Induction of bystander effects by UVA, UVB, and UVC radiation in human fibroblasts and the implication of reactive oxygen species

Free Radical Biology and Medicine 68 (2014) 278–287

Contents lists available at ScienceDirect

Free Radical Biology and Medicine journal homepage: www.elsevier.com/locate/freeradbiomed

Original Contribution

Induction of bystander effects by UVA, UVB, and UVC radiation in human fibroblasts and the implication of reactive oxygen species Maria Widel n, Aleksandra Krzywon, Karolina Gajda, Magdalena Skonieczna, Joanna Rzeszowska-Wolny Biosystems Group, Institute of Automatic Control, Faculty of Automatics, Electronics, and Informatics, Silesian University of Technology, 44-100 Gliwice, Poland

art ic l e i nf o

a b s t r a c t

Article history: Received 21 June 2013 Received in revised form 5 December 2013 Accepted 18 December 2013 Available online 27 December 2013

Radiation-induced bystander effects are various types of responses displayed by nonirradiated cells induced by signals transmitted from neighboring irradiated cells. This phenomenon has been well studied after ionizing radiation, but data on bystander effects after UV radiation are limited and so far have been reported mainly after UVA and UVB radiation. The studies described here were aimed at comparing the responses of human dermal fibroblasts exposed directly to UV (A, B, or C wavelength range) and searching for bystander effects induced in unexposed cells using a transwell co-incubation system. Cell survival and apoptosis were used as a measure of radiation effects. Additionally, induction of senescence in UV-exposed and bystander cells was evaluated. Reactive oxygen species (ROS), superoxide radical anions, and nitric oxide inside the cells and secretion of interleukins 6 and 8 (IL-6 and IL-8) into the medium were assayed and evaluated as potential mediators of bystander effects. All three regions of ultraviolet radiation induced bystander effects in unexposed cells, as shown by a diminution of survival and an increase in apoptosis, but the pattern of response to direct exposure and the bystander effects differed depending on the UV spectrum. Although UVA and UVB were more effective than UVC in generation of apoptosis in bystander cells, UVC induced senescence both in irradiated cells and in neighbors. The level of cellular ROS increased significantly shortly after UVA and UVB exposure, suggesting that the bystander effects may be mediated by ROS generated in cells by UV radiation. Interestingly, UVC was more effective at generation of ROS in bystanders than in directly exposed cells and induced a high yield of superoxide in exposed and bystander cells, which, however, was only weakly associated with impairment of mitochondrial membrane potential. Increasing concentration of IL-6 but not IL-8 after exposure to each of the three bands of UV points to its role as a mediator in the bystander effect. Nitric oxide appeared to play a minor role as a mediator of bystander effects in our experiments. The results demonstrating an increase in intracellular oxidation, not only in directly UV-exposed but also in neighboring cells, and generation of proinflammatory cytokines, processes entailing cell damage (decreased viability, apoptosis, senescence), suggest that all bands of UV radiation carry a potential hazard for human health, not only due to direct mechanisms, but also due to bystander effects. & 2013 Elsevier Inc. All rights reserved.

Keywords: Ultraviolet radiation Bystander effect Human fibroblasts Apoptosis Premature senescence Reactive oxygen and nitrogen species Mitochondrial membrane potential Interleukins 6 and 8 Free radicals

Radiation-induced bystander effects, which appear in nontargeted cells mainly as cell-damaging events (decreased viability, reduction of clonogenic survival, induction of apoptosis, and cytogenetic damage), are well known phenomena in the case of ionizing radiation [1–4], but knowledge of bystander effects after ultraviolet radiation (UVR) is limited. UVR comprises three different wavelength bands, long-wave UVA (320–400 nm), middlewave UVB (290–320 nm), and short-wave UVC (200–290 nm) [5,6]. The main source of UVR in the environment is solar radiation,

n

Corresponding author. Fax: þ48 32 237 2127. E-mail address: [email protected] (M. Widel).

0891-5849/$ - see front matter & 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.freeradbiomed.2013.12.021

of which about 95% is UVA and
5% is UVB; UVC is almost completely absorbed in the upper part of the stratosphere [5] unless it traverses an ozone hole in this layer. Bystander effects [7–9] and related genomic instability [10,11] have been reported after UVA and UVB radiation, but very limited data are available on UVC-induced bystander effects, probably because of less interest because this wavelength does not reach the earth. The short-wave radiations (UVB under 300 nm and UVC) are especially dangerous for cells because their bands coincide with the absorption spectra of DNA, RNA, and proteins and they can damage DNA by forming cyclobutane pyrimidine dimers (CPDs) and 6-4 photoproducts (6-4 PPs), which can lead directly or indirectly to DNA strand breaks [12–14] and possibly to mutation and neoplastic transformation [15].

M. Widel et al. / Free Radical Biology and Medicine 68 (2014) 278–287

UVR is responsible for the induction and promotion of basal and squamous cell skin cancer [16] and is also an important etiological factor in malignant melanoma [17,18]. Data about the possible significance of bystander effects in UV carcinogenesis come exclusively from in vitro studies. Human keratinocytes whose precursor generations were exposed to UVA showed a reduction of clonogenic cell survival [9] and persistent genomic instability [7]. A reduction of clonogenic cell survival, genomic instability, and delayed mutation has been also observed in bystander Chinese hamster fibroblasts after exposure to UVA and UVB [10,11]. Furthermore, apoptosis was observed in bystander human keratinocytes after both UVA and UVB exposure [8], although in another study [19] no bystander effect was found after UVB radiation. Bystander effects induced by ionizing radiation [20–22] as well as ultraviolet radiation [7,11,23] are reduced in the presence of antioxidants and may be linked to oxidative stress. Each region of the ultraviolet spectrum induced the formation of 8-oxo-7,8dihydro-20 -deoxyguanosine in calf thymus DNA and in HeLa cells in a fluence-dependent manner, with singlet oxygen (1O2) playing the predominant role [24]. UVC induced DNA double-strand breaks measured by γ-H2AX and 53BP1 foci formation in bystander human fibroblasts more effectively than in irradiated cells, and this effect was mediated by reactive oxygen species (ROS) [23]. Here we report that all three UV wavelength bands, UVA, UVB, and UVC, induce bystander effects in human dermal fibroblasts with a pattern of responses that differs for each. Studies of the levels of ROS and reactive nitrogen species (RNS) and changes in mitochondrial membrane potential suggest that ROS are implicated in this induction. It is known that ROS induced by UV radiation can damage DNA and lead to various skin diseases and carcinogenesis [reviewed 25]. The high production of interleukin 6 (IL-6), a mediator of inflammation, points to its possible role in the induction of UV radiation-induced bystander effects. Although further studies are required to gain knowledge of the detailed nature of the mediators and their interactions, the present results suggest that all bands of UVR carry a potential hazard for human health not only due to direct mechanisms, but also due to bystander effects.

