Orthopedic Wear Debris Mediated Inflammatory Osteolysis Is Mediated in Part by NALP3 Inflammasome Activation

Transcription

1 Orthopedic Wear Debris Mediated Inflammatory Osteolysis Is Mediated in Part by NALP3 Inflammasome Activation Lyndsey Burton, 1 Daniel Paget, 1 Nikolaus B. Binder, 1 Krista Bohnert, 1 Bryan J. Nestor, 1 Thomas P. Sculco, 1 Laura Santambrogio, 2 F. Patrick Ross, 1 Steven R. Goldring, 1 P. Edward Purdue 1 1 Hospital for Special Surgery, 535 East 70th Street, New York, New York 10021, 2 Department of Pathology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, New York, New York Received 27 March 2012; accepted 25 June 2012 Published online 29 August 2012 in Wiley Online Library (wileyonlinelibrary.com). DOI /jor ABSTRACT: Activation of myeloid cells by orthopedic particulate debris is a key event in the pathogenesis of periprosthetic osteolysis and implant loosening after total joint replacement (TJR). Several lines of evidence implicate NACHT, LRR, and PYD domains-containing protein 3 (NALP3) inflammasome-mediated production of interleukin 1 beta (IL-1b) in the pathogenesis of clinical disorders ascribable to foreign particulate materials, including asbestos, silica, and urate crystals. Recent reports indicate that orthopedic polymer products and metallic particulates and ions may activate the same pathway. Here, we investigated the contribution of the NALP3 inflammasome to the pathogenesis of peri-implant osteolysis. Pharmaceutical and genetic perturbations of caspase-1 and inflammasome components were used to assess the role of the NALP3 inflammasome in IL-1b production and osteoclast formation by human monocytes and mouse macrophages in response to polymethylmethacrylate (PMMA) particle phagocytosis. The role of caspase-1 in a mouse calvarial model of particle-mediated osteolysis was assessed using mct. Phagocytosis of PMMA particles induces caspase-1 dependent release of IL-1b from human monocytes and mouse macrophages. Importantly, using macrophages from mice deficient in components of the NALP3 inflammasome, we show PMMA-induced IL-1b production is strictly dependent on these components. Mice lacking caspase-1, the sole effector of the NALP3 inflammasome, show reduced orthopedic wear particle-induced calvarial osteolysis compared to wild-type controls. Absence of NALP3 inflammasome components fails to alter osteoclast formation in vitro. Our findings identify the NALP3 inflammasome as a critical mediator of orthopedic wear-induced osteolysis and as a viable therapeutic target for the treatment of periprosthetic osteolysis. ß 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 31:73 80, 2013 Keywords: NALP3 inflammasome; periprosthetic osteolysis; interleukin-1b; caspase-1 Aseptic loosening and the accompanying periprosthetic osteolysis represent the major long-term complication of total joint replacement (TJR) surgery. Incidence rates have generally been reported to lie between 5% and 20%, but can be as high as 40% in some studies. 1,2 These rates are predicted to substantially increase over time, a factor of particular importance in light of the fact that 35 45% of all TJR patients are now under 65 years of age. Currently over 600,000 total hip and knee replacements are performed annually in the United States, and this number is projected to rise by about sevenfold by Thus, periprosthetic osteolysis represents an important and growing public health problem. 3 The initiating cellular event underlying periprosthetic osteolysis is a granulomatous inflammatory response of macrophages to particles of orthopedic wear debris. Higher wear rates are seen in patients with osteolysis compared with controls, 4,5 and large numbers of wear particles are associated with the periprosthetic interfacial membrane removed during revision surgery. 6 8 The interfacial membrane of patients with osteolysis shows extensive macrophage infiltration, 9 and the presence of particles within these cells is indicative of an active phagocytic process. 6 The proinflammatory consequences of interaction of macrophages with wear debris are proposed to lead to Correspondence to: P. Edward Purdue (T: ; F: ; purduee@hss.edu) ß 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. increased recruitment and differentiation of osteoclast pre-cursors. In vitro and in vivo experimental approaches underscore the importance of particle mediated proinflammatory activation of myeloid cells in osteolysis. Macrophage lineage cells and cell lines exhibit the capacity to phagocytose wear particles, and this event is accompanied by the induction of proinflammatory mediators such as prostaglandin E 2 (PGE 2 ), the pleitropic cytokine interleukin 6 (IL-6) and, importantly, tumor necrosis factor alpha (TNF-a) and IL-1b. 10,12,14 18 The specific nature of this response depends on numerous parameters including the composition, size, 11,22 shape, 23 volume, and surface area 21 of the particulate debris. In vivo, a murine model of calvarial osteolysis has been used to confirm a critical role for TNF signaling in inflammatory osteolysis 15,24 and an intra-medullary implant model implicated IL-1 in the same process. 