|Home | About | Journals | Submit | Contact Us | Français|
The potential for decellularized aortic heart valves (AVs) as heart valve replacements is based on the assumption that the major cellular immunogenic components have been removed, and that the remaining extracellular matrix (ECM) should retain the necessary mechanical properties and functional design. However, decellularization processes likely alter the ECM mechanical and structural properties, potentially affecting long term durability. In the present study we explored the effects of an anionic detergent (SDS), enzymatic agent (Trypsin), and a non-ionic detergent (Triton X-100) on the mechanical and structural properties of AV leaflets (AVLs) to provide greater insight into the initial functional state of the decellularized AVL. The overall extensibility represented by the areal strain under 60 N/m increased from 68.85% for the native AV to 139.95%, 137.51%, and 177.69% for SDS, Trypsin, and Triton X-100, respectively, after decellularization. In flexure, decellularized AVLs demonstrated a profound loss of stiffness overall, and also produced a nonlinear moment-curvature relation compared to the linear response of the native AVL. Effective flexural moduli decreased from 156.0±24.6 kPa for the native AV to 23.5±5.8 kPa, 15.6±4.8 kPa, and 19.4±8.9 kPa for SDS, Trypsin, and Triton X-100 treated leaflets, respectively. While the overall leaflet fiber architecture remained relatively unchanged, decellularization resulted in substantial microscopic disruption. In conclusion, changes in mechanical and structural properties of decellularized leaflets were likely associated with disruption of the ECM, which may impact the durability of the leaflets.
The aortic valve (AV) consists of three leaflets (often referred to in the literature as cusps due to their shape) that permit unidirectional blood to flow from the left ventricle to the aorta. Structurally, the AV leaflet (AVL) consists of the fibrosa, spongiosa, and ventricularis layers. The fibrosa layer, which faces the aorta, is primarily composed of Type I collagen fibers with a strong preferred circumferential orientation. The ventricularis, which faces the left ventricle, is composed of elastin and collagen. The spongiosa, located between the fibrosa and the ventricularis, is mainly composed of glycosaminoglycans and water . The AV is capable of withstanding 30–40 million cycles per year, resulting in a total of ~3 billion cycles in a single lifetime . The astonishing performance of the AV can be interrupted by AV disease, and needs to be repaired or replaced to avoid cardiopulmonary failure or death . At least 60,000 substitute valves are implanted in the United States and 170,000 world wide each year .
Currently, mechanical and bioprosthetic heart valves (BHV) are the major valve substitute types. Mechanical valves have a functional life span of at least 25 years but are associated with a substantial risk of thromboembolism and life-long anticoagulation treatment . BHVs have better hemodynamic characteristics and avoid long-term anticoagulation therapies but suffer from structural dysfunction due to progressive tissue deterioration . Also, all clinically used tissue valve substitutes are nonviable, and thus, they have no potential to grow, to repair, or to remodel. Therefore, their capability is limited, especially in growing children [6, 7].
Cellular remnants within treated BHV tissues have been found to form the nidus for calcification and related immunological responses [8–14]. Thus, decellularization of the AV potentially attenuates calcification and immunological responses [10, 11, 13, 14]. Moreover, it was found that natural ligands and ECM constituents of decellularized valve tissues benefit cell attachment, endothelialization, and tissue reconstitution . Based on these results, decellularized AVs, wherein the interstitial cells have been extracted through tissue decellularization, have been explored as used for valve replacement [15–22]. Decellularized native valve tissues, unlike synthetic scaffold approaches [6, 23, 24], have the potential to be readily used for AV replacements since they are assumed to have the necessary mechanical strength and the inherent functional design [15–22].
