Skip to main content
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Aug 31;44(9):e70271. doi: 10.1002/jor.70271

The Human ACL Sheath Exhibits Distinct Transcriptomic Signatures and Functional Properties

James Rogot 1, Jennifer A Kunes 2, Kurt S Holuba 2, Roshan P Shah 2, Christopher S Ahmad 2, Stavros Thomopoulos 1,2, Clark T Hung 1,2, Beth G Ashinsky 2,✉
PMCID: PMC13528642  PMID: 42671303

ABSTRACT

The anterior cruciate ligament (ACL) has poor intrinsic healing capacity due to exposure to the intra‐articular synovial environment of the knee. Therefore, tendon auto‐ or allograft reconstruction is the mainstay of treatment for a torn ACL. Following reconstruction, there is a high rate of post‐traumatic osteoarthritis, likely due to alteration of joint biomechanics, which supports the clinical need to advance primary ACL repair strategies. The ACL is enclosed by a vascularized, synovial‐like sheath (ACL‐s), which supplies nutrients to the central core (ACL‐c). The ACL‐s may act as a selective barrier to solute diffusion, protecting the ACL‐c from damaging synovial enzymes and/or hyaluronic acid, while allowing nutrient diffusion. Like the synovium, the ACL‐s is rich with collagens and resident macrophages, thus may be susceptible to change of function following injury and inflammation. We investigated the healthy and inflammed ACL‐s structure, diffusivity, and cellular profile in comparison to the ACL‐c and synovium to guide the development and optimization of primary repair technologies. Diffusion and histologic assessments revealed the ACL‐s functions as a semi‐permeable barrier through lower solute diffusivity and unique collagen organization. RNAseq demonstrated differential expression of extracellular matrix, inflammatory, and immune‐related genes across ACL‐s, ACL‐c, and synovium, with the ACL‐s uniquely showing upregulated expression of CX3CR1 and CLDN5 suggestive of an immunologic barrier. This study establishes the potential of the ACL‐s to serve an immunologic and functional barrier that protects the underlying ACL‐c from the inflammatory milieu following injury, and therefore, should be recapitulated when developing primary ACL repair technologies.

1. Introduction

Anterior cruciate ligament (ACL) tears are one of the most common orthopedic injuries, with nearly 200,000 reconstructions performed each year [1]. Reconstruction utilizing a tendon auto‐ or allograft is the current gold standard of treatment and is successful at restoring gross biomechanics to the knee joint. However, following ACL reconstruction, there is a high incidence of developing post‐traumatic osteoarthritis (PTOA), as early as 10 years following surgery. This is likely secondary to a combination of factors, including micromechanical instability, altered proprioception, and biochemical sequelae of the inflammatory milieu which accompany knee injuries [2]. Over the past decade, there has been a resurgence in research efforts to advance primary ACL suture repair, as well as scaffold augmentation, to promote intrinsic healing of the native ligament. However, the ACL has poor intrinsic healing capacity, likely due to the exposure of intra‐articular enzymes and inflammatory mediators, and as such, primary ACL repair has high failure rates and remains a challenging clinical problem [3, 4, 5, 6, 7].

The ACL is surrounded by the ACL sheath (ACL‐s), which is speculated to be an extension of the synovial lining of the knee [8]. The ACL‐s is supplied by a network of blood vessels that branch off from the middle genicular artery and is responsible for diffusion of nutrients into the ACL proper, or core (ACL‐c). This central portion of the ACL is less vascularized than both the proximal and distal ends and thus is primarily reliant on the ACL‐s for blood supply [9]. Furthermore, much of the nutrient supply to the ACL‐c is reliant on diffusion of small molecules from the synovial fluid into the tissue [10]. As such, the sheath may serve as a semi‐permeable barrier to regulate diffusion of nutrients or other soluble factors into the ACL‐c. Previous work has proposed a putative protective role for the ACL‐s for the ACL‐c against soluble collagenases in the synovial fluid following trauma and inflammation [6, 11]. In contrast, following chronic inflammation, diffusion through the synovium has been shown to significantly increase, potentially leading to its breakdown and a leaky barrier [12, 13]. However, the diffusivity of the ACL‐s in healthy and pathologic conditions remains poorly understood, despite its critical role in maintaining the health of the ACL.

The structure of the ACL‐s is comprised primarily of a network of disorganized collagen type IV fibrils, and has been shown to stain positively for a small fraction of collagen type I in the rabbit [14]. This pattern can be contrasted against the highly organized collagen type I structure of the ACL‐c which underlies the sheath [15, 16, 17, 18, 19]. Similar to the cellular makeup of the synovium, the ACL‐s contains resident macrophages and fibroblasts [20]. These cells may further contribute to the barrier function of the sheath, especially in inflamed joints. Culemann et al. characterized a population of CX3 motif chemokine receptor‐1+ (CX3CR1+) macrophages that play a direct immuno‐regulatory response to inflammation through tight junctions in the synovial cavity in rheumatoid mice [21]. It is unknown whether these synovial resident macrophages are conserved in humans, or in non‐inflammatory arthritic knees.

While recent clinical studies highlight the need of the ACL‐s in ACL healing during primary repair, further investigation is needed to characterize the role of the ACL‐s in regulating homeostasis [22]. In the current study, we aim to define the structure, diffusivity, and cellular response of the ACL‐s in contrast to the ACL‐c and the synovium. Establishing the importance of the sheath for protecting the ACL‐c has critical implications for advancing primary repair strategies. We hypothesize that the ACL‐s will be highly impermeable to diffusion and can be further differentiated from the ACL and synovium through gene and protein expression in healthy and inflamed states.

2. Methods

This was a case‐control study with Level of Evidence III.

2.1. Collection of Tissues

Intact human ACLs were recovered from patients undergoing primary total knee arthroplasty (n = 3, IRB AAAQ2703) or from the Musculoskeletal Transplant Foundation tissue bank (n = 3) and split into two groups: “OA” and “healthy” (non‐OA). Exclusion criteria for the healthy group were history of ACL tear, scleroderma, or cancer. Tissues were rinsed in PBS (ThermoFisher Scientific) containing 1% antibiotic‐antimycotic (AA, Gibco) to remove surgical debris and then divided into three sections for histology, diffusion assay, and isolation of RNA/cells. ACL‐s and ACL‐c were kept intact for histology and diffusivity measurements, but separated prior to cell isolation. Samples designated for RNAseq were equilibrated in RNAlater (Invitrogen) overnight at 4°C before being flash frozen and stored at −80°C. Synovium samples were also recovered for histology, diffusion, and gene expression assessments. Due to a limited availability of donor tissues, n = 3 healthy ACL and synovium were only used for RNAseq and representative histology analyzes while n = 2 was used for diffusivity measurements.

