Mechanism and effects of STING–IFN-I pathway on nociception: A narrative review

Since the discovery of STING in 2008, numerous studies have investigated its functions in immunity, inflammation, and cancer. STING activates downstream molecules including IFN-I, NLRP3, and NF-κB. The STING–IFN-I pathway plays a vital role in nociception. After receiving the upstream signal, STING is activated and induces the expression of IFN-I, and after paracrine and autocrine signaling, IFN-I binds to IFN receptors. Subsequently, the activity of ion channels is inhibited by TYK2, which induces an acute antinociceptive effect. JAK activates PIK3 and MAPK–MNK–eIF4E pathways, which sensitize nociceptors in the peripheral nervous system. In the mid-late stage, the STING–IFN-I pathway activates STAT, increases pro-inflammatory and anti-inflammatory cytokines, inhibits ER-phagy, and promotes microglial M1-polarization in the central nervous system, leading to central sensitization. Thus, the STING–IFN-I pathway may exert complex effects on nociception at various stages, and these effects require further comprehensive elucidation. Therefore, in this review, we systematically summarized the mechanisms of the STING–IFN-I pathway and discussed its function in nociception.


Introduction
Pain is defined by the International Association for the Study of Pain as an unpleasant sensory and emotional experience associated with actual or potential tissue or other damage (Merskey and Spear, 1967). After sensing physical and chemical stimuli, nociceptors produce and transmit information to the central nervous system (CNS). Notably, the brain can generate pain without a message from nociceptors or the spinal cord, such as in phantom limb pain (Loeser and Melzack, 1999;Julius and Basbaum, 2001). Multiple molecules are involved in the production of pain, such as G proteincoupled receptors, cyclic nucleotides, capsaicin, and acid (Julius and Basbaum, 2001).
Stimulator of interferon genes (also called STING, MITA, MPYS, ERIS, and TMEM173) was first discovered in 2008 (Ishikawa and Barber, 2008;Zhong et al., 2008). STING could regulate antimicrobial response, autoimmune disease, and cancer progression (Ahn et al., 2012;Li et al., 2013;Zheng et al., 2020). The stimulator of interferon genes (STING)interferon-I (IFN-I) pathway can control nociception Wang et al., 2021). In neuropathic pain models including bone cancer pain, chemotherapy-induced peripheral neuropathy, and nerve injury, administration of STING agonists activates STING, increases the expression of IFN-I, and inhibits the excitability of nociceptors in the peripheral nervous system (PNS) . These effects induce transient, short-term, and dose-dependent antinociception at an early stage . However, the antinociceptive effect was not substantial 11 days after the injection . Similarly, activation of the STING-IFN-I pathway induces nociception or neuropathic pain at a late stage (Liu et al., 2022;Wu et al., 2022). The exact effects of STING-IFN-I remain controversial, and its differential role in different sexes, neuropathic pain models, cells, and stages requires further research. Previous studies hypothesized that this pathway might be a potential therapeutic target for pain management.
In this review, we systematically summarize the mechanisms of the STING-IFN-I pathway and discuss its function in nociception.

Research progress of STING-IFN-I pathway Antimicrobial response
Microbial DNA invasion triggers a series of immune responses. STING is essential for detecting exogenous microbial DNA . Activation of STING consequently activates the transcription factors NF-κB and interferon regulatory factor 3 (IRF3) to induce cytokines and IFN-I expression (Ishikawa and Barber, 2011). STING is required by fibroblasts, macrophages, dendritic cells, and myeloid cells to induce IFN-I production against vaccinia virus (VACV), cytomegalovirus (HCMV), baculovirus, several strains of herpes simplex virus-1 (HSV1), and Listeria monocytogenes (Ishikawa and Barber, 2008;Ishikawa et al., 2009).

Autoimmune disease
In addition to exogenous DNA, STING can detect self-DNA. Undigested DNA from apoptotic cells triggers DNA sensors, which increase the expression of cytokines and result in autoimmune diseases (Nagata, 2010;Ahn et al., 2012). The exonuclease, three prime repair exonuclease 1 (TREX1), degrades cytosolic DNA (Mazur and Perrino, 1999;Crow et al., 2006) and its deficiency leads to multiple inflammatory and autoimmune diseases such as systemic lupus erythematosus, Aicardi-Goutieres syndrome, and familial chilblain lupus (Rice et al., 2015). In a TREX1deficient rat model, cGAS activated STING through cGAMP production and mediated inflammatory disease and death in mice (Gao et al., 2015). Similarly, STING triggered by apoptotic or necrotic DNA promoted the expression of cytokines, whereas its deficiency abrogated the production of cytokines activated by self-DNA in a DNase II-deficient model (Ahn et al., 2012).

