OnabotulinumtoxinA effects on trigeminal nociceptors

Abstract

Background

OnabotulinumtoxinA (onabotA) is approved globally for prevention of chronic migraine; however, the classical mechanism of action of onabotA in motor and autonomic neurons cannot fully explain the effectiveness of onabotulinumtoxinA in this sensory neurological disease. We sought to explore the direct effects of onabotulinumtoxinA on mouse trigeminal ganglion sensory neurons using an inflammatory soup-based model of sensitization.

Methods

Primary cultured trigeminal ganglion neurons were pre-treated with inflammatory soup, then treated with onabotulinumtoxinA (2.75 pM). Treated neurons were used to examine transient receptor potential vanilloid subtype 1 and transient receptor potential ankyrin 1 cell-surface expression, calcium influx, and neuropeptide release.

Results

We found that onabotulinumtoxinA cleaved synaptosomal-associated protein-25 kDa in cultured trigeminal ganglion neurons; synaptosomal-associated protein-25 kDa cleavage was enhanced by inflammatory soup pre-treatment, suggesting greater uptake of toxin under sensitized conditions. OnabotulinumtoxinA also prevented inflammatory soup-mediated increases in TRPV1 and TRPA1 cell-surface expression, without significantly altering TRPV1 or TRPA1 protein expression in unsensitized conditions. We observed similar inhibitory effects of onabotulinumtoxinA on TRP-mediated calcium influx and TRPV1- and TRPA1-mediated release of calcitonin gene-related peptide and prostaglandin 2 under sensitized, but not unsensitized control, conditions.

Conclusions

Our data deepen the understanding of the sensory mechanism of action of onabotulinumtoxinA and support the notion that, once endocytosed, the cytosolic light chain of onabotulinumtoxinA cleaves synaptosomal-associated protein-25 kDa to prevent soluble N-ethylmaleimide-sensitive factor attachment protein receptor-mediated processes more generally in motor, autonomic, and sensory neurons.

Introduction

OnabotulinumtoxinA (onabotA; BOTOX®, Allergan, an AbbVie company, Irvine, CA, USA) is currently approved as a preventive treatment in those with chronic migraine. However, a deeper appreciation of the mechanism of action (MoA) of onabotA in migraine would contribute to understanding its treatment benefit. The botulinum neurotoxin type A (BoNT/A) complex is a 900 kDa protein complex composed of the 150 kDa neurotoxin protein, which is made up of an ∼100 kDa heavy chain (H_C/A) and an ∼50 kDa light chain (L_C/A), that is associated with several non-toxic neurotoxin-associated proteins (1). As part of BoNT/A’s classical MoA, the receptor binding domain of the H_C/A first binds to polysialoganglioside (PSG), synaptic vesicle protein 2 (SV2), and potentially fibroblast growth factor receptor 3 (FGFR3) receptors on peripheral nerve terminals before being endocytosed (2,3). Once endosome-bound, the L_C/A dissociates from the H_C/A and translocates through the vesicular membrane into the cytosol where it enzymatically cleaves synaptosomal-associated protein-25 kDa (SNAP25), preventing soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE)-mediated exocytosis of small synaptic vesicles containing acetylcholine. This interference with the fundamental vesicular recycling process is ubiquitous and reflects the classical MoA. In peripheral motor and autonomic neurons, this leads to temporarily decreased muscle and/or gland activation and underpins the effects and utilization of BoNT/A for myriad clinical conditions (2,4).

Since migraine — including chronic migraine — is considered a sensory disease, the acetylcholine inhibition-focused MoA of BoNT/A and subsequent muscle relaxation do not readily explain the effectiveness of onabotA in treating this disease. The treatment pattern includes injecting onabotA in the region of the peripheral trigeminal nerve (5). Although the pathophysiology of migraine has not been fully elucidated, current evidence suggests that the initiation of the headache phase of a migraine attack occurs when sustained activation of meningeal nociceptors sensitizes peripheral trigeminal neurons. This peripheral sensitization facilitates central sensitization and subsequent activation of multiple brain regions involved in processing sensory information and pain, leading to headache and cortically mediated symptoms of migraine (e.g., allodynia, phonophobia, photophobia) (6). Additionally, increasing evidence suggests that similar intracellular targets/mechanisms present in motor and autonomic neurons are also present in sensory neurons. Indeed, SNAP25 is expressed in sensory neurons (7,8) and BoNT/A cleavage of SNAP25 in primary sensory neurons inhibits SNARE-mediated processes that have been implicated in migraine pathophysiology, including neuropeptide release (e.g., calcitonin gene-related peptide [CGRP], substance P) and increased receptor surface expression (e.g., Transient Receptor Potential Vanilloid-subfamily member 1 [TRPV1], TRPA1) (9–17). Furthermore, a prior study showed that BoNT/A reduced TRPV1 expression in trigeminal ganglion (TG) neurons by interfering with plasma membrane trafficking, independent of transcriptional regulation (18). Thus, increasing evidence suggests that the MoA of BoNT/A is not limited to blocking neurotransmitter release (e.g., acetylcholine, norepinephrine, glutamate) (19–21); rather, once internalized, BoNT/A cleaves SNAP25 to prevent SNARE-mediated synaptic vesicle exocytosis more generally in neurons, agnostic of vesicular content (i.e., altering both neurotransmitter release and receptor trafficking).

