|Home | About | Journals | Submit | Contact Us | Français|
Micron-sized particles of poorly soluble nickel compounds, but not metallic nickel, are established human and rodent carcinogens. In contrast, little is known about the toxic effects of a growing number of Ni-containing materials in the nano-sized range. Here, we performed physicochemical characterization of NiO and metallic Ni nanoparticles and examined their metal ion bioavailability and toxicological properties in human lung epithelial cells. Cellular uptake of metallic Ni and NiO nanoparticles, but not metallic Ni microparticles, was associated with the release of Ni(II) ions after 24–48 h as determined by Newport Green fluorescence. Similar to soluble NiCl2, NiO nanoparticles induced stabilization and nuclear translocation of hypoxia-inducible factor 1α (HIF-1α) transcription factor followed by upregulation of its target NRDG1 (Cap43). In contrast to no response to metallic Ni microparticles, nickel nanoparticles caused a rapid and prolonged activation of the HIF-1α pathway that was stronger than that induced by soluble Ni (II). Soluble NiCl2 and NiO nanoparticles were equally toxic to H460 human lung epithelial cells and primary human bronchial epithelial cells; metallic Ni nanoparticles showed lower toxicity and Ni microparticles were nontoxic. Cytotoxicity induced by all forms of Ni occurred concomitant with activation of an apoptotic response, as determined by dose- and time-dependent cleavage of caspases and poly (ADP-ribose) polymerase. Our results show that metallic Ni nanoparticles, in contrast to micron-sized Ni particles, activate a toxicity pathway characteristic of carcinogenic Ni compounds. Moderate cytotoxicity and sustained activation of the HIF-1α pathway by metallic Ni nanoparticles could promote cell transformation and tumor progression.
Inhalation of nickel compounds is an occupational hazard associated with development of lung, nasal, and paranasal sinus cancers (IARC, 1990; Straif et al., 2009). Inhalation exposure to poorly soluble metallic nickel, nickel sulfides, and nickel oxides occurs in production industries, while exposures in nickel-use industries are complex and include poorly soluble and soluble nickel compounds (Zhao et al., 2009). Welding fumes are a potential source of metallic nickel particles in a range of sizes (0.11–2 μm mass mean aerodynamic diameter) with variable chemistry and crystallinity. Mixed exposures to metallic nickel and its compounds in the workplace complicate assessment of carcinogenicity of individual components (Sivulka, 2005; Straif et al., 2009). Metallic nickel particles and welding fumes are classified as possibly carcinogenic to humans (IARC, 1990). In rodent inhalation studies, Ni3S2 particles (2.0–2.2 μm) and NiO particles (2.2–2.5 μm) induce lung tumors in contrast to soluble NiSO4 6H2O aerosols that did not induce lung tumors (Dunnick et al., 1995). Metallic nickel particles produce lung tumors in rats following intratracheal instillation (IARC, 1990) but not after inhalation of metallic nickel powder (1.8 μm mass mean aerodynamic diameter) for 24 months (Oller et al., 2008).
Nickel nanoparticles are manufactured for multiple applications including catalysts, sensors, and energy storage devices (Zhang et al., 1998). Nickel is also used as a growth catalyst for synthesis of some types of carbon nanotubes (Donaldson et al., 2006), and although these catalyst residues appear to be encapsulated inside the carbon shells, a fraction is typically bioavailable (Liu et al., 2007). Inhalation exposure to nickel and nickel oxide nanoparticles may emerge as a significant occupational hazard as the nanotechnology industry expands (Cho et al., 2010; Lu et al., 2009). Metallic nickel and nickel oxide nanoparticles have been shown to induce significant lung toxicity and inflammation following intratracheal instillation in rats (Ogami et al., 2009; Zhang et al., 1998). In both of these studies, ultrafine nickel (20 nm) and nickel oxide nanoparticles (800 nm) were more toxic than fine nickel (5 μm) or fine nickel oxide (4.8 μm) particles. Lung inflammation induced by nickel oxide nanoparticles was comparable to the effects of crystalline silica (Min-U-Sil-5) after 6 months of exposure (Ogami et al., 2009). Potential carcinogenicity of metallic nickel or nickel oxide nanoparticles has not yet been assessed.
