Neurobiological mechanisms of botulinum neurotoxin-induced analgesia for neuropathic pain

Abstract

Botulinum neurotoxins (BoNTs) are a family of neurotoxins produced by Clostridia and other bacteria that induce botulism. BoNTs are internalized into nerve terminals at the site of injection and cleave soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins to inhibit the vesicular release of neurotransmitters. BoNTs have been approved for multiple therapeutic applications, including the treatment of migraines. They have also shown efficacies for treating neuropathic pain, such as diabetic neuropathy, and postherpetic and trigeminal neuralgia. However, the mechanisms underlying BoNT-induced analgesia are not well understood. Peripherally administered BoNT is taken up by the nerve terminals and reduces the release of glutamate, calcitonin gene-related peptide, and substance P, which decreases neurogenic inflammation in the periphery. BoNT is retrogradely transported to sensory ganglia and central terminals in a microtubule-dependent manner. BoNTs decrease the expression of pronociceptive genes (ion channels or cytokines) from sensory ganglia and the release of neurotransmitters and neuropeptides from primary afferent central terminals, which likely leads to decreased central sensitization in the dorsal horn of the spinal cord or trigeminal nucleus. BoNT-induced analgesia is abolished after capsaicin-induced denervation of transient receptor potential vanilloid 1 (TRPV1)-expressing afferents or the knockout of substance P or the neurokinin-1 receptor. Although peripheral administration of BoNT leads to changes in the central nervous system (e.g., decreased phosphorylation of glutamate receptors in second-order neurons, reduced activation of microglia, contralateral localization, and cortical reorganization), whether such changes are secondary to changes in primary afferents or directly mediated by trans-synaptic, transcytotic, or the hematogenous transport of BoNT is controversial. To enhance their therapeutic potential, BoNTs engineered for specific targeting of nociceptive pathways have been developed to treat chronic pain. Further mechanistic studies on BoNT-induced analgesia can enhance the application of native or engineered BoNTs for neuropathic pain treatment with improved safety and efficacy.

1. Introduction

Botulinum neurotoxins (BoNTs) are a family of neurotoxins produced by Clostridia and other bacteria to induce botulism, and their effects are characterized by the flaccid paralysis of skeletal muscles and dysautonomia. Once early researchers discovered the peripheral mechanism of action of BonTs (Burgen et al., 1949), numerous studies have been conducted to better understand the complete mechanisms of action of this toxin, as its use in clinical settings is increasing. BoNTs comprise two chains and three functional domains: a light chain (LC) (a zinc-dependent metalloproteinase), a heavy chain (the N-terminal membrane translocation domain and the C-terminal receptor binding domain), and an intermediate region which links both chains. The receptor binding domain of the heavy chain binds to the presynaptic gangliosides of the cell surface, and the toxin is taken up after binding to another surface receptor: synaptotagmin or glycosylated Sv2. After internalization, it resides inside the vesicles. Acidification of the vesicles cleaves the disulfide bond connecting the heavy and light chains, and the toxin changes into its active form (Choudhury et al., 2021). The LC then translocates to the cytoplasm from inside the vesicles, which is facilitated by the N terminal of the heavy chain. After translocation, the LC is freed from the remainder of the toxin and cleaves and deactivates soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins, which regulate the presynaptic docking of vesicles containing neurotransmitters. Different serotypes of BoNTs cleave different SNARE proteins: BoNT-A, BoNT-C, and BoNT-E target the SNAP-25 protein, whereas BoNT B, D, F, and G act on vesicular protein isoforms of synaptobrevin (also called vesicle-associated membrane protein (VAMP)). Thus, BoNT inhibits the release of neurotransmitters such as acetylcholine (Choudhury et al., 2021; Sam & Bordoni, 2023).

Based on the canonical mechanism of action (Burgen et al., 1949), in 1980, an ophthalmologist used BoNT-A for the treatment of blepharospasm (Scott, 1980), which was approved by the FDA in 1989, soon after it was approved for the treatment of strabismus. Since then, BoNT-A has been approved for multiple diseases and different applications, including cosmetic medicine, cervical dystonia, urinary incontinence, lower limb spasticity, and sialorrhea. BoNT is also approved for treating migraines in the UK and the US (S. Chen, 2012), and possibly applicable for treating other chronic pain conditions. Interestingly, many BoNT-A applications have arisen from observations where the treatment of one condition improved another. For example, the treatment of blepharospasm reduced glabellar frown lines, and, in turn, patients treated for hyperfunctional lines of the face described relief in migraine and chronic headache episodes (Binder et al., 1998; Whitcup, 2021). The finding that BoNT-A was effective in treating different pathologies where the primary etiology was not muscle contraction, suggested that BoNT-A acted through non-cholinergic systems. Currently, it is known that BoNT-A blocks neurotransmissions from synapses other than cholinergic ones by blocking the Ca²⁺ evoked exocytosis of neurotransmitters, including those involved in nociception (Pellett et al., 2015; Popoff & Poulain, 2010).

According to the International Association for the Study of Pain, neuropathic pain is caused by a lesion or disease of the somatosensory nervous system. Neuropathic pain is one of the most debilitating chronic pain conditions. It is often resistant to conventional therapies, and the currently used drugs for its treatment can present important side effects (Ilari et al., 2022). Therefore, developing novel, effective treatments is necessary. Currently, BoNT-A has been tested in patients with peripheral neuropathic pain and has shown some efficacy (Table 1). BoNT-A has been tested using different application methods (e.g., subcutaneous or intradermal) and doses. It reduces the intensity of perceived pain, measured via a visual analog scale score or a numeric pain rating scale, in patients with diabetic neuropathy (Ghasemi et al., 2014; Salehi et al., 2019; Taheri et al., 2020), postherpetic neuralgia (Apalla et al., 2013; L. Chen et al., 2022; Xiao et al., 2010), carpal tunnel syndrome (Tsai et al., 2006), and trigeminal neuralgia (Shehata et al., 2013; C.-J. Wu et al., 2012; Zhang et al., 2014, 2017; Zúñiga et al., 2013). BoNT-A also reduces pain intensity when administered peripherally in neuropathic pain due to central nervous system (CNS) injuries, such as spinal cord injuries (De Icco et al., 2019; Han et al., 2016). Differences in the characteristics of neuropathic pain that are improved after BoNT-A treatment can be observed across studies, with some finding only some parameters improved (Ranoux et al., 2008; Salehi et al., 2019; Taheri et al., 2020), while others have found all parameters are improved (Ghasemi et al., 2014). A comparison of the effectiveness of radiofrequency and BoNT-A has found that both treatments show similar efficacy in all studied parameters, although BoNT-A is cheaper than radiofrequency treatment (L. Chen et al., 2022). However, BoNT-A is superior to lidocaine for postherpetic neuralgia and reduces the number of patients needing to take opioids (Xiao et al., 2010). Furthermore, BoNT-A treatment not only lessens pain intensity but improves sleep quality (Apalla et al., 2013; Attal et al., 2016; L. Chen et al., 2022; Salehi et al., 2019) and anxiety scores (Apalla et al., 2013; Attal et al., 2016; Ranoux et al., 2008), without differences in patient depression scores (Apalla et al., 2013; Attal et al., 2016; Ranoux et al., 2008).