The goal of this study was to deepen our current understanding of the onabotA MoA and address the extent to which onabotA alters SNARE-mediated processes in sensory neurons involving the migraine-relevant TRP channels TRPV1 and TRPA1 (9,15–17). To do this, we treated primary cultured TG neurons with an inflammatory soup (IS), a cocktail of inflammatory substances that has been shown in various models to produce sensitized conditions (22–28), alone or followed by exposure to onabotA. Treated neurons were then used to examine cell-surface expression of TRP channel protein and TRP channels’ response to selective activation (i.e., calcium influx, neuropeptide release), mechanisms generally associated with migraine (11,14–16,18,29). We show that onabotA cleaves SNAP25 within a subset of TRPV1- and TRPA1-expressing TG sensory neurons. Furthermore, in sensitized conditions, onabotA decreases the cell-surface expression of TRPV1 and TRPA1; their downstream signaling, as measured by Ca²⁺ influx; and the resulting release of neuro-inflammatory peptide and prostanoid mediators associated with migraine pathogenesis.

Methods

Primary trigeminal neuron culture

Wild-type C57BL/6J mice were purchased from Jackson Laboratory or bred in-house. Pirt-GCaMP3 mice were bred in-house (30). Male and female mice were housed in climate-controlled rooms on a 12/12-hour light/dark cycle with water and standardized rodent diet available ad libitum. Animal protocols were approved by the Duke University Institutional Animal Care and Use Committee (IACUC) in compliance with National Institutes of Health (NIH) guidelines.

Detailed culture methods are available in supplemental materials. Briefly, TG of mice pups (postnatal days [P]7-P21) were dissected, dissociated into single-cell suspension, and pooled. Cells were then treated, centrifuged, and re-suspended before being plated on Poly-D-lysine- and laminin-coated coverslips in DH10 Media. For every dissected pup, four wells were plated with the neuron suspension. After two hours, media was changed to Neurobasal Plus (ThermoFisher, A3582901) supplemented with B-27 Plus, Glutamax, CultureOne, Pen-Strep, GDNF (50 ng/mL), and NGF (50 ng/mL). After four to six days of culture, neurons were treated with inflammatory soup (IS) (0.5 µM each of bradykinin, serotonin, histamine, substance P and 0.05 µM of prostaglandin E2 [PGE2]; pH 7.4), which has been shown to induce sensitized conditions in animal models (22–26) and in vitro cell culture models (31,32). The next day, selected wells were treated with BoNT/A (onabotulinumtoxinA; onabotA; 2.75 pM [50 U/mL; 2.5 ng/mL]; Allergan, an AbbVie company, Irvine, CA, USA). This generated study treatment groups with no treatment, IS, onabotA, or IS + onabotA. Twenty-four or 48 hours later, neurons were examined by methods of protein-biochemistry (i.e., immunoblot, ELISA), immunocytochemistry, RT-qPCR, and ion channel function (Ca²⁺ imaging).

A random sampling of treated cells found that >95% of cells expressed the neuronal marker PGP9.5 (data not shown), indicating that the majority of cultured TG cells were neurons. Viability was tested using the ThermoFisher LIVE/DEAD cytotoxicity kit 24 hours after onabotA treatment and 48 hours after IS treatment; viability was 90–95% (data not shown).

Western blot

Full western blotting methodology is available in supplemental methods. Briefly, to measure onabotA cleavage of SNAP25 in TG neurons cultured from wild-type mice, treated cultured TG neurons were lysed using modified RIPA buffer, sonicated, and centrifuged to be used for western blot analysis. Protein was quantified for equal sample loading (BCA Protein Assay Kit, ThermoFisher Scientific), samples were electrophoresed on polyacrylamide SDS gels (Biorad), and proteins were transferred onto polyvinylidene difluoride (PVDF) membranes. Blots were blocked in buffer for one hour at room temperature (RT) before overnight incubation at 4°C with the following primary antibodies in blocking buffer: mouse anti-β-Actin (C4) (Santa Cruz Biotechnology, sc-47778) and recombinant human anti-SNAP25₁₉₇ (Allergan plc, Ab632), a highly selective antibody for the cleaved epitope of SNAP25 (33). Blots were washed, incubated with horseradish peroxidase-conjugated secondary antibodies (1:5000; Jackson ImmunoResearch), and developed by enhanced chemiluminescence.

Enzyme-linked immunosorbent assay (ELISA)

Forty-eight hours after onabotA treatment, TG neurons were incubated for 30 minutes at 37°C using 0.25 mL per well of Hanks' balanced salt solution (HBSS) containing 0.1% bovine serum albumin (BSA) for baseline measurements. Working dilutions of agonists were prepared in HBSS/BSA on the day of use from stock solutions to 1 µM for capsaicin (TRPV1) and to 10 µM for JT010 (TRPA1). After 30 minute stimulation with relevant agonist, remaining fluids were aspirated, and the cells lysed in RIPA buffer. The supernatants were stored at −20°C until the day of assay. CGRP and PGE2 levels were measured using commercially available ELISA kits (Cayman Chemical Company, 589001 and 514010) according to manufacturer’s instructions.

