Kirsten Gerloff

Cytotoxicity and oxidative DNA damage by nanoparticles in human intestinal Caco-2 cells

Abstract The use of engineered nanoparticles in the food sector is anticipated to increase dramatically, whereas their potential hazards for the gastrointestinal tract are still largely unknown. We investigated the cytotoxic and DNA-damaging effects of several types of nanoparticles and fine particles relevant as food additives (TiO2 and SiO2) or for food packaging (ZnO and MgO) as well as carbon black on human intestinal Caco-2 cells. All particles, except for MgO, were cytotoxic (LDH and WST-1 assay). ZnO, and to lesser extent SiO2, induced significant DNA damage (Fpg-comet), while SiO2 and carbon black were the most potent in causing glutathione depletion. DNA damage by TiO2 was found to depend on sample processing conditions. Interestingly, application of different TiO2 and ZnO particles revealed no relation between particle surface area and DNA damage. Our results indicate a potential hazard of several food-related nanoparticles which necessitate investigations on the actual exposure in humans.

Distinctive Toxicity of TiO2 Rutile/Anatase Mixed Phase Nanoparticles on Caco-2 Cells

Titanium dioxide has a long-standing use as a food additive. Micrometric powders are, e.g., applied as whiteners in confectionary or dairy products. Possible hazards of ingested nanometric TiO(2) particles for humans and the potential influence of varying specific surface area (SSA) are currently under discussion. Five TiO(2)-samples were analyzed for purity, crystallinity, primary particle size, SSA, ζ potential, and aggregation/agglomeration. Their potential to induce cytotoxicity, oxidative stress, and DNA damage was evaluated in human intestinal Caco-2 cells. Only anatase-rutile containing samples, in contrast to the pure anatase samples, induced significant LDH leakage or mild DNA damage (Fpg-comet assay). Evaluation of the metabolic competence of the cells (WST-1 assay) revealed a highly significant correlation between the SSA of the anatase samples and cytotoxicity. The anatase/rutile samples showed higher toxicity per unit surface area than the pure anatase powders. However, none of the samples affected cellular markers of oxidative stress. Our findings suggest that both SSA and crystallinity are critical determinants of TiO(2)-toxicity toward intestinal cells.

Influence of simulated gastrointestinal conditions on particle-induced cytotoxicity and interleukin-8 regulation in differentiated and undifferentiated Caco-2 cells

Abstract Novel aspects of engineered nanoparticles offer many advantages for optimising food products and packaging. However, their potential hazards in the gastrointestinal tract require further investigation. We evaluated the toxic and inflammatory potential of two types of particles that might become increasingly relevant to the food industry, namely SiO2 and ZnO. The materials were characterised for their morphology, oxidant generation and hydrodynamic behaviour. Cytotoxicity and interleukin-8 mRNA and protein expression were evaluated in human intestinal Caco-2 cells. Particle pretreatment under simulated gastric and intestinal pH conditions resulted in reduced acellular ROS formation but did not influence cytotoxicity (WST-1 assay) or IL-8 expression. However, the differentiation status of the cells markedly determined the cytotoxic potency of the particles. Further research is needed to determine the in vivo relevance of our current observations regarding the role of particle aggregation and the stage of intestinal epithelial cell differentiation in determining the hazards of ingested particles.

Neutrophil-derived ROS contribute to oxidative DNA damage induction by quartz particles

The carcinogenicity of respirable quartz is considered to be driven by reactive oxygen species (ROS) generation in association with chronic inflammation. The contribution of phagocyte-derived ROS to inflammation, oxidative stress, and DNA damage responses was investigated in the lungs of C57BL/6J wild-type and p47(phox-/-) mice, 24h after pharyngeal aspiration of DQ12 quartz (100 mg/kg bw). Bone-marrow-derived neutrophils from wild-type and p47(phox-/-) mice were used for parallel in vitro investigations in coculture with A549 human alveolar epithelial cells. Quartz induced a marked neutrophil influx in both wild-type and p47(phox-/-) mouse lungs. Significant increases in mRNA expression of the oxidative stress markers HO-1 and γ-GCS were observed only in quartz-treated wild-type animals. Oxidative DNA damage in lung tissue was not affected by quartz exposure and did not differ between p47(phox-/-) and WT mice. Differences in mRNA expression of the DNA repair genes OGG1, APE-1, DNA Polβ, and XRCC1 were also absent. Quartz treatment of cocultures containing wild-type neutrophils, but not p47(phox-/-) neutrophils, caused increased oxidative DNA damage in epithelial cells. Our study demonstrates that neutrophil-derived ROS significantly contribute to pulmonary oxidative stress responses after acute quartz exposure, yet their role in the associated induction of oxidative DNA damage could be shown only in vitro.

