microRNA Sequencing of CD34+ Sorted Adipose Stem Cells Undergoing Endotheliogenesis

Introduction

Advancements in the fields of tissue engineering and regenerative medicine have led to the development of engineered tissues for the treatment of various musculoskeletal disorders [1–3]. Similarly, for breast cancer patients who have undergone mastectomy and for cosmetic and plastic reconstructive surgeries, autologous fat grafting is often used to achieve the desired outcome [4–7]. However, the successful integration of either engineered tissue grafts or transplanted autologous fat grafts mainly depends on the proper development and synchronization between the adipose vasculature with the surrounding arterial and venous structures.

Endotheliogenesis (differentiation of progenitors into mature endothelial cells) is the first step in the vascularization process followed by vasculogenesis (formation of new blood vessels from endothelial cells) and finally angiogenesis (growing of the blood vessel network) [8–10]. Without proper vascularization of the transplanted tissues, the cells deep inside the tissue structure lack a sufficient supply of nutrients and oxygen and eventually die as a consequence of necrosis. Therefore, understanding the process of angiogenesis is crucial for successful outcome of transplanted tissue grafts and has the potential to beneficially impact the field of tissue engineering and regenerative medicine.

In normal tissues, neovascularization is initiated by the endothelial progenitor cells. Nevertheless, it is cumbersome to access large numbers of these cells to construct and integrate the vasculature of engineered tissues since their major source is the bone marrow, a depot with limited accessibility and volume. Alternatively, endothelial progenitors expressing CD34+ and CD31− can be isolated from the stromal vascular fraction (SVF) of adipose tissue [11]. This particular cell subpopulation within the SVF has the potential to differentiate along the endothelial lineage when exposed to vascular endothelial growth factors present within serum [11–13].

Adipose derived stromal/stem cells (ASCs) are another subpopulation from SVF that are multipotent and can be differentiated into adipogenic, chondrogenic, and osteogenic lineages [14,15]. Due to their multipotent nature, higher yield quantities, and ease of isolation relative to other adult stromal/stem cells, ASCs have become prominent tools in the field of tissue engineering. However, induction of endothelial differentiation of ASCs requires subjecting them to physiological shear force (to mimic the natural blood flow in capillaries) in the presence of endothelial growth supplements, vascular endothelial growth factors (VEGFs) [16,17], and/or use of complex three-dimensional cell culture systems such as spheroid culture, polyglycolic acid/polylactic acid scaffolds [18,19].

MicroRNAs (miRNAs) are short noncoding RNA nucleotides which regulate various physiological and pathological cellular mechanisms by binding to mRNA sequences and blocking their translation into proteins [20,21]. Cellular differentiation is one of the vital processes that is regulated by miRNAs. For example, studies of adipose and/or bone marrow stromal/stem cells have determined that the upregulation of miRNA 148b [22] and miR-146a [23] results in the induction of osteogenesis, upregulation of miR-150 [24] in adipogenesis, upregulation of miR-410 [25] and miR-218 [26] in chondrogenesis, and downregulation of miR-145 [27] in endotheliogenesis. However, few if any miRNAs have been reported to induce endothelial differentiation in ASCs.

In this article, we used next-generation sequencing method to sequence and identify the miRNAs expressed in CD34+ sorted ASCs during the period of endothelial differentiation at different time points. Using bioinformatic analytical techniques, we identified several miRNAs (both upregulated and downregulated) as potential candidates to induce endotheliogenesis in ASCs. The goal of this article is to present the bioinformatic findings of differentially expressed miRNAs on various time points during the period of endotheliogenesis so that it can serve as an open repository source. In vitro testing of miRNAs identified on day 3 was performed by transfecting CD34+ ASCs with either miRNA mimics (for upregulation) or anti-miRNAs (to downregulate) of the candidate miRNAs identified by Database for Annotation, Visualization and Integrated Discovery (DAVID) bioinformatics tools.

