Abstract

Opioids have been used as potent analgesics in clinics for decades; however, their long-term administration leads to tolerance. Two possible mechanisms for drug tolerance are postulated as within-system and between-systems adaptation. The within-system tolerance is involved in the signal transduction of opioid receptors, including downregulation of opioid receptors, uncoupling of G-protein from opioid receptors, and β-arrestin recruitment to opioid receptors, which causes receptor desensitization and internalization/endocytosis. The between-systems tolerance comprehends the glutamatergic receptor system and glial activation with the release of proinflammatory cytokines, and thus the analgesic effect of morphine is reduced. Tumor necrosis factor-α (TNF-α) is a vital proinflammatory cytokine and exerts either a neurotoxic or neuroprotective effect on different diseases of the central nervous system. TNF-α has also been demonstrated to correlate with neuronal plasticity via activation of spinal glial cells and enhancement of glutamatergic transmission. Previous studies had revealed an increased expression of TNF-α in morphine tolerance. This review article focuses on the role of TNF-α in neuroinflammation and the glutamatergic receptor system in morphine tolerance. It may provide another adjuvant therapy for morphine tolerance, which extends the effectiveness of opioids in clinical pain management.

Keywords

drug tolerance; inflammation: neural; morphine; synaptic transmission: glutamatergic; tumor necrosis factor-alpha;

1. Introduction

Opioids have a potent analgesic effect and have widely been used for decades. Long-term opioid administration develops tolerance, which limits its clinical efficacy. Koob and Bloom¹ described two possible mechanisms for drug tolerance: within-system and between-systems adaptation. The within-system tolerance occurs by eliciting an opposite action within the same system; the between-systems tolerance is characterized by alterations in the primary drug-sensitive system adaptation not directly involved in the drug's primary action system. Investigations of the mechanisms of within-system included modulation of intracellular adenylyl cyclase (AC) and cAMP-dependent protein kinase A (PKA),2, 3 uncoupling of G-protein signaling,4, 5 increased binding of β-arrestin to opioid receptors,6, 7 and μ-opioid receptor oligomerization.8, 9 As to the mechanisms of the between-systems, N-methyl-d-aspartate (NMDA) receptors,10, 11 glutamate transporters (GTs),¹² and glial activation with the release of proinflammatory cytokines stand out.13, 14

Tumor necrosis factor-α (TNF-α) was actively bound to TNF-α receptors, which are constitutively expressed on both neurons and glial cells in the central nervous system (CNS).¹⁵ The complexity of TNF-α responses is related to the receptor context, cell type, and functional status.¹⁶ Two different TNF-α receptors (p55/TNF-R1 and p75/TNF-R2) have been recognized¹⁷; they act in distinct pathways to mediate different cellular responses.¹⁸ Morphine tolerance shares similar neuronal plasticity mechanisms associated with hyperalgesia and allodynia in neuropathic pain, which are modulated by proinflammatory cytokines including TNF-α.¹¹ Song and Zhao¹³ demonstrated that glial cells react to chronic morphine treatment by astroglial hypertrophy and increase glial fibrillary acidic protein expression in the spinal cord, posterior cingulated cortex, and hippocampus. Glutamate homeostasis is important in controlling the synaptic activity and further cellular dialogue between neurons, astrocytes, and microglia; the involvement of TNF-α was observed, via NF-κB activation, in the inhibition of GT expression and potentiation of glutamate neurotoxicity in organotypic brain slice culture.¹⁹ TNF-α enhances synaptic strength by increasing surface expression of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, which is prevented by the blockade of the TNF-α signaling pathway in cultured hippocampal neurons.²⁰ These findings prove that TNF-α plays a significant role in the modulation of glutamatergic transmission. This review addresses the role of the proinflammatory cytokine TNF-α in the development of morphine tolerance and the interactions between neuroinflammation and the glutamatergic receptor system.

