INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

MACROPHAGE MIGRATION INHIBITORY FACTOR (MIF) was originally described as a T cell-derived lymphokine that prevented the random migration of macrophages (7). More recent data indicate that MIF has a much broader role within the immune response. For example, MIF increased inducible nitric oxide synthase expression and decreased footpad lesions after Leishmania major challenge in mice (19). Furthermore, administration of anti-MIF antibodies to mice attenuated the frequency of rheumatoid arthritis development in mice and decreased the secretion of collagen II antibodies (12). MIF is secreted by cells other than T lymphocytes, further supporting its broader role in the inflammatory response. Constitutive expression of MIF has been detected in many tissues, including eye (18), kidney (11), skin (15), and gonads (17).

MIF also appears to be a neuroendocrine modulator of systemic inflammation. Administration of recombinant MIF to mice potentiated endotoxemia, whereas anti-MIF antibodies prevented the development of endotoxic shock (4). Furthermore, mice lacking the MIF gene were resistant to the lethal effects of high doses of lipopolysaccharide (LPS) and had lower plasma tumor necrosis factor- (TNF-) levels than wild-type mice (5).

A feedback mechanism exists between the hypothalamic-pituitary-adrenal (HPA) axis and mononuclear cells, such that glucocorticoids (cortisol) inhibit inflammatory cytokine secretion during inflammation, thereby limiting the inflammatory response. Constitutive preformed MIF is stored by corticotropic cells of the anterior pituitary and is released during stress (6) or in response to LPS stimulation (4). Thus MIF may also be involved in the regulation of the inflammatory response by the HPA axis. Interestingly, Calandra et al. (6) demonstrated that MIF could counteract glucocorticoid-induced inhibition of inflammatory cytokine secretion [interleukin (IL)-1, IL-6, IL-8, and TNF-].

The mechanism by which MIF counteracts glucocorticoid-induced inhibition of inflammation has not been fully elucidated. However, there are several possible pathways where MIF and glucocorticoids may interact. One such pathway involves the activation of the transcriptional factor, NF-B. Many inflammatory mediators are regulated by NF-B, including inflammatory cytokines (IL-1 and TNF-), adhesion molecules, immunoreceptors, and acute-phase proteins (3). NF-B is a ubiquitous transcriptional factor in immune cells and is activated by inflammatory stimuli (cytokines, LPS, and viruses). The classical NF-B is a heterodimeric protein that consists of p50 and p65 subunits, which reside in the cytosol. Upon activation, NF-B is released from an inhibitory protein (IB), translocates to the nucleus, and activates transcription (2, 16). Glucocorticoids prevent NF-B activation in part by increasing the expression of IB (1, 13), which keeps NF-B bound in the cytosol and thus prevents gene expression of inflammatory mediators (Fig. 1).

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Key aspects of intracellular signaling via the NF-B pathway. Binding of various inflammatory mediators to plasma membrane receptors triggers a cascade of phosphorylation events, ultimately causing dissociation and breakdown of IB. NF-B is then free to enter the nucleus and activate transcription of several inflammatory cytokines. Glucocorticoids (GC) enter the cell and bind to cytosolic receptors; the complex moves into the nucleus and increases expression of IB.

The purpose of this study was to examine the antagonism between glucocorticoids and MIF at the intracellular level in human mononuclear cells. Specifically, the hypothesis that MIF counteracts the influence of hydrocortisone on cytosolic IB concentrations and NF-B activation was tested. In addition, experiments were performed to determine whether MIF enhanced LPS-induced breakdown of IkB [experiment 2 (Exp-2) in Fig. 1], inhibited hydrocortisone-induced synthesis of IkB [experiment 3 (Exp-3)], or bound directly to hydrocortisone, interfering with cellular binding [experiment 4 (Exp-4)].

