Gelatin Promotes Cell Retention Within Decellularized Heart Extracellular Matrix Vasculature and Parenchyma

Abstract

Introduction

Recellularization of organ decellularized extracellular matrix (dECM) offers a potential solution for organ shortage in allograft transplantation. Cell retention rates have ranged from 10 to 54% in varying approaches for reseeding cells in whole organ dECM scaffolds. We aimed to improve recellularization by using soluble gelatin as a cell carrier to deliver endothelial cells to the coronary vasculature and cardiomyocytes to the parenchyma in a whole decellularized rat heart.

Methods

Rat aortic endothelial cells (RAECs) were perfused over decellularized porcine aorta in low (1%) and high (5%) concentrations of gelatin to assess attachment to a vascular dECM model. After establishing cell viability and proliferation in 1% gelatin, we used 1% gelatin as a carrier to deliver RAECs and neonatal rat cardiomyocytes (NRCMs) to decellularized adult rat hearts. Immediate cell retention in the matrix was quantified, and recellularized hearts were evaluated for visible contractions up to 35 days after recellularization.

Results

We demonstrated that gelatin increased RAEC attachment to decellularized porcine aorta; blocking integrin receptors reversed this effect. In the whole rat heart gelatin (1%) increased retention of both RAECs and NRCMs respectively, compared with the control group (no gelatin). Gelatin was associated with visible contractions of NRCMs within hearts (87% with gelatin vs. 13% control).

Conclusions

Gelatin was an effective cell carrier for increasing cell retention and contraction in dECM. The gelatin-cell-ECM interactions likely mediated by integrin.

Electronic supplementary material

The online version of this article (10.1007/s12195-020-00634-z) contains supplementary material, which is available to authorized users.

Introduction

The number of new patients diagnosed with heart failure is increasing, especially as the population ages.63,67 The definitive treatment for end-stage heart failure is heart transplantation, but the shortage of acceptable donor organs for transplant is a severe limiting factor.24 Thus, there is a significant need for new treatment options for this devastating disease.17 One approach to address the donor shortage is to create a bioartificial heart by reseeding a bioscaffold, such as a decellularized cadaveric heart, with human cells.2,31,45 We proposed that approach a decade ago when we reported that we had decellularized and reseeded a cadaveric rat heart with neonatal rat cardiomyocytes (NRCMs).45

However, adequate reseeding of a whole organ decellularized extracellular matrix (dECM) has been challenging, whether the bioscaffold is rat,45,65 pig,26,65 or human16 ECM. In some cases, as few as 10 to 15% of the cells have been preserved in the dECM after 7 days of perfusion with media and growth factors,34 whereas other studies have shown cell retention rates of 10 to 54% in whole decellularized hearts.34,45 These findings point to a major limitation in reseeding dECM: the large number of cells that do not attach to the scaffold but rather flow out of the heart.

To address the issue of cell retention in the heart, we looked at popular materials currently used in tissue engineering and found soluble gelatin to be a promising candidate as a cell carrier based on its temperature-dependent hydrogel properties. Gelatin, a naturally derived protein generated from the hydrolysis of collagen, is both biodegradable and biocompatible, with three beneficial properties: solubility, viscosity, and conductivity.65 Gelatin can be manipulated by changing its concentration and temperature.9 For example, increasing the temperature of a solution increases gelatin solubility, and decreases viscosity, and conductivity, while increasing gelatin concentration decreases solubility and increases viscosity and conductivity.15 Gelatin, a water-soluble protein, can be injected as a solution to control its area of application, and evidence suggest that gelatin allows cell-released cytokines to remain active.39,68 These features of gelatin permit its use in a wide range of biomedical applications, including as a drug delivery vehicle or carrier. We reasoned that these same properties could facilitate a more hospitable environment for cells to attach and proliferate when utilizing gelatin as a cell carrier to dECM.

We designed the current study to address the tissue engineering challenge of cell delivery and retention during recellularization of a cardiac dECM bioscaffold. We evaluated whether soluble gelatin, a naturally-derived protein, could be used as a cell carrier to increase endothelial cell and cardiomyocyte retention in a whole decellularized rat heart. We hypothesized that soluble gelatin would be a better cell carrier than medium alone for improving the retention of endothelial cells delivered to the dECM vascular tree and cardiomyocytes injected into dECM parenchyma.

