ARTICLE X-ray Phase Contrast Imaging of Calcified Tissue and Biomaterial Structure in Bioreactor Engineered Tissues Alyssa A. Appel,1,2 Jeffery C. Larson,1,2 Alfred B. Garson III,3 Huifeng Guan,3 Zhong Zhong,4 Bao-Ngoc B. Nguyen,5 John P. Fisher,5 Mark A. Anastasio,3 Eric M. Brey1,2 1
Department of Biomedical Engineering, Illinois Institute of Technology, 3255 South Dearborn St, Chicago, Illinois 60616 2 Research Services, Edward Hines Jr. VA Hospital, 5000 S. 5th Avenue, Hines, Illinois 60141; telephone: þ312-567-5098; fax: þ312-567-5707; e-mail:
[email protected] 3 Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, Missouri 4 National Synchrotron Light Source, Brookhaven National Laboratory, Upton, New York 5 Fischell Department of Bioengineering, University of Maryland, College Park, Maryland
ABSTRACT: Tissues engineered in bioreactor systems have been used clinically to replace damaged tissues and organs. In addition, these systems are under continued development for many tissue engineering applications. The ability to quantitatively assess material structure and tissue formation is critical for evaluating bioreactor efficacy and for preimplantation assessment of tissue quality. Techniques that allow for the nondestructive and longitudinal monitoring of large engineered tissues within the bioreactor systems will be essential for the translation of these strategies to viable clinical therapies. X-ray Phase Contrast (XPC) imaging techniques have shown tremendous promise for a number of biomedical applications owing to their ability to provide image contrast based on multiple X-ray properties, including absorption, refraction, and scatter. In this research, mesenchymal stem cell-seeded alginate hydrogels were prepared and cultured under osteogenic conditions in a perfusion bioreactor. The constructs were imaged at various time points using XPC microcomputed tomography (mCT). Imaging was performed with systems using both synchrotron- and tubebased X-ray sources. XPC mCT allowed for simultaneous three-dimensional (3D) quantification of hydrogel size and mineralization, as well as spatial information on hydrogel structure and mineralization. Samples were processed for histological evaluation and XPC showed similar features to Correspondence to: E.M. Brey Contract grant sponsor: Veterans Administration Contract grant sponsor: National Science Foundation Grant numbers: CBET-1263994; IIS-1125412 Contract grant sponsor: National Institute of Health Grant numbers: R01EB009715; R01AR061460 Received 23 July 2014; Revision received 10 September 2014; Accepted 18 September 2014 Accepted manuscript online xx Month 2014; Article first published online in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/bit.25467
ß 2014 Wiley Periodicals, Inc.
histology and quantitative analysis consistent with the histomorphometry. These results provide evidence of the significant potential of techniques based on XPC for noninvasive 3D imaging engineered tissues grown in bioreactors. Biotechnol. Bioeng. 2014;9999: 1–9. ß 2014 Wiley Periodicals, Inc. KEYWORDS: X-ray phase contrast; micro-computed tomography; bioreactor; alginate microbeads; synchrotron; analyzer-based imaging
Introduction Tissue engineering strategies often focus on the growth of tissue in vitro prior to implantation. Bioreactor systems are designed to provide controlled environments for engineering tissues of appropriate structure for implantation. Bioreactorbased approaches have been used recently to successfully engineer organs and tissues for clinical application, including bladder, trachea, and vascular grafts (Atala et al., 2006; Gonfiotti et al., 2014; McAllister et al., 2009; Wystrychowski et al., 2013). The success of these studies supports the broad potential for bioreactor-based strategies for engineering replacement tissues and organs for clinical application. In these settings, it is critical that tools are available that allow assessment of three-dimensional (3D) tissue development within the bioreactor to determine readiness for implantation. In addition, these tools could be used to gain insight into the basic mechanisms of tissue formation in bioreactor environments. Presently, histological evaluations of engineered tissues are considered the “gold standard” to determine the success of a given strategy. However, these
Biotechnology and Bioengineering, Vol. 9999, No. xxx, 2014
1
