Itaconate as a Promising Therapy for Neuropathic Pain and Inflammatory Pain


 Background: Chronic pain is a complex experience that often leads to multiple complications. Neuroinflammation in the immune system as well as in the peripheral and central nervous system contributes to the development and persistence of chronic pain. The metabolite itaconate, recently emerged as a regulator of immune cell functions, has features of anti-inflammation and immunomodulation via the activation of Nrf2. However, its effect on chronic pain is unclear. Methods: Chronic C57BL/6 neuropathic pain and inflammatory pain were induced followed by dimethyl itaconate (DI) treatment, an itaconate derivative. We evaluated the effect of DI on the changes of pain behaviors induced by chronic pain model. The effect of DI on inflammatory cytokine release, glial cells activation, ERK1/2 phosphorylation, inflammatory cell infiltration and Nrf2 expression in the DRGs, spinal cord and hind paw tissues was determined using the histochemistry, RT-PCR and western blot. Microglia was cultured to study the effect of DI on microglial inflammatory response and Nrf2 level.Results: DI reduced the secretion of inflammatory cytokines in DRGs, spinal cord and hind paw tissues, suppressed the activation of glial cells (like microglia and astrocytes) in spina dorsal horn and infiltration of inflammatory cells in the hind paw tissues, and decreased the phosphorylation of ERK1/2, but boosted Nrf2 levels in the DRGs and spinal dorsal horn. Similarly, administration of DI potently reversed the LPS-induced inflammatory effect in the microglia. Reduction of endogenous itaconate pretreated with irg1 siRNA prevented the expression of Nrf2, which did not exert the analgesic and anti-inflammatory effects in vitro experiment. Conclusions: Our findings verify for the first time that DI alleviates painful behaviors and suppresses neuroinflammation in neuropathic pain and inflammatory pain models through molecular mechanisms, suggesting that DI elicits sustained chronic pain relief, which will be regarded as a novel therapeutic agent for chronic pain treatment.


Introduction
Chronic pain is a kind of intractable pain, which persists for a long period, even though the initial tissue damage is cured (1,2). Chronic pain often leads to multiple complications and is a major clinical problem (1). It is characterized by spontaneous pain (such as burning) and evoked pain to noxious (hyperalgesia) or non-noxious (allodynia) stimuli (3) and affects up to 30% of adults worldwide (4) with an annual economic burden of treatment exceeding $600 billion in the USA (4,5). Due to the complex pathogenesis, currently available treatments have little effect on chronic pain (6). Growing body of evidence demonstrates that the occurrence of neuroin ammation plays an indispensable role in the development and maintenance of chronic pain (7). Its characteristic features involve the activation of glial cells (e.g. microglia and astrocytes), activation and in ltration of leukocytes, and the release of a wide range of in ammatory mediators, such as interleukin (IL)-1β, IL-6, nitric oxide (NO), tumor necrosis factor (TNF)-α both in dorsal root ganglion (DRGs) and spinal cord. Activation of glial cells and immune cells leads to the development of peripheral and central sensitization (2,(8)(9)(10)(11), which is closely associated with the establishment and maintenance of chronic pain (12,13). Those in ammatory factors, as powerful neuromodulators, can regulate synaptic transmission and plasticity (14) and play an adequate role in inducing hyperalgesia and allodynia in the peripheral nervous system (PNS) and central nervous system (CNS) (3). Given the lack of effective treatments for chronic pain, interventions targeting the neuroin ammatory cascades are considered as promising therapeutic strategies.
In the present study, we con rmed the analgesic effect of itaconate in chronic pain model. We reported for the rst time that intraperitoneal (i.p.) administration of DI ameliorated mechanical allodynia and thermal hyperalgesia in neuropathic pain induced by SNL and in ammatory pain induced by Complete Freund's adjuvant (CFA) in mice. In addition, we identi ed that DI not only suppressed the activation of microglia and astrocytes in the spinal cord, but also inhibited the in ltration of in ammatory cells in the hind paws. Besides, DI blocked the phosphorylation of ERK1/2, enhanced the Nrf2 level and suppressed the production of in ammatory cytokines in DRGs, spinal cord and plantar tissues. In vitro experiments also showed that DI exerted its neuroprotective effect.