Materials and methods

Animal Experiments

All experiments were performed in accordance with the US Animal Welfare Act and were approved by the Institutional Animal Care and Use Committee at the Texas Heart Institute.

Rat Heart Harvest and Decellularization

Fischer rats (Charles River, Wilmington, MA) were anesthetized with 5% isofluorane and then heparinized with 1000 UI/ml sodium heparin intraperitoneally. The thoracic cavity was opened through a median sternotomy to expose the heart and its branching vessels. The brachiocephalic artery was ligated and cut. The aorta, superior vena cava, inferior vena cava, left pulmonary artery, and pulmonary veins were cut to remove the heart. The aorta was cannulated with a 1-mm cannula, and the heart was flushed with phosphate-buffered saline (PBS). Decellularization was modified from the previously described protocol.45 Hypertonic (500 mM NaCl), hypotonic (20 mM NaCl), and 1% sodium dodecyl sulfate (SDS) solutions were perfused sequentially through an aortic cannula at a constant pressure of 70 mmHg for 10, 10, and 30 h, respectively. Lastly, PBS was perfused at a flow rate of 5 ml/min for 24 h. DNA, GAG, and SDS content was evaluated at the end of the protocol and the data is presented in Supplementary Fig. 1. The decellularized hearts were stored in PBS at 4 °C for recellularization at a later timepoint.

Porcine Aorta Decellularization

We obtained the ascending aorta from euthanized pigs. The aorta was cannulated at both ends and anchored at 45° with 10% of pre-stretch. Decellularization was achieved by sequentially perfusing hypertonic, hypotonic, and ionic detergent solutions through the aorta at 120 mmHg constant pressure; the perfusion was driven by a peristaltic pump controlled by a proportional-integral-derivative controller (Harvard Apparatus, USA). The following solutions were perfused sequentially over 72 h: hypertonic (500 mM) NaCl for 4 h, hypotonic (20 mM) NaCl for 2 h, 1% SDS for 60 h, and PBS for 6 h. Decellularization was confirmed as previously described.33

Rat Aortic Endothelial Cell Culture

Commercially available rat aortic endothelial cells (RAECs, Cell Application, Inc., USA) were cultured in rat endothelial cell growth medium (Cell Application, Inc., USA) supplemented with 1% penicillin/streptomycin (Sigma-Aldrich, USA) and 10% fetal bovine serum. The culture medium was changed 3 times per week. RAECs were kept below 8 passages before the experiments.35 For in vitro experiments with gelatin, RAECs were trypsinized and resuspended in 1 or 5% porcine skin gelatin (w/w, Sigma-Aldrich) that was dissolved in medium. To determine the optimized viscose of gelatin with limited compromise in cell viability, we evaluated the cell survival rate in two temperature settings of room temperature and 37 °C, at two conditions of stationary and laminar flow, i.e., conditions cells will face during recellularization. For the laminar flow condition, flow rate was kept constant in both temperature settings. The experimental design is presented in Supplementary Fig. 2A and 2B.

Cell Attachment Under Laminar Flow In Vitro

After decellularization, the porcine aorta was dissected longitudinally and flattened to evaluate cell attachment to its surface under laminar flow. A sticky-slide flow chamber was mounted on the decellularized aortic luminal surface to form 6 channels that were 0.4 mm high and 17 mm long. RAECs were labeled with 1,1′-dioctadecyl-6,6′-di(4-sulfophenyl)-3,3,3′,3′-tetramethylindocarbocyanine (SP-DilC_18,24 Thermo Fisher Scientific, USA) and suspended in standard medium without any gelatin (control) and in media with low (1%) and high (5%) gelatin concentrations. Cell suspensions were infused through the laminar flow chamber at 1 ml/min for 3 min using a syringe pump (Harvard Apparatus, USA) at room temperature. After the flow was stopped, the aorta was removed from the flow chamber, and the cells were fixed in 4% paraformaldehyde in situ for 20 min. When the experiment was repeated, a subset of cells was incubated with an antibody to block integrin α5β1 (anti-integrin α5β1 antibody, clone JBS5, Millipore, USA)46 for 1 h to determine if integrin contributed to cell attachment to the aorta dECM in 1% gelatin. We obtained images of the cells attached to the aortic luminal surface by using a fluorescent microscope (Nikon, Japan). We counted attached cells per high-power field by using ImageJ software (NIH Image, National Institutes of Health, USA). The experimental design is presented in Supplementary Fig. 2C.

