BACKGROUND: OmpF porin is a trimeric integral membrane protein responsible for the passive transport of small hydrophilic molecules, such as nutrients and waste products, across the outer membrane of Escherichia coli. Very few membrane proteins have been crystallized in three dimensions, yet this stable protein can be obtained in several crystal forms. Comparison of the structures of the same membrane protein in two different packing environments is of major interest, because it allows us to explore the integrity of the structure outside the natural membrane environment. RESULTS: The structure of OmpF porin in a tetragonal crystal form with two trimers per asymmetric unit has been determined at 3.2 A resolution and compared with that obtained previously in a trigonal crystal form. The lattice contacts involve only polar atoms, whereas extensive hydrophobic protein-protein interactions were found in the trigonal lattice. The trimer structure is virtually identical in both. CONCLUSIONS: Our comparison reveals that the overall structure of OmpF is not influenced by crystal lattice constraints and, thus, presumably bears close resemblance to the in vivo structure. The tetragonal crystal structure has provided the starting model for the phasing of neutron diffraction data obtained from this crystal form, as described in an accompanying article.
BACKGROUND: OmpF porin is a trimeric integral membrane protein responsible for the passive transport of small hydrophilic molecules, such as nutrients and waste products, across the outer membrane of Escherichia coli. Very few membrane proteins have been crystallized in three dimensions, yet this stable protein can be obtained in several crystal forms. Comparison of the structures of the same membrane protein in two different packing environments is of major interest, because it allows us to explore the integrity of the structure outside the natural membrane environment. RESULTS: The structure of OmpF porin in a tetragonal crystal form with two trimers per asymmetric unit has been determined at 3.2 A resolution and compared with that obtained previously in a trigonal crystal form. The lattice contacts involve only polar atoms, whereas extensive hydrophobic protein-protein interactions were found in the trigonal lattice. The trimer structure is virtually identical in both. CONCLUSIONS: Our comparison reveals that the overall structure of OmpF is not influenced by crystal lattice constraints and, thus, presumably bears close resemblance to the in vivo structure. The tetragonal crystal structure has provided the starting model for the phasing of neutron diffraction data obtained from this crystal form, as described in an accompanying article.