journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
Available online at www.sciencedirect.com
www.elsevier.com/locate/jmbbm
Research Paper
Recrystallization improves the mechanical properties of sintered electrospun polycaprolactone M. Tyler Nelsona, Lagnajit Pattanaikb, Marcia Allenb, Matthew Gerbichb, Kelvin Huxb, Matthew Allenc, John J. Lannuttib,n a
Department of Biomedical Engineering, Ohio State University, Columbus, OH 43210, United States Department of Materials Science and Engineering, Ohio State University, Columbus, OH 43210, United States c Department of Veterinary Clinical Sciences, Ohio State University, Columbus, OH 43210, United States b
ar t ic l e in f o
abs tra ct
Article history:
Background: Resorbable electrospun polycaprolactone (PCL) scaffolds for tissue reconstruc-
Received 12 August 2013
tion can provide physicians with an “off the shelf” product tailored to the patient's specific
Received in revised form
tissue architecture. However, many tissue-engineering platforms do not possess the
31 October 2013
necessary long-term mechanical stability needed to properly support tissue development.
Accepted 2 November 2013
Objective: Sintering has been explored as a means of altering the properties of electrospun
Available online 15 November 2013
PCL. However, crystallinity-driven changes in mechanical properties following thermal treatment have not been previously investigated. Methods: PCL nanofibers were produced by electrospinning and subsequently thermally sintered (at 55, 56 and 58 1C) to enhance their long-term mechanical integrity in response to representative biological milieux. Results: Scaffolds initially sintered at 56 1C displayed 6-fold increases in compressive strength and 3-fold increases in modulus, while displaying 10-fold increases in energy dissipation with increasing sintering temperature. Sintering just below the Tm resulted in amorphization of the 55 1C sample as indicated by the 20-fold lower XRD peak intensities. Although crystallinity is suppressed, the polymer chains likely retain chain alignment from electrospinning and are apparently highly susceptible to recrystallization. After only 1 d PBS exposure, the 55 1C samples recover a substantial fraction of the as-spun crystallinity; 7 d of exposure fully restores as-spun peak intensities. The mechanical properties of all three (55, 56, or 58 1C) scaffolds displayed peak values of compressive strength and modulus following 7 d exposure. Conclusion: In contrast with the current state-of-the-art which assumes that tissue engineering scaffolds only grow weaker following exposure, in these scaffolds maximum values of compressive strength and modulus were observed after 7 d of aqueous immersion. This suggests that polymeric recrystallization can be used to increase or optimize mechanical properties in vitro/in vivo. Scaffolds that increase their mechanical integrity during biological exposures constitute a new pathway enabling clinical advances. & 2013 Elsevier Ltd. All rights reserved.
n
Corresponding author. Tel./fax: þ1 614 688 3182. E-mail address: [email protected] (J.J. Lannutti).
1751-6161/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jmbbm.2013.11.004
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
1.
Introduction
Engineered tissue substitutes as bone, blood vessel, or skin grafts can provide a convenient alternative to allografts. However, many engineered substitutes have been hampered by losses of mechanical integrity requiring additional surgeries or, in more extreme cases, significant increases in mortality (Bryers et al., 2012; Johnson et al., 2009, 2007; Ma et al., 2005; Venugopal et al., 2005; Pektok et al., 2008; Drilling et al., 2009; Nam et al., 2007; Lee et al., 2008). Since the first successful decellularized donor reconstruction in 2008 many have suggested that this technology presents the best alternative for future scaffolds (Baiguera et al., 2011). However, given donor scarcity and degradation of mechanical properties resulting from the decellularization process (Baiguera et al., 2011), this pathway may not, in fact, be widely applicable. Other naturally-derived tissue engineered constructs have also struggled to provide the necessary mechanically stable, chemically inert environment allowing the desired cell/tissue maturation (Kim et al., 2010; Kobayashi et al., 2010; Komura et al., 2008, 2010; Liu et al., 2010; Tada et al., 2008, 2012). Synthetic tissue engineered constructs have long been proposed as a means of eliminating the need for donor tissues. Synthetic materials provide design-specific tailoring options making them ideal for tissue engineering applications. Many studies have shown the efficacy of producing porous, non-biodegradable, synthetic substitutes for reconstruction(Suh et al., 2001; Jungebluth et al., 2012; Gustafsson et al., 2012). In 2011, the first fully synthetic, non-resorbable tissue-engineered trachea substitute inoculated with mononuclear stem cells was implanted successfully in a human with no significant complications at 12-months follow up (Jungebluth et al., 2011). However, loss of mechanical integrity in these synthetics can render the scaffolds unsuited to tissue development (Jungebluth et al., 2011). Since then, much of the focus has been directed toward engineering substitutes capable of providing the initial mechanical support and stability needed for autologous tissue/cartilage maturation. To our knowledge, there have been no studies that utilize increases in mechanical properties during aqueous exposure to meet clinical needs. The fibrous microstructure and chemistry found in decellularized donor tissue provides a native environment promoting cellular attachment, proliferation, and tissue-directed differentiation crucial to the development of mechanically proficient TE cartilage (Baiguera et al., 2011, 2010a). Electrospinning offers tremendous potential for reproducing the unique ECM microstructure while allowing for the ability to tailor the mechanical properties of a scaffold using synthetic, natural, or blends of natural and synthetic polymers. Electrospun scaffolds can reproduce the fibrous microstructure, porosity, and micro/nano-scales of native tissue ECM allowing them to be personalized for a wide range of TE applications involving skin, bone, blood vessel, heart valve, and nerve target tissues (Drilling et al., 2009; Nam et al., 2007, 2011; Bolgen et al., 2005; Chen et al., 2008; Heath et al., 2010). In this study, thermally sintered electrospun biodegradable polycaprolactone (PCL) scaffolds were investigated to examine their mechanical stability and microstructure with the hope of providing a fully resorbable tissue-engineering
151
platform for tissue reconstruction. As an attempt to provide a fully resorbable tissue engineered scaffold, PCL was chosen to be electrospun due to its relative resistance (versus other resorbable polymers (Bryers et al., 2012)) to hydrolytic degradation. A key advantage over non-resorbable materials, such resorbable scaffolds provide surgeons with a construct capable of growing and maturing with pediatric patients in vivo while slowly degrading (Bryers et al., 2012). To introduce a potentially wide-ranging means of improving mechanical properties, thermally driven sintering was investigated to gauge its effects on compressive strength, modulus, and energy dissipation. Following the observed improvements in mechanical properties, exposures to aqueous biological milieu revealed a previously unreported phenomenon: amorphization followed by recrystallization. Sintering temperatures near the melting temperature (Tm) of electrospun PCL produced microstructures and crystalline behavior capable of maintaining mechanical properties similar to native tissue. As expected, however, the process greatly alters the underlying microstructure. Future investigations must be directed towards selective sintering of electrospun scaffolds by incorporating thermally inert regions capable of maintaining the highly porous electrospun microstructure necessary for cellular infiltration, nutrient in-flow, release of metabolic products and tissue maturation.
2.
Materials and methods
2.1.
Electrospinning
14 wt/wt% polycaprolactone (PCL, ffi80,000 Mw, SigmaAldrich, St. Louis, MO) in acetone (99% purity, Sigma-Aldrich, St. Louis, MO) solution was prepared by mixing at 55 1C. Once the solution was fully dissolved and transparent it was added to a 60 mL syringe (Becton-Dickinson, Franklin Lakes, NJ) fitted with a 20 gauge blunt tipped, stainless steel needle (Nordson EFD, West Lake, OH). A solution filled syringe was then placed in a syringe pump set to a flow rate of 15 ml/h. Using a working distance from the needle tip to the grounded aluminum collector of 20 cm (Gaumer et al., 2009), a positive, DC voltage of 25 kV was applied to the needle tip. The collector was a 2.5 cm diameter by 10.2 cm long cylindrical, aluminum mandrel rotated at a linear velocity of 5 m/s. In order to acquire 1.5 mm-thick electrospun sample, two 60 mL syringes dispensed flow continuously for a 4-h period. Upon completion of electrospinning, a 1.5 mm thick by 10.2 cm long by 2.5 cm diameter hollow tube formed on the mandrel. For testing purposes, 1 cm wide rings were cut from the electrospun PCL tube. In total, 35 such tubes were produced to enable a statistically significant test.
2.2.
Thermal sintering
As-electrospun PCL ring samples were sintered at 55, 56, or 58 1C for 24 h using a standard recirculating water bath. Samples were first chemically isolated from the surrounding water inside securely sealed screw-capped glass vials. To preserve a constant 2.5 cm diameter, the ring samples were placed on a custom 2.5 cm diameter stainless steel rod during
152
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
the sintering process. After sintering, these rings were removed from the water bath and placed in a plastic bag at room temperature 25 1C) until use. As-spun PCL rings were used as a control, and were placed in sealed plastic bags immediately after electrospinning.
2.3.
Mechanical testing
Compressive mechanical testing was conducted on the asspun and sintered electrospun PCL rings using a load frame (Type R, TestResources Inc, Shakopee, MN). 50% compressive strain (1.25 cm total deformation) was applied to the rings at a crosshead rate of 25.0 mm/min. The loading and unloading profiles of these rings was then recorded. Samples were tested at room conditions 25 1C) in air, and in phosphate buffered saline (PBS) containing 0.1% sodium azide at a temperature of 37 1C maintained by a heating coil and constant stirring. Sodium azide was added to prevent microbial or fungal infection that could result in the destruction and/or degradation of electrospun PCL rings imparting nonrandom variation in the observed mechanical properties. The ultimate compressive strength (UCS) was calculated as the maximum load recorded at 50% strain, and the compressive modulus was measured from the slope of the loading curve prior to detectable yielding. Energy dissipation was also recorded for each of the sintered conditions and the asspun PCL scaffolds by measuring the area in between the loading and unloading hysteresis curves using Matlab mathematic toolbox software (Mathworks, Natick, MA).
2.4.
