Scholarly article on topic 'AMECM/DCB scaffold prompts successful total meniscus reconstruction in a rabbit total meniscectomy model'

AMECM/DCB scaffold prompts successful total meniscus reconstruction in a rabbit total meniscectomy model Academic research paper on "Economics and business"

CC BY-NC-ND
0
0
Share paper
Academic journal
Biomaterials
OECD Field of science
Keywords
{Meniscus / Scaffold / "ECM (extracellular matrix)" / "DCB (demineralized cancellous bone)" / " In vivo test"}

Abstract of research paper on Economics and business, author of scientific article — Zhiguo Yuan, Shuyun Liu, Chunxiang Hao, Weimin Guo, Shuang Gao, et al.

Abstract Tissue-engineered meniscus regeneration is a very promising treatment strategy for meniscus lesions. However, generating the scaffold presents a huge challenge for meniscus engineering as this has to meet particular biomechanical and biocompatibility requirements. In this study, we utilized acellular meniscus extracellular matrix (AMECM) and demineralized cancellous bone (DCB) to construct three different types of three-dimensional porous meniscus scaffold: AMECM, DCB, and AMECM/DCB, respectively. We tested the scaffolds' physicochemical characteristics and observed their interactions with meniscus fibrochondrocytes to evaluate their cytocompatibility. We implanted the three different types of scaffold into the medial knee menisci of New Zealand rabbits that had undergone total meniscectomy; negative control rabbits received no implants. The reconstructed menisci and corresponding femoral condyle and tibial plateau cartilage were all evaluated at 3 and 6 months (n = 8). The in vitro study demonstrated that the AMECM/DCB scaffold had the most suitable biomechanical properties, as this produced the greatest compressive and tensile strength scores. The AMECM/DCB and AMECM scaffolds facilitated fibrochondrocyte proliferation and the secretion of collagen and glycosaminoglycans (GAGs) more effectively than did the DCB scaffold. The in vivo experiments demonstrated that both the AMECM/DCB and DCB groups had generated neomeniscus at both 3 and 6 months post-implantation, but there was no obvious meniscus regeneration in the AMECM or control groups, so the neomeniscus analysis could not perform on AMECM and control group. At both 3 and 6 months, histological scores were better for regenerated menisci in the AMECM/DCB than in the DCB group, and significantly better for articular cartilage in the AMECM/DCB group compared with the other three groups. Knee MRI scores (Whole-Organ Magnetic Resonance Imaging Scores (WORMS)) were better in the AMECM/DCB group than in the other three groups at both 3 and 6 months. At both 3 and 6 months, RT-PCR demonstrated that aggrecan, Sox9, and collagen II content was significantly higher, and mechanical testing demonstrated greater tensile strength, in the AMECM/DCB group neomenisci compared with the DCB group.

Academic research paper on topic "AMECM/DCB scaffold prompts successful total meniscus reconstruction in a rabbit total meniscectomy model"

Biomaterials

Contents lists available at ScienceDirect

Biomaterials

journal homepage: www.elsevier.com/locate/biomaterials

AMECM/DCB scaffold prompts successful total meniscus reconstruction in a rabbit total meniscectomy model

Zhiguo Yuan a, Shuyun Liu a, Chunxiang Hao b, Weimin Guo a, Shuang Gao c, Mingjie Wang a, Mingxue Chen a, Zhen Sun a, Yichi Xu a, Yu Wang a, Jiang Peng a, Mei Yuan a, Quan-Yi Guo a' *

a Key Lab of Musculoskeletal Trauma&War Injuries, PLA, Beijing Key Lab of Regenerative Medicine in Orthopedics, Chinese PLA General Hospital, Beijing, 100853, China

b Department of Anesthesia, Chinese PLA General Hospital, Beijing, 100853, China

c Center for Biomaterial and Tissue Engineering, Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, 100871, China

CrossMark

ARTICLE INFO

Article history:

Received 20 May 2016

Received in revised form

25 September 2016

Accepted 25 September 2016

Available online 28 September 2016

Keywords: Meniscus Scaffold

ECM (extracellular matrix)

DCB (demineralized cancellous bone)

In vivo test

ABSTRACT

Tissue-engineered meniscus regeneration is a very promising treatment strategy for meniscus lesions. However, generating the scaffold presents a huge challenge for meniscus engineering as this has to meet particular biomechanical and biocompatibility requirements. In this study, we utilized acellular meniscus extracellular matrix (AMECM) and demineralized cancellous bone (DCB) to construct three different types of three-dimensional porous meniscus scaffold: AMECM, DCB, and AMECM/DCB, respectively. We tested the scaffolds' physicochemical characteristics and observed their interactions with meniscus fibrochondrocytes to evaluate their cytocompatibility. We implanted the three different types of scaffold into the medial knee menisci of New Zealand rabbits that had undergone total meniscectomy; negative control rabbits received no implants. The reconstructed menisci and corresponding femoral condyle and tibial plateau cartilage were all evaluated at 3 and 6 months (n = 8). The in vitro study demonstrated that the AMECM/DCB scaffold had the most suitable biomechanical properties, as this produced the greatest compressive and tensile strength scores. The AMECM/DCB and AMECM scaffolds facilitated fibrochon-drocyte proliferation and the secretion of collagen and glycosaminoglycans (GAGs) more effectively than did the DCB scaffold. The in vivo experiments demonstrated that both the AMECM/DCB and DCB groups had generated neomeniscus at both 3 and 6 months post-implantation, but there was no obvious meniscus regeneration in the AMECM or control groups, so the neomeniscus analysis could not perform on AMECM and control group. At both 3 and 6 months, histological scores were better for regenerated menisci in the AMECM/DCB than in the DCB group, and significantly better for articular cartilage in the AMECM/DCB group compared with the other three groups. Knee MRI scores (Whole-Organ Magnetic Resonance Imaging Scores (WORMS)) were better in the AMECM/DCB group than in the other three groups at both 3 and 6 months. At both 3 and 6 months, RT-PCR demonstrated that aggrecan, Sox9, and collagen II content was significantly higher, and mechanical testing demonstrated greater tensile strength, in the AMECM/DCB group neomenisci compared with the DCB group.

© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND

license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

* Corresponding author.

E-mail addresses: yzgad@163.com (Z. Yuan), clear_ann@163.com (S. Liu), doctorguo@163.com (C. Hao), 603580442@qq.com (W. Guo), 591685142@qq.com (S. Gao), waikejie@163.com (M. Wang), chenmingxueplagh@hotmail.com (M. Chen), sunzhen0614@126.com (Z. Sun), 393632369@qq.com (Y. Xu), wangwangdian628@126.com (Y. Wang), pengjdxx@126.com (J. Peng), 13051529919@126.com (M. Yuan), doctorguo_301@163.com (Q.-Y. Guo).