279

Cell survival assays The proportion of viable cells was determined using the 3-(4,5dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium salt (MTS) reduction assay (Cell Titer 96 AQueous One Solution Cell Proliferation Assay, Promega). MTS is bioreduced by the dehydrogenase enzymes present in live, metabolically active cells to the colored product formazan. The quantity of formazan measured by the absorbance is directly proportional to the number of surviving cells. The manufacturer0 s protocol was adapted for our experimental system. Briefly, cells were harvested separately from wells and inserts by trypsinization, spun down, washed in phosphate-buffered saline (PBS), and loaded with MTS reagent. The suspensions were transferred to 96-well plates and incubated for 60 min in a humidified CO2 incubator, and absorbance was measured at 490 nm using a universal plate reader (Epoch, Biotek Instruments). Survival of control, irradiated, and bystander cells is presented as mean absorbance 7SD from three independent experiments. Apoptosis and necrosis assays Apoptosis and necrosis were assessed by flow cytometry using the Dead Cell Apoptosis Kit with annexin V–FITC and propidium iodide (PI; Invitrogen). Annexin V is bound to phosphatidylserine, which is translocated from the inner to the external membrane layer at an early stage of apoptosis [27]. Cells were exposed to 20 kJ/m2 UVA, 10 kJ/m2 UVB, or 200 J/m2 UVC and co-incubated with unexposed cells for the desired time. Cells were harvested separately from wells and inserts, spun down, washed with PBS, suspended in staining buffer, and incubated for 15 min with annexin V–FITC according to the manufacturer0 s protocol. PI, which stains necrotic cells, was added and the distribution of living, apoptotic, and necrotic cells was measured by flow cytometry (BD FACSAria III) using excitation/emission maxima of 494/518 nm for annexin V–FITC and 535/617 nm for PI. Ten thousand cells were counted. Results are presented as mean fluorescence intensities7SD from three independent experiments. ROS assay

Materials and methods Cells and experimental procedure Neonatal human dermal fibroblasts (NHDF-Neo, Lonza, Poland) in early (10–13) passages were grown in Dulbecco’s modified Eagle’s medium/Nutrient Mixture F-12 Ham medium (Sigma– Aldrich), supplemented with 12% fetal bovine serum (PAA, Immuniq, Poland) and 80 mg/ml gentamycin (Krka, Poland). Irradiated and nonirradiated cells were co-incubated in six-well dishes with an insert separating the two cell populations by a 0.4-mm-pore membrane (BD Immunogen) to allow diffusion of medium components between them, as described previously [26]. About 20 h before irradiation cells were seeded into wells (1  105 cells/well in 2 ml medium) and those not to be irradiated (bystander cells) were seeded on inserts. Before irradiation the medium was removed and the cells in wells were irradiated (covers opened) at room temperature (21 1C) with various doses of UVA (365 nm), UVB (302 nm), or UVC (254 nm) generated by UV crosslinkers (CL1000 models, UVP, Upland, CA, USA). We used doses of 5–20 kJ/m2 (UVA), 2–10 kJ/m2 (UVB), and 10–200 J/m2 (UVC). Immediately after irradiation, 2 ml of fresh medium was added to the wells, and then the inserts, also with medium changed, were inserted and the cells were cocultured in a CO2 incubator (standard conditions: 5% CO2, 80% humidity, 37 1C) for the desired period.

Total cellular ROS were assayed as described elsewhere [26] using 20 ,70 -dichlorofluorescein diacetate (DCFH-DA; Sigma), which was deacetylated to nonfluorescent DCFH by intracellular esterases and then converted by cellular ROS to oxidized, fluorescent DCF. After co-incubation for 3, 6, 12, or 24 h, irradiated and bystander cells were harvested, suspended in growth medium, and loaded with DCFH-DA (final concentration 30 mM) for 30 min at 37 1C in the dark. After the cells were washed in PBS to remove extracellular dye, suspended in PBS, and incubated for 15 min on ice in the dark, ROS were determined in 10,000 cells by flow cytometry using the FITC configuration with excitation and emission wavelengths of 488 and 530 nm, respectively. Results are expressed as mean fluorescence intensities7SD from three independent experiments. Measurement of superoxide radical anions The MitoSOX red mitochondrial superoxide indicator (Invitrogen), which permeates into live cells and selectively targets mitochondria [28,29], where it is rapidly oxidized by superoxide but not by other ROS or RNS and emits red fluorescence, was used according to the manufacturer0 s protocol. Cells from wells and inserts were harvested separately, spun down, suspended in medium, loaded with MitoSOX (final concentration 5 μM), and incubated for 20 min at 37 1C. Fluorescence was measured in

280

M. Widel et al. / Free Radical Biology and Medicine 68 (2014) 278–287

10,000 cells by flow cytometry using the phycoerythrin configuration (488-nm laser line, LP mirror 566, BP filter 585/42). Results are presented as mean fluorescence intensities 7SD from three independent experiments.

Determination of IL-6 and IL-8 in culture media Two series of experiments were done, one with cells irradiated and incubated without cells in the inserts, whereas in the second configuration irradiated cells were co-incubated with bystander cells in inserts. The media were collected (in bystander configuration media from wells and inserts were collected together), centrifuged (250g, 10 min, 4 1C) to remove any cells and particles, and immediately stored at  20 1C until the cytokine assay. IL-6 and IL-8 concentrations were determined by the quantitative sandwich enzyme-linked immunosorbent assay (ELISA), using immunoassay kits (R&D Systems, supplied by Biokom). The ELISAs were performed according to the manufacturer0 s protocol using 96-well plates coated with antibodies to IL-6 or IL-8. Optical density of the standard solutions and the samples was measured at 450 nm using a microplate reader. Standard curves were generated and concentrations of interleukins in samples were calculated and expressed in pg/ml.

Measurement of mitochondrial membrane potential Cells harvested by trypsinization were washed in PBS, suspended in fresh growth medium, and loaded with 50 nM tetramethylrhodamine ethyl ester (Sigma–Aldrich), a mitochondrial membrane-specific agent [30]. After 30 min incubation at 37 1C they were washed in PBS to remove extracellular dye, resuspended in PBS, and analyzed by flow cytometry with excitation and emission at 547 and 585 nm, respectively; 10,000 cells were counted. Results from three independent experiments are presented as means7SD (in arbitrary units related to fluorescence intensity). Assay of nitric oxide

Analysis of senescent cells 0

0

The indicator 4-amino-5-methylamino-2 ,7 -difluorescein diacetate (DAF-FM diacetate; Invitrogen), which is deacetylated to DAF-FM by intracellular esterases and emits fluorescence at excitation/emission maxima of 495/515 nm when it reacts with nitric oxide (NO), was used to measure intracellular NO [31]. Cells harvested from wells and inserts were suspended in growth medium and incubated with DAF-FM (final concentration 1 μM) for 30 min at 37 1C and the fluorescence intensity was measured in 10,000 cells by flow cytometry with the same configuration as for ROS assays. Results are presented as mean fluorescence intensities 7SD from three independent experiments.

