Xu-Hong Wei

Peripheral Nerve Injury Leads to Working Memory Deficits and Dysfunction of the Hippocampus by Upregulation of TNF-α in Rodents

Patients with chronic pain usually suffer from working memory deficits, which may decrease their intellectual ability significantly. Despite intensive clinical studies, the mechanism underlying this form of memory impairment remains elusive. In this study, we investigated this issue in the spared nerve injury (SNI) model of neuropathic pain, a most common form of chronic pain. We found that SNI impaired working memory and short-term memory in rats and mice. To explore the potential mechanisms, we studied synaptic transmission/plasticity in hippocampus, a brain region critically involved in memory function. We found that frequency facilitation, a presynaptic form of short-term plasticity, and long-term potentiation at CA3–CA1 synapses were impaired after SNI. Structurally, density of presynaptic boutons in hippocampal CA1 synapses was reduced significantly. At the molecular level, we found that tumor necrosis factor-α (TNF-α) increased in cerebrospinal fluid, in hippocampal tissue and in plasma after SNI. Intracerebroventricular or intrahippocampal injection of recombinant rat TNF mimicked the effects of SNI in naive rats, whereas inhibition of TNF-α or genetic deletion of TNF receptor 1 prevented both memory deficits and synaptic dysfunction induced by SNI. As TNF-α is critical for development of neuropathic pain, we suggested that the over-production of TNF-α following peripheral nerve injury might lead to neuropathic pain and memory deficits, simultaneously.

p38 activation in uninjured primary afferent neurons and in spinal microglia contributes to the development of neuropathic pain induced by selective motor fiber injury

Compelling evidence shows that the adjacent uninjured primary afferents play an important role in the development of neuropathic pain after nerve injury. The underlying mechanisms, however, are largely unknown. In the present study, the selective motor fiber injury was performed by L5 ventral root transection (L5 VRT), and p38 activation in dorsal root ganglia (DRG) and L5 spinal dorsal horn was examined. The results showed that phospho-p38 immunoreactivity (p-p38-IR) was increased in both L4 and L5 DRGs, starting on day 1 and persisting for nearly 3 weeks (P<0.05) following L5 VRT and that the activated p38 was confined in neurons, especially in IB4 positive C-type neurons. L5 VRT also induced p38 activation in L5 spinal dorsal horn, occurred at the first day after the lesion and lasted for 2 weeks (P<0.05). The activated p38 is restricted entirely in spinal microglia. In contrast, selective injury of sensory neurons by L5 dorsal root transection (L5 DRT) failed to induce behavioral signs of neuropathic pain and activated p38 only in L5 DRG but not in L4 DRG and L5 spinal dorsal horn. Intraperitoneal injection of thalidomide, an inhibitor of TNF-alpha synthesis, prevented p38 activation in DRG and spinal cord. Intrathecal injection of p38 inhibitor SB203580, starting before L5 VRT, inhibited the abnormal pain behaviors. Post-treatment with SB203580 performed at the 1st day or at the 8th day after surgery also reduced established neuropathic pain. These data suggest that p38 activation in uninjured DRGs neurons and in spinal microglia is necessary for the initiation and maintenance of neuropathic pain induced by L5 VRT.

TNF-α contributes to up-regulation of Nav1.3 and Nav1.8 in DRG neurons following motor fiber injury

