CCFold: rapid and accurate prediction of coiled-coil structures and application to modelling intermediate filaments

Abstract

Accurate molecular structure of the protein dimer representing the elementary building block of intermediate filaments (IFs) is essential towards the understanding of the filament assembly, rationalizing their mechanical properties and explaining the effect of diseaserelated IF mutations. The dimer contains a ∼300-residue long α-helical coiled coil which is not assessable to either direct experimental structure determination or modelling using standard approaches. At the same time, coiled coils are well-represented in structural databases. Here we present CCFold, a generally applicable threading-based algorithm which produces coiled-coil models from protein sequence only. The algorithm is based on a statistical analysis of experimentally determined structures and can handle any hydrophobic repeat patterns in addition to the most common heptads. We demonstrate that CCFold outperforms general-purpose computational folding in terms of accuracy, while being faster by orders of magnitude. By combining the CCFold algorithm and Rosetta folding we generate representative dimer models for all IF protein classes. The source code is freely available at https://github.com/biocryst/IF

Production and analysis of a mammalian septin hetero‐octamer complex

“…Using these methods, we also modeled a 1:1:1:1 tetramer of SEPT2/SEPT6/SEPT7/SEPT3 (Figure 2d,e) and a SEPT2/SEPT6/SEPT7 hexamer (Figure S5B). We note that these models combine multiple septin structural models ("Human septin 2 in complex with GDP," n.d.; Brognara, Pereira, Brandao‐Neto, Araujo, & Garratt, 2019; Castro et al, 2020; Guzenko & Strelkov, 2018; Sirajuddin et al, 2007) by aligning structures using the hetero‐trimer structure as a template (Sirajuddin et al, 2007) and includes models for the coiled‐coil regions that are computationally derived (Guzenko & Strelkov, 2018; Ludwiczak, Winski, Szczepaniak, Alva, & Dunin‐Horkawicz, 2019). The computationally derived parallel coiled‐coil model is consistent with FRET between the SEPT6 and SEPT7 C‐termini (Low & Macara, 2006).…”

“…Thus, a SEPT2/6/7/3 tetramer G‐domain backbone was produced. To model the coiled‐coils, which are absent from the G‐domain structures and from 2QAG, we built a model for the C‐terminal SEPT6‐SEPT7 CC hetero‐dimer and SEPT2 homo‐dimer using CCfold (Guzenko & Strelkov, 2018). These were built slightly outside of the identified coiled‐coil region (Ludwiczak et al, 2019) such that we could extend the final α‐helix in the SEPT7 G‐domain to orient the coiled‐coil model.…”

Ligand-induced conformational rearrangements regulate the switch between membrane-proximal and distal functions of Rho kinase 2

“…Five such segments were identified. Structures for individual segments were constructed using the CCFold program 40 . Missing side chains were added using the Profix tool of the Jackal package 41 .…”

Optimal data-driven parameterization of coiled coils

“…As a starting point, all parallel dimeric CC fragments of 15 residues per chain were extracted from the CC+ database [10] and truncated to polyalanine. Following the procedures described in [11], a non-redundant 45 set of 14175 CC fragments was obtained. Here non-redundancy was defined in terms of stuctural dissimilarity, so that the root mean square deviation (RMSD) between any two fragments upon their optimal superposition was at least 0.2Å.…”

“…In such cases, more sophisticated approaches should be used that take such irregularities into account. The 225 available tools include a specialized fold-and-dock protocol of the popular Rosetta suite [24] and our CCFold program [11], which implement datadriven modelling algorithms designed for CC structures in particular.…”