journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

Available online at www.sciencedirect.com

www.elsevier.com/locate/jmbbm

Research Paper

Designed composites for mimicking compressive mechanical properties of articular cartilage matrix Youjia Zhua,1, Hua Wub,1, Shaofa Sunc,n, Ting Zhoud, Jingjing Wud, Ying Wand,nn a

Department of Stomatology, Zhongnan Hospital, Wuhan University, 430071, PR China Department of Nuclear Medicine and Minnan PET Center, The First Affiliated Hospital of Xiamen University, Xiamen 316003, PR China c School of Nuclear Technology, Chemistry and Life Science, Hubei University of Science and Technology, Xianning 437100, PR China d College of Life Science and Technology, Huazhong University of Science and Technology, Wuhan 430074, PR China b

ar t ic l e in f o

abs tra ct

Article history:

Collagen, chitosan–polycaprolactone (CH–PCL) copolymer with PCL content of around 40 wt%

Received 1 February 2014

and chondroitin sulfate (CS) were mixed together at various ratios to prepare collagen/CH–PCL/

Received in revised form

CS composites and the resulting composites were used to build stratified porous scaffolds that

5 April 2014

are potentially applicable for articular cartilage repair. The ternary composites were designed in

Accepted 7 April 2014

such a way that collagen content in the scaffolds decreased from the top layer to the bottom

Available online 18 April 2014

layer while the content of CH–PCL and CS altered in a reversed trend in order to reach partial

Keywords:

similarity to cartilage matrix in the composition of main components. Porous structures inside

Composite

collagen/CH–PCL/CS scaffolds were constructed using a low-temperature deposition processing

Porous scaffold

technique and graded average pore-size and porosity for the scaffolds were established. Such

Stratiform structure

produced scaffolds were further crosslinked using 1-ethyl-3-(3-dimethyl aminopropyl) carbo-

Compressive mechanical property

diimide under optimized conditions, and the obtained scaffolds showed well-defined elastic

Articular cartilage matrix

compressive properties. Compressive modulus (E) and stress at 10% strain (s10) of full scaffolds in wet state reached about 2.8 MPa and 0.3 MPa, respectively, and meanwhile, E and s10 of layers inside hydrated scaffolds changed in a gradient-increased manner from the top layer to the bottom layer with significant differences between contiguous layers, which partially mimics compressive mechanical properties of cartilage matrix. In addition, in vitro culture of cell– scaffold constructs exhibited that scaffolds were able to well support the ingrowth and migration of seeded cells, and cells also showed relatively uniform distribution throughout the scaffolds. These results suggest that the presently developed collagen/CH–PCL/CS scaffolds have promising potential for applications in articular cartilage repair. & 2014 Elsevier Ltd. All rights reserved.

n

Corresponding author. Corresponding author. Tel.: þ86 27 87792147; fax: þ86 27 87792234. E-mail addresses: [email protected] (S. Sun), [email protected] (Y. Wan). 1 Authors contributed equally.

nn

http://dx.doi.org/10.1016/j.jmbbm.2014.04.003 1751-6161/& 2014 Elsevier Ltd. All rights reserved.

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

1.

Introduction

Articular cartilage is noted for a limited capacity for self-repair upon damage caused by traumatic injuries or degenerative diseases, especially for those defects correlated to relatively large areas of cartilage surface (Chiang and Jiang, 2009). The difficulty in repairing articular cartilage defects can be mainly ascribed to its avascular characteristic and relatively low cellular metabolic activity (Jackson et al., 2001). Clinical treatments for articular cartilage repair commonly involve microfractures, subchondral bone drilling, arthroscopic lavage with or without corticosteroids, abrasion arthroplasty, mosaicplasty and total joint replacement (Marsano et al., 2007; Thiede et al., 2012). Although improved short-term results from clinical therapies have been reported, these strategies have limited success in which they are deficient in long-term repair or lead to the formation of mechanically inferior fibro-cartilaginous tissue in the cartilage defect void (Hunziker, 2002; Shapiro et al., 1993). Many efforts have thus been made to search for alternative therapies that can yield correctly reparative tissue in cartilage defects, and on the other hand, achieve long-term recovered function of regenerated articular cartilage (Jackson et al., 2001; Chiang and Jiang, 2009). Nowadays, tissue engineering, which usually involves in combination of cells, porous scaffolds and bioactive agents, has emerged as a new method for generating functional new tissues to replace damaged articular cartilage (Oliveira et al., 2007; Tan et al., 2007). It is known that articular cartilage has a stratified structure and its composition and properties appear to change in an apparently anisotropic manner. Histological examinations point out that articular cartilage has four diacritical layers, generally being named as superficial layer, intermediate layer, deep layer and calcified layer (Castro et al., 2012; Leong et al., 2008). From the superficial zone to the calcified region, the content of water and type-II collagen progressively decreases whereas proteoglycans show an inverse trend, and meanwhile, collagen fibers and chondrocytes are organized in a certain oriented manner in different layers, which results in varied compressive properties between layers. In addition, the diameter of type-II collagen and the compressive mechanical modulus of articular cartilage matrix also change from the superficial layer to the calcified layer. Although tissue engineering strategies have potential for the repair of articular cartilage lesions, fabrication of satisfactory scaffolds faces various degrees of difficulties owing to hierarchical structures and anisotropic properties of cartilage matrix. Nevertheless, to date, many attempts have been made to build scaffolds with various compositions and structures that are approximately similar to that of cartilage matrix in order to achieve improved results for articular cartilage repair (Dormer et al., 2010; Harley et al., 2010a, 2010b; Kim et al., 2012; Leong et al., 2008; Levingstone et al., 2014). Many types of biodegradable materials, including naturally occurring and synthetic polymers or their combinations, have been investigated for fabrication of scaffolds that have designed compositions, structures and functions (Guo et al., 2012; Miao and Sun, 2010). In the case of articular cartilage tissue engineering, type-II collagen has received specific

33

attention because it is a key component in cartilage matrix which plays important roles in supporting chondrogenesis and maintaining chondrocytic phenotype (Chang et al., 2006; Levingstone et al., 2014; Moutos et al., 2007). Up to now, type-II collagen has indeed been extensively investigated for articular cartilage repair (Chiang and Jiang, 2009; Jackson et al., 2001). However, collagen alone seems not to be adequately competent for repairing articular cartilage lesions because of its high swelling, fast degradation and poor wet-state mechanical strength even though doable crosslinkers have been utilized (Berthiaume et al., 2011; Tierney et al., 2009). Chondroitin sulfate (CS) is a type of natural glycosaminoglycans (GAGs) mainly found in connective tissues such as bone, skin, and cartilage (Pieper et al., 1999). As a main GAG in cartilage matrix, CS is responsible for water retention due to its charged feature (Servaty et al., 2001), and it also plays a role in intracellular signaling, cell recognition and connecting different matrix components to the cell-surface glycoproteins and collagen (Vazquez et al., 2013). Taking into account the importance of CS, a logical strategy seems to build scaffolds for articular cartilage repair by directly using CS. But in fact, CS is usually employed as an accessory component in different types of scaffolds because it accounts for a relatively low percentage in native cartilage matrix as compared to type-II collagen (Chung and Burdick, 2008; Vazquez et al., 2013), and additionally, it generally shows inferior wet-state mechanical strength. Besides CS, another natural polysaccharide, chitosan, has also been widely studied for applications in articular cartilage repair because it has good ability to support chondrogenic activity and cartilage matrix expression by chondrocytes besides its structural similarity to GAGs (Suh and Matthew, 2000). Despite various advantages for biomedical applications, unmodified chitosan also shows poor wet-state mechanical properties (Wan et al., 2004a) and has fast in vivo degradation (Wan et al., 2010). One of effective approaches to regulating degradation and mechanical properties of natural polymers is to use them together with some other biodegradable polyesters in the form of blends or composites (LaPorta et al., 2012; Seidi et al., 2011; Smith et al., 2005; Wan et al., 2008). Of biodegradable polyesters, polycaprolactone (PCL) has been extensively investigated because it has adjustable mechanical strength and shows soft- and hard-tissue compatible properties (Wu et al., 2010). PCL is a semi-crystalline, linear and aliphatic polyester with a low melting point (ca. 60 1C), which allows easy processing. Apart from certain applications in sutures, wound dressings and drug delivery vehicles, PCL has also been commonly used for articular cartilage repair (MartinezDiaz et al., 2010; Matsiko et al., 2013). Grafting is also a versatile means to modify natural polymers in addition to blend or composite-based physical modification (Bhattacharya and Misra, 2004). Grafting polycaprolactone side chains onto chitosan backbone can generate some chitosan-polycaprolactone (CH–PCL) copolymers that would have tailorable mechanical and degradation properties as well as potential solubility in aqueous media, depending on its PCL content (Wan et al., 2010). Therefore, it would be feasible to blend collagen, CH–PCL and CS together to produce applicative composites and further to process

34

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

those into porous scaffolds with certain designed structures. By doing so, PCL in CH–PCL is able to potentially regulate mechanical and degradation characteristics of collagen/CH– PCL/CS scaffolds, and on the other hand, CS may also contribute to the mechanical enhancement in collagen/CH– PCL/CS scaffolds due to the interactions between sulfate anions in CS and amino groups in CH–PCL. In particular, it would be possible to designedly organize the compositions and structures of porous collagen/CH–PCL/CS scaffolds using low-temperature deposition processing techniques because of soluble properties of three components. In the present study, an attempt was made to build a type of stratified porous collagen/CH–PCL/CS composite scaffolds so that the scaffolds have gradient-changed compositions and graded pore-size and porosity partially similar to that of cartilage matrix while having mimetic compressive properties in comparison to cartilage matrix. Results for fabrication and characterization as well as compressive mechanical measurements of such designed scaffolds were reported.

