Search

A huge amount of genetic information is available thanks to the recent advances in sequencing technologies and the larger computational capabilities, but the interpretation of such genetic data at phenotypic level remains elusive. One of the reasons is that proteins are not acting alone, but are specifically interacting with other proteins and biomolecules, forming intricate interaction networks that are essential for the majority of cell processes and pathological conditions. Thus, characterizing such interaction networks is an important step in understanding how information flows from gene to phenotype. Indeed, structural characterization of protein-protein interactions at atomic resolution has many applications in biomedicine, from diagnosis and vaccine design, to drug discovery. However, despite the advances of experimental structural determination, the number of interactions for which there is available structural data is still very small. In this context, a complementary approach is computational modeling of protein interactions by docking, which is usually composed of two major phases: (i) sampling of the possible binding modes between the interacting molecules and (ii) scoring for the identification of the correct orientations. In addition, prediction of interface and hot-spot residues is very useful in order to guide and interpret mutagenesis experiments, as well as to understand functional and mechanistic aspects of the interaction. Computational docking is already being applied to specific biomedical problems within the context of personalized medicine, for instance, helping to interpret pathological mutations involved in protein-protein interactions, or providing modeled structural data for drug discovery targeting protein-protein interactions.

A huge amount of genetic information is available thanks to the recent advances in sequencing technologies and the larger computational capabilities, but the interpretation of such genetic data at phenotypic level remains elusive. One of the reasons is that proteins are not acting alone, but are specifically interacting with other proteins and biomolecules, forming intricate interaction networks that are essential for the majority of cell processes and pathological conditions. Thus, characterizing such interaction networks is an important step in understanding how information flows from gene to phenotype. Indeed, structural characterization of protein–protein interactions at atomic resolution has many applications in biomedicine, from diagnosis and vaccine design, to drug discovery. However, despite the advances of experimental structural determination, the number of interactions for which there is available structural data is still very small. In this context, a complementary approach is computational modeling of protein interactions by docking, which is usually composed of two major phases: (i) sampling of the possible binding modes between the interacting molecules and (ii) scoring for the identification of the correct orientations. In addition, prediction of interface and hot-spot residues is very useful in order to guide and interpret mutagenesis experiments, as well as to understand functional and mechanistic aspects of the interaction. Computational docking is already being applied to specific biomedical problems within the context of personalized medicine, for instance, helping to interpret pathological mutations involved in protein–protein interactions, or providing modeled structural data for drug discovery targeting protein–protein interactions. Spanish Ministry of Economy grant number BIO2016-79960-R; D.B.B. is supported by a predoctoral fellowship from CONACyT; M.R. is supported by an FPI fellowship from the Severo Ochoa program. We are grateful to the Joint BSC-CRG-IRB Programme in Computational Biology.

Community-wide blind prediction experiments such as CAPRI and CASP provide an objective measure of the current state of predictive methodology. Here we describe a community-wide assessment of methods to predict the effects of mutations on protein-protein interactions. Twenty-two groups predicted the effects of comprehensive saturation mutagenesis for two designed influenza hemagglutinin binders and the results were compared with experimental yeast display enrichment data obtained using deep sequencing. The most successful methods explicitly considered the effects of mutation on monomer stability in addition to binding affinity, carried out explicit side-chain sampling and backbone relaxation, evaluated packing, electrostatic, and solvation effects, and correctly identified around a third of the beneficial mutations. Much room for improvement remains for even the best techniques, and large-scale fitness landscapes should continue to provide an excellent test bed for continued evaluation of both existing and new prediction methodologies.

An important aspect of protein functionality is the formation of specific complexes with other proteins, which are involved in the majority of biological processes. The functional characterization of such interactions at molecular level is necessary, not only to understand biological and pathological phenomena but also to design improved, or even new interfaces, or to develop new therapeutic approaches. X-ray crystallography and NMR spectroscopy have increased the number of 3D protein complex structures deposited in the Protein Data Bank (PDB). However, one of the more challenging objectives in biological research is to functionally characterize protein interactions and thus identify residues that significantly contribute to the binding. Considering that the experimental characterization of protein interfaces remains expensive, time-consuming, and labor-intensive, computational approaches represent a significant breakthrough in proteomics, assisting or even replacing experimental efforts. Thanks to the technological advances in computing and data processing, these techniques now cover a vast range of protocols, from the estimation of the evolutionary conservation of amino acid positions in a protein, to the energetic contribution of each residue to the binding affinity. In this chapter, we review several existing computational protocols to model the phylogenetic, structural, and energetic properties of residues within protein-protein interfaces.

