Immunocompetent Intestine-on-Chip Model for Analyzing Gut Mucosal Immune Responses

Organ-on-Chip (OoC) systems represent an emerging technique of 3D cell culture that is capable of bridging the gap between conventional 2D cell culture and animal models. OoC platforms typically consist of one or more compartments containing tissue-specific cells grown on a wide range of scaffolds such as membranes or hydrogels¹. The models are capable of mimicking one or more defined organotypic functions. Pumps enable continuous microfluidic perfusion of cell culture medium for removal of cellular waste products, supply with nutrition and growth factors for improved cellular differentiation, and recreating essential in vivo conditions. With the integration of immune cells, OoC systems can mimic human immune response in vitro². To date, a wide range of organs and functional units have been presented¹. These systems include models of the vasculature³, lung⁴, liver^2,5, and intestine⁶ that can be facilitated for drug testing^5,7 and infection studies^6,8.

We here present a human intestine-on-chip model integrating human epithelial cells forming an organotypic 3D topography of villus-like and crypt-like structures combined with an endothelial lining and tissue-resident macrophages. The model is cultured in a microfluidically perfused biochip in the format of a microscopic slide. Each biochip consists of two separate microfluidic cavities. Each cavity is divided by a porous polyethylene terephthalate (PET) membrane into an upper and lower chamber. The membrane itself also serves as the scaffold for the cells to grow on each side. The pores of the membrane enable cellular crosstalk and cell migration between cell layers. Each chamber can be accessed by two female luer lock-sized ports. Optionally, an additional mini-luer lock-sized port can provide access to the upper or lower chamber (Figure 1).

The OoC platform offers a number of readouts that can be obtained from a single experiment. The intestine-on-chip is tailored towards combining perfused 3D cell culture, effluent analysis, and fluorescence microscopy to assess cell marker expression, metabolization rates, immune response, microbial colonization and infection, and barrier function^3,6,8. The model includes tissue-resident immune cells and direct contact of living microorganisms with the host tissue, which is a benefit compared to other published models⁹. Further, epithelial cells self-organize into three-dimensional structures that provide a physiologically relevant interface for the colonization with a living microbiota⁶.

This protocol requires access to ~20 mL of fresh blood per biochip from healthy donors to isolate primary human monocytes. All donors gave written, informed consent to participate in this study, which was approved by the ethics committee of the University Hospital Jena (permission number 2018-1052-BO). For details about the materials, refer to the Table of Materials. For details about the composition of all solutions and media, refer to Table 1.

1. General biochip handling remarks

1. Carefully separate a strip of reservoirs and detach the lids by using a heated knife to obtain single reservoirs and lids. Widen the hole of the lid so the silicon tubing fits snugly.
2. The silicone tubing has an inner diameter of 0.5 mm, is asymmetrical, and is separated into longer (20 cm) and shorter (12 cm) sides by two peristaltic pump stoppers. Assemble two tubes of each symmetry per biochip by attachment of a tube to a male luer lock-connecter and the lid on the opposing side of the tube. Also assemble four reservoirs per biochip.
   NOTE: Prepare tubing and reservoirs in advance and sterilize by autoclaving before use. As the silicone tubing has a limited lifetime, exchange the tubing after 3-5 experiments. For certain research interests, such as drug testing, it is advised to prepare new tubing for each experiment. Many steps of this protocol run in parallel; refer to the overview figure, which highlights the different steps performed on one day as shown in Figure 2.

Figure 1: Schematic representation of intestine-on-chip model. (A) The biochip is presented in a cross-sectional view. (B) The dimension of the whole biochip as well as of the flat, removable PET membrane is visible. The total volume of the upper chamber including the female luer lock-sized ports is 290 µL and 270 µL for the lower chamber respectively. (C) A schematic composition of the intestinal-biochip, the three-dimensional outgrowth epithelium resembling villus-like and crypt-like structures including differentiated immune cells and a mucus layer can be seen. The other side of the PET membrane is covered by an endothelial monolayer. Please click here to view a larger version of this figure.

Figure 2: Schematic overview of model buildup timeline and experimental setup. This figure shows the schematic overview of the presented protocol. Important procedures, such as the seeding of cells and the epithelial challenge with LPS are indicated by arrows. Abbreviations: HUVECs = human umbilical venous endothelial cells; LPS = lipopolysaccharide. Please click here to view a larger version of this figure.

