E2F1 regulates autophagy and the transcription of autophagy ... - Nature

Oncogene (2008) 27, 4860–4864

& 2008 Macmillan Publishers Limited All rights reserved 0950-9232/08 $30.00 www.nature.com/onc

SHORT COMMUNICATION

E2F1 regulates autophagy and the transcription of autophagy genes S Polager1, M Ofir1 and D Ginsberg The Mina and Everard Goodman Faculty of Life Science, Bar Ilan University, Ramat Gan, Israel

The retinoblastoma pathway is often inactivated in human tumors resulting in deregulated E2F activity that can induce both proliferation and cell death. Although the role of E2F in apoptosis is well characterized, little is known regarding its putative participation in other cell death pathways. We show here that activation of E2F1 upregulates the expression of four autophagy genes— microtubule-associated protein-1 light chain-3 (LC3), autophagy-related gene-1 (ATG1), ATG5 and damageregulated autophagy modulator (DRAM). E2F1-mediated induction of LC3, ATG1 and DRAM is direct and indeed, endogenous E2F1 can be found bound to regions encompassing the promoters of these genes. Regulation of ATG5 by E2F1 is indirect. Importantly, we demonstrate that E2F1 activation enhances autophagy and conversely, reducing endogenous E2F1 expression inhibits DNA damage-induced autophagy. These studies identify E2F1 as a transcriptional regulator of autophagy, and for the first time establish a role for E2F1 in DNA damageinduced autophagy. Oncogene (2008) 27, 4860–4864; doi:10.1038/onc.2008.117; published online 14 April 2008 Keywords: E2F; autophagy; LC3; ATG1; DRAM

E2Fs are transcription factors best known for their involvement in the timely regulation of gene expression required for cell cycle progression (DeGregori and Johnson, 2006). The product of the retinoblastoma (RB) tumor suppressor gene, pRB, exerts growth suppression mainly via inhibitory interactions with E2Fs (DeGregori and Johnson, 2006). The pivotal role of this RB/E2F relationship in normal cellular proliferation is highlighted by the common incidence in human cancer of mutations in the RB pathway that result in deregulated E2F activity (Sherr, 1996). In addition to being fundamental regulators of proliferation, E2Fs modulate diverse cellular functions, such as DNA repair, differentiation and development (Dimova and Dyson, 2005). At least one member of the E2F family, namely E2F1, can also mediate apoptosis Correspondence: Professor D Ginsberg, The Mina and Everard Goodman Faculty of Life Science, Bar Ilan University, Ramat Gan 52900, Israel. E-mail: [email protected] 1 These authors contributed equally to this study. Received 30 August 2007; revised 21 January 2008; accepted 9 March 2008; published online 14 April 2008

(programmed cell death type I) (Ginsberg, 2002). Programmed cell death (PCD) is an evolutionarily conserved phenomenon that is a necessary and important feature of various cellular processes. Apoptosis is the most characterized type of PCD; however, in recent years, alternative PCD types have been given more attention. In particular, autophagic cell death (PCD type II) is now considered an important cell death pathway that is alternative or complementary to apoptosis (Edinger and Thompson, 2004; Gozuacik and Kimchi, 2007). Autophagy, a vesicular trafficking process that mediates the degradation of long-lived proteins, is stimulated rapidly in response to various stresses, such as nutrient or growth factor deprivation (Klionsky and Emr, 2000). Upon autophagy induction, a two-layered membrane structure containing cytoplasm and organelles is formed, called an autophagosome. Ultimately, autophagosomes fuse with lysosomes leading to degradation of their content, which results in the liberation of amino acids and fatty acids that can be metabolized or recycled to promote cell survival (Baehrecke, 2005). This autophagic machinery can be utilized also for cellular suicide, a caspase-independent form of PCD (Gozuacik and Kimchi, 2007). Thus, autophagy has dichotomous roles in cell death and cell survival and may play either a tumor suppressive or oncogenic part in tumor development. Autophagy is induced following genotoxic stress (Shimizu et al., 2004; Crighton et al., 2007). Exposure of cells to genotoxic agents results commonly in either growth arrest or PCD. The latter is considered to prevent the propagation of mutations and represents a key mechanism for averting malignancy. E2F1 is a major player in the cellular responses to genotoxic stress, in particular, in DNA damage-induced apoptosis (Stevens and La Thangue, 2004). Therefore, we hypothesized that E2F1 may play a role also in DNA damage-induced autophagy. Here, data are presented that indicate that E2F1 activates the transcription of autophagy genes and, more generally, has the capacity to regulate autophagy. To identify novel features of E2F1-induced cell death, we performed a DNA microarray analysis using cells containing an inducible E2F1, ER-E2F1 (Vigo et al., 1999). In addition to known E2F target genes, the MAP1LC3 (LC3; microtubule-associated protein-1 light chain-3) and ATG1 (autophagy-related gene-1, also named ULK1 in human) genes, which encode critical components of autophagy, were noted to exhibit

