Ibudilast suppresses TNFα production by glial cells functioning mainly as type III phosphodiesterase inhibitor in the CNS

Tumor necrosis factor α (TNFα) is considered to play a critical role in the development of various pathological processes in the central nervous system (CNS), such as neuronal degeneration, demyelination and HIV-related pathology. In order to search for the agents which suppress TNFα production in the CNS for future treatment of these pathological conditions, we examined the effects of ibudilast on TNFα production by murine microglia and astrocytes. Some actions of ibudilast are reportedly mediated by inhibition of type IV phosphodiesterase (PDE). Type IV PDE inhibitor has been shown to be the most effective for experimental autoimmune inflammatory demyelination. Therefore, we also determined the subtype of PDE inhibited by ibudilast. Ibudilast significantly and selectively suppressed TNFα production by microglia in a dose-dependent manner, without affecting their viability. The inhibition assay indicated that ibudilast is a rather selective inhibitor for type III PDE purified from brain, heart and kidney with moderate inhibitory activity against types I, II and IV PDEs from various tissues. Although it required 10 μM or higher concentrations to effectively suppress TNFα production in vitro, the combination of ibudilast with other subtypes of PDE inhibitors synergistically suppressed TNFα and nitric oxide production by microglia at 1 μM, a similar concentration that could be obtained in vivo at usual therapeutic dose. Thus, ibudilast, when used in a combination with other PDE inhibitors, will be useful for future strategies to treat intractable neurological diseases in which TNFα may play a causative role.     

Acute but not chronic stimulation of glial cells in rat spinal cord by systemic injection of lipopolysaccharide is associated with hyperalgesia

We have analyzed development of mechanical hyperalgesia after repeated systemic lipopolysaccharide (LPS) injections and correlated these findings with stimulation of astrocytes and microglia in spinal cord. Male Lewis rats received a single or seven intraperitoneal injections of LPS. Mechanical hyperalgesia was measured as rat hindpaw withdrawal thresholds (PWTs). We observed that a single LPS injection elicited a specific change of PWTs while stimulated spinal glial activation was identified by immunoreactivities of specific markers, ED1, P2X4 receptor, endothelial monocyte activating polypeptide II (EMAP II) and glial fibrillary acidic protein (GFAP), respectively; multiple LPS treatments induced tolerance to mechanical hyperalgesia, whereas expression of ED1 and GFAP were further increased. In conclusion, we have demonstrated that the number of activated spinal glial cells was increased as an acute effect of LPS correlating with increased sensitivity to mechanical stimulation. However chronic exposure to LPS can develop a tolerance to mechanical hyperalgesia despite ongoing signs of CNS glial activation.     

Glial cell-specific mechanisms of TGF-β1 induction by IL-1 in cerebral cortex

Transforming growth factor beta-1 (TGF-β1) immunoreactive product (IRP) has recently been detected in autopsied brains of individuals who died with central nervous system diseases and/or fever but not in normal individuals or in normal rodent brain. However, the mechanism(s) of induction of TGF-β1 in brain and the identity of cells expressing TGF-β1 need to be understood before a role, if any, for this potent pleiotropic cytokine in neuropathogenesis can be discerned. Towards this end we determined that IL-1 stimulated the production of TGF-β1 IRP in cells and TGF-β1 activity in culture fluids of all glial cells, astrocytes, microglial cells, and oligodendrocytes, derived from neonatal rat cortex and grown in cell type-enriched cultures. TGF-β1 production in vitro varied with the cell type and isoform of IL-1. Oligodendrocytes produced the most and astrocytes the least amount of TGF-β1. IL-1α stimulated TGF-β1 production in all glial cell types, whereas IL-1β did not. In vivo, TGF-β1 IRP was detected in human tissues from cerebral frontal cortex and subcortical white matter only when interleukin-1 (IL-1) was elevated in the same tissues. Moreover, the amount of detectable TGF-β1 was positively correlated with the amount of detectable IL-1 (rho = 0.605; P = 0.003), as determined by morphometry. Double-labelling of cells for their phenotypic markers and expression of TGF-β1 indicated that all glial cells, but not neurons, expressed TGF-β1. IL-1α and IL-1β IRPs were also detected in all three glial cell types, most frequently in astrocytes and least frequently in microglial cells. The cells containing both cytokine IRPs were also detected. These results indicate that TGF-β1 may be induced by IL-1 in all glial cells of the frontal cortex, by both autocrine and paracrine mechanisms.     

