A hypertension patient-derived iPSC model demonstrates a role for G protein-coupled estrogen receptor in hypertension risk and development

Abstract

Hypertension (HTN) is a polyfactorial disease that can manifest severe cardiovascular pathologies such as heart failure or stroke. Genome-wide association studies (GWAS) of HTN indicate that single-nucleotide polymorphisms (SNPs) contribute to increased risk for HTN and resistance to some HTN drug regimens (Hiltunen TP et al., J Am Heart Assoc 4: e001521, 2015; Le MT et al., PLoS One 8: e52062, 2013; McDonough CW et al., J Hypertens 31: 698–704, 2013; Vandell AG et al., Hypertension 60: 957–964, 2012). However, cellular mechanistic insights of such SNPs remain largely unknown. Using a bank of induced pluripotent stem cells (iPSCs) derived from patients with HTN and CRISPR/Cas9-mediated gene-editing approach, we investigated the effects of a female HTN risk-associated SNP (rs1154431) of the G protein-coupled estrogen receptor (GPER) (Bassuk SS, Manson JE., Clin Chem 60: 68–77, 2014) in vascular endothelial cells. Although GPER1 deletion reduced endothelial nitric oxide synthase (eNOS) activation in iPSC-derived endothelial cells (iECs), the polymorphism itself did not significantly affect eNOS and NO production in a comparison of isogenic hemizygous iECs expressing either normal (P16) or HTN-associated (L16) GPER. Interestingly, we demonstrate for the first time that GPER plays a role in regulation of adhesion molecule expression and monocyte adhesion to iECs. Moreover, the L16 iECs had higher expression of inflammation genes than P16 iECs, implying that the risk variant may affect carrier individuals through increased inflammatory activity. This study further indicates that iPSCs are a useful platform for exploring mechanistic insights underlying hypertension GWAS endeavors.

INTRODUCTION

Hypertension (HTN) is clinically defined at a sustained resting systolic blood pressure over 130 mmHg (41). Diagnosis of hypertension can only occur after significant, observable onset of high blood pressure readings over multiple clinical visits. Subsequently, the treatment of HTN must still undergo a burdensome repertoire of dose titration and a “wait and see” approach to alter treatment type and dose on a per-patient basis. These approaches shift an undue burden of the pathology to the patient, and increase exposure to HTN, thereby augmenting risk for secondary cardiovascular disease (CVD) such as stroke or heart failure to develop. Ergo, novel and efficient methods of determining HTN risk and treatment are needed to advance clinical strategy in this field. Recently, precision medicine initiatives (13, 19, 21, 33) have produced a wealth of genomic and pharmacogenomic data for cardiovascular diseases and ushered in a revolutionary phase of research in hypertension genetics to better stratify patient risk or response to HTN diagnosis and treatment based on gene variations, such as single-nucleotide polymorphisms (SNPs). In particular, multiple genes involved in HTN have been identified (26, 35, 52); however, their contribution to the changes in cellular function and net effect in risk or severity of HTN are not well understood, and they lack convincing genotype-to-phenotype readouts (33). As such, there is a present need for new methodologies to translate genetic etiologies into mechanistic targets to provide better prognostic insight or new therapeutic targets. The recent advent of patient-derived induced pluripotent stem cells (iPSCs) has provided a unique platform to assess changes at the cellular level from an isogenic renewable cell source from a patient that has the flexibility of undergoing differentiation into the tissue of interest for pathobiology research (17). Here we propose a model system using iPSCs and CRISPR/Cas9 to demonstrate the functional role of SNPs in the modulation of vascular cell function that culminates in HTN risk. A recent study outlined a role of an SNP in the gene for G protein-coupled estrogen receptor (GPER1), rs11544331. This SNP results in a coding mutation in the 3rd exon of the gene and substitutes a thymine for cytosine, which results in an amino acid change from proline (CCA) to leucine (CTA) at the 16th amino acid, abbreviated as P16L. This SNP manifested a modest increase in mean arterial blood pressure (MABP) in young healthy women, but not men, in a discovery cohort study (11). Given the literature on observed sex-related differences in cardiovascular health and disease between men and women (2, 5, 9, 18, 32, 34) and the implicated role of estrogen and GPER in cardiovascular function (3), it is possible that the effects of this GPER1 SNP may yield important mechanistic understanding of estrogen signaling in cardiovascular tissues. Several publications have outlined multiple functions in GPER that are dependent on cell type. Historically, the salutary role of GPER in estrogen signaling has been implicated as primarily vasoprotective, in part through the activation of endothelial nitric oxide synthase (eNOS) and subsequent NO production by endothelial cells to result in increased vasodilation (12, 14, 23, 40). Several laboratories have demonstrated that GPER exerts protective effects in cardiovascular function via eNOS-dependent vasodilation in arterial myography of healthy vessels from rats, mice, humans, and pigs (1, 14, 30, 37), exerts beneficial protection in several rodent models of hypertension (20, 29, 31), and exerts anti-inflammatory effects in mouse models of atherosclerosis (40, 44). Moreover, there is a marked increase in the rate of CVD in postmenopausal women, which rapidly equalizes to that of age-matched men, suggesting that loss of sex hormones (primarily estrogen) may be a significant factor in the increased incidence of CVD (5, 9, 18, 34). To this end, several clinical studies have addressed the role of estrogen replacement therapy (ERT) in cardiovascular health of postmenopausal women. However, evidence from these trials has yielded mixed results at best and may even suggest that ERT could be detrimental in some contexts (15, 47, 54). The reasons for unclear evidence linking ERT with CVD risk range from the health of the patient to the timing of the ERT or to underlying genetics (15, 18, 22). In an effort to examine genetic parameters of estrogen in vascular function, we propose to examine the role of the GPER1 SNP in estrogen signaling in endothelial function by utilizing iPSC-derived endothelial cells (iECs) from three women with HTN who carry the GPER1 SNP.

To this end, we designed a study utilizing patient-derived iPSCs from heterozygous female carriers of the GPER1 SNP and CRISPR/Cas9 gene editing to elucidate the role of the GPER gene in a translational in vitro model of cellular mechanisms in HTN (48). Previous studies implicated a hypofunctional effect of the GPER1 SNP but fell short of providing the physiologic context of relevant pathways known to mediate HTN onset that would be altered by the GPER P16L SNP. The purpose of this study is twofold: to demonstrate an in vitro system capable of screening patient SNPs for cellular function modifications in HTN-related phenotypes and to investigate the role of the GPER1 SNP in endothelial cell function. Since previous reports indicate a role for GPER signaling in mediating eNOS activation in primary endothelial cells (1, 14, 40, 43), we hypothesize here that a hypofunctional SNP would manifest a net decrease in eNOS activity and reduce endothelial functionality through impaired NO production.

