DNA hypermethylation contributes to colorectal cancer metastasis by regulating the binding of CEBPB and TFCP2 to the CPEB1 promoter
Clinical Epigenetics volume 13, Article number: 89 (2021)
Aberrant DNA methylation has been firmly established as a factor contributing to the pathogenesis of colorectal cancer (CRC) via its capacity to silence tumour suppressor genes. However, the methylation status of multiple tumour suppressor genes and their roles in promoting CRC metastasis are not well characterised.
We explored the methylation and expression profiles of CPEB1 (the gene encoding cytoplasmic polyadenylation element-binding protein 1), a candidate CRC tumour suppressor gene, using The Cancer Genome Atlas (TCGA) database and validated these results in both CRC cell lines and cells from Han Chinese CRC patients (n = 104). The functional role of CPEB1 in CRC was examined in experiments performed in vitro and in vivo. A candidate transcription factor capable of regulating CPEB1 expression was predicted in silico and validated by luciferase reporter, DNA pull-down, and electrophoretic mobility shift assays.
Hypermethylation and decreased expression of CPEB1 in CRC tumour tissues were revealed by TCGA database. We also identified a significant inverse correlation (Pearson’s R = − 0.43, P < 0.001) between promoter methylation and CPEB1 expression. We validated these results in CRC samples and two CRC cell lines. We also demonstrated that up-regulation of CPEB1 resulted in significantly decreased tumour growth, migration, invasion, and tumorigenicity and promoted tumour cell apoptosis both in vitro and in vivo. We identified the transcription factors CCAAT enhancer-binding protein beta (CEBPB) and transcription factor CP2 (TFCP2) as critical regulators of CPEB1 expression. Hypermethylation of the CPEB1 promoter resulted in a simultaneous increase in the capacity for TFCP2 binding and a decreased likelihood of CEBPB binding, both of which led to diminished expression of CPEB1.
Our results identified a novel tumour-suppressive role of CPEB1 in CRC and found that hypermethylation of the CPEB1 promoter may lead to diminished expression due to decreased chromatin accessibility and transcription factor binding. Collectively, these results suggest a potential role for CPEB1 in the diagnosis and treatment of CRC.
Colorectal cancer (CRC) is the third most common cancer diagnosed and the second leading cause of cancer-related deaths in the USA [1, 2]. Recent reports document similar epidemiology for CRC in China . Malignant transformation and especially distant metastasis, particularly those found in the lung and liver, are amongst the major causes of CRC-associated mortality. Invasive metastatic disease in the liver or other organs occurs in 30–60% of patients diagnosed with CRC. The five-year survival rate for CRC patients with hepatic metastasis has been estimated at 12–30%, although it can be as low as 5% in patients who do not undergo treatment [4, 5]. There are very few effective therapeutic strategies that can be used to treat CRC patients with distant metastases. This is primarily due to a lack of molecular drug targets and/or any clear understanding of molecular mechanisms underlying the emergence of its invasive potential and the establishment of metastatic disease.
DNA methylation is a form of epigenetic modification that has been identified as critical for numerous physiological processes, including embryonic development [6, 7], histone modification , and gene imprinting . In general, methylation at a gene promoter serves to repress its expression by inhibiting transcription factor (TF) binding. By contrast, methylation within the body of a gene may enhance its expression [8, 10, 11]. Widespread alterations in DNA methylation patterns have been identified and are considered to be amongst the driving forces promoting growth and metastasis of multiple tumour types. Specifically, DNA hypermethylation can silence tumour suppressor genes (TSGs), thereby leading to tumorigenicity, development, and metastasis . Although altered DNA methylation patterns of some CRC-associated TSGs are already well characterised, the status of others remains largely unknown [13,14,15]. An extensive study of these TSGs might provide insight into new drug targets to be used for treatment or early diagnosis of CRC.
Cytoplasmic polyadenylation element-binding protein 1 (CPEB1) encodes a sequence-specific RNA-binding protein that regulates mRNA polyadenylation and translation [16,17,18] and has been linked to cancer progression and metastasis [19, 20]. CPEB1 could target the 3′-UTR of SIRT1 mRNA (encoding Sirtuin 1), thereby controlling the length of the poly(A) tail and suppressing its capacity to mediate cancer cell stemness both in vitro and in vivo [21, 22]. Numerous reports have documented a role for CPEB1 in the development of gastric cancer , breast cancer , glioma [20, 23], and hepatocellular carcinoma . However, the role of CPEB1 and its methylation status in CRC remains unclear. In this study, we screened The Cancer Genome Atlas (TCGA) database to determine the methylation status and expression levels of CPEB1 in CRC. We validated these findings using targeted bisulfite sequencing to evaluate CPEB1 expression in tumour and para-tumour tissues collected from Han Chinese CRC patients. Furthermore, to confirm the role of CPEB1 in the development of CRC, we performed gain- and loss-of-function experiments to determine the effects of CPEB1 expression on cell proliferation, migration, invasion, and apoptosis both in vitro and in vivo. Finally, we verified promoter hypermethylation as the underlying cause of CPEB1 down-regulation using dual-luciferase reporter, chromatin immunoprecipitation (ChIP), DNA pull-down, mass spectrometry, and electrophoretic mobility shift (EMSA) assays.
