Monocyte-derived neutrophil chemotactic factor/interleukin-8 is a potential mediator of crystal-induced inflammation

The physical interaction of particulates with resident mononuclear phagocytes is a consistent feature in certain forms of crystal-induced inflammation. In this study, we observed that monosodium urate crystals stimulated the rapid release of neutrophil chemotactic activity from monocytes, and that this activity steadily increased over 24 hours. Because the release of monocyte-derived neutrophil chemotactic activity was markedly diminished by pretreatment of the monocytes with cycloheximide, and was completely removed from conditioned media by adsorption to heparin-agarose, we addressed the possibility that monocyte-derived neutrophil chemotactic factor/interleukin-8 (IL-8), a heparin-binding neutrophil-activating polypeptide, might modulate these activities. Urate crystal-induced IL-8 secretion from monocytes was verified by radioimmunoassay. In addition, an IL-8-specific antibody markedly inhibited the neutrophil-activating capacity of the conditioned media from monocytes activated by urate crystals, as well as by inflammatory silica crystals. Last, IL-8 was significantly increased in gouty synovial fluids (range 3.0-16.8 ng/ml, mean 8.4 ng/ml, n = 6) relative to osteoarthritic synovial fluids (range 1.1-1.7 ng/ml, mean 1.5 ng/ml, n = 6) (P = 0.006). We conclude that microcrystal-induced secretion of IL-8 by mononuclear phagocytes may mediate a number of forms of crystal-induced inflammation.     

Regulation of interleukin-12 expression in human monocytes: selective priming by interferon-gamma of lipopolysaccharide-inducible p35 and p40 genes

Interleukin-12 (IL-12) is a monocyte/macrophage-derived cytokine that is critical for T lymphocyte and natural killer cell activities and functions. In this study, we examined the regulation of IL-12 expression by human monocytes in response to bacterial lipopolysaccharide (LPS). Several novel aspects of IL-12 induction from monocytes were shown. Optimal expression of IL-12 mRNA and bioactivity required specific priming of monocytes by interferon-gamma (IFN-gamma) before LPS stimulation. Granulocyte-macrophage colony-stimulating factor (GM-CSF) provided an equivalent priming stimulus for LPS-induced tumor necrosis factor (TNF) and IL-12 p40 mRNA, but primed poorly for LPS-inducible p35 message and secreted IL-12 activity. Macrophage colony-stimulating factor (M-CSF), although a potent survival factor for monocytes, showed no priming activity for IL-12 production. Time course experiments demonstrated independent regulation of p40 and p35 by IFN-gamma and LPS. LPS inducibility of p40 expression required only a brief exposure to IFN-gamma (2 hours), while prolonged exposure (+/- 24 hours) to IFN-gamma resulted in diminishing levels of p40 mRNA. p35 inducibility (by LPS) required a longer exposure to IFN-gamma (8 to 16 hours), and continued to be inducible up to 40 hours following IFN- gamma priming. Both mRNAs were rapidly induced (1 to 2 hours) in IFN- gamma-primed monocytes; p35 message reached a plateau by 2 hours, while p40 continued to accumulate. Finally, both p40 and p35 were directly induced by LPS in the presence of cycloheximide. These results indicated that both p40 and p35 are LPS-inducible in monocytes following IFN-gamma pretreatment, and that the regulated expression of p35 controls the level of active IL-12 protein in purified human monocytes. The selectivity of priming by IFN-gamma is in accord with a putative role for IL-12 in the initiation and amplification of TH1-type responses.     

Th1 and Th2 T-helper cells exert opposite regulatory effects on procoagulant activity and tissue factor production by human monocytes

