C60 fullerene accumulation in human leukemic cells and perspectives of LED‐mediated photodynamic therapy

Abstract Recent progress in nanobiotechnology has attracted interest to a biomedical application of the carbon nanostructure C60 fullerene since it possesses a unique structure and versatile biological activity. C60 fullerene potential application in the frame of cancer photodynamic therapy (PDT) relies on rapid development of new light sources as well as on better understanding of the fullerene interaction with cells. The aim of this study was to analyze C60 fullerene effects on human leukemic cells (CCRF‐CEM) in combination with high power single chip light‐emitting diodes (LEDs) light irradiation of different wavelengths: ultraviolet (UV, 365 nm), violet (405 nm), green (515 nm) and red (632 nm). The time‐dependent accumulation of fullerene C60 in CCRF‐CEM cells up to 250 ng/106 cells at 24 h with predominant localization within mitochondria was demonstrated with immunocytochemical staining and liquid chromatography mass spectrometry. In a cell viability assay we studied photoexcitation of the accumulated C60 nanostructures with ultraviolet or violet LEDs and could prove that significant phototoxic effects did arise. A less pronounced C60 fullerene phototoxic effect was observed after irradiation with green, and no effect was detected with red light. A C60 fullerene photoactivation with violet light induced substantial ROS generation and apoptotic cell death, confirmed by caspase3/7 activation and plasma membrane phosphatidylserine externalization. Our work proved C60 fullerene ability to induce apoptosis of leukemic cells after photoexcitation with high power single chip 405 nm LED as a light source. This underlined the potential for application of C60 nanostructure as a photosensitizer for anticancer therapy. Graphical abstract Figure. No Caption available. HighlightsC60 fullerene and high power LED light exhibited toxicity against leukemic cells.C60 showed high intracellular uptake and predominant mitochondrial localization.LED irradiated C60 fullerene induced oxidative stress and apoptosis.


Introduction
Photodynamic therapy is a non-surgical approach aimed on the selective elimination of cancer cells. The main idea of PDT is to combine two non-toxic componentsphotosensitizing molecule and visible lightwhich in the presence of oxygen gain a pronounced toxicity [1][2][3]. Anticancer PDT effects are realized directly through the induction of cancer cell death and/or indirectly when damage of the vascular system and activation of the immune response are provoked [4,5]. Rapid development of endoscopic fiber optic devices [6,7] allows to test PDT in treatment not only for skin malignacies, but for brain, lung, esophagus, colon, pancreas, liver, bile duct, breast, bladder, prostate and neck cancers as well [1,4,8].
Over the past decade, the application of nanoparticulate agents has been established both in pharmaceutical research and in clinical settings [4,9]. The constantly increasing interest in novel nanotechnology platforms for biomedical applications stimulated the investigation of carbon nanomaterials, including fullerenes and their most prominent representative -C 60 fullerene [10]. Pristine non-modified C 60 fullerene is a lipophilic, spheroidal shaped and symmetrical molecule with 0.72 nm in diameter [11,12]. Due to the specific packing of atoms in penta-and hexagon units the surface of C 60 is three times smaller than expected for biological molecules with the same molecular weight of 720 Da. The unusual structure of C 60 fullerene determines its unique physicochemical properties and biological activity [9,[11][12][13][14].
Today considerable attention is devoted to C 60 fullerene as potential T regulator of oxidative balance in biological systems. Since a C 60 molecule consists entirely of sp 2 -hybridized carbon atoms [14], it is able to generate reactive oxygen species (ROS) after UV-vis light irradiation with a quantum yield of 1.0 [15]. C 60 fullerene advantages compared to conventional photosensitizing molecules are the higher photostability and lesser photobleaching [2]. There are two ways of ROS production by photoexcited C 60 : by energy (type I) or electron (type II) transfer from photoexcited C 60 to oxygen ( Fig. 1) [2,[15][16][17][18][19]. The produced reactive oxygen species are excellent oxidizing agents that react with a wide range of biological targets. Oxidative stress, which occurs when ROS generation overwhelms the cell antioxidant defense system can lead to cell death by apoptosis [16][17][18]. Mitochondria have been found to be an important subcellular target for many photosensitizing drugs due to its role in apoptosis induction [3,17]. C 60 fullerene-mediated PDT efficiency in vitro and in vivo was shown to a large extent with its hydrophilic derivatives hydroxy- [14,17,20], carboxy- [16], PEG [10,12,17] -C 60 and C 60 with various organic substitutes [15,18,[20][21][22][23]. Functionalization of C 60 improves its water solubility and increases its biocompatibility by decreasing the aggregate size [14], but on the other hand, inhibits its interaction with cellular lipid membranes and changes the pattern of cellular uptake [14,18,19,22,23]. C 60 diffuses through bilayered membrane from six [22] to nine [24] orders of magnitude faster as compared with its hydrophilic derivatives, which interact with polar groups on the membrane surface instead of entering the cell. Higher lipophilicity promotes diffusion of pharmaceutical agents across the plasma membrane and further relocation to other cellular membranes, thus facilitating intracellular uptake [17]. Hydrophobic drugs are shown to attack the cancer cells mainly by direct interactions, weather hydrophilic agents act indirectly by damaging blood vessels [18]. Moreover, the presence of functional groups on C 60 fullerene surface decreases the quantum yield of singlet oxygen production after molecule photoexcitation [15,16]. So, the cellular uptake and further biological effects of pristine C 60 and its derivatives could be different. Pristine C 60 fullerene may be applied in PTD in the form of liposome-based delivery systems [2,8,16,25,26] or water colloidal C 60 solutions [16,[27][28][29][30]. Previously, a negligible toxicity of pristine C 60 stable colloid solution [30] against normal cells was shown [29,30]. At the same time a pronounced proapoptotic effect was detected in leukemic cells treated with pristine C 60 fullerene and irradiated with UV/Vis light in the range of 320-600 nm [11,27,31]. These data indicate the potential of C 60 as an effective photosensitizer in cancer therapy.
The use of high power single chip light-emitting diodes is expected to promote PDT application, since they have a higher portability and extremely lower cost, compromising the efficiency [32,33] of lasers as the classical PDT light sources. LED-based equipment has a high potential to simplify PDT's technical part and to reduce costs [34]. In this paper, we report first data concerning the pristine C 60 fullerene accumulation and localization inside human leukemic cells and its phototoxic effects potentially induced by UV, violet, green and red high power single chip LEDs light irradiation.