2.

Experimental

2.1.

Materials

Chitosan powder (deacetylation degree: 91.4(72.3)%; viscosityaverage molecular weight: 3.1(70.18)  105, measured following previously reported methods (Wan et al., 2004a)) was supplied by Aladdin Inc. Type-II collagen from bovine nasal septum, chondroitin sulfate (CS) caprolactone, N-hydroxysuccinimide (NHS), 2-(N-morpholino) ethane sulfonic acid sodium salt (MES) and 1-ethyl-3-(3-dimethyl aminopropyl) carbodiimide (EDC) were purchased from Sigma-Aldrich. All other reagents were of analytical grade and purchased from different commercial sources in China. CH–PCL copolymers were synthesized using methods as previously described (Liu et al., 2006; Wan et al., 2010). In brief, chitosan powder was reacted with phthalic anhydride to produce phthaloylchitosan (PHCH) in which the amino groups at the C-2 sites of chitosan units were protected. PHCH was then grafted with caprolactone at the C-6 sites of chitosan backbone. Afterwards, the resulting PHCH-g-PCL copolymers were deprotected by eliminating phthaloyl groups using hydrazine monohydrate to obtain CH–PCLs. In considering required solubility of resulting CH–PCLs in aqueous media as well as the mechanical and degradation properties of designed scaffolds, CH–PCL containing polycaprolactone (PCL) percentage at around 40 wt% was used for the fabrication of collagen/CH–PCL/CS scaffolds.

2.2. Preparation of collagen/CH–PCL/CS composite scaffolds Collagen/CH–PCL/CS scaffolds were fabricated via a lowtemperature deposition technique using some custom-made circular Teflon molds that have movable stainless steel bottom. A sample holder was set in a cooling bath (CC520w, Huber) for sample preparation. 1 wt% CH–PCL solution in 1% acetic acid, 1 wt% collagen solution in 3% acetic acid and 1 wt% CS solution in distilled water were prepared, respectively. In a typical procedure, 10 ml of collagen/CH–PCL/CS blend was concentrated at 50 1C using a Tekmar Tissuemiser 410 (Tekmar Co.) to attain glutinous mixture. The mixture was cooled down to room temperature and spread onto the Teflon mold as a thin layer (ca. 1 mm in thickness). The mold was then placed onto the pre-cooled sample holder (  30 1C) to solidify the sample in a low temperature environment generated by stainless steel bottom of the mold. After that, the surface of the layer was slightly heated using a heat lamp that can generate various temperatures changing from around 30 to 40 1C on the surface of the layer, depending on the distance between the surface and the lamp. After integrally moving the layer down around 1 mm together with the movable bottom of the mold, a new layer was spread onto the top of the existing layer and solidified again. Assembly was repeated until the required scaffold was obtained. After being lyophilized at 75 1C, the resultant scaffolds were neutralized using 0.1 M Na2HPO4, washed to neutrality with deionized water and freeze-dried again. By prescribing the composition proportions for each layer, pro-cooling the sample holder between  55 1C and 30 1C, changing the concentrations of glutinous collagen/ CH–PCL/CS mixtures, and maintaining other processing parameters for each layer constant, the average pore-size and porosity of each layer were regulated. The resulting four layers from the top to the bottom of the scaffold were named as L1, L2, L3 and L4, respectively. The composition of each layer was designed in such a way that components in scaffolds changed in a gradient manner and had different variation tendencies as the layers changed. Relevant data for the composition of each layer are summarized in Table 1. Some scaffolds were further crosslinked using EDC/NHS. In brief, 100 mg dry scaffold was pretreated in 40 ml of 40% ethanol solution containing 50 mM MES at room temperature for 30 min. Subsequently, the scaffold was immersed in 50 ml MES (50 mM) containing 14 mM EDC and 5.8 mM NHS at room temperature for 24 h. After that, scaffolds were repeatedly washed with distilled water and lyophilized at  75 1C again. By mainly changing the amount of EDC in a region between 8

Table 1 – Composition of stratified collagen/CH–PCL/CS composite scaffolds. Name of layers

Type-II collagen (wt%)

CH–PCLa (wt%)

Chondroitin sulfate (wt%)

L1 L2 L3 L4

90 60 30 0

8 32 56 80

2 8 14 20

a

PCL content in CH–PCL is 40.270.91 wt%.

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

and 16 mM and keeping the ratio of EDC to NHS constant, variously crosslinked scaffolds were obtained.

2.3.

Characterization

PCL weight percent in CH–PCLs was measured using an elemental analyzer (Perkin-Elmer 2400). Scaffolds were cut into almost isopachous sheets (about 1 mm in thickness) with the help of Teflon molds mentioned earlier, and porous structures in vertical and cross-sections of sheets were viewed using scanning electron microscopy (SEM) after gold–palladium coating. 100 pores at 100 lattices in an SEM image for each specimen were measured, and pore-distribution and average pore-sizes were calculated using image analysis software (AxioVision LE). Porosity of each sheet was determined using a reported method (Wan et al., 2004b). Swelling index (SI) of sheets or full scaffolds was measured with a gravimetric method. Sheet samples were immersed in PBS at 37 1C for 4 h, and they were then transferred into some tubes with sintered glass filter bottom for removal of excess water via centrifugation at 2000 rpm for 2 min. SI of scaffolds was calculated using the following equation: SI ¼ ½ðWs Wd Þ=Wd   100%

ð1Þ

where Ws and Wd are the weight of swollen and dried scaffold, respectively.

2.4.

Compressive mechanical testing

Dry scaffolds were punched into cylinders (5 mm in diameter and ca. 4.6 mm in average thickness) using a biopsy punch and the samples were measured in unconfined mode using a testing machine (Instron ElectroPuls) equipped with an auxiliary sensing unit. Measurements were performed with accuracy for load down to 0.5 g and for deformation less than 0.05 mm. To measure compressive parameters of different layers inside scaffolds, cylindrical scaffolds were cut into approximately isopachous sheets (ca. 1.1 mm in thickness) with the aid of a mold having a movable bottom, and each sheet was measured using the same machine. To measure compressive parameters of hydrated samples, dry samples were immersed in PBS solution (pH 7.4) for 4 h, and excess water inside samples was removed by centrifugation using the same method described in SI measurements. Compressive modulus (E) was determined using the slope of the initial linear elastic region of stress–strain curves, and the stress at 10% strain (s10), which is another common indicator for the compressive mechanical properties of porous materials (Wu and Ding, 2004; Wu et al., 2006), was recorded for estimating compressive strength.

2.5.

Culture of cell–scaffold constructs

The animal experiments were conducted according to National Institutes of Health standards as set forth in the Guide for the Care and Use of Laboratory Animals. Articular cartilage taken from rabbit knee joints (New Zealand white rabbits, 2-week old) was diced into pieces and the detrital sample was digested with 0.2% collagenase in DMEM at 37 1C for 6 h or longer. The isolated chondrocytes

35

were seeded to ventilated 75 cm2 monolayer cell culture flasks at an initial density of 105 cells/cm2 in a humidified atmosphere of 5% CO2 at 37 1C, supplementing with 10% fetal bovine serum, 50 μg/ml L-ascorbic acid, 100 U/ml penicillin and 100 μg/ml streptomycin. The culture medium was replaced every other day. Cells were detached using trypsin(0.25%)–EDTA(0.02%) in PBS when cell confluence reached around 85%, and they were resuspended in DMEM for further experiments. Cylindrical dry scaffolds with a dimension of 5 mm in diameter and ca. 4.5 mm in average thickness were sterilized in 75% ethanol over night. Prior to cell seeding, scaffolds were well rinsed in PBS, blotted dry and placed in wells of 24-well culture plates for subsequent cell seeding. Chondrocytes were seeded onto scaffolds at a density of 8  106 cells/scaffold. To help cells enter the interior of scaffolds, the full amount of cells was divided into two portions and seeded onto scaffolds twice. One half of cells were first seeded on the top of the scaffold, and after attachment for 30 min, the scaffold was turned over for seeding another half of cells. After attachment for 2 h at 37 1C, cell–scaffold constructs were added with complete medium and cultured in a humidified incubator at 37 1C with 5% CO2. Culture medium was changed every 2 days until harvest.

