It has long been recognized that hyperuricemia does not inevitably lead to acute gout (1). In patients with gout, monosodium urate monohydrate (MSU) crystals are readily detectable in asymptomatic joints (2) and during the intercritical phase (3). In contrast with joints of patients with acute gout, which have intense neutrophilic infiltration of the synovium, asymptomatic joints of patients with gout have a predominantly mononuclear cell infiltrate (4). Moreover, crystals can be detected within the cytoplasm of mononuclear cells in the absence of symptomatic gout (5). These observations suggest that cells of the monocyte/macrophage lineage might play a beneficial role in maintaining asymptomatic hyperuricemia.
A beneficial role of monocyte/macrophages, however, is difficult to reconcile with the fact that monocytes, when challenged with urate crystals, secrete tumor necrosis factor α (TNFα) (6, 7), interleukin 1β (IL-1β) (8), IL-6 (9), and IL-8 (10). This in turn activates expression of endothelial cell adhesion molecules (E-selectin, intercellular adhesion molecule 1, and vascular cell adhesion molecule 1), which leads to secondary neutrophil recruitment to sites of crystal deposition (11–13). In order to resolve these apparently disparate observations, we hypothesized that the state of differentiation of mononuclear phagocytes may determine how they respond to urate crystal uptake. Evidence in support of this idea has been demonstrated in a study of mouse monocyte/macrophage cell lines, which showed that incompletely differentiated monocytic cell lines synthesized TNFα and activated endothelial cells upon challenge with urate crystals, whereas well-differentiated macrophagic lines did not (14).
Here, we extend this hypothesis to humans and demonstrate that, in contrast to freshly isolated monocytes, in vitro–differentiated macrophages can ingest MSU crystals in the absence of concomitant proinflammatory cytokine synthesis, endothelial cell activation, or secondary neutrophil recruitment. This suggests that differentiated macrophages may be involved in maintaining asymptomatic hyperuricemia and in resolving acute attacks of gout.
- Top of page
- MATERIALS AND METHODS
Using an in vitro model of acute gout, we demonstrated that human macrophages, obtained by differentiation of peripheral blood monocytes for a minimum of 3–5 days in the presence of autologous serum, did not secrete TNFα, IL-1β, IL-6, or any other factors capable of inducing endothelial cell E-selectin expression or promoting secondary neutrophil recruitment under hydrodynamic shear flow. In contrast, freshly isolated monocytes from the same donor challenged with MSU crystals secreted these proinflammatory cytokines, induced E-selectin expression, and promoted rolling and adhesion of neutrophils on HUVECs. These observations support and extend our previous observations made with a panel of mouse monocytic cell lines, which showed divergent responses to MSU crystals according to the state of differentiation (14). Our results suggest that the process of macrophage differentiation may break the cycle of inflammation triggered by urate crystals, preventing proinflammatory cytokine synthesis, endothelial cell activation, and secondary neutrophil recruitment.
Several factors control the incidence, duration, and severity of gout in hyperuricemia. The capacity to nucleate and grow crystals is clearly an important variable. However, other factors besides the presence of crystals must be involved, because crystals can be detected in asymptomatic joints (2–5) and at sites of experimental intracutaneous crystal injection after erythema has resolved (12). One possibility is that crystals are coated with protective proteins, such as apolipoproteins, that reduce the inflammatory potential of crystals following uptake by polymorphonuclear cells (18–20). Dyslipidemia may therefore be another variable affecting the inflammatory response of given individuals to urate crystals. Another possibility is that products of the hypothalamic–pituitary axis may exert direct antiinflammatory effects on monocytes involved in MSU crystal clearance (21). The in vitro evidence presented here suggests that proper differentiation into macrophages should be considered as another variable.
