|Home | About | Journals | Submit | Contact Us | Français|
Clinical management of chronic pain after nerve injury (neuropathic pain) and tumor invasion (cancer pain) is a real challenge due to our limited understanding of the cellular mechanisms that initiate and maintain chronic pain. It has been increasingly recognized that glial cells, such as microglia and astrocytes in the central nervous system play an important role in the development and maintenance of chronic pain. Notably, astrocytes make very close contacts with synapses and astrocyte reaction after nerve injury, arthritis, and tumor growth is more persistent than microglial reaction and displays a better correlation with chronic pain behaviors. Accumulating evidence indicates that activated astrocytes can release proinflammatory cytokines (e.g., IL-1β) and chemokines (e.g., MCP-1/CCL2) in the spinal cord to enhance and prolong persistent pain states. IL-1β can powerfully modulate synaptic transmission in the spinal cord by enhancing excitatory synaptic transmission and suppressing inhibitory synaptic transmission. IL-1β activation (cleavage) in the spinal cord after nerve injury requires the matrix metalloprotease-2 (MMP-2). In particular, nerve injury and inflammation activate the c-Jun N-terminal kinase (JNK) in spinal astrocytes, leading to a substantial increase in the expression and release of MCP-1. MCP-1 increases pain sensitivity via direct activation of NMDA receptors in dorsal horn neurons. Pharmacological inhibition of the IL-1β, JNK, MCP-1, or MMP-2 signaling via spinal administration has been shown to attenuate inflammatory, neuropathic, or cancer pain. Therefore, interventions in specific signaling pathways in astrocytes may offer new approaches for the management of chronic pain.
Pain is an unpleasant sensory experience and normally plays a protective role by warning us about potential harm to our body and enabling us to quickly remove the body part from noxious stimuli and further learn to avoid them in the long run. Upon noxious peripheral stimulation, pain information is mainly transmitted by thin myelinated Aδ fibers and unmyelinated C fibers to the dorsal horn in the spinal cord, where second order nociceptive neurons are activated by neurotransmitters, such as glutamate and neuropeptides [e.g., substance P (SP) and calcitonin gene-related peptide (CGRP)] that are released from the primary afferents.1 The information is further relayed to the thalamus, and finally reaches to the parietal lobe of cerebral cortex for pain perception.2, 3 This type of pain is transient and referred to as acute or physiological pain.
However, under injury conditions pain can be dissociated from its normal physiological role. It can persist for months to years, even after the original injury or inflammation has largely healed. This type of pain is called chronic or pathological pain, as the consequence of damage or dysfunction of the peripheral nervous system (PNS) and central nervous system (CNS) (neuropathic pain), peripheral tissue damage or inflammation (inflammatory pain), and tumor invasion (cancer pain).4–6 Chronic pain does not convey any useful information. Under injury conditions, painful pressure and thermal stimuli are grossly amplified (hyperalgesia). Even light touch is perceived painful (allodynia). Chronic pain creates considerable suffering for people affected, and is extremely costly for the individual and for the community; the estimated cost in the United States alone is over $100 billion every year7.
Chronic pain is a maladaptive pain, resulting from the development of neural plasticity in the PNS (peripheral sensitization) and CNS (central sensitization).8–10 It was generally believed for a long time that only neurons and their neural circuits were responsible for the development and maintenance of chronic pain, which led to the development of current therapeutics that have been focusing on neuronal targets, including drugs such as N-methyl-D-aspartic acid (NMDA) receptor antagonists, selective serotonin/norepinephrine reuptake inhibitors, opioid analgesics, and sodium channel blockers. Although these drugs have shown some effects in some patients,11 they often produce a brief pain relief via transient blockade of neurotransmission. Notably, the side effects of these drugs, often CNS-related, such as nausea, sedation, drowsiness, dizziness, as well as development of analgesic tolerance and addiction after opioid treatment, have greatly limited their universal use.11, 12 Therefore, research on other means of chronic pain treatment is in an urgent demand. As a consequence, studies on non-neuronal cells, especially glial cells in chronic pain conditions, have increased exponentially in the last decade.
