Animals

A total of 51 female Sprague Dawley rats (8 weeks old, weighing 200–250 g) were purchased from Koatech Bio, South Korea. The rats were housed in ventilated cages with autoclaved wood chip bedding under standard settings, with a temperature of 23 °C, a 12-h cycle of light and dark, and a humidity level of 30%. Except during experimental sessions, free access to standard rodent food and water was supplied. Between 9:00 am and 6:00 pm, surgical procedures and behavioral testing were carried out.

Surgical procedures

The chronic constriction of the infraorbital nerve (ION), as previously described, generated the TN model [22]. The rats were randomly assigned into two groups-TN (n = 23) and sham (n = 23). General anesthesia was used for all surgical operations, and intraperitoneal injections of 9 mg/kg xylazine (Rompun®, Bayer AG, Leverkusen, Germany) and 15 mg/kg Zoletil (Zoletil50®, Virbac Laboratories, Carros, France) were administered. Briefly, the infraorbital nerve was exposed through a tiny incision above the eye area. Using Dumont forceps for blunt dissection, the ION was isolated from the nearby connective tissue and muscle. The left ION bundle was wrapped in two loose ligations with a gap of 2 mm, and the incision was closed with 3–0 silk sutures. The contralateral sides were left intact in every rat. Rats that underwent sham surgery did not have their nerves clipped. The entire surgical process was performed in an aseptic manner. Following surgery, one rat was kept in each cage for recovery, with sterile bedding. In Additional file 1 (Figure S1), the experimental timetable is illustrated.

AAV virus injection for optogenetic manipulation

We employed adeno-associated virus (AAV) infused with a red fluorescent protein (mCherry) and astrocyte promoter (GFAP) to specifically target the astrocyte population in the TNC. We used AAV8-GFAP-eNpHR3.0-mCherry, which targets the expression of halorhodopsin in the astrocyte populations, for optogenetic investigations. AAV8-GFAP-mcherry was used as control or null virus. The KIST viral facility (Korea Institute of Science and Technology, Seoul, Republic of Korea) provided the viruses. The optogenetic and null viruses exhibited concentrations of 6.93 × 1013 GC/ml and 2.01 × 1014 GC/ml, respectively.

Following ION constriction or sham surgery, anesthetized rats were mounted on a stereotaxic platform for virus injection into our designated brainstem region (TNC). The rat's body temperature was maintained at 36 °C with the aid of a heating pad, and each group received the appropriate viral injection following the previous protocol [23, 24]. The coordinates were defined as mediolateral (ML) from the midline, anterior–posterior (AP) from the bregma, and dorsoventral (DV) from the brain's pial surface. Briefly, the cranial adjacent skin was slit to expose the skull, and a hole of 1 mm diameter was drilled into the skull (-13.8 mm AP and -2.5 mm ML, -7.8 mm DV). Using the Paxinos rat brain atlas [25] and previous publications [26, 27], the injection site coordinates for the TNC was determined and the tip of the Hamilton syringe was moved slowly into the TNC to prevent excessive tissue damage. Using an automated microsyringe pump (KD Scientific Legato® 130 Syringe Pump, Harvard Apparatus, Holliston, MA, USA), the optogenetic virus was administered to TN or sham rats (designated TN/eNpHR + or sham/eNpHR + further) while a null virus (designated TN/eNpHR- or sham/eNpHR- further) was administered to the other group. The infusion rate was 0.2 μL/min, and the injection volume per animal was 2 μL. For proper virus diffusion and to avoid leakage, the needle was left in the target for five minutes. The incision site was stitched once the needle was carefully withdrawn. The animal was removed from the stereotaxic frame and put back inside its home cage.

Behavioral assessment

A researcher who was unaware of the interventions performed the behavioral analysis. Each rat was handled and accustomed to the experimenter's hand and the testing area before each test. A day before (baseline test) and weekly for three weeks following ION constriction or sham treatment, the behavioral responses in response to mechanical and thermal stimulation were assessed.

