|Home | About | Journals | Submit | Contact Us | Français|
Alpha-1-antitrypsin (A1AT) deficiency is characterized by increased neutrophil elastase (NE) activity and oxidative stress in the lung. We hypothesized that NE exposure generates reactive oxygen species by increasing lung non-heme iron. To test this hypothesis, we measured bronchoalveolar lavage (BAL) iron and ferritin levels, using inductively coupled plasma (ICP) optical emission spectroscopy and an ELISA respectively, in A1AT-deficient patients and healthy subjects. To confirm the role of NE in regulating lung iron homeostasis, we administered intratracheally NE or control buffer to rats and measured BAL and lung iron and ferritin. Our results demonstrated that A1AT-deficient patients and rats post-elastase exposure have elevated levels of iron and ferritin in the BAL. To investigate the mechanism of NE-induced increased iron levels, we exposed normal human airway epithelial cells to either NE or control vehicle in the presence or absence of ferritin, and quantified intracellular iron uptake using calcein fluorescence and ICP mass spectroscopy. We also tested whether NE degraded ferritin in vitro using ELISA and western analysis. We demonstrated in vitro that NE increased intracellular non-heme iron levels and degraded ferritin. Our results suggest that NE digests ferritin increasing the extracellular iron pool available for cellular uptake.
Inhaled exposures (e.g. microbials, particles, and fibers) can mobilize host iron, disrupt the normal homeostasis of this metal in the lower respiratory tract, and increase its availability to participate in an oxidative stress. Subsequently, host protective mechanisms against iron-induced oxidative injury are necessary for lung health. These focus on host re-acquisition of its own metal with subsequent transport and storage to a catalytically less reactive state1. There are several pathways of iron acquisition employed by living systems. Siderophores and ferrireductases are prominent among these 2. However, a third approach has been described with the recognition that microbes can utilize proteases in the acquisition of iron. Proteases may cleave iron-transport and -storage proteins allowing use of the metal by the microbe 3, 4. Reflecting potential interactions between metal availability and proteases, increased iron concentrations can impact expression and activity of collagenase 5, elastase 6, alkaline proteinase7, and metalloproteases 8.
Animal and human studies similarly suggest a participation of proteases in iron homeostasis. The intratracheal instillation of a single dose of neutrophil elastase in an animal model increased lung iron concentrations approximately 18 months later 9. Iron homeostasis is disrupted among cystic fibrosis patients in whom airway elastase content and activity is excessive; CF patients have elevated iron and ferritin concentrations in both the sputum and bronchoalveolar lavage (BAL) 10, 11. Finally, alpha-1-antitrypsin deficient patients have a propensity to develop cirrhosis which may be related to abnormal iron homeostasis 12, 13; this observation raises the possibility that unopposed neutrophil elastase activity alters iron homeostasis in both the lungs and systemically among A1AT-deficient patients.
We tested the hypothesis that a disruption of the normal balance between proteases and anti-proteases can be associated with an increased availability of iron in the lower respiratory tract of A1AT-deficient patients, in the lower respiratory tract of a rat model exposed to intratracheal neutrophil elastase, and in normal human bronchial epithelial cells exposed to neutrophil elastase. We suggest that among the mechanisms of increased iron availability would be a proteolytic cleavage of iron-transport and -storage proteins with subsequent increase in non-heme iron.
The characterization of A1AT-deficient and control subjects is reported in Table 1. Alpha-1-antitrypsin patients (Beaumont Hospital, Dublin, Ireland) had the diagnosis of A1AT deficiency and PiZZ phenotype confirmed by nephelometry and isoelectric focusing. The screening procedures for each subject included a history and physical examination, routine hematologic and biochemical tests, and pulmonary function tests. Fiberoptic bronchoscopy and bronchoalveolar lavage (BAL) for A1AT-deficient patients was performed following patient consent and according to standardized guidelines as approved by the Beaumont Hospital Review Board committee.
Healthy non-smoking volunteers (control subjects) underwent fiberoptic bronchoscopy with BAL at the National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency (EPA) Research Triangle Park, NC; the protocol and consent form were approved by the University Of North Carolina School Of Medicine Committee on the Protection of the Rights of Human Subjects. After adequate sedation and analgesia, the fiberoptic bronchoscope was wedged into a segment of the lingula or the right middle lobe and lavaged with five sequential 50 mL aliquots of sterile saline which were infused quickly with no dwell time between infusion and aspiration. BAL fluid from both lobes was combined and the percentage of recovered volume measured. Cells were separated from the BAL fluid at 2000 rpm, 15 min. BAL was separated and stored at −70° C.
