Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Synovial fluid proteome changes in ACL injury-induced posttraumatic osteoarthritis: Proteomics analysis of porcine knee synovial fluid

  • Ata M. Kiapour ,

    Roles Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing – original draft

    Ata.Kiapour@childrens.harvard.edu

    Affiliation Department of Orthopaedic Surgery, Boston Children's Hospital, Harvard Medical School, Boston, MA, United States of America

  • Jakob T. Sieker,

    Roles Formal analysis, Investigation, Methodology, Supervision, Writing – review & editing

    Affiliation Department of Orthopaedic Surgery, Boston Children's Hospital, Harvard Medical School, Boston, MA, United States of America

  • Benedikt L. Proffen,

    Roles Formal analysis, Investigation, Methodology, Writing – review & editing

    Affiliation Department of Orthopaedic Surgery, Boston Children's Hospital, Harvard Medical School, Boston, MA, United States of America

  • TuKiet T. Lam,

    Roles Data curation, Formal analysis, Investigation, Methodology, Resources, Supervision, Writing – review & editing

    Affiliations Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, CT, United States of America, MS & Proteomics Resource, W.M. Keck Biotechnology Resource Laboratory, Yale University, New Haven, CT, United States of America

  • Braden C. Fleming,

    Roles Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Visualization, Writing – original draft

    Affiliation Department of Orthopaedics, Warren Alpert Medical School of Brown University & Rhode Island Hospital, Providence, RI, United States of America

  • Martha M. Murray

    Roles Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing – original draft

    Affiliation Department of Orthopaedic Surgery, Boston Children's Hospital, Harvard Medical School, Boston, MA, United States of America

Abstract

Surgical transection of the anterior cruciate ligament (ACL) in the porcine model leads to posttraumatic osteoarthritis if left untreated. However, a recently developed surgical treatment, bridge-enhanced ACL repair, prevents further cartilage damage. Since the synovial fluid bathes all the intrinsic structures of knee, we reasoned that a comparative analysis of synovial fluid protein contents could help to better understand the observed chondroprotective effects of the bridge-enhanced ACL repair. We hypothesized that post-surgical changes in the synovial fluid proteome would be different in the untreated and repaired knees, and those changes would correlate with the degree of cartilage damage. Thirty adolescent Yucatan mini-pigs underwent unilateral ACL transection and were randomly assigned to either no further treatment (ACLT, n = 14) or bridge-enhanced ACL repair (BEAR, n = 16). We used an isotopically labeled high resolution LC MS/MS-based proteomics approach to analyze the protein profile of synovial fluid at 6 and 12 months after ACL transection in untreated and repaired porcine knees. A linear mixed effect model was used to compare the normalized protein abundance levels between the groups at each time point. Bivariate linear regression analyses were used to assess the correlations between the macroscopic cartilage damage (total lesion area) and normalized abundance levels of each of the identified secreted proteins. There were no significant differences in cartilage lesion area or quantitative abundance levels of the secreted proteins between the ACLT and BEAR groups at 6 months. However, by 12 months, greater cartilage damage was seen in the ACLT group compared to the BEAR group (p = 0.005). This damage was accompanied by differences in the abundance levels of secreted proteins, with higher levels of Vitamin K-dependent protein C (p = 0.001), and lower levels of Apolipoprotein A4 (p = 0.021) and Cartilage intermediate layer protein 1 (p = 0.049) in the ACLT group compared to the BEAR group. There were also group differences in the secreted proteins that significantly changed in abundance between 6 and 12 months in ACLT and BEAR knees. Increased concentration of Ig lambda-1 chain C regions and decreased concentration of Hemopexin, Clusterin, Coagulation factor 12 and Cartilage intermediate layer protein 1 were associated with greater cartilage lesion area. In general, ACLT knees had higher concentrations of pro-inflammatory proteins and lower concentrations of anti-inflammatory proteins than BEAR group. In addition, the ACLT group had a lower and declining synovial concentrations of CILP, in contrast to a consistently high abundance of CILP in repaired knees. These differences suggest that the knees treated with bridge-enhanced ACL repair may be maintaining an environment that is more protective of the extracellular matrix, a function which is not seen in the ACLT knees.

Introduction

Anterior cruciate ligament (ACL) injuries are has been linked to increased risk of posttraumatic osteoarthritis (OA) in humans and animal models [13]. Synovial fluid has been an attractive source to identify new biomarkers for monitoring joint health and a better understanding of the disease pathophysiology. This attraction is primarily due to the fact the synovial fluid bathes all the intrinsic structures of diarthrodial joints, including articular cartilage and synovium, both of which have shown to be actively involved in OA development [4, 5]. Additionally, alterations in these structures due to OA may be directly reflected in the composition of synovial fluid, which could be correlated to disease severity and progression. Recent advances in high-throughput and sensitive mass spectrometry (MS)-based approaches have facilitated protein profiling of complex biological fluids including synovial fluid. As such, this technology has emerged as a powerful and reproducible technique to identify proteins involved in disease etiology and pathogenesis, as well as potential biomarkers for a range of diseases, including arthritis [611].

Recently, a biologically augmented ACL repair procedure, bridge-enhanced ACL repair, has shown to be successful in reducing macroscopic evidence of posttraumatic OA following ACL injury in porcine knees [12]. This new surgical technique uses a combination of a novel extracellular matrix-based scaffold to augment a suture repair of the torn ACL [13]. Using the porcine ACL transection model, where the untreated knee progresses to posttraumatic OA and the repaired knee promotes cartilage preservation, allows us to compare these two different consequences of an identical surgical injury and potentially identify mechanisms and biomarkers for posttraumatic OA development.

Compared to other animal models, the porcine knee has been shown to be closest to the human based on its size, anatomy and functional dependency on the ACL [14]. Furthermore, porcine knee develops posttraumatic OA following ACL transection in a pattern similar to that reported in humans, but at a faster rate, with the joint changes at 1 year reflective of those seen at 10–15 years after ACL reconstruction in humans [12]. This faster onset of post-traumatic OA allows for more rapid assessment of factors that may influence the development of posttraumatic OA following ACL injury and treatment. The high degree of similarity between human and porcine synovial fluid [15]. further justifies the porcine knee as a suitable model to study the biology of OA.

In the current work, we used an isotopically labeled high resolution LC MS/MS-based proteomics approach to analyze the protein profile of synovial fluid at 6 and 12 months after ACL transection in untreated and repaired porcine knees. Our primary aim was to determine how the synovial fluid proteome differs between the two groups in an effort to identify candidate proteins that may be associated with the development of posttraumatic OA. We hypothesized that the development of macroscopic cartilage damage following surgical ACL transection would be accompanied by differential changes in synovial fluid proteome in untreated knees compared to repaired joints.

