Crystal structure of the EphA4 protein tyrosine kinase domain in the apo- and dasatinib-bound state

MolLoG: A Molecular Level Interpretability Model Bridging Local to Global for Predicting Drug Target Interactions

Developing new pharmaceuticals is a costly and time-consuming endeavor fraught with significant safety risks. A critical aspect of drug research and disease therapy is discerning the existence of interactions between drugs and proteins. The evolution of deep learning (DL) in computer science has been remarkably aided in this regard in recent years. Yet, two challenges remain: (i) balancing the extraction of profound, local cohesive characteristics while warding off gradient disappearance and (ii) globally representing and understanding the interactions between the drug and target local attributes, which is vital for delivering molecular level insights indispensable to drug development. In response to these challenges, we propose a DL network structure, MolLoG, primarily comprising two modules: local feature encoders (LFE) and global interactive learning (GIL). Within the LFE module, graph convolution networks and leap blocks capture the local features of drug and protein molecules, respectively. The GIL module enables the efficient amalgamation of feature information, facilitating the global learning of feature structural semantics and procuring multihead attention weights for abstract features stemming from two modalities, providing biologically pertinent explanations for black-box results. Finally, predictive outcomes are achieved by decoding the unified representation via a multilayer perceptron. Our experimental analysis reveals that MolLoG outperforms several cutting-edge baselines across four data sets, delivering superior overall performance and providing satisfactory results when elucidating various facets of drug-target interaction predictions.

Pharmacological Targeting of Senescence with Senolytics as a New Therapeutic Strategy for Neurodegeneration

Cellular senescence is a state of permanent cell-cycle arrest. Early in life, senescence has a physiologic role in tumor suppression and wound healing. However, gradually, as these senescent cells accumulate over the lifespan of an organism, they contribute to inflammation and the progression of age-related diseases, including neurodegeneration. Targeting senescent cells using a class of drugs known as "senolytics" holds great promise for the management of Alzheimer's and Parkinson's disease. Already, several senolytic compounds have been shown to ameliorate cognitive deficits across several preclinical models of neurodegeneration. Most of these senolytics (e.g., dasatinib) are repurposed clinical or experimental anticancer drugs, which trigger apoptosis of senescent cells by interfering with pro-survival pathways. However, outside of their senolytic function, many first-generation senolytics also have other less appreciated neuroprotective effects, such as potent antioxidant and anti-inflammatory activity. In addition, some senolytic drugs may also have negative dose-limiting toxicities, including thrombocytopenia. In this review, we discuss the various biologic pathways targeted by the leading senolytic drugs, namely dasatinib, quercetin, fisetin, and navitoclax. We further evaluate the clinical transability of these compounds for neurodegeneration, assessing their adverse effects, pharmacokinetic properties, and chemical structure. SIGNIFICANCE STATEMENT: Currently, there are no effective disease-modifying treatments for the most prevalent neurodegenerative disorders, including Alzheimer's and Parkinson's disease. Some of the drugs currently available for treating these diseases are associated with unwanted side-effects and/or become less efficacious with time. Therefore, researchers have begun to explore new innovative treatments for these belligerent diseases, including senolytic drugs. These agents lead to the apoptosis of senescent cells thereby preventing their deleterious role in neurodegeneration.

Crystal structure of clinically reported mutations Gly656Arg, Gly656Glu and Asp751His identified in the kinase domain of EphA7

Erythropoietin producing hepatocellular (Eph) forms the largest family of receptor tyrosine kinases (RTK). As a family, Eph regulates physiological events such as cell-cell interaction, cell migration, and adhesion. The Kinase domain is the catalytic core of the Eph receptor and is highly conserved sequentially. EphA7 has been recently regarded as a cancer driver gene and comprises several clinically important mutations. Three of the EphA7 mutations Gly656Glu, Gly656Arg, and Asp751His, present in the kinase domain, are predicted to be highly pathogenic. Furthermore, Gly656Glu and Gly656Arg are reported to be hotspot mutations. Considering the importance of mutations, crystals structure of EphA7 Gly656Glu, Gly656Arg, and Asp751His mutants has been explored. Changes in folding pattern and intramolecular interactions were observed in mutant structures. Secondary structural changes were observed in the hinge region of EphA7 Gly656Arg and Asp751His structure, affecting the transition of kinase domain between open and closed conformations. EphA7 Asp751His mutant structure shows a distorted nucleotide-binding groove. Differences were observed in hydrogen bonding and hydrophobic interactions between the catalytic and highly conserved DFG motif in the EphA7 mutants, which may influence the catalytic activity of kinase domain.

