Utilizing multiple biologically based techniques and state-of-the-art in vivo non-traumatic approaches (e.g., positron emission tomography, magnetic resonance imaging scans, and magnetic resonance spectroscopy), astrocytes have been thoroughly studied, and their roles encompass a broad variety of dynamic physiological compositions and functionalized features: water and ion homeostasis, metabolic specialization, and brain oxidation (Volterra and Meldolesi, 2005; Haydon and Carmignoto, 2006; Romero-Sandoval et al., 2008; Fellin, 2009; Jiang et al., 2020).

Astrocytes play a role in neural tissue, closely combining with neural networks, and control the homeostatic balance between individual molecules in the CNS and individual tissues of the entire organ through the transportation of major ions and protons, clearance and decomposition of neurotransmitters, and release of neurotransmitter precursors and reactive oxygen scavengers. Astrocytes maintain neurotransmission as a consequence of providing neurotransmitter precursors for neurons and control cellular homeostasis through embryonic and adult neurogenesis. Moreover, regulating metabolic homeostasis by synthesizing glycogen and providing energy substrates for neurons is also a momentous routine of astrocytes. Disguised as glial cells, astrocytes constitute a cortical covering that envelops the CNS, manage the blood-brain barrier, and serve as chemoreceptors, thereby promoting dynamic homeostasis throughout the body. Traumatic brain injury, infection, and other diseases can increase astrocyte reactivity, thus transforming resting astrocytes into reactive astrocytes with abnormal characteristic behavior. Thus, astrocytes potentially represent the primary defense system of the CNS by increasing reactivity (Figure 1).

These routine structural, functional, and physiological strengths of astrocytes play crucial roles in various levels of CNS function, such as growth, experience-dependent alignment, and maturation (Verkhratsky and Nedergaard, 2018; Tsuda, 2019; Giovannoni and Quintana, 2020).

During the development of neurological diseases, astrocytes often exhibit different pathological manifestations successively or simultaneously, such as, dysfunctional astrocyte atrophy, pathological remodeling of astrocytes, and reactive astrocyte hyperplasia (Pekny et al., 2016). Disorders or loss of function may form the basis of various forms of neurological dysfunction and pathology, including, but not limited to, trauma, stroke, multiple sclerosis, and Alzheimer’s disease, among others (Weggen et al., 1998; Simpson et al., 2010).

In addition, they have been linked to a worse prognosis of CNS injury, such as excitatory toxic neurodegeneration due to inadequate glutamate intake (Bush et al., 1999; Swanson et al., 2004) or a higher risk of invasion or injury resulting from astrocyte barrier dysfunction (Bush et al., 1999; Faulkner et al., 2004; Drögemüller et al., 2008; Herrmann et al., 2008; Voskuhl et al., 2009). Extensive molecular and cellular biological evidence has currently demonstrated that the harmful effects of reactive astrogliosis and glial scarring are typically manifested by the inhibition of axon regeneration (Silver and Miller, 2004). Furthermore, astrocytes have appeared in a variety of domestic and international studies on different disease pathological states (Swanson et al., 2004; Brambilla et al., 2005; Jansen et al., 2005; Tian et al., 2005; Hamby et al., 2006; Farina et al., 2007; Brambilla et al., 2009). Some researchers have that found seizures in epilepsy may be directly related to the dysfunction and denaturation of astrocytes triggered by abnormalities in cytotoxic T lymphocytic autoimmunity (Whitney and McNamara, 2000; Bauer et al., 2007). Other autoimmune inflammatory states involving the CNS have also been reasonably speculated to be associated with astrocytes, such as systemic lupus erythematosus; however, they have not been explicitly studied (Tomita et al., 2004). Furthermore, astrocytes have now been firmly considered to make a significant contribution to the pathophysiology of hepatic encephalopathy (Norenberg et al., 2005).

Tsuda demonstrated that astrocyte proliferation in neuropathic pain is regulated by the JAK-STAT3 signaling pathway. Moreover, the application of the janus kinase (JAK) inhibitors AG490 and JAK inhibitor I to the spine resulted in the inhibition of astrocyte proliferation during astrocyte restriction and STAT3 activation. Furthermore, the inhibition of astrocyte proliferation by JAK-STAT3 signaling inhibitors potentially leads to the recovery of haptic pain (Tsuda et al., 2011). Therefore, it is reasonable to conclude that the inhibition of JAK-STAT3 signaling leads to that of astrocyte proliferation in the dorsal horn.

