Small-Diameter Artery Decellularization: Effects of Anionic Detergent Concentration and Treatment Duration on Porcine Internal Thoracic Arteries

Abstract

Engineered replacement materials have tremendous potential for vascular applications where over 400,000 damaged and diseased blood vessels are replaced annually in the United States alone. Unlike large diameter blood vessels, which are effectively replaced by synthetic materials, prosthetic small-diameter vessels are prone to early failure, restenosis, and reintervention surgery. We investigated the differential response of varying 0–6% sodium dodecyl sulfate (SDS) and sodium deoxycholate (SDC) anionic detergent concentrations after 24 and 72 hours in the presence of DNase using biochemical, histological, and biaxial mechanical analyses to optimize the decellularization process for xenogeneic vascular tissue sources, specifically the porcine internal thoracic artery (ITA). Detergent concentrations greater than 1% were successful at removing cytoplasmic and cell surface proteins but not DNA content after 24 hours. A progressive increase in porosity and decrease in glycosaminoglycan (GAG) content was observed with detergent concentration. Augmented porosity was likely due to the removal of both cells and GAGs and could influence recellularization strategies. The treatment duration on the other hand, significantly improved decellularization by reducing DNA content to trace amounts after 72 hours. Prolonged treatment times reduced laminin content and influenced the vessel’s mechanical behavior in terms of altered circumferential stress and stretch while further increasing porosity. Collectively, DNase with 1% detergent for 72 hours provided an effective and efficient decellularization strategy to be employed in the preparation of porcine ITAs as bypass graft scaffolding materials with minor biomechanical and histological penalties.

Introduction

A major challenge confronting the field of regenerative medicine and tissue engineering is the design, fabrication, and validation of suitable scaffolds that direct the repair and regeneration of damaged tissues.^1–3 Advanced biomaterials including hydrogels, membranes, micro/nanofibers, and micro/nanoparticles have emerged as innovative platforms for tissue engineering purposes.^4–6 New generation biomaterials and novel technologies are making substantial headway into the engineering of complex tissue scaffolds. This is in part due to the ability to modulate the composition and microarchitecture of these materials to improve biocompatibility, promote desirable cellular events, and optimize degradation kinetics and mechanical properties. Advances in biomaterials are making the engineering of replacement tissues a reality; however, several hurdles remain before this technology reaches its full potential. These hurdles are particularly evident in the application of tissue-engineered blood vessels as a platform for coronary and peripheral bypass grafting procedures. With approximately 400,000 annual bypass surgeries annually in the United States alone, there is a pressing need to optimize replacement blood vessels for grafting procedures.^7,8

Autografts are the preferred choice in small-diameter vascular bypass procedures due to their biocompatibility and lack of thrombogenicity.⁹ However, they are inherently limited by the number of bypasses and any chronic pathological conditions that impact tissue patency such as atherosclerosis or diabetes. Prosthetic grafts or patches (e.g., Dacron or expanded PTFE) have been implanted in large numbers of patients over the past several decades and these are relatively effective for the repair of large arteries such as the aorta. However, these prosthetic grafts are limited for use in bypass procedures because an increased rate of infection and thrombotic events have been detected when used as small-diameter blood vessel replacements.^10,11 Decellularized blood vessels have gained popularity as a biomimetic scaffold by harnessing innate tissue-specific properties to direct cellular functions including differentiation and proliferation while inducing a minimal immune response.^12–14 Such scaffolds can be created from human or animal tissues and are a low-cost and readily available platform for tissue engineering. In their native form, decellularized scaffolds are mechanically-robust but extracellular matrix materials can also be solubilized and reconstituted into infinite size/shape configurations and tailored to the functional demands of the specific tissue.^15–18

Effective decellularization of blood vessels can be achieved through physical, chemical, or enzymatic processes, however, each has unique limitations and disadvantages such as the presence of residual chemicals or enzymes and the destruction of essential extracellular matrix (ECM) proteins.^19–21 Chemical decellularization, for example, has been successfully achieved through the use of anionic detergents such as sodium dodecyl sulfate (SDS) and sodium deoxycholate (SDC). These homogeneously disrupting surfactants solubilize cell membranes by inducing curvature stress and thinning the hydrophobic core of the membrane.²² Additional decellularization methodologies have combined different detergents or other chemical agents with enzymes to effectively target both the cell membrane and nuclear material.^13,23 Enzymes including proteases and nucleases are advantageous in decellularization protocols because of their substrate specificities; however, exposure to some enzymes at high concentrations or for prolonged periods can damage collagen or significantly reduce the glycosaminoglycan (GAG) content.^23–25 These materials play a vital role in the mechanical properties of blood vessels which are deterministic in graft procedure outcomes.²⁶ Likewise, the analysis of comprehensive biaxial properties (i.e., inflation-extension) from a consistent material and anatomical location is fundamental to the efficacy of this biomaterial²⁷ and deviations in host-graft properties lead to atherosclerotic and inflammatory promoting perturbed flow profiles, shear stresses, and stress concentrations at the anastomosis site.²⁸

