Systemic inflammation sensitizes the neonatal brain to excitotoxicity through a pro-/anti-inflammatory imbalance: Key role of TNFα pathway and protection by etanercept

Systemic inflammation sensitizes the perinatal brain to an ischemic/excitotoxic insult but the mechanisms are poorly understood. We hypothesized that the mechanisms involve an imbalance between pro- and anti-inflammatory factors.A well characterized mouse model where a systemic injection of IL-1β during the first five postnatal days (inflammatory insult) is combined with an intracerebral injection of the glutamatergic analogue ibotenate (excitotoxic insult) at postnatal day 5 was used.Following the inflammatory insult alone, there was a transient induction of IL-1β and TNFα, compared with controls measured by quantitative PCR, ELISA, and Western blot.Following the combined inflammatory and excitotoxic insult, there was an induction of IL-1β, TNFα, and IL-6 but not of IL-10 and TNFR1, indicating an altered pro-/anti-inflammatory balance after IL-1β sensitized lesion.We then tested the hypothesis that the TNFα pathway plays a key role in the sensitization and insult using TNFα blockade (etanercept) and TNFα^?/? mice. Etanercept given before the insult did not affect brain damage, but genetic deletion of TNFα or TNFα blockade by etanercept given after the combined inflammatory and excitotoxic insult reduced brain damage by 50%. We suggest this protective effect was centrally mediated, since systemic TNFα administration in the presence of an intact blood–brain barrier did not aggravate the damage and etanercept almost abolished cerebral TNFα production.In summary, sensitization was, at least partly, mediated by an imbalance between pro- and anti-inflammatory cytokines. Cerebral TNFα played a key role in mediating brain damage after the combined inflammatory and excitatory insult.     

Myelin contains neutral sphingomyelinase activity that is stimulated by tumor necrosis factor-α

Purified myelin from mouse brain was found to contain two forms of neutral sphingomyelinase, one Mg²⁺ dependent and the other Mg²⁺ independent. The former had a pH optimum of 7.5 and K_m of 0.35 mM, whereas the corresponding values for the latter were pH 8.0 and K_m 3.03 mM. Specific activity of the Mg²⁺-dependent enzyme showed a rostral-caudal gradient, ranging from 75 nmol/mg protein/hr in myelin from cerebral hemispheres to 21 nmol/mg protein/hr in myelin from spinal cord. Relative specific activity was approximately 20% that of brain stem or cerebral hemisphere homogenate. Treatment of myelin with taurocholate or high salt concentration did not significantly reduce activity of the Mg²⁺-dependent enzyme. The activity of that enzyme did not change with time or in the presence or absence of protease inhibitors; by contrast, that of the Mg²⁺-independent enzyme decreased sharply in the absence of protease inhibitors but rose in their presence. To test for the effect of tumor necrosis factor-α (TNFα) on myelin sphingomyelinase, mouse brain myelin was labeled in vivo by intracerebral injection of [³H]acetate into 18–20-day-old mice. After 40 hr, brain stems were removed, minced, and treated with TNFα in Krebs-Ringer solution, after which myelin was immediately isolated. Separation and counting of individual lipids revealed TNFα treatment to cause increased labeling of myelin ceramide and cholesterol ester with concomitant decrease in myelin sphingomyelin. Western blotting of myelin proteins using antibodies to the two TNFα receptors as probes revealed the presence of the p75 receptor. Implications of these findings in relation to possible mechanisms of autoimmune demyelination are discussed. J. Neurosci. Res. 50:466–476, 1997. © 1997 Wiley-Liss, Inc.     

Genetic contribution of the tumor necrosis factor region in guillain-barré syndrome

We studied genetic polymorphisms in the tumor necrosis factor (TNF) region in 81 Japanese patients with Guillain-Barré syndrome (GBS) and 85 controls. A significantly higher frequency of the 100–base pair (bp) (TNFa2) allele of the TNFa microsatellite marker, which is associated with high TNFα production, existed in Campylobacter jejuni-positive (Cj⁺) GBS patients than in controls, suggesting the involvement of a genetic predisposition to high TNFα secretion in the development of c. jejuni-related GBS.     

