Yuguang Mu
@ntu.edu.sg
Nanyang Technological University
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Yunju Zhang, Jingjing Guo, Yang Liu, Yuanyuan Qu, Yong-Qiang Li, Yuguang Mu, and Weifeng Li
Elsevier BV
Zhiqing Chen, Xiaopan Xie, Chunyan Jia, Qishuo Zhong, Qiuping Zhang, Daoxin Luo, Yin Cao, Yuguang Mu, and Changliang Ren
Wiley
AbstractIn nature, ceramides are a class of sphingolipids possessing a unique ability to self‐assemble into protein‐permeable channels with intriguing concentration‐dependent adaptive channel cavities. However, within the realm of artificial ion channels, this interesting phenomenon is scarcely represented. Herein, we report on a novel class of adaptive artificial channels, Pn‐TPPs, based on PEGylated cholic acids bearing triphenylphosphonium (TPP) groups as anion binding motifs. Interestingly, the molecules self‐assemble into chloride ion channels at low concentrations while transforming into small molecule‐permeable nanopores at high concentrations. Moreover, the TPP groups endow the molecules with mitochondria‐targeting properties, enabling them to selectively drill holes on the mitochondrial membrane of cancer cells and subsequently trigger the caspase 9 apoptotic pathway. The anticancer efficacies of Pn‐TPPs correlate with their abilities to form nanopores. Significantly, the most active ensembles formed by P5‐TPP exhibits impressive anticancer activity against human liver cancer cells, with an IC50 value of 3.8 μM. While demonstrating similar anticancer performance to doxorubicin, P5‐TPP exhibits a selectivity index surpassing that of doxorubicin by a factor of 16.8.
Swaleeha Jaan Abdullah, Yuguang Mu, and Surajit Bhattacharjya
MDPI AG
Gram-negative bacteria are intrinsically more resistant to many frontline antibiotics, which is attributed to the permeability barrier of the outer membrane, drug efflux pumps and porins. Consequently, discovery of new small molecules antibiotics to kill drug-resistant Gram-negative bacteria presents a significant challenge. Thanatin, a 21-residue insect-derived antimicrobial peptide, is known for its potent activity against Enterobacter Gram-negative bacteria, including drug-resistant strains. Here, we investigated a 15-residue N-terminal truncated analog PM15 (P1IIYCNRRTGKCQRM15) of thanatin to determine modes of action and antibacterial activity. PM15 and the P1 to Y and A substituted variants PM15Y and PM15A delineated interactions and permeabilization of the LPS–outer membrane. In antibacterial assays, PM15 and the analogs showed growth inhibition of strains of Gram-negative bacteria that is largely dependent on the composition of the culture media. Atomic-resolution structures of PM15 and PM15Y in free solution and in complex with LPS micelle exhibited persistent β-hairpin structures similar to native thanatin. However, in complex with LPS, the structures of peptides are more compact, with extensive packing interactions among residues across the two anti-parallel strands of the β-hairpin. The docked complex of PM15/LPS revealed a parallel orientation of the peptide that may be sustained by potential ionic and van der Waals interactions with the lipid A moiety of LPS. Further, PM15 and PM15Y bind to LptAm, a monomeric functional variant of LptA, the periplasmic component of the seven-protein (A-G) complex involved in LPS transport. Taken together, the structures, target interactions and antibacterial effect of PM15 presented in the current study could be useful in designing thanatin-based peptide analogs.
Tao Shen, Fuxu Liu, Zechen Wang, Jinyuan Sun, Yifan Bu, Jintao Meng, Weihua Chen, Keyi Yao, Yuguang Mu, Weifeng Li,et al.