The role of T-cell subsets in the induction of tissue factor (TF) production by human monocytes in vitro was investigated. Mitogen stimulation enabled both unfractionated T cells and their CD4+ or CD8+ subsets to promote procoagulant activity (PCA). After mitogen or antigen activation, all seven T-cell clones with Th1 cytokine profile, but none of seven Th2 clones, induced TF production and PCA. T-cell blasts from four Th1 activated clones were fixed with paraformaldehyde and added to monocytes in the presence of medium alone or their supernatants. Addition of either fixed Th1 cells or their supernatants induced low TF production (0.2 to 0.6 ng/mL), whereas addition of both resulted in much higher TF synthesis (1.8 to 3.4 ng/mL). Among Th1-type cytokines, only interferon-gamma (IFN-gamma) induced minimal TF production (0.1 to 0.4 ng/mL). No TF synthesis was induced by activated and fixed Th2 cells and/or their supernatants, whereas combined addition of fixed Th2 cells and Th1 supernatants or IFN-gamma induced noticeable TF production. The addition of either anti-IFN-gamma antibody or Th2 supernatants to monocytes stimulated with activated and fixed Th1 cells plus their supernatant resulted in a dose-dependent inhibition of TF synthesis, which was partially restored by neutralization of interleukin-4 (IL-4) or IL-10. Addition of recombinant IL-4, IL-13, or IL-10, but not IL-5, inhibited the Th1- induced TF production by monocytes. Data indicate that both CD8+ and CD4+ Th1, but not Th2, T cells can help TF production and PCA. Both cell-to-cell contact with activated T cells and Th1-type cytokines, in particular IFN-gamma, are required for optimal TF synthesis, whereas Th2-derived cytokines (IL-4, IL-13, and IL-10) are inhibitory. This may be of potential interest for future therapeutic strategies.     

Effect of granulocyte-macrophage colony-stimulating factor and interleukin-3 on interleukin-8 production by human neutrophils and monocytes

Interleukin-8 (IL-8) is a major neutrophil chemoattractant and functional stimulant that is induced by IL-1, tumor necrosis factor alpha (TNF alpha), and lipopolysaccharide (LPS). We report that recombinant human (rh) granulocyte-macrophage colony-stimulating factor (GM-CSF) and rhIL-3 are also potent inducers of IL-8 messenger RNA (mRNA) accumulation and protein secretion by normal peripheral blood monocytes. Neutrophils produce IL-8 in response to GM-CSF but not to IL- 3. In contrast, recombinant human granulocyte-CSF (rhG-CSF), at concentrations as high as 100 ng/mL, does not induce IL-8 in either cell type. rhGM-CSF also induces IL-8 mRNA expression and IL-8 protein in the promonocytic cell line, U-937, whereas rhG-CSF does not. IL-8 secretion by monocytes was stimulated within 2 hours after incubation with rhGM-CSF or rhIL-3. Stimulation of neutrophils with rhGM-CSF resulted in an increase in cell-associated IL-8 at 4 hours. At 24 hours, cell-associated IL-8 levels declined, whereas secreted IL-8 levels increased. In contrast, virtually all IL-8 induced in monocytes appeared as secreted protein. Neither rhGM-CSF nor rhIL-3 induced detectable secretion of IL-1, TNF alpha, or IL-6 protein by monocytes. rhGM-CSF, and to a lesser degree rhIL-3, potently stimulated IL-8 secretion in cultures of heparinized whole blood, whereas rhG-CSF had no significant effect on IL-8 secretion. Induction of IL-8 by GM-CSF may be physiologically important in enhancing the acute inflammatory response.     

Interferon-gamma production by human cord blood monocyte-derived dendritic cells

Interferon (IFN)-gamma is produced by T cells and natural killer cells and activates monocytes and dendritic cells (DCs). Recently, IFN-gamma has been shown to be produced by mouse DCs following stimulation with interleukin (IL)-12, which is markedly augmented by the addition of IL-18. We here analyzed whether human DCs secrete IFN-gamma in response to IL-12 and/or IL-18. Human immature DCs, generated from cord blood CD14(+) monocytes by treating with granulocyte-macrophage colony-stimulating factor and IL-4, were incubated with IL-12 and/or IL-18 and assayed for IFN-gamma production. IL-12, but not IL-18, weakly induced IFN-gamma production, while IL-12 together with IL-18 induced high levels of IFN-gamma production. Similar results were obtained with mature DCs, although levels of IFN-gamma production were less than those in immature DCs. Also with mature and immature DCs, IL-12 upregulated the expression of IL-18 receptor alpha (Ralpha), and costimulation with IL-12 and IL-18 upregulated CD40 expression. Anti-IL-18Ralpha antibody abrogated both the IFN-gamma induction and the CD40 upregulation by IL-12 plus IL-18. These findings suggest that IL-12 upregulates IL-18Ralpha expression on human DCs and acts synergistically with IL-18 to induce high levels of IFN-gamma, which subsequently enhances CD40 expression on DCs in an autocrine manner.     

