Article Text

Extended report
Cartilage-specific deletion of mTOR upregulates autophagy and protects mice from osteoarthritis
  1. Yue Zhang1,
  2. Faezeh Vasheghani1,
  3. Ying-hua Li1,
  4. Meryem Blati1,
  5. Kayla Simeone1,
  6. Hassan Fahmi1,
  7. Bertrand Lussier2,
  8. Peter Roughley3,
  9. David Lagares4,
  10. Jean-Pierre Pelletier1,
  11. Johanne Martel-Pelletier1,
  12. Mohit Kapoor1,5
  1. 1Osteoarthritis Research Unit, University of Montreal Research Centre (CRCHUM), Montreal, Quebec, Canada
  2. 2Department of Clinical Sciences, Faculty of Veterinary Medicine, University of Montreal, Montreal, Quebec, Canada
  3. 3Genetics Unit, Shriners Hospital, McGill University, Montreal, Quebec, Canada
  4. 4Pulmonary and Critical Care Unit and Center for Immunology and Inflammatory Diseases, Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Massachusetts, USA
  5. 5Division of Genetics and Development, The Toronto Western Research Institute, Toronto Western Hospital, The University Health Network (UHN), Toronto, Ontario, Canada
  1. Correspondence to Dr Mohit Kapoor, Division of Genetics and Development, The Toronto Western Research Institute, Toronto Western Hospital, The University Health Network (UHN), 60 Leonard Avenue, Toronto, Ontario, Canada M5T 2S8; mkapoor{at}uhnresearch.ca

Abstract

Objectives Mammalian target of rapamycin (mTOR) (a serine/threonine protein kinase) is a major repressor of autophagy, a cell survival mechanism. The specific in vivo mechanism of mTOR signalling in OA pathophysiology is not fully characterised. We determined the expression of mTOR and known autophagy genes in human OA cartilage as well as mouse and dog models of experimental OA. We created cartilage-specific mTOR knockout (KO) mice to determine the specific role of mTOR in OA pathophysiology and autophagy signalling in vivo.

Methods Inducible cartilage-specific mTOR KO mice were generated and subjected to mouse model of OA. Human OA chondrocytes were treated with rapamycin and transfected with Unc-51–like kinase 1 (ULK1) siRNA to determine mTOR signalling.

Results mTOR is overexpressed in human OA cartilage as well as mouse and dog experimental OA. Upregulation of mTOR expression co-relates with increased chondrocyte apoptosis and reduced expression of key autophagy genes during OA. Subsequently, we show for the first time that cartilage-specific ablation of mTOR results in increased autophagy signalling and a significant protection from destabilisation of medial meniscus (DMM)-induced OA associated with a significant reduction in the articular cartilage degradation, apoptosis and synovial fibrosis. Furthermore, we show that regulation of ULK1/adenosine monophosphate-activated protein kinase (AMPK) signalling pathway by mTOR may in part be responsible for regulating autophagy signalling and the balance between catabolic and anabolic factors in the articular cartilage.

Conclusions This study provides a direct evidence of the role of mTOR and its downstream modulation of autophagy in articular cartilage homeostasis.

  • Osteoarthritis
  • Chondrocytes
  • Arthritis
View Full Text

Statistics from Altmetric.com

Introduction

Osteoarthritis (OA) is the most common form of arthritis whose exact pathophysiology is still unknown. The major features of OA are cartilage degradation, synovial inflammation and subchondral bone remodelling. Loss of chondrocyte cellularity within the articular cartilage is one of the prominent events that contribute to its degradation. However, it is still uncertain as to what mechanisms control the fate of chondrocytes within the articular cartilage during normal versus OA conditions.

Mammalian target of rapamycin (mTOR), a serine/threonine protein kinase, is a key regulator of cell growth, metabolism, survival and lifespan of organisms.1 mTOR associates with raptor and rictor to form the mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2). mTOR is known to regulate protein synthesis by phosphorylation and inactivation of translational repressor 4E-binding protein (4E-BP1), and through the phosphorylation and activation of S6 kinase (S6K1).2 It has been previously shown that TOR deficiency in the nematode Caenorhabditis elegans is able to double its lifespan.3 In mice, treatment with mTORC1 inhibitor (rapamycin) extends lifespan of both male and female mice,4 suggesting mTOR as a key mediator of lifespan regulation. One of the key functions of mTOR is the suppression of autophagy, a cell survival mechanism.5 Autophagy is an essential homeostatic process by which cells break down their own components and is essential for survival, differentiation, development and homeostasis.5 Hypoxia, reactive oxygen species and deprivation of nutrients and energy are among the key inducers of autophagy process6 and when mTOR is inhibited.7 It has been shown that mTOR prevents ULK1 activation by phosphorylating ULK1 and disrupting the interaction between ULK1 and AMPK.8

