A Structurally-Validated Multiple Sequence Alignment of 497 Human Protein Kinase Domains

Druggable cavities and allosteric modulators of the cell division cycle 7 (CDC7) kinase

Cell division cycle 7 kinase (CDC7) has been found overexpressed in many cancer cell lines being also one of the kinases involved in the nuclear protein TDP-43 phosphorylation in vivo. Thus, inhibitors of CDC7 are emerging drug candidates for the treatment of oncological and neurodegenerative unmet diseases. All the known CDC7 inhibitors are ATP-competitives, lacking of selectivity enough for success in clinical trials. As allosteric sites are less conserved among kinase proteins, discovery of allosteric modulators of CDC7 is a great challenge and opportunity in this field.Using different computational approaches, we have here identified new druggable cavities on the human CDC7 structure and subsequently selective CDC7 inhibitors with allosteric modulation mainly targeting the pockets where the interaction between this kinase and its activator DBF4 takes place.

Evolutionary sequence and structural basis for the distinct conformational landscapes of Tyr and Ser/Thr kinases

Protein kinases are molecular machines with rich sequence variation that distinguishes the two main evolutionary branches - tyrosine kinases (TKs) from serine/threonine kinases (STKs). Using a sequence co-variation Potts statistical energy model we previously concluded that TK catalytic domains are more likely than STKs to adopt an inactive conformation with the activation loop in an autoinhibitory "folded" conformation, due to intrinsic sequence effects. Here we investigated the structural basis for this phenomenon by integrating the sequence-based model with structure-based molecular dynamics (MD) to determine the effects of mutations on the free energy difference between active and inactive conformations, using a novel thermodynamic cycle involving many (n=108) protein-mutation free energy perturbation (FEP) simulations in the active and inactive conformations. The sequence and structure-based results are consistent and support the hypothesis that the inactive conformation "DFG-out Activation Loop Folded", is a functional regulatory state that has been stabilized in TKs relative to STKs over the course of their evolution via the accumulation of residue substitutions in the activation loop and catalytic loop that facilitate distinct substrate binding modes in trans and additional modes of regulation in cis for TKs.

Evolutionary sequence and structural basis for the distinct conformational landscapes of Tyr and Ser/Thr kinases

Protein kinases are molecular machines with rich sequence variation that distinguishes the two main evolutionary branches - tyrosine kinases (TKs) from serine/threonine kinases (STKs). Using a sequence co-variation Potts statistical energy model we previously concluded that TK catalytic domains are more likely than STKs to adopt an inactive conformation with the activation loop in an autoinhibitory "folded" conformation, due to intrinsic sequence effects. Here we investigated the structural basis for this phenomenon by integrating the sequence-based model with structure-based molecular dynamics (MD) to determine the effects of mutations on the free energy difference between active and inactive conformations, using a novel thermodynamic cycle involving many (n=108) protein-mutation free energy perturbation (FEP) simulations in the active and inactive conformations. The sequence and structure-based results are consistent and support the hypothesis that the inactive conformation "DFG-out Activation Loop Folded", is a functional regulatory state that has been stabilized in TKs relative to STKs over the course of their evolution via the accumulation of residue substitutions in the activation loop and catalytic loop that facilitate distinct substrate binding modes in trans and additional modes of regulation in cis for TKs.

The intrinsic substrate specificity of the human tyrosine kinome

Phosphorylation of proteins on tyrosine (Tyr) residues evolved in metazoan organisms as a mechanism of coordinating tissue growth1. Multicellular eukaryotes typically have more than 50 distinct protein Tyr kinases that catalyse the phosphorylation of thousands of Tyr residues throughout the proteome1-3. How a given Tyr kinase can phosphorylate a specific subset of proteins at unique Tyr sites is only partially understood4-7. Here we used combinatorial peptide arrays to profile the substrate sequence specificity of all human Tyr kinases. Globally, the Tyr kinases demonstrate considerable diversity in optimal patterns of residues surrounding the site of phosphorylation, revealing the functional organization of the human Tyr kinome by substrate motif preference. Using this information, Tyr kinases that are most compatible with phosphorylating any Tyr site can be identified. Analysis of mass spectrometry phosphoproteomic datasets using this compendium of kinase specificities accurately identifies specific Tyr kinases that are dysregulated in cells after stimulation with growth factors, treatment with anti-cancer drugs or expression of oncogenic variants. Furthermore, the topology of known Tyr signalling networks naturally emerged from a comparison of the sequence specificities of the Tyr kinases and the SH2 phosphotyrosine (pTyr)-binding domains. Finally we show that the intrinsic substrate specificity of Tyr kinases has remained fundamentally unchanged from worms to humans, suggesting that the fidelity between Tyr kinases and their protein substrate sequences has been maintained across hundreds of millions of years of evolution.

Multiplexed, multimodal profiling of the intracellular activity, interactions, and druggability of protein variants using LABEL-seq

Multiplexed assays of variant effect are powerful tools for assessing the impact of protein sequence variation, but are limited to measuring a single protein property and often rely on indirect readouts of intracellular protein function. Here, we developed LAbeling with Barcodes and Enrichment for biochemicaL analysis by sequencing (LABEL-seq), a platform for the multimodal profiling of thousands of protein variants in cultured human cells. Multimodal measurement of ~20,000 variant effects for ~1,600 BRaf variants using LABEL-seq revealed that variation at positions that are frequently mutated in cancer had minimal effects on folding and intracellular abundance but could dramatically alter activity, protein-protein interactions, and druggability. Integrative analysis of our multimodal measurements identified networks of positions with similar roles in regulating BRaf's signaling properties and enabled predictive modeling of variant effects on complex processes such as cell proliferation and small molecule-promoted degradation. LABEL-seq provides a scalable approach for the direct measurement of multiple biochemical effects of protein variants in their native cellular context, yielding insight into protein function, disease mechanisms, and druggability.

Machine Learning Models to Interrogate Proteome-Wide Covalent Ligandabilities Directed at Cysteines

Machine learning (ML) identification of covalently ligandable sites may accelerate targeted covalent inhibitor design and help expand the druggable proteome space. Here, we report the rigorous development and validation of the tree-based models and convolutional neural networks (CNNs) trained on a newly curated database (LigCys3D) of over 1000 liganded cysteines in nearly 800 proteins represented by over 10,000 three-dimensional structures in the protein data bank. The unseen tests yielded 94 and 93% area under the receiver operating characteristic curves for the tree models and CNNs, respectively. Based on the AlphaFold2 predicted structures, the ML models recapitulated the newly liganded cysteines in the PDB with over 90% recall values. To assist the community of covalent drug discoveries, we report the predicted ligandable cysteines in 392 human kinases and their locations in the sequence-aligned kinase structure, including the PH and SH2 domains. Furthermore, we disseminate a searchable online database LigCys3D (https://ligcys.computchem.org/) and a web prediction server DeepCys (https://deepcys.computchem.org/), both of which will be continuously updated and improved by including newly published experimental data. The present work represents an important step toward the ML-led integration of big genome data and structure models to annotate the human proteome space for the next-generation covalent drug discoveries.

Identification and structural characterization of small molecule inhibitors of PINK1

Mutations in PINK1 and Parkin cause early-onset Parkinson's Disease (PD). PINK1 is a kinase which functions as a mitochondrial damage sensor and initiates mitochondrial quality control by accumulating on the damaged organelle. There, it phosphorylates ubiquitin, which in turn recruits and activates Parkin, an E3 ubiquitin ligase. Ubiquitylation of mitochondrial proteins leads to the autophagic degradation of the damaged organelle. Pharmacological modulation of PINK1 constitutes an appealing avenue to study its physiological function and develop therapeutics. In this study, we used a thermal shift assay with insect PINK1 to identify small molecules that inhibit ATP hydrolysis and ubiquitin phosphorylation. PRT062607, an SYK inhibitor, is the most potent inhibitor in our screen and inhibits both insect and human PINK1, with an IC50 in the 0.5-3 µM range in HeLa cells and dopaminergic neurons. The crystal structures of insect PINK1 bound to PRT062607 or CYC116 reveal how the compounds interact with the ATP-binding pocket. PRT062607 notably engages with the catalytic aspartate and causes a destabilization of insert-2 at the autophosphorylation dimer interface. While PRT062607 is not selective for PINK1, it provides a scaffold for the development of more selective and potent inhibitors of PINK1 that could be used as chemical probes.

