Id2 Reverses Cell Cycle Arrest Induced by γ-Irradiation in Human HaCaT Keratinocytes*

Id2 plays a key role in epithelial cells, regulating differentiation, the cell cycle, and proliferation. Because human skin constantly renews itself and is the first target of irradiation, it is of primary interest to evaluate whether such a gene may be regulated in keratinocytes exposed to ionizing radiation. We show here that Id2 is induced in response to γ-irradiation and have investigated the consequence of this regulation on cell fate. Using RNA interference, we observed that Id2 extinction significantly reduces cell growth in human keratinocytes through the control of the G1-S transition of the cell cycle. We have investigated whether the impact of Id2 on the cell cycle may have a physiological role on the cell's ability to cope with radiative stress. Indeed, when Id2 is down-regulated through interfering RNA, cells are more sensitive to irradiation. Conversely, when Id2 is overexpressed, this somehow protects the cell. We propose that Id2 favors reentering the cell cycle after radiation-induced cell cycle arrest to permit the recovery of keratinocytes exposed to ionizing radiation.

Id2 plays a key role in epithelial cells, regulating differentiation, the cell cycle, and proliferation. Because human skin constantly renews itself and is the first target of irradiation, it is of primary interest to evaluate whether such a gene may be regulated in keratinocytes exposed to ionizing radiation. We show here that Id2 is induced in response to ␥-irradiation and have investigated the consequence of this regulation on cell fate. Using RNA interference, we observed that Id2 extinction significantly reduces cell growth in human keratinocytes through the control of the G 1 -S transition of the cell cycle. We have investigated whether the impact of Id2 on the cell cycle may have a physiological role on the cell's ability to cope with radiative stress. Indeed, when Id2 is down-regulated through interfering RNA, cells are more sensitive to irradiation. Conversely, when Id2 is overexpressed, this somehow protects the cell. We propose that Id2 favors reentering the cell cycle after radiation-induced cell cycle arrest to permit the recovery of keratinocytes exposed to ionizing radiation.
Id proteins are negative regulators that inhibit helix loop helix (HLH) 1 transcription factors from binding to DNA through protein-protein interactions and also act as major inhibitors of cell differentiation (1,2). The HLH family of transcription factors comprises more than 200 members, from yeast to human (3). HLH proteins function in the coordination of mammalian cell lineage regulation of gene expression, controlling the cell cycle. An essential role has been established for a number of HLH proteins in hematopoietic, myogenic, pancreatic, and neurogenic cell lineage commitment and cell differentiation. Four main groups of HLH proteins can be distinguished on the basis of the presence or absence of additional functional domains. Members of a distinct subfamily of HLH proteins, the Id proteins, lack a DNA binding region and instead form dimers with other transcriptional regulators, principally those of bHLH type. Such ID-bHLH heterodimers are unable to bind DNA, and hence Id proteins act as dominant negative regulators of bHLH proteins (4). Because most bHLH transcription factors positively regulate sets of genes involved in cell differentiation, the term Id conveniently refers to the ability of these proteins to inhibit both DNA binding and differentiation.
Among Id proteins, Id2 plays a particular role in cell cycle regulation. Indeed, Id2 is also a dominant negative antagonist of pRb. A dual connection links Id2 and pRb: when overexpressed, Id2 overrides the tumor suppressor function of RB, allowing binding of E2F to its target genes, promoting entry into S phase and cell proliferation (5).
Human skin is the first tissue exposed to irradiation, including during radiotherapeutic treatment. It is therefore of primary interest to analyze the consequences of this genotoxic stress on human keratinocytes. Contradictory effects of ionizing radiation on cell differentiation have been described in epithelial cells. For example, ␥-rays increased the calciuminduced differentiation of mouse epidermal cells (6), whereas they induced dedifferentiation and proliferation in normal human skin cells and carcinoma cells as well as in mouse skin (7,8).
Because of its major role in differentiation, we have analyzed whether Id2 protein was involved in the response to ionizing radiation in human keratinocytes. We observed induction of Id2 gene expression. We then further analyzed the biological effect of the overexpression or extinction of Id2 proteins in human keratinocytes and, in response to irradiation, investigated the physiological role of Id2 induction in response to ␥-irradiation.