Myelination of axons within cytosine arabinoside treated mouse cerebellar explants by cultured rat oligodendrocytes

Cell suspensions of cultured purified rat oligodendrocytes prepared by the differential substrate adhesion method were applied to neonatal mouse cerebellar explant cultures in which myelination and oligodendrocyte maturation had been irreversibly inhibited by exposure to cytosine arabinoside. Myelination of Purkinje cell axons within 92% of the host explants was observed 2-5 days after oligodendrocyte application. Ultrastructurally, mature oligodendrocytes and axons surrounded by compact myelin, as well as spherules of compact myelin membranes without axons, were present within the cerebellar explants. It is evident that cultured dissociated purified oligodendrocytes retain the ability to myelinate appropriate axons. Such oligodendrocytes may be hyperreactive with regard to myelin membrane formation, as suggested by the presence of spheres of compact myelin without axons.     

Locomotor activity and anxiety status, but not spatial working memory, are affected in mice after brief exposure to cuprizone

Chronic long-term exposure to cuprizone causes severe brain demyelination in mice,which leads to changes in locomotion,working memory and anxiety.These findings suggest the importance of intact myelin for these behaviors.This study aimed to investigate the possible behavioral changes in mice with mild oligodendrocyte/myelin damage that parallels the white matter changes seen in the brains of patients with psychiatric disporders.We used the cuprizonetreated mouse model to test both tissue changes and behavioral functions（locomotor activity,anxiety status,and spatial working memory）.The results showed that mice given cuprizone in their diet for 7 days had no significant myelin breakdown as evaluated by immunohistochemical staining for myelin basic protein,while the number of mature oligodendrocytes was reduced.The number and length of Caspr protein clusters,a structural marker of the node of Ranvier,did not change.The locomotor activity of the cuprizonetreated mice increased whereas their anxiety levels were lower than in normal controls;spatial working memory,however,did not change.These results,for the first time,link emotion-related behavior with mild white matter damage in cuprizone-treated mice.     

Mechanism of phospholipase A2-induced conduction block in bullfrog sciatic nerve. I. Electrophysiology and morphology

The effects of exogenously added phospholipase A2 (PLA2) and its hydrolytic products in isolated bullfrog sciatic nerve were investigated. Nerves were pretreated for 3 h with a dose of trypsin which did not affect conduction in order to enhance penetration of the added agents. Treatment of nerves with beta-glucosidase, neuraminidase or chymotrypsin had no effect on conduction. Whereas incubation of the nerves with normal Ringers for 2 h had no significant effect on conduction, incubation with PLA2 in Ringers caused decrements in the height of the compound action potential in a dose-related manner. In addition, incubation of the nerves with 10 mg/ml lysolecithin, arachidonic acid, or docosahexaenoic acid caused marked decrements in the height of the compound action potential. Electron microscopic analysis of nerves after each treatment which caused conduction block revealed varying levels of myelin damage. Although myelin was damaged at the paranodal and/or internodal region, depending on the agents used, the axonal membrane appeared to be intact at the ultrastructural level. It was concluded that the block in conduction resulting from PLA2 was due to the formation of lysolecithin and long chain polyunsaturated fatty acids.     

The octapeptide repeat PrP(C) region and cobalamin-deficient polyneuropathy of the rat

Introduction: Cobalamin (Cbl) deficiency affects the peripheral nervous system (PNS) morphologically and functionally. We investigated whether the octapeptide repeat (OR) region of prion protein (PrP^C) (which is claimed to have myelinotrophic properties) is involved in the pathogenesis of rat Cbl‐deficient (Cbl‐D) polyneuropathy. Methods: We intracerebroventricularly administered antibodies (Abs) against the OR region (OR‐Abs) to Cbl‐D rats to prevent myelin damage and maximum nerve conduction velocity (MNCV) abnormalities, and PrP^Cs to normal rats to reproduce PNS Cbl‐D–like lesions. We measured nerve PrP^C levels and MNCV. Results: The OR‐Abs normalized myelin ultrastructure, MNCV values, and tumor necrosis factor (TNF)‐α levels in the sciatic and tibial nerves of Cbl‐D rats. PrP^C levels increased in Cbl‐D nerves. The nerves of the PrP^C‐treated rats showed typical Cbl‐D lesions, significantly decreased MNCV values, and significantly increased TNF‐α levels. Conclusions: OR‐Abs prevent the myelin damage caused by increased OR regions, and excess TNF‐α is involved in the pathogenesis of Cbl‐D polyneuropathy. Muscle Nerve 2011     

