The structural analysis of protein–protein interactions by NMR spectroscopy

Reconciling ASPP-p53 binding mode discrepancies through an ensemble binding framework that bridges crystallography and NMR data

ASPP2 and iASPP bind to p53 through their conserved ANK-SH3 domains to respectively promote and inhibit p53-dependent cell apoptosis. While crystallography has indicated that these two proteins employ distinct surfaces of their ANK-SH3 domains to bind to p53, solution NMR data has suggested similar surfaces. In this study, we employed multi-scale molecular dynamics (MD) simulations combined with free energy calculations to reconcile the discrepancy in the binding modes. We demonstrated that the binding mode based solely on a single crystal structure does not enable iASPP's RT loop to engage with p53's C-terminal linker-a verified interaction. Instead, an ensemble of simulated iASPP-p53 complexes facilitates this interaction. We showed that the ensemble-average inter-protein contacting residues and NMR-detected interfacial residues qualitatively overlap on ASPP proteins, and the ensemble-average binding free energies better match experimental KD values compared to single crystallgarphy-determined binding mode. For iASPP, the sampled ensemble complexes can be grouped into two classes, resembling the binding modes determined by crystallography and solution NMR. We thus propose that crystal packing shifts the equilibrium of binding modes towards the crystallography-determined one. Lastly, we showed that the ensemble binding complexes are sensitive to p53's intrinsically disordered regions (IDRs), attesting to experimental observations that these IDRs contribute to biological functions. Our results provide a dynamic and ensemble perspective for scrutinizing these important cancer-related protein-protein interactions (PPIs).

Fluorescence-Based Protein Stability Monitoring-A Review

Proteins are large biomolecules with a specific structure that is composed of one or more long amino acid chains. Correct protein structures are directly linked to their correct function, and many environmental factors can have either positive or negative effects on this structure. Thus, there is a clear need for methods enabling the study of proteins, their correct folding, and components affecting protein stability. There is a significant number of label-free methods to study protein stability. In this review, we provide a general overview of these methods, but the main focus is on fluorescence-based low-instrument and -expertise-demand techniques. Different aspects related to thermal shift assays (TSAs), also called differential scanning fluorimetry (DSF) or ThermoFluor, are introduced and compared to isothermal chemical denaturation (ICD). Finally, we discuss the challenges and comparative aspects related to these methods, as well as future opportunities and assay development directions.

Integrating unsupervised language model with multi-view multiple sequence alignments for high-accuracy inter-chain contact prediction

Accurate identification of inter-chain contacts in the protein complex is critical to determine the corresponding 3D structures and understand the biological functions. We proposed a new deep learning method, ICCPred, to deduce the inter-chain contacts from the amino acid sequences of the protein complex. This pipeline was built on the designed deep residual network architecture, integrating the pre-trained language model with three multiple sequence alignments (MSAs) from different biological views. Experimental results on 709 non-redundant benchmarking protein complexes showed that the proposed ICCPred significantly increased inter-chain contact prediction accuracy compared to the state-of-the-art approaches. Detailed data analyses showed that the significant advantage of ICCPred lies in the utilization of pre-trained transformer language models which can effectively extract the complementary co-evolution diversity from three MSAs. Meanwhile, the designed deep residual network enhances the correlation between the co-evolution diversity and the patterns of inter-chain contacts. These results demonstrated a new avenue for high-accuracy deep-learning inter-chain contact prediction that is applicable to large-scale protein-protein interaction annotations from sequence alone.

Chromatin and non-chromatin immunoprecipitations to capture protein-protein and protein-nucleic acid interactions in living cells

Proteins are expressed from genes via sequential biological processes of transcription, mRNA processing, export and translation, and play their roles in maintaining cellular functions via interactions with proteins, DNAs or RNAs. Thus, it is important to study the protein interactions during biological processes in living cells towards understanding their mechanisms-of-action in real time. Methodologies have been developed over the years to study protein interactions in vivo. One state-of-the-art approach is formaldehyde crosslinking-based immuno- or chemi-precipitation to analyze selective as well as genome/proteome-wide interactions in living cells. It is a popular and widely used methodology for cellular analysis of the protein-protein and protein-nucleic acid interactions. Here, we describe this approach to analyze protein-protein/nucleic acid interactions in vivo.

Structure of the Sec14 domain of Kalirin reveals a distinct class of lipid-binding module in RhoGEFs

Gated entry of lipophilic ligands into the enclosed hydrophobic pocket in stand-alone Sec14 domain proteins often links lipid metabolism to membrane trafficking. Similar domains occur in multidomain mammalian proteins that activate small GTPases and regulate actin dynamics. The neuronal RhoGEF Kalirin, a central regulator of cytoskeletal dynamics, contains a Sec14 domain (Kal^bSec14) followed by multiple spectrin-like repeats and catalytic domains. Previous studies demonstrated that Kalirin lacking its Sec14 domain fails to maintain cell morphology or dendritic spine length, yet whether and how Kal^bSec14 interacts with lipids remain unknown. Here, we report the structural and biochemical characterization of Kal^bSec14. Kal^bSec14 adopts a closed conformation, sealing off the canonical ligand entry site, and instead employs a surface groove to bind a limited set of lysophospholipids. The low-affinity interactions of Kal^bSec14 with lysolipids are expected to serve as a general model for the regulation of Rho signaling by other Sec14-containing Rho activators.

Attogram-level light-induced antigen-antibody binding confined in microflow

The analysis of trace amounts of proteins based on immunoassays and other methods is essential for the early diagnosis of various diseases such as cancer, dementia, and microbial infections. Here, we propose a light-induced acceleration of antigen-antibody reaction of attogram-level proteins at the solid-liquid interface by tuning the laser irradiation area comparable to the microscale confinement geometry for enhancing the collisional probability of target molecules and probe particles with optical force and fluidic pressure. This principle was applied to achieve a 10²-fold higher sensitivity and ultrafast specific detection in comparison with conventional protein detection methods (a few hours) by omitting any pretreatment procedures; 47–750 ag of target proteins were detected in 300 nL of sample after 3 minutes of laser irradiation. Our findings can promote the development of proteomics and innovative platforms for high-throughput bio-analyses under the control of a variety of biochemical reactions.

Attogram-level of proteins can be detected using light-induced acceleration of antigen-antibody interactions in a microchannel platform.

