Article Text

Download PDFPDF

microRNA-181a-5p antisense oligonucleotides attenuate osteoarthritis in facet and knee joints
  1. Akihiro Nakamura1,2,3,
  2. Yoga Raja Rampersaud1,4,
  3. Sayaka Nakamura1,2,
  4. Anirudh Sharma1,2,
  5. Fanxing Zeng1,2,
  6. Evgeny Rossomacha1,2,
  7. Shabana Amanda Ali1,2,
  8. Roman Krawetz5,
  9. Nigil Haroon1,2,3,
  10. Anthony V Perruccio1,4,6,
  11. Nizar N Mahomed1,4,
  12. Rajiv Gandhi1,4,
  13. Jason S Rockel1,2,
  14. Mohit Kapoor1,2,4,7
  1. 1 Arthritis Program, University Health Network, Toronto, Ontario, Canada
  2. 2 Division of Genetics and Development, Krembil Research Institute, University Health Network, Toronto , Ontario, Canada
  3. 3 Division of Rheumatology, University Health Network, Toronto Western Hospital, Toronto, Ontario, Canada
  4. 4 Department of Surgery, University of Toronto, Ontario, Canada
  5. 5 McCaig Institute for Bone and Joint Health, University of Calgary, Calgary, Alberta, Canada
  6. 6 Institute of Health Policy, Management & Evaluation, Dalla Lana School of Public Health, University of Toronto, Ontario, Canada
  7. 7 Department of Laboratory Medicine and Pathobiology, University of Toronto, Ontario, Canada
  1. Correspondence to Professor Mohit Kapoor, Arthritis Program, University Health Network, Department of Surgery and Department of Laboratory Medicine and Pathobiology, University of Toronto, Division of Genetics and Development, Krembil Research Institute, 60 Leonard Avenue, Toronto, Ontario M5T 2S8, Canada; mkapoor{at}uhnresearch.ca

Abstract

Objectives We recently identified microRNA-181a-5p (miR-181a-5p) as a critical mediator involved in the destruction of lumbar facet joint (FJ) cartilage. In this study, we tested if locked nucleic acid (LNA) miR-181a-5p antisense oligonucleotides (ASO) could be used as a therapeutic to limit articular cartilage degeneration.

Methods We used a variety of experimental models consisting of both human samples and animal models of FJ and knee osteoarthritis (OA) to test the effects of LNA-miR-181a-5p ASO on articular cartilage degeneration. Histopathological analysis including immunohistochemistry and in situ hybridisation were used to detect key OA catabolic markers and microRNA, respectively. Apoptotic/cell death markers were evaluated by flow cytometry. qPCR and immunoblotting were applied to quantify gene and protein expression.

Results miR-181a-5p expression was increased in human FJ OA and knee OA cartilage as well as injury-induced FJ OA (rat) and trauma-induced knee OA (mouse) cartilage compared with control cartilage, correlating with classical OA catabolic markers in human, rat and mouse cartilage. We demonstrated that LNA-miR-181a-5p ASO in rat and mouse chondrocytes reduced the expression of cartilage catabolic and chondrocyte apoptotic/cell death markers in vitro. Treatment of OA-induced rat FJ or mouse knee joints with intra-articular injections of in vivo grade LNA-miR-181a-5p ASO attenuated cartilage destruction, and the expression of catabolic, hypertrophic, apoptotic/cell death and type II collagen breakdown markers. Finally, treatment of LNA-miR-181a-5p ASO in cultures of human knee OA chondrocytes (in vitro) and cartilage explants (ex vivo) further demonstrated its cartilage protective effects.

Conclusions Our data demonstrate, for the first time, that LNA-miR-181a-5p ASO exhibit cartilage-protective effects in FJ and knee OA.

  • osteoarthritis
  • treatment
  • chondrocytes
  • microRNA

Statistics from Altmetric.com

Introduction

Osteoarthritis (OA) is the most common form of arthritis and a leading cause of chronic pain and disability.1 The prevalence of OA is increasing2; however, the aetiology of OA is not fully understood. Currently, OA drugs only target joint pain, and there are no approved therapies in clinical practice that modify disease progression. Although OA is a total joint disease affecting all of the joint tissues including articular cartilage, subchondral bone and synovium, the central characteristic is articular cartilage degeneration. Increased expression of cartilage degrading enzymes such as matrix metalloproteinase-13 (MMP13) and a disintegrin and metalloproteinase with thrombospondin motifs-4 and 5 (ADAMTS4 and 5) in articular chondrocytes catabolise the major cartilage extracellular matrix (ECM) components including type II collagen and aggrecan, respectively, leading to cartilage degeneration.3–6 Chondrocyte loss is a hallmark of OA and contributes to the catabolic phenotype as chondrocyte cell death precedes cartilage degeneration.7–9 In addition, chondrocytes in the non-calcified zone of osteoarthritic cartilage undergo hypertrophy, marked by expression of type X collagen (COL10) and accompanying MMPs and ADAMTSs,10 11 which further contributes to cartilage degeneration. With increased chondrocyte catabolic gene expression, hypertrophy and cell death, the remaining cells are unable to produce sufficient, appropriate articular cartilage ECM, leading to cartilage degradation.

