Xiaoyong Dai
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Xiaoyong Dai has a total of 107 co-authors. Here are the top co-authors by number of shared publications:
- Laiqiang Huang
School of Life Sciences, Tsinghua University, Beijing, China
7 shared publications - Huaqing Chen
Precision Medicine and Healthcare Research Center, Center for Biotechnology and Biomedicine, Shenzhen Key Laboratory of Gene and Antibody Therapy, State Key Laboratory of Chemical Oncogenomics, State Key Laboratory of Health Sciences and Technology, Tsinghua-Berkeley Shenzhen Institute (TBSI), Institute of Biopharmaceutical and Health Engineering, Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China
5 shared publications - Shaohua Ma
Institute of Biopharmaceutical and Health Engineering, Tsinghua Shenzhen International Graduate School (SIGS), Tsinghua University, Shenzhen, China
5 shared publications Yibo Hou
Key Laboratory of Active Proteins and Peptides Green Biomanufacturing of Guangdong Higher Education Institutes Institute of Biopharmaceutical and Health Engineering Shenzhen Key Laboratory of Gene and Immune Therapy Tsinghua Shenzhen International Graduate School Shenzhen Guangdong China
5 shared publicationsAlbert Yu
Key Laboratory of Active Proteins and Peptides Green Biomanufacturing of Guangdong Higher Education Institutes Institute of Biopharmaceutical and Health Engineering Shenzhen Key Laboratory of Gene and Immune Therapy Tsinghua Shenzhen International Graduate School Shenzhen Guangdong China
4 shared publications- Bing Guo
Shenzhen Key Laboratory of Advanced Functional Carbon Materials Research and Comprehensive Application, School of Science, Harbin Institute of Technology, Shenzhen, China
4 shared publications
Latest Publications
- RetractionOpen Access
- 15 September 2026
- Published by Springer Nature in Journal of Nanobiotechnology
- Vol. 24 (1) , 876
Retraction Note: Liposome-lentivirus for miRNA therapy with molecular mechanism study
- 15 September 2026
- Published by Springer Nature in Journal of Nanobiotechnology
- Vol. 24 (1) , 876
RetractionOpen AccessRetraction Note: Liposome-lentivirus for miRNA therapy with molecular mechanism study
- 9 June 2026
- Published by IOP Publishing in Biofabrication
- Vol. 18 (3)
A Novel Drug Screening Strategy Based on Non-Invasive Electrochemical Detection of Multiple Tumor Markers Secreted by Patient-Derived Microgel Organoids
Rapid and effective drug screening is a key step for personalized cancer therapy. Recently, microgel organoids have emerged as a promising in vitro drug screening model that ensures not only the uniform manufacture, but also the integration of various cells, active factors and matrix materials to mimic the real 3D tumor tissue. However, their corresponding drug sensitivity and efficacy analysis remains challenging, since traditional optical methods are invasive/destructive to tumor organoids and thus require many clinical samples for drug screening. Herein we propose a novel strategy based on non-invasive electrochemical detection of multiple tumor markers secreted by patient-derived microgel organoids. Specifically, a detection electrode is constructed using Ti3C2 MXene, Au nanoparticles (AuNPs) and aptamers (conjugated with redox probes) to detect breast cancer biomarkers through the specific recognition between aptamer and biomarkers. Our data reveal that high sensitivity and specificity of biomarker sensing can be both achieved for drug screening models prepared using breast cancer cell line, clinical breast cancer and colorectal cancer samples, which can be converted into relative cell viability and half-maximal inhibition concentration (IC50). Similar results between optical and electrochemical methods prove the feasibility and potential application of our strategy in high-throughput drug screening for diverse cancers.
