- Email: [email protected]
Oxidative Damage to Nucleic Acids Photosensitized by Titanium Dioxide
Free Radical Biology & Medicine, Vol. 23, No. 6, p. 851–858, 1997 Published by Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/97 $0.00 / .00
PII S0891-5849(97)00068-3
Original Contribution OXIDATIVE DAMAGE TO NUCLEIC ACIDS PHOTOSENSITIZED BY TITANIUM DIOXIDE
WAYNE G. WAMER, JUN-JIE YIN, and RONG RONG WEI Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, 200 C St., SW, Washington, DC 20204, USA (Received 3 May 1996; Revised 17 March 1997; Accepted 17 March 1997)
Abstract—The semiconductor TiO2 is known to have photobiological activity in prokaryotic and eukaryotic cells. Applications of this photobiological activity have been suggested including sterilization of waste water and phototherapy of malignant cells. Here, several model and cellular systems were used to study the mechanism of photocatalysis by TiO2 . Treatment of TiO2 (anatase, 0.45 mm), suspended in water containing a spin trap 5,5-dimethyl1-pyrroline-N-oxide (DMPO), with UV radiation (320 nm) resulted in an electron spin resonance (ESR) signal characteristic of the hydroxyl radical. Irradiation of solutions containing calf thymus DNA and TiO2 with UVA (320–400 nm) radiation resulted in hydroxylation of guanine bases. The degree of hydroxylation was dependent on both UVA fluence and amount of TiO2 in suspension. Human skin fibroblasts, preincubated 18 h with 10 mg/cm2 TiO2 and then UVA-irradiated (0–58 KJ/m 2 ), showed dose dependent photocytoxicity. RNA, isolated from similarly treated fibroblasts, contained significant levels of photooxidation, measured as hydroxylation of guanine bases. However, no oxidative damage was detectable in cellular DNA. These results suggest that nucleic acids are a potential target for photooxidative damage sensitized by TiO2 , and support the view that TiO2 photocatalyzes free radical formation. Published by Elsevier Science Inc. Keywords—DNA damage, Free radicals, 8-oxodG, 8-Oxo-7,8-dihydro-2 *-deoxyguanosine, Oxidative damage, Photooxidation, Phototoxicity, RNA damage, Titanium dioxide
violet radiation having an energy greater than the optical band gap of TiO2 (3.2 eV for the anatase polymorph of TiO2 ) is absorbed and results in the formation of an electron-hole pair. The electron, generated following absorption of a UV photon, is a reducing agent whereas the concomitantly formed hole is a powerful oxidizing agent. It has been shown that photoexcited TiO2 is a catalyst for a variety of redox reactions including decarboxylation of carboxylic acids 2 and hydrogen production from carbohydrates 3 and water.4 In addition, it has been reported that TiO2 is photobiologically active. When exposed to UVA (320–400 nm) radiation, TiO2 exhibits antibacterial activity.5,6 Recently, it has been proposed that TiO2 may be useful for photodynamic therapy of cancer.7 The mechanism(s) underlying the photobiological activity of TiO2 is not yet well understood. Although evidence substantiates in vitro photooxidative damage sensitized by TiO2 , the intracellular target(s) have not
INTRODUCTION
The utility of TiO2 in a wide range of consumer products derives from its photophysical properties. TiO2 does not absorb radiation in the visible region of the spectrum.1 Therefore, the photophysical properties of TiO2 in the visible region of the spectrum are predominately due to light scattering. This leads to its use as a white pigment and masking or opacifying agent. Ultraviolet radiation, however, is both scattered and absorbed by TiO2 .1 Because of the effective attenuation of UV radiation by TiO2 , TiO2 has been used as a physical sunscreen. Absorption of UV radiation by TiO2 results from the electronic band structure of this semiconductor. UltraAddress correspondence to: Wayne G. Wamer, Food and Drug Administration, Cosmetics Toxicology Branch (HFS-128), 200 C Street, S.W., Washington, DC 20204; E-Mail: [email protected] 851
/ 2b2f 5042 Mp 851 Monday Aug 11 11:32 AM EL–FRB 5042
852
W. G. WAMER et al.
been determined. Investigators have shown that the bactericidal activity photocatalyzed by TiO2 is accompanied by cell membrane damage observed as loss of potassium ions, proteins and RNA from bacterial cells.8 Recently, Sakai et al. have reported that sublethal treatment of T24 cells with TiO2 and UVA radiation resulted in a significant increase in intracellular calcium.9 Following lethal exposure to TiO2 and UVA radiation, a further increase in intracellular calcium was observed. These authors concluded that cellular membrane damage, resulting in greater permeability to calcium, may play a role in photocytotoxicity sensitized by TiO2 . Therefore, the cellular membrane may be a significant target for TiO2-photosensitized damage in both prokaryotic and eukaryotic cells. Here, we examine whether nucleic acids are potential targets for photooxidative damage sensitized by TiO2 . Investigators have demonstrated that measurement of hydroxylation of guanine bases, using HPLC with electrochemical detection, 10 provides a sensitive indicator of oxidative damage to nucleic acids resulting from normal physiological processes such as aerobic metabolism11 and aging 12 as well as exposure to exogenous oxidants.13 In addition, hydroxylation of guanine, leading to formation of 8oxo-7,8-dihydro-2 *-deoxyguanosine ( 8-oxodG ) in DNA and 8-oxo-7,8-dihydroguanosine ( 8-oxoG ) in RNA, is initiated by an array of reactive oxygen species including 1O2 , iOH and O2 i0 .13 Therefore, hydroxylation of guanine can serve as a sensitive biomarker for oxidative stress induced through diverse mechanisms.
ESR spectra were recorded, stored and manipulated using an IBM / PC computer. Photooxidation of calf thymus DNA Calf thymus DNA (100 mg/ml, Sigma Chemical Co., St. Louis, MO) was prepared in 10 mM phosphate buffer, pH 7.2. A suspension of TiO2 (anatase, 0.45 mm, Aldrich Chemical Co., Milwaukee, WI) in distilled water was then added to yield the selected concentration of TiO2 . Samples were then irradiated in open Petri dishes by using a UVA (320–400 nm) light source consisting of two GE F40BL bulbs filtered through 3 mm of soft glass. The spectral irradiance of this source, monitored with an IL1700 Research Radiometer fitted with a SED033 probe (International Light, Inc., Newburyport, MA), was determined to be 1.62 1 10 02 KW/m 2 . After irradiation, samples were centrifuged to remove TiO2 . DNA was precipitated from the supernatant by addition of 12 volume of 7.5 M sodium acetate and 2 volumes of ethanol. DNA pellets were collected by centrifugation, washed with ethanol and dried under reduced pressure. DNA pellets were then reconstituted in 100 ml of 10 mM tri(hydroxymethyl)-aminomethane hydrochloride (Tris-HCl), 10 mM MgCl2 and 1 mM ZnCl2 , pH 7.4. DNA was enzymatically hydrolyzed by addition of 240 Dornase units of DNase 1 and 3.2 units of calf intestinal alkaline phosphatase (Gibco BRL, Grand Island, NY) and incubation at 377C overnight. Hydrolysates were then centrifuged and analyzed by HPLC as described below. Cell culture
MATERIALS AND METHODS
ESR spectroscopy DMPO was purchased from Sigma Chemical Co. ( St. Louis, MO ) and was purified with activated charcoal prior to use. Deionized water was used to prepare all solutions and suspensions. ESR measurements were carried out at room temperature ( 227C ) by using a Varian E-109 spectrometer operating at 9.5 GHz with 100 KHz and 1G field modulation. Spectra were obtained with 15 mW microwave power. Magnetic field and microwave frequency measurements were made with a DTM-141 Digital Teslameter ( GMW Magnetic System, Redwood City, CA ) and EIP-25B Frequency Counter ( EIP Microwave Inc., Milpitas, CA ) . Samples were irradiated in the microwave cavity by using a 1 Kwatt xenon lamp ( Schoeffel, Kratos Inc., Westwood, NJ ) with a single monochromator ( GM 252 ) set at 320 nm. All
A human skin fibroblast cell line (ATCC CRL 1634) was obtained from American Type Culture Collection, Rockville, MD. Cells were cultured in Dulbecco’s modified Eagle medium containing 10% fetal bovine serum (Gibco BRL, Grand Island, NY), antibiotics (penicillin at 100 U/ml and streptomycin at 100 mg/ml), 4.5 mg/ml glucose and 4 mM L-glutamine. Cultures were incubated at 377C in a humidified atmosphere containing 5% CO2 . Cell survival A suspension of TiO2 in water was sterilized by heating at 1007C for 30 min. Confluent monolayers of human skin fibroblasts in 60 mm dishes were treated for 18 h with 4 ml of medium containing 47.2 mg/ml TiO2 in suspension. Measurement of the change in the absorbance of medium at 320 nm showed that more than 99% of the particulate TiO2 had settled from the medium after 18 h. This observation was concomitant
/ 2b2f 5042 Mp 852 Monday Aug 11 11:32 AM EL–FRB 5042
Titanium dioxide sensitized oxidation
with the appearance of a fine precipitate on the fibroblast monolayer. This treatment results in exposure of the fibroblast monolayer (18.9 cm2 ) to 10 mg/cm2 TiO2 . Treated monolayers were washed twice in phosphate buffered saline (PBS), covered with PBS and irradiated at 257C with UVA. The maximum exposure to UVA (58 KJ/m 2 ) required irradiation for 1 h. Cells were then removed from the substratum by trypsinization. Cells (500–1000) in 4 ml of media were then seeded into 60 mm plastic dishes. After incubation for 7–10 days at 377C, cells were fixed with methanol, stained with crystal violet, and cell colonies were counted. While it is known that TiO2 is insoluble in aqueous solution and not bioavailable when given orally or intravenously, 14 no information is available about the solubility of TiO2 in cell culture medium. Therefore, an experiment was performed to determine whether TiO2 or impurities, solubilized by cell culture medium, contribute to the photocytotoxicity sensitized by TiO2 . A suspension containing 47.2 mg/ml TiO2 in medium was prepared as described above. This suspension was then incubated with gentle mixing for 18 h at 377C. Particulate TiO2 was removed by centrifuging twice for 5 min at 5,000 1 g. The supernatant was used to treat fibroblast monolayers. Photocytotoxicity was then assessed (vide ante).
