Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Iron accumulates within a variety of organ systems in association with the onset of senescence in normal tissue (1). Despite this, the body has no known regulatory mechanism for the disposal of excess iron. This is an important point of consideration because although this transitional metal is essential to life, increasing tissue concentrations of iron leads to a significantly increased risk of infection (2, 3), fibrosis (4), and neoplasia (5, 6).

The human lung is no exception to age-related "iron loading": both alveolar structures (7, 8) and pulmonary macrophages (9) accumulate iron. Cigarette smoking escalates this phenomenon by a variety of mechanisms, including the delivery of iron particles in the smoke that is inhaled (10, 11). We have recently reported significant regional variation in the accumulation of alveolar iron in the lungs of smokers compared with those of nonsmokers (12). Concentrations of iron and iron-binding proteins in the alveoli of the upper lobes were significantly greater when compared with those of lower lobes. The significance of these findings with respect to the pathogenesis of lung diseases that have a predilection for the upper lung region, such as smoking-related cancer and emphysema (13), remains unclear. However, the possibility that there could be a cause-and-effect relationship has to be considered because we (16) and others (17) have shown that iron can affect the production of a variety of macrophage-derived inflammatory mediators, including interleukin (IL)-1.

In light of the emerging realization that iron apparently has affects on lipopolysaccharide (LPS)-induced production of IL-1, together with our data that show regional variability in the pulmonary distribution of iron, we postulated that we could (1) demonstrate regional differences in the release of IL-1 from human alveolar macrophages and (2) influence the production of IL-1 in human macrophages by altering intracellular iron concentrations. To test the first of these two hypotheses, alveolar macrophages were obtained via independent lavage of both the upper and lower lobes of healthy volunteers (both smokers and nonsmokers), after which the ability of each population to secrete IL-1, as well as tumor necrosis factor (TNF)-, IL-6, and IL-8, was quantified. Additionally, we established an in vitro model of "iron-loaded" cells using the human myelomonocytic cell line THP-1. This model allowed us to examine more directly the effect of iron and its chelation by deferoxamine (DFA) on the secretion of IL-1.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

Human THP-1 cells (TIB 202) were obtained from the American Type Culture Collection (ATCC; Rockville, MD). RPMI-1640 medium was from BioWhittaker (Walkersville, MD). Fetal bovine serum was purchased from HyClone (Logan, Utah). Streptomycin sulfate, penicillin, L-glutamine, 2-mercaptoethanol, deferoxamine mesylate, LPS (Escherichia coli serotype 0111: B4), Kodak XAR5 film, ethylenediaminetetraacetate (EDTA), H₂O₂, and L-ascorbic acid were all purchased from Sigma (St. Louis, MO). The sodium salicylate was from Aldrich (Milwaukee, WI). Heta-bound DFA was a generous gift from Biomedical Frontiers (Minneapolis, MN). The BCA Protein Assay was from Pierce (Rockford, IL). Dibasic and monobasic sodium phosphate and all cell culture plates were purchased from Fischer Scientific (Fair Lawn, NJ). The 96-well, black microfluor plates were from Dynex Technologies (Chantilly, VA). The IL-1, TNF-, IL-6, and IL-8 enzyme-linked immunosorbent assay (ELISA) kits were from R&D Systems (Minneapolis, MN). TRIzol reagent was obtained from GIBCO, Life Technologies (Gaithersburg, MD). The nylon membrane Nytran was from Schleicher & Schuell (Keene, NH). QuikHyb, the prehybridization solution, was from Stratagene (LaJolla, CA). The complementary DNA (cDNA) probe specific for IL-1 (pGEM 1-based plasmid cleaved with EcoRI in our laboratory) was a generous gift from Stephen Gillis at Immunex (Seattle, WA). The Multiprime DNA labeling system was purchased from Amersham (Chicago, IL). Quick Spin Columns were purchased from Boehringer Mannheim (Indianapolis, IN). Cytospin slide preparation supplies were purchased from Shandon Southern Products, Ltd. (Cheshire, UK). Diff-Quik kits were obtained from Harleco (Philadelphia, PA).