2.2. Histology

OA ACLs were placed onto a #1 thickness coverslip (Corning) and imaged for collagen organization using second harmonic generation (SHG) imaging on a confocal microscope (Nikon Instruments). SHG images were captured at 880 nm and 30% laser power. Tissues were then fixed in 10% formalin phosphate (Fisher Scientific), embedded in paraffin and sectioned to 6μm thickness (Columbia Molecular Pathology Core Facility). Slides were stained with H&E (cellularity) and picrosirius red (PSR, collagen) before being imaged through transmitted and polarized light.

2.3. Assessment of Transport Properties

Diffusivity was assayed using fluorescent recovery after photobleaching (FRAP) of high‐molecular‐weight‐hyaluronic‐acid (HMWHA) on a confocal microscope. Tissues were first incubated overnight in PBS supplemented with 100 μg/mL fluorescein isothiocyanate‐ (FITC‐) HMWHA (HAWorks). Tissues were blot dried and a small region in either the core, sheath, or synovium was photobleached at 488 nm for 30 s at 100% laser power. Diffusion of HMWHA back into the photobleached area was recorded and plot as a function of intensity over time. The one‐dimensional diffusivity of HMWHA through the tissues was calculated from a least‐squares curve‐fitting algorithm as a function of the fluorescent intensity, photobleached area, and time [23]. The synovium and ACL‐c from one healthy donor were not assayed due to insufficient tissue available during recovery. Data were reported as mean ± standard deviation and analyzed using a one‐way ANOVA with Tukey's HSD post‐hoc test (α = 0.05).

2.4. Total RNA Isolation

Frozen tissues were thawed on ice, minced, and 15 mg of each sample was transferred to a RNase‐/DNase‐free U‐bottom tube (ThermoFisher Scientific). 1 mL of ice‐cold TRIzol (Invitrogen) was added to each tube and a mini tissue homogenizer with a S‐18‐N dispersing tool (IKA) was used to homogenize tissues.

Homogenized tissues were transferred to phase separation gel tubes (Invitrogen), incubated with chloroform and centrifuged at 4°C and 12,000 G for 15 min. The clear, aqueous phase was pipetted into a RNase‐/DNase‐free microcentrifuge tube and combined with 250 µL isopropanol and 250 µL high‐salt solution (1.2 M NaCl and 0.8 M Sodium Citrate in RNase‐/DNase‐free water) to aid in separation of nucleic acids and chelate metal ions that activate RNase and DNase. This mixture was then moved into a RNeasy spin column (Qiagen) for elution of total RNA.

Total RNA was quantified by sending samples to the Columbia Molecular Pathology Core Facility. Diluted RNA was measured for concentration, RIN, and fraction of RNA greater than 200 nucleotides in length (DV200). Samples with RIN > 6 and DV200 > 70% were sent for bulk RNA sequencing (Azenta Life Sciences).

2.5. Bulk RNA Sequencing

Total RNA was sequenced at a depth of 20 × 106 reads with a mean quality score of 38.25% and 91.47% of bases > 30. Per the Azenta sequencing protocol, reads were trimmed to removed adapter sequences and poor‐quality nucleotides using Trimmomatic v.0.36 [24]. Trimmed reads were mapped to the homo sapiens GRCh38 reference genome using the STAR aligner v.2.5.2b [25]. Gene hit counts were calculated with featureCounts from the Subread package v.1.5.2 [26]. Downstream analysis of differential expression (DE) was completed with DESeq2 [27]. A Wald test (p < 0.05) was used to calculate p‐values from comparisons between groups and across replicates. Data with |(log2[Fold Change])| > 1 and p < 0.05 was reported as differentially expressed. GeneSCF v.1.1 was used to perform gene ontology (GO) analysis and clustering of the set of significantly expressed DE genes based on the goa_human GO list [28]. Clusters of related genes based on their biological processes (enriched GO) were created and evaluated for statistical differences using Fisher's Exact Test.

2.6. Immunostaining of Tissues

Blank tissue sections were stained (1:100) with primary antibodies to Claudin‐5 (CLDN5) tight‐junction protein (Abcam), CX3CR1 (Santa Cruz Biotechnology), and counterstained with DAPI (1:1000, Vector Laboratories). Fluorescent images were captured on a confocal microscope (Zeiss). Mean integrated density of each image channel was calculated in ImageJ to compare the distribution of CX3CR1+ cells and tight junctions between healthy and OA tissues.

2.7. Cell Culture

ACL‐c and ACL‐s were separated, rinsed, and then incubated in collagenase type II (Worthington Biochemical) at 37°C for 8 h. Synovium samples were debrided of adipose tissue prior to digestion. Digested cells were passed through a 40 μm mesh strainer before plating into tissue‐culture treated flasks at a density of 2 × 105 cells in minimum essential medium‐alpha (αMEM, Gibco) supplemented with 10% vol/vol fetal bovine serum (FBS, ThermoFisher Scientific) and 1% vol/vol AA (Gibco). Cells were expanded until passage 2–3 before further treatment.

Fibroblasts were passaged (Sigma‐Aldrich) and plated in 12‐well plates at a density of 5 × 104 cells/well. Media ± 20 ng/mL interleukin‐1β (IL1β, ThermoFisher Scientific) was refreshed after an overnight attachment period and cells were cultured for 24 h. Inflammatory dosing and treatment time course were selected based on data from previous work which demonstrated that tendon fibroblasts are optimally responsive after being primed with 20 ng/mL IL1β for 24 h [29].