Lavage (0.5 mL) was added to 0.5 mL 6 N HCl/20% trichoroacetic acid. This was hydrolyzed at 70° C for 18 hours, and centrifuged at 20,000 × g for ten minutes. Metal concentrations were determined in the supernatants using inductively coupled plasma optical emission spectroscopy (ICPOES; Model Optima 4300D, Perkin Elmer, Norwalk, CT) operated at a wavelength of 238.204 nm14. Ferritin concentrations in BAL supernatant were quantitated by enzyme immunoassay according to the manufacturer's instructions (Microgenics, Concord, CA)14.
The Environmental Protection Agency's Institutional Animal Care and Use Committee reviewed and approved all procedures on animals. After anesthesia with 2 to 5% halothane (Aldrich Chemicals, Milwaukee, WI), sixty-day old (250 g) male Sprague-Dawley rats (total n = 96) were intratracheally instilled with either 0.5 mL buffer or 50 μg human neutrophil elastase in 0.5 mL buffer. At 7, 14, and 28 days after exposure, rats were again anesthetized with halothane and euthanized. Specimens acquired included tracheal lavage with saline (35 mL/kg of body weight; n= 8/exposure/time point) and inflation-fixed lung (10% formalin; n= 4/exposure/time point).
BAL protein and lactate dehydrogenase (LDH) were evaluated as measures of NE-induced lung injury. Lavage protein was determined using the Pierce Coomassie Plus Protein Assay Reagent (Pierce, Rockford, IL). Bovine serum albumin served as the standard. Lavage LDH concentration was measured using a commercially prepared kit (Sigma). Both assays were modified for automated measurement (Cobas Fara II centrifugal analyzer).
To quantify the inflammatory cell influx in the lung following animal exposures, a modified Wright's stain (Diff-Quick stain; American Scientific Products, McGaw Park, IL) was used and cell differentials were expressed as the percentage of total cells recovered. After hydrolysis in an equivalent volume of 6 N HCl/20% trichloroacetic acid, BAL non-heme iron concentrations were quantified using ICPOES'; Model Optima 4300D, (Perkin Elmer, Norwalk, CT) operated at a wavelength of 238.204 nm. Lavage ferritin concentrations were measured using an enzyme immunoassay (Microgenics Corporation). Non-heme iron in resected lung tissue was measured after adding 10.0 mL 3 N HCl/10% trichloroacetic acid/gram tissue, heating to 70° C for 18 hours, and centrifuging at 20,000 g for ten minutes. Metal concentrations were determined in the supernatants using ICPOES. Lavage cells (at 1.0 × 106/1.0 mL) were cytocentrifuged (0.2 mL) onto slides and stained for iron using Perl's Prussian Blue for iron. Lungs were inflation-fixed with 10% formalin for 24 hours, paraffin embedded, and 4 μm sections prepared for histology and immunohistochemistry. Iron was detected using Perls’ Prussian blue stain. Immunohistochemical staining for ferritin was performed following blocking of endogenous peroxidase activity with 30% hydrogen peroxide in 30 mL methanol for 8 minutes. After treatment with Cyto Q Background Buster (Innovex Biosciences) for 10 min at room temperature, slides were incubated with the primary antibody (rabbit anti-human L ferritin antibody; 1:100 dilution in 1% BSA in PBS) (Dako, Carpenteria, CA) at 37° C, 45 min. Slides were then incubated with biotinylated anti-rabbit IgG antibody from Stat Q Staining System (Innovex Biosciences) for 10 min at room temperature, washed in PBS, and labeled with peroxidase enzyme label from Stat Q Staining System (Innovex Biosciences). Slides were developed with 3,3' diaminobenzidine tetrahydrochloride for 3 min at room temperature and counterstained with hematoxylin 15.
Primary normal human bronchial epithelial (NHBE) cells were harvested from human tracheobronchial tissues of donors obtained from the Lung Transplant Program and the Department of Pathology, Duke University Medical Center. The protocol was approved by the Institutional Review Board for Clinical Investigations, Duke University Medical Center. After initial harvest and expansion, cells were cultured submerged on 12 well plastic plates in a small airway basal medium (SABM; Clonetics/Lonza, Walkersville, MD): Dulbecco's modified Eagles Medium (DMEM) [1:1 ratio; Invitrogen, Carlsbad, CA] supplemented with twelve factors: insulin (4 μg/mL; Sigma, St. Louis, MO), holo-transferrin (5 μg/mL; Sigma), EGF (0.5 ng/mL; BD Biosciences, San Jose, CA), dexamethasone (0.1 μM; Sigma), cholera toxin (20 ng/mL; List Biological Laboratories, Inc., Campbell, CA), bovine hypothalamic extract (1:500; Pel-Freeze Biologicals, Rogers, AR), nystatin (20 U/mL; Sigma), gentamicin (50 μg/mL; Invitrogen), amphotericin B (250 ng/ml; Invitrogen), HEPES (1.5mM; Invitrogen), T3 (6.5 ng/mL; Sigma), and BSA (0.5 mg/mL; Sigma). NHBE cells were grown to confluence for all studies.
