Epigallocatechin Gallate Facilitates Extracellular Elastin Fiber Formation in Induced Pluripotent Stem Cell Derived Vascular Smooth Muscle Cells for Tissue Engineering

Abstract

Tissue engineered vascular grafts possess several advantages over synthetic or autologous grafts, including increased availability and reduced rates of infection and thrombosis. Engineered grafts constructed from human induced pluripotent stem cell derivatives further offer enhanced reproducibility in graft production. One notable obstacle to clinical application of these grafts is the lack of elastin in the vessel wall, which would serve to endow compliance in addition to mechanical strength. This study establishes the ability of the polyphenol compound epigallocatechin gallate, a principal component of green tea, to facilitate the extracellular formation of elastin fibers in vascular smooth muscle cells derived from human induced pluripotent stem cells. Further, this study describes the creation of a doxycycline-inducible elastin expression system to uncouple elastin production from vascular smooth muscle cell proliferative capacity to permit fiber formation in conditions conducive to robust tissue engineering.

1. Introduction

Tissue engineered vascular grafts (TEVGs) hold great potential to treat cardiovascular diseases as an effective alternative to autologous or synthetic vascular grafts. To date, mechanically robust TEVGs have successfully been generated from primary vascular smooth muscle cells (VSMCs) [1] and fibroblasts [2], as well as from VSMCs derived from induced pluripotent stem cells (iPSC-VSMCs) [3–5]. In particular, utilizing human iPSCs (hiPSCs) as a source for vascular tissue engineering is promising, as this cell type is self-renewable and capable of differentiating into VSMCs, endothelial cells, and fibroblasts. hiPSC-TEVG constructs could therefore be made readily available for large diameter (≥6 mm) vascular interventions, as well as in small diameter (2–4 mm) bypass applications following endothelialization to prevent coagulation.

Despite recent advances in TEVG generation, a major obstacle remaining is the ability to form elastic fibers in these tissues [6,7]. Elastin is the principal load-bearing component of the arterial wall, and endows blood vessels with the recoil necessary to return vessels to their original configurations following distension via hemodynamic flow. Further, extracellular elastin fibers are known to promote maintenance of the contractile phenotype of VSMCs [8,9], which is critical to prevent neointimal hyperplasia resulting from over-proliferation of synthetic VSMCs [10,11] (Fig. 1A).

Figure 1. EGCG Facilitates Elastin Fiber Assembly and Stability in hiPSC-VSMCs.

(A) Overview of the significance of elastin fiber production in tissue engineered vessels. As both disease and the natural process of aging lead to the degradation and fragmentation of elastin in the arterial wall, humans will experience arterial stiffening, hypertension, and an increased risk of stenosis or aneurysm. Utilizing EGCG in the production of TEVGs from hiPSC-VSMCs could ultimately lead to elastin enriched vessels for clinical transplantation, providing an ideal and universal model for these procedures. (B) Schematic of proposed mechanism of action of EGCG-facilitated elastin fiber formation. Upon secretion to the extracellular space, EGCG can coordinate monomeric tropoelastin through polar interactions between positively charged lysine residues on the protein and hydroxyl side chains on the polyphenol, allowing time for coacervation and productive deposition onto microfibrillar scaffolds for crosslinking by lysyl oxidase into insoluble elastin polymers. (C) Immunofluorescence imaging of phalloidin stained intracellular filamentous actin (red) and elastin (green) in hiPSC-VSMCs cultured in DMEM medium containing 3% FBS and 1 ng/mL TGF-β with and without addition of 7.5 μg/mL EGCG, and with and without decellularization. DNA nuclei were counterstained by DAPI. Scale bar: 200 μm. (D) TEM images of hiPSC-VSMCs with and without addition of 7.5 μg/mL EGCG. White arrowheads indicate areas of coated fibers in the extracellular space between cells. Scale bar: 5 μm (left) and 3 μm (right). (E) Quantification of insoluble elastin protein by the Fastin assay (Two-tailed, unpaired Student’s t-test was performed; Mean values and S.E.M. indicated by the error bars are shown; n = 3; *: p<0.05). (F and G) qRT-PCR quantification of relative fold change in gene expression between hiPSC-VSMCs cultured with and without addition of 7.5 μg/mL EGCG for lysyl oxidase (LOX), fibrillin-1 (FBN-1), fibulin-4 (FBLN4), fibulin-5 (FBLN5), elastin (ELN), and matrix metalloproteinases 2, 9, and 12 (MMP2, MMP9, and MMP12) (Two-tailed, unpaired Student’s t-test was performed; Mean values and S.E.M. indicated by the error bars are shown; n = 4 for LOX and FBN1, n = 3 for other markers; *: p<0.05; ***: p<0.001; ns: not significant).

Insoluble elastin polymers are formed in the extracellular matrix (ECM) following secretion of soluble tropoelastin monomers from the cell, coacervation at the cell surface into aggregate spherules, deposition of these spherules onto microfibrillar fibrillin-1 glycoprotein scaffolds, and crosslinking of lysine-enriched domains within the tropoelastin monomers by lysyl oxidase [12]. Further, the ECM proteins fibulin-4 and fibulin-5 play essential roles in coordinating this assembly process, maintaining the integrity of nascent elastin fibers, and in maturation of these fibers into higher order organizational networks [12–14]. Deficiency in these elastin complex proteins can lead to perinatal lethality, as well as a propensity for developing ascending or thoracic aortic aneurysms [15–17], neointimal hyperplasia [18], cutis laxa [13,19–21], or stenosis [8]. Although the elastin assembly process occurs effectively in vivo, this process has proved difficult to replicate under in vitro settings.

