Evaluation of Culture Conditions for In vitro Meniscus Repair Model Systems using Bone Marrow-Derived Mesenchymal Stem Cells

Abstract

Purpose:

Meniscal injury and loss of meniscus tissue lead to osteoarthritis development. Therefore, novel biologic strategies are needed to enhance meniscus tissue repair. The purpose of this study was to identify a favorable culture medium for both bone marrow-derived mesenchymal stem cells (MSCs) and meniscal tissue, and to establish a novel meniscus tissue defect model that could be utilized for in vitro screening of biologics to promote meniscus repair.

Materials and Methods:

In parallel, we analyzed the biochemical properties of MSC-seeded meniscus-derived matrix (MDM) scaffolds and meniscus repair model explants cultured in different combinations of serum, dexamethasone (Dex), and TGF-β. Next, we combined meniscus tissue and MSC-seeded MDM scaffolds into a novel meniscus tissue defect model to evaluate effects of chondrogenic and meniscal media on the tissue biochemical properties and repair strength.

Results:

Serum-free medium containing TGF-β and Dex was the most promising formulation for experiments with MSC-seeded scaffolds, whereas serum-containing medium was the most effective for meniscus tissue composition and integrative repair. When meniscus tissue and MSC-seeded MDM scaffolds were combined into a defect model, the chondrogenic medium (serum-free with TGF-β and Dex) enhanced the production of proteoglycans and promoted integrative repair of meniscus tissue. As well, cross-linked scaffolds improved repair over the MDM slurry.

Conclusions:

The meniscal tissue defect model established in this paper can be used to perform in vitro screening to identify and optimize biological treatments to enhance meniscus tissue repair prior to conducting preclinical animal studies.

Introduction

The menisci are critical knee structures that are essential for joint function and homeostasis by providing joint stability, congruity, load bearing capabilities, shock absorption, lubrication, and nutrition^1–5. Meniscus injuries result in pain and disability, and also lead to the development of osteoarthritis (OA) in approximately two-thirds of meniscus-injured patients within 5–15 years^6–13. Therefore, current treatment approaches focus on the preservation of meniscus tissue and repair of meniscus tears whenever possible, as well as replacing the lost or damaged tissue in an attempt to restore tissue function and biomechanics¹⁴. In vitro models show that meniscus repair can be successful in explanted tissue from either the avascular inner zone or the vascularized outer zone of the meniscus¹⁵. This suggests that tissue from both regions has the capacity to heal if provided with the appropriate environment. However, in patients, meniscus tissue healing is more favorable in the outer vascularized region and generally not recommended for tears in the inner avascular region, which are commonly treated with partial meniscectomy. Despite a focus on more conservative treatments, partial meniscectomy remains the most common orthopaedic procedure¹⁶. Thus, new biologic strategies are needed to promote improved meniscus tissue regeneration and repair.

Tissue engineering involves the use of cells, scaffolds, and biologic factors to repair or regenerate tissues. In particular for meniscus tissue engineering, primary meniscal cells^17–24, adipose-derived stem cells^25–27, synovial-derived stem cells^28–30, meniscus-derived stem cells³¹, bone-marrow derived mesenchymal stem cells (MSCs)^{19, 22, 23, 26, 32–35}, and MSCs co-cultured with meniscus cells^36–39 have been investigated^40–43. Stem cells are an attractive choice for meniscus tissue engineering because of their expansion capabilities, potential for fibrochondrogenic differentiation, ability to secrete repair-promoting growth factors, ability to migrate to sites of injury, and their role in immune system modulation^{31, 44–46}. Furthermore, MSCs have been studied for meniscus tissue engineering applications and tissue repair^{19, 22, 23, 26, 32–35, 37, 40–42, 47}. However, the optimal culture medium for MSCs co-cultured with meniscus tissue has not been established.

Biomaterial scaffolds are promising tools to augment meniscus repair procedures. In particular, scaffolds studied for meniscus tissue engineering have been composed of polymers^{19, 24, 28, 32, 34, 48–51}, hydrogels^{25, 30, 43}, extracellular matrix (ECM) components^{20, 26, 27, 52}, tissue-derived materials^{18, 21, 29, 36, 43, 53, 54}, and composites of the above materials^{22, 23, 32, 33, 35, 40, 41}. Several preclinical studies have shown improved repair with MSC seeding of scaffolds prior to implantation into a meniscus defect^{22, 26, 34, 35, 41}. Recently, we have shown that meniscus-derived matrix (MDM) scaffolds promote cellular infiltration of both endogenous meniscus cells and exogenous MSCs in an in vitro meniscus repair model⁵⁴. In this model in which MDM scaffold cores are surrounded by meniscus tissue, the MDM scaffolds alone increased the integrative shear strength of repair when compared to control meniscus tissue. The seeding of exogenous MSCs onto the MDM scaffolds in this repair model system caused a trend towards higher shear strength of repair than control meniscus tissue⁵⁴. The addition of MSCs to the MDM scaffolds presents an opportunity to deliver therapeutic agents to the meniscus defect but it may be necessary to use exogenous growth factors that are not present in the MDM scaffolds to further promote fibrochondrogenic differentiation of MSCs and enhance repair.

