Hardware: Patient-Derived Microfluidic Colorectal Organoid-on-a-Chip Design

1. OVERVIEW

To simulate the complex intestinal physiology and multi-layered fluid environment of the human colon, we designed a three-layer microfluidic organoid-on-a-chip device, primarily fabricated from polydimethylsiloxane (PDMS). This chip analyzes the polarization of macrophages induced by engineered bacteria and the adhesion behavior of biofilm formation within a tumor-mimetic colonic lumen, as illustrated in Fig. 1. The device co-cultures human colonic epithelium, vascular endothelium, macrophages, and cancerous colonic cells in a unified three-dimensional microenvironment, reconstructing the vascular-intestinal wall-lumen tri-layer liquid phase and its expansion-contraction peristalsis. Consequently, the chip achieves a comprehensive, patient-specific simulation of physicochemical and biological conditions, enabling verification of the feasibility of therapeutic mechanisms[1].

Fig. 1 Schematic of the microfluidic organoid-on-a-chip structure

The main body of the microfluidic chip is made of PDMS, a transparent, silicone-based cross-linked polymer with a modulus similar to human tissue, facilitating real-time visualization of full physiological processes and ensuring the feasibility of live fluorescence microscopy. The chip system applies periodic vacuum driving to mimic physiological peristalsis. The top layer enables real-time vascular drug perfusion, the middle layer allows real-time bacterial biofilm induction, and the bottom layer continuously collects effluent for biomarker and cellular analyses, quantifying macrophage polarization markers and biofilm adhesion kinetics. Benefiting from the excellent optical transparency of PDMS, fluorescent labeling can be applied in real time to verify the effects of biofilm formation on bacterial adhesion and encapsulation. This platform eliminates the need for animal models, reduces sample consumption, and faithfully recapitulates patient-specific, multi-physical–biochemical coupled intestinal scenarios, providing a high spatiotemporal resolution, repeatable, and animal-ethics-friendly in vitro verification platform.

As of the Wiki upload, the microfluidic chip structural design has been completed, as shown below:

Fig. 2 Schematic of the microfluidic chip structure

The chip structure was drawn with SolidWorks® 2018 (Fig. 2). Overall size: 40 mm × 90 mm × 6 mm, consisting of three layers—top flow plate, middle aluminum foil, and bottom base plate. Top and bottom layers are 7 mm thick; the middle layer is 3.5 mm thick. Liquids enter each layer through three 1-mm-diameter inlet holes, converge into a 2 mm × 35 mm × 2.5 mm main channel via 1.6-mm-wide feeder channels, widen mid-channel to 4 mm × 11 mm × 2.5 mm, then run through identical outlet channels into a single 40.5 µL reservoir capped by a 1-mm-diameter hole for effluent collection. Adjacent layers are separated by PET semi-permeable membranes. On each side of the main channel, a 3.5 mm × 32 mm × 9.1 mm vacuum chamber is machined, with a 2.5-mm-diameter port on top for connection to a vacuum pump.

2. MICROFLUIDIC CHIP FABRICATION

2.1 PDMS Mold Fabrication and Casting

The PDMS substrate used as the microfluidic chip body was prepared by casting against a 3D-printed mold. After 3D printing, the mold was rinsed with IPA for 30 s to remove surface particles. Upon completion of printing, an acrylic clear coat (~3 µm) was sprayed first to seal surface micropores; a PTFE dry-film release agent (~1 µm) was then sprayed. Both steps are completed at room temperature without subsequent curing. PDMS (Sylgard 184) was mixed at a 10:1 ratio, manually stirred, and centrifugally degassed for 2 min. The mixture was poured into the mold cavity, vacuum-degassed, and then cured at 80 °C for 1.5 h. After natural cooling, the mold was inverted for demolding. Each microfluidic chip contains A, B, and C three-layer structures; the same process is repeated three times for each chip, completing the preparation of the three-layer PDMS chip substrates, as shown in Fig. 3.

Fig. 3 Schematic of the three-layer PDMS chip substrate，and semipermeable membrane

2.2 Assembly of PDMS-Based Microfluidic Chip

After the three-layer structures are completed, the microfluidic chip is assembled in sequence:

1. The top, middle, and bottom PDMS layers are each subjected to 60 s plasma treatment;
2. A stretchable porous membrane is used: membrane 1 between layers A and B is seeded with endothelial cells on the upper surface and colonic epithelial cells on the lower surface, simulating the human colonic-capillary interface; membrane 2 between layers B and C is a PET porous semi-permeable membrane that allows only molecular exchange, on which patient-derived 2D colonic organoids are grown, representing normal and cancerous colonic tissues from the same patient.
3. The layers are aligned immediately after treatment, with micro-channels and chambers perpendicular to the long axis;
4. After alignment, the layers are pressed together and checked under an optical microscope for alignment. Thus, the complete chip assembly is finished[2].