Wiley
AbstractWe introduce a deep learning‐based ligand pose scoring model called zPoseScore for predicting protein–ligand complexes in the 15th Critical Assessment of Protein Structure Prediction (CASP15). Our contributions are threefold: first, we generate six training and evaluation data sets by employing advanced data augmentation and sampling methods. Second, we redesign the “zFormer” module, inspired by AlphaFold2's Evoformer, to efficiently describe protein–ligand interactions. This module enables the extraction of protein–ligand paired features that lead to accurate predictions. Finally, we develop the zPoseScore framework with zFormer for scoring and ranking ligand poses, allowing for atomic‐level protein–ligand feature encoding and fusion to output refined ligand poses and ligand per‐atom deviations. Our results demonstrate excellent performance on various testing data sets, achieving Pearson's correlation = 0.783 and 0.659 for ranking docking decoys generated based on experimental and predicted protein structures of CASF‐2016 protein–ligand complexes. Additionally, we obtain an averaged local distance difference test (lDDT pli = 0.558) of AIchemy LIG2 in CASP15 for de novo protein–ligand complex structure predictions. Detailed analysis shows that accurate ligand binding site prediction and side‐chain orientation are crucial for achieving better prediction performance. Our proposed model is one of the most accurate protein–ligand pose prediction models and could serve as a valuable tool in small molecule drug discovery.
Jia Yi Kimberly Low, Xiangyan Shi, Venkatraman Anandalakshmi, Dawn Neo, Gary Swee Lim Peh, Siew Kwan Koh, Lei Zhou, M. K. Abdul Rahim, Ketti Boo, JiaXuan Lee,et al.
Springer Science and Business Media LLC
AbstractTGFBI-related corneal dystrophy (CD) is characterized by the accumulation of insoluble protein deposits in the corneal tissues, eventually leading to progressive corneal opacity. Here we show that ATP-independent amyloid-β chaperone L-PGDS can effectively disaggregate corneal amyloids in surgically excised human cornea of TGFBI-CD patients and release trapped amyloid hallmark proteins. Since the mechanism of amyloid disassembly by ATP-independent chaperones is unknown, we reconstructed atomic models of the amyloids self-assembled from TGFBIp-derived peptides and their complex with L-PGDS using cryo-EM and NMR. We show that L-PGDS specifically recognizes structurally frustrated regions in the amyloids and releases those frustrations. The released free energy increases the chaperone’s binding affinity to amyloids, resulting in local restructuring and breakage of amyloids to protofibrils. Our mechanistic model provides insights into the alternative source of energy utilized by ATP-independent disaggregases and highlights the possibility of using these chaperones as treatment strategies for different types of amyloid-related diseases.
Jingjing Guo, Yiqiong Bao, Mengrong Li, Shu Li, Lili Xi, Pengyang Xin, Lei Wu, Huanxiang Liu, and Yuguang Mu
Wiley
Hilbert Yuen In Lam, Robbe Pincket, Hao Han, Xing Er Ong, Zechen Wang, Jamie Hinks, Yanjie Wei, Weifeng Li, Liangzhen Zheng, and Yuguang Mu
Springer Science and Business Media LLC
Irene Julca, Daniela Mutwil‐Anderwald, Vaishnervi Manoj, Zahra Khan, Soak Kuan Lai, Lay K. Yang, Ing T. Beh, Jerzy Dziekan, Yoon P. Lim, Shen K. Lim,et al.
Wiley
Plants accumulate a vast array of secondary metabolites, which constitute a natural resource for pharmaceuticals. Oldenlandia corymbosa belongs to the Rubiaceae family, and has been used in traditional medicine to treat different diseases, including cancer. However, the active metabolites of the plant, their biosynthetic pathway and mode of action in cancer are unknown. To fill these gaps, we exposed this plant to eight different stress conditions and combined different omics data capturing gene expression, metabolic profiles and anti-cancer activity. Our results show that O. corymbosa extracts are active against breast cancer cell lines and that ursolic acid is responsible for this activity. Moreover, we assembled a high-quality genome and uncovered two genes involved in the biosynthesis of ursolic acid. Finally, we also revealed that ursolic acid causes mitotic catastrophe in cancer cells and identified three high-confidence protein binding targets by Cellular Thermal Shift Assay (CETSA) and reverse docking. Altogether, these results constitute a valuable resource to further characterize the biosynthesis of active metabolites in the Oldenlandia group, while the mode of action of ursolic acid will allow us to further develop this valuable compound. This article is protected by copyright. All rights reserved.
