Modified Wen Luo Tong Alleviated Chemotherapy-Induced Peripheral Neuropathy via Regulating the CX3CL1/CX3CR1 Axis Through Inhibiting NF-κB and STAT3 Pathways

Abstract

Background:

Chemotherapy-induced peripheral neuropathy (CIPN) is a serious clinical problem with no widely applicable solutions. Modified Wen Luo Tong (mWLT) was designed specifically for paclitaxel-related CIPN, yet its efficacy and mechanisms remain unclear. This study aimed to investigate its therapeutic effects and potential mechanisms via network pharmacology and animal experimental validation.

Methods:

Paclitaxel-induced CIPN rat models were treated with mWLT pediluvium. The effect of mWLT was estimated by behavior test. The components and targets of 5 herbs in mWLT were screened from TCMSP and TCMIP databases. CIPN-related targets were retrieved from Genecards and DisGeNET. Networks of gene ontology and pathway associations related to intersection targets were constructed and visualized. A pharmacological network encompassing the intersecting genes and active components was mapped out. A protein-protein interaction network was established for these intersecting targets and visualized using Cytoscape software. Finally, the findings derived from network pharmacology were validated through a series of in vivo experiments, including ELISA, Western Blot, immunohistochemistry and RT-qPCR. Molecular docking was used to predict binding sites between small molecules of mWLT and CX3CR1.

Results:

mWLT ameliorates mechanical withdrawal threshold of CIPN model rats. Three hundred and three targets of mWLT against CIPN were identified through intersection analysis, and 8 hub targets such as IL6, TNF and STAT3 were pinpointed. Enrichment analysis of intersection targets highlighted cellular response to cytokine stimulus, JAK-STAT3 pathway and NF-κB pathway. Thus, we speculated that mWLT may exert its effects by acting on IL6 and TNF, subsequently regulating IL6-JAK-STAT3 and TNFα-NF-κB signaling pathway, ultimately mitigating CIPN. Experimental validation demonstrated that mWLT significantly decreased the levels of IL-6, IL-1β and TNF-α in both spinal cord and plasma. Additionally, mWLT downregulated the phosphorylation of JAK, STAT3 and NF-κB in spinal cord. Further analyses using Immunohistochemistry, Western Blot and ELISA confirmed that mWLT reduced the protein expression of CX3CL1. RT-qPCR results revealed downregulation of Cx3cl1 and Cx3cr1 mRNA level in spinal cord and dorsal root ganglia. Molecular docking predicts 4 potential of compounds derived from mWLT to treat CIPN targeting CX3CR1.

Conclusion:

mWLT exerts therapeutic effects in the treatment of CIPN by inhibiting CX3CL1/CX3CR1 axis and modulating JAK-STAT3 and NF-κB pathway.

Background

Chemotherapy-induced peripheral neuropathy (CIPN) is a serious clinical problem that typically occurs in patients with cancer who have received specific chemotherapy agents, like taxanes and platinums. ¹ Taxanes, including paclitaxel and docetaxel, along with their liposomal or nanoparticle albumin-bound formulations, serve as the cornerstone of antineoplastic chemotherapy. They are widely used in the treatment of malignant tumors and have significantly improved the survival rates of patients with cancer. However, paclitaxel-related CIPN is highly prevalent, with an incidence rate ranging from 57% to 83%. ² It represents a significant dose-limiting toxicity, characterized by distal, symmetrical, sensory peripheral neuropathy. ² Common symptoms include, numbness and/or pain. As the cumulative dose of chemotherapy drugs increases, these symptoms tend to worsen progressively, leading to dose reduction, or even the suspension of chemotherapy. Therefore, CIPN poses a substantial challenge for clinical practice.^1,3 Unfortunately, there is limited evidence-based data available for prophylaxis and treatment of CIPN. Clinicians might recommend acupuncture, ⁴ cryotherapy, compression therapy^5,6 for prevention and serotonin-noradrenalin reuptake inhibitor such as duloxetine or anticonvulsants such as gabapentin/pregabalin ³ for treatment. However, considering common adverse reactions and potential drug interactions of these drugs, these are not widely applicable solution for CIPN.^1,3

Wen Luo Tong (WLT) Decoction is a traditional Chinese medicine (TCM) herbal formula specifically designed for topical treatment of CIPN. This formula was developed by Dr. Liaqun Jia at China-Japan Friendship Hospital. It has been widely applied in clinical practice to manage CIPN over 2 decades. A randomized, double-blind, controlled clinical study involving 102 patients with CIPN confirmed that WLT effectively alleviated pain and reduce the severity grading of CIPN. ⁷ Previous study indicated that the active components of WLT microemulsion may affect peripheral nerves via 2 pathways: by directly acting on the nerve fiber within the skin or penetrating through skin into the circulation. ⁸ Patients with CIPN show different symptoms due to chemotherapy agents; for instance, oxaliplatin-induced neuropathy is quite sensitive to cold stimuli while paclitaxel neuropathy is not.^3,9 Thus, personalized treatment (or precision treatment) is needed, which is also in line with the Principle of Syndrome Differentiation and Treatment of TCM. Given that paclitaxel chemotherapy often induces significant fatigue associated with Qi deficiency in TCM,^9,10 the herb Astragali Radix, known for its Qi-tonifying properties, has been added to the original WLT formula to create modified WLT (mWLT). However, the efficacy of mWLT in treating paclitaxel-induced peripheral neuropathy requires further evaluation, and its pharmacological mechanisms remain to be elucidated.

