Cellular and molecular regulation of tumor necrosis factor-alpha production by pentoxifylline

Tumor necrosis factor-alpha (TNF), a mononuclear phagocyte (MO)-derived peptide, is increasingly being recognized for its pleomorphic immunologic effects. A number of investigations have demonstrated that lipopolysaccharide (LPS) can induce TNF synthesis, yet mechanisms that regulate TNF expression at the cellular and molecular levels have not been fully elucidated. In this study, we present data demonstrating pentoxifylline, a methylxanthine, is efficacious in suppressing LPS-induced MO-derived TNF at the level of both TNF mRNA accumulation and TNF supernatant bioactivity. Pentoxifylline, at a dose of 1 x 10(-5)M, suppressed the production of both biologically active TNF and TNF mRNA expression by more than 50%. Furthermore, additional methylxanthines and dibutyryl cAMP have similar effects on TNF expression. These data support the mechanism for this suppressive effect is via the generation of intracellular cAMP.     

Equine luteal function regulation may depend on the interaction between cytokines and vascular endothelial growth factor: an in vitro study

We hypothesized that cytokines influence luteal angiogenesis in mares, while angiogenic factors themselves can also regulate luteal secretory capacity. Therefore, the purpose of this study was to evaluate the role of cytokines--tumor necrosis factor alpha (TNF), interferon gamma (IFNG) and Fas ligand (FASL)--on in vitro modulation of angiogenic activity and mRNA level of vascular endothelial growth factor A (VEGF), its receptor VEGFR2, thrombospondin 1 (TSP1), and its receptor CD36 in equine corpus luteum (CL) throughout the luteal phase. After treatment, VEGF protein expression was determined in midluteal phase (mid) CL cells. The role of VEGF on regulation of luteal secretory capacity was assessed by progesterone (P(4)) and prostaglandin E(2) (PGE(2)) production and by mRNA levels for steroidogenic enzymes 3-beta-hydroxysteroid dehydrogenase (3betaHSD) and PGE synthase (PGES). In early CL cells, TNF increased angiogenic activity (bovine aortic endothelial cell viability) and VEGF and VEGFR2 mRNA levels and decreased CD36 (real-time PCR relative quantification). In mid-CL cells, TNF increased VEGF mRNA and protein expression (Western blot analysis) and reduced CD36 mRNA levels, while FASL and TNF+IFNG+FASL decreased VEGF protein expression. In late CL cells, TNF and TNF+IFNG+FASL reduced VEGFR2 mRNA, but TNF+IFNG+FASL increased TSP1 and CD36 mRNA. VEGF treatment increased mRNA levels of 3betaHSD and PGES and secretion of P(4) and PGE(2). In conclusion, these findings suggest a novel auto/paracrine action of cytokines, specifically TNF, on the up-regulation of VEGF for angiogenesis stimulation in equine early CL, while at luteolysis, cytokines down-regulated angiogenesis. Additionally, VEGF stimulated P(4) and PGE(2) production, which may be crucial for CL establishment.     

Prostaglandin E2 inhibits TNF production in murine bone marrow-derived dendritic cells

Exposure to pathogens induces dendritic cells to release inflammatory cytokines and chemokines. The inflammatory response is controlled by endogenous agents such as anti-inflammatory cytokines, glucocorticoids, anti-inflammatory neuropeptides, and lipid mediators. This study is the first report on the inhibition by prostaglandin E2 (PGE2) of TNF release from bone marrow-derived dendritic cells stimulated with lipopolysaccharide (LPS), a TLR4 ligand, or peptidoglycan, a TLR2 ligand. The inhibition of TNF occurs at both mRNA and protein level. The inhibitory effect of PGE2 is mediated by the EP2 and EP4 receptors, and involves both PKA signaling and mediation by DC-derived IL-10. Intraperitoneal administration of PGE2 together with LPS results in a reduction in serum TNF and intracellular TNF in peritoneal exudate cells, compared to LPS alone. In addition, administration of PGE2 in vivo reduces the numbers of CD11c+ DCc that accumulate in the peritoneal cavity in response to LPS. The various implications of the PGE2-induced reduction in TNF are discussed.     

