Skip to main content
NCBI home page
Toxicology Research logoLink to Toxicology Research
. 2015 Dec 22;5(2):482–491. doi: 10.1039/c5tx00384a

Physicochemical insights of irradiation-enhanced hydroxyl radical generation from ZnO nanoparticles†

Qingbo Yang a, Tien-Sung Lin b, Casey Burton a, Sung-Ho Park b, Yinfa Ma a,✉
PMCID: PMC6061809  PMID: 30090362

graphic file with name c5tx00384a-ga.jpgThe widespread use of zinc oxide nanoparticles (ZnO NPs) has raised environmental and human health concerns owing to their significant cytotoxicity.

Abstract

The widespread use of zinc oxide nanoparticles (ZnO NPs) has raised environmental and human health concerns owing to their significant cytotoxicity. Although their cytotoxic effects have been associated with reactive oxygen species (ROS), the physicochemical mechanism underlying this phenomenon remains poorly understood. In this study, the physicochemical properties of ZnO NPs were systematically investigated in relation to their effect on ROS generation. Factors that were found to affect hydroxyl radical (˙OH) generation included: NP concentration, irradiation, NP hydrodynamic size, localized pH, ionic strength, NP zeta-potential, and dissolved oxygen levels. The mechanism by which ˙OH was generated under alkaline conditions was found to obey first-order reaction kinetics that followed the conversion of OH– anions and dissolved O2 to ˙OH. Based on these findings, we propose that ZnO NP cytotoxicity involves ˙OH adsorption to the nanoparticle surface, creating a highly localized source of ROS capable of potentiating oxidative damage to cellular structures. This hypothesis was evaluated with time-resolved intracellular calcium [Ca]i imaging that irradiated ZnO NPs triggered cytoplasmic calcium influxes and facilitated nuclear degradation. Together these findings present a novel physicochemical mechanism for ˙OH generation from ZnO NPs with significant implications for nanoparticle cytotoxicity and their relation to human health.

Introduction

Zinc oxide nanoparticles (ZnO NPs) remain one of the most widely used nanomaterials with applications spanning research and industrial fields such as semiconductors, cosmetics, drug delivery agents, and more.1 However, a growing body of literature has indicated that ZnO NPs possess unique physicochemical properties that may lead to adverse biological effects.2,3 For example, ZnO NPs adversely affect cells through membrane disruption,4 increased lipid peroxidation,5 reactive oxygen species (ROS) generation,6 and destruction of important organelles including mitochondria,6 lysosomes,6 and even nuclear degradation.7 Prior research has attributed ZnO NP cytotoxicity to the release of intracellular zinc ions8,9 that contribute to oxidative stress, dysregulated calcium [Ca2+],10 mitochondrial malfunction,11 as well as interleukin (IL)-8 production.12,13 Further efforts have attempted to correlate NP cytotoxicity with hydrodynamic size,14 dosage,5 and exposure conditions.5,12 However, these factors alone overlook many of the unique physicochemical properties of ZnO. For example, ZnO is a semiconductor with remarkable photo-catalytic properties,15 especially at micro- and nano-sized levels. For this reason, ZnO NPs have been widely used for microorganism sterilization and wastewater treatment.16

Photocatalytic ROS generation from nanomaterials represents another potential mode of cytotoxicity.17 Jaeger and Bard demonstrated as early as the 1970s that irradiated TiO2 generates hydroxyl radicals (˙OH) and perhydroxyl radicals (HO2˙).18 Similarly, our group and others have successfully demonstrated the formation of reactive ˙OH from irradiated ZnO NPs.19,20 To this end, electron spin resonance (ESR) spectroscopy has proved to be an invaluable technique for studying the formation of free hydroxyl radical generation and irradiated ZnO NPs.12,20,21 The underlying mechanism is thought to arise from the transfer of absorbed photon energy through electron–hole pairs on the nanoparticle surface, although the extent to which this process generates ROS remains unclear. Current research in understanding this mechanism has not yet considered secondary interactions that arise from the surrounding environment.22 For example, ZnO NP suspensions behave as a pseudo-colloidal system,23 which has important biological ramifications including nanoparticle stability, dispersivity, surface charge, and aggregation and precipitation potential in biological matrices. This potential for aggregation results in a net loss of reactive surface area and will consequently limit ROS generation by the nanoparticles.

We hypothesize that ˙OH generated from irradiated ZnO NPs is the causative agent for ZnO NP cytotoxicity. The lack of a comprehensive understanding of the effects of physicochemical properties on ˙OH formation inspired us to investigate these salient features in this study. This study was designed to monitor ˙OH formation using ESR spectroscopy combined with spin-trapping techniques, which have been recognized as powerful techniques for identifying and quantifying transient free radicals.24 The ESR spin-trapping reagent 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) is particularly useful for sensitive and highly selective ˙OH quantification.25 The systematic study of physicochemical factors including NP dosage, irradiation time, NP hydrodynamic size as well as buffer pH, ionic strength, and oxygen abundance by this technique enabled a comprehensive examination of the mechanisms underlying ROS generation from ZnO NPs. These new insights will lead to improved understanding of ZnO NP cytotoxicity and its relation to human health.

Experimental

Chemicals and reagents

ZnO NPs of 10 nm size were purchased from NanoScale Materials (Manhattan, Kansas, USA), while ZnO NPs of 50–70 nm size in addition to micro-sized ZnO powder (420 nm) and titanium(iv) oxide (TiO2; 40 nm), cerium(iv) oxide (CeO2; 20 nm) and silicon dioxide (SiO2; 46 nm) nanoparticles were purchased from Sigma-Aldrich (Saint Louis, MO, USA) at 99.0% purity. The 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) spin-trapping agent and penicillin–streptomycin reagents were purchased from Sigma-Aldrich Chemical (St. Louis, MO, USA). Fetal bovine serum was purchased from American Type Culture Collection (ATCC) (Manassas, VA, USA). Ham's F-12K medium with added l-glutamine was purchased from Fisher Scientific (Pittsburgh, PA, USA). Fura-2 AM intracellular Calcium fluorescent dye, Trypsin–EDTA (0.25%), and 0.1 M phosphate buffered saline (PBS) were purchased from Life Technology Co. (Carlsbad, CA, USA). Ultra-pure water was generated with a Milli-Q system (Millipore, Bedford, MA, USA).