techniques require the destruction of samples and often result in elimination of the biomaterial. In addition, histology does not provide an accurate 3D representation of the sample. Currently, many optical imaging techniques including confocal or multiphoton microscopy, optical coherence tomography, laminar optical tomography among others have been investigated to overcome the shortcomings of histological methods (Ahearne et al., 2008; Bagnaninchi et al., 2007; Hofmann et al., 2011, 2012; Mason et al., 2004). These techniques allow for nondestructive 3D analysis of samples; however, they are limited in regards to the volumes of tissues that can be imaged (several millimeters) and many materials and tissues generate little contrast with these methods. Samples that involve calcified tissues or mineralization of biomaterials present particular challenges for optical imaging techniques due to their high absorption of the light used in these techniques. Magnetic Resonance Imaging (MRI) has also been employed to monitor tissue engineering in vitro (Kotecha et al., 2012, 2013; Othman et al., 2012; Ramaswamy et al., 2008, 2009; Ravindran et al., 2012; Reiter et al., 2012). MRI is a valuable noninvasive imaging technique with significant potential for longitudinal imaging of engineered tissues. In MRI, spatial resolution is inversely proportional to the field of view, which limits its use in evaluating large volumes in applications where high spatial resolution is needed. This limitation and poor contrast for mineralized tissues restrict its application for assessment of bone formed in a bioreactor. Conventional absorption-based X-ray imaging techniques are commonly used to visualize calcified tissue and mineralization in engineered tissues (Cartmell et al., 2004; Chatterjee et al., 2010; Porter et al., 2007; Sikavitsas et al., 2005). These techniques allow high resolution, on micron level, imaging of large tissue volumes, several centimeters. However, biomaterials with low X-ray attenuation generate little or no contrast with these techniques without the use of exogenous contrast agents. While many contrast agents exist to increase the visibility of these biomaterials and soft tissues in X-ray images (Metscher, 2009), they are often harmful or toxic to cells making their use in monitoring tissue development in a bioreactor nonideal. X-ray Phase Contrast (XPC) imaging techniques overcome limitations of conventional absorption-based X-ray imaging by relying on alternative X-ray contrast mechanisms. XPC imaging permits visualization of biomaterial and soft tissue structures based on differences in X-ray refractive indices. Moreover, XPC imaging facilitates the identification of different biomaterials based on their unique X-ray signatures (Appel et al., 2012; Brey et al., 2010; Izadifar et al., 2014; Langer et al., 2010; Zhu et al., 2011). XPC imaging methods allow biomaterial and soft tissue contrast while retaining the high spatial resolution (micron level), large penetration depths (several centimeters), and calcified tissue contrast of conventional methods based on X-ray absorption. By relying on contrast resulting from X-ray refraction, XPC can reduce the X-ray dose applied to less than 10% of that used for traditional microcomputed tomography (mCT) (Lewis, 2004; Parham
2
Biotechnology and Bioengineering, Vol. 9999, No. xxx, 2014
et al., 2009; Zhou and Brahme, 2008). This may enable simultaneous imaging of biomaterial scaffold and calcified tissue structure in large volume samples. There are four basic schemes that can be used to implement XPC imaging: Propogation-based PC imaging, X-ray interferometry, Differential PC imaging using X-ray gratings, and Analyzer-based imaging using crystals (Appel et al., 2011; Zhou and Brahme, 2008). Image contrast is produced by differences in the X-ray attenuation, refraction, and, in some cases, scattering properties of the tissue or material. Depending on the setup, these properties can be displayed in separate images or as a single mixed-contrast image containing both XPC and absorption effects. In this study, we investigated the use of two XPC imaging techniques. The first technique, an analyzer-based method known as multiple image radiography (MIR), was implemented at a synchrotron imaging facility. The MIR technique produces three separate images that depict the X-ray absorption, refraction, and ultra-small-angle-X-ray-scatter (USAXS) properties of the samples. This enables the identification of the specific contrast mechanisms of the tissue and materials under investigation. However, MIR is difficult to implement using an X-ray tube source limiting its broad application. The second technique was propagation-based XPC imaging, which has a similar setup to conventional absorption-based X-ray imaging; however, the detector is placed farther downstream. The act of free space wavefield propagation induces XPC effects into the measured intensity image. This results in an edge enhancement effect at image locations corresponding to the projected tissue and material interfaces. This technique was implemented with an X-ray tube source as an illustration of in-lab XPC imaging of engineered tissues to demonstrate a more translatable technology. The main objective of this study was to investigate techniques based on XPC for the simultaneous 3D imaging of alginate hydrogel structure and mineralization in engineered bone tissues generated in a perfusion bioreactor. Absorptionbased mCT systems are used routinely to image mineralization in engineered tissues, but are unable to simultaneously evaluate hydrogel scaffold structure. XPC is expected to enable imaging of both hydrogel scaffold and tissue structures in 3D. The engineered tissues are imaged without the introduction of exogenous contrast agents in order to evaluate imaging based on contrast resulting from material and tissue properties alone.