An important aspect of protein functionality is the formation of specific complexes with other proteins, which are involved in the majority of biological processes. The functional characterization of such interactions at molecular level is necessary, not only to understand biological and pathological phenomena but also to design improved, or even new interfaces, or to develop new therapeutic approaches. X-ray crystallography and NMR spectroscopy have increased the number of 3D protein complex structures deposited in the Protein Data Bank (PDB). However, one of the more challenging objectives in biological research is to functionally characterize protein interactions and thus identify residues that significantly contribute to the binding. Considering that the experimental characterization of protein interfaces remains expensive, timeconsuming, and labor-intensive, computational approaches represent a significant breakthrough in proteomics, assisting or even replacing experimental efforts. Thanks to the technological advances in computing and data processing, these techniques now cover a vast range of protocols, from the estimation of the evolutionary conservation of amino acid positions in a protein, to the energetic contribution of each residue to the binding affinity. In this chapter, we review several existing computational protocols to model the phylogenetic, structural, and energetic properties of residues within protein–protein interfaces.

We present the results for CAPRI Round 30, the first joint CASP-CAPRI experiment, which brought together experts from the protein structure prediction and protein–protein docking communities. The Round comprised 25 targets from amongst those submitted for the CASP11 prediction experiment of 2014. The targets included mostly homodimers, a few homotetramers, and two heterodimers, and comprised protein chains that could readily be modeled using templates from the Protein Data Bank. On average 24 CAPRI groups and 7 CASP groups submitted docking predictions for each target, and 12 CAPRI groups per target participated in the CAPRI scoring experiment. In total more than 9500 models were assessed against the 3D structures of the corresponding target complexes. Results show that the prediction of homodimer assemblies by homology modeling techniques and docking calculations is quite successful for targets featuring large enough subunit interfaces to represent stable associations. Targets with ambiguous or inaccurate oligomeric state assignments, often featuring crystal contact-sized interfaces, represented a confounding factor. For those, a much poorer prediction performance was achieved, while nonetheless often providing helpful clues on the correct oligomeric state of the protein. The prediction performance was very poor for genuine tetrameric targets, where the inaccuracy of the homology-built subunit models and the smaller pair-wise interfaces severely limited the ability to derive the correct assembly mode. Our analysis also shows that docking procedures tend to perform better than standard homology modeling techniques and that highly accurate models of the protein components are not always required to identify their association modes with acceptable accuracy. We are most grateful to the PDBe at the European Bioinformatics Institute in Hinxton, UK, for hosting the CAPRI website. Our deepest thanks go to all the structural biologists and to the following structural genomics initiatives: Northeast Structural Genomics Consortium, Joint Center for Structural Genomics, NatPro PSI:Biology, New York Structural Genomics Research Center, Midwest Center for Structural Genomics, Structural Genomics Consortium, for contributing the targets for this joint CASP-CAPRI experiment. MFL acknowledges support from the FRABio FR3688 Research Federation “Structural & Functional Biochemistry of Biomolecular Assemblies.”

The sixth CAPRI edition included new modeling challenges, such as the prediction of protein-peptide complexes, and the modeling of homo-oligomers and domain-domain interactions as part of the first joint CASP-CAPRI experiment. Other non-standard targets included the prediction of interfacial water positions and the modeling of the interactions between proteins and nucleic acids. We have participated in all proposed targets of this CAPRI edition both as predictors and as scorers, with new protocols to efficiently use our docking and scoring scheme pyDock in a large variety of scenarios. In addition, we have participated for the first time in the servers section, with our recently developed webserver, pyDockWeb. Excluding the CASP-CAPRI cases, we submitted acceptable models (or better) for 7 out of the 18 evaluated targets as predictors, 4 out of the 11 targets as scorers, and 6 out of the 18 targets as servers. The overall success rates were below those in past CAPRI editions. This shows the challenging nature of this last edition, with many difficult targets for which no participant submitted a single acceptable model. Interestingly, we submitted acceptable models for 83% of the evaluated protein-peptide targets. As for the 25 cases of the CASP-CAPRI experiment, in which we used a larger variety of modeling techniques (template-based, symmetry restraints, literature information, etc.), we submitted acceptable models for 56% of the targets. In summary, this CAPRI edition showed that pyDock scheme can be efficiently adapted to the increasing variety of problems that the protein interactions field is currently facing. Proteins 2017; 85:487-496. © 2016 Wiley Periodicals, Inc.

