This study was approved by the Ethics Committee at Henan Provincial People’s Hospital, Zhengzhou University. Thirty patients were prospectively enrolled at the Spinal Surgery Department and included 20 patients with LDH, suffering from radicular leg pain for less than 3 months and 10 patients with scoliosis or lumbar burst fracture as the control without leg painful symptomatology or degenerative disc disease. The LDH patient group had a mean age of 43 years and consisted of 11 men and 9 women. The control patient group had a mean age of 36 years and consisted of seven men and three women. Patients with LDH had assessed the intensity of leg pain on a 0–10 (0, no pain; 10, worst pain) visual analog scale (VAS) 1 day before discectomy. Patients with LDH were divided into two groups according to their preoperative pain scores (VAS ≤ 3 as mild pain group; VAS ≥ 4 as severe pain group). Intraoperative-collected herniated disc (HD) tissues, obtained during discectomy in patients with LDH and during orthopedic surgery in control patients, were collected. Immediately upon collection, tissues were divided into two parts. One part was flash frozen in liquid nitrogen and stored at −80°C for further use, and the other part was fixed in 4% formaldehyde solution for histopathological assessment.

Male Sprague Dawley rats (200–220 g) were obtained from the Institute of Experimental Animals of Zhengzhou University. All rats were housed in a temperature- and humidity-controlled environment on a 12/12-h light/dark cycle and provided with food and water ad libitum. All experimental procedures were approved by the Institutional Animal Care Committee of Zhengzhou University and were carried out in accordance with the guidelines of the National Institutes of Health Guide for the care and use of laboratory animals. Efforts were made to minimize animal suffering and to reduce the number of animals used.

Surgery for the lumbar disc herniation model was performed as previously described by Anzai et al. (2002) and Liu et al. (2017). In brief, rats were anesthetized intraperitoneally with sodium pentobarbital (50 mg/kg), and laminectomies were performed in which the left L5 nerve roots and corresponding dorsal root ganglion (DRG) were exposed. Autologous NP harvested from the coccygeal intervertebral disc was applied to the left L5 nerve roots just proximal to the corresponding DRG. The surgical procedure in the sham group was identical to the LDH group except for the application of NP to the left L5 nerve roots. TNF-ɑ inhibitor (500 μg Etanercept) was injected into the implanted NP after the application of NP to the left L5 nerve roots. The dose of Etanercept was determined based on the results from previous experiments (Horii et al., 2011; Inage et al., 2016). Special care was taken to prevent infection and minimize the influence of inflammation.

The 50% withdrawal threshold was assessed using von Frey hairs as described previously (Chaplan et al., 1994). Briefly, each rat was loosely restrained beneath a plastic box on a metal mesh for at least 15 min one time daily for 3 separate days, and mechanical allodynia in the LDH and sham groups were examined 1 day before surgery. Following 7 days of recovery, the test was performed weekly until 4 weeks postoperatively. Mechanical allodynia was assessed by the hind paw withdrawal threshold in response to probing with a series of von Frey filaments (bending force from 0.55 to 20.30 g) for 6 s or until the rat withdrew. A nociceptive response was defined as a brisk paw withdrawal or flinching of the paw, following von Frey filament application. Each test was repeated two to three times at approximately 2 min intervals, and the average value of von Frey filament force was determined as the force to evoke a withdrawal response. The experimenter who conducted the behavioral test was blinded to all treatments.

The herniated disc tissue was collected and immediately stored at −80°C until use. The tissue was homogenized on ice. Protein samples were separated by gel electrophoresis (SDS-PAGE) and transferred onto a PVDF membrane. The blots were incubated with a primary antibody against GLO1 (1:200, ABCAM, United States) and β-actin (1:2,000, Cell Signaling Technology, United States) overnight at 4°C according to the instructions of the manufacturer. The blots were then incubated with a secondary antibody. ECL (Pierce, United States) was used to detect the immune complex. After exposure for 2 min, the bands were achieved under Chemiluminescence and Fluorescence Imaging System (G:BOX XT4, Syngene, United Kingdom). The bands were quantified with a computer-assisted imaging analysis system (NIH Image J).

Nucleus pulposus cells were collected from the lumbar disc of 10 male Sprague-Dawley rats (3 months old). All experimental procedures described below were reviewed and approved by the Ethics Committee at Henan Provincial Peoples’ Hospital, Zhengzhou University, China. In brief, the rats were killed by an intraperitoneal overdose injection of 10% chloral hydrate, and the NP tissue was collected from the coccygeal intervertebral disc under aseptic conditions. After cutting the tissues into 1 × 1 mm³ sections, 0.2% type 2 collagenase (Sigma-Aldrich, St. Louis, MO, United States) was added and digested for 4 h. After washing with phosphate-buffered saline (PBS) and centrifuging for 5 min at 1,500 g, the isolated cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum (FBS) and antibiotics (100 U/ml penicillin and 100 U/ml streptomycins) at 37°C in a 5% CO₂ incubator. The total number of cells was less than 1,000. NP cells from the second passage were treated with different doses of TNF-ɑ.

