Abstract
While the clinical outcomes of acetabular labral repair and reconstruction have been reported, comparative cellular responses between common allografts to clinically relevant load are less clear. This study aims to compare acetabular labrum (LAB), anterior tibialis tendon (TEN), and knee meniscus (MEN) cellular responses to biaxial tensile strain during in vitro culture. LAB, TEN, and MEN tissues were recovered from skeletally mature research hounds (n = 5). Primary LAB, TEN, and MEN fibroblast cell cultures were established. Using a bioreactor, cells were loaded at 0%, 4%, or 10% biaxial tensile strain for 5 days. RNA was extracted and reverse transcription-polymerase chain reaction (RT-PCR) was used to determine relative gene expression levels. Cells were then tested for various stress-induced biomarkers. Statistical analysis determined significance between groups for relative gene expression levels and biomarker concentrations. LAB in the 0% strain group had more viable cells compared to the 4% and 10% strain groups, and MEN fibroblasts in the 4% strain group had more viable cells compared to the 10% strain group. LAB and MEN were associated with higher concentrations of MMP-3 in the 10% strain compared to the 0% strain group. The characterization of acetabular labrum cellular responses to clinically relevant loads of force more closely matches those of meniscal allograft tissue than tibialis anterior allograft tissue. This has potential implications for labral reconstruction, as meniscal allograft tissues may be more suited than tibialis anterior tissues to withstand compressive forces necessary for stabilization and chondroprotection within the hip.
INTRODUCTION
The acetabular labrum plays integral roles in the function of the hip, including contributions to joint stability, load distribution, protection of articular cartilage, and maintenance of the suction-seal effect [1–3]. Pathology of the acetabular labrum is associated with degeneration of femoral head and acetabular articular cartilage and progression to symptomatic hip osteoarthritis with severity increasing with age [4, 5]. Labral resection, repair, and reconstruction procedures are commonly used to treat labral pathology that is unresponsive to physical therapy or activity modification [6]. Resection may alleviate pain but fails to restore the important functions of the labrum, resulting in increased stress on the acetabulum [7]. Labral repair is indicated when the tissue is amenable to functional preservation via arthroscopy [6]. Reconstruction using various autografts and allografts, including tendons and meniscus, can be performed in labral deficient patients who are not candidates for labral repair [1, 2, 8, 9]. Each of these treatment options requires effective cellular responses from remaining labrum and/or grafts in order to preserve hip joint function long-term [10]. Immense heterogeneity exists in allograft tissue choice among surgeons [8]. Philippon et al. [11] reported an acetabular labral reconstruction has a survival rate of 61% at 10 years using iliotibial band autograft. It may be possible to decrease this failure rate by choosing allograft tissue that demonstrates similar biomechanical and biochemical properties to acetabular labral tissue.
The acetabular labrum provides both a suction seal and load distribution effect [1, 2]. Similarly, the knee meniscus provides load distribution and contributes to the stability of the knee [12]. Its collagen network is arranged in a multilaminar weave which resists hoop stresses through both axial and shears loading [12]. Previous research has found knee meniscus to have morphologic and metabolic similarities with native acetabular labral tissue [13]. This comparison, however, did not address the role of tensile strain in these tissues. Mechanical strain has been shown to impact gene expression and enzymatic activity in labral tissue and its candidate graft types, playing an important role in tissue regulation [10, 14, 15]. While these studies represent significant strides in understanding biomechanical and cellular properties of the acetabular labrum and its grafts, direct comparisons between tissue types can be challenging due to differences in study design. The effects of mechanical load are highly variable depending on frequency, magnitude, duration, and type of strain [14]. To the authors’ knowledge, no previous studies compare acetabular labrum (LAB), anterior tibialis tendon (TEN), and knee meniscus (MEN) cellular responses to biaxial tensile strain during in vitro culture. The study was designed to test the hypothesis that knee meniscus cell viability and extracellular matrix, degradative, and inflammatory biomarker responses to clinically relevant tensile loading would more closely match those of acetabular labrum than anterior tibialis tendon.
MATERIALS AND METHODS
Cell culture
With institutional animal care and use committee approval (#16200), knee menisci, acetabular labrums, and anterior (cranial) tibialis tendons were recovered aseptically from skeletally mature, purpose bred mix breed research hounds (n = 5) euthanatized for reasons unrelated to this study. All tissue processing and cell culture work was performed under sterile conditions in a BSL2 laminar flow hood. Tissues were minced into 0.5–1 cm pieces and digested overnight in 0.5% collagenase II (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) solution. The digests were centrifuged to pellet cells, washed with phosphate-buffered saline (PBS) to remove residual collagenase II, and filtered through a cell strainer to remove undigested tissue debris. The cells were seeded onto 25 cm2 flasks and cultured with supplemented high glucose Dulbecco's Modified Eagle Medium (DMEM) (0% Serum Pluss II (Sigma), 0.002% penicillin, 100 μg/ml streptomycin, 25 μg/ml amphotericin B, 0.002% L-ascorbate, and 0.01% L-glutamine) and cultured at 5% CO2, 37°C, and 95% humidity. The media were replaced every 3–4 days until cells were >90% confluent. To expand cell numbers, cells from each tissue from each dog were passaged onto a 75 cm2 flask and then a 150 cm2 flask cultured as above until they were >90% confluent at each passage. Once the passage 2 cells on the 150 cm2 flask reached confluence, they were used for tension cell culture as described below. During culture, cells were monitored for contamination grossly (colour of the media pH indicator, clarity of the media) and microscopically (presence of bacterial, yeast, or fungal colonies). No contamination was observed during initial cell culture or during loading culture described below.
