Apport de la génomique dans la tuberculose multirésistante : de l’étude des bases moléculaires de la dissémination à la découverte de nouveaux mécanismes de résistance

Background: Tuberculosis (TB) remains one of the leading causes of death worldwide. The spread of multidrug-resistant (MDR) strains is alarming. Second-line drugs classified in WHO group A (fluoroquinolones (FQ), bedaquiline (BDQ) and linezolid) are the key drugs. Unfortunately, recent years have seen the emergence of extensively drug-resistant (XDR) strains that are MDR and resistant to FQ and one of the other group A drugs. Controlling the emergence and spread of these resistant forms relies on: (i) early and effective treatment of patients; it is thus essential to diagnose resistance to anti-tuberculosis drugs (anti-TB) as well as possible, by optimizing molecular diagnosis and (ii) understanding the mechanisms of dissemination of MDR strains. This work responds to these challenges by a genomic approach, with the following objectives: to improve the diagnosis of FQ resistance (axis 1), and to investigate the molecular mechanisms of dissemination of a particular cluster of MDR strains (axis 2). Beyond the experimental work necessary to achieve these objectives, works and publications related to the context of the subject and allowing me to understand my thesis project have been realized. These publications cited in the bibliographic part of this thesis manuscript (articles 1 to 6), will not be detailed here. Axis 1: we investigated the molecular mechanisms of resistance to FQ not related to mutations in the DNA gyrase. We analyzed the genomes of nearly 1500 FQ-resistant M. tuberculosis (Mtb) strains without DNA gyrase mutations, both clinical strains (database of the CNR des Mycobactéries and CRyPTIC consortium) and strains selected in vivo (TB mouse model). The genomic analysis revealed: (i) the putative role of the mfpAE operon, and (ii) a novel mechanism in Mtb, namely the presence of massive genomic duplications containing this operon. These in silico hypotheses were then explored by an experimental approach. Axis 2: Among the strains collected at the CNR over the last 20 years, an MDR cluster of lineage 4 was distinguished by a specific combination of resistance mutations: KatG-A110V and EthA-Q269*; strains with few gap mutations (< 8), although isolated over a period of 20 years, testifying to a low evolutionary speed. We explored the molecular characteristics of this cluster, particularly in terms of virulence to explain this persistence. Genomic analysis revealed a specific mutation in ESX-1, a key element of virulence in Mtb, upstream of espK. To explore the hypothesis of the localization of this mutation within the espK promoter, we performed (i) an in-silico promoter prediction study and (ii) qPCR experiments to assess the impact of this mutation on the transcription of espK, but also of esxA, which encodes the virulence factor ESAT-6. We showed that there would be an inverse correlation between the level of transcription of espK and esxA. Conclusion: this work has allowed (i) to improve the knowledge on FQ resistance, and (ii) to better understand the molecular basis of the persistence of an MDR cluster linked to ESX-1. In the continuity of this work, I will continue the study of the molecular mechanisms involved in the resistance to the new antituberculosis drugs (bedaquiline, delamanid and pretomanid).