Computational Approaches to Predict Protein-Protein Interactions in Crowded Cellular Environments.

Investigating protein-protein interactions is crucial for understanding cellular biological processes because proteins often function within molecular complexes rather than in isolation. While experimental and computational methods have provided valuable insights into these interactions, they often overlook a critical factor: the crowded cellular environment. This environment significantly impacts protein behavior, including structural stability, diffusion, and ultimately the nature of binding. In this review, we discuss theoretical and computational approaches that allow the modeling of biological systems to guide and complement experiments and can thus significantly advance the investigation, and possibly the predictions, of protein-protein interactions in the crowded environment of cell cytoplasm. We explore topics such as statistical mechanics for lattice simulations, hydrodynamic interactions, diffusion processes in high-viscosity environments, and several methods based on molecular dynamics simulations. By synergistically leveraging methods from biophysics and computational biology, we review the state of the art of computational methods to study the impact of molecular crowding on protein-protein interactions and discuss its potential revolutionizing effects on the characterization of the human interactome.

Computational evidences of a misfolding event in an aggregation-prone light chain preceding the formation of the non-native pathogenic dimer.

Antibody light chain amyloidosis is a disorder in which protein aggregates, mainly composed of immunoglobulin light chains, deposit in diverse tissues impairing the correct functioning of organs. Interestingly, due to the high susceptibility of antibodies to mutations, AL amyloidosis appears to be strongly patient-specific. Indeed, every patient will display their own mutations that will make the proteins involved prone to aggregation thus hindering the study of this disease on a wide scale. In this framework, determining the molecular mechanisms that drive the aggregation could pave the way to the development of patient-specific therapeutics. Here, we focus on a particular patient-derived light chain, which has been experimentally characterized. We investigated the early phases of the aggregation pathway through extensive full-atom molecular dynamics simulations, highlighting a structural rearrangement and the exposure of two hydrophobic regions in the aggregation-prone species. Next, we moved to consider the pathological dimerization process through docking and molecular dynamics simulations, proposing a dimeric structure as a candidate pathological first assembly. Overall, our results shed light on the first phases of the aggregation pathway for a light chain at an atomic level detail, offering new structural insights into the corresponding aggregation process.

Role of the membrane anchor in the regulation of Lck activity.

Theoretical work suggests that collective spatiotemporal behavior of integral membrane proteins should be modulated by boundary lipids sheathing their membrane anchors. Here, we show evidence for this prediction while investigating the mechanism for maintaining a steady amount of the active form of integral membrane protein Lck kinase (LckA) by Lck trans-autophosphorylation regulated by the phosphatase CD45. We used super-resolution microscopy, flow cytometry, and pharmacological and genetic perturbation to gain insight into the spatiotemporal context of this process. We found that LckA is generated exclusively at the plasma membrane, where CD45 maintains it in a ceaseless dynamic equilibrium with its unphosphorylated precursor. Steady LckA shows linear dependence, after an initial threshold, over a considerable range of Lck expression levels. This behavior fits a phenomenological model of trans-autophosphorylation that becomes more efficient with increasing LckA. We then challenged steady LckA formation by genetically swapping the Lck membrane anchor with structurally divergent ones, such as that of Src or the transmembrane domains of LAT, CD4, palmitoylation-defective CD4 and CD45 that were expected to drastically modify Lck boundary lipids. We observed small but significant changes in LckA generation, except for the CD45 transmembrane domain that drastically reduced LckA due to its excessive lateral proximity to CD45. Comprehensively, LckA formation and maintenance can be best explained by lipid bilayer critical density fluctuations rather than liquid-ordered phase-separated nanodomains, as previously thought, with "like/unlike" boundary lipids driving dynamical proximity and remoteness of Lck with itself and with CD45.

Dynamical changes of SARS-CoV-2 spike variants in the highly immunogenic regions impact the viral antibodies escaping.