Analysis of senescent cells [32] was based on β-galactosidase expression using a Senescence Cells Histochemical Staining Kit (Sigma–Aldrich). Cells growing in wells (irradiated) and in inserts (bystander) were stained in situ after 24 h co-incubation according to the manufacturer0 s protocol. Estimation of senescence frequency was performed using an inverted microscope (Zeiss, Germany) and counting at least 1000 cells. Three independent experimental sets (wellsþ inserts) were assayed and data are presented as means7SD. Results Survival of irradiated and bystander cells

Determination of superoxide dismutase (SOD) activity in cell extracts Normal human dermal fibroblasts were used in these experiments because they represent one of the cell types present in dermal tissue, which is regularly exposed to UV radiation. To obtain comparable results for three bands of UV radiation that have different energies, we chose different dose ranges of 5–20 kJ/m2 for UVA, 2–10 kJ/m2 for UVB, and 50–200 J/m2 for UVC, partly based on doses used in published studies concerning genomic instability [10,11] and DNA damage [33,34]. The wavelength range of UVA used was lower than that used in other studies of bystander effects and genomic instability [e.g., 7–9]. The survival of control, UV-exposed, and bystander fibroblasts is presented in Figs. 1–3. The survival of control cells growing in wells and in inserts was similar, and therefore common control curves are presented. The measurement of survival was started from 15 min to allow for at least a short co-incubation of irradiated with bystander cells.

SOD activity was determined by a colorimetric method using a High Throughput Superoxide Dismutase Assay Kit (Trevigen, supplied by Biokom, Poland). Briefly, cells were harvested separately from wells and inserts by trypsinization, washed with cold PBS, and centrifuged at 250g for 10 min at 4 1C. The pellets of cells were lysed on ice for 30 min in cell extraction buffer contained in the kit and centrifuged at 10,000g for 10 min at 4 1C. Supernatants were transferred to tubes precooled at  80 1C and frozen at 80 1C until SOD measurement. Measurement was performed according to the protocol provided by the manufacturer. SOD activity per microgram of protein was calculated and data are presented as percentage change in relation to the control taken as 100% at a given point in time; the values for the control samples were in the range from 0.04 to 0.065 U/μg protein.

UVA 5 kJ/m2

1

10 kJ/m2

20 kJ/m2

Absorbance

0.8 0.6 0.4 0.2 0

15 min

24 h

48 h

72 h

15 min

CT

IR

24 h

48 h

BY 72 h

15 min

24 h

48 h

72 h

Incubation time Fig. 1. Survival of NHDF fibroblasts estimated by MTS assay; CT, control cells; IR, cells exposed to various doses of UVA; and BY, unexposed bystander cells co-incubated with them. Results are expressed as mean absorbance 7 SD from three independent experiments.

M. Widel et al. / Free Radical Biology and Medicine 68 (2014) 278–287

UVB 2 kJ/m2

1

281

10 kJ/m2

5 kJ/m2

Absorbance

0.8 0.6 0.4 0.2 0

15 min

24 h

48 h

72 h

15 min

CT

IR

24 h

48 h

BY 72 h

15 min

24 h

48 h

72 h

Incubation time Fig. 2. Survival of NHDF fibroblasts estimated by MTS assay; CT, control cells; IR, cells exposed to various doses of UVB; and BY, unexposed bystander cells co-incubated with them. Results are expressed as mean absorbance 7SD from three independent experiments.

UVC 50 J/m2

1

200 J/m2

100 J/m2

Absorbance

0.8 0.6 0.4 0.2 0

15 min

24 h

48 h

72 h

15 min

CT

IR

BY

24 h

48 h

72 h

15 min

24 h

48 h

72 h

Incubation time Fig. 3. Survival of NHDF fibroblasts estimated by MTS assay; CT, control cells; IR, cells exposed to various doses of UVC; and BY, unexposed bystander cells co-incubated with them. Results are expressed as mean absorbance 7SD from three independent experiments.

UVR-exposed cells showed a decline in survival with time and with increasing dose, which was seen clearly after 72 h for all three wavelengths, whereas survival of control cells increased with time because of proliferation. The highest survival at 72 h was observed after exposure to UVA (
50% after 20 kJ/m2, Fig. 1), whereas for UVB (10 kJ/m2) viability dropped to
10% (Fig. 2) and for UVC (200 J/m2) to
20% (Fig. 3). Bystanders of UVA-exposed cells show a diminution of survival with increasing time compared to controls, which at 48 h was even more marked than for irradiated cells (Fig. 1). In the cases of UVB (Fig. 2) and UVC (Fig. 3), the survival of bystander cells measured at 72 h dropped after the highest doses to about 60 and 50% of control, respectively. Apoptosis and necrosis To obtain reliable quantitation of apoptotic cells and of ROS and RNS in cells, the highest UVR doses and the time span of 24 h were used. The frequencies of apoptosis in control cells growing in wells and in inserts were comparable (
4%), and a common control is shown (Fig. 4). UVA (20 kJ/m2) caused a statistically significant approximately twofold increase in the frequency of apoptotic cells, which persisted at a slightly lower level up to 12 h. Apoptosis also appeared in UVA bystander cells but with some delay, reaching a greater than twofold increase by 6 h (Fig. 4a). UVB (10 kJ/m2) induced a significant greater than twofold increase in apoptosis by 12 h and an approximately fivefold increase by 24 h (Fig. 4b). In UVB bystander cells the apoptosis frequency increased slightly, although significantly, after 3 h and persisted at a comparable level up to 24 h (Fig. 4b). UVC (200 J/m2) induced apoptosis in irradiated cells with a frequency nearly twice

than in control cells at 3 h, which slightly decreased at 6 h and dropped almost to the control level after 12 h. However, in UVCirradiated cells apoptosis increased again after 24 h, probably as a delayed consequence of the ROS elevation seen at 12 h (Fig. 4c). In bystander cells the apoptosis frequency did not change significantly except for a small but significant increase at 6 h (Fig. 4c). The insets in Fig. 4 show the cell survival measured by MTS assay at the same time points at which apoptosis was assessed. For UVA, the data are quite consistent with the data for apoptosis at 6 and 12 h; the increase in apoptosis, particularly in bystander cells, is accompanied by a considerable reduction in survival (Fig. 4a). The good agreement between apoptosis and survival can also be seen at 12 and 24 h for UVB and UVC (Figs. 4b and c). Neither UVA nor UVB induced necrosis within the experimental period, but interestingly UVC (200 J/m2) induced a low frequency (
0.5%, not significantly different from the control) after 24 h (data not shown). Cell senescence Recently, attention has been paid to stress-induced premature senescence [35–38]. Hallmarks of senescent cells include an essentially irreversible growth arrest and expression of senescence-associated β-galactosidase and p16INK4a. We evaluated the expression of β-galactosidase, the known marker of senescence [32], after staining the cells in situ in wells and inserts after 24 h co-incubation. The microscopic analyses indicated a large difference in the efficiency of senescence induction by various UV bands. It was striking that UVC induced senescence very effectively in both irradiated and

