Protein-protein Interaction Sites Research Articles

Pair-EGRET: Enhancing the prediction of protein-protein interaction sites through graph attention networks and protein language models.

Proteins are responsible for most biological functions, many of which require the interaction of more than one protein molecule. However, accurately predicting protein-protein interaction (PPI) sites (the interfacial residues of a protein that interact with other protein molecules) remains a challenge. The growing demand and cost associated with the reliable identification of PPI sites using conventional experimental methods call for computational tools for automated prediction and understanding of PPIs. We present Pair-EGRET, an edge-aggregated graph attention network that leverages the features extracted from pre-trained transformer-like models to accurately predict PPI sites. Pair-EGRET works on a k-nearest neighbor graph, representing the three-dimensional structure of a protein, and utilizes the cross-attention mechanism for accurate identification of interfacial residues of a pair of proteins. Through an extensive evaluation study using a diverse array of experimental data, evaluation metrics, and case studies on representative protein sequences, we demonstrate that Pair-EGRET can achieve remarkable performance in predicting PPI sites. Moreover, Pair-EGRET can provide interpretable insights from the learned cross-attention matrix. Pair-EGRET is freely available in open source form at the GitHub Repository https://github.com/1705004/Pair-EGRET. Supplementary data are available at Bioinformatics online.

DGCPPISP: a PPI site prediction model based on dynamic graph convolutional network and two-stage transfer learning

BackgroundProteins play a pivotal role in the diverse array of biological processes, making the precise prediction of protein–protein interaction (PPI) sites critical to numerous disciplines including biology, medicine and pharmacy. While deep learning methods have progressively been implemented for the prediction of PPI sites within proteins, the task of enhancing their predictive performance remains an arduous challenge.ResultsIn this paper, we propose a novel PPI site prediction model (DGCPPISP) based on a dynamic graph convolutional neural network and a two-stage transfer learning strategy. Initially, we implement the transfer learning from dual perspectives, namely feature input and model training that serve to supply efficacious prior knowledge for our model. Subsequently, we construct a network designed for the second stage of training, which is built on the foundation of dynamic graph convolution.ConclusionsTo evaluate its effectiveness, the performance of the DGCPPISP model is scrutinized using two benchmark datasets. The ensuing results demonstrate that DGCPPISP outshines competing methods in terms of performance. Specifically, DGCPPISP surpasses the second-best method, EGRET, by margins of 5.9%, 10.1%, and 13.3% for F1-measure, AUPRC, and MCC metrics respectively on Dset_186_72_PDB164. Similarly, on Dset_331, it eclipses the performance of the runner-up method, HN-PPISP, by 14.5%, 19.8%, and 29.9% respectively.

Deciphering the Genetic Variation: A Comparative Analysis of Parental and Attenuated Strains of the QXL87 Vaccine for Infectious Bronchitis.

The QXL87 live attenuated vaccine strain for infectious bronchitis represents the first approved QX type (GI-19 lineage) vaccine in China. This strain was derived from the parental strain CK/CH/JS/2010/12 through continuous passage in SPF chicken embryos. To elucidate the molecular mechanism behind its attenuation, whole-genome sequencing was conducted on both the parental and attenuated strains. Analysis revealed 145 nucleotide mutations in the attenuated strain, leading to 48 amino acid mutations in various proteins, including Nsp2 (26), Nsp3 (14), Nsp4 (1), S (4), 3a (1), E (1), and N (1). Additionally, a frameshift mutation caused by a single base insertion in the ORFX resulted in a six-amino-acid extension. Subsequent comparison of post-translational modification sites, protein structure, and protein-protein binding sites between the parental and attenuated strains identified three potential virulence genes: Nsp2, Nsp3, and S. The amino acid mutations in these proteins not only altered their conformation but also affected the distribution of post-translational modification sites and protein-protein interaction sites. Furthermore, three potential functional mutation sites-P106S, A352T, and L472F, all located in the Nsp2 protein-were identified through PROVEAN, PolyPhen, and I-Mutant. Overall, our findings suggest that Nsp2, Nsp3, and S proteins may play a role in modulating IBV pathogenicity, with a particular focus on the significance of the Nsp2 protein. This study contributes to our understanding of the molecular mechanisms underlying IBV attenuation and holds promise for the development of safer live attenuated IBV vaccines using reverse genetic approaches.

