Real-time monitoring of the membrane-binding and insertion properties of the cholesterol-dependent cytolysin anthrolysin O from Bacillus anthracis
Cocklin, Simon × Jost, Monika Robertson, Noreen M Weeks, Stephen Weber, Hans-Walter Young, Emily Seal, Samar Zhang, Can Mosser, Elise Loll, Patrick J Saunders, Aleister J Rest, Richard F Chaiken, Irwin M #
Journal of molecular recognition : JMR vol:19 issue:4 pages:354-362
Bacillus anthracis has recently been shown to secrete a potently hemolytic/cytolytic protein that has been designated anthrolysin O (ALO). In this work, we initiated a study of this potential anthrax virulence factor in an effort to understand the membrane-binding properties of this protein. Recombinant anthrolysin O (rALO35-512) and two N-terminally truncated versions of ALO (rALO390-512 and rALO403-512) from B. anthracis were overproduced in Escherichia coli and purified to homogeneity. The role of cholesterol in the cytolytic activity of ALO was probed in cellular cholesterol depletion assays using mouse and human macrophage-like lines, and also Drosophila Schneider 2 cells. Challenging the macrophage cells with rALO35-512, but not rALO390-512 or rALO403-512, resulted in cell death by lysis, with this cytolysis being abolished by depletion of the membrane cholesterol. Drosophila cells, which contain ergosterol as their major membrane sterol, were resistant to rALO-mediated cytolysis. In order to determine the molecular mechanism of this resistance, the interaction of rALO with model membranes comprised of POPC alone, or with a variety of structurally similar sterols including ergosterol, was probed using Biacore. Both rALO35-512 and rALO403-512 demonstrated robust binding to model membranes composed of POPC and cholesterol, with amount of protein bound proportional to the cholesterol content. Ergosterol supported greatly reduced binding of both rALO35-512 and rALO403-512, whereas other sterols tested did not support binding. The rALO403-512--membrane interaction demonstrated an equilibrium dissociation constant (KD) in the low nanomolar range, whereas rALO35-512 exhibited complex kinetics likely due to the multiple events involved in pore formation. These results establish the pivotal role of cholesterol in the action of rALO. The biosensor method developed to measure ALO recognition of cholesterol in a membrane environment could be extended to provide a platform for the screening of inhibitors of other membrane-binding proteins and peptides.