Molecular basis for differential recognition of an allosteric inhibitor by receptor tyrosine kinases

Understanding kinase-inhibitor selectivity continues to be a major objective in kinase drug discovery. We probe the molecular basis of selectivity of an allosteric inhibitor (MSC1609119A-1) of the insulin-like growth factor-I receptor kinase (IGF1RK), which has been shown to be ineffective for the homologous insulin receptor kinase (IRK). Specifically, we investigated the structural and energetic basis of the allosteric binding of this inhibitor to each kinase by combining molecular modeling, molecular dynamics (MD) simulations, and thermodynamic calculations. We predict the inhibitor conformation in the binding pocket of IRK and highlight that the charged residues in the histidine-arginine-aspartic acid (HRD) and aspartic acid-phenylalanine-glycine (DFG) motifs and the nonpolar residues in the binding pocket govern inhibitor interactions in the allosteric pocket of each kinase. We suggest that the conformational changes in the IGF1RK residues M1054 and M1079, movement of the ⍺C-helix, and the conformational stabilization of the DFG motif favor the selectivity of the inhibitor toward IGF1RK. Our thermodynamic calculations reveal that the observed selectivity can be rationalized through differences observed in the electrostatic interaction energy of the inhibitor in each inhibitor/kinase complex and the hydrogen bonding interactions of the inhibitor with the residue V1063 in IGF1RK that are not attained with the corresponding residue V1060 in IRK. Overall, our study provides a rationale for the molecular basis of recognition of this allosteric inhibitor by IGF1RK and IRK, which is potentially useful in developing novel inhibitors with improved affinity and selectivity.

Global and local functions of the Fused kinase ortholog CdaH in intracellular patterning in Tetrahymena

Ciliates assemble numerous microtubular structures into complex cortical patterns. During ciliate division, the pattern is duplicated by intracellular segmentation that produces a tandem of daughter cells. In Tetrahymena thermophila, the induction and positioning of the division boundary involves two mutually antagonistic factors: posterior CdaA (cyclin E) and anterior CdaI (Hippo kinase). Here, we characterized the related cdaH-1 allele, which confers a pleiotropic patterning phenotype including an absence of the division boundary and an anterior-posterior mispositioning of the new oral apparatus. CdaH is a Fused or Stk36 kinase ortholog that localizes to multiple sites that correlate with the effects of its loss, including the division boundary and the new oral apparatus. CdaH acts downstream of CdaA to induce the division boundary and drives asymmetric cytokinesis at the tip of the posterior daughter. CdaH both maintains the anterior-posterior position of the new oral apparatus and interacts with CdaI to pattern ciliary rows within the oral apparatus. Thus, CdaH acts at multiple scales, from induction and positioning of structures on the cell-wide polarity axis to local organelle-level patterning.

PL-PatchSurfer3: Improved Structure-Based Virtual Screening for Structure Variation Using 3D Zernike Descriptors

Structure-based virtual screening (SBVS) is a widely used method in silico drug discovery. It necessitates a receptor structure or binding site to predict the binding pose and fitness of a ligand. Therefore, the performance of the SBVS is affected by the protein conformation. The most frequently used method in SBVS is the protein-ligand docking program, which utilizes atomic distance-based scoring functions. Hence, they are highly prone to sensitivity towards variation in receptor structure, and it is reported that the conformational change significantly drops the performance of the docking program. To address the problem, we have introduced a novel program of SBVS, named PL-PatchSurfer. This program makes use of molecular surface patches and the Zernike descriptor. The surfaces of the pocket and ligand are segmented into several patches by the program. These patches are then mapped with physico-chemical properties such as shape and electrostatic potential before being converted into the Zernike descriptor, which is rotationally invariant. A complementarity between the protein and the ligand is assessed by comparing the descriptors and geometric distribution of the patches in the molecules. A benchmarking study showed that PL-PatchSurfer2 was able to screen active molecules regardless of the receptor structure change with fast speed. However, the program could not achieve high performance for the targets that the hydrogen bonding feature is important such as nuclear hormone receptors. In this paper, we present the newer version of PL-PatchSurfer, PL-PatchSurfer3, which incorporates two new features: a change in the definition of hydrogen bond complementarity and consideration of visibility that contains curvature information of a patch. Our evaluation demonstrates that the new program outperforms its predecessor and other SBVS methods while retaining its characteristic tolerance to receptor structure changes. Interested individuals can access the program at kiharalab.org/plps3.

Contrasting Effects of Cancer-Associated Mutations in EphA3 and EphB2 Kinases

Erythropoietin-producing hepatoma (Eph) receptors are a family of tyrosine kinases that can act as tumor promoters or tumor suppressors, depending on the receptor and cancer cell type. Cancer-associated somatic mutations have been identified in all Eph receptors, but in most cases, the functional effects of the mutations are unknown. In this study, we expressed and purified the kinase domains of wild-type (WT) EphA3 and EphB2 along with 16 cancer-associated mutants. We identified mutations that decrease EphA3 activity and both activating and inhibitory mutations in EphB2. To shed light on the mechanisms by which the mutations altered kinase activity, we measured the thermal stabilities of the enzymes and performed steady-state kinetic experiments. We also expressed the full-length receptors in HEK293T cells to determine the cellular effects. WT EphB2 promoted downstream ERK signaling, while a kinase-inactive mutant (S706F) was similar to the control cells. In contrast, WT EphA3 (but not loss-of-function mutants) inhibited ERK signaling. The reciprocal effects of EphB2 and EphA3 on ERK phosphorylation in HEK293T cells were also evident in Ras-GTP loading. Thus, consistent with the dual roles of Eph receptors as tumor promoters and tumor suppressors, somatic mutations have the potential to increase or decrease Eph function, resulting in changes in the downstream signaling transduction.

A covalent fragment-based strategy targeting a novel cysteine to inhibit activity of mutant EGFR kinase

Several generations of ATP-competitive anti-cancer drugs that inhibit the activity of the intracellular kinase domain of the epidermal growth factor receptor (EGFR) have been developed over the past twenty years. The first-generation of drugs such as gefitinib bind reversibly and were followed by a second-generation such as dacomitinib that harbor an acrylamide moiety that forms a covalent bond with C797 in the ATP binding pocket. Resistance emerges through mutation of the T790 gatekeeper residue to methionine, which introduces steric hindrance to drug binding and increases the Km for ATP. A third generation of drugs, such as osimertinib were developed which were effective against T790M EGFR in which an acrylamide moiety forms a covalent bond with C797, although resistance has emerged by mutation to S797. A fragment-based screen to identify new starting points for an EGFR inhibitor serendipitously identified a fragment that reacted with C775, a previously unexploited residue in the ATP binding pocket for a covalent inhibitor to target. A number of acrylamide containing fragments were identified that selectively reacted with C775. One of these acrylamides was optimized to a highly selective inhibitor with sub-1 μM activity, that is active against T790M, C797S mutant EGFR independent of ATP concentration, providing a potential new strategy for pan-EGFR mutant inhibition.

