Vascularization of PEGylated Fibrin Hydrogels Increases Proliferation of Human iPSC-Cardiomyocytes

Abstract

Studies have long sought to develop engineered heart tissue for the surgical correction of structural heart defects, as well as other applications and vascularization of this tissue has presented a challenge. Recent studies suggest that vascular cells and a vascular network may have regenerative effects on implanted cardiomyocytes and nearby heart tissue separate from perfusion of oxygen and nutrients. The goal of this study was to test whether vascular cells or a formed vascular network in a fibrin-based hydrogel would alter the proliferation of human iPSC-derived cardiomyocytes. First, vascular network formation in a slowly degrading PEGylated fibrin hydrogel was optimized by altering the cell ratio of human umbilical vein endothelial cells (HUVEC) to human dermal fibroblasts (hDF), the inclusion of growth factors, and the total cell concentration. An endothelial to fibroblast ratio of 5:1 and a total cell concentration of 1.1x10⁶ cells/mL without additional growth factors generated robust vascular networks while minimizing the number of cells required. Using this optimized system, human iPSC-derived cardiomyocytes were cultured on hydrogels without vascular cells, hydrogels with unorganized encapsulated vascular cells, or hydrogels with encapsulated vascular cells organized into networks for 7 days. Cardiomyocyte proliferation and gene expression were assayed following 7 days of culture on the hydrogels. The presence of vascular cells in the hydrogel, whether unorganized or in vascular networks, significantly increased cardiomyocyte proliferation compared to an acellular hydrogel. Hydrogels with unorganized vascular cells resulted in lower cardiomyocyte maturity evidenced by decreased expression of cardiac troponin t (TNNT2), myosin light chain 7 (MYL7), and phospholamban (PLN) compared to hydrogels without vascular cells and hydrogels with vascular networks. Altogether, this study details a robust method of forming rudimentary vascular networks in a fibrin-based hydrogel and shows that a hydrogel containing endothelial cells and fibroblasts can induce proliferation in adjacent cardiomyocytes, and these cells do not hinder cardiomyocyte gene expression when organized into a vascular network.

Introduction

Congenital heart defects (CHD) are the most common congenital disorders, affecting 1% of all live births with 25% requiring surgery in the first year of life.^{1, 2} Currently, patches composed of synthetic polymers or fixed pericardium are used to close septal defects and reconstruct stenotic or overriding outflow tracts.³ These patches become encased in scar tissue, which inhibits the electrical and mechanical coordination of the heart muscle, leading to increased risk of arrhythmias, heart failure, and sudden cardiac death.^4,5 An engineered patch that initiates a regenerative response in the surrounding tissue and remodels to become functional heart tissue would provide better mechanical and electrical integration of the defect area, improving long-term function over the current standard of care.

The highly metabolic myocardium requires continuous vascularization to support cell survival and growth. Studies of regeneration in injured zebrafish hearts have shown that coronary re-vascularization is vital to cardiomyocyte (CM) re-population of the injury site.⁶ In apical resection of neonatal rodent hearts, angiogenesis preceded CM migration into the defect and CM infiltration spatially followed the growing vasculature.⁷ Additionally, inclusion of vascular cells in cardiac implants accelerated the ingrowth of host vasculature and formed perfusable vessels containing native red blood cells in vivo.^8–13 Therefore, we hypothesized that a scaffold containing a vascular network can allow for rapid vascularization that will drive CM regeneration in a cardiac patch.

Fibrin hydrogels have been extensively studied for their ability to induce wound healing and angiogenesis but are typically limited by rapid degradation.^14–18 We previously investigated a novel method of creating a fibrin hydrogel from fibrinogen covalently decorated with polyethylene glycol (PEG) and formed using thrombin in supraphysiological saline concentration of 250mM NaCl.¹⁹ This PEGylated high salt hydrogel had a reduced degradation rate compared to a standard fibrin hydrogel and was able to support the assembly of endothelial cells (ECs) and human dermal fibroblasts (hDF) into vascular networks. However, the reduced degradation rate also altered the dynamics of vascular network formation when using previously reported cell concentrations and EC:stromal cell ratios.^{20, 21} Consequently, in this study we sought to optimize the formation of vascular networks in the previously described PEGylated fibrin hydrogels. In these gels, the fibrinogen is partially crosslinked and decorated with a bi-functional amine reactive PEG before thrombin-induced gelation, though the PEG-fibrinogen remains soluble and does not form a hydrogel without thrombin. Vascular endothelial growth factor (VEGF) and platelet-derived growth factor (PDGF), both known to stimulate endothelial proliferation and angiogenesis,^{22, 23} were tested along with varying cell ratios and concentrations.