282

M. Widel et al. / Free Radical Biology and Medicine 68 (2014) 278–287

UVA Ct

Apoptotic cells (% total)

25

UVB IR

UVC BY

20

15

10

5

0

3

6

12

24

3

6

12

24

3

6

12

24

Incubation time (h) Fig. 4. Frequency of apoptotic cells in control cultures (Ct), cells exposed to UV (IR), and unexposed bystander cells co-incubated with cells exposed to UV (BY). Doses were UVA, 20 kJ/m2; UVB, 10 kJ/m2; and UVC, 200 J/m2. Data are means 7 SD from three independent experiments. nSignificant difference from the control level of apoptotic cells (po 0.05, Student’s t test). The insets show survival assay data obtained from MTS assays at the same time points and doses used for apoptosis assays.

Percentage of senescent cells

10 8 6

interesting that UVC increased the level in bystander cells significantly, especially after 12 h, although it did not generate such a highly significant amount of ROS in irradiated cells (Fig. 6c). The ROS level in control cells also increased after 24 h. We suppose that undistorted proliferation of cells in the wells and the inserts leads to faster acidification of medium and changes in the microenvironment, which result in increased ROS generation.

4

Superoxide radicals 2 0

Fig. 5. Percentage of senescent cells in wells and inserts after 24 h co-incubation assessed on the basis of expression of β-galactosidase. Data are means 7 SD from three independent experimental sets (wellsþinserts) in which 1000 cells were counted. nStatistically significant difference from the control at p o 0.05 (Student’s t test). UV doses were the same as those in Fig. 4.

neighboring cells (Fig. 5). In experiments using UVB, senescence was induced in bystander cells, whereas in directly exposed cells an extremely high yield of apoptosis was seen at that time (Fig. 4b). In contrast, UVA generated senescence in directly exposed cells, but not in bystanders. These differences suggest that the signals secreted by cells in response to different UV bands differ, although their nature and interactions remain to be elucidated. ROS levels In cells irradiated with UVA (20 kJ/m2) the level of total cellular ROS increased to almost threefold that in control cells by 3 h (Fig. 6a), and in bystander cells the increase was approximately twofold by 3 and 6 h. After 12 h the ROS level in both irradiated and bystander cells decreased to the control level, followed by a second but statistically insignificant increase after 24 h (Fig. 6a). UVB (10 kJ/m2) was somewhat less effective than UVA (20 kJ/m2); the yield of ROS reached almost twofold the level in control cells at 3 h and was even higher in bystander cells (Fig. 6b). It is

UVA (20 kJ/m2) generated a significant level of superoxide at 3 and 6 h in irradiated cells (Fig. 7a), whereas after UVB exposure (10 kJ/m2) a significant increase was observed after 24 h (Fig. 7b). The highest level of superoxide was generated by UVC (200 J/m2); it reached
5-fold the control level after 3 h and then rapidly declined (Fig. 7c), and a relatively high level of superoxide (
2.5fold the control) was measured in bystander cells at 3 h. Mitochondrial membrane potential UVA (20 kJ/m2) led to a significant decrease in mitochondrial membrane potential in irradiated cells between 3 and 12 h (Fig. 8a). In bystander cells the potential increased by 12 h and then decreased to under the control level by 24 h. At this time point the mitochondrial potential in UVA-exposed cells increased, but was still significantly lower than in control (Fig. 8a). After UVB (10 kJ/m2) the potential showed a rather stable level in exposed cells and some fluctuation in bystanders, e.g., a slight increase at 3 h and a decline to below the control level at 6 h (Fig. 8b). In UVC-exposed cells the membrane potential had increased at 12 h but was comparable to that in control cells at 24 h, whereas in bystanders the mitochondrial membrane potential remained constant (Fig. 8c). Nitric oxide The level of cellular NO did not change markedly after exposure to any of the wavelength bands of UVR during the whole range of time, with the exception of 12 h, at which it was reduced in bystander cells after UVA and UVB, and a significant reduction in UVC-exposed cells and their bystanders was seen (Fig. 9).

M. Widel et al. / Free Radical Biology and Medicine 68 (2014) 278–287

UVA

2500

283

UVB IR

Ct

UVC BY

ROS level (a.u.)

2000 1500 1000 500 0

3

6

12

24

3

6

12

24

3

6

12

24

Incubation time (h) Fig. 6. Intracellular ROS levels in control cells (Ct), cells exposed to UVR (IR), and unexposed bystander cells co-incubated with cells exposed to UVR (BY). Data are means7SD in arbitrary units (a.u.) from three independent experiments. nSignificant difference from the control level (po0.05, Student’s t test). UV doses were the same as those shown in Fig. 4.

UVA

2000

UVB

Superoxide level (a.u)

Ct

IR

UVC BY

1500

1000

500

0

3

6

12

24

3

6

12

24

3

6

12

24

Incubation time (h) Fig. 7. Superoxide radicals in control cells (Ct), cells exposed to UVR (IR), and unexposed bystander cells co-incubated with cells exposed to UVR (BY). Data are means 7 SD in arbitrary units (a.u.) from three independent experiments. nSignificant difference from the control level of superoxide radicals (p o 0.05, Student’s t test). UV doses were the same as those shown in Fig. 4.

UVA

1500

UVB IR

BY

mit (fluorescence

a.u.)

Ct

UVC

1000

500

0

3

6

12

24

3

6

12

24

3

6

12

24

Incubation time (h) Fig. 8. Mitochondrial membrane potential expressed as fluorescence signal in arbitrary units in control cells (Ct), cells exposed to UVR (IR), and unexposed bystander cells co-incubated with cells exposed to UVR (BY). Data are means 7 SD from three independent experiments. nSignificant difference from the control level (p o 0.05, Student’s t test). UV doses were the same as those in Fig. 4.

