A novel co-target of ACY1 governing plasma membrane translocation of SphK1 contributes to inflammatory and neuropathic pain

Summary Previous studies validate that inhibiting sodium channel 1.8 (Nav1.8) effectively relieves inflammatory and neuropathic pain. However, Nav1.8 blockers have cardiac side effects in addition to analgesic effects. Here, we constructed a spinal differential protein expression profile using Nav1.8 knockout mice to screen common downstream proteins of Nav1.8 in inflammatory and neuropathic pain. We found that aminoacylase 1 (ACY1) expression was increased in wild-type mice compared to Nav1.8 knockout mice in both pain models. Moreover, spinal ACY1 overexpression induced mechanical allodynia in naive mice, while ACY1 suppression alleviated inflammatory and neuropathic pain. Further, ACY1 could interact with sphingosine kinase 1 and promote its membrane translocation, resulting in sphingosine-1-phosphate upregulation and the activation of glutamatergic neurons and astrocytes. In conclusion, ACY1 acts as a common downstream effector protein of Nav1.8 in inflammatory and neuropathic pain and could be a new and precise therapeutic target for chronic pain.


INTRODUCTION
According to the latest epidemiological data, chronic pain severely affects the quality of life of > 30% of the world's population. 1 However, treatment of chronic pain remains a challenge. It is well known that inflammatory pain and neuropathic pain are the two common types of chronic pain. Despite the differences in etiology and associated mechanisms between inflammatory and neuropathic pain, each pain state shares common mechanisms. 2,3 To date, pharmacological strategies for the treatment of these two types of chronic pain are currently diverse. However, there is still a lack of drugs that can effectively treat both inflammatory pain and neuropathic pain with few adverse reactions. Recently, emerging evidence suggests common pathways, such as common genetic factors, activation of glial cell, and enhanced glutamate release in inflammatory and neuropathic pain. 2,4 However, the shared mechanism underlying inflammatory and neuropathic pain remains largely unknown. Therefore, exploring the common target mechanism of inflammatory and neuropathic pain is of great clinical significance for the treatment of chronic pain.
Our proteomics results showed that 5655 and 5265 proteins were detected in the CFA-induced inflammatory pain models and SNI-induced neuropathic pain models, respectively ( Figure 1C). The 55 and 58 proteins in inflammatory and neuropathic pain, respectively, showed significantly differential expression profiles in each pain paradigm compared with Nav1.8 knockout mice, and the abundances of 31 and 24 proteins in the spinal cord were clearly elevated in wild-type mice after CFA injection and SNI surgery, respectively. The abundances of 23 and 35 proteins in the spinal cord were dramatically downregulated in wild-type mice after CFA injection and SNI surgery, respectively. Thus, < 1% of all quantified proteins presented remarkable upregulation and downregulation in each pain model, indicating high specificity of the pain model. Therefore, these differentially expressed proteins may be the downstream effector molecules of Nav1.8 in the different pain models when Nav1.8 gene knockout mice were used as controls in the proteomic analysis. Raw proteomics data were shown in Data S1.
Expression of ACY1 in the spinal cord was upregulated in both chronic pain models As shown in Figure 1C, the abundance of ACY1 was significantly altered in both pain models among all differentially expressed proteins, according to our proteomic results. We also confirmed the expression level of ACY1 in the original protein lysis buffer by Western blot analysis. As shown in Figure 2A, the abundance of ACY1 in the spinal cord was significantly upregulated in both CFA and SNI-induced chronic pain models in wild-type mice compared with Nav1.8 knockout mice. The verified results are consistent with the trend of the proteomics results, indicating that the proteomics results have high credibility. To clarify the role of ACY1 in chronic pain, we reconstructed two chronic pain models in wild-type mice and detected the expression of ACY1 at different time points. As shown in Figure 2B, we observed a significant upregulation of ACY1 in the spinal cord from day 5-14 after CFA injection and SNI surgery. These results indicate that ACY1 may be an important downstream candidate of Nav1.8 in chronic inflammatory and neuropathic pain. First, we examined the cellular localization of ACY1 in the spinal dorsal horn. We conducted double immunofluorescence of ACY1 with the neuronal marker NeuN, astrocyte marker GFAP, and microglial marker IBA1 to determine the localization of ACY1 in the spinal dorsal horn. As shown in Figure 2C, ACY1 was mainly colocalized with neuronal marker NeuN and sparsely colocalized with GFAP or IBA1 in the dorsal horn of the spinal cord ( Figure S1). We then performed double immunofluorescence staining of ACY1 with Nav1.8. As shown in Figure 2D, ACY1 colocalized with Nav1.8, suggesting that ACY1 was expressed in Nav1.8-positive cells.