Immunocytochemistry

Trafficking of TRP channels to the plasma membrane occurs via SNARE-mediated regulated exocytosis and prior research suggests that the ability of BoNT/A to reduce TRPV1-mediated craniofacial pain is due to BoNT/A interfering with trafficking of TRP channels (18,34). To extend these findings, we examined the effect of onabotA on cell-surface expression of TRP channel protein (ecto-TRP channels staining) under sensitized conditions. TG neurons were treated as above before being incubated in the primary antibodies polyclonal anti-TRPV1 (1:50; Alomone Labs, ACC-030) or anti-TRPA1 (1:50; Alomone Labs, ACC-037) for 30 minutes at 37°C to label extracellular immunogen on live neurons. Neurons were then washed with 1X PBS, fixed with 4% paraformaldehyde (15 minutes), then permeabilized by 0.1% Triton X-100 in PBS (15 minutes). The neurons were blocked in 10% normal donkey serum in PBS/0.1% Triton X at RT before overnight incubation with human anti-SNAP25₁₉₇ (1:1000; Allergan, an AbbVie company, Ab632) and sheep anti-PGP9.5 (1:400; R&D Systems, AF6007) in blocking buffer. After three to five rinses with PBS, neurons were incubated for one hour with the secondary antibodies (1:1000; Jackson ImmunoResearch) anti-rabbit Alexa Fluor 488, anti-sheep Alexa Fluor 350, or anti-human Alexa Fluor 594 in blocking buffer. Coverslips containing the neurons were rinsed and mounted.

Quantitative reverse transcription polymerase chain reaction (RT-qPCR)

To determine TRP channel mRNA expression by RT-qPCR, RNA was extracted (TRIzol, Invitrogen) and cDNA was synthesized (Thermo Scientific, K1682). The following Taqman Gene Expression Assays (Applied Biosystems) containing predesigned primers were used: Mm01730183_gH Eif4a2, Mm01246301_m1 Trpv1, Mm01227437_m1 Trpa1, and Mm99999915_g1 Gapdh. Real-time PCR was performed using a QuantStudio 3 System (Applied Biosystems). Transcript quantification was performed using the comparative calculated cycle threshold method.

Intracellular Ca²⁺ imaging

Cultures of TG neurons from Pirt-GCaMP3 mice (N = 16; n = 8 per agonist), in which expression of the genetically encoded Ca²⁺ indicator GCamMP3 is driven by the Pirt promotor in primary sensory neurons in the dorsal root ganglion (DRG) and TG (30), were treated with onabotA, IS, or IS+onabotA as above. The coverslips were removed from the media and washed once with Calcium Imaging Buffer (CIB) containing the following (in mM): 130 NaCl, 20 D-(+)-Glucose, 2.1 MgCl₂-6H₂O, 2.4 KCl, 10 HEPES, and 2 CaCl₂ (pH 7.4 adjusted with NaOH, adjusted to 300–310 mOsm with D-mannitol). Coverslips were placed in CIB and held at 37°C prior to imaging. Agonists (capsaicin [100 nM; TRPV1]; AITC [100 uM; TRPA1]) were dissolved in CIB. Neurons were depolarized with 100 mM KCl in CIB as positive control for Ca²⁺ influx. Cells that did not respond to KCl or displayed abnormal responses to agonist (traces consistently below 0 [ΔF/F₀] or late response [>60 seconds] to agonist) were removed from analysis.

Imaging and analysis

Details on image analysis can be found in supplemental methods. Briefly, immunocytochemistry slides were imaged and processed using a Keyence BZ-X800 fluorescent microscope and software. All images in an experiment were acquired under identical conditions. ImageJ (35), software developed by the NIH, was used to identify and count how many cells express a given protein.

For calcium imaging, fluorescent measurements were made on an inverted epi-fluorescence/phase contrast microscope and conducted using 470 nm light for excitation (Xenon UV/Vis arc lamp), recording emissions with specific filter sets, and 1.00 second exposure. Images were taken at a rate of 54 images/minute with an observation wavelength of 560 nm. Data are presented as mean area under the curve (AUC) or maximum fluorescence intensity (ΔF/F0).

Statistical analyses

For all quantitative analyses, the statistical significance of treatment effects was examined using one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test; p-values <0.05 were considered statistically significant. All statistical analyses were completed in GraphPad Prism 9 (GraphPad Software, San Diego, CA).

