Human collagen-based multilayer scaffolds for tendon-to-bone interface tissue engineering Beob Soo Kim,1,2 Eun Ji Kim,1 Ji Suk Choi,1 Ji Hoon Jeong,3 Chris Hyunchul Jo,4 Yong Woo Cho1,2 1
Department of Chemical Engineering, Hanyang University, Hanyangdaehak-ro 55, Ansan, Kyeonggi-do 426-791, Republic of Korea 2 Center for Theragnosis, Biomedical Research Institute, Korea Institute of Science and Technology (KIST), Hwarangno 14-gil 5, Seoul 136-791, Republic of Korea 3 School of Pharmacy, Sungkyunkwan University, Suwon 440-746, Republic of Korea 4 Department of Orthopedic Surgery, Joint and Spine Center, SMG-SNU Boramae Medical Center, Seoul National University College of Medicine, 20 Boramae-ro 5-gil, Seoul 156-707, Republic of Korea Received 23 October 2013; accepted 6 December 2013 Published online 00 Month 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.a.35057 Abstract: The natural tendon-to-bone region has a gradient in structure and composition, which is translated into a spatial variation of chemical, physical, and biological properties. This unique transitional tissue between bone and tendon is not normally recreated during natural bone-to-tendon healing. In this study, we have developed a human collagen-based multilayer scaffold mimicking the tendon-to-bone region. The scaffold consists of four different layers with the following composition gradient: (a) a tendon layer composed of collagen; (b) an uncalcified fibrocartilage layer composed of collagen and chondroitin sulfate; (c) a calcified fibrocartilage layer composed of collagen and less apatite; (d) a bone layer composed of collagen and apatite. The chemical, physical, and mechanical properties of the scaffold were characterized by a scanning
electron microscope, porosimeter, universal tensile machine, Fourier transform infrared spectrometer, energy dispersive X-ray analysis apparatus, and thermogravimetric analysis apparatus. The multilayer scaffold provided a gradual transition of the physical, chemical, and mechanical environment and supported the adhesion and proliferation of human fibroblasts, chondrocytes, and osteoblasts toward each corresponding matrix. Overall, our results suggest the feasibility of a human collagen-based multilayer scaffold for regeneration C 2014 Wiley Periodicals, Inc. J of hard-to-soft interface tissues. V Biomed Mater Res Part A: 00A:000–000, 2014.
Key Words: adipose tissue, collagen, tendon-to-bone, fibrocartilage, calcification
How to cite this article: Kim BS, Kim EJ, Choi JS, Jeong JH, Jo CH, Cho YW. 2014. Human collagen-based multilayer scaffolds for tendon-to-bone interface tissue engineering. J Biomed Mater Res Part A 2014:00A:000–000.
INTRODUCTION
Traditional tissue engineering uses cells, biomaterials, and environmental factors to regenerate specific tissues such as skin, bone, cartilage, and nerve. Recent advances in tissue engineering have led to the engineering of functional interface tissues, which bridge the gap between two tissues that differ in their physical, chemical, and biological properties. Interface tissue engineering may need scaffolds made of multilayer or graduated biomaterials for the regeneration of continuous transitional regions. Indeed, recent gradient interfacial designs of scaffolds have addressed mineralization,1–3 matrix composition,3–6 pore size distribution or porosity,6,7 and bioactive signals8,9 for resolution of transitioning orthopedic interface tissues such as the cortical–trabecular bone interface, the cartilage–bone (osteochondral) interface, and the ligament/tendon-to-bone interface. In particular, the ligament/tendon-to-bone interface has been
explored recently.10 Ligaments and tendons have similar structure and composition, including a layering of type I collagen and proteoglycans. Both tissues also lack a good blood supply and do not heal well, requiring surgery if torn or injured. Therefore, many studies are being performed to formulate a gradient scaffold design for use in orthopedic surgery.11–14 The tendon-to-bone interface is characterized by a transition region of generally four types of tissue, with controlled spatial variations in cell type and matrix composition. The first region is an area of fibrous connective tissue found in normal tendon composed of aligned collagen fibrils with fibroblasts embedded throughout a type I and type III collagen matrix. The second is an uncalcified fibrocartilage layer, which contains ovoid chondrocytes, collagen types I and II, and aggrecan. The third is a calcified fibrocartilage region, which is composed of hypertrophic, highly
Correspondence to: Y.W. Cho; e-mail: [email protected] and C.H. Jo; e-mail: [email protected] Contract grant sponsor: National Research Foundation of Korea; contract grant number: 2011-0019774
C 2014 WILEY PERIODICALS, INC. V
1
circular chondrocytes and collagen types I, II, and X.15 The last region is the subchondral bone layer. The bone tissue contains osteoblasts, osteocytes, and osteoclasts with calcified type I collagen fibers. In the all regions of the interface, however, there are dense networks of collagen fibrils interconnecting heterogeneous tissues. The mechanical properties of tendon and bone are significantly different. Tendons have a high tensile strength, whereas bone is optimized for compressive loading. In contrast, the transitional regions such as the uncalcified and calcified fibrocartilage layers are superior in transferring compressive load, and the insertion zones show an increase in Young’s modulus across the interface.10 Thus, the intermediate regions bridge the gap between the mechanical properties of tendon and bone.16 This controlled matrix heterogeneity serves to minimize the formation of stress concentrations and promote a gradual transition of mechanical loading between the soft and hard tissue.17 Similarly, in the design of engineered scaffolds for regeneration of multitissue interfaces, the existence of intermediate layers is vital for soft-to-hard tissue integration, fixation, and stability. The goal of this study is to design, fabricate, and in vitro evaluate human collagen-based, multilayer scaffolds for regeneration of the interface between tendon and bone. We hypothesized that the functional properties of engineered tendon-to-bone interface tissue would depend on the chemical composition of the scaffolds. Therefore, multilayer scaffolds with the following four separate layers were fabricated: (i) an upper layer of collagen matrix, mimicking the tendon region; (ii) a first intermediate layer of crosslinked aggrecan–collagen resembling the uncalcified fibrocartilage region; (iii) a second intermediate layer of partially calcified collagen resembling the calcified fibrocartilage region; (iv) a lower layer of calcified collagen with a high degree of mineralization corresponding to subchondral bone. Thereafter, the in vitro responses of cells to the engineered multilayer scaffolds were investigated in terms of cell adhesion, viability, and proliferation using human fibroblasts, chondrocytes, and osteoblasts. It is anticipated that this collagen-based multilayer scaffold composed of four different layers will result in a successful graft for tendon-tobone interface tissue engineering. MATERIALS AND METHODS
Material processing and scaffold fabrication Collagen was extracted from human adipose tissue, according to a previously developed protocol.18 In brief, adipose tissue obtained by liposuction was washed several times with distilled water to remove blood components. The crude extracellular matrix (ECM) was isolated from adipose tissue by homogenization, centrifugation, and rinsing.19 The ECM was solubilized in 0.5M acetic acid containing 0.1% pepsin (EC 3.4.23.1 powder; Sigma-Aldrich, St. Louis, MO) after alkaline (0.1M NaOH) and alcohol (70% isopropanol) treatments. The resulting viscous solution was centrifuged, and the supernatant was selectively precipitated with 0.9M NaCl. Precipitated collagen was dissolved in 0.5M acetic acid and desalted by a centrifugal filter device (Amicon Ultra-15;
2
KIM ET AL.
Millipore, Billerica, MA). The concentration of collagen was measured using a soluble collagen assay kit (Sircol, Biocolor, Carrickfergus, Northern Ireland), and the purified collagen was freeze-dried. To mimic the tendon-to-bone interface with a composition gradient, a scaffold consisting of the following four different layers was prepared: (a) a tendon layer composed of collagen; (b) an uncalcified fibrocartilage layer composed of collagen and chondroitin sulfate; (c) a calcified fibrocartilage layer composed of collagen and apatite (low calcification); (d) a bone layer composed of collagen and apatite (high calcification). Calcification was performed on the collagen matrices to precipitate hydroxyapatite (HA) by a calcium-phosphate dipping method.20 Mimicking the tendon layer. Collagen dispersion of 0.8% w/v was prepared. The respective mass of collagen was soaked in 0.02N acetic acid for 24 h at 4oC. To prevent heat build-up, the mixture was homogenized on a bed of ice for 10 min using a homogenizer (T10 basic Ultra-Turrax; IKA, Staufen, Germany). Air bubbles were removed by centrifugation at 1000 rpm for 20 min at 4oC. The collagen dispersion was cast into a mold, frozen at 220oC, and freeze dried. The prepared collagen matrix was disk-shaped with a thickness of 1 mm and a diameter of 10 mm. The freeze-dried collagen matrix was cross-linked by 1-ethyl-3(3-dimethyl aminopropyl) carbodiimide (EDC) and N-hydroxysuccinimide (NHS). Mimicking the uncalcified fibrocartilage layer. The freezedried collagen matrix was cross-linked in the presence of chondroitin sulfate (CS; Tokyo Chemical Industry, Tokyo, Japan) by EDC and NHS. Collagen matrices were immersed and incubated in 40% v/v ethanol containing 50 mM 2-morpholinoehane sulfonic acid (MES) for 30 min at room temperature. Subsequently, matrices were incubated in 40% v/v ethanol containing 50 mM MES and 2% w/v CS for 4 h at room temperature with 5 mg/mL EDC to yield a collagen–CS composite at a ratio of 9:1.21 NHS was added in an EDC:NHS ratio of 4:1. After the reaction, excess EDC and CS were rinsed from the matrix using 0.1M Na2HPO4 for 1 h. Mimicking the calcified fibrocartilage layer. Collagen matrices were immersed in 500 mM calcium chloride dehydrate solution (CaCl22H2O; Sigma-Aldrich) for 5 min and rinsed for 1 min in distilled water. Subsequently, the matrices were immersed in 300 mM disodium hydrogen phosphate solution (Na2HPO4; Duksan Pure Chemical, Ansan, Korea) for 5 min and rinsed for 1 min in distilled water. Both calcium and phosphate solutions contained 0.1M potassium chloride (KCl; Sigma-Aldrich) to maintain ionic stability and were adjusted to pH 8.22 The Ca/P molar ratio of the precipitates at pH 8 exhibited a close stoichiometric composition (Ca/P 5 1.67) of HA. Mimicking the bone layer. The above calcification procedure represented a single cycle. Collagen matrices were subjected to three cycles of the calcification treatment. Individually prepared layers were rinsed with distilled water and freeze-dried.
