Antituberculous immunity induced in mice by vaccination with living cultures of attenuated tubercle bacilli

Migration inhibitory factor and interferon in the circulation of mice with delayed hypersensitivity

When mice infected with Mycobacterium tuberculosis strain BCG were inoculated intravenously with old tuberculin (OT) or living BCG cells, both migration inhibitory factor (MIF) and interferon appeared in the circulation within a few hours. In such animals, which showed delayed hypersensitivity by footpad tests, as little as 1.5 mg of OT or as few as 1.7 x 10(6) bacteria per mouse were capable of eliciting circulating MIF and interferon. Uninfected animals inoculated with large doses of OT or living BCG cells did not produce MIF or interferon. When nonspecific stimuli such as bacterial lipopolysaccharide (LPS; from Salmonella typhimurium strain LT-2), heat-killed Brucella abortus, Newcastle disease virus (NDV), and polyinosinic acid:polycytidilic acid (poly I:C) were inoculated intravenously into BCG-infected mice, MIF was produced in the circulation of animals challenged with LPS or Brucella but not in those challenged with NDV or poly I:C, although all the stimuli were capable of eliciting an interferon response. The interferon elicited in BCG-infected mice by specific antigen differed in at least one important property from the viral inhibitor produced by the nonspecific stimuli. The interferon which appeared after injection of OT or living BCG cells was destroyed by treatment at pH 2 for 24 hr at 4C, whereas the interferons produced after injection of the nonspecific stimuli were stable under the same conditions. The MIF activity in plasma from sensitized mice inoculated with specific antigen was also destroyed by treatment at pH 2. When mouse plasma containing both MIF and interferon activity was filtered through Sephadex G-100, both mediators were excluded in the same peak fractions. Sensitization of mice with complete Freund adjuvant instead of infection with BCG cells produces a different pattern of response. Although hypersensitive to specific antigen by footpad swelling tests, mice sensitized with complete Freund adjuvant failed to produce MIF or interferon when they were inoculated intravenously with OT or living BCG cells.

BCG--different strains, different vaccines?

After nearly a century of use, BCG vaccines continue to generate controversy and confusion. Their ability to prevent tuberculosis in studies has been inconsistent. When they have been protective, no clear mechanism of action has been established. Furthermore, the existence of different BCG strains has been described since the 1940s. These strains vary according to several laboratory properties, which may or may not translate into a discernible effect on vaccination. With recent genomic comparisons, it is now clear that different BCG vaccine strains have evolved and differ from each other and from the original BCG first used in 1921. Some of these genetic alterations explain certain variations in laboratory properties of BCG. However, these mutations in BCG strains have yet to be shown to affect BCG-associated protection and/or adverse effects. Continuing research is attempting to assess the effect of these genetic alterations on properties of BCG strains, with the goals of suggesting the ideal BCG for vaccination and providing avenues for improvement on existing BCG vaccines.

A point mutation in the mma3 gene is responsible for impaired methoxymycolic acid production in Mycobacterium bovis BCG strains obtained after 1927

BCG vaccines are substrains of Mycobacterium bovis derived by attenuation in vitro. After the original attenuation (1908 to 1921), BCG strains were maintained by serial propagation in different BCG laboratories (1921 to 1961). As a result, various BCG substrains developed which are now known to differ in a number of genetic and phenotypic properties. However, to date, none of these differences has permitted a direct phenotype-genotype link. Since BCG strains differ in their abilities to synthesize methoxymycolic acids and since recent work has shown that the mma3 gene is responsible for O-methylation of hydroxymycolate precursors to form methoxymycolic acids, we analyzed methoxymycolate production and mma3 gene sequences for a genetically defined collection of BCG strains. We found that BCG strains obtained from the Pasteur Institute in 1927 and earlier produced methoxymycolates in vitro but that those obtained from the Pasteur Institute in 1931 and later all failed to synthesize methoxymycolates, and furthermore, the mma3 sequence of the latter strains differs from that of Mycobacterium tuberculosis H37Rv by a point mutation at bp 293. Site-specific introduction of this guanine-to-adenine mutation into wild-type mma3 (resulting in the replacement of glycine 98 with aspartic acid) eliminated the ability of this enzyme to produce O-methylated mycolic acids when the mutant was cloned in tandem with mma4 into Mycobacterium smegmatis. These findings indicate that a point mutation in mma3 occurred between 1927 and 1931, and that this mutant population became the dominant clone of BCG at the Pasteur Institute.