Materials and methods

Study design

This study was approved by the Institutional Animal Care and Use Committee at Boston Children's Hospital. The study was designed as a randomized controlled large animal trial with cross-sectional outcome assessments at two post-injury time points. The study was designed as a randomized controlled large animal trial with cross-sectional outcome assessments at two post-injury time points. A total of 30 adolescent Yucatan mini-pigs (ages 14–16 months) underwent unilateral ACL transection and were randomly treated with bridge-enhanced ACL repair (BEAR, n = 16) or left untreated (ACLT, n = 14). Animals were obtained from LoneStar Laboratory Swine (Sioux Center, IA, USA). A computer based random permutation of the complete set of animal identifiers was used to randomize treatment allocation and side of unilateral surgery. Half of the animals within each treatment group were euthanized at 6 months, while the other half was followed up to 12 months post-injury. The histological, biomechanical and OA-related outcomes for these animals have been previously reported by our group [12, 16]. For the current study, synovial fluid from an additional 6 healthy intact age-matched adolescent Yucatan minipigs was also used as a control.

Surgical procedure

All animals were acclimated to the animal care facility environment for a minimum of 7 days prior to any surgical procedures. Animals allocated to ACLT or BEAR groups underwent ACL surgery under general anesthesia and postoperative housing as previously reported [12, 16]. In brief, the ACL was transected surgically and for the animals assigned to BEAR group, the torn ACL was repaired using an extracellular matrix–based scaffold soaked with autologous blood as previously described [17]. After surgery, all animals were housed in individualized pens and checked multiple times a day for 4 weeks. They were then shipped to a farm for long-term care (Coyote Consulting Corp Inc, Douglas, Massachusetts). Full weightbearing status was achieved within 48 to 72 hours and the animals were allowed ad libitum activity. Two animals (both in the ACLT group) developed subcutaneous abscesses near the jaw that were treated with short-term oral antibiotics. These animals were included in the study as there was no visual evidence of joint synovitis at the time of dissection. In the BEAR group, one animal (12 months) was shipped to the external holding facility at 2 weeks rather than 4 weeks postoperatively, a deviation in the postoperative rehabilitation, thus it was excluded from the analysis. Also, one BEAR animal (6 months) had a missing synovial fluid sample, thus was excluded from this study. The total number of animals, analyzed in this study, was 7 in each group at each time point.

After 6 and 12 months, synovial fluid was aspirated from the injured knee joints under anesthesia. For control animals, synovial fluid was obtained under anesthesia within 1 week of arrival to the animal care facility. Samples were centrifuged at 3,000g at room temperature for 10 min and the supernatant stored at −80°C. Following aspiration, animals were euthanized and the knees were harvested and stored at -20°C.

Proteomics analysis

Sample preparation.

A multiplexed Isobaric Tag for Relative and Absolute Quantitation (iTRAQ) [18] coupled with liquid chromatography with tandem mass spectrometry (LC-MS/MS) was used to identify the changes in synovial fluid protein levels. iTRAQ is a chemical labeling method, which uses multiplexed isobaric tagging reagent to label peptide mixtures which allows post label mixing of the samples (after enzymatic digestion) without adding complexity to the MS analysis [19]. This will help minimizing the quantitation variability during sample preparation prior to LC MS/MS data collection. The iTRAQ-based quantitative proteomics approach has been widely used to identify biomarkers for several diseases including end stage OA and rheumatoid arthritis [7, 8, 11].

Comparison of the peak areas and resultant peak ratios for MS/MS reporter ions was used to measure the relative abundance of proteins. Synovial fluid from each individual sample was reduced with DTT, alkylated with iodoacetamide, and digested with trypsin at 37°C overnight. Digested samples were separated across five different sets for comparison with each set consisting of seven unique samples (S1 Table). A pooled sample consisting of equal amount of proteins from all 35 total samples was made to serve as a common eighth sample across the five different sets of comparison. Per each comparison set, equal amounts of total peptides were aliquoted and labeled with the 8-plex iTRAQ reagents (iTRAQ Reagents Multiplex kit, Applied Biosystems, CA), then pooled, dried and stored until analyzed by mass spectrometry. Total digested peptides in each sample were quantified by Amino Acid Analysis (AAA; Hitachi Model L-8900) prior to iTRAQ labeling to ensure equal amounts of total peptides for each sample.

LC-MS/MS analysis.

LC-MS/MS analysis of the iTRAQ labeled samples peptides were carried out using AB SCIEX TripleTOF 5600 mass spectrometer (SCIEX, Framingham, MA) coupled to a NanoACQUITY UPLC system (Waters Inc., Milford, MA). The dried peptide samples were dissolved in 70% formic acid/water, diluted with 13 μl 0.1% trifluoroacetic acid (TFA) and quantitated at A280 absorption on a Nanodrop (ThermoFisher Scientific, Wilmington, DE). Peptides were loaded on a NanoACQUITY Symmetry C18 UPLC Trap column, washed in 9% Buffer A (0.1% formic acid in water) then run for 160 minutes over a linear gradient 99%A to 65% A at 500 nL/min. The parameters utilized on the 5600 Triple-TOF were: 2300 V IonSpray Voltage Floating, Ion Source Gas 1 of 10, Curtain Gas of 20, Interface Heater Temperature 120, with a declustering potential of 60. MS cycle time was 250 milliseconds, with a maximum 20 MS/MS spectra taken each with a 100-millisecond accumulation time, and collision energy adjusted for iTRAQ reagents.

Protein abundance level calculation and analysis.

The ProteinPilot peptide summary was filtered to only include peptides with zero mis-cleavages, high confidence peptide identifications, and two or more iTRAQ reporter ion ratios measured. We then performed a cyclic Lowess normalization [20] on the remaining peptides to compensate for any differences between iTRAQ labels. All peptides across the 5 sets were merged and normalized to the common pooled sample. This was done by subtracting the common pooled sample, replicated in each set, from the other 7 samples in each set. The purpose of this adjustment is to remove set-specific effects to facilitate comparisons between the experimental groups (i.e. Intact, ACLT and BEAR) and time points (i.e. 6 and 12 months) across all the sets. A linear mixed effect model was used to compare the normalized protein abundance levels between the groups at each time point (limma package in R v3.3.1, The R Foundation for Statistical Computing) [21]. Benjamini and Hochberg method was used to adjust the p-values to control the false discovery rate [22]. P≤0.05 considered statistically significant and used to select differentially abundant proteins between the groups (i.e. Intact, ACLT 6 months, ACLT 12 months, BEAR 6 months and BEAR 12 months). Bivariate linear regression analyses were used to assess the correlations between the macroscopic cartilage damage (total lesion area assessed at 6 or 12 months after ACLT and BEAR) and normalized abundance levels of each identified secreted proteins.

Pathway enrichment analysis.

The differentially abundant proteins were used to analyze the enrichment of specific pathways using the Process Networks ontology in the MetaCore bioinformatics suite (Thomson Reuters) [23]. Sus scrufa Ensembl identifiers and, if not recognized, gene symbols of the corresponding gene for each protein were used for as input for the analysis. Pathways were considered significantly enriched when they contain at least 2 unique proteins and adjusted P values (false discovery rate; FDR) were less than 0.05.

