Legionella pneumophila promotes intracellular growth by moving bacterial proteins across membranes via the Icm/Dot system. A strategy was devised to identify large numbers of Icm/Dot translocated proteins, and the resulting pool was used to identify common motifs that operate as recognition signals. The 3' end of the sidC gene, which encodes a known translocated substrate, was replaced with DNA encoding 200 codons from the 3' end of 442 potential substrate-encoding genes. The resulting hybrid proteins were then tested in a high throughput assay, in which translocated SidC antigen was detected by indirect immunofluorescence. Among translocated substrates, regions of 6-8 residues called E Blocks were identified that were rich in glutamates. Analysis of SidM/DrrA revealed that loss of three Glu residues, arrayed in a triangle on an α-helical surface, totally eliminated translocation of a reporter protein. Based on this result, a second strategy was employed to identify Icm/Dot substrates having carboxyl terminal glutamates. From the fusion assay and the bioinformatic queries, carboxyl terminal sequences from 49 previously unidentified proteins were shown to promote translocation into target cells. These studies indicate that by analysing subsets of translocated substrates, patterns can be found that allow predictions of important motifs recognized by Icm/Dot.
