中文导读
肾脏是药物毒性最易累及的器官之一。本文介绍肾小球与肾小管芯片模型及其在肾毒性筛选中的应用。
以下为英文全文（English full text）
1. Introduction: The Kidney as a Critical Drug Safety Target
The kidney is among the most vulnerable organs to drug toxicity, with nephrotoxicity contributing to approximately 19% of drug failures in Phase III clinical trials and 24% of preclinical attrition [1]. The renal proximal tubule is particularly susceptible due to its high metabolic activity, extensive active transport systems, and concentrated exposure to filtered xenobiotics. Current preclinical models—2D cell cultures and animal studies—exhibit poor predictive value for human nephrotoxicity, underscoring the urgent need for physiologically relevant human kidney models [2].
Kidney-on-chip (KoC) technology addresses this gap by recreating the structural and functional units of the kidney—the glomerular filtration barrier (GFB) and the renal tubules—within microfluidic devices. These platforms enable real-time monitoring of barrier integrity, drug transport, and cytotoxicity under precisely controlled flow conditions that replicate the hemodynamic forces experienced by renal cells in vivo [2].
2. Renal Physiology and Microarchitecture
The kidney's functional unit, the nephron, comprises the glomerulus (responsible for blood filtration) and the tubule system (responsible for selective reabsorption and secretion). Key physiological parameters for KoC design include:
Glomerular Filtration Barrier (GFB)
- Glomerular capillary hydrostatic pressure: 50–60 mmHg
- Filtration slit diaphragm width: 25–60 nm
- Glomerular basement membrane (GBM) thickness: 300–400 nm
- Podocyte foot process width: 200–500 nm
- Normal glomerular filtration rate (GFR): 90–120 mL/min/1.73m²
- Filtration fraction: 15–20%
Renal Proximal Tubule
- Tubular diameter: 50–60 μm
- Brush border microvilli height: 1–2 μm
- Fluid shear stress at apical surface: ~0.2–0.5 dyn/cm²
- Reabsorption rate: ~65–70% of filtered water and sodium
- Transporter density: High expression of OATs, OCTs, MATEs, and PEPT transporters
Renal Vasculature
- Peritubular capillary blood flow velocity: 100–300 μm/s
- Wall shear stress in peritubular capillaries: 1–5 dyn/cm²
- Oxygen tension in medulla: 10–40 mmHg (steep gradient)
3. Glomerular Filtration Barrier-on-Chip Design
3.1 Membrane-Based GFB Models
The glomerular filtration barrier comprises three layers: fenestrated glomerular endothelial cells (gECs), the basement membrane, and interdigitated podocytes with slit diaphragms. KoC devices recreate this architecture using porous membranes (0.4–1.0 μm pore size) with gECs seeded on the upper (blood) side and podocytes on the lower (urinary) side. Petrosyan et al. (2019) developed a glomerulus-on-chip using a track-etched membrane with 0.4 μm pores, achieving selective permeability that retained albumin (66 kDa) while permitting passage of inulin (5 kDa), consistent with in vivo GFB selectivity [3].
3.2 iPSC-Derived Podocyte Models
Musah et al. (2017) demonstrated that mature iPSC-derived human podocytes, when cultured on microfluidic chips with physiological shear stress, reconstituted kidney glomerular-capillary-wall function with appropriate albumin permeability and nephrin expression [4]. The use of patient-specific iPSC-derived podocytes enables modeling of genetic glomerular diseases such as congenital nephrotic syndrome and Alport syndrome.
3.3 Transwell-Integrated Microfluidic Platforms
Recent advances have developed SLAS-standard format microfluidic GFB-on-chip platforms using conventional 24-well transwell inserts. These systems align human glomerular mesangial cells (gMCs), podocytes, and gECs on each side of a transwell membrane, enabling pump-less perfusion and compatibility with high-throughput screening workflows. Flow rates of 15 μL/min generate wall shear stresses of ~0.16 dyn/cm², within the subdyne range targeted in kidney-on-chip studies [5].
4. Proximal Tubule-on-Chip Design and Construction
4.1 Membrane-Based Tubular Models
The landmark kidney proximal tubule-on-chip developed by Jang et al. (2013) used a two-channel PDMS device separated by a porous PDMS membrane [2]. Human primary proximal tubule epithelial cells (RPTECs) were seeded on the top channel (apical/luminal side), while human renal microvascular endothelial cells populated the bottom channel (basolateral/blood side). The apical channel was perfused with tubular fluid at 1 μL/min, generating shear stress of ~0.2 dyn/cm², while the basolateral channel received medium perfusion at 2 μL/min. This configuration maintained polarized transport for over 2 weeks, with cells forming brush borders, tight junctions, and expressing apical sodium-glucose cotransporter 2 (SGLT2) and basolateral Na⁺/K⁺-ATPase [2].