Notably, there has been significant progress in in vitro elastogenesis in primary VMSCs [22,23] and fibroblasts [24–27], though due to donor-to-donor functional variability, as well as limitations in the access and finite expandability of primary cells, stem cell derived populations may ultimately provide a more effective source for vascular tissue engineering [28]. Importantly, synthesis of robust elastin fiber formation in hiPSC-VSMCs has previously been shown in a quiescent medium containing 0.5% fetal bovine serum (FBS) and 1 ng/mL TGF-β, demonstrating the efficacy of extracellular elastin deposition in vascular stem cell derivatives [29]. However, it should also be noted that such a low serum concentration is not ideal for tissue engineering purposes, in which high serum concentration is necessary to promote cellular proliferation for scaffold repopulation [4].

Here we demonstrate that the polyphenol compound epigallocatechin gallate (EGCG) facilitates the synthesis of extracellular elastin fibers in vitro in tissue engineering grade hiPSC-VSMCs under conditions of elevated serum concentration. EGCG is a principal component of green tea, and has been shown to convey several vascular protective and cardiovascular benefits [30], including antioxidant [31–34], reduced lipid absorption [35,36], anti-inflammatory [37,38], anti-hypertensive [39,40], anti-thrombogenic [41,42], and anti-proliferative effects [43–46], the last of which has led to additional research into EGCG as an anti-cancer agent [47–49]. Several of these effects are conducive to the synthesis and maturation of elastic fibers, including the activation of endothelial nitric oxide synthase [39], as nitric oxide has previously been shown to stimulate elastin expression in VSMCs [50]. Indeed, EGCG or related polyphenolic compounds have been demonstrated to possess elasto-regenerative and elasto-protective properties in primary VSMC and dermal fibroblast in vitro culture systems [51,52]. EGCG has been shown to increase the rate of tropoelastin coacervation in primary cell culture [51], and is thought to do so through coordinating beneficial polar interactions between positively charged lysine residues on tropoelastin and the hydroxyl side chains of the polyphenol (Fig. 1B). The present study leverages these previous findings to translate EGCG-dependent extracellular elastin synthesis to hiPSC-VSMC culture, setting the stage for the future development of elastin-enriched hiPSC-TEVGs for clinical application (Fig. 1A).

2. Results

2.1. EGCG Facilitates Extracellular Elastin Fiber Formation in hiPSC-VSMCs

VSMCs were differentiated from hiPSCs following established protocols from our group [3–5,53]. For assessing extracellular elastin fiber formation, hiPSC-VSMCs were cultured in medium containing 3% FBS supplemented with 1 ng/mL TGF-β. This medium was chosen as TGF-β is known to enhance elastin gene expression and prolong elastin mRNA half-life [54–57], and builds on previous successful generation of elastin using 0.5% FBS with 1 ng/mL TGF-β [29], with the goal of steadily increasing serum concentration for better efficacy for tissue engineering [4]. An EGCG dose of 7.5 μg/mL was selected for elastin fiber formation experiments based on the results of a TUNEL cytotoxicity assay (Fig. S1). hiPSC-VSMCs were cultured with either DMSO vehicle or 7.5 μg/mL EGCG for 12 days, and subsequently fixed or decellularized [58] for immunofluorescence imaging or electron microscopy. In contrast to untreated hiPSC-VSMCs, EGCG-supplemented cells displayed abundant extracellular elastin fibers by both immunofluorescence and TEM (Fig. 1C, D). Quantification using the Fastin assay similarly revealed a significant increase in insoluble elastin present in the cell membrane or ECM of EGCG-treated hiPSC-VSMCs compared to untreated controls (Fig. 1E).

Quantitative PCR analysis further revealed ancillary effects of EGCG treatment complementary to elastin fiber synthesis. Namely, hiPSC-VSMCs treated with EGCG demonstrated significantly elevated expression of the elastin crosslinking enzyme lysyl oxidase (LOX), the scaffold glycoprotein fibrillin-1 (FBN1), and of the essential accessory ECM proteins fibulin-4 (FBLN4) and fibulin-5 (FBLN5) (Fig. 1F). Further, EGCG addition served to significantly decrease the expression of elastin-degrading matrix metalloproteinases (MMPs) 2, 9 and 12 (Fig. 1G), suggesting an elasto-protective downstream effect of EGCG treatment [51,59]. EGCG similarly downregulated candidate genes involved in cancer and inflammatory pathways, including NF-κB, DNMT-1, HDAC-1, IL-6, and VEGF-A, as suggested in the literature (Fig. S2) [48]. Note that elastin gene expression was unaffected by EGCG addition (Fig. 1F), suggesting that there may be positive feedback associated with enhanced coacervation by EGCG, leading to the observed increased expression of elastin complex factors, rather than a direct increase in tropoelastin itself. Together, these results suggest that EGCG addition facilitates the biogenesis of elastin fibers in vitro through coordination and stabilization of secreted tropoelastin monomers, upregulating the expression of factors essential to the development and maturation of elastin, and downregulating MMPs which would degrade nascent elastin fibers.