Subjects

Thirteen healthy, smoking volunteers and seven nonsmoking volunteers underwent bronchoalveolar lavage (BAL). Mean ages were 39 ± 2 yr for smokers and 29 ± 1 yr for nonsmokers. Smokers used at least one pack per day (mean pack years of 10 ± 2). None of the subjects was taking medication, had a history of pulmonary disease, nor had a recent upper respiratory tract infection. Subjects had physical exams that were normal, and there was no evidence of lung disease, as judged by pulmonary function tests. All gave informed consent and the institutional human subjects committee approved the protocol.

BAL and Alveolar Macrophage Recovery and Culture

All subjects were premedicated with meperidine and/or midazolam while reclining at a 45-degree angle. After anesthetizing the oral cavity with aerosolized tetracaine, a fiberoptic bronchoscope was inserted orally and wedged into a segment of the right and/or left upper and lower lobes. BAL was performed using five 20-ml aliquots of sterile saline. Patients tolerated the procedure well.

Lavage fluid was filtered through four layers of sterile gauge and centrifuged (400 × g for 10 min). The cell pellet was washed with RPMI 1640 and recentrifuged a total of three times. The final pellet was resuspended in RPMI 1640 medium supplemented with 100 µg/ml streptomycin, 100 U/ml penicillin, and 10% fetal bovine serum. Cell viability was determined by trypan blue exclusion, and cells were counted in a hemacytometer. A differential cell count was determined using a cytospin slide preparation stained with Diff-Quik. Cells were plated in 35-mm culture dishes at population densities of 10⁵ cells/plate for lactate dehydrogenase determination and cytokine evaluation. The cells were incubated at 37°C in air plus 5% CO₂ for 1 h. Plates were then washed gently with medium to remove nonadherent cells. Unstimulated cell cultures were replenished with fresh medium. Stimulated cell cultures received medium to which LPS had been added at a concentration of 1.0 µg/ml. Cell viability was verified at the end of incubation, again by trypan blue exclusion.

Cell Culture and Iron Loading of THP-1 Cells

THP-1 cells were maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum (heat-inactivated at 56°C for 1 h), 2 mM L-glutamine, 50 µM 2-mercaptoethanol, 100 U/ml penicillin, and 10 µg/ml streptomycin sulfate. THP-1 cells were cultured per ATCC guidelines, which included the 2-mercaptoethanol supplementation.

Ferric nitrate was added to medium at the concentrations specified in the figures (from 5 to 45 µM), and cells were maintained in 5% CO₂ at 37°C for 7 d. The THP-1 cells tolerated higher concentrations of iron "loading"; however, upon stimulation with LPS, cell survival was decreased in cells grown in concentrations above 50 µM ferric nitrate (data not shown). We used the prolonged exposure in order to stabilize the state of the iron within the macrophage before stimulation; the 7-d "load" resulted in the optimal reproducibility of intracellular iron data. After incubation in iron-supplemented medium, but before treatment with either 0.001 to 10 µg/ml LPS, 0.1 to 25 mM DFA, or 0.1 mM Heta-bound DFA, cells were first collected by centrifugation and resuspended in medium. Suspensions were seeded at 1 × 10⁶ cells/2 ml on 60-mm plates. The time interval of treatment varied, as indicated in the figures. Cells were then harvested for the preparation of lysates, as described subsequently.

Analysis of Intracellular Iron

After at least 7 d in medium or iron-supplemented medium, THP-1 cells were collected by centrifugation (250 × g for 20 min; 4°C), and pellets were carefully washed two times with normal saline at room temperature. Pellets were resuspended in 100 µM EDTA (250 µL/ 10⁶ cells). Three freeze (70°C for  1 h) /thaw (37°C for 1 h) cycles were conducted, after which the cell debris were removed by centrifugation (10,000 × g for 2 min; 4°C). After the quantification of protein, lysates were stored at 20°C for future analysis.