A variety of biologic factors have previously been used to promote and enhance meniscus tissue formation⁵⁵. In particular for MSCs, the commonly used medium to promote fibrochondrogenic differentiation is composed of insulin/transferrin/selenium (ITS), dexamethasone (Dex), transforming growth factor (TGF)-β, proline, and ascorbic acid^{19, 22, 23, 31–35, 37–39, 47}. On the other hand, meniscus growth medium is frequently composed of serum, proline, and ascorbic acid^{15, 54, 56–64}. However, the optimal differentiation and growth conditions for co-culture of reparative MSCs and meniscus tissue have not been established and are needed to allow in vitro evaluation of biological and mechanical factors influencing integrative meniscus tissue repair. Therefore, the goals of the present study were 1) to identify a favorable culture medium for both MSC-seeded MDM scaffolds and meniscal tissue explants, and 2) to establish a novel meniscus tissue defect model for use in determining the effects of chondrogenic and meniscal media on the biochemical properties of the tissue and integrative repair. In parallel, we first analyzed the biochemical properties of MSC-seeded MDM scaffolds and the biochemical and mechanical properties of meniscus repair model explants cultured in different combinations of serum, Dex, and TGF-β. Subsequently, we combined the meniscal tissue and MSC-seeded MDM scaffolds into a novel meniscus tissue defect model to evaluate the effects of chondrogenic and meniscal media on the tissue biochemical properties and the shear strength of repair.

Materials and Methods

Experiment 1: Medium Optimization for MSCs Seeded in MDM Scaffolds and Meniscal Tissue

Media Preparation

The base medium for all samples was composed of high glucose Dulbecco’s Modified Eagle Medium (DMEM-HG; Sigma-Aldrich, St. Louis, MO), 1% antibiotic-antimycotic (PSF; Sigma-Aldrich), 1% non-essential amino acid solution (Sigma-Aldrich), 1% HEPES buffer solution (Sigma-Aldrich), 40 μg/mL proline (Sigma-Aldrich), and 50 μg/mL L-ascorbic acid 2-phosphate (Sigma-Aldrich). The following supplements were added to the base medium to generate four different media for testing in experiment 1: 1) 10% HyClone characterized fetal bovine serum (FBS; AVH79983, GE Healthcare Bio-Sciences, Pittsburgh, PA) (meniscal medium); 2) 100nM Dex (Sigma-Aldrich) and 1% ITS+ premix (Corning, Corning, NY); 3) 100nM Dex, 1% ITS+ premix, and 10ng/mL recombinant human TGF-β3 (R&D Systems, Minneapolis, MN) (chondrogenic medium); and 4) 10% FBS, 100nM Dex, and 10ng/mL rhTGF-β3.

MSC Seeding of MDM Scaffolds

MDM scaffolds were generated from skeletally mature female porcine medial menisci that were minced, frozen overnight at −80°C, and then lyophilized (FreeZone 2.5 L, Labconco, Kansas City, MO) for 24 hours, as previously described⁵⁴. The meniscus tissue was then pulverized (5 min pre-cool, 5 cycles of 1 min at 5 Hz and 2 min cool) in a 6770-freezer mill (SPEX SamplePrep, Metuchen, NJ) and sieved to remove particles larger than 500 μm. The MDM powder was suspended in ultra-pure distilled water to an 8% by weight solution. The solution was then homogenized (PRO260, PRO Scientific Inc., Oxford, CT) at 30,000 rpm for 2 cycles of 2 min run, 2 min cool on ice^{54, 65} and transferred to a 5mL syringe. The homogenized solution was then injected into a delrin mold to generate scaffolds that were 4 mm in diameter and 2 mm thick. The scaffolds were frozen overnight at −80°C, lyophilized for 24 hours, and sterilized with ethylene oxide.

Human bone marrow-derived mesenchymal stem cells (MSCs) were isolated from excess bone marrow aspirate from de-identified adult bone marrow donors in a Duke University IRB-exempt waste tissue protocol as previously described^{54, 65}. Cells were pooled into a superlot (n=3 donors) and cultured in MSC expansion medium containing low glucose Dulbecco’s Modified Eagle Medium (DMEM-LG; Sigma-Aldrich), 10% FBS, 1% PSF, and 2 ng/mL recombinant human fibroblast growth factor (FGF; R&D Systems) at 37°C and 5% CO₂.

Passage 4 MSCs were resuspended in expansion medium to achieve 7.8 × 10⁶ cells/mL. MDM scaffolds were placed into individual wells of a 24-well polystyrene ultra-low attachment plate (Corning) and 16 μL of cell suspension was pipetted onto the top of each scaffold. The cell suspension was allowed to wick into the scaffold for 10 minutes, resulting in 1.25×10⁵ cells seeded per scaffold (Figure 1A). MSC expansion medium (1 mL per well) was carefully added to avoid disturbing scaffolds. Seeded scaffolds were pre-cultured for 7 days at 37°C and 5% CO₂ prior to being allocated into one of the four media formulations described above (n=6/group). Scaffolds were then cultured at 37°C and 5% CO₂ in the different media formulations for 28 days.