Methods

Magnetic bead sorting of CD34+ cells

All protocols were reviewed and approved by the Louisiana State University Institutional Research Board (protocols E9379 and E9119). Cryopreserved ASCs from three different donors originally harvested at P0 (Obatala Sciences, Inc., New Orleans, LA) were thawed, tested for viability, counted, and pooled together in equal amounts. The sorting of CD34+ cells from ASCs was done using CD34 MicroBead Kit (Miltenyi Biotec) according to the manufacturer's instructions. Briefly, the ASC pellet was suspended in a 300 μL of buffer (pH 7.2) containing 0.5% bovine serum albumin (BSA) and 2 mM ethylenediaminetetraaceticacid. This was followed by the addition of 100 μL of FcR blocking reagent and 100 μL of CD34 microbeads, mixed well, and incubated at 4°C for 30 min. The phycoerythrin (PE) conjugated CD34 antibody (BD Biosciences) was added to this mixture, incubated for another 5 min at 4°C, and subsequently used to characterize the CD34+ population by flow cytometry. After incubation the cells were washed and resuspended in buffer, which is then added to the magnetic separation column that is on the magnetic stand. The flow through the column contains CD34 depleted ASCs, which was collected. Later the column was removed from the magnetic stand, and more buffer was added to the column. The CD34+ cells were flushed, and the numbers of CD34+ and CD34 depleted ASCs were counted by hemocytometer. The sorted cells were then plated in six-well plates at a density of 8,000 cells per cm² and cultured till 80%–90% confluence was obtained in basal media, which is composed of 10% fetal bovine serum (FBS), 1% antibiotics in dulbecco's modified eagle medium (DMEM)/F12 media.

CD34 surface marker characterization

The CD34+ sorted cells and CD34 depleted ASCs were stained with CD34-PE antibody during the magnetic bead sorting process as explained in the above section. Both the sorted cell populations were analyzed by flow cytometry for the expression of CD34 antibody (Supplementary Fig. S1).

Induction of endothelial differentiation

Upon reaching confluence the CD34+ cells were treated with endothelial medium containing 50 ng/mL of VEGF, 2% FBS, and 1% antibiotic in DMEM/F12 media. The CD34+ cells and CD34 depleted ASCs treated with basal medium served as controls. The media was replaced every 3 days for up to 14 days.

Endothelial staining

To confirm endothelial genesis, cells were stained using UEA I lectin conjugated with fluorescein isothiocyanate (FITC) or Von Willebrand Factor (VWF) primary antibody (Abcam; ab6994) and a secondary antibody conjugated with Alexa Fluor 488 (Abcam; ab150077). Briefly, the cells on days 3, 9, and 14 of culture were washed with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde for 20 min at room temperature. Later, they were permeabilized with 0.1% Triton X-100 for 10 min at room temperature and washed with cold PBS thrice. Unspecific binding sites were blocked with 1% BSA in phosphate-buffered saline with Tween detergent for 30 min at room temperature. UEA I lectin FITC was then added to the cells in the dilution of 1:200 and incubated for 1 h at 4°C in the dark. After this step the cells were washed several times with PBS and imaged using fluorescence microscope. For VWF staining, VWF antibody was then added to the cells in the dilution of 1:400 and incubated for 1 h at 4°C in dark. After this step the cells were washed several times with PBS, and secondary antibody conjugated with Alexa Fluor 488 was added in a dilution of 1:200 in 1% BSA before incubating for 1 h. The cells were later washed several times with PBS. For counter staining, 4′,6-diamidino-2-phenylindole stain was added to the cells, incubated for 5 min, washed with PBS, and imaged using fluorescence microscope.

miRNA isolation and quantitation

miRNAs on the respective days were isolated using PureLink miRNA Isolation Kit (Thermo Fisher Scientific; K157001) following manufacturer's instructions. The quality and purity of miRNAs were analyzed using fragment analyzer (Advanced Analytical Technologies, Inc.) (Supplementary Fig. S2).

miRNA library preparation and sequencing

cDNA libraries were constructed using the Ion Total RNA-Seq Kit v2 (Catalog number: 4475936), protocol for the small RNA libraries. Templates for RNASeq were prepared with the Ion PI™ Hi-Q OT2 200 Template Kit (A26434) using the Ion OneTouch™ 2 System for Ion Proton™ System semiconductor sequencing. These templates were sequenced using the Ion PI Hi-Q Sequencing 200 Kit (A26433) and Ion PI Chip v3 on the Ion Proton Sequencer. Preliminary analysis of the ensuing sequencing data pertaining to quality of the run, read lengths, and coverage was performed using the torrent suite software.