We have recently shown that nickel (II) ion mobilization from metallic nickel powders or carbon nanotubes is enhanced from nickel nanoparticles relative to microparticles (Liu et al., 2007). Nickel carcinogenicity has been associated with intracellular mobilization and delivery of nickel (II) ions to target cells (Goodman et al., 2011). Soluble nickel salts show minimal uptake, while poorly soluble nickel compounds persist in particle form and are more readily taken up by phagocytosis in cultured cells (Costa and Mollenhauer, 1980). Intracellular nickel (II) ions activate multiple signaling pathways that may contribute to carcinogenicity, including generation of reactive oxygen species and activation of calcium-dependent and stress-induced signaling cascades (Barchowsky and O’Hara, 2003). Epigenetic alterations in histone acetylation, chromatin organization, DNA methylation (Arita and Costa, 2009), and interference with DNA repair pathways (Beyersmann and Hartwig, 2008) may contribute to nickel carcinogenicity (Salnikow et al., 2000; Salnikow and Zhitkovich, 2008). Nickel (II) ions can deplete intracellular ascorbate (Karaczyn et al., 2006; Salnikow et al., 2004) and inactivate prolyl hydroxylases (Costa et al., 2005) leading to induction of hypoxia-inducible factor 1α (HIF-1α) and expression of hypoxia-inducible genes (Maxwell and Salnikow, 2004). By mimicking hypoxia, nickel may provide selective pressure for growth of cells with altered energy metabolism, altered growth regulation, and resistance to apoptosis, promoting tumor development (Salnikow and Zhitkovich, 2008). The ability of metallic nickel nanoparticles to activate hypoxic signaling pathways has not yet been assessed.
In this work, we sought to test the hypothesis that a shift to the nano-size range in nickel-containing particles affects the toxicological properties that may be associated with Ni(II) carcinogenesis. We used an in vitro model of human lung epithelial cells and a panel of well-characterized metallic nickel and nickel oxide nanoparticles to investigate the cellular responses to micro-sized versus nano-sized nickel-containing particles. Specifically, we investigated particle internalization by transmission electron microscopy, nickel mobilization in both cell-free and cell-based systems by inductively coupled plasma atomic emission spectrometer (ICP-AES) and Newport Green fluorescence, respectively, and intracellular responses to nickel-containing particles, including HIF-1α pathway activation and cytotoxic responses. We demonstrate that nickel-containing nanoparticles are taken up by human lung epithelial cells, resulting in efficient delivery of intracellular nickel (II) ions, activation of the HIF-1α pathway, and expression of hypoxia-inducible genes. Additionally, exposure to nickel-containing nanoparticles, but not microparticles, induces a cytotoxic response in human lung epithelial cells, in part due to activation of an apoptotic program as indicated by immunoblot analysis of cleaved caspases and poly (ADP-ribose) polymerase (PARP), that is more robust than that induced by equivalent doses of micro-sized nickel nanoparticles.
The human lung epithelial cell line NCI-H460 (H460) was obtained from ATCC and cultured in RPMI 1640 (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS; Atlanta Biologicals, Atlanta, GA) and 1% penicillin/streptomycin (Invitrogen) at 37°C in a humidified chamber containing 6% CO2/94% air. Normal human bronchial epithelial (NHBE) cells were obtained from Lonza (Walkersville, MD) and cultured in bronchial epithelial growth medium (BEGM), supplemented with SingleQuot growth factors and retinoic acid. NHBE cells were used for cytotoxicity experiments at passage 3–5. Nickel (II) chloride hexahydrate (NiCl2·6H2O), metallic nickel nanoparticles (<100 nm Ni0, 99.9% purity), and micron-sized particles (~3 μm Ni0, 99.7% purity) were purchased from Sigma-Aldrich (St Louis, MO). Nickel (II) oxide (green) nanoparticles (<100 nm) were purchased from NanoAmor (Houston, TX). NiCl2 was diluted in deionized water and sterile-filtered prior to use, while metallic nickel nano- and micro-powders were sterilized at 400°C for 15 min in a tube furnace while purging with nitrogen to prevent oxidation. NiO nanoparticles were baked overnight in air to inactivate endotoxin. Particles were sonicated for 1 h prior to the start of each experiment to improve dispersion. H460 cells were exposed to soluble nickel and insoluble nickel particles in RPMI 1640/2% FBS/1% P/S, while NHBE cells were treated in BEGM.