Consistent with other traditional Chinese herbal formulas, mWLT consists of 5 herbs, which may contain multiple active constituents that synergistically interact by targeting multiple pathways involved in CIPN. However, this complexity also presents significant challenges for elucidating the mechanisms of Chinese herbal formulas. In recent years, the advancements in systems biology and bioinformatics have given rise to network pharmacology, an emerging approach that bridges the gap between TCM theory and modern pharmacological research. These emerging approaches have become increasingly effective and provides a novel paradigm to uncover and visualize the intricate interaction networks of TCMs against complex diseases.^11,12 By constructing comprehensive networks of “herbs–components–targets–pathways–diseases,” network pharmacology provides a novel paradigm to systematically elucidate the multi-component, multi-target mechanisms of TCM. This approach has been successfully applied to clarify the potential mechanisms of various TCM formulas, such as Mo Luo Dan for chronic atrophic gastritis, ¹³ Xuefu Zhuyu Decoction for pulmonary hypertension, ¹⁴ and Xiao Chai Hu Decoction for chronic prostatitis. ¹⁵

In this study, we validated the beneficial effects of mWLT in CIPN treatment by behavioral test. Next, network pharmacology approach was employed to predict main components, related targets, potential biological processes and candidate pathways of mWLT in the treatment of CIPN. We finally validated these targets using a rat model and elucidated the mechanisms by which mWLT regulates the CX3CL1/CX3CR1 axis and inhibits the NF-κB and STAT3 pathways. This research provides theoretical support and experimental evidence for the further development and application of mWLT.

Materials and Methods

mWLT Preparation

The mWLT formula comprises 5 herbs: Astragali Radix, Cinnamomi Ramulus, Carthami Flos, and Epimedii Folium and Erodii Herba Geranii Herba. The prescription was prepared into Chinese medicinal granula with individual packaging, obtained and authenticated by Efong Pharmaceutical (Foshan, China). Details of the mWLT prescription, including pinyin names, plant sources, dosages and batch numbers, are listed in Table 1. The Latin names and its corresponding plant source of each herb were referenced from the Pharmacopeia of the People’s Republic of China (2020).

Table 1.

Composition of mWLT.

Pinyin name, initial	Latin name	Plant source	Herb (g)	Batch no. of granule
Huang Qi, HQ	Astragali Radix	Astragalus membranaceus (Fisch.) Bge.var.mongholicus (Bge.) Hsiao	30	0051013
Gui Zhi, GZ	Cinnamomi Ramulus	Cinnamomum cassia Presl	15	0025352
Hong Hua, HH	Carthami Flos	Carthamus tinctorius L.	10	0051803
Yin Yang Huo, YYH	Epimedii Folium	Epimedium brevicornu Maxim.	30	0452343
Lao Guan Cao, LGC	Erodii Herba Geranii Herba	Geranium wilfordii Maxim.	20	0035543

Animals and Experimental Design

Twenty-four male Sprague-Dawley rats weighing 180 to 220 g were purchased from Beijing Vital River Laboratory Animal Technology Co. Ltd. (Beijing, China). All rats were housed in an environment with automatically controlled conditions, using a 12 hours light–dark cycle with free access to food and water. After a 3-day acclimation period, the rats were randomly divided into 3 groups: Control, PTX and mWLT. Since we have previously conducted a dose-response experiment of WLT, ¹⁶ only 1 dose group of mWLT was set in this study.

To induce CIPN models, rats in PTX and mWLT groups were administered paclitaxel (8 mg/kg, cumulative dose of 24 mg/kg, equivalent to average dose of clinical practice) intraperitoneally on days 1, 4, and 7. Paclitaxel was diluted with saline at a ratio of 1:3 before injection. ¹⁷ Rats in Control group received an equivalent volume of saline injection.

For CIPN treatment, rats in the mWLT group received foot soaking therapy with mWLT solution for 30 minutes twice a day. The mWLT solution was prepared by dissolving 1 package granules into 1000 ml deionized water and maintaining the temperature at approximately 40°C. The intervention began 1 day before paclitaxel administration and continued for 11 days (days 0-10). Rats in PTX and Control groups underwent foot-soaking with deionized water instead.^16,18 The rats were placed in a transparent immobilizer with holes at the bottom. The immobilizer was submerged into a shallow tray containing the herbal solution, allowing only the rats’ paws to be immersed in the liquid (Figure 1A).

Figure 1.

Schematic representation and therapeutic effects of modified Wen Luo Tong (mWLT) on paclitaxel-induced peripheral neuropathy (CIPN) in rats. (A) Experimental design for the application of mWLT in a rat model of CIPN. Three groups contain Ctrl (Saline i.p. + water pediluvium), PTX (paclitaxel i.p. + water pediluvium) and mWLT (paclitaxel i.p. + mWLT pediluvium). CIPN rat model was induced by paclitaxel injected intraperitoneally (i.p.) at 8 mg/kg on d1, 4, 7. mWLT treatment was given topically by pediluvium twice a day for 11 consecutive days (d0-10). The behavioral test was monitored every other day (d0, 2, 4, 6, 8, 10). Samples were collected on day 10. (B) Graphical representation of the mechanical paw withdrawal threshold (PWT) over the course of the study. Error bars indicate SEM. n = 8; Compared with ctrl group, ***P < .001; ****p < 0.0001; Compared with PTX group, ^#p < 0.05; ^##p < 0.01; ^###p < 0.001.