MicroRNAs (miRNAs) are an evolutionarily conserved group of small non-coding RNAs that regulate gene expression.12 13 Mature miRNAs are structurally stable and essential regulators of many physiological processes.14 Increasing evidence suggests that through regulation of catabolism, cell death and ECM production, miRNAs play a significant role in OA pathogenesis.15–17 Recent studies suggest that targeting pathologically expressed miRNAs could be a therapeutic option to treat knee OA.18 19 However, there is sparse evidence supporting the targeting of miRNAs as a genuine therapeutic option for OA. Demonstrating that a specific miRNA can be regulated by a drug with high persistence, target affinity and efficacy in animal models of OA and human OA tissues is necessary for translation to clinical application. We recently identified miR-181a-5p as a critical mediator of facet cartilage degeneration during facet joint (FJ) OA.20 Specifically, our in vitro and in vivo studies showed that miR-181a-5p induces articular cartilage degeneration by promoting inflammation, cartilage catabolism and apoptosis/cell death. These observations prompted us to further test if antisense oligonucleotides (ASO) against miR-181a-5p can exhibit cartilage-protective effects in FJ and other joints such as the knee.

Inhibition of miR-181a-5p in human knee OA chondrocytes has been reported as a potential therapeutic target in OA, with direct activity against target genes known to suppress apoptosis.21 22 Locked nucleic acid (LNA) ASO have been successfully applied in animal studies and some have been tested in human clinical trials.23–25 LNA ASO have ribose rings that are “locked’ in an optimal conformation for increased target binding affinity, stability and resistance to endonucleases. In the present study, we tested whether LNA-miR-181a-5p ASO exhibited cartilage-protective effects using a combination of human in vitro (chondrocyte culture), human ex vivo (cartilage explant culture), in vivo rat FJ OA and mouse knee OA models. Our findings, for the first time, provide crucial evidence that LNA-miR-181a-5p ASO could be a potential therapy to attenuate cartilage destruction in OA.

Results

MiR-181a-5p expression is increased in degenerated human and rat facet OA cartilage

We first employed in situ hybridisation (ISH) to determine the expression of miR-181a-5p in human FJ OA cartilage. To do this, we investigated FJ OA cartilage exhibiting a moderate to severe degree of cartilage degeneration compared with control FJ cartilage exhibiting no detectable or mild degeneration.20 We determined that there was a significant increase in the number of miR-181a-5p-positive chondrocytes in FJ OA cartilage compared with control cartilage (figure 1A-C). Expression of miR-181a-5p was also significantly increased in human FJ OA cartilage compared with control cartilage, as determined by qPCR (figure 1D), validating our previous microarray and qPCR results showing that the expression of miR-181a-5p is markedly elevated with increased severity of facet cartilage degeneration.20 In addition to increased expression of miR-181a-5p, we also observed concomitant increased expression of MMP13, COL10, chondrocyte cell death/apoptosis markers poly (ADP-ribose) polymerase p85 (PARP p85) and cleaved caspase 3, and the collagen type I/II cleavage neoepitope, C1,2C, in FJ OA cartilage compared with control cartilage, as determined by immunohistochemistry (IHC; figure 1E–N; online supplementary figure 1).

Figure 1

MiR-181a-5p expression is increased in degenerated facet joint (FJ) cartilage. (A) Representative safranin O/fast green-stained images of human FJ control cartilage (non or mildly degenerated FJ cartilage) and human FJ osteoarthritis (OA) cartilage (degenerated FJ cartilage). (B) Representative images of in situ hybridisation (ISH)-stained human FJ control and FJ OA cartilage probed for miR-181a-5p, U6 (positive control) and scramble (negative) control. (A and B) Scale bars: 100 µm. (C) Quantification of the percentage of cells positive for miR-181a-5p from sections stained by ISH (n=3/group). (D) Expression of miR-181a-5p in human FJ control cartilage and FJ OA cartilage assessed by quantitative real-time PCR (qPCR) (n=5/group). (E–J) Representative immunohistochemistry (IHC) images of FJ control and FJ OA cartilage stained for matrix metalloproteinase-13 (MMP13; E), type X collagen (COL10; F), cell death (apoptosis) markers PARP p85 (G) and cleaved caspase 3 (H), collagen type I/II cleavage (C1,2C; I) and negative control (J). n=4 FJs/group; scale bar: 100 µm. (K–N) Quantification of the percentage of cells positive for MMP13 (K), COL10 (L), PARP p85 (M) and cleaved caspase 3 (N) from IHC images. (O) Representative images of safranin O/fast green-stained control rat FJ cartilage and OA cartilage. (P) Representative images of ISH-stained rat FJ control and OA tissue sections for miR-181a-5p, U6 and scramble control (n=3/group). (O and P) Scale bars: 100 µm. (Q) Quantification of the percentage of cells positive for miR-181a-5p from sections stained by ISH (n=3/group). (C, D, K–N and Q) Data are presented as scatter dot plots (error bars denote means±SD). Significant differences in the number of positive cells or levels of expression between the control and OA groups were determined using two-tailed Student’s t-tests. *P<0.05; **p<0.01.