Various detergents or enzymatic agents have been used to remove cells and cellular debris in valved conduits. Commonly used decellularization methods are a non-ionic detergent, Triton X-100, (tert-octylphenylpolyoxyethylen) [16, 21, 25], an anionic detergent, SDS (sodium dodecyl sulfate) , and an enzymatic agent, Trypsin [15, 18, 22, 26, 27]. To understand the mechanics of decellularized valve leaflets, Korossis et al  measured the leaflet strips cut along the circumferential or the radial direction and found that after treatment with SDS (hypotonic buffer), extensibility and failure strain of circumferential strips significantly increased, while in the radial strips, there was some increase but not significantly . Although the tensile modulus decreased, ultimate tensile strengths of both circumferential and radial specimens did not decrease after SDS treatment . Similarly, Spina et al.  reported that after decellularization with either Triton X-100 (with cholate) or N-cetylpyridinium extraction, only circumferential specimens exhibited ~20% higher extensibility and had a ~10% lower stiffness. Other parameters such as extensibility and stiffness in radial direction and failure strain, failure strength, and relaxation slope of both circumferential and radial specimen were not significantly different .
The above studies suggest that decellularization procedures introduce some changes, but may not compromise tissue mechanical strength and performance [20, 28]. However, even if the ultimate tensile strength was not compromised [20, 28], it does not mean that acellular valvular tissues will perform the same way as the native tissue in the physiological range. In general, it is unclear as to how valvular function is affected by the decellularization process. Further, if recellularization and tissue maintenance do not occur in vivo, it is crucial to determine how long the decellularized AV will function.
To address these issues, an understanding of the mechanical and microstructural defects induced by decellularization of the AVL, and establishing functional limits of the decellularized AV. Thus, there is a need for in-depth studies of the mechanical and structural properties of the decellularized leaflets in the physiological functional range. In the present study, we conducted biaxial and flexural mechanical testing, SALS, and histological analyses to investigate the defects caused by the decellularization processes.
Fresh porcine AVs were harvested from hearts obtained from a local slaughter house, placed into chilled phosphate buffered saline (PBS), and transported to the laboratory, then frozen temporarily in PBS at −80°C for later use. At time of testing, the AVs were quickly thawed, and the leaflets were dissected from the aortic root. The AVLs were then placed in a decellularization solution.
A detailed description of the biaxial testing device has been previously presented [29, 30]. Briefly, a 10 mm × 10 mm square was dissected from the central belly region of the decellularized leaflet. Four fiducial markers were placed in the center of the square. Two loops of 000 nylon suture of equal length were attached to each side of the specimens with four stainless steel hooks. The specimens were then mounted onto the biaxial deice with the radial and circumferential directions aligned with the x1 and the x2 stretch axes of the device, respectively. All testing was performed with the specimen completely immersed in PBS (pH 7.4) at room temperature.
During testing, membrane tension (force/unit length) was applied along each axis and was ramped slowly from a pre-stress, ~0.5 N/m, to a peak value that depended upon the protocol using a triangular waveform. Specimens were first preconditioned for 10 contiguous cycles, following seven loading protocols: Tcc:TRR = 10:60, 30:60, 45:60, 60:60, 60:45, 60:30, and 60:10 N/m, where TCC and TRR are the applied tensions in the circumferential and radial directions, respectively . A peak tension of 60 N/m was used because it corresponds to in vivo diastolic pressure . Net extensibility was characterized by areal strain under 60 N/m tension
where λrr and λcc were the maximum stretches along radial and circumferential directions, respectively.