Decreased capacity of recombinant 45/47-kDa molecules (Apa) of Mycobacterium tuberculosis to stimulate T lymphocyte responses related to changes in their mannosylation pattern

The Apa molecules secreted by Mycobacterium tuberculosis, Mycobacterium bovis, or BCG have been identified as major immunodominant antigens. Mass spectrometry analysis indicated similar mannosylation, a complete pattern from 1 up to 9 hexose residues/mole of protein, of the native species from the 3 reference strains. The recombinant antigen expressed in M. smegmatis revealed a different mannosylation pattern: species containing 7 to 9 sugar residues/mole of protein were in the highest proportion, whereas species bearing a low number of sugar residues were almost absent. The 45/47-kDa recombinant antigen expressed in E. coli was devoid of sugar residues. The proteins purified from M. tuberculosis, M. bovis, or BCG have a high capacity to elicit in vivo potent delayed-type hypersensitivity (DTH) reactions and to stimulate in vitro sensitized T lymphocytes of guinea pigs immunized with living BCG. The recombinant Apa expressed in Mycobacterium smegmatis was 4-fold less potent in vivo in the DTH assay and 10-fold less active in vitro to stimulate sensitized T lymphocytes than the native proteins. The recombinant protein expressed in Escherichia coli was nearly unable to elicit DTH reactions in vivo or to stimulate T lymphocytes in vitro. Thus the observed biological effects were related to the extent of glycosylation of the antigen.

Deglycosylation of the 45/47-kilodalton antigen complex of Mycobacterium tuberculosis decreases its capacity to elicit in vivo or in vitro cellular immune responses

A protection against a challenge with Mycobacterium tuberculosis is induced by previous immunization with living attenuated mycobacteria, usually bacillus Calmette-Guérin (BCG). The 45/47-kDa antigen complex (Apa) present in culture filtrates of BCG of M. tuberculosis has been identified and isolated based on its ability to interact mainly with T lymphocytes and/or antibodies induced by immunization with living bacteria. The protein is glycosylated. A large batch of Apa was purified from M. tuberculosis culture filtrate to determine the extent of glycosylation and its role on the expression of the immune responses. Mass spectrometry revealed a spectrum of glycosylated molecules, with the majority of species bearing six, seven, or eight mannose residues (22, 24, and 17%, respectively), while others three, four, or five mannoses (5, 9, and 14%, respectively). Molecules with one, two, or nine mannoses were rare (1.5, 3, and 3%, respectively), as were unglycosylated species (in the range of 1%). To eliminate the mannose residues linked to the protein, the glycosylated Apa molecules were chemically or enzymatically treated. The deglycosylated antigen was 10-fold less active than native molecules in eliciting delayed-type hypersensitivity reactions in guinea pigs immunized with BCG. It was 30-fold less active than native molecules when assayed in vitro for its capacity to stimulate T lymphocytes primed in vivo. The presence of the mannose residues on the Apa protein was essential for the antigenicity of the molecules in T-cell-dependent immune responses in vitro and in vivo.

Characterization of murine monoclonal antibodies specific for the 45/47 kDa antigen complex (APA) of Mycobacterium tuberculosis, M. bovis and BCG

The alanine-proline antigen (APA), representing less than 2% of the released or excreted material during Mycobacterium tuberculosis or bacillus Calmette-Guérin (BCG) growth, is a dominant antigen present during M. tuberculosis infection or BCG immunization. To obtain new tools to dissect the major epitopes of the APA molecules, seven monoclonal antibodies (mAbs) against the purified molecules were developed. Epitope maps of the mAbs were obtained on APA molecules absorbed on plastic surfaces or in solution (BIAcore technology). The mAbs were found to be independent or to be different despite binding to adjacent or overlapping epitopes. In Western blot assays some proteins secreted in culture fluid by M. avium, M. kansasii, M. smegmatis or M. xenopi were also labelled by six of the seven antibodies. Conversely one antibody was specific for the proteins from the M. tuberculosis complex (I10-0,3) demonstrating that the APA molecules have some properties or general conformations that are specific for M. tuberculosis and M. bovis.