To determine if elevated expression of miR-181a-5p and increased cartilage degeneration was conserved in a rat model of FJ OA, we induced FJ OA by needle puncture injury into two levels of the lumbar spine (L4/5 and L5/6).26 FJs harvested at 12-week post-needle puncture were analysed by ISH. We found a significant increase in the number of miR-181a-5p-positive chondrocytes in FJ OA cartilage compared with sham-operated (unpunctured) control cartilage (figure 1O–Q). Collectively, these results show that miR-181a-5p expression is elevated in both human and rat degenerated facet cartilage in FJ OA.

MiR-181a-5p expression is increased in degenerated human and mouse knee OA cartilage

The expression of miR-181a-5p in human knee and hip is increased in OA compared with normal chondrocytes21 27. Thus, we investigated if miR-181a-5p was upregulated in degenerated OA cartilage of the knee, one of the most common joints affected by OA. We demonstrated that the number of miR-181a-5p-positive chondrocytes (ISH) as well as the expression of miR-181a-5p (qPCR) was significantly increased in human knee OA cartilage tissue compared with control knee cartilage (figure 2A–D), consistent with expression in human facet chondrocytes.

Figure 2

MiR-181a-5p expression is increased in human and mouse knee osteoarthritis (OA) cartilage. (A) Representative safranin O/fast green-stained human knee control and OA cartilage. (B) Representative in situ hybridisation (ISH)-stained images of human knee control and OA cartilage probed for miR-181a-5p, U6 and scramble control. (A and B) Scale bars: 100 µm. (C) Quantification of the percentage of cells positive for miR-181a-5p from sections stained by ISH. n=3/group. (D) Expression of miR-181a-5p in human knee control and OA cartilage assessed by quantitative real-time PCR (qPCR). n=5/group. (E) Representative safranin O/fast green stained images of mouse control and destabilisation of the medial meniscus-induced-OA tibias. (F) Representative ISH-stained images of mouse control and DMM-induced OA tibias for miR-181a-5p, U6 and scramble control. (E and F) Scale bars: 100 µm. (G) Quantification of the percentage of cells positive for miR-181a-5p from mouse tibias stained by ISH. n=3/group. (C, D and G) Data are presented as scatter dot plots (error bars denote means±SD). Significant differences in the levels of expression between the control and OA groups were determined using two-tailed Student’s t-tests. *P<0.05.

Next, we investigated the expression of miR-181a-5p in the surgical destabilisation of the medial meniscus (DMM) mouse model of OA. Joint tissues harvested at 10 weeks post-surgery were subjected to ISH to determine the expression of miR-181a-5p. We found a significant increase in miR-181a-5p-positive chondrocytes in mouse knee OA cartilage compared with sham control cartilage (figure 2E–G). These results indicate that, similar to FJ OA cartilage, miR-181a-5p expression is increased in human and mouse knee OA cartilage.

LNA-miR-181a-5p ASO partially rescue interleukin (IL)-1β-induced cartilage catabolic phenotypes in rat FJ and mouse knee chondrocytes

Since we determined that expression of miR-181a-5p is increased in rat FJ and mouse knee OA cartilage compared with control cartilage, we next tested the effect of ASO against miR-181a-5p (LNA-miR-181a-5p ASO), which have perfect sequence complementary to rat, mouse and human miR-181a-5p. Scramble control oligonucleotides (SCO) with similar sequence length and LNA design to the ASO but lacking homology to any known mouse, rat or human miRNA or mRNA sequence were used as a control.

IL-1β is a major proinflammatory cytokine implicated in OA.28 Rat FJ and mouse knee chondrocytes were treated with or without IL-1β in the presence of LNA-miR-181a-5p ASO or SCO (online supplementary figure 2A). The expression of miR-181a-5p was significantly enhanced in response to IL-1β treatment in both rat FJ and mouse knee chondrocytes (online supplementary figure 2B). In the presence of SCO, IL-1β treatment significantly increased the expression of Mmp13 in both rat FJ and mouse knee chondrocytes; however, treatment with LNA-miR-181a-5p ASO significantly attenuated IL-1β-induced Mmp13 (online supplementary figure 2C and D). It should be noted that even though the reduction of Mmp13 expression by LNA-miR-181a-5p ASO was significant, the overall attenuation was only partial and not profound.