MATERIALS AND METHODS
Cell Culture and Irradiation-HaCaT is a non-tumorigenic, spontaneously transformed human keratinocyte cell line (9) generously provided by N. E. Fusenig (German Cancer Research Center, Heidelberg, Germany). Cells were grown at 37°C in a 5% CO 2 humidified atmosphere in Dulbecco's modified Eagle's medium containing 4.5 g/liter glucose supplemented with 10% fetal calf serum, 100,000 units/liter penicillin, 50 mg/liter streptomycin, and 200 mM glutamine. HaCaT differentiation was induced by cultivating cells seeded in plastic flasks at 10,000 cells/cm 2 for 10 days as described previously (9). HaCaT cells were irradiated with 2, 6, and 10 Gy from a 60 Co source at a dose rate of 0.3 Gy/min. Irradiated and sham-irradiated cells (used as control) were then returned to the incubator until they were harvested for analysis.
Human primary keratinocytes were obtained from human mammary skin biopsy. Briefly, keratinocytes were isolated by overnight trypsinization in 0.5% trypsin (Invitrogen)/5% penicillin/streptomycin (Eurobio) in phosphate-buffered saline at 4°C, followed by scraping with a scalpel; then cells were cultured in semi-defined KGM2 medium (Clonetics) on flasks coated with collagen type I (Falcon Biocoat) at 37°C and 5% CO 2 . Third passage cultures were used for all experiments (around 12 cell doublings). For the irradiation experiment, cells were seeded at 40,000 cells/cm 2 , reached confluence at day 5, and were further incubated for 3 days after confluence in order to reach a morphological differentiated state. Primary cells were then irradiated with 2 Gy following the same procedure as mentioned above.
Microarray Experiments and Analysis-Microarrays were performed as described previously (10). A complete description of the microarrays used in this study including the protocols for production and postprocessing of slides has been deposited into the GEO data base (www.ncbi.nlm.nih.gov/geo/). This information is available on request under the following GEO accession numbers: GSE1369, GSE1370, GSM22167-GSM22214, and GSM22117-GSM22140.
Plasmid Constructions for Id2 and Lamin A/C Overexpression or Extinction-For overexpression experiments, the Id2 coding region was amplified by PCR using forward primer 5Ј-ggt-cag-cat-gaa-agc-ctt-ca-3Ј and reverse primer 5Ј-atg-aac-acc-gct-tat-tca-gcc-3Ј. The PCR product was inserted into the HindIII and NotI sites of pRc/RSV (Invitrogen) to make pRc-Id2. The Id2 sequence was verified by sequencing. For extinction experiments, short hairpin RNA (shRNA) specific for Id2 and lamin A/C gene was generated as described on-line (sitemaker. umich.edu/dlturner.vectors/files/u6_hairpin_sirna_6_21_02.doc) using the mU6pro vector (12). Briefly, three duplexes of oligonucleotides covering the sequences 5Ј-tgt-gga-cga-ccc-gat-gag-c-3Ј, 5Ј-atc-gcc-ctggac-tcg-cat-c-3Ј, and 5Ј-aga-acc-agc-gct-cca-gga-c-3Ј of the Id2 gene and one duplex of oligonucleotides covering the 5Ј-aac-tgg-act-tcc-aga-agaaca-t-3Ј sequence of the lamin A/C gene were cloned into mU6pro vector digested by BbsI and XbaI to construct shRNA Id2A, Id2B, Id2C, and lamin A/C. Transient Transfections-Transfections were performed using the JetPEI TM transfection reagent (Qbiogene, Carlsbad, CA). The day before transfection, HaCaT cells were seeded at 20,000 cells/cm 2 in a 24-well plate. On the day of transfection, 1 g of plasmid was diluted into 50 l of 150 mM NaCl and added to another 50 l of 150 mM NaCl containing 2 l of JetPEI TM . Fifteen minutes later, the mix was added to the cells together with 900 l of fresh medium. Cells were incubated for 4 h at 37°C and washed with phosphate-buffered saline, and fresh medium was added.
