Chemokine CXCL1 enhances inflammatory pain and increases NMDA receptor activity and COX-2 expression in spinal cord neurons via activation of CXCR2

Abstract

Recent studies have shown that CXCL1 upregulation in spinal astrocytes is involved in the maintenance of neuropathic pain. However, whether and how CXCL1 regulates inflammatory pain remains unknown. Here we show that intraplantar injection of CFA increased mRNA and protein expressions of CXCL1 and its major receptor CXCR2 in the spinal cord at 6h and 3days after the injection. Immunofluorescence double staining showed that CXCL1 and CXCR2 were expressed in spinal astrocytes and neurons, respectively. Intrathecal injection of CXCL1 neutralizing antibody or CXCR2 antagonist SB225002 attenuated CFA-induced mechanical and heat hypersensitivity on post-CFA day 3. Patch-clamp recordings showed that CXCL1 potentiated NMDA-induced currents in lamina II neurons via CXCR2, and this potentiation was further increased in CFA-treated mice. Furthermore, intrathecal injection of CXCL1 increased COX-2 expression in dorsal horn neurons, which was blocked by pretreatment with SB225002 or MEK (ERK kinase) inhibitor PD98059. Finally, pretreatment with SB225002 or PD98059 decreased CFA-induced heat hyperalgesia and COX-2 mRNA/protein expression and ERK activation in the spinal cord. Taken together, our data suggest that CXCL1, upregulated and released by spinal astrocytes after inflammation, acts on CXCR2-expressing spinal neurons to increase ERK activation, synaptic transmission and COX-2 expression in dorsal horn neurons and contributes to the pathogenesis of inflammatory pain.

Keywords: Astrocytes; Astroglial–neuronal interaction; COX-2; CXCL1; CXCR2; Chemokines; Complete Freund's adjuvant; ERK; Inflammatory pain.

Figure 1

CFA induces CXCL1 mRNA and protein expression in the spinal cord. (A) Real-time PCR results show the increase of CXCL1 mRNA expression in the spinal cord at 6 h and 3 d after CFA injection. * P < 0.05, compared to naïve. # P < 0.05, compared to saline. n = 4 mice/group. (B) Elisa results show that CXCL1 protein expression was also increased in the spinal cord after CFA. ** P < 0.01, compared to naïve. ## P < 0.01, compared to saline. n = 5 mice/group. (C–F) Immunostaining shows the CXCL1 expression in the spinal cord in naïve (C) and CFA (D) animals. CFA increased CXCL1-IR in the spinal cord at 3 days after injection (D). (E–G) Double staining shows CXCL1 was colocalized with astrocytic marker, GFAP (E), but not with neuronal marker NeuN (F) or microglial marker CD11b (G).

Figure 2

CFA induces CXCR2 mRNA and protein expression in the spinal cord. (A) Real-time PCR results show the increase of CXCR2 mRNA expression in the spinal cord at 6 h and 3 d after CFA injection. ** P < 0.01, compared to naïve. ## P < 0.01, compared to saline. n = 4 mice/group. (B) Western blot show CXCR2 protein expression in the spinal cord. CXCR2 increased 1.3-fold after CFA injection compared to naïve (P < 0.05). n = 3 mice/group. (C, D) Immunostaining shows the CXCR2 expression in the spinal cord in naïve (C) and CFA (D) animals. CXCR2-IR was enhanced at 3 days after CFA (D). (E–G) Double staining shows CXCR2 was colocalized with neuronal marker NeuN.

Figure 3

Intrathecal injection of CXCL1 neutralizing antibody or SB225002, a potent and selective CXCR2 antagonist, reduces CFA-induced mechanical allodynia and heat hyperalgesia at 3 days after CFA. (A, B) CXCL1 neutralizing antibody at a lower dose (4 μg) had mild effect on CFA-induced pain hypersensitivity, whereas at a higher dose (8 μg) the neutralizing antibody reversed CFA-induced mechanical allodynia (A) and heat hyperalgesia (B) for more than 24 hours. * P < 0.05, ** P < 0.01,*** P < 0.001 compared to control serum. BL, baseline. n = 6 mice/group. (C, D) SB2205002 dose-dependently attenuated CFA-induced mechanical allodynia (C) and heat hyperalgesia (D) at 3 days, with longer effect on the heat hyperalgesia. * P < 0.05; ** P < 0.01; *** P < 0.001 compared to vehicle. n = 6 mice/group.

Figure 4

Bath application of CXCL1 increases NMDA-induced currents in lamina IIo neurons in spinal cord slices. (A) Patch-clamp recording demonstrates increase in NMDA-induced currents after CXCL1 treatment (100 ng/ml, 2 min), which is blocked by SB225002 (0.1 μM). (B) Histogram shows the amplitude of NMDA-induced currents. * P < 0.05 compared to NMDA alone. # P < 0.05, compared to NMDA + CXCL1 treatment. n = 4–7 neurons/group. (C) Patch-clamp recording shows CXCL1 increases NMDA-induced current in naïve and CFA-treated animals. (D) Histogram shows the NMDA-induced inward current. * P < 0.05 compared to naïve. (E) Histogram shows the increased magnitude of NMDA-induced currents. * P < 0.05 compared to pretreatment baseline. # P < 0.05, CFA 3 days compared to naïve. n = 5–7 neurons/group.

Figure 5

CXCL1 increases the expression of COX-2 in spinal cord neurons. (A–C) Intrathecal injection of CXCL1 (100 ng) increases COX-2 expression at 2 h after injection. (D) Majority of COX-2 immunoreactive cells in spinal cord express the neuronal marker NeuN. (E) CXCL1 also increased COX-2 mRNA expression in the spinal cord, which is blocked by pretreatment with SB225002 (20 μg) or PD98059. * P < 0.05, ** P < 0.01 compared to PBS. # P < 0.05 compared to CXCL1, one-way ANOVA followed by Newman-Keuls post hoc test. n = 4 mice/group.

Figure 6

CFA induces CXCR2/ERK dependent COX-2 increase in the spinal cord. (A) Intrathecal pretreatment with SB225002 or PD98059 partly prevented CFA-induced heat hyperalgesia. *** P < 0.001 compared to vehicle. Student's t-test. n=6 mice/group. (B) Intrathecal pretreatment with SB225002 or PD98059 decreased CFA-induced COX-2 mRNA increase in the spinal cord. *** P < 0.001 compared to naive. # P < 0.01, compared to vehicle. n=6 mice/group. (C) Intrathecal pretreatment with SB225002 or PD98059 reduced CFA-induced COX-2 protein expression and pERK1/2 activation in the spinal cord. (D, E) Density of COX-2 bands (D) and pERK1, pERK2 bands (E), which is normalized to GAPDH loading control. *** P < 0.001 compared to naive. ## P < 0.01, ### P < 0.001 compared to vehicle. One-way ANOVA followed by Newman-Keuls post hoc test. n = 4 mice/group.