The prolonged circulation of the SARS-CoV-2 virus resulted in the emergence of several viral variants, with different spreading features. Moreover, the increased number of recovered and/or vaccinated people introduced a selective pressure toward variants able to evade the immune system, developed against the former viral versions. This process results in reinfections. Aiming to study the latter process, we first collected a large structural dataset of antibodies in complex with the original version of SARS-CoV-2 Spike protein. We characterized the peculiarities of such antibodies population with respect to a control dataset of antibody-protein complexes, highlighting some statistically significant differences between these two sets of antibodies. Thus, moving our attention to the Spike side of the complexes, we identify the Spike region most prone to interaction with antibodies, describing in detail also the energetic mechanisms used by antibodies to recognize different epitopes. In this framework, fast protocols able to assess the effect of novel mutations on the cohort of developed antibodies would help establish the impact of the variants on the population. Performing a molecular dynamics simulation of the trimeric form of the SARS-CoV-2 Spike protein for the wild type and two variants of concern, that is, the Delta and Omicron variants, we described the physicochemical features and the conformational changes experienced locally by the variants with respect to the original version. Hence, combining the dynamical information with the structural study on the antibody-spike dataset, we quantitatively explain why the Omicron variant has a higher capability of escaping the immune system than the Delta variant, due to the higher conformational variability of the most immunogenic regions. Overall, our results shed light on the molecular mechanism behind the different responses the SARS-CoV-2 variants display against the immune response induced by either vaccines or previous infections. Moreover, our analysis proposes an approach that can be easily extended to both other SARS-CoV-2 variants or different molecular systems.

Thermometer: a webserver to predict protein thermal stability.

MOTIVATION: Thermal properties of proteins are of great importance for a number of theoretical and practical implications. Predicting the thermal stability of a protein is a difficult and still scarcely addressed task. RESULTS: Here, we introduce Thermometer, a webserver to assess the thermal stability of a protein using structural information. Thermometer is implemented as a publicly available, user-friendly interface. AVAILABILITY AND IMPLEMENTATION: Our server can be found at the following link (all major browser supported): http://service.tartaglialab.com/new_submission/thermometer_file. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Binding site identification of G protein-coupled receptors through a 3D Zernike polynomials-based method: application to C. elegans olfactory receptors.

Studying the binding processes of G protein-coupled receptors (GPCRs) proteins is of particular interest both to better understand the molecular mechanisms that regulate the signaling between the extracellular and intracellular environment and for drug design purposes. In this study, we propose a new computational approach for the identification of the binding site for a specific ligand on a GPCR. The method is based on the Zernike polynomials and performs the ligand-GPCR association through a shape complementarity analysis of the local molecular surfaces. The method is parameter-free and it can distinguish, working on hundreds of experimentally GPCR-ligand complexes, binding pockets from randomly sampled regions on the receptor surface, obtaining an Area Under ROC curve of 0.77. Given its importance both as a model organism and in terms of applications, we thus investigated the olfactory receptors of the C. elegans, building a list of associations between 21 GPCRs belonging to its olfactory neurons and a set of possible ligands. Thus, we can not only carry out rapid and efficient screenings of drugs proposed for GPCRs, key targets in many pathologies, but also we laid the groundwork for computational mutagenesis processes, aimed at increasing or decreasing the binding affinity between ligands and receptors.

Structural analysis of SARS-CoV-2 genome and predictions of the human interactome.

Specific elements of viral genomes regulate interactions within host cells. Here, we calculated the secondary structure content of >2000 coronaviruses and computed >100 000 human protein interactions with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The genomic regions display different degrees of conservation. SARS-CoV-2 domain encompassing nucleotides 22 500-23 000 is conserved both at the sequence and structural level. The regions upstream and downstream, however, vary significantly. This part of the viral sequence codes for the Spike S protein that interacts with the human receptor angiotensin-converting enzyme 2 (ACE2). Thus, variability of Spike S is connected to different levels of viral entry in human cells within the population. Our predictions indicate that the 5' end of SARS-CoV-2 is highly structured and interacts with several human proteins. The binding proteins are involved in viral RNA processing, include double-stranded RNA specific editases and ATP-dependent RNA-helicases and have strong propensity to form stress granules and phase-separated assemblies. We propose that these proteins, also implicated in viral infections such as HIV, are selectively recruited by SARS-CoV-2 genome to alter transcriptional and post-transcriptional regulation of host cells and to promote viral replication.