Cancer progression
Nuclear and mitochondrial DNA are easily damaged in tumor cells, inducing IFN-I through the cGAS-STING-IRF3dependent pathway (Woo et al., 2014;Chen Y. A.et al., 2017;Mackenzie et al., 2017). IFN-I is a mediator of STING and exerts adaptive antitumor effects (Zheng et al., 2020). It can promote cross-presentation by stimulating the maturation of DCs, slowing down the endosome-lysosome acidification process to prevent phagocytic tumor antigen clearance, and increasing the expression of cell surface MHC I molecules, which accelerates DC migration to lymph nodes to crosstrigger tumor-specific CD8 + T cells (Reboulet et al., 2010;Diamond et al., 2011;Lorenzi et al., 2011;Zheng et al., 2020). In addition, IFN-I can induce the expression of multiple chemokines (Padovan et al., 2002;Takashima et al., 2016). For instance, CXCL9 and CXCL10 are involved in cytotoxic T lymphocyte transfer and infiltration, whereas CCL5 and CXCL10 promote the recruitment and activation of NK cells and T cells in tumors (Padovan et al., 2002;Takashima et al., 2016). By contrast, the cGAS-STING pathway can induce the senescence-associated secretory phenotype (SASP) (Loo et al., 2020). The SASP factor induces immune surveillance and acts as a tumor suppressor. However, continuous exposure to SASP may cause tissue damage and chronic inflammation associated with tumor growth (Loo et al., 2020). Nevertheless, long-term activation of STING may promote tumor growth and metastasis, and this effect is associated with tumor stage, CIN status, and the degree of STING activation (Zheng et al., 2020). STING agonists including cyclic dinucleotides and their derivatives, DMXAA and its analogs, and small-molecule agonists are widely studied as cancer treatment agents (Corrales and Gajewski, 2015;Ablasser and Chen, 2019;Zheng et al., 2020).

Mechanism of STING-IFN-I pathway with respect to pain Peripheral nociceptors and pain
Cell bodies of nociceptors are distributed in the dorsal root ganglia (DRG) and trigeminal ganglion (Basbaum et al., 2009). Most nociceptors contain unmyelinated C fibers (Woolf and Ma, 2007). However, initial and acute pain is mediated by nociceptors with A fibers (Djouhri and Lawson, 2004). After sensing physical and chemical stimuli, peripheral nociceptors are activated to produce pain through different signal transduction pathways (Julius and Basbaum, 2001;Donnelly et al., 2020). Particularly, TRP channels recognize noxious heat, and the ENaC/DEG channel family senses mechanical stimuli (Lingueglia et al., 1997;Tominaga et al., 1998). Nociceptors can convert receptor potentials into action potentials through voltage-gated channels (including sodium, calcium, and potassium channels) (Dubin and Patapoutian, 2010). Primary nociceptors transmit noxious stimuli to projection neurons located in the cornu dorsalis medullae spinalis (Basbaum et al., 2009). Harmful information is transmitted to the somatosensory cortex through the thalamus, indicating the location and intensity of the pain (Basbaum et al., 2009). Other projection neurons contact the cingulate gyrus and insular cortex through the brain stem and amygdala, forming emotional elements of pain experiences (Basbaum et al., 2009).