Table 1:

Clinical trials investigating the efficacy of BoNT-A on neuropathic pain, quality of life, mental health, and sleep

References	Dose/Administration	Groups (N)	Study duration	Outcomes	Results
POSTHERPETIC NEURALGIA
(L. Chen et al., 2022)	Maximum dose 200 U Subcutaneous	BoNT-A (n=50) Radiofrequency (n=50)	Follow-up times: 4 weeks 12 weeks 24 weeks	Outcomes: Intensity of pain (NRS) Quality of sleep (measured via 5 questionnaires) Anxiety and depression (GAD-7 and PHQ-9 questionnaires) Characteristics of pain (ID Pain questionnaire) Costs	NRS: Both groups improved similarly and obtained similar results at 24 weeks post-treatment. Quality of sleep: Both groups improved similarly and obtained similar scores by the end of the treatment. ID Pain questionnaire: At 24 weeks, the incidence of numbness in the two groups was increased compared with baseline values, while the incidence of stabbing pain, burning pain, electric shock pain, allodynia, and spontaneous pain decreased similarly between the groups. Stabbing pain was lower in the radiofrequency group, while burning pain was lower in the BoNT-A group. GAD-7 and PHQ-9 questionnaires: GAD-7 scores were more reduced in the BoNT-A group. The PHQ-9 questionnaire scores improved in both groups similarly. Costs: Costs were higher for radiofrequency treatment.
(Apalla et al., 2013)	100 U Subcutaneous	BoNT-A (n=15) Placebo, IS (n=15)	4 weeks placebo vs. BoNT-A 20 weeks follow up (crossover design)	Pain intensity (VAS score) Quality of sleep (5 item questionnaire)	VAS score: Considerable reduction in pain intensity until week 10, when it increased until the end of the experiment (week 24). Sleep Score: Sleep score improved by week 2 and remained unchanged through week 12.
(Xiao et al., 2010)	Maximum dose 200 U	BoNT-A (n=20) Lidocaine (n=20) Placebo, IS (n=20)	3 months	Pain intensity (VAS score) Sleep time (hours) Percent opioid use	VAS: Decreased in all groups, although improvement in the BoNT-A group was greater than after IS or lidocaine injection. Sleep time: Sleep hours increased in all groups, although, the improvement in the BoNT-A group was greater than in the other two groups. Percent opioid use: More patients in the IS and lidocaine groups took opioids than in the BoNT-A group.
NEUROPATHIC PAIN DIABETES
(Taheri et al., 2020)	150 U BoNT-A Subcutaneous	BoNT-A on one foot (150 U) BoNT-A on both feet (75 U/foot) Placebo, IS in both feet All groups received duloxetine 30 mg/day	3 months	Pain intensity (VAS score) Intensity of neuropathic pain characteristics (NPS)	VAS: All groups demonstrated improved VAS scores, although improvement was superior in the BoNT-A group. NPS: All scored parameters improved in all of the groups except for dull and cold sensation. Improvement was greater in BoNT-A treated groups. BoNT-A injection in both feet did not improve outcomes when compared to one injection, except for heat sensation.
(Salehi et al., 2019)	100 U BoNT-A Intradermal	BoNT-A on one foot (n=16) Placebo, IS on one foot (n=16)	12 weeks	Health related quality of life (SF-36) Pain intensity (VAS score) Intensity of pain characteristics (NPS) Sleep quality (Pittsburgh Sleep Quality Index)	VAS, SF-36, and the Pittsburgh Sleep Quality Index: BoNT-A improved the scores for pain intensity, sleep quality and those in the physical dimension of the overall health questionnaire. . NPS: Improvements occurred in all analyzed parameters, except for sharp, sensitive, and deep sensation scores.
(Ghasemi et al., 2014)	100 U BoNT-A Intradermal	BoNT-A on one foot (n=20) Placebo, IS on one foot (n=20)	3 weeks	Pain intensity (VAS score) Intensity of pain characteristics (NPS)	VAS: BoNT-A significantly reduced the intensity of pain by the end of the study and when compared to the placebo group. NPS: The BoNT-A group improved on all parameters of the questionnaire. No improvements were observed in the placebo group.
CARPAL TUNNEL SYNDROME
(Tsai et al., 2006)	60 U (30 U on each side)	BoNT-A group (n=5)	3 months	Pain intensity (VAS score)	VAS: Improvement in three patients (mean improvement 1.3).
TRIGEMINAL NEURALGIA
(Zhang et al., 2017)	Single dose (50–70 U) (n=44) Two doses (100–140 U) (n=37)	Single dose BoNT-A Two doses BoNT-A (Time between administrations: 2 weeks)	6 months	Daily diary: Provoking factors, Frequency of attacks, pain intensity (VAS scale). Efficacy with proportion of responders (defined as patients with equal or more than a 50% reduction in mean pain scores)	No significant difference between groups were observed in the rate of TN occurrence. The times to drug effect and peak efficacy of the two groups were statistically similar. However, the duration of efficacy in the single-dose group was significantly longer than that of the repeated dose group.
(Zhang et al., 2014)	Maximum 75 U	Placebo (IS) (n=28) BoNT-A 25 U (n=27) BoNT-A 75 U (n=29)		Pain intensity (VAS) Response rates Patient's belief about the efficacy of treatment.(PGIC scale)	VAS: BoNT-A groups had reduced VAS scores compared to the placebo group. No differences were observed between the BoNT-A groups. Response rates: Response rates were significantly higher in the BoNT-A groups when compared to placebo. There were no significant differences between the BoNT-A treated groups. PGIC scale: PGIC scores improved in the BoNT-A groups when compared to placebo. There were no differences between the BoNT-A treated groups.
(Zúñiga et al., 2013)	50 U BoNT-A Subcutaneous	BoNT-A (n=20) Placebo (n=16)	3 months	Pain Intensity and Impact on function by the presence or absence of mild, moderate, severe, or disabling pain when performing activities of daily living (VAS) Health related quality of life (SF-36)	VAS: After 3 months, the BoNT-A treated group had reduced its VAS score more than the placebo group SF-36: No differences were observed between the groups.
(C.-J. Wu et al., 2012)	75 U BoNT-A Intradermal/submucosal	BoNT-A (n=22) Placebo (IS) (n=20)	12 weeks	Pain intensity (VAS) Frequency of TN attacks Patient's belief about the efficacy of treatment. (PGIC scale) Response to treatment (defined as a 50% decrease in pain scores) from baseline to endpoint	VAS: The BoNT-A group had reduced mean VAS scores, which was sustained throughout the study. Frequency of TN attacks: At the study endpoint, BoNT-A was significantly superior to placebo in reducing attack frequency. PGIC: 77.27% of patients in the BoNT-A group reported their symptoms as much or very much improved vs. 20% in the placebo group. Response to treatment was much higher in the BoNT-A group compared to the placebo group.
(Shehata et al., 2013)	100 U BoNT-A	BoNT-A (10) Placebo (n=10)	12 weeks	Pain intensity (VAS) Number of painful paroxysms QoL	VAS: VAS scores at endpoint LOCF relative to baseline for the BoNT-A group showed a decrease of 6.5 compared with a decrease of 0.3 for the placebo group. QoL: There was an increase in the QoL functioning scale and a reduction in the number of weekly acute medications and frequency of paroxysms per day in the BoNT-A group compared to placebo.
COMBINED ETHIOLOGIES
(De Icco et al., 2019)	Maximum 200 U Intramuscular BoNT-A_Stroke (n=10) BoNT-A_Multiple sclerosis (n= 10) BoNT-A_Spinal cord injury (n=5)	All patients receive BONT-A. Comparisons are performed vs baseline	30 days	Pain intensity (NRS) Intensity of neuropathic pain characteristics (NPS)	NRS: NRS was slightly reduced. NPS: The score for NPS was reduced but did not reach statistical significance…
(Attal et al., 2016)	Max 300 U BoNT-A Post-traumatic/postsurgical (n=46) Polyneuropathy (n=14) Postherpetic neuralgia (n=6)	Patients are randomized to: BoNT-A administration (n=34) Placebo, IS (n=34) At 12 weeks (2nd injection IS (n=26)/BoNT-A (n=32))	24 weeks	Pain intensity (NRS) Intensity of neuropathic pain characteristics (NPS) Hospital anxiety and depression scale Quality of Sleep index	NRS: Decrease in pain intensity in the BoNT-A group with enhanced efficacy after the 2nd injection. The efficacy was greater in patients who presented greater mechanical allodynia and less thermal deficits at the beginning of treatment. Greater intra-epidermal nerve fiber density at baseline appeared to be a predictor for greater response to BoNT-A. Neuropeptide concentration at the skin-punch biopsy was not a predictor for BoNT-A efficacy. NPS: BoNT-A improved paroxysmal pain and allodynia. Hospital anxiety and depression scale and quality of sleep: Both scores were improved after BoNT-A treatment.
(Ranoux et al., 2008)	Maximum 200 U Intracutaneous Postherpetic neuralgia (n=4)/postoperative neuropathies (n=25)	IS (n=14) BoNT-A group (n=15)	24 weeks	Self-reported pain intensity (0-10 numerical scale) Intensity of neuropathic pain characteristics (NPS) Brief pain inventory interference Anxiety and depression (Hospital anxiety and Depression scale)	Pain intensity: Pain intensity was reduced in the BoNT-A group and remained stable for 14 weeks. Intensity and area of allodynia to brush and thresholds to cold stimuli were reduced but not to thermal and mechanical allodynia. NPSI: Improvement in burning, paroxysmal pain, allodynia and number of paroxysms in the BoNT-A group. Brief pain inventory interference: Improvement in general activity and mood by week 12. Hospital anxiety and Depression scale: Improvement in anxiety scores in the BoNT-A group while the placebo group’s scores worsened. No differences in depression scores between groups.