- 9 June 2026
- Published by IOP Publishing in Biofabrication
- Vol. 18 (3)
A Novel Drug Screening Strategy Based on Non-Invasive Electrochemical Detection of Multiple Tumor Markers Secreted by Patient-Derived Microgel Organoids
Rapid and effective drug screening is a key step for personalized cancer therapy. Recently, microgel organoids have emerged as a promising in vitro drug screening model that ensures not only the uniform manufacture, but also the integration of various cells, active factors and matrix materials to mimic the real 3D tumor tissue. However, their corresponding drug sensitivity and efficacy analysis remains challenging, since traditional optical methods are invasive/destructive to tumor organoids and thus require many clinical samples for drug screening. Herein we propose a novel strategy based on non-invasive electrochemical detection of multiple tumor markers secreted by patient-derived microgel organoids. Specifically, a detection electrode is constructed using Ti3C2 MXene, Au nanoparticles (AuNPs) and aptamers (conjugated with redox probes) to detect breast cancer biomarkers through the specific recognition between aptamer and biomarkers. Our data reveal that high sensitivity and specificity of biomarker sensing can be both achieved for drug screening models prepared using breast cancer cell line, clinical breast cancer and colorectal cancer samples, which can be converted into relative cell viability and half-maximal inhibition concentration (IC50). Similar results between optical and electrochemical methods prove the feasibility and potential application of our strategy in high-throughput drug screening for diverse cancers.
- Open Access
- 24 April 2026
- Published by Springer Nature in Microsystems & Nanoengineering
- Vol. 12 (1) , 151
Programmable viscoelastic hydrogels exhibit antimicrobial and regenerative properties to promote cell migration, wound healing, and tissue remodeling
The growing challenges of conventional ECM bio-inks for 3D cell culture underscore the development of novel hydrogels that fully recapitulate specialized in vivo cell microenvironments. We developed HA-gel-dex hydrogels by double-crosslinking ECM components (hyaluronic acid and gelatin) with synthetic dextran. Engineered with irreversible amide bonds and dynamic imine crosslinks, the HA-gel-dex enables viscoelastic property modifications to replicate native ECM characteristics, with the highest yield ratio reaching ~1000%. This programmability derive from tunable crosslinking architectures and network dynamics, where controlled HA-gel to dex-CHO ratios and peptide functionalization yield optimized hydrogel viscoelasticity and predictable bioprinting performance. Functionalization with selective cell-ligation groups (RGD, collagen, and laminin peptides) strengthened cell-matrix interaction, promoting increased cell proliferation, differentiation, and micro-organo-sphere formation. Moreover, rheological analyses revealed significantly enhanced stress-relaxing and self-healing hydrogel properties; while showing cytocompatibility, increased cell viability, bioprinting capacity, significant biofilm inhibition properties, and synergism with antimicrobials. Further assays on BALB/c mice showed remarkably improved wound healing, hair follicle regeneration, and nature ECM production abilities, the performance of which are further augmented with the presence and encapsulation of therapeutic mesenchymal stem cells (MSCs). The study shows their future promise as wound dressings and potentially antibiotic treatment scaffolds for infection-associated wounds. Above all, HA-gel-dex hydrogels provide an improved ECM-like, reproducible, and cost-effective alternative to traditional ECM bio-inks for 3D cell culture, 3D bioprinting, and tissue repair studies – advancing synthetic ECM alternatives for translational applications. The alternative text for this image may have been generated using AI.
