Oxidatively damaged DNA in rats exposed by oral gavage to C60 fullerenes and single-walled carbon nanotubes.

BACKGROUND
C60 fullerenes and single-walled carbon nanotubes (SWCNT) are projected to be used in medicine and consumer products with potential human exposure. The hazardous effects of these particles are expected to involve oxidative stress with generation of oxidatively damaged DNA that might be the initiating event in the development of cancer.


OBJECTIVE
In this study we investigated the effect of a single oral administration of C60 fullerenes and SWCNT.


METHODS
We measured the level of oxidative damage to DNA as the premutagenic 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodG) in the colon mucosa, liver, and lung of rats after intragastric administration of pristine C60 fullerenes or SWCNT (0.064 or 0.64 mg/kg body weight) suspended in saline solution or corn oil. We investigated the regulation of DNA repair systems toward 8-oxodG in liver and lung tissue.


RESULTS
Both doses of SWCNT increased the levels of 8-oxodG in liver and lung. Administration of C60 fullerenes increased the hepatic level of 8-oxodG, whereas only the high dose generated 8-oxodG in the lung. We detected no effects on 8-oxodG in colon mucosa. Suspension of particles in saline solution or corn oil yielded a similar extent of genotoxicity, whereas corn oil per se generated more genotoxicity than the particles. Although there was increased mRNA expression of 8-oxoguanine DNA glycosylase in the liver of C60 fullerene-treated rats, we found no significant increase in repair activity.


CONCLUSIONS
Oral exposure to low doses of C60 fullerenes and SWCNT is associated with elevated levels of 8-oxodG in the liver and lung, which is likely to be caused by a direct genotoxic ability rather than an inhibition of the DNA repair system.

Humans have been exposed to particulate matter during much of evolution, but tech nologic achievements such as combustion engines and nano technology have yielded unique types of particles. Engineered nano technologic materials are projected to be used, for instance, in electronics, cosmetics, clean ing materials, coatings, food packaging, and medicines, with increasing human exposure. In addition, consumer products containing nano materials will inevitably end up as waste, and subsequent processing or deposition may liberate particles to the environment, where they might accumulate along the food chain because some are highly persistent (Helland et al. 2007). Reinforced attention to the haz ardous properties of engineered nanoparticles has been evoked by a report showing that a single intra peritoneal application of carbon nano tubes generated mesotheliomas in p53 +/mice (Takagi et al. 2008). These data are supported by another recent report showing that a single injection of carbon nano tubes into the peritoneal cavity of mice elicited an inflammatory reaction and granulomas at the peritoneal surface of the diaphragm (Poland et al. 2008). Similarly, pulmonary exposure to singlewalled carbon nanotubes (SWCNT) produced granulomas in the lung of rodents (Lam et al. 2004;Shvedova et al. 2005; Warheit et al. 2004). It is possible that such nanotubes possess the same hazardous effects as other fibrous materials such as asbestos. On the other hand, particles such as C 60 fuller enes might be less hazardous . For example, in a recent inhalation experiment, Baker et al. (2008) found that exposure to C 60 fullerenes resulted in little pulmonary toxicity.
Generation of reactive oxygen spe cies (ROS) and oxidative stress is generally accepted to be an important mechanism of action of nano particles (Ayres et al. 2008;Nel et al. 2006;Oberdörster et al. 2005). Exposure to combustion particles in urban air or diesel exhaust particles (DEP) is associated with oxi dative stress, generated by particles themselves or by cellmediated inflammatory responses and increased formation of oxidatively dam aged DNA (Knaapen et al. 2004;Risom et al. 2005). 8Oxo7,8dihydro2´deoxyguanosine (8oxodG) is considered to be an important oxidatively generated DNA lesion in carcino genesis because a) it is mutagenic; b) mamma lian cells have a highly versatile repair system for its removal; and c) the level of 8oxodG is elevated in several types of tumor tissue (Evans et al. 2004). In addition, high urinary excretion of 8oxodG is associated with increased risk of lung cancer among non smokers . The relevance of 8oxodG is further strengthened by obser vations that this lesion is elevated in animal tissues and human blood cells upon exposure to urban air pollution, diesel exhaust, or DEP . In particular, a single oral dose of DEP was associated with increased lev els of 8oxodG in liver, lung, and colon tissue (Danielsen et al. 2008b). Effect modification by the DNA repair system is likely to occur by continuous exposure. For example, inges tion of DEP in the diet for 3 weeks led to an upregulation of the DNA repair system and unaltered levels of 8oxodG in the liver and colon mucosa cells (Dybdahl et al. 2003), whereas measurements of lung tissue indicated unaltered regulation of the DNA repair sys tem and elevated levels of oxidatively damaged DNA (Müller et al. 2004).
