Tumor necrosis factor identified in multiple sclerosis brain.

Frozen brain specimens from patients with multiple sclerosis (MS) and other neurologic diseases were analyzed using immunocytochemical techniques for the presence of TNF. In brain lesions in MS, and subacute sclerosing panencephalitis, TNF+ cells were demonstrated. At the lesion site in MS, TNF+ staining is associated with both astrocytes and macrophages. These observations were not made in Alzheimer's disease or normal brain tissue. The presence of TNF in MS lesions suggests a significant role for cytokines and the immune response in disease progression.

Human immunodeficiency virus 1 envelope proteins induce interleukin 1, tumor necrosis factor alpha, and nitric oxide in glial cultures derived from fetal, neonatal, and adult human brain.

Although microglia are the only cells found to be productively infected in the central nervous system of acquired immunodeficiency disease syndrome (AIDS) patients, there is extensive white and gray matter disease nonetheless. This neuropathogenesis is believed to be due to indirect mechanisms other than infection with human immunodeficiency virus 1 (HIV-1). Cytokines and toxic small molecules have been implicated in the clinical and histopathological findings in CNS AIDS. Previously, we have demonstrated in rodent glial cultures the presence of biologically active epitopes of gp120 and gp41 that are capable of inducing interleukin 1 and tumor necrosis factor alpha. In this study, we map the HIV-1 envelope epitopes that induce nitric oxide, inducible nitric oxide synthase, interleukin 1, and tumor necrosis factor alpha in human glial cultures. Epitopes in the carboxy terminus of gp120 and the amino terminus of gp41 induce these proinflammatory entities. In addition, we compare HIV-1 infection and pathology in glial cells derived from human brain taken at different states of maturation (fetal, neonatal, and adult brain) in an effort to address some of the clinical and histological differences seen in vivo. This study demonstrates that, in the absence of virus infection and even in the absence of distinct viral tropism, human glia respond like rodent glia to non-CD4-binding epitopes of gp120/gp41 with cytokine and nitric oxide production. Differences among fetal, neonatal, and adult glial cells' infectivity and cytokine production indicate that, in addition to functional differences of glia at different stages of development, cofactors in vitro and in vivo may also be critical in facilitating the biological responses of these cells to HIV-1.

Proliferation of astroglia and oligodendroglia in response to human T cell-derived factors.

Human T lymphocytes transformed by human T cell leukemia-lymphoma viruses or activated by lectins were found to produce stimulating factors that promoted both proliferation and maturation of oligodendroglial and astroglial cells in vitro.

Dual infection of the central nervous system by AIDS viruses with distinct cellular tropisms.

Human immunodeficiency virus (HIV) is the causative agent of the acquired immune deficiency syndrome (AIDS). A large number of AIDS patients show evidence of neurologic involvement, known as AIDS-related subacute encephalopathy, which has been correlated with the presence of HIV in the brain. In this study, two genetically distinct but related viruses were isolated from one patient from two different sources in the central nervous system: brain tissue and cerebrospinal fluid. Both viruses were found to replicate in peripheral blood lymphocytes, but only virus from brain tissue will efficiently infect macrophage/monocytes. The viruses also differ in their ability to infect a brain glioma explant culture. This infection of the brain-derived cells in vitro is generally nonproductive, and appears to be some form of persistent or latent infection. These results indicate that genetic variation of HIV in vivo may result in altered cell tropisms and possibly implicate strains of HIV with glial cell tropism in the pathogenesis of some neurologic disorders of AIDS.

Intercellular signaling in glial cells: calcium waves and oscillations in response to mechanical stimulation and glutamate.

Intercellular Ca2+ signaling in primary cultures of glial cells was investigated with digital fluorescence video imaging. Mechanical stimulation of a single cell induced a wave of increased [Ca2+]i that was communicated to surrounding cells. This was followed by asynchronous Ca2+ oscillations in some cells. Similar communicated Ca2+ responses occurred in the absence of extracellular Ca2+, despite an initial decrease in [Ca2+]i in the stimulated cell. Mechanical stimulation in the presence of glutamate induced a typical communicated Ca2+ wave through cells undergoing asynchronous Ca2+ oscillations in response to glutamate. The coexistence of communicated Ca2+ waves and asynchronous Ca2+ oscillations suggests distinct mechanisms for intra- and intercellular Ca2+ signaling. This intercellular signaling may coordinate cooperative glial function.

