Cadmium-mediated toxicity of lung epithelia is enhanced through NF-κB-mediated transcriptional activation of the human zinc transporter ZIP8

Abstract

Cadmium (Cd), a toxic heavy metal and carcinogen that is abundantly present in cigarette smoke, is a cause of smoking-induced lung disease. SLC39A8 (ZIP8), a zinc transporter, is a major portal for Cd uptake into cells. We have recently identified that ZIP8 expression is under the transcriptional control of the NF-κB pathway. On the basis of this, we hypothesized that cigarette-smoke induced inflammation would increase ZIP8 expression in lung epithelia, thereby enhancing Cd uptake and cell toxicity. Herein we report that ZIP8 is a central mediator of Cd-mediated toxicity. TNF-α treatment of primary human lung epithelia and A549 cells induced ZIP8 expression, resulting in significantly higher cell death attributable to both apoptosis and necrosis following Cd exposure. Inhibition of the NF-κB pathway and ZIP8 expression significantly reduced cell toxicity. Zinc (Zn), a known cytoprotectant, prevented Cd-mediated cell toxicity via ZIP8 uptake. Consistent with cell culture findings, a significant increase in ZIP8 mRNA and protein expression was observed in the lung of chronic smokers compared with nonsmokers. From these studies, we conclude that ZIP8 expression is induced in lung epithelia in an NF-κB-dependent manner, thereby resulting in increased cell death in the presence of Cd. From this we contend that ZIP8 plays a critical role at the interface between micronutrient (Zn) metabolism and toxic metal exposure (Cd) in the lung microenvironment following cigarette smoke exposure. Furthermore, dietary Zn intake, or a lack thereof, may be a contributing factor in smoking-induced lung disease.

chronic obstructive pulmonary disease (COPD) is a major cause of morbidity and mortality in the United States and worldwide, affecting 4–6% of all people over the age of 45 yr. The prevalence of COPD is expected to rise, and the World Health Organization predicts that it will be the third leading cause of death by 2020 (35, 39). COPD is characterized by progressive airflow obstruction, which is not fully reversible, resulting from an abnormal inflammatory response in the lungs to noxious gases and particles (11). Cigarette smoking is the single most important risk factor for developing COPD in men and women in the United States (23). Individuals that develop COPD also have a much higher likelihood of developing lung cancer (36, 53). Cadmium (Cd), a major component of cigarette smoke, contributes to both disorders. Each cigarette contains approximately 2 μg of Cd, of which, 10% is directly transferred to lung tissue following inhalation of primary cigarette smoke (37). Of the Cd present in primary inhaled cigarette smoke, nearly 50% is absorbed from the lungs into the systemic circulation during active smoking (24, 47). Cd, relative to other components in cigarette smoke, is unique with a biological half-life in humans of 15–20 yr (25, 52). Therefore, it is not surprising that smokers typically carry Cd blood and whole body burdens more than double those of nonsmokers. Importantly, it remains unclear how Cd enters the lung, thereby causing pathological manifestations that lead to COPD and lung cancer.

Analysis of over 6,700 subjects derived from the third National Health and Nutrition Examination Survey (NHANES III) revealed that adult smokers with low zinc (Zn) intakes were significantly more likely to have a higher Cd burden and increased risk of developing COPD (31). On the basis of these findings, the investigators postulated that a dynamic interplay exists between Zn and Cd; however, the mechanism(s) by which this occurs “warrant further study”. Importantly, malnutrition is common in patients with COPD (45), and small case-controlled studies have shown an association between Zn deficiency and COPD (22, 27) although this has never been thoroughly investigated. Relative to Zn metabolism and COPD pathogenesis, we propose a novel role for the mammalian Zn transporter, ZIP8, as a key mediator of Cd-mediated toxicity in lung epithelia, target cells in COPD pathogenesis. Recently ZIP8 was identified as a primary transporter of Cd, in addition to Zn (12, 20, 32). Our group reported that ZIP8 is unique, relative to all other ZIP family members, in that ZIP8 expression in lung epithelia is induced by proinflammatory cytokines (6) via transcriptional upregulation by NF-κB, a signal transduction pathway that is persistently activated in the lung of chronic smokers (49). Taken together, we contend that chronic cigarette exposure, which creates an inflammatory microenvironment in the lung (35, 39, 45), upregulates ZIP8 expression in lung epithelia, thereby increasing the extent of Cd uptake, tissue burden, and lung-related pathology. On the basis of the work of Lin (31), we predict that Zn, a known cytoprotectant (2), will also influence ZIP8-mediated Cd uptake such that nutritional Zn deficiency will augment lung injury, whereas Zn sufficiency will minimize injury by preventing Cd uptake into lung tissue through ZIP8.

In the present study, using primary human lung epithelia and a related cell line, we determined to what extent Zn metabolism impacts Cd-induced toxicity. Accordingly, we sought to determine whether TNF-α, a cytokine present in the lung of chronic smokers, induced changes in ZIP8 expression in lung epithelia and whether changes in transporter expression was necessary to mediate Cd-induced toxicity. Knowing that Zn and Cd compete for cellular uptake through ZIP8, we further modified Zn concentrations to simulate both Zn-deficient and -sufficient states and observed that the relative abundance of this key micronutrient is vital in maintaining cell viability. Cell culture findings were validated following inspection of lung specimens obtained from chronic smokers and nonsmokers. Consistent with emerging clinical evidence obtained from the NHANES III Trial, our findings indicate that Zn and a factor central to its metabolism impact the extent by which Cd enters the lung, subsequently causing cell death. We believe that these basic observations begin to fill a critical gap in our understanding of how Cd contributes to COPD pathogenesis.