&NA; A large body of evidence has demonstrated that the ectopic discharges of action potentials in primary afferents, resulted from the abnormal expression of voltage gated sodium channels (VGSCs) in dorsal root ganglion (DRG) neurons following peripheral nerve injury are important for the development of neuropathic pain. However, how nerve injury affects the expression of VGSCs is largely unknown. Here, we reported that selective injury of motor fibers by L5 ventral root transection (L5‐VRT) up‐regulated Nav1.3 and Nav1.8 at both mRNA and protein level and increased current densities of TTX‐S and TTX‐R channels in DRG neurons, suggesting that nerve injury may up‐regulate functional VGSCs in sensory neurons indirectly. As the up‐regulated Nav1.3 and Nav1.8 were highly co‐localized with TNF‐&agr;, we tested the hypothesis that the increased TNF‐&agr; may lead to over‐expression of the sodium channels. Indeed, we found that peri‐sciatic administration of recombinant rat TNF‐&agr; (rrTNF) without any nerve injury, which produced lasting mechanical allodynia, also up‐regulated Nav1.3 and Nav1.8 in DRG neurons in vivo and that rrTNF enhanced the expression of Nav1.3 and Nav1.8 in cultured adult rat DRG neurons in a dose‐dependent manner. Furthermore, inhibition of TNF‐&agr; synthesis, which prevented neuropathic pain, strongly inhibited the up‐regulation of Nav1.3 and Nav1.8. The up‐regulation of the both channels following L5‐VRT was significantly lower in TNF receptor 1 knockout mice than that in wild type mice. These data suggest that increased TNF‐&agr; may be responsible for up‐regulation of Nav1.3 and Nav1.8 in uninjured DRG neurons following nerve injury.

ATP induces long-term potentiation of C-fiber-evoked field potentials in spinal dorsal horn: The roles of P2X4 receptors and p38 MAPK in microglia

Many studies have shown that adenosine triphosphate (ATP), as a neurotransmitter, is involved in plastic changes of synaptic transmission in central nervous system. In the present study, we tested whether extracellular ATP can induce long‐term potentiation (LTP) of C‐fiber‐evoked field potentials in spinal dorsal horn. The results showed the following: (1) ATP at a concentration of 0.3 mM induced spinal LTP of C‐fiber‐evoked field potentials, lasting for at least 5 h; (2) spinal application of 2′,3′‐O‐(2,4,6‐trinitrophenyl)adenosine‐5‐triphosphate (TNP‐ATP; an antagonist of P2X1–4 receptors), but not pyridoxal‐phosphate‐6‐azophenyl‐2′,4′‐disulphonic acid (PPADS; an antagonist of P2X1,2,3,5,7 receptors), 30 min before ATP blocked ATP‐induced LTP, indicating that ATP may induce spinal LTP by activation of P2X4 receptors; (3) at 60 min after LTP induction the level of phospho‐p38 mitogen‐activated protein kinase (p‐p38 MAPK) was significantly elevated and at 180 min after LTP the number of P2X4 receptors increased significantly; both p‐p38 and P2X4 receptors were exclusively co‐located with the microglia marker, but not with neuronal or astrocyte marker; (4) spinal application of TNP‐ATP but not PPADS prevented p38 activation; (5) spinal application of SB203580, a p38 MAPK inhibitor, prevented both spinal LTP and the upregulation of P2X4 receptors. The results suggested that ATP may activate p38 MAPK by binding to intrinsic P2X4 receptors in microglia, and subsequently enhance the expression of P2X4 receptors, contributing to spinal LTP.

The up-regulation of IL-6 in DRG and spinal dorsal horn contributes to neuropathic pain following L5 ventral root transection.

Our previous works have shown that pro-inflammatory cytokine tumor necrosis factor-alpha (TNF-α) plays an important role in neuropathic pain produced by lumber 5 ventral root transection (L5-VRT). In the present work we evaluate the role of interleukin-6 (IL-6), another key inflammatory cytokine, in the L5-VRT model. We found that IL-6 was up-regulated in the ipsilateral L4 and L5 dorsal root ganglian (DRG) neurons and in bilateral lumbar spinal cord following L5-VRT. Double immunofluorescence stainings revealed that in DRGs the increased immunoreactivity (IR) of IL-6 was almost restricted in neuronal cells, while in the spinal dorsal horn IL-6-IR up-regulated in both glial cells (astrocyte and microglia) and neurons. Intrathecal administration of IL-6 neutralizing antibody significantly delayed the induction of mechanical allodynia in bilateral hindpaws after L5-VRT. Furthermore, inhibition of TNF-α synthesis by intraperitoneal thalidomide prevented both mechanical allodynia and the up-regulation of IL-6 in DRGs following L5-VRT. These data suggested that the increased IL-6 in afferent neurons and spinal cord contribute to the development of neuropathic pain following motor fiber injury, and that TNF-α is responsible for the up-regulation of IL-6.