Construction of a hepatocytes-related and protein kinase-related gene signature in HCC based on ScRNA-Seq analysis and machine learning algorithm

With recent advancements in single-cell sequencing and machine learning methods, new insights into hepatocellular carcinoma (HCC) progression have been provided. Protein kinase-related genes (PKRGs) affect cell growth, differentiation, apoptosis, and signaling during HCC progression, making the predictive relevance of PKRGs in HCC highly necessary for personalized medicine. In this study, we analyzed single-cell data of HCC and used the machine learning method of LASSO regression to construct PKRG prediction models in six major cell types. CDK4 and AURKB were found to be the best PKRG prognostic signature for predicting the overall survival of HCC patients (including TCGA, ICGC, and GEO datasets) in hepatocytes. Independent clinical factors were further screened out using the COX regression method, and a nomogram combining PKRGs and cancer status was created. Treatment with Palbociclib (CDK4 Inhibitor) and Barasertib (AURKB Inhibitor) inhibited HCC cell migration. Patients classified as PKRG high- or low-risk groups showed different tumor mutation burdens, immune infiltrations, and gene enrichment. The PKRG high-risk group showed higher tumor mutation burdens and gene set enrichment analysis indicated that cell cycle, base excision repair, and RNA degradation pathways were more enriched in these patients. Additionally, the PKRG high-risk group demonstrated higher infiltration levels of Naïve CD8+ T cells, Endothelial cells, M2 macrophage, and Tregs than the low-risk group. In summary, this study established the hepatocytes-related PKRG signature for prognostic stratification at the single-cell level by using machine learning algorithms in HCC and identified potential HCC treatment targets based on the PKRG signature.

Affimer-mediated locking of p21-activated kinase 5 in an intermediate activation state results in kinase inhibition

Kinases are important therapeutic targets, and their inhibitors are classified according to their mechanism of action, which range from blocking ATP binding to covalent inhibition. Here, a mechanism of inhibition is highlighted by capturing p21-activated kinase 5 (PAK5) in an intermediate state of activation using an Affimer reagent that binds in the P+1 pocket. PAK5 was identified from a non-hypothesis-driven high-content imaging RNAi screen in urothelial cancer cells. Silencing of PAK5 resulted in reduced cell number, G1/S arrest, and enlargement of cells, suggesting it to be important in urothelial cancer cell line survival and proliferation. Affimer reagents were isolated to identify mechanisms of inhibition. The Affimer PAK5-Af17 recapitulated the phenotype seen with siRNA. Co-crystallization revealed that PAK5-Af17 bound in the P+1 pocket of PAK5, locking the kinase into a partial activation state. This mechanism of inhibition indicates that another class of kinase inhibitors is possible.

Computational analysis of regulatory regions in human protein kinases

Eukaryotic proteins often feature modular domain structures comprising globular domains that are connected by linker regions and intrinsically disordered regions that may contain important functional motifs. The intramolecular interactions of globular domains and nonglobular regions can play critical roles in different aspects of protein function. However, studying these interactions and their regulatory roles can be challenging due to the flexibility of nonglobular regions, the long insertions separating interacting modules, and the transient nature of some interactions. Obtaining the experimental structures of multiple domains and functional regions is more difficult than determining the structures of individual globular domains. High-quality structural models generated by AlphaFold offer a unique opportunity to study intramolecular interactions in eukaryotic proteins. In this study, we systematically explored intramolecular interactions between human protein kinase domains (KDs) and potential regulatory regions, including globular domains, N- and C-terminal tails, long insertions, and distal nonglobular regions. Our analysis identified intramolecular interactions between human KDs and 35 different types of globular domains, exhibiting a variety of interaction modes that could contribute to orthosteric or allosteric regulation of kinase activity. We also identified prevalent interactions between human KDs and their flanking regions (N- and C-terminal tails). These interactions exhibit group-specific characteristics and can vary within each specific kinase group. Although long-range interactions between KDs and nonglobular regions are relatively rare, structural details of these interactions offer new insights into the regulation mechanisms of several kinases, such as HASPIN, MAPK7, MAPK15, and SIK1B.

KinScan: AI-based rapid profiling of activity across the kinome

Kinases play a vital role in regulating essential cellular processes, including cell cycle progression, growth, apoptosis, and metabolism, by catalyzing the transfer of phosphate groups from adenosing triphosphate to substrates. Their dysregulation has been closely associated with numerous diseases, including cancer development, making them attractive targets for drug discovery. However, accurately predicting the binding affinity between chemical compounds and kinase targets remains challenging due to the highly conserved structural similarities across the kinome. To address this limitation, we present KinScan, a novel computational approach that leverages large-scale bioactivity data and integrates the Multi-Scale Context Aware Transformer framework to construct a virtual profiling model encompassing 391 protein kinases. The developed model demonstrates exceptional prediction capability, distinguishing between kinases by utilizing structurally aligned kinase binding site features derived from multiple sequence alignment for fast and accurate predictions. Through extensive validation and benchmarking, KinScan demonstrated its robust predictive power and generalizability for large-scale kinome-wide profiling and selectivity, uncovering associations with specific diseases and providing valuable insights into kinase activity profiles of compounds. Furthermore, we deployed a web platform for end-to-end profiling and selectivity analysis, accessible at https://kinscan.drugonix.com/softwares/kinscan.

Architecture of the MKK6-p38α complex defines the basis of MAPK specificity and activation

The mitogen-activated protein kinase (MAPK) p38α is a central component of signaling in inflammation and the immune response and is, therefore, an important drug target. Little is known about the molecular mechanism of its activation by double phosphorylation from MAPK kinases (MAP2Ks), because of the challenge of trapping a transient and dynamic heterokinase complex. We applied a multidisciplinary approach to generate a structural model of p38α in complex with its MAP2K, MKK6, and to understand the activation mechanism. Integrating cryo-electron microscopy with molecular dynamics simulations, hydrogen-deuterium exchange mass spectrometry, and experiments in cells, we demonstrate a dynamic, multistep phosphorylation mechanism, identify catalytically relevant interactions, and show that MAP2K-disordered amino termini determine pathway specificity. Our work captures a fundamental step of cell signaling: a kinase phosphorylating its downstream target kinase.

Investigating the conformational landscape of AlphaFold2-predicted protein kinase structures

Summary

Protein kinases are a family of signaling proteins, crucial for maintaining cellular homeostasis. When dysregulated, kinases drive the pathogenesis of several diseases, and are thus one of the largest target categories for drug discovery. Kinase activity is tightly controlled by switching through several active and inactive conformations in their catalytic domain. Kinase inhibitors have been designed to engage kinases in specific conformational states, where each conformation presents a unique physico-chemical environment for therapeutic intervention. Thus, modeling kinases across conformations can enable the design of novel and optimally selective kinase drugs. Due to the recent success of AlphaFold2 in accurately predicting the 3D structure of proteins based on sequence, we investigated the conformational landscape of protein kinases as modeled by AlphaFold2. We observed that AlphaFold2 is able to model several kinase conformations across the kinome, however, certain conformations are only observed in specific kinase families. Furthermore, we show that the per residue predicted local distance difference test can capture information describing structural flexibility of kinases. Finally, we evaluated the docking performance of AlphaFold2 kinase structures for enriching known ligands. Taken together, we see an opportunity to leverage AlphaFold2 models for structure-based drug discovery against kinases across several pharmacologically relevant conformational states.

Availability and implementation

All code used in the analysis is freely available at https://github.com/Harmonic-Discovery/AF2-kinase-conformational-landscape.

AlphaFold2 models of the active form of all 437 catalytically competent human protein kinase domains

Humans have 437 catalytically competent protein kinase domains with the typical kinase fold, similar to the structure of Protein Kinase A (PKA). Only 155 of these kinases are in the Protein Data Bank in their active form. The active form of a kinase must satisfy requirements for binding ATP, magnesium, and substrate. From structural bioinformatics analysis of 40 unique substrate-bound kinases, we derived several criteria for the active form of protein kinases. We include requirements on the DFG motif of the activation loop but also on the positions of the N-terminal and C-terminal segments of the activation loop that must be placed appropriately to bind substrate. Because the active form of catalytic kinases is needed for understanding substrate specificity and the effects of mutations on catalytic activity in cancer and other diseases, we used AlphaFold2 to produce models of all 437 human protein kinases in the active form. This was accomplished with templates in the active form from the PDB and shallow multiple sequence alignments of orthologs and close homologs of the query protein. We selected models for each kinase based on the pLDDT scores of the activation loop residues, demonstrating that the highest scoring models have the lowest or close to the lowest RMSD to 22 non-redundant substrate-bound structures in the PDB. A larger benchmark of all 130 active kinase structures with complete activation loops in the PDB shows that 80% of the highest-scoring AlphaFold2 models have RMSD < 1.0 Å and 90% have RMSD < 2.0 Å over the activation loop backbone atoms. Models for all 437 catalytic kinases are available at http://dunbrack.fccc.edu/kincore/activemodels. We believe they may be useful for interpreting mutations leading to constitutive catalytic activity in cancer as well as for templates for modeling substrate and inhibitor binding for molecules which bind to the active state.