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Subjects. All subjects were healthy, were not taking any medication, and gave informed consent to participate in this study. The first series of experiments, examining MIF and hydrocortisone interactions in LPS-stimulated mononuclear cells, included seven subjects (27.6 ± 3.4 yr of age). Later studies examining pairwise combinations of LPS, MIF, and hydrocortisone involved mononuclear cells from six subjects (33 ± 10.3 yr of age). All procedures were approved by the Pennsylvania State University Human Investigation Review Committee.

Blood samples, cell isolation, and culture. A venous blood sample was drawn from each subject between 8:00 and 9:00 AM via antecubital vein into a heparinzed syringe. All subjects had fasted for ~12 h before having their blood drawn. Peripheral blood mononuclear cells (PBMC) were isolated by density centrifugation with Ficoll-Hypaque (Histopaque; Sigma, St. Louis, MO). The mononuclear cell layer was aspirated and washed three times with nonpyrogenic 0.9% NaCl. The cells were resuspended in phenol red-free RPMI supplemented with 100 U/ml penicillin, 100 µg/ml streptomycin, 100 mM Hepes, 0.2 mM L-glutamine (all from Sigma), and 2% TCH serum replacement (ICN Pharmaceuticals, Costa Mesa, CA). The cells were cultured at a density of 5 × 10⁶ cells/ml and were preincubated with sodium hydrocortisone (Solu-Cortex, Upjohn, Kalamazoo, MI) in concentrations of 0, 50, 100, and 200 ng/ml with 0, 0.1, or 1 ng/ml of recombinant human MIF (R & D Systems, Minneapolis, MN) for 1 h. The cells were then stimulated with 1 µg/ml LPS (Escherichia coli 055:B5, Sigma) or remained unstimulated. For analysis of the NF-B/IB signal transduction process, cell cultures were incubated for 30 min at 37°C in a humidified 5% CO₂ chamber, followed by extraction of the nuclear and cytoplasmic fractions as described in the next section. All containers used for cell culture were disposable and endotoxin-free: all solutions were injectable grade or endotoxin tested.

Cytoplasmic and nuclear cell extraction. Cytoplasmic and nuclear extracts were prepared using a modified method by Dignam et al. (9). After incubation with LPS, the cells were transferred to sterile 1.5-ml microcentrifuge tubes and placed on ice. Cells were centrifuged for 8 min at 750 g and washed with 1 ml of sterile ice-cold PBS. The packed cell pellet was resuspended in 100 µl of solution A [10 mM HEPES, 1.5 mM MgCl₂, 10 mM KCl, and 0.5 mM dithiothreitol (DTT)] and a protease inhibitor cocktail [100 mM 4-(2-aminoethyl)benzenesulfonyl fluoride, 1 µg/ml pepstatin A, 3 µg/ml trans-epoxysuccinyl-L-leucylamido-(4-guanidino)butane, 4 µg/ml bestatin, 10 µg/ml leupeptin, and 3 µg/ml aprotinin, all from Sigma] and was placed on ice for 10 min to allow the cells to swell. Nonidet-40 (Boehringer Mannheim, Indianapolis, IN) was added to all cells at a final concentration of 0.6%, and the cells were gently agitated to disrupt the cell membrane. The nuclei were pelleted by centrifugation for 3 min at 500 g. The supernatant containing the cytosolic extract was transferred to a new microcentrifuge tube and centrifuged for 10 min at 18,000 g. The supernatant was collected, assayed for total protein content by the Bio-Rad D_c protein assay (Bio-Rad, Hercules, CA), immediately frozen in liquid nitrogen, and stored at 70°C. The nuclear pellet was washed with 500 µl of solution A and transferred to a 0.5-ml microcentrifuge tube and centrifuged; packed nuclei were resuspended in 20 µl of solution B (20 mM HEPES, 20% glycerol, 1.5 mM MgCl₂, 0.2 mM EDTA, 0.5 mM DTT, and protease inhibitor cocktail) and placed on a rocker at 4°C for 1 h to extract the nuclear proteins. The samples were centrifuged for 10 min at 18,000 g, and the supernatant was collected and assayed for total protein content by the Bio-Rad protein assay, immediately frozen in liquid nitrogen, and stored at 70°C.

