Research Article Differential Role of Transcription-Coupled Repair in UVB–Induced Response of Human Fibroblasts and Keratinocytes 1 2 1 3 2 Mariarosaria D’Errico, Massimo Teson, Angelo Calcagnile, Tiziana Nardo, Naomi De Luca, 4 4 2 3 1 Chiara Lazzari, Silvia Soddu, Giovanna Zambruno, Miria Stefanini, and Eugenia Dogliotti 1 Department of Environment and Primary Prevention, Istituto Superiore di Sanità, Viale Regina Elena, Rome, Italy; 2Laboratory of Molecular and Cell Biology, Istituto Dermopatico dell’Immacolata, IRCCS, Via dei Monti di Creta, Rome, Italy; 3Istituto di Genetica Molecolare, Consiglio Nazionale delle Ricerche, Via Abbiategrasso, Pavia, Italy; and 4Molecular Oncogenesis Laboratory, Department of Experimental Oncology, Regina Elena Cancer Institute, Rome, Italy which selectively repairs the transcribed strand of active genes (reviewed in refs. 1–3). The dramatic consequences of a defect in NER are reflected in individuals with the inherited syndromes, xeroderma pigmentosum (XP) and Cockayne syndrome (CS). XP patients are sun-sensitive and show a dramatic increase of UVinduced skin cancer incidence. Cell fusion studies have identified seven XP complementation groups (XP-A through XP-G). The XP gene products are all involved in specific steps of the NER process (reviewed in refs. 4–6). More specifically, cells from patients with XP belonging to the A, B, D, F, and G groups are defective in both TCR and GGR, whereas XP-C and XP-E cells are defective only in GGR playing a role as sensors proteins for DNA damage in nontranscribed sequences. Like patients with XP, patients with CS show hypersensitivity to sunlight but have no predisposition to skin cancer. Their distinctive features are severe developmental and neurologic abnormalities as well as premature aging. CS is caused by mutations in either the CSA or the CSB genes that lead to defects in only one pathway, TCR (reviewed in refs. 6, 7). However, additional roles outside NER have been suggested for CS proteins. CSB is involved in chromatin remodeling (8), in general transcription (9–12) and in rRNA synthesis (13). Additionally, CSB seems to be implicated in TCR (14) and GGR (15, 16) of oxidative DNA damage. Like CSB, CSA has been reported to physically interact with TFIIH (17), thus suggesting an additional role in general transcription, but less is known about its function (reviewed in ref. 5). Recently, the CSBdependent translocation of the CSA protein to the nuclear matrix after DNA damage has been described (18). The characterization of the response to UV light of healthy subjects as well as patients with XP and CS have been mainly done by using skin fibroblasts. We have recently shown that the response to UVB damage of normal human keratinocytes differs significantly from that of fibroblasts from the same donors (19), supporting the idea that the response to DNA damage is cell type–specific (20). In addition, evidence has been accumulated that within UV radiation, different types of DNA damage and repair are induced depending on the wavelength (21). In order to learn more about the response to solar radiation of the target cells for skin cancer, we established primary cultures of keratinocytes from normal, XP-C, and CS-A individuals, and we compared UVB-induced effects in these cells with those caused in fibroblasts obtained from the same biopsy. Here, we show that the very efficient repair of UVB-induced damage by keratinocytes (19) is due to a more efficient GGR in this cell type as compared with that of fibroblasts. We also show that a defect in the CSA gene leads to massive apoptosis and lack of p53-mediated mdm2 transactivation in fibroblasts but not in keratinocytes. The efficient GGR of keratinocytes might operate as a back-up system to remove Abstract Most solar radiation–induced skin cancers arise in keratinocytes. In the human epidermis, protection against cancer is thought to be mediated mainly by nucleotide excision repair (NER) of UVB-induced cyclobutane pyrimidine dimers, and by elimination of the damaged cells by apoptosis. NER consists of two subpathways: global genome repair (GGR) and transcription-coupled repair (TCR). Here, we investigate the impact of defects in NER subpathways on the cellular response to UVBinduced damage by comparing primary human keratinocytes and fibroblasts from normal, XP-C (GGR-defective), and CS-A (TCR-defective) individuals. We show that human keratinocytes are more resistant to UVB killing than fibroblasts and present higher levels of UVB-induced DNA repair synthesis due to a more efficient GGR. The CS-A defect is associated with a strong apoptotic response in fibroblasts but not in keratinocytes. Following an UVB dose of 1,000 J/m2, no p53-mediated transactivation of mdm2 is observed in CS-A fibroblasts, whereas the p53-mdm2 circuit is fully activated in CS-A keratinocytes. Thus, in fibroblasts, the signal for apoptosis originates from DNA photoproducts in the transcribed strand of active genes, whereas in keratinocytes, it is largely TCRindependent. This study shows that the response to UVB radiation is cell type–specific in humans and provides the first evidence that a deficiency in TCR has a different impact depending on the cell type. These findings have important implications for the mechanism of skin cancer protection after UVB damage and may explain the lack of skin cancer in patients with Cockayne syndrome. (Cancer Res 2005; 65(2): 432-8) Introduction Chronic exposure to sunlight is a causative factor in skin cancer development. Modifications of DNA and other cellular components by the higher-energy shorter solar wavelengths comprising the UVB spectra (290-320 nm) are the most damaging to the skin. Sunlightinduced cancer develops from damaged epidermal cells as a result of a multistep process initiated by DNA damage. To counteract the catastrophic effects of UV-induced DNA damage, cells