Assessment of Decellularized Pericardial Extracellular Matrix (ECM) and Poly(propylene fumarate) (PPF) Biohybrid for Small-Diameter Vascular Graft Applications

Abstract

Autologous grafts are the current gold standard of care for coronary artery bypass graft surgeries, but are limited by availability and plagued by high failure rates. Similarly, tissue engineering approaches to small diameter vascular grafts using naturally derived and synthetic materials fall short, largely due to inappropriate mechanical properties. Alternatively, decellularized extracellular matrix from tissue is biocompatible and has comparable strength to vessels, while poly(propylene fumarate) (PPF) has shown promising results for vascular grafts. This study investigates the integration of decellularized pericardial extracellular matrix (dECM) and PPF to create a biohybrid scaffold (dECM+PPF) suitable for use as a small diameter vascular graft. Our method to decellularize the ECM was efficient at removing DNA content and donor variability, while preserving protein composition. PPF was characterized and added to the dECM, where it acted to preserve the dECM against degradative effects of collagenase without disturbing the material’s overall mechanics. A transport study showed that diffusion occurs across dECM+PPF without any effect from collagenase. dECM+PPF matched the modulus of human coronary arteries and saphenous veins. dECM+PPF vascular grafts demonstrated ample circumferential stress, burst pressure, and suture retention strength to survive in vivo. An in vivo study showed re-endothelialization and tissue growth. Overall, the dECM+PPF biohybrid presents a robust solution to overcome the limitations of current methods of treatment for small diameter vascular grafts.

Graphical Abstract

Introduction

There are about 400,000 coronary artery bypass graft surgeries conducted each year in the United States, as a common approach to treat cardiovascular disease (CVD).¹ This surgery utilizes a small diameter conduit to divert blood flow around blockages in the coronary artery.² Autologous grafts, typically from the saphenous vein or alternatively the internal mammary artery, are the current gold standard of care.^3,4 Autologous grafts are limited by the quality, size, and availability from the donor, cause significant donor site morbidity, and increase surgical cost.^5–8 Furthermore, bypasses commonly fail because of additional blockages from thrombosis and the irregular performance of vein grafts.⁹ The current CVD treatments are not sufficient, as evident by the high failure rates of autologous grafts; saphenous vein grafts have a 50% failure rate after 10 years.⁸ Due to the limitations of the current method of treatment, alternative solutions using tissue engineering approaches have been investigated.

Natural and synthetic materials have been unsuccessfully employed to produce viable small diameter vascular grafts (<6 mm). Natural scaffolds have substantially lower strengths than natural vessels and can degrade faster than regeneration can repair the implant.^8,10 Although decellularized extracellular matrix from tissue has comparable strength to vessels, is biocompatible, and supports regeneration, the use of decellularized vessels has led to thrombosis and infection, which have prevented their widespread clinical use.^11–13 Alternatively, in addition to matching the elastic modulus of human arteries, bovine pericardium is nonthrombogenic, resists infection, and supports endothelial cell attachment and growth.^14–17 Synthetic grafts fail due to chronic inflammatory responses to the foreign materials and compliance mismatch, stemming from the difference in mechanical properties of the graft and tissue.^18,19 Synthetic materials also have poor re-endothelialization, which contributes to the high incidences of thrombosis.^5,20 Poly(propylene fumarate) (PPF) has shown promising results for vascular grafts, and more generally, preserves mechanical properties of pericardium and reduces chronic inflammation associated with native bovine pericardium.^21–23

Here, we propose a decellularized bovine pericardium (dECM) vascular graft, combined with PPF, to create a biohybrid (dECM+PPF) that has the ability to serve as a practical alternative to autologous grafts due to its short manufacturing times and off-the-shelf availability.^10,24 The proposed graft accomplishes these objectives largely due to the materials it is composed of: bovine pericardium dECM and PPF. Here, we present an abbreviated protocol to decellularize bovine pericardium, a readily available material, in 49 hours. Additionally, PPF is synthesized in 3 days and can be stored for months.²⁵ After 1 day of dehydration, dECM+PPF vascular grafts can be manufactured within a couple hours. This entire process takes only a week, a stark contrast to those created after months of cell culture²⁶, while the raw materials are low in cost and can be stored along the way. These grafts have the potential to make a huge impact on the manufacturing of coronary artery bypass grafts and ultimately, the treatment of patients.

The innovation of the dECM+PPF graft is two-fold. First, bovine pericardium has yet to be utilized as a xenographic small diameter vascular graft. Bovine pericardium has been commonly used for cardiovascular applications²⁷, including cardiac patches²⁸, vascular regeneration^29,30, and valve repair³¹. Equine pericardium has been utilized to cover stents to treat coronary perforations³² and coronary artery thrombus³³, while canine pericardium has been used as an autologous coronary artery graft. Despite the successful employment of pericardium in cardiovascular applications, bovine pericardium has not been investigated as a xenograph small diameter vascular graft for humans. Second, although our group has previously developed and characterized a pericardium/PPF biohybrid^21,22,34, it has yet to be implemented for a specific application or function; the innovation lies not in the biohybrid material, but in its application. Therefore, we will utilize this new material to build a small diameter vascular graft. In addition, the previous biohybrid used native pericardium, where the layer of PPF modulated the immune response by the slow exposure of pericardium to the body.^21,22 However, it is critical to decellularize the pericardium before implantation to remove the cellular content, help minimize any inflammatory response, and delay biodegradation of the scaffold.^35–39 It has been shown that acellular pericardium has no immune response and demonstrates promising signs of vascularization, while the native material induces an immune response with significantly less recellularization.⁴⁰ Therefore, this project will utilize decellularized pericardium to avoid in vivo immune responses, while promoting vascular graft remodeling.⁴¹

This study investigates the development and use of a decellularized pericardium and poly(propylene fumarate) biohybrid as a vascular graft. We believe a vascular graft constructed from dECM+PPF will provide adequate strength for its application, without causing vascular graft failure associated with synthetic materials. We accomplished this goal through first, developing a new protocol to decellularize bovine pericardium, then investigating how the ideal processing conditions of PPF and how the addition of PPF affects the degradation of dECM, and finally, assessing the ability of the biohybrid to function as a vascular graft, both in vitro and in vivo.