MR2CPPIS: Accurate prediction of protein–protein interaction sites based on multi-scale Res2Net with coordinate attention mechanism

Proteins play a vital role in various biological processes and achieve their functions through protein–protein interactions (PPIs). Thus, accurate identification of PPI sites is essential. Traditional biological methods for identifying PPIs are costly, labor-intensive, and time-consuming. The development of computational prediction methods for PPI sites offers promising alternatives. Most known deep learning (DL) methods employ layer-wise multi-scale CNNs to extract features from protein sequences. But, these methods usually neglect the spatial positions and hierarchical information embedded within protein sequences, which are actually crucial for PPI site prediction. In this paper, we propose MR2CPPIS, a novel sequence-based DL model that utilizes the multi-scale Res2Net with coordinate attention mechanism to exploit multi-scale features and enhance PPI site prediction capability. We leverage the multi-scale Res2Net to expand the receptive field for each network layer, thus capturing multi-scale information of protein sequences at a granular level. To further explore the local contextual features of each target residue, we employ a coordinate attention block to characterize the precise spatial position information, enabling the network to effectively extract long-range dependencies. We evaluate our MR2CPPIS on three public benchmark datasets (Dset 72, Dset 186, and PDBset 164), achieving state-of-the-art performance. The source codes are available at https://github.com/YyinGong/MR2CPPIS.

Structural studies of a serum amyloid A octamer that is primed to scaffold lipid nanodiscs.

Serum amyloid A (SAA) is a highly conserved acute-phase protein that plays roles in activating multiple pro-inflammatory pathways during the acute inflammatory response and is commonly used as a biomarker of inflammation. It has been linked to beneficial roles in tissue repair through improved clearance of lipids and cholesterol from sites of damage. In patients with chronic inflammatory diseases, elevated levels of SAA may contribute to increased severity of the underlying condition. The majority of circulating SAA is bound to lipoproteins, primarily high-density lipoprotein (HDL). Interaction with HDL not only stabilizes SAA but also alters its functional properties, likely through altered accessibility of protein-protein interaction sites on SAA. While high-resolution structures for lipid-free, or apo-, forms of SAA have been reported, their relationship with the HDL-bound form of the protein, and with other possible mechanisms of SAA binding to lipids, has not been established. Here, we have used multiple biophysical techniques, including SAXS, TEM, SEC-MALS, native gel electrophoresis, glutaraldehyde crosslinking, and trypsin digestion to characterize the lipid-free and lipid-bound forms of SAA. The SAXS and TEM data show the presence of soluble octamers of SAA with structural similarity to the ring-like structures reported for lipid-free ApoA-I. These SAA octamers represent a previously uncharacterized structure for lipid-free SAA and are capable of scaffolding lipid nanodiscs with similar morphology to those formed by ApoA-I. The SAA-lipid nanodiscs contain four SAA molecules and have similar exterior dimensions as the lipid-free SAA octamer, suggesting that relatively few conformational rearrangements may be required to allow SAA interactions with lipid-containing particles such as HDL. This study suggests a new model for SAA-lipid interactions and provides new insight into how SAA might stabilize protein-lipid nanodiscs or even replace ApoA-I as a scaffold for HDL particles during inflammation.

MEG-PPIS: a fast protein-protein interaction site prediction method based on multi-scale graph information and equivariant graph neural network.

Protein-protein interaction sites (PPIS) are crucial for deciphering protein action mechanisms and related medical research, which is the key issue in protein action research. Recent studies have shown that graph neural networks have achieved outstanding performance in predicting PPIS. However, these studies often neglect the modeling of information at different scales in the graph and the symmetry of protein molecules within three-dimensional space. In response to this gap, this article proposes the MEG-PPIS approach, a PPIS prediction method based on multi-scale graph information and E(n) equivariant graph neural network (EGNN). There are two channels in MEG-PPIS: the original graph and the subgraph obtained by graph pooling. The model can iteratively update the features of the original graph and subgraph through the weight-sharing EGNN. Subsequently, the max-pooling operation aggregates the updated features of the original graph and subgraph. Ultimately, the model feeds node features into the prediction layer to obtain prediction results. Comparative assessments against other methods on benchmark datasets reveal that MEG-PPIS achieves optimal performance across all evaluation metrics and gets the fastest runtime. Furthermore, specific case studies demonstrate that our method can predict more true positive and true negative sites than the current best method, proving that our model achieves better performance in the PPIS prediction task. The data and code are available at https://github.com/dhz234/MEG-PPIS.git.