Integrated structure-based protein interface prediction

Background

Identifying protein interfaces can inform how proteins interact with their binding partners, uncover the regulatory mechanisms that control biological functions and guide the development of novel therapeutic agents. A variety of computational approaches have been developed for predicting a protein’s interfacial residues from its known sequence and structure. Methods using the known three-dimensional structures of proteins can be template-based or template-free. Template-based methods have limited success in predicting interfaces when homologues with known complex structures are not available to use as templates. The prediction performance of template-free methods that only rely only upon proteins’ intrinsic properties is limited by the amount of biologically relevant features that can be included in an interface prediction model.

Results

We describe the development of an integrated method for protein interface prediction (ISPIP) to explore the hypothesis that the efficacy of a computational prediction method of protein binding sites can be enhanced by using a combination of methods that rely on orthogonal structure-based properties of a query protein, combining and balancing both template-free and template-based features. ISPIP is a method that integrates these approaches through simple linear or logistic regression models and more complex decision tree models. On a diverse test set of 156 query proteins, ISPIP outperforms each of its individual classifiers in identifying protein binding interfaces.

Conclusions

The integrated method captures the best performance of individual classifiers and delivers an improved interface prediction. The method is robust and performs well even when one of the individual classifiers performs poorly on a particular query protein. This work demonstrates that integrating orthogonal methods that depend on different structural properties of proteins performs better at interface prediction than any individual classifier alone.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-022-04852-2.

HDIContact: a novel predictor of residue-residue contacts on hetero-dimer interfaces via sequential information and transfer learning strategy

Proteins maintain the functional order of cell in life by interacting with other proteins. Determination of protein complex structural information gives biological insights for the research of diseases and drugs. Recently, a breakthrough has been made in protein monomer structure prediction. However, due to the limited number of the known protein structure and homologous sequences of complexes, the prediction of residue-residue contacts on hetero-dimer interfaces is still a challenge. In this study, we have developed a deep learning framework for inferring inter-protein residue contacts from sequential information, called HDIContact. We utilized transfer learning strategy to produce Multiple Sequence Alignment (MSA) two-dimensional (2D) embedding based on patterns of concatenated MSA, which could reduce the influence of noise on MSA caused by mismatched sequences or less homology. For MSA 2D embedding, HDIContact took advantage of Bi-directional Long Short-Term Memory (BiLSTM) with two-channel to capture 2D context of residue pairs. Our comprehensive assessment on the Escherichia coli (E. coli) test dataset showed that HDIContact outperformed other state-of-the-art methods, with top precision of 65.96%, the Area Under the Receiver Operating Characteristic curve (AUROC) of 83.08% and the Area Under the Precision Recall curve (AUPR) of 25.02%. In addition, we analyzed the potential of HDIContact for human-virus protein-protein complexes, by achieving top five precision of 80% on O75475-P04584 related to Human Immunodeficiency Virus. All experiments indicated that our method was a valuable technical tool for predicting inter-protein residue contacts, which would be helpful for understanding protein-protein interaction mechanisms.

Predicting Protein-Protein Interactions via Random Ferns with Evolutionary Matrix Representation

Protein-protein interactions (PPIs) play a crucial role in understanding disease pathogenesis, genetic mechanisms, guiding drug design, and other biochemical processes, thus, the identification of PPIs is of great importance. With the rapid development of high-throughput sequencing technology, a large amount of PPIs sequence data has been accumulated. Researchers have designed many experimental methods to detect PPIs by using these sequence data, hence, the prediction of PPIs has become a research hotspot in proteomics. However, since traditional experimental methods are both time-consuming and costly, it is difficult to analyze and predict the massive amount of PPI data quickly and accurately. To address these issues, many computational systems employing machine learning knowledge were widely applied to PPIs prediction, thereby improving the overall recognition rate. In this paper, a novel and efficient computational technology is presented to implement a protein interaction prediction system using only protein sequence information. First, the Position-Specific Iterated Basic Local Alignment Search Tool (PSI-BLAST) was employed to generate a position-specific scoring matrix (PSSM) containing protein evolutionary information from the initial protein sequence. Second, we used a novel data processing feature representation scheme, MatFLDA, to extract the essential information of PSSM for protein sequences and obtained five training and five testing datasets by adopting a five-fold cross-validation method. Finally, the random fern (RFs) classifier was employed to infer the interactions among proteins, and a model called MatFLDA_RFs was developed. The proposed MatFLDA_RFs model achieved good prediction performance with 95.03% average accuracy on Yeast dataset and 85.35% average accuracy on H. pylori dataset, which effectively outperformed other existing computational methods. The experimental results indicate that the proposed method is capable of yielding better prediction results of PPIs, which provides an effective tool for the detection of new PPIs and the in-depth study of proteomics. Finally, we also developed a web server for the proposed model to predict protein-protein interactions, which is freely accessible online at http://120.77.11.78:5001/webserver/MatFLDA_RFs.

Solid-State NMR Dipolar and Chemical Shift Anisotropy Recoupling Techniques for Structural and Dynamical Studies in Biological Systems

With the development of NMR methodology and technology during the past decades, solid-state NMR (ssNMR) has become a particularly important tool for investigating structure and dynamics at atomic scale in biological systems, where the recoupling techniques play pivotal roles in modern high-resolution MAS NMR. In this review, following a brief introduction on the basic theory of recoupling in ssNMR, we highlight the recent advances in dipolar and chemical shift anisotropy recoupling methods, as well as their applications in structural determination and dynamical characterization at multiple time scales (i.e., fast-, intermediate-, and slow-motion). The performances of these prevalent recoupling techniques are compared and discussed in multiple aspects, together with the representative applications in biomolecules. Given the recent emerging advances in NMR technology, new challenges for recoupling methodology development and potential opportunities for biological systems are also discussed.