E2F1 regulates autophagy S Polager et al

4861

elevated mRNA levels in response to E2F1 activation (data not shown). To validate these microarray data and to survey whether E2F1 regulates the expression of other autophagy genes, we investigated the expression of various autophagy genes before and after induction of E2F1. Activation of ER-E2F1 by 4-hydroxytamoxifen (OHT) treatment is associated with significant increases in the mRNA levels of three key regulators of autophagy, ATG1, ATG5 and LC3 (Figure 1a). Similar treatment of parental U2OS cells that does not contain ER-E2F1 did not affect the mRNA levels of these genes (Supplementary Figure 1). ATG1 is a kinase that regulates autophagy induction. ATG5 is part of a ubiquitin-like pathway necessary for formation of the autophagic vesicle, and LC3 is also involved in autophagic vesicle formation. An E2F1-mediated increase in protein levels of ATG5 and LC3 was also detected (Figure 1b). Importantly, expression of the adenovirus oncoprotein E1A, which disrupts RB/E2F complexes, similarly upregulates the expression of ATG1, ATG5 and LC3. This finding suggests that

Figure 1 Transcription of autophagy genes is induced by E2F1. (a) U2OS cells expressing ER-E2F1 were treated with OHT (300 nM) for 16 h ( þ ) or left untreated (). Total RNA was extracted and RT–PCR performed using primers specific for the ATG1, ATG5, LC3 and GAPDH genes. (b) U2OS cells expressing ER-E2F1 were treated with OHT (300 nM) for the indicated times (in hours). Protein levels of LC3, ATG5 and tubulin were assessed by western blot analysis using anti-LC3 (gift of Professor Zvulun Elazar)-, anti-ATG5 (A0731; Sigma)- and anti-tubulin (T9026; Sigma)-specific antibodies. Total levels of LC3 (I þ II) relative to its levels in untreated cells are indicated at the bottom. (c) WI38 cells were infected with a control retrovirus (vec) or a retrovirus expressing E1A (E1A). Total RNA was extracted and RT–PCR performed as in (a). (d and e) U2OS cells (d) and H1299 cells (e) expressing ER-E2F1 were treated with OHT for the indicated times (in hours). Total RNA was extracted and RT–PCR performed using primers specific for the DRAM and GAPDH genes. DRAM, damage-regulated autophagy modulator; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; LC3, microtubule-associated protein-1 light chain-3; OHT, 4-hydroxytamoxifen; RT–PCR, reversetranscription PCR.