Glial cells are 10 to 50 times as numerous as neurons and consist of three major groups: astrocytes, microglia, and oligodendrocytes.13 Microglia are the resident macrophage-like cells of the CNS. Oligodendrocytes, which are derived from neuroectoderm, produce myelin to enshealth neuronal axons. Astrocytes are the most abundant cells in term of their number and volume and constitute 40–50% of all glial cells.14 In normal conditions, microglia and astrocytes are relatively resting or quiescent (but see15). After injury or under disease conditions, they can be converted to reactive states and participate in the pathogenesis of neurological disorders.16–18 Increasing evidence has shown that microglia and astrocytes play important roles in the development of chronic pain.18–27 Unlike microglia and oligodendrocytes, astrocytes form networks with themselves and are closely associated with neurons and blood vessels. It is estimated that a single astrocyte enwraps 4–6 neuronal somata and contacts 300–600 neuronal dendrites.28 A close contact with neurons and synapses makes it possible for astrocytes to support and nourish neurons, and regulate the external chemical environment of neurons during synaptic transmission. In this review, we will discuss recent progress on astrocyte control of pain.
Glia activation is emerging as a powerful concept for understanding cellular mechanisms underlying chronic pain. Unfortunately, the term “glia activation” is poorly defined. In the pain research field, astrocyte activation is often referred to GFAP upregulation and astrogliosis (hypertrophy of astrocytes, as manifested by enlarged cell bodies and thick processes). The active astrocytes with gliosis are also called reactive astrocytes. Thus, in the following discussion we refer to this activation state as the reactive state, in order to separate from other activation states. It is well known that after peripheral nerve injury or inflammation or tumor invasion, astrocytes in the CNS especially the spinal cord undergo various biochemical, translational, transcriptional, and morphological changes. Therefore, astrocytes could display various activation states after peripheral sensory stimuli and injury. Some activation states occur within minutes, such as increases in intracellular Ca2+ and phosphorylation of signaling molecules. Some activation states occur after tens of minutes (e.g., translational regulation) and hours (e.g., transcriptional regulation). Other activation states may occur after hours or even days, such as astrocyte hypertrophy or astrogliosis.
Astrocyte reaction (GFAP upregulation and hypertrophy) has been found in various injury conditions that are associated with enhanced pain states. These conditions include (a) peripheral nerve injury such as chronic constriction injury (CCI),29 spinal nerve ligation (SNL, Fig. 1),30, 31 and infraorbital nerve ligation,32, 33 (b) tissue injury/inflammation produced by intraplantar injection of complete Freund's adjuvant,34 formalin,35 zymosan,35 and (c) tumor growth in the skin36–38 and bone marrow.39–41 While astrocyte reaction can occur at supraspinal areas, such as the rostral ventramedial medulla after chronic constriction injury of the rat infraorbital nerve,32, 33 the forebrain after CFA injection,34 the gracile nucleus after partial sciatic nerve ligation,42 most studies focus on the spinal cord dorsal horn.19
Notably, astroglial reaction after nerve injury is more persistent than microglial reaction (e.g., upregulation of the microglial markers CD11b/OX-42 and Iba-1 and hypertrophy of mciroglia). Astroglial reaction can last more than 150 days after nerve injury.43 In most cases, microglial reaction precedes astrocytic reaction34, 44, 45 and is likely to lead to astrocyte reaction.46 Interestingly, nerve injury induces an increase in IL-18 and IL-18 receptor in reactive microglia and astrocytes, respectively, in the dorsal horn, suggesting an interaction between microglia and astrocytes in neuropathic pain.47 However, astrocyte reaction is not always preceded by microglial reaction. Hald et al40 showed that bone cancer resulted in marked spinal astroglial reaction without microglial reaction.