25 In addition to periprosthetic osteolysis, several other human disorders are triggered by particle-induced macrophage activation. Examples include asbestosis, silicosis, and gout, which are initiated by interaction of macrophages with particulate asbestos, silica, or urate crystals, respectively. These materials either undergo phagocytosis, or, in the case of larger particles, attach to the macrophage surface, leading to so-called frustrated phagocytosis In each case the outcome is assembly of a protein complex called the NALP3 inflammasome, comprising NALP3, apoptotic speck protein containing a C-terminal caspase recruitment domain (ASC), and the nascent protease caspase-1, 73

2 74 BURTON ET AL. which is activated and induces processing of pro-il-1b and secretion of mature, cleaved IL-1b. The extracellular cytokine initiates an autocrine pro-inflammatory cascade, leading to secretion of additional pro-inflammatory products, including TNF-a. Targeted ablation of components of the NALP3 inflammasome prevents development of silicosis and asbestosis in mouse models, further implicating this pathway as a major contributor to particle-induced disease pathogenesis. Although several molecularly distinct inflammasome complexes have been identified and shown to be involved in sensing danger signals in myeloid cells, the NALP3 inflammasome has thus far proved to be of especial importance in responses to such non-biodegradable phagocytic challenges. 31 Given that orthopedic wear particles undergo phagocytosis by macrophages or induce frustrated phagocytosis, a reasonable hypothesis is that NALP3 inflammasome activation plays an important role in the pathogenesis of periprosthetic osteolysis. Indeed, we reported earlier that exposure of human monocytes to Simplex bone cement particles activates processing of pro-il-1b and secretion of the active cytokine. 32 Importantly, recent reports from two separate research groups have shown that the NALP3 inflammasome can mediate myeloid cell responses to orthopedically relevant particulates. 33,34 Caicedo et al. 33 initially demonstrated that depletion of NALP3 or ASC diminished particle induced release of IL-1b by THP1 monocytic cells, and St. Pierre et al. 34 demonstrated genetically that macrophage release of IL-1b in response to titanium particles was dependent on NALP3, ASC, and caspase-1. In the current work, we extend these studies using in vitro and in vivo osteolysis models, including mice unable to mount a NALP3 response. Our results demonstrate that inflammasome-mediated signaling contributes to both the proinflammatory effects of orthopedic wear particles in vivo and subsequent resultant osteolysis. The identification of inflammasome signaling as a novel mediator of myeloid responses to orthopedic particulate materials has important implications for understanding the cellular mechanisms underlying osteolysis, and for the development of therapeutic strategies for treating this disorder. MATERIALS AND METHODS Cell Culture Human monocytes were isolated as previously described from de-identified normal blood donors, and incubated at cells/ml in a-mem supplemented with 10% FBS, 1% penicillin streptomycin (Invitrogen, Carlsbad, CA) and 10 ng/ml human M-CSF (Peprotech, Rocky Hill, NJ) in 12- well tissue culture dishes (1 ml/well). 18 Bone marrow macrophages (BMM) from C57BL6, NALP3 /, ASC /, and Caspase-1 / mice were isolated by culture of bone marrow cells for 3 days on petri dishes in a-mem medium supplemented with 10% FBS, 1% penicillin streptomycin (Invitrogen) and 10% (equivalent to 140 ng/ml M-CSF) conditioned medium from the M-CSF overproducing cell line CMG. 35 Adherent cells were collected by trypsinization and then cultured in multi-well tissue culture plastic dishes. For inflammasome signaling studies, cells were cultured at cells/ml in a-mem supplemented with 10% FBS, 1% penicillin streptomycin (Invitrogen) and 2% (equivalent to 28 ng/ml M-CSF) CMG medium in 12-well dishes (1 ml per well). For osteoclastogenesis, cells were cultured at cells/ml in the same medium with GST-RANKL added to 50 ng/ml, in 96-well dishes (200 ml per well). Where appropriate, slices of bovine cortical bone were inserted into the wells prior to addition of cells. Medium was changed after 3 days. Osteoclastogenesis was assessed on day 4 by TRAP staining (Sigma Aldrich, St. Louis, MO, kit 