Towards Biohybrid Lung: Induced Pluripotent Stem Cell Derived Endothelial Cells as Clinically Relevant Cell Source for Biologization

Abstract

In order to provide an alternative treatment option to lung transplantation for patients with end-stage lung disease, we aim for the development of an implantable biohybrid lung (BHL), based on hollow fiber membrane (HFM) technology used in extracorporeal membrane oxygenators. Complete hemocompatibility of all blood contacting surfaces is crucial for long-lasting BHL durability and can be achieved by their endothelialization. Autologous endothelial cells (ECs) would be the ideal cell source, but their limited proliferation potential excludes them for this purpose. As induced pluripotent stem cell-derived ECs enable the generation of a large number of ECs, we assessed and compared their capacity to form a viable and confluent monolayer on HFM, while indicating physiologic EC-specific anti-thrombogenic and anti-inflammatory properties. ECs were generated from three different human iPSC lines, and seeded onto fibronectin-coated poly-4-methyl-1-pentene (PMP) HFM. Following phenotypical characterization, ECs were analyzed for their thrombogenic and inflammatory behavior with or without TNFα induction, using FACS and qRT-PCR. Complementary, leukocyte- and platelet adhesion assays were carried out. The capacity of the iPSC-ECs to reendothelialize cell-free monolayer areas was assessed in a scratch assay. ECs sourced from umbilical cord blood (hCBECs) were used as control. iPSC-derived ECs formed confluent monolayers on the HFM and showed the typical EC-phenotype by expression of VE-cadherin and collagen-IV. A low protein and gene expression level of E-selectin and tissue factor was detected for all iPSC-ECs and the hCBECs, while a strong upregulation of these markers was noted upon stimulation with TNFα. This was in line with the physiological and strong induction of leukocyte adhesion detected after treatment with TNFα, iPSC-EC and hCBEC monolayers were capable of reducing thrombocyte adhesion and repopulating scratched areas. iPSCs offer the possibility to provide patient-specific ECs in abundant numbers needed to cover all blood contacting surfaces of the BHL with a viable, non-thrombogenic and non-inflammatory monolayer. iPSC-EC clones can differ in terms of their reendothelialization rate, and pro-inflammatory response. However, a less profound inflammatory response may even be advantageous for BHL application. With the proven ability of the seeded iPSC-ECs to reduce thrombocyte adhesion, we expect that thrombotic events that could lead to BHL occlusion can be avoided, and thus, justifies further studies on enabling BHL long-term application.

Keywords: EC activation; biohybrid lung; endothelialization; hemocompatibility; hollow fiber membrane; induced pluripotent stem cells; membrane oxygenator.

Figure 1

Imaging of HFM after endothelialization with iPSC-EC1. (A,B) Viable calcein-positive iPSC-ECs (green) seeded on fibronectin-coated HFM form a confluent monolayer around the individual fibers of the HFM. A: Top view, B: bottom view. (C) iPSC-ECs are interconnected via VE-cadherin (green). (D) Basal-lamina-like matrix envelops iPSC-EC-seeded hollow fibers, indicated by de novo synthesized collagen type-IV (magenta). (E,F) Nuclei counterstained with DAPI and presented in false colors (yellow).

Figure 2

iPSC-derived ECs seeded on fibronectin-coated PMP membranes show upregulation of markers of inflammation and hemostasis upon stimulation with 10 ng/mL TNFα for 6 h. (A) Flow cytometry analysis of E-selectin (CD62E) surface expression. Representative histograms of untreated (blue histograms) vs. TNFα-treated (red histogram) ECs. (B) Quantification of mean fluorescence intensities (MFI, represents geometric mean normalized to corresponding isotype controls) of E-selectin expression with flow cytometry (left) and calculated fold-change of untreated vs. TNFα-treated ECs (right). (C) Representative histograms of flow cytometry analysis ECs for tissue factor (CD142) surface expression. (D) Quantification of mean fluorescence intensity (MFI) of tissue factor surface expression with flow cytometry (left) and calculated fold-change of untreated vs. TNFα-treated ECs (right). (E) qPCR quantification of activation-relevant marker genes E-selectin, tissue factor, ICAM-1 and VCAM-1. The mRNA expression was normalized to the expression of β-actin using the ΔCt-method. All data are representative of 3 technical replicates for each EC type. Reported values are given as means ± standard error of mean. Differences were considered significant at p < 0.05. # Significantly different TNFα-treated vs. untreated. * Fold change significantly different vs. hCBECs.

Figure 3

iPSC-ECs on PMP membranes show different levels of leukocyte adhesion after TNFα treatment. (A) Exemplary images of cell tracker red-stained HL-60 leukocytes (red) on iPSC-EC endothelialized gas exchange membranes under standard condition and after treatment with TNFα for 6 h. Except for monolayers established from iPSC-EC1s, appreciably increased numbers of HL-60 cells adhered to EC-monolayers incubated with TNFα. Scalebar: 250 µm. (B) Quantification of adhered HL-60 leukocytes. Reported values are given as means ± standard deviation. Significant differences between untreated and TNFα-stimulated groups were marked with * at p < 0.05, ** when p < 0.01 and with *** when p < 0.001.

Figure 4

Reduced platelet adhesion to PMP membranes endothelialized with iPSC-ECs. Significantly less thrombocytes adhered to endothelialized samples, compared to blood plasma coated PMP membranes (p < 0.05). A higher platelet count could be detected when iPSC-EC2, iPSC-EC3 and hCBEC were activated with TNFα for 6 h, while differences were statistically not significant. All data represent the means of independent triplicates for each EC type and treatment. Dunnet’s multiple comparison test, differences to the blood plasma group were considered significant at p < 0.05 (*).

Figure 5

Regenerative capacity of endothelial monolayers on PMP gas exchange membranes. (A) Recovered area per minute. Results are given as means from independent triplicates with standard deviation. Analysis for statistically significant differences was performed using one-way ANOVA with Tukey’s post hoc test. Differences were considered significant at p < 0.01 (**) and p < 0.0001 (****). (B) Representative phase contrast pictures showing the scratched area at start (t = 0) of the experiment and after 12 h. Yellow lines indicate the cell-free area. Scale: 250 µm.