Regulation of murine mononuclear phagocyte inflammatory products by macrophage colony-stimulating factor. Lack of IL-1 and prostaglandin E2 production and generation of a specific IL-1 inhibitor

The influence of macrophage (M)-CSF on the production of inflammatory mediators has been examined in murine peritoneal macrophages. Cultures of macrophages treated with up to 30,000 U/ml of human rM-CSF or with 10,000 U/ml of L929-derived M-CSF did not reveal either PGE2, IL-1, or IL-6 secretion. In contrast, LPS, which served as a positive control, stimulated production of significant levels of PGE2, IL-1, and IL-6. Furthermore, Northern blot analysis of macrophage RNA revealed a strong induction of IL-1 alpha and IL-6 mRNA by LPS but not by M-CSF. Conditioned medium from macrophage cultures treated with purified L929 or human rM-CSF in combination with LPS exhibited a significant reduction of IL-1 bioactivity as compared with an LPS challenge alone. To investigate the mechanism involved in this M-CSF-dependent reduction of IL-1 bioactivity, we measured IL-1 alpha gene expression. The addition of M-CSF to LPS-treated macrophages did not affect IL-1 alpha mRNA levels suggesting that M-CSF may regulate production of an IL-1 inhibitor. This hypothesis was shown to be valid because removal of IL-1 alpha from conditioned medium of LPS plus M-CSF-treated cells allowed the detection of a nondialyzable factor that blocked IL-1-dependent thymocyte proliferation. The inhibitor appeared to be specific because it did not inhibit IL-2 and TNF bioactivities. Furthermore, this IL-1 inhibitor, which binds to cells and not to IL-1, competed with the binding of radioactive IL-1 alpha or beta to EL-4.6.1 cells. The results demonstrate that M-CSF alone does not induce proinflammatory mediators and PGE2 as was previously published. The data also suggest that M-CSF may play a role in the down-regulation of inflammatory responses.     

Prostaglandin E(2) increases bovine leukemia virus tax and pol mRNA levels via cyclooxygenase 2: regulation by interleukin-2, interleukin-10, and bovine leukemia virus

Prostaglandin E(2) (PGE(2)), produced by macrophages, has important immune regulatory functions, suppressing a type 1 immune response and stimulating a type 2 immune response. Type 1 cytokines (interleukin-2 [IL-2], IL-12, and gamma interferon) increase in freshly isolated peripheral blood mononuclear cells (PBMCs) of animals with an early disease stage of bovine leukemia virus (BLV) infection, while IL-10 increases in animals with a late disease stage. Although IL-10 has an immunosuppressive role in the host immune system, IL-10 also inhibits BLV tax and pol mRNA levels in vitro. In contrast, IL-2 stimulates BLV tax and pol mRNA and p24 protein expression in cultured PBMCs. The inhibitory effect of IL-10 on BLV expression depends on soluble factors secreted by macrophages. Thus, we hypothesized that PGE(2), a cyclooxygenase 2 (COX-2) product of macrophages, may regulate BLV expression. Here, we show that the level of COX-2 mRNA was decreased in PBMCs treated with IL-10, while IL-2 enhanced the level of COX-2 mRNA. Addition of PGE(2) stimulated BLV tax and pol mRNA levels and reversed the IL-10 inhibition of BLV mRNA. In addition, the specific COX-2 inhibitor, NS-398, inhibited the amount of BLV mRNA detected. Addition of PGE(2) increased BLV tax mRNA regardless of NS-398 addition. PGE(2) inhibited antigen-specific PBMC stimulation, suggesting that stimulation of BLV tax and pol mRNA levels by PGE(2) is independent of cell proliferation. These findings suggest that macrophage-derived COX-2 products, such as PGE(2), regulate virus expression and disease progression in BLV infection.     

Prostaglandin E2 suppresses chemokine production in human macrophages through the EP4 receptor

