Chronic Arsenic Exposure in Nanomolar Concentrations Compromises Wound Response and Intercellular Signaling in Airway Epithelial Cells

Abstract

Paracrine ATP signaling in the lung epithelium participates in a variety of innate immune functions, including mucociliary clearance, bactericide production, and as an initiating signal in wound repair. We evaluated the effects of chronic low-dose arsenic relevant to U.S. drinking water standards (i.e., 10 ppb [130nM]) on airway epithelial cells. Immortalized human bronchial epithelial cells (16HBE14o-) were exposed to 0, 130, or 330nM arsenic (as Na-arsenite) for 4–5 weeks and examined for wound repair efficiency and ATP-mediated Ca²⁺ signaling. We found that chronic arsenic exposure at these low doses slows wound repair and reduces ATP-mediated Ca²⁺ signaling. We further show that arsenic compromises ATP-mediated Ca²⁺ signaling by altering both Ca²⁺ release from intracellular stores (via metabotropic P2Y receptors) and Ca²⁺ influx mechanisms (via ionotropic P2X receptors). To better model the effects of arsenic on ATP-mediated Ca²⁺ signaling under conditions of natural exposure, we cultured tracheal epithelial cells obtained from mice exposed to control or 50 ppb Na-arsenite supplemented drinking water for 4 weeks. Tracheal epithelial cells from arsenic-exposed mice displayed reduced ATP-mediated Ca²⁺ signaling dynamics similar to our in vitro chronic exposure. Our findings demonstrate that chronic arsenic exposure at levels that are commonly found in drinking water (i.e., 10–50 ppb) alters cellular mechanisms critical to airway innate immunity.

Arsenic is recognized as an environmental human health hazard, with chronic ingestion of contaminated drinking water being the major source of exposure (ASTDR, 2007). Drinking-water arsenic has been correlated with both malignant and nonmalignant lung disease (reviewed by Guha Mazumder [2007], Kapaj et al. [2006], and Schuhmacher-Wolz et al. [2009]). Exposure to drinking-water arsenic has been shown to increase the lung diseases bronchiectasis and chronic obstructive pulmonary disease (COPD), decrease lung function, and increase respiratory symptoms (Dauphiné et al., 2011; Guo et al., 2007; Mazumder et al., 2000; Milton and Rahman, 2002; Parvez et al., 2008; Smith et al., 1998, 2006). Although the majority of evidence for arsenic-induced lung disease in both epidemiologic and laboratory reports comes from relatively high levels of exposure (i.e., > 100 ppb [or µg/l]), evidence is accumulating that chronic low-dose (i.e., ≤ 50 ppb) arsenic can also have detrimental effects on lung biology (Kozul et al., 2009b; Lantz et al., 2007). In fact, mice exposed chronically to low-dose arsenic-supplemented drinking water had alterations in lung gene expression at concentrations as low as 0.1 ppb arsenic (Andrew et al., 2007). Based on human health risks for arsenic, both the U.S. Environmental Protection Agency (EPA) and the World Health Organization reduced maximum contaminant levels of arsenic from 50 ppb to 10 ppb effective in 2006 for the United States (U.S. EPA, 2001).

Despite strong associations between arsenic exposure and respiratory illness, the pathophysiological mechanisms by which arsenic acts on the lung remain largely unknown. A role for drinking-water arsenic in altered immune function in the lung that may help to explain some of the nonmalignant lung pathology has emerged from multiple studies (Kozul et al., 2009a, b; Liao et al., 2011; Rahman et al., 2011; Raqib et al., 2009; Smith et al., 2011). Loss of innate immune defense can lead to chronic bacterial infections and eventual obstructive airway disease, such as COPD and bronchiectasis (Hogg et al., 2004; King, 2009; Randell and Boucher, 2006). At the cellular and tissue level, purinergic receptors have been shown to contribute to several innate immune defense mechanisms in the airway epithelium; these include ciliary beat, mucin secretion, bactericide production, and epithelial wound repair (Davis and Lazarowski, 2008; Fischer, 2009; Schwiebert and Zsembery, 2003; Wesley et al., 2007; Zsembery et al., 2003). Defects in wound repair can compromise the basic barrier function of the airway epithelium (Crosby and Waters, 2010; Puchelle et al., 2006). We have shown that acute (24h) arsenic in the micromolar range (0.8–3.9µM; 60–290 ppb) alters wound repair mechanisms and signaling pathways critical to lung innate immune defense in airway epithelial cells (Olsen et al., 2008; Sherwood et al., 2011). From these studies, we identified MMP-9 secretion and functional changes in purinergic receptor (P2Y₂ and P2X)–dependent Ca²⁺ signaling as mechanistic targets of acute micromolar arsenic exposure. Although these findings strongly suggest intercellular signaling as a target for arsenic toxicity, the levels and timing of arsenic application make it difficult to interpret how these findings might fit in with the most prevalent source of arsenic in environmental exposures, via drinking water.

In this report, we hypothesized that arsenic exposure at levels consistent with those commonly found in U.S. drinking water supplies (i.e., 10–50 ppb) when applied chronically is sufficient to compromise components of airway epithelial wound repair and signaling. To test this hypothesis, we used an in vitro model in which we first exposed human airway epithelial cells (16HBE14o-) chronically (4–5 weeks) to low-dose arsenic at concentrations of 130nM (10 ppb) or 330nM (25 ppb). We complemented this approach with an in vivo exposure model where primary cultured cells were obtained from mice fed arsenic-free or arsenic-supplemented (50 ppb) drinking water for 4 weeks. In both models, and independent of the route of exposure, we found that fundamental ATP-mediated Ca²⁺ signaling mechanisms were compromised by arsenic exposure. We propose that arsenic-induced disruption of paracrine ATP signaling in the airway plays a role in compromised airway innate immunity under chronic low-dose conditions. In turn, arsenic exposure, even at very low levels, may lead to increased lung infections and the potential for nonmalignant respiratory disease.

MATERIALS AND METHODS

Materials.