2.6.

Cell viability and distribution

Viability of chondrocytes cultured on scaffolds was evaluated using a calcein-AM/ethidium homodimer-1(EthD-1) Live/ Dead Kit (Molecular Probes) following the manufacturer's instructions. This technique stains viable cells green and dead cells red. At the end of preset culture intervals, cell– scaffold constructs were withdrawn, and each was horizontally sectioned into four approximately isopachous sheets (ca. 1.2 mm in thickness)corresponding to the matched layer using Teflon molds mentioned earlier. After staining, the cells were visualized from top surface of each layer in a downward manner at an image plane depth of about 100 μm using confocal scanning microscopy. The DNA content in scaffolds was determined using a Quant-IT PicoGreen dsDNA Kit (Invitrogen) with a SpectraMax M5 Plate Reader. In brief, scaffolds cultured for various days up to 14 days were harvested, washed in PBS and digested overnight at 45 1C, according to the manufacturer's instructions. After centrifugation, the supernatants were removed and diluted for quantifying DNA content in scaffolds. Sample fluorescence was measured at an excitation of ca. 480 nm and an emission of ca. 520 nm. Two-dimensional culture was used as control. Some cell–scaffold constructs cultured for a period up to 3 weeks were fixed in 4% phosphate-buffered paraformaldehyde at 4 1C for 1 h, dehydrated in a graded ethanol series to 100% ethanol, embedded in paraffin and vertically sectioned into slices with a thickness of 10 μm. The resulting slices were mounted onto glass slides, dried at 50 1C for 1 h and stained with hematoxylin and eosin (H/E).

2.7.

Statistical analysis

Data were presented as mean7standard deviation. In the case of normal population, means were compared by Student's

36

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

t-tests, and otherwise, analysis of variance was performed with po0.05 as the criteria for statistical significance.

3.

Results

3.1.

Fabrication of collagen/CH–PCL/CS scaffolds

In view of structure and property of articular cartilage (Castro et al., 2012), a type of four-layer porous collagen/CH–PCL/CS scaffolds was fabricated in order to attain a similar porous structure like cartilage matrix. A schematic representation of the layers inside the scaffolds is illustrated in Fig. 1(A). Three components in the scaffolds were arranged in such a way that collagen content changed from the top layer to the bottom layer in a gradient descent trend contrary to that of CH–PCL and CS. Considering processing feasibility together with properties of each component, collagen/CH–PCL/CS composites for different layers in scaffolds were optimized to achieve certain designed properties. Formulation for the composition in each layer is summarized in Table 1. All scaffolds in this study were built using this formulation unless otherwise stated. Fig. 1(A) shows a typical photo for a hierarchical-organized collagen/CH–PCL/CS scaffold crosslinked by EDC. To see internal pores and interface of scaffolds, some full scaffolds were vertically sectioned into thick slices, and the resulting slices were viewed using SEM. Several representative SEM images for interfaces between different layers in an EDC-crosslinked collagen/CH–PCL/CS scaffold are represented in Fig. 1(B) (a, b and c). These images exhibit that interface zones between adjacent layers were porous and approvingly interconnected, and there were no discernible crannies observed in the interface zones. These results confirm that the pores of different layers in the scaffold were well interconnected.

3.2.

Characterization of the construct

It is known that articular cartilage matrix has a graded porous structure with various porosities changing from the superficial layer to the calcified layer (Castro et al., 2012; Dormer et al., 2010). Therefore, besides formulated multi-level compositions in collagen/CH–PCL/CS scaffolds, preferred scaffolds should also be constructed into gradient-organized porous architecture while having suitable compressive mechanical properties comparative to that of cartilage matrix. By mainly controlling the concentration of collagen/CH–PCL/CS mixtures and optimizing a few key factors, such as pro-cooled temperatures of cooling bath, surface-heating time and the amount of crosslinker used, scaffolds with graded porous structures were successfully built. Fig. 2 shows a few representative SEM images for four layers inside a layered collagen/CH–PCL/CS scaffold. These images show that (1) pores in each layer were well interconnected and they had varied sizes changing from several tens of microns to hundreds of microns; and (2) pore-size and porosity for different layers identifiably increased from L1 to L4, revealing gradient-

Fig. 1 – A photo of a EDC-crosslinked collagen/CH–PCL/CS scaffold and schematic representation of four layers (A, blue dash lines indicate potential interface zones of layers); and several representative SEM images for porous interfaces between contiguous layers (B, dotted white lines in these images (a–c) indicate interface zones)(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

37

Fig. 2 – Some representative SEM images for four layers inside a EDC-crosslinked collagen/CH–PCL/CS scaffold (a, b, c and d respectively correspond to L1, L2, L3 and L4 layers).

changed trends for pore-sizes and porosities. Two sets of EDC-crosslinked and uncrosslinked collagen/ CH–PCL/CS scaffolds were measured for their pore-size, porosity and SI, and relevant data are shown in Fig. 3(A), (B) and (C), respectively, to examine the effect of crosslinker on these parameters while keeping other processing conditions constant. It can be seen from Fig. 3(A) and (B) that average pore-size and porosity for EDC-crosslinked and uncrosslinked scaffolds gradually increased from L1 to L4 with significant differences (po0.05), and crosslinking had significant impacts on the average pore-size but not on the porosity of scaffolds. Fig. 3(C) indicates that SI for both EDC-crosslinked and uncrosslinked scaffolds exhibited significantly decreased trends (po0.05) from L1 to L4, and SI of each layer in EDCcrosslinked scaffolds was greatly reduced when compared to the matched one in uncrosslinked scaffolds.

3.3. Compressive properties of collagen/CH–PCL/CS scaffolds Fig. 4(A) and (B) illustrates typical compression stress–strain curves for uncrosslinked collagen/CH–PCL/CS scaffolds in both dry and wet states. All curves for dry layers and the scaffold showed a few distinct characteristics: linear deformation at very small strain, flexure deformation at higher strain, followed by a plateau region at apparently larger strain and a solidifying region in which the stress sharply increases at very large strain. In addition, the curves for dry samples displayed that as layers changed from L1 to L4 the plateaus

zone in the matched curve gradually shortened and the corresponding compressive stress increased. With respect to hydrated samples, it can be observed that (1) large decrease in stress and more extended plateaus zones in stress–strain curves were registered for the wet layers and the wet scaffold in comparison to matched dry samples, and (2) the stress of hydrated layers also showed an increasing trend from L1 to L4 while plateaus zone in their stress–strain curves gradually became shorter from L1 to L4. Two series of dry or wet uncrosslinked scaffolds are measured and their average E and s10 are shown in Fig. 5(A) and (B). It can be seen that E and s10 of layers inside dry scaffolds varied in a significant uptrend (po0.01) as the layers changed from L1 to L4; and with regard to hydrated layers, their E and s10 considerably reduced in comparison to corresponding dry layers but their E and s10 still showed an increasing trend from L1 to L4 with significant differences (po0.01) between contiguous layers. Representative compression curves for dry and wet EDCcrosslinked collagen/CH–PCL/CS scaffolds are shown in Fig. 6(A) and (B), respectively. By comparing each curve in Fig. 6(A) with the matched one in Fig. 4(A), it can be concluded that dry EDCcrosslinked samples showed greatly strengthened compressive stress while maintaining similar stress–strain characteristics as compared to the dry uncrosslinked samples. In the case of hydrated EDC-crosslinked samples, great enhancement in compressive stress was also achieved when comparing to uncrosslinked samples in wet state (see Figs. 6(B) and 4(B)).