MATERIALS AND METHODS

Cell culture.

The human lung epithelial A549 cell line (catalog no. CCL-185; American Type Culture Collection, Manassas, VA) was maintained under standard culture conditions in DMEM supplemented with 10% FBS, 0.1 mg/ml streptomycin, 100 IU/ml penicillin, and 1% nonessential amino acids at 37°C in a 5% CO₂-humidified incubator. All cells were used between passages 6 and 20. Cells were maintained in serum-free conditions 24 h before and throughout the duration of experiments to minimize absorptive cadmium loss due to serum protein binding. Primary, differentiated, polarized human upper airway epithelial cells (HUAECs) were isolated and cultured as previously reported (2, 28). Results in this investigation are derived from three different donors. HUAECs were maintained in 1:1 DMEM and Ham's F-12 media (DMEM/F-12) supplemented with 2% Ultroser G (BioSepra, Villeneuve, France) and antibiotics unless otherwise stated. Human lungs were collected with approval from The Ohio State University Institutional Review Board.

Cadmium exposure and analysis of cytotoxicity.

A549 cells were seeded and then serum starved for 24 h followed by overnight treatment with 100 ng/ml TNF-α or vehicle control. Cultures were then exposed to increasing concentrations of CdCl₂ for 24 h. All experiments, unless otherwise stated, were performed in triplicate. Culture supernatants were then collected, and lactate dehydrogenase (LDH) activity was measured using the Cytotoxicity Detection Kit (Roche Diagnostics, Mannheim, Germany). Samples were compared with a positive control treatment group, generated for each experiment by treatment with 2% Triton-X (Sigma Chemical, St. Louis, MO) in DMEM for 10 min to yield 100% cell death. The same protocol was conducted in primary cultures; however, TNF-α was administered to the basolateral surface of polarized cultures, whereas cadmium was administered either basolaterally or apically. In addition to LDH release, the integrity of primary cultures was monitored by measuring transepithelial electrical resistance (TEER) using a portable ohmmeter (Millicell-ERS; Millipore, Billerica, MA). For TEER measurements, 200 μl of media was placed on the apical surface and then immediately removed following recording. Baseline TEER measurements were determined at the beginning of each experiment and then following cadmium exposure. In our model, fully differentiated cultures that maintain an air-to-liquid interface typically exhibit TEER measurements >400 Ω.

Inhibition of NF-κB and ZIP8 expression.

A549 cells were initially seeded and serum starved for 24 h followed by treatment with 20 μM Bay 11–7082, a pharmacological inhibitor of the NF-κB pathway that irreversibly binds to the phosphorylation site of IκB-α, or DMSO as a vehicle control. Following a 1-h exposure, fresh medium was then replaced containing Bay 11–7082 or DMSO and TNF-α, and then cells were incubated for an additional 24 h. Cultures were then exposed to increasing concentrations of CdCl₂ (0 to 25 μM) for an additional 24 h, after which cytotoxicity was determined by measuring LDH release.

ZIP8 expression was inhibited using a 21-mer short interfering (si)RNA target sequence (QIAGEN, Valencia, CA) following transfection into A549 cells using HiPerFect transfection reagent (QIAGEN). ZIP8 expression was typically decreased by >70% following TNF-α stimulation. ZIP8 siRNA-treated cultures were compared with a nonsilencing control siRNA that did not affect ZIP8 expression. Following siRNA treatment, cultures were exposed to TNF-α and then increasing concentrations of CdCl₂ (0 to 25 μM) for an additional 24 h, after which cytotoxicity was determined by measuring LDH release.

Zinc coculture study.

Cells were seeded and serum starved for 24 h and then stimulated with TNF-α or vehicle control for 24 h. Cultures were then exposed to a fixed concentration of CdCl₂ (10 μM) with or without increasing concentrations of ZnCl₂ (0, 10, 20, and 40 μM) and then incubated for an additional 24 h. Cytotoxicity was again determined by measuring LDH release.

Western analysis.