Intravascular infusions of soluble β-amyloid compromise the blood–brain barrier, activate CNS glial cells and induce peripheral hemorrhage

Vascular wall levels of soluble β-amyloid_1–40 (Aβ_1–40) are elevated in Alzheimer's disease (AD). Moreover, plasma Aβ levels are increased in familial AD, as well as in some cases of sporadic AD. To determine the histopathologic and behavioral consequences of elevated vascular Aβ levels, Aβ_1–40 (50 μg in distilled water) or vehicle was intravenously infused twice daily into 3-month old male Sprague–Dawley rats for 2 weeks. Intravenous Aβ infusions impaired blood–brain barrier integrity, as indicated by substantial perivascular and parenchyma IgG immunostaining within the brain. Also evident in Aβ-infused animals was an increase in GFAP immunostaining around cerebral blood vessels, and an enhancement of OX-42 microglial immunostaining in brain white matter. Gross pulmonary hemorrhage was noted in most Aβ-infused animals. All the observed changes occurred in the absence of Congo red birefringence. No significant cognitive deficits were present in Aβ-infused animals during water maze acquisition and retention testing, which was conducted during the second week of treatment. These results indicate that circulating Aβ can: (1) induce vessel dysfunction/damage in both the brain and the periphery without complex Aβ fibril formation/deposition, and (2) induce an activation of brain astrocytes and microglia. Taken together, our results suggest that if circulating Aβ is elevated in AD, it is likely to have a pathophysiologic role.     

Effects of TNFα on immature and mature oligodendrocytes and their progenitors in vitro

Tumor necrosis factor α (TNFα) appears to take part in the pathogenesis of multiple sclerosis and to contribute to the degeneration of oligodendrocytes as well as neurons. TNFα is produced by microglia and astrocytes, which also produce hormones and cytokines that influence its biological activity. Thus, in mixed cultures the effects of exogenous TNFα might be modified by products of astrocytes and microglia. The effects of TNFα in oligodendrocyte-enriched cultures are reported below. We prepared the cultures by shaking oligodendrocytes off primary mixed glial-cell cultures from brains of 2-day-old rats at 7 days in vitro and plating them (0 days post-shake, DPS). Platelet-derived growth factor and fibroblast growth factor were included in the media at 1–5 DPS in order to encourage proliferation. At 2 DPS media were added with no TNFα (controls) or 1000, 2000 or 5000 U/ml of TNFα, and at 5 DPS media were replaced with fresh serum-free media. Cultures were fixed with 4% paraformaldehyde at 5, 7, 9 and 12 DPS and immunostained. Oligodendrocyte progenitors were not reduced in numbers immediately after the incubation with TNFα (i.e. at 5 DPS). However, after an additional 4 days in culture fewer progenitors remained in the cultures that had been treated with TNFα than in the untreated cultures. In the absence of the growth factors there were fewer progenitors, but their numbers also were reduced by TNFα. Maturation to the myelin basic protein (MBP)-positive stage was inhibited by about 36% at 9 DPS by 1000–2000 U/ml of TNFα, while numbers of O4+/MBP− precursors were unaffected. It is interesting that the steady-state number of O4-positive precursors was unchanged by TNFα at 9 DPS, when there were reductions in the numbers of A2B5-positive progenitors and MBP-positive mature oligodendrocytes. That observation suggests that the rates of proliferation, death and maturation are controlled by multiple factors, with a particularly vulnerable time at the maturation to the MBP-positive stage. At 5000 U/ml TNFα the specific effect on maturation was overtaken cytotoxicity. These data and a summary of the literature suggest that inhibition of MBP expression is sensitive to lower TNFα concentrations and incubation times than is cell survival. Specific effects on numbers of MBP-positive cells, morphology and MBP expression occur at 1000–2000 U/ml for 48–72 h or at up to 10 000 U/ml for≤24 h, and the deficits remain after removal of the TNFα.     