NMR spectroscopy of wastewater: A review, case study, and future potential

NMR spectroscopy is arguably the most powerful tool for the study of molecular structures and interactions, and is increasingly being applied to environmental research, such as the study of wastewater. With over 97% of the planet's water being saltwater, and two thirds of freshwater being frozen in the ice caps and glaciers, there is a significant need to maintain and reuse the remaining 1%, which is a precious resource, critical to the sustainability of most life on Earth. Sanitation and reutilization of wastewater is an important method of water conservation, especially in arid regions, making the understanding of wastewater itself, and of its treatment processes, a highly relevant area of environmental research. Here, the benefits, challenges and subtleties of using NMR spectroscopy for the analysis of wastewater are considered. First, the techniques available to overcome the specific challenges arising from the nature of wastewater (which is a complex and dilute matrix), including an examination of sample preparation and NMR techniques (such as solvent suppression), in both the solid and solution states, are discussed. Then, the arsenal of available NMR techniques for both structure elucidation (e.g., heteronuclear, multidimensional NMR, homonuclear scalar coupling-based experiments) and the study of intermolecular interactions (e.g., diffusion, nuclear Overhauser and saturation transfer-based techniques) in wastewater are examined. Examples of wastewater NMR studies from the literature are reviewed and potential areas for future research are identified. Organized by nucleus, this review includes the common heteronuclei (¹³C, ¹⁵N, ¹⁹F, ³¹P, ²⁹Si) as well as other environmentally relevant nuclei and metals such as ²⁷Al, ⁵¹V, ²⁰⁷Pb and ¹¹³Cd, among others. Further, the potential of additional NMR methods such as comprehensive multiphase NMR, NMR microscopy and hyphenated techniques (for example, LC-SPE-NMR-MS) for advancing the current understanding of wastewater are discussed. In addition, a case study that combines natural abundance (i.e. non-concentrated), targeted and non-targeted NMR to characterize wastewater, along with in vivo based NMR to understand its toxicity, is included. The study demonstrates that, when applied comprehensively, NMR can provide unique insights into not just the structure, but also potential impacts, of wastewater and wastewater treatment processes. Finally, low-field NMR, which holds considerable future potential for on-site wastewater monitoring, is briefly discussed. In summary, NMR spectroscopy is one of the most versatile tools in modern science, with abilities to study all phases (gases, liquids, gels and solids), chemical structures, interactions, interfaces, toxicity and much more. The authors hope this review will inspire more scientists to embrace NMR, given its huge potential for both wastewater analysis in particular and environmental research in general.

Fluorescence resonance energy transfer in revealing protein-protein interactions in living cells

Genes are expressed to proteins for a wide variety of fundamental biological processes at the cellular and organismal levels. However, a protein rarely functions alone, but rather acts through interactions with other proteins to maintain normal cellular and organismal functions. Therefore, it is important to analyze the protein-protein interactions to determine functional mechanisms of proteins, which can also guide to develop therapeutic targets for treatment of diseases caused by altered protein-protein interactions leading to cellular/organismal dysfunctions. There is a large number of methodologies to study protein interactions in vitro, in vivo and in silico, which led to the development of many protein interaction databases, and thus, have enriched our knowledge about protein-protein interactions and functions. However, many of these interactions were identified in vitro, but need to be verified/validated in living cells. Furthermore, it is unclear whether these interactions are direct or mediated via other proteins. Moreover, these interactions are representative of cell- and time-average, but not a single cell in real time. Therefore, it is crucial to detect direct protein-protein interactions in a single cell during biological processes in vivo, towards understanding the functional mechanisms of proteins in living cells. Importantly, a fluorescence resonance energy transfer (FRET)-based methodology has emerged as a powerful technique to decipher direct protein-protein interactions at a single cell resolution in living cells, which is briefly described in a limited available space in this mini-review.

Expression, Purification, and Solution-State NMR Analysis of the Two Human Single-Stranded DNA-Binding Proteins hSSB1 (NABP2/OBFC2B) and hSSB2 (NAPB1/OBFC2A)

Single-stranded DNA-binding proteins (SSBs) are essential to all living organisms as protectors and guardians of the genome. Apart from the well-characterized RPA, humans have also evolved two further SSBs, termed hSSB1 and hSSB2. Over the last few years, we have used NMR spectroscopy to determine the molecular and structural details of both hSSBs and their interactions with DNA and RNA. Here we provide a detailed overview of how to express and purify recombinant versions of these important human proteins for the purpose of detailed structural analysis by high-resolution solution-state NMR.

A Practical Perspective on the Roles of Solution NMR Spectroscopy in Drug Discovery

Solution nuclear magnetic resonance (NMR) spectroscopy is a powerful tool to study structures and dynamics of biomolecules under physiological conditions. As there are numerous NMR-derived methods applicable to probe protein–ligand interactions, NMR has been widely utilized in drug discovery, especially in such steps as hit identification and lead optimization. NMR is frequently used to locate ligand-binding sites on a target protein and to determine ligand binding modes. NMR spectroscopy is also a unique tool in fragment-based drug design (FBDD), as it is able to investigate target-ligand interactions with diverse binding affinities. NMR spectroscopy is able to identify fragments that bind weakly to a target, making it valuable for identifying hits targeting undruggable sites. In this review, we summarize the roles of solution NMR spectroscopy in drug discovery. We describe some methods that are used in identifying fragments, understanding the mechanism of action for a ligand, and monitoring the conformational changes of a target induced by ligand binding. A number of studies have proven that ¹⁹F-NMR is very powerful in screening fragments and detecting protein conformational changes. In-cell NMR will also play important roles in drug discovery by elucidating protein-ligand interactions in living cells.

Rapid 3-dimensional shape determination of globular proteins by mobility capillary electrophoresis and native mass spectrometry

Established high-throughput proteomics methods provide limited information on the stereostructures of proteins. Traditional technologies for protein structure determination typically require laborious steps and cannot be performed in a high-throughput fashion. Here, we report a new medium throughput method by combining mobility capillary electrophoresis (MCE) and native mass spectrometry (MS) for the 3-dimensional (3D) shape determination of globular proteins in the liquid phase, which provides both the geometric structure and molecular mass information of proteins. A theory was established to correlate the ion hydrodynamic radius and charge state distribution in the native mass spectrum with protein geometrical parameters, through which a low-resolution structure (shape) of the protein could be determined. Our test data of 11 different globular proteins showed that this approach allows us to determine the shapes of individual proteins, protein complexes and proteins in a mixture, and to monitor protein conformational changes. Besides providing complementary protein structure information and having mixture analysis capability, this MCE and native MS based method is fast in speed and low in sample consumption, making it potentially applicable in top–down proteomics and structural biology for intact globular protein or protein complex analysis.

Using native mass spectrometry and mobility capillary electrophoresis, the ellipsoid dimensions of globular proteins or protein complexes could be measured efficiently.

Metallothionein Cd4S11cluster formation dominates in the protection of carbonic anhydrase

Results from ESI-MS and stopped flow kinetics show that apo-MT protects from toxic metalation of apo-CA with Cd²⁺due to the protein–protein interactions in solution.

