Sequestration of mitochondrial iron by silica particle initiates a biological effect

Andrew J. Ghio, Haiyan Tong, Joleen M. Soukup, Lisa A. Dailey, Wan-Yun Cheng, James M. Samet, Matthew J. Kesic, Philip A. Bromberg, Jennifer L. Turi, Daya Upadhyay, G. R. Scott Budinger, Gökhan M. Mutlu

Abstract

Inhalation of particulate matter has presented a challenge to human health for thousands of years. The underlying mechanism for biological effect following particle exposure is incompletely understood. We tested the postulate that particle sequestration of cell and mitochondrial iron is a pivotal event mediating oxidant generation and biological effect. In vitro exposure of human bronchial epithelial cells to silica reduced intracellular iron, which resulted in increases in both the importer divalent metal transporter 1 expression and metal uptake. Diminished mitochondrial 57Fe concentrations following silica exposure confirmed particle sequestration of cell iron. Preincubation of cells with excess ferric ammonium citrate increased cell, nuclear, and mitochondrial metal concentrations and prevented significant iron loss from mitochondria following silica exposure. Cell and mitochondrial oxidant generation increased after silica incubation, but pretreatment with iron diminished this generation of reactive oxygen species. Silica exposure activated MAP kinases (ERK and p38) and altered the expression of transcription factors (nF-κB and NF-E2-related factor 2), proinflammatory cytokines (interleukin-8 and -6), and apoptotic proteins. All of these changes in indexes of biological effect were either diminished or inhibited by cell pretreatment with iron. Finally, percentage of neutrophils and total protein concentrations in an animal model instilled with silica were decreased by concurrent exposure to iron. We conclude that an initiating event in the response to particulate matter is a sequestration of cell and mitochondrial iron by endocytosed particle. The resultant oxidative stress and biological response after particle exposure are either diminished or inhibited by increasing the cell iron concentration.

  • particulate matter
  • oxidants
  • inflammation
  • quartz

inhalation of suspended particulate matter (PM) has presented a challenge to human health for thousands of years (43). Several prominent causes of global deaths currently reported by the World Health Organization are positively associated with PM concentration (60); such particle-associated human mortality includes both respiratory (infections, chronic obstructive pulmonary disease, and respiratory cancers) and extrapulmonary disease (coronary heart disease and stroke and other cerebrovascular diseases). Although the sources, size distributions, and composition of inhaled PM vary widely, oxidative stress is considered fundamental in the biological effect (42). However, the specific mechanism responsible for the biological effect following PM exposure remains incompletely understood.

Following particle exposure, oxygen-containing functional groups at the PM surface provide a capacity to bind cell cations. In silica and silicate particles, these functional groups include silanol groups, whereas in ambient air pollution particles, diesel exhaust particles, wood stove particles, and PM associated with cigarette smoking and burning of biomass, such surface groups include alcohols, diols, epoxides, ethers, aldehydes, ketones, carboxylates, esters, and phenols. Among the cellular cations available for complexation by the particle surface, iron is abundant and kinetically preferred as a result of its electropositivity and high affinity for oxygen-containing functional groups. Retained PM consistently demonstrates a capacity to complex and accumulate host iron (e.g., the ferruginous body), providing in vivo evidence of an interaction between particle and endogenous metal (24). Furthermore, endpoints reflecting oxidative stress and biological effect correspond to elevated tissue iron concentrations following in vivo particle exposure (24).

It was proposed that iron, sequestered from host cells and complexed on the particle surface, supported oxidant generation and was responsible for biological function. However, cell exposure to particle with iron directly complexed to the surface demonstrated diminished biological effect relative to exposure to the particle itself (23). Accordingly, we tested the postulate that 1) particle binds and sequesters cell and mitochondrial sources of iron initiating a disruption of metal homeostasis, 2) particle-initiated iron loss from the cell and mitochondria increases superoxide production by the organelle facilitating ferrireduction and reversal of metal deficiency, and 3) oxidant generation following cell interaction with PM activates MAP kinases and transcription factors, resulting in a release of inflammatory mediators, inflammation, and apoptosis (Fig. 1).

Fig. 1.

Schematic for changes in iron homeostasis, oxidant generation, and biological effect following particle exposure. Functional groups at the surface of the particle sequester cell iron, including mitochondrial sources (A). In response to a reduction in intracellular iron, the cell generates superoxide as a ferrireductant and upregulates importers [e.g., divalent metal transporter 1 (DMT1)] and storage (ferritin) proteins in an attempt to reacquire requisite metal (B). The result of particle exposure is an altered iron homeostasis with an accumulation of the metal in the cell (C). If adequate iron concentrations are not imported, protracted oxidant generation activates MAP kinases and transcription factors, resulting in a release of inflammatory mediators, inflammation, and apoptosis (D).

METHODS

Materials.

