ABSTRACT Cytokine expression in the brain has been suggested to mediate various sickness behaviors. Here we report that intraperitoneal injection of a Corynebacterium parvum antigen in C57BL/6 mice was followed by prolonged upregulation of cytokines in the cerebral cortex and subcortical structures in a time course that coincided with reduced spontaneous running activity.

Infection by human granulocytic ehrlichiosis (HGE) is characterized clinically by cytopenias out of proportion to the number of cells seen to be infected directly. To study the pathogenic role of inflammatory mediators in HGE infection, cytokine production by untreated and dimethyl sulfoxide-treated HL-60 cells, which demonstrate enhanced infection because of granulocytic differentiation, and by normal bone marrow cells was measured using modified sandwich ELISA assays on samples obtained sequentially after inoculation with the HGE agent. All infected cells produced physiological concentrations of CC (monocyte chemotactic protein-1, macrophage inflammatory protein-1alpha and -beta, and RANTES) and CXC (interleukin [IL]-8) chemokines in amounts significantly greater than those produced by uninfected controls. In contrast, infected cells did not secrete the classic proinflammatory cytokines IL-1, IL-6, or tumor necrosis factor-alpha. The striking production of chemokines, powerful leukocyte chemoattractants capable of suppressing hematopoiesis, by susceptible target cells, is likely to be of pathogenic importance both in the observed cytopenias and in mediation of inflammation and host defenses during infection.

The B-chemokine RANTES has recently been implicated in the neuropathogenesis of the human immunodefiency virus. Based upon previous studies of the effects of morphine on microglial cell production of cytokines and chemotaxis towards the activated complement component C5a, we tested the hypothesis that this opiate would alter the production of and migration towards RANTES by human microglia. Treatment of highly purified microglial cell cultures with morphine (10–8–10–6M) potently inhibited RANTES production by lipopolysaccharide-and interleukin-1[.beta]-stimulated cells. Using a chemotaxis chamber to assess directed migration towards RANTES, treatment of microglial cells with morphine (10–10–10–6M) was found to suppress chemotaxis. The inhibitory effects of morphine on RANTES production and on chemotaxis were blocked by naloxone and [.beta]-funaltrexamine, indicating that morphine mediated its suppressive effects via activation of microglial µ-opioid receptors. Morphine's inhibitory effect on chemotaxis did not appear to be associated with an alteration in RANTES-induced [Ca2+]i mobilization. While the clinical significance of these in-vitro findings is unknown, they suggest that µ-opioid receptor agonists could alter certain neurodegenerative and inflammatory processes within the brain.

Polymorphonuclear neutrophils (PMNs) are known to mediate brain inflammation following hypoxia/reoxygenation (H/R), but the precise mechanisms leading to PMN recruitment are undefined. The α-chemokine macrophage inflammatory protein-2 (MIP-2) has specificity for the recruitment of PMNs. In this study, we found that 8 or 12 h of hypoxia followed by 24-h reoxygenation (H8/R24 or H12/R24) induced MIP-2 secretion in cultures of enriched microglia or mixed glia, respectively. Microglia, however, could not survive longer duration (>12 h) of hypoxia. Astrocytes did not produce any significant amount of MIP-2 even though astrocytes maintained 98–99% viability following H12/R24. We also found that microglia survived the H/R treatment better (following H24) in the presence of astrocytes (mixed glial culture) than in microglia-enriched culture. Reoxygenation for prolonged periods (3 and 5 days) following H24 resulted in progressively larger increases in MIP-2 production (20- and 60-fold, respectively) in mixed glial cultures. Immunocytochemical staining revealed that the cells expressing MIP-2 in response to H/R were microglia rather than astrocytes in mixed glial cultures. Examination of MIP-2 mRNA expression showed that H/R upregulated MIP-2 gene expression. Taken together, our data suggest that microglial cells are an important source of MIP-2 production and suggest a potential injury mechanism involving brain-derived production of MIP-2 in H/R. GLIA 32:155–164, 2000. © 2000 Wiley-Liss, Inc.