endogenous E2Fs are capable of influencing the regulation of these autophagy genes (Figure 1c). Recent studies have identified a gene encoding a lysosomal protein, the DRAM (damage-regulated autophagy modulator) gene, as a critical effector of p53induced autophagy (Crighton et al., 2006). As E2F1, like p53, plays a role in the response to DNA damage, we tested next whether E2F1 affects expression of DRAM. Following activation of ER-E2F1 by the addition of OHT, DRAM mRNA levels are significantly higher (Figure 1d). In contrast to the increases in ATG1, ATG5 and LC3 mRNA levels that are detectable only 8–16 h after E2F1 activation, DRAM mRNA upregulation occurs faster, evident 4 h after E2F1 activation (Figures 1a and d and data not shown). Notably, activation of ER-E2F1 in p53-deficient human non-small cell lung carcinoma cells was found similarly to result in elevated levels of DRAM mRNA (Figure 1e). Taken together, these data support the premise that DRAM transcription is activated by E2F1 in a p53-independent manner. To determine whether E2F1 affects directly the transcription of autophagy genes, we evaluated the influence of cycloheximide, a protein synthesis inhibitor, on transcriptional induction by ER-E2F1. As the ERE2F1 fusion protein is activated by OHT in the absence of protein synthesis, any mRNAs upregulated upon ERE2F1 activation in the presence of cycloheximide represent genes that are direct transcriptional targets of E2F1. DRAM mRNA expression is induced upon ER-E2F1 activation in both the absence and the presence of cycloheximide, indicating that DRAM is a direct target of E2F1 (Figure 2a). In contrast, increase in ATG5 expression in response to ER-E2F1 activation is inhibited by cycloheximide, indicating that ATG5 is not a direct target of E2F1 (Figure 2b). Expression of LC3 is elevated following ER-E2F1 activation in both the absence and the presence of cycloheximide (Figure 2b). However, LC3 mRNA levels increase in response to cycloheximide addition alone (suggesting that LC3 mRNA levels are regulated by a short-lived repressor) and, therefore, it is not possible to determine whether LC3 is a direct target of E2F1 using this method. Inconclusive data were obtained regarding the effect of cycloheximide on ATG1 expression (data not shown). To explore further the possibility that the DRAM, LC3 and ATG1 genes are direct transcriptional targets of E2F, we searched the genomic sequences 1000 bp upstream of these genes for consensus E2F-binding sites. This in silico analysis revealed two and six consensus E2F-binding sites in the LC3 and ATG1 upstream regions, respectively (Figure 2c). No consensus E2Fbinding sites were found in the human DRAM promoter; however, a putative E2F-binding site is present in the first exon of the human DRAM gene at position þ 73 to 81. To further test whether E2F1 regulates DRAM expression via transcriptional activation, a fragment spanning the 867/ þ 111 region of the human DRAM gene was cloned into a luciferase reporter plasmid. Co-transfection of an E2F1 expression plasmid with this reporter into U2OS cells resulted in a 12-fold E2F1-induced activation, indicating that this Oncogene


4862

Figure 2 ATG1, LC3 and DRAM are direct transcriptional targets of E2F1. (a) RT–PCR was performed as in Figure 1 using total RNA extracted from U2OS cells expressing ER-E2F1 and treated with OHT for 8 h ( þ ) or not treated (), in the presence ( þ ) or absence () of 10 mg/ml cycloheximide (CHX). (b) RT–PCR was performed as in Figure 1 using total RNA extracted from U2OS cells expressing ER-E2F1 and treated with OHT for 16 h ( þ ) or not treated (), in the presence ( þ ) or absence () of 10 mg/ml cycloheximide (CHX). (c) A schematic representation of the human ATG1, LC3 and DRAM promoters. The E2F-binding sites are represented as boxes and 8-mer nucleotide sequences. Transcription start sites are indicated by arrows. (d) U2OS cells were transiently co-transfected with a CMV-b-Gal plasmid and either a vector-Luc plasmid or a DRAM-Luc construct together with either empty vector (vector), E2F1 expression vector (E2F1) or an E2F1E132 expression vector. Luciferase and b-galactosidase activities were measured 24 h after transfection. Results are depicted as fold activation after normalization for b-Gal activity. This experiment is representative of three experiments, each performed in duplicate. (e) ChIP analysis was performed using U2OS cells. Crosslinked chromatin was precipitated using antibodies specific to E2F1, E2F4 and HA. Then ATG1, LC3, DRAM and b-actin promoter fragments were amplified by PCR. Input DNA represents 0.5% of total chromatin. ChIP, chromatin immunoprecipitation; DRAM, damage-regulated autophagy modulator; LC3, microtubule-associated protein-1 light chain-3; OHT, 4-hydroxytamoxifen; RT–PCR, reverse-transcription PCR.

