Animals

C57BL/6 mice were obtained from the Institute of Experimental Animals of Sun Yat-sen University. Atf4^+/− and Trpm3^−/− mice on the C57BL/6 background were purchased from Cyagen Biosciences Inc. All animals were housed in separate cages in a temperature-controlled (24 ± 1 °C) and humidity-controlled (50-60%) room under a 12/12-h light/dark cycle. The mice had ad libitum access to sterile water and standard laboratory chow. All animal experimental procedures were approved by Research Ethics Committee of Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences and were carried out in accordance with the guidelines of the National Institutes of Health on animal care and ethics⁵⁷. All animals were assigned randomly to different experimental or control groups.

Animal pain models and intrathecal injection

To produce inflammatory pain, CFA (20 µl) was injected into the plantar surface of the hindpaw. To produce neuropathic pain, the mice were anaesthetised, and the left L5 spinal nerve was isolated adjacent to the vertebral column and tightly ligated with 6-0 silk sutures distal to the DRG and proximal to the formation of the sciatic nerve. The L5 spinal nerves of sham-operated mice were identically exposed but not ligated. CIM0216 (Tocris Bioscience, catalogue no.: 5521), PS (R&D, catalogue no.: 5376) or capsaicin (Tocris Bioscience, catalogue no.: 0462) was injected intraplantarly (2.5 nmol/20 µl/paw, 2.5 nmol/20 µl/paw, or 1 nmol20 µl/paw, respectively), or intrathecally (1.25 nmol/10 µl/mouse) to induce spontaneous pain. Intrathecal administration was performed by a polyethylene-10 catheter that was inserted into the subarachnoid space of the mouse through the L5 to L6 intervertebral space so that the tip of the catheter was located at the L5 level, and 10 µl reagent was delivered into the cerebrospinal fluid.

Behavioural tests

The animals were habituated to the environment for at least 2 days before testing. All the behaviours tests were performed in a blinded manner.

Von Frey test

The mice were placed in plastic chambers on a mesh floor. Von Frey filaments with increasing grades of force were applied to the hindpaws of the mice. Each filament was applied 5 times during the test. The lowest filament force that elicited paw withdrawal more than 3 times during the test was defined as the mechanical threshold.

Dynamic mechanical test

Mice were acclimated to von Frey chambers for 1 h. The lateral side of the hindpaw was gently stroked with a 5/0 brush from heel to toe. Responses were scored as follows: 0 = no response; 1 = very short, fast movement/lifting of the paw; 2 = sustained lifting of the paw for more than 2 s towards the body or strong lateral lifting above the body level; and 3 = flinching, licking, or flicking of the affected paw. The average score of three trials per mouse was reported as the allodynia score.

Tape response test

The mice were conditioned to a round plexiglass container for at least 5 min. A 1-inch piece of laboratory tape was gently applied to the bottom centre of the mouse’s back. The mice were observed for 5 min, and the total number of responses to the tape was recorded. Biting or grabbing the tape or making an obvious “wet dog shake” movement to remove the tape from the back was considered a response.

Tail clip test

A small alligator clip (force, 700 g) was applied 1 cm from the base of the tail. The latency to attack/bite the clip was measured with a stopwatch. Upon attack, the clip was removed, and the animals were returned to their cages.

Pinprick test

The mice were acclimated to a von Frey chamber for 30 min, and 27 needles were applied to the glabrous skin of the hindpaw, taking care not to pierce the skin. Each mouse was tested 10 times at an interval of 1 min. Paw withdrawal, shaking, or licking was scored as a response, and the percentage of responses to the total number of tests was recorded.

Tail-flick test

The mice were gently restrained inside a cloth/cardboard pocket with their tails outside the pocket. The distal half of the tail was immersed in a 48 °C, 50 °C or 52 °C water bath, and the latency to vigorous withdrawal of the tail from the water was measured.

Hargreaves test

Thermal hypersensitivity was measured using a plantar test according to the method described by Hargreaves⁵⁸. Briefly, a radiant heat source beneath a glass floor was aimed at the fat pad on the plantar surface of the hindpaw. The latency to withdraw the hindpaw was measured. The hind paw of each mouse was soaked in a 43 °C water bath for 30 s, and then the mice were immediately placed in a plastic chamber. Two, 5, 10, 20 or 40 min later, the behavioural experiment was performed. The mice were tested individually.