HUMAN COLLAGEN-BASED MULTI-LAYER SCAFFOLDS
ORIGINAL ARTICLE
Scaffold characterization The morphology of each layer was observed by scanning electron microscopy (SEM; S-4800 FE-SEM, Hitachi, Tokyo, Japan) at an accelerating voltage of 10 kV, after platinum coating by a sputter coater (Emitech k575x; Quorum Technologies, Wellington, UK). Fourier transform infrared (FTIR) spectra were obtained in an attenuate total reflectance mode with a Nicolet 6700 FTIR spectrometer equipped with a diamond probe (Thermo Fisher Scientific, Waltham, MA). The results are an average of 256 scans in a spectral range between 1800 and 800 cm21. Pore size and the porosity of each layer were determined using an automated mercury porosimeter (Autopore IV 9500, Micromeritics, Norcross, GA). The compressive strength of each layer was measured using a universal tensile machine (H100KS, Tinius Olsen, Horsham, PA) with a 50-N load cell. The weight fraction of HA in a composite was measured by a thermogravimetric analysis (TGA) apparatus (TG-DTA 2000SA; Bruker AXS, Billerica, MA) upon heating from 25 to 800oC at 10oC/min under N2 atmosphere. The HA composites were examined by SEM equipped with an energy dispersive X-ray analysis apparatus (EDX; EX-250, Horiba, Kyoto, Japan). Cell culture and seeding Human synovial membrane fibroblasts, human chondrocytes, and human osteoblasts were purchased from Applied Biological Materials (abm; NY; T4018, T0020, and T4015). Cells were cultured and expanded in T25 ECM-coated flasks (G299, abm). Fibroblasts were cultured in Progrow III medium (TM003, abm) supplemented with 10% fetal bovine serum (FBS; TM999, abm) and 1% penicillin/streptomycin (P/S; Invitrogen, Carlsbad, CA). Chondrocytes and osteoblasts were also cultured in complete medium containing Progrow IV medium (TM004, abm), FBS, and P/S (1/0.1/ 0.01 volume ratio). Cells were fed every 3 days and cultured to confluence at 37oC under 5% CO2 in a humidified atmosphere. Passage of five to seven cells was used throughout. After being sterilized by ethylene oxide gas, the four types of layers (thickness 1 mm, diameter 10 mm) were incubated in culture medium for 2 h before cell seeding. The equilibrated layer was placed into each well of a non-treated 24-well polystyrene plate (32024, SPL Life Sciences, Pocheon, Korea). All three types of cells were trypsinized and resuspended at a concentration of 5 3 105 cells per 50 mL media. Ten microliters of cell suspension (1 3 105 cells) was added to each layer. After 3 h, 1 mL of the culture medium was added to each well. Cell-seeded layers were incubated at 37oC and 5% CO2 and cultured for 3 weeks. Media were changed every 2 or 3 days for the duration of all experiments. Cell adhesion and proliferation The proliferation of human cells in a scaffold was analyzed by a water soluble tetrazolium salts (WST) assay, which is based on the reduction of a stable tetrazolium salt to a soluble violet formazan product by viable cells. After 1, 7, 14, and 21 days of culture, the cell-seeded layers were rinsed with phosphate-buffered saline and then incubated with
1 mL of 10% WST (Hoffmann-La Roche, Basel, Switzerland) solution in the dark at 37oC under 5% CO2 for 30 min. After gentle pipetting, the 100 lL solution was transferred to a 96-well plate. The optical density was then measured at 440 nm using a microplate spectrophotometer (PowerWave XS; Bio-Tek Instruments, Winooski, VT). For the cell viability analysis, the commercially available live/dead viability/cytotoxicity kit (Molecular Probes, Eugene, OR) was used. Green fluorescence caused by the reaction of calcein with intracellular esterase indicated live cells, whereas red fluorescence caused by ethidium homodimers that bound to nucleic acids indicated dead cells. Cell-seeded layers were stained with the combined live/dead reagents (Molecular Probes) at 37oC under 5% CO2 for 30 min and observed using a fluorescence microscope (IX81, Olympus, Tokyo, Japan). The morphology of the cells attached to the layers was examined by SEM. Cell-layer constructs were fixed in 2.5% glutaraldehyde for 30 min. After cell fixation, they were dehydrated through a gradient series of ethanol solutions from 50 to 100% in increments of 10% for 10 min each and freezedried under vacuum for 2 days. Samples were coated with platinum by sputtering and were then observed by SEM at an accelerating voltage of 15 kV. Histological evaluation was performed after 28 days of culture. Cell-scaffold constructs were composed of an osteoblast–bone layer, a chondrocytecalcified fibrocartilage layer, a chondrocyte-uncalcified fibrocartilage layer, and a fibroblast–tendon layer at 21 days postseeding. The four layers were piled up and then cultured for 7 days in Dulbecco’s modified Eagle’s medium (DMEM; Gibco-BRL, Life Technologies, Carlsbad, CA) supplemented with 10% FBS and 1% P/S. The samples were fixed with 4% formalin, embedded in paraffin, and stained with hematoxylin-eosin (HE; Sigma-Aldrich). Statistical analysis All data are represented as means 6 standard deviation. Statistical computations used the Tukey method of one-way analysis of variance. p-values less than 0.05 were considered significant. RESULTS
Isolated collagen fibrils from human adipose tissue are shown in a high-magnification SEM image (Fig. 1). The characteristic D-band of collagen fibrils was clearly observed, and the diameter of the collagen fibrils was approximately 300 lm. Figure 2 shows the cross-sectional morphologies of each layer of the multilayer scaffold. All layers exhibited the typical porous structure of scaffolds prepared by a freezedrying method. There were no significant differences in the appearance of the structures between the tendon layer composed of collagen and the uncalcified fibrocartilage layer composed of collagen and CS, whereas the calcified fibrocartilage layer and the bone layer showed distinct morphologies because of HA crystals on the surface of the collagen matrices. Of note, the bone layer exhibited a smaller pore size because the collagen matrices were infiltrated by many HA crystal agglomerates. These results were also confirmed
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MONTH 2014 VOL 00A, ISSUE 00
3
FIGURE 1. Scanning electron micrographs of collagen isolated from human adipose tissue. Scale bar represents 1 lm.
using a mercury intrusion porosimeter [Fig. 3(A)]. The proportion of material with a pore size below 40 lm in the bone layer was 71.08%, whereas that in the tendon layer was 34.79%. A pore size of under 40 lm was observed most commonly for all layers, whereas a size of 180–350 lm was the least common. The average porosity for all layers was above 93%. The mechanical properties measured by compressive experiments are shown in Figure 3(B). The mechanical property of each layer was dependent on the chemical composition, such as the content of collagen, CS,
and HA. The elastic moduli of the tendon, uncalcified fibrocartilage, calcified fibrocartilage, and bone layer were 303 6 131, 1181 6 333, 2954 6 1143, and 4469 6 1650 kPa, respectively, whereas the typical elongation at break values were 112.5 6 13.1, 105.8 6 13.7, 82.0 6 8.8, and 71.2 6 13.8%, respectively. These results obviously show that the varying ingredients of each layer provide different mechanical properties at different layers of the graded collagen-based scaffolds. The FTIR spectra of tendon, uncalcified fibrocartilage, and bone layers are shown in Figure 4. All layers had the three major amide absorption bands of collagen, namely amide I, II, and III at 1628–1647, 1539–1560, and 1233– 1243 cm21, respectively. They were comparable to the spectra of collagen from other sources.18 The characteristic absorbance bands for CS are 1603–1611 and 1411–1413 cm21 (COO–), 1225–1258 cm21 (SO42–), and sugar groups in the range of 983–1285 cm21.23–25 The spectrum of an uncalcified fibrocartilage layer containing CS also displayed peaks corresponding to COO2(1408 cm21), SO42– (1263 cm21), and sugar (970–1120 cm21) groups [Fig. 4(A)]. On the other hand, the carboxylate absorbance band (approximately 1610 cm21) was not found because of overlap with the amide I band of collagen. In the bone layer containing HA, the vibrational bands of phosphate and carbonate were visible at 1020 and 873 cm21, respectively [Fig. 4(B)].22,26 Figure 5 shows calcification of collagen matrices evident on SEM, the quantification of HA concentration by TGA, and the
FIGURE 2. Cross-sectional images of the tendon layer (A), uncalcified fibrocartilage layer (B), calcified fibrocartilage layer (C), and bone layer (D). Scale bar represents 200 lm.
4
KIM ET AL.