It has been shown that GFAP expression after inflammation or nerve injury requires NMDA receptor48, 49 and neuronal activity.32, 49 GFAP expression is also critical for morphological changes of astrocytes (astrogliosis)34, 35 and often correlated with enhanced pain states,29, 30, 50 (but see31). Although intrathecal GFAP antisense oligonucleotide treatment in nerve injured animals was shown to reduce neuropathic pain behaviors,51 this contribution of GFAP to chronic pain could be indirect via unknown mechanisms. It is generally believed that astrocytes control pain states by producing neuromodulators/pain mediators, such as cytokines, chemokines, and growth factors,22, 32, 33, 52, 53 (also see discussion below). The production and release of these mediators are not directly controlled by GFAP, rather by some key intracellular signaling pathways, such as the MAP kinase pathway. Remarkably, nerve injury and inflammation induce a persistent phosphorylation of c-Jun N-terminal kinase (JNK) in astrocytes, which may represent a different activation state of astrocytes that is not only correlated with pain hypersensitivity but also an underlying cause of this hypersensitivity (Fig. 1).31, 54–56
Several lines of evidence suggest that activated astrocytes are sufficient to produce chronic pain symptoms. Hofstetter et al. reported that implantation of neural stem cells into the injured spinal cord causes allodynic-like hypersensitivity of the forepaws, which is mainly attributed to the conversion of the stem cells into astrocytes.57 Indeed, the allodynia is prevented when the neural cells are transfected with neurogenin-2 before transplantation to suppress the generation of astrocytes.57 Davies et al. demonstrated that transplantation of GRP (glial-restricted precursor)-derived astrocytes promotes the onset of mechanical allodynia.58 In particular, our recent data showed that intrathecal injection of TNF-α-activated astrocytes is sufficient to induce the chronic pain hall-mark, mechanical allodynia, in naïve animals by releasing the chemokine CCL2.59
Further studies indicate that astrocytes are also required for the generation of persistent pain. Fluoroacetate and its metabolite fluorocitrate are general inhibitors for glial cells especially astrocytes. Low doses of fluorocitrate specially disrupt astrocytic metabolism by blocking the glial-specific enzyme aconitase. Intrathecal injection of fluorocitrate or fluoroacetate has been shown to alleviate pain behaviors in animal models of inflammatory pain, neuropathic pain and postoperative pain.60–65 Of interest fluorocitrate fails to inhibit muscle pain, a pain condition that does not show obvious glial reaction.66 L-alpha-aminoadipate (L-α-AA) is another relative specific cytotoxin for astrocytes.67–69 Intrathecal injection of L-α-AA produces a dose-dependent attenuation of nerve injury-induced mechanical allodynia.31, 70
There is an increasing list of signaling molecules in astrocytes that have been implicated in persistent pain (Table 1). The glial glutamate transporter 1 (GLT-1) is abundantly expressed in astrocytes71 and contributes to the clearance of glutamate from synaptic clefts and the extracellular space.72, 73 The altered expression and function of glutamate transporters could modulate glutamatergic transmission74, 75 and neuronal plasticity such as long-term potentiation.76, 77 It has been demonstrated that nerve injury induces an initial increase78, 79 followed by a persistent decrease of GLT1 and GLAST in the spinal cord.78–81 Inhibition of glutamate transporters causes an elevation in spinal extracellular glutamate concentrations and elicits spontaneous nociceptive behaviors and hypersensitivity to mechanical and thermal stimuli.82, 83 Gene transfer of GLT-1 into spinal cord has no effect on acute mechanical and thermal nociceptive responses in naive animals but attenuates inflammatory and neuropathic pain.84 These studies indicate a potential role of astroglial glutamate transporters in the recovery of chronic pain. However, the role of glutamate transporters in persistent inflammatory pain conditions could be different, since these transporters are not down-regulated after inflammation. Trigeminal pain following tooth pulp inflammation is attenuated by intrathecal superfusion of methionine sulfoximine, an inhibitor of the astroglial enzyme glutamine synthetase that is involved in the glutamate-glutamine shuttle.85
