Vascular inflammation on a chip: A scalable platform for trans-endothelial electrical resistance and immune cell migration

Abstract

The vasculature system plays a critical role in inflammation processes in the body. Vascular inflammatory mechanisms are characterized by disruption of blood vessel wall permeability together with increased immune cell recruitment and migration. There is a critical need to develop models that fully recapitulate changes in vascular barrier permeability in response to inflammatory conditions. We developed a scalable platform for parallel measurements of trans epithelial electrical resistance (TEER) in 64 perfused microfluidic HUVEC tubules under inflammatory conditions. Over 250 tubules where exposed to Tumor necrosis factor alpha (TNFα) and interferon gamma (INF-γ) or human peripheral blood mononuclear cells. The inflammatory response was quantified based on changes TEER and expression of ICAM and VE-cadherin. We observed changes in barrier function in the presence of both inflammatory cytokines and human peripheral blood mononuclear cells, characterized by decreased TEER values, increase in ICAM expression as well changes in endothelial morphology. OrganoPlate 3-lane64 based HUVEC tubules provide a valuable tool for inflammatory studies in an automation compatible manner. Continuous TEER measurements enable long term, sensitive assays for barrier studies. We propose the use of our platform as a powerful tool for modelling endothelial inflammation in combination with immune cell interaction that can be used to screen targets and drugs to treat chronic vascular inflammation.

Keywords: TEER; TNFα, INF-γ and OrganoPlate; endothelium; immune cell migration; organ-on-a-chip; vascular inflammation.

Figure 1

A screening platform for automated seeding of an organ-on-a-chip endothelial model combined with high throughput TEER measurements. (A) A OrganoPlate 3-lane 64 tissues in a standardized 384-titerplate format. Here, the Organoplate 3-lane 64 seeded on a Biomek i5 automated liquid handling platform. The expanded portion of the plate highlights the top view of one individual chip covering six microtiter wells. One microfluidic chip is enlarged to show the 3 individual microchannels. The perfusion channels used for vessel formation are filled with red dye and the the gel channel is filled with blue dye. (B) The bottom view of the OrganoPlate 3-lane 64 highlighting the microfluidic channels, making up 64 individual chips attached to the bottom of the microtiter plate, allowing for high content imaging of each individual chip. One microfluidic chip is enlarged to highlight the 3 individual microchannels. The perfusion channels used for vessel formation are highlighted in red and the gel channel is highlighted in blue. (C) The OrganoTEER, a commercially available automated TEER measurement system compatible with previous OrganoPlate 3-lane 40 based tubular models (28). (D) OrganoTEER software used to perform TEER measurements and perform automated analysis of the results on an OrganoPlate 3-lane 64.

Figure 2

Immunofluorescent staining and TEER measurements of HUVEC endothelial tubules in OrganoPlate 3-lane 64. (A) Montage of Immunofluorescent images (VE-cadherin (green) and nucleus (blue)) of 64 HUVEC endothelial tubes (right perfusion channel) cultured against a collagen I ECM layer (center channel) in the OrganoPlate 3-lane 64 (scale bar 1000 µm). Bottom three rows were Staurosporine treated. One chip from the montage is blown up to illustrate how the chip looks (scale bar 150 µm). Another zoom of the endothelial tube is shown to demonstrate the cell morphology (scale bar 50 µm). (B) Confocal reconstruction of a HUVEC tubule using VE-Cadherin (green) and DAPI (blue) staining. (C) Schematic illustration of how a microfluidic chip and the positioning of the TEER electrodes. Both current carrying and voltage sensing loops are formed across the gel and perfusion channel via four pairs of electrodes, shorted pairwise in the inlets and outlets of their perfusion channels. The bottom schematic depicts a side view of the chip in the X-Z plane showing how the endothelial cells will grow to form a tubule against the gel, while the left channel remains empty. (D) values of the HUVEC tube at 0, 24, and 48 hours with the addition of Staurosporine to interrupt the barrier (n= 4-6). Scale bars are 100µm.

