Type II Secretion Substrates of Legionella pneumophila Translocate Out of the Pathogen-Occupied Vacuole via a Semipermeable Membrane

Legionella pneumophila replicates in macrophages in a host-derived phagosome, termed the Legionella-containing vacuole (LCV). While the translocation of type IV secretion (T4S) effectors into the macrophage cytosol is well established, the location of type II secretion (T2S) substrates in the infected host cell is unknown. Here, we show that the T2S substrate ProA, a metalloprotease, translocates into the cytosol of human macrophages, where it associates with the LCV membrane (LCVM). Translocation is detected as early as 10 h postinoculation (p.i.), which is approximately the midpoint of the intracellular life cycle. However, it is detected as early as 6 h p.i. if ProA is hyperexpressed, indicating that translocation depends on the timing of ProA expression and that any other factors necessary for translocation are in place by that time point. Translocation occurs with all L. pneumophila strains tested and in amoebae, natural hosts for L. pneumophila It was absent in murine bone marrow-derived macrophages and murine macrophage cell lines. The ChiA chitinase also associated with the cytoplasmic face of the LCVM at 6 h p.i. and in a T2S-dependent manner. Galectin-3 and galectin-8, eukaryotic proteins whose localization is influenced by damage to host membranes, appeared within the LCV of infected human but not murine macrophages beginning at 6 h p.i. Thus, we hypothesize that ProA and ChiA are first secreted into the vacuolar lumen by the activity of the T2S and subsequently traffic into the macrophage cytosol via a novel mechanism that involves a semipermeable LCVM.IMPORTANCE Infection of macrophages and amoebae plays a central role in the pathogenesis of L. pneumophila, the agent of Legionnaires' disease. We have previously demonstrated that the T2S system of L. pneumophila greatly contributes to intracellular infection. However, the location of T2S substrates within the infected host cell is unknown. This report presents the first evidence of a L. pneumophila T2S substrate in the host cell cytosol and, therefore, the first evidence of a non-T4S effector trafficking out of the LCV. We also provide the first indication that the LCV is not completely intact but is instead semipermeable and that this occurs in human but not murine macrophages. Given this permeability, we hypothesize that other T2S substrates and LCV lumenal contents can escape into the host cell cytosol. Thus, these substrates may represent a battery of previously unidentified effectors that can interact with host factors and contribute to intracellular infection by L. pneumophila.