Reaching for high-hanging fruit in drug discovery at protein-protein interfaces

PepExplainer: An explainable deep learning model for selection-based macrocyclic peptide bioactivity prediction and optimization

Macrocyclic peptides possess unique features, making them highly promising as a drug modality. However, evaluating their bioactivity through wet lab experiments is generally resource-intensive and time-consuming. Despite advancements in artificial intelligence (AI) for bioactivity prediction, challenges remain due to limited data availability and the interpretability issues in deep learning models, often leading to less-than-ideal predictions. To address these challenges, we developed PepExplainer, an explainable graph neural network based on substructure mask explanation (SME). This model excels at deciphering amino acid substructures, translating macrocyclic peptides into detailed molecular graphs at the atomic level, and efficiently handling non-canonical amino acids and complex macrocyclic peptide structures. PepExplainer's effectiveness is enhanced by utilizing the correlation between peptide enrichment data from selection-based focused library and bioactivity data, and employing transfer learning to improve bioactivity predictions of macrocyclic peptides against IL-17C/IL-17 RE interaction. Additionally, PepExplainer underwent further validation for bioactivity prediction using an additional set of thirteen newly synthesized macrocyclic peptides. Moreover, it enabled the optimization of the IC50 of a macrocyclic peptide, reducing it from 15 nM to 5.6 nM based on the contribution score provided by PepExplainer. This achievement underscores PepExplainer's skill in deciphering complex molecular patterns, highlighting its potential to accelerate the discovery and optimization of macrocyclic peptides.

Development of covalent chemogenetic K2P channel activators

K2P potassium channels regulate excitability by affecting cellular resting membrane potential in the brain, cardiovascular system, immune cells, and sensory organs. Despite their important roles in anesthesia, arrhythmia, pain, hypertension, sleep, and migraine, the ability to control K2P function remains limited. Here, we describe a chemogenetic strategy termed CATKLAMP (covalent activation of TREK family K+ channels to clamp membrane potential) that leverages the discovery of a K2P modulator pocket site that reacts with electrophile-bearing derivatives of a TREK subfamily small-molecule activator, ML335, to activate the channel irreversibly. We show that CATKLAMP can be used to probe fundamental aspects of K2P function, as a switch to silence neuronal firing, and is applicable to all TREK subfamily members. Together, our findings exemplify a means to alter K2P channel activity that should facilitate molecular and systems level studies of K2P function and enable the search for new K2P modulators.

14-3-3 Protein-Protein Interactions: From Mechanistic Understanding to Their Small-Molecule Stabilization

Protein-protein interactions (PPIs) are of utmost importance for maintenance of cellular homeostasis. Herein, a central role can be found for 14-3-3 proteins. These hub-proteins are known to bind hundreds of interaction partners, thereby regulating their activity, localization, and/or stabilization. Due to their ability to bind a large variety of client proteins, studies of 14-3-3 protein complexes flourished over the last decades, aiming to gain greater molecular understanding of these complexes and their role in health and disease. Because of their crucial role within the cell, 14-3-3 protein complexes are recognized as highly interesting therapeutic targets, encouraging the discovery of small molecule modulators of these PPIs. We discuss various examples of 14-3-3-mediated regulation of its binding partners on a mechanistic level, highlighting the versatile and multi-functional role of 14-3-3 within the cell. Furthermore, an overview is given on the development of stabilizers of 14-3-3 protein complexes, from initially used natural products to fragment-based approaches. These studies show the potential of 14-3-3 PPI stabilizers as novel agents in drug discovery and as tool compounds to gain greater molecular understanding of the role of 14-3-3-based protein regulation.

Structure-Based Design and Discovery of a Potent and Cell-Active LC3A/B Covalent Inhibitor

Autophagy is a highly conserved cellular homeostasis maintenance mechanism in eukaryotes. Microtubule-associated protein light chain 3 (LC3) plays a crucial role in autophagy. It has multiple pairs of protein-protein interactions (PPIs) with other proteins, and these PPIs have an effect on the regulation of autophagosome formation and the recruitment of autophagic substrates. In our previous work, a small molecule covalent inhibitor DC-LC3in-D5 which could inhibit LC3A/B PPIs was identified, but a detailed study of structure-activity relationships (SARs) was lacking. Herein, a new molecule LC3in-C42 was discovered utilizing the hybridization of advantageous fragments, whose potency (IC50 = 7.6 nM) had been greatly improved compared with that of DC-LC3in-D5. LC3in-C42 inhibits autophagy at the cellular level and its efficacy far exceeds that of DC-LC3in-D5. Thus far, LC3in-C42 stands as the most potent LC3A/B small molecule inhibitor. LC3in-C42 could serve as a powerful tool for LC3A/B protein and autophagy research.

Molecular Modeling of Single- and Double-Hydrocarbon-Stapled Coiled-Coil Inhibitors against Bcr-Abl: Toward a Treatment Strategy for CML

The chimeric oncoprotein Bcr-Abl is the causative agent of virtually all chronic myeloid leukemias and a subset of acute lymphoblastic leukemias. As a result of the so-called Philadelphia chromosome translocation t(9;22), Bcr-Abl manifests as a constitutively active tyrosine kinase, which promotes leukemogenesis by activation of cell cycle signaling pathways. Constitutive and oncogenic activation is mediated by an N-terminal coiled-coil oligomerization domain in Bcr (Bcr-CC), presenting a therapeutic target for inhibition of Bcr-Abl activity toward the treatment of Bcr-Abl+ leukemias. Previously, we demonstrated that a rationally designed Bcr-CC mutant, CCmut3, exerts a dominant negative effect upon Bcr-Abl activity by preferential oligomerization with Bcr-CC. Moreover, we have shown that conjugation to a leukemia-specific cell-penetrating peptide (CPP-CCmut3) improves intracellular delivery and activity. However, our full-length CPP-CCmut3 construct (81 aa) is encumbered by an intrinsically high degree of conformational variability and susceptibility to proteolytic degradation relative to traditional small-molecule therapeutics. Here, we iterate a new generation of CCmut3 inhibitors against Bcr-CC-mediated Bcr-Abl assembly designed to address these constraints through incorporation of all-hydrocarbon staples spanning i and i + 7 positions in α-helix 2 (CPP-CCmut3-st). We utilize computational modeling and biomolecular simulation to evaluate single- and double-stapled CCmut3 candidates in silico for dynamics and binding energetics. We further model a truncated system characterized by the deletion of α-helix 1 and the flexible loop linker, which are known to impart high conformational variability. To study the impact of the N-terminal cyclic CPP toward model stability and inhibitor activity, we also model the full-length and truncated systems devoid of the CPP, with a cyclized CPP, and with an open-configuration CPP, for a total of six systems that comprise our library. From this library, we present lead-stapled peptide candidates to be synthesized and evaluated experimentally as our next iteration of inhibitors against Bcr-Abl.

Structure-activity relationships of middle-size cyclic peptides, KRAS inhibitors derived from an mRNA display

Cyclic peptides are attracting attention as therapeutic agents due to their potential for oral absorption and easy access to tough intracellular targets. LUNA18, a clinical KRAS inhibitor, was transformed-without scaffold hopping-from the initial hit by using an mRNA display library that met our criteria for drug-likeness. In drug discovery using mRNA display libraries, hit compounds always possess a site linked to an mRNA tag. Here, we describe our examination of the Structure-Activity Relationship (SAR) using X-ray structures for chemical optimization near the site linked to the mRNA tag, equivalent to the C-terminus. Structural modifications near the C-terminus demonstrated a relatively wide range of tolerance for side chains. Furthermore, we show that a single atom modification is enough to change the pharmacokinetic (PK) profile. Since there are four positions where side chain modification is permissible in terms of activity, it is possible to flexibly adjust the pharmacokinetic profile by structurally optimizing the side chain. The side chain transformation findings demonstrated here may be generally applicable to hits obtained from mRNA display libraries.

