Adoptive transfer of M2 macrophages reduces neuropathic pain via opioid peptides

Abstract

Background: During the inflammation which occurs following nerve damage, macrophages are recruited to the site of injury. Phenotypic diversity is a hallmark of the macrophage lineage and includes pro-inflammatory M1 and anti-inflammatory M2 populations. Our aim in this study was to investigate the ability of polarized M0, M1, and M2 macrophages to secrete opioid peptides and to examine their relative contribution to the modulation of neuropathic pain.

Methods: Mouse bone marrow-derived cells were cultured as unstimulated M0 macrophages or were stimulated into an M1 phenotype using lipopolysaccharide and interferon-γ or into an M2 phenotype using interleukin-4. The macrophage phenotypes were verified using flow cytometry for surface marker analysis and cytokine bead array for cytokine profile assessment. Opioid peptide levels were measured by radioimmunoassay and enzyme immunoassay. As a model of neuropathic pain, a chronic constriction injury (CCI) of the sciatic nerve was employed. Polarized M0, M1, and M2 macrophages (5 × 105 cells) were injected perineurally twice, on days 14 and 15 following CCI or sham surgery. Mechanical and heat sensitivity were measured using the von Frey and Hargreaves tests, respectively. To track the injected macrophages, we also transferred fluorescently stained polarized cells and analyzed the surface marker profile of endogenous and injected cells in the nerves ex vivo.

Results: Compared to M0 and M1 cells, M2 macrophages contained and released higher amounts of opioid peptides, including Met-enkephalin, dynorphin A (1-17), and β-endorphin. M2 cells transferred perineurally at the nerve injury site reduced mechanical, but not heat hypersensitivity following the second injection. The analgesic effect was reversed by the perineurally applied opioid receptor antagonist naloxone methiodide. M2 cells did not affect sensitivity following sham surgery. Neither M0 nor M1 cells altered mechanical and heat sensitivity in CCI or sham-operated animals. Tracing the fluorescently labeled M0, M1, and M2 cells ex vivo showed that they remained in the nerve and preserved their phenotype.

Conclusions: Perineural transplantation of M2 macrophages resulted in opioid-mediated amelioration of neuropathy-induced mechanical hypersensitivity, while M1 macrophages did not exacerbate pain. Therefore, rather than focusing on macrophage-induced pain generation, promoting opioid-mediated M2 actions may be more relevant for pain control.

Keywords: Analgesia; Dynorphin; Endorphin; Enkephalin; Macrophage; Neuropathic pain.

Figures

Fig. 1

Analysis of surface markers on cultured macrophages following polarization. a Representative dot blots showing cell populations expressing F4/80 and CD163. The blots represent the CD163 FMO-negative control (far left), M0-polarized cells (second left), M1 cells (third left), and M2 cells (far right). The percentage of macrophages negative for CD163 (F4/80+CD163−) are shown in top left quadrants, while macrophages expressing CD163 (F4/80+CD163+) are shown in top right quadrants. b Representative dot blots showing cell populations expressing F4/80 and MHC II. The blots show the MHC II FMO-negative control (far left), the M0 cells (second left), the M1 cells (third left), and the M2 cells (far right). The percentage of macrophages negative for MHC II (F4/80+MHC II−) are shown in top left quadrants, while macrophages expressing MHC II (F4/80+MHC II+) are shown in top right quadrants. c Percentage of polarized M0, M1, and M2 macrophages expressing CD163. d Percentage of polarized M0, M1, and M2 macrophages expressing MHC II. The data were analyzed using flow cytometry and FlowJo software. In c and d, **p < 0.01, ***p < 0.001 (one-way RM ANOVA, Bonferroni’s test). Data show the mean ± SEM (n = 8 samples per group)

Fig. 2

Pro-inflammatory cytokine release profiles of cultured macrophages following polarization. Released levels of MCP-1 (a), TNF-α (b), IL-6 (c), and IFN-γ (d). The insert in d shows the graph with smaller Y-axis scale since IFN-γ levels were very low; only levels for M1 were above detection limit. The cytokine concentrations were measured in media using CBA, ~18 h after polarization was completed. *p < 0.05, **p < 0.01, ***p < 0.001 (one-way RM ANOVA, Bonferroni’s test). Data show the mean ± SEM (n = 8 samples per group)

Fig. 3

Opioid peptide content and release from cultured macrophages following polarization. Intracellular content (a) and extracellular, released levels (b) of Met-enkephalin, dynorphin, and β-endorphin. Opioid peptide levels were measured by RIA (Met-enkephalin and dynorphin) or EIA (β-endorphin), ~18 h after polarization was completed. *p < 0.05, **p < 0.01, ***p < 0.001 (one-way RM ANOVA, Bonferroni’s test). Data show the mean ± SEM (n = 8 samples per group)

Fig. 4

Effects of polarized macrophages transferred at the nerves on sensitivity following CCI and sham surgery. a, b Effects of M0, M1, and M2 macrophages on mechanical (a) and heat hypersensitivity (b) following CCI. c, d Effects of M0, M1, and M2 macrophages on mechanical (c) and heat sensitivity (d) following sham surgery. Macrophages or control DMEM were injected at the site of CCI or sham surgery on days 14 and 15 following surgeries (indicated by arrows). Mechanical and heat sensitivity were measured using von Frey filaments and Hargreaves test, respectively. ***p < 0.001 vs. control (DMEM); two-way RM ANOVA, Bonferroni’s test. Data show the mean ± SEM (n = 9 animals per group)

Fig. 5

Contribution of peripheral opioid receptors to M2 macrophage-mediated analgesia in neuropathy-induced mechanical hypersensitivity. M2 macrophages were injected on days 14 and 15 following CCI, and opioid receptor antagonist NLXM was applied on day 16 after CCI (i.e., 24 h after the second injection of macrophages). All injections (indicated by arrows) were performed at the CCI site, and the effects were assessed using von Frey filaments. ***p < 0.001 vs. vehicle; ###p < 0.001 vs. baseline measured 14 days after CCI (two-way RM ANOVA, Bonferroni’s test). Data show the mean ± SEM (n = 6 animals per group)

Fig. 6

CFSE labeling of cultured macrophages following polarization. CFSE (0.5 μM) was added to cells immediately after polarization and CFSE staining was measured ~24 h later using flow cytometry. p > 0.05 (one-way RM ANOVA). Data show the mean ± SEM (n = 8 samples per group)

Fig. 7

Quantification of CD45+ cells in injured nerves following perineural transfer of CFSE-labeled polarized macrophages. a Representative dot blots of CD45 staining: (left panel) CD45 FMO-negative control; (middle panel) CD45+ cells after one injection of DMEM only; (right panel) CD45+ cells following one injection of M2 macrophages. b Total number of CD45+ cells per nerve after one injection of DMEM only or CFSE-labeled M0, M1, and M2 cells. c Total number of CD45+ cells per nerve after two injections of DMEM only or CFSE-labeled M0, M1, and M2 cells. In all experiments, DMEM- or CFSE-labeled macrophages were injected at the injury site on day 14 (one injection) or on days 14 and 15 (two injections) after CCI. Animals were killed 24 h after the first injection and 24 h after the second injection, and cells from the nerves were analyzed using flow cytometry and FlowJo software. *p < 0.05, **p < 0.01, ***p < 0.001 (one-way ANOVA, Bonferroni’s test); ##p < 0.01; ###p < 0.001 compared to same treatment 24 h after the first injection (unpaired t test). Data show the mean ± SEM (n = 6 animals per group)

Fig. 8

Quantification of endogenous and injected macrophages in injured nerves following perineural transfer of CFSE-labeled polarized macrophages. a Representative dot blot of F4/80 and CFSE staining after one injection of DMEM only, showing endogenous macrophages (F4/80+CFSE−) in the top left quadrant, and the lack of injected macrophages (F4/80+CFSE+) in the top right quadrant. b Representative dot blot of F4/80 and CFSE staining after one injection of CFSE-labeled M2 cells, showing endogenous macrophages (F4/80+CFSE−) in the top left quadrant and injected macrophages (F4/80+CFSE+) in the top right quadrant. c Total number of endogenous macrophages (F4/80+CFSE−) after one injection of DMEM alone or CFSE-labeled M0, M1, and M2 cells. d Total number of endogenous macrophages (F4/80+CFSE−) after two injections of DMEM alone or CFSE-labeled M0, M1, and M2 cells. e Total number of injected macrophages (F4/80+CFSE+) after one injection of DMEM alone or CFSE-labeled M0, M1, and M2 cells. f Total number of injected macrophages (F4/80+CFSE+) after two injections of DMEM alone or CFSE-labeled M0, M1, and M2 cells. In all experiments, DMEM- or CFSE-labeled macrophages were injected at the injury site on day 14 (one injection) or on days 14 and 15 (two injections) after CCI. Animals were killed 24 h after the first injection and 24 h after the second injection, and cells from the nerves were analyzed using flow cytometry and FlowJo software. *p < 0.05, **p < 0.01, ***p < 0.001 (one-way ANOVA, Bonferroni’s test); ##p < 0.01, ###p < 0.001 compared to the same treatment 24 h after the first injection (unpaired t test). Data show the mean ± SEM (n = 6 animals per group)

Fig. 9

Analysis of CD163 and MHC II expression in endogenous macrophages in injured nerves following perineural transfer of CFSE-labeled polarized macrophages. Endogenous macrophages were identified as F4/80+CFSE− cells. a Percentage of endogenous macrophages expressing CD163 following one injection of DMEM only or CFSE-labeled M0, M1, and M2 cells. b Percentage of endogenous macrophages expressing CD163 following two injections of DMEM only or CFSE-labeled M0, M1, and M2 cells. c Percentage of endogenous macrophages expressing MHC II following one injection of DMEM only or CFSE-labeled M0, M1, and M2 cells. d Percentage of endogenous macrophages expressing MHC II following two injections of DMEM only or CFSE-labeled M0, M1, and M2 cells. In all experiments, DMEM- or CFSE-labeled macrophages were injected at the injury site on day 14 (one injection) or on days 14 and 15 (two injections) after CCI. Animals were killed 24 h after the first injection and 24 h after the second injection, and cells from the nerves were analyzed using flow cytometry and FlowJo software. *p < 0.05, **p < 0.01, ***p < 0.001 (one-way ANOVA, Bonferroni’s test); #p < 0.05, ##p < 0.01 compared to the same treatment 24 h after the first injection (unpaired t test). Data show the mean ± SEM (n = 6 animals per group)

Fig. 10

Analysis of CD163 and MHC II expression in perineurally injected CFSE-labeled polarized macrophages in injured nerves. Injected macrophages were identified as F4/80+CFSE+ cells. a Percentage of injected macrophages expressing CD163 following one injection of DMEM only or CFSE-labeled M0, M1, and M2 cells. b Percentage of injected macrophages expressing CD163 following one injections of DMEM only or CFSE-labeled M0, M1, and M2 cells. c Percentage of injected macrophages expressing MHC II following one injection of DMEM only or CFSE-labeled M0, M1, and M2 cells. d Percentage of injected macrophages expressing MHC II following two injections of DMEM only or CFSE-labeled M0, M1, and M2 cells. In all experiments, DMEM medium or CFSE-labeled macrophages were injected at the injury site on day 14 (one injection) or on days 14 and 15 (two injections) after CCI. Animals were killed 24 h after the first injection and 24 h after the second injection, and cells from the nerves were analyzed using flow cytometry and FlowJo software. **p < 0.01, ***p < 0.001 (one-way ANOVA, Bonferroni’s test); #p < 0.05 compared to the same treatment 24 h after the first injection (unpaired t test). Data show the mean ± SEM (n = 6 animals per group)