The world urgently needs a better tuberculosis vaccine. Bacille Calmette-Gu
erin (BCG), an attenuated strain of Mycobacterium bovis, has been very wide
ly used as a vaccine for many years but has had no major effect on reducing
the incidence of tuberculosis. A number of alternative living and non-livi
ng vaccines are being investigated. Live vaccine candidates include genetic
ally modified forms of BCG, genetically attenuated strains of the Mycobacte
rium tuberculosis complex and genetically engineered vaccinia virus and Sal
monella strains. Non-living vaccine candidates include killed mycobacterial
species, protein subunits and DNA vaccines. One requirement for acceptance
of any new vaccine will be a favourable comparison of the protection it in
duces relative to BCG in a range of animal models, some of which may need f
urther development. Molecular genetic techniques are now available that ena
ble production of live attenuated strains of the M. tuberculosis complex wi
th vaccine potential. In the first of two broadly different approaches that
are being used, large numbers of mutants are produced by transposon mutage
nesis or illegitimate recombination and are screened for properties that co
rrelate with attenuation. In the second approach, putative genes that may b
e required for virulence are identified and subsequently inactivated by all
elic exchange. In both approaches, mutants that are attenuated need to be i
dentified and subsequently tested for their vaccine efficacy in animal mode
ls. Many mutants of the M. tuberculosis complex have now been produced and
the vaccine properties of a substantial number will be assessed in the next
3 years.