Imbalance of proresolving lipid mediators in persistent allodynia dissociated from signs of clinical arthritis

Abstract

Rheumatoid arthritis-associated pain is poorly managed, often persisting when joint inflammation is pharmacologically controlled. Comparably, in the mouse K/BxN serum-transfer model of inflammatory arthritis, hind paw nociceptive hypersensitivity occurs with ankle joint swelling (5 days after immunisation) persisting after swelling has resolved (25 days after immunisation). In this study, lipid mediator (LM) profiling of lumbar dorsal root ganglia (DRG), the site of sensory neuron cell bodies innervating the ankle joints, 5 days and 25 days after serum transfer demonstrated a shift in specialised proresolving LM profiles. Persistent nociception without joint swelling was associated with low concentrations of the specialised proresolving LM Maresin 1 (MaR1) and high macrophage numbers in DRG. MaR1 application to cultured DRG neurons inhibited both capsaicin-induced increase of intracellular calcium ions and release of calcitonin gene-related peptide in a dose-dependent manner. Furthermore, in peritoneal macrophages challenged with lipopolysaccharide, MaR1 reduced proinflammatory cytokine expression. Systemic MaR1 administration caused sustained reversal of nociceptive hypersensitivity and reduced inflammatory macrophage numbers in DRG. Unlike gabapentin, which was used as positive control, systemic MaR1 did not display acute antihyperalgesic action. Therefore, these data suggest that MaR1 effects observed after K/BxN serum transfer relate to modulation of macrophage recruitment, more likely than to direct actions on sensory neurons. Our study highlights that, in DRG, aberrant proresolution mechanisms play a key role in arthritis joint pain dissociated from joint swelling, opening novel approaches for rheumatoid arthritis pain treatment.

Figure 1.

K/BxN serum-transfer arthritis is associated with mechanical hypersensitivity and macrophage joint swelling in DRG when joint swelling is resolving in the hind paw. (A) Clinical signs of arthritis after K/BxN serum transfer (2 × 50-µL intraperitoneal injections on days 0 and 2) evaluated using a 12-point clinical arthritis scoring of mouse paws (B) Mechanical hypersensitivity assessed using von Frey filaments after K/BxN serum transfer. *P < 0.05 or ***P < 0.001 vs same day control, two-way repeated-measures ANOVA, post hoc Tukey. Data are expressed as mean ± SEM; n = 10 male mice per group. Representative images of F4/80⁺ profiles (arrows) (macrophages) infiltrated in DRG on (C and D) day 5 and (E and F) day 25 after serum transfer. Scale bars, 50 μm. (G) Quantification of macrophages in lumbar DRG at 5 and 25 days after K/BxN serum transfer. *P < 0.05 vs same day control, one-way ANOVA, post hoc Tukey. Data are expressed as mean ± SEM; n = 4 mice per group. ANOVA, analysis of variance; DRG, dorsal root ganglia.

Figure 2.

Distinct lipid mediator profiles in lumbar DRG at day 5 and day 25 after K/BxN serum transfer. Orthogonal partial least square to latent structure-discriminant analysis was used to generate 2 principal component functions from bioactive lipid mediator levels. (A and B) 2D score plot, gray ellipse denotes 95% confidence regions and (C and D) corresponding loading plots, LM coloured displayed a variable importance in projection coefficient ≥1. (A and C) 5 days or (B and D) 25 days after K/BxN serum transfer. (E and F) Quantification of selected lipid mediator levels at 5 and 25 days after K/BxN serum transfer. *P<0.05, Unpaired Student t test. Data are mean ± SEM; n = 5 mice per group. DRG, dorsal root ganglia; LM, lipid mediator.

Figure 3.

