Massage ameliorates lumbar disc herniation-related radicular pain in rats by suppressing TLR4/NLRP3 inflammasome signaling transduction

Abstract

Background

Lumbar disc herniation (LDH) is a common spinal disease that can cause severe radicular pain. Massage, also known as Tuina in Chinese, has been indicated to exert an analgesic effect in patients with LDH. Nonetheless, the mechanism underlying this effect of massage on LDH remains unclarified.

Methods

Forty Sprague-Dawley rats were randomly divided into four groups. A rat LDH model was established by autologous nucleus pulpous (NP) implantation, followed by treatment with or without massage. A toll-like receptor 4 (TLR4) antagonist TAK-242 was administrated to rats for blocking TLR4. Behavioral tests were conducted to examine rat mechanical and thermal sensitivities. Western blotting was employed for determining TLR4 and NLRP3 inflammasome-associated protein levels in the spinal dorsal horn (SDH). Immunofluorescence staining was implemented for estimating the microglial marker Iba-1 expression in rat SDH tissue.

Results

NP implantation induced mechanical allodynia and thermal hyperalgesia in rat ipsilateral hindpaws and activated TLR4/NLRP3 inflammasome signaling transduction in the ipsilateral SDH. Massage therapy or TAK-242 administration relieved NP implantation-triggered pain behaviors in rats. Massage or TAK-242 hindered microglia activation and blocked TLR4/NLRP3 inflammasome activation in ipsilateral SDH of LDH rats.

Conclusion

Massage ameliorates LDH-related radicular pain in rats by suppressing microglia activation and TLR4/NLRP3 inflammasome signaling transduction.

Keywords

massage intervertebral disc degeneration microglia nucleus pulposus TLR4/NLRP3

Introduction

Lumbar disc herniation (LDH) is a common spinal disease that can damage spinal nerve roots and lead to severe radicular pain, with features of allodynia, hyperalgesia, and spontaneous pain.¹ Radicular pain is a major source of chronic pain among the adult population, which adversely impairs patients’ quality of life and brings great economic burdens.² Evidence suggests that the autologous implantation of nucleus pulposus (NP) to spinal nerve roots can result in radicular pain-related behaviors in rodent models.^3,4 Thus, we established a rat LDH model by autologous NP implantation.

Currently, both surgical and conservative treatments are available for LDH. Percutaneous endoscopic lumbar discectomy is a commonly used minimally invasive therapeutic strategy for LDH due to its advantages in several aspects such as less soft tissue injury, shorter operation time, and faster recovery.⁵ Nonetheless, the incidence of surgery-related complications is 15–30%, and the risk of recurrent LDH after surgery is reported to be 2–25%.⁶ Clinically, conservative approaches are frequently employed as the first choice for LDH treatment and have achieved certain favorable outcomes.^6,7 Massage (Tuina in Chinese), a traditional Chinese manual therapy, has been used as one of the conservative approaches in treating LDH.⁸ Massage exhibits an analgesic effect on neuropathic pain in clinical practice.⁹ Cao et al. revealed that massage alleviates pain and improves dysfunction in LDH patients.¹⁰ Moreover, Huang et al. demonstrated that massage attenuates neuropathic pain by remodeling the synaptic structure in the spinal cord dorsal horn of a rat model.¹¹ Despite these findings, the mechanism by which massage exerts its analgesic effect on LDH remains unclarified.

Toll-like receptor 4 (TLR4) belongs to the TLR family and acts as a critical regulator in pathogen recognition and innate immunity.¹² Nod-like receptor family pyrin domain containing 3 (NLRP3) inflammasome is an intracellular multiprotein complex comprising NLRP3, ASC, and caspase-1.¹³ Moreover, TLR4 can activate NLRP3 inflammasome, and evidence has suggested the significant role of the TLR4/NLRP3 inflammasome signaling pathway in many diseases.^14,15 Elevated TLR4 levels have been observed after NP implantation, and TLR4 inhibitor prominently ameliorates LDH-related radicular pain in rats.¹⁶ Zhang et al. proposed that NLRP3 inflammasome was activated in dorsal root ganglion neurons of a rat LDH model.¹⁷ Nonetheless, it remains unclear whether TLR4 and NLRP3 inflammasome participate in the protective effect of massage against LDH.

