide synthesis genes, indicating that M. tuberculosis produces a related toxin, which may be responsible for the tissue destruction associated with human disease. Finding an M. tuberculosis homolog of this toxin will be a focus of future research.

Kris Novak

Understanding why infection with Mycobacterium tuberculosis imparts only a 10% chance of developing clinical disease could advance the search for a treatment for tuberculosis. Although the environmental factors involved in infection are evident—malnutrition, poverty and overcrowding—studies on human vulnerability to tuberculosis have demonstrated a genetic component. The balance between resistance and susceptibility involves many genes inherited in a complex manner.

A recent paper in PNAS adds to this knowledge. Using a mouse model of tuberculosis, scientists at Harvard School of Public Health identified a locus on mouse chromosome 1 that has a considerable effect on susceptibility to Mycobacterium tuberculosis. The locus, sst1 (for susceptibility to tuberculosis 1), is distal to a previously identified gene that controls resistance of mice to attenuated BCG—Nramp1. The phenotypic expression of the two differed in that sst1 controlled progression of infection in a lung-specific manner. Crossover experiments showed that, despite its ability to control infection, the resistant allele of sst1 was not sufficient to confer complete protection against the bacterium, indicating that additional genes outside the locus are also involved in susceptibility/resistance to tuberculosis infection.

Karen Birmingham

The incidence of resistance to anti-tuberculosis drugs—particularly isoniazid and rifampicin—is increasing worldwide and presents a genuine problem in controlling this potentially deadly disease through the use of direct observed short-course treatment. Given the spread of resistant strains of tuberculosis, the development of assays for their rapid detection has taken on a new sense of urgency.

Studies have shown that the DNA sequence of Mycobacterium tuberculosis is highly conserved and that mutations in its genome are almost always associated with drug resistance. To determine whether commonly associated mutations are linked with clinical drug resistance, a team of scientists has used genotypic analysis in two different human populations using molecular �beacons�—fluorogenic reporter molecules within a closed-tube PCR assay. They found that rifampicin-resistance mutations were detected extremely well in isolates from both populations but that detection of isoniazid-resistance mutations was less specific. This may be because the latter seem to be due to a stepwise accumulation of mutations. The authors concluded that, although expensive to do now, molecular �beacon� analysis is a simple and rapid method for the detection of certain resistant M. tuberculosis isolates, and the speed of the technique could help to gain valuable treatment ground, given the length of time it takes to determine drug resistance by conventional culture-based tests.

Karen Birmingham

The current tuberculosis vaccine (bacille Calmette and Guérin, or BCG) was originally derived from a Mycobacterium bovis isolate attenuated by laboratory passage between 1908 and 1921, and since then it has spawned many substrains. The vaccine�s checkered history of protection efficacy has been postulated to be due to changes over time. The advent of the post-genomic tuberculosis era has now enabled researchers at McGill University in Montreal to undertake a global analysis of genetic differences between M. tuberculosis, M. bovis and various BCG substrains encompassing almost every open reading frame. They found that more than 90 M. tuberculosis genes were not present in all of the M. bovis isolates analyzed. Unexpectedly, among the genes present in M. bovis but deleted from BCG strains, there seemed to be no classical virulence elements, but there did seem to be an over-representation of transcriptional regulators, both repressors and activators. The authors speculated that these transcriptional regulators may be involved in adapting to environmental change, such as laboratory conditions, and may account for the progressive loss of immunogenicity in BCG vaccine isolates documented over the course of the twentieth century. Extensions of global analyses such as these will lead to better understanding of virulence determination and should steer efforts to create a better tuberculosis vaccine.

Simon Noble

Macrophages are the main host cell for mycobacterial infection and constitute the first line of defense against spread of infection. When successful, macrophages eliminate mycobacterial infection through a complex network of events involving tumor necrosis factor (TNF)-a production, leading to apoptosis of infected macrophages. Mycobacteria enter macrophages using various receptors, including complement receptors (CR1, CR3 and CR4), mannose receptor, CD14, Fc receptors and scavenger receptors. But a recent study by Fratazzi et al. has now identified a cell surface mucin, CD43, the main sialoglycoprotein on leukocytes, as an important co-receptor and/or cofactor in the mycobacterial uptake process. Mycobacterium avium bound in high numbers to CD43-transfected but not untransfected HeLa cells, and this effect was mycobacterium-specific, as Salmonella typhimurium and Shigella flexneri showed no such difference. Macrophages from wild-type CD43+/+ mice bound and ingested high numbers of M. avium, whereas this binding was negligible in macrophages from CD43-gene-deleted mice. M. tuberculosis and M. bovis also bound to CD43+/+ macrophages but not to CD43–/– macrophages. CD43 was also required for the production of TNF-a by macrophages exposed to M. avium, indicating a functional link between the action of CD43 in mycobacterial binding and/or uptake and induction of TNF-a production. Elucidating such host–microbe interactions will be central to future tuberculosis control efforts.

Charlotte Wang