Targeting of non-coding RNAs encoded by novel MYC enhancers inhibits the proliferation of human hepatic carcinoma cells in vitro

The proto-oncogene MYC is important for development and cell growth, however, its abnormal regulation causes cancer. Recent studies identified distinct enhancers of MYC in various cancers, but any MYC enhancer(s) in hepatocellular carcinoma (HCC) remain(s) elusive. By analyzing H3K27ac enrichment and enhancer RNA (eRNA) expression in cultured HCC cells, we identified six putative MYC enhancer regions. Amongst these, two highly active enhancers, located ~ 800 kb downstream of the MYC gene, were identified by qRT-PCR and reporter assays. We functionally confirmed these enhancers by demonstrating a significantly reduced MYC expression and cell proliferation upon CRISPR/Cas9-based deletion and/or antisense oligonucleotide (ASO)-mediated inhibition. In conclusion, we identified potential MYC enhancers of HCC and propose that the associated eRNAs may be suitable targets for HCC treatment.

To seek direct functional evidence for the suspected enhancer activities in HCC cell lines, we examined the activity of a luciferase reporter in transiently transfected Huh7 cells. Each enhancer region was cloned into the minimal promoter vector pGL4.26 immediately upstream of the luciferase gene. For comparison, we included lung adenocarcinoma (LUAD)-R3 and LUAD-R4, two SE of MYC known to exhibit high and low activity, respectively, in lung adenocarcinoma cells 29 . As expected, transfection of R2 and R3 increased luciferase activity by approximately tenfold. In contrast, LUAD-R3 and LUAD-R4 did not show an enhanced activity (Fig. 2B). Next, we analyzed 500 bp fragments from within the R2 and R3 regions. Of the R2 fragments, the R2-3-containing plasmid showed the highest enhancer activity (Fig. 2C), while of the R3 fragments, R3-2-and R3-3-containing plasmid were most active (Fig. 2D). These results suggest R2 (R2-3) and R3 (R3-2 and R3-3) as candidate regulators of the transcriptional activation of MYC in HCC cells. eRNA of putative MYC enhancers in HCC cells. Next, we analyzed the R2 and R3 regions of the HepG2 cells for eRNA expression using qRT-PCR (Supplementary Table S1). Based on the GRO-seq data, six different sets of primers were designed (Fig. 3A). We found that the RNAPII transcription elongation inhibitor DRB caused a significant reduction in sense eRNA expression in regions R2, R3, R4, and R6 but not R1 and R5 (Fig. 3B). R2 and R3 were, therefore, further studied for expression changes through treatment with BET inhibitors. As expected, eRNA expression in regions R2 and R3 was significantly decreased (Fig. 3C). Additionally, R1, R4, R5, and R6 eRNAs were decreased in BET inhibitor-treated HCC cells (Fig. S1). Together, these results indicate that there is a correlation between the activity of enhancers and eRNA transcription. From this, we hypothesized that the BET inhibitors suppressed MYC expression through the regulation of eRNA expression.