853
layers of skin fibroblasts in 150 mm Petri dishes (approximately 1 1 10 7 cells) were lysed in situ by addition of 15 ml of a solution containing 6 M guanidinium hydrochloride, 0.5 M sodium acetate, 1.3% N-lauroylsarcosine and 1.5 mg proteinase K. The lysate was transferred to a tightly capped tube and incubated for 1 h at 407C. The lysate was then allowed to cool to room temperature and was layered underneath 30 ml ethanol in a 50 ml tube. The tube was tightly capped and slowly turned on its side to avoid mixing of the lysate and ethanol layers. Tubes were then rolled until a white thread appeared (2–4 min). This thread, containing both cellular RNA and DNA, was then washed with 80% ethanol/water, transferred to a 1.5 ml Eppendorf tube and briefly dried under vacuum. Dried pellets were then resuspended in 100 ml 10 mM Tris-HCl, pH 7.4, 10 mM MgCl2 , 1 mM ZnCl2 , 240 units of DNase 1 and 20 units of nuclease P1. Hydrolysis was allowed to proceed 2 h at 377C. Samples were then briefly centrifuged to remove any TiO2 isolated with RNA and DNA. Following centrifugation, 3.2 units of calf intestinal alkaline phosphatase were added to the supernatant. Samples were then allowed to hydrolyze overnight at 377C. Hydrolysates were centrifuged before analysis by HPLC.
HPLC analysis Treatment of cells with TiO2 and UVA for assessment of photooxidative damage to RNA and DNA Human skin fibroblasts were grown to confluency in 150 mm dishes that have a plating area of 143 cm2 . Fibroblast monolayers were treated with 20 ml media supplemented with a sterilized suspension of 71.4 mg/ ml TiO2 for 18 h. This treatment results in exposure of the fibroblast monolayer to 10 mg/cm2 TiO2 . Monolayers were then washed twice with PBS and exposed to UVA radiation as described above. Additionally, skin fibroblasts were treated with medium which had been preincubated for 18 h with 71.4 mg/ml TiO2 . Particulate TiO2 was removed by centrifugation prior to treating fibroblast monolayers. After 18 h treatment, fibroblast monolayers were exposed to UVA radiation. Isolation and enzymatic hydrolysis of cellular RNA and DNA All steps following irradiation of fibroblasts were performed in a darkened room. Cellular RNA and DNA were isolated using a modification of the method described by Xu et al.15 Immediately following treatment with photosensitizer and irradiation, confluent mono-
Nucleoside mixtures, resulting from enzymatic hydrolysis of RNA and DNA, were analyzed by reversedphase HPLC. A 100 ml aliquot was injected onto an Adsorbosphere HS C18 (4.6 mm 1 15 cm, 3 mm) column (Alltech Associates, Deerfield, IL) eluted with 7% methanol in 0.1 M phosphate buffer, pH 4.6 at 0.7 ml/ min. Unmodified nucleosides were monitored at 254 nm by using a Waters Model 400 detector (Millipore Corp., Milford, MA). 8-OxodG and 8-oxoG were quantitated using a Dionex pulsed amperometric detector (Sunnyvale, CA) connected immediately downstream from the UV detector. The glassy carbon working electrode of the electrochemical detector was set at 400 mV relative to the Ag reference electrode. Standards of 8-oxodG and 8-oxoG were prepared as previously described.16 Results are expressed as the fraction of hydroxylated guanine bases (i.e., 8-oxodG/10 5 dG for DNA, 8-oxoG/10 5 G for RNA). The limits of detection were determined to be 0.95 8-oxodG/10 5 dG, and 0.36 8-oxoG/10 5 G in hydrolysates of cellular DNA and RNA. We have previously demonstrated that the sensitivity of this assay is sufficient to detect oxidative damage to cellular RNA and DNA produced by UVA radiation 17 as well as photodynamic sensitizers and visible light.18,19
/ 2b2f 5042 Mp 853 Monday Aug 11 11:32 AM EL–FRB 5042
854
W. G. WAMER et al.
Fig. 1. ESR spectra recorded at 227C after 3 min irradiation with UV radiation of 320 nm. The samples, containing 0.05 M DMPO, received no treatment (A), 3 min irradiation at 320 nm (B), 100 mg/ ml TiO2 (C) or 100 mg/ml followed by 3 min irradiation at 320 nm (D) (aN Å aH Å 14.9 G).
RESULTS
ESR spectra of DMPO spin adducts formed in a suspension of TiO2 irradiated with UVA Figure 1 illustrates the ESR spectrum of the spin adduct formed when a suspension of TiO2 is irradiated at 320 nm. The appearance of the ESR signal was dependent on the presence of both TiO2 and UVA radiation. The ESR spectrum observed consists of the 1:2:2:1 quartet pattern (aN Å aH Å 14.9 G) character-
Fig. 3. Hydroxylation of guanine in calf thymus DNA photocatalyzed by TiO2 . Levels of hydroxylation were dependent both on the fluence of UVA (3A) and concentration of TiO2 in suspension (3B). Data points are the means { standard deviation of three determinations.
istic of the DMPO-hydroxyl radical adduct.20 The time course for development of the ESR signal intensity of the DMPO-hydroxyl radical adduct is shown in Fig. 2. Dose-dependent formation of 8-oxodG in calf thymus DNA Treatment of calf thymus DNA with TiO2 and UVA resulted in significant formation of 8-oxodG. Levels of 8-oxodG were dependent on both the fluence of UVA (Fig. 3A) and amount of TiO2 present (Fig. 3B). Photocytotoxicity of TiO2 Fig. 2. Time dependence of ESR signal intensity of DMPO-OH adduct obtained during UV irradiation at 227C. Sample contained 0.05 M DMPO and 400 mg/ml TiO2 .
It has been reported that human carcinoma cells, incubated with TiO2 , incorporate TiO2 into the outer plasma membrane and cytoplasm.9 We observed that
/ 2b2f 5042 Mp 854 Monday Aug 11 11:32 AM EL–FRB 5042
Titanium dioxide sensitized oxidation
855
Fig. 4. Photograph at 4001 magnification of human skin fibroblasts after incubation for 18 h with 10 mg/cm2 TiO2 . TiO2 appears as bright, highly refractile, particles in the cellular cytoplasm.
skin fibroblasts, cultured 18 h with TiO2 , similarly incorporated TiO2 (Fig. 4). Titanium dioxide particles were incorporated into the cytoplasm and frequently had a perinuclear distribution (Fig. 4). Titanium dioxide particles were not observed in cell nuclei. No cytotoxicity resulted from treatment with TiO2 or UVA irradiation alone (Fig. 5). In contrast, treatment of skin fibroblasts with TiO2 followed by irradiation with UVA resulted in significant, UVA fluencedependent, cytotoxicity (Fig. 5). Also, we determined that solubilization of TiO2 by cell culture medium does not account for the photocytotoxicity sensitized by TiO2 . No photocytotoxicity was observed following 18 h treatment with medium preincubated with TiO2 (Fig. 5). This result demonstrates that particulate TiO2 is required for photocytotoxicity.
observed in cellular DNA following treatment with TiO2 and UVA (Fig. 6B). Similarly, levels of 8-oxodG in DNA isolated from cells receiving TiO2 or UVA alone were not significantly elevated (Fig. 6B). Also, no photooxidative damage in cellular RNA or DNA
Photooxidation of cellular RNA and DNA The methods used here allow measurement of oxidative damage to both cellular RNA and DNA derived from a single cell monolayer. Significant levels of oxidative damage in cellular RNA were found following exposure of skin fibroblast to TiO2 and UVA radiation. Treatment with TiO2 and UVA, resulting in 85% cytotoxicity, produced a 3-fold increase in 8-oxoG (Fig. 6A). Exposure to UVA or TiO2 alone caused no significant increase in levels of 8-oxoG (Fig. 6A). No significant hydroxylation of 2 *-deoxyguanosine was
Fig. 5. Photocytotoxicity of TiO2 . UVA controls ( h ) were irradiated at 257C with the indicated fluences of UVA. Dark controls ( s ) were treated 18 hr with 10 mg/cm2 TiO2 , covered with PBS and incubated in the dark at 257C for a time equivalent to samples receiving UVA. Controls to assess the photocytotoxicity of any medium-soluble components of TiO2 ( , ) were treated with medium, which was preincubated with TiO2 as described in Materials and Methods, and then irradiated with UVA. Samples received 10 mg/cm2 particulate TiO2 and the indicated fluence of UVA ( j ). Data points are the means { the standard deviation of six determinations.