Animals.
Male C57BL/6 mice aged 6-8weeks were purchased from Nanjing Institute of Biomedicine and adapted to the study habitat for one week prior to experiment. All animals were subjected to a 12 h∶12 h light∶dark cycle and fed with free accessed food and water. These mice are randomly divided into groups.

Reagents
DI (Cat.no.592498, Sigma, USA), Lipopolysaccharide (LPS) (Cat.no.L2630, Sigma, USA) and CFA (Cat.no.F5881, Sigma, USA) were purchased from Sigma-Aldrich. DI was dissolved in PBS to nal concentrations of 10 mg/500 µl and 20 mg/500 µl for in vivo experiment. A volume of 500 µl DI solution was injected into the mice via intraperitoneal route according to previously reported methods (28,29). DI was dissolved in dimethyl Sulfoxide (DMSO) and diluted to a concentration of 250 µM for in vitro experiment. LPS was diluted in PBS to a concentration of 100 ng/mL.
3. Neuropathic pain model L4 spinal nerve ligation (SNL) model, as a kind of neuropathic pain model, was induced according to previously reported methods (30). The surgery was done on mice after deeply anesthetized with 1% sodium pentobarbital. In SNL-operated group, isolate the left L4 spinal nerve and tightly ligate it with 6 − 0 silk thread using sterilized operating instruments (Fig. 1, A and B). In sham-operated mice, the left L4 spinal nerve was isolated without ligation. The skin and muscles were nally closed in layers.

CFA Model of In ammatory Pain
In ammatory pain model was established by CFA (20 µL) that was injected into the plantar surface of the left hind paw (31). The vehicle group received a same volume of saline injection into the same position (Fig. 1C).

Behavioral testing
Animals were allowed to habituate to the testing environment for 30min before testing. Mechanical and thermal pain tests were carried out as previously described (32). To test mechanical allodynia, paw withdrawal frequencies (PWF) in response to mechanical stimuli (von Frey laments) were measured according to the frequency of withdrawal response (the percentage of response) to two calibrated von Frey laments (0.07 g and 0.4 g, 10 trials of stimulation for each), which were used to stimulate the left hind paw for approximately 1s with at least 5 min interval between two adjacent trials. Paw withdrawal latencies (PWL) to noxious thermal stimulation was tested through the Hargreaves method (IITC Life Science). Withdrawal latency was recorded as the time from the start of irradiation to withdrawal of the tested hind paw. To avoid tissue damage, the cutoff time of heat stimulus was set at 20 sec. The average of three measurements that were repeated at a 15 min interval was regarded as the nal PWL for the mouse.
Blots were detected by chemiluminescence (ECL) and exposure to lm. The optical densities of protein bands were measured for grey scale densitometric analysis.

Real-time quantitative PCR (RT-PCR)
L4 DRGs and L4 spinal cord in SNL model and L4-5 spinal cord and left hind paw tissues in CFA model were collected for RNA extraction. The RNA was isolated using TRIzol reagent (Invitrogen, CA, USA), and the cDNA was synthesized from total RNA (1 µg) using 5 iScript Reaction Mix Reagents (Takara, Japan).