Materials and Methods

Bovine Pericardium Decellularization Protocol

Sheets of native bovine pericardium (Innovative Research, Novi, MI) (nECM) were cut into 5 × 5 cm samples. Samples were placed in 20 mL of 1,4-piperazinediethanesulfonic acid (PIPES) buffer (8 mM pH 6.8 PIPES (bioWORLD, Dublin, OH), 1 M NaCl (Thermo Fisher Scientific, Waltham, MA), and 25 mM ethylenediaminetetraacetic acid (EDTA, Sigma-Aldrich, St. Louis, MO) in phosphate buffered saline (PBS)) and agitated on a plate shaker at 100 rpm at room temperature for 15 h. Samples were then washed in PBS three times for 15 min to remove the PIPES buffer, placed in 20 mL of sodium dodecyl sulfate (SDS, Sigma-Aldrich) buffer (1.8 mM SDS, 1 M NaCl, and 25 mM EDTA in PBS), and agitated on a heated plate shaker at 37 °C at 70 rpm for 7 h. Exposure to SDS was limited in concentration and duration due to its potential effects on collagen.⁴² All samples were then rinsed in PBS three times and washed with Medium 200 (Thermo Fisher Scientific, with manufacturer’s required Low Serum Growth Supplement), supplemented with 12% fetal bovine serum (FBS, Gibco, Gaithersburg, MD) and 1% Pen Strep (Gibco) for 24 h. Finally, the samples were rinsed thrice with PBS. This procedure resulted in decellularized pericardial extracellular matrix tissue (dECM).

DAPI Staining

Three samples (3 × 3 mm) from nECM and from dECM were collected from three donors each (n=9). These tissues were stained with DAPI solution (1.5 μg/mL in PBS, Invitrogen, Carlsbad, CA) for 10 min. Samples were removed from DAPI solution, rinsed with PBS twice, and imaged with an Eclipse Ti2 microscope (Nikon, Tokyo, Japan) under DAPI excitation (340–380 nm) and emission (435–485 nm).

DNA Quantification

Three samples from nECM and from dECM were collected from three donors each and lyophilized. 5 mg of lyophilized tissue was digested using the DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany), according to the manufacturer’s protocol. The Quant-iT™ PicoGreen™ dsDNA Assay Kit (Thermo Fisher Scientific) was used to quantify DNA in the processed samples versus the kit DNA standard and buffer AE (Thermo Fisher Scientific). Each sample was diluted 1:16 in buffer AE and run in technical triplicate (n=27). 100 μL of the diluted sample was combined with 100 μL of 0.5% PicoGreen in 20X TE buffer (diluted to 1X, Thermo Fisher Scientific) and agitated on a plate shaker at room temperature at 50 rpm for 5 min. The fluorescence was read using an M5 SpectraMax plate reader (Molecular Devices, San Jose, CA) at 485 nm excitation and 538 nm emission. The DNA content was quantified from fluorescence based on the DNA standard, and then normalized to the starting dry mass of tissue.

Gel Electrophoresis of Proteins

Three samples (3 × 3 mm) from nECM and from dECM were collected from three donors each and lyophilized (n=9). 20 mg of diced, lyophilized tissue was digested in 2 mL of pepsin solution (1 mg pepsin (Thermo Fisher Scientific) in 1 mL pH 2 Milli-Q). Samples were placed on a shaker at room temperature for 3 d. Samples were neutralized with 1/9 volume of 0.1 N NaOH and then 1/9 of the total volume of 10X PBS was added. The digested samples were strained with 40 μm nylon cell strainers (Corning, Corning, NY) so that they could be pipetted. Samples were diluted 1:1 in PBS. An 11.5% collagen type I (rat tail high concentration, Corning) solution in 0.5 M acetic acid (Sigma-Aldrich) and a pepsin solution of 5 mg/mL in PBS were prepared as controls. 10 μL of 4x laemmli protein sample buffer (Bio-Rad, Hercules, CA) with 10% 2-mercaptoethanol (Sigma-Aldrich) was added to 30 μL of each sample. NaOH was added to the collagen sample to adjust its pH until its color matched that of all other samples. Samples were placed in a water bath at 95 °C for 5 min. 5 μL of Precision Plus Protein™ WesternC™ Blotting Standards (Bio-Rad) was loaded, alongside 20 μL of each sample, into Mini-PROTEAN® TGXGels™ Gels (4–15%, Bio-Rad). The gels were run in 10X Tris/Gline/SDS Buffer (Bio-Rad), diluted to 1X, at 120 V until completion. They were stained with Coomassie blue (Bio-Rad) and rinsed with DI water to remove excess dye before imaging on Odyssey CLx Imager (LI-COR, Lincoln, NE).

PPF Synthesis and Application to dECM

PPF was synthesized from propylene glycol (Sigma-Aldrich) and diethyl fumarate (DEF, Sigma-Aldrich) according to a previously published protocol.²⁵ The resulting PPF (number average molecular weight of 1500) was diluted with DEF in a ratio of 2 parts PPF: 1 part DEF by mass.²¹ 20 μL of 40% wt/vol bis(2,4,6-trimethylbenzoyl) phenylphosphine oxide (BAPO, Irgacure 819, BASF Corporation, Florham, NJ) in dichloromethane (Thermo Fisher Scientific) was added to 1.25 g of diluted PPF before UV crosslinking by free radical polymerization⁴³. To create thin films of PPF, PPF was spread on a microscope slide and then covered with a second slide, which resulted in films roughly 0.05 mm thick. The slides were placed in the UV box to be crosslinked for 45 min, unless otherwise specified.

To add the PPF onto decellularized tissue, a method derived from Bracaglia, et al. was used.²¹ dECM samples were secured on wooden frames to prevent them from shrinking during dehydration. Samples were dehydrated using serial ethanol washes (20%, 40%, 60%, 80%, and 100%) for 5 min each. Samples were placed in 3 additional 100% ethanol baths for 5 min each and then allowed to air dry. PPF (20 μL of 40% wt/vol BAPO in dichloromethane per 1.25 g PPF) was added to the dECM. Glass microscope slides were placed under the dECM and on the PPF. Binder clips were placed around the entirety of the microscope slides to secure the materials. The PPF was exposed to UV light to allow crosslinking for 45 min, unless otherwise specified. All 8 ethanol washes were repeated in descending order. The samples were then rinsed with PBS and stored in PBS.

Tensile Testing

Three dogbone samples in accordance with ASTM D638 from dECM and from dECM+PPF were collected from three donors each (n=3 per donor, n=9). Nine dogbones were cut from thin film PPF (n=9). All sample groups were duplicated, and the second set was subjected to collagenase enzymatic degradation for 4 h, described below. All dogbones were then rinsed with PBS and stored overnight at 4 °C in PBS until testing. Dogbone samples were tested at room temperature on an Instron 5565 mechanical tester (Instron, Norwood, MA) at a constant displacement rate of 10 mm/min, with a preload of 0.1 N (strain=0). Tensile tests for biological samples were approximately 60 s long to ensure that the samples remained hydrated. Samples that broke outside of the gauge length were excluded and replaced. The ultimate tensile strength was recorded as the maximum stress that the sample reached. The ultimate strain was recorded as the strain at the point of the maximum stress. The E2 modulus was the slope of the stress-strain curve measured between 40% and 80% of the maximum stress.