region of the DRAM gene is responsive to E2F1 (Figure 2d). An E2F1 mutant that does not bind to DNA, E2F1E132, did not activate this DRAM-luciferase reporter (Figure 2d). To substantiate that endogenous E2Fs regulate expression of ATG1, LC3 and DRAM, we performed chromatin immunoprecipitation analyses using chromatin from proliferating U2OS cells and antibodies directed against E2F1 and E2F4. The promoter fragments amplified were ATG1, 414/169; LC3, 314/ 20 and DRAM, 599/255. We found that endogenous E2F1 and E2F4 are associated detectably with all these promoters (Figure 2e). No binding of endogenous E2Fs was detected to an unrelated genomic fragment (Figure 2e). On the basis of these data, we conclude that Oncogene

the autophagy genes DRAM, LC3 and ATG1 are novel direct transcriptional targets of E2F. Having discovered the capacity of E2F1 to regulate autophagy genes at the transcriptional level, next we set out to check whether E2F1 influences autophagy at the cellular level. We exploited observations that vesicular accumulation of endogenous LC3 is a marker of the autophagic process (Kirisako et al., 1999). Specifically, in most unstimulated cells, LC3 exists in the LC3I form, which exhibits a diffuse distribution across the cytoplasm. In contrast, when autophagy is induced, LC3I is lipidated to become the LC3II form, which integrates into the autophagosome membrane and displays a distinctly punctate distribution (Kabeya et al., 2000). The expression pattern of LC3 was examined before and


4863

after OHT addition in cultured parental tumor cells and in derivatives containing the inducible ER-E2F1. In the absence of OHT, as expected, only occasional puncta were observed, representing a basal level of autophagy. No increase in the number of puncta was observed when parental cells were treated with OHT (Figure 3a). In contrast, following E2F1 activation by addition of OHT to cells containing ER-E2F1, a notable increase in the number of puncta was detected (Figure 3a). This E2F1dependent increase in the number of puncta is modest but reproducible and suggests that E2F1 does indeed regulate autophagy. The modest influence of E2F1 activation on autophagy in this model system implies that E2F1 alone is not sufficient for strong induction of autophagy. However, following ER-E2F1 activation, we could also detect conversion of LC3I into the rapidly migrating lipidated form LC3II (Figure 1b), further demonstrating autophagy regulation by E2F1. Autophagy can be induced by DNA damage (Shimizu et al., 2004; Crighton et al., 2007) and E2F1 is known to be involved in the DNA-damage response (Stevens and La Thangue, 2004). Therefore, we tested whether endogenous E2F1 influences autophagy in response to DNA damage. U2OS cells were infected with retroviruses that encode either two distinct E2F1-specific short hairpin RNAs (shRNAs) or a nonspecific shRNA.

As expected, endogenous E2F1 protein levels are reduced upon infection with the E2F1-specific shRNAcontaining retroviruses (Figure 3c). Also, as expected, treatment of cells with the DNA-damaging agent etoposide results in a marked induction in the appearance of autophagosomes. Importantly, this etoposideinduced increase in the number of autophagosomes is much smaller in cells expressing the E2F1-specific shRNAs than in cells expressing the nonspecific shRNA (Figures 3b and c). Knockdown of E2F4 had no effect on etoposide-induced autophagy (data not shown). Taken together, these data indicate that E2F1, but not E2F4, plays a role in DNA damage-induced autophagy. The observed ability of E2F1-specific shRNA expression to reduce DNA damage-dependent induction of autophagy is consistent with our finding that ectopic expression of E2F1 enhances autophagy. Taken together, these data indicate that endogenous E2F1 plays a role in autophagy, specifically in DNA damageinduced autophagy. The induction of autophagy in response to stress may determine cell fate and therefore, must be tightly regulated. Extensive study of the autophagic process has indeed identified a number of regulatory mechanisms, featuring mainly phosphorylation- and ubiquitination-like events (Gozuacik and Kimchi, 2007). However,

