Drs. Kanneganti and van der Meer contributed equally to this work.
Engagement of fatty acids with toll-like receptor 2 drives interleukin-1β production via the ASC/caspase 1 pathway in monosodium urate monohydrate crystal–induced gouty arthritis
Article first published online: 29 OCT 2010
Copyright © 2010 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 62, Issue 11, pages 3237–3248, November 2010
How to Cite
Joosten, L. A. B., Netea, M. G., Mylona, E., Koenders, M. I., Malireddi, R. K. S., Oosting, M., Stienstra, R., van de Veerdonk, F. L., Stalenhoef, A. F., Giamarellos-Bourboulis, E. J., Kanneganti, T.-D. and van der Meer, J. W. M. (2010), Engagement of fatty acids with toll-like receptor 2 drives interleukin-1β production via the ASC/caspase 1 pathway in monosodium urate monohydrate crystal–induced gouty arthritis. Arthritis & Rheumatism, 62: 3237–3248. doi: 10.1002/art.27667
- Issue published online: 29 OCT 2010
- Article first published online: 29 OCT 2010
- Manuscript Accepted: 13 JUL 2010
- Manuscript Received: 18 FEB 2010
- NIH. Grant Number: AR-056296
- American Lebanese and Syrian Associated Charities
- Vici grant from the Netherlands Organization for Scientific Research
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
The concept that intraarticular crystals of uric acid by themselves trigger episodes of painful gouty arthritis is inconsistent with the clinical reality. Patients with large deposits of monosodium urate monohydrate (MSU) crystals (tophi) do not necessarily experience gouty attacks. In fact, it is the excessive consumption of food or alcohol that elicits the inflammation of the acute gout attack. The aim of this study was to identify the precise mechanism that initiates flares of gouty arthritis.
Human peripheral blood mononuclear cells (PBMCs) and murine macrophages were stimulated in vitro with MSU, free fatty acids (FFAs), or both in combination. Thereafter, production of interleukin-1β (IL-1β) and activation of caspase 1 were determined. Gouty arthritis was induced in mice with deficiencies in the genes for caspase 1, ASC, NALP3, or IL-1β, and the lack of inflammasome activity during joint swelling or other joint pathologic features was investigated in these mice.
MSU crystals had no biologic effects on PBMCs from healthy subjects, whereas the FFA C18:0 in the presence of MSU crystals induced the release of large amounts of IL-1β following engagement of Toll-like receptor 2 (TLR-2). Interaction of FFAs, but not alcohol, with TLR-2 synergized with MSU crystals to induce an inflammatory reaction. An important event of MSU/FFA-induced acute joint inflammation is the activation of the inflammasome. MSU/FFA-induced release of IL-1β was dependent on activation of caspase 1 and ASC, but surprisingly, not NALP3.
The synergistic effect between FFAs and MSU crystals leads to ASC/caspase 1–driven IL-1β release. This mechanism could explain how constitutionally derived metabolic events initiate attacks of gout via the induction of IL-1β–mediated joint inflammation.
Gout is a chronic inflammatory disease that is characterized by recurrent attacks of acute joint inflammation and is regarded as the prototypical crystal-induced arthropathy (1). It is unknown why only a small number of individuals with hyperuricemia develop gout, which is attributable to the deposition of monosodium urate (MSU) crystals in the joints, and why the attacks of inflammation in patients with gout are sporadic, despite continuous deposition of uric acid crystals in the joints. In addition, it is well known that attacks of gout are often related to a copious meal, consumption of alcohol, or an infectious process. How food components or their metabolites after an abundant meal precipitate acute joint inflammation in the presence of MSU crystals remains to be elucidated.
