Characterization of G-Quadruplex Motifs in espB, espK, and cyp51 Genes of Mycobacterium tuberculosis as Potential Drug Targets

Mishra, Subodh Kumar; Shankar, Uma; Jain, Neha; Sikri, Kriti; Tyagi, Jaya Sivaswami; Sharma, Tarun Kumar; Mergny, Jean-Louis; Kumar, Amit

Considering the suitability of G-quadruplexes as an evolutionarily conserved promising drug target, we searched the genomic sequences of all available strains of M. tuberculosis in the NCBI database for the presence of PTQs (MTB-PGQs) by employing a G-quadruplex (G4) prediction algorithm previously developed by our group. All predictions were then confirmed by tools developed by other groups (Tables S1, S2, S3, S4, S5, S6, S7, S8, S9, and S10). [30] [31] [32] Because G4 structure folding relies on different patterns and length of consecutive G-tracts and loops ( Figure S2 ), we set the constraint of 3 or 4 nt as the minimum length of G-tract, and 0 and 20 nt as the least and highest length of loops, respectively. The G4 prediction algorithm with the above set of parameters was employed on the 160 completely sequenced M. tuberculosis strains (Table S7) , allowing us to identify more than 4,000 potential G4-forming sequences (Tables S1 and S2). Considering that the likeness between two sequences may associate with similar functions, we further sought to correlate the sequence similarity between each of the MTB-PGQs and their function. The unweighted pair group method with arithmetic mean (UPGMA) clustering method was utilized to measure the similarity between a large set of MTB-PGQs that defines the likeness hierarchy between the given set of strings. 33 This clustering analysis clusters PGQs that have almost identical sequences, similar function, and are susceptible to form G4s with the same topology (Tables S3 and S4 ). For combating multidrug-resistant bacterial infection, evolutionary conservation is a requisite that satisfies one of the criteria to work as a promising drug target. Therefore, we employed the BLAST algorithm to calculate the occurrence of each cluster in all 160 available strains of M. tuberculosis. To demonstrate the conservation of these clusters, we have employed the following equation: p = (n O N) Â 100, where p is the frequency of occurrence, n is the total number of strains in which a particular PGQ is present, and N is the total number of M. tuberculosis strains. Table S8 demonstrated the details of the MTB-PGQs occurring with >98% frequency and present in the essential genes of the 160 strains of M. tuberculosis. For further analysis, we have chosen 100% conserved MTB-PGQs because of their presence in the genome of both drug-susceptible and drug-resistant strains of bacteria. The clustering analysis followed by the conservation mapping revealed three MTB-PGQs as 100% conserved in (Table S9 ) and present in three essential genes of the bacteria, namely, espK (MTB-PGQ1), espB (MTB-PGQ2), and cyp51 (MTB-PGQ3). Figure S3 displays the consensus sequence of 100% conserved MTB-PGQs obtained from the R programming-based DECIPHER package. Figures 1A and 1B provide the schematic location, gene ID, locus tag, genome strand of G4, and the direction of the PGQ-possessing genes. A phylogenetic tree obtained from the neighbor-joining methods depicted the conservation of MTB-PGQs among the members of the mycobacteria family that include Mycobacterium bovis, Mycobacterium canettii, Mycobacterium microti, and Mycobacterium haemophilum ( Figure S4 ).