Interferon-γ inhibits DNA synthesis and insulin-like growth factor-II expression in human neuroblastoma cells

Interferon-γ (IFN-γ) is known to be an antiproliferative, differentiation agent in many cell types, including neuroblastoma. In this study, we determined the effects of IFN-γ on cellular growth and expression of insulin-like growth factor II (IGF-II) and IGF receptors in the human neuroblastoma cell line SH-SY5Y. Incubation of SH-SY5Y cells in IFN-γ (20–100 U/ml) induced the formation of long neuritic processes. IFN-γ treatment also induced decreases in [³H]TdR incorporation, as well as serum-dependent changes in cell number. Treatment with IFN-γ reduced cell number 33% in the presence of serum but had no effect on cell number in the absence of serum. IGF-II mRNA content was 60% inhibited by IFN-γ, and was not serum dependent. The concentration of immunoreactive IGF-II in SH-SY5Y conditioned medium was also reduced in the presence of IFN-γ, to less than half of control levels. In contrast, type I IGF receptor mRNA content was increased more than three-fold after treatment with IFN-γ and serum. Co-incubation in IFN-γ (20–100 U/ml) and IGF-II on (3–10 nM) prevented the inhibitory effects of IFN-γ on [³H]TdR ncorporation in serum-free media. Our results suggest that IFN-γ may inhibit DNA synthesis and cell growth by interfering with an IGF-II/type I IGF receptor autocrine growth or survival mechanism.     

Protection against blood–brain barrier disruption in focal cerebral ischemia by the type IV phosphodiesterase inhibitor BBB022: a quantitative study

We examined the effect of BBB022, a type IV phosphodiesterase inhibitor, on blood–brain barrier (BBB) integrity after transient middle cerebral artery occlusion (MCAo) in rats. Male Sprague–Dawley rats were anesthetized with halothane and subjected to 120 min of temporary MCAo by retrograde intraluminal insertion of a nylon suture coated with poly-

Full-size image

-lysine. The drug (BBB022 in saline, 1 mg kg⁻¹ h⁻¹, i.v.) or vehicle (0.9% saline, 1–2 ml kg⁻¹ h⁻¹) was administered by infusion after the onset of MCAo. Four animal groups were studied: Groups A and B were treated by infusion of vehicle or drug over 5 h, and groups C and D over 48 h. Damage to the BBB was judged by extravasation of Evans blue (EB) dye, which was administered i.v. at 3 h after the onset of MCAo in groups A and B; and at 46 h in groups C and D. Fluorometric quantitation of EB was performed 1 or 2 h later in six brain regions. In the 5-h infusion series (group B), BBB022 decreased dye extravasation in the ipsilateral cortex, striatum and hemisphere (hemisphere mean±S.E.M.: 41.2±5.4 vs. 82.4±9.2 μg/g, p=0.005) compared to the vehicle-treated group (A). The 48-h infusion of BBB022 (group D) also decreased dye extravasation in the ipsilateral cortex (7.4±2.5 vs. 29.0±8.3 μg/g, p=0.05), striatum (17.2±2.2 vs. 50.8±12.1 μg/g, p=0.03) and hemisphere (30.7±4.0 vs. 93.2±18 μg/g, p=0.01) compared to the vehicle-treated group (C). BBB022 also significantly improved the neurological score at 3 and 5 h after MCAo (in the 5-h infusion group) and at 60 min, 24 h and 48 h (in the 48-h infusion group) compared to the vehicle groups. These data indicate that BBB022 prevents ischemic damage to the BBB after focal cerebral ischemia in rats.     