Cell Viability and Proliferation In Vitro

Cell viability and proliferation were quantified 7 days after seeding RAECs on tissue culture plates and culturing the cells in medium with different concentrations of gelatin (0, 1, and 5%). To assess cell viability, we used cell-permeant glycyl-phenyl-alanyl-aminofluorocoumarin (GF-AFC) (Promega BioSciences, USA), according to the manufacturer instructions.25 Cell proliferation was measured using a standard 3-(4, 5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide (MTT) proliferation assay.10

Re-endothelialization of Decellularized Rat Heart

Before re-endothelialization, the aortic cannula was reestablished and hearts were placed in PBS with penicillin (100 U/mL), streptomycin (100 μg/mL), and amphotericin B (2.5 μg/mL, Sigma-Aldrich, USA) at 37 °C for at least 8 h, and then stabilized with medium at 37 °C for at least 8 h. On the basis of our results of cell viability, cell proliferation, and von Willebrand Factor (vWF) expression of RAECs in 1 and 5% gelatin, we used cells suspended in medium containing either no gelatin (control) or 1% gelatin for all ex vivo experiments. On the day of re-endothelialization, 40 million RAECs were trypsinized and suspended in 30 mL of medium containing 1% gelatin or medium only (control) at room temperature (20 °C). The control and 1% gelatin cell solutions were infused through the cannula inserted into the aorta by using a syringe pump at 1 ml/min for 30 min. During this infusion, the flow rate of medium through the left atrial cannula was 0.5 ml/min.49 After RAEC infusion was complete, the flow through both cannulas was stopped for 1 h to allow for cell attachment to the matrix. We collected the solution that perfused through the heart during cell infusion and counted the cells in the solution to quantify cell loss. Cell retention defined as cells infused minus cells lost was expressed as a percentage of total cells infused.

After the cell attachment period, the flow rate through the aortic cannula was set to 3 ml/min to supply oxygen and nutrients to the heart via coronary perfusion. The re-endothelialized heart was maintained at 37° for 7 days under constant flow in a Langendorff perfusion system (Harvard Apparatus), in which media was perfused through the aorta. The experimental design is presented in Supplementary Fig. 2D.

Viscosity of Whole Heart Perfusate

After the whole hearts were re-endothelialized, perfusate samples were collected every day for the first 5 days; after that, samples were collected whenever media in the whole heart was changed, which was approximately every other day. We measured the viscosity of the perfusate collected within the first 5 days by using an SV-10 viscometer (A&D Company, Japan), as previously described.3 Briefly, 10 ml of each perfusate sample was placed in the sample bath of the vibro-viscometer at 37 °C. The instrument measured the electric current needed to vibrate the sensor plates and output the associated fluid viscosity of the sample. Viscosity measurements in triplicate were done for hearts from which we had perfusate samples (n = 3 for control group, n = 3 for gelatin group).

Isolation and Preparation of Neonatal Rat Cardiomyocytes

Neonatal Sprague–Dawley rats (1–3 days old) (Texas Animal Specialties, USA) were euthanized with an intraperitoneal injection of sodium pentobarbital (100 mg/kg) and heparin (150 mg/kg). The hearts were excised and immediately placed into PBS on ice. Then, the hearts were minced and placed into Hank’s Balanced Salt Solution with trypsin (10,000 U/mL, Sigma-Aldrich, USA) at 4 °C overnight. Tissue pieces were digested in a solution of collagenase type II (200 U/mL) (Worthington Biochemical Co., USA) and 0.06% pancreatin (wt/wt, Sigma-Aldrich) in a Celstir (Wheaton, USA) at 37 °C until no tissue remained, approximately 35 min. The digested cells were collected and passed through a 100-µm cell strainer. The cells were pre-plated in T-175 flasks and incubated for 1 h to allow fibroblasts to attach to the flask. The supernatant was collected and plated for another hour to further reduce the number of fibroblasts. The freshly isolated NRCMs of > 90% purity were suspended in medium with or without 1% gelatin at a final concentration of 100 million cells/ml.