METHODS

iPSC generation.

iPSCs were established from peripheral blood mononuclear cells of three female Caucasian patients over 45 yr old (HTN-03, HTN-20, and HTN-21) who were participants in the Pharmacogenomic Evaluation of Antihypertensive Responses (PEAR) pharmacogenomics study cohort (19, 21), carriers for the GPER1 SNP rs11544331 (C/T*), and clinically diagnosed as hypertensive, as we described previously (6, 48).

CRISPR/Cas9 gene editing.

Guide RNAs targeting GPER1 gene exon 3 were designed and purchased from GeneCopoeia. A homology arm-driven recombination strategy was employed, and a selection cassette carrying puromycin and enhanced green fluorescent protein (eGFP) or neomycin resistance with homology arm sequences was transfected with Cas9 plasmid and guide RNA to drive Cas9 cleavage at a protospacer adjacent motif (PAM) site 16 bp downstream from the GPER1 SNP C/T* rs11544331. Briefly, cells were harvested and dissociated into a single-cell suspension using StemPro Accutase (cat. no. A1110501; Thermo Fisher), centrifuged at 300 g, and counted with trypan blue to determine viability. Viable cells were resuspended and divided into samples of 500,000 cells per cuvette for electroporation using an Amaxa transfection machine. Cells were mixed with 2 μg homology arm donor plasmid and 2 μg Cas9 plasmid for a total of 4 μg DNA in Amaxa transfection buffer (83 μL) and Amaxa supplement (17 μL) per cuvette and briefly electroporated on Amaxa program 006 for iPSCs. Cells were immediately resuspended in mTeSR media (cat. no. 05825; StemCell Technologies) supplemented with 10 μM Rho-associated protein kinase inhibitor (ROCK) inhibitor Y27632 and seeded into 1 well of a 12-well plate (500,000 cells/cm²) precoated with growth factor-reduced, animal product-free Matrigel (cat. no. 356230; Corning). After overnight incubation (16–20 h), cells were gently washed with warm PBS and given mTeSR media supplemented with puromycin at a concentration of 0.6 μg/mL. Cells were cultured in puromycin for 7 days changed daily. Colonies that formed were selected by hand using a 27.5-gauge needle under a dissection scope in a sterile hood and subcultured for an additional 7 days in mTeSR with puromycin. At 80–90% confluence, cells were split 1:2 for passage 1. Upon 80–90% confluency in passage 1, one well was harvested and cryogenically frozen, and the other well was harvested for genomic DNA extraction and PCR to confirm insert of selection plasmid, followed by Sanger-sequenced for gene-editing results. Clones that were successfully edited were thawed and expanded. Edited cells with nonhomologous end joining (NHEJ) products were discarded. Primers used to confirm sequences are as follows: GPER1 unedited (targeting outside homology arm regions), forward 5′-GCTCCTGACACACCCAGACT-3′ and reverse 5′-CTCCCCGAGAAAGCTGTAGA-3′; GPER puromycin selection cassette, forward 5′-AACTGGGAAAGTGATGTCGTG-3′ and reverse 5′-GTCAATTCCGGGGCTAAAC-3′; and GPER neomycin selection cassette, forward 5′-GGGCCACTCGATGAGACTTC-3′ and reverse 5′-CAACTTCTCGGGGACTGTGG-3′.

Differentiation of vascular endothelial cells.

On day 0, iPSCs were dissociated into a single-cell suspension using Accutase for 7–10 min at 37°C and plated at a density of 5,000 cells/cm² onto ES-qualified Matrigel-coated plates with mTeSR media supplemented with 10 ng/mL ROCK inhibitor. At 30–50% confluence, monolayers were washed with PBS, and mesoderm induction medium (MIM, cat. no. 05220; StemCell Technologies) was added at a volume of 300 μL/cm² (day 1). Cells were incubated for 48 h in MIM, and the medium was changed every 24 h. On day 3, the medium was exchanged for APEL2 (cat. no. 05270; StemCell Technologies) supplemented with 5 ng/μL bone morphogenetic protein 4 (BMP-4; Peprotech) and 10 ng/uL VEGF165 (Peprotech). Cells were incubated in supplemented APEL2 for 4 days, and media were changed every 48 h. On day 7, cells were dissociated and separated using CD34 EasySep selection cocktail (cat. no. 17856; StemPro) and EasySep magnet (cat. no. 18000; StemPro) for CD34⁺ and CD34⁻ populations. CD34⁺ cells were plated onto human fibronectin-coated plates (1 µg/cm²) at a density of 7,000 cells/cm², cultured in Endothelial Cell Growth Medium-2 (EGM2, cat. no. CC-3162; Lonza) supplemented with an additional 10 ng/mL VEGF165 peptide (Peprotech) for 2 wk, and then split at a ratio of 1:2 up to passage 6.

Flow cytometry.

After 7–9 days in culture, cells were assessed for specific markers of terminal differentiation into endothelial cells (iECs). Cells were dissociated into a single-cell suspension using Accutase (~100 µL/cm²) for 7 min at 37°C, washed twice with PBS, and passed through a 30-μm strainer. Cells were resuspended at a density of 5,000 cells/mL, and up to 100,000 cells were used for each stain. Cells were incubated on ice for 45 min in the dark with anti-CD31-Alexa Fluor 488 (clone M89D3; BD Biosciences) at a concentration of 1.0 μL per 100,000 cells and anti-CD144-phycoerythrin (PE, clone 55-7H1; BD Biosciences) at a concentration of 0.5 μL per 100,000 cells. Cells were washed with 1,000 μL FACS buffer, centrifuged at 300 g for 5 min, and resuspended in 300 μL FACS buffer for flow cytometry. Labeled cells were analyzed with an Accuri c6 flow cytometer using blue and red laser excitations for proper fluorophore detection. Generally, cells were in excess of 95% double positive for CD31 and CD144. In most cases, primary human aortic endothelial cells (HAECs, cat. no. CC-2535; Lonza) were used as a positive control for endothelial phenotype and functional assays.

eNOS stimulation in unedited and GPER-null iECs.

iPSC-derived endothelial cells were seeded into six-well plates at a density of 10,000 cells/cm². When cells were 80–90% confluent, cells were washed once with PBS and replaced with low-serum (1% FBS) basal endothelial media (EBM2; Lonza) for 3 h. Cells were then stimulated with 1.0 µM G-1 and 0.1% DMSO vehicle for 20 min. Cells were immediately washed with ice-cold PBS and lysed on ice with ice-cold radioimmunoprecipitation assay (RIPA) lysis buffer supplemented with 10% SDS, 10 µM NaF, 1.0 µM NaVO₄, and protease inhibitor cocktail (Sigma) to preserve phosphorylated and normal proteins.