Melvin Yong, Zhi Y. Kok, Chong H. Koh, Wenbin Zhong, Justin TY. Ng, Yuguang Mu, Mary B. Chan-Park, and Yunn-Hwen Gan
American Society for Microbiology
The treatment of bacterial infections is becoming increasingly challenging with the emergence of antimicrobial resistance. Thus, the development of antimicrobials with novel mechanisms of action is much needed.
Bingchen Che, Dan Sun, Chen Zhang, Jiaqing Hou, Wei Zhao, Guangyin Jing, Yuguang Mu, Yaoyu Cao, Liang Dai, and Ce Zhang
American Chemical Society (ACS)
Mechanically induced chromosome reorganization plays important roles in transcriptional regulation. However, the interplay between chromosome reorganization and transcription activities is complicated, such that it is difficult to decipher the regulatory effects of intranuclear geometrical cues. Here, we simplify the system by introducing DNA, packaging proteins (i.e., histone and protamine), and transcription factor NF-κB into a well-defined fluidic chip with changing spatical confinement ranging from 100 to 500 nm. It is uncovered that strong nanoconfinement suppresses higher-order folding of histone- and protamine-DNA complexes, the fracture of which exposes buried DNA segments and causes increased quantities of NF-κB binding to the DNA chain. Overall, these results reveal a pathway of how intranuclear geometrical cues alter the open/closed state of a DNA-protein complex and therefore affect transcription activities: i.e., NF-κB binding.
Shu Hui Hiew, Yang Lu, Hao Han, Rui A Gonçalves, Serena Rosa Alfarano, Raffaele Mezzenga, Atul N. Parikh, Yuguang Mu, and Ali Miserez
American Chemical Society (ACS)
The occurrence of modular peptide repeats in load-bearing (structural) proteins is common in nature, with distinctive peptide sequences that often remain conserved across different phylogenetic lineages. These highly conserved peptide sequences endow specific mechanical properties to the material, such as toughness or elasticity. Here, using bioinformatic tools and phylogenetic analysis, we have identified the GX8 peptide with the sequence GLYGGYGX (where X can be any residue) in a wide range of organisms. By simple mutation of the X residue, we demonstrate that GX8 can be self-assembled into various supramolecular structures, exhibiting vastly different physicochemical and viscoelastic properties, from liquid-like coacervate microdroplets to hydrogels to stiff solid materials. A combination of spectroscopic, electron microscopy, mechanical, and molecular dynamics studies is employed to obtain insights into molecular scale interactions driving self-assembly of GX8 peptides, underscoring that π-π stacking and hydrophobic interactions are the drivers of peptide self-assembly, whereas the X residue determines the extent of hydrogen bonding that regulates the macroscopic mechanical response. This study highlights the ability of single amino-acid polymorphism to tune the supramolecular assembly and bulk material properties of GX8 peptides, enabling us to cover a broad range of potential biomedical applications such as hydrogels for tissue engineering or coacervates for drug delivery.
Zechen Wang, Liangzhen Zheng, Sheng Wang, Mingzhi Lin, Zhihao Wang, Adams Wai-Kin Kong, Yuguang Mu, Yanjie Wei, and Weifeng Li
Oxford University Press (OUP)
Abstract The recently reported machine learning- or deep learning-based scoring functions (SFs) have shown exciting performance in predicting protein–ligand binding affinities with fruitful application prospects. However, the differentiation between highly similar ligand conformations, including the native binding pose (the global energy minimum state), remains challenging that could greatly enhance the docking. In this work, we propose a fully differentiable, end-to-end framework for ligand pose optimization based on a hybrid SF called DeepRMSD+Vina combined with a multi-layer perceptron (DeepRMSD) and the traditional AutoDock Vina SF. The DeepRMSD+Vina, which combines (1) the root mean square deviation (RMSD) of the docking pose with respect to the native pose and (2) the AutoDock Vina score, is fully differentiable; thus is capable of optimizing the ligand binding pose to the energy-lowest conformation. Evaluated by the CASF-2016 docking power dataset, the DeepRMSD+Vina reaches a success rate of 94.4%, which outperforms most reported SFs to date. We evaluated the ligand conformation optimization framework in practical molecular docking scenarios (redocking and cross-docking tasks), revealing the high potentialities of this framework in drug design and discovery. Structural analysis shows that this framework has the ability to identify key physical interactions in protein–ligand binding, such as hydrogen-bonding. Our work provides a paradigm for optimizing ligand conformations based on deep learning algorithms. The DeepRMSD+Vina model and the optimization framework are available at GitHub repository https://github.com/zchwang/DeepRMSD-Vina_Optimization.