The aim of the present study was to inves tigate the effect of a single oral adminis tration of C 60 fullerenes and SWCNT. C 60 fullerenes consist of 60 carbon atoms arranged in an aro matic soccerball structure with a nano sized diameter. SWCNT consists of two dimen sions < 100 nm, whereas the axial dimen sion is much larger. Both C 60 fullerenes and SWCNT are hydro phobic and hence difficult to suspend in saline solution, but they suspend more easily in oils. Consequently, rats received the particles in either saline solution or corn oil by oral gavage. We then assessed the level of 8oxodG, a highly validated biomarker [European Standards Committee on Oxidative DNA Damage (ESCODD) et al. 2005]. We investigated alterations in the regu lation of DNA repair by mRNA levels of 8oxoguanine DNA glycosylase (OGG1), nei endonuclease VIIIlike 1 (E. coli) (NEIL1), mutY homolog (E. coli) (MUTYH), and nudix (nucleoside Background: C 60 fullerenes and single-walled carbon nanotubes (SWCNT) are projected to be used in medicine and consumer products with potential human exposure. The hazardous effects of these particles are expected to involve oxidative stress with generation of oxidatively damaged DNA that might be the initiating event in the development of cancer. oBjective: In this study we investigated the effect of a single oral administration of C 60 fullerenes and SWCNT. Methods: We measured the level of oxidative damage to DNA as the premutagenic 8-oxo-7,8dihydro-2´-deoxyguanosine (8-oxodG) in the colon mucosa, liver, and lung of rats after intragastric administration of pristine C 60 fullerenes or SWCNT (0.064 or 0.64 mg/kg body weight) suspended in saline solution or corn oil. We investigated the regulation of DNA repair systems toward 8-oxodG in liver and lung tissue. results: Both doses of SWCNT increased the levels of 8-oxodG in liver and lung. Administration of C 60 fullerenes increased the hepatic level of 8-oxodG, whereas only the high dose generated 8-oxodG in the lung. We detected no effects on 8-oxodG in colon mucosa. Suspension of particles in saline solution or corn oil yielded a similar extent of genotoxicity, whereas corn oil per se generated more genotoxicity than the particles. Although there was increased mRNA expression of 8-oxoguanine DNA glycosylase in the liver of C 60 fullerene-treated rats, we found no significant increase in repair activity. conclusions: Oral exposure to low doses of C 60 fullerenes and SWCNT is associated with elevated levels of 8-oxodG in the liver and lung, which is likely to be caused by a direct genotoxic ability rather than an inhibition of the DNA repair system. diphosphate linked moiety X)type motif 1 (NUDT1), which are involved in the removal of 8oxodG in DNA and the nucleotide pool and may be upregulated by oxidative stress (Hah et al. 2007;Risom et al. 2003aRisom et al. , 2007. We assessed the expression of heme oxyge nase 1 (HO1) mRNA as a general marker of oxidative stress elicited by exposure to particu late matter (Ayres et al. 2008).

Materials and Methods
Particle exposure of animals. We obtained 84 female Fisher 344 rats from Taconic (Ry, Denmark). Animals were acclimatized for at least 1 week before entering the experi ments. The rats were housed in a temperature controlled (22-24°C) and moisturecontrolled (40-70%) room with a 12hr light/12hr dark cycle. All animals had free access to tap water and Standard Altromin no. 1314 rat chow (Altromin, Lage, Germany) during the accli mation and housing/treatment periods. Rats were sacrificed at 9 weeks of age. Animals were treated humanely and with regard for alleviation of suffering. All animal proce dures followed the guidelines for the care and handling of laboratory animals established by the Danish government, and the Animal Experiment Inspectorate, Ministry of Justice, approved the study (no. 2006/5611161).
