Why is a microbial agent that prevents phagosome-lysosome fusion more virulent as an intracellular pathogen than one that is not?