Wei Zhang, Yuanyuan Gou, Li Cheng, Kaiwei Dong, Yijie Sheng, Chao Ye, Xianqing Yang, and Yuguang Mu
Royal Society of Chemistry (RSC)
The disruption of phosphorene oxide (PO) nanosheets to the protein structure is enhanced with increasing oxidation concentration of PO, while PO’s oxidation mode has very little effect on the PO-HP35 interaction.
Wahyu Surya, Maria Queralt-Martin, Yuguang Mu, Vicente M. Aguilella, and Jaume Torres
Springer Science and Business Media LLC
AbstractA global pandemic is underway caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The SARS-CoV-2 genome, like its predecessor SARS-CoV, contains open reading frames that encode accessory proteins involved in virus-host interactions active during infection and which likely contribute to pathogenesis. One of these accessory proteins is 7b, with only 44 (SARS-CoV) and 43 (SARS-CoV-2) residues. It has one predicted transmembrane domain fully conserved, which suggests a functional role, whereas most variability is contained in the predicted cytoplasmic C-terminus. In SARS-CoV, 7b protein is expressed in infected cells, and the transmembrane domain was necessary and sufficient for Golgi localization. Also, anti-p7b antibodies have been found in the sera of SARS-CoV convalescent patients. In the present study, we have investigated the hypothesis that SARS-2 7b protein forms oligomers with ion channel activity. We show that in both SARS viruses 7b is almost completely α-helical and has a single transmembrane domain. In SDS, 7b forms various oligomers, from monomers to tetramers, but only monomers when exposed to reductants. Combination of SDS gel electrophoresis and analytical ultracentrifugation (AUC) in both equilibrium and velocity modes suggests a dimer-tetramer equilibrium, but a monomer–dimer–tetramer equilibrium in the presence of reductant. This data suggests that although disulfide-linked dimers may be present, they are not essential to form tetramers. Inclusion of pentamers or higher oligomers in the SARS-2 7b model were detrimental to fit quality. Preliminary models of this association was generated with AlphaFold2, and two alternative models were exposed to a molecular dynamics simulation in presence of a model lipid membrane. However, neither of the two models provided any evident pathway for ions. To confirm this, SARS-2 p7b was studied using Planar Bilayer Electrophysiology. Addition of p7b to model membranes produced occasional membrane permeabilization, but this was not consistent with bona fide ion channels made of a tetrameric assembly of α-helices.
Jie Shen, Yongting Gu, Lingjie Ke, Qiuping Zhang, Yin Cao, Yuchao Lin, Zhen Wu, Caisheng Wu, Yuguang Mu, Yun-Long Wu,et al.
Springer Science and Business Media LLC
AbstractCholesterol-enhanced pore formation is one evolutionary means cholesterol-free bacterial cells utilize to specifically target cholesterol-rich eukaryotic cells, thus escaping the toxicity these membrane-lytic pores might have brought onto themselves. Here, we present a class of artificial cholesterol-dependent nanopores, manifesting nanopore formation sensitivity, up-regulated by cholesterol of up to 50 mol% (relative to the lipid molecules). The high modularity in the amphiphilic molecular backbone enables a facile tuning of pore size and consequently channel activity. Possessing a nano-sized cavity of ~ 1.6 nm in diameter, our most active channel Ch-C1 can transport nanometer-sized molecules as large as 5(6)-carboxyfluorescein and display potent anticancer activity (IC50 = 3.8 µM) toward human hepatocellular carcinomas, with high selectivity index values of 12.5 and >130 against normal human liver and kidney cells, respectively.
