Neuroanatomical evidence and a mouse calcitonin gene-related peptide model in line with human functional magnetic resonance imaging data support the involvement of peptidergic Edinger-Westphal nucleus in migraine

Abstract

The urocortin 1 (UCN1)-expressing centrally projecting Edinger-Westphal (EWcp) nucleus is influenced by circadian rhythms, hormones, stress, and pain, all known migraine triggers. Our study investigated EWcp's potential involvement in migraine. Using RNAscope in situ hybridization and immunostaining, we examined the expression of calcitonin gene-related peptide (CGRP) receptor components in both mouse and human EWcp and dorsal raphe nucleus (DRN). Tracing study examined connection between EWcp and the spinal trigeminal nucleus (STN). The intraperitoneal CGRP injection model of migraine was applied and validated by light-dark box, and von Frey assays in mice, in situ hybridization combined with immunostaining, were used to assess the functional-morphological changes. The functional connectivity matrix of EW was examined using functional magnetic resonance imaging in control humans and interictal migraineurs. We proved the expression of CGRP receptor components in both murine and human DRN and EWcp. We identified a direct urocortinergic projection from EWcp to the STN. Photophobic behavior, periorbital hyperalgesia, increased c-fos gene-encoded protein immunoreactivity in the lateral periaqueductal gray matter and trigeminal ganglia, and phosphorylated c-AMP-responsive element binding protein in the STN supported the efficacy of CGRP-induced migraine-like state. Calcitonin gene-related peptide administration also increased c-fos gene-encoded protein expression, Ucn1 mRNA, and peptide content in EWcp/UCN1 neurons while reducing serotonin and tryptophan hydroxylase-2 levels in the DRN. Targeted ablation of EWcp/UCN1 neurons induced hyperalgesia. A positive functional connectivity between EW and STN as well as DRN has been identified by functional magnetic resonance imaging. The presented data strongly suggest the regulatory role of EWcp/UCN1 neurons in migraine through the STN and DRN with high translational value.

Figure 1.

Calcitonin receptor (Calcr) and receptor activity–modifying protein 1 (Ramp1) mRNA expression. (A) Representative fluorescence images showing the Calcr (white) and Ramp1 (red) mRNA transcripts coexpressed with urocortin 1 peptide (UCN1, green) in the mouse centrally projecting Edinger–Westphal (EWcp) nucleus. (B) Neurons (neuronal marker NeuN, green) of the mouse spinal trigeminal nucleus express both of Calcr (white) and Ramp1 (red) mRNA. (C) Representative fluorescence images showing the CALCR (white) and RAMP1 (red) mRNA transcripts coexpressed with UCN1 (green) in the human EWcp. (D) In the mouse dorsal raphe nucleus (DRN), the serotonin (5-HT, green) immunoreactive cells also contained Calcr (white) and Ramp1 (red) mRNA transcripts. (E) In the human DRN, the 5-HT (green) immunoreactive cells also contained CALCR (white) and RAMP1 (red) mRNA transcripts. The yellow cytoplasmic area in E corresponds to lipofuscin autofluorescence. Nuclear counterstaining was performed with 4′,6-diamidino-2-phenylindole (DAPI, blue).

Figure 2.

Calcitonin-like receptor (CLR) expression. (A–A″) Representative immunofluorescence images showing the CLR (red) coexpressed with urocortin 1 peptide (UCN1, green) in the mouse centrally projecting Edinger–Westphal (EWcp) nucleus. (B–B″) Representative immunofluorescence images showing the CLR (red) coexpressed with urocortin 1 peptide (UCN1, green) in the human EWcp. (C–C″) In the mouse dorsal raphe nucleus (DRN), the serotonin (5-HT, green) immunoreactive cells also contained CLR (red). (D–D″) In the human DRN, the 5-HT (green) immunoreactive cells coexpressed CLR (red).

Figure 3.