2. Biochip sterilization

1. Fill a sterile glass Petri dish, possessing a diameter of 15 cm, with 70% undenatured ethanol. Place the biochip inside so that all ports of the biochip are completely covered by the ethanol solution.
2. Two times per port, pull 1 mL of 70% ethanol through all chambers of the chip. Incubate for 45-60 min at room temperature (RT).
   NOTE: Ensure no air is trapped inside the biochip. From this point forward, no air should enter the biochip system, and the cavities should remain filled with liquid.
3. Remove the ethanol in the Petri dish and replace it with sterile double-distilled water (ddH₂O) until all ports are fully covered. Again, two times per port, draw 1 mL of ddH₂O through the biochip cavity. Refresh the ddH₂O in the Petri dish and repeat the procedure.
4. Remove all liquid from the Petri dish. Hereafter, keep the fully sterilized biochips inside the closed Petri dish whenever they are outside a sterile environment. Add a small reservoir (e.g., the lid of a 50 mL tube) of 2-5 mL ddH₂O to the Petri dish to reduce evaporation of liquid inside the biochip.
   NOTE: Biochips can be sterilized up to 3 days in advance if they are kept in a sterile environment until use. This allows for flexibility in the workload in a single day.

3. HUVECs harvest and seeding

NOTE: The human umbilical venous endothelial cells (HUVECs) were isolated from umbilical cords as published before¹⁰.

1. Before seeding the HUVECs, coat the membrane with human collagen IV. For this, prepare a 1:100 dilution of a collagen stock solution (Table 1) in Dulbecco's phosphate-buffered saline containing magnesium and calcium (PBS +/+). Add 350 µL of the diluted stock solution to the respective chamber. Incubate for 5 min at RT.
   NOTE: If handling the biochip in the sterile hood, we recommend placing a sterile tissue underneath it to collect excessive medium.
2. Flush all chambers twice with 350 µL of PBS +/+ to wash out the remaining collagen and acetic acid. Then, add 350 µL of endothelial cell growth (EC)-medium to each chamber.
   NOTE: From here, biochips are ready for cell seeding and can be stored at 37 °C until use.
3. Use HUVECs at passages 1-3 at 80-90% cell confluency. Cultivate HUVECs in EC-medium containing a defined supplement mix provided by the manufacturer. A representative brightfield image of a HUVEC cell culture is presented in Figure 3A.
   NOTE: Depending on the donor, HUVECs of higher passages can start to dedifferentiate and cannot reliably form a dense and confluent monolayer inside the biochip. Use of antibiotics, i.e.,100 U/mL penicillin and 100 µg/mL streptomycin, is optional but recommended as a supplement to the EC-medium to prevent microbial contamination.
4. Remove the cell culture medium from a T25 cell culture flask and wash the cells gently with 3-5 mL of Dulbecco's phosphate-buffered saline without magnesium and calcium (PBS -/-). Remove the PBS -/- and add 1 mL of trypsin dissociation reagent (Table 1). Incubate for 5 min at 37 °C until the cells detach from the cell culture flask.
5. Transfer the detached cells into a tube using 9 mL of 5% fetal bovine serum (FBS) in PBS -/-. Centrifuge at 350 × g for 5 min at RT. Remove the supernatant, resuspend in 1 mL of EC-medium, and determine the number of cells. Adjust the cell concentration to 0.4 × 10⁶ cells per 150 µL (seeding in lower chamber) or per 250 µL (seeding in upper chamber).
6. Add the respective volume of cells to the chamber. If seeding in the lower chamber, close all ports and immediately position the biochip upside-down so the cells fall onto the PET membrane. Incubate the biochips in a humidified incubator at 37 °C and 5% CO₂.
7. Perform a medium exchange of the HUVEC-containing chamber with 350 µL of EC-medium after 24 h. The medium in the opposing chamber does not need to be replaced.

4. Human serum collection and peripheral blood mononuclear cell (PBMC)-derived monocyte isolation

NOTE: The PBMCs were isolated as described in Mosig et al.¹¹.