Bioavailable iron, that is, the chelatable pool of loosely bound iron, was evaluated by measuring Fe³⁺-EDTA using a redox cycling, fluorometric assay developed in our laboratory. Redox cycling proceeds via reduction of Fe³⁺-EDTA with L-ascorbate and oxidation of Fe²⁺-EDTA back to Fe³⁺-EDTA with H₂O₂. The hydroxyl radicals (HO^.) formed during the H₂O₂-mediated oxidation of Fe²⁺-EDTA react with sodium salicylate, the 2,5-dihydroxybenzoic acid hydroxylation product of which is detected fluorimetrically (18). Assay specificity is assured by using EDTA, which increases the redox activity of Fe³⁺ and chelates the major competing transition metal, copper, thus blocking redox activity of the latter. Release of ferritin iron during this assay is minimal; the Fe³⁺ measured when ferritin is added to the reaction mixture is about 3% of the Fe³⁺ detected after ferritin is pretreated with 0.1 N HCl. Extensive washing of glassware and new plastic 96-well plates with deionized water were necessary to remove contaminating iron. Reagent contamination below 0.02 µM Fe³⁺ was essential for full use of the assay sensitivity range. The assay reagent was 20 mM phosphate buffer, pH 6.6 (dibasic and monobasic sodium phosphate), 20 mM EDTA, 20 mM H₂O₂, and 90 µM sodium salicylate. To remove metals from the phosphate buffer, 8-hydroxyquinoline was used as described (19), with CCl₄ substituted for chloroform in the procedure. In the studies presented, 15 to 30 µl of lysate, standards, or blanks (100 µM EDTA) were added to 200 to 250 ml of assay reagent per well in a 96-well, black microfluor plate. The plates were preread in a fluorescent plate reader set at 360 nm excitation and 455 nm emission wavelengths. The cycling reaction was started with the addition of 800 µM L-ascorbic acid and was allowed to continue for 1 h at 50°C. The plate was then cooled to room temperature and reread in the plate reader. The concentration of Fe³⁺ in the sample was determined from a standard curve covering a range of 0.02 to 0.32 µM.

Release of IL-1 Protein

THP-1 cells were plated for 20 h at 5 × 10⁵ cells/ml in 35-mm-diameter plates that contained medium either with or without LPS and in the presence or absence of various concentrations of DFA. A range of 0.1 to 25 mM DFA was used, based on previous studies in which these concentrations were found to be effective for our purposes and nontoxic in cell culture (16). The suspensions were then collected and the cells pelleted by centrifugation (as previously described). Supernatants were removed to fresh tubes and frozen at 20°C for analysis of released protein. Concentrations of human IL-1, TNF-, IL-6, and IL-8 protein were determined using a sandwich ELISA, as described by the supplier.

RNA Extraction and Northern Blot Analysis

THP-1 cells (normal or iron-loaded) were seeded into 100-mm-diameter plates at 5 × 10⁶ cells/10 ml in medium either with or without LPS and in the presence or absence of DFA (at specified concentrations), and incubated in 5% CO₂ at 37°C. Cells were harvested by centrifugation after 20 h of incubation, the time of maximal IL-1 messenger RNA (mRNA) accumulation, and pellets were completely resuspended in 0.5 ml TRIzol reagent, and stored at 20°C. Total RNA was extracted by phenol:chloroform, followed by precipitation with isopropanol at 20°C, and a wash in 75% ethanol. Pellets were resuspended in water that had been treated with diethyl pyrocarbonate. The RNA samples were quantified spectrophotometrically at 280/260 nm, and 3 to 5 µg/ sample were electrophoresed through an agarose/formaldehyde gel (20). Gels were stained with ethidium bromide (21) and the 28S and 18S bands of ribosomal RNA were photographed before RNA was transferred to a nylon membrane by capillary action. The membranes were then baked, briefly irradiated at 254 nm, incubated for 1 h at 65°C in prehybridization solution, and hybridized for 1 h at 68°C with a cDNA probe specific for IL-1. The probe was labeled with [³²P]deoxycytidine triphosphate using the Multiprime DNA labeling system, and subsequently purified on a Quick Spin column. The membranes were washed twice in 2× saline sodium citrate (SSC)/0.1% sodium dodecyl sulfate (SDS), and twice more with 0.1% SSC/0.1% SDS at 60°C. The membranes were then exposed to Kodak XAR5 film for 16 to 20 h, after which autoradiograms were subjected to laser scanning and densitometric analysis. Results were normalized against the negative image of the 18S ribosomal RNA band.