This study aims to investigate the potential mechanism by which massage protects against LDH-related radicular pain in a rat model. We hypothesized that the TLR4/NLRP3 inflammasome pathway was involved in this protection effect of massage.

Materials and methods

Induction of rat LDH model

Male Sprague-Dawley rats (7-week-old, 220–240 g; Cavens, Changzhou, China) were kept under 12-h light/dark cycles in a temperature- and humidity-controlled (22 ± 2°C, 50–60%) environment with free access to food and water.^18,19 All rats were given 1-week acclimatization before the experiments. 40 rats were randomly grouped as: (1) Sham group; (2) Model group; (3) Model + TAK-242 group; and (4) Model + Massage group, with 10 rats per group.

The rat LDH model was established by autologous nucleus pulpous (NP) implantation according to previous reports.^2,4 In brief, rats were anesthetized by intraperitoneal injection of 3% sodium pentobarbital (40 mg/kg) and fixed in a prone position. The dorsal hair was shaved followed by routine disinfection. Then, a midline dorsal incision was made over the lumbar spine. The paraspinous muscles were carefully dissected from the spinous processes to expose the left transverse processes. Then, the left laminectomy was conducted in the L4–L5 segment to expose the nerve roots. Coccygeal NP (approximately 10 mg) was harvested from the second and third coccygeal intervertebral discs of rats, and then immediately implanted to the exposed lumbar nerve roots without compression (Figure 1(a)). The incision was sutured layer by layer. Rats in the Sham group received the same surgical procedure without NP implantation. No surgical complications were found in different groups.

Figure 1.

(a) Induction of LDH in the rat. (b) The device used for massage therapy at Shenshu (BL23) acupoints.

Intervention method

For the Model + Massage group, massage therapy was performed from day 2 after the surgery. The rats were placed in a massage manipulation finger-cot device for 30-min acclimatization before the intervention. Then the thumb and the index finger were used to press and knead the Shenshu (BL23) acupoints (located adjacent to the second lumbar vertebra)²⁰ for 10 mins once a day for 19 days (Figure 1(b)). A tactile sensor system (FingerTPS, Pressure Profile System, Los Angeles, CA) was employed to monitor the intensity of massage operation, with a pressure of 5 N and a frequency of 120 times/min. For the Model + TAK-242 group, rats were fixed for 10 mins without massage therapy and intraperitoneally injected with 3 mg/kg TAK-242, a TLR4 inhibitor (HY-11,109, MedChemExpress, Shanghai, China)¹⁶ 1 h before the surgery and then once daily for 7 days. Rats in other groups were injected with the same amount of normal saline as a vehicle. All experimental procedures were performed as per the NIH Guide for the Care and Use of Laboratory Animals, and approval was obtained from the Ethics Committee of Wuhan Traditional Chinese Medicine Hospital.

Behavioral test

Rats were allowed to move freely for 30 mins to adapt to the testing environment before behavioral tests. Mechanical sensitivity was evaluated using a Von-Frey test (IITC Life Science, Woodland Hills, CA) with an up-down method according to previous reports.²¹ Each rat underwent three trials (5-min interval) by measuring the paw withdrawal threshold (PWT) in response to external force, and the average was calculated. A thermal hyperalgesia test was conducted using a thermal plantar tester (IITC Life Science). The radiant heat source would be automatically deactivated when the rat retracted its paw, and the paw withdrawal latency (PWL) was recorded. Only one hindpaw was examined in each test session, with 20-s cut-off time to avoid tissue damage in case the rat did not retract its paw. Each animal was examined three times with a 5-min interval between tests, and the average value was taken as the result. The tests were conducted by two investigators blinded to animal treatments.

Immunofluorescence staining

At the end of the experiment, rats (n = 5) were euthanized under anesthesia by an overdose of sodium pentobarbital (100 mg/kg) and perfused with saline, followed by 4% paraformaldehyde in PBS. Then the L4–L5 spinal cord was excised, fixed with 4% paraformaldehyde for 3 h and soaked in 30% sucrose solution at 4°C for 48 h. The spinal tissue was transversely sectioned (25-μm-thick) using a cryostat microtome (Leica Biosystems, Shanghai, China). After treating with the blocking buffer (P0102, Beyotime, Shanghai, China), the spinal sections were incubated with anti-Iba-1 primary antibody (ab178846, 0.1 μg/mL, Abcam, Shanghai, China) overnight at 4°C, followed by three washes with PBS and incubation with the secondary antibody (ab150077, 1:200, Abcam) at room temperature for 1 h. The results were observed under a fluorescence microscope (Leica Microsystems, Shanghai, China).

