Follistatin‐Like 3 Enhances the Function of Endothelial Cells Derived from Pluripotent Stem Cells by Facilitating β‐Catenin Nuclear Translocation Through Inhibition of Glycogen Synthase Kinase‐3β Activity

Abstract The fight against vascular disease requires functional endothelial cells (ECs) which could be provided by differentiation of induced Pluripotent Stem Cells (iPS Cells) in great numbers for use in the clinic. However, the great promise of the generated ECs (iPS‐ECs) in therapy is often restricted due to the challenge in iPS‐ECs preserving their phenotype and function. We identified that Follistatin‐Like 3 (FSTL3) is highly expressed in iPS‐ECs, and, as such, we sought to clarify its possible role in retaining and improving iPS‐ECs function and phenotype, which are crucial in increasing the cells’ potential as a therapeutic tool. We overexpressed FSTL3 in iPS‐ECs and found that FSTL3 could induce and enhance endothelial features by facilitating β‐catenin nuclear translocation through inhibition of glycogen synthase kinase‐3β activity and induction of Endothelin‐1. The angiogenic potential of FSTL3 was also confirmed both in vitro and in vivo. When iPS‐ECs overexpressing FSTL3 were subcutaneously injected in in vivo angiogenic model or intramuscularly injected in a hind limb ischemia NOD.CB17‐Prkdcscid/NcrCrl SCID mice model, FSTL3 significantly induced angiogenesis and blood flow recovery, respectively. This study, for the first time, demonstrates that FSTL3 can greatly enhance the function and maturity of iPS‐ECs. It advances our understanding of iPS‐ECs and identifies a novel pathway that can be applied in cell therapy. These findings could therefore help improve efficiency and generation of therapeutically relevant numbers of ECs for use in patient‐specific cell‐based therapies. In addition, it can be particularly useful toward the treatment of vascular diseases instigated by EC dysfunction. Stem Cells 2018;36:1033–1044


Immunofluorescence staining
The procedure used for immunofluorescent staining was similar to that described previously 2 . Briefly, cells were fixed with 4% paraformaldehyde or cold methanol and permeabilised with 0.1% Triton X-100 in PBS for 10 minutes and blocked in 5% goat or donkey serum in PBS for 30 minutes at 37˚C. The cells were incubated with primary antibodies for 1 hour at 37˚C. The bound primary antibody was revealed by incubation with the secondary antibody; anti-mouse Alexa 488, and anti-rabbit Alexa 488, anti-rabbit Alexa 568, anti-goat Alexa 568, at 37˚C for 45 min. Cells were counterstained with 4',6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich), mounted in Vectashield (Vector Laboratories, Inc. USA), and examined with a fluorescence microscope (Axioplan 2 imaging; Zeiss) or SP5 confocal microscope (Leica, Germany).

Immunoblotting
The method used was similar to that described previously 2 . Cells were harvested and washed with cold PBS, re-suspended in lysis buffer (25mM Tris-Cl pH 7.5, 120mM NaCl, 1 mM EDTA pH 8.0, 0.5% Triton X100) supplemented with protease inhibitors (Roche) and lysed by ultra-sonication (twice, 6 seconds each) (Bradson Sonifier150) to obtain whole cell lysate. The protein concentration was determined using the Biorad Protein Assay Reagent.
50 μg of whole lysate was applied to SDS-PAGE and transferred to Hybond PVDF membrane (GE Health), followed by standard western blot procedure. The bound primary antibodies were detected by the use of horseradish peroxidase (HRP)-conjugated secondary antibody and the ECL detection system (GE Health). The densitometry analysis of the bands in the western blots was done using the software Image J by National Institutes of Health (NIH).