Electrophoresis and immunoblot. For electrophoresis, 10 µg of total protein from individual samples were loaded into the lanes of a 10% SDS denaturing polyacrylamide gel and elecrophoretically separated. Proteins were then electrophoretically transferred onto Immobilin-P membranes (Millipore). The membranes were blocked with Tris-buffered saline (TBS), pH 6.7, containing 0.5% Tween 20 and 5% nonfat dry milk (BLOTTO) for a minimum of 2 h. For IB detection, the membranes were incubated overnight at 4°C with rabbit polyclonal anti-human IB (epitope corresponding to the amino-terminal domain of IB; Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1:800 in BLOTTO. After the membranes were washed twice for 7 min each in TBS containing 0.5% Tween 20 (TBST), they were incubated for 30 min in secondary horseradish peroxidase-labeled anti-rabbit IgG (Santa Cruz Biotechnology) diluted 1:7,500 in BLOTTO, and then washed three times for 5 min in TBST followed by one wash in TBS. The proteins were detected by treating the membranes with enhanced chemiluminescence assay reagents (Amersham Life Science, Arlington Heights, IL) and exposing them to film for 90 s. IB levels were determined by densitometry using the ImageQuant analysis program (Molecular Dynamics, Sunnyvale, CA).

Electrophorectic mobility shift assay. To determine DNA-protein interactions, 2 µg of each nuclear extract were added to reaction buffer (25 mM HEPES, pH 7.6, 100 mM KCl, 0.1 mM EDTA, and 10% glycerol) containing 0.5 mg poly (dI-dC), and 1 mM DTT, and randomly primed ³²P-labeled NF-B oligonucleotide probe (AAAGAAATTCCAAAGAGT) containing ~2 × 10⁵ counts/min (cpm). The NF-B probe was a generous gift from Dr. Shao-Cong Sun, Department of Microbiology and Immunology, Pennsylvania State University, Hershey Medical Center. After incubation for 10 min at room temperature, 1.5 µl of loading buffer (0.25% bromophenol blue in 10 mM Tris-1 mM EDTA buffer, pH 8.0) were added to all samples to stop the reaction. Samples were loaded onto a 5% native acrylamide gel and prerun for 1 h. The samples were electrophoresed for 3 h, and the gel was dried for 1 h at 80°C, followed by exposure to X-ray film for periods ranging from 4 h to overnight. NF-B levels were determined by densitometry by use of the ImageQuant analysis program (Molecular Dynamics). For supershifting assays, 1 µl of anti-p50 or anti-p65 antibodies (Santa Cruz Biotechnology) was added before addition of loading buffer, and the mixture was incubated for 30 min followed by addition of 1.5 ml of loading buffer and electrophoresis, as described above. Additionally, for competition assays, a 200 molar excess of unlabeled probe was added to the reaction mixture.

[³H]hydrocortisone binding. MIF (100 ng/ml) was incubated with 4.5 ng/ml [³H]hydrocortisone (1 µCi) in PBS for 2 h at room temperature. After incubation, 20 µl of the mixture were applied to a PD-10 column (Pharmacia, Piscataway, NJ), and 0.5-ml fractions were collected. The fractions were analyzed by liquid scintillation on a Beckman LS6500 counter (Fullerton, CA). In a second run, a cytochrome c standard (mol wt 12,400) was applied to the column, and its elution volume was determined spectrophotometrically at A₂₈₀. The molecular mass of MIF is 12,000 Da, and that of [³H]hydrocortisone is 375 Da.