Western blotting

After euthanasia by overdoses of sodium pentobarbital (100 mg/kg), the ipsilateral and contralateral spinal dorsal horns (SDH) were collected and stored in lipid nitrogen until use. RIPA lysis buffer (Beyotime) was employed for protein isolation from rat SDH. A BCA assay kit (Beyotime) was utilized for quantifying protein concentration. Then protein samples were separated on 10% SDS-PAGE and transferred to polyvinylidene fluoride membranes (Beyotime), followed by blocking with 5% defatted milk. Next, the membranes were incubated with the primary antibodies (Table 1) at 4°C overnight and washed thrice in TBST before incubation with the HRP-conjugated secondary antibody (ab97080, 1:5000, Abcam). Finally, blot signals were detected using BeyoECL Plus (Beyotime) and estimated using ImageJ software.

Table 1.

Primary antibodies used in western blotting.

Target	Cat. no.^a	Dilution
TLR4	ab217274	1:1000
NLRP3	ab263899	1:1000
ASC	ab309497	1:1000
Caspase-1	ab286125	1:5000
GAPDH	ab181602	1:10000

^aFrom Abcam, Shanghai, China.

Statistical analysis

Data are presented as the mean ± standard deviation and analyzed using GraphPad Prism 8.0 software (GraphPad, San Diego, CA). Difference comparisons were conducted by one-way ANOVA followed by Tukey’s post hoc analysis. Two-way ANOVA was performed for the analysis of behavioral tests. p < .05 indicated statistical significance.

Results

LDH rats shows mechanical allodynia and thermal hyperalgesia in ipsilateral hindpaws

A rat LDH model was established by autologous NP implantation, and behavioral tests of rats were conducted. In comparison to the sham-operated rats, the model rats exhibited a significantly decreased PWT in ipsilateral hindpaws for day 2 to day 20 after NP implantation (Figure 2(a)). In parallel, PWL was prominently reduced in ipsilateral hindpaws of the model rats from day 5 to day 20 as compared to that in the sham-operated rats (Figure 2(b)). In addition, there were no significant differences in PWT and PWL of the contralateral side between the model and sham groups (Figure 2(a) and (b)). These results revealed that NP implantation elicited persistent mechanical allodynia and thermal hyperalgesia in rat ipsilateral hindpaws.

Figure 2.

Behavioral tests of rats in the sham and model groups. (a)-(b) Mechanical allodynia test (a) or thermal hyperalgesia test (b) for evaluating PWT and PWL, respectively, in ipsilateral and contralateral hindpaws of rats at indicated time points. n = 10. Data are shown as mean ± standard deviation. **p < .01, ***p < .001. PWT: paw withdrawal threshold; PWL: paw withdrawal latency.

Activation of the TLR4/NLRP3 inflammasome pathway in LDH rats

Western blotting was performed to determine the expression levels of TLR4/NLRP3 signaling-related proteins in ipsilateral and contralateral SDH tissues of rats. Notably, relative to the sham group, the model group showed markedly higher protein levels of TLR4, NLRP3, ASC and caspase-1 in ipsilateral SDH tissues (Figure 3(a) and (b)), suggesting the activation of the TLR4/NLRP3 inflammasome pathway in the ipsilateral SDH of LDH rats. Additionally, the levels of these proteins in contralateral SDH tissues showed no significant differences between the sham and model groups (Figure 3(c) and (d)). Hence, ipsilateral SDH tissues of rats were used for further analysis.

Figure 3.

Activation of the TLR4/NLRP3 signaling pathway in LDH rats. (a), (c) Representative images of western blotting for determining protein levels of TLR4, NLRP3, ASC, Caspase-1 in ipsilateral (a) and contralateral (c) SDH of rats. (b), (d) Quantification of protein expression in ipsilateral (b) and contralateral (d) SDH of rats. n = 5. Data are shown as mean ± standard deviation. ***p < .001. SDH: spinal dorsal horn.