386A) followed by counting of TRAP-positive, multinuclear (3þ nuclei) cells, and resorption of bone slices was assessed by removal of cells with 2 M NaOH, and incubation for 30 min with 20 mg/ml peroxidase-conjugated wheat germ agglutinin (Sigma Aldrich) followed by 30 min with 3,3 0 -diaminobenzidine substrate solution (Vector Laboratories, Burlingame, CA). Inflammasome Activation Human monocytes ( cells/well) were cultured for 6 h in the presence of LPS (80 pg/ml), PMMA (500 mg), or both LPS and PMMA. Time course studies revealed that the 6-h time point was optimal for detection of production of IL-1b in the conditioned medium. Mouse BMM ( cells/ml) were primed for 18 h with 100 ng/ml E. coli LPS (Sigma Aldrich), following which PMMA particles (500 mg) or ATP (5 mm) was added for an additional 6 h. PMMA particles (average diameter 6.06 mm) were obtained from Polysciences Inc. (catalog # 19130, Warrington, PA). PMMA particles were sterilized by treatment with 70% Ethanol for 48 h and suspended in a-mem medium. 18 Only endotoxin-free particles as determined by Limulus assay (E-toxate, Sigma Aldrich) were used in this study. Where appropriate, LPS primed cells were pre-treated for 30 min with zvad (40 mm) or cytochalasin D (0.2 mg/ml) prior to addition of PMMA or ATP. IL-1b, IL-6 and TNF-a in conditioned media were determined by ELISA (OptEIA kits, BD Bioscience, Franklin Lakes, NJ). Extraction and immunoblotting of cellular protein (0.05 mg/ml anti-human IL1-b (ab18642, abcam, Cambridge, MA)) and RNA extraction were as described. 18 RT- PCR (28 cycles) was performed using PCR Platinum Supermix (Invitrogen) and primers for murine HPRT (AGCTACTGTAATGATCAGTCAACG/AGAGGTCCTTTTCAC CAGCA) and IL-1b (GACCTTCCAGGATGAGGACA/AGG CCACAGGTATTTTGTCG). Calvarial Osteolysis Model All procedures were approved by the Hospital for Special Surgery Institutional Animal Care and Usage Committee (IACUC). Caspase-1 deficient (NOD.129S2(B6)-Casp1 tm/sesh / LtJ) and wild-type control (NOD/LtJ) mice were obtained from Jackson Laboratories (Bar Harbor, ME). For introduction of wear particles, a 1 cm midline sagittal incision was made above the calvarium using a #15 blade scalpel, and a 1cm 1 cm area of the calvarium exposed. The periosteum was scraped laterally from the sagittal suture off the surface of both parietal bones using the #15 blade scalpel particles in 25 ml sterile phosphate-buffered saline was implanted evenly over the bone surface using ml Graduated Filter Tips and a manual pipettor. The skin edges were closed with 4-0 Ethilon nylon sutures (Ethicon, Somerville, NJ). Control mice received sham surgery and 25 ml

3 NALP3 INFLAMMASOME IN PERIPROSTHETIC OSTEOLYSIS 75 sterile phosphate-buffered saline (no particles). Consistent with an earlier description of this model, 15 we found that resorption was maximal at 1 week following particle implantation. One week following surgery, calvaria were retrieved and fixed in 10% formalin for h and then scanned in 70% ethanol in a Scanco mct35 (Scanco Medical, Brüttisellen, Switzerland) mct system with a resolution of 15 mm at regular contrast conditions (55 kv p, 145 ma, angular rotation step). 3D reconstruction of the scanned volumes and bone background segmentation was performed using the system reconstruction software and pit area over total area ratio was calculated using NIH imagej software. Two parietal bone rectangular zones extending 4 mm from the midline were defined as regions of interest for analysis, for which the bone thickness and bone surface/bone volume ratio were calculated. RESULTS PMMA Particles Induce IL-1b Secretion by Human Monocytes As a starting point for these studies we asked if release of IL-1b by primary human monocytes was stimulated following exposure to orthopedically relevant polymethylmethacrylate (PMMA) wear particles. We compared the effects of PMMA particles to the TLR4 ligand LPS that potently activates inflammatory gene expression in monocytes but does not activate inflammasomes or proteolytic processing of pro-il-1b precursor protein. As expected, and in agreement with previous results and other studies, 32,36 LPS alone did not induce IL-1b secretion despite driving synthesis and intracellular accumulation of pro-il-1b, but did induce robust secretion of TNF-a and IL-6, cytokines whose production and secretion are not dependent on inflammasome activation (Fig. 1). In contrast to LPS, PMMA particles alone resulted in small but detectable increases in extracellular levels of IL-1b (Fig. 1), Figure 1. Cytokine production by human monocytes after challenge with LPS and PMMA. Monocytes were cultured for 6 h with LPS (80 pg/ml), PMMA (500 mg), or LPS and PMMA. Conditioned medium was analyzed for cytokine production by ELISA (n ¼ 4). p < 0.05 compared to LPS alone. Cell lysates were analyzed for pro-il-1b expression by immunoblotting. suggesting that PMMA particles can activate inflammasomes to generate mature IL-1b. Strikingly, LPSco-treatment markedly increased IL-1b secretion in response to PMMA phagocytosis in human monocytes, which is consistent with the established paradigm that TLR engagement primes cells by inducing expression of IL-1b mrna and intracellular pro-il-1b pre-cursor protein, which is then proteolytically processed and mature IL-1b secreted in response to factors that activate inflammasomes and associated caspase-1 activity. 