The ethics committee has approved our study protocol for collecting patient clinical information from Affiliated Hangzhou Chest Hospital, Zhejiang University School of Medicine. This protocol was strictly followed throughout the entire experimental process. Informed consent has been obtained from the patients.

Overall, 12 eligible patients with TB were recruited from Affiliated Hangzhou Chest Hospital, Zhejiang University School of Medicine, between September 2021 and December 2021. The diagnostic criteria for linezolid-induced peripheral neuropathy were as follows: a history of MDR-TB with typical TB symptoms; imaging and laboratory evidence; and linezolid administration for >3 months at a daily dose of ≥600 mg. The exclusion criteria were as follows: 1) HIV positivity; 2) presence of severe heart, liver, kidney, and other organ diseases contraindicating anti-TB drug administration; 3) age <14 years; 4) non-TB branch Bacillus infection; and 5) loss to follow-up after treatment. Critical clinical information on pain type, duration, and pain characteristics of linezolid-induced peripheral neuropathy was collected from patients using questionnaires.

All animal experimental procedures were approved by the Laboratory Animal Ethics Committee of Zhejiang University. Adult male SD rats (200–250 g) were obtained from the Laboratory Animal Center of Zhejiang University. Five rats per cage were housed under the following controlled environmental conditions: 23°C ± 2°C temperature, 35%–60% humidity, and a 12:12 h light: dark cycle. The animals had free access to water and food. After 1 week of acclimatization, all the experimental animals were used for animal experiments. The rats were then randomly divided into four groups: Sham group (n = 4), 250 mg/kg-linezolid treated group (n = 5), 125 mg/kg-linezolid treated group (n = 5), 50 mg/kg-linezolid treated group (n = 5). The body weight of the rats was measured weekly. Four weeks later, the rats were performed behavioral tests. Afterward, the rats were anesthetized with pentobarbital sodium (50 mg/kg) and sacrificed. For drug administration, linezolid (Pfizer, New York, United States) was dissolved in saline, and the rats were treated through intragastric administration at the dose of 250, 125, and 50 mg/kg/d for 4 weeks. These doses in rats have been described previously (De Vriese et al., 2006).

The Basso Beattie and Bresnahan (BBB) scale was adopted to evaluate the locomotor ability of the hind limbs in spinal cord injury models and is also applicable to peripheral nerve injury models (Pertici et al., 2014; Watzlawick et al., 2014; Li et al., 2018). The scores were divided into 22 levels of bilateral hind limb motor function. A score of 0 indicates total paralysis, whereas a score of 21 indicates normal walking. A double-blind method was conducted to experiment. The main observation parameters included a range of motion, number of joint movements, degree of weight-bearing, body balance, and front and hind limb coordination. After administration with linezolid for 4 weeks, the rats were placed in an open and quiet field and allowed to explore autonomously. After the exploration activity was over, each rat was independently observed for 5 min. At the same time, the motor function of both hind limbs was assessed, and the observation scores were recorded.

Four weeks after linezolid administration, the rats were performed walking track analysis (Amado et al., 2008; Tang et al., 2013). The rats’ hind limbs were dipped in dark ink and allowed to walk independently through a black tunnel (100 cm * 10 cm). White paper at the bottom of the tunnel recorded the behavior of the rats’ hind limbs. The same investigator who never participated in the behavioral experiments repeated the experimental process thrice.

Sciatic nerves were immersed in 4% paraformaldehyde (PFA; Gibco, California, United States) for 24–48 h, after which the sciatic nerves were dehydrated with ethanol gradient and embedded in paraffin. 5 μm sciatic nerve sections were cut from paraffin and rehydrated. Then the sections were incubated in 3% H₂O₂ for 15 min and blocked by 5% bovine serum albumin (BAS; Beyotime, Shanghai, China) for 30 min. Subsequently, the sections were incubated with primary antibody at 4°C overnight. After washing with phosphate-buffered saline (PBS; Gibco, California, United States) with 0.1% Tween-20 (Aladdin, Shanghai, China), the sections were incubated with Alexa Fluor 647 (1:1000, Abcam) or Alexa Fluor 488 (1:1000, Abcam) as secondary antibodies at 37°C for 60 min. Cellular nuclei were stained with 4ʹ,6-diamidino-2-phenylindole (DAPI; Yesen, Shanghai, China). All fluorescence images were obtained under the Nikon ECLIPSE 80i (Nikon, Tokyo, Japan). The following primary antibody were used: MBP (1:1000, CST), NF-200 (1:1000, CST), LC3B (1:1000, CST).