possess a repair system which removes UV photoproducts, the nucleotide excision repair (NER) mechanism. NER can operate via two pathways: the global genome repair (GGR), which repairs damage over the entire genome, and the transcription coupled repair (TCR), Requests for reprints: Eugenia Dogliotti, Department of Environment and Primary Prevention, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy. Phone: 39-064990-2580; Fax: 39-064990-3650; E-mail: dogliott@iss.it. I2005 American Association for Cancer Research. Cancer Res 2005; 65: (2). January 15, 2005 432 www.aacrjournals.org Downloaded from cancerres.aacrjournals.org on March 9, 2016. © 2005 American Association for Cancer Research. UVB Response of Human Keratinocytes by terminal transferase-mediated dUTP-biotin nick end labeling (TUNEL) assay using the in situ cell death detection kit (Roche Molecular Biochemicals, Mannheim, Germany) according to the supplier’s instructions. At least 100 cells were scored per experimental point. Apoptosis of normal keratinocytes were also analyzed by a fluorimetric assay for detection of caspase-3 using the Apoalert Caspase-3 Fluorescent Kit (BD Biosciences, Clontech, Palo Alto, CA). Unscheduled DNA Synthesis Analysis. The response to UVB irradiation was analyzed by measuring Unscheduled DNA Synthesis (UDS) as described previously (29). Briefly, cells were exposed to UVB (500-2,000 J/m2), incubated in medium containing 10 ACi/mL [3H]-thymidine (3H-TdR, specific activity 35 Ci/mmol; ICN, Irvine, CA, USA) and fixed 1 or 3 hours later. UDS was determined on autoradiographic preparations by counting the number of grains on at least 50 non–S-phase cells. Western Blotting. Whole-cell extracts were prepared by lysing the cells in radioimmunoprecipitation assay buffer 0.5% NP40 in 50 mmol/L Tris-HCl, 150 nmol/L NaCl (pH 7.4) containing protease inhibitors. Total proteins (100 Ag) were resolved by SDS-PAGE (12.5% or 8% gels), transferred into polyvinylidene difluoride membranes (Millipore, Bedford, MA), and blocked with 5% non fat dry milk in TBS-T. Membranes were incubated with primary antibodies: anti-p53 and p21 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and anti-mdm2 monoclonal antibodies (Ab-1; Oncogene Research Products, Boston, MA, USA). In order to normalize the expression levels of the proteins of interest in the different samples, the membranes were incubated with anti-h actin (Santa Cruz Biotechnology) or anti-a tubulin monoclonal antibodies (TU-01) (Immunological Sciences, Rome, Italy) in the case of fibroblasts, and with anti–14-3-3 ~ (Santa Cruz Biotechnology) in the case of keratinocytes. Detection was through enhanced chemiluminescence (Amersham; Amersham Bioscience, Buckinghamshire, United Kingdom). Quantitation of protein bands was done by using Gel Doc 2002 analysis program (Bio-Rad, Hercules, CA, USA). Indirect Immunofluorescence. Cells (1
104) were cytocentrifuged on slides, fixed with 2% formaldehyde in PBS, permeabilized with 0.25% Triton-X 100 in PBS, and subjected to indirect immunofluorescence with the mouse anti-MDM2 monoclonal antibodies (Ab1; Oncogene Research Products), and FITC-conjugated anti-mouse serum (Cappel, West Chester, PA). Nuclei were counterstained with 1 Ag/mL Hoechst 33258 dye (Sigma). transcription-blocking lesions, thus providing keratinocytes from patients with CS of an efficient protection from skin cancer. Materials And Methods Cell Cultures and UVB Irradiation. Primary human fibroblast and keratinocyte cultures were established from biopsies of unaffected skin obtained from two patients, XP26PV and CS6PV, affected by XP and CS, respectively, and from two age-matched controls, coded KN1RO and KN2RO. By complementation analysis, XP26PV was assigned to the XP-C group (22), whereas CS6PV was assigned to the CS-A group. 5 Some of the experiments were carried out with primary fibroblasts obtained from another CS-A patient, coded CS4PV (23). Fibroblasts were grown in F10 medium supplemented with 10% fetal calf serum. Keratinocytes were cultivated on a feeder-layer of lethally irradiated 3T3-J2 fibroblasts (a gift from H. Green, Harvard Medical School, Boston, MA) and passaged at the stage of subconfluence as previously described (24). Cells were cultured at 37jC in a 10% carbon dioxide atmosphere. All experiments were carried out with cells at passages two to four. Cells were irradiated with UVB (TL20W12 sunlamps, Philips, Monza, Italy) and the doses were determined using a DM-300HA radiometer (Spectronics Corporation, Westbury, NY, USA). Plasmids and Transfections. Cells (5
105) were plated on 60 mm Petri dishes and transfected for 6 hours by the BES-modified calcium phosphate method (25) with 10 Ag of pcDNA3 plasmid (Invitrogen, Life Technologies, SRI, Milan, Italy), or with the same vector in which the cDNA of the human MDM2 gene was subcloned (pcMDM2) and 1 Ag of pBabe-puro, carrying the puromycin resistance gene (26). Cells were selected with 2 Ag/mL of puromycin (Sigma-Aldrich, St. Louis, MO) for 24 hours to transiently enrich for MDM2-expressing cells. Cytotoxicity Assay. For clonal analysis, subconfluent cell cultures were exposed to UVB (50-1,000 J/m2) and then plated at increasing density as a function of UVB dose (500 to 105 cells and from 500 to 8