Results

OnabotA induces SNAP25 cleavage in cultured primary trigeminal ganglion neurons

Immunocytochemistry and protein biochemistry showed that in primary cultured TG sensory neurons from wild-type mice, onabotA treatment resulted in cleavage of SNAP25 (Figure 1), as reported previously in rats (18,36). SNAP25₁₉₇ was not detected in control neurons or neurons in the sensitized condition (IS), which were not treated with toxin. Treatment with onabotA resulted in small, but significant, expression of SNAP25₁₉₇-positive cells (4%; p < 0.05 vs control) which was further significantly increased in the IS+onabotA treatment group (10%; p < 0.01 vs onabotA) (Figure 1A–C). Additionally, >97% of the SNAP25₁₉₇-positive cells co-labeled with PGP9.5, indicating their neuronal phenotype. Western blot analysis also showed the presence of cleaved SNAP25₁₉₇ in TG neurons treated with onabotA or IS+onabotA (Figure 1D, E). It is important to emphasize that these assays measure the product of the onabotA light chain’s (Lc/A) enzymatic activity (i.e., cleaved SNAP25), which has been shown to be shuttled to areas beyond the synaptic terminal (e.g., somal membrane, other areas of potential synaptic contact) even after cleavage by onabotA (37–40).

Figure 1.

OnabotA Leads to SNAP25 Cleavage in Mouse Cultured Primary Trigeminal Ganglion Neurons. TG neurons were pretreated with or without IS for 24 hours, then treated overnight with onabotA (2.75 pM). OnabotA was removed and neurons were grown for a further 48 hours. (a) Representative images of TG neurons showing immunocytochemical staining of cleaved SNAP25₁₉₇ in red and the neuronal marker PGP9.5 in green 48 hours after initial onabotA treatment. Magnification 40X, scale bar 50 µm. (b) Digital magnification showing SNAP25₁₉₇-IR+ neurons located primarily in the periphery of the cells. (c) Quantification of the percentage of SNAP25₁₉₇-IR+ cells out of the total number of neurons 48 hours after initial toxin treatment; one-way ANOVA with Tukey’s test, F(3,55) = 15.83, p < 0.0001. n = 6 independent experiments, consisting of pooled TG neurons from 4–10 animals per experiment. (d) Immunoblot and (e) densitometric analysis of bands showing that onabotA induces SNAP25₁₉₇ cleavage 48 hours after initial toxin treatment; one-way ANOVA with Tukey’s test, F(3,8) = 4.093, p < 0.05. Data represent mean + SEM. *p < 0.05, **p < 0.01, ****p < 0.0001. Background was subtracted out before analysis.

OnabotA modulates surface expression of TRPV1 and TRPA1 in primary trigeminal ganglion neurons under sensitized conditions

Given the role of TRPV1 and TRPA1 in trigeminal pain and migraine pathophysiology (15–17,41), as well as prior studies showing BoNT/A-mediated decreases in expression and trafficking of these channels to the plasma membrane (18,26), we first wanted to investigate whether onabotA affected cell-surface expression of TRPV1 and TRPA1 in our in vitro model of sensitization (26,42,43). Using ecto-TRP channel staining (13,44), we found that the proportion of TRPV1-IR positive neurons was similar in control and onabotA-treated neurons at both 24 and 48 hours after treatment (Figure 2A, B). At the 24 hour timepoint, there was a significantly higher proportion of TRPV1-IR positive neurons in the IS condition (41.0%) than in control cells (19.8%; p < 0.0001) and this enhancement was attenuated in the IS+onabotA condition (20.5%; p < 0.0001 vs IS). Similar differences and effects were observed 48 hours after treatment (Figure 2A, B). Figure 2C depicts representative triple-fluorescent micrographs of neurons in each treatment group 48 hours post-toxin treatment. We observed the colocalization of SNAP25₁₉₇ in a subset of TRPV1-IR positive neurons, indicating that onabotA is affecting receptor cell-surface expression in neurons that are taking up the toxin.

Figure 2.

OnabotA Modulates Surface Expression of TRPV1 in Primary Trigeminal Ganglion Neurons Under Inflammatory Conditions. (a, b) Quantitation of the proportion of TRPV1-IR+ TG cells relative to the total number of cells per area. At 24 hours, one-way ANOVA with Tukey’s test, F(3,124) = 21.24, p < 0.0001. At 48 hours, one-way ANOVA with Tukey’s test, F(3,163) = 7.058, p < 0.001. n = 3 independent experiments at 24 hours, consisting of pooled TG neurons from 6–7 animals per experiment. n = 6 independent experiments at 48 hours; consisting of pooed TG neurons from 4–10 animals per experiment. Data represent mean + SEM. **p < 0.01, ***p < 0.001, ****p < 0.0001. (c) Representative images showing TG neurons labeled for surface-expressed TRPV1. TRPV1 (green), SNAP25₁₉₇ (red), and PGP9.5 (blue). Arrowheads denote SNAP25₁₉₇-IR+ cells; full arrows denote cells not expressing SNAP25₁₉₇. Magnification, 60X. Scale bar, 50 µm.

