Dr. Sandell owns stock or stock options in ISTO Technologies and receives royalties from Merck/Millipore for a type IIA collagen N-propeptide enzyme-linked immunosorbent assay.
Robert H. Brophy
Washington University School of Medicine at Barnes–Jewish Hospital, St. Louis, Missouri
Department of Orthopaedic Surgery, Washington University School of Medicine at Barnes–Jewish Hospital, 14532 South Outer Forty Drive, Chesterfield, MO 63017. E-mail: email@example.com Dr. Brophy has served as an expert witness on behalf of Genzyme.
The content of this publication is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Arthritis and Musculoskeletal and Skin Diseases or the National Institutes of Health.
Meniscus tears are associated with a heightened risk of osteoarthritis. This study aimed to advance our understanding of the metabolic state of injured human meniscus at the time of arthroscopic partial meniscectomy through transcriptome-wide analysis of gene expression in relation to the patient's age and degree of cartilage chondrosis.
The degree of chondrosis of knee cartilage was recorded at the time of meniscectomy in symptomatic patients without radiographic osteoarthritis. RNA preparations from resected menisci (n = 12) were subjected to transcriptome-wide microarray and QuantiGene Plex analyses. Variations in the relative changes in gene expression with age and chondrosis were analyzed, and integrated biologic processes were investigated computationally.
We identified a set of genes in torn menisci that were differentially expressed with age and chondrosis. There were 866 genes that were differentially regulated (≥1.5-fold difference and P < 0.05) with age and 49 with chondrosis. In older patients, genes associated with cartilage and skeletal development and extracellular matrix synthesis were repressed, while those involved in immune response, inflammation, cell cycle, and cellular proliferation were stimulated. With chondrosis, genes representing cell catabolism (cAMP catabolic process) and tissue and endothelial cell development were repressed, and those involved in T cell differentiation and apoptosis were elevated.
Differences in age-related gene expression suggest that in older adults, meniscal cells might dedifferentiate and initiate a proliferative phenotype. Conversely, meniscal cells in younger patients appear to respond to injury, but they maintain the differentiated phenotype. Definitive molecular signatures identified in damaged meniscus could be segregated largely with age and, to a lesser extent, with chondrosis.
The meniscus is an essential component of the tibiofemoral articulation that contributes to the complex biomechanics of the knee joint, protecting the underlying articular cartilage. The role of the meniscus in the initiation and progression of osteoarthritis (OA) is thought to be minimal unless it is injured ([1, 2]). The location of the meniscus and its role in shock absorption and load transmission predispose the meniscus to traumatic and degenerative injuries; thus, meniscus tears are one of the most common intraarticular injuries of the knee ([3, 4]). There is an age bias in terms of the mechanisms of meniscal injury (). Younger people are more likely to have acute tears due to trauma, whereas older people are more likely to have tears due to degeneration ([6-9]). Since the majority of meniscus tears are unsuitable for repair ([3, 10, 11]), partial meniscectomy is used to redress most meniscus tears ([7, 12]).
A large body of information suggests that damage to or loss of the meniscus is associated with changes in cartilage ([13-16]). It has also been reported that following partial meniscectomy, there is an elevated risk of functional deterioration of cartilage ([11, 17, 18]), often resulting in the development of OA and irreparable joint damage ([2, 10, 14, 17, 19-21]). Despite the fact that the meniscus and articular cartilage of the knee have several characteristics in common and the mechanisms of meniscus and knee cartilage degeneration have been demonstrated ([22, 23]), the molecular pathogenesis of traumatic and degenerative meniscus tears and their association with degenerative changes in the articular cartilage are not known.
We have previously shown, through a candidate gene approach, that the expression of various OA- and obesity-related genes diverged mainly with age ([24, 25]) and to some extent with the body mass index (BMI) () in the injured meniscus. As these studies provided information on only a handful of genes (known for their specific role in OA and obesity), we sought to determine the transcriptome-wide gene expression profile in meniscus tears in patients undergoing arthroscopic partial meniscectomy in order to identify a comprehensive set of targets for future studies. The goal was to identify changes in the gene expression profile of the torn meniscus that may be relevant to age-associated meniscal and cartilage degeneration. We believe that this information will allow for a better understanding of why and how degenerative changes occur in the meniscus and how injury to the meniscus, whether traumatic or degenerative, contributes to the initiation and progression of OA in the knee.
