Evasion of UVC-Induced Apoptosis by Photorepair of Cyclobutane Pyrimidine Dimers

EXPERIMENTAL CELL RESEARCH ARTICLE NO.

244, 43–53 (1998)

EX984180

Evasion of UVC-Induced Apoptosis by Photorepair of Cyclobutane Pyrimidine Dimers Reiko Nishigaki,1 Hiroshi Mitani, and Akihiro Shima Laboratory of Radiation Biology, Department of Biological Sciences, School of Science, University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan

other deleterious effects. Mutation of certain genes might transform the cell into an immortalized cell and thereby induce a tumor. Thus elimination by apoptosis of cells that have sustained irreparable sever damage is an effective mechanism by which multicellular organisms maintain their order of cellular society. For example, the UV light contained in sunlight can precipitate a series of genetic events in skin leading to cancer. A single mutated cell can expand into a clone in the absence of recognition and elimination of aberrant cells by apoptosis [1]. After cell treatment with UV radiation, cell-cycle delay or apoptosis can occur. These responses are attributed not only to the perturbation of the physiological cell environment but also to an active process. There are variety of such responses, and the selection of activated response depends on the type of damage. UV light generates biological damage that can be largely divided into two groups: membrane damage and DNA damage [2]. UV light induces oxidative stress and production of free radicals near the cell membrane. These molecules elicit changes in membrane-associated proteins or lipids [3]. Irradiation with short-wavelength UV light induces damage to DNA primarily by the formation of cyclobutane pyrimidine dimers (CPDs). Although there have been studies of UV-induced apoptosis mediated by membraneassociated cytoplasmic cascades [4], only few studies have attempted to clarify the apoptotic pathway triggered by UV-induced DNA damage [5]. CPD photolyase is known to reverse pyrimidine dimers specifically. CPD photolyase is found in various organisms from E. coli to higher eukaroytes, while some species, such as placental mammals, utterly lack it. This enzyme reverses pyrimidine dimer more rapidly than other repair systems, such as that of nucleotide excision repair. Furthermore, the photorepair (PR) by this enzyme only occurs during exposure of the cell to light in the visible range of the wavelength (320 – 410 nm), which makes this system a convenient tool for controlling the amount of cyclobutane pyrimidine dimer after UV irradiation. In this report, we compare the frequency of apoptosis between the cell line which overexpresses the CPD photolyase gene and the progenitor cell line. It is

Cyclobutyl pyrimidine dimer (CPD) photolyase is known to reverse pyrimidine dimers specifically under illumination with visible light. OCP13, a Medaka cell line showing a high level expression of the gene for CPD photolyase, completely reversed pyrimidine dimers induced by 20 J/m2 UVC by 1 h of photorepair. When OCP13 cells were irradiated with 20 J/m2 UVC, morphological changes such as shrinkage of cells, distorted nuclear shape, and decrease in the number of nucleoli appeared 2 to 4 h after UVC irradiation. Thereafter, the irradiated cells began to detach from the substratum, and DNA ladders were observed in the DNA extracted from detached cells. Thus, these changes in cells after UVC exposure were used to characterize the progression of UV-induced apoptosis in OCP13 cells. Although formation of DNA ladders and cell detachment were blocked by cycloheximide treatment prior to UVC exposure, the morphological changes were not. With photorepair treatment, even after the morphological changes appeared cells were still able to restore their normal morphological features and remained attached. On the other hand, the cell-cycle progression in UVC-irradiated cells was arrested even after photorepair of pyrimidine dimers. Thus, photorepair can rescue cells from UVinduced apoptosis, although DNA damage other than that of pyrimidine dimers, as well as additional nonDNA damage, possibly remained, and DNA replication was left inhibited. Among the various kinds of damage induced by UVC irradiation, the presence of pyrimidine dimers is proposed to be the major trigger for UVCinduced apoptosis. © 1998 Academic Press

INTRODUCTION

Cells with genotoxic damage arrest their cell cycle to gain time for repair. However, failure to repair all the genotoxic damage sometimes results in mutations or 1 To whom correspondence and reprint requests should be addressed at Laboratory of Radiation Biology, Department of Biological Sciences, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 1130033, Japan. Fax: 081-3-3816-1965. E-mail: [email protected].

43

0014-4827/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.

