Osteoarthritis-associated basic calcium phosphate crystals activate membrane proximal kinases in human innate immune cells

Abstract

Background: Osteoarthritis (OA) is a chronic debilitating joint disorder of particularly high prevalence in the elderly population. Intra-articular basic calcium phosphate (BCP) crystals are present in the majority of OA joints and are associated with severe degeneration. They are known to activate macrophages, synovial fibroblasts, and articular chondrocytes, resulting in increased cell proliferation and the production of pro-inflammatory cytokines and matrix metalloproteases (MMPs). This suggests a pathogenic role in OA by causing extracellular matrix degradation and subchondral bone remodelling. There are currently no disease-modifying drugs available for crystal-associated OA; hence, the aim of this study was to explore the inflammatory pathways activated by BCP crystals in order to identify potential therapeutic targets to limit crystal-induced inflammation.

Methods: Primary human macrophages and dendritic cells were stimulated with BCP crystals, and activation of spleen tyrosine kinase (Syk), phosphoinositide-3 kinase (PI3K), and mitogen-activated protein kinases (MAPKs) was detected by immunoblotting. Lipopolysaccharide (LPS)-primed macrophages were pre-treated with inhibitors of Syk, PI3K, and MAPKs prior to BCP stimulation, and cytokine production was quantified by enzyme-linked immunosorbent assay (ELISA). Aa an alternative, cells were treated with synovial fluid derived from osteoarthritic knees in the presence or absence of BCP crystals, and gene induction was assessed by real-time polymerase chain reaction (PCR).

Results: We demonstrate that exposure of primary human macrophages and dendritic cells to BCP crystals leads to activation of the membrane-proximal tyrosine kinases Syk and PI3K. Furthermore, we show that production of the pro-inflammatory cytokines interleukin (IL)-1α and IL-1β and phosphorylation of downstream MEK and ERK MAPKs is suppressed following treatment with inhibitors of Syk or PI3K. Finally, we demonstrate that treatment of macrophages with BCP crystals induces the production of the damage-associated molecule S100A8 and MMP1 in a Syk-dependent manner and that synovial fluid from OA patients together with BCP crystals exacerbates these effects.

Conclusions: We identify Syk and PI3K as key signalling molecules activated by BCP crystals prior to inflammatory cytokine and DAMP expression and therefore propose that Syk and PI3K represent potential targets for the treatment of BCP-related pathologies.

Keywords: BCP crystals; Inflammation; PI3K; S100 proteins; Syk.

Fig. 1

BCP crystals activate Syk and PI3K in primary macrophages and DC. a, e Human macrophages and b, f DC (2 × 10⁶ cells/well) were stimulated with BCP crystals (50 μg/ml) for the indicated time points, and phosphorylation of Syk and PI3K was detected by immunoblotting using phospho-specific antibodies. Representative blots of three independent experiments are shown. c, d, g, h Densitometric analysis of three blots was performed using ImageJ software. Bar graphs illustrate the mean (± SEM) increase in phosphorylation, relative to the untreated sample (0) and normalised to total Syk/PI3K protein. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001

Fig. 2

Inhibition of Syk and PI3K reduces BCP crystal-induced IL-1 production in primary macrophages. Human macrophages (0.5 × 10⁶ cells/well) were primed with LPS (100 ng/ml) for 2 h prior to treatment with a–c piceatannol (10 μM, 25 μM, and 50 μM) or d–f LY294002 (10 μM, 25 μM, and 50 μM) for 1 h and stimulation with BCP crystals (50 μg/ml) for 6 h. Cell supernatants were assessed for IL-1β, IL-1α, and TNF-α by ELISA. Results indicate mean (± SEM) of three independent experiments. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 vs LPS + BCP

Fig. 3

BCP crystal-induced IL-1 production occurs via lipid raft formation in primary macrophages. LPS-primed human macrophages were treated with a–c an Fc receptor blocking antibody (5 μg/ml) or d–f M-βCD (10 mM) for 1 h prior to BCP crystal stimulation for 6 h. Cell supernatants were assessed for IL-1β, IL-1α, and TNF-α by ELISA. Results indicate mean (± SEM) of three independent experiments. ***P < 0.001

Fig. 4

BCP crystals activate MEK and ERK downstream of Syk/PI3K in primary macrophages and DC. a,b Human macrophages and c, d DC (2 × 10⁶ cells/well) were stimulated with BCP crystals (50 μg/ml) for the indicated time points. Phosphorylation of MEK and ERK was detected by immunoblotting using phospho-specific antibodies. e,f Human macrophages and g,h DC were pre-treated with piceatannol (50 μM, 100 μM; lanes 3, 4), R788 (5 μM; lane 5), or LY294002 (50 μM; lane 6) for 30 min prior to stimulation with BCP crystals for 15 min. Phosphorylation of MEK and ERK was detected by immunoblotting. Representative blots of three independent experiments are shown. Densitometric analysis of three blots was performed using ImageJ software. Bar graphs illustrate the mean (± SEM) increase in phosphorylation, relative to the untreated sample (0) and normalised to total MEK/ERK protein. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001

Fig. 5

BCP crystals induce expression of damage-associated molecules and MMPs in primary human macrophages in a Syk- and PI3K-dependent manner. Human macrophages (0.5 × 10⁶ cells/well) were stimulated with BCP crystals (50 μg/ml) for 3, 6, and 24 h, and mRNA levels of a S100A8, b S100A12, c MMP1, d MMP2, e MMP9, and f TIMP1 were analysed by real-time PCR. Alternatively, supernatants were harvested; whole cell lysates (WCL) were prepared and both were analysed for the presence of g S100A8, h S100A12, and i MMP1 protein by immunoblotting. Representative blots of three independent experiments are shown. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001

Fig. 6

BCP crystals induce expression of S100A8, S100A12, and MMP1 in primary macrophages in a Syk-, PI3K-, and ERK-dependent manner. Human macrophages were pre-treated with a piceatannol (50 μM, 100 μM), b LY294002 (25 μM, 50 μM), or c PD98059 (5 μM, 10 μM) for 30 mins prior to stimulation with BCP crystals for 24 h. mRNA levels of S100A8, S100A12, MMP1, and TIMP1 were analysed by real-time PCR. Results indicate mean (± SEM) of three independent experiments. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001

Fig. 7

S100A8 primes human macrophages for BCP-induced IL-1β production. Human macrophages were primed with recombinant human S100A8 (1 μg/ml) for 3 h prior to stimulation with BCP for 6 h. Cell supernatants were assessed for IL-1β, IL-1α, and TNF-α by ELISA. Results indicate mean (± SEM) of three independent experiments. ***P ≤ 0.001

Fig. 8

BCP crystals exert synergistic effects with OA synovial fluid on primary macrophages. Human macrophages (0.5 × 10⁶ cells/well) were pre-treated with R788 (2.5 μM) prior to treatment with BCP crystals alone, synovial fluid from one of three OA patients (a, b or c), or OA synovial fluid and BCP crystals together for 24 h. Expression of (upper panel) S100A8 and (lower panel) MMP1 were analysed by real-time PCR. Results indicate mean (± SEM) of at least three healthy macrophage donors treated with synovial fluid from one OA patient. *P ≤ 0.05, **P ≤ 0.01