Resistance to therapeutic drugs is a major challenge in the treatment of cancers, including breast cancer. Long noncoding RNAs (lncRNA) are known to have diverse physiologic and pathophysiologic functions, including in cancer. In searching for lncRNA responsible for cancer drug resistance, we identified an intergenic lncRNA ERINA (estrogen inducible lncRNA) as a novel lncRNA highly expressed in multiple cancer types, especially in estrogen receptor–positive (ER+) breast cancers. Expression of ERINA was inversely correlated with survival of patients with ER+ breast cancer and sensitivity to CDK inhibitor in breast cancer cell lines. Functional characterization established ERINA as an oncogenic lncRNA, as knockdown of ERINA in breast cancer cells inhibited cell-cycle progression and tumor cell proliferation in vitro and xenograft tumor growth in vivo. In contrast, overexpression of ERINA promoted cell growth and cell-cycle progression. ERINA promoted cell-cycle progression by interacting with the E2F transcription factor 1 (E2F1), which prevents the binding of E2F1 to the tumor suppressor retinoblastoma protein 1 (RB1). ERINA also functioned as an estrogen and ER-responsive gene, and an intronic ER-binding site was identified as an enhancer that mediates the transactivation of ERINA. In summary, ERINA is an estrogen-responsive oncogenic lncRNA that may serve as a novel biomarker and potential therapeutic target in breast cancer.

Significance:

These findings identify ERINA as an estrogen-responsive, oncogenic lncRNA, whose elevated expression may contribute to drug resistance and poor survival of patients with ER+ breast cancer.

Breast cancer is the most commonly diagnosed cancer and the second most frequent cause of cancer-related death in women (1, 2). Drug therapies remains to be the primary treatment for breast cancer before or after the surgery. Drugs that treat breast cancers include chemotherapeutic drugs that target cell cycles and metabolites (3), antihormone drugs, such as tamoxifen and aromatase inhibitors (4, 5), as well as the more recently developed CDK inhibitors (6, 7). However, the effectiveness of breast cancer drug therapies is limited by the development of resistance in the tumor tissues, which can be due to various genetic and epigenetic reasons (8, 9). The development of intrinsic and acquired drug resistance often leads to adverse prognosis of breast cancer. It is conceivable that a better understanding of the molecular basis of drug resistance will help to develop novel therapeutic strategies to overcome treatment failure.

Although mammalian genomes are extensively transcribed, the vast majority of these transcripts are noncoding RNAs (ncRNA), among which, are the long noncoding RNAs (lncRNA) with a length of over 200 nucleotides (10). Many lncRNAs have been functionally associated with human diseases including cancers. Indeed, genome-wide characterization of the human cancer transcriptome revealed that lncRNAs are among the most prevalent transcriptional changes in cancers (11, 12). Moreover, dysregulations of lncRNAs have been reported to impact cellular functions of cancer cells, such as proliferation, drug resistance, angiogenesis, metastasis, and evasion of tumor suppressors (13). LncRNAs exert their biological functions as regulatory RNAs, serving as signals, guides, decoys, or scaffolds to regulate the expression of a wide range of target genes (14). As examples, lncRNAs can recruit or block transcriptional factors, leading to the regulation of genes at the transcriptional level (15).

Knowing genetic factors play a major role in cancer drug resistance, many of the previous characterizations of genetic factors in drug resistance have been focused on protein-coding genes. There is a considerable lack of understanding of whether lncRNAs can mediate intracellular pathways that lead to resistance and if so, whether lncRNAs regulate drug resistance through their interactions with transcriptional factors.

In this study, we identified a novel long intergenic noncoding RNA, ERINA, whose elevated expression is highly correlated with poor survival of patients with ER+ breast cancer and sensitivity to CDK inhibitors in breast cancer cell lines. Mechanistically, ERINA is estrogen inducible, and can promote cell-cycle progression by targeting the transcriptional factor E2F1.

Preprocessing of lncRNAs expression data

For cancer cell lines, the expression of 2,614 cancer-related lncRNAs annotated by miTranscriptome (12) across 505 cancer cell lines from Cancer Cell Line Encyclopedia (CCLE) was downloaded from Expression Atlas (16). For patient samples, the expression of 2,614 cancer-related lncRNAs in 9,305 The Cancer Genome Atlas (TCGA) patient samples across 18 cancer types was downloaded from Genomic Data Commons (GDC) data portal. Expression levels of lncRNAs are logarithmic transformed and Z-score normalized for both cell lines and patients.