Western blot.

Cells were washed twice with ice-cold PBS and lysed and scraped on ice using 80 μL per well of a six-well plate with RIPA buffer supplemented with 10% SDS and protease inhibitor (Sigma-Aldrich), 1 μM NaF, and NaVO₄ to inhibit proteases. Lysates were centrifuged at 13,000 rpm for 10 min at 4°C. Supernatants were collected and stored at −80°C until processed further for Western blots. Lysates were blotted on 10% SDS-PAGE, transferred to a PVDF membrane, and blocked for 1 h with 3% BSA Tris-buffered saline-0.1% Tween 20 (TBS-T) blocking buffer. Blots were probed with appropriate antibodies (1:500 eNOS S1179 mouse IgG, BD Biosciences; 1:5,000 β-actin rabbit IgG, Cell Signaling Technology; and 1:10,000 GPER clone 8073 targeting NH₂-terminal epitope, rabbit IgG) overnight at 4°C, washed three times with TBS-T, and incubated with horseradish peroxidase (HRP)-linked secondary antibody at 1:5,000 in blocking buffer for 1 h at room temperature. Blots were washed three times with TBS-T, developed using equal parts PicoWest developer solutions (Thermo Fisher) for 1 min, and imaged on Carestream X-OMAT LS autoradiography X-ray film.

Immunofluorescence.

Cells were seeded onto type 00 coverslips coated with 0.1% gelatin in a 24-well plate at 10,000 cells per well. Cells were washed twice with PBS and fixed for 15 min with 4% paraformaldehyde (PFA) at room temperature and then washed three times with PBS and blocked for 1 h at room temperature with 3% BSA PBS-0.1% Triton X-100 (PBS-T). Coverslips were then delivered antibodies diluted at 1:100 in blocking buffer overnight at 4°C. Antibodies used in this study include the following: rabbit IgG platelet endothelial cell adhesion molecule (PECAM-1)-Alexa Fluor 488 conjugate (cat. no. 563607; BD Biosciences), mouse IgG anti-eNOS (cat. no. 610297; BD Biosciences), rabbit IgG anti-von Willebrand factor (cat. no. ab6994; Abcam), and rabbit polyclonal anti-GPER clone 8073 targeting the COOH terminus (provided by E. R. Prossnitz, University of New Mexico), which was used at a dilution of 1:10,000. Negative control coverslips were incubated with rabbit IgG. The next day, cells were washed three times with PBS-T and incubated with appropriate secondary antibody at 1:500 dilution in blocking buffer for 1 h at room temperature, washed three times with PBS-T (0.01% Triton X-100) and treated with 10 nM DAPI in PBS for 20 min at room temperature in the dark, washed three times with PBS followed by a three second wash with nano-pure water and mounted onto glass slides with 12 µL ProLong Gold (cat. no. P36930; Thermo Fisher) mounting medium. Coverslips were imaged with a Zeiss LSM 710 META confocal microscope. In all cases, negative controls (secondary antibody only) were used to set background signal of confocal lasers before labeled cells were imaged.

RNA, cDNA, and quantitative PCR.

Primary HAECs and unedited and gene-edited iECs were seeded into six-well plates at passage 2 and grown to 80% confluence (usually 48–72 h). Cells were washed twice on ice with cold PBS and lysed with RNA lysis buffer (70 μL per well of a 6-well plate), and raw lysates were passed with a 25-gauge needle to further lyse samples and were stored at −80°C. RNA was isolated using an RNAqueous extraction kit (cat. no. AM1912; Ambion), and DNA digest was performed using TURBO DNA DNase kit (cat. no. AM1907; Thermo Fisher). Nucleic acids were quantified, and 200 ng of RNA per sample were used for cDNA synthesis with Applied Biosystems First Strand Reverse Transcription Kit.

Quantitative PCR (qPCR) was performed using Power SYBR Green Master Mix (cat. no. 4367659; Applied Biosystems) with appropriate primers (purchased from Integrated DNA Technologies) on an AB StepOne thermocycler machine with a 60°C annealing temperature. Gene expression was calculated by the comparative threshold cycle (ΔΔCT) method. The following human primer sequences for qPCR were used: tyrosine-protein kinase receptor 2 (Tie-2), forward 3′-TCTGCAGTGCAATGAAGCAT-5′ and reverse 5′-ATCCTTGGAAGCGATCACA-3′; vascular cell adhesion molecule 1 (VCAM1), forward 5′-CCGGATTGCTGCTCAGATTGGA-3′ and reverse 5′-AGCGTGGAATTGGTCCCCTCA-3′; GAPDH, forward 5′-CTCTGCTCCTCCTGTTCGAC-3′ and reverse 5′-GCGCCCAATACGACCAAATC-3′; eNOS, forward 5′-TCAGTGGCTGGTACATGAGC-3′ and reverse 5′-TATCCAGGTCCATGCAGACA-3′; von Willebrand factor (VWF), forward 5′-TTGACGGGGAGGTGAATGTG-3′ and reverse 5′-ATGTCTGCTTCAGGACCACG-3′; monocyte chemoattractant protein 1 (MCP-1), forward 5′-TGCAGAGGCTCGCGAGCTA-3′ and reverse 5′-CAGGTGGTCCATGGAATCCTGA-3′; and GPER1, forward 5′-GTACCCAGAAGTGAGCAGCT-3′ and reverse 5′-CAGGTGGTCCATGGAATCCTGA-3′.

Diaminofluorescein-2 diacetate nitric oxide detection.

Cells were seeded at a density of 10,000 cells/cm² onto recombinant human fibronectin-coated coverslips (1.0 µg/cm²) in a 24-well plate and grown for 48 h. On the day of experiment, cells were washed with PBS and given serum-free media for 3 h. Cells were then washed with prewarmed sterile PBS and treated with prewarmed 5 µM diaminofluorescein-2 diacetate (DAF-2 DA) in 430 µL HEPES-buffered physiological saline solution (HPSS) supplemented with 0.1% FBS for 2 h in the dark at 37°C. At the 1-hour time point, cells were treated with 1 µM G-1, 1 µM E2, or 0.1% DMSO vehicle for the next 60 min of DAF-2 DA incubation. In the final 10 min of DAF-2 DA incubation, 70 μL of NucBlue (cat. no. R37605; Thermo Fisher) were added to each well used to label cell nuclei. Wells were then gently aspirated, and cells were gently washed once with warm PBS, fixed with 4% PFA at room temperature for 20 min, and then washed with PBS twice. Coverslips were mounted with ProLong Gold (Thermo Fisher) onto glass slides and imaged on a Zeiss LSM 710 META confocal microscope for FITC and DAPI signal at ×40. In all instances, a negative control of basal (unstimulated cells) DAF-2 DA fluorescence was used to set argon laser intensity and signal gain before imaging different treatment groups of DAF-2 DA on the same exact settings. TIFF imaging files were converted to 8-bit format, and the pixel intensity of DAF-2 DA fluorescence was calculated using ImageJ (National Institutes of Health, Bethesda, MD). Pixel intensity values were divided by number of cell nuclei for quantitation of DAF-2 DA fluorescence. At least three images per treatment group for unedited, hemizygous P16 (−/P16), hemizygous L16 (−/L16), or GPER-null iECs from two to three different patients were used to calculate average intensity.