Nuclear Run-On Analysis

The isolation of THP-1 nuclei, extraction and radiolabeling of nuclear RNA, and slot-blot hybridization of radiolabeled RNA to the immobilized DNA probe for IL-1 (see previously), -actin, were performed as described previously (22). Blots were autoradiographed at 70°C. Signals from nuclei isolated from cultures that were treated for either 1 or 3 h were analyzed with a Molecular Dynamics PhosphorImager, normalizing against signals obtained using -actin as a probe.

Data Analysis

The experimental results were analyzed for their statistical significance by the paired, two-tailed Student's t test. A confidence level of P < 0.05 was taken to represent a significant difference between two means.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Alveolar Macrophage and BAL Recovery

Upper lobe and lower lobe alveolar macrophages (AM) were recovered by BAL from healthy smokers and nonsmokers from the bronchoscopies used in our report describing regional differences in iron within the lungs of smokers (12). Recoveries of lavage fluids, and AM contained therein, are shown in Table 1. As previously reported (12), AM counts from upper and lower lobes were significantly higher in smokers compared with nonsmokers. There were no regional differences in either the recovery of cells or the return of lavage fluid from the lobes of smokers or nonsmokers. Neither did we find significant differences in differential analysis of cell populations obtained from upper and lower lobes.

AM acquired from upper lobes released less IL-1 compared with those AM acquired from lower lobes. To determine whether or not regional differences existed in the production of cytokines, AM isolated from both upper and lower lobes were stimulated with LPS. Supernatants were then assayed quantitatively for the presence of IL-1, TNF-, IL-6, and IL-8 after 20 h of treatment. Unstimulated AM from both smokers and nonsmokers produced insignificant amounts of IL-1, TNF-, IL-6, and IL-8 after adherence. The LPS-stimulated AM from smokers (n = 13) released less IL-1 when acquired from the upper lobe than did those acquired from the lower lobe (P < 0.05, Table 2). LPS-stimulated AM from nonsmokers (n = 7) also released less IL-1 if acquired from the upper lobe than did those acquired from the lower lobe (Table 2). The difference between the latter was not statistically significant. However, the release of IL-1 from AM acquired from upper lobes was consistently lower (55 to 60%) when compared with such from lower lobes, regardless of smoking history. Based on these observations, we postulated that the data obtained demonstrate regional variability in the release of IL-1 from AM. We propose that the diminished release of IL-1 from AM acquired from areas of lung that accumulate more iron was an example of a negative biologic correlation between iron and IL-1.

Intracellular Iron

The growth of THP-1 cells in ferric nitrate increased bioavailability of intracellular iron. To create a model by which we could evaluate the availability of iron and its relationship to IL-1 production, THP-1 cells were cultured in either the presence or absence of ferric nitrate (5 to 45 mM) for 7 d. To confirm that we were affecting concentrations of intracellular iron, THP-1 lysates were assayed for chelatable ferric iron before and after incubation. As expected, culturing THP-1 cells in ferric nitrate caused a dose-related increase in concentrations of intracellular iron (Table 3).

DFA decreased bioavailability of intracellular iron. To verify the fact that we could decrease concentrations of intracellular iron by exposing THP-1 cells to 0.1 mM DFA, we analyzed lysates of THP-1 cells for ferric iron before and after DFA treatment. As expected, a significant decrease in concentrations of intracellular iron, compared with controls, was obtained within 1 h of exposure of THP-1 cells to DFA (Table 4). Reduced levels were maintained thereafter for at least 20 h in medium + DFA. Because we were interested in distinguishing between the effects of intracellular and extracellular chelation of iron, we also used DFA bound to Heta-starch (Heta-DFA). The irreversible covalent polymerization of DFA to the large starch molecules precluded access to the intracellular space but did not interfere with the chelation of iron extracellularly (23). DFA and Heta-DFA chelated all detectable iron in supernatants, but the Heta-DFA did not alter intracellular concentrations of iron (Table 4).