2.2. EGCG-Treated Vascular Tissue Rings Indicate Enhanced Elasticity

Vascular tissue rings were generated from hiPSC-VSMCs following established protocols from our group [5,53,60]. Rings were cultured for 12 days in serum reduced, elastin-promoting medium supplemented with either DMSO vehicle or 7.5 μg/mL EGCG, and subjected to mechanical assessment (Fig. 2A). Tensile stress testing revealed a significantly increased failure strain and a decreased Young’s modulus for vascular tissue rings treated with EGCG compared to vehicle (Fig. 2B). Both of these parameters are indicative of an enhanced elasticity and decreased stiffness, setting the stage for future production of elastin-enriched hiPSC-TEVGs (Fig. 1A). Interestingly, the maximum tensile stress was lower in EGCG-treated rings compared to vehicle (Fig. 2B), potentially due to the anti-proliferative effects of EGCG [43–46] and shorter time course for this proof-of-principle experiment compared to the 8-week growth period of hiPSC-TEVGs [4].

Figure 2. Mechanical Assessment of EGCG Treated Vascular Tissue Rings.

(A) Representative tensile stress versus strain plots from DMSO vehicle (left) and EGCG treated (right) hiPSC-VSMC tissue rings. (B) Maximum tensile stress (left), failure strain (middle), and Young’s modulus (right) quantification for vascular tissue rings treated with either vehicle or EGCG (Two-tailed, unpaired Student’s t-test was performed; Mean values and S.E.M. indicated by the error bars are shown; n = 3; *: p<0.05; **: p<0.01).

2.3. Creation of a Doxycycline-Inducible Construct for Elastin Expression

In order to generate robust TEVGs from hiPSC-VSMCs, the cells must be provided growth factors and nutrition to promote proliferation and collagen deposition [4]. However, elastin mRNA levels are known to be coupled to and inversely correlated with the proliferative state of VSMCs, with cells in G0 or growth arrest displaying maximal mRNA expression [61]. In order to uncouple these processes, thereby allowing efficient elastin mRNA expression at any stage of the cell cycle, we designed a doxycycline-inducible construct for elastin expression (Fig. 3A). Using this system, the addition of doxycycline to the culture medium would allow constitutive expression of elastin mRNA, maintaining elevated steady-state expression of these transcripts, and subsequently impacting the steady-state expression of tropoelastin protein for assembly into extracellular fibers [62,63].

Figure 3. Creation and Efficacy of a Doxycycline-Inducible Elastin Expression Construct and Cell Line.

(A) Overview of elastin expression under various conditions. Under low serum, with VSMCs exhibiting a contractile, mature phenotype, endogenous elastin is well expressed, but this expression is greatly reduced under conditions of high serum, where VSMCs switch to a synthetic, proliferative phenotype conducive to robust TEVG generation. Similarly, under high serum conditions in the absence of DOX, the transgenic elastin promoter will be inactive. However, upon the addition of DOX, transgenic elastin will be expressed even under high serum conditions. (B) qRT-PCR quantification of relative fold change in expression of total (transgenic and endogenous) elastin in both parental and DOX-ELN hiPSCs treated with and without 2 μg/mL doxycycline (One-way ANOVA with Tukey’s multiple comparisons test; Mean values and S.E.M. indicated by the error bars are shown; n = 3; **: p<0.01). (C) Immunofluorescence imaging for VSMC markers α-smooth muscle actin (SMA) and calponin (CNN1) of DOX-ELN hiPSC-VSMCs cultured in 1% FBS-supplemented DMEM maturation medium. DNA nuclei were counterstained by DAPI. Scale bar: 200 μm. (D) qRT-PCR quantification of relative fold change in expression of total (transgenic and endogenous) elastin in both parental and DOX-ELN hiPSC-VSMCs treated with and without 2 μg/mL doxycycline (One-way ANOVA with Tukey’s multiple comparisons test; Mean values and S.E.M. indicated by the error bars are shown; n = 3; **: p<0.01). (E) Representative western blot images of tropoelastin protein expression levels normalized to GAPDH in DOX-ELN hiPSC-VSMCs treated with and without 2 μg/mL doxycycline. (F) Quantification of normalized tropoelastin protein expression in DOX-ELN hiPSC-VSMCs treated with and without 2 μg/mL doxycycline (Two-tailed, unpaired Student’s t-test was performed; Mean values and S.E.M. indicated by the error bars are shown; n = 3; *: p<0.05).

Elastin cDNA was cloned into the doxycycline-inducible vector AAVS1-TRE3G-EGFP for ectopic expression. This vector targets the AAVS1 safe-harbor gene locus, an ideal site for robust transgene expression in hiPSCs and their derivatives [64], using a transcription activator-like effector nuclease (TALEN) gene editing approach. The vector also contains a puromycin resistance gene, allowing for antibiotic selection of edited clones. Insertion of elastin cDNA into the vector was confirmed via restriction digestion (Fig. S3A). Using hiPSC-specific electroporation, this doxycycline-inducible elastin expression vector was transfected into dissociated, single-cell hiPSCs, utilizing puromycin selection to select for TALEN-edited clones. Puromycin resistant hiPSC colonies obtained from electroporated single cells were expanded and analyzed for presence of the construct via PCR at the AAVS1 locus (Fig. S3B). Results show PCR amplicons for the parental cell line only for the unedited, native AAVS1 locus, while edited hiPSCs containing the doxycycline-inducible elastin expression vector (DOX-ELN hiPSCs) show DNA bands only for the elastin cDNA insert at the AAVS1 locus (Fig. S3C). Karyotyping of DOX-ELN hiPSCs was performed and revealed normal chromosomal banding patterns (Fig. S3D). Further, normal expression of the pluripotency markers OCT4, Nanog, and TRA-1-60 was confirmed in DOX-ELN hiPSCs (Fig. S3E). Additionally, DOX-ELN hiPSCs cultured in the presence of 2 μg/mL doxycycline demonstrated nearly 3000-fold greater induction of elastin mRNA measured by qRT-PCR compared to untreated cells and empty vector controls (Fig. 3B), demonstrating the efficacy of the system.