Additionally, the presence of endothelial cells and fibroblasts has been shown to impact CMs in direct co-culture. One study found that the presence of ECs increased the proliferation of immature CMs as well as increased the maturation of non-dividing CMs.^{8, 24} Multiple groups have assessed the drug response of multi-cell culture systems demonstrating that vascularized cardiac tissues have significantly different toxicity responses than monoloayers and CM monocultures as a result of increased expression of mature contractile proteins and improved calcium handling.^25–29 These studies have demonstrated the effect of direct co-culture of CMs with endothelial and stromal cell types but do not show whether signaling from a vascularized scaffold can have similar impacts on CMs in the surrounding tissue. We recognize that a paracrine mechanism is necessary to induce a regenerative response in the surrounding heart tissue. Further, we showed in a previous study that pre-vascularizing a fibrin-based cardiac patch increased muscularization of the patch area compared to an acellular patch in a right ventricle replacement model despite no survival of the implanted vascular cells.³⁰

In this study, we used a structured approach to optimize vascular network formation by varying the inclusion of angiogenic growth factors (VEGF and PDGF), cell ratio of ECs to hDFs, and total cell concentration in a slowly degrading PEGylated fibrin hydrogel. We then cultured hIPSC-derived CMs on acellular gels, gels with disorganized vascular cells, and gels with a vascular network and measured CM proliferation and gene expression.

Materials and Methods

PEGylated Fibrin Hydrogel Formulation

PEGylated fibrin hydrogels were formulated as previously described ¹⁹ with slight modifications. Briefly, high salt phosphate buffered saline (PBS) was created by dissolving sodium chloride (NaCl) in PBS (Corning) to bring the NaCl concentration of the PBS up to 565 mM. Sterile fibrinogen (Millipore Sigma) was dissolved in the high salt PBS at 80 mg/mL for 30 minutes at 37°C. NHS bi-functional PEG (3.4 kDa, SUNBRIGHT, NOF America Corporation) was dissolved in the high salt PBS at 8 mg/mL, syringe filtered, and mixed with the fibrinogen at a 1:1 volume ratio, creating a 10:1 weight ratio of fibrinogen to PEG. The fibrinogen and PEG solution was incubated at 37°C for 1 hour. For vascularized hydrogels, either human umbilical vein endothelial cells (HUVEC) (Lonza) or GFP expressing HUVEC (Angio Proteomie) and human dermal fibroblasts (hDF) (Lonza) were encapsulated in the hydrogels. Following cell culture, the cells were dissociated using TrypLE Express^® (Gibco), counted, and resuspended at 8x the desired final cell concentration in media. HUVEC and hDF cell solutions were combined at a 1:1 volume ratio to generate a vascular cell solution. The PEGylated fibrinogen was combined with the vascular cell solution at 1:1 volume ratio, mixed by pipetting, and added to the wells of a cell culture plate. For acellular control hydrogels, PEGylated fibrinogen was combined with EGM- 2 cell media (Lonza) instead of a cell solution. Thrombin from human plasma (Millipore Sigma) was resuspended in cold calcium chloride solution (11.1 mM CaCl2, 145 mM NaCl, pH 7.4 in DI water) at 20 U/mL. Thrombin solution was added to the PEGylated fibrinogen and cell solution 1:1 by volume and thoroughly mixed by pipetting before incubating for 10 minutes at 37°C to allow gelation. Following gelation, hydrogels were cultured in EGM-2. The final concentrations in the hydrogels were 11 mg/mL PEGylated fibrinogen and 250 mM NaCl. Vascular network hydrogels were cultured in EGM-2 for 7 days to promote the formation of endothelial cell networks. All vascularization studies were performed using the same lot of cells with four repeats per experiment to account for variability between cell plating and time. For experiments with the addition of growth factors, VEGF-165 (Shenandoah) was added to the media at 50 ng/mL for the first 2 days followed by PDGF-bb (Shenandoah) at 25 ng/mL for days 4-6. Network formation was quantified using the open-source software Angiotool, provided by the National Cancer Institute.³¹ Angiotool provides nine various measurements of an identified vascular network as well as analysis of these per unit area. For this study, all images analyzed were of the same size and as a measure of total amount of vascularity, average vessel length, total vessel length, and number of junctions are reported here.