Downstream signals of STING
After binding with CDNs, STING is transferred from the ER to the Golgi complex via perinuclear vesicles (Ishikawa et al., 2009). STING forms oligomers in the ER-Golgi membrane and exposes its C terminal (Tanaka and Chen, 2012;Ergun et al., 2019). The C terminal of STING recruits TANK-binding kinase 1 (TBK1), and the STING dimer accesses the active site of TBK1 for its phosphorylation (Barber, 2011;Zhang et al., 2019). In addition, two TBK1 dimers can be mutually activated via transautophosphorylation (Zhang et al., 2019). The phosphorylated tail of STING recruits IRF3 and transports it to TBK1 for phosphorylation (Tanaka and Chen, 2012;Zhang et al., 2019). Notably, the interaction between STING and TBK1 enhances the binding of TBK1 and IRF3 (Zhong et al., 2008;Tanaka and Chen, 2012;Zhang et al., 2019). Phosphorylated IRF3 dimers access the nucleus and activate the transcription of IFN and inflammatory factor genes (Liu et al., 2015;. Thus, IFN-I synthesis increases notably. After paracrine and autocrine signaling, IFN-I binds with IFN receptors on sensory neurons to generate nociceptive effects Tan et al., 2021). IFN receptors comprise IFNAR1 and IFNAR2 . IFNAR1 plays a vital role in acute nociceptive functions Wang et al., 2021). Inhibition of tyrosine kinase (TYK2) eliminates the analgesic effect of IFN-β . Thus, TYK2 associated with IFNAR1 inhibits the activity of sodium (Na v 1.7) and calcium channels Tan et al., 2021). Furthermore, the low excitability of nociceptors is attributed to the loss of function of sodium (Na v 1.7) and calcium channels (Binshtok et al., 2007;McDermott et al., 2019). Therefore, IFN-I can induce acute and short-term antinociception via TYK2.

Regulatory mechanism of STING-IFN-I pathway
Regulation of the STING-IFN-I pathway mostly depends on STING activity. Posttranslational modifications including phosphorylation, ubiquitination, and palmitoylation play Mechanism of the STING-IFN-I pathway in nociception. After detecting both foreign and autologous cytosolic DNAs, STING exposes its C terminal and recruits TBK1. Activated STING transfers from the ER to the Golgi complex and recruits IRF3. IRF3 phosphorylated by TBK1 forms dimers and accesses the nucleus to activate the transcription of IFN. Secretion of IFN-I increases. Then, IFN-I binds with IFN receptors. TYK2 associated with IFNAR1 promotes the acute antinociceptive effect. JAK associated with IFNAR2 activates the MAPK-MNK-eIF4E pathway and PIK3 at a later stage, causing nociceptor sensitization in DRG. In the mid-late stage, IFN-I induces a delayed effect in the CNS.
vital roles in regulating STING activity . Activation of STING requires palmitoylation in the Golgi complex (Mukai et al., 2016;Haag et al., 2018). After recruiting TBK1 and activating IRF3, negative feedback is triggered. STING is subsequently phosphorylated by serine/threonine UNC-51-like kinase, and IRF3 activity is inhibited (Konno et al., 2013).
Among posttranslational modifications, ubiquitination is essential for STING activity. Some molecules have been shown to play essential roles in STING regulation. AMFR facilitates K27-linked polyubiquitination of STING through the ER membrane protein INSIG1 and promotes the recruitment and activation of TBK1 . EIF3S5, OTUD5, CYLD, and ubiquitin-specific protease (USP) 44 (USP44) are deubiquitinases that remove K48-linked polyubiquitination to maintain the stability of STING (Luo et al., 2016;Zhang et al., 2018;Zhang H. Y. et al., 2020;Guo et al., 2021). TRIM32 promotes K63-linked polyubiquitination of STING and increases the production of IFN-I (Cui et al., 2017). iRhom2 recruits the translocon-associated protein (TRAPβ) and the deubiquitination enzyme (EIF3S5) to promote STING trafficking from the ER to perinuclear microsomes (Luo et al., 2016). USP13 deconjugates polyubiquitin chains on STING to prevent recruitment of TBK1 (Sun et al., 2017), while USP21 hydrolyzes the K27/63-linked polyubiquitin chain on STING to negatively regulate the production of IFN-I (Table 1; Chen Y. et al., 2017). The regulatory mechanism of STING is complex and warrants further investigation.

Effects of STING-IFN-I pathway on nociception
Limited studies indicated that the STING-IFN-I pathway has dual effects on nociception ( Table 2).

