Organ-on-a-chip systems can replicate human physiological functionsin vitroby simulating the dynamicin vivomicroenvironment, therefore offering great potential for applications in drug screening, disease research, and personalized medicine. Multi-channel microfluidic chips are the core physical components of organ-on-a-chip systems, which often incorporate structures such as stripes, micro-pillars, and porous membranes to confine gels within specific channels, thereby providing a three-dimensional extracellular matrix environment for reconstruction of tissue barrier modelsin vitro. However, current multi-channel microfluidic chips confront challenges such as the unintended absorption of molecules, dependence on complex multi-material and multi-step fabrication processes, and instability in confining liquids. To address these challenges, we propose a multi-channel microfluidic chip with bilateral stripe structures, which can be mass-produced using single cyclic olefin copolymer material through injection molding. The bilateral stripe structures can effectively confine liquids with different wettabilities within the central channel by leveraging the edge effect. To demonstrate the versatility of the microfluidic platform, we successfully constructed tubular endothelial and renal tubule barriers on this chip, showcasing its potential for high-throughput, standardized organoid culture. This innovative microfluidic platform enables the construction of variousin vitroorgan models, offering a powerful tool for preclinical research and drug development.

Respiratory viral infections pose persistent global health threats, yet traditionalin vitroand animal models inadequately recapitulate human tissue microenvironments. Organ-on-a-chip technology integrates microfluidic engineering with cell biology to recreate three-dimensional architectures, mechanical forces, and multicellular interactions of the human respiratory system. This review systematically summarizes recent advances in organ-on-a-chip platforms for modeling respiratory viral infections and host immune responses. We highlight their unique capabilities in simulating alveolar-capillary barriers, lymphoid follicle formation, and multi-organ axes including lung-brain and gut-lung communication. Furthermore, we discuss applications in antiviral drug screening, vaccine evaluation, and personalized medicine, while addressing current challenges and future directions toward standardized, multi-organ integrated, and intelligently monitored systems.

Polarized epithelia integrate barrier sealing, vectorial ion-water transport, and cytoskeletal mechanics, yet scalable assays that report this coupled functional state in real time remain limited. Here, we introduce photonic-crystal hydraulic manometry (PCHM), which quantifies out-of-plane mechanical states from single-frame reflection images in standard culture formats. Using PCHM, we find that epithelial monolayers maintain a kilopascal-scale basal compression (∼3 kilopascals) at the cell-substrate interface. A systematic perturbation panel spanning ion transport and actomyosin contractility defines the sensitivity, dynamic range, and reversibility of the readout, establishing basal compression as an actionable state variable of epithelial physiology. Leveraging this physiology-anchored metric, we detect early infection with coxsackievirus as a collapse of basal compression within 2 hours, well before cytopathic effects become apparent (48 hours). In severe acute respiratory syndrome coronavirus 2 pseudovirus neutralization assays, PCHM provided a 2-hour readout that was consistent with matched 48-hour luciferase results. In viral titration assays, the same 2-hour PCHM readout extended the detectable low-input range by approximately one order of magnitude relative to the matched 48-hour luciferase end point. Together, PCHM links epithelial transport and mechanics to a scalable, stain-free assay framework for epithelial pathophysiology and therapeutic screening.

Chronic kidney disease (CKD), driven largely by renal fibrosis, lacks effective therapies due to the limited predictive capacity of existing preclinical models. To address this, we developed a human tubuloid-on-a-chip model integrating tubuloids, endothelial cells, and immune cells within a microfluidic system to recapitulate the key pathophysiology of renal fibrosis. Induction of fibrosis with TGF-β1 in this system recapitulated key pathological features, including extracellular matrix deposition, epithelial-mesenchymal transition, and loss of epithelial polarity. Functional assessments revealed impaired tubular reabsorption, including reduced albumin uptake and glucose transport, alongside elevated oxidative stress, mirroring clinical observations in CKD patients. The model’s pharmacological relevance was validated by the therapeutic effects of nintedanib, which attenuated fibrotic phenotypes. Taken together, this tubuloid-on-a-chip platform demonstrates the potential to model complex fibrotic pathologies in vitro and may serve as a useful tool for CKD research and anti-fibrotic drug development, potentially accelerating therapeutic discovery for renal fibrosis.

