Endothelial progenitor cells (EPCs) are heterogeneous population of mononuclear cells (MNCs) that originate and reside in the bone marrow (BM), they are circulating in (mobilized to) adult peripheral (PB) or umbilical cord blood (UCB) [1]. EPCs have been discovered by Asahara and his coworkers in 1997 [2]. They express endothelial antigens like CD31, von Willebrand factor (vWF), endothelial nitric oxide synthase (eNOS), VE-cadherin and VEGFR2 [3, 4]. EPCs constitute 1-5% of the total BM cells and >0.0001–0.01% of PB circulating MNCs [5]. They are implicated in homeostasis, neovascularization, vascular repair, endothelial regeneration, and in angiogenesis processes [6]. There are two distinct subpopulations of EPCs. Early EPCs which give rise to heterogeneous colonies that appear in culture after 3-5 days, they are obtained by negative selection on fibronectin, they are round cells surrounded by spindle-shaped cells in morphology, they have a slow-proliferation and their in vitro growth peak is reached after 2-3 weeks [7-10]. Moreover, early EPCs do not form vascular tubes in vitro but they have a strong paracrine activity (secrete a plethora of angiogenic factors) that contributes effectively to neovascularization [11,12], they have high expression of both hematopoietic and endothelial markers (VEGFR-2, CD31, vWf, able to uptake acLDL and bind UEA-1) [13,14], they are most likely derived from hematopoietic stem cells and had resemblance to myeloid progenitors [15], and hence they are also named "Hematopoietic EPCs" [16]. Early EPCs generate the endothelial cell colony forming units (CFU-ECs) in vitro [8, 17]. Interestingly, early EPCs [18] are also termed circulating angiogenic cells (CACs) [19]. On the other hand, the other subtype of EPCs is termed "late EPCs" [18], they are more homogenous colonies that appear after 2-4 weeks in culture, they are isolated by positive selection on collagen I, they are elongated cells that form a cobblestone-morphology monolayer in vitro which is characteristic of endothelial cells, they could be maintained in culture for ~12 weeks (up to 15 passages) and they have higher proliferative & clonogenic potential compared with early EPCs [12, 17, 20]. Moreover, late EPCs could easily form tubular/capillary-like structures in vitro, they possess high vasculogenic & angiogenic potential and in vivo they could incorporate in the existing endothelium where they form stable vessels and continue to differentiate into mature endothelial cells [17,21,22]. Noteworthy, late EPCs are phenotypically similar to mature endothelium, they are present/circulate in both PB and UCB, importantly they are not only closer to endothelium phenotypically but also by exhibiting no hematopoietic (CD45) or monocyte markers (CD14 & CD115) expression in contrast to early EPCs, whereas they express many endothelial cells (ECs) antigens (CD31, VEGFR-2, CD105, CD144, CD146, vWf, CD34, higher eNOS, Tie-2, VE-cadherin, able to uptake acLDL and bind UEA-1) [22,23]. Collectively, late EPCs are termed "Non-hematopoietic EPCs" [16,24] and thus they are considered the "EPCs" subtype that comply the most with the original endothelial phenotype and function to be the legitimate endothelial progenitor cells bearing almost all of the endothelial cells characteristics [15]. Further, late EPCs generate in vitro "endothelial colony forming cells or ECFCs" [25] and they are also called "outgrowth endothelial cells or OECs"[20, 26].
There were a number of proposed combinations of surface anntigens for identifying EPCs in human, this include (but not restricted to): CD34+, CD31+, CD133+, VEGFR2+, CD144+, CD146+, CD45-/+, CD14+, VEGFR1+, FGFR1+ [16, 24, 27].
The vast variation in the surface antigens for EPCs is possibly attributed to identifying different EPCs' subpopulations at various matuartion/differentiation phases. The term "EPCs" has been haphazardly used to refer to both circulating (late EPCs) and cultured cells (ECFCs). In addition, the accumulating literature did not provide one consolidated definition of EPCs nor a specific EPCs phenotype or a unified isolation & culture protocol of them. Accordingly, different isolation techniques and culturing methods applied resulted in EPCs with various phenotypes [28]. Therefore, we aimed herein using in silico data to reach a possible novel EPCs marker or a combination of markers that could specifically characterize EPCs.
In the current manuscript, we are adding to the already ongoing efforts for the characterization analyses of EPCs by presenting a new approach for finding novel marker(s) of EPCs in peripheral blood.