Statistical analysis. Densitometric values for NF-B activation from the electrophorectic mobility shift assay (EMSA) and IB levels from the Western immunoblot were determined for all subjects and normalized to the appropriate control condition for each experiment. Means were calculated for the normalized values, and the data are expressed as means ± SE. The data examining the interaction of hydrocortisone and MIF in LPS-stimulated cells were analyzed by a two-factor ANOVA, with Dunnett's post hoc analysis to determine significant differences from control. All other data were analyzed by a one-way ANOVA, using linear contrasts to determine significant differences between the groups. Significance was accepted at P < 0.05 for all data. Data were analyzed using SuperAnova v1.11 software (Abacus Concepts, Berkeley, CA).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Influence of hydrocortisone on NF-B. Incubation of mononuclear cells with 1 µg/ml of LPS for 30 min resulted in activation of NF-B as demonstrated by an increased mobility shift of the NF-B oligo probe above the control (unstimulated) condition in the nuclear fractions of mononuclear cells (lanes 1 and 2, Fig. 2A). Preincubation of the cells with physiological concentrations of hydrocortisone attenuated the LPS-stimulated activation of NF-B (lanes 3-5). Densitometric analysis indicated that LPS caused a 120% increase in NF-B activation in the nuclear fraction above resting, unstimulated conditions (Fig. 2B). Additionally, preincubation of the cells with hydrocortisone at concentrations similar to normal circulating plasma levels (50-250 ng/ml) decreased LPS-stimulated activation of NF-B. However, these concentrations of hydrocortisone did not completely prevent NF-B activation (i.e., return to the control condition). Probe specificity was verified by competition and supershift experiments, as shown in Fig. 3.

View larger version (41K): [in this window] [in a new window]   Fig. 2.   Effect of hydrocortisone on NF-B DNA binding and cytosolic IB in lipopolysaccharide (LPS)-stimulated peripheral blood mononuclear cells (PBMC). Cells were preincubated with hydrocortisone ranging from 0 to 200 ng/ml for 1 h, followed by stimulation with 1 µg/ml of LPS for 30 min. A: a representative electrophorectic mobility shift assay (EMSA) of NF-B DNA binding from 1 subject. B: hydrocortisone (HC) decreased LPS-stimulated NF-B DNA binding. Mean densitometric units for NF-B activity were obtained from the EMSA for all subjects (n = 7). Densitometry units for NF-B DNA binding were normalized to the unstimulated control condition for each subject (* P < 0.002 compared with unstimulated control;  P < 0.035 compared with LPS-stimulated condition without HC). C: representative Western blot for cytosolic IB from 1 subject. D: HC inhibited the LPS-stimulated decrease in cytosolic IB. Mean densitometry units were obtained from each Western immunoblot and normalized to the unstimulated control condition for all subjects (n = 7).

View larger version (44K): [in this window] [in a new window]   Fig. 3.   EMSA demonstrating competition and supershifting of the NF-B oligo probe. Nuclear extracts from LPS-stimulated PBMC were incubated with either 32P-labeled probe alone (lanes 1, 3, and 5) or labeled probe plus an excess of unlabeled probe (lane 2). Additionally, extracts from LPS-stimulated PBMC were incubated with labeled probe followed by incubation with antibodies (Ab) against NF-B subunits before electrophoresis. Anti-p65 inhibited probe binding (lane 4), and anti-p50 supershifted the probe (lane 6).

Influence of hydrocortisone on IB. Incubation of mononuclear cells with LPS resulted in reduced cytosolic IB concentrations compared with the unstimulated control (lanes 1 and 2, Fig. 2C). Preincubation of the cells with hydrocortisone attenuated the LPS-stimulated decreases in cytosolic IB (lanes 3-5). Densitometric analysis revealed that LPS decreased cytosolic IB by 25% compared with the unstimulated control, whereas preincubation with hydrocortisone diminished this response (Fig. 2D). There was a reciprocal correspondence between the amount of NF-B in the nucleus and the amount of IB detected in the cytosol (Fig. 2, B and D); however, there was a large degree of variation in the influence of hydrocortisone on IB. Therefore, subsequent analyses of the data examining the antagonism between MIF and hydrocortisone were normalized to each dose of hydrocortisone.