IntroductionAbnormal pigmentation plays an important role in various skin diseases and in studies of whitening efficacy.Three-dimensional pigmented epidermis-on-a-chip models provide a crucial in vitro platform for exploring melanin production and regulation in skin. However, dynamic and non-invasive quantitative assessment of melanin distribution remains difficult with traditional histological methods.MethodsIn this study, an AI-assisted objective evaluation framework was established for three-dimensional pigmented epidermis-on-a-chip models based on brightfield images. Melanin regions were segmented using the MEM-ViT algorithm, and their morphological features were extracted to build a multi-indicator comprehensive analysis system for determining the “good/poor” status of the model.ResultsThe results showed 98% consistency between algorithmic predictions and manual annotations, demonstrating the reliability and generalization capability of the proposed method. The framework enabled accurate segmentation of melanin regions and standardized evaluation of model quality without staining.DiscussionThis method provides a rapid, non-invasive, and standardized approach for evaluating 3D pigmented epidermis-on-a-chip models. It offers a useful technical pathway for drug efficacy research, whitening mechanism analysis, and objective assessment of skin pigmentation-related disorders.

Impaired wound healing and pathological scarring remain major clinical challenges, with immune cell dysregulation being a key driver of disease progression. Conventional in vitro models fail to recapitulate human immune responses, limiting their translational relevance. In recent years, advances in tissue engineering and microfluidic technologies have driven growing efforts to incorporate immune cells into in vitro models, thereby improving their ability to mimic pathological microenvironments. Among these, organ-on-a-chip technology stands out for its capacity to replicate dynamic perfusion, mechanical stimulation, and multicellular crosstalk-features critical for modeling immune-mediated wound repair. This review systematically summarizes recent progress in immune cell-integrated models of aberrant wound healing, including two-dimensional co-cultures, three-dimensional static cultures, organoid systems, and organ-on-a-chip platforms. We highlight core strategies for immune cell integration and their roles in recapitulating key pathological processes such as inflammation and fibrosis. Despite ongoing challenges in cell source stability, model standardization, and long-term culture viability, emerging strategies (e.g., organ-on-a-chip combined with three-dimensional bioprinting or modular design) offer new opportunities for creating biomimetic, high-throughput platforms for wound research. This review aims to facilitate the adoption of immune-integrated in vitro models in wound healing research, deepen mechanistic understanding of immune-driven pathology, and accelerate the development of precision therapeutics.

Traditional toxicology, with its reliance on animal models and oversimplified cell cultures, often fails to predict human responses due to interspecies differences and limited physiological relevance. Organ-on-a-chip (OoC) technology, as a microengineering breakthrough, enables reconstruction of human-relevant organ functions, providing a powerful tool for toxicity testing. However, OoC remains largely regarded as a technological platform rather than a distinct research discipline. In this review, we propose organ-on-a-chip toxicology (OCT) as a groundbreaking interdisciplinary paradigm that integrates advanced engineering, toxicological science, and biomedical research to redefine toxicological assessment. OCT transcends conventional OoC technology by providing a unified framework for elucidating toxicity effects and mechanisms at molecular, cellular, and organ levels. It uniquely enables comprehensive systemic toxicity modeling, incorporating full absorption-distribution-metabolism-excretion pathways and inter-organ signaling. Leveraging cutting-edge bioengineering, organoid-driven cellular fidelity, and AI-enhanced data analytics, OCT delivers unparalleled precision in drug safety evaluation, personalized toxicology, environmental hazard assessment, and food health. Despite current challenges in standardization, scalability, and regulatory acceptance, OCT holds the potential to revolutionize toxicological science by offering predictive, ethical, and human-centric insights, minimizing animal testing while advancing global health risk assessments.

Diabetic kidney disease has increasingly emerged as a global public health concern, yet substantial challenges persist in its mechanistic research and drug development. Traditional 2-dimensional cell cultures and animal models frequently lack the capacity to faithfully recapitulate human pathophysiological conditions, and the use of animal models is increasingly limited by ethical and policy-related hurdles. In vitro models such as organoids and organs-on-a-chip have demonstrated remarkable advantages, enabling more faithful simulation of the in vivo microenvironment. The advancement of 3-dimensional bioprinting, microfluidic, and vascularization technologies has further propelled the maturation and progress of in vitro kidney models. This review summarizes and discusses in vitro kidney models and their advances, including kidney cell lines, spheroids, kidney organoids, and kidneys-on-a-chip. It also elaborates on the establishment of diabetic kidney disease models based on these in vitro platforms, which provides robust support for both basic and clinical research.

