Lack of production of the 19-kDa glycolipoprotein in certain strains of Mycobacterium tuberculosis.

The 19-kDa glycolipoprotein of Mycobacterium tuberculosis (PT19) is a prominent antigen recognized by both T cells and antibodies from tuberculosis patients. We report here that two strains, I2646 and S1, when grown either in bacteriological culture or during infection of mice, do not produce this constituent, as judged by ELISA and Western blot assays. Southern blot analysis of the chromosomal DNA showed that both strains displayed the restriction fragment as in H37Rv DNA, suggesting the lack of gross gene alterations. Sequence analysis revealed multiple microlesions including small deletions, point mutations and nucleotide insertions, leading to either premature termination or alteration of open reading frame in both strains. Transformation of both mutant strains with the wild-type gene on a multicopy plasmid resulted in overproduction of native PT19. Infection of mice suggested that the I2646 is of low virulence and that the transformant-producing native PT19 exhibited higher virulence, as assessed by viable counts and gross lesions in the infected organs. The mechanisms and significance of the lack of PT19 production in certain M. tuberculosis strains is discussed.

Guinea-pig alveolar macrophages kill Mycobacterium tuberculosis in vitro, but killing is independent of susceptibility to hydrogen peroxide or triggering of the respiratory burst.

Alveolar macrophages from the lungs of guinea-pigs that had been vaccinated, boosted and then intravenously challenged with Mycobacterium microti or Mycobacterium bovis BCG killed tubercle bacilli phagocytosed in vitro. The killing was modest, about 40% of phagocytosed bacilli were killed in a day, but alveolar macrophages from animals that had been vaccinated and boosted but had not received the intravenous challenge did not kill bacilli. Different strains of tubercle bacilli had different degrees of susceptibility to these activated macrophages but there was no correlation between killing by macrophages and mycobacterial susceptibility to killing by hydrogen peroxide. The different strains of tubercle bacilli triggered peroxide release from these macrophages but there was no correlation with susceptibility to killing by macrophages or with virulence in the guinea-pig. However, phagocytic uptake of these strains by the activated macrophages was inversely correlated with virulence, and uptake by activated macrophages was less than uptake by normal macrophages.

Specificity of antibodies to immunodominant mycobacterial antigens in pulmonary tuberculosis.

A serological survey was performed in groups of patients with active sputum smear-positive or smear-negative pulmonary tuberculosis, healthy household contacts, and controls. Sera were tested for titers of antibodies which bound to each of five purified mycobacterial antigens by enzyme immunoassay and for competition of binding to single epitopes, using six radiolabeled monoclonal antibodies directed toward corresponding molecules. The evaluation of diagnostic specificity was based on a positive score represented by titers above the cutoff point of 2 standard deviations above the mean titer of a control group. For smear-positive samples, the best sensitivity (83%) was achieved by exclusive use of the 38-kilodalton (kDa) antigen or its corresponding monoclonal antibodies. For smear-negative samples, levels of antibodies binding to the 19-kDa antigen showed a lower sensitivity of 62% compared with the control group or 38% compared with the contact group. Titers of antibody binding to the 14-kDa antigen were raised in Mycobacterium bovis BCG-vaccinated contacts, indicating that the greatest potential of this antigen may be in the detection of infection in a population for which tuberculin testing is unreliable. The results demonstrated the differing antibody responses to each of the tested antigens and distinct associations with the stage of infection or disease.

Activation of macrophages for antimycobacterial activity.

The role of macrophage hydrogen peroxide in defense against mycobacterial infection is discussed - not because this is necessarily any more important than other macrophage products (reviews: [1, 2]) but more as a reflection of recently available information. First, we will describe an accumulation of evidence that peroxide has a role, then we will stage an enquiry into whether interferon-gamma (IFN gamma) activates macrophages for peroxide release, and finally we will give an appraisal of direct peroxide toxicity as the mechanism of killing Mycobacterium tuberculosis by activated macrophages.