Adhesion assay.

iECs at passages 1–3 were seeded at 30,000 cells per well of a 24-well plate precoated with 1 μg/cm² fibronectin (coated for 30 min at 37°C washed once with 1X PBS). At 24 h postseeding, cells were washed with serum-free media and treated overnight with TNF-α (10 ng/mL) in 1% FBS serum. The following morning, the THP-1 monocyte cell line was loaded with 10 μM Cell Tracker Green (Thermo Fisher) for 30 min at 37°C in the dark. Labeled THP-1 cells were added to each well (150,000 cells per well) containing an endothelial monolayer, and the plate(s) was incubated in the dark on a rocking platform at 37°C for 30 min in a total volume of 500 μL per well. Wells were gently aspirated, washed twice with 1 mL of warm PBS, and replaced with 500 μL of warm (37°C) Hank’s buffered salt solution. Wells were immediately imaged on an Olympus X-100 epifluorescence scope with an FITC filter at ×10 magnification. The number of THP-1 cells per field over four to five fields of vision per well were counted per well using ImageJ analysis software (National Institutes of Health, Bethesda, MD).

Statistics.

Analysis was performed and graphed in GraphPad Prism version 7.0. Data were analyzed using a paired two-way ANOVA with Tukey’s post hoc test. In some instances, a paired t test was used to compare −/P16 to −/L16 effects separately. A P value of <0.05 was accepted as statistically significant. Gene-edited iECs from all three patient-derived cell lines were used to calculate a response per genotype. At least two clones for the GPER null genotype and three clones for hemizygous genotypes of each patient with HTN were used to calculate an average response.

RESULTS

Gene editing of patient iPSCs.

The gene-editing schematic is illustrated in Fig. 1A. Following iPSC transfection with CRISPR donor and Cas9 plasmids, edited clones of hemizygous P16 (−/P16) and hemizygous L16 (−/L16) were derived from each parental iPSC line (HTN-03, HTN-20, and HTN-21) by selecting an average of 50 clones after 7 days of puromycin treatment at 0.6 μg/mL changed daily. GPER-null (−/−) clones were obtained by generating a subsequent CRISPR/Cas9 knockout of the other allele using a donor DNA with a neomycin cassette and treatment with G418 for 3 days at 0.3 μg/mL changed daily. At least two clones of GPER1 null (−/−) and three clones of hemizygous normal (−/P16) or hemizygous polymorphic (−/L16) GPER1 were selected from each patient iPSC line and used for future experiments. Successful gene knockout was initially examined by PCR and gel electrophoresis to determine amplicon product size using primers targeting specific gene-edited sequences (Fig. 1B) and confirmed by Sanger sequencing (Fig. 1C). Chromatographs indicated the remaining allele in hemizygous gene in −/P16 or −/L16 edited iPSCs (Fig. 1C). The average rate of homology-driven repair with no NHEJ over three patient-sourced iPSC lines for gene editing was 25.0% for −/P16 hemizygous allele, 23.5% for −/L16 hemizygous allele, and 8.8% for GPER1 null (Fig. 1D). The proliferation, survival, and morphology of iPSCs was not affected by gene editing (data not shown). In line with the deletion of one allele, qPCR for GPER1 expression demonstrated a trending reduction in GPER1 mRNA in hemizygous differentiated iECs (−/P16 and −/L16) and negligible signal in GPER1-null (−/−) differentiated iECs (Fig. 1E).

Fig. 1.

Gene editing of the G protein-coupled estrogen receptor (GPER) polymorphism rs1154331 results in hemizygous and double knockout induced pluripotent stem cell (iPSC) clones. A: schematic of the gene-editing design for homology-driven recombination. The polymorphism located is in the third exon (NM_001039966.1). A plasmid designed with puromycin (Puro) and green fluorescent protein (GFP) selection sequences is flanked by homology arms and is driven by the human elongation factor 1-alpha (EF1α) promoter. In some cases, a neomycin (Neo) selection cassette was used to replace the remaining allele in a second round of gene editing. The single-nucleotide polymorphism (SNP) rs1154331 (C/T) and gRNA (3′-GCAGTACGTGATCGGCCTGT-5′) sequences are located +931 bp and +1,274 bp from the ATG start site, respectively. Fwd, forward; Rev, reverse. B: PCR gel detecting insertion of the CRISPR donor sequence in iPSCs. Primers were designed to target sequences outside of the homology arm region during PCR amplification to ensure integration of the selection cassette when homology-driven recombination (HDR) was successful. Primer target sequences are denoted for their PCR products. For detection of native GPER1 sequence, forward primer 1 (fp1) and reverse primer2 (rp2); for detection of puromycin insert, forward primer 3 (fp3) and reverse primer 4 (rp4); and for detection of neomycin insert, forward primer 5 (fp5), and reverse primer 6 (rp6). C: example chromatograph of unedited heterozygous patient-derived iPSCs (left) and the remaining allele left as either C (P16, normal allele) or T (L16 polymorphic allele). D: table of CRISPR genotyping results for iPSC clones from each of three female patients with hypertension (HTN). E: quantitative PCR of differentiated gene-edited endothelial cells demonstrates decreases in GPER1 expression when CRISPR gene editing produced hemizygous and/or homozygous knockout. −/L16, hemizygous polymorphic; −/P16, hemizygous normal; P/L16, expressing both alleles; −/−, GPER1 null. Nonsignificant effect between −/P16 and −/L16 iPSC-derived endothelial cell (iEC) clones by paired t test. *P < 0.05 vs. unedited by two-way ANOVA; ns, nonsignificant between –/P16 and –/L16 iEC clones by paired t test.

Differentiation of endothelial cells.