Hot plate test

The mice were placed into a clear Plexiglas cylinder on top of a metal surface maintained at 50 °C, 52 °C or 55 °C. The latency to lick or shake either hindpaw was measured.

Evaporative cooling test

The mice were habituated to a plastic chamber on a mesh floor. A drop (10-20 μl) of acetone was applied to the hindpaws of the mice. The duration of flinching or licking behaviours within 1 min was measured.

Rotarod test

The mice were tested on a rotarod with a velocity that increased from 4 rpm to 40 rpm within 5 min. The mice were pre-trained for 2 days for adaptation. Then, the amount of time that each mouse spent on the rotarod before it fell off was recorded.

Spontaneous pain test

Spontaneous pain induced by intraplantar injection of CIM0216, PS or capsaicin was measured by counting the time and number of the mouse licked the affected paw or flinched within 2 min, spontaneous pain induced by intrathecal injection of CIM0216 or PS was assessed by counting the amount of time each mouse spent licking, flicking, biting, or flinching within 5 min after injection.

Cell culture and transfection

Mouse DRGs were freed from their connective tissue sheaths and broken into pieces with a pair of sclerotic scissors in DMEM/F12 (Gibco) at a low temperature. Following enzymatic and mechanical dissociation, the DRG neurons were plated on glass coverslips coated with poly-l-lysine (Sigma) in a humidified atmosphere (5% CO₂, 37 °C). The cells were used for electrophysiological recordings approximately 4 h to 24 h after plating. HEK293T cells were cultured in MEM with 10% foetal bovine serum. Full-length TRPM3-Flag, ATF4-His and KIF17-Flag plasmids were obtained from Synbio Technologies, and ATF4-His (aa 1–116, 117–232, 233–349) plasmids were modified from the full-length ATF4 plasmid. The Flag and His fusion proteins were co-expressed in HEK293T cells. The cells were transfected with 1–2 μg plasmid per 35-mm dish or 2–3 μg plasmid per 60-mm dish using Lipofectamine 2000 reagent (Invitrogen). The cells were used for subsequent experiments 24–48 h after transfection.

Electrophysiological recordings

Whole-cell patch-clamp recordings were performed with an EPC-10 amplifier and PULSE software (HEKA Electronics, Lambrecht) at RT^22,59. Currents were recorded with glass pipettes (1–3 MΩ resistance) fabricated from borosilicate glass capillaries using a Sutter P-97 puller (Sutter Instruments, Novato, CA). Voltage errors were minimised using 80-90% series resistance compensation. TRPM3 currents in DRG neurons were recorded with a ramp potential of −75 mV to +75 mV. The voltage for current analysis was ±75 mV. Neurons with a leak current of >500 pA or a series resistance of >10 MOhm were excluded. The extracellular solution contained (in mM) 140 NaCl, 3.5 KCl, 1 MgCl₂, 1.25 NaH₂PO₄, 10 HEPES, and 10 D-glucose (adjusted to pH 7.4 with NaOH). The pipette solution contained (in mM) 140 CsCl, 1 CaCl₂, 2 MgATP, 2 Na₂ATP, 10 EGTA, and 5 TEA-Cl (adjusted to pH 7.2 with CsOH). The osmolality of all solutions was adjusted to 310 mOsm.

Surface protein biotinylation

Plasma membrane protein expression was detected by cell surface biotinylation using a Cell Surface Protein Isolation Kit (Pierce, catalogue no.: 89881) according to the manufacturer’s instructions. Briefly, cells were washed with PBS and biotinylated with Sulfo-NHS-SS-Biotin in PBS for 30 min at 4 °C. After quenching, the cells were lysed, and labelled proteins were isolated by incubation with NeutrAvidin Agarose beads for 60 min at RT. After washing, the proteins were eluted by heating the beads for 5 min at 95 °C and prepared for immunoblotting.