Cytokines in inflammatory brain lesions: helpful and harmful.

Multiple sclerosis (MS) is thought to be an autoimmune disease. In healthy individuals, the T cells of the immune system, when activated by an infectious agent, regularly traffic across an intact blood-brain barrier, survey the CNS and then leave. In MS, for reasons that are only gradually being understood, certain events in the peripheral immune response and in the brain cause some autoreactive T cells to stay in the CNS. Their presence initiates infiltration by other leukocytes and activation and recruitment of endogenous glia to the inflammatory process, ultimately leading to the destruction of myelin and the myelin-producing cell, the oligodendrocyte, and the dysfunction of axons. The key mediators in the subsequent cycles of histological damage and repair, and clinical relapse and remission are thought to be adhesion molecules, chemokines and cytokines.

Expression of oligodendrocyte-associated genes in cell lines derived from human gliomas and neuroblastomas.

Two putative human oligodendroglioma cell lines were examined for the expression of the oligodendrocyte-associated genes, 2',3'-cyclic nucleotide-3'-phosphodiesterase, myelin basic protein, myelin proteolipid proteins, and myelin-associated glycoprotein. The expression of these genes also was examined in control astrocytoma and neuroblastoma cell lines. In addition, the expression of the non-oligodendrocyte-specific genes, glial fibrillary acidic protein (GFAP), neuron-specific enolase and neurofilaments (NF) NF-L and NF-M also were examined. All the cell lines expressed 2',3'-cyclic nucleotide 3'-phosphodiesterase, neuron-specific enolase, and vimentin, and none expressed myelin-associated glycoprotein. The "oligodendrocyte-specific" myelin proteolipid protein mRNAs and the "neuron-specific" NF-L mRNA were expressed in the two astrocytoma cell lines, which also expressed GFAP. Expression of intermediate filament protein genes was more restricted. The astrocytoma, neuroblastoma, and oligodendroglioma cell lines expressed only GFAP, NF-M, and cytokeratin K7, respectively. These results: (a) provide molecular data confirming the classification of the two cell lines as oligodendrogliomal and suggest that their molecular profiles are indicative of immature oligodendrocytes; (b) demonstrate the expression of cytokeratins in oligodendrogliomal cell lines and suggest that apparent GFAP expression in oligodendrogliomas detected by immunocytochemical methods may be due to cross-reactivity with cytokeratins, with which they share common polypeptide sequence; and (c) indicate that astrocytoma cell lines can exhibit a "mixed" phenotype, expressing genes associated with fully differentiated oligodendrocytes and neurons.

The role of nitric oxide in multiple sclerosis.

During the past decade nitric oxide has emerged as an important mediator of physiological and pathophysiological processes. Elevated nitric oxide bio-synthesis has been associated with nonspecific immune-mediated cellular cytotoxicity and the pathogenesis of chronic, inflammatory autoimmune diseases including rheumatoid arthritis, insulin-dependent diabetes, inflammatory bowel disease, and multiple sclerosis. Recent evidence suggests, however, that nitric oxide is also immunoregulatory and suppresses the function of activated proinflammatory macrophages and T lymphocytes involved in these diseases. This article reviews the role of nitric oxide in the biology of central nervous system glial cells (astrocytes and microglia) as it pertains to the pathogenesis of multiple sclerosis in humans and experimental allergic encephalitis, the animal model of this disease. Although nitric oxide has been clearly implicated as a potential mediator of microglia-dependent primary demyelination, a hallmark of multiple sclerosis, studies with nitric oxide synthase inhibitors in the encephalitis model have been equivocal. These data are critically reviewed in the context of what is know from clinical research on the nitric oxide pathway in multiple sclerosis. Specific recommendations for future preclinical animal model research and clinical research on the nitric oxide pathway in patients are suggested. These studies are necessary to further define the role of nitric oxide in the pathology of multiple sclerosis and to fully explore the potential for nitric oxide synthase inhibitors as novel therapeutics for this disease.

