Medial Osteochondral Defect Drives Matrix and Cell Pathology in Compartment-Matched Meniscus

Abstract

Patients with cartilage defects often experience increased meniscal degeneration. It remains unclear whether meniscal damage occurs concurrently with cartilage injury or due to later joint pathology. Limited data exists on how isolated cartilage injuries affect meniscal structure and degeneration. In osteoarthritis models, alterations to the structure and composition of meniscal ECM components have been observed, including meniscus hypertrophy characterized by excessive glycosaminoglycan deposition and fibrochondrocyte rounding. Although proteoglycan deposition increases in early OA, the timing of GAG changes relative to collagen disruption remains unclear. This study examined the correlation between changes in local proteoglycan deposition, cell morphology, and the collagen network in the meniscus following cartilage damage using an in vivo rabbit model. A medial osteochondral defect was created on the femoral condyle of New Zealand white male rabbits, and menisci were harvested 12 weeks later. Our results indicate that a medial osteochondral defect drives pathology in the underlying meniscus, likely due to altered loading conditions. The medial menisci of defect joints exhibited increased proteoglycan deposition and hypertrophy, with increased cell roundness and area in regions of elevated GAGs. Local collagen architecture showed increased fiber diameter in the medial menisci of defect joints, which positively correlated with increased GAG coverage. Abnormal collagen structures were observed, including wider variations in fiber diameters and areas of small fibers with low second harmonic generation signals, indicating poorly organized collagen. A deeper understanding of GAG regulation and fibrochondrocyte pathology in injured meniscus tissue could aid in the development of therapeutics and inform disease progression.

1 ∣. Introduction

Articular cartilage and the menisci are soft tissues in the knee joint that serve important mechanical functions, but have poor capacity for self-repair. The meniscus carries up to 70% of the knee's total compressive load [1], and injury is associated with an increased risk for osteoarthritis (OA) [2] where > 75% of patients with symptomatic OA have a meniscal injury [3]. Cartilage defects are often associated with increased meniscal degeneration, but it remains unclear whether meniscal damage occurs simultaneously with cartilage injury or as a consequence of subsequent pathologic changes in the joint [4]. Generally, studies assess how meniscus damage predisposes underlying cartilage to degeneration and OA progression. However, limited information exists on how isolated cartilage injuries impact local meniscal structure and degeneration.

While clinical evidence suggests untreated cartilage defects contribute to meniscus degeneration [4], to our knowledge no studies have investigated how a focal cartilage defect impacts the medial and lateral menisci. Healthy knees have a balance between cartilage and meniscus loading, but with meniscal injury, excess load is transferred to the articular cartilage [5]. Finite element models reported a focal cartilage defect increases cartilage strains underneath the meniscus by twofold, but it remains unclear how local meniscus strains are altered [6]. Additionally, despite the known importance of physiologic loading driving meniscal repair [7], it is unknown how aberrant mechanical loads due to cartilage injury translate into structural and compositional changes in meniscus tissue and the timescale of these changes.

Clinically and in preclinical models, the structure and composition of the extracellular matrix (ECM) in the meniscus are altered in OA [8]. Specifically, loss of collagen has been observed, in both quantity and organization [8]. Destabilization of the medial meniscus (DMM) in mice increased mechanical stress, often closest to the injury site, resulting in increased collagen fiber thickness, dysregulation of collagen fiber formation, and changes in collagen expression [9, 10]. Additionally, abnormal cell clusters and meniscus hypertrophy, characterized by excess glycosaminoglycan (GAG) deposition and rounding of meniscal fibrochondrocytes, have been observed widely in OA and after meniscal injury [11], particularly in the middle and deep zones [8]. However, the time scale of these ECM changes in meniscus degeneration is poorly understood.