Tumor-associated leukemia inhibitory factor and IL-6 skew monocyte differentiation into tumor-associated macrophage-like cells

Tumor-associated macrophages (TAMs), the most abundant immunosuppressive cells in the tumor microenvironment, originate from blood monocytes and exhibit an IL-10(high)IL-12(low) M2 profile. The factors involved in TAM generation remain unidentified. We identify here leukemia inhibitory factor (LIF) and IL-6 as tumor microenvironmental factors that can promote TAM generation. Ovarian cancer ascites switched monocyte differentiation into TAM-like cells that exhibit most ovarian TAM functional and phenotypic characteristics. Ovarian cancer ascites contained high concentrations of LIF and IL-6. Recombinant LIF and IL-6 skew monocyte differentiation into TAM-like cells by enabling monocytes to consume monocyte-colony-stimulating factor (M-CSF). Depletion of LIF, IL-6, and M-CSF in ovarian cancer ascites suppressed TAM-like cell induction. We extended these observations to different tumor-cell line supernatants. In addition to revealing a new tumor-escape mechanism associated with TAM generation via LIF and IL-6, these findings offer novel therapeutic perspectives to subvert TAM-induced immunosuppression and hence improve T-cell-based antitumor immunotherapy efficacy.     

Human cytomegalovirus increases constitutive production of interleukin- 6 and leukemia inhibitory factor by bone marrow stromal cells

Human cytomegalovirus (CMV) infection is often associated with myelosuppression and acute inflammatory reaction in immunocompromised patients. We have previously documented that CMV exposure of bone marrow (BM) stromal cells reduces the capacity of these cells to support hematopoiesis because of a decreased production of colony- stimulating factors. This study examines the potential role of CMV on constitutive and lipopolysaccharide (LPS)-stimulated production of cytokines involved in inflammatory reaction, interleukin-6 (IL-6) and leukemia inhibitory factor (LIF) by BM stromal cells. The release of IL- 6 was already detectable 2 hours post CMV-infection (2.5-fold increase in production) and the cumulative production of IL-6 after 5 days of infection was 23 +/- 1.2 ng/mL (ninefold increase in production). CMV was also able to induce a time-dependent production of LIF that was maximal 8 hours after CMV infection (2.5-fold increase in production). Concomitantly, there was no detectable release of granulocyte colony- stimulating factor (G-CSF) and granulocyte-macrophage CSF (GM-CSF) by CMV-infected stromal cells. The similar IL-6 and LIF production in the presence of polymyxin B ruled out the possibility that this increase could be caused by contamination of the viral stock by endotoxin. In addition, ultraviolet-inactivated virus behaved similarly to live virus and caused the release of IL-6 and LIF. However, heat-inactivated CMV was unable to induce IL-6 and LIF secretion by BM stromal cells. The production of IL-6 and LIF was also evaluated after stimulation by LPS. After 5 days of CMV exposure, the LPS-stimulated production of IL-6 and LIF was significantly lower than uninfected controls. This LPS-induced release of cytokine production was found to be dependent of viral replication. The experiments have shown that CMV is a potent inducer of IL-6 and LIF with differential effect on constitutive and LPS- stimulated cytokine production by stromal cells; we suggest that CMV induction of IL-6 and LIF during the first hours of infection could play a role in CMV-induced inflammatory reaction. Moreover, our results show that human CMV can disturb the balanced cytokine network involved in the regulation of hematopoiesis.     

Leprosy patients with lepromatous disease have an up-regulated IL-8 response that is unlinked to TNF-alpha responses

Tumor necrosis factor (TNF-alpha) in conjunction with interferon-gamma (IFN-gamma) plays an important role in lymphocyte recruitment and granuloma formation in mycobacterial diseases. Lepromatous leprosy infections are typically associated with low to absent T cell responses and the absence of INF-gamma secretion. Chemokines such as IL-8, MCP-1, and MIP-1beta, have also been shown to recruit neutrophils and lymphocytes to the site of mycobacterial infections. We have studied IL-8 expression in relation to TNF-alpha and TGF-beta in monocytes from lepromatous patients (LL) as compared with healthy endemic controls. In endemic controls, no spontaneous expression of IL-8, TNF-alpha, and TGF-beta was observed, but BCG and M. leprae induced activation of all three cytokines. Lepromatous leprosy monocytes spontaneously expressed high levels of IL-8 and TGF-beta but negligible levels of TNF-alpha. A further increase in IL-8 secretion or gene expression by BCG or M. leprae was not significant. BCG, but not M. leprae, was able to stimulate TNF-alpha activation in lepromatous leprosy subjects. TGF-beta responses in LL were parallel to those of IL-8. This suggests a vigorous and active ongoing IL-8 response in lepromatous disease that is independent of TNF-alpha activation. Therefore, in the absence of IFN-gamma and TNF-alpha activation, IL-8 may assume a pivotal role in cell recruitment in leprosy patients with disseminated mycobacterial infections.     