31 These results show that PMMA particles can induce IL-1b secretion in human monocytes and suggest that PMMA particles activate inflammasome-mediated generation of mature IL-1b. Phagocytosis of PMMA Particles Induces Macrophage Production of IL-1b We wished to use genetic approaches to investigate the mechanism of the results shown for the human system and to this end turned to murine models. We therefore prepared and LPS-primed wild-type mouse bone marrow derived macrophages (BMMs), and evaluated the effects of PMMA phagocytosis on cytokine release by these cells. As seen for human monocytes (Fig. 1), LPS alone was insufficient to induce production of detectable levels of secreted IL-1b, although it did robustly induce expression of IL-1b mrna (Fig. 3c). However, addition of phagocytosable particles of PMMA to LPS-primed cells for 6 h substantially increased production of secreted IL-1b (Fig. 2). To investigate whether the PMMA-induced IL-1b secretion was dependent upon activity of caspase-1, the effector protease for inflammasome mediated IL-1b production, LPS-primed cells were pre-treated with the caspase inhibitor zvad prior to particle addition. This resulted in significant inhibition of PMMA-induced IL-1b production (Fig. 2). In addition, PMMA particle induction of IL-1b production was significantly inhibited by pretreatment of LPS-primed cells with cytochalasin D, an inhibitor of actin polymerization and cytoskeletal reorganization, suggesting that phagocytosis of PMMA is a pre-requisite for the ability of these particles to induce IL-1b secretion (Fig. 2). As a positive control we treated cells with the known NALP3 inflammasome activator ATP, which functions via a phagocytosisindependent mechanism through interaction with the cell surface purinergic receptor P2X7, 37 and found that the level of caspase-1 dependent IL-1b secretion from LPS-primed cells was similar to that induced by PMMA phagocytosis (Fig. 2). Consistent with expectations, ATP induction of IL-1b release was not affected by inhibition of phagocytosis with cytochalasin D. Together, the results shown in Figure 2 demonstrate that phagocytosis of orthopedically relevant wear particles initiates caspase-dependent IL-1b production by LPS-primed murine BMMs to a comparable extent to that seen with the known NALP3 inflammasome activator ATP.

4 76 BURTON ET AL. Figure 2. IL-1b production by mouse macrophages after particle phagocytosis. BMM were primed overnight with LPS (100 ng/ ml) then incubated with PMMA (500 mg) or ATP (5 mm). After 6 h, conditioned medium was isolated and analyzed for IL1-b by ELISA (n ¼ 4). Where indicated, LPS-primed cells were pretreated for 30 min with zvad (40 mm) or cytochalasin D (0.2 mg/ ml), p < 0.05 versus no treatment. p < 0.05 versus PMMA or ATP alone. PMMA-Induced IL-1b Production Requires Components of the NALP3 Inflammasome To directly address the requirement for the NALP3 inflammasome pathway in regulation of IL-b production by PMMA particle phagocytosis, we compared responses to PMMA of wild-type BMMs with those of BMMs prepared from mice strains lacking NALP3, ASC, or caspase-1, which are the essential components of the NALP3 inflammasome complex. In common with wild-type BMMs, LPS-primed cells from each of the mutant strains showed no release of IL-1b (Fig. 3a), but did produce normal levels of secreted IL- 6 and TNF-a (Fig. 3b and data not shown), indicating that, in line with expectations, LPS signaling is retained in the absence of NALP3 inflammasome components. In contrast, absence of any of the NALP3 inflammasome components significantly reduced the ability of PMMA to induce IL-1b production from LPSprimed cells, without affecting IL-6 release (Fig. 3a,b). To rule out defective priming of IL-1b expression in the inflammasome-deficient cells we measured IL-1b mrna expression levels, which revealed, as expected, that wild-type and mutant cells all showed robust induction of IL-1b expression in response to TLR priming (Fig. 3c). Together, these results demonstrate that the production of IL-1b by murine BMMs in response to PMMA phagocytosis is strictly dependent upon the presence of the