|Home | About | Journals | Submit | Contact Us | Français|
Exposure to pollutant particles increased respiratory morbidity and mortality. The alveolar macrophages (AMs) are one cell type in the lung directly exposed to particles. Upon contact with particles, AMs are activated and produce reactive oxygen species, but the scope of this oxidative stress response remains poorly defined. In this study, we determined the gene expression profile in human AMs exposed to particles, and sought to characterize the global response of pro- and antioxidant genes. We exposed AMs obtained by bronchoscopy from normal individuals to Chapel Hill particulate matter of 2.5-μm diameter or smaller (PM2.5; 1 μg/ml) or vehicle for 4 hours (n = 6 independent samples). mRNAs were extracted, amplified, and hybridized to Agilent human 1A microarray. Significant genes were identified by significance analysis of microarrays (false discovery rate, 10%; P ≤ 0.05) and mapped with Gene Ontology in the Database for Annotation, Visualization, and Integrated Discovery. We found 34 and 41 up- and down-regulated genes, respectively; 22 genes (~30%) were involved in metal binding, and 11 were linked to oxidative stress, including up-regulation of five metallothionein (MT)-1 isoforms. Exogenous MT1 attenuated PM2.5-induced H2O2 release. PM2.5 premixed with MT1 stimulated less H2O2 release. Knockdown of MT1F gene increased PM2.5-induced H2O2 release. PM2.5 at 1 μg/ml did not increase H2O2 release. Mount St. Helens PM2.5 and acid-extracted Chapel Hill PM2.5, both poor in metals, did not induce MT1F or H2O2 release. Our results show that PM2.5 induced a gene expression profile prevalent with genes related to metal binding and oxidative stress in human AMs, independent of oxidative stress. Metals associated with PM may play an important role in particle-induced gene changes.
Exposure to particulate matter (PM) is consistently associated with increased morbidity and mortality, attributable, in part, to respiratory illnesses (1, 2). These adverse effects include increased hospital admissions, asthma attacks, pulmonary infection, and mortality (3, 4). The mechanisms by which PM increases morbidity and mortality are not entirely clear, but pulmonary oxidative stress and inflammation induced by PM appear to play an important role (5–11).
The alveolar macrophage (AM) is one major cell type in the lung constantly exposed to ambient pollutants. Upon contact with environmental particulate pollutants, AMs are activated, and produce a large quantity of reactive oxygen species (ROS) from various enzymatic sources (5, 12–16). PM exposure may also increase or decrease antioxidant defense mechanisms in the lung, which further modulates oxidative stress and enhances pulmonary and systemic inflammation (13, 14, 17–25). The scope of this pro- and antioxidant response, however, remains poorly defined, in part because of the multiple enzymatic sources of ROS, and the numerous antioxidant enzymes that can be affected by PM.
To better understand how this complex network of pro- and antioxidant enzymes may be affected by PM, we examined the gene expression profile of human AM exposed to Chapel Hill fine particles of 2.5-μm diameter or smaller (PM2.5). The gene expression analysis highlighted the role of genes involved in metal binding and oxidative stress. The functional role of metallothionein (MT)-1 in response to PM was verified by RNA interference.
All chemicals were purchased from Sigma, Inc. (St. Louis, MO) unless otherwise stated.
Particles used in this study were collected in April of 2002 in Chapel Hill, North Carolina, outside the U.S. Environmental Protection Agency Human Studies Facility using a ChemVol Model 2,400 high-volume cascade impactor (Rupprecht and Patashnick Co., Albany, NY) (5). PM2.5 on polyurethane foam (McMaster-Carr, Atlanta, GA), which were previously cleaned with methanol and water, and dried under sterile conditions, were retrieved weekly. The foam or filter was pre-etted with a small amount of 70% ethanol, and the endotoxin-free water was added to a total volume of 40 ml. The particles were removed from the foam or filter by sonication for 1 hour in a water bath (FS220, Fisher Scientific, Atlanta, GA). The foam was removed, and particles were then lyophilized. The metal constituents of PM2.5 of April 2002 were: Al, 930.0 ng/mg; As, 44.7 ng/mg; Cr, 10.2 ng/mg; Cu, 77.7 ng/mg; Fe, 1,713.0 ng/mg; Pb, 133.2 ng/mg; Ni, 26.7 ng/mg; Se, 70.8 ng/mg; Si, 729.0 ng/mg; Ti, 21.0 ng/mg; V, 66.0 ng/mg; and Zn, 522 (5).