1. Withdraw human venous blood from healthy donors. Per biochip, procure a minimum of 10 mL of whole blood in silicate-containing blood collection tubes for serum collection. After full coagulation, centrifuge the blood collection tubes at 2,500 × g for 10 min at RT. Collect the serum, aliquot, and store at -20 °C until further use.
2. Procure a minimum of 10 mL of whole blood from the same donor in EDTA-containing blood collection tubes for PBMC isolation. Gently mix the non-coagulated blood 1:1 with iso-buffer (Table 1) by inversion and slowly layer 35 mL of this mixture atop 15 mL of a density gradient medium with a density of 1.077 g/mL in a 50 mL tube.
3. Centrifuge at 800 × g for 20 min without brake at RT. Carefully withdraw the resulting immune cell layer, appearing on top of the density gradient medium, and transfer it into a new 50 mL tube. Fill up to 50 mL with cold iso-buffer and wash the cells by gentle inversion.
4. Centrifuge at 200 × g for 8 min without brake at 4 °C. Discard the supernatant and resuspend in 10 mL of iso-buffer per density-gradient. Optional: Pool PBMCs of one donor if several gradients are run in parallel.
5. Centrifuge at 150 × g for 8 min at 4 °C. Discard the supernatant and resuspend the pellet in 10 mL of iso-buffer per density gradient. Repeat centrifugation step 4.4. Finally, discard the supernatant and resuspend the cells in 2 mL of monocyte differentiation medium (Table 1).
   NOTE: The addition of M-CSF and GM-CSF enforces the differentiation of the isolated monocytes towards monocyte-derived macrophages and monocyte-derived dendritic cells (in combination with lipopolysaccharide [LPS], which is added at a later point of this protocol). Use of antibiotics, i.e., 100 U/mL penicillin and 100 µg/mL streptomycin, is optional but recommended as a supplement of the medium to prevent microbial contamination.
6. Determine the cell number and seed ~10 × 10⁶ cells per well of a 6-well plate in 2 mL of monocyte differentiation medium (Table 1). Incubate in a humidified incubator at 37 °C for 1 h to allow attachment of monocytes to the plastic of the 6-well plate.
7. Carefully discard the supernatant and wash 2x with prewarmed 2 mL of hematopoietic cell medium to remove the unbound cells. Incubate at 37 °C for another 24 h in monocyte differentiation medium.
8. To harvest the monocytes, carefully discard the supernatant and wash once with 2 mL of prewarmed PBS -/-. See Figure 3B for an example of a brightfield image of the monocyte culture at this point. Then, incubate the cells for 7 min in 1 mL of prewarmed monocyte detachment reagent (Table 1) at 37 °C to enforce detachment of the monocytes from the plastic of the 6-well plate.
9. Transfer the detached monocytes to a low-binding tube. Optional: to achieve a higher cell yield, carefully wash the 6-well plate several times with PBS -/-.
10. Centrifuge at 300 × g for 8 min at RT. Discard the supernatant and resuspend in EC-conditioned medium (Table 1). Determine the cell number and adjust the cell concentration to 0.1 × 10⁶ cells per 150 µL (seeding in the lower cavity) or per 250 µL (seeding in the upper cavity).
    NOTE: Be gentle at all steps of the immune cell isolation and reduce shear forces to prevent immune cell activation. When establishing this isolation, check for the purity of the PBMC-derived monocytes (e.g., via flow cytometry). More than 95% of all cells should be positive for typical monocyte markers such as CD14.

5. Monocyte seeding

1. Perform a medium exchange in the HUVEC-containing chamber with 350 µL of prewarmed EC-conditioned medium.
2. Add 150 µL (lower chamber) or 250 µL (upper chamber) of the prepared monocyte suspension (see step 4.10) to the same chamber. If seeding into the lower chamber, close all ports and immediately position the biochip upside-down for the cells to fall onto the HUVEC layer. Incubate the biochip in a humidified incubator at 37 °C and 5% CO₂.
3. Perform a medium exchange in the HUVEC + monocyte-containing chamber with 350 µL of EC-conditioned medium every 24 h.

6. C2BBe1 harvest and seeding

NOTE: Caco-2 brush border expressing cells 1 (C2BBe1)¹² are used up to passage 35 and are taken from flasks of 80-90% confluency. A representative brightfield image of a C2BBe1 culture is presented in Figure 3C.