Neutrophils augment LPS-mediated pro-inflammatory signaling in human lung epithelial cells.

BACKGROUND The role of polymorphonuclear neutrophils in pulmonary host defense is well recognized. The influence of a pre-existing inflammation driven by neutrophils (neutrophilic inflammation) on the airway epithelial response toward pro-inflammatory exogenous triggers, however, is still poorly addressed. Therefore, the aim of the present study is to investigate the effect of neutrophils on lipopolysaccharide (LPS)-induced pro-inflammatory signaling in lung epithelial cells. Additionally, underlying signaling pathways are examined. METHODS Human bronchial epithelial cells (BEAS-2B) were co-incubated with human peripheral blood neutrophils or bone-marrow derived neutrophils from either C57BL/6J wild type or nicotinamide adenine dinucleotide phosphate (NADPH)-oxidase deficient (p47(phox-/-)) mice. Upon stimulation with LPS, interleukin (IL)-8 production and reactive oxygen species (ROS) generation were measured. Additionally, activation of the extracellular signal-regulated kinases (ERK) 1/2 and nuclear factor (NF)-κB signaling pathways was analyzed. RESULTS Our studies show that the presence of neutrophils synergistically increases LPS-induced IL-8 and ROS production by BEAS-2B cells without inducing cytotoxicity. The observed IL-8 response to endotoxin increases in proportion to time, LPS-concentration and the number of neutrophils present. Moreover, this synergistic IL-8 production strongly correlated with the chemotactic properties of the co-incubations and significantly depended on a functional neutrophilic NADPH oxidase. The presence of neutrophils also augments LPS-induced phosphorylation of ERK1/2 and IκBα as well as NF-κB RelA DNA binding activity in BEAS-2B cells. CONCLUSIONS Our results indicate that the pro-inflammatory effects of LPS toward lung epithelial cells are amplified during a pre-existing neutrophilic inflammation. These findings support the concept that patients suffering from pulmonary neutrophilic inflammation are more susceptible toward exogenous pro-inflammatory triggers.

Oxidative stress and DNA damage responses in rat and mouse lung to inhaled carbon nanoparticles

Abstract We have investigated whether short-term nose-only inhalation exposure to electric spark discharge-generated carbon nanoparticles (∼60 nm) causes oxidative stress and DNA damage responses in the lungs of rats (152 μg/m3; 4 h) and mice (142 μg/m3; 4 h, or three times 4 h). In both species, no pulmonary inflammation and toxicity were detected by bronchoalveolar lavage or mRNA expression analyses. Oxidative DNA damage (measured by fpg-comet assay), was also not increased in mouse whole lung tissue or isolated lung epithelial cells from rat. In addition, the mRNA expressions of the DNA base excision repair genes OGG1, DNA Polβ and XRCC1 were not altered. However, in the lung epithelial cells isolated from the nanoparticle-exposed rats a small but significant increase in APE-1 mRNA expression was measured. Thus, short-term inhalation of carbon nanoparticles under the applied exposure regimen, does not cause oxidative stress and DNA damage in the lungs of healthy mice and rats.

Investigation of the cytotoxic and proinflammatory effects of cement dusts in rat alveolar macrophages.

