Outcome of the First wwPDB Hybrid/Integrative Methods Task Force Workshop

Abstract

Structures of biomolecular systems are increasingly computed by integrative modeling that relies on varied types of experimental data and theoretical information. We describe here the proceedings and conclusions from the first wwPDB Hybrid/Integrative Methods Task Force Workshop held at the European Bioinformatics Institute in Hinxton, UK, on October 6 and 7, 2014. At the workshop, experts in various experimental fields of structural biology, experts in integrative modeling and visualization, and experts in data archiving addressed a series of questions central to the future of structural biology. How should integrative models be represented? How should the data and integrative models be validated? What data should be archived? How should the data and models be archived? What information should accompany the publication of integrative models?

Figure 1. Examples of recently determined integrative structures

The molecular architecture of INO80 was determined with a 17 Å resolution cryo-EM map and 212 intra-protein and 116 inter-protein crosslinks (Russel et al., 2009). The molecular architecture of Polycomb Repressive Complex 2 (PRC2) was determined with a 21 Å resolution negative-stain EM map and ~60 intra-protein and inter-protein crosslinks (Shi et al., 2014). The molecular architecture of the large subunit of the mammalian mitochondrial ribosome (39S) was determined with a 4.9 Å resolution cryo-EM map and ~70 inter-protein cross-links (Ward et al., 2013). The molecular architecture of the RNA polymerase II transcription pre-initiation complex was determined with a 16 Å resolution cryo-EM map plus 157 intra-protein and 109 inter-protein crosslinks (Alber et al., 2008). The atomic model of Type III secretion system needle was determined with a 19.5 Å resolution cryo-EM map and solid-state NMR data (Loquet et al., 2012). Molecular architecture of the productive HIV-1 reverse transcriptase:DNA primer-template complex in the open educt state was determined by FRET positioning and screening (FPS) using a known HIV-1 reverse transcriptase structure (Kalinin et al., 2012). The structure of HIV-1 capsid protein was determined using residual dipolar couplings (RDC) and SAXS data (Deshmukh et al., 2013). The human genome architecture was determined based on tethered chromosome conformation capture (TCC) and population-based modeling (Kalhor et al., 2012). The structural model of α-globin gene domain was determined based on Chromosome Conformation Capture Carbon Copy (5C) experiments (Bau et al., 2011). The molecular architecture of the proteosomal lid was determined using native MS and 28 cross-links (Politis et al., 2014). Atomic resolution conformations of ESCRT-I complex were determined with SAXS, double electron-electron transfer (DEER), and FRET (Boura et al., 2011). Integrative model of actin and the cardiac myosin binding protein C was developed from a combination of crystallographic and NMR structures of subunits and domains, with positions and orientations optimized against SAXS and SANS data to reveal information about the quaternary interactions (Whitten et al., 2008). The ensemble of [ψ^CD]₂ NMR structures were fitted into the averaged cryo-electron tomography map (Miyazaki et al., 2010). Integrative model of the cyanobacterial circadian timing KaiB-KaiC complex was obtained based on H/D exchange and collision cross section data from MS (Snijder et al., 2014). The prepore and pore conformations of the pore-forming toxin aerolysin were obtained combining cryo-EM data and molecular dynamics simulations (Degiacomi and Dal Peraro, 2013; Degiacomi et al., 2013). Segment of a pleurotolysin pore map (~11 Å resolution) with an ensemble of conformations showing the trajectory of β-sheet opening during pore formation (Lukoyanova et al., 2015). A SAXS-based rigid body model of a ternary complex of the iron-sulphur cluster assembly proteins desulfurase (orange) and scaffold protein Isu (blue) with bacterial orthologue of frataxin (yellow) was validated by NMR chemical shifts and mutagenesis (Prischi et al., 2010). The molecular architecture of the SAGA transcription coactivator complex was determined with 199 inter- and 240 intra-subunit crosslinks, several comparative models based on X-ray crystal structures, and a TFIID core EM map at 31 Å resolution (Han et al., 2014). Structural Organization of the bacterial (T. aquaticus) RNA polymerase-promoter open complex obtained by FRET (Mekler et al., 2002), subsequently validated by a crystal structure (Zhang et al., 2012). The RNA ribosome-binding element from turnip crinkle virus genome, determined using NMR, SAXS, and EM data (Gong et al., 2015). The molecular architecture of the complex between RNA polymerase II and transcription factor IIF was determined using a deposited crystal structure of RNA polymerase II, homology models of some domains in transcription factor IIF, as well as 95 intra-protein and 129 inter-protein cross links (Chen et al., 2010).

Figure 2. The four stages of integrative structure determination

The approach is illustrated by its application to the heptameric Nup84 subcomplex of the Nuclear Pore Complex (Shi et al., 2014).