1. Cultivate C2BBe1 cells in C2-medium (Table 1).
   NOTE: Use of antibiotics, i.e., 20 µg/mL gentamicin, is optional but recommended as a supplement of the C2-medium to prevent microbial contamination.
2. Remove the cell culture medium of a T25 cell culture flask and wash the cells gently with 3-5 mL of PBS -/-. Remove the PBS -/- and add 1 mL of trypsin dissociation reagent (Table 1). Incubate for 5 min at 37 °C until the cells detach from the cell culture flask.
3. Transfer the detached cells to a tube by using 9 mL of 5% fetal bovine serum (FBS) in PBS -/-. Centrifuge at 350 × g for 5 min at RT. Remove the supernatant, resuspend in 1 mL of C2-medium, and determine the number of cells. Adjust the cell concentration to 0.5 × 10⁶ cells per 150 µL (seeding in lower chamber) or per 250 µL (seeding in upper chamber).
4. Before seeding of the C2BBe1, gently wash the respective chamber with 350 µL of C2-medium.
5. Add 150 µL (lower chamber) or 250 µL (upper chamber) of the prepared C2BBe1-suspension (see step 6.3) to the respective chamber. If seeding into the lower chamber, close all ports and immediately position the biochip upside-down for the cells to fall onto the PET membrane. Incubate the biochips in a humidified incubator at 37 °C and 5% CO₂.

Figure 3: Cell morphology of HUVECs, monocytes, and C2BBe1 before seeding in the biochip. This figure shows representative brightfield images of the different cell sources used throughout the protocol. Images were taken with a reverse brightfield microscope using 10x magnification. All cell types, (A) HUVECs, (B) monocytes, and (C) C2BBe1 were cultivated in 2D mono-layer cell culture as described in their specific protocol sections. Scale bars = 200 µm. Abbreviations: HUVECs = human umbilical venous endothelial cells; PBMCs = peripheral blood mononuclear cells. Please click here to view a larger version of this figure.

7. Connection to peristaltic pump and circular perfusion

1. Prepare an empty incubator with the addition of a peristaltic pump. Thoroughly clean all areas of the incubator and pump with disinfectant to provide a quasi-sterile environment.
   NOTE: Peristaltic pumps can produce a lot of heat while working. In well-insulated incubators or poorly air-conditioned laboratories, the number of usable pumps per incubator might be limited as incubators tend to overheat. Two peristaltic pumps per incubator should be adequate.
2. Before attaching the sterilized tubes to the biochip, flush each tubing with 700 µL of PBS +/+ followed by 500 µL of C2-medium or EC-conditioned medium. Prepare one tubing of each symmetry per medium (see step 1.2). Use the tubing with the short distance from the luer lock to the peristaltic pump stopper for the left cavity and the tubing with the other symmetry for the right cavity.
3. Retrieve the biochip from the incubator and perform a medium exchange with 350 µL for each chamber. Remove all plugs and fill all ports to the very top.
4. Starting at the left cavity, connect the first tubing to the right port of the upper chamber by inserting the luer lock adaptor into the port of the biochip. Then, connect the second tubing to the left port of the lower chamber. Repeat this procedure for the right microfluidic cavity.
5. Take a reservoir and add a small drop of cell culture medium to the bottom of the reservoir. Then, insert the reservoir to the opposite side of the first tubing and repeat for the other chamber. Once all ports are connected to a tubing or reservoir, fill the reservoirs with 3.5 mL of cell culture medium.
6. Place the loose side of the tubing, which has the lid attached to it, on top of the reservoir to close the microfluidic system of each chamber. In this state, transport the biochip to the peristaltic pump.
   NOTE: Depending on the distance to the incubator and the laboratory surroundings, a previously cleaned and autoclaved box can be used to transfer the chips to the incubator.
7. Use the peristaltic pump stoppers to connect the tubing to the pump. Connect each tubing to the peristaltic pump in such a way that the medium will flow from the reservoir into the cavity, into the tubing, and via the pump back into the reservoir (Figure 4). The reservoir serves as a bubble trap in the circular perfusion and prevents air from getting trapped in the system. Perfuse each chamber with a flow rate of 50 µL/min resulting in a shear stress of 0.013 dyn/cm² in the upper chamber and 0.006 dyn/cm² in the lower chamber⁸.
   NOTE: If the medium of the lower and upper cavities is moved in opposing directions, a higher three-dimensional outgrowth of the intestinal tissue can be achieved¹³. Hence, the reservoirs of the upper and lower cavities are placed on opposing sides (Figure 4). The circular perfusion reduces the amount of cell culture medium needed but could potentially result in the enrichment of cytokines and metabolites. If desired, a linear perfusion of the biochip is also possible.
8. Perfuse the biochip for 72 h at 37 °C and 5% CO₂.