DNA hypermethylation of the CPEB1 promoter in colorectal cancer
To examine the DNA methylation status of the CPEB1 gene in CRC, we obtained the DNA methylation microarray dataset from TCGA database, including 387 CRC tumours and 45 samples of para-tumour tissue. We found that all nine of the CpG sites located within 1,500 base pairs (bps) of the transcription start site (TSS1500, including the promoter region) were significantly hypermethylated in CRC tumours compared to the para-tumour tissue (Fig. 1a, b). We validated these findings by targeted bisulfite sequencing of 104 paired samples of tumours and para-tumour tissue from Chinese Han CRC patients (Table 1). To confirm the robustness of the targeted bisulfite sequencing method, we examined the methylation status of SEPTIN9, a gene that has been widely reported to be hypermethylated in CRC (Fig. 1f). We found significantly higher levels of methylation of the CPEB1 upstream region (located at chr15:83316804–83316986), which is a region close to the CPEB1 promoter (chr15:83316688–83316747), in CRC tumours than that in the accompanying para-tumour tissues (P < 0.0001, Fig. 1c, d). Collectively, these results suggest that the CPEB1 promoter region is significantly hypermethylated in CRC tissue. We also examined the potential utility of CPEB1 hypermethylation in the diagnosis of CRC. A logistic regression model using the CpG sites of CPEB1 promoter as variables revealed excellent sensitivity (79%) and specificity (96%) with an area under the receiving operator characteristic curve (AUC) of 0.88. Taken together, these results suggest that hypermethylation of CPEB1 might be developed as a potential diagnostic biomarker for CRC (Fig. 1e).
CPEB1 promoter hypermethylation negatively regulates gene expression
Multiple lines of evidence have suggested that promoter hypermethylation regulates gene expression. Findings available in TGCA dataset revealed significant CPEB1 down-regulation in CRC tumours (Fig. 2a). Moreover, the degree of methylation of the CPEB1 promoter inversely correlated with gene transcription (Pearson’s R = − 0.44, P < 0.0001; Fig. 2b). We validated this finding in our sample set consisting of 46 paired CRC tumour and para-tumour tissues which revealed both the down-regulation of CPEB1 expression and the inverse correlation between transcription and CPEB1 promoter methylation (Pearson’s R = − 0.49, P < 0.0001, Fig. 2c, d). We confirmed the inverse correlation between CPEB1 expression and methylation of its promoter in the SW480 and HCT116 CRC cancer cell lines that were treated with the methylation inhibitor, 5-Aza-dC (5-aza-2′-deoxycytidine, DAC). We found significantly decreased CPEB1 promoter methylation in cells treated with DAC (Fig. 2e, f) accompanied by considerable up-regulation of CPEB1 expression (Fig. 2g). Moreover, levels of immunoreactive CPEB1 protein were also significantly higher in both cell lines after DAC treatment (Fig. 2h). Collectively, these results validated the hypermethylated status of the CPEB1 promoter in CRC tissues and confirmed the inverse correlation between gene expression and promoter methylation of CPEB1.
CPEB1 overexpression plays a tumour-suppressing role in CRC
To determine the suppressive effects of CPEB1 in CRC, we examined the results of CPEB1 overexpression in the SW480 and HCT116 cell lines. CPEB1 overexpression in pcDNA3.1-CPEB1-transfected cells (CPEB1 group) resulted in significant increases in both CEPB1 mRNA and CEPB1 protein (P < 0.0001; Fig. 3a, b) compared to cells transfected with the empty pcDNA3.1 vector alone (pcDNA3.1 group) or the untransfected control cells (control group). Cells in the CPEB1 groups grew more slowly than the cells in the pcDNA3.1 and control group after three days in culture (Fig. 3c, d). Likewise, the proliferative capacity of the cells in the CPEB1 group was markedly decreased compared to those in the pcDNA3.1 and control groups (Fig. 3f). The results of in vitro scratch wound healing experiments revealed that CPEB1 overexpression in SW480 and HCT116 cells significantly inhibited cell migration (Fig. 3e and Additional file 1: Figure S4). Moreover, cells in the CPEB1 group exhibited a diminished capacity for invasion (Fig. 3g). We also found that a significantly higher fraction of SW480, DLD-1, and HCT116 cells in the CPEB1 group exhibited early apoptosis (Fig. 3h). Collectively, the results of these in vitro experiments suggest that expression of CPEB1 inhibits CRC tumour cell growth, proliferation, and metastasis and promotes apoptosis.
Increased expression of CPEB1 in the CRC tumour xenograft model significantly inhibited metastasis
To determine whether CPEB1 functions as a CRC TSG in vivo, we established a xenograft mouse model. HCT116 cells transfected with either the plasmid pcDNA3.1-CPEB1 (CPEB1 group) or pcDNA3.1 (pcDNA3.1 group) were inoculated into BALB/c nude mice. The mice were sacrificed 36 days later and tumour volumes and wet weights were recorded. The tumours of the mice in the CPEB1 group were visibly smaller and of significantly smaller volume than those in the pcDNA3.1 and control groups (Fig. 4a, b, c). However, we detected no significant difference in total mouse body weight amongst the three groups at any time point during this study (Fig. 4g). We observed no abnormalities with respect to daily food and water consumption or any other adverse effects, such as altered behaviours and haematuria.
Tumour tissue was subjected to H&E staining. Our results from this analysis suggested that higher expression of CPEB1 resulted in significantly diminished carcinogenesis in this in vivo model of CRC (Fig. 4d). To demonstrate the anti-proliferative and anti-metastatic role of CPEB1 in vivo, we examined the levels of immunoreactive Ki67, matrix metallopeptidase-9 (MMP-9), and E-cadherin in mouse tumour tissues by immunohistochemistry (IHC) and Western blotting (WB). We found that levels of immunoreactive Ki67 and MMP-9 proteins were diminished in tumours isolated from the CPEB1 group. By contrast, the levels of E-cadherin, a signal biomarker for epithelial–mesenchymal transition, were significantly increased in the CPEB1 group (Fig. 4e, f). Taken together, the results from the xenograft model result strongly support a role for CPEB1 as a TSG with the capacity to prevent CRC tumorigenesis and metastasis.