The degradation, deposition, and remodeling of ECM proteins in the meniscus is a dynamic process that changes with age and pathology. During meniscus development, proteoglycans and collagen secretion and organization are established on different time scales, with proteoglycans deposited after collagen fiber architecture has been established [12]. Similarly, in tissue engineered menisci, degeneration of GAGs during tissue maturation was associated with improved collagen fibrillogenesis and architecture [13]. This relationship between proteoglycan content and collagen organization has also been observed in the clinic where injured menisci had decreased collagen organization in areas with increased proteoglycan content [14]. In meniscectomy patients, proteomics data showed that collagen I expression was elevated in older patients with more severe OA [15]. Additionally, proteoglycan deposition is increased in early OA [8], but the temporal aspect of GAG changes relative to when the collagen network is disrupted remains unclear. Therefore, we are interested in investigating relationships between local proteoglycan content, meniscal hypertrophy, and collagen network changes in healthy and pathological meniscus tissue.

Since cartilage defects and meniscal damage often result from the same injury event, it is challenging to distinguish if hypertrophy and GAG deposition are due to meniscal injury or broader joint changes, and how these changes affect the local collagen matrix. This study assessed whether excess GAG deposition and hypertrophy occur solely due to osteochondral injury, and whether these local compositional changes in the meniscus are associated with alterations in the collagen network 12 weeks after cartilage injury. To examine the local matrix changes and investigate these aims, a well-characterized model of an osteochondral defect in the New Zealand white rabbit was used [16, 17]. We hypothesized that menisci from joints with cartilage defects, particularly medial menisci, would exhibit altered proteoglycan content, which would correlate with changes in cell morphology and collagen fiber organization. To test this hypothesis, the objectives of this study were to quantify proteoglycan area coverage and characterize local cell morphology through histological analysis. Additionally, we examined collagen fiber structure using histology and second harmonic generation imaging to provide a comprehensive overview of tissue composition and organization at the local level.

2 ∣. Materials and Methods

2.1 ∣. In Vivo Study Design

The NYU Institutional Animal Care and Use Committee approved all surgical procedures and experimental designs (IACUC ID PROTO202000084). As described previously, bilateral 3 mm diameter osteochondral (OC) defects were performed on the medial femoral condyle of 6 New Zealand female white rabbits, followed by bone marrow stimulation (BMS) [17]. Joints were randomized for the OC defect in an unblinded manner. Confounders such as the order of treatments and measurements were not controlled for in this study. Saline (100 mL) was injected 4 weeks after surgery in defect joints, and control limbs received no defect and no injection [17]. Animals were euthanized 12 weeks after surgery and the medial and lateral menisci were collected. One OC defect joint was excluded from the analysis due to an exposed joint capsule before tissue collection, otherwise no animals or data points were excluded from analysis. Four groups of menisci were studied (Figure 1C,D): (i) Control–Medial (no defect in condyle, n = 4), (ii) Control–Lateral (no defect in condyle, n = 4), (iii) Osteochondral (OC) Defect–Medial (in contact with defected condyle, n = 2), (iv) OC Defect - Lateral (adjacent compartment to defected condyle, n = 2).

FIGURE 1 ∣.

(A) New Zealand white rabbit depicted with an osteochondral defect on the medial condyle. (B) Timeline of study with injury occurring at Day 0, saline injection (defect only) occurring at 4 weeks, and euthanasia at 12 weeks. Age-matched control animals received no surgery and no saline injection. (C) Cartoon depicting menisci from osteochondral (OC) defect joints (left) and menisci from control (no defect) joints (right). (D) Table indicating group names and whether they have defects or no defects and their laterality.

2.2 ∣. Extracellular Matrix Composition

To assess the effects of an osteochondral defect on the relative proteoglycan and collagen content of the underlying meniscus tissue, histological assessment and second harmonic generation (SHG) imaging were performed. Menisci were removed from the joint, fixed in 10% (vol/vol) buffered formalin for 48 h, and stored in 70% (vol/vol) ethanol [18]. Prepared samples were dehydrated, embedded in paraffin, and sectioned for histological analysis.

