Drug Discovery Research Articles

Bridging the gap between target-based and phenotypic-based drug discovery.

The unparalleled progress in science of the last decades has brought a better understanding of the molecular mechanisms of diseases. This promoted drug discovery processes based on a target approach. However, despite the high promises associated, a critical decrease in the number of first-in-class drugs has been observed. This review analyses the challenges, advances, and opportunities associated with the main strategies of the drug discovery process, i.e. based on a rational target approach and on an empirical phenotypic approach. This review also evaluates how the gap between these two crossroads can be bridged toward a more efficient drug discovery process. The critical lack of knowledge of the complex biological networks is leading to targets not relevant for the clinical context or to drugs that present undesired adverse effects. The phenotypic systems designed by considering available molecular mechanisms can mitigate these knowledge gaps. Associated with the expansion of the chemical space and other technologies, these designs can lead to more efficient drug discoveries. Technological and scientific knowledge should also be applied to identify, as early as possible, both drug targets and mechanisms of action, leading to a more efficient drug discovery pipeline.

Differential Gene Analysis of Langerhans Cell Histiocytosis and the Significance of MMP1-Targeted Drug Repositioning.

Langerhans cell histiocytosis (LCH) is a rare condition predominantly affecting young children. Activation of the MAPK pathway has offered key new insights into the pathogenesis of LCH; however, the precise mechanisms underlying its occurrence and development are still far from being completely elucidated. There is still a relapse/reactivation rate in patients with multisystem LCH. Therefore, this study aimed to investigate other potential LCH pathophysiologies and prospective therapeutic targets. The gene expression omnibus (GEO) database was used to retrieve gene expression profiles of LCH (GSE16395). Three distinct types of analyses were performed after identifying the common differentially expressed genes (DEGs) in LCH: hub gene identification, functional annotation, module construction, drug repositioning, and expression analysis via immunohistochemistry (IHC). We identified 417 common DEGs and 50 central hub genes. This functional study highlighted the significance of keratinization, skin development, and inflammation. In addition, we predicted new drug candidates (RS2 drugs targeting matrix metalloprotease1, MMP1) that could be used for LCH treatment. Finally, gene-miRNA and gene-TF networks and immune cell infiltration were analyzed for MMP1-related genes. MMP1 expression levels in LCH tissues were validated by IHC. Our study identified the central communal genes and novel drug candidates. These shared pathways and hub genes offer new perspectives on future mechanisms of action and therapeutic targets.

Enantioselective Synthesis of Sulfinamidines via Asymmetric Nitrogen Transfer from N-H Oxaziridines to Sulfenamides.

Sulfinamidines are promising aza-SIV chiral building blocks in asymmetric synthesis and drug discovery. However, no report has documented their enantioselective synthesis. Here we present an enantioselective synthesis of sulfinamidines via electrophilic amination of sulfenamides using an enantiopure N-H oxaziridine. The resulting enantiomerically enriched primary sulfinamidines are configurationally stable at 90 oC in solution and show remarkable stability against organic acids and bases under non-aqueous conditions. We also demonstrate a one-pot, three-component, enantioselective synthesis of sulfinamides using N-H oxaziridine reagents.

Hits and misses with animal models of narcolepsy and the implications for drug discovery.

Narcolepsy is a chronic and rare neurological disorder characterized by disordered sleep. Based on animal models and further research in humans, the dysfunctional orexin system was identified as a contributing factor to the pathophysiology of narcolepsy. Animal models played a larger role in the discovery of some of the pharmacological agents with established benefit/risk profiles. In this review, the authors examine the phenotypes observed in animal models of narcolepsy and the characteristics of clinically used pharmacological agents in these animal models. Additionally, the authors compare the effects of clinically used pharmacological agents on the phenotypes in animal models with those observed in narcolepsy patients. Research in canine and mouse models have linked narcolepsy to the O×R2mutation and orexin deficiency, leading to new diagnostic criteria and a drug development focus. Advancements in pharmacological therapies have significantly improved narcolepsy management, with insights from both clinical experience and from animal models having led to new treatments such as low sodium oxybate and solriamfetol. However, challenges persist in addressing symptoms beyond excessive daytime sleepiness and cataplexy, highlighting the need for further research, including the development of diurnal animal models to enhance understanding and treatment options for narcolepsy.