Extended exposure testing
Rings from each of the sintered conditions and as-spun rings were subjected to exposure to 37 1C PBS over a 28-d period (Johnson et al., 2009). Ring samples were placed directly into polystyrene conical tubes (Becton-Dickinson, Franklin Lakes, NJ) filled with 40 ml of PBS solution containing 0.1% sodium azide. These rings were then kept in the solution for 1, 7, 14, or 28 days at 37 1C in an incubator prior to being mechanically tested under compression. Rings were removed from the PBS solution at the end of each time point and tested immediately. All samples remained saturated with PBS solution during the test. Results are displayed as the mean values with 95% interval of confidence for ultimate compressive strength (UCS), compressive modulus, and energy dissipation. Significant factor interaction between sintering temperature and exposure time was detected by 2-way ANOVA statistical analysis (Minitab 16); interaction plots for each of the response variables have been inserted.
2.5.
Scanning electron microscope (SEM) analysis
Analysis of microstructure was conducted for as-spun and all sintered conditions prior to and after PBS solution exposure. All samples were fully dried for 24 h at atmospheric room temperature conditions prior to SEM preparation. SEM samples were cut from the electrospun PCL rings and mounted on aluminum studs (SPI Supplies/Structure Probe Inc., West Chester, PA) covered with carbon tape (SPI Supplies/Structure Probe Inc., West Chester, PA). Samples were then coated with
15 nm of gold using a gold sputter coater at a plasma current of 15 mA. An environmental scanning electron microscope (ESEM, Philips XL-30 ESEM-FEG, Surrey, UK) operated at an accelerating voltage of 15 kV was used to image the as-spun and sintered samples. Porosity and Pore-size measurements were extracted from five different SEM images for each condition (As-spun, 55, 56, and 58 1C) using NIH ImageJ software. To determine porosity SEM images of the scaffolds were thresholded to display only the void areas of the image and the resulting areas were measured as a percentage of the total area to estimate scaffold porosity. Pore-size was quantified by measuring the long dimension and short dimension of space between fiber–fiber boundaries and taking the average length.
2.6.
X-ray diffraction
Using a Bragg-Brentano configuration, X-ray diffraction patterns were acquired before and after 7 days of PBS exposure for sintered and as-spun electrospun PCL. Rigaku Ultima-III diffractometer (Rigaku Americas Corp., Allison Park, PA) with a Cu Kα source and a diffracted beam monochromator analyzed electrospun specimens mounted on aluminum disks over a 2θ angle range of 19 to 251 with a step-size of 0.041. Each sample was electrospun for equal time to produce a 1.5 mm thick disc. To ensure consistent X-ray absorption, the mass equivalency of each sample verified to within 0.025 g prior to XRD. The disks were rotated at 120 rpm during data acquisition with an integration time at each step of 24 s. A constant linear background was used in the profile fitting to eliminate potential background noise and scattering when comparing the patterns from different samples. In addition, 55 1C sintered PCL samples were analyzed using XRD before and after 1 d and 28 d of exposure.
2.7.
Statistical analysis
All results are displayed as mean 795% confidence interval calculated using Minitab 16 statistical software. Mechanical testing results were acquired using a single factor random experimental design model analyzed using a one-way ANOVA and determining statistical significant differences via a Tukey's test with a 95% confidence interval. Exposure testing results were analyzed using 2-way ANOVA and a Tukey's full factor interaction test with a 95% confidence interval.
3.
Results
3.1.
Electrospun microstructure
SEM images of the electrospun PCL microstructure for asspun and each of the sintered conditions are displayed in Fig. 1. As-spun PCL fiber displays cylindrical, individual fibers and minimal fiber–fiber connections providing a highly porous microstructure with extensive pore interconnectivity (Fig. 1: As-spun). With increasing sintering temperature the nanofiber morphology exhibits more fiber–fiber connections, bonding, and loss of porosity (Figs. 1 and 2(A and B)). Samples sintered at 55 1C depict increases in fiber diameter, and increased fiber–fiber bonding as compared to as-spun PCL
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
153
Fig. 1 – SEM images of (A) as-spun and (B) 55 1C, (C) 56 1C, and (D) 58 1C sintered electrospun PCL nanofibers. (A) Depicts independent fiber morphologies, free of fiber–fiber connections. With increasing temperature (B) to (D), the images exhibit more fiber–fiber bonding, a reduction in porosity, and less independent fiber.
Fig. 2 – Porosity and pore-size quantification from SEM images depicts that with increasing sintering temperature significant reductions in free volume or porosity are observed (A), however the pore-size remains constant up to the experimental melting temperature (58 1C) (B). n Denotes statistical outliers, and # denotes statistically significant results (po0.05).
samples. Significant alterations to the electrospun fiber morphology and scaffold microstructure are not observed until the sintering temperature exceeded 55 1C, as is the case for samples sintered at 56 or 58 1C. Very little of the original electrospun fiber morphology is retained after scaffolds are sintered at 56 1C as severe fiber–fiber bonding and significant loss of the open, highly porous microstructure are observed. 58 1C-sintered scaffolds exhibit a completely closed pore, dense, film-like microstructure. As-spun and 55 1C sintered samples maintain the unique electrospun fiber microstructure that more closely resembles this natural microstructure.
electrospun PCL samples at 50% strain are displayed in Figs. 3 and 4, respectively. The UCS and compressive modulus of electrospun PCL samples significantly increase with sintering temperature. Brief exposures (less than 1 min) to PBS (as a biological testing environment) demonstrated no significant differences versus testing under dry, atmospheric room conditions (data not shown). Fig. 5 presents energy dissipation data for each of the electrospun PCL scaffold conditions. In a very similar manner, energy dissipation increased 40-fold versus thermal exposure.
3.3. 3.2.
Exposure
Mechanical properties
The ultimate compressive strength (UCS) and compressive modulus for each of the sintered conditions and the as-spun
To evaluate the long-term mechanical integrity of sintered electrospun PCL following exposure to aqueous environments, both as-spun and sintered samples were exposed to
154
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
Fig. 3 – Ultimate compressive stress versus electrospun PCL thermal processing condition. Samples sintered at 56 or 58 1C exhibited significantly greater UCS values as compared to as-spun or samples sintered at 55 1C. n Denotes statistical outliers, and # denotes statistically significant results (po0.05).
Fig. 4 – Compressive modulus stress versus electrospun PCL thermal processing condition. The modulus of PCL nanofibers sintered at 56 1C were 6-fold greater and those sintered at 58 1C exhibited 18-fold greater values as compared to as-spun or samples sintered at 55 1C. n Denotes statistical outliers, and # denotes statistically significant results (po0.05).
PBS for a period of 28 d. Fig. 6 displays the UCS for as-spun and sintered samples tested following 1, 7, 14, or 28 d of exposure. As-spun, 55, and 56 1C sintered PCL samples displayed significant decreases in UCS after 1 day of exposure, followed by a significant increase after 7 days. Compressive moduli exhibited trends similar to – but not as pronounced as – those of UCS (Fig. 7). Both UCS and compressive moduli for samples sintered at 58 1C maintained these significant increases at day 14 while for all samples 28 d of exposure resulted in a decrease of mechanical properties to preexposure levels. Energy dissipation versus 28 d of exposure decreased for all samples versus the as-electrospun zero exposure condition (Fig. 8). A 55 and 56 1C sintered samples displayed significant increases in energy dissipation after 7 d, followed by a sharp reduction at 14 d that brought the values back to pre-exposure levels.
Fig. 5 – Bar-plot of energy dissipation versus electrospun PCL thermal processing condition. Energy dissipation of sintered PCL nanofibers was greater than the as-spun condition following for all thermal exposures. 56 1C were 20-fold and 58 1C were 30-fold greater then as-spun PCL nanofiber. n Denotes statistical outliers, and # denotes statistically significant results (po0.05).
Fig. 6 – Ultimate compressive strength (UCS) versus exposure for as-spun and sintered electrospun PCL nanofibers. UCS always increased with sintering temperature. All samples experienced a decrease in strength after 1 day followed by a significant increase after 7 days of exposure. UCS for all samples, after 28-d, was equal or nearly equal to as-spun conditions. n Denotes statistical outliers, and # denotes statistically significant results with respect to the interaction between sintering temperature and exposure time (po0.05).
3.4.
X-ray diffraction
The XRD patterns for as-spun and sintered samples were acquired before and after PBS exposure. Fig. 9 displays the patterns for each of the sintering conditions (55, 56, and 58 1C) and as-spun PCL nanofiber before and after 7 d of PBS exposure. In all cases sintering initially greatly reduced the as-spun crystalline peak intensity. Conversely, 7 d of exposure to PBS resulted in significant increases in crystalline peak intensity. Samples sintered at 55 1C presented the most dramatic decrease in crystalline peak intensity after sintering, followed by the greatest rebound in crystalline peak
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
Fig. 7 – Compressive moduli versus exposure time for as-spun and sintered electrospun PCL nanofibers. Modulus results follow trends similar to that of UCS results, indicating that with increasing temperature samples became stiffer. n Denotes statistical outliers, and # denotes statistically significant results with respect to the interaction between sintering temperature and exposure time (po0.05).
Fig. 8 – Energy dissipation versus exposure time for as-spun and sintered electrospun PCL nanofibers. Sintered samples displayed greater energy dissipation values as compared to as-spun PCL nanofibers for any exposure time. n Denotes statistical outliers, and # denotes statistically significant results with respect to the interaction between sintering temperature and exposure time (po0.05).
intensity after PBS exposure. To determine if this effect occurs following much shorter exposures, a 1-d PBS exposure followed by XRD analysis was conducted. Fig. 10 displays the XRD patterns for as-spun PCL before and after sintering at 55 1C, and the 55 1C specimen following 1 and 28 days of PBS exposure. One day of exposure significantly increases crystalline peak intensities at 21.5 and 23.751 2Θ. Interestingly, lower sintering temperatures resulted in the most significant decreases in crystalline peak intensity with the lowest temperature (55 1C) producing nearly amorphous PCL while higher temperature exposures (56 and 58 1C) display considerably more crystallinity (Fig. 9). Increases in crystalline peak intensity were seen following PBS exposure (even for the as-spun condition) and decreased with increasing sintering
155
Fig. 9 – XRD profiles of as-spun and sintered (denoted by temperature) PCL nanofiber before and after 7-d of PBS exposure (denoted by temperature and PBS). Thermal sintering resulted in significant decreases in crystalline peak intensities, with 55 1C sintered samples exhibiting the greatest loss. After subsequent exposure to PBS, all samples exhibited an increase of crystalline peak intensity.
