Introduction
The meniscus of the knee joint serves a variety of critical functions that include shock absorption, cartilage protection, and joint stability [
1‐
3]. The capacity to perform these functions is by virtue of its extracellular matrix (ECM) composition and assembly, which is accomplished entirely by meniscus fibrochondrocytes [
4,
5]. The ECM consists predominantly of type I collagen throughout, type II collagen in the inner meniscus, and proteoglycans, of which aggrecan is predominant [
6‐
8]. Unfortunately, the reparative capacity of the meniscus is hindered by limited vascularization [
9]. In human meniscus, the capillary plexus supplies only the outer one third [
10] whereas the inner two thirds are avascular; if left untreated, defects in this portion do not heal and may lead to further joint degeneration [
11,
12]. Current treatment options include partial and total meniscectomies, depending on the extent of meniscal injury [
13]. However, these procedures are major risk factors for the early development of osteoarthritis (OA) [
13‐
16]. Cell-based regenerative medicine and tissue engineering have been advocated options to produce functional substitutes to aid repair or replace damaged tissue [
17‐
28]. However, current protocols suffer from several drawbacks that include insufficient numbers of differentiated meniscus cells and loss of ECM-forming phenotype of
in vitro-propagated meniscus cells. Therefore, alternative cell sources or cell-based strategies are of interest in meniscus tissue engineering.
Adult-derived mesenchymal stromal stem cells (MSCs) are of particular interest in meniscus tissue engineering because of their capacity to undergo differentiation into a variety of mesenchymal lineages, including cartilage and bone [
17,
29,
30]. Additionally, MSCs secrete soluble trophic factors that are capable of promoting cell proliferation and differentiation and suppressing local immune system through both autocrine and paracrine mechanisms [
31‐
35]. However, specific factors known to induce MSCs toward the meniscus cell phenotype are unknown. Co-culture of MSCs with primary meniscus cells is a stratagem that both provides inductive factors and mitigates the need for meniscus cell expansion and associated dedifferentiation of meniscus cells during
in vitro culture expansion [
32,
34‐
38]. Recent work in our laboratory demonstrated that supplementation of human primary meniscus cells with human bone marrow MSCs results in increased meniscus matrix-forming phenotype [
39]. Similar work by Cui and colleagues [
40] additionally demonstrated that co-cultures of primary human meniscus cells with bone marrow MSCs result in suppression of hypertrophic differentiation of bone marrow MSCs. However, both studies used meniscus cells taken from the whole meniscus and it is not clear whether the cells from the outer and inner menisci will interact similarly or distinctly with bone marrow MSCs. Previous studies have shown that the meniscus cells from the outer and inner regions of the meniscus have distinct ECM gene profiles as well as different multi-lineage differentiation characteristics [
41‐
43]. Moreover, inner meniscus cells display a more chondrogenic phenotype relative to outer meniscus cells on the basis of inherently higher collagen II and aggrecan content [
42,
43]. Furthermore, outer meniscus cells displayed a more chondrogenic response to hypoxic stimuli by enhanced expression of chondrogenic genes, collagen II and Sry-related HMG box 9 (sox9), than cells isolated from the inner and avascular regions of the meniscus [
44]. In this study, we explored the concepts that (a) cells of outer and inner menisci have the capacity to enhance matrix formation of meniscus cells after co-culture with MSCs and (b) cells from the outer and self-healing regions of the meniscus may serve as a clinical source of cells in co-culture tissue engineering strategies with MSCs to recapitulate the functional matrix-forming phenotype of the inner meniscus, which has a limited reparative capacity.