2-APQC, a small-molecule activator of Sirtuin-3 (SIRT3), alleviates myocardial hypertrophy and fibrosis by regulating mitochondrial homeostasis

Sirtuin 3 (SIRT3) is well known as a conserved nicotinamide adenine dinucleotide+ (NAD+)-dependent deacetylase located in the mitochondria that may regulate oxidative stress, catabolism and ATP production. Accumulating evidence has recently revealed that SIRT3 plays its critical roles in cardiac fibrosis, myocardial fibrosis and even heart failure (HF), through its deacetylation modifications. Accordingly, discovery of SIRT3 activators and elucidating their underlying mechanisms of HF should be urgently needed. Herein, we identified a new small-molecule activator of SIRT3 (named 2-APQC) by the structure-based drug designing strategy. 2-APQC was shown to alleviate isoproterenol (ISO)-induced cardiac hypertrophy and myocardial fibrosis in vitro and in vivo rat models. Importantly, in SIRT3 knockout mice, 2-APQC could not relieve HF, suggesting that 2-APQC is dependent on SIRT3 for its protective role. Mechanically, 2-APQC was found to inhibit the mammalian target of rapamycin (mTOR)-p70 ribosomal protein S6 kinase (p70S6K), c-jun N-terminal kinase (JNK) and transforming growth factor-β (TGF-β)/ small mother against decapentaplegic 3 (Smad3) pathways to improve ISO-induced cardiac hypertrophy and myocardial fibrosis. Based upon RNA-seq analyses, we demonstrated that SIRT3-pyrroline-5-carboxylate reductase 1 (PYCR1) axis was closely assoiated with HF. By activating PYCR1, 2-APQC was shown to enhance mitochondrial proline metabolism, inhibited reactive oxygen species (ROS)-p38 mitogen activated protein kinase (p38MAPK) pathway and thereby protecting against ISO-induced mitochondrialoxidative damage. Moreover, activation of SIRT3 by 2-APQC could facilitate AMP-activated protein kinase (AMPK)-Parkin axis to inhibit ISO-induced necrosis. Together, our results demonstrate that 2-APQC is a targeted SIRT3 activator that alleviates myocardial hypertrophy and fibrosis by regulating mitochondrial homeostasis, which may provide a new clue on exploiting a promising drug candidate for the future HF therapeutics.

Thyroid hormone analogues: Promising therapeutic avenues to improve the neurodevelopmental outcomes of intrauterine growth restriction.

Intrauterine growth restriction (IUGR) is a pregnancy complication impairing fetal growth and development. The compromised development is often attributed to disruptions of oxygen and nutrient supply from the placenta, resulting in a number of unfavourable physiological outcomes with impaired brain and organ growth. IUGR is associated with compromised development of both grey and white matter, predisposing the infant to adverse neurodevelopmental outcomes, including long-lasting cognitive and motor difficulties. Cerebral thyroid hormone (TH) signalling, which plays a crucial role in regulating white and grey matter development, is dysregulated in IUGR, potentially contributing to the neurodevelopmental delays associated with this condition. Notably, one of the major TH transporters, monocarboxylate transporter-8 (MCT8), is deficient in the fetal IUGR brain. Currently, no effective treatment to prevent or reverse IUGR exists. Management strategies involve close antenatal monitoring, management of maternal risk factors if present and early delivery if IUGR is found to be severe or worsening in utero. The overall goal is to determine the most appropriate time for delivery, balancing the risks of preterm birth with further fetal compromise due to IUGR. Drug candidates have shown either adverse effects or little to no benefits in this vulnerable population, urging further preclinical and clinical investigation to establish effective therapies. In this review, we discuss the major neuropathology of IUGR driven by uteroplacental insufficiency and the concomitant long-term neurobehavioural impairments in individuals born IUGR. Importantly, we review the existing clinical and preclinical literature on cerebral TH signalling deficits, particularly the impaired expression of MCT8 and their correlation with IUGR. Lastly, we discuss the current evidence on MCT8-independent TH analogues which mimic the brain actions of THs by being metabolised in a similar manner as promising, albeit underappreciated approaches to promote grey and white matter development and improve the neurobehavioural outcomes following IUGR.

TransFOL: A Logical Query Model for Complex Relational Reasoning in Drug-Drug Interaction.

