Vasculature-on-a-chip platform with innate immunity enables identification of angiopoietin-1 derived peptide as a therapeutic for SARS-CoV-2 induced inflammation

Abstract

Coronavirus disease 2019 (COVID-19) was primarily identified as a novel disease causing acute respiratory syndrome. However, as the pandemic progressed various cases of secondary organ infection and damage by severe respiratory syndrome coronavirus 2 (SARS-CoV-2) have been reported, including a breakdown of the vascular barrier. As SARS-CoV-2 gains access to blood circulation through the lungs, the virus is first encountered by the layer of endothelial cells and immune cells that participate in host defense. Here, we developed an approach to study SARS-CoV-2 infection using vasculature-on-a-chip. We first modeled the interaction of virus alone with the endothelialized vasculature-on-a-chip, followed by the studies of the interaction of the virus exposed-endothelial cells with peripheral blood mononuclear cells (PBMCs). In an endothelial model grown on a permeable microfluidic bioscaffold under flow conditions, both human coronavirus (HCoV)-NL63 and SARS-CoV-2 presence diminished endothelial barrier function by disrupting VE-cadherin junctions and elevating the level of pro-inflammatory cytokines such as interleukin (IL)-6, IL-8, and angiopoietin-2. Inflammatory cytokine markers were markedly more elevated upon SARS-CoV-2 infection compared to HCoV-NL63 infection. Introduction of PBMCs with monocytes into the vasculature-on-a-chip upon SARS-CoV-2 infection further exacerbated cytokine-induced endothelial dysfunction, demonstrating the compounding effects of inter-cellular crosstalk between endothelial cells and monocytes in facilitating the hyperinflammatory state. Considering the harmful effects of SARS-CoV-2 on endothelial cells, even without active virus proliferation inside the cells, a potential therapeutic approach is critical. We identified angiopoietin-1 derived peptide, QHREDGS, as a potential therapeutic capable of profoundly attenuating the inflammatory state of the cells consistent with the levels in non-infected controls, thereby improving the barrier function and endothelial cell survival against SARS-CoV-2 infection in the presence of PBMC.

Figure 1:. Establishing an organ-on-a-chip vascular model to study coronavirus induced vascular dysfunction using common cold virus HCoV-NL63.

(A) Photo of the InVADE platform. Fluorescence image illustrates autofluorescent polymeric vessel spanning a well in the platform. (B) Illustration of the gravity-assisted flow for long-term perfusion. Three wells (inlet, tissue well, and outlet) are connected with bioscaffold (blue) to make one perfusable unit. (C) Experimental timeline showing the sequence of HCoV-NL63 treatment. (D) Fluorescent microscope images of HUVECs treated with MOI 1 and MOI 10 of HCoV-NL63 48 hours after infection. Scale bar=200μm. (E) Representative confocal fluorescent images of HUVEC VE-cadherin immunostaining (red). Cells were treated with MOI 1 and MOI 10. Scale bar=200μm. (F) Endothelial cell barrier function in HUVEC under perfusion as measured using the 70kDa TRITC-dextran based tracer assay. Data are mean ± SD, N=3 with n=2–4 from each experiment. One-way ANOVA with Holm-Sidak post hoc test, p<0.05; *significant relative to the untreated control. (G) Quantification of LDH concentration in culture media under HCoV-NL63 infection. Data are mean ± SD, N=3 with n=2–4 from each experiment. One-way ANOVA with Holm-Sidak post hoc test. No statistical difference was observed. Analysis of pro-inflammatory cytokine profile upon HCoV-NL63 infection at MOI 1 after 48 hours. Concentration of (H) GM-CSF, (I) IL-1β, (J) IL-6, (K) IL-8, (L) MCP-1 measured by EVE technology. Data are mean ± SD, N=3 with n=1–3 from each experiment. One-way ANOVA with Holm-Sidak post hoc test, p<0.05; *significant related to the untreated control. Analysis of endothelial cell activation markers upon HCoV-NL63 infection after 48 hours. Concentration of soluble (M) endothelin-1, (N) ICAM-1, (O) E-selectin, (P) angiopoietin-2, and (Q) TREM-1 measured by ELLA. Data are mean ± SD, N=3 with n=1–2 from each experiment. One-way ANOVA with Holm-Sidak post hoc test, p<0.05; *significant related to the untreated control.

Figure 2:. Organ-on-a-chip vascular model allows for tracking of monocyte homing to the coronavirus infected endothelium and monocyte recruitment.

(A) Illustration of monocyte labeling with CellTracker dye and subsequent perfusion after removal of HCoV-NL63 inoculum. (B) The number of monocytes attaching on the endothelial cells 48 hours after HCoV-NL63 infection at MOI 1. Data are mean ± SD, N=3 with n=1–3 from each experiment. One-way ANOVA with Holm-Sidak post hoc test, p<0.05; *significant related to non-virus treated monocyte control. Representative confocal fluorescent images of (C) CellTracker labelled monocyte (white) and (D) HUVEC VE-cadherin (red) 48 hours after HCoV-NL63 infection at MOI 1. Scale bar = 200μm.