PATIENTS AND METHODS
Patients and meniscus samples
The study protocol was approved by the Institutional Review Board of Washington University and complied with the Health Insurance Portability and Accountability Act. All subjects furnished written informed consent to allow the use of their tissues for research purposes. The dysfunctional fragment of the human injured meniscus resected from patients with symptomatic, nonrepairable tears during arthroscopic partial meniscectomy was used for gene expression analysis.
Characteristics of the study population
The characteristics of the study population are presented in Table 1. Human meniscal tissues were collected from 12 patients (6 ≤40 years and 6 >40 years; 5 with chondrosis and 7 without chondrosis) at the time of arthroscopic meniscal resection. The average age of patients in the younger study group (age ≤40 years) was 31.3 years (range 16−40 years), and the average age of those in the older study group (age >40 years) was 50.0 years (range 43−55 years). Most injuries were complex tears of the posterior horn of the medial meniscus, with surgery occurring an estimated 1–12 months after injury. Only 1 patient had a concomitant ligament injury, which was a chronic anterior cruciate ligament tear. In patients without chondrosis, no cartilage changes (grade 0) were observed in any of the 3 knee compartments. In patients with chondrosis, however, cartilage damage (grade ≥2) was observed in at least 1 compartment of the knee. The mean age of patients with chondrosis was 48.40 years (range 40–55 years), and the mean age of those without chondrosis was 35.14 years (range 16–48 years).
Table 1. Clinical description of meniscus tears, chondrosis, and RNA quality in the 12 study patients*
Type of tear
Location of tear
Time from injury (estimated)
RIN = RNA integrity number; ACL = anterior cruciate ligament.
Arthroscopic assessment and grading of chondral changes
At the time of arthroscopy, changes in the cartilage were recorded by the operating surgeon (RHB; fellowship trained in academic orthopedic sports medicine), assessing 3 compartments of the knee: same compartment from which the injured meniscus was resected, the other weight-bearing compartment, and the patellofemoral compartment. The chondrosis was graded according to a modified Outerbridge scoring system (). For the purpose of this analysis, we simplified the chondrosis grade to a binomial variable, assigning a value of 1 if any compartment had grade 2 or greater chondrosis and assigning a value of 0 if no compartment had grade 2 or greater chondrosis.
Preparation of RNA from resected meniscus
Meniscus tissues received in the laboratory were processed and used for RNA isolation as has been described previously (). The total RNA concentration and purity were determined by use of an Agilent 2100 Bioanalyzer according to the manufacturer's instructions. The RNA integrity number for each sample is shown in Table 1. RNA samples were sent to the Washington University Genome Technology Access Center (GTAC) for microarray and QuantiGene Plex assays.
Transcriptome analysis was done at the GTAC according to a standard protocol. Briefly, 100 ng of RNA transcripts was amplified by T7 linear amplification using an ABI/Ambion MessageAmp TotalPrep amplification kit (Life Technologies). A total of 750 μg of amplified RNA was applied to HumanHT-12 v4 Expression BeadChips (Illumina) and hybridized at 58°C for 16–20 hours with high humidity. Immobilized, biotinylated amplified RNAs were then detected by staining with Cy3–streptavidin (1 μg of Cy3–streptavidin per 1 ml of Illumina Block E1) for 10 minutes at room temperature. Arrays were washed, dried, and scanned in an Illumina BeadArray Reader according to the recommended protocols. Images were quantified with Illumina BeadScan software version 3.0, and the data were imported into Illumina GenomeStudio data analysis software system. Finally, on-slide spot replicates were averaged with the use of GenomeStudio, and an individual spot probe was reported.
Comparisons and calculations of results
The gene expression differences were compared between young (≤40 years) and old (>40 years) patients and between patients with and without concomitant chondrosis. We used the age cutoff of 40 years based on our previous studies in which we found significant differences in gene expression between the age groups of >40 years and ≤40 years ([24, 25]).