Preprocessing of drug response and patient clinical data

Drug response data of 265 agents across 1,001 cancer cell lines were downloaded from the Genomics of Drug Sensitivity in Cancer (GDSC) database (17). These 265 agents include 49 clinical drugs, 76 drugs in clinical development and 141 experimental agents. Our analyses are restricted to the response of 54 chemotherapy agents in this database. The drug response in each cell line is indicated by logarithmic transformed IC50s. For the patient clinical data, survival and molecular subtype information were parsed from the NCI Genomic Data Commons (GDC) using the R functions “GDCquery,” “GDCDownload,” and “GDCprepare_clinic” provided by TCGAbiolinks (18).

Association analysis between lncRNA expression and drug response

We first calculated the Spearman correlation between lncRNA expression and drug response across 505 cancer cell lines. To identify significant genes in the correlation analysis, we applied the Benjamini–Hochberg method to convert the correlation P values to FDRs and set FDR < 0.05 as a threshold of significance. On the basis of the significance threshold and the correlation coefficients, we converted the correlation matrix into a signed Boolean matrix indicating whether a given lncRNA is positively or negatively associated with a given drug at the given significance level. Next, we applied Uniform Manifold Approximation and Projection (UMAP) algorithm to the lncRNAs based on the signed Boolean matrix. The obtained two-dimensional embedding of lncRNAs revealed that many lncRNAs are associated with a drug response of multiple chemotherapy agents targeting different pathways.

Elastic-net regression modeling of drug response by lncRNA expression

For more details, see Supplementary Methods.

Cell lines

Human breast epithelial cell line MCF10A, and human breast cancer cell lines, BT-20, HCC1937, HS578T, MDA-MB-MB231, ZR-75–1, BT-474, MDA-MB-231, MCF7, T47D, and human embryonic kidney (HEK) 293T cells were obtained from the ATCC. MCF10A cells were cultured in complete MEBM medium [MEBM with 10% FBS, 100 ng/mL cholera toxin (APExBIO, B8326), and 1% penicillin and streptomycin (Sigma, P4333)]. BT-20 cells were cultured in EMEM medium (EMEM with 10% FBS and 1% penicillin and streptomycin. HCC1937, ZR-75–1, and T47D cells were cultured in RPMI1640 medium (RPMI1640 medium with 10% FBS and 1% penicillin and streptomycin). HS578T, MDA-MB-231, BT-474, and MCF7 cells were cultured in DMEM medium (DMEM medium with 10% FBS and 1% penicillin and streptomycin). All cell lines were tested for Mycoplasma contamination using the MycoAlert Mycoplasma Detection Kit (Lonza, LT07–118).

5′ and 3′ rapid amplification of cDNA ends analysis

For more details, see Supplementary Methods.

Cloning, shRNA construction, and lentiviral transduction

ERINA was amplified using PCR with iProof High-Fidelity DNA Polymerase (BIO-RAD), and subsequently cloned into XbaI and EcoRI digested pCDH-CMV-MCS-EF1-Puro (System Biosciences, CD510B-1). The shRNA sequences, as listed in Supplementary Table S1, were designed using BLOCK-iT RNAi Designer (Thermo Fisher Scientific; https://rnaidesigner.thermofisher.com/) and ordered from Thermo Fisher Scientific. shRNA oligos were inserted into AgeI and EcoRI digested pLKO.1 TRC cloning vector (Addgene, 10878). The ERINA and shRNA expressing lentiviruses were generated in HEK293T cells as described previously (19, 20). Briefly, ERINA and shRNA-expressing vectors were separately cotransfected with psPAX2 and pMD2.G vectors into 293T cells. Supernatant was collected at 48 hours posttransfection. The culture medium containing the lentiviruses was filtered through a 0.45-μm filter. Target cells were infected with viruses in the presence of polybrene and selected using puromycin to establish stable cells.

qRT-PCR and subcellular fractionation

Total RNA was extracted using the RNeasy Mini Kit (Qiagen, 74104). Reverse transcription from 1 μg RNA to cDNA was performed using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, 4368813). Real-time PCR was performed with Power SYBR Green PCR Master Mix (Applied Biosystems, 4367659) on a QuantStudio 6 Flex Real-Time PCR System (Applied Biosystems). Results were analyzed using the comparative Ct method normalizing to GAPDH. The sequences of qRT-PCR are listed in Supplementary Table S2. Separation of nuclear and cytosolic fractions was performed using the PARIS Kit (Ambion). Briefly, approximately 5 × 106 cells were first lysed with cell fractionation buffer on ice and spun down. Supernatants were kept for the cytoplasmic fraction. Pellets were further lysed with cell disruption buffer on ice for the nuclear fraction. Cytoplasmic and nuclear fractions were separated for RNA and protein extraction.