Many tissue-engineered decellularization strategies that fail to preserve the biochemical and biomechanical properties of native tissues possess poor cell repopulation and differentiation rates.²⁹ Assuming that the optimal decellularized vascular graft retains properties similar to its native configuration but in the absence of immunogenic materials, the present studies were aimed at systematically evaluating the effects of anionic detergent concentrations and treatment durations on decellularization potential. Using the porcine internal thoracic artery (ITA) as a novel but well-characterized match for a graft target in the coronary circulation, the biochemical, histological, and mechanical properties of these tissues following enzymatic-detergent decellularization was analyzed to determine if there were meaningful modifications of these decellularized scaffolds. Our findings contribute to the broader field of tissue engineering and xenograft development for applications in vascular bypass grafting.

Materials & Methods

Arterial decellularization

Adult porcine ITAs were obtained from a local abattoir and transported to the laboratory in phosphate-buffered saline containing antibiotics (100 units/mL penicillin, 100 mg/mL streptomycin, 1 mg/mL amphotericin-B, and 10 ng/mL gentamicin). Due to variations in the structural and mechanical properties along the length of the porcine ITA ²⁷, vessels were separated into proximal and distal halves with only the proximal half used in this study. Vessels were cut into 3 cm lengths, placed into 50-mL conical tubes, and rinsed in three changes of PBS for 10 minutes each on a rotator at room temperature (20 revolutions per minute) to remove any residual blood. Vessel pieces were then incubated overnight in PBS (control group) or distilled water on a rotator at 4°C (the control sample was incubated in PBS for the duration of experimental treatments).²¹ Following overnight incubation, the samples were rinsed in PBS twice for 20–30 minutes each. The samples were incubated again overnight in DNase 1 (Roche Diagnostics cat# 10104159001) at a final concentration of 1 mg/mL in DNase digestion buffer (10 mM Tris (pH 7.4), 2.5 mM MgCl₂, 0.5 mM CaCl₂) on a rotator at 4°C. Samples were subsequently rinsed twice for 20–30 minutes in PBS. Rinsed samples were placed in a detergent solution composed of equal concentrations of SDS and SDC.³⁰ Detergent concentrations included: 0% SDS and SDC (hereafter referred to as 0% detergent), 0.5% SDS plus 0.5% SDC (hereafter referred to as 1% detergent), 1% SDS plus 1% SDC (hereafter referred to as 2% detergent), 1.5% SDS plus 1.5% SDC (hereafter referred to as 3% detergent) and 3% SDS plus 3% SDC (hereafter referred to as 6% detergent). Samples were incubated in the respective detergent concentrations for 24 or 72 hours at room temperature on a conical tube rotator. All solutions for decellularization contained antibiotics to mitigate contamination. Although the nuclease-detergent sequence used in our study differed from previous reports,^31–33 additional experiments were performed in our lab to test DNase application before or after detergents or when omitted from the protocol (n=8 to 12 per group).^21,30 Our findings (Supplemental Section) revealed similar residual DNA concentrations irrespective of enzyme-detergent order but were significantly less than when DNase was omitted from the protocol altogether.

DNA Quantification

Control and decellularized vessels were homogenized in DNAzol Reagent (Thermo Fisher Scientific) and the genomic DNA was precipitated from the lysate with ethanol. Then the DNA was solubilized in 8 mM sodium hydroxide. Quantitative measurements of the total DNA content were then determined by the Quant-IT PicoGreen dsDNA Assay Kit (Invitrogen), following the manufacturer’s specifications.

Western blot analyses

Biochemical analyses were carried out to assess the relative loss of cytoplasmic (α-smooth muscle actin), cell surface (β1 integrin), and basement membrane (laminin) components during the decellularization protocols. Samples were pulverized in liquid nitrogen and incubated in RIPA solution (150 mM sodium chloride, 1% Triton X 100, 0.5% deoxycholate, 0.1% sodium dodecyl sulfate, 1.5 mM ethylenediaminetetraacetic acid, 50 mM Tris, pH 8.0) containing Pierce Protease Inhibitor Mini-Tablets (Thermo Fisher Scientific). Samples were incubated at 60°C for 10 minutes and were mixed by inversion three times during incubation. Samples were then centrifuged for 20 minutes at 20,000 g (4°C). Supernatants were then moved to new tubes and total protein concentration was determined with the Pierce BCA (bicinchoninic acid) protein assay (Thermo Fisher Scientific).

Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose. Nitrocellulose was rinsed in tris-buffered saline containing 0.05% TWEEN 20 (TBS-T) and blocked in TBS-T containing 5% powdered milk. Nitrocellulose was rinsed in TBS-T and incubated overnight with validated primary polyclonal antibodies against α-smooth muscle actin (Abcam, #ab5694), laminin (Abcam, #ab11575) and β1 integrin (Sigma-Aldrich, #ab1952). The α-smooth muscle actin and laminin primary antibodies were used at a dilution of 1:500 and β1 integrin at a dilution of 1:1,000 in TBS-T containing 1% powdered milk. After three washes in TBS-T, nitrocellulose was incubated in 1:10,000 HRP-conjugated anti-rabbit IgG (Sigma-Aldrich) secondary antibody for 1–2 hours. Following additional rinses in TBS-T, immunoblots were developed with the Pierce SuperSignal Western blot detection reagent (Thermo Fisher Scientific) and exposed to x-ray film. Protein standards were used to provide a reliable molecular weight estimation of the protein signals on the transferred blots. The films were scanned using Adobe Photoshop and signals were quantified using Alpha Innotech software.

Scanning electron microscopy

Samples were obtained by cutting approximately 5 mm cross-sectional segments from each vessel/scaffold. Samples were fixed in McDowell Trump’s fixative for 2 h followed by rinsing in PBS (3 × 15 minutes) and ultrapure water (3 × 10 minutes). Samples were dehydrated via a graded ethanol series as follows: 15 minutes in 35%, 50%, and 75% ethanol; 2 × 15 min in 95% ethanol; and 3 × 30 minutes in 100% ethanol. Samples were then immersed in fresh hexamethyldisilazane (HMDS) twice for 15 minutes. HMDS was decanted, and samples were left to dry overnight. Samples were then mounted to allow imaging of the vessel wall in cross-section, gold sputter-coated, and imaged on a JEOL JSM-1610PLUS/LA SEM. At least two representative regions of each sample were imaged. The effects of experimental treatment on the generation of spaces within the samples were analyzed from SEM images using the NIH ImageJ program.

Histology and immunohistochemistry

Approximately 2 mL cross-sectional slices of the treated vessels were fixed overnight at 4°C in 4% paraformaldehyde prepared in PBS. Fixed vessels were rinsed in PBS, processed for paraffin embedding, and sectioned at approximately 5 μm. Tissue sections were stained with 4’,6-diamidino-2-phenylindole (DAPI) to analyze DNA content, Picrosirius Red (PSR) to assay collagen, and Verhoeff-Van Gieson (VVG) stain to assess elastic fibers. For quantitative analyses, eight random photomicrographs were taken circumferentially around the blood vessel wall. The effects of decellularization treatment parameters were quantified using the NIH ImageJ program.

DAPI staining

The fluorescent stain, DAPI, was used to evaluate the DNA content of control and decellularized tissue sections. The histological sections were heated in a hybridization oven at 60°C for 10 minutes, deparaffinized in xylene, and rehydrated through a descending alcohol series (100% 95%, 70%). Tissue sections were rinsed twice for 10 minutes each in PBS then incubated in DAPI (1:1000 dilution) in the dark for 30 minutes at room temperature. The sections were subsequently rinsed twice in PBS for 20 minutes each. For quantitative analysis, eight random photomicrographs were taken circumferentially around the vessel wall on a Nikon E600 fluorescence microscope at an exposure time of 2.5 ms. The effects of the decellularization treatment parameters on DAPI-positive content were quantified using the ImageJ program from the NIH.