C 60 fullerene synthesis
The pristine C 60 fullerene aqueous colloid solution was prepared as described in [28] by C 60 fullerene transfer from toluene to water using continuous ultrasound sonication. Obtained C 60 colloid solution was characterized by high C 60 fullerene concentration (2 × 10 -4 М, purity 99%), stability and homogeneity.

Spectrophotometric analysis
Samples (100 μl of C 60 colloid solution) were placed into a 96-well plate Sarstedt (Nümbrecht, Germany), C 60 absorbance spectrum was measured with the multimode microplate spectrometer Tecan infinite M200 PRO (Grödig, Austria) at the following parameters: wavelength range 200-900 nm, wavelength step size: 2 nm, number of flashes per well: 10.

Cell culture
The human cancer cell line of leucosis origin -CCRF-CEM (ACC 240)was purchased from the Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures. Cells were maintained in RPMI 1640 medium supplemented with 10% Fetal Bovine Serum, 1% Penicillin/Streptomycin and 2 mM Glutamine, using 25 cm 2 flasks at a 37°C with 5% CO 2 in a humidified incubator Binder (Tuttlingen, Germany). The number of viable cells was counted using 0.1% trypan blue staining and a Roche Cedex XS Analyzer (Basel, Switzerland).  Schematic mechanism of C 60 fullerene photodynamic cancer therapy. An absorbed photon excites C 60 fullerene to the first excited singlet state S 1 , that relaxes to the more long lived triplet state T 1 . The C 60 triplet interacts with oxygen either through type I or type II, resulting in the intensification of reactive oxygen species generation and induction of apoptotic cell death. 0.2% Triton X100 for 10 min at RT and washed again with PBS. Blocking was performed using 10% BSA for 20 min with following washing in PBS. The primary monoclonal antibody IgG (developed in mouse) against C 60 fullerene conjugated to thyroglobulin of bovine origin (dilution ratio of 1:30 in PBS/1.5%BSA, 1-10F-A8 Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) was added to the CCRF-CEM cells and incubated overnight at 4°C in a humidified chamber. Then CCRF-CEM cells were incubated for 3 h at RT with a FITC-labeled polyclonal antibody against mouse IgG developed in rabbit (dilution ratio of 1:200 in PBS/1.5%BSA, F7506 Sigma-Aldrich Co., St-Louis, USA). Slides were washed between each step in three shifts of PBS for 15 min each. The coverslips were rinsed with dH 2 O, incubated with nucleus staining antifade solution (0.6 µM DAPI, 90 mM p-Phenylenediamine in glycerol/PBS) for 2 h in the dark and sealed with slides. CCRF-CEM cells were observed using a Fluorescence Microscope Keyence BZ-9000 BIOREVO (Osaka, Japan) equipped with blue (λ ex 377 nm, λ em 447 nm), green (λ ex 472 nm, λ em 520 nm) and red (λ ex 543 nm, λ em 593 nm) filters with the acquisition software Keyence BZ-II Viewer (Osaka, Japan). The merged images and single cell fluorescence intensity profiles were processed with the Keyence BZ-II Analyzer software (Osaka, Japan).