Materials morphology and electron diffraction (ED) characterization were carried out on a JEOL JEM-2010 high-resolution transmission electron microscope (TEM) at 200 kV and a Philips 420 TEM at 120 kV. Particle crystallinity was determined by X-ray diffraction (XRD) on a Bruker AXS D8 Advance X-ray Diffractometer. Zeta potentials and particle size distribution were measured using a Zetasizer Nano-ZS 900 in ultrapure water (18.3 MΩ), phosphate-buffered saline (PBS), and treatment medium (phenol red-free RPMI 1640/1% penicillin/streptomycin/2% FBS) following a 30-min sonication. Particle size and size distribution of <100-nm nickel nanoparticles were also measured in the three different solutions by dynamic light scattering spectroscopy. Surface areas of both nano and micro nickel particles were determined from the Brunauer-Emmett-Teller (BET) N2 vapor adsorption isotherms obtained at 77 K with an Autosorb-1 (Quantachrome Instruments, Boynton Beach, FL), using the Brunauer-Emmett-Teller (BET) model.
Nano-sized and micron-sized metallic nickel particles and nickel oxide nanoparticles were suspended in endotoxin-free water, centrifuged, and the endotoxin levels in the supernatant were determined using the Limulus Ameobocyte Lysate assay with Escherichia coli O113:H10 (Associates of Cape Cod Inc.) as a standard. Endotoxin levels in the nickel particle preparations (<0.03 EU/ml) were not significantly different from the negative control (<0.03 EU/ml).
Cells were fixed at room temperature for 30 min in 2% glutaraldehyde in 0.1M sodium cacodylate buffer, pH 7.4. Fixed samples were stored in cacodylate buffer with 8% sucrose at 4°C. Samples were subsequently treated with 1% osmium tetroxide in cacodylate buffer and dehydrated through a series of graded ethyl alcohols. The samples were infiltrated overnight with a 1:1 solution of 100% ethanol:Spurr embedding medium (Electron Microscopy Science, Hatfield, PA) and then infiltrated with fresh Spurr medium for 4 h, embedded in molds, and polymerized at 60°C. Blocks were sectioned at a thickness of 80 nm on a Reichert Ultramicrotome with a diamond knife, placed on copper grids, stained with uranyl acetate and lead, and viewed on a Phillips 410 TEM equipped with an Advantage HR CCD camera and Advanced Microscopy Techniques imaging software.
Nickel mobilization assays were carried out by adaptation of the method of Liu et al. (2007). Briefly, nickel particles suspended at 250 μg/cm2 were sonicated for 1 h prior to dilution in 10 ml medium at 5 or 20 μg/cm2 in triplicate and incubated at 37°C for 2–72 h, using identical experimental conditions as for cell culture experiments. Suspensions were centrifuged for 1 h at 4000 rpm, 4°C in 3000 NMWL Amico centrifugal filter devices (Millipore, MA) to remove any free nickel particles. The concentration of mobilized nickel was measured using a Jobin Yvon JY2000 Ultrace ICP-AES at a wavelength of 221.647 nm. Intensities were calibrated using six standards ranging in concentration from 0 to 5 ppm, which were prepared from a 1000 ppm stock solution in treatment medium. Both nickel-free blanks and Ni salt standards were inserted in the test sequence after every 5–10 samples, and extensive rinsing after each measurement ensured no carryover from high concentration samples.
H460 cells were plated onto glass coverslips at 12,500 cells/cm2 and treated as indicated. Cells were washed with 1mM EDTA/Hanks’ Balanced Salt Solution (HBSS) to chelate exogenous ionic nickel and loaded with 5μM Newport Green DCF diacetate (Invitrogen)/Pluronic F127 (Invitrogen)/2% FBS/HBSS. After a 30-min recovery period in 2% FBS/HBSS, Newport Green fluorescence was visualized with a Nikon Eclipse E800 fluorescence microscope. Images were obtained using SPOT software.
Cells were plated at 12,500 cells/cm2, treated as indicated, and adherent and floating cells were collected for Western blot analysis as described in Pietruska and Kane (2007). Primary antibodies used were as follows: HIF-1α (1:500; BD Pharmingen), NDRG1 (1:1000; Cell Signaling), cleaved caspase-3 (1:1000; Cell Signaling), cleaved caspase-7 (1:1000; Cell Signaling), PARP (1:1000; Cell Signaling), and β-actin (1:2000, AC-15; Sigma-Aldrich). Membranes were incubated with horseradish peroxidase (HRP)-conjugated goat anti mouse IgG (Millipore) or HRP-conjugated goat anti-rabbit IgG (Cell Signaling) and processed for enhanced chemiluminescence (Amersham, GE Life Sciences, Piscataway, NJ).