Proinflammatory and antiinflammatory cytokines in multiple sclerosis and central nervous system acquired immunodeficiency syndrome.

While certain cytokines such as interleukin-1 (IL-1), tumor necrosis factor-alpha (TNF-alpha), IL-6, and transforming growth factor-beta (TGF-beta) may be involved in pro- and anti-inflammatory events in the central nervous system (CNS) of patients with multiple sclerosis (MS) and CNS acquired immunodeficiency syndrome (AIDS), there is not uniform consensus as to whether they are elevated or even detectable in all compartments of the body such as serum, cerebrospinal fluid (CSF), and tissue. Furthermore, if they are elevated in these diseases, there are no data as to whether they regulate the disease process itself. Myelin damage in MS is a punctate demyelination; in AIDS, it is a diffuse myelin pallor or dysmyelination. Oligodendrocytes are destroyed in MS but not CNS AIDS, suggesting a different mechanism for myelin loss in the two diseases. These different pathologies may provide clues about the role of macrophages, microglia, and/or the toxic products they produce in putatively giving rise to myelin damage. The stimuli that trigger such a destructive response by macrophages or glial cells and/or the regulation of the toxic events in the two diseases we would predict to be different. In MS, an effector cell-mediated lesion production and oligodendrocyte cell destruction seem to occur. We hypothesize that the effector is the inflammatory blood macrophage and/or microglial cell induced and promoted to its cytotoxic activity by a collaboration of neurotransmitters and cytokines. In CNS AIDS, virus-induced toxic products of glia and their diffusion through white and gray matter areas of the brain have been suggested. Such soluble mediators would then compromise metabolic processes of neurons and glia without widespread target cell loss.

Remyelination by oligodendrocytes stimulated by antiserum to spinal cord.

The new synthesis of myelin and the proliferation of oligodendrocytes was stimulated by serum from syngeneic mice immunized with homogenized spinal cord (SCH). Treatment with this antiserum produced a 10-fold increase in the area of remyelination in spinal cords that had become demyelinated previously as a result of infection by Theiler's murine encephalomyelitis virus. Inflammation was decreased in regions of white matter that showed remyelination. Oligodendrocytes exposed to anti-SCH in vitro incorporated three to five times more [3H]thymidine than resting cells did and expressed more myelin basic protein in their cytoplasm, suggesting stimulation of myelinogenesis. Thus, there is a factor present in anti-SCH antiserum that stimulates central nervous system-type remyelination. This finding may provide clues for the therapy of patients with demyelinating disorders such as multiple sclerosis.

Failure to detect human T-cell leukemia virus-related sequences in multiple sclerosis blood.

We tested 11 patients with multiple sclerosis for the presence of human T-cell leukemia virus type I (HTLV-I)- or type II (HTLV-II)-related sequences. DNA from blood mononuclear cells was analyzed by the polymerase chain reaction utilizing three different oligonucleotide primer pairs. Two of these primer pairs detect sequences shared between HTLV-I and HTLV-II in either p24, gag protein, or in p21, env transmembrane protein. The third primer pair was synthesized based on regions in the pol gene where amino acid sequences are conserved between HTLV-I, HTLV-II, and the related bovine leukemia virus. The multiple sclerosis samples were consistently negative while appropriate control samples were positive. We conclude that viruses related to HTLV-I, HTLV-II, or bovine leukemia virus are not present in the blood of patients with multiple sclerosis and, therefore, that HTLV-bovine leukemia virus-related viruses are not likely to be involved in the pathogenesis of multiple sclerosis.

Pentoxifylline is not a promising treatment for multiple sclerosis in progression phase.

Fourteen MS patients took pentoxifylline at varying doses for up to 24 months. In vitro production of tumor necrosis factor alpha was reduced in patients taking 2,400 to 3,200 mg/day of pentoxifylline for 12 weeks or more. Twelve of the 14 patients experienced worsening of the disease during the study according to clinical, MRI, or visual evoked potential criteria. These results provide no hint of efficacy for pentoxifylline as a treatment for MS in progression phase.

Regulation of natural killer cell cytotoxicity by prostaglandin E in the peripheral blood and cerebrospinal fluid of patients with multiple sclerosis and other neurological diseases. Part 2. Effect of exogenous PGE1 on spontaneous and interferon-induced natural killer.

