Improving the characterization of endothelial progenitor cell subsets by an optimized FACS protocol

The characterization of circulating endothelial progenitor cells (EPCs) is fundamental to any study related to angiogenesis. Unfortunately, current literature lacks consistency in the definition of EPC subsets due to variations in isolation strategies and inconsistencies in the use of lineage markers. Here we address critical points in the identification of hematopoietic progenitor cells (HPCs), circulating endothelial cells (CECs), and culture-generated outgrowth endothelial cells (OECs) from blood samples of healthy adults (AB) and umbilical cord (UCB). Peripheral blood mononuclear cells (PBMCs) were enriched using a Ficoll-based gradient followed by an optimized staining and gating strategy to enrich for the target cells. Sorted EPC populations were subjected to RT-PCR for tracing the expression of markers beyond the limits of cell surface-based immunophenotyping. Using CD34, CD133 and c-kit staining, combined with FSC and SSC, we succeeded in the accurate and reproducible identification of four HPC subgroups and found significant differences in the respective populations in AB vs. UCB. Co-expression analysis of endothelial markers on HPCs revealed a complex pattern characterized by various subpopulations. CECs were identified by using CD34, KDR, CD45, and additional endothelial markers, and were subdivided according to their apoptotic state and expression of c-kit. Comparison of UCB-CECs vs. AB-CECs revealed significant differences in CD34 and KDR levels. OECs were grown from PBMC-fractions We found that viable c-kit+ CECs are a candidate circulating precursor for CECs. RT-PCR to angiogenic factors and receptors revealed that all EPC subsets expressed angiogenesis-related molecules. Taken together, the improvements in immunophenotyping and gating strategies resulted in accurate identification and comparison of better defined cell populations in a single procedure.

• View at Publisher
• View at Scopus

Free Full Text ( Final Version , 9mb )