The Long Pentraxin 3 Contributes to Joint Inflammation in Gout by Facilitating the Phagocytosis of Monosodium Urate Crystals

Gout is the most common form of inflammatory arthritis worldwide among men and postmenopausal women. Its incidence and prevalence have increased mainly because of lifestyle changes, such as a higher alcohol and purine-rich diet consumption, and it is closely associated to the metabolic syndrome (1, 2). Joint inflammation in gout occurs because of the deposition of monosodium urate (MSU) crystals predominantly in peripheral joints and surrounding tissues and mainly in individuals with chronic hyperuricemia. Acute gout attacks, although characterized by a self-limited inflammation, are extremely painful and can lead to joint disability. Prolonged deposition of MSU crystals can result in irreversible joint damage with bone erosion and development of s.c. tophi (3).

The initial events of MSU crystal–triggered inflammation occur after the contact and phagocytosis of the crystals by synovial phagocytes. The internalization of MSU crystals results in the assembly of NACHT, LRR, and PYD domains–containing protein 3 (NLRP3)/apoptosis-associated speck-like protein containing a CARD (ASC)/caspase-1 inflammasome that culminates in the release of the mature form of IL-1β (4), a cytokine with a crucial role in gouty arthritis. IL-1β promotes the production of different chemoattractants that cause early neutrophil swarm to the joint and lead to joint inflammation, damage, and pain. However, the mechanisms underlying the recognition of MSU crystals by phagocytes have not yet been completely elucidated. Some studies have demonstrated that MSU crystals bind to plasma proteins, such as complement components IgG and IgM (5–7). The opsonisation of MSU crystals by these molecules enables direct contact with their receptors on the leukocyte surface, as demonstrated in neutrophils through CR3 (CD11b/CD18) and FcγRIIIB (CD16), which bind crystal-bound iC3b and IgG, respectively (8). However, the receptors involved in the phagocytosis of MSU crystals by mononuclear phagocytes need to be demonstrated.

The soluble pattern recognition molecules, including complement system, natural Abs, and pentraxins constitute the humoral arm of the innate immune response (9). Pentraxins are a superfamily of evolutionarily conserved multimeric proteins, which is divided into short (C-reactive protein and serum amyloid P component) and long pentraxins. Pentraxin 3 (PTX3), the prototype of the long pentraxin family, is produced and released by a variety of cell types, including phagocytes, dendritic cells, fibroblasts, and endothelial cells under different stimuli, such as LPS, IL-1, and TNF-α (10–12). PTX3 can interact with a variety of pathogens (e.g., Aspergillus fumigatus and Pseudomonas aeruginosa) and has opsonic activity facilitating their phagocytosis (13–15) through interaction with FcγR, which have been identified as pentraxin receptors (16). Moreover, PTX3 is involved in the pathogenesis of acute and chronic sterile inflammatory diseases, including rheumatoid arthritis (RA) and ischemia/reperfusion (17, 18). However, the involvement of PTX3 in gout has not yet been described.

The present study was designed to investigate the role of PTX3 in gouty inflammation. The analysis of samples obtained from gouty patients demonstrated an increase of PTX3 levels and a positive correlation of this protein with IL-1β. Using genetically deficient mice, we demonstrate that PTX3 leads to an increase of MSU crystal phagocytosis by mononuclear phagocytes by a mechanism dependent on FcγR. This event culminates in increased inflammasome activation and IL-1β secretion by these cells. Altogether, our results suggest that PTX3 is an important molecule that facilitates MSU crystal–driven inflammation in gouty arthritis.

Two Brazilian groups of patients diagnosed with acute gout flares (<48 h after gout attack), according to the 2015 Gout Classification Criteria, were used in this study: one from Rio de Janeiro (A) and other from Ribeirão Preto (B). One Brazilian group (from Ribeirão Preto) of patients diagnosed with osteoarthritis of the knee, according to the Classification of American Rheumatism Association, was also used in this study. Synovial fluid collected from the knee and blood samples were collected in EDTA tubes, centrifuged (2.000 × g for 10 min), and the supernatant kept at −20°C. A fresh sample of synovial fluids from gout and osteoarthritis underwent MSU crystal identification by compensated polarized light microscopy. All patients provided informed consent to participate in the study, which was approved by Comitê de Ética em Pesquisa under protocol number 1.297.041, and by Certificado de Apresentação para Apreciação Ética (Comissão Nacional de Ética em Pesquisa) under protocol number 50373815.4.0000.5259 in Rio de Janeiro and 4971/2012 in Ribeirão Preto.