Methylation of the CEBPB-binding site inhibits the transcription of CPEB1
Altered DNA methylation at the CPEB1 promoter region could regulate gene expression by inhibiting TF binding. However, it was not clear which TFs were crucial for the regulation of CPEB1 expression in CRC. We identified three candidate TFs using PROMO and JASPAR software, including CCAAT Enhancer-Binding Protein Beta (CEBPB), GATA-binding factor 2 (GATA2), and tumour protein P53 (TP53) (Additional file 1: Figure S1). We then performed dual-luciferase reporter assays to determine which of these TFs could regulate the expression of CPEB1. As shown in Fig. 5a, the transcriptional activity in cells transfected with both CPEB1-WT and pGL3-basic was significantly increased in response to CEBPB. Although the luciferase activities detected in both of these groups increased in response to CEBPB, the relative activity detected in the CEBPB + CPEB1-WT group was significantly higher than that detected in the CEBPB + pGL3-basic group. These results suggested that CEBPB was capable of significant activation of CPEB1 expression. By contrast, the introduction of mutations into the TF-binding region (CPEB1-Mut) substantially reduced the transcriptional activation of CPEB1 by CEBPB. These results indicated that the identified TF-binding region is essential for CEBPB-mediated activation of CPEB1 expression. In contrast to our findings for CEBPB, neither GATA2 nor TP53 could enhance the transcriptional activity of CPEB1 (Fig. 5b, c). Collectively, our results suggested that CEBPB might be critical TF regulating CPEB1 expression in CRC cells.
We then performed dual-luciferase reporter assays to determine whether hypermethylation at the CEBPB-binding site would have an impact on the expression of CPEB1. For these experiments, we generated luciferase reporter constructs that included either non-methylated or hypermethylated fragments of the upstream region of CPEB1 that included the CEBPB-binding site. We found that hypermethylation of the CEBPB-binding site (i.e. CPEB1 WT + CEBPB + M.SssI) resulted in a significant decrease in luciferase activity. These results suggest that hypermethylation of the CEBPB-binding site could inhibit CPEB1 expression via interference with the recruitment of CEBPB to the CPEB1 promoter region (Fig. 5d). To evaluate this possibility, we performed a ChIP assay that demonstrated direct CEBPB binding to its cognate site in the upstream fragment of CPEB1 using the cultured cells transfected with pcDNA3.1-CEBPB vector (Fig. 5e). Specifically, we found that the -993 to -779 upstream region of the CPEB1 promoter region bound CEBPB more effectively than the IgG control. These results suggest that CEBPB was capable of direct binding to the -993 to -779 region of the CPEB1 promoter and thus could regulate CPEB1 transcription (Fig. 5f).
Hypermethylation of the CPEB1 promoter results in increased binding of TFCP2 instead of CEBPB
We then performed DNA pull-down/mass spectrometry (MS) experiments to explore TF binding to the hypermethylated CPEB1 promoter. Consistent with our previous results, we confirmed that hypermethylation of the TF-binding site resulted in markedly reduced binding of CEBPB (Fig. 6a, b). We also found that the notable supershifted band on electromobility shift assay (EMSA) representing CEBPB interactions with the wild-type CPEB1 upstream region probe could not be detected in experiments performed with either mutant or hypermethylated probes (Fig. 6c). These results add strength to our conclusion that CEBPB can bind to the CPEB1 upstream region directly and that hypermethylation of the CEBPB-binding region results in reduced binding and diminished expression of CPEB1.
Finally, we analysed the proteins in the complexes that were pulled down by the methylated CPEB1 probe using MS. We identified 236 candidate binding proteins in this experiment (Additional file 1: Table S6). We then selected genes that were also included in the list of candidate binding proteins and the predicted TFs capable of binding to the CPEB1 upstream region from the ChIP-seq database. As a result of this analysis, we identified six TFs that might be capable of binding to the hypermethylated CPEB1 promoter. Amongst them, TFCP2 was the only TF that interacts with a cognate site found in the CPEB1 TF-binding region; these results suggested that TFCP2 might be capable of substituting for CEBPB and binding to the methylated upstream region of CPEB1 (Fig. 6d). We performed an EMSA assay to determine whether TFCP2 was capable of binding to the methylated form of the CPEB1 upstream region. The EMSA assay revealed that hypermethylation of the CPEB1 TF-binding region enhanced the binding interaction between TFCP2 and CPEB1 (Fig. 6e). Collectively, our findings suggest that TFCP2 may bind to the methylated TF-binding region of CPEB1 and thereby repress CPEB1 transcription.
Materials and methods
Human tissues, cell lines, transfection, and drug treatment
We collected the pairs of tumour tissue specimens and their corresponding para-tumour tissues (the corresponding adjacent normal tissue resected at least 5 cm distant from the tumour tissue) from Han Chinese patients diagnosed with CRC (n = 104). The samples were obtained from the Department of General Surgery in Affiliated Hospital of Nantong University. Tumours were classified according to Tumour Node Metastasis (TNM)/Union for International Cancer Control (UICC) criteria following histopathological examination. The SW480, DLD1, and HCT116 cell lines were purchased from the Cell Bank of the Type Culture Collection of Chinese Academy of Sciences (Shanghai, China). The SW480 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with high glucose (Hyclone, USA). The HCT116 cells were grown in McCoy’s 5A medium (Thermo Fisher Scientific, USA), supplemented with 10% fetal bovine serum (FBS; Gibco, USA) at 37 °C in 5% CO2. The full-length cDNA of CPEB1 (NM_030594.4) was cloned in the pcDNA3.1-EGFP (enhanced green fluorescent protein) vector to create the pcDNA3.1-CPEB1 overexpression vector. This vector was transfected into cells using Lipofectamine 2000 (Invitrogen, USA). 5-Aza-2′-deoxycytidine (DAC; Sigma-Aldrich, USA) was used for DNA demethylation as previously described .
DNA methylation status of CPEB1 in TCGA database and our validation cohort
Publicly available high-throughput DNA methylation microarray data (Illumina Human Methylation 450 K) of 387 CRC tumours and 45 samples of para-tumour tissue were obtained from the TCGA database (level-3). We identified nine candidate CpG sites within the CPEB1 promoter region. The extent of methylation (β value) at each CpG site was calculated based on the hypermethylated (M) and non-methylated (U) intensities as shown:
The β values calculated for CRC tumours and para-tumours were compared at each candidate CpG site. The mean methylation level of the CpG sites for each tissue sample was calculated as overall CPEB1 methylation. We then compared mean methylation levels of individual CRC tumours and para-tumours.
