LncRNA Anxa10-203 enhances Mc1r mRNA stability to promote neuropathic pain by recruiting DHX30 in the trigeminal ganglion

The animal study was approved by the Ethics Committee of West China Hospital of Stomatology Sichuan University (Number: WCHSIRB-D-2022-021) and complies with the ARRIVE 2.0 guidelines. A total of 273 wild-type male C57BL/6 mice and 89 female C57BL/6 mice (6-8 weeks, 20-25g) were used for this study. Every single animal was defined as an experimental unit. All animals were kept in a specific pathogen-free experimental animal center (West China Hospital Experimental Animal Center) on a 12:12 light-dark cycle at 22 ± 1 °C. Food and water were available at liberty. In all assays, the animals were randomly allocated to the experimental groups using a random number table. The experimenters were blinded for the treatment groups during the analysis of the data in each assay.

Bilateral TG was dissociated from connective tissue and minced. Then the tissue was enzymatically digested with 8 U/ml papain and incubated at 37°C for 20 minutes. An equal volume of complete medium was added to stop digestion and gently blow the digested tissue fluid several times with a sterile Pasteur pipette to mechanically dissociate neurons. Cell precipitation was obtained by centrifugation and resuspended in the neurobasal medium (Thermo Fischer, USA) containing 2% B27 (Invitrogen, USA) supplements and 1% GlutaMAX (Invitrogen, USA), and then plated in a twelve-well plate coated by poly-D-lysine (Invitrogen, USA) in 37℃ in humidified incubators with 5% CO2 and 95% air. The LV-Anxa10-203 (MOI = 5) and DHX30 siRNA or negative control siRNA (10 nM) were transfected into TG neurons (TGNs) on the second day. The Anxa10-203 oe+siDHX30 was the experimental group, and the Anxa10-203 oe+siNC (negative control siRNA) served as the control group. The LV-Anxa10-203 was purchased from Sangon (Shanghai, China). The siRNA and siNC were synthesized by GenePharma (Shanghai, China) and were transfected with HieffTrans in vitro siRNA/miRNA Transfection Reagent (Yeasen, China) following the manufacturer’s protocol. The sequences of siRNA are listed in Table 1.

Mice were anesthetized with isoflurane. For the CCI-ION group, two chronic gut ligatures (5-0) were placed loosely around the infraorbital nerve (ION) [26]. For the Sham group, the same surgical procedures were operated without ligation of the ION. The incision was sutured with 5-0 silk. The mice were placed on a heat recovery pad until they awoke from the anesthesia. Animals were monitored at least once daily after surgery.

The TG microinjection of mice was performed according to our previous study [27]. Briefly, after anesthesia, the siRNA Anxa10-203 (0.5 ug/μl), LV-Anxa10-203 (MOI = 5) or AAV2/9-CMV-GFP-shDHX30 (2E+9 vg/μl) were injected into the TG (coordinates: AP + 1 mm, ML + 4 mm, and DV -6.5 mm relative to bregma) under the guidance of a stereotaxic apparatus, and the maximum injection volume was 2 μl. A negative control siRNA or AAV2/9-CMV-GFP-shNC was used as control. The AAVs were synthesized by Xuanzun Bioscience (Chongqing, China). The siRNA was synthesized by GenePharma (Shanghai, China) and was transfected with Entranster^TM-in vivo (Engreen Biosystem, China) following the manufacturer’s protocol.

The Von Frey test was used to measure the 50% head withdrawal threshold (HWT) in response to mechanical stimuli according to our previous studies [27]. In brief, a series of Von Frey filaments were applied to the whisker pad area providing a force ranging from 0.04 g to 4 g. A positive response was recognized as sharp head withdrawal. The test was performed blinded.

RNAiso regent (Takara, Japan) was used to obtain total RNA. cDNA was synthesized using PrimeScript™ RT reagent Kit (Takara, Japan). The RT-qPCR was performed by QuantStudio 7 Flex System (Applied Biosystems, USA). The primer sequences used are listed in Table 2. The relative expression of the target gene was normalized to GAPDH and calculated with the 2^-ΔΔCT method.

The cytoplasmic and nuclear fractions of the TG were collected using a Nuclear and Cytoplasmic Extraction Reagents Kit (Beyotime, China) following the manufacturer’s protocol. RNA from cytoplasmic and nuclear fractions was extracted with RNAiso regent and analyzed by RT-qPCR. U6 and Gapdh were defined as nuclear and cytoplasmic control respectively.

