Background
Recurrence and metastasis of hepatocellular carcinoma (HCC) depend on the persistent proliferative activity of cancer cells. BIRC5, also called Survivin, has been shown to play a pivotal role in cancers by influencing cell division and proliferation and by inhibiting apoptosis [
1]. Many studies using clinical specimens have shown that BIRC5 is invariably overexpressed in a majority of human cancers and is linked to poor patient prognosis but is rarely expressed in normal tissues [
2]. Based on the abnormally high activation of BIRC5 during carcinogenesis in various types of cancers, treatment that targets BIRC5 has been increasingly recognized as a promising therapy for various cancers. However, when the anti-BIRC5 agent is used alone, the long-term efficacy remains uncertain and is variable for different types of cancers; tumors have always relapsed and regrown in later stages after treatment. Many factors are involved in the regulation of BIRC5 expression and function, and all of these factors influence the efficacy of BIRC5-targeting strategies. We have found that P16 reactivation in HCC cells down-regulates BIRC5 expression and limits CDK4 import into nuclei, and then exhibits the effect of cell cycle arrest and the induction of detachment-induced apoptosis [
3]. Another research group has reported that the octamer-binding transcription factor 4 (OCT4) regulates BIRC5 expression, which was dramatically decreased in OCT4 knockdown murine embryonic stem cells [
4].
OCT4, a member of the POU-domain transcription factor family, plays a pivotal role in the regulation and maintenance of the cellular pluripotent state [
5,
6]. More recently, the expression of OCT4 in human cancer cells has been demonstrated [
7‐
9]. OCT4 activates the transcription of downstream target genes via its octamer motif (5
′-ATGCAAAT-3
′) [
10], and various genes have been reported to have the OCT4 binding sites, including fibroblast growth factor 4 (FGF-4) [
11]. Both global chromatin immunoprecipitation assays and global expression profiling have been used to characterize the gene regulatory network governed by OCT4, and a large list of candidate target genes, whose regulatory sequences are recognized by OCT4, has been generated [
12,
13]. OCT4 also can bind to other similar sequence motifs [
13,
14]. However, many downstream target genes do not have OCT4 motifs but might still be valid candidates as putative indirect targets of OCT4. The BIRC5 promoter was reported to not have binding sites for OCT4, although OCT4 knockdown in murine embryonic stem cells has been shown to decrease the expression level of BIRC5 protein, suggesting an indirect effect of OCT4 on BIRC5 expression [
4].
In this study, we investigated the regulatory mechanism and significance of OCT4 on BIRC5 and CCND1 expression in HCC. Although the roles for BIRC5 and OCT4 in cancers are well-recognized in a number of previous studies, we gave the first evidence that OCT4 indirectly manipulates the expression and function of BIRC5, and also directly upregulates the expression of CCND1. These factors collude to promote cancer cell proliferation and resist cancer cell apoptosis. This innovative finding provides new insight into the regulation of OCT4 on CCND1 expression through a previously unidentified mechanism and indicates a variety of novel biological and prognostic markers, as well as potential therapeutic targets, for cancer diagnosis and treatment.
Methods
Vectors and adenoviruses
Vectors expressing the specific small hairpin RNA (shRNA), including BIRC5-shRNA, OCT4-shRNA and Dual-shRNA, were synthesized by and purchased from Wuhan Genesil Biotechnology Co., Ltd. (Wuhan, China). The 19-nt sense DNA of BIRC5-shRNA (5′-GAAAGTGCGCCGTGCCATC-3′) targets base pairs 436–454 of the BIRC5 gene (HSU75285), and the 19-nt sense DNA of OCT4-shRNA (5′-CCCTCACTTCACTGCACTG-3′) targets base pairs 1233–1253 of the OCT4 gene (DQ486513.1). Both gene elements were controlled by the U6 promoter. A mock control shRNA vector (Ctrl-shRNA, 5′-GACTTCATAAGGCGCATGC-3′) was concomitantly constructed.