Hilbert Yuen In Lam, Jia Sheng Guan, and Yuguang Mu
MDPI AG
Monkeypox is an emerging epidemic of concern. The disease is caused by the monkeypox virus and an increasing global incidence with a 2022 outbreak that has spread to Europe amid the COVID-19 pandemic. The new outbreak is associated with novel, previously undiscovered mutations and variants. Currently, the US Food and Drug Administration (FDA) approved poxvirus treatment involves the use of tecovirimat. However, there is otherwise limited pharmacopoeia and research interest in monkeypox. In this study, virtual screening and molecular dynamics were employed to explore the potential repurposing of multiple drugs previously approved by the FDA or other jurisdictions for other applications. Several drugs are predicted to tightly bind to viral proteins, which are crucial in viral replication, including molecules which show high potential for binding the monkeypox D13L capsid protein, whose inhibition has previously been demonstrated to suppress viral replication.
Ezekiel Ze Ken Cheong, Jun Ping Quek, Liu Xin, Chaoqiang Li, Jing Yi Chan, Chong Wai Liew, Yuguang Mu, Jie Zheng, and Dahai Luo
Elsevier BV
Kanagavel Deepankumar, Qi Guo, Harini Mohanram, Jessica Lim, Yuguang Mu, Konstantin Pervushin, Jing Yu, and Ali Miserez
Wiley
The underwater adhesive prowess of aquatic mussels has been largely attributed to the abundant post-translationally modified amino acid l-3,4-dihydroxyphenylalanine (Dopa) in mussel foot proteins (MFPs) that make up their adhesive threads. More recently, it has been suggested that during thread fabrication, MFPs form intermediate fluidic phases such as liquid crystals or coacervates regulated by a liquid-liquid phase separation (LLPS) process. Here, it is shown that Dopa plays another central role during mussel fiber formation, by enabling LLPS of Pvfp-5β, a main MFP of the green mussel Perna viridis. Using residue-specific substitution of Tyrosine (Tyr) for Dopa during recombinant expression, Dopa-substituted Pvfp-5β is shown to exhibit LLPS under seawater-like conditions, whereas the Tyr-only version forms insoluble aggregates. Combining quantum chemistry calculations and solution NMR, a transient H-bonding network requiring the two hydroxyl groups of Dopa is found to be critical to enable LLPS in Dopa-mutated Pvfp-5β. Overall, the study suggests that Dopa plays an important role in regulating LLPS of MFPs, which may be critical to concentrate the adhesive proteins at the plaque/substrate interface and therefore produce a more robust adhesive. The findings also provide molecular-level lessons to guide biomanufacturing of protein-based materials such as bioadhesives and load-bearing fibers.
Liangzhen Zheng, Jintao Meng, Kai Jiang, Haidong Lan, Zechen Wang, Mingzhi Lin, Weifeng Li, Hongwei Guo, Yanjie Wei, and Yuguang Mu
Oxford University Press (OUP)
Abstract Scoring functions are important components in molecular docking for structure-based drug discovery. Traditional scoring functions, generally empirical- or force field-based, are robust and have proven to be useful for identifying hits and lead optimizations. Although multiple highly accurate deep learning- or machine learning-based scoring functions have been developed, their direct applications for docking and screening are limited. We describe a novel strategy to develop a reliable protein–ligand scoring function by augmenting the traditional scoring function Vina score using a correction term (OnionNet-SFCT). The correction term is developed based on an AdaBoost random forest model, utilizing multiple layers of contacts formed between protein residues and ligand atoms. In addition to the Vina score, the model considerably enhances the AutoDock Vina prediction abilities for docking and screening tasks based on different benchmarks (such as cross-docking dataset, CASF-2016, DUD-E and DUD-AD). Furthermore, our model could be combined with multiple docking applications to increase pose selection accuracies and screening abilities, indicating its wide usage for structure-based drug discoveries. Furthermore, in a reverse practice, the combined scoring strategy successfully identified multiple known receptors of a plant hormone. To summarize, the results show that the combination of data-driven model (OnionNet-SFCT) and empirical scoring function (Vina score) is a good scoring strategy that could be useful for structure-based drug discoveries and potentially target fishing in future.