Minimum essential medium with Earle’s salts (MEM), Lechner and LaVeck basal media (LHC), Hanks’ Balanced Saline Solution, glutamax, penicillin, streptomycin, TRIzol, Quant-iT OliGreen cDNA quantification kit, Platinum SYBR Green, and qPCR SuperMix-UDG and GAPDH were from Invitrogen (Carlsbad, CA). Fibronectin and type I collagen were from Becton-Dickinson (Franklin Lakes, NJ). Dulbecco’s Modified Eagle’s medium (DMEM) and Ham’s F12 were from Mediatech Inc. (Manassas, VA). Fura 2-acetoxymethyl ester (fura 2-AM) and fura-2 were purchased from Calbiochem (La Jolla, CA). ATP, fetal bovine serum (FBS), and protease were from Sigma-Aldrich (St Louis, MO). iScript cDNA synthesis kit was from Bio-Rad (Hercules, CA). Real-time RT-PCR primers were purchased from IDT-DNA (Coralville, IA). All other chemicals were from Sigma-Aldrich or Fisher Scientific (Pittsburgh, PA) and were of Molecular Grade or higher in quality.

Immortalized human bronchial epithelial cell culture.

Growth conditions for 16HBE14o- cells (Gruenert et al., 1995) have been described (Olsen et al., 2008; Sherwood et al., 2011). Briefly, 16HBE14o- cells were grown on a collagen/fibronectin/BSA (CFB) matrix. Cells were expanded in flasks prior to passage onto 15-mm glass coverslips for microscopy studies. In both cases, cells were grown in a controlled growth medium (CGM) that consisted of MEM supplemented with 10% FBS, 2mM Glutamax, penicillin, and streptomycin at 37°C in a 5% CO₂ atmosphere. CGM alone (arsenic-free) or CGM supplemented with 130 or 330nM arsenic was replaced every other day. Arsenic-supplemented and matched controls were passaged 3–4 times for a total of 4–5 weeks prior to experimentation.

Scrape wound repair assays.

16HBE14o- cells were grown to confluence on CFB-treated glass bioptech dishes (Bioptech, Butler, PA) in CGM as described above. Bioptech dishes were mounted onto an Olympus IX70 microscope in a temperature control stage (Bioptech). At the time of experimentation, cell media were changed to arsenic-free or arsenic-supplemented DMEM without NaHCO₃ and supplemented with 2% FBS. The temperature of the microscope stage was maintained at 37°C throughout the experiment. A glass micropipette (tip diameter ~1 µm) was positioned immediately above a single 16HBE14o- cell with a micromanipulator (Siskiyou, Inc., Grants Pass, OR) under motorized control. The probe was briefly lowered to puncture an individual cell then dragged across the field of view and back again for approximately 4 s to dislodge cells at which point the probe was raised above the confluent culture. Wound area in each image was measured using the TScratch add on (Gebäck et al., 2009) to Image J (NIH, Bethesda, MD) and quantified as percent of wound closure over time compared with the original wound area. These values are expressed in graphs as percentage of original wound area ± SEM.

Intracellular Ca²⁺ concentration ([Ca²⁺]_i) measurements.

Cell monolayers were washed with a modified Hanks’ balanced saline solution (HBSS; 1.3mM CaCl₂, 5.0mM KCl, 0.3mM KH₂PO₄, 0.5mM MgCl₂, 0.4mM MgSO₄, 137.9mM NaCl, 0.3mM Na₂PO₄, and 1% glucose additionally buffered with 25mM HEPES, pH 7.4) and loaded for 45min in 5µM fura 2-AM in HBSS or HBSS supplemented with arsenic. Cells were removed from fura 2-AM loading solution and placed back in matching HBSS w/wo arsenic for at least 20min before Ca²⁺ imaging. For studies in Ca²⁺-free (5.33mM KCl, 0.44mM KH₂PO₄, 137.9mM NaCl, 0.34mM Na₂PO₄, and 1% glucose additionally buffered with 25mM HEPES, pH 7.4) or Na-free (140mM N-methyl-D-glucamine; KCl, 5.0mM; CaCl₂, 3.0mM; and additionally buffered with 10mM Hepes, pH 7.9) solutions, cells were transferred for 3–5min in the respective solution, then washed again with this solution prior to imaging. Fura-2 fluorescence was observed on an Olympus IX70 microscope with a 40× oil immersion objective after alternating excitation between 340 and 380nm by a 75W Xenon lamp linked to a Delta Ram V illuminator (PTI, London, Ontario) and a gel optic line. Images of emitted fluorescence above 505nm were recorded by an ICCD camera (PTI) and simultaneously displayed on a 21″ Vivitron color monitor. The imaging system was under software control (ImageMaster, PTI) and collected a ratio approximately every 0.6 s. Intracellular Ca²⁺ concentration ([Ca²⁺]_i) was calculated by ratiometric analysis of fura-2 fluorescence using equations originally published in Grynkiewicz et al. (1985). A typical field of view contained 80–110 cells at a resting [Ca²⁺]_i estimated to be ≤ 75nM. A change in [Ca²⁺]_i was considered positive if the cell increased [Ca²⁺]_i to 200nM or more.

Localized mechanical cellular wounding.

Glass coverslip cultures of fura-2 loaded cell monolayers were placed on the microscope described above and viewed in differential interference contrast mode. Glass micropipettes were maneuvered as described above, optics were switched to Ca²⁺ imaging mode and at the appropriate time, and the glass probe was lowered to puncture an individual cell (~0.25 s) and immediately retracted to a position well above the monolayer. Single and double cell wounds were characterized by a rapid loss of fura-2 dye. If no loss of dye was recorded or if more than two cells exhibited dye loss, the experiment was excluded from analysis.

Real-time RT-PCR.

