Astrocytes inhibit microglial surface expression of dendritic cell-related co-stimulatory molecules through a contact-mediated process.

Murine microglia cultured in isolation were treated sequentially with granulocyte/monocyte colony-stimulating factor (GM-CSF) (5 days) and lipopolysaccharide (LPS) (2 days) to elicit a mature dendritic cell-like (DC-like) phenotype. Examined by flow cytometry microglia thus isolated show high surface expression of CD11c together with the co-stimulatory molecules CD40, CD80, and CD86 that are necessary for T-cell activation. In contrast, microglia co-cultured with astrocytes fail to achieve a mature DC-like phenotype. Contact with the astrocytic environment is necessary for the inhibition. Failure was not because of a more rapid degradation of protein. Bone marrow-derived cells, like microglia, were prevented by astrocytes from attaining a mature DC phenotype. Although GM-CSF pre-treatment substantially increases mRNA of co-stimulatory molecules and major histocompatibility complex (MHC) Class II in isolated microglia, co-cultured microglia await treatment with LPS to up-regulate them. In contrast, western blot and immunocytochemical analysis revealed that it is not a failure of transcription or translation, nor is it a more rapid degradation of mRNA that is responsible for the low surface expression; rather microglia co-cultured with astrocytes produce mRNA and protein but do not traffic the protein onto the cell surface.

Antioxidants N-acetylcysteine (NAC) and 2-mercaptoethanol (2-ME) affect the survival and differentiative potential of cholinergic precursors from the embryonic septal nuclei and basal forebrain: involvement of ras signaling.

We investigated the effects of antioxidants N-acetylcysteine (NAC) and 2-mercaptoethanol (2-ME) on the expression of choline acetyltransferase (ChAT) in cultured cholinergic precursors from the embryonic rat septal nuclei and basal forebrain. Carboxy-dichlorofluorescein fluorescence confirmed that 2-ME inhibited intracellular oxidation. Low micromolar concentrations of 2-ME produce as much as a 12-fold increase in ChAT; this is enhanced further by inclusion of nerve growth factor (NGF). NAC effects are biphasic: 0.15 mM produces profound increases in ChAT while 1.5 mM has no effect. Immature (E16) cultures respond with increases in ChAT while more highly differentiated cultures (E18) do not. Labeling of single precursors with a lacZ-expressing retrovirus reveals that the increase in ChAT is due primarily to an increased number and size of clones, not an increase in cholinergic neurons per clone, suggesting an effect on precursor survival. Inhibition of ras farnesylation inhibits both 2-ME and NAC induction of ChAT suggesting a ras-mediated pathway. Inclusion of the MEK inhibitor PD98059 does not affect low doses of NAC, but at doses of NAC that fail to increase ChAT activity, inhibition of the pathway actually raises ChAT. Immunocytochemical investigation of the cultures indicates that cells exposed to low doses of NAC develop healthy neuronal arbors in the apparent absence of glial support. At higher concentrations of NAC, neurons were found in association with astrocytes, making contact via elaborate varicose fibers. Treatment of the cultures with PD98059 to inhibit MEK returned cultures to a 'low-dose' phenotype. These data suggest that redox status of basal forebrain precursors affect both their survival and differentiative potential.

Vasoactive intestinal peptide and pituitary adenylyl cyclase-activating polypeptide inhibit tumor necrosis factor-alpha production in injured spinal cord and in activated microglia via a cAMP-dependent pathway.