Overexpression of spinal ACY1-induced mechanical allodynia and neuronal hyperexcitability in naive mice
To assess the role of ACY1 in pain modulation, we constructed AAV9-Acy1 to specifically upregulate the expression of ACY1 and detect the mechanical thresholds of naive mice following intradorsal horn injection of AAV9-Acy1 versus the vehicle. The recombinant AAV9-Acy1 was fused to enhanced green fluorescent protein (EGFP) C-terminus to provide a fluorescent tag to identify neurons expressing ACY1. As shown in Figure 3B,  iScience Article demonstrated successful transduction of AAV9-Acy1. Western blot and immunofluorescence analyses were performed 21 days after injection to validate successful transduction of the spinal cord. As shown in Figure 3C, EGFP provides a fluorescent tag for identifying neurons expressing the ACY1 peptide aptamer (hereafter referred to as EGFP-ACY1). In addition, results of Western blot analysis showed that the protein expression of ACY1 in the spinal cord was markedly increased 21 days after AAV9-Acy1 injection, suggesting successful virus transfection ( Figure 3E). As shown in Figure 3D, the baseline mechanical thresholds were not statistically different between the vehicle and AAV9-Acy1 groups. However, the mechanical thresholds significantly declined from 14 to 21 days after AAV9-Acy1 injection. These data indicate that overexpression of spinal ACY1 is sufficient to contribute to mechanical allodynia in normal animals.
To determine the role of ACY1 in regulating the excitability of neurons in mice spinal cord, we tested the activity of spinal neurons in lamina II 21 days after treatment with an AAV9-Acy1 intradorsal horn injection using whole-cell patch-clamp recordings. As shown in Figures 3F and 3G, rheobase was significantly decreased in mice with spinal ACY1 overexpression compared with naive mice and CON323 mice, suggesting that neurons in the spinal cord more easily initiated an action potential after the AAV9-Acy1 injection. Additionally, as shown in Figures 3H and 3I, spinal ACY1 overexpression significantly increased the number of APs. Notably, rheobase and the AP frequency did not differ significantly between naive mice and CON323 mice, indicating that adeno-associated viruse (AAV) did not affect the excitability of spinal neurons. These data indicate that the overexpression of spinal ACY1 increases the activity of spinal dorsal horn neurons.

Suppressing spinal ACY1 alleviated CFA-induced inflammatory and spinal hyperexcitability
For ACY1 knockdown, we expressed AAV-Acy1-RNAi in the spinal cord via intradorsal horn microinjection. Behavioral data showed that the baseline mechanical thresholds showed no statistical differences between the groups. The mechanical thresholds significantly decreased after CFA injection. Additionally, reducing the expression of ACY1 using AAV9-Acy1-RNAi significantly alleviated CFA-induced mechanical allodynia ( Figures 4A and 4B). Western blot analysis results revealed that the protein expression of ACY1 was significantly increased after CFA injection. Furthermore, the expression of ACY1 in the spinal cord of mice pre-injected with AAV9-Acy1-RNAi was reduced after CFA injection ( Figure 4C). Moreover, as shown in Figures 4D and 4E, the rheobase of spinal cord neurons was signally decreased in CFA mice in contrast with control mice. However, spinal ACY1 knockdown significantly increased rheobase levels after CFA injection. In addition, as shown in Figures 4F and 4G, the number of APs in AAV9-Acy1-RNAi injected mice with CFA was significantly lower than that in CFA mice and CON533+ CFA mice. Notably, the rheobase and AP frequency did not differ significantly between CFA mice and CON533+ CFA mice, indicating that AAV did not affect the excitability of spinal neurons. These data indicate that spinal ACY1 knockdown could decrease the activity of spinal dorsal horn neurons and alleviate inflammatory mechanical hyperalgesia.