Some experiments have demonstrated that the proliferative capacity of astrocytes intervened by AG490 or with STAT3 defects is reduced to a certain extent (Washburn and Neary, 2006; Sarafian et al., 2010). Liu et al. (2021) demonstrated that not only were the activation levels of JAK2 and STAT3 significantly reduced by miR-135-5p mimics intervention but also GFAP, IL-1β, and TNF-α expression, which was consistently decreased. Moreover, functional in vitro studies have also confirmed that miR-135-5p potentially inhibits the JAK2-STAT3 signaling pathway in bone cancer pain, and histological staining has proven that miR-135-5p can directly interrupt the activity process of JAK2 and its upstream and downstream molecules in astrocytes (Liu et al., 2021). These studies demonstrate that JAK-STAT3 signaling has a key contribution in the occurrence and progression of neuropathic pain via the modulation of astrocyte activation.

Oligodendrocyte transcription factor 2 (Olig2) is an essential transcription factor required for the proliferation and derivation of developmental astrocytes. Olig2-labeled subsets of astrocytes have been detected in clusters and in greater numbers in certain fixed zones of the thalamus, midbrain, and spinal cord of rodents (Wang et al., 2021). By ablating Olig2, Chen found reactive astrocyte proliferation to be reduced, demonstrating that Olig2 plays a key role in reactive astrocyte proliferation (Chen et al., 2008). A BrdU labeling experiment suggested that Notch molecules may contribute to the numerical increase and functional specialization of reactive astrocytes by regulating the nuclear cytoplasmic translocation of Olig2 (Marumo et al., 2013). Mice with exclusive Notch 1 knockout in GFAP-positive reactive astrocytes also exhibited a significant reduction in proliferative reactive astrocytes (Shimada et al., 2011). Both mechanical and thermal hyperalgesia induced by CCI were effectively reduced by inhibiting the mRNA expression of Olig2, a marker of astrocyte activation (Bertozzi et al., 2017). The above findings suggest that the Olig2 signaling pathway is highly involved in astrocyte activation and the progression of neuropathic pain.

Chen et al. (2019) found that sphingosine-1-phosphate (S1P), which is produced in the spinal dorsal horn of CCI and SNI models, actuates neuropathic pain through targeted activation of S1P receptor subtype 1 (S1PR1) present in astrocytes. CCI and SNI models treated with an astrocyte-specific S1PR1 knockout were not observed to experience neuropathic pain. Therefore, astrocytes have been identified as the major cellular substrates for S1PR1 activity. In addition, their study also found that the relief of neuropathic pain induced by S1PR1 inhibition is mediated by interleukin-10 (IL-10), a cellular factor capable of participating in neuroprotection and fighting against inflammatory infections, thus identifying the S1PR1 axis as a key link in the developmental process and signaling pathways of neuropathic pain (Chen et al., 2019). Therefore, S1PR1 is expected to be a novel therapeutic target for neuropathic pain.

Transforming growth factor betas (TGF-βs), as versatile growth factors, are involved in critical events in the body that regulate growth, illness, and wound healing. TGF-β1 has been extensively known to be a cytokine implicated in injury, especially in the activation of astrocytes and scar production (Diniz et al., 2019). TGFβ is immediately triggered following CNS damage and stimulates the Sma- and Mad-related protein (SMAD) family of transcription factors in astrocytes (Koyama, 2014). Ralay Ranaivo et al. (2010) successfully activated the SMAD pathway downstream of TGFβR using albumin in primary astrocyte cultures. Other studies have found that the knockdown of repulsive guidance molecule a (RGMa) eliminates TGFβ1-induced astrocyte proliferation and activation and proved that RGMa potentially promotes TGFβ1/SMAD2/3 signal transduction by binding to activin receptor-like kinase 5 and SMAD2/3 simultaneously, constituting a composite (Zhang et al., 2018). The above findings suggest that the TGF-SMAD signaling pathway is expected to become an important target for astrocyte activation and neuropathic pain treatment.

