INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Iron is required for normal cell function; however, unbound iron catalyzes the generation of highly reactive hydroxyl radicals from less reactive oxygen species and promotes oxidative cell injury (1). Although sequestration of intracellular iron within ferritin protects cells from iron-catalyzed oxidative injury, extracellular ferritin-bound iron can promote oxidative injury to nearby cells because iron can be mobilized from ferritin by reducing agents (2). In healthy, nonsmoking humans, the lower respiratory tract contains only small amounts of extracellular iron, which is bound predominantly to transferrin, an iron transport protein that generally prevents iron mobilization (3).

Large numbers of neutrophils are present in the lungs of patients with cystic fibrosis, even during periods of clinical stability (4). Neutrophils can generate reactive oxygen species and long-lived oxidants are present in sputum produced by patients with cystic fibrosis (5). Free or ferritin-bound iron within the lower respiratory tract of patients with cystic fibrosis could promote oxidant lung injury in these patients by catalyzing the generation of highly reactive hydroxyl radicals from less reactive oxygen species (1). Pseudomonas-derived proteases have been shown to cleave transferrin and lactoferrin in the lungs of patients with cystic fibrosis, inhibiting the capacity of these iron-binding proteins to protect against the iron-catalyzed generation of hydroxyl radicals (6, 7). This impaired protection by cleaved transferrin and lactoferrin would promote iron-catalyzed generation of hydroxyl radicals in the lungs of patients with cystic fibrosis.

Ferritin is composed of 24 subunits consisting of two types: a light (L) ferritin (MW 19,000) and a heavy (H) ferritin (MW 21,000) (8). Ferritin composed predominantly of L or H subunits is termed L-type or H-type ferritin, respectively. These isoferritins (L- and H-type ferritin) are functionally different, with H-ferritin taking up iron more quickly than L-ferritin and storing iron in a more metabolically available form (8). In addition, H-type ferritin has biologic effects that are unrelated to iron binding, such as the capacity to inhibit cell growth and lymphocyte proliferation (9, 10).

In the current study we used bronchoalveolar lavage to compare concentrations of extracellular iron, L-type and H-type ferritin, and transferrin in the lower respiratory tract of stable patients with cystic fibrosis, with concentrations in healthy subjects as well as heavy cigarette smokers, who are known to accumulate extracellular ferritin-bound iron within their lungs (11, 12). Our findings demonstrate that the lungs of patients with stable cystic fibrosis contain high concentrations of extracellular ferritin-bound iron, with substantial amounts of H-type ferritin present.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Subjects

Eight adult subjects with cystic fibrosis were recruited from the Adult Cystic Fibrosis Clinic at the University of Kansas Medical Center. The diagnosis of cystic fibrosis was based on accepted clinical criteria including a typical clinical history, altered pulmonary function, and elevated levels of sodium and chloride in repeated sweat tests. All of the patients with cystic fibrosis had moderate to moderately severe disease as determined on the basis of FEV₁ and FVC (Table 1). Bronchoalveolar lavage was also performed in healthy adults as control subjects, including eight lifelong nonsmokers and eight heavy smokers (> 1 pack/d). Subject characterization is provided in Table 1. All subjects gave informed written consent and the protocol was approved by the institutional review boards for human subjects.

Collection and Treatment of Bronchoalveolar Fluid

Subjects received topical anesthesia with tetracaine (1%) and were premedicated with meperidine (25 to 50 mg). Fiberoptic bronchoscopy was performed transorally with bronchoalveolar lavage of the left upper lobe and right middle lobe, using three aliquots (50 ml) of normal saline. The return from each site was collected and samples from each site were analyzed separately. Grossly bloody samples were excluded from analysis. Samples were filtered through four layers of sterile gauze and the filtrate was centrifuged (2,000 rpm, 10 min, 10° C). The supernatant was stored at 70° C until analysis.

Cell Counts and Protein Content

The pellet was resuspended in normal saline and a cell count determined by using a hemacytometer. A cell differential was determined on cytospin preparations of cells stained with Diff-Quik (American Scientific Products, McGaw Park, IL). The protein content of recovered bronchoalveolar lavage fluid was determined using a bicinchoninic acid protein assay reagent (BCA kit; Pierce, Rockford, IL) with bovine serum albumin used as a standard.