Differential roles of TNFR1 and TNFR2 signaling in adult hippocampal neurogenesis

Tumor necrosis factor alpha (TNFα) is a potent inhibitor of neurogenesis in vitro but here we show that TNFα signaling has both positive and negative effects on neurogenesis in vivo and is required to moderate the negative impact of cranial irradiation on hippocampal neurogenesis. In vitro, basal levels of TNFα signaling through TNFR2 are required for normal neural progenitor cell proliferation while basal signaling through TNFR1 impairs neural progenitor proliferation. TNFR1 also mediates further reductions in proliferation and elevated cell death following exposure to recombinant TNFα. In vivo, TNFR1^−/− and TNFα^−/− animals have elevated baseline neurogenesis in the hippocampus, whereas absence of TNFR2 decreases baseline neurogenesis. TNFα is also implicated in defects in neurogenesis that follow radiation injury but we find that loss of TNFR1 has no protective effects on neurogenesis and loss of TNFα or TNFR2 worsened the effects of radiation injury on neurogenesis. We conclude that the immunomodulatory signaling of TNFα mediated by TNFR2 is more significant to radiation injury outcome than the proinflammatory signaling mediated through TNFR1.     

Rat optic nerve: electrophysiological, pharmacological and anatomical studies during development

Changes in conduction properties and in morphology were studied during rat optic nerve growth from birth (when no myelin is present and the glia have not differentiated) to adulthood (when the optic nerve is essentially 100% myelinated). Myelination begins around the sixth postnatal day and proceeds rapidly so that 85% of the fibers are myelinated at 28 days of age. Mean diameter of optic nerve axons remains about 0.2 micron for the first week and then increases rapidly if the fiber is being myelinated. Those axons not being myelinated remain about 0.2-0.3 micron in diameter. At birth the compound action potential has a single negative peak and a conduction velocity of about 0.2 m/s. The increase in conduction velocity prior to myelination is considerably greater than can be accounted for on the basis of increase in axonal diameter. There is no clear step increase in the velocity of the shortest latency peak correlated with the onset of myelination. During myelination the compound action potential develops multiple short latency components, which evolve into the adult-like 3 component compound action potential by 3-4 weeks of age. Durations of the relative refractory period and supernormal period decrease as age increases, but are not related to myelination in a simple manner. Sodium appears to be the only significant carrier of inward current at all ages. A measureable calcium conductance is not present at any age. Voltage-dependent potassium conductance contributes to the compound action potential at all ages, but the response to 4-aminopyridine in rapidly conducting fibers is apparently smaller than that in slowly conducting fibers. These results show that conduction can occur before myelination or the differentiation of glial cells. Moreover, changes in conduction velocity do not depend entirely on myelination or increases in axonal size. Finally, these results suggest a reorganization of axonal membrane properties during the development of rat optic nerve.     

Mechanism of phospholipase A2-induced conduction block in bullfrog

The effects of exogenously added phospholipase A₂ (PLA₂) and its hydrolytic products in isolated bullfrog sciatric nerve were investigated. Nerves were pretreated for 3 h with a dose of trypsin which did not affect conduction in order to enhance penetration of the added agents. Threatment of nerves with β-glucosidase, neuraminidase or chymotrypsin had no effect on conduction. Whereas incubatin of the nerves with normal Ringers for 2 h had no significant effect on conduction, incubation with PAL₂ in Ringers caused decrements in the height of the compound action potential in dose-related manner. In addition, incubation of the nerves with 10 mg/ml lysolecithin, arachidonic acid, or docosahexaenoic acid caused marked decrements in the height of the compound action potential. Electron microscopic analysis of nerves after each treatment which caused conduction block revealed varying levels of myelin damage. Although myelin was damaged at the paranodal and/or internodal region, depending on the agents used, the axonal membrane appeared to be intact at the ultrastructural level. It was concluded that the block in conduction resulting from PLA₂ was due to the formation of lysolecithin and long chain polyunsaturated fatty acids.     