To identify differentially expressed genes in injured menisci, a one-way analysis of variance (ANOVA) model was performed on the probe sets using Partek Genomics Suite software version 6.6. To control the false-positive response rate, a false discovery rate (FDR) was used to adjust the P value. Genes were considered differentially expressed if their probes passed the following tests: minimum expression, FDR, and fold change. An FDR of 0.05 and statistical significance at P < 0.05 were used. An arbitrary cutoff of absolute fold change ≥1.5 was applied to narrow the number of differentially regulated genes.
Hierarchical clustering and Venn diagrams
Hierarchical clustering was done with Partek Genomics Suite software version 6.6, using the full-featured hierarchical clustering option on genes that were significantly up- or down-regulated with age as well as in the presence or absence of chondrosis. This analysis allowed for dual clustering of genes and samples, interactive branch flipping, and other advanced features for clustering and coloring the resulting dendrograms and heat maps. Similarly, Venn diagrams were created with the Partek list manager to show the number of genes that were differentially up- or down-regulated with age and chondrosis and their potential overlap.
Validation of selected genes by QuantiGene Plex assay
To confirm the consistency and reliability of the microarray (transcriptome) data, we selected 45 target genes and 3 housekeeping genes for validation. First, we selected the top 15 genes that were up-regulated with age and the top 15 genes that were down-regulated with age. Additionally, we handpicked another 15 genes known to be important to cartilage homeostasis and OA. Validation of gene expression was done by QuantiGene Plex assay (Panomics), which has proven to be a reliable tool with which to validate microarray data (). This is an unbiased assay that is based on branched-chain DNA technology and provides a novel approach for gene expression analysis by analyzing the reporter gene (signal amplification), rather than a target amplification as achieved by polymerase chain reaction ([28, 29]). A list of genes validated by QuantiGene Plex assay with some of their features is provided in Supplementary Table 1 (available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/doi/10.1002/art.37984/abstract). Quantification of RNA was carried out in biologic and technical replicates using a QuantiGene Plex 2.0 assay kit. Individual bead-based oligonucleotide probe sets for each target were developed by Panomics (plex set no. 312184). All steps were performed according to the protocols supplied by GTAC, as described elsewhere ().
Biologic processes and functional network analyses for differentially expressed genes
To determine the biologic significance of differentially expressed genes, we analyzed gene sets by GeneGo MetaCore pathway analysis software to analyze the gene ontology. The altered cellular and biologic processes (gene ontology distribution) were ranked based upon significant −logP values. GeneGo MetaCore software was also used to investigate functional and molecular networks that were overrepresented by the differentially expressed genes. In addition, the gene expression data were used to generate functional and molecular networks by using GeneGo MetaCore pathways analysis software. Genes that were differentially expressed with age and chondrosis were subjected to gene networking to examine their association with each other in canonical biologic pathways.
Findings of the transcriptome analysis
We analyzed changes in gene expression in the transcriptome-wide gene expression profiles of the meniscus tears. A principal components analysis was performed to see how the samples segregated in a 3-dimensional space based on patient's age and the presence or absence of chondrosis (results not shown). The samples from younger patients clustered together, and those from older patients clustered together. Likewise, the samples also segregated according to the presence or absence of chondrosis. This segregation suggested that there would be molecular differences between the samples based on age and chondrosis.
We next identified differentially expressed genes by one-way ANOVA. Of the 47,432 transcripts detectable with the Illumina HumanHT-12 v4 BeadChips kit, 4,730 transcripts (∼10%) were significantly (P < 0.05) and differentially up- or down-regulated with age and 1,861 (4%) with chondrosis. Hierarchical clustering analyses of significantly differentially regulated genes in the torn meniscus from young versus old patients and from patients with versus those without chondrosis are shown in Figures 1A and B, respectively. Hierarchical clustering also indicated the molecular differences between meniscal samples based on age or chondrosis.
Findings of the analysis of differentially expressed genes that vary with age
Differentially expressed genes
We identified a total of 4,730 genes in the injured meniscus that were significantly differentially expressed between patients who were ≤40 years and those who were >40 years of age. Among these, 2,401 genes were significantly up-regulated and 2,329 genes were significantly down-regulated with age. Because of the large number of genes, we set an arbitrary cutoff of ≥1.5-fold to narrow the number of differentially expressed genes. This filtering criterion allowed us to limit the number of differentially expressed genes to 1,006; following removal of the nonannotated and duplicate genes, the number of differentially expressed genes dropped to 866. Among these 866 genes, 373 (Supplementary Table 2) were up-regulated with age and 493 (Supplementary Table 3) were down-regulated with age (Supplementary Tables 2 and 3 are available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/doi/10.1002/art.37984/abstract).