Interaction of MIF and hydrocortisone on NF-B. To test the hypothesis that MIF counteracts hydrocortisone activity, PBMC were preincubated with MIF (0, 0.1, 1 ng/ml) along with hydrocortisone before LPS stimulation. As shown in Fig. 4A, MIF partially reversed the hydrocortisone-mediated inhibition of NF-B. To quantify the antagonism of MIF on hydrocortisone activity, densitometric analyses were normalized to the hydrocortisone response without MIF for each dose of hydrocortisone. As shown in Fig. 4B, both doses of MIF increased NF-B activation above that observed with hydrocortisone and LPS (P < 0.05, across all doses of hydrocortisone); however, neither dose of MIF returned NF-B to the levels observed with LPS alone. MIF alone (without LPS) did not change NF-B from the unstimulated control level: densitometric analysis of the NF-B activation from such cultures incubated with 1 ng/ml was 117 ± 17% of the control condition (n = 6, data not shown).

View larger version (50K): [in this window] [in a new window]   Fig. 4.   Interaction of HC and migration inhibitory factor (MIF) on NF-B DNA binding in LPS-stimulated PBMC. Cells were preincubated with HC ranging from 0 to 200 ng/ml and with MIF ranging from 0 to 1 ng/ml for 1 h, followed by stimulation with 1 µg/ml of LPS. A: a representative EMSA from NF-B DNA binding from 1 subject. All lanes are from a single gel but are rearranged to match the order in other figures. The LPS-alone condition was run on 1 lane in the gel, but the image is shown 3 times (left in each group) to facilitate comparisons. B: MIF antagonized the action of HC. Mean densitometric units for NF-B were obtained from the EMSA and normalized to the LPS + HC condition for each dose of HC (50, 100, and 200 ng/ml) for each subject (n = 7). Across the 3 concentrations of HC, MIF significantly increased NF-B binding compared with control conditions without MIF (* P < 0.05, control vs. 0.1 ng/ml MIF;  P < 0.05, control vs. 1 ng/ml MIF).

Interaction of MIF and hydrocortisone on IB. Hydrocortisone is reported to inhibit NF-B activation by increasing cytosolic IB, thereby preventing NF-B translocation to the nucleus. Therefore, we sought to determine whether MIF prevented hydrocortisone-induced increases in cytosolic IB. Preincubation of cells with hydrocortisone and MIF before LPS stimulation resulted in levels of IB that were lower than those observed with preincubation of hydrocortisone (Fig. 5A). To quantify the antagonism of MIF on hydrocortisone activity, densitometric analyses were normalized to the hydrocortisone response without MIF for each dose of hydrocortisone. As shown in Fig. 5B, MIF attenuated the hydrocortisone-mediated preservation of cytosolic IB. The higher dose of MIF (1 ng/ml) significantly decreased cytosolic IB across the doses of hydrocortisone (P < 0.05), whereas 0.1 ng/ml had no statistically significant effect.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Effect of HC and MIF on cytosolic IB in LPS-stimulated PBMC. A: a representative Western blot for cytosolic IB for 1 subject. B: MIF antagonized the HC-induced increase in cytosolic IB. Mean densitometric units for cytosolic IB were obtained from the Western blot and normalized to the LPS + HC condition for each dose of HC (50, 100, and 200 ng/ml) from all subjects (n = 7). Across the 3 concentrations of HC, MIF significantly reduced cytosolic IB compared with control conditions without MIF (* P < 0.05, control vs. 1 ng/ml MIF).

View larger version (11K): [in this window] [in a new window]   Fig. 6.   Effect of MIF and LPS on cytosolic IB in the absence of HC. PBMC were incubated with MIF and LPS for 30 min, and cytosolic IB was examined by Western blot. LPS significantly decreased cytosolic IB (* P < 0.03), whereas MIF did not enhance this effect of LPS on IB. Densitometric units for each blot were normalized to the unstimulated control condition (n = 6).

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Effect of HC and MIF on cytosolic IB in the absence of LPS. PBMC were incubated with MIF and HC for 1 h and examined for cytosolic IB by Western blot. HC (200 ng/ml) increased cytosolic IB, and 1 ng/ml of MIF significantly inhibited the effect of HC on IB. Densitometric units for each blot were normalized to the unstimulated control condition (n = 6). *P = 0.03.