All reagents are from Sigma (St. Louis, MO) unless specified otherwise. Silica was chosen as a prototypical particle for study as a result of its historical and public health significance and the ease with which its surface metal concentration can be manipulated. The silica used was Min-U-Sil (5 μm diameter from U.S. Silica, Berkeley Springs, WV). For selected endpoints, the effect of silica was compared with silica with surface-complexed iron. Silica and iron-loaded silica (silica-Fe) were prepared by washing Min-U-Sil in 1 N HCl for 1 h, agitating the particle in either H2O or 1,000 μM ferric ammonium citrate (FAC) for 1 h, centrifuging at 1,000 g for 10 min, washing in H2O, and dispersing into PBS at 20 mg/ml. Ionizable iron concentration of the two particles was quantified as that displaced in 1 N HCl after 1 h agitation using inductively coupled plasma optical emission spectroscopy (ICPOES; Model Optima 4300D; Perkin Elmer, Norwalk, CT) operated at a wavelength of 238.204 nm. The iron concentration associated with silica particle was below detectable limits while the silica-Fe particle contained 6.00 ± 0.58 μg/g; there was no release of iron over 24 h by the latter particle. After exposure to 500 μM 57Fe FAC for 1 h and washing, ionizable 57Fe on silica and silica-Fe were demonstrated to be 225 ± 17 and 149 ± 21 parts/billion by ICP mass spectroscopy (ICPMS; Elan DRC II; Perkin Elmer). This confirmed a diminished capacity of silica-Fe to coordinate additional sources of the metal onto its surface.

Cell culture.

Primary human bronchial epithelial (HBE) cells were studied after harvesting them from healthy, nonsmoking adult volunteers by cytological brushing during bronchoscopy on a protocol approved by the University of North Carolina Institutional Review Board. For those studies requiring transfection, BEAS-2B cells (an immortalized line of normal human bronchial epithelium) were employed. Cytotoxicity was assayed using release of lactic dehydrogenase and trypan blue exclusion. There was no significant cytotoxicity with 24 h exposure to 100 μg/ml silica.

Cell iron, zinc, and silicon concentrations.

HBE cells were exposed to 200 μM FAC, 100 μg/ml silica, both FAC and silica, 500 μM deferoxamine, and both FAC and deferoxamine. After 15 min and 4 h incubation, the media and exposure were removed, and the cells were washed with PBS and scraped into 1.0 ml 3 N HCl/10% trichloroacetic acid (TCA). After hydrolysis at 70°C for 24 h with precipitation of heme in the 10% TCA, iron (nonheme) and zinc concentrations in the supernatant were determined using ICPOES at λ = 238.204 and 206.200 nm, respectively. Zinc was quantified as a control metal.

In addition, 15 min, 30 min, 1 h, and 4 h following HBE cell incubation with 100 μg/ml silica, the media was removed, and the cells were washed with PBS and scraped in 1.0 ml concentrated HNO3 to digest both cells and associated particle. After hydrolysis for 24 h, silicon was measured using ICPOES (λ = 251.611 nm).

RT-PCR.

Total RNA was isolated using a Qiagen kit (Qiagen, Valencia, CA) and reverse transcribed to generate cDNA using a High Capacity cDNA Reverse Transcription kit (Applied Biosystems, Foster City, CA). Oligonucleotide primer pairs and fluorescent probes for divalent metal transporter 1 (DMT1), interleukin (IL)-6, IL-8, caspase 3, c-myc, and GAPDH were designed using a primer design program (Primer Express; Applied Biosystems) and obtained from Integrated DNA Technologies (Coralville, IA). Quantitative fluorogenic amplification of cDNA was performed using the ABI Prism 7500 Sequence Detection System (Applied Biosystems), primer/probe sets of interest, and TaqMan Universal PCR Master Mix (Applied Biosystems). The relative abundance of GAPDH mRNA was used to normalize mRNA levels.

Cell ferritin concentrations.

HBE exposures to 200 μM FAC, 100 μg/ml silica, both FAC and silica, 500 μM deferoxamine, and both FAC and deferoxamine were repeated using 24 h incubation. After the media was removed, cells were washed with PBS, scraped into 1.0 ml PBS, and disrupted using five passes through a gauge 25 needle. The concentrations of ferritin in the lysates were quantified using an immunoturbidimetric assay (Kamiya Biomedical, Seattle, WA).

Particle-associated metal concentrations after cell exposure.

HBE cells were grown in 75-cm2 flasks and exposed to 100 μg silica/ml for 1 h. After the media was removed, cells were washed with PBS and scraped into NaOCl solution. The lysate was centrifuged and treated again with NaOCl solution, and the particle was isolated, dried, and weighed. The particle from three different flasks was pooled, and the ionizable iron concentration was quantified as that displaced into 1 N HCl after 1 h agitation using ICPOES. Control exposure was a flask with media but no cells.

Mitochondrial iron and zinc concentrations.

HBE cells were grown in 75-cm2 flasks and exposed to 100 μg silica/ml and 500 μM deferoxamine for 5 min. Nuclear and mitochondrial fractions were isolated (26) and hydrolyzed in 1.0 ml 3 N HCl/10% TCA, and nonheme iron and zinc concentrations were measured using ICPOES. HBE cells in 75-cm2 flasks were then exposed to 1.0 μM 57Fe FAC for 4 h. The cells were washed with PBS and exposed to 100 μg silica/ml and 500 μM deferoxamine for 5 min. Nuclear and mitochondrial fractions were collected and hydrolyzed in 1 N HCl/10% TCA for 24 h. 57Fe in the fraction supernatants were measured using ICPMS. HBE cells were incubated in media with 200 μM FAC for 4 h and exposed to 100 μg silica/ml and 500 μM deferoxamine for 5 min. Nuclear and mitochondrial fractions were isolated and hydrolyzed in 1.0 ml 3 N HCl/10% TCA, and nonheme iron and zinc were measured using ICPOES. Finally, HBE cells were grown in 75-cm2 flasks, incubated in media with 200 μM FAC for 4 h, and exposed to 1.0 μM 57Fe FAC for 4 h. The cells were washed with PBS and exposed to 100 μg silica/ml and 500 μM deferoxamine for 5 min. Nuclear and mitochondrial fractions were collected and hydrolyzed in 1 N HCl/10% TCA for 24 h, and 57Fe in the fraction supernatants was measured.