RSC 96 (ScienCell Research Laboratories, Shanghai, China), which is the rat SCs line, was cultured in Dulbecco’s Modified Eagle Medium (DMEM; Gibco, California, United States) containing 1% penicillin/streptomycin solution (Gibco, California, United States) and 10% fetal bovine serum (FBS; Gibco, California, United States) in a humidified incubator (37°C, 5% CO₂). In the confirmation phase of linezolid, experiments were divided into four groups: control group, 250, 125, and 62.5 μg/ml. Schwan cells were inoculated in 96-well plates (3× 10³ cells/well) and 6-well plates (1× 10³ cells/well) for cell viability analysis and colony-forming cell assays, respectively. Subsequently, SCs were seeded into 6-well plates at a density of 1 × 10⁵ cells/well to explore the mechanism. The selected wells were given CQ (40 μM; Selleck Chemicals, Houston, Texas, United States) and linezolid (250 μg/ml) for 72 h. Otherwise, SCs were treated with CQ 2 h before linezolid treatment.

Cell viability assay was evaluated with cell counting kit-8 (CCK-8; MCE, Monmouth Junction, NJ, United States). Discarded the blank medium and drug-containing medium in the 96-well plate, washed it three times with PBS, and added 100 μl of fresh medium. After that, 10 μl of CCK-8 solution was added to each well. Placed in the incubator for a proper time. When the color changed to orange, a microplate reader measured the absorbance values via spectrophotometry at 450 nm. At least three replicate wells were set for each group of cells.

SCs in each group at the logarithmic growth stage were collected to prepare suspension for cell counting and dilution. The cells were then inoculated in a 6-well plate and gently rotated so that the cells were evenly dispersed. After cell adhesion, different concentrations of linezolid solutions were added, while the control group was added PBS. Incubate in an incubator for 14 days, and replace fresh medium every 3 days. The culture was terminated after 2 weeks, and the medium was abandoned. The cells were fixed with methanol for 15 min and stained with Giemsa dye for 10 min. After washing away the excess dye, the cells were counted, and the number of clones formed was observed under a light microscope.

Proteins were extracted from RSC96 cells using a lysis buffer containing protease and phosphatase inhibitor cocktail (10 μl/ml, Beyotime, Shanghai, China). After centrifugation for 10 min at 12,000 rpm, the supernatant was taken, and the protein concentration was detected by the Bradford kit (Abcam, Cambridgeshire, England). Protein lysates containing 40 μg of protein were separated on 10% or 12% SDS–polyacrylamide gels and transferred to polyvinylidene fluoride membranes (Merck Millipore, Darmstadt, Germany). After blocking with 5% skim milk for 90 min at room temperature, the membranes were incubated with primary antibody at 4°C overnight. Three times washing with TBST (TBS with 0.05% tween 20), the membranes were incubated with the corresponding secondary antibody. The primary antibodies used in this experiment were as follows: p-AKT (1:1000, CST), p62 (1:1000, CST), LC3A/B (1:1000, CST), and GAPDH (1:1000, CST).

All data obtained in this study were the average of at least three repetitions and were plotted as means ± SEM. Comparisons between two groups were performed using a one-way analysis of variance followed by Turkey’s multiple comparison test, with p < 0.5 indicating statistical significance.