- 24 April 2026
- Published by Springer Nature in Microsystems & Nanoengineering
- Vol. 12 (1) , 151
Open AccessProgrammable viscoelastic hydrogels exhibit antimicrobial and regenerative properties to promote cell migration, wound healing, and tissue remodeling
The growing challenges of conventional ECM bio-inks for 3D cell culture underscore the development of novel hydrogels that fully recapitulate specialized in vivo cell microenvironments. We developed HA-gel-dex hydrogels by double-crosslinking ECM components (hyaluronic acid and gelatin) with synthetic dextran. Engineered with irreversible amide bonds and dynamic imine crosslinks, the HA-gel-dex enables viscoelastic property modifications to replicate native ECM characteristics, with the highest yield ratio reaching ~1000%. This programmability derive from tunable crosslinking architectures and network dynamics, where controlled HA-gel to dex-CHO ratios and peptide functionalization yield optimized hydrogel viscoelasticity and predictable bioprinting performance. Functionalization with selective cell-ligation groups (RGD, collagen, and laminin peptides) strengthened cell-matrix interaction, promoting increased cell proliferation, differentiation, and micro-organo-sphere formation. Moreover, rheological analyses revealed significantly enhanced stress-relaxing and self-healing hydrogel properties; while showing cytocompatibility, increased cell viability, bioprinting capacity, significant biofilm inhibition properties, and synergism with antimicrobials. Further assays on BALB/c mice showed remarkably improved wound healing, hair follicle regeneration, and nature ECM production abilities, the performance of which are further augmented with the presence and encapsulation of therapeutic mesenchymal stem cells (MSCs). The study shows their future promise as wound dressings and potentially antibiotic treatment scaffolds for infection-associated wounds. Above all, HA-gel-dex hydrogels provide an improved ECM-like, reproducible, and cost-effective alternative to traditional ECM bio-inks for 3D cell culture, 3D bioprinting, and tissue repair studies – advancing synthetic ECM alternatives for translational applications. The alternative text for this image may have been generated using AI.
- Open Access
- 7 April 2026
- Published by MDPI AG in International Journal of Molecular Sciences
- Vol. 27 (7) , 3327
Correction: Dai et al. Mallotucin D, a Clerodane Diterpenoid from Croton crassifolius, Suppresses HepG2 Cell Growth via Inducing Autophagic Cell Death and Pyroptosis. Int. J. Mol. Sci. 2022, 23, 14217
- 7 April 2026
- Published by MDPI AG in International Journal of Molecular Sciences
- Vol. 27 (7) , 3327
Open AccessCorrection: Dai et al. Mallotucin D, a Clerodane Diterpenoid from Croton crassifolius, Suppresses HepG2 Cell Growth via Inducing Autophagic Cell Death and Pyroptosis. Int. J. Mol. Sci. 2022, 23, 14217
- Open Access
- 27 February 2026
- Published by Springer Nature in Journal of Translational Medicine
- Vol. 24 (1)
Microfluidic-based patient-derived organoids recapitulate thyroid cancer heterogeneity and reveal NF-κB-driven maturation for precision therapy
Inter-tumor heterogeneity poses significant challenges for precision therapy in thyroid cancer (TC). The conventional organoid models are limited by ineffi
- 27 February 2026
- Published by Springer Nature in Journal of Translational Medicine
- Vol. 24 (1)
Open AccessMicrofluidic-based patient-derived organoids recapitulate thyroid cancer heterogeneity and reveal NF-κB-driven maturation for precision therapy
Inter-tumor heterogeneity poses significant challenges for precision therapy in thyroid cancer (TC). The conventional organoid models are limited by ineffi
- Open Access
- 6 January 2026
- Published by Wiley in Interdisciplinary Medicine
- Vol. 4 (1)
Anti‐CD47 antibody‐decorated and cisplatin‐containing mesoporous silica nanoparticles for targeted chemoimmunotherapy of colorectal cancer in a patient‐derived organoids xenograft model
For colorectal cancer (CRC), treatment outcomes and prognosis remain poor even when using first‐line chemotherapeutics and immunotherapeutics. The main reasons include (i) lack of tumor specificity, severe off‐target cytotoxicity leading to adverse side effects, and occurrence of chemoresistance, and (ii) the immunosuppressive “cold” tumor microenvironment (TME). Herein, we demonstrate an example of highly efficient targeted CRC chemoimmunotherapy with cisplatin (CDDP)‐containing nanomedicine (CDDP@MSN‐Anti‐CD47) in a patient‐derived organoids xenograft model. The CDDP@MSN‐Anti‐CD47 is composed of CDDP‐loaded mesoporous silica nanoparticles with superficially functionalized anti‐CD47 antibodies to target cluster of differentiation CD47, a transmembrane integrin‐associated protein that is over‐expressed in CRC cells and can bind with signal regulatory protein α on tumor‐associated macrophages (TAMs) to maintain the immunosuppressive TME. Our results demonstrated that CDDP@MSN‐Anti‐CD47 achieved targeted delivery of CDDP to CRC cells with extended release, and induced reactive oxygen species overproduction to inhibit CRC cells and CRC patient‐derived organoids growth via activating caspase‐3‐related apoptosis and immunogenic cell death. DAMPs released following CDDP@MSN‐Anti‐CD47 treatment promoted the polarization of TAMs from immunosuppressive M2 to pro‐inflammatory M1 phenotype and also synergized with the functionalized anti‐CD47 antibodies to restore macrophage phagocytic activity against CRC cells. In both CRC cell and patient‐derived organoid xenograft models, CDDP@MSN‐Anti‐CD47 could specifically target the tumor, showing strong anti‐tumor effects, prolonged survival rate, and low systemic cytotoxicity. Overall, CDDP@MSN‐Anti augments combination chemotherapy and immunotherapy, and can be developed as a novel and promising strategy for the clinical treatment of CRC.