38

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

600

uncrosslinked scaffold EDC-crosslinked scaffold *

300 * *

250

* *

200

L4 (uncrosslinked, dry) L3 (uncrosslinked, dry) full scaffold (uncrosslinked, dry) L2 (uncrosslinked, dry) L1 (uncrosslinked, dry)

500

*

Compressive stress (kPa)

Average pore size (μm)

350

*

150 100 50

400

300

200

100

0 L1

L2

L4

L3

0 0

Layer

10

20

30

40

50

Strain (%) uncrosslinked scaffold EDC-crosslinked scaffold

160 *

100

*

*

Porosity (%)

*

*

*

80

L4 (uncrosslinked, wet) L3 (uncrosslinked, wet) full scaffold (uncrosslinked, wet) L2 (uncrosslinked, wet) L1 (uncrosslinked, wet)

140

Compressive stress (kPa)

120

60

40

20

120 100 80 60 40 20

0 L2

L1

L3

L4 0

Layer

0

10

20

30

40

50

60

70

Strain (%) 180

150

Swelling index (%)

Fig. 4 – Representative compression stress–strain curves for uncrosslinked collagen/CH–PCL/CS scaffolds (A, dry sample; and B, hydrated sample).

uncrosslinked scaffold EDC-crosslinked scaffold ** **

120

* **

90

* * 60

30

0 L1

L2

L3

L4

Full scaffold

Name of sample

Fig. 3 – Average pore-size (A), porosity (B) and swelling index (C) of uncrosslinked and EDC-crosslinked collagen/CH–-PCL/ CS scaffolds (*po0.05; **po0.01; n¼ 6).

Data for E and s10 of dry and wet EDC-crosslinked collagen/CH–PCL/CS scaffolds are shown in Fig. 7(A) and (B), respectively. It can be observed that gradient-increasing E and s10 from L1 to L4 with significant differences (po0.01) for adjacent layers in dry or wet EDC-crosslinked scaffolds were

still clearly remained. It is specially worth noticing that an around 8-fold increase in both E and s10 was reached for wet layers in EDC-crosslinked scaffolds when compared to matched ones in wet uncrosslinked scaffolds (see Figs. 7(B) and 5(B)). These results suggest that hydrated collagen/CH– PCL/CS scaffolds with appropriate EDC-crosslinking can reach high enough compressive strength that is comparatively equal to that of cartilage matrix (Gannon et al., 2012; Treppo et al., 2000), and meanwhile, they can well mimic the gradient-changed compressive modulus of cartilage matrix (Castro et al., 2012).

3.4. Viability and distribution of chondrocytes inside scaffolds Collagen/CH–PCL/CS scaffolds were seeded with chondrocytes to assess the effect of scaffolds on viability of chondrocytes, and some representative images of stained chondrocytes in different layers after in vitro incubation for 7 days are shown in Fig. 8(A). It is observed that very few dead cells were imaged in each layer, meaning that seeded chondrocytes had high

39

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

6

0.35

L4 (EDC-crosslinked, dry) L3 (EDC-crosslinked, dry) full scaffold (EDC-crosslinked, dry) L2 (EDC-crosslinked, dry) L1 (EDC-crosslinked, dry)

2.5

** ** ** 3

0.20

* 0.15

σ10 (MPa)

E (MPa)

0.25

**

4

2 0.10 1

0.05

0

Compressive stress (MPa)

0.30

**

5

2.0

1.5

1.0

0.5

0.00 L1

L2

L3

L4

Full scaffold

0.0

Name of sample

0

10

20

30

40

Strain (%) 600

**

1.2

100 500

60

**

300 **

40

**

200

σ10 (kPa)

**

20

100

0

0 L1

L2

L3

L4

Compressive stress (MPa)

1.0

80 400

E (kPa)

L4 (EDC-crosslinked, wet) L3 (EDC-crosslinked, wet) full scaffold (EDC-crosslinked, wet) L2 (EDC-crosslinked, wet) L1 (EDC-crosslinked, wet)

**

0.8

0.6

0.4

0.2

Full scaffold

Name of sample 0.0

Fig. 5 – Variations of compressive modulus and the stress at 10% strain of uncrosslinked collagen/CH–PCL/CS scaffolds (A, dry sample; and B, hydrated sample; *po0.05; **po0.01; n¼ 6).

0

10

20

30

40

50

60

Strain (%)

Fig. 6 – Representative compression stress–strain curves for EDC-crosslinked collagen/CH–-PCL/CS scaffolds (A, dry sample; and B, hydrated sample). viability during 7-day culture. In addition, the appropriate appearance of stained viable cells in different layers showed similarity in their size and number, indicating that each layer has a similar ability to maintain the viability of chondrocytes. Fig. 8(B) shows measured DNA content in scaffolds during various culture periods up to 14 days, which reveals the growth trend of chondrocytes in scaffolds. The bar-graphs explicate that the growth of chondrocytes inside scaffolds roughly experienced two phases: fewer cells grown from day 1 to day 5; and cells grown relatively fast after day 5. In addition, it can also be observed that fewer cells grown in scaffolds in the first a few days with significant differences measured at day 3 and day 5 as compared to control but there were no significant differences in the proliferated cells between the scaffolds and control since day 7. Considering the initial number of seeded cells and proliferated cells after culture, results in Fig. 8(B) suggest that these scaffolds are able to well support the growth of chondrocytes. To see the distribution of chondrocytes inside scaffolds, some slices cut from cultured chondrocyte–scaffold constructs were stained using H/E. To view the vertical sections of full scaffolds with higher magnification and clearly show the cells inside the scaffold, photos were taken from top half and bottom

half of the full sections and they were then arranged into full images for the sections. Several representative images are shown in Fig. 9. After 2-week culture, seeded cells were shown to be well distributed throughout all layers inside the scaffold without typical blank areas or discontinuous zones, and several points can be drawn from these images: (1) EDC-crosslinked collagen/CH–PCL/CS scaffolds were able to support ingrowth of chondrocytes; (2) pore-size and porosity of each layer inside the scaffolds were large enough for the migration of chondrocytes; and (3) the interfaces between different layers should be porous so that chondrocytes were able to migrate from one layer to another layer. After extended culture up to 3 weeks, more cells with significant differences in cell counts (po0.01) were observed from the image when comparing to matched one cultured for 2 weeks, indicating proliferation of chondrocytes.

4.

Discussion

Articular cartilage is a type of soft tissue that overlies the articulating bony ends in diarthrodial joints, and it endues

40

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

2.5

**

15

**

2.0

*

10

1.5

* *

1.0

σ10 (MPa)

E (MPa)

*

5 0.5

0.0

0 L1

L2

L3

L4

Full scaffold

Name of sample

0.8 4

*

0.7

**

0.6 3

0.4

*

2

**

0.3 0.2

1

0.1 0

DNA (μg/sample)

**

σ10 (MPa)

*

0.5

E (MPa)

control collagen/CH-PCL/CS scaffold

2.5

2.0

* 1.5

*

1.0

0.0 L1

L2

L3

L4

Full scaffold

0.5

Name of sample

Fig. 7 – Variations of compressive modulus and the stress at 10% strain of EDC-crosslinked collagen/CH–PCL/CS scaffolds (A, dry sample; and B, hydrated sample; *po0.05; **po0.01; n¼ 6).

joints with certain crucial mechanical functions such as very low wear resistance, load bearing and shock absorption. Since articular cartilage in load-bearing joints needs to withstand and transfer high loads that are much higher than body weight in many cases (Mow et al., 1992), articular cartilage has specific mechanical properties apart from its anisotropic compositions and structures (Jurvelin et al., 2003). Compressive mechanical properties of articular cartilage are usually related to instantaneous (and/or transient) and equilibrium responses of the articular cartilage to the compressive loading, and mechanical parameters of articular cartilage are primarily measured by unconfined compression, confined compression and indentation (Armstrong and Mow , 1982; Kerin et al., 1998). In the present study, some efforts were made to fabricate composite scaffolds with certain compressive properties that can partially mimic the uniaxial compressive-deformation characteristics of the articular cartilage ECM. It is known that type-II collagen is naturally abundant in articular cartilage and it progressively alters in a trend contrary to that of GAG and proteoglycans (Castro et al., 2012; Jackson et al., 2001; Leong et al., 2008). In the case of articular cartilage tissue engineering, a logical strategy for building suitable scaffolds is to select correct materials and to organize the

0.0 0

1

3

5

7

10

14

Culture time (day)

Fig. 8 – Representative confocal microscope z-stack images of chondrocytes in different layers inside a EDC-crosslinked collagen/CH–PCL/CS scaffold (A, culture time: 7 days; green: live cells; red: dead cells; scale bar: 80 μm; a, b, c and d respectively correspond to L1, L2, L3 and L4 layers); DNA content in full scaffolds during the 14-day culture period (B, control: 2D culture with 24-well plate; *po0.05; n ¼4).

employed materials into proper 3-dimensional architecture with necessitated properties (Chen et al., 2006). In general, scaffolds for articular cartilage repair have to endure in vivo degradation for a few weeks with required mechanical strength. In the present study, besides collagen component, CH–PCL was selected as another main component in order to endue collagen/CH–PCL/CS scaffolds with enhanced wet-state mechanical strength and slowdown degradation rate because PCL has mechanically strong characteristics and very slow in vivo degradation rate (Panadero et al., 2013; Wan et al., 2010). Although different CH–PCLs with various PCL percentages up to 70 wt% can be synthesized under present synthesis conditions (Liu et al., 2006; Wan et al., 2010), PCL content in CH–PCLs needs to be well balanced because CH–PCLs with higher PCL content, for example, around 47 wt% or higher, will be insoluble in aqueous media, and as a result, presented processing techniques would be inapplicable for the manufacture of designed composite scaffolds. In view of the merits related to

