Objective Netrin-1 is a chemorepulsant and matrix protein expressed during and required for osteoclast differentiation, which also plays a role in inflammation by preventing macrophage egress. Because wear particle-induced osteolysis requires osteoclast-mediated destruction of bone, we hypothesised that blockade of Netrin-1 or Unc5b, a receptor for Netrin-1, may diminish this pathological condition.
Methods C57BL/6 mice, 6-8 weeks old, had 3 mg of ultrahigh-molecular-weight polyethylene particles implanted over the calvaria and then received 10 µg of monoclonal antibodies for Netrin-1 or its receptors, Unc5b and deleted in colon cancer (DCC), injected intraperitoneally on a weekly basis. After 2 weeks, micro-computed tomography and histology analysis were performed. Netrin-1 expression was analysed in human tissue obtained following primary prosthesis implantation or after prosthesis revision for peri-implant osteolysis and aseptic implant loosening.
Results Weekly injection of anti-Netrin-1 or anti-Unc5b-antibodies significantly reduced particle-induced bone pitting in calvaria exposed to wear particles (46±4% and 49±3% of control bone pitting, respectively, p<0.001) but anti-DCC antibody did not affect inflammatory osteolysis (80±7% of control bone pitting, p=ns). Anti-Netrin-1 or anti-Unc5b, but not anti-DCC, antibody treatment markedly reduced the inflammatory infiltrate and the number of tartrate resistance acid phosphatase (TRAP)-positive osteoclasts (7±1, 4±1 and 14±1 cells/high power field (hpf), respectively, vs 12±1 cells/hpf for control, p<0.001), with no significant changes in alkaline phosphatase-positive osteoblasts on bone-forming surfaces in any antibody-treated group. Netrin-1 immunostaining colocalised with CD68 staining for macrophages. The peri-implant tissues of patients undergoing prosthesis revision surgery showed an increase in Netrin-1 expression, whereas there was little Netrin-1 expression in soft tissues removed at the time of primary joint replacement.
Conclusions These results demonstrate a unique role for Netrin-1 in osteoclast biology and inflammation and may be a novel target for prevention/treatment of inflammatory osteolysis.
- Bone Mineral Density
Statistics from Altmetric.com
Total hip and knee replacements are successful surgical interventions with success rates close to 90% at 10–25 years after surgery in terms of reducing pain, improving function and enhancing quality of life in patients.1 However, despite improvements in surgical techniques, material and implant designs, nearly 40 000 hip arthroplasties have to be revised each year in the USA,1 and it is expected that the rates of revision will increase by 137% for total hip and by 601% for total knee revisions over the next 25 years.2 Periprosthetic inflammatory bone destruction due to wear particles is the most common cause of prosthesis failure requiring revision.3
Wear particles consist of debris shed from joint prostheses, whether polymeric, metallic or ceramic, and these particles are phagocytosed by macrophages and other local cells, which become activated.4 ,5 Further recruitment of both macrophages and osteoclasts (bone-resorbing cells) to the site6 ,7 leads to upregulation of pathways leading to bone destruction, and downregulation of bone formation.8 ,9 To date, no biological approaches to limiting the chronic inflammatory reaction associated with wear particles have proved clinically successful.10
Netrin-1, a secreted laminin-related protein, is a member of the axonal guidance protein family that inhibits migration of monocytes, neutrophils and lymphocytes by activation of its receptors A2BR (adenosine A2B receptor) and Unc5.11–13 Acting through the Unc5b receptor, Netrin-1 reduces renal ischaemia–reperfusion injury and the associated renal inflammation.14 Netrin-1 is expressed on vascular endothelium, where it is regulated by infection and inflammatory cytokines, and inhibits inflammatory cell migration into tissues and its downregulation at the onset of sepsis/inflammation may facilitate leucocyte recruitment.11 Netrin-1 also promotes chronic inflammation in atherogenesis and diet-induced obesity; tissue macrophages in atherosclerotic plaques and obese adipose tissue increase their expression of Netrin-1,15 ,16 which promotes macrophage retention and survival17 further increasing chronic inflammation. Recently we have demonstrated that Netrin-1 is an autocrine and paracrine regulator of osteoclast differentiation.18 In osteoclast precursors, the binding of Netrin-1 to its receptor Unc5b, but not to its alternative receptor deleted in colon cancer (DCC), triggers the signalling cascade involved in the activation of the small GTPase RhoA by a mechanism that has not previously been observed in osteoclast precursors. Activation of RhoA via LARG (regulator of Rho-GEF subfamily) and RGMa (repulsive guidance molecule) leads to the cytoskeletal rearrangements required for osteoclast fusion and differentiation.18
Here we tested the hypothesis that blockade of Netrin-1 or Unc5b might be a novel approach to the prevention of osteoclast-mediated bone resorption and inflammation at the site of prosthesis shedding of wear particles.