Extraction of plasma membrane proteins

For membrane protein preparation, samples were homogenised on ice with a plasma membrane protein extraction kit (Invent Biotechnologies, catalogue no.: SM-005). The detailed protocol was as follows: DRG tissues or cultured DRG neurons were lysed with 200–500 µl buffer A. The filter cartridge was capped and centrifuged at 14,000 r.p.m. (16,000 g) for 30 s. The filter was discarded, and the pellet was resuspended by vortexing for 10 s and centrifuged at 3000 r.p.m. (700 g) for one min (the pellet contained the intact nuclei). The supernatant was transferred to a new tube and centrifuged at 4 °C for 10–30 min at 16,000 g. The supernatant (the cytosolic fraction) was removed, and the pellet (the total membrane protein fraction including organelles and plasma membranes) was saved. The total membrane protein fraction was resuspended in 200 µl buffer B by vortexing and centrifuged at 10,000 r.p.m. (7800 g) for 5 min at 4 °C. The pellet contained the organelle membrane proteins (in our study, cytoplasmic protein comprised the cytosolic fraction and organelle membrane fraction). The supernatant was carefully transferred to a fresh 2.0-ml microcentrifuge tube, and 1.6 ml cold PBS was added. The sample was mixed a few times by inverting and centrifuged at 14,000 r.p.m. (16,000 g) for 15–30 min. The supernatant was discarded, and the pellet (isolated plasma membrane proteins) was saved. Protein samples of different fractions were denatured and prepared for immunoblotting.

Preparation of the GST-fused proteins

GST-Fused Proteins were expressed in E. coli BL21. The bacteria grew in 2 × YTA media and the expression of proteins were induced by 1 mM Isopropyl β-D-thiogalactopyranoside (IPTG). The bacteria were then centrifuged, resuspended and ultrasonic treated to release protein. The proteins were purified by Glutathione-Sepharose beads (GE Healthcare), concentrated and quantified before use.

Co-immunoprecipitation and GST pull-down

Transfected HEK293T cells or DRG tissues were lysed in cold co-IP RIPA buffer [20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.1% Triton X-100, 1% sodium deoxycholate, 10 mM NaF, 1 mM EDTA, 1 mM PMSF, and 1 mg/ml leupeptin]. The lysates were centrifuged, and 5% of each supernatant was taken for the input sample. The remaining supernatants were incubated with 5–10 μg ATF4, TRPM3 or His antibody at 4 °C overnight and then with protein A/G beads (GE Healthcare) at 4 °C for 4 h. The immunoprecipitated samples were denatured and prepared for immunoblotting. Immunoprecipitation was performed with antibodies against ATF4, TRPM3, KIF17, KIF3A, KIF3B, KIF5A, KIF5B, KIFC2, Flag and His. To evaluate the direct interaction between ATF4 and TRPM3 or KIF17, 5 μg purified ATF4 protein (Proteintech, Ag16665) was incubated with 5 μg TRPM3-GST-Flag or KIF17-GST-Flag purified protein in 400 μl RIPA buffer. The proteins were incubated with the cell lysate at 4 °C overnight. Then, the GST-fused protein was precipitated with 10 μl of Glutathione-Sepharose beads at 4 °C for 2 h. The precipitant was washed, denatured and prepared for immunoblotting.