16HBE14o- cells were grown for 4–5 weeks as described above in arsenic-free or arsenic-supplemented (130, 330nM) media. RNA was isolated from confluent flasks using TRIzol reagent according to the manufacturer’s protocol and quantified with a NanoDrop ND-1000 (Thermo Fisher Scientific, Waltham, MA). cDNA was synthesized using iScript cDNA synthesis kit according to the manufacturer’s protocol on an iCycler thermocycler (Bio-Rad). cDNA was quantified using Quant-iT OliGreen quantification kit according to the manufacturer’s instructions on a TBS-380 mini-fluorimeter (Turner BioSystems, Sunnyvale, CA). Total cDNA, 100ng, per reaction was amplified with Platinum SYBR Green qPCR SuperMix-UDG kit according to the manufacturer’s instructions in a Rotor-Gene 3000 real-time thermal cycler (Corbett Robotics, San Francisco, CA) under the following conditions: initial hold for 2min at 50°C and hold for 2min at 95°C followed by 45 cycles consisting of denature 15 s at 94°C; anneal 30 s at 60°C for GAPDH, P2Y₂, and P2Y₄ or 54°C for GAPDH, P2X₄, and P2X₅. Human gene-specific primer pairs were designed using IDT-DNA Primer Quest, Primer Bank (Wang and Seed, 2003), and/or MacVector Software. All primers were purchased from IDT-DNA and are listed in Supplementary table 1. Individual analyses were performed in triplicate on cDNA samples obtained from at least three separate isolations for each experiment.

Physiological response to ATP using the xCELLigence Real-Time Cell Analyzer.

Methods for physiological response have been described (Sherwood et al., 2011). Briefly, 16HBE14o- cells were plated in full medium onto 96-well E-plates (Roche Applied Science, Indianapolis, IN) coated with CFB solution and allowed to grow at 37°C and 5% CO₂, where impedance at the well surface was continuously monitored (Abassi et al., 2009; Xi et al., 2008). As per manufacturer’s instructions, relative impedance is expressed as a Cell Index where: Cell Index = (Z _i-Z ₀)/15Ω; and Z _i is impedance at a given time point during the experiment (i.e., post ATP addition), and Z ₀ is impedance before the addition of agonist. For reference, activation of GPCR/G_q, such as that occurs following ATP activation of purinergic receptors, results in an increase in Cell Index. After the time of the experimental exposure, cells were washed with prewarmed HBSS (supplemented with appropriate arsenic concentrations) and placed at room temperature (RT), room CO₂ for 30–45min. Cells were then exposed to a dose response of ATP (300nM–1µM in ½ log steps) in HBSS and appropriate arsenic concentrations. Cell Index responses to ATP were recorded every 30 s for 4h. Cell responses were collected in triplicate and adjusted to a baseline by ratioing with recordings from cells washed with HBSS alone. Total integrated cellular responses were calculated from the area under the curve for each ATP dose and converted into dose-response curves. Graph Pad Prism (San Diego, CA) software was used to calculate EC₅₀s.

Animal exposure.

All animals were treated humanely and with regard for alleviation of suffering using protocols approved by the University of Arizona Institutional Animal Care and Use Committee. Male C57Bl6 mice (Jackson Laboratories, Bar Harbor, ME), 3 months of age, were exposed to arsenic through drinking water for 4 weeks. A total of eight animals were split into two exposure groups: 50 ppb arsenic-supplemented drinking water and a control group (drinking water only, pH 7.0). Water for the 50 ppb treatment group was prepared from sodium arsenite (Na-arsenite), and the pH was adjusted to 7.0 with hydrochloric acid and sodium hydroxide. Water was administered ad libitum. Water was changed every 3 days.

Primary mouse tracheal epithelial cell culture.

The procedure used for isolation of mouse tracheal epithelial (MTE) cells was adapted from methods described by Davidson et al. (2000, 2004) and You et al. (2002). Briefly, mice were killed by cervical dislocation. Tracheas were removed, cut lengthways, and washed in PBS for 5min at RT, then transferred to collection media (1:1 mixture of DMEM and Ham’s F12 with 1% penicillin-streptomycin) at 37°C. Tracheas were then incubated at 37°C for 120min in a dissociation media (44mM NaHCO₃, 54mM KCL, 110mM NaCl, 0.9mM NaH₂PO₄, 0.25µM FeN₃O₉, 1µM sodium pyruvate, 42µM phenol red, pH 7.5, and supplemented with 1% penicillin-streptomycin, 1.4mg/ml pronase). Enzymatic digestion was stopped by adding 20% FBS to the dissociation media. Epithelial cells were dissociated by gentle agitation followed by physical removal from the tracheas. Epithelial cells were then centrifuged at 100 × g for 5min at RT. Cell pellets were washed in a base culture medium (1:1 mixture of DMEM and Ham’s F12 with 1% penicillin-streptomycin, 5% FBS) and centrifuged at 100 × g for 5min at RT. MTE cells were resuspended in a full culture medium (1:1 mixture of DMEM and Ham’s F12 supplemented with 1% penicillin-streptomycin, 5% FBS; 15mM Hepes, 26.1mM sodium bicarbonate, 4mM L-glutamine, 10 µg/ml insulin, 5 µg/ml transferrin, 25ng/ml epidermal growth factor, 30 µg/ml bovine pituitary extract) seeded onto glass coverslips coated with rat tail collagen (Dirksen et al., 1995) and cultured at 37°C with 5% CO₂. MTE cells were fed every other day in full culture medium. Medium was not supplemented with arsenic in cultures from either the arsenic-free or arsenic-supplemented mice.

Statistics.

Data were compared using a one-way ANOVA with Tukey’s multiple comparison test unless otherwise noted. A value of p < 0.05 was used to establish significant differences between samples. Figures are graphed ± SEM.

RESULTS

Chronic Low-Dose Arsenic Compromises Human Airway Epithelial Wound Repair

16HBE14o- cells were exposed 4–5 weeks (3–4 passages) to arsenic-supplemented (as Na-arsenite) media (130nM [equivalent to 10 ppb] or 330nM [25 ppb]). We assessed epithelial wound response following chronic low-dose arsenic exposure using a scrape wound assay that was followed over time with video microscopy. Representative images from the time of the scrape to 60min of wound healing are shown in Figure 1A and in Supplementary movie 1. There were no significant differences in initial wound sizes between arsenic-supplemented and arsenic-free controls in all experiments. The untreated cells closed 47.7±2.0% of the wound within 30min, neared complete closure of the wound within 60min (90.0±6.0%), and fully closed the wound within 2h in all experiments (Fig. 1B). The rate of wound closure during the first hour was significantly reduced for cells chronically exposed to either 130 or 330nM arsenic. The cells treated with 130nM arsenic displayed minimal wound closure at 30min (11.0±4.7% wound closure) and only reached 42.1±13.5% wound closure at 60min. The cells treated with 330nM arsenic displayed a similar delay in wound repair, with only 6.7±4.2% of the wound closed in the first 30min, and 26.3±8.2% wound closure within 60min.