Some of the important differentially expressed genes (top 15 up- or down-regulated plus 15 other genes selected for their significance in meniscus and cartilage metabolism and OA), and their respective QuantiGene Plex fold-change validation, are shown in Table 2. The results of the QuantiGene Plex assay showed a similar fold change for genes that were down-regulated with age and a somewhat lower fold change for those that were up-regulated with age; however, the overall trend remained the same between the microarray and the QuantiGene Plex assays.
Table 2. Genes differentially regulated with age and with chondrosis*
QGP assay, fold change
Genes that were up- or down-regulated with chondrosis were not validated by QuantiGene Plex (QGP) assay. NA = not applicable (no record found).
Down-regulated with age (top 15)
Collagen, type II, α1
Collagen, type XI, α2
Collagen, type IX, α2
C-type lectin domain family 3, member A
Fibroblast growth factor receptor 3
SHC (Src homology 2 domain–containing) family, member 4
Cartilage intermediate layer protein 2
SRY (sex determining region Y)–box 8
S100 calcium binding protein A1
S100 calcium binding protein A13
Additional selected candidate genes
Hyaluronan and proteoglycan link protein 1
SRY (sex determining region Y)–box 9
Vascular endothelial growth factor A
Sprouty homolog 4
Fibroblast growth factor 18
Wingless-type MMTV integration site family, member 16
Up-regulated with age (top 15)
Major histocompatibility complex, class II, DRβ5
Major histocompatibility complex, class II, DRβ1
Actin, γ2, smooth muscle, enteric
Ribosomal protein L14
Fc fragment of IgE, high-affinity I, receptor for; α polypeptide
Preferentially expressed antigen of melanoma family member 10
Rho GTPase–activating protein 8
Chromosome 21 open-reading frame 100
Dapper, antagonist of β-catenin, homolog 2 (Xenopus laevis)
Transmembrane 4 L6 family member 5
Ras-related associated with diabetes
Breakpoint cluster region pseudogene 2
Solute carrier family 22, member 10
Kruppel-like factor 14
Baculoviral inhibitor of apoptosis repeat–containing 5
Mesenchyme homeobox 1
Tubby-like protein 2
Period homolog 2 (Drosophila)
Ribosomal protein L10 pseudogene
Up-regulated with chondrosis
Olfactory receptor, family 52, subfamily I, member 2
Mal, T cell differentiation protein
Small nucleolar RNA, C/D box 89
Olfactory receptor, family 7, subfamily C, member 1
Ets variant 3–like
Golgin A6 family, member B
DEAD/H (Asp-Glu-Ala-Asp/His)–box helicase 11
Solute carrier family 5 (sodium/glucose cotransporter), member 12
Integrin β1 binding protein 1
Hepatocyte nuclear factor 1 homeobox B
Sphingosine 1-phosphate receptor 5
Myosin, light chain 4, alkali; atrial, embryonic
Tubulin, β1 class VI
Naked cuticle homolog 2 (Drosophila)
Ran binding protein 2–like and GRIP domain–containing 5
Ikaros family zinc-finger 3 (Aiolos)
Carbonic anhydrase I
Biologic processes and pathway analyses
Genes that were up-regulated ≥1.5-fold with age or were down-regulated ≥1.5-fold with age were subjected to gene ontology analysis to determine the biologic processes they represent. Figures 2A and B show the top 20 ontologies related to biologic processes, based on significant –logP values for genes found to be up- or down-regulated with age. The biologic processes that appeared to be repressed with age included skeletal development, cartilage development, and cartilage extracellular matrix synthesis (Figure 2A). Genes representing these processes mainly included COL2A1, CHAD, COL11A2, COL9A2, CLEC3A, SCIN, FGFR3, SHC4, CILP2, SOX8, SBSPON, S100A1, GLDN, MT1E, S100A13, HAPLN1, LEPREL1, SOX9, VEGFA, ACAN, SPRY4, SULF2, FGF18, and WNT16. Conversely, the biologic processes that were elevated (Figure 2B) with age included cell cycle and cell division (PCNA, DLGAP5, and CDC25C), and immune response and inflammation (HLA–DRB5, HLA–DRB1, IL7R, CX3CR1, and CCL8) pathways.