Delivery of Neonatal Rat Cardiomyocytes to Re-endothelialized Rat Heart Extracellular Matrix

We used a 27-guage needle to deliver the 1-ml NRCMs with and without 1% gelatin to the re-endothelialized hearts at day 7 after RAEC seeding. To determine if injection of NRCMs into the parenchyma would benefit from gelatin in the same manner as the RAECs did, RAECs were infused without gelatin in both groups injected with NR7CM. Five separate injections were spaced approximately 1-mm apart and parallel to the surface of the left ventricular free wall of the re-endothelialized hearts at room temperature. The injection speed was set at 20 µl/min using a syringe pump (88-1050, Harvard Apparatus). The hearts recellularized with RAECs (delivered without gelatin) plus injection of NRCMs (delivered with or without 1% gelatin) were cultured for 15 days paced with 2-ms long square waveform pulses of 10–15 V at 1 Hz beginning at 1 day after addition of NRCMs (day 8). One recellularized heart was maintained in vitro for 35 days to determine the longevity of the observed visual contractions. The experimental design is presented in Supplementary Fig. 2E.

Histology

At day 15 after re-endothelialization, recellularized rat hearts (with RAECs plus NRCMs) were sectioned into thirds from the base to the apex and fixed in 10% neutral buffered formalin at room temperature. The samples were dehydrated, embedded in paraffin, sectioned (6 μm), and placed on microscope slides for staining. After rehydration, slides from each of the 3 transverse sections of the heart (base, middle, and apex) were stained with Masson’s trichrome staining.52 For analysis of cellularity, representative images of each transverse section of the hearts were obtained with alight microscope (M205 FCA, Leica, Germany). Three images were obtained for each heart section (base, middle, and apex) from a total of four hearts. Images of the whole cross-section of the heart were evaluated with ImageJ by identifying red pixels as cells and blue pixels as matrix. The cellularity was determined as the percentage of cell area to matrix area, as previously described.6

For immunofluorescence studies, slides were stained using an anti-integrin β1 antibody (Abcam, USA) and either anti-rat endothelial cell antigen marker (RECA1, Abcam) or cardiac troponin T (cTnT, Abcam) primary antibodies. Fluorescent-labeled secondary antibodies, goat anti-mouse IgG, Alexa Fluor 488 conjugate antibody, and goat anti-rabbit IgG, Alexa Fluor 488 conjugate antibody (Thermo Fisher Scientific), were used. Slides were then stained with 4′,6-diamidino-2-phenylindole (DAPI). The matrix was detected by using reflected light signals,62 which enabled visualization of the autofluoresence of the ECM that originated primarily from collagen and elastin. Images were obtained with a Leica SP5 confocal microscope (Leica, Germany). The images were quantified with ImageJ (NIH, MA).

Statistical Analysis

All continuous variables (viability, fluorescence absorbance, cell number per square millimeter, and percentage of cells retained) were expressed as mean ± SEM. A Student t test was used for two-group comparisons. One-way ANOVA was used for comparing multiple concentrations of gelatin. A p value less than 0.05 was considered significant. All data for statistical analysis were processed with Graph Pad Prism 5.0 (Graph Pad Software Inc., USA). Additional descriptions of the methods are available in the Supplementary Methods.

Results

Gelatin Increased the Attachment of Rat Aortic Endothelial Cells to Decellularized Aorta Surface Under Laminar Flow

We demonstrated that the presence of gelatin in the medium delayed cell attachment in regular cell plates in a concentration-dependent manner (Supplementary Fig. 3a) under an stationary condition. To test whether gelatin improve the attachment of RAECs to a biologically relevant substrate, we quantified cell attachment to the surface of a decellularized porcine aorta in a low (1%) and high concentration (5%) of gelatin. A fluid suspension of RAECs was perfused over decellularized aorta at a rate of 1 ml/min for 3 min, mimicking the minimum flow rate in the vasculature of a child43 (Fig. 1a). Cell attachment to the aorta surface increased in the presence of gelatin in a concentration-dependent manner from 46.5 to 381.1 cells per square millimeter in 0 to 5% gelatin (Figs. 1b and 1c), p < 0.001, n = 11 images analyzed).

Figure 1.

Gelatin increases cell attachment to decellularized porcine aorta under laminar flow. (a) Schematic of the in vitro laminar flow-through chamber. Red arrows indicate the direction of fluid flow over the decellularized porcine aorta (gray rectangle). Non-attached yellow cells are round when perfused over the aorta. Some cells attach to the matrix and spread out over time, as indicated by the yellow oval cell. (b) Representative images of RAECs attached to the vessel surface after 3 min of perfusion of medium (Control), medium plus 5% Gelatin, medium with 1% Gelatin, and medium with 1% Gelatin and integrin blocking antibody (1% Gelatin + anti-integrin). Scale bar: 100 µm. (c) Quantification of RAEC attached to decellularized aorta. *p < 0.001 vs. Control; ** p < 0.05 vs. 1% Gelatin.

