Complete Genome Sequencing of Mycobacterium bovis SP38 and Comparative Genomics of Mycobacterium bovis and M. tuberculosis Strains

Abstract

Mycobacterium bovis causes bovine tuberculosis and is the main organism responsible for zoonotic tuberculosis in humans. We performed the sequencing, assembly and annotation of a Brazilian strain of M. bovis named SP38, and performed comparative genomics of M. bovis genomes deposited in GenBank. M. bovis SP38 has a traditional tuberculous mycobacterium genome of 4,347,648 bp, with 65.5% GC, and 4,216 genes. The majority of CDSs (2,805, 69.3%) have predictive function, while 1,206 (30.07%) are hypothetical. For comparative analysis, 31 M. bovis, 32 M. bovis BCG, and 23 Mycobacterium tuberculosis genomes available in GenBank were selected. M. bovis RDs (regions of difference) and Clonal Complexes (CC) were identified in silico. Genome dynamics of bacterial groups were analyzed by gene orthology and polymorphic sites identification. M. bovis polymorphic sites were used to construct a phylogenetic tree. Our RD analyses resulted in the exclusion of three genomes, mistakenly annotated as virulent M. bovis. M. bovis SP38 along with strain 35 represent the first report of CC European 2 in Brazil, whereas two other M. bovis strains failed to be classified within current CC. Results of M. bovis orthologous genes analysis suggest a process of genome remodeling through genomic decay and gene duplication. Quantification, pairwise comparisons and distribution analyses of polymorphic sites demonstrate greater genetic variability of M. tuberculosis when compared to M. bovis and M. bovis BCG (p ≤ 0.05), indicating that currently defined M. tuberculosis lineages are more genetically diverse than M. bovis CC and animal-adapted MTC (M. tuberculosis Complex) species. As expected, polymorphic sites annotation shows that M. bovis BCG are subjected to different evolutionary pressures when compared to virulent mycobacteria. Lastly, M. bovis phylogeny indicates that polymorphic sites may be used as markers of M. bovis lineages in association with CC. Our findings highlight the need to better understand host-pathogen co-evolution in genetically homogeneous and/or diverse host populations, considering the fact that M. bovis has a broader host range when compared to M. tuberculosis. Also, the identification of M. bovis genomes not classified within CC indicates that the diversity of M. bovis lineages may be larger than previously thought or that current classification should be reviewed.

Global distribution and evolution ofMycobacterium bovislineages

“…There were only three M. bovis genomes without CC classification isolated from humans in Germany that appeared closely related to genomes of the CC Eu2 (Figure 2 - indicated as **, part of isolates also emerging from node C). The CC Eu2 has been described as dominant in the Iberian Peninsula and also detected in other countries of Europe and in Brazil (Rodriguez-Campos et al, 2012; Zimpel et al, 2017a). Its actual worldwide occurrence, however, is unknown.…”

“…Nevertheless, attained results suggest that M. bovis strains are likely to cluster based on geographical location (Dippenaar et al, 2017; Kraemer Zimpel et al, 2017; Lasserre et al, 2018). In our previous study, we have also shown that few M. bovis genomes do not carry any CC genetic marker (Zimpel et al, 2017a), a phenomenon that was recently observed in M. bovis isolates from one cattle herd in the USA and from slaughterhouse cattle in Eritrea (Ghebremariam et al, 2018; Orloski et al, 2018b). These findings suggest that CCs are unlikely to represent the whole diversity of M. bovis strains, warranting further evaluation of M. bovis molecular lineages (Zimpel et al, 2017a; Lasserre et al, 2018).…”

“…Two Brazilian M. bovis isolates obtained from a captive European bison ( Bison bonasus ) (Zimpel et al, 2017b), and a captive llama ( Llama glama ) (provided by the Laboratory of Bacterial Zoonosis of the College of Veterinary Medicine, University of São Paulo, Brazil) were reactivated in Stonebrink medium and a single colony was sub-cultured for DNA extraction using a previously described protocol (Zimpel et al, 2017a). DNA quality was evaluated using Nanodrop 2000c (Thermo Scientific, MA, USA) and Agilent 2100 High Sensitivity Chip DNA Bioanalyzer (Agilent Technologies, CA, USA).…”

“…However, no studies to date have comprehensively analyzed M. bovis genomes at a global scale. Few studies that have compared transboundary M. bovis strains analyzed bacterial isolates obtained from a reduced number of countries (n<9) and included small sample sizes (Dippenaar et al, 2017; Patané et al, 2017; Zimpel et al, 2017a; Ghebremariam et al, 2018; Lasserre et al, 2018). Nevertheless, attained results suggest that M. bovis strains are likely to cluster based on geographical location (Dippenaar et al, 2017; Kraemer Zimpel et al, 2017; Lasserre et al, 2018).…”

“…Few studies that have compared transboundary M. bovis strains analyzed bacterial isolates obtained from a reduced number of countries (n<9) and included small sample sizes (Dippenaar et al, 2017; Patané et al, 2017; Zimpel et al, 2017a; Ghebremariam et al, 2018; Lasserre et al, 2018). Nevertheless, attained results suggest that M. bovis strains are likely to cluster based on geographical location (Dippenaar et al, 2017; Kraemer Zimpel et al, 2017; Lasserre et al, 2018). In our previous study, we have also shown that few M. bovis genomes do not carry any CC genetic marker (Zimpel et al, 2017a), a phenomenon that was recently observed in M. bovis isolates from one cattle herd in the USA and from slaughterhouse cattle in Eritrea (Ghebremariam et al, 2018; Orloski et al, 2018b).…”