Fig. 10 – Presents the XRD profiles of electrospun PCL nanofiber before and after 55 1C sintering and 1- and 28 d of PBS exposure. As observed previously, sintering resulted in loss of crystalline peaks, nearly exhibiting completely amorphous behavior. After only 1-d of exposure, recrystallization was observed, resulting in significant crystalline peak reformation. Following 7 d of exposure, recrystallization reaches a substantial fraction of as-spun values; 28 d of exposure greatly decreases crystallinity. temperature. The 55 1C-sintered samples displayed the greatest crystalline peak intensity rebound. Conversely, the 28-d exposure resulted a substantial decrease in crystallinity versus the 7-d exposure condition.
4.
Discussion
With the emergence of tissue engineering, regenerative medicine now offers options for tissue replacement centered
156
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
on the concept of “personalized medicine” (Baiguera et al., 2011, 2010a, 2010b; Jungebluth et al., 2012; Baiguera and Macchiarini. 2011; Delaere et al., 2012). Recent clinical successes following implantation of electrospun, non-resorbable engineered tracheal constructs for total laryngotracheal transplantation has triggered a concomitant resurgence of interest in tissue engineered organs (Fountain, 2012). Attempts to create resorbable tissue engineered scaffold substitutes have met with limited clinical acceptance partly due to risks associated with the lack of convincing, long-term mechanical property data (Jungebluth et al., 2012). As is often stated but not yet demonstrated, the rate of biomaterial degradation should match the rate of new tissue formation to compensate for the associated decreases in mechanical properties (Kai et al., 2012; Alexander et al., 1996; Cooke et al., 1996; Coury et al., 1996). However, if tissue formation takes place purely in vitro it is usually beset by over-deposition of ECM resulting in disorganized, weak scaffolding (Kim et al., 2010; Komura et al., 2010; Baiguera et al., 2012). Long term, many researchers have established that remodeling of a pre-existing ECM structure provides in vivo-like mechanical integrity and organization capable of promoting tissue maturation to meet the desired organ's functional needs (Jungebluth et al., 2012; Courtney et al., 2006). Given that the initial properties of such scaffolds must be optimized to meet clinical standards, this study characterizes the effects of thermal sintering on the microstructural and mechanical properties of tubular, electrospun polycaprolactone. Tubular electrospun PCL has been widely explored as small diameter (o6 mm) blood vessels appropriate for coronary artery replacement (Ma et al., 2005; Venugopal et al., 2005; Pektok et al., 2008; Drilling et al., 2009). Although sintering has been explored previously to alter the microstructure and properties of electrospun PCL (Nam et al., 2007; Johnson et al., 2007; Lee et al., 2008), little has been done to characterize the changes in crystallinity either after thermal treatment or the exposure of such treated microstructures to aqueous environments. This overlooks the important effects that changing crystallinity clearly might have on overall mechanical properties. The unique fibrous microstructure and extensive ( Z80%) inter-connected porosity of these electrospun PCL matrices (Fig. 1A) closely mimics the morphology and architecture of native ECM. Increasing sintering temperature resulted in the expected losses of independent fiber morphology and the unique porous microstructure. Also interesting is the definite effects that sintering will have on the ability of polymer nanofibers to rotate and translate in response to applied mechanical stresses (Johnson et al., 2007). While the sinteringdriven transition to a more highly point-bonded structure (Johnson et al., 2007) clearly imparts considerable gains in mechanical integrity (Figs. 4 and 5) by the creation of a more “net-like“ structure resisting fiber–fiber motion, these benefits come at the cost of the large scale (potentially hundreds of microns) rearrangement of which as-electrospun structures are capable (Johnson et al., 2007; Stella et al., 2010) given their lack of regular, repeating structure. Macroscopic benefits measurable by current technology do not reflect changes at the micro scale properties that may or may not be significant in controlling cell response (Stella et al., 2010).
The 55 1C scaffold displayed very little fiber–fiber bonding resulting from sintering compared to the 56 or 58 1C sintered scaffolds. As a result, the fibers within the 55 1C scaffold were able to undergo relatively large amounts of interfiber motion (Ma et al., 2005; Venugopal et al., 2005; Pektok et al., 2008) allowing the overall scaffold to deform easily with little energy dissipation. While the 58 1C-sintered PCL samples clearly exhibit undesirable changes in microstructure and morphology, samples sintered at 56 1C still possess considerable porosity capable of promoting efficient nutrient transport and cell viability. The associated UCS is 6-fold and the modulus 3-fold greater than the 55 1C samples and nearly 15-fold greater versus the as-spun condition. Long-term mechanical stability of sintered samples was maintained over a 28-d period (Figs. 5–7). More importantly, the sintered scaffolds were able to maintain their shape and structural integrity following 50% strain as shown by significantly greater energy dissipation. Exploring the connection between the XRD results (Figs. 8 and 9) and observed increases in UCS and compressive modulus reveals interesting correlations. As-spun PCL fiber exhibits significant crystalline XRD peaks at 21 and 23.751 2θ. However, following sintering (especially at 55 1C), these peaks were significantly reduced or nearly eliminated. Even more intriguing is the fact that PBS exposure caused these samples to recover a substantial degree of their as-spun crystallinity after 7 d of exposure (Fig. 10). This correlates to a direct increase in UCS and compressive moduli (Johnson et al., 2007; Lee et al., 2008; Bolgen et al., 2005; Coury et al., 1996; Gan et al., 1999). Longer-term exposures cause a substantial drop in crystallinity that parallels the observed property decreases. Sintering at temperatures just below the Tm of PCL resulted in significant decreases in XRD crystalline peak intensity likely due to the melting of poor quality crystals created during the electrospinning process (Lee et al., 2008). The 55 1C sample exhibited XRD peak intensities 20-fold lower than the as-spun condition. We suspect that the thermal exposure suppresses crystallinity but that the polymer chains retain the alignment associated with electrospinning and thus are entropically more susceptible to recrystallization. In regards to the decrease in mechanical properties observed following 1 d of in vitro exposure, although PCL is considered to be a hydrophobic polymer, the nature of it and many other hydrolytically unstable aliphatic esters is that they do allow water diffusion to take place. Once sufficient water is present within the polymeric structure, PCL is rapidly plasticized especially given its high surface area, nanoscaled form and becomes correspondingly less resistant to the application of mechanical stress. After only one day of exposure, the 55 1C samples recover a substantial fraction of their initial crystallinity. Seven days of exposure resulted in nearly full restoration of pre-sintering XRD crystalline peak intensities. The aqueous environment likely plays a role, diffusing water allowing already aligned polymer chains to rearrange more easily and reform crystals (Bolgen et al., 2005; Coury et al., 1996). Increased thermal exposures (56 and 58 1C) could lead to loss of polymer alignment and the potential for recrystallization. As a result of these increases in crystallinity, the mechanical properties of all three (55, 56, or 58 1C) sintered scaffolds displayed maximum values of UCS and modulus following 7 d of exposure.
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
In the absence of any in vitro deposition, the recovered crystallinity and the associated resistance to degradation provided the sintered PCL samples with the necessary integrity needed to maintain their initial mechanical properties up to at least 28 d of PBS exposure. In vivo studies have indicated that at least a two-week lag time in cartilage tissue development is necessary to provide sufficient mechanical properties for maturation (Jungebluth et al., 2012; Baiguera et al., 2010a, 2010b). Therefore, a scaffold that can utilize sintering to better maintain – or even improve – its mechanical integrity long term without significant degradation will provide a more suitable resorbable tissue engineering platform. Concerns regarding densification that could result in a loss of cell infiltration/adhesion can be readily addressed by either sintering in the presence of a gel like material to maintain the porosity and prevent shrinkage (Lee et al., 2008), or by inclusion of sintering-resistant compositions.
5.
Conclusions
Electrospun nanofibers have been used extensively in tissue engineering applications, due their inherent ability to provide physical dimensions and morphological aspects closely mimicking native tissue ECM (Gustafsson et al., 2012). Like tissue ECM, such engineered grafts provide the framework, support, and scaffolding for cellular in-growth, proliferation, differentiation, and eventual tissue maturation. However, many previous attempts to create fully resorbable, tissue engineered grafts have been met with mechanical failure following their exposure to biological milieu (Jungebluth et al., 2012). As an attempt to provide a fully resorbable tissue engineered scaffold, PCL was chosen to be electrospun due to its relative resistance (versus other resorbable polymers (Bryers et al., 2012)) to aqueous degradation. Sintering at temperatures just below Tm provides significant increases in the mechanical strength. However, as a result of aqueous exposures PCL polymer chains were able to crystallize to result in previously unreported increases in strength potentially enabling electrospun PCL to better maintain its initial mechanical properties out to 28 d. PCL's tendency to crystallize in response to aqueous exposure appears to increase both mechanical properties and microstructural integrity. Further advancements in the design of resorbable electrospun scaffolds will involve enhancing the structural and chemical aspects of the scaffold to provide a more suitable environment for cell integration and tissue development. While sintering can increase mechanical properties, the concomitant loss of porosity and available surface area that cells require for attachment and proliferation reveals the need for considerable improvement. Future developments must examine selective sintering, growth factor delivery, pore maintenance, and further enhancement of the structural integrity of resorbable electrospun nanofiber scaffolds.