Figure 1.

Schematic of cell seeded scaffolds and in vitro meniscus repair model explants utilized in experiment 1. (A) Meniscus-derived matrix (MDM) scaffolds were seeded with bone marrow-derived mesenchymal stem cells (MSCs). (B) Meniscus tissue repair model explants were generated by harvesting 8 mm diameter meniscus explants from porcine mensci. Then a 3 mm diameter inner core was removed and re-inserted. The MSC-seeded MDM scaffolds and meniscus tissue repair model explants were cultured in four different experimental media: 1) serum (meniscal), 2) dexamethasone (Dex) + ITS, 3) Dex + TGF-β3 + ITS (chondrogenic), and 4) serum + Dex + TGF-β3.

Meniscus Tissue Repair Model Explants

Medial meniscus tissue was harvested from female porcine knee joints obtained from a local abattoir. Explants were harvested from the centerline of medial menisci to generate 8 mm diameter punches that were cut to 2 mm thickness using a cutting block. An inner 3 mm diameter core was removed from the explant and immediately returned to the defect to generate meniscus repair model explants containing a meniscus tissue core surrounded by a meniscus tissue ring (Figure 1B), as described previously^{15, 56–62}. The meniscus repair model explants (n=6/group) were incubated at 37°C and 5% CO₂ for 28 days in the four different media described above.

Outcome Measures

Cell Viability.

Mitochondrial dehydrogenase activity, which correlates to the number of metabolically active cells, was measured using the BioVision Quick Cell Proliferation Assay Kit. MSC-seeded scaffolds and meniscus repair model explants were incubated in fresh media containing 1% WST-1 reagent at the end of culture. Absorbance was measured at 450 nm with a reference wavelength of 620 nm⁶⁶.

Histology.

Histological analyses were performed to assess the structure and composition of the MSC-seeded MDM scaffolds and meniscus tissue repair model explants that were cultured in different media formulations. Samples were fixed in 10% neutral-buffered formalin at 4°C overnight, dehydrated, cleared using xylene and embedded in paraffin overnight at 60°C. Samples were then sectioned and stained with Harris hematoxylin with glacial acetic acid (Poly Scientific, Bay Shore, NY), 0.02% aqueous fast green solution (Electron Microscopy Sciences), and 0.1% Safranin O solution (Sigma-Aldrich).

Integrative Shear Strength of Repair.

A push–out test was performed on the meniscus tissue repair model explants^{15, 56, 57, 59–62} to determine the shear strength of repair at the interface between the inner core and outer ring of meniscus tissue. Images of the samples were captured using a Genie camera (Teledyne Dalsa, Waterloo, ON, Canada) with a 50 mm lens (Tamron, Cologne, Germany), and Image J (NIH) was used to determine average sample thickness. Each sample was placed in the center of a custom made fixture with a 4 mm diameter hole in the center and placed in a mechanical testing frame (840L TestResources, Shakopee, MN). A 2 mm diameter piston was attached to a load cell and centered directly over the tissue inner core. A 0.25 g tare load was applied until tissue equilibration and then the inner core was pushed-out at a rate of 0.0833 mm/sec. Force-displacement curves were generated and the shear strength of repair was calculated as the peak force divided by the surface area of the interface.

Biochemical Analyses.

MSC-seeded scaffolds and meniscus repair model inner cores were digested in 0.5 mL or 1.5 mL, respectively, of 125 μg/mL papain overnight at 65°C^{54, 67, 68}. DNA content was measured using the picogreen assay (Invitrogen, Carlsbad, CA). Sulfated glycosaminoglycan (sGAG) content was measured using the dimethylmethylene blue assay⁶⁹ and bovine chondroitin sulfate as a standard (Sigma-Aldrich). Collagen content was assessed with the hydroxyproline (OHP) assay⁷⁰, using trans-4-hydroxy-L-proline (Sigma-Aldrich) as a standard, as previously described^{54, 67, 68}.

Experiment 2: Development of Meniscus Tissue Defect Model and Medium Optimization for Co-culture of MSCs and Meniscus Tissue

Media Preparation

Based on the results from experiment 1, we tested the serum-containing medium (meniscal) and serum-free medium supplemented with ITS, Dex, and TGF-β (chondrogenic) in a novel meniscus tissue defect model that combines both MSC-seeded MDM scaffolds and meniscus tissue. These media were chosen for subsequent analyses given that the serum-containing medium appeared to be the most favorable for the meniscus tissue explants, and the serum-free medium supplemented with ITS, Dex, and TGF-β was most favorable for the MSC-seeded MDM scaffolds.