Exposure to cement dust, a specifically alkaline and irritant dust, is one of the most common occupational dust exposures worldwide. Although several adverse respiratory health effects have been associated with cement dust exposure, the evidence is not conclusive. In the current study, cytotoxic and pro-inflammatory effects as well as oxidative stress elicited by a number of cement dusts, including a limestone and cement clinker sample, were tested using the NR8383 rat alveolar macrophage cell line and primary rat alveolar macrophages. DQ12 quartz and TiO(2) were included as positive and negative controls, respectively. Cytotoxicity was determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide assay and the lactate dehydrogenase assay, oxidative stress was determined by measurement of the depletion of total cellular glutathione, and electron spin resonance was applied to determine reactive oxygen species (ROS) generation. The release of the cytokines tumor necrosis factor-alpha (TNFalpha), interleukin-1 beta (IL-1 beta), and macrophage inflammatory protein-2 (MIP-2) was determined by enzyme-linked immunosorbent assay. None of the dust samples were found to cause toxicity to the macrophages or notable glutathione depletion when compared to DQ12. The cement samples also failed to activate macrophages for the generation of ROS and the production of inflammatory cytokines IL-1 beta and MIP-2. In contrast, however, most of the cement dusts were found to activate macrophage TNFalpha production, and this was significantly associated with their content of CaO. Further research is needed to determine the relevance of these in vitro observations for occupational cement dust exposure settings.

Oxidant generation, DNA damage and cytotoxicity by a panel of engineered nanomaterials in three different human epithelial cell lines

Due to the steeply increased use of nanomaterials (NMs) for commercial and industrial applications, toxicological assessment of their potential harmful effects is urgently needed. In this study, we compared the DNA-damaging properties and concurrent cytotoxicity of a panel of 10 engineered NMs in three different cell lines in relation to their intrinsic oxidant generating properties. The human epithelial cell lines A549, HK-2 and HepG2 were chosen to represent relevant target organs for NMs in the lung, kidney and liver. Cytotoxicity, evaluated by WST-1 assay in the treatment concentration range of 0.3–80 µg/cm2, was shown for Ag and ZnO NM in all three cell lines. Cytotoxicity was absent for all other NMs, i.e. five types of TiO2 and two types of multiwalled carbon nanotubes. DNA damage, evaluated by the alkaline comet assay, was observed with Ag and ZnO, albeit only at cytotoxic concentrations. DNA damage varied considerably with the cell line. The oxidant generating properties of the NMs, evaluated by electron spin resonance spectroscopy in cell free conditions, did not correlate with their cytotoxic or DNA-damaging properties. DNA damage by the nanosilver could be partly attributed to its surfactant-containing dispersant. The coating of a TiO2 sample with the commercial surfactant Curosurf augmented its DNA-damaging properties in A549 cells, while surface modification with serum tended to reduce damage. Our findings indicate that measurement of the intrinsic oxidant-generating capacity of NMs is a poor predictor of DNA damage and that the cytotoxic and DNA-damaging properties of NMs can vary substantially with experimental conditions. Our study also underlines the critical importance of selecting appropriate cell systems and aligned testing protocols. Selection of a cell line on the mere basis of its origin may provide only poor insight on organ-specific hazards of NMs.

Particles which may occur in food or food packaging can exert cytotoxicity and (oxidative) DNA damage in target cells of the human intestine