Massage reduces ipsilateral PWT and PWL and suppresses microglia activation in LDH rats

As depicted in Figure 3(a) and (b), massage therapy or TAK-242 treatment prominently counteracted the reduction in PWT and PWL in ipsilateral hindpaws of the model rats, suggesting that massage could attenuate mechanical allodynia and thermal hyperalgesia in LDH rats. Moreover, IF staining was performed to determine the distribution of Iba-1, a microglial marker, in the ipsilateral SDH of rats in each group. Notably, Iba-1 expression in the Model group was markedly higher than that in the sham group (Figure 4(c)). However, this effect was prominently abated after massage therapy or TAK-242 treatment in the model rats (Figure 4(c)), indicating that massage repressed microglia activation in the ipsilateral SDH of rats with LDH.

Figure 4.

Massage reduces ipsilateral PWT and PWL and suppresses microglia activation in LDH rats. (a)-(b) Mechanical allodynia test (a) or thermal hyperalgesia test (b) for evaluating PWT and PWL, respectively, in ipsilateral hindpaws of rats at indicated time points. n = 10. (c) Representative images of IF staining for detecting Iba-1 expression in rat ipsilateral SDH on day 20. n = 5. Data are shown as mean ± standard deviation. **p < .01, ***p < .001 versus sham group; #p < .05, ##p < .01, ###p < .001 versus model group. SDH: spinal dorsal horn.

Massage suppresses the TLR4/NLRP3 inflammasome activation in LDH rats

To investigate whether massage therapy affected the TLR4/NLRP3 inflammasome pathway in LDH rats, we examined the expression levels of associated proteins in ipsilateral SDH of rats. Additionally, TAK-242, a TLR4 inhibitor, was applied to inhibit this pathway. As expected, TAK-242 administration prominently decreased protein levels of TLR4, NLRP3, ASC and caspase-1 in ipsilateral SHD of the model rats (Figure 5(a)–(e)). Similar results were observed in massage-treated group (Figure 5(a)–(e)), indicating that massage therapy blocked NP implantation-evoked activation of the TLR4/NLRP3 inflammasome pathway in ipsilateral SDH of rats.

Figure 5.

Massage suppresses the TLR4/NLRP3 signaling pathway in LDH rats. (a) Representative images of western blotting for detecting TLR4, NLRP3, ASC, Caspase-1 protein levels in ipsilateral SDH of each group. (b)–(e) Quantification of protein expression in each group. n = 5. Data are shown as mean ± standard deviation. ***p < .001 versus sham group; #p < .05, ##p < .01 versus model group. SDH: spinal dorsal horn.

Discussion

The present study reveals that NP implantation induced severe mechanical allodynia and thermal hyperalgesia in ipsilateral hindpaws of rats and activated the TLR4/NLRP3 inflammasome pathway in rat ipsilateral SDH. Importantly, massage therapy or TLR4 antagonist prominently alleviated NP implantation-triggered paresthesia in ipsilateral hindpaws of rats, suppressed microglia activation and TLR4/NLRP3 inflammasome activation in ipsilateral SDH of the model rats.

Previous reports have shown the occurrence of mechanical allodynia and thermal hyperalgesia in animal model of LDH induced by application of NP to spinal nerve roots.²² Consistent results were observed in the present study, confirming that NP implantation LDH could cause radicular pain. As aforementioned, massage can relieve the neuropathic pain in patients with LDH.^9,11 Similarly, our results presented that massage could mitigate NP implantation-triggered radicular pain in rats with LDH, as manifested with the alleviation of mechanical allodynia and thermal hyperalgesia in rat ipsilateral hindpaws.

Microglia are the resident immune cells in central nervous system which mediate various pathological events.²³ Importantly, evidence suggests that microglia activation facilitates LDH-related radicular pain.^4,24 Previous studies have indicated that spinal microglia were primarily activated in the ipsilateral spine of animals with LDH, paralleled with the occurrence of pain behaviors in ipsilateral hindpaws.^2,16 Consistently, our study depicted the microglia activation in the ipsilateral SDH of rats with NP implantation, as evidenced by the increased expression of microglial marker Iba-1. Moreover, massage therapy markedly reduced Iba-1 expression in the ipsilateral SDH of LDH rats, indicating that massage might relieve LDH-related radicular pain by suppressing microglia activation.