Acknowledgments This work is supported by a research grant from the National Science Foundation under Grant No. EEC-0425626. Any opinions,
157
findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.
r e f e r e nc e s
Alexander, H., Brunski, J.B., Cooper, S.L., Hench, L.L., Hergenrother, R.W., Hoffman, A.S., Kohn, J., Langer, R., Peppas, N.A., Ratner, B.D., Shalaby, S.W., Visser, S.A., Yannas, I.V., 1996. Classes Mater. Used Med.. Baiguera, S., Macchiarini., P., 2011. Trachea. Baiguera, S., Jungebluth, P., Burns, A., Mavilia, C., Haag, J., De Coppi, P., Macchiarini., P., 2010a. Tissue engineered human tracheas for in vivo implantation. Biomaterials, 31. Baiguera, S., Birchall, M.A., Macchiarini., P., 2010b. Tissueengineered tracheal transplantation. Transplantation, 89. Baiguera, S., Lula, K.D., Macchiarini., P., 2011. Detergent-Enzym. Method. Bioeng. Hum. Airways. Baiguera, S., Del Gaudio, C., Jaus, M.O., Polizzi, L., Gonfiotti, A., Comin, C.E., Bianco, A., Ribatti, D., Taylor, D.A., Macchiarini, P., 2012. Long-term changes to in vitro preserved bioengineered human trachea and their implications for decellularized tissues. Biomaterials 33, 3662–3667. Bolgen, N., Menceloglu, Y.Z., Acatay, K., Vargel, I., Piskin, E., 2005. In vitro and in vivo degradation of non-woven materials made of poly(epsilon-caprolactone) nanofibers prepared by electrospinning under different conditions. J. Biomater. Sci. Polym. Ed., 16. Bryers, JD, Giachelli, CM, Ratner, BD, 2012. Engineering biomaterials to integrate and heal: the biocompatibility paradigm shifts. Biotechnol. Bioeng. 109, 1898–1911. Chen, Z.C.C., Ekaputra, A.K., Gauthaman, K., Adaikan, P.G., Yu, H., Hutmacher, D.W., 2008. In vitro and in vivo analysis of coelectrospun scaffolds made of medical grade poly(epsiloncaprolactone) and porcine collagen. J. Biomater. Sci. Polym. Ed., 19. Cooke, F.W., Lemons, J.E., Ratner, B.D., 1996. Prop. Mater. Courtney, T., Sacks, M.S., Stankus, J., Guan, J., Wagner, W.R., 2006. Design and analysis of tissue engineering scaffolds that mimic soft tissue mechanical anisotropy. Biomaterials 27, 3631–3638. Coury, A.J., Levy, R.J., McMillin, C.R., Pathak, Y., Ratner, B.D., Schoen, F.J., Williams, D.F., Williams, R.L., 1996. Degrad. Mater. Biol. Environ. Delaere, P.R., Vranckx, J.J., Meulemans, J., Vander Poorten, V., Segers, K., Van Raemdonck, D., De Leyn, P., Decaluwe, H., Dooms, C., Verleden, G., 2012. Learning curve in tracheal allotransplantation. Am. J. Transplant., 12. Drilling, S., Gaumer, J., Lannutti, J., 2009. Fabrication of burst pressure competent vascular grafts via electrospinning: effects of microstructure. J. Biomed. Mater. Res. Part A. 88A, 923–934. Fountain, H., 2012. Synthetic Windpipe Is Used to Replace Cancerous One. The New York TimesA12 (January 13, 2012). Gan, Z.H., Yu, D.H., Zhong, Z.Y., Liang, Q.Z., Jing, X.B., 1999. Enzymatic degradation of poly(epsilon-caprolactone)/poly(DLlactide) blends in phosphate buffer solution. Polymerization, 40. Gaumer, J., Prasad, A., Lee, D., Lannutti., J., 2009. Source-toground distance and the mechanical properties of electrospun fiber. Acta Biomater. 5, 1552–1561. Gustafsson, Y., Haag, J., Jungebluth, P., Lundin, V., Lim, M.L., Baiguera, S., Ajalloueian, F., Del Gaudio, C., Bianco, A., Moll, G., Sj├╢qvist, S., Lemon, G., Teixeira, A.I., Macchiarini, P., 2012. Viability and proliferation of rat MSCs on adhesion proteinmodified PET and PU scaffolds. Biomaterials 33, 8094 (103).
158
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
Heath, D.E., Lannutti, J.J., Cooper, S.L., 2010. Electrospun scaffold topography affects endothelial cell proliferation, metabolic activity, and morphology. J. Biomed. Mater. Res. Part A, 94A. Johnson, J., Ghosh, A., Lannutti, J., 2007. Microstructure-property relationships in a tissue-engineering scaffold. J. Appl. Polym. Sci. 104, 2919–2927. Johnson, J., Niehaus, A., Nichols, S., Lee, D., Koepsel, J., Anderson, D., Lannutti., J., 2009. Electrospun PCL in vitro: a microstructural basis for mechanical property changes. J. Biomater. Sci. Polym. Ed. 20, 467–481. Jungebluth, P., Alici, E., Baiguera, S., Le Blanc, K., Blomberg, P., Bozoky, B., Crowley, C., Einarsson, O., Grinnemo, K., Gudbjartsson, T., Le Guyader, S., Henriksson, G., Hermanson, O., Juto, J.E., Leidner, B., Lilja, T., Liska, J., Luedde, T., Lundin, V., Moll, G., Nilsson, B., Roderburg, C., Stromblad, S., Sutlu, T., Teixeira, A.I., Watz, E., Seifalian, A., Macchiarini, P., 2011. Tracheobronchial transplantation with a stem-cell-seeded bioartificial nanocomposite: a proof-of-concept study. Lancet, 378. Jungebluth, P., Moll, G., Baiguera, S., Macchiarini., P., 2012. Tissueengineered airway: a regenerative solution. Clin. Pharmacol. Ther., 91. Kai, D., Prabhakaran, M.P., Stahl, B., Eblenkamp, M., Wintermantel, E., Ramakrishna, S., 2012. Mechanical properties and in vitro behavior of nanofiber–hydrogel composites for tissue engineering applications. Nanotechnology 23, 095705. Kim, D.Y., Pyun, J., Choi, J.W., Kim, J., Lee, J.S., Shin, H.A., Kim, H., Lee, H., Min, B., Cha, H.E., Kim., C., 2010. Tissue-engineered allograft tracheal cartilage using fibrin/hyaluronan composite gel and its in vivo implantation. Laryngoscope 120, 30–38. Kobayashi, K., Suzuki, T., Nomoto, Y., Tada, Y., Miyake, M., Hazama, A., Wada, I., Nakamura, T., Omori., K., 2010. A tissueengineered trachea derived from a framed collagen scaffold, gingival fibroblasts and adipose-derived stem cells. Biomaterials 31, 4855–4863. Komura, M., Komura, H., Kanamori, Y., Tanaka, Y., Suzuki, K., Sugiyama, M., Nakahara, S., Kawashima, H., Hatanaka, A., Hoshi, K., Ikada, Y., Tabata, Y., Iwanaka, T., 2008. An animal model study for tissue-engineered trachea fabricated from a biodegradable scaffold using chondrocytes to augment repair of tracheal stenosis. J. Pediatr. Surg., 43. Komura, M., Komura, H., Kanamori, Y., Tanaka, Y., Ohatani, Y., Ishimaru, T., Sugiyama, M., Hoshi, K., Iwanaka, T., 2010. Study of mechanical properties of engineered cartilage in an in vivo
culture for design of a biodegradable scaffold. Int. J. Artif. Organs., 33. Lee, S.J., Oh, S.H., Liu, J., Soker, S., Atala, A., Yoo., J.J., 2008. The use of thermal treatments to enhance the mechanical properties of electrospun poly (epsilon-caprolactone) scaffolds. Biomaterials, 29. Liu, L., Wu, W., Tuo, X., Geng, W., Zhao, J., Wei, J., Yan, X., Yang, W., Li, L., Chen., F., 2010. Novel strategy to engineer trachea cartilage graft with marrow mesenchymal stem cell macroaggregate and hydrolyzable scaffold. Artif. Organs, 34. Ma, Z.W., He, W., Yong, T., Ramakrishna, S., 2005. Grafting of gelatin on electrospun poly(caprolactone) nanofibers to improve endothelial cell spreading and proliferation and to control cell orientation. Tissue Eng. 11, 1149–1158. Nam, J., Huang, Y., Agarwal, S., Lannutti, J., 2007. Improved cellular infiltration in electrospun fiber via engineered porosity. Tissue Eng. 13, 2249–2257. Nam, J., Johnson, J., Lannutti, J.J., Agarwal, S., 2011. Modulation of embryonic mesenchymal progenitor cell differentiation via control over pure mechanical modulus in electrospun nanofibers. Acta Biomater., 7. Pektok, E., Nottelet, B., Tille, J., Gurny, R., Kalangos, A., Moeller, M., Walpoth, B.H., 2008. Degradation and healing characteristics of small-diameter poly(epsilon-caprolactone) vascular grafts in the rat systemic arterial circulation. Circulation 118, 2563–2570. Stella, J.A., D’Amore, A., Wagner, W.R., Sacks, M.S., 2010. On the biomechanical function of scaffolds for engineering loadbearing soft tissues. Acta Biomater. 6, 2365–2381. Suh, S., Kim, J., Baek, C., Han, J., Kim, H., 2001. Replacement of a tracheal defect with autogenous mucosa lined tracheal prosthesis made from polypropylene mesh. ASAIO J. 47, 496–500. Tada, Y., Suzuki, T., Takezawa, T., Nomoto, Y., Kobayashi, K., Nakamura, T., Omori, K., 2008. Regeneration of tracheal epithelium utilizing a novel bipotential collagen scaffold. Ann. Otol. Rhinol. Laryngol., 117. Tada, Y, Takezawa, T, Tani, A, Nakamura, T, Omori, K, 2012. Collagen vitrigel scaffold for regenerative medicine of the trachea: experimental study and quantitative evaluation. Acta Otolaryngol, 132. Venugopal, J., Zhang, Y.Z., Ramakrishna, S., 2005. Fabrication of modified and functionalized polycaprolactone nanofibre scaffolds for vascular tissue engineering. Nanotechnology 16, 2138–2142.