Cellular oxidant generation.

Oxidant generation by HBE cells was determined using Amplex Red (Molecular Probes, Eugene, OR) fluorescence. Cells grown on 96-well Co-Star (Corning, Corning, NY) white-walled tissue culture plates to confluence were preloaded with the dye before exposure for 20 min at 37°C/5% CO2. The reported value is fold change over control cells that have been preloaded with dye and exposed to PBS only.

Mitochondrial oxidant generation.

A plasmid carrying dMitoHyPer, a genetically encoded probe with submicromolar sensitivity and specificity for hydrogen peroxide, was purchased from Evrogen (pHyPer-dMito; Axxora, San Diego, CA). HyPer was created by an insertion of yellow fluorescent protein into the regulatory domain of the H2O2-sensitive Escherichia coli transcription factor OxyR and has been shown to be insensitive to other oxidants, such as superoxide, nitric oxide, and peroxynitrite. MitoHyper includes a duplicated mitochondrion targeting sequence present in the coding sequence for subunit VIII of human cytochrome c oxidase. dMitoHyPer was transfected in BEAS-2B cells using Fugene 6 (Roche, Indianapolis, IN) according to the manufacturer's protocol. Analyses were carried out on a Nikon Eclipse C1Si confocal microscope equipped with a TE 2000 microscope and quantified with Nikon EZ-C1 and Nikon Elements software (Nikon Instruments, Melville, NY). The ratio of fluorescence intensity emitted under 488 and 404 nm excitations was used for the relative quantification of H2O2 concentrations.

MAP kinase activation.

Protein kinase phosphorylation was analyzed by Western blotting. HBE cells were incubated with and without 200 μM FAC for 4 h. After being washed with PBS, cells were exposed to 100 μg/ml silica for 60 min and lysed with RIPA lysis buffer. After normalization for protein content, cell extracts were subjected to SDS-PAGE on 11% gradient PAGE gels with a Tris-glycine-SDS buffer. Proteins were then electroblotted onto nitrocellulose. The membranes were incubated with nonfat milk, washed briefly, and incubated with 1:1,000 phosphospecific ERK and p38 antibodies (Cell Signaling Technology, Danvers, MA) in 5% bovine serum albumin overnight at 4°C followed by incubation with 1:2,000 HRP-conjugated secondary antibody for 1 h at room temperature. Bands were detected using enhanced chemiluminescence (ECL) detection reagents 1 and 2 and high-performance chemiluminescence ECL films (Amersham Pharmacia Biotech) with a model SRX-101 Konica medical film processor (Konica).

Transcription factor activation.

To measure transactivation of NF-κB and NF-E2-related factor 2 (Nrf2) ARE promoters, BEAS-2B cells were cotransfected with either 0.1 μg of κB-luciferase reporter or Nrf2 ARE-luciferase reporter plasmid along with 0.02 μg of thymidine kinase-Renilla luciferase using Fugene 6 (Roche) according to the manufacturer's recommendations. The total amount of DNA was kept constant. After 24 h, cells were incubated with and without 200 μM FAC for 4 h and then exposed to PBS or 100 μg silica/ml PBS for 1 h. Cells were then pelleted and lysed in passive lysis buffer (Promega, Madison, WI). The transactivation activity was measured as luciferase light units as described previously (32).

IL-6 and IL-8 release.

IL-6 and IL-8 concentrations in cell media were measured using ELISA (R&D Systems, Minneapolis, MN).

Indexes of apoptosis.

Uptake after Hoechst staining was used as an index of apoptosis; wavelengths for excitation and emission were 350/50 and 460/50 (Chroma Technology, Bellows Falls, VT). RT-PCR was performed to detect RNA for caspase 3 and c-myc. In addition, immunoblotting and detection of caspase-3 and Bax was performed (55). The primary antibody used was from Cell Signaling Technology.

Measurement of mitochondrial membrane potential.

Mitochondrial membrane potential was monitored using the fluorescent indicator JC-1 (Invitrogen, Carlsbad, CA). In the presence of physiological mitochondrial membrane potentials, JC-1 forms aggregates that fluoresce with an emission peak at 588 nm. Loss of membrane potential favors the monomeric form of JC-1 that has an emission peak at 530 nm. Cells were labeled with 5 μM JC-1 in DMEM supplemented with 10% FBS and 1 μg/ml gentamicin at 37°C. After 15 min incubation, cells were washed with PBS two times and placed in the chamber with 500 μl PBS glucose. JC-1 fluorescence intensity was monitored with dual excitation at 488 and 561 nm and an emission scan range of 490–650 nm (32 channels, 5 nm/channel). Mitochondrial membrane potential was inferred from the ratio of fluorescence intensity of emission maximum at 593 and 538 nm, representing the J-aggregate and monomeric form, respectively.