1. Introduction

The knee meniscus is the crescent-shaped connective tissue between the femoral condyle and tibial plateau. It distributes the body weight, absorbs shock, reduces friction between the tibial surface and femoral condyle, and stabilizes the joint during flexion and extension [1]. Damage to the meniscus is among the most common injuries to the knee joint, often resulting from an impact during sport, or simply due to joint degeneration. Suture repair is

http://dx.doi.org/10.1016/j.biomaterials.2016.09.017

0142-9612/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

feasible for simple longitudinal tears at the periphery of the meniscus, because only the outer part of the meniscus is vascu-larized, that can make the meniscus periphery self-healing available [2]. However, for serious meniscus damage, partial or total meniscectomy is the most common treatment strategy, although this can hasten the degeneration of articular cartilage and induce osteoarthritis [3]. Meniscal allograft transplantation can reduce pain and enhance function in the short term. However, the long-term efficacy of meniscal allografts is uncertain [4].

Tissue engineering has been proposed as a promising therapy for meniscal regeneration [5]. Recently, a number of studies have investigated the possibility of using synthetic scaffolds to completely replace the meniscus [6]. Chang Lee and co-workers [7] used a three-dimensional (3D) printed biomaterial scaffold that releases two recombinant proteins in a spatially and temporally controlled manner to regenerate sheep meniscus. They reported that the biomaterial scaffold promoted knee meniscus regeneration in a large animal model at 3-month follow-up, but the study lacked longer-term evaluation. Merriam and co-workers [8,9] used a novel fiber-reinforced scaffold to regenerate the meniscus in an ovine model, and reported that this scaffold can act as a functional meniscus replacement, while protecting the articular cartilage of the knee after a total meniscectomy. Because of their excellent mechanical properties, polymer scaffolds can perform very well in animal studies in the short term. However, the biocompatibility of polymers is low compared with natural materials and their long-term efficacy after the polymer begins to degrade is unknown [10]. The only natural biological material which has been applied in patients is collagen meniscus implants (CMI) [10], which is made from purified type I collagen derived from bovine Achilles tendon. After preliminary tests on animals, CMI (Menafiex; ReGen Biologics, Hackensack, NJ, USA) was used in human implants. The results demonstrated that this scaffold can improve clinical outcomes in patients with a chronic meniscal injury, and it has subsequently been used in other clinical studies [11,12]. Currently, the CMI scaffold is distributed by Ivy Sports Medicine (ISM) (formerly ReGen Biologics). Other natural materials, including silk and bacterial cellulose gel, have been used to synthesize meniscal scaffolds for in vitro studies only [13,14].

The microenvironment of the meniscal extracellular matrix (ECM) plays an important role in regulating cell behavior, and this may make it a good material for meniscal tissue engineering applications [15]. The meniscus comprises 72% water, 22% collagen and 0.8% glycosaminoglycans (GAGs). Type I collagen predominates (>90% meniscal collagen content) and type II collagen is only found in the inner meniscal regions [16]. Numerous studies have demonstrated that ECM-derived protein molecules can have a positive effect on meniscal tissue regeneration [17-20]. However, none have focused on meniscal ECM as a scaffold for meniscal tissue engineering in vivo. In this study, we utilize meniscus ECM to synthesize a meniscal scaffold and prompt meniscus regeneration in a rabbit total meniscectomy model.

We also considered the possibility that meniscus ECM-derived scaffold may not possess the required biomechanical properties for a complete meniscus implant in vivo, while another natural material, demineralized cancellous bone (DCB), does have these properties (natural 3D pore structure, good biocompatibility, and mechanical strength). DCB has been widely used as a scaffold for tissue-engineered bone/cartilage regeneration, and may also provide a good scaffold for tissue-engineered meniscus regeneration [21].

In this study, we utilized a physical and enzymatic method to decellularize menisci and generate acellular meniscus extracellular matrix (AMECM). A modified version of Urist's protocol was used to obtain DCB. Additionally, the AMECM and DCB were combined to

form an AMECM/DCB scaffold. Therefore, in the present work, we generated three types of meniscus scaffold: an AMECM scaffold, a DCB scaffold, and an AMECM/DCB scaffold. We then tested the physicochemical properties of the three scaffolds, seeded them with meniscus fibrochondrocytes and implanted each subcutane-ously into rats to evaluate their biocompatibility. Finally, we implanted the meniscus scaffolds in a total meniscectomy rabbit model to evaluate their ability to promote meniscus regeneration.

2. Materials and methods

2.1. Scaffold preparation and synthesis

The menisci derived from swine were purchased from a butcher and washed with electrolyzed oxidizing water, distilled water and phosphate-buffered saline (PBS; Sigma-Aldrich Ltd., St. Louis, MO, USA). They were then cut into pieces using scissors, treated with hydrogen peroxide (Sigma-Aldrich), and washed with distilled water again. The minced menisci were treated with pepsin (Sigma-Aldrich) and acetic acid (Sigma-Aldrich), and then homogenized (Kinematica AG, Lucerne, Switzerland) at low temperature (4 °C). A differential centrifugation method with a high-speed, low-temperature centrifuge (Thermo, Osterode, Germany) was used to decellularize the menisci as follows: samples were centrifuged (260 g, 10 min), then the pellet was removed and the supernatant recentrifuged (2400 g, 10 min), with the pellet was removed again and the supernatant recentrifuged (6700 g, 30 min); finally, the supernatant was removed and the remaining meniscal slurry retained. The meniscal slurry was placed into crescent-shaped molds, and freeze-dried to synthesize the AMECM scaffold (Fig. 1A).

Bovine bone was purchased from a butcher and cancellous bone was removed for use in the protocol. We used a modified version of Urist's protocol to prepare DCB as follows: cancellous bone was immersed in ethanol for 3 h, and then in diethyl ether for 1.5 h before being washed with distilled water. It was then stirred in HCl at room temperature for 72 h, washed in distilled water, and immersed in ethanol for 3 h. Finally, the cancellous bone was immersed in diethyl ether for 1.5 h, washed with distilled water and freeze-dried to obtain the DCB. Crescent-shaped molds were used to form the DCB scaffold (Fig. 1A).

We immersed DCB in the meniscal slurry to generate the AMECM and demineralized bone matrix and, after freeze-drying, poured the mixture into crescent-shaped molds to form the AMECM/DCB scaffold (Fig. 1A).

All three types of scaffold were crosslinked using ethyl dimethyl aminopropyl carbodiimide (0.6 mol/l EDAC; Sigma-Aldrich) and sterilized using ethylene oxide. The success of decellularization was evaluated by Hoechst staining and a biochemical analysis was performed to compare the amount of DNA and major ECM components (i.e., GAGs and collagen).

2.2. Characterization of scaffolds

2.2.1. Biochemical analysis

The TIANamp Genomic DNA kit (TIANamp, Beijing, China) was used to extract DNA. To quantify DNA in the three different types of scaffold, the PicoGreen DNA assay kit was used according to the manufacturer's instructions (Invitrogen, Carlsbad, CA, USA). Total collagen was measured by the hydroxyproline assay kit according to the manufacturer's protocol (Nanjing Jiancheng, Jiangsu, China). GAG content was determined by 1, 9-dimethylmethylene blue (DMMB) assay using the Tissue GAG Total Content DMMB Color-imetry kit according to the manufacturer's instructions (Genmed Scientific Inc., Shanghai, China).

Fig. 1. A. Flow chart of the preparation of the three different scaffolds. B. Overview of experimental design. (Abbreviations: AMECM - acellular meniscus extracellular matrix; DCB -demineralized cancellous bone; AMECM/DCB - acellular meniscus extracellular matrix/demineralized cancellous bone).