Dynamic Mechanical Analyzer

PPF Thin Films

Five rectangular samples (25.4 × 4 mm) were collected from thin films of PPF for the corresponding UV exposure time (n=5). Samples were tested immediately after preparation at room temperature on the Q800 dynamic mechanical analyzer (DMA, TA Instruments, New Castle, DE). A ramp displacement test was run at a speed of 1000 μm/min until failure. The preload of the machine was set to 0.010 N. The modulus was calculated from the linear region of the stress-strain curve.

dECM and dECM+PPF Samples

Three rectangular samples (25.4 × 4 mm) from dECM and from dECM+PPF were collected from three donors each (n=9). These sample groups were duplicated, and the second set was subjected to collagenase enzymatic degradation for 4 h. The samples were then rinsed with PBS and stored overnight at 4 °C in PBS until testing. The test was run at room temperature on the Q800 DMA with a preload of 0.010 N and at a strain rate of 75%/min until failure. The DMA was selected for the E1 modulus test due to its high sensitivity in force measurements. The E1 modulus was calculated from the slope of the stress-strain curve until the R² value reached 0.95 and 0.98, respectively.

Cell Metabolic Activity Assay (XTT)

PPF Thin Films

6 square samples (6.5 × 6.5 mm) were collected from thin films of PPF for the corresponding UV exposure time and for each time point of the assay. Upon crosslinking, the PPF slides were kept under sterile conditions. Half of the squares were used after crosslinking and half of the squares were washed with ethanol. The ethanol protocol consisted of four 100% ethanol rinses for 5 min and then serial ethanol washes (80%, 60%, 40%, and 20%) for 5 min each. All samples were soaked in Minimum Essential Medium (MEM, Gibco) supplemented with 10% horse serum (ATCC, Manassas, VA) and 1% Pen Strep (Gibco) for 30 min. Cells from a fibroblast cell line (L929, ATCC) were seeded in a 48 well plate at a density of 20,000 cells/well in 0.4 mL of medium. Cells were given 4 h to adhere to the well before 0.5 mL of additional medium and the PPF samples were added. Wells were left without PPF samples as a control. The electron coupling and XTT labeling reagents from the Cell Proliferation Kit II (XTT) (Roche, Mainheim, Germany) were mixed according to the manufacturer’s protocol and added to the cell medium in a ratio of (1:2). Cell medium was removed from the wells and the XTT solution was added. It was incubated for 4 h and 100 μL was extracted in triplicate (n=9). The absorbance was measured with an M2e SpectraMax plate reader (Molecular Devices, Sunnyvale, CA). Net absorbance was calculated (A₄₇₅-A₆₅₀) for each sample and XTT fold change was calculated by normalizing the net absorbance to that at 8 h.

dECM and dECM+PPF

3 square samples (6.5 × 6.5 mm) from dECM and from dECM+PPF were collected from three donors each for each time point of the assay. All samples were UV sterilized for 30 min and soaked in Medium 200 (with LSGS), supplemented with 1% Pen Strep, for 3 h. Human umbilical vein endothelial cells (HUVECs, ThermoFisher Scientific) were seeded in a 48-well plate at a density of 50,000 cells/well in 0.4 mL of media. Cells were given 4 h to adhere before 0.5 mL of additional medium and the dECM or dECM+PPF samples were added. 9 wells were left without dECM or dECM+PPF samples as controls. The Cell Proliferation Kit II (XTT) (Roche) was used as described above (n=27). Absorbance was measured using a Spark Multimode Microplate Reader (Tecan, Männedorf, Switzerland). Net absorbance was calculated (A₄₇₅-A₆₅₀) for each sample and XTT fold change was calculated by normalizing the net absorbance to that at 24 h.

Collagenase Enzymatic Degradation Study

One rectangular sample (25.4 × 6.5 mm) from dECM and from dECM+PPF were collected from three donors each for each time point, as well as three samples of PPF for each time point (n=3). The initial width of each sample was measured, then lyophilized and weighed (initial mass). They were then submerged in 12 mL of 117 U/mL collagenase II (Worthington Biochemical Corporation, Lakewood, NJ) in PBS. Collagenase was used to mimic in vivo degradation.^14,44 The samples were agitated on a temperature controlled plate shaker at 37 °C at 73 rpm 2, 4, 8, 12, 16, 20, 24, 28, and 32 h. They were rinsed with PBS and the width was recorded (final width). Samples were lyophilized and the dry mass was recorded (final mass). Both the original mass and width, normalized to the initial time point, were calculated using the following equation

$$\mathit{\text{original }}x,\mathit{\text{normalized }}tothe\mathit{\text{initial time }}\mathit{\text{point}}=100-\frac{x_{\mathit{\text{initial}}}-x_{\mathit{\text{final}}}}{x_{\mathit{\text{initial}}}}*100$$

where x is either mass or width. Digestion of other test samples in collagenase was conducted for 4 h.

Permeability Measurement

The permeability of fluorescein isothiocyanate (FITC), a model small molecule, in the sample materials was assessed by a molecular transport study. Three circular samples (22.5 mm diameter) from dECM and from dECM+PPF were collected from three donors each (n=3). Three circular samples (12.5 mm diameter) were cut from three thin films of PPF (n=3). All sample groups were duplicated, and the second set was subjected to collagenase enzymatic degradation for 4 h. The samples were placed in 12-well transwells (Corning) with a hollow cylindrical stopper (8 mm inner diameter) to hold the sample in place.⁴⁵ 3 mL of PBS was added to the well plate and 0.5 mL of FITC solution (1 mg/mL in PBS, Sigma-Aldrich) was added to the top of the transwell, inside of the stopper. 4 μL was sampled from the donor (top) and receiver (bottom) at each time point and diluted in 196 μL of PBS. The sampling led to a 6% volume loss from the donor and 1% volume loss from the receiver, thus we assumed constant volume throughout the test. 50 μL of the diluted sample was run in triplicate. A high concentration ladder (100 to 0 μg/mL) and a low concentration ladder (2.5 μg/mL) of FITC were constructed using serial dilutions in PBS. The FITC concentration in each well was quantified by interpolation based on standard curves of known FITC content. We assumed that the concentration profile across the membrane is constant to use a pseudo-steady state model to calculate the permeability (P) of the samples by plotting

$$-ln\left(\frac{C_d-C_r}{C_{d0}-C_{r0}}\right)=\frac{PtA}{l}\left(\frac{1}{V_d}+\frac{1}{V_r}\right)$$

where C_d and C_r are the concentrations of the donor and receiver, respectively, C_d0 and C_r0 are the initial concentrations of the donor and receiver, P is permeability, t is time, A is surface area, l is thickness, and V_d and V_r are the volumes of the donor and receiver, as previously described.⁴⁵

