Nerve injury-induced changes in Homer/glutamate receptor signaling contribute to the development and maintenance of neuropathic pain

Abstract

While group 1 metabotropic glutamate receptors (mGluRs) and ionotropic N-methyl-d-aspartate (NMDA) receptors regulate nociception, the precise molecular mechanism(s) contributing to glutamate signaling in chronic pain remain unclear. Here we not only confirmed the key involvement of Homer proteins in neuropathic pain, but also distinguished between the functional roles for different Homer family members and isoforms. Chronic constriction injury (CCI) of the sciatic nerve induced long-lasting, time-dependent increases in the postsynaptic density expression of the constitutively expressed (CC) isoforms Homer1b/c and/or Homer2a/b in the spinal dorsal horn and supraspinal structures involved in nociception (prefrontal cortex, thalamus), that co-occurred with increases in their associated mGluRs, NR2 subunits of the NMDA receptor, and the activation of downstream kinases. Virus-mediated overexpression of Homer1c and Homer2b after spinal (intrathecal) virus injection exacerbated CCI-induced mechanical and cold hypersensitivity, however, Homer1 and Homer2 gene knockout (KO) mice displayed no changes in their neuropathic phenotype. In contrast, overexpression of the immediate early gene (IEG) Homer1a isoform reduced, while KO of Homer1a gene potentiated neuropathic pain hypersensitivity. Thus, nerve injury-induced increases in CC-Homers expression promote pain in pathological states, but IEG-Homer induction protects against both the development and maintenance of neuropathy. Additionally, exacerbated pain hypersensitivity in transgenic mice with reduced Homer binding to mGluR5 supports also an inhibitory role for Homer interactions with mGluR5 in mediating neuropathy. Such data indicate that nerve injury-induced changes in glutamate receptor/Homer signaling contribute in dynamic but distinct ways to neuropathic pain processing, which has relevance for the etiology of chronic pain symptoms and its treatment.

Keywords: Group 1 metabotropic glutamate receptors; Homer proteins; NMDA receptors; Neuropathic pain; Spinal cord.

Fig. 1.

Chronic constriction injury (CCI) of the sciatic nerve produces time-dependent changes in CC-Homers and glutamate receptors expression mainly within the postsynaptic density ([PSD] LP1; A, C, E), than within the P1 (nuclear and large debris) fraction (B, D, F) of the ipsilateral spinal dorsal horn (A, B), thalamus (C, D), and prefrontal cortex (E, F). Each graph represents summary of the change in protein expression obtained 1, 7, and 14 days after the CCI and expressed as a percentage of the average protein expression of the naïve controls. Data are presented as mean ± SEM, n = 10–12 of B6 mice per each experimental group. The intensity of the bands for each antibody was normalized with the intensity of its appropriate calnexin signal. The asterisk (*) denotes significance vs naïve controls; *P < 0.05 (one-way analysis of variance, followed by Tukey’s comparison post hoc test). Representative immunoblots correspond to the total protein levels of Homer1b/c, Homer2a/b, mGluR1a, mGluR5, NR2a, and NR2b.

Fig. 2.

Chronic constriction injury (CCI) of the sciatic nerve elevates protein expression of the members of the PI3K/MAPK signaling pathway mainly 1 and/or 2 weeks after the injury within the postsynaptic density ([PSD] LP1; A, C, E) and the P1 fraction (B, D, F) of the ipsilateral spinal dorsal horn (A, B), thalamus (C, D), and prefrontal cortex (E, F). Each graph represents summary of the change in protein expression obtained 1, 7, and 14 days after the CCI and expressed as a percentage of the average protein expression of the naïve controls. Data are presented as mean ± SEM, n = 10–12 of B6 mice per each experimental group. The intensity of the bands for each antibody was normalized with the intensity of its appropriate calnexin signal. The asterisk (*) denotes significance vs naïve controls; *P < 0.05 (one-way analysis of variance, followed by Tukey’s comparison post hoc test). Representative immunoblots correspond to the total protein levels of protein kinase C (PKC)ε, phospho-PKCε, PI3K, phospho(Tyr)p85α PI3K binding motif [p(Tyr)p85α], ERK1/2 and phospho-ERK1/2.

Fig. 3.

Adeno-associated vector (AAV)-mediated overexpression of CC-Homer isoforms potentiates neuropathic pain hypersensitivity. (A) Representative images of green fluorescent protein (hrGFP) epifluorescence (AAV-control vector) or immunostaining for the HA tag (AAV-Homer1c and -Homer2b cDNA vectors) using a specific antibody on L4–L6 spinal dorsal horn sections (80 μm) at 7–8 weeks after AAV infusion. HA staining was conducted also on spinal cord tissue from AAV-control animals as a control for nonspecific staining. Scale bar = 200 μm. (B–E) AAV-mediated overexpression of 2 CC-Homer isoforms, Homer1c (B, C), and Homer2b (D, E) after spinal delivery exacerbates the injury-induced mechanical (B, D) and cold (C, E) hypersensitivity in neuropathic pain. AAV-Homer1c and AAV-Homer2b were injected intrathecally in B6 mice. Mechanical sensitivity was assessed using von Frey filaments; cold sensitivity was determined by the acetone test. The measurements were assessed before chronic constriction injury (CCI) of the sciatic nerve and then every 2nd day for 2 weeks following the injury. The same control group was used in B and C, as well as in C and E. Data are presented as means ± SEM, n = 8–10 in each group. The asterisk (*) denotes significance vs control; *P < 0.05 (2-way analysis of variance, followed by Bonferroni comparison post hoc test).

Fig. 4.

Deletion of Homer1 and Homer2 does not affect the development and maintenance of neuropathic pain. Each graph demonstrates mechanical (A, C) or cold (B, D) hypersensitivity induced by chronic constriction injury (CCI) of the sciatic nerve in Homer1 and Homer2 gene knockout mice (KO), heterozygous (HET), and wild-type controls (WT). Mechanical sensitivity was assessed using von Frey filaments; cold sensitivity was determined by the acetone test. The measurements were assessed before CCI and then every 2nd day for 2 weeks following the injury. Data are presented as means ± SEM, n = 10–12 in each group. The asterisk (*) denotes significance vs WT; *P < 0.05 (2-way analysis of variance, followed by Bonferroni comparison post hoc test).

Fig. 5.

Homer1a protects against mechanical and cold hypersensitivity following chronic constriction injury (CCI) of the sciatic nerve. (A) Representative images of immunostaining for the HA tag (adeno-associated vector [AAV]-Homer1a cDNA vectors) using a specific antibody on L4–L6 spinal dorsal horn sections (80 μm) at 7–8 weeks after AAV infusion. HA staining was conducted also on spinal cord tissue from AAV-control animals as a control for nonspecific staining. Scale bar = 200 μm. (B–G) Each graph demonstrates mechanical and cold hypersensitivity induced by CCI in transgenic mice with disrupted binding of CC-Homer to mGluR5, mGluR5^F1128R (B, C), in Homer1a gene knockout mice (KO), heterozygous (HET), and their respective wild-type controls (WT) (D, E), and in B6 mice after intrathecal injection of AAVs carrying Homer1a (F, G). Mechanical sensitivity was assessed using von Frey filaments; cold sensitivity was determined by the acetone test. The measurements were assessed before CCI and then every 2nd day for 2 weeks following the injury. Data are presented as means ± SEM, n = 10–12 in each group. The asterisk (*) denotes significance vs WT/control; *P < 0.05 (2-way analysis of variance, followed by Bonferroni comparison post hoc test).