Astrocytes express proteases such as tissue type plasminogen activator (tPA) and matrix metalloproteases (MMPs, see below) that may be critical for the cleavage and release of signaling molecules from astrocytes. tPA is an extracellular serine protease and converts the plasminogen into the serine protease plasmin. Kozai et al.86 showed that L4/5 root injury induces marked induction of tPA in activated astrocytes and a resultant increase of proteolytic enzymatic activity in the dorsal horn. Moreover, intrathecal administration of tPA inhibitor suppresses dorsal root ligation-induced mechanical allodynia. tPA-plasmin system may alter the excitability of dorsal horn neurons and pain transmission through the activation of growth factors87, 88 and modification of the NMDA receptors.89, 90
Astrocytes are characterized by forming gap junction-coupled networks, which could transmit Ca2+ signaling in the form of oscillations through the networks.91, 92 The major structural components of gap junctions are connexins. In the mammalian nervous system, at least six connexins (Cx26, Cx29, Cx30, Cx32, Cx36 and Cx43) have been identified. Among them, Cx30 and Cx43 are specifically expressed by astrocytes.93, 94 Interestingly, the expression of Cx43 increases markedly in response to facial nerve lesion,95 spinal cord injury,96 and CFA-induced inflammation,32 indicating a role of connexin in chronic pain. Inhibition of gap junction function by carbenoxolone— a nonselective gap junction inhibitor— produces analgesia in different pain models.97–99 Particularly, intrathecal injection of carbenoxolone reduces sciatic nerve inflammation-induced mechanical allodynia in the contralateral paw, suggesting a role of astrocytes network and gap junction in the spread of pain beyond the injury site.98
In addition, astrocytes also express phosphorylated JNK and JNK1 (Fig. 1),31, 56 phosphorylated ERK,53, 100 endothelin receptor-B,101 TNF-α,33 bFGF (Fig. 2),102, 103 neurokinin-2 receptor,104 IL-18 receptor,47 IL-1β32, 33, 53, 100 and monocyte chemoattractant proetine-1 (MCP-1),52, 105 in response to nerve injury or inflammation. Importantly, pharmacological inhibition of these signaling molecules via spinal cord administration has been shown to reduce chronic pain symptoms (Table 1).
IL-1β is a major proinflammatory cytokine and upregulated in the spinal cord under different chronic pain conditions.35, 62, 106 Specifically, several studies have shown IL-1β upregulation in astrocytes after bone cancer,41 nerve injury,33 hindpaw inflammation53, 107 and masseter inflammation.32 IL-1β was also found in neurons in the spinal cord.108, 109 Several lines of evidence support an important role of IL-1β for pain sensitization. Inhibition of spinal IL-1β signaling with intrathecal IL-1 receptor antagonist (IL-1ra) or neutralizing antibody has been shown to alleviate inflammatory, neuropathic, and cancer pain.32, 33, 62, 106, 107, 110, 111 Neuropathic pain is also markedly reduced in mouse strains with deletion of the IL-1 receptor type I or transgenic over-expression of IL-1ra.112 Conversely, intrathecal injection of IL-1β is sufficient to elicit pain hypersensitivity.113–117