The experiments with mice were performed in two different laboratories: Italy and Brazil. Eight- to twelve-week-old mice were housed in a controlled environment and had free access to commercial chow and filtered water. All mice used were on C57BL/6 genetic background. All efforts were made to minimize the number of animals used and their suffering. Italy: Ptx3-deficient mice were generated as described (13). FcRγ-deficient mice were purchased from The Jackson Laboratory, Bar Harbor, ME. All colonies were housed and bred in the specific pathogen-free animal facility of Humanitas Clinical and Research Center in individually ventilated cages. Wild-type (WT) mice were obtained from Charles River Laboratories, Calco, Italy. Procedures involving animals’ handling and care were conformed to protocols approved by the Humanitas Clinical and Research Center (Rozzano, Milan, Italy) in compliance with national (Decreto Legislativo n.116, Gazzetta Ufficiale, suppl. 40, February 18, 1992; Decreto Legislativo n.26, March 4, 2014) and international law and policies (European Economic Community Council Directive 2010/63/EU, official journal L 276/33, 22-09-2010; National Institutes of Health Guide for the Care and Use of Laboratory Animals, US National Research Council, 2011). The study was approved by the Italian Ministry of Health (approval 71/2012-B). Brazil: FcγRIIB- and FcγRIII-deficient mice were generated as previously described. These mice were a kind gift of Dr. J. M. Penninger and were bred in our facilities. WT mice were purchased from the animal facility of Federal University of Minas Gerais, Brazil. The experiments were made according to the ethical principles in animal experimentation of our institution, and the experimental protocol was approved by the Ethics Committee in Animal Experimentation of the University (protocol 35/2016-Comissão de Ética no Uso de Animais/Universidade Federal de Minas Gerais).

Mice were placed under anesthesia (80:15 mg/kg ketamine:xylazine, i.p.; Syntec, São Paulo, Brazil) and were injected with MSU crystals (100 μg/cavity) into the tibiofemoral knee joint. MSU crystals were prepared as previously described (19). Inflammatory parameters were evaluated at 6 or 15 h after injection of MSU crystals. Mice were euthanized in a CO₂ chamber, followed by cervical dislocation, and the articular cavity was washed with PBS containing 3% BSA for leukocyte counts. The number of cells was determined in a Neubauer chamber after staining with Turk solution. Differential counts were performed by Shandon CytoSpin III (Thermo Shandon, Frankfurt, Germany) preparations by evaluating the percentage of each leukocyte on a slide after staining with May-Grunwald–Giemsa (20). Periarticular tissue was removed from the joints for cytokine evaluation by ELISA.

Periarticular tissues were collected from joints and homogenized with PBS containing antiproteases (21). Samples were centrifuged, and the supernatant was used to quantify PTX3, IL-1β, and CXCL1. Plasma and synovial fluid from patients were used for determination of PTX3 and IL-1β levels. A commercially available ELISA was used in accordance with the manufacturer’s instructions (R&D Systems, Minneapolis, MN).

One hundred microliters of mouse whole blood were incubated with MSU crystals (150 μg/ml) for 15 min at 37°C. Previously, these crystals were incubated or not with recombinant PTX3 (rPTX3) for 1 h. We tested different concentrations of rPTX3 and presented those results that had significance compared with the groups without rPTX3. Human cells were isolated with a density gradient centrifugation in accordance with the manufacturer’s instructions (Histopaque 1077 and Histopaque 1119; Sigma-Aldrich, St. Louis, MO). PBMC were plated for 2 h to allow the adherence, and then these cells or isolated neutrophils (5 × 10⁵) were stimulated as murine cells described above. Purity was checked using cytospin preparation, and all samples used exceeded 90% purity. After stimulation with MSU crystals, samples were placed on ice to block phagocytosis, and red cells were lysed by adding cold ammonium chloride lysis solution (pH 7.2). As a control of phagocytosis, cells were preincubated with 20 μM cytochalasin D (CytD) (Sigma-Aldrich) for 1 h before the stimulation with MSU crystals. Human monocytes were stained for CD14. Murine samples were stained for monocytes (CD45⁺CD11b⁺Ly6C^high) and neutrophils (CD45⁺CD11b^high Ly6G^high) (eBiosciences, BD Pharmingen). Samples were then fixed with 1% paraformaldehyde. The phagocytosis of MSU crystals was determined by analysis of side scatter change in flow cytometry, as previously described (22). The values were represented by relative percentage to the group of cells incubated only with MSU crystals. The analyses were performed with FlowJo software (Tree Star).