Unraveling the folding-assisted unbinding mechanism of TCF with its binding partner β-catenin

This study utilizes molecular dynamics simulations aided with multiple walker parallel bias metadynamics to investigate the TCF unbinding mechanism from the β-catenin interface. The results, consistent with experimental binding affinity calculations, unveil a folding-assisted unbinding mechanism.

Bidirectional linkage of DNA barcodes for the multiplexed mapping of higher-order protein interactions in cells

Capturing the full complexity of the diverse hierarchical interactions in the protein interactome is challenging. Here we report a DNA-barcoding method for the multiplexed mapping of pairwise and higher-order protein interactions and their dynamics within cells. The method leverages antibodies conjugated with barcoded DNA strands that can bidirectionally hybridize and covalently link to linearize closely spaced interactions within individual 3D protein complexes, encoding and decoding the protein constituents and the interactions among them. By mapping protein interactions in cancer cells and normal cells, we found that tumour cells exhibit a larger diversity and abundance of protein complexes with higher-order interactions. In biopsies of human breast-cancer tissue, the method accurately identified the cancer subtype and revealed that higher-order protein interactions are associated with cancer aggressiveness.

Nonsymmetrically Substituted 1,1'-Biphenyl-Based Small Molecule Inhibitors of the PD-1/PD-L1 Interaction

Therapeutic antibodies directed against either programmed cell death-1 protein (PD-1) or its ligand PD-L1 have demonstrated efficacy in the treatment of various cancers. In contrast with antibodies, small molecules have the potential for increased tissue penetration; better pharmacology; and therefore, improved antitumor activity. A series of nonsymmetric C2 inhibitors were synthesized and evaluated for PD-1/PD-L1 interaction inhibition. These compounds induced PD-L1 dimerization and effectively blocked PD-L1/PD-1 interaction in a homogeneous time-resolved fluorescence (HTRF) assay with most inhibitors exhibiting IC50 values in the single-digit nM range and below. Their high inhibitory potency was also demonstrated in a cell-based coculture PD-1 signaling assay where 2 exhibited an EC50 inhibitory activity of 21.8 nM, which approached that of the PD-L1 antibody durvalumab (EC50 = 0.3-1.8 nM). Structural insight into how these inhibitors interact with PD-L1 was gained by using NMR and X-ray cocrystal structure studies. These data support further preclinical evaluation of these compounds as antibody alternatives.

Identification of Small Molecule Ligand Binding Sites On and In the ARNT PAS-B Domain

Transcription factors are generally challenging to target with small molecule inhibitors due to their structural plasticity and lack of catalytic sites. Notable exceptions include several naturally ligand-regulated transcription factors, including our prior work with the heterodimeric HIF-2 transcription factor which showed that small molecule binding within an internal pocket of the HIF-2α PAS-B domain can disrupt its interactions with its dimerization partner, ARNT. Here, we explore the feasibility of similarly targeting small molecules to the analogous ARNT PAS-B domain itself, potentially opening a promising route to simultaneously modulate several ARNT-mediated signaling pathways. Using solution NMR screening of an in-house fragment library, we previously identified several compounds that bind ARNT PAS-B and, in certain cases, antagonize ARNT association with the TACC3 transcriptional coactivator. However, these ligands have only modest binding affinities, complicating characterization of their binding sites. We address this challenge by combining NMR, MD simulations, and ensemble docking to identify ligand-binding 'hotspots' on and within the ARNT PAS-B domain. Our data indicate that the two ARNT/TACC3 inhibitors, KG-548 and KG-655, bind to a β-sheet surface implicated in both HIF-2 dimerization and coactivator recruitment. Furthermore, while KG-548 binds exclusively to the β-sheet surface, KG-655 can additionally bind within a water-accessible internal cavity in ARNT PAS-B. Finally, KG-279, while not a coactivator inhibitor, exemplifies ligands that preferentially bind only to the internal cavity. All three ligands promoted ARNT PAS-B homodimerization, albeit to varying degrees. Taken together, our findings provide a comprehensive overview of ARNT PAS-B ligand-binding sites and may guide the development of more potent coactivator inhibitors for cellular and functional studies.

CytoSIP: an annotated structural atlas for interactions involving cytokines or cytokine receptors

Therapeutic agents targeting cytokine-cytokine receptor (CK-CKR) interactions lead to the disruption in cellular signaling and are effective in treating many diseases including tumors. However, a lack of universal and quick access to annotated structural surface regions on CK/CKR has limited the progress of a structure-driven approach in developing targeted macromolecular drugs and precision medicine therapeutics. Herein we develop CytoSIP (Single nucleotide polymorphisms (SNPs), Interface, and Phenotype), a rich internet application based on a database of atomic interactions around hotspots in experimentally determined CK/CKR structural complexes. CytoSIP contains: (1) SNPs on CK/CKR; (2) interactions involving CK/CKR domains, including CK/CKR interfaces, oligomeric interfaces, epitopes, or other drug targeting surfaces; and (3) diseases and phenotypes associated with CK/CKR or SNPs. The database framework introduces a unique tri-level SIP data model to bridge genetic variants (atomic level) to disease phenotypes (organism level) using protein structure (complexes) as an underlying framework (molecule level). Customized screening tools are implemented to retrieve relevant CK/CKR subset, which reduces the time and resources needed to interrogate large datasets involving CK/CKR surface hotspots and associated pathologies. CytoSIP portal is publicly accessible at https://CytoSIP.biocloud.top , facilitating the panoramic investigation of the context-dependent crosstalk between CK/CKR and the development of targeted therapeutic agents.

Integration of protein sequence and protein-protein interaction data by hypergraph learning to identify novel protein complexes

Protein-protein interactions (PPIs) are the basis of many important biological processes, with protein complexes being the key forms implementing these interactions. Understanding protein complexes and their functions is critical for elucidating mechanisms of life processes, disease diagnosis and treatment and drug development. However, experimental methods for identifying protein complexes have many limitations. Therefore, it is necessary to use computational methods to predict protein complexes. Protein sequences can indicate the structure and biological functions of proteins, while also determining their binding abilities with other proteins, influencing the formation of protein complexes. Integrating these characteristics to predict protein complexes is very promising, but currently there is no effective framework that can utilize both protein sequence and PPI network topology for complex prediction. To address this challenge, we have developed HyperGraphComplex, a method based on hypergraph variational autoencoder that can capture expressive features from protein sequences without feature engineering, while also considering topological properties in PPI networks, to predict protein complexes. Experiment results demonstrated that HyperGraphComplex achieves satisfactory predictive performance when compared with state-of-art methods. Further bioinformatics analysis shows that the predicted protein complexes have similar attributes to known ones. Moreover, case studies corroborated the remarkable predictive capability of our model in identifying protein complexes, including 3 that were not only experimentally validated by recent studies but also exhibited high-confidence structural predictions from AlphaFold-Multimer. We believe that the HyperGraphComplex algorithm and our provided proteome-wide high-confidence protein complex prediction dataset will help elucidate how proteins regulate cellular processes in the form of complexes, and facilitate disease diagnosis and treatment and drug development. Source codes are available at https://github.com/LiDlab/HyperGraphComplex.