As a member of the Ras homolog gene family, the small G protein RhoA is widely expressed in astrocytes and participates in cytoskeletal regulation together with its downstream effector Rho-dependent kinase (ROCK) by regulating myosin and actin (Tönges et al., 2011; Maldonado et al., 2017; Wen et al., 2019; Lu et al., 2021). The expression of RhoA and Rac1 in the glial scar area is significantly increased after injury (Erschbamer et al., 2005). Several experimental results reveal that the activation of RhoA/ROCK and its upstream and downstream molecules facilitates the occurrence and sustainment of neuropathic pain (Tatsumi et al., 2005; Hang et al., 2013). The inhibition of ROCK leads to rapid and reversible astral alignment of cultured astrocytes and increases their migratory activity. In one experiment, Rac1-KD or Rac1-KO astrocytes displayed significantly delayed cell cycle progression and decreased cell migration as well as the diminished expression of G1 to S phase transition 1 (GSPT1). In addition, the decreased expression and response of GSPT1 to lipopolysaccharide (LPS) treatment were observed. This suggests that GSPT1 is a downstream target of Rac1 (Ishii et al., 2017). Rac1-GSPT1, as a novel signaling axis in astrocytes, is an important factor in astrocyte activation after CNS injury. However, whether GSPT1 directly leads to neuropathic pain still warrants further research.

The MAPK family, which is composed of P38, extracellular signal-regulated kinase (ERK), and C-Jun N-terminal kinase (JNK), plays a crucial role in pain regulation (Gao and Ji, 2010b). Different members exhibit specific expressions in various pain states (Ji et al., 2009); nonetheless, all of them have been proven effective in alleviating neuropathic pain in various pain models (Zhuang et al., 2005; Gao et al., 2009; Wang et al., 2012).

Phosphorylated ERK has been observed in spinal astrocytes during the late stage of pain induced by complete Freund’s adjuvant (Weyerbacher et al., 2010). In an experiment, Yu used the mitogen-activated ERK kinase inhibitor U0126 to inhibit astrocyte activation and moderate the release of proinflammatory cytokines (IL-1β and IL-6) by activated astrocytes. The ERK1/2 pathway was found to potentially mediate astrocyte activity through the PPARγ pathway in CCI rats (Zhong et al., 2019).

Choi et al. (2019) demonstrated that the neurosteroid-metabolizing enzyme cytochrome P450c17 stimulates astrocytes through the P38 phosphorylation pathway, eventually contributing to the occurrence of CCI-induced mechanical nociception in mice.

Janus kinase has been shown to engage in the regulation and sustainment of nociception by astrocytes after nerve damage (Cao et al., 2015). P-JNK inhibitors relieve neuropathic pain in mice by inhibiting JNK signaling in spinal cord astrocytes (Li et al., 2020). Tetramethylpyrazine selectively inhibits JNK activity, inhibits astrocyte activation, and reduces matrix metalloproteinase-2/9 (MMP-2/9) expression, thus alleviating the maintenance of neuropathic pain (Jiang et al., 2017). This suggests that JNK-MMP-2/9 is a potentially promising new focus for neuropathic pain relief.

To date, relatively few ion channels and receptors that are involved in the activation of astrocytes, thus ultimately leading to the development of neuropathic pain, have been identified (Table 1).

Toll-like receptor 4 (TLR-4) is widely present throughout astrocytes. Overexpression of the high mobility group box-1 (HMGB1) protein potentially downregulates the paw withdrawal mechanical threshold and paw withdrawal thermal latency in CCI rodents by enhancing the activation of astrocytes, the TLR4/NF-κB axis, and its upstream and downstream molecules, resulting in the enhancement of neuropathic pain (Zhao et al., 2020). A subject of experimental observations in mice treated with toll-like receptor 2 (TLR2) knockout demonstrated that TLR2 expression is essential for neuropathic pain development. Neurodamage-induced astrocyte activation was also reduced in TLR2-knockout mice. The expression of TLRs has been found to enable astrocytes to release inflammatory mediators closely related to neuropathic pain, including, but not limited to, IL-6, monocyte chemoattractant protein-1 (MCP-1), and nitric oxide (NO) (Thakur et al., 2017). Thus, both TLR2 and TLR4 facilitate the progression and sustainment of pain caused by nerve damage (Kim et al., 2007).

