What makes gouty inflammation so variable?

Recent discoveries illuminate the complex alignments of opposing endogenous and exogenous factors at several steps in the gouty inflammation process (Table 1). The canonical NLRP3 inflammasome response to the ‘second signal’, provided by urate crystals, requires ‘first signal’ priming events, including increases in activation of NADPH oxidase and NF-κB, expression of pro-caspase-1, pro-IL-1β, and NLRP3 mRNA, and particle phagocytic capacity [21‐23]. Certain first signal NLRP3 inflammasome activators can promote and sustain chronic low-grade inflammation, clinically evident chronic synovitis in gout, and acute inflammatory flares in response to deposited urate crystals. These include C5a [24], GM-CSF [25], and the TLR4 ligand and abundant neutrophil granular protein heterodimer S100A8/A9 (also known as calprotectin) [26]. These mediators are among the factors that further amplify inflammation by promoting phagocyte movement into the inflammatory locus in gout, and activation of phagocytes at those sites (Fig. 1).

Exogenous, dietary-induced, first signal NLRP3 inflammasome activators include long-chain saturated fatty acids such as the TLR2 and TLR4 ligand palmitate [23]; they also appear to include spikes in systemic levels of the short-chain fatty acid acetate [27]. Such acetate spikes can be pronounced (e.g., 20-fold rise) rapidly after alcohol intake. Increased acetate, which signals through the G protein coupled receptor GPR43 [27], could contribute to the association of acute gout flares with high level consumption of all forms of alcohol [28].

New findings highlight the genetic impact on the clinical expression of the gouty arthritis phenotype (Table 1), including arthritis flares frequently associated with dietary excesses [29]. The nutritional biosensor peroxisome proliferator-activated receptor (PPAR)-γ is a ligand-dependent nuclear receptor and transcription factor that regulates insulin sensitivity, adipocyte differentiation, and glucose homeostasis. PPAR-γ also exerts a variety of anti-inflammatory effects, in part by transrepression of numerous NF-κB target genes. These effects include suppression of experimental gouty inflammation by some PPAR-γ agonists, mediated in part by suppression of TNF-α [30, 31]. Notably, urate crystals rapidly induce PPAR-γ in monocytes in vitro [31]. However, the natural ligand(s) of PPAR-γ that could limit gouty arthritis remain unclear.

In a study of a Chinese cohort [32], PPAR-γ co-activator 1B (PPARGC1B) gene variance permissive for inflammation was increased in those with gout, with the PPARGC1B risk A allele rs45520937 approximately doubling the risk of gout. The PPARGC1B risk allele A is common (13% in Han Chinese, 4% in Whites) [32]. Transfecting cDNA with this risk allele into cultured human macrophage lineage cells augments inflammatory responsiveness to urate crystals [32].

A small but significant increase in the risk of gout has been linked with functional variants in the inflammasome component caspase activation and recruitment domain (CARD) gene CARD8, IL1B, and the TLR4 and TLR2 receptor complex co-receptor molecule CD14 [33]. Multiplicative interactions of the gout-associated IL1B risk genotype with that of CARD8 amplify the risk of gout [33]. In contrast, genetic variance in TLR4 has been linked with gout risk in some populations, but with widely differing results, and the impact is less clear for the heritable variance of TLR4 in gout [34].

Epigenetic effects are increasingly well understood in the regulation of inflammation, exemplified by the pro-inflammatory effects of some class I histone deacetylases involved in the ability of urate crystals to initiate phagocyte activation [35]. Significantly, phagocyte activation for inflammation is mediated by miR-155. Indeed, peripheral blood mononuclear cells (PBMCs) from gout patients and mouse tissues from experimental acute gout have been shown to have increased miR-155 compared to healthy controls [36]. Urate crystals induce miR-155 in phagocytes and, in turn, elevated miR-155 levels increase the capacity of urate crystals to induce IL-1β and TNF-α; decreased SH2-containing inositol 5’-phosphatase-1 has been shown to be implicated in this mechanism [36].