IL-1β released from astrocytes could directly modulate neuronal activity. Immunostaining shows that IL-1 receptor colocalizes with the NMDA receptor NR1 subunits in neurons of the spinal cord,107 trigeminal nucleus,32 and rostral ventromedial medulla.33 In primary cultured neurons, IL-1β regulates the phosphorylation of the NMDAR NR2B and NR1 subunit.32, 118 IL-1β-mediated enhancement of NR1 subunit phosphorylation in the spinal cord may facilitate inflammatory pain and bone cancer pain.107, 119 In particular, our ex vivo electrophysiological study using patch clamp recordings in lamina II neurons demonstrated that bath application of IL-1β onto isolated spinal cord slices can markedly enhance NMDA-induced current.106 Perfusion of spinal slices with IL-1β also increases the frequency and amplitude of spontaneous postsynaptic currents (sEPSCs) in dorsal horn neurons, indicating that IL-1β can directly enhance excitatory synaptic transmission.106 While the frequency increase of sEPSCs results from increased glutamate release from presynaptic terminals, the amplitude increase is caused by enhanced signaling of glutamate receptor (AMPA-subtype) in postsynaptic sites. IL-1β also increases the excitability of nociceptors via IL-1R that is expressed in small size primary sensory neurons,120 leading to increased glutamate release in nociceptor central terminals in the spinal cord. Strikingly, IL-1β can further modulate inhibitory synaptic transmission in dorsal horn neurons. Bath application of IL-1β reduces the frequency and amplitude of spontaneous inhibitory postsynaptic currents (sIPSCs) and inhibits GABA-and glycine-induced currents in lamina II neurons,106 which will contribute to disinhibition (loss of inhibition), an important mechanism that is increasingly appreciated for the generation of neuropathic pain.121, 122 Collectively, these studies suggest that IL-1β powerfully modulates synaptic transmission by a) enhancing excitatory synaptic transmission and b) reducing inhibitory synaptic transmission. In addition, IL-1β also produces long-term neuronal plasticity in the pain circuit by inducing the phosphorylation of the transcription factor CREB9, 106 and expression of COX-2 in spinal cord neurons.123
IL-1β is synthesized as a precursor and requires a protease for its activation via cleavage to produce biological function. Notably, caspase-1 is not the only enzyme for IL-1β cleavage.114 MMPs have been implicated in the cleavage of extracellular matrix proteins, cytokines, and chemokines to control inflammation and tissue remodeling associated with various neurodegenerative diseases.124–127 Several studies showed that MMP-9 and MMP-2 are involved in IL-1β cleavage.114, 125, 128 Particularly, MMP2 is persistently induced in astrocytes after spinal nerve ligation.114 Treatment of MMP-2 siRNA in the late-phase of nerve injury blocks IL-1β cleavage in the spinal cord and reduces mechanical allodynia.114 These data suggest that astrocyte-derived MMP-2 may maintain neuropathic pain by active cleavage of IL-1β.
MCP-1 (also called CCL2) is the chemokine that is highly produced by astrocytes. MCP-1 expression is increased in spinal cord astrocytes after spinal nerve ligation52 and spinal cord contusion injuries.105 Several studies demonstrate that activated astrocytes in vitro also produce MCP-1.52, 129–132 MCP-1 was found in astrocytes in the brain after demyelinating lesions,133, 134 mechanical injury,135 entorhinodentate axon transaction,136 and focal cerebral ischemia.137
CCR2, the major receptor of MCP-1, is expressed in DRG neurons and increased in these neurons after nerve injury.138 CCR2 is also constitutively expressed in spinal cord neurons,52, 139 which is upregulated after nerve injury.52 Our recent study indicated a direct action of MCP-1 on spinal cord neurons. In isolated spinal cord slices, perfusion of MCP-1 immediately increases the frequencies of sEPSCs and the amplitude in lamina II neurons of the dorsal horn.52 MCP-1 also rapidly (< 2 min) enhances NMDA- and AMPA-induced inward currents.52, indicating a potentiation of glutamatergic synaptic transmission, which has been strongly implicated in central sensitization and hyperalgesia,9, 122 Additionally, Gosselin et al.139 demonstrated in neonatal cultures that MCP-1 inhibits GABA-induced currents in spinal neurons without affecting the electrical properties of these neurons. Thus, MCP-1 also modulates inhibitory synaptic transmission in spinal cord neurons.