Hoechst 33342 and Propidium Iodide Staining for Microscopic Analysis of Cell Death-One day after transfection, living HaCaT cells were washed once in phosphate-buffered saline and stained with Hoechst 33342 dye (Sigma-Aldrich) and propidium iodide (Sigma-Aldrich) to final concentrations of 1 and 2.5 g/ml, respectively. After a 15-min incubation period, cells were washed once again, and coverslips were mounted for fluorescent staining analysis with an Olympus 1X70 microscope.
Cell Cycle Analysis-Per fraction, 10 5 cells were collected and processed for cell cycle analysis. Cells were fixed and permeabilized in 70% ethanol (Ϫ20°C) at 4°C for 20 min, washed with phosphate-buffered saline and 5% fetal calf serum, and treated with 40 g/ml RNase for 20 min at 37°C. Cells were incubated in 3 g/ml propidium iodide (Sigma-Aldrich) for 5 min at room temperature and placed on ice before flow analysis in a MoFlo (DakoCytomation).
Proliferation Assays-Proliferation assays were performed using the Vialight TM HS test, a bioluminescent assay monitoring ATP present in metabolically active cells (BioWhittaker, Walkersville, MD). One day after transfection, cells were collected and transferred to a 96-well plate at a concentration of 5000 cells/well. Four days later, 100 l of Nucle-FIG. 1. Effect of ␥-rays on Id2 messenger and protein expression in human primary keratinocytes and HaCaT cells. A, differentiated human primary keratinocytes were exposed to a dose of 2 Gy and recovered 3, 6, 15, and 24 h later. Gene expression variations were monitored by microarrays. For each time point, an irradiated sample was hybridized versus its sham-irradiated counterpart. Hybridizations were replicated four times, and significant variations (p Ͻ 0.05) are indicated by an asterisk. B and C, differentiated HaCaT cells were either not exposed (0 h) or exposed to a dose of 2 Gy and recovered 3, 6, 18, and 24 h later. Id2 (B) and c-myc (C) messenger expressions were analyzed in semiquantitative real-time PCR experiments. Total RNA extracted from the different incubation time points was reverse-transcribed and submitted to amplification with primers specific to Id2, c-myc, and 18S rRNA (control). Id2 versus 18S rRNA and c-myc versus 18S rRNA levels of expression are represented as the ratio of their expression calculated at each time point based on real-time PCR efficiency of the primers and the crossing point of each transcript (11). D, Id2 protein expression was analyzed by Western blot. Protein extracts recovered at the indicated times were subjected to immunoblot analysis with anti-Id2 antibodies. The same membrane was re-probed with anti-␣-tubulin antibodies as a loading control. Semiquantitative PCR and Western blots experiments were performed several times, and the most representative one is presented.
otide Releasing Reagent were added to each well. After a 5-min incubation, 180 l of the supernatant were mixed into 20 l of ATP Monitoring Reagent, and the ATP content present in the sample was immediately quantified in a MicroLumatPlus LB96V luminometer (Berthold Technologies, Bad Wildbad, Germany).