Following treatment, monolayers were washed with sterile PBS before being processed with the RNeasy Kit (Qiagen) to precipitate RNA. Two‐step RT‐qPCR was performed to measure gene expression of IL6, MMP3, cyclo‐oxygenase‐2 (COX2), CX3CR1, CLDN5, and tenascin‐C (TNC). RNA was first converted to cDNA using the High‐Capacity cDNA Reverse Transcription Kit (Fisher Scientific). Gene master mixes were then created using custom primers (Table 1, IDT) and SYBR Green (Applied Biosystems). Gene expression was calculated using the 2−ΔΔCt method and normalized to GAPDH and untreated controls.

Table 1.

List of human qPCR primers.

Gene Forward primer Reverse primer
IL6 AATCTGGGTTCAATCAGGCGA GTGTTTGTGGCTGGAGTGGT
MMP3 ACAAAGGATACAACAGGGACCA GGAACCGAGTCAGGTCTGTG
COX2 TGAGCATCTACGGTTTGCTG TGCTTGTCTGGAACAACTGC
CX3CR1 TCCGTAACCCAACTCTCCTG GGTGACTGGGAGGGGATATT
CLDN5 CGGGTGTCAGACTGAGGATT AAGTAAGGCAGCAGCCAAGA
TNC GGTACAGTGGGACAGCAGGT GAAGGATCTGCCATTGTGGT
GAPDH CAAGAGCACAAGAGGAAGAGAG CTACATGGCAACTGTGAGGAG

3. Results

3.1. Human ACL Structure

Longitudinal sections of the intact ACL showed two distinct regions in OA and healthy tissues: one consisted of organized extracellular matrix (ECM) populated with fibroblasts and the other of relatively disorganized ECM and was innervated, vascular, and highly cellular (Figure 1A). The OA and healthy tissue structures both stained positive for collagen through PSR, although there were local differences when comparing regions of the tissues (Figure 1B). The central tissue structure, or ACL‐c, displayed thick, parallel‐oriented collagen bundles, as seen through polarized light of PSR‐stained sections in healthy and OA tissues and SHG images in OA tissues (Figure 1C,D). There was a notably dimmer polarized light signal when the ACL‐s was imaged under the same conditions, indicating a lack of clear organization, despite the presence of collagen fibers. This pattern was further observed under SHG of OA ACL through the random alignment of collagen in the ACL‐s (Figure 1D). The ACL‐s was characterized by a heterogeneous arrangement of collagen fibers with interspersed sinusoidal vasculature structures (Figure 1A,B,C arrows). The elongated fibroblasts were found sparsely populated in the ACL‐c, along the crimped pattern of the collagen fibers. In contrast, the high cellularity of both fusiform and circular cells in the ACL‐s were seen ubiquitously throughout the tissue. Further differentiating the ACL‐s and the ACL‐c was the presence of sinusoidal infiltrations in the sheath, which stopped at the border between the two tissues (Figure 1A). In contrast, the ACL‐c displayed a dense, organized matrix of parallel‐aligned collagen bundles with minimal vascularization, emphasizing the structural and functional differentiation between these regions (Figure 1A).

Figure 1.

Figure 1

Representative histology of the OA ACL‐s and ACL‐c. Hematoxylin and eosin (H&E), picrosirius red (PSR), polarized light imaging of PSR, and single harmonic generation (SHG) images were captured of the ACL‐s (Teal) and ACL‐c (Purple) from OA and healthy donors. Scale bars represent 50 μm.

3.2. ACL‐s Barrier Function

Representative regions in the OA ACL‐c and ACL‐s (Figure 2A) display the degree to which HMWHA saturation and subsequent photobleaching occurred in key tissues (Figure 2B). A custom MATLAB program was used to assess bleached intensity across the field of view and then plot the recovery of signal in the bleached area over time (Figure 2C,D). The OA ACL‐c exhibited the highest diffusion coefficient (49.82 ± 3.50 μm2/s) compared to the other tissues. The OA ACL‐s was significantly less permeable to HA (29.28 ± 4.81 μm2/s, p < 0.05), demonstrating its effect as a semi‐permeable membrane around the OA ACL‐c, similar to that of the OA synovial lining (Figure 2E). The pattern of diffusivity between the ACL‐c, ACL‐s, and synovium was upheld in healthy tissues. The healthy ACL‐s was less permeable than the healthy ACL‐c (24.01 ± 10.05 μm2/s vs. 35.19 ± 11.25 μm2/s) and more permeable than the synovium (17.16 ± 6.02 μm2/s), however these differences were not statistically significant (Figure 2E).

Figure 2.

Figure 2

Diffusivity measurements of the OA ACL‐c, ACL‐s, and Synovium. (A) Representative sample areas from the ACL‐c (Purple), ACL‐s (Teal), and Synovium (Green). (B) Representative fields of view of the FITC‐saturated ACL‐c (left) and ACL‐s (right). (C) Example of the Gaussian distribution of bleached intensity. (D) Example of the recovery in photobleached area over time. Diffusivity was calculated as the slope of the line. (E) Diffusion coefficients measured from OA and healthy donors in all three tissues. One‐way ANOVA with Tukey's HSD post‐hoc test (α = 0.05) was used to make comparisons. *p < 0.05, ***p < 0.001.

3.3. Expression Patterns in the ACL and Synovium

The mean quality score of the sequenced reads was 38.25% with 91.47% of bases greater than or equal to 30 bp in length. There was an average of 33,121,710.6 ± 4,954,806.8 total reads per sample and 32,234,542.4 ± 4,864,225.8 total mapped reads to the homo sapiens GRCh38 reference genome (97.31 ± 0.51% map rate). Principal component analysis (PCA) between each relevant comparison was made and projected onto a 2D plane spanned by their first two principal components. The total number of DE genes between different tissue structures (ACL‐c, ACL‐s, synovium) and across disease states (healthy vs. OA) was calculated after trimming of the sequenced RNA (Table 2).

Table 2.

Summary of DE genes. Upregulated genes are expressed more in the first entity than the second entity. Downregulated genes are expressed less in the first entity than the second entity.