- 6 January 2026
- Published by Wiley in Interdisciplinary Medicine
- Vol. 4 (1)
Open AccessAnti‐CD47 antibody‐decorated and cisplatin‐containing mesoporous silica nanoparticles for targeted chemoimmunotherapy of colorectal cancer in a patient‐derived organoids xenograft model
For colorectal cancer (CRC), treatment outcomes and prognosis remain poor even when using first‐line chemotherapeutics and immunotherapeutics. The main reasons include (i) lack of tumor specificity, severe off‐target cytotoxicity leading to adverse side effects, and occurrence of chemoresistance, and (ii) the immunosuppressive “cold” tumor microenvironment (TME). Herein, we demonstrate an example of highly efficient targeted CRC chemoimmunotherapy with cisplatin (CDDP)‐containing nanomedicine (CDDP@MSN‐Anti‐CD47) in a patient‐derived organoids xenograft model. The CDDP@MSN‐Anti‐CD47 is composed of CDDP‐loaded mesoporous silica nanoparticles with superficially functionalized anti‐CD47 antibodies to target cluster of differentiation CD47, a transmembrane integrin‐associated protein that is over‐expressed in CRC cells and can bind with signal regulatory protein α on tumor‐associated macrophages (TAMs) to maintain the immunosuppressive TME. Our results demonstrated that CDDP@MSN‐Anti‐CD47 achieved targeted delivery of CDDP to CRC cells with extended release, and induced reactive oxygen species overproduction to inhibit CRC cells and CRC patient‐derived organoids growth via activating caspase‐3‐related apoptosis and immunogenic cell death. DAMPs released following CDDP@MSN‐Anti‐CD47 treatment promoted the polarization of TAMs from immunosuppressive M2 to pro‐inflammatory M1 phenotype and also synergized with the functionalized anti‐CD47 antibodies to restore macrophage phagocytic activity against CRC cells. In both CRC cell and patient‐derived organoid xenograft models, CDDP@MSN‐Anti‐CD47 could specifically target the tumor, showing strong anti‐tumor effects, prolonged survival rate, and low systemic cytotoxicity. Overall, CDDP@MSN‐Anti augments combination chemotherapy and immunotherapy, and can be developed as a novel and promising strategy for the clinical treatment of CRC.
