​연구개발 성과

• Objectives: An integrated pipeline combining UHPLC-MS/MS metabolomics, computational methods (network pharmacology, molecular docking, and dynamics simulation), and in vitro bioassays was established to efficiently discover and mechanistically characterize anti-aging compounds from novel botanical sources. This pipeline was applied to identify and evaluate Tephroseris flammea, a previously unassessed plant, to characterize its comprehensive metabolite profile and verify its anti-aging effects.

• Methods: The study utilized a whole plant extract of T. flammea collected from the Korean Peninsula. Untargeted UHPLC-MS/MS profiling was performed for compound annotation, and network pharmacology was conducted using multiple databases (GeneCards, Open Targets, and SenSkin™) to identify skin-aging-related targets. Molecular docking and 100 ns molecular dynamics simulations were used to predict binding stability, followed by experimental validation through MTT, radical scavenging, and ELISA assays for inflammatory and matrix-degrading markers.

• Results: Metabolomic profiling identified 21 compounds, including flavonoids, phenylpropanoids, and pyrrolizidine alkaloids, which were linked to 226 skin-aging-related targets, notably AKT1, RELA, and MAPK3. The extract demonstrated significant antioxidant activity and effectively suppressed key inflammatory mediators (IL-6, TNF-$\alpha$, COX-2) and MMP-1 levels in UVB-exposed fibroblasts without detectable cytotoxicity.

• Conclusions: This study validates the "from-selection-to-validation" pipeline for mechanistically characterizing complex botanicals, establishing T. flammea as a promising candidate for natural anti-aging applications. While the presence of certain unsaturated pyrrolizidine alkaloids highlights a toxicological challenge, the findings suggest a rational path for developing safe, functional ingredients through optimized purification of bioactive, non-toxic constituents like Thesinine.

• Objectives: We developed and applied an integrated analytical pipeline combining UHPLC-MS/MS metabolomics, computational analyses (network pharmacology, molecular docking, and molecular dynamics simulation), and experimental bioassays to efficiently identify and characterize novel natural products with anti-aging potential. This workflow was applied to Melampyrum roseum Maxim. to elucidate its bioactive potential against skin aging.

• Methods: The workflow comprised identification and sourcing of M. roseum, untargeted UHPLC-MS/MS metabolomic profiling and compound annotation, and network pharmacology to identify skin-aging-related targets. We further utilized molecular docking and molecular dynamics simulations to predict ligand-target interactions and binding stability, followed by in vitro bioassays using human dermal fibroblasts to validate antioxidant and anti-inflammatory effects.

• Results: UHPLC-MS/MS profiling annotated 13 secondary metabolites, which were linked to 172 potential skin-aging-associated targets mainly within inflammatory and oxidative-stress pathways. The M. roseum extract demonstrated significant antioxidant activity and effectively suppressed key inflammatory mediators (IL-6, TNF-α, COX-2) and MMP-1 levels in UVB-exposed fibroblasts without significant cytotoxicity.

• Conclusions: We established M. roseum as a previously unrecognized source of multifunctional compounds that modulate key inflammatory and matrix-regulatory pathways. Our findings provide preliminary mechanistic evidence for its potential as a promising candidate for natural anti-aging applications and demonstrate the utility of the integrated informatics-to-bioassay pipeline.

The Hippo signaling pathway is critical for regulating cell growth, tissue homeostasis, and organ size. Dysregulation of this pathway has been associated with a range of pathologies, especially cancer, through its modulation of downstream effectors—Yes-associated protein (YAP) and the transcriptional coactivator with PDZ-binding motif (TAZ). These proteins bind to transcriptional enhanced associate domain (TEAD) proteins and function as transcription factors in the nucleus, producing oncogenic target genes such as CTGF and CYR61. TEAD proteins require palmitoylation via a covalent bond with cysteine in the central pocket to bind YAP/TAZ. Therefore, competitive inhibition that prevents palmitoylation could serve as an effective anticancer strategy. In this study, we analyzed the crystal structures of the known inhibitor VT-105 bound to TEAD3 to identify new binding spots that were previously unexplored, with the aim of discovering more potent compounds using structure-based drug design. Consequently, we identified a novel hydrogen-bonding site and discovered C-2, which effectively binds to this site, as confirmed by X-ray crystallography. Furthermore, C-2 exhibited stable pharmacokinetic properties and demonstrated impressive efficacy in a mouse xenograft model.