Granulocyte-macrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF) secretion by adherent monocytes measured by quantitative immunoassays.

The kinetics of GM-CSF and G-CSF secretion by purified adherent human monocytes were studied by quantitative immunoassays. Interleukin-1 (IL-1); 4-40 ng/ml and E. coli lipopolysaccharide (LPS); 0.1-1.00 ng/ml, were the most effective stimuli and induced dose-dependent secretion of both GM-CSF and G-CSF. Secretion of newly synthesized CSF was detectable 3-6 hours after stimulation and continued for approximately 24 h. Twenty minutes pulse exposure to LPS was sufficient to induce half maximum secretion of GM-CSF, and after 24-36 h the adherent monocytes could not be restimulated. Neither GM-CSF nor TNF could down-regulate the secretion of GM-CSF. IL-3 induced a minor secretion of GM-CSF whereas TNF, G-CSF, M-CSF and IFN-gamma were unable to induce GM-CSF secretion. In addition to LPS and IL-1, GM-CSF and to a minor degree TNF induced G-CSF secretion. Enriched T lymphocytes secreted GM-CSF, but not G-CSF, after stimulation with PHA or staphylococcal enterotoxin A (SEA), whereas LPS and IL-1 were without stimulatory effects. We also noted that enriched T lymphocytes added to LPS-stimulated adherent monocytes at ratios of 1:10 or more inhibited, in a dose-dependent fashion, GM-CSF secretion by 13-55%. These findings add new quantitative data on CSF secretion by human monocytes.     

Oncostatin M induces procoagulant activity in human vascular smooth muscle cells by modulating the balance between tissue factor and tissue factor pathway inhibitor.

Oncostatin M (OSM) is a cytokine of the interleukin-6 (IL-6) family secreted by activated monocytes, and is expressed in atherosclerotic plaque. Smooth muscle cells (SMC), by expressing tissue factor (TF) and tissue factor pathway inhibitor (TFPI) can contribute to the thrombogenicity of atherosclerotic plaque. Consequently, the aim of this study was to evaluate the effects of OSM on the procoagulant activity of SMC. We observed that OSM induced in a concentration-dependent manner a potent procoagulant activity (PCA) that was related in part to an increased synthesis of TF, both at the cell membrane and in SMC lysates. The increased expression of TF on SMC membrane induced by OSM was sustained and was still observed 24 h after stimulation by OSM. IL-6 and leukaemia inhibitory factor (LIF), two OSM-related cytokines, did not significantly modify TF expression at the surface of SMC. In addition to its effects on TF, OSM decreased the secretion of TFPI in the supernatants of SMC, as well as in the lysates, but was devoid of effect on TFPI bound at the membrane of SMC. IL-6 and LIF reduced also TFPI secretion, which could explain why the PCA of SMC lysates treated by IL-6 or LIF was increased, despite an absence of effect on TF expression. In conclusion, these data support the hypothesis that by increasing the PCA of SMC, OSM might be involved in the thrombotic complications associated with plaque rupture.     

白细胞介素-18对慢性乙型肝炎患者外周血单个核细胞的作用及意义

目的 探讨白细胞介素-18(IL-18)对慢性乙型肝炎患者外周血单个核细胞(PBMC)的作用及体外对转染乙型肝炎病毒(HBV)基因的人肝癌细胞系2.2.15(HepG2.2.15)细胞HBV DNA分泌的影响。方法 分离25例正常人及25例慢性乙型肝炎患者的PBMC,在乙型肝炎核心抗原(HBcAg)及不同浓度IL-18刺激下培养72h,收集其培养上清液,采用酶联免疫试验办法检测其中干扰素γ(IFN-γ)的含量;收集1例慢性乙型肝炎患者的PBMC,与已培养了24 h的HepG2.2.15细胞共孵育,经过不同浓度的IL-18刺激培养96h后,收集其培养上清液,采用聚合酶链反应(PCR)检测HBV DNA的含量。结果 PBMC在HBcAg联合0.2、1.0、5.0ng/ml的IL—18刺激下,慢性肝炎组培养上清液中IFN-γ的水平均较正常对照组明显增高(0.2ng/ml组:t 11.7,P<0.01;1.0ng/ml组:t-16.19,P<0.01;5.0ng/ml组:t-20.12,P<0.01);在HBcAg联合IL-18和IL-12刺激下,慢性肝炎IFN-γ的水平亦较正常对照组明显增高,分别为(1313.20±187.76)pg/ml和(390.75±43.23)pg/ml,t—23.94,P<0.01;在慢性肝炎轻、中、重度患者中,单独HBcAg刺激组与HBcAg联合不同浓度的IL-18刺激组相比,联合刺激组的IFN-γ的水平均较单独刺激组增高;HBcAg联合IL-18刺激组与联合IL-12和同一浓度的IL-18刺激组相比,IFN-γ的水平后者显著高     