Cell number was evaluated using the PicoGreen dsDNA Kit (Invitrogen) according to the manufacturer’s instructions. H460 cells and NHBE cells were plated at densities of 12,500 cells/cm2 or 23,400 cells/cm2, respectively, in 96-well plates and treated as indicated prior to lysis in 200 μg/ml proteinase K/1× Tris-ethylenediaminetetraacetic acid (TE) overnight at 37°C. For H460 cells, which are hyperdiploid, the lysate was diluted 1:4 in 1× TE to avoid assay saturation. Using these assay conditions, PicoGreen fluorescence was linear with respect to cell number (R2 = 0.988) and DNA content (R2 = 0.999) (data not shown). For primary diploid NHBE cells, the entire lysate was used for viability determination. PicoGreen was added at a final concentration of 1× and fluorescence was measured in a SpectraMax M2 fluorescence microtiter plate reader (excitation 480 nm/emission 520 nm). Data are expressed relative to a mock-treated control at each time point.
Statistical significance was determined using one-way ANOVA in conjunction with the Tukey post hoc test (Prism 5 GraphPad Software, La Jolla, CA), and the data were considered significant when p < 0.05.
Morphologies of metallic nickel and nickel oxide particles are shown in Figure 1. Metallic nickel nanoparticles (<100 nm) are approximately spherical (Fig. 1A, panels i and ii), while micron-sized metallic nickel particles (3 μm) are nonuniform in shape (Fig. 1B, panels i and ii). High-resolution TEM of both <100-nm and 3-μm particles indicates thin (<5 nm) surface oxide layers, which were characterized using ED patterns (Fig. 1A, panel iii and Fig. 1B, panel iii). Nano-sized metallic nickel particles exhibit rings representative of face-centered cubic (FCC) metallic nickel, including Ni (111), (200), (220), and (222), with weak FCC nickel (II) oxide (NiO) patterns, including NiO (200), (220), and (311) (Fig. 1A, panel iv). ED rings for 3-μm particles were indexed as FCC Ni (111), (220), and (311) with weak patterns present at NiO (200),(220), and (311) (Fig. 1B, panel iv). Among these ED patterns, Ni (111)/NiO (200) and Ni (220)/NiO (311) overlapped and are difficult to distinguish. However, the presence of NiO patterns suggests that the particles are slightly oxidized. The XRD patterns of nano-sized and micron-sized nickel particles are shown in Figure 1D. Both nickel samples exhibit three sharp peaks near 2θ = 44.6°, 51.9°, 76.4°, which can be readily indexed as the (111), (200), and (220) crystal planes of cubic phase Ni (JCPDS Card No. 04-0850). Interestingly, small NiO peaks were visible (Fig. 2D, arrow, 2θ = 43.3°, which can be indexed as crystalline NiO [JCPDS 47-1049]). The presence of small NiO peaks in the metallic nickel samples indicates slight oxidation of metallic Ni nanoparticles, either during fabrication or during handling, consistent with TEM and ED analysis.
The physicochemical properties of metallic nickel particles were characterized in the dry state and in aqueous solutions (Table 1), including ultrapure water, PBS, and the cell culture medium (RPMI 1640 with 2% FBS) in which all cell-based assays were performed. Metallic nickel particles aggregate in aqueous solutions, resulting in an increase in mean particle diameter relative to those observed for the dry particles using TEM (Table 1, Fig. 1E). Zeta potential is seen to be both particle dependent and medium dependent but similar for all materials in the medium most relevant to in vitro cellular toxicology studies (RPMI with 2% FBS).
