How Organoid Chips Achieve Precise Reproduction of 0%–21% Oxygen Gradients, Revealing the Mechanisms of Hypoxic Metabolic Reprogramming in Tumors and Ischemic Diseases

1. Overview

The goal is to construct and stably maintain a spatial oxygen gradient ranging from normoxia (≈21% O₂ gas phase) to anoxia/zero oxygen (0% O₂ gas phase) on a reproducible and quantifiable organoid/tumor-on-chip platform, and to achieve real-time/periodic monitoring and biological sampling to elucidate hypoxia-induced metabolic reprogramming (such as HIF pathways, glycolysis/lactate production, mitochondrial function changes, etc.). The implementation methods commonly use two types of technical routes (which can be used separately or in combination):

1. Gas-channel control + PDMS/membrane diffusion method: Set gas channels above the fluid layer to generate the desired gas phase O₂ distribution through precise mixing (air/nitrogen or oxygen), allowing O₂ to diffuse into the culture medium layer through a permeable membrane (usually PDMS), thereby establishing a liquid phase oxygen gradient. This method can produce spatial oxygen gradients of any shape and achieve 0.1% level control.
2. Chemical oxygen consumption/oxygen-consuming reactions or physical barriers (diffusion barrier) to generate gradients: In a single-layer channel, a steady-state gradient is formed through local oxygen consumption (e.g., using oxygen-consuming chemical reactions or setting high-density cell/gel areas) and barrier layers, which is structurally simpler and facilitates optical imaging.

2. Required Materials and Equipment

Materials

• Elastic polymer sheets (PDMS) or equivalent gas-permeable membranes (recommended thickness variable 50–500 µm)
• Glass bottom plates or injection-molded plates (for microscopic observation)
• Microfluidic channel molds (silicon or aluminum molds) or laser cutting/CNC machined parts
• Basic cell culture consumables: culture medium, serum-free/serum-containing solutions depending on cell lines, basic ECM (basement membrane matrix gel or synthetic hydrogel)
• Precision gas mass flow controllers (MFC) or multi-channel gas mixers (capable of accurately mixing air/O₂/N₂ to 0.1% resolution)
• Oxygen sensors (choose one or a combination):
• Fluorescent/phosphorescent oxygen-sensitive membranes (based on Ru- or Pt-porphyrin type probes), which can be attached to channels or embedded in ECM.
• Or micro Clark electrodes (for instantaneous point measurement).
• Optical imaging equipment: inverted microscope (with fluorescence channel), phase contrast/confocal for cell morphology, HIF and other protein immunostaining imaging.
• Microinjection pump or syringe pump (flow rate range 0.01–100 µL/min)
• Temperature chamber/cell culture incubator (37°C, 5% CO₂, or adjusted according to cell needs)
• Nitrogen (or argon) and compressed air or oxygen cylinders (gas source)
• Oxygen calibration solutions or chemical oxygen consumption standards (for sensor calibration)
• Common molecular biology reagents: RNA extraction reagents, protein extraction, lactate/glucose detection kits, metabolic flux measurement (if needed for transfer plates like Seashore, etc.)

Optional

• Temperature/humidity online monitoring module (to ensure stable gas flow and culture)
• Integrated microfluidic valves or bubble exclusion modules (to reduce bubble interference)
• Data acquisition system (for real-time recording of oxygen concentration, temperature, flow rate in each channel)

3. Device Design

3.1Two-layer PDMS (gas channel above/liquid channel below) — Recommended for precise 0–21% control

• Layer structure: From top to bottom: gas inlet and outlet → gas channel (height: 100–300 µm, width: 200–1000 µm) → PDMS membrane (thickness: 50–200 µm) → liquid culture channel (height: 100–500 µm).
• Alignment of gas and liquid channels: Place the gas channel directly above the liquid channel, ensuring the gas channel length matches the liquid channel length to achieve a stable one-dimensional gradient.
• Recommended membrane thickness: 50–150 µm can establish and respond to O₂ changes within minutes to hours; thinner membranes respond faster but are prone to mechanical weaknesses, while thicker membranes require longer to establish steady state (can be estimated through calculations or experiments).
• Scale and gradient spatial scale: If a linear gradient is desired within a spatial scale of 1–5 mm, the liquid channel width is recommended to be 1–5 mm (across direction), with the gas mixer supplying gas in a linear distribution (see below).
• Gas flow rate: The gas laminar flow rate in the gas channel does not need to be very high (recommended 5–200 sccm, depending on the gas channel cross-section), the key is to maintain precise and stable mixing ratios. Literature shows that this structure can control liquid phase O₂ to 0.1% level.

3.2Single-layer channel + isolation barrier (diffusion barrier) — Simplified version, easy for imaging

• Structure: A single-layer liquid channel with an oxygen-impermeable “barrier/obstacle” placed in the middle or with isolation walls with ventilation holes on both sides, with cells or gel placed in the center of the channel. Gradients are established by different gases or oxygen-consuming areas on both sides (diffusing towards the center).
• Applicable scenarios: Suitable for gradient experiments that pursue easy processing and low-cost implementation; effective when there are no extreme requirements for response speed.