A shortened isomeric form of tyrosine receptor kinase B (TrkB) has recently been found to be implicated in the pathological process of neuropathic pain (Renn et al., 2009). The TrkB.T1 isomeric form is generated through alternative splicing of TrkB. As the only isoform expressed in astrocytes (Rose et al., 2003; Dorsey et al., 2006; Ohira et al., 2007; Holt et al., 2019), Trkb.T1 has exhibited an increased expression in various experimental models of pain (Renn et al., 2009; Wu et al., 2013; Matyas et al., 2017) and attracted increasing attention. Antiretroviral treatment causes abnormal mechanical pain in mice through triggering the activity of brain-derived neurotrophic factor (BDNF) and consequently resulting in polyneuronal hyperactivity (Renn et al., 2011). Trkb.t1-specific knockout mice (Dorsey et al., 2006) were found to be relieved of abnormal pain sensation and heat hyperalgesia in the neuropathic pain model system (Renn et al., 2009). After RNA sequencing work on astrocytes extracted from TrkB.T1 KO mice, the researchers detected a downregulation in the ability of their astrocytes to divide, differentiate, and migrate (Cao et al., 2020). These results suggest that the imbalance in the physiological state of astrocytes caused by the involvement of TrkB.T1 promotes neuropathic pain development.

Considering the prevalence of neuropathic pain and incurable suffering associated with it, the study of ion channels and receptors associated with astrocytes is a potentially new opportunity to bring hope to affected people worldwide. AV411 (ibudilast), an antagonist of TLR4, is currently undergoing phase II clinical trials and has previously been shown to exert satisfactory efficacy in various models of neuropathic pain by inhibiting cytokines produced by glial cells as well as promoting the production of the anti-inflammatory cytokine IL-10 and neurotrophic factors (Connolly and O’Neill, 2012). This presents new directions and possibilities to research on targeted therapy for neuropathic pain. The TLR family and Trkb.T1 deserve further investigation and deeper exploration as key components of astrocytes that are closely linked to neuropathic pain.

Astrocytes form a network of gap junctions through Ca2+ signals in the form of calcium oscillations (Blomstrand et al., 1999; Haydon, 2001). The main structural components of gap junctions are connexins (CX), such as Cx26, Cx29, Cx30, Cx32, and Cx36, among others. Cx30 and Cx43 are specifically expressed in astrocytes (Giaume and McCarthy, 1996; Nagy et al., 2004). Astrocytic connexins not only regulate gap junction function but also act as semi-channels to transfer and release chemicals such as ATP from inside to outside the cell (Xing et al., 2019), thus regulating synaptic transmission and further leading to pain by directly interacting with injurious neurons (D’Hondt et al., 2014). Cx43 has been shown to effectively regulate the production and release of the chemokine CXCL12 in bone marrow stromal cells (Schajnovitz et al., 2011). Transfection of Cx43 can inhibit LPS-triggered hyperexcitability and upregulation of proinflammatory cellular factors and chemokines (Tsuchida et al., 2013). In addition, intrathecal injection of HIV1 GP120 is followed by a massive proliferation of IL-1β and IL-6 in the cerebrospinal fluid. Carboxy ketone (CBX), a non-specific gap junction antagonist, can block this in a timely and effective manner (Spataro et al., 2004). These behaviors are directly related to neuropathic pain.

IL-1β is a well-known proinflammatory cytokine. IL-1β upregulation has been observed in spinal cord astrocytes following peripheral nerve lesions. Increasing evidence suggests that IL-1β is involved in pain sensitivity (Cao and Zhang, 2008; Kiguchi et al., 2012). Intrathecal administration of IL-1β receptor antagonists alleviates inflammatory and neuropathic pain. Neuropathic pain was also significantly reduced in mice with IL-1β type I receptor knockout or those deliberately treated with an IL-1β antagonist. In contrast, IL-1β injected into the spinal canal induces pain hypersensitivity (Gao and Ji, 2010b; Ji et al., 2013). Further experiments have also shown that IL-1β is directly sensitive to thermally and chemically sensitive cationic channels in the transient receptor potential cation channel subfamily V member 1 (Kiguchi et al., 2012).

