Objectives The study of the proinflammatory role of uric acid has focused on the effects of its crystals of monosodium urate (MSU). However, little is known whether uric acid itself can directly have proinflammatory effects. In this study, we investigate the priming effects of uric acid exposure on the cytokine production of primary human cells upon stimulation with gout-related stimuli.
Methods Peripheral blood mononuclear cells (PBMCs) were harvested from patients with gout and healthy volunteers. Cells were pretreated with or without uric acid in soluble form for 24 h and then stimulated for 24 h with toll-like receptor (TLR)2 or TLR4 ligands in the presence or absence of MSU crystals. Cytokine production was measured by ELISA; mRNA levels were assessed using qPCR.
Results The production of interleukin (IL)-1β and IL-6 was higher in patients compared with controls and this correlated with serum urate levels. Proinflammatory cytokine production was significantly potentiated when cells from healthy subjects were pretreated with uric acid. Surprisingly, this was associated with a significant downregulation of the anti-inflammatory cytokine IL-1 receptor antagonist (IL-1Ra). This effect was specific to stimulation by uric acid and was exerted at the level of gene transcription. Epigenetic reprogramming at the level of histone methylation by uric acid was involved in this effect.
Conclusions In this study we demonstrate a mechanism through which high concentrations of uric acid (up to 50 mg/dL) influence inflammatory responses by facilitating IL-1β production in PBMCs. We show that a mechanism for the amplification of IL-1β consists in the downregulation of IL-1Ra and that this effect could be exerted via epigenetic mechanisms such as histone methylation. Hyperuricaemia causes a shift in the IL-1β/IL-1Ra balance produced by PBMCs after exposure to MSU crystals and TLR-mediated stimuli, and this phenomenon is likely to reinforce the enhanced state of chronic inflammation.
Statistics from Altmetric.com
Gout is one of the oldest described rheumatic diseases which affects approximately 1% of the world's population1 and reaches 2.5%–3.9% prevalence in developed countries.2 ,3 Gout is characterised by painful, recurrent and initially self-limited attacks of acute inflammation with long-term progression towards chronic tophaceous gout in some patients.4 The biological culprit of gout is represented by monosodium urate (MSU) crystals,5 which are formed during hyperuricaemia and elicit inflammatory events.4 The main proinflammatory cytokine that has been proven to strongly mediate acute gouty inflammation is interleukin (IL)-1β.6–10 MSU crystals have been associated with the activation of the NLRP3 inflammasome.7 This effect is exerted via microtubule-mediated co-localisation of apoptosis-associated speck-like protein containing a caspase recruitment domain (ASC) to the site of inflammasome formation.11 Nevertheless, second signals are equally required to induce pro-IL-1β and synergise with MSU crystals9 ,10 in line with the clinical situation where gouty flares are precipitated in specific environmental situations, despite continuous deposition of MSU crystals in the joint.12
The single factor that is significantly associated with gout susceptibility and represents a necessary cause in gout development is hyperuricaemia.13 This is the elevation of serum uric acid levels above the threshold of 0.36 mM14 when uric acid crystallisation ensues. Uric acid is the end product of purine metabolism in humans and higher primates due to the evolutionary loss of uricase activity, an enzyme that metabolises uric acid to the more soluble product allantoin.15 Recently, it has been shown that the loss of uricase activity is gradually lost during evolution, allowing adaptation to less enzymatic activity and slow rise of uric acid levels in the blood.16 Additionally, several uric acid transporters are involved in the reabsorption of approximately 90% of the total urate that has been filtered via the glomeruli.17 On the one hand, these two mechanisms have led to several hypotheses of evolutionary advantage that might be conferred by higher serum urate levels in primates.18 ,19 On the other hand, they make humans more prone to develop hyperuricaemia and gout. Consistently with uric acid being the major risk factor for gout, the most clear genetic associations with gout susceptibility were obtained for genetic variations in genes encoding urate transporters.20 ,21
Up to now, major lines of research investigating the proinflammatory effects of uric acid have focused on MSU crystal-induced processes.22 However, emerging data suggest that uric acid in a soluble form might also have proinflammatory effects. It has been observed that hyperuricaemic mice have a higher cytokine production upon lipopolysaccharide (LPS) challenge compared with control animals,23 and different studies have shown nuclear factor-κB activation in renal and pancreatic tissue of mice receiving intraperitoneal injections with uric acid.24 ,25
In this study, we hypothesise that uric acid might exert proinflammatory properties by predisposing to MSU crystallisation, tissue deposition and acute inflammation, and also through a direct effect on human primary peripheral blood mononuclear cells (PBMCs). Here, we show that cells originating from patients with gout differ in their cytokine production capacity from control volunteers and that these differences correlate with serum uric acid levels. Furthermore, we describe enhanced proinflammatory cytokine production by primary cells exposed to high uric acid concentrations in vitro and show that this effect is due to downregulation of IL-1Ra, the natural antagonist of the IL-1 receptor type I.