BoNT-A: botulinum neurotoxin; IS: isotonic saline; GAD-7: Generalised Anxiety Disorder Assessment; QoL: quality of life; NPS: neuropathic pain scale; NPSI: Neuropathic Pain Symptom Inventory; NRS: Numeric rating scale; PGIC: Patients' Global Impression of Change; PHQ-9: Patient Health Questionnaire; SF-36: 36-Item Short Form Health Survey; TN: trigeminal neuralgia; VAS: visual analogue scale

Despite therapeutic efficacy, the mechanisms underlying the resulting analgesia for neuropathic pain following BoNT injections are not fully understood. In preclinical animal studies, the effectiveness of BoNT-A on neuropathic pain has primarily been investigated using the chronic constriction injury (CCI) model, which has been highly effective in reducing neuropathic pain in spinal and craniofacial areas (Table 2). The effectiveness of BoNT has also been assessed in diabetic neuropathy (Bach-Rojecky et al., 2010; Favre-Guilmard et al., 2017a) and peripheral neuropathy induced by paclitaxel (Favre-Guilmard et al., 2009; Waskitho et al., 2021). The pathogenesis of different neuropathic pain conditions is heterogeneous; therefore, BoNT-induced analgesia should be mediated through diverse peripheral and central mechanisms. A better understanding of the neurobiological mechanisms of BoNT-induced analgesia for neuropathic pain should enhance its application in chronic pain treatment and contribute to developing more effective treatment strategies for neuropathic pain. In the present narrative review, the effects of BoNT on peripheral and central nociceptive transmission in peripheral neuropathy models will be discussed. In Figure 1, we summarize the proposed mechanisms of analgesia for neuropathic pain produced by peripherally administered BoNT. We also summarize novel approaches used to target neuropathic pain through new forms of engineered BoNTs.

Table 2.

Studies determining the effects of botulinum neurotoxin (BoNT) on hyperalgesia from peripheral neuropathy in rodents

References	Animal model	Animals type	Site of injury	BoNT usage	Behavior Tests	Findings/Highlights
(Luvisetto et al., 2007)	CCI	M	Sciatic nerve	i.pl. pre- and post-CCI	Mechanical sensitivity (dynamic plantar aesthesiometer)	BoNT-A only reduced pain symptoms once neuropathic pain had already developed, but was unable to prevent the onset of neuropathic pain.
(Marinelli et al., 2010)	CCI	M, R	Sciatic nerve	i.pl. for mice; i.pl./i.t. for rats	Mechanical sensitivity (VF), cold plate test, walking track analysis, weight bearing analysis	Anti-allodynic and anti-hyperalgesic effects of BoNT-A along with enhanced regenerative processes of the injured nerve.
(Mika et al., 2011)	CCI	R	Sciatic nerve	i.pl.	Mechanical sensitivity (VF and dynamic plantar aesthesiometer), cold plate test	BoNT/A altered gene expression in the DRG, and inhibited microglia and macrophages in the spinal cord.
(Vacca et al., 2013)	CCI	M	Sciatic nerve	i.pl.	Mechanical sensitivity (VF), heat sensitivity	BoNT-A increased the analgesic effects of morphine and countered morphine-induced tolerance during chronic morphine treatment.
(Zychows ka et al., 2016)	CCI	R	Sciatic nerve	i.pl.	Motor function (Rotarod test), exploratory (open field test), paw pressure test (Randall-Selitto Test), mechanical sensitivity (VF), Hargreave's test, cold plate test	BoNT-A did not influence motor function, but significantly attenuated pain-related behaviors and microglial activation.
(J. Wang et al., 2019)	CCI	R	Sciatic nerve	i.t.	Mechanical sensitivity (VF), motor function (Rotarod test)	SNAP-25 could be a potential downstream target of PKA and p-CREB. Its involvement contributed to neuropathic pain induced by CCI by modulating the expression of VGluT2 and activating astrocytes.
(Shi et al., 2023)	CCI	R	Sciatic nerve	i.t.	Mechanical sensitivity (VF), motor function (Rotarod test)	BoNT-A produced significant analgesic activity by regulating the expression of VNUT in the spinal cord of rats. BoNT-A did not affect the motor function of rats.
(Favre-Guilmard et al., 2017b)	CCI Diabetic neuropathy Carrageenan injection	R	Unilateral sciatic nerve; Streptozotocin i.p. injection	i.pl.	Pressure pain (modified Randall-Selitto method)	Bilateral analgesic effects of BoNT-A following unilateral administration across pain modalities.
(Drinovac et al., 2013)	Sciatic nerve partial transection	R	Sciatic nerve; paw	i.pl.	Paw pressure test, VF, cold allodynia	BoNT-A-induced pain reduction was mediated by the μ-opioid receptor, which is involved in central antinociceptive activity.
(Bach-Rojecky et al., 2005)	Sciatic nerve partial transection	R	Sciatic nerve	i.pl.	Modified unilateral hot plate test, modified paw-pressure test	Significant analgesic activity of BoNT-A lasted for more than 10 days, but was evident 5 days after peripheral application of the toxin.
(Drinovac et al., 2013)	Sciatic nerve partial transection Formalin injection	R	Sciatic nerve	i.pl.	Nocifensive behaviors (licking, flinching and shaking of the injected paw) Mechanical sensitivity (modified Randall-Selitto Test and VF), cold allodynia,	The central antinociceptive action of BoNT-A might be associated with the activity of the endogenous opioid system (involving the μ-opioid receptor).
(Drinovac et al., 2014)	Sciatic nerve partial transection Formalin injection	R	Sciatic nerve	i.pl.	Nocifensive behaviors (licking, flinching and shaking of the injected paw) Mechanical sensitivity (VF)	BoNT-A’s antinociceptive effects was blocked by bicuculline, suggesting an association of the GABA-A receptors and BoNT/A analgesia.
(Matak et al., 2017)	Partial sciatic nerve ligation model CFA injection Formalin injection	M	Sciatic nerve	i.pl.	Mechanical sensitivity (dynamic plantar aesthesiomete r), nocifensive behaviors (paw lifting and licking)	BoNT-A did not reduce pain in substance P and neurokinin 1 receptor knockout mice.
(J. Wang et al., 2017)	SNI	R	Peroneal and tibial nerves	i.pl.	Mechanical sensitivity (VF), cold plate test	The new biotherapeutic LC/E-BoNT-A was an efficacious, locally-applied, and long-acting anti-hyperalgesic, which left normal somatosensation unaltered in sham animals.
(Leese et al., 2023)	SNI	R	Peroneal and tibial nerves	i.pl.	Mechanical sensitivity (VF)	iBoNT enabled production of safe neuronal modulators lacking the adverse effects of muscle paralysis, which opened a new avenue for developing treatments for neuropathic pain.
(Filipović et al., 2012)	ION-CCI Formalin injection	R	Infraorbital nerve	injected into the vibrissal pad	Mechanical sensitivity (VF), nocifensive behaviors (facial rubbing time)	Both pain and dural neurogenic inflammation could be prevented by a single BoNT-A peripheral injection. Bilateral effects of BoNT-A and dependence on retrograde axonal transport suggest a central site of action.
(C. Wu et al., 2016)	ION-CCI	R	Infraorbital nerve	Subcutaneously into the whisker pad	Mechanical sensitivity (VF), motor behavior (Rotarod test)	BoNT-A attenuated mechanical hyperalgesia without affecting motor coordination. BoNT-A may act on the Vc via axonal transport, inhibit the upregulation of TRPA1, TRPV1 and TRPV2, and reduce central sensitization.
(Piovesan et al., 2016)	ION-CCI	R	Infraorbital nerve	Injected into the area of nerve ligation	Cold plate test	BoNT/A had an antinociceptive effect in sensitized animals and a pronociceptive effect in nonsensitized animals.
(W.-J. Chen et al., 2021)	ION-CCI	M	Distal infraorbital nerve	s.c. into the whisker pad	Mechanical sensitivity (VF), open field test, elevated plusmaze test, forced swimming test	Unilateral injection of BoNT-A attenuated bilateral mechanical hypersensitivity and anxiety-like behaviors, which may be associated with the inhibition of TLR2-mediated neuroinflammation in the Vc.
(Favre-Guilmard et al., 2009)	Paclitaxel-induced peripheral neuropathy Carrageenan injection	R	Hindpaw	i.pl.	Pressure pain (Randall-Selitto test)	BoNT-A reduced hyperalgesia in both the ipsilateral and the contralateral paws.