MolTrans: Molecular Interaction Transformer for drug-target interaction prediction

Motivation

Drug–target interaction (DTI) prediction is a foundational task for in-silico drug discovery, which is costly and time-consuming due to the need of experimental search over large drug compound space. Recent years have witnessed promising progress for deep learning in DTI predictions. However, the following challenges are still open: (i) existing molecular representation learning approaches ignore the sub-structural nature of DTI, thus produce results that are less accurate and difficult to explain and (ii) existing methods focus on limited labeled data while ignoring the value of massive unlabeled molecular data.

Results

We propose a Molecular Interaction Transformer (MolTrans) to address these limitations via: (i) knowledge inspired sub-structural pattern mining algorithm and interaction modeling module for more accurate and interpretable DTI prediction and (ii) an augmented transformer encoder to better extract and capture the semantic relations among sub-structures extracted from massive unlabeled biomedical data. We evaluate MolTrans on real-world data and show it improved DTI prediction performance compared to state-of-the-art baselines.

Availability and implementation

The model scripts are available at https://github.com/kexinhuang12345/moltrans.

Supplementary information

Supplementary data are available at Bioinformatics online.

Dasatinib-SIK2 Binding Elucidated by Homology Modeling, Molecular Docking, and Dynamics Simulations

Salt-inducible kinases (SIKs) are calcium/calmodulin-dependent protein kinase (CAMK)-like (CAMKL) family members implicated in insulin signal transduction, metabolic regulation, inflammatory response, and other processes. Here, we focused on SIK2, which is a target of the Food and Drug Administration (FDA)-approved pan inhibitor N-(2-chloro-6-methylphenyl)-2-(6-(4-(2-hydroxyethyl)piperazin-1-yl)-2-methylpyrimidin-4-ylamino)thiazole-5-carboxamide (dasatinib), and constructed four representative SIK2 structures by homology modeling. We investigated the interactions between dasatinib and SIK2 via molecular docking, molecular dynamics simulation, and binding free energy calculation and found that dasatinib showed strong binding affinity for SIK2. Binding free energy calculations suggested that the modification of various dasatinib regions may provide useful information for drug design and to guide the discovery of novel dasatinib-based SIK2 inhibitors.

The role of Eph receptors in cancer and how to target them: novel approaches in cancer treatment

INTRODUCTION

Erythropoietin-producing human hepatocellular (Eph) receptors are among the largest family of tyrosine kinases that are divided into two classes: EphA and EphB receptors. Over the past two decades, their role in cancer has become more evident.

AREAS COVERED

There is a need for new anticancer treatments and more insight in the emerging role of Eph receptors in cancer. Molecular mechanisms underlying the pro-tumorigenic effects of Eph receptors could be exploited for future therapeutic strategies. This review describes the variability in expression levels and different effects on oncogenic and tumor suppressive downstream signaling of Eph receptors in various cancer types, and the small molecules, antibodies and peptides that target these receptors.

EXPERT OPINION

The complexity of Eph signaling is a challenge for the definition of clear targets for cancer treatment. Nevertheless, numerous drugs that target EphA2 and EphB4 are currently in clinical trials. However, some Eph targeted drugs also inhibit other tyrosine kinases, so it is unclear to what extent the targeting of Eph receptors contributes to their efficacy. Future research is warranted for an improved understanding of the full network in which Eph receptors function. This will be critical for the improvement of the anticancer effects of drugs that target the Eph receptors.

Direct Phosphorylation of SRC Homology 3 Domains by Tyrosine Kinase Receptors Disassembles Ligand-Induced Signaling Networks

Phosphotyrosine (pTyr) signaling has evolved into a key cell-to-cell communication system. Activated receptor tyrosine kinases (RTKs) initiate several pTyr-dependent signaling networks by creating the docking sites required for the assembly of protein complexes. However, the mechanisms leading to network disassembly and its consequence on signal transduction remain essentially unknown. We show that activated RTKs terminate downstream signaling via the direct phosphorylation of an evolutionarily conserved Tyr present in most SRC homology (SH) 3 domains, which are often part of key hub proteins for RTK-dependent signaling. We demonstrate that the direct EPHA4 RTK phosphorylation of adaptor protein NCK SH3s at these sites results in the collapse of signaling networks and abrogates their function. We also reveal that this negative regulation mechanism is shared by other RTKs. Our findings uncover a conserved mechanism through which RTKs rapidly and reversibly terminate downstream signaling while remaining in a catalytically active state on the plasma membrane.