- Open Access
- 28 October 2025
- Published by Frontiers Media SA in Frontiers in Cell and Developmental Biology
- Vol. 13, 1700658
Editorial: Organoids as advanced model in cancer biology: the drug screening and clinical applications
The field of cancer research has long been driven by the pursuit of more physiologically relevant and predictive model systems. Traditional two-dimensional cell cultures and animal models, while invaluable, often fall short of recapitulating the complex heterogeneity, spatial architecture, and dynamic microenvironment of human tumors. This gap has propelled the emergence of three-dimensional organoid cultures as a transformative technology. This Research Topic, "Organoids as Advanced Model in Cancer Biology: The Drug Screening and Clinical Applications", was conceived to explore the rapidly evolving role of cancer organoids in refining drug discovery pipelines and bridging the daunting gap between preclinical findings and clinical utility. The articles published within this collection collectively underscore the significant potential of organoid technology while also acknowledging the challenges that remain on the path to its widespread standardization and application. This special issue significantly advances cancer organoid research by demonstrating their critical role in elucidating resistance mechanisms and enhancing drug screening. The study "Autophagy inhibition improves sensitivity to the multi-kinase inhibitor regorafenib in preclinical mouse colon tumoroids " utilizes mouse colon organoids to reveal that autophagy inhibition overcomes regorafenib resistance by suppressing EMT and Erk1/2 activation, providing a novel combinatorial strategy for colorectal cancer. The other article "Repurposing piroxicam enhances the antineoplastic effects of docetaxel and enzalutamide in prostate cancer cells using 2D and 3D in vitro culture models" employs prostate cancer organoids and 2D/3D models to show that repurposing piroxicam synergistically enhances the efficacy of conventional docetaxel and enzalutamide therapies, effectively reducing cell proliferation and organoid growth. The paper "Organoids technology in cancer research: from basic applications to advanced ex vivo models" reviews the broader utility of patient-derived organoids (PDOs), highlighting innovative techniques like decellularized ECM scaffolds to better recapitulate the tumor microenvironment (TME) for improved preclinical drug testing and personalized therapy prediction. Collectively, these contributions underscore organoids as physiologically relevant models that bridge basic research and clinical applications, offering powerful platforms for mechanistic discovery and therapeutic optimization in oncology. In a broader context, this Research Topic illustrates a field in a state of exciting maturation. The published articles move beyond simply extolling the virtues of organoids and instead present critical data and analyses that probe their practical applications and limitations. The collective findings affirm that organoids serve as a crucial intermediate model, offering a much-needed human-relevant system that sits between conventional cell lines and patient trials. However, the journey is far from complete. Future directions must address the need to incorporate non-epithelial components—such as immune cells, fibroblasts, and vasculature—to create more complex tumor microenvironments that can be used to study immunotherapy and stromal interactions. Standardization of protocols for organoid generation, biobanking, and drug screening assays is also paramount to ensure reproducibility and allow for multi-institutional comparisons. Moreover, the ethical considerations surrounding the use of patient-derived tissues and the potential for these living biobanks to guide patient care require ongoing dialogue. In conclusion, the work presented in this Topic solidifies the standing of organoids as an advanced and indispensable model in cancer biology, poised to make significant contributions to drug development and the realization of personalized cancer medicine.
- 28 October 2025
- Published by Frontiers Media SA in Frontiers in Cell and Developmental Biology
- Vol. 13, 1700658
Open AccessEditorial: Organoids as advanced model in cancer biology: the drug screening and clinical applications