Membrane protein fractions were generated by suspending cells in a lysis buffer composed of Buffer A (20 mM Tris·HCl, pH 7.5, 5 mM MgCl₂, 1 mM ethylene glycol tetra-acetic acid, 20 mM β-glycerophosphate, 1 mM phenylmethanesulfonylfluoride, 2 μg/ml aprotinin, 2 μg/ml leupeptin, 1 mM sodium vanadate) followed by sonication for 5 s four times while on ice. The lysate was then centrifuged at 2,900 revolution/min for 5 min. The supernatant was then centrifuged at 55,000 revolution/min for 30 min with a TLA 120.2 rotor using the Optima TLX-120 Ultracentrifuge (Beckman Coulter, Brea, CA). The pellet was resuspended in Buffer A and centrifuged for 5 min at 2,900 revolution/min. The remaining pellet was resuspended in Buffer A containing 1% NP-40 and agitated for 1 h at 4°C followed by centrifugation at 55,000 revolution/min for 30 min, giving yield to the membrane extract in the resulting supernatant. The whole cell and membrane lysates were quantified using a protein assay (Bio-Rad, Hercules, CA) and then mixed in Laemmli buffer (Bio-Rad) containing 5% (vol/vol) 2-mercaptoethanol, boiled for 5 min, separated on 10% SDS-PAGE gel (Bio-Rad), and then transferred to a nitrocellulose membrane (Amersham Biosciences, Little Chalfont, UK). Membranes were blocked with 5% milk (wt/vol) in PBS 0.1% Tween 20 (PBS-T) for 1 h at room temperature and then incubated with primary antibody overnight at 4°C. After being washed, the membranes were incubated with secondary antibody for 1 h at room temperature. The signal was detected with an ECL Kit (Amersham Biosciences) and a Fluor-S Multi-Imager Max/Bio-quantity one (Bio-Rad). The following antibodies were used in our experiments: rabbit anti-ZIP8 (1:2,000; Covance, Princeton, NJ), mouse anti-β-actin (1:2,000; MP Biomedicals, Aurora, OH), goat anti-rabbit IgG-horseradish peroxidase (HRP) (1:3,000; Cell Signaling, Beverly, MA), and horse anti-mouse IgG-HRP (1:3,000, Cell Signaling).

Intracellular cadmium measurements.

A549 cells were subjected to a 24-h TNF-α stimulation with or without ZIP8 siRNA and then treated with increasing concentrations of Cd as previously described. Cell supernatants were then collected and centrifuged to further collect detached cells. The Measure-iT Lead and Cadmium Assay Kit (Invitrogen, Carlsbad, CA) was used to measure intracellular cadmium concentration. DMSO was added directly to each well, and then cells were scraped and combined with pelleted cells in the supernatant, followed by vortexing to lyse the cells. Then 10 μl of sample was added to a 96-well plate followed by 200 μl of the Measure-iT kit reagent. The fluorescence intensity was recorded for each well at 520 nm (λ_ex: 490). Samples were analyzed in triplicate, and three readings were performed for each sample. A cadmium calibration curve with a range between 5 and 200 nM CdCl₂ was used per manufacturer guidelines to determine intracellular Cd concentration within each sample. Cd content of samples was standardized to protein content as measured in the lysates using the Pierce BCA Protein Assay Kit (Thermo Scientific, Rockford, IL).

Analysis of apoptosis and necrosis.

A549 cells were detached with trypsin and pooled with cells already suspended during culture and then fixed using 100% methanol and stained with an M30 CytoDEATH, Fluorescein-conjugated antibody (Boehringer Mannheim, Indianapolis, IN), a monoclonal antibody that specifically detects caspase-cleaved human cytokeratin-18 (CK-18). Concomitant nuclear staining was also conducted using 0.5 mg/ml 4′6-diamidino-2-phenylindole dihydrochloride (DAPI; Roche Molecular Biochemicals, Indianapolis, IN). Upon staining, cells were collected and cytospun onto slides and analyzed by fluorescent microscopy. Apoptotic cells (M30-positive cells with fragmented nuclei) were enumerated by a blinded observer who randomly selected 14 fields of view per treatment condition. Data are presented as the average percentage of apoptotic cells divided by the total number of cells per viewing area. Lung epithelial cultures were also evaluated by flow cytometric analysis to evaluate cell death. Briefly, A549 cells were detached with trypsin and combined with cells already suspended during culture and then pelleted. The pellet was resuspended and washed with PBS and again pelleted. Pellets were then resuspended in annexin V (AV) binding buffer with an AV antibody and then incubated in the dark for 15 min. An additional volume of AV binding buffer was added, and then samples were filtered through 0.2-μm filters into tubes suited for flow cytometry. Propidium iodide (PI) was added immediately before flow analysis.

Immunohistochemical analysis of ZIP8 expression.

The apical surface of HUAEC cultures grown on 24-well Transwells were first washed with PBS to remove debris and then fixed with 4% formaldehyde. Monolayers were then blocked with 10% goat serum for 2 h at room temperature in permeabilizing buffer and then incubated with primary rabbit anti-ZIP8 antibody overnight at 4°C. Following the wash, membranes were incubated with secondary antibody (Alexa Fluor 488 goat anti-rabbit antibody, Invitrogen) for 1 h. Nuclear DNA was detected with DAPI. Slides were mounted with Citifluor antifadent mounting medium (AF1; Electron Microscopy Science, Fort Washington, PA) and then examined using a disk-scanning confocal microscope at ×600 (Olympus BX61). The magnification of all images was performed using ×10 (WHN10X) and ×60 (Olympus 60X/1.42 Oil PlaneApon or 60X/0.90N LUMPLANF1) objectives. The z-section images were obtained and analyzed using Slidebook (Intelligent Imaging Innovations, Denver, CO) software.

mRNA analysis of human lung tissue.