In addition to protein-protein docking, this CAPRI edition included new challenges, like protein-water and protein-sugar interactions, or the prediction of binding affinities and ΔΔG changes upon mutation. Regarding the standard protein-protein docking cases, our approach, mostly based on the pyDock scheme, submitted correct models as predictors and as scorers for 67% and 57% of the evaluated targets, respectively. In this edition, available information on known interface residues hardly made any difference for our predictions. In one of the targets, the inclusion of available experimental small-angle X-ray scattering (SAXS) data using our pyDockSAXS approach slightly improved the predictions. In addition to the standard protein-protein docking assessment, new challenges were proposed. One of the new problems was predicting the position of the interface water molecules, for which we submitted models with 20% and 43% of the water-mediated native contacts predicted as predictors and scorers, respectively. Another new problem was the prediction of protein-carbohydrate binding, where our submitted model was very close to being acceptable. A set of targets were related to the prediction of binding affinities, in which our pyDock scheme was able to discriminate between natural and designed complexes with area under the curve = 83%. It was also proposed to estimate the effect of point mutations on binding affinity. Our approach, based on machine learning methods, showed high rates of correctly classified mutations for all cases. The overall results were highly rewarding, and show that the field is ready to move forward and face new interesting challenges in interactomics. Proteins 2013; 81:2192-2200. © 2013 Wiley Periodicals, Inc.

The covalent attachment of adenosine monophosphate (AMP) to proteins, a process called AMPylation (adenylylation), has recently emerged as a novel theme in microbial pathogenesis. Although several AMPylating enzymes have been characterized, the only known virulence protein with de-AMPylation activity is SidD from the human pathogen Legionella pneumophila. SidD de-AMPylates mammalian Rab1, a small GTPase involved in secretory vesicle transport, thereby targeting the host protein for inactivation. The molecular mechanisms underlying Rab1 recognition and de-AMPylation by SidD are unclear. Here, we report the crystal structure of the catalytic region of SidD at 1.6 Å resolution. The structure reveals a phosphatase-like fold with additional structural elements not present in generic PP2C-type phosphatases. The catalytic pocket contains a binuclear metal-binding site characteristic of hydrolytic metalloenzymes, with strong dependency on magnesium ions. Subsequent docking and molecular dynamics simulations between SidD and Rab1 revealed the interface contacts and the energetic contribution of key residues to the interaction. In conjunction with an extensive structure-based mutational analysis, we provide in vivo and in vitro evidence for a remarkable adaptation of SidD to its host cell target Rab1 which explains how this effector confers specificity to the reaction it catalyses., Author Summary The covalent attachment of adenosine monophosphate (AMP) to proteins, a process called AMPylation (adenylylation), has recently emerged as a novel theme in microbial pathogenesis. While AMPylases from various pathogenic microorganisms have recently been characterized, the only virulence protein with de-AMPylation activity known to date is the Legionella pneumophila effector SidD which catalyzes AMP removal from the host GTPase Rab1. Thus, both AMPylation and de-AMPylation constitute a novel catalytic mechanism to precisely control the function and membrane dynamics of a host Rab GTPase. In spite of this pivotal role, the molecular mechanism of AMP removal and the structural determinants for Rab1 recognition by SidD have remained largely unexplored. Here, we present the crystal structure of the de-adenylylation domain of SidD and reveal the catalytic mechanism of Rab1 de-adenylylation. Surprisingly, the structure of SidD is not related to the other known enzyme with de-AMPylation activity, the Escherichia coli GS-ATase. Instead, the catalitic domain of SidD is remarkably similar to that of the metal-dependent protein phosphatases (PPMs), however with distinctive structural features to distinguish AMPylated Rab1 from similarly modified substrates. Importantly, we provide a model for the SidD-Rab1 complex which sheds light into the specific details of substrate recognition and catalysis by this virulence factor.

To understand cellular processes at the molecular level we need to improve our knowledge of protein−protein interactions, from a structural, mechanistic, and energetic point of view. Current theoretical studies and computational docking simulations show that protein dynamics plays a key role in protein association and support the need for including protein flexibility in modeling protein interactions. Assuming the conformational selection binding mechanism, in which the unbound state can sample bound conformers, one possible strategy to include flexibility in docking predictions would be the use of conformational ensembles originated from unbound protein structures. Here we present an exhaustive computational study about the use of precomputed unbound ensembles in the context of protein docking, performed on a set of 124 cases of the Protein−Protein Docking Benchmark 3.0. Conformational ensembles were generated by conformational optimization and refinement with MODELLER and by short molecular dynamics trajectories with AMBER. We identified those conformers providing optimal binding and investigated the role of protein conformational heterogeneity in protein−protein recognition. Our results show that a restricted conformational refinement can generate conformers with better binding properties and improve docking encounters in medium-flexible cases. For more flexible cases, a more extended conformational sampling based on Normal Mode Analysis was proven helpful. We found that successful conformers provide better energetic complementarity to the docking partners, which is compatible with recent views of binding association. In addition to the mechanistic considerations, these findings could be exploited for practical docking predictions of improved efficiency.