Picrosirius Red staining

Picrosirius Red (PSR) stain was used to evaluate vascular collagen content of control and decellularized tissue sections. Histological sections were heated to 60°C in an incubator for 45 minutes and deparaffinized using xylenes, a descending alcohol series (100%, 90%, 70%), and a brief 1-minute bath in distilled water. The sections were stained with 0.2% Phosphomolybdic Acid and then rinsed in distilled water. The sections were then stained with PSR (12 g Picric Acid, 400 mL water, 0.4 g Sirius Red) followed by incubation in 0.1 M HCL. Ascending alcohol (70%, 90%, 100%) and xylene series were used to dehydrate the sections. The sections were mounted with Depex. The stained sections were imaged and the collagen volume was determined as previously described.³⁴

Verhoeff-Van Gieson staining

Verhoeff-Van Gieson stain was used to evaluate elastic fiber content of control and decellularized tissue sections. Histological sections were heated at 60°C in a hybridization oven for 30 minutes and then deparaffinized using xylenes and a descending alcohol series (100%, 95%, and 70%). The sections were rinsed in distilled water twice, followed by incubation with Working Elastin Stain (10 mL Hematoxylin Solution, 1.5 mL Ferric Chloride Solution, 4 mL Weigert’s Iodine Solution, 2.5 mL deionized water) for 10 minutes. Following incubation, the sections were rinsed in distilled water and differentiated in Working Ferric Chloride Solution (3 mL Ferric Chloride Solution and 37 mL distilled water) for 1–2 minutes. The sections were sequentially rinsed briefly in tap water, 95% alcohol to remove the iodine, and deionized water. After this rinse, the sections were stained with Van Gieson for 1–3 minutes. They were again rinsed in 95% alcohol, dehydrated in xylene, and coverslipped. The stained sections were imaged, and the vascular elastic fiber density was determined using ImageJ.

Biochemical quantification of glycosaminoglycan content

A Dimethylmethylene Blue (DMMB) assay was used to evaluate the sulfated-glycosaminoglycan (s-GAG) content within treated ITA samples. The tissue samples were dried in glass tubes at 65°C overnight in a hybridization oven and the dried tissues were weighed. Tissues were digested with papain (Invitrogen cat# 10108014001) made in papain extraction buffer (400 mg Sodium Acetate, 200 mg EDTA, 40 mg Cysteine HCl – added to 50 mL 0.2 M NaH₂PO₄) overnight at 65°C. The liquid was transferred to Eppendorf tubes and centrifuged for 10 minutes at 10,000 g (4°C). Aliquots of each sample were added to DMMB Reagent (16 mg DMMB, 3.04 g Glycine, 1.6 g Sodium Chloride, 95 mL 0.1 M Acetic acid – for 1L). s-GAG concentrations were determined by comparison to a standard curve using a BioRad Benchmark Plus Microplate Spectrophotometer at 525 nm.

Collagen quantification

A hydroxyproline assay was used to biochemically evaluate the collagen content within the treated ITA samples. Small pieces of tissues were dried overnight at 60°C and subsequently, dry weights were obtained. Tissues were digested overnight at 60°C with papain (Invitrogen cat# 10108014001 30 units/mg). Samples were subsequently incubated at 120°C in 4N NaOH for 15 minutes. Samples were returned to room temperature and 4N HCl was added to each sample to neutralize the pH. Chloramine-T solution (0.05 M Chloramine-T, 0.629 M NaOH, 0.140 M Citric Acid, 0.453 Sodium Acetate, 0.112 M Acetic Acid – in 74% water/26% 2-Propanol) was added to each sample and the sample was incubated at room temperature for 20 minutes. Ehrlich’s Reagent (1 M DMAB in 30% HCl/70% 2-Propanol) was added and the samples were incubated at 65°C for 20 minutes. The reaction was quenched by immersing the tubes in cool water and samples read on a BioRad Benchmark Plus Microplate Spectrophotometer at 550 nm. Hydroxyproline concentration was determined by comparison to a standard curve created from purified collagen.

Biaxial Mechanical Testing

The mechanical properties of the fresh, control, and decellularized ITAs were found through biaxial inflation-extension testing using a Bose BioDynamic mechanical testing device following protocols described in.²⁷ The tissue sections were trimmed to comply with the testing device’s 10 mm maximum displacement range, mounted onto two luer-fittings, and then fixed in place with 3–0 braided sutures. Then tissues were submerged in a testing bath and perfused with 1% PBS and sodium nitroprusside (10⁻⁵ M) to elicit a fully passivated state.³⁵ The “in vivo axial stretch ratio” was found by axial stretching the vessel to a point that yielded a constant axial force value in response to pressurization. Every decellularized tissue sample then underwent five cycles of extension and inflation preconditioning to minimize viscous dissipation and ensuring reproducible results. For data collection, all vessels were inflated from 0 to 200 mmHg while measurements of force, axial stretch, and outer diameter were collected every 20 mmHg at the in vivo axial stretch ratio as well as 10% above and below that value.