Luciferase Reporter Assay
For the luciferase reporter assays, iPS-ECs were seeded into 12-well plates and cotransfected with EX-FSTL3, and control plasmids with the FSTL3 human promoter reporter (217HPRM16010-PG02) (Genecopoeia), or TopFlash promoter 2 . Briefly, 0.33 μg/well of the reporter plasmids were co-transfected with EX-FSTL3, EX-OCT4, EX-KLF4 3 and controls; EX-GFP or EX-mCherry (0.17 μg/well) using Fugene 6 or Endofectin Max according to the manufacturer's protocol. pGL3-Luc Renilla (0.1μg/well) was included in all transfection assays as an internal control. Luciferase and Renilla (Promega) activity assays were detected 48 hours after transfection using a standard protocol 3 . Relative luciferase units (RLU) were defined as the ratio of luciferase activity to Renilla activity with that of control set as 1.0.

FSTL3 treatment
iPS-ECs were treated by addition of human recombinant FSTL3 (25-50 ng/ml) in the cell culture media and the cells were harvested 48 hours later and subjected to further analysis.

Angiogenesis Array
Cell culture supernatants were used for these arrays. iPS-ECs were transfected with either EX-mCherry, or EX-FSTL3 using Endofectin Max as described before. magnification. Quantification of angiogenesis progression was accomplished using the angiogenesis analyzer in Image J by counting the total master segments length, total segments, total meshes area and capillary tube branch points that formed after 8 hours.

Lentiviral particle transduction
Lentiviral particles were produced using MISSION shFSTL3 plasmid DNA (Sigma-Aldrich) according to the protocol provided and as previously described 2 . The shRNA Non-Targeting

LDL Uptake
To detect acetylated low-density lipoprotein (LDL) uptake by iPS-ECs, cells were incubated with Dil-ac-LDL (Molecular Probes) for 4 hours and were examined and photographed under a fluorescent microscope.

In vitro tube formation assay
24-well plates were coated with 289 μl/ml of Matrigel Matrix (10 mg/ml). The plates were incubated at 37ºC for 30 minutes, after which the remaining liquid was removed. 1.2x10 5 cells were plated in each well at a concentration of 4 x10 5 cells/ml. The cells were incubated for up to 18 hours at 37ºC & 5% CO 2 . Staining of the tubes was performed as described in the immunofluorescence staining section.

In vivo Matrigel Plug assay
In in vivo angiogenesis assays iPS-ECs overexpressing FSTL3 (EX-FSTL3) or control plasmid

Experimental hindlimb ischemia
The mouse hindlimb ischemia model was performed as previously described 3,4 . iPS-ECs overexpressing FSTL3 (EX-FSTL3) or control plasmid (EX-mCherry) trypsinised and injected intramuscularly into the adductors of ischemic NOD.CB17-Prkdcscid/NcrCrl mice. PBS-CTL was used as an additional control. Tissue blood flow of both legs was sequentially assessed by Laser Doppler imaging (moorLDL2-IR). Fourteen days later, mice were sacrificed, and hindlimb muscles were harvested following in situ perfusion fixation at physiological pressure, frozen in liquid nitrogen, and cryo-sectioned for assessment of neoangiogenesis.
Sections of adductor muscles were stained with CD144 antibody and capillary density was expressed as capillary number per mm 2 . In brief, specimens were placed in a humidified chamber and blocked in 5% donkey serum in PBS for 30 minutes at 37˚C and incubated with primary antibodies rabbit anti-CD144, prior to immunostaining, as described above. The bound primary antibodies were revealed by incubation with the secondary antibody; antirabbit Alexa488, at 37 o C for 45 minutes. Specimens were counterstained with 4',6diamidino-2-phenylindole (DAPI; Sigma-Aldrich), mounted in Floromount-G (Cytomation; DAKO, Glostrup, Denmark), and examined with a fluorescence microscope (Axioplan 2 imaging; Zeiss) or SP5 confocal microscope (Leica, Germany). Immunostaining was assessed and capillary density was calculated as capillary number/mm 2 . Cell engraftment ability was assessed by counting cells double positive for mCherry and EC marker (CD144) at ten randomly selected microscopic fields (at x40).

Statistical Analysis
Data is expressed as mean±SEM and analyzed using GraphPad Prism 5 software with a twotailed Student's t test for two groups or pairwise comparisons or ANOVA. A value of *p<0.05, **p<0.01, ***p<0.001 was considered significant.