44

NISHIGAKI, MITANI, AND SHIMA

shown that UVC-induced apoptosis can be evaded when pyrimidine dimers are reversed by photorepair. This indicates that UVC-induced pyrimidine dimer is essential for UVC-induced apoptosis. MATERIALS AND METHODS Cell line and culture conditions. A cell line, OL32, derived from the tail fin of the inbred strain HB32 of the Medaka (Oryzias latipes) at exponentially growing phase was used [6]. OCP13, a cell line with high expression of the Medaka photolyase gene expression plasmid, was also used [7]. OL32 is a progenitor cell line of OCP13. Cells were cultured at 33°C in a 100-mm-diameter plastic dish with L15 medium (Irvine Scientific, Santa Ana, CA) containing 10 mM Hepes and 15% fetal bovine serum (Sera-lab, Crawley Down, Sussex, England) (pH 7.6). UVC irradiation and photorepair (PR) treatment of cells. Cells were inoculated on a plastic dish and incubated. Cells were then washed thoroughly with RPMI saline [100 mM NaCl, 11 mM glucose, 5.6 mM Na2HPO4 z 12H2O, 5.4 mM KCl, 0.4 mM Ca(NO3)2 z 4H2O, 0.4 mM MgSO4 z 7H2O]. Cells were irradiated with UVC, then left in RPMI saline for 1 h in the dark, illuminated with fluorescent light as a PR treatment, or grown in the growth medium immediately after UVC irradiation. The UVC source consisted of four GL-10 germicidal lamps (Toshiba, Tokyo, Japan). The fluence rate was 0.6 J/m2/s as determined by a UVX radiometer (Ultra-Violet Products Inc., San Gabriel, CA). For PR treatment, the cells were illuminated with six FL15W white lamps (Toshiba) through two 2-mm-thick polystyrene plates. The intensity of fluorescent light was 29 J/m2/s as determined by a SPI-5 photocell illuminometer (Toshiba). Endonuclease sensitive site (ESS) assay. The procedure of ESS assay was essentially the same as described previously [8]. Cultured cells were lysed and then DNA was extracted, and sites of pyrimidine dimers were nicked by Micrococcus luteus UV endonuclease at 37°C for 1 h. Samples were electrophoresed on 0.35% alkaline agarose gel for 16 h at 20 V. The gel was stained with ethidium bromide and photographed on a TF-35L transilluminator (Virber Lourmat, Marne la Vallee, France) with Type 665 positive-negative film (Polaroid Corp., Cambridge, MA). Detection of DNA fragmentation. After UVC irradiation and PR treatment, cells were grown in fresh medium. Four hours after UVC irradiation, cells detached from the substratum were collected from the medium while the attached cells were collected by trypsinization separately. DNA was extracted as previously described [9], electrophoresed on 2% agarose gel containing ethidium bromide, and then visualized on a TF-35L transilluminator (Virber Lourmat). Cycloheximide (CHX)-treated cells were grown in medium containing 0.2 mg/ml CHX for 24 h in advance of experiments. Histochemical analysis of preapoptotic cells. A total of 3.5 3 105 OCP13 cells in 2 ml of the medium was seeded on a coverslip in a 35-mm-diameter plastic dish. Four hours after inoculation, cells attached to the coverslip. Then the medium was changed to 2 ml of the fresh medium. Cells were irradiated with UVC in RPMI saline containing 10 mg/ml hydroxyurea (HU). PR treatment was done by illumination with fluorescent light for 1 h in the same medium. Then, cells were grown in the growth medium containing HU until they were fixed. The coverslip was removed from the dish and washed twice with PBS (2) [137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4 z 7H2O, 1.4 mM KH2PO4]. Cells were then fixed with 3% formaldehyde and washed thoroughly with water prior to further staining. Morphological changes in Giemsa-stained nuclei were examined under a light microscope. A cell with dense staining, reduced number of nucleoli per nucleus, and distorted nuclear shape was defined as a “preapoptotic” cell. Three or more randomly selected areas (.0.14 3 3 mm2) were observed for each coverslip and 200 or more attached cells were examined to determine the frequency of preapoptotic cells. Fixed cells were also stained by DAPI to reveal their nuclear shape

FIG. 1. Increased ability for photorepair of cyclobutane pyrimidine dimers in OCP13 cells detected by ESS. OL32 and OCP13 cells were irradiated with 20 J/m2 UVC, immediately followed by a 1-h PR treatment. DNA was extracted and an ESS assay carried out. Lane 1, molecular markers l HindIII; lane 2, DNA from OL32 cells without UVC; lane 3, DNA from OL32 cells with 20 J/m2 UVC; lane 4, DNA from OL32 cells with 20 J/m2 UVC followed by PR treatment; lane 5, DNA from OCP13 cells without UVC; lane 6, DNA from OCP13 cells with 20 J/m2 UVC; lane 7, DNA from OCP13 cells with 20 J/m2 UVC followed by PR treatment.