We recorded similar effects of IS and onabotA treatment on cell-surface expression of TRPA1. Twenty-four hours after toxin treatment, TRPA1 cell-surface expression was present in 27.4% and 31.9% of control and onabotA-treated neurons, respectively; although one-way ANOVA revealed a significant main effect of treatment condition, Tukey pairwise comparisons did not show significant differences between control and IS (p = 0.058) or IS vs IS+onabotA (p = 0.073) (Figure 3A). Forty-eight hours after toxin treatment, TRPA1 cell-surface expression was present in 24.9% and 22.3% of control and onabotA-treated neurons, respectively. Treatment with IS significantly enhanced the proportion of TRPA1-IR positive neurons to 38.2% (p < 0.0001 vs control) and this effect was attenuated in neurons treated with IS+onabotA (19.9%; p < 0.0001 vs IS) (Figure 3B). Figure 3C depicts representative neurons in each treatment group 48 hours post-toxin treatment. Once again, we saw colocalization of SNAP25₁₉₇ with a subset of TRPA1-IR positive neurons, indicating that onabotA affects receptor expression in the neurons taking up the toxin.

Figure 3.

OnabotA Modulates Surface Expression of TRPA1 in Primary Trigeminal Ganglion Neurons Under Inflammatory Conditions. (a, b) Quantitation of the proportion of TRPA1-IR+TG cells relative to the total number of cells per area. At 24 hours, one-way ANOVA with Tukey’s test, F(3,90) = 2.789, p < 0.05. At 48 hours, one-way ANOVA with Tukey’s test, F(3.193) = 17.84, p < 0.0001. n = 3 independent experiments at 24 hours, consisting of pooled TG neurons from 6–7 animals per experiment. n = 6 independent experiments at 48 hours; consisting of pooled TG neurons from 4–10 animals per experiment. Data represent mean + SEM. **p < 0.01, ***p < 0.001, ****p < 0.0001. (c) Representative images showing TG neurons labeled for surface-expressed TRPA1. TRPA1 (green), SNAP25₁₉₇ (red), and PGP9.5 (blue). Arrowheads denote SNAP25₁₉₇-IR+ cells; full arrows denote cells not expressing SNAP25₁₉₇. Magnification, 60X. Scale bar, 50 µm.

Recorded changes in surface protein expression of TRPV1 and TRPA1 were not accompanied by concomitant changes in transcription, as RT-qPCR showed that there were no statistically significant differences in TRPV1 or TRPA1 mRNA expression in neurons treated with IS, onabotA, or IS+onabotA compared to control at the time points tested (Supplemental Figure 1A, B). This is consistent with prior findings (18).

OnabotA modulates TRPV1- and TRPA1-Mediated Ca²⁺ influx in primary cultured trigeminal ganglion neurons under sensitized conditions

To investigate the broader functional effects of onabotA in our in vitro model of sensitization, we next examined the effect of onabotA treatment on TRPV1 and TRPA1 ion channel function as measured by Ca²⁺ influx in response to selective activation. TG neurons were cultured from Pirt-GCaMP3 mice, in which the Ca²⁺ sensor GCaMP3 is expressed selectively in sensory neurons within the DRG and TG (30), allowing the study of calcium influx without dye loading. Stimulation of these sensory neurons with the TRPV1 agonist capsaicin (100 nM) produced Ca²⁺ influx, as measured by mean area under the curve (AUC), although IS did not significantly sensitize stimulated Ca²⁺ influx in our studies compared with control cultures (p = 0.058 vs control). Mean maximum intensity of fluorescence was not significantly sensitized by IS treatment. However, neurons treated with IS+onabotA had significantly lower TRPV1-mediated Ca²⁺ influx (AUC) than neurons treated with IS alone (p < 0.01) (Figure 4A–C). Similar findings were observed with the TRPA1 agonist, AITC (100 µM), with IS+onabotA significantly decreasing Ca²⁺ influx compared to IS (p < 0.01), although IS alone did not produce significant sensitization of the Ca²⁺ response compared to control cultures (Figure 4D–E). Importantly, application of KCl confirmed responsiveness of TG sensory neurons, regardless of treatment.

Figure 4.

OnabotA Modulates TRPV1- and TRPA1-Mediated Ca²⁺ Influx in Primary Cultured Trigeminal Ganglion Neurons. (a) Ca²⁺ influx via TRPV1 channels stimulated by capsaicin (100 nM) in TG sensory neurons that were cultured from Pirt-GCaMP3 mice and treated with onabotA, IS, or IS + onabotA. Control = blue, onabotA = orange, IS = red, IS + onabotA = green. n = 3 independent experiments, 37–39 neurons per experiment from n = 8 animals. Quantification of the (B) mean area under the curve (AUC) and (c) mean maximum fluorescent intensities for capsaicin stimulation. One-way ANOVA with Tukey’s test for AUC: F(3,148) = 3.799, p < 0.05. One-way ANOVA with Tukey’s test for mean maximum fluorescence intensity: F(3,148) = 3.626, p < 0.05 Data are presented as mean ± SEM. *p < 0.05. (d) Ca²⁺ influx via TRPA1 channels stimulated by AITC (100 µM) in TG sensory neurons that were cultured from Pirt-GCaMP3 mice and treated with onabotA, IS, or IS + onabotA. Control = blue, onabotA = orange, IS = red, IS + onabotA = green. n = 3 independent experiments, 27–29 neurons per experiment from n = 8 animals. Quantification of the (e) mean AUC and (f) mean maximum fluorescent intensities for AITC stimulation. One-way ANOVA with Tukey’s test for AUC: F(3,119) = 4.564, p < 0.01. One-way ANOVA with Tukey’s test for mean maximum fluorescence intensity: F(3,112) = 4.014, p < 0.01. Data are presented as mean ± SEM. P-values <0.05 were considered significant. **p < 0.01.