Total RNA was isolated using Trizol reagent from human lung tissue samples that were received from the Lung Tissue Research Consortium of the National Heart Lung and Blood Institute that included seven chronic smokers (COPD Stage 0) and five control, life-time nonsmokers. The ThermoScript RT-PCR kit (Invitrogen) was used to generate cDNA. Primer pairs were designed for ZIP8 and GAPDH using Primer Express software version 2.0 (Applied Biosystems, Foster City, CA) as previously reported (8). The 7900HT Fast Real-Time PCR system (Applied Biosystems) using SYBR Green Master Mix (2x, Applied Biosystems) was used to perform real-time quantitative PCR. All samples were standardized to average cycle threshold number of the GAPDH gene. Messenger expression was reported as the average relative copy number as follows: 2^−ΔC_t ×100, where ΔC_t is the C_t value standardized to GAPDH (18).

Immunodetection of ZIP8 in human lung tissue.

A portion of the same human specimens described above were also fixed and mounted on slides. After deparaffinization, endogenous peroxidases were blocked for 15 min with hydrogen peroxide, then blocked with 10% milk in TBS for 10 min, and then with 10% normal goat serum in TBS for 2 h. Primary antibody (anti-hZIP8) was then applied in 5% BSA in TBS and incubated overnight at 4°C. Following three washes with PBS-T, slides were incubated with goat anti-rabbit biotin-conjugated antibody for 45 min at room temperature. Slides were washed with PBS-T four times and then incubated in avidin-biotin-HRP in PBS-T for 45 min. After being washed three times in PBS-T, slides were developed using diaminobenzidine diluted in 0.05M Tris·HCl and then rinsed in distilled water and counterstained with hematoxylin. After dehydration with alcohols and xyline, slides were mounted on a permanent coverslip and evaluated by light microscopy.

Statistical analysis.

All data are expressed as means ± SD. For comparisons involving multiple variables and observations, a two- and three-way ANOVA (GraphPad, La Jolla, CA) were used. Having passed statistical significance by ANOVA, individual comparisons were made with the Bonferroni multiple-comparison test. Statistical significance was defined as a P value <0.05.

RESULTS

TNF-α enhances Cd toxicity in lung epithelia.

ZIP8 expression is typically low in lung epithelia but highly induced by proinflammatory mediators (3, 6). On the basis of this and knowing that ZIP8 is a transporter of Cd, we predicted that induction of ZIP8 expression by TNF-α, a relevant proinflammatory factor present in the lung of smokers (11), would increase Cd toxicity in lung epithelia. To investigate the transporter's contribution to Cd-induced toxicity, A549 cells were first stimulated with TNF-α for a time sufficient to increase ZIP8 expression and then exposed to increasing concentrations of Cd for 24 h. A549 cells stimulated with TNF-α before Cd challenge had a significant increase in cell death, as determined by LDH release, compared with cultures that were exposed only to Cd (Fig. 1A). Western analysis confirmed a 6.5-fold induction of membrane-bound ZIP8 following TNF-α stimulation (Fig. 1B). These results indicate that lung epithelia become more vulnerable to Cd-mediated toxicity following activation by TNF-α. There was no toxicity associated with just TNF-α exposure in agreement with our previously reported findings (2).

Fig. 1.

A549 cell toxicity in response to Cd and TNF-α treatment. A: A549 cells were first exposed to TNF-α (100 ng/ml) for 24 h and then exposed to increasing concentrations of Cd for an additional 24 h. A significant increase in cell toxicity was observed in TNF-α-stimulated cultures exposed to Cd. Toxicity was standardized relative to detergent-lysed control cells (100% cell death). Data are expressed in triplicate and representative of at least 3 experiments (***P < 0.001, 2-way ANOVA). B: A549 cells were exposed to TNF-α for 24 h, and cell membrane fractions were analyzed by Western blotting for human ZIP8. Autoradiographs were quantified by standard densitometry and standardized to β-actin for each sample to measure fold induction of ZIP8 expression compared with untreated control samples.

NF-κB and ZIP8 inhibition decrease Cd-induced cell toxicity.

We next determined whether Cd-induced toxicity in lung epithelia is dependent on the induction of ZIP8 expression. Our group recently reported that ZIP8 expression is transcriptionally activated by NF-κB (RelA/p65) (unpublished observations, M. Liu). Knowing this, we stimulated lung epithelia cultures with TNF-α but in the presence of Bay 11–7082, a compound that prevents IκB-α phosphorylation, thereby inactivating the canonical NF-κB pathway. Inhibition of the NF-κB pathway resulted in a significant decrease in cell toxicity in cultures exposed to a combination of TNF-α and Cd compared with the DMSO vehicle control treatment group (Fig. 2A). Knowing that Cd-mediated toxicity is NF-κB dependent, we then pursued a similar study in conjunction with siRNA designed to suppress the induction of ZIP8 expression. A549 cells were first treated with ZIP8-specific siRNA or a corresponding scrambled siRNA control, then stimulated with TNF-α, and then once again subject to increasing concentrations of Cd. Cultures in which ZIP8 expression was inhibited exhibited a significant decrease in cell toxicity relative to cultures treated with the siRNA control (Fig. 2B). Western blotting confirmed an ∼70% knockdown of ZIP8 in cultures treated with either Bay 11–7082 or siRNA (Fig. 2C). Collectively, these results support our hypothesis that the induction of ZIP8 expression by a proinflammatory factor associated with chronic cigarette smoke exposure significantly increases lung epithelial vulnerability to Cd.