It is well-recognized that ULT can precipitate more acute gout flares. This may be mediated by inflammatory effects through ULT-induced changes in tissue urate equilibrium and remodeling of articular urate crystal deposits, leading to greater accessibility of crystals to inflammatory cells. Essentially, urate crystal deposits formed at different sites within the joint (including synovial fluid, the articular cartilage surface, and within tendons and ligaments, often at insertion sites) have different physical characteristics [37]. Urate crystals at sites of initial nucleation on templates of collagen and proteoglycan-rich connective tissues and urate crystals with a dense protein coating may have a distinct inflammatory potential from crystals at sites of secondary nucleation [37]. It is likely that dynamic changes in urate crystal morphology develop not only with ULT, but also with spikes in the level of hyperuricemia. Moreover, hyperuricemia markedly increases the inflammatory potential of exogenous urate crystals in experimental gout in mice in vivo [38].

A high level of soluble urate, apart from promoting urate crystal deposition, appears to exert a variety of priming effects on inflammation [38‐40]. PBMCs of subjects with asymptomatic hyperuricemia were observed to generate greater amounts of multiple inflammatory cytokines than cells from healthy controls when stimulated ex vivo [38, 39]. Moreover, elevated soluble urate suppresses mononuclear phagocyte autophagy and IL-1ra expression and, conversely, activates multiple inflammatory pathways in mononuclear phagocytes, including NF-κB activity and, via the AKT (protein kinase B) pathway and the proline-rich AKT substrate 40 kDa (PRAS 40), activity of the mammalian target of rapamycin [38, 39]. Soluble urate may also activate the NLRP3 inflammasome under hypoxic conditions [41]. However, the impact of hyperuricemia alone on inflammation requires further investigation. In this context, physiologic concentrations of soluble urate have been suggested to have anti-inflammatory as well as chondroprotective effects [42].

Replicative senescence not only develops with aging, but also with recurring and chronic inflammation. Senescence promotes tissue inflammation, particularly by differentiation to the senescence-associated secretory cellular phenotype [43]. A recent study suggests linkage of gouty arthritis and cardiovascular disease in gout to leukocyte senescence [44]. Specifically, PBMC senescence has been assessed by quantification of telomere length and telomerase activity in a discovery cohort of Dutch patients with gout (and healthy controls), along with a New Zealand replication cohort [44]. Gout patients have significantly shorter telomeres, with telomere erosion being higher at all ages in patients with the disease and being correlated with gout flare frequency [44]. The shortest telomeres are seen in gout patients with cardiovascular disease [44].

PPAR-γ is one of the various metabolic regulators that influence gouty inflammation. Specifically, AMPK is a nutritional biosensor that promotes healthy metabolic adaptations to stress, including glucose transport, insulin sensitization, lipid metabolism including β-oxidation of fatty acids, mitochondrial function and biogenesis, balanced cell growth, and autophagy [45]. AMPK is not simply a metabolic ‘super-regulator’ but also suppresses oxidative stress and inflammation [46, 47]. Active AMPK also mediates some of the therapeutic effects of metformin and the anti-inflammatory effects of methotrexate, salicylates, and high-dose aspirin [46]. Activated AMPK markedly suppresses mononuclear phagocyte responses to urate crystals in vitro, including NLRP3 inflammasome activation and IL-1β and chemokine release [47]. AMPK also promotes anti-inflammatory M2 macrophage polarization and autophagy [47]. Genetic AMPK α1 chain deficiency markedly increases mouse inflammatory responses to urate crystals in vivo [47]. Conversely, a pharmacologic AMPK activator suppresses experimental gout-like inflammation in vivo [47].