Figure 4: Biochip connected to peristaltic pump. An example of a biochip connected to a peristaltic pump is presented. Epithelial C2BBe1 cells are cultivated in the lower chamber (red C2-medium is in the reservoirs at the front) whilst the HUVECs are cultivated in the upper chamber (yellowish EC-conditioned-medium is in the reservoirs at the back). The different cell culture media are not mixing due to the barrier function of the grown tissue. The biochip is connected to the peristaltic pump in such a way that the medium flows from the reservoir into the cavity. From here, the medium flows back into the reservoir through the tubing via the pump. Please click here to view a larger version of this figure.

8. LPS-conditioning of the epithelial barrier

1. After 72 h of preperfusion, stop the peristaltic pump and remove the lid connected to the tube of each reservoir. Place it on a sterile tissue next to the pump.
2. Remove all medium and fill the reservoirs with 2 mL of freshly prepared medium. For the epithelial side containing the C2BBe1-cells, add 100 ng/mL LPS to the medium.
   NOTE: The LPS increases the barrier function of the tissue, stimulates the monocyte-derived macrophages to migrate into the epithelial tissue, and allows monocyte-derived dendritic cell differentiation.
3. Reconnect the tubing and lids to the reservoir and continue the circular perfusion at a flow rate of 50 µL/min for an additional 24 h.
   NOTE: From this point forward, the chip model can be used in experiments-compound testing or infection studies. We recommend a medium exchange of 2 mL per reservoir every 24 h.

9. Access to the tissue for different readout methods

1. Collect cell-culture medium supernatants from the reservoirs at all times of the perfusion. Open the reservoir and collect the desired volume (see steps 8.1-8.3). Use these supernatants for the detection of metabolites, cytokines, or other molecules.
2. To access the tissue, use a scalpel to make a precise cut along the outside of the upper chamber and remove the bonding foil to open the microfluidic cavity. The tissue of the intestinal-biochip model is now accessible. Carefully cut along the outside of the microfluidic chamber to detach the membrane from the biochip. Collect the tissue-containing membrane using tweezers.
   CAUTION: Be aware of finger placement during this step and work carefully to prevent accidents. Cut-resistant gloves are recommended.
3. Alternatively, harvest the cells from separate layers inside the biochip using enzyme solutions, i.e., trypsin or cells lysed using Triton X-100-containing buffers.

10. Permeability assessment via FITC-dextran diffusion

NOTE: The barrier function of the tissue can be analyzed via a FITC-dextran permeability assay after disconnection of the peristaltic pump. The FITC-dextran permeability assessment was adapted from Deinhardt-Emmer et al.⁴.

1. Prepare a stock solution of fluorescein isothiocyanate (FITC)-dextran (molecular weight of 3-5 kDa, Table 1).
2. Empty the reservoirs and disconnect the chip from the perfusion.
3. Perform a medium exchange in the upper and lower chamber with phenol red-free medium.
   NOTE: This step is not necessary if phenol red-free medium was already used during the experiment.
4. Add 350 µL of 1 mg/mL FITC-dextran solution to the chamber containing the C2BBe1 cells.
5. Close the ports and incubate the chip for 60 min at 37 °C with the epithelial side facing upwards.
6. After the incubation time, collect the culture medium from both chambers of the chip separately and store at 4 °C, protected from light until the measurement.
7. For the measurement, prepare a standard curve in C2-medium and the EC-conditioned medium without phenol red in the range of 1,000 µg/mL to 0 µg/mL FITC-dextran with 11 consecutive 1:2 serial dilutions.
8. Transfer 200 µL of each sample into a black 96-well plate with a clear bottom. Measure the fluorescence with a microplate reader at an excitation wavelength of 495 nm and an emission wavelength of 517 nm.
9. Use the standard curve to calculate the FITC-dextran concentration of the samples and thereby, the permeability coefficient.