Proliferation assay

A total of 2 × 103 cells per well were seeded in 150 μL culture medium into 96-well plates. After 4 days of growth, MTT assay was performed using the CellTiter 96 Non-Radioactive Cell Proliferation Assay Kit (Promega, G4100). Relative cell growth was determined by normalizing to the values at day 0. Each time point represents means ± SEM of at least three replicates.

In vitro drug resistance by means of the MTT assay and clone formation assay

A modified MTT colorimetric assay was used to detect the sensitivity of cells to anticancer drugs in vitro as described previously (21). In brief, cells were trypsinized, resuspended, and seeded at 5 × 103 cells/well into 96-well plates and incubated for 24 hours. The cells were then treated with palbociclib (0–1 μmol/L) for 72 hours before the MTT assay. For the clone formation assay, shCtrl cells and ERINA knockdown cells were seeded in 12-well plates at 1 × 104 cells/per well. Cells were treated with palbociclib at the concentration of 0.1 μmol/L. The drug supplemented medium was replaced every 2 days. After 10 days, cells were subjected to formaldehyde fixation, crystal violet staining, photographing, and counting.

Soft agar colony formation assay

Agar (1.2%) was melted in a microwave and cooled to 40°C in a water bath. Tissue culture medium containing 20% FBS was warmed to 40°C in a water bath. Mix equal volumes of the two solutions to give 0.6% Agar + 1X medium + 10% FBS. Add 2 mL of mixture to each well of 6-well plate and set aside for 10 minutes to allow agar to solidify. Then, 1 mL of cell mixture containing 1 × 104 cells in culture medium containing 0.35% agarose with 10% FBS was carefully plated on top of the bottom layer. The plates were incubated at 37°C in a humidified incubator for 2 to 3 weeks. Cells were fed with fresh cell culture media 1 to 2 times per week. Plates were stained with 0.5 ml of 0.005% crystal violet for more than 1 hour. Colonies were counted under a microscope.

Cell-cycle analysis

Cells were harvested and washed in PBS. Cells were fixed in cold 70% ethanol for at least 2 hours at −20°C. Cells were then washed twice in PBS and once in BD Pharmingen stain buffer (BD Biosciences, 554656). The supernatant was discarded after spinning down. Cell pellets were resuspended in 0.5 mL of BD Pharmingen PI/RNase staining buffer (BD Biosciences, 550825) and incubated for 15 minutes at room temperature, and cells were immediately analyzed using an LSRFORTESSA X-20 flow cytometer (BD Biosciences). The data was analyzed with the FlowJo software.

Cell line–derived xenograft model

Five- to 6-week-old female athymic nude mice (Charles River Laboratories) were used for the xenograft experiments. A total of 1 × 106 shRNA infected and selected T47D and BT474 cells were resuspended in Matrigel (Corning, 356231) and injected subcutaneously into the right flank of the mice. Mice were monitored twice a week for tumor growth, and tumor size was measured using a caliper. Tumor volume was calculated using the formula: tumor volume = 0.5 × (width)2 × length. For the tumor growth curve, mean ± SEM tumor volume was plotted for each experimental group at each of the time points. Tumor weight on the final day of measurement was also plotted. All animal studies were performed in accordance with the institutional guidelines, and the experiments followed the protocols approved by the Institutional Animal Care and Use Committee of the University of Pittsburgh (Pittsburgh, PA).

RNA sequencing analysis

The RNAs of control and ERINA knockdown cells were extracted using the RNeasy Mini Kit (QIAGEN, 74106). RNA sequencing was performed at the Health Sciences Sequencing Core at the Children's Hospital of Pittsburgh (Pittsburgh, PA). Differential expression analysis was performed by gene set enrichment analysis (GSEA). We used STAR and RSEM to profile RNA-seq data of T47D cell lines after ERINA knockdown. The gene expression was quantified in log2-transformed FPKM and annotated based on the human reference genome GRCh38. The RNA-seq data can be downloaded from GEO (accession number GSE147901). To interpret the function of regulated genes after ERINA shRNA treatment, GSEA was performed using the 50 cancer hallmark gene sets and C2 curated studies on the log2-fold change of gene expression.

E2F overexpression plasmids and siRNA knockdown of E2F1 and RB

The full-length E2F1 plasmid was purchased from GenScript (NM_005225.3. The E2F1 deletion mutant constructs were kindly provided by Dr. Jiandong Chen from H. Lee Moffitt Cancer Center and Research Institute (Tampa, FL). The siRNAs of E2F1 (sc-29297) and RB (sc-29468) were purchased from Santa Cruz Biotechnology.