Among the up-to-date ‘‘-omics’’, "gene-expression profiling" or "transcriptomics", is currently the most widely used tool for the characterization and functional analysis of cells, moreover, transcriptomics has provided a better understanding for EPCs' characterization analyses in an unbiased manner [28].
Large genomic data from large tissue sample collections are difficult to analyze; however if we use the individual transcriptomic data coming from the tissue-representing or "single-cell" level, this would render mass analysis of bulk single-cell(s) data to be fast, non-tedious [29, 30] and thus would introduce new insights about the ontogeny of new and rare cell types and the relation-ships between various cell lineages [31]. Collectively, single cell transcriptomics would help herein to improve our knowledge for the identification and characterization of EPCs in peripheral blood.
Using Gene ontology and literature survey, we assembled five groups of EPCs' molecules/factors/markers that have been specifically chosen for being of special interest and importance to the EPCs biology.
The categorization and choice of various factors were based on grouping different molecules/factors into groups involved in similar EPCs and ECs functions. The first group is involved in developmental angiogenesis, tumor angiogenesis and vascular development, this group comprises neuropilins (NRP1 & NRP2), semaphorins (3A, 3B, 3D, 3E, 3F, 4A, 4D, 5A & 6A), and VEGFR1, 2 &3 [32-35]. The second group is implicated in ECs/EPCs-immune cells interaction, proliferation, migration, survival, apoptosis, angiogenesis, immunogenicity and immune-modulation. It includes TNF-α, TNFR2/P75, TNFR1/P55 and TRAIL (Tumor Necrosis Factor Related Apoptosis Inducing Ligand) [36-40]. The third group of factors is engaged in proliferation, survival, migration and differentiation of vascular stem/progenitor cells which includes closely-related cells co-enhabiting the vascular niche; namely they are EPCs, smooth muscle cells (SMCs), pericytes & mesenchymal stem cells (MSCs). The representing candidates of this group were PDGF- (A, B & C), BMP (2,4 & 9), Wnt (1, 4, 11 & 5A), VEGF (A & C), TGF β, FGF2, IFG-1 and EGF [41-42]. Group 4 comprises microRNAs which are small, non-coding, single-stranded RNAs with regulatory activities. Recent studies showed that microRNAs play an important role in regulating EPC functions which includes proliferation, senecence, apoptosis & autophagy, mobilization & migration, tube formation & angiogenic capacity and differentiation. We have chosen representative microRNAs that could be involved in one or more biological process; the chosen candidates were microRNA-221/222, 34a, 126, 16,107, 150, 22, 21 & 130 a [43-45]. The fifth group is involved in the internalization (of ligands from the extracellular matrix to be recycled back to the endosomal compartment), endocytosis, migratory and/or invasive capacity and motility. It comprises urokinase plasminogen activator (uPA), urokinase plasminogen activator receptor(uPAR), urokinase plasminogen activator receptor associated protein(uPARAP), tissue-type plasminogen activator (tPA), Neuropilin-1 NRP1, Neuropilin-2 NRP2, VEGFR1, 2 & 3, PECAM-1, ICAM-1, VE-cadherin , Ephrin-B2, EphB4 and EGFL7 [46-57].
Herein, our main objective is to search for novel markers of EPCs in peripheral blood. Thus, we have created a short list divided into five groups of EPCs factors/molecules using pubmed literature, gene ontology, and other sources. This list was used for both the transcriptomic and single cell analyses. In transcriptome analyses, the list was used to compare the relative expression of various EPCs genes (involved within this list) between ECFCs, HUVECs and two adult ECs from skin and adipose tissue. Moreover, EPCs chosen-genes were used for functional enrichment on mouse phenotype and STRING protein-protein interaction network database to decipher the involvement of these factors in endothelial and vascular development & morphogenesis. Additionally, we built a digital matrix of healthy donors' PBMCs (33 thousand transcriptomes) and analyzed the expression of the short list of EPCs factors and more specifically EPCs molecules that have shown to be significantly regulated between ECFCs and the other three adult ECs in the transcriptome analyses.
The current study has identified novel markers, which include secreted factors, miRNAs and growth factors. Among these markers we have analyzed, some of them could be used for better cytometric analyses and an optimized characterization of EPCs sub-population in peripheral blood.