The Upregulation of Translocator Protein (18 kDa) Promotes Recovery from Neuropathic Pain in Rats

At present, effective drug for treatment of neuropathic pain is still lacking. Recent studies have shown that the ligands of translocator protein (TSPO, 18 kDa), a peripheral receptor for benzodiazepine, modulate inflammatory pain. Here, we report that TSPO was upregulated in astrocytes and microglia in the ipsilateral spinal dorsal horn of rats following L5 spinal nerve ligation (L5 SNL), lasting until the vanishing of the behavioral signs of neuropathic pain (∼50 d). Importantly, a single intrathecal injection of specific TSPO agonists Ro5-4864 or FGIN-1-27 at 7 and 21 d after L5 SNL depressed the established mechanical allodynia and thermal hyperalgesia dramatically, and the effect was abolished by pretreatment with AMG, a neurosteroid synthesis inhibitor. Mechanically, Ro5-4864 substantially inhibited spinal astrocytes but not microglia, and reduced the production of tumor necrosis factor-α (TNF-α) in vivo and in vitro. The anti-neuroinflammatory effect was also prevented by AMG. Interestingly, TSPO expression returned to control levels or decreased substantially, when neuropathic pain healed naturally or was reversed by Ro5-4864, suggesting that the role of TSPO upregulation might be to promote recovery from the neurological disorder. Finally, the neuropathic pain and the upregulation of TSPO by L5 SNL were prevented by pharmacological blockage of Toll-like receptor 4 (TLR4). These data suggested that TSPO might be a novel therapeutic target for the treatment of neuropathic pain.

Activation of p38 signaling in the microglia in the nucleus accumbens contributes to the acquisition and maintenance of morphine-induced conditioned place preference

Several lines of evidence have suggested that activated glia contributes to morphine-induced reward (conditioned place preference, CPP). Compared to well-defined roles of astrocyte in morphine CPP, the role of microglia in the nucleus accumbens (NAc) remains poorly characterized. The aim of the present study was to investigate the distinct role of microglia in morphine-induced CPP. Systemic administration of morphine (7.5 mg/kg for 5 days) induced significant preference for the morphine-paired compartment in rats, which lasted for at least 6 days after cessation of morphine treatment. Immunohistochemistry results showed that activation of p38 in the NAc microglia induced by chronic morphine treatment maintained on day 11. Bilateral intra-NAc injection of minocycline, a putative microglia inhibitor, or SB203580, an inhibitor of p38, prior to morphine administration not only inhibited p38 activation in the microglia but impaired the acquisition of CPP. On the day following the acquisition of morphine CPP, a single injection of minocycline or SB203580 failed to block the expression of CPP. Notably, pretreatment with minocycline or SB203580 for 5 days following the acquisition of morphine CPP significantly suppressed the activation of p38 and attenuated the maintenance of morphine CPP. Collectively, our present study indicates that the p38 signaling in the NAc microglia may play an important role in the acquisition and maintenance but not the expression of morphine CPP, and provides new evidence that microglia might be a potential target for the therapy of morphine addiction.

Inhibition of NF-kappaB prevents mechanical allodynia induced by spinal ventral root transection and suppresses the re-expression of Nav1.3 in DRG neurons in vivo and in vitro.