At 2 wk post-differentiation, iPSC-derived vascular endothelial cells from patients with hypertension (HTN-iECs) presented classical cobblestone morphology under bright-field (BF) microscopy (Fig. 2A). Immunofluorescence microscopy of HTN-iECs compared with primary human aortic endothelial cells (HAECs) demonstrated similar expression patterns of PECAM (CD31), VE-cadherin (CD144), and von Willebrand factor (VWF; Fig. 2A). To further quantify endothelial differentiation efficiency, flow cytometry of iECs (Fig. 2B, right) routinely demonstrated an excess of >90% positive staining for CD144 (VE-cadherin, y-axis) and CD31 (PECAM, x-axis), similar to expression in HAECs (Fig. 2B, left). qPCR showed that iECs had positive expression of endothelial-specific genes such as eNOS and Tie 2 (Fig. 2, C and D) compared with HAECs and no expression of stem cell pluripotency genes octamer-binding transcription factor 4 (Oct 4) and Nanog (Fig. 2, E and F) compared with iPSCs. Preliminary functional characterization showed that primary HAECs and differentiated iECs demonstrated similar time-dependent increases in mRNA expression of MCP-1 in response to the inflammation agent TNF-α (10 ng; Fig. 2G). Cells regularly proliferated through passage 7; however, experiments in this study utilized cells between passages 1 and 4.

Fig. 2.

Induced pluripotent stem cell (iPSC)-derived endothelial cells (iECs) reveal expected phenotype. A: immunofluorescence of iECs presented similar patterns of both cell membrane and cytosolic markers of endothelial phenotyping such as platelet endothelial cell adhesion molecule (PECAM), VE-cadherin, and von Willebrand factor (VWF) compared with primary human aortic endothelial cells (HAECs). BF, bright field. B: flow cytometry revealed that iEC differentiation results in similar purity of CD31 (y-axis) and CD144 (x-axis) double-positive populations at 2 wk postdifferentiation compared with primary human endothelial cells in excess of 95% double-positive populations. HTN-iECs, iECs from a patient with hypertension; PE, phycoerythrin. C–F: quantitative PCR demonstrated similar expression levels of endothelial nitric oxide synthase (eNOS; C) and Tie 2 (D) endothelial genes compared with HAECs, and no longer express stem cell-specific pluripotency genes Oct 4 (E) and Nanog (F), indicating complete differentiation of iPSCs; *P < 0.05 vs. iPSCs. G: functional test of iECs showed upregulation of monocyte chemoattractant protein-1 (MCP-1) in response to TNF-α (10 ng/mL) over time and to a similar extent observed in HAECs; *P < 0.05 vs. basal control per cell type.

GPER expression and localization in differentiated iECs.

Further to qPCR detection of GPER1 mRNA in edited iPSCs (Fig. 1E), we compared GPER1 expression in iEC clones expressing either both alleles (P/L16), neither allele (−/−), or the P16 or L16 allele of the GPER gene. GPER was detected by immunofluorescence in unedited and hemizygous iECs, but signal was not detected above negative IgG background in GPER1-null edited iECs (Fig. 3). Given recent evidence that the GPER1 SNP results in nuclear sequestration of GPER protein in breast cancer carcinoma cells (45), we investigated whether this observation could be reproduced in our system. DAPI (blue) was used to counterstain nuclear space, and an antibody targeting eNOS (red) was used to identify the cytosolic space in endothelial cells. Contrary to data in cancer cells, confocal microscopy of gene-edited iECs revealed that GPER in both P16 and L16 iECs is similarly localized in the cytoplasmic space and did not present a pattern of nuclear sequestration of GPER of the variant (−/L16) compared with normal allele (−/P16) in HTN iECs (Fig. 3).

Fig. 3.

Immunofluorescence of G protein-coupled estrogen receptor (GPER) expression and localization in gene-edited induced pluripotent stem cell (iPSC)-derived endothelial cells (iECs). GPER expression was not detectable in GPER-null (−/−) edited iECs (third row) and did not differ from IgG negative control images (fourth row). When comparing subcellular localization between nuclear (DAPI, blue) or cytoplasmic [endothelial nitric oxide synthase (eNOS), red] compartments, confocal detection of GPER (green) did not appear to differ between hemizygous normal (−/P16) (second row) and hemizygous polymorphic (−/L16) (third row) genotypes. HTN-iECs, iECs from a patient with hypertension; P/L16, expressing both alleles (first row).

eNOS phosphorylation in response to GPER agonist is lost in GPER1-null iECs.

Since it has been shown that the GPER-specific agonist G-1 can increase eNOS phosphorylation at residue serine 1177 (S1177) in endothelial cells (14), we tested efficacy of GPER gene editing in our patient-derived iEC lines. Western blot detected an increase in S1177 phosphorylation to G-1 in unedited iECs; however, quantifiable changes in S1177 phosphorylation in response to G-1 were not observed in GPER1-null HTN-iECs (Fig. 4A). These data establish a functional loss of GPER signaling in our CRISPR gene-edited patient-derived iEC model. Functional characterization of endothelial NO was also examined in GPER −/P16 or −/L16 hemizygous iECs. Western blot of eNOS phosphorylation at residue serine 1177 did not demonstrate a significant difference in densitometry values of eNOS phosphorylation to the GPER-specific agonist G-1 between normal and polymorphic hemizygous iECs (0.841 ± 0.101 −/P16 vs. 0.765 ± 0.145 −/L16, means ± SEM, P = 0.237; Fig. 4, A and B).

Fig. 4.

Impacts of G protein-coupled estrogen receptor (GPER) genotypes on endothelial nitric oxide synthase (eNOS) activation in induced pluripotent stem cell (iPSC)-derived endothelial cells (iECs) from patients with hypertension (HTN). A: Western blot of iEC lysates from (left to right) unedited (P/L16), GPER1-null (−/−), hemizygous normal (−/P16), and hemizygous polymorphic (−/L16) iECs treated with G-1 (100 nM for 20 min) show that eNOS phosphorylation (S1177; p-eNOS) increases in unedited and hemizygous iECs but has no effect on GPER1-null edited iECs. B: densitometry revealed no statistical difference in S1177 phosphorylation upon stimulation with the GPER-specific agonist G-1 between −/P16 and −/L16 single-nucleotide polymorphism (SNP) variants in iECs from all three patients with hypertension (representative blot of genotype clones for patient HTN-03 iECs).

NO production in GPER1-edited iECs.