2.4. Expression of Elastin is Inducible in DOX-ELN hiPSC-VSMCs

DOX-ELN hiPSCs were subsequently differentiated into VSMCs following established protocols from our group [3–5,53], with additional supplementation of 10 μg/mL ascorbic acid, to further enhance the proliferative capacity of hiPSC-derived progenitors [65]. VSMC marker expression of α-smooth muscle actin and calponin was confirmed by immunofluorescence in DOX-ELN hiPSC-VSMCs (Fig. 3C). Inducible expression of elastin mRNA was assessed via qRT-PCR following treatment with or without 2 μg/mL doxycycline. Results show that DOX-ELN hiPSC-VSMCs demonstrated a 14-fold greater induction of elastin mRNA over baseline elastin expression in the absence of doxycycline (Fig. 3D), which was comparable to elastin expression in the empty vector control, indicating the inducible expression is tightly controlled by the presence of doxycycline. As the baseline expression of elastin mRNA would be expected to be much greater in differentiated cells than at the stem cell stage, it is promising to see the AAVS1-targeted construct maintain efficacy in differentiated cell types. Further, western blotting revealed close to 4-fold increased abundance of tropoelastin protein in DOX-ELN hiPSC-VSMCs treated with doxycycline compared to cells treated with vehicle (Fig. 3E, F), indicating that elevated elastin mRNA transcript expression indeed led to increased translation of tropoelastin [62,63].

2.5. Extracellular Elastin Synthesis in Serum Enriched Culture Medium

We next sought to leverage our EGCG-facilitated elastin fiber synthesis with our DOX-ELN hiPSC-VSMCs to augment elastin deposition in serum enriched culture medium conducive to TEVG generation. DOX-ELN hiPSC-VSMCs were cultured with either 7.5 μg/mL EGCG alone or both EGCG and 2 μg/mL doxycycline in medium supplemented with either 3%, 5%, or 10% FBS and 1 ng/mL TGF-β, or in a 20% FBS-supplemented collagen promoting medium previously utilized for hiPSC-TEVG generation by our group [4]. Vehicle only and doxycycline only controls were additionally used for comparison (Fig. S4). TGF-β was included in all cases due to previously established elasto-generative and elasto-protective effects [54–57]. Results demonstrate that although there was notable extracellular elastin deposition in both EGCG alone and EGCG with doxycycline conditions in the 3% FBS and TGF-β culture medium (Fig. 4, leftmost panels), replicating earlier results with the parental hiPSC-VSMC line (Fig 1C), this was not the case for the higher serum cultures. For DOX-ELN hiPSC-VSMCs cultured in the presence of 5%, 10%, or 20% FBS-supplemented media, only the combination of EGCG and doxycycline produced extracellular elastin deposition (Fig. 4), while EGCG alone under serum enriched conditions produced matrices similar to an untreated control (Fig. 1C, Fig. S4). These results suggest that our doxycycline-inducible construct for elastin expression effectively produces tropoelastin protein in hiPSC-VSMCs, even during growth stages of the cell cycle [61], thereby providing EGCG sufficient substrate for extracellular elastin fiber formation.

Figure 4. Extracellular Elastin Expression in DOX-ELN hiPSC-VSMCs.

Representative immunofluorescence imaging of decellularized matrices from DOX-ELN hiPSC-VSMCs treated with either 7.5 μg EGCG alone (top row) or both 7.5 μg EGCG and 2 μg/mL doxycycline (bottom row). Columns represent increasing serum enrichment of the culture medium from 3% FBS to 20% FBS to assess the efficacy of the construct. All serum conditions were supplemented with 1 ng/mL TGF-β. Scale bar: 200 μm.

3. Discussion

This study establishes the utility of the polyphenol EGCG in ECM elastin fiber formation in hiPSC-VSMCs for future tissue engineering. Elastin is predominantly produced during the late embryonic and neonatal periods of life, with dramatically reduced expression throughout adulthood [66]. With increasing age, elastin fibers undergo natural degradation, resulting in stiffer arteries and clinical hypertension [67,68]. Therefore, engineering hiPSC-derived vessels enriched with elastin fibers would be highly beneficial for vascular graft development in both disease contexts and for the general aging population (Fig. 1A). As EGCG possesses numerous cardiovascular benefits [30,39,40], including anti-oxidant [31–34] and anti-inflammatory [37,38] activity, which would otherwise exacerbate the degradation of elastin, it serves as an optimal compound for facilitating elastin fiber formation in hiPSC-TEVGs.