2.2.2. Histological and immunohistochemical staining

The three different scaffolds were fixed in 4% (v/v) para-formaldehyde solution (Sigma), embedded in paraffin, and sectioned into 7 mm slices. The sections were stained with tolu-idine blue. Safranin-O was used to stain sulfated proteoglycans and Sirius Red was used to stain total collagen in the matrix. For immunohistochemical staining with polyclonal antibodies against collagen I and II (Abcam, Cambridge, UK), the sections were deparaffinized, hydrated, and permeabilized. Sections were then washed in PBS before microwave antigen retrieval was applied. Sections were washed again in PBS, treated with 0.5% (v/v) hydrogen peroxide for 10 min, and incubated with primary antibodies overnight at 4 °C before being washed and incubated with horseradish-conjugated secondary antibodies (ABC kit; Vector Laboratories Inc., Burlingame, CA, USA). The signal was developed using diaminobenzidine (DAB) (Vector Laboratories) and nuclei were counterstained with hematoxylin.

2.2.3. Scanning electron microscopy (SEM)

The scaffolds were fixed with 2.5% (v/v) glutaraldehyde buffered with PBS, dehydrated using a graded series of ethanol washes, and dried to a critical point (EM CPD300; Leica, Wetzlar, Germany) using carbon dioxide (CO2). Samples were sputter coated with gold prior to SEM observation. The pore size and microstructure of the scaffolds were observed using SEM (S-4800 field emission scanning electron microscope; Hitachi, Tokyo, Japan), and the mean pore size is calculated through SEM.

2.2.4. Water absorption and porosity

Water absorption of the scaffolds was measured as follows: Dry scaffolds were weighed (W0) then immersed in PBS buffer at room temperature overnight to allow free liquid absorption. The wet scaffolds were removed and weighed (W1) and the percentage water absorption was calculated according to the following equation:

%Water absorption = [(W1 - W0)/W0] x 100%

The porosity of the scaffolds was measured as follows: the volumes of ethanol vials were measured (V1), scaffolds were placed in the vials for 10 min and the volumes were measured (V2). Scaffolds were then removed and the volume of the remaining ethanol measured (V3). This procedure was repeated five times and the mean values were used to calculate the porosity of the scaffolds according to the following equation:

Porosity = [(V1 - V3)/(V2 - V3)] x 100%

2.2.5. Mechanical testing

Scaffold specimens of different sizes were prepared for mechanical testing. For unconfined compressive strength, 5 x 5 x 5 mm cubes of each scaffold were tested using a BOSE biomechanical testing machine (BOSE 5100; TE Instruments, New Castle, DE, USA). For tensile strength, 2 x 5 x 10 mm specimens were tested using a uniaxial materials testing machine (model 5969; Instron, High Wycombe, UK). All specimens were kept moist using PBS throughout these tests.

2.3. In vitro cytocompatibility studies

2.3.1. Rabbit meniscus fibrochondrocyte isolation and culture

Rabbit meniscal fibrochondrocytes were isolated as follows: Menisci were dissected from rabbit knees and sliced into 1 x 1 x 1 mm prills. These were digested for 2 h on a magnetic stirrer using collagenase, with 100 lg/mL penicillin and 100 lg/mL streptomycin, in Dulbecco's modified Eagle's medium (DMEM) (Corning, Glendale, AZ, USA). The cells were isolated by centrifu-gation (1500 rpm for 5 min) and cultured in DMEM with 10% FBS (Sigma-Aldrich) in 75 cm2 fiasks (BD Bioscience, San Jose, CA, USA). After 90% confiuence was reached, the cells were diluted

(1:3) and passage 3 fibrochondrocytes were used for the cyto-compatibility studies.

2.3.2. Cell seeding on scaffolds

Before seeding with cells, scaffolds were cut into discs (2 mm thick, 5 mm in diameter); they were then sterilized, washed in sterile PBS, and treated with DMEM (Corning) overnight. Each scaffold was seeded with 1 x 106 fibrochondrocytes in 100 pL medium (DMEM + 10% FBS). Cells were allowed to adhere for 2 h; then, more media was added and this was changed every 2 d over the following 2 weeks.

2.3.3. Scanning electron microscopy

We utilize the SEM to observe the microstructure of the cell-scaffold composites and the growth of fibrochondrocytes cultured in vitro on the scaffolds. Cell-scaffold composites were acquired for 3 d after seeding the cells. Samples were fixed in 2.5% (v/v) glutaraldehyde, buffered with PBS and treated as described above. After coating with gold, cell-scaffold composites were observed using an S-4800 field emission SEM (Hitachi).

2.3.4. Cell viability staining

To evaluate the cytotoxicity of the three different types of scaffold using fibrochondrocytes, cell viability was observed with a Live/Dead Assay kit (Invitrogen) after cells were seeded and cultured for 3 days. The cell/scaffold constructs were washed with sterile PBS (3 x 5 min) and incubated in PBS solution with 2 mM calcein AM and 4 mM ethidium homodimer-1 for 1 h at room temperature. After another wash with sterile PBS, constructs were observed using a Leica TCS-SP8 confocal microscope (Leica) and the images analyzed with Imaris software (ver. 7.4.; Bitplane, Zurich, Switzerland). Cell viability was calculated as follows: (live cells/ total cells) x 100%.

2.3.5. Biochemical assays for DNA, GAG and collagen

After 3, 7 and 14 d of culture, the cell/scaffold constructs were assayed for DNA, GAG and collagen. The TIANamp Genomic DNA kit was used to extract DNA (TIANamp), and this was quantified using the PicoGreen DNA assay kit (Invitrogen), all as described above.

For sGAG quantification, the 1, 9-dimethylmethylene blue (DMMB) assay and Tissue GAG Total Content DMMB Colorimetry kit (GenMed) were used according to the manufacturer's instructions, and as described above. The sGAG secreted by fibrochondrocytes were measured in two parts: the sGAG in cell/scaffold constructs and the sGAG in the media. For quantification of collagen, the hy-droxyproline assay kit was used according to the manufacturer's protocol (Nanjing Jiancheng). Collagen was also measured in two parts: collagen in cell/scaffold constructs and collagen in the media.

2.4. In vivo animal studies

2.4.1. Surgical procedure

This study was performed under a protocol approved by the Institutional Animal Care and Use Committee at PLA General Hospital. In total, 32 3-month old New Zealand white rabbits were randomly allocated into four groups of 8 rabbits as follows: a control group with meniscectomy only and no graft implant; an AEMCM group with meniscectomy and AMECM scaffold implant; a DCB group with meniscectomy and DCM scaffold implant; and an AEMCM/DCB group with meniscectomy and AMECM/DCB scaffold implant. Under anesthetic with intramuscular injections of 160 mg ketamine and 12 mg xylazine, the medial collateral ligament was cut down to expose the posterior horn of the medial meniscus. Meniscectomy was performed in both stifle joints of all rabbits and the test groups were implanted with the appropriate scaffolds. The

control group were not implanted. The implants were sutured to the capsule at the level of the original meniscal rim, and the anterior and posterior horns were fixed to the meniscal ligaments. The capsule was then closed and the medial collateral ligament reconstructed using resorbable sutures. After the operation, each rabbit was given intramuscular penicillin injections to prevent infection, and returned to its cage where it was free to move. All rabbits were euthanized and evaluated at the appropriate time point (3 or 6 months).