Several mechanisms have been proposed for the induction of joint inflammation by MSU crystals. Previous findings reported in the literature have implicated disruption of phagolysomes and killing of neutrophils, complement activation, production of oxygen radicals, and eicosanoid release as mechanisms of action (2–4). Induction of the release of proinflammatory cytokines and chemokines by MSU crystals has also been suggested to play an important role (5, 6). Among these mediators, interleukin-1β (IL-1β) is a pivotal proinflammatory cytokine strongly associated with the inflammation in gout (7–9). Consequently, treatments that block IL-1 activity, using either recombinant IL-1 receptor antagonist (IL-1Ra) (anakinra) or anti–IL-1β antibodies, are beneficial in several inflammatory disorders (10). Recently, treatment with IL-1Ra or an IL-1 trap has also been shown to be effective in gout (11, 12), which prompted the hypothesis that IL-1β release is a central event in the inflammatory reaction of a gout attack.
IL-1β is one of the very few cytokines that lacks a signal peptide, and its processing and secretion depend on cleavage by proteolytic cysteine proteases such as caspase 1 (13). Caspase 1 in turn is activated within protein platforms called inflammasomes (9). Several inflammasomes are able to activate caspase 1, all comprising proteins of the NOD-like receptor family (14). MSU crystals have been reported to activate caspase 1 and to induce IL-1β production via the NALP3 inflammasome (8, 9, 15). Interestingly, however, in all of the in vitro studies identifying MSU crystals as an inflammasome activator, either lipopolysaccharide (LPS) or phorbol myristate acetate (PMA) was used to prime cells (7, 8, 15). Recently, we have shown that purified MSU crystals cannot induce IL-1β by themselves, and that a second stimulus is needed (e.g., a bacterial component such as LPS) (16). Although this could indeed explain the triggering role of infections in a gout attack (in which microbial components released during infection synergize with MSU crystals already present in the joint), this could not explain the induction of an attack by food intake or alcohol consumption.
In the present study, we demonstrate that inflammasome activation and IL-1β release take place after exposure to MSU crystals and free fatty acids (FFAs), and this explains the triggering effects of copious meals or alcohol consumption on gout attacks in patients. We demonstrate that C18:0 acting on Toll-like receptor 2 (TLR-2) strongly synergizes with MSU crystals to induce the release of IL-1β and to enhance inflammation, and this effect can be reversed in mice defective in components of the inflammasome. This demonstrates that the release of FFAs after food ingestion or alcohol consumption represents the “missing link” between metabolic changes, inflammasome activation, and gout attacks.
MATERIALS AND METHODS
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
Male C57BL/6 mice were obtained from Charles River Wiga. IL-1β gene–deficient (IL-1β−/−) mice were kindly provided by J. Mudgett (Merck, NJ). TLR2−/− and TLR4−/− mice were provided by S. Akira (Osaka University, Japan). ASC−/−, Nalp3−/−, and caspase 1−/− mice were kindly provided by A. Coyle, J. Bertin, E. Grant (all from Millennium Pharmaceuticals), G. Nunez (University of Michigan), and R. Flavell (Yale University). Mice were bred at the Central Animal Laboratory of Radboud University Nijmegen Medical Centre (RUNMC) or St. Jude Children's Research Hospital. All mice were housed in filter-topped cages, with water and food supplied ad libitum, and used at 10–12 weeks of age. All animal experiments were approved by the animal ethics committee of the RUNMC or St. Jude Children's Research Hospital Committee on Use and Care of Animals.
Preparation of MSU crystals.
MSU crystals were prepared according to the method described by Seegmiller et al (17). Briefly, a 0.03M solution of MSU at a volume of 200 ml was prepared after diluting 1.0 gram of uric acid (Sigma) in 200 ml of sterile water containing 24 grams of NaOH. The pH was adjusted to 7.2 after the addition of HCl, and the solution became pyrogen free after heating for 6 hours at 120°C. The solution was left to cool at room temperature and stored at 4°C. Crystals produced were 5–25 μm in length. On each day of the experiment, a small amount of the crystals was weighed under sterile conditions for application. LPS contamination was controlled by Limulus amebocyte cell lysate assay.
Isolation of mouse peritoneal macrophages and stimulation of cytokine production.