Passage 2 fibroblasts (n = 5/group/tissue type) were seeded on Collagen Type I-coated BioFlex® plates (1 × 105 cells/well) in supplemented DMEM described above. Cells were incubated overnight before strain was applied. Fibroblasts were subjected to continuous mechanical stimulation (2-s strain and 10-s relaxation at a 0.5 Hz frequency) with a biaxial sinusoidal waveform with three different elongation strains that simulated three different clinical situations (mechanical stress deprivation—0%, physiologic strain—4%, and high strain—10%) for 5 days using the Flexcell FX-4000T strain system (Flexcell International, NC, USA. Strain levels were selected based on a previous study of human acetabular labral strain [16]. After 5 days of culture, a sample of the media was collected and stored at −20°C for biomarker analysis and then the relative level of cell activity each well was determined using the resazurin assay as described below.
Cell viability
On Day 5 of culture, the media in the well were replaced with 3 ml of fresh media containing 10 μg/ml resazurin sodium salt. Additionally, the resazurin media were added to two wells without cells as a blank control. Cells were incubated 2 h at for 5% CO2, 37°C, and 95% humidity, and after 2 h a sample of the resazurin media was collected for analysis. The level of fluorescence (ex: 560 nm, em: 590 nm) in the sample and blanks was determined using a Synergy HT plate reader. The level of fluorescence measured in the blank well was averaged and subtracted from the mean level of fluorescence in each sample to obtain the standardized fluorescent level for each sample. For this assay, increased fluorescence is indicative of higher cell number. After the resazurin assay, the cells were processed for mRNA extraction and gene expression analysis as described below.
Gene expression
Resazurin media were removed, and cells were washed with PBS (Table 1). The cells were lysed using the Qiagen RLT cell lysis buffer, and the RNA was extracted from the cells using the Qiagen RNeasy mini-kit according to the manufacture’s protocol. The quality and quantity of RNA in the sample was determined by absorbance. The RNA was transcribed to cDNA using the Quantitect Reverse Transcription kit (Qiagen). Real-time PCR was performed using a LightCycler 480 and the QuantiNova SYBR Green PCR kit (Qiagen) to determine gene expression level of collagen (COL) I, COL III, cartilage oligomeric protein (COMP), decorin, transforming growth factor (TGF)-β, cyclooxygenase (COX)-2, a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS)5, matrix metalloproteinase (MMP)-2, MMP-3, tissue inhibitor of metalloproteinases (TIMP)-1, TIMP-2, and TIMP-3 relative to the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH) using species specific primer sets (Table 1). GAPDH has been previously used as internal control previous studies examining meniscus, labral, and tendon tissues under tensile load [10, 15, 17, 18]. Relative gene expression level was determined for each gene using the ∆ct method.
Table 1.
PCR primers used for gene expression analysis.
| Gene | Primer | Sequence | Product length |
|---|---|---|---|
| COL1A1 | Forward | GTCTGGTACGGCGAGAGCAT | 154 |
| Reverse | CCACGCTGTTCTTGCAGTGG | ||
| COL3A1 | Forward | GTTTGCCCACAGCCTCCAAC | 115 |
| Reverse | TACCGGGGTCGCCATTTCTC | ||
| COMP | Forward | GACGCGCAGATAGATCCCAA | 119 |
| Reverse | TCGAAGTCCACGCCATTGAA | ||
| Decorin | Forward | GGCGGAGCATAAGTACATCCA | 140 |
| Reverse | GCACTGGGTTGCTGAAAAGG | ||
| TGF-β | Forward | CACTGGAGTCGTGAGGCAGT | 124 |
| Reverse | ACTGAACCCTGCGTTAATGTCT | ||
| COX-2 | Forward | CAGGCAAATTGCTGGCAGGG | 137 |
| Reverse | TCAGCCTAAAGCGTTTGCGA | ||
| ADAMTS5 | Forward | GTCAACTCAGTCACCAGCCA | 154 |
| Reverse | CTAACTTGCGGTTTGCGTCC | ||
| MMP-2 | Forward | CACCCCAACTCTGGGACCTG | 142 |
| Reverse | CATGGGCTTGTTCCGTGGTG | ||
| MMP-3 | Forward | CCTTCGACGCAATCAGCACC | 135 |
| Reverse | AGCCTGAAGGAAGGGATGGC | ||
| TIMP-1 | Forward | GGGACCGCAGAAGTCAACCA | 100 |
| Reverse | TGTCCGAGGCATTCCCCAAG | ||
| TIMP-2 | Forward | TTCTGCAACGCGGACGTAGT | 91 |
| Reverse | GCTTGATGGGGTTGCCGTAG | ||
| TIMP-3 | Forward | GGCAAGATGTACACAGGACT | 154 |
| Reverse | CACTCGTTCTTGGAGGTCAC | ||
| GAPDH | Forward | AAGGCTGTGGGCAAGGTCAT | 160 |
| Reverse | CCTCCGATGCCTGCTTCACT |
Biomarkers
The Day 5 media were tested for Prostaglandin E2 (PGE2) (Cayman Chemical), COMP, procollagen Type I N-propeptide (PINP) (Kendall Scientific), C-telopeptide of Type I collagen (CTX-I) (Kendall Scientific), MMP-3 (Kendall Scientific), MMP-13 (Kendall Scientific), and ADAMTS4 (Kendall Scientific) using commercially available canine specific enzyme-linked immunosorbent assay (ELISA) assays. The concentrations of interleukin (IL)-6, IL-8, monocyte chemoattractant protein (MCP)-1, keratinocyte-derived chemokine (KC), and tumour necrosis factor (TNF)-α were determined using a canine specific Luminex assay (MilliporeSigma). The concentration of MMP-2 was determined using a Human Luminex assay (R&D Systems) that was found to cross react with canine samples in previous studies [19, 20]. The concentration of proteoglycans (GAG) in the media was determined using the 1,9-dimethylmethylene blue as previously described [21].