/ 2b2f 5042 Mp 855 Monday Aug 11 11:32 AM EL–FRB 5042
856
W. G. WAMER et al.
90% of these electron/hole pairs recombine within 10 nanoseconds.21 Within approximately 250 nanoseconds after formation, the remaining photogenerated electrons and holes are trapped, predominately at surface sites of particulate TiO2 .21 Evidence suggests that photogenerated electrons are trapped at Ti /4 sites, at or near the surface of TiO2 particles.21,22 Photogenerated holes (h / ) are thought to be trapped as iOH, adsorbed to the surface of TiO2 particles: 23 h / / H2O r iOH / H /
Fig. 6. Photooxidation of cellular RNA (6A) and DNA (6B) sensitized by TiO2 . Data points are the means { the standard deviations of at least four determinations. Confluent monolayers of skin fibroblasts were treated at 257C with UVA ( h ), 10 mg/cm2 TiO2 ( s ), or both UVA and 10 mg/cm2 TiO2 ( j ).
resulted from irradiating fibroblasts that had been treated with medium preincubated with TiO2 (data not shown). This result demonstrates that particulate TiO2 is required for photosensitized damage to cellular nucleic acids. DISCUSSION
Much is known about the photophysical events which follow absorption of UV radiation by TiO2 . Absorption of UV radiation having an energy greater than the bandgap of TiO2 results in the formation of paired electrons and holes which reside, respectively, in the conduction and valence bands of TiO2 . Approximately
A number of reactive oxygen species may be subsequently generated in aqueous solutions. Using spin trapping and ESR detection, Jaeger and Bard have observed ESR spectra consistent with formation of iOH and O2 i0 following absorption of UV radiation by TiO2 (anatase).24 Generation of iOH was attributed to oxidative decomposition of water resulting from the hole produced in photoexcited TiO2 . Jaeger and Bard suggested that O2 i0 may be formed by reduction of O2 by photogenerated electrons in UV irradiated TiO2 particles. In our hands, spin adducts formed during UV irradiation of TiO2 with the spin trap yielded an ESR signal characteristic of DMPO-OH spin adduct alone. The difference between results reported by Jaeger and Bard and in this study may be due to differences in spin traps, pH and composition of solvent.25 The results of our spin trapping studies do not exclude the possibility that O2 i0 , formed following UV irradiation TiO2 , is subsequently interconverted to radicals which give rise to the DMPO-OH spin adduct. Additional studies are necessary to fully understand the mechanism of radical formation by photoexcited TiO2 . In addition, it has been shown that H2O2 can be formed on TiO2 electrodes under aerobic conditions.26 In more complex cellular systems, where enzymatic and metal-catalyzed interconversion of radicals is expected, the relative importance of the reactive oxygen species formed by photoexcited TiO2 is still unknown. Evidence has recently appeared supporting the involvement of both O2 i0 and H2O2 in killing of cancer cells photosensitized by TiO2 . In support for the involvement of O2 i0 , Cai et al. have reported that addition of superoxide dismutase, which converts the relatively unreactive O2 i0 to H2O2 , greatly increases toxicity to HeLa cells photosensitized by TiO2 .27 It has also been demonstrated that addition of catalase, a scavenger of H2O2 , to HeLa cells prior to photosensitization by TiO2 significantly reduced cytotoxicity.7 In addition, L-tryptophan, a quencher of iOH, reduced the photocytotoxicity of TiO2 .7 Therefore, the photocytotoxicity sensitized by TiO2 appears to be the result of intracellular damage involving an array of reactive oxygen species.
/ 2b2f 5042 Mp 856 Monday Aug 11 11:32 AM EL–FRB 5042
Titanium dioxide sensitized oxidation
Our results demonstrate that cellular nucleic acids, particularly RNA, are targets for photooxidative damage sensitized by TiO2 . We have found that measurement of 8-oxodG and 8-oxoG provides a versatile, sensitive and unambiguous indication of oxidative damage to nucleic acids. This approach is versatile since oxidation proceeding through a wide range of reactive intermediates results in guanine hydroxylation. The reactive species generated by photoexcited TiO2 are amongst those known to react with nucleic acids to yield guanine hydroxylation. It has been established that iOH, formed via radiolysis, 28 UV-induced decomposition of H2O2 29 and the Fenton reaction, 30 readily hydroxylates guanine in calf thymus DNA. Similarly, O2 i0 31 and H2O2 30 have been shown to induce formation of 8-oxodG in calf thymus DNA, although oxidation due to these reactive oxygen species proceeds through a transition metal dependent mechanism. In addition, direct interaction of photogenerated holes, trapped at the surface of TiO2 , with guanine in nucleic acids may be possible. The resulting radical cation of guanine has been shown to yield guanine hydroxylation.32 However, oxidative damage to nucleic acids via direct interaction with trapped holes would require adsorption of nucleic acids onto particles of TiO2 . Although our results clearly demonstrate that nucleic acids are targets for photooxidation sensitized by TiO2 , these results do not allow us to distinguish between the above mechanisms for photooxidation. Oxidative damage to cellular RNA and DNA, assessed by measurement of 8-oxoG and 8-oxodG, may be viewed as an indication of oxidative stress in the cytoplasmic and nuclear compartments of cells. Our results indicate that treatment of human skin fibroblasts with TiO2 and UVA produces significant cytoplasmic oxidative stress, measured as formation of 8-oxoG in RNA. In contrast, no significant oxidative stress within cell nuclei, assessed as formation of 8-oxodG in DNA, was observed. Cellular RNA has similarly been found to be more susceptible to oxidative damage than DNA both in vivo 33 and in vitro.18,19 Several reasons for higher levels of oxidative damage in cellular RNA, relative to DNA, may be offered. It has been clearly demonstrated that hydroxylated guanine bases are removed from cellular DNA via an efficient enzymatic repair process.34,35 The half life for removal of 8-oxodG from DNA in human lymphoblast cells, at 377C, has been reported to be 55 min.35 No analogous mechanism for repair of oxidatively damaged cellular RNA has been reported. While DNA repair may result in reduced levels of 8-oxodG under conditions used in our study, it is unlikely that DNA repair can account for the dramatic difference we observed between levels of oxidative damage in cellular RNA and DNA. The intra-
857
cellular location of RNA and DNA may also determine their susceptibility to photosensitized oxidative damage. Cellular RNA is primarily located in the cytoplasm, the site where reactive oxygen species are frequently generated. Indeed, the oxidative damage photocatalyzed by TiO2 would be expected to involve predominately cytoplasmic components. The results presented here, and those of other investigators, 9 indicate that microcrystalline TiO2 is incorporated into the cellular membranes and cytoplasm of mammalian cells in culture. Short-lived reactive oxygen species, generated upon photoexcitation of this intracellular TiO2 , would more readily interact with cytoplasmic targets. In contrast, DNA would be less accessible to shortlived reactive oxygen species generated within the cytoplasm because of compartmentalization within the cell nucleus and the protective role of nuclear chromatin.36 The results we have obtained may imply that longer lived intermediates, such as H2O2 , play a minor role in damaging nucleic acids after photosensitization with TiO2 . In conclusion, TiO2 has been reported to photocatalyze a number of functional changes in cells including altered permeability of cellular membranes to potassium and calcium ions, release of RNA and proteins, and cytotoxicity. To our knowledge, we are the first to report direct measurement of photooxidative damage to cellular components sensitized by TiO2 . Additional studies are necessary to determine the mechanism for photooxidative damage to nucleic acids sensitized by TiO2 . The methods described in this report should prove useful in future studies of TiO2 and other inorganic photosensitizers. Acknowledgements—The authors wish to thank Dr. William Obermeyer, Jr. for providing access to tissue culture facilities. We also gratefully acknowledge Dr. Sandra Bell and Dr. Andrija Kornhauser for helpful discussions.