Recent studies by Caramés et al9 ,10 and Sasaki et al11 have shown dysregulation in the expression of key autophagy genes in OA pathogenesis. It has been shown that the expression of ULK1 (the most upstream autophagy inducer), microtubule-associated protein 1 light chain 3 (LC3B; autophagy structural and functional factor) and Beclin1 (autophagy regulator) is reduced during OA.9 Treatment with rapamycin (mTORC1 Inhibitor) has also been shown to reduce the severity of experimental OA in mice.10

The specific in vivo mechanism of mTOR signalling in OA pathophysiology is not fully characterised. This study was designed to first determine the expression of mTOR and autophagy genes in human OA cartilage as well as in mouse and dog models of experimental OA. To elucidate the specific in vivo role of mTOR in OA pathophysiology and regulation of autophagy during OA, we created cartilage-specific mTOR knockout (KO) mice using LoxP Cre technology and subjected these mice to DMM model of OA surgery to determine the effects of cartilage-specific ablation of mTOR on autophagy signalling and kinetics of OA in vivo. This study provides direct evidence on the role of mTOR and its downstream modulation of autophagy in articular cartilage homeostasis.

Methods

Human specimens

Normal human cartilage was obtained from both femoral condyles and tibial plateaus, at autopsy within 24 h of death, from individuals with no history of arthritic diseases (n=15, 60.4±15.4 years (mean±SD)). To ensure that only normal tissue was used, cartilage specimens were thoroughly examined both macroscopically and microscopically as previously described.12 Human OA cartilage was obtained from patients undergoing knee replacement (n=32, 65.2±17.6 years). In all patients, OA was diagnosed on the basis of criteria developed by the American College of Rheumatology Diagnostic Subcommittee for OA.13 At the time of surgery, the patients had symptomatic disease requiring medical treatment in the form of non-steroidal anti-inflammatory drugs or selective COX-2 inhibitors. None had received intra-articular steroid injections within 3 months prior to surgery. The Institutional Ethics Committee Board approved the use of the human articular tissues.

Generation of inducible cartilage-specific mTOR knockout mice

Inducible cartilage-specific mTOR KO mice were generated by mating mice containing a mTOR gene flanked by LoxP sites (C57BL/6- mTORfl/fl, Jackson Laboratory) with C57BL/6 Col2-rt-TA-Cre transgenic mice14 (obtained from Dr. Peter Roughley, McGill University, Montreal). Five weeks old mTORfl/fl Cre mice were fed doxycycline (Sigma-Aldrich Inc., Oakville, ON) dissolved at 10 µg/µL in phosphate buffer saline (PBS), pH 7.4 by oral gavage with the dose of 80 µg/g body weight for 7 days. mTORfl/fl Cre mice without doxycycline (only PBS) treatment were used as control mice. Routine genotyping of tail DNA followed by confirmation of cartilage-specific loss of mTOR expression upon treatment with doxycycline was confirmed by western blotting and immunohistochemistry as described before.14 All animal procedure protocols were approved by the Comité Institutionnel de protection des animaux of CRCHUM.

Chondrocyte culture studies

Human OA chondrocytes were released from cartilage by sequential enzymatic digestion, as previously described.15 Cells were seeded at 3×105/well in 6-well culture plates in DMEM supplemented with 10% FBS and cultivated at 37°C for 48 h. Cells were washed and incubated for an additional 24 h in DMEM containing 0.5% FBS before treatment with rapamycin (100 nM) for 24 h. Cells were then harvested for RNA or protein isolation.

RNA extraction and quantitative real-time PCR (qPCR)

Total RNA was isolated by Trizol, then the RNeasy Mini kit (QIAGEN) was used including on-column DNase digestion to eliminate DNA as mentioned above (RNase-Free DNase Set, QIAGEN). RNA quantification was then performed using the QuantiTect Reverse Transcription PCR Kit (QIAGEN) on the Rotor Gene 3000 real-time PCR system (Corbett Research, Mortlake, Australia) according to the manufacturer's protocol and the fold increase in PCR products by 2−ΔΔCt method and calculated using the housekeeping gene GAPDH.16 All experiments were performed in triplicate for each sample, and the primers were designed using Primer3 online software.

Immunohistochemistry and TUNEL staining

The immunohistochemistry analysis was performed as previously described.16 The percentage of chondrocytes staining positive for each specific antigen was determined as previously reported.17 To detect chondrocyte apoptosis, TUNEL staining and Poly (ADP-ribose) polymerase (PARP)p85 immunostaining was performed in cartilage specimens obtained from normal and OA human cartilage as well as control and mTOR KO mice. TUNEL assay was performed using ApopTag Plus Peroxidase In Situ Apoptosis Detection Kit according to the manufacturer's recommendations (Millipore, Ontario, Canada).