ZnO particles characterization and suspension preparation

The ZnO NPs used in this study have previously been characterized.5,19 ZnO NP suspensions were freshly prepared in either PBS buffer solution, ultra-pure (MQ) water, or serum-free culture medium based on experimental design. The selection of PBS buffer solution, despite its detrimental effects on photocatalytic oxidation by TiO2, ZnO as well as FeOx, was premised on its biological relevance.26 Nanoparticles were dispersed using an ultrasonicator (FS-60H, Fisher Scientific, Pittsburg, PA, USA) for 15 minutes. Because ZnO NPs tend to aggregate or precipitate in suspension, hydrodynamic sizes and zeta potentials of ZnO NPs in various dispersants (e.g. cell medium, ultra-pure water, PBS buffer) at various concentrations were examined using dynamic light scattering (nano series Malvern Zetasizer ZEN 3690, Malvern Instruments Ltd, Worcestershire, UK). The UV–Vis spectra were recorded over the range of 350–900 nm with a Cary 50 UV–Vis spectrophotometer (Agilent Technologies, Santa Clara, CA). All pH measurements were conducted using an Accumet AB15 Plus pH meter (Thermo Fisher Scientific Inc., Waltham, MA, USA).

Electron spin resonance (ESR) experiments

DMPO spin adducts are stable nitroxides with unique ESR spectral patterns for each free radical (R˙) added to the 2-position (β-carbon) of DMPO (reaction (1)) with particular sensitivity for ˙OH.20,27,28 Here, a 100 μL addition of 100 mM DMPO was added to freshly prepared and ultrasonicated ZnO NPs in either ultra-pure water or PBS. The resulting mixture was introduced immediately into a tip-sealed disposal long tip pipette (OD = 1.2 mm, Pyrex, Fisher Scientific, Pittsburgh, PA, USA). The pipette was then inserted into the cavity of an X-band ESR spectrometer (Model JES-FA 100, JEOL Inc., Peabody, MA, USA). The JEOL rectangular resonator has a 0.5 inch hole in the front for light access. Spectrometer parameters included: microwave frequency: 8.8–9.0 GHz, power: 2–5 mW and modulation width: 0.01–0.05 mT. Single measurements were taken, as opposed to signal averaging, owing to the high sensitivity of the measurements. Moreover, daily standards were freshly prepared and examined as reference values with inter-daily precision less than 5% relative standard deviation (RSD). A combination of 100 W tungsten/halogen and 150 W xenon/mercury lamps were used as irradiation sources. ESR signals were recorded before, during, and after irradiation to examine the effect of radiation on free radical generation, and all experiments were performed at room temperature. The ESR experiments were confined to approximately 10 minutes in order to minimize heat generation caused by the irradiation. Control samples were additionally analyzed to consider the presence of interfering ESR artifacts such as oxaziranes that isomerized from spin trap nitrones.25

Since dissolved oxygen in aqueous solutions may affect free radical generation, solutions purged at 50 mL gas per min with either nitrogen or oxygen were examined to study this effect. The sample tube was capped with a rubber septum with two hypodermic needles inserted (inlet and outlet) during the gas purging. Following purging, the needles were removed and the samples remained sealed with septa under a N2 atmosphere during the ESR measurements.

Another interfering reaction involves the non-radical, nucleophilic reaction through the so-called Forrester–Hepburn mechanism.29 This reaction entails the formation of DMPO/˙SO3– artifacts from bisulfite that is typically present under real biological conditions. In this study, cell extracts were not investigated with ESR spectroscopy, so this potential interference was not anticipated.

Inductively-coupled plasma-mass spectrometry (ICP-MS) measurements

An Elan DRCe ICP-MS (PerkinElmer, Waltham, MA, USA) was used to quantify zinc ions released from ZnO NPs. Zinc dissolution was measured in ZnO NP suspensions (500 μg mL–1) prepared in 0.1 M PBS that were pH adjusted (1–14). Solutions were ultrasonicated for 15 minutes followed by vortex mixing for 3 minutes. Samples were then centrifuged at 12 000 g-force (5810 R Centrifuge, Eppendorf, Germany) and the supernatants were immediately filtered twice using disposable 0.22 μm nylon filters (Fisher Scientific, Pittsburgh, PA, USA). Samples were finally diluted using 1% nitric acid (trace metal grade, Fisher Scientific, Pittsburgh, PA, USA) prior to HPLC-ICP-MS analysis.

Cell culture and treatment with ZnO NPs

The human alveolar carcinoma-derived cell line (A549) was purchased from ATCC (Manassas, VA, USA) and used as an in vitro cytotoxicity model in this study. This cell line has been widely used in particulate matter-related pulmonary toxicity studies,30–32 and was used in our previous work to demonstrate irradiation-enhanced ZnO NP cytotoxicity.19 Cells were maintained in Ham's F-12 K medium supplemented with 5% fetal bovine serum, 100 units per mL penicillin, 100 μg mL–1 streptomycin, and grown at 37 °C in a 5% CO2 humidified environment. In each test, cells were seeded and allowed to attach for 48 hours prior to nanoparticle exposure. Cell densities between 5 × 104 and 1 × 105 cells per milliliter were used for analysis. Cells without ZnO NP exposure were used as the control group in each experiment.