Manufacturing of dECM+PPF Vascular Grafts

Sheets of dECM were secured on wooden frames to prevent shrinking during dehydration. Framed samples were dehydrated using serial ethanol washes (20%, 40%, 60%, 80%, and 100% ethanol in DI water) for 5 min each. Samples were then placed in 3 additional 100% ethanol baths for 5 min each and then allowed to air dry. The dehydrated pericardium was cut into a rectangle approximately 3.5 cm by 1.75 cm and rolled. PPF (20 μL of 40% wt/vol BAPO in dichloromethane per 1.25 g PPF) was applied to the overlap seam and then over the outer surface of the graft. The graft was placed inside PVC clear vinyl tubing (Home Depot, Atlanta, GA), UV crosslinked for the specified amount of time (15, 45, or 60 min), rotated 180°, and UV crosslinked again for the same amount of time. All 8 ethanol washes were repeated in descending order. The samples were then rinsed with PBS and stored overnight at 4 °C in PBS until testing.

Circumferential Stress Measurement

Vascular grafts were manufactured in triplicate from three different sheets of dECM and soaked in PBS overnight at 4 °C until testing. All sample groups were duplicated, and the second set was subjected to collagenase enzymatic degradation for 4 h. Three rings were cut from each vascular graft (n=9). Each ring was mounted using two U-shaped metal fixtures, as described in literature.⁴⁶ They were tested at room temperature on an Instron 5565 mechanical tester at a constant displacement rate of 10 mm/min. Samples were preloaded to 0.1 N (strain=0). The maximum force was recorded and used to calculate the maximum stress from the following equation:

$${\sigma }_{\mathit{\text{circumferential}}}=\frac{F_{max}}{2A}$$

where F_max is the maximum force and A is the area, as a function the thickness of the wall and width of the ring. Additionally, the right coronary artery (RCA) was isolated and dissected from four porcine hearts (Animal Biotech Industries, Danboro, PA), preserved in PBS at 4 °C, and tested in the same manner as graft samples within 24 h of harvest (n=12).

Burst Pressure Testing

Vascular grafts were manufactured in triplicate from three different sheets of dECM and soaked in PBS overnight at 4 °C until testing. All sample groups were duplicated, and the second set was subjected to collagenase enzymatic degradation for 4 h. Each vascular graft was attached to a burst pressure testing system, as described in ANSI/AAMI VP20—1994, using a suture and an elastic tie. The burst pressure system was composed of a NE-300 syringe pump (New Era Pump System, Farmingdale, NY) with a 6 mL syringe that is connected to the PRTemp1000 pressure gauge (MadgeTech, Warner, NH), and ultimately, connected to the vascular graft. Methyl cellulose (15 cP, 1% in PBS, Sigma-Aldrich) was used in the system to increase viscosity. The methyl cellulose solution was initially flowed through the graft to remove air, after which the open end of the graft was clamped shut. The pressure of the system was measured as the methyl cellulose was dispensed at a rate of 500 μL/min, which corresponded to a rate of approximately 5 mmHg/s. The test was stopped when the graft was unable to retain pressure. The maximum pressure obtained was recorded (n=3).

Suture Retention Strength

Three rectangular samples (25 × 10 mm) were cut from three sheets of dECM+PPF (n=9). All sample groups were duplicated, and the second set was subjected to collagenase enzymatic degradation for 4 h. 3/0 sutures (Ethicon, Somerville, NJ) were looped through each sample, leaving 2 mm of tissue perpendicular to the 10 mm edge of the sample and the suture site, as described in ANSI/AAMI VP20—1994. Each suture was then triple-knotted to form a secure loop. All samples were soaked in PBS overnight at 4 °C until testing. Samples were loaded with the suture in the top grip and the dECM+PPF rectangle in the lower grip and tested at room temperature on an Instron 5565 mechanical tester at a constant displacement rate of 0.1 mm/sec. The maximum force was reported as the suture retention strength (N). Additionally, the left anterior descending artery (LADA) was isolated and dissected from four porcine hearts, preserved in PBS at 4 °C, and tested in the same manner as dECM+PPF samples within 24 h of harvest (n=4).

In Vivo Assessment of Patency in a Murine Animal Model

Two vascular grafts were manufactured from individual sheets of dECM. Each graft was implanted into a female Lewis nude rat (8–10 weeks) as an abdominal aorta interposition conduit (n=2). This model was used because it is commonly employed to model mid-term immune response and graft remodeling for new vascular graft materials.^4,47 The surgical procedures were conducted in a manner consistent with previous publications.^23,48 The Institutional Animal Care and Use Committee of Johns Hopkins University approved the study, and all animals were treated in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Rats were anesthetized, placed in the supine position, and an abdominal midline incision was made. The abdominal aorta was exposed, cross-clamped, and excised. Grafts approximately 5 mm in length were inserted and secured using a 10–0 nylon suture for the proximal and distal anastomoses. Rats were recovered from surgery and maintained without antiplatelet or anticoagulation therapies. Two weeks after the procedure, anesthetized rats were sacrificed.

Histology

Grafts were excised and fixed in 10% neutral buffered formalin and embedded in optimal cutting temperature compound for frozen sectioning. After cryosectioning, sections were stained with hematoxylin and eosin (H&E) stain and Masson’s trichrome. To characterize vascular neotissue in vascular grafts, sections were blocked for endogenous peroxidase activity (0.3% H₂O₂ in MeOH) and nonspecific background staining (Background Sniper, Biocare Medical, Concord, CA). Antigens were retrieved using the citrate buffer method (pH 6.0, 90 °C). Primary antibodies used were: anti-α-Smooth Muscle Actin (1:500, M0851, Dako, Carpinteria, CA) and anti-CD31 (1:50, Ab28364, Abcam, Cambridge, MA). Slides were imaged with an Eclipse Ni microscope (Nikon).

Statistics

All quantitative assessments of two experimental groups were statistically compared using a two-sample t-test. All quantitative assessments of three or more experimental groups were statistically compared using a one-way ANOVA test, followed by a post hoc Tukey’s test. All tests assumed equal variance and conducted with 95% confidence intervals (p<0.05). Capital and lowercase letters differentiate between separate ANOVA analysis (typically, original and digested).