Cardiomyocyte Differentiation

CMs were differentiated from human induced pluripotent stem cells (hiPSC) using small molecule modulation of the Wnt pathway. A human primary dermal fibroblast derived iPSC line from a 62-year-old female, ic4-4, was purchased from the Gates Institute for Regenerative Medicine with verification as a normal karyotype and contamination-free.³² Cells were single cell passaged using Accutase^® (Millipore Sigma) into a 12- well plate coated with Matrigel^® (Corning) and cultured in mTeSR Plus medium (Stem Cell Technologies) for 3 days. Differentiation was initiated by switching to RPMI1640 media (Gibco) supplemented with B27^® minus insulin (Gibco) and 4uM CHIR99021 (SelleckChem) for 48hrs. CHIR99021 was then removed and RPMI/B27^® minus insulin was added for 24hrs. Media was then switched to RPMI/B27^® minus insulin with 5uM IWR-1 (SelleckChem) added for 48hrs followed by just RPMI/B27^® minus insulin for 2 days. Finally, media was switched to RPMI/B27^® (with insulin) on Day 7 and media was refreshed every 2-3 days thereafter. Metabolic purification was performed from Days 10-14 by switching to RPMI1640 without glucose (Gibco) supplemented with B27^®. Purified cardiomyocytes were maintained in RPMI (with glucose)/B27^®.

Cardiomyocyte Culture on Hydrogels

Following 25-30 days from the start differentiation, hiPSC-derived CMs were lifted using Accutase^® and transferred to the top of 3 different hydrogel conditions. Vascular network hydrogels were formed 7 days prior to the addition of CMs with encapsulated HUVEC and hDF at a 5:1 ratio and a total cell concentration of 1.1 × 10⁶ cells/mL. Vascular cell hydrogels were made 1 day prior to introduction of CMs with the same vascular cell ratio and concentration as the network hydrogels. Acellular (CM only) hydrogels were formed without any encapsulated vascular cells 1 day prior to introduction of CMs. CMs were cultured on top of these 3 hydrogel conditions at approximately 30,000 cells/cm² for 7 days in RPMI/B27. Proliferation and gene expression analysis were performed on day 7 (Figure 1).

Figure 1. Schematic of experimental procedure.

Vascular network hydrogels were cultured in EGM-2 for 7 days prior to adding CMs. Vascular cell hydrogels and acellular hydrogels were formed 1 day prior to adding CMs. CM analysis was performed after 7 days of culture on the hydrogels.

Immunofluorescence Analysis

Hydrogels were fixed in 10% Formalin for 10 min. Cells were permeabilized with 0.5% Triton-X (Thermo Fisher) and non-specific antibody binding was blocked with 3% bovine serum albumin (BSA) in PBS. For CM proliferation experiments, Click-iT EdU proliferation kit (Invitrogen) was used according to manufacturer’s instructions. EdU was added to the media for the final 24h of culture on the hydrogels. A cardiac troponin T (1:400 dilution, Life Technologies, product number MA5-12960, clone 13-11) primary antibody was used to label CMs. The primary antibody was detected using an Alexa Fluor 546- conjugated goat anti-mouse secondary antibody (1:1000 dilution, Invitrogen, product number A11030). Cells were counterstained with 4,6-diamidino-2-phenylindole (DAPI, Biotium) and imaged using a Zeiss ObserverZ.1 fluorescent microscope.