Comparison Upregulated genes Downregulated genes Total significantly DE genes
Healthy_sheath‐vs‐healthy_core 138 133 271
Healthy_sheath‐vs‐healthy_synovium 986 603 1589
Healthy_sheath‐vs‐OA_sheath 1015 1499 2514
Healthy_core‐vs‐OA_core 66 146 212
OA_sheath‐vs‐OA_core 1 0 1
OA_sheath‐vs‐OA_synovium 251 68 319

PCA of the healthy ACL‐s versus ACL‐c comparison showed two distinct clusters of patient samples with 45.44% variance across the first PC and 30.05% variance across the second PC (Figure 3A). 271 DE genes were found between the healthy ACL‐s and ACL‐c, pointing to significant differences in expression patterns in the tissues (Figure 3B). Genes involved in ECM organization, skeletal system development, and inflammatory response pathways were all significantly different between the ACL‐s and ACL‐c (Figure 3C,D). PCA clustering of the healthy ACL‐s and synovium was clearer along the first principal component with 54% variance (Figure 4A). There were 1589 DE genes between the healthy ACL‐s and synovium, affiliated with inflammatory response, cell signaling, cell adhesion, and ECM organization pathways. Notably, markers of matrix remodeling, such as ADAMTS, and OA progression (Wnt) were DE between the tissues. Markers of immune cell chemotaxis, such as CCL2 and CXCL3 were present in both tissues (Figure 4B,C,D).

Figure 3.

Figure 3

Healthy ACL‐s versus ACL‐c analysis. (A) PCA comparison. (B) Volcano plot of DE patterns. (C) Key DE genes of the ACL‐c (Teal) versus ACL‐s (Purple). Scale bar represents log2 (fold change). (D) GO analysis.

Figure 4.

Figure 4

Healthy ACL‐s versus Synovium analysis (A) PCA Comparison. (B) Volcano plot of DE patterns. (C) Key DE genes of the ALC‐s (Purple) and synovium (Green). Scale bar represents log2 (fold change). (D) GO analysis.

In the OA ACL‐s, DE of 2514 genes were found in comparison to the healthy ACL‐s, representing a sharp change following the onset of chronic inflammation and degeneration.

Markers of fibrotic ECM organization, inflammation, and immune response were all significantly changed (Supporting Information Figure S1). There were 212 DE genes in the OA ACL‐c compared to non‐diseased groups, mostly associated with cellular responses to secondary messengers (cyclic adenosine monophosphate, calcium ions, protein kinases) and foreign substances (lipopolysaccharide, morphine, peptides). Interestingly, no significant changes in ECM or structural GO terms were observed (Supporting Information Figure S2). When comparing differences between diseased tissues, only 1 DE gene was found between the OA ACL‐s and ACL‐c (Supporting Information Figure S3). No other significant differences were found between these two tissues. Differences between the OA ACL‐s and OA synovium stemmed from changes in metabolic processes, such as insulin and glucose responses (Supporting Information Figure S4).

3.4. Macrophage and Tight Junction Markers in the ACL‐s and Synovium

Immunostaining of the OA ACL revealed a population of CX3CR1+ macrophages throughout the ACL‐s, localized around the blood vessels and throughout the tissue (Figure 5). There was no evidence of infiltration of the macrophages into the ACL‐c. Closely associated with the macrophage population was evidence of the tight junction protein CLDN5. Staining of CLDN5 was more widespread within the ACL‐s than CX3CR1 and was not necessarily confined to areas near the vasculature. Similarly, the human synovial lining expressed a barrier structure of CX3CR1+ macrophages and tight junctions along the intimal lining, but not the blood vessels throughout the tissue. The healthy ACL‐s demonstrated a similar pattern of expression for both CX3CR1 and CLDN5 localized around the vasculature but absent from the ACL‐c (Figure 5). While a population of CX3CR1+ macrophages were found along the intima of healthy synovium, there was a distinct lack of expression of CLDN5.

Figure 5.

Figure 5

Protein expression in the healthy and ACL and Synovium. ACL‐c (Purple), ACL‐s (Teal), and synovium (Green). Scale bar represents 50 μm.

3.5. Differential Gene Expression in the ACL‐c, ACL‐s, and Synovium

ACL‐c, ACL‐s, and synovium‐derived fibroblasts from OA donors all responded to the cytokine challenge by a significant upregulation of pro‐inflammatory markers IL6, MMP3, and COX2 compared to untreated controls (Figure 6A). CX3CR1, CLDN5, and TNC expressions were upregulated in the ACL‐s and synovium but downregulated in the ACL‐c. Differential expression of CLDN5 between the ACL‐c and the ACL‐s (p < 0.01) and synovium (p < 0.0001) was observed, pointing to the ability of those tissues to form new tight junctions in response to inflammation. TNC expression was significantly upregulated in the synovium compared to the ACL‐c (p < 0.01), indicative of fibrosis in the synovium (Figure 6B).

Figure 6.

Figure 6

Gene Expression of OA Fibroblasts from the ACL‐c, ACL‐s, and Synovium following IL1β challenge. Two‐way ANOVA with Tukey's HSD post‐hoc test (α = 0.05) used to compare differences between groups. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

4. Discussion

Previous work has shown that the ACL‐s exhibits a protective role to the ACL‐c against degradative enzymes and hemarthrosis following trauma to the knee [30]. Our findings suggest that the sheath may serve as a physical barrier by hindering the diffusion of molecules through the OA ACL‐s into the OA ACL‐c, and as a biologic barrier through tight‐junction and macrophage mediation. The OA ACL‐s is comprised of significantly less solute‐penetrable ECM than the ACL‐c, potentially pointing to its ability to act as a semi‐permeable membrane. Notably, this trend was also observed in healthy tissues; however, the difference did not reach statistical significance, suggesting that OA fibrosis may amplify an underlying baseline property of the sheath. This ability of the ACL‐s may be due, in part, to the random alignment of the fibrotic collagen fibers when compared to the ACL‐c. While the thicker, longitudinally‐oriented fibers in the ACL‐c provide axial tensile structural support, the looser fibers in the ACL‐s likely serve a non‐structural role [11, 14].