HUMAN COLLAGEN-BASED MULTI-LAYER SCAFFOLDS
ORIGINAL ARTICLE
FIGURE 3. Pore size distribution (A) and mechanical properties (B) of the tendon, uncalcified fibrocartilage, calcified fibrocartilage, and bone layers. Data presented are means 6 standard deviation (p < 0.05).
ratio of calcium to phosphate values by EDX. SEM images of the calcified fibrocartilage and bone layer [Fig. 5(A,B), respectively] qualitatively show an increase in the homogenous calcification proportional to the number of times the material was dipped in the calcium and phosphate solutions. The higher magnification SEM image showed the D-band of collagen fibrils and HA nano-crystals on the collagen framework [Fig. 5(C)]. TGA comparatively showed the HA content of the calcified fibrocartilage and bone layers [Fig. 5(D)]. The first weight loss observed between 25 and 100oC is associated with the removal of water in each composite,27 whereas the second weight loss occurring at 250–550oC corresponds to the loss of organic content. The collagen/HA (organic/inorganic) mass ratios of the calcified fibrocartilage and bone layers were 51/49 and 27/73, respectively. EDX analyses showed rich calcium phosphate deposits, and the Ca/P ratios of deposits in the calcified fibrocartilage and bone layer were 1.63 6 0.14 and 1.62 6 0.06, respectively [Fig. 5(E,F)], which were close to biological apatite (Ca/P ratio 1.67). Cell adhesion and proliferation in each layer of the gradient scaffold was evaluated using a WST assay as shown in Figure 6. The proliferation of fibroblasts on the tendon layer
FIGURE 4. ATR-FTIR spectra of the tendon layer composed of collagen, the uncalcified fibrocartilage layer composed of collagen and chondroitin sulfate (A), and the bone layer composed of collagen and HA (B). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
was significantly higher (p 0.05) than those of chondrocytes and osteoblasts on day 21. Similarly, the proliferation of chondrocytes on the uncalcified fibrocartilage layer and that of osteoblasts on the bone layer were the highest compared with other cell types on day 21. On the other hand, no significant difference was observed between chondrocytes and osteoblasts on the calcified fibrocartilage layer on day 21. However, the percentage increases in the proliferation rates of chondrocytes and osteoblasts on the calcified fibrocartilage layer from day 7 to day 21 were found to be 263 and 193%, respectively. Overall, every type of cell showed good adhesion and proliferation on all the prepared layers, and all cell numbers significantly increased during the experimental period. The left column of Figure 7 shows the live/dead staining images on day 7, and the right column shows SEM micrographs on day 14, fibroblasts on the tendon layer [Fig. 7(A)], chondrocytes on uncalcified
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MONTH 2014 VOL 00A, ISSUE 00
5
FIGURE 5. SEM images of the calcified fibrocartilage layer (A) and the bone layer (B). (C) is higher magnification of (A) and shows that the nano-HA particles were formed on the collagen fibrils. TGA curves of collagen/HA composites (D). EDX spectra of HA on the collagen matrices of the calcified fibrocartilage layer (E) and bone layer (F). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
fibrocartilage [Fig. 7(B)], and calcified fibrocartilage layers [Fig. 7(C)] and osteoblasts on the bone layer [Fig. 7(D)]. In the live/dead assay, live cells were stained green by calceinAM and dead cells were stained red by an ethidium homodimer. After 7 days of culture, a uniform cell distribution was observed in each layer and most cells expressed green fluorescence. The SEM results showed that all human cells at a high density spread well on the surface of each scaffold. The histological appearance of the cell-seeded multilayer scaffold was observed by H&E staining (Fig. 8). All sections of the cell-seeded scaffold displayed dense structures and a large amount of secreted and accumulated ECM when compared with unseeded scaffolds.
6
KIM ET AL.
DISCUSSION
Kovacevic and Rodeo28 reported the histological images of a normal supraspinatus tendon insertion site, clearly showing that there were four distinct zones in the tendon-to-bone interface tissue: tendon, unmineralized fibrocartilage, mineralized fibrocartilage, and bone. Our research goal was to engineer a multilayer scaffold to mimic the natural tendonto-bone interface tissue. In an effort to achieve this goal, a scaffold composed of four layers with different chemical compositions was fabricated. We hypothesized that different constituents of each layer could provide graded mechanical properties in a scaffold and favorable environment for each cell type. The stratified scaffold system is based on human
HUMAN COLLAGEN-BASED MULTI-LAYER SCAFFOLDS
ORIGINAL ARTICLE
FIGURE 6. Proliferation of human synovial membrane fibroblasts, chondrocytes, and osteoblasts in the tendon (A), uncalcified fibrocartilage (B), calcified fibrocartilage (C), and bone layers (D) for 21 days examined by a WST assay. Data presented are means 6 standard deviation. * p < 0.05.
collagen, the main component of the natural tendon-to-bone interface tissue. Collagen possesses attractive cell binding properties required for cells to attach to the scaffold. In a previous study, we successfully isolated and purified human collagen from adipose tissue.18 The highly porous collagen matrix with high interconnectivity was prepared to resemble the tendon layer [Fig. 2(A)]. High porosity and pore interconnections affect not only cell infiltration, binding, and migration, but also the transport of oxygen, nutrients, and waste exchange.29 Then, a collagen–CS layer was fabricated to mimic a native uncalcified fibrocartilage region. EDC and NHS catalyze a reaction between amine groups of collagen and carboxylic acid groups of CS to form crosslinks.30 EDC activates a carboxylic acid of CS and then reacts with NHS to produce an NHS-activated carboxylic acid. Finally, the NHS-activated carboxylic acid reacts with an amine group of collagen to produce a crosslink. EDC/NHS cross-linked collagen–CS improves mechanical stability, and creates a favorable environment for chondrocytes.21 The results of this study also indicate that the presence of CS cross-linked onto a collagen matrix affects mechanical properties [Fig. 3(B)]. To mimic the calcified fibrocartilage and the bone region, the composites of both layers were composed of collagen
and HA. HA improves not only the mechanical properties, but also the bioactivity of the scaffold by providing a source of calcium and phosphate ions. The nucleation of HA onto a collagen matrix results from a high affinity for Ca21 ions and the formation of a hydrogen-bonded hypercomplex of PO42– ions to link the carboxylic acid groups of collagen.3,31 HA synthesis generally requires sintering at increased temperature, usually in excess of 1000oC,32 but our calcification process was conducted at room temperature to avoid collagen denaturation. Nevertheless, EDX profiles of the matrices confirmed that the values of the Ca/P ratio of the nucleated HA were not in the range of amorphous HA, but rather in the range of crystalline HA.33 This result shows that HA with a moderate Ca/P ratio can be created by control of pH values. The microstructures of a scaffold such as pore size, porosity, and interconnectivity have been shown to significantly influence cell behaviors such as adhesion, migration, and proliferation.34 Several studies have shown that the optimal pore sizes for cell ingrowth and fixation strength of tendon, fibrocartilage, and bone tissues are 20–250, 150– 500, and 50–400 lm, respectively.35–40 In this work, a layered scaffold with pore volume gradients between pore size
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MONTH 2014 VOL 00A, ISSUE 00
7
FIGURE 7. Fluorescence micrographs on day 7 and SEM images on day 14. Human synovial membrane fibroblasts on a tendon layer (A). Human chondrocytes on an uncalcified fibrocartilage layer (B) and a calcified fibrocartilage layer (C). Human osteoblasts on a bone layer (D). Scale bar represents 100 lm. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
8
KIM ET AL.