2. TNF-α and neuroinflammation in the development of morphine tolerance

2.1. Role of neuroinflammation in morphine tolerance

Clinical evidence has showed that heroin addicts are prone to infection.²¹ It had been demonstrated that morphine compromises immune function by promoting macrophage apoptosis in opiate addicts²² and inducing Fas expression in lymphocytes.²³ This immune compromise by morphine administration was postulated to result from a direct interaction between opioid receptors and immune cells, or indirectly through the activation of opioid receptors in the CNS, which modulate the hypothalamic pituitary adrenal axis and the autonomic function.²⁴ Drug abusers, including those of cocaine and morphine, have been shown to have significant T-cell-dependent antibody responses with increasing IL-4 and shifting of the T-cell function to Th2-type cells via an opioid receptor-mediated action.25, 26 An acceleration of morphine tolerance development in animals with neuropathic pain suggests similar cellular mechanisms in tolerance and pain hypersensitivity.11, 27 A previous study demonstrated that glia-derived proinflammatory cytokines inhibit the antinociceptive effect of morphine by sensitizing pain-transmission neurons in animals with morphine tolerance and neuropathic pain.²⁸ Both central and peripheral administration of proinflammatory cytokines TNF-α, IL-1β, and IL-6 facilitate pain transmission,29, 30 and this reduction of the antinociceptive effect of morphine can be reversed by inhibition of glial metabolism, antagonism of IL-1 receptors, and induction of anti-inflammatory cytokine IL-10 expression.13, 31 In our previous study, chronic morphine infusion induced a significant increase in TNF-α, IL-1β, and IL-6 mRNA expressions in the spinal cord dorsal horn of tolerant rats; inhibition of the proinflammatory cytokine expression attenuated the morphine tolerance; and administration of the TNF-α inhibitor etanercept reduced proinflammatory cytokines production and microglial activation, thus preserving the antinociceptive effect of morphine.³² Taken together, blockade of TNF-α signaling seems to be an effective biological target in the management of morphine tolerance.

2.2. Glial activation and morphine tolerance

Microglia have been shown to play a main role in cytokine release upon activation in the CNS.³³ Compelling evidence has suggested that non-neuronal cells, instead of neurons, are crucial for the development of morphine tolerance.³¹ Repeated morphine administration leads to activation of calcitonin gene-related peptide (CGRP) receptors, which differentially control the synthesis and release of proinflammatory cytokines TNF-α, IL-1β, and IL-6 from microglia and astrocytes as well by extracellular signal-regulated kinase and p38MAPK.³⁴ The blockade of CGRP receptor signaling in microglia and astrocytes prevents proinflammatory cytokine release and subsequent tolerance development. Based on previous studies, pharmacological modulation of spinal glial activity using amitriptyline and etanercept attenuated morphine tolerance.32, 35 Investigation of other potent anti-inflammatory drugs for morphine tolerance management is thus warranted to further explicate the features of morphine tolerance.

2.3. Effect of the TNF-α inhibitor etanercept in morphine tolerance

The first clinical use of TNF inhibitors for the treatment of rheumatoid arthritis was in 1998.³⁶ Etanercept is a recombinant soluble p75 receptor (p75), which is linked to the Fc portion of human IgG. Nonsoluble but membraneous TNF-α receptors can elicit biological activity upon TNF-α binding. Etanercept binds two circulating TNF-α, preventing it from interacting with its membrane receptors, and restrains the following signaling.³⁷ Application of etanercept in the treatment of other inflammatory disorders such as ankylosing spondilytis,³⁸ Crohn disease,³⁹ and Alzheimer disease⁴⁰ has expanded in recent years. Perispinal administration of etanercept attenuated TNF-α effect on synaptic dysfunction, which rapidly improved cognitive function for Alzheimer patients.⁴⁰ In our study, intrathecal pretreatment with etanercept 50 μg suppressed effectively microglia activation and mRNA expression of proinflammatory cytokines TNF-α, IL-1β, and IL-6 in the spinal cords of morphine-tolerant rats.³² It restored the antinociceptive effect of morphine with a maximum of 60% recovery in morphine-tolerant rats.

3. Glutamatergic receptor system and morphine tolerance

3.1. Excitatory amino acids and GTs

The acidic amino acid glutamate is a major excitatory neurotransmitter in the mammalian CNS including spinal cord. Excess of glutamate in the synaptic cleft may result in massive Ca²⁺ influx, causing neuronal plasticity and even excitotoxicity.41, 42 Hence, maintenance of glutamate homeostasis is crucial for highly efficient synaptic communication and prevention of neurotoxicity.⁴³ In physiological conditions, glutamate is rapidly removed from the synaptic cleft by diffusion and reuptake.44, 45 The latter is accomplished by a family of high-affinity Na⁺-dependent GTs localized in the cytoplasmic membrane of astrocytes and neurons.⁴⁶ Five GTs have been cloned and characterized. EAAC1 (excitatory amino acid carrier/transporter; EAAT3) and EAAT4 are predominantly localized in neurons, and EAAT5 is enriched in retinal tissue, whereas glutamate–aspartate transporter (GLAST; EAAT1) and GLT-1 (EAAT2) are generally expressed in astrocytes.⁴⁷

