Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Apr 1.
Published in final edited form as: Osteoarthritis Cartilage. 2018 Jan 31;26(4):588–599. doi: 10.1016/j.joca.2018.01.017

Molecular Influence of Anterior Cruciate Ligament Tear Remnants on Chondrocytes: A Biologic Connection between Injury and Osteoarthritis

Nobuaki Chinzei 1, Robert H Brophy 1, Xin Duan 1, Lei Cai 1, Ryan M Nunley 1, Linda J Sandell 1,2,3, Muhammad Farooq Rai 1,2,*
PMCID: PMC5871587  NIHMSID: NIHMS938226  PMID: 29391276

Abstract

Objective

Anterior cruciate ligament (ACL) injury initiates a cascade of events often leading to osteoarthritis. ACL reconstruction does not alter the course of osteoarthritis, suggesting that heightened osteoarthritis risk is likely due to factors in addition to the joint instability. We showed that torn ACL remnants express periostin (POSTN) in the acute phase of injury. Considering that ACL injury predisposes to osteoarthritis and that POSTN is associated with cartilage metabolism, we hypothesize that ACL injury affects chondrocytes via POSTN.

Design

Cartilage was obtained from osteoarthritic patients and ACL remnants were collected from patients undergoing ACL reconstruction. Crosstalk between remnants and chondrocytes was studied in a transwell co-culture system. Expression of POSTN and other anabolic and catabolic genes was assessed via real-time PCR. Immunostaining for periostin was performed in human and mouse cartilage. The impact of exogenous periostin and siRNA-mediated ablation of periostin on matrix metabolism and cell-migration was examined. Furthermore, the effect of anabolic (TGF-β1) and catabolic (IL-1β) factors on POSTN expression was investigated.

Results

ACL remnants induced expression of POSTN, MMP13 and ADAMTS4. Periostin levels were significantly higher in osteoarthritic compared to normal cartilage. Exogenous periostin induced MMP13 expression and cell-migration, repressed COL1A1 expression while POSTN-knockdown inhibited expression of both anabolic and catabolic genes and impeded cell-migration. TGF-β1 and IL-1β treatment did not alter POSTN expression but influenced chondrocyte metabolism as determined by anabolic and catabolic genes.

Conclusions

ACL remnants can exert paracrine effects on cartilage, altering cellular homeostasis. Over time, this metabolic imbalance could contribute to osteoarthritis development.

Keywords: periostin, chondrocytes, anterior cruciate ligament, osteoarthritis, matrix degradation, ACL tear

INTRODUCTION

Osteoarthritis (OA) is a total joint disease resulting from an intricate and complex disruption of the normal interaction between different tissues in the knee. Clinical data show that anterior cruciate ligament (ACL) deficient knees have altered biomechanics that likely lead to cartilage damage[1] and eventually OA[2, 3]. Evidence supports an increasing likelihood of cartilage damage with greater time from ACL injury[4, 5] and delayed ACL reconstruction increases the risk of meniscus and cartilage injuries due to persistent instability[6]. However, ACL reconstruction does not prevent the progression of OA[7, 8], suggesting that the increased risk of OA following ACL injury is not due to joint instability alone[9].

Torn ACL remnants have been shown to express high levels of periostin (POSTN), particularly in the first 3 months after injury[10]. Periostin is a matricellular secretory protein expressed in multiple collagen-rich fibrous connective tissues in addition to the ACL, including periodontal ligament, bone, periosteum and tendons[11, 12]. POSTN plays an important role in tissue repair but its role in matrix degradation and disease is not well-defined[13]. While inadequate levels of POSTN result in impaired tissue remodeling[14], overexpression may lead to chronic diseases as it can induce catabolic factors, especially MMP13, and may exacerbate OA pathology[15, 16].

Considering that ACL injury is a predisposing factor for OA and ACL remnants exhibit elevated expression of POSTN in the acute phase[10], our goal is to examine the paracrine effect of ACL remnants on chondrocytes. We hypothesize that factors secreted from ACL remnants influence chondrocyte metabolism in a manner that may contribute to OA development.

METHOD

The Institutional Review Board (#201104119) and Institutional Animal Care and Use Committee (#20150224) approved this study. All study participants provided written informed consent.

Reagents

Transwells, DMEM/F-12, HBSS, FBS, recombinant human periostin, recombinant human TGF-β1, scrambled- and POSTN-siRNA oligonucleotides, TRIzol-reagent, RNAlater, SuperScript II, and SYBR Green PCR Master-Mix were purchased from ThermoFisher Scientific. Recombinant human IL-1β, antibiotics, proteinase K, and trypsin EDTA were procured from Sigma-Aldrich. Antibodies were bought from Abcam unless indicated otherwise. Collagenase P and Pronase were purchased from Roche. TransIT-siQUEST Transfection Reagents were acquired from Mirus-Bio. RNeasy spin-columns were bought from Qiagen. Fluoro-Gel II with DAPI (4′,6-diamidino-2-phenylindole) was procured from Electron Microscopy Sciences. Invitrogen synthesized PCR primers.

ACL tear remnants

Acute (<3 months of injury, N=5) and chronic (>12 months of injury, N=6) ACL remnants were obtained from patients undergoing ACL reconstruction. One-half of each specimen was preserved in RNAlater, while other half was used for co-culture experiments.

Chondrocyte isolation and culture

Cartilage was obtained from patients undergoing total knee replacement due to end-stage primary OA (N=41). Healthy-looking cartilage was collected from non-OA patients undergoing lower extremity amputation (N=3). Cartilage from OA and non-OA subjects was harvested from the grossly intact areas of the medial and lateral tibial plateaus for consistency. Chondrocytes were isolated via enzymatic digestion as previously described[17]. Briefly, cartilage fragments were collected in DMEM/F-12 containing 10% heat-inactivated FBS, 2% antibiotics penicillin 5000-U/mL, and streptomycin 5-mg/mL (DMEM/F-12++). Fragments were digested using an enzyme cocktail [0.025% collagenase P (1.5-U/mg) and 0.025% pronase (7-U/mg)] in a spinner-flask containing DMEM/F-12++. The digest was filtered through nylon-mesh (pore-size 70-μm) and centrifuged 3 times at 1500 rpm for 5 min. The resulting pellet was washed with HBSS (without Ca++/Mg++) and suspended in DMEM/F-12++ containing L-ascorbic acid. Cells were cultured in 10-cm culture dishes in DMEM/F12++. Attached cells were collected by trypsin-EDTA and seeded in a 6-well plate (0.5×105 cells/cm2). Cells from healthy cartilage were used for RNA isolation to compare OA and normal chondrocytes. For all other experiments, only OA chondrocytes were used. For each experiment below, we used 0.5×106 chondrocytes per well (except for migration assay).

Transwell co-culture system

A transwell culture system was established using ACL remnants and chondrocytes (Fig-1). In this system, chondrocytes (0.5×106/well) from separate donors (for each experiment) were plated and transwells were placed into 6-well plates overlying the responder chondrocytes. Acute (N=5) or chronic (N=6) ACL remnants were added to the apical chamber for 48h. The membrane allowed free exchange of media and soluble molecules.