HUMAN COLLAGEN-BASED MULTI-LAYER SCAFFOLDS
ORIGINAL ARTICLE
FIGURE 8. Macroscopic images and histological analysis of a collagen-based multilayer scaffold without cells (A) and with cells in culture media on day 28 (B). The sections were stained with hematoxylin and eosin (1–4); tendon layer (A1), uncalcified fibrocartilage layer (A2), calcified fibrocartilage layer (A3), bone layer (A4), tendon layer with fibroblasts (B1), uncalcified fibrocartilage layer with chondrocytes (B2), calcified fibrocartilage layer with chondrocytes (B3), and bone layer with osteoblasts (B4). Black and green scale bars represent 500 and 100 lm, respectively. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
of 20 and 350 lm was fabricated by CS crosslinking and HA nucleation on a collagen matrix for distribution of different cells in tendon-to-bone tissues [Fig. 3(A)]. Four different layers corresponding to the tendon, uncalcified fibrocartilage, calcified fibrocartilage, and bone provided adequate support for cellular adhesion and proliferation, which indirectly demonstrated that the human collagen-based multilayer scaffolds had good biocompatibility. The proliferation of human synovial membrane fibroblasts was superior to that of human chondrocytes and osteoblasts in the tendon layer for 21 days [Fig. 6(A)]. In addition, human osteoblasts showed outstanding proliferation within the bone layer
compared with other cell types [Fig. 6(D)]. Similarly, human chondrocytes demonstrated higher growth rates in the fibrocartilage layer than fibroblasts and osteoblasts. In particular, fibroblasts in the uncalcified fibrocartilage layer adjacent to the tendon layer proliferated more rapidly than osteoblasts [Fig. 6(B)]. The number of osteoblasts in the calcified fibrocartilage layer adjoining a bone layer, on the other hand, was somewhat higher than that of fibroblasts for 21 days [Fig. 6(C)]. Furthermore, each cell type in each layer of the multilayer scaffold showed good viability, spreading, and growth as confirmed by microscopic images and histological analysis (Figs. 7 and 8). These results
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MONTH 2014 VOL 00A, ISSUE 00
9
demonstrate that by engineering different chemical and physical properties for each layer one can achieve a microenvironment tailored for each cell type, and human collagen-based multilayer scaffolds could support the formation of tendon-to-bone interface tissue. CONCLUSIONS
We have reported the fabrication and in vitro study of a multilayer scaffold designed to regenerate tendon-to-bone interface tissue. The stratified scaffold was composed of four layers with different chemical compositions. Each layer closely resembled natural tendon, fibrocartilage, and bone regions and was prepared by CS crosslinking and artificial calcification of a porous collagen matrix. The chemically graded scaffold was made with hybrid constructs and provided a gradual transition in microenvironment and mechanical properties. The main cell types in the tendon-tobone interface tissue—human fibroblasts, chondrocytes, and osteoblasts—showed robust proliferation in each corresponding matrix. Overall, our results show great promise for the application of functional multilayer collagen-based scaffolds to help tendon-to-bone healing in arthroscopic surgery. REFERENCES 1. Liu L, Xiong Z, Zhang R, Jin L, Yan Y. A novel osteochondral scaffold fabricated via multi-nozzle low-temperature deposition manufacturing. J Bioact Compat Polym 2009;24:18–30. 2. Wahl DA, Sachlos E, Liu C, Czernuszka J. Controlling the processing of collagen-hydroxyapatite scaffolds for bone tissue engineering. J Mater Sci Mater Med 2007;18:201–209. 3. Tampieri A, Sandri M, Landi E, Pressato D, Francioli S, Quarto R, Martin I. Design of graded biomimetic osteochondral composite scaffolds. Biomaterials 2008;29:3539–3546. 4. Mimura T, Imai S, Kubo M, Isoya E, Ando K, Okumura N, Matsusue Y. A novel exogenous concentration-gradient collagen scaffold augments full-thickness articular cartilage repair. Osteoarthr Cartilage 2008;16:1083–1091. 5. Chatterjee K, Lin-Gibson S, Wallace WE, Parekh SH, Lee YJ, Cicerone MT, Young MF, Simon CG Jr. The effect of 3D hydrogel scaffold modulus on osteoblast differentiation and mineralization revealed by combinatorial screening. Biomaterials 2010;31:5051– 5062. 6. Liu L, Xiong Z, Yan Y, Zhang R, Wang X, Jin, L. Multinozzle lowtemperature deposition system for construction of gradient tissue engineering scaffolds. J Biomed Mater Res B 2009;88:254–263. 7. Hsu YH, Turner IG, Miles AW. Fabrication of porous bioceramics with porosity gradients similar to the bimodal structure of cortical and cancellous bone. J Mater Sci Mater Med 2007;18:2251–2256. 8. Teng SH, Lee EJ, Wang P, Jun SH, Han CM, Kim HE. Functionally gradient chitosan/hydroxyapatite composite scaffolds for controlled drug release. J Biomed Mater Res B 2009;90:275–282. 9. Dormer NH, Singh M, Wang L, Berkland CJ, Detamore MS. Osteochondral interface tissue engineering using macroscopic gradients of bioactive signals. Ann Biomed Eng 2010;38:2167–2182. 10. Moffat KL, Sun WH, Pena PE, Chahine NO, Doty SB, Ateshian GA, Hung CT, Lu HH. Characterization of the structure–function relationship at the ligament-to-bone interface. Proc Natl Acad Sci USA 2008;105:7947–7952. 11. Samavedi S, Guelcher SA, Goldstein AS, Whittington AR. Response of bone marrow stromal cells to graded co-electrospun scaffolds and its implications for engineering the ligament-bone interface. Biomaterials 2012;33:7727–7735. 12. Samavedi S, Olsen Horton C, Guelcher SA, Goldstein AS, Whittington AR. Fabrication of a model continuously graded coelectrospun mesh for regeneration of the ligament–bone interface. Acta Biomater 2011;7:4131–4138.
10
KIM ET AL.
13. Spalazzi JP, Dagher E, Doty SB, Guo XE, Rodeo SA, Lu HH. In vivo evaluation of a multiphased scaffold designed for orthopaedic interface tissue engineering and soft tissue-to-bone integration. J Biomed Mater Res A 2008;86A:1–12. 14. Spalazzi JP, Doty SB, Moffat KL, Levine WN, Lu HH. Development of controlled matrix heterogeneity on a triphasic scaffold for orthopedic interface tissue engineering. Tissue Eng 2006;12:3497–3508. 15. Niyibizi C, Sagarrigo Visconti C, Gibson G, Kavalkovich K. Identification and immunolocalization of type X collagen at the ligament-bone interface. Biochem Biophys Res Commun 1996; 222:584–589. 16. Benjamin M, Kumai T, Milz S, Boszczyk BM, Boszczyk AA, Ralphs JR. The skeletal attachment of tendons - Tendon ‘entheses’. Comp Biochem Phys A 2002;133:931–945. 17. Benjamin M, Evans EJ, Copp L. The histology of tendon attachments to bone in man. J Anat 1986;149:89–100. 18. Kim BS, Choi JS, Kim JD, Yoon HI, Choi YC, Cho YW. Human collagen isolated from adipose tissue. Biotechnol Prog 2012;28:973–980. 19. Kim BS, Choi JS, Kim JD, Choi YC, Cho YW. Recellularization of decellularized human adipose-tissue-derived extracellular matrix sheets with other human cell types. Cell Tissue Res 2012;348: 559–567. 20. Ravichandran R, Venugopal JR, Sundarrajan S, Mukherjee S, Ramakrishna S. Precipitation of nanohydroxyapatite on PLLA/ PBLG/Collagen nanofibrous structures for the differentiation of adipose derived stem cells to osteogenic lineage. Biomaterials 2012;33:846–855. 21. Cao H, Xu SY. EDC/NHS-crosslinked type II collagen-chondroitin sulfate scaffold: Characterization and in vitro evaluation. J Mater Sci Mater Med 2008;19:567–575. 22. Sachlos E, Gotora D, Czernuszka JT. Collagen scaffolds reinforced with biomimetic composite nano-sized carbonate-substituted hydroxyapatite crystals and shaped by rapid prototyping to contain internal microchannels. Tissue Eng 2006;12:2479–2487. 23. Potter K, Kidder LH, Levin IW, Lewis EN, Spencer RG. Imaging of collagen and proteoglycan in cartilage sections using Fourier transform infrared spectral imaging. Arthritis Rheum 2001;44:846–855. 24. Cavalcanti OA, Da Silva CC, Pineda EAG, Hechenleitner AAW. Synthesis and characterization of phosphated crosslinked chondroitin sulfate: Potential ingredient for specific drug delivery. Acta utica Bonaerense 2005;24:234–238. Farmace 25. Servaty R, Schiller J, Binder H, Arnold K. Hydration of polymeric components of cartilage - An infrared spectroscopic study on hyaluronic acid and chondroitin sulfate. Int J Biol Macromol 2001; 28:121–127. 26. Marelli B, Ghezzi CE, Mohn D, Stark WJ, Barralet JE, Boccaccini AR, Nazhat SN. Accelerated mineralization of dense collagennano bioactive glass hybrid gels increases scaffold stiffness and regulates osteoblastic function. Biomaterials 2011;32:8915–8926. 27. Tampieri A, Celotti G, Landi E, Sandri M, Roveri N, Falini G. Biologically inspired synthesis of bone-like composite: Selfassembled collagen fibers/hydroxyapatite nanocrystals. J Biomed Mater Res 2003;67:618–625. 28. Kovacevic D, Rodeo SA. Biological augmentation of rotator cuff tendon repair. Clin Orthop Relat Res 2008;466:622–633. 29. van Tienen TG, Heijkants RG, Buma P, de Groot JH, Pennings AJ, Veth RP. Tissue ingrowth and degradation of two biodegradable porous polymers with different porosities and pore sizes. Biomaterials 2002;23:1731–1738. 30. Haugh MG, Murphy CM, McKiernan RC, Altenbuchner C, O’Brien FJ. Crosslinking and mechanical properties significantly influence cell attachment, proliferation, and migration within collagen glycosaminoglycan scaffolds. Tissue Eng Part A 2011;17:1201–1208. 31. Bonzani IC, George JH, Stevens MM. Novel materials for bone and cartilage regeneration. Curr Opin Chem Biol 2006;10:568–575. 32. Pattanayak DK, Divya P, Upadhyay S, Prasad RC, Rao BT, Rama Mohan TR. Synthesis and evaluation of hydroxyapatite ceramics. Trends Biomater Artif Organs 2005;18:87–92. 33. Bose S, Tarafder S. Calcium phosphate ceramic systems in growth factor and drug delivery for bone tissue engineering: A review. Acta Biomater 2012;8:1401–1421.