• Objectives: We conducted a comprehensive comparative analysis of flavonoid diversity and antioxidant capacities between wild soybean (WB) and cultivated soybean (CB). The study aimed to identify unknown antioxidant flavonoids using advanced mass spectrometry techniques and to evaluate WB as a potential genetic resource for functional food ingredients.

• Methods: An integrated ion-filtering strategy (IIFS)—including precursor ion list (PIL), mass defect filter (MDF), and diagnostic fragment ions (DFIs)—combined with LC-HRMSn analysis was established for reliable qualitative profiling. We quantified 72 identified flavonoids using 19 standard compounds and assessed antioxidant capacities through both in vitro (ABTS, DPPH) and intracellular (DCFH-DA) assays using PC-12 and SH-SY5Y cell lines.

• Results: A total of 72 flavonoids were identified, with 11 being reported in soybeans for the first time. WB exhibited 3.5-fold higher total flavonoid content and significantly higher antioxidant capacity compared to CB. These superior effects were primarily attributed to the high concentrations of flavanols (especially epicatechin), procyanidins, and anthocyanins in WB.

• Conclusions: We demonstrated that WB is a valuable genetic resource with potent antioxidant activity, far exceeding that of CB in specific flavonoid classes. This study also validates the effectiveness of the IIFS-combined HRMSn strategy for the rapid and systematic identification of bioactive compounds in complex botanical matrices.

• In Vitro and In Vivo Anti-inflammatory and Chondroprotective Effects of Standardized Hot Water Extract from Hydrangea serrata (Thunb.) Ser. via Modulation of NF-κB and MAPK Pathways

• Objectives: This study aimed to identify the most effective sea cucumber body part for inhibiting lipid accumulation in adipocytes and to elucidate the specific compounds responsible for this effect. The research focused on filling the gap in knowledge regarding which body part of Stichopus japonicus (whole body, ovaries, or gut) exhibits the highest anti-obesity potential.

• Methods: Ethanol extracts from different body parts were tested on 3T3-L1 murine preadipocytes to assess adipogenic differentiation using Oil Red O staining and protein expression analysis. An in vivo study was conducted using C57BL/6 mice fed a high-fat diet (HFD) supplemented with sea cucumber ovary 80% ethanol extract (SCOE) for eight weeks. Saponins were isolated from SCOE using column chromatography and identified through nuclear magnetic resonance (NMR) and mass spectrometry.

• Results: SCOE demonstrated the highest efficacy in inhibiting adipocyte differentiation without significant cytotoxicity. In HFD-fed mice, SCOE treatment (0.2%) significantly reduced body weight gain by approximately 30.4% and improved serum lipid profiles and fatty liver conditions. Eight saponins were isolated, four of which inhibited adipogenesis. Notably, three active saponins—holotoxins A, B, and D1—were newly identified as potent inhibitors of adipocyte differentiation.

• Conclusions:The findings establish SCOE and its constituent saponins as effective anti-obesity agents. The identification of novel holotoxins provides a promising foundation for the development of natural product-based treatments for obesity and associated metabolic disorders using sea cucumber ovaries.

The   Hippo pathway controls organ size and homeostasis and is linked to numerous   diseases,including cancer. The transcriptional enhanced associate domain   (TEAD) family of transcriptionfactors acts as a receptor for downstream   effectors, namely yes-associated protein (YAP) and tran-scriptional   co-activator with PDZ-binding motif (TAZ), which binds to various   transcription factorsand is essential for stimulated gene transcription.   YAP/TAZ-TEAD facilitates the upregulation ofmultiple genes involved in evolutionary   cell proliferation and survival. TEAD1–4 overexpressionhas been observed in   different cancers in various tissues, making TEAD an attractive target   fordrug development. The central drug-accessible pocket of TEAD is crucial   because it undergoes apost-translational modification called   auto-palmitoylation. Crystal structures of the C-terminal TEADcomplex with   small molecules are available in the Protein Data Bank, aiding   structure-based drugdesign. In this study, we utilized the fragment molecular   orbital (FMO) method, molecular dynamics(MD) simulations, shape-based   screening, and molecular mechanics–generalized Born surface area(MM-GBSA)   calculations for virtual screening, and we identified a novel non-covalent   inhibitor—BC-001—with IC50 = 3.7 μM in a reporter assay. Subsequently, we optimized several   analogs of BC-001and found that the optimized compound BC-011 exhibited an   IC50 of 72.43 nM. These findings can beused to design effective TEAD   modulators with anticancer therapeutic implications.