Lactoferrin Inhibition of the Complex Formation between ACE2 Receptor and SARS CoV-2 Recognition Binding Domain.

The present investigation focuses on the analysis of the interactions among human lactoferrin (LF), SARS-CoV-2 receptor-binding domain (RBD) and human angiotensin-converting enzyme 2 (ACE2) receptor in order to assess possible mutual interactions that could provide a molecular basis of the reported preventative effect of lactoferrin against CoV-2 infection. In particular, kinetic and thermodynamic parameters for the pairwise interactions among the three proteins were measured via two independent techniques, biolayer interferometry and latex nanoparticle-enhanced turbidimetry. The results obtained clearly indicate that LF is able to bind the ACE2 receptor ectodomain with significantly high affinity, whereas no binding to the RBD was observed up to the maximum "physiological" lactoferrin concentration range. Lactoferrin, above 1 µM concentration, thus appears to directly interfere with RBD-ACE2 binding, bringing about a measurable, up to 300-fold increase of the KD value relative to RBD-ACE2 complex formation.

Prediction of Time Series Gene Expression and Structural Analysis of Gene Regulatory Networks Using Recurrent Neural Networks.

Methods for time series prediction and classification of gene regulatory networks (GRNs) from gene expression data have been treated separately so far. The recent emergence of attention-based recurrent neural network (RNN) models boosted the interpretability of RNN parameters, making them appealing for the understanding of gene interactions. In this work, we generated synthetic time series gene expression data from a range of archetypal GRNs and we relied on a dual attention RNN to predict the gene temporal dynamics. We show that the prediction is extremely accurate for GRNs with different architectures. Next, we focused on the attention mechanism of the RNN and, using tools from graph theory, we found that its graph properties allow one to hierarchically distinguish different architectures of the GRN. We show that the GRN responded differently to the addition of noise in the prediction by the RNN and we related the noise response to the analysis of the attention mechanism. In conclusion, this work provides a way to understand and exploit the attention mechanism of RNNs and it paves the way to RNN-based methods for time series prediction and inference of GRNs from gene expression data.

Assessing the accuracy of contact and distance predictions in CASP14.

We present the results of the assessment of the intramolecular residue-residue contact and distance predictions from groups participating in the 14th round of the CASP experiment. The performance of contact prediction methods was evaluated with the measures used in previous CASPs, while distance predictions were assessed based on a new protocol, which considers individual distance pairs as well as the whole predicted distance matrix, using a graph-based framework. The results of the evaluation indicate that predictions by the tFold framework, TripletRes and DeepPotential were the most accurate in both categories. With regards to progress in method performance, the results of the assessment in contact prediction did not reveal any discernible difference when compared to CASP13. Arguably, this could be due to CASP14 FM targets being more challenging than ever before.

proABC-2: PRediction of AntiBody contacts v2 and its application to information-driven docking.

MOTIVATION: Monoclonal antibodies are essential tools in the contemporary therapeutic armory. Understanding how these recognize their antigen is a fundamental step in their rational design and engineering. The rising amount of publicly available data is catalyzing the development of computational approaches able to offer valuable, faster and cheaper alternatives to classical experimental methodologies used for the study of antibody-antigen complexes. RESULTS: Here, we present proABC-2, an update of the original random-forest antibody paratope predictor, based on a convolutional neural network algorithm. We also demonstrate how the predictions can be fruitfully used to drive the docking in HADDOCK. AVAILABILITY AND IMPLEMENTATION: The proABC-2 server is freely available at: https://wenmr.science.uu.nl/proabc2/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Insights on protein thermal stability: a graph representation of molecular interactions.