In addition to determining tissue extensibility, we also examined the maximum tangent moduli (MTM) at the peak tension of 60 N/m along the circumferential and radial axes. Due to the highly coupled behavior of aortic valve tissues [2, 31], the MTM does not have a unique value but instead is highly dependent on the tissue loading state. We thus decided to determine MTM values for each axis using the specific test protocols wherein the loading along the axis was the greatest (i.e. Tcc:Trr = 60:10 and Tcc:Trr = 10:60 for the circumferential and radial directions, respectively). We also compared the inter-fiber mechanical interactions and fiber rotations, providing further insight into the structural changes caused by decellularization. As in our previous study  the degree of axial cross-coupling was defined as the percent change in peak stretch ratio along each axis, as the biaxial load was changed from equibiaxial (TCC:TRR = 1:1) to non-equibiaxial conditions (e.g. TCC:TRR = 1:0.17) and calculated by
where was the peak stretch at 60 N/m and Δλ was the difference between the peak stretch at non-equibiaxial condition and
Bending is also a major leaflet deformation mode in native  and bioprosthetic heart valves . Moreover, exposing AVL to bidirectional flexural deformations will subject the AVL layers to tensile and compressive forces , allowing one to investigate the mechanical properties of the different layers of the AVL and how these properties were affected by decellularization. A detailed description of the bending device was previously presented [35, 36]. Briefly, rectangular strips were dissected from the central belly region, with the strips cut parallel to the circumferentially oriented collagen fibers. Each sample was marked with 3 fiducial graphite markers placed on the transmural surface of the sample with cyanoacrylate. The ventricularis was mounted upwards and the fibrosa faced the bottom of the bath. Hollow posts were mounted on both cross-sectional surfaces of a rectangular sample. Two anchor points, one placed on the fixed post and one placed on the bending bar were used as the final two markers used to compute the curvature. During testing, the fixed post was moved towards the bending bar. As the sample deformed, the bending bar deflected, which represented the amount of force exerted on the sample. The deformation was measured in terms of change of curvature calculated from the fiducial markers and two anchor points. Each specimen was flexed in two directions, with curvature (WC), which subjected the ventricularis to tension and against curvature (AC), which subjected the fibrosa to tension.
The overall effective stiffness was found by relating the moment and change of curvature to the flexural rigidity using the Euler-Bernoulli relationship:
This approach has been previously presented by Mirnajafi et al. . Following this approach, the following exponential function was used
Thus, the effective stiffness, Eeff, was computed using
where a and b are parameters determined by least squares fitting to the moment –curvature data.
A detailed description of the SALS technique has been previously presented . Briefly, a 4 mW HeNe continuous unpolarized laser (λ = 632.8 nm) was passed through the tissue specimen. The spatial intensity distribution of the resulting scattered light represented the sum of all structural information within the light beam envelope. The angular distribution of scattered light pattern, I(Φ), which represents distribution of fiber angles within light beam envelope, was obtained. Quantifiable information based on I(Φ) includes orientation index (OI) and preferred fiber direction (Φc). The orientation index (OI) was defined as the angle that contains one half of the total area under the I(Φ) distribution. Normalized orientation index (NOI) was calculated using:
where NOI ranged from 0% for a complete random network to 100% for a perfectly aligned network .
To prepare specimens for SALS testing, the leaflets were first fixed with glutaraldehyde in order to prevent tissue degradation during tissue preparation and fixation is required for histological preparation. The glutaraldehyde-fixed leaflets were dehydrated in graded solutions of glycerol/saline of 50, 75, 87, and 100% for an hour each. This process cleared the leaflets, which was necessary for accurate SALS measurements. It has been shown that the graded glycerol solution did not measurably distort the leaflet shape . SALS measurements were then conducted over the entire leaflet surface in order to quantify the gross fiber structure of the leaflets. To quantify regional variations, the SALS data were subdivided into upper (Regions A and B) and lower (Regions C and D) commissure regions and upper (Regions 1, 2, and 3) and lower (Region 4 and 5) belly regions .
The following histological analyses were performed to assess overall dimensional changes due to decellurization, verify that no cell nuclei was present, assess the degree of preservation of the leaflet trilayered structure, and to observe changes in the collagen crimp structure. Before and after decellularization of the AVLs, thickness measurements were taken with a dial caliper with an accuracy of ±0.0254 mm. Three thickness measurements were taken in the belly region, and two thickness measurements were taken in the commissure regions. The area of the valve leaflet was measured by plotting the leaflet contour and then quantifying with imaging software (Image Pro Plus) with an accuracy of ±0.1mm2.