SOD activity To check whether the generation of reactive oxygen species, including superoxide, entails changes in the activity of one of the

key antioxidant enzymes, superoxide dismutase, SOD activity measurements were done in cell extracts soon after irradiation and after 3, 12, and 24 h of exposure. Whereas SOD activity in control cells was at a fairly even level (0.04–0.06 U/μg protein),

284

M. Widel et al. / Free Radical Biology and Medicine 68 (2014) 278–287

UVA

900

UVB IR

CT Nitric oxide (a.u.)

UVC BY

600

300

0

3

6

12

24

3

6

12

24

3

6

12

24

Incubation time (h) Fig. 9. Nitric oxide levels in control cells (CT), cells exposed to UVR (IR), and unexposed bystander cells co-incubated with cells exposed to UVR (BY). Data are means7SD in arbitrary units (a.u.) from three independent experiments. nSignificant difference from the level in control cells (po0.05, Student’s t test). UV doses were the same as those in Fig. 4.

UVA

800

UVB Ct

Relative SOD activity

700

IR

UVC BY

600 500 400 300 200 100 0

15 min

3h

12 h

24 h

15 min

3h

12 h

24 h

15 min

3h

12 h

24 h

Incubation time Fig. 10. The relative activity of superoxide dismutase in cell extracts, expressed as a percentage of controls at each time point (the measured values for control samples were in the range from 0.04 to 0.065 U/μg protein). Data are the means7 SD from measurement of two independent samples performed in duplicate for each time point. n Statistical significance at p o0.05 (Student’s t test). UV doses were the same as those in Fig. 4.

exposure of cells to all UV bands caused upregulation of this enzyme practically immediately after irradiation, but to varying degrees (Fig. 10). UVA at a dose of 20 kJ/m2 was most effective, and after 15 min a twofold increase in activity was observed, which after 3 h reached a level more than five times higher than the control (Fig. 10a). At this time SOD activity was also doubled in bystander cells, but later the activity of SOD declined even to under the control level in exposed and bystander cells. UVB also caused upregulation of SOD very quickly in irradiated and bystander cells, which peaked at 3 h (Fig. 10b). In UVC-exposed cells a slight increase in SOD activity was seen between 15 min and 12 h, and a significant increase to about threefold of the control by 24 h. The increase in SOD activity in bystander cells was also shifted in time to 12 and 24 h (Fig. 10c).

but although UVA induced IL-6 to comparable levels in both systems, the level of IL-6 in UVB and UVC co-incubation systems increased significantly, which indicates that IL-6 must be also generated by bystander cells. A significant decrease in the level of IL-6 by 24 h was associated with an increase in apoptosis at this time. In the case of IL-8, production of this cytokine over the control level was observed only for UVA in an experimental system without co-incubation (Fig. 11c). In UVB and UVC the IL-8 concentration in culture medium was permanently below the control level. However, in the co-incubation system the IL-8 concentration in the medium was lower than in the control for all three bands of UV and this reduction was statistically significant between 15 min and 6 h.

Generation of interleukins

Discussion

Measurement of IL-6 and IL-8 cytokines as potential mediators of bystander effect was performed in culture medium collected from irradiated cells incubated alone or co-incubated with bystander cells. In Fig. 11 we see that the concentration of IL-6 increased immediately after irradiation with all UV wavelengths under both experimental conditions and remained significantly higher compared to the control until 6 h. UVA seemed to be most effective,

In this study we investigated the occurrence of bystander effects in normal human fibroblasts after exposure to three different UV bands and we looked for differences and similarities in the effectiveness at reducing the viability of cells, inducing apoptosis and senescence, and generation of putative mediators. Normal human fibroblasts exposed to any of the three bands of ultraviolet radiation, UVA (365 nm), UVB (302 nm), and UVC

M. Widel et al. / Free Radical Biology and Medicine 68 (2014) 278–287

IL-6 (pg/ml)

800

400 200 0

15 min

1500

3h

6h

12 h

24 h

15 min

900

3h

6h

12 h

24 h

6h

12 h

24 h

Ct + BY UVA + BY UVB + BY UVC + BY

Ct UVA UVB UVC

1200

IL-8 (pg/ml)

Ct + BY UVA + BY UVB + BY UVC + BY

Ct UVA UVB UVC

600

285

600 300

0

15 min

3h

6h

12 h

24 h

Incubation time

15 min

3h

Incubation time

Fig. 11. Concentration of IL-6 (top) and IL-8 (bottom) in culture medium collected from control and UV-exposed cells. (a, c) Medium from cells incubated without neighbors in inserts; (b, d) medium from cells co-incubated with cells in inserts. UV doses were the same as those shown in Fig. 4. Data are means 7 SD from two independent experiments performed in duplicate. A significant increase in IL-6 concentration relative to control was measured between 15 min and 6 h in both systems (p o0.05, Student’s t test).

(254 nm), induced bystander effects manifested as a reduction in cell survival and an increased frequency of apoptosis in cells of the same type separated from them by a 0.4-mm-pore membrane, which allowed diffusion of medium components. UVA appeared to be relatively the most effective, even at doses well below the minimal erythema dose (MED; 1 MED corresponds to 750 kJ/m2 [39]). The highest dose of UVA we used, 20 kJ/m2, reducing cell survival and inducing apoptosis in directly exposed and bystander fibroblasts, was low compared to that (100 kJ/m2) which reduced clonogenic cell survival in bystander human keratinocytes and fibroblasts [19]. UVB at 10 kJ/m2, reducing survival even more efficiently, was equivalent to 1 MED [39]. This dose was higher than that used in a similar co-incubation system [19] in which 400 J/m2 did not induce a bystander effect, but it corresponds to those used in other studies of bystander effect [11] and chromosomal damage [34]. Published data on bystander effect induction after UVB exposure are not consistent and may depend on the cell type, the evaluation system, the dose, and the endpoints; for example, an increased apoptosis frequency was reported in bystander keratinocytes exposed to 300 J/m2 UVB [8], a dose lower than that used in [19], which did not induce a bystander effect. In our study, UVB at 10 kJ/m2 was extremely effective at inducing apoptosis in irradiated cells and showed a similar effect in bystander cells (Fig. 4b). UVC at 200 J/m2 reduced the survival of bystander cells even more effectively than UVB at 10 kJ/m2 (compare Figs. 2 and 3) and was effective in apoptosis induction in irradiated cells, but not in bystander cells (Fig. 4c). At the same time UVC radiation induced efficiently senescence in the irradiated and neighboring cells (Fig. 5). This increase in senescence appears to be associated with increased levels of ROS, but secretion of IL-6 by senescent cells seems probable. The mechanism of action of UVR on cells is different for the three wavelength bands [14;reviewed in 40]. UVA acts mainly through generation of ROS such as singlet oxygen and hydroxyl radicals, which can induce oxidative damage to DNA, proteins, and lipids [7,9,14]; UVB also generates ROS [8,11], but UVC rather