The assay for intracellular iron concentration is based on the fluorescent and metal chelating properties of calcein- acetoxymethyl ester (Calcein-AM; Invitrogen). Calcein-AM is a non-fluorescent lipophilic ester that penetrates the cell membrane and is cleaved by cytosolic esterases resulting in intracellular capture of a fluorescent alcohol with the capacity to chelate catalytically active iron. Upon iron chelation, the calcein green fluorescence is quenched, and this property can be used to quantitate the intracellular labile iron pool 16. Fluorescence decreases as the amount of non-heme iron increases. Cells were loaded with calcein- AM (0.5 μM, 1h) in the media described above. Following loading, the media was removed and cells were treated in one of five conditions in an iron-free media (RPMI; Invitrogen): control, neutrophil elastase (NE, 100 nM; 875 U/mg protein, Elastin Products, Co., Owensville, MO) alone, ferric ammonium citrate (FAC, positive control, 200 μM; Sigma), ferritin alone (500 ng/mL; horse spleen ferritin, Sigma, Cat # F4503), and ferritin plus NE. Cells were incubated for 4h at 37°C. After incubation, media was removed and cells were rinsed twice with PBS. Then 500 μL of PBS (Invitrogen) were added to each well and the plate was analyzed for fluorescence in a fluorescent plate reader (Safire II, TECAN USA, Research Triangle Park, NC). The data is expressed as a percentage decrease in absorbance compared to control treated cells.
Horse spleen apo-ferritin (Sigma, St. Louis) was loaded with 57FeCl2 as previously described 17. Briefly, an apo-ferritin solution (0.25 μM, 2.5 mL) was prepared in 0.1 M Mops buffer pH 7.4, 0.05 M NaCl, and iron ions were added from a 0.010 M 57FeCl2 stock solution to a final concentration of 50 μM (~200 Fe molecules/ferritin molecule). The iron loading into ferritin was monitored spectrophotometrically at 350 nm. This process was repeated 9 times to achieve a theoretical loading of 2000 Fe molecules/ferritin molecule. After the loading of iron, unbound iron was removed using Amicon Ultra centrifugal filter devices with a 30,000 molecular weight cutoff. 57Fe-loaded ferritin was diluted in 0.1 M MOPS solution and protein concentrations were determined using the BioRad assay (BioRad, Hercules, CA).
Cells were treated in one of four conditions in an iron free media (RPMI): control, NE (100 nM) alone, 57Fe- loaded ferritin (100 ng/mL), and 57Fe- loaded ferritin plus NE. Cells were incubated for 4h at 37°C, media was removed, and cells were collected using 10% Tricarboxylic acid/HCl. 57Fe concentration was quantified on a Perkin Elmer Elan 6000 inductively coupled plasma mass spectrometer (ICPMS) with a Scott cross-flow nebulizer to minimize oxide formation. The settings were: power = 700 W, nebulizer flow = 1.3 l/min, lens = 3 V (static). Co was added at a level of 50 ppb as an internal standard using an online internal standard addition kit (Perkin Elmer, N0690673).