High performance liquid chromatography-electro spray ionization-mass spectrometry (HPLC-ESI-MS)
The liquid chromatography separation and mass spectrometric detection were achieved by employing the Nexera HPLC system coupled to the LCMS-8040 Tandem Quadrupole Mass Spectrometer, equipped with an ESI source (Shimadzu, Kyoto, Japan). Chromatographic separation was performed using the column Eclipse XDV-C8 150 mm × 4,6 mm, 5 µm (Agilent, Santa Clara, USA). Optimized HPLS-ESI-MS conditions described in [35]. For data processing the software LabSolutions LCMS (Shimadzu, Kyoto, Japan) was used.

C 60 fullerene extraction
CCRF-CEM cells (2 × 10 5 /ml) were seeded in 6-well plate Sarstedt (Nümbrecht, Germany). After 24 h cells were incubated for 0-48 h in the presence of 20 µM C 60 , washed with PBS three times and transferred to the dH 2 0. The freeze-thawing cycle was repeated three times. The probes were dried at 80°C under reduced pressure. Toluene/2-propanol (6:1, v/v) was added in the final volume 0.5 ml, the mixture was sonicated for 1 h and centrifuged (70 min, 20238g). The toluene layer was analyzed with HPLC-ESI-MS.

Isolation of mitochondria
CCRF-CEM cells were incubated for 24 h in the presence of 20 µM C 60 and the mitochondria fraction was isolated accordingly to [36]. Briefly, cell suspension (5 × 10 6 /4 ml) was centrifuged at 600g at 4°C for 10 min, cells were resuspended in 3 ml of ice cold isolation buffer (IB: 0.01 M Tris-MOPS, 1 mM EGTA/Tris, 0.2 M sucrose, pH 7.4) and homogenized in the teflon-glass potter on ice. The homogenate was centrifuged at 600g at 4°C for 10 min. The collected supernatant (S1) was centrifuged at 7000g at 4°C for 10 min. The pellet (P2) was resuspended in 200 μl ice-cold IB and centrifuged at 7000g at 4°C for 10 min. The mitochondrial fraction obtained in pellet (P3) was used for extraction of C 60 fullerene as well as for measurements of protein concentration [37] and succinate-reductase activity as mitochondrial marker [38]. The data concerning mitochondrial purity test available in [35].

Photodynamic therapy in vitro and cell viability assay
For cell viability assay, 10 4 cells/well were cultured in 96-well cell culture plates (Sarstedt, Nümbrecht, Germany) at for 24 h and then incubated for 24 h with 20 µM C 60 and washed with PBS. Light irradiation was applied at the following wavelengths: UV -365 nm Nichia SMD LED UV NCSU275, 140.6 mW/cm 2 (LUMITRONIX LED-Technik GmbH, Hechingen, Germany); violet -405 nm high power single chip LED VL400-EMITTER, 108.3 mW/cm 2 ; green -515 nm high power single chip LED APG2C1-515, 50.9 mW/cm 2 ; red -650 nm, ELD-650-523, 5.1 mW/cm 2 (Roithner Lasertechnik GmbH, Vienna, Austria); red -632 nm helium-neon 30 mW laser, 90 mW/cm 2 (Melles Griot, New York, USA). The light fluence was used in the range of 1-8 J/cm 2 for UV, 5-20 J/cm 2 for violet and green light and 1-80 J/cm 2 for red light. PBS was replaced with fresh medium immediately after irradiation. Control cells were incubated without exposure to fullerene treatment or light irradiation. After 24 h incubation cell viability was determined with MTT reduction assay. 10 μl of MTT solution (5 mg/ml in PBS) was added to each well and cells were incubated for 2 h at 37°C. The culture medium was then replaced with 100 μl of DMSO, diformazan formation was determined by measuring absorption at 570 nm with a microplate reader Tecan Infinite M200 Pro (Männedorf, Switzerland).