To validate these findings, we extracted DNA samples from both tumour and paired para-tumour tissue which was subjected to bisulfite conversion. After careful evaluation of the CG percentage, poly(T), and single-nucleotide polymorphisms (SNPs) in the potentially differentially methylated regions (DMRs), we designed primers to detect the methylation of CPEB1 upstream region using next-generation sequencing (NGS). The amplified DNA fragments were sequenced by Illumina Hiseq 2000 (Illumina, San Diego, CA, USA). To verify the robustness of our methodology, we also examined the methylation status of the SEPTIN9, a gene known to be hypermethylated in CRC tissues (Additional file 1: Table S1). The detailed protocol used for targeted bisulfite sequencing was described previously . We applied BSseeker2 to map reads and to identify methylated sites. Samples with high missing rates (> 30%) and CpG sites with high missing rates (> 20%) were removed from the dataset. The methylation rate (β value) for each CpG site was calculated according to the number of reads (C or T) detected at each CpG site.
Similarly, the mean methylation rate at each of these CpG sites in each sample was considered as the overall methylation of CPEB1. Overall methylation of CPEB1 was compared in both CRC tumours and their associated para-tumour tissue. The methylation rate (β value) for each CpG site and for each sample determined in the validation stage is shown in Additional file 2: Table S7.
RNA extraction and real-time quantitative PCR
Total RNA from CRC tissues and cell lines was isolated with TRIzol (Thermo Fisher Scientific, USA) following the manufacturer’s instructions. First-strand complementary DNA (cDNA) was synthesised from 1 μg total RNA using an M-MLV Reverse Transcriptase Kit (ELK Biotech., China). Real-time quantitative PCR was performed on an Applied Biosystems 7500 real-time PCR system (ABI, USA) with SYBR green I dye (ELK Biotech., China). The expression levels of ACTB (encoding β-actin) and GAPDH served as internal references.
Cell proliferation assay
SW480 and HCT116 cells transfected with pcDNA3.1-CPEB1 and pcDNA3.1-EGFP vector-only control were incubated for 0, 24, 48, and 72 h in 96-well plates at 4,000 cells/well. Cell proliferation was detected by adding 10 µL CCK-8 solution (Biolite Biotech Co., China) to each well followed by the optical density (OD 450 nm) measurements 3 h later using a microplate reader (Thermo Fisher Scientific).
Cell clone formation assay for cell proliferation and growth
Cells were seeded into 24-well ultra-low attachment round-bottom plates (Excel Scientific, USA) at a density of 100 cells per well in cell culture medium together with pcDNA3.1-CPEB1 and pcDNA3.1-EGFP vectors, respectively, and cultured for 14 days. Wells were then washed twice with phosphate-buffered saline (PBS), fixed for 15 min with 500 μL 4% paraformaldehyde, and stained with Giemsa dye for 10 min. The number of violet-stained cell spheres (nb: only cell spheres with > 10 cells were included) was determined in each well. The efficiency of cell colony formation was calculated as:
Transwell assays to examine cell metastasis and invasion
A cell invasion assay was performed using an invasion chamber (Corning, USA) in a 24-well tissue culture plate with 12-cell culture inserts. Suspensions of cells (105 cells/mL) that were transfected with pcDNA3.1-CPEB1 or pcDNA3.1-EGFP or mock-transfected with medium alone were added to the interior of the inserts in 200 μL serum-free medium. Five hundred μL of medium containing 20% FBS was added to the lower chamber followed by incubation at 37 °C in 5% CO2 for 24 h. Invasive cells detected on the lower surface of the membrane by staining with 0.1% crystal violet for 10 min were photographed. The number of invasive cells was calculated from images photographed at high magnification (400-fold). Relative invasion was defined as follows:
Wound healing assay to evaluate cell migration
Cells were seeded into a six-well tissue culture plate at a density that will permit them to reach 70%–80% confluence after 24 h of growth. The monolayer was scratched gently and slowly across the centre of the well with a 200-µL pipette tip. After scratching, the well was gently washed twice with tissue culture medium to remove the detached cells, and the distance between the two edges of the gap was measured (t = 0 h). Cells were incubated for an additional 24 and 48 h; the distances between the edges of the gap distance were assessed at these times to determine the rate of wound healing. Migration distance = gap distance at T0—gap distance at Tx (Tx = 24 h or 48 h). The gap distance was measured using ImageJ software (v1.53c). The wound width (%) was calculated as follows:
Assessment of apoptosis by flow cytometry (FCM)
Cells were removed from tissue culture plates with trypsin and washed twice in ice-cold PBS. A total of 5 × 105 cells were resuspended in 500 µL of binding buffer (KeyGen, China). This was followed by the addition of 5 µL of Annexin V-APC (KeyGen) and 5 µL of the counterstain, 7-amino actinomycin D (7-AAD). The mixture was incubated at room temperature for 10 min in the dark, followed by quantitative analysis of apoptosis using a FlowSight flow cytometer (Merck, Germany).
Dual-luciferase reporter gene assay
A dual-luciferase reporter assay system was purchased from Promega, USA. Three TFs (CEBPB, GATA, and TP53) were identified as candidate transcriptional regulators of CPEB1. The TF-binding regions were predicted for each factor within the upstream region of CPEB1. We then constructed wild-type and the mutant fragments of these regions for use in the dual-luciferase assay (Additional file 1: Table S4). The wild-type or mutant fragments of the upstream region of CPEB1 were cloned into the pGL3-basic vector (Promega, USA) to generate CPEB1-WT and CPEB1-Mut, respectively. The sequences of all construct inserts were verified by Sanger sequencing. The full-length cDNAs of these three TFs were cloned individually into the pcDNA3.1 expression vector to generate pcDNA3.1-CEBPB, pcDNA3.1-GATA, and pcDNA3.1-TP53, respectively. HCT116 cell cultures plated on 24-well plates were transfected individually with each of the TF expression vectors (0.81–1.07 μg/μL) together with wild-type pGL3-CPEB1 (1.64–2.01 μg/μL) or mutant pGL3-CPEB1 (1.30–1.71 μg/μL) using Lipofectamine 2000. In each transfection, 0.2 μg/μL of a pRL-TK vector (Renilla luciferase vector) was used to normalise the transfection efficiency. The pGL3-basic + pcDNA3.1 group served as a negative control.