Luciferase reporter assay

The E2F1-responsive 3x-E2F-Luc luciferase vector was constructed by inserting three copies of the E2F-binding motifs into the pGL3 backbone vectors as described previously (22). The CCNA2-luc luciferase vector containing the CCNA2 gene promoter was described previously (23). The 3x-wtERE-Luc and its mutant variant 3x-mutERE-Luc vectors were constructed by inserting three copies of the wild-type and mutant estrogen receptor (ER) response element (ERE) into the tk-Luc vector, respectively. Cells were transiently transfected with the luciferase vectors with the cotransfection of other appropriate plasmids using Lipofectamine 2000 (Invitrogen, 11668). A β-gal expression vector was cotransfected as an internal control to normalize the transfection efficiency. After 48 hours, the luciferase and β-gal activities were measured using a Victor2 Microplate Reader (PerkinElmer) as described previously (24). The luciferase activities were normalized to the β-gal activities. Data were shown as fold changes over the control groups.

Biotin RNA pull-down assay

RNA pulldown assays were performed as we have described previously (25). In brief, biotinylated full-length or truncated ERINA were synthesized by T7 RNA polymerase using the Biotin RNA Labeling Mix (Roche, 11685597910), and then incubated with the RNA structure buffer (10 mmol/L Tris-HCl pH 7.0, 0.1 mol/L KCl, 10 mmol/L MgCl2) to allow the formation of the RNA secondary structures. The nucleus of cells were collected and lysed with cell lysis buffer (150 mmol/L KCl, 25 mmol/L Tris-HCl pH 7.4, 0.5 mmol/L DTT, 0.5% NP-40, 1 mmol/L PMSF, 1× Superase-in, and 1× protease inhibitor cocktail). Cell lysates were precleared with the magnetic beads (Thermo Fisher Scientific, 65001). Then, the lysates were incubated with biotinylated DNA probes at 4°C for 4 hours with rotation. Magnetic beads were then added to the lysate and incubated at room temperature for an additional hour with rotation. Beads–probe–RNA–protein complexes were collected and washed 5 times. Proteins were eluted from the beads complexes for Western blot analysis.

RNA immunoprecipitation analysis

RNA immunoprecipitation (RIP) was performed as we have described previously (26). Briefly, cells were washed with PBS and fixed with 0.3% formaldehyde, and quenched with 0.125 mol/L glycine. The cells were subsequently washed twice with cold PBS, and the pellets were resuspended in RIPA buffer [50 mmol/L Tris-HCl, pH 7.4, 150 mmol/L NaCl, 1 mmol/L EDTA, 0.1% SDS, 1% NP-40, 0.5% sodium deoxycholate, 0.5 mmol/L DTT, 1 mmol/L PMSF, and 1× protease inhibitor cocktail (Sigma, P8340)] and incubated on ice for 30 minutes with shaking. The cleared lysates were incubated with 5 μg of anti-E2F1 antibody or IgG bounded to the magnetic beads (Thermo Fisher Scientific, 88847) at 4°C overnight. Beads were washed six times with RIPA buffer (50 mmol/L Tris-HCl, pH 7.4, 1 mol/L NaCl, 1 mmol/L EDTA, 0.1% SDS, 1% NP-40, and 0.5% sodium deoxycholate) before elution, reverse cross-linking, and subsequent RNA extraction. The recovered RNA was subjected to qRT-PCR analysis, and relative fold enrichment and percentage of input was calculated using the comparative Ct method normalizing to input sample.

Chromatin immunoprecipitation assay

This was performed as we have described previously (27). Briefly, 1 × 107 cells were cross-linked with a final concentration of 1.42% formaldehyde in tissue culture medium for 15 minutes at room temperature, and cross-linking was quenched by the addition of glycine to a final concentration of 125 mmol/L and incubated for 5 minutes at room temperature. Cells were rinsed twice with cold PBS, harvested in IP buffer (50 mmol/L pH 7.5 Tris-HCl, 150 mmol/L NaCl, 5 mmol/L EDTA, 0.5% NP-40, and 1% Triton X-100) supplemented with 1 mmol/L PMSF and 1× protease inhibitor cocktail, and sonicated to shear the chromatin to yield DNA fragment sizes of 0.5 to 1 kb. Samples were cleared by centrifuging at 12,000 × g for 10 minutes at 4°C and preincubated for 1 hour with 40 μL of protein A/G agarose beads. A portion of the precleared samples was used as the input DNA. Then, approximately 5 μg of E2F antibody, ERα antibody, or the control IgG was added to the remainder of the samples and incubated for 1 hour at 4°C before 40 μL of protein A/G agarose beads (Thermo Fisher Scientific, 20421) were added, and the mixture was incubated for 4 hours at 4°C. Beads were washed six times with cold IP buffer, and DNA was isolated with 10% Chelex, and the total input DNA was also isolated. Quantification of the recovered DNA was performed using real-time PCR with SYBR Green Master Mix. Control IgG and input DNA signal values were used for normalization. The chromatin immunoprecipitation (ChIP) primer sequences are listed in Supplementary Table S2.