α-Hydrazino Acid Insertion Governs Peptide Organization in Solution by Local Structure Ordering

In this work, we have applied the concept of α-hydrazino acid insertion in a peptide sequence as a means of structurally organizing a potential protein-protein interactions (PPI) inhibitor. Hydrazino peptides characterized by the incorporation of an α-hydrazino acid at specific positions introduce an additional nitrogen atom into their backbone. This modification leads to a change in the electrostatic properties of the peptide and induces the restructuring of its hydrogen bonding network, resulting in conformational changes toward more stable structural motifs. Despite the successful use of synthetic hydrazino oligomers in binding to nucleic acids, the structural changes due to the incorporation of α-hydrazino acid into short natural peptides in solution are still poorly understood. Based on NMR data, we report structural models of p53-derived hydrazino peptides with elements of localized peptide structuring in the form of an α-, β-, or γ-turn as a result of hydrazino modification in the peptide backbone. The modifications could potentially lead to the preorganization of a helical secondary peptide structure in a solution that is favorable for binding to a biological receptor. Spectroscopically, we observed that the ensemble averaged rapidly interconverting conformations, including isomerization of the E-Z hydrazide bond. This further increases the adaptability by expanding the conformational space of hydrazine peptides as potential protein-protein interaction antagonists.

Multi-level bioinformatics resources support drug target discovery of protein-protein interactions

Drug discovery often begins with a new target. Protein-protein interactions (PPIs) are crucial to multitudinous cellular processes and offer a promising avenue for drug-target discovery. PPIs are characterized by multi-level complexity: at the protein level, interaction networks can be used to identify potential targets, whereas at the residue level, the details of the interactions of individual PPIs can be used to examine a target's druggability. Much great progress has been made in target discovery through multi-level PPI-related computational approaches, but these resources have not been fully discussed. Here, we systematically survey bioinformatics tools for identifying and assessing potential drug targets, examining their characteristics, limitations and applications. This work will aid the integration of the broader protein-to-network context with the analysis of detailed binding mechanisms to support the discovery of drug targets.

Genome-scale annotation of protein binding sites via language model and geometric deep learning

Revealing protein binding sites with other molecules, such as nucleic acids, peptides, or small ligands, sheds light on disease mechanism elucidation and novel drug design. With the explosive growth of proteins in sequence databases, how to accurately and efficiently identify these binding sites from sequences becomes essential. However, current methods mostly rely on expensive multiple sequence alignments or experimental protein structures, limiting their genome-scale applications. Besides, these methods haven't fully explored the geometry of the protein structures. Here, we propose GPSite, a multi-task network for simultaneously predicting binding residues of DNA, RNA, peptide, protein, ATP, HEM, and metal ions on proteins. GPSite was trained on informative sequence embeddings and predicted structures from protein language models, while comprehensively extracting residual and relational geometric contexts in an end-to-end manner. Experiments demonstrate that GPSite substantially surpasses state-of-the-art sequence-based and structure-based approaches on various benchmark datasets, even when the structures are not well-predicted. The low computational cost of GPSite enables rapid genome-scale binding residue annotations for over 568,000 sequences, providing opportunities to unveil unexplored associations of binding sites with molecular functions, biological processes, and genetic variants. The GPSite webserver and annotation database can be freely accessed at https://bio-web1.nscc-gz.cn/app/GPSite.

Small Molecule Antagonists of the DNA Repair ERCC1/XPA Protein-Protein Interaction

The DNA excision repair protein ERCC1 and the DNA damage sensor protein, XPA are highly overexpressed in patient samples of cisplatin-resistant solid tumors including lung, bladder, ovarian, and testicular cancer. The repair of cisplatin-DNA crosslinks is dependent upon nucleotide excision repair (NER) that is modulated by protein-protein binding interactions of ERCC1, the endonuclease, XPF, and XPA. Thus, inhibition of their function is a potential therapeutic strategy for the selective sensitization of tumors to DNA-damaging platinum-based cancer therapy. Here, we report on new small-molecule antagonists of the ERCC1/XPA protein-protein interaction (PPI) discovered using a high-throughput competitive fluorescence polarization binding assay. We discovered a unique structural class of thiopyridine-3-carbonitrile PPI antagonists that block a truncated XPA polypeptide from binding to ERCC1. Preliminary hit-to-lead studies from compound 1 reveal structure-activity relationships (SAR) and identify lead compound 27 o with an EC50 of 4.7 μM. Furthermore, chemical shift perturbation mapping by NMR confirms that 1 binds within the same site as the truncated XPA67-80 peptide. These novel ERCC1 antagonists are useful chemical biology tools for investigating DNA damage repair pathways and provide a good starting point for medicinal chemistry optimization as therapeutics for sensitizing tumors to DNA damaging agents and overcoming resistance to platinum-based chemotherapy.

Substituted Oligosaccharides as Protein Mimics: Deep Learning Free Energy Landscapes

Protein-protein complexes power the majority of cellular processes. Interfering with the formation of such complexes using well-designed mimics is a difficult, yet actively pursued, research endeavor. Due to the limited availability of results on the conformational preferences of oligosaccharides compared to polypeptides, the former have been much less explored than the latter as protein mimics, despite interesting ADMET characteristics. In this work, the conformational landscapes of a series of 956 substituted glucopyranose oligomers of lengths 3 to 12 designed as protein interface mimics are revealed using microsecond-time-scale, enhanced-sampling molecular dynamics simulations. Deep convolutional networks are trained on these large conformational ensembles, to predict the stability of longer oligosaccharide structures from those of their constituent trimer motifs. Deep generative adversarial networks are then designed to suggest plausible conformations for oligosaccharide mimics of arbitrary length and substituent sequences that can subsequently be used as input to docking simulations. Analyzing the performance of the neural networks also yields insights into the intricate collective effects that dominate oligosaccharide conformational dynamics.

Peptide-based allosteric inhibitor targets TNFR1 conformationally active region and disables receptor-ligand signaling complex

Tumor necrosis factor (TNF) receptor 1 (TNFR1) plays a pivotal role in mediating TNF induced downstream signaling and regulating inflammatory response. Recent studies have suggested that TNFR1 activation involves conformational rearrangements of preligand assembled receptor dimers and targeting receptor conformational dynamics is a viable strategy to modulate TNFR1 signaling. Here, we used a combination of biophysical, biochemical, and cellular assays, as well as molecular dynamics simulation to show that an anti-inflammatory peptide (FKCRRWQWRMKK), which we termed FKC, inhibits TNFR1 activation allosterically by altering the conformational states of the receptor dimer without blocking receptor-ligand interaction or disrupting receptor dimerization. We also demonstrated the efficacy of FKC by showing that the peptide inhibits TNFR1 signaling in HEK293 cells and attenuates inflammation in mice with intraperitoneal TNF injection. Mechanistically, we found that FKC binds to TNFR1 cysteine-rich domains (CRD2/3) and perturbs the conformational dynamics required for receptor activation. Importantly, FKC increases the frequency in the opening of both CRD2/3 and CRD4 in the receptor dimer, as well as induces a conformational opening in the cytosolic regions of the receptor. This results in an inhibitory conformational state that impedes the recruitment of downstream signaling molecules. Together, these data provide evidence on the feasibility of targeting TNFR1 conformationally active region and open new avenues for receptor-specific inhibition of TNFR1 signaling.