2.3 ∣. Assessment of GAG Content and Cell Roundness via Histology

Prepared histological slides were stained using Safranin-O and fast green with Weigert's hematoxylin. Slides were imaged under brightfield to observe GAG content and images were collected using a Nikon Eclipse TE2000-S microscope (Nikon Instruments, Melville, NY) with a SPOT RT camera (Diagnostic Instruments, Sterling Heights, MI). Images were thresholded to maintain consistency and the percentage of area with GAGs was quantified using ImageJ (blinded by SL) [17]. Assessment of cell roundness and cell area were performed manually using ImageJ by outlining cell boundaries and calculating area and comparing the area to the perimeter for roundness (n = 7–10 ROIs/meniscus). GAG content was assessed globally and locally for each meniscus section (0.27 mm × 0.2 mm, n = 7–10/meniscus) and matched with cell roundness and cell area measurements.

2.4 ∣. Assessment of Collagen Structure via Histology and SHG

To understand the collagen structure, both picrosirius red histology imaged under polarized light and SHG imaging were performed. Picrosirius red histological samples imaged under polarized light display birefringence, which is associated with the size of collagen fibers [19]. SHG enables visualization of collagen at a small length scale [20]. Prepared histological slides were stained using Picrosirius red with Weigert's hematoxylin. Slides were imaged under polarized light to observe collagen structure and images were collected using a Nikon Eclipse TE2000-S microscope (Nikon Instruments, Melville, NY) with a SPOT RT camera (Diagnostic Instruments, Sterling Heights, MI).

Safranin-O stained slides were used for SHG imaging to allow for direct comparison of Safranin-O and collagen organization. An LSM 880 confocal/multiphoton inverted microscope with 10×/1.2 N.A. C-Apochromat water immersion objective were used to measure collagen fiber reflectance between 437 and 464 nm emission [21-23]. Five ROIs were set for each sample (0.25 mm × 0.25 mm) and were matched with five ROIs from the Safranin-O stained slides. The fiber alignment index was calculated for each ROI using a custom MATLAB code, as described previously [21, 24]. This code uses a series of 2-dimensional fast Fourier transforms (FFTs) to determine the maximum degree of alignment of the fibrils in each SHG image. The alignment index was calculated using the intensity ± 20° from the maximum degree of alignment and ranged from 1 (unaligned) and 4.5 (completely aligned).

The average SHG intensity and collagen fiber diameters for each ROI were calculated using ImageJ (blinded by SL). For fiber diameter analysis, ROIs (0.25 mm × 0.25 mm) were divided into a 4 × 4 grid and the diameters of fibrils were measured until the average remained constant (~100 fibrils per micrograph). Histograms of fibers from each group show a peak around 2 μm (Supporting Information S1: Figure S1). For larger fibers, one standard deviation from the peak extends to 8 μm. Based on these observations, 2 and 8 μm were chosen as the thresholds to assess both smaller and larger fibril/fiber bundles. Intensity histograms were obtained using ImageJ from SHG menisci tile scans. Fiber area coverage was calculated as the percentage of area fraction that the SHG signal covered within a given ROI on a thresholded image to account for background noise.

2.5 ∣. Statistics

Wilcoxon Rank-Sum test was used to assess differences between laterality and treatment in GAG area, cell roundness, cell area, fiber diameter, alignment index, SHG intensity, percentage of fibers below 2 μm, and percentage of fibers above 8 μm. This nonparametric test was chosen because the data did not meet the assumption of normality, as assessed using a Q-Q plot and a Shapiro-Wilk test. A Bonferroni correction was used to adjust for multiple comparisons. Values in this study were reported as mean ± standard deviation and p < 0.05 was considered significant. Pearson correlation coefficients were computed to assess the linear relationship between cell morphology metrics, collagen fiber structure, and GAG coverage in the meniscus. Pearson correlations were performed on individual ROIs where each data point represents one ROI, with a matched sample for each measurement. Statistical analysis was performed using R Statistical Software and GraphPad Prism.

Due to the limited availability of osteochondral defect samples, this study included two medial and two lateral menisci from defect joints, alongside four medial and four lateral menisci from control joints. While multiple regions of interest (ROIs) were analyzed per sample to increase measurement precision, the number of biological replicates remains small, limiting statistical power. A sensitivity power analysis was performed using R studio to estimate the minimum detectable effect size given our sample size. Due to differences in number of animals per group, the harmonic mean of 2.67 samples per group size was used as the effective sample size. The study was powered (80% at α = 0.05) to detect only very large effect sizes (Cohen's f > 1.36), corresponding to group differences explaining more than 65% of the variance (R² > 0.65). Smaller effects may not be detected with this sample size. This limitation is acknowledged and considered when interpreting the results. Findings that reach statistical significance are therefore likely to reflect large, biologically meaningful effects, while subtle differences may remain undetected. Future studies with larger numbers of biological replicates will be important to validate and extend these findings.

