Genomic diversity of drug-resistant mycobacterium tuberculosis isolates in Lisbon Portugal: towards tuberculosis genomic epidemiology

Multidrug- (MDR) and extensively drug resistant (XDR) tuberculosis (TB) present a challenge to disease control and elimination goals. Lisbon, Portugal, has a high TB incidence rate and, unusual and successful XDR-TB strains that are found in circulation for almost two decades. In the present study, 56 Mycobacterium tuberculosis isolates, mostly recovered in Lisbon, were genotyped by 24-loci Mycobacterial Interspersed Repetive Unit – Variable Number of Tandem Repeats (MIRU-VNTR) and the genomes sequenced using a next generation sequencing platform – Illumina HiSeq 2000. The genotyping data revealed three major clusters associated with MDR-TB (Lisboa3-A, Lisboa3-B and Q1), two of which associated with XDR-TB (Lisboa3-B and Q1). Whilst the genomic data contributed to elucidate the phylogenetic positioning of circulating MDR-TB strains, showing a high predominance of a single SNP cluster group 5. Furthermore, a genome-wide phylogeny analysis from these strains, together with 19 publicly available genomes of Mycobacterium tuberculosis clinical isolates, revealed two major clades responsible for M/XDR-TB in the region: Lisboa3 and Q1. On the overall, 9419 different SNPs were identified, ranging between 488 – 1465 per isolate (mean: 928 SNPs/isolate). The data presented by this study contributes to the expanding knowledge of Mycobacterium tuberculosis genomic diversity yielding insights on microevolution and identification of novel compensatory mutations associated with rifampicin resistance in rpoB and rpoC. The screening for other structural variations revealed putative clade-defining variants. One deletion in PPE41, found among Lisboa3 isolates, is proposed to contribute to immune evasion and as a selective advantage. Insertion sequence (IS) mapping has also demonstrated the role of IS6110 as a major driver in mycobacterial evolution by affecting gene integrity and regulation. A total of 251 candidate insertion sites were detected, of which 105 were intergenic and 64 were predicted to have a putative upregulatory effect. Additionally, the analysis of non-synonymous/synonymous ratios revealed heterogeneities across the chromosome, genotype and Clusters of Orthologous Groups, highlighting possible and different evolution strategies. Globally, our data supports the notion of a growing genomic diversity facing both setting and host adaptation