Identification of outer membrane proteins with emulsifying activity by prediction of beta-barrel regions

Microbial bioemulsifiers are secreted by many bacteria and are important for bacterial interactions with hydrophobic substrates or nutrients and for a variety of biotechnological applications. We have recently shown that the OmpA protein in several members of the Acinetobacter family has emulsifying properties. These properties of OmpA depend on the amino acid composition of four putative extra-membrane loops, which in various strains of Acinetobacter, but not in E. coli, are highly hydrophobic. As many Acinetobacter strains can utilize hydrophobic carbon sources, such as oil, the emulsifying activity of their OmpA may be important for the utilization and uptake of hydrocarbons. We assumed that if outer membrane proteins with emulsifying activity are physiologically important, they may exist in additional oil degrading bacteria. In order to identify such proteins, it was necessary to obtain bioinformatics-based predictions for hydrophobic extra-membrane loops. Here we describe a method for using protein sequence data for predicting the hydrophobic properties of the extra-membrane loops of outer membrane proteins. The feasibility of this method is demonstrated by its use to identify a new microbial bioemulsifier - OprG - an outer membrane protein of the oil degrading Pseudomonas putida KT2440.