A comprehensive review of protein-centric predictors for biomolecular interactions: from proteins to nucleic acids and beyond

Proteins interact with diverse ligands to perform a large number of biological functions, such as gene expression and signal transduction. Accurate identification of these protein-ligand interactions is crucial to the understanding of molecular mechanisms and the development of new drugs. However, traditional biological experiments are time-consuming and expensive. With the development of high-throughput technologies, an increasing amount of protein data is available. In the past decades, many computational methods have been developed to predict protein-ligand interactions. Here, we review a comprehensive set of over 160 protein-ligand interaction predictors, which cover protein-protein, protein-nucleic acid, protein-peptide and protein-other ligands (nucleotide, heme, ion) interactions. We have carried out a comprehensive analysis of the above four types of predictors from several significant perspectives, including their inputs, feature profiles, models, availability, etc. The current methods primarily rely on protein sequences, especially utilizing evolutionary information. The significant improvement in predictions is attributed to deep learning methods. Additionally, sequence-based pretrained models and structure-based approaches are emerging as new trends.

Arginine (315) is required for the PLIN2-CGI-58 interface and plays a functional role in regulating nascent LDs formation in bovine adipocytes

Perilipin-2 (PLIN2) can anchor to lipid droplets (LDs) and play a crucial role in regulating nascent LDs formation. Bimolecular fluorescence complementation (BiFC) and flow cytometry were examined to verify the PLIN2-CGI-58 interaction efficiency in bovine adipocytes. GST-Pulldown assay was used to detect the key site arginine315 function in PLIN2-CGI-58 interaction. Experiments were also examined to research these mutations function of PLIN2 in LDs formation during adipocytes differentiation, LDs were measured after staining by BODIPY, lipogenesis-related genes were also detected. Results showed that Leucine (L371A, L311A) and glycine (G369A, G376A) mutations reduced interaction efficiencies. Serine (S367A) mutations enhanced the interaction efficiency. Arginine (R315A) mutations resulted in loss of fluorescence in the cytoplasm and disrupted the interaction with CGI-58, as verified by pulldown assay. R315W mutations resulted in a significant increase in the number of LDs compared with wild-type (WT) PLIN2 or the R315A mutations. Lipogenesis-related genes were either up- or downregulated when mutated PLIN2 interacted with CGI-58. Arginine315 in PLIN2 is required for the PLIN2-CGI-58 interface and could regulate nascent LD formation and lipogenesis. This study is the first to study amino acids on the PLIN2 interface during interaction with CGI-58 in bovine and highlight the role played by PLIN2 in the regulation of bovine adipocyte lipogenesis.

Conformational diversity and protein-protein interfaces in drug repurposing in Ras signaling pathway

We focus on drug repurposing in the Ras signaling pathway, considering structural similarities of protein-protein interfaces. The interfaces formed by physically interacting proteins are found from PDB if available and via PRISM (PRotein Interaction by Structural Matching) otherwise. The structural coverage of these interactions has been increased from 21 to 92% using PRISM. Multiple conformations of each protein are used to include protein dynamics and diversity. Next, we find FDA-approved drugs bound to structurally similar protein-protein interfaces. The results suggest that HIV protease inhibitors tipranavir, indinavir, and saquinavir may bind to EGFR and ERBB3/HER3 interface. Tipranavir and indinavir may also bind to EGFR and ERBB2/HER2 interface. Additionally, a drug used in Alzheimer's disease can bind to RAF1 and BRAF interface. Hence, we propose a methodology to find drugs to be potentially used for cancer using a dataset of structurally similar protein-protein interface clusters rather than pockets in a systematic way.

Computational analysis of protein-protein interactions of cancer drivers in renal cell carcinoma

Renal cell carcinoma (RCC) is the most common type of kidney cancer with rising cases in recent years. Extensive research has identified various cancer driver proteins associated with different subtypes of RCC. Most RCC drivers are encoded by tumor suppressor genes and exhibit enrichment in functional categories such as protein degradation, chromatin remodeling, and transcription. To further our understanding of RCC, we utilized powerful deep-learning methods based on AlphaFold to predict protein-protein interactions (PPIs) involving RCC drivers. We predicted high-confidence complexes formed by various RCC drivers, including TCEB1, KMT2C/D and KDM6A of the COMPASS-related complexes, TSC1 of the MTOR pathway, and TRRAP. These predictions provide valuable structural insights into the interaction interfaces, some of which are promising targets for cancer drug design, such as the NRF2-MAFK interface. Cancer somatic missense mutations from large datasets of genome sequencing of RCCs were mapped to the interfaces of predicted and experimental structures of PPIs involving RCC drivers, and their effects on the binding affinity were evaluated. We observed more than 100 cancer somatic mutations affecting the binding affinity of complexes formed by key RCC drivers such as VHL and TCEB1. These findings emphasize the importance of these mutations in RCC pathogenesis and potentially offer new avenues for targeted therapies.

Design and evaluation of α-helix-based peptide inhibitors for blocking PD-1/PD-L1 interaction

The current research in tumor immunotherapy indicates that blocking the protein-protein interaction (PPI) between PD-1 and its ligand, PD-L1, may be one of the most effective treatments for cancer patients. The α-helix is a common elements of protein secondary structure and is often involved in protein interaction. Thus, α-helix-based peptides could mimic proteins involved in such interactions and are also capable of modulating PPI in vivo. In this study, starting from a potential α-helix-rich protein, we designed a series of α-helix-based peptide candidates to block PD-1/PD-L1 interaction. These candidates were first screened using molecular docking and molecular dynamics simulations, and then their capacities to inhibit PD-1/PD-L1 interactions and to restore antitumor immune activities were investigated using the HTRF assay, SPR assay, cellular co-culture experiments and animal model experiments. Two peptides exhibited the best anti-tumor effects and the strong ability to restore the immunity of tumor-infiltrating T-cells. Further D-amino acid substitution was employed to improve the serum stability of peptide candidate, making the intravenous administration easier while maintaining the therapeutic efficacy. The resultant peptides showed promise as checkpoint inhibitors for application in tumor immunotherapy. These findings suggested that our strategy for developing peptides starting from an α-helical structure could be used in the design of bioactive inhibitors to potential block protein-protein interactions.

Underlying features for the enhanced electrostatic strength of the extremophilic malate dehydrogenase interface salt-bridge compared to the mesophilic one

Malate dehydrogenase (MDH) exists in multimeric form in normal and extreme solvent conditions where residues of the interface are involved in specific interactions. The interface salt-bridge (ISB) and its microenvironment (ME) residues may have a crucial role in the stability and specificity of the interface. To gain insight into this, we have analyzed 218 ISBs from 42 interfaces of 15 crystal structures along with their sequences. Comparative analyses demonstrate that the ISB strength is ∼30 times greater in extremophilic cases than that of the normal one. To this end, the interface residue propensity, ISB design and pair selection, and ME-residue's types, i.e., type-I and type-II, are seen to be intrinsically involved. Although Type-I is a common type, Type-II appears to be extremophile-specific, where the net ME-residue count is much lower with an excessive net ME-energy contribution, which seems to be a novel interface compaction strategy. Furthermore, the interface strength can be enhanced by selecting the desired mutant from the net-energy profile of all possible mutations of an unfavorable ME-residue. The study that applies to other similar systems finds applications in protein-protein interaction and protein engineering.Communicated by Ramaswamy H. Sarma.