Materials and methods
A detailed Materials and methods version is provided in the online supplementary information of the manuscript.
PBMC isolation and stimulation
Peripheral blood was drawn from healthy volunteers or patients with gout (for basic parameters see online supplementary table S1), after informed consent. PBMCs were separated using Ficoll–Paque and suspended in Roswell Park Memorial Institute (RPMI) culture medium. Cells were stimulated for 24 h with Pam3Cys or C16:0 with or without MSU crystals. Separately, cells were incubated for 24 h with culture medium, uric acid or allantoin (priming). After priming, culture medium was removed and the remaining adherent cells were restimulated with medium, Pam3Cys, LPS with or without MSU crystals.
Cells primed for 24 h uric acid were stained with annexin V-FITC (BioVision) and propidium iodide (PI) (Invitrogen) for assessment of cell death. Fluorescence was measured using Cytomics FC500 (Beckman Coulter).
Cytokine concentrations were determined using ELISA kits for IL-1β, tumour necrosis factor α (TNFα), IL-1α, IL-1Ra (R&D Systems), IL-6 and IL-10 (Sanquin).
Samples stimulated for 4 h were treated with TRIzol Reagent (Invitrogen) for total RNA purification. Isolated RNA was transcribed into complementary DNA using iScript (Bio-Rad), followed by quantitative-PCR using Sybr Green.
Data were analysed using Kruskal–Wallis, Mann–Whitney or Wilcoxon signed rank test according to the number of datasets and experimental design, α<0.05.
Enhanced cytokine production in PBMCs of patients with gout compared with healthy volunteers
To assess the cytokine production of PBMCs of patients with gout compared with healthy controls, 24 h stimulations were performed using MSU crystals alone or in combination with TLR2 ligands Pam3Cys and saturated fatty acid palmitate (C16:0). MSU crystals alone did not induce IL-1β or IL-6 on their own, but significantly increased the effects of Pam3Cys or C16:0 in both patients and controls (figure 1A, B). Nevertheless, the overall cytokine levels observed in PBMCs originating from patients with gout were higher compared with controls (figure 1A, B). Of high interest, a positive relation was observed between serum uric acid levels and ex-vivo IL-1β secretion in cells from patients with gout: patients with hyperuricaemia (>0.36 mM) have a higher response than patients with normouricaemia (<0.3 mM) (figure 1C, D and online supplementary figure S1).
Uric acid pretreatment enhances proinflammatory cytokine production in primary PBMCs of healthy volunteers
To mimic in vitro the effects of hyperuricaemia, cells were exposed to soluble uric acid or were left untreated for 24 h. Subsequently, they were stimulated with MSU crystals in the presence or absence of TLR2 ligand Pam3Cys or TLR4 ligand LPS. Uric acid alone after 24 h priming did not induce detectable IL-1β or TNF concentrations but slightly increased IL-6 (figure 2A–C). Nevertheless, after stimulation, significantly increased cytokine levels were measured in uric acid pretreated cells compared with RPMI pretreated cells, effects that occurred in a dose-dependent (figure 2A–E) and time-dependent manner (see online supplementary figure S2A). To assess mRNA modifications, after the first 24 h priming with uric acid, cells were restimulated for another 4 h. Uric acid pretreated cells exhibited higher relative IL-1β mRNA compared with control conditions (figure 2F) and this was also observed in a dose-dependent manner (see online supplementary figure S2B). TNF and IL-6 mRNA levels were also enhanced (see online supplementary figure S2C, D). This effect of uric acid was most prominent at the highest concentration (50 mg/dL). Nevertheless, lower uric acid levels show the same tendency towards increased IL-1β production (see online supplementary figure S3A). The effect on IL-1β was replicated in purified monocyte cell suspensions isolated from PBMCs (see online supplementary figure S4).