Degradation of ferritin by neutrophil elastase was evaluated by ELISA and western analysis. For the ELISA, 50 ng/ml ferritin (type V from human spleen; Sigma, Cat#F6879, Lot 078K1463)/ mL Dulbecco's PBS (Invitrogen) was incubated for 1h at 37°C with neutrophil elastase (0 to 250 nM). At the end of incubation, the sample was frozen at −80° C. After thawing, ferritin in the sample was measured using an ELISA (Alpha Diagnostics International, San Antonio, TX). This ELISA utilizes two different mouse monoclonal antibodies to capture and detect intact human ferritin but the specific epitopes of each antibody are unknown. For the western analysis, horse spleen ferritin (Sigma, Cat # F4503, Lot 098K7009) was vortexed vigorously for 30 seconds prior to dilution to 1ug/ml in Dulbecco's PBS (+Ca/ +Mg). Neutrophil elastase (1 μM) or control vehicle was added to 20ul of this diluted ferritin and incubated for 1 h, at 37°C with shaking. SDS-PAGE loading buffer was added to the incubation rxeactions and mixed but not boiled, and all samples were separated on a 4-15% Tris-HCl SDS-polyacrylamide gel (Bio-Rad, Hercules, CA). After transfer to nitrocellulose (Bio-Rad), membranes were blocked with 5% milk in TBS-Tween 20 (0.01%), 1h, room temperature. Ferritin was detected by incubating overnight with rabbit anti-horse spleen ferritin antibody (1:1000 dilution; Sigma, Cat # F6136, Lot 019K4787) in 5% BSA in TBS-Tween 20. After washing, membranes were incubated with peroxidase-conjugated goat anti-rabbit IgG used as the secondary antibody (1:2000 dilution; Cell Signaling, Danvers, MA) and developed with ECL-Plus according to the manufacturer's instructions (GE Healthcare Biosciences Corp, Piscataway, NJ). Autoradiographs were scanned and densitometric analysis was performed with ImageQuant TL software (GE Healthcare Bio-sciences Corp.).
A1AT-deficient patients have increased oxidative stress that has been attributed to both inflammation and loss of alpha-1-antitrypsin activity 18. We hypothesized that this oxidative stress may be due to increased levels of catalytically active iron in the lung. Using ICP optical emission spectroscopy to quantitate iron, and an ELISA for ferritin, we observed that A1AT-deficient patients have increased concentrations of both iron and ferritin in the BAL compared to normal control subjects (Figure 1).
To determine whether NE exposure directly increased airway non-heme iron, we exposed Sprague-Dawley rats to intratracheal neutrophil elastase and evaluated whether, over 7 to 28 days following exposure, there were changes in BAL non-heme iron and ferritin. Relative to instillation of buffer only, lavage concentrations of non-heme iron were increased 7 days following elastase (Figure 2B). Concentrations of BAL non-heme iron remained elevated throughout the duration of study. Similarly, ferritin concentrations in the lavage were increased (Figure 2A). Lung non-heme iron concentrations almost doubled after elastase exposure (Figure 2C); levels remained elevated at 28 days after instillation. Lavage concentrations of both protein and LDH were significantly elevated at 7 days after exposure to elastase (Figures 2D and 2E). These results reflect minor injury which resolved by 14 and 28 days post instillation.
Iron-laden macrophages were detected in NE-treated rats starting at 7 days and these persisted at 28 days post-NE instillation (Figures 3A-D); no sideromacrophages were observed in those animals exposed to buffer. Lung section immunohistochemistry similarly revealed increased ferritin staining in macrophages after elastase exposure (Figures 3E and F). There was no evidence of pulmonary hemorrhage/ alveolar red blood cells at time points > 7 days (data not shown).
Importantly, BAL cell differentials revealed an inflammatory response to NE at 7 days. Neutrophils were significantly increased at this time point (21.6 ± 9.2% and 1.2 ± 0.8 of total cell count after NE and buffer respectively). By 14 days, this inflammatory influx had resolved. These results were consistent with the findings from A1AT-deficient patients suggesting that unopposed airway NE activity could induce increased airway iron and ferritin.
We have previously shown that NE increases reactive oxygen species (ROS) in A549 lung cancer cells and exposure to desferrioxamine, an iron chelator, blocked this ROS production19. To evaluate whether NE increased cell iron levels, we loaded cells with the fluorescent probe, calcein-AM, which is modified by cytosolic esterases resulting in an intracellular fluorescent probe. Calcein chelates non-heme iron resulting in quenching of fluorescence and has been used to quantify the labile iron pool in the cytosol (9). NHBE cells exposed to NE had increased non-heme iron concentrations as detected by decreased calcein fluorescence (Figure 4). A positive control, ferric ammonium citrate, also caused significant quenching of calcein fluorescence consistent with increased intracellular non-heme iron. The response to NE required the presence of both the protease and the iron-containing protein, ferritin, in the media; NE alone in iron-free, protein-free medium had no effect on intracellular non-heme iron content. Therefore, we hypothesized that NE degrades iron-containing proteins in the extracellular milieu resulting in the release of iron and its uptake into the cell. To test this hypothesis, we exposed NHBE to 57Fe-loaded ferritin in the presence or absence of NE (Figure 5). Although in the presence of 57Fe-loaded ferritin alone there was a small increase in intracellular 57Fe, the increase in intracellular 57Fe was significantly greater in the presence of NE.