3.2. Morphine tolerance and NMDA receptors

Glutamatergic receptors, in particular NMDA receptors, are critically involved in chronic opioid-induced neuronal adaptations, such as opioid tolerance, dependence, and withdrawal,10, 48, 49 and in chronic pain-associated hyperalgesia.50, 51, 52 Emerging evidence suggests that opioid tolerance and abnormal pain sensitivity, the opioid-induced hyperalgesia, may share common cellular mechanisms and mediate, at least in part, through NMDA receptors12, 53 and GT system.⁵⁴ Our previous studies have demonstrated that coinfusion of the NMDA antagonist MK-801 with morphine shifted the AD₅₀ (analgesic dose) of morphine to 11.2 μg compared to 83.8 μg in morphine tolerant-rats.⁵⁵ In addition, dexamethasone retained the antinociceptive effect of morphine in chronic morphine-infused rats via inhibition of downregulation of GTs and the proceeding increased of EAAs.⁵⁶ We therefore assume that GT regulation plays an important role in morphine tolerance, by inhibiting the accumulation of excitatory amino acid in the synaptic cleft and influencing the neurons, microglia, and astrocyte activity of surrounding cells.

3.3. TNF-α regulates glutamatergic receptor activity

As known, NF-κB and AP-1 are important transcriptional factors related to the TNF-α signaling cascade.⁵⁷ Central glucocorticoid receptors modulate downregulation of spinal EAAC1 after peripheral nerve injury through diminished NF-κB expression.⁵⁸ Intriguingly, we found that amitriptyline upregulates GT expressions, including GLAST, GLT-1, and EAAC1, through NF-κB-dependent activation.⁵⁹ At present, the knowledge of EAAC1 regulation by NF-κB pathway is still limited and controversial. Chronic morphine exposure induced E3 ligase activation via cAMP/PKA signaling and resulted in a ubiquitin–proteasome system that mediated degradation of EAAC1.⁶⁰ This means that more mechanisms are involved in the EAAC1 regulation after chronic morphine exposure; further studies are required to elucidate these underlying signaling mechanisms. Moreover, Sitcheran et al⁶¹ have demonstrated that TNF-α induces I kappa B degradation, which triggers NF-κB nuclear translocation and suppresses GLT-1 expression in H4 astroglioma cells. Nevertheless, differential effect of TNF-α on GLT-1 expression is suggested; it evokes GLT-1 expression in spinal microglia of rats by lipopolysaccharide treatment⁶² and also inhibits glutamate uptake by primary human fetal astrocyte.⁴⁶ The reason for this apparent discrepancy is uncertain. Similarly, Korn et al demonstrated that increases in TNF-α might reduce GLAST protein expression, in a dose-dependent manner, which was prevented by preincubation with a neutralizing TNF-α antibody. It was assumed that downregulation of GLAST resulted in an excess of extracellular glutamate and triggered excitotoxicity through NMDA receptors in neurons and AMPA/kainite receptors in oligodendrocytes, and they concluded that TNF-α played an important role in CNS inflammatory disorders via reduction of glutamate uptake.⁶³