Available online at www.sciencedirect.com
www.elsevier.com/locate/jmbbm
Research Paper
Recrystallization improves the mechanical properties of sintered electrospun polycaprolactone M. Tyler Nelsona, Lagnajit Pattanaikb, Marcia Allenb, Matthew Gerbichb, Kelvin Huxb, Matthew Allenc, John J. Lannuttib,n a
Department of Biomedical Engineering, Ohio State University, Columbus, OH 43210, United States Department of Materials Science and Engineering, Ohio State University, Columbus, OH 43210, United States c Department of Veterinary Clinical Sciences, Ohio State University, Columbus, OH 43210, United States b
ar t ic l e in f o
abs tra ct
Article history:
Background: Resorbable electrospun polycaprolactone (PCL) scaffolds for tissue reconstruc-
Received 12 August 2013
tion can provide physicians with an “off the shelf” product tailored to the patient's specific
Received in revised form
tissue architecture. However, many tissue-engineering platforms do not possess the
31 October 2013
necessary long-term mechanical stability needed to properly support tissue development.
Accepted 2 November 2013
Objective: Sintering has been explored as a means of altering the properties of electrospun
Available online 15 November 2013
PCL. However, crystallinity-driven changes in mechanical properties following thermal treatment have not been previously investigated. Methods: PCL nanofibers were produced by electrospinning and subsequently thermally sintered (at 55, 56 and 58 1C) to enhance their long-term mechanical integrity in response to representative biological milieux. Results: Scaffolds initially sintered at 56 1C displayed 6-fold increases in compressive strength and 3-fold increases in modulus, while displaying 10-fold increases in energy dissipation with increasing sintering temperature. Sintering just below the Tm resulted in amorphization of the 55 1C sample as indicated by the 20-fold lower XRD peak intensities. Although crystallinity is suppressed, the polymer chains likely retain chain alignment from electrospinning and are apparently highly susceptible to recrystallization. After only 1 d PBS exposure, the 55 1C samples recover a substantial fraction of the as-spun crystallinity; 7 d of exposure fully restores as-spun peak intensities. The mechanical properties of all three (55, 56, or 58 1C) scaffolds displayed peak values of compressive strength and modulus following 7 d exposure. Conclusion: In contrast with the current state-of-the-art which assumes that tissue engineering scaffolds only grow weaker following exposure, in these scaffolds maximum values of compressive strength and modulus were observed after 7 d of aqueous immersion. This suggests that polymeric recrystallization can be used to increase or optimize mechanical properties in vitro/in vivo. Scaffolds that increase their mechanical integrity during biological exposures constitute a new pathway enabling clinical advances. & 2013 Elsevier Ltd. All rights reserved.
n
Corresponding author. Tel./fax: þ1 614 688 3182. E-mail address: [email protected] (J.J. Lannutti).
1751-6161/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jmbbm.2013.11.004
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
1.
Introduction
Engineered tissue substitutes as bone, blood vessel, or skin grafts can provide a convenient alternative to allografts. However, many engineered substitutes have been hampered by losses of mechanical integrity requiring additional surgeries or, in more extreme cases, significant increases in mortality (Bryers et al., 2012; Johnson et al., 2009, 2007; Ma et al., 2005; Venugopal et al., 2005; Pektok et al., 2008; Drilling et al., 2009; Nam et al., 2007; Lee et al., 2008). Since the first successful decellularized donor reconstruction in 2008 many have suggested that this technology presents the best alternative for future scaffolds (Baiguera et al., 2011). However, given donor scarcity and degradation of mechanical properties resulting from the decellularization process (Baiguera et al., 2011), this pathway may not, in fact, be widely applicable. Other naturally-derived tissue engineered constructs have also struggled to provide the necessary mechanically stable, chemically inert environment allowing the desired cell/tissue maturation (Kim et al., 2010; Kobayashi et al., 2010; Komura et al., 2008, 2010; Liu et al., 2010; Tada et al., 2008, 2012). Synthetic tissue engineered constructs have long been proposed as a means of eliminating the need for donor tissues. Synthetic materials provide design-specific tailoring options making them ideal for tissue engineering applications. Many studies have shown the efficacy of producing porous, non-biodegradable, synthetic substitutes for reconstruction(Suh et al., 2001; Jungebluth et al., 2012; Gustafsson et al., 2012). In 2011, the first fully synthetic, non-resorbable tissue-engineered trachea substitute inoculated with mononuclear stem cells was implanted successfully in a human with no significant complications at 12-months follow up (Jungebluth et al., 2011). However, loss of mechanical integrity in these synthetics can render the scaffolds unsuited to tissue development (Jungebluth et al., 2011). Since then, much of the focus has been directed toward engineering substitutes capable of providing the initial mechanical support and stability needed for autologous tissue/cartilage maturation. To our knowledge, there have been no studies that utilize increases in mechanical properties during aqueous exposure to meet clinical needs. The fibrous microstructure and chemistry found in decellularized donor tissue provides a native environment promoting cellular attachment, proliferation, and tissue-directed differentiation crucial to the development of mechanically proficient TE cartilage (Baiguera et al., 2011, 2010a). Electrospinning offers tremendous potential for reproducing the unique ECM microstructure while allowing for the ability to tailor the mechanical properties of a scaffold using synthetic, natural, or blends of natural and synthetic polymers. Electrospun scaffolds can reproduce the fibrous microstructure, porosity, and micro/nano-scales of native tissue ECM allowing them to be personalized for a wide range of TE applications involving skin, bone, blood vessel, heart valve, and nerve target tissues (Drilling et al., 2009; Nam et al., 2007, 2011; Bolgen et al., 2005; Chen et al., 2008; Heath et al., 2010). In this study, thermally sintered electrospun biodegradable polycaprolactone (PCL) scaffolds were investigated to examine their mechanical stability and microstructure with the hope of providing a fully resorbable tissue-engineering
151
platform for tissue reconstruction. As an attempt to provide a fully resorbable tissue engineered scaffold, PCL was chosen to be electrospun due to its relative resistance (versus other resorbable polymers (Bryers et al., 2012)) to hydrolytic degradation. A key advantage over non-resorbable materials, such resorbable scaffolds provide surgeons with a construct capable of growing and maturing with pediatric patients in vivo while slowly degrading (Bryers et al., 2012). To introduce a potentially wide-ranging means of improving mechanical properties, thermally driven sintering was investigated to gauge its effects on compressive strength, modulus, and energy dissipation. Following the observed improvements in mechanical properties, exposures to aqueous biological milieu revealed a previously unreported phenomenon: amorphization followed by recrystallization. Sintering temperatures near the melting temperature (Tm) of electrospun PCL produced microstructures and crystalline behavior capable of maintaining mechanical properties similar to native tissue. As expected, however, the process greatly alters the underlying microstructure. Future investigations must be directed towards selective sintering of electrospun scaffolds by incorporating thermally inert regions capable of maintaining the highly porous electrospun microstructure necessary for cellular infiltration, nutrient in-flow, release of metabolic products and tissue maturation.
2.
Materials and methods
2.1.
Electrospinning
14 wt/wt% polycaprolactone (PCL, ffi80,000 Mw, SigmaAldrich, St. Louis, MO) in acetone (99% purity, Sigma-Aldrich, St. Louis, MO) solution was prepared by mixing at 55 1C. Once the solution was fully dissolved and transparent it was added to a 60 mL syringe (Becton-Dickinson, Franklin Lakes, NJ) fitted with a 20 gauge blunt tipped, stainless steel needle (Nordson EFD, West Lake, OH). A solution filled syringe was then placed in a syringe pump set to a flow rate of 15 ml/h. Using a working distance from the needle tip to the grounded aluminum collector of 20 cm (Gaumer et al., 2009), a positive, DC voltage of 25 kV was applied to the needle tip. The collector was a 2.5 cm diameter by 10.2 cm long cylindrical, aluminum mandrel rotated at a linear velocity of 5 m/s. In order to acquire 1.5 mm-thick electrospun sample, two 60 mL syringes dispensed flow continuously for a 4-h period. Upon completion of electrospinning, a 1.5 mm thick by 10.2 cm long by 2.5 cm diameter hollow tube formed on the mandrel. For testing purposes, 1 cm wide rings were cut from the electrospun PCL tube. In total, 35 such tubes were produced to enable a statistically significant test.
2.2.
Thermal sintering
As-electrospun PCL ring samples were sintered at 55, 56, or 58 1C for 24 h using a standard recirculating water bath. Samples were first chemically isolated from the surrounding water inside securely sealed screw-capped glass vials. To preserve a constant 2.5 cm diameter, the ring samples were placed on a custom 2.5 cm diameter stainless steel rod during
152
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
the sintering process. After sintering, these rings were removed from the water bath and placed in a plastic bag at room temperature 25 1C) until use. As-spun PCL rings were used as a control, and were placed in sealed plastic bags immediately after electrospinning.
2.3.
Mechanical testing
Compressive mechanical testing was conducted on the asspun and sintered electrospun PCL rings using a load frame (Type R, TestResources Inc, Shakopee, MN). 50% compressive strain (1.25 cm total deformation) was applied to the rings at a crosshead rate of 25.0 mm/min. The loading and unloading profiles of these rings was then recorded. Samples were tested at room conditions 25 1C) in air, and in phosphate buffered saline (PBS) containing 0.1% sodium azide at a temperature of 37 1C maintained by a heating coil and constant stirring. Sodium azide was added to prevent microbial or fungal infection that could result in the destruction and/or degradation of electrospun PCL rings imparting nonrandom variation in the observed mechanical properties. The ultimate compressive strength (UCS) was calculated as the maximum load recorded at 50% strain, and the compressive modulus was measured from the slope of the loading curve prior to detectable yielding. Energy dissipation was also recorded for each of the sintered conditions and the asspun PCL scaffolds by measuring the area in between the loading and unloading hysteresis curves using Matlab mathematic toolbox software (Mathworks, Natick, MA).
2.4.
