Mycobacteria attenuate nociceptive responses by formyl peptide receptor triggered opioid peptide release from neutrophils

Abstract

In inflammation, pain is regulated by a balance of pro- and analgesic mediators. Analgesic mediators include opioid peptides which are secreted by neutrophils at the site of inflammation, leading to activation of opioid receptors on peripheral sensory neurons. In humans, local opioids and opioid peptides significantly downregulate postoperative as well as arthritic pain. In rats, inflammatory pain is induced by intraplantar injection of heat inactivated Mycobacterium butyricum, a component of complete Freund's adjuvant. We hypothesized that mycobacterially derived formyl peptide receptor (FPR) and/or toll like receptor (TLR) agonists could activate neutrophils, leading to opioid peptide release and inhibition of inflammatory pain. In complete Freund's adjuvant-induced inflammation, thermal and mechanical nociceptive thresholds of the paw were quantified (Hargreaves and Randall-Selitto methods, respectively). Withdrawal time to heat was decreased following systemic neutrophil depletion as well as local injection of opioid receptor antagonists or anti-opioid peptide (i.e. Met-enkephalin, beta-endorphin) antibodies indicating an increase in pain. In vitro, opioid peptide release from human and rat neutrophils was measured by radioimmunoassay. Met-enkephalin release was triggered by Mycobacterium butyricum and formyl peptides but not by TLR-2 or TLR-4 agonists. Mycobacterium butyricum induced a rise in intracellular calcium as determined by FURA loading and calcium imaging. Opioid peptide release was blocked by intracellular calcium chelation as well as phosphoinositol-3-kinase inhibition. The FPR antagonists Boc-FLFLF and cyclosporine H reduced opioid peptide release in vitro and increased inflammatory pain in vivo while TLR 2/4 did not appear to be involved. In summary, mycobacteria activate FPR on neutrophils, resulting in tonic secretion of opioid peptides from neutrophils and in a decrease in inflammatory pain. Future therapeutic strategies may aim at selective FPR agonists to boost endogenous analgesia.

Figure 1. Neutrophils attenuate inflammatory pain by local opioid peptide release.

[A] In a rat model, inflammation was induced by intraplantar injection of heat-inactivated Mycobacterium butyricum (complete Freund's adjuvant). Prior to induction of inflammation, rats were pretreated with i.v. anti-neutrophil serum (open triangles). Control animals received non-immune rabbit serum (filled triangle). Two hours after complete Freund's adjuvant inoculation thermal nociceptive thresholds (i.e. paw withdrawal latency) were significantly decreased in neutropenic rats (n = 6–8; * p<0.05, t-test). [B, C] To examine the role of endogenous opioids in local inflammatory pain control, rats with complete Freund's adjuvant-induced inflammation were intraplantarly injected with the opioid receptor antagonist naloxone (0.56 ng, control solvent only; open and filled circle respectively; baseline, BL [B]) (n = 9–14), with an anti-opioid peptide antibody (i.e. anti-Met-enkephalin, 1.25 µg, or anti-β-endorphin, 2 µg) or with control IgG; open and filled squares, respectively [C, D]) and nociceptive thresholds were determined. Nociceptive thresholds were significantly decreased following both treatments (* p<0.05 all one way ANOVA, Dunn's method). Data are presented as means±SEM.

Figure 2. Opioid peptide release from neutrophils is triggered by mycobacteria and FPR agonists but not by TLR-2 or TLR-4 agonists.

[A–C] Rat and human neutrophils as well as CD14⁺ human monocytes were incubated with heat-inactivated Mycobacterium butyricum (Myco. but.), and Met-enkephalin (ENK) release was quantified by radioimmunoassay (n = 7–13 * p<0.05, one way RM ANOVA, Student-Newman-Keuls Method). [D, E] Expression of TLR-2, TLR-4 [D] and FPR [E] was determined on human neutrophils by flow cytometry (dotted line: unstained control, grey histogram: isotype control, black line: anti-TLR-2-PE, black histogram: fMLP-FITC or anti-TLR-4-PE). [F–H] Human (n = 7–16) and [I–K] rat neutrophils (n = 7–21) were stimulated with the TLR-2 agonist peptidoglycan, the TLR-4 agonist lipopolysaccharide or the FPR agonist fMLP, and Met-enkephalin (ENK) release was measured in the supernatant (* p<0.05; one way RM ANOVA, Student-Newman-Keuls Method). Data are presented as means±SEM.

Figure 3. Formyl peptides inhibit inflammatory pain through release of opioid peptides from neutrophils.

[A, C] Rats received intraplantar injections of fMLP into inflamed paws (2 h post complete Freund's adjuvant: filled circles; noninflamed paw: open circles). Paw withdrawal latency [A] or paw pressure thresholds [C] were determined 5–7 min after injections (n = 5–6, * p<0.05 one way ANOVA, Duncan's and Dunnett's method, respectively as well as p<0.001 linear regression analysis for dose-dependency of fMLP-induced analgesia). [B] Two hours post complete Freund's adjuvant rats received intraplantar fMLP (3 ng) together with either control IgG (filled circles), antibody against Met-enkephalin (open circles) or naloxone. Paw withdrawal latency was measured 5–7- min thereafter (baseline, BL; n = 6;* p<0.05 one way ANOVA, Duncan's method). [D] Subcutaneous paw tissue was analyzed for FPR expression by flow cytometry. Cells were first gated on CD45⁺ hematopoetic cells (left panel) followed by gating on RP1-PE⁺ neutrophils (upper middle panel) or ED1-PE⁺ macrophages (lower middle panel). Staining with fMLP-FITC (solid lanes) was analyzed in comparison to unstained controls (grey histograms) (right panel). [E] Systemic neutrophil depletion by intravenous injection of anti-neutrophil serum 18 h before induction of inflammation (anti-neutrophil, open triangles; control: rabbit IgG, closed triangles; n = 6, * p<0.05 one way RM ANOVA, Duncan's method) abolished fMLP-induced analgesia. [F–H] Similarly, concomitant intraplantar injection of fMLP with either the opioid receptor antagonist naloxone ([F] 0.56 ng, open diamond, control: solvent, filled diamond; n = 6, * p<0.05 one way ANOVA, Dunnett's method), CTOP ([G] 50 µg, filled triangles), NTI ([G] 20 µg, open circles; n = 6, * p<0.05 one way ANOVA, Student–Newman-Keuls) or anti–Met-enkephalin antibody ([H] 1.25 µg open circles; control: rabbit IgG, filled circles; n = 4–5) resulted in significant inhibition of analgesia (* p<0.05 one way ANOVA, Duncan's method). All data are means±SEM.