Calciphylaxis (calcific uremic arteriolopathy, CUA) is a rare, fatal disorder primarily affecting chronic kidney disease patients, characterized by microvascular calcification, thrombosis, and skin necrosis. In a discovery cohort (3 CUA, 10 uremic), plasma proteomics identified Thrombospondin-1 (THBS1) as the top upregulated hub in CUA, significantly reduced after human amnion-derived mesenchymal stem cell (hAMSC) therapy, alongside latent TGF-β binding protein 1, both linked to coagulation and wound healing. In vitro proteomics indicated that THBS1/TGF-β1 blockade impaired CUA serum-induced endothelial adhesion and coagulation. ELISA in combined discovery and validation cohorts (8 CUA, 20 uremic) confirmed this reduction post-treatment (6 patients), independent of systemic inflammation. Multiplex immunofluorescence revealed THBS1 and CD47 co-localized with CD31 and integrin β3 in injured microvessels. A human microvascular chip showed that THBS1 inhibition or hAMSC-conditioned medium alleviates injury. These findings implicate THBS1 as a key factor and potential biomarker in calciphylaxis, suggesting hAMSC therapy as a promising mechanism-based approach. Video Abstract:

Micro- and nano-plastics (MNPs) have emerged as ubiquitous environmental contaminants and are increasingly implicated in adverse cardiovascular outcomes, yet their induced cardiotoxicity and potential mechanisms remain poorly understood. This study integrated in vivo mouse models, AC16 cardiomyocytes and a human cardiac organoid-on-a-chip (COoC) platform to multi-dimensionally evaluate polystyrene nanoparticles (PS-NPs)-induced cardiac injury and clarify its key molecular mechanisms. We found that PS-NPs exposure induced pronounced structural and functional cardiac injury in mice and caused impaired myocardial contraction, disrupted calcium transients and increased injury biomarkers in vitro. Notably, PS-NPs exposure perturbed myocardial energy metabolism, producing a metabolic reprogramming profile characterized by suppressed fatty acid oxidation (FAO) and enhanced glycolytic activity. Metabolic interventions further showed that activation of FAO or promotion of mitochondrial pyruvate oxidation improved myocardial energy status and alleviated cardiotoxicity, whereas direct inhibition of glycolysis aggravated energy depletion and cellular injury, suggesting that enhanced glycolysis provided partial energetic compensation but was insufficient to offset impaired oxidative metabolism. Mechanistically, our findings indicated a functional role of the SDHA/succinate/HIF-1α signaling axis in this metabolic reprogramming. PS-NPs-induced SDHA downregulation promoted succinate accumulation and HIF-1α stabilization, thereby rewiring myocardial energy metabolism and contributing to cardiac dysfunction. Collectively, we revealed myocardial metabolic reprogramming as an important mechanism underlying PS-NPs-induced cardiotoxicity and identified the SDHA/succinate/HIF-1α axis as a potential molecular link between PS-NPs exposure and cardiac injury.

Skin aging results from a combination of intrinsic factors and exogenous stimuli, leading to changes in the structure and components of the extracellular matrix (including the skin basement membrane), which directly influence the aging process. In vitro models are powerful tools for exploring skin aging and overcoming inter-species differences and ethical issues associated with animal models, thus demonstrating powerful potential in skin aging research and anti-aging drug development. In this review, the advantages and disadvantages of in vitro models are discussed, including 2D monolayer models, 3D static reconstructed human skin models, 3D bioprinting models, organoid models, and Skin-on-Chip models for studying skin aging and anti-aging drug development. Finally, concepts and perspectives for the next-generation skin aging models are proposed. These models are expected to provide innovative tools for investigating the mechanisms of skin aging in depth, as well as skin aging repair and prevention.

Melanoma, an aggressive and lethal form of skin cancer, poses significant challenges in treatment because of its complex biological mechanisms and poor therapeutic response. However, the construction of biomimeticin vitromodels of melanoma with organs-on-a-chip micro-environment have not been reported. This study aimed to design and biofabricate a novel melanoma-containing skin-on-a-chip model for the evaluation of drug responses, and to investigate the effects of ultraviolet (UV) irradiation on melanoma progression. Our results demonstrated pronounced heterogeneity in the tumor immune microenvironment of melanoma patients. The melanoma organoids derived from patients effectively preserve the native immune components, and organoids retained both the histological features and genetic characteristics of the original tumors. Drug response assessments indicated that the combination of dabrafenib and trametinib significantly inhibited melanoma cell proliferation and induced apoptosis; while UV exposure significantly promoted melanoma progression, highlighting the role of environmental factors in tumor development. Our research suggested the biofabrication of a biomimetic melanoma organoid-containing skin-on-a-chip model could elucidate patients' specific melanoma biology, tumor immune microenvironment, and corresponding drug sensitivity.