[10], and Tunel staining was performed to detect fragmentation of DNA histochemically as described previously [11]. Cells were observed under a phase-contrast microscope, using a green filter to avoid the light with effective wavelengths for photorepair. Cell-cycle analysis. A total of 2.0 3 105 OCP13 cells in 3 ml of the medium was seeded in a 60-mm-diameter dish. Cells were incubated in the dark for 1 or 2 days. Cells were further grown in medium containing 10 mg/ml HU for 24 h in advance of the experiments to synchronize the cell cycle at the S phase. The medium was exchanged for a medium without HU, and cells were grown for 4 h to synchronize the cell cycle at the G2/M phase. The doubling time of OCP13 cells was 17.4 h, and the percentages of each cell-cycle phase in asynchronized cells were G0/G1, 71.5%; S, 14.7%; and G2/M, 13.8%, indicating that the average length of the S phase of OCP13 cells is about 2.6 h. Each sample was subjected to UVC irradiation and PR treatment in RPMI saline. Four hours after UVC irradiation, cells grown in growth medium were collected by trypsinization and fixed/ stained using a Cycle Test Plus DNA reagent kit (Becton-Dickinson Immunocytometry Systems, Mountain View, CA) according to the manufacturer’s instructions. Flow-cytometric analysis was carried out using a FACScan (Becton-Dickinson), a single laser flow system. Data were analyzed using CellFit software (Becton-Dickinson). Red fluorescence of propidium iodide (PI) was measured as an index of DNA content.

RESULTS

Photorepair Ability of OCP13 and OL32 Cells The difference in photorepair ability between OL32 and OCP13 cells was examined (Fig. 1). Cells were irradiated with 20 J/m2 UVC followed by a 1-h PR treatment. To estimate the overall frequency of pyrimidine dimers in the genome, an ESS assay was performed on extracted DNA using UV endonuclease of M. luteus. Electrophoresed on alkaline agarose gels, the

EVASION OF UV-INDUCED APOPTOSIS BY PHOTOREPAIR

45

control cells. With or without UVC irradiation, cells attached to the substratum did not show a DNA ladder (Fig. 2, lanes 5–7). Cells grown in the medium containing 0.2 mg/ml cycloheximide (CHX) 24 h in advance of UVC irradiation did not show a DNA ladder and contained only high molecular weight DNA (Fig. 2, lane 4). UVC-Induced Preapoptotic Cells

FIG. 2. DNA ladder formation after UVC irradiation. OCP13 cells were irradiated with 20 J/m2 UVC, cells detached/attached from the substratum during 4 h of incubation after irradiation were collected separately, and DNA was extracted. CHX treatment was done by adding CHX to the medium (0.2 mg/ml) for 24 h in advance of UVC irradiation. After UVC irradiation, cells were grown in medium without CHX. The initial number of cells used for each experiment was the same. Lane 1, molecular markers l HindIII; lane 2, DNA from detached cells without UVC; lane 3, DNA from detached cells with 20 J/m2 UVC; lane 4, DNA from detached cells with 20 J/m2 UVC after CHX treatment; lane 5, DNA from attached cells without UVC; lane 6, DNA from attached cells with 20 J/m2 UVC; lane 7, DNA from attached cells with 20 J/m2 UVC after CHX treatment.

smear pattern of DNA corresponds with the number of endonuclease-sensitive sites (pyrimidine dimers), which was detected as a decrease in the molecular weight of the enzyme-treated DNA. In both cell lines, no dark repair of pyrimidine dimers was observed under this experimental condition (Fig. 1, lanes 3 and 6). In OL32 cells, the progenitor of OCP13 cells, photorepair of the pyrimidine dimers induced by 20 J/m2 UVC could not be detected by a 1-h PR treatment (Fig. 1, lane 4). On the other hand, OCP13 cells with high expression of the photolyase gene completely photorepaired the dimers induced by the same dose of UVC with the same PR treatment (Fig. 1, lane 7). DNA Ladder and UVC-Induced Apoptosis Detachment of both OCP13 and OL32 cells started at 4 h after 20 J/m2 UVC irradiation, and most cells detached within 24 h. The cells which remained attached to the dish after 24 h were able to begin proliferating again (data not shown). DNA was extracted from OCP13 cells 4 h after 20 J/m2 UVC irradiation. DNA extracted from the cells which detached from the substratum after UVC irradiation showed a DNA ladder (Fig. 2, lane 3) whereas DNA from nonirradiated control cells did not (Fig. 2, lane 2). However, when a higher concentration of DNA from nonirradiated control cells was electrophoresed, a DNA ladder was detected (data not shown), demonstrating the low frequency of apoptosis occurring in