miRNA-Seq data alignment and data analysis

The FastQ files obtained from sequencing were used to generate output alignments translated into transcript coordinates by STAR program that were later used as an input for RSEM program to find expression [28]. The output bam files were sorted and visualized using Integrated Genomics Viewer to check for proper alignment [29]. Later, bedtools program was used to extract the miRNA data by interspersing the bam files with mature human miRNA sequences from miRbase.org The miRNAs with low counts <25 were filtered out, and the differential expression was calculated using edgeR program in R [30]. The miRNAs with P values <0.05 were considered significant for further analysis. The miRNAs with fold change (FC) above 1.2 or below 0.8 are only considered as either upregulated or downregulated. The data discussed in this publication have been deposited in National Center for Biotechnology Information's Gene Expression Omnibus (GEO) [31] and are accessible through GEO Series accession number GSE151561(https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE151561)

Kyoto Encyclopedia of Genes and Genomes analysis of experimentally validated miRNA targets

Experimentally validated miRNA targets for each differentially expressed miRNA at various time points were obtained through miRWalk [32,33]. miRWalk algorithm finds out the experimentally validated published interactions on miRNA-gene targets. To obtain the miRNA targets, the mirbase database was selected in the miRWalk program. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis was performed on these miRNA targeted genes using DAVID bioinformatics [34,35]. The miRNA-gene targets and the respective pathways are presented in Supplementary Table S4.

miRNAs or anti-MiRNA mimics transfection

The sequences for miRNA and anti-miRNA mimics, listed in Supplementary Tables S1 and S2, were purchased from IDTdna Technologies. Transfection of these was transfected into CD34+ ASCs at concentrations of 400, 800, 1,600, and 3,200 nM individually using RNAiMAX transfecting reagent as per manufacturer's guidelines (Thermo Fisher Scientific; 13778030). After 24 h of transfection the media was replaced and cells were cultured for 14 days at which time the cells were stained with VWF antibody as explained above in the “Endothelial staining” section and examined under fluorescence microscopy.

RNA isolation, reverse transcription, and quantitative real time–polymerase chain reaction

RNA was isolated using the PureLink RNA Kit (Life Technologies, Carlsbad, CA) according to the manufacturer's instructions on day 14 for adipogenic samples and on day 21 for osteogenic samples. The quality and quantity of isolated RNA were measured using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA). The first strand cDNA synthesis was done using a High Capacity cDNA Synthesis Kit (Applied Biosciences, Beverly Hills, CA). For quantitative real time–polymerase chain reaction (qRT-PCR), a SYBR Green (Applied Biosystems) Kit was used as per the manufacturer's instructions for the ABI-7900 qRT-PCR machine. The primer sequences used in this study are listed in Supplementary Table S3. The FC in gene expression was normalized to glyceraldehyde 3-phosphate dehydrogenase and FC calculated by 2^ΔΔCt method [36].

Results

Endothelial differentiation

The sorted CD34+ ASCs were cultured until reaching 90% confluence and then treated with endothelial media containing 50 ng/mL of VEGF for 14 days. Positive fluorescence staining of UEA I lectin was observed on days 9 and 14 with CD34+ ASCs treated with endothelial media, whereas no staining was observed in the control samples (treated with basal media), as well as samples on day 3 (Fig. 1). The cells imaged on day 3 did not show any fluorescence. The qRT-PCR evaluation of endothelial gene biomarkers [endothelial nitric oxide synthase (eNOS), VWF, and CD31] on days 3, 9, and 14 indicated an increase in expression (Fig. 2). The expression of endothelial genes was found to be increasing from day 3 and 9 with a maximum on day 14. Thus, UEA I lectin fluorescence staining and qRT-PCR endothelial gene evaluation studies indicated the differentiation of CD34+ ASCs along the endothelial pathway.