The dry powder of C 60 fullerenes was described by the manufacturer to be a 99.9% pure preparation with a primary par ticle size of 0.7 nm (SigmaAldrich, Brøndby, Denmark). The dry powder of SWCNT was described by the manufacturer to have a pri mary particle size of 0.9-1.7 nm and a fiber length < 1 µm (Thomas Swan and Co Ltd, Consett, UK). We suspended the particles in either saline or corn oil (SigmaAldrich) by sonication at 70 W and 42 kHz (Branson 1510, VWR-Bie & Berntsen A/S, Herlev, Denmark) in a 5day period for 10 hr each day and again 30 min before administra tion. For C 60 fullerene and SWCNT, the gas exchange surface areas were < 20 m 2 /g and 731 ± 2 m 2 /g and the average pore sizes were 0 and 15 nm respectively (Jacobsen et al. 2008b). We used dynamic light scattering, as described by Jacobsen et al. (2008b), to meas ure the particle size in suspensions at the same concentration as administered to the rats by oral gavage. In general, it was difficult to determine the presence of nanoparticles in both types of solutions by dynamic light scat tering because the solutions contain agglom erates with larger particle sizes. The particle sizes of C 60 fullerenes in the saline solution were 407 nm in the low dose and 621 and 5,117 nm in the high dose. In saline solution, the particle sizes of SWCNT were 195, 797, and 5,457 nm (low dose); the particle size in the highest dose could not be determined by dynamic light scattering. The particles were easier to suspend in corn oil than in saline. We obtained size modes of the C 60 fuller enes in corn oil by dynamic light scattering as follows: 234 nm (low dose); 40, 713, and 3,124 nm (high dose). The SWCNT had size modes as follows: 34 and 178 nm (low dose) and 1,015 nm (high dose). The SWCNT con tained transition metals (2% iron and traces of cobalt, nickel, and manganese) and poly cyclic aromatic hydrocarbons [PAHs; 417 ng/g of the U.S. Environmental Protection Agency (EPA) priority PAH compounds (U.S. EPA 1991)], whereas neither transition metals nor PAH could be detected in C 60 fullerenes.
The rats received a single intragastric dose of the particle preparations by oral gavage (0.064 and 0.64 mg/kg body weight sus pended in saline or corn oil; n = 8), saline solution (control; n = 10), or corn oil (con trol; n = 10). Each rat received 200 µL fluid. The rats were sacrificed by cervical dis location 24 hr after intragastric administration. The liver, lung, and colon tissues were snapfrozen in liquid nitrogen and stored at -80°C. We investigated effects in these organs because they are along the likely local and systemic exposure route and we have data from similar experiments on DEP as Standard Reference Material 2975 (SRM2975; National Institute of Standards and Technology, Gaithersburg, MD, USA), which was previously reported to be associated with increased levels of DNA damage 24 hr after a single intra gastric application (Danielsen et al. 2008b). For all experiments, rats from each group (0, 0.064, and 0.64 mg particles/kg body weight) were treated the same day.
ROS-generating ability. The ability of the C 60 fullerenes, SWCNT, SRM2975, and Printex 90 carbon black (Degussa Hüls, Frankfurt, Germany) to generate ROS in aqueous solution was determined by oxidation of 2´,7´dichloro dihydro fluorescin (Molecular Probes, Portland, OR, USA). The particle suspensions were prepared in Hank's balanced saline solution and sonicated immediately before incubation, as described previously by Jacobsen et al. (2008b). The oxidation product (2´,7´dichloro fluorescein) was determined by fluorescence spectrometry with excitation at 490 nm and emission at 520 nm on a fluorescence spectro photometer (Victor Wallac2 1420; PerkinElmer, Hvidovre, Denmark).