Abstract P52: Combination of 5-fluorouracil and Menadione Exerts Synergistic Effect on Colorectal Cancer Cells by Targeting Wnt/β-catenin Signaling Pathway

Abstract Colorectal cancer is the third most common type of cancer and the second most in terms of mortality rate worldwide. 5-fluorouracil (5-FU), the fluoropyrimidine drug, that interferes with DNA replication and repair processes, is the first-line-of-treatment against colorectal cancer. However, cells gain resistance over time and relapses occur upon continuous 5-FU therapy. Interestingly, 5-FU promotes stemness in colorectal cancer via Wnt signaling pathway which is critical for its progression. Studies have pointed out the possibility of using a Wnt signaling inhibitor along with 5-FU as a better strategy to overcome the poor outcomes of current 5-FU based therapies. We have previously shown that menadione (vitamin K3) suppresses Wnt signaling pathway in colorectal cancer cells. In the present study, we attempted to evaluate if the combination of menadione and 5-FU can exert a synergistic effect in the human colorectal cancer cell lines HCT116 and SW620. Analysis of the cell viability assay results using CompuSyn software confirmed that the combination indeed acts synergistically. The combination efficiently downregulated the expression of the Wnt signal transducer protein, β-catenin, and the expression of other proteins in the pathway such as phospho-GSK3β in the cell lines HCT116 and SW620. Docking of the individual drugs and their combination on β-catenin using Autodock Vina showed that 5-FU and menadione in combination bind to key protein-protein interaction (PPI) sites of β-catenin whereby they may inhibit β-catenin-TCF interaction or activate the β-catenin destruction complex. Taken together, our data possibly explains that the synergistic action of 5-FU and menadione in colorectal cancer cells may be due to the simultaneous binding of the two drugs to multiple PPI sites thereby blocking the activity of β-catenin. Citation Format: Vidya P Warrier, Sankaran Venkatachalam, Sakthivel Ramasamy, M Michael Gromiha, Devarajan Karunagaran. Combination of 5-fluorouracil and Menadione Exerts Synergistic Effect on Colorectal Cancer Cells by Targeting Wnt/β-catenin Signaling Pathway [abstract]. In: Proceedings of Frontiers in Cancer Science; 2023 Nov 6-8; Singapore. Philadelphia (PA): AACR; Cancer Res 2024;84(8_Suppl):Abstract nr P52.

Frustraevo: a web server to localize and quantify the conservation of local energetic frustration in protein families.

According to the Principle of Minimal Frustration, folded proteins canonly have a minimal number of strong energetic conflicts in their native states. However, not all interactions are energetically optimized for folding but some remain in energetic conflict, i.e. they are highly frustrated. This remaining local energetic frustration has been shown to be statistically correlated with distinct functional aspects such as protein-protein interaction sites, allosterism and catalysis. Fuelled by the recent breakthroughs in efficient protein structure prediction that have made available good quality models for most proteins, we have developed a strategy to calculate local energetic frustration within large protein families and quantify its conservation over evolutionary time. Based on this evolutionary information we can identify how stability and functional constraints have appeared at the common ancestor of the family and have been maintained over the course of evolution. Here, we present FrustraEvo, a web server tool to calculate and quantify the conservation of local energetic frustration in protein families.

5-Aminothiazoles Reveal a New Ligand-Binding Site on Prolyl Oligopeptidase Which is Important for Modulation of Its Protein-Protein Interaction-Derived Functions.

A series of novel 5-aminothiazole-based ligands for prolyl oligopeptidase (PREP) comprise selective, potent modulators of the protein-protein interaction (PPI)-mediated functions of PREP, although they are only weak inhibitors of the proteolytic activity of PREP. The disconnected structure-activity relationships are significantly more pronounced for the 5-aminothiazole-based ligands than for the earlier published 5-aminooxazole-based ligands. Furthermore, the stability of the 5-aminothiazole scaffold allowed exploration of wider substitution patterns than that was possible with the 5-aminooxazole scaffold. The intriguing structure-activity relationships for the modulation of the proteolytic activity and PPI-derived functions of PREP were elaborated by presenting a new binding site for PPI modulating PREP ligands, which was initially discovered using molecular modeling and later confirmed through point mutation studies. Our results suggest that this new binding site on PREP is clearly more important than the active site of PREP for the modulation of its PPI-mediated functions.

