Computational Prediction of Extracellular Loops of the Por39 Outer Membrane Porin of Suitable for Epitope Surface Display

Outer membrane porins from Gram-negative bacteria are established vehicles for the production of vaccines. Typically, one or more of the extracellular loops of a porin are replaced by a peptide encoding a foreign epitope, and recombinant porin is then used as a vaccine. However, many host strains are potentially pathogenic, and also produce toxic lipopolysaccharide (LPS), both of which are undesirable for safety reasons. In contrast, the outer membrane porins from photosynthetic, purple bacteria have no known human pathology and produce only weakly toxic LPS. The purple bacterium is well-suited for large-scale biotechnology, and expresses a major porin, Por39, which is a candidate for a vaccine platform. Unfortunately, the atomic structure of Por39 could not be determined so far, and Por39 shows only a weak homology to other porins of known structure, making the assignment of external loops difficult. Here, we construct a knowledge-based model of Por39 using secondary structure constraints from both the low sequence homology to the 2POR porin from , for which the X-ray structure is known, as well as those obtained using secondary structure prediction packages. The secondary structure predictions were used to constrain a three-dimensional model created using the I-TASSER package. The modelling procedure was validated by predicting the structure of 2POR using the same strategy, but excluding the 2POR X-ray structure from the I-TASSER database. The final Por39 model allows three external loops to be defined precisely, and could also be used to obtain an initial model for the closely related Por41 using molecular modelling. These structures provide a good starting point for the insertion of epitopes with vaccine potential.