Predicting drug-drug interaction (DDI) plays a crucial role in drug recommendation and discovery. However, wet lab methods are prohibitively expensive and time-consuming due to drug interactions. In recent years, deep learning methods have gained widespread use in drug reasoning. Although these methods have demonstrated effectiveness, they can only predict the interaction between a drug pair and do not contain any other information. However, DDI is greatly affected by various other biomedical factors (such as the dose of the drug). As a result, it is challenging to apply them to more complex and meaningful reasoning tasks. Therefore, this study regards DDI as a link prediction problem on knowledge graphs and proposes a DDI prediction model based on Cross-Transformer and Graph Convolutional Networks (GCN) in first-order logical query form, TransFOL. In the model, a biomedical query graph is first built to learn the embedding representation. Subsequently, an enhancement module is designed to aggregate the semantics of entities and relations. Cross-Transformer is used for encoding to obtain semantic information between nodes, and GCN is used to gather neighbour information further and predict inference results. To evaluate the performance of TransFOL on common DDI tasks, we conduct experiments on two benchmark datasets. The experimental results indicate that our model outperforms state-of-the-art methods on traditional DDI tasks. Additionally, we introduce different biomedical information in the other two experiments to make the settings more realistic. Experimental results verify the strong drug reasoning ability and generalization of TransFOL in complex settings. Data and code are available at https://github.com/Cheng0829/TransFOL.

DeepPepPI: A deep cross-dependent framework with information sharing mechanism for predicting plant peptide-protein interactions

Peptide-protein interactions (PepPIs) play a crucial role in various fundamental biological activities in plants. As traditional experimental technologies are usually time-consuming and labor-intensive, it is indispensable to develop computational methods for PepPI identification, which has potential applications across diverse fields including food safety and peptide drug discovery. However, existing methods have received relatively little attention in adequately representing the interaction information between peptides and proteins, consequently limiting the prediction performance. To this end, we present DeepPepPI, a novel deep cross-dependent framework for accurate prediction of plant PepPIs. Concretely, a data-driven context-embedded representation (DCR) module is developed as the peptide feature extractor that can capture rich contextual semantic information, even from short sequences. To fully characterize inherent properties of proteins, we propose a bi-level self-correlation search (BSS) module, which integrates the primary sequence and secondary structure into a unified space to learn their potential relationships. In addition, a cross-dependent feature integration (CFI) module with information sharing mechanism is introduced, aimed at providing a comprehensive feature representation to portray the intricate interaction patterns between peptides and proteins from a global perspective. Extensive experimental results conducted on the benchmark datasets show that DeepPepPI achieves superior prediction performance compared to other state-of-the-art methods.

AIGO-DTI: Predicting Drug-Target Interactions Based on Improved Drug Properties Combined with Adaptive Iterative Algorithms.

Artificial intelligence-based methods for predicting drug-target interactions (DTIs) aim to explore reliable drug candidate targets rapidly and cost-effectively to accelerate the drug development process. However, current methods are often limited by the topological regularities of drug molecules, making them difficult to generalize to a broader chemical space. Additionally, the use of similarity to measure DTI network links often introduces noise, leading to false DTI relationships and affecting the prediction accuracy. To address these issues, this study proposes an Adaptive Iterative Graph Optimization (AIGO)-DTI prediction framework. This framework integrates atomic cluster information and enhances molecular features through the design of functional group prompts and graph encoders, optimizing the construction of DTI association networks. Furthermore, the optimization of graph structure is transformed into a node similarity learning problem, utilizing multihead similarity metric functions to iteratively update the network structure to improve the quality of DTI information. Experimental results demonstrate the outstanding performance of AIGO-DTI on multiple public data sets and label reversal data sets. Case studies, molecular docking, and existing research validate its effectiveness and reliability. Overall, the method proposed in this study can construct comprehensive and reliable DTI association network information, providing new graphing and optimization strategies for DTI prediction, which contribute to efficient drug development and reduce target discovery costs.

The hydrazine moiety in the synthesis of modified nucleosides and nucleotides.

Synthetic nucleoside mimics are re-emerging as crucial contenders for antiviral and anticancer medications. While, Ribavirin stands out for its unique antiviral properties, predominantly associated with its distinctive triazole heterocycle as a nucleobase, the exploration of alternative nitrogen-based aromatic heterocycles hold great promises for the discovery of novel bioactive nucleoside mimics. Although nucleoside derivatives synthesised from hydrazine-ribose units have been in development for many decades, they have been little evaluated biologically and even less for their antiviral properties. With the aim of taking a closer look at these under-explored derivatives and investigating their synthetic pathways, this concise review provides an overview of the molecular design, the chemical synthesis, and the biological activity, when available, of these nucleoside analogues. Overall, the entire body of work already done motivates further exploration of theses analogues and encourages us of formulating structurally novel nucleoside drug candidates featuring innovative mode of action.