Behavioral Test

Mechanical paw withdrawal threshold (PWT) was measured before, during, and after paclitaxel administration (on days 0, 2, 4, 6, 8, and 10) by an experimenter blinded to the grouping. Rats were placed in an elevated transparent chamber with a wire mesh floor and were acclimated for 15 minutes before measurement. The PWT was tested using calibrated von Frey filaments according to the “up and down” method. ¹⁹

Blood and Tissue Collection

Blood and tissue samples were collected on day 10, after the completion of behavioral tests and interventions. Rats were anesthetized with intraperitoneal injection of 3% pentobarbital sodium (40 mg/kg). Blood was collected from the abdominal aorta and then centrifuged to extract plasma. The L4-L6 segments of the spinal cord and dorsal root ganglion (DRG) were either stored at −80°C or fixed in 4% paraformaldehyde for further analysis.

Screening Candidate Compounds and Targets of mWLT

The compounds of each herb were screened in the Traditional Chinese Medicine Systems Pharmacology database (TCMSP, http://tcmspw.com/tcmsp.php) using the Pinyin names of the herbs and filtered based on the criteria: drug-likeness (DL) ≥0.18 and oral bioavailability (OB) ≥10%. The Mol IDs of the selected compounds were then entered into TCMSP to retrieve corresponding targets with full protein names. These protein names were conserved into gene symbols by search and matching in the Uniprot database (https://www.uniprot.org/) and the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov/). Additionally, the Integrative Pharmacology-based Research Platform of Traditional Chinese Medicine (TCMIP, http://www.tcmip.cn/TCMIP/index.php) was queried using Pinyin names. Only Chemical Components with a drug-likeness Grading of “good” and their Candidate Target Genes are included in the study.

Screening Genes Related to CIPN

The genes related to CIPN were collected by searching “CIPN” or “paclitaxel induced peripheral neuropathy” or “chemotherapy induced peripheral neuropathy” in following databases: GeneCards (https://www.genecards.org/) and DisGeNET (https://www.disgenet.org/). All results were integrated and the duplications were removed.

Potential Action Targets of mWLT Against CIPN and Compound-target Network Construction

To identify the potential therapeutic targets of mWLT for CIPN, we intersected the candidate targets of mWLT with CIPN-related genes using the Draw Venn Diagram (http://bioinformatics.psb.ugent.be/webtools/Venn/). The overlapping targets and their associated candidate compounds were identified as potential therapeutic targets and active components of mWLT for CIPN treatment. Additionally, candidate compounds interacting with these targets were also recognized as potential active components. These compounds and their corresponding targets were then used to construct a compound-target interaction network, which was visualized using Cytoscape 3.10.1.

Construction of PPI Network and Identification of Hub Targets

To construct the protein–protein interaction (PPI) network of mWLT’s potential targets against CIPN, we imported the targets into the STRING database (https://string-db.org/), selected Homo sapiens, set the interaction score to 0.900, and removed disconnected nodes. The resulting PPI network was exported in TSV format and visualized in Cytoscape 3.10.1. We analyzed the network topology using Cytoscape’s built-in tools and identified hub targets with cytoHubba, ²⁰ focusing on Degree, Closeness, and Betweenness centrality measures. ²¹ The top 15 targets ranked by each algorithm were intersected to determine the final hub targets for mWLT against CIPN.

GO Function and KEGG Pathway Enrichment Analyses

To elucidate the biological functions and implications of the potential action targets of mWLT, we performed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses using KEGG Orthology database (https://www.genome.jp/kegg/). The common targets identified through STRING analysis were imported into the Metascape database (https://metascape.org/) for GO and KEGG enrichment analysis, and the relevant data of Cellular Component (CC), Molecular Function (MF), Biological Process (BP), KEGG pathways were obtained. The top 10 GO terms and the top 20 KEGG pathways with the smallest P-values were selected and visualized as bar and bubble charts using the bioinformatics platform (http://www.bioinformatics.com.cn/).

Enzyme-Linked Immunosorbent Assay (ELISA)

Proteins were extracted from the frozen spinal cord and DRG tissues using lysis buffers to obtain a homogenate. The levels of TNF-α, IL-1β, IL-6, and CX3CL1 in plasma and/or homogenate were detected using ELISA kits (Raybiotech, USA) according to the manufacturer’s protocols.

Immunohistochemistry (IHC)

The fixed spinal cord tissues were sectioned into 5-μm-thick slices and processed for immunohistochemical analysis. The sections were incubated with a primary antibody against CX3CL1 (Abcam Cat# ab25088, RRID:AB_448600) overnight at 4°C. Subsequently, the sections were treated with a horseradish peroxidase-conjugated secondary antibody for 4 hours at room temperature. The staining was visualized using 3,3′-diaminobenzidine (DAB) as the chromogen.

Reverse Transcription Quantitative Polymerase Chain Reaction (RT-qPCR)

Total RNA was extracted from the spinal cord and DRG samples, and isolated using Beyotime RNAeasy™ Kit (Beyotime, China). The cDNA was synthesized using the Vazyme Script RT SuperMix (Vazyme, China). ABI StepOne™ System (Applied Biosystems, Foster City, CA, USA) was used to perform RT-qPCR with Vazyme SYBR Green Master Mix (Vazyme, China). The 2^−ΔΔCT method was used to calculate the relative expression levels of target genes. Sequences of the primer for quantitative real-time RT-qPCR were listed in Table 2.