Integrins, as a class of heterodimeric transmembrane cell receptors and composed of one α subunit and one β subunit, involve in cell–cell and cell-ECM interactions.28 The motif of gelatin, Arg-Gly-Asp (RGD), ligates several integrins, but primarily α5β1 and αvβ3.50 To determine whether the cell attachment evolves an integrin-dependent pathway, we incubated cells in 1% gelatin with an α5β1 integrin-blocking antibody before cell perfusion. The number of RAECs that attached to the aortic scaffold decreased to 40% of pre-incubation values (p < 0.05) and was not significantly different from the number of RAECs attached when gelatin was not used (p = 0.54; Fig. 1c).

Rat Aortic Endothelial Cell Survival, Proliferation, and von Willebrand Factor Expression Were Retained in the Presence of 1% Gelatin

To determine a concentration of gelatin that promoted cell attachment to the dECM without altering cell proliferation and survival, RAECs were cultured in the presence of 1 and 5% gelatin in medium for 7 days, after which cell viability, cell proliferation, and von Willebrand expression (vWF) were evaluated. A gelatin concentration of 1% did not alter cell viability; however, 5% gelatin decreased RAEC survival by 29.9% (p < 0.05, Fig. 2a). Cell proliferation significantly decreased in 5% gelatin (p < 0.05, Fig. 2b) but was not significantly affected in 1% gelatin. Finally, vWF expression on RAECs was not altered upon visual inspection in 1% gelatin but was qualitatively decreased in 5% gelatin (Fig. 2c). Because these data showed lower cell viability, cell proliferation, and vWF expression in 5% gelatin, we chose to use 1% gelatin in all ex vivo studies to increase cell attachment to ECM.

Figure 2.

RAECs survive, proliferate, and retain von Willebrand Factor expression in 1% gelatin. (a) Viability of RAECs cultured in medium (Control), medium plus 1% Gelatin, and medium plus 5% Gelatin. Hydrogen Peroxide was used to induce cell death. (b) Proliferation assay of RAECs cultured in medium (Control), medium plus 1% Gelatin, and medium plus 5% Gelatin for 1, 3, and 7 days. *p < 0.05 vs. Control. (c) RAEC morphology in Control, 1% gelatin, and 5% gelatin. von Willebrand factor (vWF, green), Nuclei DAPI (blue). Scale bar: 100 µm.

Gelatin Increased Endothelial Cell Retention in Rat Heart dECM

As described in the methods section, the re-endothelization procedure occurs at room temperature and takes about 30 min. To better simulate this procedure in vitro, we quantified cell retention during re-endothelialization over 60 min at room temperature conditions. Adding 1% gelatin in the medium not only delayed cell attachment but also prevent cell death, which is shown by an increase of number of cells after 45 min (Supplementary Fig. 3b).

Quantification of cell attachment over 60 min for medium only (Control), medium plus 0.1, 0.5, 1, 3, and 5% gelatin at room temperature.

To test whether gelatin improved the cell retention of a re-endothelialized whole organ, cells were delivered to the cardiac vascular tree suspended in a media solution with and without 1% gelatin (control). RAECs (42 ± 3 million) were perfused into the coronary arteries of the rat whole heart dECM scaffold at a flow rate of 1 ml/min. Cell retention was significantly increased with 1% gelatin (86 ± 2%, n = 9) compared with no gelatin (58 ± 0%, n = 14; p < 0.0001; Fig. 3a).

Figure 3.

Endothelial cells and cardiomyocytes attach to decellularized cardiac matrix when delivered with gelatin. (a) RAEC retention 1 h after infusion. *p < 0.0001 vs. Control. (b) Viscosity of media perfusate 1, 2, 3, 4, and 5 days after re-endothelization without gelatin (Control) or with 1% Gelatin. (c) Presence of endothelial cells and (d) cardiomyocytes within tissue engineered hearts in hearts recellularized with gelatin.

We measured the viscosity of media perfusate samples collected from the hearts daily for the first 5 days after re-endothelialization. A slight increase in the viscosity of the perfusate from the 1% gelatin group was observed only in the first day after RAEC perfusion.