The frozen sections were washed and permeabilized. The slices were blocked with 10% goat serum and then incubated with primary antibody overnight at 4℃. After washing with PBS, the sections were then incubated with a secondary antibody for 1 h. The images were captured with a laser scanning confocal microscope (LSCM, Olympus FV3000, Japan). The antibodies used are listed in Table 3.

The samples were homogenized in a lysis buffer with protease inhibitors. The proteins were separated on 10% SDS/PAGE gel and then transferred to PVDF membranes. The membranes were blocked by 10% non-fat milk and incubated with the primary antibodies at 4℃ overnight, and further incubated with a secondary antibody. The antibodies used are listed in Table 3. Quantification of immunoblots was measured by ImageJ. The relative expression level of the target protein in all groups were normalized to α-Tubulin.

The RNAs of the Anxa10-203, Mc1r, and four Anxa10-203 sequence segments were in vitro transcribed (Vazyme, China) and biotinylated (Thermo Scientific, USA). The Pierce magnetic RNA-protein pull-down kit (Thermo Scientific, USA) was used. Briefly, the biotinylated RNA was bonded to streptavidin magnetic beads (Thermo Scientific, USA). The TG lysates from CCI-ION mice were added into RNA-conjugated beads and incubated overnight at 4℃ with agitation. The RNA-interacting proteins were eluted and then analyzed by liquid chromatography-mass spectrometry (LCMS) and/or Western Blot. The input of the antisense RNA was defined as positive control and negative control separately.

Magna RIP Kit (Millipore, USA) was used to conduct the RNA-binding protein immunoprecipitation assay according to the manufacturer’s instructions. Briefly, the TG of CCI-ION mice was lysed by RIP lysis buffer. The magnetic beads conjugated with anti-DHX30 antibody were incubated with the tissue lysates on a rotator at 4℃ overnight. The magnetic beads conjugated with anti-IgG served as a negative control. The RNA was eluted from the beads, and the expression of Anxa10-203 and Mc1r was analyzed by RT-qPCR.

Click-iT Plus OPP Protein Synthesis Assay Kit (Invitrogen, USA) was applied. The Anxa10-203 oe+siDHX30 was the experimental group, and the Anxa10-203 oe+siNC served as the control group. The OPP working solution was added to the culture medium and incubated for 30 minutes. Then the cells were fixed and permeabilized. After washing, the OPP reaction cocktail was added and incubated for 30 minutes at room temperature. Images were acquired by LSCM (Olympus FV3000, Japan) and the relative fluorescence unit was assessed by a microplate reader (SpectraMax iD3, Molecular Devices, USA).

The electrophysiology was recorded by an Axopatch-700B amplifier and a Digidata 1440 digitizer (Molecular Devices, USA). The MC1R agonist BMS-470539 (BMS, 100 nM, MedChemExpress, China) was used, and the extracellular solution without BMS was defined as control [28]. The extracellular solution consisted of (in mM) NaCl 150, KCI 5, CaCl₂ 2.5, MgCl₂ 2, HEPES 10, and D-glucose 10 (pH = 7.4 by NaOH, 330 mOsm/L). The intracellular pipette solution contained (in mM) KCl 140, CaCl₂ 1, MgCl₂ 2.5, HEPES 10 and D-glucose 10, EGTA 11, Mg-ATP 5 (pH = 7.2 by NaOH, 310 mOsm/L). The resting membrane potential (RMP) was recorded for 3 min. The currents from -120 pA to 170 pA with an increment of 10 pA were injected into TGNs to evoke action potential (AP) [29]. The rheobase current was defined as the minimum current required to evoke the first AP. The AP threshold was obtained at dV/dt= 10 mV/ms. The action potential-related parameters including AP latency, the AP amplitude, AP half-width, afterhyperpolarization (AHP) time, and AHP amplitude are measured.

