A Microfluidic System for Modeling Endothelial Dysfunction under Combined Physiological Pulsatile Shear Stress and Oscillatory Hyperglycemia

Summary

Here, we present a protocol to fabricate and operate a microfluidic system that exposes endothelial cells to synchronized oscillatory hyperglycemia and pulsatile shear stress. This approach provides a physiologically relevant in vitro model for studying diabetic endothelial dysfunction. This protocol enables quantitative measurement of oxidative stress and endothelial cell responses.

Abstract

Endothelial dysfunction in diabetic vascular complications involves complex interactions between metabolic disturbances, specifically oscillatory glucose (OG) and pulsatile shear stress (PSS). Although these factors have been investigated in isolation, conventional models remain limited to static high-glucose or simplified laminar flow, lacking the capacity to mimic the integrated spatiotemporal coupling found in the diabetic vasculature. By utilizing a programmable microfluidic platform, this protocol provides a valuable platform to investigate the integrated effects of synchronized oscillatory hyperglycemia and physiological PSS on endothelial cells. This protocol aims to model endothelial dysfunction under physiologically relevant coupled metabolic and mechanical conditions. The use of a polydimethylsiloxane (PDMS) chip combined with a pressure-driven control system allows for the precise, independent modulation of flow waveforms and glucose concentration profiles. The hemodynamic fidelity of the system is validated by Micro-Particle Image Velocimetry (Micro-PIV), while the cellular response is quantitatively characterized by monitoring intracellular reactive oxygen species (ROS) levels and cell viability. Representative results demonstrate that physiological PSS effectively attenuates the oxidative injury induced by OG. This effect is demonstrated by reduced intracellular ROS levels and improved cell viability under combined stimulation conditions. Depending on the research question, parameters such as shear stress patterns, glucose oscillation frequencies, and channel geometries can be adjusted. This method serves as a versatile tool for mechanistic studies of mechanobiological pathways and drug screening for therapeutic interventions in diabetic vascular disease.

Introduction

Endothelial dysfunction is a critical early event in the development of diabetic vascular complications, driven by both metabolic disturbances and abnormal hemodynamic forces^1,2. OG levels, rather than sustained hyperglycemia alone, are increasingly recognized as potent inducers of oxidative stress and inflammatory signaling in endothelial cells, leading to vascular injury and impaired homeostasis^3,4,5. Concurrently, hemodynamic shear stress is a major determinant of endothelial phenotype, with PSS under physiological exercise conditions promoting nitric oxide production, antioxidant defense, and vascular protection^6,7,8. Therefore, understanding the integrated effects of oscillatory hyperglycemia and hemodynamic stimulation is essential for elucidating mechanisms of endothelial dysfunction.

Conventional in vitro models, which often rely on static high-glucose conditions or simplified laminar flow, fail to replicate the complex, dynamic interplay between metabolic and mechanical cues present in vivo^9,10,11. While animal models capture systemic aspects of diabetes, they lack the cellular-level resolution needed to dissect signaling under controlled microenvironmental conditions¹². Microfluidic technology offers an effective alternative, providing precise spatiotemporal control of biochemical and biomechanical stimuli in physiologically relevant geometries¹³, and has driven significant recent advancements across various innovative biomedical applications^14,15. For example, Chen et al.¹⁶ designed a microfluidic device with spatial and temporal wall shear stress and ATP signaling for studying endothelial cell intracellular Ca²⁺ dynamics. Yu et al.¹¹ developed a hemodynamic microfluidic chip system for detecting the cellular responses under different combinations of physiological PSS and high glucose concentrations. Although these systems have been used to study endothelial responses to shear stress or single-factor metabolic perturbations, few have successfully integrated both glucose oscillations and PSS within a single system. Therefore, a standardized, reproducible protocol that synchronizes both factors in a controlled microfluidic setting remains lacking.