Design, Synthesis, and Evaluation of Dasatinib-Amino Acid and Dasatinib-Fatty Acid Conjugates as Protein Tyrosine Kinase Inhibitors

Derivatives of the tyrosine kinase inhibitor dasatinib were synthesized by esterification with 25 carboxylic acids, including amino acids and fatty acids, thereby extending the drug to interact with more diverse sites and to improve specificity. The dasatinib-l-arginine derivative (Das-R, 7) was found to be the most potent of the inhibitors tested, with IC50 values of 4.4, <0.25, and <0.45 nm against Csk, Src, and Abl kinases, respectively. The highest selectivity ratio obtained in our study, 91.4 Csk/Src, belonged to compound 18 (Das-C10 ) with an IC50 value of 3.2 μm for Csk compared with 35 nm for Src. Furthermore, many compounds displayed increased selectivity toward Src over Abl. Compounds 15 (Das-glutamic acid) and 13 (Das-cysteine) demonstrated the largest gains (10.2 and 10.3 Abl/Src IC50 ratios). Das-R (IC50 =2.06 μm) was significantly more potent than the parent dasatinib (IC50 =26.3 μm) against Panc-1 cells, whereas both compounds showed IC50 <51.2 pm against BV-173 and K562 cells. Molecular modeling and binding free energy simulations revealed good agreements with the experimental results and rationalized the differences in selectivity among the studied compounds. Integration of experimental and computational approaches in the design and biochemical screening of dasatinib derivatives facilitated rational engineering and diversification of the dasatinib scaffold, providing useful insight into mechanisms of kinase selectivity.

EPHA4 is overexpressed but not functionally active in Sézary syndrome

EPHA4 belongs to the largest subfamily of receptor tyrosine kinases. In addition to its function during development, overexpression of EPHA4 in tumors has been correlated with increased proliferation, migration and poor survival. Several genome-wide transcription profiling studies have demonstrated high EPHA4 expression in Sézary syndrome (SS), a leukemic variant of cutaneous CD4+ T-cell lymphoma (CTCL) with an aggressive clinical course and poor prognosis. In this study we set out to explore the functional role of EPHA4 in SS. Both high EPHA4 mRNA and protein expression was found in circulating SS-cells of patients compared to healthy CD4+ T-cells. However, using a phosphospecific EPHA4 antibody, phosphorylation of the EPHA4 kinase domain was not detected in either circulating or skin residing SS cells. Moreover, treatment with the phosphatase inhibitor pervanadate did not result in detectable phosphorylation of the EPHA4 kinase domain, in either SS cells or in healthy CD4+ T-cells. Thus, the results from our study confirm high EPHA4 expression in SS cells both on the mRNA and protein levels, making EPHA4 a good diagnostic marker. However, the overexpressed EPHA4 does not appear to be functionally active and its overexpression might be secondary to other oncogenic drivers in SS, like STAT3 and TWIST1.

Genome-wide association study identifies new susceptibility loci for adolescent idiopathic scoliosis in Chinese girls

Adolescent idiopathic scoliosis (AIS) is a structural deformity of the spine affecting millions of children. As a complex disease, the genetic aetiology of AIS remains obscure. Here we report the results of a four-stage genome-wide association study (GWAS) conducted in a sample of 4,317 AIS patients and 6,016 controls. Overall, we identify three new susceptibility loci at 1p36.32 near AJAP1 (rs241215, P_combined=2.95 × 10⁻⁹), 2q36.1 between PAX3 and EPHA4 (rs13398147, P_combined=7.59 × 10⁻¹³) and 18q21.33 near BCL-2 (rs4940576, P_combined=2.22 × 10⁻¹²). In addition, we refine a previously reported region associated with AIS at 10q24.32 (rs678741, P_combined=9.68 × 10⁻³⁷), which suggests LBX1AS1, encoding an antisense transcript of LBX1, might be a functional variant of AIS. This is the first GWAS investigating genetic variants associated with AIS in Chinese population, and the findings provide new insight into the multiple aetiological mechanisms of AIS.