We next tested whether LNA-miR-181a-5p ASO could protect chondrocytes from IL-1β-induced cell death. Flow cytometry analysis was performed with rat chondrocytes stained with near-infrared (IR) dead cell stain, which detects live and dead cells based on membrane permeability. Results showed that IL-1β-treatment increased cell death in rat FJ chondrocytes while cotreatment with LNA-miR-181a-5p ASO reduced the IL-1β-induced increase in cell death (online supplementary figures 2E,F and 3). To assess cell death via apoptosis in mouse chondrocytes, we used Annexin V and 7-aminoactinomycin D (AAD). Annexin V+/7-AAD− cells are indicative of early apoptotic cells, while Annexin V+/7-AAD+label late apoptotic or dead cells. Similar to rat chondrocytes, IL-1β-induced mouse chondrocyte apoptosis and cell death was also significantly attenuated by treatment with the LNA-miR-181a-5p ASO (online supplementary figures 2G,H and 4). Collectively, these results indicate that LNA-miR-181a-5p ASO partially reverse IL-1β-induced catabolic and cell death activities in chondrocytes from multiple species (rat and mouse) and joints (facet and knee) in vitro.

In vivo grade LNA-miR-181a-5p ASO attenuate the severity of cartilage degeneration in vivo using a rat FJ model of OA

Needle puncture of the rat FJ produces OA-like pathologies including significant degeneration of facet cartilage, loss of chondrocyte cellularity and proteoglycan depletion compared with sham surgery (figure 3A & B, online supplementary figure 5) in addition to increased expression of Mmp13 and decreased expression of type II collagen (Col2a1) (online supplementary figure 6A and B). Thus, to determine whether LNA-miR-181a-5p ASO have a protective effect on cartilage degeneration in vivo, we intra-articularly injected in vivo grade LNA-miR-181a-5p ASO into rat lumbar FJ following needle puncture injury-induced FJ OA or sham control (unpunctured) surgery. At 3 and 6 weeks post-needle puncture, we injected in vivo grade LNA-miR-181a-5p ASO (right side of FJ; 2 µL of 5 µg/µL per FJ) or in vivo grade SCO (left side of FJ; 2 µL of 5 µg/µL per FJ). At 12 weeks post-needle puncture, FJ tissues were collected and assessed by histology and IHC (figure 3A).

Figure 3

Intra-articular injection of in vivo grade LNA-miR-181a-5p antisense oligonucleotides (IVG ASO) attenuate facet joint (FJ) cartilage degeneration in a rat model of FJ osteoarthritis (OA). (A) Schematic of the needle puncture (injury)-induced FJ OA model and injection time-course. IVG scramble control oligonucleotides (IVG SCO) or IVG ASO were injected at 3 and 6 weeks post-needle puncture and harvested at 12 weeks. (B) Representative safranin O/fast green-stained tissue sections from sham and injury-induced rat FJs (L4/5). Scale bar: 100 μm. (C) Representative in situ hybridisation (ISH)-stained images probed for miR-181a-5p, U6 (positive control), and scramble (negative) control of rat FJ cartilage with puncture and IVG ASO injection compared with puncture and IVG SCO injection. n=3/group. Scale bars: 100 μm. (D) Quantification of the percentage of cells positive for miR-181a-5p from injury-induced rat FJs injected with IVG SCO or IVG ASO. n=3/group. (E) Representative images of safranin O/fast green-stained tissue sections from injury-induced rat FJs (L4/5 and L5/6) injected with either IVG SCO or IVG ASO. n=10 FJs/group. Scale bars: 100 μm. (F and G) Osteoarthritis Research Society International (OARSI) score (F) and cellularity (G) of sham (n=6), needle puncture without injection (n=6), needle puncture with IVG SCO and needle puncture with IVG ASO tissue sections (n=10 /group). (D, F and G) Data are presented as scatter dot plots (error bars denote means±SD). Significant differences in the levels of expression between the groups were determined using two-tailed Student’s t-tests or one-way analysis of variance followed with Tukey’s post-hoc tests. *P<0.05; **p<0.01.

We used ISH to first confirm whether in vivo grade LNA-miR-181a-5p ASO had any effect on the detection of miR-181a-5p in FJ cartilage. We found a significant reduction in the number of miR-181a-5p positive cells in FJs injected with the in vivo grade LNA-miR-181a-5p ASO as compared with those injected with in vivo grade SCO at 12 weeks post-injury (figure 3C and D). Using qPCR analysis of the FJ cartilage, we also detected lower levels of miR-181a-5p at 6 weeks post-needle puncture in joints injected with in vivo grade LNA-miR-181a-5p ASO compared with those injected with in vivo grade SCO (online supplementary figure 7A and B).