REFERENCES 1. Sayre, R. M.; Kollias, N.; Roberts, R. L.; Baqer, A.; Sadiq, I. Physical sunscreens. J. Soc. Cosmet. Chem. 41:103–109; 1990. 2. Kraeulter, B.; Bard, A. J. Heterogeneous photocatalytic decomposition of saturated carboxylic acids on TiO2 powder. Decarboxylative route to alkanes. J. Am. Chem. Soc. 100:5985–5992; 1978. 3. Kawai, T.; Sakata, T. Conversion of carbohydrate into hydrogen fuel by a photocatalytic process. Nature 286:474–476; 1980. 4. Fujishima, A.; Honda, K. Electrochemical photolysis of water at a semiconductor electrode. Nature 238:37–38; 1972. 5. Nagame, S.; Oku, T.; Kambara, M.; Konishi, K. Antibacterial effect of the powdered semiconductor TiO2 on the viability of oral micro-organisms. J. Dent. Res. 68:1696–1697; 1989. 6. Ireland, J. C.; Klostermann, P.; Rice, E. W.; Clark, R. M. Inactivation of Escherichia coli by titanium dioxide photocatalytic oxidation. Appl. Environ. Microbiol. 59:1668–1670; 1993. 7. Cai, R.; Kubota, Y.; Shuin, T.; Sakai, H.; Hashimoto, K.; Fu-
/ 2b2f 5042 Mp 857 Monday Aug 11 11:32 AM EL–FRB 5042
858
8.
9.
10.
11. 12.
13.
14. 15. 16.
17. 18. 19. 20. 21.
W. G. WAMER et al. jishima, A. Induction of cytotoxicity by photoexcited TiO2 particles. Cancer Res. 52:2346–2348; 1992. Saito, T.; Iwase, T.; Horie, J.; Morioka, T. Mode of photocatalytic bactericidal action of powdered semiconductor TiO2 on mutans streptococci. J. Photochem. Photobiol. B: Biol. 14:369– 379; 1992. Sakai, H.; Ito, E.; Cai, R.; Yoshioka, T.; Kubota, Y.; Hashimoto, K.; Fujishima, A. Intracellular Ca /2 concentration change of T24 cell under irradiation in the presence of TiO2 ultrafine particles. Biochim. Biophys. Acta 1201:259–265; 1994. Floyd, R. A.; Watson, J. J.; Wong, P. K.; Altmiller, D. H.; Rickard, R. C. Hydroxyl free radical adduct of deoxyguanosine: Sensitive detection and mechanisms of formation. Free Rad. Res. Commun. 1:163–172; 1986. Richter, C.; Park, J.; Ames, B. N. Normal oxidative damage to mitochondrial and nuclear DNA is extensive. Proc. Natl. Acad. Sci. USA 85:6465–6467; 1988. Sai, K.; Takagi, A.; Umemur, T.; Hasegawa, R.; Kurokawa, Y. Changes of 8-hydroxydeoxyguanosine levels in rat organ DNA during the aging process. J. Environ. Pathol. Toxicol. Oncol. 11:139–143; 1992. Kasai, H.; Nishimura, S. Formation of 8-hydroxyguanosine in DNA by oxygen radicals and its biological significance. In: Sies, H., ed. Oxidative stress: Oxidants and antioxidants. New York: Academic Press; 1991:98–116. World Health Organization. Environmental Health Criteria Vol. 24 Titanium. Geneva: World Health Organization; 1982. Xu, H.; Jevnikar, A. M.; Rubin-Kelley, V. E. A simple method for the preparation of chromosomal DNA from cell culture. Nucleic Acids Res. 18:4943; 1990. Wamer, W. G.; Timmer, W. C.; Wei, R. R.; Miller, S. A.; Kornhauser, A. Furocoumarin-photosensitized hydroxylation of guanosine in RNA and DNA. Photochem. Photobiol. 61:336–340; 1995. Wamer, W. G.; Wei, R. R. In vitro photooxidation of nucleic acids by ultraviolet A radiation. Photochem. Photobiol. 65:560– 563; 1997. Wamer, W. G.; Wei, R. R.; Kornhauser, A. Oxidative damage to cellular RNA and DNA accompanying phototoxicity in vitro. Photochem. Photobiol. 59:30S abstr.; 1994. Wei, R. R.; Wamer, W. G.; Bell, S.; Kornhauser, A. Oxidative damage to cellular RNA and DNA photosensitized by fluorescein dyes. Photochem. Photobiol. 61:80S abstr.; 1995. Janzen, E. G.; Nutter, D. E. Jr.; Davis, E. R.; Blackburn, B. J.; Poyer, J. L.; McCay, P. B. On spin trapping hydroxyl and hydroperoxyl radicals. Can. J. Chem. 56:2237–2242; 1978. Serpone, N.; Lawless, D.; Khairutdinov, R. Subnanosecond relaxation dynamics in TiO2 colloidal sols (particle sizes Rp Å 1.0– 13.4 nm). Relevance to heterogeneous photocatalysis. J. Phys. Chem. 99:16655–16661; 1995.
22. Howe, R. F.; Gra¨tzel, M. EPR observation of trapped electrons in colloidal TiO2 . J. Phys. Chem. 89:4495–4499; 1985. 23. Lawless, D.; Serpone, N.; Meisel, D. Role of OH i radicals and trapped holes in photocatalysis. A pulse radiolysis study. J. Phys. Chem. 95:5166–5170; 1991. 24. Jaeger, C. D.; Bard, A. J. Spin trapping and electron spin resonance detection of radical intermediates in photodecomposition of water at TiO2 particulate systems. J. Phys. Chem. 83:3146– 3152; 1979. 25. Finkelstein, E.; Rosen, G. M.; Rauckman, E. J. Spin trapping of superoxide and hydroxyl radical: Practical Aspects. Arch. Biochem. Biophys. 200:1–16; 1980. 26. Clechet, P.; Martelet, C.; Martin, J. R.; Olier, R. Photoelectrochemical behaviour of TiO2 and the formation of hydrogen peroxide. Electrochim. Acta. 24:457–461; 1979. 27. Cai, R.; Hashimoto, K.; Kubota, Y.; Fujishima, A. Increment of photocatalytic killing of cancer cells using TiO2 with the aid of superoxide dismutase. Chem. Lett. :427–430; 1992. 28. Dizdaroglu, M. Formation of an 8-hydroxyguanine moiety in deoxyribonucleic acid on g irradiation in aqueous solution. Biochemistry 24:4476–4481; 1985. 29. Floyd, R. A.; West, M. S.; Eneff, K. L.; Hogsett, W. E.; Tingey, D. T. Hydroxyl free radical mediated formation of 8-hydroxyguanine in isolated DNA. Arch. Biochem. Biophys. 262:266– 272; 1988. 30. Blakely, W. F.; Fuciarelli, A. F.; Wegher, B. J.; Dizdaroglu, M. Hydrogen peroxide-induced base damage in deoxyribonucleic acid. Radiat. Res. 121:338–343; 1990. 31. Aruoma, O. I.; Halliwell, B.; Dizdaroglu, M. Iron ion-dependent modification of bases in DNA by the superoxide radical-generating system hypoxanthine/xanthine oxidase. J. Biol. Chem. 264:13024–13028; 1989. 32. Kasai, H.; Yamaizumi, Z.; Berger, M.; Cadet, J. Photosensitized formation of 7,8-dihydro-8-oxo-2 *-deoxyguanosine (8-hydroxy2 *-deoxyguanosine) in DNA by riboflavin: A non singlet oxygen mediated reaction. J. Am. Chem. Soc. 114:9692–9694; 1992. 33. Fiala, E. S.; Conaway, C. C.; Mathis, J. E. Oxidative DNA and RNA damage in the livers of Sprague–Dawley rats treated with the hepatocarcinogen 2-nitropropane. Cancer Res. 49:5518– 5522; 1989. 34. Demple, B.; Harrison, L. Repair of oxidative damage to DNA: Enzymology and biology. In Richardson, C. C.; Abelson, J. A.; Meister, A.; Walsh, C. T., eds. Annu. Rev. Biochem. Vol. 63. Palo Alto, CA: Annual Reviews Inc.; 1994:915–948. 35. Jaruga, P.; Dizdaroglu, M. Repair of products of oxidative DNA base damage in human cells. Nucleic Acids Res. 24:1389–1394; 1996. 36. Enright, M.; Miller, M. J.; Hays, R.; Floyd, R. A.; Hebbel, R. P. Preferential targeting of oxidative base damage to internucleosomal DNA. Carcinogenesis 17:1175–1177; 1996.