MaR1 reduces capsaicin-induced increase of intracellular Ca²⁺ in cultured DRG neurons through GPCRi/o pathway-mediated mechanisms. (A) Calcium imaging trace of neurons loaded with fura-2-AM (3 µM, 30 minutes) and then exposed to 6 capsaicin pulses (500 nM, 15 seconds/pulse) (red dots) at 5-minute intervals, before final application of KCl (50 mM, 15 seconds) (black dot). (B) Quantification of calcium responses after 6 consecutive applications of capsaicin. Data are mean ± SEM; n = 3 experiments. (C and D) Representative calcium traces of neurons exposed to HBSS buffer control (90 seconds) or MaR1 (3 ng/mL, 90 seconds) with the fourth capsaicin pulse. (E) Capsaicin response ratio (response 4/3) in the presence of MaR1 (0.83-8.3 nM for 90 seconds) during fourth capsaicin pulse. Data are mean ± SEM; n = 4 experiments. (F) Capsaicin response ratio in the presence of MaR1 (8.3 nM for 90 seconds) during fourth capsaicin pulse in neurons that were preincubated in pertussis toxin (GPCRi/o inhibitor) (500 ng/mL, 18 hours). *P < 0.05, **P < 0.01, one-way ANOVA, post hoc Tukey. Data are expressed as mean ± SEM; n = 4 experiments. ANOVA, analysis of variance; DRG, dorsal root ganglia.

Figure 4.

MaR1 reduces expression of proinflammatory markers in macrophages challenged with lipopolysaccharide. Cultured peritoneal macrophages incubated with or without LPS (100 ng/mL for 3 hours), followed by MaR1 (3 ng/mL = 8.3 nM) or vehicle for 5 hours. Quantification of mRNA for (A-D) proinflammatory markers and (E-G) anti-inflammatory markers assessed by qPCR. *P < 0.05, **P < 0.01, ***P < 0.001 compared with medium/vehicle controls; #P < 0.05, one-way ANOVA, post hoc Tukey. Data are mean ± SEM; n = 4 to 6 experiments. ANOVA, analysis of variance; LPS, lipopolysaccharide; qPCR, quantitative polymerase chain reaction.

Figure 5.

Repeated systemic administration of MaR1 results in sustained reversal of K/BxN serum-transfer–associated allodynia. (A) Timeline of experiments in which MaR1 (100 ng/mouse i.p) was repeatedly administered either at days 5, 7, 9, and 11 (protocol 1) or days 19, 21, and 23 (protocol 2) after serum transfer. (B) Clinical scoring of hind paws following MaR1 or saline administration. Control mice received control serum. (C) Reversal of mechanical hypersensitivity following third and fourth doses of MaR1 on days 9 and 11 after serum transfer. (D) Reversal of mechanical hypersensitivity after second and third doses of MaR1 on days 21 and 23 after serum transfer. Arrows indicate treatment days. **P < 0.01, ***P < 0.001 vs same day K/BxN-Saline group (closed circles), two-way ANOVA, post hoc Tukey. Data are mean ± SEM; n = 8 animals per group. ANOVA, analysis of variance.

Figure 6.

MaR1 treatment is associated with a decrease of immune cells and M1 macrophages recruitment in DRG after K/BxN serum transfer. (A–C) Representative scatterplots of immune cells sorted from lumbar and cervical DRG dissected at day 25 after transfer of control serum, K/BxN serum or K/BxN serum and MaR1 (4 doses, days 5-11 after K/BxN serum transfer, protocol 1). Cells were gated on CD45⁺, F4/80⁺, and CD11b⁺. Macrophages were defined as CD45⁺F4/80⁺CD11b⁺ and further analysed for M1 (CD11c⁺ CD206⁻) and M2 (CD11c⁻CD206⁺). (D–F) Bar charts representing numbers of leukocytes (CD45⁺ cells), macrophages (CD11b⁺F4/80⁺) and M1 macrophages (CD206⁻CD11c⁺) 14 days after last MaR1 dose (protocol 1); *P < 0.05, **P < 0.01, one-way ANOVA, post hoc Tukey. Data are mean ± SEM; n = 3 to 5 animals per group. (G–I) Bar charts representing numbers of leukocytes (CD45⁺ cells), macrophages (CD11b⁺F4/80⁺), and M1 macrophages (CD206⁻CD11c⁺) 2 days after last MaR1 dose (protocol 2); *P < 0.05, **P < 0.01, one-way ANOVA, post hoc Tukey. Data are mean ± SEM; n = 6 to 8 animals per group. ANOVA, analysis of variance; DRG, dorsal root ganglia.