Table 2.

Sequences of the Primer For Quantitative Real-Time RT-qPCR.

Gene	Sequences
Cx3cl1	F	5′-TCATTCAGAAGCTGCCAGGA-3′
	R	5′-AGAGTCCCTTCCAGAACACG-3′
Cx3cr1	F	5′-GCTGAGGCCTGTTATTTGGG-3′
	R	5′-GACCGAACGTGAAGACAAGG-3′
Gadph	F	5′-TGGCCTTCCGTGTTCCTACC-3′
	R	5′-TCTTCCACCACTTCGTCCGC-3′

Western Blotting (WB)

The L4-L6 segments of the rat spinal cords were lysed in RIPA lysis buffer. The tissue lysates (40 μg/lane) were separated via 10% SDS-polyacrylamide gel electrophoresis and were transferred to PVDF membranes. The membranes were blocked with 5% skim milk in TBST for 1 hour at room temperature. Subsequently, the membranes were incubated with primary antibody against CX3CL1 (Abcam Cat# ab25088, RRID:AB_448600), P-JAK2 (Abcam Cat# ab32101, RRID:AB_775808), p-STAT3 (Abcam Cat# ab32143, RRID:AB_2286742, p-NF-kB (Abcam Cat# ab76302, RRID:AB_1524028) and GAPDH (Abcam Cat# ab8245, RRID:AB_2107448) overnight at 4°C. On the second day, the membranes were washed 3 times with TBS and then incubated with appropriate secondary antibodies (Zhong Shan Jin Qiao Co. Ltd., China) for 1 hour at room temperature, Immunoreactive bands were visualized using an enhanced chemiluminescent reagent. The intensity of the band was quantified using a computer-assisted imaging analysis system (Gel-Pro Analyzer, Media Cybernetics, USA).

Molecular Docking

Based on the percutaneous components of mWLT microemulsion identified by UPLC-Q-TOF/MS in previous literature, ⁸ and the common components among various herbs of mWLT screened by network pharmacology, the 9 active component compounds of mWLT were determined. The files of sdf. structure of mWLT active ingredient compounds were downloaded from PubChem database, and imported into ChemBio3D 14.0 software to adjust the spatial conformation of active ingredients, calculate the optimization of energy, and save in mol2 format. Compounds of mWLT were selected as the ligands and the hub gene CX3CR1 was selected as the receptor. The three-dimensional crystal structure of the hub gene was downloaded from the PDB protein database (https://www.rcsb.org/). PyMOL 1.7.2.1 software was utilized to remove water molecules and impurity from the receptor protein, followed by the separation of the original ligand to obtain a standardized receptor. Next, the ligand and receptor were imported into Auto Dock Tools 1.5.6 software. Subsequently, the Grid tool was utilized to automatically identify the Grid Box for the docking and the dimension of Grid Box was manually adjusted based on the protein volume until the receptor was completely enveloped. Finally, molecular docking was performed using AutoDock Vina 4.2.6 software, and the docking results were visualized and assessed for hydrogen bond formation using PyMOL 1.7.2.1 software. Usually, binding energy lower than − 7.0 kcal/mol indicates excellent binding activity between receptor and ligand. ²²

Statistical Analysis

Statistical analysis was performed using GraphPad Prism 8 software. Data are presented as mean ± standard deviation (SD) unless otherwise indicated. Two-way ANOVA was employed to analyze the behavioral test data, whereas one-way ANOVA was applied for comparisons among other datasets. P < .05 was considered statistically significant.

Results

mWLT Ameliorates Mechanical Withdrawal Threshold of CIPN Model Rats

To evaluate the therapeutic effects of mWLT on CIPN, we conducted a classic behavioral test by measuring the mechanical paw withdrawal threshold (PWT).^{16 -18} As shown in Figure 1B, the mean PWT value of Control group was 24.78-26.00 g. Following paclitaxel administration, the PWT of PTX group was significantly reduced compared to the Control group from day 2 to 10, indicating marked mechanical allodynia induced by paclitaxel. Specifically, the PWT of PTX group decreased significantly from 26 to 18.2 ± 2.54 g (P < .001) on day 2 and continued to decline to 4.22 ± 0.78 g (P < .0001) by the end of the observation period. In contrast, rats co-administrated with mWLT exhibited a less pronounced decrease in PWT, with values ranging from 26 to 22.3 ± 1.83 and 8.22 ± 0.52 g. A statistically significant difference between mWLT and PTX group was observed since day 2 (P < .05). These results suggest that mWLT effectively alleviates mechanical allodynia in rats with paclitaxel-induced peripheral neuropathy, consistent with our previous findings. ¹⁶

Candidate Compounds and Targets of mWLT

To elucidate the mechanism of action of mWLT, we conducted a network pharmacology study. In TCMSP and TCMIP databases, 146 compounds were obtained, such as calycosin, 3,4-dihydroxybenzoic acid, and icariin. Specifically, there were 36, 9, 61, 52and 9 compounds in Huang Qi (HQ), Gui Zhi (GZ), Hong Hua (HH), Yin Yang Huo (YYH) and Lao Guan Cao (LGC), respectively. Some of these compounds were found in 2 or more herbs. Additionally, 429 targets of the compounds were acquired from TCMSP and TCMIP databases.