Full-length OCT4 cDNA was cloned into pDC315 (Microbix Biosystems, Ontario, Canada) at the EcoRI and SalI sites to generate pDC315-OCT4. Sequences of the shRNA loop were digested from shRNA vectors and then inserted into pDC315 at the BamHI and SalI sites to obtain pDC315-shBIRC5, pDC315-shOCT4 and pDual-shRNA. The plasmids pDC315-OCT4, pDC315-shBIRC5, pDC315-shOCT4 and pDual-shRNA were transfected into HEK293 cells (Microbix Biosystems, Ontario, Canada) using the Lipofectamine 2000 reagent (Invitrogen Corporation Shanghai Representative Office, Shanghai, China) together with the type 5 adenovirus packaging plasmid pBHGloxdelE13cre (Microbix Biosystems, Ontario, Canada) to generate a set of adenoviruses named Ad5-OCT4, Ad5-shBIRC5, Ad5-shOCT4 and AdDual-shRNA.
The luciferase plasmid pSRVN-Luc, in which luciferase expression was under the control of the BIRC5 promoter (nucleotides 1824–2800, GenBank U75285), was kindly provided by Himanshu Garg (Center of Excellence for Infectious Disease, Texas Tech University Health Sciences Center, TX). The BIRC5 promoter was amplified with the indicated primers (P1: 5′-cgGCTAGCcatagaaccagag-3′; P2: 5′-gaAGATCTgccgccgccgccacct-3′) and inserted into an EGFP plasmid at the NheI and BglII sites to yield pSRVN-EGFP.
The wild type CCND1 promoter (wPro; nucleotides 2501–3178, GenBank Z29078.1) was amplified from HepG2 genomic DNA with the indicated primers (P3: 5′-cgGGATCCagattctttggccgtctgtc-3′; P4: 5′-cgGAATTCAAGCTTggctggggctcttc ctg-3′) and then inserted into the luciferase plasmid at the BglII and HindIII sites to generate pGL3wPro-Luc. The base pairs “ATTTGCAT” in wPro from −252 to −245 were replaced by base pairs “ATCTGTAT” and “ATTTGAAATGCAAAT (PORE motif)” to construct the motif-mutated promoter (mPro) and motif-enhanced promoter (ePro), respectively. The mPro-controlled luciferase plasmid (pGL3mPro-Luc) and the ePro-controlled luciferase plasmid (pGL3ePro-Luc) were generated.
Animal experiments
Hep3B cells were subcutaneously injected into the right flanks of BALB/c (nu/nu) mice (107 cells per mouse) (Shanghai Experimental Animal Center, Chinese Academy of Sciences, Shanghai, China) to establish xenograft tumors. Ten weeks later, mice were separated randomly into 4 groups (Dual-shRNA, BIRC5-shRNA, OCT4-shRNA and blank control groups) with 5 mice per group. Mice in the virus-treated groups were given 5 viral intratumoral injections once every other day for a total dosage of 109 pfu virus per mouse. Mice in the control group were given the same volume of viral preservation solution (10 mmol/L Tris–HCl pH 8.0, 2 mmol/L MgCl2, 4% sucrose). Tumor size was measured regularly, and the tumor volume was estimated with the formula “a × b
2 × 0.5”, in which a and b represent the maximal and minimal diameters, respectively. The animal welfare guidelines for the care and use of laboratory animals were approved by the Animal Care Committee of Second Military Medical University (No. SCXK2009-0003).
Statistical analysis
The experimental data were statistically analyzed using the student's
t test, and two-way analysis of variance (ANOVA) according to the properties of the data. All tests were performed using the PASW Statistics 18.0 software.
P < 0.05 was considered statistically significant. The detailed methods is available at Journal’s website as Additional file
1: Supplementary Methods.
