Ferulic acid alleviates sciatica by inhibiting neuroinflammation and promoting nerve repair via the TLR4/NF‐κB pathway

Abstract Introduction Sciatica causes intense pain. No satisfactory therapeutic drugs exist to treat sciatica. This study aimed to probe the potential mechanism of ferulic acid in sciatica treatment. Methods Thirty‐two SD rats were randomly divided into 4 groups: sham operation, chronic constriction injury (CCI), mecobalamin, and ferulic acid. We conducted RNA sequencing, behavioral tests, ELISA, PCR, western blotting, and immunofluorescence analysis. TAK‐242 and JSH23 were administered to RSC96 and GMI‐R1 cells to explore whether ferulic acid can inhibit apoptosis and alleviate inflammation. Results RNA sequencing showed that TLR4/NF‐κB pathway is involved in the mechanism of sciatica. CCI induced cold and mechanical hyperalgesia; destroyed the sciatic nerve structure; increased IL‐1β, IL‐6, TNF‐α, IL‐8, and TGF‐β protein levels and IL‐1β, IL‐6, TNF‐α, TGF‐β, TLR4, and IBA‐1 mRNA levels; and decreased IL‐10 and INF‐γ protein levels and IL‐4 mRNA levels. Immunohistochemistry showed that IBA‐1, CD32, IL‐1β, iNOS, nNOS, COX2, and TLR4 expression was increased while S100β and Arg‐1 decreased. CCI increased TLR4, IBA‐1, IL‐1β, iNOS, Myd88, p‐NF‐κB, and p‐p38MAPK protein levels. Treatment with mecobalamin and ferulic acid reversed these trends. Lipopolysaccharide (LPS) induced RSC96 cell apoptosis by reducing Bcl‐2 and Bcl‐xl protein and mRNA levels and increasing Bax and Bad mRNA and IL‐1β, TLR4, Myd88, p‐NF‐κB, and p‐p38MAPK protein levels, while ferulic acid inhibited cell apoptosis by decreasing IL‐1β, TLR4, Myd88, p‐NF‐κB, and p‐p38MAPK levels and increasing Bcl‐2 and Bcl‐xl levels. In GMI‐R1 cells, Ferulic acid attenuated LPS‐induced M1 polarization by decreasing the M1 polarization markers IL‐1β, IL‐6, iNOS, and CD32 and increasing the M2 polarization markers CD206, IL‐4, IL‐10 and Arg‐1. After LPS treatment, IL‐1β, iNOS, TLR4, Myd88, p‐p38MAPK, and p‐NF‐κB levels were obviously increased, and Arg‐1 expression was reduced, while ferulic acid reversed these changes. Conclusion Ferulic acid can promote injured sciatic nerve repair by reducing neuronal cell apoptosis and inflammatory infiltration though the TLR4/NF‐κB pathway.


| INTRODUC TI ON
Sciatica, a common type of neuropathic pain attributed to impingement of the sciatic nerves or injury to the sciatic nerves, is experienced by up to 10% of patients with chronic lower back pain, with a reported lifetime incidence ranging from 10% to 40% 1 or even up to 70%. 2 According to the National Institute for Health & Clinical Excellence (NICE) guidelines, noninvasive treatments, including specific exercises, psychological therapy, and nonsteroidal anti-inflammatory drugs, can be used to treat sciatica. 3 Gabapentinoids, other antiepileptics, oral corticosteroids or benzodiazepines are not used to manage sciatica because there is no overall evidence of their benefit. 3 Although nonsteroidal antiinflammatory drugs can help relieve sciatica, it is necessary to take into account their potential gastrointestinal, liver and cardio-renal toxicity. 3 Moreover, opioids have a risk of addiction in patients. 4 Thus, it is urgent to seek new therapeutic drugs for the efficacious treatment of sciatica.
Nerve injury and neuroinflammation play vital roles in sciatica.
Following injury to the nerve, axonal breakdown is initiated, and the products of degenerated neural tissue stimulate microglia and resident macrophages to secrete chemokines and cytokines, which promote neuroinflammation and axonal breakdown. Activation of microglia leads to the progression of neuropathic pain by interfering with neuronal function. 5 Inhibiting microglial activation reduces hyperalgesia after nerve damage. 6 Similarly, Schwann cells undergo dramatic reprogramming from highly quiescent, mature, differentiated myelinating cells to proliferative, prorepair cells after nerve injury and exhibit Wallerian degeneration. In addition, the proliferation and migration of Schwann cells and the inhibition of cell aging and apoptosis can restore the structure of injured peripheral nerves.
TLR4 can induce neuroinflammation and neuralgia. TLR4 is expressed on the cell surface as well as in endosomes, mainly in immune and glial cells. 7 Myeloid differentiation primary response 88 (MyD88) is the most common adaptor protein that interacts with the intracellular domain of TLR4, which can activate transcription factors such as nuclear factor κ light-chain enhancer of activated B cells (NF-κB) and mitogen-activated protein kinase (MAPK). 8 Following injury, TLR4 activation on microglia and macrophages contributes to their shift towards an inflammatory phenotype and thus their release of inflammatory factors, including IL-1β, TNFα, and IL-6. 9 Intrathecal administration of TLR4 antagonists and siRNAmediated suppression of TLR4 signaling prevents activation of the NF-κB pathway and production of TNF and IL-1β, which attenuates mechanical allodynia and thermal hyperalgesia in a chronic constriction injury (CCI)-induced pain model. 10,11 Ferulic acid exhibits a potential advantage in the treatment of sciatica. Ferulic acid can decrease the levels of oxidative stress, inflammation and apoptosis markers in the sciatic nerves of patients with diabetes. 12 Ferulic acid exerts a neuroprotective effect against radiation-induced nerve damage by targeting the NLRP3 inflammasome to enhance learning and memory ability and ameliorate pathological changes in the hippocampal tissues of irradiated mice. 13 In this study, ferulic acid was found to relieve pain in CCI rats, and we aimed to identify the related mechanisms. The therapeutic effect of ferulic acid on CCI of the sciatic nerve was assessed via behavioral tests, pathological examination, and immunohistochemistry. Next, we investigated the underlying mechanism at the cellular level to provide additional experimental evidence supporting the application of ferulic acid for the treatment of sciatica.  CCI, ferulic acid, GMI-R1, inflammation, NF-κB, pain, RSC96, sciatica, TLR4 and RNase-free water (AG11701, AG11602, AG11012) were ob-

