P2X3 receptor upregulation in trigeminal ganglion neurons through TNFα production in macrophages contributes to trigeminal neuropathic pain in rats

Background Trigeminal neuralgia is a characteristic disease that manifests as orofacial phasic or continuous severe pain triggered by innocuous orofacial stimulation; its mechanisms are not fully understood. In this study, we established a new animal model of trigeminal neuralgia and investigated the role of P2X3 receptor (P2X3R) alteration in the trigeminal ganglion (TG) via tumor necrosis factor alpha (TNFα) signaling in persistent orofacial pain. Methods Trigeminal nerve root compression (TNC) was performed in male Sprague-Dawley rats. Changes in the mechanical sensitivity of whisker pad skin, amount of TNFα in the TG, and number of P2X3R and TNF receptor-2 (TNFR2)-positive TG neurons were assessed following TNC. The effects of TNFR2 antagonism in TG and subcutaneous P2X3R antagonism on mechanical hypersensitivity following TNC were examined. Results TNC induced unilateral continuous orofacial mechanical allodynia, which was depressed by carbamazepine. The accumulation of macrophages showing amoeboid-like morphological changes and expression of TNFα in the TG was remarkably increased following TNC treatment. The number of P2X3R- and TNFR2-positive TG neurons innervating the orofacial skin was significantly increased following TNC. TNFα was released from activated macrophages that occurred in the TG following TNC, and TNFR2 antagonism in the TG significantly diminished the TNC-induced increase in P2X3R-immunoreactive TG neurons. Moreover, subcutaneous P2X3R antagonism in the whisker pad skin significantly depressed TNC-induced mechanical allodynia. Conclusions Therefore, it can be concluded that the signaling of TNFα released from activated macrophages in the TG induces the upregulation of P2X3R expression in TG neurons innervating the orofacial region, resulting in orofacial mechanical allodynia following TNC.


Background
Many current studies have evaluated the mechanism underlying trigeminal neuropathic pain using animal models of trigeminal nerve injury [1][2][3]. For example, chronic constriction injury or partial ligation of the infraorbital nerve is known to cause mechanical allodynia in the whisker pad skin [3,4]. Various molecules such as nitric oxide or neuropeptides are also generated in the injured trigeminal ganglion (TG) neurons and released from TG neurons, resulting in the sensitization of TG neurons [5,6]. These mechanisms are believed to be involved in orofacial neuropathic pain associated with trigeminal nerve injury.
Alternatively, it is well known that the trigeminal nerve root compression (TNC) by a cerebellar artery is clinically implicated as a cause of trigeminal neuralgia (TN) [7]. The widely used trigeminal neuropathic pain model induced by injury to the terminal branches of the trigeminal nerve is not a perfect model to mimic TN, but sheds light on the mechanisms underlying TN. To clarify the detailed mechanisms underlying TN, a novel animal model of TNC-mimicked TN needs to be developed. It has been reported that p38 phosphorylation in microglial cells is involved in TN using the TNC model developed by dental agar injection in the trigeminal nerve root [8,9]. However, many questions regarding the mechanisms underlying TN remain unclear.
Activated macrophages are known to produce various molecules such as tumor necrosis factor alpha (TNFα), and TNFα released from macrophages causes direct modulation of neuronal excitability [2]. TNFα is also known to increase in the cerebrospinal fluid and blood of patients with chronic neuropathic pain [10,11]. In animal experiments, TNFα upregulation in the dorsal root ganglion (DRG) contributes to mechanical allodynia following spinal nerve ligation [12,13]. Inferior alveolar nerve injury induces the infiltration of macrophages from which TNFα is released in the TG, which appears to be involved in the mechanical hypersensitivity in the whisker pad skin [2].
The ATP-gated P2X 3 receptor (P2X 3 R), which is expressed in nociceptive primary afferents, plays a significant role in facilitating transduction of noxious mechanical stimuli to the central nervous system [14]. Painful diabetic neuropathy is caused by membrane P2X 3 R upregulation in dorsal root ganglion neurons via protein kinase C activation in rats [15]. Mechanical allodynia following nerve injury caused by experimental lumbar disc herniation is elicited by neuronal hyperactivity in the DRG, attributable to the enhancement of P2X 3 R expression and function [16]. Nevertheless, the mechanisms responsible for the enhancement of membrane P2X 3 R expression and function in sensory neurons following peripheral nerve injury are not well understood. Together, TNFα, which is elevated in the sensory ganglion, may be listed as one of the likely candidate molecules to regulate membrane P2X 3 R expression and function in sensory neurons in neuropathic pain conditions.
In this study, we hypothesized that (1) TNC induces orofacial mechanical hyperalgesia, (2) TNC induces the infiltration of macrophages that release TNFα into the TG, and (3) TNFα signaling in the TG contributes to the increase in P2X 3 R expressing TG neurons and orofacial mechanical allodynia following TNC. To develop an effective therapy for TN, it is critical to establish an appropriate animal model and to elucidate the mechanisms of orofacial mechanical hypersensitivity using this model.