CFA, complete Freund’s adjuvant; CCI, chronic constriction injury; DRG, dorsal root ganglion; iBoNT,; ION, infraorbital nerve; i.pl; intraplantar; i.t., intrathecal; M, mice; PKA, protein kinase A; R, rats; SNI, spared nerve injury; s.c., subcutaneous; TLR, toll-like receptor; TRPA1, transient receptor potential cation channel subfamily A member 1; TRPV, transient receptor potential vanilloid; Vc, trigeminal nucleus caudalis; VF, Von Frey test; VGluT2, vesicular glutamate transporter 2; VNUT, vesicular nucleotide transporter.

Figure 1. Proposed mechanisms of analgesia for neuropathic pain by the peripheral administration of botulinum neurotoxin (BoNT).

After the peripheral administration of BoNT in rodents with peripheral neuropathic injuries (A), BoNT is taken up by the peripheral terminals of peptidergic neurons that are likely transient receptor potential vanilloid subtype 1 (TRPV1)-expressing afferents. The active BoNT interferes with Ca²⁺-dependent exocytosis of vesicles containing glutamate or neuropeptides, such as calcitonin gene-related peptide (CGRP) or substance P (SubP) (B), leading to decreased neurogenic inflammation and peripheral sensitization. BoNT is also internalized in mast cells (MC) and reduces degranulation, which also reduces peripheral inflammation. BoNT is retrogradely transported to the sensory ganglia and the central terminals of the primary afferents, which is abolished by colchicine, a microtubule depolymerizer (C). In the sensory ganglia, BoNT suppresses the upregulation of pronociceptive ion channels (e.g., Nav1.7, TRPV1, Transient Receptor Potential Ankyrin subtype 1) and pronociceptive cytokines (interleukin-1β, interleukin-6, interleukin-18, and tumor necrosis factor), or increases the expression of antinociceptive cytokines (interleukin-10 and interleukin-1 receptor antagonist protein) (D). BoNT, which is transported into the central terminals in the dorsal horn of the spinal cord or trigeminal nucleus, inhibits the release of neurotransmitters and neuropeptides (E). Peripheral administration of BoNT suppresses microglia activation and the phosphorylation of the N-methyl-D-aspartate (NMDA) receptor in dorsal horn neurons, which likely leads to reduced central sensitization. Knockouts of neurokinin 1 (NK1R) ablate BoNT-induced analgesia.

2. Peripheral mechanisms

Given that the major biological mechanism of action of BoNTs is disrupting neurotransmitter release from nerve terminals, it is logical that it attenuates neuropathic pain by reducing pronociceptive neurotransmitters upon injection. Such effects can occur at the peripheral terminals of primary afferents at the injection site. Additionally, BoNT can interfere with neurotransmitter release at the central terminals of primary afferents, as peripheral administration of BoNT can lead to SNAP-25 cleavage in the dorsal horn of the spinal cord (Matak et al., 2011, 2012). Peripheral administration of BoNT-A reduces glutamate release evoked by intraplantar formalin injections in the hindpaw and the dorsal horn of the spinal cord (Cui et al., 2004; Matak et al., 2011). Early studies have demonstrated that after peripheral administration of radiolabeled BoNT-A, radioactivity can be seen in the spinal cord. However, this was prevented when the antitoxin was given simultaneously with the toxin in rats and cats (Habermann, 1974; Wiegand et al., 1976). More recently, radiolabeled BoNT-A with technetium-99m was administered to the bladder wall of healthy female rats, which was detected in the dorsal root ganglion (DRG) and the spinal cord six hours after its administration (Papagiannopoulou et al., 2016). Similarly, BoNT-A was detected in the brains of rats after intramuscular administration (Tang-Liu et al., 2003). Evidence has supported the presumption that BoNT-A is taken up from the nerve terminal after peripheral administration and transported retrogradely through the axon. Taking into account that the mechanism of action of BoNT-A is via cleavage of the SNAP-25 protein, studies have found cleaved SNAP-25 protein in the dorsal horn of the spinal cord or trigeminal nucleus after peripheral application of BoNT-A, which is further prevented when colchicine is administered, indicating that BoNT-A is transported to the spinal cord through a microtubule-dependent mechanism (Bu et al., 2022; G. He et al., 2024; Matak et al., 2011, 2012; T. Wu et al., 2016). Retrograde transport of BoNT from peripheral terminals to central terminals causes analgesia as colchicine-mediated disruption of axonal transport in sensory neurons prevents the analgesic effects of BoNT (Bach-Rojecky & Lacković, 2009; Matak et al., 2011). Additionally, cleaved SNAP-25 proteins in the ventral horn of the spinal cord were prevented after the administration of lumbar intrathecal application of BoNT-A-neutralizing antitoxin (Matak, 2020). Peripheral BoNT-induced cleavage of SNAP-25 in central terminals within the spinal cord or trigeminal nucleus caudalis is prevented by chemical denervation of transient receptor potential vanilloid 1 (TRPV1)-expressing afferents using capsaicin (Matak et al., 2014), suggesting that BoNT-mediated analgesia is largely dependent on the disruption of the neurotransmission of TRPV1-expressing afferents. The contribution of TRPV1+ afferents to the maintenance of neuropathic pain is convincing. Defunctionalization of nociceptors induced by peripheral administration of capsaicin, a ligand of TRPV1, has been widely used for treating peripheral neuropathic pain in patients (Arora et al., 2021). In CCIs of the trigeminal nerve in mice, peripheral administration of capsaicin produces long-lasting analgesia for mechanical allodynia and ongoing pain, which depends on the ablation of TRPV1+ nerve terminals at the site of capsaicin administration (Arora et al., 2022; S. Wang et al., 2020). Therefore, it is reasonable to presume that BoNT attenuates neuropathic pain by preventing the neurotransmission of TRPV1+ nociceptors.

