Sensitization to UV-induced apoptosis by the histone deacetylase inhibitor trichostatin A (TSA)

Sensitization to UV-induced apoptosis by the histone deacetylase inhibitor trichostatin A (TSA)

Experimental Cell Research 306 (2005) 94 – 102 Sensitization to UV-induced apoptosis by the histone deacetylase inhibit...

474KB Sizes 0 Downloads 14 Views

Experimental Cell Research 306 (2005) 94 – 102

Sensitization to UV-induced apoptosis by the histone deacetylase inhibitor Trichostatin A (TSA) Myoung Sook Kimb,1, Jin Hyen Baek1, Devulapalli Chakravartya, David Sidranskyb, France Carriera,T a

Biochemistry and Molecular Biology Department, School of Medicine, and Greenebaum Cancer Center, University of Maryland, 108 North Greene Street, Room 330, Baltimore, MD 21201-1503, USA b Department of Otolaryngology-Head and Neck Surgery, Head and Neck Cancer Research Division, Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA Received 4 October 2004, revised version received 31 January 2005 Available online 19 March 2005

Abstract UV-induced apoptosis is a protective mechanism that is primarily caused by DNA damage. Cyclobutane pyrimidine dimers (CPD) and 6-4 photoproducts are the main DNA adducts triggered by UV radiation. Because the formation of DNA lesions in the chromatin is modulated by the structure of the nucleosomes, we postulated that modification of chromatin compaction could affect the formation of the lesions and consequently apoptosis. To verify this possibility we treated human colon carcinoma RKO cells with the histone deacetylase inhibitor trichostatin A (TSA) prior to exposure to UV radiation. Our data show that pre-treatment with TSA increased UV killing efficiency by more than threefold. This effect correlated with increased formation of CPDs and consequently apoptosis. On the other hand, TSA treatment after UV exposure rather than before had no more effect than UV radiation alone. This suggests that a primed (opened) chromatin status is required to sensitize the cells. Moreover, TSA sensitization to UV-induced apoptosis is p53 dependent. p53 and acetylation of the core histones may thus contribute to UV-induced apoptosis by modulating the formation of DNA lesions on chromatin. D 2005 Elsevier Inc. All rights reserved. Keywords: UV; Apoptosis; Histone acetylation; p53

Introduction Living cells are constantly exposed to different sources of DNA damaging agents. In order to maintain their genomic integrity and survive, cells have evolved a complex response that either leads to apoptosis or cell cycle arrest. Several genes that are activated by stress responsive protein kinases regulate these two cellular events. In general, the genes that are activated by the stress response play a protective role against the cellular insults [1]. UV radiation is a major source of DNA damage that can activate the stress response. In addition of causing the formation of DNA

T Corresponding author. Fax: +1 410 706 8297. E-mail address: [email protected] (F. Carrier). 1 Present address: Johns Hopkins University, Baltimore, MD 21205, USA. 0014-4827/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2005.02.013

lesions such as cyclobutane pyrimidine dimers (CPD) and 64 photoproducts (6-4PPs), UV radiation can also trigger plasma membrane and mitochondria events that can lead to apoptosis [2,3]. However, the main source of UV-induced apoptosis is the formation of DNA lesions [4]. This was best illustrated in cells that are deficient for nucleotide excision repair (NER), the mechanism responsible for the repair of UV-induced DNA damage. These cells are dramatically hypersensitive to UV-induced apoptosis, which indicates that DNA damage is the major stimulus for the apoptotic response [5]. The capacity to induce apoptosis following exposure to UV radiation is believed to be a protective mechanism that allows the elimination of severely damaged cells that could potentially lead to malignant transformation. Even though primary DNA lesions such as CPDs and 64PPs are the predominant initiators of UV-induced apoptosis, they do not cause apoptosis by themselves [6]. These