Assessment of axonal dysfunction in an in vitro model of acute compressive injury to adult rat spinal cord axons

An in vitro model of spinal cord injury was developed to study the pathophysiology of posttraumatic axonal dysfunction. A 25 mm length of thoracic spinal cord was removed from the adult male rat (n = 27). A dorsal column segment was isolated and pinned in a recording chamber and superfused with oxygenated (95%O₂/5% CO₂) Ringer. The cord was stimulated with a bipolar electrode, while two point responses were recorded extracellularly. Injury was accomplished by compression with a modified aneurysm clip which applied a 2 g force for 15 s. With injury the compound action potential (CAP) amplitude decreased to 53.7 ± 5.4% (P < 0.001), while the latency increased to 115.6 ± 3.1% (P < 0.0025) of control values. The absolute refractory period increased with injury from 1.7 ± 0.1 ms to 2.1 ± 0.1 ms (P < 0.001). With train stimulation (200 and 400 Hz), injured axons showed evidence of high frequency conduction failure (P < 0.05). The infusion of 5 mM 4-aminopyridine (4-AP), a blocker of voltage-sensitive ‘fast’ K channels confined to internodal regions, resulted in broadening of the CAP of injured axons to 114.9 ± 3.1% of control (P < 0.05). Ultrastructural analysis of the injured dorsal column segments revealed marked axonal and myelin pathology, including considerable myelin disruption.In conclusion, we have developed and characterized an in vitro model of mammalian spinal cord injury which simulates many of the features of in vivo trauma. Injured axons display characteristic changes in physiological function including a shift in refractory period and high frequency conduction failure. The ultrastructural data and response of injured axons to 4-AP suggest that myelin disruption with exposure of ‘fast’ K⁺ channels contributes to posttraumatic axonal dysfunction.     

Tumour necrosis factor α induces neuroinflammation and insulin resistance in immortalised hypothalamic neurones through independent pathways

The links between obesity, inflammation and insulin resistance, which are all key characteristics of type 2 diabetes mellitus, are yet to be delineated in the brain. One of the key neuroinflammatory proteins detected in the hypothalamus with over‐nutrition is tumour necrosis factor (TNF)α. Using immortalised embryonic rat and mouse hypothalamic cell lines (rHypoE‐7 and mHypoE‐46) that express orexigenic neuropeptide Y and agouti‐related peptide, we investigated changes in insulin signalling and inflammatory gene marker mRNA expression after TNFα exposure. A quantitative polymerase chain reaction array of 84 inflammatory markers (cytokines, chemokines and receptors) demonstrated an increase in the expression of multiple genes encoding inflammatory markers upon exposure to 100 ng mL^‐1 TNFα for 4 hours. Furthermore, neurones pre‐exposed to TNFα (50 ng mL^‐1) for 6 or 16 hours exhibited a significant reduction in phosphorylated Akt compared to control after insulin treatment, indicating the attenuation of insulin signalling. mRNA expression of insulin signalling‐related genes was also decreased with exposure to TNFα. TNFα significantly increased mRNA expression of IκBα, Tnfrsf1a and IL6 at 4 and 24 hours, activating a pro‐inflammatory state. An inhibitor study using an inhibitor of nuclear factor kappa B kinase subunit β (IKK‐β) inhibitor, PS1145, demonstrated that TNFα‐induced neuroinflammatory marker expression occurs through the IKK‐β/nuclear factor‐kappa B pathway, whereas oleate, a monounsaturated fatty acid, had no effect on inflammatory markers. To test the efficacy of anti‐inflammatory treatment to reverse insulin resistance, neurones were treated with TNFα and PS1145, which did not significantly restore the TNFα‐induced changes in cellular insulin sensitivity, indicating that an alternative pathway may be involved. In conclusion, exposure to the inflammatory cytokine TNFα causes cellular insulin resistance and inflammation marker expression in the rHypoE‐7 and mHypoE‐46 neurones, consistent with effects seen with TNFα in peripheral tissues. It also mimics insulin‐ and palmitate‐induced insulin resistance in hypothalamic neurones. The present study provides further evidence that altered central energy metabolism may be caused by obesity‐induced cytokine expression.     