Western blotting

Cells were lysed in Tris buffered saline (TBS) containing 0.1% sodium dodecyl sulfate (SDS), and the protein content of the lysates was determined using a bicinchoninic acid protein assay reagent (Thermo Fisher Scientific), with bovine serum albumin (BSA) as the standard. Cell lysates were adjusted to equal equivalents of protein and then were applied to SDS–polyacrylamide gels for electrophoresis as described before.16

Murine model of OA

OA was surgically induced in 10-week-old control and mTOR KO male mice by DMM or sham surgery (control) in the right knee, as previously described.18 ,19 Histological and biochemical analysis were performed at 5 and 10 weeks post-OA surgery.

At 5 weeks post-OA surgery, right knee joint cartilage from femoral condyles and tibial plateaus was removed and primary chondrocytes were prepared from the mice as previously described.16

Assessment of progression and severity of osteoarthritis

Mouse knee joints sections (5 µm) at 5 and 10 weeks post-OA surgery were stained with Safranin-O/Fast Green (Sigma-Aldrich, Oakville, Ontario) according to the manufacturer's recommendations. Slides were evaluated by two independent readers in a blinded fashion. To determine the extent of cartilage deterioration, joint sections were stained with Safranin-O/Fast Green and histological scoring method issued by Osteoarthritis Research Society International (OARSI) was used for analysis as previously described.20 To determine degree of synovial fibrosis, 5-week post-OA knee joint sections were stained with Masson's Trichrome stain. Stained sections were blindly scored for degree of fibrosis on a scale of 0–3 (0, no fibrosis; 1, mild fibrosis; 2, moderate fibrosis; and 3, severe fibrosis).

Dog model of OA

Adult mixed breed Mongrel dogs (LAKA, St-Basile-le-Grand, Quebec, Canada) weighing 25±3 kg underwent surgical sectioning of the anterior cruciate ligament of the right knee as previously described,21 and at 8 weeks postsurgery, cartilage from femoral condyles and tibial plateaus was removed and processed for RNA extraction as described above. Cartilage from femoral condyles and tibial plateaus of the left (contralateral) knee was used as control cartilage. All experimental procedures were approved by the Institutional Ethics Committee Board.

Gene silencing

Briefly, human OA chondrocytes were plated at 3×105/well in 6-well culture plates in DMEM supplemented with 10% FBS and pretreated with rapamycin (or DMSO as control) for 6 h. Cells were then transfected with ULK1 siRNA (20 nM, QIAGEN) or non-targeting (random) CTRL-ALLstars siRNAs (QIAGEN) using Lipofectamine RNAiMax Reagent (Invitrogen) for 48 h and protein and RNA were harvested for analysis.

Statistical analysis

The data are expressed as mean±SEM. The significance of differences in the levels of expression between the control (normal) and OA groups was determined using a two-tailed Mann–Whitney U test. For significance of differences between rapamycin-treated OA chondrocytes and DMSO-treated (control) chondrocytes, as well as siCTRL and siULK-treated OA chondrocytes, two-tailed t test was used. For autophagy PCR array analysis, two-tailed T test was used. p<0.05 was considered statistically significant.

Results

Increased expression of mTOR in human OA as well as mouse and dog models of experimental OA

We first determined the expression of mTOR in human OA cartilage compared with normal human cartilage. Western blotting and immunohistochemical analysis showed enhanced protein expression of mTOR in human OA cartilage compared with normal human cartilage (figure 1A, B). Further, we also observed a significant upregulation (p<0.05) in the mRNA expression of mTOR in human OA cartilage compared with normal human cartilage (figure 1C). Similarly, mTOR expression was also significantly upregulated (p<0.05) in both dog OA cartilage and mouse OA chondrocytes compared with control dog cartilage and control sham surgery chondrocytes, respectively (figure 1D, E).

Figure 1

Enhanced expression of mTOR during OA. (A and B) Protein expression of mTOR was enhanced in human OA cartilage (n=4) compared to normal human cartilage (n=4) as determined by Western blotting and immunohistochemistry. Original magnification = ×25. (C) A significant increase (*p<0.05) in the mRNA expression of mTOR was observed in human OA cartilage (n=5) compared to normal human cartilage (n=5). (D) In dog OA cartilage (n=5), a significant increase (*p<0.05) in the mRNA expression of mTOR was observed compared to control contralateral cartilage (n=5). (E) A significant increase (*p<0.05) in the mRNA expression of mTOR was also observed in mouse OA chondrocytes (n=8) compared to sham-surgery control chondrocytes (n=8).

Aberrant expression of autophagy genes in human OA cartilage compared with normal human cartilage

Since mTOR is a master negative regulator of autophagy and we observed increased expression of mTOR in human OA cartilage compared with normal human cartilage, we further determined the expression of known autophagy genes in human OA cartilage compared with normal human cartilage. The expression patterns of 84 key autophagy genes and 5 housekeeping genes in human OA cartilage compared with normal human cartilage were profiled using a human autophagy PCR array.