CIPN Genes and Action Targets of mWLT Against CIPN

A total of 5503 genes associated with CIPN were identified in GeneCards and DisGeNET databases. Subsequently, we determined the intersection between these CIPN-related genes and the candidate targets of mWLT, resulting in 303 targets. These targets are considered to be the key action targets through which mWLT exerts its therapeutic effects against CIPN (Figure 2A).

Figure 2.

Identification of potential targets and interaction network for mWLT in the context of CIPN. (A) Venn diagram illustrating the overlap of potential targets between CIPN and mWLT. (B) “Herbs-compounds-potential targets” network. Ellipticals above represent compounds in mWLT (“hq” represent compounds from Astragali Radix, “hh” for Carthami Flos, “gz” for Cinnamomi Ramulus, “yyh” for Epimedii Folium, “lgc” for Erodii Herba Geranii Herba), compounds that derived from 2 or more herbs were defined as “CF” (means “repetitive”). Yellow rectangles below represent 303 intersecting targets. The edges represent the interactions between compounds and targets.

Compound-Target Network, PPI Network and Hub Targets

Then, the compound-target network was constructed using the Cytoscape 3.10.1 software. As depicted in Figure 2B, the network comprised 93 compounds (eg, kaempferol, quercetin, lupeol, astragalin and apigenin) and 303 targets (eg, TNF, IL6 and CX3CL1) in the network. It was observed that 1 compound could regulate multiple targets, and 1 target could be modulated by multiple compounds, suggesting that mWLT might exert its efficacy through a combination of multiple compounds and multiple targets. Additionally, the action targets of mWLT against CIPN were imported into the STRING database to construct a PPI network. After hiding disconnected nodes, the network contained 266 nodes and 1153 edges (Figure 3A). Next, the top 15 targets ranked by Degree, Closeness and Betweenness were intersected, and we ultimately identified 9 targets, including TP53, AKT1, STAT3, JUN, HSP90AA1, TNF, ESR1, and IL6 (Figure 3B and C), which were considered as the key in the mechanism of action of mWLT against CIPN.

Figure 3.

PPI network and hub targets of mWLT against CIPN. (A) PPI network. A node represents a target. The color and the size of the node represent the value of the degree. Blue→yellow→orange indicates that the degree value is from low to high, and the bigger the circle, the higher the degree value. (B) The intersecting 8 targets of the top 15 targets ranked by the 3 topological algorithms (Degree, Closeness, and Betweenness) were identified as the hub targets of mWLT against CIPN. Degree refers to the number of nodes connected by a node; Closeness is used to measure the average distance between a node and other nodes; Betweenness is an indicator of node importance characterized by the number of shortest paths passing through a node. (C) PPI network of the 8 hub targets. The color of the node represents the value of the degree. Yellow→ orange→ pink indicates that the degree value is from low to high, and the bigger the circle, the higher the degree value.

GO and KEGG Enrichment Analysis of mWLT Targets

The 303 intersecting targets identified were subjected to enrichment analysis with a significance threshold of P < .05, resulting in the identification of 230 KEGG pathways and 3129 GO functional entries. The top 30 results from the GO enrichment analysis are illustrated in Figure 4A. Specifically, there were 2608 BP results, including response to oxygen levels, cellular response to cytokine stimulus, and regulation of phosphorylation; 176 CC results, including membrane raft, receptor complex, transcription regulator complex, and perinuclear region of cytoplasm; 345 MF results, including kinase binding, transcription factor binding, nuclear receptor activity, and protein kinase activity. The top 20 pathways from the KEGG enrichment analysis, as shown in Figure 4B, yielded 230 pathways, included pathways in cancer, lipids and atherosclerosis, NF-κB signaling pathway and JAK-STAT signaling pathway.

Figure 4.

GO term and KEGG Pathway enrichment of the potential action targets of mWLT against CIPN. (A) GO enrichment analysis results, categorized by biological process (BP), cellular component (CC), and molecular function (MF). (B) Top 20 of KEGG pathways. The color gradient indicates the q-value, with more purple hues representing more significant enrichment. The size of the dots corresponds to the count of genes associated with each GO term or pathway.

mWLT Inhibited Pro-inflammatory Cytokines in CIPN Model Rats

According to the PPI interaction network, the hub targets included IL6, STAT3 and TNF, which is consistent with the biological process of cellular response to cytokine stimulus by GO analysis. It is widely recognized that inflammation in the spinal cord and DRG, along with increased expression of chemokines and cytokines such as IL-1β, IL-8, and TNF-α, account for the main pathomechanism of CIPN. ²³ As a result, we focused on the pathway of pro-inflammatory cytokines. In plasma samples, ELISA analysis showed that levels of IL-6 in PTX rats significantly increased (Figure 5A). Similar situation was also observed in the results of IL-1β and TNF-α (Figure 5B and C). In tissue samples from the spinal cord, these pro-inflammatory cytokines were also higher in PTX rats (Figure 5D-F). Compared with PTX group, the levels of IL-6, IL-1β, and TNF-α in the mWLT group were significantly lower in both plasma and spinal cord samples. Notably, these changes indicated that mWLT could mitigate the production and release of pro-inflammatory cytokines induced by paclitaxel administration.