2.1. Reducing the release of neuropeptides and neurotransmitters

BoNTs can induce analgesia for neuropathic pain by modulating the release of neurotransmitters and neuropeptides from peripheral and central terminals of nociceptive afferents. In rats with neuropathic injuries, dissociated sensory neurons from BoNT-A show a substantial reduction of KCl-induced vesicular release (Kitamura et al., 2009). Calcitonin gene-related peptide (CGRP) plays a crucial role in the transmission and modulation of nociceptive signals. It promotes vasodilation and increases the release of other neuropeptides involved in pain transmission (Benemei et al., 2009). The development of CGRP-targeted therapies has clear benefits for migraine patients (Edvinsson et al., 2018), and the therapeutic effects of BoNT-A on migraine headaches can be attributed to reduced CGRP release (Burstein et al., 2020). BoNT-A inhibits the release of CGRP from rat trigeminal ganglia neurons (Durham et al., 2004) and reduces the level of CGRP in the plasma and saliva of patients (Cady et al., 2014; Cernuda-Morollón et al., 2015) with migraine headaches. Although the contribution of CGRP to neuropathic pain has been reported (Iyengar et al., 2017) in preclinical studies (Jang et al., 2004; Lee & Kim, 2007; Nitzan-Luques et al., 2013), the efficacy of anti-CGRP monoclonal antibodies in clinical neuropathic pain needs to be determined. A recent study showed that anti-CGRP antibodies attenuate neuropathic pain in patients with migraine headaches (Kang & Govindarajan, 2021). Nevertheless, BoNT-A may inhibit neuropathic pain by modulating the effects of CGRP. Therefore, enhancing the efficacy of BoNT-A in regulating CGRP-mediated signaling may be beneficial. For example, BoNT-A cannot abolish the CGRP1 receptor-mediated effect of capsaicin or its elevation of CGRP release from trigeminal ganglion (TG) neurons in culture. This has been overcome by using a recombinant chimera of BoNT-A and BoNT-E, which involves /EA targeting of nociceptive neurons and inhibiting CGRP release in vitro and in situ (Meng et al., 2009).

Substance P is a member of the tachykinin family of peptides, which is released from immune cells and terminals of primary sensory neurons in response to stimulation. The primary function of substance P is to transmit nociceptive signals from peripheral tissues to the CNS by binding to specific receptors, such as neurokinin-1 (NK1) receptors. NK1 receptor antagonists attenuate neuropathic pain in rodents (Jang et al., 2004; Lee & Kim, 2007), and the interference of substance P signaling can be analgesic for neuropathic pain. The contribution of substance P in BoNT-induced analgesia has been determined using genetic models. In wild-type (WT) mice, acute, formalin-evoked pain and chronic neuropathic and inflammatory pain were reduced by peripherally injected BoNT-A. In contrast, tachykinin 1 (Tac1) knockout (KO) and Tac1r (an NK1 receptor) KO mice exhibited no BoNT-A-induced analgesic effects (Matak et al., 2017). As SNAP-25 cleavage occurs in WT and KO mice, deficient analgesia in KO mice is not due to the lack of BoNT enzymatic action, supporting the idea that substance P-NK1R neurotransmission modulates BoNT-A antinociceptive activity.

Glutamate is important for nociceptive transmission in the spinal cord. Glutamate within peripheral tissues is pro-nociceptive and plays a role in the development and progression of pain (Chung & Ro, 2020). Therefore, decreasing glutamate release or synthesis from primary afferents can lead to anti-nociceptive effects. A recent study supports this notion in the context of neuropathic pain (J. Wang et al., 2019). In rats with chronic constriction neuropathic injury, intraplantar administration of BoNT/A attenuated mechanical allodynia. Peripheral BoNT-A downregulated SNAP-25 and vesicular glutamate transporter 2 (VGluT2) in the dorsal horn of the spinal cord and reduced glutamate concentration in the spinal cord. Virally mediated overexpression of VgluT2 abolishes the analgesia by BoNT-A for neuropathic pain. Therefore, the analgesic effect of BoNT-A on neuropathic pain is attributable to reduced VgluT2 and glutamate (J. Wang et al., 2019).

2.2. Regulation of peripheral inflammation at the injection site

Peripheral administration of BoNT-A decreases inflammation at the injection site. In rats with ankle joint inflammation, BoNT-A was found to reduce the expression of interleukin-1β and tumor necrosis factor-α (TNF-α) in synovial tissues (L. Wang et al., 2017). These effects are likely secondary to BoNT-A interfering with the release of neuropeptides from sensory nerve terminals, which might reduce neurogenic inflammation at the injection site. However, it is also possible that BoNT-A directly interferes with the function of immune cells at the injection site. Treatment of human or mouse mast cells with BoNT-A or BoNT-B reduces the cleavage of SNAP-25 in mast cells and decreases degranulation, leading to decreased skin inflammation (Choi et al., 2019). Given the pivotal roles of mast cells in inflammatory pain (Green et al., 2019), their direct regulation by BoNT-A could be an important mechanism of BoNT-A-induced analgesia. The anti-inflammatory effects of BoNT-A at the injection site (Gfrerer et al., 2022) have also been shown in humans. Peripheral administration of BoNT-A regulates gene expression at the injection site. BoNT-A was injected into the muscles around the occiput of patients with chronic bilateral occipital headaches. During pre-scheduled occipital nerve decompression surgery, tissues around the injection site were collected and used for targeted transcriptomics. BoNT-A injection reduced the activation of inflammatory pathways (such as adaptive and innate immune responses, lymphocyte activation, cytokine, chemokine, NF-kB, TNF-α, and toll-like receptor signaling) and immune cells (neutrophils, macrophages, cytotoxic T cells, B cells, natural killer (NK) cells, and dendritic cells) (Gfrerer et al., 2022). These results suggest that BoNT-A analgesia for headaches may be attributable to its anti-inflammatory effects at the injection site. However, the contribution of these effects of BoNT/A at the injection site to analgesia for neuropathic pain remains unknown.