Results

As previously reported, macroscopic cartilage damage was observed in the ACLT knees, primarily across the medial femoral condyle [12]. While at 6 months no difference in total cartilage lesion area was observed between ACLT and BEAR knees, ACLT knees had a significantly larger total lesion area at 12 months compared to the BEAR knees (Fig 1) [12].

thumbnail
Fig 1. Development of macroscopic cartilage damage following ACL transection.

Total cartilage lesion area at 6 and 12 months after untreated ACL injury (ACLT) and bridge-enhanced ACL repair (BEAR). Data is presented as Mean ± SD. P value is derived from a one-way ANOVA with posthoc Bonferroni correction for multiple comparisons (adopted and modified with permission from Murray MM and Fleming BC, Am J Sports Med 2013 [12]) Macroscopic cartilage damage was assessed by measuring the cartilage lesions across the femoral condyles and tibial plateau in medial and lateral compartments using India ink staining and calipers [12].

https://doi.org/10.1371/journal.pone.0212662.g001

Most abundant secreted proteins found in the synovial fluid after ACL transection

A total of 1,340 unique peptides corresponding to 232 proteins were identified across all 34 samples (S2 Table). Of the 232 detected proteins, 80 (34%) were secreted extracellular proteins. The subcellular localization patterns of the identified proteins were similar to those reported in knee synovial fluid proteins of humans with knee OA [24]. The top 20 most abundant secreted proteins in the synovial fluid of each surgical group are listed in Table 1. Of the 232 detected proteins, 65 (28%) were significantly differentially abundant in at least one of the comparisons between the experimental groups (i.e. ACLT, BEAR and Intact) and / or between the time points (S3 Table), and 33 (51%) of these were secreted extracellular proteins.

thumbnail
Table 1. Top 20 most abundant secreted proteins detected in the synovial fluid in each surgical group.

The uniquely abundant proteins in each group are highlighted in bold.

https://doi.org/10.1371/journal.pone.0212662.t001

Treatment related changes in synovial fluid proteome

At 6 months, a total of 27 secreted proteins were significantly differentially abundant between intact group and surgical groups (ACLT and/or BEAR; Table 2). There were no significant differences in the normalized abundance levels of the detected secreted proteins between the ACLT and BEAR groups at 6 months.

thumbnail
Table 2. Secreted proteins that are significantly different in abundance in ACLT and/or BEAR knees compared to intact group at 6 months.

Fold change values represent the ratio of detection in the ACLT or BEAR group versus intact group; thus a fold change greater than 1 reflects higher protein abundance in the synovial fluid of the ACLT or BEAR group and a fold change less than 1 reflects higher protein abundance in the synovial fluid of the intact group.

https://doi.org/10.1371/journal.pone.0212662.t002

At 12 months, a total of 16 secreted proteins were significantly differentially abundant between intact group and surgical groups (ACLT and/or BEAR; Table 3). Compared to the BEAR group, the ACLT group had higher normalized abundance levels of vitamin K-dependent protein C (by 2.4 fold, P = 0.00126) along with lower normalized abundance levels of apolipoprotein A4 (by 0.7 fold, P = 0.0213) and cartilage intermediate layer protein (by 0.35 fold, P = 0.049).

thumbnail
Table 3. Secreted proteins that are significantly different in abundance in ACLT and/or BEAR knees compared to intact group at 12 months.

Fold change values represent the ratio of detection in the ACLT or BEAR group versus intact group; thus a fold change greater than 1 reflects higher protein abundance in the synovial fluid of the ACLT or BEAR group and a fold change less than 1 reflects higher protein abundance in the synovial fluid of the intact group.

https://doi.org/10.1371/journal.pone.0212662.t003

There were 20 secreted synovial fluid proteins with significant fold changes between the 6 to 12-month time points in ACLT and/or BEAR knees (Table 4). Nine proteins had significant fold changes from 6 to 12 months after surgery in both groups. Five proteins had a significant fold change from 6 to 12 months only after untreated ACL transection. Six proteins had a significant fold change from 6 to 12 months only after bridge-enhanced ACL repair.

thumbnail
Table 4. Secreted proteins that are significantly different in abundance between 6 and 12 months after surgery.

https://doi.org/10.1371/journal.pone.0212662.t004

Differentially abundant proteins enriched 7 biological pathways (Process Network ontology) including terms related to blood coagulation, proteolysis, inflammation and cell adhesion (Table 5).

thumbnail
Table 5. Biological pathways enriched by differentially abundant proteins at 6 months.

https://doi.org/10.1371/journal.pone.0212662.t005

Associations between protein levels and total area of macroscopic articular cartilage damage

Among all the identified secreted proteins, notable associations (R2 ≥ 0.1) were only found between the total area of the macroscopic cartilage damage across the femoral condyles and tibial plateau and the normalized abundance levels of Ig lambda-1 chain C regions (r = 0.52), Hemopexin (r = -0.40), Clusterin (r = -0.35), Coagulation factor 12 (r = -0.32) and Cartilage intermediate layer protein 1 (r = -0.33); Fig 2.

thumbnail
Fig 2. Associations between macroscopic cartilage damage and secreted proteins concentrations.

Correlations between macroscopic cartilage damage area and the normalized abundance levels of (A) Ig lambda-1 chain C regions, (B) Hemopexin, (C) Clusterin, (D) Coagulation factor 12 and (E) Cartilage intermediate layer protein 1 for both ACLT and BEAR groups at 6 and 12 months.

https://doi.org/10.1371/journal.pone.0212662.g002

Discussion

The results of this study support our hypothesis that development of macroscopic cartilage damage following an ACL injury and subsequent surgery are accompanied with changes in synovial fluid proteome with different responses in untreated knees compared to repaired joints. There was no significant difference in cartilage lesion area or quantitative protein abundance levels between the two treatment groups at 6 months. However, by 12 months after surgery, there was significantly greater cartilage damage in the knees with an untreated ACL transection than in those knees that underwent bridge-enhanced ACL repair (Fig 1). This damage was accompanied by differences in the proteome between the groups at 12 months, with the ACLT group having a higher concentration of Vitamin K-dependent Protein C (2.4 fold), and a decreased content of Apolipoprotein A4 (0.7 fold) and Cartilage intermediate layer protein 1 (0.35 fold) when compared to the BEAR knees. In addition, there was also a difference in the proteins which significantly changed in abundance between 6 and 12 months (Table 4), with Insulin-like growth factor-binding protein 2 (IGFBP2), Ig lambda-1 chain C regions (IGLC1), Hemopexin and CILP changing in the ACLT group, and significant changes in Annexin A1, alpha-1 antitrypsin, Complement C7, Apolipoprotein C3, Leukocyte elastase inhibitor and Peptidylprolyl isomerase A (PPIA) found in the BEAR group. Increased levels of IGLC1 and decreased levels of Hemopexin, Clusterin, Coagulation factor 12 and CILP were also found to correlate with greater cartilage lesion area when the knees from both groups were pooled (Fig 2).