M.S. Kim et al. / Experimental Cell Research 306 (2005) 94 – 102

lesions need to be converted into more critical secondary lesions, probably DNA double strand breaks (DSBs), through replication. This is supported by the fact that electroporation of the restriction enzyme PvuII into mouse fibroblasts is sufficient to induce apoptosis [7]. Formation of DNA strand breaks is also sufficient to activate the tumor suppressor p53 [8]. p53 is the most mutated gene in human cancers and sits at an important decisional fork to regulate both cell cycle arrest and apoptosis in response to DNA damage [9]. In response to low levels of DNA damage, p53 can activate the transcription of regulatory genes such as p21 to arrest the cell cycle and presumably allow time to repair the DNA. On the other hand, if damage is beyond repair, p53 can trigger an apoptotic response to eliminate the damaged cells and prevent transmission of potentially mutagenic material to the next generation of cells. In response to UV radiation, p53 can mediate an apoptotic response but its role is rather complex and seems to be cell type specific. For example, lymphoblastoid cells that are wild type for p53 are more sensitive to UV-induced apoptosis than their p53 mutated counterparts while p53 is not required to induce apoptosis in response to UV radiation in fibroblasts [6]. This may reflect differing basal patterns of expression of pro- and anti-apoptotic regulators that are under the control of p53 in these cells [10]. Another possible mechanism by which p53 could affect UV-induced apoptosis is through its recently described chromatin relaxation activity in response to UV radiation [11]. p53 mediates this effect by recruiting the histone acetylase p300 to allow acetylation of histone H3 at sites of NER [11]. Post-translational modifications, especially acetylation, of core histone have been associated with a looser, more open, chromatin structure that facilitates accessibility to transcription, replication, and repair machinery [12]. Acetylation of the amino-terminal tails of the core histones occurs on lysine residues and is mediated by histone acetyltransferases (HATs). However, acetylation is a reversible process that is catalyzed by the histone deacetylases (HDACs) enzymes. Deacetylation of the core histone thus results in a repressed (compacted) chromatin structure that is inhibitory to most cellular processes [12]. The formation of UV-induced DNA lesions is thus likely to be affected by the structure of the chromatin (compact vs. open) that is dictated by the nucleosomes acetylation status. This is because the formation of the lesions is not random on the chromatin. 64PPs are predominantly formed in linker DNA while CPDs occur at sites where the DNA minor grooves face the outside of the nucleosome cores [13]. To verify the possibility that modulation of chromatin through inhibition of histone deacetylases could affect UVinduced apoptosis, we treated human colon carcinoma cells with the histone deacetylase inhibitor trichostatin A (TSA) prior to exposing the cells to UV radiation. Our data show that pre-treatment with TSA decreased the survival of colon cancer cells exposed to UV radiation by more than threefold. This effect correlates with increased formation of CPDs and


apoptosis as measured by TUNEL assay and DNA ladder formation. Moreover, the increased apoptosis was not observed in cells expressing reduced p53 levels thus indicating that the TSA effect is p53 dependent. Because TSA treatment after rather than before UV radiation did not result in a similar synergy, we conclude that a primed (opened) chromatin structure is required to sensitize the cells to UV radiation. Collectively these data suggest that p53 and acetylation of the core histone could contribute to UV-induced apoptosis by modulating the formation of DNA lesions on chromatin.

Materials and methods Cell culture and treatments The human colon carcinoma RKO and RKO-E6 cells, the human breast cancer cells MCF-7, and the normal human skin fibroblasts Malme-3 cells were purchased from American Type Culture Collection (ATCC, Manassas, VA). RKO, RKO-E6, and MCF-7 cells were grown in RPMI 1640 supplemented with 10% fetal bovine serum in the absence of antibiotics. The Malme-3 cells were grown in DMEM:Ham’s F-12 containing 10% Chelex-treated horse serum. UV radiation was performed with a UVC Philips 30-W germicidal lamp emitting at 253.7 nm and intensities of the lamp were determined with a UVX radiometer (UVP Inc., Upland, CA). Colony formation assay was performed as described before [14]. Briefly, the cells were plated at 2 – 3  102 cells/60-mm-diameter Petri dish, rinsed with phosphate buffer, and exposed dried to the different doses of UV radiation (0, 5, and 15 J m 2). The cells were replenished with the original media and colonies were fixed, stained, and counted 10 days later. Cells viability was evaluated by measuring mitochondrial metabolic activity in short-term colorimetric assays (MTT) as described previously [14]. The assays were performed 72 h after treatments with tetrazolium salt (0.25 mg/ml). Where indicated the cells were treated with the histone deacetylase inhibitor trichostatin A (TSA, Sigma, St. Louis, MO) 100 ng/ml for 4 h. Cell morphology was evaluated by phase contrast on an inverted fluorescence microscope (Nikon TE 200, HG-100W mercury lamp) with a 10 Plan fluor objective. Western blots analyses Proteins (50 Ag) extracted from RKO cells treated as indicated were run on 10– 15% SDS – PAGE, transferred on nitrocellulose membrane, and reacted with the different antibodies as follows: anti-p53 mouse monoclonal antibody (Pab421; Oncogene, Cambridge, MA) at a 1:1000 dilution, anti-Gadd45 rabbit polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) at a 1:500 dilution, antiMDM2 mouse monoclonal antibody (Santa Cruz) at a 1:500 dilution, anti-HDAC1 rabbit polyclonal antibody (Santa