In Silico Search for Drug Candidates Targeting the PAX8–PPARγ Fusion Protein in Thyroid Cancer

In Silico Search for Drug Candidates Targeting the PAX8–PPARγ Fusion Protein in Thyroid Cancer

Structural and mechanistic insights into Quinolone Synthase to address its functional promiscuity

Quinolone synthase from Aegle marmelos (AmQNS) is a type III polyketide synthase that yields therapeutically effective quinolone and acridone compounds. Addressing the structural and molecular underpinnings of AmQNS and its substrate interaction in terms of its high selectivity and specificity can aid in the development of numerous novel compounds. This paper presents a high-resolution AmQNS crystal structure and explains its mechanistic role in synthetic selectivity. Additionally, we provide a model framework to comprehend structural constraints on ketide insertion and postulate that AmQNS’s steric and electrostatic selectivity plays a role in its ability to bind to various core substrates, resulting in its synthetic diversity. AmQNS prefers quinolone synthesis and can accommodate large substrates because of its wide active site entrance. However, our research suggests that acridone is exclusively synthesized in the presence of high malonyl-CoA concentrations. Potential implications of functionally relevant residue mutations were also investigated, which will assist in harnessing the benefits of mutations for targeted polyketide production. The pharmaceutical industry stands to gain from these findings as they expand the pool of potential drug candidates, and these methodologies can also be applied to additional promising enzymes.

Homologous-targeting biomimetic nanoparticles co-loaded with melittin and a photosensitizer for the combination therapy of triple negative breast cancer.

Melittin (Mel) is considered a promising candidate drug for the treatment of triple negative breast cancer (TNBC) due to its various antitumor effects. However, its clinical application is hampered by notable limitations, including hemolytic activity, rapid clearance, and a lack of tumor selectivity. Here, we designed novel biomimetic nanoparticles based on homologous tumor cell membranes and poly(lactic-co-glycolic acid) (PLGA)/poly(beta-aminoester) (PBAE), denoted MDM@TPP, which efficiently coloaded the cytolytic peptide Mel and the photosensitizer mTHPC. Both in vitro and in vivo, the MDM@TPP nanoparticles effectively mitigated the acute toxicity of melittin and exhibited strong TNBC targeting ability due to the homologous targeting effect of the tumor cell membrane. Under laser irradiation, the MDM@TPP nanoparticles showed excellent photodynamic performance and thus accelerated the release of Mel by disrupting cell membrane integrity. Moreover, Mel combined with photodynamic therapy (PDT) can synergistically kill tumor cells and induce significant immunogenic cell death, thereby stimulating the maturation of dendritic cells (DCs). In 4T1 tumor-bearing mice, MDM@TPP nanoparticles effectively inhibited the growth and metastasis of primary tumors and finally prevented tumor recurrence by improving the immune response.

Deuterated Alkyl Sulfonium Salt Reagents; Importance of H/D Exchange Methods in Drug Discovery.

Deuterated drugs (heavy drugs) have recently been spotlighted as a new modality for small-molecule drugs because the pharmacokinetics of pharmaceutical drugs can be enhanced by replacing C-H bonds with more stable C-D bonds at metabolic positions. Therefore, deuteration methods for drug candidates are a hot topic in medicinal chemistry. Among them, the H/D exchange reaction (direct transformation of C-H bonds to C-D bonds) is a useful and straightforward method for creating novel deuterated target molecules, and over 20 reviews on the synthetic methods related to H/D exchange reactions have been published in recent years. Although various deuterated drug candidates undergo clinical trials, approved deuterated drugs possess CD3 groups in the same molecule. However, less diversification, except for the CD3 group, is a problem for future medicinal chemistry. Recently, we developed various deuterated alkyl (dn-alkyl) sulfonium salts based on the H/D exchange reaction of the corresponding hydrogen form using D2O as an inexpensive deuterium source to introduce CD3, CH3CD2, and ArCH2CD2 groups into drug candidates. This concept summarises recent reviews related to H/D exchange reactions and novel reagents that introduce the CD3 group, and our newly developed electrophilic dn-alkyl reagents are discussed.

Reductionism and proxy failure: From neuroscience to target-based drug discovery.

Reductionist methodologies reduce phenomena to some of their lower-level components. Researchers gradually shift their focus away from observing the actual object of study toward investigating and optimizing such lower-level proxies. Following reductionism, these proxies progressively diverge further from the original object of study. We vividly illustrate this in the evolution of target-based drug discovery from rational and phenotypic drug discovery.

Quantifying the Hardness of Bioactivity Prediction Tasks for Transfer Learning.