Materials and Methods Preparation of Alginate Hydrogel Scaffolds Alginate microbeads were prepared using standard methods (Thompson et al., 2009; Yeatts and Fisher, 2011; Yeatts et al., 2011, 2012; Yoon and Fisher, 2008; Yoon et al., 2007). Briefly, a 2.0% w/v alginate solution was prepared by adding alginic acid sodium salt from brown algae (Sigma–Aldrich, St. Louis, MO) to a solution of 0.15 M NaCl (Sigma–Aldrich),
and 0.025 M HEPES (Sigma–Aldrich) in deionized water. The alginate solution was then sterilized via autoclaving. Human mesenchymal stem cells (hMSCs) were suspended in the alginate solution at a density of 1.25–2.5 106 cells/mL. The alginate/cell solution was added dropwise through an 18gauge needle into a stirred solution of 0.1 M calcium chloride (Sigma–Aldrich). Beads were incubated for 15 min then transferred into six-well plates in control medium for 24 h before loading into the bioreactor (Yeatts and Fisher, 2011). Control media consisted of DMEM (Gibco, Carlsbad, CA) supplemented with 10% fetal bovine serum (Gibco), 1.0% v/v penicillin/streptomycin (Gibco), 0.1 mM nonessential amino acids (Gibco), and 4 mM L glutamine (Gibco) (Yeatts et al., 2012). Beads were cultured in osteogenic medium in a perfusion bioreactor system previously described (Yeatts and Fisher, 2011; Yeatts et al., 2011, 2012) and prepared for analysis at 1, 14, 21, and 28 days. Osteogenic media was formulated by supplementing control media with 100 nM dexamethasone (Sigma–Aldrich), 10 mM b-glycerophosphate, and 173 mM ascorbic acid (Sigma–Aldrich) (Yeatts et al., 2012). The bioreactor system is the same system that has been previously described (Yeatts and Fisher, 2011; Yeatts et al., 2011, 2012) and consisted of a tubular growth chamber and medium reservoir connected via a tubing circuit. The flow of media was driven by an L/S Multichannel Pump System (Cole Parmer, Vernon Hills, IL). The circuit was comprised of platinum-cured silicone tubing (Cole Parmer) for all areas except those that pass through the pump, which was composed of Pharmed BPT tubing (Cole Parmer). Tubing was connected using silver ion-lined microbial resistant tubing connectors (Cole Parmer). The growth chamber consisted of a length of platinum-cured silicone tubing (Cole Parmer) with an inner diameter of 6.4 mm, an outer diameter of 11.2 mm, and a wall thickness of 2.4 mm. The growth chamber was 13 cm in length and was packed with 30 cell seeded alginate beads. Growth chamber tubing connectors were modified by adding 60 mesh stainless steel screens (Fisher Scientific, Pittsburgh, Pennsylvania) to restrict bead movement. After loading, the autoclaved tubing was fully assembled inside a cell culture hood and then placed in a cell culture incubator at 37 C and 5% CO2. Fifty milliliters of the osteogenic medium was loaded into 125 mL Erlenmeyer flask and topped with a rubber stopper. The medium was withdrawn and replaced from the reservoir through two tubes that penetrate the stopper and changed every 3 days. Beads were removed from the bioreactor by moving the entire bioreactor system into the hood, disconnecting one tubing circuit and flushing beads out of the growth chamber with phosphate-buffered saline (PBS) (Yeatts and Fisher, 2011; Yeatts et al., 2011, 2012). XPC imaging Multiple Image Radiography (MIR) Alginate microbeads submerged in calcium chloride solution within plastic cuvettes were imaged using an MIR imaging
system at the National Synchrotron Light Source at Brookhaven National Laboratory (Beamline X15A) (Zhong et al., 2000). A monochromated 20 keV beam was utilized for X-ray imaging. The detector employed was an X-ray Imager VHR 1:1, CCD (Photonic Science Limited, East Sussex, United Kingdom) sensor (400 4,008 pixels) with a detector pixel size of 9 mm. Using a [333] analyzer crystal reflection, the measurement data were acquired at 11 angular positions of the analyzer crystal ranging from 4 mrad to þ4 mrad to generate a rocking curve for each pixel in the detector for each tomographic view angle (Appel et al., 2012; Brankov et al., 2006; Chou et al., 2007). One hundred tomographic views were acquired over a 180 angular range. The acquisition time at each analyzer-crystal orientation was one second making total acquisition time 21/2 h. At each tomographic view angle, three parametric MIR images that represent projected absorption, refraction, and USAXS properties of the sample were computed (Wernick et al., 2003). From knowledge of the three MIR images computed at each tomographic view angle, an OS-SART-FISTA reconstruction method (Guan et al., 2014) with a binning factor of 2 was employed to reconstruct volumetric images of the three MIR properties resulting in a final spatial resolution of 36 mm/pixel. Propagation-Based XPC Benchtop studies of propagation-based XPC imaging were conducted at Washington University in St. Louis using the same system described by Zysk et al. (2012). This imaging system employs a microfocus X-ray tube with a Tungsten anode (Thermo Kevex PXS10-65W; 7-100 micron spot size, 65 W tube power, 45–130 kVp) and a deep cooled highresolution