Iron, L-ferritin, H-ferritin, and Transferrin Assays

The iron content of bronchoalveolar lavage fluid was determined by a method based on the use of ferrozine, as described by Fish (13). This method involves the use of an iron-releasing reagent (0.6 N HCl and 2.25% [wt/vol] KMnO₄), which releases iron complexed in biologic samples. This assay measures both free iron and iron bound to the iron-binding proteins ferritin and transferrin. The sensitivity of the method is 1 µg/dl. Concentrations of L- and H-ferritin in bronchoalveolar lavage fluid were determined by enzyme-linked immunosorbent assay (ELISA), using specific monoclonal antibodies as previously described (14, 15). The assay for L-ferritin was developed using recrystallized human liver ferritin and the sensitivity of this assay is 0.7 ng/ ml. The assay for H-ferritin was developed using human heart as the ferritin source as previously described, and this assay has a sensitivity of 0.7 ng/ml. Recrystallized human liver ferritin was completely nonreactive in the assay for H-ferritin at any concentration. The proportion of H-ferritin reacting in the L-ferritin assay was 5.2%, which represents the minimum L-subunit composition of H-rich ferritin. The transferrin content of lavage fluid was determined by a two-sided enzyme immunoassay, using a combination of polyclonal and monoclonal antibodies to human transferrin as previously described (16). The sensitivity of this assay is 10 µg/L.

Cell Culture

The human lung epithelium-like adenocarcinoma A549 cell line was obtained from the American Type Culture Collection (Rockville, MD). RPMI 1640 medium and fetal calf serum were purchased from JRH Biosciences (Lenexa, KS). A549 cells have multiple characteristics of Type 2 cells, including synthesis and secretion of surfactant, and have been used in prior studies on lung epithelium cells (17). Ferrous sulfate, penicillin, and streptomycin were obtained from Sigma (St. Louis, MO) and recombinant human tumor necrosis factor  (TNF-) was obtained from R&D Systems (Minneapolis, MN).

A549 cells (10⁶/ml) were allowed to adhere to 35-mm culture dishes in medium supplemented with 10% fetal calf serum and antibiotics. Nonadherent cells were removed and medium was refreshed. In some cultures medium was supplemented with ferrous sulfate (10 to 40 µM iron content) or iron (40 µM) together with human recombinant tumor necrosis factor  (5 to 20 ng). The concentration of FeSO₄ used in studies with TNF- (40 µM) was determined on the basis of preliminary studies indicating no significant cytotoxicity (lactate dehydrogenase [LDH] release) was caused by this concentration, whereas higher iron concentrations were associated with significant cytotoxicity. The studies with in vitro iron exposure were performed with FeSO₄ rather than ferric iron, on the basis of prior reports that non-transferrin-dependent iron accumulation required reduction of ferric iron to ferrous iron in some cell populations (18). A549 cells were incubated in unsupplemented medium or medium supplemented with iron or iron and cytokines for either 4 or 24 h at 37° C in 5% CO₂.

Statistics

Data are expressed as means ± SE. Differences between groups were analyzed by the Mann-Whitney rank sum test. In all tests, significance was identified at the p < 0.05 level.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Recovery of Cells and Protein

The recovery of total cells by bronchoalveolar lavage was significantly increased in patients with cystic fibrosis compared with both healthy nonsmokers and smokers (Table 2). The increase in total cells was largely attributable to an increased recovery of neutrophils, which constituted more than 40% of recovered cells in patients with cystic fibrosis. There were also significant increases in absolute numbers of macrophages and lymphocytes recovered by lavage in cystic fibrosis patients compared with recovery from nonsmokers and heavy smokers (data not shown). Erythrocytes comprised less than 1% of recovered cells in all lavage samples used in these studies. Protein concentrations in bronchoalveolar lavage fluid were significantly higher in patients with cystic fibrosis compared with either healthy nonsmokers or smokers (Table 2).

Tumor Necrosis Factor Concentrations

There were detectable concentrations of TNF- in bronchoalveolar lavage fluid recovered from all patients with cystic fibrosis, whereas detectable TNF- was not present in any healthy nonsmoker or smoker (Table 2). The range of TNF- concentrations present in lavage fluid from patients with cystic fibrosis was 14 to 59 pg/ml.