The joint protective effects of in vivo grade LNA-miR-181a-5p ASO were evident in our injury-induced FJ OA model. Histopathological analysis of FJ cartilage using safranin O/fast green staining at 12 weeks post-injury showed marked reduction in the degree of FJ cartilage degeneration [as determined by Osteoarthritis Research Society International (OARSI) scoring] associated with reduced proteoglycan loss and increased chondrocyte cellularity in the in vivo grade LNA-miR-181a-5p ASO-injected group compared with SCO-injected group (figure 3E–G). Consistent with our previous report showing that miR-181a-5p mimic promotes FJ cartilage degeneration in vivo,20 these results demonstrate that in vivo grade LNA-miR-181a-5p ASO protect FJ cartilage from injury-induced degeneration.

In vivo grade LNA-miR-181a-5p ASO attenuate the severity of cartilage degeneration in vivo using the mouse DMM model of OA

After identifying the protective effects of in vivo grade LNA-miR-181a-5p ASO in a rat model of FJ OA, we next examined its ability to attenuate cartilage degeneration in a mouse model of trauma-induced knee OA. We used one of the most validated models of knee OA, which is the mouse DMM model. DMM surgery was performed on the right knees of 12-week-old C57BL/6J male mice, as previously reported.29–31 Similar to our in vivo FJ OA study, we intra-articulary injected in vivo grade LNA-miR-181a-5p ASO (3 µL of 1 µg/µL per knee joint) or in vivo grade SCO (3 µL of 1 µg/µL per knee joint) into mouse knee joints at 2 and 4 weeks post-DMM or sham surgery and collected the knee joint tissues at 10 weeks post-surgery (figure 4A).

Figure 4

Intra-articular injection of in vivo grade LNA-miR-181a-5p antisense oligonucleotides (IVG ASO) attenuate cartilage degeneration in a model of mouse knee osteoarthritis (OA). (A) Schematic of the injection time-course of the mouse destabilisation of the medial meniscus (DMM)-induced knee OA model. IVG scramble control oligonucleotides (IVG SCO) or IVG ASO were injected at 2 and 4 weeks post-surgery and harvested at 10 weeks. (B) Representative in situ hybridisation-stained images of mouse knee cartilage from DMM-surgery injected with IVG SCO or IVG ASO probed for miR-181a-5p, U6 (positive control), and scramble (negative) control. n=3/ group. Scale bar: 100 μm. (C) Quantification of the percentage of cells positive for miR-181a-5p from DMM mouse knee joints injected with IVG SCO or IVG ASO. n=3/group. (D) Representative images of safranin O/fast green-stained tissue sections of sham surgery and SCO-injected or ASO-injected joints. Scale bars: 100 μm. (E) Representative images of safranin O/fast green-stained tissue sections of DMM-induced OA mouse knee joints injected with IVG SCO or IVG ASO. Scale bars: 100 μm. n=5/group. (F and G) Osteoarthritis Research Society International (OARSI) score (F) and cellularity (G) of DMM-induced OA mouse tissue sections injected with IVG SCO or IVG ASO. n=5/group. (C, F and G) Data are presented as scatter dot plots (error bars denote means±SD). Significant differences between groups were determined using two-tailed Student’s t-tests. *P<0.05; **p<0.01.

Similar to the rat FJ cartilage, detection of miR-181a-5p was significantly reduced in cartilage from knee joints injected with the in vivo grade LNA-miR-181a-5p ASO compared with in vivo grade SCO, as assessed by ISH (figure 4B,C). Mouse knee cartilage at 4 weeks post-surgery (after an injection at 2 weeks; online supplementary figure 8A) also showed reduced detection of miR-181a-5p after joints were injected with the LNA-miR-181a-5p ASO, as assessed by qPCR, when compared with those injected with SCO (online supplementary figure 8B).

We next determined the effect of in vivo grade LNA-miR-181a-5p ASO on the pathogenesis of knee OA. There was no marked difference in the severity of cartilage degeneration between in vivo grade SCO and in vivo grade LNA-miR-181a-5p ASO in the sham surgery groups (figure 4D). In contrast, we found significant attenuation of DMM-induced cartilage degeneration (by OARSI scoring) and a significant reduction in chondrocyte loss in knee joints injected with in vivo grade LNA-miR-181a-5p ASO compared with the in vivo grade SCO-injected knees (figure 4E–G). Overall, in vivo grade LNA-miR-181a-5p ASO exhibit cartilage-protective effects in both facet and knee joints during OA.

We also performed histopathological analysis to assess the severity of synovitis in the knee OA model in response to in vivo grade LNA-miR-181a-5p ASO injection compared with the in vivo grade SCO-injected knees. We did not observe any significant differences in the degree of synovitis between the injection groups (online supplementary figure 9).

