中文导读
人体是多器官协同的整体系统。本文介绍多器官芯片的系统集成、生理互联与全身水平药代研究。
以下为英文全文（English full text）
1. Introduction: From Single Organs to Systemic Physiology
The human body operates as an integrated system where organs communicate through circulating blood, shared metabolites, hormones, and immune signals. A drug absorbed in the intestine is metabolized by the liver, excreted by the kidney, and may exert toxic effects on the heart—none of which can be fully understood by studying organs in isolation. This systemic complexity is a primary reason why approximately 90% of drug candidates that succeed in preclinical studies fail in human clinical trials [1].
Multi-organ-on-chip (multi-OoC) technology addresses this fundamental limitation by interconnecting multiple organ-specific microphysiological systems within a common fluidic circuit. These platforms enable the study of organ-organ crosstalk, first-pass metabolism, systemic toxicity, and pharmacokinetic/pharmacodynamic (PK/PD) relationships in a human-relevant context. Often termed "body-on-a-chip" or "human-on-a-chip," these systems represent the frontier of in vitro modeling and are increasingly recognized by regulatory agencies as acceptable alternatives to animal testing [2].
2. Principles of Multi-Organ Integration
Scaling and Physiological Fidelity
Successful multi-OoC platforms must address the challenge of scaling—matching fluid volumes, flow rates, and cell numbers to human physiological ratios. The human body contains approximately 5 L of blood, with organ-specific blood volumes and flow rates that determine drug distribution and organ exposure. Microphysiological systems cannot replicate absolute scale but must preserve relative scaling (e.g., liver blood flow vs. kidney blood flow vs. cardiac output) to generate physiologically meaningful PK data [1].
Key scaling principles include:
- Volume scaling: Media volume per cell number should approximate in vivo blood volume to tissue mass ratios
- Flow rate scaling: Perfusion rates should match organ-specific blood flow (e.g., liver: 25% of cardiac output; kidney: 20%)
- Residence time: The time a drug spends in contact with each organ should approximate in vivo perfusion times
- Metabolite clearance: Systemic clearance of metabolites should not be artificially limited by small media volumes
Fluidic Interconnection Strategies
- Tubing-based serial connections: Individual organ chips are connected via microfluidic tubing in series or parallel configurations. This approach offers flexibility but introduces dead volumes and requires peristaltic or syringe pumps.
- Microfluidic motherboards: Centralized microfluidic distribution manifolds connect multiple organ modules through standardized ports, reducing tubing complexity and enabling "plug-and-play" modularity [3].
- Recirculating loops: Media is continuously circulated through organ modules and a reservoir, mimicking systemic circulation. A "liver equivalent" volume of 1–5 mL per 10⁶ hepatocytes is typical.
- Pumpless gravity-driven systems: Height differences between reservoirs create hydrostatic pressure for flow, eliminating pump-induced pulsation and simplifying operation. However, flow rates are less precisely controlled [4].
3. Platform Architectures and Commercial Systems
3.1 Modular Multi-Organ Systems
TissUse's HUMIMIC platform and MIMETAS' OrganoPlate represent modular approaches where individual organ modules (liver, kidney, intestine, brain, heart) are interconnected via a central perfusion system. The modular design allows researchers to customize organ combinations and introduce disease-specific modifications. Standardized ISO 22916:2022 connections support interoperability between different manufacturers' devices [3].
3.2 Integrated Body-on-Chip Devices
The four-organ-chip developed by the Technical University of Berlin integrates liver, intestine, skin, and kidney compartments on a single PDMS device with on-chip peristaltic micropumps. This platform demonstrated repeated-dose toxicity testing capabilities for up to 28 days, with automated media recirculation and sampling [4].
3.3 Vascularized Multi-Organ Platforms
Ronaldson-Bouchard et al. developed a vascularized multi-organ chip interconnecting heart, liver, bone, and skin tissues via a biomimetic vascular network. The system maintained functionality for four weeks, enabling chronic drug toxicity studies and demonstrating that vascular perfusion enhanced tissue maturation and metabolic activity across all organ modules [5].
3.4 Robotic Multi-Organ Systems
Advanced platforms integrate robotic liquid handling for automated media exchange, compound dosing, and sample collection. These systems support high-throughput screening workflows while maintaining the physiological connectivity of multi-organ interactions. Integration with PBPK modeling software enables real-time prediction of human plasma drug concentrations based on in-chip measurements [6].