Mechanisms underlying human immunodeficiency virus-1 encephalopathy are not completely known; however, recent studies suggest that the viral protein gp41 may be neurotoxic via activation of inducible nitric oxide synthase (iNOS) in glial cells. In the present study, we investigated the NO-generating activity of primary human fetal astrocytes in response to gp41 and the relationship to microglial cell production of interleukin-1 (IL-1). Gp41 failed to trigger iNOS mRNA expression in highly enriched (>99%) astrocyte or microglial cell cultures. However, gp41-treated microglia released a factor(s) that triggered iNOS mRNA expression and NO production in astrocytes. Because IL-1 receptor antagonist protein blocked gp41-induced NO production, a pivotal role was suggested for microglial cell IL-1 production in astrocyte iNOS expression. Also, gp41 induced IL-1beta mRNA expression and IL-1 production in microglial cell but not astrocyte cultures. Using specific inhibitors, we found that gp41-induced IL-1beta production in microglia was mediated via a signaling pathway involving protein-tyrosine kinase. These data support the hypothesis that gp41 induces astrocyte NO production indirectly by triggering upregulation of microglial cell IL-1 expression.

OBJECTIVES:To characterize replication patterns and cytopathic effects during human cytomegalovirus (HCMV) infection of brain cells.DESIGN:Primary human mixed glial/neuronal cells, as well as purified microglial, astroglial, and enriched neuronal cell cultures, were infected with HCMV strains AD169 and RC256 to determine the ability of the different brain cell types to support viral replication.RESULTS:Mixed glial/neuronal cell cultures were fully permissive for viral replication. Based on previous studies, we hypothesized that human microglial cells would preferentially support productive HCMV replication. However, HCMV did not replicate or display genomic expression in microglial cells. In contrast, primary astrocytes were fully permissive and displayed HCMV-induced cytopathic effects resulting in cell death. In highly enriched neuronal cultures, productive infection and viral expression occurred only in scattered astrocytes. Early in the infection, apoptotic plasma membrane changes were induced in astrocytes. However, nuclear fragmentation was not apparent until later during the course of infection.CONCLUSIONS:These results suggest that HCMV possesses astrocytotropic properties that confer preferential expression and cytopathic replication in astrocytes over microglia or neuronal cells. Apoptotic cell death, which is a result of HCMV infection, appears to be delayed until peak viral replication has occurred.

ACTIVATION of the immune system has been associated with the development of fatigue of unknown cause. We were interested in brain energy stores (e.g., phosphocreatine (PCr) and creatine kinase) after immune activation to investigate whether this system was altered. In this model, fatigue (defined as > 50% reduction in spontaneous running) was induced in C57BL/6 mice after a single injection of Corynebacterium parvum antigen. Maximal fatigue (about 86% reduction on day 10 post injection) was associated with reduced (about 29%) brain PCr/gamma-ATP and increased creatine kinase levels (similar to 31%), suggesting an active process of brain ATP depletion and replenishment. These findings need to be further delineated to establish the relationship between immune activation, reduced brain energy pools and fatigue. (C) 1999 Lippincott Williams & Wilkins.

Endogneous δ and κ opioid peptides possess a variety of immunomodulatory properties, and κ-opioid receptor ligands recently were shown to suppress the expression of human immunodeficiency virus type 1 (HIV-1) in microglial cells, the resident macrophages of the brain. To determine whether the newly discovered endogenous μ-opioid receptor ligands endomorphin-1 and -2 would affect HIV-1 replication, these peptides were added to acutely infected brain cell cultures. Endomorphin-1 potentiated viral expression, in a bell-shaped dose-response manner with maximal enhancement≅35% at 10−10 M, in both mixed glial/neuronal cell and purified microglial cell cultures. Endomorphin-1’s amplifying effect was blocked by pretreatment of brain cells with either the μ-opioid receptor selective antagonist β-funaltrexamine or the G protein inhibitor pertussis toxin. However, the classical μ receptor agonists morphine and DAMGO (Tyr-d-Ala-Gly-N-Me-Phe-Gly-ol) had no effect on viral expression or on endomorphin-1’s amplifying effect. Taken together, these findings suggest that in this in vitro model of HIV-1 brain infection, endomorphin-1 potentiates viral expression via activation of an atypical μ-selective opioid receptor. They also provide evidence, for the first time, that an endogenous μ-opioid peptide has neuroimmunomodulatory activity.