Extended exposure testing
Rings from each of the sintered conditions and as-spun rings were subjected to exposure to 37 1C PBS over a 28-d period (Johnson et al., 2009). Ring samples were placed directly into polystyrene conical tubes (Becton-Dickinson, Franklin Lakes, NJ) filled with 40 ml of PBS solution containing 0.1% sodium azide. These rings were then kept in the solution for 1, 7, 14, or 28 days at 37 1C in an incubator prior to being mechanically tested under compression. Rings were removed from the PBS solution at the end of each time point and tested immediately. All samples remained saturated with PBS solution during the test. Results are displayed as the mean values with 95% interval of confidence for ultimate compressive strength (UCS), compressive modulus, and energy dissipation. Significant factor interaction between sintering temperature and exposure time was detected by 2-way ANOVA statistical analysis (Minitab 16); interaction plots for each of the response variables have been inserted.
2.5.
Scanning electron microscope (SEM) analysis
Analysis of microstructure was conducted for as-spun and all sintered conditions prior to and after PBS solution exposure. All samples were fully dried for 24 h at atmospheric room temperature conditions prior to SEM preparation. SEM samples were cut from the electrospun PCL rings and mounted on aluminum studs (SPI Supplies/Structure Probe Inc., West Chester, PA) covered with carbon tape (SPI Supplies/Structure Probe Inc., West Chester, PA). Samples were then coated with
15 nm of gold using a gold sputter coater at a plasma current of 15 mA. An environmental scanning electron microscope (ESEM, Philips XL-30 ESEM-FEG, Surrey, UK) operated at an accelerating voltage of 15 kV was used to image the as-spun and sintered samples. Porosity and Pore-size measurements were extracted from five different SEM images for each condition (As-spun, 55, 56, and 58 1C) using NIH ImageJ software. To determine porosity SEM images of the scaffolds were thresholded to display only the void areas of the image and the resulting areas were measured as a percentage of the total area to estimate scaffold porosity. Pore-size was quantified by measuring the long dimension and short dimension of space between fiber–fiber boundaries and taking the average length.
2.6.
X-ray diffraction
Using a Bragg-Brentano configuration, X-ray diffraction patterns were acquired before and after 7 days of PBS exposure for sintered and as-spun electrospun PCL. Rigaku Ultima-III diffractometer (Rigaku Americas Corp., Allison Park, PA) with a Cu Kα source and a diffracted beam monochromator analyzed electrospun specimens mounted on aluminum disks over a 2θ angle range of 19 to 251 with a step-size of 0.041. Each sample was electrospun for equal time to produce a 1.5 mm thick disc. To ensure consistent X-ray absorption, the mass equivalency of each sample verified to within 0.025 g prior to XRD. The disks were rotated at 120 rpm during data acquisition with an integration time at each step of 24 s. A constant linear background was used in the profile fitting to eliminate potential background noise and scattering when comparing the patterns from different samples. In addition, 55 1C sintered PCL samples were analyzed using XRD before and after 1 d and 28 d of exposure.
2.7.
Statistical analysis
All results are displayed as mean 795% confidence interval calculated using Minitab 16 statistical software. Mechanical testing results were acquired using a single factor random experimental design model analyzed using a one-way ANOVA and determining statistical significant differences via a Tukey's test with a 95% confidence interval. Exposure testing results were analyzed using 2-way ANOVA and a Tukey's full factor interaction test with a 95% confidence interval.
3.
Results
3.1.
Electrospun microstructure
SEM images of the electrospun PCL microstructure for asspun and each of the sintered conditions are displayed in Fig. 1. As-spun PCL fiber displays cylindrical, individual fibers and minimal fiber–fiber connections providing a highly porous microstructure with extensive pore interconnectivity (Fig. 1: As-spun). With increasing sintering temperature the nanofiber morphology exhibits more fiber–fiber connections, bonding, and loss of porosity (Figs. 1 and 2(A and B)). Samples sintered at 55 1C depict increases in fiber diameter, and increased fiber–fiber bonding as compared to as-spun PCL
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
153
Fig. 1 – SEM images of (A) as-spun and (B) 55 1C, (C) 56 1C, and (D) 58 1C sintered electrospun PCL nanofibers. (A) Depicts independent fiber morphologies, free of fiber–fiber connections. With increasing temperature (B) to (D), the images exhibit more fiber–fiber bonding, a reduction in porosity, and less independent fiber.
Fig. 2 – Porosity and pore-size quantification from SEM images depicts that with increasing sintering temperature significant reductions in free volume or porosity are observed (A), however the pore-size remains constant up to the experimental melting temperature (58 1C) (B). n Denotes statistical outliers, and # denotes statistically significant results (po0.05).
samples. Significant alterations to the electrospun fiber morphology and scaffold microstructure are not observed until the sintering temperature exceeded 55 1C, as is the case for samples sintered at 56 or 58 1C. Very little of the original electrospun fiber morphology is retained after scaffolds are sintered at 56 1C as severe fiber–fiber bonding and significant loss of the open, highly porous microstructure are observed. 58 1C-sintered scaffolds exhibit a completely closed pore, dense, film-like microstructure. As-spun and 55 1C sintered samples maintain the unique electrospun fiber microstructure that more closely resembles this natural microstructure.
electrospun PCL samples at 50% strain are displayed in Figs. 3 and 4, respectively. The UCS and compressive modulus of electrospun PCL samples significantly increase with sintering temperature. Brief exposures (less than 1 min) to PBS (as a biological testing environment) demonstrated no significant differences versus testing under dry, atmospheric room conditions (data not shown). Fig. 5 presents energy dissipation data for each of the electrospun PCL scaffold conditions. In a very similar manner, energy dissipation increased 40-fold versus thermal exposure.
3.3. 3.2.
Exposure
Mechanical properties
The ultimate compressive strength (UCS) and compressive modulus for each of the sintered conditions and the as-spun
To evaluate the long-term mechanical integrity of sintered electrospun PCL following exposure to aqueous environments, both as-spun and sintered samples were exposed to
154
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
Fig. 3 – Ultimate compressive stress versus electrospun PCL thermal processing condition. Samples sintered at 56 or 58 1C exhibited significantly greater UCS values as compared to as-spun or samples sintered at 55 1C. n Denotes statistical outliers, and # denotes statistically significant results (po0.05).
Fig. 4 – Compressive modulus stress versus electrospun PCL thermal processing condition. The modulus of PCL nanofibers sintered at 56 1C were 6-fold greater and those sintered at 58 1C exhibited 18-fold greater values as compared to as-spun or samples sintered at 55 1C. n Denotes statistical outliers, and # denotes statistically significant results (po0.05).
PBS for a period of 28 d. Fig. 6 displays the UCS for as-spun and sintered samples tested following 1, 7, 14, or 28 d of exposure. As-spun, 55, and 56 1C sintered PCL samples displayed significant decreases in UCS after 1 day of exposure, followed by a significant increase after 7 days. Compressive moduli exhibited trends similar to – but not as pronounced as – those of UCS (Fig. 7). Both UCS and compressive moduli for samples sintered at 58 1C maintained these significant increases at day 14 while for all samples 28 d of exposure resulted in a decrease of mechanical properties to preexposure levels. Energy dissipation versus 28 d of exposure decreased for all samples versus the as-electrospun zero exposure condition (Fig. 8). A 55 and 56 1C sintered samples displayed significant increases in energy dissipation after 7 d, followed by a sharp reduction at 14 d that brought the values back to pre-exposure levels.
Fig. 5 – Bar-plot of energy dissipation versus electrospun PCL thermal processing condition. Energy dissipation of sintered PCL nanofibers was greater than the as-spun condition following for all thermal exposures. 56 1C were 20-fold and 58 1C were 30-fold greater then as-spun PCL nanofiber. n Denotes statistical outliers, and # denotes statistically significant results (po0.05).
Fig. 6 – Ultimate compressive strength (UCS) versus exposure for as-spun and sintered electrospun PCL nanofibers. UCS always increased with sintering temperature. All samples experienced a decrease in strength after 1 day followed by a significant increase after 7 days of exposure. UCS for all samples, after 28-d, was equal or nearly equal to as-spun conditions. n Denotes statistical outliers, and # denotes statistically significant results with respect to the interaction between sintering temperature and exposure time (po0.05).
3.4.
X-ray diffraction
The XRD patterns for as-spun and sintered samples were acquired before and after PBS exposure. Fig. 9 displays the patterns for each of the sintering conditions (55, 56, and 58 1C) and as-spun PCL nanofiber before and after 7 d of PBS exposure. In all cases sintering initially greatly reduced the as-spun crystalline peak intensity. Conversely, 7 d of exposure to PBS resulted in significant increases in crystalline peak intensity. Samples sintered at 55 1C presented the most dramatic decrease in crystalline peak intensity after sintering, followed by the greatest rebound in crystalline peak
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
Fig. 7 – Compressive moduli versus exposure time for as-spun and sintered electrospun PCL nanofibers. Modulus results follow trends similar to that of UCS results, indicating that with increasing temperature samples became stiffer. n Denotes statistical outliers, and # denotes statistically significant results with respect to the interaction between sintering temperature and exposure time (po0.05).
Fig. 8 – Energy dissipation versus exposure time for as-spun and sintered electrospun PCL nanofibers. Sintered samples displayed greater energy dissipation values as compared to as-spun PCL nanofibers for any exposure time. n Denotes statistical outliers, and # denotes statistically significant results with respect to the interaction between sintering temperature and exposure time (po0.05).
intensity after PBS exposure. To determine if this effect occurs following much shorter exposures, a 1-d PBS exposure followed by XRD analysis was conducted. Fig. 10 displays the XRD patterns for as-spun PCL before and after sintering at 55 1C, and the 55 1C specimen following 1 and 28 days of PBS exposure. One day of exposure significantly increases crystalline peak intensities at 21.5 and 23.751 2Θ. Interestingly, lower sintering temperatures resulted in the most significant decreases in crystalline peak intensity with the lowest temperature (55 1C) producing nearly amorphous PCL while higher temperature exposures (56 and 58 1C) display considerably more crystallinity (Fig. 9). Increases in crystalline peak intensity were seen following PBS exposure (even for the as-spun condition) and decreased with increasing sintering
155
Fig. 9 – XRD profiles of as-spun and sintered (denoted by temperature) PCL nanofiber before and after 7-d of PBS exposure (denoted by temperature and PBS). Thermal sintering resulted in significant decreases in crystalline peak intensities, with 55 1C sintered samples exhibiting the greatest loss. After subsequent exposure to PBS, all samples exhibited an increase of crystalline peak intensity.