Very little is known about the state of monocyte/macrophage differentiation in joints, either in health or disease. Histologic studies in mice suggest that synovial lining macrophages exist in a mature state of differentiation (22). The presence of end-stage macrophages in any given tissue is controlled by the availability of growth factors and cytokines, such as macrophage colony-stimulating factor and IL-6 (23, 24), that promote differentiation into the macrophagic end point. It is possible that polymorphisms in genes that control monocyte differentiation, such as the G174C polymorphism in IL-6 that reduces IL-6 levels (25), may prevent adequate differentiation of monocytes into macrophages, thus predisposing such individuals toward a more proinflammatory response to MSU crystals. Because the noninflammatory phenotype we have described here is closely linked with a macrophagic but not dendritic cell phenotype, a prediction arising from the present work is that mononuclear cells recovered from synovial fluid aspirates during the intercritical phase would be expected to express markers characteristic of differentiated macrophages, such as CD64, CD163, and RFD7 antigen.
Besides failing to induce TNFα release in macrophages, MSU crystals also actively suppressed secretion of TNFα in response to zymosan. This was not attributable to steric hindrance between MSU and zymosan particles, which suggested that the 2 stimuli were taken up by distinct phagocytic receptor pathways. Although zymosan is thought to be taken up via the glucan binding site on Mac-1 (CR3) (26), at present it is unknown how monocyte/macrophages phagocytose MSU crystals. Our present hypothesis is that phagocytosis of MSU and zymosan occurs through distinct receptor pathways, leading to separate but potentially interactive signaling events. It also remains possible, based on studies in neutrophils, that contact with the cell membrane alone, before phagocytosis has taken place, may be sufficient to trigger signaling events resulting in the eventual suppression of the zymosan response (27, 28).
The resolution phase of gout has previously been recognized to involve a contribution from macrophages based on their engulfment of apoptotic neutrophils, thus giving rise to Reiter's cells in synovial fluid (29, 30). Here, we have provided evidence that MSU uptake can initiate antiinflammatory responses in macrophages ab initio. This does not involve elaboration of IL-10 (or soluble p55 TNF receptor or IL-1 receptor antagonist [data not shown]) but may involve activation of antiinflammatory signaling pathways and/or elaboration of other antiinflammatory cytokines.
Although these possibilities have yet to be tested, the study of apoptotic cell removal, the only other phagocytic process thus far described to be uncoupled from inflammatory cytokine synthesis, has confirmed the existence of such antiinflammatory mechanisms in macrophages. Apoptotic cell removal is linked to crkII/Dock180/Rac 1 signaling (31) and antiinflammatory cytokine synthesis, including transforming growth factor β, prostaglandin E2, and platelet-activating factor (16). Regardless of whether similar antiinflammatory mechanisms are invoked following MSU crystal uptake, it remains likely that apoptotic neutrophil removal and MSU crystal phagocytosis will proceed hand-in-hand during the resolution phase of an acute gout attack.
Another relevant observation from the apoptotic neutrophil field is that the normal noninflammatory clearance mechanisms of macrophages can be subverted by concomitant engagement of Fc receptors (32). Coating of MSU crystals with IgG may thus predispose certain individuals to a more proinflammatory reaction to crystal deposition in their joints. This is consistent with other research showing that crystal-specific antibodies can help nucleate MSU crystallization (33), and that IgG coating of crystals can enhance the reactive oxygen burst elicited from polymorphonuclear phagocytes (34).
We have postulated a possible mechanism whereby monocyte/macrophage differentiation may influence the inflammatory response to MSU crystal precipitation in joint fluids (35). In this scenario, resident synovial macrophages would be expected to provide joint protection in hyperuricemia through the continual noninflammatory removal of insoluble MSU crystals. However, any of the known precipitants of gout, such as mechanical trauma, infection, or surgical stress, could tip the balance toward inflammation by triggering a fresh wave of monocyte and neutrophil recruitment from the bloodstream into the affected joint. Spontaneous resolution of an acute attack may be reinforced by differentiation of recently infiltrated monocytes into macrophages, which safely dispose of apoptotic neutrophils and MSU crystals through noninflammatory and antiinflammatory clearance mechanisms.