11. Immunofluorescence staining

NOTE: The living tissue can be investigated microscopically. For easier handling, we recommend the detachment of the biochip from the peristaltic pump and the use of long-distance objectives on an inverted microscope. As an endpoint analysis, the tissue can be fixated inside the biochip for procedures like immunofluorescence staining.

1. Stop the peristaltic pump and open the reservoirs of all cavities. Empty the reservoirs and disconnect the tubing, as well as the reservoirs from the biochip.
2. Twice per chamber, wash the microfluidic cavities with 500 µL of cold PBS +/+. Add 500 µL of ice-cold methanol to all cavities and incubate for 15 min at -20 °C. Then, twice per cavity, wash the microfluidic chamber with 500 µL of PBS +/+.
   NOTE: Other fixation methods, such as fixation with 4% paraformaldehyde or Carnoy's fixative, are also suitable. After fixation, the chips can be stored at 4 °C or proceed directly to immune fluorescence staining. CAUTION: Fixating chemicals such as methanol or paraformaldehyde are toxic. Perform the respective tasks under a fume hood and collect the waste accordingly.
3. Open the chip as described in step 9.2 to access the tissue. Cut the tissue-containing PET membrane into up to three pieces to stain in parallel with different immunopanels.
4. Transfer each of the membrane pieces to a separate 24-well plate containing a blocking and permeabilizing solution (Table 1) using precision tweezers. Ensure that the cell layer of interest always faces upwards during the whole staining process. Incubate the membrane pieces for 30 min at RT.
   NOTE: The best staining results are obtained when matching the serum to the secondary antibody. For example, if secondary antibodies are obtained from goat species, we recommend the use of normal goat serum.
5. Transfer the membrane pieces onto a clean glass slide inside a humid chamber. Prepare the primary antibody panel in the staining solution (Table 1) and add 50 µL to each membrane piece. Incubate overnight at 4 °C.
   NOTE: The optimal antibody concentration and staining efficacy can differ between manufacturers and clones. We recommend testing the staining panels beforehand in a 2D cell culture.
6. After incubation, transfer the samples to a 24-well plate and gently wash the membranes for 3 x 5 min with wash solution (Table 1).
7. Again, transfer the membrane pieces onto a clean glass slide inside a humid chamber. Prepare the secondary antibody panel in staining solution (Table 1) and add 50 µL to each membrane piece. If required, add a nuclear counterstain such as 4',6-diamidino-2-phenylindole (DAPI) or Hoechst. Incubate for 30 min at RT.
   NOTE: While working with fluorophores, keep samples protected from light to prevent photobleaching to increase the image quality.
8. After incubation, transfer the samples to a 24-well plate and gently wash the membranes 2 x 5 min with wash solution (Table 1). Then, wash once with PBS +/+ for 5 min.
9. Mount the membrane pieces on a clean glass slide using a fluorescence mounting medium and a cover glass. Store at 4 °C until microscopic imaging.

These representative results show the distinct tissue layers of the intestine-on-chip model. They are immunofluorescent stained as described in protocol section 11. The images were taken with an epifluorescence or confocal fluorescence microscope as z-stacks and processed to an orthogonal projection. See the Table of Materials for details about the microscopical setup and software. Figure 5 shows the vascular layer, a barrier-forming endothelial monolayer, consisting of HUVECs and macrophages. Representative cellular markers such as VE-Cadherin and von Willebrand factor for the endothelial cells and CD68 for the macrophages are stained. The three-dimensional epithelial layer consisting of C2BBe1 cells is shown in Figure 6 and Figure 7. The tissue integrity is confirmed by staining of E-Cadherin and ZO-1. A detailed analysis of all expected tissue markers and tissue polarization can be found in Maurer et al.⁶. Figure 8 shows an exemplary image lacking three-dimensional outgrowth of intestinal villus-like structures. Compared to a typical three-dimensional cell layer (Figure 6), the number of stained nuclei is highly reduced, and cells are only present in a monolayer. This specific phenotype was created by the addition of a histone deacetylase inhibitor to the cell culture medium which inhibits cell proliferation. The barrier permeability is assessed by a permeation assay using FITC-dextran. Permeability increased upon infection with the opportunistic pathogenic C. albicans, as shown in Figure 9A. Infection results in the increased release of cytokines IL-1β, IL-6, IL-8, and IL-10 into the vascular medium supernatant (Figure 9B)⁸.

