Production of avirulent mutants of Mycobacterium bovis with vaccine properties by the use of illegitimate recombination and screening of stationary-phase cultures.

A better tuberculosis vaccine is urgently required to control the continuing epidemic. Molecular techniques are now available to produce a better live vaccine than BCG by producing avirulent strains of the Mycobacterium tuberculosis complex with known gene deletions. In this study, 1000 illegitimate recombinants of Mycobacterium bovis were produced by illegitimate recombination with fragments of mycobacterial DNA containing a kanamycin resistance gene. Eight recombinant strains were selected on the basis of their inability to grow when stationary-phase cultures were inoculated into minimal medium. Five of these recombinants were found to be avirulent when inoculated into guinea pigs. Two of the avirulent recombinants produced vaccine efficacy comparable to BCG against an aerosol challenge in guinea pigs with M. bovis. One of these recombinants had an inactivated glnA2 gene encoding a putative glutamine synthetase. Transcriptional analysis showed that inactivation of glnA2 did not affect expression of the downstream glnE gene. The other recombinant had a block of 12 genes deleted, including the sigma factor gene sigG. Two avirulent recombinants with an inactivated pckA gene, encoding phosphoenolpyruvate carboxykinase which catalyses the first step of gluconeogenesis, induced poor protection against tuberculosis. It is clear that live avirulent strains of the M. tuberculosis complex vary widely in their ability as vaccines to protect against tuberculosis. Improved models may be required to more clearly determine the difference in protective effect between BCG and potential new tuberculosis vaccines.