Oxidatively damaged DNA. We obtained suspensions of colon epithelial cells by scrap ing off the cells on the luminal side of the colon with a glass slide in icecold Merchant buffer, as described previously by Dybdahl et al. (2003). The DNA was extracted from colon mucosa cells, liver, and lung tissue according to the procedure recommended by ESCODD et al. (2005). We isolated nuclei in buffer containing deferoxamine mesylate (SigmaAldrich) to prevent spurious oxida tion; this was followed by lysis, RNase, and proteinase treatment (SigmaAldrich). We extracted DNA in buffer containing 40 mM Tris, 20 mM Na 2 EDTA, 7.6 M NaI, and 0.3 mM deferoxamine mesylate, pH 8.0 (SigmaAldrich), 2propanol, and ethanol. We digested DNA extracts to nucleosides using nuclease P1 and alkaline phosphatase (Merck, Darmstadt, Germany) and measured 8oxodG and dG by HPLC with electro chemical and ultra violet detection, respectively.
mRNA expression of HO1, MUTYH, NEIL1, NUDT1, and OGG1. We purified total RNA from liver and lung tissue using TRIzol Reagent (Invitrogen, San Diego, CA, USA) and DNase treatment according to the manufacturer's protocol (SV Total RNA isola tion kit; Promega, Madison, WI, USA). For cDNA synthesis, 100-200 ng RNA was used in a reaction volume of 20 µL using GeneAmp RNA PCR (polymerase chain reaction) Kit (Applied Biosystems, Naerum, Denmark) as recommended by the manufacturer.
We analyzed mRNA levels of the follow ing genes: HO1 [GeneID 24451 (National Center for Biotechnology Information 2008)], MUTYH (GeneID 170841), NEIL1 (GeneID 367090), NUDT1 (GeneID 117260), OGG1 (GeneID 81528) on a Taqman ABI7900 (Applied Biosystems) as described by Risom et al. (2003b). We used Taqman probes and primers MUTYH (Rn00591196), NEIL1 (Rn01422330), NUDT1 (Rn00589097), and Taqman 18S probe (Euk 18S rRNA FAM/ MGB probemix) from Applied Biosystems. All probes and primers span exon junctions and are cDNA specific. We used oligo nucleo tides of HO1 and OGG1 as follows (final concentra tions are shown in parentheses): • RnHO1 forward primer (900 nM), 10F: 5´CCA CAG CTC GAC AGC ATG T3´; For the PCR reactions, we mixed 4 µL of the cDNA preparation with master mix and water to a final volume of 228 µL. Aliquots were mixed with probes and primers to a final volume of 36 µL; and then transferred to three wells at 10 µL/well. For the PCR reaction, we used the following protocol: activation of the polymerase, 95°C for 20 sec; followed by 45 cycles of 95°C for 1 sec; and finally 60°C for 20 sec. We meas ured the expressions of refer ence (18S) and target genes in separate tubes. The expression levels are reported as target mRNA normalized to 18S, calculated by the comparative cycle threshold (C t ) method 2 −Δ Ct . The SD of the triplicates was < 0.5 (based on C t values). Each run included a standard (for verification of the PCR efficiency) and negative (no template) and positive controls.
OGG1 repair activity. We obtained extracts of liver tissue by using a plunger to force the tissue through a sieve in one end of a stainless steel cylinder (0.5 cm in diameter, mesh size 0.4 mm) while the cylinder was submerged in 2 mL icecold buffer (45 mM HEPES, 0.4 M KCl, 1 mM EDTA, 0.1 mM DTT, 10% glycerol, pH 7.8), as described previously (Folkmann et al. 2007). The protein concen trations of the liver tissue extracts were meas ured using the Coomassie Plus -The Better Bradford Assay Reagent as recommended by the manufacturer (Pierce, Rockford, IL, USA) using bovine serum albumin as standard. We diluted liver tissue extracts to a final protein concentration of 3 mg/mL.
We analyzed OGG1 repair activity as the number of incisions in substrate nuclei (Guarnieri et al. 2008). We prepared substrate nuclei containing 8oxodG by treating THP1 human monocyte cells (American Type Culture Collection, Manassas, VA, USA) with 1 µM Ro198022 in phosphatebuffered saline and irradiating them with white light for 4 min. The Ro198022 photo sensitizer was a gift from F. HoffmannLa Roche Ltd (Basel, Switzerland). Substrate nuclei were embedded in lowmeltingpoint gel (0.75%; SigmaAldrich), applied onto GelBond films (Cambrex; Medinova Scientific A/S, Hellerup, Denmark), and lysed in icecold lysis solution (2.5 M NaCl, 0.1 mM Na 2 EDTA, 10 mM Tris, 1% Triton X100, pH 10) for 1 hr at 4°C. We applied liver tissue extracts onto the gels and incubated them for 20 min at 37°C in a humid box. Negative controls (buffer) and positive controls (formamido pyrimidine DNA glycosylase; gift from A. Collins, University of Oslo, Oslo, Norway) were incu bated for 45 min.