MOTIVATION: Understanding the molecular mechanisms of thermal stability is a challenge in protein biology. Indeed, knowing the temperature at which proteins are stable has important theoretical implications, which are intimately linked with properties of the native fold, and a wide range of potential applications from drug design to the optimization of enzyme activity. RESULTS: Here, we present a novel graph-theoretical framework to assess thermal stability based on the structure without any a priori information. In this approach we describe proteins as energy-weighted graphs and compare them using ensembles of interaction networks. Investigating the position of specific interactions within the 3D native structure, we developed a parameter-free network descriptor that permits to distinguish thermostable and mesostable proteins with an accuracy of 76% and area under the receiver operating characteristic curve of 78%. AVAILABILITY AND IMPLEMENTATION: Code is available upon request to edoardo.milanetti@uniroma1.it. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Quantitative Characterization of Binding Pockets and Binding Complementarity by Means of Zernike Descriptors.

In this work, we describe the application of the Zernike formalism to quantitatively characterize the binding pockets of two sets of biologically relevant systems. Such an approach, when applied to molecular dynamics trajectories, is able to pinpoint the subtle differences between very similar molecular regions and their impact on the local propensity to ligand binding, allowing us to quantify such differences. The statistical robustness of our procedure suggests that it is very suitable to describe protein binding sites and protein-ligand interactions within a rigorous and well-defined framework.

Simulated Epidemics in 3D Protein Structures to Detect Functional Properties.

The outcome of an epidemic is closely related to the network of interactions between individuals. Likewise, protein functions depend on the 3D arrangement of their residues and the underlying energetic interaction network. Borrowing ideas from the theoretical framework that has been developed to address the spreading of real diseases, we study for the first time the diffusion of a fictitious epidemic inside the protein nonbonded interaction network, aiming to study network features and properties. Our approach allows us to probe the overall stability and the capability of propagating information in complex 3D structures, proving to be very efficient in addressing different problems, from the assessment of thermal stability to the identification of functional sites.

Counter-regulation of regulatory T cells by autoreactive CD8+ T cells in rheumatoid arthritis.

The mechanisms whereby autoreactive T cells escape peripheral tolerance establishing thus autoimmune diseases in humans remain an unresolved question. Here, we demonstrate that autoreactive polyfunctional CD8+ T cells recognizing self-antigens (i.e., vimentin, actin cytoplasmic 1, or non-muscle myosin heavy chain 9 epitopes) with high avidity, counter-regulate Tregs by killing them, in a consistent percentage of rheumatoid arthritis (RA) patients. Indeed, these CD8+ T cells express a phenotype and gene profile of effector (eff) cells and, upon antigen-specific activation, kill Tregs indirectly in an NKG2D-dependent bystander fashion in vitro. This data provides a mechanistic basis for the finding showing that AE-specific (CD107a+) CD8+ T killer cells correlate, directly with the disease activity score, and inversely with the percentage of activated Tregs, in both steady state and follow-up studies in vivo. In addition, multiplex immunofluorescence imaging analyses of inflamed synovial tissues in vivo show that a remarkable number of CD8+ T cells express granzyme-B and selectively contact FOXP3+ Tregs, some of which are in an apoptotic state, validating hence the possibility that CD8+ Teff cells can counteract neighboring Tregs within inflamed tissues, by killing them. Alternatively, the disease activity score of a different subset of patients is correlated with the expansion of a peculiar subpopulation of autoreactive low avidity, partially-activated (pa)CD8+ T cells that, despite they conserve the conventional naïve (N) phenotype, produce high levels of tumor necrosis factor (TNF)-α and exhibit a gene expression signature of a progressive activation state. Tregs directly correlate with the expansion of this autoreactive (low avidity) paCD8+ TN cell subset in vivo, and efficiently control their differentiation rather their proliferation in vitro. Interestingly, autoreactive high avidity CD8+ Teff cells or low avidity paCD8+ TN cells are significantly expanded in RA patients who would become non-responders or patients who would become responders to TNF-α inhibitor therapy, respectively. These data provide evidence of a previously undescribed role of such mechanisms in the progression and therapy of RA.

On the activation and deactivation pathways of the Lck kinase domain: a computational study.

Here we report the description of the conformational pathways connecting the Lck active and inactive states by means of all-atoms molecular dynamics simulations coupled to an enhancing sampling methodology. By such an approach, we describe the major structural determinants characterizing these large conformational transitions and compare such pathways to those obtained for a similar kinase, i.e. c-Src. Our results show that both the activation and deactivation processes could follow distinct pathways, differentiated by the order by which the A-loop and the C-helix regions rearrange.