Disruption of MYC enhancers affects MYC-related gene expression and cell growth in HCC cells.
Since our eRNA expression experiments showed enhancer activity of R3, we tested whether its deletion affects MYC gene expression. We performed the deletion in the Huh7 cell line using the CRISPR/Cas9 system. The targeted sequences are located on chromosome 8 (Chr 8: 128,556,059-128,557,653) downstream of the MYC gene (Supplementary Table S3). After plating the transfected suspension at limiting dilution density, genomic PCR revealed a deletion of the R3 region in one of the wells showing growth. DNA sequencing revealed a 357 bp deletion on Chr 8:128,556,457-128,556,814 (Fig. 4A). Agarose gel electrophoresis confirmed the R3 region deletion (Fig. S2). Note that we did not rule out that due to either delayed Cas9 activity or contamina- www.nature.com/scientificreports/ The R3-deleted cells showed reduced MYC gene expression relative to wild-type cells (Fig. 4B). Using qRT-PCR analysis, we found that R2 eRNA (R2-S) and R3 eRNA (R3-S) expression were significantly decreased (Fig. 4C), and so was the expression of lncRNA CCAT1 (Chr 8: 127,207,382-127,219,268), a gene that is not part of the same TAD (Topologically associated domains) as MYC but is known to be regulated by MYC in HCC 30 . By contrast, the expression of the neighboring lncRNA gene, PVT1, which is in the same TAD as MYC, was unaffected (Chr 8: 127,795,799-128,101,256). Similarly, FAM49B, which is not in the same TAD, was scarcely affected (Chr 8: 129,841,470-130,016,672) 31 (Fig. 4D,E). We then analyzed the expression of two more MYCrelated genes, IRF2 and TERT, which are known to be repressed and activated by MYC, respectively 32,33 . In agreement, we found that IRF2 was upregulated in R3-deleted cells; however, TERT was downregulated (Fig. 4F). www.nature.com/scientificreports/ Next, we performed cell proliferation and colony formation assays. The R3-deleted cells exhibited a decreased proliferation and colony formation ability (Fig. 4G,H). Similarly, the R3 deletion reduced the spheroid formation by Huh7 cells (Fig. 4I).
It is known that MYC deletion is incompatible with proliferation 34 . It was therefore not surprising that our attempts to delete R2 and R3 did not result in well growing homogeneous cell lines. In fact, we could not recover R2-deleted cell cultures, suggesting a strong positive effect of R2 on MYC expression (see also further below the section on ASO inhibition). In contrast, we obtained passagable cultures that showed at least a significant proportion of R3-deleted cells (Fig. S2). These cultures exhibited a reduction of both the expression of MYC (Fig. 4B) and of their proliferation (Fig. 4G).
These results confirm that enhancer deletion influences cancer cell growth by reducing MYC expression 35 .  Table S1), and transfected into Huh7 cells. As expected, both eRNAs were specifically decreased by the corresponding ASOs (Fig. 5B,C). Importantly, both transfections also resulted in a significantly reduced MYC gene expression relative to the control (ASO NC) (Fig. 5D). Furthermore, when R2-S was targeted, the expression levels of the lncRNAs, PVT1, CCAT1, and FAM49B, were altered similarly as in the MYC-R3 deletion (Fig. 5E,   Fig. 4E). In addition, the expression of IRF2 and ICAM1, another gene known to be repressed by MYC 36 , was increased by both ASOs (Fig. 5F, also compared with Fig. 4F).We next analyzed the proliferation and spheroid-forming abilities of the ASO-transfected Huh7 cells. Both ASO R2-P1 and ASO R3-P2 significantly reduced cell proliferation compared to ASO NC (Fig. 5G), again mimicking the R3 deletion (compare with Fig. 4G). In addition, the spheroid-forming assay showed that inhibition of R2-and R3-induced eRNAs by ASO treatment negatively affected growth (Fig. 5H). These results indicate that inhibition of eRNA can mimic a direct deletion of the corresponding chromosomal DNA (compare with Fig. 4).