For isolation of mouse peritoneal macrophages and spleen cells, groups of TLR2−/− and TLR4−/− mice (n = 5 per group) and their control littermates were killed, and resident peritoneal macrophages were harvested by injecting the joints with 4 ml of sterile phosphate buffered saline (PBS) containing 0.38% sodium citrate. After washing, the cells were resuspended in RPMI 1640 containing 1 mM pyruvate, 2 mML-glutamine, 100 μg/ml gentamicin, and 2% fresh mouse plasma. Cells were cultured in 96-well microtiter plates (Greiner) at 105 cells per well, in a final volume of 200 μl. The cells were stimulated with either control medium, ethanol, MSU, C18:0 (Sigma), MSU/C18:0, or highly purified Escherichia coli LPS (10 ng/ml). After 24 hours of incubation at 37°C, the plates were centrifuged (500g for 10 minutes), and the supernatant was collected and stored at −80°C until cytokine assays were performed.
Generation of bone marrow–derived macrophages (BMDMs) or bone marrow–derived dendritic cells (BMDCs) and caspase 1 activation.
Bone marrow from mice (ages 8–20 weeks) was flushed out after dissection of the mouse legs. Differentiation into BMDMs occurred in 5 days, at 37°C in 5% CO2, in the presence of Iscove's modified Dulbecco's medium supplemented with 30% of L929 supernatant containing 10% fetal calf serum (FCS) (Invitrogen), 100 units/ml penicillin, and 100 mg/ml streptomycin. For isolation of BMDCs, bone marrow cells were cultured in RPMI 1640 medium supplemented with 10% FCS, 100 units/ml penicillin, 100 μg/ml streptomycin, 50 μM 2-mercaptoethanol, 2 mM sodium pyruvate, and 20 ng/ml granulocyte–macrophage colony-stimulating factor (PeproTech). The medium was replaced with fresh medium on day 3. On days 6, 8, and 10, nonadherent cells were harvested, washed, and replated in fresh medium. Cells were treated with either ultrapure LPS (purchased from Invitrogen), which was used in a concentration of 10 μg/ml, or ATP (purchased from Sigma), which was used in a final concentration of 3 mM. For immunoblotting, cells were washed twice with PBS and lysed in buffer (150 mM NaCl, 10 mM Tris, pH 7.4, 5 mM EDTA, 1 mM EGTA, 0.1% Nonidet P40), which was supplemented with a protease inhibitor cocktail (Roche). After clarification and denaturation with sodium dodecyl sulfate (SDS) buffer, samples were boiled for 5 minutes. Separation of the proteins was done using SDS–polyacrylamide gel electrophoresis. Thereafter, the proteins were transferred to a nitrocellulose membrane. These membranes were coated with primary antibodies, and active caspase 1 was detected using a secondary anti-rabbit antibody conjugated to horseradish peroxidase, followed by evaluation using an enhanced chemiluminescence assay (18).
Isolation of human peripheral blood mononuclear cells (PBMCs) and stimulation of cytokine production.
Isolation of human PBMCs was performed as described elsewhere (19), with minor modifications. PBMCs (5 × 105) in a 100-μl volume were added to round-bottomed 96-well plates (Greiner) and incubated with either 100 μl of culture medium or the various stimuli (highly purified MSU crystals at 30–300 μg/ml, ultra pure fatty acid [C18:0] at 200 μM, highly purified E coli LPS at 1–10 ng/ml, and palmitoyl-3-cysteine [Pam3Cys] at 10 μg/ml). In additional experiments, several TLR or NOD ligands were used as the stimulus. In some experiments, cells were pretreated with or without double-extracted Bartonella quintana LPS (1 μg/ml), which is a TLR-4 antagonist (20), or anti–TLR-2 antibodies (10 μg/ml), 30 minutes before treatment with MSU/C18:0 (300 μg/ml–200 μM). After 24 hours of culture, the supernatants were collected, centrifuged at 1,200 revolutions per minute, and stored at −80°C until assayed.
Induction of joint inflammation by MSU crystals.
Joint inflammation was induced by intraarticular injection of highly pure MSU in a range of doses (30–300 μg), C18:0 (200 μM), MSU/C18:0 (300 μg/200 μM), or streptococcal cell wall (rhamnose content; 25 μg) in 10 μl of PBS into the right knee joint of naive mice. Four hours after intraarticular injection, swelling of the knee joint was determined, synovial tissue was isolated, and the knee joints were removed for histologic assessment.