Figure 5: Fixated vascular side of the intestine-on-chip. Representative immunofluorescence image of the vascular barrier after five days of perfusion. The tissue was fixated with methanol. A z-stack of 1 µm sections was taken with an epifluorescence microscope, and an orthogonal maximum projection of this z-stack is displayed. (A) DNA within nuclei was stained with DAPI. (B) Fully differentiated monocyte-derived macrophages express CD68 and are spread all over the vascular tissue. (C) HUVECs create a confluent layer showing tissue integrity by the formation of adherens junctions such as VE-Cadherin. Additionally, (D) von Willebrand factor is highly expressed by HUVECs. Finally, (E) all markers are shown in an overlay. Scale bars = 100 µm. Abbreviations: DAPI = 4',6-diamidino-2-phenylindole; HUVECs = human umbilical venous endothelial cells. Please click here to view a larger version of this figure.

Figure 6: Fixated epithelial side of the intestine-on-chip. Representative immunofluorescence image of the polarized epithelial cell layer after 5 days of perfusion expressing the junction proteins E-Cadherin and ZO-1. The tissue was fixated with methanol. A z-stack of 1 µm sections was taken with an epifluorescence microscope. An orthogonal maximum projection of this z-stack is displayed. (A) DNA in nuclei was stained with DAPI. A tight epithelial barrier is developed by the formation of (B) adherens junctions, stained with E-Cadherin, and (C) apical tight junctions, stained with ZO-1. Tight junctions and adherens junctions co-localize at the apical cell membrane, as seen in the (D) overlay image. Scale bars = 100 µm. Abbreviation: DAPI = 4',6-diamidino-2-phenylindole. Please click here to view a larger version of this figure.

Figure 7: Fixated epithelial side of the intestine-on-chip with villus-like outgrowth. Representative immunofluorescence image of the three-dimensional epithelial cell layer after six days of perfusion. The tissue was fixated with Carnoy's fixative. A z-stack of 1 µm sections was taken with a confocal laser scanning microscope, and (A) an orthogonal maximum projection of this z-stack is displayed. DNA in nuclei was stained with Hoechst. Additionally, the tissue was stained with wheat germ agglutin, a lectin, which stains human mucin. (B) The three-dimensional growth of villus-like and crypt-like structures of the epithelium can be detected in the x- and y-sections of the z-stack. Scale bars = 50 µm. Abbreviation: WGA = wheat germ agglutinin. Please click here to view a larger version of this figure.

Figure 8: Fixated epithelial side of the intestine-on-chip without villus-like outgrowth. Representative immunofluorescence image of the epithelial cell layer treated with suberoylanilide hydroxamic acid (SAHA), a histone deacetylase inhibitor, preventing cell proliferation. The tissue was fixated with methanol. A z-stack of 1 µm sections was taken with an epifluorescence microscope. An orthogonal maximum projection of this z-stack is displayed. (A) DNA in nuclei was stained with DAPI. In (B), adherens junctions are stained with E-Cadherin, and in (C), tight junctions are stained with ZO-1. Tight junctions and adherens junctions co-localize, as seen in the overlay image (D). Scale bars = 100 µm. Abbreviation: DAPI = 4',6-diamidino-2-phenylindole. Please click here to view a larger version of this figure.

Figure 9: Pathogen-dependent immune response of integrated innate immune cells: This figure is modified from Kaden et al.⁸. The intestine-on-chip model uninfected (-) or infected (+) with C. albicans SC5314 was analyzed for (A) permeability with FITC-Dextran (3-5 kDa) and (B) cytokine secretion. Please click here to view a larger version of this figure.