The results show that in OA cartilage 20 autophagy-related genes were significantly downregulated and 5 autophagy-related genes were significantly upregulated compared with normal human cartilage (p<0.05) (see online supplementary table 1). Also, no significant change in the expression of 53 autophagy-related genes was observed in OA cartilage compared with normal human cartilage. 6 autophagy-related genes were undetected in both normal and OA cartilage samples. Results obtained from the PCR arrays showed that the key autophagy-related genes such as ATG3, ATG5, ATG12, ULK1, LC3B, Beclin-1 and GABA(A)receptor-associated protein like 1 (GABARAPL1) involved in initiating autophagy, autophagic vacuole formation and phagophore extention were downregulated in OA cartilage compared with normal human cartilage (see online supplementary table 1). Also, coregulators of autophagy and apoptosis including BNIP3, cyclin-dependent kinase inhibitor 1b (CDKN1B), TNF receptor superfamily member 6 (FAS) were downregulated in OA cartilage. Chaperone-mediated autophagy-related genes, such as HSP90AA1 and HSPA8, were also significantly downregulated in OA cartilage. Array data also revealed that critical regulators of cell death/apoptosis mechanisms including APP, CTSB, BCL2 and BCL2-associated agonist of cell death (BAD) were upregulated in OA cartilage compared with normal cartilage.

Increase in cell death and reduction in the production/expression of key autophagy markers in human OA cartilage versus normal cartilage

We performed TUNEL assay to determine the degree of cell death in human OA cartilage compared with normal human cartilage. Results clearly show greater percentage of TUNEL positive cells in OA cartilage compared with normal human cartilage (figure 2A). Immunostaining using PARPp85 antibody further confirmed increased number of apoptotic cells in human OA cartilage compared with normal human cartilage.

Figure 2

Increased chondrocyte apoptosis and decreased expression of autophagy genes during OA. (A) TUNEL assay and immunohistochemical analysis for PPAP p85 demonstrated a significant increase in the percentage (%) of apoptotic cells in human OA cartilage compared to normal human cartilage (n=6, *p<0.05). (B) Immunohistochemical analysis showed decreased % of positive cells for LC3B, ATG5, BNIP3 and ULK1 in human OA cartilage (n=5) compared to normal human cartilage (n=5, *p<0.05). Original magnification = ×25. (C) Significant down-regulation in the mRNA expression of autophagy genes including LC3B, ATG5, BNIP3, and ULK1 in human OA cartilage (n=8) compared to normal human cartilage (n=8; *p<0.05). (D) A significant reduction (n=5; *p<0.05) in the mRNA expression of LC3B and ATG5 (with no significant change in the expression of BNIP3 and ULK1) was observed in the dog OA cartilage (n=5) compared to control contralateral cartilage (n=5). (E) A significant (n=8; *p<0.05) reduction in the mRNA expression levels of LC3B, ATG5 and ULK1 (with no significant change in the expression of BNIP3) was observed in mouse OA chondrocytes (n=8) compared to sham-surgery chondrocytes (n=8).

Since autophagy PCR array data showed that several autophagy genes involved in the induction of autophagy process were downregulated in human OA cartilage compared with normal human cartilage, we further analysed the expression of four crucial autophagy markers including ULK1 (most upstream autophagy inducer), LC3B (an autophagy structural and functional factor), ATG5 (an autophagy regulator) and BNIP3 (an interactor of LC3 in autophagy) by immunohistochemistry and qPCR analysis. Immunohistochemistry showed that all four autophagy genes (ULK1, LC3B, ATG5 and BNIP3) were constitutively expressed throughout the normal human cartilage. However, a significant (p<0.01) reduction in the number of positive cells for ULK1, LC3B, ATG5 and BNIP3 was observed in the human OA cartilage compared with normal human cartilage (figure 2B). qPCR analysis also showed a significant reduction in the mRNA expression of ULK1, LC3B, ATG5 and BNIP3 (p<0.05) in OA cartilage compared with normal human cartilage (figure 2C).

Reduction in the expression of autophagy markers in dog and mouse experimental OA

We further assessed the expression of LC3B, ATG5, BNIP3 and ULK1 in dog and mouse models of experimental OA. In the cartilage obtained from the knees of dogs subjected to the instability model of experimental OA (n=5) compared with cartilage obtained from the non-surgery contralateral knee joints (control, n=5), we observed a significant downregulation (p<0.05) in the mRNA expression levels of LC3B and ATG5 with no significant differences in the expression of BNIP3 and ULK1 in dog OA cartilage compared with control cartilage (figure 2D). In mouse model of experimental OA, we isolated chondrocytes from the knee joints of mouse subjected to DMM OA surgery (n=8) and control sham surgery (n=8). As observed in human OA cartilage, the mRNA expression of LC3B, ATG5 and ULK1 (but not BNIP3) was significantly (p<0.05) reduced in mouse OA chondrocytes compared with sham surgery chondrocytes (figure 2E). These results indicate the complexity and differential regulation of certain autophagy genes across species.