5α-Reductase activity in isolated and cultured neuronal and glial cells of the rat

The distribution of the 5α-reductase, the enzyme which converts testosterone into its ‘active’ metabolite dihydrotestosterone (DHT), has been studied in neurons, astrocytes and oligodendrocytes isolated from the brain of male rats by density gradient ultracentrifugation and in neurons and glial cells grown in cultures. Purity of cellular preparations was examined by electron and light microscopy. Purified neurons, astrocytes and oligodendrocytes, obtained from the brain of adult male rats, are all able to form DHT from testosterone and consequently possess a 5α-reductase activity. Among the 3 cell types studied, neurons appear to be more active than oligodendrocytes and astrocytes. Moreover, between the two population of glial cells, the oligodendrocytes seem to possess a slightly higher enzymatic activity than that present in the astrocytes. Neurons appeared more active in metabolizing testosterone than glial cells also in cell culture experiments. It is presently believed that the 5α-reduction of testosterone to DHT provides one of the mechanisms through which the hormone becomes effective in the CNS. This is supported by the present findings, which indicate that neurons are the cell population in which the 5α-reductase is more concentrated. However, the presence of a considerable 5α-reductase activity in glial cells indicates that also non-neuronal cells might participate in androgen-mediated events occurring in the brain.     

Interferon-γ secretion by peripheral blood T-cell subsets in multiple sclerosis: Correlation with disease phase and interferon-β therapy

Interferon-γ (IFN-γ) is implicated as a participant in the immune effector and regulatory mechanisms considered to mediate the pathogenesis of multiple sclerosis (MS). We have used an intracellular cytokine staining technique to demonstrate that the proportion of ex vivo peripheral blood CD4 and CD8 T-cell subsets expressing IFN-γ is increased in secondary progressing (SP) MS patients, whereas the values in untreated relapsing-remitting (RR) MS patients are reduced compared with those of controls. Patients treated with interferon-β (IFN-β) have an even more significant reduction in the percentage of IFN-γ–secreting cells. The finding that the number of IFN-γ–expressing CD8 cells is increased in SPMS patients, a group with reduced functional suppressor activity, and is most significantly reduced by IFN-β therapy, which increases suppressor activity, indicates that IFN-γ secretion by CD8 T cells and functional suppressor defects attributed to this cell subset in MS can be dissociated. Ann Neurol 1999;45:247–250     

β-Amyloid stimulates glial cells in vitro to produce growth factors that accumulate in senile plaques in Alzheimer's disease

The effects of a synthetic homolog of β-amyloid (β1–42) on the secretion of interleukin-1 (IL-1) and basic fibroblast growth factor (bFGF) from cultures of microglia and astrocytes, cells that surround β-amyloid-containing plaques in Alzheimer's disease, were examined. Our results show that β-amyloid not only enhances glial cell secretion of these factors, it stimulates the proliferation and morphological transformation of microglia. Since IL-1 and bFGF are known to elevate the synthesis of the β-amyloid precursor protein and other plaque components, it is suggested that in this way, cascades may arise that contribute to the process of plaque development.     

Induction of human immunodeficiency virus type 1 replication in human glial cells after proinflammatory cytokines stimulation: Effect of IFNγ, IL1β, and TNFα on differentiation and chemokine production in glial cells