Tumor necrosis factor-alpha (TNF-alpha) production accompanies CNS insults of all kinds. Because the neuropeptide vasoactive intestinal peptide (VIP) and the structurally related peptide pituitary adenylyl cyclase-activating polypeptide (PACAP) have potent anti-inflammatory effects in the periphery, we investigated whether these effects extend to the CNS. TNF-alpha mRNA was induced within 2 hr after rat spinal cord transection, and its upregulation was suppressed by a synthetic VIP receptor agonist. Cultured rat microglia were used to examine the mechanisms underlying this inhibition because microglia are the likely source of TNF-alpha in injured CNS. In culture, increases in TNF-alpha mRNA resulting from lipopolysaccharide (LPS) stimulation were reduced significantly by 10(-7) m VIP and completely eliminated by PACAP at the same concentration. TNF-alpha protein levels were reduced 90% by VIP or PACAP at 10(-7) m. An antagonist of VPAC(1) receptors blocked the action of VIP and PACAP, and a PAC(1) antagonist blocked the action of PACAP. A direct demonstration of VIP binding on microglia and the existence of mRNAs for VPAC(1) and PAC(1) (but not VPAC(2)) receptors argue for a receptor-mediated effect. The action of VIP is cAMP-mediated because (1) activation of cAMP by forskolin mimics the action; (2) PKA inhibition by H89 reverses the neuropeptide-induced inhibition; and (3) the lipophilic neuropeptide mimic, stearyl-norleucine(17) VIP (SNV), which does not use a cAMP-mediated pathway, fails to duplicate the inhibition. We conclude that VIP and PACAP inhibit the production of TNF-alpha from activated microglia by a cAMP-dependent pathway.

Macrophage cell-conditioned medium promotes cholinergic differentiation of undifferentiated progenitors and synergizes with nerve growth factor action in the developing basal forebrain.

Conditioned medium from stimulated microglia and from the monocyte/macrophage cell line (RAW 264.7; MC-CM) promotes the differentiation of cholinergic neurons from undifferentiated progenitors in the septal nuclei and adjacent basal forebrain (BF). We have studied the regulation of this process by measuring the activity of choline acetyltransferase (ChAT) in cultured BF taken from embryonic day 16 rat brain. Inhibition of either xanthine oxidase with allopurinol or nitric oxide synthase with N(G)-monomethyl-l-arginine produces a small but significant improvement in the efficacy of MC-CM while inclusion of pyrrolidine dithiocarbamate, a hydroxyl radical scavenger widely used as an antioxidant, lowers MC-CM-induced ChAT activity. Addition of nerve growth factor (NGF) but not brain-derived neurotrophic factor or glial-derived neurotrophic factor together with MC-CM has a synergistic effect on both ChAT activity and ChAT mRNA, raising ChAT activity as much as 29-fold and ChAT mRNA almost 15-fold. While MC-CM raised mRNA for trkA, the effect was not synergistic with NGF. mRNA for the common neurotrophin receptor (p75NTR) showed a modest synergistic increase. Blockade of the Ras/Raf/ERK [extracellular signal-regulated kinase, also known as mitogen-activated protein [(MAP) kinase] signal transduction pathway with either PD28059 (an inhibitor of MAP kinase/ERK kinase kinase or MEK) or N-acetyl-S-farnesyl-l-cysteine (an inhibitor of Ras farnesylation and, hence, activation) inhibited the action of MC-CM. Moreover, a subpopulation of cells responded rapidly to MC-CM with an increased appearance of phosphorylated ERK. Because NGF also utilizes this pathway, synergy may occur along this signal transduction pathway.

Neurotrophins and the anti-inflammatory agents interleukin-4 (IL-4), IL-10, IL-11 and transforming growth factor-beta1 (TGF-beta1) down-regulate T cell costimulatory molecules B7 and CD40 on cultured rat microglia.

Microglia are essential for T cell activation in the CNS. Since T cell activation requires costimulation by B7 and/or CD40, we examined the regulation by cytokines of B7-1, B7-2 and CD40 mRNA expression in cultured rat microglia in serum-free medium. All three ligands are expressed constitutively, but are profoundly up-regulated by granulocyte-macrophage colony-stimulating factor (GM-CSF). By contrast, interferon-gamma raises only B7-2 and CD40 mRNA, and the B7-2 increase is inhibited by IL-10. IL-4, transforming growth factor-beta1, and nerve growth factor (NGF) repress GM-CSF-induced B7-2 and CD40, but not B7-1. NGF also down-regulates its own high-affinity trkA receptor. IL-11, unrecognized for its effect on antigen presentation, represses GM-CSF-induced B7-2.

Transforming growth factor-alpha expands progenitor cells of the basal forebrain, but does not promote cholinergic differentiation.