Fig. 2.

ZIP8 inhibition reduces Cd-mediated toxicity. A: A549 cultures were first treated with the NF-κB inhibitor Bay 11–7082 (20 μM) or DMSO as vehicle control for 60 min and then exposed to TNF-α for 24 h followed by exposure to increasing concentrations of Cd for an additional 24 h. NF-κB inhibition resulted in a significant decrease in cell toxicity compared with the DMSO (vehicle) control in TNF-α- + Cd-treated cultures. Data are expressed in triplicate and representative of 3 separate experiments (***P < 0.001; 2-way ANOVA). B: similarly, inhibition of ZIP8 expression with a ZIP8-specific siRNA resulted in a significant decrease in cell toxicity compared with the siControl treatment group. Data are expressed in triplicate and representative of 3 experiments (***P < 0.001, 2-way ANOVA). C: membrane fractions were also obtained from samples in A and B and analyzed by Western blotting with a primary antibody against ZIP8. Densitometry was standardized to actin and used to determine the percent knockdown of ZIP8.

Zn decreases Cd-induced cell toxicity.

ZIP8 was first identified as a Zn importer and then subsequently discovered to also be an avid transporter of Cd (5,12). Knowing that Zn acts as a cytoprotectant in lung epithelia (6), we wanted to determine whether physiologically relevant concentrations of extracellular Zn can prevent Cd-mediated toxicity. A549 cells were again stimulated with TNF-α and then exposed to a constant concentration of Cd but in the presence of increasing concentrations of Zn. Strikingly, lung epithelial cell toxicity induced by Cd was decreased in the presence of increasing concentrations of Zn, which was most notable when the molar ratio between Cd and Zn was in favor of Zn (Fig. 3C). Cell toxicity was completely inhibited with a fourfold molar excess of zinc relative to Cd. Consistent with these findings, we observed an increase in intracellular Cd concentrations in TNF-α-treated cells and a decrease in intracellular concentrations when treated in conjunction with a ZIP8 siRNA (Fig. 3, A and B). On the basis of the minimum detection threshold of intracellular Cd measurements (5 nM), we were not able to evaluate the impact of Cd doses below 10 μM. Taken together, these findings further demonstrate that ZIP8 is an important regulator of Cd uptake and that the vital micronutrient Zn plays an important role in preventing Cd-mediated lung epithelia toxicity, which we believe to be relevant when considering that a substantial number of patients with COPD are also malnourished.

Fig. 3.

ZIP8 increases Cd and Zn uptake in lung epithelia. A: A549 cells were stimulated with TNF-α for 24 h and then exposed to increasing amounts of Cd (0–25 μM) for 24 h. Intracellular Cd was significantly increased in TNF-α-treated cells. Data are expressed in triplicate and representative of 3 experiments (***P < 0.001, *P < 0.05, 2-way ANOVA). B: A549 cells were transfected with scrambled or ZIP8 siRNA and then stimulated with TNF-α for 24 h. Cells were then exposed to Cd (0–25 μM) for 12 h. A decrease in intracellular Cd was observed in cells treated with the ZIP8 siRNA although not statistically significant. Data are expressed in triplicate and representative of 3 experiments. C: A549 cells were first exposed to TNF-α and then increasing concentrations of Zn (10–40 μM) in combination with a fixed amount of Cd (10 μM). A significant decrease in cell toxicity was observed as Zn concentration increased, relative to Cd, in TNF-α-stimulated cultures. Data are expressed in triplicate and representative of 4 experiments (***P < 0.001, **P < 0.01, 2-way ANOVA).

Cd induces apoptosis and necrosis.

Having established that Cd induces toxicity in lung epithelia in a ZIP8-dependent manner, we also wanted to determine whether cell death was a consequence of necrosis, apoptosis, or both under these conditions. We first evaluated cells for the presence of caspase-cleaved cytokeratin-18 to identify apoptotic cells following combined TNF-α and Cd exposure using the M30 apoptotic marker and DAPI. Cells were considered apoptotic only in the presence of diffuse M30 staining throughout the cytosol and in the presence of a condensed nucleus (Fig. 4A). Combined TNF-α and Cd exposure increased the frequency of apoptotic cells, achieving statistical significance at the highest Cd exposure (Fig. 4B). Using flow cytometry in conjunction with AV and PI staining, we next determined the extent of both necrotic and apoptotic cell populations. Briefly, cells that did not stain positive for either dye were considered viable. AV-positive/PI-negative cells were considered early apoptotic; AV-positive/PI-positive were identified as mixed late apoptotic/necrotic; and PI-positive/AV-negative cells were considered to be necrotic. Consistent with previous findings, the combination of TNF-α and Cd exposure resulted in an increase in both the necrotic and apoptotic populations (Fig. 4C). Combined TNF-α and Cd exposure increased the frequency of PI- and AV-positive cells, achieving statistical significance at both Cd concentrations (Fig. 4D).

Fig. 4.