Calcitonin gene–related peptide receptor component (Crcp) mRNA expression. (A) Representative fluorescence images showing the Crcp mRNA transcripts (red) coexpressed with urocortin 1 peptide (UCN1, green) in the mouse centrally projecting Edinger–Westphal (EWcp) nucleus. (B) In the mouse dorsal raphe nucleus (DRN), the tryptophan hydroxylase-2 (TPH2, green) immunoreactive cells also contained Crcp mRNA transcripts (red). (C) Neurons (neuronal marker NeuN, green) of the mouse spinal trigeminal nucleus express also Crcp mRNA (red). (D) Representative fluorescence images showing the CRCP mRNA transcripts (red) coexpressed with UCN1 (green) in the human EWcp. (E) In the human DRN, the TPH2 (green) immunoreactive cells also contained CRCP mRNA transcripts (red). Nuclear counterstaining was performed with 4′,6-diamidino-2-phenylindole (DAPI, blue).

Figure 4.

Calcitonin receptor-like receptor protein expression in mouse spinal trigeminal nucleus (STN). (A and B) The NeuN (green) labeling combined with CLR staining (red) revealed that a part of the STN neurons show also CLR immunoreactivity. Yellow arrowheads in the higher magnification images (B) depict the boxed areas in (A). (C and D) GFAP-CLR double labeling demonstrates that the GFAP (green)-immunoreactive astrocytes show also some CLR (red) immunopositivity. White arrowheads in the high magnification images in (D) demonstrating the marked areas in (C). The overlay images (A″–D″) in the right column demonstrate the colocalization of the labeled antigens. CLR, calcitonin-like receptor; GFAP, Glial fibrillary acidic protein.

Figure 5.

Urocortinergic afferentation from the centrally projecting Edinger–Westphal (EWcp) nucleus to the spinal trigeminal nucleus (STN). (A) Representative images showing the injection site of AAV8 Syn-enhanced green fluorescent protein (EGFP) by GFP (green) immunofluorescence in the EWcp urocortinergic (urocortin 1 [UCN1], white) neurons. All of the virus-infected cells are neurons (neuronal marker: NeuN, red), note the UCN1 and NeuN colocalization (yellow). (B) Fluorescence labeling for cholera toxin subunit B (CTB) in the dorsal horn of STN. (C) The colocalization of UCN1 (red) with GFP (green) in a nerve fiber in the STN and (D) the colocalization of UCN1 (red) with CTB (green) in the EWcp neurons. (E) Cells coexpressing neuronal marker mRNA (NeuN, white) and corticotropin-releasing hormone receptor 1 mRNA (Crhr1, red) as well as (F) corticotropin-releasing hormone receptor 2 (Crhr2, red) receiving a UCN1 (green)-positive afferentation in the STN. Nuclear counterstaining was performed with 4′,6-diamidino-2-phenylindole (DAPI, blue).

Figure 6.

Model validation. (A) Columns show the ratio of time spent in the dark compartment of the light–dark box device, 30 minutes after saline (control) or calcitonin gene–related peptide (CGRP) injection (n = 11-15, *P = 0.02; Student t test). (B) Columns show the facial withdrawal threshold (g) in von Frey test, 30 minutes after saline (control) or CGRP injection (n = 13, ***P = 8.25531 × 10⁻¹³; Student t test). (C) Columns show the number of FOS positive neurons in the TRG (n = 6; ***P = 0.0004; Student t test). (C′ and C″) Representative fluorescence images showing the expression of FOS, as a marker of early neural activation in the trigeminal ganglion (TRG), 4 hours after saline (control) and CGRP injection. Nuclei of FOS-positive activated neurons (red) are highlighted by the white arrowheads. (D) Columns show the number of phosphorylated c-AMP–responsive element binding protein (P-CREB) positive neurons in the spinal trigeminal nucleus (STN) (n = 6; **P = 0.003; Student t test). (D′ and D″) Representative fluorescence images showing the expression of P-CREB, as a marker of neural activation in the STN, 4 hours after saline (control) and CGRP injection. Nuclei of P-CREB–positive activated neurons (red) are colocalized (yellow) with the neuronal marker (NeuN, green). (E) Columns show the number of FOS-positive neurons in the lateral periaqueductal gray matter (lPAG; n = 6; **P = 0.002; Student t test). (E′ and E″) Representative images showing the nuclei (brown dots) of activated FOS-positive neurons in the lPAG in control and CGRP-treated mice. FOS, c-fos gen–encoded protein; STN, spinal trigeminal nucleus.