While there were no statistically significant differences in protein abundance between the surgical groups at 6 months after surgery, the makeup of the synovial fluid proteome, and specifically the top 20 most abundant proteins in the fluid, were similar but not identical in the two surgical groups. Seventeen of the 20 most abundant secreted proteins (Table 1) were the same in ACLT and BEAR groups, 15 of which have been previously reported to be abundant in synovial fluid of human knees with OA [8, 11, 24]. For the six differences (three in each group) in the top 20 secreted proteins, the ACLT knees had 2 additional pro-inflammatory proteins (Haptoglobin and Pituitary adenylate cyclase-activating polypeptide) while the BEAR group had two protease inhibitors at 6 months (Plasminogen activator inhibitor 1 and Elafin). The pro-inflammatory role of Haptoglobin, Pituitary adenylate cyclase-activating polypeptide (PACAP), Plasminogen activation and Elastase are well documented [2528]. High synovial fluid concentrations of Haptoglobin, Plasminogen (or Plasmin) and Elastase have been reported in arthritic knees [11, 24, 2931]. PACAP-deficiency in mice has shown to result in lower inflammation in early arthritis followed by increased immune cell function and bone formation (e.g. osteophytes) in late arthritis [32]. The presence of Haptoglobin and PACAP among the most abundant proteins in the ACLT group in addition to the presence of Plasminogen activator inhibitor 1 and Elafin among the most abundant proteins in the BEAR group at 6 months might be indicative of a more pro-inflammatory environment in ACLT knees at early stages of the disease. Further studies would be needed to validate this hypothesis.

In addition to proteases and protease inhibitors, Apolipoprotein C3 and Aggrecan core protein (Aggrecan) were the other two secreted proteins which were uniquely present in the BEAR or ACLT top 20 abundance lists, respectively, at 6 months (Table 1). Apolipoprotein C3 has previously been identified in the synovial fluid of human knees with OA [24], and its plasma levels have been found to be elevated in patients with non-progressing radiographic knee OA compared to those who progress [33]. This is reflected by our finding of Apolipoprotein C3 among the most abundant proteins in the non-progressive BEAR group and not in the progressive ACLT group at 6 months (Table 1). Aggrecan has also been previously detected in synovial fluid of knees with OA [11, 24]; however, in this case, increased Aggrecan in the synovial fluid has been associated with greater cartilage damage in osteoarthritic joints [3437]. This is reflected in our study where Aggrecan was among the most abundant proteins in the ACLT group and not in BEAR or intact groups (Table 1).

At 12 months after surgery, there were significant differences in protein abundance levels between the BEAR and ACLT groups, with an increase in Vitamin K-dependent protein C and decreases in Apolipoprotein A4 and CILP seen in the ACLT group. Vitamin K-dependent protein C is an activator of several matrix metalloproteinases (MMPs), including MMP2, 9 and 13, and has shown to be significantly elevated in the synovium and synovial fluid of patients with rheumatoid arthritis (RA) and OA [38]. While Apolipoprotein A4 has been detected in the synovial fluid of human knees with OA [8, 24], their role in OA pathogenesis is not well understood and warrants further studies. Considering recent reports on the role of lipid metabolism in the OA development [39, 40], it is possible that apolipoprotein proteins affects knee OA progression by regulating the lipid metabolism and efflux. CILP is a large secreted glycoprotein with major role in cartilage scaffolding and new cartilage formation [41]. Increased CILP synthesis by chondrocytes has also been reported in early-stage OA [42, 43]. which can be indicative of joint response to repair damaged cartilage [44]. The consistently higher abundance of CILP in BEAR knees compared to intact (at 6 and 12 months) and ACLT (at 12 months) knees may suggest a more chondroprotective environment in repaired knees (Tables 2 & 3).

In addition, at 12 months, the top 20 most abundant proteins in the ACLT knees had 5 additional pro-inflammatory proteins (Vitamin K-dependent protein C, Prothrombin, PPIA and Chitinase-3-like protein 1) and 3 fewer anti-inflammatory proteins (Clusterin, Alpha-2-HS glycoprotein and Annexin A1; Table 1). The pro-inflammatory role of Vitamin K-dependent protein C, Prothrombin, PPIA and Chitinase-3-like protein 1, as well as anti-inflammatory role of Clusterin, Alpha-2-HS glycoprotein (Fetuin-A) and Annexin A1 in arthritic joints are well documented [2528, 4561]. Elevated levels of these pro-inflammatory proteins have been reported in serum and synovial fluid of patients with RA and OA [8, 24, 49, 6265]. Prophenin-2 is a porcine specific Cathelicidin antimicrobial peptide, with similar pro-inflammatory properties as LL-37 (the only identified Cathelicidin antimicrobial peptide in humans), which is regulated by macrophages, polymorphonuclear leukocytes, and keratinocytes [66]. A recent study has shown strong overexpression of Cathelicidin in joints of patients with RA and rats with pristane-induced arthritis [67]. A recent proteomic analysis has also reported LL-37 among identified synovial fluid proteins in the patients with knee OA [24]. The presence of pro-inflammatory proteins among the most abundant proteins in the ACLT group and the anti-inflammatory proteins among the most abundant proteins in the BEAR group at 12 months suggests there may be a greater pro-inflammatory environment in ACLT knees at a later stage of the disease. This is further supported by the observed significantly lower abundance of the Vitamin K-dependent protein C in BEAR knees compared to ACLT and intact groups as well as higher abundance of Annexin A1 in the BEAR knees compared to intact group.

In comparing the proteomes at 6 and 12 months after surgery (Table 4), there were significant increases in IGFBP2 and IGLC1 levels, as well as significant decreases in CILP and Hemopexin levels in the synovial fluid of ACLT knees from 6 to 12 months. IGFP2 is synthesized by synoviocytes and chondrocytes and regulates Insulin-like growth factor (IGF) [68, 69], a key promoter of cartilage growth [70, 71]. Increased levels of IGFBPs have been reported in the synovial fluid of arthritic joints [72, 73], and serum concentration has been correlated with RA disease severity [7476]. Hemopexin has a negative modulatory function on T-cell differentiation [77, 78], and exogenous administration of Hemopexin has recently been shown to result in increased glycosaminoglycan deposition [79]. In contrast to increased synovial concentrations of pro-inflammatory proteins (IGFBP2 and IGLC1) in the ACLT knees, there were significant decreases in the synovial concentrations of two pro-inflammatory proteins (PPIA and complement C7) in the BEAR knees from 6 to 12 months.

Biological pathways enriched by the changes in synovial fluid protein content were related to blood coagulation, proteolysis, inflammation and cell adhesion (Table 5). Among those, proteolysis (ECM remodeling) and inflammation (complement pathway) pathways have consistently shown to play an active role in OA pathophysiology [8, 8083]. The role of remaining 5 enriched pathways in the development of OA is less studied. Recent comparative analysis of synovial fluid in patients with healthy knees with those with knee OA (both early and late stages) have shown significant enrichment of blood coagulation pathway by differentially abundant proteins between the groups [8]. Our group has also shown enrichment of cell adhesion (platelet-endothelium-leucocyte interactions) and proteolysis (connective tissue degradation) pathways by differentially expressed genes in the synovium and cartilage within 2 weeks after ACL injury in a minipig model of posttraumatic OA [80, 81]. These findings, in addition to abovementioned evidence supporting the involvement of the proteins / genes, which enriched these pathways, in OA pathophysiology underscore their importance as potential targets to mediate the biologic joint response to injury in an effort to lower the risk of posttraumatic OA.