3 ∣. Results

3.1 ∣. Glycosaminoglycan Assessment of Menisci

Safranin-O staining revealed the medial menisci of OC defect joints show increased GAG presence compared to controls (n = 2–4, Figure 2A-D). Menisci from control joints had 5.3% ± 4.5% and 9.2% ± 11.5% of GAG area coverage in medial and lateral menisci, respectively. Notably, there was a 5.7 fold increase in GAG area % in medial menisci from OC defect joints compared to controls, but no difference between lateral menisci (Figure 2E). In the defect group, enriched Safranin-O staining was found primarily in the inner zone and mid-substance region compared to the surface, especially in the medial menisci. Safranin-O staining in the menisci of control joints was also found in the middle or outer regions, and in the mid-substance rather than closer to the surface.

FIGURE 2 ∣.

(A–D) Representative Safranin-O stained micrographs of medial meniscus in joint with defect, lateral meniscus in joint with defect, medial meniscus in joint with no defect, and lateral meniscus in joint with no defect, respectively. Scale bars = 1 mm. (E) Area quantification of GAG area percentage of coverage in meniscus as a whole.

3.2 ∣. Cell Assessment of Menisci and Correlation With GAG Content

Cell area and cell roundness were assessed using histology micrographs and subsequently correlated with GAG area % to determine effects of enriched GAG content on these cell properties (Figure 3A). Note that the micrographs in Figure 3A represent GAG rich and GAG poor areas and do not constitute the vast majority of the meniscus, as seen in Figure 2A-D. Cell area quantification did not reveal any differences between medial or lateral menisci of defect and nondefect joints (Figure 3B). However, the cell area was highly variable and appeared to be bimodal in the menisci in defect joints. In the lateral defect menisci, there was one cluster around 30 μm² and another around 75 μm². Collectively, these data average to around 50 μm², which is similar to controls, rendering them not statistically significant. Cell roundness was found to be increased in medial menisci of OC defect joints compared to controls (p = 0.01, Figure 3B). Similar to cell area, roundness in the medial menisci of OC defect joints had high variability with an apparent bimodal distribution. Cell area and roundness had positive correlations with GAG area % across all samples R = 0.64, p < 0.001 and R = 0.49, p < 0.001, respectively, Figure 3E). While the differences between compartment (medial vs. lateral) and injury (control vs. OC defect) were only significant in between injury groups in the medial menisci, GAG area coverage correlated with local differences in both cell area and roundness, regardless of injury or laterality.

FIGURE 3 ∣.

(A) Safranin-O stained regions of interest in a GAG poor region (top) and GAG rich region (bottom) of menisci. Scale bars = 100 μm. (B) Roundness measurements of cells in ROIs in each group. ROIs from separate animals are shown in varying shades of red or black (defect and no defect, respectively). (C) Cell area measurements of cells in ROIs in each group. Data points represent one ROI with 7–10 per meniscus and shades of color represent different animals (B and C) (D) Pearson correlation between GAG area % and cell roundness (each data point represents one ROI, matched sample for each measurement). (E) Pearson correlation between GAG area % and cell area (each data point represents one ROI, matched sample for each measurement). Pearson correlation coefficients (R) scale from −1 to 1, where −1 indicates a perfectly negative linear correlation, 0 indicates no linear correlation, and 1 indicates perfectly positive linear correlation between two variables.