Results

Decellularization of Bovine Pericardium

An abbreviated decellularization method was developed from a previously published protocol⁴⁹ to decellularize bovine pericardium, as summarized in Figure 1A. The effectiveness of this new decellularization protocol was qualitatively examined using DAPI staining (Figures 1B and 1C). A qualitative reduction in DNA content was observed in dECM when compared to nECM. This difference was quantified with a PicoGreen assay (Figure 1D). The original DNA content of nECM varies significantly between the donors (p<0.05; Tukey multiple comparison; n=9). After the tissues were decellularized, the dECM from each donor contained below 12 ng residual DNA/mg tissue, a significant decrease from nECM for all donors (p<0.001; two-sample t-test; n=9) and independent of donor (p>0.05; Tukey multiple comparison; n=9). This decellularization protocol reduces the amount of DNA from an average of 2005.3 to 11.2 ng DNA/mg dry tissue, a reduction of over 99%. The decellularization process does not appear to remove significant tissue proteins as assessed using gel electrophoresis (Figure 1E). Protein bands present in the nECM can still be observed in the dECM. In addition to the standard ladder, collagen I and pepsin were run alongside the samples, and both proteins are visible in nECM and preserved in dECM. When considering mechanical changes, the ultimate tensile strength of both nECM and dECM is independent of donor (p>0.05; Tukey multiple comparison; n=3) (Figure 1F). Additionally, there is no significant change in the ultimate tensile strength caused by the decellularization process (p>0.05; two-sample t-test; n=3).

Figure 1: Decellularization of bovine pericardium.

(A) Native pericardium (nECM) was washed using sequential rinses of PIPES, SDS, and media/FBS/Pen Strep to produce decellularized pericardium (dECM). DAPI fluorescent imaging was used to qualitatively confirm the reduction in DNA content in the tissues, by comparison of (B) nECM and (C) dECM samples, indicative of complete decellularization (scale bars=200 μm). (D) DNA quantification from PicoGreen assays show significant reduction in DNA content of dECM to below the 50 ng DNA/mg tissue threshold (*p<0.001; two-sample t-test), independent of variability in initial donor value (p< 0.05; Tukey multiple comparison). (E) Electrophoresis gel displaying protein bands for nECM (lanes 5–7) and dECM (lanes 8–10), compared to the standard ladder (lane 1), collagen I (lane 2), pepsin (lane 3). Lane 4 was intentionally left blank. Protein bands in dECM indicate that nECM protein composition is preserved through decellularization, including collagen I. (F) Ultimate tensile strength data of nECM and dECM show no significant differences due to the decellularization process, regardless of donor (p>0.05; Tukey multiple comparison, two-sample t-test). Groups with different letters indicate statistical difference. All data is presented as mean + standard deviation.

Mechanical Properties of Crosslinked PPF and its Addition to dECM

The elastic modulus of PPF was measured as a function of UV exposure time (Figure 2A), demonstrating that the elastic modulus of PPF increases with UV exposure time (p<0.05; Tukey multiple comparison; n=5). The elastic modulus experiences an insignificant upward trend from 2 to 45 min, and reaches a plateau at 45 min. The metabolic activity of L929 cells exposed to PPF was measured as a function of UV exposure time in minutes, with and without graduated ethanol washes, over 24, 48, and 72 hours (Figures 2B–D). The normalized values of the metabolic activity of the cells in the ethanol washed PPF group and the cell control increase with time, while those in the unwashed PPF group do not (p<0.05, p>0.05, respectively; Tukey multiple comparison; n=9). Furthermore, the UV crosslinking time does not affect the XTT fold change with any discernable trend (p>0.05; Tukey multiple comparison; n=9). We utilized these results to apply PPF to pericardium and introduce graduated ethanol washes for rehydration after UV crosslinking (Figure 2E).

Figure 2: Application of PPF to dECM.

(A) Elastic modulus (MPa) exhibited by PPF, crosslinked under different UV exposure times. The elastic modulus of PPF trends upward between 2 and 45 min and reaches a plateau at 45 min (p> 0.05, Tukey multiple comparison). (B-D) Fold change of the XTT metabolic activity as a function of time (normalized to 8 h) for L929 cells exposed to PPF of various UV exposure times (reported in minutes). Ethanol washes (denoted “e”) after UV crosslinking of PPF improves the metabolic activity of cells (p< 0.05, Tukey multiple comparison). (E) dECM was dehydrated with graduated ethanol washes in order to apply PPF. The system was then crosslinked under UV light for different periods of time, followed by rehydration in reverse graduated ethanol washes, to produce a biohybrid material of decellularized pericardium reinforced with PPF (dECM+PPF). Groups with different letters indicate statistical difference. Data is presented as mean + standard deviation, except for the elastic modulus where data is mean ± standard deviation.

Characterization of the dECM+PPF Biohybrid

To model dimensional changes during degradation in vivo, dECM, dECM+PPF, and PPF samples were placed in collagenase II (Figure 3A and 3B). There is significantly more mass remaining for samples of dECM+PPF and PPF than for those of dECM at 4, 8, 12, 16, and 32 h (p<0.05; Tukey multiple comparison; n=3). Similarly, the width change of dECM is significantly larger than that of dECM+PPF and PPF at 12, 16, 20, 24, and 28 h (p<0.05; Tukey multiple comparison; n=3). Figures 3C–E and Figures 3F–H show images of the dECM, dECM+PPF, and PPF before and after 32 h collagenase degradation, respectively. The images show the qualitative changes in mass and width (arrows) for each material during the degradation study, as quantified in Figures 3A and 3B. The concentration of FITC, normalized to initial concentration at 0 h, was measured as a function of time for dECM, dECM+PPF, and PPF (Figure 3I) in order to calculate permeability (Figure 3J) and determine how the addition of PPF affects the permeability of dECM. The permeability of the dECM+PPF is 3.67E-8 ± 1.12E-8 cm²/s, which is comparable to that of dECM and PPF (p>0.05; Tukey multiple comparison; n=3).

Figure 3: Enzymatic degradation of dECM, dECM+PPF, and PPF.

(A) Original mass (%) and (B) original width (%) of material samples, normalized to the initial time point, as a function of time after digestion in 117 U/mL collagenase II solution. When compared to dECM, both dECM+PPF and PPF experience significantly less weight loss and change in width after collagenase digestion, as indicated (*p< 0.05; Tukey multiple comparison). Representative images of (C) dECM, (D) dECM+PPF, and (E) PPF before collagenase digestion and (F) dECM, (G) dECM+PPF, and (H) PPF after 32 h collagenase degradation, arrows indicate the measured width. dECM loses its shape with digestion, while both dECM+PPF and PPF retain their shape. The images display qualitative changes in mass and width of each material with collagenase degradation. (I) FITC concentration (%) normalized to initial concentration at 0 h, as a function of time. (J) The addition of PPF to dECM does not significantly impact the permeability (cm²/s) of dECM+PPF. Groups with different letters indicate statistical difference. Data is presented as mean + standard deviation, except for the permeability where data is mean + standard deviation.