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

2-week culture

3-week culture

Top of scaffold

Middle of scaffold

Bottom of scaffold

Fig. 9 – Chondrocyte distribution inside EDC-crosslinked collagen/CH–PCL/CS scaffolds (H/E staining; scale bar:100 μm; see Fig. 1(A) for “top” and “bottom” of scaffold).

chitosan and PCL, CH–PCL containing ca. 40 wt% PCL was selected for the fabrication of scaffolds. CS is known to be an anionic polymer and it was previously used as a counterpart for building up multilayer CS/chitosan polyelectrolytes via a layer-by-layer processing technique based on interactions between negative charges in CS and positive charges in chitosan (Liu et al., 2005). Recently, an effort has been made to build hierarchical CS/chitosan scaffolds with a prescribed ratio of CS to chitosan at 1:1 using PEI-modified paraffin spheres as porogen (Silva et al., 2013). In our cases, PCL was designedly grafted onto the C-6 sites of chitosan units rather than C-2 sites using a group-protection method to leave amino groups at the C-2 sites of chitosan backbone free (Liu et al., 2006; Wan et al., 2010). Consequently, free amino groups in CH–PCLs were able to interact with sulfate anions and carboxyl groups in CS to form into CH–PCL/CS polyelectrolytes in which CS actually acted as a

41

type of physical crosslinker, like the cases mentioned in CS/chitosan complexes (Vazquez et al., 2013). In fact, in this study, CS was used not only as a component for the resulting collagen/CH–PCL/CS scaffolds but also as a counterpart to CH–PCL to regulate the mechanical property of the scaffolds. The total content of CS inside collagen/CH–PCL/CS scaffolds was prescribed as a relatively low amount as compared to two other components mainly considering the native ratio of collagen to CS inside articular cartilage matrix (Chang et al., 2006). In addition, it was found that higher CS content in collagen/CH–PCL/CS composites would result in poor scaffold fabrication since some visual crannies in interface zones could occur due to the enhanced interactions between CH– PCL and CS in this ternary system if CS content was beyond a threshold. Based on many trials, the ratio of CH–PCL to CS was optimized as 4:1. The compositional proportions of components for each layer in collagen/CH–PCL/CS scaffolds were formulated using orthogonal experimental design while taking account of several key factors such as (1) graded compositions of components; (2) processing feasibility; and (3) gradient compressive modulus. To take full advantage of three components while effectively controlling the structures and properties of the resulting scaffolds, the compositions for different layers were optimized and data are shown in Table 1. On the basis of formulation listed in Table 1, collagen content inside scaffolds decreased from L1 to L4, and on the contrary, CS content and the chitosan percentage in CH–PCL followed an increasing trend, revealing that the composition of scaffolds became partially similar to that of cartilage matrix. To enhance the compressive strength of the scaffolds, collagen/CH–PCL/CS scaffolds were further crosslinked using EDC. To date, various types of covalent crosslinkers have been used for crosslinking collagen, chitosan or CS (Hennink and Nostrum, 2012). In the present instance, EDC was selected as a crosslinker because it has good biocompatibility (Cao and Xu, 2008; Pieper et al., 1999) and is also widely used for crosslinking amino groups and carboxyl groups (Berger et al., 2004; Hennink and Nostrum, 2012). In principle, it seems to be possible to fabricate presently designed composite scaffolds by directly using EDC as a crosslinker. Despite our trials, direct use of covalent crosslinkers for the fabrication of desired collagen/CH–PCL/CS scaffolds was unsuccessful. It was found that collagen/CH–PCL/CS gels would become a sort of crosslinked elastomer in a short period of time during the gel condensation if a prescribed amount of EDC was directly added into this ternary system, and as a result, the layers inside the scaffolds were commonly separated by visualized cracks, or otherwise, nonporous or oligoporous interfaces between adjacent layers very frequently occurred inside the scaffolds. Soluble properties of collagen, CH–PCL and CS in aqueous medium make it possible to fabricate porous collagen/CH–PCL/ CS scaffolds with certain graded structures using presented processing technique. Up to now, freeze-drying method has been attracting much attention in fabrication of porous polymer scaffolds due to its biofriendly feature and the ease of processing (O'Brien et al., 2004). In the freeze-drying process of scaffolds, pores inside scaffolds are generated as negative

42

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

replicate of ice crystals that are formed during freezing. In principle, concentration of polymer solution can significantly affect growth of ice crystals and thus regulate the pore size and porosity due to various thermal conduction rates (Yang et al., 2008). In addition, freezing rate and annealing during lyophilisation can also modulate the size of ice crystals, and in turn, regulate pore sizes and porosity (Yang et al., 2008). By mainly controlling the concentration of collagen/CH–PCL/CS composite gels, regulating the temperature of cooling bath and optimizing the amount of crosslinker used, stratified porous collagen/CH–PCL/CS scaffolds were built. The photo in Fig. 1(A) confirms that the obtained scaffold had a regular cylindrical shape without visual defects on the surface. SEM images in Fig. 1(B) show that the combinatorial processing technique developed in the present study is proved to be successful to promote establishment of porous linkages between adjacent layers, and on the other hand, to avoid formation of cracks or clefts in the interface zones. It is known that articular cartilage matrix has graded pores and its porosity changes from the superficial layer to the calcified layer in a region between 70% and 90% (Castro et al., 2012; Jackson et al., 2001; Leong et al., 2008). Therefore, layered porous architecture with gradient-organized features for collagen/CH– PCL/CS scaffolds also needs to be constructed besides considering their multi-level compositions. SEM images in Fig. 2 show that the presently developed collagen/CH–PCL/CS scaffold can be endued with gradient-changed pore-sizes and porosity. Results shown in Figs. 1 and 2 confirm that designed collagen/CH–PCL/ CS composites can be practically processed into hierarchical scaffolds with graded compositions and necessitated porous structures using presented processing techniques and prescribed formulation. In general, composition, processing conditions and amount of crosslinker can significantly affect the structure or property of porous polymer scaffolds (Chen et al., 2007). In the present study, the main attempt is to build a certain type of collagen/ CH–PCL/CS scaffolds so that they have appropriate compositions and structures partially similar to that of cartilage matrix while possibly mimicking the compressive modulus and strength of cartilage matrix. By keeping processing conditions constant and following the formulation shown in Table 1, the crosslinker amount and crosslinking conditions for collagen/ CH–PCL/CS scaffolds were optimized. To achieve the required wet-state compressive strength for full scaffolds, and at the same time, gradient-changed trends for E and s10 from L1 to L4 with significant differences between adjacent layers, the used amount of EDC and NHS was selected as 14 mM and 5.8 mM, respectively, as mentioned in the experimental section. As mentioned earlier, gradient-changed average pore-size and porosity for the uncrosslinked scaffolds were first constructed by mainly controlling the concentration of collagen/ CH–PCL/CS composite gels for each layer and regulating the pre-cooled temperature of cooling bath. After EDC-crosslinking, collagen/CH–PCL/CS scaffolds might be integrally contracted to some extent because EDC can covalently crosslink carboxyl groups in collagen or CS and amino groups in CH– PCL. As a result, the average pore-size of EDC-crosslinked scaffolds could be significantly reduced while their porosity remained nearly unchanged in comparison to that of uncrosslinked scaffolds, as indicated in Fig. 3(A) and (B).