Hydrogen peroxide and superoxide release by alveolar macrophages from normal and BCG-vaccinated guinea-pigs after intravenous challenge with Mycobacterium tuberculosis.

The release of H2O2 and .O2- by alveolar macrophages from normal and BCG-vaccinated guinea-pigs was measured 3 h, 3 days and 6 days after i.v. challenge infection with Mycobacterium tuberculosis H37Ra. Vaccination did not affect the release of H2O2 or .O2- form macrophages that were removed from guinea-pigs 3 h after i.v. infection and tested as monolayers without a phagocytic stimulus. However, macrophages that were removed from vaccinated animals on the third and sixth days after i.v. infection released progressively more than macrophages that were removed after 3 h. This was not seen with cells from i.v.-infected normal animals. Exposure of macrophage monolayers to phorbol myristate acetate (PMA) and opsonized H37Ra caused increased release of H2O2 and .O2-. There was no difference in the response to PMA either between macrophages from normal animals and those from vaccinated animals or between macrophages taken 3 h, 3 days and 6 days after i.v. infection. Thus the response with PMA gave no indication of the development of local immunity. In contrast, with H37Ra as a phagocytic stimulus in vitro the amounts of H2O2 and .O2- released per cell-associated bacillus increased with the time elapsed since i.v. infection. This increase was greater with the macrophages from vaccinated animals than those from normal animals. The results support the hypothesis that H2O2 production by macrophages is involved in killing M. tuberculosis in vivo.

The contribution of hydrogen peroxide resistance to virulence of Mycobacterium tuberculosis during the first six days after intravenous infection of normal and BCG-vaccinated guinea-pigs.

The course of infection with Mycobacterium tuberculosis strains H37Rv, H37Ra and their isoniazid-resistant, hydrogen peroxide-susceptible mutants in guinea-pig spleen and lung were assessed by measuring changes in number of viable bacteria during the first and second 3-day intervals after i.v. infection of normal and BCG-vaccinated animals. Vaccination had no effect on bacterial survival in the first 3 days of infection. The peroxide-susceptible mutants were killed or inhibited more than their parent strains; in normal animals this enhanced susceptibility was expressed equally during the first and second 3-day intervals while in vaccinated animals the effect was greater in the second 3-day interval. The results suggest that hydrogen peroxide is generated in significant amounts in the environment of tubercle bacilli lodged in normal tissues and in enhanced amounts when acquired immunity becomes expressed after a few days' lodgement in the tissues of vaccinated animals. Thus hydrogen peroxide resistance may contribute to virulence by protecting against both normal resident and immunologically activated macrophages.

The susceptibility of strains of Mycobacterium tuberculosis to catalase-mediated peroxidative killing.

At low pH and with continuous low concentrations of hydrogen peroxide generated in situ, catalase was able to replace peroxidase in the peroxidase/hydrogen peroxide/iodide microbicidal system. The system was effective against Escherichia coli and Mycobacterium tuberculosis. Iodide could not be replaced by chloride. The system was effective in lactate buffer, but not in citrate/phosphate buffer. Strains of M. tuberculosis with high and low virulence were equally susceptible. The observations are discussed in the context of an involvement of host-cell catalase in a possible intracellular killing mechanism against M. tuberculosis.

Analytical subcellular fractionation of rabbit alveolar macrophages with particular reference to the subcellular localisation of pyridine nucleotide-dependent superoxide-generating systems and superoxide dismutase.

Normal and Bacillus Calmette-Guerin (BCG) vaccine-induced rabbit alveolar macrophage homogenates were fractionated by isopycnic density gradient centrifugation. Superoxide dismutase-inhibitable NAD(P)H-dependent nitro-blue tetrazolium reductase was found localised to endoplasmic reticulum and mitochondria. The normal macrophages tended to contain more of this activity than the BCG-induced macrophages. Two superoxide dismutases were found: cyanide-sensitive superoxide dismutase was predominantly present in the cytosol, with a small proportion in mitochondria; cyanide-resistant superoxide dismutase was found confined to mitochondria. Neither differed in specific activity betw-en the normal and BCG-induced macrophages.