Transforming growth factor-alpha (TGF-alpha), a member of the epidermal growth factor (EGF) family, binds to the EGF-receptor (EGF-R). The early expression and widespread distribution of TGF-alpha and EGF-R in the developing central nervous system (CNS) suggest that TGF-alpha may play a role in the developing CNS. To study possible effects of TGF-alpha on cholinergic differentiation in the basal forebrain, we cultured septal nuclei with adjacent basal forebrain from embryonic rat brain in the presence and absence of TGF-alpha. At the highest dose of TGF-alpha used (100 ng/mL), activity of choline acetyltransferase (ChAT; EC 2.3.1.6) and the number of cholinergic neurons doubled. However, because protein levels tripled, specific ChAT activity actually declined. To determine the mechanism accounting for the increase in ChAT, we labeled dividing precursors present in the cultures with a replication-deficient retrovirus expressing beta-galactosidase in the presence and absence of TGF-alpha. By staining the cultures for both LacZ and ChAT, we determined that the precursor population expanded in size (individually labeled clones contained more cells), but the percentage of cholinergic neurons present in the clones was unchanged. Therefore, while TGF-alpha expands the precursor pool, it does not promote cholinergic differentiation. Interleukin-9, included to prompt neuronal differentiation, did not by itself increase ChAT activity, nor did it enhance the action of TGF-alpha. This was true even when basic fibroblast growth factor was included.

Interleukin-1 involvement in the induction of leukemia inhibitory factor mRNA expression following axotomy of sympathetic ganglia.

Axotomy of superior cervical (sympathetic) ganglia (SCG) results in increased neuropeptide gene expression. In vitro, neuropeptide gene expression is similarly increased by exposure to the inflammatory cytokine interleukin-1 (IL-1). The effect of IL-1 in-vitro has been shown to be mediated by leukemia inhibitory factor (LIF). Since IL-1 regulates neuropeptide expression via LIF in vitro, we asked whether axotomy in vivo produces an increase in LIF mRNA, and whether that increase is regulated by IL-1 activity. Within 6 h following axotomy, ganglionic LIF mRNA is substantially elevated. Moreover, axotomy produces a rapid and transient increase in intraganglionic IL-1 beta mRNA, followed rapidly by an increase in ICAM-1 mRNA, thereby suggesting a local source of IL-1 activity. Pretreatment with the anti-inflammatory agent dexamethasone (DEX) reduces the increases of both IL-1 beta and LIF mRNAs following axotomy. mRNA encoding the specific signal-transducing Type I IL-1 receptor is present in unlesioned SCG in vivo, and increases following axotomy. Local application of IL-1 beta in vivo induces LIF mRNA even in uninjured ganglia, though not to the extent seen with axotomy. DEX treatment blocks this IL-1 beta-mediated increase in LIF mRNA. Therefore, DEX blocks the induction of LIF mRNA by inhibiting both the production of IL-1 and its action on LIF gene expression. Axotomy of a homozygous IL-1 receptor type I gene knockout mouse leads to a delayed and/or diminished induction of LIF mRNA in SCG, but does not prevent LIF mRNA expression. We conclude that while IL-1 is likely to be involved in the cascade of gene expression that follows axotomy, it alone is not sufficient to mediate the full induction of LIF mRNA by axotomy.

Interleukin-1 beta increases leukemia inhibitory factor mRNA levels through transient stimulation of transcription rate.

Interleukin-1 beta (IL-1 beta) induces leukemia inhibitor factor (LIF) expression in a number of cell types including non-neuronal cells of the sympathetic superior cervical ganglion (SCG). Upregulation of LIF by inflammatory cytokines is usually associated with injury response. We characterized the molecular mechanism of LIF mRNA regulation by IL-1 beta in explanted neonatal rat SCG and a Schwann cell line. IL-1 beta increases LIF mRNA levels by interacting with IL-1 receptors in SCG, since this induction could be diminished by inclusion of either soluble IL-1 receptors or IL-1 receptor antagonist. The antiinflammatory glucocorticoid dexamethasone also inhibits LIF mRNA induction by IL-1 beta. LIF mRNA encodes a 3' AU-rich mRNA stability control sequence, but IL-1 beta does not appear to regulate the decay of LIF mRNA by this mechanism. IL-1 beta does not raise LIF gene transcription rate in cultured SCG 6 or 24 h after addition of IL-1 beta as measured by nuclear run-on assays. LIF gene transcription is induced repidly and transiently in an immortalized Schwann cell line, returning to uninduced rates by 1 h after induction. These results suggest that the IL-1 beta induction of LIF gene expression is at least partially transcriptional, but that LIF mRNA increases to a greater extent than LIF transcription, suggesting the possibility of posttranscriptional regulation as well.