Table 1: List and composition of solutions used in this protocol. Abbreviation: PBS -/- = Dulbecco's phosphate-buffered saline without magnesium and calcium.

The presented protocol details the necessary steps for generating an immunocompetent intestine-on-chip model. We described specific techniques and possible readout methods such as immunofluorescence microscopy, cytokine and metabolite analysis, flow cytometry, protein and genetic analysis, and permeability measurement.

The described model consists of primary HUVECs, monocyte-derived macrophages, and monocyte-derived dendritic cells co-cultured with a 3D layer of intestinal epithelial cells representing aspects of mucus-secreting, absorptive, enteroendocrine, and Paneth cell-like populations, as described before^6,14. Some critical points need to be kept in mind while performing the protocol steps. At all steps, air bubble formation should be prevented within the microfluidic system. Therefore, always pipette reverse (take up the volume from the second stop and dispense only until the first stop) when handling the biochip or the tubing. To prevent high pressure or shear forces on cells within the biochip, always pipette gently and keep the opposing chamber closed when pipetting in the biochip (e.g., when exchanging the medium in the upper chamber, close the lower chamber with the respective plugs). The biochip allows the seeding of cells on two sides of a PET membrane. Hence, the model can be set up with intestinal villus-like structures underneath the PET membrane or growing on the membrane upwards. Depending on the research interest, one or the other orientation of the cells has certain benefits in terms of physical interaction, such as gravity-aiding cell attachment of circulating cells suspended in the flow medium.

Due to the microfluidic medium flow, the polarized intestinal cells form a three-dimensional tissue resembling crypt-like and villus-like structures, which shape a profound apical tissue barrier by a dense network of tight junctions and adherens junctions⁶. Additionally, the integrated innate immune cells increase the intestinal barrier function and allow for the detailed investigation of cytokine and immune cell responses toward various stimuli, including donor-specific immune responses^3,6. The model offers a physiologically relevant platform to study microbial commensals (e.g., Lactobacillus rhamnosus⁶) and pathogens (e.g., Candida albicans^6,8 ) of the human gut. The model was leveraged to investigate mechanisms of gut microbial composition changes and regulation of the mucosal immune response⁶. In the study, we demonstrated how to recreate the physiological immune tolerance of the intestinal lumen and support homeostatic colonization by living microorganisms. Furthermore, pre-colonization of the luminal side of the model with probiotic Lactobacillus rhamnosus reduced tissue invasion of Candida albicans and limited its translocation to the vascular compartment, similar to the in vivo situation. The study proved the model's potential for studying microbial interactions, immune responses, and pathogenicity mechanisms of the intestine under physiologically relevant conditions in vitro.

Still, this chip platform has limitations, namely, the use of cancer-derived intestinal cells, which are not fully capable of representing the human gut, and the use of primary endothelial and immune cells. Hypoxic culture conditions can be established by culturing the model in a hypoxia incubator under perfused conditions to better reflect the in vivo situation of the human intestine.

In comparison to other OoC platforms, such as the "gut-on-a-chip" model published by Kim et al.^14,15 the initial cell number forming the presented tissue model is about 10-fold higher. This, for example, enables the simultaneous analysis of up to three immunofluorescence staining panels within a singular experiment. The larger cell numbers further streamline endpoint analyses, including flow cytometry, western blot analysis, and other standard off-the-shelf assays, which require cell culture supernatants such as cytokine profiling and lactate dehydrogenase (LDH) measurement. Gaining multiple readouts from a single experiment reduces overall costs and experimental time. Off-the-shelf hardware solutions from a broad spectrum of manufacturers are available due to the standardized biochip platform that features industry interfaces and footprints (microscopic slide format of the chip, ports in luer lock format, and aligned on the 96 well matrix). This also includes many membranes of different materials and pore sizes. The pore size can influence cell growth and tissue integrity¹⁶ as well as cell migration.

In summary, this protocol presents a detailed and adaptable intestine-on-chip model that replicates crucial elements of the human gut environment. It provides a scalable and effective platform for conducting extensive research on gut microbiota and the intestinal host tissue, with a particular focus on host immune responses and intestinal pathophysiology. Thus, the model is capable of bridging the gap between traditional in vitro models and human biology, making it an essential tool for in-depth studies on intestinal biology.