Scaffold Preparation and MSC Seeding

As shown in Figure 2A, three different types of MDM scaffolds were used for these experiments: a slurry, prefabricated non cross-linked (non X-linked), and prefabricated cross-linked (X-linked) scaffold rings. In order to make the slurry, MDM powder was sterilized with ethylene oxide, and was suspended under sterile conditions in ultra-pure distilled water and homogenized to create a 16% by weight MDM slurry. In parallel, MDM powder was reconstituted to a 16% by weight solution and injected into a 3 mm wide gel casting apparatus (BioRad, Hercules, CA) using a 5 mL syringe. The scaffold in the gel casting apparatus was frozen overnight at −80°C and lyophilized for 24 hours to generate an MDM sheet. This sheet was divided, and half was left untreated and half was dehydrothermally cross-linked by heating at 120°C for 24 hours in a dry oven^{54, 71}. Using a biopsy punch, 4 mm diameter scaffolds were removed from both non cross-linked and cross-linked MDM sheets. A 3 mm inner core was removed from each of these 4 mm scaffolds using a biopsy punch to obtain a prefabricated scaffold ring with an inner diameter of 3 mm and an outer diameter of 4 mm. All prefabricated scaffolds were gas-sterilized using ethylene oxide.

Figure 2.

Schematic of scaffold preparation and meniscus tissue defect model utilized in experiment 2. (A) MDM scaffolds were prepared by reconstituting MDM powder to 16% to generate a slurry that was either seeded with MSCs to generate a cell-seeded slurry or was injected into a cassette, frozen, and lyophilized to form a MDM sheet. The MDM sheet was divided and left non-cross-linked (Non X-link) or was dehydrothermally cross-linked (X-link). Scaffolds were obtained using a 4 mm diameter biopsy punch. A 3 mm diameter core was removed from each of these prefabricated scaffolds. (B) The meniscus tissue defect model was generated using 8 mm diameter meniscus explants from which a 4 mm diameter core was removed. Next, a 3 mm diameter inner core tissue was cut from this tissue and the resulting tissue ring (4 mm outer diameter and 3 mm inner diameter) was discarded to generate the tissue defect. The defect was filled with non X-link or X-link scaffolds or slurry and the inner 3 mm diameter tissue core was replaced. Meniscus tissue defect model explants were cultured in either meniscal or chondrogenic medium.

Passage 4 MSCs were suspended in expansion medium to achieve a cell suspension of 1 × 10⁷ cells/mL. The non X-linked and X-linked MDM scaffold rings were seeded with 1.6 × 10⁵ cells per scaffold by placing 16μl of cell suspension in the center of the MDM ring and allowing the cell suspension to wick through the MDM scaffolds. The MSC-seeded scaffolds (n = 6/group) were pre-cultured in 1 mL of expansion medium at 37°C and 5% CO₂ for 7 days prior to implantation into the meniscus tissue defect model.

Meniscus Tissue Defect Model

As shown in Figure 2B, 8 mm diameter explants were harvested from the centerline of porcine medial menisci. A 4 mm diameter core was removed from these 8 mm explants. Then, we explanted a 3 mm inner core from this 4 mm tissue. The resulting 4 mm outer diameter and 3 mm inner diameter tissue ring was discarded, leaving a 0.5 mm meniscus tissue defect that was filled with either the prefabricated non X-linked or X-linked MSC-seeded scaffolds, or MDM slurry containing 1.6 × 10⁵ MSCs (n = 6/group). Next, the 3 mm inner core meniscus tissues were replaced in all samples, taking care that the prefabricated scaffold ring surrounded the inner core and scaffold material filled the gap between the inner core tissue and outer ring meniscus tissue. Meniscus tissue defect model samples were cultured in meniscal medium (serum) or chondrogenic medium (serum-free + Dex + TGF-β) for 14 days (n = 3/group) at 37°C and 5% CO₂.

Outcome Measures

Integrative Shear Strength of Repair.

A push-out test was performed to assess the integrative shear strength of repair^{15, 54, 56, 57, 59–62} between the inner core meniscus tissue and MDM scaffolds or slurry using an ElectroForce 3220 Series III mechanical test instrument (TA Instruments, New Castle, DE), as described above.

Biochemical Analyses.

After mechanical testing, the inner core tissue of all meniscus tissue defect model samples was digested in 1 mL of papain^{54, 67, 68} after which DNA, sGAG⁶⁹, and OHP content^{54, 67, 68, 70} were measured as detailed above.

Statistical Analyses

Statistical analyses were performed using Statistica 7.0 (Tibco). For experiment 1, MSC-seeded scaffolds and tissue repair model explants were analyzed by one-way ANOVA and Fisher LSD post hoc tests to determine significant effects of the different media formulations on all outcome measures (α=0.05). For experiment 2, factorial ANOVA and Fisher LSD post hoc tests were performed to determine significant group differences, as well as the interactive effects of medium formulation and scaffold type in the meniscus tissue defect model.