Pathway analysis showed that genes that were down-regulated with age demonstrated significant interactions with each other, suggesting that these interactions act in concert to down-regulate the extracellular matrix genes (Figure 3A). In addition, genes that were up-regulated with age were also intercorrelated, but for the sake of clarity, this group of processes was divided into 2 components: one component mainly depicting the networking and intercorrelation of genes associated with cellular proliferation and inflammation (Figure 3B), and the other component depicting the interaction between genes that lead to the up-regulation of the immune response (Figure 3C).
These results suggest that the differential regulation of these genes is not a random occurrence, but could have important biologic consequences. For example, genes down-regulated with age ranged from transcription factors (SOX9) in the nucleus with direct links to genes involved in extracellular matrix development (COL2A1, ACAN, etc.) that were also down-regulated, implicating an interlinked degenerative mechanism and loss of potential repair genes in meniscal cells.
Findings of the analysis of differentially expressed genes that vary with chondrosis
Differentially expressed genes
A total of 1,861 genes (of 47,432 transcripts examined) showed significant differential expression (P < 0.05) with chondrosis. As stated above, using a minimal fold change of ≥1.5 as an additional criterion, only 50 genes (49 after removing 1 nonannotated gene) were identified as having significant differential expression with chondrosis. Among these 49 genes, 26 were down-regulated (25 after removing a duplicate gene) in patients with chondrosis and 23 were up-regulated in patients with chondrosis. All differentially regulated genes, along with their fold changes, are listed in Table 2.
Gene ontology analysis demonstrated that with chondrosis, cellular catabolism including cAMP and cyclic nucleotide catabolic processes were repressed and biologic processes representing cell differentiation and cell apoptosis were stimulated. The top 20 biologic processes in each category are shown in Figures 2C and D, respectively. The pathway analysis showed no direct gene interactions, suggesting that the genes belonged to disparate pathways that did not overlap. Moreover, the variance in patient age and degree of chondrosis also implies that overlap would have been unlikely.
Findings of the analysis of genes commonly regulated with age and chondrosis
As cartilage damage is correlated with age ([31, 32]), we sought to determine whether we could identify genes that changed with both age and chondrosis. As shown in the Venn diagram in Figure 1C, we found 38 genes in common for age and chondrosis: 15 were up-regulated and 23 were down-regulated by the combined effect of age and chondrosis (Supplementary Table 4, available on the Arthritis & Rheumatism web site at http://onlinelibrary.wiley.com/doi/10.1002/art.37984/abstract). However, there were no common genes for age and chondrosis at a ≥1.5-fold difference (Figure 1D). These findings suggest that separate sets of genes and mechanisms are involved in age- and chondrosis-related molecular changes in meniscal tissues.
In this study, we identified a larger number of differentially expressed genes in injured human meniscus tissues that diverged mainly with age and to some extent with the presence of chondrosis of the articular cartilage. This divergence was observed for individual genes, as well as for sets of coordinately expressed genes, which provides a number of possible pathways of interest in meniscus tissues derived from an aging population. Since aging affects both meniscus tears and OA ([9, 33-35]), we and other investigators have shown age-related differences in gene expression in OA for diverse orthopedic tissues, including cartilage ([34, 36]), injured meniscus ([24, 25]), cartilage and meniscus combined (), and all knee joint tissues (). Taken together, these studies have demonstrated that the degenerative processes in joint tissues are accelerated with age. In addition to differences in gene expression with age, the severity of OA (defined by the degree of chondrosis) also increases with age ([4, 37]). It has been reported that meniscus tears are associated with changes in cartilage ([13, 14]) that are thought to accelerate OA progression ([2, 10, 14, 17, 19-21, 33]). The findings of our study show that the metabolic state of the torn meniscus is associated mainly with age and less with the degree of chondrosis characterized arthroscopically (i.e., early degenerative changes in cartilage).