/ 2b2f 5042 Mp 858 Monday Aug 11 11:32 AM EL–FRB 5042
PII S0891-5849(97)00068-3
Original Contribution OXIDATIVE DAMAGE TO NUCLEIC ACIDS PHOTOSENSITIZED BY TITANIUM DIOXIDE
WAYNE G. WAMER, JUN-JIE YIN, and RONG RONG WEI Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, 200 C St., SW, Washington, DC 20204, USA (Received 3 May 1996; Revised 17 March 1997; Accepted 17 March 1997)
Abstract—The semiconductor TiO2 is known to have photobiological activity in prokaryotic and eukaryotic cells. Applications of this photobiological activity have been suggested including sterilization of waste water and phototherapy of malignant cells. Here, several model and cellular systems were used to study the mechanism of photocatalysis by TiO2 . Treatment of TiO2 (anatase, 0.45 mm), suspended in water containing a spin trap 5,5-dimethyl1-pyrroline-N-oxide (DMPO), with UV radiation (320 nm) resulted in an electron spin resonance (ESR) signal characteristic of the hydroxyl radical. Irradiation of solutions containing calf thymus DNA and TiO2 with UVA (320–400 nm) radiation resulted in hydroxylation of guanine bases. The degree of hydroxylation was dependent on both UVA fluence and amount of TiO2 in suspension. Human skin fibroblasts, preincubated 18 h with 10 mg/cm2 TiO2 and then UVA-irradiated (0–58 KJ/m 2 ), showed dose dependent photocytoxicity. RNA, isolated from similarly treated fibroblasts, contained significant levels of photooxidation, measured as hydroxylation of guanine bases. However, no oxidative damage was detectable in cellular DNA. These results suggest that nucleic acids are a potential target for photooxidative damage sensitized by TiO2 , and support the view that TiO2 photocatalyzes free radical formation. Published by Elsevier Science Inc. Keywords—DNA damage, Free radicals, 8-oxodG, 8-Oxo-7,8-dihydro-2 *-deoxyguanosine, Oxidative damage, Photooxidation, Phototoxicity, RNA damage, Titanium dioxide
violet radiation having an energy greater than the optical band gap of TiO2 (3.2 eV for the anatase polymorph of TiO2 ) is absorbed and results in the formation of an electron-hole pair. The electron, generated following absorption of a UV photon, is a reducing agent whereas the concomitantly formed hole is a powerful oxidizing agent. It has been shown that photoexcited TiO2 is a catalyst for a variety of redox reactions including decarboxylation of carboxylic acids 2 and hydrogen production from carbohydrates 3 and water.4 In addition, it has been reported that TiO2 is photobiologically active. When exposed to UVA (320–400 nm) radiation, TiO2 exhibits antibacterial activity.5,6 Recently, it has been proposed that TiO2 may be useful for photodynamic therapy of cancer.7 The mechanism(s) underlying the photobiological activity of TiO2 is not yet well understood. Although evidence substantiates in vitro photooxidative damage sensitized by TiO2 , the intracellular target(s) have not
INTRODUCTION
The utility of TiO2 in a wide range of consumer products derives from its photophysical properties. TiO2 does not absorb radiation in the visible region of the spectrum.1 Therefore, the photophysical properties of TiO2 in the visible region of the spectrum are predominately due to light scattering. This leads to its use as a white pigment and masking or opacifying agent. Ultraviolet radiation, however, is both scattered and absorbed by TiO2 .1 Because of the effective attenuation of UV radiation by TiO2 , TiO2 has been used as a physical sunscreen. Absorption of UV radiation by TiO2 results from the electronic band structure of this semiconductor. UltraAddress correspondence to: Wayne G. Wamer, Food and Drug Administration, Cosmetics Toxicology Branch (HFS-128), 200 C Street, S.W., Washington, DC 20204; E-Mail: [email protected] 851
/ 2b2f 5042 Mp 851 Monday Aug 11 11:32 AM EL–FRB 5042
852
W. G. WAMER et al.
been determined. Investigators have shown that the bactericidal activity photocatalyzed by TiO2 is accompanied by cell membrane damage observed as loss of potassium ions, proteins and RNA from bacterial cells.8 Recently, Sakai et al. have reported that sublethal treatment of T24 cells with TiO2 and UVA radiation resulted in a significant increase in intracellular calcium.9 Following lethal exposure to TiO2 and UVA radiation, a further increase in intracellular calcium was observed. These authors concluded that cellular membrane damage, resulting in greater permeability to calcium, may play a role in photocytotoxicity sensitized by TiO2 . Therefore, the cellular membrane may be a significant target for TiO2-photosensitized damage in both prokaryotic and eukaryotic cells. Here, we examine whether nucleic acids are potential targets for photooxidative damage sensitized by TiO2 . Investigators have demonstrated that measurement of hydroxylation of guanine bases, using HPLC with electrochemical detection, 10 provides a sensitive indicator of oxidative damage to nucleic acids resulting from normal physiological processes such as aerobic metabolism11 and aging 12 as well as exposure to exogenous oxidants.13 In addition, hydroxylation of guanine, leading to formation of 8oxo-7,8-dihydro-2 *-deoxyguanosine ( 8-oxodG ) in DNA and 8-oxo-7,8-dihydroguanosine ( 8-oxoG ) in RNA, is initiated by an array of reactive oxygen species including 1O2 , iOH and O2 i0 .13 Therefore, hydroxylation of guanine can serve as a sensitive biomarker for oxidative stress induced through diverse mechanisms.
ESR spectra were recorded, stored and manipulated using an IBM / PC computer. Photooxidation of calf thymus DNA Calf thymus DNA (100 mg/ml, Sigma Chemical Co., St. Louis, MO) was prepared in 10 mM phosphate buffer, pH 7.2. A suspension of TiO2 (anatase, 0.45 mm, Aldrich Chemical Co., Milwaukee, WI) in distilled water was then added to yield the selected concentration of TiO2 . Samples were then irradiated in open Petri dishes by using a UVA (320–400 nm) light source consisting of two GE F40BL bulbs filtered through 3 mm of soft glass. The spectral irradiance of this source, monitored with an IL1700 Research Radiometer fitted with a SED033 probe (International Light, Inc., Newburyport, MA), was determined to be 1.62 1 10 02 KW/m 2 . After irradiation, samples were centrifuged to remove TiO2 . DNA was precipitated from the supernatant by addition of 12 volume of 7.5 M sodium acetate and 2 volumes of ethanol. DNA pellets were collected by centrifugation, washed with ethanol and dried under reduced pressure. DNA pellets were then reconstituted in 100 ml of 10 mM tri(hydroxymethyl)-aminomethane hydrochloride (Tris-HCl), 10 mM MgCl2 and 1 mM ZnCl2 , pH 7.4. DNA was enzymatically hydrolyzed by addition of 240 Dornase units of DNase 1 and 3.2 units of calf intestinal alkaline phosphatase (Gibco BRL, Grand Island, NY) and incubation at 377C overnight. Hydrolysates were then centrifuged and analyzed by HPLC as described below. Cell culture
MATERIALS AND METHODS
ESR spectroscopy DMPO was purchased from Sigma Chemical Co. ( St. Louis, MO ) and was purified with activated charcoal prior to use. Deionized water was used to prepare all solutions and suspensions. ESR measurements were carried out at room temperature ( 227C ) by using a Varian E-109 spectrometer operating at 9.5 GHz with 100 KHz and 1G field modulation. Spectra were obtained with 15 mW microwave power. Magnetic field and microwave frequency measurements were made with a DTM-141 Digital Teslameter ( GMW Magnetic System, Redwood City, CA ) and EIP-25B Frequency Counter ( EIP Microwave Inc., Milpitas, CA ) . Samples were irradiated in the microwave cavity by using a 1 Kwatt xenon lamp ( Schoeffel, Kratos Inc., Westwood, NJ ) with a single monochromator ( GM 252 ) set at 320 nm. All
A human skin fibroblast cell line (ATCC CRL 1634) was obtained from American Type Culture Collection, Rockville, MD. Cells were cultured in Dulbecco’s modified Eagle medium containing 10% fetal bovine serum (Gibco BRL, Grand Island, NY), antibiotics (penicillin at 100 U/ml and streptomycin at 100 mg/ml), 4.5 mg/ml glucose and 4 mM L-glutamine. Cultures were incubated at 377C in a humidified atmosphere containing 5% CO2 . Cell survival A suspension of TiO2 in water was sterilized by heating at 1007C for 30 min. Confluent monolayers of human skin fibroblasts in 60 mm dishes were treated for 18 h with 4 ml of medium containing 47.2 mg/ml TiO2 in suspension. Measurement of the change in the absorbance of medium at 320 nm showed that more than 99% of the particulate TiO2 had settled from the medium after 18 h. This observation was concomitant
/ 2b2f 5042 Mp 852 Monday Aug 11 11:32 AM EL–FRB 5042
Titanium dioxide sensitized oxidation
with the appearance of a fine precipitate on the fibroblast monolayer. This treatment results in exposure of the fibroblast monolayer (18.9 cm2 ) to 10 mg/cm2 TiO2 . Treated monolayers were washed twice in phosphate buffered saline (PBS), covered with PBS and irradiated at 257C with UVA. The maximum exposure to UVA (58 KJ/m 2 ) required irradiation for 1 h. Cells were then removed from the substratum by trypsinization. Cells (500–1000) in 4 ml of media were then seeded into 60 mm plastic dishes. After incubation for 7–10 days at 377C, cells were fixed with methanol, stained with crystal violet, and cell colonies were counted. While it is known that TiO2 is insoluble in aqueous solution and not bioavailable when given orally or intravenously, 14 no information is available about the solubility of TiO2 in cell culture medium. Therefore, an experiment was performed to determine whether TiO2 or impurities, solubilized by cell culture medium, contribute to the photocytotoxicity sensitized by TiO2 . A suspension containing 47.2 mg/ml TiO2 in medium was prepared as described above. This suspension was then incubated with gentle mixing for 18 h at 377C. Particulate TiO2 was removed by centrifuging twice for 5 min at 5,000 1 g. The supernatant was used to treat fibroblast monolayers. Photocytotoxicity was then assessed (vide ante).
