Effects of human recombinant tumor necrosis factor-alpha and interleukin 1 on the synthesis of glycosaminoglycan and DNA in cultured rat costal chondrocytes

We examined effects of human rTNF alpha on the synthesis of glycosaminoglycan and DNA in cultured rat costal chondrocytes. The effects of human recombinant IL-1 alpha and IL-1 beta were also given attention. rTNF alpha, as well as rIL-1 alpha and rIL-1 beta, decreased the incorporation of [35S]sulfate into glycosaminoglycan to about 10% of the levels in the control. The half-maximal doses of rTNF alpha, rIL-1 alpha or rIL-1 beta required for the suppression of glycosaminoglycan synthesis (by rTNF alpha, rIL-1 alpha, and rIL-1 beta) were 2 ng/ml, 30 ng/ml, or 5 ng/ml, respectively. rTNF alpha stimulated incorporation of [3H]thymidine in the chondrocytes in a dose- and time-dependent manner. DNA synthesis was increased to about threefold over the control cultures in the presence of 1 microgram/ml rTNF alpha for 72 hr. The stimulatory effect of rTNF alpha on DNA synthesis was observed in both subconfluent and confluent cultures, whereas rIL-1 alpha and rIL-1 beta had no stimulatory activity on DNA synthesis. The addition of rTNF alpha to the cultures of chondrocytes stimulated DNA synthesis, even in medium containing no fetal calf serum. The fetal calf serum acted synergistically with rTNF alpha in increasing DNA synthesis. We propose that both TNF and IL-1 may be involved in inflammatory diseases of cartilage, and that TNF alpha, but not IL-1, may have some physiologic growth factor function for chondrocytes.     

Tumor necrosis factor-alpha enhances inhibitory effects of interleukin-1 beta on Leydig cell steroidogenesis

Human recombinant tumor necrosis factor-alpha (rTNF alpha) alone (up to 1000 units/ml) did not alter either basal or human chorionic gonadotropin (hCG)-induced testosterone formation in primary culture of rat Leydig cells. However, concomitant addition of rTNF alpha with human recombinant interleukin-1 beta (rIL-1 beta) enhanced the inhibitory effects of rIL-1 beta. The rIL-1 beta dose response curve was shifted to the left (IC50 changed from 1 ng/ml to 0.3 ng/ml). Even though rTNF alpha had no effect on testosterone formation, hCG-stimulated cyclic AMP formation was inhibited by rTNF alpha in a dose dependent manner. In the presence of both rTNF alpha and rIL-1 beta, hCG-induced cyclic AMP formation and binding of [125I]-hCG to Leydig cells were further inhibited. Testicular macrophages represent about 20% of the interstitial cells. TNF alpha and IL-1 may be produced locally by interstitial macrophages and have paracrine effects on Leydig cell function.     

Blockade of prostaglandin biosynthesis in intact mice dramatically augments the expansion of committed myeloid progenitor cells (colony-forming units-granulocyte, macrophage) after acute administration of recombinant human IL-1 alpha