Exploring the Druggable Conformational Space of Protein Kinases Using AI-Generated Structures

Protein kinase function and interactions with drugs are controlled in part by the movement of the DFG and ɑC-Helix motifs, which enable kinases to adopt various conformational states. Small molecule ligands elicit therapeutic effects with distinct selectivity profiles and residence times that often depend on the kinase conformation(s) they bind. However, the limited availability of experimentally determined structural data for kinases in inactive states restricts drug discovery efforts for this major protein family. Modern AI-based structural modeling methods hold potential for exploring the previously experimentally uncharted druggable conformational space for kinases. Here, we first evaluated the currently explored conformational space of kinases in the PDB and models generated by AlphaFold2 (AF2) (1) and ESMFold (2), two prominent AI-based structure prediction methods. We then investigated AF2's ability to predict kinase structures in different conformations at various multiple sequence alignment (MSA) depths, based on this parameter's ability to explore conformational diversity. Our results showed a bias within the PDB and predicted structural models generated by AF2 and ESMFold toward structures of kinases in the active state over alternative conformations, particularly those conformations controlled by the DFG motif. Finally, we demonstrate that predicting kinase structures using AF2 at lower MSA depths allows the exploration of the space of these alternative conformations, including identifying previously unobserved conformations for 398 kinases. The results of our analysis of structural modeling by AF2 create a new avenue for the pursuit of new therapeutic agents against a notoriously difficult-to-target family of proteins.

Significance Statement

Greater abundance of kinase structural data in inactive conformations, currently lacking in structural databases, would improve our understanding of how protein kinases function and expand drug discovery and development for this family of therapeutic targets. Modern approaches utilizing artificial intelligence and machine learning have potential for efficiently capturing novel protein conformations. We provide evidence for a bias within AlphaFold2 and ESMFold to predict structures of kinases in their active states, similar to their overrepresentation in the PDB. We show that lowering the AlphaFold2 algorithm's multiple sequence alignment depth can help explore kinase conformational space more broadly. It can also enable the prediction of hundreds of kinase structures in novel conformations, many of whose models are likely viable for drug discovery.

Pathogenic mutation hotspots in protein kinase domain structure

Control of eukaryotic cellular function is heavily reliant on the phosphorylation of proteins at specific amino acid residues, such as serine, threonine, tyrosine, and histidine. Protein kinases that are responsible for this process comprise one of the largest families of evolutionarily related proteins. Dysregulation of protein kinase signaling pathways is a frequent cause of a large variety of human diseases including cancer, autoimmune, neurodegenerative, and cardiovascular disorders. In this study, we mapped all pathogenic mutations in 497 human protein kinase domains from the ClinVar database to the reference structure of Aurora kinase A (AURKA) and grouped them by the relevance to the disease type. Our study revealed that the majority of mutation hotspots associated with cancer are situated within the catalytic and activation loops of the kinase domain, whereas non-cancer-related hotspots tend to be located outside of these regions. Additionally, we identified a hotspot at residue R371 of the AURKA structure that has the highest number of exclusively non-cancer-related pathogenic mutations (21) and has not been previously discussed.

Colorectal cancer-associated mutations impair EphB1 kinase function

Erythropoietin-producing hepatoma (Eph) receptor tyrosine kinases regulate the migration and adhesion of cells that are required for many developmental processes and adult tissue homeostasis. In the intestinal epithelium, Eph signaling controls the positioning of cell types along the crypt-villus axis. Eph activity can suppress the progression of colorectal cancer (CRC). The most frequently mutated Eph receptor in metastatic CRC is EphB1. However, the functional effects of EphB1 mutations are mostly unknown. We expressed and purified the kinase domains of WT and five cancer-associated mutant EphB1 and developed assays to assess the functional effects of the mutations. Using purified proteins, we determined that CRC-associated mutations reduce the activity and stability of the folded structure of EphB1. By mammalian cell expression, we determined that CRC-associated mutant EphB1 receptors inhibit signal transducer and activator of transcription 3 and extracellular signal-regulated kinases 1 and 2 signaling. In contrast to the WT, the mutant EphB1 receptors are unable to suppress the migration of human CRC cells. The CRC-associated mutations also impair cell compartmentalization in an assay in which EphB1-expressing cells are cocultured with ligand (ephrin B1)-expressing cells. These results suggest that somatic mutations impair the kinase-dependent tumor suppressor function of EphB1 in CRC.

Comprehensive molecular docking and dynamic simulations for drug repurposing of clinical drugs against multiple cancer kinase targets

Drug repurposing is a method to identify novel therapeutic agents from the existing drugs and clinical compounds. In the present comprehensive work, molecular docking, virtual screening and dynamics simulations were carried out for ten cancer types viz breast, colon, central nervous system, leukaemia, melanoma, ovarian, prostate, renal and lung (non-small and small cell) against validated eighteen kinase targets. The study aims to understand the action of chemotherapy drugs mechanism through binding interactions against selected targets via comparative docking simulations with the state-art molecular modelling suits such as MOE, Cresset-Flare, AutoDock Vina, GOLD and GLIDE. Chemotherapeutic drugs (n = 112) were shortlisted from standard drug databases with appropriate chemoinformatic filters. Based on docking studies it was revealed that leucovorin, nilotinib, ellence, thalomid and carfilzomib drugs possessed potential against other cancer targets. A library was built to enumerate novel molecules based on the scaffold and functional groups extracted from known drugs and clinical compounds. Twenty novel molecules were prioritised further based on drug-like attributes. These were cross docked against 1MQ4 Aurora-A Protein Kinase for prostate cancer and 4UYA Mitogen-activated protein kinase for renal cancer. All docking programs yielded similar results but interestingly AutoDock Vina yielded the lowest RMSD with the native ligand. To further validate the final docking results at atomistic level, molecular dynamics simulations were performed to ascertain the stability of the protein-ligand complex. The study enables repurposing of drugs and lead identification by employing a host of structure and ligand based virtual screening tools and techniques.Communicated by Ramaswamy H. Sarma.

A conserved arginine within the αC-helix of Erk1/2 is a latch of autoactivation and of oncogenic capabilities

Eukaryotic protein kinases (EPKs) adopt an active conformation following phosphorylation of a particular activation loop residue. Most EPKs spontaneously autophosphorylate this residue. While structure-function relationships of the active conformation are essentially understood, those of the "prone-to-autophosphorylate" conformation are unclear. Here, we propose that a site within the αC-helix of EPKs, occupied by Arg in the mitogen-activated protein kinase (MAPK) Erk1/2 (Arg84/65), impacts spontaneous autophosphorylation. MAPKs lack spontaneous autoactivation, but we found that converting Arg84/65 of Erk1/2 to various residues enables spontaneous autophosphorylation. Furthermore, Erk1 molecules mutated in Arg84 are oncogenic. Arg84/65 thus obstructs the adoption of the "prone-to-autophosphorylate" conformation. All MAPKs harbor an Arg that is equivalent to Arg84/65 of Erks, whereas Arg is rarely found at the equivalent position in other EPKs. We observed that Arg84/65 of Erk1/2 interacts with the DFG motif, suggesting that autophosphorylation may be inhibited by the Arg84/65-DFG interactions. Erk1/2s mutated in Arg84/65 autophosphorylate not only the TEY motif, known as critical for catalysis, but also on Thr207/188. Our MS/MS analysis revealed that a large proportion of the Erk2R65H population is phosphorylated on Thr188 or on Tyr185 + Thr188, and a small fraction is phosphorylated on the TEY motif. No molecules phosphorylated on Thr183 + Thr188 were detected. Thus, phosphorylation of Thr183 and Thr188 is mutually exclusive suggesting that not only TEY-phosphorylated molecules are active but perhaps also those phosphorylated on Tyr185 + Thr188. The effect of mutating Arg84/65 may mimic a physiological scenario in which allosteric effectors cause Erk1/2 activation by autophosphorylation.