4. Organ Combinations and Physiological Coupling
Gut-Liver Axis
The intestine-liver combination is one of the most extensively studied multi-OoC configurations. The intestinal module (Caco-2 or intestinal organoids) absorbs oral compounds, which then flow to the liver module for first-pass metabolism. This arrangement is critical for predicting oral bioavailability and identifying gut-wall metabolites that may differ from hepatic metabolism. Yang et al. constructed a gut-liver chip that modeled bidirectional regulatory feedbacks in non-alcoholic fatty liver disease metabolism, demonstrating the unique advantage of multi-organ coupling in disease mechanism studies [7].
Liver-Kidney Axis
The liver-kidney combination enables assessment of hepatic metabolite nephrotoxicity and renal clearance of parent drugs and metabolites. Theobald et al. (2019) demonstrated metabolic activation of vitamin D3 in a liver module followed by renal clearance and bioactivation in a kidney module, showcasing the sequential processing capability of multi-OoC systems [8].
Liver-Heart Axis
Cardiotoxicity of hepatically metabolized drugs (e.g., terfenadine, astemizole) can only be predicted when liver and heart modules are functionally coupled. iPSC-derived liver and heart microtissues on a shared circuit predicted unsafe drug-drug interactions that static cultures failed to identify, demonstrating the value of metabolic activation in systemic toxicity assessment [9].
Gut-Liver-Immune System
Kim et al. developed a gut-liver-immune chip incorporating macrophages and T-cells, successfully recapitulating the entire process of gut barrier dysfunction, aberrant lipid transport, and immune response activation in metabolic disease. This tri-organ system demonstrated that multi-organ platforms with immune components can model complex disease mechanisms inaccessible to single-organ models [10].
Liver-Skin Axis
Troglitazone, a withdrawn antidiabetic drug, causes hepatotoxicity that may be preceded by skin reactions. A liver-skin multi-organ chip replicated the systemic distribution of troglitazone and its metabolites, enabling identification of skin-specific toxicities and cross-organ biomarker signatures [4].
5. Technical Specifications and Operating Parameters
| Parameter | Typical Range | Considerations |
|---|---|---|
| Number of organs | 2–10 modules | Common: 4–6 organ types |
| Media volume | 1–20 mL total | Scale to human blood volume |
| Flow rate per organ | 5–100 μL/min | Organ-specific scaling |
| Recirculation rate | 0.5–5 mL/hour | Maintain physiological residence times |
| Media change | Every 24–72 hours | Or continuous replenishment |
| Culture duration | 2–8 weeks | Chronic toxicity studies |
| Connection tubing | 0.5–1.0 mm ID | Minimize dead volume |
| Temperature | 37°C | Uniform across all modules |
| pH monitoring | 7.2–7.4 | Continuous or endpoint |
| Oxygen monitoring | 5–95 mmHg | Organ-specific requirements |
6. ADME and Systemic Toxicity Applications
Pharmacokinetic Modeling
Multi-OoC platforms generate time-course concentration data from multiple organ compartments simultaneously. When coupled with physiologically-based pharmacokinetic (PBPK) modeling, these data enable prediction of human plasma concentration-time profiles, tissue distribution, and clearance pathways. The European Commission's Joint Research Centre has published guidelines for PBPK-OoC integration, supporting regulatory use of these combined approaches [1].
Repeated-Dose Toxicity
Traditional acute toxicity assays fail to identify toxicities that emerge only after chronic exposure. Multi-OoC platforms maintain organ viability for weeks, enabling repeated-dose studies that match the duration of regulatory animal toxicity studies. The four-organ-chip demonstrated 28-day repeated-dose toxicity testing with automated sampling, generating histopathological, transcriptomic, and metabolic endpoints comparable to standard in vivo studies [4].
Drug-Drug Interactions (DDIs)
When multiple drugs are administered simultaneously, metabolic competition and enzyme induction/inhibition can dramatically alter drug exposure. Multi-OoC platforms with liver modules predict the magnitude of clinical DDI risk by quantifying changes in metabolite production across the systemic circuit. Incorporation of pharmacogenetically diverse iPSC lines enables population-level DDI prediction [6].