BIPSPI: a method for the prediction of partner-specific protein-protein interfaces

Motivation

Protein-Protein Interactions (PPI) are essentials for most cellular processes and thus, unveiling how proteins interact is a crucial question that can be better understood by identifying which residues are responsible for the interaction. Computational approaches are orders of magnitude cheaper and faster than experimental ones, leading to proliferation of multiple methods aimed to predict which residues belong to the interface of an interaction.

Results

We present BIPSPI, a new machine learning-based method for the prediction of partner-specific PPI sites. Contrary to most binding site prediction methods, the proposed approach takes into account a pair of interacting proteins rather than a single one in order to predict partner-specific binding sites. BIPSPI has been trained employing sequence-based and structural features from both protein partners of each complex compiled in the Protein-Protein Docking Benchmark version 5.0 and in an additional set independently compiled. Also, a version trained only on sequences has been developed. The performance of our approach has been assessed by a leave-one-out cross-validation over different benchmarks, outperforming state-of-the-art methods.

Availability and implementation

BIPSPI web server is freely available at http://bipspi.cnb.csic.es. BIPSPI code is available at https://github.com/bioinsilico/BIPSPI. Docker image is available at https://hub.docker.com/r/bioinsilico/bipspi/.

Supplementary information

Supplementary data are available at Bioinformatics online.

Structural and functional analyses of the N-terminal domain of the A subunit of a Bacillus megaterium spore germinant receptor

Germination of Bacillus spores is induced by the interaction of specific nutrient molecules with germinant receptors (GRs) localized in the spore's inner membrane. GRs typically consist of three subunits referred to as A, B, and C, although functions of individual subunits are not known. Here we present the crystal structure of the N-terminal domain (NTD) of the A subunit of the Bacillus megaterium GerK₃ GR, revealing two distinct globular subdomains bisected by a cleft, a fold with strong homology to substrate-binding proteins in bacterial ABC transporters. Molecular docking, chemical shift perturbation measurement, and mutagenesis coupled with spore germination analyses support a proposed model that the interface between the two subdomains in the NTD of GR A subunits serves as the germinant binding site and plays a critical role in spore germination. Our findings provide a conceptual framework for understanding the germinant recruitment mechanism by which GRs trigger spore germination.

FRETting about the affinity of bimolecular protein-protein interactions

Fluorescence resonance energy transfer (FRET) is a powerful tool to study macromolecular interactions such as protein-protein interactions (PPIs). Fluorescent protein (FP) fusions enable FRET-based PPI analysis of signaling pathways and molecular structure in living cells. Despite FRET's importance in PPI studies, FRET has seen limited use in quantifying the affinities of PPIs in living cells. Here, we have explored the relationship between FRET efficiency and PPI affinity over a wide range when expressed from a single plasmid system in Escherichia coli. Using live-cell microscopy and a set of 20 pairs of small interacting proteins, belonging to different structural folds and interaction affinities, we demonstrate that FRET efficiency can reliably measure the dissociation constant (K_D ) over a range of mM to nM. A 10-fold increase in the interaction affinity results in 0.05 unit increase in FRET efficiency, providing sufficient resolution to quantify large affinity differences (> 10-fold) using live-cell FRET. This approach provides a rapid and simple strategy for assessment of PPI affinities over a wide range and will have utility for high-throughput analysis of protein interactions.

Lead Identification Through the Synergistic Action of Biomolecular NMR and In Silico Methodologies

The combination of virtual screening with biomolecular NMR can be a powerful approach in the first steps toward drug discovery. Here, we describe how computational methodologies to screen large databases readily available for testing small molecules, in synergy with NMR techniques focused on protein-ligand interactions, can be used in the early lead compound identification process against a protein drug target.

Characterizing Protein-Protein Interactions Using Solution NMR Spectroscopy

In this chapter, we describe how NMR chemical shift titrations can be used to study the interaction between two proteins with emphasis on mapping the interface of the complex and determining the binding affinity from a quantitative analysis of the experimental data. In particular, we discuss the appearance of NMR spectra in different chemical exchange regimes (fast, intermediate, and slow) and how these regimes affect NMR data analysis.

Unspinning chromatin: Revealing the dynamic nucleosome landscape by NMR

NMR is an essential technique for obtaining information at atomic resolution on the structure, motions and interactions of biomolecules. Here, we review the contribution of NMR to our understanding of the fundamental unit of chromatin: the nucleosome. Nucleosomes compact the genome by wrapping the DNA around a protein core, the histone octamer, thereby protecting genomic integrity. Crucially, the imposed barrier also allows strict regulation of gene expression, DNA replication and DNA repair processes through an intricate system of histone and DNA modifications and a wide range of interactions between nucleosomes and chromatin factors. In this review, we describe how NMR has contributed to deciphering the molecular basis of nucleosome function. Starting from pioneering studies in the 1960s using natural abundance NMR studies, we focus on the progress in sample preparation and NMR methodology that has allowed high-resolution studies on the nucleosome and its subunits. We summarize the results and approaches of state-of-the-art NMR studies on nucleosomal DNA, histone complexes, nucleosomes and nucleosomal arrays. These studies highlight the particular strength of NMR in studying nucleosome dynamics and nucleosome-protein interactions. Finally, we look ahead to exciting new possibilities that will be afforded by on-going developments in solution and solid-state NMR. By increasing both the depth and breadth of nucleosome NMR studies, it will be possible to offer a unique perspective on the dynamic landscape of nucleosomes and its interacting proteins.

Monitoring Interactions Between S100B and the Dopamine D2 Receptor Using NMR Spectroscopy

S100B is a dimeric EF-hand protein that undergoes a calcium-induced conformational change and interacts with a wide range of proteins to modulate their functions. The dopamine D2 receptor is one potential S100B binding partner that may play a key role in neurological processing. In this chapter, we describe the use of NMR spectroscopy to examine the interaction between calcium-bound S100B and the third intracellular loop (IC3) from the dopamine D2 receptor. We provide details that allow the strength of the interaction (K _d) between the two proteins to be determined and the IC3 site of interaction on the structure of S100B to be identified. Both these characteristics can be identified from a single series of nondestructive experiments.

Exploring the potential of 3D Zernike descriptors and SVM for protein-protein interface prediction

Background

The correct determination of protein–protein interaction interfaces is important for understanding disease mechanisms and for rational drug design. To date, several computational methods for the prediction of protein interfaces have been developed, but the interface prediction problem is still not fully understood. Experimental evidence suggests that the location of binding sites is imprinted in the protein structure, but there are major differences among the interfaces of the various protein types: the characterising properties can vary a lot depending on the interaction type and function. The selection of an optimal set of features characterising the protein interface and the development of an effective method to represent and capture the complex protein recognition patterns are of paramount importance for this task.