Fig. 1.

Wound healing assay of 16HBE14o- cells exposed chronically (4–5 weeks) to 0, 130, or 330nM arsenic. (A) Representative images of initial wounding (left column), 30min post-wounding (middle column), and 1h post-wounding (right column). White scale bar represents 50 µm. (B) Quantification of percent wound closure over time at 30 and 60min post-wounding. “*” indicates significant difference from untreated controls by a one-way ANOVA (p < 0.05).

Chronic Low-Dose Arsenic Reduces Wound-Induced Inter- and Intracellular Ca²⁺ Signaling in Human Airway Epithelial Cells

We investigated wound-induced Ca²⁺ signaling as a potential mechanism involved in arsenic-compromised airway epithelial wound repair. To quantify Ca²⁺ signaling, we used localized wounding (1–2 cells) of 16HBE14o- monolayers and monitored [Ca²⁺]_i changes in all cells near the wound over time. Typical [Ca²⁺]_i changes in all cells following response to localized wounding in arsenic-free or arsenic-supplemented media are shown in Figure 2A. Under arsenic-free conditions, localized wounds caused a rapid increase of [Ca²⁺]_i in cells adjacent to the wounded area that resulted in an intercellular Ca²⁺ wave of 45.7±2.6 cells (n = 20; Fig. 2B, Supplementary movie 2). There was a significant reduction in the number of cells participating in the Ca²⁺ wave following treatment with 130nM arsenic (27.3±2.8 cells; n = 20). The propagated Ca²⁺ wave was further reduced in cells treated with 330nM arsenic (23.0±2.1 cells; n = 34). To better quantify cellular Ca²⁺ signaling, we evaluated responses in cells directly adjacent to the wounded cell(s) that were postulated to receive similar concentrations of wound-induced signaling at controlled time points. In arsenic-free cultures, 97.8±1.2% (n = 20; Fig. 2C) of wound-adjacent cells responded. However, in the 130nM arsenic-treated cultures, the wound-adjacent cells displayed a significant reduction in response (89.7±2.5%; n = 20). The percentage of wound-adjacent cells responding with an increase in [Ca²⁺]_i was further reduced in the 330nM arsenic-treated cultures (84.3±3.6%; n = 34).

Fig. 2.

Localized wound-induced Ca²⁺ signaling dynamics in 16HBE14o- cells exposed to chronic low-dose arsenic. (A) Representative images of 16HBE14o- cell monolayers subjected to a localized wound (1–2 cells; grey area) at ~3 s and monitored for [Ca²⁺]_i over 60 s. Color scale at bottom indicates approximate [Ca²⁺]_i (nM); cells are outlined with white borders; white bar represents 50 µm. (B) Number of cells demonstrating increase in [Ca²⁺]_i (i.e., [Ca^2+] _i ≥ 200nM) following 1–2 cell wound for each arsenic exposure. (C) Percentage of wound-adjacent cells with a Ca²⁺ response. (D) Average peak [Ca²⁺]_i in wound-adjacent cells for control and arsenic-treated cultures. (E) Time required to reach a baseline [Ca²⁺]_i of 200nM following wound of adjacent cell. “*” indicates significant difference from untreated controls by a Student t-test (p < 0.05) for the 130nM arsenic treatment group in (C) and (E).

In order to quantify intracellular Ca²⁺ signaling dynamics after chronic low-dose arsenic exposure, components of the [Ca²⁺]_i changes in wound-adjacent cells were assayed for arsenic-free (n = 102) and arsenic-treated (130nM, n = 102; 330nM, n = 99) 16HBE14o- cells. In the arsenic-free cells, the mean peak [Ca²⁺]_i reached 639.3±15nM (Fig. 2D). This value was significantly greater than that observed in either 130nM arsenic-treated cells (567.5±16nM) or 330nM arsenic-treated cells (563.1±18nM). We also found that the time necessary to reach a signaling level of [Ca²⁺]_i (200nM) in wound-adjacent cells was significantly increased in both arsenic treatment groups (4.6±0.4 s for 130nM; 5.1±0.3 s for 330nM) compared with controls (3.2±0.3 s; Fig. 2E).

Chronic Low-Dose Arsenic Reduces ATP-Mediated Ca²⁺ Signaling in Human Airway Epithelial Cells

Wound-induced Ca²⁺ signaling in 16HBE14o- cells is largely driven by extracellular ATP release from the wounded cell (Sherwood et al., 2011; Wesley et al., 2007). The reduction in Ca²⁺ signaling observed above could be due to a loss of ATP recognition at the cellular surface or as the result of a decrease in intracellular ATP levels. To evaluate ATP signaling under conditions of chronic arsenic exposure, we first used the high throughput xCELLigence Real Time Cell Analysis (RTCA) physiological assay to quantify an ATP dose response (Fig. 3A; Sherwood et al., 2011). The arsenic-free cells responded with a calculated EC₅₀ of 6.43µM ATP (95% Confidence Interval (CI) 4.45–9.28µM), whereas the EC₅₀s for the arsenic-treated cells were clearly increased (10.2µM, CI: 7.91–13.2µM for 130nM arsenic-treated cells; 9.8µM, CI: 7.96–12.2µM for 330nM arsenic-treated cells). To assay whether these physiological changes were reflected in ATP-mediated Ca²⁺ signaling, we next applied defined concentrations of ATP to 16HBE14o- cultures exposed to chronic low-dose arsenic and untreated controls and measured [Ca²⁺]_i responses (Figs. 3B and C). When a supraphysiological amount of ATP (10µM) was applied to 16HBE14o- cell monolayers, no significant changes in [Ca²⁺]_i between arsenic-free (n = 6) and arsenic-treated (130nM; n = 6; 330nM; n = 5) cells were detected (Fig. 3B), confirming that arsenic-treated cells maintained the ability to respond to ATP with Ca²⁺ signaling events. When applied ATP concentrations were lowered to 1µM to be in the range of in vivo airway epithelial signaling amounts (Okada et al., 2006), both arsenic-treated groups had a significantly reduced percentage of cells responding with a [Ca²⁺]_i change in comparison with arsenic-free controls (Fig. 3C). Under this condition, 43.3±4.8% (n = 10) of the arsenic-free cells responded, whereas only 23.6±5.5% (n = 9) of the 130nM and 15.1±3.5% (n = 10) of the 330nM arsenic-treated cells responded. Representative panels from a typical 1µM ATP experiment are depicted in Figure 3D.