Statistical analysis
To account for variations in cell number between wells, the concentration of each media biomarker was standardized to the level of fluorescence measured in the resazurin assay for each sample and natural log transformed for statistical analysis. Significant (P < .05) differences in media biomarker concentrations among load groups within each cell type and among cell types within each load group were determined using one-way analysis of variance (ANOVA) with Tukey post hoc analyses. Significant (P < .05) differences in relative gene expression level among load groups within each cell type and among cell types within each load group were determined using Kruskal-Wallace tests with Bonferroni corrections.
RESULTS
Cell viability
Within tissue types, LAB fibroblasts in the 0% strain group had a higher number of viable cells compared to the 4% (P = .03) and 10% (P = .005) strain groups, and MEN fibroblasts in the 4% strain group had a higher number of viable cells compared to the 10% (P = .045) strain group (Table 2). No significant differences in cell viability were noted among strain levels for TEN fibroblasts.
Table 2.
Cell viability between tissue types across strain levels.
| Strain group | ||||||
|---|---|---|---|---|---|---|
| Cell type | 0% | 4% | 10% | |||
| Mean | SD | Mean | SD | Mean | SD | |
| LAB | 47720.2 | 6 594 | 24786.8 | 13 491 | 16622.4 | 15 020 |
| MEN | 53 361 | 5 358 | 80626.2 | 14 165 | 49304.6 | 27 574 |
| TEN | 47975.75 | 20 104 | 49008.75 | 21 996 | 48853.5 | 26 936 |
When comparing among cell types, MEN fibroblasts had a higher number of viable cells compared to LAB (P = .038) and TEN (P < .001) fibroblasts in the 4% strain group. No significant differences in cell viability were noted among cell types in the 0% or 10% strain groups. Refer to Table 2 for a complete list of cell viability data.
Extracellular matrix responses
In response to biaxial tensile strain, LAB fibroblasts expressed significantly (P = .011) lower levels of COL I in the 10% strain group compared to the 0% strain group (Figs 1, 2 and Tables 3, 4). MEN fibroblasts expressed lower levels of COL I in the 10% (P = .012) and 4% (P = .015) strain groups compared to the 0% strain group. The expression of COL I by TEN fibroblasts was not significant among strain groups. The expression levels of decorin were higher in LAB fibroblasts in the 4% strain group compared to the 0% strain group (P = .003). All relative gene expression levels are displayed in Table 3.
Figure 1.
Representative box and whisker plots for the relative gene expression level of the extracellular matrix proteins Collagen I, Collagen III, COMP, and Decorin on Day 5 of in vitro culture. (*) Significantly higher than other loads within tissues. (‡) Significantly higher than other tissues between tissues.
Figure 2.
Representative box and whisker plots for the concentration of the extracellular matrix proteins CTX-I, PINP, COMP, and GAG in the media on Day 5 of in vitro culture. (*) Significantly higher than other loads by tissue type. (‡) Significantly higher than other tissues by load.
Table 3.
Relative gene expression across strain levels.
| Strain group | |||||||
|---|---|---|---|---|---|---|---|
| Gene | Cell type | 0% | 4% | 10% | |||
| Mean | SD | Mean | SD | Mean | SD | ||
| COL I | LAB | 3.463 | 2.796 | 1.729 | 1.143 | 0.894 | 0.766 |
| MEN | 4.28 | 2.122 | 1.647 | 0.979 | 1.543 | 0.568 | |
| TEN | 1.789 | 0.888 | 1.777 | 1.295 | 1.225 | 0.734 | |
| COL III | LAB | 0.06 | 0.039 | 0.021 | 0.015 | 0.035 | 0.034 |
| MEN | 0.091 | 0.111 | 0.051 | 0.037 | 0.046 | 0.015 | |
| TEN | 0.047 | 0.038 | 0.028 | 0.016 | 0.033 | 0.046 | |
| COMP | LAB | 0.049 | 0.047 | 0.064 | 0.057 | 0.047 | 0.033 |
| MEN | 0.029 | 0.026 | 0.021 | 0.015 | 0.034 | 0.019 | |
| TEN | 0.021 | 0.016 | 0.041 | 0.016 | 0.032 | 0.012 | |
| DEC | LAB | 0.239 | 0.103 | 0.074 | 0.016 | 0.161 | 0.14 |
| MEN | 0.232 | 0.147 | 0.125 | 0.087 | 0.243 | 0.136 | |
| TEN | 0.808 | 0.96 | 0.268 | 0.243 | 0.279 | 0.237 | |
| MMP-2 | LAB | 0.038 | 0.035 | 0.08 | 0.027 | 0.063 | 0.016 |
| MEN | 0.032 | 0.026 | 0.151 | 0.11 | 0.094 | 0.026 | |
| TEN | 0.191 | 0.207 | 0.111 | 0.079 | 0.101 | 0.061 | |
| MMP-3 | LAB | 0.034 | 0.037 | 0.076 | 0.025 | 0.058 | 0.009 |
| MEN | 0.024 | 0.027 | 0.162 | 0.125 | 0.102 | 0.03 | |
| TEN | 0.215 | 0.25 | 0.109 | 0.078 | 0.094 | 0.056 | |
| ADAMTS5 | LAB | 0.014 | 0.01 | 0.003 | 0.002 | 0.005 | 0.003 |
| MEN | 0.011 | 0.009 | 0.002 | 0.002 | 0.003 | 0.001 | |
| TEN | 0.007 | 0.004 | 0.003 | 0.001 | 0.004 | 0.002 | |
| TIMP-1 | LAB | 0.357 | 0.254 | 0.105 | 0.098 | 0.079 | 0.075 |
| MEN | 0.435 | 0.177 | 0.449 | 0.142 | 0.18 | 0.09 | |
| TEN | 0.167 | 0.122 | 0.207 | 0.237 | 0.207 | 0.274 | |
| TIMP-2 | LAB | 0.032 | 0.023 | 0.049 | 0.009 | 0.032 | 0.015 |
| MEN | 0.045 | 0.062 | 0.076 | 0.03 | 0.032 | 0.009 | |
| TEN | 0.064 | 0.034 | 0.087 | 0.043 | 0.08 | 0.029 | |
| TIMP-3 | LAB | 0.001 | 0.001 | 0.001 | 0 | 0.001 | 0.001 |
| MEN | 0.002 | 0.002 | 0.001 | 0 | 0 | 0 | |
| TEN | 0.001 | 0.001 | 0.001 | 0.001 | 0.002 | 0.003 | |
| COX-2 | LAB | 0.006 | 0.01 | 0.014 | 0.022 | 0.004 | 0.004 |
| MEN | 0.004 | 0.002 | 0.006 | 0.005 | 0.008 | 0.005 | |
| TEN | 0.002 | 0.002 | 0.007 | 0.007 | 0.062 | 0.095 | |
| TGF-β | LAB | 0.004 | 0.003 | 0.004 | 0 | 0.003 | 0.001 |
| MEN | 0.006 | 0.005 | 0.004 | 0.003 | 0.003 | 0.001 | |
| TEN | 0.003 | 0.002 | 0.005 | 0.002 | 0.003 | 0.002 | |
Table 4.