Dimerization and autophosphorylation of the MST family of kinases are controlled by the same set of residues

The Hippo pathway controls tissue growth and regulates stem cell fate through the activities of core kinase cassette that begins with the Sterile 20-like kinase MST1/2. Activation of MST1/2 relies on trans-autophosphorylation but the details of the mechanisms regulating that reaction are not fully elucidated. Proposals include dimerization as a first step and include multiple models for potential kinase-domain dimers. Efforts to verify and link these dimers to trans-autophosphorylation were unsuccessful. We explored the link between dimerization and trans-autophosphorylation for MST2 and the entire family of MST kinases. We analyzed crystal lattice contacts of structures of MST kinases and identified an ensemble of kinase-domain dimers compatible with trans-autophosphorylation. These dimers share a common dimerization interface comprised of the activation loop and αG-helix while the arrangements of the kinase-domains within the dimer varied depending on their activation state. We then verified the dimerization interface and determined its function using MST2. Variants bearing alanine substitutions of the αG-helix prevented dimerization of the MST2 kinase domain both in solution and in cells. These substitutions also blocked autophosphorylation of full-length MST2 and its Drosophila homolog Hippo in cells. These variants retain the same secondary structure as wild-type and capacity to phosphorylate a protein substrate, indicating the loss of MST2 activation can be directly attributed to a loss of dimerization rather than loss of either fold or catalytic function. Together this data functionally links dimerization and autophosphorylation for MST2 and suggests this activation mechanism is conserved across both species and the entire MST family.

Molecular Dynamics and Machine Learning Give Insights on the Flexibility-Activity Relationships in Tyrosine Kinome

Tyrosine kinases are a subfamily of kinases with critical roles in cellular machinery. Dysregulation of their active or inactive forms is associated with diseases like cancer. This study aimed to holistically understand their flexibility-activity relationships, focusing on pockets and fluctuations. We studied 43 different tyrosine kinases by collecting 120 μs of molecular dynamics simulations, pocket and residue fluctuation analysis, and a complementary machine learning approach. We found that the inactive forms often have increased flexibility, particularly at the DFG motif level. Noteworthy, thanks to these long simulations combined with a decision tree, we identified a semiquantitative fluctuation threshold of the DGF+3 residue over which the kinase has a higher probability to be in the inactive form.

Elucidation of genome-wide understudied proteins targeted by PROTAC-induced degradation using interpretable machine learning

Proteolysis-targeting chimeras (PROTACs) are hetero-bifunctional molecules that induce the degradation of target proteins by recruiting an E3 ligase. PROTACs have the potential to inactivate disease-related genes that are considered undruggable by small molecules, making them a promising therapy for the treatment of incurable diseases. However, only a few hundred proteins have been experimentally tested for their amenability to PROTACs, and it remains unclear which other proteins in the entire human genome can be targeted by PROTACs. In this study, we have developed PrePROTAC, an interpretable machine learning model based on a transformer-based protein sequence descriptor and random forest classification. PrePROTAC predicts genome-wide targets that can be degraded by CRBN, one of the E3 ligases. In the benchmark studies, PrePROTAC achieved a ROC-AUC of 0.81, an average precision of 0.84, and over 40% sensitivity at a false positive rate of 0.05. When evaluated by an external test set which comprised proteins from different structural folds than those in the training set, the performance of PrePROTAC did not drop significantly, indicating its generalizability. Furthermore, we developed an embedding SHapley Additive exPlanations (eSHAP) method, which extends conventional SHAP analysis for original features to an embedding space through in silico mutagenesis. This method allowed us to identify key residues in the protein structure that play critical roles in PROTAC activity. The identified key residues were consistent with existing knowledge. Using PrePROTAC, we identified over 600 novel understudied proteins that are potentially degradable by CRBN and proposed PROTAC compounds for three novel drug targets associated with Alzheimer's disease.

A critical evaluation of protein kinase regulation by activation loop autophosphorylation

Phosphorylation of proteins is a ubiquitous mechanism of regulating their function, localization, or activity. Protein kinases, enzymes that use ATP to phosphorylate protein substrates are, therefore, powerful signal transducers in eukaryotic cells. The mechanism of phosphoryl-transfer is universally conserved among protein kinases, which necessitates the tight regulation of kinase activity for the orchestration of cellular processes with high spatial and temporal fidelity. In response to a stimulus, many kinases enhance their own activity by autophosphorylating a conserved amino acid in their activation loop, but precisely how this reaction is performed is controversial. Classically, kinases that autophosphorylate their activation loop are thought to perform the reaction in trans, mediated by transient dimerization of their kinase domains. However, motivated by the recently discovered regulation mechanism of activation loop cis-autophosphorylation by a kinase that is autoinhibited in trans, we here review the various mechanisms of autoregulation that have been proposed. We provide a framework for critically evaluating biochemical, kinetic, and structural evidence for protein kinase dimerization and autophosphorylation, and share some thoughts on the implications of these mechanisms within physiological signaling networks.

AiKPro: deep learning model for kinome-wide bioactivity profiling using structure-based sequence alignments and molecular 3D conformer ensemble descriptors

The discovery of selective and potent kinase inhibitors is crucial for the treatment of various diseases, but the process is challenging due to the high structural similarity among kinases. Efficient kinome-wide bioactivity profiling is essential for understanding kinase function and identifying selective inhibitors. In this study, we propose AiKPro, a deep learning model that combines structure-validated multiple sequence alignments and molecular 3D conformer ensemble descriptors to predict kinase-ligand binding affinities. Our deep learning model uses an attention-based mechanism to capture complex patterns in the interactions between the kinase and the ligand. To assess the performance of AiKPro, we evaluated the impact of descriptors, the predictability for untrained kinases and compounds, and kinase activity profiling based on odd ratios. Our model, AiKPro, shows good Pearson's correlation coefficients of 0.88 and 0.87 for the test set and for the untrained sets of compounds, respectively, which also shows the robustness of the model. AiKPro shows good kinase-activity profiles across the kinome, potentially facilitating the discovery of novel interactions and selective inhibitors. Our approach holds potential implications for the discovery of novel, selective kinase inhibitors and guiding rational drug design.

Current and emerging therapeutic approaches for colorectal cancer: A comprehensive review

Colorectal cancer (CRC) affects 1 in 23 males and 1 in 25 females, making it the third most common cancer. With roughly 608000 deaths worldwide, CRC accounts for 8% of all cancer-related deaths, making it the second most common cause of death due to cancer. Standard and conventional CRC treatments include surgical expurgation for resectable CRC and radiotherapy, chemotherapy, immunotherapy, and their combinational regimen for non-resectable CRC. Despite these tactics, nearly half of patients develop incurable recurring CRC. Cancer cells resist the effects of chemotherapeutic drugs in a variety of ways, including drug inactivation, drug influx and efflux modifications, and ATP-binding cassette transporter overexpression. These constraints necessitate the development of new target-specific therapeutic strategies. Emerging therapeutic approaches, such as targeted immune boosting therapies, non-coding RNA-based therapies, probiotics, natural products, oncolytic viral therapies, and biomarker-driven therapies, have shown promising results in preclinical and clinical studies. We tethered the entire evolutionary trends in the development of CRC treatments in this review and discussed the potential of new therapies and how they might be used in conjunction with conventional treatments as well as their advantages and drawbacks as future medicines.

Deep Learning-Based Modeling of Drug–Target Interaction Prediction Incorporating Binding Site Information of Proteins

Chemogenomics, also known as proteochemometrics, covers various computational methods for predicting interactions between related drugs and targets on large-scale data. Chemogenomics is used in the early stages of drug discovery to predict the off-target effects of proteins against therapeutic candidates. This study aims to predict unknown ligand–target interactions using one-dimensional SMILES as inputs for ligands and binding site residues for proteins in a computationally efficient manner. We first formulate a Deep learning CNN model using one-dimensional SMILES for drugs and motif-rich binding pocket subsequences of proteins as inputs. We evaluate and compare the proposed deep learning model trained on expert-based features against shallow feature-based machine learning methods. The proposed method achieved better or similar performance on the MSE and AUPR metrics than the shallow methods. Additionally, We show that our deep learning model, DeepPS is computationally more efficient than the deep learning model trained on full-length raw sequences of proteins. We conclude that a beneficial research approach would be to integrate structural information of proteins for modeling drug-target interaction prediction of large datasets for more interpretability, high throughput, and broad applicability.