Fig. 3.

ATP-mediated Ca²⁺ signaling in 16HBE14o- cells exposed to chronic low-dose arsenic. (A) ATP dose response measured in a Real-Time Cell Analysis physiological assay (RTCA, see Materials and Methods). EC₅₀ is 6.42µM ATP for 0nM arsenic, 10.2µM ATP for 130nM arsenic-treated cells, and 9.8µM ATP for 330nM arsenic-treated cells. (B) Percentage of cells with a Ca²⁺ response following application of supraphysiological levels of ATP (i.e., 10µM). (C) Percentage of cells with a Ca²⁺ response following application of endogenous signaling [ATP] (i.e., 1µM). (D) Representative images of 16HBE14o- monolayers that were monitored for [Ca²⁺]_i changes in response to exogenous application of 1µM ATP over 180 s. Color scale at bottom indicates approximate [Ca²⁺]_i (nM); white scale bar represents 50 µm. “*” in C indicates significant difference from untreated controls by a one-way ANOVA (p < 0.05).

Chronic Low-Dose Arsenic Reduces P2Y Receptor Function and Molecular Expression

P2Y receptors are G protein–coupled receptors that respond to ATP agonism with inositol triphosphate (IP₃)–mediated release of Ca²⁺ from internal stores (Burnstock, 2007). In order to isolate arsenic-mediated alterations in P2Y receptor function, and specifically, changes in [Ca²⁺]_i due to Ca²⁺ release, we implemented our localized wound model in the absence of extracellular Ca²⁺. Ca²⁺ signaling dynamics were assayed by aligning the average peak [Ca²⁺]_i in cells immediately adjacent to the wound cell(s) (Fig. 4A) and estimating the total relative Ca²⁺ change in response by calculating the area under the Ca²⁺ response curve down to a baseline signaling concentration (grey area in Fig. 4A). Cells treated with 330nM arsenic (n = 63) displayed a significant reduction in total Ca²⁺ response to wounding compared with arsenic-free (n = 66) or 130nM arsenic-treated cells (n = 84) (Fig. 4B). Unlike experiments conducted in Ca²⁺-containing buffer (Fig. 2), there were no differences in average peak [Ca²⁺]_i between the arsenic-free (787.9±13.4nM [n = 66]) and the arsenic-exposed cells (774.3±11.8nM [n = 84] and 763.3±12.0nM [n = 63] for 130 and 330nM arsenic-treated cells, respectively; Fig. 4C). Also in contrast to Ca²⁺-containing solutions, there were no significant differences in the ability for the arsenic-treated cells to respond to the local wound as demonstrated in the time required to reach a threshold signaling [Ca²⁺]_i (Fig. 4D). In addition to wound-induced ATP Ca²⁺ signaling, we tested arsenic effects on P2Y receptor function with exogenously applied ATP in Ca²⁺-free buffer (Fig. 4E). Similar to that observed in the wound-generated ATP signaling above (Fig. 4B), the arsenic-free and 130nM arsenic-treated cells displayed full Ca²⁺ responses following 500nM ATP application (96.8±2.8%, n = 4; 94.9±1.6%, n = 5, respectively); however, there was a significant reduction in Ca²⁺ response in the 330nM arsenic-treated cells (75.5±10.2%, n = 3). When we repeated these experiments with the more selective but less potent P2Y₂/P2Y₄ agonist UTP, we obtained similar activation patterns (Fig. 4F). Application of 5µM UTP resulted in 80.1±1.9% (n = 5) activation under arsenic-free conditions and 71.6±4.8% (n = 5) in the 130nM treated cultures. However, a significant reduction in Ca²⁺ response (50.9±5.0%, n = 5) was evident in the cultures chronically treated with 330nM arsenic. An identical pattern was observed when 2.5µM UTP was used as an agonist where a 42.4±6.0% Ca²⁺ response was observed in control cells (n = 5), a 25.5±4.3% response was observed in 130nM arsenic-treated cells (n = 4), and a significant reduction in Ca²⁺ response (6.1±1.6%, n = 4) was observed in 330nM arsenic-treated cells.

Fig. 4.

Isolation of arsenic effects on P2Y receptor function and molecular expression. Localized wound assay was repeated in the absence of extracellular Ca²⁺ to isolate metabotropic P2Y receptor function. (A) All [Ca²⁺]_i traces from wound-adjacent cells reaching a threshold of ≥ 200nM Ca²⁺ were aligned at their peak [Ca²⁺]_i to evaluate Ca²⁺ signaling dynamics under Ca²⁺-free buffer conditions. Each data point is plotted ± SEM; grey area under the curve represents relative Ca²⁺ signal. (B) Total relative Ca²⁺ change quantified under Ca²⁺-free buffer conditions. (C) Average peak [Ca²⁺]_i in wound-adjacent cultures under Ca²⁺-free conditions. (D) Time required to reach threshold [Ca²⁺]_i (200nM) under Ca²⁺-free buffer conditions. (E) Percentage of cells with a Ca²⁺ response following application of 500nM ATP under Ca²⁺-free buffer conditions. (F) Percentage of cells displaying Ca²⁺ response following application of 5 and 2.5µM UTP. (G) Expression of P2Y₂ and P2Y₄ receptor mRNA.