Extracellular matrix biomarker expression across strain levels.
| Strain group | |||||||
|---|---|---|---|---|---|---|---|
| Biomarker | Cell type | 0% | 4% | 10% | |||
| Mean | SD | Mean | SD | Mean | SD | ||
| GAG (μg/resa) | LAB | 0.142 | 0.033 | 0.502 | 0.426 | 2.248 | 3.547 |
| MEN | 0.121 | 0.025 | 0.168 | 0.076 | 0.404 | 0.438 | |
| TEN | 0.15 | 0.104 | 0.212 | 0.121 | 0.413 | 0.433 | |
| COMP (μg/resa) | LAB | 0.2 | 0.083 | 0.894 | 0.703 | 2.748 | 4.367 |
| MEN | 0.143 | 0.067 | 0.198 | 0.055 | 0.278 | 0.352 | |
| TEN | 0.191 | 0.137 | 0.324 | 0.138 | 0.688 | 0.821 | |
| PINP (μg/resa) | LAB | 2.822 | 1.278 | 12.766 | 5.686 | 53.963 | 100.915 |
| MEN | 1.731 | 0.757 | 2.118 | 0.98 | 5.545 | 8.772 | |
| TEN | 2.939 | 1.499 | 4.975 | 1.389 | 10.746 | 8.213 | |
| CTX-I (μg/resa) | LAB | 0.016 | 0.003 | 0.039 | 0.03 | 0.193 | 0.266 |
| MEN | 0.017 | 0.003 | 0.008 | 0.004 | 0.043 | 0.061 | |
| TEN | 0.02 | 0.012 | 0.023 | 0.02 | 0.025 | 0.032 | |
The media concentrations of the Collagen I synthesis protein biomarker, PINP, were higher in LAB (P = .046) and TEN (P = .018) 10% strain groups compared to the respective 0% strain groups. The concentrations of COMP (P = .018), GAG (P = .025), and CTX-I (P = .024) in the media were higher in LAB fibroblasts in the 10% strain group compared to the 0% strain group. The concentrations of CTX-I (P = .037) were higher in MEN fibroblasts in the 10% strain group compared to the 4% strain group. Refer to Table 4 for more information on extracellular matrix biomarker levels.
For comparisons among cell types within each load group, the only difference in gene expression levels was higher expression of COL I by MEN fibroblasts compared to TEN fibroblasts in the 0% strain group (P = .018). For protein biomarkers, concentrations of PINP (P ≤ .001), CTX-I (P = .018), and COMP (P = .003) were higher for LAB fibroblasts, and the concentration of PINP (P = .017) was higher for TEN fibroblasts, when compared to MEN fibroblasts in the 4% strain group. The concentrations of PINP (P = .018) and COMP (P = .045) were higher for LAB fibroblasts compared to TEN fibroblasts in the 4% strain group.
Degradative responses
The application of 10% biaxial tensile strain to LAB fibroblasts was associated with higher concentrations of ADAMTS4 (P = .018), MMP-2 (P = .004), and MMP-3 (P = .005) protein biomarkers when compared to the 0% strain group (Figs 3, 4 and Tables 3, 5). Concentrations of MMP-3 (P ≤ .002) for MEN fibroblasts in the 10% strain group were higher when compared to the 0% and 4% strain groups.
Figure 3.
Representative box and whisker for the relative gene expression level of MMP-2, MMP-3, ADAMTS5, TIMP-1, TIMP-2, and COX-2 on Day 5 of in vitro culture. (*) Significantly higher than other loads by tissue type. (‡) Significantly higher than other tissues by load.
Figure 4.
Representative box and whisker plots for the concentration of MMP-2, MMP-3, ADAMTS4, IL-8, MCP-1, and PGE2 in the media on Day 5 of in vitro culture. (*) Significantly higher than other loads by tissue type. (‡) Significantly higher than other tissues by load.
Table 5.