We used the H460 human lung epithelial cell line to investigate the effects of nickel particles in a cell type relevant to nickel carcinogenesis. H460 cells accurately recapitulate the responses of primary human cells to carcinogenic metals (Reynolds and Zhitkovich, 2007; Reynolds et al., 2009) and are appropriate for studying internalization of particles and fibers (Pietruska et al., 2010). We first verified that H460 cells were capable of taking up nickel particles since we expect the toxicity and carcinogenicity of insoluble nickel particles to depend on their uptake by target cells (Costa and Mollenhauer, 1980). H460 cells internalized both nanoparticles and microparticles, and particles could be visualized within cytoplasmic vacuoles after a 24-h exposure at a dose of 5 μg/cm2 (Fig. 2). In order to determine whether ionic nickel was mobilized from NiO nanoparticles or metallic nickel particles, nickel concentration in cell-free culture medium was measured after 2- to 72-h incubation with nickel particles using identical experimental conditions as for cell-based experiments. Nickel was readily mobilized from NiO particles, and nickel mobilization reached a plateau at 12–24 h (Fig. 3A). Incubation with equivalent doses of metallic nickel nanoparticles resulted in mobilization of over 40-fold less soluble nickel, and there was minimal nickel mobilization from metallic microparticles. When nickel mobilization was expressed as a percentage of the total available nickel, we observed that approximately 50% of the nickel in NiO nanoparticles is mobilized within 24 h, while only 1–3% of the available nickel in metallic nickel nanoparticles is mobilized, even after 72 h (Fig. 3B).
While these data indicate that ionic nickel can be mobilized from nickel nanoparticles into the extracellular medium, we wanted to determine whether exposure to nickel nanoparticles also resulted in intracellular accumulation of ionic nickel in H460 cells. In order to evaluate qualitatively the presence of intracellular mobilized nickel, cells were exposed to NiCl2 or nickel particles and loaded with Newport Green, a cell-permeant dye that fluoresces in the presence of ionic nickel. Intracellular ionic nickel release was detected by Newport Green fluorescence within 24 h of exposure to NiCl2 or NiO nanoparticles and within 48 h of exposure to metallic nickel nanoparticles (Fig. 3C).
In order to investigate cellular responses to intracellular mobilized nickel, we evaluated activation of HIF-1α, which is stabilized in response to nickel as well as in response to hypoxia and induces transcription of numerous target genes involved in hypoxia signaling (Salnikow and Zhitkovich, 2008). Compared to control, NiCl2 induced HIF-1α stabilization as early as 4–6 h, which persisted throughout a 24-h exposure. NiO nanoparticles also induced robust HIF-1α stabilization, although HIF-1α expression decreased at later time points due to cytotoxicity (Figs. 4A and and5B).5B). Exposure to metallic nickel nanoparticles, but not to microparticles, resulted in HIF-1α stabilization, albeit it to a lesser extent than mass equivalent doses of NiO and with a delayed response (Fig. 4A). In all cases, stabilized HIF-1α localized exclusively to cell nuclei (Fig. 4B). In order to confirm that stabilized HIF-1α was functional, we evaluated expression of N-myc downstream-regulated gene 1 (NDRG1), a HIF-1α target that is induced following nickel exposure (Zhou et al., 1998). NRDG1 was upregulated within 24 h of HIF-1α stabilization following exposure to NiCl2, NiO nanoparticles, and metallic nickel nanoparticles but not to metallic micron-sized particles. After 72 h, increased NDRG1 expression was also induced in control cultures, consistent with previous reports of increased NDRG1 over time in culture (Karaczyn et al., 2006). Taken together, these results indicate that ionic nickel mobilized from nickel oxide and metallic nickel nanoparticles activates the HIF-1α pathway in H460 cells.
In order to evaluate the cytotoxic effects of nickel particles, we evaluated fluorescence of the DNA-binding dye PicoGreen as a surrogate measure of cell number. We first confirmed that NiCl2 and nickel particles did not interfere with the fluorescence of PicoGreen (Fig. 5A). NiCl2 induced a dose-dependent and time-dependent reduction in cell number during a 72-h exposure (Fig. 5B). Mass equivalent doses of nickel particles exhibited different cytotoxic potential; NiO nanoparticles and metallic nickel nanoparticles, albeit it to a lesser extent, exhibited dose-dependent and time-dependent toxicity, as indicated by reduced cell number. Micron-sized metallic nickel particles were not cytotoxic at any of the tested doses or times (Fig. 5B). When normalized by the total amount of nickel in the assay on a molar basis, we observed that NiCl2 and NiO were approximately equitoxic, metal nickel nanoparticles had intermediate toxicity, and metal nickel microparticles induced no significant change in cell number (Fig. 5C). This differential toxicity was not limited to the H460 cell line as the relative toxicity of each particle was conserved in primary human bronchial epithelial cells (Fig. 5C).