We next used real-time RT-PCR to determine whether arsenic was altering cellular levels of P2Y₂ and P2Y₄, the major P2Y receptors expressed in 16HBE14o- cells (Sherwood et al., 2011). We quantified relative mRNA levels of P2Y₂ (highly expressed) and P2Y₄ (minimally expressed) in arsenic-free and arsenic-treated 16HBE14o- cultures and found P2Y₂ mRNA was significantly decreased in both arsenic-treated cultures (130, 330nM), whereas P2Y₄ levels were unaffected (Fig. 4G). Immunoblots and immunocytochemistry of P2Y₂ protein confirmed expression of this protein in both arsenic-free and arsenic-treated cultures; however, quantification of changes in expression was not possible (data not shown).

Chronic Low-Dose Arsenic Reduces P2X Receptor Function in Human Airway Epithelial Cells

The noted differences in Ca²⁺ signaling gleaned from comparing local wound responses in Ca²⁺-free and Ca²⁺-containing buffers suggest that arsenic is targeting mechanisms of Ca²⁺ influx. For example, when Ca²⁺ influx can occur (e.g., in Ca²⁺-containing buffers), there is a significant increase in the time from mechanical wounding to the cellular response in the arsenic-exposed cells (Fig. 2E). However, when these experiments are repeated in Ca²⁺-free buffer, this lag is eliminated in 130nM arsenic-treated cells and significantly reduced in the 330nM arsenic-treated cells (Fig. 4D). The P2X receptors are a family of ligand-gated, nonspecific cation channels that are agonized by ATP, leading to a cellular influx of Ca²⁺ and other ions from the extracellular milieu (North, 2002; Schwiebert and Zsembery, 2003). We analyzed the effects of chronic low-dose arsenic on P2X receptor function by applying exogenous Zn (2.5µM) in a Na⁺-free buffer to monolayers of 16HBE14o- cells to activate P2X receptors and monitored cells for changes in [Ca²⁺]_i (Sherwood et al., 2011; Zsembery et al., 2003). Representative cell traces of [Ca²⁺]_i in response to Zn are shown for arsenic-exposed cultures and untreated controls (Fig. 5A). Following Zn application, the time necessary for cells to reach a threshold [Ca²⁺]_i (i.e., > 200nM) in arsenic-free cultures (18.9±1.1 s [n = 6]; Fig. 5B) was significantly faster than that observed in the 130nM (23.7±1.7 s, n = 6) or the 330nM arsenic-treated cells (28.5±2.5 s, n = 4).

Fig. 5.

Chronic arsenic exposure effects on P2X receptor function and molecular expression. (A) Representative single-cell traces showing changes in [Ca²⁺]_i following exogenous application of 2.5µM Zn in Na⁺-free buffer. (B) P2X receptor activation with 2.5µM Zn as indicated by an increase in time to threshold [Ca²⁺]_i (200nM). (C) P2X₄ and P2X₅ receptor mRNA. “*” indicates significant difference (p < 0.05) from untreated controls by one-way ANOVA (330nM) and Student’s t-test (130nM).

We next analyzed molecular expression of P2X receptors in 16HBE14o- cells following chronic low-dose arsenic exposure. We quantified relative mRNA levels of P2X₄ (highly expressed) and P2X₅ (minimally expressed; Sherwood et al., 2011) in arsenic-free and arsenic-treated (130, 330nM) 16HBE14o- cultures and found no differences (Fig. 5C). Immunoblots of P2X₄ protein levels also showed no significant differences in total protein expression between arsenic-exposed and arsenic-free cultures (data not shown). The lack of changes in receptor expression patterns suggests that the arsenic-induced changes in P2X receptors occur posttranslationally.

In Vivo Chronic Low-Dose Arsenic Exposure Compromises Wound-Induced Ca²⁺ Signaling Dynamics in Primary MTE Cells

To better assess whether our in vitro results model drinking-water arsenic exposure conditions, we evaluated Ca²⁺ signaling responses in airway epithelial cells from mice treated chronically (4 weeks) with arsenic-free (0 ppb) or arsenic-supplemented (50 ppb) drinking water. After the exposure period, MTE cells were cultured for 4 days prior to experimentation and subjected to localized wounding. All [Ca²⁺]_i traces from wound-adjacent cells reaching a threshold of ≥ 200nM Ca²⁺ were aligned at peak [Ca²⁺]_i to evaluate Ca²⁺ signaling dynamics (Fig. 6A). In cells from mice exposed in vivo to 50 ppb arsenic, the total relative Ca²⁺ signal was reduced to 79.8% (n = 80) of that observed in cells from unexposed mice (n = 29; Fig. 6B). Peak [Ca²⁺]_i in cells from mice exposed in vivo to 50 ppb rose to 568.2±17.5nM (n = 80), which significantly reduced from the response observed in cells cultured from unexposed mice (719.9±31.9nM, n = 29; Fig. 6C). The primary cultured MTE cells from in vivo exposed mice also displayed significant reductions in Ca²⁺ signaling.

Fig. 6.

Localized wound-induced [Ca²⁺]_i changes in primary cultured tracheal epithelial cells from mice exposed chronically (4 weeks) to low-dose (50 ppb) arsenic. MTE cell monolayers were subjected to a 1–2 cell wound at ~3 s and monitored for [Ca²⁺]_i over 60 s. (A) All [Ca²⁺]_i traces from wound-adjacent cells reaching a [Ca²⁺]_i threshold (200nM) were aligned at the peak [Ca²⁺]_i to evaluate Ca²⁺ signaling. Each data point is plotted ± SEM. Grey area under the curve represents relative Ca²⁺ signal. (B) Total relative [Ca²⁺]_i change. (C) Average peak [Ca²⁺]_i in wound-adjacent cells. “*” indicates significant difference from untreated controls by a Student’s t-test (p < 0.05).