Degradative biomarkers expression across strain levels.
| Strain group | |||||||
|---|---|---|---|---|---|---|---|
| Biomarker | Cell type | 0% | 4% | 10% | |||
| Mean | SD | Mean | SD | Mean | SD | ||
| MMP-2 (pg/resa) | LAB | 1120.745 | 186.627 | 1809.873 | 308.529 | 3092.328 | 1855.989 |
| MEN | 1222.043 | 491.928 | 1138.66 | 376.143 | 1686.407 | 235.921 | |
| TEN | 2 177 | 1696.75 | 2115.945 | 920.825 | 2103.228 | 900.967 | |
| MMP-3 (ng/resa) | LAB | 0.141 | 0.062 | 0.728 | 1.077 | 1.692 | 1.968 |
| MEN | 0.081 | 0.006 | 0.102 | 0.035 | 0.389 | 0.326 | |
| TEN | 0.148 | 0.113 | 0.217 | 0.118 | 0.385 | 0.369 | |
| MMP-13 (ng/resa) | LAB | 0.027 | 0.018 | 0.077 | 0.053 | 0.224 | 0.286 |
| MEN | 0.021 | 0.013 | 0.016 | 0.01 | 0.017 | 0.02 | |
| TEN | 0.031 | 0.018 | 0.028 | 0.027 | 0.065 | 0.072 | |
| ADAMTS4 (ng/resa) | LAB | 0.025 | 0.008 | 0.1 | 0.082 | 0.453 | 0.771 |
| MEN | 0.024 | 0.008 | 0.015 | 0.004 | 0.06 | 0.074 | |
| TEN | 0.033 | 0.021 | 0.033 | 0.021 | 0.056 | 0.049 | |
The gene expression levels of ADAMTS5 for LAB (P = .042) and MEN (P = .012) fibroblasts were higher in the 0% strain groups when compared to the 4% strain groups. The expression levels of MMP-2 (P ≤ .022) and MMP-3 (P ≤ .018) for MEN fibroblasts were higher in the 4% and 10% strain groups when compared to the 0% strain group.
When comparing among tissue types, the concentrations of MMP-3 (P = .032) and ADAMTS4 (P = .004) protein biomarkers for LAB fibroblasts were higher than for MEN fibroblasts at 4% strain.
The gene expression levels of TIMP-1 (P = .03) for LAB fibroblasts were lower than for MEN fibroblasts in the 4% strain group, and the expression levels of TIMP-2 (P = .023) for LAB fibroblasts were lower than for TEN fibroblasts in the 10% strain group. TEN fibroblasts expressed higher levels of MMP-2 (P = .043) and MMP-3 (P = .046) compared to MEN fibroblasts in the 0% strain group.
Inflammatory responses
The concentrations of IL-8 (P = .002) and MCP-1 (P = .033) protein biomarkers for LAB fibroblasts were higher in the 10% strain group compared to the 0% strain group (Figs 3, 4 and Table 6). The concentrations of PGE2 and IL-8 protein biomarkers for MEN fibroblasts were higher in the 10% strain group compared to the 0% (P ≤ .033) and 4% (P ≤ .032) strain groups. There were no statistically significant differences noted for measured inflammatory protein biomarkers among strain groups for TEN fibroblasts.
Table 6.
Inflammatory biomarkers across strain levels.
| Strain group | |||||||
|---|---|---|---|---|---|---|---|
| Biomarker | Cell type | 0% | 4% | 10% | |||
| Mean | SD | Mean | SD | Mean | SD | ||
| PGE2 (pg/resa) | LAB | 2.862 | 0.629 | 41.409 | 73.153 | 52.474 | 83.218 |
| MEN | 2.851 | 0.827 | 2.606 | 0.599 | 11.526 | 13.75 | |
| TEN | 5.431 | 2.413 | 6.179 | 5.17 | 5.984 | 6.625 | |
| IL-8 (pg/resa) | LAB | 5.085 | 3.512 | 114.479 | 208.556 | 1017.686 | 1757.665 |
| MEN | 13.057 | 7.17 | 35.317 | 18.447 | 189.564 | 190.705 | |
| TEN | 18.352 | 13.026 | 39.03 | 32.562 | 173.386 | 237.698 | |
| MCP-1 (pg/resa) | LAB | 24.461 | 2.983 | 61.365 | 48.161 | 196.665 | 250.006 |
| MEN | 21.757 | 5.159 | 18.891 | 3.797 | 48.289 | 57.017 | |
| TEN | 36.196 | 28.032 | 37.024 | 23.787 | 43.17 | 45.757 | |
| KC (pg/resa) | LAB | 0.059 | 0.088 | 0 | 0 | 0 | 0 |
| MEN | 0.109 | 0.121 | 0.112 | 0.104 | 0.092 | 0.097 | |
| TEN | 0.166 | 0.282 | 0.14 | 0.137 | 0.041 | 0.081 | |
| TNF-α (pg/resa) | LAB | 0 | 0 | 0.158 | 0.353 | 0.103 | 0.23 |
| MEN | 0.026 | 0.059 | 0.05 | 0.047 | 0.048 | 0.069 | |
| TEN | 0 | 0 | 0.065 | 0.13 | 0 | 0 | |
There were no statistically significant differences noted for measured inflammatory protein biomarkers between cell types at the strain levels tested.
DISCUSSION
The present study provides important insight as to how meniscus and tendon fibroblast stress responses liken to acetabular labrum fibroblasts in vitro. The authors compared the changes in cell viability, extracellular matrix, degradative, and inflammatory biomarker responses to clinically relevant tensile loading among fibroblasts cultured from acetabular labrum, anterior tibialis tendon, and knee meniscus. Meniscus and labral fibroblasts demonstrated comparable strain-dependent responses in cell viability, collagen expression, inflammatory responses, and collagen breakdown markers. Taken together, these data suggest that meniscus metabolically responds to tensile load in a manner similar to labrum.
Cell viability
This study found a strain-dependent decrease in cell viability in labral and meniscus fibroblasts. The similarities between tendon and meniscus fibroblasts in this study are not surprising given the morphological and physiological similarities between meniscus and labrum [13].
In contrast, the tibialis anterior tendon, which typically receives uniaxial force through its associated muscle belly [22], did not demonstrate changes in cell viability in response to load. This finding is consistent with a previous study assessing tendon-derived stem cells under various tensile load conditions [23]. This resistance may be due protective effects of its tendon’s high levels of Type I collagen, which stabilizes the tendon primarily against tensile forces [23, 24].