A dual-luciferase reporter assay was also performed to determine whether hypermethylation at the predicted binding site for CEBPB had an impact on the binding interaction. HCT116 cells plated on 24-well plates were transfected with 160 ng of the pcDNA3.1-CEBPB expression vector and 200 ng wild-type pCpGL-CPEB1 (i.e. the wild-type fragment of the CPEB1 promoter region cloned into the pCpGL-basic vector), 200 ng of mutant pCpGL-CPEB1, or 200 ng of hypermethylated wild-type and mutated pCpGL-CPEB1 using Lipofectamine 2000. The pCpGL-basic vector engineered as devoid of CpG sites was a gift from Professor M. Wang, Soochow University, China. The pCpGL-CPEB1 reporter vector was hypermethylated with M.SssI (CpG) methyltransferase as per the manufacturer’s instructions (Thermo Fisher Scientific). In each transfection, 40 ng of a pRL-TK vector was used to normalise the transfection efficiency. Transfection with each of the pCpGL-basic reporter vectors together with each of the pcDNA3.1 expression vectors served as negative controls. The relative luciferase activity in each group was calculated as follows:
A chromatin immunoprecipitation (ChIP) assay was performed as previously described . Briefly, HCT116 cells were transfected with pcDNA3.1-CEBPB or the control pcDNA3.1 vector. After transfection for 48 h, cells were crosslinked with 1% formaldehyde for 10 min at room temperature. Chromatin was fragmented using a sonication device (Scientz Biotechnology Company, China) to obtain DNA fragments of approximately 100–500 bp in length. Anti-CEBPB antibody (Abcam, USA) was added to form an anti-CEBPB–CEBPB protein–DNA chromatin complex. Protein A beads (Beaver Biosciences Inc., China) were used to pull down the chromatin complex. Crosslinks between DNA and proteins were destroyed by overnight incubation at 65 °C with proteinase K (Sigma-Aldrich, USA). The resulting DNA fragments were purified using the MinElute PCR Purification Kit (Aidlab Bio., China) to serve as PCR templates. We used Emboss (The European Molecular Biology Open Software Suite) software to identify the core TF-binding region of CPEB1. This program predicted that the region between nucleotide -993 and -779 of the CPEB1 promoter as containing the core TF-binding region. To validate and extend this prediction, we divided the promoter region of CPEB1 into three different parts: PR1, PR2 (containing the core TF-binding region), and PR3. Moreover, to identify the core TF-binding region by real-time quantitative PCR, we divided the core TF-binding region (PR2) into three smaller parts: PCR1, PCR2, and PCR3 (Fig. 5f). DNA purified from the ChIP experiment was used as a template to amplify the core TF-binding region of CPEB1. The primers used in the ChIP-PCR assay are shown in Additional file 1: Table S2. Input sample represented the fragmented, diluted, and pre-cleared chromatin DNA that was introduced into the immunoprecipitation experiment. The responses to 1% input served as the internal reference for data normalisation, and IgG was included as a negative control. Relative input (%) represented the level of CPEB1 expression from DNA in the ChIP experiment that was precipitated with anti-CEBPB or anti-IgG.
DNA pull-down and mass spectrometry (MS)
A DNA pull-down assay was performed as reported previously . The CPEB1 promoter region was amplified by PCR using a 5′-biotin-labelled primer (Additional file 1: Table S3); the non-methylated biotinylated double-stranded DNA was methylated with M.SssI (CpG) methyltransferase. We prepared cell lysates from HCT116 cells that were transiently transfected with the pcDNA3.1-CEBPB vector. Cell lysates (100 μg) and hypermethylated or non-methylated biotinylated probes (0.6 pmol) were incubated for 2 h at room temperature in the presence of streptavidin–agarose beads (Beaver Bio., China) to form DNA–protein–beads complexes; these were placed on a magnetic rack to promote the separation of the DNA–protein–beads complexes. The pulled-down complex was washed six times with 1 mL ice-cold tris-buffered saline and eluted in 80 μL protein elution buffer for 5 min at 95 °C. Proteins were separated by gel electrophoresis and identified by WB and mass spectrometry (MS) using LC–MS/MS (Eksigent ekspert nanoLC; AB Sciex TripleTOF 5600-plus) as described previously . The MS data were analysed with Mascot software and mapped in the NCBInr database by Proteinpilot software.
Electrophoretic mobility shift assay (EMSA)
Nucleoproteins from HCT116 cells were obtained using the nucleus protein extraction kit supplemented with the protease inhibitor phenylmethanesulfonyl fluoride (PMSF) at a final concentration of 1% (Beyotime, China). Protein quantification was determined using the bicinchoninic acid (BCA) method (Beyotime). A synthetic 5′-end biotin-labelled probe that was consistent with the TF-binding sequence was prepared. The 5′-end biotin-labelled probe, an unlabelled specific competitive probe, and the non-specific competitive mutation probe were generated by GeneCreate Biological Engineering Co (Wuhan, China) as previously described . The sequences of all probes are shown in Additional file 1: Table S5. Briefly, in each 20 μL reaction, 20 fmol labelled probes were incubated with 2 μg nucleoprotein, together with 4 pmol specific competitive probe or non-specific mutation probe in an ionising environment, and 0.4–0.6 μg anti-CEBPB (Abcam, USA) and anti-TFCP2 (Proteintech, USA) antibodies. Migrating bands representing protein–DNA complexes were developed with a mixture of stable peroxide and luminol/enhancer solutions (LightShift Chemiluminescent EMSA kit, Thermo Fisher). An unlabelled probe at 50 × excess concentration was used as a specific competitor.