The RNAseq expression patterns of the healthy ACL‐c, ACL‐s, and synovium were distinct from each other. While there were overlapping motifs, the differential expression of ECM organization, cell adhesion, signal transduction, and inflammatory response genes defined distinct characteristics of the three tissue types. The genes encoding complement C3,C7, and CXCL12 were differentially expressed in the healthy ACL‐s, pointing to an innate ability of the tissue to induce macrophage recruitment following inflammation [31, 32, 33]. DE of MMPs and COLs varied, as the structural organization and basal remodeling capacity of the ACL‐s was greater than the ACL‐c [14, 34]. The ACL‐c significantly upregulated Mohawk (MKX) expression relative to the ACL‐s, while SOX9 expression was downregulated, further differentiating the structure and propensity to respond to OA degeneration between the tissues [35, 36]. Comparing the healthy ACL‐s with the healthy synovium, COL1, COL2, COL3, COL4, COL5, and COL8 were all found to be DE. As both tissues stained intensely positive for PSR, the relationship between the COL expression patterns must be unique. While differences in COL1 and COL4 expression were previously reported between the ACL‐c and ACL‐s [14], the current study expands our understanding by identifying the uniqueness of the ACL‐s from the synovium through gene expression patterns. DE of genes from the ADAMTS family were also found between the ACL‐s and synovium, indicating a variation in matrix remodeling capability [37]. WNT expression mediates joint homeostasis in both healthy and OA phenotypes and was found to be significantly upregulated in the ACL‐s compared to the synovium [38]. Previous work explored the Wnt signaling cascade in OA cartilage, bone and synovium; the current work adds data on WNT expression by ACL cells [39]. The results point to the importance of exploring OA pathology as a whole joint disease.

In patients who experienced end‐stage OA, there was a dampened DE of genes between the ACL‐s versus the ACL‐c and ACL‐s versus the synovium. The cells within the ACL have been shown to become senescent following PTOA, which may explain this phenomenon [40]. While the tissue structures of the ACL‐s and ACL‐c remain distinct through histology and diffusivity outcome measures, chronic catabolism and degeneration of the intraarticular environment may have primed the cellular expression within each tissue towards a similar phenotype. PCA revealed a clustering, rather than a clear separation, of 5 of 6 donors and a single DE gene was found between the OA ACL‐s and OA ACL‐c (Supporting Information Figure S3). Similarly, there was a decrease in the number of DE genes between the OA ACL‐s and synovium (Supporting Information Figure S4). Significant changes in GO analysis were centered around cellular metabolism rather than ECM structure or inflammation. The cells from these OA tissues likely experienced significant pro‐inflammatory cues which drove similar catabolic expression patterns. This effect was seen in the ubiquitous upregulation of IL6, MMP3, and COX2 in all three tissue types when isolated fibroblasts were challenged with IL1β.

The synovium plays a crucial immunological role within intraarticular tissues, mediated by resident macrophages participating in the inflammatory response following the onset of OA [41, 42, 43]. Since the current work did not employ cell sorting or enrichment, the RNAseq data reflects a mixed population of fibroblasts and macrophages. The synovial macrophages can secrete chemokines CCL2 and CXCL3, both of which were upregulated in the healthy ACL‐s and synovium. This suggests an innate capacity for immune cell recruitment by macrophages in both tissues [44, 45, 46]. DE of CCL2 and CXCL3 between the ACL‐s and synovium demonstrates that there is a tissue‐specific mechanism of action for immune response. Higher expression of both markers in the synovium indicates a greater propensity for inflammatory response than the ACL‐s. Upregulation of CCL2 in the synovium could reflect its readiness to recruit macrophages during early inflammatory responses, whereas expression in the ACL‐s may be more tightly regulated to preserve tissue integrity. In order to explore the translation of these gene expression patterns in the ACL‐s and synovium, we performed tissue immunohistochemistry.

As the OA ACL‐s and synovium exhibit the barrier‐forming tight‐junction protein, CLDN5, with CX3CR1+ macrophages, they may both serve a transport governing function, into the ACL‐c and knee joint, respectively. We observed the presence of CX3CR1 throughout the ACL‐s and synovium under both healthy and inflammatory conditions. This expands our understanding of synovial resident macrophages that can be observed under pro‐inflammatory conditions by demonstrating their presence within the synovium regardless of disease‐state [21]. While CLDN5 was present in the ACL‐s of both OA and healthy donors, non‐OA synovium did not stain positively in either the intima or subintima. The differential expression of CLDN5 between the healthy ACL‐s and synovium further points to the tissues being distinct structures under physiologic conditions that serve different purposes within the knee.

Fibroblasts derived from the OA ACL and cultured in vitro responded strongly to a cytokine challenge with IL1β. Pro‐inflammatory genes IL6, MMP3, and COX2 were all significantly upregulated following exposure to IL1β, indicative of its effect on OA‐related pathways [47]. Furthermore, differential expression of the ACL‐s from the ACL‐c was exhibited by an upregulation in the expression of CLDN5, pointing to a conserved relationship between gene and protein expression in the sheath. The ACL‐s still underwent a significant change in response to injury of the ACL despite passaging of the cells, which demonstrates a conservation of their ability to respond to pro‐inflammatory cytokines in vitro. The synovium‐derived fibroblasts differentially expressed TNC, pointing to the fibrotic response of synovium to inflammation [48].

There are several limitations associated with the work in this paper. First, the ACL‐c and ACL‐s were studied in isolation from the multi‐tissue crosstalk environment within the knee. In this regard, investigations using more biofidelic in vitro co‐culture models may be required to elucidate the ability of the ACL‐s to differentially express a protective barrier phenotype through CX3CR1+ macrophages. Additionally, the cells and tissues used in the current study were sourced from end‐stage OA patients. It is possible that the demonstrated dynamic effects of the ACL‐s could be better characterized with a larger sample of non‐diseased or mildly arthritic tissues. However, recovery of non‐diseased tissues from multiple, diverse, donors, remains a ubiquitous research challenge, as fresh young cadavers are scarce and whole ACL tissues are not routinely sacrificed in non‐arthroplasty knee surgery. One potential solution may lie in RNAseq of embedded tissues, which has recently been demonstrated [49]. This would allow for sequencing of younger, non‐diseased, fresh‐fixed ACL‐c and ACL‐s. Another possible avenue of investigation could lie in the ACL remnants from the population of patients who have suffered an isolated ACL rupture and are awaiting reconstructive surgery. Future work that will continue to expand on these findings to investigate the presence of CX3CR1+ macrophages in non‐diseased synovium and ACL‐s is ongoing. Immunostaining or proteomic analyzes of SF samples for chemokines CCL2 and CXCL3 will be used to determine the degree to which CX3CR1+ macrophages respond to proinflammatory signaling. Although CX3CR1 expression was not significantly different across the three tissue types, this result may not be surprising due to the nature of in vitro culture in this study. The expansion and passaging of cells derived from the ACL‐s and synovium has been shown to select for fibroblasts, which might not express CX3CR1 [50]. Selection for the macrophage population through flow cytometry could be a future avenue to pursue to compare the expression patterns of the cell populations in the ACL‐s and synovium. Nevertheless, this study demonstrates the activation of a previously uncharacterized, innate protective system within the structure of the human ACL following the onset of OA.