For histological analysis, transverse sections were cut from the belly regions, where the sections were aligned so that the long edge was parallel to the local preferred fiber direction. A hematoxylin and eosin (H&E) stain was used to validate that no cell nuclei was present after decellularization of the AVLs. To elucidate the trilayered leaflet structure, a Movat Pentachrome stain was used to identify collagen, elastin, and glycosaminoglycans with yellow, black and blue color, respectively. A picrosirius red stain was used to investigate the effects of decellularization on the collagen crimp structure. Under polarized light microscopy, collagen fibers of the AV displayed periodic light-distinguishing bands that corresponded to collagen crimp periods. All stained leaflet sections were imaged using bright field microscopy.
For scanning electron microscopy (SEM), a square section was dissected from the central belly region of the 3 different decellularized leaflets. The square section was then cut into 2 smaller squares, in order to examine the fibrosa of the first square and the ventricularis of the second square. Each section was fixed with 2.5% glutaraldehyde for 24 hours followed by three 15-minute washes in PBS. Then, each section was washed in sequential washes of 30%, 50%, 70%, 90%, and 100% ethanol. The specimens were left in each wash for 15 minutes and the 100% washed was performed 3 times. Next, specimens were subjected to critical point drying and gold sputter coating. Finally, SEM was used to view the ultrastructure of the fibrosa and the ventricularis of the decellularized leaflets.
A Statistical analysis was performed with One Way Analysis of Variances (ANOVA) (SigmaStat 3.0, SPSS Inc., Chicago, IL). The Holm-Sidak Test, which can be used for pair wise comparisons and comparisons versus a control group, was used for the post hoc comparison. Results were considered significantly different at p < 0.05.
The net extensibility, represented by areal strain under 60 N/m equibiaxial tension, was found to increase after all three decellularization treatments when compared to the native AV  (Fig. 1-a). Among the three decellurization protocols, Triton X-100 showed the largest increase (p=0.003, Fig. 1-a). Maximum stretch ratios were evaluated in both the circumferential and the radial directions at 60 N/m equibiaxial tension (Fig. 1-b), which showed that Triton X-100 had the greatest effect on the extensibility in the radial direction (p < 0.001). Overall, the MTM of the radial direction was not affected by the decellularization process (p = 0.736, Fig. 2). However, the MTM of the circumferential direction was increased in tissue decellularized with Triton X-100 (p=0.003, Fig. 2). Mechanical coupling of the stretch axes was affected in all decellularization protocols (Fig. 3), indicative of tissue structural alterations. In general, all decellurization protocols induced larger changes in stretch along the opposite stretch axis. This is likely due to increased fiber mobility after decellularization resulting from a collagen network less bounded after decellularization.
All decellularization treatments were found to significantly affect the flexural behavior of valve leaflets. As in our previous study, native AVLs demonstrated a near linear moment-curvature relationship (Fig 4-a, ). However, the decellurized tissues did demonstrate a profound stiffness loss, which decreased with increasing flexure (Fig. 4-b). Moreover, this nonlinear response was observed when the specimen was flexed in both WC and AC directions (Fig. 4). The resulting moment-curvature response was fit well by equation (4) to determine Eeff (Δκ) (Table 1). To determine a mean specimen stiffness, Eeff at Δκ=0.01 was computed for each specimen based on the fit. The resulting Eeff of decellularized AVL was then compared with the native AV. The mean Eeff of the native AV was 156.0±24.6 kPa when flexed WC and 133.7±20.7 kPa when flexed AC.
The gross collagen fiber structure was still preserved after decellularization (Fig. 5). Evaluation of tissue dimensions revealed that decellularization of the AV caused variable results. Decellularization with SDS did not cause any statistical differences (p=0.192) in area or thickness (Table 2). Decellularization with Trypsin caused tissue area and thickness dimensions to increase (p=0.011 and p=0.002, respectively (Table 2), and decellularization with Triton-X caused tissue area and thickness dimensions to decrease (p ≤ 0.001 and p ≤ 0.001, respectively) (Table 2).