Materials and methods
Collection of bone marrow specimens and culture of bone marrow stem cells
Bone marrow aspirates were obtained from surgically discarded material after approval and a waiver of informed consent of the local ethical committee of the University of Alberta (Edmonton, AB, Canada) during routine orthopedic procedures from the iliac crest of two female donors (ages of 23 and 59 years). The number of nucleated cells in the aspirates was determined by crystal violet nuclei staining and counting on a hemocytometer. Thereafter, 15 million nucleated cells per 150-cm
2 tissue culture flask were seeded. Culture medium was alpha minimum essential medium supplemented with 10% heat-inactivated fetal bovine serum, 1 mM sodium pyruvate, 100 mM HEPES buffer, 1 mM sodium pyruvate, 100 U/mL penicillin, 100 μg/mL streptomycin, and 0.29 mg/mL L-glutamine (all from Invitrogen, Mississauga, ON, Canada) and 5 ng/mL of basic fibroblast growth factor 2 (FGF-2) (Humanzyme-Medicorp Inc., Montreal, QC, Canada). Nucleated cells were allowed to adhere and grow for 7 days before the first media change under normal oxygen tension (21% O
2 and 95% air) at 37°C in a humidified incubator with 5% CO
2. Thereafter, media change was implemented twice per week until 70% to 80% cell confluence was attained. These adherent bone marrow MSCs were detached by using trypsin-ethylenediaminetetraacetic acid (trypsin-EDTA) (0.05% wt/vol) and expanded until passage 2 prior to experimental use. We included FGF-2 throughout MSC cultivation as it maintains human MSCs in an immature state for
in vitro expansion [
45].
Isolation of human meniscus cells
Menisci were obtained from surgically discarded material after approval and a waiver of informed consent of the local ethical committee of the University of Alberta. Both lateral and medial menisci were harvested from the knee joint of four female donors (age of 55 to 71 years and mean age of 62 ± 6.8 years) undergoing total knee arthroplasty. Tissue from the outer and inner regions represented pieces taken from one third and two thirds of the radial distance, respectively. Meniscus cells were released via treatment with type II collagenase (0.2% wt/vol; Worthington, Lakewood, NJ, USA) after 16-hour digestion of tissue at 37°C in a standard medium-high glucose Dulbecco's modified Eagle's medium containing 4.5 mg/mL D-Glucose (DMEM-HG), 0.1 mM non-essential amino acids, 1 mM sodium pyruvate, 100 mM HEPES buffer, 1 mM sodium pyruvate, 100 U/mL penicillin, 100 μg/mL streptomycin, and 0.29 mg/mL L-glutamine (all from Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen). Meniscus cells from the outer meniscus region were labeled as MCO, and those from the inner portion of the meniscus were designated as MCI. The cell suspension obtained after digestion was passed through a 70-μm nylon-mesh filter. Isolated cells were plated at 104 cells/cm2 and cultured in standard medium for 24 hours under normal oxygen tension (21% O2 and 95% air) at 37°C in a humidified incubator with 5% CO2 before experimental use. Non-adherent cells were aspirated off while adherent cells were detached with trypsin-EDTA (0.05% wt/vol) (Invitrogen).
Chondrogenic differentiation
Mono-culture in pellets of MSCs, MCO, and MCI were formed in 250 μL of serum-free chondrogenic culture medium consisting of standard medium supplemented with 0.1 mM ascorbic acid 2-phosphate, 10
−5 M dexamethasone, 1x ITS+1 premix (Sigma-Aldrich, Oakville, ON, Canada), and 10 ng/mL transforming growth factor-beta 1 (TGF-β1) (Humanzyme-Medicorp Inc.) as described previously [
35]. A total of 2.5 × 10
5 cells were spun in 1.5 mL of sterile conical polypropylene microfuge tubes (Enzymax LLC, Lexington, KY, USA) at 1,500 revolutions per minute for 3 minutes to form spherical cell pellets. Co-cultures in pellet form consisting of MCO/MSCs or MCI/MSCs were formulated by mixing the two cell types at a 1:3 ratio, respectively. Preliminary studies with a variety of ratios had shown a ratio of 1:3 to reproducibly result in enhanced matrix formation, but identification of the optimal cell-cell ratio remains to be determined [
46]. After 3 weeks of culture with media changes three times per week, pellets were processed biochemically for glycosaminoglycan (GAG) and DNA contents, histologically, and immuno-histochemically and by quantitative reverse transcription-polymerase chain reaction for gene expression analysis. For each experiment (four in total), there were two experimental groups (co-culture of MCO/MSCs and co-culture of MCI/MSCs) and three control groups (mono-culture of MSCs, mono-culture of MCO, and mono-culture of MCI). At least two or three replicate pellets per control or experimental group were assessed for each experiment.