Positive effect
The STING-IFN-I pathway is associated with acute antinociceptive effects. In a rat model, deficiency of the STING-IFN-I pathway increased the excitability of nociceptors . In the chronic constriction injury model of rats, knockdown of the D-type protein tyrosine phosphatase receptor increased the expression of STING and IFN-α, which attenuated pain (Sun et al., 2022). STING agonists can relieve neuropathic pain in peripheral neuropathy induced by paclitaxel chemotherapy  and pain induced by nerve injury . Similarly, they may inhibit bone cancer pain and maintain motor function by reducing tumor burden and inhibiting cancer-induced osteoclast generation Wang et al., 2021). Moreover, STING agonists can attenuate fracture-induced pain in tumor-free mice . Notably, after injecting STING agonists in rats, IFN-I levels in serum, DRG tissues, and bone marrow lysates were significantly upregulated 1000-fold in 4 h and maintained for up to 24 h Wang et al., 2021;Sun et al., 2022). Therefore, the STING-IFN-I pathway may promote short-term antinociception.

Negative effect
Several studies have also reported contradictory results, wherein the STING-IFN-I pathway exerted negative effects. A case series reported that IFN-I induced by STING causes neuropathic pain in young patients (Papa et al., 2021). Similarly, in patients with hepatitis C virus infection, the use of IFN-α leads to somatic pain (Lin et al., 2020).
In the spared nerve injury (SNI) model, inhibiting the cGAS-STING pathway can restrain microglial M1-polarization and attenuate neuropathic pain . M1polarization microglia express CD16 and induce TNF-α and IL-1β synthesis, which may cause central sensitization (Mesquida-Veny et al., 2021). In a rat SNL model, ketamine and dexmedetomidine induced ER-phagy and alleviated ER stress to provide antianxiety and antinociceptive effects by inhibiting the STING-TBK pathway in the spinal cord (Liu et al., 2022).

Underlying reasons for the dual effects
There are several possible explanations for these contradictions. First, the sex of the animals may have caused this discrepancy. This pathway more likely has a negative effect on male rats . Second, animal experiments were used to create different neuropathic pain models to explore its effects. However, different animal models may exhibit various neuropathies. Third, different injection methods may also have caused bias. Peripheral administration of IFN-I induced pain behavior in rat models (Barragán-Iglesias et al., 2020). However, intrathecal injection of IFN-α inhibited mechanical hypersensitivity caused by intraplantar . Fourth, the different effective times influenced the results. Short-term activation of this pathway led to transient and acute antinociception, which was maintained for up to 24 h . However, consecutive and repeated administration of STING agonists caused central sensitization and nociception . Fifth, the STING-IFN-I pathway does not participate in the physiological regulation of pain sensitivity and is only involved in the regulation of pain after nerve injury (Sun et al., 2022;Wu et al., 2022). Therefore, observing a positive effect in normal rat models injected with STING agonists or IFN-I is challenging. Finally, the STING-IFN-I pathway may play distinct roles in different parts of the PNS and CNS. A study has demonstrated that after STING agonist DMXAA treatment in mouse models, bone cancer-induced cold and

FIGURE 3
Different roles of the STING-IFN-I pathway in the peripheral nerve system (PNS) and central nerve system (CNS). In the PNS, the pathway inhibits the excitability of nociceptors at an early stage. Subsequently, it may cause nociceptor sensitization. At the mid-late stage, it plays a dominant role in the CNS, which leads to central sensitization. Under normal physiological conditions, antinociception and nociception of the STING-IFN-I pathway maintain homeostasis. A nerve injury disrupts the balance and leads to antinociception or nociception. mechanical allodynia were reduced at an early stage but not at the mid-late stage . Therefore, at an early stage, it induces antinociception and reduces pain by restraining the activity of ion channels and the excitability of nociceptors in the PNS. Subsequently, this pathway may induce nociceptor sensitization via the MAPK-MNK-eIF4E pathway and PIK3. At the mid-late stage, it can cause central sensitization in several ways (Figure 3).