X-ray camera (Princeton Instruments Quad RO 4096; 15 mm pitch, 16-bit quantization) mounted on a vibration-isolated optical table and housed in a shielded room. The sample and detector were mounted to a longtravel optical rail assembly (Thorlabs XT95SP-1000) for translation in the beam propagation direction. The sample was positioned by a pair of computer-controlled linear translation stages (Thorlabs LTS150) and a computercontrolled rotation stage (Thorlabs NR360S). Source control and image acquisition were performed via computer at a remote location outside the shielded room (Zysk et al., 2012). The beads were submerged in calcium chloride solution in 1 ml syringes and placed 39.9 cm from the source and 56.7 cm from the detector resulting in a magnification of 2.42. The X-ray tube voltage was set at 45 kVp with a current of 160 mA corresponding to a nominal focal spot size of 9 mm. The system effective pixel size was calculated as 6.2 mm. The exposure time was 15 s for each tomographic view angle making total CT scan time 3 h. Four hundred tomographic views were collected over a 200 angular range in order to correct for the cone beam geometry (Kak and Slaney, 1999). An OS-SART-FISTA reconstruction method (Guan et al., 2014) with a binning factor of 2 was employed to reconstruct volumetric images of the alginate beads resulting in a final spatial resolution of 25 mm/pixel.
Appel et al.: 3D Imaging of Tissues Engineered in Bioreactor Biotechnology and Bioengineering
3
Histology Following imaging, the alginate beads were dehydrated in an ethanol series (Leica TP1020) and paraffin embedded using routine histological methods. Serial sections (5 mm) were cut using a Leica RM2200 microtome and stained for calcium using Alizarin Red S or von Kossa using standard protocols. Stained sections were imaged using an Axiovert 200 inverted microscope (Carl Zeiss MicroImaging, Inc. Gottingen, Germany) equipped with an AxioCam MRc5 color digital camera (Carl Zeiss). The camera and computer controlled X-Y-Z stage allowed tiling of multiple images to build high resolution images of entire bead cross sections. Axiovision 4.2 image analysis software (Carl Zeiss) allowed automated control of all aspects of acquisition and processing.
Calculation of Bead Volume and Mineralization Percent mineralization of alginate beads was calculated from images produced by all the imaging techniques at all time points. Percent mineralization was calculated as volume/area of mineralization after thresholding divided by volume/area of beads. For the synchrotron MIR data, the reconstructed refraction volume was used to calculate bead volumes and the reconstructed absorption volume was globally thresholded and used to calculate the volumes of mineralization within the beads. Because all beads from all time points were imaged in the same scan, a single threshold was selected for the entire volume. For the benchtop propagation-based data, only one volume is produced, that is, a combination of refraction and absorption contrast. This volume was used to measure the
Figure 1.
total bead volumes and globally thresholded to calculate the mineralization volume. Finally, bead area and mineralization area were calculated using both Alizarin Red S and von Kossa stained sections. Bead area and mineralization from histology were only two-dimensional measurements. Bead area was manually measured in Axiovision while area of mineralization measurements were calculated after image thresholding. Two sections of each bead were included for the measurements in the histology samples. Statistical Analysis All statistical data are expressed as means standard deviation. Statistical significance was determined using ANOVA with a Tukey’s post test, with a P value of less than 0.05 considered significant. Coefficient of Variation (Cv),defined as standard deviation divided by mean, was also calculated for percent mineralization for each method in order to quantify the relative magnitude of deviation in the measurements.
Results MSC-seeded alginate beads were cultured in osteogenic media in a perfusion bioreactor system. These culture conditions are known to lead to a temporal increase in mineralization resulting from osteogenic differentiation (Yeatts et al., 2012). Three to five beads from the bioreactor were imaged with two XPC imaging techniques following culture for 1, 14, 21, and 28 days. All samples were imaged with both XPC systems to allow for comparison.
(a, d) Absorption, (b, e) refraction, and (c, f) USAXS projection views of cell seeded alginate beads cultured in the perfusion bioreactor for 1 day (a, b, c) and 21 days (d, e, f). Beads without mineralization are invisible in the absorption image (a) but can be seen in the refraction image (b). Mineralized beads produce contrast in the absorption image (d) while the change from spherical shape can be seen in the refraction image (e). Scale bar ¼ 1 mm.
4
Biotechnology and Bioengineering, Vol. 9999, No. xxx, 2014
Figure 3.
Average diameter of alginate beads cultured in perfusion bioreactor measured from MIR data. denotes P-value