In our previous RNA-seq study [12], the top 10 up-regulated and 10 down-regulated DE lncRNAs were predicted (adjusted P value < 0.05, top 10 absolute log2-fold change in DE lncRNAs). To evaluate these DE lncRNAs in the TG of mice with orofacial pain, we first established the CCI-ION mice model. Von Frey test was applied to assess the mechanical pain threshold, and the results showed that CCI-ION induced significantly decreased 50% HWT on 3, 7, and 14 days (Fig. 1a). The expression of the top 10 up-regulated and 10 down-regulated DE lncRNAs was verified by RT-qPCR (data of Cdh4-204, Dleu2-212, Ninl-203, Gm28403-201 were abandoned for lack of specific primers), in which, TCONS_00081846 was significantly down-regulated, while Anxa10-203, A630023A22Rik-202, and TCONS_00032016 were significantly up-regulated (Fig. 1b-c). Considering the most striking fold changes of Anxa10-203, we selected Anxa10-203 for the following research. Then the expression of Anxa10-203 at different time points after CCI-ION was assessed. It was revealed that Anxa10-203 increased significantly on days 3, 7, and 14, and reached the peak on day 7 in the CCI-ION group (Fig. 1d). We further investigated whether TG microinjection of Anxa10-203 siRNA affected the development of CCI-ION-induced mechanical nociceptions. The behavioral tests demonstrated that the 50% HWT of mice increased after the siAnxa10-203 injection, and a significant difference was observed on the fourth day (Fig. 1e). The expression of Anxa10-203 in the TG of CCI-ION mice was significantly decreased four days after siAnxa10-203 injection (Fig. 1f). The data above indicated that CCI-ION induced increased Anxa10-203 expression in the TG, and Anxa10-203 knockdown relieved the pain induced by CCI-ION.

The function of lncRNAs usually depends on their subcellular distribution, thus we further explored Anxa10-203 distribution in the TG [30]. To define the subcellular localization of Anxa10-203, the online algorithm LncLorator was used to predict first. The results showed that the score of cytoplasm is the highest, which is about 0.87 (Table 4). Consistent with the predicted results, the subcellular fraction assay further verified that Anxa10-203 was mainly expressed in the cytoplasm (Fig. 2a). Besides, RNA-FISH also showed that Anxa10-203 was mainly distributed in the cytoplasm of the cultured TGNs in vitro (Fig. 2b). In vivo, Anxa10-203 was co-located with MAP2 (a marker for neurons) but not with glutamine synthetase (GS, a specific marker for satellite glial cells) in the TG (Fig. 2c). Furthermore, The localizations of Anxa10-203 in non-peptidergic (labeled by the specific marker IB4) and peptidergic nociceptive neurons (labeled by the specific marker CGRP) were also observed (Fig. 2c). According to the data above, it was suggested that Anxa10-203 was mainly distributed in the cytoplasm of TGNs, which indicated that Anxa10-203 may play a role in post-transcriptional regulation.

LncRNAs-RBPs interaction is an important regulatory pathway of lncRNAs in post-transcriptional regulation, thus we focused on the LncRNA-protein interaction [13, 31]. To identify the binding proteins of Anxa10-203, RNA pull-down-LCMS and RNA-protein interaction online algorithms (CatRAPID and RBPsuite) were used. Totally, 3041, 1093, and 154 proteins were predicted by RNA pull-down/LC-MS, CatRAPID, and RBPsuite respectively, of which 102 proteins were overlapped (Fig. 3a). These overlapped proteins were notably enriched in RNA processing and RNA binding (Fig. 3b). Among the overlapped proteins, DHX30 with the highest score of catRAPID interaction propensity appeared to be the most likely binding protein of Anxa10-203. The co-localization of DHX30 protein and Anxa10-203 RNA in the cytoplasm of TGNs was demonstrated, indicating that there was spatial proximity (Fig. 3c). In the RNA pull-down assay, DHX30 protein was significantly pulled down by Anxa10-203 RNA compared with the antisense RNA (Fig. 3d). The interaction between Anxa10-203 and DHX30 in the TG of CCI-ION mice was further verified by RIP (Fig. 3e). RBPsuite was applied to predict the binding sites between DHX30 and Anxa10-203, and 5 binding sites were predicted and scored (Fig. 3f). Subsequently, serial sequence deletion of Anxa10-203 showed that the exon 8 of Anxa10-203 deletion decreased the binding of DHX30 protein (Fig. 3g), indicating that the exon 8 of Anxa10-203 was the main binding region of DHX30. The data shown above demonstrated the interaction between Anxa10-203 and DHX30 protein in the TG.