Figure 5.

mWLT inhibited inflammatory cytokines in plasma and spinal cord of CIPN model rats. (A) IL-6 in plasma. (B) IL-1β in plasma. (C) TNF-α in plasma. (D) IL-6 in spinal cord. (E) IL-1β in spinal cord. (F) TNF-α in spinal cord. n = 8, by ELISA. ****P < .0001

mWLT Inhibits the Phosphorylation of JAK, STAT3 and NF-κB in CIPN Model Rats

Next, we investigated the potential mechanisms of mWLT based on the predicted results from KEGG analysis. The JAK-STAT3 pathway and NF-κB pathway can be activated by cytokines, including IL-6, IL-1β and TNF-α.^24,25 Given that phosphorylation is the critical step of pathway activation, we further examined whether mWLT could inhibit phosphorylation of key proteins within these pathways in lumbar spinal cord (L4-L6). Western blot analysis showed that paclitaxel administration increased the expression of p-JAK, p-STAT3 and p-NF-κB in spinal cords, while mWLT application reversed these effects. Quantitative analysis confirmed significant differences (P < .0001). The results are presented in Figure 6.

Figure 6.

mWLT inhibited the expression of p-JAK, p-STAT3 and p-NF-κB in spinal cord of CIPN model rats. (A) Western blot images and quantitive results of p-JAK. (B) p-STAT3. (C) p-NF-κB. n = 8. ****P < .0001

mWLT Attenuates the Expression of CX3CL1 and CX3CR1 in CIPN Model Rats

According to previous studies, IL6-JAK-STAT3 pathway and TNFα/IL1β-NF-κB pathway are regulated by CX3CL1/CX3CR1 axis.^{26 -29} We subsequently examined CX3CL1 protein level in the spinal cord. As depicted in Figure 7A, CX3CL1 protein was detected through IHC staining and observed under a light microscope, appearing as brownish-yellow. Compared to the control group, the PTX group exhibited a significantly larger area of brownish-yellow staining, indicating vigorous expression, which was notably reduced in the mWLT group. Western Blot and ELISA analyses further confirmed that mWLT treatment decreased CX3CL1 protein levels in spinal cord and or in DRG (Figure 7B and C). To assess the transcriptional impact of mWLT, we measured the mRNA expression of CX3CL1 and CX3CR1 in spinal cord and DRG by RT-qPCR. Compared to the control group, PTX rats showed significantly increased expression levels of Cx3cl1 and Cx3cr1 mRNA in both spinal cord and DRG, which were mitigated by mWLT at the transcriptional level (Figure 7D and E). These findings suggest that the therapeutic effects of mWLT in paclitaxel-induced peripheral neuropathy likely involve inhibition the CX3CL1/CX3CR1 axis, leading to the downregulation of NF-κB and JAK-STAT3 pathways.

Figure 7.

mWLT inhibited CX3CL1/CX3CR1 in CIPN model rats. (A) Immunohistochemistry (IHC) staining showed that mWLT decreased insoluble CX3CL1 in spinal cord. (B) Western blot images and quantitive results of CX3CL1 protein in spinal cord. (C) ELISA analysis of CX3CL1 in spinal cord and dorsal root ganglia (DRG). (D) Relative Cx3cl1 mRNA in spinal cord and DRG. (E) Relative Cx3cr1 mRNA expression in spinal cord and DRG. n = 8. ****P < .0001.

Molecular Docking Verification

To identify potential small-molecule compounds in mWLT that may alleviate paclitaxel-induced peripheral neuropathy pain by targeting CX3CR1, we ranked 9 molecules in ascending order according to binding energy in Table 3. Top 8 of them are the binding energy lower than − 7.0 kcal/mol. Those compounds showing excellent binding activity were majorly derived from Epimedii Folium, Carthami Flos and Astragali Radix. Then, the docking structures of the top 4 compounds ranked by affinity and the number of hydrogen bond were visualized (Figure 8). Myricetin formed 1 hydrogen bond with SER147 and 2 with ARG150 in CX3CR1 respectively (Figure 8A). As shown in Figure 8B, quercetin formed 3 hydrogen bonds with LEU318, ARG150, SER147 and CYS148. Kaempferol formed 5 hydrogen bonds with LEU318, ARG150, SER147, MET101 in CX3CR1(Figure 8C). Figure 8D shows that luteolin interacted with VAL276, ARG150, LYS210 and HIS213 of CX3CR1 through 5 hydrogen bonds. In short, molecular docking revealed the potential binding interactions between compounds derived from mWLT and CX3CR1, further indicating the material basis for mWLT to target CX3CR1 and inhibit the CX3CL1/CX3CR1 axis.

Table 3.

Top 9 Compounds Ranked by Affinity and Number of Hydrogen Bonds with CX3CR1 in Molecular Dockings.

Compounds	PubChem ID	Affinity (kcal/mol)	Number of hydrogen bonds	Structural classification	Source (Latin name)
Myricetin	5281672	−10.122	3	Flavonoids	Erodii Herba Geranii Herba/Carthami Flos
Quercetin	5280343	−10.052	3	Flavonoids	Astragali Radix/Carthami Flos/Epimedii Folium
Kaempferol	5280863	−9.834	5	Flavonoids	Carthami Flos/Epimedii Folium
Luteolin	5280445	−9.477	5	Flavonoids	Carthami Flos/Epimedii Folium
Beta-Sitosterol	222284	−9.259	0	Terpenoids	Cinnamomi Ramulus/Carthami Flos
Icariin	5318997	−9.253	2	Flavonoids	Epimedii Folium
Calycosin	5280448	−8.9	4	Flavonoids	Astragali Radix
Hydroxysafflor Yellow A	6443665	−8.204	5	Flavonoids	Carthami Flos
3,4-dihydroxybenzoic acid	72	−6.261	5	Phenylpropanoids	Cinnamomi Ramulus

Figure 8.

