器官芯片技术：原理、平台与预测性应用

以下为英文全文（English full text）

1. Introduction: A Paradigm Shift in Preclinical Research

Organ-on-chip (OoC) technology represents one of the most significant advances in biomedical engineering over the past two decades, offering a transformative approach to modeling human physiology in vitro. By integrating microfluidic engineering with cell biology and tissue engineering, OoC platforms recreate the dynamic microenvironments of human organs on miniature devices, often no larger than a standard microscope slide. These systems provide physiologically relevant alternatives to traditional two-dimensional (2D) cell cultures and animal models, which have long been criticized for their limited predictive value in human drug development and disease modeling [1].

The concept of microphysiological systems (MPS) emerged from the recognition that biological function depends critically on tissue architecture, mechanical forces, and dynamic fluid interactions—elements that conventional static cultures fail to recapitulate. OoC devices address these limitations by incorporating living cells within microengineered environments that mimic tissue-tissue interfaces, vascular perfusion, mechanical stretching, and chemical gradients characteristic of native organs [1]. The technology has rapidly evolved from academic proof-of-concept demonstrations to commercially available platforms increasingly adopted by pharmaceutical companies, regulatory agencies, and research institutions worldwide.

2. Historical Development and Milestones

The origins of OoC technology can be traced to pioneering work in microfluidics during the 1990s. In 1998, George and coworkers reported a microfluidic method for drug screening and clinical diagnostics that enabled simultaneous execution of multiple trials with reduced variability [2]. The field gained substantial momentum with the introduction of polydimethylsiloxane (PDMS) as a soft, optically transparent elastomer ideally suited for biological applications. In 2010, Donald Ingber's group at the Wyss Institute demonstrated a landmark microfluidic device that replicated human lung functions using a thin flexible PDMS membrane, establishing the lung-on-chip paradigm [3]. This work demonstrated that mechanical forces—specifically, the cyclic stretching of alveolar-capillary interfaces—could be recreated in vitro and were essential for physiological responses to pathogen infection and nanoparticle exposure.

Subsequent years witnessed the development of organ-specific chips modeling the liver, kidney, intestine, heart, brain, and skin. The National Institutes of Health (NIH), Defense Advanced Research Projects Agency (DARPA), and Food and Drug Administration (FDA) provided significant funding through programs such as the NIH Tissue Chips in Space initiative, accelerating standardization and commercialization. By the mid-2010s, companies including Emulate, MIMETAS, CN Bio Innovations, and TissUse had launched commercial OoC platforms, while the FDA's ISTAND (Innovative Science and Technology Approaches for New Drugs) pilot program began accepting organ-chip data in regulatory submissions [4].

3. Microfluidic Technology Principles

The core operational principle of OoC technology is microfluidics—the precise manipulation of fluids at the microscale (typically 1–1000 micrometers). At these dimensions, fluid behavior is dominated by laminar flow rather than turbulent mixing, enabling highly controlled delivery of nutrients, drugs, and signaling molecules to cultured cells. The Reynolds number (Re) in microfluidic channels typically ranges from 0.001 to 100, ensuring predictable flow profiles and precise spatial control of chemical gradients [5].

Key microfluidic components in OoC devices include:

• Channel Networks: Microfabricated channels (typically 50–500 μm in width and 10–200 μm in height) guide fluid flow through cell culture chambers. Channel geometry—whether straight, branched, or networked—determines flow distribution, shear stress profiles, and mass transport characteristics. Computational fluid dynamics (CFD) modeling is routinely employed during device design to optimize flow patterns and ensure uniform perfusion.
• Porous Membranes: Many OoC devices incorporate thin, porous membranes (pore sizes 0.4–10 μm) that separate distinct tissue compartments while enabling molecular exchange. PDMS membranes (typically 10–50 μm thick) are most common, though polyester track-etched (PETE) and polycarbonate (PC) membranes are also widely used. These membranes recreate tissue-tissue interfaces such as the alveolar-capillary barrier or the blood-brain barrier (BBB) [1].
• Perfusion Systems: Continuous media perfusion is achieved through external syringe pumps, peristaltic pumps, or gravity-driven flow. Flow rates typically range from 1 μL/min to 1 mL/min, generating wall shear stresses of 0.1–15 dyn/cm² that match physiological values in blood vessels (1–15 dyn/cm²), renal tubules (~0.2 dyn/cm²), and liver sinusoids (~0.5 dyn/cm²) [6].
• Gradient Generators: Christmas-tree or serpentine microchannel structures create stable concentration gradients for studying chemotaxis, drug dose-response, and morphogen patterning. These devices exploit the predictable diffusion-limited mixing at the microscale to establish linear or logarithmic gradients across cell culture regions [5].