Intracellular ROS-generation
To determine ROS production 2,7-dichlorofluorescin diacetate (DCFH-DA) (Sigma-Aldrich Co., St-Louis, USA) was applied. A 5 mM stock solution of DCFH-DA was prepared in DMSO, stored at −20°C and diluted with PBS immediately before use. 10 4 cells were incubated in RPMI with 20 µM C 60 for 24 h, irradiated (10 J/cm 2 405 nm LED) as indicated above, and washed once with PBS at 1 h and 3 h of further incubation. 5 µM DCF-DA was added and the fluorescence (λ ex 488 nm, λ em 520 nm) was recorded every 5 min during 50 min with the microplate reader Tecan Infinite M200 Pro (Männedorf, Switzerland). After 60 min incubation fluorescent images of cells were obtained with the Fluorescence Microscope Keyence BZ-9000 BIOREVO (Germany), equipped with a green filter (λ ex 472 nm, λ em 520 nm).

Caspase 3/7 activity
The CCRF-CEM cells were seeded into 96-well plates (10 4 cells/ well) and incubated for 24 h. The cells were treated with 20 µM C 60 for 24 h and irradiated (405 nm, 10 J/cm 2 ) irradiation as described above. Activity of caspases 3/7 was determined during 6 h period after light exposure using the Promega Caspase-Glo® 3/7 Activity assay kit (Madison, USA) according to the manufacturer's instructions. Briefly, the plates were removed from the incubator and allowed to equilibrate to RT for 30 min. After treatment, an equal volume of Caspase-Glo 3/7 reagent was added followed by gentle mixing with a plate shaker at 300 rpm for 1 min. The plate was then incubated at RT for 2 h. The luminescence of each sample was measured with the microplate reader Tecan Infinite M200 Pro (Männedorf, Switzerland).

Flow cytometry analysis
CCRF-CEM cells were seeded onto 6-well plates at a cell density of 2 × 10 5 cells/well in 2 ml of culture medium, incubated for 24 h, than treated with 20 µM C 60 for 24 h and irradiated with 405 nm LED as described above. At 6 and 24 h incubation period the cells were harvested. Apoptosis was detected by Annexin V-fluorescein isothiocyanate/propidium iodide (Annexin V-FITC/PI) apoptosis detection kit (eBioscience™, San Diego, USA) according to the manufacturer's instructions. Briefly, cells were harvested and washed with Binding buffer. After addition of FITC-conjugated Annexin V cells were incubated for 15 min at RT in dark. Cells were washed with Binding buffer and at 10 min after propidium iodide addition were analyzed with the flow cytometer BD FACSJazz™ (Singapore). A minimum of 20,000 cells per sample were acquired and analyzed with the BD FACS™ software (Singapore).

Statistical analysis
All experiments were carried out with a minimum of four replicates. Data analysis was performed with the use of the GraphPad Prism 7 (GraphPad Software Inc., USA). Paired Student's t-tests were performed. Differences values p < 0.05 were considered to be significant.

C 60 fullerene uptake by leukemic cells and its intracellular distribution
The first requirement for any photosensitizing agent is an extensive penetration into the cancer cells since otherwise ROS generation is not sufficient to induce cell death. Effective cellular uptake of pristine C 60 fullerene was shown in vitro on different cell lines, including human keratinocytes HaCaT and human lung carcinoma cells A549 [39], human monocyte-derived macrophage cells [40], mouse macrophages RAW 264.7 [41], human mammary epithelial cells MCF10A and MDA MB 231 [42]. Though the intracellular accumulation of nanostructures was proved, still little is known about its subcellular localization, as well as its ability to relocate and to realize effects at the level of intracellular compartments in cells of different types.