Conditioned medium from activated microglia promotes cholinergic differentiation in the basal forebrain in vitro.

In earlier studies we found that treatment with interferon-gamma (IFN-gamma) produced an 8- to 11-fold increase in choline acetyltransferase (ChAT) in cultured cells taken from Embryonic Day 16 (E16) septal nuclei with adjacent basal forebrain (SN/BF). Since younger cultures responded even more profoundly to IFN treatment, we have tested the possibility that the action of IFN (or its intermediate; see below) is to prompt the cholinergic differentiation of neuronal precursors. SN/BF cultures of various ages were labeled with a retrovirus engineered to express beta-galactosidase (Lac-Z), and ChAT-positive descendants of the retrovirally labeled precursors were counted. IFN-gamma treatment of cultures caused as much as an 8.8-fold increase in the proportion of ChAT-positive cells present in Lac-Z-positive clones, suggesting that IFN promoted cholinergic differentiation in precursor populations. By contrast, bFGF increased clone size but did not change the proportion of ChAT-positive cells. NGF affected neither. Only ameboid microglia present in the cultures responded to IFN with characteristic nuclear translocation of the signal transducing molecule p91, suggesting that a microglial-derived molecule may mediate the action of IFN. Consistent with this hypothesis, conditioned media from cultures of enriched, activated microglia also increased ChAT activity in a dose-dependent fashion. Conditioned media from an unstimulated macrophage/monocyte cell line (RAW 264.7) also proved extremely efficacious in raising ChAT activity. In addition, conditioned media from both activated microglia and RAW 264.7 cells increased the proportion of ChAT-positive cells in retrovirally labeled clones to the same extent as IFN itself, suggesting the possibility that they contain the molecule(s) that mediates the action of IFN. Preliminary characterization of this molecule suggests that it is a very stable and large protein. Together these data suggest that a molecule promoting cholinergic differentiation is produced by activated microglia and other macrophage-like cells. The identity of this molecule and its precise role in normal development await its further purification.

Tumor necrosis factor-alpha induces substance P in sympathetic ganglia through sequential induction of interleukin-1 and leukemia inhibitory factor.

The present study establishes that tumor necrosis factor-alpha (TNF-alpha) induction of sympathetic substance P (SP) requires sequential induction of both interleukin (IL-1) and leukemia inhibitory factor (LIF). TNF-alpha dose-dependently induces SP, an induction that is secondary to an increase in the SP precursor, preprotachykinin (PPT), mRNA. Since TNF-alpha conditioned medium (CM) mimics the effect of TNF-alpha by raising SP, actions that are not antagonized by a neutralizing TNF-alpha antibody, TNF-alpha induction of SP is mediated by a soluble intermediate or intermediates. The blockade of TNF-alpha action by a specific IL-1 receptor antagonist and the induction of IL-1 mRNA by TNF-alpha suggest that IL-1 is one of the intermediates. Moreover, because immunoprecipitation with LIF antibodies decreases SP-inducing activity of TNF-alpha CM, and because LIF mRNA is also induced by TNF-alpha, LIF is a second intermediate. Furthermore, TNF-alpha-induced LIF mRNA is blocked by the IL-1 receptor antagonist, whereas IL-1-induced LIF mRNA is not affected by TNF-alpha antibodies, suggesting that TNF-alpha first induces IL-1, and IL-1 subsequently induces LIF. These data suggest that TNF-alpha induces SP in sympathetic ganglia through the sequential inductions of IL-1 and LIF.

Identification of mRNAs regulated by interferon-gamma in cultured rat astrocytes by PCR differential display.

We have used RNA differential display to identify specific mRNA species that are regulated by interferon-gamma (IFN gamma) treatment of enriched astrocyte cultures. Following a 24-hour treatment with IFN, mRNA was subjected to differential display using 15 different random primers. 105 upregulated and 5 downregulated cDNAs were selected for further sequencing and identification. Northern blot analysis confirmed the upregulation of 13 genes identical or highly similar to: intercrine adrenomedullin, H-rev 107, CAP-like protein, ATP synthase epsilon-subunit, complement C3, S-100 beta, Ca2+ ATPase, mg11, IFN-upregulated 56-kD protein mRNA, laminin receptor-like protein, protein tyrosine phosphatase, and zic. These data suggest that exposure to IFN gamma results in a complex change in the pattern of astrocyte gene expression.