Binding interactions between MIF and hydrocortisone. The possibility that MIF modulated hydrocortisone by acting as a binding protein that prevented hydrocortisone association with target cells was tested in two ways: first, by determining whether MIF directly bound [³H]hydrocortisone in solution, and second, by determining whether MIF interfered with specific binding of [³H]hydrocortisone in PBMC.

View larger version (17K): [in this window] [in a new window]   Fig. 8.   A: elution of a mixture of MIF and [3H]hydrocortisone from a PD-10 column after 2-h incubation at room temperature. Fraction volume was 0.5 ml. B: binding of [3H]hydrocortisone to human PBMC after incubation with various concentrations of MIF. NSB, nonspecific binding; CPM, counts/min. Values are means ± SD for a single experiment.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The major finding of this study is that MIF counteracts the hydrocortisone-mediated inhibition of NF-B. The mechanism for the counterregulatory action of MIF involves prevention of hydrocortisone-induced increases in cytosolic IB. As a result, NF-B can translocate to the nucleus and activate transcription, even in the presence of normal physiological concentrations of hydrocortisone. Additionally, hydrocortisone, in concentrations normally found in the plasma, inhibits NF-B activation in LPS-stimulated PBMC. This extends previous findings based on pharmacological concentrations of glucocorticoids (13).

The present study tested several cellular pathways that could account for the counterregulatory effect of MIF on hydrocortisone activity. MIF could activate an intracellular pathway similar to LPS, leading to IB kinase (IKK) activation and ultimate IB degradation, thus overriding the hydrocortisone induced-inhibition of NF-B activation. If this were the case, then MIF should have enhanced IB degradation in LPS-stimulated cells. However, this study demonstrated that LPS stimulation resulted in a 28% decrease in cytosolic IB, and preincubation with MIF did not significantly change the IB levels compared with LPS alone (see Fig. 6). Thus, it is unlikely that MIF is functioning through an intracellular pathway, similar to LPS, that leads to IB degradation. Furthermore, MIF alone did not alter cytosolic IB levels.

Alternatively, MIF may interfere with hydrocortisone-induced IB synthesis. Hydrocortisone increases IB expression, thus preventing NF-B translocation to the nucleus and activation of transcription (1, 13). In the present study, incubation of unstimulated mononuclear cells with 200 ng/ml of hydrocortisone significantly increased cytosolic IB. The addition of MIF was able to prevent this hydrocortisone-induced increase in IB. Thus one counterregulatory effect of MIF is the prevention of hydrocortisone-induced IB synthesis.

Hydrocortisone also prevents NF-B from binding to DNA in the nucleus. Hydrocortisone bound to the glucocorticoid receptor migrates to the nucleus, where it competes for DNA binding sites with NF-B, thereby preventing NF-B-dependent transcription (14). However, the direct interference of glucocorticoids with NF-B DNA binding has only been demonstrated in transient cotransfection experiments (8). Whether this mechanism occurs in primary cells, such as PBMC, remains to be determined. This mechanism was not examined in our studies and therefore cannot be ruled out as a mode of interaction of MIF with hydrocortisone.

Glucocorticoid-induced increases in IB synthesis are believed to be a direct influence of the glucocorticoid receptor-hormone complex on the transcription rate of IB (13). It is possible that MIF may bind hydrocortisone (similar to the plasma glucocorticoid binding proteins), thereby preventing it from crossing either the cellular or the nuclear membrane and associating with the intracellular glucocorticoid receptor and activating IB transcription. However, we could not demonstrate that MIF either bound hydrocortisone or prevented it from associating with mononuclear cells. Alternatively, MIF may activate another cellular factor or pathway and thus indirectly interfere with the action of hydrocortisone on IB synthesis. This indirect effect would suggest that MIF binds to a membrane receptor, inducing secondary factors. However, receptors for MIF have not been discovered to date. Therefore, the cellular pathway through which MIF inhibits hydrocortisone induction of IB remains an open question.

Perspectives

7