Measurement of joint swelling.
The extent of inflammation in the mouse knee joints was measured by either macroscopic scoring or using the 99mTc uptake method. Macroscopic joint swelling was scored on a scale ranging from 0 to 3. After the skin was removed, the knee joint was scored for the extent of swelling (0 = no swelling to 3 = severe swelling). The 99mTc uptake method was performed as previously described (21, 22). Values for joint swelling were expressed as the ratio of the 99mTc uptake in the inflamed joint to that in the control joint (left knee joint); all ratio values >1.10 are considered to represent joint swelling.
RNA isolation and polymerase chain reaction (PCR) amplification.
Immediately after cervical dislocation, synovial tissue was isolated from the inflamed knee joints. The samples of synovium were immediately stored in N2 until total RNA isolation. RNA was extracted as described previously (21). To obtain complementary DNA, standard reverse transcription–PCR was performed using oligo(dT) primers. Subsequently, quantitative PCR was performed using the ABI Prism 7000 Sequence Detection System (Applied Biosystems). PCR analyses of murine GAPDH, IL-1β, IL-6, and cytokine-induced neutrophil chemoattractant (KC; the murine homolog of IL-8) were performed with SYBR Green PCR Master Mix (Applied Biosystems). Quantification of the PCR signals of each sample was performed by comparing the threshold cycle (Ct) values of the gene of interest, obtained in duplicate, with the Ct values of the GAPDH housekeeping gene. We validated all primers according to the protocol, and the standard curves were all within the tolerable range (21–23).
Protein levels of murine IL-1β or KC were measured in patella washouts. Four hours after injection of MSU/C18:0, the patellae were isolated from inflamed knee joints and cultured for 1 hour at room temperature in RPMI 1640 medium containing 0.1% bovine serum albumin (200 μl/patella). For determination of intracellular IL-1β levels, the patellae were frozen directly after isolation. After repeated freeze-thawing, the levels of IL-1β were determined. Human IL-1β and IL-8 were measured using either specific or commercial enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems) and Pelikine Compact (Sanquin). The sensitivity of both ELISAs is below 5 pg/ml (22, 23). Mouse cytokines were determined by Luminex technology. Kits for IL-1β and KC were obtained from Bio-Rad.
Mice were killed by cervical dislocation. Whole knee joints were removed and fixed in 4% formaldehyde for 7 days before decalcification in 5% formic acid and processing for paraffin embedding. Tissue sections (7 μm) were stained with hematoxylin and eosin. Histopathologic changes in the knee joints were scored in the patella/femur region on 5 semiserial sections. Scoring was performed on decoded slides by 2 separate observers (LABJ and MO), using scoring parameters to measure the inflammatory cell influx, in which the amount of cells infiltrating the synovial lining and the joint cavity was scored from 0 to 3 (21–23).
Differences between experimental groups were tested using the Mann-Whitney U test. Data are expressed as the mean ± SEM, unless stated otherwise. P values less than 0.05 (in comparison with wild-type control mice) were considered significant.
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
Lack of induction of IL-1β production or joint inflammation by highly purified MSU crystals.
Although it has been reported that exposure of mononuclear cells to MSU crystals leads to the production of active IL-1β (16), many of these studies used LPS- or PMA-primed cells to observe these effects. In the present series of experiments, we confirmed that highly pure MSU crystals could not induce IL-1β production in human PBMCs, even at high concentrations. Likewise, murine peritoneal macrophages did not respond to highly pure MSU crystals, but did respond to LPS (Figures 1A and 2C).
Several reports have indicated that MSU crystals will induce acute inflammation when injected locally, e.g., into the peritoneal cavity of mice (8). Since gout manifests as an acute joint inflammation, we injected highly pure MSU crystals directly into murine knee joints. As demonstrated in Figure 1B, even at high doses (up to 300 μg MSU per joint), MSU did not induce joint swelling, when inflammation of the knee joint was measured with a highly sensitive method (99mTc uptake). Histologic assessment of the joints confirmed the lack of joint inflammation after local injections of MSU crystals.