This method addresses this gap by establishing a reproducible microfluidic system capable of delivering coupled OG and PSS. The primary goal of this approach is to model oxidative stress, impaired cell viability and mechanobiological responses underlying endothelial dysfunction in a physiologically relevant manner. Unlike static culture or conventional perfusion chambers, this approach improves upon existing models by enabling simultaneous and controlled modulation of both metabolic and mechanical stimuli, more closely approximating the diabetic vascular microenvironment. Furthermore, the system is compatible with live-cell imaging, real-time monitoring, and molecular analysis, making it suitable for a wide range of applications including mechanistic studies, disease modeling, and drug screening for endothelial function and injury.

Protocol

Ethics statement:

NOTE: All reagents, devices, and software used in this protocol are listed in the Table of Materials.

1. Microfluidic chip fabrication

1. Use computer-aided design (CAD) software to design the structure of the microfluidic chip (W x L x H = 2 mm x 30 mm x 0.1 mm, Figure 1A) and make the SU-8 master mold through a commercial company.
2. Mix PDMS curing agent and prepolymer at a 10:1 ratio by weight in a clean mixing cup and vigorously stir manually for 5 min.
3. Place the container in a vacuum chamber under -0.08 MPa for a minimum of 15 min to remove air bubbles trapped during the mixing process.
   NOTE: To expedite bubble removal and prevent overflow, temporarily vent the chamber to atmospheric pressure to burst the large bubbles, and then immediately reapply the vacuum. Repeat this venting cycle 2–3 times until the bulk mixture becomes completely transparent and free of macroscopic bubbles.
4. Pour the PDMS onto the wafer placed in a glass Petri dish.
5. Degas the PDMS poured onto the wafer in a desiccator for an additional 15 min to remove any bubbles that might have formed during casting.
   NOTE: If any stubborn micro-bubbles remain trapped near or within the SU-8 master mold structures, gently blow them to the edge or burst them using a manual rubber air blower. Complete removal of air bubbles is crucial to prevent structural defects in the microchannels.
6. Place the petri dish containing mold and polymer in a hot oven set to 80 °C for 2 h.
   CAUTION: Use heat-resistant gloves when handling hot materials
7. Carefully peel the solidified polymer slab from the master mold using a scalpel.
8. Use a core punch tool (e.g., 1.2 mm diameter) to create inlet and outlet ports at the designated channel ends on the polymer slab.
9. Clean the glass coverslip and the patterned face of the polymer slab with adhesive tape.
   PAUSE POINT: The cured and punched PDMS slabs can be stored in a clean container (e.g., a sealed Petri dish) at room temperature for several weeks before plasma treatment and bonding.
10. Treat both the polymer slab and the glass coverslip with a plasma cleaner at 65W power under room air for 60 s.
11. Immediately align the polymer channel face to the treated side of the glass coverslip.
12. Press gently to create a permanent, closed microfluidic channel.