For the detection of active caspase-1, FAM FLICA Caspase-1 Assay Kit (ImmunoChemistry) was used. Monocytes isolated from human whole blood (as previously described) were primed with LPS (100 ng/ml) for 1 h and then stimulated with MSU crystals (MSU 150 μg/ml) for 6 h. Cells were stained with FLICA in accordance with the manufacturer’s instructions and also for CD14 for identification of monocytes. The analyses were performed with FlowJo software (Tree Star).

Mice received an i.p. injection of a 3% thioglycolate solution, and macrophages were collected by peritoneal lavage 4 d later. Cells were plated for 2 h, and later, nonadherent cells were removed. Cells were cultured in RPMI medium complemented with 10% FCS, penicillin/streptomycin, and l-glutamine and kept in a humidified incubator at 37°C with 5% CO₂ overnight. Cells were primed for 1 h with LPS (100 ng/ml) (Sigma-Aldrich) and then stimulated with MSU crystals (150 μg/ml) for 6 h. The supernatant was used for IL-1β detection by ELISA.

All results are presented as the means ± SEM. Results were tested for normality, and differences between groups were evaluated using one-way ANOVA followed by Newman–Keuls posttest, Student t test, Mann–Whitney U test, or Pearson correlation as specified. The level of significance was set at p < 0.05. Statistics were calculated with GraphPad Prism software version 6 (GraphPad Software, La Jolla, CA).

To address the relevance of PTX3 in gout, we first evaluated the concentration of PTX3 in plasma and synovial fluid of patients diagnosed with acute gout. In group A, there was an increase of PTX3 levels in plasma of patients with gout compared with healthy volunteers. Interestingly, the levels of PTX3 were higher in synovial fluid as compared with plasma in those patients (Fig. 1A). In group B, the concentration of PTX3 in synovial fluid was much higher in gout patients compared with osteoarthritic (OA) patients (Fig. 1B). Because IL-1β is a key proinflammatory marker in gout (23–25), we quantified this cytokine in those samples. In both groups, IL-1β was detected in synovial fluid of gout patients (data not shown). Importantly, there was a positive correlation between PTX3 and IL-1β in synovial fluid during acute gout in group A (Fig. 1C) and group B (Fig. 1D). However, there was no positive correlation between PTX3 and IL-1β in plasma of group A (data not shown). Collectively, the increased level of PTX3 during gout flares and its association with IL-1β indicate a participation of PTX3 in the early events of gouty inflammation.

To investigate whether PTX3 contributed to the inflammatory response observed in gout, we used an experimental model of acute gout in mice through the injection of MSU crystals into the tibiofemoral joint (20). In this model, the peak of inflammation ranges between 6 and 15 h after the insult. Injection of MSU crystals caused an increase of PTX3 levels in inflamed joints 6 h postinjection when compared with saline-injected joints, and this increase was even greater at the 15th hour (Fig. 2A). MSU crystals also caused an increase of IL-1β, CXCL1, and influx of neutrophils to the joint compared with control group (Fig. 2B–D). Ptx3-deficient mice had a reduction in these inflammatory parameters following the injection of MSU crystals as compared with WT mice (Fig. 2B–D). Of note, the injection of rPTX3 alone was not able to cause an inflammatory response in the joint (data not shown). Altogether, these data strengthen the contribution of PTX3 to MSU crystals–induced inflammation.