853
layers of skin fibroblasts in 150 mm Petri dishes (approximately 1 1 10 7 cells) were lysed in situ by addition of 15 ml of a solution containing 6 M guanidinium hydrochloride, 0.5 M sodium acetate, 1.3% N-lauroylsarcosine and 1.5 mg proteinase K. The lysate was transferred to a tightly capped tube and incubated for 1 h at 407C. The lysate was then allowed to cool to room temperature and was layered underneath 30 ml ethanol in a 50 ml tube. The tube was tightly capped and slowly turned on its side to avoid mixing of the lysate and ethanol layers. Tubes were then rolled until a white thread appeared (2–4 min). This thread, containing both cellular RNA and DNA, was then washed with 80% ethanol/water, transferred to a 1.5 ml Eppendorf tube and briefly dried under vacuum. Dried pellets were then resuspended in 100 ml 10 mM Tris-HCl, pH 7.4, 10 mM MgCl2 , 1 mM ZnCl2 , 240 units of DNase 1 and 20 units of nuclease P1. Hydrolysis was allowed to proceed 2 h at 377C. Samples were then briefly centrifuged to remove any TiO2 isolated with RNA and DNA. Following centrifugation, 3.2 units of calf intestinal alkaline phosphatase were added to the supernatant. Samples were then allowed to hydrolyze overnight at 377C. Hydrolysates were centrifuged before analysis by HPLC.
HPLC analysis Treatment of cells with TiO2 and UVA for assessment of photooxidative damage to RNA and DNA Human skin fibroblasts were grown to confluency in 150 mm dishes that have a plating area of 143 cm2 . Fibroblast monolayers were treated with 20 ml media supplemented with a sterilized suspension of 71.4 mg/ ml TiO2 for 18 h. This treatment results in exposure of the fibroblast monolayer to 10 mg/cm2 TiO2 . Monolayers were then washed twice with PBS and exposed to UVA radiation as described above. Additionally, skin fibroblasts were treated with medium which had been preincubated for 18 h with 71.4 mg/ml TiO2 . Particulate TiO2 was removed by centrifugation prior to treating fibroblast monolayers. After 18 h treatment, fibroblast monolayers were exposed to UVA radiation. Isolation and enzymatic hydrolysis of cellular RNA and DNA All steps following irradiation of fibroblasts were performed in a darkened room. Cellular RNA and DNA were isolated using a modification of the method described by Xu et al.15 Immediately following treatment with photosensitizer and irradiation, confluent mono-
Nucleoside mixtures, resulting from enzymatic hydrolysis of RNA and DNA, were analyzed by reversedphase HPLC. A 100 ml aliquot was injected onto an Adsorbosphere HS C18 (4.6 mm 1 15 cm, 3 mm) column (Alltech Associates, Deerfield, IL) eluted with 7% methanol in 0.1 M phosphate buffer, pH 4.6 at 0.7 ml/ min. Unmodified nucleosides were monitored at 254 nm by using a Waters Model 400 detector (Millipore Corp., Milford, MA). 8-OxodG and 8-oxoG were quantitated using a Dionex pulsed amperometric detector (Sunnyvale, CA) connected immediately downstream from the UV detector. The glassy carbon working electrode of the electrochemical detector was set at 400 mV relative to the Ag reference electrode. Standards of 8-oxodG and 8-oxoG were prepared as previously described.16 Results are expressed as the fraction of hydroxylated guanine bases (i.e., 8-oxodG/10 5 dG for DNA, 8-oxoG/10 5 G for RNA). The limits of detection were determined to be 0.95 8-oxodG/10 5 dG, and 0.36 8-oxoG/10 5 G in hydrolysates of cellular DNA and RNA. We have previously demonstrated that the sensitivity of this assay is sufficient to detect oxidative damage to cellular RNA and DNA produced by UVA radiation 17 as well as photodynamic sensitizers and visible light.18,19
/ 2b2f 5042 Mp 853 Monday Aug 11 11:32 AM EL–FRB 5042
854
W. G. WAMER et al.
Fig. 1. ESR spectra recorded at 227C after 3 min irradiation with UV radiation of 320 nm. The samples, containing 0.05 M DMPO, received no treatment (A), 3 min irradiation at 320 nm (B), 100 mg/ ml TiO2 (C) or 100 mg/ml followed by 3 min irradiation at 320 nm (D) (aN Å aH Å 14.9 G).
RESULTS
ESR spectra of DMPO spin adducts formed in a suspension of TiO2 irradiated with UVA Figure 1 illustrates the ESR spectrum of the spin adduct formed when a suspension of TiO2 is irradiated at 320 nm. The appearance of the ESR signal was dependent on the presence of both TiO2 and UVA radiation. The ESR spectrum observed consists of the 1:2:2:1 quartet pattern (aN Å aH Å 14.9 G) character-
Fig. 3. Hydroxylation of guanine in calf thymus DNA photocatalyzed by TiO2 . Levels of hydroxylation were dependent both on the fluence of UVA (3A) and concentration of TiO2 in suspension (3B). Data points are the means { standard deviation of three determinations.
istic of the DMPO-hydroxyl radical adduct.20 The time course for development of the ESR signal intensity of the DMPO-hydroxyl radical adduct is shown in Fig. 2. Dose-dependent formation of 8-oxodG in calf thymus DNA Treatment of calf thymus DNA with TiO2 and UVA resulted in significant formation of 8-oxodG. Levels of 8-oxodG were dependent on both the fluence of UVA (Fig. 3A) and amount of TiO2 present (Fig. 3B). Photocytotoxicity of TiO2 Fig. 2. Time dependence of ESR signal intensity of DMPO-OH adduct obtained during UV irradiation at 227C. Sample contained 0.05 M DMPO and 400 mg/ml TiO2 .
It has been reported that human carcinoma cells, incubated with TiO2 , incorporate TiO2 into the outer plasma membrane and cytoplasm.9 We observed that
/ 2b2f 5042 Mp 854 Monday Aug 11 11:32 AM EL–FRB 5042
Titanium dioxide sensitized oxidation
855
Fig. 4. Photograph at 4001 magnification of human skin fibroblasts after incubation for 18 h with 10 mg/cm2 TiO2 . TiO2 appears as bright, highly refractile, particles in the cellular cytoplasm.
skin fibroblasts, cultured 18 h with TiO2 , similarly incorporated TiO2 (Fig. 4). Titanium dioxide particles were incorporated into the cytoplasm and frequently had a perinuclear distribution (Fig. 4). Titanium dioxide particles were not observed in cell nuclei. No cytotoxicity resulted from treatment with TiO2 or UVA irradiation alone (Fig. 5). In contrast, treatment of skin fibroblasts with TiO2 followed by irradiation with UVA resulted in significant, UVA fluencedependent, cytotoxicity (Fig. 5). Also, we determined that solubilization of TiO2 by cell culture medium does not account for the photocytotoxicity sensitized by TiO2 . No photocytotoxicity was observed following 18 h treatment with medium preincubated with TiO2 (Fig. 5). This result demonstrates that particulate TiO2 is required for photocytotoxicity.
observed in cellular DNA following treatment with TiO2 and UVA (Fig. 6B). Similarly, levels of 8-oxodG in DNA isolated from cells receiving TiO2 or UVA alone were not significantly elevated (Fig. 6B). Also, no photooxidative damage in cellular RNA or DNA
Photooxidation of cellular RNA and DNA The methods used here allow measurement of oxidative damage to both cellular RNA and DNA derived from a single cell monolayer. Significant levels of oxidative damage in cellular RNA were found following exposure of skin fibroblast to TiO2 and UVA radiation. Treatment with TiO2 and UVA, resulting in 85% cytotoxicity, produced a 3-fold increase in 8-oxoG (Fig. 6A). Exposure to UVA or TiO2 alone caused no significant increase in levels of 8-oxoG (Fig. 6A). No significant hydroxylation of 2 *-deoxyguanosine was
Fig. 5. Photocytotoxicity of TiO2 . UVA controls ( h ) were irradiated at 257C with the indicated fluences of UVA. Dark controls ( s ) were treated 18 hr with 10 mg/cm2 TiO2 , covered with PBS and incubated in the dark at 257C for a time equivalent to samples receiving UVA. Controls to assess the photocytotoxicity of any medium-soluble components of TiO2 ( , ) were treated with medium, which was preincubated with TiO2 as described in Materials and Methods, and then irradiated with UVA. Samples received 10 mg/cm2 particulate TiO2 and the indicated fluence of UVA ( j ). Data points are the means { the standard deviation of six determinations.