A salutary role of the endothelium is production of the vasodilator nitric oxide (NO) via eNOS activation. Further to our observations of decreased eNOS phosphorylation, we assessed NO production via DAF-2 DA fluorescence. Unedited and GPER1-null HTN-iECs were stimulated with the GPER-specific agonist G-1 or nonspecific estrogen receptor ligand E2 for 20 min. Confocal imaging of DAF-2 DA fluorescence revealed that GPER-null iECs had reduced NO production in response to E2 compared with unedited iECs (3,185 ± 114 P/L16 + E2 vs. 2,275 ± 177 P/L16 GPER −/− + E2; Fig. 4, A and B). GPER-specific agonist G-1 elicited an increase in NO production in unedited iECs, but to a lesser extent than E2 (1,057 ± 269 P/L16 + G-1 vs. 3,185 ± 114 P/L16 + E2, P < 0.05). GPER deletion in iECs from patients with HTN abrogated any significant NO production to G-1 (294 ± 74 GPER −/− vehicle vs. 463 ± 164 GPER −/− + G-1, P = 0.083; Fig. 4, A and B). These results align with our molecular data from Western blotting and establish that GPER contributes to estrogen-mediated NO production in HTN-iECs. However, in line with observations of negligible differences in phospho-eNOS blots of the hemizygous iECs treated with G-1, DAF-2 DA fluorescence showed no statistically significant change in NO production in response to G-1 between the normal and polymorphic alleles (3,181 ± 59 −/P16 vs. 3,094 ± 180 −/L16; Fig. 5, A and B).

Fig. 5.

Effects of G protein-coupled estrogen receptor (GPER) single-nucleotide polymorphism (SNP) variants on nitric oxide generation in induced pluripotent stem cell (iPSC)-derived endothelial cells (iECs) of patients with hypertension (HTN). A: diaminofluorescein-2 diacetate (DAF-2 DA) fluorescence intensity as a measure of NO production increased in response to both G-1 and E2 in unedited (P/L16) iECs. NO production via G-1 is lost in GPER-null (−/−) iECs; however, NO production in response to nonselective estrogen activation (E2) was not statistically significantly different between unedited and GPER-null iECs. No statistically significant difference in the production of NO was demonstrated between hemizygous normal (−/P16) and hemizygous polymorphic (−/L16) SNP iECs. Scale bar, 20 µm. B: graphical quantification of DAF-2 fluorescence intensity. KO, knockout; ns, not significant; WT, wild type. *P < 0.05 vs. basal.

GPER1 deletion increases expression of inflammation genes in response to TNF-α.

In addition to vasotone control via NO, the endothelium acts as a first line of regulation of the inflammatory response in arterial networks via recruitment of lymphocytes for extravasation to the intimal layers (40, 42). Estrogen receptors including GPER have been implicated in the inflammatory response in vivo (16, 24, 40). Therefore, we sought to investigate the impact of GPER1 deletion in the regulation of genes involved in endothelial inflammation responses related to leukocyte attachment (MCP-1), barrier function (VCAM), and coagulation (VWF). Using qPCR, we determined that basal expression of MCP-1, VCAM, and VWF was not significantly different between unedited, hemizygous, or GPER-null edited iECs over three patient-derived iEC lines (Fig. 6, A–C). An initial phenotyping result of iEC differentiation demonstrated that unedited iECs derived from HTN females revealed a time-dependent increase in MCP-1 RNA expression in response to TNF-α treatment (10 ng/mL) that was similar to primary aortic endothelial cells (HAECs; Fig. 2G). However, when we repeated this assay in GPER1-null HTN-iECs, we observed a marked increase (32 ± 1.24 P/L16 vs. 112 ± 11 GPER1 −/−, P = 0.04) in MCP-1 mRNA expression in response to TNF-α (Fig. 7A). VCAM gene expression also increased in response to TNF-α in unedited HTN-iECs, and this response was also exacerbated in GPER-null HTN-iECs (Fig. 7C; 9.7 ± 3.8 P/L16 vs. 1,425 ± 412 GPER1 −/−, P < 0.05). VWF gene expression increased upon TNF-α stimulation in unedited HTN-iECs, and this response was again significantly magnified in GPER1-null HTN-iECs (Fig. 7E; 20 ± 4 P/L16 vs. 126 ± 27 GPER −/−). Inflammation gene expression was also assessed by qPCR in hemizygous P16 or L16 iECs challenged with TNF-α. Interestingly, MCP-1 mRNA markedly increased upon TNF-α treatment in −/L6 compared with −/P16 iECs (303.1 ± 39 −/L16 vs. 19.47 ± 7 −/P16, P < 0.0001; Fig. 7B). VCAM and VWF mRNA expression also increased in response to TNF-α, and polymorphic iECs had slightly greater expression in response to TNF-α compared with the normal allele (Fig. 7, D and F, respectively). Pretreatment of iECs with the GPER-specific agonist G-1 for 30 min dampened MCP-1 expression in response to TNF-α in both normal and polymorphic iECs (31.5 ± 2.5 −/P16 + TNF-α + G1 vs. 19.5 ± 7.5 −/L16 + TNF-α + G-1, P < 0.0001).

Fig. 6.

Basal expression of inflammation genes is no different in induced pluripotent stem cell (iPSC)-derived endothelial cells (iECs) of different G protein-coupled estrogen receptor (GPER1) genotypes. Relative quantification by quantitative PCR shows that between patient lines of iECs, genotype [unedited (P/L16), hemizygous normal (−/P16), hemizygous polymorphic (−/L16), or GPER null (−/−)] did not result in a statistically significant difference in expression of monocyte chemoattractant protein 1 (MCP-1; A), VCAM (B), or von Willebrand factor (VWF; C).

Fig. 7.

G protein-coupled estrogen receptor (GPER)-edited induced pluripotent stem cell (iPSC)-derived endothelial cells (iECs) have an increased expression of inflammatory response genes. However, GPER1-null (−/−) iECs and polymorphic hemizygous (−/L16) iECs showed a marked increase in monocyte chemoattractant protein 1 (MCP-1) production in response to TNF-α compared with unedited (P/L16) iECs and normal allele-expressing (−/P16) iECs (A and B). VCAM (C and D) and von Willebrand factor (VWF; E and F) were also increased in GPER1-null and −/L16 iECs in response to TNF-α to a greater extent than unedited or −/P16 iECs. A: MCP-1 mRNA increased in response to TNF-α (10 ng/mL) in both −/P16 and −/L16 iECs. B: MCP-1 expression was significantly higher in −/L16 single-nucleotide polymorphism (SNP) variant-expressing iECs. This increase was mitigated by pretreatment with the GPER-specific agonist G-1 in both −/P16 and −/L16 hemizygous genotypes. C and D: iEC expression of VCAM and VWF mRNA (respectively) also increased in response to TNF-α, and this increase was diminished by G-1 pretreatment in both hemizygous genotypes. *P < 0.05 vs. vehicle, #P < 0.05 vs. TNF treated.