Discussion
As a member of the inhibitors of apoptosis protein (IAPs) family, BIRC5 is preferentially expressed in human cancer cells and has multiple functions, including the inhibition of cell apoptosis [
1], control of the cell cycle [
15,
16], promotion of tumor angiogenesis [
17,
18], resistance to chemotherapy or radiotherapy [
19], acceleration of metastasis and recurrence [
20,
21], and regulation of cancer cell autophagy [
22], all of which favour cancer cell survival and tumor maintenance. Therefore, multiple strategies have been employed to target BIRC5 for cancer therapy by silencing BIRC5 expression with small interfering RNA [
23] or antisense oligonucleotides [
24], inhibiting the BIRC5 promoter activity with small-molecule antagonists [
25], and interfering BIRC5 function with dominant-negative mutant forms of the protein [
26]. Some of these strategies are being applied in clinical trials at various phases, and the initial results are promising when combined with other treatments, such as chemotherapy or radiotherapy [
16,
27]. Although certain strategies for cancer therapy targeting BIRC5 have shown a varied extent of antitumor efficacy, the potential benefit of single anti-BIRC5 treatment in different types of cancers is uncertain. Although the down-regulation of BIRC5 expression by anti-BIRC5 agents can inhibit the growth of cancer, tumors consistently obtain growth capabilities in later stages, demonstrating that this treatment approach remains poorly characterized and requires further study.
BIRC5 expression is precisely regulated at transcriptional and post-translational levels. The signal transducer and activator of transcription 3 (Stat-3), β-catenin-activated T-cell factor (TCF) transcription factor, hypoxia-inducible factor-1 alpha (HIF-1
α) and Sp1 transcription factor promote BIRC5 expression by increasing BIRC5 promoter activity [
28‐
31]. Sp1-mediated BIRC5 expression can be suppressed by p53 [
32]. The stability of BIRC5 protein represents another potential method of regulating BIRC5 function. BIRC5 protein is phosphorylated at Thr34 by cdc2 kinase, which prevents BIRC5 proteosome-mediated clearance or degradation [
33]. Recently, OCT4 was reported to have a regulatory effect on BIRC5 expression [
4].
OCT4 belongs to the family of POU-domain transcription factors, which are involved in the regulation of cell growth and differentiation in a variety of tissues [
11,
34]. Many studies have shown that OCT4 expression is restricted to germline and pregastrulation embryos and also to embryonal carcinomas and testicular germ cell tumors [
7], but not expressed in mature somatic cells. Further evidence has shown that some cancer cells, such as breast, bladder, prostate, liver, head and neck squamous cell cancer and non-small cell lung cancers, are positive for OCT4 expression [
7‐
9,
35‐
39]. Therefore, OCT4 acts as a multifunctional factor not only in stem cells but also in many cancers, and the expression of OCT4 causes more malignant histological phenotypes, including rapid progression, great metastasis, and short cancer-related survival. However, one study unexpectedly found that adult human peripheral blood mononuclear cells, which are genetically stable and mainly terminally differentiated cells with a limited lifespan, express OCT4; this finding challenges the paradigm of OCT4 as a marker of pure stem cells and provides novel insight into the role of OCT4 in fully differentiated cells [
11]. OCT4 functions by directly or indirectly activating a series of downstream target genes. By characterizing the genes in OCT4-mediated regulatory networks, it has been found that many candidate target genes that are directly regulated by OCT4 have an OCT4-binding octamer motif [
10]. However, a large number of target genes, such as BIRC5 [
4], have no OCT4 motifs and might be indirectly regulated by OCT4. Therefore, transcriptional regulation of target genes by OCT4 is very complicated, and it is necessary to understand the key gene network that maintains cell pluripotency in embryo development and governs cell differentiation and proliferation in cancer progression.
To clarify the relationship between OCT4 and BIRC5 in HCC, we first analyzed the OCT4 and BIRC5 expression levels in HCC cell lines, including Hep3B, HepG2, PLC/PRF5, SMMC-7721, BEL-7402 and BEL-7404. All cell lines were positive for BIRC5 expression, although only the Hep3B, HepG2 and PLC/PRF5 cells were positive for OCT4 expression; SMMC-7721 cells were weakly positive for OCT4 expression. OCT4 and BIRC5 expression was also investigated by immunohistochemistry in 49 pairs of cancer and liver tissues taken from HCC patients. They were overexpressed in HCC compared with the corresponding liver tissues (Additional file
2: Table S1). BIRC5 immunostaining was mainly localized in cancer cell cytoplasm and nuclei, and OCT4 expression was localized in cancer cell nuclei (Additional file
3: Figure S1).