Fig 1. Transwell co-culture system.

Fig 1

Open in a new tab

ACL remnants were obtained from patients undergoing ACL reconstruction. A fragment of ACL tissue was used for RNA isolation and the remaining tissue was cultured with chondrocytes using a transwell co-culture system. In this co-culture system, transwell inserts were placed in 6-well culture plates seeded with chondrocytes. Small fragments of ACL remnants (Acute=5 biological replicates; Chronic=6 biological replicates) were placed in the apical chamber of transwells. Transwell inserts separated by 0.4-μm permeable membrane allowed free exchange of media and soluble molecules. Cells were collected after 48h of incubation and lysed for RNA isolation followed quantification of various gene transcripts by quantitative PCR.

Exogenous periostin and siRNA-ablation

To demonstrate the role of exogenous periostin, chondrocytes cultured in 6-well plates were treated with recombinant human periostin at two different concentrations (1-μg/mL and 10-μg/mL) for 48h. Furthermore, to induce specific removal of POSTN and to assess the functional consequences of POSTN-knockdown, we used siRNA oligonucleotides. A scrambled-siRNA was used as control to identify any possible off-target effects. Chondrocytes were transfected using 2.5-nM POSTN-siRNA oligonucleotides (sense: CUGACAUCAUGACAACAAATT, antisense: UUUGUUGUCAUGAUGUCAGAT) using TransIT-siQUEST transfection reagents. After incubation for 6h, cells were incubated for another 18h in fresh media.

Anabolic (TGF-β1) and catabolic (IL-1β) factors

We treated chondrocytes with human recombinant TGF-β1 (2-ng/mL or 10-ng/mL) for 6h, 24h and 48h or IL-1β for 24h at 0, 10, and 100-ng/mL to examine concentration dependence[18]. Exposure to 10-ng/mL of IL-1β for 24h has been shown to induce catabolic pathways in chondrocytes[17, 19], so we used this dose and a higher dose (100-ng/mL) to examine its effects on POSTN and other genes.

RNA isolation and real-time PCR

ACL remnants were homogenized in TRIzol-reagent using Polytron System (Kinematica AG). RNA from the homogenized tissues and lysed cells (from above experiments) was prepared using a combination of TRIzol:chloroform and RNeasy spin column method as previously described[20]. RNA was treated with DNase-I and then reverse-transcribed with SuperScript-II reverse-transcriptase to synthesize cDNA using random primers. Real-time PCR was performed with 20-μL of reaction mixture containing 10-μL of SYBR Green PCR Master Mix, 1.5-μl cDNA and 200-nM primers on the 7500 Fast Real-Time PCR System (Applied Biosystems)[20] with custom-designed primers (Table-1). GAPDH was used as the housekeeping gene. Samples were amplified with an initial activation step at 95°C for 10 min, followed by 40 cycles of denaturation at 95°C for 15 sec and annealing at 60°C for 1 min. The comparative gene expression was calculated using the 2− ΔΔCt approach.

Table 1.

Characteristics of the primers for real-time PCR

Accession # Symbol Gene name Forward Reverse Location Length (bp)
NM_006475.2 POSTN Periostin 5′-caacgcagcgctattctgac-3′ 5′-ccaagttgtcccaagcctca-3′ 453–554 101
NM_002427.3 MMP13 Matrix metallopeptidase 13 5′-tggtccaggagatgaagacc-3′ 5′-tcctcggagactggtaatgg-3′ 827–923 96
NM_005099.5 ADAMTS4 A disintegrin and metalloproteinase with thrombospondin motif 4 5′-ggctaaagcgctacctgcta-3′ 5′-gagtcacaccaccaagctgaca- 1136–1224 88
NM_000660.6 TGFB1 Transforming growth factor beta 1 5′-tcgccagagtggttatcttttg-3′ 5′-aggagcagtgggcgctaag-3′ 1413–1512 99
NM_000576.2 IL1B Interleukin 1 beta 5′-tccaggagaatgacctgagc-3′ 5′-gtgatcgtacaggtgcatcg-3′ 341–451 110
NM_000088.3 COL1A1 Collagen type I alpha 1 5′-gtgctaaaggtgccaatggt-3′ 5′-accaggttcaccgctgttac-3′ 1310–1437 127
NM_001844.4 COL2A1 Collagen type II alpha I 5′-cccagaggtgacaaaggaga-3′ 5′-caccttggtctccagaagga-3′ 3521–3638 117
NM_002046.3 GAPDH Glyceraldehyde 3-phosphate dehydrogenase 5′-acccagaagactgtggatgg-3′ 5′- \1–\2gaggcagggatgatgttctg-3′ 652–731 79
Open in a new tab

bp = base pair

Scratch migration assay

A scratch was produced with a 200-μl pipet tip[21] and the cells (1.0×106 chondrocytes in 6-well plates) were grown with and without recombinant human periostin (10-μg/mL) for 0, 6, 24, 48, and 72h. In a separate set of experiments, 1.0×106 cells were transfected in the same manner as described above using POSTN-siRNA. Briefly, chondrocytes were transfected using 2.5-nM POSTN-siRNA oligonucleotides using TransIT-siQUEST transfection reagents. Scrambled-siRNA and non-transfected chondrocytes were used as control. After incubation for 6h, cells in fresh media were monitored for migration. Cells cultured with either recombinant human periostin or POSTN-siRNA were incubated and images were taken at 0, 6, 24, 48, and 72h. Images for the same view were obtained and the migration distance was quantified at three different locations using ECLIPSE (Nikon) and QCapture-Pro v6.0 (QImaging).

Periostin immunostaining

Normal cartilage was collected from an additional 3 patients undergoing above the knee amputation and from 6 patients undergoing total knee arthroplasty. Full-thickness cartilage specimens were fixed in 10% neutral buffered formalin, paraffin embedded and sectioned (5-μm). Paraffin-embedded sections were also available from both DMM- (destabilization of medial meniscus) and sham-operated knees in 10-wks old male C57BL/6J mice (N=3)[22]. Briefly, mice were anesthetized by an intra-peritoneal administration of ketamine (100-mg/kg), xylazine (20-mg/kg) and acepromazine (10-mg/kg). Using appropriate aseptic measures, the joint capsule medial to the patellar tendon was opened in the right knee and the anterior attachment of the medial meniscotibial ligament (MMTL) was transected. The joint capsule was closed with 6-0 absorbable polypropylene sutures and the skin was closed with Vetclose skin-glue. The left knee, in which a capsular incision was made but MMTL was not severed, served as a sham. Joints were harvested 8-wks after surgery. Three sections from each knee were stained with toluidine blue staining for evaluation of their summed OA score using the OARSI scoring in a blinded fashion[22]. Sections were deparaffinized, rehydrated and treated with Proteinase-K (10-μg/mL)for antigen retrieval. After blocking with 10% goat serum, slides were incubated with the primary antibodies: rabbit anti-periostin (1:100) and rat anti-type-II collagen (1:200; in-house)[23] in 2% goat serum overnight at 4°C. Slides were then incubated with the secondary antibodies: goat anti-rabbit IgG (1:200; Alexa-Fluor 488) and goat anti-rat IgG (1:200; Alexa-Fluor 594) in 2% goat serum for 1h. Finally, slides were mounted using Fluoro-Gel II with 4′,6-diamidino-2-phenylindole (DAPI).