miRNA expression by miRNA-Seq

Heat maps indicated the expression of miRNAs on day 3, 9, and 14 with P values <0.05 at one or more time points (Figs. 3 and 4). The top five upregulated miRNAs on day 3 are miRs 330-5p FC (FC 2.16), let7c-3p (FC 2.13), 15b-5p (FC 1.9), 99a-5p (FC 1.86), and 335-5p (FC 1.76), while top five downregulated miRNAs are miRs 342-5p (FC 0.43), 32-3p (FC 0.48), 455-3p (FC 0.51), 32-5p (FC 0.54), and let7a-2-3p (FC 0.56). Similarly for day 9, the upregulated miRNAs are 190a-5p (FC 6.16), 126-3p (FC 4.84), 1301-3p (FC 4.80), 195-5p (FC 4.55), and let7c-5p (FC 4.46), and the downregulated ones are miRNAs 181a-2-3p (FC 0.19), 4638-3p (FC 0.24), 574-5p (FC 0.25), let-7a-3p (FC 0.38), and let-7e-3p (FC 0.39). On day 14, the miRNAs 126-3p (FC 11.86), 204-5p (FC 8.41), 675-5p (FC 3.89), 99a-5p (FC 3.27), and 335-5p (FC 3.04) are upregulated, and the miRNAs 143-5p (FC 0.32), 342-5p (FC 0.36), 155-5p (FC 0.36), 146a-5p (0.41), and 145-5p (FC 0.42) are downregulated. Furthermore, the miRNAs shared in common across all time points were displayed using Venn diagrams (Fig. 5; Tables 1 and 2).

KEGG analysis of experimentally validated miRNA targets

All the miRNA targets that are used for the KEGG analysis are presented in a supplemental excel file in multiple worksheets. The enriched pathways modulated by the miRNAs at each time point were determined (Figs. 6–8). The P-values and number of genes within each pathway targeted by the miRNA are shown in Tables 1 and 2. The enriched pathways found from KEGG analysis included ErbB signaling, NF-KB signaling, Notch signaling, hypoxia inducible factor (HIF) signaling, phosphoinositide 3-kinase (PI3K)-AKT signaling, P53 signaling, VEGF signaling, Hippo signaling, Ras signaling, Rap1 signaling, mTOR signaling, tumor necrosis factor (TNF) signaling, transforming growth factor-beta (TGF-β) signaling, Insulin signaling, mitogen-activated protein kinase (MAPK) signaling, cGMP-PKG signaling, and cAMP signaling (Figs. 6–8; Tables 3–8). The genes associated with the specific miRNAs at various time points are shown in the Supplementary Table S4.

Experimental miRNA and anti-miRNA validation

Based on the validated miRNA targets and their KEGG analysis of all the downregulated miRNAs, six downregulated miRNAs with antiangiogenic properties were selected as potential candidates to induce endothelial genesis: miR-145-5p [37–39], miR-155-5p [40], miR-193a-3p [41], miR-125a-5p [42–44], miR-221-5p [45,46], and miR-222-3p [45,46]. Supporting evidence that these miRNAs are antiangiogenic can be found in the cancer literature where they have each been associated with suppression of angiogenesis. The following preliminary experimental studies tested the hypothesis that downregulation of the abovementioned miRNAs in the CD34+ ASCs by transfection with their respective antagomirs could result in the induction of endothelial genesis.

Similarly, six upregulated miRNAs (miR-181a-5p, miR-330-5p, miR-335-3p, miR-15b-5p, miR-99a-5p, and miR-199a-5p) based on the KEGG analysis of all the upregulated miRNAs were chosen to induce endotheliogenesis in ASCs. The following studies tested the hypothesis that upregulation of the abovementioned miRNAs by transfection will promote endotheliogenesis. The initial staining studies indicate that cells treated with miR-335-5p (400 nM), miR-330-5p (800 nM), miR-181a-5p (400, 800, and 1,600 nM), anti-miR-125a-5p (1,600 and 3,200 nM), anti-miR-145-5p (3,200 nM), and anti-miR-155-5p (1,600 nM) mimics showed extensive VWF staining (Supplementary Figs. S3 and S4). Therefore, these miRNAs and anti-miRNA mimics are selected for further conformational studies. Each of these miRNAs and anti-miRNA mimics are transfected into cells in 12-well plates at the respective concentrations as mentioned above and imaged. In addition, mRNA was isolated from all the samples, and qRT-PCR was performed for the expression of endothelial genes (VEGF, CD31, and VWF). Robust fluorescence imaging of VWF antibody staining was observed in the samples transfected with miR-330-5p (800 nM), miR-181a-5p (1,600 nM), anti-miR-125a-5p (1,600 nM), anti-miR-125a-5p (3,200 nM), and anti-miR-145-5p (3,200 nM) mimics (Fig. 9). qRT-PCR analysis determined that cells transfected with miR-335-5p (400 nM), miR-330-5p (800 nM), and miR-181a-5p (400 nM) mimics induced the expression of all three endothelial associated genes (VEGF, CD31, and VWF), while miR-181a-5p mimic (800 nM) upregulated CD31 and VWF genes. Highest gene expression of all the three endothelial genes was induced by miR-181a-5p mimic (400 nM) (Fig. 10). However, transfection with anti-miR mimics did not significantly induce the expression of the endothelial biomarker genes. Interestingly, the anti-miR mimics, anti-miR-125a-5p (at 1,600 and 3,200 nM), and anti-miR-145-5p (3,200 nM) resulted in robust positive staining for VWF (Fig. 9).