Fig. 10 – Presents the XRD profiles of electrospun PCL nanofiber before and after 55 1C sintering and 1- and 28 d of PBS exposure. As observed previously, sintering resulted in loss of crystalline peaks, nearly exhibiting completely amorphous behavior. After only 1-d of exposure, recrystallization was observed, resulting in significant crystalline peak reformation. Following 7 d of exposure, recrystallization reaches a substantial fraction of as-spun values; 28 d of exposure greatly decreases crystallinity. temperature. The 55 1C-sintered samples displayed the greatest crystalline peak intensity rebound. Conversely, the 28-d exposure resulted a substantial decrease in crystallinity versus the 7-d exposure condition.
4.
Discussion
With the emergence of tissue engineering, regenerative medicine now offers options for tissue replacement centered
156
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
on the concept of “personalized medicine” (Baiguera et al., 2011, 2010a, 2010b; Jungebluth et al., 2012; Baiguera and Macchiarini. 2011; Delaere et al., 2012). Recent clinical successes following implantation of electrospun, non-resorbable engineered tracheal constructs for total laryngotracheal transplantation has triggered a concomitant resurgence of interest in tissue engineered organs (Fountain, 2012). Attempts to create resorbable tissue engineered scaffold substitutes have met with limited clinical acceptance partly due to risks associated with the lack of convincing, long-term mechanical property data (Jungebluth et al., 2012). As is often stated but not yet demonstrated, the rate of biomaterial degradation should match the rate of new tissue formation to compensate for the associated decreases in mechanical properties (Kai et al., 2012; Alexander et al., 1996; Cooke et al., 1996; Coury et al., 1996). However, if tissue formation takes place purely in vitro it is usually beset by over-deposition of ECM resulting in disorganized, weak scaffolding (Kim et al., 2010; Komura et al., 2010; Baiguera et al., 2012). Long term, many researchers have established that remodeling of a pre-existing ECM structure provides in vivo-like mechanical integrity and organization capable of promoting tissue maturation to meet the desired organ's functional needs (Jungebluth et al., 2012; Courtney et al., 2006). Given that the initial properties of such scaffolds must be optimized to meet clinical standards, this study characterizes the effects of thermal sintering on the microstructural and mechanical properties of tubular, electrospun polycaprolactone. Tubular electrospun PCL has been widely explored as small diameter (o6 mm) blood vessels appropriate for coronary artery replacement (Ma et al., 2005; Venugopal et al., 2005; Pektok et al., 2008; Drilling et al., 2009). Although sintering has been explored previously to alter the microstructure and properties of electrospun PCL (Nam et al., 2007; Johnson et al., 2007; Lee et al., 2008), little has been done to characterize the changes in crystallinity either after thermal treatment or the exposure of such treated microstructures to aqueous environments. This overlooks the important effects that changing crystallinity clearly might have on overall mechanical properties. The unique fibrous microstructure and extensive ( Z80%) inter-connected porosity of these electrospun PCL matrices (Fig. 1A) closely mimics the morphology and architecture of native ECM. Increasing sintering temperature resulted in the expected losses of independent fiber morphology and the unique porous microstructure. Also interesting is the definite effects that sintering will have on the ability of polymer nanofibers to rotate and translate in response to applied mechanical stresses (Johnson et al., 2007). While the sinteringdriven transition to a more highly point-bonded structure (Johnson et al., 2007) clearly imparts considerable gains in mechanical integrity (Figs. 4 and 5) by the creation of a more “net-like“ structure resisting fiber–fiber motion, these benefits come at the cost of the large scale (potentially hundreds of microns) rearrangement of which as-electrospun structures are capable (Johnson et al., 2007; Stella et al., 2010) given their lack of regular, repeating structure. Macroscopic benefits measurable by current technology do not reflect changes at the micro scale properties that may or may not be significant in controlling cell response (Stella et al., 2010).
The 55 1C scaffold displayed very little fiber–fiber bonding resulting from sintering compared to the 56 or 58 1C sintered scaffolds. As a result, the fibers within the 55 1C scaffold were able to undergo relatively large amounts of interfiber motion (Ma et al., 2005; Venugopal et al., 2005; Pektok et al., 2008) allowing the overall scaffold to deform easily with little energy dissipation. While the 58 1C-sintered PCL samples clearly exhibit undesirable changes in microstructure and morphology, samples sintered at 56 1C still possess considerable porosity capable of promoting efficient nutrient transport and cell viability. The associated UCS is 6-fold and the modulus 3-fold greater than the 55 1C samples and nearly 15-fold greater versus the as-spun condition. Long-term mechanical stability of sintered samples was maintained over a 28-d period (Figs. 5–7). More importantly, the sintered scaffolds were able to maintain their shape and structural integrity following 50% strain as shown by significantly greater energy dissipation. Exploring the connection between the XRD results (Figs. 8 and 9) and observed increases in UCS and compressive modulus reveals interesting correlations. As-spun PCL fiber exhibits significant crystalline XRD peaks at 21 and 23.751 2θ. However, following sintering (especially at 55 1C), these peaks were significantly reduced or nearly eliminated. Even more intriguing is the fact that PBS exposure caused these samples to recover a substantial degree of their as-spun crystallinity after 7 d of exposure (Fig. 10). This correlates to a direct increase in UCS and compressive moduli (Johnson et al., 2007; Lee et al., 2008; Bolgen et al., 2005; Coury et al., 1996; Gan et al., 1999). Longer-term exposures cause a substantial drop in crystallinity that parallels the observed property decreases. Sintering at temperatures just below the Tm of PCL resulted in significant decreases in XRD crystalline peak intensity likely due to the melting of poor quality crystals created during the electrospinning process (Lee et al., 2008). The 55 1C sample exhibited XRD peak intensities 20-fold lower than the as-spun condition. We suspect that the thermal exposure suppresses crystallinity but that the polymer chains retain the alignment associated with electrospinning and thus are entropically more susceptible to recrystallization. In regards to the decrease in mechanical properties observed following 1 d of in vitro exposure, although PCL is considered to be a hydrophobic polymer, the nature of it and many other hydrolytically unstable aliphatic esters is that they do allow water diffusion to take place. Once sufficient water is present within the polymeric structure, PCL is rapidly plasticized especially given its high surface area, nanoscaled form and becomes correspondingly less resistant to the application of mechanical stress. After only one day of exposure, the 55 1C samples recover a substantial fraction of their initial crystallinity. Seven days of exposure resulted in nearly full restoration of pre-sintering XRD crystalline peak intensities. The aqueous environment likely plays a role, diffusing water allowing already aligned polymer chains to rearrange more easily and reform crystals (Bolgen et al., 2005; Coury et al., 1996). Increased thermal exposures (56 and 58 1C) could lead to loss of polymer alignment and the potential for recrystallization. As a result of these increases in crystallinity, the mechanical properties of all three (55, 56, or 58 1C) sintered scaffolds displayed maximum values of UCS and modulus following 7 d of exposure.
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
In the absence of any in vitro deposition, the recovered crystallinity and the associated resistance to degradation provided the sintered PCL samples with the necessary integrity needed to maintain their initial mechanical properties up to at least 28 d of PBS exposure. In vivo studies have indicated that at least a two-week lag time in cartilage tissue development is necessary to provide sufficient mechanical properties for maturation (Jungebluth et al., 2012; Baiguera et al., 2010a, 2010b). Therefore, a scaffold that can utilize sintering to better maintain – or even improve – its mechanical integrity long term without significant degradation will provide a more suitable resorbable tissue engineering platform. Concerns regarding densification that could result in a loss of cell infiltration/adhesion can be readily addressed by either sintering in the presence of a gel like material to maintain the porosity and prevent shrinkage (Lee et al., 2008), or by inclusion of sintering-resistant compositions.
5.
Conclusions
Electrospun nanofibers have been used extensively in tissue engineering applications, due their inherent ability to provide physical dimensions and morphological aspects closely mimicking native tissue ECM (Gustafsson et al., 2012). Like tissue ECM, such engineered grafts provide the framework, support, and scaffolding for cellular in-growth, proliferation, differentiation, and eventual tissue maturation. However, many previous attempts to create fully resorbable, tissue engineered grafts have been met with mechanical failure following their exposure to biological milieu (Jungebluth et al., 2012). As an attempt to provide a fully resorbable tissue engineered scaffold, PCL was chosen to be electrospun due to its relative resistance (versus other resorbable polymers (Bryers et al., 2012)) to aqueous degradation. Sintering at temperatures just below Tm provides significant increases in the mechanical strength. However, as a result of aqueous exposures PCL polymer chains were able to crystallize to result in previously unreported increases in strength potentially enabling electrospun PCL to better maintain its initial mechanical properties out to 28 d. PCL's tendency to crystallize in response to aqueous exposure appears to increase both mechanical properties and microstructural integrity. Further advancements in the design of resorbable electrospun scaffolds will involve enhancing the structural and chemical aspects of the scaffold to provide a more suitable environment for cell integration and tissue development. While sintering can increase mechanical properties, the concomitant loss of porosity and available surface area that cells require for attachment and proliferation reveals the need for considerable improvement. Future developments must examine selective sintering, growth factor delivery, pore maintenance, and further enhancement of the structural integrity of resorbable electrospun nanofiber scaffolds.