In the case of uncrosslinked collagen/CH–PCL/CS scaffolds, some electrostatic interactions would occur among collagen, CH–PCL and CS since there are various amounts of amino groups, carboxyl groups and sulfate anions in the respective component. In terms of the formulation listed in Table 1, it is rational to believe that PCL content in CH–PCL is able to regulate SI of each layer to a certain extent because PCL is nearly nonswelling in aqueous media (Wan et al., 2010), and thus, rising PCL content in CH–PCL from L1 to L4 would result in descending SI for the scaffolds; in addition, physical linkages between amino groups in CH–PCL and sulfate anions in CS would progressively enhance from L1 to L4 due to gradually increasing amount of CH–PCL and CS, which also contributes to decreasing SI of layers from L1 to L4. Accordingly, the significant differences in SI between adjacent layers should be ascribed to the synergetic effect of mentioned two factors. With regard to EDC-crosslinked collagen/CH–PCL/CS scaffolds, many amino groups in CH–PCL and carboxyl groups in collagen or CS were covalently crosslinked, and thus, chemical linkages formed between amino groups and carboxyl groups would certainly limit the swelling of full EDCcrosslinked scaffolds and notably reduce the differences in SI between adjacent layers as well. In principle, greater SI for a hydrophilic scaffold should denote higher water content inside the scaffold. It is worth noticing that SI of EDC-crosslinked collagen/CH–PCL/CS scaffolds followed a clear downtrend with significant differences from L1 to L4, meaning that water content of layers inside the scaffolds changed in a gradient-decreased manner, which is closely similar to the variation tendency of water content in articular cartilage matrix (Jackson et al., 2001; Leong et al., 2008). Compressive mechanical properties of scaffolds used for repairing cartilage defects located in load-bearing joints such as hip, knee and ankle are of particular importance since they are strongly linked to the shape-persistency and durability of scaffolds in practical operations (Buckwalter, 1998; Waldman et al., 2004), and at the same time, to the modulation of seeded cells (Dormer et al., 2010; Raghunath et al., 2007). In the light of a fact that axial compressive modulus of cartilage matrix gradually increases from the superficial layer to the calcified layer (Castro et al., 2012; Leong et al., 2008), an intended effort of present study is to fabricate certain types of scaffolds with required compressive modulus similar to that of cartilage matrix. Typical compression stress–strain curves in Fig. 4 signify that the uncrosslinked collagen/CH– PCL/CS scaffolds in both dry and wet states had elastic features which resemble the compression curves of elastomers (Harley et al., 2010a), and the slope of initial linear region as well as the height of plateau region in curves for matched layers significantly increased from L1 to L4. Fig. 5 further verifies that there were significant differences in E and s10 of dry and wet layers. Gradient-varied E and s10 shown in Fig. 5 can be attributed to the regulation of PCL content in CH–PCL as well as interactions between amino groups in CH–PCL and sulfate anions in CS. As listed in Table 1, PCL content in CH–PLC increased from around 3 wt % in L1 to about 30 wt% in L4, which potentially enables the resulting scaffolds to have gradient-changed mechanical strength because of mechanically strong characteristics of

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

PCL (Cruz et al., 2008; Panadero et al., 2013). On the other hand, CS was arranged to increase from 2 wt% in L1 to 20 wt% in L4, implying that interaction intensity between CH–PCL and CS would also be enhanced from L1 to L4. As a synergetic result, dry or wet uncrosslinked collagen/CH–PCL/CS scaffolds would be endued with gradient-varied E and s10. Considering that the implanted scaffolds will be in wet state under physiological conditions, the E and s10 for hydrated scaffolds are of more significance as compared to that of dry scaffolds. Fig. 5(B) indicates that E of wet uncrosslinked scaffolds was around 0.4 MPa, which is lower than that of cartilage in hip, knee or ankle (Shepherd and Seedhom, 1999; Jurvelin et al., 2003). To enhance compressive strength of scaffolds in wet state, collagen/CH–PCL/CS scaffolds were further crosslinked using EDC. In comparison to wet uncrosslinked scaffolds illustrated in Fig. 5(B), Fig. 7(B) exhibits that E and s10 of wet full EDC-crosslinked scaffolds considerably increased and their E appeared to be close to 3 MPa, reaching a comparative value in comparison to that of native cartilage in knee (Jurvelin et al., 2003; Treppo et al., 2000). In addition, it can also be seen that E and s10 of layers inside wet EDC-crosslinked scaffolds increased with significant differences (po0.05) between adjacent layers as the layers changed from L1 to L4 (see Fig. 7(B)), suggesting that the presently designed EDC-crosslinked collagen/CH–PCL/CS scaffolds can partially mimic the compressive mechanical properties of cartilage matrix while having required compressive strength in wet state. In general, higher porosity and larger average pore-size in a porous polymer scaffold with homogeneous composition may endow the scaffold with lower compressive modulus and s10 as compared to that having exactly the same composition but lower porosity and smaller average pore-size (Altala et al., 1997; Reis and Cohn, 2002). In our cases, collagen/CH–PCL/CS scaffolds had nonhomogeneous compositions from L1 to L4, and in particular, PCL content in CH–PCL increased from L1 to L4. With respect to the single effect of porosity and average pore-size of layers in wet EDC-crosslinked scaffolds, E and s10 of layers might be down-regulated from L1 to L4; on the other hand, with regard to the effect of PCL content in each layer, E and s10 of layers should have an increasing trend from L1 to L4 in considering mechanically strong characteristics of PCL. Results shown in Fig. 7(B) confirm that these two factors synergistically resulted in gradient-increasing E and s10 from L1 to L4, suggesting that gradient-changed mechanical properties as well as graded average pore-size and porosity for presently designed scaffolds can be concomitantly achieved. Although compressive strength of wet EDC-crosslinked scaffolds can be further enhanced by using an increasing amount of EDC, it was found that clearly gradient-increasing trends in E and s10 with significant differences between adjacent layers could be lost if the applied EDC amounts was higher than a threshold. The used EDC amount was therefore optimized to achieve required porous structures and necessitated wet state compressive properties while maintaining significant differences between adjacent layers for compressive modulus and swelling index of scaffolds, as mentioned earlier. Calcein-AM/EthD-1 double-staining was used to identify live or dead cells (Li and Zhang, 2005;Li et al., 2008). Images in

43

Fig. 8(A) reveal that very few cells were stained by EthD-1 in each layer after 7-day culture, indicating that good cell viability over 90% for seeded cells is remained for these layers. In addition, it can also be noticed that cell population and size denoted by Calcein-AM staining show appropriate similarity between layers, suggesting that each layer in the scaffold has a similar ability to support the growth of seeded cells. Quantitative measurements for DNA content in scaffolds, as shown in Fig. 8(B), confirm that after 5-day culture collagen/CH–PCL/CS scaffolds were able to well support the proliferation of cells and function as effective as control. Since the DNA content in scaffolds is associated with the number of cells inside scaffolds, the lower DNA content in the first a few days means that a smaller cell number in scaffolds as compared to control. This may be ascribed to (1) some cells might drop into the well rather than staying inside the scaffolds during the cell seeding because the scaffolds were turned over for uniform seeding and some seeded cells were not able to firmly attached to the scaffolds; and (2) the seeded cells needed a certain period of time to migrate throughout the full scaffolds. After 5-day culture, advantages in 3-dimensional culture might become significant so that the number of proliferated cells relatively fast increased and appeared to be comparative to the control. Besides capability for supporting the growth of seeded cells, the EDC-crosslinked collagen/CH–PCL/CS scaffolds also showed proven abilities to facilitate the migration and distribution of the seeded cells throughout the full scaffolds, as shown in Fig. 9. These results can be correlated to components and designed structures of collagen/CH–PCL/CS scaffolds. Collagen, CH–PCL and CS are quite hydrophilic and each of them contains a large number of polar groups. Based on the organization of three components in each layer (see Table 1), there were certain amounts of residual polar groups in each layer except for those consumed by EDC-crosslinking. In addition, each layer inside the EDC-crosslinked collagen/ CH–PCL/CS scaffolds was highly porous (see Fig. 2) and adjacent layers were interconnected by porous interface (see Fig. 1(B)). Accordingly, the hydrophilic surface of pores inside scaffolds was favorable to the movement of seeded cells, polar groups on the surface of pores inside scaffolds could facilitate cell attachment (Cima et al., 1991) and highly interconnected pores in scaffolds allowed cells to migrate through layers. Consequently, the synergistic effect of these factors lead to relatively homogeneous distribution of chondrocytes. It has been suggested that good cell spreading is a favorable characteristic for cell survival (Ruoslahti, 1997) and uniform cell distribution inside a scaffold is desirable for obtaining uniform mechanical strength of engineered cartilage tissue (Grande et al., 1997; Ishaug-Riley et al., 1999). Results in Figs. 8 and 9 confirm that seeded cells can grow on EDC-crosslinked scaffolds with high viability and relatively uniform distribution, suggesting that presently developed collagen/CH–PCL/CS scaffolds are suitable to function as an approving chondrocyte-carrier for articular cartilage repair. Since it is very difficult to fully mimic the structure and properties of articular cartilage ECM, main efforts in this manuscript were made to build such a type of scaffold that has graded composition while showing gradient compressive

44

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

properties. Further studies on evaluation of chondrocyte– scaffold constructs for their protein synthesis and products of proteoglycans and glycoproteins as well as in vivo repair of articular cartilage lesions are now in progress, and relevant results will be presented in separate reports.

5.