Injection of human rIL-1 alpha in intact normal mice has positive and negative effects on myelopoiesis. Within 6 h postinjection, peripheral neutrophilia can be demonstrated. However, bone marrow and spleen cells capable of inhibiting CFU-granulocyte macrophage proliferation are detected between 6 and 48 h postinjection. These myelopoietic suppressor cells belong to the monocytic lineage and are identical to inhibitory cells induced by PGE2. Treatment of mice with indomethacin, a PG synthesis inhibitor, completely blocked the generation of IL-1-alpha-induced myelopoietic suppressor cells, and significantly enhanced femoral and splenic CFU-GM proliferation after a single injection of 0.4 microgram/mouse IL-1. The peripheral blood neutrophilia observed within 6 h after IL-1 injection was delayed to 18 to 24 h postinjection in indomethacin-pretreated mice. In mice treated with four consecutive daily injections of 0.4 microgram IL-1, a sustained peripheral neutrophilia was observed. IL-1 had little effect on femoral CFU-GM in these animals, however, splenic CFU-GM was increased 7- to 10-fold by 4 to 7 days postinjection. In IL-1 plus indomethacin-treated mice, sustained peripheral neutrophilia was observed although to a lesser degree than with IL-1 alone. Marrow CFU-GM were relatively unaffected, however, splenic CFU-GM were increased by 27-fold. These results indicate that the in vivo administration of IL-1 results in neutrophilia and generation of myelopoietic suppressive effects, mediated by cyclo-oxygenase pathway products. Blockade of PG synthesis by using the cyclo-oxygenase inhibitor indomethacin abrogates the myelopoietic suppressive effects associated with IL-1 administration and optimizes its myelopoietic stimulatory capacity. The inclusion of a cyclo-oxygenase inhibitor may have significant relevance to the clinical use of IL-1.     

Species differences in human and rat islet sensitivity to human cytokines. Monoclonal anti-interleukin-1 (IL-1) influences on direct and indirect IL-1-mediated islet effects

Species differences in sensitivity to human recombinant cytokines were observed when human or rat islets were co-cultured with human recombinant cytokines for 6 days. Suppression of both human and rat islet insulin secretion resulted from co-culture with recombinant interleukin-1 alpha (rIL-1 alpha) or interleukin-1 beta (rIL-1 beta); however, direct rIL-1 alpha and rIL-1 beta cytotoxicity was seen with rat islets but not with human islets. Human islet insulin secretion was also suppressed during co-culture with recombinant tumor necrosis factor (rTNF) or interferon (rIFN), but not with lymphotoxin (rLT) or rIL-6; rat islet insulin secretion was not suppressed by any of these cytokines. No direct cytotoxic effects resulted from co-culture of human islets with rLT, rTNF, rIFN, or rIL-6; rLT was slightly cytotoxic for rat islets. Human islet cytotoxic synergy occurred between rLT and rIL-1 alpha, rIL-1 beta, or rIFN; synergy in suppression of human islet insulin secretion occurred between rLT and rIL-1 beta, and between rIFN and rTNF. Pretreatment of rIL-1 with monoclonal antibody (mAb) specific for non-crossreactive epitopes on rIL-1 alpha (H43 and H12) or rIL-1 beta (H34 and H21) prevented islet cytotoxic synergy between rIL-1 alpha or rIL-1 beta, respectively, and rLT. Although all four mAb's neutralize the thymocyte and fibroblast stimulatory activities of rIL-1 alpha or rIL-1 beta, mAb H21 does not neutralize rIL-1 beta activity against rat islets. Implications for cytokine-mediated islet cytotoxicity and suppression of insulin secretion are discussed.     

Cytokine activation of vascular endothelium. Effects on tissue-type plasminogen activator and type 1 plasminogen activator inhibitor

Regulation of the fibrinolytic system of cultured human umbilical vein endothelial cells (HUVECs) by recombinant interleukin 1 beta (rIL-1 beta) and tumor necrosis factor alpha (rTNF alpha) was investigated. Functional and immunologic assays indicated that both cytokines decreased HUVEC tissue-type plasminogen activator (tPA) and increased type 1 plasminogen activator inhibitor (PAI-1) in a dose- and time-dependent manner. Maximal effects (50% decrease in tPA antigen; 300-400% increase in PAI-1 activity) were achieved with 2.5 units/ml rIL-1 beta and 200 units/ml rTNF alpha. Combinations of rIL-1 beta and rTNF alpha were not additive at these maximal concentrations. After a 24-h pretreatment with rIL-1 beta, HUVECs secreted tPA at one-quarter of the rate of control cells and released PAI-1 at a rate that was 5-fold higher than controls. Neither the basal rate of PAI-1 release nor the increased rate of release of PAI-1 in response to rIL-1 beta was affected by subsequently treating the cells with secretagogues (e.g. phorbol myristate acetate) suggesting that PAI-1 is not contained within a rapidly releasable, intracellular storage pool. Northern blot analysis using a PAI-1 cDNA probe indicated that the cytokines increased the steady-state levels of the 3.2- and 2.3-kb PAI-1 mRNA species, but with a preferential increase in the larger mRNA form. The fact that both rIL-1 beta and rTNF alpha act in a similar manner strengthens the hypothesis that the local development of inflammatory/immune processes could reduce endothelial fibrinolytic activity.     