Although evidence for human immunodeficiency virus 1 (HIV-1) presence in the central nervous system (CNS) of infected patients is well established, the intensity of viral replication within the brain is not usually known. In vitro, human embryonic microglial cells internalized HIV-1 through a CD4-dependent pathway but were not permissive to viral replication. We observed that HIV replication was induced when CNS cell cultures were stimulated for 14 days by a combination of proinflammatory cytokines including IFNγ, IL1β, and TNFα. After long-term cytokine stimulation, morphologically differentiated glial cells appeared, in which HIV-1 tat antigen was detected after infection. Thus, variations in the stage of maturation/activation of CNS cells under inflammatory conditions probably play a major role in facilitating massive production of HIV-1. We then studied the effect of prolonged cytokine stimulation on the secretion of inflammatory mediators by glial cells. An early increased secretion of prostaglandin F2α and chemokines (RANTES>>MIP-1α>>MIP-1β) was observed, due to both microglia and astrocytes. In contrast to persistent PGF2α production, an extinction of RANTES and MIP-1β but not of MIP-1α secretion occurred during the 14 days of stimulation and was inversely correlated with the ability of glial cells to replicate HIV-1. The study of the secretory factors produced in response to a persistent inflammation could provide a better understanding of the modulation of HIV replication in glial cells. GLIA 23:304–315, 1998. © 1998 Wiley-Liss, Inc.     

Japanese Encephalitis Virus infection induces IL-18 and IL-1beta in microglia and astrocytes: correlation with in vitro cytokine responsiveness of glial cells and subsequent neuronal death

IL-1β and IL-18 are members of the IL-1 family of ligands, and their receptors are members of the IL-1 receptor family. Although several biological properties overlap for these cytokines, differences exist. In order to assess functional importance of these two cytokines in viral encephalitis, we have exploited an experimental model of Japanese Encephalitis (JE) and subsequent in vitro cell culture system. We report for the first time that in Japanese Encephalitis, microglia and astrocytes both produce IL-18 and IL-1β. In vitro, these two cytokines differentially activate microglia and astrocyte, and also alter the by stander neuronal survival following treatment with these two cytokines.     

Differential suppression of interferon-γ-induced Ia antigen expression on cultured rat astroglia and microglia by second messengers

The roles of intracellular second messengers in interferon-γ (IFN-γ)-induced Ia antigen (Ag) expression by astroglia and microglia were examined. Ia Ag on both glia types was induced by IFN-γ. Reagents known to increase intracellular cAMP or activate intracellular protein kinase C (PKC) reduced IFN-γ-induced Ia Ag expression by astroglia. In contrast, increasing intracellular cAMP had no suppressive effect on Ia Ag expression by microglia. These results indicate (1) cAMP and PKC negatively regulate IFN-γ-induced Ia expression on astroglia, and (2) Ia expression is regulated differentially in astroglia vs. microglia. These findings may explain the frequent observation of Ia⁺ microglia (or macrophages) but not astroglia in various neurodegenerative diseases.     

Aging and glial responses to lipopolysaccharide in vitro: greater induction of IL-1 and IL-6, but smaller induction of neurotoxicity

Glial activation during aging was analyzed in primary glia cultured from brain regions sampled across the life span. An initial study showed that phenotypes of activated astrocytes and microglia from aging rat cerebral cortex persisted in primary cultures (Neurobiol. Aging 19 (1998), 97). We extend these findings by examining effects of age on the activation of glial cultures from adult rat brain in response to lipopolysaccharide (LPS), an inflammatory stimulus. Mixed glia from F344 male rats, aged 3 and 24 months, cultured from cerebral cortex (Cx), hippocampus (Hc), and striatum (St), were assayed for cytokines implicated in Alzheimer's disease: IL-1alpha, IL-1beta, IL-6, and TNF-alpha. Regional differences across all age groups included consistently lower expression of these cytokines in glia derived from Cx than Hc and St. Aging increased basal IL-6 mRNA and secretion by >or=3-fold in glia from Cx and Hc. Aging also increased LPS-induced IL-1 and IL-6 in Hc more than in Cx, whereas no significant effects of age were seen in St-derived glial cytokines. TNF-alpha secretion did not differ by donor age (basal or LPS-induced). Nitric oxide production by microglia from Cx of aging brains showed a smaller induction in response to LPS, with proportionately less neurotoxicity. Thus, glial activation during aging shows regional selectivity in cytokine expression, with opposite effects of aging on the increased inducibility of IL-1 and IL-6 vs the decreased production of nitric oxide.