LPS-mediated stimulation increased bioavailability of intracellular iron. To identify possible shifts in the bioavailability of iron after treatment with LPS, intracellular iron was measured after THP-1 cells were exposed to this stimulus (Figure 1A). At 1 h postexposure to LPS, there was a significant increase in intracellular iron in THP-1 cells compared with untreated control cells (P < 0.05). Values then returned to, and remained at, 115% of the untreated control throughout the 20-h treatment period. From these data, it was concluded that LPS-mediated stimulation caused rapid, but transient, mobilization of iron to the chelatable pool within THP-1 cells.

View larger version (14K): [in this window] [in a new window]   Figure 1.   Time course of intracellular iron bioavailability after cellular treatment in the presence or absence of LPS and/or DFA. (A) THP-1 cells were exposed to either 1.0 µg/ml LPS or a combination of LPS with 0.1 mM DFA over a 20-h time course. Samples were collected for the preparation of lysates at the times indicated, and intracellular iron was measured. Values are reported as the percent of control (cells incubated in medium alone) ± SEM. *P < 0.05. Control, squares; LPS, diamonds; DFA (0.1 mM), circles; LPS/DFA (0.1 mM), triangles. (B) THP-1 cells were exposed to either 1.0 µg/ml LPS or a combination of LPS with 25 mM DFA over a 20-h time course. Samples were collected for the preparation of lysates at the time indicated, and intracellular iron was measured. Values are reported as the percent of control (cells incubated in medium alone) ± SEM. *P < 0.05. Control, squares; LPS, diamonds; DFA (25 mM), solid squares; LPS/DFA (25 mM), solid diamonds.

THP-1 cells treated with 0.1 mM DFA quickly demonstrated a reduction in concentrations of intracellular iron that persisted throughout the 20-h period of incubation. The simultaneous addition of 0.1 mM DFA and LPS eliminated the LPS-associated early increase in iron, and reduced concentrations of intracellular iron to levels similar to those of THP-1 treated with DFA alone. However, after the early decrease in available iron, LPS + DFA- treated THP-1 cells continued to mobilize iron throughout the treatment period. By 6 h, THP-1 cells had compensated for the early decrease, and iron levels returned to those of untreated control cells. This rebound continued throughout the 20-h incubation period, resulting in iron concentrations that were 220% that of LPS-stimulated THP-1 cells incubated in the absence of DFA. Thus, while the addition of DFA to THP-1 cells caused a reduction in concentrations of intracellular iron in both LPS-treated and -untreated groups, cotreatment with DFA + LPS appeared to cause a compensatory supermobilization of iron to the "chelatable pool." We concluded from these experiments that LPS, directly or indirectly, has the capacity to cause both an early and late mobilization of iron to the chelatable pool within the THP-1 cell dependent upon the presence or absence of DFA.

Increasing the concentration of DFA to 25 mM caused a decrease in the level of intracellular iron by 90%, whether or not LPS was present throughout the 20-h incubation period (Figure 1B). It appeared that at higher concentrations, DFA overcame the ability of the LPS-stimulated THP-1 cells to compensate and mobilize iron to the chelatable pool, implicating a finite amount of iron that is available for mobilization.

Cytokine Production by THP-1 Cells

Iron-loading decreased the release of IL-1 from THP-1 cells. To evaluate the effect of iron-loading on the release of IL-1 by cultured THP-1 cells, cells were stimulated with 1.0 µg/ml LPS, and the release of both IL-1 and TNF- was assayed after 20 h. The comparison of IL-1 and TNF- was done to determine whether the effects of iron were specific to IL-1 or indicative of a general suppression of cytokine production. Increasing the iron-load of THP-1 cells was associated with a decrease in the release of IL-1 from LPS-stimulated THP-1 cells (Table 5). However, significant decreases in the release of TNF- were not observed under these same conditions (data not shown). From these data, it was concluded that production of IL-1 was inversely related to the accumulation of iron within THP-1 cells. This conclusion validated our hypothesis that there exists a cause-and-effect relationship between the concentration of intracellular iron and the release of IL-1.