Suppressing spinal ACY1 alleviated SNI-induced pain behavior and spinal hyperexcitability
As shown in Figures 5A and 5B, mechanical thresholds were decreased after SNI surgery, and AAV9-Acy1-RNAi pretreatment significantly alleviated SNI-induced mechanical allodynia. Western blot analysis results iScience Article exhibited that the expression of ACY1 was significantly increased after SNI surgery. Moreover, the SNI surgery-induced upregulation of ACY1 in mice pre-injected with AAV9-Acy1-RNAi in the spinal cord was significantly suppressed ( Figure 5C). Furthermore, as shown in Figures 5D and 5E, the rheobase of SNI mice with spinal ACY1 knockdown was increased compared with SNI mice without AAV injection. In addition, as shown in Figures 5F and 5G, the number of APs in spinal neurons significantly decreased in SNI mice iScience Article with AAV9-Acy1-RNAi injection. Notably, the rheobase and AP frequency did not differ between SNI mice and CON533+ SNI mice, indicating that AAV did not affect the excitability of spinal neurons. These data indicated that spinal ACY1 knockdown could decrease the activity of neurons in the spinal dorsal horn to alleviate neuropathic mechanical hyperalgesia.

ACY1 interacted with SphK1 and promoted its membrane translocation in both chronic pain models
Our results indicated that ACY1 plays an important role in the development of chronic inflammatory and neuropathic pain. However, the underlying mechanisms remain unclear. Therefore, we explored the involvement of spinal ACY1 in chronic pain. Recently, sphingosine kinase 1 (SphK1) and its product, sphingosine-1-phosphate (S1P), were found to be involved in pain modulation according to a growing body of evidence. SphK1 translocates from the cytoplasm to the membrane after activation to catalyze formation of the important bioactive sphingolipid mediator S1P. 27 ACY1 can interact with and activate SphK1 and increase S1P secretion. 28 We hypothesized that spinal ACY1 interacts with SphK1 and promotes its plasma membrane translocation, resulting in the upregulation of S1P and the development of chronic pain. To determine the cellular localization of ACY1 and SphK1 in the spinal dorsal horn, we first conducted immunofluorescence of ACY1 and SphK1 in spinal cord sections of mice that were pre-transfected with AAV9-Acy1. As previously mentioned, recombinant AAV9-Acy1 was fused to the C-terminus of EGFP, which provided a fluorescent tag for identifying AAV-targeted neurons expressing the ACY1 peptide aptamer. As shown in Figure 6A, EGFP-ACY1 was highly colocalized with SphK1 in the spinal dorsal horn. Next, we examined the relationship between ACY1 and SphK1 using co-immunoprecipitation. Co-immunoprecipitation results showed that there was an interactive relationship between ACY1 and SphK1 in the spinal cord of mice ( Figure 6B).
Next, we performed Western blot analysis on the subcellular fraction to confirm the subcellular expression of SphK1. As shown in Figure 7A, the expression of SphK1 was significantly increased in the plasma membrane on day 21 after AAV9-Acy1 spinal injection in naive mice, suggesting that ACY1 promotes the translocation of SphK1 from the cytoplasm to the plasma membrane. As shown in Figures 7B and 7C, the expression of SphK1 was upregulated in the plasma membrane on day 14 after CFA injection and SNI surgery. This finding suggests that the plasma membrane translocation of SphK1 was increased in mice with chronic pain. Next, we examined whether plasma membrane translocation of SphK1 in mice with chronic pain could be regulated by ACY1. As expected, suppression of spinal ACY1 by pretreatment with AAV9-Acy1-RNAi significantly decreased membrane translocation of SphK1 in both chronic pain models.