We further developed a doxycycline-inducible elastin expression vector to be used in tandem with EGCG treatment. Although previous work demonstrated elastin fiber deposition in hiPSC-VSMCs cultured in 0.5% FBS with 1 ng/mL TGF-β [29], this medium is highly restrictive for cell growth, impeding its utility for vascular tissue engineering [4]. As the expression of elastin is inversely correlated to the growth state of VSMCs [61], and EGCG treatment showed no significant increase in elastin expression (Fig. 1F), we sought to uncouple elastin expression from proliferative status through constitutive expression via doxycycline induction. We constructed a doxycycline-inducible system using elastin cDNA and genetically edited this into single-cell hiPSCs at the AAVS1 safe-harbor locus [64]. Doxycycline-stimulated expression of elastin was apparent in both edited DOX-ELN hiPSCs (Fig. 3B) and VSMC derivatives at both the mRNA and protein level (Fig. 3 D–F). We then combined our EGCG discovery with the DOX-ELN hiPSC-VSMCs to assess the ability to synthesize ECM elastin in progressively increasing concentrations of serum, thereby establishing a solid foundation for use in future tissue engineering. We observed that EGCG treatment alone was insufficient to produce extracellular elastin fibers at serum enriched conditions, more conducive to cell growth (Fig. 4, top panels). However, upon activating the expression of elastin mRNA and protein with 2 μg/mL doxycycline, EGCG treatment effectively led to the appearance of extracellular elastin (Fig. 4, bottom panels), perhaps suggesting a baseline abundance of tropoelastin may be required for EGCG-facilitated elastin fiber expression to occur. This is critical information for future tissue engineering efforts, as VSMC growth and collagen deposition are essential for robust TEVG generation, but often occur at the cost of elastin expression [4,61]. Importantly, the relative intensity of elastin immunofluorescence in the EGCG with doxycycline treatment condition in 20% FBS-supplemented collagen promoting medium appeared lower than in the 5% and 10% FBS with TGF-β cultures (Fig. 4), suggesting that while the combination of EGCG and doxycycline appears promising for elastin expression in tissue engineering, further optimization could be pursued to enhance this effect.

Our results also suggest an enhanced elasticity in vascular tissue rings treated with EGCG, due to an increased failure strain and decreased Young’s modulus (Fig. 2). Although these rings show a reduced maximum tensile strength, this could likely be attributed to the lower serum conditions (3% vs. 20% FBS) and shorter time course (12 days vs. 8 weeks) of these proof-of-principle ring experiments compared to hiPSC-TEVGs, as well as potential anti-proliferative effects of EGCG [43–49]. As such, utilizing our doxycycline-inducible construct and high serum for a longer time course may likely overcome the anti-proliferative effects of EGCG and build mechanically robust, elastic hiPSC-TEVGs. Further, dynamic culture using EGCG is likely to result in enhanced organization of ECM deposited elastin fibers [4,29,69], and could be explored in the future. Additionally, it is expected that the EGCG approach presented here could be combined with the approaches for in vitro generation of elastin in primary VSMCs in follow-up studies for enhanced success [22–27]. Ultimately, following recent efforts at modulating the expression of human leukocyte antigens (HLAs) of hiPSCs [70–72], thereby making these cells and their derivates hypoimmunogenic, elastin-enriched, mechanically robust hiPSC-TEVGs suitable for clinical transplantation to any patient recipient could be produced.

9. Methods

9.1. Derivation of hiPSC-VSMCs

hiPSC-VSMCs were obtained following an established embryoid body (EB)-based approach in our group [3–5,53]. Briefly, hiPSCs were expanded on growth factor reduced (GFR)-Matrigel-coated tissue culture plates (Corning) until 80% confluency, and subsequently dissociated with ReLeSR (Stemcell Technologies), resuspended in mTeSR1 (Stemcell Technologies) supplemented with 1:100 (v/v) GFR-Matrigel and 5 μM ROCK inhibitor (Y-27632; Millipore), and transferred to a 6-well low attachment plate (Corning). Overnight incubation in this medium allowed the formation of EBs of uniform size, which were progressively mixed with an EB medium [DMEM high glucose (ThermoFisher) supplemented with 10% FBS (Gemini), 2 mM L-glutamine (ThermoFisher), 1% (v/v) NEAA (ThermoFisher), 1% (v/v) pen/strep (ThermoFisher), and 0.012 mM β-mercaptoethanol (Sigma Aldrich]), as previously described. After 6 days in suspended culture, EBs were collected and seeded onto gelatin-coated culture dishes for an additional 6 days in EB medium. To induce the cells to a VSMC lineage, adherent EB-derived cells were dissociated by 0.05% Trypsin-EDTA (ThermoFisher), seeded onto GFR-Matrigel-coated dishes at 20,000 cells/cm², and cultured in SmGM-2 medium (Lonza) expansion medium. To further induce the maturation phenotype of hiPSC-VSMCs for characterization of marker expression, cells were subcultured in VSMC maturation medium [DMEM with 1% FBS, 1% (v/v) NEAA, 2 mM L-glutamine, and 1 ng/mL TGF-β (Peprotech)] for six days.

9.2. Generation of Extracellular Elastin

To produce extracellular elastin in vitro, hiPSC-VSMCs were cultured in a serum-reduced elastin-promoting culture medium [DMEM with 3% FBS, 1% (v/v) pen/strep, 1 ng/mL TGF-β, and 7.5 μg/mL EGCG (Sigma Aldrich)] for 12 days. In order to maximize the effects of EGCG addition, while providing sufficient nutrition to the cells, medium was replenished once every 72 hours.

9.3. Decellularization

Following established protocols [58], cell monolayers were removed from the ECM using a decellularization buffer [DPBS with Ca²⁺ and Mg²⁺ (ThermoFisher), 40 mM NH-₄OH (JT Baker Chemical Company), and 0.5% (v/v) Triton-X 100 (AmericanBIO)]. Cells were washed with DPBS with Ca²⁺ and Mg²⁺, then submerged in decellularization buffer and incubated for 2 minutes at 37°C. This buffer was then discarded and decellularized matrices were gently washed with DPBS with Ca²⁺ and Mg²⁺ twice for 5 minutes each, with the first of these incubations occurring at 37°C. Matrices were then incubated overnight at 4°C in DPBS with Ca²⁺ and Mg²⁺ supplemented with 1:1000 DNase (Roche) to remove any remaining debris.