Yunju Zhang, Liangzhen Zheng, Yanmei Yang, Yuanyuan Qu, Yong-Qiang Li, Mingwen Zhao, Yuguang Mu, and Weifeng Li
Royal Society of Chemistry (RSC)
Quantitative energy decomposition analysis of the main proteinase of SARS-CoV-2 revealed key residues that mediate the dimerization process. This provides important targets for the design of anti-SARS-CoV-2 drugs through inhibiting activity.
Thang Cong Do, Jun Wei Lau, Caixia Sun, Songhan Liu, Khoa Tuan Kha, Seok Ting Lim, Yu Yang Oon, Yuet Ping Kwan, Jia Jia Ma, Yuguang Mu,et al.
American Association for the Advancement of Science (AAAS)
Epigenetic mediation through bromodomain and extraterminal (BET) proteins have progressively translated protein imbalance into effective cancer treatment. Perturbation of druggable BET proteins through proteolysis-targeting chimeras (PROTACs) has recently contributed to the discovery of effective therapeutics. Unfortunately, precise and microenvironment-activatable BET protein degradation content with promising tumor selectivity and pharmacological suitability remains elusive. Here, we present an enzyme-derived clicking PROTACs (ENCTACs) capable of orthogonally cross-linking two disparate small-molecule warhead ligands that recognize BET bromodomain-containing protein 4 (BRD4) protein and E3 ligase within tumors only upon hypoxia-induced activation of nitroreductase enzyme. This localized formation of heterobifunctional degraders promotes specific down-regulation of BRD4, which subsequently alters expression of epigenetic targets and, therefore, allows precise modulation of hypoxic signaling in live cells, zebrafish, and living mice with solid tumors. Our activation-feedback system demonstrates compelling superiorities and may enable the PROTAC technology with more flexible practicality and druggable potency for precision medicine in the near future.
Neslihan A. Kaya, Jianbin Chen, Hannah Lai, Hechuan Yang, Liang Ma, Xiaodong Liu, Jacob Santiago Alvarez, Jin Liu, Axel M. Hillmer, David Tai,et al.
Ivyspring International Publisher
Hepatocellular carcinoma (HCC) is one of the deadliest cancer types with diverse etiological factors across the world. Although large scale genomic studies have been conducted in different countries, integrative analysis of HCC genomes and ethnic comparison across cohorts are lacking. Methods: We first integrated genomes of 1,349 HCC patients from five large cohorts across the world and applied multiple statistical methods in identifying driver genes. Subsequently, we systematically compared HCC genomes and transcriptomes between Asians and Europeans using the TCGA cohort. Results: We identified 29 novel candidate driver genes, many of which are infrequent tumor suppressors driving late-stage tumor progression. When we systematically compared ethnic differences in the genomic landscape between Asian and European HCCs using the TCGA cohort (n = 348), we found little differences in driver frequencies. Through multi-modal integrative analysis, we found higher genomic instability in Asians together with a collection of molecular events ranging from tumor mutation burden (TMB), copy number alterations as well as transcriptomic subtypes segregating distinctively between two ethnic backgrounds. Strikingly, we identified an Asian specific transcriptomic subtype with multiple ethnically enriched genomic alterations, in particular chromosome 16 deletion, leading to a clinically aggressive RNA subgroup unique to Asians. Integrating multi-modal information, we found that survival models predict patient prognosis much better in Asians than in Europeans, demonstrating a higher potential for precision medicine applications in Asia. Conclusion: For the first time, we have uncovered an unprecedented amount of genomic differences segregating distinctively across ethnicities in HCC and highlighted the importance of differential disease biology and management in HCC across ethnic backgrounds.