Fig. 4 Schematic of the assembly of PDMS-based microfluidic chip

3. SCHEMATIC OF ORGANOID

3.1 3D Construction and Culture of Patient-Derived Colorectal Organoids

To construct colonic organoid models, cancerous and adjacent non-cancerous area tissues were first resected from the same CRC patient surgical specimen, with healthy controls. The tissues were then placed in 5 mM EDTA/PBS at 4 °C for 30–45 min incubation, followed by mechanical shaking and 70 µm filtration to enrich Lgr5⁺ crypt base cells[3][4].

• For normal colonic organoids, 200–500 crypts were mixed with 30–50 µL Matrigel or Cultrex, dropped into a 24-well plate, and solidified at 37 °C for 10 min to form 3D structures. Finally, the culture medium was supplemented with Advanced DMEM/F12, 1× B27/N2, 10 mM HEPES, 1 mM N-acetylcysteine, 10 mM nicotinamide, 100 ng/mL Wnt3a, 500 ng/mL R-spondin-1, 50 ng/mL EGF, 100 ng/mL Noggin, 500 nM A83-01, 10 µM SB202190 to maintain stem cell characteristics and initiate expansion.
• For colonic cancer organoids, the above medium without Wnt3A (all other factors unchanged) was used for selective expansion under 1 % O₂ hypoxia; whole-exome sequencing was performed every 3–5 passages to verify retention of APC, KRAS, and TP53 mutations[5].

3.2 2D Conversion, Induction of Differentiation, and Seeding of Patient-Derived Colorectal Organoids

To obtain polarized and differentiated colonic monolayers, when 3D spheroids reached 300 µm in diameter, cancerous/non-cancerous area-derived colonic organoids were digested with 0.25 % Trypsin-EDTA at 37 °C for 5 min and gently pipetted to a single-cell suspension. Cells were seeded at 1×10⁵ cells cm⁻² on a Type-IV collagen-coated 0.4 µm PET membrane and cultured to 100 % confluence in complete stem-cell medium containing Wnt3a, R-spondin-1, and EGF. The apical medium was then removed to establish an air–liquid interface, and differentiation was induced by withdrawing Wnt3a/R-spondin-1 and supplementing BMP-4 (10 ng mL⁻¹) and IL-22 (20 ng mL⁻¹) for 5–7 days. After differentiation, TEER ≥ 500 Ω cm² and 4 h FITC-dextran paracellular leakage < 5 % were verified to confirm barrier integrity. Qualified membranes were encapsulated into the PDMS chip prepared in Section 2.2, with sides B and C continuously perfused with differentiation medium at 30 µL h⁻¹; after 24 h equilibration, compounds, bacterial suspensions, or inflammatory stimuli were added to side A, and effluent was collected at regular intervals for downstream functional analyses[6].

4. FUNCTIONAL STIMULATION

4.1 Simulation of Intestinal Behavior Using Microfluidic Organoid-on-a-Chip

After the patient-derived colorectal organoid-on-a-chip was assembled, to simulate the complex multi-layer liquid-phase biochemical environment and characteristic biomechanical behavior of the intestine, the chip operated as follows:

• Layer 1 simulates blood flow, driven by a peristaltic pump to mimic pulse rhythm; continuous perfusion of inducer-containing medium simulates intravenous drug injection.
• A PET semi-permeable membrane between layers 1 and 2 is seeded with endothelial cells/colonic epithelial cells on opposite surfaces, constructing the colonic-capillary interface.
• Layer 2 is used to culture normal or cancerous 2D organoids and is supplied with engineered bacteria and medium to reconstruct the tissue/tumor microenvironment (see Sections 3.1 and 3.2 for organoid construction).
• A PET semi-permeable membrane between layers 2 and 3 acts as a 2D organoid scaffold while enabling solid-liquid separation of small-molecule liquids and cells.
• Layer 3 is perfused with standard medium containing macrophages, enabling polarization detection without tissue/immune cell separation.
• Additionally, strain channels on both sides of the flow path are connected to a vacuum pump; periodic pressure drop-recovery cycles stretch the intermediate flow path and attached cell layers/organoids, achieving dynamic adhesion simulation under colonic peristalsis[7][8].

Fig. 5 Cross section of the organoid chamber

4.2 Macrophage Polarization Detection Based on Organoid-on-a-Chip

After the microfluidic organoid-on-a-chip was completed, to verify whether the established biofilm-induced polarization model was correct, quantitative detection of biofilm adhesion and macrophage polarization was performed. Layers A and B were operated as described in Section 4.1, accompanied by rhythmic vacuum cavity decompression-contraction and pressure recovery. Layer 3 was perfused with standard medium containing unpolarized macrophages, and effluent was collected at the outlet. Flow cytometry and cytokine ELISA analysis of the collected effluent (containing M1/M2 macrophages) were performed to complete quantitative detection of macrophage polarization. Due to the size-selectivity of the PET semi-permeable membrane between layers 2 and 3, layer 3 maintains a physicochemical microenvironment similar to layer 2, yet the collected effluent is free of interference from cancerous/non-cancerous colorectal organoid tissue cells. A single filtration step yields purified macrophages, simplifying the workflow and significantly reducing interference from colorectal tissue cells in flow cytometry, improving accuracy and reliability of results[9].