Results

In this work we investigate the potential of a novel local surface descriptor based on 3D Zernike moments for the interface prediction task. Descriptors invariant to roto-translations are extracted from circular patches of the protein surface enriched with physico-chemical properties from the HQI8 amino acid index set, and are used as samples for a binary classification problem. Support Vector Machines are used as a classifier to distinguish interface local surface patches from non-interface ones. The proposed method was validated on 16 classes of proteins extracted from the Protein–Protein Docking Benchmark 5.0 and compared to other state-of-the-art protein interface predictors (SPPIDER, PrISE and NPS-HomPPI).

Conclusions

The 3D Zernike descriptors are able to capture the similarity among patterns of physico-chemical and biochemical properties mapped on the protein surface arising from the various spatial arrangements of the underlying residues, and their usage can be easily extended to other sets of amino acid properties. The results suggest that the choice of a proper set of features characterising the protein interface is crucial for the interface prediction task, and that optimality strongly depends on the class of proteins whose interface we want to characterise. We postulate that different protein classes should be treated separately and that it is necessary to identify an optimal set of features for each protein class.

Electronic supplementary material

The online version of this article (10.1186/s12859-018-2043-3) contains supplementary material, which is available to authorized users.

The adherens junction-associated LIM domain protein Smallish regulates epithelial morphogenesis

Cell–cell adhesion and cell shape are regulated at adherens junctions during embryonic morphogenesis. Beati et al. show that the Drosophila LIM domain protein Smallish interacts with Bazooka, Canoe, and Src42A at adherens junctions. Loss-of-function and gain-of-function phenotypes reveal a function for Smallish in regulation of actomyosin contractility and cell shape.

In epithelia, cells adhere to each other in a dynamic fashion, allowing the cells to change their shape and move along each other during morphogenesis. The regulation of adhesion occurs at the belt-shaped adherens junction, the zonula adherens (ZA). Formation of the ZA depends on components of the Par–atypical PKC (Par-aPKC) complex of polarity regulators. We have identified the Lin11, Isl-1, Mec-3 (LIM) protein Smallish (Smash), the orthologue of vertebrate LMO7, as a binding partner of Bazooka/Par-3 (Baz), a core component of the Par-aPKC complex. Smash also binds to Canoe/Afadin and the tyrosine kinase Src42A and localizes to the ZA in a planar polarized fashion. Animals lacking Smash show loss of planar cell polarity (PCP) in the embryonic epidermis and reduced cell bond tension, leading to severe defects during embryonic morphogenesis of epithelial tissues and organs. Overexpression of Smash causes apical constriction of epithelial cells. We propose that Smash is a key regulator of morphogenesis coordinating PCP and actomyosin contractility at the ZA.

Predictive and Experimental Approaches for Elucidating Protein-Protein Interactions and Quaternary Structures

The elucidation of protein-protein interactions is vital for determining the function and action of quaternary protein structures. Here, we discuss the difficulty and importance of establishing protein quaternary structure and review in vitro and in silico methods for doing so. Determining the interacting partner proteins of predicted protein structures is very time-consuming when using in vitro methods, this can be somewhat alleviated by use of predictive methods. However, developing reliably accurate predictive tools has proved to be difficult. We review the current state of the art in predictive protein interaction software and discuss the problem of scoring and therefore ranking predictions. Current community-based predictive exercises are discussed in relation to the growth of protein interaction prediction as an area within these exercises. We suggest a fusion of experimental and predictive methods that make use of sparse experimental data to determine higher resolution predicted protein interactions as being necessary to drive forward development.

Solution NMR Spectroscopy in Target-Based Drug Discovery

Solution NMR spectroscopy is a powerful tool to study protein structures and dynamics under physiological conditions. This technique is particularly useful in target-based drug discovery projects as it provides protein-ligand binding information in solution. Accumulated studies have shown that NMR will play more and more important roles in multiple steps of the drug discovery process. In a fragment-based drug discovery process, ligand-observed and protein-observed NMR spectroscopy can be applied to screen fragments with low binding affinities. The screened fragments can be further optimized into drug-like molecules. In combination with other biophysical techniques, NMR will guide structure-based drug discovery. In this review, we describe the possible roles of NMR spectroscopy in drug discovery. We also illustrate the challenges encountered in the drug discovery process. We include several examples demonstrating the roles of NMR in target-based drug discoveries such as hit identification, ranking ligand binding affinities, and mapping the ligand binding site. We also speculate the possible roles of NMR in target engagement based on recent processes in in-cell NMR spectroscopy.

Overcoming bottlenecks in the membrane protein structural biology pipeline

Membrane proteins account for a third of the eukaryotic proteome, but are greatly under-represented in the Protein Data Bank. Unfortunately, recent technological advances in X-ray crystallography and EM cannot account for the poor solubility and stability of membrane protein samples. A limitation of conventional detergent-based methods is that detergent molecules destabilize membrane proteins, leading to their aggregation. The use of orthologues, mutants and fusion tags has helped improve protein stability, but at the expense of not working with the sequence of interest. Novel detergents such as glucose neopentyl glycol (GNG), maltose neopentyl glycol (MNG) and calixarene-based detergents can improve protein stability without compromising their solubilizing properties. Styrene maleic acid lipid particles (SMALPs) focus on retaining the native lipid bilayer of a membrane protein during purification and biophysical analysis. Overcoming bottlenecks in the membrane protein structural biology pipeline, primarily by maintaining protein stability, will facilitate the elucidation of many more membrane protein structures in the near future.

DeepPPI: Boosting Prediction of Protein-Protein Interactions with Deep Neural Networks

The complex language of eukaryotic gene expression remains incompletely understood. Despite the importance suggested by many proteins variants statistically associated with human disease, nearly all such variants have unknown mechanisms, for example, protein-protein interactions (PPIs). In this study, we address this challenge using a recent machine learning advance-deep neural networks (DNNs). We aim at improving the performance of PPIs prediction and propose a method called DeepPPI (Deep neural networks for Protein-Protein Interactions prediction), which employs deep neural networks to learn effectively the representations of proteins from common protein descriptors. The experimental results indicate that DeepPPI achieves superior performance on the test data set with an Accuracy of 92.50%, Precision of 94.38%, Recall of 90.56%, Specificity of 94.49%, Matthews Correlation Coefficient of 85.08% and Area Under the Curve of 97.43%, respectively. Extensive experiments show that DeepPPI can learn useful features of proteins pairs by a layer-wise abstraction, and thus achieves better prediction performance than existing methods. The source code of our approach can be available via http://ailab.ahu.edu.cn:8087/DeepPPI/index.html .