4. Chip Materials and Fabrication Considerations

The selection of materials for OoC fabrication critically influences device optical properties, biocompatibility, gas permeability, and drug adsorption characteristics. The most widely used materials include:

• Polydimethylsiloxane (PDMS): PDMS remains the most prevalent material in academic OoC research due to its optical transparency (transmittance >90% at 400–700 nm), gas permeability (oxygen diffusion coefficient ~4×10⁻⁵ cm²/s), and ease of fabrication through soft lithography [7]. However, PDMS has significant limitations: it absorbs small hydrophobic molecules (including many drugs and hormones), exhibits batch-to-batch variability in mechanical properties, and is challenging to manufacture at industrial scale. Surface treatments such as plasma oxidation, poly(ethylene glycol) (PEG) grafting, or protein coating are often required to improve hydrophilicity and cell adhesion [7].
• Polymethyl Methacrylate (PMMA): PMMA offers excellent optical clarity, rigidity, and compatibility with laser micromachining and injection molding. It is less permeable to gases than PDMS but is more resistant to small-molecule absorption, making it suitable for drug metabolism studies where compound adsorption must be minimized. PMMA is also significantly more cost-effective for mass production [8].
• Glass: Glass microfluidic devices provide exceptional optical clarity, well-defined surface chemistry, and resistance to organic solvents and high pressures. They are preferred for applications requiring high-resolution imaging, electroosmotic flow control, or chemical resistance. However, fabrication of high-aspect-ratio glass structures requires specialized etching techniques (wet etching with HF or dry etching with reactive ion etching), increasing manufacturing complexity [8].
• Thermoplastics and Emerging Materials: Polycarbonate (PC), cyclic olefin copolymer (COC), and polystyrene (PS) are increasingly used for commercially manufactured OoC devices due to their compatibility with injection molding and thermoforming. Polyurethanes (PU) have gained attention as alternatives to PDMS because they resist absorption of small hydrophobic molecules while maintaining biocompatibility and flexibility. Recent innovations include biodegradable elastomers, photocurable resins, and paper-based microfluidics for low-cost applications [8].

5. Cell Sources: From Primary Cells to iPSC-Derived Organoids

The physiological relevance of OoC models depends fundamentally on the cell types employed. Four major categories of cells are used:

• Primary Cells: Human primary cells isolated directly from tissues (e.g., primary hepatocytes, renal proximal tubule epithelial cells, pulmonary microvascular endothelial cells) offer the closest approximation to in vivo physiology. Primary hepatocytes, for instance, maintain drug-metabolizing enzyme activities (CYP450 isoforms) that are frequently lost in immortalized cell lines. However, primary cells have limited lifespan in vitro, exhibit donor-to-donor variability, and may be difficult to source in sufficient quantities [6].
• Immortalized Cell Lines: Established cell lines such as HepG2 (hepatocarcinoma), Caco-2 (colorectal adenocarcinoma), and A549 (lung carcinoma) provide unlimited, homogeneous cell populations ideal for standardized assays and high-throughput screening. Their limitations include reduced metabolic enzyme expression, altered gene expression profiles, and aneuploidy. Co-culture strategies and microfluidic conditioning can partially restore differentiated phenotypes in these cell lines [1].
• Induced Pluripotent Stem Cells (iPSCs): iPSCs offer a renewable source of patient-specific cells that can be differentiated into virtually any somatic cell type. iPSC-derived hepatocytes, cardiomyocytes, neurons, and endothelial cells have been successfully integrated into OoC platforms, enabling disease modeling with genetic backgrounds and pharmacogenomic profiling [9]. iPSC-derived cells are particularly valuable for modeling rare diseases, evaluating inter-individual variability in drug responses, and generating immune-compatible tissues for transplantation research. Challenges include batch variability, immaturity of differentiated phenotypes, and the complexity and duration of differentiation protocols.
• Organoids and Organoid-on-Chip: Organoids are self-organizing, three-dimensional tissue cultures derived from stem cells that recapitulate organ architecture and multicellular composition. When integrated into microfluidic chips (organoid-on-chip), these structures benefit from perfusion-mediated nutrient delivery, removal of metabolic waste, and application of mechanical forces such as shear stress and cyclic stretch. This integration addresses key limitations of static organoid culture, including necrotic core formation, limited size, and lack of tissue-tissue interfaces [10]. Organoid-on-chip platforms have demonstrated enhanced vascularization, maturation, and functional performance compared to traditional organoid culture methods.