HUMAN COLLAGEN-BASED MULTI-LAYER SCAFFOLDS
ORIGINAL ARTICLE
34. O’Brien FJ, Harley BA, Yannas IV, Gibson LJ. The effect of pore size on cell adhesion in collagen-GAG scaffolds. Biomaterials 2005;26:433–441. 35. Bobyn JD, Wilson GJ, MacGregor DC, Pilliar RM, Weatherly GC. Effect of pore size on the peel strength of attachment of fibrous tissue to porous-surfaced implants. J Biomed Mater Res 1982;16: 571–584. 36. Freeman JW. Tissue engineering options for ligament healing. Bone Tissue Regen Insights 2009;2:13–23. 37. Klompmaker J, Jansen HW, Veth RP, Nielsen HK, de Groot JH, Pennings AJ. Porous implants for knee joint meniscus reconstruc-
tion: A preliminary study on the role of pore sizes in ingrowth and differentiation of fibrocartilage. Clin Mater 1993;14:1–11. 38. Spector M, Michno MJ, Smarook WH, Kwiatkowski GT. A highmodulus polymer for porous orthopedic implants: Biomechanical compatibility of porous implants. J Biomed Mater Res 1978;12: 665–677. 39. Bobyn JD, Pilliar RM, Cameron HU, Weatherly GC. The optimum pore size for the fixation of porous surfaced metal implants by the ingrowth of bone. Clin Orthop Relat Res 1980;150:263–270. 40. Clemow AJT, Weinstein AM, Klawitter JJ, Koeneman J, Anderson J. Interface mechanics of porous titanium implants. J Biomed Mater Res 1981;15:73–82.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MONTH 2014 VOL 00A, ISSUE 00
11
Department of Chemical Engineering, Hanyang University, Hanyangdaehak-ro 55, Ansan, Kyeonggi-do 426-791, Republic of Korea 2 Center for Theragnosis, Biomedical Research Institute, Korea Institute of Science and Technology (KIST), Hwarangno 14-gil 5, Seoul 136-791, Republic of Korea 3 School of Pharmacy, Sungkyunkwan University, Suwon 440-746, Republic of Korea 4 Department of Orthopedic Surgery, Joint and Spine Center, SMG-SNU Boramae Medical Center, Seoul National University College of Medicine, 20 Boramae-ro 5-gil, Seoul 156-707, Republic of Korea Received 23 October 2013; accepted 6 December 2013 Published online 00 Month 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.a.35057 Abstract: The natural tendon-to-bone region has a gradient in structure and composition, which is translated into a spatial variation of chemical, physical, and biological properties. This unique transitional tissue between bone and tendon is not normally recreated during natural bone-to-tendon healing. In this study, we have developed a human collagen-based multilayer scaffold mimicking the tendon-to-bone region. The scaffold consists of four different layers with the following composition gradient: (a) a tendon layer composed of collagen; (b) an uncalcified fibrocartilage layer composed of collagen and chondroitin sulfate; (c) a calcified fibrocartilage layer composed of collagen and less apatite; (d) a bone layer composed of collagen and apatite. The chemical, physical, and mechanical properties of the scaffold were characterized by a scanning
electron microscope, porosimeter, universal tensile machine, Fourier transform infrared spectrometer, energy dispersive X-ray analysis apparatus, and thermogravimetric analysis apparatus. The multilayer scaffold provided a gradual transition of the physical, chemical, and mechanical environment and supported the adhesion and proliferation of human fibroblasts, chondrocytes, and osteoblasts toward each corresponding matrix. Overall, our results suggest the feasibility of a human collagen-based multilayer scaffold for regeneration C 2014 Wiley Periodicals, Inc. J of hard-to-soft interface tissues. V Biomed Mater Res Part A: 00A:000–000, 2014.
Key Words: adipose tissue, collagen, tendon-to-bone, fibrocartilage, calcification
How to cite this article: Kim BS, Kim EJ, Choi JS, Jeong JH, Jo CH, Cho YW. 2014. Human collagen-based multilayer scaffolds for tendon-to-bone interface tissue engineering. J Biomed Mater Res Part A 2014:00A:000–000.
INTRODUCTION
Traditional tissue engineering uses cells, biomaterials, and environmental factors to regenerate specific tissues such as skin, bone, cartilage, and nerve. Recent advances in tissue engineering have led to the engineering of functional interface tissues, which bridge the gap between two tissues that differ in their physical, chemical, and biological properties. Interface tissue engineering may need scaffolds made of multilayer or graduated biomaterials for the regeneration of continuous transitional regions. Indeed, recent gradient interfacial designs of scaffolds have addressed mineralization,1–3 matrix composition,3–6 pore size distribution or porosity,6,7 and bioactive signals8,9 for resolution of transitioning orthopedic interface tissues such as the cortical–trabecular bone interface, the cartilage–bone (osteochondral) interface, and the ligament/tendon-to-bone interface. In particular, the ligament/tendon-to-bone interface has been
explored recently.10 Ligaments and tendons have similar structure and composition, including a layering of type I collagen and proteoglycans. Both tissues also lack a good blood supply and do not heal well, requiring surgery if torn or injured. Therefore, many studies are being performed to formulate a gradient scaffold design for use in orthopedic surgery.11–14 The tendon-to-bone interface is characterized by a transition region of generally four types of tissue, with controlled spatial variations in cell type and matrix composition. The first region is an area of fibrous connective tissue found in normal tendon composed of aligned collagen fibrils with fibroblasts embedded throughout a type I and type III collagen matrix. The second is an uncalcified fibrocartilage layer, which contains ovoid chondrocytes, collagen types I and II, and aggrecan. The third is a calcified fibrocartilage region, which is composed of hypertrophic, highly
Correspondence to: Y.W. Cho; e-mail: [email protected] and C.H. Jo; e-mail: [email protected] Contract grant sponsor: National Research Foundation of Korea; contract grant number: 2011-0019774
C 2014 WILEY PERIODICALS, INC. V
1
circular chondrocytes and collagen types I, II, and X.15 The last region is the subchondral bone layer. The bone tissue contains osteoblasts, osteocytes, and osteoclasts with calcified type I collagen fibers. In the all regions of the interface, however, there are dense networks of collagen fibrils interconnecting heterogeneous tissues. The mechanical properties of tendon and bone are significantly different. Tendons have a high tensile strength, whereas bone is optimized for compressive loading. In contrast, the transitional regions such as the uncalcified and calcified fibrocartilage layers are superior in transferring compressive load, and the insertion zones show an increase in Young’s modulus across the interface.10 Thus, the intermediate regions bridge the gap between the mechanical properties of tendon and bone.16 This controlled matrix heterogeneity serves to minimize the formation of stress concentrations and promote a gradual transition of mechanical loading between the soft and hard tissue.17 Similarly, in the design of engineered scaffolds for regeneration of multitissue interfaces, the existence of intermediate layers is vital for soft-to-hard tissue integration, fixation, and stability. The goal of this study is to design, fabricate, and in vitro evaluate human collagen-based, multilayer scaffolds for regeneration of the interface between tendon and bone. We hypothesized that the functional properties of engineered tendon-to-bone interface tissue would depend on the chemical composition of the scaffolds. Therefore, multilayer scaffolds with the following four separate layers were fabricated: (i) an upper layer of collagen matrix, mimicking the tendon region; (ii) a first intermediate layer of crosslinked aggrecan–collagen resembling the uncalcified fibrocartilage region; (iii) a second intermediate layer of partially calcified collagen resembling the calcified fibrocartilage region; (iv) a lower layer of calcified collagen with a high degree of mineralization corresponding to subchondral bone. Thereafter, the in vitro responses of cells to the engineered multilayer scaffolds were investigated in terms of cell adhesion, viability, and proliferation using human fibroblasts, chondrocytes, and osteoblasts. It is anticipated that this collagen-based multilayer scaffold composed of four different layers will result in a successful graft for tendon-tobone interface tissue engineering. MATERIALS AND METHODS
Material processing and scaffold fabrication Collagen was extracted from human adipose tissue, according to a previously developed protocol.18 In brief, adipose tissue obtained by liposuction was washed several times with distilled water to remove blood components. The crude extracellular matrix (ECM) was isolated from adipose tissue by homogenization, centrifugation, and rinsing.19 The ECM was solubilized in 0.5M acetic acid containing 0.1% pepsin (EC 3.4.23.1 powder; Sigma-Aldrich, St. Louis, MO) after alkaline (0.1M NaOH) and alcohol (70% isopropanol) treatments. The resulting viscous solution was centrifuged, and the supernatant was selectively precipitated with 0.9M NaCl. Precipitated collagen was dissolved in 0.5M acetic acid and desalted by a centrifugal filter device (Amicon Ultra-15;
2
KIM ET AL.
Millipore, Billerica, MA). The concentration of collagen was measured using a soluble collagen assay kit (Sircol, Biocolor, Carrickfergus, Northern Ireland), and the purified collagen was freeze-dried. To mimic the tendon-to-bone interface with a composition gradient, a scaffold consisting of the following four different layers was prepared: (a) a tendon layer composed of collagen; (b) an uncalcified fibrocartilage layer composed of collagen and chondroitin sulfate; (c) a calcified fibrocartilage layer composed of collagen and apatite (low calcification); (d) a bone layer composed of collagen and apatite (high calcification). Calcification was performed on the collagen matrices to precipitate hydroxyapatite (HA) by a calcium-phosphate dipping method.20 Mimicking the tendon layer. Collagen dispersion of 0.8% w/v was prepared. The respective mass of collagen was soaked in 0.02N acetic acid for 24 h at 4oC. To prevent heat build-up, the mixture was homogenized on a bed of ice for 10 min using a homogenizer (T10 basic Ultra-Turrax; IKA, Staufen, Germany). Air bubbles were removed by centrifugation at 1000 rpm for 20 min at 4oC. The collagen dispersion was cast into a mold, frozen at 220oC, and freeze dried. The prepared collagen matrix was disk-shaped with a thickness of 1 mm and a diameter of 10 mm. The freeze-dried collagen matrix was cross-linked by 1-ethyl-3(3-dimethyl aminopropyl) carbodiimide (EDC) and N-hydroxysuccinimide (NHS). Mimicking the uncalcified fibrocartilage layer. The freezedried collagen matrix was cross-linked in the presence of chondroitin sulfate (CS; Tokyo Chemical Industry, Tokyo, Japan) by EDC and NHS. Collagen matrices were immersed and incubated in 40% v/v ethanol containing 50 mM 2-morpholinoehane sulfonic acid (MES) for 30 min at room temperature. Subsequently, matrices were incubated in 40% v/v ethanol containing 50 mM MES and 2% w/v CS for 4 h at room temperature with 5 mg/mL EDC to yield a collagen–CS composite at a ratio of 9:1.21 NHS was added in an EDC:NHS ratio of 4:1. After the reaction, excess EDC and CS were rinsed from the matrix using 0.1M Na2HPO4 for 1 h. Mimicking the calcified fibrocartilage layer. Collagen matrices were immersed in 500 mM calcium chloride dehydrate solution (CaCl22H2O; Sigma-Aldrich) for 5 min and rinsed for 1 min in distilled water. Subsequently, the matrices were immersed in 300 mM disodium hydrogen phosphate solution (Na2HPO4; Duksan Pure Chemical, Ansan, Korea) for 5 min and rinsed for 1 min in distilled water. Both calcium and phosphate solutions contained 0.1M potassium chloride (KCl; Sigma-Aldrich) to maintain ionic stability and were adjusted to pH 8.22 The Ca/P molar ratio of the precipitates at pH 8 exhibited a close stoichiometric composition (Ca/P 5 1.67) of HA. Mimicking the bone layer. The above calcification procedure represented a single cycle. Collagen matrices were subjected to three cycles of the calcification treatment. Individually prepared layers were rinsed with distilled water and freeze-dried.