Calcium imaging and three-dimensional (3-D) images plot

Calcium imaging was performed using Fura-2 dye labels for intracellular calcium in A549 cells. Cells were pre-seeded onto confocal petri dishes in whole culture medium (with serum) 24 hours prior to ZnO NP exposure (serum free). A series of calcium images that displayed Fura-2 fluorescence intensity were continuously taken for up to 6 hours. Ratiometric data analysis was conducted to identify cells with intracellular calcium levels that exceeded a preselected threshold (100 nM [Ca2+]i). 3-D plots that represented dynamic changes in intracellular Ca2+ spatial distributions were drawn using ImageJ (National Institutes of Health, USA).

Results and discussion

The effect of irradiation on ˙OH generation

Free hydroxyl radicals have extremely short life spans and are therefore ill-suited for conventional quantitative techniques.33 In this study, the highly sensitive and selective spin-trapping reagent, 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) was used to trap the transient ˙OH as stable DMPO-spin adducts (reaction (1)).

graphic file with name c5tx00384a-u1.jpg 1

The spin-trapping spectrum of aqueous ZnO NPs using a 150 W Xe/Hg lamp for irradiation is shown in Fig. 1. Briefly, the 4-line spectrum characterized by its 1 : 2 : 2 : 1 pattern (Fig. 1a) is attributed to DMPO/˙OH spin adducts with the following parameters: g = 2.0046, aN = aHβ = 1.49 mT, which agrees with literature values for DMPO/˙OH spin adducts.34,35 A time profile of the third peak before and during irradiation was then recorded for eight minutes (Fig. 1b). The stabilization of the ESR signal during irradiation implied limited free radical formation that decreased linearly following cessation of irradiation. This gradual, linear decrease was attributed to the relative stability of the DMPO/˙OH spin adducts which have a lifetime of several hours. A secondary explanation for this decrease is ZnO NP aggregation and precipitation,36 which is a phenomenon that has been previously observed.19

Fig. 1. ESR recording of ˙OH radical generation by 50–70 nm ZnO NPs in aqueous solution. (a) Representative spectrum after 2 minutes irradiation, and (b) a time profile of ESR recording of the third peak for eight minutes. Instrumentation settings: frequency = 8905.758 MHz, field center = 317.500 mT, width (±) = 4.000 mT, MOD: Fq = 100.00 kHz, width = 0.1000 mT, power = 2 mW, sweep time = 2.0 min, mod amplitude: CH1 = 1000.0, CH2 = 2.0.

Fig. 1

Open in a new tab

The effect of nanoparticle concentration on ˙OH generation

ESR measurements were performed on 50–70 nm ZnO NPs with doses ranging from 0.1 to 30 mg mL–1 (Fig. 2 and S1†). Our results demonstrated a non-linear relationship between ˙OH generation and ZnO NP concentration during the initial six minute irradiation period. Curiously, an inflection point was observed for 1 mg mL–1 ZnO NP doses (Fig. 2a, 1st irradiation), which suggested an equilibrium process between: (1) internal exciton formation and (2) interfacial electron transfer. The irradiation-induced electron–hole pairs will either combine together to form an exciton internally or are charge transferred by interfacial reactions. Higher concentrations of nanoparticles will effectively shield inner nanoparticles, leading to decreased interfacial reactions and higher rates of internal exciton formation. This understanding has been attributed to the inflection point observed during this experiment.

Fig. 2. Continuous ESR recording of ˙OH by 50–70 nm ZnO NPs. (a) 1st irradiation and dark (lamp off) periods recording at NP doses of 0.1, 0.5, 1, 3, 10 and 30 mg mL–1; (b) 2nd irradiation at ZnO NP doses of 0.5 and 3 mg mL–1. Raw ESR data was normalized to (1) the temporal linear propagation rate for the first and fourth ESR peaks, (2) ESR instrumental parameters, and (3) a daily calibration standard using 1 mg mL–1 50–70 nm ZnO NP MQ water suspension, and the solid trend lines were drawn based on a 5th order of polynomial function.

Fig. 2

Open in a new tab

ZnO NPs were then irradiated for six minutes followed by dark conditions (Fig. 2a, lamp off). Notably, the ESR signals under dark conditions failed to correlate with NP dosage. This could be readily explained by the nitroxide decomposition on the NP surface. A second irradiation period was then studied to determine whether NP aggregation would affect the result during the first dark cycle. Indeed, our hypothesis was supported by the observation in lower concentrated NPs (upper plot) shown in Fig. 2b, that the aggregation/precipitation did not significantly affect the ˙OH generation because a similar ˙OH generation was observed. However, this was not the case with higher NP concentration (bottom plot in Fig. 2b). The observed inflecting effect (Fig. 2a) of NP dosage to the radical generation though, may also have an alternative explanation: higher NP concentration will induce stronger light scattering and weaker light absorption by DMPO. This will yield lower amount of the isomerized nitrone, and thus lower adduct concentration.

The effect of hydrodynamic size on ˙OH generation

Previous NP toxicity studies have primarily focused on the relationship between nanoparticle size and cytotoxicity;37,38 however, recent efforts have proposed that hydrodynamic size more accurately reflects nanoparticle cytotoxicity.39 In this study, hydrodynamic size was measured using a real-time light scattering method coupled with ESR detection under irradiation (Fig. 3). Non-linear decreases in hydrodynamic size, as indicated by weight-averaged distribution peaks (1600 nm to 1200 nm) and relative peak intensities (7.5 to 4.7), suggested nanoparticle aggregation processes (Fig. 3a and b). This trend continued until the critical size for precipitation was reached. Notably, this observation would imply that only small ZnO NPs were present at sufficiently long time periods (>80 min.) during the experiment described above. This aggregative process was characterized using UV/Vis spectrophotometry, wherein time-based decreases in the absorbance and band gap energy shift (from 3.36 to 3.29 eV) were observed (Fig. S2†). This phenomenon, according to H. Weller, was a quantization effect, and the physicochemical properties largely correspond to the nature of the particle size and surface.40 Together these results suggest that spontaneous aggregation of nanoparticles must be considered in the evaluation of their hydrodynamic size.