Results

Experiment 1: Medium Optimization for MSCs Seeded in MDM Scaffolds and Meniscal Tissue

Biochemical Analyses

MSC-seeded scaffolds had the highest DNA content when cultured in serum-free medium containing Dex and TGF-β (Figure 3A), and the DNA content was significantly reduced in scaffolds cultured without TGF-β (p < 0.00001). The addition of serum to the medium (serum, Dex, and TGF-β) decreased the DNA content as compared to Dex and TGF-β (p < 0.00001). However, DNA content was significantly higher in the scaffolds cultured in serum, Dex, and TGF-β as compared to both groups cultured in the absence of TGF-β (p < 0.00001). In addition, cell metabolic activity was highest in the MSC-seeded scaffolds cultured in serum-free medium with Dex and TGF-β (Figure 3B), and there was a significant decrease in metabolic activity in MSC-seeded scaffolds cultured with Dex alone (p < 0.00001). Furthermore, cell metabolic activity was the lowest in the serum-containing medium and upon supplementation with Dex and TGF-β there was a significant increase in metabolic activity (p < 0.01). The sGAG content of the MSC-seeded scaffolds was significantly elevated in scaffolds cultured in Dex and TGF-β, as compared to scaffolds cultured in all other media (Figure 3C, p< 0.00001). In addition, collagen content was highest in MSC-seeded scaffolds cultured in Dex and TGF-β (Figure 3D) and the addition of serum caused a significant decrease in the collagen content (p < 0.00005). There was a significant decrease in collagen content in scaffolds cultured in Dex alone, as compared to Dex and TGF-β (p < 0.0005). Collagen content was lowest in the scaffolds treated with serum alone; however, there was a trend towards an increase in collagen content with the addition of Dex and TGF-β (p = 0.06).

Figure 3.

The effects of different combinations of serum, dexamethasone (Dex), and TGF-β3 on the (A) DNA content, (B) cell metabolic activity [based on the absorbance (abs) of WST-1], (C) sulfated glycosaminoglycan (sGAG) content, and (D) collagen (OHP) content of the MSC-seeded MDM scaffolds. The serum-free media contain ITS. Data are expressed as the mean + SEM. Groups not sharing the same letters have p-values < 0.05.

Meniscus tissue DNA content was highest in samples cultured in serum or in serum + Dex + TGF-β (Figure 4A). DNA content was significantly lower in samples cultured without serum (p< 0.01), and these effects were independent of the presence of TGF-β. Furthermore, cell metabolic activity was highest in the meniscus repair model explants cultured with serum, or with serum + Dex + TGF-β (Figure 4B). Cell metabolic activity was significantly lower in samples cultured without serum (p < 0.00001). Meniscus tissue cultured in Dex alone had the lowest metabolic activity, but metabolic activity was significantly increased with TGF-β supplementation (p < 0.005). On the other hand, sGAG content was significantly elevated in meniscus tissues cultured in Dex and TGF-β, as compared to all other media (Figure 4C, p <0.001). No significant differences were detected in the collagen content of the meniscus tissue repair model explants (Figure 4D).

Figure 4.

The effects of different combinations of serum, dexamethasone (Dex), and TGF-β3 on the (A) DNA content, (B) cell metabolic activity [based on the absorbance (abs) of WST-1], (C) sulfated glycosaminoglycan (sGAG) content, and (D) collagen (OHP) content of the meniscus repair model explants. The serum-free media contain ITS. Data are expressed as the mean + SEM. Groups not sharing the same letters have p-values < 0.05.

Histological Analyses

Histological analyses revealed that the scaffolds and meniscus tissue were composed of a predominantly collagen extracellular matrix (Figure 5). In the MSC-seeded scaffolds, TGF-β + Dex and serum + Dex + TGF-β resulted in a denser ECM, as compared to the scaffolds cultured with serum or Dex alone. Strong proteoglycan staining was detectable in both the scaffolds and meniscus tissue cultured in Dex and TGF-β. However, the proteoglycan staining was diminished in scaffolds and tissues cultured in the presence of serum, Dex, and TGF-β.

Figure 5.

Histological staining of the MSC-seeded MDM scaffolds and meniscus tissue in different combinations of serum, dexamethasone (Dex), and TGF-β3. The serum-free media contain ITS. Sections were stained with Safranin-O (red: proteoglycans), fast green (blue: collagen), and hematoxylin (black: nuclei). Scale bar is 100 μm.

Mechanical Analyses

The integrative shear strength of repair was highest in the meniscus repair model explants cultured with serum (Figure 6). The tissues cultured with Dex alone had the lowest shear strength of repair and had significantly reduced strength of repair when compared to explants cultured in either serum or serum + Dex + TGF-β (p < 0.05).

Figure 6.

The effects of different combinations of serum, dexamethasone (Dex), and TGF-β3 on the integrative shear strength of repair of the meniscus tissue repair model explants. The serum-free media contain ITS. Data are expressed as the mean + SEM. Groups not sharing the same letters have p-values < 0.05.