PTX3 is known to interact with different microorganisms. It has opsonic activity and contributes to pathogen phagocytosis and clearance (13–15). Thus, we investigated whether PTX3 could be involved in the phagocytosis of MSU crystals. To set up an assay of phagocytosis of MSU crystal (Supplemental Fig. 1) and its consequence on IL-1β release, we first treated murine whole blood with CytD, an inhibitor of actin polymerization that impairs phagocytosis, before MSU crystal stimulation and performed the analysis of phagocytosis of MSU crystals by flow cytometry. CD45⁺CD11b⁺Ly6C^high monocytes phagocytosed MSU crystals, and this was significantly reduced after CytD treatment (Fig. 3A). MSU crystals stimulated the release of IL-1β by LPS-primed peritoneal macrophages. Incubation with CytD showed decreased release of IL-1β levels in the supernatant, confirming that phagocytosis of MSU crystals is necessary for IL-1β maturation and release (Fig. 3B). This effect was indeed dependent on the phagocytosis of MSU crystals because CytD did not reduce the release of IL-1β when cells were incubated with ATP, a well-known NLRP3 inflammasome activator (data not shown). Ptx3-deficient monocytes showed defective phagocytic activity when compared with WT cells. However, the incubation of rPTX3 in whole blood increased the phagocytosis of MSU crystals by monocytes in both WT and Ptx3-deficient cells (Fig. 3C). Peritoneal macrophages incubated with rPTX3 also increased the release of IL-1β after MSU crystal stimulation in both WT and Ptx3-deficient cells (Fig. 3D). Of note, macrophages stimulated with rPTX3 alone did not produce IL-1β (data not shown). CD45⁺CD11b^highLy6G^high neutrophils from Ptx3-deficient mice had reduced phagocytosis of MSU crystals when compared with WT neutrophils, a phenotype that was rescued by the addition of rPTX3 (Supplemental Fig. 2A).

A similar profile was also observed in human cells. The presence of rPTX3 increased the phagocytosis of MSU crystals by blood CD14⁺ monocytes from healthy donors (Fig. 4A). Also, there was an increase of IL-1β in the supernatant of cells incubated with rPTX3 and MSU crystals (Fig. 4B). As the cleavage and release of IL-1β under MSU crystal stimulation is dependent on NLRP3/ASC/caspase-1 (20), we measured the cleavage of caspase-1 in CD14⁺ monocytes. As shown in Fig. 4C, the incubation with rPTX3 increased MSU crystal–dependent caspase-1 activation when compared with cells treated only with MSU crystals. Along the same line, isolated human neutrophils incubated with MSU crystals in the presence of rPTX3 showed greater phagocytosis of the MSU crystals (Supplemental Fig. 2C). Thus, PTX3 facilitates the phagocytosis of MSU crystals and consequently favors the production of IL-1β in mononuclear phagocytes.

It has been reported that the interaction of PTX3 with FcγRs mediates part of the biological activities of PTX3 (16). Thus, we investigated the contribution of FcγRs to PTX3-induced MSU crystal phagocytosis and their role in gouty inflammation. Monocytes from FcγR-deficient mice had lower phagocytosis rate in comparison with cells from WT mice. Interestingly, the addition of rPTX3 to MSU crystal–stimulated FcγR^−/− cells did not increase the phagocytosis of the crystals, in contrast with what we observed with WT cells (Fig. 5A). Next, we tested this response in cells from FcγRIIB (an inhibitory FcγR)-deficient and FcγRIII (a stimulatory FcγR)-deficient mice. Monocytes from both FcγRIIB^−/− and FcγRIII^−/− mice showed the same levels of phagocytosis of MSU crystals as WT mice. However, cells from WT and FcγRIIB^−/−, but not from FcγRIII^−/−, mice showed increased phagocytosis after the incubation with rPTX3 (Fig. 5B). We found a similar profile when analyzing IL-1β release (Fig. 5C). WT-, FcγR^−/−-, FcγRIIB^−/−-, and FcγRIII^−/−-primed macrophages stimulated with MSU crystals released more IL-1β levels in the supernatant (Fig. 5D) than nonstimulated cells. However, the incubation with rPTX3 increased the amount of IL-1β further only in WT and FcγRIIB^−/− cells but did not affect IL-1β release by FcγR^−/− and FcγRIII^−/− cells (Fig. 5C, 5D).