Extracellular matrix responses
In this study, labral and meniscus fibroblasts expressed decreased levels of COL I at higher strain levels, while tendon did not exhibit differences in COL I expression. This finding contrasts a study of bovine labrums that failed to detect a change in COL1 expression levels under 10% uniaxial tensile strain [10]. All three tissue types demonstrated increased markers of extracellular matrix (ECM) remodelling at higher strain levels. Multiple meniscal tissue engineering studies have taken advantage of this strain dependent ECM response to promote meniscal repair [14]. One study hypothesized that meniscal response to mechanical load is due to intracellular calcium signalling pathway upregulating production of GAG [25]. Another study hypothesized that labral ECM also exhibited mechanosensitivity through upregulation or downregulation of COMP, a bridging protein, at various load conditions [10]. While the exact mechanisms of mechanosensitivity on ECM changes are not entirely elucidated, labral and meniscus fibrochondrocytes demonstrated analogous changes ECM under supraphysiologic tensile load.
Degradative responses
Meniscus and labral cells demonstrated increased degradative responses at higher strain levels, as evidenced by elevated matrix metalloproteinases. ADAMTS5 in labral and meniscus fibroblasts was higher in the absence of strain. Together, these findings suggest that the meniscus and labrum may be sensitive to both loading and unloading conditions. Notably, labral tissue had higher concentrations of MMP-3 and ADAMTS4, as well as lower expression levels of TIMP-1, than meniscus at physiologic strain. These findings imply labral tissue experiences increased degradation relative to meniscus under physiologic strain. Previous research has demonstrated comparably elevated degradative markers between labrum and meniscus fibroblasts [13]. To the authors’ knowledge, no studies have examined the labrum and meniscus degradative responses with the three tensile load conditions used in this study. One such explanation of this finding is the labrum’s relatively lower level of Type I collagen, a resistor of tensile force, than meniscus [13].
In the unloaded condition, tendon tenocytes had increased presence of matrix metalloproteinases. Conversely, tendon fibroblasts expressed higher levels of TIMP-2, a metalloproteinase inhibitor, than labral fibroblasts under supraphysiologic strain, suggesting a higher tolerance against matrix degradation in response to tensile load. These findings indicate that increased strain has an inhibitory function on tendon fibroblast degradative responses under clinically relevant loads.
Inflammatory responses
Meniscus and labral fibroblasts demonstrated upregulation of inflammatory markers under supraphysiologic strain. Similar patterns of inflammation between these tissues are expected as both tissues are fibrochondrocytes, capable to reacting to and expressing pro-inflammatory stimuli [13]. Dhollander et al. [13] hypothesized these inflammatory capabilities may contribute to hip arthritis in the context of native labral abnormalities. Multiple in vivo studies have demonstrated a differential effect of load on pro- and anti-inflammatory responses of the meniscus [14]. In a rabbit model, passive motion red. On the contrary, tendon tenocytes did not demonstrate a load-dependent inflammatory response. This contrasts with Bayer et al. [15] who reported increased inflammatory in human tendon tissue when tension was removed.
Clinical implications
This study’s findings have potential clinical implications for acetabular labral reconstruction and postoperative rehabilitation protocols. When selecting a graft, its mechanical and biochemical properties should mimic those of native tissue [13]. In the immediate period following labral reconstruction, the mechanical properties of the selected graft type are responsible for maintaining stability of the graft; however, the long-term maintenance of a graft requires that it produce ECM responses that closely resemble native tissue [10]. Labral cells must effectively respond to load in a largely avascular tissue to maintain these functions and preserve joint health [16]. When labral reconstruction is performed, cells retained within and/or repopulating the tissue used for reconstruction must assume these functions. As such, autograft and fresh allograft cells in the fibroblast lineage have direct impact on hip labrum reconstruction success or failure. In this study, supraphysiologic strain (10%) was associated with the most potential detrimental effects across tissue types. These findings have clinical relevance for design of rehabilitation protocols following acetabular labrum resection, repair, and reconstruction, as well as regarding graft choice when labral reconstruction is indicated.
Limitations
This study has several limitations. Firstly, this is an in vitro analysis relying on processing and manipulation of cultured cells, which may differ from the in vivo setting, particularly in the postoperative state. It is unclear how closely the responses of allograft fibroblasts must mimic those of acetabular labrum to represent a clinically relevant difference. Furthermore, this analysis was performed on canine tissues rather than human tissues. Prior research into ligamentous stress responses using a similar model has demonstrated viable translational validity, but it is possible that human tissue responses may differ [19, 26–28]. The number of animals used in this study was modelled on prior translational studies performed at our institution [19, 27, 28]. Furthermore, sample size decision-making was based on the availability of tissues from unrelated terminal studies in concordance with the 3 R’s of ethical research animal use [29]. While statistical significance was attained during this study, post hoc powers did not always exceed 0.8. A larger sample size may have minimized the risk of Type II error. Finally, meniscus tissue typically used for allograft reconstructions in humans is preserved at tissue banks to maintain viability (fresh), whereas tibialis anterior tendons are typically used as fresh-frozen allografts [28]. As such, the results from the present study in which all cell types were recovered from directly recovered tissues may differ from those other processing or preservation methods.
CONCLUSION
The characterization of acetabular labrum cellular responses to clinically relevant loads of force more closely match those of meniscal allograft tissue than tibialis anterior allograft tissue. This has potential implications for labral reconstruction, as meniscal allograft tissues may be more suited than tibialis anterior tissues to withstand the forces necessary for joint stabilization and secondarily chondroprotection within the hip joint. The meniscus and labrum’s analogous roles in joint stability and chondroprotection may make the meniscus a more resilient allograft tissue in the setting of labral reconstruction than tibialis anterior tendon tissue.