Graphical abstract

Biochemical Studies of Systemic Lupus Erythematosus-Associated Mutations in Nonreceptor Tyrosine Kinases Ack1 and Brk

Tyrosine kinases (TKs) play essential roles in signaling processes that regulate cell survival, migration, and proliferation. Dysregulation of tyrosine kinases underlies many disorders, including cancer, cardiovascular and developmental diseases, as well as pathologies of the immune system. Ack1 and Brk are nonreceptor tyrosine kinases (NRTKs) best known for their roles in cancer. Here, we have biochemically characterized novel Ack1 and Brk mutations identified in patients with systemic lupus erythematosus (SLE). These mutations are the first SLE-linked polymorphisms found among NRTKs. We show that two of the mutants are catalytically inactive, while the other three have reduced activity. To understand the structural changes associated with the loss-of-function phenotype, we solved the crystal structure of one of the Ack1 kinase mutants, K161Q. Furthermore, two of the mutated residues (Ack1 A156 and K161) critical for catalytic activity are highly conserved among other TKs, and their substitution in other members of the kinase family could have implications in cancer. In contrast to canonical gain-of-function mutations in TKs observed in many cancers, we report loss-of-function mutations in Ack1 and Brk, highlighting the complexity of TK involvement in human diseases.

Strategy toward Kinase-Selective Drug Discovery

Kinase drug selectivity is the ground challenge in cancer research. Due to the structurally similar kinase drug pockets, off-target inhibitor toxicity has been a major cause for clinical trial failures. The pockets are similar but not identical. Here, we describe a transformation invariant protocol to identify distinct geometric features in the drug pocket that can distinguish one kinase from all others. We integrate available experimental structures with the artificial intelligence-based structural kinome, performing a kinome-wide structural bioinformatic analysis to establish the structural principles of kinase drug selectivity. We generate the structural landscape from the experimental kinase-ligand complexes and propose a binary network that encapsulates the information. The results show that all kinases contain binary units that are shared by less than seven other kinases in the kinome. 331 kinases contain unique binary units that may distinguish them from all others. The structural features encoded by these binary units in the network represent the inhibitor-accessible geometric space that may capture the kinome-wide selectivity. Our proposed binary network with the unsupervised clustering can serve as a general structural bioinformatic protocol for extracting the distinguishing structural features for any protein from their families. We apply the binary network to epidermal growth factor receptor tyrosine kinase inhibitor selectivity by targeting the gate area and the AKT1 serine/threonine kinase selectivity by binding to the αC-helix region and the allosteric pocket. Finally, we develop the cross-platform software, KDS (Kinase Drug Selectivity), for customized visualization and analysis of the binary networks in the human kinome (https://github.com/CBIIT/KDS).

Conservation of land plant-specific receptor-like cytoplasmic kinase subfamily XI possessing a unique kinase insert domain

The number of genes encoding receptor-like kinases (RLKs) has expanded in the plant lineage. Their expansion has resulted in the emergence of diverse domain architectures that function in signaling cascades related to growth, development, and stress response. In this study, we focused on receptor-like cytoplasmic kinase subfamily XI (RLCK XI) in plants. We discovered an exceptionally long kinase insert domain (KID), averaging 280 amino acids, between subdomains VII and VIII of the conserved protein kinase domain. Using sequence homology search, we identified members of RLCK XI with the unique KID architecture in terrestrial plants, up to a single copy in several hornwort and liverwort species. The KID shows a high propensity for being disordered, resembling the activation segment in the model kinase domain. Several conserved sequence motifs were annotated along the length of the KID. Of note, the KID harbors repetitive nuclear localization signals capable of mediating RLCK XI translocation from the plasma membrane to the nucleus. The possible physiological implication of dual localization of RLCK XI members is discussed. The presence of a KID in RLCK XI represents a unique domain architecture among RLKs specific to land plants.

Tree visualizations of protein sequence embedding space enable improved functional clustering of diverse protein superfamilies

Protein language models, trained on millions of biologically observed sequences, generate feature-rich numerical representations of protein sequences. These representations, called sequence embeddings, can infer structure-functional properties, despite protein language models being trained on primary sequence alone. While sequence embeddings have been applied toward tasks such as structure and function prediction, applications toward alignment-free sequence classification have been hindered by the lack of studies to derive, quantify and evaluate relationships between protein sequence embeddings. Here, we develop workflows and visualization methods for the classification of protein families using sequence embedding derived from protein language models. A benchmark of manifold visualization methods reveals that Neighbor Joining (NJ) embedding trees are highly effective in capturing global structure while achieving similar performance in capturing local structure compared with popular dimensionality reduction techniques such as t-SNE and UMAP. The statistical significance of hierarchical clusters on a tree is evaluated by resampling embeddings using a variational autoencoder (VAE). We demonstrate the application of our methods in the classification of two well-studied enzyme superfamilies, phosphatases and protein kinases. Our embedding-based classifications remain consistent with and extend upon previously published sequence alignment-based classifications. We also propose a new hierarchical classification for the S-Adenosyl-L-Methionine (SAM) enzyme superfamily which has been difficult to classify using traditional alignment-based approaches. Beyond applications in sequence classification, our results further suggest NJ trees are a promising general method for visualizing high-dimensional data sets.

An atlas of substrate specificities for the human serine/threonine kinome

Protein phosphorylation is one of the most widespread post-translational modifications in biology1,2. With advances in mass-spectrometry-based phosphoproteomics, 90,000 sites of serine and threonine phosphorylation have so far been identified, and several thousand have been associated with human diseases and biological processes3,4. For the vast majority of phosphorylation events, it is not yet known which of the more than 300 protein serine/threonine (Ser/Thr) kinases encoded in the human genome are responsible3. Here we used synthetic peptide libraries to profile the substrate sequence specificity of 303 Ser/Thr kinases, comprising more than 84% of those predicted to be active in humans. Viewed in its entirety, the substrate specificity of the kinome was substantially more diverse than expected and was driven extensively by negative selectivity. We used our kinome-wide dataset to computationally annotate and identify the kinases capable of phosphorylating every reported phosphorylation site in the human Ser/Thr phosphoproteome. For the small minority of phosphosites for which the putative protein kinases involved have been previously reported, our predictions were in excellent agreement. When this approach was applied to examine the signalling response of tissues and cell lines to hormones, growth factors, targeted inhibitors and environmental or genetic perturbations, it revealed unexpected insights into pathway complexity and compensation. Overall, these studies reveal the intrinsic substrate specificity of the human Ser/Thr kinome, illuminate cellular signalling responses and provide a resource to link phosphorylation events to biological pathways.

Evolutionary divergence in the conformational landscapes of tyrosine vs serine/threonine kinases

Inactive conformations of protein kinase catalytic domains where the DFG motif has a "DFG-out" orientation and the activation loop is folded present a druggable binding pocket that is targeted by FDA-approved 'type-II inhibitors' in the treatment of cancers. Tyrosine kinases (TKs) typically show strong binding affinity with a wide spectrum of type-II inhibitors while serine/threonine kinases (STKs) usually bind more weakly which we suggest here is due to differences in the folded to extended conformational equilibrium of the activation loop between TKs vs. STKs. To investigate this, we use sequence covariation analysis with a Potts Hamiltonian statistical energy model to guide absolute binding free-energy molecular dynamics simulations of 74 protein-ligand complexes. Using the calculated binding free energies together with experimental values, we estimated free-energy costs for the large-scale (~17-20 Å) conformational change of the activation loop by an indirect approach, circumventing the very challenging problem of simulating the conformational change directly. We also used the Potts statistical potential to thread large sequence ensembles over active and inactive kinase states. The structure-based and sequence-based analyses are consistent; together they suggest TKs evolved to have free-energy penalties for the classical 'folded activation loop' DFG-out conformation relative to the active conformation, that is, on average, 4-6 kcal/mol smaller than the corresponding values for STKs. Potts statistical energy analysis suggests a molecular basis for this observation, wherein the activation loops of TKs are more weakly 'anchored' against the catalytic loop motif in the active conformation and form more stable substrate-mimicking interactions in the inactive conformation. These results provide insights into the molecular basis for the divergent functional properties of TKs and STKs, and have pharmacological implications for the target selectivity of type-II inhibitors.