OnabotA modulates TRPV1- and TRPA1-mediated neuropeptide release in cultured primary trigeminal ganglion neurons under sensitized conditions

We next assessed whether onabotA-mediated reductions in TRPV1 and TRPA1 surface expression and consequent Ca²⁺ influx would lead to diminished release of the neuropeptide CGRP and the prostanoid PGE2, inflammatory mediators known to play a role in migraine pathophysiology, including migraine pain (7,8,14,45,46). For this, cultured TG neurons from wild-type mice were exposed to IS and 24 hours later treated with onabotA to generate groups of control, onabotA-, IS-, and IS+onabotA-treated cells. Forty-eight hours later, TRPV1 or TRPA1 were stimulated using the agonists capsaicin (1 µM; 30 minute) or JT010 (10 µM; 30 minute), respectively, and secretion of CGRP and PGE2 were measured via ELISA. Capsaicin induced CGRP secretion that was greater in cells sensitized by IS treatment (201.6 pg/mL; p < 0.05 vs control) (Figure 5A). Cells treated with IS+onabotA had significantly less capsaicin-stimulated CGRP secretion than cells treated with IS alone (108.8 pg/mL; p < 0.05 vs IS) (Figure 5A). OnabotA had similar effects on TRPA1-mediated CGRP release. JT010 stimulated CGRP secretion that was increased in cells sensitized by IS (143.7 pg/mL; p < 0.01 vs control), although onabotA reduced, but did not significantly alter this effect (Figure 5B).

Figure 5.

OnabotA Modulates TRPV1- and TRPA1-Mediated Neuropeptide Release in Cultured Primary Trigeminal Ganglion Neurons Under Inflammatory Conditions. TG neurons were treated with onabotA, IS, or IS+onabotA. Forty-eight hours later, neurons were stimulated with the TRPV1 agonist capsaicin (1 µM; 30 minutes) or the TRPA1 agonist JT010 (10 µM; 30 minutes) and neuropeptide release was analyzed via ELISA. (a, b). Quantification of CGRP release stimulated by capsaicin (left; one-way ANOVA with Tukey’s test, F(3,31) = 3.808, p < 0.05) or JT010 (right; one-way ANOVA with Tukey’s test, F(3,12) = 8.846, p < 0.01) and (c, d) Quantification of PGE2 release stimulated by capsaicin (left; one-way ANOVA with Tukey’s test, F(2,9) = 9.431, p < 0.01) or JT010 (right; one-way ANOVA with Tukey’s test, F(2,20) = 7.075, p < 0.01). n = 4 independent experiments, consisting of pooled TG neurons from 9–15 animals per experiment. Data represent mean+SEM. *p < 0.05, **p < 0.01.

IS treatment also significantly enhanced TRPV1-stimulated PGE2 release (64.4 pg/mL; p < 0.05 vs control) and this enhancement was attenuated in cells treated with IS+onabotA (20.3 pg/mL; p < 0.05 vs IS) (Figure 5C). Similarly, stimulation of TRPA1 with JT010 induced PGE2 secretion in control TG neurons that was significantly enhanced by IS treatment (442.3 pg/mL; p < 0.01 vs control). Once again, onabotA attenuated this IS-induced enhancement (203.7 pg/mL; p < 0.01 vs IS) (Figure 5D).

Discussion

Multiple controlled and real-world studies have demonstrated the efficacy of onabotA in chronic migraine prevention (47–49); however, the classical MoA of onabotA, involving inhibition of acetylcholine from motor and autonomic neurons, cannot fully explain the effectiveness of onabotA in preventing migraine attacks since it is considered a sensory neurological disease. OnabotA treatment for migraine involves injections in the region of the trigeminal nerve, which has been shown to be important in the pathogenesis of migraine (5). Indeed, a prior study demonstrated that peripheral injections of onabotA produced functional changes, as evidenced by TRPV1-specific antinociception and decreased TRPV1 expression in dural neurons, for which the course is through the TG (18). Similarly, peripheral injections of onabotA in the region of the trigeminal nerve resulted in functional consequences on TRPV1 and TRPA1 when stimulated with their agonists, capsaicin and mustard oil, respectively (50). Here, we extend these findings and provide evidence to show that onabotA at a concentration that is comparable to clinical applications (50 U/mL; 2.75 pM) (51) has direct effects on TG sensory neurons expressing TRPV1 and TRPA1 receptors; reduces trafficking of TRP channels to the plasma membrane in a sensitized state seen in trigeminal pain such as migraine (52), without concomitant changes in transcription; and significantly inhibits their signaling and subsequent secretion of inflammatory mediators involved in migraine pain and pathophysiology.