Limitations

Unbiased LC MS/MS proteomics techniques are primarily limited to detection of more highly abundant proteins within a complex biological matrix. Antibody-based depletion methods could have been used to improved our ability to quantify lower abundance level proteins by removing the top abundance level proteins. However, they can also result in removal of lower abundance level proteins and ultimately pose challenges in accurate quantification of protein contents [84]. In our discovery workflow, we wanted to capture a “global” representation of proteins in the synovial fluid, as such chose to avoid antibody-based depletion. To better compensate and delve deeper into the synovial fluid proteome, additional in-depth and/or targeted approaches are required to further study less abundant proteins that are likely present but not detectable or quantifiable using this technique (i.e. Lubricin, matrix metalloproteinases and cytokines). We have previously used similar MS-based proteomics approach to study the synovial fluid proteome of healthy porcine knees [15]. The substantial overlap between those findings and identified proteins here is reassuring and supports our MS-based protein discovery approach. However, further efforts are essential for orthogonal validation of the identified proteins in individual samples compared to the pooled samples in independent cohorts. Another limitation is the cross-sectional nature of the study. While the cross-sectional time points have enabled us to macroscopically assess the cartilage health, longitudinal analyses are required to better characterize the changes in synovial fluid proteome in response to joint degeneration over time. Finally, the current results only support associations between cartilage damage and certain synovial fluid proteins in early posttraumatic OA. Further functional studies are required to investigate the mechanistic nature of these associations and to determine the potential application of suggested targets as biomarkers for OA progression or as disease modifiers for development of novel therapeutics.

Conclusions

Knees treated with a surgical procedure that results in posttraumatic OA (ACLT group) had higher concentrations of pro-inflammatory proteins and lower concentrations of anti-inflammatory proteins than knees treated with a surgical procedure that does not result in post-traumatic OA (bridge-enhanced ACL repair). In addition, the group developing posttraumatic OA had a lower and declining synovial concentrations of CILP, in contrast to a consistently high abundance of CILP in repaired knees. These differences suggest that the knees treated with bridge-enhanced ACL repair may be maintaining an environment that is more protective of the extracellular matrix, a function which is not seen in the ACLT knees.

Supporting information

S1 Table. iTRAQ labeling of the synovial fluid samples.

https://doi.org/10.1371/journal.pone.0212662.s001

(XLSX)

S2 Table. Detected synovial fluid proteins.

https://doi.org/10.1371/journal.pone.0212662.s002

(XLSX)

S3 Table. Significant differentially abundant proteins between groups.

https://doi.org/10.1371/journal.pone.0212662.s003

(XLSX)

Acknowledgments

Authors would like to thank: Patrick Vavken (Boston Children's Hospital) for helping with surgical procedures, and Edward Voss and Jean Kanyo (MS & Proteomics Resource within the Keck Biotechnology Resource Laboratory at Yale University) for proteomics sample preparation and collecting the mass spectrometric proteomics data, respectively.