Biochemical analysis
All
in vitro cultivated pellets were rinsed in phosphate-buffered saline (Invitrogen) and digested in proteinase K (1 mg/mL in 50 mM Tris with 1 mM EDTA, 1 mM iodoacetamide, and 10 mg/mL pepstatin A; all from Sigma-Aldrich) for 16 hours at 56°C. The sulphated GAG content was measured by 1,9 dimethylmethylene blue binding (Sigma-Aldrich) by using chondroitin sulphate (Sigma-Aldrich) as standard [
47]. The DNA content was determined by using the CyQuant cell proliferation assay kit (Invitrogen) with supplied bacteriophage λ DNA as standard. Based on experimental GAG values of mono-culture of MSC, MCO, or MCI pellets, the expected GAG values were calculated as a linear function of the proportion (percentage) of MSCs and MCO or MSCs and MCI by using the following equations [
35,
39]:
The interaction index or measure of chondro-induction was then calculated as the ratio of measured GAG (that is, GAG
measured) in co-cultured pellets to expected GAG (that is, GAG
expected). If the value of the ratio is greater than 1, then chondro-induction, an enhanced chondrogenic matrix formation, is considered to have taken place [
35,
39].
Histology and immuno-histochemistry
Tissues generated from the pellet cultures were fixed in 4% phosphate-buffered formalin, processed into paraffin wax, sectioned at 5 μm, and stained with 1% Alcian blue and counterstained with 1% neutral red stain to reveal sulphated proteoglycan (GAG) matrix depositions. Other sections were probed with antibodies raised against collagen types I and II. Sections were digested with trypsin and then incubated with antibodies against collagen I (MAB3391) from Millipore (Temecula, CA, USA) or collagen II (II-II6B3) from the Developmental Studies Hybridoma Bank at the University of Iowa (Iowa City, IA, USA). Immuno-localized antigens were visualized with goat anti-mouse IgG biotinylated secondary antibody (Dako Canada Inc., Mississauga, ON, Canada) and a streptavidin-horseradish peroxidase labeling kit with 3,3'-diaminobenzidine (Dako Canada Inc.). Images were captured by using an Omano OM159T biological trinocular microscope (Microscope Store, Roanoke, VA, USA) fitted with an Optixcam summit series 5MP digital camera and Optixcam software and assembled in Adobe Photoshop (Adobe Systems Inc., Mountain View, CA, USA).
Gene expression analysis
Total RNA was extracted from pellets by using Tri-Reagent (Sigma-Aldrich) after grinding with Molecular Grinding Resin (Geno Technology Inc., St. Louis, MO, USA) in combination with the use of an Aurum Total RNA Fatty and Fibrous Tissue Kit (Bio-Rad, Mississauga, ON, Canada) and after removal of contaminating genomic DNA from the pellets by DNase treatment. To minimize changes in gene expression during cell pellet harvest, cell pellets were immediately (less than 1 minute) transferred into Tri-Reagent. Total RNA (100 ng) in a 40-μL reaction was reverse-transcribed to cDNA by using GoScript reverse transcriptase primed with oligo(dT)
15 primer (Fisher Scientific, Whitby, ON, Canada). Quantitative real-time polymerase chain reaction was performed in DNA Engine Opticon II Continuous Fluorescence Detection System (Bio-Rad) by using hot start Taq and SYBR Green detection (Eurogentec North America Inc., San Diego, CA, USA). Primer sequences for collagen I (
COL1A2), collagen II (
COL2A1), collagen × (
COL10A1), aggrecan, versican, SOX9, and β-actin (Table
1) were taken from previously published work or were custom designed by using Primer Express software (Applied Biosystems, Foster City, CA, USA) [
44,
48,
49]. All primers were obtained from Invitrogen. Gene (mRNA) expression levels for each primer set were normalized to the expression level of β-actin by the 2-
Δct method [
50].