Human AMs and bronchial epithelial cells (BECs) were obtained by bronchoalveolar lavage and bronchial brushings from normal individuals according to procedures described previously (26, 27). Subjects were informed of the procedures and potential risks, and each signed an informed consent. The protocol was approved by the University of North Carolina School of Medicine Committee on Protection of the Rights of Human Subjects.
Procedures for human AM isolation and culture have been described previously (5, 13). Bronchoalveolar lavage samples were put on ice immediately and centrifuged at 300 × g for 10 minutes at 4°C. The lavaged cells were washed once with ice-cold RPMI 1640 medium with 20 mg/ml gentamicin (Life Technologies, Rockville, MD). Cell counts were performed with a hemacytometer. Cytocentrifuge slides were prepared and stained with Diff Quick (Leukostat Solution; Fisher Scientific) to check for AM purity. The cell preparation consisted of 85–95% AMs. The viability of AMs was determined by trypan blue exclusion. Viability exceeded 85% in all samples. AMs were suspended in 1 ml of RPMI 1,640 supplemented with 2.5% FBS (Life Technologies, Gaithersburg, MD) in 5-ml polypropylene tubes. The suspended cells were then incubated with vehicle or PM2.5.
Procedures for human BEC isolation and culture have been described previously (5). BECs adherent to the brushes were treated for 15 minutes at room temperature with 1 mM DTT in BEC growth medium (BEGM; Clonetics, San Diego CA) supplemented with bovine pituitary extract, 5 μg/ml insulin, 0.5 μg/ml hydrocortisone, 50 μg/ml gentamicin, 0.1 ng/ml retinoic acid, 10 μg/ml transferrin, 6.5 ng/ml triiodothyrodine, 0.5 μg/ml epinephrine, and 0.5 ng/ml human epidermal growth factor, and dislodged from the brushes by pipeting. The cells were then plated at a density of 1 × 105 viable cells/cm2 in six-well cluster plates (surface area, 9.6 cm2) that were precoated with 50 mg/ml human placental collagen type IV and used after passage 2 or 3.
A suspension of AMs (1 × 106 cells) were incubated with control vehicle or Chapel Hill PM2.5 (1 μg/ml) for 4 hours in a static condition. At the end of the 4-hour incubation, total cellular RNA was extracted from AMs with Trizol reagent (GIBCO BRL/Life Technologies, Gaithersburg, MD), and further purified with phenol/chloroform. The RNA was amplified with a fluorescent linear amplification kit (no. 5184-3523; Agilent Technologies, Inc., Palo Alto, CA) according to the manufacturer's recommended procedures. The RNA integrity was assessed with an Agilent 2,100 bioanalyzer (Agilent Technologies, Inc.). The 260 nm:280 nm ratios for all RNAs were greater than 2.0.
Total RNAs (250 ng) were labeled using a Low Input Linear Amplification kit (Agilent Technologies) according to the manufacturer's protocol. Briefly, first- and second-strand cDNA synthesis was immediately followed by an in vitro transcription reaction, during which Cy3- or Cy5-UTP were incorporated. Labeled cRNA was purified on RNAEasy columns (Qiagen, Valencia, CA) and quantified using NanoDrop ND-1,000 UV-Vis spectrophotometer; 1 μg each of Cy3- and Cy5-cRNA were combined, defragmented, and hybridized on Agilent's In Situ Oligo Human 1A version 1 arrays per Agilent's protocol. Hybridizations were performed overnight (16–18 h) at 60°C in a rotisserie oven. Slides were washed using Agilent's sodium chloride–sodium citrate wash buffers and procedures, and scanned on an Axon 4000B confocal scanner (Molecular Devices, Sunnyvale, CA).