103 cells per 100 mm dish in the case of keratinocytes and fibroblasts, respectively). Colonies were fixed 14 days later, stained and scored under microscope. The number of colonies in the irradiated samples were expressed as percentages of those in unirradiated samples. Keratinocyte colonies were scored as progressively growing or aborted, as described in ref. 27. Kinetics of Removal of UVB-Induced DNA Photoproducts. Primary human fibroblasts and keratinocytes were irradiated with 1,000 J/m2 of UVB and harvested after different post irradiation incubation times. DNA was extracted with the Qiagen kit (Genenco, Florence, Italy). The level of photoproducts were measured in microtiter plates, coated with protamine sulfate [10-30 ng for cyclobutane pyrimidine dimers (CPD) and 100-400 ng for 6-4 pyrimidine-pyrimidone photoproducts (PP)] using TDM-2 and 6-4 M2 monoclonal antibodies (a kind gift from O. Nikaido, Division of Radiation Biology, Kanazawa University, Kanazawa, Japan) in a standard ELISA technique as previously described (28). Briefly, after extensive washing with 0.05% Tween 20/PBS, DNA were incubated with the antibodies for 90 minutes at 37jC, washed again with 0.05% Tween 20/PBS, and then incubated with biotinylated F(abV)2 fragment of anti-mouse IgG (Zymed, San Francisco, CA; 1:1,000 in PBS) for 90 minutes at 37jC. After washing with 0.05% Tween 20/PBS, DNAs were incubated with horseradish peroxidasestreptavidin (Zymed; 1:10,000 in PBS) for 90 minutes at 37jC, washed with citrate-phosphate buffer, and incubated with 0.4 mg/mL o-phenylene diamine buffer and 0.007% hydrogen peroxide/citrate-phosphate buffer (pH 5) for 30 minutes at 37C. The reaction was stopped with 2 mol/L sulfuric acid. Detection of Apoptosis. The induction of apoptosis was measured in normal and defective cells following exposure at different UVB doses (250-1,500 J/m2). Apoptotic keratinocytes and fibroblasts were detected 5 Results Global Genome Repair Is the Main Determinant for UVB Survival in Human Keratinocytes. Primary cultures of normal, XP-C, and CS-A keratinocytes and fibroblasts were exposed to UVB (50-1,000 J/m2) and cell survival was determined by measuring colony-forming ability. As shown in Fig. 1A, in general, UVB induced a decrease of cell viability in all cell strains but, at equal dose exposure, keratinocytes were more resistant to the lethal effects of UVB than fibroblasts from the same skin biopsy. XP-C and CS-A defects conferred marked hypersensitivity to UV in both fibroblasts and keratinocytes. The survival hierarchy was the same in both cell types (i.e., normal>CS-A>XP-C). These data indicate that the capacity to perform GGR is the main determinant for UVB survival in keratinocytes, as previously reported for fibroblasts. To verify whether the differential sensitivity to UVB of the two cell types might be affected by differences in the level of DNA damage, monoclonal antibodies directed either against CPD or 6-4 PP were used to detect UV photoproducts in cellular DNA. Fibroblasts and keratinocytes were exposed to 1,000 J/m2 of UVB and the amount of CPD and 6-4 PP was determined on the extracted DNA by ELISA using the specific antibodies. The yield of both DNA lesions was approximately 1.5-fold higher in fibroblasts than in keratinocytes but the ratio of CPD to 6-4 PP was similar in the two cell types (Fig. 1B). Although keratinocytes present a lower level of UV photoproducts than fibroblasts, this difference might only partially account for their UVB resistance. M. Stefanini and T. Nardo, unpublished results. www.aacrjournals.org 433 Cancer Res 2005; 65: (2). January 15, 2005 Downloaded from cancerres.aacrjournals.org on March 9, 2016. © 2005 American Association for Cancer Research. Cancer Research Figure 1. A, cell survival after UVB irradiation of human primary fibroblasts (open symbols ) and keratinocytes (closed symbols ) from normal (4, E), XP-C (o, .), and CS-A (5, n) donors. Survival was determined by colony formation assay. The reported values are the mean of at least two independent experiments each done in triplicate with standard error (SE) always < 10%. The values for keratinocytes include both proliferative and aborted colonies. In all keratinocyte strains, the relative number of proliferative colonies was slightly decreased and that of aborted colonies increased after UVB (data not shown) as previously reported [Otto et al. (30)]. B, induction of UV photoproducts in human primary fibroblasts (4) and keratinocytes (E). The levels of CPD and 6-4 PP lesions present on DNA isolated immediately after UVB irradiation (1,000 J/m2) from fibroblasts and keratinocytes were measured by ELISA. Genomic DNAs were monitored for the presence of UV photoproducts by using either anti-CPD (TDM-2) or anti-6-4 PP (6-4M-2) antibodies. The data are the mean of four independent experiments. Bars, SE. Global Genome Repair Is Responsible for the Efficient Repair of UVB Damage by Human Keratinocytes. The speed and efficiency of repair might also play a role in UVB resistance. In a previous study (19), accelerated removal of UV photoproducts, as detected by using specific antibodies, was observed in keratinocytes as compared with fibroblasts. To gain insight into the mechanism of DNA repair, the capacity to perform DNA repair synthesis (UDS) following UVB irradiation was analyzed at single cell level by autoradiography in normal and NER-defective keratinocytes and fibroblasts. UDS is an efficient measure of the overall repair (GGR), whereas the contribution of TCR is low because it occurs in a minor part of the genome. Accordingly, TCRdefective CS fibroblasts (CS-A and CS-B) were characterized by UDS levels similar to those of normal cells. UDS levels observed 1 hour (data not shown) and 3 hours (Fig. 2) after UVB exposure were time-dependent, but the pattern of response in each cell type was substantially the same. In both fibroblasts and keratinocytes, the UDS levels were UVB dose–dependent (Fig. 2A and B). In normal keratinocytes, the UDS values were higher than those measured in fibroblasts. The significance of the increased UDS levels of keratinocytes is strengthened by the observation that, after the same UVB dose, keratinocytes present lower levels of UV photoproducts than fibroblasts (Fig. 1B; ref. 30). Similarly to CS-A fibroblasts, the UDS levels of CS-A keratinocytes approached those observed in normal cells. Therefore, also in the absence of TCR, the repair efficiency of keratinocytes was higher than that of fibroblasts. In