Computer-Aided Drug Design and Drug Discovery: A Prospective Analysis

In the dynamic landscape of drug discovery, Computer-Aided Drug Design (CADD) emerges as a transformative force, bridging the realms of biology and technology. This paper overviews CADDs historical evolution, categorization into structure-based and ligand-based approaches, and its crucial role in rationalizing and expediting drug discovery. As CADD advances, incorporating diverse biological data and ensuring data privacy become paramount. Challenges persist, demanding the optimization of algorithms and robust ethical frameworks. Integrating Machine Learning and Artificial Intelligence amplifies CADDs predictive capabilities, yet ethical considerations and scalability challenges linger. Collaborative efforts and global initiatives, exemplified by platforms like Open-Source Malaria, underscore the democratization of drug discovery. The convergence of CADD with personalized medicine offers tailored therapeutic solutions, though ethical dilemmas and accessibility concerns must be navigated. Emerging technologies like quantum computing, immersive technologies, and green chemistry promise to redefine the future of CADD. The trajectory of CADD, marked by rapid advancements, anticipates challenges in ensuring accuracy, addressing biases in AI, and incorporating sustainability metrics. This paper concludes by highlighting the need for proactive measures in navigating the ethical, technological, and educational frontiers of CADD to shape a healthier, brighter future in drug discovery.

Transcription factors in the development and treatment of immune disorders

Immune function is highly controlled at the transcriptional level by the binding of transcription factors (TFs) to promoter and enhancer elements. Several TF families play major roles in immune gene expression, including NF-κB, STAT, IRF, AP-1, NRs, and NFAT, which trigger anti-pathogen responses, promote cell differentiation, and maintain immune system homeostasis. Aberrant expression, activation, or sequence of isoforms and variants of these TFs can result in autoimmune and inflammatory diseases as well as hematological and solid tumor cancers. For this reason, TFs have become attractive drug targets, even though most were previously deemed "undruggable" due to their lack of small molecule binding pockets and the presence of intrinsically disordered regions. However, several aspects of TF structure and function can be targeted for therapeutic intervention, such as ligand-binding domains, protein-protein interactions between TFs and with cofactors, TF-DNA binding, TF stability, upstream signaling pathways, and TF expression. In this review, we provide an overview of each of the important TF families, how they function in immunity, and some related diseases they are involved in. Additionally, we discuss the ways of targeting TFs with drugs along with recent research developments in these areas and their clinical applications, followed by the advantages and disadvantages of targeting TFs for the treatment of immune disorders.

A Multimodal Deep Learning Framework for Predicting PPI-Modulator Interactions

Protein-protein interactions (PPIs) are essential for various biological processes and diseases. However, most existing computational methods for identifying PPI modulators require either target structure or reference modulators, which restricts their applicability to novel PPI targets. To address this challenge, we propose MultiPPIMI, a sequence-based deep learning framework that predicts the interaction between any given PPI target and modulator. MultiPPIMI integrates multimodal representations of PPI targets and modulators and uses a bilinear attention network to capture intermolecular interactions. Experimental results on our curated benchmark data set show that MultiPPIMI achieves an average AUROC of 0.837 in three cold-start scenarios and an AUROC of 0.994 in the random-split scenario. Furthermore, the case study shows that MultiPPIMI can assist molecular docking simulations in screening inhibitors of Keap1/Nrf2 PPI interactions. We believe that the proposed method provides a promising way to screen PPI-targeted modulators.

Streamlined Protein-Protein Interface Loop Mimicry

Cyclic peptides comprising endocyclic organic fragments, "cyclo-organopeptides", can be probes for perturbing protein-protein interactions (PPIs). Finding loop mimics is difficult because of high conformational variability amongst targets. Backbone Matching (BM), introduced here, helps solve this problem in the illustrative cases by facilitating efficient evaluation of virtual cyclo-organopeptide core-structure libraries. Thus, 86 rigid organic fragments were selected to build a library of 602 cyclo-organopeptides comprising Ala and organic parts: "cyclo-{-(Ala)n -organo-}". The central hypothesis is "hit" library members have accessible low energy conformers corresponding to backbone structures of target protein loops, while library members which cannot attain this conformation are probably unworthy of further evaluation. BM thereby prioritizes candidate loop mimics, so that less than 10 cyclo-organopeptides are needed to be prepared to find leads for two illustrative PPIs: iNOS ⋅ SPSB2, and uPA ⋅ uPAR.

DeepProSite: structure-aware protein binding site prediction using ESMFold and pretrained language model

MOTIVATION

Identifying the functional sites of a protein, such as the binding sites of proteins, peptides, or other biological components, is crucial for understanding related biological processes and drug design. However, existing sequence-based methods have limited predictive accuracy, as they only consider sequence-adjacent contextual features and lack structural information.

RESULTS

In this study, DeepProSite is presented as a new framework for identifying protein binding site that utilizes protein structure and sequence information. DeepProSite first generates protein structures from ESMFold and sequence representations from pretrained language models. It then uses Graph Transformer and formulates binding site predictions as graph node classifications. In predicting protein-protein/peptide binding sites, DeepProSite outperforms state-of-the-art sequence- and structure-based methods on most metrics. Moreover, DeepProSite maintains its performance when predicting unbound structures, in contrast to competing structure-based prediction methods. DeepProSite is also extended to the prediction of binding sites for nucleic acids and other ligands, verifying its generalization capability. Finally, an online server for predicting multiple types of residue is established as the implementation of the proposed DeepProSite.

AVAILABILITY AND IMPLEMENTATION

The datasets and source codes can be accessed at https://github.com/WeiLab-Biology/DeepProSite. The proposed DeepProSite can be accessed at https://inner.wei-group.net/DeepProSite/.

Modular Nanotransporters Delivering Biologically Active Molecules to the Surface of Mitochondria

Treatment of various diseases, in particular cancer, usually requires the targeting of biologically active molecules at a selected subcellular compartment. We modified our previously developed modular nanotransporters (MNTs) for targeting mitochondria. The new MNTs are capable of binding to the protein predominantly localized on the outer mitochondrial membrane, Keap1. These MNTs possessing antiKeap1 monobody co-localize with mitochondria upon addition to the cells. They efficiently interact with Keap1 both in solution and within living cells. A conjugate of the MNT with a photosensitizer, chlorin e6, demonstrated significantly higher photocytotoxicity than chlorin e6 alone. We assume that MNTs of this kind can improve efficiency of therapeutic photosensitizers and radionuclides emitting short-range particles.

Computational Modeling of Stapled Coiled-Coil Inhibitors Against Bcr-Abl: Toward a Treatment Strategy for CML