We found that the expression levels of OCT4 in HCC cell lines were consistent with the percentages of CD133-positive cells, suggesting that OCT4 expression might be related to CD133 expression. By manipulating the expression of OCT4 and BIRC5, we found that BIRC5 expression silencing did not influence OCT4 expression in Hep3B and BEL-7404 cells. However, down-regulation of OCT4 expression inhibited BIRC5 expression, even in the OCT4-negative BEL-7404 cells, and increasing OCT4 expression by infection with adenovirus carrying the OCT4 gene in BEL-7404 cells up-regulated BIRC5 expression. In exploring the superior-subordinate relationship between BIRC5 and OCT4, we found that the relative activity of the BIRC5 promoter in HCC cells was controlled by OCT4. These results seemingly demonstrated that BIRC5 is a downstream target gene of OCT4.
Functional binding sites for the transcription factors SP1, KLF5, HIF-1α, Rb/E2F, TCF4 and Egr1 have been found in the BIRC5 gene promoter, suggesting that these factors regulate BIRC5 gene expression [
40]. However, a binding site for OCT4 is not found in the BIRC5 promoter region, suggesting that OCT4 may indirectly regulate BIRC5 expression [
13,
40]. In addition to the Rb suppressor and E2F activators (i.e., E2F1, E2F2 and E2F3) that bind directly to the BIRC5 promoter and regulate BIRC5 transcription [
41], the regulatory proteins CDK4, SKP2, Rad51, BRCA2, E2F-DP1, CCND1, Stat3, Rb and p21 can activate the SP1 promoter [
42], which then indirectly leads to an increase in BIRC5 expression. These factors are all involved in cell cycle regulation. In addition, OCT4 modulates the cell cycle by up-regulating CDKN1B, CDKN1C, CDK6 and MAPK4 [
10]. Coincidentally, by screening the binding sites in the promoter regions of these cell cycle regulators, we found an octamer motif for OCT4 at −252 to −245 in the CCND1 proximal promoter. Further studies have confirmed that CCND1 expression and promoter activity is strictly correlated with OCT4 expression levels in OCT4-positive Hep3B cells. When the OCT4 motif in the CCND1 promoter was mutated or modified with a PORE motif that could bind two OCT4 molecules, the promoter activity was suppressed or enhanced, respectively. We also observed high CCND1 promoter activity in OCT4-negative BEL-7404 cells. These results suggested that the OCT4 motif might participate in the regulation of CCND1 promoter activity, and that there are other factors that regulate CCND1 promoter activity in HCC cells.
In the
in vitro experiments, silencing of BIRC5 expression effectively induced apoptosis and cell cycle arrest in HCC cells, thereby inhibiting cancer cell proliferation and decreasing cancer cell viability. Co-suppression of OCT4 and BIRC5 further enhanced the inhibitory effect on cancer cell proliferation. In the
in vivo experiments, BIRC5-shRNA expression inhibited the growth of HCC xenograft tumors by inducing cell apoptosis, although tumor growth was restored in the late stage after the adenovirus injections ceased. The Dual-shRNA that targeted both OCT4 and BIRC5 inhibited tumor growth with great efficiency for a long period of time. These results showed that OCT4 and BIRC5 collusively educe cell proliferation. Clinical follow-up information also demonstrated that the HCC patients who showed co-expression of OCT4 and BIRC5 in cancer tissues had poorer disease-free survival (DFS) and overall survival (OS) than patients who were negative for both OCT4 and BIRC5 (Additional file
4: Figure S2).
Competing interests
The authors declare no conflict of interest.
Authors’ contributions
Conceived and designed the experiments: CS; Performed the experiments: LC, CL, YY, DW, LF, HQ, JW and ZL; Contributed reagents/materials/analysis tools and analyzed the data: LC, CL, XJ, MW, JZ and CS; Wrote the paper: LC, CL, LF, JW and CS. All authors read and approved the final manuscript.