Cardiac pericyte reprogramming by MEK inhibition promotes arteriologenesis and angiogenesis of the ischemic heart

Pericytes (PCs) are abundant yet remain the most enigmatic and ill-defined cell population in the heart. Here, we investigated whether PCs can be reprogrammed to aid neovascularization. Primary PCs from human and mouse hearts acquired cytoskeletal proteins typical of vascular smooth muscle cells (VSMCs) upon exclusion of EGF/bFGF, which signal through ERK1/2, or upon exposure to the MEK inhibitor PD0325901. Differentiated PCs became more proangiogenic, more responsive to vasoactive agents, and insensitive to chemoattractants. RNA sequencing revealed transcripts marking the PD0325901-induced transition into proangiogenic, stationary VSMC-like cells, including the unique expression of 2 angiogenesis-related markers, aquaporin 1 (AQP1) and cellular retinoic acid–binding protein 2 (CRABP2), which were further verified at the protein level. This enabled us to trace PCs during in vivo studies. In mice, implantation of Matrigel plugs containing human PCs plus PD0325901 promoted the formation of αSMA+ neovessels compared with PC only. Two-week oral administration of PD0325901 to mice increased the heart arteriolar density, total vascular area, arteriole coverage by PDGFRβ+AQP1+CRABP2+ PCs, and myocardial perfusion. Short-duration PD0325901 treatment of mice after myocardial infarction enhanced the peri-infarct vascularization, reduced the scar, and improved systolic function. In conclusion, myocardial PCs have intrinsic plasticity that can be pharmacologically modulated to promote reparative vascularization of the ischemic heart.


LMOD1
Leiomodin List of top genes unique to PC or DPC. The antigens later used for histological analyses of the mouse hearts were selected according to the following criteria: • Intracellular/membrane marker suitable for microscopy imaging; • Exclusion of soluble factors because unsuitable to localise in PC in situ; • High identity between the human and mouse proteins; • Availability of antibodies for histology.

Online extended methods
Where not otherwise stated, all chemicals were purchased from Sigma-Aldrich.

Isolation and primary cultures of cardiac pericytes (PC)
Human pericytes: Myocardial samples were collected in sterile, cold DMEM (Gibco) and processed within 24 h. The cell extraction protocol was as described previously (1). Briefly, samples were finely minced using scissors and scalpel until nearly homogenous and digested with Liberase enzyme (Roche) for up to 1 hour at 37 C, with gentle rotation. The digest was passed through 70-, 40and 30-μm strainers. Finally, the cells were recovered and sorted using anti-human CD31 and antihuman CD34 microbeads (Miltenyi) to deplete the population of CD31+ endothelial cells and select CD31-CD34+ cells (these are the PC). Cardiac PC were cultured onto plastic plates with full Endothelial Cell Growth Medium 2 (ECGM2, PromoCell) including recombinant human EGF (5 ng/mL), bFGF (10 ng/mL), VEGF (0.5 ng/mL), R3-IGF1 (5 ng/mL), and 2% v/v FBS. Cells were fed every three days and passaged at 90% confluence. Optimal seeding density was 8,000 cells/cm 2 . Frozen stocks of cells were preserved with 90% v/v FBS + 10% v/v DMSO. Centrifugation speed and time were 300g x 6 min. Unless otherwise stated, all experiments were performed between passage 4 and 7.
Mouse pericytes: C57BL/6 mice (n=5) were humanely killed under schedule 1 procedure and the heart harvested for cell extraction. Each heart was processed separately. Briefly, samples were finely minced using scissors and scalpel until nearly homogenous and digested with Liberase enzyme for up to 1 hour at 37 C, with gentle rotation. The digest was passed through 70-, 40-and 30-μm strainers. Final cells were recovered and sorted using anti-mouse CD31 microbeads (Miltenyi) and antimouse biotinylated-CD34 (Invitrogen) followed by anti-biotin microbeads (Miltenyi) to deplete CD31+ endothelial cells and select CD31-CD34+ cells (these are the PC). Cardiac PC were cultured with basal ECBM2 medium (Promocell) supplemented with 10% v/v FBS, recombinant mouse EGF (5 ng/mL) and bFGF (10 ng/mL) (both from Peprotech) and Amphotericin B (Gibco, final concentration of 0.25 μg/mL), onto plastic plates pre-coated with collagen from bovine skin (final concentration of 30 μg/mL). Cells were fed every 3 days and passaged at 90% confluence. Optimal seeding density was 20,000 cells/cm 2 . Frozen stocks of cells were preserved with 90% v/v FBS + 10% v/v DMSO. Centrifugation speed and time were 300g x 5 min. Unless otherwise stated, all experiments were performed between passage 4 and 7.

