Xenobiotic Exposure and Migraine-Associated Signaling: A Multimethod Experimental Study Exploring Cellular Assays in Combination with Ex Vivo and In Vivo Mouse Models

Abstract

Background: Mechanisms for how environmental chemicals might influence pain has received little attention. Epidemiological studies suggest that environmental factors such as pollutants might play a role in migraine prevalence. Potential targets for pollutants are the transient receptor potential (TRP) channels ankyrin 1 (TRPA1) and vanilloid 1 (TRPV1), which on activation release pain-inducing neuropeptide calcitonin gene-related peptide (CGRP).

Objective: In this study, we aimed to examine the hypothesis that environmental pollutants via TRP channel signaling and subsequent CGRP release trigger migraine signaling and pain.

Methods: A calcium imaging-based screen of environmental chemicals was used to investigate activation of migraine pain-associated TRP channels TRPA1 and TRPV1. Based on this screen, whole-cell patch clamp and in silico docking were performed for the pesticide pentachlorophenol (PCP) as proof of concept. Subsequently, PCP-mediated release of CGRP and vasodilatory responses of cerebral arteries were investigated. Finally, we tested whether PCP could induce a TRPA1-dependent induction of cutaneous hypersensitivity in vivo in mice as a model of migraine-like pain.

Results: A total of 16 out of the 52 screened environmental chemicals activated TRPA1 at 10 or $100\mu M$. None of the investigated compounds activated TRPV1. Using PCP as a model of chemical interaction with TRPA1, in silico molecular modeling suggested that PCP is stabilized in a lipid-binding pocket of TRPA1 in comparison with TRPV1. In vitro, ex vivo, and in vivo experiments showed that PCP induced calcium influx in neurons and resulted in a TRPA1-dependent CGRP release from the brainstem and dilation of cerebral arteries. In a mouse model of migraine-like pain, PCP induced a TRPA1-dependent increased pain response ($N_{\text{total}}=144$).

Discussion: Here we show that multiple environmental pollutants interact with the TRPA1-CGRP migraine pain pathway. The data provide valuable insights into how environmental chemicals can interact with neurobiology and provide a potential mechanism for putative increases in migraine prevalence over the last decades. https://doi.org/10.1289/EHP12413.

Figure 1.

Calcium imaging of environmental chemicals using hTRPA1-HEK and hTRPV1-HEK cell lines. Calcium imaging was performed using fluo-4 following acute exposure to 52 chemicals. Cells were seeded at a density of 25,000 cells/well in a 96-well plate 1 day prior to imaging and recordings were performed every 0.3 s for 70.6 s using a NOVOstar microplate reader. (A) Summary of calcium imaging data of chemical screen ($100\mathrm{\mu }\mathrm{M}$). Numeric data are found in Table S1. (B) Summary of calcium imaging data of pesticide screen ($10\mathrm{\mu }\mathrm{M}$). Numeric data are found in Table S2. For (A) and (B) calcium imaging was performed using hTRPA1-HEK (yellow circle), hTRPV1-HEK (blue square), and control cells (green triangle). Responses were normalized to the ionomycin response ($10\mathrm{\mu }\mathrm{M}$). (C) Calcium imaging response normalized to supercinnamaldehyde (SCA) response for PCP and HCP in hTRPA1-HEK. Numeric data are found in Table S3. (A–C) The screen was repeated twice with 3–6 technical replicates within each plate. The data are presented as a weighted mean from repeated experiments, and whiskers represent the combined SD from repeated screens. TRPA1 agonist supercinnamaldehyde (SCA) and TRPV1 agonist capsaicin (caps) were tested in all three cell lines. For abbreviations of chemicals see Tables 1 and 2. Note: HCP, hexachlorophene; PCP, pentachlorophenol.

Figure 2.