Fig. 3. Correlation of ˙OH generation and ZnO NP hydrodynamic sizes. First, time-lapsed measurement of 1 mg mL–1 50–70 nm ZnO NP hydrodynamic size were shown as (a) full spectrum distribution and (b) peak analysis. Inset plot in (b) indicates weighed-averaged particle size distribution. (c) Comparison of ˙OH generation by nano- (50–70 nm) and micro- (420 nm) sized ZnO NPs. (d) Comparison of ˙OH generation by 10 nm and 50–70 nm ZnO NPs at 1 mg mL–1 or 10 mg mL–1 dosage.

Fig. 3

Open in a new tab

To determine the effect of hydrodynamic size on ˙OH generation, ˙OH generation was measured as a function of ZnO NP hydrodynamic size using bulk-form ZnO powder (420 nm) as a control (Fig. 3c) in which ˙OH generation from ZnO NPs greatly exceeded that of bulk ZnO. Ultrasonication of ZnO NPs similarly elicited greater ˙OH generation compared with NPs that were allowed to spontaneously aggregate (Fig. S3†). Three groups of ZnO NPs prepared at 1 mg mL–1 (50–70 nm), 10 mg mL–1 (10 nm), and 10 mg mL–1 (50–70 nm) were then studied to determine the extent to which hydrodynamic size influenced ˙OH generation relative to ZnO NP concentration (Fig. 3d and S4†). Hence, the hydrodynamic size has a sizeable effect on ˙OH generation from ZnO NPs.

The effect of pH on ˙OH generation

We further evaluated the correlation between ˙OH formation and environmental physicochemical factors including radical quenching capacity, pH, matrix effect and ionic strength (IS).41,42 The colloidal stability of the ZnO NP suspension under a wide pH range was initially studied (Fig. 4). Fresh suspensions in PBS under varied pH conditions (from 1 to 14) were ultrasonicated followed by zeta (ζ)-potential measurements (Fig. 4a). Ultrasonication minimally affected suspension pH, where only a 0.3 pH unit basic shift following ultrasonication was observed (Fig. S5†). A typical colloidal ζ-potential stabilization range between pH 5 and 11 for both 10 nm and 50–70 nm ZnO NPs that was maintained between –15 and –25 mV was observed. These measurements also indicated that ZnO NPs dissolve beyond this pH range, which supported the ICP-MS analysis results (Fig. 4b). Orthogonal UV/Vis absorption analysis (Fig. S5-b†) showed a strong positive correlation between pH and ZnO NP absorption. Note that a good linear correlation can be found under the basic pH conditions. Correspondingly, basic pH conditions resulted in increased hydroxyl radical generation as indicated by ESR analysis (Fig. 4c). A first-order reaction kinetics between hydroxide anions and free hydroxyl radical formation was similarly observed under alkaline conditions (Fig. 4d), which was plotted based on the first four-minute's irradiation generated ESR signal intensity. This finding indicates a direct pH-dependent process of hydroxyl radical generation by the ZnO NPs.

Fig. 4. Physicochemical properties of ZnO NPs as functions of pH environment that presented through: (a) ζ-potential, (b) released zinc ions, and (c) ˙OH generation. (d) Plot of initial four-minute irradiation hydroxyl radical production rate (as accumulated ESR signal intensity per minutes) vs. hydroxide anion concentration in varied pH buffers. A linear fitting was drawn to see if it falls into a first-order reaction kinetics. Data shown in (a) to (c) are presented as mean ± standard deviation after triplicate (or at least duplicate) measurements.

Fig. 4

Open in a new tab

Effect of ionic strength and oxygen content on ˙OH generation

Total ionic strength (IS)42 and the effective oxygen vacancy sites on ZnO NP surfaces43 have also been reported to modulate ROS generation. This understanding was explored using ultra-pure water (IS = 0.01 mM), tap water (IS = 2 mM), and PBS buffer (pH = 7, IS = 70 mM). Results shows that ionic strength generally exhibited an inverse relationship with ˙OH generation (Fig. 5a) that displayed non-linear complexities that were later attributed to radical quenching by inorganic anions.44,45 Specifically, inorganic anions will compete for the photocatalytic oxidizing sites on the nanoparticle surface to form inorganic anion radicals that will not contribute to the ESR signal shown (Fig. 5a) as per the Langmuir–Hinshelwood kinetic mode.26

Fig. 5. Influences of radical production from (a) variable ionic strengths (IS) of ultra-pure water, tap water and a phosphate buffer solution (PBS), (b) varied oxygen abundances of non-aerated control (normal), N2 de-oxygenation, and O2 oxygenation treatments, and (c) non-centrifuged, centrifuged supernatant clear solution and re-suspended centrifugation–precipitation.

Fig. 5

Open in a new tab

Surface defects and oxygen vacancies on the ZnO NP surfaces46 further correlated with ˙OH formation.47 In this study, ZnO NP suspensions underwent a five-minute pre-treatment of either (1) N2 de-oxygenation, (2) O2 oxygenation, or (3) no pre-treatment. Then DMPO was immediately added followed by ESR analysis for 10 minutes (Fig. 5b). Heightened ˙OH formation within the oxygenated group indicated dissolved oxygen facilitates ˙OH generation. This finding supported experimental evidence related to mitochondrial dysfunctions following ZnO NP exposure.48 Finally, the non-aerated control group contributed to the greatest ˙OH formation, although this occurrence is attributed to the poor stability of the NP suspension following aeration.