Figure 3 E2F1 regulates autophagy. (a) Parental U2OS cells and U2OS cells expressing ER-E2F1 were treated with OHT for 16 h ( þ ) or not treated (). Cells were then immunostained using an anti-LC3-specific antibody (gift of Professor Zvulun Elazar) and the appearance of LC3 puncta assessed. Cells with 10 or more LC3 puncta were scored as LC3-positive cells and the percentage of LC3positive cells is presented. The data are the average of two independent experiments. In each treatment, at least 200 cells were analysed. *Samples in which a significant difference (Po0.05) exists after E2F1 activation. (b and c) U2OS cells containing either shRNAs specific to E2F1 (pRS-E2F11 and pRS-E2F12), vector (pRS) or nonspecific shRNA (pRS-NS) were treated with 1 mM etoposide for 24 h or not treated. (b) Cells were immunostained with an anti-LC3 antibody to visualize LC3 puncta. (c) Immunostained cells were scored for the appearance of LC3 puncta as in (a), except that in each treatment at least 100 cells were analysed. The results are depicted graphically as the percentage of cells that scored LC3-positive and represent the average of four independent experiments. Lower panel: protein levels of E2F1 and tubulin were assessed by western blot analysis using an anti-E2F1 antibody (Santa Cruz, Santa Cruz, CA, USA; sc-251) and anti-tubulin antibody (T9026; Sigma, Rehovot, Israel). LC3, microtubule-associated protein-1 light chain-3; OHT, 4-hydroxytamoxifen; shRNA, short hairpin RNA. Oncogene


4864

recently, it has become clear that autophagy is regulated not only at the post-translational level, but also at the level of transcription. p53 activation promotes autophagy (Feng et al., 2005) and depends on p53-induced activation of the DRAM gene (Crighton et al., 2006). DRAM transcription is also upregulated by the p53 family member, p73 (Crighton et al., 2007). In addition, c-Myc overexpression induces autophagy (Tsuneoka et al., 2003), and ectopic expression of either c-Myc or E2F-1 leads to an increase in levels of the autophagy inducer smARF (Reef et al., 2006). Furthermore, Macleod and co-workers identified BNIP3 as an E2Fregulated gene that affects hypoxia-induced autophagic cell death (Tracy et al., 2007). Also, E2F1 has been shown to bind with the promoter of the autophagy gene Beclin1, although an effect of E2F1 on Beclin1 expression remains to be demonstrated (Weinmann et al., 2001). Our data provide further support for the premise that autophagy is regulated at the transcriptional level by demonstrating for the first time that E2F1 regulates the expression of four autophagy genes—LC3, ATG1, ATG5 and DRAM. Both E2F1 and p53 upregulate directly DRAM gene expression (Figures 1 and 2 and Crighton et al. (2006)).

Notably, co-regulation by E2F1 and p53 has been documented for several pro-apoptotic genes, including ApafI, PUMA, Noxa and SIVA (Ginsberg, 2002; Fortin et al., 2004; Hershko and Ginsberg, 2004). Moreover, p53 and E2F1 cooperate to mediate apoptosis (Wu and Levine, 1994), but it remains to be seen whether and how they cooperate in inducing autophagy. Our data taken together with the well-established role of E2F1 in apoptosis suggest that E2F1 regulates both apoptosis and autophagy. Thus, it is conceivable that autophagy and apoptosis are co-regulated at the transcriptional level. Future studies must address in more detail the relationships between E2F1-induced autophagy and autophagic cell death and between E2F1-induced autophagy and E2F1-induced apoptosis. Acknowledgements We are grateful to Dalia Hacohen for excellent technical assistance and Yehuda Brody for statistical analysis. We thank Zvulun Elazar for helpful discussions and reagents and Uri Nir for critical reading of the manuscript. This work was supported by grants from the Israel Science Foundation (ISF), The Israel Cancer Association and The Israeli Ministry of Health to DG.