2.4.2. Macroscopic observations

The tibial plateau with the meniscus and the femoral condyles were observed and photographed [22,23].

2.4.3. Histological and immunohistochemical analyses

The regenerated meniscal tissue was fixed in 4% para-formaldehyde for 3 days, and embedded in paraffin. The corresponding distal femur and proximal tibia were decalcified in 10% EDTA solution for 28 days after being fixed in 4% paraformaldehyde for 3 days, and then also embedded in paraffin. The regenerated meniscus and bone-cartilage were sectioned into 7 mm slices and stained with H&E, toluidine blue, safranin-O and Sirius Red. Meniscal regeneration was evaluated quantitatively using the Ish-ida score [24]. Cartilage degeneration of the femoral condyle and tibial plateau were evaluated using the Mankin score [25]. The native meniscus and articular cartilage derived from healthy rabbits were compared as a control.

2.4.4. Magnetic resonance imaging (MRI)

The MRI scans were performed using a 7.0 T Bruker Biospec system (Bruker, Ettlingen, Germany). The rabbit knee joints were positioned straight (i.e., not flexed) in an MRI-compatible device. A small animal-specific knee coil was used as the transmitter coil and a separate quadrature surface coil (Bruker) was placed above the knee joint to achieve maximal signal reception. T2-weighted imaging (T2WI) was performed (repetition time = 3200 ms/echo time = 65 ms; slices = 15, slice thickness = 1 mm, replicate measurements = 3). In addition to the experimental and control groups, healthy rabbit knee MRI scans were also conducted as a positive control.

2.4.5. RT-PCR

Total RNA was extracted from the regenerated menisci using a standard Trizol procedure (Invitrogen), and the concentration and purity of the RNA was checked with a NanoDrop ND-2000 Spec-trophotometer (Thermo Fisher, Wilmington, MA, USA). To ensure the purity of the RNA, only samples with an absorbance 260/ 280 nm ratio of >1.8 were analyzed. The mRNA were reverse-transcribed into cDNA using a ReverTra Ace kit (Toyobo, Osaka, Japan). Reactions were conducted at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15s, 58 °C for 15s, 72 °C for 35s, and finally 60 °C for 1 min. Gene expression of aggrecan, SOX9 and collagen types I and II were quantified using real-time PCR on a LightCycler 480 system (Roche Applied Science, Indianapolis, IN, USA). Target genes were amplified using specific primers: aggrecan (XM_002-723376.1, forward: 5'-GGAGGAGCAGGAGTTTGTCAA-3' and reverse: 5'-TGTCCATCCGACCAGCGAAA-3'), SOX9 (XM_002719499, forward: 5'-GCGGAGGAAGTCGGTGAAGAAT-3' and reverse: 5'-AAGATGGC-GTTGGGCGAGAT-3'), collagen I (NM_001195668.1, forward: 5'-GCCACCTGCCAGTCTTTACA-3' and reverse: 5'-CCATCATCACCATCT-CTGCCT-3'), and collagen II (XM_002723438.1, forward: 5'-CACGCTCAAGTCCCTCAACA-3', and reverse: 5'-TCTATCCAGTAGT-CACCGCTCT-3'). The house-keeping gene, glyceraldehyde-3-phosphatedehydrogenase (GAPDH), was used as a reference gene (NM_001082253.1, forward: 5'-CAAGAAGGTGGTGAAGCAGG-3' and

reverse: 5'-CACTGTTGAAGTCGCAGGAG-3'). The mRNA levels of aggrecan, SOX9, collagen I and collagen II were all normalized to the value of GAPDH at the corresponding time points.

2.4.6. Tensile strength test

To test tensile strength, regenerated menisci from months 3 and 6 were cut into tubular sections (2 x 3 x 10 mm) along the long axis. The sections were tested using a materials testing machine (model 5969, Instron). Sections were preloaded to 0.5 N, preconditioned from 0% to 2% strain for 15 cycles, and tension was increased at a rate of 0.5% per second. The tensile modulus was determined by the stress-strain curve.

2.5. Statistical analysis

Statistical analysis was performed with one-way analysis of variance (ANOVA) using PASW for Windows software (ver. 18.0; SPSS Inc., Chicago, IL, USA). All data were expressed as means ± standard deviation (SD), and overall significance was set at p < 0.05.

3. Results

3.1. Physicochemical and biological characteristics of the scaffolds

3.1.1. Scaffold microstructure

A macroscopic photograph of the three different types of scaffold is shown in Fig. 2A, and SEM images in Fig. 2D show circular pores in the size range of 192.6 ± 11.5 mm for the AMECM scaffold, 316.5 ± 40.4 mm for the DCB scaffold and 122.0 ± 26.6 mm for the AMECM/DCB scaffold (Table 1). The porosity of each scaffold was also measured: 90.54 ± 0.71% (n = 5) for the AMECM scaffold, 95.05 ± 0.86% (n = 5) for the DCB scaffold and 90.60 ± 1.49% (n = 5) for the AMECM/DCB scaffold (Table 1). In addition, percentage water absorption was recorded: 3151.23 ± 68.15% for the AMECM scaffold, 956.56 ± 95.34% for the DCB scaffold and 1431.78 ± 56.49% for the AMECM/DCB scaffold (Table 1).

Hoechst 33258 staining demonstrated that there were no residual nuclei in any of the three scaffold types (Fig. 2B), and this was confirmed by residual DNA quantification in all three scaffold types. The residual DNA concentration was less than 20 pg/mg (Fig. 2C), indicating that most cells were removed after decellularization.

3.1.2. Histological staining and biochemical analysis

The three different types of scaffold were analyzed by histo-logical staining and biochemical analysis. Toluidine blue and safranin-O staining were positive in the AMECM and AMECM/DCB scaffolds, but negative in the DCB scaffold (Fig. 2E). A positive result with these stains indicates that GAG content derived from the meniscal ECM remains after decellularization. Sirius red staining (Fig. 2E) shows abundant type I collagen as red or yellow fibers in all three scaffolds. A small amount of type II collagen appears, but is only loosely held-together, in the AMECM and AMECM/DCB scaffolds, while no type II collagen is seen at all in the DCB scaffold. This result was confirmed by immunohistochemical analysis for collagen type I and type II in the three scaffold types (Fig. 2E). The biochemical analysis of total GAG and collagen in the three different scaffolds is shown in Fig. 2F and G, and this is consistent with the histological result. The GAG content in the AMECM and AMECM/ DCB scaffolds was higher than in the DCB scaffold (p < 0.05), but there was no significant difference in collagen content among the three different scaffold types.

3.1.3. Mechanical characterization

To evaluate biomechanical properties, compressive and tensile

strength testing was used to compare the three different scaffolds. The compressive modulus was approximately 2.44 ± 0.78 kPa for the AMECM scaffold, 25.85 ± 3.92 kPa for the DCB scaffold and 86.34 ± 8.65 kPa for the AMECM/DCB scaffold (Fig. 2H). The tensile modulus was approximately 0.24 ± 0.16 kPa for the AMECM scaffold, 9.57 ± 3.11 kPa for the DCB scaffold and 19.09 ± 2.38 kPa for the AMECM/DCB scaffold (Fig. 2F). Therefore, the AMECM/DCB scaffold is clearly stronger (both in terms of compressive and tensile strength) than the AMECM and DCB scaffolds.