Histochemistry
Animals were anesthetized with 1% sodium pentobarbital and slowly perfused with PBS through the ascending aorta, then rapidly perfused with 4% paraformaldehyde. The spinal cord segments (L4) and the left paw tissues were removed and post xed in 4% paraformaldehyde for 24 h. Subsequently the spinal cord was allowed to equilibrate in 30% sucrose in PBS at 4°C overnight. The spinal cord sections were cut and processed for immuno uorescence. The sections were incubated in PBS containing 4% normal goat serum and 0.3% TritonX-100 at room temperature for 1 h. The samples were incubated with primary antibodies: mouse GFAP (1:200, Millipore, USA) and rabbit IBA-1 (1:200, Wako, USA) at 4°C overnight, washed with PBS for 3 times, followed incubated by donkey anti-rabbit and anti-mouse secondary antibodies conjugated with Alexa Fluor 594 or 488 (1:500; Thermo Fisher Scienti c, Waltham, MA, USA) at room temperature for 1 h in the dark. The nuclei were stained by 4′,6-diamidino-2phenylindole (DAPI, Sigma-Aldrich, USA). Images were captured using a uorescent microscope (DMIRB, Leica, Germany). The left paw tissues were embedded in para n and sections (5µm) were stained by hematoxylin and eosin (H&E), and visualized using a light microscope. Eagle Medium/F12 (HyClone™) supplemented with 10% fetal bovine serum (FBS, Gibco, USA), 100 U/ml penicillin, and 100 µg/ml streptomycin in a 5% CO 2 atmosphere at 37°C. BV 2 cells were used under passage 20. After 12 h of treatment with DI followed by the stimulation of LPS for 2h, BV 2 were collected and centrifuged at 5000 rpm for 5 min. Cell pellets were then used in follow-up experiment. 10. siRNA transfection of microglia Cells were planted at 2×10 5 cells per ml in 6-well plates to incubate at 37° C in a CO 2 incubator until the cells are 60-80% con uent. SiRNA was dissolved in RNase-free water at 10 µM as a stock solution. Then, Irg1 siRNA (sc-146287, Santa Cruz Biotechnology, USA) and control siRNA (sc-37007, Santa Cruz Biotechnology, USA) were separately mixed with the transfection reagent (sc-29528, Santa Cruz Biotechnology, USA) to incubate the cells 5-7 hours at 37° C in a CO 2 incubator. Remove the transfection mixture and replace with 1x normal growth medium. Incubate the cells for an additional 18-24 hours.
All data were expressed as mean ± SEM. Two-way repeated-measures ANOVA followed by Tukey's posthoc test was used for analysis of behavioral data. One-way ANOVA was used for the statistical analyses in other tests. A P value of less than 0.05 was considered statistically signi cant.

DI alleviated neuropathic pain after SNL.
The mouse SNL model mimics neuropathic pain induced by nerve injury in the clinical setting (33). In this model, neuropathic pain symptoms such as mechanical allodynia and heat hyperalgesia sustained for many weeks (34). To assess the effect of DI on behavioral signs of neuropathic pain, we continuously administered DI (10 mg or 20 mg) once daily by i.p. injection from day 4 till day 32 after SNL. PWF and PWL was evaluated before surgery (baseline), and on day 1, 3, 5, 7, 11, 18, 25 and 32 day after SNL (Figure 2A). We found that both doses of DI produced an increasing inhibition of SNL-induced mechanical allodynia and thermal hyperalgesia starting from 1 day after injection (Figure 2, B-D).
Notably, this antinociceptive effect remained during the 4 weeks period of DI treatment without signs of tolerance. There were no differences in among groups in either mechanical or heat thresholds on the contralateral side of SNL mice ( Figure S1, A-C).

