Closely and Grossly: ESX-1 Mediated Membrane Disruption

Abstract

Mycobacterium tuberculosis, the causative agent of Tuberculosis (TB), is an obligate human pathogen. Its ability to survive and spread is dependent upon its ability to survive within human hosts. The RD1 locus, which contains the Type VII ESX-1 secretion system, mediates this subversion at multiple stages in both Mycobacterium tuberculosis and Mycobacterium marinum. Mutants in this locus are defective for an array of virulence phenotypes, including intramacrophage growth, phagosomal permeabilization, and granuloma formation. These phenotypes have been proposed to be all dependent upon the membrane permeabilizing activity of the ESX-1 system, with the secreted substrate ESAT-6 acting as a pore forming toxin. Using Mycobacterium marinum as a model for Mtb, we found that we could measure ESX-1 mediated membrane disruption in vitro using a hemolysis assay. We demonstrate that membrane permeabilization requires ESX-1 secretion, but that this membrane permeabilization is absolutely contact-dependent. Analysis of lysed membranes via TEM shows that the manner in which Mm causes gross disruptions in host membranes, as opposed to the structured pores that are usually observed with pore forming toxins. We find that the pore forming activity previously ascribed to ESAT-6 is solely due to the presence of the hemolytic detergent ASB-14, to the extent that ESAT-6 preparations that are totally digested retain hemolytic activity. While we do confirm that ESAT-6 is capable of lysing liposomes in vitro at acidic pH, we do not find this correlates with phagosomal permeabilization. Instead we find that this activity is pH independent, suggested that ESX-1 mediated membrane disruption is operating through some still unknown mechanism. Concurrently, to improve our understanding of membrane disruption, we screened for small molecule inhibitors using hemolysis as an in vitro measurement of ESX-1 activity, and identified the seleno-organic compound Ebselen. Ebselen inhibits mycobacterial membrane disruption by inhibiting secretion of the ESX-1 substrates ESAT-6 and CFP-10. Comparisons of ebselen with an inactive analogue suggest that this inhibition is mediated via its cysteine-modifying activity. We find that Ebselen is capable of inhibiting the ESX-1 mediated virulence phenotypes of intramacrophage growth and phagosomal permeabilization. Mass spectrometry analysis finds that the inhibition of secretion is likely restricted to the ESX-1 locus., indicating that ebselen has a possible future as an antivirulence compound.