From September 2021 to December 2021, we collected a total of 12 valid questionnaires. The average age of the patients was 34 years. The ratio of men to women respondents was 1:2 (Figure 1A). Through the questionnaires, we determined that 10 patients developed linezolid-induced pain, which mainly manifested as mild pain (Figure 1B). In addition to the sensation of pain, 83.0% of patients developed numbness (Figure 1C). Besides, 66.6% of patients reported that linezolid-induced pain was mainly in the feet (Figure 1D). This led us to consider that linezolid caused clinical symptoms similar to diabetic peripheral neuropathy (DPN) (Liu J. et al., 2022; Fang, Wang, Song, Wang, & Zhang, 2022).

Behavioral tests were performed after administration of linezolid for 4 weeks. BBB scores and walking track analysis showed that no abnormal motor function was observed in each group (Figures 2A,B). During the weekly monitoring of rats’ body weight, we found that the rats in the 250-mg/kg dose group exhibited moderate growth during the first 2 weeks of feeding (Figure 2C). However, compared with the sham group, the 250-mg/kg doses group presented with a significant difference in body weight on the fourth week (Figure 2D). Moreover, we also fortuitously observed that the 250-mg/kg dose group showed decreased activity and depression after 4 weeks of continuous administration.

The damage to neurons and myelin sheaths is a fatal blow to the nervous system (Swire et al., 2021; Liu Q. et al., 2022). In the current study, we found that nerve fibers and myelin sheaths exhibited a clear sciatic nerve structure in rats from the sham group. They displayed myelin sheath tightly wrapped laterally around the nerve fibers (Figure 3A). In the linezolid-treated group, nerve fibers and myelin sheaths were arranged sparsely, and some degree of loss in neurons and myelin sheaths was observed, which was more severe in the 250-mg/kg dose group (Figures 3A–C). Our previous studies have shown that appropriate autophagy has a neuroprotective effect and promotes peripheral nerve regeneration (Wu et al., 2019; Li et al., 2020). Through immunofluorescence staining, we observed a significant decrease in autophagic structural protein (LC3B) expression in the sciatic nerve after linezolid administration (Figures 3D,E), suggesting that linezolid contributes to the blockage of autophagy flux.

To better elucidate the significance of autophagy in linezolid-induced peripheral neuropathy, SCs were selected for subsequent experimental studies. Different concentrations of linezolid (62.6, 125, and 250 μg/ml) were exposed to SCs across different durations (24, 48, and 72 h) (Figures 4A–C). The results of CCK-8 assay showed that linezolid did not inhibit the growth of SCs at a concentration of 62.5 μg/ml. At the concentration of 250 μg/ml, linezolid could significantly inhibit the growth of SCs when the incubation time was 48 and 72 h, and the inhibition effect of 72 h-incubation is stronger than 48 h-incubation. After 72 h-incubation, the pseudopodia of SCs were shortened and the shape of cells was round under the light microscope (Figure 4D). The results of monoclonal experiments also confirmed SCs were significantly reduced in proliferative capacity at dose 250 μg/ml (Figures 4E,F). Given that 250 μg/ml of linezolid overtly induced toxicity in SCs at 72 h, this same concentration was used in in-vitro experiments.

SCs autophagy plays an essential role in peripheral neuropathy (Belgrad, De Pace, & Fields, 2020); therefore, we detected the autophagy function in our in-vitro model. After administering different concentrations of linezolid, we found p-AKT protein expression levels to be associated with linezolid in network pharmacology. The expression levels of p-AKT were significantly upregulated after administration, whereas the autophagy pathway of SCs was inhibited (Figures 5A,B). To further elucidate its mechanism of action, we added CQ for verification. We found that after the addition of CQ, the SCs shape changed from spindle to spherical (Figure 5C). Notably, our CCK8 assay findings further revealed that when linezolid and CQ act on SCs simultaneously, their proliferation ability is more strongly inhibited (Figure 5D). This indicated that the survival of SCs was associated with the level of autophagy and that linezolid inhibited the proliferation of SCs by inhibiting the autophagy pathway. Finally, we added linezolid and CQ at the same time and detected changes in the expression level of autophagy marker protein LC3A/B. We found that both linezolid and CQ could reduce LC3A/B expression, with the simultaneous use of linezolid and CQ promoting a more significant inhibition of LC3 A/B (Figures 5E,F).