Figure 3:. SARS-CoV-2 treatment in the organ-on-a-chip vascular model results in profound endothelial cell activation and inflammatory cytokine secretion.

(A) Endothelial cell barrier function in HUVEC under perfusion as measured using 70kDa TRITC-dextran based tracer assay 48- and 72-hours after SARS-CoV=2 infection at MOI 1. Data are mean±SD, N=3 with n=2–4 from each experiment. Two-way ANOVA with Holm-Sidak post hoc test, *p<0.05, ****p<0.0001. (B) Quantification of LDH concentration in culture media. Data are mean±SD, N=3 with n=2–3 from each experiment. One-way ANOVA with Holm-Sidak post hoc test, *p<0.05. (C) Representative confocal fluorescent images of HUVEC VE-cadherin 48- and 72-hours after SARS-CoV-2 infection at MOI 1. Scale bar = 200μm. Analysis of endothelial cell activation markers upon SARS-CoV-2 infection at MOI 1 after 48 and 72 hours. The concentration of (ng-2, (E) E-selectin, (F) IL-6, (G) IL-8, (H) endothelin-1, and (I) ICAM-1 as measured by ELLA. Data are mean±SD, N=3 with n=1–3 from each experiment. Two-way ANOVA with Holm-Sidak post hoc test, *p<0.05, **p<0.01.

Figure 4:. Comparison of Transwell and InVADE system for studies of SARS-CoV-2 induced vascular dysfunction.

(A) Endothelial cell barrier function in HUVEC in both InVADE and Transwell system as measured using the 70kDa TRITC-dextran based tracer assay. Data are mean±SD, n=3. Two-way ANOVA with Holm-Sidak post hoc test, p<0.05. (B) The number of monocytes attaching on the endothelial cells upon SARS-CoV-2 infection for 72 hours at MOI 1 in Transwell. The monocyte numbers are normalized to surface area of the respective system. Data are mean±SD, N=3. One-way ANOVA with Holm-Sidak post hoc test. (C) Representative fluorescent images of CellTracker labelled monocytes 72 hours after SARS-CoV-2 infection at MOI 1. Scale bar=100μm. Analysis of endothelial cell activation markers upon peptide treatment in InVADE system. Concentration of (D) Ang-2, (E) E-selectin, (F) IL-6, (G) IL-8, (H) ICAM-1, (I) endothelin-1, and (K) TREM-1. Data are mean±SD. N=3. One-way ANOVA with Holm-Sidak post hoc test, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Figure 5:. Potential immunomodulatory role of QHREDGS peptide as a therapeutic against SARS-CoV-2 infection in the presence of PBMC.

(A) Experimental timeline showing the sequence of HUVEC seeding and addition of QHREDGS peptide in the presence of PBMCs. (B) The number of monocytes attached to the endothelial cells upon SARS-CoV-2 infection for 72 hours at MOI 1. Data are mean ± SD, N=3 with n=1–3 from each experiment. One-way ANOVA with Holm-Sidak post hoc test, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. (C) Endothelial cell barrier function in HUVEC under perfusion as measured using the 70kDa TRITC-dextran based tracer assay 72 hours after SARS-CoV-2 infection at MOI 1 in the presence of monocyte and Q-peptide. Red shading represents the range of values obtained from TRITC dextran with SARS-CoV-2 infection at MOI 1 after 72 hours. Data are mean±SD, N=3 with n=2–4 from each experiment. One-way ANOVA with Holm-Sidak post hoc test, p<0.05; *significant related to non-virus control. (D) Quantification of LDH concentration in culture media under Q-peptide treatment. Red shading: LDH release with SARS-CoV-2 infection at MOI 1 after 72 hours. Data are mean±SD, N=3 with n=2–3 from each experiment. One-way ANOVA with Holm-Sidak post hoc test, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Representative confocal fluorescent images of (E) monocytes and (F) HUVEC VE-cadherin 72-hours after SARS-CoV-2 infection at MOI 1 for non-virus treated control, non-virus-treated PBMC control, virus-PBMC and Q-peptide, and virus-PBMC and scrambled (sQ)-peptide samples. Scale bar = 200μm. Analysis of endothelial cell activation markers upon SARS-CoV-2 infection at MOI 1 after 72 hours. Concentration of (G) Ang-2, (H) E-selectin, (I) IL-6, (J) IL-8, (K) ICAM-1, (L) endothelin-1 and (M) TREM-1. (G - M) Red shading represents the range of values obtained from cytokine release with SARS-CoV-2 infection at MOI 1 after 72 hours. Data are mean±SD, N=3 with n=2–4 from each experiment. Two-way ANOVA with Holm-Sidak post hoc test, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.