To the best of our knowledge, this study is the first to report unbiased microarray analysis data for global gene expression patterns in human meniscus tears. There are, however, 2 studies that have shown gene expression changes in harvested meniscal cells (i.e., not meniscal tissues) through microarray analysis. In one study (), the investigators compared gene expression signatures in normal meniscal cells to those in OA meniscal cells and found that genes associated with inflammation and cytokine production were up-regulated in OA meniscal cells, whereas genes representing DNA repair processes were down-regulated in OA meniscal cells. In the other study (), the investigators compared a handful of genes to determine associations of gene expression with pathologic deposition of calcium in the meniscus. This study demonstrated the up-regulation of genes associated with biomineralization (e.g., ENPP1 and ANKH) in meniscal cells from OA patients. This finding indicates that menisci from OA patients are more prone to pathologic calcification than are menisci from non-OA patients.
Our study clearly demonstrates elevated expression of genes related to immune response, inflammation, cell cycle, and cellular proliferation in the injured menisci from older patients (>40 years). In addition, our study shows that biologic processes such as skeletal development, cartilage development, and cartilage extracellular matrix development are repressed in older patients. The loss of extracellular matrix genes indicates a decreased ability to repair the injured meniscus and an increase in degenerative changes in the injured meniscus. Both of these findings indicate that the torn meniscus is metabolically active in young and old patients, but with an age-dependent molecular degenerative state. Therefore, a combination of both the metabolic activity of the injured meniscus and the loss of extracellular matrix synthesis could drive meniscus-associated cartilage damage, as the end result of all of these molecular changes may lead to early cartilage damage and the initiation of OA ([2, 10, 14, 17, 19-21, 33]).
Our findings on the down-regulation of extracellular matrix genes and the up-regulation of inflammation and proliferation genes with age are very relevant to the clinical picture seen in arthritic tissues, where loss of extracellular matrix gene expression is observed in cartilage degeneration ([35, 40]). These findings suggest that there are decelerated anabolic and accelerated catabolic (molecular) activities taking place in the injured meniscus with age. As with age, important biologic processes were detected with chondrosis, which included cellular catabolism, cell differentiation, and apoptosis, although the number of genes was much lower and the pathways affected were more diverse. The variation in gene expression by age and chondrosis taken together indicates a dysregulation of cell cycle and cell metabolism with age.
A recent study () showed a zonally dependent degenerative phenotype (characterized by collagenolysis and aggrecanolysis) of the meniscus after exogenous induction of proinflammatory cytokines. Thus, we also suspect that decreased expression of extracellular matrix genes in the injured meniscus, as seen in this study, might contribute to meniscal degeneration. Furthermore, the finding that inflammation-related genes were up-regulated with age is consistent with the findings of studies documenting the role of intermittent inflammation in OA ([42-46]). Our current study also demonstrates the loss of extracellular matrix gene expression and the increased involvement of the immune system and a chronic low-grade proinflammatory state in the injured meniscus with age. These two aspects of aging meniscus tears are very conducive to exacerbating degenerative-type meniscus tears.
Previous studies have demonstrated a positive link between the severity of cartilage degeneration and the degree of meniscus degeneration ([1, 38, 47, 48]), and a few studies have reported that meniscus degeneration precedes cartilage degeneration ([15, 16]). In light of these studies, it is puzzling that our study did not show any overlap between differentially regulated genes when age, as opposed to chondrosis (presence or absence), was used as a variable. One reason for the lack of this overlap could be that the transcriptome analysis is a molecular analysis of the injured meniscus tissues and not of the cartilage upon which the observation of chondrosis is based. Furthermore, the torn meniscus tissues were harvested at the time of meniscectomy, which is definitely a preradiographic phase of OA and perhaps an early phase of cartilage degeneration. A molecular analysis of the cartilage specimens taken at the time of meniscectomy or a followup study on the chondrosis in the same patients may allow for more molecular overlaps between meniscus tears associated with aging and the presence or absence of chondrosis. Finally, in our study of 12 patients, only 3 of those who did not have chondrosis were over the age of 40 years, and this did not allow for statistically significant associations between aging and chondrosis in relation to meniscus tears.