In summary, this work highlights the distinct structural and functional properties of the ACL‐s, with implications for understanding its response to injury and disease. A significantly lower diffusivity among the OA tissues combined with the immunologic barrier forming proteins CX3CR1 and CLDN5 indicate that the ACL‐s serves an innate protective role for the ACL‐c. During surgical primary repair or reconstruction, consideration should be given to maintain the integrity or recapitulate the ACL‐s to protect the ACL‐c from damaging cytokines following injury and during the progression of PTOA. Future work will aim to optimize tissue‐engineered strategies to harness the structure and function of the sheath for translation to clinical practice.

Author Contributions

James Rogot: experimental design, development of tissue and RNA isolation techniques, performance of experiments, data acquisition and analysis, writing, drafting, and revising article. Jennifer A. Kunes, Kurt S. Holuba: experimental design, performance of experiments, data acquisition and analysis, and final approval of version to be published. Roshan P. Shah, Christopher S. Ahmad: experimental design, tissue recovery, and final approval of version to be published. Stavros Thomopoulos and Clark T. Hung: experimental design, data analysis, drafting and revising article. Beth G. Ashinsky: experimental design, development of tissue and RNA isolation techniques, performance of experiments, data acquisition and analysis, drafting and revising article.

Supporting information

Figure S1: Healthy vs. OA ACL‐s (A) PCA plot. (B) Volcano plot of DE genes. (C) GO analysis.

JOR-44-0-s004.pdf (172.5KB, pdf)

Figure S2: Healthy vs. OA ACL‐c (A) PCA plot. (B) Volcano plot of DE genes. (C) GO analysis.

JOR-44-0-s003.pdf (168.9KB, pdf)

Figure S3: OA ACL‐s vs. OA ACL‐c. (A) PCA plot. (B) Volcano plot of DE genes.

JOR-44-0-s002.pdf (98.5KB, pdf)

Figure S4: OA ACL‐s vs. OA synovium. (A) PCA plot. (B) Volcano plot of DE genes. (C) GO analysis.