Confocal microscopy

Confocal microscopy was used to quantify the expression of periostin in OA cartilage. Images were acquired using ECLIPSE (Nikon) with MetaMorph v7.7 (Molecular Devices) and LSM-880 Confocal Laser Scanning Microscope (Zeiss). The quantification was performed with ZEN v2.3 (Zeiss). Briefly, for human cartilage sections, two randomly chosen fields (60x objective) per slide with two slides per patient were examined. We then selected two fields from each section to evaluate for superficial and deeper zone chondrocytes. About 4–10 chondrocytes per field were used for quantification of signals, which were represented as fluorescent intensity per cell. For mouse knee sections, we selected a site immediately outside of the cartilage damaged area and examined one field (60x) per slide. About 9–13 chondrocytes per field were studied for quantification of signal, represented as fluorescent intensity per cell.

Statistical analyses

Statistical analyses were performed using paired t-tests, Mann-Whitney tests, Kruskal-Wallis with Dunn’s multiple comparison tests and 2-way ANOVA with Bonferroni multiple tests in GraphPad Prism 6 (GraphPad software). All values are presented as mean ± standard deviation unless indicated otherwise. Where applicable, mean Ct values for each gene of interest are included in each graph.

RESULTS

Expression profile of ACL remnants

RNA analysis of acute and chronic ACL remnants confirmed our previous findings that acute remnants had significantly (P=0.004) higher expression of POSTN mRNA (2.83-fold) than chronic remnants. Moreover, acute remnants showed significantly (P=0.004) higher expression of MMP13 (5.02-fold), ADAMTS4 (5.10-fold) and COL1A1 (4.37-fold). There was no significant difference in the expression of TGFB1, IL1B and COL2A1 between acute and chronic ACL remnants (Fig-2A).

Fig 2. Gene expression profile of acute and chronic ACL remnants and the effect of these remnants on chondrocytes.

Fig 2

Open in a new tab

A) Injured ACL remnants were collected from acute (<3 months from injury, 5 biological replicates) and chronic (>12 months from injury, 6 biological replicates) tears from patients undergoing ACL reconstruction surgery. Gene expression of periostin along with other anabolic and catabolic gene transcripts was assessed by real-time PCR. Mann-Whitney test. B) To understand how chondrocyte respond to paracrine signaling from ACL remnants, we used an in-vitro co-culture system consisting of acute (5 biological replicates) or chronic ACL remnants (6 biological replicates) and primary OA chondrocytes (0.5×106 cells/well of a 6-well plate) as shown in Fig. 1. For control, cells were cultured with transwells but without any tissues in the apical chamber. Chondrocytes were collected 48h after co-culture in TRIzol-reagent for RNA preparation and real-time PCR quantification of various gene transcripts. Kruskal-Wallis with Dunn’s multiple comparison test. Similar lower-case letters in each graph represent statistical significance from each other at P<0.05. Values in parentheses above graphs represent mean Ct values for each gene of interest.

Findings from co-culture system

We observed that soluble mediators released from acute and chronic ACL remnants significantly induced POSTN mRNA expression in chondrocytes (Fig-2B). There was a 2-fold increase in expression of POSTN with acute remnants (Dunn’s P=0.006) and 3-fold increase in expression with chronic remnants (Dunn’s P=0.0005). There was a slight increase (1.3-fold) in the expression of TGFB1 with chronic remnants compared to both controls (Dunn’s P=0.006) and acute remnants (Dunn’s P=0.021). In contrast, there was a significant reduction (2.4-fold) in IL1B expression in chondrocytes co-cultured with chronic remnants compared to controls (Dunn’s P=0.003). The expression of MMP13 was significantly higher (6.4-fold) in chronic compared to control group (Dunn’s P=0.040). Likewise, both acute and chronic remnants significantly induced the expression of ADAMTS4: was induced with acute (4.7-fold; Dunn’s P=0.007) and chronic (4.8-fold; Dunn’s P=0.024) remnants compared to control. COL2A1 expression was significantly decreased in chondrocytes co-cultured with acute (2.2-fold, Dunn’s P=0.003) remnants but increased (4.5-fold; Dunn’s P=0.045) in chondrocytes cultured with chronic remnants compared to controls. Expression of COL1A1 did not change in the presence or absence of remnants.

Expression profile in normal and OA cartilage

A comparison between normal and OA chondrocytes demonstrated that OA chondrocytes had significantly higher POSTN (5.7-fold; P=0.002) expression, along with that of other OA-related genes such as TGFB1 (2.9-fold; P=0.003), IL1B (76-fold; P=0.009), MMP13 (11.8-fold; P=0.021), ADAMTS4 (5.7-fold; P=0.020), and COL1A1 (14.7-fold; P=0.047) (Fig-3). conversely, COL2A1 expression was significantly (P=0.002) less in OA chondrocytes compared to normal chondrocytes (3.3-fold).

Fig 3. Comparison of normal and OA chondrocytes for mRNA expression of POSTN and other gene transcripts.

Fig 3

Open in a new tab

To determine whether POSTN is highly expressed in OA chondrocytes, we compared the expressions profile between normal (3 biological replicates) and OA chondrocytes (16 biological replicates). Using 6-well plates, chondrocytes (0.5×106 cells per well) were cultured and then collected in Trizol-reagent for RNA preparation and real-time PCR. Mann-Whitney test. Similar lower-case letters in each graph represent statistical significance from each other at P<0.05. Values in parentheses above graphs represent mean Ct values for each gene of interest.

Periostin immunostaing in human and mouse cartilage

Protein-level differences between normal and OA cartilage by immunostaining revealed that human OA cartilage had significantly increased periostin (median=393.3; IQR=254.4) compared to normal cartilage (median=120.9; IQR=69.4). (Fig-4A–B). DMM-operated mouse knees showed significant OA as determine by the OARSI scoring system (median=62.5; IQR=1.0) compared to sham-operated knees (median=5.5; IQR=0.5) Periostin was elevated in the cartilage of DMM-operated mouse knees (median=438.7; IQR=214.0) compared to that of sham-operated knees (median=234.8; IQR=75.3) (Fig-4C–D).

Fig 4. Comparison of normal and OA cartilage for periostin immunostaining.