3.2. Cytocompatibility analysis of scaffolds using meniscus fibrochondrocytes

3.2.1. Fibrochondrocyte attachment and viability on the scaffolds

The attachment of fibrochondrocytes to the three different

scaffolds was observed using SEM (Fig. 3A). The fibrochondrocytes attached to all three scaffolds and migrated well between interconnecting pores in the AMECM and AMECM/DCB scaffolds over a 3-day culture period. Fibrochondrocytes were less well distributed between interconnecting pores in the DCB scaffold after 3 days of culture. This is probably because pore size in the DCB scaffold was greater than in the AMECM and AMECM/DCB scaffolds, making strong attachment to the interconnecting pores more difficult.

Fibrochondrocyte viability on the three different scaffolds was evaluated by live/dead cell staining after 3 days of culture. For the AMECM and AMECM/DCB scaffolds, most cells were stained fluorescent green (living cells), with very few red (dead) cells. Most cells were also alive on the DCB scaffold; however, the proportion of dead cells increased. The quantitative cell viability analysis (n = 5) suggested that cell viability on the AMECM and AMECM/DCB scaffolds may be higher than on the DCB scaffold, but did not demonstrate any significant differences (Fig. 3B).

3.2.2. Biochemical analysis

Quantification of DNA, total GAG, and collagen was used to measure the proliferation of fibrochondrocytes and ECM deposition. DNA content measurements suggested that fibrochondrocytes increased by approximately 37%, 28%, and 36% at 7 days and by approximately 91%, 78%, and 89% at 14 days compared with the number of cells at 3 days, for the AMECM, DCB, and AMECM/DCB scaffolds, respectively (Fig. 3C1).

Both sGAG and collagen content increased over time in each of the three scaffolds (p < 0.01) (day 14 vs. day 7; day 7 vs. day 3). Total sGAG content (scaffold + media) increased by approximately 70%, 62%, and 79% at 7 days and by approximately 233%, 210%, and 251% at 14 days compared with sGAG secreted by the fibrochondrocytes at 3 days, for the AMECM, DCB, and AMECM/DCB scaffolds, respectively (Fig. 3C2). To evaluate the sGAG secreted by fibrochondrocytes, we calculated sGAG normalized for DNA content. This increased over time, as measured at days 3, 7, and 14 (p < 0.01) (Fig. 3C3). sGAG deposition in the AMECM and AMECM/DCB scaffolds was higher than in the DCB scaffold for 7 and 14 day-old cultures, and this was confirmed by normalized sGAG per DNA content analysis. Therefore, the AMECM and AMECM/DCB scaffolds can promote fibrochondrocyte proliferation and sGAG synthesis.

As collagen can also be secreted into the medium while fibro-chondrocytes are being cultured, total collagen was estimated both as that present in the media and that deposited in the scaffolds. Total collagen from the fibrochondrocytes increased by approximately 80%, 57%, and 85% at 7 days and by approximately 218%, 181%, and 237% at 14 days compared with collagen secreted by fibrochondrocytes at 3 days, for the AMECM, DCB, and AMECM/DCB scaffolds, respectively (Fig. 3C4). Similarly, total collagen normalized for DNA content in the AMECM and AMECM/DCB scaffolds was significantly higher than in the DCB scaffold at 7 days and 14 days

Fig. 2. The physicochemical and histological properties of the three different scaffolds. A, Macroscopic features of the three different scaffolds. B, Hoechst staining of the three different scaffolds. C, Residual DNA quantitative analysis of the three different scaffolds; values are presented as means ± standard deviation (n = 5). D, Scanning electron micrographs of the three different scaffolds. E, Histological analysis of the three different scaffolds and native meniscus; sections from the three different scaffolds and native meniscus stained with toluidine blue (B), Safranin-O (SO), Sirius Red, and immunohistochemical analysis for type I collagen (Col 1) and type II collagen (Col 2). F, Glycosaminoglycan (GAG) content of the three different scaffolds; values are presented as means ± standard deviation (n = 5; *p < 0.05 and **p < 0.01). G, Total collagen content of the three different scaffolds and native meniscus (n = 5). H, Comparative graph showing the compressive modulus of the three different scaffolds; data are expressed as means ± standard deviation (n = 5; **p < 0.01). I, Comparative graph showing tensile modulus of the three different scaffolds; data are expressed as means ± standard deviation (n = 5; **p < 0.01).

respectively (Fig. 3C5). All data indicate that the AMECM and AMECM/DCB scaffolds can promote the secretion of collagen by fibrochondrocytes.

3.3. The AMECM/DCB scaffold promoted meniscus regeneration

To compare their effect on meniscus regeneration, we implanted the AMECM, DCB and AMECM/DCB scaffolds in rabbits that had undergone total meniscectomy (Fig. 4A). Macroscopically, the control and AMECM groups displayed no obvious meniscal tissue regeneration, and the corresponding tibial plateau and femoral condyle cartilage showed significant damage at 3 and 6 months post-operation. In contrast, the DCB and AMECM/DCB groups did have regenerated meniscal tissue and the corresponding cartilage was not as seriously damaged as it was in the control and AMECM groups. Additionally, we found that meniscus regenerated in the AMECM/DCB group was more similar to native meniscus than that of the DCB group, and that regenerated meniscus in the AMECM/ DCB group was the most similar of all, in terms of appearance, to native meniscus at 6 months post-operation (Fig. 4B). As there is no obvious neomeniscus emerge in AMECM and control group, the relevant testing could not be performed.

Histological H&E staining of the DCB group neomenisci 3 months after surgery (Fig. 5A) demonstrated the appearance of abundant elongated fibroblast-like cells, while no chondrocyte-like cells were apparent. Toluidine blue staining was negative. Collagen I immunohistochemical staining was positive, while collagen II staining was negative. Picrosirius Red staining also highlighted a tangled arrangement of collagen fibers. H&E staining of the DCB group neomenisci 6 months after surgery shows the appearance of small, rounded chondrocyte-like cells. Toluidine blue staining is weakly positive; Collagen I immunohistochemical staining is positive, while collagen II staining is weakly positive. Picrosirius red staining still shows that the arrangement of collagen fibers is disordered.

Histological H&E staining of the AMECM/DCB group neomenisci 3 months after surgery showed small chondrocyte-like cells emerging. Toluidine blue staining is weakly positive; collagen I immunohistochemical staining is positive, while collagen II staining is weakly positive. Picrosirius red staining shows a disordered arrangement of collagen fibers. In contrast, H&E staining of the AMECM/DCB group neomenisci 6 months after surgery shows abundant chondrocyte-like cells and cartilage lacuna emerging. Toluidine blue staining is strongly positive. Collagen I immuno-histochemical staining is positive, collagen II staining is also positive. Picrosirius red staining now shows a well-organized arrangement of collagen.

We also examined one 3-month old healthy rabbit as a control. This comparison indicated that the histological characteristics of neomeniscus in the AMECM/DCB group at 6 months is the most similar to native meniscus.