JOR-44-0-s001.pdf (153KB, pdf)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  • 1. Sherman S. L., Calcei J., Ray T., et al., “ACL Study Group Presents the Global Trends in ACL Reconstruction: Biennial Survey of the ACL Study Group,” Journal of ISAKOS 6, no. 6 (2021): 322–328, 10.1136/jisakos-2020-000567. [DOI] [PubMed] [Google Scholar]
  • 2. Guilak F., “Biomechanical Factors in Osteoarthritis,” Best Practice and Research. Clinical Rheumatology 25, no. 6 (2011): 815–823, 10.1016/j.berh.2011.11.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Kiapour A. M. and Murray M. M., “Basic Science of Anterior Cruciate Ligament Injury and Repair,” Bone and Joint Research 3, no. 2 (2014): 20–31, 10.1302/2046-3758.32.2000241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Kaplan N., Wickiewicz T. L., and Warren R. F., “Primary Surgical Treatment of Anterior Cruciate Ligament Ruptures: A Long‐Term Follow‐Up Study,” American Journal of Sports Medicine 18, no. 4 (1990): 354–358, 10.1177/036354659001800404. [DOI] [PubMed] [Google Scholar]
  • 5. Strand T., Mølster A., Hordvik M., and Krukhaug Y., “Long‐Term Follow‐Up After Primary Repair of the Anterior Cruciate Ligament: Clinical and Radiological Evaluation 15–23 Years Postoperatively,” Archives of Orthopaedic and Trauma Surgery 125, no. 4 (2005): 217–221, 10.1007/s00402-004-0766-2. [DOI] [PubMed] [Google Scholar]
  • 6. Marshall J. L., Warren R. F., Wickiewicz T. L., and Reider B., “The Anterior Cruciate Ligament: A Technique of Repair and Reconstruction.” https://oce-ovid-com.ezproxy.cul.columbia.edu/article/00003086-197909000-00014/HTML.
  • 7. O'donoghue D. H., Frank G. R., Jeter G. L., Johnson W., Zeiders J. W., and Kenyon R., “Repair and Reconstruction of the Anterior Cruciate Ligament in Dogs: Factors Influencing Long‐Term Results,” Journal of Bone and Joint Surgery 53, no. 4 (1971): 710–718. [PubMed] [Google Scholar]
  • 8. Yoo H. and Marappa‐Ganeshan R., “Anatomy, Bony Pelvis and Lower Limb, Knee Anterior Cruciate Ligament.” StatPearls (StatPearls Publishing, 2024), http://www.ncbi.nlm.nih.gov/books/NBK559233/. [PubMed] [Google Scholar]
  • 9. Toy B. J., Yeasting R. A., Morse D. E., and McCann P., “Arterial Supply to the Human Anterior Cruciate Ligament,” Journal of Athletic Training 30, no. 2 (1995): 149–152. [PMC free article] [PubMed] [Google Scholar]
  • 10. Murakami Y., Ochi M., Ikuta Y., and Higashi Y., “Permeation From the Synovial Fluid as Nutritional Pathway for the Anterior Cruciate Ligament in Rabbits,” Acta Physiologica Scandinavica 158, no. 2 (1996): 181–187, 10.1046/j.1365-201X.1996.537301000.x. [DOI] [PubMed] [Google Scholar]
  • 11. Haus J. and Refior H. J., “A Study of the Synovial and Ligamentous Structure of the Anterior Cruciate Ligament,” International Orthopaedics 11, no. 2 (1987): 117–124, 10.1007/BF00266696. [DOI] [PubMed] [Google Scholar]
  • 12. Guang Y., McGrath T. M., Klug N. R., et al., “Combined Experimental Approach and Finite Element Modeling of Small Molecule Transport Through Joint Synovium to Measure Effective Diffusivity,” Journal of Biomechanical Engineering 142, no. 4 (2019): 0410101, 10.1115/1.4044892. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Stefani R. M., Halder S. S., Estell E. G., et al., “A Functional Tissue‐Engineered Synovium Model to Study Osteoarthritis Progression and Treatment,” Tissue engineering. Part A 25, no. 7–8 (2019): 538–553, 10.1089/ten.tea.2018.0142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Chen B., Zhang J., Nie D., Zhao G., Fu F. H., and Wang J. H.‐C., “Characterization of the Structure of Rabbit Anterior Cruciate Ligament and its Stem/Progenitor Cells,” Journal of Cellular Biochemistry 120, no. 5 (2019): 7446–7457, 10.1002/jcb.28019. [DOI] [PubMed] [Google Scholar]
  • 15. Arnoczky S. P., “Anatomy of the Anterior Cruciate Ligament,” Clinical Orthopaedics and Related Research 172, no. 172 (1983): 19–25. [PubMed] [Google Scholar]
  • 16. Duthon V. B., Barea C., Abrassart S., Fasel J. H., Fritschy D., and Ménétrey J., “Anatomy of the Anterior Cruciate Ligament,” Knee Surgery, Sports Traumatology, Arthroscopy 14, no. 3 (2006): 204–213, 10.1007/s00167-005-0679-9. [DOI] [PubMed] [Google Scholar]
  • 17. Petersen W. and Tillmann B., “Anatomy and Function of the Anterior Cruciate Ligament,” Orthopade 31, no. 8 (2002): 710–718, 10.1007/s00132-002-0330-0. [DOI] [PubMed] [Google Scholar]
  • 18. Petersen W. and Zantop T., “Anatomy of the Anterior Cruciate Ligament With Regard to Its Two Bundles,” Clinical Orthopaedics and Related Research 454 (2007): 35–47, 10.1097/BLO.0b013e31802b4a59. [DOI] [PubMed] [Google Scholar]
  • 19. Skelley N. W., Castile R. M., Cannon P. C., Weber C. I., Brophy R. H., and Lake S. P., “Regional Variation in the Mechanical and Microstructural Properties of the Human Anterior Cruciate Ligament,” American Journal of Sports Medicine 44, no. 11 (2016): 2892–2899, 10.1177/0363546516654480. [DOI] [PubMed] [Google Scholar]
  • 20. Rogot J., Kunes J. A., Kim M. K., et al., “Biologic and Transport Functions for the ACL Sheath Under Inflammatory Conditions,” Transactions of the Orthopedic Research Society, (2024).
  • 21. Culemann S., Grüneboom A., Nicolás‐Ávila J. Á., et al., “Locally Renewing Resident Synovial Macrophages Provide a Protective Barrier for the Joint,” Nature 572, no. 7771 (2019): 670–675, 10.1038/s41586-019-1471-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Ateschrang A., Schreiner A. J., Ahmad S. S., et al., “Improved Results of ACL Primary Repair in One‐Part Tears With Intact Synovial Coverage,” Knee Surgery, Sports Traumatology, Arthroscopy 27, no. 1 (2019): 37–43, 10.1007/s00167-018-5199-5. [DOI] [PubMed] [Google Scholar]
  • 23. Albro M. B., Rajan V., Li R., Hung C. T., and Ateshian G. A., “Characterization of the Concentration‐Dependence of Solute Diffusivity and Partitioning in a Model Dextran‐Agarose Transport System,” Cellular and Molecular Bioengineering 2, no. 3 (2009): 295–305, 10.1007/s12195-009-0076-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Bolger A. M., Lohse M., and Usadel B., “Trimmomatic: A Flexible Trimmer for Illumina Sequence Data,” Bioinformatics 30, no. 15 (2014): 2114–2120, 10.1093/bioinformatics/btu170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Dobin A., Davis C. A., Schlesinger F., et al., “STAR: Ultrafast Universal RNA‐Seq Aligner,” Bioinformatics 29, no. 1 (2013): 15–21, 10.1093/bioinformatics/bts635. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Liao Y., Smyth G. K., and Shi W., “Featurecounts: An Efficient General Purpose Program for Assigning Sequence Reads to Genomic Features,” Bioinformatics 30, no. 7 (2014): 923–930, 10.1093/bioinformatics/btt656. [DOI] [PubMed] [Google Scholar]