Western blotting (WB) and Immunohistochemistry (IHC)
Total protein was extracted from cells and tissues using radioimmunoprecipitation (RIPA) lysis buffer (Beyotime, China) and quantitated using a BCA protein assay (Beyotime, China) as previously described . Protein samples were denatured by boiling in a sodium dodecyl sulfate (SDS) sample buffer, separated by 12% SDS–polyacrylamide gel electrophoresis (PAGE), transferred to PVDF (polyvinylidene fluoride) membrane (Millipore, USA), and incubated with antibodies including anti-CPEB1 (Abcam, USA), anti-E-cadherin (Cell Signaling Technology (CST), USA), and anti-MMP-9 (CST, USA). GAPDH was used as the loading control. An IHC assay to detect Ki67 protein (Aoquan Med., China) was performed to evaluate cell proliferation in situ as previously described . IHC staining was evaluated by two independent gastrointestinal pathologists. The staining intensity of each sample was scored as follows: 0, no staining; 1, weak staining; 2, moderate staining; and 3, strong staining. The percentage of positively stained tumour cells was also scored as follows: 1, < 10%; 2, 10% – 50%; 3, 51 – 75%; 4, > 75%. The IHC staining score was calculated as follows:
IHC staining score = IHC staining intensity score × score based on the number of positively stained tumour cells.
Xenograft tumour mouse model
All animal experiments were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Affiliated Hospital of Nantong University. Four male nude BALB/c mice at 5 weeks of age (n = 4 per group; Nantong University Laboratory Animal Center, China) were anaesthetised via inhalation of a mixture of isoflurane/propylene glycol. HCT116 cells transfected with recombinant plasmid constructs containing pcDNA3.1-CPEB1 or pcDNA3.1 were injected subcutaneously into the back of each mouse (2.0 × 106 cells in 200 μL PBS). The tumour sizes were assessed every 3 days by measuring in two dimensions; tumour volumes were estimated as the volume = (tumour length) × (tumour width)2/2. Mice were killed 36 days after implantation, and the tumours were collected and weighed. Slides were prepared and stained with hematoxylin and eosin (H&E). Levels of immunoreactive CPEB1, Ki67, MMP-9, and E-cadherin were determined by WB and IHC assays.
Statistical assessments of differential DNA methylation and gene expression in CRC tumours and para-tumour tissue were performed using R (v3.6.3). All data from the functional experiments are shown as the mean ± standard deviation (SD). A value of P < 0.05 was considered statistically significant in two-tailed t-tests performed using SPSS 20.0. The false discovery rate (FDR) correction was used for multiple test corrections, as applicable. The bottoms, middles, and tops of the boxes in box-and-whisker plots represent the 25th (Q1), median (Q2), and 75th percentile (Q3), respectively.
The upper and the lower whiskers shown in the boxplots were calculated as follows:
where the interquartile range (IQR) = Q3–Q1.
Genome-wide alterations in DNA methylation have been widely reported in CRC in association with tumorigenesis and metastasis. However, the DNA methylation status of numerous TSGs and the contributions of hypermethylation to the pathogenesis of CRC remain largely unknown. Our study examined the hypermethylation status and expression of CPEB1, a putative TSG in CRC. The functional characterisation presented in this study confirmed the CRC tumour suppressor function of the CPEB1 gene. We found that overexpression of CPEB1 in CRC cell lines suppressed viability, colony formation, cellular invasion, and migration. CPEB1 overexpression also resulted in the increased capacity for cell transformation, larger tumour volumes, and increased rates of tumour cell apoptosis in experiments performed both in vitro and in vivo. Furthermore, we found the TF, CEBPB could bind to the core promoter region of CPEB1 and thereby regulate gene transcription. The hypermethylated form of the CPEB1 promoter found in CRC cells did not support CEBPB binding, although this epigenetic modification facilitated interactions with TFCP2. These interactions resulted in diminished expression of CPEB1 and enhanced CRC tumorigenesis and metastasis (Fig. 7).
Results from previous studies revealed that CPEB1 played a major role in tumour suppression via its capacity to maintain p53 mRNA polyadenylation and translation  and thus the functionality of the p53 gene . CPEB1 has also been identified as a participant in the p53-mediated senescence of glioma cells . Of note, CPEB1 may also inhibit glioblastoma growth via its capacity to promote poly-A tail elongation and translation of the cell cycle inhibitor, p27 . CPEB1 might also target the 3′-UTR of SIRT1 to suppress the growth of hepatic carcinoma . However, the functional significance of CPEB1 in CRC has not been well characterised. Moreover, although previous studies have identified components of the CPEB1 downstream signalling pathways involved in tumour suppression, the corresponding upstream regulatory events remain undefined. In this study, we used a multifaceted experimental approach to explore the molecular mechanisms employed by CPEB1 as a TSG in CRC. Of note, we have demonstrated that CPEB1 suppression was associated with CRC malignant transformation and that this role is dependent on hypermethylation of the TF-binding site in its promoter region. Collectively, our results suggest a role for demethylation drugs as a future therapeutic approach to CRC.
The TF, TFCP2, plays a crucial role in cancer incidence and development . The TFCP2/TFCP2L1/UBP1 TF subfamily is involved in various aspects of cancer development , including roles as pro-oncogenic factors in hepatocellular carcinoma as well as pancreatic and breast cancer. The actions of TFCP2 may also promote cervical carcinogenesis and CRC . TFCP2 was identified as involved in the epithelial-mesenchymal transition and might also enhance angiogenesis . Nonetheless, TFCP2 can also act as a TSG, as exemplified by its capacity to inhibit melanoma growth . These contrasting observations suggest that further research might be needed to further clarify the role of this TF subfamily in oncogenesis. In this study, our findings revealed that TFCP2 recognises and interacts with the hypermethylated CPEB1 promoter and can suppress CPEB1 transcription. Interestingly, TFCP2 binds to the methylated upstream region of CPEB1; this suggests that TFCP2 might be a novel DNA methylation reader. However, the conclusion was based on only the in vitro studies only. Future in vivo experiments focused on the role of TFCP2 in CPEB1 regulation and CRC metastasis might provide important verification of this finding.