In vivo grade LNA-miR-181a-5p ASO attenuate the expression of OA phenotypic makers

Since we observed a reduction in the loss of proteoglycans and chondrocyte cellularity in both rat FJ OA cartilage and mouse knee OA cartilage injected with the in vivo grade LNA-miR-181a-5p ASO, we examined the effect of in vivo grade LNA-miR-181a-5p ASO on phenotypic markers of OA. We found a significant decrease in the percentage of cells positive for MMP13, COL10, PARP p85 and cleaved caspase 3, as well as a reduction in the expression of type II collagen breakdown marker (C1,2C) in the in vivo grade LNA-miR-181a-5p ASO-injected joints compared with in vivo grade SCO-injected joints in both rat FJ and mouse knee joints, as assessed by IHC (figure 5A–T; online supplementary figures 10 and 11). These findings suggest that in vivo grade LNA-miR-181a-5p ASO may impart cartilage-protective effects by reducing chondrocyte hypertrophy, cell death and catabolic activity in vivo.

Figure 5

In vivo injection of in vivo grade LNA-miR-181a-5p antisense oligonucleotides (IVG ASO) decrease expression of cartilage catabolic, hypertrophic, cell death and type II collagen breakdown markers in rat facet joint (FJ) cartilage and mouse knee cartilage. (A–F) Representative images of immunohistochemistry (IHC)-stained tissues of needle puncture (injury)-induced FJs (at 12 weeks post-injury) injected with IVG scramble control oligonucleotides (IVG SCO) or IVG ASO for matrix metallopeptidase 13 (MMP13; A), type X collagen (COL10; B), PARP p85 (C), cleaved caspase 3 (D) and collagen type I/II cleavage (C1,2C; E) and negative control (F). n=4 animals/group (online supplementary figure 10). Scale bar: 100 μm. (G–J) Quantification of the percentage of cells positive for MMP13 (G), COL10 (H), PARP p85 (I) and cleaved caspase 3 (J) from IHC-stained tissues of injury-induced rat FJs injected with IVG SCO or IVG ASO. n=4/group. (K–P) Representative images of IHC-stained DMM-induced OA mouse knee joint sections (at 10 weeks post-surgery) injected with IVG SCO or IVG ASO for MMP13 (K) , COL10 (L), PARP p85 (M), cleaved caspase 3 (N), C1,2C (O) and negative control (P). n=4 animals/group (online supplementary figure 11). Scale bar: 100 μm. (Q-T) Quantification of the percentage of cells positive for MMP13 (Q), COL10 (R), PARP p85 (S) and cleaved caspase 3 (T) from IHC-stained tissues of DMM-induced mouse knee OA joints injected with IVG SCO or IVG ASO. n=4/group. (G–J, Q–T) Data are presented as scatter dot plots (error bars denote means±SD). Significant differences in the quantification of percentage of positive cells were determined using two-tailed Student’s t-tests. *P<0.05; **p<0.01.

LNA-miR-181a-5p ASO reduce the expression of catabolic and cell death markers in human articular chondrocytes and cartilage

To evaluate the clinical relevance of LNA-miR-181a-5p ASO, we next investigated whether the LNA-miR-181a-5p ASO could reduce the expression of catabolic and cell death markers in cultured human OA chondrocytes. Chondrocytes were isolated from knee cartilage of patients with OA undergoing total knee replacement (TKR) surgery and treated with in vitro LNA-miR-181a-5p ASO or SCO in the presence or absence of IL-1β. We found that chondrocytes significantly increased the expression of miR-181a-5p in response to IL-1β treatment in the presence of SCO (online supplementary figure 12). In contrast to SCO-treated cultures, treatment with LNA-miR-181a-5p ASO suppressed the IL-1β-induced expression of MMP13 and PARP p85, (figure 6A–C) and partially rescued the expression of COL2A1 in response to IL-1β treatment (figure 6A). Furthermore, flow cytometry analysis showed a significant decrease in IL-1β-induced chondrocyte apoptotic/cell death markers (for Annexin V/7-AAD) in cultures treated with LNA-miR-181a-5p ASO compared with those treated with SCO (online supplementary figure 13).