Immune cytokine regulation of sympathetic ganglion response to injury.

Without appropriate response to traumatic injury, the nervous system cannot survive. The superior cervical (sympathetic) ganglion of the neonatal rat has been a uniquely valuable model system in which to study neuronal response to traumatic injury. Explantation of the ganglion into culture involves severing both presynaptic fibers (compromising preganglionic input) and axotomy (damaging efferent axons of ganglionic neurons). Since traumatic injury of any kind involves the elaboration of inflammatory cytokines, we have been interested in examining the action of these inflammatory cytokines on neurotransmitter gene expression in injured sympathetic ganglia.

Interactions of the nervous and immune systems in development, normal brain homeostasis, and disease.

Neurotransmitters, neuropeptides, neurotrophins, and neuroendocrine hormones have traditionally been assigned functions in normal development and homeostasis of neuronal networks; cytokines and adhesion molecules have been assigned functions within the peripheral immune system. Molecular dissection of the presence and function of these receptors and ligands during development of the immune and nervous systems, in normal healthy adult central and peripheral nervous tissue, and in the pathological response of immune elements in the brain and neuroelements in the immune system has forced us to alter these long-held concepts. Examples of how glia and neurons function in relationship to these paracrine and autocrine stimuli in health and disease are provided in this short review.

The interleukin-1-induced increase of substance P in sympathetic ganglia is not mediated by ciliary neurotrophic factor.

Interleukin-1 (IL-1) induction of substance P (SP) in cultured sympathetic ganglia requires a soluble intermediate molecule that is present in IL-1 conditioned medium (IL-1CM). One of the required intermediates is leukemia inhibitory factor (LIF; Shadiack et al., J Neurosci 13:2601-2609, 1993). In the present study we have examined the possibility that ciliary neurotrophic factor (CNTF) is another intermediate involved in the IL-1 induction of sympathetic SP. CNTF mimics the action of IL-1CM by raising both SP and choline acetyltransferase activity--actions that are blocked by a specific neutralizing antiserum for CNTF. However, IL-1CM and CNTF differ in their response to depolarizing agents: while KCl (40 mM) blocks the action of IL-1CM (and LIF), it enhances the action of CNTF. Furthermore, neither CNTF bioactivity nor CNTF protein is detected in IL-1CM. Neutralizing antiserum to CNTF fails to block the action of either IL-1 or IL-1CM, suggesting that neither a soluble nor a membrane-bound form of the molecule is active in direct response to IL-1 action. While Northern blots confirm the presence of both CNTF and CNTF receptor mRNA in neonatal ganglia, neither culturing nor IL-1 treatment alters these mRNA levels. These data taken together suggest that while CNTF is present and possibly active in sympathetic ganglia, it is not a mediator of the IL-1 induction of SP.

Interferon-gamma promotes cholinergic differentiation of embryonic septal nuclei and adjacent basal forebrain.

In cultured rat embryonic septal nuclei with adjacent basal forebrain, murine interferon-gamma (IFN gamma) produces a striking increase in choline acetyltransferase (ChAT) activity and mRNA. The effect of IFN gamma on cholinergic differentiation is more potent in E14 cultures than in older cultures. IFN gamma does not cause a change in the affinity of ChAT for choline, nor does it affect cell proliferation. Whereas IFN gamma doubles neuronal cell number, the cholinergic cell number increases more than 7-fold. Ameboid microglia respond to IFN gamma with the translocation of p91 to the nucleus. The action of IFN gamma is not mediated by NGF or bFGF. The enhancement of cholinergic expression that occurs with increased cell density may be partly attributable to an endogenous IFN gamma-like molecule, since antibodies to IFN gamma offset the effects of increased cell density.

Lipopolysaccharide induces substance P in sympathetic ganglia via ganglionic interleukin-1 production.