Qualitative analysis
The intracellular uptake and distribution of C 60 fullerene was studied by fluorescent immunostaining of human leukemic CCRF-CEM cells using a FITC-labeled antibody against C 60 . DNA-binding dye DAPI was used as a cell nucleus marker and MitoTracker Orange as a mitochondrial marker. No significant unspecific green FITC-fluorescence was observed in the control cells incubated in the absence of C 60 fullerene. On Fig. 2B the images of CCRF-CEM cells stained after 24 h incubation with C 60 are presented. No changes in nucleus fluorescence as compared with controls were found, whereas green fluorescent punctuated dots surrounding the nucleus were detected (Fig. 2B). The data showed that C 60 fullerene could diffuse through the cell plasma membrane and locate in the perinuclear region of leukemic cells. Next, we have evaluated whether C 60 fullerene could localize in the mitochondria membranes. The data of fluorescent microscopy verified a partial co-localization of C 60 antibodies and the mitochondrial marker. The intensity profiles of the three fluorescence channels (Fig. 2C) for individual cells ( Fig. 2A, B arrows) revealed the overlap of the green C 60 fullerene and the red mitochondria signals. These data support the evidence of C 60 fullerene localization in mitochondria of human leukemic cells.

Quantitative analysis
To study the accumulation dynamics we have extracted C 60 fullerene from the cell homogenate as well as from the mitochondrial fraction and carried out liquid chromatography mass-spectrometry analysis. This method together with electro-spray ionization was previously reported to be an effective tool for C 60 fullerene quantification in water samples [43], zebrafish embryo [44] and human skin keratinocytes HaCaT [25]. The method [35] enabled the quantitative analysis of C 60 fullerene accumulation in CCRF-CEM cells. According to the data presented on Fig. 2D the intracellular content of C 60 fullerene reached its maximum of < 250 ng/10 6 cells after 24 h of incubation. A subsequent minor decrease of C 60 fullerene content in leukemic cells extract at 48 h could be accounted by its partial efflux from the cancer cells.
HPLC-ESI-MS analysis of C 60 fullerene content in the mitochondria fraction showed accumulation of the nanostructure at a level of < 180 ng/10 6 cells at 24 h that amounted to 72% of its overall content in cell extract. This data demonstrate that C 60 fullerene predominantly accumulates within mitochondria. This could be explained by a high electronegativity of C 60 fullerene and a resulting affinity to the mitochondria-associated proton pool [22,45]. According to density functional theory simulations, C 60 fullerene diffuses into the protonated mitochondrial intermembrane space, where it interacts with up to 6 protons, acquiring a positive charge [34]. This phenomenon is common for other negatively charged carbon nanoparticles such as single walled carbon nanotubes [46,47]. These were shown to be localized in mitochondria of different cells, too (ASTC-a-1, MCF 7, COS 7, EVC304 and RAW264.7) [48].