ZIP8-mediated Cd uptake induces apoptosis and necrosis. A: cells were treated with TNF-α and increasing concentrations of Cd (0–25 μM) for 24 h and subsequently stained with an M30 antibody specific for caspase-cleaved cytokeratin 18. Representative photomicrographs of M30-positive A549 cells exposed to TNF-α and Cd are shown. Cells were designated as apoptotic if they stained green with condensed nuclei (M30-positive green, DAPI blue). B: increase in apoptosis was observed in cells exposed to TNF-α and Cd. Percent apoptosis was calculated by dividing the number of apoptotic cells by the total number of cells in the field of view. Data are representative of 4 experiments. A statistically significant increase in apoptotic cells was observed in cells exposed to higher Cd concentrations in combination with TNF-α (***P < 0.001, 2-way ANOVA). C: A549 cells treated under the same conditions with TNF-α and Cd were also stained with annexin V (AV) and propidium iodide (PI) and enumerated by flow cytometry. The incidence of both necrosis (N) (PI+/AV+, top, right) and apoptosis (A) (PI-/AV+, bottom, right) was highest in cells exposed to both TNF-α and Cd. Data in each panel are representative of 5 experiments. D: increase in cell death was observed in cells exposed to TNF-α and Cd. There was a statistically significant increase in PI+/AV+ cells following exposure to Cd in combination with TNF-α (***P < 0.001, 2-way ANOVA). PI+/AV+ cells represent a mixed population of late apoptotic and necrotic cells.

ZIP8 is preferentially expressed at the apical surface and mediates Cd-induced toxicity in primary human lung epithelia.

Initial studies were conducted in fully differentiated and polarized HUAEC monolayers to determine whether ZIP8 preferentially translocates to the apical or basolateral membranes following transcriptional activation by TNF-α. Confocal analysis of TNF-α-stimulated HUAEC cultures established that ZIP8 protein preferentially but not completely localized to the apical membrane upon cell activation (Fig. 5A), whereas little if any ZIP8 was observed in nonstimulated cultures. Next, we determined whether polarized cultures were more sensitive to Cd following apical or basolateral exposure. Treatment conditions were similar to past studies; however, TNF-α was administered basolaterally, as previously reported by our group (7), followed by Cd exposure at either the apical or basolateral surface. Consistent with past observations, basolateral and apical Cd exposure resulted in increased cell toxicity and more so in TNF-α-stimulated cells as measured by LDH release. Strikingly, Cd exposure at the apical surface resulted in significantly more cell damage compared with basolateral exposure (Fig. 5B). A similar significant reduction in TEER (R_t) was also observed, demonstrating that apical Cd exposure in TNF-α-activated cultures resulted in the largest reduction in epithelial barrier function (Fig. 5C). Taken together, our findings demonstrate that ZIP8 protein expression is rapidly induced in primary lung epithelia and preferentially localizes to the apical membrane, an anatomical location ideally suited for Cd uptake following cigarette exposure.

Fig. 5.

Polarized ZIP8 expression in primary lung epithelia increases cell toxicity in response to TNF-α and Cd treatment. A: immunofluorescent staining was conducted on confluent polarized primary HUAEC cultures using an hZIP8 antibody and then visualized by z-stack using a disc-scanning confocal microscope. TNF-α-treated cells exhibited an increase in ZIP8 expression that was preferentially localized to the apical membrane, whereas nonstimulated cultures showed minimal evidence of ZIP8 expression. ZIP8 staining is indicated by red fluorescence, and DAPI nuclear staining is indicated by blue fluorescence. B: fully differentiated, polarized, primary human upper airway epithelial cells were treated with TNF-α for 24 h, followed by Cd exposure for 48 h. LDH release was measured, and percent cell death was determined using a detergent-positive control (100% cell death). Toxicity was significantly increased in cells treated with TNF-α and then apically exposed to cadmium. Data are expressed in triplicate and representative of 4 experiments (*P < 0.05, 2-way ANOVA). C: in a similar experiment, cultures were treated as previously described, and transepithelial membrane resistance (R_t) was measured (Ω). The decrease in R_t, indicative of compromised membrane integrity, was more significant following apical Cd exposure compared with basolateral Cd exposure. Data are expressed in triplicate and representative of 3 experiments (*P < 0.05, **P < 0.01, 2-way ANOVAs).

ZIP8 is increased in the lungs of chronic smokers.

On the basis of our findings obtained from human lung epithelial cell models, we next sought to evaluate the expression of ZIP8 in human lung tissue. Lung tissue samples were obtained from lifetime nonsmokers (n = 5) and chronic smokers (n = 7) through the NIH-sponsored Lung Tissue Research Consortium. Quantitative analysis of ZIP8 mRNA levels revealed a consistent and significant increase in ZIP8 mRNA transcripts in the lungs of smokers compared with nonsmokers (Fig. 6A). Consistent with RNA findings, immunohistochemical analysis of the same biopsy specimens revealed an increase in ZIP8 protein throughout parenchymal tissue that appeared to be most prominent in upper airway and alveolar epithelia (Fig. 6B). The increase of both ZIP8 protein and mRNA transcripts in the lungs of smokers strongly supports our cell culture findings indicating that chronic cigarette smoke exposure increases ZIP8 expression, thereby enhancing the capacity of lung tissue to obtain cadmium.

Fig. 6.