To test whether ferritin could function as a target for the proteolytic action of NE, in vitro incubations of NE and ferritin were conducted. A human ferritin ELISA, based on two mouse monoclonal antibodies, was used to measure intact ferritin concentrations. Significant decrements in intact human spleen ferritin concentrations recognized by ELISA were noted after one hour with 150 or 250 nM neutrophil elastase, supporting ferritin degradation (Figure 6A). To confirm that NE degrades ferritin, potentially releasing the metal, degradation of horse spleen ferritin was detected by western analysis using a rabbit affinity purified anti-horse spleen ferritin antibody. Western analysis results confirm ELISA results and demonstrate that ferritin is digested by NE from a large multimer (~250 kD) to at least 2 degradation products (~30 and 25 kD) following NE treatment for 1 h (Figures 6B and 6C). This was associated with a decrease in the density of the multimer band. Together, these results support a role for NE to act in the extracellular milieu to degrade iron-containing proteins, releasing iron and potentially increasing concentrations of non-heme iron available for cellular uptake.
Our results confirm that A1AT-deficient patients have elevated iron and ferritin levels in the BAL. These results are similar to those found in CF patients 10, 20. Interestingly, both patient groups are characterized by excessive NE activity and protease capacity in the airway surface lining fluid. These observations raise the possibility that there is a link between high levels of airway protease and high concentrations of non-heme iron in the airway. To test this relationship directly, we exposed rats to intratracheal NE and found persistent elevations in iron concentrations of the lavage and lung.
We then evaluated in vitro for a mechanism by which NE protease activity could increase available iron levels. Our experiments demonstrate that NE degrades iron-containing proteins in the extracellular fluid resulting in release of non-heme iron. The action of NE to release protein-bound iron is not unique to NE. Other proteases, serine and non-serine proteases, including pseudomonas elastase, porcine pancreatic elastase, and trypsin can degrade transferrin 3 and lactoferrin 4 releasing iron.
This non-heme iron is likely taken up by cells resident in the lower respiratory tract including epithelial cells and macrophages. Cell import is predicted to be both transferrin and non-transferrin dependent 21. This transport likely results in an increase in the intracellular labile iron pool, a cytosolic domain of non-heme, catalytically active iron 22. Increased iron in the labile iron pool is sensed by iron regulatory proteins that upregulate ferritin translation. Ferritin is a major intracellular storage protein that sequesters iron, diminishing catalytic activity 21. However, ferritin can be released from macrophages into the extracellular space 23. This mechanism may explain the increased level of ferritin in the lining fluid and sputum of patients with A1AT-deficiency and CF. Such sequestration of iron by ferritin in the lower respiratory tract does not absolutely block the potential to mount an oxidative stress as ferritin can present an additional target for proteases or reductants to release iron 23.
NE has been reported to increase reactive oxygen species by several mechanisms including increased mitochondrial oxidant release 24, or activation of oxidoreductases such as dual oxidase (DUOX) 25 and NADP(H) quinone oxidoreductase 1 (NQO1) 26 . In this report we present evidence that NE also increases iron concentrations in cells and this will contribute to oxidative stress. Non-heme iron is a catalyst for the generation of reactive oxygen species via the Haber-Weiss reaction including superoxide, hydrogen peroxide and hydroxyl radical 21.
Iron is essential for life but must be strictly sequestered in heme or iron-binding proteins to prevent host injury. Iron participates in numerous lung injuries including those after exposure to pollutants including silica dust, coal dust, oil fly ash, welding dust, tobacco smoke, and diesel particles 27. In addition, available iron can participate in lung disease 27 such as acute respiratory distress syndrome, post- cardiopulmonary bypass, pulmonary hemorrhage, and pulmonary alveolar proteinosis 14. Importantly, increased airway iron promotes infection with microorganisms including bacteria, protozoans, and fungus. Iron is scavenged from the host using siderophores/receptors and ferrireductases present on microbial membranes. Host sources of iron, such as transferrin and lactoferrin, can then be utilized as sources of metal required for microbial proliferation 2. Therefore, host responses to increased airway non-heme iron are critical to protect against oxidative stress and to prevent infection. Our report suggests NE is another participant in the host battle for essential iron. Elastase in the lower respiratory tract may degrade iron containing proteins including ferritin, lactoferrin and transferrin, releasing catalytically active iron. This results in an alteration of iron homeostasis, accumulation of metal required for microbial proliferation and increased risk of infection. Furthermore the release of the iron by degradation of the storage proteins promotes extracellular-to-intracellular oxidative stress in the airway epithelium. Importantly, our findings suggest a new therapeutic target in the treatment of A1AT deficiency.
Financial Support: HL 082504 (JAV), HL 081763 (BMF)