Chemotactic activity of recombinant human granulocyte colony- stimulating factor

Recombinant human granulocyte colony-stimulating factor (rhG-CSF) induced migration across polycarbonate filters of human polymorphonuclear leukocytes (PMN). rhG-CSF was active in inducing PMN migration at concentrations greater than or equal to 10 to 100 U/mL (7 to 70 ng/mL). rhG-CSF did not contain appreciable levels of endotoxin contamination as assessed by Limulus amebocyte assay, and Polymixin B did not affect the chemotactic activity of rhG-CSF. A monoclonal anti-G- CSF antibody blocked the induction of migration by G-CSF, thus establishing that the cytokine was responsible for the activity of the recombinant preparation. Checkerboard analysis was performed by seeding different concentrations of G-CSF above and/or below the filter and revealed that the migratory response to this cytokine was best observed in the presence of a positive concentration gradient between the lower and upper compartments of the chamber, thus indicating an actual chemotactic effect. When different migrating cells were examined, rhG- CSF was inactive on large granular lymphocytes and endothelial cells under conditions in which appropriate reference attractants were active. In contrast, rhG-CSF elicited a chemotactic response in monocytes inhibited by specific antibody. Thus, G-CSF is a chemotactic signal for phagocytes. This cytokine, when produced at inflammatory sites, may contribute to the recruitment of phagocytes from the blood compartment to amplify resistance against certain noxious agents.     

C5a- and tumor necrosis factor-alpha-induced leukocytosis occurs independently of beta 2 integrins and L-selectin: differential effects on neutrophil adhesion molecule expression in vivo

We previously demonstrated in rabbits that various neutrophil chemotactic factors share an ability to induce recruitment of polymorphonuclear leukocytes (PMN) from bone marrow when administered intravenously (Jagels and Hugli, J Immunol 148:1119, 1992). In the study reported here, we investigated the effects of chemotactic factors on the expression of beta 2 integrins and L-selectin in vivo and the roles of these adhesionmolecules in the recruitment process. Leukocytosis was induced by infusion of either C5a (5 micrograms/kg), N- formyl-methionyl-leucyl-phenylalanine (f-MLP; 2.5 micrograms/kg), or tumor necrosis factor alpha (TNF-alpha; 100 ng/kg). C5a increased the expression of CD18 (the common subunit of beta 2 integrins) on PMN by nearly twofold and decreased levels of L-selectin by 50% within 15 minutes after administration. Levels of beta 2 integrins returned to baseline 2 to 3 hours after induction of leukocytosis. L-selectin remained depressed for more than 5 hours, demonstrating that shedding was induced in the recruited bone marrow leukocytes as well as in circulating PMN. In contrast to the response to C5a, TNF-alpha did not cause upregulation of CD18 or shedding of L-selectin. Levels of L- selectin were consistently increased 60 minutes after administration of TNF-alpha, coinciding with a rapid rise in the number of band-form PMN in the circulation. Intact IgG and F(ab)2 forms of the anti-CD18 monoclonal antibody IB4 or the anti-L-selectin antibody DREG-200 were administered intravenously 15 minutes before induction of leukocytosis by the chemotactic factors. Neither IB4 nor its F(ab)2 fragments blocked leukocytosis induced by C5a, f-MLP, or TNF-alpha. DREG-200 also did not block leukocytosis induced by f-MLP, C5a, or TNF-alpha. These results suggest that leukocyte emigration from the bone marrow into the circulation proceeds through interactions distinct from those involved in neutrophil chemotaxis and diapedesis. Shedding of L-selectin from C5a-recruited bone marrow leukocytes demonstrates activation of these cells in the recruitment process and may reflect a potential mechanism for their release. The dissimilar effects of C5a and TNF-alpha on expression of adhesion molecules may result from distinct stimulatory pathways and suggests differential activation states for cellular recruitment by these inflammatory factors.     

Interleukin-8 is a major neutrophil chemotactic factor in pleural liquid of patients with empyema.