Iron, Isoferritin, and Transferrin Concentrations

There was not detectable iron in bronchoalveolar lavage (BAL) fluid recovered from nonsmokers (lower limit of detection, 1 µg/dl) whereas BAL fluid from seven of eight patients with cystic fibrosis had detectable iron, as did the BAL fluid of eight of eight heavy smokers (Figure 1). The mean iron concentration in bronchoalveolar lavage fluid from patients with cystic fibrosis was approximately four times as high as concentrations in heavy smokers. Concentrations of L-type ferritin in bronchoalveolar lavage fluid were increased in patients with cystic fibrosis and heavy smokers compared with nonsmokers, with differences between patients with cystic fibrosis and smokers not being significant (Figure 2). In both nonsmokers and heavy smokers, there were only small amounts of H-type ferritin in recovered bronchoalveolar lavage fluid (Figure 3). In contrast, bronchoalveolar lavage fluid recovered from patients with cystic fibrosis contained substantial concentrations of H-type ferritin. The increase in both L- and H-type ferritin within the lungs of patients with cystic fibrosis resulted in a marked increase in lung content of total ferritin (L- plus H-ferritin) in these patients (Figure 4). Concentrations of iron and H-ferritin in recovered lavage fluid were also increased in patients with cystic fibrosis compared with other groups if values were expressed as a ratio to lavage fluid total protein content (data not shown), although differences were less because there were higher concentrations of protein in lavage fluid recovered from the lungs of patients with cystic fibrosis compared with other groups.

View larger version (9K): [in this window] [in a new window]   Figure 1.   Concentrations of iron in bronchoalveolar lavage fluid recovered from patients with cystic fibrosis, heavy smokers, or healthy nonsmoking volunteers (n = 8 for each group). *p < 0.05 compared with heavy smokers; +p = 0.0005 compared with nonsmokers.

View larger version (10K): [in this window] [in a new window]   Figure 2.   Concentrations of L-type ferritin in bronchoalveolar lavage fluid recovered from patients with cystic fibrosis, heavy smokers, or healthy nonsmoking volunteers (n = 8 for each group). Each value represents the average concentration from two different locations in one subject. *p < 0.0005 compared with nonsmokers.

View larger version (10K): [in this window] [in a new window]   Figure 3.   Concentrations of H-type ferritin in bronchoalveolar lavage fluid recovered from patients with cystic fibrosis, heavy smokers, or healthy nonsmoking volunteers (n = 8 for each group). Each value represents the average concentration from two different locations in one subject. *p < 0.0005 compared with heavy smokers.

View larger version (11K): [in this window] [in a new window]   Figure 4.   Concentrations of total ferritin (H-type plus L-type) in bronchoalveolar lavage fluid recovered from patients with cystic fibrosis, heavy smokers, or healthy nonsmoking volunteers (n = 8 for each group). Each value represents the average concentration from two different locations in one subject. *p < 0.005 compared with heavy smokers; +p < 0.005 compared with nonsmokers.

Analysis of transferrin concentrations in bronchoalveolar lavage fluid demonstrated values that were similar to concentrations in nonsmokers and heavy smokers (Figure 5). In contrast to marked variability in iron and isoferritin concentrations in patients with cystic fibrosis, concentrations of transferrin in bronchoalveolar lavage fluid were found to be within a relatively narrow range. The concentrations of transferrin found in the lungs of nonsmokers and smokers in this study are consistent with our prior findings (10, 11).

View larger version (10K): [in this window] [in a new window]   Figure 5.   Concentrations of transferrin in bronchoalveolar lavage fluid recovered from patients with cystic fibrosis, heavy smokers, or healthy nonsmoking volunteers (n = 8 for each group). Each value represents the average concentration from two different locations in one subject. Differences between groups were not significant.

Isoferritin Accumulation in A549 Cells Induced by Iron and TNF-

Incubation of A549 cells in RPMI 1640 with 10% fetal calf serum was not associated with significant changes in cell content of H- or L-type ferritin (data not shown). Incubation of A549 cells in medium supplemented with iron (10-40 µM) for 24 h was associated with a dose-dependent accumulation of L-type ferritin with only small amounts of H-type ferritin accumulating (Table 3). Although there was significant synthesis of ferritin, particularly L-type ferritin, by 4 h, accumulation of ferritin occurred predominantly between 4 and 24 h of treatment. In vitro exposure of A549 cells to concentrations of FeSO₄ higher than 40 µM was associated with significant cytotoxicity as indicated by decreased ferritin synthesis and increased release of lactate dehydrogenase (data not shown). Exposure of A549 cells to both FeSO₄ and tumor necrosis factor  (5 to 20 ng/ml) was associated with substantially greater accumulation of H-type ferritin, as well as L-type ferritin, with a dose- dependent effect of tumor necrosis factor  on the accumulation of H-ferritin (Figure 6).