Culture of other human cells
Cardiac fibroblasts were purchased from PromoCell and Lonza, and expanded in complete Fibroblast Growth Medium 2 (PromoCell). Coronary artery EC (CAEC) were purchased from PromoCell and expanded in full Endothelial Cell Growth MicroVascular medium 2 (ECGMV2, PromoCell) according to manufacturer guidelines. Coronary artery smooth muscle cells (CASMC) were purchased from PromoCell and cultured with the Smooth Muscle Cell Growth Medium 2 (SMCGM2, PromoCell) according to the vendor instructions. All cells were used between passage 4 and 6.

Mycoplasma testing
All human cells used in this study tested negative for mycoplasma contamination (assessed using the PCR Mycoplasma Test Kit I/C, PromoCell, cat# PK-CA91-1096).

In vitro cellular assays
Immunocytochemistry analysis of human and mouse cells Cells were rinsed with PBS and fixed with 4% w/v PFA in PBS for 15 min at 20°C. After washing with PBS, the cells were permeabilized with 0.1% v/v Triton-X100 in PBS for 10 min at 20°C as required. Cells were blocked with 10% v/v FBS and incubated with the antibodies as reported in Supplemental tables 8 and 9, for 16 h at 4°C. Secondary antibodies (conjugated with Alexa 488, Alexa 568, Alexa 647) were all purchased from ThermoFisher Scientific and used at a dilution of 1:200, for 1 h at 20°C, in the dark. Nuclei were counterstained using Hoechst (1:10,000 in PBS, 3 min at 20°C). Cells were mounted using Fluoromount-G medium (ThermoFisher).

2D-Matrigel in vitro angiogenesis assay
CAEC were seeded on the top of Matrigel (Corning® Matrigel® Growth Factor Reduced Basement Membrane Matrix, cat# 356231) either alone (4,000 cells/well) or in co-culture with PC (4,000 CAEC + 1,500 PC /well), using Angiogenesis μ-Slides (IBIDI, UK). Images were snapped after 5 hours, and the total tube length per imaging field was measured. To assess the interaction between PC and CAEC, PC were labelled with the red fluorescent tracker Vybrant™ DiI Cell-Labeling Solution (Invitrogen) according to manufacturer instructions. In experiments with only PC (naïve or differentiated using PD0325901), 4,000 cells/well were seeded on the top of Matrigel. In experiments with only CAEC preconditioned with PD0325901 or vehicle for 5 days, 6,000 cells/well were seeded on the top of Matrigel.

Differentiation of human and mouse pericytes into vascular smooth muscle cells (VSMC) based on GF
For experiments aiming to assess the role of the different GF in the differentiation of PC, species-specific GF were added to, or removed from, the culture medium as indicated in the single experiments. While the concentration of FBS was always kept constant. Cells were cultured for a continuous period of 10 days with full media replacement every 48 h. Differentiation of human CASMC based on GF CASMC grow in SMCGM2 medium, which contains recombinant human EGF and bFGF and 5% v/v FBS. For differentiation experiments, EGF and bFGF were depleted from the medium and the FBS concentration reduced to 2% (equal to experiments with PC). Cells were cultured for a continuous period of 10 days with full media replacement every 48h. Pilot studies to determine PD0325901 efficacy and cytotoxicity PD0325901 (Sigma-Aldrich) was reconstituted with DMSO to get a final stock concentration of 500 μM. The compound was diluted in the culture medium as required, and the same amount of DMSO used as vehicle control. Human PC were cultured for either three or ten days with increasing concentrations of the drug (125 nM, 250 nM, 500 nM, 1 μM, 2 μM). To determine the cell viability, either the live cells Calcein-AM/EthDIII staining (Biotium) or Caspase Glo 3/7 assay (Promega) were used. For this latter, the relative luminescence units (RLU) were normalized against the total intracellular protein content. Based on these findings, we determined the maximum, non-toxic dose usable for the in vitro and in vivo cell transplantation experiments. Also, Western blotting analysis of ERK1/2 phosphorylation was used to determine the minimum dose of the compound that effectively prevented ERK1/2 activation during stimulation with GF.