ACKNOWLEDGEMENTS
None declared.
Contributor Information
Jarod A Richards, Department of Orthopaedic Surgery, University of Missouri, 1100 Virginia Avenue, Columbia, MO 65212, United States; Department of Orthopaedic Surgery, University of Louisville, 550 S. Jackson St., 1st Floor ACB, Louisville, KY 40202, United States.
Muhammad-Amin H Munshi, Department of Orthopaedic Surgery, University of Missouri, 1100 Virginia Avenue, Columbia, MO 65212, United States.
Aaron M Stoker, Department of Orthopaedic Surgery, University of Missouri, 1100 Virginia Avenue, Columbia, MO 65212, United States; Thompson Laboratory for Regenerative Orthopaedics, University of Missouri, 1100 Virginia Avenue, Columbia, MO 65212, United States.
James L Cook, Department of Orthopaedic Surgery, University of Missouri, 1100 Virginia Avenue, Columbia, MO 65212, United States; Thompson Laboratory for Regenerative Orthopaedics, University of Missouri, 1100 Virginia Avenue, Columbia, MO 65212, United States.
Brett D Crist, Department of Orthopaedic Surgery, University of Missouri, 1100 Virginia Avenue, Columbia, MO 65212, United States.
Steven F DeFroda, Department of Orthopaedic Surgery, University of Missouri, 1100 Virginia Avenue, Columbia, MO 65212, United States.
CONFLICT OF INTEREST
A.M.S. receives IP royalties from Musculoskeletal Transplant Foundation. J.L.C. receives research support from AANA; receives research support from AO Trauma; receives IP royalties, is a paid consultant and receives research support from Arthrex, Inc.; is a paid consultant for Bioventus; is a paid consultant for Boehringer Ingelheim; is a paid consultant and receives research support from Collagen Matrix Inc.; receives research support from GE Healthcare; is on the editorial or governing board for the Journal of Knee Surgery; is a board or committee member for Midwest Transplant Network; is a board or committee member, receives IP royalties and research support from Musculoskeletal Transplant Foundation; receives research support from the National Institutes of Health (NIAMS and NICHD); receives research support from OREF; receives research support from Orthopaedic Trauma Association; receives research support from PCORI; receives research support from Regenosine; receives research support from SITES Medical; receives publishing royalties, financial or material support from Thieme; is a paid consultant for Trupanion; and receives research support from U.S. Department of Defence. B.D.C. is a board or committee member for AO Trauma North America; receives other financial or material support from Arthrex, Inc.; is a paid consultant, paid presenter or speaker for Curvafix; is a paid presenter or speaker for DePuy, A Johnson & Johnson Company; is a board or committee member for Fragility Fracture Network—USA; receives IP royalties from Globus Medical; is a board or committee member for International Geriatric Fracture Society; is on the editorial or governing board for the Journal of Hip Preservation; is on the editorial or governing board for the Journal of Orthopaedic Trauma; is a paid consultant, paid presenter or speaker for KCI; is a board or committee member for Orthopaedic Trauma Association; is an unpaid consultant for Osteocentric; has stock or stock options from RomTech; is on the editorial or governing board from SLACK Incorporated; is a paid consultant and receives research support from Synthes; is an unpaid consultant for Urgo Medical. S.F.D. is a board or committee member for American Orthopaedic Society for Sports Medicine; is a paid presenter or speaker for AO North America; receives research support for Arthrex, Inc.; is on the editorial or governing board of Arthroscopy; is a board or committee member for Arthroscopy Association of North America; and receives publishing royalties, financial or material support Springer. J.A.R. and M.A.H.M. have no conflicts of interest to disclose.
FUNDING
This work was supported by the Thompson Laboratory for Regenerative Orthopaedics.
DATA AVAILABILITY
The data underlying this article will be shared on reasonable request to the corresponding author.
REFERENCES
- 1. Nepple JJ, Philippon MJ, Campbell KJ et al. The hip fluid seal—part II: the effect of an acetabular labral tear, repair, resection, and reconstruction on hip stability to distraction. Knee Surg Sports Traumatol Arthrosc 2014;22:730–6. 10.1007/s00167-014-2875-y [DOI] [PubMed] [Google Scholar]
- 2. Philippon MJ, Nepple JJ, Campbell KJ et al. The hip fluid seal—part I: the effect of an acetabular labral tear, repair, resection, and reconstruction on hip fluid pressurization. Knee Surg Sports Traumatol Arthrosc 2014;22:722–9. 10.1007/s00167-014-2874-z [DOI] [PubMed] [Google Scholar]
- 3. Storaci HW, Utsunomiya H, Kemler BR et al. The hip suction seal, part I: the role of acetabular labral height on hip distractive stability. Am J Sports Med 2020;48:2726–32. 10.1177/0363546520941855 [DOI] [PubMed] [Google Scholar]
- 4. McCarthy JC, Noble PC, Schuck MR et al. The role of labral lesions to development of early degenerative hip disease. Clin Orthop Relat Res 2001;393:25–37. 10.1097/00003086-200112000-00004 [DOI] [PubMed] [Google Scholar]