Immediately following biaxial testing, a 1 mm thick ring segment was cut from the middle of the tested vessel and a cross-sectional image was captured. The residual strain in the vessel was removed by making a radial cut and allowing the tissue to equilibrate for 30 minutes in 1% PBS. An image of this zero-stress state was captured using a Canon EOS 60d and ImageJ image analysis software was used to measure the geometries of both images. This configuration enables calculation of wall thickness H and opening angle Φ so that,

$$H=\frac{2A}{L_i+L_o}\text{and}\Phi =\pi -\frac{L_o-L_i}{2H}$$ (1–2)

where L_i and L_o are the inner and outer arc lengths of the open ring segment, respectively. The cross-sectional area A was found by calculating the difference between the area within the outer wall and the luminal area of the intact ring segment before introducing the stress-relieving cut. The biaxial inflation-extension apparatus controlled the luminal pressure P and axial displacement while measuring the outer diameter and axial force F at 20 mmHg pressure increments for three axial displacements. Under the assumption of tissue incompressibility, the inner radius r_i could be found at any deformed configuration given that

$$r_i=\sqrt{{r_o}^2–\frac{A}{\pi {\lambda }_z}}$$ (3)

where r_o is the outer radii and λ_z the axial stretch ratio. The axial and circumferential λ_θ stretch ratios are then easily calculated via

$${\lambda }_z=\frac{l}{L}\text{and}{\lambda }_{\theta }=\frac{2\pi \left(r_i+r_o\right)}{L_i+L_o}$$ (4–5)

where l and L are the suture-to-suture deformed and undeformed lengths, respectively. The average circumferential σ_θ and axial σ_z stresses were (see also Prim et al. 2018):

$${\sigma }_{\theta }=\frac{P{r}_i}{r_o–r_i}\text{and}{\sigma }_z=\frac{F}{\pi \left(r_o^2-r_i^2\right)}$$ (6–7)

where P, is the luminal pressure and F the axial force. Finally, the area compliance was represented by

$$C=\pi \frac{\Delta r_i^2}{\Delta P}$$ (8)

where Δr_i and ΔP were measured around a 100 mmHg operating point

Statistics

Comparisons of the histological images, SEM images, and the mechanical properties between the anionic detergent concentrations and controls were made via one-way ANOVA. Data were imported into Prism GraphPad for statistical analyses. Statistically significant differences were taken at a level of p<0.05.

Results

Decellularization efficiency

The effects of anionic detergent (SDS and SDC) concentration and treatment duration on decellularization efficiency were evaluated by staining of tissue sections with DAPI to assess DNA content (Figure 1). Treatment of porcine ITAs with DNase 1 followed by anionic detergents for 24 hours resulted in substantial residual DAPI-positive material in the scaffolds that were not statistically different from untreated controls (Figure 1A). Unlike control samples, however, the DAPI-positive material was no longer in compact nuclei form but was more diffusely organized. Anionic detergent treatment for 72 hours resulted in essentially undetectable DAPI staining (Figure 1B) with no significant difference between the concentrations of anionic detergent. The PicoGreen assay also failed to reveal significant differences in DNA concentration for any groups after only 24 hours (Figure 1E). The 72 hour treatment, however, resulted in a significant decrease in the measured DNA concentration between all treated and untreated controls (Figure 1F) with no significant differences between the detergent concentrations. To further evaluate decellularization efficiency and to examine differential effects on specific tissue components (i.e., cytoplasmic, cell surface, and extracellular matrix), western blot analyses were carried out of tissue lysates following the treatment process (Figure 2). For these experiments, western blots were performed of representative cytoplasmic (α-smooth muscle actin), cell surface (β1 integrin), and basement membrane (laminin) proteins. Quantitative assessment was not performed due to a lack of normalization control. Despite this limitation, it was clear that the inclusion of high concentrations of an anionic detergent (2–6%) effectively removed cytoplasmic and cell surface proteins after 24 and 72 hours of treatment. After 24 hours of detergent treatment, a high concentration of laminin was still detected in the tissue lysates, but this diminished substantially following 72 hours of treatment.

Figure 1.

DAPI nuclear stain and quantification of DNA content following (A) 24 hours and (B) 72 hours of anionic detergent treatment. The insets are representative images of DAPI-stained sections of control, 6% detergent-treated, and 1% detergent-treated tissues at respective time points. Area fraction quantification of DAPI positive pixels for (C) 24 hours and (D) 72 hours. Quantified DNA concentrations from the PicoGreen assays for (E) 24 hours and (F) 72 hours. Statistical significance between detergent concentrations relative to untreated controls was determined by one-way ANOVA and is indicated by (*) at p<0.05. Mean ± SD, n=4 for each group.