Prediction of the permeability of neutral drugs inferred from their solvation properties.

MOTIVATION: Determination of drug absorption is an important component of the drug discovery and development process in that it plays a key role in the decision to promote drug candidates to clinical trials. We have developed a method that, on the basis of an analysis of the dynamic distribution of water molecules around a compound obtained by molecular dynamics simulations, can compute a parameter-free value that correlates very well with the compound permeability measured using the human colon adenocarcinoma (Caco-2) cell line assay. RESULTS: The method has been tested on twenty-three neutral drugs for which a consistent set of experimental data is available. We show here that our method reproduces the experimental data better than other existing tools. Furthermore it provides a detailed view of the relationship between the hydration and the permeability properties of molecules. CONTACT: anna.tramontano@uniroma1.it SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Lactoferrins in Their Interactions with Molecular Targets: A Structure-Based Overview.

Lactoferrins and lactoferrin-derived peptides display numerous functions linked to innate immunity in mammalians, spanning from antimicrobial to anti-inflammatory and immunomodulatory actions, and even demonstrate antitumor properties. To date, the proposed mechanisms for their biological actions are varied, although the molecular basis that governs lactoferrin interactions with molecular targets has been clarified only in a limited number of specific cases. However, key in silico methods have recently moved the topic to the fore, thus greatly expanding the possibilities of large-scale investigations on macromolecular interactions involving lactoferrins and their molecular targets. This review aims to summarize the current knowledge on the structural determinants that drive lactoferrin recognition of molecular targets, with primary focus on the mechanisms of activity against bacteria and viruses. The understanding of the structural details of lactoferrins' interaction with their molecular partners is in fact a crucial goal for the development of novel pharmaceutical products.

High-Throughput Interactome Determination via Sulfur Anomalous Scattering.

We propose a novel approach for detecting the binding between proteins making use of the anomalous diffraction of natively present heavy elements, e.g., sulfurs, inside molecular three-dimensional structures. In particular, we analytically and numerically show that the diffraction patterns produced by the anomalous scattering of the sulfur atoms in a given direction depend additively on the relative distances between all couples of sulfur atoms. Thus, the differences in the patterns produced by bound proteins with respect to their nonbonded states can be exploited to rapidly assess protein complex formation. On the basis of our results, we suggest a possible experimental procedure for detecting protein-protein binding. Overall, the completely label-free and rapid method we propose may be readily extended to probe interactions on a large scale, thus paving the way for the development of a novel field of research based on a synchrotron light source.

A recurrent Hopfield network for estimating meso-scale effective connectivity in MEG.

The architecture of communication within the brain, represented by the human connectome, has gained a paramount role in the neuroscience community. Several features of this communication, e.g., the frequency content, spatial topology, and temporal dynamics are currently well established. However, identifying generative models providing the underlying patterns of inhibition/excitation is very challenging. To address this issue, we present a novel generative model to estimate large-scale effective connectivity from MEG. The dynamic evolution of this model is determined by a recurrent Hopfield neural network with asymmetric connections, and thus denoted Recurrent Hopfield Mass Model (RHoMM). Since RHoMM must be applied to binary neurons, it is suitable for analyzing Band Limited Power (BLP) dynamics following a binarization process. We trained RHoMM to predict the MEG dynamics through a gradient descent minimization and we validated it in two steps. First, we showed a significant agreement between the similarity of the effective connectivity patterns and that of the interregional BLP correlation, demonstrating RHoMM's ability to capture individual variability of BLP dynamics. Second, we showed that the simulated BLP correlation connectomes, obtained from RHoMM evolutions of BLP, preserved some important topological features, e.g, the centrality of the real data, assuring the reliability of RHoMM. Compared to other biophysical models, RHoMM is based on recurrent Hopfield neural networks, thus, it has the advantage of being data-driven, less demanding in terms of hyperparameters and scalable to encompass large-scale system interactions. These features are promising for investigating the dynamics of inhibition/excitation at different spatial scales.