Cloning, sequencing, and expression of the apa gene coding for the Mycobacterium tuberculosis 45/47-kilodalton secreted antigen complex

Effective protection against a virulent challenge with Mycobacterium tuberculosis is induced mainly by previous immunization with living attenuated mycobacteria, and it has been hypothesized that secreted proteins serve as major targets in the specific immune response. To identify and purify molecules present in culture medium filtrate which are dominant antigens during effective vaccination, a two-step selection procedure was used to select antigens able to interact with T lymphocytes and/or antibodies induced by immunization with living bacteria and to counterselect antigens interacting with the immune effectors induced by immunization with dead bacteria. A Mycobacterium bovis BCG 45/47-kDa antigen complex, present in BCG culture filtrate, has been previously identified and isolated (F. Romain, A. Laqueyrerie, P. Militzer, P. Pescher, P. Chavarot, M. Lagranderie, G. Auregan, M. Gheorghiu, and G. Marchal, Infect. Immun. 61:742-750, 1993). Since the cognate antibodies recognize the very same antigens present in M. tuberculosis culture medium filtrates, a project was undertaken to clone, express, and sequence the corresponding gene of M. tuberculosis. An M. tuberculosis shuttle cosmid library was transferred in Mycobacterium smegmatis and screened with a competitive enzyme-linked immunosorbent assay to detect the clones expressing the proteins. A clone containing a 40-kb DNA insert was selected, and by means of subcloning in Escherichia coli, a 2-kb fragment that coded for the molecules was identified. An open reading frame in the 2,061-nucleotide sequence codes for a secreted protein with a consensus signal peptide of 39 amino acids and a predicted molecular mass of 28,779 Da. The gene was referred to as apa because of the high percentages of proline (21.7%) and alanine (19%) in the purified protein. Southern hybridization analysis of digested total genomic DNA from M. tuberculosis (reference strains H37Rv and H37Ra) indicated that the apa gene was present as a single copy on the genome. The N-terminal identity or homology of the M. tuberculosis and M. bovis BCG purified molecules and their similar global and deduced amino acid compositions demonstrated the perfect correspondence between the molecular and chemical analyses. The presence of a high percentage of proline (21.7%) was confirmed and explained the apparent higher molecular mass (45/47 kDa) determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis resulting from the increased rigidity of molecules due to proline residues.(ABSTRACT TRUNCATED AT 400 WORDS)

Isolation of a proline-rich mycobacterial protein eliciting delayed-type hypersensitivity reactions only in guinea pigs immunized with living mycobacteria

Effective protection against a virulent challenge with Mycobacterium tuberculosis is induced only by a previous immunization with living attenuated mycobacteria, usually bacillus Calmette-Guérin (BCG). Living and killed bacteria share a number of common antigens. To identify and to purify molecules that are dominant antigens during immunization with living bacteria, a two-step selection procedure was used. Quantitative delayed-type hypersensitivity (DTH) reactions elicited in guinea pigs immunized either with living or with killed BCG were used to select or counterselect antigens present in BCG culture filtrates. Each major fraction eluted from a series of HPLC columns (gel filtration, DEAE, reverse-phase chromatography) was assayed and titrated on guinea pigs of each group. A protein with an unusual amino acid composition (40% proline, 12% threonine) was purified and N-terminally sequenced. To our knowledge, the sequence Thr-Pro-Pro-Xaa-Glu-Xaa-Pro-Pro-Pro-Pro-Gln-Xaa-Val-Xaa-Leu has not been previously reported. The protein was 100-fold more potent on guinea pigs immunized with living bacteria than on guinea pigs immunized with dead bacteria to elicit a DTH reaction.

The antigen 85 complex: a major secretion product of Mycobacterium tuberculosis

The large number of different proteins synthesized by the mycobacterial cell are currently classified and studied in terms of groups of proteins with certain common properties such as physical and chemical characteristics, function, and localization in the mycobacterial cell. Proteins that are actively secreted during culture on synthetic media represent a particular group of great current interest. At least eight proteins secreted by Mycobacterium tuberculosis have been isolated and characterized to various extents. The genes coding for five proteins secreted from M. tuberculosis and/or Mycobacterium bovis BCG have been cloned and sequenced. All of them contain typical signal sequences. The proteins of the antigen 85 complex, which form the main subject of this review, are often the most common proteins in M. tuberculosis culture fluid. The constituents denoted 85A, 85B, and 85C are encoded by three genes located at different sites in the mycobacterial genome and show extensive cross-reactivity as well as homology at amino acid and gene levels. The proteins differ slightly in molecular mass in the 30- to 31-kDa region, and all of them are fibronectin-binding proteins, but the significance of the latter observation and the role of these proteins in mycobacterial physiology and interaction with the infected host remain to be elucidated. The antigen 85 complex proteins are strongly immunogenic in natural and experimental mycobacterial infections in terms of both induction of antibody synthesis and T-cell-mediated reactions. The well-recognized difference in the efficacy of live and dead mycobacterial vaccines should be considered in relation to the group of secreted antigens. After inoculation, live bacteria in vaccines such as BCG multiply in the host, probably releasing several constituents belonging to the class of secreted proteins and hence resulting in more efficient stimulation of the immune system. Secreted mycobacterial antigens are expected to be of particular significance in induction of various immune responses that are responsible for development of protective immunity in some individuals and for clinical symptoms and complications of the ensuing disease in others.