The data discussed in this article have been deposited in the Gene Expression Omnibus (accession no. GSE10394; http://www.ncbi.nlm.nih.gov/geo/). All analyses were performed using TM4 software suite (28). Intensities were extracted from raw tiff images using TIGR Spotfinder. Cy3 and Cy5 intensities were lowess normalized for each individual hybridization using TIGR MIDAS. Dye flip replicas were filtered, merged to produce a single expression ratio measure for each gene, and log–base 2 transformed. Differentially expressed genes were identified using significance analysis of microarrays with 100 permutations (29). FDR was set at 10%.
We also mapped the significant genes with Gene Ontology in the Database for Annotation, Visualization, and Integrated Discovery (DAVID Bioinformatic Resources 2007, National Institute of Allergy and Infectious Disease, http://apps1.niaid.nih.gov/david/) (30).
Quantitative PCR (Q-PCR) was performed for selected genes. cDNAs were synthesized from 0.4 μg of total RNA in 100 μl of buffer containing 5 μM random hexaoligonucleotide primers (Pharmacia, Piscataway, NJ), 10 U/μl Moloney murine leukemia virus reverse transcriptase (GIBCO BRL/Life Technologies), 1 U/μl RNase inhibitor (RNasin; Promega, Madison, WI), 0.5 mM deoxynucleotide triphosphate solution (Pharmacia), 50 mM KCl, 3 mM MgCl2, and 10 mM Tris-HCl (pH 9.3) for 1 hour at 39°C. Reverse transcriptase was heat inactivated at 94°C for 4 minutes.
Q-PCR of specimen and standard cDNA was completed using TaqMan predeveloped assay reagents. Quantitative fluorogenic amplification of cDNA was performed using the ABI Prism 7,500 Sequence Detection System, primers and probes of interest, and TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, CA). The relative abundance of mRNA levels was determined from standard curves generated from a serially diluted standard pool of cDNA prepared from control human pulmonary artery endothelial cell cultures. The relative abundance of glyceraldehyde-3-phosphate dehydrogenase (Unigene accession no. 544,577) mRNA was used to normalize levels of the mRNAs of interest. For Q-PCR verification, RNA from six additional experiments was collected. RNA samples for microarray and Q-PCR were collected from different experiments, and do not represent the same sample.
Release of H2O2 into the medium was measured by the Amplex red reagent (10-acetyl-3,7-dihydroxyphenoxazine) (Molecular Probes/Invitrogen Corp., Carlsbad, CA). Amplex red reacts with H2O2 to produce highly fluorescent resorufin in the presence of horseradish peroxidase. The fluorescent signals were measured over 120 minutes with a Bioassay HTS7,000 plate reader (Perkin Elmer, Inc., Wellesley, MA) with HTSoft v1.0 software (PE Applied Biosystems, Weiterstadt, Germany). The excitation and emission wavelengths were 530 and 590 nm, respectively. Generation of H2O2 was calculated by subtracting the fluorescence signal at 0 minutes (baseline) from that at 120 minutes, and expressed as optical density of resorufin. Positive control samples using H2O2 (0.5–2.0 μM) were included for all Amplex red assays. Human AMs were also incubated with PM2.5 with or without rabbit liver zinc–containing MT1 (Alexis Biochemicals, San Diego, CA). MT was added 20 minutes before Amplex red.