Cancer Drug Screening
Multi-organ chips connect tumor modules with liver, kidney, and bone marrow compartments to assess systemic anticancer drug effects. This configuration enables simultaneous evaluation of tumor efficacy, hepatic metabolism, renal clearance, and off-target hematopoietic toxicity. Patient-derived tumor organoids integrated into multi-OoC platforms offer personalized cancer medicine applications [5].
Metabolic Disease Modeling
Type 2 diabetes, non-alcoholic steatohepatitis (NASH), and metabolic syndrome involve multi-organ dysfunction (pancreatic islet failure, hepatic insulin resistance, adipose inflammation, muscle glucose uptake defects). Multi-OoC platforms coupling pancreatic islets, liver, adipose tissue, and skeletal muscle recreate disease-specific metabolic phenotypes and enable evaluation of multi-targeted therapeutics [7].
7. Validation and Regulatory Pathways
The validation of multi-OoC platforms for regulatory decision-making requires rigorous benchmarking against human clinical data. Key validation principles include:
Reference Compound Sets
Well-characterized drugs with known clinical PK/PD and toxicity profiles are used to validate platform performance. A platform must accurately predict the safety and metabolism of training compounds before it can be used for novel drug assessment [2].
Cross-Laboratory Reproducibility
Multi-site studies demonstrating reproducible results across different laboratories are essential for regulatory acceptance. The NIH NCATS Tissue Chips program has funded extensive cross-validation studies to establish reliability standards [3].
Biomarker Concordance
Platform-derived biomarkers (albumin, urea, creatinine, troponin, LDH, cytokines) must correlate with clinical biomarkers of organ function and injury. FDA qualification of novel biomarkers for regulatory use is facilitated through the ISTAND and CDER biomarker qualification programs [2].
Good Cell Culture Practice (GCCP)
Adherence to OECD and ISSCR guidelines for good cell culture practice ensures reproducibility, traceability, and data quality. Standardized operating procedures (SOPs), batch-to-batch quality control, and documented cell authentication are mandatory for regulatory submissions [1].
8. Future Perspectives and Challenges
Despite remarkable progress, multi-OoC technology faces challenges that must be addressed for widespread adoption:
Scalability and Throughput
Current multi-OoC platforms accommodate relatively small numbers of cells per organ, limiting statistical power and complicating omics analyses. High-throughput multi-OoC systems (HT-OoC) under development aim to parallelize 96 or more multi-organ circuits for screening applications [6].
Immune System Integration
The immune system modulates drug responses, inflammation, and tissue repair. Incorporating circulating immune cells (macrophages, T-cells, neutrophils) and tissue-resident immune populations into multi-OoC platforms remains technically challenging but is essential for modeling immunotherapy, autoimmune disease, and infection [10].
Microbiome Interfaces
The gut microbiome influences drug metabolism, immune activation, and systemic inflammation. Co-culture of intestinal organoids with defined bacterial communities within multi-OoC platforms will enable mechanistic studies of microbiome-drug-host interactions [7].
Vascularization and Innervation
Functional blood vessels and nerve networks are essential for organ maturation and systemic integration. Advances in microvascular engineering, including sprouting angiogenesis from endothelial cells and 3D bioprinting of vascular networks, will enhance the physiological fidelity of multi-OoC platforms [5].
Artificial Intelligence Integration
Machine learning algorithms trained on multi-OoC data can predict clinical outcomes, optimize drug dosing regimens, and identify novel therapeutic targets. The combination of multi-organ experimental data with AI-driven analysis promises to create "digital twins" of patient physiology for personalized medicine [6].
9. Conclusion
Multi-organ-on-chip technology represents the culmination of organ-on-chip development, enabling systemic drug assessment that captures the complexity of human physiology. By interconnecting liver, kidney, intestine, heart, lung, brain, and immune modules within microfluidic circuits, these platforms predict ADME properties, drug-drug interactions, and organ-specific toxicities with unprecedented accuracy. As regulatory frameworks evolve and standardization advances, multi-OoC systems will become indispensable tools in pharmaceutical development. GBiowit provides comprehensive multi-organ-on-chip platforms, modular organ modules, specialized culture media, and integrated drug screening services to advance your systemic pharmacology research.