Figure 2.

Western blot analysis of representative cytoplasmic (α-smooth muscle actin), cell surface (β1 integrin), and basement membrane (laminin) proteins. Representative images of western blots illustrating the effects of anionic detergent concentration and treatment duration. Lanes 1 – 6 are untreated control, 0% detergent, 1% detergent, 2% detergent, 3% detergent and 6% detergent, respectively.

Analysis of scaffold structure

Scanning electron microscopy was performed to analyze the effects of anionic detergent concentration and treatment duration on the structure of resulting scaffolds (Figure 3). Treatment of porcine ITAs resulted in the formation of void space or pores between extracellular matrix fibers in the tunica media. At 24 and 72 hours, the area fraction of void space within the tissue (porosity) was significantly greater in the samples treated with detergent compared to the non-detergent treated samples (Figures 3C and 3D). For the 24 hour detergent treatment, the highest doses of detergent (3 and 6%) created a significantly greater porosity when compared to the lower doses (1 and 2%). This difference was no longer apparent at 72 hours of treatment and all detergent doses were statistically the same.

Figure 3.

Representative scanning electron microscopic images of the (A,C) 72 hour control sample at low and high power (B,D) 1% detergent for 72 hours at low and high power. Quantification of the total porosity in tissues after (E) 24 hours and (F) 72 hours for all anionic detergents was performed on high-powered images. Statistical significance between detergent concentrations relative to untreated controls was determined by one-way ANOVA and is indicated by (*) at p<0.05. Mean ± SD, n=3 for each group.

Analysis of extracellular matrix content

To assess the effects of anionic detergent concentration and treatment duration on the vascular ECM, various assays qualitatively measured the collagen, elastin, and s-GAG content. Overall collagen content was inferred by analyzing the hydroxyproline content for all experimental groups, while collagen distribution and area fraction were evaluated histologically with picrosirius red staining. There was substantial heterogeneity in the effects of detergent treatment on collagen content with both assays (Figure 4). Due in part to the level of variability, there was no significant effect of detergent concentration and treatment duration on the overall collagen content (Figure 4). Through quantitative analysis of the VVG staining, it is evident that there was no significant effect of detergent concentration nor treatment duration on elastic fiber content (Figure 5). Interestingly, there was a trend towards increased elastic fiber density following treatment for 24 or 72 hours with higher concentrations of detergent suggesting compaction of elastic fibers due to the removal of cells or other components. The most notable alteration to the vascular ECM was the significant reduction in s-GAG concentration with all decellularization treatments compared to the control samples (Figure 6). Reduction in s-GAG concentration was also seen in samples only treated with DNase 1 (0% detergent).

Figure 4.

Quantification of hydroxyproline concentrations after (A) 24 hours and (B) 72 hours of anionic detergent treatment. Collagen percent volume quantified by image thresholding of picrosirius red-stained tissue sections after treatment of (C) 24 hours and (D) 72 hours. No statistically significant differences were found following ANOVA analysis. Mean ± SD, n=5 for each group.

Figure 5.

Quantitative analysis of elastic fibers following anionic detergent treatment. Representative microscopic images of control samples, 6% detergent, and 1% detergent for (A) 24 hours and (B) 72 hours, respectively. Elastic fiber area fraction was quantified from Verhoeff-Van Gieson-stained tissue following anionic detergent treatment for (C) 24 hours and (D) 72 hours. No statistically significant differences were found following ANOVA analysis. Mean ± SD, n=5 for each group.

Figure 6.

Analysis of dimethylmethylene blue (DMMB) concentrations of all anionic detergent concentrations following (A) 24 and (B) 72 hours of treatment. DMMB concentration was indicative of the glycosaminoglycan concentration present in all tissue samples. Statistical significance between detergent concentrations relative to untreated controls was determined by one-way ANOVA and is indicated by (*) at p<0.05. Mean ± SD, n=5 for each group.