Monocyte adhesion alteration in GPER1-edited endothelial cells.

Since increased MCP-1 expression promotes lymphocyte adhesion to endothelial surfaces, we sought to determine whether GPER1-null HTN-iECs exhibit increased lymphocyte adhesion as a consequence of increased inflammation gene transcription. We assessed THP-1 monocyte adhesion to unedited and GPER1-null HTN-iEC monolayers (bright-field and FITC image examples, Fig. 8C). Counts of THP-1 adhesion to untreated endothelial monolayers were higher in GPER1-null iECs but not statistically different compared with unedited (P/L16) iECs (Fig. 8, A and B; 19 ± 8 P/L16 vs. 34 ± 17 GPER −/−, P = 0.53). Upon TNF-α stimulation (10 ng/mL overnight), there was a significant increase in THP-1 adhesion in both unedited and GPER1-null HTN-iEC monolayers (416 ± 17 P/L16 + TNF-α and 765 ± 128 GPER1 −/− + TNF-α). The increase in monocyte adhesion in TNF-α-treated GPER1-null iECs was 45.62% greater than in TNF-α-treated unedited iECs. GPER-specific agonist G-1 reduced THP-1 adhesion in unedited iECs by 66.58% but did not affect THP-1 adhesion in GPER1-null iECs (139 ± 57 P/L16 + G-1 and 797 ± 75 GPER1 −/− + G-1; Fig. 8B). These results imply that the changes in expression of inflammation genes in GPER1-null HTN-iECs align with increased monocyte adhesion. Nonselective estrogen receptor agonist E2 significantly reduced THP-1 adhesion in unedited HTN-iECs by an average of 71.94% and in GPER1-null HTN-iECs by an average of 65.12%. The difference between total THP-1 adhesion counts in the two genotypes was not statistically different for E2 treatment (360 ± 15 P/L16 + E2, and 246 ± 79 GPER1 −/− + E2; Fig. 8A) suggesting that E2 pretreatment normalized the GPER1-null iEC adhesion phenotype response to TNF-α. Subsequent to gene expression of inflammation targets in polymorphic iECs, THP-1 monocyte adhesion revealed that L16 hemizygous iECs had a marginal increase in monocyte adhesion compared with P16 iECs in response to TNF-α, but this did not reach statistical significance (Fig. 8B; 579 ± 83 −/P16 + TNF-α vs. 657 ± 93 −/L16 + TNF-α). Pretreatment with estrogen (E2) abrogated this increase in both genotypes, but G-1 pretreatment had a smaller effect on reducing THP-1 adhesion in hemizygous −/L16 compared with −/P16 (208.8 ± 66 for −/P16 vs. 437 ± 37 for −/L16, P < 0.05 by paired t test). These data suggest that whereas the GPER1 SNP may alter inflammatory pathways that could increase risk for high blood pressure, it could be corrected with estrogenic agonists.

Fig. 8.

Effects of G protein-coupled estrogen receptor (GPER) single-nucleotide polymorphism (SNP) expression and deletion on monocyte adhesion to induced pluripotent stem cell (iPSC)-derived endothelial cells (iECs). A: when monocyte cell line (THP-1) adhesion to iEC monolayer was evaluated, GPER-null (−/−) iECs displayed increased rates of THP-1 adhesion compared with unedited (P/L16) patient iECs. Adhesion of THP-1 monocytes to hemizygous normal (−/P16) or hemizygous polymorphic (−/L16) iEC monolayers stimulated with TNF-α revealed that polymorphic hemizygous iECs had a statistically significant increase in THP-1 adhesion compared with normal allele, and this increase could be diminished with pretreatment of G-1 or E2. B: quantitation of adhered, labeled THP-1 cells to iEC monolayers. C: representative images of bright-field and FITC channels for labeled THP-1 adhesion on an iEC monolayer. HTN-iECs, iECs from a patient with hypertension. *P < 0.05 vs. basal, #P < 0.05 vs. vehicle + TNF-α for the same genotype. %P < 0.5, −/L16 + G-1 vs. −/P16 + G-1 determined by separate t test.

DISCUSSION

New tools are needed to model hypertension, which is a polyfactorial pathology involving several tissues and thus is difficult to personalize (25, 34). Our study here demonstrates that we can utilize patient-derived iPSCs to establish an in vitro model of vascular cell assays to investigate genotype-phenotype relationships in vascular cell pathobiology. Our model of GPER deletion in patient-sourced HTN-iECs is a novel approach to understand the effects of estrogen receptor signaling in hypertension. This system has the potential to provide critical insights that recapitulate the genetic backgrounds of cohorts of patients with HTN while bypassing the need to procure patient tissue (a limited resource) with invasive methods.

Original studies of the GPER P16L SNP used adenoviral overexpression of GPER L16 in an immortalized rat vascular smooth muscle cell line (A7r5) to determine the functional consequence(s) of the L16 variant, which was a decreased rate of ERK1/2 phosphorylation (11). Feldman and colleagues also reported a modest, statistically significant increase in MABP in a candidate population of premenopausal Caucasian women (10, 11). On the basis of the initial findings, we postulated that a hypofunctional GPER1 SNP may lead to increased MABP through endothelial mechanisms.

Although we observed no striking effects of the L16 SNP on acute eNOS phosphorylation and subsequent NO formation, we uncovered that the L16 variant reduced the anti-inflammatory effects of GPER in iECs. qPCR data indicated that basal levels of inflammation genes MCP-1, VCAM, and VWF were similar between GPER genotypes but that TNF-α stimulation revealed an increase in inflammation-related gene expression in GPER1-null and −/L16 HTN-iECs. These data implicate an anti-inflammatory role for GPER and expand upon findings that GPER contributes to anti-inflammation mechanisms in the brain (36), diabetes (49), pulmonary arterial hypertension (51), and Crohn’s disease (53). However, GPER L16 HTN-iECs still responded to the GPER-specific agonist G-1, which mitigated the increase in mRNA expression of these inflammation genes. This study highlights the concomitant roles of GPER and endothelium in vascular inflammation, demonstrates a feasible research model to bridge mechanistic gaps underlying the GPER1 SNP in HTN risk, and suggests that the GPER SNP may associate with inflammatory gene transcription and could be counteracted with the GPER-specific agonist G-1. Although our in vitro iEC system showed that the GPER P16L SNP altered the endothelial cell inflammation response, the net manifestation of the functional effects of this SNP may be difficult to interpret depending on the health and age of the patient. The biological context of the GPER P16L SNP is vital to understanding the pathobiology of HTN and other cardiovascular diseases, such as age, cell or tissue type, and environmental factors such as exogenous damage to the endothelium from smoking or a high-fat diet (13, 35).