Experiment 2: Development of Meniscus Tissue Defect Model and Medium Optimization for Co-culture of MSCs and Meniscus Tissue

Biochemical Analyses

In the meniscus tissue defect model, no significant differences were observed in the DNA content of the inner core tissue due to either culture medium or scaffold type (Figure 7A). On the other hand, the chondrogenic medium caused a significant increase in the sGAG content of the tissue, as compared to the meniscal medium (Figure 7B, p = 0.05). However, there was no effect of scaffold type on the sGAG content of the inner core meniscus tissue. In addition, no significant differences were noted in the collagen content across culture media or scaffold types (Figure 7C). There were no interactive effects of medium formulation and scaffold type on the biochemical properties.

Figure 7.

The effects of meniscal (serum) or chondrogenic (Dex + TGF-β3 + ITS) medium and different types of MSC-seeded scaffolds [slurry or prefabricated non cross-linked (Non X-link) or prefabricated cross-linked (X-link) scaffolds] on (A) DNA content, (B) sulfated glycosaminoglycan (sGAG) content, and (C) collagen (OHP) content of meniscus tissue defect inner cores. Data are expressed as mean ± SEM. sGAG content is higher in the chondrogenic medium than the meniscal medium (p = 0.05). No significant differences are detected by scaffold type.

Mechanical Analyses

The integrative shear strength of repair was significantly increased in meniscus tissue defect model explants that were cultured in the chondrogenic medium, as compared to explants cultured in the meniscal medium (Figure 8, p < 0.001). In addition, there was a significant effect of scaffold type, revealing that defects treated with the prefabricated X-linked scaffolds had a higher shear strength of repair than meniscus defects treated with the slurry (p < 0.05). There were no interactive effects of medium formulation and scaffold type on the shear strength of repair.

Figure 8.

The effects of meniscal (serum) or chondrogenic (Dex + TGF-β3 + ITS) medium and different types of MSC-seeded scaffolds [slurry or prefabricated non-cross-linked (Non X-link) or prefabricated cross-linked (X-link) scaffolds] on the integrative shear strength of repair of meniscus tissue defect model explants. Data are expressed as mean ± SEM. The chondrogenic medium improves shear strength of repair over the meniscal medium (p < 0.001). Meniscal defects treated with the X-link prefabricated scaffolds have a higher shear strength of repair than defects treated with the slurry (p < 0.05).

Discussion

In the present study, we established favorable culture conditions for in vitro meniscus tissue engineering utilizing bone marrow-derived MSCs, which will allow efficient and rapid screening of factors that may modulate meniscus tissue repair. The serum-free medium containing both TGF-β and Dex, which is commonly utilized for MSC chondrogenesis^{19, 22, 23, 31–35, 37–39, 47}, was the most promising formulation for future experiments with MSC-seeded MDM scaffolds. In contrast, the serum-containing medium, which is commonly used for meniscus tissue culture^{15, 54, 56–64}, was the most effective medium for meniscus tissue culture and integrative repair of native meniscus tissue. When native meniscus tissue and MSC-seeded MDM scaffolds were combined in a co-culture meniscus tissue defect model, the chondrogenic medium (serum-free medium containing both TGF-β and Dex) enhanced proteoglycan production and promoted the integrative repair of meniscus tissue.

Serum-free medium supplemented with both TGF-β and Dex yielded the most favorable biochemical properties for the MSC-seeded scaffolds from experiment 1. Our results suggest that TGF-β and Dex promoted MSC proliferation and increased production of the major meniscus ECM components, collagens and proteoglycans. In particular, the elevated DNA and sGAG content in the Dex and TGF-β treated samples appears to be driven predominantly by TGF-β, since MSC-seeded scaffolds cultured with Dex alone did not show these increases. However, samples treated with TGF-β alone would be required to determine if there is a synergistic effect between Dex and TGF-β. Prior work screening 32 different media formulations for MSC chondrogenesis revealed that differentiation medium composed of TGF-β and Dex promoted the most favorable chondrogenic gene expression⁷², which is consistent with our findings. In other prior studies, the addition of TGF-β to chondrogenic medium containing Dex upregulated type II collagen and aggrecan gene expression and type I collagen and GAG production in MSC cell pellets³⁷. Likewise, transduction of MSCs with cDNA encoding TGF-β increased expression of type I and II collagen and sGAG production²³. In this study, we tested the effects of Dex alone as it has previously been shown to increase sGAG and collagen production by MSCs⁷³. Furthermore, Dex treatment of MSCs has been shown to suppress catabolism, including aggrecanase activity⁷⁴ and matrix metalloproteinase (MMP)-13 expression, and the pro-inflammatory mediator prostaglandin E2⁷⁵, factors which are key mediators of joint tissue degradation and OA progression.

Serum-supplemented medium resulted in the lowest DNA content, metabolic activity and proteoglycan and collagen content of the MSC-seeded MDM scaffolds. Furthermore, the addition of serum to the Dex and TGF-β containing medium attenuated the beneficial effects of TGF-β on collagen and proteoglycan content and decreased the DNA content. Prior work demonstrated that serum prevented MSC aggregate formation⁷⁶. More recent studies have shown that serum-free MSC chondrogenic medium increased type II collagen and sGAG production, but the addition of serum caused increased type I collagen expression and staining⁷⁷, suggesting a fibrochondrogenic phenotype. However, in this study, the general suppression of metabolic activity and ECM production in the serum-containing medium suggests the presence of inhibitors of TGF-β and/or Dex signaling in the serum that prevent MSC proliferation and differentiation.