Discussion
The enhancer is a DNA region in which the E1A-binding proteins p300 (p300), RNAPII, and TFs are enriched through increased accessibility caused by histone modification. Enhancers help transcriptional activation directly through looping or indirectly affect transcription by expressing eRNAs 37,38 . Previous studies have revealed the presence of cancer-specific enhancers, the expression of eRNA in cancer cells, and the associated abnormal expression of oncogenes 39,40 . Since lineage-specific TFs affect the tumorigenesis function of their enhancers, they can also be cancer type-specific 41 . Therefore, we used ChIP-seq data from HepG2 cells as well as GRO-seq www.nature.com/scientificreports/ data that included nascent transcripts to identify eRNA expression, an active enhancer marker 42,43 . Based on these data, we found that the specific MYC enhancers R2 and R3 in HCC cell lines are significantly different from those in other cancers in terms of their location and activity 29,39,44 . In acute myeloid leukemia (AML), MYC expression is regulated by a SE that consists of five distinct small enhancers located 1.7 Mb downstream of the MYC promoter 40 . In colorectal cancer (CRC) and prostate cancer, the MYC enhancer is located 335 kb upstream of MYC 5 . In addition, the Zhang group deleted approximately 1.5 kb of the MYC enhancer located 450 kb downstream of the 3′ end in lung adenocarcinoma cells. MYC expression was reduced by 70%, and clonogenic growth was inhibited by approximately 50%. Note that we epigenetically identified the novel MYC enhancers in HepG2 cells (which are derived from a 15 years-old patient) 45 but functionally confirmed these enhancers in Huh7 cells (which are derived from a 57 years-old patient) 46 . This coincidence strongly suggests the general significance of these newly found enhancers in hepatocyte tumorigenesis. It will be interesting to see whether the same enhancers may also be active in normal proliferating liver cells (such as in embryogenesis or liver regeneration). eRNAs are generally upregulated in various cancers compared to normal tissues, and they can be used as pan-cancer diagnostic markers 47 . In addition, tissue-specific highly expressed eRNAs, such as CCAT1 in colorectal cancer and androgen receptor (AR)-induced Kallikrein-related peptidase 3 (KLK3) eRNA (KLK3e) in prostate cancer, are considered new targets for treating various cancers 14,48,49 . Although the function of eRNAs has not been fully elucidated, eRNA depletion reduces the transcription of target genes by affecting alterations in chromatin structure and contributing to transcriptional initiation of target genes 49 . eRNA transcription can be regulated by inhibiting enhancer activity or effectively targeting ASO to control target gene expression and cancer cell progression 50,51 .
Previous studies have found that the inhibition of MYC-related and other eRNAs by using ASOs can effectively inhibit tumor cell progression, suggesting that eRNAs can be helpful as therapeutic targets [51][52][53] . For example, Epstein-Barr virus (EBV) super-enhancer (ESE) RNAs facilitated the expression of the MYC oncogene in lymphoma, and targeting ESE eRNA showed a therapeutic effect on EBV-related malignancies 53 . Similarly, in our results, targeting MYC-IEANC RNAs transcribed from MYC enhancers R2 and R3 in Huh7 cells caused an effective decrease in MYC expression along with reduction of proliferation and spheroid formation, suggesting a therapeutic effect on HCC. A previous study demonstrating that downregulation of MYC suppressed spheroid growth of colon CSCs and tumor growth in vivo 54 , lends support to our suggestion. Of note, the ASO-mediated inhibition of MYC eRNA expression may be superior compared to an alternative strategy, in which BRD4 inhibitors also suppressed tumor cell-associated MYC expression 5,55-57 . Unlike BET inhibitors, the selective targeting of MYC eRNAs probably avoids side effects such as toxicity and resistance.

Conclusion
In this study, we identified the putative MYC enhancers of HCC cells. Enhancer activity and eRNA transcription were analyzed to determine the region involved in MYC expression, and it was found that both the deletion of the MYC enhancers and the ASO-mediated inhibition of the corresponding eRNAs suppressed the proliferation and reduced spheroid formation in HCC cell lines. Thus, our study suggests that for HCC, a strategy for reducing MYC expression through specific targeting with ASO has therapeutic potential without the side effects of gene editing or BRD4 inhibition. Future work will need to evaluate the role of the MYC enhancers identified here in HCC in vivo.