References Baehrecke EH. (2005). Autophagy: dual roles in life and death? Nat Rev Mol Cell Biol 6: 505–510. Crighton D, O’Prey J, Bell HS, Ryan KM. (2007). p73 regulates DRAM-independent autophagy that does not contribute to programmed cell death. Cell Death Differ 14: 1071–1079. Crighton D, Wilkinson S, O’Prey J, Syed N, Smith P, Harrison PR et al. (2006). DRAM, a p53-induced modulator of autophagy, is critical for apoptosis. Cell 126: 121–134. DeGregori J, Johnson DG. (2006). Distinct and overlapping roles for E2F family members in transcription, proliferation and apoptosis. Curr Mol Med 6: 739–748. Dimova DK, Dyson NJ. (2005). The E2F transcriptional network: old acquaintances with new faces. Oncogene 24: 2810–2826. Edinger AL, Thompson CB. (2004). Death by design: apoptosis, necrosis and autophagy. Curr Opin Cell Biol 16: 663–669. Feng Z, Zhang H, Levine AJ, Jin S. (2005). The coordinate regulation of the p53 and mTOR pathways in cells. Proc Natl Acad Sci USA 102: 8204–8209. Fortin A, MacLaurin JG, Arbour N, Cregan SP, Kushwaha N, Callaghan S M et al. (2004). The proapoptotic gene SIVA is a direct transcriptional target for the tumor suppressors p53 and E2F1. J Biol Chem 279: 28706–28714. Ginsberg D. (2002). E2F1 pathways to apoptosis. FEBS Lett 529: 122–125. Gozuacik D, Kimchi A. (2007). Autophagy and cell death. Curr Top Dev Biol 78: 217–245. Hershko T, Ginsberg D. (2004). Up-regulation of Bcl-2 homology 3 (BH3)only proteins by E2F1 mediates apoptosis. J Biol Chem 279: 8627–8634. Kabeya Y, Mizushima N, Ueno T, Yamamoto A, Kirisako T, Noda T et al. (2000). LC3, a mammalian homologue of yeast Apg8p, is localized in autophagosome membranes after processing. EMBO J 19: 5720–5728.

Kirisako T, Baba M, Ishihara N, Miyazawa K, Ohsumi M, Yoshimori T et al. (1999). Formation process of autophagosome is traced with Apg8/Aut7p in yeast. J Cell Biol 147: 435–446. Klionsky DJ, Emr SD. (2000). Autophagy as a regulated pathway of cellular degradation. Science 290: 1717–1721. Reef S, Zalckvar E, Shifman O, Bialik S, Sabanay H, Oren M et al. (2006). A short mitochondrial form of p19ARF induces autophagy and caspase-independent cell death. Mol Cell 22: 463–475. Sherr CJ. (1996). Cancer cell cycles. Science 274: 1672–1677. Shimizu S, Kanaseki T, Mizushima N, Mizuta T, Arakawa-Kobayashi S, Thompson CB et al. (2004). Role of Bcl-2 family proteins in a non-apoptotic programmed cell death dependent on autophagy genes. Nat Cell Biol 6: 1221–1228. Stevens C, La Thangue NB. (2004). The emerging role of E2F-1 in the DNA damage response and checkpoint control. DNA Repair (Amst) 3: 1071–1079. Tracy K, Dibling BC, Spike BT, Knabb JR, Schumacker P, Macleod KF. (2007). BNIP3 is an RB/E2F target gene required for hypoxiainduced autophagy. Mol Cell Biol 27: 06229–06242. Tsuneoka M, Umata T, Kimura H, Koda Y, Nakajima M, Kosai K et al. (2003). c-myc induces autophagy in rat 3Y1 fibroblast cells. Cell Struct Funct 28: 195–204. Vigo E, Muller H, Prosperini E, Hateboer G, Cartwright P, Moroni MC et al. (1999). CDC25A phosphatase is a target of E2F and is required for efficient E2F-induced S phase. Mol Cell Biol 19: 6379–6395. Weinmann AS, Bartley SM, Zhang T, Zhang MQ, Farnham PJ. (2001). Use of chromatin immunoprecipitation to clone novel E2F target promoters. Mol Cell Biol 21: 6820–6832. Wu X, Levine AJ. (1994). p53 and E2F-1 cooperate to mediate apoptosis. Proc Natl Acad Sci USA 91: 3602–3606.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc).

Oncogene