components of the NALP3 inflammasome complex, strongly implicating the involvement of this complex in myeloid cell responses to orthopedic wear particles. Caspase-1 Deficiency Reduces Wear Particle-Induced Calvarial Osteolysis In Vivo The preceding results provide evidence for involvement of the NALP3 inflammasome in IL-1b release triggered by myeloid cell phagocytosis of wear Figure 3. IL-1b production by mouse macrophages after challenge with particles is dependent upon components of the NALP3 inflammasome. BMM from mice deficient in NALP3, ASC, or caspase-1 and wild type strains were primed overnight with LPS (100 ng/ml) then incubated with or without PMMA (500 mg/ml). After 6 h, conditioned medium was isolated and analyzed for IL1-b (a) and IL-6 (b) by ELISA (n ¼ 3). p < versus equivalent treatment of wild-type cells. RNA was extracted from the cells and analyzed for IL-1b and HPRT expression by RT-PCR (c). particles in vitro. To investigate the involvement of the NALP3 inflammasome in particle-mediated osteolysis in vivo, we next employed a mouse calvarial implantation model that previously has been validated as recapitulating the osteolytic process observed with orthopedic wear debris in human subjects, and widely used to identify and characterize key signaling mediators of osteolysis. 4,15 For these studies, particles of PMMA or Simplex bone cement were implanted over

5 NALP3 INFLAMMASOME IN PERIPROSTHETIC OSTEOLYSIS 77 the calvarial surface and 1 week later calvaria were retrieved and analyzed for bone resorption by mct. To evaluate the potential involvement of inflammasome signaling in this model we compared responses in mice lacking caspase-1, the effector molecule of the inflammasome, to those seen in wild-type mice. As illustrated in Figure 4a, sham operated animals of both strains in which PBS lacking wear particles was implanted over the calvarium showed no bone resorption. In contrast, both PMMA and Simplex particles induced distinct regions of bone resorption in the calvaria of wild-type mice. Deficiency of caspase-1 markedly reduced osteolysis observed with both types of wear particles, although bone destruction was not completely prevented (Fig. 4a). Quantitation of the percentage decrease in bone thickness and increase in bone surface over total volume (BS/TV) verified that the bone loss induced by orthopedic particulate debris was significantly reduced in the absence of caspase-1 (Fig. 4b). These results show that caspase-1, the inflammasome effector caspase previously shown to be required for particle induced IL-1b production by myeloid cells in vitro, is also required for optimal osteolytic responses to multiple types of orthopedic wear debris in vivo. This represents the first evidence that inflammasome mediated responses to orthopedic wear particles can impact pathological bone resorption, and therefore are of significant importance in validating inflammasome signaling as a valid target for treatment of periprosthetic osteolysis. NALP3 Inflammasome Components are Dispensable for Osteoclastogenesis Induced by M-CSF and RANKL The attenuated osteolytic response to particle implantation in mice lacking caspase-1 suggests that the diminished IL-1b response in these animals may protect against inflammatory recruitment and activation of osteoclasts to the bone surface. However, an alternative explanation could be that one or more members of the inflammasome complex is required in a cell autonomous manner for generation and/or activation of osteoclasts. To evaluate the possible involvement of inflammasome components in osteoclast differentiation and activation, macrophages from wild-type controls and mouse strains lacking NALP3, ASC, or caspase-1 were isolated and treated with M-CSF and RANKL to induce osteoclast formation. As shown in Figure 5a,b, absence of these individual NALP3 components did not significantly affect formation of osteoclasts compared to wild type cells. Furthermore, osteoclasts from each of the various strains were able to generate resorption pits when cultured on bone slices, demonstrating that osteoclast activation is not affected by NALP3 inflammasome deficiency (Fig. 5c). These results suggest that absence of NALP3 inflammasome components has no detectable cell autonomous effect on osteoclastogenesis, suggesting that the reduced in vivo osteolytic response of caspase-1 deficient mice represents an indirect consequence of reduced inflammation, not a direct effect on osteoclast formation. Figure 4. Particle mediated calvarial osteolysis is partially dependent upon caspase-1. Simplex or PMMA was implanted over calvaria of caspase-1 knockout and wild type mice. One week later, calvaria were retrieved and analyzed by mct. (a) Representative images and (b) Quantitation of the percentage increase in bone surface/total volume and decrease in bone thickness relative to sham controls (n ¼ 5). p < 0.05 versus PBS in mice of the same strain. ns ¼ not significant versus PBS. [Color figure can be seen in the online version of this article, available at wileyonlinelibrary.com/journal/jor] DISCUSSION Although there is considerable evidence that inflammatory responses of myeloid cells to orthopedic implant particulate debris represent an important initiating event in periprosthetic osteolysis, the specific

