Identification of variable regions in the genomes of tubercle bacilli using bacterial artificial chromosome arrays

Whole-genome comparisons of the tubercle bacilli were undertaken using orde red bacterial artificial chromosome (BAC) libraries of Mycobacterium tuberc ulosis and the vaccine strain, Mycobacterium bovis BCG-Pasteur, together wi th the complete genome sequence of M. tuberculosis H37Rv, Restriction-diges ted BAC arrays of M. tuberculosis H37Rv were used in hybridization experime nts with radiolabelled M. bovis BCG genomic DNA to reveal the presence of 1 0 deletions (RD1-RD10) relative to M. tuberculosis. Seven of these regions, RD4-RD10, were also found to be deleted from M, bovis, with the three M. b ovis BCG-specific deletions being identical to the RD1-RD3 loci described p reviously. The distribution of RD4-RD10 in Mycobacterium africanum resemble s that of M, tuberculosis more closely than that of M. bovis, whereas an in termediate arrangement was found in Mycobacterium microti, suggesting that the corresponding genes may affect host range and virulence of the various tubercle bacilli. Among the known products encoded by these loci are a copy of the proposed mycobacterial invasin Mce, three phospholipases, several P E, PPE and ESAT-6 proteins, epoxide hydrolase and an insertion sequence. In a complementary approach, direct comparison of BACs uncovered a third clas s of deletions consisting of two M. tuberculosis H37Rv loci, RvD1 and RvD2, deleted from the genome relative to M. bovis BCG and M, bovis, These delet ions affect a further seven genes, including a fourth phospholipase, plcD. In summary, the insertions and deletions described here have important impl ications for our understanding of the evolution of the tubercle complex.