Activation of nucleus factor-kappaB (NF-κB) in the dorsal root ganglia (DRG) is critical for development of neuropathic pain. The underlying mechanisms, however, are largely unknown. In the present work we tested if the activation of NF-κB is required for re-expression of Nav1.3, which is important for development of neuropathic pain, in uninjured DRG neurons. We found that intrathecal injection of pyrrolidine dithiocarbamate (PDTC), a NF-κB inhibitor, completely blocked the mechanical allodynia induced by L5 ventral root transection (L5-VRT), when applied 30 min before or 8h after operation, but at 7d after L5-VRT the same manipulation had no effect on established allodynia. Pre-treatment with PDTC also prevented the re-expression of Nav1.3 induced by L5-VRT. As our previous work has shown that up-regulation of tumor necrosis factor-alpha (TNF-α) in DRG is responsible for the re-expression of Nav1.3 in uninjured DRG neurons following L5 ventral root injury, we investigated whether activation of NF-κB is essential for the up-regulation of Nav1.3 by TNF-α. Results showed that application of rat recombinant TNF-α (rrTNF) into the cultured normal adult rat DRG neurons increased the immunoreactive (IR) of Nav1.3 localized mainly around the cell membrane and pre-treatment with PDTC blocked the change dose-dependently. The data suggested that injury to ventral root might lead to neuropathic pain and the re-expression of Nav1.3 in primary sensory neurons by activation of NF-κB.

Differential effects of adenosine A1 receptor on pain-related behavior in normal and nerve-injured rats

This study investigated the effects of N6-cyclopentyladenosine (CPA), a potent and selective adenosine A1 receptor (A1R) agonist in normal and nerve-injured rats and mechanisms of its action by behavioral tests and electrophysiological technique. The results showed: (1) In normal rats, intraperitoneal administration of CPA (1mg/kg) increased paw withdrawal latencies, in a way blocked by a selective A1R antagonist 8-cyclopentyl-1, 3-dipropylxanthine (DPCPX, 3mg/kg, i.p.), but had no influence on the threshold of mechanical stimulation. (2) In rats with neuropathic pain induced by spinal nerve ligation (SNL), CPA reduced thermal hyperalgesia and mechanical allodynia, which could last 6h and 10h, respectively (n=6/group, P<0.05). Both of the effects could be blocked by pretreatment of DPCPX intraperitoneally. (3) The baseline of C-fiber but not A-fiber evoked field potentials was depressed by spinal application of CPA (0.01 mM), and this effect was prevented by application of DPCPX (0.02 mM) 30 min before CPA. (4) Spinal application of CPA depressed long-term potentiation (LTP) of A- and C-fiber evoked field potentials, and both the depression could be blocked by pretreatment of DPCPX 30 min before CPA. These results suggested that the activation of A1R has different influences on normal and neuropathic rats probably due to the absence and presence of central sensitization in spinal dorsal horn.

Limited BDNF contributes to the failure of injury to skin afferents to produce a neuropathic pain condition

&NA; Although a large body of evidence has shown that peripheral nerve injury usually induces neuropathic pain, there are also clinical studies demonstrating that injury of the sural nerve, which almost only innervates skin, fails to do so. The underlying mechanism, however, is largely unknown. In the present work, we found that the transection of either the gastrocnemius–soleus (GS) nerve innervating skeletal muscle or tibial nerve supplying both muscle and skin, but not of the sural nerve produced a lasting mechanical allodynia and thermal hyperalgesia in adult rats. High‐frequency stimulation (HFS) or injury of either the tibial nerve or the GS nerve induced late‐phase long‐term potentiation (L‐LTP) of C‐fiber‐evoked field potentials in spinal dorsal horn, while HFS or injury of the sural nerve only induced early‐phase LTP (E‐LTP). Furthermore, HFS of the tibial nerve induced L‐LTP of C‐fiber responses evoked by the stimulation of the sural nerve and the heterotopic L‐LTP was completely prevented by spinal application of TrkB‐Fc (a BDNF scavenger). Spinal application of low dose BDNF (10 pg/ml) enabled HFS of the sural nerve to produce homotopic L‐LTP. Finally, we found that injury of the GS nerve but not that of the sural nerve up‐regulated BDNF in DRG neurons, and that the up‐regulation of BDNF occurred not only in injured neurons but also in many uninjured ones. Therefore, the sural nerve injury failing to produce neuropathic pain may be due to the nerve containing insufficient BDNF under both physiological and pathological conditions.