Animal exposures.

Experiments utilized protocols reviewed and approved by the Institutional Animal Care and Use Committees at the National Health and Environmental Effects Research Laboratory, The treatment and care of animals was conducted under the direction of the Institute for Animal Care and Use Committee, National Health and Environmental Effects Research Laboratory, Environmental Protection Agency (Research Triangle Park, NC). Sixty-day-old (250–300 g) male Sprague-Dawley rats were housed in temperature- and humidity-controlled rooms and fed a standard diet (Rat Chow 5001; Ralston Purina, St. Louis, MO). After anesthesia with 2–5% halothane (Aldrich Chemicals, Milwaukee, WI), animals were intratracheally instilled with: 0.5 ml HBSS (n = 12), 500 μg silica in 0.5 ml HBSS (n = 12), or 500 μg silica-Fe in 0.5 ml HBSS (n = 12). Twenty-four hours after exposure, rats were again anesthetized with halothane, killed by exsanguination, and tracheally lavaged (35 ml/kg body wt). The percentage neutrophils and protein concentration (Pierce Coomassie Plus Protein Assay Reagent; Pierce, Rockford, IL) were determined.

Animal exposures were repeated with tracheal instillations including: 1) 0.5 ml HBSS followed within 15 s by 0.5 ml HBSS (n = 12), 2) 0.5 ml 1,000 μM FAC in HBSS followed by 0.5 ml HBSS (n = 12), 3) 0.5 ml HBSS followed by 500 μg silica in 0.5 ml HBSS (n = 12), or 4) 0.5 ml 1,000 μM FAC in HBSS followed by 500 μg silica in 0.5 ml HBSS (n = 12). Twenty-four hours after exposure, rats were anesthetized, killed, and tracheally lavaged. Percentage neutrophils and protein concentrations were determined.

Statistics.

Data are expressed as mean values ± SE unless specified otherwise. The minimum number of replicates for all measurements was three; those experiments using HBE cells were collected from at least three different volunteers. Differences between two and multiple groups were compared using t-tests of independent means and one-way analysis of variance, respectively. The post hoc test employed was Duncan's Multiple-Range test. Two-tailed tests of significance were employed. Significance was assumed at P < 0.05.

RESULTS

Relative to HBE cells exposed to PBS, those exposed to 100 μg/ml silica for 15 min showed no detectable cellular silicon content as measured by ICPOES after cell digestion in concentrated HNO3 (Table 1). This lack of a measurable cell silicon concentration suggests an absence of significant particle endocytosis before 15 min. Cell nonheme iron concentration was decreased following 15 min silica exposure (0.25 ± 0.02 and 0.17 ± 0.02 parts/million for PBS and silica, respectively). At exposure times greater than 15 min, particle endocytosis was evidenced by increased cell silicon levels (Table 1). Four hour exposure to 200 μM FAC confirmed iron import by HBE cells (Fig. 2A). In contrast to 15 min exposure, nonheme iron concentrations in HBE cells exposed to 100 μg/ml silica for 4 h increased with the source of the imported metal being the media (initial iron concentration in the media was ∼10 μM iron). Coincubation of HBE cells with both silica and 200 μM FAC was associated with a greater accumulation of iron relative to either FAC or silica exposure alone. Deferoxamine exposure decreased cell iron levels (Fig. 2A). Cell zinc concentrations were unchanged with all exposures.

View this table:
Table 1.

Silicon concentrations in HBE cells

Fig. 2.

Cell and mitochondrial iron concentrations after exposures to silica and deferoxamine. A: with ferric ammonium citrate (FAC) exposure, cells imported iron (*). Inclusion of silica also elevated (**) cell iron concentrations, and both FAC and silica increased metal levels further (***). Incubation with deferoxamine diminished baseline iron levels (+) and blunted the increase in nonheme iron after FAC exposure (++). There were no differences in cell zinc with any exposure. B: exposures were repeated for 24 h, and ferritin was measured in cell lysate using an immunoturbidimetric assay. Exposures to FAC (*), silica (**), and both (***) increased cell ferritin concentration. Deferoxamine did not decrease baseline cell ferritin while inclusion of both the chelator and FAC modestly increased (+) levels of the storage protein. C: human bronchial epithelial (HBE) cells in flasks were exposed to silica and deferoxamine for 5 min. After silica exposure, the nuclear fraction showed elevated nonheme iron concentrations after silica exposure (*), but metal was diminished in the mitochondrial fraction (**). Deferoxamine decreased iron in both fractions. Zinc was unchanged in any fraction after silica and deferoxamine. D: cells were incubated with 57Fe FAC and exposed to silica and deferoxamine. Collected fractions were hydrolyzed, and ICP mass spectroscopy was employed to measure 57Fe. After silica exposure, 57Fe concentrations were increased in the nuclear fraction (*) and decreased in the mitochondrial fraction (**). Deferoxamine decreased 57Fe in both fractions. E: HBE cells were incubated with FAC and exposed to silica and deferoxamine. Silica exposure increased iron concentrations in the nuclear fraction (*) but had no effect on levels of mitochondrial metal. Deferoxamine decreased iron in the nuclear fraction (**). F: finally, cells were incubated with FAC and with 57Fe FAC, and the exposures were repeated. Relative to PBS exposure, there were no differences in 57Fe in either the nuclear or the mitochondrial fraction after silica. Deferoxamine decreased the 57Fe in the nuclear fraction (**).