Compared to normal and other neurological disease (OND) controls, multiple sclerosis (MS) pre nylon wool (pre NW) and nylon wool passed (NWP)-peripheral blood cells' natural killer (NK) activity was more sensitive to prostaglandin E (PGE1); it was suppressed to a greater degree and at lower concentrations of PGE1. At the single cell level this was reflected by lower numbers of target-binding cells (TBCs) and fewer killers among the TBCs. ONDs and normal controls were equally sensitive to PGE1. Though PGE-producing cells were depleted in the NWP population of normal and control ONDs, MS patients still had indomethacin-sensitive NK suppressors in the NWP population; these apparently did not suppress at the single cell effector level but at the level of recycling. MS and OND cerebrospinal fluid (CSF) cells' NK activity could not be 'enhanced' by indomethacin. Depression of interferon (IFN)-induced NK by PGE1 was greater in MS than in OND or normal controls perhaps through its effect on IFN-induced recycling. All subjects' cells maintained sensitivity to PGE1 after overnight incubation in the presence of PGE-producing cells (pre NW) or exogenous PGE1. In sharp contrast to normal and OND controls, MS NWP cells were still inhibited by PGE1 even after overnight incubation in the absence of PGE1.

Cytotoxic activity of peripheral blood and cerebrospinal fluid lymphocytes from patients with multiple sclerosis and other neurological diseases. Analysis at the single cell level using morphological and surface marker phenotype criteria.

The large granular lymphocyte (LGL) has been identified in normal individuals' MNC population as the NK-K cell (Ault and Weiner 1978; Timonen et al. 1981); it bears the OKM1 surface antigen (Breard et al. 1981; Fast et al. 1981) and while negative for T cell antigens OKT4 and OKT8, is a low avidity E-rosette forming cell. However, in a unfractionated nylon wool passed peripheral blood lymphocyte (NWP PBL) population, we show that not more than 50% of KN activity in normal or OND control NWP PBL and 30% NK activity in MS NWP PBL can be attributed to this cell. Nevertheless, 100% of control K cell activity and 50% of MS K cell activity can be mediated by an LGL. MS patients have normal proportions of LGLs in their NWP PBL. The proportion of LGLs in CSF of MS and OND patients is too low to account for the number of CSF K cells. While in control NWP PBLs, the LGLs are OKM1+ and mediate NK and ADCC, in MS the LGL NK effectors are probably different from LGL-K cell effectors. In MS both populations include effector cells with cell surface antigens. Thus, the OKM1+ LGL characteristics may not be used in analysis of NK and K cells in multiple sclerosis.

Regulation of natural killer cell cytotoxicity by prostaglandin E in the peripheral blood and cerebrospinal fluid of patients with multiple sclerosis and other neurological diseases. Part 1. Association between amount of prostaglandin produced, natural killer, and endogenous interferon.

MS patients' peripheral blood mononuclear cells (MNC) spontaneously produced more prostaglandin E1&2 (PGE) in vitro than did OND or normal control MNC correlating with lower levels of NK activity and endogenous interferon (IFN) production while there were no differences between OND and normal controls. MS neat CSF contained significantly higher levels of measurable PGE with concomitantly lower NK activity and % IFN-positive cells than was seen in OND CSF. The PGE producing cells were depleted in the nylon wool-passed (NWP) and enriched in the nylon wool-adherent (NWAd) populations in direct correlation with depletion and enrichment of esterase-positive monocyte/macrophages, respectively, suggesting that at least some of the PGE-producing cells were adherent monocyte/macrophages. At 24 and 48 h after stimulation with PHA, MS unfractionated MNC produced significantly more PGE than OND and normal MNC; at 24 and 72 h, MS NWAd cells produced significantly more PGE than normals and ONDs. MS and ONDs produced the same amount of PGE at 48 h in the stimulated NWAd fraction.

The mapping of HIV-1 gp160 epitopes required for interleukin-1 and tumor necrosis factor alpha production in glial cells.