Animals
All experiments in this study were performed using male Sprague-Dawley rats (n = 125, 160-180 g, Japan SLC, Shizuoka, Japan). Transparent polycarbonate cages (length: 48 cm; width: 26.5 cm; height: 21 cm) with bedding made with paper shavings individually housed all rats. The temperature of the breeding facility was controlled at a constant temperature (23°C) and light-dark cycle (light on at 7: 00, off at 19:00). All rats were allowed free access to food and water in their cages. All procedures in this study adhered to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the guidelines of the International Association for the Study of Pain [17]. The Animal Experimentation Committee at Nihon University approved all procedures (AP18DEN005).

Trigeminal nerve root compression
Rats were anesthetized with intraperitoneal (i.p.) butorphanol (2.5 mg/kg, Meiji Seika Pharma, Tokyo, Japan), medetomidine (0.375 mg/kg, Zenoaq, Koriyama, Japan), and midazolam (2.0 mg/kg, Sandoz, Tokyo, Japan), and TNC was conducted. Briefly, a drill hole (diameter: 1 mm) in the skull right above the location of the left trigeminal root was made after mounting the rat onto a stereotaxic frame. The glass rod that was passed through the drill hole depressed the left trigeminal root (11 mm below the skull surface, 1.7 mm anterior from the posterior fontanelle, 2.8 mm lateral to the sagittal suture). The rod was anchored to the skull bone using three stainless steel screws with dental cement. As a sham control, the glass-rod was anchored to the skull bone without the TNC (7 mm below the skull surface, 1.7 mm anterior from the posterior fontanelle, 2.8 mm lateral to the sagittal suture). The dissected scalp wound was stitched together using a 6 − 0 silk suture.

Assessment of the whisker pad skin mechanical sensitivity
Rats were seasoned to keep their snout protruding from a cage that was equipped with a small fenestration previously for 5 to 7 days preceding the assessment of the  3 -IR TG neurons were counted using a BZ-9000 system (Keyence, Osaka, Japan). Immunofluorescence, which is more than twice as large as the background level, was defined as IR and labeled. There was no IR using the same procedure in the absence of the primary antibody. All immunohistochemical analyses were performed under blinded conditions.

Western blotting
On day 7 after TNC or sham treatment, transcardial perfusion using saline was performed under the abovedescribed anesthesia. Ipsilateral immediate-removed TG was homogenized in lysis buffer (137 mM NaCl; 20mM Tris-HCl, pH 8.0; 1 % NP40; 10 % glycerol; 1mM phenylmethylsulfonyl fluoride; 10 µg/mL aprotinin; 1 g/mL leupeptin; 0.05 mM sodium vanadate) at 4°C. The homogenized TG was centrifuged at 15,000 rpm for 10 min, and the protein concentration in the supernatant was analyzed using a protein assay kit (Bio-Rad, Hercules, CA, USA) at 4°C. Each protein sample (30 µg) obtained from the heat-denatured supernatants was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis using a 10 % acrylamide gel (Bio-Rad). Following the transfer of the protein samples to a polyvinylidene difluoride membrane (Trans-Blot Turbo Transfer Pack, Bio-Rad) using a Trans-Blot Turbo (Bio-Rad), the membrane was incubated in 3 % bovine serum albumin (BSA; Bovogen, Essendon, Australia). Afterwards, the membrane was incubated with rabbit anti-TNFα polyclonal antibody (1:200 diluted in Tris-buffered saline mixed with 0.1 % Tween 20 and 3 % BSA, Cat#ab66711, Abcam) for 12 h at 4°C. Next, the membrane was incubated with donkey horseradish peroxidase-conjugated anti-rabbit antibody (Cell Signaling, Danvers, MA, USA) for 2 h at 25°C. The band intensity was examined using a ChemiDoc MP system (Bio-Rad) after the detection of antibody binding by Western Lightning ELC Pro (Perki-nElmer, Waltham, MA, USA). The examined band intensity was normalized to the β-actin band intensity reprobed with anti-β-actin antibody (1:200; Santa Cruz, Santa Cruz, CA, USA) following bound protein removal by a stripping agent (Thermo Scientific).