Analysis of scaffold biomechanical properties

Inflation and extension biaxial tests were performed to quantify the physical characteristics of the scaffolds and used to determine how anionic detergent concentrations impacted the vascular wall mechanical properties. All specimens were tested at common axial stretches to facilitate comparisons between groups. Overall, the control and fresh vessels exhibited similar mechanical properties and the presence of detergents impacted the slope of the stress-strain curve (stiffness) but in a concentration-independent manner (Figure 7). That is, the presence of anionic detergents had some effect, but no obvious trends emerged between the varying concentrations on mechanical properties. When the anionic detergent treatment duration was increased to 72 hours the tissue appeared to exhibit a more compliant stress-strain profile. Comparing the structural and mechanical properties at common loading conditions revealed a significant increase in circumferential stress and circumferential stretch for tissues treated with 1% detergent for 72 hours relative to the control group (Figure 8). Similarly, short treatment durations of high detergent concentrations (6% detergent) exhibited a significantly higher axial stress value in comparison to the control group. Conversely, a significant decrease in the area compliance was observed between tissues treated with 6% detergent and the DNase enzyme only (0% Det) treated sample. The area compliance value of the ITA sample treated with 1% detergent for 72 hours was elevated, albeit insignificantly, above all other samples (Figure 8). Data taken from cross-sectional images and stress-free ring sector images reported a statistically significant difference between the wall thicknesses of the 6% detergent sample and the fresh tissue that was biomechanically tested immediately after harvest (Figure 9).

Figure 7.

Biaxial mechanical data for all decellularization groups and fresh porcine ITA tissue. (A) Pressure-inner diameter, (B) axial force-pressure, (C) circumferential stress-stretch all plotted at λ_z = 1.45. (D) Axial stress-stretch at 100 mmHg. Mean ± SD, n=4 for each group.

Figure 8.

Biaxial mechanical data of decellularized porcine ITAs plotted at common loading conditions of 100 mmHg and λ_z = 1.45. (A) Inner radius, (B) thickness, (C) circumferential stress, (D) axial stress, (E) circumferential stretch, and (F) area compliance. Statistical significance between different decellularized treatments is indicated by (*) at p<0.05. Statistical significance between a decellularized ITA group and the control or fresh tissue is indicated by (#) and (**), respectively. Mean ± SD, n=4 for each group.

Figure 9.

Unloaded conditions of decellularized and fresh porcine ITAs. (A) Average unloaded wall thickness and (B) opening angle. Statistical significance between a decellularized ITA group and the fresh tissue group is indicated by (**) at p<0.05. Mean ± SD, n=4 for each group.

Discussion

Decellularized vascular xenografts present a promising platform for small-diameter tissue-engineered blood vessel replacements.^36,37 A variety of physical, biological, and chemical methods have been used in the past to decellularize diverse tissues and organs.^20,21,38,39 Yet each of these methods, although largely effective at the removal of cellular material, can negatively impact the ECM microarchitecture limiting the tissue’s inherent biophysical qualities. Starting with a novel but well-characterized match for the coronary vasculature ²⁷ we chose the porcine ITA as a decellularization starting point and assessed the effects of varying DNase and anionic (SDS and SDC) detergents in a time-dependent manner. Our immersion-based chemical decellularization procedure retained tissue in an unloaded configuration to help minimize loading effects on the existing microarchitecture. Our findings suggest that DNase and 1% anionic detergent for 72 hours is the most effective strategy at removing cellular material while minimizing the consequences to the vessel’s histoarchitecture.

The ITA composition has been characterized as an elastic and musculoelastic artery, depending on the relative distance from the heart.^27,40,41 Heterogeneous distribution of elastin and collagen fibers influences the vascular wall mechanics and highlights the importance of donor tissue selectivity.^9,27 Histomechanical similarities exist between the ITA and the left anterior descending artery (LAD), with both arteries exhibiting similar compliance and a higher amount of elastin than other common autograft tissue sources.²⁶ This histomechanical graft-target matching plays a vital role in long-term patency. Our investigation of decellularized ITAs revealed similar histological, mechanical, and morphological properties to their native state. Xeno-bioprosthetics are favorable as a scaffold for cell repopulation due to the consistency of microstructural protein networks across mammalian species.^17,42 Conservation of key biomechanical properties such as mean circumferential stress, mid-wall circumferential stretch, and mean axial stress across several species further supports the use of xenogeneic tissue sources as future grafting scaffolds.⁴²