Uric acid specifically downregulates the production of the anti-inflammatory cytokine IL-1Ra
To further decipher the mechanism of uric acid induction of proinflammatory cytokines, the effect on anti-inflammatory cytokines IL-1 receptor antagonist (IL-1Ra) and IL-10 was investigated. Of interest, the IL-1Ra concentrations were found to be significantly decreased in PBMCs treated with uric acid compared with medium pretreated cells (figure 3A), whereas IL-10 levels were not affected (figure 3B). This surprising effect of uric acid on IL-1Ra production was observed at the level of transcription after 4 h uric acid exposure (figure 3C) and after 24 h priming followed by 4 h stimulation (figure 3D). This implies a specific modulation of uric acid on IL-1 natural inhibition. IL-1Ra downregulation was also obvious at lower uric acid doses (see online supplementary figure S3B) and was present in selected monocytes (see online supplementary figure S4). Addition of exogenous IL-1Ra (recombinant protein) dose-dependently reversed the effects of uric acid (see online supplementary figure S5).
Modulation of immune responses by uric acid is independent of allantoin, myeloperoxidase inhibition or cell death
To determine whether the effect of uric acid is specific or whether it can also be induced by related metabolites, allantoin was used as control in the same experimental conditions as uric acid. Uric acid induced significantly higher IL-1β production and strongly suppressed IL-1Ra, which was not seen in allantoin-treated PBMCs (figure 4A and online supplementary figure S6A, B). Myeloperoxidase (MPO) inhibitor, 4-ABH, was used to assess whether oxidation products resulting after MPO-mediated oxidation of uric acid might play a role in the effects observed. However, the effect of uric acid on cytokine production was not modified by MPO inhibition (figure 4B and online supplementary figure S6C, D). Moreover, annexinV/PI staining after uric acid treatment did not show differences in cell death due to uric acid exposure (figure 4C).
PBMCs of patients with reveal to be less responsive to uric acid priming for IL-1β induction when compared with PBMCs from healthy controls
Furthermore, we investigated whether uric acid priming is a possible mechanism explaining the higher cytokine production observed in cells of patients with gout (as described in figure 1). For testing this hypothesis, the experimental setup of 24 h uric acid priming and 24 h stimulation of PBMCs was replicated in a larger cohort of healthy controls and patients with gout aiming at overruling bias due to cytokine variation. Indeed, in consistency with previous data shown in figure 1, the absolute cytokine concentrations determined in cells from patients with gout were significantly higher than in controls (white bars, figure 5A). The enhancement of IL-1β (figure 5A) and suppression of IL-1Ra (figure 5B) due to uric acid were observed in both groups. However, the degree of IL-1β induction by uric acid was significantly lower in patients with gout, as observed by relative induction of IL-1β by uric acid (figure 5C). CD14+ monocytes of patients with gout also exhibited higher steady-state IL-1β mRNA levels than healthy controls in unstimulated conditions (figure 5D), suggesting that cells of patients with gout might have encountered factors in vivo inducing a facilitated state for IL-1β production. IL-1Ra production was significantly decreased upon uric acid priming, with a lower basal level of IL-1Ra in unstimulated cells from patients with gout compared with controls (figure 5B). However, the level of IL-1Ra inhibition by uric acid in patients with gout does not significantly differ from that of controls (figure 5C).