Seven days after endothelial cell infusion, we observed viable RAECs (CMFDA, green) with nuclear staining (Hoechst, blue) throughout the coronary vascular tree (Supplementary Fig. 4A). In hearts re-endothelialization with endothelial cells (no NRCMs), the RAECs were detected with DAPI staining throughout the vascular tree (Supplementary Fig. 4B). Integrin β1 expression was observed between RAECs and matrix in re-endothelialized hearts (Supplementary Fig. 5).

Gelatin Promoted Retention of Cardiomyocytes in the Parenchyma of the Re-endothelialized Heart

To determine whether gelatin would increase cardiomyocyte retention in the cardiac parenchyma, we injected NRCMs (72 ± 6 million) infused with and without 1% gelatin into the left ventricular free wall of decellularized rat hearts at 7 days after injection with RAECs in the absence of gelatin. After a period of 30 min without perfusion, more NRCMs were retained in the recellularized rat hearts when gelatin was present than when NRCMs were injected in medium alone (Table 1). On the other days, the viscosity of the perfusate in the control and the 1% gelatin group was not significantly different and remained at an average of 0.90 mPas (Fig. 3b), indicating that gelatin was no longer present in the solution.

Table 1.

Rat hearts recellularized with neonatal rat cardiomyocytes.

Recellularized hearts*	Total number of hearts injected	NRCMs retained after 30 min (%)	Recellularized beating hearts n (%)
Control	8	74.1 ± 3.2	1 (12.5)
1% Gelatin	6	85.0 ± 1.8**	5 (83.3)

*Hearts were recellularized with neonatal rat cardiomyocytes (NRCMs) injected in medium only (Control) or in medium with gelatin (1% Gelatin)

**P < 0.05, vs. Control

Hearts were maintained for 15 days to observe contractility with one heart contracting for 35 days until stopped for analysis. Contractility was observed in 5 of the 6 hearts (83%) that were re-endothelialized and subsequently recellularized with NRCMs delivered in 1% gelatin, but in only 1 of the 8 hearts (13%) from the control group with RAECs and NRCMs delivered without gelatin (Table 1, Supplementary Video 1). Visible contractions started at day 5 in both groups and were maintained up to 15 days when hearts were harvested; in one case, the heart showed contractions for 35 days until harvested. Endothelial cells lined the vessels, whereas the cardiomyocytes were found in the parenchyma (Figs. 3c and 3d, respectively).

Gelatin Aided in Placing Cells in Desired Locations Throughout the Heart During Recellularization

We examined cell distribution throughout the base, middle, and apex regions of the hearts, which were previously fixed and stained with Masson’s trichrome staining. In control hearts, few cells were found in the basal region, but groups of cells were found in the apex (Fig. 4a). However, in hearts recellularized with 1% gelatin, cells appeared more evenly distributed throughout the base, middle, and apex areas of the hearts (Fig. 4b). Qualitative comparisons of trichrome staining images taken from the crosssection of hearts showed differences in cell distribution among base, middle and apex (p < 0.05, n = 4 hearts) (Fig. 4c).

Figure 4.

Cardiomyocyte distribution within the recellularized rat heart. Masson’s trichrome staining of cross-section images from base, middle, and apex regions of rat recellularized (NRCMs) hearts using control medium (a) or 1% Gelatin (b). Scale bar: 200 μm for cross-section images (top) and 100 μm for close-up images. (c) Quantification of cellularity in the base, middle and apex of control (black bars) and 1% Gelatin (white bars) recellulalized hearts. *p < 0.05%, n = 4.

Discussion

This study showed that delivering cells in gelatin—a biocompatible, viscous material—improves cell retention in a thick, vascularized, whole organ scaffold to generate a contracting heart. We demonstrated that under the same cell density (less confounder), gelatin promoted cell adhesion on the surface of a decellularized aorta, increased immediate retention of RAECs delivered through decellularized vessel conduits into whole heart dECM, and improved immediate retention rate of NRCMs in decellularized hearts re-endothelialized with RAECs. Moreover, with gelatin, the better cell distribution increased the efficiency of matrix coverage at the apex, and especially in the middle area and base of hearts which were most affected by gravity. To avoid confusion about the indirect or direct impact of gelatin, RAECs without the presence of gelatin were delivered to the recellularized heart group prior to injection of NRCMs with and without 1% gelatin. Using our improved cell delivery method with the addition of gelatin, we report the highest cell retention values in cardiac dECM published to date, to the best of our knowledge.7,34,45 Our in vitro experiments showed that the increased gelatin-associated RAEC attachment in decellularized aorta was blocked by using a specific antibody against α5β1, indicating that integrin likely contributed to the improved cell attachment and retention observed in the dECM scaffold.