For the open field test (OFT), each rat was kept in a black arena-sized 70 cm × 70 cm × 30 cm and videotaped for 5 min with an overhead camera that was positioned vertically 1 m above the test field. The number of explored areas and mobility rate were then determined from the video recordings utilizing automated open-source executable software (ToxTrac) for image-based tracking [28]. Post OFT, the mechanical hypersensitivity and allodynia of the orofacial region were assessed using the graded series of von Frey (VF) filaments (bending forces: 0.4, 1.4, 6, 8,15 g, 26 g) and the air puff test (APT), respectively. In both the ipsilateral and contralateral vibrissal pads, the filaments were evaluated in ascending order of force. A quick head retraction and/or an attack/escape reaction were counted as the responses. The mechanical response threshold was defined as the smallest force required to cause at least one of these responses when applied through von Frey filaments [29]. APT is based on the notion that when the injured orofacial region is exposed to continuous air puffs of varying pressures, the face withdraws. After 10 consecutive trials of continuous air-puff pressure (4 s long and 10 s apart) were applied to the trigeminal territory, withdrawal behavioral reactions such as an escape from air-puff or aggressive behaviors like biting were witnessed [30, 31]. The air puff pressure and intervals were regulated by a pneumatic pump module (BH2 system, Harvard Apparatus, USA). A 26-gauge metal tube (length: 10 cm) held at a 90-degree angle to the skin was used to deliver air puffs. For air puffs, the pressure cutoff was 40 psi (pounds per square inch). A considerable drop in the pressure threshold relative to baseline readings was used to characterize mechanical allodynia. A couple of drops of cold acetone were applied to the vibrissal pad region to measure the extent of the cold allodynia subsequently developed after the injury. Within the first two minutes following acetone delivery, nocifensive behavior such as rubbing or scratching on the ipsilateral whisker pad was noted [32]. Other than asymmetric orofacial grooming, body rubbing was not included. The number of responses was added up after each of the three behavioral tests, which were spaced out by five minutes.

In vivo electrophysiology and fiber optic cannula implantation

After three weeks of virus injection, in vivo recording and fiber cannula insertion were carried out for optimum viral expression. Extracellular recordings from anesthetized rats (15 mg/kg Zoletil® and 9 mg/kg xylazine as the anesthetic dose) were conducted in a Faraday cage in a quiet setting with low lighting. Rats were positioned inside a stereotaxic device, and a craniotomy was carried out above the thalamic ventral posteromedial (VPM) region (AP, 3.5 mm; ML, 2.8 mm, DV, -6.0 mm). A fiber optic cannula (MFC_200/230–0.48_###_ZF2.5_A45, Doric Lenses, Quebec City, Quebec, Canada) was stereotaxically implanted 0.2 mm dorsal to the viral injection site. Dental cement and superbond (Ortho-jet Pound Package, Lang Dental, Wheeling, IL, USA) were used to secure the cannula to the rat skull. A quartz-insulated carbon microelectrode (E1011-20, Carbostar-1, Kation Scientific, Minneapolis, MN, USA) was then placed contralaterally (DV = -5.6 mm) to the cannula implantation site. The attached screws served as anchors for the ground and reference wires that connected the electrode to the electronic interface board (EIB-36, Neuralynx, USA). A Cheetah Acquisition System (Neuralynx, USA) was implemented to acquire VPM thalamic output through the EIB-36's 36-channel headstage and preamplifiers. In a Digital Lynx SX data-acquisition system (Neuralynx, Bozeman, USA), neuronal signal bands were filtered at 0.9–6 kHz and the sampling frequency was set to 32 kHz. With simultaneous light-off and light-on instances in the TNC, acute recordings (spontaneous and evoked triggered with 10 g von Frey filament provocation on the ipsilateral whisker pad) from the VPM were obtained. Without supplementary anesthesia, the recording period for each rat lasted approximately two hours. Before the subsequent experiment, the animals were placed back in their home cage for a seven-day recovery period. Spikesort 3D software was used to further analyze the raw data to identify and isolate clusters. Waveforms with separable clusters and distinctive properties were referred to as single units and exported to NeuroExplorer (version 5, Nex Technologies, Colorado Springs, USA), which offers a full range of spike train analysis tools, including rate histograms, rasters, and burst analysis.