Histological scores for regenerated menisci in the AMECM/DCB group were better than those in the DCB group at 3 and 6 months (Fig. 5C).

3.4. Menisci regenerated using the scaffolds prevented cartilage degeneration

Next, we evaluated the chondroprotective effect of regenerated meniscus by histological examination (toluidine blue) of the corresponding femoral condyle and tibial plateau cartilage (Fig. 5B). Histological examination yielded results similar to the macroscopic findings. Cartilage degeneration progressed in the control and AMECM group over 6 months, whereas it was better preserved in the DCB and AMECM/DCB groups, with cartilage in the AMECM/

Table 1

Comparison of the pore size, porosity and water absorption of the three scaffolds.

AMECM DCB AMECM/DCBS

Pore size (mm) 192.6 ± 11.5 316.5 ± 40.4 122.0 ± 26.6 (*)

Porosity (%) 90.54 ± 0.71 95.05 ± 0.86 90.60 ± 1.49 (*)

Water absorption (%) 3151.23 ± 68.15 956.56 ± 95.34 1431.78 ± 56.49 (*)

AMECM, acellular meniscus extracellular matrix; DCB, demineralized cancellous bone.

(n = 5, *p < 0.05).

DCB group being the best preserved over the 6 month period. Mankin scores for both the femoral condyle (Fig. 5D) and tibial plateau (Fig. 5E) cartilage in the AMECM/DCB group at 3 and 6 months were significantly better than those of the other three groups.

3.5. MRI assessment of rabbit knees after scaffold implants

We evaluated the degeneration of rabbit knee joints MRI. MRI evaluation (Fig. 6A), showed massive inflammatory signals emerging from the knee joints of the control and AMECM groups, weak inflammatory signals in the DCB group, and no obvious inflammatory signals in the AMECM/DCB group. While there were clear signals indicating the appearance of meniscus-like tissue in the meniscus triangle region of the coronal and sagittal projection in the AMECM/DCB and DCB groups, only inflammatory signals were observed in the meniscus triangle region of the control and AMECM groups. When semi-quantitative Whole-Organ Magnetic Resonance Imaging Scores (WORMS) [26]were used to grade the different groups, the AMECM/DCB group knees were significantly healthier than those of the other three groups (Fig. 6B).

3.6. The regenerated meniscus restored biomechanical properties and gene expression

All rabbits that underwent operations resumed weight bearing and locomotion after surgery. Gene expression in the regenerating menisci was shown to vary. Aggrecan, SOX9, and collagen II gene expression levels increased in regenerating menisci over 6 months in the DCB and AMECM/DCB groups, whereas for collagen I the increase in expression level was not statistically significant. The aggrecan, SOX9, and collagen II expression levels in regenerated menisci from the AMECM/DCB group were significantly higher than in the DCB group at 3 and 6 months (Fig. 7A).

The tensile modulus of regenerated meniscus increased over 6 months in the DCB and AMECM/DCB groups, and was significantly higher in the AMECM/DCB group than in the DCB group at 3 and 6 months. The tensile modulus of regenerated meniscus in the AMECM/DCB group at 6 months is similar to that of native meniscus (Fig. 7B).

4. Discussion

In this study, we used natural meniscus ECM and DCB as scaffolds for meniscus engineering. The ECM provides the necessary microenvironment to regulate cell behaviors, including cell migration and matrix synthesis [16]. The DCB have natural 3D pore structure, good biocompatibility, especially, the well mechanical strength. And both the in vitro and in vivo study confirm that the AMECM/DCB combination is greater than either on its own. Compare to polymers materials, which have been widely used in meniscus regeneration due to their biodegradability and biome-chanical properties [27,28], the AMECM/DCB scaffold has the better biocompatibility and can preferably facilitate cell proliferation and

Fig. 3. Cytocompatibility analysis of the three different scaffolds. A, Scanning electron micrographs of the three different scaffolds on which rabbit fibrochondrocytes were seeded for 3 days. B, Live/dead cell analysis for the three different scaffolds on which rabbit fibrochondrocytes were seeded for 3 days; representative images show dead (red) cells, live (green) cells and 3D reconstruction images of the fibrochondrocyte distribution in the three different scaffolds, and the viability analysis for the fibrochondrocytes on the three different scaffolds. C, Biochemical assay results showing: (Cj) DNA content; (C2) total GAG content; (C3) GAG/DNA; (C4) total collagen; (C5) collagen/DNA, estimated in three scaffolds individually seeded with fibrochondrocytes after days 3, 7 and 14. Data are expressed as means ± standard deviation (n = 4; *p < 0.05, **p < 0.01).

Fig. 4. Macroscopic analyses of regenerated menisci and the corresponding femoral condyles. A, Surgical strategy and study schema. B, Macroscopic observations; red dashed line indicates regenerated meniscus; black dashed line indicates native meniscus.

differentiation, additionally, it is relatively easy to observe the complete meniscus repair process in vivo as both meniscus ECM and DCB are easily degraded, and the degradation products may themselves be used for meniscus reconstruction by meniscus cells, while the polymer degradation may interfere with meniscus regeneration [29]. Compare to other collagenic meniscus scaffolds, such as CMI, derived from purified type I collagen of bovine Achilles tendon, the AMECM/DCB scaffold may provide the better microbioenvironment, which contain the meniscus original "soil", meniscus ECM.

The in vitro experiments found that the AMECM/DCB composite scaffold had better biomechanical properties than the AMECM or DCB scaffolds. This may be because the AMECM scaffold consists of thin collagen fibers and glycosaminoglycan particles (analogous to a concrete filler), and the DCB scaffold consists of bulky fibers (similar to a bar framework), so that the AMECM/DCB scaffold can be regarded as analogous to reinforced concrete, greatly enhancing its biomechanical properties. Fibrochondrocytes grow well on all three types of meniscus scaffold. Both the AMECM and AMECM/ DCB scaffolds promote fibrochondrocyte proliferation and

Fig. 5. Histological analyses of regenerated menisci and corresponding cartilage. A, Neomeniscus sections stained with H&E, TB, Picrosirius Red (PR), and immunostained for Col I and Col II. One 3 month old healthy rabbit was used as a control. B, Sagittal sections from the medial femoral condyle cartilage and medial tibial plateau cartilage of rabbits from each group, stained with toluidine blue 3 months and 6 months after surgery. One 3 month old normal rabbit was used as a control. C, Ishida histological score for regenerated meniscus. D, Mankin scores for medial femoral condyle cartilage. E, Mankin scores for medial tibial plateau cartilage. Data are expressed as mean ± standard deviation (n = 3; *p < 0.05).

Fig. 6. Magnetic Resonance Imaging (MRI) scans. A, MRI of rabbit knees 3 months and 6 months after surgery. The left panels show coronal scanning and the right panels show sagittal scanning. One 3 month old healthy rabbit was used as a control. The white arrow indicates synovial fluid effusion, The red arrow indicates the regenerated menisci. B, The Whole-Organ Magnetic Resonance Imaging Score (WORMS) for MRI. Data are expressed as means ± standard deviation (n = 8; *p < 0.05, **p < 0.01).

secretion of sGAG and collagen better than the DCB scaffold. This may be because both the AMECM and AMECM/DCB scaffolds consist of meniscus ECM molecules that play essential roles in regulating cell behaviors, including cell migration, proliferation and differentiation.

