Mycobacterium tuberculosis Toxin CpnT Is an ESX-5 Substrate and Requires Three Type VII Secretion Systems for Intracellular Secretion.

CpnT, a NAD+ glycohydrolase, is the only known toxin that is secreted by Mycobacterium tuberculosis CpnT is composed of two domains; the C-terminal domain is the toxin, whereas the N-terminal domain is required for secretion. CpnT shows characteristics of type VII secretion (T7S) substrates, including a predicted helix-turn-helix domain followed by a secretion motif (YxxxE). Disruption of this motif indeed abolished CpnT secretion. By analyzing different mutants, we established that CpnT is specifically secreted by the ESX-5 system in Mycobacterium marinum under axenic conditions and during macrophage infection. Surprisingly, intracellular secretion of CpnT was also dependent on both ESX-1 and ESX-4. These secretion defects could be partially rescued by coinfection with wild-type bacteria, indicating that secreted effectors are involved in this process. In summary, our data reveal that three different type VII secretion systems have to be functional in order to observe intracellular secretion of the toxin CpnT.IMPORTANCE For decades, it was believed that the intracellular pathogen M. tuberculosis does not possess toxins. Only fairly recently it was discovered that CpnT is a potent secreted toxin of M. tuberculosis, causing necrotic cell death in host cells. However, until now the secretion pathway remained unknown. In our study, we were able to identify CpnT as a substrate of the mycobacterial type VII secretion system. Pathogenic mycobacteria have up to five different type VII secretion systems, called ESX-1 to ESX-5, which play distinct roles for the pathogen during growth or infection. We were able to elucidate that CpnT is exclusively secreted by the ESX-5 system in bacterial culture. However, to our surprise we discovered that, during infection studies, CpnT secretion relies on intact ESX-1, ESX-4, and ESX-5 systems. We elucidate for the first time the intertwined interplay of three different and independent secretion systems to secrete one substrate during infection.

The mannose cap of mycobacterial lipoarabinomannan does not dominate the Mycobacterium-host interaction.

Pathogenic mycobacteria have the ability to persist in phagocytic cells and to suppress the immune system. The glycolipid lipoarabinomannan (LAM), in particular its mannose cap, has been shown to inhibit phagolysosome fusion and to induce immunosuppressive IL-10 production via interaction with the mannose receptor or DC-SIGN. Hence, the current paradigm is that the mannose cap of LAM is a crucial factor in mycobacterial virulence. However, the above studies were performed with purified LAM, never with live bacteria. Here we evaluate the biological properties of capless mutants of Mycobacterium marinum and M. bovis BCG, made by inactivating homologues of Rv1635c. We show that its gene product is an undecaprenyl phosphomannose-dependent mannosyltransferase. Compared with parent strain, capless M. marinum induced slightly less uptake by and slightly more phagolysosome fusion in infected macrophages but this did not lead to decreased survival of the bacteria in vitro, nor in vivo in zebra fish. Loss of caps in M. bovis BCG resulted in a sometimes decreased binding to human dendritic cells or DC-SIGN-transfected Raji cells, but no differences in IL-10 induction were observed. In mice, capless M. bovis BCG did not survive less well in lung, spleen or liver and induced a similar cytokine profile. Our data contradict the current paradigm and demonstrate that mannose-capped LAM does not dominate the Mycobacterium-host interaction.