Effect of intra-TG etanercept administration on mechanical sensitivity
Under the foregoing i.p. anesthesia, a drill hole (diameter: 1 mm) was bored in the skull immediately above the ipsilateral TG (2.8 mm anterior from the posterior fontanelle, 2.7 mm lateral to the sagittal suture) following skull stabilization onto the stereotaxic frame like the preceding TNC treatment. A guide cannula that was passed through the drill hole compressed the ipsilateral TG (9 mm below the skull surface). Subsequently, the cannula was anchored to the skull bone using dental cement, together with three screws implanted in the skull bone. To confirm the tip of the guide cannula positioning, multi-unit activities were induced by mechanical stimulation of the whisker pad skin using an electrode inserted into the guide cannula. On day 4 after the guide cannula installation, the TNC procedure was conducted as described above. Five microliters of etanercept (1 µg; Pfizer, New York, NY, USA) dissolved in saline or vehicle (saline) was administrated into the TG through the guide cannula under light anesthesia using 2 % isoflurane on day 4 after TNC or sham treatment. The measurement of whisker pad skin MHWT was performed before TNC or sham treatment (Pre), and on days 3, 7, 9, 11, and 18 after TNC or sham treatment.

Statistical analysis
For the analysis of mechanical sensitivity in the whisker pad skin, nonparametric procedures were selected because the normality and homogeneity of variances was not fulfilled. Data are presented as median and interquartile range (25-75 %). The upper and lower whiskers represent the maximum and minimum values, respectively. The Mann-Whitney or Kruskal-Wallis test followed by Dunn's multiple comparisons were used as non-parametric procedures. For the analysis of immunohistochemistry and Western blotting, Student's t-test or one-way analysis of variance with Tukey's multiple comparison test used. Data are expressed as the mean ± standard error. All statistical analyses were performed using GraphPad Prism ver. 8 (GraphPad Prism Software Inc., San Diego, CA, USA). N represents the number of rats that were tested. Statistical significance was defined as p < 0.05.

Change in mechanical sensitivity after TNC
There was no significant difference in MHWT of the bilateral whisker pad skin between the TNC and sham groups before the TNC and sham treatments. TNC induced a significant reduction in MHWT of the whisker pad skin ipsilateral to TNC compared to sham, and in addition contralateral to TNC from day 3 to day 7 (median on day 7, sham: 60 g; TNC (ipsilateral): 15 g, TNC (contralateral): 37.5 g) ( Fig. 1a and b). The reduction in MHWT of the whisker pad skin ipsilateral to TNC was recovered from day 10 onwards, and there was no significant difference in MHWT of the bilateral whisker pad skin between the TNC and sham groups on days 10, 14, and 21. The MHWT reduction of the whisker pad skin was significantly suppressed at 60 min after i.p. administration of carbamazepine on day 7 after TNC (median, vehicle: 10 g; carbamazepine: 43 g) (Fig. 1c). No behavioral signs of orofacial spontaneous pain were observed during the experimental period (data not shown).

Iba1 and TNFα expression in the depressed trigeminal root and TG
Iba1-IR cells were detected in the compressed trigeminal root on day 7 after TNC, and the mean number of Iba1-IR cells in the ipsilateral TG was significantly greater in the TNC (231.3 ± 31.9, n = 7) than in the sham group (121.3 ± 24.7, n = 6) (p < 0.05, Fig. 2a). Iba1-IR cells were also detected in the ipsilateral TG on day 7 after TNC or sham treatment, and many of them were also TNFα-IR (Fig. 2b). A small number of Iba1-IR and TNFα-IR cells was observed in sham-treated rats. TNFα protein level in ipsilateral TG was significantly higher on day 7 after TNC compared to that of sham or naive animals (naive: 0.55 ± 0.05; sham: 0.66 ± 0.11; TNC: 1.13 ± 0.12, n = 7 in each) (p < 0.01, Fig. 2c).