ZIP8 mRNA and protein expression are elevated in smokers. A: total RNA was extracted from human lung tissue samples that were obtained from chronic smokers (stage 0 COPD) and lifetime nonsmokers, and ZIP8 mRNA expression levels were measured. Quantitative RT-PCR analysis consistently revealed higher ZIP8 mRNA levels in the lung tissue of smokers (n = 7) compared with tissues obtained from nonsmokers with the exception of one outlier in the control group (n = 5) (Student's t-test; **P < 0.01). RCN, relative copy number. B: immunohistochemical staining for ZIP protein was then conducted on tissue sections obtained from 2 subjects within the chronic smoker group. Consistent with mRNA expression, both specimens exhibited an increase in ZIP8 throughout the lung epithelia (staining indicated by brown areas at ×200 and ×400 magnification of donor 224671). ZIP8 immunostaining of 2 lifetime nonsmokers did not reveal evidence of increased ZIP8 expression (donor 013011 shown at ×200 and ×400 magnification). As a negative control, the same specimens were stained with the secondary rabbit IgG control antibody (donors 224671; smoker, and 013011; nonsmoker), at ×200.

DISCUSSION

Smoking is responsible for 90% of all COPD cases in the United States. Cigarette smoke itself contains over 2,000 xenobiotic compounds that can damage lung tissue, resulting in chronic bronchitis and emphysema. Specifically, chronic smoke inhalation creates an inflammatory environment within the lung through the elaboration of cytokines and chemokines by parenchymal cells and alveolar macrophages. Activation of resident and recruited cell populations result in a significant increase in TNF-α production, as well as other inflammatory mediators, both within the lung and throughout the body (4, 9–11, 17). Consistent with these observations, the lungs of smokers and patients with COPD exhibit a persistent increase in NF-κB activity, a transcriptional pathway that is central to the innate immune system and activation of the inflammatory response (49, 55). Whether persistent inflammation in the lung directly impacts the uptake of xenobiotics present in tobacco smoke, particularly Cd, remains unknown. Importantly, recent analysis of the NHANES III study that involved 6,726 patients revealed that higher urine Cd and therefore body levels positively correlate with smoking-related lung dysfunction and even more so in subjects with insufficient Zn intakes (31). Taken together, these findings suggest that a dynamic interplay between Cd and Zn may exist in the context of cigarette smoke-induced lung disease.

Zn is an essential micronutrient, and its metabolism in humans is primarily controlled by Zn transporters, a composite of 24 proteins that are further divided into two families based on their ability to transport zinc into or out of the cytosol. There are 14 SLC39A importers (ZIPs) and 10 SLC30A exporters (ZnTs), of which all 24 are highly conserved between humans and mice (13, 33). Together, Zn transporters function coordinately throughout the body to direct Zn biodistribution in response to nutritional intake and body demand. ZIP8 is unique, relative to most other transporters, in that its expression is under the transcriptional control of inflammatory mediators (6). Specifically, our group was the first to discover that the ZIP8 promoter is activated by NF-κB (p65) (unpublished observations, M. Liu), a transcription factor that, as already mentioned, plays a central role in coordination of innate immune function and host defense. With this in mind, it is plausible to consider that ZIP8 expression is induced in the lung as a consequence of chronic cigarette smoke exposure. Consistent with this postulate, we observed an increase in ZIP8 mRNA and protein expression in the lungs of smokers compared with nonsmokers (Fig. 6). Furthermore, ZIP8 was most prominent throughout epithelia, lining both the airway and alveolar regions. This observation is remarkable when considering that ZIP8 was first identified as the Cd toxicity gene (12). Indeed, in addition to the unique ability of ZIP8 to become transcriptionally activated by inflammatory stimuli, it is a transporter of Cd (K_m = 0.48 μM), exhibiting a K_m comparable to its endogenous ligand Zn (K_m = 0.26 μM) (20, 32). Furthermore, consistent with our own observations (Fig. 5A), ZIP8 is preferentially expressed on the apical surface of polarized epithelia as a plasma membrane protein, making it ideally situated to transport recognizable trace metals into lung epithelia (12, 20, 21, 32). Taken together, this would suggest that the induction of ZIP8 expression in lung epithelia of a smoker creates a potential disadvantage by enhancing Cd uptake into tissues following its repeated inhalation subsequent to cigarette smoke exposure.

Zn deficiency as a consequence of insufficient nutritional intake affects nearly 2 billion people worldwide primarily within developing nations (43). Adolescence, aging, and lower socioeconomic status are significant risk factors for poor nutrition and Zn deficiency (1, 42, 43). Perhaps not coincidentally, COPD is also associated with lower socioeconomic status and age (8, 23) and often occurs with comorbidities, of which nutritional deficiency is a common manifestation (22, 27, 45). In particular, muscle wasting, cachexia, and appetite suppression are common symptoms associated with COPD linked to nutritional deficiencies. Although the incidence of Zn deficiency has never been formally studied within the COPD population, we predict that its prevalence may be relatively high considering the incidence of malnutrition. Clearly, epidemiological evidence exists within the United States to indicate that lower Zn intakes are commonly encountered within smokers and that it increases the risk of developing COPD following prolonged cigarette smoke (31). It then becomes plausible to consider that Zn supplementation in smokers has the potential to prevent or delay the progression of Cd-dependent pathogenesis within the lung. Our findings (Fig. 3C) and those of others (32) demonstrate that Cd and Zn are both substrates for uptake via ZIP8 into cells and that differences in the uptake of either trace metal has profound influence on cell viability. In particular, when Zn concentrations are in relative excess, Zn uptake relative to Cd is most likely favored, thereby enhancing cell survival. Joshi et al. (26) reported a reduction of Zn within the bronchoalveolar lavage fluid of alcohol-fed rats that was restored upon dietary Zn supplementation (26). This suggests that there exists an available pool of Zn within the airway that is dependent on dietary intake. However, little is known regarding homeostatic regulation of Zn within the airway in humans and requires further study. Although not yet investigated in our model, we believe that the preferential uptake of Zn results in beneficial intracellular affects by influencing signaling events and redox balance that maintain a more favorable environment, unlike Cd. Indeed, ZIP8-mediated Zn uptake acts as a cytoprotective in lung epithelia in the setting of inflammation, whereas Cd has been shown to alter formation of reactive oxygen species and derivatives, leading to cell toxicity in the setting of inflammation (29, 41, 56).

