The ∆MSMEG4722 mutant also showed a slightly slower growth rate than the parental mc2155
strain (Figure 2C; the OD600 values at 24h correspond to 2×108 and 107 colony forming units/ ml
for mc2155 and ∆MSMEG4722 respectively). This wasn’t surprising as a similar growth defect
was observed when the homologous gene was deleted in C. glutamicum, and a genome-wide
transposon screen predicted that the loss of the homologous gene in M. tuberculosis would result
in a slow growth phenotype (Lea-Smith et al., 2007; Sassetti et al., 2003). Additionally, when
grown in LB broth, the mutant showed an increased sensitivity to the lipophilic antibiotic
rifampicin (MIC=0.125 μg∕ml) as compared to the parental strain mc2155 (MIC=16 μg∕ml), but
not to hydrophilic antibiotics like isoniazid and ethambutol. Wild type characteristics were
restored on complementation of the ∆MSMEG4722 mutant with plasmid-borne MSMEG4722
indicating that the observed phenotypes in the mutant strain were solely due to the loss of
MSMEG4722 (Figure 2B and C).
The ∆MSMEG4722 mutant failed to synthesize mature mycolic acids
The predicted role of MSMEG4722 in mycolic acid motif formation and the observed changes in
the colony morphology of the ∆MSMEG4722 mutant prompted us to examine mycolic acids in
the mutant strain. If MSMEG4722 was indeed the reductase catalyzing the conversion of the
post Pks13, α-alkyl, β-oxo fatty acyl intermediate, then the ∆MSMEG4722 mutant would be
expected to accumulate this unreduced intermediate of mycolic acid biosynthesis (Figure 1A). A
standard procedure for release of mycolic acids from mycobacteria involves base hydrolysis of
cells using tetrabutyl ammonium hydroxide (TBAH). This is followed by phase-transfer
catalyzed derivatisation using methyl iodide that results in the formation of mycolic acid methyl