Fig 4

Open in a new tab

To confirm that periostin protein is present in OA cartilage, we performed the immunostaining for periostin. A) Immunostaining of human normal (3 biological replicates with each measurement made 7–10 times) (a) and OA (6 biological replicates with each measurement made in duplicates) (e) cartilage for periostin. White dotted boxes indicate area of interest in both superficial (b) and deep (c) zones of normal and superficial (f) and deep (g) zones of OA cartilage. White arrows show cells expressing periostin. 3-dimensional images were also taken by ZEISS LSM 880 Confocal Laser Scanning Microscope as shown on the left panel (d) for normal and right panel for OA cartilage (h). B) Quantification of periostin expression as measured by ZEN V2.3 software. Mann-Whitney test. C) Knee joint sections obtained from mice that underwent DMM (3 biological replicates) or sham (3 biological replicates) procedure. All measurements were made 11–13 times. DMM knees had significantly higher summed OA score based on OARSI scoring system. Paired t-test. D) Sham and DMM sections were stained for periostin at 8 weeks after surgery. White dotted boxes in sham (i) and DMM (l) knees show the area magnified in the upper right panels (j, m), while 3-dimensional images are shown in the lower right panels (k, n). Scale bar=50-μm, red=type-II collagen, green=periostin, blue=DAPI. Paired t-test E) Quantification of periostin expression was measured by ZEN V2.3 software. Paired t-test. Similar lower-case letters in each graph represent statistical significance from each other at P<0.05.

Effects of exogenous periostin and POSTN-siRNA knockdown on chondrocytes

We used recombinant human periostin and POSTN-siRNA to study the effect of periostin in chondrocytes as described below.

Exogenous periostin

We found that treatment of chondrocytes with recombinant periostin did not have any influence on the expression of POSTN. Likewise, it did not significantly alter the expression of TGFB1, COL2A1, ADMATS4, TGFB1, and COL1A1. Interestingly, it significantly increased the expression of MMP13 (1.9-fold; Dunn’s P=0.002) only at 10-μg/mL concentration (Fig-5A).

Fig 5. Chondrocytes treatment with recombinant human periostin or siRNA-mediated periostin knockdown.

Fig 5

Open in a new tab

A) Chondrocytes (0.5×106 cells per well) were treated with recombinant human periostin at two different doses (1-μg/mL and 10-μg/mL) for 48h while not-treated cells were used as control (7 biological replicates). Real-time PCR was used to assess gene expression. Kruskal-Wallis with Dunn’s multiple comparison test. B) Chondrocytes (0.5×106 cells per well) were transfected with 2.5 nM siRNA oligonucleotide directed towards POSTN using TransIT-siQUEST for 6h while non-treated and scrambled-siRNA-treated cells acted as controls (6 biological replicates). Real-time PCR was used to measure gene expression. Kruskal-Wallis with Dunn’s multiple comparison test. Similar lower-case letters in each graph represent statistical significance from each other at P<0.05. Values in parentheses above graphs represent mean Ct values for each gene of interest.

POSTN-siRNA

When we examined whether POSTN-ablation could reverse the catabolic effects observed by POSTN, POSTN knockdown significantly (3.4-fold; Dunn’s P=0.008) suppressed its own expression compared to scrambled-siRNA-treated chondrocytes (Fig-5B). Interestingly, knockdown of POSTN significantly suppressed the expression of TGFB1 (2-fold; Dunn’s P=0.018 compared to control and 2-fold; Dunn’s P=0.034 compared to scrambled-siRNA), MMP13 (1.9-fold; Dunn’s P=0.0008 compared to scrambled-siRNA) and ADAMTS4 (1.8-fold; Dunn’s P=0.0034 compared to scrambled-siRNA) expression.

Effect of TGF-β1 (anabolic protein)

We observed that there was no significant time by dose interaction of for POSTN. However, there was a significant time by dose interaction for TGFB1 (P=0.001), IL1B (P=0.005), MMP13 (P<0.0001), ADAMTS4 (P<0.0001), COL1A1 (P<0.0001) and COL2A1 (P=0.01) (Fig-6A). Interestingly, 2-ng/mL of TGF-β 1 significantly increased TGFB1 expression at 48h compared to controls (4.7-fold; Bonferroni P<0.0001). While 10-ng/mL significantly increased TGFB1 expression at 6h (5.2-fold; Bonferroni P<0.0001) and 24h (4.8-fold; Bonferroni P=0.0001), there was a significant decline in TGFB1 expression at 48h compared to 6h (2.1-fold; Bonferroni P=0.024). For IL1B expression, there was only one significant change, its expression significantly decreased (12.8-fold; Bonferroni P=0.005) in 10-ng/mL dose from 6h to 24h. Unlike TGFB1, expression of MMP13 significantly decreased in a time and dose dependent manner, while ADAMTS4 showed a similar pattern of expression as TGFB1. At 6h, the expression of COL1A1 and COL2A1 showed similar pattern, at 24h there was no effect of TGF-β1 treatment on either COL1A1 or COL2A1. At 48h, 2-ng/mL resulted in a significant increase in COL1A1 expression (8.3-fold; Bonferroni P<0.0001), which significantly dropped from 2-ng/mL to 10-ng/mL (3.1-fold; Bonferroni P<0.0001). At 48h, COL2A1 expression decreased significantly from control to 2-ng/mL (3.6-fold; Bonferroni P=0.017) and then increased significantly with 10-ng/mL concentration compared to 2-ng/mL concentration (3.6-fold; Bonferroni P=0.045).

Fig 6. Chondrocyte treatment with recombinant human TGF-beta 1 or interleukin 1 beta.

Fig 6

Open in a new tab

A) Chondrocytes were treated with recombinant human TGF-β1 (2-ng/mL or 10-ng/mL) for 6h, 24h and 48h (4–5 biological replicates). Gene expression was measured by quantitative PCR. 2-way Analysis of Variance with Bonferroni multiple comparison test. B) Chondrocytes were treated with recombinant IL-1β at two different doses (10-ng/mL and 100-ng/mL) for 24h while non-treated chondrocytes were used as control (3 biological replicates). Real-time PCR was used to measure gene expression. Kruskal-Wallis with Dunn’s multiple comparison test. Similar lower-case letters in each graph represent statistical significance from each other at P<0.05. Values in parentheses above graphs represent mean Ct values for each gene of interest (for A, the mean Ct values in each row represent control, TGF-β 1 [2-ng/mL] and TGF-β 1 [10-ng/mL] respectively).

Effect of IL-1β

The presence of IL-1β appeared to have no significant effect on the expression of POSTN, TGFB1 and COL1A1 (Fig-6B). However, as expected IL-1β resulted in significant up-regulation of IL1B (668.4-fold, Dunn’s P=0.019), MMP13 (78.9-fold, Dunn’s P=0.019) and ADAMTS4 (140.5-fold, Dunn’s P=0.019) expression at 100-ng/mL concentration and suppression of COL2A1 (2.9-fold, Dunn’s P=0.046) expression at 10-ng/mL concentration.