Abstract 6239: Human computer interface guided approach to accelerate anticancer drug discovery: Application to human estrogen receptor interactome

Abstract Structure-based drug discovery of target specific drugs greatly rely on the existence of high resolution X-ray crystal structure of the target proteins. Flexible and dynamic regions including hinges and loops which constitute major protein-protein interaction sites as well as allosteric sites are often beyond the scope of current tools and techniques available for protein folding and modeling. Human ability in recognizing patterns and to solve complex puzzles are far superior to any existing computer program at folding these atypical regions of proteins. We have utilized an unconventional combination of using dynamic three-dimensional protein models as physical human computer interface (HCI) devices and integrated proteomics data to predict flexible and dynamic protein-protein interfaces and allosteric pockets of key regulatory proteins to accelerate compound discovery. To this end, we have successfully utilized 1) Flexible HCI devices to generate an ensemble of dynamic three dimensional structures which includes a subset of biologically active conformations among others (thereby exploring the viable chemical space) and 2) Structure refinement and efficient filtering of biologically active conformations can be accomplished by integrating protein-protein interface and fold proteomics data. Here we present AI ready, HCI guided exploration of human estrogen receptor interactome and early in vitro confirmatory analyses. AI approachs to predict 3D protein structures from amino acid sequences like AlphaFold are often on par with experimentally generated structures and AI-guided structure prediction or docking approaches have been reported to significantly reduce computer time utilization. Streamlining of HCIguided tools to enable access to dynamic druggable pockets in protein targets will accelerate drug discovery. Citation Format: Rajendram V. Rajnarayanan, Aya Alsabagh. Human computer interface guided approach to accelerate anticancer drug discovery: Application to human estrogen receptor interactome [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6239.

Structure-based drug design and molecular dynamics studies of an allosteric modulator targeting the protein-protein interaction site of PDK1.

Protein-protein interaction interfaces play a major role in cell signaling pathways. There is always agreat interest in developing protein-protein interaction (PPI) inhibitors of kinases, as they are challenging due to their hydrophobicity, flat surface, specificity, potency, etc. 3 Phosphoinositide-dependent kinase-1 (PDK1), which is involved in the PI3K/PDK1/AKT pathway, is a cancer target that has gained insight for the past two decades. PDK1 possesses a protein interaction fragment (PIF) pocket, which is a well-known PPI that targets allosteric modulators. This work focusses on energy-based pharmacophore model development which on virtual screening could yield novel scaffolds towards the drug designing objective for the kind of PDK1 modulators. A novel pyrazolo pyridine molecule was identified as an allosteric modulator that binds to the PPI site. The metadynamics simulations with free energy profiles further revealed the conformational allosteric changes stimulated on the protein structure upon ligand binding. The cytotoxic activity (IC50-20 μM) of the identified compound against five different cancer cell lines and cell cycle analysis supported the anticancer activity of the identified compound. All the computational works were carried out by the most commonly used Schrodinger Suite software. The pharmacophore was validated by Receiver Operation Characteristics (ROC) and screening against allosteric Enamine database library. The Optimized Potential Liquid Simulations (OPLS-2005) was used to minimize the structures for molecular docking calculations, and inbuilt scoring method of ranking the compounds based on docking score and Glide energy was used. Molecular dynamics simulations were conducted by Desmond implemented in Maestro. The hit compound was purchased from Enamine and tested against different cancer cell lines by MTT assay, apoptosis by western blotting technique, and by flow cytometry analysis.

Analysis of the Healthy Platelet Proteome Identifies a New Form of Domain-Specific O-Fucosylation

