Molecular recognition, which is the process of biological macromolecules interacting with each other or various small molecules with a high specificity and affinity to form a specific complex, constitutes the basis of all processes in living organisms. Proteins, an important class of biological macromolecules, realize their functions through binding to themselves or other molecules. A detailed understanding of the protein–ligand interactions is therefore central to understanding biology at the molecular level. Moreover, knowledge of the mechanisms responsible for the protein-ligand recognition and binding will also facilitate the discovery, design, and development of drugs. In the present review, first, the physicochemical mechanisms underlying protein–ligand binding, including the binding kinetics, thermodynamic concepts and relationships, and binding driving forces, are introduced and rationalized. Next, three currently existing protein-ligand binding models—the “lock-and-key”, “induced fit”, and “conformational selection”—are described and their underlying thermodynamic mechanisms are discussed. Finally, the methods available for investigating protein–ligand binding affinity, including experimental and theoretical/computational approaches, are introduced, and their advantages, disadvantages, and challenges are discussed.

Modeling antigenic variation in influenza (flu) virus A H3N2 using amino acid sequences is a promising approach for improving the prediction accuracy of immune efficacy of vaccines and increasing the efficiency of vaccine screening. Antigenic drift and antigenic jump/shift, which arise from the accumulation of mutations with small or moderate effects and from a major, abrupt change with large effects on the surface antigen hemagglutinin (HA), respectively, are two types of antigenic variation that facilitate immune evasion of flu virus A and make it challenging to predict the antigenic properties of new viral strains. Despite considerable progress in modeling antigenic variation based on the amino acid sequences, few studies focus on the deep learning framework which could be most suitable to be applied to this task. Here, we propose a novel deep learning approach that incorporates a convolutional neural network (CNN) and bidirectional long-short-term memory (BLSTM) neural network to predict antigenic variation. In this approach, CNN extracts the complex local contexts of amino acids while the BLSTM neural network captures the long-distance sequence information. When compared to the existing methods, our deep learning approach achieves the overall highest prediction performance on the validation dataset, and more encouragingly, it achieves prediction agreements of 99.20% and 96.46% for the strains in the forthcoming year and in the next two years included in an existing set of chronological amino acid sequences, respectively. These results indicate that our deep learning approach is promising to be applied to antigenic variation prediction of flu virus A H3N2.

Cold-adapted enzymes feature a lower thermostability and higher catalytic activity compared to their warm-active homologues, which are considered as a consequence of increased flexibility of their molecular structures. The complexity of the (thermo)stability-flexibility-activity relationship makes it difficult to define the strategies and formulate a general theory for enzyme cold adaptation. Here, the psychrophilic serine hydroxymethyltransferase (pSHMT) from Psychromonas ingrahamii and its mesophilic counterpart, mSHMT from Escherichia coli, were subjected to μs-scale multiple-replica molecular dynamics (MD) simulations to explore the cold-adaptation mechanism of the dimeric SHMT. The comparative analyses of MD trajectories reveal that pSHMT exhibits larger structural fluctuations and inter-monomer positional movements, a higher global flexibility, and considerably enhanced local flexibility involving the surface loops and active sites. The largest-amplitude motion mode of pSHMT describes the trends of inter-monomer dissociation and enlargement of the active-site cavity, whereas that of mSHMT characterizes the opposite trends. Based on the comparison of the calculated structural parameters and constructed free energy landscapes (FELs) between the two enzymes, we discuss in-depth the physicochemical principles underlying the stability-flexibility-activity relationships and conclude that (i) pSHMT adopts the global-flexibility mechanism to adapt to the cold environment and, (ii) optimizing the protein-solvent interactions and loosening the inter-monomer association are the main strategies for pSHMT to enhance its flexibility.

The receptor-binding domain (RBD) of the SARS-CoV-2 spike protein (RBD_CoV2) has a higher binding affinity to the human receptor angiotensin-converting enzyme 2 (ACE2) than the SARS-CoV RBD (RBD_CoV). Here, we performed molecular dynamics (MD) simulations, binding free energy (BFE) calculations, and interface residue contact network (IRCN) analysis to explore the mechanistic origin of different ACE2-binding affinities of the two RBDs. The results demonstrate that, when compared to the RBD_CoV2-ACE2 complex, RBD_CoV-ACE2 features enhanced dynamicsand inter-protein positional movements and increased conformational entropy and conformational diversity. Although the inter-protein electrostatic attractive interactions are the primary determinant for the high ACE2-binding affinities of both RBDs, the significantly enhanced electrostatic attractive interactions between ACE2 and RBD_CoV2 determine the higher ACE2-binding affinity of RBD_CoV2 than of RBD_CoV. Comprehensive comparative analyses of the residue BFE components and IRCNs between the two complexes reveal that it is the residue changes at the RBD interface that lead to the overall stronger inter-protein electrostatic attractive force in RBD_CoV2-ACE2, which not only tightens the interface packing and suppresses the dynamics of RBD_CoV2-ACE2, but also enhances the ACE2-binding affinity of RBD_CoV2. Since the RBD residue changes involving gain/loss of the positively/negatively charged residues can greatly enhance the binding affinity, special attention should be paid to the SARS-CoV-2 variants carrying such mutations, particularly those near or at the binding interfaces with the potential to form hydrogen bonds and/or salt bridges with ACE2.

Owing to the rapid changes in the antigenicity of influenza viruses, it is difficult for humans to obtain lasting immunity through antiviral therapy. Hence, tracking the dynamic changes in the antigenicity of influenza viruses can provide a basis for vaccines and drug treatments to cope with the spread of influenza viruses. In this paper, we developed a novel quantitative prediction method to predict the antigenic distance between virus strains using attribute network embedding techniques. An antigenic network is built to model and combine the genetic and antigenic characteristics of the influenza A virus H3N2, using the continuous distributed representation of the virus strain protein sequence (ProtVec) as a node attribute and the antigenic distance between virus strains as an edge weight. The results show a strong positive correlation between supplementing genetic features and antigenic distance prediction accuracy. Further analysis indicates that our prediction model can comprehensively and accurately track the differences in antigenic distances between vaccines and influenza virus strains, and it outperforms existing methods in predicting antigenic distances between strains.