Wear particle preparation
Ultrahigh-molecular-weight polyethylene (UHMWPE) particles, a gift of P H Wooley (Via Christi Regional Medical Center), had a mean particle size (equivalent circle diameter) of 1.74±1.43 μm (range 0.05–11.06), with more than 34% of the particles smaller than 1 μm. For decontamination from endotoxins, the particles were washed twice in 70% ethanol for 24 h at room temperature. The particles were washed in phosphate-buffered saline and dried in a desiccator.
All protocols followed internationally recognised guidelines and were approved by the NYU SoM Institutional Animal Care and Use Committee. Osteolysis was induced by implantation of UHMWPE particles over the mouse calvaria, as we have previously described.19 Briefly, male C57BL/6 mice, 6–8 weeks old, were anesthetised by intraperitoneal injection of 100 mg/kg ketamine and 10 mg/kg xylazine, and a 1-cm midline sagittal incision was made over the calvaria anterior to the line connecting both ears as previously described.19 Five mice received no particles (sham, untreated mice) and the incision was closed without any further intervention. The remaining animals (n=40) received 3 mg of dried UHMWPE particles. Of these animals, 10 mice received intraperitoneally 100 µL of 0.9% saline (control) and the rest received (in a final volume of 100 µL) 10 µg of mouse monoclonal Netrin-1, Unc5b or DCC antibodies (n=10 each), beginning immediately before surgery and continuing on a weekly basis until sacrifice (2 weeks later). Water and food were given ad libitum until sacrifice. Animals were sacrificed after 14 days in a CO2 chamber and the calvaria were removed, fixed, and prepared for micro-computed tomography (CT), histological staining and cytokine measurements.
See online supplementary text.
Micro-X-ray CT (micro-CT) analysis
After sacrifice, five calvaria per treatment group were fixed in 70% ethanol and prepared for high-resolution micro-CT as previously described to perform qualitative and quantitative analyses of resorbing areas in murine calvarial bone.19 Analyses were performed in the NYU College of Dentistry micro-CT core with a Skyscan 1172 micro-CT (Bruker, Wisconsin, USA) using the previously described parameters20: 60 kV, 167 μA, 9.7 μ pixel size, 2000×1332 matrix, 0.3° rotation steps, six averages, movement correction of 10, 0.5 mm Al filter and two segments scanned per sample (56 min/segment). Images were reconstructed using the Skyscan NRECON software (histogram range 0–0.065, beam hardening correction of 35, Gaussian smoothing (factor 1), ring artefact correction of seven). For quantitative analysis of UHMWPE particle-induced osteolysis, a square-shaped region of interest across the parietal bone of approximately 4 mm right and left of the midline suture of the skull was placed in one of the two-dimensional reconstructed slices, as described previously,19 ,21 and a Matlab software application was used to analyse calvarial bone resorption, as previously described.19
Calvariae were fixed in 4% paraformaldehyde for 48 h, followed by decalcification in 10% EDTA for 4 weeks and paraffin embedding (n=5 per treatment). Sections (5 µm) were cut and H&E staining was performed. Inflammatory infiltration in mid-sagittal suture areas was quantified from five images per animal using Sigma Scan Pro Image V.5.0.0 software.