2.3. Transcriptional regulation of pronociceptive genes in sensory ganglia

RNAseq analysis of the DRG in experimental knee arthritis in rats with intra-articular injections of BoNT-A showed changes in multiple pathways, including the PI3K-Akt signaling pathway, metabolic pathways, extracellular matrix-receptor interactions, focal adhesion, and cytokine-cytokine receptor interactions (Li et al., 2021). At the individual gene level, S100A9, S100A8, and MMP8 transcripts were decreased in the DRG of rats with complete Freund’s adjuvant (CFA)-induced knee arthritis, while the injection of BoNT-A into the knee joints upregulated their expression (Li et al., 2021). Although the anti-inflammatory roles of these genes have been demonstrated in other contexts (Owen et al., 2004; Pagano et al., 2002), their contribution to BoNT-A-induced analgesia is unknown. Nonetheless, this report indicates that the peripheral administration of BoNT-A can modify gene expression in sensory ganglia, which could contribute to analgesia for neuropathic pain. A recent study found transcriptional changes in sensory ganglia after the viral expression of BoNT-A LC protease in sensory neurons using the sensory neuron-specific Pirt promoter. The overexpression of the BoNT-A subunit decreases the transcripts of pronociceptive genes in cultured sensory neurons stimulated with supernatant from activated macrophages. The downregulated pronociceptive genes included those encoding substance P, CGRP, 5-hydroxytryptamine receptor 3A, Nav1.7, TRPV1, and transient receptor potential cation channel subfamily A member 1 (TRPA1). Therefore, BoNT-induced analgesia can be partly attributable to the transcriptional regulation of nociceptive genes. Mechanisms of transcriptional regulation by BoNT-A and their contribution to analgesia in vivo need to be studied. Peripheral administration of BoNT-A can also upregulate transcripts of antinociceptive genes, such as prodynorphin, in the DRG of rats with sciatic nerve injury (Mika et al., 2011), which could contribute to analgesia.

2.4. Post-transcriptional regulation of the pronociceptive ion channels in sensory ganglia

Peripheral injections of BoNT-A inhibit the expression of multiple pronociceptive genes associated with neuropathic pain. For example, a cutaneous injection of BoNT-A inhibits the upregulation of Nav1.7 in the trigeminal ganglia in a rat model of trigeminal neuropathic pain produced by malpositioned dental implants (K. Y. Yang et al., 2016). Suppression of Nav1.7 is known to produce analgesia for neuropathic pain (Moreno et al., 2021; Moutal et al., 2020). Intraplantar injections of BoNT-A also reduce neuropathic pain induced by ventral root transection in rats, accompanied by the inhibition of TRPV1 or P2X3 upregulation in the DRG (Xiao et al., 2011, 2013). As previously mentioned, suppressing pronociceptive gene expression by BoNT-A can be mediated by transcriptional regulation within the sensory ganglia. BoNT-A also post-translationally regulates pronociceptive molecules. In a rat model of knee-joint arthritis, intra-articular injections of BoNT-A decreased the TRPV1 protein level in the DRG without changes in mRNA levels (Fan et al., 2017). Consistently, subcutaneous BoNT-A injections into the face of rats decreased TRPV1-immunoreactive neurons in the TG and TRPV1-immunoreactive afferent terminals, which was not mediated by transcriptional changes (Shimizu et al., 2012). Interestingly, TRPV1 in the plasma membrane was decreased in TG neurons after BoNT-A administration. Furthermore, TRPV1 Y200F (a mutant TRPV1 that lacks membrane trafficking) showed increased proteasome-mediated proteolysis. These results suggest that BoNT-A can reduce TRPV1 in the sensory ganglia by decreasing membrane trafficking and increasing proteosome-mediated degradation. Furthermore, BoNT-A prevents the upregulation of TRPV1 cell surface expression after exposing cultured sensory neurons to inflammatory soup or TNF-α (Moore et al., 2023; Nugent et al., 2018). Since TRPV1 localized in the central terminals of trigeminal afferents within the trigeminal nucleus caudalis contributes to the maintenance of neuropathic pain (Kim et al., 2014), decreased membrane translocalization and the expression of TRPV1 by peripheral administration BoNT-A can lead to neuropathic pain analgesia. BoNT-A also reduces the surface localization of TRPA1 in cultured sensory ganglia neurons (Moore et al., 2023; Nugent et al., 2018). Peripheral injections of BoNT-A decreases transcripts and protein levels of TRPA1 in the DRG (Cao et al., 2017). In rats with a CCI of the infraorbital nerve, TRPV1 and TRPA1 protein levels are upregulated in the trigeminal nucleus caudalis, which is reduced by BoNT-A (C. Wu et al., 2016). The inhibitory effects of BoNT-A on TRPA1 upregulation in sensory ganglia can lead to analgesia for neuropathic pain, as neuronal and non-neuronal TRPA1 plays a role in neuropathic pain (Iannone et al., 2023). Considering the preferential targeting of TRPV1-expressing afferents by BoNT-A, the regulation of TRPV1 and TRPA1 by BoNT-A should be an important mechanism of analgesia for neuropathic pain.

2.5. Regulation of proinflammatory cytokines and glia in sensory ganglia

Peripheral administration of BoNT-A leads to the reduction of proinflammatory cytokines within the sensory ganglia. For example, the expressions of IL-1β, IL-6, and TNF-α in the TG are reduced by the peripheral administration of BoNT-A in rats with trigeminal root compression (Cho et al., 2022). In rats with a CCI, peripheral BoNT-A suppresses the upregulation of IL-18 and IL-1β, whereas it increases the level of anti-inflammatory cytokines, such as IL-10 and IL-1RA, in the DRG (Zychowska et al., 2016). Robust changes of intraganglionic cytokines after nerve injury indicate activation of glial and immune cells within the sensory ganglia, which involves intricate interactions among neuronal and non-neuronal cells (Shinoda et al., 2021). Intraganglionic mechanisms contribute to persistent and ectopic hyperalgesia. Therefore, regulating intraganglionic proinflammatory cytokines by BoNT-A can lead to analgesia. However, mechanistically it is unknown whether the altered activation of the glia or decreased cytokines is mediated by the direct effects of BoNT-A in the glia within the sensory ganglia in vivo. BoNT-A most likely regulates the glia via the regulation of sensory neurons, e.g., by preventing transmitters or neuropeptide release. This may also apply to the situation in the CNS. A more detailed review of this topic is available elsewhere (Luvisetto, 2022).

3. Central mechanisms

3.1. Transport of BoNT into the CNS after peripheral injection

Although there is little doubt that BoNT-A can be transported along primary afferents and to the dorsal horn of the spinal cord or trigeminal nucleus, whether it is transported to higher structures remains controversial. Previous studies have found that BoNT-A can be trans-synaptically transported across neurons. While it has been shown that BoNT-A-truncated SNAP-25 proteins appear at distant sites, which project to the infusion site after injections in the hippocampus and the superior colliculus, no SNAP-25 fragments have been observed in unconnected areas (Antonucci et al., 2008). Similar results were observed when BoNT-A was peripherally injected into whisker muscles, and cleaved SNAP-25 protein was observed in the facial nucleus (Antonucci et al., 2008). These results have been replicated by other researchers using the rat retinotectal pathway (Restani et al., 2011) and the nasolabial musculature, where BoNT-A was retrogradely transported to the facial nucleus and further transcytosed to upstream cholinergic neurons (Caleo et al., 2018). In contrast, Matak et al., observed that after BoNT-A was administered to the whisker pad, SNAP-25 cleaved protein was only observed in the caudalis and oralis trigeminal nucleus, but not in other superior structures such as the thalamus, hypothalamus, sensory cortex, or locus coeruleus (Matak et al., 2014). Nonetheless, it is unknown whether BoNT-A-induced analgesia for neuropathic pain is attributable to trans-synaptic distribution and concomitant direct effects on higher brain structures.

