CTSL2 Is a Pro-Apoptotic Target of E2F1 and a Modulator of Histone Deacetylase Inhibitor and DNA Damage-Induced Apoptosis
Total Page:16
File Type:pdf, Size:1020Kb
Oncogene (2014) 33, 1249–1257 & 2014 Macmillan Publishers Limited All rights reserved 0950-9232/14 www.nature.com/onc ORIGINAL ARTICLE CTSL2 is a pro-apoptotic target of E2F1 and a modulator of histone deacetylase inhibitor and DNA damage-induced apoptosis CH Wong1,5,ZWu2,5 and Q Yu1,3,4 Aberrant regulation of the pRB/E2F1 pathway has been invariably linked to inappropriate proliferation and/apoptosis in human cancers. Therefore, understanding the intricacies of the signaling pathway and identification of novel E2F1 targets involved in apoptosis could pave way for new therapeutic manipulation. Here, we identified CTSL2 (cathepsin L2/cathepsin V) as a novel E2F1 target that participates in E2F1-dependent apoptosis. We showed that E2F1 directly binds to CTSL2 promoter and that CTSL2 is regulated by both exogenous and endogenous E2F1. RNAi-mediated depletion of CTSL2 effectively abrogated ectopic E2F1-induced apoptosis, coupled with reduced lysosomal membrane permeabilization (LMP) and mitochondrial membrane depolarization. CTSL2 knockdown also inhibited apoptosis mediated by the endogenous E2F1 activated by DNA damage. Furthermore, we showed that CTSL2 depletion in cancer cells resulted in inhibition of histone deacetylase inhibitor (HDACi)- induced apoptosis, and conversely ectopic overexpression of CTSL2-sensitized cancer cells to HDACi. This study uncovered a novel E2F1 target implicated in LMP and apoptosis activation, as well as in the modulation of HDACi and chemotherapeutic drugs response. Oncogene (2014) 33, 1249–1257; doi:10.1038/onc.2013.72; published online 1 April 2013 Keywords: E2F1; apoptosis; CTSL2 INTRODUCTION Here, through a large-scale gene expression analysis using an E2F transcription factor family members have diverse roles in cellular E2F1-confined system, we identified CTSL2 gene, encoding physiology and homeostasis.1,2 Among them, E2F1 (and perhaps cathepsin like 2 (cathepsin L2/cathepsin V), as a novel E2F3) possesses a unique ability to induce apoptosis,3 which has E2F1 target. We showed that CTSL2 is not only required for been regarded as a fail-safe mechanism to protect cells from E2F1-induced apoptosis by engaging LMP but its level also aberrant oncogenic activation of Rb/E2F1 pathway.4,5 Therefore, modulates the cellular sensitivity to HDACi. These data provide a cancer cells with aberrant activity of E2F1 may preferentially undergo role for CTSL2 as a novel pro-apoptotic target of E2F1, which apoptosisincertaincontextorupondrugtreatment,thusproviding might be potentially important in determining the therapeutic a therapeutic window for tumor eradication. Although it is not response of HDACi. exactly clear how E2F1 possesses biphasic properties on cellular proliferation and death, the critical concentration of E2F1 has been proposed to be a critical regulatingpoint,withhighE2F1level RESULTS preferentially driving cellular proliferation, while an excessive E2F1 Exogenous E2F1-dependent transcriptional profiling identifies level tips the cellular balance to death.6,7 CTSL2 as a potential E2F1 target It is well known that E2F1-induced cell death relies on its ability Induction of E2F1 could lead to transactivation of its downstream to transactivate various target genes of which protein products targets, many of which are potent inducers of apoptosis. are prominent cell death mediators,5,8 in both p53-dependent To systematically identify E2F1 targets, we made use of a and p53-independent manner. For example, E2F1-mediated trans- p53-deficient Saos-2 osteosarcoma cell line that expresses an activation of p19ARF would result in p53 stabilization and exogenous E2F1 fused to the 4-hydroxytamoxifen (4-OHT)- eventual execution of p53-mediated apoptosis,9 whereas the responsive ligand-binding domain of the estrogen receptor (ER). activation of various cell death mediators, such as p73,10,11 In this system, addition of 4-OHT leads to translocation of ER-E2F1 APAF1,12 ASK1,13 caspases14 and pro-apoptotic Bcl-2 family to the nucleus, thereby allowing the induction of E2F1 target members,15,16 may lead to apoptosis in the absence of p53. genes.17 Using this E2F1-confined system, we performed Illumina However, relatively little is known about E2F1 apoptotic targets gene expression analysis and profiled E2F1-dependent gene that are functionally important for the therapeutic response of expression (Figure 1a). Using a twofold cutoff, we identified a anticancer drugs. cluster of genes that are differentially regulated in Saos-2 pBabe 1Cancer Biology and Pharmacology, Genome Institute of Singapore, Agency for Science, Technology and Research, Singapore, Singapore; 2State Key Laboratory of Animal Nutrition, College of Animal Science and Technology, China Agricultural University, Beijing, PRC; 3Department of Physiology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore and 4Cancer and Stem Cell Biology, DUKE-NUS Graduate Medical School of Singapore, Singapore, Singapore. Correspondence: Dr Q Yu, Cancer