Tartrate resistance acid phosphatase (TRAP) staining was carried out in a TRAP buffer (0.1 M acetate buffer, 0.3 M sodium tartrate 10 mg/mL Naphtol AS-MX phosphate, 0.1% Triton X-100, 0.3 mg/mL Fast Red Violet LB (Sigma-Aldrich, Missouri, USA)). After deparaffinisation and acetate buffer incubation, samples were incubated in TRAP buffer for 30 min and counterstained with Fast Green.
Immunohistochemistry analysis for Netrin-1, Unc5b, DCC, alkaline phosphatase (ALP), cathepsin K and CD68 were carried out as previously described19 (see online supplementary text).
Images were observed in a Leica microscope equipped with SlidePath Digital Image Hub V.3.0 software, or under light microscope (Nikon) equipped with Nis Elements F3.0 SP7 software.
Netrin-1 expression in implant biopsies
Netrin-1 immunostaining was performed in human tissue obtained from donors following primary prosthesis implantation or after prosthesis revision for peri-implant osteolysis and aseptic implant loosening (Hospital for Special Surgery, New York, New York, USA).22 Tissue was associated with reactions to orthopaedic implant wear debris. Specimens of soft tissue and bone were collected from regions of bone resorption during joint revision surgery. The specimens were fixed in freshly prepared 4% paraformadehyde, followed by demineralisation with 14% EDTA, processed and embedded in paraffin and 5 μm sections were prepared. Sections of bone and soft tissue taken at the time of primary hip implant were obtained from fixed, paraffin-embedded tissue. Paraffin-embedded sections (n=5 each) were deparaffinised and after rehydration, Netrin-1 immunostaining was performed following the protocol described above. TRAP staining was also performed as described above. All procedures were approved by the Institutional Review Boards of the Hospital for Special Surgery and NYU School of Medicine.
Measurement of inflammatory cytokines
See online supplementary text.
Statistical significance for differences between groups was determined by use of one-way analysis of variance and Bonferroni post hoc test or Student's t test, as appropriate. All statistics were calculated using GraphPad software (GraphPad, San Diego, California, USA).
Netrin-1 is highly expressed in human periarticular soft tissue removed during revision
We have recently reported that during osteoclast maturation, expression of Netrin-1 and its receptor Unc5b is increased in osteoclast precursors, and required for osteoclast differentiation.18 In parallel, activated tissue macrophages were reported to induce Netrin-1 expression in atherosclerotic plaques and diet-induced obesity15 ,17 where Netrin-1 contributes to inflammation.
To determine if these findings are relevant to orthopaedic wear particle-induced osteolysis in humans, we studied the expression of Netrin-1 in peri-implant tissues retrieved from patients undergoing revision surgery for peri-implant osteolysis and aseptic implant loosening. As seen in figure 1A there is little expression of Netrin-1 in cells populating the soft tissues removed at the time of primary joint replacement but high levels of Netrin-1 expression in the matrix and the cells populating the peri-implant tissues from patients undergoing revision surgery. TRAP staining revealed the presence of osteoclasts in patches on the bone surfaces in both specimens from initial joint replacement and revision implant loosening samples.
Netrin-1 and Unc5b are induced in the inflammatory infiltrate after wear particle exposure in mice
The results described above suggested that Netrin-1 might play a critical role in inflammatory osteolysis and contribute to the progression of inflammatory osteolysis. To test this hypothesis we employed a model of orthopaedic wear particle-induced osteolysis19 in which UHMWPE particles are placed under the periosteum over the calvaria. In sham-operated animals we observed negligible levels of Netrin-1-, Unc5b- and DCC-positive cells in the bone and soft tissue at the surgical site (figure 1B). In contrast, in the calvaria of mice exposed to UHMWPE particles there was a marked increase in Netrin-1 and Unc5b expression (118±3 Netrin-1-positive cells/high power field (hpf) and 60±3 Unc5b-positive cells/hpf vs 8±1 Netrin-1-positive cells/hpf and 20±2 Unc5b-positive cells/hpf for sham operated, p<0.001, n=5), primarily in the cells of the inflammatory infiltrate which was composed almost entirely of macrophages, as previously described.19 In addition, we observed a modest increase in DCC-positive cells (59±2 DCC-positive cells/hpf vs 45±2 positive cells/hpf for sham, p<0.001, n=5) in the inflammatory infiltrate of mice exposed to UHMWPE particles (figure 1B).