The field of cancer research has long been driven by the pursuit of more physiologically relevant and predictive model systems. Traditional two-dimensional cell cultures and animal models, while invaluable, often fall short of recapitulating the complex heterogeneity, spatial architecture, and dynamic microenvironment of human tumors. This gap has propelled the emergence of three-dimensional organoid cultures as a transformative technology. This Research Topic, "Organoids as Advanced Model in Cancer Biology: The Drug Screening and Clinical Applications", was conceived to explore the rapidly evolving role of cancer organoids in refining drug discovery pipelines and bridging the daunting gap between preclinical findings and clinical utility. The articles published within this collection collectively underscore the significant potential of organoid technology while also acknowledging the challenges that remain on the path to its widespread standardization and application. This special issue significantly advances cancer organoid research by demonstrating their critical role in elucidating resistance mechanisms and enhancing drug screening. The study "Autophagy inhibition improves sensitivity to the multi-kinase inhibitor regorafenib in preclinical mouse colon tumoroids " utilizes mouse colon organoids to reveal that autophagy inhibition overcomes regorafenib resistance by suppressing EMT and Erk1/2 activation, providing a novel combinatorial strategy for colorectal cancer. The other article "Repurposing piroxicam enhances the antineoplastic effects of docetaxel and enzalutamide in prostate cancer cells using 2D and 3D in vitro culture models" employs prostate cancer organoids and 2D/3D models to show that repurposing piroxicam synergistically enhances the efficacy of conventional docetaxel and enzalutamide therapies, effectively reducing cell proliferation and organoid growth. The paper "Organoids technology in cancer research: from basic applications to advanced ex vivo models" reviews the broader utility of patient-derived organoids (PDOs), highlighting innovative techniques like decellularized ECM scaffolds to better recapitulate the tumor microenvironment (TME) for improved preclinical drug testing and personalized therapy prediction. Collectively, these contributions underscore organoids as physiologically relevant models that bridge basic research and clinical applications, offering powerful platforms for mechanistic discovery and therapeutic optimization in oncology. In a broader context, this Research Topic illustrates a field in a state of exciting maturation. The published articles move beyond simply extolling the virtues of organoids and instead present critical data and analyses that probe their practical applications and limitations. The collective findings affirm that organoids serve as a crucial intermediate model, offering a much-needed human-relevant system that sits between conventional cell lines and patient trials. However, the journey is far from complete. Future directions must address the need to incorporate non-epithelial components—such as immune cells, fibroblasts, and vasculature—to create more complex tumor microenvironments that can be used to study immunotherapy and stromal interactions. Standardization of protocols for organoid generation, biobanking, and drug screening assays is also paramount to ensure reproducibility and allow for multi-institutional comparisons. Moreover, the ethical considerations surrounding the use of patient-derived tissues and the potential for these living biobanks to guide patient care require ongoing dialogue. In conclusion, the work presented in this Topic solidifies the standing of organoids as an advanced and indispensable model in cancer biology, poised to make significant contributions to drug development and the realization of personalized cancer medicine.
- Open Access
- 27 October 2025
- Published by Wiley in Advanced Science
- Vol. 13 (1) , e10508
EM2, a Natural Product MST1/2 Kinase Activator, Suppresses Non‐Small Cell Lung Cancer via Hippo Pathway Activation
Lung cancer, 85% of which is non‐small cell lung cancer (NSCLC), is the cancer with the highest incidence and mortality rate worldwide. Despite recent advancements in therapeutic approaches, the efficacy of conventional radiotherapy and chemotherapy remains suboptimal, highlighting the urgent need for more effective treatment strategies. Dysregulation of kinases MST1 and MST2 (MST1/2) is implicated in the progression of NSCLC, positioning MST1/2 as a potential therapeutic target. However, no high‐selectivity and high‐efficacy MST1/2 activator is identified to date. In this study, by using computer‐aided virtual screening combined with cell experiments, EM2 is identified as a promising MST1/2‐binding candidate. Subsequent experimental validation demonstrates that EM2 significantly suppresses the proliferation, migration, and invasion of NSCLC cells by directly targeting MST1/2 and enhancing its kinase activity, thereby activating the Hippo signaling pathway and reducing nuclear translocation of the downstream effector YAP. Both in vivo xenograft models and organoid models demonstrates that EM2 effectively suppresses NSCLC tumor growth. In summary, this study not only reaffirms MST1/2 as a viable therapeutic target for NSCLC but also provides compelling experimental evidence supporting EM2 as a highly effective and promising anti‐cancer agent.