Differentiation of human and mouse cells into VSMC with PD0325901
Cells were cultured with media supplemented with either 250 nM PD0325901 or the same volume of DMSO (vehicle) for ten days, with complete medium exchange every 48 h.

Role played by the cell cycle arrest in human PC differentiation
To investigate whether the PC differentiation was a consequence of the cell cycle arrest, PC were cultured with full GF medium supplemented with either 1 or 10 μM Ribociclic (TOCRIS) for ten days, with complete medium exchange every other day. The same volume of DMSO was used as control. The effective inhibition of cell proliferation was confirmed using an EdU assay (below). EdU was added to the cells during say 9 and 10 of the protocol. Expression of VSMC markers was assessed using immunocytochemistry.

Treatment of human cardiac PC with the PI3K inhibitor LY294002
To investigate if the AKT signaling impacted on PC differentiation, PC were cultured with full GF medium supplemented with LY294002 (FOCUS BIOMOLECULES), a PI3K inhibitor. Three concentrations were tested: 1, 5 or 10 μM (TOCRIS). Cells were cultured for ten days, with complete medium exchange every other day. The same volume of DMSO was used as control. The effective inhibition of AKT phosphorylation was confirmed using Western blotting (described below, see signaling studies). Expression of VSMC markers was assessed using immunocytochemistry.

EdU proliferation assay
The Click-iT EdU Cell Proliferation Kit for imaging (C10337 -ThermoFisher Scientific) was used to assess cell proliferation. Cells were incubated with EdU for 24 or 48 h and then analyzed. Fluorescence staining for EdU was performed according to the manufacturer instructions. Contraction assay A Cell Contraction Assay (CBA-201, Cell Biolabs) was used to assess the capacity of PC to contract upon stimulation with vasoconstrictors. At the end of the 10-days differentiation protocol, differentiated and control PC were embedded in collagen gels following the manufacturer instructions. Endothelin-1 (ET1) was used at a concentration of 0.1 μM to stimulate cell contraction. Butanedione Monoxime (BDM) -contraction inhibitor provided in the kit -was added to the culture medium at a 1:100 dilution. Gels were released and the area of the gel measured at baseline and after 24h. Results are expressed as percentage of gel contraction (measuring the gel area).

Intracellular calcium flux assay
The Fluo-4 AM fluorescent calcium indicator (ThermoFisher Scientific) was employed to study the intracellular calcium flux in naïve PC and differentiated PC following stimulation with Endothelin-1. Cell media was replaced with a buffer made of 20 mM HEPES, 137 mM NaCl, 5 mM KCl, 2 mM MgCl2, 1.8 mM CaCl2, 5.6 mM Glucose, 1 mg/mL bovine serum albumin, 0.5 mM NaH2PO4 and 1 mM NaHCO3, final pH = 7.4. The Fluo-4 dye (2 μM) was loaded into the cells for 15 min. HOECHST was used to counterstain the nuclei (1:1000 dilution). At the end of this incubation time, cells were stimulated with 0.1 μM ET-1 or vehicle (dH20). Images were recorded every 20 sec for 10 min using an INCell Analyzer 2200 microscope (GE Healthcare) equipped with a 20 X objective. ET1 was introduced in the cellular system after 120 sec using a built-in injection needle. The intracellular calcium was measured as relative fluorescence units (RFU). Data analysis was carried out using the dedicated software In Cell Analyser Workstation 3.7.

