The variability in the genome sequence of M. tuberculosis between clinical isolates has been analysed earlier and variability in the number and site of integration of transposable element IS6110 is well documented [1]. There are also reports on the analysis of whole genome SNPs in mycobacteria [2]. Compared to many other bacterial species, M. tuberculosis exhibits very little genomic sequence variation [3]. However, there is increasing evidence that even this limited inter-strain genetic variability is biologically significant [4]. M. tuberculosis infection in animal models has shown a range of immune responses and variable degrees of virulence depending on the infecting strain [5, 6]. In the majority of humans, an effective immune response develops after infection with M. tuberculosis and restricts the spread of the pathogen and clinical manifestation of the disease is seen in less than 10% of those infected. Clinical tuberculosis is influenced by variability in the host's genetic background, immune status, diet, social and environmental factors [7, 8]. However, little is known about the bacterial factors, especially, genetic diversity in bacterial virulence factors contributing to variable host responses.

The expression of mce genes is of importance for the virulence of mycobacteria [9, 10]. The presence of four copies of mce genes in four operons each consisting of eight genes [11] and the differential expression of mce1 and mce4 operons points towards functional importance of these operons [9, 12]. Interestingly, the domain organization in the genes of all the four operons is similar. This conservative arrangement may be of strategic significance to the biology of M. tuberculosis. The antigenic and immunogenic effects of mce proteins in nature suggest that the variation in amino acid sequence of these proteins may affect host response, apart from their effect on functions of these proteins [13, 14]. In the light of these observations, we initiated the present study to understand the possible importance of genetic diversity in the mce operon genes which have a role in the pathogenesis of M. tuberculosis. Polymorphism in the genes of mce1 and mce4 operons in 112 clinical isolates of M. tuberculosis was analysed to understand and relate the effect of the genetic variability to structural changes in the proteins by computational methods.

It has been observed that severity of tuberculosis varies in different patients. It is possible that clinical isolates of M. tuberculosis encountering the human hosts with individual immune systems need to accordingly modulate their virulence associated biological factors to survive within the host. Therefore, it is important to understand the biology of the pathogen at the genetic level. Genetic polymorphisms in the bacterial hosts have been shown to significantly influence the biology of the organisms [17]. In M. tuberculosis, most of the polymorphisms have been studied in the transposable elements and drug resistant genes [1, 18]. A study of the genetic mutations in the genes coding for virulence factors interacting with host's immune system would help us in understanding the ways in which various strains of M. tuberculosis adapt to different hosts.

The sequencing and Sequenom MassARRAY analysis presented here have revealed that mce4 operon is significantly more polymorphic than mce1 operon. Seven out of eight genes of mce4 operon were found to be polymorphic. The mce4 operon is expressed in the stationary phase of in vitro broth culture and in the late phase of infection [9, 12] and has role in the survival of the pathogen inside the host body [19]. This is the time when the bacterium has established itself efficiently in the host and it is possible that the bacterium then permits itself to undergo genetic substitutions to evade the host immune response. The detailed analysis of codon usage for synonymous changes observed in both mce1 and mce4 operons revealed that codons of amino acids were changed to the next preferred codon which would alter the expression of proteins. Our observation of more codon bias in mce4 operon that may lead to less expression of proteins further supports the possibility that such diversity facilitates better survival of M. tuberculosis inside the host's body. Our results further reveal that more than 25% of clinical isolates have SNPs in yrbE4A and lprN genes of mce4 operon. The lprN gene of mce4 operon codes for lipoprotein precursor [20]. The lipoproteins of M. tuberculosis are known to be effectively antigenic in nature [21]. Thus, high polymorphism in lprN gene (both synonymous and nonsynonymous) further supports our hypothesis that such polymorphisms favour intracellular survival of the pathogen. Drug resistance itself makes the organism in a better position to survive within the hostile intracellular environment. But DS isolates being drug susceptible do not have this advantage. Therefore antigenic variation is a tool utilized by DS clinical isolates. For example, the function of PPE proteins is unknown. However several observations and results support that many are cell surface associated and recognized by the host immune system. The possibility of high antigenic variation associated with these highly antigenic PE and the PPE family proteins have also been reported [22]. The PGRS member Rv1759 is a fibronectin-binding protein of relative molecular mass 55,000 Da [23] that elicits a variable antibody response, indicating either that individuals mount different immune responses or that this PGRS protein may vary between strains of M. tuberculosis. Bioinformatics analysis have indicated that LprN is also a cell surface associated protein. Therefore it is possible that SNP observed in this gene could be translated into antigenic variation in the LprN protein to facilitate the intracellular survival of mycobacteria.