Cartilage-specific deletion of mTOR results in increased expression of autophagy markers

Since we observed dysregulation in the expression of mTOR and autophagy genes in OA conditions, we then created inducible cartilage-specific mTOR KO mice to elucidate the in vivo role of mTOR signalling in OA pathophysiology. Routine genotyping of tail DNA followed by confirmation of cartilage-specific loss of mTOR expression upon treatment with doxycycline was confirmed by western blotting and immunohistochemistry (figures 3A, B and 4A). Inducible mTOR KO mouse develop normally exhibit no developmental defects with no phenotypic changes in the articular cartilage till 10 weeks postbirth. First, we examined the effect of cartilage-specific ablation of mTOR on mTOR signalling pathway. Loss of mTOR in the cartilage resulted in decreased phosphorylation of ribosomal protein S6 Kinase (rpS6 K), a downstream target of mTORC12 (figure 3B). Furthermore, mTOR genetic deletion resulted in increased expression of ULK1, AMPK1, ATG5, BNIP3 and increase in total LC3 expression as well as the conversion from LC3BI to LC3BII. These results show an increased autophagy signalling upon genetic deletion of mTOR in the articular cartilage.

Figure 3

Inducible cartilage-specific mTOR KO mice exhibit protection from DMM-induced OA. (A) Characterization of inducible cartilage-specific mTOR KO mouse: Genotyping confirms the presence of the Cre transgene in mTOR heterozygote (mTORfl/w) and homozygote (mTORfl/fl) mice. The mTOR KO band is detected at 533 bp, the wild-type (WT) band at 349 bp, and the Cre band at 700 bp. (B) mTOR deficient cartilage exhibits increased expression of autophagy genes: Western blotting results showed abolishment of mTOR protein expression, decreased phosphorylation of P70S6 kinase and increased expression of ULK1, AMPK1, ATG5, BNIP3 and total LC3 expression as well as the conversion from LC3BI to LC3BII, in articular chondrocytes isolated from inducible cartilage-specific mTOR KO mice compared to control mice chondrocytes. (C) mTOR KO mice are protected from DMM-induced OA: Inducible cartilage-specific mTOR KO mice and control mice were subjected to DMM surgery and kinetics of cartilage degradation were assessed at 5 and 10 weeks post OA surgery. Safranin O-fast green stained sections showed that all mTOR KO mice exhibited significant protection from OA compared to control mice at 5 and 10 weeks post OA associated with reduced cartilage degradation/fibrillation, proteoglycan loss and reduced loss of articular chondrocyte cellularity. Original magnification = ×25. (D) A significant reduction in the OARSI scale was observed at both medial tibial plateau and medial femoral chondyle in mTOR KO mice compared to control mice (n=5/genotype/time point). (E) mTOR KO mice are protected from synovial fibrosis during OA: Trichrome staining showed decreased synovial fibrosis in mTOR KO mice (n=5) compared with control mice (n=5) 5 weeks post OA surgery. Original magnification = ×25. (F) A significant reduction in the fibrosis score in mTOR KO mice synovium compared to control mice synovium (n=5, **p<0.01). (G and H) Immunohistochemical analysis showing a significant decrease in number of p-Smad3 positive cells in the synovium of mTOR KO mice (n=4) compared to control mice (n=4). Original magnification = ×25.

Figure 4

mTOR KO mice exhibit decreased articular chondrocyte apoptosis, increased expression of autophagy markers and decreased expression of MMP-13 during OA. TUNEL assay and immunohistochemical analysis for PPAP p85 showed a decreased number of apoptotic cells (chondrocytes) in the tibial plateau of mTOR KO mice articular cartilage compared to control mice articular cartilage 10 weeks post OA surgery. Immunohistochemical analysis showed significantly increased expression of autophagy genes including ULK1 (n=5, **p<0.01), LC3B (n=5, **p<0.01) and ATG5 (n=5, *p<0.05), and decreased expression of OA catabolic factor MMP-13 (n=5, *p<0.05) and decreased staining for the type II collagen breakdown product C1-2C (n=5) in the tibial plateau of mTOR KO mice (n=5) compared to control mice (n=5) 10 weeks post OA surgery. Original magnification = ×25. (B) A significant increase in the mRNA expression levels of ULK1, LC3B and ATG5 (n×5; *p<0.05) and significant decrease in the expression of MMP-13 (n=5; *p<0.05) was observed in mTOR KO OA chondrocytes compared to control mice OA chondrocytes.