ACY1 promoted S1P-induced glutamatergic neurons and astrocyte activation in the CFA and SNI models
Altered sphingolipid metabolism is linked to nociceptive processing in an increasing number of studies.
Previous studies have validated that S1P increases glutamate release through autocrine action. 29, 30 Glutamate, the most abundant excitatory neurotransmitter, plays a key role in central sensitization and chronic pain. 31 In addition, recent animal studies from Chen et al. indicated that increased S1P in the spinal cord selectively activates astrocytes via S1PR1, resulting in the development of neuropathic pain. 32 Accordingly, we speculated that ACY1 may interact with SphK1 to drive chronic pain through S1P-induced glutamatergic neuron and astrocyte activation. Therefore, we first examined the expression level of S1P, S1PR1, the glutamatergic neuronal marker VGLUT2, and the astrocyte marker GFAP. The ELISA results showed that S1P expression was increased in the spinal cord of CFA-and SNI-induced chronic pain mice ( Figures 8A-8G, and 8B-8G). Western blot results showed that S1PR1, VGLUT2, and GFAP ( Figures 8A and 8B) also significantly increased in both chronic pain mice models. Next, we performed correlation analysis of S1PR1 between VGLUT2 and GFAP. The results showed that expression levels of VGLUT2 and GFAP were significantly determined by S1PR1 in the cytomembrane when CFA and SNI models were established.
Next, to further investigate whether ACY1 is involved in S1P-induced glutamatergic neuron and astrocyte activation, we intervened with spinal ACY1 using an intradorsal horn injection of AAV9-Acy1 and iScience Article AAV9-Acy1-RNAi. As shown in Figure 9A, the ELISA results showed that the levels of S1P were significantly increased in the spinal cord of wild-type mice with ACY1 overexpression. Additionally, the expression of S1P in CFA and SNI mice were upregulated. Suppressing spinal ACY1 by pretreating with AAV9-Acy1-RNAi could reverse the upregulation of S1P in CFA and SNI mice. These results indicated that ACY1 mediates the release of S1P in both CFA and SNI pain models. As shown in Figure 9B, the Western blot results showed that the expression of S1PR1, VGLUT2, and GFAP were upregulated in the spinal cord of wild-type mice with ACY1 overexpression. Additionally, the knockdown of spinal ACY1 by AAV9-Acy1-RNAi pretreatment reversed the upregulation of S1PR1, VGLUT2, and GFAP in CFA and SNI mice ( Figures 9C and 9D), indicating that ACY1 mediates the activation of S1P in both CFA and SNI pain models. Together, these results iScience Article suggest that ACY1 interacts with SphK1 to promote the production of S1P, resulting in glutamatergic neuron and astrocyte activation in both the CFA and SNI models.