Molecular docking between the hub target CX3CR1 and the top 4 compounds from mWLT ranked by affinity and the number of hydrogen bonds. Molecular docking between CX3CR1 and (A) Myricetin, (B) Quercetin, (C) Kaempferol, (D) Luteolin, respectively. The label represents the amino acid compositions around the ligand, the yellow line represents the hydrogen bond, and the number next to the yellow line represents the length of the hydrogen bond.

Discussion

Chemotherapy-induced peripheral neuropathy (CIPN) is a serious dose-limiting toxicity with few effective medical solution for prophylaxis and treatment.^{1 -3} Clinical practice and trial have suggested that WLT can alleviate pain and reduce the severity grading. ⁷ We have been improving the WLT formula to suit patients’ conditions and found that mWLT has better efficacy in treating paclitaxel-related CIPN according to our clinical observations. In this study, we confirmed mWLT’s therapeutic effects through serial mechanical PWT testing, explored mechanisms using a network pharmacology approach, and verified potential biological processes and candidate pathways in CIPN rat models.

Symptoms of CIPN generally manifest as paresthesia, numbness and/or pain. In traditional Chinese medicine (TCM) theory, these are believed to be caused by blood stasis blocking the meridians and collaterals.^9,10 Theoretically, the original 4 herbs of WLT work synergistically to address the condition: Gui Zhi and Hong Hua warm the meridians to unblock the collaterals, while Yin Yang Huo and Lao Guan Cao dispel wind and activate blood circulation to remove bi-syndrome (obstruction). Scientifically, the efficacy of WLT on CIPN has been demonstrated repeatedly in both clinical trials and model rats.^7,16,18 For platinum-related CIPN, we prefer adding Aconite (Wu Tou) to the formula to enhance the warming effect, specifically targeting cold hypersensitivity. For taxanes-related CIPN, we usually incorporate Astragali Radix (Huang Qi) to boost the body’s energy and address the issue of fatigue and weakness. The latter is the mWLT studied in this paper. Consistent with our previous experience, mWLT attenuated paclitaxel-induced mechanical allodynia in rats, as evidenced by a higher paw withdraw threshold. The precise efficacy of these modifications gives us confidence to further explore their pharmacological mechanisms.

In this study, we applied network pharmacology and ultimately identified 9 hub targets of mWLT, including STAT3, TNF, IL6 etc. GO enrichment analysis indicated that the cellular response to cytokine stimulus is among the top 10 biological process. Chemical agents, including paclitaxel, affect cells of the central and peripheral nervous system, induce elevated levels of inflammatory cytokines, promote glial activation and accumulation, and enhance pain sensation.^23,30 Previous studies have demonstrated that levels of TNF-α, IL-1β, and IL-6 are elevated in CIPN animal model. Neutralizing antibodies against TNF-α or IL-6 have been shown to be effective in reducing chemotherapy-induced allodynia.^31,32 Pharmacological blockade of TNF-α and IL-1β have attenuated both evoked and spontaneous pain in a paclitaxel rat model of CIPN. ³³ The above findings suggest that the inhibition of these cytokines may be effective in treating CIPN, which could also be a potential mechanism for mWLT to treat CIPN. Therefore, we measured IL-6, IL-1β and TNF-α levels in both plasma and spinal cord and confirmed the downregulation of these cytokines, which are involved in the mechanisms of mWLT’s effects on CIPN. To further decipher the material basis of mWLT in targeting CX3CR1, we employed molecular docking to simulate the interactions between bioactive compounds from mWLT and the CX3CR1 receptor. The computational predictions revealed that several key flavonoids, including myricetin, quercetin, kaempferol, and luteolin, could stably bind within the active pocket of CX3CR1 with high affinity (binding energy < −7.0 kcal/mol). Critical interactions, such as hydrogen bonds with amino acid residues SER147, ARG150, and LEU318, were identified, which are known to be important for the receptor’s function. This finding provides a structural basis for our experimental observations, suggesting that these compounds may act as potential antagonists or modulators of CX3CR1.

Previous studies and our KEGG analysis highlight NF-κB and STAT3 pathways, both classical inflammatory pathways.^34,35 TNF-α, IL-1β and IL-6 are involved in the activation of NF-κB and STAT3 in neuroinflammation.^{36 -38} Paclitaxel induces the activation of NF-κB by activating TRL4 on macrophages, thus initiating a cascade of inflammation and cytokines that induce CIPN. ³⁹ Previous studies have demonstrated that the JAK-STAT3 pathway is crucial in nerve pain and peripheral nerve injury. JAK inhibitors have been shown to significantly alleviate pain symptoms in model rats and reduce levels of inflammatory cytokines IL-1β, IL-6, and TNF-α, highlighting the pathway’s key role in nerve pain.^40,41 In this study, we took JAK-STAT3 and NF-κB as candidate pathways to be verified in model rats. We found that mWLT decreased the phosphorylation of JAK, STAT3, and NF-κB, which were elevated by paclitaxel stimulation. This suggests that mWLT may alleviate paclitaxel-induced peripheral neuropathy by inhibiting the IL6-JAK-STAT3 and TNFα/IL1β-NF-κB pathways.