Figure 6

LNA-miR-181a-5p antisense oligonucleotides (ASO) attenuate matrix metalloproteinase-13 (MMP13) expression and increase type II collagen (COL2A1) expression in human knee osteoarthritis (OA) chondrocytes in vitro and human knee OA cartilage explants ex vivo. (A) Quantitative real-time PCR (qPCR) analysis of MMP13 and COL2A1 RNA levels from cultured human knee OA chondrocytes treated with or without interleukin (IL)-1β and ASO or scramble control oligonucleotides (SCO; n=7/group). (B and C) Immunoblot (B) and densitometry analysis (C) of MMP13 and PARP p85 protein from cultured human knee OA chondrocytes treated with or without IL-1β and ASO or SCO (n=5/group). Uncut membranes are shown in online supplementary figure 14. (D) Flow cytometric analysis of annexin V and 7-aminoactinomycin D (AAD) in cultured human kneeOA chondrocytes treated with or without IL-1β and ASO or SCO (n=4/group). (E) Determination of the percentage of cells positive for Annexin V+/7-AAD+ treated with IL-1β and without IL-1β. (F) Schematic diagram of the extraction of human knee OA cartilage and treatment ex vivo (explant) with either SCO or ASO. (G) qPCR analysis of MMP13, COL10A1 and COL2A1 RNA levels from cultured human knee OA cartilage treated with ASO or SCO (n=5/group). (H and I) Immunoblot (H) and densitometry analysis (I) of MMP13 protein from human knee OA cartilage treated with ASO or SCO (n=5/group). Uncut membranes are shown in online supplementary figure 14. (A, C, E, G, I) Data are presented as scatter dot plots (error bars denote means±SD). Significant differences in the levels of expression between groups were determined using two-way analysis of variance followed byTukey’s post-hoc tests or two-tailed Student’s t tests. (A and C) **P<0.01, SCO without IL-1β versus SCO with IL-1β. ¶¶P<0.01, SCO versus ASO in the absence of IL-1β. †P<0.05, ††P<0.01, SCO versus ASO in the presence of IL-1β. (E, G, I) *P<0.05; **p<0.01, SCO versus ASO.

To test these findings in a system that reflects the chondrocyte microenvironment, we performed ex vivo culture of human knee OA cartilage obtained from TKR surgery and treated the explants with LNA-miR-181a-5p ASO or SCO for 24 hours (figure 6F). Consistent with our in vitro culture studies using human knee chondrocytes, human knee cartilage explants treated with the LNA-miR-181a-5p ASO exhibited significantly reduced expression of MMP13 and COL10A1 and increased expression of COL2A1 (figure 6G–I), as compared with explant cultures treated with SCO. Overall, these results provide crucial evidence of cartilage protective effects of the LNA-miR-181a-5p ASO in preclinical animal models (in vivo) and human OA cells/tissues (in vitro and ex vivo).

Discussion

In this study, we provide the first evidence that intra-articular injection of in vivo grade LNA-miR-181a-5p ASO can attenuate cartilage degeneration in preclinical models of FJ and knee OA. The in vivo grade LNA-miR-181a-5p ASO consist of a modified oligonucleotide targeting miR-181a-5p. In addition to greater stability and increased target affinity, in vivo grade LNA-miR ASO are shorter (12–16 nucleotides) in length, compared with other in vitro grade ASO (~20 nucleotides), enabling more efficient cellular uptake via natural mechanisms when administered to animal models in vivo.32 In vivo grade LNA-miR ASO can persist for 2–3 weeks in vital organs such as the liver and kidney after systemic injection in mice.33 This rapid uptake and long retention are features that make these types of oligonucleotides promising for therapeutic use.

According to ClinicalTrial.gov (https://clinicaltrials.gov/), two clinical trials have tested the therapeutic potential of LNA-miRNAs. Miravirsen, a short LNA-miR-122 ASO for the treatment of hepatitis C, has been studied in a completed phase IIa clinical trial (Clinical trial ID: NCT01200420).24 TargomiRs, an LNA-miR-16 mimic developed for the treatment of malignant pleural mesothelioma or non-small cell lung cancer, has recently completed phase I clinical trial (Clinical trial ID: NCT02369198).34 These clinical trials were performed with systemic drug administration, resulting in some adverse events including loss of consciousness with Miravirsen,24 and cardiomyopathy and cardiac ischaemia with TargomiRs.34 Adverse events caused by drugs are major concerns in clinical settings. Indeed, there were two phase I clinical trials with MRX34, a liposomal miR-34a mimic, as a potential treatment for advanced solid tumours (Clinical trial ID: NCT01829971) and skin cancer melanoma (Clinical trial ID: NCT02862145). However, these trials were terminated or withdrawn due to adverse events.35 Thus, local administration is of interest, if capable of being retained in the target organ, as this would limit systemic effects of the LNA-miRs and potential adverse events of the therapies.