Potency testing of BCG vaccines on white mice: influence of variables on survival time, lung findings and vaccine assessment

In protection tests on white mice vaccinated with BCG vaccine and challenged with a pathogenic strain of Mycobacterium bovis, the survival times are considerably altered by several variables. In the strains of mice used mainly in this study (NMRI and Albany), the median survival time of a group was roughly doubled in the sensitive range of the test system either by a twofold increase in the immunization period, a threefold decrease in the challenge dose or a 100-fold or less increase in the vaccine dose. The shape of the survival curve of an animal group depends on the median survival time achieved. The Gaussian distributions (sum curves) of the logarithms of the individual survival times are near linearity and parallelity in groups of animals which either survive for short or very long periods. In an intermediate range, however, the survival curves show a flatter and sometimes S-shaped course. This intermediate range of survival corresponds to the time at which the lung findings shift from acute to chronic. The occurrence of acute or chronic findings depends on the individual survival time after challenge. The autopsies show that both findings are equally frequent approximately equal to 35 days after challenge. Individual survival times should be evaluated by non-parametric methods due to their non-normal (bimodal) distribution. Evaluation of the gross lung findings supports these results but is, however, less efficient. The discriminating power of the test system can be altered by changes in any of the variables and is best when animal groups attaining less than 20 days median survival time are compared with groups attaining greater than 30 days. A twofold increase in the median survival time generally provides evidence of significance that may already be obtained 30 days after challenge. With a vaccination-challenge interval of 21 days or more, a 50 microliter vaccine dose generally induces a significant increase in the survival times of the vaccinated animals versus non-vaccinated controls. With increasing immunization periods (vaccination-challenge interval), however, a difference in the efficacy of several vaccines or vaccine doses will be evened out.

Experimental infection with Mycobacterium avium, serotype 2, in pigs. 5. The immunizing effect of BCG vaccine against M. avium infection

The immunizing effect of BCG vaccination against infection with M. avium was evaluated in pigs on the basis of clinical and pathological findings and numbers of acid-fast organisms in the tissues.

In experiments with small and large challenge doses i.v. (10⁻² and 5 mg) the vaccinated animals were found to be partially protected. As compared to non-vaccinates, a reduction of viable organisms was found in vaccinates examined 28-31 or 70-73 days after challenge (Table 4), and fewer positive tissues were found in vaccinates than in non-vaccinates (Table 3). The most obvious results were seen in the experiment with a challenge dose of 10⁻² mg i.V., where the number of organisms was consistently smaller in vaccinates than in non-vaccinates (Table 4). In contact infection experiments, the observations in both vaccinated and non-vaccinated pigs were limited and the results difficult to evaluate. There seemed to be a protection, as judged by histopathological and cultural findings. kw|Keywords|k]Mycobacterium avium, Serotype 2; k]BCG vaccination; k]challenge, intravenous, oral; k]pigs

Experimental male genital tuberculosis-appraisal of protective value of BCG vaccination

As part of an experimental study of the spread of urogenital tuberculosis, the development of tuberculous infection was compared in vaccinated and non-vaccinated guinea pigs after challenge with Mycobacterium tuberculosis strain H37Rv into the seminal vesicle. The animals were saccrified 35 days after challenge. Morphological observations indicate that tuberculous infection in an immunized animal is retarded, but spreads by the same routes as in an nonimmunized animal. On microbial enumeration significantly lower numbers of organisms were recovered from the inoculated seminal vesicle and regional iliac lymph nodes of the BCG-vaccinated guinea pigs.

Universal occurrence of antibodies to tubercle bacilli in sera from non-tuberculous and tuberculous individuals

Studies were carried out to detect and quantitate humoral antibodies to components derived from Mycobacterium tuberculosis. Using both quantitative and qualitative primary tests, binding was observed by all sera obtained from normal as well as from tuberculous subjects. There was a significant difference, however, in the amount of binding by sera from controls. There was a high incidence of both IgG and IgM antibodies in all sera.