Images were quantified by hand using ImageJ (National Institutes of Health) by counting nuclei and EdU-labeled nuclei inside cells that were stained positive for cTnT, reported in figures as %EdU+ CMs. Circularity was determined through manual outlining of cells with best judgment used in connected cells using ImageJ (National Institutes of Health) and reporting of circularity by the software, calculated as the following per ImageJ online documentation:

$$\mathit{\text{Circularity}}=\mathit{4}\mathit{\text{pi}}(\mathit{\text{area}}/\mathit{\text{perimeter}}^\mathit{2})$$

Gene Expression

After 7 days of culture on top of hydrogels, CMs were lifted using Accutase^® and RNA was extracted using TriZol (Life Technologies) and chloroform followed by RNeasy Minikit (Qiagen) according to the manufacturer’s instructions. cDNA was synthesized using a cDNA Reverse Transcription Kit (Applied Biosystems). Expression of the cardiac genes TNNT2 (HS00943911_m1), MYH6 (HS01101425_m1), MYH7 (HS01110632_m1), MYL2 (HS00166405_m1), MYL7 (HS01085598_g1), CACNA1c (HS00167681_m1), and PLN (HS01848144_s1) were measured by quantitative polymerase chain reaction and normalized to GAPDH (HS02786624_g1). Fold change was normalized to the acellular (CM only) hydrogel group.

Statistics

Statistical analysis was performed using GraphPad Prism software. One way ANOVA followed by Fisher’s LSD test was used for statistical comparisons with P<0.05 considered significant. Results are reported as mean ± SD unless otherwise indicated.

Results

Vascular Network Formation

Vascular network hydrogels were made by encapsulating GFP-HUVEC and hDF within a PEGylated fibrin hydrogel and culturing them for 7 days. Endothelial vessel lengths and branching were quantified to find the optimal cell density, ratio of endothelial cells to fibroblasts, and addition of the growth factors VEGF and PDGF. A range of cell parameters were tested, centered around the 800,000 cells/mL and a 4:1 ratio of HUVEC:hDF used in a previous study by our team.¹⁹ Increasing the HUVEC:hDF ratio increased vessel lengths and branching, plateauing at a 5:1 ratio (Figure 2). At this point, the 3:1 ratio was eliminated from future experiments. The addition of VEGF or sequential addition of VEGF and PDGF showed some increases in average vessel length at 4:1 and 5:1 ratios, but decreased the number of junctions in the networks at the 5:1 ratio (Figure 2 B, D). Therefore, the addition of growth factors was not considered to improve network formation. Finally, total cell concentration was varied from 500,000 cells/mL to 1.4 × 10⁶ cells/mL using a HUVEC:hDF ratio of 5:1. Increasing the total cell concentration resulted in significant increases in average vessel length and trended towards higher total vessel length (Figure 3). No significant differences were observed between 1.1 × 10⁶ cells/mL and 1.4 × 10⁶ cells/mL. Based on these results, experiments with hiPSC-CMs were conducted using hydrogels made with a 5:1 HUVEC:hDF ratio, a cell concentration of 1.1 × 10⁶ cells/mL, and no added growth factors.

Figure 2.

Changing the ratio of HUVEC to hDF alters endothelial network formation. (A) Fluorescent imaging of GFP-HUVEC (green) counterstained with DAPI (blue) after culture for 7 days within a fibrin hydrogel. Endothelial networks were quantified by average vessel length (B), total vessel length (C), and total number of junctions (D). #p<0.05, ##p<0.01 compared to 4:1.*p<0.05 compared to no growth factor at the same cell ratio. Scale bars = 50μm. n=4.

Figure 3.

Increasing the total number of encapsulated cells improves endothelial network formation. (A) Fluorescent imaging of GFP-HUVEC (green) counterstained with DAPI (blue) at varying initial cell concentrations. Endothelial networks were quantified by average vessel length (B), total vessel length (C), and total number of junctions (D). *p<0.05 compared to 800,000 cells/mL. Scale bars = 50μm. n=4.