9.4. Immunofluorescence

Cells were washed with DPBS (ThermoFisher) and fixed in 4% paraformaldehyde (PFA; Electron Microscopy Sciences, Hatfield, PA) for 10 minutes at room temperature. Cells were subsequently blocked using 10% normal goat serum (NGS; ThermoFisher) in PBST (DPBS with 0.3% Triton X-10) for 30 minutes at room temperature. These blocked cells were then incubated with primary antibody in 1% NGS in PBST at 4°C overnight, before being washed with DPBS and incubated for 1 hour at room temperature, protected from light, with a secondary antibody (1:1000 in 1% NGS in PBST), then washed with DPBS again prior to imaging on an inverted fluorescent microscope (Leica). Primary antibodies for immunofluorescence include elastin (1:100; Abcam, ab21610), OCT4 (1:200; Abcam, ab18976), Nanog (1:200; Peprotech, 500-P236), TRA-1-60 (1:200; Chemicon, MAB4360), α-smooth muscle actin (1:100; Sigma, A5228), and calponin (1:100; Sigma, C2687). Secondary antibodies include Alexa 488 goat anti-rabbit IgG (ThermoFisher, A11008) and Alex 555 goat anti-mouse IgG (ThermoFisher, A21426). Filamentous actin was stained with phalloidin (Alexa Fluor 594; ThermoFisher, A12381) and the nuclei were counterstained with DAPI (ThermoFisher).

9.5. TUNEL Cell Toxicity Assay

Cells were processed according to the TUNEL staining kit (Roche) manufacturer’s instructions. As a positive control, cells were pretreated with 10 U/mL recombinant DNase I for 10 minutes at room temperature. TUNEL positivity was assessed via fluorescence microscopy relative to the total number of cell nuclei.

9.6. Transmission Electron Microscopy

Cells for transmission electron microscopy (TEM) imaging were fixed in 2.5% Glutaraldehyde (EMS) in 0.1M Sodium cacodylate buffer at 7.4 pH for 30 minutes shaking at room temperature, followed by one hour at 4°C. Cells were subsequently rinsed 3 times in a 0.1M Sodium cacodylate buffer. The cells were postfixed with reduced osmium, containing 0.5% Osmium tetroxide (EMS) and 0.8% Potassium ferrocyanide in 0.1 M cacodylate buffer for 1 hour. They were then rinsed 3 times with 0.1 M Sodium cacodylate buffer followed by HPLC water. The cells were then stained en bloc with 2% aqueous Uranyl acetate (EMS) for 1 hour in the dark to prevent precipitation. The cells were rinsed 5 times with HPLC water and then incrementally dehydrated in 50%, 70%, 90%, 95% ethanol followed by 3 final changes in 100% anhydrous ethanol. The cells were slowly infiltrated with EPON resin (EMS) using an initial 1:1 (v/v) mixture of 100% anhydrous ethanol and EPON resin for ~30 mins before undergoing 2 changes in 100% EPON resin for ~ 1 hr. The cells were embedded in fresh EPON which was polymerized overnight at 60°C.

Ultrathin sections (60 nm thick) were cut on a Leica UC7 Ultramicrotome and mounted on formvar-coated Nickel hexagonal mesh grids and stained with 2% aqueous uranyl acetate followed by Reynold’s Lead citrate. The grids were imaged in an 80 kV FEI Tecnai G2 Spirit BioTWIN TEM (Thermo Fisher Scientific, Eindhoven, The Netherlands) and images were acquired using a SIS Morada 11-megapixel CCD camera.

9.7. Insoluble Elastin Protein Quantification

Cell cultures were lysed with RIPA buffer (ThermoFisher), and cellular and extracellular material was collected using cell scrapers (Biologix), sonicated, and centrifuged to separate soluble and insoluble material. Pelleted material was then subjected to two 1 hour hot 0.25 M oxalic acid treatments at 100°C to extract digested α-elastin soluble derivatives. Samples were then analyzed via the Fastin^™ assay (Biocolor), following the manufacturer’s instructions, with absorbances measured at 513 nm using a Synergy 2 multi-mode plate reader (BioTek).

9.8. Formation and Mechanical Assessment of Vascular Tissue Rings

Vascular tissue rings were fabricated using a custom agarose mold, as previously described [5,53,60]. Briefly, a 2% agarose solution was created in basal DMEM, and allowed to solidify in polycarbonate molds of 5 annular wells, each containing a 2 mm post. One million hiPSC-VSMCs were added per well and allowed to self-aggregate around each post overnight in SmGM-2 (Lonza) medium before being switched to the serum reduced elastin-promoting medium supplemented with or without 7.5 μg/mL EGCG. Medium was replenished every 72 hours for a total culture period of 12 days. The maximum tensile strain, failure strain, and Young’s modulus of vascular tissue rings were evaluated using an Instron 5960 microtester (Instron) with a 10 N load cell. The rings were mounted between two stainless steel pins, each anchored to the load cell. Each ring was cyclically pre-stretched three times at 10% strain, then progressively stretched until failure. Tensile force was normalized to cross-sectional area of the surfaces perpendicular to the applied force for each ring to yield the engineering stress of the tissue. Surface area of each face was approximated as a circle (A = π*r²) where the radius (r) is the wall thickness of the tissue ring. When calculating the engineering stress, the cross-sectional area of each ring is multiplied by two to consider both sides of the ring perpendicular to the applied tensile force. The Young’s modulus was subsequently determined from the plot of the linear region of each curve.