Western blotting

Spinal dorsal horn, dorsal root, DRG, sciatic nerve and sural nerve tissues or cultured DRG neurons were lysed and homogenised in cold RIPA buffer [50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 10 mM NaF, 1 mM EDTA, 1 mM PMSF, and 1 mg/ml leupeptin]. The protein samples were separated via gel electrophoresis (SDS-PAGE) and transferred onto PVDF membranes. The membranes were placed in blocking buffer for 1 h at RT and incubated in primary antibodies against ATF4 (goat, 1:1000, GeneTex, catalogue no.: GTX89973), TRPM3 (rabbit, 1:1000, Bioss, catalogue no.: bs-9046R; rabbit, 1:200, Alomone Labs, catalogue no.: ACC-050), TRPV1 (rabbit, 1:200, Alomone Labs, catalogue no.: ACC-030), TRPA1 (rabbit, 1:1000, ABclonal, catalogue no.: A12544), KIF17 (mouse, 1:100, Santa Cruz Biotechnology, catalogue no.: sc-137040), KIF3A (goat, 1:100, Santa Cruz Biotechnology, catalogue no.: sc-18745), KIF3B (rabbit, 1:100, Santa Cruz Biotechnology, catalogue no.: sc-50456), KIF5A (rabbit, 1:1000, Abcam, catalogue no.: ab5628), KIF5B (rabbit, 1:2000, Abcam, catalogue no.: ab167429), KIFC2 (rabbit, 1:100, Abcam, catalogue no.: ab3476), TfR (mouse, 1:1000, Thermo Fisher Scientific, catalogue no.: 13-6800), CXCR4 (rabbit, 1:1000, Abcam, catalogue no.: ab124824), CXCR7 (rabbit, 1:1000, Abcam, catalogue no.: ab72100), β-tubulin (mouse, 1:2000, Arigo, catalogue no.: ARG62347), Flag (rabbit, 1:1000, Cell Signaling Technology, catalogue no.: 14793) or His (rabbit, 1:1000, Cell Signaling Technology, catalogue no.: 12698) overnight at 4 °C. Next, the membranes were incubated with an HRP-conjugated secondary antibody (1:10,000). Enhanced chemiluminescence (ECL) solution (Millipore) was used to detect the immunocomplexes. Each band was quantified with computer-assisted imaging analysis software (Tanon Gis).

In situ hybridisation

To label KIF17, ATF4 and TRPM3, a KIF17 mRNA probe labelled with CY5, an ATF4 mRNA probe labelled with FITC, and a TRPM3 mRNA probe labelled with CY3 were constructed. The probes were designed by BersinBio, and the assay was conducted according to the manufacturer’s instructions^60,61. Paraffin sections were dewaxed in xylene, dehydrated in 70%, 80%, 95% and 100% ethanol for 5 min each. And digested in proteinase K digested at 55 °C for 10 min. The sections were soaked in preheated (78 °C) denaturing solution and denatured for 8 min. Fluorescent dye-labelled probe and probe diluent were mixed to produce a probe hybridisation mixture and denatured at 78 °C for 8 min. A total of 20–40 μl hybridisation reaction solution was dripped onto each sample, and the samples were co-denatured at 73 °C for 5–8 min and then hybridised overnight (16 h–20 h) at 53 °C. The sections were washed and photographed with a laser confocal microscope.

Luciferase assay

Twenty thousand BV-2 cells were seeded in triplicate in 48-well plates and allowed to settle for 24 h. One hundred nanograms control luciferase plasmid or luciferase reporter plasmid containing a fragment (−2052/+558) of Trpm3 promoter plus 1 ng pRL-TK Renilla plasmid (Promega) was transfected into the cells using Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer’s recommendations. Luciferase and Renilla signals were measured 36 h after transfection using the Dual Luciferase Reporter Assay Kit (Promega) according to a protocol provided by the manufacturer. To observe the effect of ATF4 on Trpm3 luciferase activity, BV-2 cells were transfected with ATF4 siRNA (30 nM) using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions 12 h before the experiments.