View larger version (15K): [in this window] [in a new window]   Figure 6.   Concentrations of H-type ferritin accumulating in A549 cells exposed to non-transferrin-bound iron (FeSO4, 40 µM) and increasing doses of TNF- for 24 h. Results represent the mean of four different experiments. *p < 0.001 compared with no iron.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The current study demonstrates that concentrations of extracellular iron and ferritin are substantially increased within the lower respiratory tract of patients with stable cystic fibrosis. Compared with the increases in extracellular iron present in the lungs of heavy cigarette smokers, increases in patients with stable cystic fibrosis were approximately fourfold higher (11, 12). Concentrations of extracellular ferritin in the lungs of patients with cystic fibrosis were threefold higher than in heavy smokers and more than 70-fold greater than concentrations in healthy nonsmokers. In addition, the large amounts of H-type ferritin together with L-type ferritin present in the lower respiratory tract of patients with cystic fibrosis differed from that present in the lungs of either nonsmokers or smokers, which was almost completely L-type ferritin. This accumulation of H-type ferritin within the lower respiratory tract of patients with cystic fibrosis may be of clinical significance because H-ferritin has known biologic effects unrelated to iron binding, including effects on cell proliferation and immune function (8, 9).

The current study extends our previous report of increased iron and L-type ferritin concentrations in sputum produced by patients with cystic fibrosis (19). Our finding of increased extracellular iron and L-ferritin concentrations in bronchoalveoalar lavage fluid recovered from heavy cigarette smokers is similar to our previous findings (10, 11). Prior studies indicate that increased extracellular iron is present in the alveolar spaces of patients with acute respiratory distress syndrome, with extracellular iron bound predominantly to transferrin (4). The increased alveolar concentrations of transferrin in patients with acute respiratory distress syndrome reduces the availability of catalytic iron and may limit iron-catalyzed oxidant injury. In patients with cystic fibrosis, however, there were greater increases in lung iron content than have been reported in patients with acute respiratory distress syndrome or in cigarette smokers, together with no increase in lung content of transferrin. In addition, the transferrin present within the lungs of patients with cystic fibrosis may not effectively sequester iron, as prior studies indicate that both transferrin and lactoferrin can be cleaved within the lungs of patients with cystic fibrosis by Pseudomonas-derived proteases (6, 7). Although the increased availability of extracellular iron within the lungs of patients with cystic fibrosis would not in itself enhance generation of reactive oxygen species, it would promote the generation of highly toxic hydroxyl radicals from less reactive oxygen species present within the lungs (1, 5).

The concentrations of L-ferritin present in cystic fibrosis-affected lungs were greater than concentrations present in heavy cigarette smokers, although the differences were not significant. However, because there were also large amounts of intrapulmonary H-type ferritin in patients with cystic fibrosis, but not in smokers, the total amount of intrapulmonary ferritin present in patients with cystic fibrosis was substantially higher than in cigarette smokers. The potential of extracellular ferritin-bound iron to be toxic to lungs of patients with cystic fibrosis is supported by studies demonstrating that intrapulmonary instillation of large amounts of L-ferritin-bound iron in experimental animals induces acute lung injury (20). Ferritin-induced acute lung injury in this model is mediated by iron release, as instillation of iron-free ferritin did not cause injury. Prior in vitro studies suggest that H-type ferritin may be more effective in inhibiting iron-catalyzed lipid peroxidation than L-type ferritin and, therefore, may be more protective (21). However, the capacity of iron bound to H-type ferritin to promote oxidative lung injury in vivo compared with L-type ferritin is uncertain.

The source of iron that accumulates within the lungs of patients with cystic fibrosis is uncertain. In an animal model of acute airway inflammation induced by exposure to ozone, the airway content of iron rapidly increased, presumably owing to an influx of serum-derived iron (22). Because airway inflammation is a feature of patients with cystic fibrosis, there may be an influx of serum-derived iron into airway structures of these patients. Prior studies in rats demonstrated that chronic pulmonary infection was associated with accumulation of L-ferritin, and presumably iron, by lung macrophage populations (23). Chronic lung infection and inflammation in patients with cystic fibrosis, such as with Pseudomonas species, may promote accumulation of iron and ferritin in lung macrophages, which then could release ferritin-bound iron as a result of cell injury or death. Some of the extracellular iron present in the lower respiratory tract of patients with cystic fibrosis may be derived from hemoglobin released from extravasated erythrocytes, although less than 1% of recovered cells were erythrocytes. However, an increase in erythrocytes would also not explain the accumulation of ferritin in the lungs of patients with cystic fibrosis, because erythrocytes contain extremely small amounts of ferritin (24).