Evidence suggests that the activation of TLR4 and NLRP3 inflammasome contributes to radicular pain in LDH.^16,17 In our study, we found that TLR4 and NLRP3 inflammasome (NLRP3, ASC, caspase-1) protein levels were prominently elevated in the model rats, which were consistent with previous evidence. Additionally, TLR4 acts as an upstream activators of NLRP3 inflammasome in many diseases.²⁵ Consistently, our results displayed that administration of TLR4 antagonist TAK-242 not only reduced TLR4 protein level but also suppressed expression of NLRP3 inflammasome-associated proteins in the ipsilateral SDH of LDH rats. Intriguingly, massage therapy showed similar effects to TAK-242 administration in LDH rats, suggesting that massage blocked the TLR4/NLRP3 inflammasome signaling transduction in ipsilateral SDH rats. Moreover, TLR4 is expressed on the surface of microglia and is critical in mediating microglia activation.²⁶ Our study depicted that inhibiting TLR4 markedly impeded microglia activation in the ipsilateral SDH of rats with NP implantation. In accord with previous reports,¹⁶ TAK-242 treatment, similar to massage therapy, mitigated pain behaviors in rats with NP implantation.

It is worth noting that this study still has some limitations. First, despite some pathological similarities to human LDH, this rat model of LDH may not exactly mimic human LDH pathology. Future studies are needed to elucidate our findings. Moreover, we only investigated the involvement of TLR4/NLRP3 signaling pathway in this study. Considering the complexity of mechanisms, more investigations are required to identify the mechanism underlying the protective effect of massage on LDH. Additionally, we performed mechanical allodynia and thermal hyperalgesia tests for pain behavior assessment. However, this may not necessarily reflect radicular pain. Further studies are needed to confirm pain as the primary effector examined in the behavioral tests by administering analgesics to LDH rats.

In conclusion, this study reveals that massage ameliorates NP implantation-evoked pain behaviors in rats possibly by suppressing microglia activation and blocking TLR4/NLRP3 signaling transduction. Our findings may provide new directions for treating LDH and relieving radicular pain.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: 1. Study on mechanism of Chinese School acupoint Massage on disc degeneration based on miR-224 mediated JAK/STAT pathway regulation of apoptosis (Grant number: ZY2023M026). 2. Based on the study of “DHPR regulation of intracellular calcium,” the mechanism of the deactivation of myofascialitis pain syndrome (Grant number: 20230202010205577). 3. The effect of meridian dredge and pain point kneading manipulation on the level of IL-17, TGF-β, TNF-α and GAP-43 in serum (Grant number: WZ20Q09).

ORCID iD

Shenghua He

References

Cho

Kim

Choi

, et al. The effect of GCSB-5 a new herbal medicine on changes in pain behavior and neuroglial activation in a rat model of lumbar disc herniation. J Korean Neurosurg Soc 2016; 59(2): 98–105.

Zhong

Huang

, et al. Puerarin alleviate radicular pain from lumbar disc herniation by inhibiting ERK-dependent spinal microglia activation. Neuropeptides 2018; 72: 30–37.

Miao

Liu

Wei

, et al. Lipoxin A4 attenuates radicular pain possibly by inhibiting spinal ERK, JNK and NF-κB/p65 and cytokine signals, but not p38, in a rat model of non-compressive lumbar disc herniation. Neuroscience 2015; 300: 10–18.

Huang

Zhong

, et al. Src-family kinases activation in spinal microglia contributes to central sensitization and chronic pain after lumbar disc herniation. Mol Pain 2017; 13: 1744806917733637.

Shan

Ren

Shi

, et al. Machine learning prediction model and risk factor analysis of reoperation in recurrent lumbar disc herniation patients after percutaneous endoscopic lumbar discectomy. Global Spine J. 2023; 21925682231173353.

Xie

Chen

Xiao

, et al. TNFAIP3 alleviates pain in lumbar disc herniation rats by inhibiting the NF-κB pathway. Ann Transl Med 2022; 10(2): 80.

Kesikburun

Eksioglu

Turan

, et al. Spontaneous regression of extruded lumbar disc herniation: correlation with clinical outcome. Pakistan J Med Sci 2019; 35(4): 974–980.

