Microvascular Mimetics for the Study of Leukocyte-Endothelial Interactions

Abstract

Introduction: The pathophysiological increase in microvascular permeability plays a well-known role in the onset and progression of diseases like sepsis and atherosclerosis. However, how interactions between neutrophils and the endothelium alter vessel permeability is often debated.

Methods: In this study, we introduce a microfluidic, silicon-membrane enabled vascular mimetic (μSiM-MVM) for investigating the role of neutrophils in inflammation-associated microvascular permeability. In utilizing optically transparent silicon nanomembrane technology, we build on previous microvascular models by enabling in situ observations of neutrophil-endothelium interactions. To evaluate the effects of neutrophil transmigration on microvascular model permeability, we established and validated electrical (transendothelial electrical resistance and impedance) and small molecule permeability assays that allow for the in situ quantification of temporal changes in endothelium junctional integrity.

Results: Analysis of neutrophil-expressed β₁ integrins revealed a prominent role of neutrophil transmigration and basement membrane interactions in increased microvascular permeability. By utilizing blocking antibodies specific to the β₁ subunit, we found that the observed increase in microvascular permeability due to neutrophil transmigration is constrained when neutrophil-basement membrane interactions are blocked. Having demonstrated the value of in situ measurements of small molecule permeability, we then developed and validated a quantitative framework that can be used to interpret barrier permeability for comparisons to conventional Transwell™ values.

Conclusions: Overall, our results demonstrate the potential of the μSiM-MVM in elucidating mechanisms involved in the pathogenesis of inflammatory disease, and provide evidence for a role for neutrophils in inflammation-associated endothelial barrier disruption.

Keywords: Endothelial permeability; Microfluidics; Neutrophil transendothelial migration; Silicon nanomembranes; Transendothelial electrical resistance.

Figure 1

Microvascular Mimetic (µSiM-MVM) platform: design and assembly. (a) 2D schematic showing apical (luminal) and basal (abluminal) compartments separated by a silicon nanomembrane, which also acts as a scaffold for incorporating type 1 collagen gel as an extracellular matrix (ECM) mimic. Top and bottom channels are accessible using independent inlet/outlet ports. Indium tin oxide (ITO) electrodes are assembled to enable electrical resistance measurement. (b) Exploded 3D view of the microfluidic system illustrating the silicone gaskets, pressure sensitive adhesive (PSA) and electrodes. (c) Assembled device without electrodes (for visual clarity) highlighting the miniaturized nature and the transparency of the platform.

Figure 2

TEER and EC morphology under shear. (a) TEER measurement was performed over 48–52 h post cell seeding in the µSiM-MVM. TEER rises during first 24 h (red curve), after which it remains stable in the absence of shear (blue curve) or progresses to an elevated level in the presence of shear (brown curve). The addition of thrombin immediately causes the TEER to decay below 50% of its maximum value in less than half an hour (black dashed line). Error bars represent standard error of mean with at least n = 5. (b) Endothelial cells grown under static condition for 24 h demonstrate cobblestone morphology as observed in the phase image on the top left, and the corresponding image analysis in MATLAB reveals the isotropic orientation of the cell population. Endothelial cells exposed to shear stress of 10 dynes cm⁻² for 24 h exhibit uniformly aligned morphology with a majority of the cells parallel to the flow direction (Y axis).

Figure 3

Neutrophil transmigration through a collagen matrix. (a) Even in the presence of endothelial cells and supplemented collagen I gel, neutrophils transmigrate to the abluminal channel. The presence of basal gel (below the nanomembrane) does not significantly reduce the amount of migrated cell population as observed from the representative images of cells on the cover-slip 3 h after addition (17.54 ± 4.29% in absence of basal gel, and 19.89 ± 5.52% in the presence of basal gel, n = 3 each; the percentage of cells calculated by dividing the number of cells in the basal chamber 3 h later to the number of cells in the top chamber at the onset of migration). In the absence of endothelial cells, neutrophils (hPMN) migrate fastest due to potential absence of any steric inhibition. Scale bar = 100 μm. (b) Scanning electron microscopy (SEM) analysis shows the ability of hPMNs to transmigrate through the collagen gel, eventually residing on the underlying nanomembrane.

Figure 4

Optical clarity of μSiM devices permit real-time observations of EC junctions. (a) Phase time lapse images of neutrophils (yellow arrow) transmigrating through a HUVEC monolayer reveal potential separations in EC junctions (blue arrow). Scale bar = 10 μm. (b) Live immunolabeling of HUVEC junctional proteins (VE-cadherin; green) further highlights the disruption of the EC barrier during a neutrophil transmigration event. Scale bar = 20 μm.

Figure 5

Neutrophils migrate uninhibited in the interstitial space in the absence of β₁ integrin blocking (a) but remain trapped in the subendothelial space in the presence of the blocking antibody against β₁ integrins (b). (c) Percent migration from the apical to basal compartments under the influence of basally added fMLP over the duration of 3 h reduces significantly in response to β₁ integrin blocking. Error bars represent standard error of mean; n = 3 for each condition. (d) The disruption of 3D migration of β₁ integrin-blocked neutrophils was again revealed under SEM. The yellow inset shows the zoomed-in portion of the membrane indicating the absence of neutrophils, which have egressed beyond the collagen gel. The entire membrane-area was thoroughly scanned to identify any neutrophils present on the porous membrane.

Figure 6

Endothelial cell monolayer disruption due to neutrophil transmigration: (a) Differential decay in the net impedance during neutrophil migration with (red curve) and without (blue curve) the blocking of β₁ integrins on the neutrophil surface. Impedance values for each individual scan was normalized to its initial value before averaging, to yield a common starting point for comparison. (For each test condition, n = 3) Error bars represent standard error of the mean. Final time points were statistically compared by t test, *p < 0.05. (b) The changes in the diffusion of FITC-dextran was monitored during the extravasation of untreated neutrophils (blue curve) or β₁ integrin blocked neutrophils (red curve). The increase in the fluorescent signal was greater for the untreated neutrophils than for antibody-treated neutrophils, confirming the permeability-regulating effects of antibody treatment. (n = 5 for blue curve, and n = 2 for red curve) Error bars represent standard error of the mean. Final time points were statistically compared by t test, *p < 0.05.

Figure 7

Quantitative analysis of in situ small molecule permeability. (a) Open-top μSiM devices were seeded with HPAECs and grown to confluency. (b) The diffusive properties of 40 kDa FITC-dextran through the monolayer were recorded in the basal compartment using an epifluorescence scope and ×40 objective. (c) Changes in normalized fluorescence intensity 50 μm from the membrane edge were compared to expected results based on conventional Transwell™ permeability values and a COMSOL simulation of the experimental configuration. Root mean square error (RMSE) represents the deviation of the experimental results from the expected. “Cell free” represents media only devices, while “Native” represents HPAEC seeded devices. (d) Additionally, conventional permeability values were determined and compared to values recorded in Transwell™ systems. A paired t-test showed a lack of statistical significance between the two experimental configurations (n = 3). Error bars represent standard error of mean.