Discussion

Since their discovery in Caenorhabditis elegans in 1993, the biogenesis of miRNAs and their functional role in various cellular and physiological processes have been extensively studied [47,48]. Each of these small noncoding RNAs regulate the transcription and translational processes, thus deciding the fate of cellular differentiation either individually or together in clusters [49,50]. Furthermore, a single miRNA can target multiple gene transcripts or a single gene can be targeted by multiple miRNAs or multiple genes targeted by miRNA clusters [49,50]. Therefore, identifying the miRNAs and their targets and the associated mechanistic pathways provides deeper insights in the cellular differentiation process.

Endothelial differentiation is the primary step in the angiogenesis process. This formation and expansion of new blood vessel network is a double-edged sword that can be either beneficial or insidious. For example, in case of novel tissue engineered constructs or autologous transplantations, it is extremely crucial for the transplanted grafts to develop new vasculature for the proper integration within the host. In contrast, new blood vessel formation facilitates the aggressive growth and expansion of cancerous tissues. Similarly, angiogenesis is also associated with the obesity where the adipose tissue stimulates the new blood vessel formation by secreting various growth factors and proangiogenic cytokines and chemokines [51]. Furthermore, the stromal microenvironment in the fat tissues in obese patients promotes angiogenesis and, consequently, provides a growth promoting environment for tumors and cancers [52].

Studying what miRNAs are expressed and how they regulate the endothelial differentiation and subsequent angiogenic pathways provides deeper insights that can be used to better understand and develop new therapeutical strategies for tissue engineering applications, cancer, and obesity. In this study, we used next generation sequencing, a high throughput method, to identify the miRNAs and targeted pathways that are associated with the endothelial differentiation of CD34+ ASCs at different time points. In addition these identified miRNAs are evaluated for their efficacy to induce endothelial differentiation in vitro.

miRNA-based sequencing approach was used to identify pathways associated with endothelial differentiation by human ASC in a temporal manner. On day 3 of exposure to endotheliogenesis factors, human ASC displayed 22 upregulated and 18 downregulated miRNAs (Figs. 2 and 3). The target mRNAs for these miRNAs are identified using miRWalk and subjected to KEGG analysis to identify the enriched pathways that involve these mRNAs. Based on the obtained results, we chose six upregulated (miR-181a-5p, miR-330-5p, miR-335-3p, miR-15b-5p, miR-99a-5p, and miR-199a-5p) and six downregulated miRNAs (miR-145-5p, miR-155-5p, miR-193a-3p, miR-125a-5p, miR-221-5p, and miR-222-3p) with the targeted mRNAs associated in the several regulatory pathways in endotheliogenesis. The enriched pathways are MAPK signaling, PI3K-AKT signaling, VEGF signaling, ErbB signaling, NF-KB signaling, Notch signaling, HIF-signaling, P53 signaling, Hippo signaling, Ras signaling, Rap1 signaling, mammalian target of rapamycin (mTOR) signaling, TNF signaling, TGF-β signaling, Insulin signaling, cGMP-PKG signaling, and cAMP signaling (Tables 3 and 4).