Circulating tumor necrosis factor-α correlates with electrodiagnostic abnormalities in Guillain-Barré syndrome

Autoimmune damage to peripheral nerves, mediated by activated T lymphocytes and macrophages, underlies the pathogenesis of inflammatory demyelination in Guillain-Barré syndrome. Both T lymphocytes and macrophages secrete tumor necrosis factor-α, a cytokine that exerts toxic effects on myelin, Schwann cells, and endothelial cells. The reportedly high serum levels of this cytokine in patients with Guillain-Barré syndrome may reflect the degree of immune activation rather than a direct pathogenic effect. We compared serum levels of tumor necrosis factor-α, interleukin-1β, and soluble interleukin-2 receptor with well-established electrodiagnostic criteria for primary demyelination in 23 patients with Guillain-Barré syndrome, to assess the relationship between these cytokines and peripheral myelin damage. High serum levels of tumor necrosis factor-α were associated with prolonged distal motor latencies and slowed motor conduction velocities, prolonged or absent F-wave responses, and reduced amplitude of distal compound muscle action potentials. No significant correlation was observed between electrodiagnostic criteria for primary demyelination and serum levels of interleukin-1β or soluble interleukin-2 receptor. These findings suggest a putative role of tumor necrosis factor-α in the pathogenesis of peripheral nerve demyelination in Guillain-Barré syndrome.     

Adolescent binge alcohol exposure induces long-lasting partial activation of microglia

Accumulating evidence indicates that the adolescent hippocampus is highly susceptible to alcohol-induced structural damage and behavioral deficits. Microglia are vitally important brain constituents needed to support and maintain proper neural function; however, alcohol's effects on microglia have only recently gained attention. The microglial response to alcohol during adolescence has yet to be studied; therefore, we examined hippocampal microglial activation in an adolescence binge alcohol exposure model. Adolescent male Sprague-Dawley rats were administered ethanol 3 times/day for 4 days and were sacrificed 2, 7, and 30 days later. Bromo-deoxy-Uridine was injected 2 days after ethanol exposure to label dividing cells. Microglia morphology was scored using the microglia marker Iba-1, while the extent of microglial activation was examined with ED-1, major histocompatibility complex-II (MHC-II), and tumor necrosis factor (TNF)-α expression. Ethanol induced significant morphological change in hippocampal microglia, consistent with activation. In addition, ethanol increased the number of BrdU+ cells throughout all regions of the hippocampus 2 days after the last dose. Confocal microscopy showed that the proliferating BrdU+ cells in each region were Iba-1+ microglia. Importantly, newly born microglia survived and retained their morphological characteristics 30 days after ethanol exposure. Ethanol did not alter hippocampal ED-1, MHC-II, or TNF-α expression, suggesting that a single period of binge ethanol exposure does not induce a full microglial-driven neuroinflammatory response. These results establish that ethanol triggers partial microglial activation in the adolescent hippocampus that persists through early adulthood, suggesting that alcohol exposure during this unique developmental time period has long-lasting consequences.     