Again, this phenotype was also observed in neutrophils. FcγR^−/− neutrophils had decreased phagocytosis of MSU crystals compared with WT cells, whereas FcγRIIB^−/− and FcγRIII^−/− neutrophils showed a similar phagocytosis rate. However, the addition of rPTX3 increased the phagocytosis rate only in WT and FcγRIIB^−/− neutrophils, keeping the control levels in FcγR^−/− and FcγRIII^−/− neutrophils (Supplemental Fig. 2A, 2B).

To evaluate the relevance of these receptors in gouty inflammation, FcγR^−/−, FcγRIIB^−/−, and FcγRIII^−/− mice were injected with MSU crystals, and the inflammatory parameters were evaluated 15 h later. As shown in Fig. 6, IL-1β, CXCL1, and neutrophils were reduced in FcγR^−/− and FcγRIII^−/− when compared with WT mice. In contrast, there was no significant alteration of these parameters between WT and FcγRIIB^−/− (Fig. 6). Altogether, these results suggest the involvement of the stimulatory FcγR FcγRIII in mice in PTX3-induced MSU crystal phagocytosis and joint inflammation.

Several mechanisms involved in the pathogenesis of gout and recognition of MSU crystals have been discovered, including MSU crystal–triggered inflammasome activation and the contribution of different molecules and cell types to joint inflammation, damage, and pain. This knowledge has potentially opened new possibilities for the therapy of gout. In the current study, we investigated the contribution of the soluble pattern recognition molecule PTX3 to gouty inflammation. The main results of this study can be summarized as follows: 1) high levels of PTX3 were found in plasma and synovial fluid of gouty patients during the acute phase of the disease with a positive correlation with IL-1β in the synovial fluid; 2) Ptx3-deficient mice showed diminished joint inflammation following the injection of MSU crystals; 3) PTX3 potentiated the phagocytosis of MSU crystals by human and murine cells, promoting an increase of inflammasome activation and IL-1β production; 4) the enhancement by PTX3 of the phagocytosis of MSU crystals and joint inflammation was dependent on FcγR.

PTX3 is an acute phase protein produced by different cells and body tissues. In the inflammatory context, a dual function of PTX3 has been proposed, either conferring host resistance against pathogens or contributing to immunopathogenesis of diseases (26). Circulating PTX3 is barely detectable in normal conditions but increases rapidly in a range of pathological conditions, including sepsis, A. fumigatus infections, and acute myocardial infarction (27–29). PTX3 is also expressed at high levels in synovial fluid during early joint inflammation in patients with systemic juvenile idiopathic arthritis (30). Moreover, synoviocytes and synovial fluid of RA patients displayed higher levels of PTX3 (17). Interestingly, a recent study demonstrated higher concentration of PTX3 in synovial fluid from RA patients testing positive for anticitrullinated protein Ab and rheumatoid factor when compared with seronegative patients. Moreover, there was a positive correlation between increased levels of (synovial fluid) PTX3 and the severity of the disease (31). In our population of gouty patients, we demonstrated an increase of PTX3 in plasma and synovial fluid during the acute phase when compared with healthy volunteers or OA patients. In addition, we showed a positive correlation between the levels of PTX3 and the amount of IL-1β in the synovial fluid of patients during the acute phase. Of note, the concentration of PTX3 was higher in synovial fluid than blood, suggesting local production of the molecule.

During the acute phase of gout, there is an elevated production and maturation of the cytokine IL-1β, a key agent for gout disease (4, 32). Because of relevance to gout, clinical trials with drugs that interfere on IL-1β–IL-1R pathway have been conducted, including rIL-1Ra (Anakinra), IL-1Trap (Rilonacept), and anti–IL-1β (Canakinumab), which provide rapid and sustained pain relief in patients with acute gouty arthritis and have revealed an impressive and sustained reduction in recurrent attacks of gout (33–36). IL-1β is a potent inducer of PTX3 production and release by a variety of cell types, including mononuclear phagocytes, neutrophils, fibroblasts, endothelial cells, synovial cells, and chondrocytes (37), which are active cells during joint inflammation. Thus, the presence of PTX3 in the synovial fluid of gouty patients could represent an indirect marker of this disease or could have a direct contribution to joint inflammation. In this study, we demonstrated that Ptx3-deficient mice have reduced inflammation following the injection of MSU crystals, suggesting an important role of this molecule in the development of gout.