4. Gas Mixing and Gradient Generation

4.1Gas sources and mixing

• Use two or three mass flow controllers (MFC): For example, Air (or 21% O₂ + balance N₂) mixed with pure N₂ (or pure O₂) in two routes, or Air/N₂/O₂ in three routes for faster adjustments.
• Objective: To accurately set the flow rates of each route through MFC to achieve the desired gas phase O₂ ratio (for example, if the two routes are Air (21% O₂) + N₂, set Air flow rate = 0–100 sccm, the relative ratio directly determines O₂%). MFC resolution is recommended to be ≤0.5 sccm to achieve ≤0.1% O₂ adjustment accuracy (depending on specific MFC specifications).
• Gas filtration: Gas pipelines should be equipped with 0.2 µm filters, and humidification (or not) is determined by experimental needs; humidification can affect the dissolved oxygen transfer rate.
• Pressure stabilization: Stabilize pressure to 0.1–0.5 psi to prevent bubbles from pushing the PDMS deformation.

4.2Gradient formation strategies

• Arbitrary shape gradients: Parallel multiple channels in the gas channel and supply different O₂% gas flows, or use solenoid valves for dynamic mixing, combined with PDMS membrane diffusion to obtain programmable gradients. Literature examples demonstrate control of arbitrary shapes of O₂ distribution.
• Linear gradient (commonly used) construction method: Supply 0% and 21% gases on both sides, and let the gases mix in a gradient through designed serial/mixing microstructures (or use electronically controlled multi-way mixers); O₂ diffuses under the membrane to form a liquid phase linear gradient. It is recommended to use numerical simulations (COMSOL or equivalent) to estimate gradients during design, and then verify through sensors.

5. Sensor Arrangement and Calibration

5.1Sensor types and installation locations

• Fluorescent/phosphorescent oxygen-sensitive membranes/arrays (recommended for microfluidics): Can be directly attached to glass slides or embedded in PDMS, non-invasive, fast response, suitable for imaging with spatial resolution of 10s–100s µm.
• Electrochemical (Clark) microelectrodes: Used for point verification but slightly larger in size, which may disturb the flow field.
• Recommended arrangement: Place a sensor point every 200–500 µm along the transverse (gradient direction) or use imaging-type oxygen-sensitive membranes for continuous spatial readings.

5.2Calibration

1. Two-point calibration (minimum):

• 0% reference: Use pure nitrogen (or gas containing 0% O₂) to flow through the gas channel and confirm with deoxygenated liquid/chemical methods (e.g., adding oxygen-consuming agents to the fluid and waiting for steady state).
• 100% reference: Air (21% gas phase) at atmospheric pressure, or use a known dissolved O₂ saturated solution (at 37°C, air saturation value ≈ 0.21 × gas phase saturation constant, specific values should follow solubility tables).

Stern–Volmer curve (for fluorescent/phosphorescent sensors): Use multiple points such as 0%/5%/10%/21% to construct response curves and linear fit to backtrack local O₂.

Temperature control and salinity correction: Dissolved oxygen is sensitive to temperature (calibration is recommended at experimental temperature 37°C).

6. Step Analysis (From Chip Preparation to Biological Sampling)

Below is a complete example process (gas channel + PDMS membrane scheme), with numbers and times being suggested values that can be adjusted according to cell types.

6.1Chip Preparation (Day 0)

1. Manufacture/prepare the chip: Cast PDMS according to the design mold (recommended base:cure = 10:1 ratio), degas, and cure at 80°C for 2 hours. Cut and clean.
2. Drill and install connectors: Drill holes at the gas inlet and outlet and assemble Luer connectors or tubing connectors, and install fluid connectors at the liquid channel end.
3. Plasma surface treatment and bonding with glass: Plasma treatment for 30–60 seconds, immediately bond with glass, and place in an oven at 60°C for 30 minutes to enhance bonding.
4. Attach oxygen-sensitive membranes/arrays (if using imaging probes): Attach the sensing membrane at predetermined locations (or mix fluorescent probe particles during PDMS casting). Ensure the sensing points face the liquid phase.

6.2Pre-operation and gas line debugging (Day 0)

1. Connect the gas source and use MFC to mix gases to set 21% and 0% at both ends, first introducing air and nitrogen into the left and right gas channels respectively, confirming no leaks.
2. Slowly perfuse the liquid channel with cell-free culture medium (0.5–5 µL/min), expelling bubbles (using a foam exclusion device or briefly backflushing).
3. Run for 1–2 hours, observe oxygen-sensitive membrane readings and perform calibration (see section 5). Record stabilization time (usually 10–60 minutes depending on membrane thickness).