The GelBonds were subsequently treated in icecold alkaline solution (300 mM NaOH, 1 mM Na 2 EDTA, pH > 13.0) for 40 min and electrophoresed 20 min in the same solu tion at 25 V (0.83 V/cm) and 300 mA. The nuclei were visualized using fluorescence microscope (Olympus, Ballerup, Denmark) at 40× magnification after staining with the YOYO1 fluorescent dye (Molecular Probes). We scored the level of DNA damage accord ing to five classes of damage of 100 randomly selected nuclei from each sample; we exam ined two slides for each sample. The samples were coded before scoring. We determined the repair activity of the tissue extracts as the difference in score (arbitrary units) between parallel gels incubated with extract and control solution, respectively.
Statistics. We analyzed the results sta tistically by twofactor analysis of variance (ANOVA) using the dose of particles and vehicle as categorical variables. We found no interactions between the categorical factors; thus, the pvalues correspond to singlefactor effects. We tested the homogeneity of vari ance by Levene's test. To achieve homoge neity of variance, we log 10 transformed all results, except those for 8oxodG in the lung.
We found no interactions between type of vehicle and particle exposure, indicating that particles in corn oil generated the same level of DNA damage as particles in saline. The gen eration of 8oxodG is based on data from both types of vehicles (corresponding to estimates of singlefactor effects). In the liver, the expo sure to C 60 fullerenes was associated with 17% (95% CI, 4-34) and 25% (95% CI, 11-41) increases of 8oxodG in groups treated with the low dose and high dose of particles, respec tively. Exposure to SWCNT increased the level of 8oxodG in the liver by 22% (95% CI, 8-38) and 20% (95% CI, 7-36) in the rats given the low dose and high dose, respec tively. There were increased levels of 8oxodG in the lung after exposure to SWCNT at both the low [21% (95% CI, 9-33)] and high doses    23% (95% CI,], whereas only the high dose of C 60 fullerenes was associated with significantly elevated levels of 8oxodG in the lung [18% (95% CI, 4-31)]. Oral exposure to particles was not associated with increased levels of 8oxodG in colon mucosa tissue. mRNA expression of HO1, MUTYH, NEIL1, NUDT1, and OGG1. Table 1 out lines the gene expression of OGG1, HO1, NEIL1, MUTYH, and NUDT1 mRNA in liver and lung tissue. The type of vehicle did not affect the level of gene transcription (p > 0.05, singlefactor effect in ANOVA). The low and high doses of C 60 fullerenes increased the gene expression of OGG1 by 1.30fold (95% CI, 0.9-1.9) and 1.80fold (95% CI, 1.2-2.6), respectively, in the liver, whereas a similar effect of SWCNT did not reach sta tistical significance. In contrast, we found no significant effects in the expression of HO1, NEIL1, MUTYH, and NUDT1 mRNA in liver and lung tissue. Colon tissue was not analyzed because there was unaltered level of 8oxodG and limited amounts of material were available.
OGG1 repair activity. Based on the obser vation that the gene expression of OGG1 was increased in the liver after oral exposure to particles, we also measured the OGG1 repair activity of liver extracts (Figure 3). Neither the exposure to particles nor the vehicle were asso ciated with a significantly altered level of OGG1 repair activity, although a doserelated effect of C 60 fullerenes is suggested (p = 0.20, ANOVA).

Discussion
In this study we found increased levels of oxidatively damaged DNA in liver and lung tissue 24 hr after oral administration of C 60 fullerenes and SWCNT in saline or corn oil. There were virtually no alterations of the repair activity of oxidized base lesions in the same tissues, indicating that the level of DNA damage is not underestimated as a conse quence of increased repair.