2. Endothelial cell preparation and seeding

1. Place the bonded microfluidic chip in a boronized glass bottle filled with deionized water and sterilize by autoclaving at 121 °C and 15 psi for 20 min.
   NOTE: Submerge the chip in water during autoclaving to equalize pressure and prevent deformation of PDMS microchannels. This also helps remove trapped air bubbles and primes the channels for subsequent fluid introduction.
   PAUSE POINT: The sterilized microfluidic chips can be safely stored in the sealed glass bottle at room temperature for up to several weeks until they are needed for cell seeding.
2. Place the sterilized chip into a 60 mm culture dish filled with a 1:1 mixture of phosphate buffer saline (PBS) and low glucose Dulbecco’s Modified Eagle Medium (DMEM).
3. Introduce 100 µg·mL^-1 fibronectin into the channel using a 1 mL syringe and incubate at 37 °C for 4 h.
   NOTE: Ensure that the microfluidic chip is placed below the liquid level. Connect the adapter hose to the syringe needle, fill the hose with liquid and insert it below the liquid level, connect the inlet and then inject to ensure that no bubbles enter.
4. Add 24.5 mM D-mannitol to low glucose DMEM (5.5mM glucose) to elevate osmotic pressure. Add 5 mM D-glucose to high glucose DMEM (25mM glucose) to a final 30 mM glucose. Make normal glucose (NG) culture medium and high-glucose (HG) culture medium supplemented with 20% fetal bovine serum (FBS), 1% penicillin/streptomycin (P/S), respectively.
5. Warm 1× PBS, trypsin, NG and HG culture media in a water bath at 37 °C. Harvest the Human Umbilical Vein Endothelial Cells (HUVECs) from the culture dish at 90% confluency. Detach the cells using 0.25% Trypsin-EDTA solution and incubate at 37 °C for 30 s. Count the harvested cells and adjust the concentration to 1 x 10⁶ cells/mL in NG medium and HG medium, respectively.
6. Inject 100 µL of the cell suspension into the inlet port of a microfluidic chip using a 1 mL syringe, utilizing NG medium for the control group and HG medium for the experimental group. Place microfluidic chips in a humidified incubator at 37 °C with 5% CO₂ for 48 h.
   NOTE: Allow cells to adhere for at least 4 h before further processing. Manually alternate the two culture media every 2 h to stimulate oscillatory hyperglycemia during the static culture phase. Keep the cells at 5.5 mM glucose overnight to mimic the in vivo situation (Figure 1B).

3. Dynamic stimulation setup

1. Fill two 100 mL reservoirs with NG and HG culture media, respectively, and connect them to the corresponding channels of the programmable pressure control system.
2. Link the reservoirs, the two-position three-way magnetic valve, and the microfluidic chip using silicone tubing.
3. Use 1.3 mm diameter 90° angled stainless-steel connectors to attach the inlet and outlet ports of the PDMS chip to the tubing.
   NOTE: Sterilize the 100 mL reservoirs, silicone tubing and stainless-steel connectors in a high-pressure, high-temperature autoclave beforehand.
4. Calculate the flow rate (Q) required to achieve the target PSS (τ). Use the formula τ = (6µQ)/(wh²), where µ the viscosity, w is the channel width, and h is the channel height.
5. Set a target physiological PSS (τ = 10 ± 2 dyn·cm^-2, 1 Hz), which corresponds to a calculated flow-rate of 200 ± 40 µL·min^-1.
6. Program the programmable air pump to generate a sine wave function to maintain the PSS. Program the valve controller to switch channels every 5 min to generate a square-wave glucose oscillation during the dynamic stimulation.
7. Start the programmable pump system to deliver the combined OG and PSS for a total of 30 min.

4. Micro-PIV for flow field verification

1. Prepare the NG and HG media containing a 0.05% volume concentration of fluorescent microparticles (1 μm). Ensure the particles are evenly dispersed and free of aggregation.
2. Place a clean, cell-free microfluidic chip (identical to the one used in the cell experiments) onto the stage of the inverted fluorescence microscope.
3. Connect all components in the system (Figure 1C).
   1. Connect the pressure source to the pressure controller.
   2. Connect the pressure controller to the valve controller, which is linked to the computer for automated control of pressure waveforms.
   3. Connect the pressure controller outlets to the HG and NG reservoirs via separate tubing lines.
   4. Connect both reservoirs to the 3/2 pressure valve, which switches between HG and NG media according to the programmed oscillatory glucose protocol.
   5. Connect the outlet of the 3/2 pressure valve to the inlet of the microfluidic chip via tubing.
   6. Connect the outlet of the microfluidic chip to the waste reservoir.
   7. Place the microfluidic chip on the microscope stage for live-cell imaging throughout the experiment.
      NOTE: Ensure all tubing connections are secure and free of leaks before initiating flow.
4. Trigger the high-speed camera (operating at 15 Hz in double-frame mode) after the flow is steady to capture sequential image pairs of the flowing microparticles. Set the precise time delay (Δt) between the two frames of each pair to 80 μs.
5. Capture images for Region of Interest (ROI) within the channel. Select the ROI at the center of the straight channel section (Figure 2A), equidistant from both the inlet and outlet, to avoid end effects and ensure representative flow measurements.
6. Process the acquired image pairs using PIV analysis software. Use an adaptive multi-pass cross-correlation algorithm, starting with an initial interrogation area of 64 × 64 pixels and refining to a final area of 16 × 16 pixels with an 8 × 8 pixel grid step size, to determine the displacement vector (Δx) of the particles between the two frames.
7. Calculate the velocity vector (v) for each interrogation area using the formula v = Δx / Δt.