| Sciatica model
The CCI model (sciatica model) was constructed as described in previous studies. [14][15][16][17][18][19] Before the mice were anesthetized with pentobarbital sodium (3%; 40 mg/kg) and fixed to the operation table, the rats were fasted for 12 h. After exposing the sciatic nerve under a microscope, the right sciatic nerve was tied 4 times with 4.0 sutures at intervals of approximately 1 mm. At this point, we observed a small twitch in the operated hind limb. The rats in the sham group did not undergo nerve ligation. Finally, gentamicin (10 mg/ml, i.m.) was injected.

| Treatment programs
Thirty-two rats were randomly divided into four groups: the sham operation group, the CCI group, the mecobalamin group, and the ferulic acid group. The rats in the sham operation group and the CCI group were given saline (0.9%, 12 ml/kg), the rats in the mecobalamin group received a gavage of mecobalamin (20 mg/kg), and the rats in the ferulic acid group received ferulic acid (100 mg/kg) by gavage. 20 The drugs were administered for 21 days.

| Behavioral tests
To assess cold hyperalgesia, we chose the acetone experiment. A total of 100 μl of acetone was dropped on the plantar surface of the right paw. Next, we noted the total number of times that the rat lifted or clutched its right hind paw within 120 s. The von Frey test was used to evaluate mechanical hyperalgesia (50% mechanical withdrawal threshold [MWT]). The hind paw was stimulated 10 times with each filament (2.0-26.0 g) beginning with the 2-g filament, and paw lifting was considered a positive response. If we detected a positive response, we calculated the pain threshold (50% g threshold = 10 [Xf+kδ]/10,000 ). We assessed hyperalgesia of the right paw on the 1st, 4th, 7th, 14th, and 21st days after surgery. All tests were repeated three times at 10-min intervals for each paw, and the mean was calculated.

| RNA sequencing
The sciatic nerves of rats in the sham operation group and CCI group were collected 21 days postinjury. Total RNA was isolated using RNAiso Plus. Subsequently, the concentration and quality of the total RNA were assessed using a Nano Drop and Agilent 2100 bioanalyzer (Thermo Fisher Scientific). After the mRNA was purified with oligo(dT)-attached magnetic beads, it was fragmented into small pieces with fragment buffer at the appropriate temperature. Then, first-strand cDNA was generated using random hexamer-primed reverse transcription, followed by secondstrand cDNA synthesis. Afterwards, A-Tailing Mix and RNA Index Adapters were added by incubation for end repair. The cDNA fragments obtained in the previous step were amplified by PCR, and the products were purified by Ampure XP Beads and then dissolved in EB solution. The products were validated with an Agilent Technologies 2100 bioanalyzer for quality control. The doublestranded PCR products obtained in the previous step were heated, denatured and circularized by the splint oligo sequence to obtain the final library. Single-strand circular DNA (ssCir DNA) was formatted as the final library. The final library was amplified with phi29 to make DNA nanoballs (DNBs), which had more than 300 copies of a single molecule. DNBs were loaded into the patterned nanoarray, and paired-end 150-base reads were generated on the DNBSEQ-T7 platform by Tsingke Biotechnology Co., Ltd. The raw reads were filtered using the Trim Galore method (https://ccb. jhu.edu/softw are/hisat 2/index.shtml) to obtain clean reads for subsequent analysis and to ensure the quality of the information analysis. The clean reads obtained after filtering were compared with the reference database annotations (the Rno6 version of the rat genome was selected) using HISAT2 software (https://ccb.jhu. edu/softw are/hisat 2/index.shtml). Differentially expressed genes (DEGs) were screened using Sangerbox, and two criteria were used for screening the DEGs: a false discovery rate (FDR) ≤0.05 and |Log2-fold change (FC) | ≥ 1. Then, we used DAVID to conduct KEGG enrichment analysis.