DISCUSSION
In this study, Nav1.8 knockout mice served as the control to screen for a differential protein expression profile in the spinal cord of wild-type chronic pain mice. In the preliminary models, it seemed that Nav1.8 KO had a more prominent effect on the neuropathic model. However, the paw withdrawal threshold (PWT) of the CFA group declined during the first day, and the drop of PWT was even greater in the CFA group. Thus, pain sensitization occurs earlier and present more severe hypersensitivity for wild-type mice in the CFA iScience Article model compared to the SNI model. Moreover, significant differences exist with the methods for constructing the SNI and CFA models. Whether Nav1.8 may be more effective for neuropathic treatment needs further investigation. Through proteomic analysis, we found that < 1% of all quantified proteins were significantly upregulated and downregulated in each pain model, indicating the high specificity of these differentially expressed proteins downstream of Nav1.8. Among these differentially expressed proteins, the abundance of ACY1 was significantly increased in both inflammatory and neuropathic pain mice when compared with Nav1.8 knockout mice with chronic pain. Then, we examined the protein levels of ACY1 in the original protein lysis buffer using Western blot analysis, and the validation results were consistent with the proteomic results.
ACY1 is the most widely expressed aminoacylase that catalyzes amino acid deacylation in protein degradation. 33 Reportedly, ACY1 is involved in various human diseases. 34-36 ACY1 deficiency due to mutations in   iScience Article studies have also demonstrated that the overexpression of ACY1 improves cell growth, invasion, migration, and tumorigenesis. 36,40 Hence, we speculated that ACY1 might play a role in promoting neuronal excitability in the nervous system. However, the effects of ACY1 on neuronal excitability and pain modulation have not yet been described. Thus, considering its expression in neurons and its role in neurological diseases, we chose ACY1 as a potential candidate to perform further study. To verify the involvement of ACY1 in chronic pain, we established two chronic pain models using wild-type mice and examined the expression patterns of ACY1. Expression of spinal ACY1 was significantly upregulated from day 5-14 after CFA injection and SNI surgery, suggesting that ACY1 may be implicated in inflammatory and neuropathic pain signaling. In a previous study, ACY1 is abundantly expressed in the kidneys and central nervous system. 33 Through cellular localization of spinal ACY1, we found that Nav1.8 and ACY1 were mainly colocalized with NeuN in the dorsal horn of the spinal cord. These results suggest that increased expression of ACY1 in the spinal cord is involved in the SNI and CFA procedures.
To further test the central role of ACY1 in pain modulation, we constructed AAV9-Acy1 and AAV9-Acy1-RNAi to specifically upregulate and knockdown the expression of ACY1, respectively. We found that overexpression of ACY1 by intradorsal horn injection of AAV9-Acy1 resulted in mechanical hyperalgesia in normal mice. In addition, the overexpression of ACY1 enhanced the excitability of spinal neurons. Furthermore, suppression of spinal ACY1 by AAV9-Acy1-RNAi decreased the activity of spinal dorsal horn neurons and reversed mechanical allodynia in mice with inflammatory and neuropathic pain. Taken together, these results experimentally validate the critical role of ACY1 in chronic pain. However, the precise mechanisms by which ACY1 is involved in the development of chronic pain remain unclear.
In recent decades, alterations in sphingolipid metabolism have been associated with numerous neurological diseases. 41 Sphingolipids are a critical class of bioactive lipids that are abundantly expressed in the central nervous system and comprise ceramides and their metabolites, such as sphingosine and S1P. 42 SphK1, a key enzyme in maintaining sphingolipid metabolism homeostasis, is translocated from the cytoplasm to the membrane after activation to catalyze the formation of the important bioactive sphingolipid mediator S1P. 27 Indeed, ACY1 can interact with SphK1 and improve its membrane translocation. 28 More recently, preclinical and clinical studies have clearly shown that the dyshomeostasis of S1P contributes to the development of chronic pain. 43 Therefore, we speculated that ACY1 may participate in the development of chronic pain by affecting the SphK1-dependent spinal sphingolipid metabolism balance.
To test our hypothesis, we first examined the relationship between ACY1 and SphK1 in the spinal cord, and the immunofluorescence results showed that ACY1 was highly co-labeled with SphK1 in the spinal dorsal horn. Additionally, there was protein interaction between ACY1 and SphK1 in the spinal cord of mice. Further, our findings showed that plasma membrane SphK1 was significantly increased after ACY1 was overexpressed in the spinal cord of naive mice. Inhibition of ACY1 significantly decreased membrane translocation of SphK1 in both chronic pain models. Meanwhile, using ELISA, we found that the expression level of S1P was significantly upregulated after ACY1 overexpression. These findings suggest a facilitation effect of ACY1 on the translocation of SphK1, resulting in an increase in S1P in chronic pain.
In recent years, S1P signaling has been shown as a vital modulator of pain pathways according to accumulating evidence. 43,44 S1P also performs strong inflammatory and nociceptive actions. 44 S1P acts in an autocrine or paracrine manner after synthesis, and the cellular functions of S1P are mainly mediated through its receptors, comprising five G protein-coupled receptors (S1PR1-5). 43 In addition, S1P can trigger glutamate secretion and enhance neuronal excitability. 45,46 Glutamate, the most abundant excitatory neurotransmitter, plays an important role in central sensitization and chronic pain. 31 Our recent study and other studies have indicated that the activation of glutamatergic neurons plays a critical role in the development of chronic pain. [47][48][49] Our results also showed that overexpression of spinal ACY1 increased the excitability of mouse spinal neurons. It is noteworthy to mention that Chen et al. reported that increased S1P in the spinal cord selectively activated astrocytes via S1PR1, resulting in the development of neuropathic pain. 32 Glial cell activation plays an indispensable role in the maintenance of chronic pain. 50 Therefore, we hypothesized that ACY1-induced upregulation of S1P would trigger glutamate iScience Article secretion and enhance neuron excitability in an autocrine manner, as well as activate astrocytes in a paracrine manner.
To confirm this hypothesis, we detected the protein levels of the glutamatergic neuron marker VGLUT2 and astrocyte marker GFAP. As expected, the expressions of S1P, S1PR1, VGLUT2, and GFAP were significantly increased in both mouse models of chronic pain. In addition, we analyzed the correlation of S1PR1 between VGLUT2 and GFAP, indicating that S1PR1 had a significant effect on the expression levels of VGLUT2 and GFAP in both CFA and SNI models. Furthermore, overexpressing ACY1 upregulated the expression of S1P, S1PR1, VGLUT2, and GFAP in the spinal cord of naive mice, and suppressing spinal ACY1 by pretreating with AAV-Acy1-RNAi could reverse the expression of these proteins in both chronic pain mouse models.
As reported in a previous study, S1P-based drugs have been approved for numbers of clinical indications and can also be used for chronic pain treatment. 43 For example, proof-of-concept clinical trials of FTY720 in patients with breast cancer have been initiated to explore its effectiveness in preventing and/or treating chemotherapy-induced neuropathic pain. 43 However, the clinical application of S1P-based drugs in pain therapy has not yet made breakthrough progress, which may be attributed to the diversity of receptor types, drug specificity, and side effects. 51 Combined with our results that targeted ACY1 could directly inhibit the increase of S1P, which may have a more effective analgesic effect, suggesting that ACY1 is a potential analgesic target.