Phagolysosome formation, cyclic adenosine 3':5'-monophosphate and the fate of Salmonella typhimurium within mouse peritoneal macrophages.

Salmonella typhimurium did not inhibit fusion of lysosomes with the phagocytic vacuoles in infected macrophages and caused no increase in cyclic adenosine 3':5'-monophosphate. Glutaraldehyde-killed bacteria showed rapid ultrastructural degeneration within the phagolysosomes. In contrast, untreated bacteria were resistant to digestion by lysosomal enzymes. Intracellular survival of this species appears to depend on resistance to, and not evasion of, lysosomal enzymes.

Phagosome-lysosome fusion and cyclic adenosine 3':5'-monophosphate in macrophages infected with Mycobacterium microti, Mycobacterium bovis BCG or Mycobacterium lepraemurium.

When ingested by mouse peritoneal macrophage monolayers, live Mycobacterium microti caused a sustained increase in monolayer cyclic AMP content and fusion of lysosomes with the bacterium-containing phagosomes was impaired. Ingested live M. bovis BCG caused a transient increase in cyclic AMP and the defect in phagolysosome formation was less pronounced. Dead mycobacteria and live M. lepraemurium neither enhanced monolayer cyclic AMP content nor inhibited phagolysosome formation. Mycobacterium microti and BCG exceeded M. lepraemurium in cyclic AMP-synthesizing activity in vitro but the question of whether bacterial cyclic AMP contributed substantially to the increments in infected macrophages was not resolved. Antibody-coated BCG retained the ability to synthesize cyclic AMP and to enhance monolayer cyclic AMP but lost the ability to inhibit phagolysosome formation in macrophages, The observations are discussed in terms of possible control of phagolysosome formation by cyclic nucleotides.

Virulence of Mycobacterium tuberculosis and susceptibility to peroxidative killing systems.

At sub-bactericidal concentrations of hydrogen peroxide, Mycobacterium tuberculosis was killed by hydrogen peroxide/peroxidase/halide microbicidal systems. The halide cofactor could be either iodide or, with much lower efficiency, chloride. Omission of any one of the reactants eliminated the tuberculocidal effect. Differences in susceptibility between different strains of M. tuberculosis did not correlate with virulence differences. The observations are discussed in the context of host defence mechanisms against tuberculosis.

Virulence and resistance to superoxide, low pH and hydrogen peroxide among strains of Mycobacterium tuberculosis.

Six strains of Mycobacterium tuberculosis of different virulence in guinea-pigs were compared with regard to their resistance to low pH, to hydrogen peroxide (H2O2) at different pH values and to superoxide (O2-). Low virulence was associated with susceptibility to H2O2 in native and isoniazid-resistant strains but not in laboratory-attenuated strain H37Ra. H2O2 resistance was only partly related to catalase content. Low virulence was not associated with susceptibility to an acid environment but the tuberculocidal effect of H2O2 was significantly increased at low pH. The strains were uniformly resistant to O2- and contained similar amounts of superoxide dismutase. The implications of these observations are discussed in the context of mechanisms of host defence in tuberculosis.

The relationship between nicotinamide adenine dinucleotide concentration and antibacterial activity of isoniazid in Mycobacterium tuberculosis.

The relationship between the antibacterial effect of isoniazid and the intracellular concentration of nicotinamide adenine dinucleotide (NAD) was investigated in Mycobacterium tuberculosis strain H37Rv given continuous and pulsed exposures to the drug. Depletion of NAD to a plateau value occurred rapidly during exposure, and recovery after a pulse of isoniazid was also rapid. It seemed unlikely that NAD depletion was the direct cause of the antibacterial activity because (1) insufficient depletion occurred at low isoniazid concentrations; (2) antibacterial activity, but not NAD depletion, was proportional to the product of isoniazid concentration and the exposure period; and (3) NAD depletion was not related to antibacterial activity in cultures of differing physiologic state.