In parallel with electrophysiological evidence, behavioral evidence shows that spinal injection of MCP-1 induces rapid heat hyperalgesia, starting at 15 min, peaking at 30 min, and recovering at 24 h.52 Moreover, incubation of spinal cord slice with MCP-1 induces a rapid (within 5 minutes) phosphorylation of the extracellular signal-regulated kinase (pERK) in superficial dorsal horn neurons52, which is regarded as a marker for spinal nociceptive neuron sensitization (central sensitization)140. Thus, the rapid phosphorylation of ERK in dorsal horn neurons by MCP-1 supports a direct action of MCP-1 on spinal cord neurons and its involvement in central sensitization. In neuropathic pain models, MCP-1 neutralizing antibody reduces mechanical allodynia induced by SNL52 or CCI.141 Nerve injury-evoked mechanical allodynia is also reduced by CCR2 antagonist or in mice lacking CCR2.142–145 Taken together, these studies demonstrate an important role of MCP-1/CCR2 in chronic pain via astrocyte-neuron interaction (Fig. 3). In additional to a direct action of neurons, astrocyte-produced MCP-1 may also act on microglia to induce proliferation and migration of microglia in the spinal cord, which can further enhance pain.143
Mounting evidence has demonstrated important roles of mitogen-activated protein kinases (MAPKs)—ERK, p38, and JNK—in chronic pain sensitization.146 Of interest these MAPKs are differentially activated in spinal cord glial cells after nerve injury. While p38 is persistently activated in microglia at all the times examined,31, 147, 148 ERK is only activated in microglia in the early-phase (first several days) of nerve injury.100 In the late-phase (>3 weeks) of nerve injury, pERK is induced in spinal astrocytes.42, 100 Spinal inhibition of this late-phase activation of ERK by intrathecal administration of a MEK inhibitor reverses mechanical allodynia, implicating a role of astrocytic ERK in the maintenance of neuropathic pain.100 Intraplantar injection of CFA also induces pERK in spinal cord astrocytes in the late-phase of this inflammatory pain condition.149
We will focus our discussion on JNK, also called stress-activated protein kinase, which is well known for its role in regulating apoptosis and neurodegeneration.140 But JNK activation in spinal astrocytes after peripheral nerve injury is not associated with apoptosis in astrocytes.31 Rather, JNK activation in astrocytes regulates the expression and release of chemokines.52 SNL induces a persistent (> 3 weeks) increase of phosphorylated JNK (pJNK) in the spinal cord, particularly in reactive astrocytes.31 Increase in pJNK was also found in spinal astrocytes in other neuropathic conditions such as partial sciatic nerve injury42 and amyotrophic lateral sclerosis.150 pJNK is further induced in spinal cord astrocytes in inflammatory pain conditions following intraplantar injection of carrageenan55 and CFA56. In particular, CFA elicits a bilateral phosphorylation of JNK, starting at 6 hours and maintaining after 2 weeks.56 Despite there are three isoforms of JNK (JNK1, JNK2 and JNK3), JNK1 is the isoform that is expressed in spinal astrocytes and hyperphosphorylated after SNL and CFA injection.31, 56 In parallel, inflammatory pain is reduced in mice lacking JNK1 but not JNK2.56
The role of astrocyte JNK in pain control has been also evaluated by intrathecal injection of the JNK inhibitor SP600125. Administration of SP600125, either before or after nerve injury, can both attenuate neuropathic pain after SNL.31, 151 SP600125 also suppresses neuropathic pain in a diabetes model of neuropathic pain.152 The peptide inhibitor D-JNKI-1 is a more potent and selective inhibitor of JNK. A single bolus injection of D-JNKI-1 can block SNL-induced mechanical allodynia for more than 6 hours.31
How does JNK signaling in astrocytes control chronic pain? JNK activation in astrocytes results in the production of various inflammatory mediators. In cultured astrocytes there is a JNK-dependent expression of COX-2 and iNOS as well as the release of NO, PGE2 and IL-6.153 Notably, stimulation of astrocytes with TNF-α not only activates JNK but also induces a marked upregulation of several chemokines, such as MCP-1, KC, IP-10.52 Strikingly, TNF-α induces a substantial increase (>100 fold), both in the expression and release of MCP-1 in astrocyte cultures; and this increase is completely blocked by JNK inhibition.52 JNK activation in astrocytes also leads to the production of MCP-1 in vivo.52 Thus, JNK activation in astrocytes can enhance pain via producing chemokines such as MCP-1, which is known to increase the sensitivity of dorsal horn neurons.154