Polo-like   kinase 1 (PLK1) plays a pivotal role in cell division regulation and   emergesas a promising therapeutic target for cancer treatment. Consequently,   the development of small-molecule inhibitors targeting PLK1 has become a   focal point in contemporary research. The adenosinetriphosphate (ATP)-binding   site and the polo-box domain in PLK1 present crucial interaction sites   forthese inhibitors, aiming to disrupt the protein’s function. However,   designing potent and selectivesmall-molecule inhibitors can be challenging,   requiring a deep understanding of protein–ligandinteraction mechanisms at   these binding sites. In this context, our study leverages the   fragmentmolecular orbital (FMO) method to explore these site-specific   interactions in depth. Using the FMOapproach, we used the FMO method to   elucidate the molecular mechanisms of small-moleculedrugs binding to these   sites to design PLK1 inhibitors that are both potent and selective.   Ourinvestigation further entailed a comparative analysis of various PLK1   inhibitors, each characterizedby distinct structural attributes, helping us   gain a better understanding of the relationship betweenmolecular structure   and biological activity. The FMO method was particularly effective in   identifyingkey binding features and predicting binding modes for   small-molecule ligands. Our research alsohighlighted specific “hot spot”   residues that played a critical role in the selective and robust bindingof   PLK1. These findings provide valuable insights that can be used to design new   and effective PLK1inhibitors, which can have significant implications for   developing anticancer therapeutics.

Since   the Simplified Molecular Input Line Entry System (SMILES) is oriented to the   atomic-level representation ofmolecules and is not friendly in terms of human   readability and editable, however, IUPAC is the closest to natural language   and isvery friendly in terms of human-oriented readability and performing   molecular editing, we can manipulate IUPAC to generatecorresponding new   molecules and produce programming-friendly molecular forms of SMILES. In   addition, antiviral drug design,especially analogue-based drug design, is   also more appropriate to edit and design directly from the functional group   level of IUPACthan from the atomic level of SMILES, since designing analogues   involves altering the R group only, which is closer to theknowledge-based   molecular design of a chemist. Herein, we present a novel data-driven   self-supervised pretraining generative modelcalled “TransAntivirus” to make   select-and-replace edits and convert organic molecules into the desired   properties for design ofantiviral candidate analogues. The results indicated   that TransAntivirus is significantly superior to the control models in terms   ofnovelty, validity, uniqueness, and diversity. TransAntivirus showed   excellent performance in the design and optimization ofnucleoside and non-nucleoside   analogues by chemical space analysis and property prediction analysis.   Furthermore, to validate theapplicability of TransAntivirus in the design of   antiviral drugs, we conducted two case studies on the design of nucleoside   analoguesand non-nucleoside analogues and screened four candidate lead   compounds against anticoronavirus disease (COVID-19). Finally,we recommend   this framework for accelerating antiviral drug discovery.

We   construct a protein–protein interaction (PPI) targeted drug-likeness dataset   and propose a deep molecular generative frameworkto generate novel   drug-likeness molecules from the features of the seed compounds. This   framework gains inspiration from publishedmolecular generative models, uses   the key features associated with PPI inhibitors as input and develops deep   molecular generativemodels for de novo molecular design of PPI inhibitors.   For the first time, quantitative estimation index for compounds targetingPPI   was applied to the evaluation of the molecular generation model for de novo   design of PPI-targeted compounds. Our resultsestimated that the generated   molecules had better PPI-targeted drug-likeness and drug-likeness.   Additionally, our model also exhibitscomparable performance to other several   state-of-the-art molecule generation models. The generated molecules share   chemical spacewith inhibitors of protein-protein interaction database   (iPPI-DB) inhibitors as demonstrated by chemical space analysis. The   peptidecharacterization-oriented design of PPI inhibitors and the   ligand-based design of PPI inhibitors are explored. Finally, we recommendthat   this framework will be an important step forward for the de novo design of   PPI-targeted therapeutics.