Finding from migration assay

The effect of periostin on cell migration was studied by a migration assay. We observed that recombinant periostin did not effect on cell migration (Fig-7A), or migration rate except for a significant increase in the rate of migration only at 6 h (2.92±1.19 in control vs. 7.26±1.08 in periostin treated; Bonferroni P=0.037) (Fig-7B). POSTN knockdown significantly impeded the migration at 48h (321.5±46.40 in control vs. 151.8±14.39 in POSTN-siRNA; Bonferroni P=0.043) and 72h (377.9±39.74 in control vs. 196.3±33.25 in POSTN-SiRNA, Bonferroni P=0.021; 372.9±52.23 in scrambled-siRNA vs. 196.3±33.25 in POSTN-SiRNA, Bonferroni P=0.028) (Figs-7C–D). The migration rate was significantly suppressed at the 6h time point (10.1±0.68 in control vs. 4.54±0.53 in POSTN-SiRNA, Bonferroni P=0.026) compared to control (Fig-7D).

Fig 7. Scratch migration assay with periostin treatment or ablation.

Fig 7

Open in a new tab

Scratch migration assay was performed by using recombinant human periostin (10-μg/mL) (A, B) (6–7 biological replicates with each measurement made in triplicates) and siRNA (C, D) (3 biological replicates with each measurement made in triplicates) to abolish POSTN at the indicated time points (0, 6, 24, 48, and 72h). Images for the same view were obtained and the migration distance was quantified using ECLIPSE and QCapture-Pro v6.0. Yellow lines show the initial wound line (0h) and red lines shows the ones that chondrocyte migrated until the indicated time point. Scale bar=100-μm. Quantification of migration distance and migration rate was performed at 6, 24, 48 and 72h for cells treated with and without periostin (B) and cells treated with scrambled-siRNA, Periostin-siRNA and not treated cells (D). 2-way Analysis of Variance with Dunn’s multiple comparison test. All values represent mean ± standard error of the mean (S.E.M). *=significant difference from control at 6h, adjusted P value=0.036; $=significant difference between control and POSTN-siRNA groups at 48h, adjusted P value=0.044; &=significant difference between control and POSTN-siRNA groups at 72h, adjusted P value=0.021; #=significant difference between scrambled-siRNA and POSTN-siRNA groups at 72h, adjusted P value=0.029; @=significant difference between control and POSTN-siRNA groups at 6h, adjusted P=0.026.

DISCUSSION

These findings identify a plausible biologic connection between ACL injury and the development of OA, focused on the potential role of POSTN. Given the association of ACL tears with OA, this information sets the stage for further investigation into whether ACL:cartilage crosstalk via POSTN, or perhaps other mediators, is causally linked to the onset of joint degeneration.

ACL tears accelerate cartilage extracellular-matrix loss and lead to OA[24]. Moreover, alterations in ACL metabolism occurred early before radiological signs of OA in a murine model of idiopathic OA[25]. POSTN expression in ACL remnants was shown to be very sensitive to time-from-injury in a previous study, with >40-fold more in acute remnants compared to chronic[10]. Moreover, acute ACL remnants had higher expression of other genes such as MMP2, ADAMTS2, ADAMTS14, COL1A1 and TGFB1[10]. Our findings demonstrate that ACL remnants have an effect on the cartilage as they can alter the expression of a number of anabolic and catabolic genes. Since up-regulation of genes that degrade extracellular-matrix (e.g. MMP13, ADAMTS4) starts within the first few days after ACL injury[26], there may be some biological advantages to early intervention.

POSTN is highly expressed in OA chondrocytes/cartilage compared to healthy chondrocytes/cartilage, consistent with previous work in this area[15, 16, 27]. Recently, microarrays profiling of cartilage and subchondral bone revealed that the POSTN is highly expressed in OA tissues[28, 29]. Furthermore, higher periostin in plasma/synovial fluid from OA patients is associated with the radiographic severity of disease[30, 31]. In other tissues, POSTN is induced in response to mechanical loading[32–34]. One of the main etiologies of post-traumatic OA is mechanical stress[35], which may stimulate POSTN expression in OA cartilage. Conversely, OA meniscus has less mRNA expression of POSTN compared to non-OA meniscus[36] possibly suggesting a tissue specific role in the knee joint. Since POSTN interacts with COL1A1[37], it may influence collagen homeostasis through altering collagen expression. MMP13 and ADAMTS4 were higher in OA chondrocytes and are involved in OA pathogenesis[38, 39]. Moreover, COL1A1 was highly expressed in OA compared to healthy chondrocytes and COL2A1 was highly expressed in healthy compared to OA chondrocytes, which is consistent with other studies[40, 41].

Our study demonstrated that treating cells with exogenous periostin increased MMP13 expression but did not affect the expression of POSTN. This finding is in line with another study in which exposing periodontal cells to exogenous periostin did not increase expression of endogenous POSTN[42]. Furthermore, this finding is consistent with our co-culture results where acute remnants (expressing high POSTN mRNA) did not induce POSTN expression in chondrocytes. These findings suggest that exogenous periostin can act as a catabolic factor that promotes cartilage degradation but does not alter periostin mRNA expression.

We observed that exogenous periostin treatment of chondrocytes stimulated MMP13 expression while siRNA-mediated POSTN-ablation inhibited expression of all genes including POSTN and MMP13, consistent with prior results[16]. POSTN knockdown has also been shown to affect collagen synthesis[37]. Although chronic ACL tears and exogenous periostin exhibit similar effects on MMP13 and ADAMTS4 expression, they differ with regards to COL1A1 expression. This discrepancy may be attributed to the fact that ACL tears have a cocktail of factors that influences chondrocyte metabolism as compared to exogenous periostin alone. It is likely that adequate levels of POSTN are needed for COL1A1 because higher levels of POSTN result in degeneration (due to MMP13 expression) and low levels of POSTN cannot induce matrix synthesis. Nevertheless, the up-regulation of MMP13 and ADAMTS4 in both cases implies that POSTN has a catabolic role. Furthermore, since periostin is located intracellularly in chondrocytes, as observed in the current study and as has been reported in a prior study[43], it is likely that intracellular periostin plays quite different roles in chondrocyte behavior than extracellular periostin. This may explain why the effect of recombinant periostin on chondrocyte gene expression in the current study is not always opposite to the effect of POSTN-knockdown.

Taken together, our results show that while excessive periostin may lead to cartilage degradation by provoking catabolic activity, insufficient periostin may cause an imbalance between synthesis and degradation of extracellular-matrix. There may be an optimal level of POSTN expression that is essential for maintaining cartilage homeostasis, which is in line with the proposal that periostin is an important factor in the remodeling process that balances appropriate versus inappropriate tissue adaptation in response to injury[13].