6 78 BURTON ET AL. Figure 5. NALP3 inflammasome components are dispensable for osteoclastogenesis. MCSF/RANKL mediated osteoclastogenesis of BMM was assessed by TRAP staining (a), quantitation of TRAP-positive multinuclear cells (b) and staining of resorption pits (c). signaling pathways involved in these events are poorly understood. Elucidating the underlying molecular pathogenesis will be key in developing therapeutic strategies for treatment of osteolysis, for which there currently are no approved medical therapies. In this paper we provide in vitro and in vivo evidence that activation of the NALP3 inflammasome plays an important role in bone loss induced by orthopedic implant particulate debris. NALP3 inflammasome activation is the major mechanism controlling processing and release of IL-1b from myeloid cells. 28 IL-1b is synthesized within cells as an inactive pre-cursor that is cleaved and secreted only following inflammasome assembly and activation of caspase-1. A number of mediators of inflammasome activation relevant to innate immunity have now been described, including adjuvants, silica particles, asbestos crystals, and urate, the causative agent in gout. 26,27,29,38 Two recent papers from Caicedo et al. 33 and St. Pierre et al. 34 have demonstrated that metal ions and particles derived from orthopedic implant materials can also activate the NALP3 inflammasome, providing the first evidence that this pathway may be of relevance to wear particle-mediated inflammatory osteolysis. These papers showed for the first time that release of IL-1b in response to orthopedic wear particles is mediated by components of the NALP3 inflammasome. Although the intracellular pathways by which particulates induce inflammasome activation have yet to be fully resolved, the accumulated evidence implicates endosomal destabilization following phagocytosis as a critical step. 29,39 Thus, we showed that orthopedic wear particles initiate endosomal/lysosomal damage and release of lysosomal enzymes in myeloid cells 32 a finding buttressed by the recent report that titanium particle induced inflammasome activation in mouse macrophages is dependent upon lysosomal cathepsin B activity. 34 A role for NADPH/ROS in macrophage responses to orthopedic metals has also been proposed, 33 based on the observation that the NADPH inhibitor diphenylene iodinium blocked metal induced release of IL-1b from THP1 monocytes. In the present studies, which were performed using human and mouse myeloid cells, we utilized a range of in vitro pharmacological and genetic approaches to demonstrate an essential role for the NALP3 inflammasome in production of IL-1b in response to orthopedically relevant phagocytosable PMMA wear particulates. First, pre-incubation of murine macrophages with the caspase inhibitor zvad prevented particle induction of IL-1b release. Second, macrophages from mouse strains lacking NALP3 components (NALP3, ASC, or caspase-1) were deficient in IL-1b release in response to orthopedic wear particles. Particle uptake was required for inflammasome activation, since inhibition of particle phagocytosis with cytochalasin D prevented IL-1b release. As expected, IL- 1b release in response to ATP, which triggers NALP3 inflammasome activation through a non-phagocytic mechanism, was inhibited by zvad but unaffected by cytochalasin D. In common with other inflammasome activating particulates, production of substantial amounts of IL- 1b from both human and murine myeloid cells in response to PMMA stimulation in vitro requires prior priming with a pro-inflammatory stimulus, such as the TLR4 agonist LPS, to induce cellular accumulation of pro-il-1b. LPS mediated TLR4 activation directly induces IL-1b mrna, and also induces production of cytokines such as TNF, which may further induce

