Uncovering the principles of membrane effector translocation in Legionella pneumophila

Abstract

The type III and type IV secretion systems are a major factor of pathogenicity for many bacteria. Effector proteins translocated by the secretion machinery are used to hijack and subvert host cell functions in order to colonize and live within the host cell. Next to soluble effectors, transmembrane domain effectors can successfully be secreted by the type III and type IV secretion system and carry out their function within a host cell membrane. Before being translocated, effector proteins must be targeted to the respective secretion system. Generally, protein targeting starts in the bacterial cytoplasm where protein biosynthesis takes place, and depends on various signals residing within the protein. Type III effectors harbor a N-terminal secretion signal followed by a chaperone binding domain. In contrast, the translocation signal for effectors of the T4SS resides at the C-terminal end of the protein. Additionally, TMD-effectors of both secretion systems contain hydrophobic segments which are essential for their proper localization within the host cell. The presence of these two incompatible signals within the same protein poses a possible targeting conflict as transmembrane segments can be recognized by the Signal Recognition Particle (SRP) and result in subsequent inner membrane insertion. For transmembrane domain effectors of the T3SS it was shown that inner membrane (mis-)targeting is avoided by a balanced hydrophobicity of the TMS (passive avoidance) as well as protection by chaperone binding (active avoidance). In contrast, some transmembrane domain effectors of the Dot/Icm system in L. pneumophila are predicted to possess a sufficiently hydrophobic signal for targeting to the bacterial inner membrane by SRP. The aim of this study was to investigate if transmembrane domain effectors of the type IV secretion system can, similarly to effectors of the type III secretion system, avoid SRP targeting or uses a hypothetical two-step secretion pathway through the Dot/Icm machinery with an inner membrane intermediate. Using membrane fractionation by a sucrose gradient centrifugation protocol as well as urea extraction, I could show that transmembrane domain effectors in L. pneumophila can indeed be found properly integrated in the bacterial inner membrane. The same results were obtained when the type IV chaperones IcmSW were overexpressed suggesting that there is no active avoidance of SRP targeting and no direct delivery to the secretion machinery by chaperone binding. Investigation of the membrane topology of transmembrane domain effectors showed that transmembrane domain effectors are “anchored” in the bacterial inner membrane with a Nin-Cin topology and only small loops of few amino acids located in the periplasm. Furthermore, I could show that, similarly to soluble effectors, the C-terminal translocation signal located in the cytoplasm and possible internal signals as well as the presence of the chaperones IcmSW are crucial for the successful translocation of transmembrane domain effectors into host cells. Based on these results, I propose that transmembrane domain effectors in L. pneumophila follow a two-step secretion pathway with SRP XI targeting as the first step. Once “anchored” in the bacterial inner membrane, transmembrane domain effectors are recognized as substrates of the Dot/Icm system by IcmSW, resulting in their extraction towards the cytoplasmic side before being translocated into host cells.