- 27 October 2025
- Published by Wiley in Advanced Science
- Vol. 13 (1) , e10508
Open AccessEM2, a Natural Product MST1/2 Kinase Activator, Suppresses Non‐Small Cell Lung Cancer via Hippo Pathway Activation
Lung cancer, 85% of which is non‐small cell lung cancer (NSCLC), is the cancer with the highest incidence and mortality rate worldwide. Despite recent advancements in therapeutic approaches, the efficacy of conventional radiotherapy and chemotherapy remains suboptimal, highlighting the urgent need for more effective treatment strategies. Dysregulation of kinases MST1 and MST2 (MST1/2) is implicated in the progression of NSCLC, positioning MST1/2 as a potential therapeutic target. However, no high‐selectivity and high‐efficacy MST1/2 activator is identified to date. In this study, by using computer‐aided virtual screening combined with cell experiments, EM2 is identified as a promising MST1/2‐binding candidate. Subsequent experimental validation demonstrates that EM2 significantly suppresses the proliferation, migration, and invasion of NSCLC cells by directly targeting MST1/2 and enhancing its kinase activity, thereby activating the Hippo signaling pathway and reducing nuclear translocation of the downstream effector YAP. Both in vivo xenograft models and organoid models demonstrates that EM2 effectively suppresses NSCLC tumor growth. In summary, this study not only reaffirms MST1/2 as a viable therapeutic target for NSCLC but also provides compelling experimental evidence supporting EM2 as a highly effective and promising anti‐cancer agent.
- Open Access
- 2 October 2025
- Published by Springer Nature in EMBO Molecular Medicine
- Vol. 17 (11) , 2932-2965
AI-identified CD133-targeting natural compounds demonstrate differential anti-tumor effects and mechanisms in pan-cancer models
Advanced algorithms have significantly improved the efficiency of in vitro screening for protein-interactive compounds. However, target antigen (TAA/TSA)-based drug discovery remains challenging, as predictions of compound-protein interaction (CPI) based solely on molecular structure fail to fully elucidate the underlying mechanisms. In this study, we utilized deep learning, specifically TransformerCPI to screen active molecules from a Chinese herb compound library based on protein sequences. Two natural products, Polyphyllin V and Polyphyllin H, were identified as targeting the pan-cancer marker CD133. Their anti-tumor efficacy and safety were confirmed across validation in cancer cell lines, tumor patient-derived organoids, and animal models. Despite their analogous structures and binding affinity to CD133, Polyphyllin V suppresses the PI3K-AKT pathway, inducing pyroptosis and blockage of mitophagy, whereas Polyphyllin H inhibits the Wnt/β-catenin pathway and triggers apoptosis. These distinct mechanisms underscore the potential of combining AI-driven screening with biological validation. This AI-to-patient pipeline identifies Polyphyllin V and Polyphyllin H as CD133-targeted drugs for pan-cancer therapy, and reveals the limitations of virtual screening alone and emphasizes the necessity of live model evaluation in AI-based therapeutic discovery.
- 2 October 2025
- Published by Springer Nature in EMBO Molecular Medicine
- Vol. 17 (11) , 2932-2965
Open AccessAI-identified CD133-targeting natural compounds demonstrate differential anti-tumor effects and mechanisms in pan-cancer models
Advanced algorithms have significantly improved the efficiency of in vitro screening for protein-interactive compounds. However, target antigen (TAA/TSA)-based drug discovery remains challenging, as predictions of compound-protein interaction (CPI) based solely on molecular structure fail to fully elucidate the underlying mechanisms. In this study, we utilized deep learning, specifically TransformerCPI to screen active molecules from a Chinese herb compound library based on protein sequences. Two natural products, Polyphyllin V and Polyphyllin H, were identified as targeting the pan-cancer marker CD133. Their anti-tumor efficacy and safety were confirmed across validation in cancer cell lines, tumor patient-derived organoids, and animal models. Despite their analogous structures and binding affinity to CD133, Polyphyllin V suppresses the PI3K-AKT pathway, inducing pyroptosis and blockage of mitophagy, whereas Polyphyllin H inhibits the Wnt/β-catenin pathway and triggers apoptosis. These distinct mechanisms underscore the potential of combining AI-driven screening with biological validation. This AI-to-patient pipeline identifies Polyphyllin V and Polyphyllin H as CD133-targeted drugs for pan-cancer therapy, and reveals the limitations of virtual screening alone and emphasizes the necessity of live model evaluation in AI-based therapeutic discovery.
- Open Access
The application of liver cancer organoids in tumour precision medicine: A comprehensive review