Some studies have also evaluated the possible transport of BoNT-A to contralateral neurons in spinal nerves. BoNT-A has been shown to reduce the grip strength and muscle action potentials of ipsilateral and contralateral muscles, and similarly cleaved SNAP-25 protein levels were observed in the ipsilateral and contralateral dorsal and ventral horns of the spinal cord (Akaike et al., 2013; Koizumi et al., 2014; Torii et al., 2011). A recent study demonstrated bilateral analgesia by unilateral BoNT-A administration (Waskitho et al., 2021). Unilateral injection of BoNT-A in rats with cisplatin-induced neuropathy attenuates allodynia bilaterally. Contralateral injection of BoNT-A decreases allodynia in the ipsilateral to the infraorbital nerve constriction. While the mechanisms of bilateral distribution following unilateral administration of BoNT-A are not well understood, transport through the bloodstream cannot be dismissed. BoNT-A has been observed in the bloodstream and in contralateral muscles following unilateral administration (Tang-Liu et al., 2003; Waskitho et al., 2021). Torii et al. demonstrated that grip strength was not diminished in contralateral muscles when retrograde transport of BoNT-A was inhibited through the administration of colchicine or neurotomy (Akaike et al., 2013; Torii et al., 2011). However, BoNT-A is too large to permeate the blood-brain barrier, and it is questionable whether BoNT-A can be transported into the brain through hematogenous transport. While BoNT-A can be transported via blood, its pharmacokinetics at clinical doses have yet to be elucidated.

3.2. Changes in the CNS following peripheral administration of BoNT

Peripheral administration of BoNT-A produces a comprehensive modulation of neuronal and non-neuronal components in the nociceptive pathways in peripheral neuropathic pain models. It is unknown if peripherally delivered BoNT-A is trans-synaptically or transcytotically transported from primary afferents to glial or post-synaptic neurons within the CNS and whether direct modulation of glial and post-synaptic central neurons by transported BoNT-A induces analgesia for neuropathic pain. Therefore, it would be reasonable to interpret the majority of reported changes within the CNS as a consequence of modulating the primary afferents discussed above. The effects observed in the CNS through different neuroimaging techniques after peripheral administration of BoNT-A in patients with dystonia have been reviewed (Hok et al., 2021; Luvisetto, 2021). Cortical reorganization after peripheral intervention has been frequently demonstrated and may imply an indirect effect of BoNT-A. While there have been no neuroimaging studies showing brain changes following BoNT-A treatment in patients with neuropathic pain, it may produce changes in nociceptive pathways in the brain.

Peripheral administration of BoNT-A induces neurobiological changes in the CNS associated with analgesia. Glutamate is the most abundant neurotransmitter in the CNS and regulates multiple physiological processes; thus, its dysfunction has been implicated in multiple pathological processes, including chronic pain. Glutamate can bind to ionotropic α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid and N-methyl-D-aspartate (NMDA) receptors in postsynaptic neurons to mediate a rapid response in physiological circumstances, but phosphorylation of these receptors has been shown to cause long-term synaptic plasticity associated with central sensitization (Bleakman et al., 2006). The phosphorylation of the NR2B subunit of the NMDA receptor has been associated with the central sensitization process (Lin et al., 2015) and has also been found to be upregulated in the dorsal horn of the spinal cord after CCI of the sciatic nerve (J. He et al., 2023; Liu et al., 2015). The perineural application of BoNT-A inhibits long-term potentiation and reduces the expression of the phosphorylated NR2B subunit (J. He et al., 2023). These effects may be mediated by decreased glutamate release from presynaptic primary afferent terminals after peripheral administration of BoNT-A (as discussed above), rather than the direct effects of BoNT-A in postsynaptic neurons.

BoNT-A interferes with pronociceptive receptors and inhibitory nociceptive systems in the spinal cord. Spinal inhibitory signaling is important in neuropathic pain, and glycine concentration in the synapse is regulated by glycine transporters. CCI increases the expression of glycine transporter 2 (GlyT2) in the spinal cord, which is downregulated by intraplantar BoNT-A (J. Wang et al., 2022). BoNT-A-induced analgesia for neuropathic pain is prevented by virally overexpressed GlyT2 in the spinal cord. The spinal GABA-A and μ opioid receptor are also implicated in BoNT-A-induced analgesia for neuropathic pain. Intrathecally administered antagonists against GABA-A or μ opioid receptors can abolish BoNT-A-induced analgesia for neuropathic pain induced by sciatic nerve transection (Drinovac et al., 2013, 2014). Additionally, BoNT-A given a day before repeated administrations of morphine increases the antinociceptive effect of morphine and reduces morphine tolerance, possibly through the reduction in spinal glial activation and increased mu opioid expression (Vacca et al., 2013).

Purinergic signaling in the spinal cord plays an important role in the pathogenesis of neuropathic pain (Inoue & Tsuda, 2021). Vesicular nucleotide transporter (VNUT), a product of SLC17A9, regulates ATP release in the spinal cord, and VNUT-deficient mice exhibit reduced neuropathic pain (Kato et al., 2017; Masuda et al., 2016). Intrathecal injection of BoNT-A suppresses the upregulation of the expression of VNUT and ATP release in the spinal cord of rats with neuropathic injury (Shi et al., 2023). The analgesic effects of BoNT-A on neuropathic pain are prevented by virus-mediated overexpression of VNUT in the spinal cord, suggesting interference of purinergic signaling is a plausible mechanism of BoNT-A-induced analgesia. However, it is unclear whether BoNT-A targets VNUT in primary afferents or any other cells within the spinal cord.

Activation of microglia after a CCI has been well characterized and is an important mechanism for the development of the neuroinflammatory process. BoNT-A can reduce microglia activation induced by a CCI (Gui et al., 2020; J. He et al., 2023; Zychowska et al., 2016). Microglia can be activated as two different phenotypes, M1, the proinflammatory phenotype, which is characteristic of the production of proinflammatory cytokines and reactive oxygen species, and M2, the anti-inflammatory phenotype involved in wound repair and debris clearance (Cherry et al., 2014). CCIs have been shown to increase the M1 phenotype. However, in a study conducted by Gui et al., the administration of BoNT-A upregulated the M2 phenotype after a CCI, likely by promoting glial polarization to the M2 phenotype through the suppression of the glial receptor, P2X7 (Gui et al., 2020). Still, it is unknown whether peripherally administered BoNT-A is transcytotically transported to microglia and directly influences microglial function. Alternatively, microglial modulation by BoNT-A could result from the modulation of cytokines or purinergic signaling in primary afferents.

After a CCI, BoNT-A has been shown to reduce the release of the proinflammatory factors IL-18 (J. He et al., 2023), IL-1β, IL-6, TNF-α in the trigeminal nucleus (W.-J. Chen et al., 2021), and the spinal dorsal horn (X. Wang et al., 2020). The inhibition of these interleukins appears to be mediated through the suppression of TLR2/myD88 signaling. However, the cytokine changes in the CNS after peripheral administration of BoNT-A appear to be dependent on experimental conditions. For example, the expression of IL-18 and IL-6 in the spinal cord after BoNT treatment was negligible in another report (Zychowska et al., 2016).

4. New approaches for more specific and efficacious targeting to treat neuropathic pain

Despite its application for treating migraine headaches, BoNT has limited applicability in treating chronic pain: the efficacy is often modest, its effects on motor neurons produce paralysis, and the extent of targeting in the sensory nociceptive system is not clearly understood. Therefore, new methods for selectively targeting BoNT to the nociceptive system should greatly enhance our ability to manage chronic pain. To improve the therapeutic potential of BoNT, multiple types of native and engineered BoNTs have been comprehensively studied and well described in recent reviews (Rasetti-Escargueil & Popoff, 2020; Steward et al., 2021). In this section, we summarize recent efforts focused on specific targeting of BoNTs to nociceptive pathways for treating chronic pain (Table 3).

Table 3.

