Characterization of Endothelial Progenitor Cell Interactions with Human Tropoelastin

Abstract

The deployment of endovascular implants such as stents in the treatment of cardiovascular disease damages the vascular endothelium, increasing the risk of thrombosis and promoting neointimal hyperplasia. The rapid restoration of a functional endothelium is known to reduce these complications. Circulating endothelial progenitor cells (EPCs) are increasingly recognized as important contributors to device re-endothelialization. Extracellular matrix proteins prominent in the vessel wall may enhance EPC-directed re-endothelialization. We examined attachment, spreading and proliferation on recombinant human tropoelastin (rhTE) and investigated the mechanism and site of interaction. EPCs attached and spread on rhTE in a dose dependent manner, reaching a maximal level of 56±3% and 54±3%, respectively. EPC proliferation on rhTE was comparable to vitronectin, fibronectin and collagen. EDTA, but not heparan sulfate or lactose, reduced EPC attachment by 81±3%, while full attachment was recovered after add-back of manganese, inferring a classical integrin-mediated interaction. Integrin αVβ3 blocking antibodies decreased EPC adhesion and spreading on rhTE by 39±3% and 56±10% respectively, demonstrating a large contribution from this specific integrin. Attachment of EPCs on N-terminal rhTE constructs N25 and N18 accounted for most of this interaction, accompanied by comparable spreading. In contrast, attachment and spreading on N10 was negligible. αVβ3 blocking antibodies reduced EPC spreading on both N25 and N18 by 45±4% and 42±14%, respectively. In conclusion, rhTE supports EPC binding via an integrin mechanism involving αVβ3. N25 and N18, but not N10 constructs of rhTE contribute to EPC binding. The regulation of EPC activity by rhTE may have implications for modulation of the vascular biocompatibility of endovascular implants.

Fig 1. Schematic diagram showing full-length rhTE and constructs N18, N25 and N10.

Key domain features are indicated.

Fig 2. OEC characterization by flow cytometry.

A) Stained cells are shown as blue histograms, while unstained controls are shown in black. The percentage of positive cells is shown in the top right of each graph. The OECs are CD34/31/54/VEGFR2 positive and CD45/14 negative. B) Representative images of the binding of isothiocyanate-Ulex europaeus agglutinin I lectin binding (ULEX), uptake of acetylated low density lipoprotein (AcLDL) and staining for CD31 by EPCs (bottom row of panel) but not by fibroblasts (top row of panel). Together, these results are indicative of a positive endothelial cell phenotype.

Fig 3. Cell binding to recombinant human tropoelastin.

(A) Relative attachment of human dermal fibrolasts, EPCs and human coronary artery smooth muscle cells (SMC) to increasing concentrations of tropoelastin. (B) The percentage of spread EPCs on increasing concentrations of tropoelastin. (C) Phase contrast microscopy of spreading EPCs on BSA-blocked wells, tropoelastin (rhTE) collagen (CN) fibronectin (FN). Images were taken at 10x magnification. (D) EPC proliferation on days 3 and 5, respectively. Error bars represent S.E.M. of triplicate measurements.

Fig 4. Mechanism of EPC attachment to rhTE.

(A) EPCs attached to 40 μg/ml rhTE in the presence of α-lactose, β-lactose, heparan sulfate, or EDTA. (B) Attachment of EPCs to 40 μg/ml rhTE in the presence of Ca²⁺, Mg²⁺, or Mn²⁺. (C) and (D) Inhibition of EPC attachment and spreading on 40 μg/ml rhTE using antibodies to integrins α₂β₁, α₅β₁, and α_vβ₃. Error bars represent S.E.M. of triplicate measurements.

Fig 5. EPC attachment (A) and spreading (B) on tropoelastin constructs N25 and N18.

C) and (D) Proliferation of EPCs on N25 and N18 constructs, on days 3 and 5, respectively. Values were normalized to proliferation on collagen. Error bars represent S.E.M. of triplicate measurements. EPC attachment (E) and spreading (F) to tropoelastin construct N10. Attachment and spreading assays on N10 constructs was done on the same cell plates as for N25 and N18 constructs. Results are expressed separately to highlight higher maximal concentrations of N10 used. Error bars represent S.E.M. of triplicate measurements.

Fig 6. Mechanism of EPC attachment to truncated tropoelastin constructs.

(A) and (B) EPCs attached to 40 μg/ml N25 and N18 respectively, in the presence of α-lactose, β-lactose, heparan sulfate or EDTA. (C) and (D) Attachment of EPCs to 40 μg/ml N25 and N18, respectively, in the presence of Ca²⁺, Mg²⁺, or Mn²⁺. (E) and (F), Inhibition of EPC spreading on 40 μg/ml N25 and N18 respectively, using antibody that inhibits binding to integrin α_vβ₃. Error bars represent S.E.M. of triplicate measurements.