Nucleus pulposus material was dissected on dry ice and stored at −80°C until use. The samples were homogenized, and the supernatant was collected for further assays. The total protein concentration was determined using the BCA Protein Assay Kit (Pierce, United States). GLO1 activity was measured as described by Hovatta et al. (2005) and the Glyoxalase 1 Assay Kit (Sigma-Aldrich, United States) according to the instructions of the manufacturer. The GLO1 activity rate was calculated by the absorbance at 240 nm.

The GSH/GSSG ratio was measured by using the described method with a minor modification (Baig et al., 2020). Briefly, NP material was dissected on dry ice and stored at −80°C until use. The samples were homogenized in 1.5 ml of a cold homogenization buffer for 1 min, and the supernatant was collected for further assays. GSH and GSSG were quantified on a fluorescent microplate reader at an excitation/emission wavelength set to 490/520 nm. Absolute amounts of GSH and GSSG were determined using GSH and GSSG standard curves.

The concentration of methylglyoxal was determined by HPLC using a simple derivatization procedure (Liu et al., 2017). Briefly, NP material was homogenized on dry ice, and the supernatant sample was supplemented with internal standard 5-methylquinoxaline (5-MQ) and the o-phenylenediamine (o-PD) at room temperature for 4 h. Perchlorate (PCA) was added to the derived sample and incubated on ice for 10 min. Methylglyoxal (2-MQ) and the quinoxaline internal standard (5-MQ) were measured using the conditions below. The analysis conditions were applied as follows: detector wavelength, 315 nm; mobile phase flow rate, 1. ml/min; typical sample size, 15 μl; and column temperature, 20°C. Duplicate injections of each sample were made. Samples were calibrated by comparison with a 2-MQ standard. The average retention times of 2-MQ and 5-MQ were 3.76 and 7.55 min, respectively.

All results are statistically confirmed SPSS 13.0 (SPSS, United States) and expressed as mean ± SEM. Statistical differences between the two groups were analyzed by one-way ANOVA. One-way or two-way ANOVA with repeated measures followed by Tukey, Dunnett, or Bonferroni post hoc test was carried out to compare differences between more than two groups. The criterion for statistical significance was p < 0.05. Complete statistical analysis is detailed in figure legends.

In the present study, we first found that the methylglyoxal levels of the herniated disc were significantly increased in the patients who suffered from the radicular leg pain accompanied by LDH compared with the patients with no-leg pain (Figure 1A). The analysis of Pearson correlation showed a strong positive linear correlation between VAS scores and a herniated disc MG level in patients with LDH (Figure 1B). To further investigate the causal relationship between MG- and LDH-induced radicular pain, the mechanical withdrawal threshold and the MG level in exposed NP tissues were examined in a rat NP implantation-induced LDH model. These animals exhibited significant mechanical allodynia on Days 7, 14, 21, and 28 after NP implantation (Figure 1C). Meanwhile, the MG level in exposed NP tissues was also significantly increased compared with that in the native NP tissue (Figure 1D). Note that the time course of increased MG was consistent with that of mechanical allodynia.

It is well known that the glyoxalase system in the cytoplasm is the main enzyme that metabolizes methylglyoxal and Glo1 as the rate-limiting enzyme and uses L-glutathione (GSH) as a cofactor (Gaffney et al., 2020). In the present study, we found that GLO1 activity of HD decreased in the patients who suffered from the LDH-induced radicular pain compared to the patients with no-leg pain (Figure 2A). The analysis of Pearson correlation showed a strong negative linear correlation between VAS scores and herniated disc GLO1 activity in patients with LDH (Figure 2B). In comparison with the patients with no-leg pain, immunoblotting and immunohistochemistry showed GLO1 expression decreased significantly in patients with radicular pain (Figures 2C,D). However, for GSH, there was no significant difference between patients with LDH-induced radicular pain and patients with no-leg pain (Figure 2E). Moreover, both GLO1 activity and expression decreased in the rats with NP implantation compared with that in the native NP tissue (Figures 2F,G).

Research showed that stimulation with the inflammatory mediator TNF-ɑ reduced GLO1 activity in human U937 monocytes (Hanssen et al., 2014). In the present study, we found that TNF-ɑ level increased in the patients who suffered from the radicular leg pain, accompanied by LDH compared to the patients with no-leg pain (Figure 3A). Meanwhile, the analysis of Pearson correlation showed a strong negative linear correlation between TNF-ɑ and GLO1 activity in HD (Figure 3B). To further investigate the relation between TNF-ɑ and GLO1 activity, the primary NP cells were cultured. The GLO1 activity and expression of NP cells exposed to TNF-ɑ were decreased significantly compared with that in the control group (Figures 3C,D). Furthermore, etanercept treatment (500 μg), a known TNF-ɑ inhibitor, significantly inhibited the mechanical allodynia induced by NP implantation in the LDH rat model (Figure 1C). Meanwhile, etanercept treatment (500 μg) also significantly attenuated the decrease of GLO1 activity and expression and the increase of MG in the implanted NP of LDH rat (Figures 1D, 2F,G).