Interleukin 1 induces interleukin 1. II. Recombinant human interleukin 1 induces interleukin 1 production by adult human vascular endothelial cells

Interleukin 1 (IL-1) alters several potentially pathogenic endothelial cell (EC) functions. The authors report here that recombinant human IL-1 (rIL-1) alpha (0.1 to 10 ng/ml) or IL-1-beta (1 to 100 ng/ml) induce concentration- and time-dependent increases in IL-1-beta mRNA levels in EC derived from adult human saphenous vein. rIL-1 induced IL-1-alpha mRNA only in EC treated concomitantly with cycloheximide (2 micrograms/ml). IL-1-beta mRNA production began within 1 hr of exposure to rIL-1, peaked after 24 hr, and declined thereafter. Actinomycin D prevented the appearance of IL-1 mRNA in rIL-1-treated EC. rIL-1 also induced the release of biologically active IL-1 from EC, which was inhibited by cycloheximide (1 microgram/ml). When compared on the basis of their activity in the thymocyte costimulation assay, rIL-1-alpha and rIL-1-beta were equipotent as inducers of IL-1 production by EC. EC stimulated with rIL-1 produced prostaglandin E2, which inhibits IL-1 production by other cell types and also decreases the responsiveness of thymocytes to IL-1. When EC were exposed to rIL-1 in the presence of indomethacin (1 microgram/ml), which blocked prostaglandin E2 production, greater amounts of rIL-1-induced IL-1 release were detected, although the inhibitor did not affect IL-1-beta mRNA levels. IL-1-induced IL-1 production was unlikely to be caused by endotoxin contamination of tissue culture media or IL-1 preparations, because the lipopolysaccharide (LPS) antagonist polymyxin B (10 micrograms/ml) blocked LPS-induced IL-1 production by EC but did not affect IL-1 release in response to rIL-1-beta (100 ng/ml). The IL-1-inducing property of rIL-1-beta was heat-labile, whereas heated LPS stimulated EC IL-1 production. The source of IL-1 in our cultures was not monocyte/macrophages, as treatment of EC with monoclonal antibody to the monocyte antigen Mo2 under conditions that lysed adherent peripheral blood monocytes did not affect production of IL-1 by EC in response to LPS (1 microgram/ml) or rIL-1-beta (100 ng/ml). IL-1 elicits a coordinated program of altered endothelial function that increases adhesiveness for leukocytes and coagulability. IL-1-induced IL-1 gene expression in human adult EC could thus provide a positive feedback mechanism in the pathogenesis of vascular disease including atherosclerosis, vasculitis, and allograft rejection.     

Interleukin 1 and tumor necrosis factor synergistically stimulate prostaglandin synthesis and phospholipase A2 release from rat renal mesangial cells

Treatment of rat glomerular mesangial cells with recombinant human interleukin 1 alpha (rIL-1 alpha), recombinant human interleukin 1 beta (rIL-1 beta) or recombinant human tumor necrosis factor (rTNF) induces prostaglandin E2 (PGE2) synthesis and the release of a phospholipase A2 (PLA2) activity. rIL-1 beta is significantly more potent than rIL-1 alpha or rTNF in stimulating PGE2 as well as PLA2 release from mesangial cells. When given together, rTNF interacts in a synergistic fashion with rIL-1 alpha and rIL-1 beta to enhance both, PGE2 synthesis and PLA2 release. The released PLA2 has a neutral pH optimum and is calcium-dependent. Pretreatment of cells with actinomycin D or cycloheximide inhibits basal and cytokine-stimulated PGE2 and PLA2 release.     