Cardiomyocyte Proliferation on Vascular Hydrogels

Human iPSC-CMs were differentiated using previously published methods from our laboratory. The average differentiation efficiency from ongoing laboratory data is 98.93% post purification, based on cardiac troponin-t positive cells. Human iPSC-CMs were cultured for 7 days on PEGylated fibrin hydrogels containing no cells (CM Only), encapsulated HUVEC and hDF (+Vascular Cells), or encapsulated HUVEC and hDF given 7 days to assemble into networks (+Vascular Networks) (Figure 1). The impact of the presence and organization of endothelial cells and fibroblasts on the proliferation and gene expression of hiPSC-CMs was then assessed. The presence of endothelial cells and fibroblasts significantly increased the number of proliferating hiPSC-CMs from 6.4±0.9% in the acellular (CM only) group up to 20.9±8.2% for the vascular cell group and 14.9±2.7% for the vascular network group (Figure 4).

Figure 4.

The presence of HUVEC and hDF increases proliferation of hiPSC-CMs. (A, B, C, D) CMs were cultured on hydrogels containing no vascular cells (CMs only), (E, F, G, H) unorganized HUVEC and hDF (+Vascular cells), or (I, J, K, L) HUVEC and hDF assembled into networks (+Vascular networks). After 7 days, cells were stained for cardiac troponin T (red), EdU (green), and DAPI (blue). (M) CMs positive for cTnT on gels with cells had significantly higher proliferation, (N) lower circularity, and (O) lower cell numbers. *p<0.05. Scale bar = 50μm. n=4.

Cardiomyocyte Gene Expression on Vascular Hydrogels

Expression of an array of cardiac maturation markers related to contractile proteins and calcium handling were measured following 7 days of culture on top of the three different hydrogel groups. There were no significant differences in any of the measured genes between the acellular (CM only) and vascular network hydrogel groups. However, the vascular cell hydrogel group showed a trend towards lower expression of all genes, reaching significantly lower expression of the structural protein myosin light chain 7 (MYL7) and the calcium handling protein phospholamban (PLN) (Figure 5). Additionally, cardiac troponin T (TNNT2) expression was significantly lower in the vascular cell group compared to the vascular network group.

Figure 5.

Unorganized vascular cells lower CMs expression of contractile genes. CMs gene expression of cardiac troponin T (TNNT2), myosin heavy chain 6 (MYH6) and 7 (MYH7), myosin light chain 2 (MYL2) and 7 (MYL7), L-type calcium channel α1c subunit (CACNA1c), and phospholamban (PLN) after culture for 7 days on various hydrogel conditions. Fold change was normalized to CMs only group. Data shown as mean ± SEM. *p<0.05, **p<0.01. n=3.

Discussion

Vascularization is critical to the success of cardiac implants and groups have shown that pre-vascularization of a scaffold increased the rate of angiogenesis and anastomosis in vivo.^{8–11, 33} Fibrin hydrogels have been used to generate vascular networks;^{20, 21, 33} however, the optimal cell concentrations and ratios vary between hydrogel formulations and specific cell types. Here, we manipulated the total cell concentration and the ratio of HUVEC to hDF to improve the development of vascular networks within a PEGylated fibrin hydrogel. We found that a HUVEC:hDF ratio of 5:1 and at least 1.1 × 10⁶ cells/mL total generated vascular networks with higher average vessel length and similar branching and total vessel length compared to lower concentrations. We saw no significant improvement at higher cell concentrations. Our values reported here agree with those reported by Chen et al.,³³ who demonstrated accelerated anastomosis in vivo with a pre-vascularized fibrin construct.

With the goal of using this scaffold for cardiac regeneration, we analyzed whether a pre- vascularized hydrogel could induce CM proliferation and whether this would affect CM sarcomeric gene expression. Previous studies showed that culturing CMs with ECs and fibroblasts results in increased electrical and mechanical cell-cell junctions ^{24, 34} as well as more mature contractile properties.^{25–29, 35} One study found that CM proliferation increased when cultured with ECs.⁸ However, the maturation effect of ECs on CMs has been reported to be differentiation stage specific with minimal effects seen after CM differentiation day 13-18³⁶ and all previous studies involved direct mixing of vascular cells with CMs.