Considering our findings on the differences in age-related gene expression in injured menisci, we propose that the molecular events in the injured meniscus could be critical indicators of future degenerative changes in the cartilage. Age-related loss of extracellular matrix genes and gain in inflammation-related genes synergistically make the meniscus susceptible to degenerative tears and the knee cartilage more susceptible to damage from altered joint mechanics due to an impaired ability of chondrocytes to synthesize matrix (). Furthermore, the degenerative meniscus itself likely results in altered joint mechanics. In our previous studies ([24, 25]), we observed that the expression of various cytokines, chemokines, and enzyme mediators was suppressed with age and the expression of matrix genes was stimulated with age. However, except for MMP1, none of the genes was found to show a statistically significant divergence from our current findings. While it is not possible to explain these differences in MMP1 expression at this point, it is, however, believed that its expression is decreased with an advanced stage of OA and a higher degree of cartilage damage ().
Based on our current findings and those gleaned from the report by Pennock et al (), we propose a working model (Figure 4) that highlights how aging accelerates meniscal and cartilage degeneration and metabolic dysregulation with elevated levels of inflammation-related genes, resulting in the OA phenotype of the joint. In addition, the decreased expression of extracellular matrix genes with age indicates a failure of the repair potential. These findings suggest that in older patients, there is an increased risk of meniscal degeneration and tearing, which fuels the cartilage matrix degradation, accelerates the dedifferentiated phenotype, and in conjunction with inflammation, leads to development of OA. In younger individuals, in contrast, matrix synthesis is not compromised, which likely maintains the cartilage phenotype. This model further warrants the investigation of complex molecular events in the meniscus and cartilage as they relate to the overall joint health.
A limitation of our study is that we lack data on gene expression in normal (uninjured) meniscus tissues for comparison with the injured meniscus tissues and their potential relationship to age and chondrosis. While it would be informational to have data on gene expression from normal tissues, such information is not essential in our analysis for a few reasons. First, our current study was focused solely on meniscus tears, with no intention of making comparisons with uninjured meniscus tissues, but rather to identify age- and chondrosis-related differences in gene expression in injured menisci. We are interested in individuals who have meniscus tears to monitor the impact of age and chondrosis on gene expression since this is more relevant clinically in terms of the prognosis for these patients. Second, given the heightened risk of developing OA in patients with meniscus tears, our analysis could help to stratify this risk independently of the potential relationship of gene expression in the torn meniscus to gene expression in the normal meniscus. Third, age and chondrosis may have similar or dissimilar effects on gene expression in the injured versus uninjured meniscus, but that question was not the focus of our present investigation. Considering that measuring gene expression in the normal meniscus is neither paramount nor practical, the relationship, if any, between gene expression in normal versus torn meniscus tissue is interesting, but not integral to the current investigation. Furthermore, it would be of great value in the future to have histology samples of the injured meniscus to correlate morphology with gene expression.
In summary, our work identified definitive molecular signatures in damaged meniscus that could be segregated based on age and, to a lesser extent, on the degree of chondrosis in the knee. Although this is a complex interaction, a substantial number of genes found to be differentially down-regulated are known to have critical roles in cartilage extracellular matrix synthesis and maintenance, and genes that are differentially up-regulated are known for their roles in inflammation and proliferation. These findings suggest an emerging paradigm, that the metabolic state of the injured meniscus is dependent on the age, but not on chondrosis, of the articular cartilage. Taken together, our transcriptome analysis reveals that a torn meniscus has a molecular footprint that resembles OA in individuals over the age of 40 years and a healing potential in individuals ages 40 years and younger. These studies may provide a molecular rationale for the cause or frequency of degenerative meniscus tears in the aging population and the initiation and propagation of degenerative changes in the cartilage. Therefore, further investigation of metabolic activity in the injured meniscus and its relationship to cartilage degeneration and the early development of knee OA may identify new therapies by which to prevent or limit the contribution of meniscal degeneration to the initiation and progression of OA.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Brophy had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Rai, Sandell, Brophy.
Acquisition of data. Rai, Patra, Brophy.
Analysis and interpretation of data. Rai, Patra, Sandell, Brophy.
We thank staff members at the Washington University Genome Technology Access Center for their help with the transcriptome and QuantiGene Plex assays.