Using human fetal microglial cell cultures, we found that the gram-negative bacterial cell wall component lipopolysaccharide (LPS) stimulated RANTES (regulated upon activation of normal T cell expressed and secreted) production through the protein kinase C signaling pathway and that activation of transcription nuclear factor (NF)-kappaB was required for this effect. Similarly, the proinflammatory cytokines interleukin (IL)-1beta and tumor necrosis factor-alpha dose-dependently stimulated microglial cell RANTES production via NF-kappaB activation. Anti-inflammatory cytokines, IL-10, and transforming growth factor (TGF)-beta sequentially inhibited LPS- and cytokine-induced microglial cell NF-kappaB activation, RANTES mRNA expression, and protein release. Proinflammatory cytokines but not LPS also stimulated RANTES production by human astrocytes. These findings demonstrate that human microglia synthesize RANTES in response to proinflammatory stimuli, and that the anti-inflammatory cytokines IL-10 and TGF-beta down-regulate the production of this beta-chemokine. These results may have important therapeutic implications for inflammatory diseases of the brain.

Although the precise mechanisms whereby HIV-1 infection induces neurodegeneration have yet to be determined, a great deal of evidence has incriminated glial cells and the production of proinflammatory mediators in this pathologic process. For this reason, ideal therapeutic agents for the treatment of AIDS dementia would attenuate HIV-1 neuropathogenesis through both direct inhibition of viral expression and suppression of brain cell-produced immune mediators. Benzodiazepines (BDZs), such as Valium, are extensively prescribed drugs for anxiety disorders, which readily cross the blood-brain barrier and have demonstrated immunomodulatory properties. BDZs bind to primary human microglial cells, the principal site of HIV-1 replication in the brain, and inhibit lipopolysaccharide (LPS) induced tumour necrosis factor (TNF-α) production by these cells in a concentration-dependent manner. Treatment of HIV-1-infected primary human microglial, as well as mixed glial/neuronal, cell cultures with BDZs inhibits the expression of HIV-1 p24 antigen. BDZ-induced inhibition of HIV-1 expression in chronically infected promonocytic (U1) cells has been found to be associated with decreased activation of the nuclear transcription factor kappa B (NF-κB). Because HIV-1 expression is critically dependent on the cellular transcription machinery, inhibition of the activation of transcription factors, which participate in both HIV-1 expression and the production of neurotoxic immune mediators, by BDZ analog may provide new therapeutic options for AIDS dementia.

Opioids have been postulated to play an immunomodulatory role in the pathogenesis of HIV-1. Synthetic κ-opioid receptor (KOR) ligands have been found to inhibit HIV-1 expression in acutely infected microglial cell cultures. We recently found that interleukin(IL)-1β and tumor necrosis factor(TNF)-α have antiviral effects in acutely infected mixed glial/neuronal cell cultures. In the present study, we investigated whether selective KOR ligands would exert antiviral effects in acutely infected brain cell cultures. While the KOR ligand trans-3,4-dichloro-N-methyl-N[2-(1-pyrolidinyl)cyclohexyl]benzeneaceamide methanesulfonate (U50,488) alone had little anti-HIV-1 activity, this opioid potentiated in a concentration-dependent manner the antiviral activity of TNF-α, but not of IL-1β. The potentiating effect of U50,488 was detected after a 6-hr pretreatment and peaked at 24 hr. The KOR antagonist nor-binaltorphimine completely blocked the potentiating effect of U50,488, suggesting the involvement of a KOR-mediated mechanism. Antibodies to TNF-α completely blocked the potentiating effect of U50,488, suggesting a critical role for TNF-α. Antibodies to IL-1β blocked the potentiating effect of U50,488, suggesting that IL-1β was released following U50,488 treatment, which might contribute to the potentiating effect of U50,488. These in vitro findings support the notion that synthetic κ-opioids could be considered as potential adjunctive therapeutic agents in HIV-1-related brain disease.