Our recent study, in which pre-vascularized patches were used to replace part of the right ventricle free wall in a rat heart defect model, demonstrated increased muscularization in the patches containing vascular cells compared to acellular patches.³⁰ Few implanted cells could be detected at the 2-month timepoint, indicating paracrine signaling from the implanted cells as the probable cause of improved muscularization. Secreted factors from endothelial cells are known to play a role in cardiomyocyte proliferation and remodeling in both development and post-injury.^37,38 In this study, we wanted to test the hypothesis that these effects could extend to native CMs outside the implanted scaffold, that a vascular network would produce different effects than unorganized vascular cells, and that these effects would be measurable in more mature cells of 25- 30 days post differentiation.

We found that regardless of their organization into networks, the presence of endothelial cells and fibroblasts did increase the proliferation of CMs cultured on our hydrogel. Unlike in direct co-culture, the expression of genes related to cardiomyocyte maturity was not increased. Interestingly, there were significant decreases in gene expression of the calcium handling protein phospholamban (PLN) as well as the contractile proteins myosin light chain 7 (MYL7) and cardiac troponin T (TNNT2) in the vascular cell hydrogel group, which contained unorganized HUVEC and fibroblasts. Proliferating CMs disassemble their contractile structures to complete cell division,^{39, 40} which could explain the lower expression of contractile proteins and PLN in the highly proliferative vascular cell hydrogel group. The vascular cell hydrogel group had the highest amount of proliferation and trended towards lower expression of all proteins measured, which would indicate a less mature and more proliferative CM cell state in response to unorganized vascular cells. Additionally, the organization of the encapsulated vascular cells into networks did not result in decreased CM gene expression, which could be related to a change in cell signaling factors once the vascular cells have formed networks; however more investigation is necessary.

These results show that cardiomyocytes increase proliferation by approximately the same amount in culture on both cellularized and vascularized hydrogels. However, some contractile proteins genes were more highly expressed in cardiomyocytes cultured on the vascularized hydrogels. Because proliferating CMs must disassemble their sarcomeric cytoskeleton in a process referred to as dedifferentiation^{41, 42}, the proliferating cells likely express fewer sarcomeric proteins. We hypothesize that the increase in some sarcomeric gene expression in cardiomyocytes on vascularized gels reflects a separate non-proliferating and maturing subpopulation of cells that are increasing expression of contractile proteins, showing up in the bulk PCR results. We considered the hypothesis that CMs on cellularized gels undergo full cell proliferation, while CMs on vascularized cells undergo binucleation that is halted in mitosis without full cell division. However, because binucleated CMs were not observed in this study and other studies have shown that CMs disassemble sarcomeres in apparent dedifferentiation in both cell division and binucleation⁴³, we feel this explanation is unlikely.

While many studies have investigated signaling from cardiomyocytes that induces or directs angiogenesis, few have investigated signaling from vascular cells that can influence cardiac maturation and proliferation. However, studies have found that nitric oxide (NO) relaxes blood vessel smooth muscle cells and affects CM contractility⁴⁴, neuregulin-1 (NRG-1) activates receptors in CMs, promoting cardioprotection and regeneration⁴⁵, apelin regulates CM hypertrophy and responses to heart injuries^{46, 47} and endothelin-1 (ET-1) is produced by various heart cells and influences CM contractility and remodeling⁴⁸. Other potential factors like prostacyclin, periostin, and angiotensin II may also play roles in EC-CM communication as reviewed by Talman and Kivela.³⁷

We note that the time of culture of HUVEC and hDF in the hydrogels was different for the unorganized and network-forming hydrogels. The unorganized hydrogels were cultured for only 1 day while the network hydrogels had cells cultured for 7 days, allowing greater accumulation of excreted factors and additional remodeling and extracellular matrix secretion by the loaded cells. Although we hypothesize that the cell states are different when they are individually in a hydrogel compared to when they are in vascular networks, and this changes secreted factors with differing effects on CMs, we are aware that the effects observed may be due to accumulation of similar secreted factors.