Effects of pretreatment of DI on SNL-induced neuropathic Pain Behavior.
We next examined whether DI pretreatment prevents neuropathic pain. As shown in Figure 3A, two different doses of DI were injected i.p. once daily for consecutive 3 days before and after SNL model. Only the higher dose of DI (20 mg) reversed mechanical allodynia and thermal hyperalgesia during the administration. However, the analgesic effect of 20 mg DI gradually vanished once DI administration was ceased (Figure 3, B-D). Interestingly, the recovery of daily DI supply on day 7 to day 10 after SNL again resulted in the relief of neuropathic pain in a dose-dependent manner, as indicated by the ndings that 10 mg DI only reduces PWF to the stimulation of 0.4 g and PWL to thermal stimulation, whereas 20 mg DI relieved all these neuropathic pain symptoms (Figure 3, B-D). The absence of out-lasting analgesic effect may be related to the rapid metabolism of itaconate (35,36), which requires DI to be provided continuously. The mechanical and heat thresholds in the contralateral side were not altered ( Figure S2, A-Page 7/23 C). The above results demonstrated that DI can prevent and relieve neuropathic pain when given before and after nerve injury, respectively.

DI regulated neuroin ammation in DRGs and spinal dorsal horn of SNL model mice.
Neuroin ammation in the PNS and CNS has been strongly involved in the development and maintenance of neuropathic pain (37). The L4 DRGs and spinal dorsal horn were harvested on day 10 after DI injection ( Figure 4A). Quantitative PCR (qPCR) revealed that SNL led to upregulation of Il1b and downregulation of Il10 in the ipsilateral L4 DRGs. The administration of DI reversed the expression of these factors (Figure 4, Itaconate contains an electrophilic α, β-unsaturated carboxylic acids, which may alkylate protein cysteine residues through a Michael addition reaction to form 2,3-dicarboxypropyl adducts. The cysteine alkylated protein KEAP1 usually binds to Nrf2 and promotes its degradation, but the alkylation of KEAP1 cysteine residue allows Nrf2 to activate a transcriptional antioxidants and anti-in ammatory programs (18, 40). Besides, it has been shown that itaconate is required for the activation of anti-in ammatory transcription factor Nrf2 by LPS in mouse and human macrophages (28). Therefore, we further examined whether DI affects Nrf2 expression in SNL model. It was found that SNL model blocked the Nrf2 expression in the DRGs and spinal dorsal horn, but the treatment of DI promoted the production of Nrf2 in a dosedependent manner in SNL mice ( Figure 5, F-I). SNL surgery did not change the ERK1/2 phosphorylation and Nrf2 expression in the contralateral spinal dorsal horn ( Figure S3, A-D). However, DI boosted levels of Nrf2 protein in the contralateral spinal dorsal horn ( Figure S3, B and D). These results demonstrated that DI suppressed the phosphorylation of ERK1/2 and in uenced the Nrf2 expression in the DRGs and spinal dorsal horn in SNL model.