Limitations of the study
Several limitations should be considered when interpreting the current study. First, it is important to analyze the role of the dorsal root ganglion cells in pain processing because this is where Nav1.8 is majorly expressed. As a large number of studies and our previous study have explored the potential mechanism of Nav1.8 underlying chronic pain processing in the DRG, this study mainly focused on the downstream pathway of Nav1.8 in the spinal cord. 11,52 Therefore, we did not analyze the DRG in this study. Second, gender differences have a large impact on pain syndromes. One mechanism that has emerged is the influence of estrogen on the glutamate pathway, ultimately affecting neuronal physiology and behavior. 53 However, only male mice were used in our study. Therefore, the present findings need to be confirmed in female mice in further studies. Third, since we have found the protein of interest and focused on exploring the underlying mechanism of ACY1 in chronic pain, we did not perform further analysis and mining on the proteomics data in this study.
In conclusion, this study demonstrated that ACY1, a downstream effector protein of Nav1.8, regulates the plasma membrane translocation of SphK1 and promotes the production of S1P on the cell membrane. The upregulation of S1P triggers the activation of glutamatergic neurons and astrocytes, which in turn results in neuronal hyperexcitability and the development of chronic pain ( Figure 10). This study provides a novel cotarget that can be effective for both inflammatory and neuropathic pain.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

ACKNOWLEDGMENTS
We would like to thank Professor Stephen G. Waxman for kindly providing Nav1.8 knockout mice and Editage (www.editage.cn) for English language editing.  Complete Freund adjuvant (CFA)-and spared nerve injury (SNI)-induced the increased expression and activation of Nav1.8, which led to the upregulation of ACY1. ACY1 interacted with SphK1 and promoted its transportation to the plasma membrane, which resulted in the production of sphingosine-1-phosphate (S1P) on the cell membrane. Then, the upregulated S1P trigged the activation of glutamatergic neurons and astrocytes by autocrine and paracrine mode, respectively, which in turn resulted in neuronal hyperexcitability and the development of chronic pain. DRG, dorsal root ganglion; S1PR1, sphingosine-1phosphate receptor 1; Glu, glutamate; AMPA, alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid subtype receptor; NMDA, N-methyl-D-aspartate subtype receptor; mRNA, messenger RNA.