3.3 ∣. Collagen Organization

To assess the effects of a medial OC defect on meniscus collagen structure at the bulk tissue level and microscale, both polarized light microscopy of picrosirius red histology micrographs and SHG imaging were evaluated. Collagen birefringence from picrosirius red staining correlates with fiber size and orientation [20]. Picrosirius red staining showed some slight increases in collagen fiber thickness in OC defect medial menisci, as shown by the white arrow (Figure 4A,B,E,F). Additionally, OC defect lateral menisci showed increased green birefringence, indicative of lower collagen organization and smaller fibers throughout the entire tissue. SHG was used to assess fiber level structure including fiber architecture and diameter. SHG intensity histograms of bulk menisci sections were analyzed where the mean intensity for control menisci (lateral and medial) and OC defect lateral menisci were all around 25 (Figure 4J-L). However, OC defect medial menisci had a slightly lower mean SHG intensity at 23, as well as a larger spread than the other groups with an increased probability density in areas with less intensity. Similar to the fiber diameter increase in the OC defect medial menisci from picrosirius red staining, the diameter analysis from SHG micrographs showed a 51% increase in fibril diameter in the OC defect medial menisci compared to control medial menisci (Figure 5C). In all menisci, collagen organization and fiber size increased towards the outer edges of the tissues, as expected [25].

FIGURE 4 ∣.

(A–H) Representative picrosirius red histology micrographs imaged using polarized light (red–top) and SHG microscopy micrographs (bottom–yellow) of medial (left column) and lateral (right column) menisci from defect (top 4 micrographs) and no defect (bottom 4 micrographs) joints. Scale bars = 1 mm. (I–L) SHG intensity histograms of entire menisci, averaged for each group. Means and medians of averaged intensity histograms are reported.

FIGURE 5 ∣.

(A) Safranin-O stained histology micrograph with five outlined ROIs, matched with SHG micrograph. (B) SHG micrograph with five outlined ROIs, matched with Saf-O histology micrograph. (C) Quantified fiber diameter measurements from ROIs in each group. (D) Fiber diameter measurements correlated with matched GAG area coverage in each ROI. (E) Alignment index values from each ROI. (F) Alignment index correlated with matched GAG area coverage. (G) SHG intensity for each ROI. (H) SHG intensity correlated with matched GAG area coverage. Each data point represents one ROI with five per meniscus and shades of color represent different animals (C, E, G). Pearson correlation coefficients (R) scale from −1 to 1, where −1 indicates a perfectly negative linear correlation, 0 indicates no linear correlation, and 1 indicates perfectly positive linear correlation between two variables.

3.4 ∣. Correlation of Collagen Organization and GAG Area

Safranin-O stained slides were also used for SHG imaging so that GAG presence and collagen organization could be directly compared. Five ROIs (0.25 mm × 0.25 mm) were randomly selected for each sample at the outer, outer-middle, middle, middle-inner, and inner zones of the menisci. Fiber diameter of OC defect medial menisci was elevated compared to control medial menisci (51% increase, p = 0.005), and lateral meniscus fiber diameter was lower compared to medial menisci from defect joints (p = 0.008 Figure 5C). The alignment index of collagen from SHG was slightly elevated in control lateral menisci compared to all other groups but not statistically significant (Figure 5E). The SHG intensity, which is associated with overall local collagen content and organization [20, 26], was similar with high variability within all four groups (Figure 5G). Local GAG area % was a significant predictor of fiber diameter with a moderately strong positive correlation (R = 0.54, p < 0.001), but no correlation was observed with fiber alignment index and SHG intensity (R = 0.03, p = 0.84 and R = 0.11, p = 0.4, respectively; Figure 5D).

To further characterize the collagen structure, the distributions of fibers within the menisci were analyzed. Histograms of the fibers from each group reveal a cluster around 2 μm, which was selected as the threshold for small fibers (Supplemental Figure 1). For larger fibers, a threshold of 8 μm was chosen, as it corresponds to approximately one standard deviation from the peak in the histograms (Supporting Information S1: Figure S1). Using these thresholds, the percentage of small fibers (below 2 μm), the percentage of large fibers (above 8 μm), and the fiber area coverage within each ROI were evaluated. Interestingly, the medial menisci from control and OC defect joints had the highest percentage of small fibers (< 2 μm) with 68% and 63%, respectively (Figure 6A). Additionally, medial menisci from OC defect joints had the highest percentage of large fibers (> 8 μm), and was significantly higher than control medial menisci (p < 0.05, Figure 6B).