Table 1
Primer sequences used in quantitative real-time polymerase chain reaction
Beta-actin
|
5'-AAGCCACCCCACTTCTCTCTAA-3'
5'-AATGCTATCACCTCCCCTGTGT-3'
|
Forward
Reverse
| |
Aggrecan |
5'-AGGGCGAGTGGAATGATGTT-3'
5'-GGTGGCTGTGCCCTTTTTAC-3'
|
Forward
Reverse
| |
Collagen I (COL1A2) |
5'-TTGCCCAAAGTTGTCCTCTTCT-3'
5'-AGCTTCTGTGGAACCATGGAA-3'
|
Forward
Reverse
| |
Collagen II
(COL2A1) |
5'-CTGCAAAATAAAATCTCGGTGTTCT-3'
5'-GGGCATTTGACTCACACCAGT-3'
|
Forward
Reverse
| |
Collagen × (COL10A1) |
5'-CAAGGCACCATCTCCAGGAA-3'
5'-AAAGGGTATTTGTGGCAGCATATT-3'
|
Forward
Reverse
| |
IHh | 5'-CCTTGTCAGCCGTGAGGCCG-3'
5'-GCTGCCGGCTCCGTGTGATT-3'
|
Forward
Reverse
| NM_00218.3 |
MMP-13 | 5'-TGGAAGGATGCCTTTTTTTCTC-3'
5'-CACCCTCCCCAAGTATCAATAGG-3'
|
Forward
Reverse
| NM_002427 |
SOX9 |
5'-CTTTGGTTTGTGTTCGTGTTTTG-3'
5'-AGAGAAAGAAAAAGGGAAAGGTAAGTTT-3'
|
Forward
Reverse
| |
Versican
| 5'-TGGAATGATGTTCCCTGCAA-3' 5'-AAGGTCTTGGCATTTTCTACAACAG-3' |
Forward
Reverse
| |
Statistical analysis
Data are presented as mean ± standard error of mean of measurements. All statistical analyses were performed by using SPSS version 18 (PASW Statistics 18; SPSS Inc., Chicago, IL, USA) unless stated otherwise. Experimental and control groups were compared with one-way analysis of variance with Tukey's multiple comparison post hoc tests. Statistical differences between chondro-induction indices of MCO and MSC experimental groups and/or MCI and MSC experimental groups were assessed by paired two-tailed Student t test distribution. All statistical differences were considered to be significant with a P value of less than 0.05.
Discussion
Our previous study demonstrated that direct cell-to-cell co-culture of primary human meniscus cells from all regions of the meniscus with bone marrow-derived MSCs resulted in increased matrix formation [
39]. However, we did not determine if the cells from the outer-vascular and inner-avascular regions of the meniscus interacted with MSCs similarly or otherwise. The response of cells of outer and inner menisci to culture conditions and growth factors is different, and the expression of
COL2A1 and aggrecan is higher in monolayer cultures of porcine inner meniscus cells than in outer meniscus cells [
42]. We showed that, in aggregate cultures of human inner and outer meniscus cells in the presence of TGF-β1, hypoxia (5%) increased the expression of
Sox9 in outer meniscus cells but not in inner meniscus cells [
44].
In this study, we isolated cells from the inner and outer regions of the meniscus and co-cultured the cells separately with bone marrow-derived MSCs by using the scaffold-free three-dimensional pellet model of tissue formation. Co-culture of primary inner or outer meniscus cells with MSCs resulted in enhanced matrix formation since the calculated interaction indices were greater than 1. This finding was consistent with the data from our recent work on co-culture of whole meniscus cells with MSCs [
35]. Furthermore, our results corroborated the findings of similar studies that involved co-culture of chondrocytes or intervertebral discs with MSCs and that resulted in enhanced matrix synthesis [
32,
35]. The mechanism underlying the enhancement of matrix synthesis was not explored in this study. However, it was interesting to find that the cellularities of the pellets (as judged by the DNA content of our pellets) were not statistically different among experimental and control groups. In a previous study involving co-culture of primary chondrocytes and MSCs, enhanced matrix synthesis was attributed to a synergistic increase in chondrocyte proliferation and induced chondrogenesis of MSCs [
35]. However, the lack of difference between the DNA contents of the pellets in this study suggests that the underlying mechanism is most likely independent of changes in cell proliferation. Moreover, our DNA data are consistent with the findings of Furumatsu and colleagues [
43], who showed that the proliferation rates of both inner and outer human meniscus cells were the same in the presence of TGF-β1 (10 ng/mL) as used in this study. Other mechanistic possibilities may involve a synergistic cross-talk mediated by soluble factors from both meniscus cells and MSCs [
31,
52].