Acknowledgments This work is supported by a research grant from the National Science Foundation under Grant No. EEC-0425626. Any opinions,
157
findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.
r e f e r e nc e s
Alexander, H., Brunski, J.B., Cooper, S.L., Hench, L.L., Hergenrother, R.W., Hoffman, A.S., Kohn, J., Langer, R., Peppas, N.A., Ratner, B.D., Shalaby, S.W., Visser, S.A., Yannas, I.V., 1996. Classes Mater. Used Med.. Baiguera, S., Macchiarini., P., 2011. Trachea. Baiguera, S., Jungebluth, P., Burns, A., Mavilia, C., Haag, J., De Coppi, P., Macchiarini., P., 2010a. Tissue engineered human tracheas for in vivo implantation. Biomaterials, 31. Baiguera, S., Birchall, M.A., Macchiarini., P., 2010b. Tissueengineered tracheal transplantation. Transplantation, 89. Baiguera, S., Lula, K.D., Macchiarini., P., 2011. Detergent-Enzym. Method. Bioeng. Hum. Airways. Baiguera, S., Del Gaudio, C., Jaus, M.O., Polizzi, L., Gonfiotti, A., Comin, C.E., Bianco, A., Ribatti, D., Taylor, D.A., Macchiarini, P., 2012. Long-term changes to in vitro preserved bioengineered human trachea and their implications for decellularized tissues. Biomaterials 33, 3662–3667. Bolgen, N., Menceloglu, Y.Z., Acatay, K., Vargel, I., Piskin, E., 2005. In vitro and in vivo degradation of non-woven materials made of poly(epsilon-caprolactone) nanofibers prepared by electrospinning under different conditions. J. Biomater. Sci. Polym. Ed., 16. Bryers, JD, Giachelli, CM, Ratner, BD, 2012. Engineering biomaterials to integrate and heal: the biocompatibility paradigm shifts. Biotechnol. Bioeng. 109, 1898–1911. Chen, Z.C.C., Ekaputra, A.K., Gauthaman, K., Adaikan, P.G., Yu, H., Hutmacher, D.W., 2008. In vitro and in vivo analysis of coelectrospun scaffolds made of medical grade poly(epsiloncaprolactone) and porcine collagen. J. Biomater. Sci. Polym. Ed., 19. Cooke, F.W., Lemons, J.E., Ratner, B.D., 1996. Prop. Mater. Courtney, T., Sacks, M.S., Stankus, J., Guan, J., Wagner, W.R., 2006. Design and analysis of tissue engineering scaffolds that mimic soft tissue mechanical anisotropy. Biomaterials 27, 3631–3638. Coury, A.J., Levy, R.J., McMillin, C.R., Pathak, Y., Ratner, B.D., Schoen, F.J., Williams, D.F., Williams, R.L., 1996. Degrad. Mater. Biol. Environ. Delaere, P.R., Vranckx, J.J., Meulemans, J., Vander Poorten, V., Segers, K., Van Raemdonck, D., De Leyn, P., Decaluwe, H., Dooms, C., Verleden, G., 2012. Learning curve in tracheal allotransplantation. Am. J. Transplant., 12. Drilling, S., Gaumer, J., Lannutti, J., 2009. Fabrication of burst pressure competent vascular grafts via electrospinning: effects of microstructure. J. Biomed. Mater. Res. Part A. 88A, 923–934. Fountain, H., 2012. Synthetic Windpipe Is Used to Replace Cancerous One. The New York TimesA12 (January 13, 2012). Gan, Z.H., Yu, D.H., Zhong, Z.Y., Liang, Q.Z., Jing, X.B., 1999. Enzymatic degradation of poly(epsilon-caprolactone)/poly(DLlactide) blends in phosphate buffer solution. Polymerization, 40. Gaumer, J., Prasad, A., Lee, D., Lannutti., J., 2009. Source-toground distance and the mechanical properties of electrospun fiber. Acta Biomater. 5, 1552–1561. Gustafsson, Y., Haag, J., Jungebluth, P., Lundin, V., Lim, M.L., Baiguera, S., Ajalloueian, F., Del Gaudio, C., Bianco, A., Moll, G., Sj├╢qvist, S., Lemon, G., Teixeira, A.I., Macchiarini, P., 2012. Viability and proliferation of rat MSCs on adhesion proteinmodified PET and PU scaffolds. Biomaterials 33, 8094 (103).
158
journal of the mechanical behavior of biomedical materials 30 (2014) 150 –158
Heath, D.E., Lannutti, J.J., Cooper, S.L., 2010. Electrospun scaffold topography affects endothelial cell proliferation, metabolic activity, and morphology. J. Biomed. Mater. Res. Part A, 94A. Johnson, J., Ghosh, A., Lannutti, J., 2007. Microstructure-property relationships in a tissue-engineering scaffold. J. Appl. Polym. Sci. 104, 2919–2927. Johnson, J., Niehaus, A., Nichols, S., Lee, D., Koepsel, J., Anderson, D., Lannutti., J., 2009. Electrospun PCL in vitro: a microstructural basis for mechanical property changes. J. Biomater. Sci. Polym. Ed. 20, 467–481. Jungebluth, P., Alici, E., Baiguera, S., Le Blanc, K., Blomberg, P., Bozoky, B., Crowley, C., Einarsson, O., Grinnemo, K., Gudbjartsson, T., Le Guyader, S., Henriksson, G., Hermanson, O., Juto, J.E., Leidner, B., Lilja, T., Liska, J., Luedde, T., Lundin, V., Moll, G., Nilsson, B., Roderburg, C., Stromblad, S., Sutlu, T., Teixeira, A.I., Watz, E., Seifalian, A., Macchiarini, P., 2011. Tracheobronchial transplantation with a stem-cell-seeded bioartificial nanocomposite: a proof-of-concept study. Lancet, 378. Jungebluth, P., Moll, G., Baiguera, S., Macchiarini., P., 2012. Tissueengineered airway: a regenerative solution. Clin. Pharmacol. Ther., 91. Kai, D., Prabhakaran, M.P., Stahl, B., Eblenkamp, M., Wintermantel, E., Ramakrishna, S., 2012. Mechanical properties and in vitro behavior of nanofiber–hydrogel composites for tissue engineering applications. Nanotechnology 23, 095705. Kim, D.Y., Pyun, J., Choi, J.W., Kim, J., Lee, J.S., Shin, H.A., Kim, H., Lee, H., Min, B., Cha, H.E., Kim., C., 2010. Tissue-engineered allograft tracheal cartilage using fibrin/hyaluronan composite gel and its in vivo implantation. Laryngoscope 120, 30–38. Kobayashi, K., Suzuki, T., Nomoto, Y., Tada, Y., Miyake, M., Hazama, A., Wada, I., Nakamura, T., Omori., K., 2010. A tissueengineered trachea derived from a framed collagen scaffold, gingival fibroblasts and adipose-derived stem cells. Biomaterials 31, 4855–4863. Komura, M., Komura, H., Kanamori, Y., Tanaka, Y., Suzuki, K., Sugiyama, M., Nakahara, S., Kawashima, H., Hatanaka, A., Hoshi, K., Ikada, Y., Tabata, Y., Iwanaka, T., 2008. An animal model study for tissue-engineered trachea fabricated from a biodegradable scaffold using chondrocytes to augment repair of tracheal stenosis. J. Pediatr. Surg., 43. Komura, M., Komura, H., Kanamori, Y., Tanaka, Y., Ohatani, Y., Ishimaru, T., Sugiyama, M., Hoshi, K., Iwanaka, T., 2010. Study of mechanical properties of engineered cartilage in an in vivo
culture for design of a biodegradable scaffold. Int. J. Artif. Organs., 33. Lee, S.J., Oh, S.H., Liu, J., Soker, S., Atala, A., Yoo., J.J., 2008. The use of thermal treatments to enhance the mechanical properties of electrospun poly (epsilon-caprolactone) scaffolds. Biomaterials, 29. Liu, L., Wu, W., Tuo, X., Geng, W., Zhao, J., Wei, J., Yan, X., Yang, W., Li, L., Chen., F., 2010. Novel strategy to engineer trachea cartilage graft with marrow mesenchymal stem cell macroaggregate and hydrolyzable scaffold. Artif. Organs, 34. Ma, Z.W., He, W., Yong, T., Ramakrishna, S., 2005. Grafting of gelatin on electrospun poly(caprolactone) nanofibers to improve endothelial cell spreading and proliferation and to control cell orientation. Tissue Eng. 11, 1149–1158. Nam, J., Huang, Y., Agarwal, S., Lannutti, J., 2007. Improved cellular infiltration in electrospun fiber via engineered porosity. Tissue Eng. 13, 2249–2257. Nam, J., Johnson, J., Lannutti, J.J., Agarwal, S., 2011. Modulation of embryonic mesenchymal progenitor cell differentiation via control over pure mechanical modulus in electrospun nanofibers. Acta Biomater., 7. Pektok, E., Nottelet, B., Tille, J., Gurny, R., Kalangos, A., Moeller, M., Walpoth, B.H., 2008. Degradation and healing characteristics of small-diameter poly(epsilon-caprolactone) vascular grafts in the rat systemic arterial circulation. Circulation 118, 2563–2570. Stella, J.A., D’Amore, A., Wagner, W.R., Sacks, M.S., 2010. On the biomechanical function of scaffolds for engineering loadbearing soft tissues. Acta Biomater. 6, 2365–2381. Suh, S., Kim, J., Baek, C., Han, J., Kim, H., 2001. Replacement of a tracheal defect with autogenous mucosa lined tracheal prosthesis made from polypropylene mesh. ASAIO J. 47, 496–500. Tada, Y., Suzuki, T., Takezawa, T., Nomoto, Y., Kobayashi, K., Nakamura, T., Omori, K., 2008. Regeneration of tracheal epithelium utilizing a novel bipotential collagen scaffold. Ann. Otol. Rhinol. Laryngol., 117. Tada, Y, Takezawa, T, Tani, A, Nakamura, T, Omori, K, 2012. Collagen vitrigel scaffold for regenerative medicine of the trachea: experimental study and quantitative evaluation. Acta Otolaryngol, 132. Venugopal, J., Zhang, Y.Z., Ramakrishna, S., 2005. Fabrication of modified and functionalized polycaprolactone nanofibre scaffolds for vascular tissue engineering. Nanotechnology 16, 2138–2142.