4.3 Biofilm Formation and Bacterial Adhesion Detection Based on Organoid-on-a-Chip

To quantitatively analyze the adhesion of engineered E. coli to human colorectal organoids, a fluorescence labeling method was employed: the sfGFP green fluorescent protein gene was inserted into the bacterial chromosome via the mini-Tn7 plasmid system (pTNS3-Tn7-sfGFP), and bacteria were cultured overnight in LB medium containing 0.2 % arabinose to ensure uniform fluorophore expression. Day-7 ALI-differentiated organoid monolayers were gently rinsed twice with PBS and placed in antibiotic-free, phenol-red-free Advanced DMEM/F12 medium. Collected bacteria were washed and resuspended to OD₆₀₀ = 0.1 (~1×10⁷ CFU/mL), added to the Transwell apical side, and incubated at 37 ℃, 5 % CO₂for 2 h. After incubation, non-adherent bacteria were removed by gentle washing three times with PBS (200 µL, 50 rpm, 30 s). Adherent bacteria were fixed with 4 % PFA for 15 min, rinsed, and counter-stained with 1 µg/mL DAPI for 10 min[10].

Imaging was performed on an inverted laser confocal microscope (Zeiss LSM 880, 63×/1.4 NA oil objective) using 488 nm (GFP) and 405 nm (DAPI) excitation; z-stacks were acquired at 0.5 µm steps to cover the entire monolayer thickness. Maximum-intensity projections were imported into ImageJ, and adhesion density (bacteria/mm²) and adhesion index (GFP area/DAPI area) were calculated from five random fields per membrane using Otsu automatic thresholding and particle analysis, normalized to the corresponding inoculum CFU, yielding the final quantitative result of adherent CFU/mm².

References

[1]C. M. Sakolish, M. B. Esch, J. J. Hickman, M. L. Shuler, and G. J. Mahler, "Modeling barrier tissues in vitro: Methods, achievements, and challenges," EBioMedicine, vol. 5, pp. 30-39, 2016, doi: 10.1016/j.ebiom.2016.02.023.

[2] L. Mou, B. Hu, J. Zhang, and X. Jiang, "A hinge-based aligner for fast, large-scale assembly of microfluidic chips," Biomedical Microdevices, vol. 21, p. 69, 2019, doi: 10.1007/s10544-019-0404-y.

[3] M. van de Wetering et al., "Prospective derivation of a Living Organoid Biobank of colorectal cancer patients," Cell, vol. 161, no. 4, pp. 933-945, 2015, doi: 10.1016/j.cell.2015.03.053.

[4] A. Barbáchano et al., "Organoids and colorectal cancer," Cancers, vol. 13, no. 11, p. 2657, 2021, doi: 10.3390/cancers13112657.

[5] C. Tian et al., "Stem cell-derived intestinal organoids: a novel modality for IBD," Cell Death Discovery, vol. 9, no. 255, 2023, doi: 10.1038/s41420-023-01556-1.

[6] B. A. Ojo, K. L. VanDussen, and M. J. Rosen, "The promise of patient-derived colon organoids to model ulcerative colitis," Inflammatory Bowel Diseases, vol. 28, no. 2, pp. 299-308, 2021, doi: 10.1093/ibd/izab161.

[7] A. Bein et al., "Enteric coronavirus infection and treatment modeled with an immunocompetent human intestine-on-a-chip," Frontiers in Pharmacology, vol. 12, p. 718484, 2021, doi: 10.3389/fphar.2021.718484.

[8] M. E. K. Bein et al., "Microfluidic organ-on-a-chip models of human intestine," Cellular and Molecular Gastroenterology and Hepatology, vol. 5, no. 4, pp. 659-668, 2018, doi: 10.1016/j.jcmgh.2017.12.010.

[9] S. Yuan, H. Yuan, D. C. Hay, H. Hu, and C. Wang, "Revolutionizing drug discovery: The impact of distinct designs and biosensor integration in microfluidics-based organ-on-a-chip technology," Biosensors, vol. 14, no. 9, p. 425, 2024, doi: 10.3390/bios14090425.

[10] B. Hu et al., "An automated and portable microfluidic chemiluminescence immunoassay for quantitative detection of biomarkers," Lab Chip, vol. 17, pp. 2225-2234, 2017, doi: 10.1039/c7lc00249a.