BackgroundStir-fried atractylodis macrocephalae rhizoma (AMR) with aurantii fructus (AF) (SFALCA), a classical prescription of traditional Chinese medicine (TCM), has been widely used for promoting gastrointestinal health for centuries, with multiple pharmacological properties such as anti-tumor, anti-inflammatory, anti-aging, antioxidant, and antibacterial effects. Inflammatory bowel disease (IBD) is an immune-mediated chronic gastrointestinal inflammatory disorder that causes long-term distress to patients. Despite its widespread use, the specific effects of SFALCA on intestinal barrier function and underlying mechanisms remain unclear. Furthermore, interspecies discrepancies in animal studies and the physiological constraints of existing in vitro models synergistically limit the translational potential of current findings on this botanical combination.MethodsTo investigate the active components, mechanisms of SFALCA in treating IBD and assesses its safety. We designed an innovative organ-on-a-chip system to simulate the human intestinal and liver environment. By stimulating with LPS/PMA, we established an in vitro IBD model and intervened with SFALCA and its extracts. Immunofluorescence staining was used to evaluate the success of the model. TEER measurements were employed to assess the integrity of tight junctions. Alcian Blue staining characterizes the expression of acidic mucins in HT-29 cells. The levels of inflammatory cytokines and the human albumin were measured using ELISA kits. The cytotoxicity of TCM to liver was evaluated by CellTiter-Glo® 3D test according to the manufacturer’s instructions. Flow cytometric analysis was used to detect the polarization of macrophages in intestinal inflammation model after drug treatment. RNA-seq analysis was used to identify key targets and pathways.ResultsThe results showed that SFALCA and its extracts significantly increased transepithelial electrical resistance (TEER) and Zonula occludens-1 (ZO-1) expression, while inhibiting LPS/PMA-induced IL-6 levels and the proportion of M1 macrophages. Further analysis revealed that the main active components of SFALCA, Atractylenolide I and Naringin, exert anti-inflammatory effects by inhibiting the Interleukin-17C (IL-17C) mediated positive feedback loop. Additionally, organ-on-a-chip technology confirmed that SFALCA showed no significant toxicity to the liver.ConclusionIn conclusion, this study elucidates the active components and mechanisms of SFALCA in treating IBD and assesses its safety, providing a reliable in vitro platform for future therapeutic strategies.

Targeting the NLRP3 inflammasome is critical for treating fibrotic diseases; however, current NACHT domain inhibitors face mutational escape and limited isoform specificity. The autoinhibitory leucine-rich repeat (LRR) domain represents an underexplored pharmacological target. Here, we report QX-31, a novel covalent probe selectively engaging the LRR domain. QX-31 covalently binds to Cysteine 838, stabilizing NLRP3 in an inactive conformation and preventing inflammasome assembly with high selectivity over AIM2 and NLRC4. Furthermore, QX-31 reverses maladaptive immunometabolic reprogramming within the renal microenvironment by blunting the PI3K/AKT/HIF-1α axis, suppressing aberrant glycolysis and glutaminolysis. Evaluated in human kidney organoids and murine models of renal injury (UUO and IRI), QX-31 demonstrated potent antifibrotic efficacy comparable to reference drugs, alongside a favorable safety profile. Collectively, QX-31 serves as a valuable chemical tool, demonstrating that covalent targeting of this conformational regulatory domain is a viable therapeutic strategy for fibrotic diseases.

The contractile force exerted by hepatic stellate cells (HSCs) plays a critical role in both physiological and pathological processes in the liver. Endothelin-1 (ET-1), a key inducer of HSC activation and contraction, can rapidly trigger cellular contractions within minutes, placing demands on the spatiotemporal resolution of detection tools. However, existing methods for measuring HSC mechanics often fail to simultaneously achieve precise mechanical quantification, high spatiotemporal resolution, and high-throughput analysis. To address these limitations, we employed a photonic crystal cellular force microscopy system that utilizes structural color changes of a photonic crystal substrate to sensitively detect minute cellular deformations and nanoscale vertical forces. This system integrates single-cell precision, high spatiotemporal resolution, and high-throughput capabilities. Using this platform, we successfully visualized and quantified the subcellular mechanical distribution within minutes during the contraction and migration of individual HSCs. Our findings demonstrate that photonic crystal cellular force microscopy provides a real-time, intuitive, and powerful approach for investigating HSC biomechanics, offering potential mechanical insights into liver disease mechanisms and drug screening.