6. Multi-Organ-on-Chip Systems and Body-on-a-Chip

While single-organ chips provide valuable insights into organ-specific biology, the human body functions as an integrated system where organs communicate through circulating hormones, metabolites, and immune cells. Multi-organ-on-chip (multi-OoC) platforms address this complexity by interconnecting multiple organ modules via microfluidic channels, creating a microphysiological system that approximates systemic physiology [11].

Multi-OoC configurations include:

• Serial Connection: Organ modules are connected in series via tubing or microfluidic channels, with media flowing sequentially from one organ to the next. This arrangement mimics first-pass metabolism (e.g., intestine → liver) and enables study of metabolite-mediated organ crosstalk. Flow rates and media volumes must be carefully scaled to physiological blood volume ratios (QVIVE—quantitative in vitro-to-in vivo extrapolation) [11].
• Parallel Connection with Common Media: Multiple organ modules share a common circulating medium, allowing bidirectional communication. This approach better recapitulates systemic circulation but requires careful media formulation to satisfy the requirements of all co-cultured cell types.
• Physiologically-Based Pharmacokinetic (PBPK) Integration: Computational PBPK models are coupled with multi-OoC platforms to predict drug concentrations in different organs over time. This integration enables quantitative translation of in vitro findings to in vivo predictions, supporting clinical dose estimation and risk assessment [12].

Notable multi-OoC achievements include the vascularized heart-liver-bone-skin platform developed by Ronaldson-Bouchard et al., which maintained functionality for four weeks and supported chronic drug toxicity studies [13]. The "human body-on-a-chip" concept extends this approach further, aiming to incorporate ten or more interconnected organ systems with immune components, microbiome interfaces, and neural innervation [11].

7. ADMET Applications: Absorption, Distribution, Metabolism, Excretion, and Toxicity

ADMET profiling is a cornerstone of drug development, and OoC technology is revolutionizing each component:

• Absorption: Intestinal-on-chip models with villus-like structures and mucus-secreting goblet cells predict oral drug absorption more accurately than Caco-2 monolayers. Airway-on-chip models with air-liquid interface (ALI) culture assess pulmonary absorption of inhaled therapeutics [14].
• Distribution: BBB-on-chip platforms quantify drug permeability across the neurovascular interface, addressing a major bottleneck in central nervous system (CNS) drug development. These models incorporate shear stress-responsive tight junctions that closely match in vivo BBB permeability coefficients [15].
• Metabolism: Liver-on-chip platforms maintain primary hepatocyte CYP450 activity for weeks rather than days, enabling accurate prediction of metabolic clearance, drug-drug interactions (DDIs), and bioactivation of prodrugs. The combination of liver chips with mathematical modeling (PBPK) allows quantitative in vitro-to-in vivo extrapolation (QIVIVE) of metabolic parameters [16].
• Excretion: Kidney-on-chip models reconstitute tubular transport and glomerular filtration, enabling prediction of renal clearance and identification of nephrotoxic mechanisms involving active transporters and drug accumulation [17].
• Toxicity: OoC platforms have demonstrated predictive value for drug-induced liver injury (DILI), cardiotoxicity, nephrotoxicity, and pulmonary toxicity. The FDA's ISTAND program accepted the first organ-on-chip submission for predicting DILI in 2024, validating these models for regulatory decision-making [4].