M.S. Kim et al. / Experimental Cell Research 306 (2005) 94 – 102

Cruz) at a 1:1000 dilution, anti-Actin polyclonal antibody (Sigma) at a 1:5000 dilution, anti-acetylated H4 rabbit polyclonal antibody (Upstate Biotechnology, Lake Placid, NY) at a 1:2000 dilution, p53 Ser15 phosphospecific polyclonal antibody (Cell Signaling, Beverly, MA) at a 1:500 dilution, anti-PARP rabbit polyclonal antibody (Santa Cruz) at a 1:1000 dilution, anti-BAX rabbit polyclonal antibody (Cell Signaling) at a dilution of 1:1000, anti-BclxL rabbit polyclonal antibody (Cell Signaling) at a 1:1000 dilution, and anti Bcl-2 rabbit polyclonal antibody (Cell Signaling) at a dilution of 1:1000. The blots were reacted for 2 h at room temperature, incubated with their respective secondary antibodies, and detected with chemiluminescence reagents (Amersham) according to the manufacturer’s recommendation. The acid-soluble histone proteins were extracted as described before [14]. Apoptosis DNA ladders formation was detected by staining fragmented DNA with ethidium bromide on 1.5% agarose gel. Fragmented DNA was extracted 4, 12, and 24 h after treatments with the suicide track DNA ladder isolation kit (Oncogene) as recommended by the manufacturer. Free 3VOH were end-labeled with the DeadEnd fluorometric TUNEL system (Promega, Madison, WI) and fluorescence was detected with an inverted fluorescence microscope (Nikon TE 200, HG-100W mercury lamp). The apoptotic index was determined by dividing the number of fluorescent cells (TUNEL positive) by the total number of cells (as determined by conventional DAPI staining) times 100. The average of three different fields was calculated for each treatment. At least 30 cells were counted in each field. Thymine dimers formation RKO cells (1.5  106) were plated on a 10-cm dish and, where indicated, treated with TSA (100 ng/ml) for 4 h and UV (15 J m 2). The cells were harvested 4 h after exposure to UV radiation and genomic DNA was isolated with Qiazol (Qiagen) as described by the manufacturer. DNA (100 ng) was spotted on a nylon membrane (Hybond-N, Amersham Biosciences) with a dot blot apparatus (Fisher). The membrane was then baked at 80-C for 3 h. Formation of thymine dimers was measured with CPDs specific antibodies (mAbs MC-062, Kamiya Biomedical, Tukwila, WA) and chemiluminescence detection of the secondary antibody with an ECL kit (Amersham). The amounts of CPDs were evaluated by densitometry and normalized to the levels of GAPDH genomic DNA in each sample. The amounts of GAPDH were determined by hybridizing the membrane to a 32 P-labeled human GAPDH genomic DNA probe (Ambion) and autoradiography was performed on a phosphorimager (Molecular Dymanics, STORM 820). The relative fold was calculated by dividing the densitometry numbers by the autoradiography numbers times 1000. Dot blots of different

amounts of DNA (0– 150 ng) isolated from cells exposed to 15 J m 2 of UV radiation were also hybridized to demonstrate a linear relationship between DNA on the membrane and chemiluminescence signal as determined by densitometry. The experiment was performed twice in triplicate.

Results The molecular mechanisms underlying UV-induced apoptosis are fairly complex and involved the induction of several key regulatory proteins. To verify the molecular response of the human colon carcinoma RKO cells to UV radiation we exposed the cells to different (0– 40 J m 2) UV doses and performed Western blot analyses. The data shown in Fig. 1 indicate that as expected p53 protein levels increase in response to UV doses [15]. The protein levels of the growth arrest and DNA damage inducible Gadd45 protein also increase in response to UV radiation and closely correlate p53 response. Gadd45 is a p53 downstream effector gene that can promote UV-induced apoptosis in keratinocytes [16]. Another downstream effector of p53 is the mouse double minute oncogene mdm2 [17]. MDM2 can be down-regulated in response to high doses of UV radiation and contributes to the p53-dependent UV-induced apoptosis in human teratocarcinoma [18]. Our data (Fig. 1) indicate that MDM2 protein levels are up-regulated at low UV doses (<5 J m 2, lanes 1– 4) and down-regulated at higher doses (>5 J m 2, lanes 5 – 9). This is in good agreement with previous reports [18 –20] and indicates that the molecular response to UV radiation is conserved in RKO cells. Our goal is to determine whether chromatin modulation through inhibition of histone deacetylase (HDAC) could affect UV-induced apoptosis. To determine whether UV

Fig. 1. Western blots. Total proteins were extracted from RKO cells 4 h after exposure to the indicated doses of UV radiation (J/m2). Detection of the indicated proteins was performed on 50 Ag of total protein extracts as described under Materials and methods. Actin was used as a control for loading.