To our knowledge, there are no current or past clinical trials using local injection of in vivo grade LNA-miR ASO. However, previous studies of local delivery using in vivo animal studies have demonstrated encouraging therapeutic potential. For instance, intratracheally instillated in vivo grade LNA-miR-21 ASO remarkably suppressed bleomycin-induced lung fibrosis in mice.36 Furthermore, delivering in vivo grade LNA-miR-92a ASO to the heart via intracoronary catheter showed a more prominent protective effect for ischaemia/reperfusion injury than intravenous injection in a pig model.37 Consistent with the positive outcomes using a local mode of administration from these studies, intra-articular injection of in vivo grade LNA-miR-181a-5p ASO in our current study further confirmed the promising potential of in vivo grade LNA-miR ASO delivered by local injection. In vivo treatment of injury-induced rat FJ OA or trauma-induced mouse knee OA joints with intra-articular injections of in vivo grade LNA-miR-181a-5p ASO significantly attenuated cartilage destruction associated with decreased expression of OA catabolic, hypertrophic, apoptotic/cell death and type II collagen breakdown markers. The ability of in vivo grade LNA-miR-181a-5p ASO to reduce catabolic and cell death activity in the cartilage is further supported by our human cartilage explant and chondrocyte culture studies obtained from human FJ and knee OA as well as rat and mouse OA chondrocytes.

This study is not without some limitations. First, minimal off-target effects on unrelated longer RNAs is one of the features of the LNA-miRNA ASO; off-target effects mediating some of the observed study outcomes remains a possibility. Second, although we clearly observed cartilage-protective effects and significant decreases in the expression of catabolic and cell death markers as consequences of in vivo grade LNA-miR-181a-5p ASO treatment, functional assays to directly examine the inhibitory activity of the in vivo grade LNA-miR-181a-5p ASO to miR-181a-5p are needed. However, studies using luciferase activity of direct target genes of miR-181a-5p, including GPD1L and PTEN (both of which suppress apoptosis), indicate that in vitro grade miR-181a-5p ASO can attenuate miR-181a-5p-mediated loss of expression and induction of apoptosis,21 22 suggesting that direct inhibition of endogenous miR-181a-5p-target binding is likely contributing to part of the responses seen in our studies. Finally, in vivo grade LNA-miR-181a-5p ASO and ISH probes for miR-181a-5p have some sequence overlap; however, the LNA-modification renders the ASO with higher binding affinity and stability to targets. Thus, probes to detect endogenous miR-181a-5p may not be able to compete for targets already bound in a complex with in vivo grade LNA-miR-181a-5p ASO. This suggests that the expression of endogenous miR-181a-5p may not be altered but could instead form a non-functional complex with the in vivo grade LNA-miR-181a-5p ASO and simply be undetectable.

Overall, using two distinct animal OA models in addition to human cells and tissues, we provide the first evidence of cartilage protective effects of LNA-miR-181a-5p ASO. This is the first report showing that LNA-miR-181a-5p ASO attenuate cartilage degeneration during OA across joints (knee and FJ), animal models (rat FJ OA and mouse knee OA models) and species (human, rat and mouse).

Material and methods

Detailed experimental procedures are provided in the online supplemental materials and methods.

Acknowledgments

Authors would like to thank Amanda Weston, Kim Perry and Sarah Gabrial for their help collecting facet and knee cartilage. Authors would also like to acknowledge the help of members of the Arthritis Program at the Toronto Western Hospital (Toronto).

References

Footnotes

  • Handling editor Josef S Smolen

  • Contributors AN was involved in the conception and design of the study, acquisition of data, analysis and interpretation of data, drafting the article, revising it critically for important intellectual content and approved final version of the manuscript. YRR was involved in the design of the study, performed spine surgery in patients with LSS or LDH, provided tissues, performed MRI rading, drafting the article, revising it critically for important intellectual content and approved final version of the manuscript. SN and AS performed animal surgeries to induce osteoarthritis models, inject miRNA antisense oligonucleotide into rat facet joint and mouse knee joints, tissue dissections along with AN, drafting the article, revising it critically for important intellectual content and approved final version of the manuscript. FZ was involved to perform flow cytometry along with AN, drafting the article, revising it critically for important intellectual content and approved final version of the manuscript. ER was involved to perform histology for making sections and safranin O staining, drafting the article, revising it critically for important intellectual content and approved final version of the manuscript. SAA, RK, NH, AVP, NNM, RG and JSR were involved in the interpretation of data, drafting the article, revising it critically for important intellectual content and approved final version of the manuscript. MK was involved in the conception and design of the study, interpretation of data, drafting the article, revising it critically for important intellectual content and approved final version of the manuscript.

  • Funding This study was supported by grants to MK from the Krembil Foundation, the Canadian Institute of Health Research (FRN: 156299) and Campaign to Cure Arthritis via the Toronto General and Western Foundation, University Health Network, Toronto. AN and SAA are recipients of postdoctoral fellowship funding from the Krembil Research Institute (University Health Network).

  • Competing interests A US provisional patent application (62/299,305, filed 24 February 2016) and a PCT international patent application (PCT/CA2017/000019, filed 31 January 2017) have been filed in respect of therapeutic and diagnostic uses of miRNA-181a-5p.

  • Patient consent Obtained.

  • Ethics approval Human facet and knee cartilage were obtained under the institutional approval.

  • 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.