In vivo optogenetic manipulations in free-moving animals

Animals were taken to the behavior assessment area after a week of complying with the electrophysiological experiment and habituated for 30 min before the start of the test session. We used a diode-pumped solid-state 589-nm yellow DPSS laser (model: YL589T3-010FC, Shanghai Laser & Optics Century Co., Ltd., Shanghai, China) coupled with a monofiber optic patch chord that was placed over the cranial window in vivo for full-field stimulation to drive cells that were expressing mCherry and NpHR. We used a minimal irradiance of 0.4 mW/mm2 for 100% spiking inhibition on eNpHR3.0 expressing neurons during steady 593 nm laser illumination. A waveform generator (Keysight 33511B-CFG001, Santa Rosa, CA, USA) was employed to control the delivery of a 5 min yellow light pulse (593 nm, 8 mW, constant). When TNC astrocytes were silenced for five minutes using continuous yellow light, behavioral tests (OFT, VF, AT, and AP) were conducted. In both instances, a tethered fiber system and associated equipment were attached.

Immunostaining and imaging

Animals transfected with AAV8-GFAP-eNpHR3.0-mCherry or AAV8-GFAP-mCherry were optically stimulated with continuous yellow light for 10 min followed by deep anesthesia with a Zoletil/xylazine mix. PBS and 4% paraformaldehyde (PFA) were transcardially administered to rats. The brain and TG were removed, preserved in 4% PFA overnight, and then submerged in a 30% sucrose solution. The tissues were cryo-frozen and embedded in a compound with optimum cutting temperature (OCT; Tissue Tek®, Sakura, USA). Tissue  sections (20 μm) were obtained from the cryostat for immunostaining. Sections were fixed in acetone for 10 min and blocked at room temperature for 1 h in 10% normal goat serum. Slides were further incubated in mouse anti-GFAP (1:250, ab68428, Abcam), mouse anti-c-fos (1:1000, ab208942, Abcam), mouse anti-Iba1 (1: 100, ab283319, Abcam), and mouse anti-CGRP (1:50, ab81887, Abcam) overnight at 4 °C. Sections were washed and incubated for 2 h at room temperature with appropriate secondary antibodies (Alexa Fluor 488, ab150077, ab150113, Abcam). Mounting medium with DAPI (H-2000, Vectashield®, Vector Laboratories Inc., Burlingame, CA 94010) was used and coverslipped. The slides were imaged using a fluorescence microscope with cellSens Standard (Olympus Corp., Tokyo, Japan) or OlyVia 2.4 software (Olympus Corp., Tokyo, Japan). For each immunohistochemical staining and quantitative analysis, three coronal sections including the TNC and TG, were chosen. Quantifying the intensity in the red channel allowed for the identification of mCherry fluorescence. The mean fluorescence in a reference region close to the injection site was used to define and remove the background fluorescence from the image. By deducting the area of the selected cell*mean background fluorescence values from the integrated density, the fluorescent intensity or corrected total cell fluorescence was determined [33]. ImageJ software (National Institutes of Health, MD, USA) was used to compare and quantify the mean fluorescence intensity between groups.

To determine P2X3 expression in the TNC, chromogenic immunohistochemistry was used. The recombinant anti-P2X3 antibody (ab300493, Abcam) was incubated overnight at a 1:50 dilution with acetone-fixed frozen sections after preincubation for one hour in 2.5% normal horse serum. Following washing, slices were treated with biotinylated anti-mouse/rabbit IgG and then subjected to an avidin–biotin HRP procedure using an ABC kit (PK-7200, Vector Laboratories, Burlingame, CA) and diaminobenzidine (DAB substrate kit, SK-4100, Vector Laboratories, Burlingame, CA). Using the immunohistochemical image analysis toolbox of ImageJ software, the color detection method was used to quantify P2X3.

Statistical analysis

The data are presented as the mean ± standard deviation (SD). The sample sizes were established using G*Power (version 3.1.9.4, Germany) and prior experimental considerations. Rats exhibiting insufficient viral expression were excluded from the study. Two-tailed Mann–Whitney test, Welch’s t-test or an ordinary two-way analysis of variance (ANOVA), followed by Tukey’s or Sidak’s post hoc test based on the experimental terms, were used to conduct statistical analyses of the experiments to determine the significance. *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001 were used to identify significant differences. For the statistical evaluations, GraphPad Prism (version 9.1.0, Inc., San Diego, CA, USA) was utilized.