In addition, intraarticular injection of 300 μg MSU crystals did not lead to enhanced influx of granulocytes (Figures 1D and E). Even a single dose of 600 μg MSU crystals injected intraarticularly did not give rise to signs of joint inflammation (results not shown). Furthermore, MSU crystals were not able to induce a substantial amount of messenger RNA (mRNA) for IL-1β in synovial tissue (Figure 1C).
To further explore the lack of effect of highly pure MSU crystals on the induction of joint inflammation, we analyzed whether MSU could activate caspase 1. As shown in Figure 1F, there was no effect on caspase 1 activation. Thus, these data convincingly demonstrate that MSU crystals alone do not induce inflammation, and additional stimuli are needed for the activation of the inflammasome, caspase 1, and IL-1β by MSU.
Synergy between MSU and saturated fatty acids.
In patients with gout, inflammatory attacks are often triggered by ingestion of a copious meal or alcohol consumption. We hypothesized that one or more substances or metabolites released through these processes could have a stimulatory effect on the inflammatory properties of MSU crystals. One obvious candidate is alcohol concentration, which may influence the induction of inflammation by MSU crystals. However, ethanol concentrations up to 0.4% were not able to enhance the effect of MSU crystals on the production of proinflammatory cytokines such as IL-1β (Figure 2A). Moreover, increasing concentrations of ethanol actually inhibited LPS-induced IL-1β production, whereas no influence on TLR-2 signaling was found (Figure 2A).
A second important candidate is represented by FFAs that are released from adipose tissue and liver depots after ingestion of a large meal or alcohol consumption (24). Therefore, we investigated the capacity of several saturated fatty acids, i.e., C12.0, C14.0, C16.0, and C18:0, to synergize with MSU crystals for the induction of IL-1β. Interestingly, the combination of MSU crystals and saturated fatty acid C18:0 resulted in the synergistic production of IL-1β by human PBMCs (Figures 2B and C). The other saturated fatty acids examined appeared to be less potent. Similarly, exposure of PBMCs to the combination of MSU and the FFA C18:0 led to an enhanced production of the chemokine IL-8, which is able to attract neutrophils, an important feature of the gout inflammatory reaction. C18:0 alone was also able to induce IL-8 in PBMCs (results not shown). Exposure to MSU/C18:0 did not result in robust production of tumor necrosis factor α (TNFα) in human PBMCs or mouse peritoneal macrophages (2-fold increase in TNFα, compared with 30-fold increase in IL-1β) (results not shown).
Pharmacologic inhibition of caspase 1 with YVAD strongly reduced the production of IL-1β that was induced by MSU/C18:0 in human PBMCs (Figure 2D), indicating that active caspase 1 is needed for IL-1β production. Murine peritoneal macrophages were also activated by MSU/C18:0 to produce IL-1β (Figure 2E). The role of caspase 1 in MSU/C18:0-induced IL-1β production was confirmed by showing that MSU/C18:0 activated caspase 1 in murine BMDCs (Figure 2F).
Role of TLR-2 and TLR-4 in the effects of FFAs.
It has been suggested that fatty acids may trigger cell activation through both TLR-2 and TLR-4 (25). Therefore, we examined whether different known TLR ligands will synergize with MSU to induce IL-1β. As shown in Figures 3A and B, among the TLR ligands tested, only the TLR-2 agonist Pam3Cys synergized with MSU to induce IL-1β in human PBMCs. MSU did not synergize with the ligands for TLRs 3, 4, 5, 7, 8, and 9 or those for the NOD1 or NOD2 genes (Figure 3A and results not shown). The combination of a low dose of LPS (1 ng/ml) with MSU resulted in slightly enhanced IL-1β production in human PBMCs (Figure 3A), as reported previously (16).
To investigate whether TLR-2 or TLR-4 mediates the synergistic effects of FFAs on MSU crystals, peritoneal macrophages from both TLR2-deficient and TLR4-deficient mice were stimulated with MSU/C18:0. As shown in Figures 3C and D, TLR-2, and not TLR-4, is involved in the enhanced IL-1β production that occurs after exposure to the combination of MSU and C18:0. To find out whether this also occurs in human cells, PBMCs were stimulated with MSU/C18:0 in combination with an anti–TLR-2 antibody or a potent TLR-4 antagonist. The results, as shown in Figures 3E and F, confirmed that only TLR-2 ligation is needed for IL-1β production induced by the MSU/C18:0 combination.