Miao

Tong

, et al. Tuina for lumbar disc herniation: a protocol for systematic review and meta analysis. Medicine 2021; 100(1): e24203.

Meng

Xing

, et al. Analgesic effect of tuina on rat models with compression of the dorsal root ganglion pain. J Vis Exp. 2023; (197).

10.

Cao

Zhou

Zhang

, et al. Effect of traditional Chinese manual therapy on alleviating pain and dysfunction of lumbar disc herniation: a randomized controlled pilot study. Am J Transl Res 2022; 14(10): 6941–6952.

11.

Hongye

Bingqian

Shuijin

, et al. Chinese tuina remodels the synaptic structure in neuropathic pain rats by downregulating the expression of N-methyl D-aspartate receptor subtype 2B and postsynaptic density protein-95 in the spinal cord dorsal horn. J Tradit Chin Med 2023; 43(4): 715–724.

12.

Chen

Tan

Xiao

, et al. Deletion of TLR4 attenuates lipopolysaccharide-induced acute liver injury by inhibiting inflammation and apoptosis. Acta Pharmacol Sin 2021; 42(10): 1610–1619.

13.

Yuan

Wang

Bhat

, et al. Differential effects of short chain fatty acids on endothelial Nlrp3 inflammasome activation and neointima formation: antioxidant action of butyrate. Redox Biol 2018; 16: 21–31.

14.

Liu

Dai

, et al. Beta-amyloid activates NLRP3 inflammasome via TLR4 in mouse microglia. Neurosci Lett 2020; 736: 135279.

15.

Yuan

Zhao

Wang

, et al. Cucurbitacin B inhibits non-small cell lung cancer in vivo and in vitro by triggering TLR4/NLRP3/GSDMD-dependent pyroptosis. Pharmacol Res 2021; 170: 105748.

16.

Zhu

Huang

, et al. Toll-like receptor 4/nuclear factor-kappa B pathway is involved in radicular pain by encouraging spinal microglia activation and inflammatory response in a rat model of lumbar disc herniation. Korean J Pain 2021; 34(1): 47–57.

17.

Zhang

Wang

Ding

, et al. Bay11-7082 attenuates neuropathic pain via inhibition of nuclear factor-kappa B and nucleotide-binding domain-like receptor protein 3 inflammasome activation in dorsal root ganglions in a rat model of lumbar disc herniation. J Pain Res 2017; 10: 375–382.

18.

Zhan

Pei

, et al. Involvement of 5-HT(1A) receptors of the thalamic descending pathway in the analgesic effect of intramuscular heating-needle stimulation in a rat model of lumbar disc herniation. Front Neurosci 2023; 17: 1222286.

19.

Yang

Zhu

, et al. Electroacupuncture relieves chronic pain by promoting microglia M2 polarization in lumbar disc herniation rats. Neuroreport 2023; 34(12): 638–648.

20.

Yao

Guo

Huang

, et al. Manual therapy regulates oxidative stress in aging rat lumbar intervertebral discs through the SIRT1/FOXO1 pathway. Aging 2022; 14(5): 2400–2417.

21.

Chaplan

Bach

Pogrel

, et al. Quantitative assessment of tactile allodynia in the rat paw. J Neurosci Methods 1994; 53(1): 55–63.

22.

Suzuki

Inoue

Gemba

, et al. Nuclear factor-kappa B decoy suppresses nerve injury and improves mechanical allodynia and thermal hyperalgesia in a rat lumbar disc herniation model. Eur Spine J 2009; 18(7): 1001–1007.

23.

Jeon

Youn

. Spinal gap junction channels in neuropathic pain. Korean J Pain 2015; 28(4): 231–235.

24.

Cho

Ahn

Kim

, et al. Changes in the expressions of Iba1 and calcitonin gene-related peptide in adjacent lumbar spinal segments after lumbar disc herniation in a rat model. J Kor Med Sci 2015; 30(12): 1902–1910.

25.

Yang

Wise

Fukuchi

. TLR4 cross-talk with NLRP3 inflammasome and complement signaling pathways in alzheimer’s disease. Front Immunol 2020; 11: 724.

26.

Zusso

Lunardi

Franceschini

, et al. Ciprofloxacin and levofloxacin attenuate microglia inflammatory response via TLR4/NF-kB pathway. J Neuroinflammation 2019; 16(1): 148.