5. Data acquisition

NOTE: Quantify cell viability and function using the Calcein-AM/propidium iodide (PI) Kit and live-cell oxidative stress fluorescent probe to verify the damage caused by glucose fluctuations and evaluate the effect of co-stimulation.

1. Cell viability
   1. Prepare the Calcein-AM/PI working solution according to the manufacturer’s instructions at the experimental endpoint.
      CAUTION: PI is a known mutagen and a suspected carcinogen that intercalates into DNA. Handle with extreme care. Wear gloves and handle within a designated safe workspace.
   2. Rinse the channel one time with PBS and introduce the Calcein-AM/PI working solution into the channel. Incubate the device at 37 °C with 5% CO₂ for 30 min in the dark.
   3. Immediately capture fluorescent images of the stained cells using an inverted fluorescence microscope equipped with a 10× objective lens. Use the FITC (Green) filter for Calcein-AM and the TRITC or Rhodamine (Red) filter for PI.
2. Measure Intracellular ROS
   1. Prepare the live-cell oxidative stress fluorescent probe (see Table of Materials) working solution at the concentration specified by the manufacturer (5 μM) in PBS.
      CAUTION: Fluorescent probes and their common stock solvents (such as DMSO) can be toxic and may facilitate the absorption of hazardous substances through the skin. Wear gloves and handle within a designated safe workspace.
   2. Flush the channel three times with PBS and introduce the live-cell oxidative stress fluorescent probe solution into channel. Incubate the chip at 37 °C with 5% CO2 for 30 min in the dark.
   3. Flush the channel three times with PBS to remove the unbound probe. Immediately capture fluorescent images using the Far-Red filter set (excitation 640 nm, emission 665 nm) through a 20× objective lens, maintaining the exact same exposure parameters for all groups.
      NOTE: Increased fluorescence intensity indicates higher ROS levels.

6. Data analysis

PAUSE POINT: Once all raw fluorescent images and high-speed camera images are acquired and securely saved to a hard drive, the subsequent image processing, quantitative analysis, and statistical comparisons can be paused and performed at any later time.

1. Quantify cell viability
   1. Open the acquired Calcein-AM (green) and PI (red) images in ImageJ (see Table of Materials). Manually adjust the threshold for each channel to distinguish fluorescent signals from background. Use the cell counter function to count fluorescent cells, adjusting the minimum cell size according to the observed cell size, and count the number of live (N_Live) and dead (N_Dead) cells in multiple fields of view.
      NOTE: N_Live refers to the number of Calcein-AM-positive (green) cells, N_Dead refers to the number of PI-positive (red) cells.
   2. Calculate the percentage of live cells (N_Live / (N_Live + N_Dead) x 100%) for each experimental condition using spreadsheet software. Calculate the mean and standard deviation across multiple fields of view using statistical analysis software for comparison between groups.
2. Quantify intracellular ROS
   1. Drag image of choice in ImageJ. Set the scale by calibrating the pixel size using the microscope's metadata (or a known reference grid): Analyze | Set Scale.
   2. Remove a non-specific uniform background signal from the entire image using the background subtraction function: Process | Subtract Background.
   3. Manually adjust the threshold slider to segment and define the intracellular ROS signal: Image | Adjust | Threshold. Once the threshold is set, use Analyze | Set Measurements to choose Area and Mean gray value. Set the minimum cell size according to the observed cell size and measure the mean intensity of the segmented signal within the cell monolayer ROI: Analyze | Analyze Particles.
   4. Report the ROS level for each experimental condition as the means of the measured values (e.g., mean intensity) derived from at least three randomly selected fields of view per chip.
3. Performing statistical comparisons
   1. Input the quantitative results (viability percentage and ROS mean intensity) into data analysis and graphing software (see Table of Materials).
   2. Determine statistical significance by comparing the experimental groups with the control group using two-way Analysis of Variance (ANOVA) followed by Tukey’s post-hoc test, with a significance level of p = 0.05. Use the software to perform the analysis and generate plots with error bars representing standard deviation (SD).