Pro-inflammatory pathways participate in the pathogenesis of atherosclerosis. However, the role of endogenous anti-inflammatory pathways in atheroma has received much less attention. Therefore, using cDNA microarrays, we screened for genes regulated by prostaglandin E(2) (PGE(2)), a potential endogenous anti-inflammatory mediator, in lipopolysaccharide (LPS)-treated human macrophages (MPhi). PGE(2) (50 nm) attenuated LPS-induced mRNA and protein expression of chemokines including monocyte chemoattractant protein-1, interleukin-8, macrophage inflammatory protein-1alpha and -1beta, and interferon-inducible protein-10. PGE(2) also inhibited the tumor necrosis factor-alpha-, interferon-gamma-, and interleukin-1beta-mediated expression of these chemokines. In contrast to the case of MPhi, PGE(2) did not suppress chemokine expression in human endothelial and smooth muscle cells (SMC) treated with LPS and pro-inflammatory cytokines. To assess the potential paracrine effect of endogenous PGE(2) on macrophage-derived chemokine production, we co-cultured MPhi with SMC in the presence of LPS. In these co-cultures, cyclooxygenase-2-dependent PGE(2) production exceeded that in the mono-cultures, and MIP-1beta declined significantly compared with MPhi cultured without SMC. We further documented prominent expression of the PGE(2) receptor EP4 in MPhi in both culture and human atheroma. Moreover, a selective EP4 antagonist completely reversed PGE(2)-mediated suppression of chemokine production. Thus, endogenous PGE(2) may modulate inflammation during atherogenesis and other inflammatory diseases by suppressing macrophage-derived chemokine production via the EP4 receptor.     

Differential mRNA expression of prostaglandin receptor subtypes in macrophage activation

Assessing the regulation of macrophage receptors for prostaglandin (PGE2) is essential to understanding the control which that potent lipid mediator has in modulating macrophage activities. The purpose of this study was to assess the differential mRNA expression of PGE2 receptor subtypes (EP) during macrophage exposure to activating and transducing agents. RAW 264.7 macrophages constitutively expressed mRNA for EP2,EP3 and EP4 receptor subtypes. Messenger RNA for EP4 was expressed at a much higher level when compared to EP2 in unstimulated macrophages as assessed by kinetic quantitative RT-PCR. When macrophages were stimulated with LPS, EP2 m RNA levels were 12-fold higher when compared to unstimulated macrophages, while EP4 m RNA remained unchanged. Conversely, mRNA levels of both EP2 and EP4 receptors were lower after macrophages were treated with IFN-gamma. Messenger RNA levels of both receptors were lower in macrophages after treatment with PGE2 or dibutyryl (db) cAMP Addition of the PKA inhibitor H89 reversed the effects of PGE2 and dbcAMP to varying degrees. Proteosome and p38 MAP kinase inhibitors blocked the LPS-stimulated increase in EP2 mRNA levels. Those inhibitors had no effect on EP4 mRNA.Thus, activating agents such as LPS and IFN-gamma may differentially regulate mRNAfor PGE2 receptor types in macrophages but the ligand and its associated signal transducing factors probably have similar regulatory effects.     

Biochemical events accompanying macrophage activation and the inhibition of colony-stimulating factor-1-induced macrophage proliferation by tumor necrosis factor-alpha, interferon-gamma, and lipopolysaccharide.

Agents that can arrest cellular proliferation are now providing insights into mechanisms of growth factor action and how this action may be controlled. It is shown here that the macrophage activating agents tumor necrosis factor-alpha (TNF alpha), interferon-gamma (IFN gamma), and lipopolysaccharide (LPS) can maximally inhibit colony stimulating factor-1 (CSF-1)-induced, murine bone marrow-derived macrophage (BMM) DNA synthesis even when added 8-12 h after the growth factor, a period coinciding with the G1/S-phase border of the BMM cell cycle. This inhibition was independent of autocrine PGE2 production or increased cAMP levels. In order to compare the mode of action of these agents, their effects on a number of other BMM responses in the absence or presence of CSF-1 were examined. All three agents stimulated BMM protein synthesis; TNF alpha and LPS, but not IFN gamma, stimulated BMM Na+/H+ exchange and Na+,K(+)-ATPase activities, as well as c-fos mRNA levels. IFN gamma did not inhibit the CSF-1-induced Na+,K(+)-ATPase activity. TNF alpha and LPS inhibited both CSF-1-stimulated urokinase-type plasminogen activator (u-PA) mRNA levels and u-PA activity in BMM, whereas IFN gamma lowered only the u-PA activity. In contrast, LPS and IFN gamma, but not TNF alpha, inhibited CSF-1-induced BMM c-myc mRNA levels, the lack of effect of TNF alpha dissociating the inhibition of DNA synthesis and decreased c-myc mRNA expression for this cytokine. These results indicate that certain biochemical responses are common to both growth factors and inhibitors of BMM DNA synthesis and that TNF alpha, IFN gamma, and LPS, even though they all have a common action in suppressing DNA synthesis, activate multiple signaling pathways in BMM, only some of which overlap or converge.     