DISCUSSION

A primary finding from this report is that chronic arsenic exposure at nanomolar levels (130–330nM) can compromise two important components in airway epithelial cell–based innate immunity–wound repair and ATP-mediated Ca²⁺ signaling. Specific targets of arsenic-induced dysfunction in airway epithelial cells include both P2Y and P2X receptors that recognize extracellular ATP and directly contribute to Ca²⁺ signaling (Fig. 7). These data highlight the low threshold for arsenic tolerance on epithelial signaling and, given the ubiquitous importance of paracrine ATP signaling in airway epithelium, as well as in other cells and tissues, provide a target for dysfunction that requires further exploring in the eventual understanding of arsenic-induced toxicity. We hypothesize that chronic low-dose arsenic exposure results in a physical interaction with both P2Y₂ and P2X receptors, which reduces their ATP response and Ca²⁺ signaling capabilities. Chronic exposure to nanomolar levels of arsenic also results in a reduction of available P2Y₂ mRNA, which may impact the amount of protein at the membrane. Although it remains unknown how arsenic is disrupting these receptors and identifies only a subset of a larger collection of changes that can occur following arsenic exposure, the model, nonetheless, is an important step toward understanding arsenic-induced mechanistic dysfunction of innate immunity in the airway.

Fig. 7.

Proposed mechanism for arsenic-induced reductions in ATP-mediated Ca²⁺ signaling in airway epithelial wound response. ATP is released into the extracellular milieu following epithelial cellular wounding of 16HBE14o- cells and subsequently binds purinergic receptors (P2Y₂ and P2X) on neighboring cell surfaces. Upon ATP binding of the P2Y₂ receptor, Ca²⁺ is released from the endoplasmic reticulum via IP₃ binding its receptor on the surface of the endoplasmic reticulum, leading to an increase in intracellular Ca²⁺. When ATP binds P2X receptors, Ca²⁺ enters the cell from the extracellular milieu. We suggest that arsenic physically interacts with both P2Y₂ and P2X receptors, which leads to their functional reduction and causes the reduction of available P2Y₂ mRNA.

Evidence from multiple studies suggests that arsenic exposure can have important consequences for airway immunity (Kozul et al., 2009a; Liao et al., 2011; Rahman et al., 2011; Smith et al., 2011). In human studies from Bangladesh, a population of individuals exposed to arsenic in utero demonstrated significantly increased respiratory tract infections during infancy (Rahman et al., 2011), and in a Chilean population, a correlation between consumption of drinking-water arsenic and increased mortality following infection with pulmonary tuberculosis was demonstrated (Smith et al., 2011). In animal studies, where experimental paradigms are more controlled, the links between arsenic exposure and increased infections are more dramatic. For example, mice given arsenic in their drinking water for 5 weeks suffered increased morbidity and viral load after intranasal inoculation with H1N1 virus (Kozul et al., 2009b). These studies exemplify the necessity to investigate arsenic effects on lung immunity in order to elucidate mechanisms that lead to the development of obstructive airway disease related to chronic arsenic exposure.

The airway epithelium is a critical component of airway innate immunity that requires constant remodeling and repair to maintain its primary barrier function. Restoration of the airway epithelium following minor insults involves spreading and migration of wound-adjacent cells and cellular proliferation and differentiation (reviewed by Crosby and Waters [2010]). Both in vitro and in vivo epithelial repair models suggest that the initial cellular spreading and migration of wound-adjacent cells is essential to quick restoration of the epithelial layer (reviewed by Crosby and Waters [2010] and Puchelle et al. [2006]). It is well documented that Ca²⁺ signaling plays multiple roles in cell spreading and migration (reviewed by Prevarskaya et al. [2011]), including that of airway (Wesley et al., 2007). The findings that chronic low-dose arsenic alters the intracellular Ca²⁺ signaling footprint of airway epithelial cells following localized cellular wounding is a likely contributor to the loss of directed cell migration. It is important, however, to consider that airway cell migration requires an extensive array of cellular processes, and arsenic can also contribute to wound dysfunction through these alternate mechanisms (Olsen et al., 2008; Wang et al., 2012).

In previous reports using acute (24h) arsenic exposure at higher arsenic concentrations (0.8–3.9µM) on 16HBE14o- cells, a loss of wound repair via a MMP-9 regulation dysfunction (Olsen et al., 2008) and reduced purinergic (P2Y₂ and P2X₄) Ca²⁺ signaling (Sherwood et al., 2011) had been shown. In this study, we show that chronic exposure to arsenic at levels associated with drinking water standards effectively lowers the dose necessary for loss of migration and purinergic-dependent Ca²⁺ signaling. In contrast to our acute higher dose exposure model, the chronically exposed cells do not display the relatively large changes in mRNA expression (P2Y₂ and P2X₄) observed after acute exposure (Sherwood et al., 2011). Given that these studies were carried out in identical cell lines, and resulted in a similar loss of Ca²⁺ signaling despite apparent mechanistic differences, the ability to fully understand the multiple mechanisms of arsenic toxicity and especially how arsenic affects basic signaling mechanisms remains challenging.