The immune cytokine interleukin-1 (IL-1) causes a pronounced elevation in substance P (SP) immunoreactivity and the mRNA coding for its preprotachykinin precursor in cultured superior cervical (sympathetic) ganglia (SCG; Jonakait and Schotland, 1990; Freidin and Kessler, 1991; Hart et al., 1991). In this study we have investigated the possibility that the SCG can respond to other immune stimulators, notably lipopolysaccharide (LPS), a product of bacterial cell walls. LPS treatment of cultured SCG resulted in a dose-dependent increase in SP. However, LPS did not induce SP in the absence of non-neuronal cells, suggesting the necessity of a non-neuronal cell-derived intermediate. Since the LPS induction of SP was partially blocked by a specific IL-1 receptor antagonist (IL-1ra) and since LPS induced approximately an 8-fold increase in mRNA coding for IL-1 itself, we concluded that IL-1 is at least one of these LPS-induced intermediates. TNF-alpha, which also raises SP levels, may be another. IL-6, which may also be increased by LPS, does not increase levels of SP. The synthetic glucocorticoid hormone dexamethasone (DEX) blocks the LPS induction of SP with a Ki approximating 8 x 10(-11) M. The inhibition is due in part to the blockade of the LPS induction of ganglionic IL-1 mRNA. Moreover, inhibition of the LPS induction of SP by indomethacin implies mediation of the effect through prostaglandins. The inhibition by indomethacin suggests a non-monocytic cell source since prostaglandins are thought to restrict the LPS induction of monocytic IL-1.(ABSTRACT TRUNCATED AT 250 WORDS)

Neural-immune interactions in sympathetic ganglia.

Effects of immune cytokines on neuronal gene expression have recently been examined in cultured superior cervical (sympathetic) ganglia, a widely used model system for the study of neurotransmitter plasticity. Following deafferentation and explantation into culture, interleukin-1 causes an up-regulation of the neuropeptide substance P as well as of choline acetyltransferase. Tumor necrosis factor-alpha has a similar, though less potent, action. Since interleukin-1 was ineffective in raising the concentration of substance P in pure neuronal cultures, the existence of a non-neuronally derived intermediate was postulated and found to exist in interleukin-1-conditioned medium. Antibody neutralization of either nerve growth factor or ciliary neurotrophic factor failed to affect the ability of interleukin-1 to induce substance P. Inhibition of prostaglandin biosynthesis was equally ineffective. However, immunoprecipitation of leukemia inhibitory factor from interleukin-1-conditioned medium eliminated substance-P-inducing activity, suggesting leukemia inhibitory factor as a possible interleukin-1-induced intermediate. The ability of interleukin-1 to induce leukemia inhibitory factor mRNA strengthens this conclusion. Glucocorticoid hormones block the interleukin-1 induction of leukemia inhibitory factor, which explains why they block the interleukin-1 induction of substance P.

Interleukin-1 induces substance P in sympathetic ganglia through the induction of leukemia inhibitory factor (LIF).

It has become increasingly clear that immune cytokines perform growth and differentiation functions in the nervous system similar to those performed in the immune system. In previous studies we have shown that interleukin-1 beta (IL-1 beta) raises substance P (SP) and the mRNA coding for its preprotachykinin precursor in cultured sympathetic superior cervical ganglia (SCG) (Jonakait and Schotland, 1990; Hart et al., 1991a). The action of IL-1 is blocked both by depolarization of the ganglia and by glucocorticoid hormones (Hart et al., 1991a). In the present report, we have found that IL-1 does not act directly upon neurons to raise SP, but rather induces the production of a soluble intermediate molecule that raises both SP and the cholinergic-specific enzyme ChAT. Its induction by IL-1 is blocked by the synthetic glucocorticoid hormone dexamethasone; its action is compromised under depolarizing conditions. Because medium conditioned by IL-1 (IL-1CM) is functionally similar to leukemia inhibitory factor (LIF), we sought to determine whether this molecule might be an active constituent of IL-1CM. Immunoprecipitation with an antiserum directed against LIF eliminated large proportions of SP-inducing activity from IL-1CM. In addition, steady-state levels of mRNA coding for LIF are increased by IL-1 treatment of SCG. These data suggest that LIF, induced by IL-1, may ultimately be responsible for the IL-1 induction of SP.