It was still unclear whether the newly formed ˙OH bind to ZnO NP active sites or are released into the surrounding environment, a distinction with substantial cytotoxic implications that merited additional study. To address this critical question, centrifuged, suspended, and control ZnO NP suspensions were compared by using ESR to determine the localization of the radicals. One mg mL–1 50–70 nm ZnO NPs in PBS (pH = 7) were freshly prepared and divided into three groups before ESR analysis: (1) the original suspension, (2) the supernatant after centrifugation at 5000 rpm for 15 seconds, and (3) the re-suspended precipitate following centrifugation (Fig. 5c). Decreased ˙OH formation was observed in the centrifuged group which was attributed to ZnO NP precipitation from the centrifugation process. This observation would imply that DMPO/˙OH adducts form on the nanoparticle surface, which was supported by the resuspended group having higher initial DMPO/˙OH levels than the control group. Moreover, the trends in the resuspended group closely paralleled those in the control group following additional irradiation. These findings suggest that ˙OH preferentially binds to the nanoparticle surface rather than being released into the surrounding environment. We further evaluated this hypothesis by adding DMPO before and after centrifugation to locate the ˙OH. Similar ESR results were acquired in both setups, strongly indicating surface localization of the newly formed ˙OH. Similar conclusions were suggested in a recent study, where ROS generation was inversely proportional to the nanoparticle diameter, which demonstrated surface-catalyzed ROS generation.49 This new ˙OH formation mechanism poses serious concerns for ZnO NP nanotoxicity under biologically relevant conditions.50

The effects of organic solvent and NP composition on ˙OH generation

In an extended ESR test, glycerol and ethanol were used to suspend ZnO NPs instead of water (Fig. S6†). A six-peak signal was observed in both groups that was different from the typical four-peak signal of DMPO-trapped ˙OH adducts in water. We infer that it may correlate with carbon-centered radicals produced in organic environments.51,52 Moreover, we found that the chemical composition of the ZnO NPs also play an important role in ESR spectra. Nanoparticles of differing composition, but similar concentration and size, namely TiO2 (40 nm), SiO2 (46 nm) and CeO2 (20 nm) NPs produced remarkably different ESR spectra (Fig. S7†). Weak or no ˙OH signals were observed with irradiated TiO2 and SiO2 NP while ˙OH formation occurred in irradiated CeO2. Both four-peak (typical DMPO-trapped ˙OH) and six-peak signals were also shown in the CeO2 NP suspension (Fig. S7†), suggesting that ˙OH formation by CeO2 proceeds via a unique process.

Proposed cytotoxic mechanism

ZnO NPs can have several types of defects, such as interstitial atoms or vacancies, that are characterized as either anionic or cationic.53 Singly ionized O vacancy site (Inline graphic), and (2) doubly ionized O vacancy site (Inline graphic) exist in ZnO NPs, which give rise to an overall positive charge on ZnO NP surfaces alongside excess amounts of Zn2+, which have several key implications for ˙OH formation. First, Inline graphic is an ESR-active defect that has an effective monovalent positive charge with respect to regular O2– sites. It lies approximately 2 eV below the ZnO conduction band and acts as a recombination center under irradiation. Second, Inline graphic is a vacancy containing no electrons, having an effective divalent positive charge with respect to the normal O2– sites, and can be formed when a hole is trapped at a Inline graphic center. Emitted photons were mostly assigned to a shallowly trapped electron with a deeply trapped hole in a Inline graphic center. The Inline graphic and Inline graphic vacancy sites can chemo-adsorb negatively charged, polar molecules like OH–, H2O, and O2. Photon absorption (>3.6 eV) promotes an electron (ecb–)/hole (hvb+) pair, which may further react with adsorbed species, such as OH– ions to form ˙OH free radicals (reaction (2)). The observed four-fold increase in ˙OH ESR signal intensity as the solution pH was increased from 8.0 to 10.0 supported this understanding. Hence, locally adsorbed OH– ions at vacancy sites have important roles in ˙OH photogeneration.

hvb+ + OH– →˙OH 2
ecb– + O2 → O2˙– 3

Furthermore, the ecb– can also react with locally adsorbed O2 molecules to form superoxide anions (reaction (3)), which was observed by the weak set of ESR signals that were attributed to superoxide anions (Fig. 1a). These anions may be generated by nearby photo/Auger electron charge transfer.49 Although superoxide anion radicals are cytotoxic, their highly transient nature limits their capacity to directly present cytotoxic effects. Biologically, O2˙– is rapidly converted to H2O2 by superoxide dismutase (SOD) and further transformed to ˙OH via Fenton reaction assisted by trace amounts of transition metal ions (e.g. Cu+, Fe2+, Mn2+). Our ICP-MS analysis indicated the presence of these and other Fenton metals in both the 10 nm and 50–70 nm ZnO NPs (Fig. S8†). Special attention is given here to trace metal contamination across different batches and vendors of nanoparticles that may lead to confounding results. These subtle variations in nanoparticle compositions will not dramatically alter cytotoxic effects, but may lead to unique ROS generation properties. On the other hand, the trapping of “excited” electrons, ecb–, by O2 at the vacancy sites, and their conversion to O2˙– can further decelerate the recombination of ecb– with hvb+, and thereby enhance ˙OH formation. These pathways can favorably proceed as long as the surface defects are localized to the irradiated sites that trap hvb+ and ecb– so as to prevent exciton emission. Hence, irradiated ZnO NPs may generate ˙OH from multiple, distinct pathways (Scheme 1), which leads us to posit that ˙OH formation by intact NPs is a more likely route for cytotoxicity than dissolved zinc ions which is minimally released under biologically relevant conditions.

Scheme 1. Proposed mechanism of ˙OH generation by ZnO NPs mainly with assistance of irradiation and basic pH environment. SOD: superoxide dismutase.