In vascular inflammation, lymphocyte rolling and extravasation to the vascular media are initiating events regulated by the endothelium. Our studies here demonstrated that GPER-null and SNP-expressing iECs had increased monocyte adhesion. These results are in line with previous studies that have shown that GPER1 knockout mice have elevated macrophage and T cell infiltration in Paigen diet-induced atherosclerosis (40). Given the longevity of HTN disease progression, our findings suggest that even a modest elevation of inflammation in the endothelium is a likely mechanism contributing to increased MABP and risk for developing resistant HTN in the Feldman discovery cohort and the PEAR cohort, respectively.

GPER has been previously demonstrated to localize at Golgi and endoplasmic reticulum membranes in human endothelial cells, human and mouse vascular smooth muscle cells, and human breast cancer cell lines (46). Our gene-edited iECs exhibited a cytosolic dispersion of GPER protein similar to what has been shown in primary and immortalized endothelial cells (5). Recently, Pupo et al. demonstrated nuclear sequestration of GPER variant protein in L16 when overexpressed by transfection in breast cancer cells (45). Using our model of hemizygous iECs endogenously expressing only L16 or P16, our data did not recapitulate nuclear sequestration of the L16 variant. Given the marked difference between cell types and expression systems used to study the GPER1 SNP, we are unable to fully resolve a possible mechanistic explanation for the contrasting data between our studies with those of Pupo et al.

The composition of the discovery cohort for GPER P16L was mostly young healthy Caucasian women, whereas the PEAR patients from whom our iPSCs were derived were postmenopausal women over the age of 45 and already presented with progressive HTN (19), yet still were carriers of the GPER SNP. We present evidence that iPSC-derived vascular cells may help predict SNPs that will associate with risk of HTN. Given the numerous observations that CVD increases exponentially in postmenopausal women (2, 7, 34, 38, 54), it is critical to understand the role of estrogen receptors in mediating such changes that promote CVD in women and necessitate a personalized approach to CVD treatment in women.

Since the GPER1 SNP was observed to influence MABP in women but not men, we selected iPSCs from female donors to investigate mechanistic implications. However, many studies demonstrate GPER expression and functionality in arteries and tissues from both male and female mice and humans, albeit with sex-dependent variations in function and phenotype, such as eliciting protective effects against stroke in male but not female rats, inducing similar vasodilation in healthy male and female mouse arteries, and exhibiting a maximal vasodilation effect in postmenopausal human female arteries (4, 8, 38, 49). The presence of estrogen receptors in male tissue suggests that estrogen signaling may play a role in systemic tissue when endogenous testosterone is converted to E2 via constitutive aromatase activity in the relevant tissues. However, it is not possible to determine whether the GPER SNP rs11544331 may play a role in cardiovascular function in men with the currently available data.

Our endeavor to articulate mechanistic differences between hemizygous iEC models of the GPER1 SNP revealed subtle trends of difference in NO production that did not translate to statistically significant differences using acute agonist targeting (G-1). We hypothesize that long-term disruption of NO generation may be necessary to articulate a significant difference, which was not possible with our in vitro system. Further to the limits of our model system, although we did observe marked changes in gene expression of inflammation targets at the mRNA level, that did not directly translate to a dramatic increase in THP-1 cell adhesion to monolayers of polymorphic iECs. It is possible that the mRNA increase in MCP-1 expression did not immediately translate to a larger increase in protein for MCP-1, which could possibly explain our functional readout (at least for the signaling pathway of MCP-1) of THP-1 adhesion. However, the effect of G-1 pretreatment on polymorphic iEC monolayers for THP-1 adhesion was smaller compared with the G-1 protective effect in normal allele hemizygous iEC monolayers, suggesting that the GPER polymorphic SNP may be hypofunctional in the endothelial cell inflammatory response. A lower dose of TNF such as 1 ng/mL could be effective in revealing a subtler difference in this response. In any case, perhaps it is not particularly surprising that our model system did not present dramatic differences in cellular function, because the majority of SNPs implicated in HTN risk or prognosis account for ~20% of a HTN phenotype (50). Hence, our results here are more in agreement with an expected subtle change in vascular function and HTN phenotypes, as opposed to a dramatic change in cell function from a single SNP, which would seem illogical to expect in the context of hypertension and its polyfactorial origins. Nevertheless, our GPER1-null iECs clearly implicated a functional role of GPER, which serves as a conduit to understanding discrete molecular functions of GPER in the endothelium of patients with HTN. We maintain that further evaluations would be needed to discretely propose whether the polymorphic allele of GPER results in HTN risk through a specific mechanism(s) of action. However, the present study does offer insights into GPER’s role in hypertension using human genetics. Future studies must address the role of GPER and SNP(s) in additional cell types from patients with HTN, such as vascular smooth muscle cells or cardiomyocytes, where GPER has been documented to play a role (27, 28, 30, 39). Finally, this study provides additional validation that gene editing of iPSCs from patient biobanks is a valuable method to advance mechanistic studies of genome-wide association study (GWAS) findings. The ability to utilize CRISPR gene editing and iPSC tools to investigate functional alterations of SNPs in a cell-specific manner will undoubtedly clarify our understanding of HTN etiology and beckon new avenues for clinical utility and context of large genomics and pharmacogenomics studies.

GRANTS

N.T. was supported in part by National Institutes of Health (NIH) National Institute of General Medical Sciences Grant (NIGMS) GM119977 and National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Grant DK104194. N.T. and N.C.F were supported by American Heart Association Grant-in-Aid 16GRNT30980002, and NIH Grant UL1TR001427 to University of Florida Clinical and Translational Science Institute. N.C.F. is a recipient of NIDDK postdoctoral fellowship (Grant T32 DK074367). E.R.P. was supported by NIH National Cancer Institute (NCI) Grants CA163890 and CA194496, NIH NIGMS Grant P20 GM121176 to the Center of Biomedical Research Excellence in Autophagy, Inflammation and Metabolism, and NIH NCI Grant P30 CA118100 to the University of New Mexico Comprehensive Cancer Center.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

N.C.F. and N.T. conceived and designed research; N.C.F., E.M., and M.L.M. performed experiments; N.C.F., E.M., M.L.M., and E.R.P. analyzed data; N.C.F., E.M., M.L.M., E.R.P., and N.T. interpreted results of experiments; N.C.F., E.M., and M.L.M. prepared figures; N.C.F. drafted manuscript; N.C.F., E.R.P., and N.T. edited and revised manuscript; N.C.F., E.M., M.L.M., E.R.P., and N.T. approved final version of manuscript.