Histological and biochemical analyses of the pellets in this study revealed that the cells from the outer and inner regions of the meniscus interacted almost indistinguishably with MSCs. Proteoglycan staining via safranin O and Alcian blue stains were similar, and the GAG content normalized to DNA content were also similar (Figure
1). This finding was surprising as Furumatsu and colleagues [
43] recently reported that human inner meniscus cells have a higher GAG content than outer meniscus cells, even after cultivation in chondrogenic factors. The reason for this is not clear but may be associated with differences in the severity of OA in the specimens used in this study. Furumatsu and colleagues used specimens from medial-affected OA joints (that is, specimens from joints with sufficient articular cartilage of lateral femoral condyle and undegenerated lateral meniscus), whereas this study used both lateral and medial menisci from joints with macroscopically diverse, degenerated articular cartilage of lateral and medial femoral condyles. This possibility is further supported by the findings of Sun and colleagues [
53], who demonstrated that altered gene expression profiles exist between OA and normal menisci.
Collagen I (
COL1A2) is the predominant collagen type and ECM molecule in meniscus [
5,
6]. Co-culture of inner or outer meniscus cells with MSCs resulted in the upregulation of
COL1A2 transcription, as we previously reported [
39]. Similarly to previous findings, at the protein level, collagen I deposition was qualitatively the same in all pellets. Furthermore, there was no statistical difference between its transcriptional expressions in the two co-cultures. Similarly to our previous findings with co-cultures of whole meniscus cells and MSCs [
39], the levels of expression of
COL1A2 in co-cultures of inner or outer meniscus cells with MSCs appear to be orchestrated by the MSC component since the level of
COL1A2 expression in mono-cultures of MSCs was not surpassed. The underlying mechanism for this regulatory role of MSCs is not clear at this point but may be related to trophic activities of MSCs [
31,
54]. Nonetheless, our data were consistent with our previous findings that
COL1A2 expression is higher in human outer meniscus cells than in inner meniscus cells chondrogenic culture [
44].
We investigated the gene expression of aggrecan, which is the predominant proteoglycan of the meniscus [
55], and versican, another proteoglycan expressed in the meniscus [
56], to confirm our histological and biochemical findings. The relative expression levels of aggrecan and versican were not statistically different between co-cultures of inner or outer meniscus cells with MSCs, nor was there any difference between their expression in mono-cultures of inner and outer human meniscus cells (Figure
3A). These findings seem contrary to previous reports that the mRNA expression of aggrecan is higher in inner meniscus cells than in outer meniscus cells. Further contrasts to previous findings were noted in our results on collagen II (
COL2A1) expression. Our data demonstrated that the expression of
COL2A1 was least in mono-cultures of inner meniscus cells and highest in co-cultures of outer or inner meniscus cells with MSCs relative to mono-cultures of MSCs or outer meniscus cells. However, previous findings by us [
44] and Furumatsu and colleagues [
43] demonstrated that
COL2A1 and
Sox9 (a transcription factor known to promote
COL2A1) expression in inner meniscus cells was higher relative to their expression in outer meniscus cells, even in the presence of chondrogenic factors. A plausible reason for this disparity between our findings here and those of previous studies is donor-donor variability (age, gender, and so on) or, as alluded to earlier, the association with possible differences in the severity of OA among study specimens [
53,
57]. Additionally, it is probable that differences in the multi-potential characteristics of outer and inner meniscus cells play a contributing factor. Mauck and colleagues [
41] showed that bovine meniscus cells from the inner and outer regions of the meniscus exhibit different multi-lineage differentiation capabilities. Cells from the outer region were more plastic than inner meniscus cells, differentiating along all three mesenchymal lineages (adipogenesis, chondrogenesis, and osteogenesis). Mauck and colleagues also showed that, when outer or inner meniscus cells were treated with chondrogenic factors (as in this study), inner and outer cell populations expressed similar levels of aggrecan. In contrast, in the absence of chondrogenic factors, the expression of aggrecan in inner meniscus is significantly higher than its expression in outer meniscus [