To predict the downstream mRNA of the Anxa10-203/DHX30 complex, CatRAPID was applied. 499 RNAs were predicted to interact with DHX30, in which 25 RNAs were overlapped in pain-related genes (Fig. 4a). RIP was used to verify the interaction between DHX30 and the overlapped RNAs, and the result showed that Mc1r mRNA was significantly enriched in DHX30-RNAs immunoprecipitations, suggesting the DHX30/Mc1r interaction in the TG (Fig. 4b). Further, FISH and IF assays verified that Mc1r mRNA and DHX30 protein were co-located in the cytoplasm of TGNs (Fig. 4c). Subsequent RNA pull-down demonstrated that Mc1r mRNA interacted with DHX30 as well (Fig. 4d).

To demonstrate the expression pattern of MC1R, the expression of Mc1r mRNA at different time points was detected after CCI-ION. The results showed that Mc1r increased significantly at 3, 7, and 14 days after CCI-ION, and reached the peak on day 7 (Fig. 5a). Subsequently, Pearson correlation analysis verified that Mc1r mRNA was positively correlated with Anxa10-203 expression (R² = 0.7023) and the result was statistically significant (P < 0.0001) (Fig. 5b). To further confirm the association between Anxa10-203 and MC1R, we knocked down Anxa10-203 and assessed MC1R expression. The results revealed the mRNA and protein of MC1R increased in the TG 7 days after CCI-ION (Fig. 5c-e). CCI-ION-induced increases of MC1R mRNA and protein were markedly reversed by siAnxa10-203 microinjection (Fig. 5c-e). We then explored whether the up-regulation of Anxa10-203 in the TG is sufficient to induce up-regulated MC1R. The significant increase of Anxa10-203 in the TG of naive mice was verified 7 days after LV-Anxa10-203 microinjection (Fig. 5f) and the expression levels of MC1R mRNA and protein in the TG were correspondingly increased (Fig. 5g-i). Besides, to explore whether the association between Anxa10-203 and MC1R was affected by sex difference, we also performed relevant studies on female mice. The RT-qPCR data in supplementary Fig. 1 ( see Additional file 1) indicated the positive correlation between Anxa10-203 and Mc1r mRNA in female mice, similar to that in male mice. To explore the effect of MC1R activation on neuron excitability, we also performed electrophysiology in supplementary Fig. 2 (see Additional file 1), which indicated that MC1R activation promoted the intrinsic excitability of TGNs, resulting in the development of orofacial pain.

To explore the function of DHX30 recruited by Anxa10-203 in Anxa10-203/DHX30/Mc1r complex. We first knocked down DHX30 in the TG and verified the down-regulated DHX30 mRNA and protein in Fig. 6a-b. Then MC1R expression was assessed, and the results showed that compared with the Sham+AAV-shNC group, the MC1R mRNA and protein remained unchanged in the Sham+AAV-shDHX30 group, while DHX30 knockdown in the CCI-ION mice led to a decrease in MC1R expression (Fig. 6c-d). Therefore, DHX30 knockdown induced the decrease of MC1R expression under the condition of Anxa10-203 over-expression. To explore how DHX30 knockdown with Anxa10-203 over-expression induced the decrease in MC1R expression, we over-expressed Anxa10-203 and knocked down DHX30 in the TGNs to simulate the condition in vivo (Fig. 6e-f). Consistent with the results in vivo, DHX30 silencing caused a decrease in MC1R mRNA and protein expression (Fig. 6g-h). According to the cytoplasmic localization of DHX30 and Gene Ontology analysis in Fig. 3b, we speculated that DHX30 played a role in post-transcriptional processes. The effects of DHX30 on mRNA translation were evaluated by a nascent protein synthesis assay. Fluorescence images showed that nascent proteins (green dots) were in the cytoplasm of TGNs transfected with siNC or siDHX30 under the condition of Anxa10-203 over-expression (Fig. 6i). Then the green fluorescence intensity was quantitatively analyzed, and no significant difference was observed in the two groups, which indicated that DHX30 recruited by Anxa10-203 did not affect mRNA translation (Fig. 6j). At the transcription level, we blocked nascent mRNA synthesis using ActD, and the decay of Mc1r mRNA was detected by RT-qPCR and analyzed by nonlinear regression analysis. A lower level of Mc1r in the Anxa10-203 oe + siDHX30 group was observed at all time points compared with the Anxa10-203 oe + siNC group, and the half-life of Mc1r mRNA in the Anxa10-203 oe + siDHX30 group is shorter (t_1/2 (Anxa10-203 oe + siNC) = 0.49, R² =0.53; t_1/2 (Anxa10-203 oe + siDHX30) = 0.35, R² =0.89, Fig. 6k).