7 NALP3 INFLAMMASOME IN PERIPROSTHETIC OSTEOLYSIS 79 pro-il-1b production via autocrine and paracrine signaling. Although LPS is sufficient to induce limited IL-1b secretion by monocytes under certain culture conditions, 40 robust levels of secretion require an additional inflammasome-activating signal. It has been suggested that the requirement (at least in vitro) for two stimuli for production of inflammasome regulated cytokines may provide some measure of protection against inappropriate activation of this pathway by ubiquitous endogenous molecules. 36 Nevertheless, in the absence of a distinct TLR agonist, orthopedic wear particles do induce IL-1b mrna and protein production, albeit at lower levels than in primed cells (Fig. 1 and data not shown). In addition, we and others 32,41 have shown that wear particles can stimulate TLR signaling. Taken together, these results suggest that wear particles alone can mediate both pro-il-1b synthesis and its NALP3 inflammasome mediated release, consistent with the ability of these particles to induce aseptic inflammatory osteolysis in vivo. Our in vitro observations were validated and extended in vivo using a mouse calvarial model of wear particle-induced osteolysis, in which bone loss in response to PMMA and Simplex bone cement particles was significantly diminished in caspase-1 deficient mice. In vitro osteoclastogenesis of bone marrow derived macrophages was not affected by the absence of caspase-1 (or NALP3 or ASC), suggesting that the reduced osteolysis in caspase-1 deficient mice reflects diminished inflammation rather than a direct cellautonomous effect on osteoclast formation. These findings provide the first direct evidence that inflammasome signaling in response to orthopedic wear phagocytosis can modulate subsequent bone destruction, and hence are of critical importance in identification of this pathway as a critical mediator of periimplant bone destruction. Although caspase-1 deficiency provided significant protection from orthopedic wear particle induced calvarial osteolysis, it did not completely prevent bone loss, implicating the existence of additional inflammasome-independent pathways for wear particle induced bone loss. Consistent with this finding, TNF-a signaling has been shown to mediate wear particle induced osteolysis in murine models, 15,24 and TNF-a expression is significantly induced by phagocytosis of orthopedic wear particles. 18 The demonstrated dependence upon caspase-1 for maximal osteolytic responses to wear particles is also in agreement with the observation that IL-1 acts as a mediator of particle-induced inflammation in an intra-medullary osteolysis model. 25 In summary, we provide the first genetic-based evidence that wear particle induction of the inflammasome, which is a central mediator of many components of the innate immune system, is a major mechanism in implant-associated bone loss. Key effectors downstream of caspase-1 activation are the pro-inflammatory cytokines IL-1b and TNF-a. At least in the murine system IL-1 signaling and TNF-a play collaborative roles in amplifying RANKL-induced osteoclastogenesis. 42 Thus, future studies will be necessary to address whether these two cytokines interact with and modulate each other, and whether blocking both of these pathways simultaneously will be required to prevent osteolysis. ACKNOWLEDGMENTS We thank Dr. Fayyaz Sutterwala, University of Iowa, for providing bones from wild-type and inflammasome deficient mice. REFERENCES 1. Harris W Conquest of a Worldwide Human Disease: particle-induced periprosthetic osteolysis. Clin Orthop Relat Res 429: Purdue PE, Koulouvaris P, Potter HG, et al The cellular and molecular biology of periprosthetic osteolysis. Clin Orthop Relat Res 454: Paxton E, Inacio M, Slipchenko T, et al The Kaiser Permanente National Total Joint Replacement Registry. Perm J 112: Wilkinson JM, Hamer AJ, Stockley I, et al Polyethylene wear rate and osteolysis: critical threshold versus continuous dose-response relationship. J Orthop Res 23: Dumbleton JH, Manley MT, Edidin AA A literature review of the association between wear rate and osteolysis in total hip arthroplasty. J Arthroplasty 17: Schmalzried TP, Jasty M, Harris WH Periprosthetic bone loss in total hip arthroplasty. Polyethylene wear debris and the concept of the effective joint space. J Bone Joint Surg Am 74: Hirakawa K, Bauer TW, Stulberg BN, et al Comparison and quantitation of wear debris of failed total hip and total knee arthroplasty. J Biomed Mater Res 31: Margevicius KJ, Bauer TW, McMahon JT, et al Isolation and characterization of debris in membranes around total joint prostheses. J Bone Joint Surg Am 76: Willert HG Reactions of the articular capsule to wear products of artificial joint prostheses. J Biomed Mater Res 11: Blaine TA, Rosier RN, Puzas JE, et al Increased levels of tumor necrosis factor-alpha and interleukin-6 protein and messenger RNA in human peripheral blood monocytes due to titanium particles. J Bone Joint Surg Am 78: Yagil-Kelmer E, Kazmier P, Rahaman MN, et al Comparison of the response of primary human blood monocytes and the U937 human monocytic cell line to two different sizes of alumina ceramic