The chimeric oncoprotein Bcr-Abl is the causative agent of virtually all chronic myeloid leukemias (CML) and a subset of acute lymphoblastic leukemias (ALL). As a result of the so-called Philadelphia Chromosome translocation t(9;22), Bcr-Abl manifests as a constitutively active tyrosine kinase which promotes leukemogenesis by activation of cell cycle signaling pathways. Constitutive and oncogenic activation is mediated by an N-terminal coiled-coil oligomerization domain in Bcr (Bcr-CC), presenting a therapeutic target for inhibition of Bcr-Abl activity toward the treatment of Bcr-Abl+ leukemias. Previously, we demonstrated that a rationally designed Bcr-CC mutant, CCmut3, exerts a dominant negative effect upon Bcr-Abl activity by preferential oligomerization with Bcr-CC. Moreover, we have shown conjugation to a leukemia-specific cell-penetrating peptide (CPP-CCmut3) improves intracellular delivery and activity. However, our full-length CPP-CCmut3 construct (81 aa) is encumbered by an intrinsically high degree of conformational variability and susceptibility to proteolytic degradation, relative to traditional small molecule therapeutics. Here, we iterate a new generation of our inhibitor against Bcr-CC mediated Bcr-Abl assembly that is designed to address these constraints through incorporation of all-hydrocarbon staples spanning i, i + 7 positions in helix α2 (CPP-CCmut3-st). We utilize computational modeling and biomolecular simulation to design and characterize single and double staple candidates in silico, evaluating binding energetics and building upon our seminal work modeling single hydrocarbon staples when applied to a truncated Bcr-CC sequence. This strategy enables us to efficiently build, characterize, and screen lead single/double stapled CPP-CCmut3-st candidates for experimental studies and validation in vitro and in vivo. In addition to full-length CPP-CCmut, we model a truncated system characterized by deletion of helix α1 and the flexible-loop linker, which are known to impart high conformational variability. To study the impact of the N-terminal cyclic CPP toward model stability and inhibitor activity, we also model the full-length and truncated systems without CPP, with cyclized CPP, and with linear CPP, for a total of six systems which comprise our library. From this library, we present lead stapled peptide candidates to be synthesized and evaluated experimentally as our next-generation inhibitors against Bcr-Abl.

Structural Insights into Protein-Aptamer Recognitions Emerged from Experimental and Computational Studies

Aptamers are synthetic nucleic acids that are developed to target with high affinity and specificity chemical entities ranging from single ions to macromolecules and present a wide range of chemical and physical properties. Their ability to selectively bind proteins has made these compounds very attractive and versatile tools, in both basic and applied sciences, to such an extent that they are considered an appealing alternative to antibodies. Here, by exhaustively surveying the content of the Protein Data Bank (PDB), we review the structural aspects of the protein-aptamer recognition process. As a result of three decades of structural studies, we identified 144 PDB entries containing atomic-level information on protein-aptamer complexes. Interestingly, we found a remarkable increase in the number of determined structures in the last two years as a consequence of the effective application of the cryo-electron microscopy technique to these systems. In the present paper, particular attention is devoted to the articulated architectures that protein-aptamer complexes may exhibit. Moreover, the molecular mechanism of the binding process was analyzed by collecting all available information on the structural transitions that aptamers undergo, from their protein-unbound to the protein-bound state. The contribution of computational approaches in this area is also highlighted.

Validation of a New Methodology to Create Oral Drugs beyond the Rule of 5 for Intracellular Tough Targets

Establishing a technological platform for creating clinical compounds inhibiting intracellular protein-protein interactions (PPIs) can open the door to many valuable drugs. Although small molecules and antibodies are mainstream modalities, they are not suitable for a target protein that lacks a deep cavity for a small molecule to bind or a protein found in intracellular space out of an antibody's reach. One possible approach to access these targets is to utilize so-called middle-size cyclic peptides (defined here as those with a molecular weight of 1000-2000 g/mol). In this study, we validated a new methodology to create oral drugs beyond the rule of 5 for intracellular tough targets by elucidating structural features and physicochemical properties for drug-like cyclic peptides and developing library technologies to afford highly N-alkylated cyclic peptide hits. We discovered a KRAS inhibitory clinical compound (LUNA18) as the first example of our platform technology.

State of the art in protein-protein interactions within the fungi kingdom

Proteins rarely exert their function by themselves. Protein-protein interactions (PPIs) regulate virtually every biological process that takes place in a cell. Such interactions are targets for new therapeutic agents against all sorts of diseases, through the screening and design of a variety of inhibitors. Here we discuss several aspects of PPIs that contribute to prediction of protein function and drug discovery. As the high-throughput techniques continue to release biological data, targets for fungal therapeutics that rely on PPIs are being proposed worldwide. Computational approaches have reduced the time taken to develop new therapeutic approaches. The near future brings the possibility of developing new PPI and interaction network inhibitors and a revolution in the way we treat fungal diseases.

Development of covalent chemogenetic K2P channel activators

K2P potassium channels regulate excitability by affecting cellular resting membrane potential in the brain, cardiovascular system, immune cells, and sensory organs. Despite their important roles in anesthesia, arrhythmia, pain, hypertension, sleep, and migraine, the ability to control K2P function remains limited. Here, we describe a chemogenetic strategy termed CATKLAMP (Covalent Activation of TREK family K+ channels to cLAmp Membrane Potential) that leverages the discovery of a site in the K2P modulator pocket that reacts with electrophile-bearing derivatives of a TREK subfamily small molecule activator, ML335, to activate the channel irreversibly. We show that the CATKLAMP strategy can be used to probe fundamental aspects of K2P function, as a switch to silence neuronal firing, and is applicable to all TREK subfamily members. Together, our findings exemplify a new means to alter K2P channel activity that should facilitate studies both molecular and systems level studies of K2P function and enable the search for new K2P modulators.

Characterization of allosteric modulators that disrupt androgen receptor co-activator protein-protein interactions to alter transactivation-Drug leads for metastatic castration resistant prostate cancer

Three series of compounds were prioritized from a high content screening campaign that identified molecules that blocked dihydrotestosterone (DHT) induced formation of Androgen Receptor (AR) protein-protein interactions (PPIs) with the Transcriptional Intermediary Factor 2 (TIF2) coactivator and also disrupted preformed AR-TIF2 PPI complexes; the hydrobenzo-oxazepins (S1), thiadiazol-5-piperidine-carboxamides (S2), and phenyl-methyl-indoles (S3). Compounds from these series inhibited AR PPIs with TIF2 and SRC-1, another p160 coactivator, in mammalian 2-hybrid assays and blocked transcriptional activation in reporter assays driven by full length AR or AR-V7 splice variants. Compounds inhibited the growth of five prostate cancer cell lines, with many exhibiting differential cytotoxicity towards AR positive cell lines. Representative compounds from the 3 series substantially reduced both endogenous and DHT-enhanced expression and secretion of the prostate specific antigen (PSA) cancer biomarker in the C4-2 castration resistant prostate cancer (CRPC) cell line. The comparatively weak activities of series compounds in the H3-DHT and/or TIF2 box 3 LXXLL-peptide binding assays to the recombinant ligand binding domain of AR suggest that direct antagonism at the orthosteric ligand binding site or AF-2 surface respectively are unlikely mechanisms of action. Cellular enhanced thermal stability assays (CETSA) indicated that compounds engaged AR and reduced the maximum efficacy and right shifted the EC50 of DHT-enhanced AR thermal stabilization consistent with the effects of negative allosteric modulators. Molecular docking of potent representative hits from each series to AR structures suggest that S1-1 and S2-6 engage a novel binding pocket (BP-1) adjacent to the orthosteric ligand binding site, while S3-11 occupies the AR binding function 3 (BF-3) allosteric pocket. Hit binding poses indicate spaces and residues adjacent to the BP-1 and BF-3 pockets that will be exploited in future medicinal chemistry optimization studies. Small molecule allosteric modulators that prevent/disrupt AR PPIs with coactivators like TIF2 to alter transcriptional activation in the presence of orthosteric agonists might evade the resistance mechanisms to existing prostate cancer drugs and provide novel starting points for medicinal chemistry lead optimization and future development into therapies for metastatic CRPC.

