Spinal injection of TNF-α-activated astrocytes produces persistent pain symptom mechanical allodynia by releasing monocyte chemoattractant protein-1

Abstract

Accumulating evidence suggests that spinal astrocytes play an important role in the genesis of persistent pain, by increasing the activity of spinal cord nociceptive neurons, i.e., central sensitization. However, direct evidence of whether activation of astrocytes is sufficient to induce chronic pain symptoms is lacking. We investigated whether and how spinal injection of activated astrocytes could produce mechanical allodynia, a cardinal feature of chronic pain, in naïve mice. Spinal (intrathecal) injection of astrocytes, which were prepared from cerebral cortexes of neonatal mice and briefly stimulated by tumor necrosis factor-alpha (TNF-α), induced a substantial decrease in paw withdrawal thresholds, indicating the development of mechanical allodynia. This allodynia was prevented when the astrocyte cultures were pretreated with a peptide inhibitor of c-Jun N-terminal kinase (JNK), D-JNKI-1. Of note a short exposure of astrocytes to TNF-α for 15 min dramatically increased the expression and release of the chemokine monocyte chemoattractant protein-1 (MCP-1), even 3 h after TNF-α withdrawal, in a JNK-dependent manner. In parallel, intrathecal administration of TNF-α induced MCP-1 expression in spinal cord astrocytes. In particular, mechanical allodynia induced by TNF-α-activated astrocytes was reversed by a MCP-1 neutralizing antibody. Finally, pretreatment of astrocytes with MCP-1 siRNA attenuated astrocytes-induced mechanical allodynia. Taken together, our results suggest that activated astrocytes are sufficient to produce persistent pain symptom in naïve mice by releasing MCP-1.

Figure 1

Intrathecal injection of TNF-α-treated astrocytes induces persistent mechanical allodynia in naive mice via the activation of JNK. (A) Experimental protocol showing the preparation of astrocytes for intrathecal injection. The differentiated astrocytes were incubated with TNF-α (10 ng/ml, 15 min), or TNF-α together with the JNK inhibitor D-JNKI-1 (20 µM), then collected and resuspended for intrathecal injection into naïve mice. (B) Paw withdrawal thresholds after intrathecal injection of control astrocytes, TNF-α-treated astrocytes, or astrocytes treated with TNF-α plus D-JNKI-1. ^Δ P<0.05, compared with baseline; * P<0.05, compared with control group; ^# P<0.05, compared with TNF-α-treated group. n = 8 mice.

Figure 2

A brief exposure of astrocytes to TNF-α induces robust MCP-1 expression and release via the activation of JNK. (A) Experimental protocol of astrocyte preparation. (B) MCP-1 expression and release in astrocytes after a brief stimulation with TNF-α (10 ng/ml, 15 min). TNF-α evokes MCP-1 expression and release even 3 h after TNF-α withdrawal. D-JNKI-1 pretreatment (20 µM) reduces TNF-α-triggered MCP-1 expression and release. *P<0.05, *** P<0.001, n = 3.

Figure 3

Intrathecal injection of TNF-α (20 ng) induces MCP-1 expression in spinal astrocytes. (A, B) Immunohistochemistry showing MCP-1 expression in the spinal cord dorsal horn of mice receiving intrathecal injection of PBS (A) and TNF-α (B). (C, D) Immunohistochemistry showing GFAP expression in the spinal cord dorsal horn of mice receiving intrathecal injection of PBS (C) and TNF-α (D). (E, F) Double staining showing the colocalization of MCP-1 and GFAP in the spinal cord dorsal horn of mice receiving intrathecal injection of PBS (E) and TNF-α (F). E is the merge of A and C and F is the merge of B and D. Animals were sacrificed 3 hours after PBS or TNF-α injection. Scale bar, 100 µm.

Figure 4

Effects of intrathecal administration of the MCP-1 neutralizing antibody on mechanical allodynia induced by intrathecal injection of TNF-α-activated astrocytes (A) and non-activated control astrocytes (B). Note that the MCP-1 neutralizing antibody only reverses allodynia induced by the activated astrocytes. * P<0.05, compared with control serum; ^Δ P<0.05, compared with baseline, n = 4 mice.

Figure 5

MCP-1 siRNA treatment in astrocytes reduces TNF-α-induced MCP-1 expression and release and astrocytes-induced mechanical allodynia. (A, B) pretreatment of MCP-1 siRNA dose-dependently reduces TNF-α-induced MCP-1 expression (A) and release (B) in cultured astrocytes. *, P<0.05; **, P<0.01; ***, P<0.001. n = 3 separate cultures from different mice. (C) Mechanical allodynia induced by intrathecal injection of TNF-α-activated astrocytes. Pretreatment of astrocytes with MCP-1 siRNA reduces allodynia induced by activated astrocytes. *, P<0.05; **, P<0.01 vs non-targeting siRNA control. n = 5 mice.

Figure 6

Iba1 immunostaining in the spinal cord dorsal horn of naïve mice (A) and mice receiving intrathecal injection of control astrocytes (B) or TNF-α treated astrocytes (C). The morphology and density of Iba1-labeled microglia in the spinal cord are comparable among different groups. (D) Intensity of Ibal staining in the spinal cord dorsal horn. N.S., no significance, n=4 mice. Mice were sacrificed 48 hours after the intrathecal injections. Scale bar, 200µm