Biology and Pharmacology, Genome Institute of Singapore, Agency for Science, Technology and Research, 60 Biopolis Street, #02-01, Singapore 138672, Singapore. E-mail: [email protected] 5These authors contributed equally to this work. Received 26 September 2012; revised 3 December 2012; accepted 27 December 2012; published online 1 April 2013 E2F1 target CTSL2 modulates drug-induced apoptosis C Hooi Wong et al 1250 Saos-2 ER-E2F1 pBabe CTSL2: Saos-2 OHT0 8 24 48 0 8 24 48 (hr) 15 * DMSO Saos-2 ER-E2F1 APAF-1 OHT CASP3 CHX OHT - + CCND3 10 CHX+OHT CCNE1 Cyclin E1 Map3k5 EZH2 5 CTSL2 BCL2L11 CCNA1 β Relative mRNA level -Actin CTSA 0 CTSB ER-E2F1 ER-E2F1 E132 CTSK CTSL2 CTSF CTSL2: U2OS CTSE OHT 0 hr CTSL1 15 OHT 24 hr U2OS ER-E2F1 CTSS OHT 48 hr CTSZ * CTSC 10 4-OHT - + CTSW CTSG Cyclin E1 CTSL1 5 CTSO CTSL2 CTSC Relative mRNA level CTSD β-Actin 0 CTSH pBabe ER-E2F1 CTSK3 Figure 1. Ectopic E2F1 induces CTSL2 expression. (a) Gene expression heatmap of representative genes that are differentially regulated in Saos-2 cells expressing pBabe (vector control) or ER-E2F1. Saos-2 cells infected with a retrovirus expressing ER-E2F1 or the empty vector (pBabe) were treated with OHT (300 nM) for various time points as indicated. Red column represents genes with higher expression levels; green column represents genes with lower expression levels. (b) Quantitative real-time PCR (qRT-PCR) and western blot of CTSL2 mRNA and protein level in Saos-2 ER-E2F1 cells or ER-E2F1/E132 cells with or without OHT (300 nM) treatment for 24 h. Saos-2 ER-E2F1 and ER-E2F1/E132 cells were treated with or without cycloheximide (10 mg/ml) for 8 h before OHT treatment. (c) Left: qRT-PCR of CTSL2 mRNA expression in U2OS cells infected with an empty vector control (pBabe) or ER-E2F1 and treated with OHT (300 nM) for 24 h and 48 h. Total RNA was isolated and analyzed as described in Materials and methods. Right: western blot of CTSL2 protein level in U2OS ER-E2F1 cells with or without OHT (300 nM) treatment for 24 h. Results are representative of three independent experiments. *Po0.05. and ER-E2F1 cells. In addition to many previously known E2F1 and protein expression, indicating a causal role for endogenous targets, such as CCNE1, APAF1 and caspase-3, we noticed that E2F1 in regulating CTSL2 (Figure 2b). CTSL2, a previously undescribed target, was robustly induced Endogenous E2F1 can also be activated via protein stabilization following the 4-OHT treatment over different time course by DNA damage.18,19 Consistently, we showed that the treatment (Figure 1a). Of further notice, CTSL2 seemed to be the most of DNA-damaging agents, adriamycin and etoposide (also known highly induced gene by E2F1, as compared with the rest of the as VP-16) in U2OS cells, resulted in marked induction of CTSL2 at cathepsin family member. We, thus, chose to focus on CTSL2, as both mRNA and protein levels accompanying E2F1 accumulation the role for CTSL2 in apoptosis regulation has not been previously (Figure 2c), which was abolished upon E2F1 knockdown studied. The array observation was further validated by quantita- (Figure 2d). Taken together, these data provide substantial tive reverse transcriptase-PCR analysis, which shows the induction evidence that CTSL2 is regulated by E2F1 expressed both of CTSL2 mRNA upon 4-OHT treatment in ER-E2F1-expressing exogenously and endogenously. Saos-2 cells, but not in Saos-2 cells expressing a DNA-binding- To further validate the relationship between E2F1 and CTSL2 deficient ER-E2F1 mutant (E132) (Figure 1b). Moreover, induction expression in a clinically relevant setting, we interrogated a gene of CTSL2 mRNA could not be inhibited by pre-treatment of the expression data set that consists of a panel of 24 paired primary protein synthesis inhibitor, cyclohexamide, indicating that CTSL2 colorectal tumor and matched control samples.20 We found that could be a direct target of E2F1 (Figure 1b). Western blot analysis E2F1 and CTSL2 share a striking similarity in their expression in Saos-2 ER-E2F1 cells treated with 4-OHT also confirmed the pattern in these paired tumor and normal samples (Figure 2e), upregulation of CTSL2 protein concomitant with the known E2F1 with a Pearson correlation of 0.692 (Po0.001). This supports a target Cyclin E (Figure 1b, right panel). A similar observation of direct correlation between E2F1 and CTSL2 expression level in CTSL2 mRNA and protein inductions by E2F1 were also found in patient samples. Taken together, the data support the endogen- U2OS ER-E2F1 cells (Figure 1c). ous regulation of CTSL2 by E2F1 and indicate that CTSL2 might be a clinically relevant E2F1 target in human cancer. CTSL2 is regulated by endogenous E2F1 and DNA damage response E2F1 activates CTSL2 gene promoter activity We next investigated whether CTSL2 expression is also regulated Sequence analysis of the CTSL2 promoter identified a putative by endogenous E2F1. To this end, we began with the human lung E2F1-binding site (50-TTT(C/G)GGGC-30) in the proximal promoter fibroblast cells IMR90 and IMR90 cells transformed with oncopro- region (50-TTTCGCTC-30, at position À 1180/ À 1173) (Figure 3a).