Induction of gouty arthritis by MSU/C18:0.
In a followup set of experiments, we investigated whether the MSU/C18:0 combination could lead to acute arthritis in vivo. Intraarticular injection of MSU alone, C18:0 alone, or both in combination was studied in C57BL/6 mice. As shown in Figures 4A and B, only the combination of MSU crystals with C18:0 fatty acids induced joint swelling and influx of inflammatory cells in the joint cavity. Analysis of the inflamed synovium demonstrated enhanced mRNA expression and protein production for several proinflammatory cytokines (results not shown), which was not seen with MSU or C18:0 alone. The levels of both intracellular and released IL-1β were significantly elevated after intraarticular injection of MSU/C18:0 (Figure 4C).
Exposure of murine peritoneal macrophages to the MSU/C18:0 combination resulted in elevated concentrations of KC, the murine homolog of IL-8 (Figure 4D). Histologic assessments performed 4 hours after the local injection of MSU/C18:0 revealed that the MSU/C18:0 combination induced joint inflammation that was dominated by neutrophils (Figure 4F), whereas this was not evident histologically in mouse tissue extracts exposed to MSU or C18:0 alone (Figures 1E and 4E).
Role of the NALP3 inflammasome in MSU crystal– and C18 fatty acid–induced arthritis.
The NALP3 inflammasome has been reported to consist of the components NALP3, ASC, and caspase 1, and recognition of uric acid crystals by NALP3 has been suggested to activate caspase 1 (8). Using a series of gene-knockout mice, we examined the contribution of these components to MSU/C18:0-induced arthritis. Joint swelling, determined 4 hours after intraarticular injection of MSU/C18:0, was suppressed in caspase 1 and ASC gene–deficient mice. ASC-deficient mice showed the strongest reduction in the macroscopic score for joint swelling (Figure 5A). Surprisingly, NALP3 did not appear to be involved in MSU/C18:0-induced arthritis.
Release of IL-1β is crucial for the induction of joint inflammation by MSU/C18:0, as shown by the results in IL-1β−/− mice using the 99mTc uptake method (Figure 5B). The levels of both released and intracellular IL-1β were strongly reduced in synovial tissue from caspase 1 and ASC gene–deficient mice, but not in the Nalp3-deficient mice (Figures 5C and D). These results indicate that NALP3 is bypassed in the production of IL-1β induced by MSU in combination with C18:0.
Surprisingly, an increased transcription of mRNA for IL-1β was found in the Nalp3−/− mice compared with control mice (Figure 5E). In contrast, ASC deficiency led to reduced expression of IL-1β mRNA in the synovial tissue, indicating that ASC also plays a role in transcription, and not only for the processing of IL-1β. In addition, levels of the chemokine KC were decreased in ASC-deficient mice (Figure 5F). Not only were the protein concentrations of IL-6 and KC clearly reduced in caspase 1– and ASC-deficient mice, but also, at the level of transcription, a decrease in these inflammatory mediators was seen (results not shown).
Finally, histologic assessments performed at 4 hours after the injection of MSU/C18:0 showed that ASC-deficient mice were almost completely protected against arthritis (Figure 6D). Low numbers of inflammatory cells were detected in the joint cavity of the ASC-deficient mice, as compared with the extent of inflammatory cell influx in wild-type mice (Figure 6D compared with Figure 6A). Caspase 1–deficient mice also exhibited reduced joint inflammation as compared with wild-type mice (Figure 6C), but not to the same degree as that in ASC-deficient mice. Experiments in the Nalp3−/− mice showed normal influx of polymorphonuclear neutrophils into the joint cavity (Figure 6B), implying that NALP3 is not crucial for the influx of these cells. Significantly reduced numbers of polymorphonuclear neutrophils were found in the joints of caspase 1–deficient and ASC-deficient mice that received MSU/C18:0 intraarticularly (Figure 6E, left). Likewise, a strong reduction of polymorphonuclear neutrophil influx was seen in the joints of IL-1β–deficient mice (Figure 6E, right), underscoring the pivotal role of IL-1β in gout.
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
In the present report, we proposed a pathogenetic model to explain the pathways that link metabolic changes with induction of inflammation in gout, mechanisms that involve synergistic effects between the C18:0 fatty acid and MSU crystals. Pure MSU crystals do not activate caspase 1, and they cannot induce the production of active IL-1β. However, costimulatory signals induced by other stimuli can strongly synergize with MSU crystals in the induction of inflammation, and cell stimulation by C18:0 fatty acids through TLR-2 can provide such a synergistic impulse. In addition, C18:0 fatty acids in combination with MSU crystals activate caspase 1 and induce IL-1β–driven joint inflammation. The inflammasome components caspase 1 and ASC are also needed for the induction of both IL-1β and acute joint inflammation by MSU/C18:0. Interestingly, however, the induction of IL-1β through MSU/C18:0 exposure is NALP3-independent.
Several reports claim that MSU crystals are potent inducers of IL-1β production. However, in most of these studies, the experiments were performed in cell lines that were preactivated with PMA or LPS. Such preactivation is necessary for the induction of both IL-1β mRNA and proIL-1β. In experiments using pure MSU alone, even at very high doses (up to 1,000 μg/ml [results not shown]), we were unable to induce either caspase 1 activation or IL-1β production in human as well as murine cells. The addition of ATP, which is needed for optimal caspase 1 activation through NALP3 (26, 27), to the stimulation cocktail did not lead to the activation of caspase 1 (Figure 1F). Intraarticular injection of high amounts of MSU crystals in vivo in the synovial tissue of mice was also unable to induce arthritis.
These data suggest that the induction of inflammation by MSU crystals is not a one-step process that invariably leads to cytokine release. This conclusion fits with the condition in patients with gout, many of whom have deposition of tophi in the joints or tendons without experiencing continuous inflammatory reactions. In contrast, the attacks of acute arthritis occur especially in the middle of the night, and they are known to be triggered by an abundant meal, alcohol consumption, or fasting. A common metabolic consequence of these processes is represented by the release of FFAs in the circulation (24), which have recently been shown to be recognized by TLRs and to induce an inflammatory reaction (25, 28). We hypothesized that FFAs can synergize with MSU crystals in order to induce inflammasome activation and cytokine release. Indeed, while MSU crystals or FFAs alone were very weak inflammatory stimuli, the combination of C18:0 fatty acids and MSU crystals had very strong synergistic effects on cytokine release (Figure 2). Moreover, intraarticular injection of MSU and C18:0 fatty acids induced a severe arthritis, with activation of caspase 1, production of both IL-1β and KC, and subsequent neutrophil influx. In the majority of the previous studies of MSU/IL-1β production, LPS was used as the first “priming” signal and MSU as a second signal for the induction of IL-1β. However, LPS is not a clinically relevant first signal with respect to gout.
In contrast to FFAs, ethanol did not directly synergize with MSU crystals in the induction of IL-1β production (results not shown). Ethanol has been reported to modulate the activity of both TLR-2 and TLR-4 (29, 30), and this effect was corroborated in our study, in which the addition of alcohol inhibited TLR-4 signaling (Figure 2A). In addition, increasing alcohol concentrations did not affect TLR-2 signaling, which is consistent with a previous study in which it was shown that alcohol did not affect TLR-2 signaling or recruitment of TLR-2 to lipid rafts (31).
The inflammatory effects of MSU crystals have recently been suggested to involve activation of the NALP3 inflammasome (8), followed by caspase 1–dependent processing of proIL-1β. The role of IL-1β release in the pathogenesis of inflammation in gout has been strengthened by studies showing that anakinra (recombinant IL-1Ra) has therapeutic effects in gout (11). Induction of a potent IL-1β response depends on the induction of mRNA transcription on the one hand, while also involving the processing of proIL-1β into the active cytokine.