References

  1. 1. Oiestad BE, Holm I, Aune AK, Gunderson R, Myklebust G, Engebretsen L, et al. Knee function and prevalence of knee osteoarthritis after anterior cruciate ligament reconstruction: a prospective study with 10 to 15 years of follow-up. Am J Sports Med. 2010;38(11):2201–10. pmid:20713644.
  2. 2. Lohmander LS, Englund PM, Dahl LL, Roos EM. The long-term consequence of anterior cruciate ligament and meniscus injuries: osteoarthritis. Am J Sports Med. 2007;35(10):1756–69. pmid:17761605.
  3. 3. von Porat A, Roos EM, Roos H. High prevalence of osteoarthritis 14 years after an anterior cruciate ligament tear in male soccer players: a study of radiographic and patient relevant outcomes. Annals of the rheumatic diseases. 2004;63(3):269–73. Epub 2004/02/14. pmid:14962961; PubMed Central PMCID: PMC1754918.
  4. 4. Goldring SR, Goldring MB. Changes in the osteochondral unit during osteoarthritis: structure, function and cartilage-bone crosstalk. Nat Rev Rheumatol. 2016;12(11):632–44. pmid:27652499.
  5. 5. Loeser RF, Goldring SR, Scanzello CR, Goldring MB. Osteoarthritis: a disease of the joint as an organ. Arthritis Rheum. 2012;64(6):1697–707. pmid:22392533; PubMed Central PMCID: PMCPMC3366018.
  6. 6. Ritter SY, Collins J, Krastins B, Sarracino D, Lopez M, Losina E, et al. Mass spectrometry assays of plasma biomarkers to predict radiographic progression of knee osteoarthritis. Arthritis Res Ther. 2014;16(5):456. pmid:25287745; PubMed Central PMCID: PMCPMC4207325.
  7. 7. Balakrishnan L, Bhattacharjee M, Ahmad S, Nirujogi RS, Renuse S, Subbannayya Y, et al. Differential proteomic analysis of synovial fluid from rheumatoid arthritis and osteoarthritis patients. Clin Proteomics. 2014;11(1):1. pmid:24393543; PubMed Central PMCID: PMCPMC3918105.
  8. 8. Ritter SY, Subbaiah R, Bebek G, Crish J, Scanzello CR, Krastins B, et al. Proteomic analysis of synovial fluid from the osteoarthritic knee: comparison with transcriptome analyses of joint tissues. Arthritis Rheum. 2013;65(4):981–92. pmid:23400684; PubMed Central PMCID: PMCPMC3618606.
  9. 9. Prakash A, Rezai T, Krastins B, Sarracino D, Athanas M, Russo P, et al. Platform for establishing interlaboratory reproducibility of selected reaction monitoring-based mass spectrometry peptide assays. J Proteome Res. 2010;9(12):6678–88. pmid:20945832.
  10. 10. Addona TA, Abbatiello SE, Schilling B, Skates SJ, Mani DR, Bunk DM, et al. Multi-site assessment of the precision and reproducibility of multiple reaction monitoring-based measurements of proteins in plasma. Nat Biotechnol. 2009;27(7):633–41. pmid:19561596; PubMed Central PMCID: PMCPMC2855883.
  11. 11. Gobezie R, Kho A, Krastins B, Sarracino DA, Thornhill TS, Chase M, et al. High abundance synovial fluid proteome: distinct profiles in health and osteoarthritis. Arthritis Res Ther. 2007;9(2):R36. pmid:17407561; PubMed Central PMCID: PMCPMC1906814.
  12. 12. Murray MM, Fleming BC. Use of a bioactive scaffold to stimulate anterior cruciate ligament healing also minimizes posttraumatic osteoarthritis after surgery. The American journal of sports medicine. 2013;41(8):1762–70. pmid:23857883; PubMed Central PMCID: PMC3735821.
  13. 13. Murray MM, Flutie BM, Kalish LA, Ecklund K, Fleming BC, Proffen BL, et al. The Bridge-Enhanced Anterior Cruciate Ligament Repair (BEAR) Procedure: An Early Feasibility Cohort Study. Orthop J Sports Med. 2016;4(11):2325967116672176. pmid:27900338; PubMed Central PMCID: PMCPMC5120682.
  14. 14. Kiapour AM, Shalvoy MR, Murray MM, Fleming BC. Validation of porcine knee as a sex-specific model to study human anterior cruciate ligament disorders. Clin Orthop Relat Res. 2015;473(2):639–50. pmid:25269532; PubMed Central PMCID: PMCPMC4294889.
  15. 15. Bennike T, Ayturk U, Haslauer CM, Froehlich JW, Proffen BL, Barnaby O, et al. A normative study of the synovial fluid proteome from healthy porcine knee joints. J Proteome Res. 2014;13(10):4377–87. pmid:25160569; PubMed Central PMCID: PMCPMC4184458.
  16. 16. Proffen BL, Fleming BC, Murray MM. Histologic Predictors of Maximum Failure Loads Differ between the Healing ACL and ACL Grafts after 6 and 12 Months In Vivo. Orthop J Sports Med. 2013;1(6). pmid:25343145; PubMed Central PMCID: PMCPMC4203461.
  17. 17. Murray MM, Magarian EM, Harrison SL, Mastrangelo AN, Zurakowski D, Fleming BC. The effect of skeletal maturity on functional healing of the anterior cruciate ligament. J Bone Joint Surg Am. 2010;92(11):2039–49. pmid:20810854; PubMed Central PMCID: PMCPMC2924734.
  18. 18. Rauniyar N, Yates JR, 3rd. Isobaric labeling-based relative quantification in shotgun proteomics. J Proteome Res. 2014;13(12):5293–309. pmid:25337643; PubMed Central PMCID: PMCPMC4261935.
  19. 19. Ross PL, Huang YN, Marchese JN, Williamson B, Parker K, Hattan S, et al. Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol Cell Proteomics. 2004;3(12):1154–69. pmid:15385600.
  20. 20. Bolstad BM, Irizarry RA, Astrand M, Speed TP. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics. 2003;19(2):185–93. pmid:12538238.
  21. 21. Listgarten J, Emili A. Statistical and computational methods for comparative proteomic profiling using liquid chromatography-tandem mass spectrometry. Mol Cell Proteomics. 2005;4(4):419–34. pmid:15741312.
  22. 22. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the royal statistical society Series B (Methodological). 1995:289–300.
  23. 23. Bessarabova M, Ishkin A, JeBailey L, Nikolskaya T, Nikolsky Y. Knowledge-based analysis of proteomics data. BMC Bioinformatics. 2012;13 Suppl 16:S13. pmid:23176192; PubMed Central PMCID: PMCPMC3489533.
  24. 24. Balakrishnan L, Nirujogi RS, Ahmad S, Bhattacharjee M, Manda SS, Renuse S, et al. Proteomic analysis of human osteoarthritis synovial fluid. Clin Proteomics. 2014;11(1):6. pmid:24533825; PubMed Central PMCID: PMCPMC3942106.
  25. 25. Theilgaard-Monch K, Jacobsen LC, Nielsen MJ, Rasmussen T, Udby L, Gharib M, et al. Haptoglobin is synthesized during granulocyte differentiation, stored in specific granules, and released by neutrophils in response to activation. Blood. 2006;108(1):353–61. pmid:16543473.
  26. 26. Kim IS, Lee IH, Lee JH, Lee SY. Induction of haptoglobin by all-trans retinoic acid in THP-1 human monocytic cell line. Biochem Biophys Res Commun. 2001;284(3):738–42. pmid:11396964.
  27. 27. Shen H, Song Y, Colangelo CM, Wu T, Bruce C, Scabia G, et al. Haptoglobin activates innate immunity to enhance acute transplant rejection in mice. J Clin Invest. 2012;122(1):383–7. pmid:22156194; PubMed Central PMCID: PMCPMC3248290.
  28. 28. Vaudry D, Falluel-Morel A, Bourgault S, Basille M, Burel D, Wurtz O, et al. Pituitary adenylate cyclase-activating polypeptide and its receptors: 20 years after the discovery. Pharmacol Rev. 2009;61(3):283–357. pmid:19805477.