The response of specific iron-related proteins to exposure was assessed. After 4 h exposure to 100 μg/ml silica and 500 μM deferoxamine, there were 374 ± 105 and 633 ± 149% increases, respectively, for DMT1 RNA. This change in RNA for a major iron importer supports a reduction in intracellular iron following metal sequestration by the endocytosed particle. Following 24 h silica exposure, cell ferritin concentrations increased and were further elevated with coincubations of silica and FAC (Fig. 2B). Deferoxamine had no significant effect on cell ferritin (Fig. 2B).

Appropriation of host iron by silica following in vivo exposure has been demonstrated (24). After exposure of HBE cells to 100 μg/ml silica for 1 h, isolation of the particle demonstrated a concentration of iron that was measurable (0.20 ± 0.01 μg/g) and significantly elevated relative to that with no cell exposure (below detectable limits). Although the silica has increased levels of iron following endocytosis, it is not saturated with iron (6.00 ± 0.58 μg/g).

Mitochondria are central to the metabolism of iron (45). Interactions between PM exposure and mitochondrial nonheme iron concentration were therefore examined. Nonheme iron concentrations in the mitochondrial fraction decreased following 5 min exposure to both 100 μg/ml silica and 500 μM deferoxamine (Fig. 2C). After PM exposure, iron in the nuclear fraction increased. In contrast, exposure to deferoxamine diminished iron concentrations in the nuclear fraction (Fig. 2C). The stable isotope 57Fe was then used as a tracer to better define changes after particle and deferoxamine exposures. HBE cells were incubated with 1.0 μM 57Fe FAC for 4 h, washed with PBS, and then exposed for 5 min to either silica or deferoxamine. ICPMS analyses confirmed decrements in mitochondrial 57Fe in cells incubated with silica and deferoxamine (Fig. 2D). HBE cells were pretreated with 200 μM FAC for 4 h, the media were removed, and the cells were exposed to silica and deferoxamine for 5 min. FAC treatment increased iron concentrations in both nuclear and mitochondrial fractions, and there were no effects of either silica or deferoxamine on mitochondrial nonheme levels (Fig. 2E). Cells were then pretreated with 200 μM FAC for 4 h, incubated with 1.0 μM 57Fe FAC for 4 h, and exposed to silica and deferoxamine for 5 min. After fractionation, there were no decrements in mitochondrial nonheme 57Fe observed in those incubations that included silica and deferoxamine (Fig. 2F). Cell pretreatment with FAC precluded the decrement in mitochondrial iron following exposures to both silica and deferoxamine.

A fluorescence-based method showed time- and concentration-dependent oxidant generation by HBE cells following silica and deferoxamine exposures (Fig. 3, A and B, respectively). The fluorescence-based signal following HBE cell exposure to silica-Fe was decreased relative to that after silica (Fig. 3C). HBE cells were then pretreated with 200 μM FAC and exposed to either PBS or particle. Rather than increasing oxidant generation, cell pretreatment with FAC decreased oxidant production after exposure to both PBS and silica (Fig. 3D). In a comparable manner, oxidant generation after exposure to either PBS or deferoxamine was markedly inhibited by FAC pretreatment (Fig. 3E). Oxidant generation provoked by silica (Fig. 3F) and deferoxamine (Fig. 3G) was partially inhibited by both rotenone and diphenyleneiodonium chloride, consistent with some dependence on mitochondrial function. Finally, mitochondrial oxidant generation following particle exposure was confirmed in BEAS-2B cells using dMitoHyPer, a genetically encoded probe specific for hydrogen peroxide detection (Fig. 3H). Cell treatment with FAC before challenge with silica also inhibited this specific measure of mitochondrial oxidant generation.

Fig. 3.

Cell and mitochondrial oxidant generation after exposures to silica and deferoxamine. Oxidant generation was determined using Amplex Red fluorescence. The reported value is fold change over control cells, which were exposed to PBS only. Oxidant generation after silica (A) and deferoxamine (B) exposures of HBE cells was both dose and time dependent. Incubation of cells with silica-Fe caused less oxidant generation relative to the same dose of silica (C). Incubation of cells with FAC diminished the oxidant generation following exposure to PBS and both silica (D) and deferoxamine (E). In a similar manner, pretreatment of cells with rotenone and diphenyleneiodonium chloride (DPI) inhibited oxidant generation after silica (F) and deferoxamine (G). Mitochondrial oxidant generation, measured using BEAS-2B cells transfected with dMitoHyPer (Evrogen; Axxora, San Diego, CA), increased with silica exposure, whereas pretreatment with FAC inhibited this response (H). With the exception of silica exposure, all groups coalesce into a single tracing. Arrows successively refer to addition of silica, 1.0 mM H2O2, and 10 mM dithiothreitol.