To examine the direct effects of onabotA treatment on sensory neurons, we treated cultured mouse primary TG neurons for 24 hours with onabotA (2.75 pM; 50 U/mL) and used an antibody highly selective for the BoNT/A-cleaved epitope of SNAP25 (33) for immunocytochemistry and western blot analyses. Immunocytochemical analysis showed that onabotA induced significant SNAP25 cleavage in TG neurons 48 hours after toxin exposure, which is consistent with prior studies demonstrating BoNT/A-mediated cleavage of SNAP25 and inhibition of SNARE-mediated processes in sensory neurons (18,36,45,53–56). Furthermore, the percentage of neurons expressing SNAP25₁₉₇ significantly increased in cultures that were treated with IS+onabotA, which suggests a net increase in SNAP25 cleavage, potentially due to increased uptake of onabotA in sensitized neurons. It is important to emphasize that these assays are measuring the product of the onabotA light chain’s (L_C/A) enzymatic activity (i.e., cleaved SNAP25) which has been shown to translocate more broadly within the cell (37,38). Our data do not show the location of the Lc/A itself within individual neurons. We would not expect the toxin to migrate from the synapse due to its association with septins at the nerve terminal membrane, but evidence suggests that SNAP25 can be shuttled to areas beyond the synaptic terminal (e.g., somal membrane, other areas of potential synaptic contact) even after cleavage by onabotA (39,40). Further, increasing evidence refutes the occurrence of trans-synaptic transfer of onabotA and shows that the clinical effects of onabotA are the result of decreased peripheral input, rather than central effects (39,57,58).

In the family of TRP ion channels, several members are involved in initiating and sustaining pain and inflammation at the level of sensory transduction in primary sensory neurons within the DRG and TG; TRPV1 and TRPA1 are two widely-studied examples (41,59). In DRG neurons, inflammatory cytokines and prostanoids (i.e., TNFα, PGE2) increase cell-surface expression of TRPV1 and TRPA1 (29,60), and stimulation of TG neurons with the TRPV1 agonist capsaicin induces secretion of inflammatory mediators such as CGRP and PGE2 (60,61). Insertion of TRP channels into the plasma membrane has been shown to occur via SNARE-mediated processes (18,59,62), and multiple studies have demonstrated that BoNT/A inhibits trafficking and insertion of these channels into the plasma membrane (13,18,62,63). Indeed, an earlier investigation showed that BoNT/A decreased TRPV1 mobilization and insertion into the plasma membrane of TG neurons, led to proteasomal degradation of intracellular TRPV1, and produced TRPV1-specific antinociception (18). Other groups have also demonstrated that the inhibition of mechanical nociception by onabotA is due to the toxin interfering with trafficking of TRPV1 and TRPA1 within C-type mechanosensitive meningeal nociceptors (50,64).

To expand on these findings and explore the direct effects of onabotA in TG neurons, we examined the effect of onabotA treatment on cell-surface expression of TRP channels in cultured TG neurons, which were treated with inflammatory soup (IS), a mixture of substances (bradykinin, substance P, serotonin, and histamine) that has been shown in various models to be an agent that creates a sensitized state in trigeminal neurons that contributes to trigeminal pain, including migraine (25,26,28,41,64,65). Although the specific composition and concentration of each component of IS differs across studies, multiple in vitro and in vivo studies have used IS to induce sensitized conditions that model the hallmark peripheral and central sensitization of trigeminovascular neurons in migraine (6,22–25,50). We found that surface expression of TRPV1 is increased in sensitized compared to control TG neurons and that this effect is attenuated by onabotA. TRPV1 surface expression is reduced in SNAP25₁₉₇-IR positive neurons treated with IS+onabotA, which is consistent with onabotA-mediated SNAP25 cleavage reducing surface-expressed TRPV1 (59,62,64). OnabotA had similar inhibitory effects on upregulation of TRPA1 surface expression in cultures treated with IS, which indicates that onabotA inhibits the SNARE-dependent trafficking and insertion of TRPV1 and TRPA1 into the plasma membrane of TG neurons in sensitized conditions. We also show that decreased TRP channel trafficking in treated TG neurons occurred without concomitant decreases in transcription, as measured by RT-qPCR 24 and 48 hours after onabotA exposure, which for non-sensitized neurons is in keeping with previous reports (18,36), and is novel for sensitized conditions.

Our findings are in line with prior studies demonstrating that onabotA inhibits activity of neurons stimulated via TRPV1 and TRPA1 agonists, which are likely to be C-fibers (50,64). Among nociceptors, TRPV1 is expressed primarily in slower responding C-type meningeal nociceptors, while TRPA1 is expressed in both fast-responding A-delta meningeal nociceptors and C-fibers (65–67). Of note, we observed a slightly different time course of onabotA’s effects when examining TRPA1 expression than with TRPV1. At 24 hours, TRPV1 cell-surface expression had increased in response to IS treatment, and this effect was attenuated by onabotA. In contrast, treatment with IS did not significantly increase cell-surface expression of TRPA1 at the 24 hours timepoint after toxin exposure, nor did onabotA significantly alter any of the effects of IS; however, at the 48 hours timepoint, onabotA significantly attenuated an IS-mediated increase in TRPA1 expression. The differential in time course across TRPV1 and TRPA1 may be related to the imbalance of receptors across C-fibers, which express both channels and are responsive to onabotA, versus A-delta which express only TRPA1 and are not responsive to onabotA (50).