contrast, GGR-defective XP-C cells showed a drastic reduction of UDS levels in both cell types. These findings show that an efficient GGR is responsible for the more efficient CPD removal (19) and DNA repair capacity (this study) displayed by keratinocytes as compared with fibroblasts. A Defect in Transcription Coupled Repair Is Associated with a Strong Apoptotic Response in Human Fibroblasts but not in Keratinocytes. A primary mechanism to remove UVB-damaged skin cells is apoptosis. Apoptosis was measured by the TUNEL assay in both normal and NER-defective cells at different times after exposure to a UVB dose of 1,000 J/m2 (Fig. 3A). Massive Cancer Res 2005; 65: (2). January 15, 2005 apoptosis was observed in CS-A fibroblasts already at 12 hours post irradiation and 65% apoptotic cells were scored at 24 hours post-UVB. Similar results were obtained with another CS-A fibroblast strain (CS4PV) that showed 52% apoptotic cells 24 hours post–1,000 J/m2 UVB. Under the same irradiation regimen, no apoptosis was detected in normal fibroblasts that need higher doses to activate the apoptotic program (Fig. 3A). These findings provide further evidence that TCR-defective fibroblasts trigger the apoptotic pathway in response to UV light (31–33). In contrast with fibroblasts, normal keratinocytes exposed to 1,000 J/m2 UVB undergo apoptosis (Fig. 3A), in agreement with previous observations (19). The activation of an apoptotic response was observed in two primary keratinocyte cell lines and confirmed by fluorimetric detection of caspase-3 (data not shown). Surprisingly, the frequency of apoptotic cells in CS-A keratinocytes was significantly lower than that observed in fibroblasts from the same donor. An anticipation of the apoptotic response was observed (maximum level at 6 hours post irradiation) but the level of apoptotic cells at longer post irradiation times only slightly exceeded that observed in normal keratinocytes. XP-C keratinocytes (Fig. 3A) and fibroblasts (data not shown) showed the same apoptotic pattern observed in the corresponding cell type from the normal donor, confirming that UVB damage repaired by GGR is not a signal for apoptosis (23, 31, 34). To further characterize the apoptotic response of keratinocytes to UVB, the induction of apoptosis was analyzed after exposure to UVB doses ranging between 50 and 1,000 J/m2 (Fig. 3B). Normal keratinocytes showed apoptosis at doses as low as 150 J/m2 of UVB, whereas fibroblasts did not activate this process in the analyzed dose range (this study; refs. 19, 33). A higher frequency of apoptotic cells was detected in CS-A keratinocytes when compared with normal cells, particularly at low UVB doses. However, the number of apoptotic cells in CS-A keratinocytes is clearly lower compared with that in CS-A fibroblasts after equal UVB exposure and even more after doses inducing similar amounts of damage (see Fig. 1B). Besides confirming the crucial role of TCR in inducing apoptosis in fibroblasts, these findings provide clear 434 www.aacrjournals.org Downloaded from cancerres.aacrjournals.org on March 9, 2016. © 2005 American Association for Cancer Research. UVB Response of Human Keratinocytes the higher DNA repair capacity of these specialized epidermal cells (19). In XP-C cells, the p53/p21/mdm2 induction profile was very similar to that of the corresponding normal cell type (Fig. 4B), indicating that a defect in GGR does not significantly affect the p53-associated pathway neither in fibroblasts nor in keratinocytes. In the case of CS-A (Fig. 4C), the kinetics of p53 and p21 induction in fibroblasts were similar to those of normal fibroblasts, whereas a lack of mdm2 increase was detected following UVB-induced p53 stabilization. Unexpectedly, in the CSA keratinocytes, the p53-mdm2 circuit was unaffected and mdm2 expression followed the kinetics of p53/p21 induction with a consistent increase after 6 hours. Although caution should be taken when Western blotting is used for quantitative analysis, it is interesting to notice that the levels of p21 in UVB-treated CSA keratinocytes were higher and persisted longer than in normal cells. Because of the pleiotropic role of this protein in keratinocyte survival, cell cycle arrest and differentiation (reviewed in ref. 36) this observation deserves further analysis. These data indicate that the CS-A defect has a cell type– specific effect on p53-associated pathway in response to UVB: the absence of TCR does not affect the p53/mdm2 circuit in keratinocytes as it does in fibroblasts. Figure 2. UVB-induced DNA repair synthesis expressed as mean number of autoradiographic grains/nucleus of fibroblasts (A) and keratinocytes (B ). Fibroblasts (open symbols ), keratinocytes (closed symbols ) from normal (4, E), XP-C (o, .), and CS-A (5, n) donors. evidence that in human keratinocytes, the persistence of DNA lesions in the transcribed strand of active genes contributes but is not the only mechanism that triggers apoptosis. Lack of p53-Mediated mdm2 Transactivation after UVB Exposure in CS-A Fibroblasts but not in Keratinocytes. UV light activates the p53 tumor suppressor gene which controls DNA damage response by transactivating several downstream genes, such as p21 and mdm2 (reviewed in ref. 35). We analyzed the kinetics of induction of the stress response genes p53, p21, and mdm2 after UVB irradiation of NER-defective fibroblasts and keratinocytes and we compared the gene response with that of normal cells. As shown in Fig. 4A, in normal fibroblasts following 1,000 J/m2 of UVB, p53 started to accumulate 2 hours after irradiation and reached a plateau at 12 to 24 hours postirradiation. The p53 response of normal keratinocytes was significantly different from that of fibroblasts. In keratinocytes, p53 levels reached a maximum at 6 hours postirradiation and then drastically decreased to background levels at 12 hours. In both cell types, the