The in vivo study found regenerated meniscus appearing macroscopically in the AMECM/DCB and DCB groups but not in AMECM group. This may be because the poor biomechanical properties of the AMECM scaffold made it hard to secure during the operations and it was subsequently degraded too easily. After 6 months, the regenerated menisci of the AMECM/DCB group were histologically more similar to native menisci than those of the DCB group. Toluidine blue and type II collagen staining was more extensive in regenerated menisci of the AMECM/DCB group, and histological scores for neomenisci of the AMECM/DCB group were better than those of the DCB group. These findings indicate the

efficacy of the AMECM/DCB scaffold in promoting meniscus regeneration. A histological evaluation of cartilage from the femoral condyle and tibial plateau also demonstrates that regenerated menisci from the AMECM/DCB group are more effective in preventing cartilage degeneration and the progress of osteoarthritis than those from the DCB group, a finding that is confirmed by MRI scans. It was discovered that the biomechanical properties and gene expression characteristics of regenerated menisci in the AMECM/DCB and DCB groups were more similar to those of native menisci at 6 months than they were at 3 months, and that regenerated menisci from the AMECM/DCB group were more similar than those of the DCB group to native menisci. Therefore, regeneration of menisci with the necessary functional properties and tissue characteristics can be orchestrated by the AMECM/DCB scaffold, which provides an enhanced protective effect on the articular surfaces compared with the other scaffolds tested.

Fig. 7. Gene expression levels and tensile modulus of the neomeniscus. A, Real-time gene expression results showing fold increases in aggrecan, SOX9, collagen I, and collagen II gene expression in neomenisci of different groups 3 and 6 months after surgery. One native rabbit was used as a control. B, The tensile modulus of neomenisci in different groups 3 and 6 months after surgery. The 6-month-old and 9-month-old native rabbits were used as control respectively. Data are expressed as means ± standard deviation (n = 3; *p < 0.05, **p < 0.01).

Unfortunately, despite we try our best to perform the experiment, there are still some limitations in this study. Here, in order to make the paper more rigorous as well as make our research more perfect in the future, we listed the limitations in the following. A drawback of the animal model, rabbits especially 3-month-old rabbits, is its more tendency to regeneration than the large animal models, such as the sheep model, which would be used in our next experiment. Another limitation is the undone compression testing for the neomeniscus in the in-vivo study. Although there are some limitations in this paper, the study still can confirm that the AMECM/DCB scaffold can better promotes meniscus regeneration and prevents cartilage degeneration in a rabbit meniscus defect model compared with the other scaffolds tested.

In contrast to cell transplantation, the present strategy serves as a therapeutic prototype for meniscus regeneration and does not require cell seeding [30]. The bioactive scaffold recruits endogenous cells and provides an appropriate cartilage-inducing microenvironment for cell proliferation and differentiation [31]. The present study shows that an AMECM/DCB acellular biomaterial scaffold can prompt successful total meniscus regeneration in a rabbit total meniscectomy model by recruiting endogenous cells.

The source of endogenous cells is unclear but may include the synovium and vascular stem cells. Other studies have also indicated the extensive involvement of endogenous cells in meniscus regeneration. Ozeki and co-workers [23] transplanted Achilles tendons from wild-type rats into meniscal defects in LacZ-transgenic rat knees. The transplanted tissue was subsequently covered by LacZ-transgenic rat synovium, indicating that synovial coverage by the host knee joint plays an important role in supporting meniscus remodeling and regeneration. Researchers from Chang Lee's laboratory [7] used a 3D-printed meniscus scaffold and spatiotemporal release of CTGF and TGF$3 without cell seeding to prompt meniscus regeneration in a sheep partial meniscectomy model, also indicating that endogenous cells can be recruited for meniscus regeneration. Merriam and co-workers [8] implanted a novel fiber-reinforced meniscus scaffold in an ovine total menis-cectomy model, and demonstrated that the scaffold could successfully induce the formation of neomeniscus tissue that remained intact and functional, also without the need for cell seeding. These findings suggest that cell infiltration can play a crucial role in tissue engineering meniscus regeneration and can produce the appropriate proteins and structure in their native locations.

Fig. 8. Mechanism of meniscus regeneration. Meniscus scaffolds, host knee synovial tissue and endogenous stem cells contributed to meniscus regeneration in these experiments.

We also investigated possible mechanisms of meniscus regeneration in this study (Fig. 8). When meniscus defects were not treated (control group) or treated using the AMECM scaffold, only a small amount of synovial tissue proliferation was observed in situ and this had no effect on preventing cartilage degeneration. Because the DCB scaffold has good biomechanical properties and can be fixed in situ to the rabbit knee, meniscus defects treated with DCB scaffold synovial tissue and other endogenous cells supported neomeniscus tissue formation and prevented cartilage degeneration and the progression of osteoarthritis. The AMECM/DCB scaffold has even better biomechanical characteristics and ECM protein molecules can stimulate cell behaviors, including cell proliferation, migration, and differentiation that enhance the capacity of the AMECM/DCB scaffold to promote neomeniscus formation and prevent cartilage degeneration and the progress of osteoarthritis.

5. Conclusion

The AMECM/DCB scaffold has good biocompatibility and biomechanical characteristics. It promotes fibrochondrocyte proliferation and the secretion of collagen and glycosaminoglycan. It also promotes meniscus regeneration and prevents cartilage degeneration in a rabbit meniscus defect model.

Acknowledgements

This work is supported by National High Technology Research and Development Program of China, 2012AA020502, National High Technology Research and Development Program of China, 2015AA020303, National Natural Science Foundation of China (81472092), National Natural Science Foundation of China (Key Program) (21134004).

References

[1] A.L. McNulty, F. Guilak, Mechanobiology of the meniscus, J. Biomech. 48

(2015) 1469-1478.

[2] A.J. Fox, F. Wanivenhaus, A.J. Burge, R.F. Warren, S.A. Rodeo, The human meniscus: a review of anatomy, function, injury, and advances in treatment, Clin. Anat. 28 (2015) 269-287.

[3] S.N. Edd, N.J. Giori, T.P. Andriacchi, The role of inflammation in the initiation of osteoarthritis after meniscal damage, J. Biomech. 48 (2015) 1420-1426.

[4] N.A. Smith, M.L. Costa, T. Spalding, Meniscal allograft transplantation: rationale for treatment, Bone Jt. J. 97-B (2015) 590-594.

[5] M. Cucchiarini, A.L. McNulty, R.L. Mauck, L.A. Setton, F. Guilak, H. Madry, Advances in combining gene therapy with cell and tissue engineering-based approaches to enhance healing of the meniscus, Osteoarthr. Cartil./OARS Osteoarthr. Res. Soc. 24 (8) (2016 Aug) 1330-1339, http://dx.doi.org/10.1016/ j.joca.2016.03.018.

[6] G. Narayanan, V.N. Vernekar, E.L. Kuyinu, C.T. Laurencin, Poly (lactic acid)-based biomaterials for orthopedic regenerative engineering, Adv. Drug Deliv. Rev. (2016), http://dx.doi.org/10.1016Zj.addr.2016.04.015 pii: S0169-409X(16)30125-9.