  29. 29. Nakano S, Ikata T, Kinoshita I, Kanematsu J, Yasuoka S. Characteristics of the protease activity in synovial fluid from patients with rheumatoid arthritis and osteoarthritis. Clin Exp Rheumatol. 1999;17(2):161–70. pmid:10342041.
  30. 30. Momohara S, Kashiwazaki S, Inoue K, Saito S, Nakagawa T. Elastase from polymorphonuclear leukocyte in articular cartilage and synovial fluids of patients with rheumatoid arthritis. Clin Rheumatol. 1997;16(2):133–40. pmid:9093794.
  31. 31. Yamagiwa H, Sarkar G, Charlesworth MC, McCormick DJ, Bolander ME. Two-dimensional gel electrophoresis of synovial fluid: method for detecting candidate protein markers for osteoarthritis. J Orthop Sci. 2003;8(4):482–90. pmid:12898298.
  32. 32. Botz B, Bolcskei K, Kereskai L, Kovacs M, Nemeth T, Szigeti K, et al. Differential regulatory role of pituitary adenylate cyclase-activating polypeptide in the serum-transfer arthritis model. Arthritis Rheumatol. 2014;66(10):2739–50. pmid:25048575; PubMed Central PMCID: PMCPMC4320777.
  33. 33. Takinami Y, Yoshimatsu S, Uchiumi T, Toyosaki-Maeda T, Morita A, Ishihara T, et al. Identification of potential prognostic markers for knee osteoarthritis by serum proteomic analysis. Biomark Insights. 2013;8:85–95. pmid:23935359; PubMed Central PMCID: PMCPMC3735238.
  34. 34. Struglics A, Larsson S, Pratta MA, Kumar S, Lark MW, Lohmander LS. Human osteoarthritis synovial fluid and joint cartilage contain both aggrecanase- and matrix metalloproteinase-generated aggrecan fragments. Osteoarthritis Cartilage. 2006;14(2):101–13. pmid:16188468.
  35. 35. Lohmander LS, Ionescu M, Jugessur H, Poole AR. Changes in joint cartilage aggrecan after knee injury and in osteoarthritis. Arthritis Rheum. 1999;42(3):534–44. pmid:10088777.
  36. 36. Poole AR, Ionescu M, Swan A, Dieppe PA. Changes in cartilage metabolism in arthritis are reflected by altered serum and synovial fluid levels of the cartilage proteoglycan aggrecan. Implications for pathogenesis. J Clin Invest. 1994;94(1):25–33. pmid:7518830; PubMed Central PMCID: PMCPMC296278.
  37. 37. Taylor EJ. Preparation for administrative positions in schools of nursing. 1931. NLN Publ. 1991;(14–2410):57–72. pmid:1795983.
  38. 38. Jackson MT, Moradi B, Smith MM, Jackson CJ, Little CB. Activation of matrix metalloproteinases 2, 9, and 13 by activated protein C in human osteoarthritic cartilage chondrocytes. Arthritis Rheumatol. 2014;66(6):1525–36. pmid:24574263.
  39. 39. Tsezou A, Iliopoulos D, Malizos KN, Simopoulou T. Impaired expression of genes regulating cholesterol efflux in human osteoarthritic chondrocytes. J Orthop Res. 2010;28(8):1033–9. pmid:20108316.
  40. 40. Naishiro Y, Suzuki C, Kimura M, Yamamoto M, Takahashi H, Sohma H, et al. [Plasma analysis of rheumatoid arthritis by SELDI]. Nihon Rinsho Meneki Gakkai Kaishi. 2007;30(3):145–50. pmid:17603254.
  41. 41. Bomer N, den Hollander W, Suchiman H, Houtman E, Slieker RC, Heijmans BT, et al. Neo-cartilage engineered from primary chondrocytes is epigenetically similar to autologous cartilage, in contrast to using mesenchymal stem cells. Osteoarthritis Cartilage. 2016;24(8):1423–30. pmid:26995110.
  42. 42. Lorenzo P, Bayliss MT, Heinegard D. A novel cartilage protein (CILP) present in the mid-zone of human articular cartilage increases with age. J Biol Chem. 1998;273(36):23463–8. pmid:9722583.
  43. 43. Lorenzo P, Bayliss MT, Heinegard D. Altered patterns and synthesis of extracellular matrix macromolecules in early osteoarthritis. Matrix Biol. 2004;23(6):381–91. pmid:15533759.
  44. 44. Tsuruha J, Masuko-Hongo K, Kato T, Sakata M, Nakamura H, Nishioka K. Implication of cartilage intermediate layer protein in cartilage destruction in subsets of patients with osteoarthritis and rheumatoid arthritis. Arthritis Rheum. 2001;44(4):838–45. pmid:11315923.
  45. 45. Arora K, Gwinn WM, Bower MA, Watson A, Okwumabua I, MacDonald HR, et al. Extracellular cyclophilins contribute to the regulation of inflammatory responses. J Immunol. 2005;175(1):517–22. pmid:15972687; PubMed Central PMCID: PMCPMC2862457.
  46. 46. Jin ZG, Melaragno MG, Liao DF, Yan C, Haendeler J, Suh YA, et al. Cyclophilin A is a secreted growth factor induced by oxidative stress. Circ Res. 2000;87(9):789–96. pmid:11055983.
  47. 47. Yang Y, Lu N, Zhou J, Chen ZN, Zhu P. Cyclophilin A up-regulates MMP-9 expression and adhesion of monocytes/macrophages via CD147 signalling pathway in rheumatoid arthritis. Rheumatology (Oxford). 2008;47(9):1299–310. pmid:18567920.
  48. 48. Kim H, Kim WJ, Jeon ST, Koh EM, Cha HS, Ahn KS, et al. Cyclophilin A may contribute to the inflammatory processes in rheumatoid arthritis through induction of matrix degrading enzymes and inflammatory cytokines from macrophages. Clin Immunol. 2005;116(3):217–24. pmid:15993649.
  49. 49. Yao Q, Li M, Yang H, Chai H, Fisher W, Chen C. Roles of cyclophilins in cancers and other organ systems. World J Surg. 2005;29(3):276–80. pmid:15706440.
  50. 50. Recklies AD, Ling H, White C, Bernier SM. Inflammatory cytokines induce production of CHI3L1 by articular chondrocytes. J Biol Chem. 2005;280(50):41213–21. pmid:16234240.
  51. 51. Di Rosa M, Malaguarnera G, De Gregorio C, D'Amico F, Mazzarino MC, Malaguarnera L. Modulation of chitotriosidase during macrophage differentiation. Cell Biochem Biophys. 2013;66(2):239–47. pmid:23152091.
  52. 52. Di Rosa M, Malaguarnera G, De Gregorio C, Drago F, Malaguarnera L. Evaluation of CHI3L-1 and CHIT-1 expression in differentiated and polarized macrophages. Inflammation. 2013;36(2):482–92. pmid:23149946.
  53. 53. Di Rosa M, De Gregorio C, Malaguarnera G, Tuttobene M, Biazzo F, Malaguarnera L. Evaluation of AMCase and CHIT-1 expression in monocyte macrophages lineage. Mol Cell Biochem. 2013;374(1–2):73–80. pmid:23129258.
  54. 54. Vaananen T, Koskinen A, Paukkeri EL, Hamalainen M, Moilanen T, Moilanen E, et al. YKL-40 as a novel factor associated with inflammation and catabolic mechanisms in osteoarthritic joints. Mediators Inflamm. 2014;2014:215140. pmid:25132728; PubMed Central PMCID: PMCPMC4124234.
  55. 55. Szychlinska MA, Trovato FM, Di Rosa M, Malaguarnera L, Puzzo L, Leonardi R, et al. Co-Expression and Co-Localization of Cartilage Glycoproteins CHI3L1 and Lubricin in Osteoarthritic Cartilage: Morphological, Immunohistochemical and Gene Expression Profiles. Int J Mol Sci. 2016;17(3):359. pmid:26978347; PubMed Central PMCID: PMCPMC4813220.
  56. 56. Xiao J, Wang XR, Hu KZ, Li MQ, Chen JW, Ma T, et al. Serum fetuin-A levels are inversely associated with clinical severity in patients with primary knee osteoarthritis. Biomarkers. 2013;18(1):51–4. pmid:23066960.