The levels of 8oxodG were approxi mately 20% increased in the liver and lung tissue after the exposure to C 60 fullerenes and SWCNT, whereas the vehicle did not affect the particlegenerated genotoxicity. The effect of C 60 fullerenes on OGG1 mRNA expression in the liver and 8oxodG in the lung showed clear dose-response relationships, whereas other dose-response relationships were more flat. Possible agglomeration of particles affect ing uptake and effects might have influenced the dose relationships. This could also be rele vant for the lack of genotoxicity in colon mucosa cells. These cells have high turnover, which could dilute any possibly transient DNA damage, whereas liver and lung cells have a low proliferation rate, allowing dam age to accumulate. We have previously shown that rats given the same oral dose of DEP as SRM2975 had approximately 50% elevated level of 8oxodG in colon mucosa cells, liver, and lung (Danielsen et al. 2008b). Thus, our results indicate that C 60 fullerenes and SWCNT are genotoxic in rats, but the effect is lower than that observed for SRM2975. The C 60 fullerenes and SWCNT were adminis tered in rather low doses because we wanted to use the SRM2975 preparation as benchmark particles. In comparison, a recent study with intravenous administration of 150 mg pegy lated SWCNT per mouse argued that concen trations between 10 and 100fold lower were sufficient in biomedical applications (Schipper et al. 2008). Other recent studies on pulmo nary toxicity of SWCNT have used doses in the range of 20-40 µg/mouse (corresponding to about 1-2 mg/kg body weight) by pharyn geal aspiration (Han et al. 2008;Shvedova et al. 2008). The doses we used in the present study can thus be considered as being in the lower range of exposures.
To the best of our knowledge, our data are the first to demonstrate that C 60 fullerenes and SWCNT generate oxidatively damaged DNA in rodent organs. Other reports have shown that C 60 fullerenes and SWCNT exposure of cells in culture is associated with elevated levels of DNA damage measured by the comet assay (Dhawan et al. 2006;Kisin et al. 2007;Pacurari et al. 2008). These types of lesions represent general geno toxicity rather than oxi datively damaged DNA, which can be assessed by a modified version of the comet assay including digestion with DNA glycosylase or endonuclease enzymes (Møller 2006). We used this approach in a cell culture experiment and demonstrated that both C 60 fullerenes and SWCNT generated oxidatively damaged DNA (Jacobsen et al. 2008b). The oxidiz ing effect of C 60 fullerenes on bio molecules in cell culture settings could be due to gen eration of singlet oxygen induced by photo sensitization (Kamat et al. 2000;Yamakoshi et al. 2003). Oxidations of DNA in internal organs probably arise from mixed ROS gen eration, as supported by increased mortality after C 60 fullerene exposure in zebrafish in the dark and by enhanced mortality resulting   The mRNA expression is relative to the expression of 18S per 10 6 . Values are mean ± SE. *p < 0.05 (single-factor effect of the particle exposure, ANOVA).
from coexposure of C 60 fullerenes with hydro gen peroxide or glutathione depletion (Usenko et al. 2008). Hydroxyl radicals appear to be the important type of ROS generated by SWCNT in cell cultures (Manna et al. 2005;Pacurari et al. 2008). The level of transition metals is not likely to explain the differences in geno toxicity between the particles used in the pres ent study and SRM2975, because the latter has very low levels, whereas the SWCNT used here contained 2% iron, as well as traces of other metals, and the C 60 fullerenes had unde tectable levels. Moreover, SWCNT at low concentrations produced higher levels of ROS in cellfree systems than did SRM2975 and C 60 fullerenes, whereas SRM2975 appeared to induce more guanine oxidation than the engineered particles in cell culture (Danielsen et al. 2008a;Jacobsen et al. 2008aJacobsen et al. , 2008b. The SRM2975 preparation contains substan tial amounts of PAHs, which show low lev els in SWCNT and are undetectable in C 60 fullerenes. The content of PAHs could explain the higher DEPinduced level of 8oxodG possibly through generation of oxidative stress by the aldoketoreductase metabolic pathway, as recently shown in cell culture (Park et al. 2008). Nevertheless, particles such as carbon black with very low PAH and metal levels induced the highest levels of ROS in both cell free systems and in cell culture with resulting DNA oxidation and mutagenicity (Jacobsen et al. 2007), indicating that particles as such have the ability to induce oxidative stress. Accordingly, ROS generation in an acellular in vitro system does not appear to predict the ability of particles to oxidatively damage DNA in cell culture or in vivo.