STING agonists compared with opiates
Since the isolation of morphine in 1805, opioids have been widely used for pain management (Pasternak, 2014). Opioids, including morphine, interact with µ, κ, and δ receptors to produce analgesic effects, respiratory depression, and euphoria addiction. After binding with opioid receptors, opioids cause antinociception through the same mechanism as enkephalin, which involves hyperpolarization of interneurons and reduction of transmitters associated with pain (Haigler, 1987;Lipp, 1991). In addition, morphine can react with opioid receptors in supraspinal structures to activate the supraspinal system (Lipp, 1991). By contrast, STING agonists produce acute and short-term antinociception via the STING-IFN-I pathway in the PNS. Furthermore, opioids are highly addictive, which is caused by a reduction in the inhibitory function of GABAergic synapses in the neurons of the central amygdala and brain reward/motivational mesocorticolimbic circuitry (Navratilova and Porreca, 2014;Zhang et al., 2014). In contrast, the repeated use of STING agonists does not cause addiction and attenuates SNI-induced astrogliosis . In nonhuman primates, intrathecal administration of STING agonists produces longer lasting analgesic effects at lower doses than morphine (3 vs. 100 nmol) (Sjöström et al., 1987;Donnelly et al., 2021;Wang et al., 2021). Naloxone, a nonselective and shortacting opioid receptor antagonist, can reverse the analgesic effect of morphine (Drug and the Therapeutics Bulletin, 1981;van Dorp et al., 2007). By contrast, STING agonist-mediated analgesia is not affected by naloxone .
Previous studies have suggested that STING agonists have potential advantages including strong efficacy at low doses, a longer lasting effect, and non-addictive. However, the exact effects of STING-IFN-I on nociception remain unclear and require further investigation.

Discussion
Apart from inducing antimicrobial response, mediating autoimmune disease, and regulating tumor growth, the STING-IFN-I pathway can induce acute antinociception for a short period of time. Therefore, the STING-IFN-I pathway may be a potential therapeutic target for pain management.
However, the effects of STING on nociception have several issues that need to be discussed. First, Donnelly et al. (2021) demonstrated that STING agonists reduced bone cancer pain.
However,  suggested that mitochondrial DNA triggers the STING pathway, leading to peripheral neuroinflammation and sensitization . In the early stage, the STING-IFN-I pathway was dominant, which reduced bone cancer pain. In the mid-late stage, the MAPK-MNK-eIF4E pathway was activated, and the STING-NF-κB pathway increased bone cancer pain via IL-1β, IL-6, and TNF-α ( Barragán-Iglesias et al., 2020;. In addition, STING agonists have been shown to reduce bone cancer pain through immune and neuronal modulation, reducing tumor burden and inhibiting osteoclastogenesis (Amouzegar et al., 2021;Wang et al., 2021). Therefore, it is difficult to determine the true effects of STING agonists in bone cancer pain models. Second, the STING-IFN-I pathway may influence central sensitization through ER-phagy and microglial M1-polarization. Further studies are needed to confirm this hypothesis and to determine how the STING-IFN-I pathway regulates ER-phagy and microglial M1-polarization. Third, previous studies have only discussed one downstream pathway in a neuropathic pain model. However, STING has various downstream signaling components, including IFN-I, NF-κB, and NLPRS. Studies that include all downstream signals of STING are still lacking. Lastly, STING-IFN-I exists not only in peripheral and central neurons but also in immune cells. Whether STING agonists interact with these cell types to cause nociception requires further studies .
The current clinical use of STING agonists focuses on cancer immunotherapy. Several combination therapies are currently available in clinical trials (Zheng et al., 2020). Few studies have indicated the effectiveness of STING in nociception; however, the use of STING for nociception remains controversial and warrants further extensive and comprehensive studies.

Conclusion
At an early stage, the STING-IFN-I pathway can induce short-term antinociceptive effects by activating TYK2, restraining the activity of calcium and sodium channels, and inhibiting the excitability of nociceptors in the PNS. Subsequently, it activates the JAK-MAPK-MNK-eIF4E pathway and PIK3, which cause nociceptor sensitization. At the mid-late stage, it promotes microglial M1-polarization, inhibits ER-phagy, activates STAT, and increases the expression of proinflammatory and anti-inflammatory cytokines in the CNS, which leads to central sensitization. Thus, the STING-IFN-I pathway at various stages has a dual effect on nociception.

Author contributions
JY wrote the manuscript and made illustrations. HD, BS, and YZ (Fifth author) provided advice for the manuscript. YZ (Fourth author) provided the supervision and comments on the manuscript. All the authors read and approved the final manuscript.

Funding
The work was supported by the National Natural Science Foundation of China (grant no. 81901144).