M.S. Kim et al. / Experimental Cell Research 306 (2005) 94 – 102

radiation itself could affect the protein levels of HDAC we then performed Western blot analyses for HDAC1 after exposure to different doses of UV radiation. The data shown in Fig. 1 indicate that UV radiation, up to 40 J m 2, had no effect on HDAC1 protein levels. Because histone acetylation can modify the accessibility of chromatin DNA [21], we then aimed at determining whether TSA treatments could result in histone acetylation under the conditions used here. Our data (Fig. 2A, lane 2) indicate that TSA treatment for 4 h is sufficient to increase the levels of histone H4 acetylation. The levels of histone H4 acetylation were virtually the same whether TSA was added before (lanes 4 and 7) or after (lanes 5 and 8) UV radiation. Exposure to UV radiation (5 or 15 J m 2) did not affect TSA capacity to increase histone H4 acetylation (lanes 2, 4, 5, 7, and 8). As judged by the Coomassie blue staining of the acid-urea gel (Fig. 2A, bottom panel), the increased levels of histone H4 acetylation were not due to variations in the total amounts of protein loaded. Since the inhibitory effect of TSA on HDAC is transient [22], we also evaluated the effect of TSA 4 h after exposure to UV radiation. The data shown in Fig. 2B indicate that the pattern of histone H4 acetylation is essentially similar, even though not as intense, 4 h after UV radiation than right after treatment (no delay, panel A). These data indicate that TSA treatments result in histone hyperacetylation, which may contribute to open the chromatin structure [21,23] before or after UV radiation. Prolonged (48 h) exposure to high doses (500 ng/ml) of TSA results in morphological changes characteristics of apoptosis [24]. Our goal is to use TSA to relax the chromatin structure rather than trigger an apoptotic response


per se. We thus treated the cells with a lower dose (100 ng/ ml) of TSA for a brief (4 h) period of time. Our data (Fig. 3A, panels a and b) indicate that under these conditions no significant morphological changes occur in RKO cells. Exposing the cells to UV radiation (Fig. 3A, panel c) slightly increased the appearance of smaller and rounder shaped cells that for the most part appeared detached. However, pre-treating the cells with TSA before exposure to UV radiation (Fig. 3A, panel d) markedly increased the appearance of detached smaller, rounder cells. On the other hand, when TSA was added after exposing the cells to UV radiation (e), the smaller rounder cells do not appear to be detached (e). This effect was UV dose dependent since a less pronounced effect was observed at 5 J/m2 (data not shown). These data indicate that brief TSA treatments prior to UV exposure can induce morphological changes characteristic of apoptosis. To determine if the brief TSA treatment could affect the long-term capacity of the cells to survive UV radiation exposure, we then performed clonogenic survival assays. The data shown in Fig. 3B indicate that, as expected, survival decreased in a dose-dependent manner following exposure to UV radiation alone. Pre-treatment of the cells with TSA for 4 h increased by more than threefold the killing efficiency of each UV dose 10 days later. However, when the cells were treated with TSA after exposure to UV radiation, the survival rate was not different than UV radiation alone. The increased killing efficiency could be a result of increased apoptosis. To verify this possibility we first performed a TUNEL assay in RKO cells. The data shown in Fig. 4A indicate that pre-treatment with TSA (Fig. 4A,

Fig. 2. Western blots. (A) The acid-soluble proteins from RKO cells were extracted immediately (no delay) after the indicated treatments. Lane 1, no treatment; lane 2, 100 ng/ml TSA for 4 h; lane 3, UV (5 J/m2); lane 4, TSA first followed by UV (5 J/m2); lane 5, UV (5 J/m2) first followed by TSA; lane 6, UV (15 J/ m2); lane 7, TSA first followed by UV (15 J/m2); lane 8, UV (15 J/m2) first followed by TSA. Even loading of proteins was verified by Coomassie blue staining of the acid-urea gel (bottom panel). (B) Same as A except that the proteins were extracted 4 h after the UV treatment. 10 Ag of proteins was loaded on a 15% SDS – PAGE and acetylated histone H4 was detected as described in Materials and methods.


M.S. Kim et al. / Experimental Cell Research 306 (2005) 94 – 102

Fig. 3. (A) Morphology of RKO cells after the indicated treatments. RKO cells were left untreated (Control, a), exposed to TSA (100 ng/ml for 4 h, b), or UV radiation (15 J/m2) in the indicated order. Cells were examined and photograph under light microscope with a 10 objective 12 h after UV treatment. (B) Clonogenic survival after UV radiation alone (-0-), pre-treatment with 100 ng/ml TSA for 4 h before UV (- -), or UV treatment followed by TSA (-r-) in RKO cells. Colonies were counted 10 days later. Survival is expressed as the percentage of colonies obtained with untreated cells or TSA alone.


panel j) increases the formation of free 3V-OH generated by UV radiation. On the other hand adding TSA after exposure to UV radiation (Fig. 4A, panel l) had no more effect than each treatment alone (Fig. 4A, panels f and h). Formation of DNA ladder was also performed and resulted again in a synergistic effect when TSA was added before UV exposure (data not shown). To determine if this effect was cell type specific we also performed a TUNEL assay in normal human skin fibroblasts. Similar results (Fig. 4B) were obtained with that cell line. Pre-treatment with TSA resulted in a tenfold increase in the apoptotic index (72.52%, SD 17.23) compared to the index obtained with TSA added after UV radiation (7.26%, SD 4.57). The synergistic effect of TSA on the UV-induced formation of 3V-OH (Fig. 4) suggests that the increased apoptosis is most likely due to higher levels of DNA damage. To verify directly the effect of TSA pre-treatment on UV induced DNA damage, we measured the levels of CPDs formed in the genomic DNA. The data in Fig. 5 represent the results of two independent experiments and