The precise mechanisms by which innate immune molecules lead to the phagocytosis of MSU crystals are still not clear, although it is known that different serum factors can bind the crystal. MSU crystals isolated from sites of acute gouty inflammation are typically coated with Igs, mostly IgG isotype (5, 38). This interaction facilitates the engulfment of MSU crystals by macrophages and leads to the production of superoxide anion by phagocytes (39). There is evidence that MSU crystals interact with FcγRIIIB (CD16) in association with the CD11b/CD18 integrin complex in human neutrophils, whereas the receptor FcγRIIA (CD32) does not appear to play a direct role in mediating this effect (8). Importantly, FcγRs have also been identified as pentraxins receptors (16). Moreover, it is known that both MSU crystals and PTX3 bind to C1q, the first subcomponent of the C1 complex of the classical pathway of complement activation (40, 41). In this way, C1q could function as a bridge orchestrating the interaction between MSU crystals and PTX3, although further studies are necessary to clarify the nature of this interaction. It is known that PTX3 enhances phagocytosis of uropathogenic Escherichia coli (42) and A. fumigatus conidia (15) by human neutrophils. However, the contribution of PTX3 to MSU crystal phagocytosis and the involvement of FcγRs in this process by mononuclear phagocytes had not yet been demonstrated. In this study, we demonstrated that PTX3 facilitates the phagocytosis of MSU crystals by both human and mouse monocytes and neutrophils in a mechanism dependent on FcγRs in the murine context.

Humans and mice possess two classes of FcγRs: the activating and inhibitory receptors based on their cytoplasmic tyrosine-containing motifs. Both species possess only one inhibitory FcγR, FcγRIIB, which is expressed on B cells, dendritic cells, and basophils in both species, and found additionally on monocytes, macrophages, and all granulocyte populations in mice (43). In contrast, activating FcγRs have remarkable differences between both species. Although murine-activating FcγRs are represented by FcγRI, FcγRIII, and FcγRIV, the human activating FcγRs are composed by FcγRI, FcγRIIA, FcγRIIC, and FcγRIIIA (44). Based on the similarity of the extracellular portion sequence, murine FcγRIV is the ortholog of human FcγRIIIA, and murine FcγRIII is more closely related to human FcγRIIA; however, some differences between them are observed. Human FcγRIIA, for example, contains an intrinsic intracellular ITAM, whereas murine FcγRIII is noncovalently linked to the ITAM-containing FcγR adapter. In addition, there is an exclusive human receptor, FcγRIIIB, that is only expressed on neutrophils (45). Thus, the extrapolation of data regarding these receptors from animal studies to the human system must be very careful. In this study, cells obtained from mice deficient for FcγR had reduced phagocytosis when compared with WT cells. The addition of rPTX3 did not increase the engulfment of the crystals as it happened in WT cells. A similar profile was observed in murine FcγRIII, but not in FcγRIIB-deficient cells, which responded equally to WT cells. However, cells from FcγR-deficient mice showed more pronounced response than FcγRIII-deficient cells, indicating that FcγRI and/or FcγRIV could also have a role in this process. FcεRI was not included in this study because it is the high-affinity receptor for the Fc region of IgE, which is not related to the MSU crystal–induced inflammation. Thus, our results clearly show the dependence of FcγR and specifically FcγRIII to enhancement of phagocytosis of MSU crystals by PTX3. The relevance of specific human FcγR classes in this context should be studied further in the future.

In conclusion, the results reported in this study indicate that PTX3 plays an important role in gouty inflammation by promoting the phagocytosis of MSU crystals via FcγRs that culminates in increase of caspase-1 activation and IL-1β secretion. PTX3 seems to be a very important mediator in the initial process of inflammation in gout and could be a target to control excessive inflammation induced by the deposition of MSU crystals in the joint.

The authors have no financial conflicts of interest.