Ultimate tensile strength, maximum strain, E1 modulus, and E2 modulus, as defined in the methods section, are illustrated by Figure 4A. Figure 4B displays average tensile curves of dECM, dECM+PPF, PPF before and after 4 h in collagenase II. All three materials undergo significant changes in ultimate tensile strength after 4 h in collagenase II (Figure 4C). The ultimate tensile strength decreases in both dECM and dECM+PPF, but increases in PPF (p=0.003, 0.014, and 0.002, respectively; two-sample t-test; n=9). However, the maximum strain or E1 modulus (Figures 4D–4E) for all materials remain unchanged with the collagenase degradation (p>0.05; two-sample t-test; n=9). The E2 modulus of PPF significantly increases from 543 ± 180 MPa to 1170 ± 135 MPa (p<0.001, two-sample t-test; n=9) with the collagenase degradation, while that of dECM and dECM+PPF is not significantly impacted (Figure 4F). The maximum strain of dECM+PPF is significantly higher than that of dECM and over an order of magnitude higher than that of PPF (p<0.05; Tukey multiple comparison; n=9). dECM+PPF has an E1 modulus of 7.3 ± 11 MPa and an E2 modulus of 107 ± 45 MPa. There is no significant difference when comparing the E1 and the E2 modulus of dECM and dECM+PPF (p>0.05; two-sample t-test and Tukey multiple comparison, respectively; n=9). The E2 modulus of PPF is significantly higher than that of dECM and dECM+PPF (p<0.05; Tukey multiple comparison; n=9).

Figure 4: Mechanical evaluation of dECM, dECM+PPF, and PPF in tension.

(A) A typical curve for dECM to illustrate ultimate tensile strength, maximum strain, E1, and E2. (B) Average curves for the groups studied to illustrate differences in mechanical behavior. (C) dECM and dECM+PPF undergo a significant decrease, while PPF experiences a significant increase, in ultimate tensile strength after 4 h in collagenase (*p=0.003, **p=0.014, and ***p=0.002; two-sample t-test). (D) Degradation does not have a significant impact on the materials’ deformation at failure. The addition of PPF to dECM does not significantly change the (E) E1 modulus (MPa) or (F) E2 modulus (MPa). Groups with different letters indicate statistical difference. All data is presented as mean + standard deviation.

dECM+PPF as a Vascular Graft

Vascular grafts were manufactured as illustrated in Figure 5A, by applying the PPF resin onto the rolled dECM and crosslinking in a translucent tube. The circumferential stress (Figure 5B) and burst pressure (Figure 5C) of the vascular graft and suture retention strength (Figure 5D) of the dECM+PPF were tested for UV crosslinking times of 15, 45, and 60 min. There is an increase in circumferential stress from 2.24 ± 0.99 MPa to 4.20 ± 2.05 MPa between 15 and 45 min and there is no significant change between 45 and 60 min (p<0.05, p>0.05, respectively; Tukey multiple comparison; n=9). The circumferential stresses measured for the dECM+PPF grafts, independent of UV exposure time, are comparable to that of porcine RCA (p>0.05; Tukey multiple comparison; n=9 and n=12 for RCA). Additionally, there is no loss of strength after 4 h in collagenase for 15, 45, and 60 min UV exposure times. The burst pressure of the vascular grafts increases from 104 ± 160 mmHg to 404 ± 61 mmHg and 408 ± 45 mmHg from 15 to 45 and 60 min of UV exposure (p<0.05; Tukey multiple comparison; n≥3). This difference diminishes after 4 h in collagenase. In both circumferential stress and burst pressure tests, the vascular grafts failed at the overlap seam. The suture retention strength is independent of the UV exposure times and these values are significantly greater than those of porcine LADA (p<0.05; Tukey multiple comparison; n=9 and n=4 for LADA). The 45 and 60 min groups do experience significant decrease in strength over a 4 h collagenase degradation (p<0.05; two-sample t-test; n=9). The normalized value of the metabolic activity for dECM+PPF, shown in Figure 5E, is larger than that of dECM at 72 h, while both are significantly lower than that of the cell control (p<0.05; Tukey multiple comparison; n=27 for samples and n=9 for cell control).

Figure 5: dECM+PPF as a vascular graft.

(A) Process of rolling dehydrated dECM, applying PPF, and crosslinking using UV light to create a biohybrid vascular graft (e.g. graft with an inner diameter of 3 mm and 3 cm long as pictured). (B) Circumferential stress (MPa) measured from the rings of the vascular graft and porcine right coronary artery (RCA). The circumferential stress of the graft is comparable to that of RCA, is independent of UV crosslinking time, and increases with degradation (*p=0.010 and **p=0.001; two-sample t-test). (C) Burst pressure (mmHg) measured from the total length of the vascular graft increases between 15 and 45 min UV exposure time. (D) Suture retention strength (N) measured from dECM+PPF and porcine left anterior descending artery (LADA). The suture retention strength of the dECM+PPF demonstrates strengths that are comparable to that of porcine LADA, independent of UV crosslinking time. (E) Fold change of the XTT metabolic activity (normalized to 24 h) as a function of time for HUVECs exposed to dECM with and without PPF. The addition of PPF to dECM does not negatively impact the metabolic activity of cells. Data is shown as mean + standard deviation, except for XTT metabolic fold change where data is mean ± standard deviation.

Vascular grafts with a diameter of 1 mm and approximately 1.5 cm long were constructed for implantation as an abdominal aorta interposition conduit into two rats as an in vivo abdominal aortic graft model for two weeks (Figure 6A–C). Upon explantation, the vascular grafts were sectioned and stained with hematoxylin and eosin (H&E), Masson’s trichrome, smooth muscle actin, and CD31 (Figure 6D–G). All four stains demonstrate that the vascular grafts remained patent for 2 weeks. The H&E and Masson’s trichrome stains show the ECM components, pink and blue, respectively, and the cell nuclei, blue and dark brown, respectively, present in the explanted graft. The smooth muscle actin illustrates a presence of smooth muscle cells on the outside and inside of the vascular graft. The CD31 staining confirms that there are endothelial cells on the surface of the lumen.

Figure 6: In vivo evaluation of dECM+PPF as a vascular graft.