Biological effects after PM exposure are associated with an oxidative stress (40, 50). In this study, oxidant generation after silica exposure was dependent on decreased mitochondrial iron. A relationship between particle effect and endogenous iron was subsequently examined by augmenting cell iron concentrations before PM exposure. Phosphorylation of both ERK and p38, MAP kinases activated by iron chelators (6, 34), was observed in HBE cells exposed to silica (Fig. 4A). Pretreatment of the cells with FAC prevented the phosphorylation of ERK and p38 following particle exposure (Fig. 4A). To quantify transcriptional activation by silica, BEAS-2B cells were transfected with either NF-κB or Nrf2 ARE promoter-luciferase construct. Silica exposure increased NF-κB activation at least 500% in BEAS-2B cells (Fig. 4B). Whereas cell preincubation with FAC had no effect on NF-κB activation after HBSS exposure only, it diminished subsequent NF-κB activation induced by silica (Fig. 4B). In a similar manner, particle exposure elevated Nrf2 ARE activation, and this was prevented by FAC pretreatment of the BEAS-2B cells (Fig. 4C).

Fig. 4.

Activation of MAP kinases and transcription factors after exposures to silica and deferoxamine. While silica increased phosphorylation of both ERK and p38 in respiratory epithelial cells, preincubation with FAC diminished this response (A). Similarly, BEAS-2B cell incubation with FAC decreased activation of both NF-κB and NF-E2-related factor 2 (Nrf2) ARE (B and C, respectively). *Increased relative to media; **increased relative to media but decreased relative to silica or deferoxamine.

NF-κB or Nrf2 ARE influence the expression of genes involved in inflammation. Therefore, the release of inflammatory mediators after particle exposure was assessed. Silica exposure increased IL-8 and IL-6 RNA (Fig. 5, A and B) and protein (Fig. 5, C and D) in HBE cells. Cell pretreatment with FAC diminished silica-induced changes in both IL-8 and IL-6 mRNA and protein expression (Fig. 5, AD). Deferoxamine exposure of HBE cells resulted in a comparable effect with increases in cytokine RNA and protein levels that were inhibited by FAC pretreatment (Fig. 5, AD). Incubation with silica and silica-Fe demonstrated an analogous effect; silica increased both IL-8 and IL-6, whereas silica-Fe showed a diminished response (Fig. 5, E and F).

Fig. 5.

Release of proinflammatory mediators after exposures to silica and deferoxamine. Silica and deferoxamine increased both RNA for interleukin (IL)-6 (A) and IL-8 (B) and release of these mediators by HBE cells (C and D). FAC pretreatment diminished the response after both silica and deferoxamine. In a comparable manner, silica-Fe decreased levels of IL-6 and IL-8 release relative to silica (E and F, respectively). *Increased relative to media; **increased relative to media but decreased relative to silica or deferoxamine.

Apoptotic endpoints were analyzed after silica exposure of HBE cells. Silica induced nuclear condensation and fragmentation associated with the apoptotic process, whereas pretreatment with FAC inhibited these nuclear changes (Fig. 6A). Similarly, changes in RNA for caspase-3 and c-myc and protein expression of caspase-3 and BAX were decreased by FAC pretreatment (Fig. 6, BD). Taken together, these data suggest that silica incites apoptosis, in part, by depleting intracellular iron stores.

Fig. 6.

Apoptotic endpoints after exposures to silica and deferoxamine. Relative to PBS (A, top left; magnification approximates ×630), exposure to silica caused nuclear condensation and fragmentation in Hoechst staining of the cells (A, top right). While preincubation of the cells with FAC had no effect on this staining (A, bottom left), it inhibited evidence of apoptosis after silica exposure (A, bottom right). RNA for proapoptotic proteins caspase-3 and c-myc both increased after 4 h silica exposure (B and C, respectively); preincubation with FAC also decreased these elevations. Western blot analyses for caspase-3 and the proapoptotic BAX protein (with lanes moved to reflect the order provided in Fig. 3A) showed elevations of expression after silica and decrements with FAC pretreatment (D and E, respectively). *Increased relative to media; **increased relative to media but decreased relative to silica or deferoxamine.

The relationship between biological effect of PM and mitochondrial iron homeostasis was further explored. Mitochondrial permeability transition (MPT) is an early event in the response to adverse exposures. Measurement of mitochondrial membrane potential with JC-1 confirmed depolarization with MPT in HBE cells following particle exposure (Fig. 7A). Pretreatment with FAC had no effect on MPT after silica exposure. The permeability transition pore mediates some portion of mitochondrial depolarization, and this multiprotein complex between the inner and outer membranes includes voltage-dependent anion channel 1 (VDAC1) (54). VDAC1 is a major route for superoxide release by cells that can function in ferrreduction (3, 28, 53, 59). Inhibition of VDAC1 by the stilbene 4-acetamido-4′-isothiocyanatostilbene-2,2′-disulfonic acid (SITS) decreased HBE cell iron import following PM exposure (Fig. 7B). Similarly, SITS diminished iron transport to the mitochondria after incubations that included silica (Fig. 7C). Inclusion of SITS in cell incubations also diminished generation of oxidants after particle exposure (Fig. 7D). Finally, with decreased oxidant generation, SITS significantly diminished release of IL-6 and IL-8 release by HBE cells following particle exposure (Fig. 7, E and F).

Fig. 7.

