A number of bacterial proteins seem to be involved in translocating specific proteins across the bacterial cell membrane by a distinct secretory mechanism that does not require the cleavage of a signal peptide. They are evolutionary related and belong to a family [1,2,3] that currently consists of:

• flhA from various bacterial species, involved in the export of flagellar proteins.
• hrpI from Erwinia amylovora, necessary for secreting harpin, a proteinaceous elecitor of the hypersensitive response (HR) in the host of this plant-pathogen.
• hrpC2 from Xanthomonas campestris pv. vesicatoria, also involved in HR.
• hrpI from Pseudomonas syringae, also involved in HR and probably in the secretion of harpin-Pss.
• hrpO from Burkholderia solanacearum, also involved in HR.
• invA from Salmonella typhimurium, which could be involved in translocating the invE protein necessary for invading cells of the intestinal epithelium.
• lcrD from Yersinia enterocolitica and pestis, which could be involved in secreting the Yop virulence proteins.
• mxiA from Shigella flexneri, involved in secreting the Ipa invasins which are necessary for penetration of intestinal epithelial cells.

These proteins have all about 700 amino-acid residues. The N-terminal half is highly conserved and seems to contain 6 to 8 transmembrane domains. The C-terminal half is less conserved and seemingly devoid of transmembrane domains. It is possible that these proteins serve as pores for the export of specific proteins. We call this family FHIPEP, which stands for Flagella/Hr/ Invasion Proteins Export Pore.

As a signature pattern, we selected the best conserved region: a highly conserved hydrophilic domain between two transmembranes regions.