Induction of macrophage TNF alpha, IL-1, IL-6, and PGE2 production by DTH-initiating factors

The elicitation of delayed-type hypersensitivity (DTH) reactions in mice is due to the sequential action of two different, antigen-specific, Thy-1+ cells. We have previously cloned the early-acting DTH-initiating cell from nude mice that were immunized and boosted by contact sensitization with oxazolone (OX). This cell clone, WP-3.27, releases an antigen-specific factor (OX-F) that sensitizes mast cells such that specific antigen challenge will induce serotonin release which mediates the early phase of DTH. In normal mice contact sensitized with picryl chloride (PCl), a similar polyclonal factor (PCl-F) has a similar activity and is also known to bind to macrophages. Thus, we measured macrophage production of TNF alpha, IL-1, IL-6, and PGE2 in response to the hapten affinity-purified DTH-initiating factors OX-F and PCl-F. Both factors induced significant release of each cytokine and PGE2. The production of TNF alpha, IL-1, and IL-6 was measured by bioassays. Northern blot analysis showed rapid accumulation of cytokine mRNA (2-4 hr), while maximal production of PGE2 occurred at approximately 8 hr. These macrophage activating properties of OX-F and PCl-F were not due to contamination with LPS as determined by the low levels of LPS present in OX-F and PCl-F and by the failure of polymyxin B to inhibit factor-induced PGE2 and TNF alpha production. Also, macrophage activation was shown not to be due to the action of several lymphokines known to be produced by WP3.27. Separation of OX-F and PCl-F by preparative isoelectric focusing showed a similar pattern: there were two major peaks of PGE2-inducing activity observed for both factors (for PCl-F at pI of 2-3 and 5.0, and for OX-F at pI of 3.5-4 and 5.0), but not for a sham factor produced by WEHI-3 cells. The ability of DTH-initiating factors to rapidly induce macrophage cytokine release and PGE2 synthesis 4-6 hr later may suggest a role for these mediators during the respective early vascular and late cellular phases of inflammation in DTH.     

Prostaglandins as endogenous mediators of interleukin 1 production

We examined the role of cyclooxygenase (CO)-derived metabolites of arachidonic acid (AA) in the regulation of interleukin 1 (IL 1) production by lipopolysaccharide (LPS)-stimulated murine resident peritoneal macrophages. The use of LPS proved to be an efficacious probe, because it stimulated both IL 1 production and AA metabolism via only the CO pathway. The production of the CO metabolites prostaglandin E2 (PGE2) and prostaglandin I2 (PGI2; measured as its stable metabolite 6-Keto prostaglandin F1 alpha) by LPS-stimulated macrophages was demonstrated by high pressure liquid chromatography and radioimmunoassay. The addition of exogenous PGE2 or PGI2 resulted in a dose-dependent suppression of macrophage IL 1 production. Inhibitors of the CO pathway (indomethacin, piroxicam, and ibuprofen) caused a dose-dependent augmentation in the LPS-induced IL 1 response. This augmentation directly correlated with the efficacy of the compounds as CO inhibitors. Similar results were found when macrophage-derived fibroblast growth factor was assessed. The addition of exogenous IL 1 to macrophage cultures caused an increase in the levels of PGE2, over a narrow dose range (0.05 to 0.6 IL 1 units). These studies provide detailed evidence that AA metabolites synthesized via the CO pathway can modulate the production of growth factors by LPS-stimulated macrophages. In addition, our data support the concept that IL 1, as with classical hormones, can regulate its own production through a self-induced inhibitor, PGE2.     

Activated human dendritic cells express inducible cyclo-oxygenase and synthesize prostaglandin E2 but not prostaglandin D2