In contrast to a previous study[13], recombinant human TGF-β1 did not affect POSTN expression. Although it is known to be involved in OA pathogenesis[38], TGF-β1 exhibited opposite effects on the expression of MMP13 and ADAMTS4, consistent with a previous report that TGF-β1 in conjunction with IL-1β suppresses MMP13 expression[17]. This suggests that TGF-β1 has both anabolic and catabolic roles in cartilage homeostasis and OA[17, 44], and implies that MMP13 up-regulation in the presence of POSTN is independent of TGF-β1. We observed that TGF-β1 inhibited COL2A1 expression but stimulated COL1A1 expression. However, with POSTN knockdown, TGFB1 and both collagen genes showed decreased expression This may indicate that TGFB1 can influence type-II collagen synthesis[44] although its role in collagen synthesis remains controversial[45]. These observations hint towards a biphasic role of TGF-β1 in chondrocyte metabolism and warrat further investigation.

In this study, we also tested how pro-inflammatory (catabolic) IL-1β influences POSTN expression. As IL-1β is a pro-inflammatory cytokine, it induces degradative enzymes, stimulates anabolic activity through BMP-2[17, 46] and potentially represses type-I and type-II collagens[47]. Our data show that IL-1β inhibited COL1A1, COL2A1 and POSTN expression, and promoted MMP13 and ADAMTS4 expression. This indicates that POSTN expression may not follow the expression pattern of catabolic genes in an inflammatory milieu. It also implies that IL-1β may not induce POSTN expression in this setting and that low levels of POSTN may contribute to impaired collagen synthesis[13].

Appropriate cell migration to the wounded area is essential to initiate tissue repair[48] and periostin is known to be upregulated after injury[13]. Our scratch assay indicated that periostin accelerated wound closure, especially at the early time point, which is consistent with previous studies performed with human periodontal ligament cells[13, 49]. In contrast, periostin knockdown delayed wound closure. Therefore, these results strengthen our hypothesis that an optimal level of periostin is essential for maintaining cell migration.

Our study is in line with previous reports that human OA cartilage has elevated expression of periostin, which may play a role in matrix degradation[15, 16]. However, there are a couple of apparent contradictions regarding the role of periostin in chondrocyte metabolism. The results of OA chondrocyte treatment with human recombinant periostin is not consistent with the result of transwell co-culture system with ACL remnants (i.e. POSTN, IL1B, TGFB1 and COL1A1). One plausible reason is that other factors released from ACL remnants, as suggested by on the results of transcriptome-wide expression profiles of acute and chronic ACL remnants[10], may influence chondrocyte metabolism.

There are some limitations in our study. First, the findings of in-vitro study should be confirmed in an in-vivo model. It is plausible that this molecular communication influences the joint as a whole and drives OA development but this is not completely established by the current in-vitro investigation. Second, other tissues from ACL injured knees should be investigated in concert with the ACL remnants to confirm what changes can be attributed to the ACL tears alone in the development of OA after ACL injury. Nevertheless, our study demonstrated that ACL remnants can affect cartilage/chondrocyte metabolism through POSTN and we believe it is an important first step in understanding this process. Further study at the molecular-, cellular- and joint-level is necessary to assess the clinical impact of this relationship and whether therapeutic intervention is possible and beneficial. Third, the gene expression differences were mainly presented in the form of fold-change using 2−ddCt method as a relative quantification strategy, where the control group was usually set at 1. While this is the most commonly used approach, it does not reflect the actual Ct (threshold cycle) at which amplification of genes of interest occurred. Therefore, we have included mean Ct values for each gene of interest in each graph. Lastly, we used OA chondrocytes, which have higher expression of POSTN, to co-culture with ACL remnants, but the use of healthy cells might have been preferable. However, given the fact that OA chondrocytes were used for both acute and chronic ACL remnants, the significant increase in POSTN expression with chronic ACL co-culture is likely a result of the difference in the ACL tissue. Moreover, we have previously shown that many growth factor and cytokine responses of OA chondrocytes are similar to normal chondrocytes, even though they are in general more metabolically active[17].

In conclusion, our study demonstrates that ACL tear remnants can alter chondrocyte homeostasis. It further suggests that POSTN may play an important role in both cartilage homeostasis and OA and serve as a mediator of post-traumatic OA following ACL injury. ACL tears can affect cartilage homeostasis, suggesting that a chronic ACL injury contributes to the development of OA both biomechanically and biochemically. Although the effect of ACL reconstruction on OA risk is debated[8, 50], these findings imply that untreated ACL tears are not entirely benign. Further research is needed to better understand the clinical implications of ligament-cartilage crosstalk.

Acknowledgments

The authors would like to thank Britta Anderson, Hongjun Zheng, Crystal Idleburg and Samantha Coleman for their technical assistance.

Role of the funding source

Dr. Rai was supported by Pathway to Independence Award (R00-AR064837) and Dr. Chinzei was supported by AR045550 (PI: Sandell) from the National Institutes of Arthritis and Musculoskeletal and Skin Diseases (NIAMS), National Institutes of Health (NIH). The content of this publication is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the NIAMS.

Footnotes

Author contributions

The conception and design of the study, or acquisition of data, or analysis and interpretation of data: NC, RHB, XD, LC, RMN, LJS and MFR. Drafting the article or revising it critically for important intellectual content: NC, RHB, XD, LJS and MFR. Final approval of the version to be submitted: NC, RHB, XD, LC, RMN, LJS and MFR.