FIGURE 6 ∣.

(A and B) Percentage of fiber below 2 μm and above 8 μm calculated from ROIs from SHG micrographs, respectively. (C–H) Pearson correlation of percentage of fibers below 2 μm and above 8 μm compared with fiber area coverage, number of fibers, and GAG area %. Each data point represents one ROI with five per meniscus. Each sample has matched ROIs from Safranin-O and SHG micrographs. Pearson correlation coefficients (R) scale from −1 to 1, where −1 indicates a perfectly negative linear correlation, 0 indicates no linear correlation, and 1 indicates perfectly positive linear correlation between two variables.

To better understand differences in fiber distributions, correlations were performed between matrix characteristics. While fiber coverage was similar across all groups (Supporting Information S1: Figure S2), all ROIs exhibited high variability. A significant negative correlation was observed between fiber coverage and percentage of fibers below 2 μm (R = −0.33, p = 0.01, Figure 6C). Similarly, when compared with fiber number, there was a very strong negative correlation to percentage of fibers below 2 μm (R = −0.98, p < 0.001, Figure 6E). However, there was no significant trend between percentage of fibers below 2 μm and GAG area % (R = 0.04, p = 0.74, Figure 6G). For larger fibers, the percentage of fibers > 8 μm showed a positive correlation with fiber coverage (R = 0.42, p = 0.001, Figure 6D), but, interestingly, there was little correlation with the number of fibers (R = 0.14, p = 0.28, Figure 6F). Unlike the percentage of fibers < 2 μm, the larger fiber diameters did have a significant and positive correlation with GAG area % (R = 0.5, p < 0.001) (Figure 6H), matching findings from average fiber diameter data (Figure 5D).

4 ∣. Discussion

The objective of this study was to characterize the relationship between local GAG content, cell morphology, and collagen fiber organization in menisci from healthy and cartilage-injured joints. Using an OC defect model in rabbits, we examined whether changes in GAG content affected the local structure and organization of the collagen fiber network in menisci. GAG content increased 5.7-fold in the medial menisci underlying the injured medial condyle compared to uninjured controls. This increase aligns with studies of late-stage OA menisci or those with traumatic tears, where proteoglycan deposition is elevated [9]. Proteoglycans, especially large aggregating proteoglycans, contribute to the compressive resistance of the tissue [27]. While aggrecan is prevalent in the inner meniscus of healthy joints [28], injury increased proteoglycan content in this region (Figure 2). These findings support other rabbit studies examining ACLT-induced effects on meniscal injury [29]. Collectively, these data highlight pathological alterations in proteoglycan content and the correlation with collagen fiber size and cell morphology in menisci within 12 weeks of OC injury.

Several factors likely drive pathological changes in the joint after injury, including inflammation and mechanical alterations. In this study, we created a medial OC defect and evaluated both medial and lateral menisci in the joint. Inflammation or biochemical changes increase with intra-articular fractures and may contribute to increased GAG content [30]. However, it is improbable that inflammation is restricted to the medial side, making it difficult to attribute the observed GAG changes solely to inflammation. As such, altered mechanics likely drove the observed ECM changes, as an OC defect in the medial condyle increases contact pressures and aberrant loading on the underlying meniscus [6], rendering the joint more susceptible to degenerative processes. Additionally, inflammation may stretch the synovium, further disrupting joint mechanics. In healthy knees, load is balanced between the condyle and meniscus [5]. Thus, a cartilage defect may increase meniscal loading, leading to hypertrophy and increases in proteoglycan production and cell morphology changes.

To assess how an osteochondral defect and increased GAG deposition affect cell morphology, GAG area was correlated with cell area and roundness (Figure 3). While no distinct difference in cell area was observed between medial and lateral menisci of OC defect and control joints, we found a significant positive correlation between GAG area % and both cell area (R = 0.64, p < 0.001) and cell roundness (R = 0.49, p < 0.001). Further, cell roundness differed significantly between OC defect and control joints (p = 0.01). These shape changes are consistent with previous work documenting that GAG deposition and hypertrophy in human OA menisci are associated with more rounded cells [9, 31]. Interestingly, GAG increases in the rabbit menisci from OC defect joints were predominantly in the mid-substance, consistent with findings in OA menisci [8, 11]. Collectively, these data point to the critical role of altered joint mechanics in driving pathologic changes in meniscus cells and ECM.