Human BECs were grown to 60–70% confluency and then transfected with Gene Silencer transfecting agent plus (Gene Therapy System, Inc., San Diego, CA) with MT1F short interfering RNA (siRNA; 100 nM) (Ambion, Austin, TX) (cat. no. 16,708/siRNA, ID no. 44,734/lot no. ASO04QXP/Locus ID 4,494/RefSeq no. NM_005949) in serum-free basic growth medium-2 for 3 hours, according to the manufacturer's recommendations. All control cells were incubated with Gene Silencer select negative control no. 1 siRNA (part no. 4390844) (Ambion). Fresh endothelial growth medium-2 with 2% FBS was then added, and cells were cultured for an additional 48 hours. Cell lysates from some wells were collected for MT1F gene expression by RT-PCR. Other wells were used for measuring H2O2 release.
Data are presented as mean values (±SE). Data from nonmicroarray experiments were analyzed by one-way ANOVA, followed by Scheffé's test for post hoc comparisons. The statistical analysis was performed using commercially available software (Statview version 5.0.1; SAS Institute Inc., Cary, NC). A P value of less than 0.05 was taken as statistically significant.
Using the above statistical algorithm, we identified 38 and 43 probes that were up- and down-regulated, respectively, at 10% FDR in AMs exposed to PM2.5. After eliminating probes with no corresponding genes, we reduced the list to 34 up-regulated genes (Table 1) and 41 down-regulated genes (Table 2). Expression of 13 up- and down-regulated genes was confirmed by Q-PCR (Figures 1 and and2).2). We further categorized these differentially expressed genes by their molecular functions using Gene Ontology. We found that 22 genes were involved in metal binding (Tables 1 and and2)2) (GOTERM_MF_3; P = 0.049).
Many of these genes, such as MT1F, MT1E, MT1H, MT1G, MT1B, ATOX1, CYP1B1, and MMP9, are also involved in oxidative stress (Tables 1 and and2).2). Other differentially expressed genes involved in oxidative stress included up-regulation of NCF1, NQO1, and GAPD (Tables 1 and and2).2). Hierarchical clustering analyses of significant genes are shown in Figure 3, which shows two main clusters in up-regulated genes. One cluster contains all MT1 isoform genes, NCF1, NQO1, and MMP9, and the other contains CYP1B1, GAPD, AMID, and ATOX1. Because of up-regulation of multiple MT1 isoform genes, we further investigated the role of MT1F, one of the main isoforms in the lung (31–33), in oxidative stress induced by PM2.5.
Incubation of human AM with PM2.5 increased the extracellular H2O2 release detected by Amplex red in a dose-dependent manner, which was inhibited completely by catalase (200 U/ml) (Figure 4A). The enhanced release of H2O2 was also inhibited by the addition of MT1 protein in the medium (Figure 4B). These concentrations of PM2.5 were not cytotoxic, as there was no increase in lactic dehydrogenase release (data not shown). In addition, PM2.5 preincubated with 100 μM MT1 for 30 minutes before being added to the medium induced lower levels of H2O2 release compared with PM2.5 preincubated with BSA (10 μg/ml) in human AM (Figure 4C) or cell-free system (Figure 4D), indicating that metal chelation is one mechanism by which MT1 decreases PM2.5-induced H2O2 release.
To determine how MT1F regulates H2O2 release, we knocked down MT1F gene expression using siRNA in human BECs. We used human BECs instead of human AMs for these siRNA experiments, because siRNA was ineffective in human AMs. We verified that human BECs, like human AMs, increased MT1F and CYP1B1 gene expression (measured by Q-PCR) by five- and sixfold, respectively, after PM2.5 exposure (100 μg/ml or 10 μg/cm2). Incubation with siRNA reduced MT1F gene expression by approximately 50% (Figure 5A) and increased PM2.5-induced H2O2 release by approximately 30% (Figure 5B).