The MAPK pathways and PI3K pathway play a vital role in the induction and maintenance of endothelial cell differentiation of human-induced pluripotent stem cells through mediation of ETS-related Gene (ERG) and friend leukemia integration 1 transcription factor (FLI1) [53]. A critical proangiogenic factor, VEGF, promotes endothelial progenitor survival, proliferation, and invasion by activating MAPK and PI3K pathways through VEGFR-2 binding [54]. From our selected list of miRNAs for the initial in vitro study, the day 3 upregulated miRNAs 181a-5p, 335-3p, 15b-5p and the day 3 downregulated miRNAs 145-5p, 155-5p, 193a-3p, 125a-5p, 222-3p are found to be targeting MAPK and PI3K-AKT pathways by KEGG analysis (Tables 3 and 4).

The overexpression of miR-181a-5p has been shown by others to enhance tumor angiogenesis through targeting of the SRC kinase signaling inhibitor 1 (SRCIN1) in colorectal cancer [55]. It was also reported that expression of miR-181a induces the progression of ovarian cancer by facilitating the epithelial to mesenchymal transition through TGF-β signaling pathway [56]. In chondrosarcoma, the upregulation of miR-181a was found to be positively correlated with the increased expression of VEGF and matrix metalloprotease 1 (MMP1), thereby increasing angiogenesis [57]. Furthermore, miR-181a repressed RGS16 gene, which was a negative regulator of CXC chemokine receptor 4 (CXCR4) signaling associated with enhanced VEGF expression [57]. However, in MDA-MB-231 cancerous cells, miR-181a-5p attenuated the metastasis and angiogenesis by repressing the expression of MMP-14 [58]. Thus, the literature indicates that the role of miR-181a-5p in promoting or inhibiting angiogenesis is cell specific. Our current study noted increased expression of miR-181a-5p on day 3 during ASC endothelial differentiation suggesting a possible role in positively regulating endothelial genesis. KEGG analysis revealed that the enriched pathways associated with the miR-181a-5p targeted genes are known to play regulatory roles in endothelial and angiogenic differentiation. Consistent with this, transfection of miR-181a-5p at 400 nM resulted in the upregulation of VWF, VEGF, and CD31 (Fig. 10). Furthermore, increasing the concentration to 1,600 nM resulted in more robust VWF antibody staining (Fig. 9).

In the current study, miR-125a-5p was identified as downregulated using miRNA sequencing of ASCs undergoing endothelial genesis, and knockdown of miR-125a-5p expression using anti-miR-125a-5p (1,600 and 3,200 nM) resulted in extensive VWF antibody staining (Fig. 9). The miR-125a-5p has been recognized as a tumor suppressor. In breast cancer [59], ovarian cancer [60], lung cancer [61], and medulloblastoma [62], it was found that the expression of miR-125a-5p was extremely low. This low expression of miR-125a-5p correlated with the overexpression of proangiogenic factor VEGF, thereby promoting both proliferation and angiogenesis [42,43]. Conversely, Dai et al., showed that upregulation of miR-125a caused the downregulation of VEGF-A and reduced progression of gastric cancer [42]. Che et al., found that endothelial cells from older mice had increased expression of miR-125a-5p associated with defective angiogenesis, whereas endothelial cells from young mice had higher expression levels of miR-125a-5p [43]. Furthermore, inhibition of miR-125a-5p in endothelial cells from older mice significantly improved the angiogenic potential [43].

While the functional role of miR-335-5p in enhancing osteogenesis [63–65] and inhibiting the metastasis and invasion of gastric cancer has been documented in the literature [66,67], its direct role in regulating angiogenesis is less well explored. The current study determined that expression of miR-335-5p was increased on day 3 suggesting its positive role in endothelial genesis regulation, and transfection of ASCs with miR-335-5p at 400 nM concentration resulted in the upregulation of VWF and CD31 genes (Fig. 10) and similarly resulted in VWF antibody staining (Fig. 9).