/ 2b2f 5042 Mp 855 Monday Aug 11 11:32 AM EL–FRB 5042
856
W. G. WAMER et al.
90% of these electron/hole pairs recombine within 10 nanoseconds.21 Within approximately 250 nanoseconds after formation, the remaining photogenerated electrons and holes are trapped, predominately at surface sites of particulate TiO2 .21 Evidence suggests that photogenerated electrons are trapped at Ti /4 sites, at or near the surface of TiO2 particles.21,22 Photogenerated holes (h / ) are thought to be trapped as iOH, adsorbed to the surface of TiO2 particles: 23 h / / H2O r iOH / H /
Fig. 6. Photooxidation of cellular RNA (6A) and DNA (6B) sensitized by TiO2 . Data points are the means { the standard deviations of at least four determinations. Confluent monolayers of skin fibroblasts were treated at 257C with UVA ( h ), 10 mg/cm2 TiO2 ( s ), or both UVA and 10 mg/cm2 TiO2 ( j ).
resulted from irradiating fibroblasts that had been treated with medium preincubated with TiO2 (data not shown). This result demonstrates that particulate TiO2 is required for photosensitized damage to cellular nucleic acids. DISCUSSION
Much is known about the photophysical events which follow absorption of UV radiation by TiO2 . Absorption of UV radiation having an energy greater than the bandgap of TiO2 results in the formation of paired electrons and holes which reside, respectively, in the conduction and valence bands of TiO2 . Approximately
A number of reactive oxygen species may be subsequently generated in aqueous solutions. Using spin trapping and ESR detection, Jaeger and Bard have observed ESR spectra consistent with formation of iOH and O2 i0 following absorption of UV radiation by TiO2 (anatase).24 Generation of iOH was attributed to oxidative decomposition of water resulting from the hole produced in photoexcited TiO2 . Jaeger and Bard suggested that O2 i0 may be formed by reduction of O2 by photogenerated electrons in UV irradiated TiO2 particles. In our hands, spin adducts formed during UV irradiation of TiO2 with the spin trap yielded an ESR signal characteristic of DMPO-OH spin adduct alone. The difference between results reported by Jaeger and Bard and in this study may be due to differences in spin traps, pH and composition of solvent.25 The results of our spin trapping studies do not exclude the possibility that O2 i0 , formed following UV irradiation TiO2 , is subsequently interconverted to radicals which give rise to the DMPO-OH spin adduct. Additional studies are necessary to fully understand the mechanism of radical formation by photoexcited TiO2 . In addition, it has been shown that H2O2 can be formed on TiO2 electrodes under aerobic conditions.26 In more complex cellular systems, where enzymatic and metal-catalyzed interconversion of radicals is expected, the relative importance of the reactive oxygen species formed by photoexcited TiO2 is still unknown. Evidence has recently appeared supporting the involvement of both O2 i0 and H2O2 in killing of cancer cells photosensitized by TiO2 . In support for the involvement of O2 i0 , Cai et al. have reported that addition of superoxide dismutase, which converts the relatively unreactive O2 i0 to H2O2 , greatly increases toxicity to HeLa cells photosensitized by TiO2 .27 It has also been demonstrated that addition of catalase, a scavenger of H2O2 , to HeLa cells prior to photosensitization by TiO2 significantly reduced cytotoxicity.7 In addition, L-tryptophan, a quencher of iOH, reduced the photocytotoxicity of TiO2 .7 Therefore, the photocytotoxicity sensitized by TiO2 appears to be the result of intracellular damage involving an array of reactive oxygen species.
/ 2b2f 5042 Mp 856 Monday Aug 11 11:32 AM EL–FRB 5042
Titanium dioxide sensitized oxidation
Our results demonstrate that cellular nucleic acids, particularly RNA, are targets for photooxidative damage sensitized by TiO2 . We have found that measurement of 8-oxodG and 8-oxoG provides a versatile, sensitive and unambiguous indication of oxidative damage to nucleic acids. This approach is versatile since oxidation proceeding through a wide range of reactive intermediates results in guanine hydroxylation. The reactive species generated by photoexcited TiO2 are amongst those known to react with nucleic acids to yield guanine hydroxylation. It has been established that iOH, formed via radiolysis, 28 UV-induced decomposition of H2O2 29 and the Fenton reaction, 30 readily hydroxylates guanine in calf thymus DNA. Similarly, O2 i0 31 and H2O2 30 have been shown to induce formation of 8-oxodG in calf thymus DNA, although oxidation due to these reactive oxygen species proceeds through a transition metal dependent mechanism. In addition, direct interaction of photogenerated holes, trapped at the surface of TiO2 , with guanine in nucleic acids may be possible. The resulting radical cation of guanine has been shown to yield guanine hydroxylation.32 However, oxidative damage to nucleic acids via direct interaction with trapped holes would require adsorption of nucleic acids onto particles of TiO2 . Although our results clearly demonstrate that nucleic acids are targets for photooxidation sensitized by TiO2 , these results do not allow us to distinguish between the above mechanisms for photooxidation. Oxidative damage to cellular RNA and DNA, assessed by measurement of 8-oxoG and 8-oxodG, may be viewed as an indication of oxidative stress in the cytoplasmic and nuclear compartments of cells. Our results indicate that treatment of human skin fibroblasts with TiO2 and UVA produces significant cytoplasmic oxidative stress, measured as formation of 8-oxoG in RNA. In contrast, no significant oxidative stress within cell nuclei, assessed as formation of 8-oxodG in DNA, was observed. Cellular RNA has similarly been found to be more susceptible to oxidative damage than DNA both in vivo 33 and in vitro.18,19 Several reasons for higher levels of oxidative damage in cellular RNA, relative to DNA, may be offered. It has been clearly demonstrated that hydroxylated guanine bases are removed from cellular DNA via an efficient enzymatic repair process.34,35 The half life for removal of 8-oxodG from DNA in human lymphoblast cells, at 377C, has been reported to be 55 min.35 No analogous mechanism for repair of oxidatively damaged cellular RNA has been reported. While DNA repair may result in reduced levels of 8-oxodG under conditions used in our study, it is unlikely that DNA repair can account for the dramatic difference we observed between levels of oxidative damage in cellular RNA and DNA. The intra-
857
cellular location of RNA and DNA may also determine their susceptibility to photosensitized oxidative damage. Cellular RNA is primarily located in the cytoplasm, the site where reactive oxygen species are frequently generated. Indeed, the oxidative damage photocatalyzed by TiO2 would be expected to involve predominately cytoplasmic components. The results presented here, and those of other investigators, 9 indicate that microcrystalline TiO2 is incorporated into the cellular membranes and cytoplasm of mammalian cells in culture. Short-lived reactive oxygen species, generated upon photoexcitation of this intracellular TiO2 , would more readily interact with cytoplasmic targets. In contrast, DNA would be less accessible to shortlived reactive oxygen species generated within the cytoplasm because of compartmentalization within the cell nucleus and the protective role of nuclear chromatin.36 The results we have obtained may imply that longer lived intermediates, such as H2O2 , play a minor role in damaging nucleic acids after photosensitization with TiO2 . In conclusion, TiO2 has been reported to photocatalyze a number of functional changes in cells including altered permeability of cellular membranes to potassium and calcium ions, release of RNA and proteins, and cytotoxicity. To our knowledge, we are the first to report direct measurement of photooxidative damage to cellular components sensitized by TiO2 . Additional studies are necessary to determine the mechanism for photooxidative damage to nucleic acids sensitized by TiO2 . The methods described in this report should prove useful in future studies of TiO2 and other inorganic photosensitizers. Acknowledgements—The authors wish to thank Dr. William Obermeyer, Jr. for providing access to tissue culture facilities. We also gratefully acknowledge Dr. Sandra Bell and Dr. Andrija Kornhauser for helpful discussions.