SALS results for native AV exhibited significant regional variations of various degrees of alignment (Fig. 6). The region of coaptation and the mid belly region was aligned relatively well, whereas the region of the leaflet, which contained the Nodulus of Aranti, displayed poorer alignment (Fig. 6-a). The regional variations were preserved in the leaflets decellularized with SDS and Trypsin (Figs. 6-b,c), but were not preserved in leaflets decellularized with Triton X-100. The leaflets decellularized with Triton X-100 displayed a more homogenous fiber ultrastructure (Fig. 6-d). A regional analysis gave further insight into the regional variations of the decellularized leaflets. The analysis revealed that there were not significant differences between the native AV and the decellularized AVs, but, overall, decellularization caused a decrease in NOI values in each commissure and belly region (Table 3).
Histological results indicated no cell nuclei present after all three decellularization protocols (Figs. 6 b-d) as compared to the native AV (Fig. 7-a). The trilayered structure of the leaflet decellularized with SDS was very similar to the trilayered structure of the native AV (Figs. 7 e,f). However, the spongiosa was depleted with decellularization with Triton X-100 and Trypsin (Figs. 7-g, h). The macroscopically well-organized collagen crimp structure of the fibrosa was disrupted by all three decellularization protocols as compared to the native AV (Figs. 7 I–l). The collagen fibers were locally crimped (Fig. 7 j–l) but no longer showed macroscopic distinguishing bands (Fig. 7-i).
Comparing the fibrosa of all decellularized leaflets, SEM verified that leaflets decellularized with SDS preserved a dense collagen network (Fig. 8-b). Trypsin and Triton X-100’s collagen networks were loose with large pores (Figs. 8-c, d). When comparing the ventricularis of all the decellularized leaflets, the same results were found (Figs. 8f–h). Overall, leaflets decellularized with SDS contained a dense ECM network similar to the ECM of the native AV (Figs. 8a, e), whereas, leaflets decellularized with Trypsin and Triton X-100, displayed a looser ECM network.
While heart valve tissue engineering has generated much interest, many unanswered questions remain and technical barriers must be overcome before widespread clinical application can be envisioned [41–43]. Normal heart valves are vital and dynamic tissues composed of specialized cells and ECM that respond and remodel in response to changes in local mechanical forces . A successful TEHV and its components must accommodate repetitive changes in shape, dimension, and stress of the leaflet due to opening and closing of the valve leaflets. It must also have ongoing strength, flexibility, and durability, beginning at the instant of implantation and continuing indefinitely thereafter . Thus, the goal of the current study was to provide insight into these long-term goals for the decellurization approach to TEHV. Understanding the mechanical and the microstructural effects caused by decellularization of the AV tissues and establishing functional limits are essential to the future development of a functional TEHV.
All decellularization protocols (SDS, Trypsin, and Triton X-100) induced an increase in leaflet extensibility, with Triton X-100 having the greatest affect on the radial extensibility (Figs. 1-a, b). This hypothesis is supported by the results for axial cross-coupling (Fig. 3a, b) which indicated larger radial strains for all protocols for decellurized tissues. It is known that the large radial strains in the native aortic valve are enabled by large rotations of the circumferentially oriented fibers rather than more extensible fibers aligned to the radial direction [2, 31]. Thus, increased radial extensibility was likely due to the increased ability for the circumferentially oriented collagen fibers to rotate in the decellurized tissue matrix. Along similar lines, the flexural behavior changed markedly after each decellularization protocol (Fig. 4). The nonlinear response and decrease in flexural rigidity was likely associated with ECM disruption. Overall, all mechanical measurements were affected by decellularization. This implied that the decellularization process induced mechanical and microstructural defects.