While the role of miR-330-5p in the angiogenesis remains relatively under explored, it has been found to promote metastasis in hepatocellular carcinoma [68]. Interestingly, in human umbilical vein endothelial cells, miR-330-5p's expression was decreased upon exposure to angiopoietin-1, and upregulation of miR-330-5p resulted in the reduction of capillary-like tube formation [69]. The current study found that miR-330-5p was upregulated during endothelial differentiation of ASCs, and when miR-330-5p was transfected, there was both upregulation of VWF gene expression and VWF antibody staining. We found that transfection of miR-33-5p at 800 nM resulted in higher endothelial differentiation in comparison to the miRNA concentration at 400, 1,600, and 3,200 nM (Supplementary Fig. S3). This possibly suggests that miR-330-5p may have the dual activity as the higher dosage (above 1,600 nM) of miR-330-5p might be downregulating while the lower dosage could have a positive effect on endothelial differentiation. However, further studies are required to test the dose versus functional relationship of the miR-330-5p.

In breast, colon, glioma, ovary, prostate, and pancreatic cancers, miRNA-145 expression has been found to be suppressed [70]. Upregulation of miR-145 inhibited the metastasis and angiogenesis of osteosarcoma cells by downregulating VEGF expression [37]. Several studies have shown that miR-145 downregulates VEGF and Ras, thereby impairing angiogenesis [38,71]. miRNA-145 also targeted HIF-2α in neuroblastoma models, thereby inhibiting angiogenesis and metastasis [72]. Moreover, miRNA-145 downregulation using the anti-miR-145 in ASCs induced endothelial differentiation and promoted the angiogenesis by activation of endothelial transcription factor ETS1 [27]. In the current study, miR-145-5p knockdown using anti-miR-145-5p at 3,200 nM concentration resulted in extensive VWF antibody staining, in agreement with the published literature.

In the current study, in vitro analysis of the miRNAs that are upregulated or downregulated on the day 3 of endothelial differentiation is evaluated by qRT-PCR and VWF staining. Additional experimental validation studies such as matrigel assay (a gold standard to test the sprouting behavior endothelial cells) and western blotting analysis to measure the expression of endothelial proteins will be conducted in future studies to further strengthen our findings that these miRNAs or anti-miRNAs induce endotheliogenesis in ASCs. The miRNAs from day 9 to 14 can also be explored for future studies to evaluate their potential for inducing endothelial genesis. One possible strategy is to temporally transfect the miRNAs and/or anti-miRNAs expressed on different days during the course of differentiation at discrete time points to get an efficient differentiation outcome.

Conclusion

We have presented the expression of differentially expressed miRNAs identified by miRNA sequencing during the endotheliogenesis of CD34+ ASCs. The gene targets of these miRNAs were identified using miRWalk program using mirbase database, and the KEGG analysis using DAVID bioinformatics on these miRNA targets revealed the enrichment of MAPK signaling, PI3K-AKT signaling, VEGF signaling, ErbB signaling, NF-KB signaling, Notch signaling, HIF-signaling, P53 signaling, Hippo signaling, Ras signaling, Rap1 signaling, mTOR signaling, TNF signaling, TGF-β signaling, Insulin signaling, cGMP-PKG signaling, and cAMP signaling pathways. Our article is primarily aimed to serve as an open repository source of the miRNA sequencing findings of the differential temporal expression of miRNAs during endotheliogenesis of CD34+ ASCs. In vitro experimental validation of miRNAs identified on day 3 was performed by transfecting ASCs with either miRNA mimics (for upregulation) or anti-miRNAs (to downregulate) and studied by VWF staining and qRT-PCR techniques. Transfection of these day 3 miRNAs with most gene targets in the endothelial regulatory pathways showed that the overexpression of miR-335-5p, miR-330-5p, miR-181a-5p and downregulation of miR-125a-5p and mir-145-5p suggested the likelihood of the differentiation of CD34+ sorted ASCs into endothelial lineage.

Disclaimer

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Author Disclosure Statement

During the study period, J.G. was employed at LaCell, LLC and Obatala Sciences, a for-profit biotech companies cofounded and co-owned by Dr. J.G. The remaining coauthors have no conflicts to disclose.

Funding Information

Research reported in this publication was supported by the National Institute Diabetes, Digestive and Kidney Diseases of the National Institutes of Health under award R21 DK 91852 and by the National Institute of Dental and Craniofacial Research of the National Institutes of Health under award R01 DE024790.

Supplementary Material

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4