Genome analysis of 17 extensively drug-resistant strains reveals new potential mutations for resistance.

We report the whole-genome sequence of an extensively drug-resistant (XDR) tuberculosis (TB) strain of Latin American-Mediterranean (LAM) lineage. This strain is phenotypically resistant to aminoglycosides, but carries no related mutations in rrs, tlyA, and eis. Through genome analysis comparison with 16 XDR strains, we found 218 non-synonymous single nucleotide polymorphisms (SNPs) shared that could confer resistance.

GENOME ANNOUNCEMENT

For Peru, the incidence rate for tuberculosis (TB) in 2012 was 95/100,000 population, from which 0.3% were new cases of extensively drug-resistant (XDR) TB. 94% of these cases were from Lima (1). XDR-TB strains are resistant to isoniazid (INH), rifampicin (RIF), fluoroquinolones (FLQ), and at least one of the three injectable anti-TB drugs: amikacin (AMK), capreomycin (CAP), or kanamycin (KAN) (2). In 1999, Peru reported one case with this resistant pattern. However, the concept of XDR-TB appears in 2006 with the first outbreak in KwaZulu-Natal (KZN), South Africa (3).

In our quest to identify new possible mutations that can confer resistance in XDR-TB strains, we analyzed the XDR strain, INS-XDR, isolated from a Peruvian patient with active tuberculosis. We determined the strain’s lineage to be Latin American–Mediterranean (LAM) using 24 mycobacterial interspersed repetitive unit-variable number of tandem repeat (MIRU-VNTR) loci (4) and SpolPred (5). The genomic DNA was sequenced on an Illumina HiSeq 2000 with 1,140× coverage. Sequence reads for the isolates were mapped against the corrected reference genome M. tuberculosis strain H37Rv The average coverage was 82 for INS-XDR (49,793,402 mapped reads) and assembly with BWA v0.5.9-r16 (6), using H37Rv genome (AL123456.3) (7) as a reference. For annotation process, we used Rapid Annotations using Subsystem Technology (8) and the Prokaryotic Genome Annotation Pipeline. The sequencing data generated has 49,793,402 paired-ends reads. The assembling process resulted in 20 contigs that represents 99.98% of the H37Rv genome. The INS-XDR has a total of 4,391,020 pb with an average G+C content of 65.4%. We used Sanger sequencing to confirm specific mutations. Phenotypically, this strain was resistant to RIF, INH, FLQ, KAN, and CAP. However, sequence analysis of rrs and the promoter eis do not show mutations for aminoglycosides resistance (9).

To analyze possible mutations causing resistance to CAP and KAN, a polymorphism study was carried out by comparing INS-XDR against KZN605 (a XDR-TB strain of LAM linage) and 16 reported TB-XDR genomes from China, FJ05194, GuangZ0019 (10), XDR1219, XDR1221 (11), GuangZ0008, 293, FJ07070, GuangZ0017, GuangZ0026 (12); Peru, Peru_04_R0292, Peru_03_R0268 (13); Russia, CTRI-4 (14), Russia_03-R1082 (13); Malaysia, UM1072388579 (15), and South Africa, KZN_R506 (16) and KZN_605. Single nucleotide polymorphisms (SNPs) were identified using SNPsFinder (17) and SNP tree (18), and were classified in the Clusters of Orthologous Groups (COG) database (19). These analyses showed evidence of 218 non-synonymous SNPs mainly in four COGs: (1) replication, recombination and repair, L, n=14, (2) energy production and conversion, C, n=14, (3) amino acid transport and metabolism, E, n=12, and (4) secondary metabolites biosynthesis, transport, and catabolism, Q, n=11. These 218 non-synonymous SNPs belong to 125 genes. Given that CAP and KAN act by inhibiting protein synthesis and causing production of abnormal proteins, we hypothesize that the 14 non-synonymous SNPs belonging to the C category would be the principal SNPs responsible for CAP and KAN resistance. This provides evidence that it may be worthwhile to explore these new SNPs for diagnosis of XDR-TB.

Nucleotide sequence accession numbers.

This whole-genome shotgun project has been deposited at DDBJ/EMBL/GenBank under the accession JANH00000000. The version described in this paper is version JANH01000000.