Human βB2-Crystallin Forms a Face-en-Face Dimer in Solution: An Integrated NMR and SAXS Study

βγ-Crystallins are long-lived eye lens proteins that are crucial for lens transparency and refractive power. Each βγ-crystallin comprises two homologous domains, which are connected by a short linker. γ-Crystallins are monomeric, while β-crystallins crystallize as dimers and multimers. In the crystal, human βB2-crystallin is a domain-swapped dimer while the N-terminally truncated βB1-crystallin forms a face-en-face dimer. Combining and integrating data from multi-angle light scattering, nuclear magnetic resonance, and small-angle X-ray scattering of full-length and terminally truncated human βB2-crystallin in solution, we show that both these βB2-crystallin proteins are dimeric, possess C2 symmetry, and are more compact than domain-swapped dimers. Importantly, no inter-molecular paramagnetic relaxation enhancement effects compatible with domain swapping were detected. Our collective experimental results unambiguously demonstrate that, in solution, human βB2-crystallin is not domain swapped and exhibits a face-en-face dimer structure similar to the crystal structure of truncated βB1-crystallin.

Detecting reliable non interacting proteins (NIPs) significantly enhancing the computational prediction of protein-protein interactions using machine learning methods

Protein-protein interactions (PPIs) play a vital role in most biological processes. Hence their comprehension can promote a better understanding of the mechanisms underlying living systems. However, besides the cost and the time limitation involved in the detection of experimentally validated PPIs, the noise in the data is still an important issue to overcome. In the last decade several in silico PPI prediction methods using both structural and genomic information were developed for this purpose. Here we introduce a unique validation approach aimed to collect reliable non interacting proteins (NIPs). Thereafter the most relevant protein/protein-pair related features were selected. Finally, the prepared dataset was used for PPI classification, leveraging the prediction capabilities of well-established machine learning methods. Our best classification procedure displayed specificity and sensitivity values of 96.33% and 98.02%, respectively, surpassing the prediction capabilities of other methods, including those trained on gold standard datasets. We showed that the PPI/NIP predictive performances can be considerably improved by focusing on data preparation.

In Vivo Analysis of Protein-Protein Interactions with Bioluminescence Resonance Energy Transfer (BRET): Progress and Prospects

Proteins are the elementary machinery of life, and their functions are carried out mostly by molecular interactions. Among those interactions, protein-protein interactions (PPIs) are the most important as they participate in or mediate all essential biological processes. However, many common methods for PPI investigations are slightly unreliable and suffer from various limitations, especially in the studies of dynamic PPIs. To solve this problem, a method called Bioluminescence Resonance Energy Transfer (BRET) was developed about seventeen years ago. Since then, BRET has evolved into a whole class of methods that can be used to survey virtually any kinds of PPIs. Compared to many traditional methods, BRET is highly sensitive, reliable, easy to perform, and relatively inexpensive. However, most importantly, it can be done in vivo and allows the real-time monitoring of dynamic PPIs with the easily detectable light signal, which is extremely valuable for the PPI functional research. This review will take a comprehensive look at this powerful technique, including its principles, comparisons with other methods, experimental approaches, classifications, applications, early developments, recent progress, and prospects.

Heterologous expression of Cenchritis muricatus protease inhibitor II (CmPI-II) in Pichia pastoris system: Purification, isotopic labeling and preliminary characterization

Cenchritis muricatus protease inhibitor II (CmPI-II) is a tight-binding serine protease inhibitor of the Kazal family with an atypical broad specificity, being active against several proteases such as bovine pancreatic trypsin, human neutrophil elastase and subtilisin A. CmPI-II 3D structures are necessary for understanding the molecular basis of its activity. In the present work, we describe an efficient and straightforward recombinant expression strategy, as well as a cost-effective procedure for isotope labeling for NMR structure determination purposes. The vector pCM101 containing the CmPI-II gene, under the control of Pichia pastoris AOX1 promoter was constructed. Methylotrophic Pichia pastoris strain KM71H was then transformed with the plasmid and the recombinant protein (rCmPI-II) was expressed in benchtop fermenter in unlabeled or (15)N-labeled forms using ammonium chloride ((15)N, 99%) as the sole nitrogen source. Protein purification was accomplished by sequential cation exchange chromatography in STREAMLINE DirectHST, anion exchange chromatography on Hitrap Q-Sepharose FF and gel filtration on Superdex 75 10/30, yielding high quantities of pure rCmPI-II and (15)N rCmPI-II. Recombinant proteins displayed similar functional features as compared to the natural inhibitor and NMR spectra indicated folded and homogeneously labeled samples, suitable for further studies of structure and protease-inhibitor interactions.

Global De Novo Protein-Protein Interactome Elucidates Interactions of Drought-Responsive Proteins in Horse Gram (Macrotyloma uniflorum)

Inspired by the availability of de novo transcriptome of horse gram (Macrotyloma uniflorum) and recent developments in systems biology studies, the first ever global protein-protein interactome (PPI) map was constructed for this highly drought-tolerant legume. Large-scale studies of PPIs and the constructed database would provide rationale behind the interplay at cascading translational levels for drought stress-adaptive mechanisms in horse gram. Using a bidirectional approach (interolog and domain-based), a high-confidence interactome map and database for horse gram was constructed. Available transcriptomic information for shoot and root tissues of a sensitive (M-191; genotype 1) and a drought-tolerant (M-249; genotype 2) genotype of horse gram was utilized to draw comparative PPI subnetworks under drought stress. High-confidence 6804 interactions were predicted among 1812 proteins covering about one-fourth of the horse gram proteome. The highest number of interactions (33.86%) in horse gram interactome matched with Arabidopsis PPI data. The top five hub nodes mostly included ubiquitin and heat-shock-related proteins. Higher numbers of PPIs were found to be responsive in shoot tissue (416) and root tissue (2228) of genotype 2 compared with shoot tissue (136) and root tissue (579) of genotype 1. Characterization of PPIs using gene ontology analysis revealed that kinase and transferase activities involved in signal transduction, cellular processes, nucleocytoplasmic transport, protein ubiquitination, and localization of molecules were most responsive to drought stress. Hence, these could be framed in stress adaptive mechanisms of horse gram. Being the first legume global PPI map, it would provide new insights into gene and protein regulatory networks for drought stress tolerance mechanisms in horse gram. Information compiled in the form of database (MauPIR) will provide the much needed high-confidence systems biology information for horse gram genes, proteins, and involved processes. This information would ease the effort and increase the efficacy for similar studies on other legumes. Public access is available at http://14.139.59.221/MauPIR/ .