  57. 57. Albilia JB, Tenenbaum HC, Clokie CM, Walt DR, Baker GI, Psutka DJ, et al. Serum levels of BMP-2, 4, 7 and AHSG in patients with degenerative joint disease requiring total arthroplasty of the hip and temporomandibular joints. J Orthop Res. 2013;31(1):44–52. pmid:22778059.
  58. 58. Gavins FN, Hickey MJ. Annexin A1 and the regulation of innate and adaptive immunity. Front Immunol. 2012;3:354. pmid:23230437; PubMed Central PMCID: PMCPMC3515881.
  59. 59. Perretti M, D'Acquisto F. Annexin A1 and glucocorticoids as effectors of the resolution of inflammation. Nat Rev Immunol. 2009;9(1):62–70. pmid:19104500.
  60. 60. Zipfel PF, Skerka C. Complement regulators and inhibitory proteins. Nat Rev Immunol. 2009;9(10):729–40. pmid:19730437.
  61. 61. Connor JR, Kumar S, Sathe G, Mooney J, O'Brien SP, Mui P, et al. Clusterin expression in adult human normal and osteoarthritic articular cartilage. Osteoarthritis Cartilage. 2001;9(8):727–37. pmid:11795992.
  62. 62. Di Rosa M, Szychlinska MA, Tibullo D, Malaguarnera L, Musumeci G. Expression of CHI3L1 and CHIT1 in osteoarthritic rat cartilage model. A morphological study. Eur J Histochem. 2014;58(3):2423. pmid:25308850; PubMed Central PMCID: PMCPMC4194398.
  63. 63. Johansen JS, Stoltenberg M, Hansen M, Florescu A, Horslev-Petersen K, Lorenzen I, et al. Serum YKL-40 concentrations in patients with rheumatoid arthritis: relation to disease activity. Rheumatology (Oxford). 1999;38(7):618–26. pmid:10461474.
  64. 64. Johansen JS, Jensen HS, Price PA. A new biochemical marker for joint injury. Analysis of YKL-40 in serum and synovial fluid. Br J Rheumatol. 1993;32(11):949–55. pmid:8220933.
  65. 65. Buisson-Legendre N, Smith S, March L, Jackson C. Elevation of activated protein C in synovial joints in rheumatoid arthritis and its correlation with matrix metalloproteinase 2. Arthritis Rheum. 2004;50(7):2151–6. pmid:15248212.
  66. 66. Nijnik A, Hancock R. Host defence peptides: antimicrobial and immunomodulatory activity and potential applications for tackling antibiotic-resistant infections. Emerg Health Threats J. 2009;2:e1. pmid:22460279; PubMed Central PMCID: PMCPMC3167646.
  67. 67. Hoffmann MH, Bruns H, Backdahl L, Neregard P, Niederreiter B, Herrmann M, et al. The cathelicidins LL-37 and rCRAMP are associated with pathogenic events of arthritis in humans and rats. Ann Rheum Dis. 2013;72(7):1239–48. pmid:23172753.
  68. 68. Firth SM, Baxter RC. Cellular actions of the insulin-like growth factor binding proteins. Endocr Rev. 2002;23(6):824–54. pmid:12466191.
  69. 69. Mohan S, Baylink DJ. IGF-binding proteins are multifunctional and act via IGF-dependent and -independent mechanisms. J Endocrinol. 2002;175(1):19–31. pmid:12379487.
  70. 70. Luyten FP, Hascall VC, Nissley SP, Morales TI, Reddi AH. Insulin-like growth factors maintain steady-state metabolism of proteoglycans in bovine articular cartilage explants. Arch Biochem Biophys. 1988;267(2):416–25. pmid:3214163.
  71. 71. McQuillan DJ, Handley CJ, Campbell MA, Bolis S, Milway VE, Herington AC. Stimulation of proteoglycan biosynthesis by serum and insulin-like growth factor-I in cultured bovine articular cartilage. Biochem J. 1986;240(2):423–30. pmid:3545187; PubMed Central PMCID: PMCPMC1147434.
  72. 72. Busby WH Jr., Yocum SA, Rowland M, Kellner D, Lazerwith S, Sverdrup F, et al. Complement 1s is the serine protease that cleaves IGFBP-5 in human osteoarthritic joint fluid. Osteoarthritis Cartilage. 2009;17(4):547–55. pmid:18930415; PubMed Central PMCID: PMCPMC3271436.
  73. 73. Schneiderman R, Rosenberg N, Hiss J, Lee P, Liu F, Hintz RL, et al. Concentration and size distribution of insulin-like growth factor-I in human normal and osteoarthritic synovial fluid and cartilage. Arch Biochem Biophys. 1995;324(1):173–88. pmid:7503553.
  74. 74. Kormelink TG, Tekstra J, Thurlings RM, Boumans MH, Vos K, Tak PP, et al. Decrease in immunoglobulin free light chains in patients with rheumatoid arthritis upon rituximab (anti-CD20) treatment correlates with decrease in disease activity. Ann Rheum Dis. 2010;69(12):2137–44. pmid:20679475.
  75. 75. Gottenberg JE, Miceli-Richard C, Ducot B, Goupille P, Combe B, Mariette X. Markers of B-lymphocyte activation are elevated in patients with early rheumatoid arthritis and correlated with disease activity in the ESPOIR cohort. Arthritis Res Ther. 2009;11(4):R114. pmid:19627580; PubMed Central PMCID: PMCPMC2745796.
  76. 76. Gottenberg JE, Aucouturier F, Goetz J, Sordet C, Jahn I, Busson M, et al. Serum immunoglobulin free light chain assessment in rheumatoid arthritis and primary Sjogren's syndrome. Ann Rheum Dis. 2007;66(1):23–7. pmid:16569685; PubMed Central PMCID: PMCPMC1798389.
  77. 77. Lee SH, Kim EK, Kwon JE, Lee JK, Lee D, Kim SY, et al. Ssu72 attenuates autoimmune arthritis via targeting of STAT3 signaling and Th17 activation. Sci Rep. 2017;7(1):5506. pmid:28710354; PubMed Central PMCID: PMCPMC5511296.
  78. 78. McInnes IB, Schett G. Cytokines in the pathogenesis of rheumatoid arthritis. Nat Rev Immunol. 2007;7(6):429–42. pmid:17525752.
  79. 79. Kaneva MK, Greco KV, Headland SE, Montero-Melendez T, Mori P, Greenslade K, et al. Identification of Novel Chondroprotective Mediators in Resolving Inflammatory Exudates. J Immunol. 2017;198(7):2876–85. pmid:28242648.
  80. 80. Sieker JT, Ayturk UM, Proffen BL, Weissenberger MH, Kiapour AM, Murray MM. Immediate Administration of Intraarticular Triamcinolone Acetonide After Joint Injury Modulates Molecular Outcomes Associated With Early Synovitis. Arthritis Rheumatol. 2016;68(7):1637–47. pmid:26866935; PubMed Central PMCID: PMCPMC4982368.
  81. 81. Sieker JT, Proffen BL, Waller K, Chin KE, Karamachedu N, Akelman MR, et al. Transcriptional Profiling of Articular Cartilage in a Porcine Model of Early Post-traumatic Osteoarthritis. J Orthop Res. Under Review. pmid:28671352
  82. 82. Struglics A, Okroj M, Sward P, Frobell R, Saxne T, Lohmander LS, et al. The complement system is activated in synovial fluid from subjects with knee injury and from patients with osteoarthritis. Arthritis Res Ther. 2016;18(1):223. pmid:27716448; PubMed Central PMCID: PMCPMC5052889.
  83. 83. Wang Q, Rozelle AL, Lepus CM, Scanzello CR, Song JJ, Larsen DM, et al. Identification of a central role for complement in osteoarthritis. Nat Med. 2011;17(12):1674–9. pmid:22057346; PubMed Central PMCID: PMCPMC3257059.
  84. 84. Granger J, Siddiqui J, Copeland S, Remick D. Albumin depletion of human plasma also removes low abundance proteins including the cytokines. Proteomics. 2005;5(18):4713–8. pmid:16281180
  85. 85. Perez-Riverol Y, Csordas A, Bai J, Bernal-Llinares M, Hewapathirana S, Kundu DJ, et al. The PRIDE database and related tools and resources in 2019: improving support for quantification data. Nucleic acids research. 2018;47(D1):D442–D50.