DI inhibited in ammatory cytokine production and enhanced Nrf2 expression in cultured microglia.
The regulatory effects of itaconate on peripheral macrophages have been well established (28), there are few studies of its effect on microglia. Microglia, as resident macrophage in the CNS, composes the innate immune system of the CNS and is a key cellular mediators of neuroin ammatory processes (41). In Figure 4D, it has been certi ed that DI can inhibit microglial activation in the spinal dorsal horn. However, the anti-in ammatory consequence of this inhibition requires veri cation. Therefore, we did in vitro experiments on microglial cell line BV 2 . Firstly, microglial cells (1 × 10 6 cells) pretreated with DI (250 μM) for 12 h were stimulated by LPS (100 ng/ml) for 2 h, then the cells were collected for RT-PCR and western blot ( Figure 6A). As demonstrated in Figure 6B and C, DI signi cantly reduced the contents of Il1b and tnf mRNA in BV 2 cells exposed to LPS. Besides, we examined the expression of Nrf2 in BV 2 cells. The stimulation of LPS or exposure of DI alone improved the level of Nrf2 of microglia, but the coadministration of LPS and DI promoted more production of Nrf2 (Figure 6, D and E).
To further verify the role of Nrf2 in the inhibition of microglia activation by DI, we treated BV 2 cells with a speci c irg1 siRNA (4 μl) for 6 h. It has been shown that suppression of IRG1 expression that could inhibited the production of itaconate (28). This treatment resulted in a 67% reduction in irg1 production ( Figure 6F), which led to a reduction in Nrf2 expression in BV 2 cells compared with control siRNA (con-siRNA) treatment ( Figure 6, G and H). The in ammatory factors Il1b and tnf were increased after the treatment of irg1 siRNA, which was reversed by DI, although there was no difference between LPS plus irg1-siRNA group and LPS plus irg1-siRNA plus DI group, which demonstrated that exogenous itaconate derivative like DI, may have a certain effect on endogenous itaconate, even though DI could not simulate the endogenous itaconate. In addition, there was signi cant difference between LPS plus con-siRNA group and LPS plus con-siRNA plus DI group (Figure 6, I and J).
In addition to neuropathic pain model, we veri ed the analgesic e cacy of DI (20 mg) in the in ammatory pain model that was induced by CFA (20μL) via the plantar injection. Pain behavior was evaluated at baseline, 2h, 6h, 1, 2, 4, 6, 8 and 12 day after CFA injection ( Figure 7A). In this model, mechanical allodynia and thermal hyperalgesia were rapidly induced within 2 h and peaks at 1 day and began to decrease from day 2 after CFA in mice (Figure 7, B-D). The continuous administration of DI (20 mg), given day 1 after CFA, produced an inhibition of CFA-induced mechanical and thermal painful behaviors (Figure 7, B-D). The sensitivity to mechanical and thermal stimulations on the contralateral side were not altered ( Figure S4, A-C). These results demonstrated that the analgesic effect of DI is evident not only in the neuropathic pain but also the in ammatory pain.
7. DI inhibited in ammation response in the spinal cord and peripheral tissues of CFA model mice.
In ammatory pain occurs during in ammatory or immune responses. The pro-in ammatory mediators (e.g., cytokines) modulate pain sensitivity, and the accumulation of in ltrated cells induces the development of peripheral and central sensitization in the chronic pain conditions (42). PCR results revealed that CFA also caused upregulation of Il1b and tnf in the ipsilateral hind paw and L4-L5 spinal dorsal horn, and downregulation of Il10 in the ipsilateral hind paw (Figure 8, A-E). The expression changes of these mRNAs were reserved by treatment with DI (Figure 8, A-E). H&E staining result showed that the vehicle group had normal paw tissue, whereas the hind paws that received CFA injection showed massive accumulation of in ltrated cells ( Figure 8F). However, in ltrated in ammatory cell was evidently decreased by DI treatment (20 mg) on CFA day 5. This result veri ed that DI also inhibited in ammation response in the spinal cord and hind paw tissue of CFA model mice.