Whole-cell patch clamp recordings from hTRPA1-HEK cells, control cells and neurons isolated from the mouse trigeminal ganglion. Whole-cell voltage clamp recordings using a MultiClamp 700B amplifier and MultiClamp Commander. Currents were activated using a voltage ramp protocol from $-100$ to $+100\text{\hspace{0.17em}mV}$ over 325 ms from a holding potential of $-60\text{\hspace{0.17em}mV}$. Recordings were performed at 36°C in control cells (A–C) and hTRPA1-HEK cells (D–F). (A) Representative currents recorded for control cells elicited by a ramp from $-100$ to $100\text{\hspace{0.17em}mV}$ in absence and presence of $10\mathrm{\mu }\mathrm{M}$ PCP and $10\mathrm{\mu }\mathrm{M}\text{\hspace{0.17em}PCP}+100\mathrm{\mu }\mathrm{M}$ HC-030031 (HC) (TRPA1 antagonist). (B) A representative current recording from control cells. The current amplitude at $-100$ and $100\text{\hspace{0.17em}mV}$ is shown as a function of time and in absence and presence of $10\mathrm{\mu }\mathrm{M}$ PCP as well as $10\mathrm{\mu }\mathrm{M}\text{\hspace{0.17em}PCP}+100\mathrm{\mu }\mathrm{M}\text{\hspace{0.17em}HC}$. (C) Mean current amplitude ($\pm \text{SEM}$) in control cells at $100\text{\hspace{0.17em}mV}$ in absence and presence of $10\mathrm{\mu }\mathrm{M}$ PCP and $10\mathrm{\mu }\mathrm{M}\text{\hspace{0.17em}PCP}+100\mathrm{\mu }\mathrm{M}$ HC ($n=4$). One-way ANOVA followed by Bonferroni’s multiple comparisons test. Numeric data are found in Table S5. (D) Representative currents recorded from hTRPA1-HEK cells elicited by a ramp from $-100$ to $100\text{\hspace{0.17em}mV}$ in absence and presence of $10\mathrm{\mu }\mathrm{M}$ PCP and $10\mathrm{\mu }\mathrm{M}\text{\hspace{0.17em}PCP}+100\mathrm{\mu }\mathrm{M}$ HC. (E) Currents recorded from hTRPA1-HEK cells using a ramp from $-100$ to $100\text{\hspace{0.17em}mV}$ over 325 ms from a holding potential of $–60\text{\hspace{0.17em}mV}$. The current amplitude at $-100$ and $100\text{\hspace{0.17em}mV}$ is shown as a function of time and in absence and presence of $10\mathrm{\mu }\mathrm{M}$ PCP as well as $10\mathrm{\mu }\mathrm{M}\text{\hspace{0.17em}PCP}+100\mathrm{\mu }\mathrm{M}$ HC. (F) Mean current amplitude ($\pm \text{SEM}$) in TRPA1- HEK at $100\text{\hspace{0.17em}mV}$ in absence and presence of $10\mathrm{\mu }\mathrm{M}$ PCP and $10\mathrm{\mu }\mathrm{M}\text{\hspace{0.17em}PCP}+100\mathrm{\mu }\mathrm{M}$ HC ($n=3$). One-way ANOVA followed by Bonferroni’s multiple comparisons test. Numeric data are found in Table S5. (G) Membrane currents at $-60\text{\hspace{0.17em}mV}$ from hTRPA1-HEK cells with the indicated PCP concentrations [${\text{EC}}_{50}$ of $15.05\mathrm{\mu }\mathrm{M}$ (95% CI: $6.429\mathrm{\mu }\mathrm{M}$ to $63.19\mathrm{\mu }\mathrm{M}$)] ($n=3–6$). (H) Representative whole-cell current recording of a mouse trigeminal ganglia neuron exposed to $10\mathrm{\mu }\mathrm{M}$ PCP. The current was measured at a constant voltage of $-60\text{\hspace{0.17em}mV}$. (I) Mean current amplitude ($\pm \text{SEM}$) in neurons in absence or presence of $10\mathrm{\mu }\mathrm{M}$ PCP ($n=10$). Wilcoxon matched-pairs signed rank test. Numeric data are found in Table S6. The current measurements of the mouse trigeminal ganglia neurons were performed at 22°C at a constant potential of $-60\text{\hspace{0.17em}mV}$. *$p<0.05$, **$p<0.01$, ***$p<0.001$, ****$p<0.0001$. Note: ANOVA, analysis of variance; PCP, pentachlorophenol; SEM, standard error of the mean.