  • 27. Love M. I., Huber W., and Anders S., “Moderated Estimation of Fold Change and Dispersion for RNA‐Seq Data With DESeq2,” Genome Biology 15, no. 12 (2014): 550, 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Subhash S. and Kanduri C., “GeneSCF: A Real‐Time Based Functional Enrichment Tool With Support for Multiple Organisms,” BMC Bioinformatics 17, no. 1 (2016): 365, 10.1186/s12859-016-1250-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Marshall B. P., Ashinsky B. G., Ferrer X. E., et al., “The Subacromial Bursa Is a Key Regulator of the Rotator Cuff and a New Therapeutic Target for Improving Repair,” STM 16, no. 744 (2024): eadd8273, 10.1126/scitranslmed.add8273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Amiel D., Billings E., and Harwood F. L., “Collagenase Activity in Anterior Cruciate Ligament: Protective Role of the Synovial Sheath,” Journal of Applied Physiology 69, no. 3 (1990): 902–906, 10.1152/jappl.1990.69.3.902. [DOI] [PubMed] [Google Scholar]
  • 31. He W., Wang M., Wang Y., Wang Q., and Luo B., “Plasma and Synovial Fluid CXCL12 Levels Are Correlated With Disease Severity in Patients With Knee Osteoarthritis,” Journal of Arthroplasty 31, no. 2 (2016): 373–377, 10.1016/j.arth.2015.08.018. [DOI] [PubMed] [Google Scholar]
  • 32. Wang Q., Rozelle A. L., Lepus C. M., et al., “Identification of a Central Role for Complement in Osteoarthritis,” Nature Medicine 17, no. 12 (2011): 1674–1679, 10.1038/nm.2543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Xu J. X., Xu F. Z., Furbish A., et al., “Inhibition of Complement C3 Prevents Osteoarthritis Progression in Guinea Pigs by Blocking STAT1 Activation,” Communications Biology 7, no. 1 (2024): 370, 10.1038/s42003-024-06051-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Yung P. S.‐H., Lee Y.‐W., Fu S.‐C., Chen C.‐H., Rolf C. G., and Chan K.‐M., “Differential MMP 1 and MMP 13 Expression in Proliferation and Ligamentization Phases of Graft Remodeling in Anterior Cruciate Ligament Reconstruction,” Connective Tissue Research 62, no. 6 (2021): 681–688, 10.1080/03008207.2020.1862806. [DOI] [PubMed] [Google Scholar]
  • 35. Nakahara H., Hasegawa A., Otabe K., et al., “Transcription Factor Mohawk and the Pathogenesis of Human Anterior Cruciate Ligament Degradation,” Arthtitis and Rheumatism 65, no. 8 (2013), 10.1002/art.38020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Zhang Q., Ji Q., Wang X., et al., “SOX9 is a Regulator of ADAMTSs‐Induced Cartilage Degeneration at the Early Stage of Human Osteoarthritis,” Osteoarthritis and Cartilage 23, no. 12 (2015): 2259–2268, 10.1016/j.joca.2015.06.014. [DOI] [PubMed] [Google Scholar]
  • 37. Kelwick R., Desanlis I., Wheeler G. N., and Edwards D. R., “The ADAMTS (A Disintegrin and Metalloproteinase With Thrombospondin Motifs) Family,” Genome Biology 16, no. 1 (2015): 113, 10.1186/s13059-015-0676-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Cheng J., Li M., and Bai R., “The Wnt Signaling Cascade in the Pathogenesis of Osteoarthritis and Related Promising Treatment Strategies,” Frontiers in Physiology 13 (2022): 954454, 10.3389/fphys.2022.954454. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Nakamura Y., Nawata M., and Wakitani S., “Expression Profiles and Functional Analyses of Wnt‐Related Genes in Human Joint Disorders,” American Journal of Pathology 167, no. 1 (2005): 97–105, 10.1016/S0002-9440(10)62957-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Hasegawa A., Otsuki S., Pauli C., et al., “Anterior Cruciate Ligament Changes in Human Joint in Aging and Osteoarthritis,” Arthtitis and Rheumatism 64, no. 3 (2012): 696–704, 10.1002/art.33417. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Liu Y., Hao R., Lv J., et al., “Targeted Knockdown of PGAM5 in Synovial Macrophages Efficiently Alleviates Osteoarthritis,” Bone Research 12, no. 1 (2024): 15, 10.1038/s41413-024-00318-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Sanchez‐Lopez E., Coras R., Torres A., Lane N. E., and Guma M., “Synovial Inflammation in Osteoarthritis Progression,” Nature Reviews Rheumatology 18, no. 5 (2022): 258–275, 10.1038/s41584-022-00749-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Wang D., Chai X.‐Q., Hu S.‐S., and Pan F., “Joint Synovial Macrophages as a Potential Target for Intra‐Articular Treatment of Osteoarthritis‐Related Pain,” Osteoarthritis and Cartilage 30, no. 3 (2022): 406–415, 10.1016/j.joca.2021.11.014. [DOI] [PubMed] [Google Scholar]
  • 44. Fang C., Zhong R., Lu S., et al., “TREM2 Promotes Macrophage Polarization From M1 to M2 and Suppresses Osteoarthritis Through the NF‐κB/CXCL3 Axis,” International Journal of Biological Sciences 20, no. 6 (2024): 1992–2007, 10.7150/ijbs.91519. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45. Raghu H., Lepus C. M., Wang Q., et al., “CCL2/CCR2, but not CCL5/CCR5, Mediates Monocyte Recruitment, Inflammation and Cartilage Destruction in Osteoarthritis,” Annals of the Rheumatic Diseases 76, no. 5 (2017): 914–922, 10.1136/annrheumdis-2016-210426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Zheng Y., Wei K., Jiang P., et al., “Macrophage Polarization in Rheumatoid Arthritis: Signaling Pathways, Metabolic Reprogramming, and Crosstalk With Synovial Fibroblasts,” Frontiers in Immunology 15 (2024): 1394108, 10.3389/fimmu.2024.1394108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47. Daheshia M. and Yao J. Q., “The Interleukin 1β Pathway in the Pathogenesis of Osteoarthritis,” Journal of Rheumatology 35, no. 12 (2008): 2306–2312, 10.3899/jrheum.080346. [DOI] [PubMed] [Google Scholar]
  • 48. Bhattacharyya S., Midwood K. S., and Varga J., “Tenascin‐C in Fibrosis in Multiple Organs: Translational Implications,” Seminars in Cell and Developmental Biology 128 (2022): 130–136, 10.1016/j.semcdb.2022.03.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49. Matsunaga H., Arikawa K., Yamazaki M., et al., “Reproducible and Sensitive Micro‐Tissue RNA Sequencing From Formalin‐Fixed Paraffin‐Embedded Tissues for Spatial Gene Expression Analysis,” Scientific Reports 12, no. 1 (2022): 19511, 10.1038/s41598-022-23651-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50. Zhao J., Ouyang Q., Hu Z., et al., “A Protocol for the Culture and Isolation of Murine Synovial Fibroblasts,” Biomedical Reports 5, no. 2 (2016): 171–175, 10.3892/br.2016.708. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1: Healthy vs. OA ACL‐s (A) PCA plot. (B) Volcano plot of DE genes. (C) GO analysis.

JOR-44-0-s004.pdf (172.5KB, pdf)

Figure S2: Healthy vs. OA ACL‐c (A) PCA plot. (B) Volcano plot of DE genes. (C) GO analysis.

JOR-44-0-s003.pdf (168.9KB, pdf)

Figure S3: OA ACL‐s vs. OA ACL‐c. (A) PCA plot. (B) Volcano plot of DE genes.

JOR-44-0-s002.pdf (98.5KB, pdf)

Figure S4: OA ACL‐s vs. OA synovium. (A) PCA plot. (B) Volcano plot of DE genes. (C) GO analysis.

JOR-44-0-s001.pdf (153KB, pdf)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.


Articles from Journal of Orthopaedic Research are provided here courtesy of Wiley

RESOURCES