Interleukin-8 (IL-8), a potent neutrophil chemotactic peptide, has been found in association with human disease, but its contribution to chemotactic activity in humans is not yet known. We asked whether IL-8 is present in inflammatory human pleural effusions, and to what extent it contributes to pleural liquid neutrophil chemotactic activity. Because tumor necrosis factor alpha (TNF-alpha) is a strong inducer of IL-8, we also asked whether TNF-alpha was present. For this prospective study, we collected pleural liquid from 51 patients (empyema, 14; parapneumonic, four; tuberculous, eight; malignant, nine; miscellaneous exudative, seven; and transudative, nine), counted pleural neutrophils, and measured IL-8 and TNF-alpha concentrations in the supernatant. To determine the contribution of IL-8 to chemotactic activity in empyema, we measured the neutrophil migration induced by empyemic liquids before and after addition of anti-IL-8 F(ab')2 antibody fragments or control anti-IL-6 F(ab')2. We found that IL-8 concentrations were higher in empyema (61.3 +/- 21.0 ng/ml [SEM]) than in all other effusions (1.1 +/- 0.5 ng/ml) (p = 0.0001). All empyema liquids had IL-8 concentrations above 2.5 ng/ml, which was true for only three of the other 37 effusions (two parapneumonic, one tuberculous). IL-8 levels correlated with the pleural neutrophil count (r = 0.46; p = 0.007) and the neutrophil chemotactic activity of pleural liquid (r = 0.43; p = 0.008). Anti-IL-8 antibodies decreased chemotactic activity in empyema liquids by 65 +/- 5%, whereas the control antibody had no effect (0 +/- 5% decrease) (p = 0.0005).(ABSTRACT TRUNCATED AT 250 WORDS)     

In vivo and in vitro regulation of thyroid leukemia inhibitory factor (LIF): marker of hypothyroidism.

Several cytokines regulate thyroid function and may be involved in the pathogenesis of thyroid disorders, including euthyroid sick syndrome. Leukemia inhibitory factor (LIF), a neuroimmune pleiotropic cytokine, was measured to assess its role in hypothalamic-pituitary-thyroid function. Mean circulating serum LIF levels in 10 hypothyroid patients [TSH, 23+/-0.5 mIU/L (mean+/-SEM); free T4, 0.77+/-0.1 ng/dL] was 0.29+/-0.04 ng/mL, 145% higher (P < 0.04) than in 20 normal subjects (LIF, 0.20+/-0.02 ng/mL; TSH, 2.23+/-0.21 mIU/L; free T4, 1.23+/-0.04 ng/dL) but was not different from those in 10 hyperthyroid patients (LIF, 0.21+/-0.03 ng/mL; TSH, 0.01+/-0.00 mIU/L; free T4, 3.63+/-0.51 ng/dL). Serum LIF concentrations linearly correlated with serum TSH in the 40 samples (r = 0.58, P < 0.001). When T4 (1-8 microg/kg x day) was administered to cynomolgus monkeys with methimazole-induced hypothyroidism, serum T4 and T3 levels increased appropriately, and TSH and LIF concentrations decreased. When methimazole was given alone, both serum TSH (146+/-30 mIU/L) and LIF (8.84+/-0.49 ng/mL) were markedly induced. When methimazole together with T4 (>2 microg/kg x day) was administered, both serum TSH (7.5+/-1.2 mIU/L) and LIF (6.22+/-0.31 ng/mL) were lowered (P < 0.01). Monkey serum LIF levels and log TSH levels also correlated (r = 0.72, P < 0.01). Cultured thyroid carcinoma cells produced LIF (9.2 ng/10(6) cells/48 h). TSH (100 mIU/mL) and interleukin (IL)-6 (10 nmol/L) stimulated in vitro LIF secretion from the cells by 170+/-12% (P < 0.05) and 261+/-8% (P < 0.05), respectively. Dexamethasone (1 micromol/L) inhibited basal LIF concentration by 83% (P < 0.05), whereas TSH and IL-6 stimulated LIF by 52% (P = 0.04) and 42% (P = 0.03), respectively. However, using Northern blot analysis, we could not observe induction of LIF mRNA by TSH, suggesting that LIF regulation by TSH may be due to stimulation of secretion. The results show that the thyroid gland is a source of LIF production; TSH, IL-6, and glucocorticoid influence thyroid cell LIF expression. The correlation between TSH and LIF suggests that LIF may participate in the physiologic regulation of hypothalamic-pituitary-thyroid function.