Gelatin is an ideal carrier for a wide range of applications, from a stabilizer in food products to a gelling agent in medical operations.38 Gelatin is naturally derived from collagen and thus is biocompatible and biodegradable, making it ideal for use in biomedical systems.53 Many groups have used gelatin in hydrogels cross-linked with other materials, such as glyceraldehyde61 and chitosan,22 to provide additional control over material properties14,48; however, this cross-linking can introduce toxic effects, reducing biocompatibility of the resulting hydrogel.29,60 In our short-term in vitro experiments, we confirmed the cell viability of endothelial cells with low concentrations of gelatin. However, 5% gelatin decreased viability and proliferation of endothelial cells over multi-day culture, a finding that has not been previously reported. Instead of cross-linking the 5% gelatin solution, we found that a lower 1% concentration of gelatin resulted in similar viability and proliferation of endothelial cells as using media alone.

We sought to balance the effects of gelatin in a flow and no flow environment under different temperatures to optimize cell attachment and viability in a physiologically-relevant flow model. The different concentrations of gelatin allowed us to test the effects of gelatin’s viscosity on the short-term and long-term cell attachment to a vessel in vitro with limited compromise in the cell viability. Choosing the low and high concentrations of gelatin allowed us to test two extremes in our experimental setup. This led us to believe gelatin was increasing cell attachment in a physiologically-relevant flow model. As a result of our in vitro findings, we used 1% gelatin in further ex vivo experiments. Our results indicate that gelatin can be used as a biocompatible carrier without any cross-linking but needs to be used within a proper working range to ensure long-term cell viability.

One of the main advantages of gelatin is that it can be used as an injectable carrier to target cell, drug, and growth factor delivery to a specific location.55 Gelatin allows for the controlled release of drugs and growth factors,54 but the delivery method can substantially affect how the final component acts at the delivery site.8 Injected microspheres can have a controlled release,37 whereas injected gelatin reduces cell scatter but degrades quickly.27,29 For our study, we intentionally used unique delivery methods for each cell type so that infused endothelial cells would perfuse through the coronary vasculature and injected cardiomyocytes would be localized to cardiac parenchyma. As a result, we found cells in their specific cardiac compartments, with endothelial cells lining vessels and cardiomyocytes clustered in the left ventricular wall. Other investigators have used infusion and injection methods for recellularization of the whole decellularized heart,16,26,65 but none has reported the addition of gelatin to the cell solution.

While gelatin has been used as a scaffold or cell-sheet carrier in ocular,20,51 osteogenic,40 and chondrogenic19 tissue engineering, a gelatin-based scaffold would not be appropriate in the whole heart engineering setting. The essential function of the heart is to contract and pump blood to the body. Building a bioartifical heart requires a scaffold that has the mechanical and biological properties necessary to sustain beating over a patient’s lifetime. A dECM bioscaffold with the elastic properties4 and preserved glycosaminoglycans18 provides mechanical and chemical cues so that cells function as in native tissue.47 ECM derived from whole organs has the physical microstructure of the original organ, giving tissue-specific spatial cues to cells for organization.11 Using gelatin as a cell carrier in cardiovascular tissue engineering helped to target cells to a specific location, but this was just the first step. Delivering cardiomyocytes to the parenchyma with gelatin helped to maintain contractile cells within the bioscaffold, which is an even more important endpoint.69 We achieved a higher percentage of beating hearts with gelatin in the delivery solution than with media only, indicating that gelatin improves not only cell retention but also cell function, which will aid in creating a fully functioning bioartificial heart. The relationship between cell retention and contractility is most likely closely linked, as having more cardiomyocytes in the scaffold provides a greater overall contraction force.64