particles. J Orthop Res 22: Nakashima Y, Sun DH, Trindade MC, et al Signaling pathways for tumor necrosis factor-alpha and interleukin-6 expression in human macrophages exposed to titanium-alloy particulate debris in vitro. J Bone Joint Surg Am 81: Xing S, Waddell JE, Boynton EL Changes in macrophage morphology and prolonged cell viability following exposure to polyethylene particulate in vitro. Microsc Res Tech 57: Ingham E, Green TR, Stone MH, et al Production of TNF-alpha and bone resorbing activity by macrophages in response to different types of bone cement particles. Biomaterials 21: Merkel KD, Erdmann JM, McHugh KP, et al Tumor necrosis factor-alpha mediates orthopedic implant osteolysis. Am J Pathol 154:

8 80 BURTON ET AL. 16. Wimhurst JA, Brooks RA, Rushton N Inflammatory responses of human primary macrophages to particulate bone cements in vitro. J Bone Joint Surg Br 83: Maloney WJ, James RE, Smith RL Human macrophage response to retrieved titanium alloy particles in vitro. Clin Orthop Relat Res 322: Rakshit DS, Ly K, Sengupta TK, et al Wear debris inhibition of anti-osteoclastogenic signaling by Interleukin-6 and Interferon-{gamma} Mechanistic insights and implications for periprosthetic osteolysis. J Bone Joint Surg Am 88: Sethi RK, Neavyn MJ, Rubash HE, et al Macrophage response to cross-linked and conventional UHMWPE. Biomaterials 24: Haynes DR, Boyle SJ, Rogers SD, et al Variation in cytokines induced by particles from different prosthetic materials. Clin Orthop Relat Res Shanbhag AS, Jacobs JJ, Black J, et al Macrophage/ particle interactions: effect of size, composition, and surface area. J Biomed Mater Res 28: Green TR, Fisher J, Stone M, et al Polyethylene particles of a critical size are necessary for the induction of cytokines by macrophages in vitro. Biomaterials 19: Yang SY, Ren W, Park Y, et al Diverse cellular and apoptotic responses to variant shapes of UHMWPE particles in a murine model of inflammation. Biomaterials 23: Schwarz EM, Lu AP, Goater JJ, et al Tumor necrosis factor-alpha/nuclear transcription factor-kappab signaling in periprosthetic osteolysis. J Orthop Res 18: Epstein NJ, Warme BA, Spanogle J, et al Interleukin- 1 modulates periprosthetic tissue formation in an intramedullary model of particle-induced inflammation. J Orthop Res 23: Cassel SL, Eisenbarth SC, Iyer SS, et al The Nalp3 inflammasome is essential for the development of silicosis. Proc Natl Acad Sci USA 105: Dostert C, Pêtrilli V, Van Bruggen R, et al Innate immune activation through Nalp3 inflammasome sensing of asbestos and silica. Science 320: O Neill L Immunology. How frustration leads to inflammation. Science 320: Hornung V, Bauernfeind F, Halle A, et al Silica crystals and aluminum salts activate the NALP3 inflammasome through phagosomal destabilization. Nat Immunol 9: Martinon F, Petrilli V, Mayor A, et al Gout-associated uric acid crystals activate the NALP3 inflammasome. Nature 440: Cobelli N, Scharf B, Crisi GM, et al Mediators of the inflammatory response to joint replacement devices. Nat Rev Rheumatol 7: Maitra R, Clement CC, Scharf B, et al Endosomal damage and TLR2 mediated inflammasome activation by alkane particles in the generation of aseptic osteolysis. Mol Immunol 47: Caicedo MS, Desai R, McAllister K, et al Soluble and particulate Co-Cr-Mo alloy implant metals activate the inflammasome danger signaling pathway in human macrophages: a novel mechanism for implant debris reactivity. J Orthop Res 27: St Pierre CA, Chan M, Iwakura Y, et al Periprosthetic osteolysis: characterizing the innate immune response to titanium wear-particles. J Orthop Res 28: Takeshita S, Kaji K, Kudo A Identification and characterization of the new osteoclast progenitor with macrophage phenotypes being able to differentiate into mature osteoclasts. J Bone Miner Res 15: Eisenbarth SC, Flavell RA Innate instruction of adaptive immunity revisited: the inflammasome. EMBO Mol Med 1: Ferrari D, Chiozzi P, Falzoni S, et al Extracellular ATP triggers IL-1 beta release by activating the purinergic P2Z receptor of human macrophages. J Immunol 159: Eisenbarth SC, Colegio OR, O Connor W, et al Crucial role for the Nalp3 inflammasome in the immunostimulatory properties of aluminium adjuvants. Nature 453: Jin C, Flavell RA Molecular mechanism of NLRP3 inflammasome activation. J Clin Immunol 30: Ward JR, West PW, Ariaans MP, et al Temporal interleukin-1beta secretion from primary human peripheral blood monocytes by P2 7-independent and P2 7-dependent mechanisms. J Biol Chem 285: Pearl JI, Ma T, Irani AR, et al Role of the Toll-like receptor pathway in the recognition of orthopedic implant wear-debris particles. Biomaterials 32: Wei S Kitaura H, Zhou P, et al IL-1 mediates TNFinduced osteoclastogenesis. J Clin Invest 115: