Effective control of neuropathic pain by transient expression of hepatocyte growth factor in a mouse chronic constriction injury model

Abstract

Hepatocyte growth factor (HGF) is a multifunctional protein that contains angiogenic and neurotrophic properties. In the current study, we investigated the analgesic effects of HGF by using a plasmid DNA that was designed to express 2 isoforms of human HGF-pCK-HGF-X7 (or VM202)-in a chronic constriction injury (CCI) -induced mouse neuropathic pain model. Intramuscular injection of pCK-HGF-X7 into proximal thigh muscle induced the expression of HGF in the muscle, sciatic nerve, and dorsal root ganglia (DRG). This gene transfer procedure significantly attenuated mechanical allodynia and thermal hyperalgesia after CCI. Injury-induced expression of activating transcription factor 3, calcium channel subunit α2δ1, and CSF1 in the ipsilateral DRG neurons was markedly down-regulated in the pCK-HGF-X7-treated group, which suggested that HGF might exert its analgesic effects by inhibiting pain-mediating genes in the sensory neurons. In addition, suppressed CSF1 expression in DRG neurons by pCK-HGF-X7 treatment was accompanied by a noticeable suppression of the nerve injury-induced glial cell activation in the spinal cord dorsal horn. Taken together, our data show that pCK-HGF-X7 attenuates nerve injury-induced neuropathic pain by inhibiting pain-related factors in DRG neurons and subsequent spinal cord glial activation, which suggests its therapeutic efficacy in the treatment of neuropathic pain.-Nho, B., Lee, J., Lee, J., Ko, K. R., Lee, S. J., Kim, S. Effective control of neuropathic pain by transient expression of hepatocyte growth factor in a mouse chronic constriction injury model.

Keywords: DRG; VM202; gene therapy; microglia; plasmid DNA.

Figure 1

A) Diagram of pCK-HGF-X7 used in the study. The genomic–cDNA hybrid sequence of HGF is present in the pCK backbone as reported by Lee et al. (23). Major immediate–early (IE) region of HCMV (pink arrow, boxes, and wavy lines); HGF-X7 (orange box), which consists of genomic–cDNA hybrid sequence of HGF, including a truncated form of intron 4 (blue boxes and wavy lines); polyadenylation signal of bovine hormone gene (green box); the kanamycin resistant gene, phosphotransferase (purple box); and ColE1, the origin of replication, from Escherichia coli (yellow box). B) Intramuscular injection of pCK-HGF-X7 produces HGF proteins in the muscle, sciatic nerve, and DRG. Peripheral neuropathy was induced in 5-wk-old ICR mice by CCI, and 200 μg of pCK-HGF-X7 was injected intramuscularly on d 0. Time kinetics of hHGF expression. Skeletal muscles around the injection site were isolated at appropriate time points, and total proteins were prepared, followed by ELISA specific for hHGF (n = 4). C) hHGF expression in the peripheral nervous system. DRGs and sciatic nerves were isolated on d 4 after CCI and subjected to hHGF ELISA. In sham-treated and control vector lacking the HGF sequence (pCK) injected groups, hHGF was not detected (n = 4).

Figure 2

Intramuscular injection of pCK-HGF-X7 ameliorates nerve injury–induced neuropathic pain in a mouse CCI model. A–C) pCK and pCK-HGF-X7 (200 μg) were injected intramuscularly on the day of CCI, and the pain sensitivity toward mechanical and thermal stimuli was measured at appropriate times by von Frey filaments (A, B) and Hargreaves (C) tests, respectively. Each group consisted of 6 mice, and >3 independent experiments were performed (means + sem). Sham-treated mice (■); CCI + pCK (○); and CCI + pCK-HGF-X7 (●). D) Different doses of pCK-HGF-X7 were injected intramuscularly, and, 2 wk later, the pain sensitivity toward mechanical and thermal stimulation was measured by von Frey filaments (●) and Hargreaves (○) tests, respectively (n = 6/group; mean + sem). *P < 0.05, 2-way ANOVA.

Figure 3

Repeated administration of pCK-HGF-X7 further attenuates neuropathic pain, depending on the time point of injection. pCK and pCK-HGF-X7 (200 μg) were delivered by intramuscular injection on the day of CCI, and the second injection was introduced, as indicated by the red arrows, after 1 (A), 2 (B), 3 (C), and 4 (D) wk. Pain sensitivity to a non-noxious mechanical stimuli was measured by von Frey filaments (n = 6/group; means + sem;). Sham-treated mice (■); CCI + pCK (○); CCI + pCK-HGF-X7 (1) (●); and CCI + pCK-HGF-X7 (2) (X).

Figure 4

Intramuscular pCK-HGF-X7 injection attenuates ATF3 expression in DRG neurons. Intramuscular injection of 200 μg of pCK or pCK-HGF-X7 and CCI were introduced, and injured DRGs (L4–L6) were isolated 4 d later. Total RNAs, proteins, and DRG sections were prepared to perform quantitative RT-PCR (A), Western blot analysis (B, C), and immunostaining (D), respectively, using an Ab specific to ATF3. The level of mRNA was normalized to the expression of the contralateral DRG, and HPRT was used as housekeeping gene (n = 3/group; means + sem). Scale bar, 50 μm. In Western blot analysis, β-actin was used as a loading control. C, contralateral; Ip, ipsilateral. *P < 0.05, **P < 0.01, ***P < 0.001 compared with sham-treated group; ^#P < 0.05, ^##P < 0.01 compared with CCI + pCK group (1-way ANOVA).

Figure 5

Intramuscular pCK-HGF-X7 injection attenuates α2δ1 expression in DRG neurons. A–C) Intramuscular injection of pCK or pCK-HGF-X7 (200 μg) and CCI were performed, and affected DRGs were prepared 4 d later for quantitative RT-PCR (A) and Western blot analysis (B, C). D, E) Similar procedures were performed for the analysis of sciatic nerve by Western Blot analysis. The level of mRNA was normalized to the expression of the contralateral DRG, and HPRT was used as housekeeping gene (n = 3/group; means + sem). In Western blot analysis, β-actin was used as a loading control (n = 3/group). C, contralateral; Ip, ipsilateral. *P < 0.05, **P < 0.01, ***P < 0.001 compared with sham-treated mice; 1-way ANOVA.

Figure 6

Intramuscular pCK-HGF-X7 injection attenuates CSF1 expression in DRG neurons. CCI was introduced, and 200 μg of pCK/pCK-HGF-X7 were transferred by intramuscular injection simultaneously. Ipsilateral DRGs (L4–L6) were isolated 4 d later. A, B) Total RNAs and proteins were prepared for quantitative RT-PCR (A) and ELISA (B) specific for CSF1, respectively. The level of mRNA was normalized to the expression of the contralateral DRG, and HPRT was used as housekeeping gene (n = 3/group; means + sem). *P < 0.05, **P < 0.01 compared with sham-treated mice; 1-way ANOVA.

Figure 7

Nerve injury–induced spinal cord microglial activations are attenuated by pCK-HGF-X7 injection. CCI and intramuscular injection of 200 μg of pCK or pCK-HGF-X7 were performed. Affected spinal cords were removed 4 d later, and cryocut sections were prepared. Spinal cord sections were subjected to immunohistochemistry assay by using Abs that were specific to Iba1 (red; marker for microglia; n = 4). Scale bars, 200 (top row), 20 (bottom row) μm. Contra, contralateral; Ipsi, ipsilateral.

Figure 8

Nerve injury–induced spinal cord astrocyte activations are attenuated by pCK-HGF-X7 injection. CCI and intramuscular injection of 200 μg of pCK or pCK-HGF-X7 were performed. Affected spinal cords were appropriately prepared 4 d later, and cryocut sections were performed. Spinal cord sections were subjected to immunohistochemistry assay by using Abs that were specific to glial fibrillary acidic protein (GFAP; red; marker for astrocyte). Scale bars, 200 (top row), 100 (bottom row) μm. Contra, contralateral; Ipsi, ipsilateral.

Figure 9

Nerve injury–induced spinal cord glial cell activations are attenuated by pCK-HGF-X7 injection. Four days after CCI and plasmid injection, the ipsilateral spinal cord dorsal horn was isolated and total RNAs were prepared, followed by quantitative RT-PCR analysis for respective genes. The level of mRNA of Cathepsin S (A), IRF8 (B), IRF5 (C), and IBA1 (D) was measured. They were normalized against GAPDH (glyceraldehyde 3-phosphate dehydrogenase; n = 3/group; means + sem). *P < 0.05, **P < 0.01 compared with sham-treated mice; 1-way ANOVA.