While it is known that in injured and OA joints, meniscus GAG content increases and there is loss of collagen content and organization [9], the timeline of such changes is unclear. Fiber diameter was elevated in defect medial menisci by 51% (p = 0.005) compared to control medial menisci (Figure 5), consistent with other studies showing increased collagen I fiber thickness in a DMM mouse model [9]. Comparing fiber properties with local GAG area % showed that fiber diameter was positively correlated with GAG area (R = 0.54 p < 0.001), indicating that the local increase in proteoglycan deposition due to joint injury leads to changes in matched collagen structure consistent with hallmarks of meniscal OA progression. Alignment index and SHG intensity were similar across all groups and had no significant correlations with GAG area %. While we might expect a decrease in collagen organization, reflected by reduced alignment and SHG intensity in defect menisci, the menisci in this study were collected 12 weeks postinjury. This timeframe may not have been long enough to observe the full progression of disease or degeneration, including loss of collagen and collagen organization.

To determine the effect of an OC defect on meniscus collagen fiber size and distribution, the percentage of small (< 2 μm) and large (> 8 μm) fibers and the fiber area coverage within each ROI were quantified. Both medial meniscus groups had the highest percentage of small fibers (< 2 μm), indicating the OC defect did not alter the percentage of small collagen fibers. Fiber coverage and percentage of small fibers (< 2 μm) were negatively correlated (R = −0.33, p = 0.01), indicating that ROIs with a higher proportion of small fibers had a smaller area fraction covered by fibers (Figure 5C). Additionally, there was a very strong negative correlation between the percentage of small fibers (< 2 μm) and the total number of fibers (R = −0.98, p < 0.001), suggesting that regions with a higher proportion of smaller diameter fibers contained significantly fewer fibers overall (Figure 5E). Although the negative relationship between the percentage of small fibers and the number of fibers may seem counterintuitive, ROIs with a larger total number of fibers–and fewer small fibers–may represent healthier regions of the matrix. These findings underscore a potential relationship between matrix organization and fiber dimensions, providing insight into how fiber size distributions may reflect tissue health.

When correlating small fibers with GAG area percentage, no clear trend was observed. This lack of association is unsurprising, as GAG increases are typically linked to larger fiber diameters and small fibers were present across most ROIs [9]. Although it is intuitive that small fibers would appear in many ROIs, these results are noteworthy because ROIs with predominantly small fibers contained fewer fibers overall, and these fibers occupied much less space. Examining the SHG micrographs of OC defect medial menisci (Figure 4), patches of high SHG signal and larger fiber diameters are evident, alongside regions with little signal and much smaller fibers. These data suggest that collagen structure within these menisci may undergo heterogeneous changes in joints with cartilage injury.

Medial menisci from OC defect joints unsurprisingly had the highest percentage of large fibers (> 8 μm), and had 2.1-fold higher percentage of these larger fibers than control medial menisci. These larger fiber data correlated positively with fiber coverage (R = 0.41, p = 0.001). However, when compared with fiber number, there was little correlation, indicating that there weren't necessarily fewer fibers in a given ROI if there were large fibers present. Additionally, consistent with fiber diameter data, there was a positive correlation with GAG area % (R = 0.5 p < 0.001). Overall, areas with higher GAG area coverage contained larger fiber diameters. In contrast, ROIs with smaller fibers had fewer fibers and they took up less space, perhaps indicating loss in collagen organization. Both ends of this spectrum were seen primarily in medial menisci from OC defect joints. ECM changes in proteoglycan content and collagen fiber architecture were predominantly in the mid-substance for all samples, suggesting this region of the meniscus may be more sensitive to mechanical aberrations in the joint and may also be the first location of pathological changes. When correlations were analyzed separately for control and OC defect ROIs, some trends remained consistent between the two groups—such as the correlation between number of fibers and the percentage of fibers below 2 μm—while others differed substantially, including the correlation between GAG area percentage and the percentage of fibers below 2 μm (Supporting Information S1: Table S1). Further investigation is needed to fully elucidate these relationships in healthy versus pathological environments.