OCP13 cells which attached to the substratum changed their morphological features 4 h after 20 J/m2 UVC irradiation (Fig. 3): shrinkage of cells and decrease in number of nucleoli to 1 or 2 was observed by Giemsa staining, and distorted nuclear shape was seen by DAPI staining. By Tunel staining, DNA fragmentation was rarely observed in attached cells 4 h after UVC irradiation. This corresponds to the result that DNA ladders were observed not in the attached cells but only in detached cells (Fig. 2, lanes 3 and 6). A time-lapse sequence of the cell shape from 4 to 5 h after UVC irradiation is shown in Fig. 4. Cells with characteristic morphological features after UVC irradiation proceeded to shrink further, before progressing to an apoptotic phase with apoptotic bodies. The UVC-irradiated cells with morphological characteristics such as cell shrinkage, distorted nuclear shape, and decreased number of nucleoli were defined as “preapoptotic” cells. It took only a few minutes for cells to progress from an initiation of shrinkage to fragmentation into several apoptotic bodies. This transformation from preapoptosis to apoptosis occurred randomly among preapoptotic cells. When UVC-irradiated cells were subjected to PR treatment immediately after UVC irradiation, they restored their normal morphological features (Fig. 3), remained attached to the dish and, after 24 h, began to proliferate again (data not shown). Preapoptotic cells tend to be condensed and the boundary of adjacent cells to be unclear which made it hard to distinguish each cell. However, when the cells were grown for 4 h in medium with 10 mg/ml of HU after UVC irradiation, nuclei became more plump and preapoptotic cells separated from each other, which made it easier to measure the number of preapoptotic cells. HU did not have any effect on frequency of preapoptotic cells with or without UVC irradiation (Table 1). Subsequent experiments on preapoptosis were done in the presence of HU. Although CHX-inhibited DNA ladder formation (Fig. 2, lane 4), CHX did not affect the frequency of preapoptotic cells (Table 2). Dose Dependency and Time Course of UVC-Induced Preapoptosis OCP13 cells were fixed 4 h after UVC irradiation and preapoptotic cells were examined. The number of preapoptotic cells increased in a dose-dependent manner between 0 to 30 J/m2 UVC irradiation (Fig. 5). Pre-

46

NISHIGAKI, MITANI, AND SHIMA

FIG. 3. Morphology of preapoptotic cells after UVC irradiation. OCP13 cells were irradiated with 20 J/m2 UVC, immediately followed by a 1-h PR treatment. Cells were then grown in medium containing 10 mg/ml HU until fixation with formaldehyde at 4 h after UVC irradiation. Cells were stained by Giemsa staining, DAPI staining, or Tunel staining and observed under a light/fluorescent microscope. Scale bars, 20 mm.

EVASION OF UV-INDUCED APOPTOSIS BY PHOTOREPAIR

47

FIG. 4. Morphological changes of cells proceeding from preapoptosis to apoptosis at 4 h after UVC irradiation. Morphological changes of the OCP13 cells were observed under a phase-contrast microscope after irradiation with 20 J/m2 UVC. Photographs of the same field were taken every 10 min from 4 to 5 h after UVC irradiation. Scale bars, 40 mm.

48

NISHIGAKI, MITANI, AND SHIMA

TABLE 1 Effect of Hydroxyurea (HU) on UV-Induced Preapoptosis Treatment

2

0 J/m 20 J/m2 20 J/m2 1 PR

Frequency of preapoptosis (%) 2HU

1HU

5.3 6 2.7 40.6 6 5.1 13.4 6 6.8

6.3 6 1.3 31.9 6 1.4 12.7 6 2.4

Note. OCP13 cells were irradiated with 20 J/m2 UVC, immediately followed by a 1-h PR treatment. Cells were grown for 4 h in medium supplemented with 10 mg/ml HU after UVC irradiation. After UVC irradiation, cells were grown in medium without HU for 4 h until fixation.

apoptotic changes were not observed until 2 h after irradiation (Fig. 6). From 2 to 4 h, the fraction of preapoptotic cells increased. At 4 h after 20 J/m2 UVC irradiation, most cells were still attached to the substratum and, in 40% of these cells, preapoptotic changes were observed.

FIG. 5. Dose dependency of UVC-induced preapoptosis. OCP13 cells were irradiated with various doses of UVC and fixed 4 h after UVC irradiation, and preapoptotic cells were examined. Each bar represents the standard error of three independent experiments.

Effect of Photorepair on Evasion of Preapoptosis and Apoptosis

Effective Time of Photorepair on Evasion of Preapoptosis

OCP13 cells with a high capability for photorepair (Fig. 1, lane 7) were able to evade preapoptosis when PR treatment was performed immediately after UVC irradiation (Fig. 7). The frequency of preapoptosis was below 10%, which is the same as that for nonirradiated cells. On the other hand, OL32 cells hardly photorepaired pyrimidine dimers induced by 20 J/m2 UVC (Fig. 1, lane 4) and could not evade preapoptosis. Evasion of preapoptosis by photorepair also occurred in OCP2 cells, which were another independently isolated sister clone of OCP13 showing a high expression of photolyase gene (data not shown). Furthermore, with UVC irradiation, the number of cells that detached from the substratum increased, whereas with PR treatment after UVC irradiation, the number of detached cells was reduced to that of nonirradiated cells (data not shown). This would indicate that the reversal of pyrimidine dimers is the major cause of evasion of preapoptosis by photorepair.

To determine how long after UVC irradiation preapoptosis could be evaded by photorepair, UVC-irradiated cells were subjected to PR treatment immediately or at 2 or 4 h after irradiation (Table 3). The fraction of apoptotic cells was 59% at 4 h postirradiation. However, when PR treatment was performed 4 h after UVC irradiation, the fraction of apoptotic cells decreased to 6.6% (versus 66% without photorepair) at 8 h postirradiation.