The universal occurrence of humoral antibodies demonstrated in this study suggests that most, if not all, persons have been sensitized to the tubercle bacillus to one degree or another. Evidence is presented that the immunogens involved may be derived from saprophytic and/or avirulent mycobacteria.

Behavior of attenuated mycobacteria in organs of neonatal and adult mice

The consequences of Calmette-Guérin bacillus (BCG) vaccination were followed in newborn and adult mice. BCG failed to multiply in the organs of adult mice when administered peritoneally. In contrast, extensive multiplication of the vaccine occurred in both splenic and pulmonary tissue after its peritoneal administration to newborn mice. This absence of tuberculostasis occurred during the period when the animal's spleen, lung, and thymus were rapidly growing. Animals achieved similar levels of resistance to virulent respiratory challenge 10 wk after vaccination irrespective of whether the vaccine had been administered when the mice were newly born and BCG had multiplied in vivo or administered when the animals were fully grown and the BCG had remained in the dormant state. Although neonatal infection with BCG was severe, as shown by the large numbers of organisms recovered from the animals' tissues, the animals suffered no mortality or overt signs of disease. Neonatal vaccination did not significantly affect either the animal's growth rate or the gross development of its organs.

Measurement of resistance to experimental tuberculosis in albino mice. The immune phase

Enhanced resistance against experimental tuberculosis was acquired by albino mice after peritoneal vaccination with living BCG. Prolongation of life was used as a measure of resistance against massive intravenous challenge infection. The average degree of prolongation of life was found to be directly related to the size of the vaccinating dose. The ultimate numbers of BCG bacilli recovered from the animal's organs were also directly related to the quantity of vaccine initially administered. It appears therefore that the relative degree of antitubercular resistance achieved by immunized mice is closely related to the vaccinal population present in the animal's organs.

Tuberculous morbidity in Swiss albino mice immunized with BCG

Study was made of the effect of BCG vaccination on the establishment of experimental tuberculous infection in mice. To this end, mice were infected with small quantities of virulent human bacilli administered by the respiratory route. The quantity of bacilli present for inhalation was so controlled as to permit only a finite percentage of unvaccinated animals to become infected. The relative proportion of mice infected by inhaling these small numbers of virulent bacilli was not found to be lowered by prior BCG vaccination. Thus, pulmonary tissue of immunized mice was not able to prevent the initial lodgement of virulent tubercle bacilli. The growth of virulent organisms in infected lungs, however, was significantly retarded by prior vaccination.

The protective effect in bank voles of some strains of BCG

The bank vole, like the guinea-pig, can be immunized against tuberculosis with very small doses of BCG, presumably because BCG is sufficiently virulent for both these species to multiply freely until the immunization of the animal has reached a maximum. Large doses of BCG induce an earlier, but not an ultimately stronger, immunity. However, the smallest immunizing dose is not the same for all BCG strains and this is a feature that may be used to differentiate strains. In the studies reported by the authors, the smallest immunizing dose for the bank vole was estimated for 11 BCG strains, some of them widely used in man. The results indicate potency ratios of 20: 1 between the strongest and the weakest strains. The ranking of strains in terms of the smallest immunizing dose was found to be correlated in most, but not in all, cases with a previously reported ranking of the lethal effects of the various strains in golden hamsters. The implications of these findings for production requirements and for reference preparations of BCG are discussed.

The host response to Calmette-Guérin bacillus infection in mice

Heterologous organisms (L. monocytogenes and S. typhimurium) were used to study the rate of development, magnitude, and persistence of the antimicrobial resistance engendered in mice by vaccination with BCG. These same methods were used to investigate the influence of prior vaccination on the host response to reinfection. The rate of onset and magnitude of the resistance produced by BCG varied with the vaccinating dose. Increased resistance was detected within 48 hr of injecting large numbers of BCG (approximately 10(8) viable units), but concurrent treatment with isoniazid interrupted its further development. An equal number of heat-killed organisms failed to influence host resistance significantly. The development of tuberculin sensitivity was also dependent upon the continued survival of the immunizing population of BCG. When vaccinated mice were reinfected with BCG, host resistance in spleen and liver was rapidly augmented to the accompaniment of striking changes in the morphology and microbicidal activity of the peritoneal macrophages. These changes occurred most rapidly in mice with a high level of delayed hypersensitivity at the time of reinfection.