Delta-opioid receptor (DOR) transcripts and binding sites are expressed by lymphocytes and lymphoid cell lines from several species. Direct modulation of lymphocyte function through DORs affects T cell proliferation, interleukin-2 production, chemotaxis, and intracellular signaling. Moreover, in human DOR-transfected T cells (DOR-Ju.1), delta-opioids have been shown previously to mobilize intracellular calcium rapidly, to inhibit forskolin-stimulated cyclic AMP production, and to activate the mitogen-activated protein kinases ERKs 1 and 2. These observations led us to consider whether delta agonists modify T cell functions, thus affecting the expression of human immunodeficiency virus-1 (HIV-1) by CD4+ T cells. To test this hypothesis, DOR-Ju.1 cells, derived from Jurkat cells stably transfected with a cDNA encoding the neuronal DOR, were stimulated with deltorphin or benzamide, 4-[[2,5-dimethyl-4-(2-propenyl)-1-piperazinyl](3-methoxyphenyl)methyl]N- ,[2S[(S*),2alpha,5beta]]-(9Cl) (SNC-80) prior to the addition of HIV-1. Both deltorphin and SNC-80 concentration-dependently inhibited the production of p24 antigen, an index of HIV-1 expression. Inhibition was maximal with 10(-13)-10(-9) M SNC-80 (>60% reduction) or 10(-15)-10(-11) M deltorphin (>50% reduction). At higher concentrations, less inhibition of p24 antigen production was found. Naltrindole (NTI, 10(-11) M), a selective DOR antagonist, abolished the inhibitory effects of 10(-9) M SNC-80, whereas 10(-13) M NTI partially reversed the effect of SNC-80. Thus, activation of DORs expressed by CD4+ T cells significantly (P < 0.05) reduced the expression of HIV-1 by these cells. These findings suggest that opioid immunomodulation directed at host T cells may be adjunctive to standard antiviral approaches to HIV-1 infection.

Opioids have been postulated to play an immunomodulatory role in the CNS. Recently, we found that priming microglia with interferon (IFN)-γ or tumor necrosis factor (TNF)-α resulted in an enhanced production of superoxide anion, a reactive oxygen intermediate that may be pathogenic during brain inflammation. In the present study, we investigated the effects of trans-3,4-dichloro-N-methyl-N[2-(1-pyrolidinyl)cyclohexyl]benzeneaceamide methanesulfonate (U50,488), a selective κ-opioid ligand, on microglial cell superoxide production when cells were primed with cytokines or stimulated with phorbol myristate acetate. While treatment of microglial cells with U50,488 had little effect on nonstimulated or stimulated superoxide production, this opioid inhibited (by >70%) the priming effects of cytokines. Maximal inhibition of microglial cell superoxide generation by U50,488 was observed at 10 nM for the priming effect of interferon-γ and at 1 μM for tumor necrosis factor-α. Pretreatment of microglial cell cultures for 30 min with an equal concentration of the selective κ-opioid receptor antagonist nor-binaltorphimine (nor-BNI) completely blocked the inhibitory effect of U50,488. The results of this study suggest that κ-opioids may have therapeutic potential in inflammatory diseases of the CNS involving reactive oxygen intermediates produced by activated microglia.

Opioids (exogenous opiates and endogenous opioid peptides) have a diversity of effects on the immune system. Although numerous studies have shown that opioid-induced immunosuppression can be mediated indirectly via the central nervous system (CNS) or through direct interactions with immunocytes, the precise cellular mechanisms underlying the immunomodulatory effects of opioids are largely unknown. In recent years, investigations from several laboratories have indicated that opioids can operate as cytokines, the principal communication signals of the immune system. All of the major properties of cytokines are shared by opioids, i.e., production by immune cells with paracrine, autocrine, and endocrine sites of action, functional redundancy, pleiotropy and effects that are both dose- and time-dependent. Studies of the effects of opioids on peripheral blood mononuclear cells (PBMC) or brain cells cocultured with HIV-infected cells suggest that some of the immunoregulatory actions of opioids are mediated by ultrahigh affinity receptors on PBMC and glial cells. Because the CNS is populated predominantly by astroglia and microglia which have properties of immune cells, it is possible that certain of the CNS effects of opioids involve cytokine-like interactions with glial cells. Although there is mounting evidence supporting the concept that opioids are members of the cytokine family, the relative contribution of the opioids to immunoregulation remains unclear. The importance of opiate addiction in the AIDS epidemic means that gaining a better understanding of the mechanisms of opioid-induced immunomodulation is of more than academic interest.