The decrease in release of IL-1 by THP-1 cells was accompanied by a decrease in the accumulation of mRNA specific for IL-1. This became apparent after 20 h of LPS stimulation of iron-loaded THP-1 cells when compared with control cells (Figure 2).

View larger version (27K): [in this window] [in a new window]   Figure 2.   Accumulation of IL-1 mRNA. After a 1-wk period of maintenance in medium or medium supplemented with ferric nitrate at the concentrations indicated, THP-1 cells were washed and resuspended in medium and exposed to 1.0 µg/ml LPS in the presence or absence of 0.1 mM DFA for 20 h. The accumulation of IL-1 mRNA was measured. Autoradiogram and densitometric values illustrate one representative experiment out of four.

Chelation of iron increased the production of IL-1. To assess further the effect that iron has on the release of IL-1 protein from LPS-stimulated THP-1 cells, we included in our studies the presence or absence of DFA. We expected that because iron-loading of THP-1 cells had decreased the release of IL-, chelation of iron would have the opposite effect, owing to the fact that DFA decreased intracellular iron concentrations in THP-1 cells. As predicted, DFA increased the release of IL-1 from LPS-stimulated THP-1 cells; in fact, it did so to an extent that exceeded our expectations (Table 5). Ablating the early mobilization of iron with 0.1 mM DFA, but allowing a compensatory supermobilization of iron, caused a 12- to 16-fold superinduction of IL-1. And, as predicted, iron-loading THP-1 cells reduced the augmentary effect of 0.1 mM DFA on LPS-mediated release of IL-1 (Figure 3). However, ablating the ability to mobilize any iron (by 25 mM DFA) resulted in (1) an increase of only 3- to 5-fold in ability to induce IL-1 release and (2) loss of the influence by iron-loading THP-1 cells on the release of IL-1 (Table 5). Thus, iron may have an inhibitory effect during the early phase of LPS-mediated stimulation but a later augmentory effect.

View larger version (13K): [in this window] [in a new window]   Figure 3.   Release of IL-1 protein. After a 1-wk period of maintenance in medium or medium supplemented with ferric nitrate at the concentrations indicated, THP-1 cells were washed and resuspended in medium. The cells were then exposed to either LPS (from 0.001 to 1.0 µg/ml) or a combination of LPS with 0.1 mM DFA for 20 h. The release of IL-1 protein was measured from the supernatants. The data are expressed as mean IL-1 (pg/105 cells) ± SEM. *P < 0.05. Medium with LPS, open squares; medium with LPS/DFA, divided squares; 5 µM iron with LPS/DFA, triangles; 45 µM iron with LPS/DFA, divided circles.

To examine whether the pool of iron responsible for the negative effect on the release of IL-1 was intracellular, extracellular, or both, we again performed chelation experiments using Heta-DFA. As expected, Heta-DFA had no effect on the release of IL-1 by LPS-stimulated THP-1 cells (Table 6). Thus, we were able to conclude that the iron that was active in reducing the release of IL-1 was part of the intracellular pool.

To determine whether or not the positive effect of iron chelation was unique to the LPS-stimulated release of IL-1, the effect of 0.1 mM DFA on LPS-stimulated release of other cytokines was evaluated. THP-1 cells, without additional iron, treated by the combination of LPS + DFA showed an increased release of TNF- (3.9 ± 0.2-fold), IL-6 (6.7 ± 0.3-fold), and IL-8 (1.4 ± 0.5-fold) when compared with cells treated with LPS alone. However, in this same population of cells, the increase of IL-1 was 37 ± 18-fold. Thus, whereas chelation of iron had a general augmentory effect on the production of cytokines by LPS-stimulated THP-1 cells, the effect on production of IL-1 was dramatically greater in magnitude.

Chelation of iron increased the transcription rate of the IL-1 gene. To test for negative effects related to iron early in LPS-mediated stimulation, we examined the rate of transcription of the gene that encodes IL-1. Northern blot analysis showed that LPS-stimulated THP-1 cells accumulated more IL-1 mRNA in the presence of DFA than did cells treated with LPS alone (Figure 4). Transcriptional analysis by nuclear run-on assay showed that transcription of the IL-1 gene was increased by the combination of LPS and DFA (Figure 5) compared with LPS alone. Whereas transcription of the -actin gene does change with the LPS, there is no difference between the LPS and LPS + DFA groups. Thus, in contrast to the situation observed for the IL-1 gene, transcription of the -actin gene was not changed by DFA. From these data, we were able to conclude that the primary effect of iron is to downregulate transcription of the IL-1 gene.