Zn plays a critical role in immune function, redox signaling, and regulation of inflammation (38, 44, 48). Consistent with this, our group revealed that Zn-deficient mice exhibit greater NF-κB activation and subsequent inflammation in response to polymicrobial sepsis (3). Furthermore, intracellular zinc content has been shown to function as a cytoprotectant and maintain cellular homeostasis in response to inflammation (6, 51). On the basis of these findings, we postulate that intracellular Zn status plays an important role relative to the cellular response following Cd uptake. Importantly, Cd has been shown to induce TNF-α production in monocytes, macrophages, and renal and liver tissue (15, 16, 19, 30, 54). Taken together, this would suggest that ZIP8-mediated uptake of Zn relative to Cd is critical in determining the overall cellular response, in this case, to the inflammatory environment within a smoker's lung. In our model, we provide evidence that ZIP8 expression is a critical event in determining cell survival and that the amount of Zn present in the extracellular environment relative to Cd is critical in determining cell fate. Indeed, if Cd accumulation was favored over Zn, then it is plausible to consider that intracellular Cd would alter cellular signaling events, NF-κB-mediated transcription, and more ZIP8 production, thereby creating a vicious cycle of further Cd accumulation and cellular dysfunction. In our model, we did not observe substantial cell toxicity with just Cd exposure alone as shown (Fig. 1A and Fig. 3A). A major difference between past studies and findings presented herein is that Cd-mediated changes were studied in monocytes and macrophages, both principal producers of TNF-α. Epithelial cells are not considered to be significant producers of TNF-α. Furthermore, our group and others have shown that ZIP8 expression in monocytes and macrophages, similar to lung epithelia, is highly inducible in response to proinflammatory stimuli such as TNF-α (5, 7, 46). This raises important questions regarding the impact of Cd within alveolar macrophages in the setting of Zn deficiency and chronic cigarette smoke exposure and how cellular crosstalk may contribute to epithelial cell toxicity. Future investigations within our laboratory will need to focus on ZIP8-mediated effects of both epithelial and mononuclear cellular function in the setting of COPD pathogenesis.

This investigation provides novel insight into a mechanism whereby Zn metabolism influences Cd-mediated toxicity in the lung; however, findings from our model are limited in that COPD develops over decades. Disruption of the balance of cell turnover results in alveolar septal loss, but the rate and extent of cell turnover is very low. Clearly, the balance between micronutrient and trace metal uptake is more complicated in humans. In consideration of our observations, we postulate that ZIP8 may contribute to COPD pathogenesis through additional mechanisms. In particular, Cd accumulation increases epithelial injury and turnover, either via apoptosis or necrosis, wherein the latter can result in the release of damage-associated molecular patterns. In this scenario, ZIP8-mediated cadmium uptake would further promote inflammation (14, 34, 40). In addition, Cd entry into lung cells has been shown to disturb multiple signaling pathways, thereby altering normal physiological functions (50). Given the long half-life of Cd in relation to the life expectancy of lung epithelia, it is plausible that increased uptake into lung epithelia would have a sustained impact on cell function, thereby causing permanent impairment.

In conclusion, we have identified that ZIP8 is a major regulator of Cd-mediated lung toxicity in the setting of inflammation. The expression of ZIP8 is upregulated by the NF-κB pathway that in turn enhances Cd uptake into lung epithelia, causing apoptosis and necrosis. We believe that these findings fill a current gap in our understanding of Cd-mediated toxicity in the lung through interaction with a unique Zn transporter, thereby potentially influencing pathogenesis associated with COPD and possibly cancer.

GRANTS

This work was supported by R01 HLD086981-01 (D. Knoell).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: J.R.N. and D.L.K. conception and design of research; J.R.N., M.-J.L., S.B., and M.C. performed experiments; J.R.N., M.-J.L., S.B., and D.L.K. analyzed data; J.R.N. and D.L.K. interpreted results of experiments; J.R.N. prepared figures; J.R.N. and D.L.K. drafted manuscript; J.R.N., P.N.-S., E.C.-B., and D.L.K. edited and revised manuscript; J.R.N., M.-J.L., S.B., M.C., P.N.-S., E.C.-B., and D.L.K. approved final version of manuscript.