Studies with subtractive hybridization and microarrays have identified 16 regions present in M. tuberculosis H37Rv but are absent in M. bovis BCG [1, 8]. Further deletions of RvD1 to RvD5 and TbD1 genes in H37Rv are also reported . RD1 region comprises of nine genes (Rv3871 to Rv3879c) and spans a 9.5-kb region. In M. bovis BCG, RD1 deletion completely removes seven genes (Rv3872 to Rv3878) and truncates two others (Rv3871 and Rv3879c) .
Of the nine genes predicted within the 9.5 kb RD1 region, those coding for cfp 10 (Rv3874) and esat 6 (Rv3875) are immunogenic and RD1 deletion mutants of M. tuberculosis have been found to be less virulent 
Thus, generation of deletion of genes appear to be a major mechanism for creating genetic diversity.
Based on this, a study from Kerala was carried-out to screen for differences in the RD1 region. Amplification using primers that span the ORF coding for cfp10 and esat 6 was expected to give a 652 bp PCR product. But the PCR results revealed an extra amplicon of 386 bp in 97% of the clinical isolates screened. Further characterization by sequencing and homology search indicated that this region is a part of the moaA3 gene which encodes for molybdopterin cofactor protein A in M.bovis. The PCR primer that was made spanning the RD1 region was shown to be similar to portions of the moaA3 (MT3427) gene in the RvD5 region in clinical isolates and also in CDC1551 which resulted in the amplification of the 386 bp fragment. This amplicon spanned the nucleotides 575 to 960 of the moaA3 gene (MT 3355) in M. bovis 
With this background, we searched for the presence of moaA3 gene among the clinical isolates collected from Tamil Nadu and Pondicherry. Our search showed presence of 652 bp of cfp10 and esat 6 of RD1 region in all the 116 clinical isolates screened but only a limited number of 12/116 (10.3%) clinical isolates showed the presence of 386 bp amplicon of moaA3. Amplification of flanking sequence of 386 bp moaA3 showed the expected 1254 bp product in among all the isolates showed the extra 386 bp amplicon. Thus, in contrast to Kerala study, only in limited proportion of our isolates moaA3 was amplified using the designed primer.
Although Rao et al,  reported total absence of RD1 region in clinical isolates, in a study from Hyderabad, Andhrapradesh (a south eastern coastal state) all our clinical isolates showed the presence of RD1 (cfp 10 and east 6) region. In fact, few isolates we obtained from Hyderabad were found positive for RD1 region in all the isolates (data not shown). This is in concurrence with other reports [11, 4].
In H37Rv, the moaA3 amplicon was shown to be absent. The RvD5 region from which the amplicon was generated is an IS6110 mediated deletion in the type strain H37Rv . IS6110 is a powerful genetic marker for strain differentiation . In general low/no copies of IS6110 were implicated among south-east Asian strains, including India [13–15].
However, a closer scrutiny at the frequency of no/low copies of IS6110 reveals that they are commoner among Kerala strains than those from Tamil Nadu. An analysis from Kerala  reported about 24% of no copies and 39% of single copy of IS6110 in their isolates. Thus, around 62.5% of (50 out of 80) strains analyzed were not type-able by IS6110 based finger printing. This reported to be higher than the small numbers of IS6110 – deficient strains -1%–4% reported in Tamil Nadu [17, 18]. It is appropriate to underscore that all our isolates screened for IS6110 by PCR were found positive.
Adding to this genetic variations, our finding of low frequency of 10.3% of moaA3 gene among Tamil Nadu and Pondicherry isolates compared to the higher proportion of 97% among Kerala isolates, further strengthens the speculation of genetic variation among the strains of Kerala and Tamil Nadu & Pondicherry, the South Indian states.
Further in screening of different genes of RD1 region among clinical isolates from Kerala and Western and Northern India, Rv 3871 and Rv 3872 (part of RD1 region) was reported as high as 98% in Kerala isolates, where as only 30% was reported in the other isolates screened .
This further emphasizes the need to carry out systematic molecular epidemiological studies in these endemic areas to explore any other genetic variations. Further the role of IS6110 or any other insertion sequences or mobile genetic elements in the genetic variation may also be investigated.