Gap closure migration assay
The Radius TM 96-Well Cell Migration Assay (CBA-126, Cell Biolabs) was used to assess the migratory properties of cardiac PC and CASMC at baseline and after differentiation. Stimuli used to induce cell migration were recombinant human VEGF-A (10 ng/mL), human SDF1α (100 ng/mL) and human PDGF-BB (20 ng/mL) (all from Peprotech). Migration time was 24h, in the presence of an inhibitor of cells proliferation (hydroxyurea, 2 mM). Absence of stimuli served as a control for migration. At the end of the protocol, cells were fixed with PFA and stained for VSMC markers (α-SMA or Calponin) and HOECHST (1:10,000 in PBS). The area of the gap not covered by cells was measured.

Experiments with cardiac fibroblasts
Fibroblasts were cultured in full GF medium and preconditioned for 7 days with 250 nM PD0325901 or vehicle. For the wound closure assay, 15 to 20,000 cells/well were seeded in 96-well plates. After 24 h, a scratch was created in the confluent cell monolayer using a 20 μL micropipette tip. Cells were washed twice with PBS and incubated for 24 h using basal medium depleted of GF and FBS as control, and full GF + FBS medium as stimulus for migration. Media were supplemented with 2 mM hydroxyurea to stop cell proliferation. Images were snapped at baseline and after 24 h. The surface of wound closed was calculated. For phenotype characterization, fibroblasts were used for immunocytochemistry and Western blotting after the 7-day treatment.

Enzyme-linked immunosorbent assay (ELISA)
For measurement of GF in cell-conditioned medium, confluent cells were cultured for 48 h with basal media without FBS and GF. Media were then collected, centrifuged for 10 min at 5,000g, aliquoted and stored at -80 C until used in the experiments. Quantikine ELISA Kit anti-human VEGF, ANGPT-1, ANGPT-2 and HGF were purchased from R&D Systems. The amount of secreted factors was normalized against the total intracellular protein content. For experiments with PD0325901-differentiated PC, the conditioned medium was collected at the end of the 10-day differentiation protocol. Analytes included also LEP and SERPINF1 (R&D). For all experiments, conditioned media were diluted as required for OD readings to fall within the standard curve. Cardiac troponin I was measured in mouse plasma using the Mouse cTn-I ELISA kit from Elabscience (E-EL-M1203), according to manufacturer instructions.

Western blotting on human and mouse whole cell lysates
Whole-cell protein lysates were collected using RIPA buffer supplemented with 1:50 proteases inhibitors cocktail and 1:100 phosphatases inhibitors. Protein lysates were centrifuged at 10,000 g, at 4°C, for 15 min. After the assessment of protein concentration (BCA Protein Assay Kit, ThermoFisher Scientific), the supernatants were aliquoted and kept at -80°C. Equal amounts of protein samples (10 to 30 μg as required) were prepared in Laemmli loading buffer, incubated for 8 min at 98°C, resolved on 8-12% SDS-PAGE and transferred onto 0.2 μm PVDF membranes (all from Bio-Rad). Membranes were blocked using 5% w/v BSA or 5% w/v non-fat dried milk (Bio-Rad) in Tris-buffered saline (TBS, BioRad) supplemented with 0.05% v/v Tween-20 for 2 h at 20°C. Primary antibodies (listed in Supplemental Tables 8 and 9) were incubated for 16 h at 4°C on a tube rocker. β-tubulin, β-Actin or GAPDH were used as a loading controls. Anti-rabbit IgG or anti-mouse IgG HRP-conjugated antibodies were employed as secondary antibodies (both 1:5000, GE Healthcare). Membrane development was performed by an enhanced chemiluminescence-based detection method (ECL™ Prime Western Blotting Detection Reagent, GE Healthcare) and observed using a ChemiDoc-MP system (Bio-Rad). Proteins with similar MW were assessed on different gels. No more than one stripping procedure was performed on an individual membrane when required (RestoreTM Plus Western Blot Stripping Buffer; Thermo Fisher Scientific). Western blot data were collected using the BioRad ImageLab software and analyzed using the free ImageJ software. Western blotting on mouse liver and heart samples Freshly frozen liver and heart samples were homogenized in RIPA buffer supplemented with proteases and phosphatases inhibitors (as described above) using gentleMACS M tubes (Miltenyi). Tissue lysates were centrifuged at 10,000 g, at 4°C, for 15 min and processed as described above. Signaling studies in human cardiac PC For the study of signaling activated by specific GF, human PC were cultured for 48 h with medium depleted of all GF and FBS. Cells were then stimulated with specific combinations of GF or nothing for control, for the time described in the different experiments (1, 6, 24 h). The whole-cell protein lysates were collected for Western blotting or protein arrays. A Human Phospho-kinase array kit (R&D Systems, catalogue number ARY003B) was used for quick screening of 43 kinase phosphorylation sites. For experiments employing small inhibitory molecules (PD0325901 and LY294002), cells were treated with the drug for 1 hour before stimulation with GF. Gene expression analysis by real-time qPCR Extracted total RNA was reverse-transcribed into single-stranded cDNA using a High Capacity RNA-to-cDNA Kit (ThermoFisher Scientific) according to manufacturer instructions. The RT-PCR was performed using first-strand cDNA with TaqMan Fast Universal PCR Master Mix (ThermoFisher Scientific). The assay numbers for the endogenous controls and target transcripts are listed in Supplemental Table  10. TaqMan primer-probes were all obtained from ThermoFisher Scientific. Quantitative PCR was performed on a QuantStudio TM 5 System (ThermoFisher). All reactions were performed in a 10 μL volume in triplicate, using 7.5 ng cDNA per reaction. The mRNA expression levels were normalized against the UBC gene and determined using the 2 −ΔΔCt method.(2) For cardiac PC, the transcriptional data were further normalized versus basal CASMC for two reasons: (i) To show the difference between naïve PC and CASMC -that means the intrinsic PC commitment to the VSMC lineage; (ii) Some transcripts in PC were not expressed at baseline (e.g. MYH11 and ELN were not detected in most samples and assigned Ct = 40).