TABLE 2 Effect of Cycloheximide (CHX) on UV-Induced Preapoptosis Treatment

2

0 J/m 20 J/m2

Frequency of preapoptosis (%) 2CHX

1CHX

2.8 6 0.98 49.7 6 4.7

6.2 6 3.3 47.3 6 3.6

Note. OCP13 cells were grown in medium supplemented with 0.2 mg/ml CHX for 24 h in advance of irradiation with 20 J/m2 UVC. After UVC irradiation, cells were grown in medium without CHX for 4 h until fixation.

FIG. 6. Time course of induction of preapoptosis. OCP13 cells were irradiated with 20 J/m2 UVC, immediately followed by a 1-h PR treatment. Cells were then fixed at various times after UVC irradiation and stained with Giemsa for observation under a light microscope. Each bar represents the standard error of three independent experiments.

49

EVASION OF UV-INDUCED APOPTOSIS BY PHOTOREPAIR

FIG. 7. Evasion of UVC-induced preapoptosis of OCP13 cells by photorepair treatment. OL32 and OCP13 cells were irradiated with 20 J/m2 UVC, immediately followed by a 1-h PR treatment. Cells were fixed 4 h after UVC irradiation, stained by Giemsa, and observed under a microscope. Each bar represents the standard error of three independent experiments. An asterisk indicates significant difference (P , 0.05, Student’s t test).

FIG. 8. Difference in frequency of preapoptosis at various cellcycle phases. OCP13 cells were synchronized at each phase of the cell cycle, then irradiated with 20 J/m2 UVC, and immediately followed by a 1-h PR treatment. Cells were fixed 4 h after UVC irradiation, stained by Giemsa, and observed under microscope. Each bar represents the standard error of three samples in one experiment.

Cell-Cycle Arrest after UVC Irradiation

ble 4). On the other hand, PR treatment alone had no effect on progression of the cell cycle. Thus, UVC with or without PR treatment had no effect on the cell-cycle progression of cells synchronized in the G2/M phase.

Synchronized OCP13 cells were subjected to UVC irradiation immediately followed by PR treatment. The frequency of preapoptosis by UVC irradiation did not differ between asynchronous, S phase- and G2 phasesynchronized cells (Fig. 8). Moreover, the evasion of preapoptosis by PR treatment also had no effect from different cell cycles. In asynchronous and S phasesynchronized cells, S phase arrest could be observed in UVC-irradiated cells, both with and without PR treatment (Fig. 9). However, the mode of DNA contents of cells in S phase moved toward the G2/M position in PR-treated cells compared to non-PR-treated cells (Ta-

DISCUSSION

Trigger of UVC-Induced Apoptosis: Membrane Damage or DNA Damage? Many biochemical and cytological effects of UV irradiation, including damage to DNA, the plasma membrane, mitochondria, and other extranuclear structures, are potential causes of the cell death induced by

TABLE 3 Evasion of Preapoptosis by Delayed PR Treatment Frequency of preapoptosis (%) Time after UVC irradiation 4h

8h

0 J/m2

3.5 6 0.6

2.3 6 1.2

20 J/m2

58.6 6 2.4

66.3 6 13.0

, UVC irradiation

, PR (1 h)

, Fix

20 J/m2 1 PR at 0 h

4.2 6 0.3

20 J/m2 1 PR at 2 h

9.3 6 5.4

20 J/m2 1 PR at 4 h

6.6 6 0.6

Note. UVC-irradiated OCP13 cells were subjected to a 1-h PR treatment immediately (0 h), 2, or 4 h after irradiation. Cells were fixed 4 or 8 h after UVC irradiation.

50

NISHIGAKI, MITANI, AND SHIMA

FIG. 9. S phase arrest by UVC or PR treatment in various cell cycles. OCP13 cells were synchronized at each phase of cell cycle using 10 mg/ml HU, then irradiated with 20 J/m2 UVC, immediately followed by a 1-h PR treatment. Four hours after UVC irradiation, cell cycle was determined by flow cytometry. Percentage of cells in the G0/G1, S, and G2/M phases of the cell cycle was calculated from flow-cytometric measurements of DNA content using SFIT.

UV irradiation. In some cell lines, the lethal effects of UVC have been reversed by photoreactivation: the decreased clonogenic survival rate by UVC irradiation was recovered with PR treatment in cultured goldfish cells [12, 13]. In these studies, the removal of pyrimi-

dine dimers by photorepair was the most likely candidate for the recovery of cell survival. A similar result was obtained with rat kangaroo cells [5]. However, the fact that some cell lines undergo apoptosis in the absence of nuclei [3] supports the hypothesis that the

EVASION OF UV-INDUCED APOPTOSIS BY PHOTOREPAIR

TABLE 4 Effect of PR Treatment on UVC-Induced S Phase Arrest S phase progression (%) Before treatment 4 h after treatment CONT UV UV 1 PR

16.2 6 7.3 100 41.2 6 5.7 70.0 6 3.0

Note. OCP13 cells synchronized at S phase was irradiated with 20 J/m2 UVC, immediately followed by a 1-h PR treatment. The cell cycle was analyzed by flow cytometry 4 h after UVC irradiation. S phase progression was calculated from the mode of the S phase using mode of DNA contents of G0/G1 as 0% and mode of DNA contents of G2/M as 100% S phase progression: S phase progression ~%! 5

M~S! 2 M~G0 /G1! 3 100 M~G2 /M! 2 M~G0 /G1!