Spatom: a graph neural network for structure-based protein-protein interaction site prediction

Accurate identification of protein-protein interaction (PPI) sites remains a computational challenge. We propose Spatom, a novel framework for PPI site prediction. This framework first defines a weighted digraph for a protein structure to precisely characterize the spatial contacts of residues, then performs a weighted digraph convolution to aggregate both spatial local and global information and finally adds an improved graph attention layer to drive the predicted sites to form more continuous region(s). Spatom was tested on a diverse set of challenging protein-protein complexes and demonstrated the best performance among all the compared methods. Furthermore, when tested on multiple popular proteins in a case study, Spatom clearly identifies the interaction interfaces and captures the majority of hotspots. Spatom is expected to contribute to the understanding of protein interactions and drug designs targeting protein binding.

Molecular Docking and Molecular Dynamics Studies Reveal the Anticancer Potential of Medicinal-Plant-Derived Lignans as MDM2-P53 Interaction Inhibitors

The interaction between the tumor suppressor protein p53 and its negative regulator, the MDM2 oncogenic protein, has gained significant attention in cancer drug discovery. In this study, 120 lignans reported from Ferula sinkiangensis and Justicia procumbens were assessed for docking simulations on the active pocket of the MDM2 crystal structure bound to Nutlin-3a. The docking analysis identified nine compounds with higher docking scores than the co-crystallized reference. Subsequent AMDET profiling revealed satisfactory pharmacokinetic and safety parameters for these natural products. Three compounds, namely, justin A, 6-hydroxy justicidin A, and 6'-hydroxy justicidin B, were selected for further investigation due to their strong binding affinities of -7.526 kcal/mol, -7.438 kcal/mol, and -7.240 kcal/mol, respectively, which surpassed the binding affinity of the reference inhibitor Nutlin-3a (-6.830 kcal/mol). To assess the stability and reliability of the binding of the candidate hits, a molecular dynamics simulation was performed over a duration of 100 ns. Remarkably, the thorough analysis demonstrated that all the hits exhibited stable molecular dynamics profiles. Based on their effective binding to MDM2, favorable pharmacokinetic properties, and molecular dynamics behavior, these compounds represent a promising starting point for further refinement. Nevertheless, it is essential to synthesize the suggested compounds and evaluate their activity through in vitro and in vivo experiments.

Modulation of a conformational ensemble by a small molecule that inhibits key protein-protein interactions involved in cell adhesion

Small molecules that regulate protein-protein interactions can be valuable drugs; however, the development of such small molecules is challenging as the molecule must interfere with an interaction that often involves a large surface area. Herein, we propose that modulating the conformational ensemble of the proteins participating in a given interaction, rather than blocking the interaction by directly binding to the interface, is a relevant strategy for interfering with a protein-protein interaction. In this study, we applied this concept to P-cadherin, a cell surface protein forming homodimers that are essential for cell-cell adhesion in various biological contexts. We first determined the crystal structure of P-cadherin with a small molecule inhibitor whose inhibitory mechanism was unknown. Molecular dynamics simulations suggest that the inhibition of cell adhesion by this small molecule results from modulation of the conformational ensemble of P-cadherin. Our study demonstrates the potential of small molecules altering the conformation ensemble of a protein as inhibitors of biological relevant protein-protein interactions.

Conformationally constrained cyclic grafted peptidomimetics targeting protein-protein interactions

Sunflower trypsin inhibitor-1 (SFTI-1) structure is used for designing grafted peptides as a possible therapeutic agent. The grafted peptide exhibits multiple conformations in solution due to the presence of proline in the structure of the peptide. To lock the grafted peptide into a major conformation in solution, a dibenzofuran moiety (DBF) was incorporated in the peptide backbone structure, replacing the Pro-Pro sequence. NMR studies indicated a major conformation of the grafted peptide in solution. Detailed structural studies suggested that SFTI-DBF adopts a twisted beta-strand structure in solution. The surface plasmon resonance analysis showed that SFTI-DBF binds to CD58 protein. A model for the protein-SFTI-DBF complex was proposed based on docking studies. These studies suggested that SFTI-1 grafted peptide can be used to design stable peptides for therapeutic purposes by grafting organic functional groups and amino acids. However, when a similar strategy was used with another grafted peptide, the resulting peptide did not produce a single major conformation, and its biological activity was lost. Thus, conformational constraints depend on the sequence of amino acids used for SFTI-1 grafting.

An efficient computational protocol for template-based design of peptides that inhibit interactions involving SARS-CoV-2 proteins

The RNA-dependent RNA polymerase (RdRp) complex of SARS-CoV-2 lies at the core of its replication and transcription processes. The interfaces between holo-RdRp subunits are highly conserved, facilitating the design of inhibitors with high affinity for the interaction interface hotspots. We, therefore, take this as a model protein complex for the application of a structural bioinformatics protocol to design peptides that inhibit RdRp complexation by preferential binding at the interface of its core subunit nonstructural protein, nsp12, with accessory factor nsp7. Here, the interaction hotspots of the nsp7-nsp12 subunit of RdRp, determined from a long molecular dynamics trajectory, are used as a template. A large library of peptide sequences constructed from multiple hotspot motifs of nsp12 is screened in-silico to determine sequences with high geometric complementarity and interaction specificity for the binding interface of nsp7 (target) in the complex. Two lead designed peptides are extensively characterized using orthogonal bioanalytical methods to determine their suitability for inhibition of RdRp complexation. Binding affinity of these peptides to accessory factor nsp7, determined using a surface plasmon resonance (SPR) assay, is slightly better than that of nsp12: dissociation constant of 133nM and 167nM, respectively, compared to 473nM for nsp12. A competitive ELISA is used to quantify inhibition of nsp7-nsp12 complexation, with one of the lead peptides giving an IC50 of 25μM . Cell penetrability and cytotoxicity are characterized using a cargo delivery assay and MTT cytotoxicity assay, respectively. Overall, this work presents a proof-of-concept of an approach for rational discovery of peptide inhibitors of SARS-CoV-2 protein-protein interactions.

DIPS-Plus: The enhanced database of interacting protein structures for interface prediction

In this work, we expand on a dataset recently introduced for protein interface prediction (PIP), the Database of Interacting Protein Structures (DIPS), to present DIPS-Plus, an enhanced, feature-rich dataset of 42,112 complexes for machine learning of protein interfaces. While the original DIPS dataset contains only the Cartesian coordinates for atoms contained in the protein complex along with their types, DIPS-Plus contains multiple residue-level features including surface proximities, half-sphere amino acid compositions, and new profile hidden Markov model (HMM)-based sequence features for each amino acid, providing researchers a curated feature bank for training protein interface prediction methods. We demonstrate through rigorous benchmarks that training an existing state-of-the-art (SOTA) model for PIP on DIPS-Plus yields new SOTA results, surpassing the performance of some of the latest models trained on residue-level and atom-level encodings of protein complexes to date.

Development of Orally Bioavailable Peptides Targeting an Intracellular Protein: From a Hit to a Clinical KRAS Inhibitor

Cyclic peptides as a therapeutic modality are attracting a lot of attention due to their potential for oral absorption and accessibility to intracellular tough targets. Here, starting with a drug-like hit discovered using an mRNA display library, we describe a chemical optimization that led to the orally available clinical compound known as LUNA18, an 11-mer cyclic peptide inhibitor for the intracellular tough target RAS. The key findings are as follows: (i) two peptide side chains were identified that each increase RAS affinity over 10-fold; (ii) physico-chemical properties (PCP) including Clog P can be adjusted by side-chain modification to increase membrane permeability; (iii) restriction of cyclic peptide conformation works effectively to adjust PCP and improve bio-activity; (iv) cellular efficacy was observed in peptides with a permeability of around 0.4 × 10^-6 cm/s or more in a Caco-2 permeability assay; and (v) while keeping the cyclic peptide's main-chain conformation, we found one example where the RAS protein structure was changed dramatically through induced-fit to our peptide side chain. This study demonstrates how the chemical optimization of bio-active peptides can be achieved without scaffold hopping, much like the processes for small molecule drug discovery that are guided by Lipinski's rule of five. Our approach provides a versatile new strategy for generating peptide drugs starting from drug-like hits.