| H&E staining, immunohistochemistry, and ELISA
Liver, kidney, and sciatic nerve tissues were fixed in 4% paraformaldehyde and then sliced at a thickness of 3 μm for H&E staining and 9 μm for immunohistochemistry. The sections were deparaffi- To measure the concentrations of serum inflammatory factors, we collected 8 ml of blood from the abdominal aorta. The blood samples were centrifuged at 4032 g for 15 min at 4°C, and 800 μl of the supernatant was retained for measurement of inflammatory factor levels. The remaining steps were performed according to the manufacturer's instructions. We used a microplate reader (Bio Tek Instruments, Inc.) to measure the optical density.

| Cell viability and cytotoxicity assays
The viability of GMI-R1 cells and RSC96 cells was determined GMI-R1 cells and RSC96 cells were cultured in 6-well plates for 24 h.
Next, the medium was discarded, and the cells were washed with PBS. Drugs and LPS were added to the medium at the same time, and then the cells were cultured for 24 h.

| Flow cytometry analysis
We collected and washed GMI-R1 cells and RSC96 cells. We first measured the percentages of M1 and M2 microglia among LPStreated GMI-R1 cells. The membrane protein CD32 was detected by direct staining. The cells were fixed using fixation buffer, permea-

| Quantitative real-time PCR
Total RNA was harvested using RNAiso Plus and synthesized into cDNA with an RT-PCR kit according to the manufacturer's instructions. The relative mRNA expression was calculated by the 2 −ΔΔCq method after normalization to the level of β-actin expression. 21 The Applied Biosystems 7900 real-time PCR (qPCR) system, SYBR® Green Premix qPCR, and primers, which are shown in Table 1, were used for quantitative real-time PCR.

| Behavioral tests
After CCI, rats exhibited cold and mechanical hyperalgesia from the 4th day to the 21st day ( Figure 1A,B; p < 0.05). Treatment with ferulic acid and mecobalamin (positive control drug) relieved neuropathic pain but did not normalize sensitivity from the 4th day to the 21st day (p < 0.05). The analgesic effects of ferulic acid were not different from those of mecobalamin (p > 0.05). These results showed a lack of full functional recovery of the injured sciatic nerves.

| Analysis of DEGs and KEGG pathway analysis
In the present study, DEGs were identified by comparing gene expression in the sciatic nerve between the sham operation group and the CCI group. A total of 14,622 genes were found, and 3777 genes with an FDR ≤0.05 and |Log2(FC)| ≥ 1 ( Figure 1C,D) were selected as DEGs.
A heatmap ( Figure 1D) was used to visualize the expression of the differentially expressed genes in each sample. Then, we performed KEGG pathway analysis of the 3777 DEGs via the DAVID database ( Figure 1E; p < 0.05). KEGG pathway analysis ( Figure 1E) revealed that the relationship between the "NF−kappa B signalling pathway" and the "Toll-like receptor signalling pathway" was quite close, and the two pathways ranked in the top 20. The "NF-kappa B signaling pathway" is ranked in the top 3, and the "TLR4/NF-κB" pathway is also involved. The "NF-kappa B signaling pathway" is involved in the "Tolllike receptor signaling pathway." We chose the TLR4/NF-κB pathway to conduct experimental verification. The differentially expressed genes associated with these two pathways are shown in Figure 1E,F.

| H&E staining
The structures of the liver, kidney, and sciatic nerve were observed via H&E staining. The neural structure of the sciatic nerve ( Figure 1G) was normal in the sham operation group but was destroyed after CCI. Ferulic acid and mecobalamin helped restore nerve structure.
Liver and kidney structures ( Figure 1H,I) were normal in all groups, which indicated that neither the drugs nor CCI had negative effects on the livers and kidneys of the rats.