Discussion
Itaconate, as a metabolite, is generated by mitochondria-associated enzyme IRG1, has recently been regarded as a regulator of macrophage activation (19,29). It has been shown that itaconate exerts an anti-in ammatory effect through promoting Nrf2 or suppressing SDH which exerts anti-oxidant, antiin ammatory and immunomodulatory effects (18). A number of studies reported the therapeutic effect with DI administration in rodent no matter it is systemic or local injection, but the analgesic effects are not clear. Therefore, in this study, our ndings rstly demonstrate that a continuous i.p. injection of different doses of DI (10mg or 20mg), either at pre-SNL time points (3 days before SNL) or at post-SNL time points (28 days after SNL), produced relief from neuropathic pain symptoms, including mechanical allodynia and heat hyperalgesia in mice. However, different administration time of DI leads to different pain relief effects. In Fig. 2, a continuous i.p. injection of DI, given starting from day 4 to 32 after SNL, produced long-term relief (> 4 weeks) of SNL-induced painful behaviors following either low dose (10mg) or high dose (20mg), which has the similar result with the in ammatory pain model caused by CFA. In previous studies, in addition to postoperative administration, DI can also be administered before the establishment of the model (24). Therefore, we tried continuous administration of DI from 3 days before the model establishment until the 3 days after SNL surgery, and found that only large doses of DI (20mg) slightly alleviated mechanical and thermal pain caused by SNL. The analgesic effect of DI would disappear after the drug was stopped. Then SNL-induced neuropathic pain was reversed when DI was provided again, as seen in Fig. 3, which this difference may be closely related to the rapid metabolism of itaconate (17).
DI treatment reduced central and peripheral in ammation response in spinal cord, DRGs and hind paws tissues, and also suppressed the phosphorylation of ERK1/2, as an important cellular molecule related to pain sensitization, in the DRGs and spinal dorsal horn. Our results demonstrate the immunomodulatory effect of DI on SNL/CFA-induced pain. Besides, LPS to stimulate the microglia mimics the pathological change in vivo. This is consistent with in vivo experiment that DI exhibited reduced in ammatory cytokines.
In this study, we found that SNL inhibited the expression of Nrf2 in the DRGs and spinal dorsal horn, which has the same result with that of Wei Di (43). But in vitro experiment, it was shown that the expression of Nrf2 was upregulated after LPS simulation, which demonstrated that after the initial injury, the body produced anti-in ammatory or analgesic substances to protect itself from injury, such as increased Nrf2 in cells, but this increase cannot completely eliminate the injury, which need to be provided exogenously. The production of Nrf2 pretreated with irg1 siRNA were compromised in the microglia. The administration of exogenous DI potently reversed in ammatory response induced by LPS. Nrf2 is a key transcription factor in the regulation of multiple antioxidants (44). Kawamura et al. have proved that inhaled hydrogen reduced lung injury by activating Nrf2 mediates oxidative stress response mechanisms (45,46). Studies have demonstrated that activation of the Nrf2 protected against in ammation-mediated tissue damage by inhibiting the production of proin ammatory mediators (47,48 vivo. In addition to Nrf2, whether the inhibition of SDH is also involved in the analgesic effect of DI, which need to be further explored. Apart from DI, 4-octyl itaconate (4-OI) is regarded as another common derivative of itaconate. It was found that DI degrades rapidly in cells, without causing the release of itaconate, indicating that it is unlikely to mimic endogenous itaconate. OI has been determined to form endogenous itaconate, making it a suitable cell-permeable itaconate surrogate. Two types of itaconate derivatives both can effectively activate Nrf2 signaling to exert their anti-in ammatory and antioxidant properties (17). Its toxicity is likely to be very low, given that itaconate is produced in the course of a natural immune response, (52). Therefore, modifying of itaconate derivatives to delay the rate of metabolism in the body and prolong its therapeutic effect or endogenous form (for example, by boosting IRG1 activity) may bring good news to clinical patients in the future.

Conclusions
In summary, we reported for the rst time that continuously i.p. injected DI not only produce neuropathic pain relief in mice and ameliorated in ammatory pain. In addition, we identi ed that DI suppressed neuroin ammation and the phosphorylation of ERK1/2 in the peripheral and central tissues, and ameliorated in ammatory response in the microglia. DI also promoted the Nrf2 expression, suggesting that DI elicits sustained chronic pain relief, which will be developed as a novel therapeutic agent for the treatment of chronic pain. Declarations contributed to study discussion and manuscript editing. MY contributed to the assessment and write the manuscript. All authors read and approved the nal manuscript.

Funding
This work is supported by National Natural Science Foundation of China (82071227 and 81771189) and partly supported by the and the Natural Science Foundation of Zhejiang Province (LQ19C090006, LY18H090006, LZ18H090001).

Availability of data and materials
The datasets of the current study are available from the corresponding author on a reasonable request.

Ethics approval and consent to participate
This study was carried out in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals. All experimental procedures were approved by the Animal Research Committee of the Second A liated Hospital, School of Medicine, Zhejiang University (Hangzhou, China).

Consent for publication
Not applicable. The in ammatory pain model cartoon caused by CFA for this study. The left hind paw of mice was injected by CFA.