Genomic epidemiology of Mycobacterium tuberculosis in Santa Catarina, Southern Brazil

Abstract

Mycobacterium tuberculosis (M.tb), the pathogen responsible for tuberculosis (TB) poses as the major cause of death among infectious diseases. The knowledge about the molecular diversity of M.tb enables the implementation of more effective surveillance and control measures and, nowadays, Whole Genome Sequencing (WGS) holds the potential to produce high-resolution epidemiological data in a high-throughput manner. Florianópolis, the state capital of Santa Catarina (SC) in south Brazil, shows a high TB incidence (46.0/100,000). Here we carried out a WGS-based evaluation of the M.tb strain diversity, drug-resistance and ongoing transmission in the capital metropolitan region. Resistance to isoniazid, rifampicin, streptomycin was identified respectively in 4.0% (n = 6), 2.0% (n = 3) and 1.3% (n = 2) of the 151 studied strains by WGS. Besides, resistance to pyrazinamide and ethambutol was detected in 0.7% (n = 1) and reistance to ethionamide and fluoroquinolone (FQ) in 1.3% (n = 2), while a single (0.7%) multidrug-resistant (MDR) strain was identified. SNP-based typing classified all isolates into M.tb Lineage 4, with high proportion of sublineages LAM (60.3%), T (16.4%) and Haarlem (7.9%). The average core-genome distance between isolates was 420.3 SNPs, with 43.7% of all isolates grouped across 22 genomic clusters thereby showing the presence of important ongoing TB transmission events. Most clusters were geographically distributed across the study setting which highlights the need for an urgent interruption of these large transmission chains. The data conveyed by this study shows the presence of important and uncontrolled TB transmission in the metropolitan area and provides precise data to support TB control measures in this region.

Figure 1

Maximum likelihood phylogenetic tree for the 151 M.tb isolates included in the study. This tree was constructed based on 17,027 core SNPs and is shown annotated with (from the outside to the inside): presence or absence of fbpC¹⁰³ LAM marker; SIT and spoligotyping clade; SNP-based sub-lineage; drug resistance associated mutations; and genotypic-based prediction for drug susceptibility. Tips are shown coloured according to the genomic cluster (see legend). NC not clustered, GC genomic cluster.

Figure 2

Minimum spanning tree (MST) of the 151 M.tb clinical isolates included in the present study. This MST is based on 17,027 core SNPs and nodes are shown coloured in function of the associated genomic cluster.

Figure 3

Geographic distribution of 142 out of the 151 TB cases (excluding 9 homeless) included in the study for which address of residence was available. These cases were geographically mapped according to the respective address of residence and are shown in the map colored according to the respective isolate’s genomic cluster. The spatial distribution of these cases and correlation with the genomic cluster shows an extensive geographical spread of the larger clusters (n ≥ 3) except for GC2, GC4 and GC10. The map was created using the online microreact tool, available at https://microreact.org/project/hrdK7GcztX8M2ZXEXjc579.