42]. Despite the dissimilarities in our aggrecan and
COL2A1 data relative to previous reports, it was interesting that the expression of
Sox9 and
COL2A1 correlated linearly in mono-cultures of MSCs and MCO and in co-cultures of MCO/MSCs but inversely in mono-cultures of MCI (Figure
3C,E). The expressions of
Sox9 and
COL2A1 often are not correlated [
57‐
59], but our findings suggest that, although
Sox9 is necessary for
COL2A1 expression, other factors may facilitate
COL2A1 transcription in inner meniscus. These factors may be particularly active in specimens derived from OA joints and perhaps reflect the similarities of the inner meniscus and articular cartilage [
30,
53,
57]. The deposition and distribution of collagen II were notably enhanced in co-cultures of inner or outer meniscus cells with MSCs and in mono-cultures of MSCs. This correlated with our previous findings involving whole meniscus cells and MSCs [
39].
Further gene expression analysis showed statistically significant differences between the expressions of hypertrophic genes,
COL10A1,
IHh, and
MMP-13, in co-cultures of MSCs with outer or inner meniscus cells.
COL10A1,
IHh, and
MMP-13 are predominantly expressed by hypertrophic chondrocytes during endochondral ossification [
60].
COL10A1 is also a hypertrophic marker for OA cartilage and OA meniscus [
61,
62]. Expression of both
IHh and
MMP-13 modulates hypertrophic differentiation of chondrocytes during endochondral ossification.
IHh is produced by hypertrophic chondrocytes and, in turn, stimulates secretion of parathyroid hormone-related protein (PTHrP), which acts directly on the chondrocytes to slow the pace of hypertrophic chondrocyte differentiation during endochondral ossification [
63]. Increased MMP-13 expression and activity have been implicated in extensive reorganization of ECM during endochondral ossification [
62]. Our results showed that outer meniscus cells had a greater potency to suppress hypertrophic differentiation of MSCs than inner meniscus cells. This finding was particularly surprising since inner meniscus cells have been shown to display a reduced osteogenic plasticity relative to outer meniscus cells [
41]. The reason for this difference is unclear, but it is plausible that outer meniscus cells have a greater capacity to secrete PTHrP than inner meniscus cells. PTHrP was shown to suppress the mRNA expression of
COL10A1 [
64]. This aspect of the study merits further investigation.
Co-culture of outer or inner meniscus cells with MSCs exhibit similar capacity to synthesis cartilaginous matrix with increased collagen II and proteoglycan content relative to mono-cultures of MSCs, outer or inner meniscus cells. This finding suggests that primary meniscus cells from the outer and reparative region of the meniscus, after supplementation with MSCs, can in principle be used to generate functional grafts or substitutes for the reconstruction of irreparable or degenerate inner meniscus. This possibility is supported by the findings of Kobayashi and colleagues [
65], who showed that explants of the outer meniscus integrated well after transplantation to the inner-avascular region of the meniscus. Additionally, our finding regarding the enhanced potency of outer meniscus cells (compared with inner meniscus cells) to suppress hypertrophic differentiation of MSCs makes it an attractive cell source for the reconstruction of the inner meniscus. Furthermore, supplementation of primary meniscus cells with MSCs will mitigate the need to expand meniscus cells for cell-based meniscus tissue engineering purposes since expanded meniscus cells lose their functional matrix-forming phenotype during
in vitro expansion.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
DJJS helped to perform experiments and data acquisition and shared responsibility for tissue procurement. AM-S helped to perform experiments and data acquisition. NMJ shared responsibility for tissue procurement, data analysis, and manuscript writing. ABA conceived and designed the study, performed experiments and data acquisition and analysis, wrote the manuscript, and supervised the entire study. All authors read and approved the final manuscript.