L-type ferritin effectively stores large amounts of iron and is preferentially expressed in tissues after iron loading (25). In contrast, H-ferritin has greater ferroxidase activity than L-ferritin and stores iron in a more metabolically available form (26). In additon, H-type ferritin has additional biologic effects that are unrelated to iron storage, including inhibition of T lymphocyte proliferation, suppression of differentiation of B lymphocytes, and inhibition of other types of cell proliferation (9, 10). Increased concentrations of H-ferritin in the lungs of patients with cystic fibrosis could, therefore, inhibit lung normal lymphocyte function and potentially impair proliferation of other cell populations involved in airway repair.

We studied only patients with stable cystic fibrosis; however, there was substantial inflammation in the lower respiratory tract of these patients as indicated by markedly increased recovery of neutrophils by bronchoalveolar lavage. This finding is consistent with a prior study indicating increased neutrophil recovery by bronchoalveolar lavage in patients with stable cystic fibrosis (4). Neutrophil accumulation in the lungs of patients with cystic fibrosis could be a source of reactive oxygen species capable of promoting oxidative lung injury. Neutrophils are also a source of the iron-binding protein lactoferrin; however, as previously noted in studies by Britigan and colleagues, the capacity of lactoferrin to sequester iron within the lungs of patients with cystic fibrosis is impaired (6).

Our finding of increased concentrations of TNF- within the lungs of patients with stable cystic fibrosis is consistent with a prior report (27). We postulated that TNF- within the lungs of patients with cystic fibrosis could promote synthesis of H-type ferritin, as TNF- induces iron-independent accumulation of mRNA for H-type ferritin in cells (28). To assess this hypothesis, we performed in vitro studies that demonstrated that exposure of alveolar epithelium-like cells (A549) to iron alone induced synthesis of predominantly L-type ferritin; however, exposure to both iron and TNF- induced the synthesis of large amounts of H-type ferritin together with L-type ferritin. The concentrations of iron that we used in vitro are consistent with the concentrations present in the lungs of patients with cystic fibrosis. The mean iron concentration in bronchoalveolar lavage fluid recovered from patients was approximately 8 µM, and we demonstrated significantly increased ferritin synthesis in vitro with iron concentrations as low as 10 µM. Because bronchoalveolar lavage fluid represents a marked dilution of alveolar constituents, the concentrations of iron that we used in vitro may be lower than alveolar iron concentrations present in vivo. The accumulation of both iron and TNF- within the lower respiratory tract, therefore, could promote synthesis of H-type ferritin by lung epithelial cells in patients with cystic fibrosis with subsequent release of ferritin stores as a consequence of cell injury or death.

There are several potential clinical implications of our findings. First, increased respiratory tract content of ferritin-bound iron may promote oxidative injury to lung structures in patients with cystic fibrosis. Second, increased availability of iron can promote bacterial growth, in particular growth of Pseudomonas species (29). Accumulation of iron within the lungs of patients with cystic fibrosis may, therefore, promote airway infection with Pseudomonas species. Third, the accumulation of H-ferritin in cystic fibrosis-affected lungs may impair local humoral and cell-mediated immune function (9). H-ferritin also inhibits cell growth and may, therefore, impair cell proliferation necessary for lung repair (10).

In summary, these studies demonstrate the presence, in the lower respiratory tract of patients with stable cystic fibrosis, increased concentrations of ferritin-bound iron that could promote oxidative lung injury. The concentrations of extracellular iron and ferritin within the lower respiratory tract of patients with cystic fibrosis are substantially greater than concentrations present in heavy cigarette smokers. Our in vitro studies suggest that TNF- present within the lungs of patients with cystic fibrosis could promote the synthesis of H-type ferritin. The accumulation of intrapulmonary H-type ferritin in patients with cystic fibrosis may impair lung lymphocyte function and inhibit cell growth involved in lung repair in these patients.