To determine whether or not the metal components of PM are important in inducing MT1F, human AMs were incubated with untreated PM2.5 (1 μg/ml), metal-poor Chapel Hill PM2.5 and Mount St. Helens (MSH) PM2.5 (1 μg/ml), which contains little metal. To prepare metal-poor PM2.5, 1 μg/ml of Chapel Hill PM2.5 was washed with 1 N HCl three times, and the pellet was resuspended in sterile water (34, 35). Human AMs incubated with acid-extracted Chapel Hill PM2.5 and MSH PM2.5 showed no increase in MT1F gene expression (Figure 6A). Neither the acid-extracted Chapel Hill PM2.5 nor MSH PM2.5 at 1 μg/ml increased H2O2 release (Figure 6B). We also tested higher doses of these metal-poor particle preparations, and there was no increase in H2O2 release at doses up to 100 μg/ml (data not shown). These results indicate that metals associated with Chapel Hill PM2.5 play an important role in gene expression changes induced by the whole particles, independent of oxidative stress.
Our findings suggest that metal components associated with PM may be important in inducing gene changes, independent of oxidative stress. Using microarray analysis and gene ontology, we found that about one-third of the differentially expressed genes in AMs challenged with PM2.5 had molecular functions related to metal binding. We also noted that five MT-1 isoform genes (MT1F, MT1E, MT1H, MT1G, and MT1B) were among the top 10 up-regulated genes. The MT1 genes are known to be involved in heavy metal binding. The MT1 proteins are also important antioxidant enzymes. Three other metal-binding genes, ATOX1, CYP1B1, and MMP9, are also linked to oxidative stress. Furthermore, four other genes not in the metal-binding category, including NCF1, NQO1, AIFM2, and GAPD, are also known to be involved in oxidative stress. Thus, a total of 11 genes could be included in the pro- and antioxidant network. The 11 up-regulated oxidative stress genes fall into two main groups that include MT1 isoform genes and CYP1B1, respectively. Seven genes are involved in calcium binding (MMP9, GCA, PROS1, ASGR1, CALM2, MRC1, and S100A8), and may further modulate oxidant signaling by altering intracellular calcium (36–39).
PM is known to produce oxidative stress in lung cells. Our microarray study identified three pro-oxidant genes that were up-regulated: NCF1, CYP1B1, and AIFM2. NCF1 encodes a 47-kD cytosolic subunit of the membrane reduced nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, an enzyme responsible for the respiratory burst. The membrane NADPH oxidase on AMs may be activated by PM exposure and produce ROS (5, 12, 16). CYP1B1 encodes a mono-oxygenase in the cytochrome p450 system localized in the endoplasmic reticulum. CYP1B1 is part of the phase-I xenobiotic enzyme system. It may be up-regulated by PM exposure, and is a major source of ROS in cells stimulated with PM (40–43). AIFM2 (AMID) encodes a mitochondrial protein with significant homology with reduced nicotinamide adenine dinucleotide oxidoreductase and apoptosis-inducing factor. Up-regulation of this gene by PM has not been reported. AIFM2 may be one of the important sources of mitochondrial ROS produced by AMs during PM exposure (13, 15, 44), and is probably linked to a caspase-independent mitochondrial effector of apoptotic cell death.
The up-regulation of multiple MT1 isoform genes is intriguing. MTs are low-molecular weight, metal and sulfur-rich proteins that are widely distributed in all organs, including the lung (32, 45). These intracellular proteins are thought to be involved in both heavy metal detoxification and the homeostasis of essential trace metals, such as zinc and copper. The metal-binding property and the abundance of cysteines (18–23 residues) in MTs have also been implicated as the mechanisms by which MTs protect cells from oxidative injury (46–49). We have shown that addition of exogenous MT1 protein attenuated the extracellular resorufin signals. In human BECs that also showed up-regulation of MT1F and CYP1B1 gene expression after PM2.5 exposure similar to human AMs, knocking down MT1F gene expression by more than 50% was associated with an approximately 30% increase in PM2.5-induced H2O2 release. In addition, particles preincubated with MT1 before being added to the cells attenuated the release of H2O2. These results confirm the capability of MT1 to scavenge ROS and bind metals in human lung cells during PM exposure. MT1 may bind to redox-active metals associated with PM via its cysteine residues, and decrease hydroxyl radicals generated from the Fenton reaction. Binding of MT1 to PM may also interfere with the engulfment of MT1–PM complex by the AMs and inhibit downstream oxidant pathways. The abundant thiol groups present in MT1 also may directly scavenge oxygen radicals, independent of its metal-binding capability (49).