Conclusions

Based on presently designed composites consisted of collagen, chitosan–polycaprolactone and chondroitin sulfate as well as a selected crosslinker, it was practical and effective to build a type of crosslinked collagen/chitosan–polycaprolactone/chondroitin sulfate composite scaffolds with stratified porous microarchitecture using a low-temperature deposition technique. Contiguous layers inside the resulting scaffolds were demonstrated to be well connected by porous interface zones without visual crannies, and hierarchically changed average pore-size and porosity for the scaffolds could be achieved using presented composites and processing methods. By devising collagen content changed in a trend contrary to that of chitosan–polycaprolactone and chondroitin sulfate and using an optimal amount of crosslinker, the obtained scaffolds could have high enough wetstate compressive strength comparative to that of native cartilage in knee while showing gradient-changed compressive modulus and swelling index, revealing that the presently developed composite scaffolds are similar to cartilage matrix in the composition of main components, and can partially mimic compressive properties and water content of articular cartilage matrix. It was also found that the optimized scaffolds were able to well support the ingrowth and migration of cells, and the seeded cells showed relatively uniform distribution throughout the scaffolds after culture for a prescribed period of time, suggesting that these scaffolds have promising potential for applications in articular cartilage repair.

Acknowledgment This work was funded by the National Natural Science Foundation of China (Grant no. 81071470).

references

Altala, A., Mooney, D.J., Vacanti, J.P., Langer, R., 1997. Synthetic Biodegradable Polymer Scaffolds. Springer/Birkhauser, Boston199–234. Armstrong, C.G., Mow, V.C., 1982. Variations in the intrinsic mechanical properties of human articular cartilage with age, degeneration, and water content. J. Bone Joint Surg. 64, 88–94. Berthiaume, F., Maguire, T.J., Yarmush, M.L., 2011. Tissue engineering and regenerative medicine: history, progress, and challenges. Ann. Rev. Chem. Biomol. Eng. 2, 403–430. Berger, J., Reist, M., Mayer, J.M., Felt, O., Peppas, N.A., Gurny, R., 2004. Structure and interactions in covalently and ionically crosslinked chitosan hydrogels for biomedical applications. Eur. J. Pharm. Biopharm. 57, 19–34.

Bhattacharya, A., Misra, B.N., 2004. Grafting: a versatile means to modify polymers, techniques, factors and applications. Prog. Polym. Sci. 29, 767–814. Buckwalter, J.A., 1998. Articular cartilage repair: injuries and potential for healing. J. Orthop. Sports Phys. Ther. 28, 192–202. Cao, H., Xu, S.Y., 2008. EDC/NHS-crosslinked type II collagenchondroitin sulfate scaffold: characterization and in vitro evaluation. J. Mater. Sci. Mater. Med. 19, 567–575. Castro, N.J., Hacking, S.A., Zhang, L.G., 2012. Recent progress in interfacial tissue engineering approaches for osteochondral defects. Ann. Biomed. Eng. 40, 1628–1640. Cima, L.G., Vacanti, J.P., Vacanti, C., Ingber, D., Mooney, D., Langer, R., 1991. Tissue engineering by cell transplantation using degradable polymer substrates. J. Biomech. Eng. 113, 143–151. Chang, C.H., Kuo, T.F, Lin, C.C., Chou, C.H., Chen, K.H., Lin, F.H., Liu, H.C., 2006. Tissue engineering-based cartilage repair with allogenous chondrocytes and gelatin-chondroitin- hyaluronan tri-copolymer scaffold: a porcine model assessed at 18, 24, and 36 weeks. Biomaterials 27, 1876–1888. Chen, Y.L., Chen, H.C., Lee, H.P., Chan, H.Y., Hu, Y.C., 2006. Rational development of GAG-augmented chitosan membranes by fractional factorial design methodology. Biomaterials 27, 2222–2232. Chen, Y.L., Lee, H.P., Chan, H.Y., Sung, L.Y., Chen, H.C., Hu, Y.C., 2007. Composite chondroitin-6-sulfate/dermatan sulfate/ chitosan scaffolds for cartilage tissue engineering. Biomaterials 28, 2294–2305. Chiang, H., Jiang, C.C., 2009. Repair of articular cartilage defects: review and perspectives. J. Formos. Med. Assoc. 108, 87–101. Chung, C., Burdick, J.A., 2008. Engineering cartilage tissue. Adv. Drug Deliv. Rev. 60, 243–262. Cruz, D.M.G., Ribelles, J.L.G., Sanchez, M.S., 2008. Blending polysaccharides with biodegradable polymers. I. Properties of chitosan/polycaprolactone blends. J. Biomed. Mater. Res. 85B, 303–313. Dormer, N.H., Berkland, C.J., Detamore, M.S., 2010. Emerging techniques in stratified designs and continuous gradients for tissue engineering of interfaces. Ann. Biomed. Eng. 38, 2121–2141. Gannon, A.R., Nagel, T., Kelly, D.J., 2012. The role of the superficial region in determining the dynamic properties of articular cartilage. Osteoarthr. Cartil. 20, 1417–1425. Grande, D.A., Halberstadt, C., Naughton, G., Schwartz, R., Manji, R., 1997. Evaluation of matrix scaffolds for tissue engineering of articular cartilage grafts. J. Biomed. Mater. Res. 34, 211–220. Guo, Y., Yuan, T., Xiao, Z., Tang, P., Xiao, Y., Fan, Y., Zhang, X., 2012. Hydrogels of collagen/chondroitin sulfate/hyaluronan interpenetrating polymer network for cartilage tissue engineering. J. Mater. Sci. Mater. Med. 23, 2267–2279. Harley, B.A., Lynn, A.K., Wissner-Gross, Z., Bonfield, W., Yannas, I.V., Gibson, L.J., 2010a. Design of a multiphase osteochondral scaffold. II. Fabrication of a mineralized collagen -glycosaminoglycan scaffold. J. Biomed. Mater. Res. 92A, 1066–1077. Harley, B.A., Lynn, A.K., Wissner-Gross, Z., Bonfield, W., Yannas, I.V., Gibson, L.J., 2010b. Design of a multiphase osteochondral scaffold III: fabrication of layered scaffolds with continuous interfaces. J. Biomed. Mater. Res. 92A, 1078–1093. Hennink, W.E., Nostrum, C.F.V., 2012. Novel crosslinking methods to design hydrogels. Adv. Drug Deliv. Rev. 64, 223–236. Hunziker, E.B., 2002. Articular cartilage repair: basic science and clinical progress. Osteoarthr. Cartil. 10, 432–463. Ishaug-Riley, S.L., Okun, L.E., Prado, G., Applegate, M.A., Ratcliffe, A., 1999. Human articular chondrocyte adhesion and proliferation on synthetic biodegradable polymer films. Biomaterials 20, 2245–2256.