Potent inhibition of gastric acid secretion by intravenous interleukin-1 beta and -1 alpha in rats.

The influence of human and rat recombinant interleukin-1 (hIL-1 beta and -1 alpha and rIL-1 beta) on acid secretion was investigated in conscious pylorus-ligated rats. Intravenous injection of either hIL-1 beta, hIL-1 alpha or rIL-1 beta dose dependently inhibited gastric acid output with an ED50 of 0.05 microgram, 0.5 microgram and 2.2 micrograms, respectively. The antisecretory action of IL-1 beta was associated with an increase in circulating levels of gastrin. hIL-1 beta-induced inhibition of acid secretion was dose dependently reversed by peripheral injection of the IL-1 receptor antagonist, IL-RA, with a dose ratio of 1:10(3) for complete reversal. The inhibitory effect of hIL-1 beta was blocked by indomethacin and was not modified by IV injections of the CRF receptor antagonist, alpha-helical CRF(9-41), or the monoclonal somatostatin antibody CURE.S6, or by systemic capsaicin pretreatment. These results show that systemic hIL-1 beta-induced inhibition of gastric acid secretion is mediated through IL-1 receptors and prostaglandin pathways, and does not involves CRF receptors, afferent fibers, or changes in circulating gastrin or somatostatin levels.     

IL-2 stimulates the production of IL-1 alpha and IL-1 beta by human peripheral blood mononuclear cells

Human PBMC were cultured in medium containing human rIL-2, and the supernatants and cell lysates were analyzed for IL-1 alpha and IL-1 beta using specific RIA. IL-2, but not the excipient detergents included in the rIL-2 preparation, induced the synthesis of both cytokines. The concentrations of IL-1 alpha and IL-1 beta in the cell lysates and supernatants depended on both the concentration of rIL-2 in the culture medium and the duration of the incubation. After 24 h of stimulation, IL-2-induced IL-1 alpha remained almost entirely cell-associated. In contrast, IL-1 beta was present in both cell lysates and supernatants and was more abundant in the latter. SDS-PAGE analysis after radioimmunoprecipitation with anti-IL-1 antibodies indicates that cell-associated IL-1 resulting from IL-2 stimulation was in the form of the 35 kDa IL-1 precursor whereas secreted IL-1 was almost entirely in the form of the mature 18 kDa product. Depletion of monocytes from the PBMC culture substantially reduced IL-2-induced IL-1 production. In addition, Leu M3+ monocytes obtained through FACS, but not CD16+ NK cells, produced both IL-1 alpha and IL-1 beta in response to IL-2. The low level of endotoxin present in the IL-2 preparation used in our studies and the selective inhibition by polymyxin B of LPS-induced, but not IL-2-induced, IL-1 production by PBMC indicate that IL-2-induced IL-1 production was not due to endotoxin contamination. Furthermore, an anti-IL-2 antiserum selectively inhibited IL-1 production in response to IL-2 stimulation. We conclude that IL-2 is a potent inducer of IL-1 synthesis and secretion in vitro and propose that IL-1 may be generated in vivo in patients undergoing IL-2 immunotherapy.     