Figure 3

Cytokine response of HUVEC endothelial tubes in the OrganoPlate 3-lane 64. (A) Representative phase contrast images of zoom in of endothelial vessel exposed to inflammatory triggers for 44hrs. Concentration of TNFα and INF-γ in ng/ml. Scale bar is 100µm. (B–D) Relative TEER timelapse of 44 hours of exposure to increasing concentrations of (B) INF-γ, (C) TNFα and (D) combination of TNFα and INFγ (N=3, n=3-5 per experiment).

Figure 4

ICAM-1 and VE-Cadherin staining of endothelial vessel. (A) ICAM-1 expression staining of endothelial vessel exposed to TNFα or/and INF-γ in ng/ml for 44hrs. Scale bar is 100µm. (B) Quantification of ICAM-1 staining. Significant difference between cytokine conditions (*p<0.05, ****p < 0.0001) as well as a dose dependent effect for TNFα and TNFα + INF-γ (****p < 0.0001) Data was analyzed using Two-way ANOVA tests, followed by a Dunnet’s multiple comparison tests and Tukey’s multiple comparison test. (C) Montage of the max projection image of VE-Cadherin staining of the bottom 10 z-steps (3µm step size). Scale bar 100µm. (D) Quantification of cell roundness determined from VE-Cadherin staining. Significant difference between cytokine conditions (***p=0.0002, ****p <0.0001) as well as a dose dependent effect for TNFα and TNFα + INF-γ (****p < 0.0001). Data was analyzed using Two-way ANOVA tests, Dunnet’s multiple comparison tests and Tukey’s multiple comparison tests. (N=2, n=3-6) Scale bars are 100µm.

Figure 5

Perfusion and extravasation of peripheral blood mononuclear cells (PBMCs) through a HUVEC tubule into an ECM gel in an OrganoPlate 64. (A) Schematic of TEER electrodes with endothelial tubule and PBMCs adhered to the endothelial tube (B) Schematic of TEER electrodes with endothelial tubule and PBMCs migrating from the endothelial tube into gel channel due to addition of chemokine (purple) in basal lateral channel. (C) Relative change in barrier of PBMCs or 10 ng/ml of TNFα + INF-γ can be observed in the 48h timelapse. (D) TEER timelapse comparing conditions with and without CLCX12. No difference was observed between comparable conditions with +/- CLCX12. (E) Number of PBMCs adhered to the HUVEC tubules 48 hour after addition. Significant difference unstimulated and stimulated PBMCs (****p < 0.0001) was analyzed using Welch’s T tests. (F) PBMC migration out of endothelial vessel into gel channel. Significant difference between PBMC and PBMCs with cytokines and CLCX12 (* p=0.04, ** p < 0.0018) was analyzed using Brown-Forsythe and Welch ANOVA tests and Dunnett’s T3 multiple comparison test. 3way ANOVA showed no significant (ns) difference between unstimulated and stimulated PBMCs in migration of PBMCs into the ECM. (G) Fluorescent based image of immune cells perfused through a HUVEC tubule and migrated into ECM. (H) Area of interest from (G) to highlight PBMC migration and staining. Shown are nucleus (blue), CellTracker™ Orange stain (yellow) and CD45 (red). (I) ICAM-1 quantification after exposure and addition of PBMCs. Significant difference due to addition of cytokines (****p < 0.0001) as well an effect of addition of stimulated PBMCs without cytokines (****p < 0.0001) Data was analyzed using Two-way ANOVA tests with Šídák’s multiple comparisons test. (J) Quantification of cell roundness determined from VE-Cadherin staining. Significant difference due to addition of cytokines (****p < 0.0001) as well a significant different between medium and stimulated PBMC and unstimulated PBMCs and stimulated PBMCs (****p < 0.0001). Data was analyzed using Two-way ANOVA tests with Tukey’s multiple comparisons test. (N=2, n=3-6).