Netrin-1 and Unc5b are required for wear particle-induced osteolysis
Since Netrin-1 is highly expressed in inflammatory infiltrates in mice and in peri-implant joint tissue in humans, we determined if targeting Netrin-1 or its receptors blocked progression of or reversed inflammatory osteolysis in vivo. Ten microgram murine monoclonal anti-Netrin-1, anti-Unc5b or anti-DCC antibodies were injected intraperitoneally on a weekly basis in the murine model of inflammatory osteolysis. Weekly injections of anti-Netrin-1 or anti-Unc5b antibodies significantly reduced the area of particle-induced bone pitting in calvaria exposed to UHMWPE (46±4 and 49±3% of control bone pitting, respectively, p<0.001, n=5, figure 2A) but anti-DCC receptor antibody did not affect UHMWPE-induced pitting and resorption (80±7% bone pitting, p=ns vs control, figure 2A). Micro-CT also revealed a significant increase in bone volume (BV) and BV/total volume ratio (BV/TV) in both anti-Netrin-1- and anti-Unc5b-treated mice (figure 2A and table 1).
Micro-CT results were confirmed by histomorphometry (see online supplementary text).
We also observed a marked reduction in the inflammatory infiltrate in calvaria from mice treated with anti-Netrin-1 and anti-Unc5b antibodies, whereas anti-DCC antibody treatment did not affect the inflammatory infiltrate (83±2% reduction for anti-Netrin-1, 81±1% reduction for anti-Unc5b and 2±2% reduction for anti-DCC, p<0.001 and p=ns, respectively, n=5) (figure 2B). Anti-Netrin-1 and anti-Unc5b treatment, but not anti-DCC treatment, markedly reduced the number of TRAP-positive osteoclasts in affected bone (7±1 TRAP-positive cells/hpf for anti-Netrin-1, 4±1 TRAP-positive cells/hpf for anti-Unc5b and 14±1 TRAP-positive cells/hpf for anti-DCC compared with 12±2 TRAP-positive cells/hpf for control, p<0.001 and p=ns, respectively, n=5) (figure 2B).
Inflammatory cytokines also contribute to inflammatory osteolysis and anti-Netrin-1 and ani-Unc5b, but not anti-DCC, antibody treatment abrogated wear particle-induced increases in interleukin (IL)-1β and tumour necrosis factor (TNF)-α (see online supplementary text).
To further confirm that Netrin-1 and Unc5b are key factors in bone resorption in vivo, immunohistological examination was performed. When osteoclasts were identified by immunohistological staining for cathepsin K, similar results as those observed for TRAP staining were found (figure 3). Treatment with anti-Netrin-1 or anti-Unc5b antibody markedly reduced the number of cathepsin K–positive cells, whereas treatment with anti-DCC antibody did not significantly change the number of cathepsin K–positive cells (10±2 cells/hpf for anti-Netrin-1, 9±1 cells/hpf for anti-Unc5b and 19±3 cells/hpf for anti-DCC compared with 24±2 cells/hpf for control, p<0.001 and p=ns, respectively, n=5) (figure 3). In contrast, there were no significant changes in ALP-positive osteoblasts on bone forming surfaces in any antibody-treated group compared with control (8±1 cells/hpf for anti-Netrin-1, 10±1 cells/hpf for anti-Unc5b and 10±2 cells/hpf for anti-DCC compared with 9±2 cells/hpf for control, p=ns, n=5) (figure 4).
As previously described,18 the only cells that express Netrin-1 in bone are osteoclasts and osteoclast precursors. This is confirmed in vivo as Netrin-1 stain colocalised with CD68 immunostaining (see online supplementary figure S2). Both anti-Netrin-1 and anti-Unc5b treatments markedly reduced the number of CD68-positive macrophages and, consequently, the immunostaining for Netrin-1, whereas anti-DCC treatment did not change the number of CD68-positive macrophages when compared with control (6±1 cells/hpf for anti-Netrin-1, 8±2 cells/hpf for anti-Unc5b and 24±3 cells/hpf for anti-DCC compared with 29±4 cells/hpf for control, p<0.001 and p=ns, respectively, n=5) (see online supplementary figure S2).