| Immunohistochemical staining of the sciatic nerve
IBA-1 (a microglial and macrophage marker), M1 polarization markers (IL-1β, CD32, and iNOS), nNOS, COX2, and TLR4 were expressed at low levels in the normal sciatic nerve in the sham operation group but were expressed at higher levels after CCI (Figure 2A; p < 0.05).
The levels of IBA-1, IL-1β, CD32, iNOS, nNOS, COX2, and TLR4 in the ferulic acid and mecobalamin groups were lower than those in the CCI group (p < 0.05). S100β (a Schwann cell marker) and Arg-1 (an M2 polarization marker) were expressed at higher levels in the sham operation group and at lower levels in the CCI group (Figure 2A; p < 0.05). Ferulic acid and mecobalamin increased the expression of S100β and Arg-1 (p < 0.05).
Ferulic acid and mecobalamin decreased the levels of these proteins. The levels of p38MAPK and NF-κB were nearly equal among the groups (p < 0.05). In addition, ferulic acid and mecobalamin restored the levels of IBA-1, IL-1β, iNOS, TLR4, Myd88, p-NF-κB, and p-p38MAPK to nearly normal levels (p < 0.05).

| Ferulic acid alleviated LPS-induced apoptosis via the TLR4/NFκ B pathway in RSC96 cells
Flow cytometry indicated that LPS induced RSC96 cell apoptosis ( Figure 3A; p < 0.05). PCR showed that LPS increased the mRNA expression levels of Bax and Bad and decreased the mRNA levels of Bcl-2 and Bcl-xl ( Figure 3B; p < 0.05). Ferulic acid, TAK-242 or JSH23 inhibited LPS-induced cell apoptosis (p < 0.05), but ferulic acid alone did not affect Schwann cell viability ( Figure 3A). In addition, ferulic acid combined with TAK-242 or JSH23 lowered the mRNA level of Bad to normal levels (p < 0.05), but it could not normalize the level of Bax or increase the levels of Bcl-2 and Bcl-xl to normal levels ( Figure 3B; p < 0.05).

| Ferulic acid promoted the transformation of M1 GMI-R1 microglia to M2 microglia following LPS treatment via the TLR4/NFκ B pathway
To explore whether ferulic acid has anti-inflammatory effects, we conducted PCR and flow cytometry to assess whether ferulic acid regulates the levels of inflammatory cytokines. Figure 4 shows that the mRNA levels of the M1 microglia-related proinflammatory cytokines IL-1β, IL-6, iNOS, and CD32 were increased after LPS treat- and p-NF-κB. These results indicated that ferulic acid may reduce inflammatory factor levels and promote nerve repair. Therefore, we verified the anti-inflammatory mechanism of ferulic acid in GMI-R1 cells and the mechanism by which it inhibits apoptosis in RSC96 cells.
S100β is a marker of Schwann cells, which secrete neurotrophic factors and provide structural support and guidance to promote nerve regeneration. 23 The autologous transplantation of Schwann cells can promote human peripheral nerve repair in 7.5-cm and 5-cm sciatic nerve injuries. 24 After CCI, S100β distribution was altered, and CCI reduced the expression of S100β, indicating that the normal IBA-1 is a marker of activated microglia and macrophage.
Activation of microglia and macrophage induces neuroinflammation. 25 Macrophages in the sciatic nerve had some characteristics of microglia. 26 Acute injury of the sciatic nerve led to a rapid infiltration of circulating monocytes, and the monocytes quickly adapted a macrophage phenotype. 27 The GMI-R1 cell (microglia) was used to conduct the vitro experiments in this study. Immunohistochemical analysis of the sciatic nerve showed that CCI led to the activation of macrophage or microglia and that ferulic acid suppressed these ac-

| CON CLUS ION
Ferulic acid can alleviate sciatica in CCI rats and inhibit neuroinflammation, promote sciatic nerve repair and exert an analgesic effect via the TLR4/NF-κB pathway.

AUTH O R CO NTR I B UTI O N S
Di Zhang, Bei Jing, Zhenni Chen, and Guoping Zhao contributed substantially to the experimental design, data analysis and experimental procedures. Huimei Shi, Xin Li, Shiquian Chang, Zhenni Chen, Li Gao, and Yachun Zheng assisted with the English writing and partial experiments. We thank Yixuan Li for her valuable comments on the statistical analysis and English writing. Guoping Zhao is the corresponding author. All data were generated in-house, and no paper mill was used. All authors agree to be accountable for all aspects of the work, ensuring its integrity and accuracy.

ACK N OWLED G M ENTS
This study was supported by the National Natural Science Foundation of China (grant No. 81874404 and 82274294).

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are available from the corresponding author upon request.