Scheme 1

Open in a new tab

Extracellular calcium influx imaging

To support this hypothesis, we conducted an in vitro study to model ZnO NP cytotoxicity using extracellular calcium influx imaging. This approach is premised on the idea that divalent calcium (Ca2+) is an important secondary messenger for cell signaling that can be used to accurately monitor cytotoxic processes in a real-time manner.54,55 Since calcium influxes can be activated by ˙OH, its use in monitoring ˙OH cytotoxicity is particularly advantageous for studying ZnO NP cytotoxicity (Fig. 6).56 Four ZnO NP concentrations (7.5, 15, 25 and 50 μg mL–1) were used to induce extracellular ionic calcium ([Ca2+]i) influx in A549 epithelial lung cancer cells. A 100 nM intracellular Ca2+ level54 was selected as the triggering threshold since [Ca2+]i is physiologically maintained at or below this level. Extracellular Ca2+ influxes and subsequent cell nucleus decomposition were observed following exposure to ZnO NPs (Fig. 6a–c, three featured calcium imaging animations are also available in the ESI†). Our findings indicated that ZnO NP exposure quickly induced Ca2+ influxes that were best modeled by a linear correlation between logarithmic NP centration and the influx triggering time (Fig. 6d), indicating first-order reaction kinetics as well. These findings also indicated that nuclear membrane damage occurred simultaneously with calcium influxes, which are tell-tale signs of cell apoptosis or autophagy.19,57 This result revealed a unique manner of ZnO NP cytotoxicity through nuclear deformation. This finding supports previously observed irradiation-enhanced nuclear decomposition by ZnO NPs.19,57

Fig. 6. The correlation between ZnO NP uptake and intracellular calcium homeostasis; 3-D plot of intracellular calcium kinetics were established based on featured-cell's 2-D fluorescent intensity under three ZnO NP dosages of (a) 7.5, (b) 15, and (c) 25 μg mL–1. (d) Linear correlation between the natural logarithmic concentration of ZnO NPs and the calcium influx triggering time. Pseudo-colored regions are selected cells that are stained with Fura-2 AM to show real-time variation of the intracellular calcium concentration. Blue-purple regions represent resting status with low [Ca2+]i, while red-yellow regions represent highly increased intracellular [Ca2+]i. Each point represents 20 averaged measurements and error bars represent standard deviations.

Fig. 6

Open in a new tab

This is coincidentally supported by a recent study that high hydroxyl radical production was observed due to the formation of a structured water layer in the vicinity of the nanoparticle which is possibly through the interaction between NP's charge and the water dipoles.22 Moreover, our results indicate that ZnO NP ˙OH generation is accelerated under high oxygen and basic pH environments, which impart significant ZnO NP cytotoxicity implications as detailedly visualized in Fig. S9.† Together these findings will likely impact toxicology research, chemical enhancement58 and nanomedicine,22 such as by presenting novel therapeutic targets to limit ZnO NP cytotoxicity following exposure.

Conclusions

In summary, the physicochemical properties of ZnO NPs were systematically studied for their effects on ROS generation. ESR analysis demonstrated that ˙OH generation is influenced by irradiation time, hydrodynamic size, dosage, and local pH. Decreases in time-lapsed ESR measurements under dark conditions were attributed to nanoparticle aggregation and precipitation, which resulted in concentration-dependent inflection points in the ESR data. Preferential ˙OH generation was observed under the following conditions: irradiation, basic pH, high dissolved oxygen, and low ionic strength. Finally, the discovery of bound ˙OH to ZnO NP surfaces suggested highly concentrated, localized ROS regions that may present a novel mechanism of cytotoxicity. Future work should aim to quantify NP surface vacancy sites and impurities to substantiate this proposed mechanism.

Conflict of interest

The authors declare no competing financial interest.

Supplementary Material

Click here for additional data file. (2.1MB, pdf)

Acknowledgments

The authors gratefully thank Dr Robert S. Aronstam and Hsiu-Jen Wang for their technical assistance with the calcium imaging experiments as well as Yongbo Dan and Kun Liu for their technical assistance with the ICP analysis. This project was supported by internal funding from the Center for Single Nanoparticle, Single Cell, and Single Molecule Monitoring (CS3M), and the Department of Chemistry at Missouri University of Science and Technology. C. Burton received research funding through a National Science Foundation Graduate Research Fellowship (#DGE-1011744).

Footnotes

†Electronic supplementary information (ESI) available: Additional data including raw ESR spectra, non-ultrasonicated ZnO NPs, time-lapse UV/Vis absorption spectra, ultrasonication influence on pH, trace elemental analysis of two different sizes of ZnO NPs, final proposed cellular cytotoxic pathways as well as three calcium imaging related animations. See DOI: 10.1039/c5tx00384a