(A-B) The dehydrated vascular grafts manufactured for the rat model had an outer diameter of 1 mm and were 1.5 cm long. (C) Successful implantation of the graft in the abdominal aorta of a rat. (D) Hematoxylin and eosin (H&E) stain, (E) Masson’s trichrome stain, and (F) immunohistochemical stain for smooth muscle actin of the vascular graft explanted after 2 weeks; all scale bars: 1000 μm. The graft is still visibly patent after 2 weeks in vivo. Remnants of the PPF structure (black arrows) are still visible. The dECM is also visible in the cross-sections (white arrows), as the mostly collagen layer observed in the different stains. The rolled composite dECM+PPF material retains its original cylindrical shape and remains patent after 2 weeks. (G) Immunohistochemical staining using CD31 as a marker of endothelial cells on the explanted graft; scale bar: 200 μm. Inset: Endothelialization observed along the inner luminal surface of the vascular graft; inset scale bar: 50 μm.

Discussion

Autologous grafts, the gold standard of care in coronary artery bypass grafting surgeries, are extremely limited and are plagued by high failure rates. Grafts of natural materials lack the strength necessary for a vascular graft application, while synthetic grafts struggle to provide the desirable biological response. The overall aim of this study was to develop a biohybrid vascular graft from dECM and PPF. First, we decellularized bovine pericardium and characterized how the decellularization protocol affects the ECM. Next, we identified the processing parameters for PPF and characterized the degradation of dECM+PPF. Last, we examined the ability of dECM+PPF to perform as a vascular graft in vitro and in vivo.

Effects of Decellularization

When developing our decellularization method, it was important not only to remove DNA content, but also reduce the time that the tissue spent in each wash. As compared to previous protocols, we decreased the amount of time of the washes from 6 d to 2 d to limit exposure to reagents that can be disruptive and destructive to the tissue proteins in order to reduce their deleterious effects.^41,49 We are confident that the newly developed decellularization protocol reliably decellularizes native tissue, as evident qualitatively by the DAPI stained images and quantitatively using the PicoGreen assay. Our results confirm that the dECM was well below the 50 ng DNA/mg dry tissue threshold commonly accepted in literature⁴¹, which suggests that we have fully decellularized the pericardium and reduced the possibility of eliciting an adverse immune response. Furthermore, we found that there is tissue source variability in the native DNA content, which we eliminated with the decellularization protocol. The idea of decreasing biological variability, as a result of donor variability, with decellularization has yet to be explicitly investigated and reported. We compared the proteins present in nECM to those in dECM and found that the protein bands present in nECM are still visible in the dECM from all donors. These findings imply that the reduction in wash time helps to preserve the protein composition in the ECM, independent of donor, which is supported by the reported loss of β-actin in longer decellularization processes⁴⁹. The decellularization protocol did not have an impact on the ultimate tensile strength of the material, independent of donor in both nECM and dECM. We believe that the variability between pericardial donors is negligible and should not be considered a factor after decellularization. We have showed that we are able to successfully decellularize bovine pericardium with minimal deleterious effects and that differences among donor post-decellularization are insignificant.

Addition of PPF to Create the dECM+PPF Biohybrid

Considering that dECM provides biological cues necessary for cellular recruitment, we selected PPF as a synthetic component to preserve the properties of dECM against the effects of degradation.⁵⁰ Assessing PPF crosslinking on its own, we investigated the influence of UV exposure time on the elastic modulus and XTT metabolic activity for PPF thin films, driven by the considerable variety of crosslinking methods that have been previously reported for PPF^21,51–54. The significant increase in the elastic modulus of PPF shows that the UV exposure impacts the mechanical properties of PPF. This trend reflects the corresponding increase in mechanical properties with crosslinking seen in literature^55,56 and suggests that we can control the mechanical behavior of our PPF thin films with the UV exposure time. Although the XTT fold change was unaffected by the UV exposure time, it was positively impacted by the ethanol wash. We expected the XTT results to be unchanged at higher UV exposure times⁵¹, but we predicted that the metabolic activity of cells would be impacted by the lower times due to larger amounts of uncrosslinked materials. Based on the elastic modulus and XTT, 45 min of UV crosslinking was selected for further assays. Additionally, the XTT results show that the ethanol washing of PPF after UV crosslinking improves the metabolic activity and thus the long-term cellular viability, likely due to the removal of detrimental materials from the PPF film that may leach into the medium. Without ethanol washes, the PPF lowers the pH of the media, as evident by the color change, which we assume stem from the leaching of uncrosslinked PPF or DEF, but we do not want to speculate without further investigation. Therefore, we incorporated the reverse graduated ethanol washes for rehydrating the biohybrid.

After developing the parameters for the addition of PPF to the dECM, the benefit of this biohybrid was characterized by the percentage of original mass and width, permeability, and mechanical properties. These parameters were assessed in vitro using enzymatic collagenase degradation as a model of tissue degradation during remodeling in vivo. The addition of PPF to dECM reduces the rate at which its mass and width are affected by enzymatic degradation. This result was expected because it is known that PPF degrades slowly through hydrolysis³⁴ and its dimensions remain consistent over time^52,57. The stable width of dECM+PPF indicates there is less swelling of this material than dECM, due to the addition of PPF. Furthermore, dECM displays an increasing amount of variability in its width with time, while dECM+PPF shows reduced variability. dECM+PPF is a more reliable biomaterial that degrades slower and experiences less swelling than dECM and therefore, is better suited for the vascular graft application. The permeability and mechanical properties of the materials were assessed at 4 h in collagenase II because this time point is the first where both dECM+PPF and PPF experience significantly different mass loss from dECM. Additionally, there is no significant change in width at this time point, so the dimensions of the samples were expected to remain constant.

We investigated the permeability of the dECM+PPF to address the concern that the addition of PPF may inhibit permeability and ultimately regeneration. We found that the permeability of the dECM does not experience a significant decrease in value with the addition of PPF, a promising trend because scaffolds with higher permeability have demonstrated more favorable in vivo regeneration.⁵⁸ This effect should preserve the allowance of nutrient transport needed for regeneration throughout the graft.^11,58 Additionally, the permeability of dECM+PPF is not significantly impacted by the collagenase digestion, meaning this property does not rapidly deteriorate with degradation. Furthermore, this phenomenon suggests that the vascular graft should maintain wall integrity to support blood flow without leakage.