T4 cell activation by immobilized phytohemagglutinin: differential capacity to induce IL-2 responsiveness and IL-2 production

Soluble mitogens, such as PHA induce accessory cell (AC)-dependent T cell proliferation. One function of the AC is to create a stimulatory matrix. Therefore, experiments were carried out to determine whether PHA immobilized onto microtiter plates could stimulate T cells in the absence of AC. Peripheral blood T4 cells were cultured under limiting dilution conditions with either soluble or immobilized PHA with or without rIL-1 beta, rIL-2, r-TNF-alpha, an anti-CD28 mAb (9.3), or irradiated EBV-transformed B cells as AC. The frequency of proliferating T4 cells was assessed by examining wells microscopically, and the frequency of T4 cells producing IL-2 was assessed by examining the ability of supernatants to support CTLL-2 proliferation. The percentage of T4 cells growing and producing IL-2 was determined by a maximum likelihood procedure. Immobilized, but not soluble, PHA induced a mean of 20.0 +/- 2.6% of T4 cells to grow in the complete absence of AC in medium supplemented with rIL-2. Whereas rIL-1 beta, rTNF-alpha, and 9.3 were unable to support T4 cell growth in the absence of rIL-2, each enhanced the percentage of T4 cells responding to immobilized PHA in the presence of rIL-2. In contrast, both soluble and immobilized PHA were unable to induce T4 cell IL-2 production in the absence of AC, even when cultures were supplemented with rIL-1 beta or 9.3. In the presence of AC, a small percentage of T4 cells (5.4 to 11.7%) was stimulated to produce detectable amounts of IL-2 by either immobilized or soluble PHA. Moreover, in the presence of AC, a very small population (approximately 1%) of PHA-stimulated T4 cells proliferated without supplemental rIL-2. The data indicate that a matrix of immobilized PHA is sufficient for some T4 cells to be activated to respond to IL-2, whereas others require additional signals provided by rIL-1 beta, rTNF alpha, 9.3, or AC. In contrast, neither soluble nor immobilized PHA is sufficient to induce T cell IL-2 production. This response requires signals provided by intact AC.     

Involvement of the liver, but not of IL-6, in IL-1-induced desensitization to the lethal effects of tumor necrosis factor.

C57BL/cnb mice were found to be protected against a lethal combination of recombinant murine (m) TNF and GalN by pretreatment with several cytokines. At certain doses, rmTNF and human (h) TNF protected completely. The clearest protection was induced by rIL-1: all four rIL-1 species (both m and h, as well as alpha and beta) protected when given 12 h before the challenge. LPS and rmIFN-gamma protected weakly, whereas rmIL-6 and rhIL-6 did not protect at all. Also adrenocorticotropic hormone, dexamethasone, or dexamethasone in combination with rhIL-6 could not protect. A single IL-1 injection also completely protected mice against a lethal dose of mTNF in the absence of GalN sensitization. The desensitization by IL-1 cannot be explained by a faster clearance of the challenge TNF. In addition, we demonstrate that the IL-1-induced desensitization was only observed when a functioning liver was present, that IL-1-pretreated animals did not show decreased numbers of hepatocyte TNF receptors, and that the amount of TNF-induced IL-6 was not reduced.     

Different sensitivity to interleukin 4 of interleukin 2- and interferon alpha-induced CD69 antigen expression in human resting NK cells and CD3+, CD4-, CD8- lymphocytes.

The effect of rIL-4 on CD69 antigen expression induced by rIL-2 or by rINF-alpha on human resting NK cells and CD3+, CD4-, CD8- T lymphocytes has been investigated. rIL-4 drastically inhibited CD69 antigen expression induced by rIL-2 in both cell types. In contrast, rIL-4 did not alter rINF-alpha-induced CD69 antigen expression. Consistent results were obtained evaluating the cytolytic activity of NK cells against the Raji target cell line: rINF-alpha-induced lytic activity was not inhibited by rIL-4, while rIL-2-induced lytic activity was drastically inhibited. Proliferative activity of NK cells induced by rIL-2, in contrast, was only slightly reduced by rIL-4. rIL-4 did not alter the expression of the beta chain of IL-2 receptor, evaluated in NK cells by indirect immunofluorescence. Expression of the alpha chain of IL-2 receptor could not be detected in NK cells by indirect immunofluorescence. It can therefore be suggested that the selective inhibitory effect of rIL-4 on rIL-2-induced activation of NK cells is not mediated by downregulation of alpha and beta chains of IL-2 receptor.     