We next investigated whether the decrease in cell number following nickel exposure was associated with induction of apoptosis. Similar to soluble NiCl2, NiO nanoparticles induced cleavage of caspase-3, caspase-7, and PARP in H460 cells within 24–48 h of exposure (Figs. 6A and 6B). The loss of PARP and β-actin observed after exposure to 20 μg/cm2 NiO nanoparticles is likely due to extreme cytotoxicity as this treatment reduced cell number by 75% (Fig. 4B). Metallic nickel nanoparticles also induced cleavage of caspase-3, caspase-7, and PARP following 48–72 h of exposure (Fig. 6C), while micron-sized metallic nickel particles did not induce cleavage of caspases or PARP during a 72-h exposure (Fig. 6D), in agreement with the differential toxicity observed following exposure of H460 cells to nickel-containing particles. Taken together, these results indicate that nickel-containing nanoparticles induce apoptosis in H460 cells, consistent with the same mechanism of cell death induced by ionic nickel.
Intracellular mobilization of nickel ions from poorly soluble nickel particles is hypothesized to contribute to nickel carcinogenicity (Goodman et al., 2011). In this study, nickel microparticles and nanoparticles were internalized into cytoplasmic vacuoles by human lung epithelial cells (Fig. 2). Bioavailability of nickel as assessed in an acellular assay was minimal from metallic microparticles compared with mobilization of up to 3% of available nickel from metallic nickel nanoparticles after 72 h and 50% of available nickel from NiO nanoparticles after 24 h (Fig. 3B). The kinetics and magnitude of intracellular nickel mobilization as assessed by Newport Green fluorescence (Ke et al., 2007) corresponded with this acellular assay, with increased fluorescence 24 h after exposure to NiO nanoparticles followed by intracellular mobilization from metallic nickel after 48 h. Lower Newport Green fluorescence was observed 48 h after exposure to metallic nickel microparticles (Fig. 3C). We conclude that intracellular mobilization of nickel is related to surface area and thus inversely related to particle diameter (Table 1; Fig. 1E) confirming our previous acellular bioavailability assays using metallic nickel particles (Liu et al., 2007). Increased toxicity and intracellular mobilization of nickel following exposure of human cells to NiO nanoparticles in comparison with NiO microparticles was reported by Horie et al. (2009). NiO nanoparticles have also been shown to have high lung toxicity in acute in vitro and in vivo and in subchronic in vivo assays (Cho et al., 2010; Lu et al., 2009; Ogami et al., 2009).
The relative potency and kinetics of intracellular nickel mobilization from these nickel particles were compared by stabilization and nuclear translocation of HIF-1α protein and induction of NRDG1, also known as Cap43, a protein induced by hypoxia and nickel (Zhou et al., 1998). Soluble nickel rapidly activated the HIF-1α pathway, while NiO nanoparticles induced early and sustained activation. Metallic nickel nanoparticles showed slower but also sustained activation, while metallic nickel microparticles showed no activation after 48–72 h (Fig. 4).
Salnikow et al. (2000) showed that sustained activation of the HIF-1α pathway is correlated with intracellular delivery of nickel (II) ions. Davidson et al. (2005) reported that soluble nickel enters cells by the divalent metal ion transporter and competes with iron sites on prolyl hydroxylases. Inhibition of HIF-prolyl-hydroxylase decreases binding of Von Hippel-Lindau protein and stabilizes HIF-1α protein, resulting in sustained activation of the HIF-1α pathway (Davidson et al., 2005, 2006). Other members of the iron-dependent and 2-oxyglutarate-dependent dioxygenase family are potential targets of nickel and may be associated with epigenetic modifications including decreased methylation, chromatin condensation, and inhibition of ABH3, a DNA repair enzyme (Chen and Costa, 2009). These epigenetic effects, in addition to activation of the HIF-1α hypoxia signaling pathway, are hypothesized to contribute to the carcinogenicity of soluble nickel and poorly soluble compounds (Costa et al., 2005). In this study, metallic nickel and nickel oxide nanoparticles also caused sustained activation of the HIF-1α hypoxia signaling pathway in human lung epithelial cells, raising concern about their potential carcinogenicity. This study shows that nanoparticles are a more efficient delivery vehicle for nickel ions and are more potent in activating the HIF-1α hypoxia signaling pathway than metallic nickel microparticles or soluble nickel. For example, a 24-h exposure to 5 μg/cm2 <100 nm NiO induced a stronger HIF-1α and NDRG1 response than a 24-h exposure to 100μM NiCl2 (Fig. 5A) doses which represent approximately equivalent amounts of nickel on a micromolar basis. Thus, of the two mechanisms of Ni ion delivery into the cell (ion transport vs. endocytosis with subsequent intracellular Ni ion release), these studies suggests that endocytosis is the more efficient delivery mechanism in this experimental model system.