LED-assisted photodynamic therapy in vitro
Photosensitizing effects of C 60 fullerene and its derivatives are induced by different visible light sources including broadband mercuryvapor [11,27,31], halogen [2,19,49,50], tungsten-xenon [51,52] and fluorescent [53] lamps as well as sharp band lasers [2,16,53]. A recent study [54] demonstrates C 60 fullerene photoinduced cytotoxic effects against leukemic cells after irradiation with the white light-emitting diode lamp with the board emission spectrum (420-700 nm). LEDs as a light-source for PDT have been explored previously and were shown to be more cost-effective and serviceable as clinical lamps or lasers [32][33][34]. The fractionation of light is a promising tool for the optimization of PDT in order to select the most effective combination of the photosensitizing agent and the light conditions such as light wavelength and fluence [1,51], that can be realized with the use of sharp spectrum LEDs. In this study we tested four high output light-emitting diode chips with 365, 405, 515 and 650 nm light irradiation.
The efficiency of light-induced excitation of photosensitizing agents substantially depends on their relative optical absorption extinction coefficients. The UV/Vis absorption spectrum (200-900 nm) of pristine C 60 fullerene aqueous colloidal solution (Fig. 3A) has three intense absorption bands typical for C 60 [28] with maxima at 220, 265, and 350 nm and a long broad tail up to the red region of the visible light. Fig. 3A demonstrates that the absorption spectrum of C 60 fullerene and the spectra of the used LEDs are overlapping, suggesting that they could be applied for C 60 photoexcitation.
After 24 h treatment with C 60 fullerene leukemic cells were irradiated and at 48 h the cell viability was estimated with the MTT assay. As shown in Fig. 3B-D, the effect of light irradiation itself on CCRF-CEM cell viability depends on light wavelength. Irradiation of cells in the ultraviolet light at 365 nm was followed by decrease of cell viability. The effect became stronger with the increase of the light fluence. After irradiation with 8 J/cm 2 the viability was only 32% compared to viability of control cells in the dark. So the high toxic effects of UV light itself makes its application unfavorable. Irradiation with visible light was followed by less cytotoxic effect in comparison with UV irradiation even considering the fact, that the visible light was used at higher fluences. After irradiation with violet light at 405 nm at a maximal dose (20 J/cm 2 ) cell viability was decreased on 16% of the control level (Fig. 3C). Light irradiation at 405 nm is used in practice for the sterilization of both clinical and nonclinical environment due to the strong bactericidal activity [55,56], but its inactivating effect against mammalian cells is slight [55,57]. No significant toxic effect was observed after cells irradiation with green light at 515 nm and red light at 650 nm with fluence rate of 20 J/cm 2 , the cell viability was 90% and 95% accordingly as compared with control ( Fig. 3D and Fig. SI 1).
We next studied cell viability after treatment with C 60 fullerene and photoexcitation of accumulated nanostructures. No cytotoxic effect was detected when CCRF-CEM cells were treated with C 60 and kept in the dark. When cells were incubated for 24 h in the presence of 20 µM C 60 fullerene and then irradiated with 365 nm or 405 nm LEDs, a substantial decrease of cell viability was observed at 24 h after light exposure ( Fig. 3B and C). Сombined treatment with C 60 fullerene and UV 365 nm light at the doses of 2 and 4 J/cm 2 decreased cell viability up to 39% and 7% respectively as compared with control cells. The increase of UV LED light fluence up to 8 J/cm 2 was followed by almost total cell death (Fig. 3B). Photoexcitation of accumulated C 60 fullerene with violet 405 nm light at 5, 10 and 20 J/cm 2 light fluence caused cell viability decrease by 73%, 54% and 10%, respectively, as compared with control cells (Fig. 3C). C 60 fullerene exhibited lower cytotoxicity under green 515 nm LED light irradiation. The viability of cells, treated with C 60 fullerene for 24 h and exposed to the 5 and 10 J/cm 2 green LED light irradiation, was estimated to be around 85% with a further 10% decrease at 20 J/cm 2 (Fig. 3D). No fullerene cytotoxicity was observed at the further shift of the light wavelength into the red region of visible spectrum (Fig. SI 1). No significant effect on the viability of leukemic cells, incubated either in the presence or the absence of 20 µM C 60 fullerene was detected under 650 nm LED light irradiation. Even  when we have applied intensities up to 80 J/cm 2 with the use of 650 nm helium-neon 30 mW laser no cytotoxic effect was observed because of low absorbance of longwavelength light by C 60 . In order to increase the red light absorbance C 60 fullerene can be modified with decacationic radicals [51] and additional red light harvesting chromophores [21,58] or excited with simultaneously delivered two near-infrared photons [2,59]. The anticancer effect of pristine C 60 fullerene and LEDs light irradiation combination against human leukemic cells was shown to be strongly dependent on the light wavelength and fluence. The shortwavelength LEDs light irradiation was followed by most substantial decrease of leukemic cells viability, loaded with C 60 fullerene. The efficiency of combined effect of C 60 fullerene and LEDs light irradiation was proved to follow an expected order 365 nm > 405 nm > 515 nm > 650 nm (no effect). The obtained results indicate, that high output single chip 405 nm LED is the most favorable light source for C 60 fullerene photodynamic therapy of human leukemic cells at intensities from 5 to 20 J/cm 2 . The application of a 405 nm sharp spectrum high output LED chip for C 60 fullerene photodynamic therapy allows to decrease the light dose in comparison with broad spectrum light sources, that are used at higher fluence rates of more than 100 J/cm 2 [2,26,50,60].