We further noted that approximately 30% of the 75 differentially expressed genes have molecular functions related to metal ion binding. Of these, the five MT1 isoforms are known to be essential in detoxifying environmental heavy metals, such as arsenic, cadmium, and mercury (50–53). MT1 levels in various tissues may also increase in animals exposed to environmental zinc (54, 55). The MT1 genes also are involved in binding of intracellular metals, such as zinc and copper (32, 56–58), and, thus, may regulate signaling mediated by these trace metals. ATOX1 encodes a copper chaperone that binds and transports cytosolic copper to ATPase proteins in the trans-Golgi network for later incorporation to the ceruloplasmin (59–61). Thus, our gene profiling study indicates that PM exposure may induce host defense mechanisms that are normally invoked to combat heavy metal toxicity, and that molecular signaling pathways mediated by zinc and copper may be important in PM-induced health effects.
Metals associated with PM likely were the primary inducer for gene changes, as metal-poor PM preparations did not induce MT1. Oxidative stress was unlikely the inducer for changes in gene expression, because PM2.5 at 1 μg/ml, which altered gene expression, did not increase H2O2 release. The induction of MT1 genes further indicates that certain metal components in PM may be the biologically active constituents for pulmonary health effects. MT1 genes may be responsive specifically to some heavy metals that are normally present in PM, such as zinc and copper (32, 50, 62). The PM2.5 sample used in this study contains approximately 500 ng/mg of zinc and 70 ng/mg of copper (5). Zinc is a major metal element detected in traffic-derived PM (63–66), presumably from tire wear. Incubation of human BECs with zinc up-regulated MT1 isoform genes (33). Copper in PM has also been linked to traffic sources, as copper is an element in car brakes (65, 67, 68). Zinc and copper may also come from certain industrial sources (63, 65). In a previous study exposing healthy volunteers to Chapel Hill PM2.5, the principal component analysis identified a factor containing zinc and copper correlating with increased blood fibrinogen (26). Also, in an in vitro study of cultured human BECs, the principal component analysis identified a copper-containing factor that was associated with Chapel Hill PM–induced IL-6 release (5).
AMs have an important role in clearing particles from the lungs (69). Binding and subsequent engulfment of particles may activate signaling pathways, resulting in production of ROS by various pro-oxidant enzymes, such as NADPH (16). Peroxinitrite may also be formed from the interaction between ROS and nitric oxide (25). PM effects may be recognized via binding receptors on AM, which activates downstream inflammatory pathways (70). Thus, interactions between PM and cell membrane proteins of AMs may be a critical event that triggers subsequent oxidative stress and inflammation. Our results demonstrate changes in the expression of genes encoding membrane proteins, such as PTPRE, MRC1, ADRB2, CLEC4A, ASGR1, MS4A4A, and MS4A6A. All of these genes, except PTPRE, were down-regulated. How PM-induced changes in these novel membrane proteins affect oxidative stress and inflammation, and thus health effects, will need to be investigated in future studies.
In summary, our gene profiling experiments identified an array of gene expression response in human AMs exposed to PM2.5, independent of oxidative stress. This profile is most notable for genes unique to heavy metal binding and oxidative stress. Metals associated with PM may be important in inducing host defense mechanisms normally invoked to combat heavy metal toxicity, and molecular signaling pathways mediated by metals may be important in PM-induced health effects.
This work was supported in part by the Intramural Research Program of the National Institutes of Health, National Heart, Lung, and Blood Institute, and the U.S. Environmental Protection Agency.
Originally Published in Press as DOI: 10.1165/rcmb.2008-0064OC on February 27, 2009
Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.