HUMAN COLLAGEN-BASED MULTI-LAYER SCAFFOLDS
ORIGINAL ARTICLE
Scaffold characterization The morphology of each layer was observed by scanning electron microscopy (SEM; S-4800 FE-SEM, Hitachi, Tokyo, Japan) at an accelerating voltage of 10 kV, after platinum coating by a sputter coater (Emitech k575x; Quorum Technologies, Wellington, UK). Fourier transform infrared (FTIR) spectra were obtained in an attenuate total reflectance mode with a Nicolet 6700 FTIR spectrometer equipped with a diamond probe (Thermo Fisher Scientific, Waltham, MA). The results are an average of 256 scans in a spectral range between 1800 and 800 cm21. Pore size and the porosity of each layer were determined using an automated mercury porosimeter (Autopore IV 9500, Micromeritics, Norcross, GA). The compressive strength of each layer was measured using a universal tensile machine (H100KS, Tinius Olsen, Horsham, PA) with a 50-N load cell. The weight fraction of HA in a composite was measured by a thermogravimetric analysis (TGA) apparatus (TG-DTA 2000SA; Bruker AXS, Billerica, MA) upon heating from 25 to 800oC at 10oC/min under N2 atmosphere. The HA composites were examined by SEM equipped with an energy dispersive X-ray analysis apparatus (EDX; EX-250, Horiba, Kyoto, Japan). Cell culture and seeding Human synovial membrane fibroblasts, human chondrocytes, and human osteoblasts were purchased from Applied Biological Materials (abm; NY; T4018, T0020, and T4015). Cells were cultured and expanded in T25 ECM-coated flasks (G299, abm). Fibroblasts were cultured in Progrow III medium (TM003, abm) supplemented with 10% fetal bovine serum (FBS; TM999, abm) and 1% penicillin/streptomycin (P/S; Invitrogen, Carlsbad, CA). Chondrocytes and osteoblasts were also cultured in complete medium containing Progrow IV medium (TM004, abm), FBS, and P/S (1/0.1/ 0.01 volume ratio). Cells were fed every 3 days and cultured to confluence at 37oC under 5% CO2 in a humidified atmosphere. Passage of five to seven cells was used throughout. After being sterilized by ethylene oxide gas, the four types of layers (thickness 1 mm, diameter 10 mm) were incubated in culture medium for 2 h before cell seeding. The equilibrated layer was placed into each well of a non-treated 24-well polystyrene plate (32024, SPL Life Sciences, Pocheon, Korea). All three types of cells were trypsinized and resuspended at a concentration of 5 3 105 cells per 50 mL media. Ten microliters of cell suspension (1 3 105 cells) was added to each layer. After 3 h, 1 mL of the culture medium was added to each well. Cell-seeded layers were incubated at 37oC and 5% CO2 and cultured for 3 weeks. Media were changed every 2 or 3 days for the duration of all experiments. Cell adhesion and proliferation The proliferation of human cells in a scaffold was analyzed by a water soluble tetrazolium salts (WST) assay, which is based on the reduction of a stable tetrazolium salt to a soluble violet formazan product by viable cells. After 1, 7, 14, and 21 days of culture, the cell-seeded layers were rinsed with phosphate-buffered saline and then incubated with
1 mL of 10% WST (Hoffmann-La Roche, Basel, Switzerland) solution in the dark at 37oC under 5% CO2 for 30 min. After gentle pipetting, the 100 lL solution was transferred to a 96-well plate. The optical density was then measured at 440 nm using a microplate spectrophotometer (PowerWave XS; Bio-Tek Instruments, Winooski, VT). For the cell viability analysis, the commercially available live/dead viability/cytotoxicity kit (Molecular Probes, Eugene, OR) was used. Green fluorescence caused by the reaction of calcein with intracellular esterase indicated live cells, whereas red fluorescence caused by ethidium homodimers that bound to nucleic acids indicated dead cells. Cell-seeded layers were stained with the combined live/dead reagents (Molecular Probes) at 37oC under 5% CO2 for 30 min and observed using a fluorescence microscope (IX81, Olympus, Tokyo, Japan). The morphology of the cells attached to the layers was examined by SEM. Cell-layer constructs were fixed in 2.5% glutaraldehyde for 30 min. After cell fixation, they were dehydrated through a gradient series of ethanol solutions from 50 to 100% in increments of 10% for 10 min each and freezedried under vacuum for 2 days. Samples were coated with platinum by sputtering and were then observed by SEM at an accelerating voltage of 15 kV. Histological evaluation was performed after 28 days of culture. Cell-scaffold constructs were composed of an osteoblast–bone layer, a chondrocytecalcified fibrocartilage layer, a chondrocyte-uncalcified fibrocartilage layer, and a fibroblast–tendon layer at 21 days postseeding. The four layers were piled up and then cultured for 7 days in Dulbecco’s modified Eagle’s medium (DMEM; Gibco-BRL, Life Technologies, Carlsbad, CA) supplemented with 10% FBS and 1% P/S. The samples were fixed with 4% formalin, embedded in paraffin, and stained with hematoxylin-eosin (HE; Sigma-Aldrich). Statistical analysis All data are represented as means 6 standard deviation. Statistical computations used the Tukey method of one-way analysis of variance. p-values less than 0.05 were considered significant. RESULTS
Isolated collagen fibrils from human adipose tissue are shown in a high-magnification SEM image (Fig. 1). The characteristic D-band of collagen fibrils was clearly observed, and the diameter of the collagen fibrils was approximately 300 lm. Figure 2 shows the cross-sectional morphologies of each layer of the multilayer scaffold. All layers exhibited the typical porous structure of scaffolds prepared by a freezedrying method. There were no significant differences in the appearance of the structures between the tendon layer composed of collagen and the uncalcified fibrocartilage layer composed of collagen and CS, whereas the calcified fibrocartilage layer and the bone layer showed distinct morphologies because of HA crystals on the surface of the collagen matrices. Of note, the bone layer exhibited a smaller pore size because the collagen matrices were infiltrated by many HA crystal agglomerates. These results were also confirmed
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MONTH 2014 VOL 00A, ISSUE 00
3
FIGURE 1. Scanning electron micrographs of collagen isolated from human adipose tissue. Scale bar represents 1 lm.
using a mercury intrusion porosimeter [Fig. 3(A)]. The proportion of material with a pore size below 40 lm in the bone layer was 71.08%, whereas that in the tendon layer was 34.79%. A pore size of under 40 lm was observed most commonly for all layers, whereas a size of 180–350 lm was the least common. The average porosity for all layers was above 93%. The mechanical properties measured by compressive experiments are shown in Figure 3(B). The mechanical property of each layer was dependent on the chemical composition, such as the content of collagen, CS,
and HA. The elastic moduli of the tendon, uncalcified fibrocartilage, calcified fibrocartilage, and bone layer were 303 6 131, 1181 6 333, 2954 6 1143, and 4469 6 1650 kPa, respectively, whereas the typical elongation at break values were 112.5 6 13.1, 105.8 6 13.7, 82.0 6 8.8, and 71.2 6 13.8%, respectively. These results obviously show that the varying ingredients of each layer provide different mechanical properties at different layers of the graded collagen-based scaffolds. The FTIR spectra of tendon, uncalcified fibrocartilage, and bone layers are shown in Figure 4. All layers had the three major amide absorption bands of collagen, namely amide I, II, and III at 1628–1647, 1539–1560, and 1233– 1243 cm21, respectively. They were comparable to the spectra of collagen from other sources.18 The characteristic absorbance bands for CS are 1603–1611 and 1411–1413 cm21 (COO–), 1225–1258 cm21 (SO42–), and sugar groups in the range of 983–1285 cm21.23–25 The spectrum of an uncalcified fibrocartilage layer containing CS also displayed peaks corresponding to COO2(1408 cm21), SO42– (1263 cm21), and sugar (970–1120 cm21) groups [Fig. 4(A)]. On the other hand, the carboxylate absorbance band (approximately 1610 cm21) was not found because of overlap with the amide I band of collagen. In the bone layer containing HA, the vibrational bands of phosphate and carbonate were visible at 1020 and 873 cm21, respectively [Fig. 4(B)].22,26 Figure 5 shows calcification of collagen matrices evident on SEM, the quantification of HA concentration by TGA, and the
FIGURE 2. Cross-sectional images of the tendon layer (A), uncalcified fibrocartilage layer (B), calcified fibrocartilage layer (C), and bone layer (D). Scale bar represents 200 lm.
4
KIM ET AL.