The pathology in central nervous system (CNS) AIDS suggests that direct infection with HIV-1 is not required for changes in glia and neurons. Induction of a variety of pathological responses in vitro in rodent brain cultures also suggests that CD4 is not the receptor for HIV-1 in the brain, given that human and rodent CD4 are not homologous. This implies that the epitopes on HIV-1 which bind glia and activate them are novel, non-CD4-binding domains. We have therefore mapped the envelope (env) regions required for production in rat glial cultures of interleukin-1 (IL-1) and tumor necrosis factor alpha (TNF alpha) which we hypothesize are important in CNS AIDS. Serially truncated deletion mutants from the gp120/gp41 carboxy terminus representing folded, glycosylated recombinant env proteins were expressed in HeLa cells via a vaccinia virus vector. These proteins, linear gp120/gp41 peptides, as well as polyclonal and monoclonal antibodies reactive to defined regions of gp120/gp41 were used to map the epitopes involved in production of IL-1 and TNF alpha. Compared to HeLa cell and wild-type vaccinia virus controls, the vaccinia recombinant env protein gp160 containing cleaved gp120 and gp41 induced both IL-1 and TNF alpha. If gp160 was not cleaved into gp120 and gp41, IL-1 but not TNF alpha induction was reduced. Peak production of TNF alpha by gp120/gp41 was at 4 h while IL-1 production was still significantly elevated at 44 h at the highest concentrations of env protein. Using the truncation deletions, the V3 loop of gp120 appeared to be critical for IL-1. Glycosylation and folding of V3 is probably important in IL-1 induction since a V3 peptide was not as active. While removal of glycosylated, folded V4 and C4 regions had no effect on IL-1, linear peptides in the region from the V4 loop to the C4 domain were strong inducers of IL-1. Non-glycosylated, linear V4 loop peptide induced more IL-1 than the V4 in protein generated in HeLa cells, suggesting that glycosylation and/or conformational structures sequester V4 inducer epitopes. Using the truncation deletions, the carboxy terminus region (V4-C5) of gp120 as well as gp41 were shown to be critical for TNF alpha production. Peptides representing linear epitopes in the V3 loop, C5 domain of gp120, and the ectodomain of gp41 were all strong inducers of TNF alpha; a protein representing almost the entire gp41 was the strongest inducer of TNF alpha.(ABSTRACT TRUNCATED AT 400 WORDS)

Differential expression, cytokine modulation, and specific functions of type-1 and type-2 tumor necrosis factor receptors in rat glia.

Tumor necrosis factor alpha (TNF alpha) and lymphotoxin alpha (LT alpha) induce pleiotropic cellular effects through low-affinity 55 kDa type-1 receptors (TNFR1, CD120a) and high-affinity 75 kDa type-2 receptors (TNFR2, CD120b). Both cytokines have potent biological effects on glial cells and are strongly implicated in the pathology of central nervous system (CNS) demyelinating diseases. However, to date, neither constitutive nor cytokine-induced TNFR expression by glial cells have been definitively characterized. We therefore characterized TNF receptors at the molecular, protein, and functional levels in rat astrocytes, microglia, and oligodendrocytes. Northern blotting demonstrated that all three types of glia constitutively transcribed a single TNFR1 mRNA. IFN gamma increased transcript levels in all three types of glia, but TNF alpha increased levels only in oligodendrocytes Microglia constitutively transcribed three distinct TNFR2 mRNAs, levels of which were increased by either IFN gamma or TNF alpha. In contrast, astrocytes and oligodendrocytes constitutively transcribed nearly undetectable levels of TNFR2 mRNAs, and levels were not affected by IFN gamma, TNF alpha, or oligodendrocyte maturation. Immunocytochemical staining of glial cells corroborated Northern data by demonstrating that glia express a parallel pattern of TNFR proteins on their cell surfaces. In co-cultures of microglia plated atop irradiated astrocytes, human TNF alpha (which, on mouse cells, binds TNFR1 exclusively) induced microglial cell proliferation, whereas murine TNF alpha (which binds both TNFRs) did not. Collectively, the data show that microglia, a primary source of TNF alpha at CNS inflammatory sites, express both TNFR1 and TNFR2, whereas astrocytes and oligodendrocytes, whose embryological origin differs from that of microglia, predominantly express TNFR1. TNF alpha increases expression of TNFR1 by oligodendrocytes whereas it increases expression of TNFR2 by microglia. Microglia proliferation data suggest that signals transduced through TNFR2 directly or indirectly inhibit signals transduced through TNFR1. Different patterns of TNFR expression by glia at sites of CNS inflammation may be critical in determining whether TNF has activational, proliferative, or cytotoxic effects on these cells.