Immunohistochemistry and SIM

Mice were perfused with 4% paraformaldehyde (PFA). The spinal cord, DRG, skin and sciatic nerve were dissected and post-fixed in 4% PFA for 1 h. Next, the tissues were dehydrated in 30% sucrose and embedded for cryostat sectioning. The cryostat sections were incubated with primary antibodies against ATF4 (rabbit, 1:200, Abcam, catalogue no.: ab23760; rabbit, 1:100, Cell Signaling Technology, catalogue no.: 11815 S; goat, 1:100, GeneTex, catalogue no.: GTX89973), TRPM3 (rabbit, 1:100, Bioss, catalogue no.: bs-9046R; rabbit, 1:100, Alomone Labs, catalogue no.: ACC-050), IB4 (1:50, Sigma, catalogue no.: L2895), CGRP (mouse, 1:200, Abcam, catalogue no.: ab81887; goat, 1:200, Abcam, catalogue no.: ab36001), NF200 (mouse, 1:200, Sigma, catalogue no.: N0142), NeuN (mouse, 1:200, Millipore, catalogue no.: MAB377; rabbit, 1:200, Abcam, catalogue no.: ab177487) or KIF17 (mouse, 1:50, Santa Cruz Biotechnology, catalogue no.: sc-137040) at 4 °C overnight and then incubated with secondary antibodies (1:400) for 1 h at RT. Three-dimensional super-resolution images were captured using a three-dimensional structured illumination microscope with the N-SIM System and a CFI SR oil immersion objective lens (Apochromat TIRF×100, 1.49 numerical aperture, Nikon, Japan), and images were post-processed with Nikon NIS-Elements software. We used Image-Pro Plus 6.0 software to analyse the colocalization percentage of images. First, we recorded the intensity of colocalization spots and then measured the intensity of single-colour spots. The intensity of the colocalization spots divided by the intensity of the single-colour spots was considered the percentage of colocalization. To confirm the specificity of the ATF4 and TRPM3 antibodies, blocking experiments were conducted in DRG sections using a mixture of anti-ATF4 antibody and immunising blocking peptide (10 times the molar concentration of the antibody, GeneTex, catalogue no.: GTX89973-PEP, sequence: EEVRKARGKKRVP, species: human) or anti-TRPM3 antibody and immunising blocking peptide (10 times the molar concentration of the antibody, Bioss, catalogue no.: bs-9046P, sequence: KLFITDDELKKVH, species: human) based on an immunising peptide blocking protocol (https://www.abcam.com/protocols/blocking-with-immunizing-peptide-protocol-peptide-competition).

Intrathecal siRNA and AAV injection

siRNAs targeting mouse ATF4 (FITC-labelled, sense: GCUAGGCAGUGAAGUUGAU), KIF17 (sense: GCCACGCAUUAAUGAAGAC), CXCR4 (sense: GCUAACCCUUAUGCAAAGA), CXCR7 (sense: CCAUGCCUAACAAGAACGU) and a nontargeting siRNA (sense: CCTAAGGTTAAGTCGCCCTCG) were purchased from Thermo Fisher Scientific. AAV overexpression vectors for ATF4 (serotype 5, rAAV-CMV-Atf4-2A-EGFP-WPRE-PA), KIF17 (serotype 5, rAAV-CMV-Kif17-2A-EGFP-WPRE-PA) and a transcriptionally inactive form of ATF4 (serotype 5, rAAV-CMV-Atf4-ΔbZIP-2A-EGFP-WPRE-PA) were purchased from BrainVTA Biotechnology. The transcriptionally inactive form of ATF4 (rAAV-CMV-Atf4-ΔbZIP-2A-EGFP-WPRE-PA) was generated by site-directed mutagenesis with 6 amino acid substitutions within the DNA-binding domain (²⁹²RYRQKKR²⁹⁸ to ²⁹²GYLEAAA²⁹⁸)^27,28,29. A mixture of 10 μg siRNA and 7.5 μg transfection reagent (chimeric rabies virus glycoprotein fragment, RVG-9R, Anaspec, USA)^62,63 in 10 μl 5% dextrose in water (D5W) was injected intrathecally 48 h prior to the experiments. rAAV-CMV-Atf4-2A-EGFP-WPRE-PA (10 μl), rAAV-CMV-Kif17-2A-EGFP-WPRE-PA (10 μl) and rAAV-CMV-Atf4-ΔbZIP-2A-EGFP-WPRE-PA (10 μl) were injected intrathecally in mice and the experiments were performed after 21 days.

Data analysis

The grey values of the western bands were quantified by Tanon Gis software. The grey value of the images and colocalization were quantified by Image-Pro Plus 6.0 software. All western blotting, immunostaining, electrophysiological and behavioural data are expressed as the mean ± SEM and were analysed with GraphPad Prism 6. The threshold for statistical significance was P < 0.05.

Reproducibility

Experiments were repeated independently with similar results at least three times. Micrographic images presented in figures are representative ones from experiments repeated independently: Fig. 1a (three times), Fig. 1c, d (four times), Fig. 5f (three times), Fig. 7g (three times), Fig. 8a–c (three times), Supplementary Fig. 1b, d, h (three times), Supplementary Fig. 1f (four times), Supplementary Fig. 2,a, b (four times), Supplementary Fig. 5l, m (three times), Supplementary Fig. 7b (three times).

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.