Cool-temperature-mediated activation of phospholipase C-γ2 in the human hereditary disease PLAID

Deletions in the gene encoding signal-transducing inositol phospholipid-specific phospholipase C-γ2 (PLCγ2) are associated with the novel human hereditary disease PLAID (PLCγ2-associated antibody deficiency and immune dysregulation). PLAID is characterized by a rather puzzling concurrence of augmented and diminished functions of the immune system, such as cold urticaria triggered by only minimal decreases in temperature, autoimmunity, and immunodeficiency. Understanding of the functional effects of the genomic alterations at the level of the affected enzyme, PLCγ2, is currently lacking. PLCγ2 is critically involved in coupling various cell surface receptors to regulation of important functions of immune cells such as mast cells, B cells, monocytes/macrophages, and neutrophils. PLCγ2 is unique by carrying three Src (SH) and one split pleckstrin homology domain (spPH) between the two catalytic subdomains (spPHn-SH2n-SH2c-SH3-spPHc). Prevailing evidence suggests that activation of PLCγ2 is primarily due to loss of SH-region-mediated autoinhibition and/or enhanced plasma membrane translocation. Here, we show that the two PLAID PLCγ2 mutants lacking portions of the SH region are strongly (>100-fold), rapidly, and reversibly activated by cooling by only a few degrees. We found that the mechanism(s) underlying PLCγ2 PLAID mutant activation by cool temperatures is distinct from a mere loss of SH-region-mediated autoinhibition and dependent on both the integrity and the pliability of the spPH domain. The results suggest a new mechanism of PLCγ activation with unique thermodynamic features and assign a novel regulatory role to its spPH domain. Involvement of this mechanism in other human disease states associated with cooling such as exertional asthma and certain acute coronary events appears an intriguing possibility.

Current Experimental Methods for Characterizing Protein-Protein Interactions

Protein molecules often interact with other partner protein molecules in order to execute their vital functions in living organisms. Characterization of protein-protein interactions thus plays a central role in understanding the molecular mechanism of relevant protein molecules, elucidating the cellular processes and pathways relevant to health or disease for drug discovery, and charting large-scale interaction networks in systems biology research. A whole spectrum of methods, based on biophysical, biochemical, or genetic principles, have been developed to detect the time, space, and functional relevance of protein-protein interactions at various degrees of affinity and specificity. This article presents an overview of these experimental methods, outlining the principles, strengths and limitations, and recent developments of each type of method.

Current Experimental Methods for Characterizing Protein-Protein Interactions

Protein molecules often interact with other partner protein molecules in order to execute their vital functions in living organisms. Characterization of protein-protein interactions thus plays a central role in understanding the molecular mechanism of relevant protein molecules, elucidating the cellular processes and pathways relevant to health or disease for drug discovery, and charting large-scale interaction networks in systems biology research. A whole spectrum of methods, based on biophysical, biochemical, or genetic principles, have been developed to detect the time, space, and functional relevance of protein-protein interactions at various degrees of affinity and specificity. This article presents an overview of these experimental methods, outlining the principles, strengths and limitations, and recent developments of each type of method.

Protein-protein interaction analysis by nuclear magnetic resonance spectroscopy

Nuclear magnetic resonance (NMR) has continued to evolve as a powerful method, with an increase in the number of pulse sequences and techniques available to study protein-protein interactions. In this chapter, a straightforward method to map a protein-protein interface and design a structural model is described, using chemical shift perturbation, paramagnetic relaxation enhancement, and data-driven docking.

Molecular determinants of tetramerization in the KcsA cytoplasmic domain

The cytoplasmic C-terminal domain (CTD) of KcsA, a bacterial homotetrameric potassium channel, is an amphiphilic domain that forms a helical bundle with four-fold symmetry mediated by hydrophobic and electrostatic interactions. Previously we have established that a CTD-derived 34-residue peptide associates into a tetramer in a pH-dependent manner (Kamnesky et al., JMB 2012;418:237-247). Here we further investigate the molecular determinants of tetramer formation in the CTD by characterizing the kinetics of monomer-tetramer equilibrium for 10 alanine mutants using NMR, sedimentation equilibrium (SE) and molecular dynamics simulation. NMR and SE concur in finding single-residue contributions to tetramer stability to be in the 0.5 to 3.5 kcal/mol range. Hydrophobic interactions between residues lining the tetramer core generally contributed more to formation of tetramer than electrostatic interactions between residues R147, D149 and E152. In particular, alanine replacement of residue R147, a key contributor to inter-subunit salt bridges, resulted in only a minor effect on tetramer dissociation. Mutations outside of the inter-subunit interface also influenced tetramer stability by affecting the tetramerization on-rate, possibly by changing the inherent helical propensity of the peptide. These findings are interpreted in the context of established paradigms of protein-protein interactions and protein folding, and lay the groundwork for further studies of the CTD in full-length KcsA channels.

Protein-protein interaction detection: methods and analysis

Protein-protein interaction plays key role in predicting the protein function of target protein and drug ability of molecules. The majority of genes and proteins realize resulting phenotype functions as a set of interactions. The in vitro and in vivo methods like affinity purification, Y2H (yeast 2 hybrid), TAP (tandem affinity purification), and so forth have their own limitations like cost, time, and so forth, and the resultant data sets are noisy and have more false positives to annotate the function of drug molecules. Thus, in silico methods which include sequence-based approaches, structure-based approaches, chromosome proximity, gene fusion, in silico 2 hybrid, phylogenetic tree, phylogenetic profile, and gene expression-based approaches were developed. Elucidation of protein interaction networks also contributes greatly to the analysis of signal transduction pathways. Recent developments have also led to the construction of networks having all the protein-protein interactions using computational methods for signaling pathways and protein complex identification in specific diseases.