References

  1. Nel A., Xia T., Madler L., Li N. Science. 2006;311:622–627. doi: 10.1126/science.1114397. [DOI] [PubMed] [Google Scholar]
  2. Xia T., Kovochich M., Liong M., Mädler L., Gilbert B., Shi H., Yeh J. I., Zink J. I., Nel A. E. ACS Nano. 2008;2:2121–2134. doi: 10.1021/nn800511k. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Patra P., Mitra S., Debnath N., Goswami A. Langmuir. 2012;28:16966–16978. doi: 10.1021/la304120k. [DOI] [PubMed] [Google Scholar]
  4. Vandebriel R. J., De Jong W. H. Nanotechnol., Sci. Appl. 2012;5:61. doi: 10.2147/NSA.S23932. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Lin W., Xu Y., Huang C.-C., Ma Y., Shannon K. B., Chen D.-R., Huang Y.-W. J. Nanopart. Res. 2009;11:25–39. [Google Scholar]
  6. Kocbek P., Teskač K., Kreft M. E., Kristl J. Small. 2010;6:1908–1917. doi: 10.1002/smll.201000032. [DOI] [PubMed] [Google Scholar]
  7. Sharma V., Shukla R. K., Saxena N., Parmar D., Das M., Dhawan A. Toxicol. Lett. 2009;185:211–218. doi: 10.1016/j.toxlet.2009.01.008. [DOI] [PubMed] [Google Scholar]
  8. Yang S.-T., Liu J.-H., Wang J., Yuan Y., Cao A., Wang H., Liu Y., Zhao Y. J. Nanosci. Nanotechnol. 2010;10:8638–8645. doi: 10.1166/jnn.2010.2491. [DOI] [PubMed] [Google Scholar]
  9. Pujalté I., Passagne I., Brouillaud B., Tréguer M., Durand E., Ohayon-Courtès C., L'Azou B. Part. Fibre Toxicol. 2011;8:1–16. doi: 10.1186/1743-8977-8-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. George S., Pokhrel S., Xia T., Gilbert B., Ji Z., Schowalter M., Rosenauer A., Damoiseaux R., Bradley K. A., Mädler L. ACS Nano. 2009;4:15–29. doi: 10.1021/nn901503q. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Li J.-h., Liu X.-r., Zhang Y., Tian F.-f., Zhao G.-y., Jiang F.-l., Liu Y. Toxicol. Res. 2012;1:137–144. [Google Scholar]
  12. De Berardis B., Civitelli G., Condello M., Lista P., Pozzi R., Arancia G., Meschini S. Toxicol. Appl. Pharmacol. 2010;246:116–127. doi: 10.1016/j.taap.2010.04.012. [DOI] [PubMed] [Google Scholar]
  13. Hackenberg S., Scherzed A., Technau A., Kessler M., Froelich K., Ginzkey C., Koehler C., Burghartz M., Hagen R., Kleinsasser N. Toxicol. in Vitro. 2011;25:657–663. doi: 10.1016/j.tiv.2011.01.003. [DOI] [PubMed] [Google Scholar]
  14. Kato H., Suzuki M., Fujita K., Horie M., Endoh S., Yoshida Y., Iwahashi H., Takahashi K., Nakamura A., Kinugasa S. Toxicol. in Vitro. 2009;23:927–934. doi: 10.1016/j.tiv.2009.04.006. [DOI] [PubMed] [Google Scholar]
  15. Xia T., Zhao Y., Sager T., George S., Pokhrel S., Li N., Schoenfeld D., Meng H., Lin S., Wang X. ACS Nano. 2011;5:1223–1235. doi: 10.1021/nn1028482. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Niazi J. H. and Gu M. B., in Atmospheric and Biological Environmental Monitoring, Springer, 2009, pp. 193–206. [Google Scholar]
  17. Hoffmann M. R., Martin S. T., Choi W., Bahnemann D. W. Chem. Rev. 1995;95:69–96. [Google Scholar]
  18. Jaeger C. D., Bard A. J. J. Phys. Chem. 1979;83:3146–3152. [Google Scholar]
  19. Yang Q., Ma Y. Int. J. Toxicol. 2014:1091581814529168. doi: 10.1177/1091581814529168. [DOI] [PubMed] [Google Scholar]
  20. Lewicka Z. A., William W. Y., Oliva B. L., Contreras E. Q., Colvin V. L. J. Photochem. Photobiol., A. 2013;263:24–33. [Google Scholar]
  21. Kavitha V., Palanivelu K. J. Photochem. Photobiol., A. 2005;170:83–95. [Google Scholar]
  22. Sicard-Roselli C., Brun E., Gilles M., Baldacchino G., Kelsey C., McQuaid H., Polin C., Wardlow N., Currell F. Small. 2014;10:3338–3346. doi: 10.1002/smll.201400110. [DOI] [PubMed] [Google Scholar]
  23. Johnston B. D., Scown T. M., Moger J., Cumberland S. A., Baalousha M., Linge K., van Aerle R., Jarvis K., Lead J. R., Tyler C. R. Environ. Sci. Technol. 2010;44:1144–1151. doi: 10.1021/es901971a. [DOI] [PubMed] [Google Scholar]
  24. He W., Kim H., Wamer W., Melka D., Callahan J., Yin J. J. Am. Chem. Soc. 2014;136:750. doi: 10.1021/ja410800y. [DOI] [PubMed] [Google Scholar]
  25. Finkelstein E., Rosen G. M., Rauckman E. J. Arch. Biochem. Biophys. 1980;200:1–16. doi: 10.1016/0003-9861(80)90323-9. [DOI] [PubMed] [Google Scholar]
  26. Abdullah M., Low G. K., Matthews R. W. J. Phys. Chem. 1990;94:6820–6825. [Google Scholar]
  27. Janzen E. G. Acc. Chem. Res. 1971;4:31. [Google Scholar]
  28. Janzen E. G. J. Magn. Reson. 1973;9:510. [Google Scholar]
  29. Ranguelova K., Mason R. P. Magn. Reson. Chem. 2011;49:152–158. doi: 10.1002/mrc.2709. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Wottrich R., Diabaté S., Krug H. F. Int. J. Hyg. Environ. Health. 2004;207:353–361. doi: 10.1078/1438-4639-00300. [DOI] [PubMed] [Google Scholar]