Today, machine learning methods are widely employed in drug discovery. However, the chronic lack of data continues to hamper their further development, validation, and application. Several modern strategies aim to mitigate the challenges associated with data scarcity by learning from data on related tasks. These knowledge-sharing approaches encompass transfer learning, multitask learning, and meta-learning. A key question remaining to be answered for these approaches is about the extent to which their performance can benefit from the relatedness of available source (training) tasks; in other words, how difficult ("hard") a test task is to a model, given the available source tasks. This study introduces a new method for quantifying and predicting the hardness of a bioactivity prediction task based on its relation to the available training tasks. The approach involves the generation of protein and chemical representations and the calculation of distances between the bioactivity prediction task and the available training tasks. In the example of meta-learning on the FS-Mol data set, we demonstrate that the proposed task hardness metric is inversely correlated with performance (Pearson's correlation coefficient r = -0.72). The metric will be useful in estimating the task-specific gain in performance that can be achieved through meta-learning.

CNSMolGen: A Bidirectional Recurrent Neural Network-Based Generative Model for De Novo Central Nervous System Drug Design.

Central nervous system (CNS) drugs have had a significant impact on treating a wide range of neurodegenerative and psychiatric disorders. In recent years, deep learning-based generative models have shown great potential for accelerating drug discovery and improving efficacy. However, specific applications of these techniques in CNS drug discovery have not been widely reported. In this study, we developed the CNSMolGen model, which uses a framework of bidirectional recurrent neural networks (Bi-RNNs) for de novo molecular design of CNS drugs. Results showed that the pretrained model was able to generate more than 90% of completely new molecular structures, which possessed the properties of CNS drug molecules and were synthesizable. In addition, transfer learning was performed on small data sets with specific biological activities to evaluate the potential application of the model for CNS drug optimization. Here, we used drugs against the classical CNS disease target serotonin transporter (SERT) as a fine-tuned data set and generated a focused database against the target protein. The potential biological activities of the generated molecules were verified by using the physics-based induced-fit docking study. The success of this model demonstrates its potential in CNS drug design and optimization, which provides a new impetus for future CNS drug development.

Artificial Intelligence in Medicine: A New Frontier

Artificial intelligence (AI) refers to the engineering and science of making intelligent machines through algorithms or rules, mimicking human cognitive functions, such as learning and problem-solving. AI has several branches, such as machine learning and deep learning, which can add intelligence to applications. Machine learning is the study of algorithms that allow computer programs to improve automatically through experience. Deep learning algorithms learn from an extensive, multi-layered collection of interconnected processes and expose these processors to many examples. In the coming years, the integration of AI in routine medical care is expected to revolutionize Medicine, potentially improving patient care and quality of life. The time required for a diagnosis can be greatly reduced, and diagnostic efficiency can be significantly enhanced when AI assists clinicians. Large language model chatbots are capable of clinical expert-level medical note-taking, consultation, and questionanswering. Chatbotscan generate human-like text, may help diagnose diseases based on medical records, and may suggest treatment options or plans. Artificial intelligence algorithms, particularly deep learning, have demonstrated remarkable progress in radiological image analysis and diagnosis and may improve radiologists’ efficiency. These algorithms may also improve diagnostic accuracy in dermatology, histopathology, fundoscopy, endoscopy, and other medical images. Natural language processing and ambient clinical intelligence automate administrative duties like recording patient visits in electronic health records, streamlining clinical workflow, and freeing up doctors to spend more time with patients. AI may also help with new drug discoveries, precision medicine, and clinical research. AI developments can revolutionize several healthcare-related fields and pave the way for a more individualized, accurate, predictive, and portable future. Bangladesh J Medicine 2024; 35: 54-60

14-3-3 Protein-Protein Interactions: From Mechanistic Understanding to their Small-molecule Stabilization.

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

Aggrescan4D: structure-informed analysis of pH-dependent protein aggregation.

Protein aggregation is behind the genesis of incurable diseases and imposes constraints on drug discovery and the industrial production and formulation of proteins. Over the years, we have been advancing the Aggresscan3D (A3D) method, aiming to deepen our comprehension of protein aggregation and assist the engineering of protein solubility. Since its inception, A3D has become one of the most popular structure-based aggregation predictors because of its performance, modular functionalities, RESTful service for extensive screenings, and intuitive user interface. Building on this foundation, we introduce Aggrescan4D (A4D), significantly extending A3D's functionality. A4D is aimed at predicting the pH-dependent aggregation of protein structures, and features an evolutionary-informed automatic mutation protocol to engineer protein solubility without compromising structure and stability. It also integrates precalculated results for the nearly 500,000 jobs in the A3D Model Organisms Database and structure retrieval from the AlphaFold database. Globally, A4D constitutes a comprehensive tool for understanding, predicting, and designing solutions for specific protein aggregation challenges. The A4D web server and extensive documentation are available at https://biocomp.chem.uw.edu.pl/a4d/. This website is free and open to all users without a login requirement.