In order to characterize the mechanical behavior of these materials in tension, we measured their ultimate tensile strength, maximum strain, E1 modulus, and E2 modulus. The ultimate tensile strength was examined as an indication of overall material strength to ensure that the graft material can withstand arterial pressures and stresses. The significant changes in this value indicate that 4 h in the degradation collagenase solution was ample time to see mechanical changes in the materials. This change in mechanical properties due to the collagenase serve as a model for enzymatic degradation of the tissue components in vivo. Although the ultimate tensile strength of dECM and dECM+PPF decreases, it is sensible that the value for PPF does not decrease. It should be noted that enzymatic degradation does not cause strength loss in PPF, as it has been previously shown that PPF slowly degrades by hydrolysis, losing approximately half of its weight after 500 days.^25,34 Next, we looked at the maximum strain to understand the deformation that occurs before the material fails, a variable of interest for us due to the physiologic elastic deformation of blood vessels. The higher maximum strain in dECM and dECM+PPF suggest that they can deform more before failure than PPF. It was important to preserve the allowable deformation of dECM with the application of PPF to ensure that the dECM+PPF material will be able to withstand the fluctuations in diameter of native vessels.^59,60 The E1 modulus was investigated because this modulus will determine the elastic mechanical behavior of the material at psychologically relevant pressures (<150 mmHg^61,62) for vascular grafts. Human coronary arteries have similar E1 and E2 modulus values, which are not frequently reported as separate values.^63–65 A range of modulus values for human coronary arteries have been reported from around 1 MPa to 4 MPa.^63,64 Higher values between 6 and 12 MPa for the modulus of human saphenous veins have been reported.^15,66 The E1 modulus value of dECM+PPF is within this range and therefore behaves in a similar manner to both coronary arteries and human saphenous veins. The moduli values of dECM+PPF are unaffected by the addition of PPF and the material behaves similarly to dECM, and thus a native vessel, avoiding vascular graft failure associated with synthetic materials.⁶⁷

Assessment of dECM+PPF Biohybrid as a Vascular Graft

The ultimate goal in developing and characterizing the dECM+PPF material was to create a dECM+PPF vascular graft and study its mechanical properties and ability to survive in vivo. We constructed a vascular graft from the dECM+PPF material by UV crosslinking the PPF inside a cylindrical mold. When varying the UV exposure time, the circumferential stress of these grafts increases from 15 to 45 min of UV exposure and is stable between 45 and 60 min, which is similar to the previous trends we have seen for PPF only in Figure 2A. Also, the 15 and 60 min UV exposure grafts experience a significant increase in circumferential strength with collagenase degradation, likely due to the increase in strength exhibited by PPF under the same conditions. These trends suggest that the circumferential strength of the vascular graft is largely a result of the PPF addition. The dECM+PPF vascular graft, regardless of UV exposure time, has circumferential stress values similar to those of porcine RCAs. The maximum circumferential stress in porcine coronary arteries is about 3 times greater than that in human arteries.⁶⁸ Based on our results, the dECM+PPF vascular grafts have the potential to withstand larger circumferential stress values than those of human coronary arteries. Burst pressure increases significantly with UV exposure time from 15 to 45 min. This result correlates to the increase in circumferential strength, and ultimately, the strength of PPF with UV exposure. The burst pressure values for 45 and 60 min of UV exposure, although lower than those reported for human vessels¹⁸, have ample strengths to support physiological blood pressures. Alternatively, suture retention strength is independent of UV exposure time and decreases with collagenase degradation, which based on our previous results, suggests that this type of strength depends largely on the dECM component of the graft. Also, the suture retention values of dECM+PPF are significantly higher than those of porcine LADAs tested here and those of human internal mammary arteries reported in literature¹⁸. The improved suture retention values suggest that the graft is less likely to mechanically fail at the site of anastomosis in vivo. The dECM+PPF graft demonstrates ample strengths to adequately perform as a vascular replacement under normal physiological conditions in vivo. Finally, the effect of the dECM+PPF material on the metabolic activity was investigated to determine if the material is biologically compatible with cells, particularly endothelial cells, since they play a vital role in the patency of vascular grafts.² The addition of PPF to the pericardium with 45 minutes of UV exposure does not negatively impact the metabolic activity of HUVECs. It is important that endothelial cells survive and proliferate on this material in order to develop a healthy endothelial barrier that reduces blood coagulation and increases the chances of patency and vessel regeneration. We concluded that the dECM+PPF vascular graft, with 45 min of UV exposure, demonstrates strengths that are comparable to native vessel function once implanted.

The grafts demonstrated ample strength to survive physiological conditions, so we proceeded with a preliminary in vivo study to assess the ability of the fabricated grafts to be successfully implanted into a vascular defect and to examine the initial tissue response to the implanted grafts. Both rats survived for 2 weeks with the implanted vascular grafts without any indications of vascular complications or adverse effects on their health. After explantation, the vascular grafts were found to be patent. The new tissue layers formed on the outside and inside surfaces may be an indication of future loss of patency. Future in vivo studies are warranted to investigate the long term patency and regenerative properties of the vascular graft, including a better understanding of the immune response and its capacity for remodeling. H&E and Masson’s trichrome stains allowed us to visualize the distribution of ECM components inside the original graft material, as well as the surrounding surfaces. First, we are able to confirm that the dECM+PPF material is still in place and holding its original patent shape because the stains show decellularized ECM (clear pink region in H&E, clear blue region in Masson’s) interlayered with PPF (synthetic material that looks white in the stains). Around this structure, new cellularized tissue had started to grow after 2 weeks, particularly smooth muscle and an endothelial layer. Smooth muscle cells play a vital role in the mechanical functionality of blood vessels.⁶⁹ Endothelialization plays an extremely important role in preserving patency and preventing thrombosis and ultimately fighting graft failure.^11,70 Overall, the deposition of new ECM, and the presence smooth muscle cells and endothelial cells are a promising indication of vascular graft integration with the adjacent vascular system.

Conclusions

The goal of this work was to construct a vascular graft composed of decellularized pericardium (dECM) and poly(propylene fumarate) (PPF). We developed a decellularization protocol for bovine pericardium that allows us to remove cellular content without causing deleterious effects on the protein content or mechanical strength, while also eliminating variability introduced by donors. Next, we studied PPF to identify the proper processing parameters to create the dECM+PPF biohybrid, which demonstrates mechanical behavior similar to that of native coronary arteries and saphenous veins, an important property for a vascular graft, an application dominated by mechanical functionality. Furthermore, preliminary results showed promise due to the patent grafts with regeneration and re-endothelialization at two weeks. In this study, we have demonstrated the potential for success of dECM+PPF as a vascular graft.

Statement of Significance.

In creating a dECM+PPF biohybrid graft, we have observed phenomena that will have a lasting impact within the scientific community. First, we found that we can reduce donor variability through decellularization, a unique use of the decellularization process. Additionally, we coupled a natural material with a synthetic polymer to capitalize on the benefits of each: the cues provided to cells and the ability to easily tune material properties, respectively. This principle can be applied to other materials in a variety of applications. Finally, we created an off-the-shelf alternative to autologous grafts with a newly developed material that has yet to be utilized in any scaffolds. Furthermore, bovine pericardium has not been investigated as a small diameter vascular graft.