After 24–72 h, NiCl2 and NiO nanoparticles were equally toxic to human lung epithelial cells followed by metallic nickel nanoparticles, but not by metallic nickel microparticles, at doses up to 20 μg/cm2 (Fig. 5). Similar ranking of toxicity was observed in the H460 lung epithelial cell line as well as in primary human bronchial epithelial cells. In order to investigate the mechanism responsible for nickel-induced toxicity, we evaluated markers of apoptosis: cleavage of caspase-3, caspase-7, and PARP after 24 and 48 h of exposure. NiCl2, NiO nanoparticles, and metallic nickel nanoparticles showed a dose- and time-dependent cleavage of caspases and PARP as markers of apoptosis; however, this does not exclude the possibility that Ni-containing particles also induce necrosis in a subset of the population. The signaling pathways responsible for nickel-induced apoptosis are complex and cell-type specific (Barchowsky and O’Hara, 2003). Regardless of the specific signaling pathways involved in nickel-induced apoptosis, high doses of soluble or poorly soluble nickel compounds are toxic for human lung epithelial cells and may provide selective pressure for survival and growth of cells resistant to nickel-induced toxicity (Salnikow and Zhitkovich, 2008).
Activation of the HIF-1α pathway following uptake of metallic nickel and nickel oxide nanoparticles by human lung epithelial cells has significant implications for potential carcinogenicity of nanomaterials containing bioavailable nickel (Maxwell and Salnikow, 2004). Soluble and poorly soluble nickel compounds have been shown to activate the HIF-1α pathway (Davidson et al., 2006; Salnikow and Zhitkovich, 2008) and to transform cells in culture (Biedermann and Landolph, 1987; Costa and Mollenhauer, 1980; Patierno et al., 1993). While soluble nickel (II) ions show acute toxicity at high concentrations, poorly soluble nickel compounds produce sustained HIF-1α activation leading to selection of transformed cells resistant to acute nickel-induced toxicity and elevated expression of HIF-1α responsive genes (Salnikow et al., 1999). Overexpression of HIF-1α is frequently found in solid human cancers, especially in regions of hypoxia (Zhong et al., 1999), which may facilitate selection of cells resistant to hypoxia and accelerate tumor progression (Semenza, 2000). A key role for HIF-1α in cell transformation was demonstrated by increased colony formation of wild-type mouse fibroblasts in soft agar following repeated exposure to soluble nickel while HIF-1α-deficient cells showed decreased growth in soft agar (Salnikow et al., 2003).
Metallic and metal oxide nanoparticles are currently evaluated for toxicity using a combination of in vitro and in vivo assays for surface reactivity, lung injury, and inflammation (Cho et al., 2010; Lu et al., 2009). The results presented in this paper suggest additional screening assays including cellular bioavailability of metal ions and activation of the HIF-1α pathway for toxicological evaluation of nanoparticles containing nickel. New and existing commercial nanomaterials of high priority for screening include nickel and nickel oxide nanoparticles (Cho et al., 2010), nickel nanowires (Byrne et al., 2009), and Ni-containing carbon nanotubes (Liu et al., 2007), as well as nickel nanocomposites for medical implants and drug delivery (Fadeel and Garcia-Bennett, 2010).
National Institutes of Health Superfund Research Program (P42 ES013660, R01 ES016178); Environmental Protection Agency STAR (R833862).
We gratefully acknowledge Sara Pacheco for her assistance with statistical analyses. A preliminary report of this work was published in the Conference Proceedings, Volume 2—Implications on “International Perspectives on Environmental Nanotechnology” sponsored by U.S. Environmental Protection Agency Region 5, Superfund Division held on October 7–9, 2008, Chicago, Illinois. This research does not necessarily reflect the views of either the National Institutes of Health or the Environmental Protection Agency.