(M, Mode of DNA contents).

major cause of UV-induced apoptosis is non-DNA damage, presumably membrane damage, which is known to activate the JNK/SAPK cascade [4]. The typical morphological changes occurring in cells during apoptosis are common among various cell types. Cell shrinkage before breaking into apoptotic bodies and nuclear condensation before nuclear fragmentation were observed in mammalian cells [14] as well as in Medaka cells undergoing apoptosis (Figs. 3 and 4). One of the characteristics of apoptosis is that it requires nascent protein synthesis, as evidenced by the fact that apoptosis can be inhibited by inhibitors of protein synthesis, such as cycloheximide. In Medaka cells, cycloheximide did not have any effect on preapoptosis, but it could inhibit apoptosis, as shown by inhibition of DNA ladder formation (Fig. 2). Furthermore, in a study by Godar et al., UV induced two kinds of apoptosis in murine lymphoma cells depending on the wavelength. UVA induced an immediate (,4 h) and a delayed (.20 h) apoptosis, while UVB or UVC induced only delayed apoptosis. This suggests that immediate apoptosis is initiated by membrane damage, since it is prevented by vitamin E, while delayed apoptosis is initiated by DNA damage [15]. In the present study, preapoptosis and apoptosis occurred in the Medaka cultured cells 4 h after UVC irradiation (Fig. 6), a time course resembling that of immediate apoptosis induced by UVA in murine lymphoma cells. However, because UVC-induced apoptosis was evaded by PR treatment in Medaka cells (Fig. 7), it is clear that pyrimidine dimers were the trigger of UVC-induced apoptosis. Our preliminary data show that UVA irradiation of Medaka cells induced apoptosis in 4 h (unpublished).

51

Role of Photolyase in Evasion of the Apoptotic Pathway Irradiation with UVC light induces damage to DNA, primarily via the formation of cyclobutane pyrimidine dimers. Photolyases, the enzymes that perform photoreactivation, bind specifically to cyclobutane pyrimidine dimers and catalyze their photoreduction by using photon energy between 320 to 410 nm. The photolyase genes characterized to date fall into two types based on sequence homology: Type I and Type II enzymes. Type I photolyase is responsible for photoreactivation in many microbial organisms. In plants, however, the blue light photoreceptors having a high homology for Type I photolyase do not show photoreactivation activity [16]. Type II photolyases are also found in prokaryotic and eukaryotic organisms, and are functionally interchangeable with the Type I photolyase. Type II photolyase of plants complements a photolyase-deficient mutant of E. coli [17]. It has been suggested that a common ancestor gave rise to both Type I and Type II photolyases, but diverged very early during evolution [18]. The Medaka photolyase gene, which has been shown to be overexpressed in the OCP13 cell line, is a Type II photolyase gene [7, 18]. Fish cells have a high ability for photorepair compared to cells of other classes. On the other hand, fish cells have a low ability for removing cyclobutane pyrimidine dimers from the genome as a whole by dark excision repair [19]. Photolyase, which is overproduced in OCP13 cells, is known only to reverse cyclobutane pyrimidine dimers. When OCP13 cells were irradiated with UVC, they developed preapoptotic morphological features, detached from the substratum, and underwent apoptosis. However, PR treatment immediately or within at least 4 h after UVC irradiation evaded both preapoptotic changes in terms of morphological features and apoptotic cell detachment (Table 3). On the other hand, OL32 cells, which are the progenitor cell line of OCP13 and express a low level of photolyase as detected by RT-PCR [7], could not evade UVC-induced preapoptosis or apoptosis by PR treatment (Fig. 7). This supports a hypothesis that pyrimidine dimer is the direct trigger for UVC-induced apoptosis. The fact that PR treatment of preapoptotic cells was able to prevent apoptosis suggests that persistent presence of pyrimidine dimers is necessary for manifestation of the morphological features of preapoptosis. UVC-Induced Preapoptosis Is Independent of CellCycle Arrest Cells were synchronized using 10 mg/ml HU and cells at S or G2/M phase and asynchronous cells were examined for frequency of preapoptosis 4 h after UVC irradiation. However, no cell-cycle dependency was ob-