Pseudokinase NRP1 facilitates endocytosis of transferrin in the African trypanosome

Trypanosoma brucei causes human African trypanosomiasis (HAT) and nagana in cattle. During infection of a vertebrate, endocytosis of host transferrin (Tf) is important for viability of the parasite. The majority of proteins involved in trypanosome endocytosis of Tf are unknown. Here we identify pseudokinase NRP1 (Tb427tmp.160.4770) as a regulator of Tf endocytosis. Genetic knockdown of NRP1 inhibited endocytosis of Tf without blocking uptake of bovine serum albumin. Binding of Tf to the flagellar pocket was not affected by knockdown of NRP1. However the quantity of Tf per endosome dropped significantly, consistent with NRP1 promoting robust capture and/or retention of Tf in vesicles. NRP1 is involved in motility of Tf-laden vesicles since distances between endosomes and the kinetoplast were reduced after knockdown of the gene. In search of possible mediators of NRP1 modulation of Tf endocytosis, the gene was knocked down and the phosphoproteome analyzed. Phosphorylation of protein kinases forkhead, NEK6, and MAPK10 was altered, in addition to EpsinR, synaptobrevin and other vesicle-associated proteins predicted to be involved in endocytosis. These candidate proteins may link NRP1 functionally either to protein kinases or to vesicle-associated proteins.

The protein common assembly database (ProtCAD)-a comprehensive structural resource of protein complexes

Proteins often act through oligomeric interactions with other proteins. X-ray crystallography and cryo-electron microscopy provide detailed information on the structures of biological assemblies, defined as the most likely biologically relevant structures derived from experimental data. In crystal structures, the most relevant assembly may be ambiguously determined, since multiple assemblies observed in the crystal lattice may be plausible. It is estimated that 10-15% of PDB entries may have incorrect or ambiguous assembly annotations. Accurate assemblies are required for understanding functional data and training of deep learning methods for predicting assembly structures. As with any other kind of biological data, replication via multiple independent experiments provides important validation for the determination of biological assembly structures. Here we present the Protein Common Assembly Database (ProtCAD), which presents clusters of protein assembly structures observed in independent structure determinations of homologous proteins in the Protein Data Bank (PDB). ProtCAD is searchable by PDB entry, UniProt identifiers, or Pfam domain designations and provides downloads of coordinate files, PyMol scripts, and publicly available assembly annotations for each cluster of assemblies. About 60% of PDB entries contain assemblies in clusters of at least 2 independent experiments. All clusters and coordinates are available on ProtCAD web site (http://dunbrack2.fccc.edu/protcad).

Molecular Recognition of FDA-Approved Small Molecule Protein Kinase Drugs in Protein Kinases

Protein kinases are key enzymes that catalyze the covalent phosphorylation of substrates via the transfer of the γ-phosphate of ATP, playing a crucial role in cellular proliferation, differentiation, and various cell regulatory processes. Due to their pivotal cellular role, the aberrant function of kinases has been associated with cancers and many other diseases. Consequently, competitive inhibition of the ATP binding site of protein kinases has emerged as an effective means of curing these diseases. Decades of intense development of protein kinase inhibitors (PKIs) resulted in 71 FDA-approved PKI drugs that target dozens of protein kinases for the treatment of various diseases. How do FDA-approved protein kinase inhibitor PKI drugs compete with ATP in their own binding pocket? This is the central question we attempt to address in this work. Based on modes of non-bonded interactions and their calculated interaction strengths by means of the advanced double hybrid DFT method B2PLYP, the molecular recognition of PKI drugs in the ATP-binding pockets was systematically analyzed. It was found that (1) all the FDA-approved PKI drugs studied here form one or more hydrogen bond(s) with the backbone amide N, O atoms in the hinge region of the ATP binding site, mimicking the adenine base; (2) all the FDA-approved PKI drugs feature two or more aromatic rings. The latter reach far and deep into the hydrophobic regions I and II, forming multiple CH-π interactions with aliphatic residues L(3), V(11), A(15), V(36), G(51), L(77) and π-π stacking interactions with aromatic residues F(47) and F(82), but ATP itself does not utilize these regions extensively; (3) all FDA-approved PKI drugs studied here have one thing in common, i.e., they frequently formed non-bonded interactions with a total of 12 residues L(3),V(11), A(15), K(17), E(24),V(36),T(45), F(47), G(51), L(77), D(81) and F(82) in the ATP binding. Many of those 12 commonly involved residues are highly conserved residues with important structural and catalytic functional roles. K(17) and E(24) are the two highly conserved residues crucial for the catalytic function of kinases. D(81) and F(82) belong to the DFG motif; T(45) was dubbed the gate keeper residue. F(47) is located on the hinge region and G(51) sits on the linker that connects the hinge to the αD-helix. It is this targeting of highly conserved residues in protein kinases that led to promiscuous PKI drugs that lack selectivity. Although the formation of hydrogen bond(s) with the backbone of the hinge gives PKI drugs the added binding affinity and the much-needed directionality, selectivity is sacrificed. That is why so many FDA-approved PKI drugs are known to have multiple targets. Moreover, off-target-mediated toxicity caused by a lack of selectivity was one of the major challenges facing the PKI drug discovery community. This work suggests a road map for future PKI drug design, i.e., targeting non-conserved residues in the ATP binding pocket to gain better selectivity so as to avoid off-target-mediated toxicity.

Exploring the Role of Chemical Reactions in the Selectivity of Tyrosine Kinase Inhibitors

A variety of diseases are associated with tyrosine kinase enzymes that activate many proteins via signal transduction cascades. The similar ATP-binding pockets of these tyrosine kinases make it extremely difficult to design selective covalent inhibitors. The present study explores the contribution of the chemical reaction steps to the selectivity of the commercialized inhibitor acalabrutinib over the Bruton's tyrosine kinase (BTK) and the interleukin-2-inducible T-cell kinase (ITK). Ab initio and empirical valence bond (EVB) simulations of the two kinases indicate that the most favorable reaction path involves a water-assisted mechanism of the 2-butynamide reactive group of acalabrutinib. BTK reacts with acalabrutinib with a substantially lower barrier than ITK, according to our calculated free-energy profile and kinetic simulations. Such a difference is due to the microenvironment of the active site, as further supported by a sequence-based analysis of specificity determinants for several commercialized inhibitors. Our study involves a new approach of simulating directly the IC50 and inactivation efficiency k_eff, instead of using the standard formulas. This new strategy is particularly important in studies of covalent inhibitors with a very exothermic bonding step. Overall, our results demonstrate the importance of understanding the chemical reaction steps in designing selective covalent inhibitors for tyrosine kinases.

KiSSim: Predicting Off-Targets from Structural Similarities in the Kinome

Protein kinases are among the most important drug targets because their dysregulation can cause cancer, inflammatory and degenerative diseases, and many more. Developing selective inhibitors is challenging due to the highly conserved binding sites across the roughly 500 human kinases. Thus, detecting subtle similarities on a structural level can help explain and predict off-targets among the kinase family. Here, we present the kinase-focused, subpocket-enhanced KiSSim fingerprint (Kinase Structural Similarity). The fingerprint builds on the KLIFS pocket definition, composed of 85 residues aligned across all available protein kinase structures, which enables residue-by-residue comparison without a computationally expensive alignment. The residues' physicochemical and spatial properties are encoded within their structural context including key subpockets at the hinge region, the DFG motif, and the front pocket. Since structure was found to contain information complementary to sequence, we used the fingerprint to calculate all-against-all similarities within the structurally covered kinome. We could identify off-targets that are unexpected if solely considering the sequence-based kinome tree grouping; for example, Erlobinib's known kinase off-targets SLK and LOK show high similarities to the key target EGFR (TK group), although belonging to the STE group. KiSSim reflects profiling data better or at least as well as other approaches such as KLIFS pocket sequence identity, KLIFS interaction fingerprints (IFPs), or SiteAlign. To rationalize observed (dis)similarities, the fingerprint values can be visualized in 3D by coloring structures with residue and feature resolution. We believe that the KiSSim fingerprint is a valuable addition to the kinase research toolbox to guide off-target and polypharmacology prediction. The method is distributed as an open-source Python package on GitHub and as a conda package: https://github.com/volkamerlab/kissim.