Pharmacological inhibition of histone-modifying enzymes reverse the priming effect of uric acid
Uric acid primes PBMCs for enhanced cytokine production at the transcription level, and we hypothesised that epigenetic modifications known to influence cytokine production, such as histone methylation and acetylation, are involved. To investigate the role of epigenetic histone modifications on uric acid-induced priming that results in aggravated IL-1β, histone methyltransferase inhibitor, MTA (5′-deoxy-5′-methylthio-adenosine), and histone acetyltransferase inhibitor, EGCG (epigallocatechin-3-gallate), were coincubated with uric acid. Of high interest, upon restimulation, the enhancement of IL-1β by uric acid was reversed by MTA (figure 6A) while staying unmodified by acetyltransferase inhibitor, EGCG (figure 6B).
In this study, we have revisited the hypothesis of uric acid acting as a possible proinflammatory agent, independently of its precipitated form in MSU crystals. As previously reported,26 here we confirm the finding that ex vivo stimulated PBMCs from patients with gout reveal enhanced inflammatory cytokine production and we show that this is linked with serum uric acid levels of the donors.
Using primary human PBMCs as well as purified monocytes, we have performed an in vitro stimulation protocol mimicking hyperuricaemia for 24 h of initial treatment of the cells (priming) followed by restimulation with TLR ligands and MSU. These experiments revealed that significantly increased levels of proinflammatory cytokines (up to threefold more IL-1β) were produced by uric acid primed cells.
We next studied the molecular level at which uric acid exerts its effect: remarkably, we show that IL-1β production is enhanced at both intracellular and extracellular compartments, suggesting that enhanced secretion is probably not the explanation of these effects, but transcriptional upregulation of cytokines is the likely mechanism. Together with IL-1β, other proinflammatory cytokines like IL-6 and TNF were also increased, probably at least in part secondarily to IL-1β induction. However, most interestingly, a specific inhibition of IL-1Ra production was observed after uric acid pre-exposure, both at protein and at transcription level. The increased IL-1β and decreased IL-1Ra levels were similarly observed in monocytes purified from the PBMC suspension. This finding is of high importance, as IL-1Ra is the natural inhibitor of IL-1RI, and it is known to be upregulated in parallel to IL-1β production in a feedback loop aiming to control IL-1β-driven and IL-1α-driven inflammation.27 To our knowledge, this is a unique finding of a signal that does not upregulate both IL-1β and IL-1Ra in the same direction, but rather decreases IL-1Ra production. Therefore, this promotes IL-1β-induced IL-1β production without the natural antagonist counterbalancing this pathway.
Exogenous IL-1Ra addition to uric acid-treated samples during the first 24 h of priming re-established the IL-1Ra levels during the first 24 h. This, however, did not have lasting effects for the second 24 h stimulation on IL-1β (see online supplementary figure S5A, B). Adding IL-1Ra, also at the moment of restimulation, dose-dependently restored IL-1Ra and diminished IL-1β (see online supplementary figure S5C, D). This proves that IL-1Ra downregulation by uric acid is an important part of the mechanism of IL-1β enhancement, not only an associated phenomenon with no functional significance.
Cells isolated from patients with gout were found to be less prone to potentiate IL-1β induction after uric acid pre-exposure, compared with healthy volunteers. This effect was due to an already enhanced state of IL-1β production, most likely due to the effects of hyperuricaemia already present in these patients. In line with this, steady-state IL-1β mRNA levels appear to be higher in patients with gout than in controls and basal levels of IL-1Ra in unstimulated cells are lower than those observed in healthy volunteers. These observations suggest that the priming effects of soluble uric acid, which enhance IL-1β while decreasing IL-1Ra, are biologically relevant and present in vivo in patients with gout.
In experiments using PBMCs of patients with gout and healthy controls for in vitro priming, uric acid was not present during the time of restimulation with TLR ligands and/or MSU crystals. This implies a long-term consequence of uric acid exposure that induces an effect reminiscent of non-specific immunological memory (‘trained immunity’) of monocytes observed after infections and vaccinations.28 ,29 Long-term adaptive effects on innate immune cells resulting in enhanced cytokine production have been shown to be mediated by epigenetic reprogramming of cells.28 ,29 The increased long-term H3K4me1 induced by inflammatory stimuli has been reported to increase cell responsiveness, and these epigenetic marks have been termed ‘latent enhancers’.30 Because uric acid exposure induces a long-term increase in IL-1β gene transcription (accompanied by less IL-1Ra), we hypothesised that this effect may be mediated by histone modifications. Using pharmacological inhibitors of either histone methylation or acetylation, two major histone modification marks with functional consequences on gene expression, we show that uric acid effects were abolished when histone methyl transferase inhibitor was used. This is the first indication that indeed uric acid might exert effects at the epigenetic level and that these effects could have long-term consequences on the individual cytokine profile. Future studies are warranted to assess the role of hyperuricaemia on the specific epigenetic marks and complete functional signature associated with high urate exposure.