4.2 3D Tubular Structures
Alternative KoC designs use hollow tubular structures formed in collagen or Matrigel hydrogels within microfluidic devices. Cells are seeded around the inner surface of the tubule, recreating the cylindrical geometry of the nephron. These models enable investigation of tubular flow dynamics, cilia-mediated mechanosensing, and three-dimensional cell-cell interactions absent in planar membrane configurations.
4.3 Co-Culture and Immune Components
Advanced KoC platforms incorporate peritubular capillary endothelial cells (PCECs) and immune cells (macrophages, T-cells) to model inflammatory nephrotoxicity. Yin et al. (2020) developed a co-culture microfluidic kidney chip with RPTECs and PCECs, demonstrating that cimetidine intervention reduced cisplatin-induced nephrotoxicity through competitive inhibition of OCT2-mediated drug uptake [6].
5. Technical Specifications and Operating Parameters
6. Functional Validation and Nephrotoxicity Assessment
Barrier Integrity Markers
- Trans-epithelial electrical resistance (TEER): 50–200 Ω·cm² for healthy tubules; >1000 Ω·cm² for tight GFB models
- Albumin permeability: <5% passage for intact GFB; increased passage indicates barrier disruption
- Fluorescein isothiocyanate (FITC)-dextran permeability (4–70 kDa molecular weight series)
- Immunostaining for tight junction proteins (ZO-1, occludin, claudins), slit diaphragm proteins (nephrin, podocin), and adherens junctions (E-cadherin)
Transport Function
- Fluorescent substrate uptake: 6-carboxyfluorescein (6-CF) for OAT1/3, 4-(4-(dimethylamino)styryl)-N-methylpyridinium (ASP⁺) for OCT2, fluorescein for MRP2/4
- SGLT2-mediated glucose uptake: Measured using [¹⁴C]-α-methyl-D-glucopyranoside (α-MG) or fluorescent glucose analogs
- Na⁺/K⁺-ATPase activity: Ouabain-sensitive ATP hydrolysis
- Aquaporin-1 and urea transporter expression by Western blot or immunofluorescence
Nephrotoxicity Endpoints
- Cell viability (ATP content, MTT, LDH release)
- Apoptosis (caspase-3/7, TUNEL staining)
- Oxidative stress (ROS, MDA, GSH/GSSG ratio)
- Mitochondrial dysfunction (TMRM, JC-1)
- DNA damage (γ-H2AX, comet assay)
- Inflammatory markers (IL-6, IL-8, TNF-α, MCP-1)
- Kidney injury molecule-1 (KIM-1) expression (early biomarker of tubular injury)
- Clusterin and osteopontin (promising urinary biomarkers)
7. Drug Screening and Nephrotoxicity Applications
Cisplatin Nephrotoxicity
Cisplatin is a widely used chemotherapeutic agent whose clinical utility is limited by dose-dependent nephrotoxicity targeting the proximal tubule. KoC models have demonstrated that cisplatin accumulation via OCT2 and MATE transporters causes mitochondrial dysfunction, oxidative stress, and tubular apoptosis. Importantly, cimetidine—a known OCT2 inhibitor—reduces cisplatin uptake and attenuates toxicity in KoC platforms, validating the model's predictive value for clinical nephroprotection strategies [6].
Aminoglycoside Toxicity
Gentamicin and other aminoglycosides accumulate in proximal tubule cells via megalin-cubilin receptor-mediated endocytosis, causing lysosomal phospholipidosis and tubular cell death. KoC models with physiological shear stress show enhanced gentamicin uptake compared to static cultures, correlating with in vivo nephrotoxicity profiles. Kim et al. (2016) demonstrated that pharmacokinetic profiling in a perfused kidney-on-chip could identify dosing regimens that minimize nephrotoxicity while maintaining therapeutic efficacy [7].
Crystal Nephropathy
Drugs such as sulfadiazine, indinavir, and ciprofloxacin can precipitate as crystals in the renal tubule, causing obstruction and tubular injury. KoC platforms with luminal flow enable real-time visualization of crystal formation and assessment of tubular damage under dynamic conditions that static cultures cannot replicate.
Immune-Mediated Nephrotoxicity
Polymyxin B, tenofovir, and certain NSAIDs cause immune-mediated tubular injury. KoC platforms incorporating immune cells (e.g., THP-1 monocytes, primary macrophages) enable modeling of inflammatory responses and cytokine-mediated tubular damage, providing mechanistic insights that are difficult to obtain from animal models [2].
8. Conclusion
Kidney-on-chip technology has emerged as a powerful platform for nephrotoxicity screening, renal transport studies, and disease modeling. By recreating the glomerular filtration barrier and tubular microenvironment with physiological flow, shear stress, and multicellular interactions, KoC devices offer predictive value that exceeds conventional in vitro models. GBiowit provides specialized kidney-on-chip platforms, renal cell-compatible media, ECM matrices, and comprehensive drug screening services to support your nephrotoxicity assessment programs.