Chemokines, a specific group of cell factors possessing over 50 family components, are widely believed to exist in the blood and immune system and have recently been found to be widespread in the CNS and particularly active in astrocytes (Kiguchi et al., 2012). Chemokines predominantly perform their biological functions by binding to their G-protein-coupled receptors. Current studies have demonstrated that chemokines and corresponding receptors are usually present in different cell types to participate in their interaction and regulation (Gao and Ji, 2010a). CCL2, also called MCP-1, is mainly derived from astrocytes. A remarkable increase in JNK-dependent CCL2 in spinal astrocytes induced by peripheral nerve injury has also been observed, and it has been found to drive neuropathic pain with direct modulation of spinal neurons. CCL2 can rapidly increase the inward currents of spinal dorsal horn neurons induced by AMPA and NMDA. This indicates increased synaptic glutamate transmission, which is strongly associated with increased central sensitization and markedly diminished nociceptive sensation (Gao and Ji, 2010b; Nakagawa and Kaneko, 2010). Furthermore, CCL2 potentially stimulates the division and differentiation of spinal microglia to further support pain development (Nakagawa and Kaneko, 2010).

CCL3 has also been found to cause neuropathic pain through IL-1β upregulation (Kiguchi et al., 2012). CXCL1, an essential cytokine, potentially acts as a mediator, lubricating the communication between neurons and astrocytes (Kiguchi et al., 2012). In spinal nerve ligation models, CXCL1 and its receptor CXCR2 have been observed to undergo upregulation in spinal astrocytes and neurons (Zhang et al., 2013). Intrathecal injection of anti-CXCl1 antibodies has been found to successfully alleviate pain hypersensitivity induced by spinal nerve ligation, and CXCR2 antagonists can definitively terminate CXCl1-induced thermal pain allergy. This implies that CXCL1, CXCR2, and CCL3 are all involved in driving neuropathic pain.

Proteases expressed in astrocytes, such as matrix metalloproteinases (MMPs) and tissue-type plasminogen activators (tPAs), have been found to play a key role in the development of neuropathic pain driven by astrocytes (Gao and Ji, 2010b). Different MMPs have been implicated in the progression of neuropathic pain at different periods of time after nerve injury. MMP-9 induces the early development of neuropathic pain by mediating the early cleavage of IL-1β and activation of microglia, while MMP-2 is engaged in the sustainment of neuropathic pain in late stages through modulating the cleavage of IL-1β and activation of astrocytes (Kawasaki et al., 2008).

As an extracellular serine protease, tPA participates in the regulation and revision of tissue components located outside the cell, leading to alterations in neuroplasticity (Kozai et al., 2007). One study found that reactive astrocytes express tPA following trauma to the spinal cord, and tPA inhibitors injected into the spinal canal successfully inhibit mechanical pain triggered by dorsal root ligature (Kozai et al., 2007).

In recent years, BDNF, which binds to TrkB, has been found to be closely associated with astrocyte activation in a rat model of spinal nerve ligation. In certain experimental studies, the BDNF inhibitor ANA-12 was found to inhibit astrocyte activation in the dorsal horn of the spinal cord and subsequently alleviate or even inhibit the onset of mechanical nociceptive hypersensitivity. Increased BDNF levels and proinflammatory cytokines have also been observed following astrocyte activation (Chiang et al., 2012). Similarly, mechanical pain alterations have been reversed using fluorotic acids that inhibit astrocyte activation to downregulate BDNF (Ding et al., 2020). These actions are all accomplished through the binding of BDNF to TrkB.T1, the only TrkB receptor subtype expressed on astrocytes (Wu et al., 2013). BDNF produced by astrocytes has also been found to enhance NMDA receptors in excitatory neurons and decrease inhibitory neuronal activity by accomplishing depolarization displacement of the transient reversal potential (E_GABA) and decreased conductance of presynaptic gamma-aminobutyric acid (GABA). These provide BDNF the ability to enhance the synaptic activity of excitatory neurons, thus promoting the development of neuropathic pain (Zhou et al., 2021).