Platelet activation induces the secretion of proteins that promote platelet aggregation and inflammation. However, detailed analysis of the released platelet proteome is hampered by platelets’ tendency to preactivate during their isolation and a lack of sensitive protocols for low abundance releasate analysis. Here, we detail the most sensitive analysis to date of the platelet releasate proteome with the detection of >1300 proteins. Unbiased scanning for posttranslational modifications within releasate proteins highlighted O-glycosylation as being a major component. For the first time, we detected O-fucosylation on previously uncharacterized sites including multimerin-1 (MMRN1), a major alpha granule protein that supports platelet adhesion to collagen and is a carrier for platelet factor V. The N-terminal elastin microfibril interface (EMI) domain of MMRN1, a key site for protein–protein interaction, was O-fucosylated at a conserved threonine within a new domain context. Our data suggest that either protein O-fucosyltransferase 1, or a novel protein O-fucosyltransferase, may be responsible for this modification. Mutating this O-fucose site on the EMI domain led to a >50% reduction of MMRN1 secretion, supporting a key role of EMI O-fucosylation in MMRN1 secretion. By comparing releasates from resting and thrombin-treated platelets, 202 proteins were found to be significantly released after high-dose thrombin stimulation. Complementary quantification of the platelet lysates identified >3800 proteins, which confirmed the platelet origin of releasate proteins by anticorrelation analysis. Low-dose thrombin treatment yielded a smaller subset of significantly regulated proteins with fewer secretory pathway enzymes. The extensive platelet proteome resource provided here (larancelab.com/platelet-proteome) allows identification of novel regulatory mechanisms for drug targeting to address platelet dysfunction and thrombosis.

RGCNPPIS: A Residual Graph Convolutional Network for Protein-Protein Interaction Site Prediction.

Accurate identification of protein-protein interaction (PPI) sites is crucial for understanding the mechanisms of biological processes, developing PPI networks, and detecting protein functions. Currently, most computational methods primarily concentrate on sequence context features and rarely consider the spatial neighborhood features. To address this limitation, we propose a novel residual graph convolutional network for structure-based PPI site prediction (RGCNPPIS). Specifically, we use a GCN module to extract the global structural features from all spatial neighborhoods, and utilize the GraphSage module to extract local structural features from local spatial neighborhoods. To the best of our knowledge, this is the first work utilizing local structural features for PPI site prediction. We also propose an enhanced residual graph connection to combine the initial node representation, local structural features, and the previous GCN layer's node representation, which enables information transfer between layers and alleviates the over-smoothing problem. Evaluation results demonstrate that RGCNPPIS outperforms state-of-the-art methods on three independent test sets. In addition, the results of ablation experiments and case studies confirm that RGCNPPIS is an effective tool for PPI site prediction.

Base-editing mutagenesis maps alleles to tune human T cell functions.

CRISPR-enabled screening is a powerful tool for the discovery of genes that control T cell function and has nominated candidate targets for immunotherapies1-6. However, new approaches are required to probe specific nucleotide sequences within key genes. Systematic mutagenesis in primary human T cells could reveal alleles that tune specific phenotypes. DNA base editors are powerful tools for introducing targeted mutations with high efficiency7,8. Here we develop a large-scale base-editing mutagenesis platform with the goal of pinpointing nucleotides that encode amino acid residues that tune primary human T cell activation responses. We generated a library of around 117,000 single guide RNA molecules targeting base editors to protein-coding sites across 385 genes implicated in T cell function and systematically identified protein domains and specific amino acid residues that regulate T cell activation and cytokine production. We found a broad spectrum of alleles with variants encoding critical residues in proteins including PIK3CD, VAV1, LCP2, PLCG1 and DGKZ, including both gain-of-function and loss-of-function mutations. We validated the functional effects of many alleles and further demonstrated that base-editing hits could positively and negatively tune T cell cytotoxic function. Finally, higher-resolution screening using a base editor with relaxed protospacer-adjacent motif requirements9 (NG versus NGG) revealed specific structural domains and protein-protein interaction sites that can be targeted to tune T cell functions. Base-editing screens in primary immune cells thus provide biochemical insights with thepotential to accelerate immunotherapy design.

AAp-MSMD: Amino Acid Preference Mapping on Protein-Protein Interaction Surfaces Using Mixed-Solvent Molecular Dynamics.