This study focused specifically on bone marrow-derived MSCs. Due to inherent differences in stem cells isolated from different sites, further studies with stem cells from other sources will be needed to identify favorable medium formulations for in vitro co-culture with meniscus tissue. Additionally, in this study, MSCs were cultured on a MDM scaffold but our findings are consistent with other studies of MSCs in alginate, agarose, and type I collagen-GAG matrices^{23, 37, 72–76}. This suggests that the chondrogenic medium is able to drive MSC differentiation across a variety of different scaffolds.

In contrast to the MSC-seeded MDM scaffolds, the meniscus repair model explants had the highest DNA content, metabolic activity, and integrative shear strength of repair when cultured in serum supplemented medium. However, we did not detect a significant difference in the shear strength of repair between media containing serum and medium containing Dex + TGF-β. Previously, we have shown that serum promotes meniscus cell proliferation and accumulation in a micro-wound repair model⁶¹. In general, meniscus tissue culture experiments are performed with 10% serum, and the effects of TGF-β supplementation on tissue repair were rather minimal in most prior experiments. In bovine meniscal repair model explants, 10 ng/mL TGF-β3 had no effect on strength of repair until 8 weeks of treatment⁶³, suggesting that longer culture times may be required to see beneficial effects of TGF-β3 on tissue repair. On the other hand, 50 ng/mL TGF-β1 treatment of in vitro meniscal defects for 8 weeks did not affect cell density or proliferation⁶⁴. In addition, we have previously shown that 1 ng/mL TGF-β1 promotes cellular accumulation at the repair interface and integrative meniscus repair but 10 ng/mL TGF-β1 did not^{58, 61}, suggesting a dose response to TGF-β1 treatment. Since both TGF-β1 and TGF-β3 isoforms promote fibrochondrogenic differentiation of MSCs^{19, 31, 78, 79}, in this study we chose to test the effect of 10 ng/mL TGF-β3 as this is the most commonly used concentration of TGF-β3 in chondrogenic medium^{19, 31}.

In the meniscus tissue repair model explants, TGF-β and Dex increased sGAG content of the tissue. This effect is likely driven by TGF-β since Dex treatment alone did not enhance sGAG tissue content. The suppressed proteoglycan content upon the addition of serum to the TGF-β and Dex medium further suggests that there may be inhibitors of TGF-β signaling in the serum. Our findings regarding TGF-β-mediated upregulation of sGAG production are consistent with prior studies, which showed that meniscal cells in monolayer, cells in alginate beads and poly-glycolic acid (PGA) scaffolds, and meniscal explants increase proteoglycan production in response to TGF-β treatment^{24, 80–82}. Transduction of meniscus cells with TGF-β or recombinant TGF-β treatment also increased collagen production as measured by radiolabeling^{24, 81, 83}. However, we did not detect differences in collagen content in our meniscus repair model explants with the different media treatments. This may be due to subtle changes in total tissue collagen content, as compared to measuring new collagen synthesis by radiolabel incorporation.

While glucocorticoid treatment is frequently used as a conservative treatment for OA pain, few studies have directly examined the effects of the synthetic glucocorticoid steroid Dex on meniscus tissue. In our study, Dex treatment showed the lowest DNA content, cell metabolic activity, sGAG content, and shear strength of repair in the meniscus repair model explants. Prior work has shown that Dex treatment of isolated meniscus cells induced autophagy and apoptosis and also upregulated the expression of aggrecanases and matrix metalloproteinases⁸⁴. Dex has also been shown to cause a significant decrease in medial meniscus cell viability in the presence of blood⁸⁵. However, in an ex vivo impact injury of porcine knee joints, Dex treatment suppressed meniscus IL-1β and MMP-3 expression and upregulated miR140, which is a negative regulator of aggrecanase activity⁸⁶. Most recently, in a rabbit surgical drill injury model, Dex treatment caused a downregulation in the meniscal expression of inflammatory and degradative markers⁸⁵. Given these conflicting findings, further studies are necessary to understand the full spectrum of effects of Dex treatment on meniscus tissue.