View larger version (26K): [in this window] [in a new window]   Figure 4.   Accumulation of IL-1 mRNA. THP-1 cells were incubated in either medium or medium with 1.0 µg/ml LPS, 0.8 mM DFA, or both LPS and DFA, for 20 h. Accumulation of IL-1 mRNA was measured. Autoradiogram and densitometric values illustrate one representative experiment out of three.

View larger version (28K): [in this window] [in a new window]   Figure 5.   Transcription of IL-1. THP-1 cells were incubated in either medium or medium with 1.0 µg/ml LPS or 1.0 µg/ml LPS with 0.1 mM DFA, for either 1 or 3 h. The rate of IL-1 transcription was measured. Autoradiogram and densitometric values illustrate one representative experiment out of two.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, we have shown that the capacity of AM to produce IL-1 is not uniform throughout the lung, a finding that parallels our previous report of regional variations in alveolar iron concentrations within the human lung. We also report a strong negative correlation between the production of human IL-1 and bioavailability of intracellular iron using the human myelomonocytic cell line THP-1. Furthermore, we demonstrated a fluctuating pool of chelatable iron that is sensitive to cellular treatment with either LPS or DFA that accounts for the alterations demonstrated. All in all, we provide both a physiologic example and in vitro data supporting a relationship between iron and the production of IL-1. These findings may have particular importance in lung diseases, such as those associated with smoking, that demonstrate a regional predominance (13).

Tissue iron is important because it has been implicated in the pathophysiology of a number of conditions, including infection (2, 3) and neoplasia (5, 6). Increases in the burden of iron in the lower respiratory tract of smokers are also well documented (7, 8). We have previously demonstrated that iron accumulation has a regional predilection for the upper lobes of the lungs of smokers. The data reported here also demonstrate regional variation in AM function. There were no differences in lavage fluid return, alveolar lavage cell counts, or differentials in upper versus lower lobes to account for these differences. Yet, AM cultured in the presence of LPS produced less IL-1 when they were acquired from the upper lobe of a subject that smoked. Significant regional variation was not found with other cytokines, including TNF, IL-6, and IL-8. Our data suggest that uneven distribution of iron throughout the lung may alter important immunomodulatory function on a regional basis.

Moving beyond the limitations of BAL in humans, we provide evidence for a cause-and-effect relationship between iron and IL-1 by the utilization of THP-1 cells. Our data demonstrated that iron-loading of the THP-1 cells, verified by intracellular analysis, decreased the LPS-induced release of IL-1 protein and the accumulation of IL-1 mRNA in a dose-dependent manner. The capacity of the THP-1 cell to produce IL-1 after stimulation with LPS was increased 20-fold when intracellular iron concentrations were decreased by the iron chelator DFA. This remarkable augmentation in the release of IL-1 protein by DFA was reduced by previously iron loading the THP-1 cells, further solidifying a relationship between iron and control of IL-1. Our findings also demonstrate that an increase in the transcription rate of the IL-1 gene as a consequence of the chelation of iron contributes to the increased release of IL-1. Whether an increase in stabilization of IL-1 mRNA may also contribute to its accumulation is not addressed in this report. Furthermore, we identify the intracellular space as the location of iron chelation that alters the production of IL-1 through comparisons between experiments involving intracellular and/or extracellular chelation. We have previously reported an augmentation in LPS-induced IL-1 release from human AM using concentrations of DFA similar to those used in this report (16). In addition, iron-related suppression of other mediators such as inducible nitric oxide synthase (24) and TNF- has been reported (17). In these reports, as in our study, an iron-related reduction in transcription rates and/or accumulation of mRNA has been the mechanism implicated. In contrast to the report by Silver and colleagues (17), we did not observe a significant reduction in the release of TNF- by iron-loaded THP-1 cells that were stimulated with LPS, either in this study or our previous report using human AM. However, considerable differences exist between our studies and those previously reported, in the methods of iron supplementation and time intervals used. In our experience, the control of IL- release is more sensitive than that of TNF- to the influence of iron in both the experiments reported here using LPS-stimulated THP-1 cells and previous experiments of ours using human AM (16). Whereas the effect of decreased iron bioavailability has a general effect on cytokine production, the increases in production of TNF-, IL-6, and IL-8 are subtle in comparison with the dramatic increase observed in the production of IL-1.