damages DNA directly by forming CPDs and 6-4 PPs [14]. Our observations of ROS generation in cells exposed to UVA and UVB, but less effectively by UVC, are consistent with this picture. Different types of cellular damage induced by different UV bands may start the cellular response from different signaling pathways, and because of that one could expect responses varying in efficiency and kinetics. However, many characteristics of cellular response to different UV bands are very similar. For all three bands of UV wavelengths bystander effects appear after relatively low doses, all induce signaling to nonirradiated neighbors and a decrease in their survival, all induce SOD activity that correlates with the increase in ROS. Several cytokines may be implicated in the bystander effect, including transforming growth factor β and tumor necrosis factor α [23]. The proinflammatory cytokines IL-6 and IL-8 have been also proposed as signaling molecules in bystander effects because they were detected in the medium after ionizing radiation [41,42] and UVB radiation [43]. In our experiments we observed an increase in IL-6 in the medium collected from cells exposed to all three UV bands or in medium collected from irradiated cells co-incubated with unexposed cells. The level of IL-6 in the medium obtained in experiments with coincubation was higher than in medium collected from irradiated cells incubated alone (Fig. 11); thus IL-6 must also be generated by the nonirradiated cells, especially in UVB and UVC experiments. Possibly senescent cells participate because an increase in IL-6 observed in the co-incubation system after UVC and UVB exposure was associated with senescence induction in bystander cells (compare Fig. 5 vs Fig. 11b). Elevation of IL-6 and IL-8 in conditioned medium has been observed for various cancer cells after ionizing radiation, and their action was associated with different profiles of bystander cell survival (a decrease or an increase) depending on the cell line [42]. This suggests that the generation, as well as the reception, of these cytokines is highly cell-type specific. The profile of secretion of IL-8 with its decrease to below control after irradiation of NHDF fibroblasts by UVB and UVC was

286

M. Widel et al. / Free Radical Biology and Medicine 68 (2014) 278–287

rather unexpected as this cytokine was proposed as one of the factors participating in bystander signaling after ionizing radiation [42] and we observed an increase in IL-8 release after UVA irradiation (Fig. 11c). The presence of nonirradiated fibroblasts in the neighborhood of irradiated ones caused inhibition of IL-8 release to below that observed in medium collected from control cells. The doses used for UVB and UVC may not stimulate IL-8 production because its production is not simply dependent on the dose, as was observed in the case of ionizing radiation [42]. However, a decrease in IL-8 under co-incubation conditions does not necessarily mean that this cytokine is not involved in bystander signaling; it may also suggest an induction of inhibitory factors in paracrine regulation mechanisms or maybe its binding by receptors on bystander cells. Expression of IL-8 is regulated by the NF-κB signaling pathway and reactive oxygen species are necessary for activation of transcription complex [44]. Thus, it is possible that an interplay between induction of ROS and intracellular antioxidants is crucial for the observed effects. All three UV wavelength bands induced reactive oxygen species and significantly increased the levels of superoxide dismutase. Superoxide radical anions were generated at a higher level by UVC compared with UVA and UVB radiation (Fig. 7). Other studies suggest that superoxide radicals formed by respiratory complex I in mitochondria, which can be converted to hydrogen peroxide by superoxide dismutase, and then further to water by catalase, or to the hydroxyl radical in the presence of transition metals [45–47], are involved in bystander effects and genomic instability because catalase inhibits these effects [7,11]. After exposure of NHDF fibroblasts to UVA, the rapid increase in the frequency of apoptosis was paralleled by increases in ROS and superoxide radicals, which activated SOD enzyme (compare Fig. 4 vs Figs. 6, 7, and 10). These increases were not accompanied by an increase in mitochondrial membrane potential but rather by a decrease (Fig. 8a), which reflects mitochondrial membrane depolarization. This, however, did not appear to reflect irreversible damage to mitochondria because after 24 h apoptosis had returned to the control level and cell survival was normal. After exposure of fibroblasts to UVB, a large increase in apoptosis at 24 h was correlated with an increase in superoxide radicals (Fig. 4b vs Fig. 7b) and a slight decrease in membrane potential (Fig. 8b). However, in some bystander cells (UVA and UVB) and in UVC-exposed cells a significant increase in mitochondrial membrane potential (ψmit) was noticed. Different cells may respond to UV radiation in various ways. Reduction of the mitochondrial membrane potential in bystander cells has been associated with increased production of ROS in human melanocytes after exposure to UVA, but not UVB [48]. Also, human keratinocytes responded, by increasing ROS and reducing mitochondrial membrane potential, to bystander signals contained in medium collected from cells exposed to γ rays [49]. In both cases a reduction in ψmit was associated with an influx of calcium ions into the cells. In another study, ionizing-radiation-induced increase in ROS in lymphoma cells was accompanied by an increase in mitochondrial membrane potential and apoptosis [50]. Furthermore, impairment of mitochondrial membrane may appear as initial membrane hyperpolarization (increase in ψmit) followed by depolarization (decrease in ψmit) [51], which probably is the case in bystander cells in our UVA and UVB experiments (Fig. 8). We can also assume that different responses of the mitochondrial membrane to stress induced by UV radiation result from different levels and/or natures of signaling molecules at a given time generated by different bands of UV. However, more studies are required to investigate the underlying mechanisms. A highly significant increase in the level of ROS in UVCirradiated and bystander cells appearing by 12 h was accompanied by an increase in mitochondrial membrane potential. At the same time, a significant decrease in nitric oxide levels occurred

in UVC-irradiated and bystander cells. These changes (reduction in NO production, increase in ROS, and increase in ψmit) seen at 12 h might be associated with cell cycle inhibition, especially as an arrest in G1 was seen after UVC radiation (G2/M:G1 ¼0,18 at 12 h vs
0.5 at the start of treatment, data not presented). However, it appears that NO plays a minor role in the action of all UV radiation bands and in bystander effects in our experimental system, agreeing with studies of effects of UVB in human keratinocytes in which NO levels were only weakly associated with apoptosis and mitochondrial dysfunction [52].