Establishment of the inflammatory pain model
The inflammatory pain model was induced using CFA according to a previously described method. 56 After anesthetizing the mice with isoflurane, their left hind paws were swabbed with iodophor and then injected with CFA (20 mL, F5881; Sigma, St. Louis, USA) or a saline solution subcutaneously using a 27-gauge needle.

Establishment of the neuropathic pain model
The neuropathic pain model was induced by spared nerve injury surgery, according to the method described previously. 57 The mice were anesthetized by intraperitoneal injection of pentobarbital sodium (50 mg/kg). Briefly, the left sciatic nerve trifurcates were exposed to blunt dissection. The common peroneal and tibial nerves were ligated using 6-0 silk suture. The distal adjacent nerve was transected at 2-4 mm. The sural nerve was protected from damage. The muscles and skin were closed using sutures. The same procedure was performed, excluding nerve ligation and transection, in the sham-operated group.

Behavioral tests
Mechanical allodynia was assayed by testing the mice's paw withdrawal threshold (PWT) in response to stimulation by von Frey filaments. 58,59 The mice were placed in isolated Plexiglas chambers on an elevated mesh floor and allowed to acclimatize for 30 min before testing. The PWT was assessed using the Dixon updown method. Briefly, the tips of the von Frey filaments (0.008 g, 0.02 g, 0.04 g, 0.16 g, 0.4 g, 0.6 g, 1.0 g, 1.4 g, and 2 g) were stuck perpendicularly to the mid-plantar surface of the hind paws of mice with gradually increasing force. Abrupt paw lifting or licking in response to the filaments was defined as a positive response. When a positive response occurred, the next descending filament was tested after resting for 5 min. When no positive response occurred, the next-ascending filament was tested. The lowest force-eliciting positive response was determined to be the PWT.

Proteomics
Isobaric tags for relative and absolute quantitation (iTRAQ)-based proteomics analysis (Genechem Co., Ltd., Shanghai, China) was performed to profile changes in the abundance of spinal cord proteins in two well-established animal pain models of chronic pain using wild-type and Nav1.8 knockout mice. Both the CFA-induced inflammatory pain and SNI-induced neuropathic pain models simulated the features of mechanical hypersensitivity in patients with chronic pain. Mice were assessed for nociceptive thresholds before and 1, 3, 5, 7, 10, and 14 days after the induction of inflammatory and neuropathic pain, and the mice were sacrificed for dissection of the lumbar spinal cord 14 days after the last behavioral test according to our preliminary results and the findings in previous studies. 11,60,61 Twelve samples were prepared for iTRAQ analysis, consisting of three biological replicates per sample. The samples included four conditions: CFA-induced inflammatory pain in wild-type mice versus CFA-induced inflammatory pain in Nav1.8 knockout mice, and SNI-induced neuropathic pain in wild-type mice versus SNI-induced neuropathic pain in Nav1.8 knockout mice. Only mice that experienced significant mechanical allodynia were included in our study. iScience Article QUANTIFICATION AND STATISTICAL ANALYSIS All data are expressed as mean G standard error of the mean. Behavioral data were analyzed using twoway repeated-measures analysis of variance (ANOVA) followed by the Bonferroni post hoc test. Western blot analysis and ELISA data were analyzed using one-way ANOVA followed by the Bonferroni post hoc test. Electrophysiology data were analyzed using one-way ANOVA with the Tukey multiple tests or twoway ANOVA with Bonferroni post hoc comparisons. Statistical analyses were performed using GraphPad Prism, version 7.0 (GraphPad Software, San Diego, CA, USA). Values with p < 0.05 were considered statistically significant.