一种新的基质金属蛋白酶抑制剂ONO-4817对实验性自身免疫性脑脊髓炎的治疗作用

目的　探讨一种新合成的含氧肟酸的基质金属蛋白酶 ( MMP)抑制剂 ONO-481 7对实验性自身免疫性脑脊髓炎 ( EAE)的治疗效果。方法　给 EAE大鼠口服 ONO-481 7,观察临床症状、T淋巴细胞增殖以及血清肿瘤坏死因子 ( TNF) -α水平。结果　 ONO-481 7能显著改善 EAE临床症状 ( P <0 .0 1 ) ,同时明显抑制 T淋巴细胞增殖 ( P <0 .0 1 ) ,显著降低大鼠血清 TNF-α水平 ( P <0 .0 5 )。结论　研究表明 ,ONO-481 7通过抑制 MMPs活性、T淋巴细胞增殖和减少 TNF-α生成 ,进而能显著减轻血脑屏障 ( BBB)的破坏 ,又可抑制炎细胞浸润和髓鞘破坏 ,从而有效缓解 EAE。     

Appearance of cytokine-associated central nervous system myelin damage coincides temporally with serum tumor necrosis factor induction after recombinant interleukin-2 infusion in rats

Endogenous tumor necrosis factor (TNF) activity, assessed by L-929 fibroblast bioassay, was determined in serum samples from rats infused intravenously with recombinant interleukin-2 (rIL-2) or rIL-2 vehicle. Parallel studies of cerebral ultrastructure were conducted in additional rats, comparably infused. Rats received rIL-2 or vehicle either one time only or 3 times daily for 3 days. TNF activity was assessed at 2, 4, and 8 h after the single or final infusion. Rats employed for ultrastructural studies were sacrificed at 4 h after the single or final infusion. Every rIL-2-infused rat exhibited unusual abnormalities of axonal ultrastructure, identical to those previously described after in vitro TNF application to living spinal cord slices. Serum samples drawn during and after the development of axonal changes revealed significantly elevated circulating TNF activity. Controls exhibited neither TNF activity nor altered axons. These studies demonstrate that, following rIL-2 infusion in rats, endogenous TNF circulates at elevated levels during the development of rIL-2-related central nervous system abnormalities similar to those produced in vitro by recombinant TNF. Whether rIL-2-induced circulating TNF is causally-related to the observed myelin damage remains to be determined but merits further investigation, particularly since blood-brain barrier function has been shown to be compromised following rIL-2 infusion.     

Different pathogenicity of encephalitic togaviruses in organotypic cultures of spinal cord slices.

The pathogenicity of two encephalitic Togaviruses, Sindbis virus (SV), an alphavirus, and West Nile virus (WNV), a flavivirus, was studied in organotypic cultures of fetal mouse spinal cord slices grown in roller tubes. After about 3 weeks in vitro, during which time the cultures became abundantly myelinated, they were infected either by 5 X 10(5) PFU SV or by 5 X 10(6) PFU WNV per culture. The viruses caused different patterns of cytopathogenicity: SV induced severe cytotoxicity in all glia cells and neurons with concomitant demyelination within 48 hr. In contrast, WNV, even 4 days after infection, caused only mild cytopathic effects mainly to neurons and astrocytes and a slight degree of damage to the myelin sheath. A most remarkable finding was the entrapment of WNV particles in the interperiod lines of the myelin sheaths. Treatment of cultures with mouse alpha and beta interferon prior to their infection with either virus protected the cultures from any viral damage. Long-term exposure of non-infected control organotypic cultures of fetal spinal cord slices to mouse interferons had no significant effect on neuronal and glial differentiation, and myelin formation.     

Peripheral sensory abnormalities in patients with multiple sclerosis.

Although multiple sclerosis primarily affects myelin within the central nervous system, both pathologic and physiological studies suggest that mild deficits in peripheral nervous system myelin may be common. To evaluate this question further, we performed near nerve studies on sural nerves of 14 patients with multiple sclerosis. Peak-to-peak amplitude and maximum conduction velocity were normal in 9 of 14 patients, while minimum conduction velocity, or the velocity of the slowest-conducting component of the sensory action potential, was abnormally reduced in 9 patients. In addition, the supernormal period was evaluated for patients and compared with a control sample; multiple sclerosis patients showed a significant reduction in the amplitude of supernormality. Both the reduction in minimum conduction velocity and the alteration in the supernormal period are consistent with a mild defect in peripheral myelin.