RESULTS
Id2 Is Induced in Response to ␥-Irradiation-We investigated global changes in gene expression in response to ␥-irradiation in human primary keratinocyte cells. Microarray results revealed a total of 639 probes shown to be modulated in response to the therapeutic dose of 2 Gy. This represents 9.39% of the probes included in the analysis. The affected genes belonged to several families according to their biological characteristics and function. Among the genes involved in RNA synthesis and modification, Id2 was found to be transiently induced 24 h after irradiation (Fig. 1A). Id2 induction was weak (1.655) but reproducible (p Ͻ 0.05). To further challenge and analyze Id2 response to ␥-radiations, we investigated Id2 expression and function in the HaCaT cell line as a model of the human keratinocyte. We have monitored Id2 expression in response to 2 Gy of irradiation in differentiated HaCaT cells by quantitative PCR. We found that Id2 regulation of expression was biphasic. We observed that Id2 transcript levels slightly decreased (25%) in HaCaT cells 3 h after treatment (Fig. 1B) and then increased significantly (up to 2.5-fold) when compared with untreated cells. Id2 messenger induction began 18 h postirradiation and was still observable 24 h after irradiation. This result was further confirmed at the protein level, with Id2 protein accumulating as soon as 6 h after irradiation (2 Gy) in HaCaT cells (Fig. 1D) and reaching a 2-fold increase after 24 h HaCaT cells plated at low density were transfected with three different shRNAs specific for Id2 and one shRNA specific for lamin A/C. A, 3 days after transfection, protein extracts were subjected to immunoblot analysis with anti-Id2 antibodies. The same membrane was re-probed with anti-lamin A/C antibodies for both loading and lamin A/C extinction controls. B, subconfluent HaCat cells transfected with shRNA-Id2A, shRNA-Id2B, shRNA-Id2C, or shRNA-lamin A/C constructs were seeded in a 96-well plate. Cell number and viability were monitored sequentially at 24, 48, and 72 h in a proliferative assay monitoring ATP content in the cells. Cell number is proportional to the level of ATP expressed in fluorescent units (RLUs). When cell growth was statistically different (p Ͻ 0.01) from the control, an asterisk was placed above the bar. C, subconfluent HaCaT cells were transiently transfected with the constructs shRNA-Id2A and shRNA-Id2C, and cellular phenotype was analyzed the next day. Transfected HaCaT cells were incubated with Hoechst 33342 dye (blue cells) and propidium iodide (red cells) to detect cell death. Cell staining was analyzed immediately. Cell mortality was counted and expressed as a percentage. Cells exposed only to the transfection reagent exhibited a mortality of 2.4%. D, 1 day post-transfection, HaCaT cells were fixed, and DNA content was stained with propidium iodide to determine the cellular cell cycle repartition of cells expressing Id2 (shRNA-Id2C) or lacking Id2 (shRNA-Id2A). The different populations sorted (G 1 , S, and G 2 -M) are expressed as percentages of the total cell population. Cell cycle analysis of non-transfected proliferating HaCaT cells (NT) is also represented to demonstrate the effect of transfection. Experiments were performed several times, and the most representative one is presented. (Fig. 1D). As a positive control of cell response to exposure to ionizing radiation, we monitored c-myc expression. A transient, sharp repression of c-myc expression was detected 3 h after exposure to 2 Gy, followed by a 2-fold increase at 24 h (Fig. 1C), thus exhibiting a pattern similar to that of expression of Id2 in response to irradiation.
Id2, a Regulator of Human Keratinocyte Growth-To determine whether the induction of Id2 was merely a consequence of its up-regulation by transcription factors such as c-myc (5) or whether Id2 induction had a physiological effect in irradiated cells, we knocked down Id2 in human keratinocytes and analyzed their response to irradiation. We blocked Id2 expression by RNA interference in HaCaT cells. We generated three different shRNAs, specific to three regions of the Id2 mRNA sequence, and monitored their gene extinction abilities. As observed in Fig. 2A, the shRNA-Id2A construct was the most effective in silencing Id2 expression, shRNA-Id2B induced only a slight diminution, whereas the plasmid shRNA-Id2C had no effect. This shRNA-Id2C construct was then used as a negative control in subsequent experiments. Expression of lamin A/C was also analyzed to confirm the specificity of the Id2 silencing we observed ( Fig. 2A). Delivery of these shRNAs in HaCaT cells in transient transfection experiments resulted in a decrease in cell growth directly proportional to the degree of Id2 extinction (Fig. 2B). Cell growth was diminished by half in shRNA-Id2Aexpressing cells. Kinetic experiments highlight a decrease of cell proliferation in HaCaT cells expressing low levels of Id2 when transfected with shRNA-Id2A and weak growth in Ha-CaT cells expressing shRNA-Id2B in comparison with shRNA-Id2C and shRNA-lamin A/C (Fig. 2B). These results suggest that Id2 is either essential to maintain cell survival or implicated in cell cycle control. To determine the potential origin of reduced growth in cells lacking Id2, we analyzed cell death and the cell cycle. Staining of dying cells was performed in HaCaT cells expressing the different shRNA constructs. HaCaT cells no longer expressing Id2 showed no particular increase in cell death 24 h post-transfection, as demonstrated by propidium iodide staining and quantification of dead cells (Fig. 2C). Therefore, we investigated the cell cycle distribution of transfected cell populations. The experiment showed a major block (Ͼ80% of cells) in the G 0 -G 1 phase of the cell cycle when Id2 is knocked down (Fig. 2D). These results indicate that Id2 protein is required for HaCaT cell growth and is a key regulator of human keratinocyte cell cycle progression at the G 1 -S transition.