Based on the beneficial effects of the serum-free medium containing TGF-β and Dex (chondrogenic medium) on the MSC-seeded scaffolds and the proteoglycan content of the meniscus tissue, and the positive effects of the serum-containing medium (meniscal medium) on the meniscus repair model explants, we evaluated the effects of both of these media on our meniscus tissue defect model. In this experiment, we did not evaluate the effects of serum, Dex, and TGF-β containing medium, which showed intermediate results in experiment 1, and thus may be a good medium formulation to consider for future experiments. Nonetheless, in this experiment when the MSC-seeded MDM scaffolds were placed into the meniscus tissue defect and co-cultured, the chondrogenic medium increased sGAG content and integrative shear strength of repair. The elevated proteoglycan content in this defect model is consistent with our findings in experiment 1. The increase in the shear strength with the chondrogenic medium is in contrast with the tissue repair results from experiment 1. While this difference could be due to the mixed species co-culture of human MSCs with porcine meniscus tissue, it more likely suggests an important contribution of the MSCs to meniscus repair. Previously, we have shown that MSC-seeded MDM scaffolds trended towards improved integrative repair when cultured as an inner core surrounded by a meniscus tissue ring in meniscal medium, but this repair was not significantly improved over a native tissue core⁵⁴. In the current study, the chondrogenic medium containing TGF-β and Dex improved integrative meniscus repair, suggesting that exogenous growth factors are necessary to enhance MSC-mediated repair using the MDM scaffold. These findings are consistent with prior work using MSC-seeded scaffolds composed of tissue-derived matrix, which required chondrogenic medium containing TGF-β to promote matrix gene expression and ECM production^{33, 47, 87, 88}.

Several studies have been performed evaluating co-cultures of MSCs and meniscus cell pellets^{36–39, 89}. These studies revealed that there is a synergy between the meniscus cells and MSCs that promotes type I collagen and GAG production and suppresses expression of hypertrophic genes during culture with chondrogenic medium^{37–39, 89}. Furthermore, co-cultures were partially protected against the pro-inflammatory cytokine IL-1, allowing retention of a fibrochondrogenic phenotype³⁹. On the other hand, 50:50 co-cultures of MSCs and meniscus cells seeded on a 3D collagen scaffold and cultured in meniscal medium had the highest sGAG retention and mechanical properties and diminished hypertrophy, as compared to all other co-culture ratios³⁶. While these studies used either the meniscal or chondrogenic medium, our experiments directly compared the effects of these media formulations on cell proliferation, ECM production, and strength of repair. In addition to utilizing MSCs to directly promote cell-mediated repair, MSCs can also be utilized to deliver growth factors or other therapeutics through viral transduction^{23, 83, 90}. Therefore, future work could explore the use of genetically modified MSCs to deliver therapeutics to further improve meniscus healing.

Prior in vitro meniscus repair models⁵⁸ created either a biopsy punch defect that was filled with a reparative scaffold^{54, 63, 91} or a meniscal tear that was reintegrated and evaluated for repair^56–63, similar to our meniscal repair model system in experiment 1. In our current study, we established a novel meniscus tissue defect model that created a 0.5 mm defect that can be filled with a reparative scaffold and evaluated for integration between two meniscus tissue explants. Although circular defects are not natural defects that occur during meniscal injuries, this model allows quantification of tissue healing (through push-out testing) between two pieces of tissue that are bridged by a tissue engineered scaffold, which is not feasible with a naturally occurring defect. As compared to a surgically created tear, our meniscal tissue defect model more closely mimics clinical meniscal defects that are generally left after a meniscus tear debridement that could benefit from tissue-engineered repair strategies to augment tissue healing. In addition to adding MSCs to biomaterial scaffolds, it is possible to modulate the properties of scaffolds through various material processing and cross-linking strategies. In our meniscus tissue defect model, the prefabricated X-linked scaffolds had improved integrative shear strength of repair, as compared to the defects filled with MDM slurry. Overall, the prefabricated scaffolds trended towards a higher shear strength of repair, which may be due to the enhanced mechanical integrity of the prefabricated scaffolds and/or improved retention of the scaffolds. Previous studies have shown that cross-linking of cartilage-derived scaffolds may be necessary to increase the mechanical properties of the scaffolds and prevent cell-mediated retraction away from the tissue interface⁷¹. However, in the current study we were unable to see a significant difference in shear strength between the non X-linked and X-linked scaffolds, but this may be due to the small sample size used in these experiments. However, in our prior work comparing scaffolds of varying MDM concentrations and cross-linking in our meniscus tissue repair model with meniscal medium, we also did not detect a difference in shear strength with dehydrothermal cross-linking⁵⁴. Nevertheless, there was improved integrative shear strength of repair in the higher percentages of MDM⁵⁴, laying the ground work for testing 16% MDM in our meniscus tissue defect model.

In conclusion, the chondrogenic serum-free medium containing both TGF-β and Dex promoted integrative meniscus repair in meniscus tissue defects repaired with bone marrow-derived MSC-seeded MDM scaffolds. With the ability to modulate the differentiation of MSCs to provide an abundant source of fibrochondrogenic cells, and an MDM scaffold that supports the growth and differentiation of MSCs that can integrate with meniscal defects to promote repair, there is great promise for successful meniscus tissue engineering and improved tissue repair. In addition to cell-based repair, MSCs can be tailored to deliver therapeutic agents to the meniscus defect to promote cellular differentiation, tissue repair, and/or suppress the pro-inflammatory microenvironment following joint injury^{23, 83, 90, 92, 93}. Finally, the meniscal tissue defect model that was established in this paper can be used to perform in vitro screening to identify and optimize biological and mechanical treatments that may be utilized to enhance meniscus tissue repair prior to moving into preclinical animal studies.