Inducible cartilage-specific mTOR KO mice exhibit protection from DMM-induced OA

mTOR KO mouse were then subjected to DMM model of OA and kinetics of OA progression and severity were assessed at 5 and 10 weeks post-OA surgery. Compared with control mice, mTOR KO mice exhibited significant protection from DMM-induced OA manifestations including significant protection from cartilage degradation, significant reduction in proteoglycan loss and loss of articular chondrocyte cellularity (figure 3C). Two blinded observers further confirmed a significant reduction in the OARSI scale at both medial tibial plateau and medial femoral chondyle in mTOR KO mice compared with control mice (figure 3D).

Interestingly, protection from DMM-induced OA in mTOR KO mouse was not limited to the articular cartilage as significant reduction in the synovial fibrosis was also observed in mTOR KO mouse compared with control mouse (figure 3E–F). Since transforming growth factor-beta (TGF-β) is a major pro-fibrotic factor and its signalling through Smads is implicated in mediating fibrosis, we investigated the mechanism related to decreased synovial fibrosis observed in mTOR KO mice by performing immunostaining for pSmad3 in the synovium of mTOR KO mice compared with control mice. In 5 weeks post-OA mice synovium, a significant decrease in the number of pSmad3 positive cells was observed in the synovium of mTOR KO mice compared with control mice (figure 3G, H), suggesting that loss of mTOR is associated with decreased TGF-β/Smad3 signalling resulting in decreased synovial fibrosis.

Inducible cartilage-specific mTOR KO mice exhibit reduced apoptosis, enhanced production of autophagy markers and reduced expression of catabolic factor MMP13 during OA

TUNEL assay and PARPp85 on 10 weeks post-OA mouse cartilage showed a reduction in the apoptotic cells in mTOR KO mouse cartilage compared with control mouse cartilage (figure 4). The protection of articular chondrocytes from apoptosis in mTOR KO cartilage co-related with the increased expression of autophagy genes including ULK1, LC3B and ATG5 and decreased expression of major OA-catabolic factor matrix metalloproteinase-13 (MMP-13) and MMP-induced type II collagen breakdown product C1,2C (figure 4A) as assessed by immunohistochemistry. qPCR analysis further confirmed the increased expression of ULK1, LC3B and ATG5 and decreased expression of MMP-13 in chondrocytes isolated from mTOR KO mice at 5 weeks post-OA surgery compared with control mice chondrocytes (figure 4B). These results show that upon genetic deletion of mTOR, increase in autophagy signalling and the reduction in chondrocyte apoptosis may in part contribute to decreased catabolic activity observed in mTOR KO mouse, resulting in protection from OA.

mTOR modulates the expression of autophagy factors and the expression of catabolic and anabolic factors implicated in OA

We treated human OA chondrocytes with rapamycin (mTORC1 inhibitor)22 and determined the effect of inhibition of mTOR signalling on the expression of autophagy-related genes as well as on the expression of key catabolic and anabolic factors implicated in OA pathophysiology. Rapamycin treatment significantly reduced rpS6K phosphorylation in OA chondrocytes compared with control (DMSO-treated) chondrocytes confirming the inhibition of mTOR signalling pathway (figure 5A). Further, rapamycin treatment resulted in a significant increase in total LC3 expression and the conversion from LC3BI to LC3BII (figure 5A) as well as increase in the mRNA expression of LC3, ULK1 and AMPK1 in rapamycin-treated OA chondrocytes compared with DMSO-treated chondrocytes (figure 5C). Interestingly, treatment of OA chondrocytes with rapamycin significantly increased the mRNA expression of aggrecan and type II collagen (two major components of extracellular matrix) and decreased the expression of MMP-13 and chemokines including CCL5/RANTES and CCL2/MCP-1 compared with DMSO-treated chondrocytes (figure 5D). Similarly, we isolated mouse OA chondrocytes from control and mTOR KO mouse at 5 weeks post-OA surgery and treated these chondrocytes with rapamycin in vitro. As expected, in control mice OA chondrocytes, rapamycin treatment resulted in a significant rescue/increase in the expression of ULK1, LC3B, ATG5 and type II collagen and a significant decrease in the expression of MMP-13, thus mimicking the effects observed in human OA chondrocytes (see online supplementary figure 1). In mTOR-deficient chondrocytes, rapamycin treatment exhibited no significant differences in the expression of autophagy markers as well as the expression of catabolic and anabolic factors, thus confirming that in the absence of mTOR (mTOR KO mice chondrocytes) rapamycin exhibited no further rescue effect.