  31. Upadhyay D., Panduri V., Ghio A., Kamp D. W. Am. J. Respir. Cell Mol. Biol. 2003;29:180. doi: 10.1165/rcmb.2002-0269OC. [DOI] [PubMed] [Google Scholar]
  32. Huang M., Khor E., Lim L.-Y. Pharm. Res. 2004;21:344–353. doi: 10.1023/b:pham.0000016249.52831.a5. [DOI] [PubMed] [Google Scholar]
  33. Lipovsky A., Tzitrinovich Z., Friedmann H., Applerot G., Gedanken A., Lubart R. J. Phys. Chem. C. 2009;113:15997–16001. [Google Scholar]
  34. Janzen E. G., Liu J. I. J. Magn. Reson. 1973;9:510–512. [Google Scholar]
  35. Finkelstein E., Rosen G. M., Rauckman E. J. J. Am. Chem. Soc. 1980;102:4994–4999. [Google Scholar]
  36. Fernandez-Cruz M. L., Lammel T., Connolly M., Conde E., Barrado A. I., Derick S., Perez Y., Fernandez M., Furger C., Navas J. M. Nanotoxicology. 2013;7:935–952. doi: 10.3109/17435390.2012.676098. [DOI] [PubMed] [Google Scholar]
  37. Hanley C., Thurber A., Hanna C., Punnoose A., Zhang J., Wingett D. G. Nanoscale Res. Lett. 2009;4:1409–1420. doi: 10.1007/s11671-009-9413-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Reddy K. M., Feris K., Bell J., Wingett D. G., Hanley C., Punnoose A. Appl. Phys. Lett. 2007;90:213902. doi: 10.1063/1.2742324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Hsiao I.-L., Huang Y.-J. Sci. Total Environ. 2011;409:1219–1228. doi: 10.1016/j.scitotenv.2010.12.033. [DOI] [PubMed] [Google Scholar]
  40. Weller H. Angew. Chem., Int. Ed. Engl. 1993;32:41–53. [Google Scholar]
  41. Navarro E., Baun A., Behra R., Hartmann N. B., Filser J., Miao A.-J., Quigg A., Santschi P. H., Sigg L. Ecotoxicology. 2008;17:372–386. doi: 10.1007/s10646-008-0214-0. [DOI] [PubMed] [Google Scholar]
  42. Ma H., Williams P. L., Diamond S. A. Environ. Pollut. 2013;172:76–85. doi: 10.1016/j.envpol.2012.08.011. [DOI] [PubMed] [Google Scholar]
  43. Senthilkumar K., Tokunaga M., Okamoto H., Senthilkumar O., Fujita Y. Appl. Phys. Lett. 2010;97:091907. [Google Scholar]
  44. Guo M. Y., Ng A. M. C., Liu F., Djurišić A. B., Chan W. K., Su H., Wong K. S. J. Phys. Chem. C. 2011;115:11095–11101. [Google Scholar]
  45. Li L., Abe Y., Kanagawa K., Usui N., Imai K., Mashino T., Mochizuki M., Miyata N. Anal. Chim. Acta. 2004;512:121–124. doi: 10.1016/j.aca.2007.08.008. [DOI] [PubMed] [Google Scholar]
  46. Kong J.-Z., Li A.-D., Zhai H.-F., Gong Y.-P., Li H., Wu D. J. Solid State Chem. 2009;182:2061–2067. [Google Scholar]
  47. Applerot G., Lipovsky A., Dror R., Perkas N., Nitzan Y., Lubart R., Gedanken A. Adv. Funct. Mater. 2009;19:842–852. [Google Scholar]
  48. Yu K.-N., Yoon T.-J., Minai-Tehrani A., Kim J.-E., Park S. J., Jeong M. S., Ha S.-W., Lee J.-K., Kim J. S., Cho M.-H. Toxicol. in Vitro. 2013;27:1187–1195. doi: 10.1016/j.tiv.2013.02.010. [DOI] [PubMed] [Google Scholar]
  49. Misawa M., Takahashi J. Nanomedicine. 2011;7:604–614. doi: 10.1016/j.nano.2011.01.014. [DOI] [PubMed] [Google Scholar]
  50. Wahab R., Mishra A., Yun S.-I., Kim Y.-S., Shin H.-S. Appl. Microbiol. Biotechnol. 2010;87:1917–1925. doi: 10.1007/s00253-010-2692-2. [DOI] [PubMed] [Google Scholar]
  51. Mazumdar A., Adak S., Chatterjee R., Banerjee R. Biochem. J. 1997;324:713–719. doi: 10.1042/bj3240713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Wilms L. C., Kleinjans J., Moonen E. J., Briedé J. J. Toxicol. in Vitro. 2008;22:301–307. doi: 10.1016/j.tiv.2007.09.002. [DOI] [PubMed] [Google Scholar]
  53. McCluskey M., Jokela S. J. Appl. Phys. 2009;106:071101. [Google Scholar]
  54. Clapham D. E. Cell. 1995;80:259–268. doi: 10.1016/0092-8674(95)90408-5. [DOI] [PubMed] [Google Scholar]
  55. Shi M., Zhang T., Sun L., Luo Y., Liu D.-H., Xie S.-T., Song X.-Y., Wang G.-F., Chen X.-L., Zhou B.-C. Apoptosis. 2013;18:435–451. doi: 10.1007/s10495-012-0786-2. [DOI] [PubMed] [Google Scholar]
  56. Ishii M., Shimizu S., Hara Y., Hagiwara T., Miyazaki A., Mori Y., Kiuchi Y. Cell Calcium. 2006;39:487–494. doi: 10.1016/j.ceca.2006.01.013. [DOI] [PubMed] [Google Scholar]
  57. Wang C.-C., Wang S., Xia Q., He W., Yin J.-J., Fu P. P., Li J.-H. J. Nanosci. Nanotechnol. 2013;13:3880–3888. doi: 10.1166/jnn.2013.7177. [DOI] [PubMed] [Google Scholar]
  58. Cheng N. N., Starkewolf Z., Davidson R. A., Sharmah A., Lee C., Lien J., Guo T. J. Am. Chem. Soc. 2012;134:1950–1953. doi: 10.1021/ja210239k. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Click here for additional data file. (2.1MB, pdf)

Articles from Toxicology Research are provided here courtesy of Oxford University Press

RESOURCES