Figure 4. fMLP-induced opioid peptide release in vitro and analgesia in vivo are dependent on activation of the FPR.

[A] FPR staining of human neutrophils with fMLP-FITC (black histogram) is dose-dependently inhibited by the FPR antagonist Boc-FLFLF (Boc, dotted line: 0.1 µM, thin black line: 1 µM, thick black line: 10 µM). The unstained control is shown in the gray histogram. [B] Neutrophils were loaded with Fura-2 and changes in [Ca²⁺]_i were analyzed after addition of fMLP (left panel) and after preincubation with Boc-FLFLF and subsequent stimulation with fMLP (right panel). Boc-FLFLF also blocked the fMLP-triggered release of Met-enkephalin (ENK) from human [C] and rat [D] neutrophils (n = 5–9; both * p<0.05 one way RM ANOVA, Student-Newman-Keuls Method). [E] In vivo, intraplantar injection of Boc-FLFLF blocked fMLP-induced analgesia (0.3 ng fMLP) in the inflamed paw (filled squares; noninflamed paws: open squares). Baseline hyperalgesia prior to fMLP injection is shown for comparison (n = 6; * p<0.05 one way ANOVA, Duncan's method, and p<0.001 linear regression analysis for dose-dependency of FPR blockage). Data are means±SEM.

Figure 5. Mycobacterium butyricum-triggered intracellular calcium elevation is FPR dependent.

Mock [A] or human FPR [B,C] transfected HEK293 cells or human neutrophils [D–F] were loaded with Fura-2. Changes in [Ca²⁺]_i were analyzed after addition of Mycobacterium butyricum (Myco. but.) at 60 s in the presence [C, E] or absence [B,D] of 10 µM Boc-FLFLF (Boc). [F] Impact of fMLP stimulation on subsequent stimulation with Mycobacterium butyricum in human neutrophils.

Figure 6. Mycobacterium butyricum-induced opioid peptide release and translocation of Met-enkephalin-containing granules from human neutrophils requires FPR stimulation, intracellular Ca²⁺ mobilization and PI3K activation.

[A, B] Met-enkephalin release induced by Mycobacterium butyricum (Myco. but. 0.66 mg/ml) was analyzed after preincubation with anti-TLR-2 and anti-TLR-4 ([A], anti-TLR2 or 4 both 10 µg/ml, n = 10–12), or the FPR antagonist Boc-FLFLF ([B], Boc 10 µM; n = 7–8, * p<0.05, both one way RM ANOVA, Student-Newman-Keuls Method). [C–E] Cytospins from freshly isolated human neutrophils previously incubated with Mycobacterium butyricum in the presence [F] or absence [E] of the FPR antagonist Boc-FLFLF were stained with anti-Met-enkephalin Ab (red; [C] solvent control only). Original magnification ×63. [F–H] Mycobacterium butyricum-induced opioid peptide release was prevented by the intracellular Ca²⁺ chelator BAPTA/AM (100 µM, [F]; n = 8) but not dependent on extracellular Ca²⁺ ([G], n = 7–14, no [Ca²⁺]_e crosshatched bars). [H] Mycobacterium butyricum-induced opioid peptide release was also blocked by the PI3K-inhibitors LY294002 (LY, 100 µM) and wortmannin (wort, 100 nM, n = 8–10). All experiments * p<0.05 one way RM ANOVA, Student-Newman-Keuls Method. Data are means±SEM.

Figure 7. Tonic opioid peptide release through FPR stimulation reduces inflammatory pain.

[A] Rat neutrophils were repeatedly stimulated with Mycobacterium butyricum (Myco. but. 0.66 mg/ml) and Met-enkephalin (ENK) content was determined in supernatants (n = 12–16; * p<0.05, one way RM ANOVA, Student-Newman-Keuls Method). [B, C] Mycobacterium butyricum-induced Met-enkephalin release was unaltered by anti-TLR-2 and -TLR-4 antibodies ([B], anti-TLR-2/4 both 10 µg/ml, n = 8–19), but was significantly reduced by FPR antagonist Boc-FLFLF ([C], Boc: 100 µM, n = 7–11, both * p<0.05 one way RM ANOVA, Student-Newman-Keuls Method. [D] Rats with 2 h complete Freund's adjuvant inflammation were intraplantarly injected with Boc-FLFLF (Boc, 3 µg, [D], n = 5–6) and nociceptive thresholds were determined (*p<0.05 one way RM ANOVA, Duncan Method). Data are means±SEM. [E] Rats (n = 6) with 2, 12 and 24 h complete Freund's adjuvant inflammation were treated with 3 µg Boc intraplantarly and paw withdrawal latency was measured afterwards (* p<0.05, Wilcoxon Signed Rank test compared to baseline (BL)).