Figure 5

mTOR controls autophagy by modulating ULK1 expression in the articular cartilage. (A) Human OA chondrocytes weretreated with rapamycin for 24 hr. Results showed decreased phosphorylation of P70S6 Kinase in rapamycin-treated OA chondrocytes (n=4) compared to DMSO-treated (control) chondrocytes (n=4). Treatment of OA chondrocytes with rapamycin resulted in an increase in total LC3 expression. (B) The conversion from LC3BI to LC3BII compared to DMSO-treated (control) OA chondrocytes. (C) mRNA expression of LC3, ULK1 and AMPK1 was significantly increased (n=6; *p<0.05) in rapamycin-treated human OA chondrocytes (n=6) compared to control (DMSO-treated) chondrocytes (n=6). (D) A significant increase in the mRNA expression of aggrecan and Type II collagen (p<0.05), and a significant decrease in the mRNA expression of MMP-13 (n=6; *p<0.05), CCL5/RANTES (n=6; *p<0.05) and CCL2/MCP-1 (n=6; **p<0.01) was observed in rapamycin-treated OA chondrocytes (n=6) compared to DMSO-treated OA chondrocytes (n=6). (E-F) Silencing efficiency of ULK1 siRNAs in human OA chondrocytes (n=6) was confirmed by qPCR and western blotting analysis. A significant decrease in the expression of ULK1 was observed at both mRNA (E) and protein levels (F) in ULK1 siRNA-treated cells compared to control siRNA-treated cells.(G-H) Silencing of ULK1 in rapamycin-treated human OA chondrocytes resulted in a significant decrease in the protein expression of LC3B determined by Western blots (total LC3 expression and the conversion from LC3BI to LC3BII), significant decrease in the mRNA expression of LC3B and ATG5 and increase in the expression of OA catabolic factors including MMP-13 and CCL2 at both mRNA (H) and protein levels (I) (n=8; *p<0.05).

Since IL-1β is a major pro-inflammatory cytokine implicated in OA pathophysiology,23 we treated human OA chondrocytes in the presence/absence of IL-β to determine whether IL-1β can alter the expression of mTOR. Indeed, our results show that IL-1β-treatment results in a significant increase in the expression of mTOR associated with a significant increase in the expression of MMP-13, CCL2 and CCL5 and a significant decrease in the expression of type II collagen (see online supplementary figure 2). These results suggest that IL-1β may play a critical role in increasing the expression of mTOR during OA.

Silencing of ULK1 can rescue the protective effects of rapamycin

Previously, it has been reported that mTOR regulates ULK1 in combination with AMPK.8 Since ULK1 is the most upstream autophagy inducer and we observed that genetic deletion of mTOR in mouse (in vivo) and rapamycin treatment in human OA chondrocytes (in vitro) resulted in the upregulation of ULK1 and AMPK1, we hypothesised that the ability of mTOR to modulate ULK1 signalling may in part be responsible for controlling autophagy as well as the balance between catabolic and anabolic processes in the articular cartilage. To test this, we pretreated OA chondrocytes with rapamycin and transfected these cells in the presence/absence of ULK1 siRNA to determine if silencing of ULK1 can rescue the protective effects of rapamycin. Indeed, silencing of ULK1 (figure 5E, F) in rapamycin-treated OA chondrocytes resulted in a significant decrease in the expression of LC3B (total LC3 expression and the conversion from LC3BI to LC3BII) (figure 5G–H) as well as mRNA expression of LC3 and ATG5 (figure 5H). Furthermore, silencing of ULK1 in rapamycin-treated OA chondrocytes resulted in a significant increase in the expression of OA catabolic factors (MMP-13 and CCL2) (figure 5H, I). These results suggest that in the articular cartilage, ability of mTOR to suppress autophagy and create imbalance between catabolic and anabolic processes may in part be regulated by shutting down ULK1 pathway.

Discussion

Adult articular cartilage comprises only one type of cells ‘chondrocytes’ that are essential for maintaining the integrity of extracellular matrix as well as imparting/maintaining adequate homeostasis and balance between catabolic and anabolic activities in the cartilage milieu.24 One of the critical events during OA is the loss of chondrocyte cellularity within the articular cartilage that can disrupt the balance between catabolic and anabolic processes, resulting in destruction of the cartilage. However, the mechanisms that control the fate of the chondrocytes in the articular cartilage during normal versus OA conditions remain to be elucidated.

Recent studies by Caramés et al9 ,10 and Sasaki et al11 have suggested a dysregulation in the process of autophagy during OA. Specifically, it has been shown that the expression of some of the key autophagy genes is downregulated in human OA and ageing-related and mouse experimental OA cartilage.9 Subsequently, rapamycin has been shown to induce autophagy and protect mice from experimentally induced OA.10 These results suggest that compromised autophagy may contribute to decreased chondroprotection and development of OA. However, it still remains to be determined what controls autophagy in the articular cartilage.

In the present study, we provide comprehensive in vivo and in vitro evidence that one of the key central factors that control autophagy signalling and articular cartilage homeostasis is mTOR. We first show that mTOR is overexpressed in human OA cartilage (compared with normal human cartilage), as well as mouse and dog models of experimental OA. This upregulation in the expression of mTOR during OA correlates with enhanced chondrocyte apoptosis and decreased expression of key genes involved in the induction of autophagy including ULK1 (most upstream autophagy inducer),25 LC3B (an autophagy structural and functional factor), ATG5 (an autophagy regulator)26 and BNIP3 (an interactor of LC3 in autophagy).27

We then created inducible cartilage-specific mTOR KO mouse (mTOR global KO mouse are not viable)28 to specifically elucidate the in vivo role of mTOR/autophagy signalling in OA pathophysiology. Our results show that the genetic loss of mTOR in the articular cartilage results in increased autophagy signalling, reduced expression of OA catabolic factor MMP-13, decreased chondrocyte apoptosis and significant protection from DMM-induced OA.