HUMAN COLLAGEN-BASED MULTI-LAYER SCAFFOLDS
ORIGINAL ARTICLE
FIGURE 3. Pore size distribution (A) and mechanical properties (B) of the tendon, uncalcified fibrocartilage, calcified fibrocartilage, and bone layers. Data presented are means 6 standard deviation (p < 0.05).
ratio of calcium to phosphate values by EDX. SEM images of the calcified fibrocartilage and bone layer [Fig. 5(A,B), respectively] qualitatively show an increase in the homogenous calcification proportional to the number of times the material was dipped in the calcium and phosphate solutions. The higher magnification SEM image showed the D-band of collagen fibrils and HA nano-crystals on the collagen framework [Fig. 5(C)]. TGA comparatively showed the HA content of the calcified fibrocartilage and bone layers [Fig. 5(D)]. The first weight loss observed between 25 and 100oC is associated with the removal of water in each composite,27 whereas the second weight loss occurring at 250–550oC corresponds to the loss of organic content. The collagen/HA (organic/inorganic) mass ratios of the calcified fibrocartilage and bone layers were 51/49 and 27/73, respectively. EDX analyses showed rich calcium phosphate deposits, and the Ca/P ratios of deposits in the calcified fibrocartilage and bone layer were 1.63 6 0.14 and 1.62 6 0.06, respectively [Fig. 5(E,F)], which were close to biological apatite (Ca/P ratio 1.67). Cell adhesion and proliferation in each layer of the gradient scaffold was evaluated using a WST assay as shown in Figure 6. The proliferation of fibroblasts on the tendon layer
FIGURE 4. ATR-FTIR spectra of the tendon layer composed of collagen, the uncalcified fibrocartilage layer composed of collagen and chondroitin sulfate (A), and the bone layer composed of collagen and HA (B). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
was significantly higher (p 0.05) than those of chondrocytes and osteoblasts on day 21. Similarly, the proliferation of chondrocytes on the uncalcified fibrocartilage layer and that of osteoblasts on the bone layer were the highest compared with other cell types on day 21. On the other hand, no significant difference was observed between chondrocytes and osteoblasts on the calcified fibrocartilage layer on day 21. However, the percentage increases in the proliferation rates of chondrocytes and osteoblasts on the calcified fibrocartilage layer from day 7 to day 21 were found to be 263 and 193%, respectively. Overall, every type of cell showed good adhesion and proliferation on all the prepared layers, and all cell numbers significantly increased during the experimental period. The left column of Figure 7 shows the live/dead staining images on day 7, and the right column shows SEM micrographs on day 14, fibroblasts on the tendon layer [Fig. 7(A)], chondrocytes on uncalcified
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MONTH 2014 VOL 00A, ISSUE 00
5
FIGURE 5. SEM images of the calcified fibrocartilage layer (A) and the bone layer (B). (C) is higher magnification of (A) and shows that the nano-HA particles were formed on the collagen fibrils. TGA curves of collagen/HA composites (D). EDX spectra of HA on the collagen matrices of the calcified fibrocartilage layer (E) and bone layer (F). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
fibrocartilage [Fig. 7(B)], and calcified fibrocartilage layers [Fig. 7(C)] and osteoblasts on the bone layer [Fig. 7(D)]. In the live/dead assay, live cells were stained green by calceinAM and dead cells were stained red by an ethidium homodimer. After 7 days of culture, a uniform cell distribution was observed in each layer and most cells expressed green fluorescence. The SEM results showed that all human cells at a high density spread well on the surface of each scaffold. The histological appearance of the cell-seeded multilayer scaffold was observed by H&E staining (Fig. 8). All sections of the cell-seeded scaffold displayed dense structures and a large amount of secreted and accumulated ECM when compared with unseeded scaffolds.
6
KIM ET AL.
DISCUSSION
Kovacevic and Rodeo28 reported the histological images of a normal supraspinatus tendon insertion site, clearly showing that there were four distinct zones in the tendon-to-bone interface tissue: tendon, unmineralized fibrocartilage, mineralized fibrocartilage, and bone. Our research goal was to engineer a multilayer scaffold to mimic the natural tendonto-bone interface tissue. In an effort to achieve this goal, a scaffold composed of four layers with different chemical compositions was fabricated. We hypothesized that different constituents of each layer could provide graded mechanical properties in a scaffold and favorable environment for each cell type. The stratified scaffold system is based on human
HUMAN COLLAGEN-BASED MULTI-LAYER SCAFFOLDS
ORIGINAL ARTICLE
FIGURE 6. Proliferation of human synovial membrane fibroblasts, chondrocytes, and osteoblasts in the tendon (A), uncalcified fibrocartilage (B), calcified fibrocartilage (C), and bone layers (D) for 21 days examined by a WST assay. Data presented are means 6 standard deviation. * p < 0.05.
collagen, the main component of the natural tendon-to-bone interface tissue. Collagen possesses attractive cell binding properties required for cells to attach to the scaffold. In a previous study, we successfully isolated and purified human collagen from adipose tissue.18 The highly porous collagen matrix with high interconnectivity was prepared to resemble the tendon layer [Fig. 2(A)]. High porosity and pore interconnections affect not only cell infiltration, binding, and migration, but also the transport of oxygen, nutrients, and waste exchange.29 Then, a collagen–CS layer was fabricated to mimic a native uncalcified fibrocartilage region. EDC and NHS catalyze a reaction between amine groups of collagen and carboxylic acid groups of CS to form crosslinks.30 EDC activates a carboxylic acid of CS and then reacts with NHS to produce an NHS-activated carboxylic acid. Finally, the NHS-activated carboxylic acid reacts with an amine group of collagen to produce a crosslink. EDC/NHS cross-linked collagen–CS improves mechanical stability, and creates a favorable environment for chondrocytes.21 The results of this study also indicate that the presence of CS cross-linked onto a collagen matrix affects mechanical properties [Fig. 3(B)]. To mimic the calcified fibrocartilage and the bone region, the composites of both layers were composed of collagen
and HA. HA improves not only the mechanical properties, but also the bioactivity of the scaffold by providing a source of calcium and phosphate ions. The nucleation of HA onto a collagen matrix results from a high affinity for Ca21 ions and the formation of a hydrogen-bonded hypercomplex of PO42– ions to link the carboxylic acid groups of collagen.3,31 HA synthesis generally requires sintering at increased temperature, usually in excess of 1000oC,32 but our calcification process was conducted at room temperature to avoid collagen denaturation. Nevertheless, EDX profiles of the matrices confirmed that the values of the Ca/P ratio of the nucleated HA were not in the range of amorphous HA, but rather in the range of crystalline HA.33 This result shows that HA with a moderate Ca/P ratio can be created by control of pH values. The microstructures of a scaffold such as pore size, porosity, and interconnectivity have been shown to significantly influence cell behaviors such as adhesion, migration, and proliferation.34 Several studies have shown that the optimal pore sizes for cell ingrowth and fixation strength of tendon, fibrocartilage, and bone tissues are 20–250, 150– 500, and 50–400 lm, respectively.35–40 In this work, a layered scaffold with pore volume gradients between pore size
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MONTH 2014 VOL 00A, ISSUE 00
7
FIGURE 7. Fluorescence micrographs on day 7 and SEM images on day 14. Human synovial membrane fibroblasts on a tendon layer (A). Human chondrocytes on an uncalcified fibrocartilage layer (B) and a calcified fibrocartilage layer (C). Human osteoblasts on a bone layer (D). Scale bar represents 100 lm. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
8
KIM ET AL.