In vivo administration with IL-1 accelerates the development of collagen-induced arthritis in mice

In an attempt to examine the in vivo proinflammatory properties of IL-1, the effects of rIL-1 beta on the development of collagen-induced arthritis in mice were investigated. The results presented in this paper demonstrated that the administration of rIL-1 beta via mini-osmotic pumps into DBA/1 mice which were suboptimally immunized with native chick type II collagen (NcII) markedly accelerated the onset as well as the progression of the arthritic disease. When IL-1-containing osmotic pumps were s.c. implanted onto mice 18 days post-collagen immunization, clinical signs of arthritis appeared within 3 to 4 days after the implant with the pumps. Maximal incidence of arthritis which was usually 80 to 100% occurred between the 6th and 7th day after the administration of rIL-1 beta. Histologic analyses revealed that the knee and ankle joints from mice which were treated with rIL-1 beta for 7 days were most severely and consistently affected. Furthermore, these IL-1-treated mice exhibited granulocytic hyperplasia within the marrow as well as marked peripheral blood neutrophilia. By contrast, arthritis was not observed during the 7-day course of the IL-1 study in the following control groups: 1) mice that were only immunized with NcII, and 2) collagen-immunized mice which received osmotic pumps containing PBS. A substantial number of these collagen-immunized mice which were not treated with IL-1 eventually developed arthritis but at later times after the incidence of arthritis had peaked in the IL-1-treated group. In addition, unimmunized mice failed to develop arthritis upon treatments with IL-1 beta. Moreover, the humoral responses to NcII were not altered in the IL-1-treated mice. Thus, these in vivo studies suggest that IL-1 is potentially capable of triggering the various inflammatory events of collagen-induced arthritis, and thereby, contribute to the pathogenesis of murine arthritis.     

Matrix metalloproteinase 9 (92-kDa gelatinase/type IV collagenase) is induced in rabbit articular chondrocytes by cotreatment with interleukin 1 beta and a protein kinase C activator.

The synthesis of an 88-kDa gelatinolytic enzyme, identified as a zymogen of matrix metalloproteinase (proMMP)-9, was induced in the primary culture of rabbit articular chondrocytes by cotreatment with recombinant interleukin 1 beta (rIL-1 beta) and the protein kinase C (PKC) agonists, phorbol 12,13-dibutyrate (PDBu) or mezerein. Negligible 88-kDa gelatinolytic activity was produced by unstimulated cells or cells treated with a PKC activator alone at concentrations up to 100 ng/ml, and only a modest induction occurred with rIL-1 beta alone at concentrations of 1-100 ng/ml. However, when these cells were treated with a PKC activator in the presence of IL-1 beta (1 ng/ml), induction was striking, with enzymic activity detectable at a concentration as low as 1 ng/ml of mezerein or 10 ng/ml of PDBu. Rabbit chondrocytes in culture constitutively produced the zymogen of MMP-2 (proMMP-2) and its production was not altered by treatment with IL-1 beta or PKC agonists alone or in combination. Recombinant tumor necrosis factor alpha (rTNF alpha) did not substitute for IL-1 beta in inducing proMMP-9 in the presence of PKC activators, nor was the combination of IL-1 beta or TNF alpha alone effective. These data indicate that rabbit articular chondrocytes have a potential to synthesize and secrete proMMP-9 under certain biological and pathological conditions but that the expression of proMMP-9 is differently regulated from that of other MMPs.     