Conclusions In conclusion, our results demonstrate that all three wavelength bands of UV radiation caused a reduced survival and an increased frequency of apoptosis in nonirradiated bystander human dermal fibroblasts co-incubated with irradiated fibroblasts, although with varying efficiency and kinetics. In addition, UVR induced premature senescence, in particular in UVC-exposed and bystander cells. Our results are consistent with the idea that these bystander effects are caused by an increase in the level of cellular ROS in irradiated cells. Increased secretion of interleukin-6 suggests its role as a molecular bystander signal released by irradiated cells, but mutual signaling between irradiated and bystander cells modulates this secretion. Further studies are required to understand the nature of the mediators of these UV-induced bystander effects, but nonetheless the present results showing that all three bands efficiently induced a damaging bystander effect through generation of ROS and proinflammatory cytokine suggest that UVR carries a potential health risk not only due to direct mechanisms, but also due to the bystander phenomenon.

Acknowledgments This work was supported by Grants NN 518 497 639 from the Polish Ministry of Science and Higher Education and DEC-2012/05/ B/ST6/03472 from the National Center of Science. Ronald Hancock (Laval University, Laval, QC, Canada) is acknowledged for critically reading and editing the English of the manuscript. References [1] Mothersill, C.; Seymour, C. B. Radiation-induced bystander effects: past history and future directions. Radiat. Res. 155:759–767; 2001. [2] Prise, K. M.; Folkard, M.; Michael, B. D. Bystander responses induced by low LET radiation. Oncogene 22:7043–7049; 2003. [3] Rzeszowska-Wolny, J.; Przybyszewski, W. M.; Widel, M. Ionizing radiationinduced bystander effect, potential targets for modulation of radiotherapy. Eur. J. Pharmacol. 625:156–167; 2009. [4] Widel, M.; Przybyszewski, W. M.; Rzeszowska-Wolny, J. Radiation-induced bystander effect: the important part of ionizing radiation response. Potential clinical implications. Postepy Hig. Med. Dosw. 63:377–388; 2009. [5] Hockberger, P. E. A history of ultraviolet photobiology for humans, animal and microorganisms. Photochem. Photobiol. 76:561–579; 2002. [6] Batista, L. F. Z.; Kaina, B.; Meneghini, R.; Menck, C. F. How DNA lesions are turned into powerful killing structures: insights from UV-induced apoptosis. Mutat. Res. 681:197–208; 2009. [7] Phillipson, R. P.; Tobi, S. E.; Morris, J. A.; McMillan, T. J. UV-A induces persistent genomic instability in human keratinocytes through an oxidative stress mechanism. Free Radic. Biol. Med. 32:474–480; 2002. [8] Banerjee, G.; Gupta, N.; Kapoor, A.; Raman, G. UV induced bystander signaling leading to apoptosis. Cancer Lett. 223:275–284; 2005. [9] McMillan, T. J.; Leatherman, E.; Ridley, A.; Shorrocks, J.; Tobi, S. E.; Whiteside, J. R. Cellular effects of long wavelength UV light (UVA) in mammalian cells. J. Pharm. Pharmacol. 60:969–976; 2008. [10] Dahle, J.; Kvam, E. Induction of delayed mutations and chromosomal instability in fibroblasts after UVA-, UVB-, and X-radiation. Cancer Res 63:1464–1469; 2003. [11] Dahle, J.; Kvam, E.; Stokke, T. Bystander effects in UV-induced genomic instability: antioxidants inhibit delayed mutagenesis induced by ultraviolet A and B radiation. J. Carcinog 4:11–19; 2005.

M. Widel et al. / Free Radical Biology and Medicine 68 (2014) 278–287

[12] Wang, T. C.; Smith, K. C. Postreplication repair in ultraviolet-irradiated human fibroblasts: formation and repair of DNA double-strand breaks. Carcinogenesis 7:389–392; 1986. [13] Slieman, T. A.; Nicholson, W. L. Artificial and solar UV radiation induces strand breaks and cyclobutane pyrimidine dimers in Bacillus subtilis spore DNA. Appl. Environ. Microbiol. 66:199–205; 2000. [14] Rastogi, R. P.; Richa; Kumar, A.; Tyagi, B.; Sinha, R. P. Molecular mechanisms of ultraviolet radiation-induced DNA damage and repair. J. Nucleic Acids 2010:592980; 2010. [15] Rünger, T. M. How different wavelengths of the ultraviolet spectrum contribute to skin carcinogenesis: the role of cellular damage responses. J. Invest. Dermatol. 127:2103–2105; 2007. [16] De Gruijl, F. R.; Sterenborg, H. J.; Forbes, D.; Davies, R. E.; Cole, C.; Kelfkens, G.; van Weelden, H.; Slaper, H.; Van der Leun, J. C. Wavelength dependence of skin cancer induction by ultraviolet irradiation of albino hairless mice. Cancer Res. 53:53–60; 1993. [17] Setlow, R. B.; Grist, E.; Thompson, K.; Woodhead, A. D. Wavelengths effective in induction of malignant melanoma. Proc. Natl. Acad. Sci. USA 90:6666–6670; 1993. [18] Wolnicka-Głubisz, A.; Płonka, M. Role of UV irradiation in aetiopathogenesis of malignant melanoma. Współczesna Onkol 11:419–429; 2007. [19] Whiteside, J. R.; McMillian, T. J. A bystander effect is induced in human cells treated with UVA radiation but not UVB radiation. Radiat. Res 171:204–221; 2009. [20] Azzam, E. I.; de Toledo, S. M.; Little, J. B. Oxidative metabolism, gap junctions and the ionizing radiation-induced bystander effect. Oncogene 22:7050–7057; 2003. [21] Harada, T.; Kashino, G.; Suzuki, K.; Matsuda, N.; Kodama, S.; Watanabe, M. Different involvement of radical species in irradiated and bystander cells. Int. J. Radiat. Biol. 84:809–814; 2008. [22] Ermakov, A. V.; Konkova, M. S.; Kostyuk, S. V.; Egolina, N. A.; Efremova, L. V.; Veiko, N. N. Oxidative stress as a significant factor for development of an adaptive response in irradiated and nonirradiated human lymphocytes after inducing the bystander effect by low dose X-radiation. Mutat. Res. 669: 155–161; 2009. [23] Dickey, J. S.; Baird, B. J.; Redon, C. E.; Sokolov, M. V.; Sedelnikova, O. A.; Bonner, W. M. Intercellular communication of cellular stress monitored by γ-H2AX induction. Carcinogenesis 30:1686–1695; 2009. [24] Zhang, X.; Rosenstein, B. S.; Wang, Y.; Lebwohl, M.; Wei, H. Identification of possible reactive oxygen species involved in ultraviolet radiation-induced oxidative DNA damage. Free Radic. Biol. Med. 23:980–995; 1997. [25] Bickers, D. R.; Athar, M. Oxidative stress in the pathogenesis of skin disease. J. Invest. Dermatol. 126:2565–2575; 2006. [26] Widel, M.; Przybyszewski, W. M.; Cieslar-Pobuda, A.; Saenko, Y. V.; Rzeszowska-Wolny, J. Bystander normal human fibroblasts reduce damage response in radiation targeted cancer cells through intercellular ROS level modulation. Mutat. Res. 731:117–124; 2012. [27] Zhang, G.; Gurtu, V.; Kain, S. R.; Yan, G. Early detection of apoptosis using a fluorescent conjugate of annexin V. BioTechniques 23:525–531; 1997. [28] Robinson, K. M.; Janes, M. S.; Pehar, M.; Monette, J. S.; Ross, M. F.; Hagen, T. M.; Murphy, M. P.; Beckman, J. S. Selective fluorescent imaging of superoxide in vivo using ethidium-based probes. Proc. Natl. Acad. Sci. USA 103:15038–15043; 2006. [29] Robinson, K. M.; Janes, M. S.; Beckman, J. S. The selective detection of mitochondrial superoxide by live cell imaging. Nat. Protoc. 3:941–947; 2008. [30] Scaduto Jr R. C.; Grotyohann, L. W. Measurement of mitochondrial membrane potential using fluorescent rhodamine derivatives. Biophys. J. 76:469–477; 1999. [31] Kojima, H.; Nakatsubo, N.; Kikuchi, K.; Kawahara, S.; Kirino, Y.; Nagoshi, H.; Hirata, Y.; Nagano, T. Detection and imaging of nitric oxide with novel fluorescent indicators: diaminofluoresceins. Anal. Chem. 70:2446–2453; 1998.