Li Na Zhao, Dibyendu Mondal, Weifeng Li, Yuguang Mu, and Philipp Kaldis
Wiley
Lignin is one of the world's most abundant organic polymers, and 2-pyrone-4,6-dicarboxylate lactonase (LigI) catalyzes the hydrolysis of 2-pyrone-4,6-dicarboxylate (PDC) in the degradation of lignin. The pH has profound effects on enzyme catalysis and therefore we studied this in the context of LigI. We found that changes of the pH mostly affects surface residues, while the residues at the active site are more subject to changes of the surrounding microenvironment. In accordance with this, a high pH facilitates the deprotonation of the substrate. Detailed free energy calculations by the empirical valence bond (EVB) approach revealed that the overall hydrolysis reaction is more likely when the three active site histidines (His31, His33 and His180) are protonated at the &ip.eop; site, however, protonation at the δ site may be favored during specific steps of the reaction. Our studies have uncovered the determinant role of the protonation state of the active site residues His31, His33 and His180 in the hydrolysis of PDC. This article is protected by copyright. All rights reserved.
Antonín Brisuda, James C. S. Ho, Pancham S. Kandiyal, Justin T-Y. Ng, Ines Ambite, Daniel S. C. Butler, Jaromir Háček, Murphy Lam Yim Wan, Thi Hien Tran, Aftab Nadeem,et al.
Springer Science and Business Media LLC
AbstractPartially unfolded alpha-lactalbumin forms the oleic acid complex HAMLET, with potent tumoricidal activity. Here we define a peptide-based molecular approach for targeting and killing tumor cells, and evidence of its clinical potential (ClinicalTrials.gov NCT03560479). A 39-residue alpha-helical peptide from alpha-lactalbumin is shown to gain lethality for tumor cells by forming oleic acid complexes (alpha1-oleate). Nuclear magnetic resonance measurements and computational simulations reveal a lipid core surrounded by conformationally fluid, alpha-helical peptide motifs. In a single center, placebo controlled, double blinded Phase I/II interventional clinical trial of non-muscle invasive bladder cancer, all primary end points of safety and efficacy of alpha1-oleate treatment are reached, as evaluated in an interim analysis. Intra-vesical instillations of alpha1-oleate triggers massive shedding of tumor cells and the tumor size is reduced but no drug-related side effects are detected (primary endpoints). Shed cells contain alpha1-oleate, treated tumors show evidence of apoptosis and the expression of cancer-related genes is inhibited (secondary endpoints). The results are especially encouraging for bladder cancer, where therapeutic failures and high recurrence rates create a great, unmet medical need.
Zechen Wang, Liangzhen Zheng, Yang Liu, Yuanyuan Qu, Yong-Qiang Li, Mingwen Zhao, Yuguang Mu, and Weifeng Li
Frontiers Media SA
One key task in virtual screening is to accurately predict the binding affinity (△G) of protein-ligand complexes. Recently, deep learning (DL) has significantly increased the predicting accuracy of scoring functions due to the extraordinary ability of DL to extract useful features from raw data. Nevertheless, more efforts still need to be paid in many aspects, for the aim of increasing prediction accuracy and decreasing computational cost. In this study, we proposed a simple scoring function (called OnionNet-2) based on convolutional neural network to predict △G. The protein-ligand interactions are characterized by the number of contacts between protein residues and ligand atoms in multiple distance shells. Compared to published models, the efficacy of OnionNet-2 is demonstrated to be the best for two widely used datasets CASF-2016 and CASF-2013 benchmarks. The OnionNet-2 model was further verified by non-experimental decoy structures from docking program and the CSAR NRC-HiQ data set (a high-quality data set provided by CSAR), which showed great success. Thus, our study provides a simple but efficient scoring function for predicting protein-ligand binding free energy.