The authors perform a genome-wide association study of adolescent idiopathic scoliosis patients of Han Chinese descent, and identify 3 new loci for disease susceptibility.

Molecular Determinants Underlying Binding Specificities of the ABL Kinase Inhibitors: Combining Alanine Scanning of Binding Hot Spots with Network Analysis of Residue Interactions and Coevolution

Quantifying binding specificity and drug resistance of protein kinase inhibitors is of fundamental importance and remains highly challenging due to complex interplay of structural and thermodynamic factors. In this work, molecular simulations and computational alanine scanning are combined with the network-based approaches to characterize molecular determinants underlying binding specificities of the ABL kinase inhibitors. The proposed theoretical framework unveiled a relationship between ligand binding and inhibitor-mediated changes in the residue interaction networks. By using topological parameters, we have described the organization of the residue interaction networks and networks of coevolving residues in the ABL kinase structures. This analysis has shown that functionally critical regulatory residues can simultaneously embody strong coevolutionary signal and high network centrality with a propensity to be energetic hot spots for drug binding. We have found that selective (Nilotinib) and promiscuous (Bosutinib, Dasatinib) kinase inhibitors can use their energetic hot spots to differentially modulate stability of the residue interaction networks, thus inhibiting or promoting conformational equilibrium between inactive and active states. According to our results, Nilotinib binding may induce a significant network-bridging effect and enhance centrality of the hot spot residues that stabilize structural environment favored by the specific kinase form. In contrast, Bosutinib and Dasatinib can incur modest changes in the residue interaction network in which ligand binding is primarily coupled only with the identity of the gate-keeper residue. These factors may promote structural adaptability of the active kinase states in binding with these promiscuous inhibitors. Our results have related ligand-induced changes in the residue interaction networks with drug resistance effects, showing that network robustness may be compromised by targeted mutations of key mediating residues. This study has outlined mechanisms by which inhibitor binding could modulate resilience and efficiency of allosteric interactions in the kinase structures, while preserving structural topology required for catalytic activity and regulation.

Distinctive Structure of the EphA3/Ephrin-A5 Complex Reveals a Dual Mode of Eph Receptor Interaction for Ephrin-A5

The Eph receptor tyrosine kinase/ephrin ligand system regulates a wide spectrum of physiological processes, while its dysregulation has been implicated in cancer progression. The human EphA3 receptor is widely upregulated in the tumor microenvironment and is highly expressed in some types of cancer cells. Furthermore, EphA3 is among the most highly mutated genes in lung cancer and it is also frequently mutated in other cancers. We report the structure of the ligand-binding domain of the EphA3 receptor in complex with its preferred ligand, ephrin-A5. The structure of the complex reveals a pronounced tilt of the ephrin-A5 ligand compared to its orientation when bound to the EphA2 and EphB2 receptors and similar to its orientation when bound to EphA4. This tilt brings an additional area of ephrin-A5 into contact with regions of EphA3 outside the ephrin-binding pocket thereby enlarging the size of the interface, which is consistent with the high binding affinity of ephrin-A5 for EphA3. This large variation in the tilt of ephrin-A5 bound to different Eph receptors has not been previously observed for other ephrins.

Bruton's TK inhibitors: structural insights and evolution of clinical candidates

Bruton's TK (BTK) is a promising biological target for therapeutic intervention of several diseases including inflammatory diseases and cancer/B cell malignancies. Numerous research groups are actively engaged in investigating the functions of BTK, and discovering potent and selective BTK inhibitors as drug candidates. Revealed by x-ray crystal structures with ligands of diverse chemical structures, the ability of BTK kinase domain to adopt various inactive conformations offers unique opportunities to identify highly potent and exquisitely selective inhibitors. Both reversible and covalent inhibitor approaches have yielded candidates demonstrating safety profiles and efficacies in multiple preclinical models of autoimmunity and oncology. Two BTK inhibitors have entered human clinical trials for oncology indications. Ibrutinib won the US FDA approval in November 2013 to become the first-in-class BTK inhibitor for treating mantle cell lymphoma. This encouraging outcome and the other on-going human studies could ultimately expand the utility of BTK inhibitors to broader autoimmune disease areas.