Human recombinant interleukin-1 beta- and tumor necrosis factor alpha-mediated suppression of heparin-like compounds on cultured porcine aortic endothelial cells

Cytokines are known to tip the balance of the coagulant-anticoagulant molecules on the endothelial cell surface toward intravascular coagulation. Their effects on endothelial cell surface-associated heparin-like compounds have not been examined yet. Incorporation of [35S]sulfate into heparan sulfate on cultured porcine aortic endothelial cells was suppressed by human recombinant interleukin-1 beta (rIL-1 beta) or tumor necrosis factor alpha (rTNF alpha) in a dose- and time-dependent manner with little effect on cell number, protein content, and [3H]leucine incorporation of cells. Maximal inhibition was achieved by incubation of cells with 100 ng/ml of rIL-1 beta or 5 ng/ml of rTNF alpha for 12-24 hours, resulting in a reduction of the synthesis of heparan sulfate on the cell surface by approximately 50%. The dose dependency was consistent with that seen in the stimulation of endothelial cell procoagulant activity by each cytokine. The suppression of heparan sulfate synthesis was sustained for at least 48 hours after pretreatment of cells with cytokines and was unchanged after the addition of indomethacin or polymyxin B. The rate of degradation of prelabeled 35S-heparan sulfate on the cell surface was not altered by cytokine treatments. Neither the size, the net negative charge, nor the proportion of the molecule with high affinity for antithrombin III of endothelial cell heparan sulfate was changed by cytokines. Furthermore, specific binding of 125I-labeled antithrombin III to the endothelial cell surface was reduced to 40-60% of control by cytokines. In parallel with reduction in binding, antithrombin III cofactor (heparin-like) activity was partially diminished in cytokine-treated endothelial cells. Thus, cytokine-mediated suppression of heparin-like substance on endothelial cells appears to be another cytokine-inducible endothelial effects affecting coagulation.     

Trace levels of bacterial lipopolysaccharide prevent interferon-gamma or tumor necrosis factor-alpha from enhancing mouse peritoneal macrophage respiratory burst capacity

Preexposure of resident mouse peritoneal macrophages for 1 hr to traces of bacterial lipopolysaccharide (LPS) (less than or equal to 1 ng/ml) rendered the cells refractory to activation by recombinant interferon-gamma (rIFN gamma) or recombinant tumor necrosis factor-alpha (rTNF alpha), as evaluated by release of H2O2 upon stimulation with phorbol myristate acetate. Inhibition persisted for at least 4 days. Fifty percent inhibition of activation mediated by rIFN gamma followed 1 hr exposure to 10 pg/ml LPS. Fifty percent inhibition of activation mediated by rTNF alpha was achieved with 1 hr exposure to 1 pg/ml LPS. Such low levels LPS exposures (concentration X time) are far below those reported for many other actions of LPS on host cells. Inhibition was partially prevented by the cyclooxygenase inhibitors indomethacin, ibuprofen, and acetylsalicylic acid. Exogenous prostaglandins PGE1 and PGE2, and the 3',5'-cyclic adenosine monophosphate analog dibutyryl cyclic adenosine monophosphate (cAMP), mimicked the inhibitory effect of LPS in a dose-dependent manner, consistent with the hypothesis that formation of endogenous cyclooxygenase products in response to LPS may elevate intracellular cAMP and that the latter may mediate the observed inhibition. In addition, neutralizing antibody against IFN alpha and IFN beta selectively prevented LPS inhibition of activation mediated by rIFN gamma, but not by rTNF alpha. This suggests that IFN alpha and/or IFN beta induced by LPS also contributed to inhibition of activation by rIFN gamma. Thus, release of LPS may afford microorganisms a means by which to interfere with immunologically mediated enhancement of the respiratory burst-dependent antimicrobial capacity of macrophages.     

IL-1 beta and prostaglandins regulate integrin mRNA expression.