287

[32] Dimri, G. P.; Lee, X.; Basile, G.; Acosta, M.; Scott, G.; Roskelley, C.; Medrano, E. E.; Linskens, M.; Rubelj, I.; Pereira-Smith, O. A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proc. Natl. Acad. Sci. USA 92:9363–9367; 1995. [33] Schuch, A. P.; Menck, C. F. M. The genotoxic effects of DNA lesions induced by artificial UV-radiation and sunlight. J. Photochem. Photobiol. B Biol. 99:111–116; 2010. [34] Emri, G.; Wenczl, E.; Van Erp, P.; Jans, J.; Roza, L.; Horkay, I.; Schothorst, A. A. Low doses of UVB or UVA induce chromosomal aberrations in cultured human skin cells. J. Invest. Dermatol. 115:435–440; 2000. [35] Suzuki, M.; Boothman, D. A. Stress-induced premature senescence (SIPS)— influence of SIPS on radiotherapy. J. Radiat. Res. 49:105–112; 2008. [36] Sabin, R. J.; Anderson, R. M. Cellular senescence—its role in cancer and the response to ionizing radiation. Genome Integr 2:7; 2011. [37] Rodier, F.; Campisi, J. Four faces of cellular senescence. J. Cell Biol. 192: 547–556; 2011. [38] Nelson, G.; Wordsworth, J.; Wang, C.; Jurk, D.; Lawless, C.; Martin-Ruiz, C.; von Zglinicki, T. A senescent cell bystander effect: senescence-induced senescence. Aging Cell 11:345–349; 2012. [39] Parrish, J. A.; Jaenicke, K. F.; Anderson, R. R. Erythema and melanogenesis action spectra of normal human skin. Photochem. Photobiol. 36:187–191; 1982. [40] Widel, M. Bystander effect induced by UV radiation: why should we be interested? Postepy Hig. Med. Dosw. 66:828–837; 2012. [41] Facoetti, A.; Mariotti, L.; Ballarini, F.; Nano, R.; Pasi, F.; Ranza, E.; Ottolenghi, A. Experimental and theoretical analysis of cytokine release for the study of radiation-induced bystander effect. Int. J. Radiat. Biol. 85:690–699; 2009. [42] Desai, S.; Kumar, A.; Laskar, S.; Pandey, B. N. Cytokine profile of conditioned medium from human tumor cell lines after acute and fractionated doses of gamma radiation and its effect on survival of bystander tumor cells. Cytokine 61:54–62; 2013. [43] Yoshizumi, M.; Nakamura, T.; Kato, M.; Ishioka, T.; Kozawa, K.; Wakamatsu, K.; Kimura, H. Release of cytokines/chemokines and cell death in UVB-irradiated human keratinocytes, HaCaT. Cell. Biol. Int. 32:1405–1411; 2008. [44] Vlahopoulos, S.; Boldogh, I.; Casola, A.; Brasier, A. R. Nuclear factor-κBdependent induction of intrleukin-8 gene expression by tumor necrosis factor? Evidence for an antioxidant sensitive activating pathway distinct from nuclear translocation Blood 94:1878–1889; 1999. [45] Kowaltowski, A. J.; de Souza-Pinto, N. C.; Castilho, R. F.; Vercesi, A. E. Mitochondria and reactive oxygen species. Free Radic. Biol. Med. 47:333–343; 2009. [46] Murphy, M. P. How mitochondria produce reactive oxygen species. Biochem. J. 417:1–13; 2009. [47] Chen, Q.; Vazquez, E. J.; Moghaddas, S.; Hoppel, C. L.; Lesnefsky, E. J. Production of reactive oxygen species by mitochondria: central role of complex III. J. Biol. Chem. 278:36027–36031; 2003. [48] Nishiura, H.; Kumagai, J.; Kashino, G.; Okada, T.; Tano, K.; Watanabe, M. The bystander effect is a novel mechanism of UVA-induced melanogenesis. Photochem. Photobiol. 88:389–397; 2012. [49] Lyng, F. M.; Seymour, C. B.; Mothersill, C. Production of a signal by irradiated cells which leads to a response in unirradiated cells characteristic of initiation of apoptosis. Br. J. Cancer 83:1223–1230; 2000. [50] Saenko, Y.; Cieslar-Pobuda, A.; Skonieczna, M.; Rzeszowska-Wolny, J. Changes of reactive oxygen and nitrogen species and mitochondrial functioning in human K562 and HL60 cells exposed to ionizing radiation. Radiat. Res. 180:360–366; 2013. [51] Russell, J.; Golovoy, D.; Vincent, A. M.; Mahendru, P.; Olzmann, J. A.; Mentzer, A.; Feldman, E. L. High glucose-induced oxidative stress and mitochondrial dysfunction in neurons. FASEB J. 16:1738–1748; 2002. [52] Paz, M. L.; Maglio, D. H. G.; Weill, F. S.; Bustamante, J.; Leoni, J. Mitochondrial dysfunction and cellular stress progression after ultraviolet B irradiation in human keratinocytes. Photodermatol. Photoimmunol. Photomed. 24:115–122; 2008.