Absence of Id2 Impairs Recovery of Human Keratinocytes Exposed to Ionizing Radiation-Because Id2 is implicated in the control of human HaCaT keratinocyte growth, we investigated the importance of Id2 expression following exposure to ␥-rays. HaCaT cells expressing shRNA-Id2A or shRNA-Id2C as a control were exposed to 2 and 10 Gy, and cell number was monitored for 8 days. Counting experiments revealed that the lack of Id2 was deleterious when HaCaT were exposed to ␥-irradiation. Indeed, cells lacking Id2 grew much slower after 2 and 10 Gy of irradiation compared with control-irradiated cells (Fig. 3A). After 8 days, the interfering RNA was no longer efficient, and non-irradiated cells were able to proliferate. These cells finally reached a cell number equal to that of cells expressing Id2. This further confirmed that the absence of Id2 induced cell cycle arrest, but not cell death. However, irradiated cells were unable to recover from radiative stress and grew normally in the absence of Id2, even after 8 days. Analysis of cells expressing shRNA-Id2A and shRNA-Id2C 4 days after a range of irradiation doses shows that the deleterious effect of the absence of Id2 is proportional to the delivered dose (Fig. 3B) and to the fraction of cells blocked in G 0 -G 1 or G 2 -M of the cell cycle (Fig. 4). These data suggest that Id2 expression could be important to permit the exit from the cell cycle pause induced by ␥-irradiation.
Id2 Overexpression Reverses Irradiation-induced Cell Cycle Arrest-Given that Id2 seems to be implicated in the recovery of cells that have been exposed to ionizing radiation, we next tested the impact of Id2 overexpression on cell growth in response to ␥ irradiation. HaCaT cells were transfected with a pRc-Id2 construct leading to overexpression of Id2 in human keratinocytes or with the empty vector, pRc, as control (Fig. 5).
Cell growth experiments show that Id2 overexpression stimulates the growth rate of HaCaT human keratinocytes. We then tested the effects of ionizing radiation on these cells. A prolif- eration assay performed 4 days post-irradiation revealed that Id2 overexpression forced HaCaT cells to grow even after irradiation at 6 Gy, a dose that normally induces a marked pause in the G 0 -G 1 phase of the cell cycle, as demonstrated in Fig. 4. Indeed, in response to ␥-rays, cells expressing pRc present a clear decrease in cell proliferation at 2, 6, and 10 Gy (Fig. 6). In contrast, cells expressing pRc-Id2 were still growing after 2 Gy of irradiation and even after 6 Gy of irradiation. At the higher dose of 10 Gy, inducing a sharp arrest in the G 2 -M phase of the cell cycle, HaCaT cells expressing either construct were no longer able to grow normally. This experiment demonstrated that Id2 overexpression was able to override the cell cycle arrest induced by ionizing radiation in human keratinocytes. To further investigate the mechanism involved, we monitored p16 and p21 expression in HaCaT cells overexpressing Id2. No significant decrease in the content of these two proteins was observed by Western blot analysis (data not shown), suggesting that p16 and p21 are still present in the cell but unable to promote their growth-inhibitory effect. DISCUSSION A moderate but reproducible induction of the Id2 gene and a similar accumulation of Id2 proteins were observed in human keratinocytes in response to ionizing radiation. To our knowledge, this is the first report of such induction of Id2 in response to ␥-irradiation. This induction is not part of the immediate response to irradiation because Id2 expression decreased slightly at 3 and 6 h after irradiation. We had to wait until 18 h and up to 24 h after 2 Gy of irradiation to measure a 2.5-fold induction of Id2 transcripts. This pattern of expression in response to irradiation is consistent with expression of the Id2 gene that has been described following, for instance, mitogenic stimulation. In these studies, Id2 expression is rapidly induced (within 1-2 h) as part of early response genes, followed by a decline in expression sustained throughout the G 1 phase of the cell cycle and further upregulation as cells enter S phase (13,14).