TFs could modulate the expression of target genes and play a key role in tumour development, differentiation, and metastasis [33, 34]. Methods that permit accurate detection of TF binding to specific target genes may have substantial potential for improved diagnosis and treatment of neoplastic disease [35, 36]. Our study predicted and confirmed that CEBPB binding to the -993 to -779 region of the CPEB1 upstream region activated CPEB1 transcription. Moreover, we demonstrated that hypermethylation of the CPEB1 promoter region resulted in significantly diminished CEBPB binding of CEBPB to its DNA recognition elements, thereby leading to diminished expression of CPEB1. These results suggest that diminished capacity for CEBPB bindings may be an important mechanism that contributes to decreased CPEB1 expression in CRC.
In summary, our findings revealed hypermethylation within the upstream regions of CPEB1 that was associated with its diminished expression in CRC tumours compared to para-tumour tissues. We identified CEBPB and TFCP2 as two critical TFs contributing to CPEB1 regulation. The epigenetic modifications of CPEB1 and subsequent transcriptional changes contribute significantly to CRC tumorigenesis and metastasis. Collectively, these results suggest that the development of CPEB1 as a candidate biomarker and molecular target for CRC treatment is a subject worthy of further consideration.
Availability of data and materials
The data of the current study are available from the corresponding author upon reasonable request.
Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer J Clin. 2019;69(1):7–34.
Siegel RL, Miller KD, Goding Sauer A, Fedewa SA, Butterly LF, Anderson JC, Cercek A, Smith RA, Jemal A. Colorectal cancer statistics, 2020. CA Cancer J Clin. 2020;70(3):145–64.
Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F, Jemal A, Yu XQ, He J. Cancer statistics in China, 2015. CA Cancer J Clin. 2016;66(2):115–32.
Dekker E, Tanis PJ, Vleugels JLA, Kasi PM, Wallace MB. Colorectal cancer. Lancet. 2019;394(10207):1467–80.
Zarour LR, Anand S, Billingsley KG, Bisson WH, Cercek A, Clarke MF, Coussens LM, Gast CE, Geltzeiler CB, Hansen L, et al. Colorectal cancer liver metastasis: evolving paradigms and future directions. Cell Mol Gastroenterol Hepatol. 2017;3(2):163–73.
Sellars M, Huh JR, Day K, Issuree PD, Galan C, Gobeil S, Absher D, Green MR, Littman DR. Regulation of DNA methylation dictates Cd4 expression during the development of helper and cytotoxic T cell lineages. Nat Immunol. 2015;16(7):746–54.
Feng Y, Rudensky AY. DNA methylation secures CD4(+) and CD8(+) T cell lineage borders. Nat Immunol. 2015;16(7):681–3.
Zhao Q, Zhang J, Chen R, Wang L, Li B, Cheng H, Duan X, Zhu H, Wei W, Li J, et al. Dissecting the precise role of H3K9 methylation in crosstalk with DNA maintenance methylation in mammals. Nat Commun. 2016;7:12464.
Bell AC, Felsenfeld G. Methylation of a CTCF-dependent boundary controls imprinted expression of the Igf2 gene. Nature. 2000;405(6785):482–5.
Peng S, Yang S, Gao M, Liao X, Liu J, Yang C, Wu C, Yu W. P-Hint-Hunt: a deep parallelized whole genome DNA methylation detection tool. BMC Genomics. 2017;18(Suppl 2):134.
Harris CJ, Scheibe M, Wongpalee SP, Liu W, Cornett EM, Vaughan RM, Li X, Chen W, Xue Y, Zhong Z, et al. A DNA methylation reader complex that enhances gene transcription. Science. 2018;362(6419):1182–6.
Dawson MA, Kouzarides T. Cancer epigenetics: from mechanism to therapy. Cell. 2012;150(1):12–27.
Ding W, Pu W, Wang L, Jiang S, Zhou X, Tu W, Yu L, Zhang J, Guo S, Liu Q, et al. Genome-Wide DNA Methylation Analysis in Systemic Sclerosis Reveals Hypomethylation of IFN-Associated Genes in CD4(+) and CD8(+) T Cells. J Invest Dermatol. 2018;138(5):1069–77.
Xu RH, Wei W, Krawczyk M, Wang W, Luo H, Flagg K, Yi S, Shi W, Quan Q, Li K, et al. Circulating tumour DNA methylation markers for diagnosis and prognosis of hepatocellular carcinoma. Nat Mater. 2017;16(11):1155–61.
Hao X, Luo H, Krawczyk M, Wei W, Wang W, Wang J, Flagg K, Hou J, Zhang H, Yi S, et al. DNA methylation markers for diagnosis and prognosis of common cancers. Proc Natl Acad Sci USA. 2017;114(28):7414–9.
Bava FA, Eliscovich C, Ferreira PG, Minana B, Ben-Dov C, Guigo R, Valcarcel J, Mendez R. CPEB1 coordinates alternative 3’-UTR formation with translational regulation. Nature. 2013;495(7439):121–5.
Hake LE, Richter JD. CPEB is a specificity factor that mediates cytoplasmic polyadenylation during Xenopus oocyte maturation. Cell. 1994;79(4):617–27.
Caldeira J, Simoes-Correia J, Paredes J, Pinto MT, Sousa S, Corso G, Marrelli D, Roviello F, Pereira PS, Weil D, et al. CPEB1, a novel gene silenced in gastric cancer: a Drosophila approach. Gut. 2012;61(8):1115–23.