9.9. Quantitative Reverse Transcription PCR

RNA was extracted and purified from cell cultures using the TRIzol^™ RNA Isolation Kit (ThermoFisher), following the manufacturer’s instructions. Total RNA was then subjected to reverse transcription using the iScript^™ cDNA Synthesis Kit (Bio-Rad). Quantitative reserve transcription PCR (qRT-PCR) was performed using the SYBR Green Supermix and respective gene primers. Expression of genes of interest was normalized to that of human B2M. The primer sequences used for qRT-PCR are listed in Table 1.

Table 1:

List of Primers for qRT-PCR

Gene	Primer Sequence
LOX-F	AGATGTCCATGTACAACCTGAG
LOX-R	GTAATGTTGATGACAACTGTGCCA
FBN1-F	GAGTGCCTTGACAATCGGGA
FBN1-R	GATTTGGTGACGGGGTTCCT
FBLN4-F	AACCGCTCCTGTGTTGATGT
FBLN4-R	GATGCAGCGGTACTGACAGA
FBLN5-F	GAATAAAACACCCGCGAGCC
FBLN5-R	GCACTGTGCCTGTGCATTC
ELN-F	AAGATGGTGCAGACACTTCC
ELN-R	AGAGCGAATCCAGCTTTGAG
MMP2-F	CCTGACCAAGGGTACAGCC
MMP2-R	GTCAATGTCAGGAGAGGCCC
MMP9-F	AGCGAGAGACTCTACACCCA
MMP9-R	CGGAGTAGGATTGGCCTTGG
MMP12-F	AAGGCCGTAATGTTCCCCAC
MMP12-R	TGGGTCTCCATACAGGGACT
Total ELN for DOX-F	GCCACCATGGCGGGTCTG
Total ELN for DOX-R	CAAGGCCAGCACCCGCAAG
B2M-F	AGCAGCATCATGGAGGTTTGA
B2M-R	AGCCCTCCTAGAGCTACCTG
NF-κB-F	CTTAGGAGGGAGAGCCCAC
NF-κB-R	TGAAACATTTGTTCAGGCCTTCC
DNMT-1-F	CCGCAGGCGGCTCAAAG
DNMT-1-R	ACTTTAGCCAGGTAGCCCTCC
HDAC-1-F	AACTGCTAAAGTATCACCAGAGGG
HDAC-1-R	AGCCCCGATATCCCGTAGG
IL-6-F	GATGGATGCTTCCAATCTGG
IL-6-R	GTTCTGGAGGTACTCTAGGT
VEGF-A-F	AGGAGGAGGGCAGAATCATC
VEGF-A-R	CGATCTCATCAGGGTACTCC

9.10. Elastin cDNA Synthesis and Cloning into Dox-Inducible AAVS1-TRE3G-EGFP Vector

The cDNA of the abundantly expressed variant of the Elastin gene (ENST00000252034.12), encoding the full-length elastin protein of 724 amino acid residues, was synthesized by the OriGene Technologies, Inc. The cDNA sequence was verified by the Sanger sequencing using cDNA specific primers. MluI and SalI endonucleases restriction sites were introduced to the 5’ and 3’ UTRs, respectively, of the Elastin cDNA to facilitate cloning into the doxycycline-inducible vector, AAVS1-TRE3G-EGFP (Addgene plasmid # 52343), which contains a puromycin selection marker. The EGFP gene in the AAVS1-TRE3G-EGFP plasmid was replaced with the 2.193 kb elastin insert to generate the AAVS1-TRE3G-ELN plasmid using MluI and SalI restriction enzymes. Cloning of the elastin construct was confirmed through restriction digestion of the AAVS1-TRE3G-ELN plasmid using the MluI and SalI enzymes (Fig. S3A).

9.11. TALEN Gene Editing to Knock-in Elastin cDNA at the AAVS1 Safe Harbor Locus

TALEN gene editing was used to knock-in the doxycycline-inducible elastin cDNA into the human safe harbor AAVS1 locus. The DNA delivery of the two TALEN plasmids (left and right TALEN), and AAVS1-TRE3G-ELN plasmid into hiPSCs was accomplished using the Human Stem Cell Nucleofector® Kit 1 (Cat. No. VPH-5002, Lonza, Bioscience) as per the vendor instructions. Briefly, 0.8–1 million single-cell suspensions of hiPSCs were prepared in 100 μL of nucleofection solution. In total 5 μg of the plasmid DNA mixture was added to the cell suspension in the following ratio: 1.5:1.5:2, Left TAL: Right TAL: AAVS1-TRE3G-ELN. DNA electroporation was facilitated using the Nucleofector® Device as per the manufacturer’s instructions and electroporated cells were seeded into a Matrigel-coated well of a 6-well tissue culture plate and placed in a hypoxia chamber to improve single cell viability. The culture medium was replaced with fresh medium the next day. After 48 hours, the culture medium was replaced with medium supplemented with 0.5 μg/mL puromycin every day for 5 days to select for resistant clones. After this 5-day period, cells were cultured in puromycin-free medium until clones became visible, about two weeks post seeding. Puromycin resistant clones were then manually dissociated for further expansion and characterization to confirm integration of the elastin construct into the AAVS1 locus. PCR with custom-designed primers spanning the TRE3G vector backbone and the AAVS1 locus outside the left and right TALEN binding sequences was used to determine elastin construct integration at the AAVS1 locus (Fig. S3B, C). The primer sequences used for PCR are listed in Table 2.