Renal infiltration of human immunodeficiency virus type 1 (HIV-1)-infected monocytes might play an important role in the development of HIV-associated nephropathy (HIVAN). In the present study, we investigated the effects of cytokines produced by cultured human mesangial cells (HMC) and proximal tubular epithelial cells (PTEC) on HIV-1 expression in chronically HIV-1-infected promonocytes (U1 cells). Human mesangial cells constitutively secreted interleukin-6 (IL-6) but not tumor necrosis factor-alpha (TNF-alpha) into the culture medium, whereas PTEC constitutively secreted both IL-6 and TNF-alpha. Coculture of U1 cells with HMC or PTEC for 72 hours markedly stimulated HIV-1 expression, with the p24 antigen concentration in the coculture supernatants ranging from approximately 200 to 1850 pg/ml. The presence of anti-IL-6 antibody in the coculture medium nearly completely blocked HIV-1 expression in the HMC/U1 cell cocultures (P < 0.05). Anti-IL-6 antibody and anti-TNF-alpha antibody blocked HIV-1 expression in the PTEC/U1 cell cocultures by 40% and 53%, respectively (P < 0.05). Moreover, the combination of anti-IL-6 and anti-TNF-alpha antibodies additively reduced coculture HIV-1 expression by 87% (P < 0.05). We conclude that renal cell production of IL-6 and TNF-alpha might provide a potent stimulus for HIV-1 expression in HIV-1-infected monocytes that infiltrate the kidney, and that this may play an important role in the pathogenesis of HIVAN.

IL-8 involvement in neutrophil activation and chemotaxis may be important in inflammatory responses within the central nervous system, secondary to meningitis, encephalitis, and traumatic injury. The source of IL-8 within the brain during these inflammatory processes, however, is unknown. To explore the role of microglia in the production of IL-8, human fetal microglia, which are the resident macrophages of the brain, were treated with LPS and pro- and anti-inflammatory cytokines to determine their effects on IL-8 production. We found that IL-8 protein levels increased in response to LPS or IL-1 beta, or to TNF-alpha, which also corresponded to elevated IL-8 mRNA levels by RT-PCR. Pretreatment with IL-4, IL-10, or TGF-beta 1 potently inhibited the stimulatory effects of these proinflammatory agents. These findings indicate that human microglia synthesize IL-8 in response to proinflammatory stimuli, and that anti-inflammatory cytokines down-regulate the production of this chemokine. These results may have important therapeutic implications for certain central nervous system insults involving inflammation.

We found that orphan opioid receptor (OR) mRNA was constitutively expressed in human fetal microglia, astrocytes, and cerebral cortical neurons. An OR antisense oligonucleotide markedly inhibited HIV-1 expression in enriched microglial cell and the mixed glial/neuronal cell cultures but not in the chronically infected promonocytic U1 cells. The sense oligonucleotide also inhibited HIV-1 expression in brain cell cultures. These findings indicate that while human brain cells constitutively express mRNA for OR, antisense oligonucleotides to OR non-specifically inhibit acute HIV-1 infection.

Nitric oxide (NO) plays an important role in host defense as well as cell injury within the CNS. In contrast to rodent species, human astrocytes are the major glial source of NO. Although interleukin (IL)-1 stimulates astrocyte inducible NO synthase (iNOS) expression, the mechanism is poorly defined. In the present study using primary human fetal astrocyte cultures, we found that IL-1 beta stimulated activation of nuclear factor kappa B (NF-kappa B) within 2 h, iNOS mRNA expression at 8 h, and maximal NO production by 5 days post-treatment. This IL-1-induced activation of astrocyte iNOS was suppressed by pyrrolidine dithiocarbamate, an inhibitor of NF-kappa B activation, suggesting involvement of a NF-kappa B mechanism.

The level of bioactive transforming growth factor-β (TGF-β) was measured in serum from patients with chronic fatigue syndrome (CFS), healthy control subjects, and patients with major depression, systemic lupus erythematosis (SLE), and multiple sclerosis (MS) of both the relapsing/remitting (R/R) and the chronic progressive (CP) types. Patients with CFS had significantly higher levels of bioactive TGF-β levels compared to the healthy control, major depression, SLE, R/R MS, and CP MS groups (P < 0.01). Additionally, no significant differences were found between the healthy control subjects and any of the disease comparison groups. The current finding that TGF-β is significantly elevated among patients with CFS supports the findings of two previous studies examining smaller numbers of CFS patients. In conclusion, TGF-β levels were significantly higher in CFS patients compared to patients with various diseases known to be associated with immunologic abnormalities and/or pathologic fatigue. These findings raise interesting questions about the possible role of TGF-β in the pathogenesis of CFS.