Figure 7.

Response of the mouse centrally projecting Edinger–Westphal (EWcp) nucleus to calcitonin gene–related peptide (CGRP) treatment. (A) Columns show the number of FOS-immunoreactive neurons in the EWcp (n = 6; *P = 0.03; Student t test). (A′ and A″) Representative images showing the nuclei of the neurons positive (brown dots) for the early neural activation marker (FOS) in the EWcp, 4 hours after saline (control) or CGRP injection. (B) Quantitative evaluation of the ratio of urocortin 1 (UCN1)/FOS double-labelled neurons from all FOS-immunoreactive neurons in the EWcp revealed the urocortinergic identity of activated (FOS-immunoreactive) neurons. (B′ and B″) Representative fluorescence images showing the colocalization of UCN1 (red) and FOS (green) in the EWcp of control and CGRP-injected mice. (C) Quantitative evaluation of Ucn1 mRNA-specific signal density (SSD) in the EWcp (n = 6; ***P = 0.00045; Student t test). (C′ and C″) Representative fluorescence images showing the expression of Ucn1 mRNA (green) in the EWcp in control and CGRP-treated mice. (D) Quantitative evaluation of UCN1 peptide SSD in the EWcp (n = 6; ***P = 0.0007; Student t test). (D′ and D″) Representative images illustrate UCN1 peptide (green) immunoreactivity in the EWcp in control and CGRP-injected mice. For nuclear counterstaining, 4′,6-diamidino-2-phenylindole (DAPI, blue) was used. a.u., arbitrary unit; FOS, c-fos gene–encoded protein.

Figure 8.

Serotonin (5-HT) and tryptophan hydroxylase-2 (TPH2) immunoreactivity in the dorsal raphe nucleus (DRN). (A) Specific signal density (SSD) of 5-HT in the DRN (n = 6; *P = 0.03; Student t test). (A′ and A″) Representative fluorescence images showing the 5-HT (green) immunoreactivity in the DRN, 4 hours after saline (control) and calcitonin gene–related peptide (CGRP) injection. (B) SSD of the TPH2 in the DRN (n = 6; *P = 0.01; Student t test). (B′ and B″) Representative fluorescence images illustrate TPH2 (red) immunosignal in the DRN. a.u., arbitrary unit.

Figure 9.

Ablation of urocortinergic (urocortin 1 [UCN1]) neurons in the centrally projecting Edinger–Westphal (EWcp) nucleus by leptin-saporin (LS). (A) Columns show the number of UCN1 neurons in the EWcp (n = 11; ***P = 0.0002; Student t test) in controls and LS-injected groups. (A′ and A″) Representative images showing the UCN1 neurons (brown) in the EWcp, in control and LS-injected mice. (B) Columns show the facial withdrawal threshold (g) in von Frey test, 30 minutes after saline or calcitonin gene–related peptide (CGRP) injection in control and LS-injected mice (n = 11, ***P = 0.0001; Tukey post hoc test upon repeated-measures ANOVA. ANOVA, analysis of variance.

Figure 10.

Functional connectivity matrix of Edinger–Westphal nucleus. Blue color represents functional connectivity of Edinger–Westphal nucleus with other brain areas: (A) sagittal reconstruction, lateral view; (B) left anterior aspect; (C) right posterior view. Significance threshold was cluster-level p(FWE) < 0.05 including at least 10 contiguous voxels. Results are corrected for age and sex. Statistical maps were visualized on the MNI 152 template brain provided in MRIcroGL (http://www.mccauslandcenter.sc.edu/mricrogl/). MNI, Montreal Neurological Institute.

Figure 11.

Brain clusters where functional connectivity of Edinger–Westphal nucleus positively correlated to migraine frequency. The green area represents the first cluster containing the angular gyrus and superior temporal gyrus, whereas red area is the other cluster containing the middle frontal and triangular part of inferior frontal gyri. Significance threshold was cluster-level p(FWE) < 0.05 including at least 10 contiguous voxels. Results are corrected for age and sex. Statistical maps were visualized on the MNI 152 template brain provided in MRIcroGL (http://www.mccauslandcenter.sc.edu/mricrogl/). MNI, Montreal Neurological Institute.