SIR: In a paper recently published inNanotoxicology we have reported our investigations on the cytotoxic and DNAdamaging effects of a set of nanoparticles and fine particles in human intestinal Caco-2 cells (Gerloff et al. 2009). As examples of possible food additives we selected TiO2 and SiO2 particles. As promising constituents of food packaging materials ZnO and MgO particles were tested, while carbon black was investigated as a model for ambient air nanoparticles. The major message of this study was that specific types of particles which may occur in food or food packaging can exert cytotoxicity and (oxidative) DNA damage in target cells of the human intestine. Among other things in this study, we also addressed the role of the specific surface area (SSA) in the observed effects. The SSA has been considered an important determinant of toxicity and inflammatory potency for poorly soluble particles of low toxicity (Oberdörster et al. 2005a; Monteiller et al. 2007), and is typically measured by means of N2-adsorption according to the method of Brunauer, Emmett and Teller (BET method). We investigated the role of the SSA by comparing the effects at equal mass concentrations of three different commercially available samples of ZnO as well as TiO2, differing in their BET values according to the informations provided by the sample suppliers. The experiments revealed that all three ZnO samples were equally effective in inducing (oxidative) DNA damage as shown by the formamidopyrimidine glycosylase-modified comet assay, whereas damage was not observed with any of the three TiO2 samples (Gerloff et al. 2009). Similarly, we could not observe a clear association between the claimed BET values of the samples and their cytotoxic effects measured by two independent assays (i.e., LDH and WST-1 tests). Taken together, this led us to conclude that the SSA may not necessarily be the best dose-metric to predict the toxicity of poorly soluble low toxicity particles in the gastrointestinal tract and/or their DNA damaging properties in general, as opposed to inflammogenic and toxic potency of such particles in the respiratory tract (Oberdörster et al. 2005a; Monteiller et al. 2007). However, after the publication of our previous study, by chance we found new and contradictory information regarding the SSA of one of the tested TiO2 samples, and this was confirmed upon our further inquiries to the supplier. Upon a detailed re-analysis of the sample by the BET method, we found out that the SSA value listed by the supplier in the catalogue as well as on the label of the specific batch that was sent to us was incorrect. Contrary to a declared SSA of >300 m/g, this TiO2 sample appeared to have a BET-value of only 53 m/g, closely resembling that of one of the two other tested TiO2 samples. Hence, our statement on the apparent lacking association between particle surface area and toxicity towards the Caco-2 cells could not be truly made for TiO2 in our paper (Gerloff et al. 2009). Since we also relied on supplier-provided BET values for the ZnO materials, we conclude that the experiments described in our previous study do not allow for any judgment on the predictive value of the particle SSA as a dose-metric for the described effects. However, the major message of our study, on the potential hazard of specific types of food-occurring nanoparticles and fine particles, in our opinion remains justified. A thorough evaluation of the physicochemical properties and composition of the materials used isSIR: In a paper recently published inNanotoxicology we have reported our investigations on the cytotoxic and DNAdamaging effects of a set of nanoparticles and fine particles in human intestinal Caco-2 cells (Gerloff et al. 2009). As examples of possible food additives we selected TiO2 and SiO2 particles. As promising constituents of food packaging materials ZnO and MgO particles were tested, while carbon black was investigated as a model for ambient air nanoparticles. The major message of this study was that specific types of particles which may occur in food or food packaging can exert cytotoxicity and (oxidative) DNA damage in target cells of the human intestine. Among other things in this study, we also addressed the role of the specific surface area (SSA) in the observed effects. The SSA has been considered an important determinant of toxicity and inflammatory potency for poorly soluble particles of low toxicity (Oberdörster et al. 2005a; Monteiller et al. 2007), and is typically measured by means of N2-adsorption according to the method of Brunauer, Emmett and Teller (BET method). We investigated the role of the SSA by comparing the effects at equal mass concentrations of three different commercially available samples of ZnO as well as TiO2, differing in their BET values according to the informations provided by the sample suppliers. The experiments revealed that all three ZnO samples were equally effective in inducing (oxidative) DNA damage as shown by the formamidopyrimidine glycosylase-modified comet assay, whereas damage was not observed with any of the three TiO2 samples (Gerloff et al. 2009). Similarly, we could not observe a clear association between the claimed BET values of the samples and their cytotoxic effects measured by two independent assays (i.e., LDH and WST-1 tests). Taken together, this led us to conclude that the SSA may not necessarily be the best dose-metric to predict the toxicity of poorly soluble low toxicity particles in the gastrointestinal tract and/or their DNA damaging properties in general, as opposed to inflammogenic and toxic potency of such particles in the respiratory tract (Oberdörster et al. 2005a; Monteiller et al. 2007). However, after the publication of our previous study, by chance we found new and contradictory information regarding the SSA of one of the tested TiO2 samples, and this was confirmed upon our further inquiries to the supplier. Upon a detailed re-analysis of the sample by the BET method, we found out that the SSA value listed by the supplier in the catalogue as well as on the label of the specific batch that was sent to us was incorrect. Contrary to a declared SSA of >300 m/g, this TiO2 sample appeared to have a BET-value of only 53 m/g, closely resembling that of one of the two other tested TiO2 samples. Hence, our statement on the apparent lacking association between particle surface area and toxicity towards the Caco-2 cells could not be truly made for TiO2 in our paper (Gerloff et al. 2009). Since we also relied on supplier-provided BET values for the ZnO materials, we conclude that the experiments described in our previous study do not allow for any judgment on the predictive value of the particle SSA as a dose-metric for the described effects. However, the major message of our study, on the potential hazard of specific types of food-occurring nanoparticles and fine particles, in our opinion remains justified. A thorough evaluation of the physicochemical properties and composition of the materials used is