HUMAN COLLAGEN-BASED MULTI-LAYER SCAFFOLDS
ORIGINAL ARTICLE
FIGURE 8. Macroscopic images and histological analysis of a collagen-based multilayer scaffold without cells (A) and with cells in culture media on day 28 (B). The sections were stained with hematoxylin and eosin (1–4); tendon layer (A1), uncalcified fibrocartilage layer (A2), calcified fibrocartilage layer (A3), bone layer (A4), tendon layer with fibroblasts (B1), uncalcified fibrocartilage layer with chondrocytes (B2), calcified fibrocartilage layer with chondrocytes (B3), and bone layer with osteoblasts (B4). Black and green scale bars represent 500 and 100 lm, respectively. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
of 20 and 350 lm was fabricated by CS crosslinking and HA nucleation on a collagen matrix for distribution of different cells in tendon-to-bone tissues [Fig. 3(A)]. Four different layers corresponding to the tendon, uncalcified fibrocartilage, calcified fibrocartilage, and bone provided adequate support for cellular adhesion and proliferation, which indirectly demonstrated that the human collagen-based multilayer scaffolds had good biocompatibility. The proliferation of human synovial membrane fibroblasts was superior to that of human chondrocytes and osteoblasts in the tendon layer for 21 days [Fig. 6(A)]. In addition, human osteoblasts showed outstanding proliferation within the bone layer
compared with other cell types [Fig. 6(D)]. Similarly, human chondrocytes demonstrated higher growth rates in the fibrocartilage layer than fibroblasts and osteoblasts. In particular, fibroblasts in the uncalcified fibrocartilage layer adjacent to the tendon layer proliferated more rapidly than osteoblasts [Fig. 6(B)]. The number of osteoblasts in the calcified fibrocartilage layer adjoining a bone layer, on the other hand, was somewhat higher than that of fibroblasts for 21 days [Fig. 6(C)]. Furthermore, each cell type in each layer of the multilayer scaffold showed good viability, spreading, and growth as confirmed by microscopic images and histological analysis (Figs. 7 and 8). These results
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MONTH 2014 VOL 00A, ISSUE 00
9
demonstrate that by engineering different chemical and physical properties for each layer one can achieve a microenvironment tailored for each cell type, and human collagen-based multilayer scaffolds could support the formation of tendon-to-bone interface tissue. CONCLUSIONS
We have reported the fabrication and in vitro study of a multilayer scaffold designed to regenerate tendon-to-bone interface tissue. The stratified scaffold was composed of four layers with different chemical compositions. Each layer closely resembled natural tendon, fibrocartilage, and bone regions and was prepared by CS crosslinking and artificial calcification of a porous collagen matrix. The chemically graded scaffold was made with hybrid constructs and provided a gradual transition in microenvironment and mechanical properties. The main cell types in the tendon-tobone interface tissue—human fibroblasts, chondrocytes, and osteoblasts—showed robust proliferation in each corresponding matrix. Overall, our results show great promise for the application of functional multilayer collagen-based scaffolds to help tendon-to-bone healing in arthroscopic surgery. REFERENCES 1. Liu L, Xiong Z, Zhang R, Jin L, Yan Y. A novel osteochondral scaffold fabricated via multi-nozzle low-temperature deposition manufacturing. J Bioact Compat Polym 2009;24:18–30. 2. Wahl DA, Sachlos E, Liu C, Czernuszka J. Controlling the processing of collagen-hydroxyapatite scaffolds for bone tissue engineering. J Mater Sci Mater Med 2007;18:201–209. 3. Tampieri A, Sandri M, Landi E, Pressato D, Francioli S, Quarto R, Martin I. Design of graded biomimetic osteochondral composite scaffolds. Biomaterials 2008;29:3539–3546. 4. Mimura T, Imai S, Kubo M, Isoya E, Ando K, Okumura N, Matsusue Y. A novel exogenous concentration-gradient collagen scaffold augments full-thickness articular cartilage repair. Osteoarthr Cartilage 2008;16:1083–1091. 5. Chatterjee K, Lin-Gibson S, Wallace WE, Parekh SH, Lee YJ, Cicerone MT, Young MF, Simon CG Jr. The effect of 3D hydrogel scaffold modulus on osteoblast differentiation and mineralization revealed by combinatorial screening. Biomaterials 2010;31:5051– 5062. 6. Liu L, Xiong Z, Yan Y, Zhang R, Wang X, Jin, L. Multinozzle lowtemperature deposition system for construction of gradient tissue engineering scaffolds. J Biomed Mater Res B 2009;88:254–263. 7. Hsu YH, Turner IG, Miles AW. Fabrication of porous bioceramics with porosity gradients similar to the bimodal structure of cortical and cancellous bone. J Mater Sci Mater Med 2007;18:2251–2256. 8. Teng SH, Lee EJ, Wang P, Jun SH, Han CM, Kim HE. Functionally gradient chitosan/hydroxyapatite composite scaffolds for controlled drug release. J Biomed Mater Res B 2009;90:275–282. 9. Dormer NH, Singh M, Wang L, Berkland CJ, Detamore MS. Osteochondral interface tissue engineering using macroscopic gradients of bioactive signals. Ann Biomed Eng 2010;38:2167–2182. 10. Moffat KL, Sun WH, Pena PE, Chahine NO, Doty SB, Ateshian GA, Hung CT, Lu HH. Characterization of the structure–function relationship at the ligament-to-bone interface. Proc Natl Acad Sci USA 2008;105:7947–7952. 11. Samavedi S, Guelcher SA, Goldstein AS, Whittington AR. Response of bone marrow stromal cells to graded co-electrospun scaffolds and its implications for engineering the ligament-bone interface. Biomaterials 2012;33:7727–7735. 12. Samavedi S, Olsen Horton C, Guelcher SA, Goldstein AS, Whittington AR. Fabrication of a model continuously graded coelectrospun mesh for regeneration of the ligament–bone interface. Acta Biomater 2011;7:4131–4138.
10
KIM ET AL.
13. Spalazzi JP, Dagher E, Doty SB, Guo XE, Rodeo SA, Lu HH. In vivo evaluation of a multiphased scaffold designed for orthopaedic interface tissue engineering and soft tissue-to-bone integration. J Biomed Mater Res A 2008;86A:1–12. 14. Spalazzi JP, Doty SB, Moffat KL, Levine WN, Lu HH. Development of controlled matrix heterogeneity on a triphasic scaffold for orthopedic interface tissue engineering. Tissue Eng 2006;12:3497–3508. 15. Niyibizi C, Sagarrigo Visconti C, Gibson G, Kavalkovich K. Identification and immunolocalization of type X collagen at the ligament-bone interface. Biochem Biophys Res Commun 1996; 222:584–589. 16. Benjamin M, Kumai T, Milz S, Boszczyk BM, Boszczyk AA, Ralphs JR. The skeletal attachment of tendons - Tendon ‘entheses’. Comp Biochem Phys A 2002;133:931–945. 17. Benjamin M, Evans EJ, Copp L. The histology of tendon attachments to bone in man. J Anat 1986;149:89–100. 18. Kim BS, Choi JS, Kim JD, Yoon HI, Choi YC, Cho YW. Human collagen isolated from adipose tissue. Biotechnol Prog 2012;28:973–980. 19. Kim BS, Choi JS, Kim JD, Choi YC, Cho YW. Recellularization of decellularized human adipose-tissue-derived extracellular matrix sheets with other human cell types. Cell Tissue Res 2012;348: 559–567. 20. Ravichandran R, Venugopal JR, Sundarrajan S, Mukherjee S, Ramakrishna S. Precipitation of nanohydroxyapatite on PLLA/ PBLG/Collagen nanofibrous structures for the differentiation of adipose derived stem cells to osteogenic lineage. Biomaterials 2012;33:846–855. 21. Cao H, Xu SY. EDC/NHS-crosslinked type II collagen-chondroitin sulfate scaffold: Characterization and in vitro evaluation. J Mater Sci Mater Med 2008;19:567–575. 22. Sachlos E, Gotora D, Czernuszka JT. Collagen scaffolds reinforced with biomimetic composite nano-sized carbonate-substituted hydroxyapatite crystals and shaped by rapid prototyping to contain internal microchannels. Tissue Eng 2006;12:2479–2487. 23. Potter K, Kidder LH, Levin IW, Lewis EN, Spencer RG. Imaging of collagen and proteoglycan in cartilage sections using Fourier transform infrared spectral imaging. Arthritis Rheum 2001;44:846–855. 24. Cavalcanti OA, Da Silva CC, Pineda EAG, Hechenleitner AAW. Synthesis and characterization of phosphated crosslinked chondroitin sulfate: Potential ingredient for specific drug delivery. Acta utica Bonaerense 2005;24:234–238. Farmace 25. Servaty R, Schiller J, Binder H, Arnold K. Hydration of polymeric components of cartilage - An infrared spectroscopic study on hyaluronic acid and chondroitin sulfate. Int J Biol Macromol 2001; 28:121–127. 26. Marelli B, Ghezzi CE, Mohn D, Stark WJ, Barralet JE, Boccaccini AR, Nazhat SN. Accelerated mineralization of dense collagennano bioactive glass hybrid gels increases scaffold stiffness and regulates osteoblastic function. Biomaterials 2011;32:8915–8926. 27. Tampieri A, Celotti G, Landi E, Sandri M, Roveri N, Falini G. Biologically inspired synthesis of bone-like composite: Selfassembled collagen fibers/hydroxyapatite nanocrystals. J Biomed Mater Res 2003;67:618–625. 28. Kovacevic D, Rodeo SA. Biological augmentation of rotator cuff tendon repair. Clin Orthop Relat Res 2008;466:622–633. 29. van Tienen TG, Heijkants RG, Buma P, de Groot JH, Pennings AJ, Veth RP. Tissue ingrowth and degradation of two biodegradable porous polymers with different porosities and pore sizes. Biomaterials 2002;23:1731–1738. 30. Haugh MG, Murphy CM, McKiernan RC, Altenbuchner C, O’Brien FJ. Crosslinking and mechanical properties significantly influence cell attachment, proliferation, and migration within collagen glycosaminoglycan scaffolds. Tissue Eng Part A 2011;17:1201–1208. 31. Bonzani IC, George JH, Stevens MM. Novel materials for bone and cartilage regeneration. Curr Opin Chem Biol 2006;10:568–575. 32. Pattanayak DK, Divya P, Upadhyay S, Prasad RC, Rao BT, Rama Mohan TR. Synthesis and evaluation of hydroxyapatite ceramics. Trends Biomater Artif Organs 2005;18:87–92. 33. Bose S, Tarafder S. Calcium phosphate ceramic systems in growth factor and drug delivery for bone tissue engineering: A review. Acta Biomater 2012;8:1401–1421.
HUMAN COLLAGEN-BASED MULTI-LAYER SCAFFOLDS
ORIGINAL ARTICLE
34. O’Brien FJ, Harley BA, Yannas IV, Gibson LJ. The effect of pore size on cell adhesion in collagen-GAG scaffolds. Biomaterials 2005;26:433–441. 35. Bobyn JD, Wilson GJ, MacGregor DC, Pilliar RM, Weatherly GC. Effect of pore size on the peel strength of attachment of fibrous tissue to porous-surfaced implants. J Biomed Mater Res 1982;16: 571–584. 36. Freeman JW. Tissue engineering options for ligament healing. Bone Tissue Regen Insights 2009;2:13–23. 37. Klompmaker J, Jansen HW, Veth RP, Nielsen HK, de Groot JH, Pennings AJ. Porous implants for knee joint meniscus reconstruc-
tion: A preliminary study on the role of pore sizes in ingrowth and differentiation of fibrocartilage. Clin Mater 1993;14:1–11. 38. Spector M, Michno MJ, Smarook WH, Kwiatkowski GT. A highmodulus polymer for porous orthopedic implants: Biomechanical compatibility of porous implants. J Biomed Mater Res 1978;12: 665–677. 39. Bobyn JD, Pilliar RM, Cameron HU, Weatherly GC. The optimum pore size for the fixation of porous surfaced metal implants by the ingrowth of bone. Clin Orthop Relat Res 1980;150:263–270. 40. Clemow AJT, Weinstein AM, Klawitter JJ, Koeneman J, Anderson J. Interface mechanics of porous titanium implants. J Biomed Mater Res 1981;15:73–82.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | MONTH 2014 VOL 00A, ISSUE 00
11
