Network analysis identifies regulators of lineage-specific phenotypes in Mycobacterium tuberculosis

The Mycobacterium tuberculosis (Mtb) complex comprises seven phylogenetically distinct human-adapted lineages exhibiting different geographical distribution and degrees of pathogenicity. Among these, Lineage 1 (L1) has been associated with low virulence whereas Lineage 2 (L2) has been linked to hyper-virulence, enhanced transmission and drug resistance. Here, we conducted multi-layer comparative analyses using whole genome sequencing data combined with quantitative transcriptomic and proteomic profiling of a set of L1 and L2 clinical strains, each grown under two different conditions in vitro. Our data revealed different degrees of correlation between transcript and protein abundances across clinical strains and functional gene categories, indicating variable levels of post-transcriptional regulation in the tested lineages. Contrasting genomic and gene expression data showed that the magnitude of the transcriptional and translational changes was proportional to the phylogenetic distance between strains, with one out of three single nucleotide polymorphisms leading to a transcriptional and/or translational change on average. We devised a new genome-scale transcriptional regulatory model and identified several master transcription factors, strongly linked to the sigma factor network, whose targets were differentially regulated between the two lineages. These differences resulted in a higher basal expression of DosR proteins and a stronger response to nitric oxide (NO) exposure in L2 compared to L1. These patterns are most likely responsible for the shorter NO-induced growth arrest in L2 observed. Given the limited genetic variation between strains, it appears that phenotypic differences in Mtb are substantially driven by differences in the regulation of biochemical networks through master transcriptional regulators.