Figure 3.

Molecular modeling of pentachlorophenol in lipid pocket of hTRPA1 and hTRPV1. (A) 3D representation of PCP in the lipid pocket of hTRPA1. Two monomeres of hTRPA1 are represented in green and magenta. The PCP is colored in cyan. (B) 2D representation of the docking pose of PCP in hTRPA1. PCP is stabilized through a polar interaction with E864 (green arrow). (C) 2D representation of the docking pose of PCP in hTRPV1. E570 is not present in the 2D representation, because it is located too far from PCP in the binding site. (B–C) The pink spheres are polar residues ($\$$) [with a red border for acidic (*)]. The green spheres are hydrophobic residues (#). The blue shadows are ligand exposure and receptor exposure. The green arrows are hydrogen bond interactions. Note: PCP, pentachlorophenol.

Figure 4.

CGRP release from trigeminal nucleus caudalis of wild type and $Trpa{\mathit{1}}^{-/-}$ mice after supercinnamaldehyde, PCP, and vehicle treatment. (A–D) The TNC from WT and $Trpa{\mathit{1}}^{-/-}$ mice was isolated and washed before incubation (37°C) with increasing concentrations of either SCA (A–B) or PCP (C–D) and corresponding vehicle treatment. Following 10 min of incubation, a sample was collected and the released CGRP amount was measured by ELISA. (A–B) CGRP release after increasing concentrations ($1\mathrm{\mu }\mathrm{M}$, $10\mathrm{\mu }\mathrm{M}$, and $100\mathrm{\mu }\mathrm{M}$) of SCA (yellow circle) and vehicle (0.1%, 0.01% and 0.001% DMSO) (black triangle) from the TNC from (A) WT (SCA: $n=7$, vehicle: $n=6$) and (B) $Trpa{\mathit{1}}^{-/-}$ mice (SCA: $n=6$, vehicle: $n=6$). Numeric data are found in Tables S7–S8. (C–D) CGRP release from the TNC from (C) WT (PCP: $n=8$, vehicle: $n=6$) and (D) $Trpa{\mathit{1}}^{-/-}$ mice (PCP: $n=7$, vehicle: $n=6$) after stimulation with increasing concentrations ($1\mathrm{\mu }\mathrm{M}$, $10\mathrm{\mu }\mathrm{M}$, and $100\mathrm{\mu }\mathrm{M}$) of PCP (green circle) and vehicle (0.1%, 0.01% and 0.001% DMSO) (black triangle). Numeric data are found in Tables S9–S10. (A–D) Data are normalized to the basal response and presented as mean CGRP release with SEMs. Nonnormalized data are found in Tables S11–S14. Multiple unpaired $t$ tests with post hoc Holm-Šidák correction. (E) The BA was isolated from WT and $Trpa{\mathit{1}}^{-/-}$ mice following mounting on wire myographs. Dilation of the BA shown as percent of precontraction to increasing concentrations of PCP (${10}^{-7}\mathrm{M}$, $3\times {10}^{-7}\mathrm{M}$, ${10}^{-6}\mathrm{M}$, $3\times {10}^{-6}\mathrm{M}$, ${10}^{-5}\mathrm{M}$, and $3\times {10}^{-5}\mathrm{M}$) from WT (black triangle, $n=4$) and $Trpa{\mathit{1}}^{-/-}$ mice (blue circle, $n=4$), and in presence of $1\mathrm{\mu }\mathrm{M}$ antibodies directed against human CGRP (CGRP ab) (green upside-down triangle, $n=4$) from WT mice. Numeric data are found in Table S15. Two-way ANOVA. *$p<0.05$, **$p<0.01$, ***$p<0.001$, ****$p<0.0001$. Note: ANOVA, analysis of variance; BA, basal artery; CGRP, calcitonin gene-related peptide; ELISA, enzyme-linked immunosorbent assay; PCP, pentachlorophenol; SCA, supercinnamaldehyde; SEM, standard error of the mean; TNC, trigeminal nucleus caudalis; WT, wild type.

Figure 5.