indicate that indeed pre-treatment with TSA increases the levels of CPDs formation. To assess the role of p53 in the TSA sensitization to UVinduced apoptosis, we measured the effect of TSA in RKO and RKO-E6 cells. The RKO-E6 cells have a reduced level of p53 due to a stably transfected human papillomavirus 16 E6 protein that degrades p53 [25]. In these cells (Fig. 6A) the level of UV-induced DNA laddering is less (lane 6) and no synergistic effect is observed with TSA pre-treatment (lanes 7– 8). The p53-dependent synergistic effect of TSA was also confirmed by increased cytotoxicity as measured by the mitochondrial metabolic activity (MTT) (Fig. 6B). This effect is not cell type specific since pre-treatment with TSA could also enhance UV-induced cytotoxicity in the breast cancer MCF-7 cells (data not shown), another p53 wild-type cell line [26,27]. Collectively these data indicate that pre-treatment with TSA can sensitize cells to UVinduced apoptosis in a p53-dependent manner. To better understand the role of p53 in TSA sensitization to UV-induced apoptosis, we measured the levels of

M.S. Kim et al. / Experimental Cell Research 306 (2005) 94 – 102


Fig. 4. Detection of free 3V-OH by TUNEL. (A) RKO cells were treated as indicated. UV radiation was with 15 J m 2. Four hours later, the cells were reacted with terminal deoxynucleotidyl transferase (TdT) and fluorescein-labeled and -unlabeled deoxynucleotides. The cell populations are shown in phase-contrast (100; A, panels a, c, e, g, i, k) and with fluorescein filter (panels b, d, f, h, j, l). (B) Malme-3 normal skin fibroblasts were treated as in A except that labeling was performed 20 h after UV exposure. The cell populations are shown with DAPI staining (200; B, panels a, c, e, g, i, k) and with fluorescein filter (panels b, d, f, h, j, l). Control: untreated cells.

p53 and its phosphorylation at Ser15 as well as four of its downstream effector genes, Bax, Bcl-2, Bcl-xL, and Gadd45 in response to the treatments described above. The data shown in Fig. 7 indicate that pre-treatment with TSA (lanes 4 and 7) increased the levels of p53 as well as its phosphorylation at Ser15 beyond the levels obtained with TSA (lanes 2), UV (lanes 3 and 6), or UV followed by TSA (lanes 5 and 8). None of the treatment used had

any significant impact on the expression levels of the proapoptotic BAX protein. On the other hand, pre-treatment with TSA before UV radiation (lanes 4 and 7) decreased the Bcl-2 levels and the UV-induced Bcl-xL levels. Downregulation of Bcl-2 and Bcl-xL, two anti-apoptotic proteins, thus correlates with the levels of p53 phosphorylation and induction. Fig. 7 indicates that the UV-induced Gadd45 protein levels are also enhanced by pre-treatment

Fig. 5. Formation of thymine dimers. Immunoassay for thymine dimmers was performed with CPDs specific antibody (MC-062) on genomic DNA from RKO cells. Control, no treatment; TSA, 100 ng/ml TSA for 4 h; TSAY15J, TSA first followed by UV (15 J m 2); 15J, UV (15 J m 2); 15JYTSA, UV first followed by TSA. The values were normalized to GAPDH content in each sample. Two experiments performed in triplicate are shown.


M.S. Kim et al. / Experimental Cell Research 306 (2005) 94 – 102

Fig. 6. (A) Internucleosomal DNA cleavage. RKO and RKO-E6 cells were treated as indicated. T, 100 ng/ml TSA for 4 h; U, either 5 or 15 J m 2 as indicated; T U U , TSA first followed by UV; T , UV first followed by TSA. The DNA was extracted and run on 1.5% agarose gel and stained with ethidium bromide. M: markers, DNA ladder. (B) Cytotoxicity determined by MTT assay. RKO cells were treated as indicated and assayed 72 h later. C, untreated control; TSA, 100 ng/ml TSA for 4 h; 5J, UV 5 J m 2; T/5J, TSA first followed by UV (5 J m 2); 5J/T, UV first followed by TSA; 15J, UV 15 J m 2; T/15J, TSA first followed by UV (15 J m 2); 15J/T, UV first followed by TSA. Each sample was assayed at least three times and results are expressed as a percentage of MTT reduction compared to untreated sample (C, control, 100%).

with TSA (lanes 4 and 7). Gadd45 protects against UV irradiation-induced skin tumors and promotes apoptosis by maintaining p38 and c-JNK MAPK activation in keratinocytes [16]. In most case apoptosis is accompanied by caspase-3mediated cleavage of the DNA repair associated enzyme Poly(ADP-ribose) polymerase (PARP), generating a 85-kDa fragment [28]. Because PARP is activated by DNA damage and can modify the p53-mediated DNA damage response [29], we then aimed at evaluating the effect of TSA pretreatment on PARP cleavage. The data shown in Fig. 7 indicate that pre-treatment with TSA (lane 7) markedly increased the amount of an 85-kDa fragment when the cells were exposed to 15 J m-2 UV radiation. Again no synergy

was observed when TSA was added after exposure to UV radiation (lanes 6 and 8).