Overall, decellurization did not change the gross appearance (Fig. 5) or collagen fiber network (Fig. 6) when compare do the native valve. This may give the initial impression that decellurization does little damage to the AV ECM. Yet, decellularization with Trypsin caused the tissue area and thickness dimentions increase and decellularization with Triton X-100 caused the tissue area and thickness dimensions to decrease (Table 2). Tissue stained with H&E demonstrated that there were no cells nuclei present after decellularization with all three protocols (Figs. 7 b–d), and a preserved the trilayered structure with SDS (Fig. 7-f). However, decellularization with Trypsin and Triton X-100 resulted in a depletion of the spongiosa layer (Fig. 7 g, h). It has been suggested that the principle function of the spongiosa is to dampen the vibrations in the fibrosa associated with leaflet flexion during closure . Due to the loss of major GAG content in the spongiosa with decellularization of Trypsin and Triton X-100, early tissue degradation may result. Moreover, decellularization with Trypsin appeared to have induced a disruption of the elastin layer (Fig. 7-g), causing a loss of its organization.
Using picrosirius staining, decellurized AV tissue revealed a disruption of the collagen crimp structure when decellularized with each protocol (Fig. 7j–l) compared to the native AV (Fig. 7-i). The collagen fibers were locally crimped (Fig. 7j–l) but no longer showed the macroscopically organized light distinguishing bands (Fig. 7-i). This may have major implications on the long term tissue durability. Loss of collagen crimp indicates a loss of a protection mechanism, which prevents the collagen from being damaged during loading.
SEM revealed that tissues decellularized with SDS had a dense collagen network with a small pore size as opposed to the loose collagen network and increased pore size of tissues decellularized with Trypsin and Triton X-100 (Fig. 8b–d). Evaluation of the ventricularis revealed the same results (Figs. 8f–h). It has been hypothesized that normal tissue homeostasis was a function of aortic valve interstitial cells (AVIC) . Although the ECM of SDS treated leaflets bore close resemblance to that of the native AV (Figs. 8a,e), the dense collagen network and small pore size may limit AVIC repopulation. If AVIC repopulation did not occur, the ECM will probably degrade over time, rendering the tissue scaffold nonfunctional.
In our experiments, heart valves were temporarily stored at −80 °C for testing. It has been demonstrated that short term storage at low temperature has minimal effects on mechanical properties of connective tissues.[46–48] A study also showed that freeze storage in saline solution only significantly changes the mechanical properties of collagenous tissue (tendon) for duration longer than 40 days. Recently, Schenke-Layland et al. showed that certain local ultrastructural disruption was caused by one week cryopreservation using multiphoton laser scanning microscope. Note that the structural deterioration after decellularization procedures were more aggressive than what observed in Schenke-Layland’s study, in which conventional histology revealed almost comparable cell and ECM formations in fresh and cryopreserved valve leaflets. Apparently, structural deterioration in our observation was caused by decellularization procedures instead of temporary freeze storage.
The potential for decellularized aortic heart valves (AVs) as heart valve replacements is based on the assumption that the major cellular immunogenic components have been removed, and that the remaining extracellular matrix (ECM) retains the necessary mechanical properties and functional design. In the present study we explored the effects of an anionic detergent (SDS), enzymatic agent (Trypsin), and a non-ionic detergent (Triton X-100) on the mechanical and structural properties of AV leaflets (AVLs) to provide greater insight into the initial functional state of the decellularized AVL. Overall, SDS appeared to maintain critical mechanical and microstructural properties. However, SEM suggested that although the ECM of SDS treated leaflets resembled that of the native AV, AVIC repopulation may not be optimal due to a dense ECM network and a small pore size. While the overall fiber architecture remained relatively unchanged, polarized light imaging and SEM revealed that decellularization resulted in substantial disruption of collagen network. In conclusion, changes in mechanical and structural properties of decellularized leaflets were likely associated with disruption of the ECM, which may impact the durability of the leaflets.
This work is supported by a Beginning Grant-in-Aid (0565346U) from the American Heart Associate (Pennsylvania/Delaware Affiliate) to JL. The authors would like to thank Jennifer Debarr of the McGowan Institute and Mark Rubin of Center for Biologic Imaging for their assistance with histology and SEM.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.