Mitochondrial membrane potential and permeability transition pore involvement in the response to silica. Silica significantly altered mitochondrial membrane potential measured using the fluorescent indicator JC-1 (A). There was no effect of pretreatment of HBE cells with 200 μM FAC for 4 h on mitochondrial membrane potential. Four-hour coincubations of PBS, 200 μM FAC, and 100 μg silica demonstrated increased import of iron following particle exposure (B); intracellular transport of iron was diminished by inclusion of 100 μM 4-acetamido-4′-isothiocyanatostilbene-2,2′-disulfonic acid (SITS). Four hours of incubation of HBE cells in 70-cm2 flasks to the same exposures revealed metal import into mitochondria after silica (C); these increases in organelle iron concentration were decreased using SITS. Oxidative stress by HBE cells, measured as Amplex Red fluorescence, after silica exposure was inhibited by inclusion of SITS (D). Finally, HBE cell release of proinflammatory mediators IL-6 and IL-8 after silica exposure was significantly diminished by inclusion of SITS (E and F, respectively). *Increased relative to PBS only; **decreased relative to incubations with no SITS.

Sprague-Dawley rats were intratracheally instilled with HBSS, 500 μg silica in HBSS, or 500 μg silica-Fe in HBSS. Tracheal lavage 24 h after exposure showed a neutrophilic inflammation in those animals exposed to silica (Fig. 8, A and B); this injury decreased following exposure to an equivalent mass of silica-Fe. Animals were then exposed to 1) 0.5 ml HBSS followed within 15 s by 0.5 ml HBSS, 2) 0.5 ml 500 μM FAC in HBSS followed by 0.5 ml HBSS, 3) 0.5 ml HBSS followed by 500 μg silica in 0.5 ml HBSS, or 4) 0.5 ml 500 μM FAC in HBSS followed by 500 μg silica in 0.5 ml HBSS. Instillation of HBSS followed by silica was associated with an increased neutrophilic influx and total protein relative to exposure to HBSS alone (Fig. 8, C and D). FAC treatment of the animals diminished the inflammatory injury following silica exposure. These in vivo data support the position that silica disrupts iron homeostasis to produce a biological effect.

Fig. 8.

In vivo neutrophilic lung injury following exposure to silica. Twenty-four hours after intratracheal exposure of animals to silica, there were elevated values of percentage neutrophils (A) and protein (B) in the lavage reflecting tissue injury. Exposure to a comparable mass of silica-Fe also was associated with injury, but both neutrophils and protein were significantly decreased relative to those observed after silica. Finally, animals were then exposed to 1) HBSS followed by HBSS (HBSS/HBSS), 2) FAC followed by HBSS (FAC/HBSS), 3) HBSS followed by silica (HBSS/Silica), or 4) FAC followed by silica (FAC/Silica). Lavage neutrophils (C) and protein (D) were both elevated after silica exposure, but this injury was significantly diminished with coexposure to FAC. *Increased relative to HBSS; **increased relative to HBSS but decreased relative to either silica or HBSS/Silica.

DISCUSSION

Ferric ion reacts with surface silanol groups of silica to produce a silicato-iron coordination complex (16). After endocytosis of the particle, nonheme iron concentrations in exposed HBE cells increased with the source of the imported metal being the media. Coincubation of HBE cells with both silica and FAC was associated with a greater accumulation of iron relative to either FAC or silica exposure alone. These results suggest a rapid entry of the particle into the cell, which presents a potential for sequestration of intracellular iron comparable to that induced by treatment of the cells with the extracellular metal chelator deferoxamine. Following in vitro exposure, the concentration of iron associated with the silica significantly increased, supporting a capacity to sequester the metal from cell sources. In response to this loss of metal by the cell to the particle, RNA for the importer DMT1 increased, and intracellular iron transport was elevated to meet increased cell requirements. A new iron homeostasis was achieved with augmented cell iron and ferritin concentrations at 4 and 24 h, respectively. Elevated cell iron concentrations after silica exposure allowed continued metal availability to the cell, allowing its continued function and survival despite sequestration of essential iron by the particle. The specific sources of iron sequestered by PM and pathways of iron export from mitochondria were not defined. Increased cell iron and ferritin concentrations after silica exposure differed from those observed after incubation with deferoxamine, which similarly appropriated cell iron. In contrast to silica particles, deferoxamine is precluded from moving intracellularly in substantial quantities and, consequently, iron did not accumulate in the cell with inclusion of this chelator but rather decreased (44).

Whereas elevated concentrations of PM-bound iron increased oxidant generation in an acellular system, augmented metal availability had the opposite effect on such production by cells and mitochondria. The mitochondrial electron transport chain is a major location of oxidant production (41); specific sites include complex I and complex III (11, 27). Cellular oxidant generation, specifically superoxide, follows exposure to iron chelators (6, 9, 14). The evidence from this study suggests that decreased cellular and mitochondrial iron concentrations, following exposure to either silica or deferoxamine, induce a mitochondrial response that includes oxidant generation. It is proposed that this mitochondrial oxidant production after PM exposure functions in the remedial response to iron loss following sequestration of the metal by the particle surface. Superoxide, produced by mitochondria in response to metal deficit, may assist in the import of required iron. Ferrireduction is an essential, and frequently limiting, reaction in such iron import and can be achieved in many cell types using superoxide (5, 47, 57). Furthermore, such ferrireduction can be accomplished with the electron transport chain in the mitochondria serving as the ultimate source of reducing equivalents (4, 46).