In addition, TNF-α was demonstrated to enhance synaptic strength by increasing the surface expression of AMPA receptors in cultured hippocampal neurons; it was prevented by the blockade of the TNF-α signaling pathway.²⁰ Stellwagen et al⁶⁴ further demonstrated that TNF-α preferentially increased the synaptic expression of GluR2-lacking AMPA receptors that are more permeable to Ca²⁺ and simultaneously decreased the surface expression of GABA_A receptors, which resulted in a reduction of the inhibitory synaptic transmission. It seems that the net effect of TNF-α is to alter the balance of excitation and inhibition. In contrast, Beattie et al²⁰ found that TNF-α enhanced synaptic strength by increasing the expression of AMPA receptor, but not the NMDA receptor; they suggested that mobility of NMDA receptors was less than that of AMPA receptors due to differential attachment.⁶⁵ Induction of GluR1 expression of AMPA receptors by TNF-α increases the neuron vulnerability to excitotoxicity, which is mediated through acid sphingomyelinase and NF-κB pathway.⁶⁶ As already known, AMPA GluR2 expression affects calcium permeability through AMPA receptors and GluR2 subunit-lacking AMPA receptors will alter its synaptic function.⁶⁷ NMDA receptors are highly permeable to Ca²⁺, and the NR1 subunit of the NMDA receptor is an essential functional unit of the NMDA receptor.⁶⁸ As already mentioned, NMDA receptors are involved in the mechanism of morphine tolerance and NMDA receptor antagonists were demonstrated to inhibit morphine tolerance.10, 12, 69 Lim et al⁷⁰ observed an increase in the expression of the NMDA receptor NR1 subunit in the spinal cord dorsal horns of chronic morphine-infused rats, which was mediated by glucocorticoid receptor activation in a time-dependent manner. Moreover, Shimoyama et al⁷¹ also found that morphine tolerance was attenuated by the deletion of NMDA receptor NR1 subunit via intrathecal administration of antisense oligonucleotide. We therefore suggested that the increase in glutamate availability by TNF-α was due to the downregulation of membrane GTs, in accordance with the upregulation of surface AMPA and NMDA receptors in chronic morphine-infused rats, which contributed to the excessive activation of glutamatergic receptors and then to the reduction in the antinociceptive effect of morphine.⁷² Conceivably, activation of glutamatergic receptor system will enhance the intracellular mechanisms of morphine tolerance, such as interaction with μ-opioid receptors through Ca²⁺ influx, which activates PKC, NO synthase, and relevant gene regulation.¹¹

4. Conclusion

In summary, the effect of neuroinflammation in the development of morphine tolerance and the possible role of etanercept in the attenuation of morphine tolerance are shown in Fig. 1. Long-term morphine administration induces microglial activation and alters glutamatergic transmission in the spinal cord. Morphine enhances the expression of proinflammatory cytokines TNF-α, IL-1β, and IL-6 in spinal cord microglia (see ① in Fig. 1) and upregulates expressions of AMPA- and NMDA-receptor subunits on the postsynaptic neuronal plasma membrane (② in Fig. 1). It therefore inhibits the membraneous GT expression in the spinal astrocytes or postsynaptic neurons, including GLAST, GLT-1, and EACC1 (③ in Fig. 1), thus facilitating the TNF-α-related signaling. According to the results of our studies and others, we suggest that the TNF-α inhibitor etanercept may restore the membraneous GTs on the spinal astrocytes in particular (④ in Fig. 1) and reduce the accumulation of synaptic excitatory amino acids, subsequently inhibiting the upregulation of AMPA- and NMDA-receptor subunits on the postsynaptic neuron plasma membrane (⑤ in Fig. 1). Recent works have focused on research on morphine tolerance and the interaction between inflammation and glutamatergic receptor system function; studies have demonstrated that repeated morphine administration leads to an increase in the TNF-α expression in spinal cord microglia.³² Watkins et al⁷³ have suggested that chronic administration of opioids activates glia as facilitators of pain. The activated glia release neuroexcitatory substances including proinflammatory cytokines, and modulation of glial activity can be the target of treating morphine tolerance. The therapeutic potential of blockade of TNF-α activity is widely used in a variety of disorders of excessive inflammation and has been approved by the US Food and Drug Administration for clinical use. Modulation of glutamatergic transmission by TNF-α suggests that inhibition of TNF-α signaling provides another potential therapy of clinical pain management, particularly in patients who need long-term morphine administration or suffer from neuropathic pain. Etanercept reduced mechanical allodynia and attenuated membraneous expression of Ca²⁺-permeable AMPA receptors by reducing GluR1 membraneous insertion.⁷⁴ However, more information is needed about the intracellular TNF-α signaling pathway and the subtypes of TNF-α receptors that are involved. This review emphasizes the clinical role of a selective TNF-α inhibitor in pain management, in particular, in the attenuation of morphine tolerance.

Download full-size image

Fig. 1. Effect of neuroinflammation on and the possible role of etanercept in the development of morphine tolerance. EAAC = excitatory amino acid carrier; EAAT = excitatory amino acid transporter; IL = interleukin; GLAST = glutamate–aspartate transporter; NMDA = N-methyl-d-aspartate; TNF = tumor necrosis factor.