Because of their effect on osteoclast differentiation we tested the effect of monoclonal anti-Netrin-1 and anti-Unc5b antibodies on bone density and found that treatment with these antibodies directly increased bone density in mice (see online supplementary text and table 2).
Here we report that Netrin-1 is highly expressed by macrophages at sites of wear particle-induced osteolysis in the inflamed peri-implant soft tissue from patients undergoing implant revision and in macrophages and osteoclasts in a murine model of wear particle-induced bone destruction. Moreover, in vivo blockade of Netrin-1 and its receptor Unc5b by murine monoclonal antibodies prevents wear particle-induced bone destruction. These results indicate that Netrin-1 plays a critical role in inflammatory osteolysis.
Multiple studies have shown that orthopaedic wear products are potent inducers of proinflammatory mediators, including IL-1β and TNFα, and other soluble stimuli for osteoclast formation and bone resorption.19 Here we demonstrated that treatment with monoclonal antibodies in both ex vivo cultures of calvaria exposed to wear particles, and in the RAW264.7 cell line, diminished both IL-1β and TNFα secretion. Macrophages represent a major component of the cellular infiltrate associated with wear particle-induced inflammation. Based on the findings in atherosclerotic plaques it is reasonable to hypothesise that Netrin-1 may contribute to the recruitment and retention of macrophages at the site of wear particle deposition and that signalling by Netrin-1 may further contribute to the osteolytic process via enhancement of osteoclast differentiation and activity. Prevention of wear particle-induced bone destruction by blockade of Netrin-1 and Unc5b by murine monoclonal antibodies indicates that Netrin-1 is acting in both an autocrine and paracrine fashion to promote bone destruction, as described in vitro.18 The demonstration that blockade of DCC had no effect on bone destruction provides strong evidence that the effects of the anti-Netrin-1 and anti-Unc5b antibodies on osteolysis are specific, since DCC is also expressed by osteoclasts and macrophages.18
Because of its chemorepulsant activities, administration of Netrin-1 has been described as an anti-inflammatory treatment in a number of settings, including inflammatory bowel disease, pancreatitis, peritonitis and pulmonary inflammation.11 ,12 ,14 ,23–29 Neuronal guidance proteins present in peripheral tissues contribute to the local control of leucocyte migration and inflammation.30 ,31 Netrin-1 administration was found to direct leucocyte traffic during acute inflammation in peripheral organs,23 ,27 to reduce local injury and inflammatory responses,27 ,31 and to stimulate resolution mechanisms and production of resolvins.32 In accordance, Netrin-1 administration regulates inflammation and infiltration of monocytes and ameliorates ischaemia reperfusion-induced kidney injury,14 ,33 by inducing overexpression of macrophage a M2-like phenotype and suppression of expression of M1 markers.34 Netrin-1 also regulates the inflammatory response of neutrophils and macrophages, and suppresses ischaemic acute kidney injury by inhibiting COX-2-mediated prostaglandin E2 and thromboxane A2 production.33 Netrin-1 is also generated within the intestinal epithelium to dampen neutrophil recruitment in acute colitis and may have a beneficial function in chronic models of intestinal inflammation.35
However, the anti-inflammatory effects of Netrin-1 do not accord with the demonstration that Netrin-1 plays a critical role in promoting atherogenesis by inhibiting macrophage emigration from plaque,15 ,17 and other findings that identify Netrin-1 as a macrophage retention signal in adipose tissue of obese mice that promotes chronic inflammation and insulin resistance. Recently, the molecular mechanism linking plaque hypoxia and macrophage infiltration has been demonstrated in mouse and human plaques.17 It involves hypoxia-inducible factor 1α–mediated upregulation of Netrin-1 and Unc5b in human and mouse atherosclerotic plaques and subsequent macrophage chemostasis and protection from apoptosis.17
Nonetheless, the observation that there is marked overexpression of Netrin-1 in cells at sites of inflammatory bone destruction both in mice and in humans suggests that Netrin-1 plays an important role in inflammatory osteolysis. Our in vitro data suggest that this may be in part attributable to the effects of Netrin-1 on enhancing osteoclast differentiation, attachment and adhesion to the bone surfaces, which are known to play a role in bone resorption.36 Moreover, histomorphometry analysis in long bones treated with monoclonal antibodies for 2 weeks suggested a trend to increased trabecular bone formation with changes in BV/TV and bone mineral density and a decrease in osteoclast number. Although the change in bone formation detected by the calcein/declomycin double-labelling technique was not significant, there was a tendency to increased bone formation. In some prior studies Netrin-1 was reported to mediate its effects on myeloid cells and inflammation by a mechanism that involves adenosine A2B receptors.13 ,23 ,27 ,37 ,38 Nonetheless, it is unlikely that adenosine A2B receptors are involved in the effects of Netrin-1 on osteoclast differentiation and function since He et al39 recently reported that stimulation of adenosine A2B receptors inhibits osteoclast differentiation. Thus, Netrin-1 most likely plays a direct role in inflammatory bone destruction by increasing or permitting osteoclast differentiation and function.