Accurate detection of tumor organoids is essential for advancing drug discovery, personalized medicine, and disease modeling, yet it remains technically challenging. Organoid images often contain thousands of objects with large variations in size and frequent overlaps, and suffer from severe resolution distortions after image compression. Existing deep learning approaches struggle to balance accuracy with computational efficiency under these conditions. To address these limitations, we developed DTONet, a deep learning-based framework designed specifically for detecting small and overlapping organoids in high-resolution images. DTONet integrates a Bidirectional Feature Pyramid Network (BiFPN) for multi-scale feature fusion, Depth-Wise Separable Excitation (DWSE) modules for lightweight computation, and a Parallel Patch Inference (PPI) strategy to accelerate panoramic image processing. Across diverse organoid datasets (liver, lung, pancreas), DTONet achieved superior precision (93.4%) and mAP50:95 (58.9%), outperforming YOLOv8, YOLOv9, and Faster R-CNN by 5.9-9.3%. PPI reduced inference time by 22% while maintaining the detection accuracy, enabling real-time analysis of images up to 7344 & times; 6144 pixels. Notably, DTONet detected organoids smaller than 50 pixels with 89.3% accuracy and effectively resolved overlaps in high-density cultures, addressing key bottlenecks in automated image analysis. By combining hierarchical feature fusion, efficient parameterization, and parallel processing, DTONet establishes a robust foundation for high-throughput organoid research and expands the potential of tumor organoid-based applications in clinical and pharmaceutical settings.

Organoids have emerged as one of the most predictive preclinical models in medical research due to their ability to closely retain the genetic and phenotypic characteristics of original tissues. Nevertheless, the application in the systematic evaluation of innovative medical devices (especially biodegradable metals) remains largely unexplored, whereas conventional analytical approaches have significant limitations in data throughput, objectivity, and reproducibility. Recent advances in deep learning-based artificial intelligence (AI) image analysis offer powerful quantitative tools to overcome this bottleneck. In this study, we established a high-throughput quantitative in vitro evaluation platform for the dynamic assessment of biodegradable metal ions by integrating patient-derived colorectal cancer organoids with an OrganoSeg-based deep learning AI image analysis system. Systematic assessments revealed that Mg2+ and Zn2+ had significant concentration-dependent effects on organoid growth, and the organoid model exhibited sensitivity profiles distinct from conventional cell lines. RNA-seq analysis further revealed that high concentrations of Mg2+ induced cell cycle arrest by activating the p53/CDKN1C signaling axis. In contrast, high concentrations of Zn2+ disrupted intracellular zinc homeostasis by regulating metallothionein family members and the zinc transporter SLC39A10, triggering a robust inflammatory response and ultimately leading to apoptosis. This work not only confirms the considerable potential of integrating organoids with AI technology in the evaluation of medical devices, but also reveals the differential mechanism of action of bioactive metal ions Mg2+ and Zn2+ in a model closer to the human environment. This study establishes a reliable and generalizable paradigm for high-throughput and high-content biomedical research.

The rapid and accurate prediction of anticancer drug responses is critical for enhancing treatment efficacy and improving clinical outcomes in cancer patients. However, the practical implementation of current predictive models is hampered by dual limitations: machine learning approaches reliant on cell line data often exhibit suboptimal accuracy, while patient-derived organoids (PDOs) platform typically lack the rapid turnaround required for timely clinical decision-making. Here, we present a deep learning framework to predict drug response in lung cancer patients by integrating patient genomic sequencing data with compound structural information, trained against phenotypic drug sensitivity profiles from lung cancer PDOs, to predict drug responses in lung cancer patients. Our model enables individualized prediction of antitumor activity across diverse chemical structures, demonstrating capabilities for predicting efficacy of both approved drugs and novel compounds, as well as facilitating drug repurposing. The framework achieved 81.6% prediction accuracy, which was experimentally validated using patient-derived organoid models. More importantly, evaluation in a clinical cohort of lung cancer patients confirmed the model's ability to accurately reflect actual treatment responses. This study represents the first successful integration of genotype, drug structure, and organoid phenotype within a unified computational framework, significantly enhancing the accuracy and biological interpretability of drug response predictions while providing a clinically applicable tool for precision oncology in lung cancer.