We next sought to examine potential functional consequences of onabotA-mediated SNAP25 cleavage and resultant reductions in TRP cell-surface expression within TG neurons. To accomplish this, we measured calcium influx in response to selective activation of TRPV1 and TRPA1 in neurons cultured from Pirt-GCaMP3 mice, in which expression of a genetically coded Ca²⁺ indicator is under the control of the Pirt promoter found selectively in peripheral sensory neurons (30). We show that onabotA attenuated both TRPV1- and TRPA1-mediated increases in Ca²⁺ influx in sensitized, but not control, sensory neurons. This may indicate that the decreased number of cells expressing TRPV1 or TRPA1 after onabotA exposure in sensitized conditions ultimately decreases function of both migraine-relevant TRP channels, and precludes the threshold of stimulation from being reached. We did not observe a statistically significant increase compared to control in TRP-mediated Ca²⁺ influx in neurons treated with IS alone, although onabotA produced similar inhibitory effects on TRP-mediated Ca²⁺ influx in neurons treated with IS, which complements our immunocytochemistry data. Although the expression of TRPV1/TRPA1 and responsiveness to TRP agonists, coupled with evidence demonstrating selective effects of onabotA on C-fiber nociceptors (50,64–68), strongly suggest that the studied TG neurons are nociceptors, a potential limitation of our study is that we did not use additional methods (e.g., electrophysiology) to further characterize these cells. Regardless, our findings give rise to the notion that onabotA-mediated decreases in cell-surface expression of TRP channels in nociceptors under sensitized conditions decreases downstream signaling of these channels. This is supportive of clinical observations that onabotA is effective in prevention of chronic migraine, which is likely due to decreased peripheral input reducing subsequent central sensitization (5).

CGRP is considered a key component of migraine pathophysiology (14). Additionally, NSAIDs that reduce prostanoids such as PGE2 are a common treatment for migraine (69,70). Prior studies have shown that stimulation of TG neurons with TRPV1 agonists induces secretion of CGRP and PGE2 (60,61) and that BoNT/A can inhibit stimulated neuropeptide release from DRG neurons (8,12). We found that onabotA exposure alone in unsensitized neurons did not influence the release of the inflammatory mediators CGRP and PGE2, but onabotA did significantly attenuate TRPV1- and TRPA1-mediated release of these neuropeptides in sensitized conditions. This is a novel observation, and demonstrates a direct connection between reduced surface-expression ofTRPV1 and TRPA1 by onabotA-mediated SNAP25 cleavage in TG neurons under sensitized conditions, decreased Ca²⁺ influx, and a resultant reduction in fusion of the large dense core vesicles necessary for CGRP and PGE2 release.

Conclusion

Overall, our findings deepen the understanding of the role of onabotA in nociceptor physiology as it pertains to trigeminal-related pain, including migraine. We show that onabotA has direct effects on mouse primary trigeminal sensory neurons. In these cells, onabotA cleaves SNAP25, prevents SNARE-mediated insertion of TRP channels into the plasma membrane and TRP-mediated increases in calcium influx in sensitized conditions without concomitant changes in transcription, and inhibits secretion of inflammatory mediators involved in migraine pathophysiology. Our data further demonstrate that the acetylcholine-focused classical MoA of onabotA is too narrow. After being endocytosed, the cytosolic light chain of onabotA cleaves SNAP25 which prevents SNARE-mediated synaptic vesicle exocytosis and receptor trafficking more generally in motor, autonomic, and sensory neurons.

Article highlights

OnabotA has direct effects on cultured trigeminal ganglion sensory neurons.

In neurons treated with inflammatory soup to mimic migraine-like sensitization, onabotA decreases cell-surface expression of TRP channels (TRPV1, TRPA1) involved in migraine pathophysiology.

OnabotA also attenuates TRP-mediated calcium influx and release of CGRP and PGE2 under sensitized conditions.

Supplemental Material

sj-jpg-1-cep-10.1177_03331024221141683 - Supplemental material for OnabotulinumtoxinA effects on trigeminal nociceptors

Supplemental material, sj-jpg-1-cep-10.1177_03331024221141683 for OnabotulinumtoxinA effects on trigeminal nociceptors by Ashley A Moore, Mariana Nelson, Christopher Wickware, Shinbe Choi, Gene Moon, Emma Xiong, Lily Orta, Amy Brideau-Andersen, Mitchell F Brin, Ron S Broide, Wolfgang Liedtke and Carlene Moore in Cephalalgia

Supplemental Material

sj-pdf-2-cep-10.1177_03331024221141683 - Supplemental material for OnabotulinumtoxinA effects on trigeminal nociceptors

Supplemental material, sj-jpg-1-cep-10.1177_03331024221141683 for OnabotulinumtoxinA effects on trigeminal nociceptors by Ashley A Moore, Mariana Nelson, Christopher Wickware, Shinbe Choi, Gene Moon, Emma Xiong, Lily Orta, Amy Brideau-Andersen, Mitchell F Brin, Ron S Broide, Wolfgang Liedtke and Carlene Moore in Cephalalgia