Effect of TNFα disruption on TRNR2 expression
TNFR2 expression was identified in FG-labeled TG neurons on day 7 after TNC or sham treatment, and many FG-labeled TG cells showed TNFR2-IR in TNC and sham rats (Fig. 3a). There was no significant difference in the number of TNFR2-IR TG neurons that innervate the whisker pad skin in the TNC and sham groups (Fig. 3b). Intra-TG administration of etanercept on day 4 after sham treatment did not change the MHWT of the ipsilateral whisker pad skin during the experimental period. TNC treatment TNC + vehicle: 10 g, n = 6; sham + Etanercept: 60 g, n = 5) (p < 0.05, Fig. 3c).

P2X 3 R expression and its signaling in the TG following TNC
P2X 3 R expression was identified in FG-labeled TG neurons on day 7 after TNC or sham treatment (Fig. 4a). Many FG-labeled cells were P2X3-IR in TNC rats, whereas a small number of them were P2X3-IR in shamtreated rats. TNC did not change the number of FGlabeled TG neurons compared with that of sham rats  (Fig. 4b). On the other hand, the number of P2X 3 R-IR TG neurons that innervate the whisker pad skin was significantly higher in TNC (18 ± 1.6 %, n = 5) the sham group (11 ± 1.6, n = 5) (p < 0.05, Fig. 4c). TNC treatment decreased the MHWT of the ipsilateral whisker pad skin on day 7, and subcutaneous A-317491 administration into the whisker pad skin significantly recovered the decreased MHWT 60 min after the administration (median, 0 min: 6 g; 60 min: 43 g, p < 0.05, n = 6) (Fig. 4d). Subcutaneous TNP-ATP administration into the whisker pad skin did not change the MHWT after administration in the sham group.