Discussion The initiation of apoptosis following UV exposure is determined by several factors including DNA damage, the activation of the tumor suppressor p53, triggering of cell death receptors, mitochondrial damage, and cytochrome c release [4]. How these different cellular events work together to orchestrate the apoptotic response is still ill defined but DNA damage appears to be the predominant factor determining whether a cell undergoes apoptosis or

Fig. 7. Western blot. RKO cells were treated as in Fig. 2 and total protein extracts were run on 10 – 15% SDS – PAGE. After transfer on a nitrocellulose membrane the proteins we reacted with indicated antibodies. (*) cleaved PARP (p85 kDa).

M.S. Kim et al. / Experimental Cell Research 306 (2005) 94 – 102

Fig. 8. Schematic representation of molecular mechanisms leading to TSA sensitization to UV-induced apoptosis. CPDs, cyclobutane pyrimidine dimmers; DSB, double strand breaks; p38, MAPK p38; AcetH3, acetylated histone H3; p300, acetyltransferase p300. See text for details.

not. The two major classes of DNA lesions produced by UV radiation are cyclobutane pyrimidine dimers and (6-4) photoproducts. Their distribution along genes is nucleotide sequence dependent. In vivo, the frequency of these lesions at specific sites is modulated by nucleosomes and other DNA binding proteins [30]. Histone –DNA interaction in nucleosomes is thus expected to modulate formation of the DNA photoproducts. To verify this possibility we treated human colon carcinoma RKO cells with the histone deacetylase inhibitor TSA either before or after exposure to UV radiation. We choose to study this effect in the RKO cell line because it has a wild-type p53 genotype [26], has conserved the molecular response to UV radiation [18 – 20], and does not undergo apoptosis with TSA treatment alone, at least at low doses (new Figs. 4 and 6). Our goal was to verify that a primed (open) chromatin structure could affect the formation of DNA lesions and consequently UV-induced apoptosis. We reasoned that if TSA is indeed affecting UV-induced apoptosis through chromatin modulation, relaxing the chromatin structure after exposure to UV radiation should not affect apoptosis. The data presented here all support this hypothesis. We observed a synergistic effect on cell survival (Fig. 3B), formation of CPDs (Fig. 5), and apoptosis (Figs. 4 and 6) only when TSA was added before UV radiation. In Fig. 8, we have summarized the molecular events that are likely to underlie the TSA sensitization to UV-induced apoptosis. By acetylating the core histone (Fig. 2) TSA can increase DNA accessibility [21], which in turn facilitates the formation of UV-induced photoproducts (Fig. 5). Upon replication, the photoproducts could result in increased


DNA strand breaks [6], which is sufficient to activate p53 by phosphorylation at Ser15 (Fig. 7) [8]. The increased UVinduced apoptosis is p53 dependent (Fig. 6) and triggers down-regulation of Bcl-2 and Bcl-xL while activates Gadd45 and PARP. PARP cleavage parallels caspase-3 activation [28] as well as internucleosomal DNA fragmentation (Fig. 6), which also depends on caspase-3 activation [31]. A role of caspase-3 in UV-induced apoptosis is best exemplified by embryonic stem cells lacking caspase-3 that are refractory to UV-induced apoptosis [32,33]. Caspase-3 cleavage and eventually PARP are part of the downstream responses to caspase-9 activation [34]. Because p53mediated apoptosis requires caspase-9 activation in response to non-cytokine cellular stress [35 –37] we can conclude that PARP cleavage is a downstream event of the p53 mediated apoptosis cascade. Even though we cannot rule out the possibility that pretreatment with TSA increases the acetylation levels of other proteins including p53, this mechanism is unlikely to be the prime mediator of the sensitization to UV-induced apoptosis since treating the cells in the reverse order (UV radiation first followed by TSA), which still leads to protein acetylation (Fig. 2), had no synergistic effect. However, phosphorylation of p53 at Ser15 facilitates its acetylation at Lys382 presumably by recruiting the acetyltransferase p300 [38]. This capacity to recruit p300 has recently been link to p53 role in the mediation of UV-induced acetylation of histone H3 and consequently UV-induced chromatin relaxation [11]. Pre-treatment with TSA could thus prime the chromatin structure directly through histone acetylation and indirectly through p53 activation. Collectively our data indicate that pre-treatment with TSA can enhance UV-induced apoptosis in a p53-dependent manner. This effect is likely to require a primed (open) chromatin structure and p53 transcriptional as well as nontranscriptional activity. Because many anticancer drugs damage DNA in a similar way as UV radiation [33], the model system applied here could be useful for providing insight into the mechanism of action of drugs targeting DNA or enzymes acting on the DNA.