Although the mechanisms by which gelatin improves cell retention, and thus contractility, have not been identified, several factors may be at play. Gelatin could improve cell retention during recellularization via its physical properties (viscosity), cell morphological changes, mechanoreceptors, or a combination of these factors. The physical properties of gelatin directly affect its ability to retain cells in a specific area. Gelatin offers tunable mechanical properties for delivering NRCMs into a specific region of interest. Increasing gelatin concentration in a solution increases the solution’s viscosity,56 as does decreasing the temperature.12 Initially, we tested high and low concentrations of gelatin to ensure suitable cell attachment, viability, and proliferation. We found that the lower concentration (1%) of gelatin fit our design criteria for increasing cell attachment to the matrix while not decreasing cell viability and proliferation. Under lower temperature, room temperature (20 °C), when 1% gelatin was first perfused and injected into the dECM, we found that gelatin stayed in the local area. After giving the cells an opportunity to attach, we increased the temperature (37 °C). At this higher temperature, the viscosity of the gelatin solution was the same as that in the media; therefore, the gelatin washed out, and the cells were able to interact with the natural environment of the dECM (Fig. 3).

In our study, when cells were infused at room temperature, adding 1% gelatin increased the viscosity of the cell suspension and increased cell retention by 28%, with cell retention rates higher than previously reported (10 to 54%).34,45 Although the no-gelatin control group showed improvement in cell retention (75%) compared to previous reports, we believe that our attribute this finding to the use of a pump-driven injection protocol that allows us to have constant-flow and control the temperature during the injection procedure. Increasing solution viscosity with gelatin decreased the chance of leaking from site of injection during cell injection, resulting in increased cell retention. This increased viscosity could also increase friction between the plasma membrane of the cell and the ECM surface, causing cells to roll rather than slide along the vessel wall.1 This rolling of the cells increases the surface area of the cell membrane that is exposed to the dECM, providing more opportunity for adherence and potentially contributing to increased endothelial cell retention in the recellularized matrix vessels. The ability to modify the physical properties of gelatin by adjusting the temperature provides significant advantages in this approach. The viscosity of gelatin at room temperature improves cell retention, and increasing the temperature leads to the removal of gelatin from the matrix. Having control over our recellularization methods is advantageous in that we can manipulate the process to retain more cells in the relevant compartments of the cardiac dECM bioscaffold and subsequently remove the cell carrier, which allows for the biologically driven remodeling of the ECM.30 This manipulation of temperature also allows important cell–matrix interactions to occur by maximizing spatiotemporal cell-ECM exposure.23

Another mechanism for improving cell attachment with gelatin may relate to morphological changes in the cells. The higher solution viscosity caused by the addition of gelatin can increase shear stress, which can promote cell retention and attachment to dECM by changing cell shape. Shear stress is the product of viscosity times the shear rate, which is determined by the diameter of the vessel, as per Newton’s Law. Given this, with a constant vessel diameter, shear stress increases when viscosity is increased. Thus, when cells are suspended in a gelatin solution, the shear stress and cell deformation increase within a fixed-diameter vessel.66 Furthermore, cell shape and morphology are critical predictors of endothelial cell spreading.21,42,44 When cells enter the flow stream in the coronary vasculature, their shape is determined by the shear stress.32,36 Therefore, this increase in shear stress combined with the viscosity of the gelatin may lead to increased spreading and attachment of endothelial cells to the dECM.

Lastly, another possible contributing factor is an alteration in the mechanosignaling at the cell surface, which can mediate cell adhesion and attachment. Shear stress can activate adhesion molecules and receptors on the endothelial cell membrane, causing downstream signaling of molecules such as receptor tyrosine kinases, integrins, G protein-coupled receptors, and stretch-activated ion channels.5,41,58 Many of these mechanoreceptors mediate cell attachment. As shear stress increased during perfusion in the vasculature with gelatin, the affinity of cells for the matrix may have also increased due to shear stress-induced activation of adhesion molecules. In addition, NRCM adhesion to matrices may be mediated by integrin receptors acting as mechanoreceptors.28,50

Together, our results build on existing applications of gelatin and suggest that gelatin, when used as a cell delivery vehicle, can improve recellularization in cardiac tissue engineering studies. Additional work is warranted to identify specific factors involved in the interactions between cells in gelatin and the dECM. We conclude that the presence of viscous gelatin promotes cell attachment and retention within the dECM vasculature and parenchyma.

Conclusion

Gelatin should be considered as a potential biocompatible cell carrier to increase cell retention during recellularization of dECM and to support re-endothelialization of the whole heart.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

dECM

Decellularized extracellular matrix

ECM

Extracellular matrix

NRCM

Neonatal rat cardiomyocyte

PBS

Phosphate-buffered saline

RAEC

Rat aortic endothelial cell

SDS

Sodium dodecyl sulfate