A link between soluble uric acid and induction of inflammation has also been suggested by previous reports where uric acid has been identified as a danger signal released in the context of cell death–inducing inflammation,31 or adaptive immunity.32 Although the initial event in these studies is also the release of uric acid and formation of a high urate environment, the main inflammatory role in this context is still in fact attributed to MSU crystals.33 In contrast, data linking uric acid with higher LPS responses,23 and NF-κB activation,24 ,25 were obtained in mice receiving intraperitoneal uric acid injections. In this study, despite the high uric acid concentrations used (approximately eight times higher than the threshold for hyperuricaemia) we were unable to detect MSU crystals being formed over the 24 h of cell culture in the presence of uric acid, arguing for the stability of the solutions used for the time and conditions of our experimental setup. Moreover, MSU crystals and soluble uric acid have totally opposite effects on IL-1Ra levels: as shown in figures 3A and 5B, MSU crystals increase the levels of IL-1Ra secreted upon MSU stimulation alone, or in synergism with other ligands, while soluble urate significantly downregulates IL-1Ra.
The high concentrations of uric acid used can be a limitation of this study. Nevertheless, we provide evidence (see online supplementary figure S3) that the same cytokine profile is observed also at lower uric acid levels, in the range of clinical hyperuricaemia and within the solubility limits of uric acid. The rationale of using a high dose was to have the ability to best observe these significant effects in an in vitro setting, allowing further experimental intervention. In figure 5, we show data supporting that similar features of uric acid priming were also present in patients with gout, sustaining the biological relevance of this finding. Further studies using in vivo models are needed to demonstrate this in more detail.
This mechanism of uric acid-induced modulation of proinflammatory cytokines is of high relevance also for other metabolic diseases. There is a large body of evidence that associates uric acid with metabolic syndrome, atherosclerosis, hypertension, type 2 diabetes and chronic kidney disease.34–38 In this context, mechanistic data on the proinflammatory effects of uric acid are likely to turn into a potential link between the diseases of modern society associated with chronic low-grade inflammation.
In conclusion, in this study, we propose a mechanism through which high uric acid concentrations mediate the metabolic modulation of inflammatory responses by facilitating IL-1β production in PBMCs. We show that a mechanism for the amplification of IL-1β is via the unique inhibitory effect on IL-1Ra that would normally counterbalance the effects of IL-1 and the IL-1β autoinduction loop. This effect is likely to be epigenetically mediated by histone marks that modify the degree of transcription for cytokine genes. Consequently, patients having hyperuricaemia could be at risk of increased reactivity of cells upon encounter of acute inflammatory stimuli (infectious or sterile danger signals) and this might induce enhanced states of (auto)inflammation. Thus, we provide here the first evidence that uric acid represents a silent modulator of cytokine production in primary human cells. This pinpoints to the role of metabolic triggers on the inflammatory properties of the circulating cells and could represent a relevant link for understanding the pathogenesis of gout and other metabolic diseases with inflammatory components.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Handling editor Tore K Kvien
Contributors TOC, MCPC, MO, HL, HT-D: designed and performed the experiments. TOC: analysed the data and wrote the initial manuscript. MCPC, MO: revised the manuscript. MGN: conceived the experiments and revised the manuscript. TLJ: diagnosed and recruited the patients and revised the manuscript. LABJ: conceived and designed the study and revised the manuscript. All authors approved the final version.
Funding This study was supported by a grant from the Dutch Arthritis Foundation (NR 12-2-303). MGN was supported by a Vici grant and a ERC Starting Grant (310372).
Competing interests None.
Ethics approval Ethics Committee of the Radboud University Medical Center Nijmegen.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.