Peptides have attracted much attention recently owing to their well-balanced properties as drugs against protein-protein interaction (PPI) surfaces. Molecular simulation-based predictions of binding sites and amino acid residues with high affinity to PPI surfaces are expected to accelerate the design of peptide drugs. Mixed-solvent molecular dynamics (MSMD), which adds probe molecules or fragments of functional groups as solutes to the hydration model, detects the binding hotspots and cryptic sites induced by small molecules. The detection results vary depending on the type of probe molecule; thus, they provide important information for drug design. For rational peptide drug design using MSMD, we proposed MSMD with amino acid residue probes, named amino acid probe-based MSMD (AAp-MSMD), to detect hotspots and identify favorable amino acid types on protein surfaces to which peptide drugs bind. We assessed our method in terms of hotspot detection at the amino acid probe level and binding free energy prediction with amino acid probes at the PPI site for the complex structure that formed the PPI. In hotspot detection, the max-spatial probability distribution map (max-PMAP) obtained from AAp-MSMD detected the PPI site, to which each type of amino acid can bind favorably. In the binding free energy prediction using amino acid probes, ΔGFE obtained from AAp-MSMD roughly estimated the experimental binding affinities from the structure-activity relationship. AAp-MSMD, with amino acid probes, provides estimated binding sites and favorable amino acid types at the PPI site of a target protein.

GHGPR-PPIS: A graph convolutional network for identifying protein-protein interaction site using heat kernel with Generalized PageRank techniques and edge self-attention feature processing block

Accurately pinpointing protein-protein interaction site (PPIS) on the molecular level is of utmost significance for annotating protein function and comprehending the mechanisms underpinning various diseases. While numerous computational methods for predicting PPIS have emerged, they have indeed mitigated the labor and time constraints associated with traditional experimental methods. However, the predictive accuracy of these methods has yet to reach the desired threshold. In this context, we proposed a groundbreaking graph-based computational model called GHGPR-PPIS. This innovative model leveraged a graph convolutional network using heat kernel (GraphHeat) in conjunction with Generalized PageRank techniques (GHGPR) to predict PPIS. Additionally, building upon the GHGPR framework, we devised an edge self-attention feature processing block, further augmenting the performance of the model. Experimental findings conclusively demonstrated that GHGPR-PPIS surpassed all competing state-of-the-art models when evaluated on the benchmark test set. Impressively, on two distinct independent test sets and a specific protein chain, GHGPR-PPIS consistently demonstrated superior generalization performance and practical applicability compared to the comparative model, AGAT-PPIS. Lastly, leveraging the t-SNE dimensionality reduction algorithm and clustering visualization technique, we delved into an interpretability analysis of the effectiveness of GHGPR-PPIS by meticulously comparing the outputs from different stages of the model.

A CNN-LSTM Ensemble Model for Predicting Protein-Protein Interaction Binding Sites.

Proteins commonly perform biological functions through protein-protein interactions (PPIs). The knowledge of PPI sites is imperative for the understanding of protein functions, disease mechanisms, and drug design. Traditional biological experimental methods for studying PPI sites still incur considerable drawbacks, including long experimental time and high labor costs. Therefore, many computational methods have been proposed for predicting PPI sites. However, achieving high prediction performance and overcoming severe data imbalance remain challenging issues. In this paper, we propose a new sequence-based deep learning model called CLPPIS (standing for CNN-LSTM ensemble based PPI Sites prediction). CLPPIS consists of CNN and LSTM components, which can capture spatial features and sequential features simultaneously. Further, it utilizes a novel feature group as input, which has 7 physicochemical, biophysical, and statistical properties. Besides, it adopts a batch-weighted loss function to reduce the interference of imbalance data. Our work suggests that the integration of protein spatial features and sequential features provides important information for PPI sites prediction. Evaluation on three public benchmark datasets shows that our CLPPIS model significantly outperforms existing state-of-the-art methods.

Protein-Protein Interaction Site Prediction Based on Attention Mechanism and Convolutional Neural Networks.

Proteins usually perform their cellular functions by interacting with other proteins. Accurate identification of protein-protein interaction sites (PPIs) from sequence is import for designing new drugs and developing novel therapeutics. A lot of computational models for PPIs prediction have been developed because experimental methods are slow and expensive. Most models employ a sliding window approach in which local neighbors are concatenated to present a target residue. However, those neighbors are not distinguished by pairwise information between a neighbor and the target. In this study, we propose a novel PPIs prediction model AttCNNPPISP, which combines attention mechanism and convolutional neural networks (CNNs). The attention mechanism dynamically captures the pairwise correlation of each neighbor-target pair within a sliding window, and therefore makes a better understanding of the local environment of target residue. And then, CNNs take the local representation as input to make prediction. Experiments are employed on several public benchmark datasets. Compared with the state-of-the-art models, AttCNNPPISP improves the prediction performance. Also, the experimental results demonstrate that the attention mechanism is effective in terms of constructing comprehensive context information of target residue.