Conflict of interest

All authors declare that there is no conflict of interest concerning this work.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Nelson F, Billinghurst RC, Pidoux I, Reiner A, Langworthy M, McDermott M, et al. Early post-traumatic osteoarthritis-like changes in human articular cartilage following rupture of the anterior cruciate ligament. Osteoarthritis Cartilage. 2006;14:114–119. doi: 10.1016/j.joca.2005.08.005. [DOI] [PubMed] [Google Scholar]
  • 2.Roos H, Adalberth T, Dahlberg L, Lohmander LS. Osteoarthritis of the knee after injury to the anterior cruciate ligament or meniscus: the influence of time and age. Osteoarthritis Cartilage. 1995;3:261–267. doi: 10.1016/s1063-4584(05)80017-2. [DOI] [PubMed] [Google Scholar]
  • 3.Neuman P, Englund M, Kostogiannis I, Friden T, Roos H, Dahlberg LE. Prevalence of tibiofemoral osteoarthritis 15 years after nonoperative treatment of anterior cruciate ligament injury: a prospective cohort study. Am J Sports Med. 2008;36:1717–1725. doi: 10.1177/0363546508316770. [DOI] [PubMed] [Google Scholar]
  • 4.Church S, Keating JF. Reconstruction of the anterior cruciate ligament: timing of surgery and the incidence of meniscal tears and degenerative change. J Bone Joint Surg Br. 2005;87:1639–1642. doi: 10.1302/0301-620X.87B12.16916. [DOI] [PubMed] [Google Scholar]
  • 5.Ohly NE, Murray IR, Keating JF. Revision anterior cruciate ligament reconstruction: timing of surgery and the incidence of meniscal tears and degenerative change. J Bone Joint Surg Br. 2007;89:1051–1054. doi: 10.1302/0301-620X.89B8.19000. [DOI] [PubMed] [Google Scholar]
  • 6.Kennedy J, Jackson MP, O’Kelly P, Moran R. Timing of reconstruction of the anterior cruciate ligament in athletes and the incidence of secondary pathology within the knee. J Bone Joint Surg Br. 2010;92:362–366. doi: 10.1302/0301-620X.92B3.22424. [DOI] [PubMed] [Google Scholar]
  • 7.Nordenvall R, Bahmanyar S, Adami J, Mattila VM, Fellander-Tsai L. Cruciate ligament reconstruction and risk of knee osteoarthritis: the association between cruciate ligament injury and post-traumatic osteoarthritis. a population based nationwide study in Sweden, 1987–2009. PLoS One. 2014;9:e104681. doi: 10.1371/journal.pone.0104681. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.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;35:1756–1769. doi: 10.1177/0363546507307396. [DOI] [PubMed] [Google Scholar]
  • 9.Li H, Chen C, Chen S. Posttraumatic knee osteoarthritis following anterior cruciate ligament injury: Potential biochemical mediators of degenerative alteration and specific biochemical markers. Biomed Rep. 2015;3:147–151. doi: 10.3892/br.2014.404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Brophy RH, Tycksen ED, Sandell LJ, Rai MF. Changes in Transcriptome-Wide Gene Expression of Anterior Cruciate Ligament Tears Based on Time From Injury. Am J Sports Med. 2016;44:2064–2075. doi: 10.1177/0363546516643810. [DOI] [PubMed] [Google Scholar]
  • 11.Horiuchi K, Amizuka N, Takeshita S, Takamatsu H, Katsuura M, Ozawa H, et al. Identification and characterization of a novel protein, periostin, with restricted expression to periosteum and periodontal ligament and increased expression by transforming growth factor beta. J Bone Miner Res. 1999;14:1239–1249. doi: 10.1359/jbmr.1999.14.7.1239. [DOI] [PubMed] [Google Scholar]
  • 12.Norris RA, Borg TK, Butcher JT, Baudino TA, Banerjee I, Markwald RR. Neonatal and adult cardiovascular pathophysiological remodeling and repair: developmental role of periostin. Ann N Y Acad Sci. 2008;1123:30–40. doi: 10.1196/annals.1420.005. [DOI] [PubMed] [Google Scholar]
  • 13.Conway SJ, Izuhara K, Kudo Y, Litvin J, Markwald R, Ouyang G, et al. The role of periostin in tissue remodeling across health and disease. Cell Mol Life Sci. 2014;71:1279–1288. doi: 10.1007/s00018-013-1494-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Oka T, Xu J, Kaiser RA, Melendez J, Hambleton M, Sargent MA, et al. Genetic manipulation of periostin expression reveals a role in cardiac hypertrophy and ventricular remodeling. Circ Res. 2007;101:313–321. doi: 10.1161/CIRCRESAHA.107.149047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Chijimatsu R, Kunugiza Y, Taniyama Y, Nakamura N, Tomita T, Yoshikawa H. Expression and pathological effects of periostin in human osteoarthritis cartilage. BMC Musculoskelet Disord. 2015;16:215. doi: 10.1186/s12891-015-0682-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Attur M, Yang Q, Shimada K, Tachida Y, Nagase H, Mignatti P, et al. Elevated expression of periostin in human osteoarthritic cartilage and its potential role in matrix degradation via matrix metalloproteinase-13. FASEB J. 2015;29:4107–4121. doi: 10.1096/fj.15-272427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Sandell LJ, Xing X, Franz C, Davies S, Chang LW, Patra D. Exuberant expression of chemokine genes by adult human articular chondrocytes in response to IL-1beta. Osteoarthritis Cartilage. 2008;16:1560–1571. doi: 10.1016/j.joca.2008.04.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Wenner CE, Yan S. Biphasic role of TGF-beta1 in signal transduction and crosstalk. J Cell Physiol. 2003;196:42–50. doi: 10.1002/jcp.10243. [DOI] [PubMed] [Google Scholar]
  • 19.Matsushita T, Sasaki H, Takayama K, Ishida K, Matsumoto T, Kubo S, et al. The overexpression of SIRT1 inhibited osteoarthritic gene expression changes induced by interleukin-1beta in human chondrocytes. J Orthop Res. 2013;31:531–537. doi: 10.1002/jor.22268. [DOI] [PubMed] [Google Scholar]
  • 20.Brophy RH, Rai MF, Zhang Z, Torgomyan A, Sandell LJ. Molecular analysis of age and sex-related gene expression in meniscal tears with and without a concomitant anterior cruciate ligament tear. J Bone Joint Surg Am. 2012;94:385–393. doi: 10.2106/JBJS.K.00919. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Liang CC, Park AY, Guan JL. In vitro scratch assay: a convenient and inexpensive method for analysis of cell migration in vitro. Nat Protoc. 2007;2:329–333. doi: 10.1038/nprot.2007.30. [DOI] [PubMed] [Google Scholar]
  • 22.Takebe K, Rai MF, Schmidt EJ, Sandell LJ. The chemokine receptor CCR5 plays a role in post-traumatic cartilage loss in mice, but does not affect synovium and bone. Osteoarthritis Cartilage. 2015;23:454–461. doi: 10.1016/j.joca.2014.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Zhu Y, Oganesian A, Keene DR, Sandell LJ. Type IIA procollagen containing the cysteine-rich amino propeptide is deposited in the extracellular matrix of prechondrogenic tissue and binds to TGF-beta1 and BMP-2. J Cell Biol. 1999;144:1069–1080. doi: 10.1083/jcb.144.5.1069. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Kiapour AM, Murray MM. Basic science of anterior cruciate ligament injury and repair. Bone Joint Res. 2014;3:20–31. doi: 10.1302/2046-3758.32.2000241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Anderson-MacKenzie JM, Billingham ME, Bailey AJ. Collagen remodeling in the anterior cruciate ligament associated with developing spontaneous murine osteoarthritis. Biochem Biophys Res Commun. 1999;258:763–767. doi: 10.1006/bbrc.1999.0713. [DOI] [PubMed] [Google Scholar]