Acknowledgment This work was supported by a Departmental Research Initiative Fund from the School of Medicine, University of Maryland, Baltimore.

References [1] S.A. Amundson, M. Bittner, A.J. Fornace Jr., Functional genomics as a window on radiation stress signaling, Oncogene 22 (2003) 5828 – 5833. [2] D. Kulms, B. Poppelmann, D. Yarosh, T.A. Luger, J. Krutmann, T. Schwarz, Nuclear and cell membrane effects contribute independently to the induction of apoptosis in human cells exposed to UVB radiation, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 7974 – 7979.


M.S. Kim et al. / Experimental Cell Research 306 (2005) 94 – 102

[3] D. Kulms, E. Zeise, B. Poppelmann, T. Schwarz, DNA damage, death receptor activation and reactive oxygen species contribute to ultraviolet radiation-induced apoptosis in an essential and independent way, Oncogene 21 (2002) 5844 – 5851. [4] D. Kulms, T. Schwarz, Molecular mechanisms of UV-induced apoptosis, Photodermatol., Photoimmunol. Photomed. 16 (2000) 195 – 201. [5] R.D. Wood, H.J. Burki, Repair capability and the cellular age response for killing and mutation induction after UV, Mutat. Res. 95 (1982) 505 – 514. [6] B. Kaina, DNA damage-triggered apoptosis: critical role of DNA repair, double-strand breaks, cell proliferation and signaling, Biochem. Pharmacol. 66 (2003) 1547 – 1554. [7] J. Lips, B. Kaina, DNA double-strand breaks trigger apoptosis in p53deficient fibroblasts, Carcinogenesis 22 (2001) 579 – 585. [8] W.G. Nelson, M.B. Kastan, DNA strand breaks: the DNA template alterations that trigger p53-dependent DNA damage response pathways, Mol. Cell. Biol. 14 (1994) 1815 – 1823. [9] L.J. Hofseth, S.P. Hussain, C.C. Harris, p53: 25 years after its discovery, Trends Pharmacol. Sci. 25 (2004) 177 – 181. [10] K.H. Vousden, X. Lu, Live or let die: the cell’s response to p53, Nat. Rev., Cancer 2 (2002) 594 – 604. [11] C.P. Rubbi, J. Milner, p53 is a chromatin accessibility factor for nucleotide excision repair of DNA damage, EMBO J. 22 (2003) 975 – 986. [12] A.E. Ehrenhofer-Murray, Chromatin dynamics at DNA replication, transcription and repair, Eur. J. Biochem. 271 (2004) 2335 – 2349. [13] F. Thoma, Light and dark in chromatin repair: repair of UV-induced DNA lesions by photolyase and nucleotide excision repair, EMBO J. 18 (1999) 6585 – 6598. [14] M.S. Kim, M. Blake, J.H. Baek, G. Kohlhagen, Y. Pommier, F. Carrier, Inhibition of histone deacetylase increases cytotoxicity to anticancer drugs targeting DNA, Cancer Res. 63 (2003) 7291 – 7300. [15] Q. Zhan, F. Carrier, A.J. Fornace Jr., Induction of cellular p53 activity by DNA-damaging agents and growth arrest, Mol. Cell. Biol. 13 (1993) 4242 – 4250. [16] J. Hildesheim, D.V. Bulavin, M.R. Anver, W.G. Alvord, M.C. Hollander, L. Vardanian, A.J. Fornace Jr., Gadd45a protects against UV irradiation-induced skin tumors, and promotes apoptosis and stress signaling via MAPK and p53, Cancer Res. 62 (2002) 7305 – 7315. [17] U.M. Moll, O. Petrenko, The MDM2-p53 interaction, Mol. Cancer Res. 1 (2003) 1001 – 1008. [18] X. Zeng, D. Keller, L. Wu, H. Lu, UV but not gamma irradiation accelerates p53-induced apoptosis of teratocarcinoma cells by repressing MDM2 transcription, Cancer Res. 60 (2000) 6184 – 6188. [19] M.E. Perry, J. Piette, J.A. Zawadzki, D. Harvey, A.J. Levine, The mdm-2 gene is induced in response to UV light in a p53-dependent manner, Proc. Natl. Acad. Sci. U. S. A. 90 (1993) 11623 – 11627. [20] L. Wu, A.J. Levine, Differential regulation of the p21/WAF-1 and mdm2 genes after high-dose UV irradiation: p53-dependent and p53-independent regulation of the mdm2 gene, Mol. Med. 3 (1997) 441 – 451. [21] T.R. Hebbes, A.L. Clayton, A.W. Thorne, C. Crane-Robinson, Core histone hyperacetylation co-maps with generalized DNase I sensitivity in the chicken beta-globin chromosomal domain, EMBO J. 13 (1994) 1823 – 1830. [22] M. Yoshida, S. Horinouchi, T. Beppu, Trichostatin A and trapoxin: novel chemical probes for the role of histone acetylation in chromatin structure and function, BioEssays 17 (1995) 423 – 430.

[23] M. Yoshida, M. Kijima, M. Akita, T. Beppu, Potent and specific inhibition of mammalian histone deacetylase both in vivo and in vitro by trichostatin A, J. Biol. Chem. 265 (1990) 17174 – 17179. [24] T. Suzuki, H. Yokozaki, H. Kuniyasu, K. Hayashi, K. Naka, S. Ono, T. Ishikawa, E. Tahara, W. Yasui, Effect of trichostatin A on cell growth and expression of cell cycle- and apoptosis-related molecules in human gastric and oral carcinoma cell lines, Int. J. Cancer 88 (2000) 992 – 997. [25] T.D. Kessis, R.J. Slebos, W.G. Nelson, M.B. Kastan, B.S. Plunkett, S.M. Han, A.T. Lorincz, L. Hedrick, K.R. Cho, Human papillomavirus 16 E6 expression disrupts the p53-mediated cellular response to DNA damage, Proc. Natl. Acad. Sci. U. S. A. 90 (1993) 3988 – 3992. [26] M.B. Kastan, Q. Zhan, W.S. el-Deiry, F. Carrier, T. Jacks, W.V. Walsh, B.S. Plunkett, B. Vogelstein, A.J. Fornace Jr., A mammalian cell cycle checkpoint pathway utilizing p53 and GADD45 is defective in ataxia – telangiectasia, Cell 71 (1992) 587 – 597. [27] P.M. O’Connor, J. Jackman, I. Bae, T.G. Myers, S. Fan, M. Mutoh, D.A. Scudiero, A. Monks, E.A. Sausville, J.N. Weinstein, S. Friend, A.J. Fornace Jr., K.W. Kohn, Characterization of the p53 tumor suppressor pathway in cell lines of the National Cancer Institute anticancer drug screen and correlations with the growthinhibitory potency of 123 anticancer agents, Cancer Res. 57 (1997) 4285 – 4300. [28] M. Enari, H. Sakahira, H. Yokoyama, K. Okawa, A. Iwamatsu, S. Nagata, A caspase-activated DNase that degrades DNA during apoptosis, and its inhibitor ICAD, Nature 391 (1998) 43 – 50. [29] M.T. Valenzuela, R. Guerrero, M.I. Nunez, J.M. Ruiz De Almodovar, M. Sarker, G. de Murcia, F.J. Oliver, PARP-1 modifies the effectiveness of p53-mediated DNA damage response, Oncogene 21 (2002) 1108 – 1116. [30] G.P. Pfeifer, Formation and processing of UV photoproducts: effects of DNA sequence and chromatin environment, Photochem. Photobiol. 65 (1997) 270 – 283. [31] R.U. Janicke, M.L. Sprengart, M.R. Wati, A.G. Porter, Caspase-3 is required for DNA fragmentation and morphological changes associated with apoptosis, J. Biol. Chem. 273 (1998) 9357 – 9360. [32] M. Woo, R. Hakem, M.S. Soengas, G.S. Duncan, A. Shahinian, D. Kagi, A. Hakem, M. McCurrach, W. Khoo, S.A. Kaufman, G. Senaldi, T. Howard, S.W. Lowe, T.W. Mak, Essential contribution of caspase 3/CPP32 to apoptosis and its associated nuclear changes, Genes Dev. 12 (1998) 806 – 819. [33] T.R. Dunkern, G. Fritz, B. Kaina, Ultraviolet light-induced DNA damage triggers apoptosis in nucleotide excision repair-deficient cells via Bcl-2 decline and caspase-3/-8 activation, Oncogene 20 (2001) 6026 – 6038. [34] K. Cain, D.G. Brown, C. Langlais, G.M. Cohen, Caspase activation involves the formation of the aposome, a large (approximately 700 kDa) caspase-activating complex, J. Biol. Chem. 274 (1999) 22686 – 22692. [35] M.S. Soengas, R.M. Alarcon, H. Yoshida, A.J. Giaccia, R. Hakem, T.W. Mak, S.W. Lowe, Apaf-1 and caspase-9 in p53-dependent apoptosis and tumor inhibition, Science 284 (1999) 156 – 159. [36] P. Sabbatini, F. McCormick, Phosphoinositide 3-OH kinase (PI3K) and PKB/Akt delay the onset of p53-mediated, transcriptionally dependent apoptosis, J. Biol. Chem. 274 (1999) 24263 – 24269. [37] A.M. Ranger, B.A. Malynn, S.J. Korsmeyer, Mouse models of cell death, Nat. Genet. 28 (2001) 113 – 118. [38] M.B. Kastan, D.S. Lim, The many substrates and functions of ATM, Nat. Rev., Mol. Cell Biol. 1 (2000) 179 – 186.