JNK is activated by the transforming growth factor (TGF)-activated kinase 1 (TAK1), a member of the MAPK kinase kinase family. Interestingly, peripheral nerve injury induces TAK1 upregulation in hyperactive astrocytes in the spinal cord.155 Intrathecal administration of TAK1 antisense oligodeoxynucleotides, either before and after nerve injury, can reduce nerve injury-induced mechanical allodynia.155
Basic fibroblast factor (bFGF) is a well-known activator of astrocytes and induces mitosis, growth, differentiation, and gliosis of astrocytes.156, 157 Spinal nerve ligation induces a substantial increase of bFGF in reactive astrocytes in the late phase (3 weeks after injury, Fig. 2). Intrathecal infusion of bFGF induces persistent JNK phosphorylation and GFAP expression in the spinal cord, which is associated with the development of mechanical allodynia.22 Conversely, intrathecal injection of a bFGF neutralizing antibody can reverse nerve injury-induced mechanical allodynia.102 Compared to a transient JNK activation by TNF-α, bFGF induces a sustained activation of JNK in astrocyte cultures.22, 52 Therefore, bFGF, produced in astrocytes in the late-phase of injury, may maintain chronic pain via sustained JNK activation in astrocytes.
In summary, we have reviewed behavioral, histochemical, and electrophysiological evidence to support a rising role of astrocytes in chronic pain sensitization. We also demonstrate how astrocytes promote chronic pain via neuronal-glial interactions (Fig. 3). After peripheral nerve injury or tissue damage in the skin, muscle, or joint (e.g., arthritis), astrocytes are activated in the spinal cord, in response to neurotransmitters/neuromodulators (e.g., ATP, glutamate, neuropeptides) and inflammatory mediators (e.g., TNF-α) released after injuries. Astrocyte activation may manifest as the activation of several intracellular signaling pathways such as the JNK and ERK pathways or/and up-regulation of GFAP and astrogliosis/hypertrophy (Fig. 3a). Activation of the JNK or/and ERK results in the production of proinflammatory cytokines and chemokines (e.g., IL-1β and MCP-1). These mediators can act at both presynaptic sites on primary afferents and post-synaptic sites on dorsal horn neurons to increase excitation and decrease inhibition of spinal cord nociceptive neurons, leading to enhanced pain states (Fig. 3b).
Given the important role of astrocytes in chronic pain facilitation, targeting astrocytes could reveal novel therapies for the management of chronic pain. However, caution must be taken when we consider strategies to target astrocytes, since astrocytes play an essential supportive and protective role in the CNS.158 Inhibition of reactive astrocytes with the toxin fluorocitrate has been shown to retard neurovascular remodeling and recovery after focal cerebral ischemia.159 Thus, it is important to target specific signaling events in astroctyes, without disrupting the overall well being of astrocytes. As discussed in Table-1, all the signaling molecules that are induced in astrocytes under chronic pain conditions and contribute to pain behaviors can be potentially targeted. In particular, JNK inhibitor not only exhibits anti-allodynic action but also has a neuroprotecitve role.160 JNK inhibitor further reduces tumor growth36 and insulin resistance161, 162, therefore, should be beneficial in pain conditions associated with cancer and diabetic neuropathy.
Finally, it is worthy to note that all the evidence we present is from animal studies. Indeed, astrocytes from human are quite different.163 The human brain appears to contain subtypes of GFAP-positive astrocytes that are not represented in rodents. Strikingly, in human cortex, astrocytes are >2 fold larger in diameter and extend 10-fold more GFAP-positive primary processes than their rodent counterparts. The domain of a single human astrocyte has been estimated to contain up 2 million synapses.164 Hence, it is reasonable to postulate human astrocytes may play a more important role in chronic pain control than rodent astrocytes.
This work was supported by NIH R01 grants NS54932, NS67686, and DE17794.