Protein-protein interactions: switch from classical methods to proteomics and bioinformatics-based approaches

Following the sequencing of the human genome and many other organisms, research on protein-coding genes and their functions (functional genomics) has intensified. Subsequently, with the observation that proteins are indeed the molecular effectors of most cellular processes, the discipline of proteomics was born. Clearly, proteins do not function as single entities but rather as a dynamic network of team players that have to communicate. Though genetic (yeast two-hybrid Y2H) and biochemical methods (co-immunoprecipitation Co-IP, affinity purification AP) were the methods of choice at the beginning of the study of protein-protein interactions (PPI), in more recent years there has been a shift towards proteomics-based methods and bioinformatics-based approaches. In this review, we first describe in depth PPIs and we make a strong case as to why unraveling the interactome is the next challenge in the field of proteomics. Furthermore, classical methods of investigation of PPIs and structure-based bioinformatics approaches are presented. The greatest emphasis is placed on proteomic methods, especially native techniques that were recently developed and that have been shown to be reliable. Finally, we point out the limitations of these methods and the need to set up a standard for the validation of PPI experiments.

Phosphorylation of human Tau protein by microtubule affinity-regulating kinase 2

Tau protein plays an important role in neuronal physiology and Alzheimer's neurodegeneration. Its abilities to aggregate abnormally, to bind to microtubules (MTs), and to promote MT assembly are all influenced by phosphorylation. Phosphorylation of serine residues in the KXGS motifs of Tau's repeat domain, crucial for MT interactions and aggregation, is facilitated most efficiently by microtubule-associated protein/microtubule affinity-regulating kinases (MARKs). Here we applied high-resolution nuclear magnetic resonance analysis to study the kinetics of phosphorylation of Tau by MARK2 and its impact on the structure and microtubule binding of Tau. We demonstrate that MARK2 binds to the N-terminal tail of Tau and selectively phosphorylates three major and five minor serine residues in the repeat domain and C-terminal tail. Structural changes induced by phosphorylation of Tau by MARK2 are highly localized in the proximity of the phosphorylation site and do not affect the global conformation, in contrast to phosphorylation in the proline-rich region. Furthermore, single-residue analysis of binding of Tau to MTs provides support for a model in which Tau's hot spots of MT interaction bind independently of each other and are differentially affected by phosphorylation.

Structural basis of a physical blockage mechanism for the interaction of response regulator PmrA with connector protein PmrD from Klebsiella pneumoniae

In bacteria, the two-component system is the most prevalent for sensing and transducing environmental signals into the cell. The PmrA-PmrB two-component system, responsible for sensing external stimuli of high Fe(3+) and mild acidic conditions, can control the genes involved in lipopolysaccharide modification and polymyxin resistance in pathogens. In Klebsiella pneumoniae, the small basic connector protein PmrD protects phospho-PmrA and prolongs the expression of PmrA-activated genes. We previously determined the phospho-PmrA recognition mode of PmrD. However, how PmrA interacts with PmrD and prevents its dephosphorylation remains unknown. To address this question, we solved the x-ray crystal structure of the N-terminal receiver domain of BeF3(-)-activated PmrA (PmrA(N)) at 1.70 Å. With this structure, we applied the data-driven docking method based on NMR chemical shift perturbation to generate the complex model of PmrD-PmrA(N), which was further validated by site-directed spin labeling experiments. In the complex model, PmrD may act as a blockade to prevent phosphatase from contacting with the phosphorylation site on PmrA.

An introduction to biological NMR spectroscopy

NMR spectroscopy is a powerful tool for biologists interested in the structure, dynamics, and interactions of biological macromolecules. This review aims at presenting in an accessible manner the requirements and limitations of this technique. As an introduction, the history of NMR will highlight how the method evolved from physics to chemistry and finally to biology over several decades. We then introduce the NMR spectral parameters used in structural biology, namely the chemical shift, the J-coupling, nuclear Overhauser effects, and residual dipolar couplings. Resonance assignment, the required step for any further NMR study, bears a resemblance to jigsaw puzzle strategy. The NMR spectral parameters are then converted into angle and distances and used as input using restrained molecular dynamics to compute a bundle of structures. When interpreting a NMR-derived structure, the biologist has to judge its quality on the basis of the statistics provided. When the 3D structure is a priori known by other means, the molecular interaction with a partner can be mapped by NMR: information on the binding interface as well as on kinetic and thermodynamic constants can be gathered. NMR is suitable to monitor, over a wide range of frequencies, protein fluctuations that play a crucial role in their biological function. In the last section of this review, intrinsically disordered proteins, which have escaped the attention of classical structural biology, are discussed in the perspective of NMR, one of the rare available techniques able to describe structural ensembles. This Tutorial is part of the International Proteomics Tutorial Programme (IPTP 16 MCP).

Exploiting conformational ensembles in modeling protein-protein interactions on the proteome scale

Cellular functions are performed through protein-protein interactions; therefore, identification of these interactions is crucial for understanding biological processes. Recent studies suggest that knowledge-based approaches are more useful than "blind" docking for modeling at large scales. However, a caveat of knowledge-based approaches is that they treat molecules as rigid structures. The Protein Data Bank (PDB) offers a wealth of conformations. Here, we exploited an ensemble of the conformations in predictions by a knowledge-based method, PRISM. We tested "difficult" cases in a docking-benchmark data set, where the unbound and bound protein forms are structurally different. Considering alternative conformations for each protein, the percentage of successfully predicted interactions increased from ~26 to 66%, and 57% of the interactions were successfully predicted in an "unbiased" scenario, in which data related to the bound forms were not utilized. If the appropriate conformation, or relevant template interface, is unavailable in the PDB, PRISM could not predict the interaction successfully. The pace of the growth of the PDB promises a rapid increase of ensemble conformations emphasizing the merit of such knowledge-based ensemble strategies for higher success rates in protein-protein interaction predictions on an interactome scale. We constructed the structural network of ERK interacting proteins as a case study.

Constructing structural networks of signaling pathways on the proteome scale

Proteins function through their interactions, and the availability of protein interaction networks could help in understanding cellular processes. However, the known structural data are limited and the classical network node-and-edge representation, where proteins are nodes and interactions are edges, shows only which proteins interact; not how they interact. Structural networks provide this information. Protein-protein interface structures can also indicate which binding partners can interact simultaneously and which are competitive, and can help forecasting potentially harmful drug side effects. Here, we use a powerful protein-protein interactions prediction tool which is able to carry out accurate predictions on the proteome scale to construct the structural network of the extracellular signal-regulated kinases (ERK) in the mitogen-activated protein kinase (MAPK) signaling pathway. This knowledge-based method, PRISM, is motif-based, and is combined with flexible refinement and energy scoring. PRISM predicts protein interactions based on structural and evolutionary similarity to known protein interfaces.