Response of human glioblastoma cells to recombinant interleukin-2.

We investigated the ability of glioma cells to respond to T cell-derived lymphokines. The growth of astrocytoma and mixed glioblastoma cell lines, as assessed by DNA synthesis, was inhibited in the presence of supernatants derived from mitogen-stimulated human T cells, an HTLV-II-transformed human T cell line, Mo, and human interleukin-2 (IL-2). The mixed glioblastoma cell line, 138-MG-C, was subjected to limiting dilution analysis, and two cell lines (5D7, 5C5) were derived which were homogeneous with respect to staining for galactocerebroside (GalC) (100%). These two GalC+ glioblastoma cell lines proliferated in the presence of high concentrations of recombinant human interleukin-2 (RIL-2). Additionally, these cell lines bear receptors for the IL-2 molecule as determined by immunofluorescent staining with various anti-IL-2 receptor antibodies.

Production of interleukin-1 by microglia in response to substance P: role for a non-classical NK-1 receptor.

Substance P (SP) is a central and peripheral neurotransmitter which has been found in multiple sclerosis plaques. SP stimulates peripheral immune cells and may play a role in some chronic inflammatory diseases. Human peripheral monocyte/macrophages have been shown to produce the inflammatory cytokines interleukin-1 (IL-1) and tumor necrosis factor alpha (TNF alpha) in response to SP. Therefore, in this study we examined rat brain microglia for the presence of SP receptors and production of IL-1 and TNF alpha in response to SP. Microglia had 4900 +/- 950 (mean +/- SE) receptors per cell fitting a two-site model. Four percent of these were high-affinity receptors with a Kd of 8.2 x 10(-8) M +/- 3.6 x 10(-8) M (mean +/- SE), and 96% of them were low-affinity receptors with a Kd of 2.1 x 10(-6) M +/- 5.2 x 10(-7) M (mean +/- SE). Competitive studies with CP 96,345 and other SP analogs demonstrate these to be non-classical NK-1 receptors. SP alone did not stimulate IL-1 or TNF alpha production. However, SP in synergy with lipopolysaccharide (LPS) quadrupled IL-1 production compared to LPS alone, but did not affect TNF alpha production. These results have implications for certain inflammatory conditions in the central nervous system.

T cell lines established from multiple sclerosis cerebrospinal fluid T cells using human retroviruses.

Cerebrospinal fluid (CSF) lymphocytes from patients with multiple sclerosis (MS) were transformed with human T cell leukemia/lymphoma virus (HTLV I and HTLV II) and the resulting cell lines characterized by cell surface phenotyping and functional assessment. The lines were predominantly of the CD4 helper/induce phenotype although the HTLV II lines contained 10-20% CD8+ cells. The lines appeared to be activated cells; the majority were TA1+, HLA-DR+, and TAC+ (CD25+). Interestingly, they were OKT10- (CD38-). Functionally, the lines contained no natural killer (NK) activity and were modestly cytotoxic in the antibody-dependent cellular cytotoxicity (ADCC) assay. They were poor proliferative responders to antigens and mitogens though the HTLV II lines did respond to interleukin 2 (IL2). The HTLV I lines were either nonresponsive to or were suppressed by IL2. Early passages of two of the lines produced IL2 but this was lost as the cells were passed in culture. The cell lines were capable of either directly or indirectly suppressing pokeweed mitogen (PWM)-driven immunoglobulin production by normal B cells. In addition, the lines were capable of producing gamma-interferon (IFN-gamma), lymphotoxin (LT), an interleukin 1 (IL1)-like factor, glial growth promoting factor (GGPF), and IL6. The advantage of these lines over clones or cell lines developed using other techniques is their growth in the absence of feeder layers or IL2 and their ability to be cloned and to grow in culture indefinitely.