The EGM-2 culture media we used for vascular gel culture contains 0.5ng/ml VEGF, which has been shown to enhance endothelial cell proliferation and vessel formation⁴⁹. To investigate effects of superphysiologic VEGF, we measured vascularization after culture with 50ng/ml and found no increase in vessel length at high ratios of endothelial cells to fibroblasts. This is consistent with previous studies that found that VEGF greatly increases endothelial cell proliferation but only increases vascular area in certain specific conditions.⁵⁰ Studies have shown increases in angiogenesis in fibrin gels after a sequential addition of VEGF and PDGF.⁵¹ In this system we found no benefit of addition of PDGF on total vessel length, though this can be confounded by PDGF secreted by the implanted hDFs⁵².

Cell density has varied by orders of magnitude in previous studies. While some studies have found significant sprouting differences at 50,000 cells/ml,⁵⁰ others have cultured 100,000 cells/ml⁵³ or up to 6 million cells/ml.⁵⁴ We saw no improvement beyond 1.1 million cells/ml and did not explore higher concentrations.

The measurement of proliferation was based on culture with EdU for 24 hours after 6 days of culture. However, analysis of cell number found that the acellular gels have significantly more cardiomyocytes on top of the gel, which may suggest better attachment or early proliferation in cells on acellular gels that shifted to higher proliferation on cellularized gels after a week of culture. As a measure of cell structure and organization, we measured circularity in all cells and found that cardiomyocytes were more elongated, with significantly lower circularity, on gels with vascular cells compared to acellular controls.

Previous studies have found that the maturation of cardiomyocytes can be enhanced through coculture in an organoid with cardiac-derived endothelial cells and fibroblasts.^53,55 One limitation of this study is the use of HUVEC as an endothelial cell source and hDF as a source of fibroblasts. Though this may limit the applicability of this system as a model for cardiac maturation during development, these cell sources are consistent with previous in vivo studies in our lab⁵⁶ and are easily obtainable.

In this study, we intentionally imaged only the top layer of the gel was imaged, using a Zeiss ObserverZ.1 equipped with a motorized Z-drive and a Zeiss ApoTome.2 that filters out out-of-plane fluorescence. However, we acknowledge the limitation that this method cannot eliminate the possibility of detection of endothelial or fibroblast nuclei that are very close to the surface layer.

iPSC-derived cardiomyocytes at this early stage without interventions for maturation generally have proliferation between 5 and 10% as shown here in controls⁵⁷ but can increase proliferation to levels even higher than 20% when measured with 24 hour EdU incorporation when exposed to proliferative factors^{58, 59}, including NRG1, released by endothelial cells⁶⁰.

One limitation of this study is the lack of high-resolution images of sarcomere structure, which were not possible to obtain due to the difficulties of imaging with vascularized gels. We acknowledge that high resolution images of sarcomeres could indicate levels of structural maturity, and this will be a focus of new systems in future studies.

Another limitation of this study is that the day 25-30 post differentiation iPSC-CMs are less mature and organized than the CMs in a developed heart and have higher baseline proliferation. While this study provides insight into how a pre-vascularized fibrin-based gel can contribute to regeneration of heart tissue, future work will explore combining pre-vascularized hydrogels with additional signaling factors to induce a more proliferative CM cell state, leading to greater muscularization of defects.

Conclusion

In summary, we systematically optimized the formation of vascular networks within slowly degrading PEGylated fibrin hydrogels and found that a total cell concentration of 1.1 × 10⁶ cells/mL and a HUVEC:hDF ratio of 5:1 consistently produced robust vascular networks. We found that hydrogels containing both HUVEC and hDF resulted in increased proliferation of hIPSC-derived CMs cultured on the gels. Additionally, we found that gels with unorganized vascular cells resulted in lower expression of some sarcomeric CM genes compared gels with a preformed vascular network and acellular gels. Overall, these results suggest that the presence on a vascularized gel may enhance nearby cardiomyocyte proliferation and sarcomeric gene expression.

Data Availability Statement

The data that support the findings of this study are openly available in Mendalay at http://doi.org/10.17632/hpzbk2rcxs.1.