Materials and methods
Cell culture of HCCs. The HepG2 and Huh7 HCC cell lines were purchased from Korean Cell Line Bank (Seoul, Korea) and maintained in minimum essential medium or RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) and penicillin (100 units/ml)/streptomycin (100 μg/ml) (Thermo Fisher Scientific, Waltham, MA, USA). The medium was replaced every 3-4 days. The cells were maintained in a humidified incubator with 95% air and a 5% CO 2 atmosphere at 37 °C. JQ1, OTX015, C646, and 5, 6-dichlorobenzimidazole 1-β-d-ribofuranoside (DRB) were purchased from Tocris Bioscience (Minneapolis, MN, USA). JQ1, OTX015, C646, and DRB were dissolved in dimethyl sulfoxide (DMSO) at a stock concentration of 10 mM. The cells were treated with different concentrations of JQ1 and DRB for different durations.
Cell proliferation assays. Relative cell numbers were assessed using a premixed water-soluble tetrazolium salt (WST-1) cell viability test (Takara, Shiga, Japan) according to the manufacturer's instructions. The cells were seeded at a density of 1 × 10 4 cells per well. WST-1 was added to each well, and the absorbance of the microplate at 450 nm was measured after an additional 4 h incubation. The data represent three independent experiments (n = 3). DNA-synthesizing cells were visualized using an Ethynyl deoxyuridine (EdU) kit (Invitrogen, CA, USA) following the manufacturer's instructions. Then, the cells were washed with phosphate-buffered saline, mounted with a 4' , 6-diamidino-2-phenylindole (DAPI)-containing mounting solution (Vectashield, Vector Laboratories, Burlingame, CA, USA), and imaged by microscopy (Nikon Eclipse 80i, Tokyo, Japan). The percentage of EdUpositive cells was examined in HCC cell lines using ImageJ (Bethesda, MD, USA) software. The data represent three independent experiments (n = 3).
Genomic data analysis. We re-analyzed two public ChIP-seq data sets in Gene Expression Omnibus (GEO) (GSE29611 and GSM1112809) according to the procedure described previously 58 , and three GRO-seq data sets in GEO (GSE92375, GSM3271003, and GSM3271012). For our re-analysis, the raw data were trimmed with Trimmomatic (version 0.36) 59 63 . The transfected cells were seeded into 96-well plates at a ratio of less than 1 cell per well. After ~ 2 weeks, cells from wells that showed growth were moved into 24-well plates. After further growth, a portion of each well was used to extract genomic DNA (gDNA) using a Wizard Genomic DNA Purification Kit (A1125, Promega, Madison, WI, USA). Then, gDNA was amplified by PCR with target-specific primers and sequenced to check properly generated deletions. The cells of one confirmed well were expanded in 100 mm dishes and then used for the various gene expression and proliferation assays. These analyses were performed within the first 10 passages.
Knockdown of eRNA using ASO. Locked nucleic acid (LNA)-modified ASOs complementary to eRNA of MYC were designed from Antisense LNA GapmeRs (Qiagen, Hilden, Germany). The ASOs were purchased from Qiagen. The sequences are listed in Supplementary Table S3. For the transfection of Huh7 cells, ASOs were mixed with RNAiMAX in serum-free Opti-MEM (Gibco, Waltham, MA, USA). At varying concentrations of ASOs, dissolved Opti-MEM was added, and the cells were incubated in a growth medium for 4 h at 37 °C and 5% CO 2 . For total RNA extraction, the cells were harvested 48 h posttransfection.
Colony formation assay. R3-deleted and WT Huh7 cells were seeded on 6-well plates in growth media at a density of 2500 cells/well and incubated in a CO 2 incubator for 10 days. Then, the cells were washed with PBS, fixed with 4% paraformaldehyde for 20 min, and washed once with PBS. The cells were stained with 1% Crystal Violet (Sigma, St. Louis, MO, USA) for 30 min. After Crystal Violet was removed, the plates were washed with DW for 5 min and dried. The stained cells were analyzed for colony formation rates using ImageJ (Bethesda, MD, UAS).

Statistical analysis.
The data are presented as the mean ± standard deviation (SD) of the mean. All statistical analyses were performed using the IBM SPSS Statistics 26.0 program (IBM). We used a one-way analysis of variance followed by Tukey's honestly significant difference post hoc test. p values < 0.05 were considered significant.