[7] C.H. Lee, S.A. Rodeo, L.A. Fortier, C. Lu, C. Erisken, J.J. Mao, Protein-releasing polymeric scaffolds induce fibrochondrocytic differentiation of endogenous cells for knee meniscus regeneration in sheep, Sci. Transl. Med. 6 (2014), 266ra171.

[8] A.R. Merriam, J.M. Patel, B.M. Culp, C.J. Gatt Jr., M.G. Dunn, Successful total meniscus reconstruction using a novel fiber-reinforced Scaffold: a 16- and 32-week study in an ovine model, Am. J. Sports Med. 43 (2015) 2528-2537.

[9] J.M. Patel, A.R. Merriam, B.M. Culp, C.J. Gatt Jr., M.G. Dunn, One-year outcomes of total meniscus reconstruction using a novel fiber-reinforced scaffold in an ovine model, Am. J. Sports Med. 44 (2016) 898-907.

[10] J.J. Rongen, T.G. van Tienen, B. van Bochove, D.W. Grijpma, P. Buma, Biomaterials in search of a meniscus substitute, Biomaterials 35 (2014) 3527-3540.

[11] R.J. Warth, W.G. Rodkey, Resorbable collagen scaffolds for the treatment of meniscus defects: a systematic review, Arthrosc. J. Arthrosc. Relat. Surg. Off. Publ. Arthrosc. Assoc. N. Am. Int. Arthrosc. Assoc. 31 (2015) 927-941.

[12] D.D. Lin, N.E. Picardo, A. Adesida, W.S. Khan, Clinical studies using biological and synthetic materials for meniscus replacement, Curr. Stem Cell Res. Ther.

(2016), http://dx.doi.org/10.2174/1574888X11666160429123110. PMID: 27133081.

[13] B.B. Mandal, S.H. Park, E.S. Gil, D.L. Kaplan, Multilayered silk scaffolds for

meniscus tissue engineering, Biomaterials 32 (2011) 639-651.

[14] B.B. Mandal, S.H. Park, E.S. Gil, D.L. Kaplan, Stem cell-based meniscus tissue engineering, Tissue Eng. Part A 17 (2011) 2749-2761.

[15] G.K. Tan, D.L. Dinnes, L.N. Butler, J.J. Cooper-White, Interactions between meniscal cells and a self assembled biomimetic surface composed of hyal-uronic acid, chitosan and meniscal extracellular matrix molecules, Biomaterials 31 (2010) 6104-6118.

[16] G.K. Tan, J.J. Cooper-White, Interactions of meniscal cells with extracellular matrix molecules: towards the generation of tissue engineered menisci, Cell Adhes. Migr. 5 (2011) 220-226.

[17] M.M. Pillai, V. Elakkiya, J. Gopinathan, C. Sabarinath, S. Shanthakumari, K.S. Sahanand, et al., A combination ofbiomolecules enhances expression ofE-cadherin and peroxisome proliferator-activated receptor gene leading to increased cell proliferation in primary human meniscal cells: an in vitro study, Cytotechnology 68 (5) (2016 Oct) 1747-1761, http://dx.doi.org/10.1007/ s10616-015-9926-1.

[18] Y.C. Chen, R.N. Chen, H.J. Jhan, D.Z. Liu, H.O. Ho, Y. Mao, et al., Development and characterization of acellular extracellular matrix scaffolds from porcine menisci for use in cartilage tissue engineering, Tissue Eng. Part C Methods 21 (2015) 971-986.

[19] G.K. Tan, D.L. Dinnes, P.T. Myers, J.J. Cooper-White, Effects of biomimetic surfaces and oxygen tension on redifferentiation of passaged human fibro-chondrocytes in 2D and 3D cultures, Biomaterials 32 (2011) 5600-5614.

[20] J. Baek, X. Chen, S. Sovani, S. Jin, S.P. Grogan, D.D. D'Lima, Meniscus tissue engineering using a novel combination of electrospun scaffolds and human meniscus cells embedded within an extracellular matrix hydrogel, J. Orthop. Res. Off. Publ. Orthop. Res. Soc. 33 (2015) 572-583.

[21] Z.Z. Zhang, D. Jiang, S.J. Wang, Y.S. Qi, J.Y. Zhang, J.K. Yu, Potential of centrifugal seeding method in improving cells distribution and proliferation on demineralized cancellous bone scaffolds for tissue-engineered meniscus, ACS Appl. Mater. Interfaces 7 (2015) 15294-15302.

[22] N. Ozeki, T. Muneta, S. Matsuta, H. Koga, Y. Nakagawa, M. Mizuno, et al.,

Synovial mesenchymal stem cells promote meniscus regeneration augmented by an autologous Achilles tendon graft in a rat partial meniscus defect model, Stem Cells 33 (2015) 1927-1938.

[23] N. Ozeki, T. Muneta, H. Koga, H. Katagiri, K. Otabe, M. Okuno, et al., Transplantation of Achilles tendon treated with bone morphogenetic protein 7 promotes meniscus regeneration in a rat model of massive meniscal defect, Arthritis Rheum. 65 (2013) 2876-2886.

[24] K. Ishida, R. Kuroda, M. Miwa, Y. Tabata, A. Hokugo, T. Kawamoto, et al., The regenerative effects of platelet-rich plasma on meniscal cells in vitro and its in vivo application with biodegradable gelatin hydrogel, Tissue Eng. 13 (2007) 1103-1112.

[25] J.A. van der Sluijs, R.G. Geesink, A.J. van der Linden, S.K. Bulstra, R. Kuyer, J. Drukker, The reliability of the Mankin score for osteoarthritis, J. Orthop. Res. Off. Publ. Orthop. Res. Soc. 10 (1992) 58-61.

[26] C.G. Peterfy, A. Guermazi, S. Zaim, P.F. Tirman, Y. Miaux, D. White, et al., Whole-organ magnetic resonance imaging score (WORMS) of the knee in osteoarthritis, Osteoarthr. Cartil./OARS Osteoarthr. Res. Soc. 12 (2004) 177-190.

[27] A. Gloria, R. De Santis, L. Ambrosio, Polymer-based composite scaffolds for tissue engineering, J. Appl. Biomate. Biomech. JABB 8 (2010) 57-67.

[28] J.H. de Groot, Polyurethane scaffolds for meniscal tissue regeneration, Med. Device Technol. 16 (2005) 18-20.

[29] R.L. Mauck, J.A. Burdick, From repair to regeneration: biomaterials to repro-gram the meniscus wound microenvironment, Ann. Biomed. Eng. 43 (2015) 529-542.

[30] K.F. Schuttler, T. Efe, Tissue regeneration in orthopedic surgery - do we need cells? Technol. Health Care Off. J. Eur. Soc. Eng. Med. 23 (2015) 403-409.

[31] T. Trzeciak, M. Richter, W. Suchorska, E. Augustyniak, M. Lach, M. Kaczmarek, et al., Application of cell and biomaterial-based tissue engineering methods in the treatment of cartilage, menisci and ligament injuries, Int. Orthop. 40 (2016) 615-624.