Results

Micro-PIV validation of flow field within the microfluidic channel
Figure 1D provides an overview of the microfluidic platform used to investigate endothelial responses to coupled metabolic and mechanical stimulation. The system integrates a straight-channel PDMS microfluidic chip with a pressure-driven flow control module, enabling the application of OG condition together with physiological PSS. A programmable valve-switching strategy was employed to generate periodic a...

Discussion

The microfluidic system described herein provides a robust and physiologically relevant platform to investigate the integrated effects of OG and PSS on endothelial cell function. The primary advantage of this method lies in its capability to precisely control and decouple mechanical and biochemical stimuli within a defined microenvironment. Unlike conventional static cultures that lack hemodynamic forces or animal models that often obscure cellular level signaling^4,17<...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
0.25% Trypsin-EDTA	NEST Biotechnology Co., Ltd.	211052	Cell dissociation
1 mL syringe	Shanghai Zhengbang Medical Science Co., Ltd.	1mL	Inject fluid
60 mm culture dish	NEST Biotechnology Co., Ltd.	705001	Cell culture
90° angled stainless-steel connector	Luoyang Mayer Trading Co.,Ltd.	1.3 mm	Adapter sealing device
AutoCAD software	Autodesk	RRID:SCR_014981	Computer-aided design software
Boronized glass bottle	Hunan Aidete Scientific Instruments Co., Ltd.	Double-pass 100mL	reservoir
Calcein-AM/PI Kit	Beyotime	byt-c2015	Staining
Cell incubator	ESCO	Esco CelMate	Cell culture
CellROX Deep Red	Invitrogen	C10422	Staining
D-glucose	Sigma Aldrich	47249	Glucose
D-mannitol	Macklin	M813423	Balance osmotic pressure
DynamicStudio	Dantec Dynamics	V8.6	PIV analysis software
Fetal bovine serum (FBS)	Gibco	14190-094	Medium
Fibronectin	Sigma Aldrich	F-2006	Cell adhesion
Fluorescent microparticles	Thermo Scientific	G0100	Micro-PIV
Glass coverslip	Citotest Scientific Co.,Ltd.	#1	Microfluidic chip
High glucose DMEM	Solarbio	12100	Medium
High-speed camera	Dantec Dynamics	FlowSense EO	Micro-PIV
ImageJ	NIH	RRID:SCR_003070	Image processing software
Inverted fluorescence microscope	Olympus	IX73	Cell viability
Inverted fluorescence microscope	Olympus	IX83	Observation and acquisition
Low glucose DMEM	Solarbio	31600	Medium
Origin	OriginLab	RRID:SCR_014212	Data analysis and graphing software
PDMS	Dow lnc.	184	Microfluidic chip
Penicillin/streptomycin (P/S)	NEST Biotechnology Co., Ltd.	211092	Medium
Phosphate buffer saline (PBS)	NEST Biotechnology Co., Ltd.	211031	Cell culture
Plasma cleaner	Jiarun Wanfeng Technology Co., Ltd.	PC-6D	Microfluidic chip
Programmable pressure control system	Elvesys	OB1 MK4	Flow control
Silicone tube	Nanjing Runze Fluid Control Equipment Co., Ltd.	96410	1mm*3mm
SU-8 master mold	Boao Biology Group Co. Ltd.	N/A	Microfluidic chip
Valve controller	Elvesys	MUX Wire V3	Flow control