In this study, chronic low-dose arsenic leads to the functional reduction of the P2Y₂ receptor as measured by ATP-induced Ca²⁺ signaling. P2Y₂ receptor has been shown to be required for ATP-induced epithelial wound repair in multiple epithelial cell types including those that line the lung (Klepeis et al., 2004; Wesley et al., 2007), and we have demonstrated that wound-induced Ca²⁺ signaling is ATP dependent in 16HBE14o- cells (Sherwood et al., 2011). P2Y₂ receptor involvement in wound-induced ATP-mediated signaling is not necessarily limited to effects directly involving intracellular Ca²⁺ but can also contribute to epithelial wound cell migration through other cellular pathways (e.g., activation of G(o) and G(12) via integrin binding) (Bagchi et al., 2005; Klepeis et al., 2004; Liao et al., 2007; Peterson et al., 2010; Weisman et al., 2005). Although we saw significant reductions in P2Y₂ mRNA expression following arsenic exposure, we cannot confirm that this translates into a reduction of available receptor at the membrane or that membrane-bound receptor quantity fully accounts for signaling toxicity. In human microvascular endothelial cells, arsenic has been shown to directly interact with G protein–coupled receptors to alter function (Straub et al., 2009). In that study, arsenic (≤ 5µM as Na-arsenite) activated the Sphingosine-1-Phosphate Type 1 receptor, and that interaction was proposed to be an allosteric activation involving arsenic directly binding to thiols within the receptor. Because arsenic is well known to bind thiols (Delnomdedieu et al., 1993), it may interact in such a way with the P2Y₂ receptor resulting in its reduced function. In human small airway carcinoma cells, chronic application of arsenite (6.7–26µM) has been shown to alter the phosphoprotein profile of a variety of receptors and their associated mitogen-activated protein kinase (MAPK) signaling pathways (Wen et al., 2010). Akt and MAPK pathways were upregulated under acute exposures of 5µM arsenic in BEAS-2B and A549 cells (Liu et al., 2011). Also, in a human airway epithelial cell line related to the 16HBE14o- cells used herein, the Cystic Fibrosis Transmembrane Conductance Regulator protein showed increased ubiquitin-mediated degradation via enhanced phosphorylation of the E3 ubiquitin ligase c-Cbl following acute arsenic (≤ 130nM) exposure (Bomberger et al., 2012). Because of the multiple possible effects of arsenic in GPCR and other plasma membrane proteins, further studies are needed to clarify the mechanism(s) by which arsenic reduces P2Y₂ receptor–mediated signaling.

In contrast to findings from the acute, high-dose arsenic studies, changes in the Ca²⁺ signaling of adjacent cells following localized wounding in our chronic exposure model were more defined by changes in Ca²⁺ influx associated with P2X dysfunction. The lack of significant changes in the mRNA or protein of the primarily expressed P2X receptors of 16HBE14o- cells (P2X₄ and P2X₅) suggests that arsenic leads to a physical alteration of the P2X receptor channel. Arsenic is known to interact with free thiol groups, and such groups are ample on the surface of P2X receptors (North, 2002; Schwiebert et al., 2005). It has been proposed that histidines may cluster with cysteines in the extracellular domain of P2X receptors to form a binding motif reminiscent of a zinc finger (Schwiebert et al., 2005). Arsenic has been reported to bind Zn finger domains on estrogen receptor-α (Kitchin and Wallace, 2005) and to compete for binding with Zn on a zinc finger domain of the DNA repair protein poly(ADP-ribose) polymerase-1 (Ding et al., 2009). Based on the abundant thiol groups at the extracellular surface of the P2X receptor and its suggested Zn finger–like domain, the P2X receptor presents as a prime target for arsenic/protein interaction.

Although P2X and P2Y were the primary molecular targets examined in this study, we cannot rule out other components of the Ca²⁺ signaling apparatus that could be altered by arsenic. Additional potential targets for arsenic disruption are the channels and pumps related to Ca²⁺ release (ryanodine receptor; inositol trisphosphate receptor) and recovery (plasma membrane calcium ATPase; the sarcoplasmic and endoplasmic reticulum calcium ATPase: SERCA). In particular, SERCAs have been shown to be denatured following arsenic exposure and lead to a cellular stress response (Senisterra et al., 1997). In a recent report on Ca²⁺ signaling in primary cultured keratinocytes, acute arsenic (1–5µM as Na-arsenite) exposure altered phosphorylation of the IP₃ receptor to limit Ca²⁺ signaling in response to 500µM ATP (Hsu et al., 2012). Further examination of arsenic effects on the Ca²⁺ signaling pathway under low exposure conditions is needed to fully define the mechanisms involved in its toxicity.

The range of arsenic concentrations used in this study is well below levels associated with inhalation and chronic drinking water studies where arsenic has been correlated with lung disease (Guha Mazumder, 2007; Kapaj et al., 2006; Schuhmacher-Wolz et al., 2009; Smith et al., 2009). Although inhalation exposure that reaches the lung can be similar to that measured in the local environment, the concentrations of arsenic that reach the lung following ingestion are less well defined. Mouse models in which arsenic was given orally have demonstrated an accumulation of arsenic into the lung within 1–4h of a single administration; these concentrations doubled if administration was repeated over 9 days (Hughes et al., 2003; Kenyon et al., 2005). Studies in populations with arsenic-associated disease symptoms demonstrated an increase in blood levels of arsenic over controls that surpassed 10 µg/l in populations with ~100 ppb arsenic in their well water (Hall et al., 2006). These studies indicate that arsenic can reach the lung rapidly and in high concentrations following acute exposures and can remain significantly elevated under chronic conditions. Our findings that wound-induced Ca²⁺ signaling remains depressed in primary cultured MTE cells isolated from mice fed 50 ppb arsenic despite 4 days of arsenic-free culture conditions (Fig. 6) further validates the in vitro model findings as an indicator of effects of low-dose, ingested arsenic. An interesting difference between the immortalized cell model and the primary cultured cell model is the ability for the immortalized cells to recover most of their ATP- and wound-induced Ca²⁺ signaling when cultured in the absence of arsenic for 7 days (data not shown). Taken together, the cell culture arsenic exposure model used herein provides an opportunity to delineate mechanistic insight into arsenic toxicity in the lung epithelium under conditions that are prevalent in real-world exposures. These insights provide molecular targets for further studies to be evaluated from primary cultures to better understand the impact of low-dose arsenic on lung toxicology. However, specific differences such as that seen in recovery from arsenic exposure highlight the impact of the exposure environment, and careful interpretation of targets with validation from primary culture, in situ, and in vivo models is important to the full understanding of arsenic toxicity.

SUPPLEMENTARY DATA

Supplementary data are available online at http://toxsci.oxfordjournals.org/.

Funding

National Institute of Environmental Health Sciences (Superfund Research Grant ES 04940); Semiconductor Research Corporation (project #425.024); National Institute of Environmental Health Sciences (Center Grant ES 06694); United States Environmental Protection Agency (STAR grant RD832095); National Heart Lung and Blood Institute (Training grant T32-HL007249); National Institute of Environmental Health Sciences (Training grant ES 007091).