  • 26.Haslauer CM, Elsaid KA, Fleming BC, Proffen BL, Johnson VM, Murray MM. Loss of extracellular matrix from articular cartilage is mediated by the synovium and ligament after anterior cruciate ligament injury. Osteoarthritis Cartilage. 2013;21:1950–1957. doi: 10.1016/j.joca.2013.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Loeser RF, Olex AL, McNulty MA, Carlson CS, Callahan MF, Ferguson CM, et al. Microarray analysis reveals age-related differences in gene expression during the development of osteoarthritis in mice. Arthritis Rheum. 2012;64:705–717. doi: 10.1002/art.33388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Karlsson C, Dehne T, Lindahl A, Brittberg M, Pruss A, Sittinger M, et al. Genome-wide expression profiling reveals new candidate genes associated with osteoarthritis. Osteoarthritis Cartilage. 2010;18:581–592. doi: 10.1016/j.joca.2009.12.002. [DOI] [PubMed] [Google Scholar]
  • 29.Chou CH, Wu CC, Song IW, Chuang HP, Lu LS, Chang JH, et al. Genome-wide expression profiles of subchondral bone in osteoarthritis. Arthritis Res Ther. 2013;15:R190. doi: 10.1186/ar4380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Honsawek S, Wilairatana V, Udomsinprasert W, Sinlapavilawan P, Jirathanathornnukul N. Association of plasma and synovial fluid periostin with radiographic knee osteoarthritis: Cross-sectional study. Joint Bone Spine. 2015;82:352–355. doi: 10.1016/j.jbspin.2015.01.023. [DOI] [PubMed] [Google Scholar]
  • 31.Rousseau JC, Sornay-Rendu E, Bertholon C, Garnero P, Chapurlat R. Serum periostin is associated with prevalent knee osteoarthritis and disease incidence/progression in women: the OFELY study. Osteoarthritis Cartilage. 2015;23:1736–1742. doi: 10.1016/j.joca.2015.05.015. [DOI] [PubMed] [Google Scholar]
  • 32.Tsai TT, Lai PL, Liao JC, Fu TS, Niu CC, Chen LH, et al. Increased periostin gene expression in degenerative intervertebral disc cells. Spine J. 2013;13:289–298. doi: 10.1016/j.spinee.2013.01.040. [DOI] [PubMed] [Google Scholar]
  • 33.Litvin J, Blagg A, Mu A, Matiwala S, Montgomery M, Berretta R, et al. Periostin and periostin-like factor in the human heart: possible therapeutic targets. Cardiovasc Pathol. 2006;15:24–32. doi: 10.1016/j.carpath.2005.09.001. [DOI] [PubMed] [Google Scholar]
  • 34.Romanos GE, Asnani KP, Hingorani D, Deshmukh VL. PERIOSTIN: role in formation and maintenance of dental tissues. J Cell Physiol. 2014;229:1–5. doi: 10.1002/jcp.24407. [DOI] [PubMed] [Google Scholar]
  • 35.Varady NH, Grodzinsky AJ. Osteoarthritis year in review 2015: mechanics. Osteoarthritis Cartilage. 2016;24:27–35. doi: 10.1016/j.joca.2015.08.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Brophy RH, Zhang B, Cai L, Wright RW, Sandell LJ, Rai MF. Transcriptome comparison of meniscus from patients with and without osteoarthritis. Osteoarthritis Cartilage. 2017 doi: 10.1016/j.joca.2017.12.004. In Press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Norris RA, Damon B, Mironov V, Kasyanov V, Ramamurthi A, Moreno-Rodriguez R, et al. Periostin regulates collagen fibrillogenesis and the biomechanical properties of connective tissues. J Cell Biochem. 2007;101:695–711. doi: 10.1002/jcb.21224. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Fang J, Xu L, Li Y, Zhao Z. Roles of TGF-beta 1 signaling in the development of osteoarthritis. Histol Histopathol. 2016;31:1161–1167. doi: 10.14670/HH-11-779. [DOI] [PubMed] [Google Scholar]
  • 39.Sandell LJ. Modern molecular analysis of a traditional disease: progression in osteoarthritis. Arthritis Rheum. 2007;56:2474–2477. doi: 10.1002/art.22760. [DOI] [PubMed] [Google Scholar]
  • 40.Cucchiarini M, de Girolamo L, Filardo G, Oliveira JM, Orth P, Pape D, et al. Basic science of osteoarthritis. J Exp Orthop. 2016;3:22. doi: 10.1186/s40634-016-0060-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Zhou Q, Wei B, Liu S, Mao F, Zhang X, Hu J, et al. Cartilage matrix changes in contralateral mobile knees in a rabbit model of osteoarthritis induced by immobilization. BMC Musculoskelet Disord. 2015;16:224. doi: 10.1186/s12891-015-0679-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Aukkarasongsup P, Haruyama N, Matsumoto T, Shiga M, Moriyama K. Periostin inhibits hypoxia-induced apoptosis in human periodontal ligament cells via TGF-beta signaling. Biochem Biophys Res Commun. 2013;441:126–132. doi: 10.1016/j.bbrc.2013.10.027. [DOI] [PubMed] [Google Scholar]
  • 43.Chen KS, Tatarczuch L, Mirams M, Ahmed YA, Pagel CN, Mackie EJ. Periostin expression distinguishes between light and dark hypertrophic chondrocytes. Int J Biochem Cell Biol. 2010;42:880–889. doi: 10.1016/j.biocel.2010.01.018. [DOI] [PubMed] [Google Scholar]
  • 44.Zhu Y, Tao H, Jin C, Liu Y, Lu X, Hu X, et al. Transforming growth factor-beta1 induces type II collagen and aggrecan expression via activation of extracellular signal-regulated kinase 1/2 and Smad2/3 signaling pathways. Mol Med Rep. 2015;12:5573–5579. doi: 10.3892/mmr.2015.4068. [DOI] [PubMed] [Google Scholar]
  • 45.Bush JR, Beier F. TGF-beta and osteoarthritis--the good and the bad. Nat Med. 2013;19:667–669. doi: 10.1038/nm.3228. [DOI] [PubMed] [Google Scholar]
  • 46.Fukui N, Zhu Y, Maloney WJ, Clohisy J, Sandell LJ. Stimulation of BMP-2 expression by pro-inflammatory cytokines IL-1 and TNF-alpha in normal and osteoarthritic chondrocytes. J Bone Joint Surg Am. 2003;85-A(Suppl 3):59–66. doi: 10.2106/00004623-200300003-00011. [DOI] [PubMed] [Google Scholar]
  • 47.Ley C, Svala E, Nilton A, Lindahl A, Eloranta ML, Ekman S, et al. Effects of high mobility group box protein-1, interleukin-1beta, and interleukin-6 on cartilage matrix metabolism in three-dimensional equine chondrocyte cultures. Connect Tissue Res. 2011;52:290–300. doi: 10.3109/03008207.2010.523803. [DOI] [PubMed] [Google Scholar]
  • 48.Wu Z, Dai W, Wang P, Zhang X, Tang Y, Liu L, et al. Periostin promotes migration, proliferation, and differentiation of human periodontal ligament mesenchymal stem cells. Connect Tissue Res. 2017:1–12. doi: 10.1080/03008207.2017.1306060. [DOI] [PubMed] [Google Scholar]
  • 49.Padial-Molina M, Volk SL, Rios HF. Periostin increases migration and proliferation of human periodontal ligament fibroblasts challenged by tumor necrosis factor-alpha and Porphyromonas gingivalis lipopolysaccharides. J Periodontal Res. 2014;49:405–414. doi: 10.1111/jre.12120. [DOI] [PubMed] [Google Scholar]
  • 50.Naendrup JH, Zlotnicki JP, Chao T, Nagai K, Musahl V. Kinematic outcomes following ACL reconstruction. Curr Rev Musculoskelet Med. 2016;9:348–360. doi: 10.1007/s12178-016-9359-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES