Hyperuricemia may result in the deposition of monosodium urate monohydrate (MSU) crystals in joints and soft tissues. When shed from deposits or precipitated de novo (1), acute inflammation may result. Although gouty arthritis was first recognized more than 4,500 years ago (2), the pathogenic mechanisms are still not fully characterized. For most patients, therapy with nonsteroidal antiinflammatory drugs (NSAIDs) or corticosteroids for acute episodes and prevention of recurrence with agents that lower the serum uric acid levels are highly effective. However, there remain many patients with chronic polyarticular gout for whom these therapies are not sufficient because of ineffectiveness, toxicity, or comorbidities. MSU crystals induce a variety of inflammatory cytokines and chemokines including tumor necrosis factor α (TNFα), interleukin-1β (IL-1β), IL-6, CXCL8 (IL-8), and CXCL1 (growth-related oncogene α) (3–6). Recently reported observations identify IL-1 as the pivotal cytokine in gouty inflammation and provide new insights into the role of a molecular complex called the inflammasome in the release of IL-1β by MSU crystals (7–10).
IL-1β synthesis, processing, and secretion
IL-1β is a proinflammatory cytokine, synthesized primarily by monocytes and macrophages, that contributes to the pathogenesis of chronic inflammatory diseases such as rheumatoid arthritis. However, in rheumatoid arthritis, IL-1β does not appear pivotal since inhibition of IL-1 generally does not result in dramatic clinical improvement in the majority of patients, while therapeutic strategies that interfere with other cytokines, such as TNFα, are highly effective. However, IL-1β does appear critical in a number of conditions including Still's disease and hereditary periodic febrile syndromes including familial cold-induced autoinflammatory syndrome, Muckle-Wells syndrome, and neonatal-onset multisystem inflammatory disease (11–13). Recent studies suggest that IL-1β may also play a central role in gout (7, 9).
IL-1β, also called endogenous pyrogen, is highly regulated both transcriptionally and posttranscriptionally, and monocytes and macrophages are the principal cells that contribute to its expression. IL-1β is induced transcriptionally through a number of pathways including through Toll-like receptors (TLRs), which may be activated by microbial pathogens, or by cytokines including TNFα and IL-1β itself. These pathways lead to activation of transcription factors, particularly NF-κB, but also CCAAT/enhancer binding protein β, PU.1, and interferon regulatory factor 4 (14–16). Transcriptional activation of the IL-1β promoter results in the expression of the 35-kd proIL-1β, which is inactive and remains within the cell (17). Although for years there was uncertainty about how IL-1β gained access to the extracellular space, it is now clear that IL-1β–converting enzyme or caspase 1 cleaves the proIL-1β, resulting in the active 17-kd IL-1β that is released from the cell (18) (Figure 1). This dual mechanism for the regulation of IL-1β, requiring first synthesis and then processing, may have developed to ensure that IL-1β is expressed only in those situations where it is truly required.
TLR/IL-1 signaling pathway
Toll receptors were first described in Drosophila, in which they are critical for normal development (19). Search for mammalian counterparts of Toll receptors have resulted in the identification of at least 10 human TLRs. The primary function of mammalian TLRs is to recognize microbial ligands in order to initiate a rapid protective response (20). TLRs consist of a leucine-rich repeat (LRR) ectodomain (21), which is important in the recognition of microbial pathogen-associated molecular patterns, a transmembrane domain, and a cytoplasmic Toll/IL-1 receptor (TIR) domain, which is structurally similar in TLRs and IL-1 receptor (IL-1R). The major downstream signaling pathway of TLRs and IL-1R involves the adaptor molecule myeloid differentiation factor 88 (MyD88), which is recruited to the TIR domain following ligation with TLR ligands or IL-1 (Figure 2). Once MyD88 is localized to a TLR or IL-1R, it recruits other molecules that lead to the activation of NF-κB and the transcription of cytokines and chemokines, such as TNFα, IL-1β, and CXCL8.
The caspase 1–mediated processing of IL-1β is mediated by a molecular platform called the inflammasome (22). The inflammasome, through the association of cytosolic proteins, promotes the oligomerization and activation of caspase 1, which then cleaves proIL-1β (Figure 1). Although there are multiple inflammasomes, it is the NALP3 inflammasome that is important for the activation of caspase 1 by MSU or calcium pyrophosphate dihydrate (CPPD) crystals (9). NALP3 possesses a NACHT domain, which has a nucleotide sequence predicted to bind ATP; an LRR domain, which, like those found in TLRs, is capable of recognizing microbial pathogen-associated molecular patterns; and a pyrin domain, which is critical for homotypic protein–protein interactions (Figure 1). An adopter molecule called ASC links NALP3 with caspase 1 (23) (Figure 1). These interactions result in the association of caspase 1 molecules that are autocatalytically cleaved, resulting in 10-kd and 20-kd fragments that are active enzymatically and which are capable of cleaving or processing proIL-1β (17).
The stimulus to activate the inflammasome may be provided by cell surface receptors that are activated by danger-associated molecular patterns or by pathogen-associated molecular patterns (Figure 1). Danger-associated molecular pattern signals occur when a cell is under stress, which results in the production and release of ATP, which binds and activates the purogenic cell surface receptor P2X7, resulting in activation of the NALP3 inflammasome (24). Additionally, factors such as pore-forming toxins expressed by pathogenic microbes, such as Streptomyces hygroscopicus, Aeromonas hydrophila, or Staphylococcus aureus (24), may provide the danger-associated molecular pattern signal to activate the NALP3 inflammasome. Additionally, pathogen-associated molecular patterns, such as peptidoglycan present on the surface of gram-positive S aureus, or microbial RNA, may activate the NALP3 inflammasome by interacting through the LRR domain. This is similar to the mechanism responsible for activation by cell surface or intracellular TLRs (25). Recent studies, to be described below, demonstrate that MSU and CPPD crystals may also activate the NALP3 inflammasome (9).
The clinical importance of the NALP3 inflammasome was demonstrated when the NALP3 from patients with certain periodic febrile syndromes, including familial cold-induced autoinflammatory syndrome, Muckle-Wells syndrome, and neonatal-onset multisystem inflammatory disease, was shown to possess gain-of-function mutations (26–28). These mutations predispose these patients with periodic febrile syndromes to intermittent episodes of fever and other symptoms related to the release of IL-1β. Macrophages from patients with Muckle-Wells syndrome spontaneously release IL-1β, and, following activation, the release of IL-1β is greater than that observed with macrophages possessing normal NALP3 (29). Therefore, patients with Muckle-Wells syndrome release IL-1β more readily than individuals without the NALP3 mutations. Recent studies have demonstrated that blocking the IL-1R with anakinra is highly effective at terminating febrile episodes and ameliorating progression of disease in patients with Muckle-Wells syndrome and in patients with neonatal-onset multisystem inflammatory disease (11, 28).
MSU-induced NALP3 inflammasome
Monocytes that encounter MSU crystals express and release cytokines, including IL-1β, TNFα, and IL-6, and the neutrophil chemokine CXCL8 (3–6). Since TNFα appears to be upstream of IL-1β in rheumatoid arthritis, studies were performed to determine whether TNFα was responsible for the crystal-induced release of IL-1β. Studies revealed that the effects of MSU crystals were independent of TNFα, since IL-1β was released prior to TNFα and inhibition of IL-1 with IL-1R antagonist suppressed the expression of TNFα (9). Therefore, studies were performed to determine whether MSU crystals were capable of activating the NALP3 inflammasome. The initial studies, which used a monocytic cell line or peripheral blood monocytes from normal volunteers, demonstrated that MSU crystals cleaved proIL-1β and released the active 17-kd molecule from the cells (9).
Subsequent studies used peritoneal macrophages from mice that were genetically deficient in the components of the NALP3 inflammasome, i.e., caspase 1, ASC, and NALP3. In contrast to the normal or wild-type mice, which contained each of these molecules, the addition of MSU crystals to primed macrophages from mice lacking caspase 1, ASC, or NALP3 failed to promote the processing of proIL-1β or the release of the active cytokine from the macrophages. Similar observations were made with CPPD crystals. Further, an in vivo mouse model of gout, crystal-induced peritonitis, also demonstrated that caspase 1 and ASC were critical, since mice deficient in either of these molecules developed significantly reduced MSU crystal–induced inflammation. These observations demonstrate that each of the components of the NALP3 inflammasome is required for the processing of proIL-1β and the induction of crystal-induced inflammation.
The effect of colchicine
Since MSU crystals induced the processing of proIL-1β, experiments were performed to examine the effect of colchicine on the NALP3 inflammasome. Colchicine was highly effective at preventing the MSU- or CPPD-induced processing of proIL-1β and the release of IL-1β (9). Therefore, colchicine is capable of preventing the crystal-induced activation of the inflammasome and the release of IL-1β, although the concentration employed was greater than that which is effective prophylactically (30), suggesting an alternative mechanism for the effectiveness of colchicine at preventing an attack of gout. Low-dose colchicine has also been shown to decrease the adhesion of neutrophils to endothelial cells by altering the distribution of adhesion molecules on endothelial cells (30). The mechanism by which colchicine prevents crystal-induced activation of the inflammasome is not known. It is possible that the ability of colchicine to inhibit microtubules may somehow prevent the crystals from entering the cell, possibly by interfering with the complement membrane attack complex or CD14 (31, 32), or possibly by preventing the crystals from interacting with NALP3 through the LRR domain.
IL-1R signaling is essential for MSU-induced inflammation
Further understanding of the mechanisms responsible for crystal-induced inflammation resulted through dissection of the TLR/IL-1R signaling pathways, using genetically deficient mice. A characteristic feature of crystal-induced arthritis is the rapid accumulation of polymorphonuclear cells (PMNs) that is associated with the ability of the crystals to induce inflammatory mediators including TNFα, IL-1β, IL-6, and CXCL8 (3–6). Because these mediators are induced by activation through the TLR pathway, 2 groups examined the role of TLRs and the downstream signaling pathways using genetically deficient mice (7, 8, 10). Although the results were different concerning the role of TLRs in initiating the response to crystals, both groups, employing peritoneal and air pouch models of MSU-induced inflammation, demonstrated that MyD88 was essential for the induction of MSU-induced inflammation in mice (7, 10). Since MyD88 is critical for TLR and IL-1R signaling, it was possible that either pathway might be important for MSU crystal–induced inflammation.
Further studies demonstrated that it was the IL-1R signaling pathway that was essential for the development of MSU-induced inflammation in vivo. MSU-induced inflammation was greatly reduced in mice lacking IL-1R (7, 9), suggesting a critical role for IL-1R signaling in gouty inflammation. However, this effect was not due to reduced IL-1β production by IL-1R–deficient cells, since macrophages isolated from IL-1R–deficient mice produced essentially the same amount of IL-1β as macrophages from mice that were not deficient in IL-1R (7). Supporting the role of IL-1 in the in vivo model, treatment of mice with antibodies to IL-1α and IL-1β greatly reduced MSU crystal–induced peritonitis (7). Therefore, IL-1R and MyD88 are essential for the induction of MSU-induced inflammation in vivo, even though macrophages lacking IL-1R were capable of synthesizing and releasing IL-1β.
To define the pathogenic mechanism for MSU-induced inflammation, Chen and colleagues (7) performed reciprocal transfer experiments of bone marrow–derived cells from normal control or IL-1R–deficient mice into IL-1R–deficient or control mice that had been lethally irradiated to deplete their bone marrow–derived cells. When bone marrow cells from IL-1R–deficient mice were transferred into control mice, MSU-induced peritonitis occurred. However, when bone marrow from the control mice was transferred into the IL-1R–deficient mice, MSU-induced inflammation was greatly reduced. Similarly, MyD88, which is the downstream target of IL-1R signaling, was not essential in bone marrow–derived cells but was necessary in resident peritoneal cells, such as fibroblasts and endothelial cells, for the induction of MSU-induced inflammation. These observations demonstrate that it is the non–bone marrow cells in the peritoneum, such as endothelial cells or fibroblasts, that respond to IL-1, while bone marrow–derived monocytes and macrophages produce the IL-1β (Figure 3). It is likely that IL-1β induces the expression of adhesion molecules, such as E-selectin, and chemokines, such as CXCL8, that are critical for the migration of PMNs into the site of acute inflammation. Once within the joint, phagocytosis of the crystals by PMNs occurs, the classic finding of gout.
Implications for therapy
Pathogenic crystals induce the expression of a number of cytokines and chemokines from monocytes and macrophages. Of these, IL-1 plays a pivotal role, since MSU crystals promote the processing and release of IL-1β, and signaling by IL-1R is critical for the development of MSU crystal–induced inflammation in vivo. Additionally, antibodies to IL-1 effectively suppressed MSU-induced inflammation. Together, these studies identify the inflammasome and IL-1 as potential targets for therapeutic intervention in crystal-induced inflammation.
The possible effectiveness of IL-1 inhibition was recently evaluated in a mouse model of MSU-induced inflammation (33). Inhibition of IL-1 by anakinra prevented MSU-induced peritoneal neutrophil accumulation, while TNF blockade was not effective. Based on these observations, a pilot, open-label study was initiated in 10 patients with gout who could not tolerate standard antiinflammatory therapies or for whom they had failed (33). All patients received 100 mg anakinra daily for 3 days. Three of the patients had chronic tophaceous gout and 7 had acute, generally polyarticular disease. All patients responded to this short course of therapy, with the patient assessments ranging from 50% to 100% (mean 79%) improvement by the third day. Together, these observations suggest that IL-1 and the inflammasome may be promising therapeutic targets in patients with gout, particularly in those patients with chronic polyarticular disease for whom NSAIDs or corticosteroids are contraindicated or too toxic.
The authors thank Tri Tran, PhD for assistance with the figures, and Drs. Peggy Wu and Ami Kurani for reviewing the manuscript.