6.3Cell/organoid seeding (Day 1)

1. Prepare organoids/tumor spheroids: Prepare 3D spheroids or organoids (recommended diameter 100–500 µm, density 50–200 spheroids/mL, controlling total number based on channel volume).
2. Mix ECM: Mix organoids with basement membrane matrix or synthetic hydrogel at a 1:1 (volume ratio) or according to actual viscosity requirements (final gel volume depends on channel volume), avoiding bubbles.
3. Inject into the channel: Use a pipette or syringe pump to fill the gel + organoids at a flow rate of 0.5–5 µL/min, allowing the gel to solidify at room temperature/37°C (time 15–30 minutes, depending on ECM).
4. Slowly perfuse culture medium: After gel solidification, perfuse culture medium at a low flow rate of 0.1–1 µL/min, without changing the gas gradient within 24 hours to allow cell adaptation.

6.4Establish and record gradients (Days 1–n)

1. Set gas percentages: Use MFC to set one side gas to 21% (air) and the other side to 0% (pure N₂ or treated with oxygen scavengers). Or set intermediate values according to experimental design (e.g., 21%→10%→5%→0% continuous gradient).
2. Record oxygen distribution: Real-time imaging with oxygen-sensitive membranes (time intervals depending on needs, recommended initially every 5–15 minutes, and after stabilization every 1–6 hours). Record data for at least 3 stable cycles to confirm repeatability.
3. Steady-state time: Typical steady-state establishment time ranges from minutes to hours (depending on membrane thickness, gas flow rate, liquid flow rate), it is recommended to record the first experiment for 6–24 hours to determine the steady-state curve.

6.5Biological sampling and functional assessment

1. Metabolic indicators: Collect outflow culture medium (volume calculated based on flow rate, collected every 24 hours) for measuring lactate, glucose consumption, ATP, pH changes, etc. (using colorimetric/fluorescent kits according to instructions).
2. Molecular detection: Sample at predetermined time points (e.g., 6 h, 24 h, 48 h, 72 h) (recover organoids from the chip or directly lyse with lysis buffer) for qPCR (HIF1A, HIF2A, GLUT1, LDHA, etc.), Western blot (HIF-1α, CA9, PDK1) or RNA-seq.
3. Functional readouts: Staining (live cell staining/immunofluorescence) to observe mitochondrial membrane potential (JC-1 type probes), ROS (DCFDA), cell death (Annexin V/PI), etc.
4. Cell morphology and migration: Use real-time microscopy to observe local cell invasion, vascularization, or stromal cell remodeling.

7. Data Interpretation and Common Expected Results

• Under stable linear gradient conditions, a central hypoxic region typically shows accumulation of HIF-1α, upregulation of GLUT1/LDHA, and increased lactate secretion within 6–48 hours; high oxygen-consuming cells (tumor cells) will show a metabolic shift to glycolysis at the hypoxic end. Reference existing microfluidic-ToC reports for comparison.
• Cross-sectional oxygen distribution maps can infer local oxygen consumption rates (OCR) — combining steady-state diffusion equations with measured sensor data can estimate regional oxygen consumption power (requires numerical fitting/simulation tools). Literature examples provide methods and models.

8. Common Issues and Solutions

1. Unable to reach 0% (readings do not drop to zero)

• Check for gas line leaks; confirm the gas source is indeed pure N₂ (or use gas with O₂ adsorbents).
• Thickening the gas-liquid barrier may reduce sensor sensitivity; refer to calibration curves and extend steady-state time.

Sensor drift or fluorescence quenching

• Recalibrate with multiple points (Stern–Volmer); avoid prolonged direct exposure of strong lasers to the sensing membrane to prevent photobleaching.

Bubble issues

• Add humidifiers/bubble catchers to the gas line or set up a Y-type exhaust branch at the fluid inlet; ensure all tubing is degassed before operation.

Poor cell/organoid survival

• Check ECM density, perfusion speed (being washed away or insufficient nutrient supply can affect survival); low flow rates (0.1–1 µL/min) are often used for 3D gel cultures to ensure nutrient supply while maintaining gradients.

Gradients are unstable or drift over time

• May be caused by temperature/pressure changes or unstable gas supply; use data loggers to synchronize temperature and pressure recording during experiments.

9. Extensions and Recommendations

• Closed-loop control: Feed oxygen sensor readings back to MFC to achieve real-time closed-loop control (adjust gas ratios as needed) to maintain long-term steady state or dynamic changes (periodic hypoxia/reoxygenation), existing platforms have reported feasibility and used for multi-organ interconnected experiments.
• Parallelization and high-throughput: Expand single channels into arrays, with each channel having separate MFC or using multi-way solenoid valves for rapid switching to support parallel control experiments under different gradient conditions.
• Combine with metabolomics/mass spectrometry: Collect outflow liquid for LC-MS metabolomics analysis, which can provide more detailed information on metabolic network remodeling.
• Modeling: It is recommended to perform COMSOL or custom finite element modeling before experiments to predict gradient shapes and steady-state times, fitting experimental data to estimate oxygen consumption rates and diffusion constants, improving interpretability (and saving repeated experiments).

10. References