Evolutionary Patterns in Coiled-Coils

• Autor(a) principal: Surkont, J.
• Data de Publicação: 2015
• Outros Autores: Pereira-Leal, J. B.
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.7/419
• topic: coiled-coil; protein evolution; phylogenetic inference; homology detection; amino acid substitutions; protein structure

Models of protein evolution are used to describe evolutionary processes, for phylogenetic analyses and homology detection. Widely used general models of protein evolution are biased toward globular domains and lack resolution to describe evolutionary processes for other protein types. As three-dimensional structure is a major constraint to protein evolution, specific models have been proposed for other types of proteins. Here, we consider evolutionary patterns in coiled-coil forming proteins. Coiled-coils are widespread structural domains, formed by a repeated motif of seven amino acids (heptad repeat). Coiled-coil forming proteins are frequently rods and spacers, structuring both the intracellular and the extracellular spaces that often form protein interaction interfaces. We tested the hypothesis that due to their specific structure the associated evolutionary constraints differ from those of globular proteins. We showed that substitution patterns in coiled-coil regions are different than those observed in globular regions, beyond the simple heptad repeat. Based on these substitution patterns we developed a coiled-coil specific (CC) model that in the context of phylogenetic reconstruction outperforms general models in tree likelihood, often leading to different topologies. For multidomain proteins containing both a coiled-coil region and a globular domain, we showed that a combination of the CC model and a general one gives higher likelihoods than a single model. Finally, we showed that the model can be used for homology detection to increase search sensitivity for coiled-coil proteins. The CC model, software, and other supplementary materials are available at http://www.evocell.org/cgl/resources (last accessed January 29, 2015).