Next-generation RNA-Sequencing
A whole-transcriptome analysis was carried out in 3 human PC differentiated for 10 days using PD0325901 (DPC) and compared with the respective naïve controls (DMSO-treated PC). CASMC (n=2) were used for internal reference. Strand specific RNA-sequencing was performed starting from total RNA (GENEWIZ). RNA-Seq was carried out on an Illumina HiSeq platform, with a 2x150bp configuration, ~20M reads per sample. Sequence reads were trimmed to remove possible adapter sequences and nucleotides with poor quality using Trimmomatic v.0.36. The trimmed reads were mapped to the Homo sapiens GRCh38 reference genome available on ENSEMBL using the STAR aligner v.2.5.2b, and BAM files generated. Unique gene hit counts were calculated by using featureCounts from the Subread package v.1.5.2. Only unique reads that fell within exon regions were counted. After extraction of gene hit counts, the gene hit counts table was used for downstream differential expression analysis. Using DESeq2, a comparison of gene expression between the groups of samples was performed. The Wald test was used to generate p-values and log2 fold changes. Genes with a false discovery rate (FDR) < 0.05 and absolute log2 fold change > 1 were called as differentially expressed genes (DEGs) for each comparison. A gene ontology (GO) analysis was performed on the statistically significant set of genes by implementing the software GeneSCF v.1.1-p2. The goa_human GO list was used to cluster the set of genes based on their biological processes and determine their statistical significance. Extended gene annotations, hierarchical clusterings, MA-plots, pathway analyses, heatmaps were generated using the open-access bioinformatics softwares iDEP 9.1 (http://bioinformatics.sdstate.edu/idep/) (3) and Morpheus (https://software.broadinstitute.org/morpheus). Venn diagrams were generated using InteractiVenn (http://www.interactivenn.net/index.html).(4) Protein-protein interaction networks were generated using the STRING software (https://stringdb.org/cgi/input?sessionId=bM1QcP9Hl6oN&input_page_show_search=on).

Discovery of genes unique to PC and DPC
For discovery of unique genes characterizing the two cell populations, we extracted from the RNA-Seq dataset a list of transcripts expressed only by either PC or DPC, and not expressed by CASMC. The top genes ranked by average transcript per million (TPM) and characterized by average TPM ≥ 200 were selected per each cell type. Genes later used for histology analysis were validated at the protein level using immunocytochemistry and Western blot.