Temporal genome-wide fitness analysis of Mycobacterium marinum during infection reveals genetic requirement for virulence and survival in amoebae and microglial cells

Tuberculosis remains the most pervasive infectious disease and the recent emergence of multiple or even fully drug-resistant strains increases the risk and emphasizes the need for more efficient and better drug treatments. A key feature of mycobacteria pathogenesis, conserved between the human pathogen Mycobacterium tuberculosis and the model pathogen Mycobacterium marinum, is the metabolic switch to lipid catabolism during infection and altered expression of virulence genes in coordination with different stages of infection. This study aims at identifying genes that are involved in the establishment and maintenance of the infection. To achieve this, we have applied Transposon Sequencing (Tn-Seq) to M. marinum, an unbiased genome-wide strategy that combines saturation insertional mutagenesis and high throughput sequencing. This approach allowed us to precisely identify the localization and relative abundance of insertions in pools of transposon mutants. The essentiality of genes and the positive and negative fitness cost of mutations were quantitatively compared between in vitro growth and different stages of infection in two evolutionary distinct host phagocytes, the amoeba Dictyostelium discoideum and the murine BV2 microglial cells. We found that 57% of TA sites in the M. marinum genome were disrupted and that 568 genes (10.2%) were essential for M. marinum, which is comparable to previous Tn-Seq studies on M. tuberculosis and M. bovis. Major pathways involved in the survival of M. marinum during infection of D. discoideum were related to DNA damage repair, vitamin metabolism, the type VII secretion system (T7SS) ESX-1 and the Mce1 lipid transport system. To probe for genetic requirements at different stages of infection, we submitted a transposon insertion library of Mycobacterium marinum to selection by infection of Dictyostelium discoideum or BV2 microglial cells in 3 to 8 replicates, quantified the insertion frequency in each sample (TnSeq), and subsequently compared them to the frequency in the initial library.