Others have shown that low concentrations of anionic detergents can effectively remove cellular material from vascular tissue with minimal residual detergent retention.^20,21 Analysis of the cellular content in our vessels revealed that the porcine ITAs retained DAPI-positive material even after 24 hours of treatment with high levels of anionic detergent treatment. This was undetectable when the treatment duration was extended to 72 hours. The quantified residual DNA concentrations using PicoGreen for both treatment durations align well with the DAPI image thresholding. This result supports our findings that nuclease and low anionic detergent concentration treatment for 72 hours was preferred over 24 hours for removing significant nuclear material. An acknowledged limitation of the current study is that we did not optimize the concentrations of SDS and SDC detergents independently. Moreover, our nuclease-detergent sequence was different than other studies;^31–33 however, DNase is often omitted entirely from decellularization protocols.^21,30 Still, qualitative analysis of cytoplasmic and surface proteins confirmed that 24 hours of detergent treatment removed a large amount of cellular material while 72 hours showed a significant decrease of the cellularity of all detergent groups in a concentration-independent manner. However, extending the treatment duration to 72 hours reduced the basement membrane protein laminin in the scaffolds. This reduction in laminin may have adverse effects on the recellularization of scaffolds as vascular cells readily adhere to this protein and coating decellularized prostheses with laminin enhances recellularization.⁴³ Although we did not test it directly, another risk of increasing detergent concentration and time is the potential for cytotoxic detergent retention, which would limit future recellularization efficacy.

Although we performed a complete mechanical characterization via state-of-the-art biaxial testing protocols using matched tissue segments, very few of the metrics used for mechanical comparison reached statistical significance. Others have shown that decellularized arteries are significantly stiffer than native tissue through uniaxial tensile testing, but these studies do not assess how decellularization impacts the multidirectional loading of these arteries.⁴⁴ A decrease in the cellularity of these engineered tissue constructs is accompanied by a decrease in cell-ECM fiber interactions. It is noteworthy to point out that longer decellularization treatment durations resulted in more compliant stress-strain profiles relative to the control group. Others have observed that extended physical decellularization procedures can significantly reduce the maximum tensile strength and Young’s modulus of small-diameter vascular tissue via uniaxial ring extension.⁴⁵ We believe the compliant response seen with prolonged detergent-based decellularization procedures (>24 hours) may be due in part to greater disruption to the ECM fiber network, decreased fiber-fiber interactions, or collagen denaturation and structural changes.^20,44,46

Analysis of the constituent makeup of the decellularized ITAs did not reveal significant differences in the two major load-bearing proteins, namely collagen and elastin, following anionic detergent and DNase treatments and explain the preservation of relevant mechanical metrics. However, there was a significant decrease in cellular material and s-GAG content and an inversely proportional increase in tissue porosity with detergent and DNase treatments that are consistent with other research groups.^23,24 Retention of these ECM proteins throughout the decellularization process is crucial for maintaining tissue integrity and providing attachment points for recellularization. Although we did not test differentially for collagen type IV compared to other types, the persistence of collagen overall and the retention of laminin at short treatment durations suggests a high potential for recellularization. We hypothesize that the recellularization process would be enhanced by the increased porosity of the tissues during culture as cells and nutrients would diffuse more readily throughout the matrix; however, the increase in porosity also presents swelling concerns and a potential site for residual detergents.^19–21 Likewise, the DNase enzyme-detergent sequence used in our study could influence the retention of residual detergents.⁴⁷ Never-the-less, it is also worth noting that the removal of constituents resulting in increased porosity has a moderate effect on the wall thickness and therefore limits the utility of the incompressibility assumption employed during biaxial testing. The porosity-compressibility relationship, therefore, needs to be further disambiguated in future studies.

Decellularized xenografts have shown potential as coronary and peripheral bypass candidates that could help alleviate donor tissue limitations. This novel tissue source was chosen based on widespread availability, relative size in comparison to human vascular tissue, and because other porcine vessels (e.g., carotids) are popular amongst xenograft studies. More importantly, the native histomechanical features closely match those of the coronary vasculature thus our optimization processes are aimed at retention of these features. In our study, short-term (24 hours) detergent treatments in conjunction with DNase failed to remove sufficient nuclear material, however, prolonged (72 hours) resulted in adequate decellularization even when using low concentrations of anionic detergents (1% detergent). Collectively these findings suggest that low detergent concentrations for 72 hours can be used in future decellularization strategies but at the cost of altered mechanical properties and laminin retention.

Conclusion

The combination of enzymatic and anionic detergent decellularization of porcine ITAs revealed the removal of cellular content proceeded in a time-dependent and concentration-independent manner. Significant differences in tissue composition and structure were found through a combination of qualitative and quantitative analyses of histology, electron microscopy, and biaxial mechanical testing. These differences can be used to optimize the decellularization process of xenogenic vascular tissue and may contribute to the development of an acellular scaffold used to investigate cellular repopulation and tissue remodeling.

Data Availability Statement

All data that support the findings of this study are presented in the article and are available from the corresponding author upon reasonable request.