The purpose of this study was to examine the effects of IL-1 beta on integrin expression in MG-63 human osteosarcoma cells. Human recombinant IL-1 beta (rIL-1 beta) produced significant increases in both alpha 2- and alpha 5-subunit mRNA levels, as well as a smaller increase in alpha v-subunit mRNA. In contrast, IL-1 beta decreased alpha 4-subunit mRNA levels by approximately 30% relative to untreated controls. These findings suggest that human IL-1 beta differentially regulates expression of integrins. When cultures were treated with both IL-1 beta and the cyclooxygenase inhibitor, indomethacin, the expression of alpha 2-, alpha 5-, and alpha v-subunit mRNA levels were dramatically increased relative to untreated controls; co-treatment with 0.5 mM prostaglandin E2 (PGE2) partially reversed this effect. Indomethacin alone did not affect integrin mRNA levels. Treatment with IL-1 beta or IL-1 beta + indomethacin also induced significant changes in MG-63 morphology (i.e., increased cell elongation) and increased the ability of cells to contract collagen gels. PGE2 reversed the above effects on cell morphology and gel contraction. These findings indicate that (a) IL-1 beta differentially regulates the expression of integrins and (b) that PGE2, which is induced by IL-1 beta, may provide a negative feedback loop which counteracts the stimulatory effect of IL-1 beta on integrin gene expression. It is suggested that products of inflammation may affect cell behavior by differentially regulating the expression of various integrins.     

Stimulation of arachidonic acid metabolism: differences in potencies of recombinant human interleukin-1 alpha and interleukin-1 beta on two cell types

Recombinant human interleukin-l (rIL-1) alpha and beta, which have 26% homology in their amino acid sequence, stimulated arachidonic acid metabolism by squirrel monkey smooth muscle cells and rat liver cells; their relative effectiveness, however, varied with the two cells. Recombinant IL-1 alpha was 3 times more effective than rIL-1 beta at stimulating arachidonic acid metabolism by the primate smooth muscle cells. Recombinant IL-1 alpha was 3 times less effective than rIL-1 beta when measured by their capacity to synergistically stimulate arachidonic acid metabolism of rat liver cells in the presence of palytoxin and anti-diuretic hormone (ADH). The rIL-1 alpha and rIL-1 beta also stimulated the release of radiolabelled arachidonic acid from the smooth muscle cells prelabelled with [3H]arachidonic acid. The two recombinant IL-1s have different heat stabilities, again when measured by their capacity to stimulate arachidonic acid metabolism; IL-1 alpha was more heat stable than IL-1 beta.     

IL-1 induces IL-1. III. Specific inhibition of IL-1 production by IFN-gamma

IL-1 possesses several biologic properties, some of which are associated with chronic inflammatory diseases. We have recently shown that IL-1 induces its own gene expression and, in the present studies, we have examined the effect of IFN-gamma on IL-1-induced IL-1 production. Whereas IFN-gamma increases the total amount of IL-1 (extracellular and cell-associated) produced after endotoxin stimulation of human PBMC, in the same cultures, IL-1-induced IL-1 production was markedly (greater than 70%) reduced in the presence of IFN-gamma. We observed this inhibition in the PBMC from over 40 human donors by employing non-cross-reacting RIA for either IL-1 beta or IL-1 alpha. IFN-gamma inhibited IL-1 beta-induced IL-1 alpha as well as IL-1 alpha-induced IL-1 beta production; furthermore, this inhibitory effect of IFN-gamma was unaffected by indomethacin. The ability of 100 U/ml of IFN-gamma to inhibit IL-1-induced IL-1 production was comparable to that accomplished by 10(-7) M dexamethasone. In contrast to its effect on IL-1 production from PBMC, IFN-gamma had no effect on the proliferative responses of T cells to IL-1. We conclude that IFN-gamma down-regulates synthesis of total IL-1-induced IL-1 production but up-regulates endotoxin-induced IL-1 production. These studies may explain the ameliorating effects of IFN-gamma in experimental models of IL-1-induced bone and cartilage degradation, in peritoneal fibrosis, and in patients with diseases associated with increased IL-1 production.