Examples of engineered botulinum neurotoxins (BoNTs) for analgesia

Name	Engineered form	Selectivity	Validation	Reference
BiTox		C fiber nociceptors? / non-paralytic	Inhibit extravasation in hindpaw / Inhibit mechanical hyperalgesia in SNI (IPL)	(Mangione et al., 2016)
el-iBoNT		A fiber nociceptors? / non-paralytic	Inhibit mechanical hyperalgesia in SNI (IPL)	(Leese et al., 2023)
BoNT/A- LC:SP		NK1R- expressing neurons	Inhibit thermal hyperalgesia in chemotherapy-induced neuropathy (ICT)	(Mustafa et al., 2013)
SP-BOT		NK1R- expressing neurons	Inhibit mechanical hyperalgesia in SNI or CFA (ITH)	(Maiarù et al., 2018)
DERM- BOT		MOR-expressing neurons	Inhibit mechanical hyperalgesia in SNI or CFA (ITH)	(Maiarù et al., 2018)
/D- CGRP 8-37		CGRP receptor-expressing cells	Inhibit substance P release from dissociated DRG neurons	(Tang et al., 2019)
LFN-LC + PA		Anthrx2- expressing nociceptors	Inhibit mechanical hyperalgesia in SNI (IPL)	(N. J. Yang et al., 2022)

Anthrx2, anthrax toxin receptor 2; Bd, binding domain of BoNT; CFA, complete Freund’s adjuvant; CGRP, calcitonin gene-related peptide; DRG, dorsal root ganglion; IPL, intraplantar injection; ICT, intracisternal injection; ITH, intrathecal injection; LC, light chain of BoNT; LFN, N-termal domain of LF of anthrax toxin; MOR, Mu opioid receptors; NKR1, neurokinin receptor 1; PA, protective antigen of anthrax toxin; SNI, spared nerve injury; Td, translocation domain

BiTox is an engineered form of BoNT-A in which the translocation domain is assembled with the receptor binding domain through the SNARE stapling system. Intraplantar (Mangione et al., 2016) injection of BiTox robustly inhibits mechanical hyperalgesia after spared nerve injury (SNI) and inflammation. Although BiTox is not developed to be targeted to a specific subset of afferents, it reduces plasma extravasation, suggesting an effect on peptidergic C fiber nociceptors. Interestingly, BiTox is not transported to the central terminals of primary afferents. A similar assembled approach has produced a more advanced BoNT, called elongated isopeptide-bonded BoNT (el-iBoNT) (Leese et al., 2023). Intraplantar injection of el-iBoNT attenuates mechanical hyperalgesia and improves gait behavior following SNIs. Furthermore, cleaved SNAP-25 in the skin was partly colocalized with TRPV1 and neurofilament 200, suggesting that el-iBoNT may preferentially affect A and C fiber nociceptors.

To target NK1 receptor-expressing second-order neurons, engineered toxins, whereby the LC of BoNT was conjugated with substance P (SP-BOT), have been developed. By administering an intracisternal or intrathecal injection (Maiarù et al., 2018; Mustafa et al., 2013), these conjugates successfully attenuate mechanical or thermal hyperalgesia in rat models of peripheral neuropathy. The analgesic effects of intrathecally injected SP-BOT on neuropathic pain appears to last up to 100 days in rats and the second injection can re-instate the analgesia (Maiarù et al., 2024). The LC of BoNT was also conjugated with dermorphin, which targets mu opioid receptor-expressing central neurons. The intrathecal injection of the conjugate effectively inhibits mechanical hyperalgesia after SNI or CFA-induced inflammation (Maiarù et al., 2018). BoNT subtype D (BoNT-D) was engineered to conjugate its heavy chain with a truncated CGRP (CGRP_8–37) that binds to CGRP receptors. When dissociated DRG neurons were treated with the conjugate, named /D-CGRP_8–37, VAMP1 cleavage was increased and depolarization-induced release of substance P was reduced (Tang et al., 2019). However, its analgesic efficacy in vivo has yet to be validated.

A recent study used an anthrax toxin subunit as cargo to deliver BoNT (N. J. Yang et al., 2022). Anthrax toxin binds to anthrax toxin receptor 2 (ANTXR2) enriched in Nav1.8+ nociceptors. ANTXR2 binds with two components of anthrax toxin, protective antigen (PA) and lethal factor (LF). Conjugation of the N-terminal domain of LF (LFN) with the LC of BoNT (LFN-LC) along with PA can specifically deliver the LC of BoNT into ANTXR2-expressing nociceptors. Intrathecal delivery of PA and LFN-LC reduces mechanical hyperalgesia following SNIs.

5. Conclusion

The current first-line treatment for neuropathic pain includes medications acting on the CNS, such as anti-convulsants or anti-depressants (Finnerup et al., 2015). However, these medications produce adverse side effects such as somnolence, dizziness, and nausea. Second and third-line treatments include topical capsaicin and BoNT-A, respectively (Finnerup et al., 2015). These two treatments commonly act on the peripheral nervous systems without direct action on the CNS. Both treatments target TRPV1-expressing nociceptors, leading to interference in nociceptive transmission, supporting the critical contributions of TRPV1-expressing nociceptors in neuropathic pain. Given the broad expression of TRPV1 in human sensory neurons (Shiers et al., 2020), interference of nociceptive transmission through TRPV1-expressing nociceptors likely produces even greater analgesia in humans than in rodents. Aside from procedural pain upon injection, topical capsaicin and BoNT-A do not produce adverse side effects. Due to the heterogeneous pathogenesis and etiology of peripheral neuropathies and neuropathic pain conditions, peripheral targeting may not work in every patient, and it is important to establish methods to identify responding patients whose chronic pain is maintained by peripheral mechanisms through sensitized nociceptors. Nevertheless, considering distinct sites of action, peripheral treatments may be used in combination with first-line ones to enhance analgesic efficacies. BoNT-A is a relatively safe analgesia whose primary site of action is restricted to primary afferents. Both clinical trials and mechanistic studies are needed to better understand the neurobiological mechanisms of BoNT-induced analgesia and enhance the use of BoNT in chronic pain treatment. A better understanding of the mechanisms involved can also expedite the development of novel engineered botulinum toxins to further improve the safety and clinical efficacy of treatment.

LIST OF ABBREVIATION

ANTXR2

anthrax toxin binds to anthrax toxin receptor 2

ATP

adenosine triphosphate

BoNT(s)

botulinum neurotoxin(s)

CCI

chronic constriction injury

CFA

complete Freund’s adjuvant

CGRP

calcitonin gene-related peptide

CNS

central nervous system

DRG

dorsal root ganglion

GABA

gamma-aminobutyric acid

GlyT2

glycine transporter 2

IL-1b/6

interleukin 1b/6

ION

infraorbital nerve

KO

knockout

LC

light chain

NF-kB

nuclear factor kappa B

NK1

neurokinin 1

NK1R

neurokinin 1 receptor

NMDA

N-methyl-D-aspartate

NR2B

N-methyl D-aspartate receptor subtype 2B

PA

protective antigen

RNAseq

ribonucleic acid sequencing

SNAP-25

synaptosome associated protein-25

SNARE

soluble N-ethylmaleimide-sensitive factor attachment protein receptor

SNI

spared nerve injury

Tac1

tachykinin 1

Tac1R

tachykinin 1 receptor

TG

trigeminal ganglion

TN

trigeminal neuralgia

TNF-a

tumor necrosis factor-a

TRPA1

transient receptor potential Ankyrin 1

TRPV1

transient receptor potential vanilloid 1

VAMP

vesicle-associated membrane protein

VAS

visual analogue scale

VF

von Frey test

VGluT2

vesicular glutamate transporter 2

VNUT

vesicular nucleotide transporter

WT

wild type