52

NISHIGAKI, MITANI, AND SHIMA

served with regard to frequency of preapoptosis. This leads to a conclusion that the sensitivity to UVC or PR treatment was not associated with the phase of cell cycle. PR treatment was effective for evasion of preapoptosis irrespective of the phase of cell cycle at the time of UVC irradiation or PR treatment. However, the cell-cycle progression was arrested at the S phase by UVC, and PR treatment immediately after UVC irradiation had no additional effect (Fig. 8). On the other hand, no G1/G0 arrest occurred in OCP13 cells after UVC irradiation (Fig. 9). Some studies have shown S phase arrest after UVC irradiation, while others have demonstrated an increase in fraction of G1 phase cells after irradiation. The checkpoint at which the cells are arrested/delayed after UVC irradiation may be dependent on cell type and/or culture conditions [20, 21]. Several processes, such as p53 cascade, are known to regulate the cell cycle and DNA repair in order to prolong the time available for repair and protect cells against the mutagenesis and clastogenesis induced by DNA-damaging agents such as radiation and chemicals. Although PR treatment successfully reversed cyclobutane pyrimidine dimer and prevented preapoptosis, the cell-cycle arrest was not recovered for at least 4 h. The S phase arrest after UVC irradiation was recovered partially by subsequent PR treatment compared with the arrest after UVC irradiation only. Thus, the S phase arrest after UVC irradiation may be due to inhibition of DNA replication by DNA damage other than that of pyrimidine dimers, which cannot be repaired by photorepair. One of the candidates for the DNA damage responsible for inhibition of DNA replication is the (6-4) photoproducts [22]. Although (6-4) photoproducts can be photorepaired [23], the (6-4) photoproduct induced by 10 J/m2 UVC irradiation has been shown to remain after 1-h PR treatment of goldfish cells [23], and in another study, these products probably remained for more than 4 h after UVC irradiation [24]. It is predicted that the (6-4) photoproducts induced by 20 J/m2 UVC irradiation in OCP13 cells will exceed endogenous photorepair ability. The reason that the remaining (6-4) photoproducts produced by 20 J/m2 UVC irradiation after PR treatment do not induce apoptosis may be because the (6-4) photoproducts are repaired rapidly by excision repair [25]. UVC-Induced Apoptotic Pathway Induced by Pyrimidine Dimers A model proposed for the evasion of apoptosis by photorepair is that the presence of the pyrimidine dimers could be monitored by some unknown mechanism throughout the apoptotic pathway (Table 3). The persistent presence of pyrimidine dimers is required for progression of the apoptotic pathway. And when the dimers disappear by photorepair or through the action

of other repair systems, the progression of the apoptotic pathway would be stopped and the cells permitted to engage exclusively in the repair of other DNA or membrane damage. Cockayne’s syndrome (CS) cells, which are defective in preferential repair of pyrimidine dimers, are more apt to undergo apoptosis by UVC irradiation than xeroderma pigmentosum (XP)-C cells, which are defective in genomic repair, or normal cells [26]. Ljungman et al. have proposed a pyrimidine dimer recognition system leading UV-exposed cells to apoptosis. Blockage of RNA polymerase II at pyrimidinedimer sites would be one of the candidates for eliciting a signal to the apoptotic pathway. However, in preferential repair, it has been shown that the arrested RNA polymerase II shields the pyrimidine dimer from recognition by photolyase [27]. From these previous studies, the following model is suggested for UVCinduced apoptosis in OCP13 cells. First, by UVC irradiation, pyrimidine dimers are produced on DNA. Second, RNA polymerases encounter pyrimidine dimers. Third the blockage of RNA polymerase elicits a signal to the apoptotic pathway. At 4 h after UVC irradiation, the apoptotic pathway is already activated and OCP13 cells change their morphological feature to preapoptotic features and then progress into apoptotic phase with cell detachment and DNA fragmentation. Although the apoptotic pathway has proceeded to preapoptotic phase, evasion of UVC-induced preapoptosis occurred by photorepair 4 h after UVC irradiation in OCP13 cells (Table 3). This result indicates the reversibility of the apoptotic pathway in OCP13 cells. One model for evasion of apoptosis by photorepair is that the pyrimidine dimers shielded by RNA polymerases in transcribed strand are repaired by preferential repair, while unshielded pyrimidine dimers in the direction of movement of RNA polymerases are repaired by photolyase. When the blocked RNA polymerases are released by preferential repair, the elongation of RNA polymerases will go successfully without any more blockage. Then, the apoptotic pathway which requires persistent blockage of RNA polymerases will be interrupted. Another model is that photorepair of pyrimidine dimers in the genome may have an important role in evasion of preapoptosis because pyrimidine dimers on transcribed strand would be unavailable for photorepair by arrested RNA polymerases, shielding the pyrimidine dimer sites. In order to determine the role of the release of blocked RNA polymerases from pyrimidine dimer sites in evasion of apoptosis by photorepair, the effects on RNA synthesis by photorepair are under study. If evasion of apoptosis by photorepair occurred under inhibitors of RNA polymerases, the latter model suggesting the importance of pyrimidine dimers on genome is supported. Comparing OCP13 cells to those of the progenitor