time course of p21 induction paralleled that observed for p53, whereas mdm2 induction was shifted in time, reaching the highest level at 24 and 12 hours post-UVB in fibroblasts and keratinocytes, respectively. The accelerated p53 response in keratinocytes as compared with fibroblasts is consistent with www.aacrjournals.org Figure 3. Apoptosis of human primary fibroblasts (open symbols ) and keratinocytes (closed symbols ) from normal (4, E, y), XP-C (.), and CS-A (5, n) donors. Frequency of apoptosis was measured by the TUNEL assay. A, cells were analyzed at different times after irradiation with 1,000 J/m2 UVB. The data relative to two primary keratinocyte cell lines from normal donors are shown for comparison (E, y). B, cells were analyzed 24 hours after irradiation with different UVB doses. The reported values are the mean of three independent experiments with SE always < 10%. 435 Cancer Res 2005; 65: (2). January 15, 2005 Downloaded from cancerres.aacrjournals.org on March 9, 2016. © 2005 American Association for Cancer Research. Cancer Research induction, typically observed in CS fibroblasts following UV exposure, is due to their defect in TCR (i.e., in the removal of transcription blocking lesions). Conversely, the normal p53mdm2 response in UVB-exposed CS keratinocytes supports the hypothesis that this cell type ‘‘tolerates’’ the defect in TCR because of an efficient GGR back-up system. Discussion Cell Survival after UVB Irradiation. UVB is responsible for the most deleterious effects of solar radiation including erythema, immunosuppression, and skin cancer. The majority of skin cancers originate from epidermal keratinocytes. These specialized cells, which are the cellular target of solar radiation, display a higher clonal cell survival following UVB irradiation than the corresponding fibroblasts which are located in the skin in the underlying dermis. As in the case of fibroblasts, UVB survival of keratinocytes seems to reflect mainly the capacity to perform GGR because TCR-defective keratinocytes (CS-A) are less sensitive to UVB damage than GGR-defective keratinocytes (XP-C). It should be taken into account that significantly lower levels (1.5-fold to 2-fold) of UV photoproducts are induced in keratinocytes as compared with fibroblasts under the same UVB dose regimen (ref. 30; this study). The presence of keratin might act as natural shielding against UVB damage thus contributing to keratinocyte UVB resistance. Cell cycle control mechanisms might also be involved because cell cycle progression of keratinocytes is unaltered after UVB doses that determine an abrupt G1-S arrest in fibroblasts (19). Figure 4. Human primary fibroblasts (open symbols ) and keratinocytes (closed symbols ) from normal (A), XP-C (B), and CS-A (C ) donors were UVB-irradiated and quantitative analysis of the level of expression of p53, p21, and mdm2 were done. Protein lysates were obtained from unirradiated cells and after 2, 6, 12, and 24 hours post-UVB exposure to 1,000 J/m2. The samples were probed with anti-p53, anti-p21, and anti-mdm2 antibodies. Appropriate housekeeping genes (h-actin and 14-3-3 ~ for fibroblasts and keratinocytes, respectively) were measured simultaneously. The Western blotting analysis was carried out in parallel on keratinocytes and fibroblasts derived from the same individual; therefore, a comparison of the intensity of the signals is allowed only between keratinocytes and fibroblasts of the same subject. In order to explore the mechanism behind the lack of mdm2 transactivation in CS-A fibroblasts, we decided to verify whether the p53-mdm2 circuit is fully functional in this cell type. CS-A fibroblasts were either transfected with an expression vector carrying the human mdm2 cDNA (pcMDM2) or with an empty vector (pcDNA.3) and then enriched for plasmid-containing cells by puromycin selection. CS-A fibroblasts transfected with pcMDM2 vector expressed high levels of mdm2 as shown by both immunofluorescence and Western blotting (Fig. 5). If the p53-mdm2 circuit of CS-A fibroblasts is functional, the high levels of exogenous mdm2 should lead to inhibition of UVBinduced p53, and therefore to decreased apoptosis. Indeed, at 24 and 48 hours post-1,000 J/m2 UVB, the CS-A fibroblasts expressing mdm2 showed a reduction (approximately 50%) in the frequency of apoptotic cells as compared with fibroblasts transfected with the empty vector (Fig. 5). The restoration of a normal p53-mdm2 circuit in CS-A fibroblasts overexpressing mdm2 provides the first direct evidence that the lack of mdm2 Cancer Res 2005; 65: (2). January 15, 2005 Figure 5. CS-A human fibroblasts were transiently transfected with an expression vector encoding for the human mdm2 protein (pcMDM2) or with the empty vector (pCDNA.3). The cell population, enriched for mdm2 expressing cells by 24 hours puromycin selection, was exposed to 1,000 J/m2 UVB and the frequency of apoptotic cells was measured by TUNEL. Cells transfected with pcMDM2 and pCDNA.3 were analyzed by immunofluorescence with anti-mdm2 antibodies (top ) and the cell extracts analyzed by Western blot with the same antibodies and with anti-tubulin antibodies as a control (bottom, left). The histogram (bottom, right ) reports the frequency of apoptotic cells in the cell population expressing mdm2 and in that transfected with the empty vector at 24 and 48 hours post-UVB. The frequency of apoptotic cells in the untreated cells (mock) is shown for comparison. A representative experiment, out of the three independent experiments done, is shown. 