Although we found that IL-1β mRNA transcription by C18:0/MSU crystals was strongly dependent on TLR-2 engagement, the role of the specific inflammasome components necessary for caspase 1 activation, particularly the role of NALP3, remained less clear. ASC-deficient mice were protected against inflammation induced by the C18:0/MSU combination, suggesting an important role of this inflammasome. Surprisingly, however, the MSU/C18:0-induced joint inflammation was NALP3-independent, since Nalp3-deficient mice developed gouty arthritis in a manner similar to that in wild-type mice. Very recently, it has been shown in a model of antigen-induced arthritis that NALP3 was not critical for the induction of joint inflammation, although this model was IL-1β–dependent (32). These data suggest that a different inflammasome platform is engaged in vivo by MSU crystals/C18:0 fatty acids, and indicates that during certain inflammatory conditions, NALP3 is bypassed for IL-1β production.
Both ingestion of an abundant meal and alcohol consumption can result in an increase in FFA concentrations in the circulation (33). Fasting is also associated with inflammatory attacks of gout, and it is well-known that during fasting, the serum concentrations of FFAs are elevated (34). Because all of these well-known risk factors for the induction of a gout attack are accompanied by an increase in circulating concentrations of FFAs, and C18:0 fatty acids, as shown herein, act in synergy with MSU crystals for the induction of inflammation, it is tempting to speculate that the engagement of TLR-2 by FFAs and the subsequent potentiation of inflammasome activation by MSU crystals represents the link between the metabolic changes preceding a gout attack and the inflammatory flare in the joints of patients. In line with this conclusion is the fact that in patients with gout, elevated levels of FFAs were found shortly after exacerbation of joint inflammation (35, 36).
An aspect that should not be omitted is the presence of a residual inflammatory reaction in caspase 1–knockout mice injected with MSU crystals and C18:0 fatty acids. At least 2 explanations could account for this finding. On the one hand, the production of other cytokines/chemokines (e.g., IL-8/KC) can also be induced by this inflammatory cocktail. On the other hand, inflammasome-independent mechanisms could also account for the partial activation of IL-1β. Indeed, in vitro studies have previously shown that neutrophil-derived serine proteases such as proteinase 3 or elastase can also process proIL-1β into active fragments (37), and we have recently demonstrated the existence of inflammasome-independent pathways of IL-1β processing in a model of acute arthritis (21). Inflammasome activation is therefore an important, but by no means the only, pathway through which MSU crystals in combination with FFAs could induce inflammatory reactions in the joints.
Thus, in the present study, we have demonstrated strong synergistic effects on inflammation in the interaction between MSU crystals and FFAs. Since MSU crystals alone are not able to display inflammatory properties that would explain the basis of gout attacks, we hypothesized the existence of a metabolic change that is induced by the risk factors known to predispose individuals to attacks of the disease and that potentiates the effects of the crystals. Release of FFAs by many of these risk factors, as well as our results corroborating their TLR-2–dependent synergistic role in MSU-induced inflammation, place FFAs at the crossroads between the metabolic changes preceding a gout attack and the occurrence of inflammation in the joint. These findings not only contribute to the understanding of the pathogenesis of gout, but also offer a new therapeutic target in patients with gout, since treatments aimed at impeding the interaction between MSU crystals and FFAs could prove of therapeutic value for both the prevention and treatment of gout attacks.
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Joosten had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Joosten, Netea, Koenders, Stienstra, Kanneganti, van der Meer.
Acquisition of data. Joosten, Mylona, Koenders, Malireddi, Oosting, van de Veerdonk, Giamarellos-Bourboulis.
Analysis and interpretation of data. Joosten, Netea, Koenders, Stalenhoef, van der Meer.
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
Prof. Dr. P. Vandenabeele is acknowledged for his kindness in providing the anti–caspase 1 p20 antibody.
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
- 34Fasting induces changes in peripheral blood mononuclear cell gene expression profiles related to increases in fatty acid β-oxidation: functional role of peroxisome proliferator activated receptor α in human peripheral blood mononuclear cells. Am J Clin Nutr 2007; 86: 1515–23., , .