EVASION OF UV-INDUCED APOPTOSIS BY PHOTOREPAIR

cell line OL32, OCP13 cells must rely solely on high photolyase activity in order to evade apoptosis by PR treatment. The results of the present study, in addition to showing for the first time that pyrimidine dimers are an essential trigger for UVC-induced apoptosis, might also represent an important step in elucidating the mechanism of UV-induced apoptosis and help to clarify the intriguing pathway of the trigger of apoptosis. It will be of interest to determine whether pyrimidine dimers pass signals down to already known apoptotic pathways, such as caspase cascades. We thank Prof. Nobuo Tsuchida and Dr. Tatsuo Hamazaki of the Tokyo Medical and Dental University for their encouragement, discussion, and advice on the flow-cytometry analysis, and Kaori Suzuki and Shiro Tochitani for their assistance with the Tunel staining. This work was supported in part by grants-in-aid from the Ministry of Education, Science, Sports, and Culture of Japan, from the Science and Technology Agency of Japan, and from the Japan Society for the Promotion of Science.

REFERENCES 1. Brash, D. E. (1997) TIG 13, 410 – 414. 2. Friedberg, E. C., Walker, G. C., and Siede, W. (1995) “DNA Repair and Mutagenesis,” ASM, Washington, DC. 3. Devary, Y., Rosette, C., DiDonato, J. A., and Karin, M. (1993) Science 261, 1442–1445. 4. Verheij, M., Bose, R., Lin, X. H., Yao, B., Jarvis, W. D., Grant, S., Birrer, M. J., Szabo, E., Zon, L. I., Kyriakis, J. M., Haimovitz-Friedman, A., Fuks, Z., and Kolesnick, R. N. (1996) Nature 380, 75–79. 5. Miyaji, E., and Menck, C. (1995) Photochem. Photobiol. 61, 454 – 458. 6. Komura, J., Mitani, H., and Shima, A. (1988) In Vitro Cell Dev. Biol. 24, 294 –298. 7. Funayama, T., Mitani, H., and Shima, A. (1996) Photochem. Photobiol. 63, 633– 638. 8. Yasuhira, S., Mitani, H., and Shima, A. (1991) Photochem. Photobiol. 53, 211–215. Received January 7, 1998 Revised version received May 8, 1998

9.

53

Ishizawa, M., Kobayashi, Y., Miyamura, T., and Matsuura, S. (1991) Nucleic Acids Res. 19, 5792.

10.

Hamada, S., and Fujita, S. (1983) Histochemistry 79, 219 –226.

11.

Gavrieli, Y., Sherman, Y., and Ben-Sasson, S. A. (1992) J. Cell Biol. 119, 493–501.

12.

Shima, A., Ikenaga, M., Nikaido, O., Takebe, H., and Egami, N. (1981) Photochem. Photobiol. 33, 313–316.

13.

Mano, Y., Kator, K., and Egami, N. (1982) Photochem. Photobiol. 35, 735–737.

14.

Wyllie, A. H., Kerr, J. F. R., and Currie, A. R. (1980) Int. Rev. Cytol. 68, 251–306.

15.

Godar, D. E., and Lucas, A. D. (1995) Photochem. Photobiol. 62, 108 –113.

16.

Lin, C., Robertson, D. E., Ahmad, M., Raibekas, A. A., Jorns, M. S., Dutton, P. L., and Cashmore, A. R. (1995) Science 269, 968 –970.

17.

Ahmad, M., Jarillo, J. A., Klimczak, L. J., Landry, L. G., Peng, T., Last, R. L., and Cashmore, A. R. (1997) The Plant Cell 9, 199 –207.

18.

Yasui, A., Eker, A., Yasuhira, S., Yajima, H., Kobayashi, T., Takao, M., and Oikawa, A. (1994) EMBO J. 13, 6143– 6151.

19.

Komura, J., Mitani, H., Nemoto, N., Ishikawa, T., and Shima, A. (1991) Mutat. Res. 254, 191–198.

20.

Laat, A., Tilvurg, M., Leun, J. C., Volten, W. A., and Gruijl, F. R. (1996) Photochem. Photobiol. 63, 492– 497.

21.

Banrud, H., Stokke, T., Moan, J., and Berg, K. (1995) Carcinogenesis 16, 1087–1094.

22.

Uchida, N., Mitani, H., Todo, T., Ikenaga, M., and Shima, A. (1997) Photochem. Photobiol. 65, 964 –968.

23.

Todo, T., Takemori, H., Ryo, H., Ihara, M., Matsunaga, T., Nikaido, O., Sato, K., and Nomura, T. (1993) Nature 361, 371– 374.

24.

Mitchell, D. L., Clarkson, J. M., Chao, C. C.-K., and Rosenstein, B. S. (1986) Photobiology 43, 595–597.

25.

Mitchell, D. L. (1988) Photochem. Photobiol. 48, 51–57.

26.

Ljungman, M., and Zhang, F. (1996) Oncogene 13, 823– 831.

27.

Donahue, B. A., Yin, S., Taylor, J., Reines, D., and Hanawalt, P. C. (1994) Proc. Natl. Acad. Sci. USA 91, 8502– 8506.