This induction of Id2 could be a direct consequence of the earlier induction in response to ␥-irradiation of genes such as c-myc. Indeed, myc is induced in response to ␥-irradiation in numerous cells (15)(16)(17) and is known to promote Id2 expression (5). It has been demonstrated that Id2 is a direct target of Myc, and there is good correlation between the expression of N-Myc and Id2 in neuroblastoma-derived cells (5). This hypothesis is supported in our model by the observation that Id2 follows the pattern of c-myc expression in response to ␥-rays.
Trying to unravel the physiological role of Id2 in response to ionizing radiation, we have used interfering RNA (18,19) to suppress Id2 expression in human keratinocytes. In this cell line, when the expression of Id2 was knocked down to 95% of expression in control cells, growth was dramatically reduced. This was not due to cell death, but rather to the arrest of cells in G 0 -G 1 phase of the cell cycle. These results show that in human keratinocytes, Id2 is necessary for correct cell cycle progression and to ensure G 1 -S transition. Even though the cell line we have used has both P53 alleles mutated, our results are consistent with former reports showing the importance of Id2 in cell cycle regulation (1,2). Id2 is a stimulator of the G 1 -S transition able to reverse the cell cycle arrest induced by each of the three members of the pRb family (20) and to enhance cell proliferation in vitro (13,14). This is also in agreement with phenotypes described in mice lacking Id2, showing that Id2 is indispensable to promote correct G 1 -S transition and normal epithelial cell growth (21).
Cells lacking Id2 were more sensitive to irradiation than cells transfected with control shRNA, showing that the presence of Id2 helped cells to cope with radiative stress. To confirm these results, we then forced Id2 expression in keratinocytes and exposed them to ionizing radiation. We observed that Id2overexpressing cells were able to recover from radiative stress and resume growth 4 days after irradiation. These results indicate that in human keratinocytes, Id2 is able to reverse the G 1 arrest induced by ␥-irradiation. Overexpression of Id2 may liberate E2F from the constraint derived from its association with pRb and pRb-related proteins and abolish the growthinhibitory effect of p16 and p21 (20). Other groups also report the ability of human normal keratinocytes to proliferate rapidly after irradiation, in contrast to fibroblasts (22). Authors noted that this difference may account for the high proportion of epithelial cells originating in skin cancer (22). However, one may wonder whether a "forced" re-entry into the cell cycle after irradiation may jeopardize appropriate DNA repair. Indeed, such phenomena have been described in human fibroblast cells overexpressing the c-myc proto-oncogene. These cells exhibited a shorter cell cycle pause in response to ionizing radiation, leading in turn to defective double-stranded break repair and accumulation of chromosomal translocations (23,24). In our study, we show that Id2 accumulates by 18 h post-irradiation, a delay that may permit the vast majority of DNA repair to process efficiently. However, we will need to address this issue in additional studies.
Taken together, our results support a proliferative response of human keratinocytes to ␥-irradiation instead of terminal differentiation, and Id2 is a potential molecular player to promote this response.
Acknowledgments-We thank Jean-Jacques Leplat and Pierre Vaigot for technical assistance with irradiations and cell cycle analysis  FIG. 6. Analysis of radioresistance in HaCaT cells overexpressing Id2. Human keratinocyte HaCaT cells transfected with either pRc or pRc-Id2 for 24 h were plated in a 96-well plate and exposed the next day to 0, 2, 6, and 10 Gy of irradiation. To monitor cell proliferation after exposure to ionizing radiation, the number of viable cells was analyzed after 1 and 5 days using the ATP monitoring test Vialight TM . experiments, respectively. We thank Marine Cegalerba for help in analysis of cell mortality. We thank Dr. D. L. Turner for the generous gift of mU6pro vector for RNA interference experiments. We thank members of Service de Génomique Fonctionnelle for critical reading of the manuscript.