REFERENCES 1. Sayre, R. M.; Kollias, N.; Roberts, R. L.; Baqer, A.; Sadiq, I. Physical sunscreens. J. Soc. Cosmet. Chem. 41:103–109; 1990. 2. Kraeulter, B.; Bard, A. J. Heterogeneous photocatalytic decomposition of saturated carboxylic acids on TiO2 powder. Decarboxylative route to alkanes. J. Am. Chem. Soc. 100:5985–5992; 1978. 3. Kawai, T.; Sakata, T. Conversion of carbohydrate into hydrogen fuel by a photocatalytic process. Nature 286:474–476; 1980. 4. Fujishima, A.; Honda, K. Electrochemical photolysis of water at a semiconductor electrode. Nature 238:37–38; 1972. 5. Nagame, S.; Oku, T.; Kambara, M.; Konishi, K. Antibacterial effect of the powdered semiconductor TiO2 on the viability of oral micro-organisms. J. Dent. Res. 68:1696–1697; 1989. 6. Ireland, J. C.; Klostermann, P.; Rice, E. W.; Clark, R. M. Inactivation of Escherichia coli by titanium dioxide photocatalytic oxidation. Appl. Environ. Microbiol. 59:1668–1670; 1993. 7. Cai, R.; Kubota, Y.; Shuin, T.; Sakai, H.; Hashimoto, K.; Fu-
/ 2b2f 5042 Mp 857 Monday Aug 11 11:32 AM EL–FRB 5042
858
8.
9.
10.
11. 12.
13.
14. 15. 16.
17. 18. 19. 20. 21.
W. G. WAMER et al. jishima, A. Induction of cytotoxicity by photoexcited TiO2 particles. Cancer Res. 52:2346–2348; 1992. Saito, T.; Iwase, T.; Horie, J.; Morioka, T. Mode of photocatalytic bactericidal action of powdered semiconductor TiO2 on mutans streptococci. J. Photochem. Photobiol. B: Biol. 14:369– 379; 1992. Sakai, H.; Ito, E.; Cai, R.; Yoshioka, T.; Kubota, Y.; Hashimoto, K.; Fujishima, A. Intracellular Ca /2 concentration change of T24 cell under irradiation in the presence of TiO2 ultrafine particles. Biochim. Biophys. Acta 1201:259–265; 1994. Floyd, R. A.; Watson, J. J.; Wong, P. K.; Altmiller, D. H.; Rickard, R. C. Hydroxyl free radical adduct of deoxyguanosine: Sensitive detection and mechanisms of formation. Free Rad. Res. Commun. 1:163–172; 1986. Richter, C.; Park, J.; Ames, B. N. Normal oxidative damage to mitochondrial and nuclear DNA is extensive. Proc. Natl. Acad. Sci. USA 85:6465–6467; 1988. Sai, K.; Takagi, A.; Umemur, T.; Hasegawa, R.; Kurokawa, Y. Changes of 8-hydroxydeoxyguanosine levels in rat organ DNA during the aging process. J. Environ. Pathol. Toxicol. Oncol. 11:139–143; 1992. Kasai, H.; Nishimura, S. Formation of 8-hydroxyguanosine in DNA by oxygen radicals and its biological significance. In: Sies, H., ed. Oxidative stress: Oxidants and antioxidants. New York: Academic Press; 1991:98–116. World Health Organization. Environmental Health Criteria Vol. 24 Titanium. Geneva: World Health Organization; 1982. Xu, H.; Jevnikar, A. M.; Rubin-Kelley, V. E. A simple method for the preparation of chromosomal DNA from cell culture. Nucleic Acids Res. 18:4943; 1990. Wamer, W. G.; Timmer, W. C.; Wei, R. R.; Miller, S. A.; Kornhauser, A. Furocoumarin-photosensitized hydroxylation of guanosine in RNA and DNA. Photochem. Photobiol. 61:336–340; 1995. Wamer, W. G.; Wei, R. R. In vitro photooxidation of nucleic acids by ultraviolet A radiation. Photochem. Photobiol. 65:560– 563; 1997. Wamer, W. G.; Wei, R. R.; Kornhauser, A. Oxidative damage to cellular RNA and DNA accompanying phototoxicity in vitro. Photochem. Photobiol. 59:30S abstr.; 1994. Wei, R. R.; Wamer, W. G.; Bell, S.; Kornhauser, A. Oxidative damage to cellular RNA and DNA photosensitized by fluorescein dyes. Photochem. Photobiol. 61:80S abstr.; 1995. Janzen, E. G.; Nutter, D. E. Jr.; Davis, E. R.; Blackburn, B. J.; Poyer, J. L.; McCay, P. B. On spin trapping hydroxyl and hydroperoxyl radicals. Can. J. Chem. 56:2237–2242; 1978. Serpone, N.; Lawless, D.; Khairutdinov, R. Subnanosecond relaxation dynamics in TiO2 colloidal sols (particle sizes Rp Å 1.0– 13.4 nm). Relevance to heterogeneous photocatalysis. J. Phys. Chem. 99:16655–16661; 1995.
22. Howe, R. F.; Gra¨tzel, M. EPR observation of trapped electrons in colloidal TiO2 . J. Phys. Chem. 89:4495–4499; 1985. 23. Lawless, D.; Serpone, N.; Meisel, D. Role of OH i radicals and trapped holes in photocatalysis. A pulse radiolysis study. J. Phys. Chem. 95:5166–5170; 1991. 24. Jaeger, C. D.; Bard, A. J. Spin trapping and electron spin resonance detection of radical intermediates in photodecomposition of water at TiO2 particulate systems. J. Phys. Chem. 83:3146– 3152; 1979. 25. Finkelstein, E.; Rosen, G. M.; Rauckman, E. J. Spin trapping of superoxide and hydroxyl radical: Practical Aspects. Arch. Biochem. Biophys. 200:1–16; 1980. 26. Clechet, P.; Martelet, C.; Martin, J. R.; Olier, R. Photoelectrochemical behaviour of TiO2 and the formation of hydrogen peroxide. Electrochim. Acta. 24:457–461; 1979. 27. Cai, R.; Hashimoto, K.; Kubota, Y.; Fujishima, A. Increment of photocatalytic killing of cancer cells using TiO2 with the aid of superoxide dismutase. Chem. Lett. :427–430; 1992. 28. Dizdaroglu, M. Formation of an 8-hydroxyguanine moiety in deoxyribonucleic acid on g irradiation in aqueous solution. Biochemistry 24:4476–4481; 1985. 29. Floyd, R. A.; West, M. S.; Eneff, K. L.; Hogsett, W. E.; Tingey, D. T. Hydroxyl free radical mediated formation of 8-hydroxyguanine in isolated DNA. Arch. Biochem. Biophys. 262:266– 272; 1988. 30. Blakely, W. F.; Fuciarelli, A. F.; Wegher, B. J.; Dizdaroglu, M. Hydrogen peroxide-induced base damage in deoxyribonucleic acid. Radiat. Res. 121:338–343; 1990. 31. Aruoma, O. I.; Halliwell, B.; Dizdaroglu, M. Iron ion-dependent modification of bases in DNA by the superoxide radical-generating system hypoxanthine/xanthine oxidase. J. Biol. Chem. 264:13024–13028; 1989. 32. Kasai, H.; Yamaizumi, Z.; Berger, M.; Cadet, J. Photosensitized formation of 7,8-dihydro-8-oxo-2 *-deoxyguanosine (8-hydroxy2 *-deoxyguanosine) in DNA by riboflavin: A non singlet oxygen mediated reaction. J. Am. Chem. Soc. 114:9692–9694; 1992. 33. Fiala, E. S.; Conaway, C. C.; Mathis, J. E. Oxidative DNA and RNA damage in the livers of Sprague–Dawley rats treated with the hepatocarcinogen 2-nitropropane. Cancer Res. 49:5518– 5522; 1989. 34. Demple, B.; Harrison, L. Repair of oxidative damage to DNA: Enzymology and biology. In Richardson, C. C.; Abelson, J. A.; Meister, A.; Walsh, C. T., eds. Annu. Rev. Biochem. Vol. 63. Palo Alto, CA: Annual Reviews Inc.; 1994:915–948. 35. Jaruga, P.; Dizdaroglu, M. Repair of products of oxidative DNA base damage in human cells. Nucleic Acids Res. 24:1389–1394; 1996. 36. Enright, M.; Miller, M. J.; Hays, R.; Floyd, R. A.; Hebbel, R. P. Preferential targeting of oxidative base damage to internucleosomal DNA. Carcinogenesis 17:1175–1177; 1996.
/ 2b2f 5042 Mp 858 Monday Aug 11 11:32 AM EL–FRB 5042