In contrast, the mce1 operon is required for the entry of the pathogen inside the host cell [24] and hence, remains less polymorphic. However, the yrbE1A gene is revealed to be highly polymorphic in mce1 operon. Since, YrbE1A has been predicted to be a transmembrane protein [20], so the observed polymorphism in its gene may influence activity of the protein.

From the computational analysis, we could infer that the results obtained on the basis of structural details (PolyPhen) and sequence details (PMut) were in tune with each other. Both the programs have predicted that the SNP observed in mce1A gene is having the highest pathological relevance. Although, programs used here were primarily designed for predicting functional consequences of nonsynonymous SNPs in human proteins. However, concurrent observations on the nonsynonymous SNPs of mce operon proteins reported by both PolyPhen and PMut substantiate our hypothesis further. Energy minimization studies on the structure of Mce1A protein show that Pro359Ser mutation resulted in the loss of α-helical structure in the mutated protein. Analysis of wild and mutated Mce1A protein structures by HB plot indicates that change in hydrogen bonding interaction pattern in the mutant protein lead to conformational changes. Mutation of proline to serine residue in proteins are known to cause structural alterations by the reduction of α-helix content of protein and decreases protein stability and increase its susceptibility to proteolysis by trypsin [25]. Yazyu et al. [26] observed that Pro122Ser mutation could bring about the alteration in the pH of the system by changing the cation specificity of melibose carrier (a membrane bound protein which mediates co transport of α-galactosides with monovalent cations) in E. coli. Pro122Ser mutant lost the ability to utilize H⁺ and made the carrier favorable for Li⁺- melibose co-transport. Serine being a hard Lewis base interacts with hard Lewis acids such as Li⁺ instead of H⁺ [26]. Mce1A protein is a cell surface protein [27] so it may be speculated that the aforementioned changes due to Pro359Ser mutation may have a diminishing effect on the stability of protein and thus on the biological function of it.

In a further analysis, we compared the SNPs in the genes of mce1 and mce4 operons in 59 drug resistant (DR) and 22 drug sensitive (DS) clinical isolates. The comparison of SNPs in the mce genes in DR and DS clinical isolates revealed that both mce1 and mce4 operon genes of DS clinical isolates were more polymorphic than DR clinical isolates. It is possible that while drug resistance provides extra edge to DR isolates, the DS isolates try to enhance their virulence mechanisms and adaptability to hostile intracellular environment by undergoing mutations in them. This is also supported by a report by Shimono et al. [28] where they have demonstrated that, unlike wild type M. tuberculosis, a strain of M. tuberculosis with disrupted mce1 operon become hypervirulent.

Further study of larger number of single and multi drug resistant isolates may give a conclusive answer to the significance of such an observation.

Taken together the SNP analysis and in silico modeling reported here predict that the SNPs in the mce1 and mce4 operons in the clinical isolates are reasonably frequent. Also, the in silico modeling of nonsynonymous SNP in the mce1A gene of mce1 operon indicates that such change may translate into altered function of the gene that may reflect on the virulence and biology of the pathogen.

In the present study, we have investigated the extent of polymorphism in the genes of mce1 and mce4 operons of M. tuberculosis in a panel of four standard reference strains (H37Rv, H37Ra, LVS (Low virulent Strain) and BCG) and 112 clinical isolates. Our results show that SNPs in the coding sequences of mce1 and mce4 operons in clinical isolates can be significantly high. Twenty SNPs were discovered in the two operons out of which 12 were nonsynonymous changes. Further analysis of pathological relevance of these changes revealed that five of the SNPs were deleterious. Overall, mce4 operon is significantly more polymorphic than mce1 operon (p < 0.001). However, nonsynonymous SNPs detected in mce1A gene of mce1 operon predict effect of such SNPs on the biology of the pathogen.