Recent studies has underscored the significance of CX3CL1/CX3CR1 signaling in neurological disorders including CIPN.^23,42 CX3CL1 (also known as fractalkine) is a member of the CX3C chemokines family, which exists as a membrane-bound or soluble molecule. Intrathecal injection of CX3CL1 can induce intense mechanical pain and thermal hypersensitivity in rats. ⁴³ The CX3CL1-CX3CR1 axis promotes chemotaxis of CX3CR1+ cells toward soluble CX3CL1 as well as adhesion of CX3CR1+ cells to membrane-bound CX3CL1, facilitating neuron-monocyte/macrophage/microglia crosstalk. ⁴⁴ It has been reported that increased expression of CX3CL1, along with other chemokines and cytokines, stimulates microglia activation and accumulation, thereby contributing to neuroinflammatory responses following exposure to paclitaxel. ²³ Similar findings were observed in rats with vincristine-induced peripheral neuropathy that CX3CR1+ monocytes infiltrated sciatic nerve and were activated by CX3CL1. This activation led to the production of reactive oxygen species, which in turn activated the TRPA1 receptor in sensory neurons, triggering pain. ⁴⁵ Previous studies have established that CX3CL1/CX3CR1 axis influences a variety of inflammatory processes via multiple signaling pathways, including JAK-STAT, NF-κB, MAPK, AKT, etc.^29,46,47 Specifically, upregulation of CX3CL1 expression through NF-κB pathway activation has been observed in paclitaxel-induced peripheral neuropathy. ¹⁷ These findings suggest that targeting CX3CL1/CX3CR1 axis could be pivotal in managing CIPN. Consequently, we extensively examined changes in CX3CL1 and CX3CR1 using various approaches at different sites, and at both gene and protein levels. The results consistently demonstrated that mWLT inhibited CX3CL1/CX3CR1 axis in CIPN rats.

As shown in Figure 9, this study elucidates the potential mechanism of mWLT mitigation of chemotherapy-induced peripheral neuropathy (CIPN). Specifically, following exposure to paclitaxel, upregulated CX3CL1 induces the migration of CX3CR1-expressing monocytes or microglia into nervous tissue where they further become activated and release pro-inflammatory cytokines, including IL-6, IL-1β, and TNF-α. Subsequently, those cytokines as well as CX3CL1/CX3CR1 axis in neurons activated JAK-STAT3 pathway and NF-κB pathway by a phosphorylation cascade, which further drive the transcription of additional cytokines and chemokines such as IL-6, IL-1β, TNF-α, and CX3CL1, perpetuating the cycle of inflammatory activation.

Figure 9.

The biological mechanism of mWLT against CIPN may involve the inhibition of CX3CL1/CX3CR1, JAK-STAT3 and NF-κBpathway.

While this study provides valuable insights into the potential mechanisms of mWLT in CIPN treatment, several limitations should be noted. The pharmacodynamic evaluation in this study did not include 2 or more dose groups of mWLT at different concentrations. There were 2 concerns while designing the study. First, the research standards for topical formulations of traditional Chinese medicine TCM decoctions are not as mature as those for oral medications, and there is currently no established paradigm for reference. Although OB/DL parameters were originally designed for oral drugs, they can be used in this study to screen components in the topical formulation that “possess potential skin absorption capacity.” Specifically, components with high DL values are more likely to penetrate the stratum corneum barrier of the skin, while the “component stability” characteristics associated with OB can predict whether they are prone to degradation during local retention in the skin, thereby affecting the local effective concentration and therapeutic efficacy. Second, the low sensitivity and high variability of behavioral tests used in the efficacy evaluation of CIPN may not be sufficient to detect differences between different dosage groups. Nevertheless, this pilot study of mWLT laid the foundation for future efforts to further optimize the composition and develop new dosage forms that have a better transdermal absorption rate and are more convenient to use.

Conclusion

Network pharmacology has empowered us to predict the potential compounds, targets, and pathways through which mWLT exerts its effects against CIPN, thereby guiding subsequent experimental validation. Our research concludes that paclitaxel triggers neuroinflammation and pain via a multitude of molecular mechanisms. These include the involvement of chemokines, cytokines, the CX3CL1/CX3CR1 axis, as well as the NF-κB and JAK-STAT3 pathways, thereby unveiling the intricate pathophysiology underlying CIPN. Notably, molecular docking simulations provided critical insights by predicting that key flavonoid components in mWLT, such as myricetin and quercetin, could directly bind to the CX3CR1 receptor with high affinity, underscoring a plausible molecular initiation point for the observed inhibition of the CX3CL1/CX3CR1 axis. The insights gleaned from these findings present promising targets for the development of therapeutic strategies aimed at combating chemotherapy-induced neuropathy.

Supplemental Material

sj-pdf-1-ict-10.1177_15347354251408836 – Supplemental material for Modified Wen Luo Tong Alleviated Chemotherapy-Induced Peripheral Neuropathy via Regulating the CX3CL1/CX3CR1 Axis Through Inhibiting NF-κB and STAT3 Pathways

Supplemental material, sj-pdf-1-ict-10.1177_15347354251408836 for Modified Wen Luo Tong Alleviated Chemotherapy-Induced Peripheral Neuropathy via Regulating the CX3CL1/CX3CR1 Axis Through Inhibiting NF-κB and STAT3 Pathways by Feize Wu, Jianjing Zhang, Hongyan Wang, Hanwen Luo, Huimiao Lyu, Yawen Zhang, Guanjing Ling, Yilin Li, Jinghua Li, Wei Wang, Qiyan Wang and Linghui Lu in Integrative Cancer Therapies