Prostaglandins (PG) are well known lipid mediators with important immunoregulatory properties. While exogenous PGE2 has the ability to modulate the function and maturation of antigen presenting cells, such as dendritic cells (DC), it is not clear whether human DC have the capacity to synthesize PGE2 and other prostaglandins themselves. We therefore examined the expression of inducible cyclo-oxygenase (COX-2) by monocyte derived DC and the production of PGE2 and PGD2. Both monocyte derived DC and freshly isolated blood myeloid DC expressed little COX-2 constitutively, though COX-2 expression was rapidly but transiently upregulated in response to lipopolysaccharide stimulation. COX-2 mRNA was detectable within 1 h of LPS exposure, peaked at 4-6 h, and rapidly declined thereafter. COX-2 expression was accompanied by DC synthesis of PGE2, with peak levels present at 6-18 h post-stimulation. In contrast, PGD2 synthesis was not detected at any time point. When DC were activated with LPS in the presence of nimesulide, a COX-2 selective inhibitor, IL-10 synthesis was inhibited, indicating that endogenous prostaglandins regulate DC cytokine production. PGE2 production by DC may therefore modulate DC and T-cell function, thereby shaping the character of the immune response.     

Lps induces IL-6 in the brain and in serum largely through TNF production

We investigated the relative contribution of IL-6 and PGE2 directly induced by LPS and indirectly induced via TNF, using in vivo and in vitro models in the mouse. In these models we have used as tools an anti-TNF antibody and a cyclooxygenase inhibitor, the S enantiomer of ketoprofen (S-KPF). Anti-TNF antibodies inhibited LPS-induced IL-6 production in three different models: IL-6 production by mouse peritoneal macrophages in vitro; serum IL-6 levels induced by intraperitoneal LPS; and brain IL-6 levels induced by an intracerebroventricular injection of LPS. However, in vitro anti-TNF antibodies, did not inhibit LPS-induced PGE2, indicating that this effect is not mediated by TNF. Since PGE2 has an opposite effect on TNF and IL-6 production, inhibiting that of TNF but inducing that of IL-6, we investigated the effect of S-KPF on TNF and IL-6 production in vivo following LPS injection. Both TNF and IL-6 induction was augmented by S-KPF, but anti-TNF antibodies abolished the augmentation of IL-6 production. Thus, the effect of anti-inflammatory drugs on IL-6 production in some models can be secondary to their effect on TNF production.     

Engagement of tumor necrosis factor mRNA by an endotoxin-inducible cytoplasmic protein.

BACKGROUND: Tumor necrosis factor (TNF) production by macrophages plays an important role in the host response to infection. TNF-alpha gene expression in RAW 264.7 macrophages is predominantly regulated at the translational level. A key element in this regulation is an AU-rich (AUR) sequence located in the 3' untranslated region (UTR) of TNF mRNA. In unstimulated macrophages, the translation of TNF mRNA is inhibited via this AUR sequence. Upon stimulation with LPS, this repression is overcome and translation occurs. In this study, we attempted to identify cellular proteins that interact with the AUR sequence and thereby regulate TNF mRNA translation. MATERIALS AND METHODS: RNA probes corresponding to portions of TNF mRNA 3' UTR were synthesized. These labeled RNAs were incubated with cytoplasmic extracts of either unstimulated or lipopolysaccharides (LPS)-stimulated RAW 264.7 macrophages. The RNA/protein complexes formed were analyzed by gel retardation. Ultraviolet (UV) cross-linking experiments were performed to determine the molecular weight of the proteins involved in the complexes. RESULTS: TNF mRNA AUR sequence formed two complexes (1 and 2) of distinct electrophoretic mobilities. While the formation of complex 1 was independent of the activation state of the macrophages from which the extracts were obtained, complex 2 was detected only using cytoplasmic extracts from LPS-stimulated macrophages. Upon UV cross-linking, two proteins, of 50 and 80 kD, respectively, were capable of binding the UAR sequence. The 50-kD protein is likely to be part of the LPS-inducible complex 2, since its binding ability was enhanced upon LPS stimulation. Interestingly, complex 2 formation was also triggered by Sendaï virus infection, another potent activator of TNF mRNA translation in RAW 264.7 macrophages. In contrast, complex 2 was not detected with cytoplasmic extracts obtained from B and T cell lines which are unable to produce TNF in response to LPS. Protein tyrosine phosphorylation is required for LPS-induced TNF mRNA translation. Remarkably, the protein tyrosine phosphorylation inhibitor herbimycin A abolished LPS-induced complex 2 formation. Complex 2 was already detectable after 0.5 hr of LPS treatment and was triggered by a minimal LPS dose of 10 pg/ml. CONCLUSIONS: The tight correlation between TNF production and the formation of an LPS-inducible cytoplasmic complex suggests that this complex plays a role in the translational regulation of TNF mRNA.