Nagaoka K, Fujii K, Zhang H, Usuda K, Watanabe G, Ivshina M, Richter JD. CPEB1 mediates epithelial-to-mesenchyme transition and breast cancer metastasis. Oncogene. 2016;35(22):2893–901.
Xiaoping L, Zhibin Y, Wenjuan L, Zeyou W, Gang X, Zhaohui L, Ying Z, Minghua W, Guiyuan L. CPEB1, a histone-modified hypomethylated gene, is regulated by miR-101 and involved in cell senescence in glioma. Cell Death Dis. 2013;4:e675.
Novoa I, Gallego J, Ferreira PG, Mendez R. Mitotic cell-cycle progression is regulated by CPEB1 and CPEB4-dependent translational control. Nat Cell Biol. 2010;12(5):447–56.
Xu M, Fang S, Song J, Chen M, Zhang Q, Weng Q, Fan X, Chen W, Wu X, Wu F, et al. CPEB1 mediates hepatocellular carcinoma cancer stemness and chemoresistance. Cell Death Dis. 2018;9(10):957.
Galardi S, Petretich M, Pinna G, D’Amico S, Loreni F, Michienzi A, Groisman I, Ciafre SA. CPEB1 restrains proliferation of Glioblastoma cells through the regulation of p27(Kip1) mRNA translation. Sci Rep. 2016;6:25219.
Jiang D, He Z, Wang C, Zhou Y, Li F, Pu W, Zhang X, Feng X, Zhang M, Yecheng X, et al. Epigenetic silencing of ZNF132 mediated by methylation-sensitive Sp1 binding promotes cancer progression in esophageal squamous cell carcinoma. Cell Death Dis. 2018;10(1):1.
Zhang J, Ding W, Kuai X, Ji Y, Zhu Z, Mao Z, Wang Z. Dermcidin as a novel binding protein of lncRNA STCAT3 and its effect on prognosis in gastric cancer. Oncol Rep. 2018;40(5):2854–63.
Ding W, Pu W, Jiang S, Ma Y, Liu Q, Wu W, Chu H, Zou H, Jin L, Wang J, et al. Evaluation of the antifibrotic potency by knocking down SPARC, CCR2 and SMAD3. EBioMedicine. 2018;38:238–47.
Ding W, Wang G, Shao K, Wang F, Huang H, Ju S, Cong H, Wang H. Amelioration of colorectal cancer using negative lipidoid nanoparticles to encapsulate siRNA against APRIL by enema delivery mode. Pathol Oncol Res POR. 2014;20(4):953–64.
Burns DM, D’Ambrogio A, Nottrott S, Richter JD. CPEB and two poly(A) polymerases control miR-122 stability and p53 mRNA translation. Nature. 2011;473(7345):105–8.
Liu Z, Gao F, Shao Z, Xie H, Zhou L, Xu X, Zheng S. TFCP2 Genetic polymorphism is associated with predisposition to and transplant prognosis of hepatocellular carcinoma. Gastroenterol Res Pract. 2017;2017:6353248.
Kotarba G, Krzywinska E, Grabowska AI, Taracha A, Wilanowski T. TFCP2/TFCP2L1/UBP1 transcription factors in cancer. Cancer Lett. 2018;420:72–9.
Zhang X, Sun F, Qiao Y, Zheng W, Liu Y, Chen Y, Wu Q, Liu X, Zhu G, Chen Y, et al. TFCP2 is required for YAP-dependent transcription to stimulate liver malignancy. Cell Rep. 2017;21(5):1227–39.
Liu Y, Zhang X, Lin J, Chen Y, Qiao Y, Guo S, Yang Y, Zhu G, Pan Q, Wang J, et al. CCT3 acts upstream of YAP and TFCP2 as a potential target and tumour biomarker in liver cancer. Cell Death Dis. 2019;10(9):644.
Lambert SA, Jolma A, Campitelli LF, Das PK, Yin Y, Albu M, Chen X, Taipale J, Hughes TR, Weirauch MT. The human transcription factors. Cell. 2018;172(4):650–65.
Ulz P, Perakis S, Zhou Q, Moser T, Belic J, Lazzeri I, Wolfler A, Zebisch A, Gerger A, Pristauz G, et al. Inference of transcription factor binding from cell-free DNA enables tumor subtype prediction and early detection. Nat Commun. 2019;10(1):4666.
Lai B, Gao W, Cui K, Xie W, Tang Q, Jin W, Hu G, Ni B, Zhao K. Principles of nucleosome organization revealed by single-cell micrococcal nuclease sequencing. Nature. 2018;562(7726):281–5.
Snyder MW, Kircher M, Hill AJ, Daza RM, Shendure J. Cell-free DNA comprises an in vivo nucleosome footprint that informs its tissues-of-origin. Cell. 2016;164(1–2):57–68.
The authors gratefully acknowledge Professor M. Wang, Soochow University, China, for his gift of the pCpGL-Basic vector.
The work was supported by the National Natural Science Foundation of China (Grant Nos. 81974313, 82003360), Jiangsu Province's Key Young Medicine Talents Program (Grant Nos. QNRC2016688, QNRC2016469), Postdoctoral Science Foundation of China (Grant No. 2019M651930), the Natural Science Foundation of Jiangsu Commission of Health (M2020065), and the Natural Science Foundation of Higher Education of Jiangsu Province (20KJB320001).
Ethics approval and consent to participate
The study was approved by the institutional review boards of Affiliated Hospital of Nantong University (No. 2018-K008). Written informed consent was obtained from each participant before sample collection.
Consent for publication
The manuscript has not been submitted elsewhere, and all authors declare their consent for publication in Clinical Epigenetics.
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Shao, K., Pu, W., Zhang, J. et al. DNA hypermethylation contributes to colorectal cancer metastasis by regulating the binding of CEBPB and TFCP2 to the CPEB1 promoter. Clin Epigenet 13, 89 (2021). https://doi.org/10.1186/s13148-021-01071-z