436 www.aacrjournals.org Downloaded from cancerres.aacrjournals.org on March 9, 2016. © 2005 American Association for Cancer Research. UVB Response of Human Keratinocytes keratinocytes were more sensitive to UVB-induced apoptosis than fibroblasts but, similarly to fibroblasts, the XP-C defect did not affect the apoptotic response as compared with normal keratinocytes. A defect in TCR was associated with an anticipation of the apoptotic response after UVB and higher levels of apoptosis at low UVB doses. Similar results have been reported for one strain of CS keratinocytes (33). However, in contrast with fibroblasts, the level of apoptosis in keratinocytes is only partially affected by the lack of TCR. p53 Regulatory Pathway after UVB. What are the factors that govern the apoptotic response in human skin cells? In a previous study, Conforti et al. (23) reported that in CS fibroblasts exposed to UVC, the p53 increase was not followed by mdm2 induction and this phenomenon was associated with the appearance of apoptosis. In this study, we have observed the same response in CS-A fibroblasts following UVB irradiation. mdm2 expression was detected at basal levels but it was not induced by UVB doses that triggered a strong apoptotic response. Unexpectedly, mdm2 induction was detected in CS-A keratinocytes after UVB injury (Fig. 4). These keratinocytes showed significantly lower UVB-induced apoptosis than the corresponding fibroblasts (Fig. 2). Is there a causal link between lack of mdm2 transactivation ( following p53 stabilization) and apoptosis? We provide strong evidence that this is the case. mdm2overexpressing CS-A fibroblasts were less prone to UVB-induced apoptosis than vector-transfected fibroblasts (Fig. 5). This finding shows that mdm2 gene regulation ensures the elimination of heavily damaged cells by apoptosis and that the persistence of transcription-blocking UVB damage in CS-A fibroblasts is responsible for the lack of mdm2 transactivation. Recently, it has been shown that induction of UV-induced gene expression in human cells is subject to a strong gene size constraint (37). mdm2, as well as other large-sized genes encoding negative regulators of p53 would act as molecular dosimeters of irreparable transcription-blocking DNA damage. In the case of keratinocytes, we have shown that apoptosis is GGR-independent and is only partially affected by a defect in TCR. Our findings support the notion that DNA damage is not the only mediator causing UVB-induced apoptosis in keratinocytes (reviewed in ref. 38). UVB directly activates death receptors, including CD95 (39) and tumor necrosis factor receptors (40), and the inhibition of this activation leads to a partial reduction in UVBinduced apoptosis (39). Additionally, UVB-induced intracellular formation of reactive oxygen species accompanied by mitochondrial dysfunction and cytochrome c release, was shown to be involved in the apoptotic program in this cell type (41). Finally, it is important to mention that UV-induced apoptosis can be influenced by cytokines. These mediators are specifically released by keratinocytes during UV exposure and affect apoptosis in a diverse manner. For instance, IL-1 enhances apoptosis (42), whereas IL-12 protects from apoptosis by inducing DNA repair (43). Cell Type–Specific Response to UVB and Skin Cancer. One of the most puzzling disparities between patients with CS and CS mouse models is the difference in cancer incidence. The mouse models are photosensitive and skin cancer–prone (44, 45), whereas patients with CS have photosensitive skin but do not develop skin cancer (7). In particular, CS mouse fibroblasts (45) and keratinocytes (46–48) are UV-sensitive, do not repair CPD in the transcribed strand of active genes, and are extremely sensitive to UV-induced apoptosis. It has been shown that in vivo in mouse keratinocytes, the signal for p53 induction and sunburn formation originates from DNA damage of actively transcribed genes (48). In this study, we show that, as in the rodent situation, CS-A fibroblasts are extremely sensitive to UV-induced apoptosis. This Figure 6. A model for cell type–specific signaling pathways involved in UVB-induced response in humans. Following UVB damage to DNA, p53 is up-regulated and induces the expression of its primary regulator, mdm2. The p53-mdm2 autoregulatory loop is involved in the control of apoptosis. In fibroblasts (left ), the signal for UVB-induced apoptosis involves DNA photoproducts in actively transcribed genes (TCR-dependent), whereas in keratinocytes (right ) this pathway has a minor role due to the presence of an efficient GGR that eventually repairs most CPD from both strands (GGR back-up system). The other pathways that are known to contribute to UVB-induced apoptosis in keratinocytes, including activation of death receptors like CD95 and induction of reactive oxygen species, are indicated. DNA Repair of UVB Damage. In a previous study, we have shown that the repair of CPD, as measured by specific antibodies, is accelerated in keratinocytes as compared with fibroblasts (19). Only 20% of residual CPD were detected on keratinocyte DNA 24 hours after irradiation, whereas 60% of CPD were still present on fibroblast DNA. In this study, by measuring UDS, we confirmed the higher efficiency of keratinocytes in repairing UVB damage when compared with fibroblasts and we were able to ascribe this phenomenon to a more efficient GGR. The skin cells from the XP-C donor showed a similar drastic decrease of UDS levels independently on the cell type, whereas a defect in TCR (CS-A cells) was associated with UDS levels higher in keratinocytes than in fibroblasts. Altogether, our data show that the increased UVB-induced DNA repair synthesis of keratinocytes reflects a more efficient damage recognition/excision from the genome overall. Apoptosis after UVB Exposure. UVB-mediated apoptosis is a highly complex process in which a variety of signaling pathways is involved. Induction of nuclear DNA damage seems to be the predominant pathway. Upon UVB irradiation, p53 is up-regulated proportionally to UV damage inferred to DNA. Accordingly, a rapid decrease in the levels of p53/p21/mdm2 was observed in keratinocytes which repair UV photoproducts more