Caloric deprivation and exercise are among the factors that elevate tissue AMPK activity levels by increasing the cellular AMP:ATP ratio [45, 46]. Conversely, AMPK activity at the tissue level is inhibited by dietary intake of palmitate, fructose, and alcohol excess, as well as by exposure of cells to IL-1β, TNF-α, and urate crystals in vitro [45, 47]. The impact of nutritional biosensing by AMPK on gout is likely substantial in a disease marked by common metabolic co-morbidities and dietary excesses. In this context, obesity, metabolic syndrome, type II diabetes, and hyperglycemia diminish tissue AMPK activity [45]. Moreover, AMPK activity limits progression of several common gout comorbidities, including non-alcoholic steatosis and chronic kidney disease [48].

Dietary (e.g., in purine and overall caloric intake) and alcohol consumption restrictions, as well as a reduction in obesity, are dietary measures that have an impact on gouty inflammation [29, 30]. Recently, serum omega-3 fatty acid concentration was suggested to be inversely proportional to the risk of frequent acute gout flares [49]. Omega-3 fatty acids blunt NLRP3-mediated caspase-1 activation and IL-1β release in response to urate crystals and other agonists [50]. Internalization of omega-3 fatty acids by the G protein-coupled receptors 120 (GPR120) and GPR40, and signaling by their downstream scaffold protein β-arrestin-2 (ARRB-2), are centrally implicated in the mechanism [50]. ARBB-2 limits NF-κB activation, and is directly associated with and inhibits NLRP3 [50]. In addition, a ketogenic diet (involving substantial carbohydrate intake restriction), inhibits experimental gouty inflammation via β-hydroxybutyrate and inhibition of NLRP3 inflammasome first-signal priming, as well as IL-1β release from phagocytes [26].

Recent work has suggested beneficial effects of short-chain fatty acids (including acetate and butyrate) in inflammation, including gouty arthritis [35, 51, 52]. Short-chain fatty acids exert a variety of anti-inflammatory effects, mediated in part by limiting leukocyte activation [35]. To date, butyrate is the best studied short-chain fatty acid and is available not simply via intestinal microbiome action on dietary fiber intake, but also from intake of milk fat of grass-eating animals (e.g., cheese, butter, cream). Butyrate inhibits urate crystal-induced expression and release of IL-1β from PBMCs in vitro, transduced by effects including inhibition of certain class I histone deacetylases [35]. Importantly, intestinal microbiome dysbiosis has been reported in two studies of Chinese adults with gout [53, 54]. Fecal material findings reported in gout subjects have included depletion of Faecalibacterium prausnitzii, which normally exerts anti-inflammatory effects via butyrate production [53], and differences in several metabolites that modulate inflammation (e.g., increased succinate, which increases IL-1β via modulation of hypoxia inducible factor-1α) [54].

In experimental gout-like inflammation in mice, a high-fiber diet has been reported to induce more rapid resolution, but not onset, of the urate crystal-induced inflammatory response [52]. These results were mimicked by acetate administration, which was effective even after injection of urate crystals into the knee joint, and at the highest level of the inflammatory response [52]. The contributing mechanisms include increased neutrophil apoptosis and inflammation-dampening uptake of apoptotic neutrophils by macrophages (‘efferocytosis’), with associated decreases in NF-κB activity, and increased production of certain anti-inflammatory mediators [52]. Thus, acetate has paradoxical effects on the inflammatory response in experimental gout [26, 27] (Table 1).

IL-37, a member of the IL-1 family, is an anti-inflammatory cytokine differentially regulated during the course of tissue inflammatory reactions, largely through expression by epithelial cells and mononuclear leukocytes [55, 56]. IL-37 suppresses multiple innate inflammatory responses in vitro and in vivo, acting partially via inhibition of the NLRP3 inflammasome and activation of suppressor of cytokine signaling 3 [55]. IL-37 actions are mediated partly by the phosphatidylserine receptor Mer receptor tyrosine kinase signaling, and enhanced expression of IL-1R8 [55]. Exogenous human IL-37 markedly suppresses experimental urate crystal-induced inflammation in mice [55]. In a recent study of PBMCs from subjects with acute gout, non-acute gout, non-acute gouty arthritis, and healthy controls, IL-37 was reported as significantly greater in the non-acute gouty arthritis patients compared to acute gout and healthy controls [56], which is in contrast to increases in IL-1β, IL-6, and TNFα in the acute gout group.