To extrapolate our in vivo mouse findings to human OA and dissect the mTOR signalling pathway operative during OA pathogenesis, we treated human OA chondrocytes with rapamycin to inhibit mTOR signalling. Rapamycin-treated OA chondrocytes exhibited suppression in the phosphorylation of rpS6K (confirming the inhibition of mTOR signalling) associated with a significant increase in the expression of AMPK1 and autophagy genes including LC3 (total LC3 expression and the conversion from LC3BI to LC3BII) and ULK1. Interestingly, rapamycin treatment elevated the expression of aggrecan and type II collagen, two major components of cartilage ECM, and decreased the expression of OA catabolic factors including MMP-13 and chemokines (CCL5/RANTES and CCL2/MCP-1), suggesting the potential of mTOR inhibition in correcting the imbalance between catabolic and anabolic factors during OA.

ULK1 is the most upstream autophagy inducer and a key initiator of autophagy process.29 It is believed that the interaction between ULK1/AMPK and mTOR signalling pathway are critical components in the regulation of autophagy.30 In mammalian cells it has been shown that activation of mTORC1 results in phosphorylation of ULK-1 resulting in disruption of interaction between AMPK and ULK-1, thus inhibiting AMPK-mediated ULK1 activation and hence loss of autophagy.8 ,31 Similarly, upon starvation mTORC1 receives stress signals that promote ULK1 complex formation, mTORC1 kinase activity is inhibited and autophagosome formation occurs.29 ,32 Our in vivo and in vitro studies using mTOR KO mouse as well as rapamycin treatment (mTORC1 Inhibition) in human OA chondrocytes respectively show upregulation in the expression of ULK1 and AMPK1 upon genetic deletion of mTOR as well as pharmacological inhibition of mTOR signalling (rapamycin) in OA chondrocytes. To confirm our hypothesis that mTOR controls autophagy in part by modulating ULK1 expression in the articular cartilage, we pretreated OA chondrocytes with rapamycin and transfected these cells in the presence/absence of ULK1 siRNA to determine if silencing of ULK1 can rescue the protective effects of rapamycin. Our results clearly show that silencing of ULK1 in rapamycin-treated OA chondrocytes resulted in a significant decrease in the expression of other autophagy regulators including LC3B and ATG5. Furthermore, silencing of ULK1 in rapamycin-treated OA chondrocytes resulted in a significant increase in the expression of OA catabolic factors (MMP-13 and CCL2). Collectively, these results suggest that in the articular cartilage, ability of mTOR to modulate ULK1/AMPK autophagy pathway may in part be responsible for the imbalance between catabolic and anabolic processes and decreased chondroprotection and ultimately cartilage destruction observed during OA. The involvement of other signalling pathways through which mTOR mediates its actions in the articular cartilage cannot be ruled out and needs further investigation.

Targeting ECM degradative products such as MMPs as well as inflammatory cytokines have yielded disappointing results. Thus far, all MMP inhibitors tested in clinical trials have been unsuccessful due to specificity issues as well as being associated with adverse effects.23 OA is currently managed by using treatments such as acetaminophen, opioids and non-steroidal anti-inflammatory drugs (NSAIDs) including selective cyclooxygenase-2 (COX-2) inhibitors, all of which only provide symptomatic relief. Thus, targeting cellular homeostasis mediators such as mTOR and its downstream signalling by autophagy pathway may be a promising therapeutic strategy to achieve chondroprotection and correct the imbalance between catabolic and anabolic processes during OA and related disorders.

Acknowledgments

We thank Mr. Changshan Geng, Francois-Cyril Jolicoeur, Frederic Pare and Francois Mineau for their technical assistance. We also thanks Dr A Robin Poole for providing C1, 2C antibody.

References

View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Handling editor Tore K Kvien

  • Contributors All authors were either involved in conception and design, or analysis and interpretation of data. Each author was involved in drafting the article and revising it critically for important intellectual content. Each author gave their final approval of the version to be published.

  • Funding This work was supported by Dr Mohit Kapoor's Canadian Institutes of Health Research Operating Grant MOP126016. Y Z is a recipient of Young Investigator Fellowship from The European Society for Clinical and Economic Aspects of Osteoporosis and Osteoarthritis (ESCEO) and Amgen.

  • Competing interests None.

  • Ethics approval CHUM Ethics Committee Board. Comité Institutionnel de protection des animaux (CIPA) de CR-CHUM.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.