efficiently than fibroblasts (19). p53 is critically involved in the apoptotic signaling in human fibroblasts that, in the absence of TCR (CS-A and CS-B fibroblasts), show p53 induction and massive apoptosis at UVC doses which are significantly lower than those required for normal cells (23, 31, 34), whereas in the absence of GGR (XP-C fibroblasts), the apoptotic response is similar to that of normal cells (23). We confirm these findings after UVB irradiation showing that the absence of TCR (CS-A fibroblasts) determines massive apoptosis at doses where no apoptosis is detected in normal and XP-C fibroblasts. When the apoptotic response of keratinocytes was analyzed, www.aacrjournals.org 437 Cancer Res 2005; 65: (2). January 15, 2005 Downloaded from cancerres.aacrjournals.org on March 9, 2016. © 2005 American Association for Cancer Research. Cancer Research Fig. 6. Altogether, our findings strongly suggest that in humans, protection from skin carcinogenesis does not rely upon a functional TCR. This would explain the lack of skin cancer in patients with CS and account for the disparity in skin cancer rates in patients with CS and CS mouse models. finding confirms that in human fibroblasts, the signal for p53 induction and apoptosis involves DNA photoproducts in actively transcribed genes. Conversely, in human keratinocytes, this pathway does not (or only to a minor extent) contribute to apoptosis because GGR repairs eventually most CPD from both strands. Therefore, in human keratinocytes, the mechanism for elimination of heavily UVB-damaged cells must rely on signaling pathways that are largely TCR/p53-independent. Indeed, it is well known that in keratinocytes UVB-induced apoptosis is mediated by several pathways that do not involve DNA damage as already discussed. A model summarizing the different responses of human keratinocytes and fibroblasts to UVB irradiation is reported in References Acknowledgments Received 5/21/2004; revised 10/30/2004; accepted 11/11/2004. Grant support: MIUR/FIRB (RBNE01RNN7), Italian Ministry of Health (Ricerca Finalizzata) and Associazione Italiana per la Ricerca sul Cancro (AIRC). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1. de Laat WL, Jaspers NGJ, Hoeijmakers JHJ. Molecular mechanism of nucleotide excision repair. Genes Dev 1999;13:768–85. 2. Hoeijmakers JH. Genome maintenance mechanisms for preventing cancer. Nature 2001;411:366–74. 3. Svejstrup JQ. Transcription repair coupling factor: a very pushy enzyme. Mol Cell 2002;9:1151–2. 4. Friedberg EC. How nucleotide excision repair protects against cancer. Nat Rev Cancer 2001;1:22–33. 5. Lehmann AR. DNA repair-deficient diseases, xeroderma pigmentosum, Cockayne syndrome and trichothiodystrophy. Biochimie 2003;85:1101–11. 6. Bootsma D, Kraemer KH, Cleaver JE, Hoeijmakers JHJ. Nucleotide excision repair syndromes: xeroderma pigmentosum, Cockayne syndrome, and trichothiodystrophy. New York: McGraw-Hill, 2002. p. 211–37. 7. Nance MA, Berry SA. Cockayne syndrome: review of 140 cases. Am J Med Genet 1992;42:68–84. 8. Citterio E, Van Den Boom V, Schnitzler G, et al. ATPdependent chromatin remodeling by the Cockayne syndrome B DNA repair-transcription-coupling factor. Mol Cell Biol 2000;20:7643–53. 9. Selby CP, Sancar A. Human transcription-repair coupling factor CSB/ERCC6 is a DNA-stimulated ATPase but is not a helicase and does not disrupt the ternary transcription complex of stalled RNA polymerase II. J Biol Chem 1997;272:1885–90. 10. Tantin D. RNA polymerase II elongation complexes containing the Cockayne syndrome group B protein interact with a molecular complex containing the transcription factor IIH components xeroderma pigmentosum B and p62. J Biol Chem 1998;273:27794–9. 11. van Gool AJ, Citterio E, Rademakers S, et al. The Cockayne syndrome B protein, involved in transcriptioncoupled DNA repair, resides in an RNA polymerase IIcontaining complex. EMBO J 1997;16:5955–65. 12. Balajee AS, May A, Dianov GL, Friedberg EC, Bohr VA. Reduced RNA polymerase II transcription in intact and permeabilized Cockayne syndrome group B cells. Proc Natl Acad Sci U S A 1997;94:4306–11. 13. Bradsher J, Auriol J, Proietti de Santis L, et al. CSB is a component of RNA pol I transcription. Mol Cell 2002; 10:819–29. 14. Le Page F, Klungland A, Barnes D, Sarasin A, Boiteux S. Transcription coupled repair of 8-oxoguanine in murine cells: the ogg1 protein is required for repair in nontranscribed sequences but not in transcribed sequences. Proc Natl Acad Sci U S A 2000;97:8397–402. 15. Dianov G, Bischoff C, Sunesen M, Bohr VA. Repair of 8-oxoguanine in DNA is deficient in Cockayne syndrome group B cells. Nucleic Acids Res. 1999;27: 1365–8. 16. Tuo J, Jaruga P, Rodriguez H, Bohr VA, Dizdaroglu M. Primary fibroblasts of Cockayne syndrome patients are defective in cellular repair of 8-hydroxyguanine and 8hydroxyadenine resulting from oxidative stress. FASEB J 2003;17:668–74. 17. Henning KA, Li L, Iyer N, et al. The Cockayne syndrome group A gene encodes a WD repeat protein that interacts with CSB protein and a subunit of RNA polymerase II TFIIH. Cell 1995;82:555–64. 18. Kamiuchi S, Saijo M, Citterio E, et al. Translocation of Cockayne syndrome group A protein to the nuclear matrix: possible relevance to transcription-coupled DNA repair. Proc Natl Acad Sci U S A 2002;99:201–6. 19. D’Errico M, Teson M, Calcagnile A, et al. Apoptosis and efficient repair of DNA damage protect human keratinocytes against UVB. Cell Death Differ 2003;10: 754–6. 20. de Waard H, de Wit J, Gorgels TG, et al. Cell typespecific hypersensitivity to oxidative damage in CSB and XPA mice. DNA Repair (Amst) 2003;2:13–25. 21. Mathonnet G, Leger C, Desnoyers J, et al. UV wavelength-dependent regulation of transcriptioncoupled nucleotide excision repair in p53-deficient human cells. Proc Natl Acad Sci U S A 2003;100:7219–24. 22. Chavanne F, Broughton BC, Pietra D, et al. Mutations in the XPC gene in families with xeroderma pigmentosum and consequences at the cell, protein, and transcript levels. Cancer Res 2000;60:1974–82. 23. Conforti G, Nardo T, D’Incalci M, Stefanini M. Proneness to UV-induced apoptosis in human fibroblasts defective in transcription coupled repair is associated with the lack of Mdm2 transactivation. Oncogene 2000;19:2714–20. 24. Zambruno G, Marchisio PC, Marconi A, et al. Transforming growth factor-h1 modulates h1 and h5 integrin receptors and induces the de novo expression of the avh6 heterodimer in normal human keratinocytes: implications for wound healing. J Cell Biol 1995; 129:853–65. 25. Chen C, Okayama H. High-efficiency transformation of mammalian cells by plasmid DNA. Mol Cell Biol 1987;7:2745–52. 26. Morgenstern JP, H. L Advanced mammalian gene transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helperfree packaging cell line. Nucl Acids Res 1990;18: 3587–96. 27. Barrandon Y, Green H. Three clonal types of keratinocyte with different capacities for multiplication. Proc Natl Acad Sci U S A 1987;84:2302–6. 28. Mori T, Rinaldy TL, Athwal RS, et al. A xeroderma pigmentosum complementation group A related gene: confirmation using monoclonal antibodies against the cyclobutane dimer and (6-4) photoproduct. Mutat Res 1993;293:143–50. 29. Stefanini M, Lagomarsini P, Giliani S, et al. Genetic heterogeneity of excision repair defect associated with trichothiodystrophy. Carcinogenesis 1993;14:1101–5. 30. Otto AI, Riou L, Marionnet C, et al. Differential behaviors toward ultraviolet A and B radiation of fibroblasts and keratinocytes from normal and DNArepair-deficient patients. Cancer Res 1999;59:1212–8. 31. Ljungman M, Zhang F. Blockage of RNA polymerase as a possible trigger for U.V. light-induced apoptosis. Oncogene 1996;13:823–31. 32. Balajee AS, Proietti DeSantis L, Brosh RM Jr, Selzer R, Bohr VA. Role of the ATPase domain of the Cockayne syndrome group B protein in UV induced apoptosis. Oncogene 2000;19:477–89. Cancer Res 2005; 65: (2). January 15, 2005 438 33. Petit-Frere C, Capulas E, Lowe JE, et al. Ultraviolet-Binduced apoptosis and cytokine release in xeroderma pigmentosum keratinocytes. J Invest Dermatol 2000;115: 687–93. 34. Yamaizumi M, Sugano T. U.V.-induced nuclear accumulation of p53 is evoked through DNA damage of actively transcribed genes independent of the cell cycle. Oncogene 1994;9:2775–84. 35. Lakin ND, Jackson SP. Regulation of p53 in response to DNA damage. Oncogene 1999;18:7644–55. 36. Missero C, D’Errico M, Dotto GP, Dogliotti E. The molecular basis of skin carcinogenesis. Totowa, NJ: Humana Press; 2002. p. 407–26. 37. McKay BC, Stubbert LJ, Fowler CC, et al. Regulation of ultraviolet light-gene expression by gene size. Proc Natl Acad Sci U S A 2004;101:6582–6. 38. Kulms D, Schwarz T. Independent contribution of three different pathways to ultraviolet-B-induced apoptosis. Biochem Pharmacol 2002;64:837–41. 39. Aragane Y, Kulms D, Metze D, et al. Ultraviolet light induces apoptosis via direct activation of CD95 (Fas/ APO-1) independently of its ligand CD95L. J Cell Biol 1998;140:171–82. 40. Rosette C, Karin M. Ultraviolet light and osmotic stress: activation of the JNK cascade through multiple growth factor and cytokine receptors. Science 1996; 274:1194–7. 41. Kulms D, Zeise E, Poppelmann B, Schwarz T. DNA damage, death receptor activation and reactive oxygen species contribute to ultraviolet radiation-induced apoptosis in an essential and independent way. Oncogene 2002;21:5844–51. 42. Kothny-Wilkes G, Kulms D, Luger TA, Kubin M, Schwarz T. Interleukin-1 protects transformed keratinocytes from tumor necrosis factor-related apoptosisinducing ligand- and CD95-induced apoptosis but not from ultraviolet radiation-induced apoptosis. J Biol Chem 1999;274:28916–21. 43. Schwarz A, Stander S, Berneburg M, et al. Interleukin12 suppresses ultraviolet radiation-induced apoptosis by inducing DNA repair. Nat Cell Biol 2002;4:26–31. 44. van der Horst GT, van Steeg H, Berg RJ, et al. Defective transcription-coupled repair in Cockayne syndrome B mice is associated with skin cancer predisposition. Cell 1997;89:425–35. 45. van der Horst GT, Meira L, Gorgels TG, et al. UVB radiation-induced cancer predisposition in Cockayne syndrome group A (Csa) mutant mice. DNA Repair (Amst) 2002;1:143–57. 46. Brash DE, Ziegler A, Jonason AS, et al. Sunlight and sunburn in human skin cancer: p53, apoptosis, and tumor promotion. J Investig Dermatol Symp Proc 1996; 1:136–42. 47. van Oosten M, Rebel H, Friedberg EC, et al. Differential role of transcription-coupled repair in UVB-induced G2 arrest and apoptosis in mouse epidermis. Proc Natl Acad Sci U S A 2000;97:11268–73. 48. Brash DE, Wikonkal NM, Remenyik E, et al. The DNA damage signal for Mdm2 regulation, Trp53 induction, and sunburn cell formation in vivo originates from actively transcribed genes. J Invest Dermatol 2001;117: 1234–40. www.aacrjournals.org Downloaded from cancerres.aacrjournals.org on March 9, 2016. © 2005 American Association for Cancer Research. Differential Role of Transcription-Coupled Repair in UVB− Induced Response of Human Fibroblasts and Keratinocytes Mariarosaria D'Errico, Massimo Teson, Angelo Calcagnile, et al. Cancer Res 2005;65:432-438. Updated version Cited articles Citing articles E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/65/2/432 This article cites 44 articles, 23 of which you can access for free at: http://cancerres.aacrjournals.org/content/65/2/432.full.html#ref-list-1 This article has been cited by 7 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/65/2/432.full.html#related-urls Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at pubs@aacr.org. To request permission to re-use all or part of this article, contact the AACR Publications Department at permissions@aacr.org. Downloaded from cancerres.aacrjournals.org on March 9, 2016. © 2005 American Association for Cancer Research.