In conclusion, our results indicate that Netrin-1 is an autocrine factor produced by osteoclast precursors that enhances osteoclast differentiation and function. The demonstration of high levels of Netrin-1 expression in cells at sites of inflammatory bone resorption furthermore suggests a contributory role in the enhanced bone resorption associated with inflammatory processes. Thus, Netrin-1 may be a novel therapeutic target for the reduction of osteoclast-mediated bone resorption and other forms of inflammatory bone destruction.
We acknowledge Dr Giogio Perino for his help with human histological sections. We thank L Mazinski and S Ahmed from the NYU Histopathology Core (supported by Cancer Center Support Grant P30CA016087 at the Laura and Isaac Perlmutter Cancer Center) for technical assistance and Drs N Dyment and D Rowe at the University of Connecticut Health Center for advice on tissue preparation and sectioning.
Handling editor Tore K Kvien
Contributors BNC and KJM have designed the experiments, wrote and revised the manuscript. AM designed the experiments and has been the primary person responsible for carrying out all experimental procedures and writing the manuscript. TW helped on surgery and animal treatments and revised the manuscript. BR performed the immunohistochemistries and revised the manuscript. PEP and SRG carried out human subject experiments and revised the manuscript. ZD and CL performed undecalcified bone preparation, sectioning, staining and scanning.
Funding This work was supported by grants from the National Institutes of Health (AR56672, AR54897, AR046121, RC1HL100815, K99 HL125667), the NYU-HHC Clinical and Translational Science Institute (UL1TR000038), the NYUCI Center Support Grant, 9NIH/NCI 5 P30CA16087-310 and grants from Celgene and Gilead Pharmaceuticals.
Competing interests AM and BNC have filed a patent on use of adenosine A2AR agonists to prevent prosthesis loosening (pending). AM, BNC, BR and KJM have filled a patent on the use of antibodies against Netrin-1 for the treatment of bone diseases. TW, PEP, SRG, ZD and CL do not have any disclosures. BNC holds patents numbers 5 932 558; 6 020 321; 6 555 545; 7 795 427; adenosine A1R and A2BR antagonists to treat fatty liver (pending); adenosine A2AR agonists to prevent prosthesis loosening (pending). BNC is a consultant for Bristol-Myers Squibb, AstraZeneca, Novartis, CanFite Biopharmaceuticals, Cypress Laboratories, Regeneron (Westat, DSMB), Endocyte, Protalex, Allos, Savient, Gismo Therapeutics, Antares Pharmaceutical, Medivector, King Pharmaceutical, Celizome, Tap Pharmaceuticals, Prometheus Laboratories, Sepracor, Amgen, Combinatorx, Kyowa Hakka, Hoffman-LaRoche and Avidimer Therapeutics. BNC has stock in CanFite Biopharmaceuticlas.
Patient consent Obtained.
Ethics approval All procedures abided by the normal human studies regulations by the Institutional Review Boards of the Hospital for Special Surgery.
Provenance and peer review Not commissioned; externally peer reviewed.
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.