Several limitations should be considered when interpreting these data. First, we had only two lateral and medial OC defect samples, limiting statistical power when comparing bulk properties. However, local analyses were performed providing larger data sets at the microscale level comparing matrix content, organization, and cell morphology (n = 7–10 measurements per sample). These meniscus samples were harvested as excess tissue from rabbits involved in a separate study and a sensitivity analysis was performed, as detailed above [17]. Increasing sample size would improve statistical power and allow for further exploration of how cartilage injury affects underlying meniscus tissue. This study evaluated a single time point postinjury at 12 weeks, limiting our understanding of meniscus remodeling dynamics. Additionally, no significant changes in SHG intensity were observed at this 12-week time point. Extending the timeline and adding more collection points could provide further insight into the progression of meniscus ECM pathology and the relationship between proteoglycan deposition and collagen degeneration and disorganization following joint injury. It is known that proteoglycan deposition increases in early OA [8] and that the collagen network is eventually disrupted [9]. However, the temporal sequence of these matrix changes, and the extent to which they occur in relation to each other, remains unclear. In this study, we observed increased proteoglycan deposition in defect joints at a 12-week postinjury time point. However, assessment of collagen organization via SHG did not reveal any significant differences between injured and uninjured tissues. It is also known that in meniscus development, collagen and proteoglycans are deposited on different time scales, with collagen networks forming before proteoglycan deposition [12]. It may be that the tissue's response to injury also follows a temporal pattern and that the 12-week postinjury time point was not sufficient to observe the full progression of degeneration, particularly with respect to collagen loss and disruption of collagen organization. In addition, most changes in the defect joints were observed in the medial meniscus. Therefore, investigation of later time points after injury may reveal changes in the lateral menisci of defect joints.

In conclusion, we found that a medial OC defect leads to localized pathological changes in the underlying meniscus, likely due to an altered loading environment. These changes were evidenced by increased proteoglycan deposition in medial meniscus following injury, along with hypertrophic cell morphology, indicated by increases in cell roundness and cell area in GAG rich regions. Effects were observed in collagen architecture, including increased fiber diameter in the medial menisci of OC defect joints. This increase in fiber diameter was positively correlated with elevated GAG coverage. Notably, the medial menisci of OC defect joints exhibited abnormal and sporadic regions of altered collagen structure, characterized by a broader range of fiber diameters, a higher average fiber diameter, and areas of small fibers with very low SHG signal, indicating poor collagen organization. These findings underscore the localized and heterogeneous nature of matrix changes, highlighting cartilage-meniscus tissue crosstalk and its impact on ECM turnover and maintenance. Importantly, the lateral menisci in joints with defects were largely unaffected, suggesting that at 12 weeks postinjury, degenerative changes remain confined to the compartment directly underlying the defect. Understanding the regulation of GAG deposition and fibrochondrocyte pathology in injured or diseased meniscus tissue helps inform further therapeutics and disease progression within the joint. These results demonstrate that an altered loading environment resulting from an osteochondral defect can lead to measurable downstream pathological effects in surrounding tissues. Therefore, early intervention after such an injury may help mitigate these effects and underscores the importance of addressing not only the damaged tissue itself but also the underlying and surrounding tissues. Our lab investigated the effects of recombinant lubricin injection on the cartilage following an osteochondral injury and observed improved quality and lubricating ability of the repair tissue compared to tissues with no intervention [1]. Future work will build upon this study, investigating how this early lubrication-based therapies may benefit the menisci and surrounding tissues after an osteochondral injury.

Supplementary Material

Supplemental Materials

Supplemental Figure 1: Fiber diameter histograms of all ROIs for each group combined. Supplemental Figure 2: Fiber coverage calculated from ROIs from SHG micrographs. Supplemental Table 1: Correlations done separately on control (top) and OC defect joints (bottom).

Additional supporting information can be found online in the Supporting Information section.