Table 2:

List of Primers for PCR at AAVS1 Locus

Target Site	Primer Sequence
AAVS-Fw2 (Left TALEN Arm)	CCCAGGCAGGTCCTGCTTTC
AAVS-Rv2 (Left TALEN Arm)	AGAGATGGCTCCAGGAAATG
Backbone-Rv (Left TALEN Arm)	CAACAGATGGAAGGCCTCCT
AAVS-Fw3 (Right TALEN Arm)	CGGTTAATGTGGCTCTGGTT
AAVS-Rv3 (Right TALEN Arm)	AGGAGAATCCACCCAAAAGG
Backbone-Fw (Right TALEN Arm)	ATTAATTGCGTTGCGCTCAC

9.12. Karyotyping

Karyotype analysis of hiPSCs was performed by Yale Cytogenetic Services using a standard G-banding chromosome analysis according to standard procedures.

9.13. Western Blotting

Soluble proteins were extracted from cell cultures with direct lysis by RIPA buffer with cOmplete^™ Protease Inhibitor Cocktail (Roche) and Phosphatase Inhibitor Cocktail 2 (Sigma Aldrich) followed by sonication and centrifugation to remove cellular debris. Samples containing 15 μg total protein were loaded into SDS-PAGE gels and western blotting was carried out with PVDF Membranes (Bio-Rad). Transferred membranes were blocked with 5% Bovine Serum Albumin (BSA; Sigma Aldrich) in TBST (American BIO) then incubated with primary antibodies overnight at 4°C. Membranes were then washed with TBST and incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies. Following secondary antibody incubation, membranes were washed again with TBST and incubated with SuperSignal^™ West Pico PLUS Chemiluminescent Substrate (Thermo Fisher) for 5 minutes and protein bands were visualized on a BOX Chemi XX6, Gel Documentation and ECL Detection System. Samples were analyzed for quantification using the ImageJ software, and protein of interest expression was normalized to that of human GAPDH. Primary antibodies for western blotting include human aortic elastin (1:500; gift from Dr. Robert Mecham of Washington University in St. Louis) and GAPDH (1:10,000; Abcam, ab128915). The secondary antibody used was an HRP-conjugated goat anti-rabbit IgG (1:5,000; Vector Laboratories Inc, PI-1000).

Supplementary Material

1

Supplementary Figure S1. TUNEL cytotoxicity analysis for hiPSC-VSMCs treated with vehicle control, EGCG of varying dose (5, 7.5, and 10 μg/mL), or DNase positive control. Green fluorescence indicates TUNEL positivity and DNA damage. DNA nuclei were counterstained by DAPI. Scale bar: 100 μm.

2

Supplementary Figure S2. qRT-PCR quantification of relative fold change in gene expression between hiPSC-VSMCs cultured with and without addition of 7.5 μg/mL EGCG for nuclear factor kappa B (NF-κB), DNA methyltransferase 1 (DNMT-1), histone deacetylase 1 (HDAC-1), interleukin-6 (IL-6), and vascular endothelial growth factor A (VEGF-A) (Two-tailed, unpaired Student’s t-test was performed; Mean values and S.E.M. indicated by the error bars are shown; n = 3 for all markers; *: p<0.05; **: p<0.01; ns: not significant).

3

Supplementary Figure S3. (A) DNA gel showing result of restriction digestion by MluI and SalI following cloning of 2.193 kb elastin cDNA into a DOX-inducible vector targeting the AAVS1 locus. (B) Schematic overview of primers used to assess insertion of DOX-inducible elastin construct into AAVS1 locus of electroporated hiPSCs by TALEN genetic engineering. Yellow and blue primer pair PCR amplification yields the result of left and right TALEN arms at the native AAVS1 locus. In the case of elastin cDNA insertion, this primer pair will be too far separated, and no band will be observed. Yellow and red or blue and red primer pair PCR amplification yields the result of left and right TALEN arms including the transgenic, inserted sequence. For these primer pairings, a band will appear only in the case of elastin cDNA introduction. (C) DNA gel showing the result of the various primer pairings described in panel B, conducted on both parental (non-electroporated) and DOX-ELN electroporated, genetically engineered hiPSCs. (D) Karyotyping analysis for chromosomal integrity and banding pattern for DOX-ELN hiPSCs. (E) Immunofluorescence imaging for pluripotency markers OCT4, Nanog, and TRA-1-60 in cultured DOX-ELN hiPSCs. Scale bar: 200 μm.

4

Supplementary Figure S4. Representative immunofluorescence imaging of decellularized matrices from DOX-ELN hiPSC-VSMCs treated with either DMSO vehicle alone (top row) or DMSO and 2 μg/mL doxycycline (bottom row). Columns represent increasing serum enrichment of the culture medium from 3% FBS to 20% FBS. All serum conditions were supplemented with 1 ng/mL TGF-β. Scale bar: 200 μm.

Highlights:

• Epigallocatechin gallate addition leads to extracellular elastin formation in vitro

• Epigallocatechin gallate-treated tissue rings appear to display enhanced elasticity

• Doxycycline-inducible construct for ectopic expression of elastin mRNA and protein

• Use of doxycycline and epigallocatechin gallate produces elastin under high serum

• Applied to stem cell-derived culture for future elastin enriched vessel engineering