The notion that C 60 fullerenes and SWCNT induce less oxidative stress com pared with DEP is supported by the unal tered regulation of HO1, which in our previous study was upregulated by exposure to SRM2975 (Danielsen et al. 2008b). In the present experiments, we observed unaltered expression levels of NEIL1, MUTYH, and NUDT1, whereas the increased expression of OGG1 mRNA in the liver of rats exposed to C 60 fullerenes was not associated with higher repair activity, which might be due to the fact that longer exposure times are required to observe increased repair activity. Indirect evidence of this effect modification comes from studies of inhalation exposure to DEP where repeated exposures on 4 consecutive days were associated with upregulation of OGG1 and unaltered 8oxodG in pulmonary tissue (Risom et al. 2003a), whereas an identi cal exposure scenario yielded higher levels of 8oxodG in OGG1 knockout mice (Risom et al. 2007). Although increased expression levels of OGG1 might be considered benefi cial to the cells, it should be emphasized that in human beings, with the variant form of OGG1 (Ser326Cys) genotype, high expres sion levels of OGG1 mRNA in leukocytes is associated with increased risk of lung cancer (Hatt et al. 2008).
Although our results strongly indicate that oral exposure to C 60 fullerenes and SWCNT is associated with increased generation of oxi dized DNA, it is still unresolved how appli cation of particulate matter in the gut can oxidize bio molecules in internal organs. A somewhat naive notion suggests that the epi thelial lining of the gut is designed to absorb substances and thus may allow passage of particulate matter. In fact, we suspended the particles in either corn oil or saline solution because of the notion that their hydropho bic nature would let them follow the regular passage of lipids in the gut. Because there was no difference in genotoxicity between the particles suspended in corn oil and saline solution, we believe that orally adminis tered particles will distribute in the chyle of the intestinal juice, despite the fact that they may reach the gastro intestinal tract as agglomerates, as indicated by aqueous par ticle suspension characterization. The results suggest that the particles are absorbed from the gastro intestinal tract to blood circulation and secondary organs. However, at pres ent the extent of trans location of particulate matter across epithelial barriers is a highly controversial issue. Studies of the pulmonary translocation of model particles to systemic circulation and secondary organs indicate only minute passage (Kreyling et al. 2002;Mills et al. 2006;Möller et al. 2008;Wiebert et al. 2006). On the other hand, wholebody inhalation exposure to ultrafine carbon par ticles have suggested some deposition in the liver, which has been speculated to originate from gastro intestinal exposure and uptake from the gut (Oberdörster et al. 2002). In addition, the uptake of C 60 fullerenes and polystyrene latex microspheres from the gastro intestinal tract into blood circulation has been estimated in the range of 1% of the applied dose (Carr et al. 1996;Yamago et al. 1995). However, it is interesting to note that in drug delivery research, there is acceptance that particulate matter can be absorbed from the gastro intestinal tract, and this feature of nano materials is being used as an approach of altering the pharmaco kinetic behavior of drugs (Florence 2004).
We observed an effect of corn oil in all three organs. This was not part of our a priori hypothesis. We chose the volume of corn oil to be identical to that of the saline solution (200 µL), which would allow a reasonable volume for the suspension in the aqueous solution. It should be emphasized that corn oil is rich in polyunsaturated fatty acids and was sonicated, which might produce geno toxic compounds. In addition, rat studies indicated that a diet rich in corn oil increased 8oxodG excretion in urine (Loft et al. 1998) as well as the levels of 5hydroxymethyl2´ deoxyuridine, another marker of oxidatively damaged DNA in blood and mammary gland tissue (Djuric et al. 2001). A dyslipidemic state, as found in apoE knockout mice, also appears to include agedependent accumu lated oxidized DNA base lesions in the liver (Folkmann et al. 2007).
In conclusion, the data obtained from the present study indicate that C 60 fullerenes and SWCNT generated oxidatively damaged DNA in liver and lung cells by a gastro intestinal route. The genotoxic effect resulting from exposure to C 60 fullerenes and SWCNT was smaller than that from DEP exposure; how ever, it may be a cause for concern in humans.