Recently, a seasonal, summer drop in the circulating protease inhibitor alpha-1 antitrypsin (and inversely higher stimulated IL-1β production) has been linked with increased summer flares of gout in a large European functional genomics consortium study [57]. Alpha-1 antitrypsin is anti-inflammatory, which acts partly by inducing IL-1ra and inhibiting proteolytic generation of IL-1β, including in response to urate crystals [58]. Moreover, the recombinant human alpha-1 antitrypsin-IgG1 Fc fusion protein markedly inhibits joint inflammation in experimental acute gouty arthritis in mice [58].

A fundamental endogenous mechanism limiting the NLRP3-mediated inflammasome pathway in vitro and in vivo is cognate adaptive immunity, which acts via effector and memory CD41 T cells and cell-to-cell contact, and is mimicked by stimulation with selected TNF family ligands (e.g., CD40L) [59]. Moreover, type I interferon suppresses NLPR3 inflammasome activation and IL-1β release [60]. As such, some adaptive immune responses might help maintain clinical inflammatory quiescence of subcutaneous tophi, in which granulomatous structures have been described to have a coronal ring that includes T cells surrounding a central core of massed crystals [61]. The presence of B cells and plasma cells in the tophus corona zone suggests that some form of cognate adaptive immunity could exist in tophi [61].

Recent research has elucidated that neutrophils are required not only to fuel experimental gouty inflammation but also to limit it in vivo [24, 62, 63]. First, at high neutrophil densities achievable in acute gouty arthritis, urate crystals stimulate extracellular trap formation (NETosis) and aggregation [62]. IL-1β is among the stimuli that can induce NETosis. Generally pro-inflammatory, NETosis promotes endothelial damage and thrombosis, in part via release of neutrophil elastase in the extracellular web enriched in chromatin, and by antimicrobial and inflammatory proteins [20, 64]. Complex aggregated NET structures develop in gouty joint fluids and tophi, and aggregated NET formation appears to promote resolution of acute neutrophilic gout-like inflammation in mice. Further, aggregated NET formation may also promote clinically quiescent tophus deposition by trapping and degrading several inflammatory mediators [62]. Conversely, failure of clearance of circulating urate microaggregates in phagocytes has been suggested to promote NETosis, which is toxic to the vasculature [20]. Second, the neutrophil chemotaxin and activator C5a have been reported to induce neutrophil-derived phosphatidylserine-positive microvesicles early in the course of inflammation, including peritonitis induced by urate crystals in mice [24]. Such microvesicles suppress C5a priming of the NLRP3 inflammasome and induction of IL-1β release and neutrophil influx via Mer receptor tyrosine kinase signaling [24]. These effects are shared by synthetic phosphatidylserine-containing liposomes and by neutrophil-derived microvesicles isolated from joint fluids of patients with gouty arthritis [24]. Third, annexin A1 promotes acceleration of neutrophil apoptosis and can combine with acetate, transglutaminase 2, and the inhibitory C-type lectin receptor 12a to promote associated efferocytosis and resolution of the acute phase of experimental gouty inflammation [63, 65, 66].

Increased miR-146a suppresses multiple urate crystal-induced inflammatory responses in vitro, including IL-1β expression [67]. In a study of patients with gout, miR-146a has been reported to be significantly increased in PBMCs between flares compared to controls with normal serum urate, hyperuricemia, and gout patients during acute flares [67]. Moreover, miR-146a expression is broadly detected in tophi [67]. Urate crystals induce miR-146a in cultured mononuclear phagocytic cells [67]. However, tissue levels of this epigenetic transcriptional suppressor of gouty inflammation decrease early in experimental urate crystal-induced inflammation in mice [67].