- 5. Ganz R, Leunig M, Leunig-Ganz K et al. The etiology of osteoarthritis of the hip: an integrated mechanical concept. Clin Orthop 2008;466:264–72. 10.1007/s11999-007-0060-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. White BJ, Herzog MM. Labral reconstruction: when to perform and how. Front Surg 2015;2:27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Lee S, Wuerz TH, Shewman E et al. Labral reconstruction with iliotibial band autografts and semitendinosus allografts improves hip joint contact area and contact pressure: an In vitro analysis. Am J Sports Med 2015;43:98–104. 10.1177/0363546514553089 [DOI] [PubMed] [Google Scholar]
- 8. DeFroda SF, Hanish S, Muhammad M et al. Graft options for hip labral reconstruction. JBJS Rev 2022;10:e22.00124. 10.2106/JBJS.RVW.22.00124 [DOI] [PubMed] [Google Scholar]
- 9. Rosenthal RM, Featherall J, Parkes CW et al. Acetabular labral reconstruction: review of techniques and outcomes. Curr Rev Musculoskelet Med 2023;16:470–9. 10.1007/s12178-023-09856-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Huber S, Günther S, Cambria E et al. Physiological stretching induces a differential extracellular matrix gene expression response in acetabular labrum cells. Eur Cell Mater 2022;44:90–100. 10.22203/eCM.v044a06 [DOI] [PubMed] [Google Scholar]
- 11. Philippon MJ, Arner JW, Crawford MD et al. Acetabular labral reconstruction with iliotibial band autograft: outcome and survivorship at a minimum 10-year follow-up. J Bone Joint Surg Am 2020;102:1581–7. 10.2106/JBJS.19.01499 [DOI] [PubMed] [Google Scholar]
- 12. Markes AR, Hodax JD, Ma CB. Meniscus form and function. Clin Sports Med 2020;39:1–12. 10.1016/j.csm.2019.08.007 [DOI] [PubMed] [Google Scholar]
- 13. Dhollander AAM, Lambrecht S, Verdonk PCM et al. First insights into human acetabular labrum cell metabolism. Osteoarthr Cartil 2012;20:670–7. 10.1016/j.joca.2012.03.023 [DOI] [PubMed] [Google Scholar]
- 14. McNulty AL, Guilak F. Mechanobiology of the meniscus. J Biomech 2015;48:1469–78. 10.1016/j.jbiomech.2015.02.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Bayer ML, Schjerling P, Herchenhan A et al. Release of tensile strain on engineered human tendon tissue disturbs cell adhesions, changes matrix architecture, and induces an inflammatory phenotype. PLoS One 2014;9:e86078. 10.1371/journal.pone.0086078 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Dy CJ, Thompson MT, Crawford MJ et al. Tensile strain in the anterior part of the acetabular labrum during provocative maneuvering of the normal hip. J Bone Jt Surg-Am Vol 2008;90:1464–72. 10.2106/JBJS.G.00467 [DOI] [PubMed] [Google Scholar]
- 17. Kanazawa T, Furumatsu T, Matsumoto-Ogawa E et al. Role of Rho small GTPases in meniscus cells. J Orthop Res 2014;32:1479–86. 10.1002/jor.22703 [DOI] [PubMed] [Google Scholar]
- 18. Kawamura Y, Tetsunaga T, Yamada K et al. Mechanical stretching induces calcification and cartilage matrix metabolism, causing degeneration of the acetabular labrum. HIP Int 2023;33:500–7. 10.1177/11207000211044675 [DOI] [PubMed] [Google Scholar]
- 19. Breshears LA, Cook JL, Stoker AM et al. The effect of uniaxial cyclic tensile load on gene expression in canine cranial cruciate ligamentocytes. Vet Surg 2010;39:433–43. 10.1111/j.1532-950X.2010.00679.x [DOI] [PubMed] [Google Scholar]
- 20. Bc G, Am S, K K et al. Using animal models in osteoarthritis biomarker research. J Knee Surg 2011;24:251–64. 10.1055/s-0031-1297361 [DOI] [PubMed] [Google Scholar]
- 21. Farndale RW, Buttle DJ, Barrett AJ. Improved quantitation and discrimination of sulphated glycosaminoglycans by use of dimethylmethylene blue. Biochim Biophys Acta 1986;883:173–7. 10.1016/0304-4165(86)90306-5 [DOI] [PubMed] [Google Scholar]
- 22. Chen J, Kadakia R, Akoh CC et al. Management of anterior tibialis tendon ruptures. J Am Acad Orthop Surg 2021;29:691–701. 10.5435/JAAOS-D-20-00802 [DOI] [PubMed] [Google Scholar]
- 23. Xu Y, Wang Q, Li Y et al. Cyclic tensile strain induces tenogenic differentiation of tendon-derived stem cells in bioreactor culture. Biomed Res Int 2015;2015:1–13. 10.1155/2015/790804 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Docking S, Samiric T, Scase E et al. Relationship between compressive loading and ECM changes in tendons. Muscles Ligaments Tendons J 2013;3:7–11. 10.11138/mltj/2013.3.1.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Eifler RL, Blough ER, Dehlin JM et al. Oscillatory fluid flow regulates glycosaminoglycan production via an intracellular calcium pathway in meniscal cells. J Orthop Res 2006;24:375–84. 10.1002/jor.20028 [DOI] [PubMed] [Google Scholar]
- 26. Cardona-Ramirez S, Stoker AM, Cook JL et al. Fibroblasts from common anterior cruciate ligament tendon grafts exhibit different biologic responses to mechanical strain. Am J Sports Med 2021;49:215–25. 10.1177/0363546520971852 [DOI] [PubMed] [Google Scholar]
- 27. Cook AE, Stoker AM, Leary EV et al. Metabolic responses of meniscal explants to injury and inflammation ex vivo. J Orthop Res 2018;36:2657–63. 10.1002/jor.24045 [DOI] [PubMed] [Google Scholar]
- 28. Schreiner AJ, Stannard JP, Cook CR et al. Comparison of meniscal allograft transplantation techniques using a preclinical canine model. J Orthop Res 2021;39:154–64. 10.1002/jor.24668 [DOI] [PubMed] [Google Scholar]
- 29. Russell WMS, Burch RL. The Principles of Humane Experimental Technique. London: Metheun, 1959. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data underlying this article will be shared on reasonable request to the corresponding author.