Chemical and Biological Strategies for Profiling Protein-Protein Interactions in Living Cells

Protein-protein interactions (PPIs) play critical roles in almost all cellular signal transduction events. Characterization of PPIs without interfering with the functions of intact cells is very important for basic biology study and drug developments. However, the ability to profile PPIs especially those weak/transient interactions in their native states remains quite challenging. To this end, many endeavors are being made in developing new methods with high efficiency and strong operability. By coupling with advanced fluorescent microscopy and mass spectroscopy techniques, these strategies not only allow us to visualize the subcellular locations and monitor the functions of protein of interest (POI) in real time, but also enable the profiling and identification of potential unknown interacting partners in high-throughput manner, which greatly facilitates the elucidation of molecular mechanisms underlying numerous pathophysiological processes. In this review, we will summarize the typical methods for PPIs identification in living cells and their principles, advantages and limitations will also be discussed in detail.

Identification of small-molecule inhibitors against the interaction of RNA-binding protein PSF and its target RNA for cancer treatment

Diverse cellular activities are modulated through a variety of RNAs, including long noncoding RNAs (lncRNAs), by binding to certain proteins. The inhibition of oncogenic proteins or RNAs is expected to suppress cancer cell proliferation. We have previously demonstrated that PSF interaction with its target RNAs, such as androgen-induced lncRNA CTBP1-AS, is critical for hormone therapy resistance in prostate and breast cancers. However, the action of protein-RNA interactions remains almost undruggable to date. High-throughput screening (HTS) has facilitated the discovery of drugs for protein-protein interactions. In the present study, we developed an in vitro alpha assay using Flag peptide-conjugated lncRNA, CTBP1-AS, and PSF. We then constructed an effective HTS screening system to explore small compounds that inhibit PSF-RNA interactions. Thirty-six compounds were identified and dose-dependently inhibited PSF-RNA interaction in vitro. Moreover, chemical optimization of these lead compounds and evaluation of cancer cell proliferation revealed two promising compounds, N-3 and C-65. These compounds induced apoptosis and inhibited cell growth in prostate and breast cancer cells. By inhibiting PSF-RNA interaction, N-3 and C-65 up-regulated signals that are repressed by PSF, such as the cell cycle signals by p53 and p27. Furthermore, using a mouse xenograft model for hormone therapy-resistant prostate cancer, we revealed that N-3 and C-65 can significantly suppress tumor growth and downstream target gene expression, such as the androgen receptor (AR). Thus, our findings highlight a therapeutic strategy through the development of inhibitors for RNA-binding events in advanced cancers.

What Makes a Good Protein-Protein Interaction Stabilizer: Analysis and Application of the Dual-Binding Mechanism

Protein-protein interactions (PPIs) are essential for biological processes including immune reactions and diseases. Inhibition of PPIs by drug-like compounds is a common basis for therapeutic approaches. In many cases the flat interface of PP complexes prevents discovery of specific compound binding to cavities on one partner and PPI inhibition. However, frequently new pockets are formed at the PP interface that allow accommodation of stabilizers which is often as desirable as inhibition but a much less explored alternative strategy. Herein, we employ molecular dynamics simulations and pocket detection to investigate 18 known stabilizers and associated PP complexes. For most cases, we find that a dual-binding mechanism, a similar stabilizer interaction strength with each protein partner, is an important prerequisite for effective stabilization. A few stabilizers follow an allosteric mechanism by stabilizing the protein bound structure and/or increase the PPI indirectly. On 226 protein-protein complexes, we find in >75% of the cases interface cavities suitable for binding of drug-like compounds. We propose a computational compound identification workflow that exploits new PP interface cavities and optimizes the dual-binding mechanism and apply it to 5 PP complexes. Our study demonstrates a great potential for in silico PPI stabilizers discovery with a wide range of therapeutic applications.

Doubly Constrained C-terminal of Roc (COR) Domain-Derived Peptides Inhibit Leucine-Rich Repeat Kinase 2 (LRRK2) Dimerization

Missense mutations along the leucine-rich repeat kinase 2 (LRRK2) protein are a major contributor to Parkinson's Disease (PD), the second most commonly occurring neurodegenerative disorder worldwide. We recently reported the development of allosteric constrained peptide inhibitors that target and downregulate LRRK2 activity through disruption of LRRK2 dimerization. In this study, we designed doubly constrained peptides with the objective of inhibiting C-terminal of Roc (COR)-COR mediated dimerization at the LRRK2 dimer interface. We show that the doubly constrained peptides are cell-permeant, bind wild-type and pathogenic LRRK2, inhibit LRRK2 dimerization and kinase activity, and inhibit LRRK2-mediated neuronal apoptosis, and in contrast to ATP-competitive LRRK2 kinase inhibitors, they do not induce the mislocalization of LRRK2 to skein-like structures in cells. This work highlights the significance of COR-mediated dimerization in LRRK2 activity while also highlighting the use of doubly constrained peptides to stabilize discrete secondary structural folds within a peptide sequence.

Proteomic discovery of chemical probes that perturb protein complexes in human cells

Most human proteins lack chemical probes, and several large-scale and generalizable small-molecule binding assays have been introduced to address this problem. How compounds discovered in such "binding-first" assays affect protein function, nonetheless, often remains unclear. Here, we describe a "function-first" proteomic strategy that uses size exclusion chromatography (SEC) to assess the global impact of electrophilic compounds on protein complexes in human cells. Integrating the SEC data with cysteine-directed activity-based protein profiling identifies changes in protein-protein interactions that are caused by site-specific liganding events, including the stereoselective engagement of cysteines in PSME1 and SF3B1 that disrupt the PA28 proteasome regulatory complex and stabilize a dynamic state of the spliceosome, respectively. Our findings thus show how multidimensional proteomic analysis of focused libraries of electrophilic compounds can expedite the discovery of chemical probes with site-specific functional effects on protein complexes in human cells.

Using a stacked ensemble learning framework to predict modulators of protein-protein interactions

Identifying small molecule protein-protein interaction modulators (PPIMs) is a highly promising and meaningful research direction for drug discovery, cancer treatment, and other fields. In this study, we developed a stacking ensemble computational framework, SELPPI, based on a genetic algorithm and tree-based machine learning method for effectively predicting new modulators targeting protein-protein interactions. More specifically, extremely randomized trees (ExtraTrees), adaptive boosting (AdaBoost), random forest (RF), cascade forest, light gradient boosting machine (LightGBM), and extreme gradient boosting (XGBoost) were used as basic learners. Seven types of chemical descriptors were taken as the input characteristic parameters. Primary predictions were obtained with each basic learner-descriptor pair. Then, the 6 methods mentioned above were used as meta learners and trained on the primary prediction in turn. The most efficient method was utilized as the meta learner. Finally, the genetic algorithm was used to select the optimal primary prediction output as the input of the meta learner for secondary prediction to obtain the final result. We systematically evaluated our model on the pdCSM-PPI datasets. To our knowledge, our model outperformed all existing models, which demonstrates its great power.