SaLT&PepPr is an interface-predicting language model for designing peptide-guided protein degraders

Protein-protein interactions (PPIs) are critical for biological processes and predicting the sites of these interactions is useful for both computational and experimental applications. We present a Structure-agnostic Language Transformer and Peptide Prioritization (SaLT&PepPr) pipeline to predict interaction interfaces from a protein sequence alone for the subsequent generation of peptidic binding motifs. Our model fine-tunes the ESM-2 protein language model (pLM) with a per-position prediction task to identify PPI sites using data from the PDB, and prioritizes motifs which are most likely to be involved within inter-chain binding. By only using amino acid sequence as input, our model is competitive with structural homology-based methods, but exhibits reduced performance compared with deep learning models that input both structural and sequence features. Inspired by our previous results using co-crystals to engineer target-binding “guide” peptides, we curate PPI databases to identify partners for subsequent peptide derivation. Fusing guide peptides to an E3 ubiquitin ligase domain, we demonstrate degradation of endogenous β-catenin, 4E-BP2, and TRIM8, and highlight the nanomolar binding affinity, low off-targeting propensity, and function-altering capability of our best-performing degraders in cancer cells. In total, our study suggests that prioritizing binders from natural interactions via pLMs can enable programmable protein targeting and modulation.

B-291 Proximity Labeling at Sites of Protein-Protein Interactions Using a Switchable DNA Catalyst

Abstract Background Extracellular protein-protein interactions (PPIs) play a crucial role in cell-cell adhesion, communication, and signaling. The dysregulation of PPIs in cancer caused by differential protein expression can drive oncogenic factors. Traditional methods to study PPIs disfavor weak or transient interacting protein partners, leaving them undiscovered. Proximity labeling (PL) is an emerging tool in which a catalyst is attached to a protein of interest (POI) and creates reactive species that tag nearby endogenous proteins. Tagged proteins are subsequently isolated and identified as potential interacting partners with the POI. If PL could be activated at a more specific subcellular location, such as the site of a PPI, more detailed information could be unveiled concerning the interactome of cancer-relevant PPIs. Recent advances in PL involve the use of a synthetic catalyst that can be localized to a POI via an antibody or ligand and avoids the need for PL enzyme-POI fusion proteins to be constructed, allowing for analysis of primary tissue samples. Methods To create a PL catalyst with pseudo-allosteric regulation, we created a switchable DNA catalyst. The DNA catalyst is composed of synthetic photocatalyst and quencher moieties covalently linked to DNA oligomers such that when hybridized together the photocatalyst is quenched and inactive. The DNA scaffold affords conformational control that is utilized to change the distance between photocatalyst and quencher, thus regulating photocatalyst activity. Split triggers composed of DNA oligos that bind to catalyst-DNA and simultaneously disrupt quencher-DNA hybridization were used to create an AND logic gate where both split triggers need to be in proximity to activate the PL catalyst. The split triggers were localized to POIs via connection to aptamers, short pieces of synthetic DNA that bind targets. Using the c-Met homodimer as a model system, we first labeled c-Met expressing HEK293T cells with aptamer-split trigger DNA oligomers as well as the quenched photocatalyst. After the catalyst was activated at the site of the PPI, we performed proximity labeling using a biotin-phenol substrate in the presence of an oxidizing agent and 450 nm light. Fluorescent streptavidin was used to visualize sites of biotinylation and detected using fluorescence microscopy. Results We observed extracellular biotin signal only of c-Met transfected cells as evidenced by the expression of a co-transfection plasmid encoding a fluorescent nuclear protein. When one of the split triggers had the aptamer motif removed, meaning it couldn’t be localized to the cell surface, no labeling was observed which shows that both split probes must be localized to the cell surface to activate labeling. Additionally, controls including omission of light, photocatalyst, or biotin phenol showed no labeling. Using a free-floating version of the photocatalyst to label cells, we confirmed that lysates contained biotinylated proteins via western blot. Conclusion We localized two split triggers to interacting proteins to activate a PL catalyst and performed proximity labeling at the site of the PPI. This technique demonstrates non-genetic split proximity labeling to enable downstream analysis of cancer cell proteomes in an ultra-specific manner and expands the toolkit for spatiotemporally controlled proximity labeling.