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

Jackson, D.W., Scheer, M.J., Simon, T.M., 2001. Cartilage substitutes: overview of basic science and treatment options. J. Am. Acad. Orthop. Surg. 9, 37–52. Jurvelin, J.S., Buschmann, M.D., Hunziker, E.B., 2003. Mechanical anisotropy of the human knee articular cartilage in compression. In: Proceedings of the Institution of Mechanical Engineers, Part H, J. Eng. Med., 217, pp.: 215–219. Kerin, A.J., Wisnom, M.R., Adams, M.A., 1998. The compressive strength of articular cartilage. In: Proceedings of the Institution of Mechanical Engineers, Part H – J. Eng. Med., 212 (H), pp.: 273–280. Kim, T.G., Shin, H., Lim, D.W., 2012. Biomimetic scaffolds for tissue engineering. Adv. Funct. Mater. 22, 2446–2468. LaPorta, T.F., Richter, A., Sgaglione, N.A., Grande, D.A., 2012. Clinical relevance of scaffolds for cartilage engineering. Orthop. Clin. N. Am. 43, 245–254. Leong, K.F., Chua, C.K., Sudarmadji, N., Yeong, W.Y., 2008. Engineering functionally graded tissue engineering scaffolds. J. Mech. Behav. Biomed. Mater. 1, 140–152. Levingstone, T.J., Matsiko, A., Dickson, G., O’Brien, F.J., Gleeson, J.P., 2014. A biomimetic multi-layered collagen-based scaffold for osteochondral repair. Acta Biomater. 10, 1996–2004. Li, Z., Zhang, M., 2005. Chitosan-alginate as scaffolding material for cartilage tissue engineering. J. Biomed. Mater. Res. 75A, 485–493. Li, Z., Gunn, J., Chen, M.H., Cooper, A., Zhang, M., 2008. On-site alginate gelation for enhanced cell proliferation and uniform distribution in porous scaffolds. J. Biomed. Mater. Res. 86A, 552–559. Liu, L., Chen, L., Fang, Y., 2006. Self-catalysis of phthaloylchitosan for graft copolymerization of ε-caprolactone with chitosan. Macromol. Rapid Commun. 27, 1988–1994. Liu, Y., He, T., Gao, C., 2005. Surface modification of poly(ethylene terephthalate) via hydrolysis and layer-by-layer assembly of chitosan and chondroitin sulfate to construct cytocompatible layer for human endothelial cells. Coll. Surf. B 46, 117–126. Marsano, A., Millward-Sadler, S.J., Salter, D.M., Adesida, A., Hardingham, T., Tognana, E., Kon, E., Chiari-Grisar, C., Nehrer, S., Jakob, M., Martin, I., 2007. Differential cartilaginous tissue formation by human synovial membrane, fat pad, meniscus cells and articular chondrocytes. Osteoarthr. Cartil. 15, 48–58. Martinez-Diaz, S., Garcia-Giralt, N., Lebourg, M., Gomez-Tejedor, J.A., Vila, G., Caceres, E., Benito, P., Pradas, M.M., Nogues, X., Ribelles, J.L.G., Monllau, J.C., 2010. In vivo evaluation of 3-dimensional polycaprolactone scaffolds for cartilage repair in rabbits. Am. J. Sports Med. 38, 509–519. Matsiko, A., Levingstone, T.J., O’Brien, F.J., 2013. Advanced strategies for articular cartilage defect repair. Materials 6, 637–668. Miao, X., Sun, D., 2010. Graded/gradient porous biomaterials. Materials 3, 26–47. Moutos, F.T., Freed, L.E., Guilak, F., 2007. A biomimetic threedimensional woven composite scaffold for functional tissue engineering of cartilage. Nat. Mater. 6, 162–167. Mow, V.C., Ratcliffe, A., Poole, A.R., 1992. Cartilage and diarthrodial joints as paradigms for hierarchical materials and structures. Biomaterials 13, 67–97. O’Brien, F.J., Harley, B.A., Yannas, I.V., Gibson, L., 2004. Influence of freezing rate on pore structure in freeze-dried collagen-GAG scaffolds. Biomaterials 25, 1077–1086. Oliveira, J.T., Crawford, A., Mundy, J.M., Moreira, A.R., Gomes, M.E., Hatton, P.V., Reis, R.L., 2007. A cartilage tissue engineering approach combining starch-polycaprolactone fibre mesh scaffolds with bovine articular chondrocytes. J. Mater. Sci. Mater. Med. 18, 295–302. Panadero, J.A., Vikingsson, L., GomezRibelles, J.L., Sencadasa, V., Lanceros-Mendez, S., 2013. Fatigue predictionin fibrin poly-ε-

45

caprolactone macroporous scaffolds. J. Mech. Behav. Biomed. Mater. 28, 55–61. Pieper, J.S., Oosterhof, A., Dijkstra, P.J., Veerkamp, J.H., van Kuppelvelt, T.H., 1999. Preparation and characterization of porous crosslinked collagenous matrices containing bioavailable chondroitin sulphate. Biomaterials 20, 847–858. Raghunath, J., Rollo, J., Sales, K.M., Butler, P.E., Seifalian, A.M., 2007. Biomaterials and scaffold design: key to tissueengineering cartilage. Biotechnol. Appl. Biochem. 46, 73–84. Reis, R.L., Cohn, D. (Eds.), 2002. Kluwer Academic Publishers, Dordrecht, Netherlands. Ruoslahti, E., 1997. Stretching is good for a cell. Science 276, 1345–1346. Seidi, A., Ramalingam, M., Elloumi-Hannachi, I., Ostrovidov, S., Khademhosseini, A., 2011. Gradient biomaterials for soft-tohard interface tissue engineerin. Acta Biomater. 7, 1441–1451. Shepherd, D.E.T., Seedhom, B.B., 1999. The ‘instantaneous’ compressive modulus of human articular cartilage in joints of the lower limb. Rheumatology 38, 124–132. Servaty, R., Schiller, J., Binder, H., Arnold, K., 2001. Hydration of polymeric components of cartilage-an infrared spectroscopic study on hyaluronic acid and chondroitin sulfate. Int. J. Biol. Macromol. 28, 121–127. Shapiro, F., Koide, S., Glimcher, M.J., 1993. Cell origin and differentiation in the repair of full-thickness defects of articular-cartilage. J. Bone Jt. Surg. 75A, 532–553. Silva, J.M., Georgi, N., Costa1, R., Sher, P., Reis, R.L., Blitterswijk, C.A.V., Karperien, M., Mano, J.F., 2013. Nanostructured 3D constructs based on chitosan and chondroitin sulphate multilayers for cartilage tissue engineering. PLos One 8, e55451. Smith, G.D., Knutsen, G., Richardson, J.B., 2005. A clinical review of cartilage repair techniques. J. Bone Jt.Surg. 87, 445–449. Suh, J.K.F., Matthew, H.W.T., 2000. Application of chitosan-based polysaccharide biomaterials in cartilage tissue engineering: a review. Biomaterials 21, 2589–2598. Tan, H., Gong, Y., Lao, L., Mao, Z., Gao, C., 2007. Gelatin/chitosan/ hyaluronan ternary complex scaffold containing basic fibroblast growth factor for cartilage tissue engineering. J. Mater. Sci. Mater. Med. 18, 1961–1968. Thiede, R.M., Lu, Y., Markel, M.D., 2012. A review of the treatment methods for cartilage defects. Vet. Comp. Orthop. Traumatol. 25, 263–272. Tierney, C.M., Haugh, M.G., Liedl, J., Mulcahy, F., Hayes, B., O’Brien, F.J., 2009. The effects of collagen concentration and crosslink density on the biological, structural and mechanical properties of collagen-GAG scaffolds for bone tissue engineering. J. Mech. Behav. Biomed. Mater. 2, 202–209. Treppo, S., Koepp, H., Quan, E.C., Cole, A.A., Kuettner, K.E., Grodzinsky, A.J., 2000. Comparison of biomechanical and biochemical properties of cartilage from humanknee and ankle pairs. J. Orthop. Res. 18, 739–748. Vazquez, J.A., Rodriguez-Amado, I., Montemayor, M.I., Fraguas, J., Gonza´lez, M.D.P., Murado, M.A., 2013. Hondroitin sulfate, hyaluronic acid and chitin/chitosan production using marine waste sources: characteristics, applications and eco-friendly processes: a review. Mar. Drugs 11, 747–774. Waldman, S.D., Spiteri, C.G., Grynpas, M.D., Pilliar, R.M., Kandel, R.A., 2004. Long-term intermittent compressive stimulation improves the composition and mechanical properties of tissue-engineered cartilage. Tissue Eng. 10, 1323–1331. Wan, Y., Creber, K.A.M., Peppley, B., Bui, V.T., 2004a. Ionic conductivity and tensile properties of hydroxyethyl and hydroxypropyl chitosan membranes. J. Polym. Sci. Part B 42, 1379–1397. Wan, Y., Huang, W., Wang, Z., Zhu, X.X., 2004b. Preparation and characterization of high loading porous crosslinked poly(vinyl alcohol) resins. Polymer 45, 71–77.

46

journal of the mechanical behavior of biomedical materials 36 (2014) 32 –46

Wan, Y., Cao, X., Wu, Q., Zhang, S., Wang, S., 2008. Preparation and mechanical properties of poly(chitosan-g-DL-lactic acid) fibrous mesh scaffolds. Polym. Adv. Technol. 19, 114–123. Wan, Y., Gao, J., Zhang, J., Peng, W., Qiu, G., 2010. Biodegradability of conducting chitosang-polycaprolactone/polypyrrole conduits. Polym. Degrad. Stab. 95, 1994–2002. Wu, H., Wan, Y., Dalai, S., Zhang, R., 2010. Response of rat osteoblasts to polycaprolactone/chitosan blend porous scaffolds. J. Biomed. Mater. Res. Part A 92A, 238–245. Wu, L., Ding, J., 2004. In vitro degradation of three-dimensional porous poly(D,L-lactide-co- glycolide) scaffolds for tissue engineering. Biomaterials 25, 5821–5830.

Wu, L., Zhang, J., Jing, D., Ding, J., 2006. “Wet-state” mechanical properties of three-dimensional polyester porous scaffolds. J. Biomed. Mater. Res., Part A 76A, 264–271. Yang, Q., Peng, J., Guo, Q., Huang, J., Zhang, L., Yao, J., Yang, F., Wang, S., Xu, W., Wang, A., Lu, S., 2008. A cartilage MATRIXderived 3-D porous acellular matrix scaffold for in vivo cartilage tissue engineering with PKH26-labeled chondrogenic bone marrow-derived mesenchymal stem cells. Biomaterials 29, 2378–2387.

Designed composites for mimicking compressive mechanical properties of articular cartilage matrix.

Collagen, chitosan-polycaprolactone (CH-PCL) copolymer with PCL content of around 40wt% and chondroitin sulfate (CS) were mixed together at various ra...
3MB Sizes 0 Downloads 3 Views