PATHOGENICITY AND IMMUNOGENICITY OF STREPTOMYCIN-DEPENDENT MUTANTS OF BRUCELLA

Simon, Ellen M. (University of Wisconsin, Madison) and David T. Berman . Pathogenicity and immunogenicity of streptomycin-dependent mutants of Brucella . J. Bacteriol. 83: 1347–1355. 1962.—Streptomycin-dependent (Sd) mutants of Brucella suis and B. abortus were avirulent for guinea pigs whether selected in the presence of streptomycin only or streptomycin and normal or immune serum. Administration of large quantities of streptomycin to guinea pigs increased the numbers of organisms which could be recovered, but did not cause the development of progressive infections. Vaccination with Sd mutants of B. abortus diminished the pathological response of guinea pigs infected with a large challenge dose of virulent B. abortus , but equal numbers of organisms were recovered from vaccinated animals and unvaccinated controls. Vaccination with Sd mutants of B. suis protected some guinea pigs from small challenge doses. Immunization by multiple injections or by one injection plus streptomycin was superior to a single inoculation of organisms.

Resistance to reinfection in experimental mouse typhoid

Mice surviving infection with a strain of Salm. typhimurium of reduced virulence developed a progressive resistance to reinfection with virulent strains.

The degree of resistance was greater than that of vaccinated mice, although the primary infection had not caused any significant degree of natural selection or higher serum titres of O antibody.

The previously infected animals responded to reinfection with a virulent strain of Salm. typhimurium by gradually restraining bacterial growth. Many of the survivors eradicated the reinfecting strain.

The possible reasons for the difference between vaccinated mice and survivors of previous infection have been discussed.

I am indebted to the Medical Research Council for the provision of a grant towards the expenses of this investigation and to Mr T. Norris and Miss Italia S. Alderton for technical assistance.

Antituberculous immunity induced by methanol extracts of tubercle bacilli; its enhancement by adjuvants

It is possible to prepare from tubercle bacilli a fraction soluble in methanol which is capable of eliciting in mice a marked degree of resistance against virulent tuberculous infection. The immunity was evident whether the infective dose was large and caused a disease with a rapid course, or was very small and caused a disease of many months duration. Active material has been obtained by extraction with methanol at 55 degrees C. of bacterial cells killed with 2 per cent phenol, and washed with acetone. The methanol extracts used in the present study have been prepared from the phenol-killed cells of a culture of BCG, and of the avirulent culture H37Ra. Vaccination of mice bas been carried out by the intraperitoneal route, and the challenge infection (with a highly virulent bovine culture), by the intravenous route. Weight for weight, the protective activity of the methanol extract is smaller than that of the bacterial cells from which it is extracted, but its primary toxicity for mice is also considerably lower. The protective activity can be increased, and the immunity prolonged, by using certain adjuvants as vehicle for injection of the vaccine. An oil adjuvant mixture, and small amounts of a highly purified preparation of the somatic antigen of typhoid bacilli, have been found capable of enhancing and prolonging the antituberculous immunity induced by the methanol extract. Under appropriate conditions the resistance resulting from intraperitoneal injection of the methanol extract is of the same order as that which follows vaccination with whole killed tubercle bacilli or with living BCG.

The protection of infants against tuberculosis

Images

Antituberculous immunity in mice vaccinated with killed tubercle bacilli

The resistance of white mice to tuberculous infection could be increased by preliminary vaccination with small amounts of tubercle bacilli killed by contact with 2 per cent phenol. Vaccine prepared from a variant strain of human tubercle bacilli unable to multiply in vivo (H37Ra) proved as active as vaccines prepared from either virulent or attenuated strains. The immunity induced by phenol-killed bacilli persisted for several weeks. Under the conditions of the experiments, however, it was never able to bring about the death of the virulent bacilli used for the challenge infection, even when the infective inoculum was very small. Its protective effect could be detected (a) by the increased survival time of mice infected with a very large dose of virulent bacilli, and particularly (b) by the lower numbers of bacilli present in the organs of mice sacrificed at various periods of time after injection of sublethal infective doses. Under the proper conditions of vaccination the immunity produced in mice by phenol-killed cells of avirulent bacilli was of the same order as that produced by BCG. The protective antigen proved to be susceptible to heat, particularly at acid reactions. It retained its activity when the bacilli were disintegrated and rendered non-acid-fast by grinding with concentrated phenol. It remained in the insoluble cellular debris when the bacilli were extracted with 88 per cent phenol. Reasons are presented to support the view that the antigenic components present in the tubercle bacilli (avirulent as well as virulent) killed with phenol play a significant part in several manifestations of increased resistance to tuberculosis.