Reactive oxygen species generation
The efficient and continuous intracellular ROS production is a critical step for realization of a photoexcited C 60 fullerene anticancer effect. ROS generation was estimated with the use of the fluorescence dye 2′,7′-dichlorodihydrofluorescein diacetate [61,62]. DCFH-DA is able to penetrate inside the cell, where it is deacyleted by esterases to its nonfluorescent form DCFH. Upon interaction with intracellular ROS DCFH is oxidized to DCF, which is characterized by a high green fluorescence. DCFH is mostly sensitive to hydroxyl radical, peroxonitrite and H 2 O 2 [62]. CCRF-CEM cells, treated with 10 and 20 µM C 60 fullerene were irradiated at 10 J/cm 2 405 nm. The ROS production was studied after 1 and 3 h of light exposure respectively. When DCFH-DA was added to untreated cells the slight continuous increase of fluorescence intensity was detected (Fig. 4A). Neither C 60 fullerene nor light exposure alone caused significant changes in ROS generation in comparison with control, while at combined treatment with C 60 fullerene and 405 nm LED light, ROS production in CCRF-CEM cells was shown to be increased dramatically. Irradiation of cells treated with 10 µM C 60 was followed by 4-fold while treated with 20 µM C 60by 8-and 10-fold increase of ROS level at 1 and 3 h respectively (Fig. 4A). The microscopy analysis of cells, presented on Fig. 4B, confirmed DCF fluorescent measurement data, thus, indicating that photoexcitation of accumulated C 60 fullerene with 405 nm LED was followed by oxidative stress in

Apoptosis induction
The photosensitizing potential of intracellular accumulated C 60 fullerene irradiated with 405 nm LED light was further studied by evaluation of the caspase 3/7 activity and plasma membrane phosphatidylserine translocation as primary markers of cell death by apoptosis. CCRF-CEM cells were incubated for 24 h in the presence or absence of C 60 fullerene, irradiated with 405 nm LED and caspase 3/7 activity was measured during further 6 h incubation. We have shown that light irradiation alone had no effect on caspase 3/7 activity, while C 60 fullerene photoexcitation was followed by 4-fold increase of caspase 3/7 activity at 3 h in comparison with control (Fig. 5A).
Exposure of phosphatidylserine on cell surface is proved to be an "eat me" signal, which facilitates phagocytic recognition of apoptotic cells and their destruction [63]. CCRF-CEM cells, treated with C 60 and either kept in the dark or irradiated with 405 nm LED, were subjected to double staining with phosphatidylserine-binding Annexin V-FITC and DNA-binding dye propidium iodide. On FACS histograms (Fig. 5B) four populations of cells are presented according to green (annexin V-FITC) and red (PI) fluorescence intensities: viable (annexin V-FITC negative, PI negative), early apoptotic (annexin V-FITC positive, PI negative), late apoptotic (annexin V-FITC positive, PI positive) and necrotic (annexin V-FITC negative, PI positive) cells. Quantitative analysis of cell populations content at 6 and 24 h after combined treatment of CCRF-CEM cells with C 60 and 405 nm light is presented on Fig. 5C. Neither treatment with 20 µM C 60 nor 405 nm light irradiation alone had significant effect on cells distribution profiles, demonstrating a viability rate around 95%. Under combined action of C 60 fullerene and 405 nm light a time-dependent increase in the content of apoptotic CCRF-CEM cells was detected, that reached level of 18% and 50% at 6 and 24 h after light exposure respectively, compared to 4% of control cells, treated with C 60 fullerene and kept in the dark (Fig. 5B, C). The obtained data allow to conclude that toxic effect of C 60 fullerene against CCRF-CEM cells after photoexcitation is realized by apoptosis induction.

Conclusions
With the use of immunocytochemical staining and mass spectrometry we have shown that pristine C 60 fullerene was taken up by human leukemic CCRF-CEM cells from the media in a time-dependent manner, reaching a maximum intracellular level of almost 250 ng/10 6 cells at 24 h of incubation and was predominantly localized within mitochondria. The comparative analysis of a potential phototoxicity of C 60 fullerene in human leukemic cells using LEDs of different wavelengths revealed that the most favorable effect was at 405 nm, 10 J/ cm 2 . No C 60 fullerene neither 405 nm LED light alone impaired cell viability, while their combined action was followed by 46% cell viability decrease, 10-fold increase of reactive oxygen species generation, 4-fold increase of apoptosis-executive caspases 3/7 activity and plasma membrane phosphatidylserine externalization in 50% of cells. Our data suggest the use of C 60 fullerene and 405 nm high power LED as a light source for photodynamic treatment of cancer cells.