Three-Dimensional Vascular Network Assembly From Diabetic Patient-Derived Induced Pluripotent Stem Cells

Abstract

Objective: In diabetics, hyperglycemia results in deficient endothelial progenitors and cells, leading to cardiovascular complications. We aim to engineer 3-dimensional (3D) vascular networks in synthetic hydrogels from type 1 diabetes mellitus (T1D) patient-derived human-induced pluripotent stem cells (hiPSCs), to serve as a transformative autologous vascular therapy for diabetic patients.

Approach and results: We validated and optimized an adherent, feeder-free differentiation procedure to derive early vascular cells (EVCs) with high portions of vascular endothelial cadherin-positive cells from hiPSCs. We demonstrate similar differentiation efficiency from hiPSCs derived from healthy donor and patients with T1D. T1D-hiPSC-derived vascular endothelial cadherin-positive cells can mature to functional endothelial cells-expressing mature markers: von Willebrand factor and endothelial nitric oxide synthase are capable of lectin binding and acetylated low-density lipoprotein uptake, form cords in Matrigel and respond to tumor necrosis factor-α. When embedded in engineered hyaluronic acid hydrogels, T1D-EVCs undergo morphogenesis and assemble into 3D networks. When encapsulated in a novel hypoxia-inducible hydrogel, T1D-EVCs respond to low oxygen and form 3D networks. As xenografts, T1D-EVCs incorporate into developing zebrafish vasculature.

Conclusions: Using our robust protocol, we can direct efficient differentiation of T1D-hiPSC to EVCs. Early endothelial cells derived from T1D-hiPSC are functional when mature. T1D-EVCs self-assembled into 3D networks when embedded in hyaluronic acid and hypoxia-inducible hydrogels. The capability of T1D-EVCs to assemble into 3D networks in engineered matrices and to respond to a hypoxic microenvironment is a significant advancement for autologous vascular therapy in diabetic patients and has broad importance for tissue engineering.

Keywords: diabetes mellitus; endothelial cells; hydrogels; induced pluripotent stem cells; vascular networks.

Figure 1. EVC Differentiation of T1D-hiPSCs

(A) LM images of the differentiating EVCs on day 1 in two different conditions. (i) Control: 5×10⁴ cells/cm² and (ii) 5×10⁴ cells/cm² and ROCK inhibitor supplement. Black arrowheads indicate spindle-like morphology. (B) Quantification of cells attachment after 24 hours in the two conditions examined. (C) LM images of the differentiating EVCs on day 1 in two different conditions. White arrowheads indicate cobble stone-like morphology. (i) Control: 5×10⁴ cells/cm² and (ii) 5×10⁴ cells/cm² and ROCK inhibitor supplement. (D) Representative flow cytometry histogram of VEcad and PDGFRβ expression in T1D-EVCs derived in the two conditions. Scale bars are 100 μm. Significance levels were set at *p<0.05, **p<0.01, and ***p<0.001.

Figure 2. Maturation of functional T1D-ECs from T1D-EVCs

(A) Representative flow cytometry histogram for VEcad expression of magnetic sorted VEcad+ cells. (B) LM images of sorted VEcad+ and VEcad- cells that were subcultured for an additional 6 days. (C) IF images of sorted and subcultured VEcad+ cells for: VEcad, CD31, vWF, eNOS, binding of Ulex europaeus fluorescein-conjugated lectin, and uptake of acLDL. Marker in red or green as indicated on each figure panel; nuclei in blue. (D) LM image of T1D-ECs forming cord on Matrigel. (E) T1D-ECs responded to TNFα via upregulation of ICAM-1 expression. Significance levels were set at *p<0.05, **p<0.01, and ***p<0.001.

Figure 3. Three-dimensional analysis of T1D-EVC networks in both non-hypoxic and hypoxic hydrogel

Representative day 3 confocal z-stack of (A) non-hypoxic and (B) hypoxic hydrogel encapsulated T1D-EVCs analyzed with Imaris Filament Tracer. Green: Phalloidin. Red lines: T1D-EVC tubes/network that are traced and analyzed. Yellow lines: an example of a continuous T1D-EVC network in the hypoxic hydrogel. Scale bars are 100 μm. Multiple aspects of the analysis based on non-hypoxic and hypoxic hydrogel confocal z-stacks are compared and presented on bar graphs or scatter plots: (C) Total tube length; (D) Mean tube length; (E) Mean tube thickness; (F) Mean Z-distance covered in hydrogel; (G) Total tube volume; and (H) Mean tube volume. Significance levels were set at *p<0.05, **p<0.01, and ***p<0.001.

Figure 4. In vivo functionality of T1D-ECs

(A) Representative zebrafish embryo showing the integration of T1D-ECs (in red) into the zebrafish host vasculature (in green) 5-day post injection. (B) High-magnification images of (A). T1D-ECs (red) and host vasculature (green). (C) Percentage of T1D-EVCs incorporation into whole zebrafish embryos (n=12) using Imaris analysis (red bar). Percentage of incorporation in the head region (yellow bar) and trunk region (yellow checkered bar).