The Ascension of Targeted Covalent Inhibitors

Covalent drugs have made a major impact on human health but until recently were shunned by the pharmaceutical industry over concerns about the potential for toxicity. A resurgence has occurred driven by the clinical success of targeted covalent inhibitors (TCIs), with eight drugs approved over the past decade. The opportunity to create unique drugs by exploiting the covalent mechanism of action has enabled clinically decisive target product profiles to be achieved. TCIs have revolutionized the treatment paradigm for non-small-cell lung cancer and chronic lymphocytic leukemia. This Perspective will highlight the clinical and financial success of this class of drugs and provide early insight into toxicity, a key factor that had hindered progress in the field. Further innovation in the TCI approach, including expanding beyond cysteine-directed electrophiles, kinases, and cancer, highlights the broad opportunity to deliver a new generation of breakthrough therapies.

Colocalization and Interaction Study of Neuronal JNK3, JIP1, and β-Arrestin2 Together with PSD95

c-Jun N-terminal kinases (JNKs) are stress-activated serine/threonine protein kinases belonging to the mitogen-activated protein kinase (MAPK) family. Among them, JNK3 is selectively expressed in the central nervous system, cardiac smooth muscle, and testis. In addition, it is the most responsive JNK isoform to stress stimuli in the brain, and it is involved in synaptic dysfunction, an essential step in neurodegenerative processes. JNK3 pathway is organized in a cascade of amplification in which signal transduction occurs by stepwise, highly controlled phosphorylation. Since different MAPKs share common upstream activators, pathway specificity is guaranteed by scaffold proteins such as JIP1 and β-arrestin2. To better elucidate the physiological mechanisms regulating JNK3 in neurons, and how these interactions may be involved in synaptic (dys)function, we used (i) super-resolution microscopy to demonstrate the colocalization among JNK3-PSD95-JIP1 and JNK3-PSD95-β-arrestin2 in cultured hippocampal neurons, and (ii) co-immunoprecipitation techniques to show that the two scaffold proteins and JNK3 can be found interacting together with PSD95. The protein-protein interactions that govern the formation of these two complexes, JNK3-PSD95-JIP1 and JNK3-PSD95-β-arrestin2, may be used as targets to interfere with their downstream synaptic events.

Switch Pocket Kinase: An Emerging Therapeutic Target for the Design of Anticancer Agents

Protein kinases are amongst the most focused enzymes in current century to design, synthesize and formulate drugs ought to be effective in the treatment of various disordered and diseased states involving either overexpression or deficiency situations. The ATP pocket on the kinases is the binding active site for most of the kinase inhibitors. However, the kinase mutations prevent the binding of kinase inhibitors to ATP pocket. The switch pocket site on this enzyme when occupied by switch pocket inhibitors, the enzyme become inactive even in the mutated state. This review comprises the detailed information on various classical protein kinases and switch pocket kinase inhibitors with their mechanism of action so that new molecules can be designed to encounter mutations in the kinase enzyme.

Comparative analysis of the kinomes of Plasmodium falciparum, Plasmodium vivax and their host Homo sapiens

Background

Novel antimalarials should be effective across all species of malaria parasites that infect humans, especially the two species that bear the most impact, Plasmodium falciparum and Plasmodium vivax. Protein kinases encoded by pathogens, as well as host kinases required for survival of intracellular pathogens, carry considerable potential as targets for antimalarial intervention (Adderley et al. Trends Parasitol 37:508–524, 2021;  Wei et al. Cell Rep Med 2:100423, 2021). To date, no comprehensive P. vivax kinome assembly has been conducted; and the P. falciparum kinome, first assembled in 2004, requires an update. The present study, aimed to fill these gaps, utilises a recently published structurally-validated multiple sequence alignment (MSA) of the human kinome (Modi et al. Sci Rep 9:19790, 2019). This MSA is used as a scaffold to assist the alignment of all protein kinase sequences from P. falciparum and P. vivax, and (where possible) their assignment to specific kinase groups/families.

Results

We were able to assign six P. falciparum previously classified as OPK or ‘orphans’ (i.e. with no clear phylogenetic relation to any of the established ePK groups) to one of the aforementioned ePK groups. Direct phylogenetic comparison established that despite an overall high level of similarity between the P. falciparum and P. vivax kinomes, which will help in selecting targets for intervention, there are differences that may underlie the biological specificities of these species. Furthermore, we highlight a number of Plasmodium kinases that have a surprisingly high level of similarity with their human counterparts and therefore not well suited as targets for drug discovery.

Conclusions

Direct comparison of the kinomes of Homo sapiens, P. falciparum and P. vivax sheds additional light on the previously documented divergence of many P. falciparum and P. vivax kinases from those of their human host. We provide the first direct kinome comparison between the phylogenetically distinct species of P. falciparum and P. vivax, illustrating the key similarities and differences which must be considered in the context of kinase-directed antimalarial drug discovery, and discuss the divergences and similarities between the human and Plasmodium kinomes to inform future searches for selective antimalarial intervention.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-022-08457-0.

Evidence that nuclear receptors are related to terpene synthases

Ligand-activated nuclear receptors (NRs) orchestrate development, growth, and reproduction across all animal lifeforms - the Metazoa - but how NRs evolved remains mysterious. Given the NR ligands including steroids and retinoids are predominantly terpenoids, we asked whether NRs might have evolved from enzymes that catalyze terpene synthesis and metabolism. We provide evidence suggesting that NRs may be related to the terpene synthase (TS) enzyme superfamily. Based on over 10,000 3D structural comparisons, we report that the NR ligand-binding domain and TS enzymes share a conserved core of seven α-helical segments. In addition, the 3D locations of the major ligand-contacting residues are also conserved between the two protein classes. Primary sequence comparisons reveal suggestive similarities specifically between NRs and the subfamily of cis-isoprene transferases, notably with dehydrodolichyl pyrophosphate synthase and its obligate partner, NUS1/NOGOB receptor. Pharmacological overlaps between NRs and TS enzymes add weight to the contention that they share a distant evolutionary origin, and the combined data raise the possibility that a ligand-gated receptor may have arisen from an enzyme antecedent. However, our findings do not formally exclude other interpretations such as convergent evolution, and further analysis will be necessary to confirm the inferred relationship between the two protein classes.

Regulatory spine RS3 residue of protein kinases: a lipophilic bystander or a decisive element in the small-molecule kinase inhibitor binding?

In recent years, protein kinases have been one of the most pursued drug targets. These determined efforts have resulted in ever increasing numbers of small-molecule kinase inhibitors reaching to the market, offering novel treatment options for patients with distinct diseases. One essential component related to the activation and normal functionality of a protein kinase is the regulatory spine (R-spine). The R-spine is formed of four conserved residues named as RS1–RS4. One of these residues, RS3, located in the C-terminal part of αC-helix, is usually accessible for the inhibitors from the ATP-binding cavity as its side chain is lining the hydrophobic back pocket in many protein kinases. Although the role of RS3 has been well acknowledged in protein kinase function, this residue has not been actively considered in inhibitor design, even though many small-molecule kinase inhibitors display interactions to this residue. In this minireview, we will cover the current knowledge of RS3, its relationship with the gatekeeper, and the role of RS3 in kinase inhibitor interactions. Finally, we comment on the future perspectives how this residue could be utilized in the kinase inhibitor design.