Biological effects after PM exposure are associated with oxidative stress (40, 50). In this study, oxidant generation after silica exposure was dependent on decreased mitochondrial iron. A relationship between particle effect and endogenous iron was subsequently examined by augmenting cell iron concentrations before PM exposure. Oxidant-induced activation of the MAP kinase cascade represents a signaling pathway by which PM exposure mediates specific biological effects (49). MAP kinase phosphorylation follows both silica exposure (1, 51) and reduction in iron levels (62). Phosphorylation of ERK and p38 following silica exposure was diminished by increasing the cell concentration of available iron. Similarly, silica activates transcription factors, including NF-κB and Nrf2 ARE (10, 18); the data suggest that the particle induces an inflammatory and adaptive response. The activation of these nuclear transcription factors by silica exposure was decreased by augmenting cell iron. These nuclear transcription factors control the activity of genes involved in both inflammation and apoptosis (10, 39, 52, 61). Changes in RNA and protein expression for IL-6 and IL-8 after silica exposure were diminished by cell pretreatment with iron. All of these results demonstrate a relationship of biological effect by particle with both iron homeostasis and mitochondrial function suggested in previous studies (15, 23).

Iron is essential for cell growth and viability, and its decreased availability (following exposure to chelators, metals, serum deprivation, and anti-transferrin receptor antibodies) leads to apoptosis (20, 31, 56). Iron chelators halt cell cycle progression at the S phase and engage MAP kinases, caspase 3, c-myc, and proto-oncogene proteins in the apoptotic process (21, 22, 30). Silica also appeared to provoke apoptosis by inducing an intracellular deficit in iron with alteration of apoptotic indexes after particle exposure inhibited by excess iron. It is likely that increased iron availability provides an intracellular rescue from metal deficiency and subsequent growth inhibition and apoptosis (36).

This proposed pathway for biological effect of PM was further investigated with examination of MPT involvement in the response to silica and then blocking VDAC1, which participates in this reaction. SITS diminished oxidant generation and iron import into the cells and mitochondria presumably by decreasing superoxide export; this function of ferrireduction by VDAC1 enables iron import, which requires a reduction of Fe3+ to Fe2+ for movement across a membrane (3, 8, 25, 28, 53, 54, 59). Previous investigation supports some participation of the electron transport chain in iron uptake comparable to our results (53).

Animals instilled with silica have repeatedly been described to suffer a neutrophilic alveolitis (19). The alveolitis following silica was significantly greater than that after silica-Fe, and coexposure of the animal to FAC similarly decreased the inflammatory response to silica. These results conflict with previous investigation (7) but are consistent with others′ in vitro results (23). These in vivo data also support the position that silica particle disrupts iron homeostasis to produce biological effect, including injury.

We conclude that an initiating event in oxidative stress and biological effect after exposure to a particle (i.e., silica) is a sequestration of cell and mitochondrial iron. Particle-initiated iron loss from the cell and mitochondria increases superoxide production by the organelle, possibly involving production of the oxidant by proteins included in the electron transport chain, which then facilitates ferrireduction and reversal of metal deficiency. If adequate iron concentrations are not imported, protracted oxidant generation activates MAP kinases and transcription factors, resulting in a release of inflammatory mediators, inflammation, and apoptosis. After particle exposure, inflammation and apoptosis are the products of a deficiency of requisite iron comparable to cell incubation with a chelator (2, 12, 13, 29, 33, 35, 37, 38). It is possible that sources of cell iron other than those in the mitochondria are both available to and appropriated by the particle. Increasing the cell and mitochondrial iron concentration by incubation with FAC protected against both the oxidative stress associated with PM exposure and biological effect. The same pathway for disruption in cell iron homeostasis, oxidative stress, and biological effect may also participate in the response to xenobiotics, fibers, and metals (17, 48, 58).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

Author contributions: A.J.G., H.T., J.M. Samet, M.J.K., J.L.T., D.U., G.S.B., and G.M.M. conception and design of research; A.J.G., H.T., J.M. Soukup, L.A.D., W.-Y.C., J.M. Samet, and M.J.K. performed experiments; A.J.G., H.T., J.M. Soukup, L.A.D., W.-Y.C., J.M. Samet, M.J.K., P.A.B., J.L.T., G.S.B., and G.M.M. analyzed data; A.J.G., H.T., J.M. Soukup, L.A.D., W.-Y.C., J.M. Samet, M.J.K., P.A.B., J.L.T., G.S.B., and G.M.M. interpreted results of experiments; A.J.G., H.T., J.M. Soukup, L.A.D., W.-Y.C., J.M. Samet, M.J.K., P.A.B., J.L.T., G.S.B., and G.M.M. prepared figures; A.J.G., H.T., J.M. Soukup, L.A.D., W.-Y.C., J.M. Samet, M.J.K., P.A.B., J.L.T., D.U., G.S.B., and G.M.M. drafted manuscript; A.J.G., H.T., J.M. Soukup, L.A.D., W.-Y.C., J.M. Samet, M.J.K., P.A.B., J.L.T., D.U., G.S.B., and G.M.M. edited and revised manuscript; A.J.G., H.T., J.M. Soukup, J.M. Samet, M.J.K., P.A.B., J.L.T., D.U., G.S.B., and G.M.M. approved final version of manuscript.

REFERENCES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
View Abstract