8. Regulatory Acceptance and the FDA Modernization Act 2.0

The regulatory landscape for OoC technology transformed dramatically with the passage of the FDA Modernization Act 2.0 in December 2022. This landmark legislation amended the Federal Food, Drug, and Cosmetic Act of 1938, which had mandated animal testing for all new drug applications. The new law explicitly permits the use of "nonclinical tests" conducted "in vitro, in silico, or in chemico, or a nonhuman in vivo test," including cell-based assays, organ chips and microphysiological systems, computer modeling, and other human biology-based methods [18].

Key regulatory developments include:

• FDA ISTAND Pilot: The FDA's Center for Drug Evaluation and Research (CDER) launched the ISTAND pilot program to facilitate the use of innovative technologies in drug development. In 2024, the FDA accepted its first organ-on-chip submission—Emulate's Liver-Chip for predicting human drug-induced liver injury (DILI)—marking a historic milestone for the regulatory acceptance of OoC data [4].
• European Initiatives: The European Organ-on-Chip Society (EUROoCS) and the Organ-on-Chip Development Project (ORCHID) have established roadmaps for standardization and validation. The European Medicines Agency (EMA) published guidelines encouraging the use of 3Rs (replacement, reduction, refinement) testing approaches, including organ chips, in regulatory submissions [19].
• International Standardization: ISO 22916:2022 established interoperability requirements for microfluidic devices, supporting the development of standardized, modular OoC platforms. The Foundation for NIH's Validation and Qualification Network brings together FDA, EPA, and industry partners to standardize New Approach Methodologies (NAMs) [20].

The regulatory shift does not ban animal testing but makes it optional when scientifically validated alternatives are available. This change is expected to accelerate pharmaceutical innovation, reduce drug development costs, and improve the human relevance of preclinical safety data.

9. Organoid-on-Chip: Converging Technologies

The integration of organoid technology with microfluidic platforms represents a rapidly advancing frontier. Organoids provide multicellular complexity and self-organizing architecture, while microfluidic chips provide physiological perfusion, mechanical forces, and real-time analytical access. Together, they overcome limitations of each approach alone [10].

Key advantages of organoid-on-chip systems include:

• Enhanced Vascularization: Microfluidic perfusion of organoids promotes vessel ingrowth and maturation. Homan et al. demonstrated that flow-enhanced vascularization significantly improved kidney organoid maturation and nephron patterning [10].
• Controlled Microenvironment: Real-time regulation of oxygen tension, nutrient gradients, and mechanical forces (shear stress, cyclic stretch) enables precise control over organoid development and function.
• Tissue-Tissue Interfaces: Organoid-on-chip platforms can position organoids adjacent to endothelial, immune, or stromal cell populations, creating physiologically relevant tissue boundaries that are absent in standard organoid culture.
• Long-Term Culture: Continuous perfusion removes metabolic waste and delivers fresh nutrients, extending organoid viability from weeks to months and enabling chronic drug exposure studies.

Applications span disease modeling (cystic fibrosis, polycystic kidney disease, neurodevelopmental disorders), personalized medicine (patient-derived organoids for drug screening), and regenerative medicine (scalable tissue production for transplantation) [21].

10. Conclusion and Future Perspectives

Organ-on-chip technology has progressed from an academic curiosity to a commercially viable platform with demonstrated regulatory acceptance. As microfabrication techniques mature, materials improve, and iPSC differentiation protocols become more robust, OoC systems will increasingly replace conventional preclinical models. The convergence of organoid technology, multi-organ integration, and artificial intelligence-driven data analysis promises to create "clinical trials in a dish" that predict human drug responses with unprecedented accuracy. For organizations seeking to adopt these technologies, GBiowit offers comprehensive organ-on-chip products, culture media, matrices, and specialized services in organoid modeling, drug screening, and advanced model development.

返回技术文章列表