Acute and chronic hind-paw sensitivity and motor function in mice treated with PCP or vehicle. Acute and chronic sensitivity to tactile stimulation on the plantar area is measured with von Frey filaments. Acute hypersensitivity (A, D) was measured 2 h and 4 h post drug administration, whereas chronic hypersensitivity (B) was measured the following day prior to drug administration. On completion of the von Frey experiments, the motor function of the mice was examined using a rotarod where the time spent on a horizontal rotating rod was measured. Mice were placed on the rotarod with a start speed of $0\text{\hspace{0.17em}rpm}$, which was increased to $30\text{\hspace{0.17em}rpm}$ with a ramp of 45 s and terminated after 150 s. The first cohort of mice is presented in A–C, and the second cohort including $Trpa{\mathit{1}}^{-/-}$ mice is presented in D–E. (A) Acute effect 2 h and 4 h post PCP ($15\mathrm{mg}/\mathrm{kg}$ p.o. in 0.5% CMC in water, $n=16$) (blue circle) or Veh (p.o., $n=16$) (black triangle) administration in WT mice. Sensitivity was measured with von Frey filaments and calculated as 50% withdrawal threshold (g). Two-way ANOVA with Šidák’s test. (B) Basal response (before PCP administration) in WT mice following daily administration of PCP ($15\mathrm{mg}/\mathrm{kg}$ p.o. in 0.5% CMC in water, $n=16$) or vehicle (p.o., $n=16$) for 10 d. Sensitivity was measured with von Frey filaments and calculated as 50% withdrawal threshold (g). Two-way ANOVA with Šidák’s test. (C) Locomotor performance was tested using a rotarod. The time spent on a rotating rod was measured in seconds, with a maximal duration of 150 s. Motor function was examined 2 and 4 h after PCP ($n=16$) or Veh ($n=16$) administration and 10 min after (Mida) (i.p. $n=16$) or saline (i.p., $n=16$) administration in WT mice. Data are presented as individual data points and medians. Multiple Mann-Whitney tests with Holm-Šidák post hoc comparison (D) Acute effect 2 h and 4 h post PCP ($15\mathrm{mg}/\mathrm{kg}$ p.o. in 0.5% CMC in water, $n=16$) or vehicle (p.o., $n=16$) administration in WT and $Trpa{\mathit{1}}^{-/-}$ mice. Sensitivity was measured with von Frey filaments and calculated as 50% withdrawal threshold (g). Two-way ANOVA with post hoc Dunnett’s correction. (E) Locomotor performance was tested using a rotarod. The time spent on a rotating rod was measured in seconds with a maximal duration of 150 s. Motor function was examined 2 and 4 h after PCP ($n=16$) or vehicle ($n=16$) administration in WT and PCP ($n=16$) administration $Trpa{\mathit{1}}^{-/-}$ mice, and 10 min after Mida (i.p., $n=24$) or saline (i.p., $n=24$) administration. Data are presented as individual data points and medians. Kruskal-Wallis with Dunn’s post hoc test for 2 h and 4 h data and Mann-Whitney for controls. (A, D) Data are presented as individual data points with $\text{means}\pm \text{SEMs}$ and calculated as 50% withdrawal threshold (g) after SQRT. (B) Data are presented as $\text{means}\pm \text{SEMs}$ and calculated as 50% withdrawal threshold (g) after SQRT. Numeric data for A–E is found in Tables S16–S20. *$p<0.05$, **$p<0.01$, ***$p<0.001$, ****$p<0.0001$. Note: ANOVA, analysis of variance; Mida, midazolam; PCP, pentachlorophenol; SEM, standard error of the mean; SQRT, square root transformation; Veh, vehicle; WT, wild type.

Figure 6.

Schematic illustration of the proposed signaling pathway involved in PCP-induced hypersensitivity. Interpretation of data from Figures 1–5. PCP was shown to activate TRPA1 with subsequent release of CGRP, resulting in dilation of blood vessels. PCP administration also induced hypersensitivity in mice via TRPA1. Created with Biorender.com. Note: CGRP, calcitonin gene-related peptide; CGRP-R, CGRP receptor; PCP, pentachlorophenol; TNC, trigeminal nucleus caudalis.