During our investigation of the intracellular pool of chelatable iron, we revealed an increase in iron bioavailability by the mobilization of iron to the chelatable pool within an hour after LPS stimulation. Additional evidence, regarding the influence of LPS on iron bioavailability, is provided by the demonstration of a decline and subsequent increase in the chelatable pool of iron within THP-1 cells treated with DFA and LPS. In contrast, cells treated with DFA alone showed a decrease in the concentration of intracellular iron that did not re-equilibrate over the 20-h treatment interval. In fact, that concentration of DFA which caused by cotreatment with LPS the greatest increase in the release of IL-1 (0.1 mM; Table 5) caused, in parallel experiments, an increase in the chelatable pool to more than twice that of controls (Figure 2). Increasing the concentration of chelator in these experiments overcame the capacity that the cell had to increase intracellular iron and was associated with the production of less IL-1. Thus, whereas an early decrease in iron bioavailability appears to augment the transcription of IL-1, we speculate that a critical role for iron exists in the release of IL-1.

That iron and its chelation have paradoxical effects on the production of IL-1 is not surprising given the discrepancies in the literature. Clearly, iron can have a pro-oxidant influence (25, 26), and IL-1 has promoter elements that are both oxidant and antioxidant activated (27). Thus, changes in iron bioavailability may have biphasic effects on transcription. We and others have previously demonstrated that the production of IL-1 is augmented by oxidant stress (30) and inhibited by antioxidants (16). However, our data supports that lowering iron, which is in effect an antioxidant influence, augments the production of IL-1, and although IL-1 has an antioxidant-sensitive promoter, active protein-1 (33), we have not been able to demonstrate augmentation by antioxidants (16) and have no knowledge that this has been reported. In addition, iron may also exert influence over promoters that are not dependent on catalytic properties. For example, iron has control of its storage and carrier proteins through iron-responsive mechanisms (34); however, this is post-transcriptionally mediated and IL-1 mRNA does not contain the recognized sequence motif homologous to the iron-response element of either ferritin or transferrin receptor genes (35). Thus, at this point, no cohesive model can be drawn for the effect of iron on the expression of the IL-1 gene.

While we propose a regulatory role for iron on the production of IL-1, additional evidence also exists that strengthens support for further integration of the production of IL-1 and the metabolism of iron. The synthesis of both subunits of ferritin, heavy and light, is increased (36, 37) in the presence of IL-1. These mechanisms appear to be mediated independent of the conventional control mechanisms determined by iron availability. However, this enhancement in ferritin production by IL-1 would be held "in check," given our data, by the limitation of IL-1 production in the presence of high iron concentrations. Whereas it is clear that the handling of iron and the production of IL-1 are interrelated, the appreciation of this relationship is in its infancy.

Data presented here substantiate our hypothesis that an intracellular, chelatable pool of iron exerts significant control over the production of IL-1. We also demonstrate that this pool of chelatable iron increases both after loading with exogenous iron and after stimulation with LPS. The increase in intracellular iron appears to suppress transcription of the IL-1 gene, whereas diminution of this pool by chelation augments transcription of IL-1 and dramatically increases the release of the cytokine. We also demonstrate that the control of IL-1 by iron is unique, at least in its magnitude of response to the alteration of iron bioavailability, compared with TNF-, IL-6, and IL-8. Finally, we recognize a regional negative correlation between excess alveolar iron and the production of human AM-derived IL-1, suggesting a relationship between iron and inflammation that may have clinical significance, especially in regards to lung diseases with a regional predominance.