Discussion
Clinically, TN can be classified into three types (classical, secondary, idiopathic) according to the diagnostic criteria issued by the international association for the study of pain [7]. Of these three types, unilateral orofacial neuropathic pain, which is caused by neuropathological changes in the trigeminal nerve root attributed to unilateral vascular compression and triggered by innocuous orofacial stimulation, is defined as the classical TN [18]. Typical patients with classical TN do not suffer from only orofacial neuropathic pain, which is typified by paroxysmal stabbing pain in the trigeminal territory, but also sustained orofacial pain [19]. At present, the first line of treatment for TN is oral administration of CBZ, which can stabilize a hypersensitized neural membrane by blocking the voltage-gated sodium channel, which is known to be effective in virtually all TN patients [20]. In this study, TNC induced unilateral transient orofacial mechanical hypersensitivity that mimics pain hypersensitivity in the trigeminal territory in TN patients, which was depressed by CBZ administration. These findings indicate that the TNC model sheds light on the mechanisms underlying orofacial neuropathic pain referred from TN. However, not all cases of TN present neurovascular contact on the symptomatic side, though CBZ treatment was confirmed to be effective in almost all TN patients [20,21]. Moreover, about half of the TN patients are afflicted with concomitant spontaneous pain, which is rarely seen in the TNC model [18]. In addition, the TNC is not pulsating but permanent [18], which is different from vascular compression frequently underlying TN patients. Therefore, it is conceivable that the developed TNC model reflects a part of the neuropathophysiological condition of TN. In the future, it will be necessary to develop an animal model that is closer to the pathophysiology of TN. Iba1 is notable as a specific marker of macrophages [22]. Moreover, the morphological features of macrophages change with an amoeboid-like appearance, indicating that their activation accumulates in the inflamed tissue [23]. Previous studies have indicated that peripheral nerve injury is conducive to increase in proliferated and/or resident macrophages in the TG or DRG [2,[24][25][26]. This is the first report that TNC, aside from peripheral nerve injury, induces the accumulation of macrophages showing amoeboid-like morphological changes into the TG, though a marked increase in OX42 (a marker of macrophages) immunoreactive cells was reportedly induced in non-neuronal structures in the spinal trigeminal nucleus in a TNC mouse model [27]. Previous studies have shown that the chemokine C-C motif ligand 2 (CCL2) released from damaged neurons, also designated as monocyte chemoattractant protein 1, and its signaling activates macrophages via the C-C chemokine receptor type 2 (CCR2), which elicits activation and proliferation of macrophages in the sensory ganglion [28,29]. Additionally, peripheral nerve damage induces the enhancement of CCL2 expression and macrophage accumulation in sensory neurons through toll-like receptor 2 (TLR2) signaling [30]. From these studies, it can be assumed that TNC facilitates the remarkable accumulation of activated macrophages in the TG, which is likely induced by CCL2 signaling through TLR2 activation. However, it is not yet known whether exogenous macrophages infiltrate into the TG or activated macrophages proliferate in the TG; further research is needed to determine the origin of the increased macrophages in the TG.
Following peripheral nerve injury activated macrophages in the sensory ganglion release TNFα [31,32], whose levels are reported to be increased as well [33,34]. TNF-α release associated with macrophage activation accelerates intracellular signal transduction pathways, including extracellular signal-regulated kinase (ERK) and p38 mitogenactivated protein kinase (MAPK) cascades [35]. Peripheral nerve injury enhances the production of substance P (SP), which is a well-known mediator of neuroimmunomodulation and released from DRG neurons [36]. SP signaling regulates not only various functions of macrophages but also accelerates ERK1/2 and p38 MAPK cascades in macrophages [37]. Furthermore, the facilitation of the p38 MAPK cascade in macrophages infiltrates the nerveinjured region, resulting in TNFα production from macrophages [38]. Based on the present results, the accumulation of activated macrophages that express TNFα in the TG was immunohistochemically confirmed. TNC remarkably increased macrophage and TNFα protein expression in the ipsilateral TGs according to TNC. Taken together, these results suggest that the signaling of SP, which was likely released from TG neurons associated with TNC, activates macrophages in the TG, resulting in TNFα release from activated macrophages via the ERK1/2 and p38 MAPK signaling cascades.
In this study, P2X 3 R expression was identified in TNFR2-positive TG neurons innervating the whisker pad skin following TNC. The number of P2X 3 R-IR TG neurons that innervate the whisker pad skin was significantly increased following TNC. TNFα release from activated macrophages occurred in the TG following TNC, and the TNFR2 antagonism in the TG significantly diminished the TNC-induced increase of P2X 3 R-IR TG neurons innervating the whisker pad skin. Moreover, subcutaneous P2X 3 R antagonism in the whisker pad skin significantly depressed TNC-induced mechanical hypersensitivity. P2X 3 R, which is a ligand-gated ion channel, is expressed in small-diameter TG neurons; the upregulation of P2X 3 R-positive TG neurons after trigeminal nerve injury reportedly leads to trigeminal neuropathic pain [39]. It has been reported that exogenous TNFα by intra-DRG administration elicits DRG neuronal depolarization and hyperexcitability [40]. TNFα signaling is known to induce ERK phosphorylation in cultured DRG neurons [41], and increased ERK signaling upregulates the transcription of the P2X 3 R gene [42]. Together, these results suggest that TNC-induced orofacial mechanical hypersensitivity is, at least in part, dependent on the upregulation of P2X 3 R expression in TG neurons innervating the orofacial region via the increased TNFR2/ERK signaling cascade. In addition, trigeminal nerve injury activates satellite glial cells through the TG via gap junctions [1]; TNFα is released from activated satellite glial cells through enhanced P2X 7 signaling following peripheral sensory nerve injury [43,44].
Alternatively, following peripheral nerve injury, the voltage-gated sodium channel activation is augmented by TNFα signaling in DRG neurons in a concentrationdependent manner depending on the upregulation of tetrodotoxin (TTX)-sensitive and resistant current densities via the increased intracellular p38 MAPK signaling [34,45,46]. Membrane potassium ion conductance is also increased by TNFα signaling in a non-voltage-gated fashion, resulting in sensory neuronal hyperexcitability [47, 48]. These mechanisms to enhance sensory neuronal excitability by TNFα signaling are also potentially involved in TNC-induced orofacial mechanical hypersensitivity. However, the detailed mechanisms remain unknown. Therefore, further studies are required.

Conclusions
In conclusion, TNC could induce unilateral continuous orofacial mechanical hypersensitivity that mimics pain hypersensitivity in the trigeminal territory in patients with TN. The signaling of TNFα released from the activated macrophages and infiltrated into and/or proliferated in the TG induces the upregulation of P2X 3 R expression in TG neurons innervating the orofacial region, resulting in orofacial mechanical allodynia following TNC. Consequently, TG neuronal P2X 3 R hyperexpression by enhanced TNFα signaling in the TG is a potential target for TN treatment.