Coimmunoprecipitation, protein isolation, and Western blotting

The cell pellets were washed with cold PBS, and the supernatant was removed after spinning down. The cell pellets were then lysed in lysis buffer (50 mmol/L Tris-HCl pH 7.5, 150 mmol/L NaCl, 1 mmol/L EDTA, 1% Triton X-100, PMSF freshly added to a final concentration of 1 mmol/L, and 1× protease inhibitor cocktail). After quantification of protein concentrations using a BCA Protein Assay Kit (Thermo Fisher Scientific, 23225), 1 mg of total protein was used for coimmunoprecipitation (Co-IP) and incubated for overnight with 2 μg of corresponding antibodies. The prewashed magnetic beads were incubated with the cell lysate for another 2 to 4 hours, and then beads were washed at least 4 times, and boiled in 1× SDS sample buffer (Bio-Rad, 161–0737) for 10 minutes before gel loading. Proteins were resolved by SDS-PAGE and transferred onto PVDF membranes (Bio-Rad, 162–0177). Membranes were blocked with 5% skim milk at room temperature, and then incubated with the appropriate primary antibodies at 4°C overnight, and then secondary antibodies at room temperature for 1 hour. Antibody-bound proteins were visualized with enhanced chemiluminescence substrate (Thermo Fisher Scientific, 32106). Antibodies used in this study are listed in Supplementary Table S3.

Statistical analysis

All statistical analyses were performed using Prism software (GraphPad_Prism_7.0). Differences between two groups were calculated using a two-tailed Student t test. Analysis involving multiple groups was performed using one-way ANOVA. Survival curves were calculated using Kaplan–Meier and log-rank tests. P values of ≤0.05 were considered statistically significant.

ERINA is a novel lncRNA overexpressed in ER+ breast cancer and is associated with poor clinical survival

To identify lncRNAs responsible for cancer drug resistance, we obtained the expression of 2,614 cancer-related lncRNAs from CCLE (16) and the drug response data of 54 chemotherapy agents across 505 cancer cell lines from the GDSC database (17). The chemotherapy agents in the GDSC dataset include FDA approved drugs and preclinical compounds that target cell cycle, cytoskeleton, DNA replication, genome integrity, and mitosis (17). As a pilot analysis of lncRNAs as predictive markers of drug resistance, we evaluated the association between lncRNA expression and drug response using Spearman correlation. In total, 1,331 of 2,614 lncRNAs showed significant association with at least one drug resistance (FDR ≤ 0.05), among which, 348 were associated with the resistance to more than 10 drugs. Through the UMAP (28) embedding of lncRNAs based on their association pattern with the drug response in cell lines, we found that many drug resistance predicting lncRNAs can correlate with the sensitivity to drugs that target many different cellular pathways (Fig. 1A). These lncRNAs include previously reported drug resistance lncRNAs such as MIR22HG (29, 30) and NEAT1 (31, 32), as well as novel lncRNAs whose functions are yet to be defined including the top-ranked ERINA (Fig. 1B). To further identify robust and nonredundant drug resistance lncRNAs, we implemented elastic net (EN) regression with bootstrapping process and build a lncRNA-based EN prediction (LENP) model for each agent using a previously reported training-testing framework (19). This analysis further identified 426 lncRNAs that can predict drug response of up to 28 agents. Among these lncRNAs, ERINA was found to predict the resistance of 28 chemotherapy agents in cancer cell lines and was ranked as the top predictor of 13 agents including the clinic drugs, such as doxorubicin (Fig. 1C) and gemcitabine (Fig. 1D). Analysis of the TCGA database showed that ERINA is overexpressed in TCGA tumor samples of several cancer types, including breast cancer (BRCA), prostate cancer (PRAD), lung cancer (LUSC), and uterine cancer (UCEC), but not in bladder cancer (BLCA) or cervical cancer (CESC; Fig. 1E). Because doxorubicin is widely used in breast cancer chemotherapy, we further looked into the expression of ERINA in patients with breast cancer and found ERINA had a particularly high expression in the ER+ patients compared with the ER cohorts (Fig. 1F). Consistent with results from the TCGA breast cancer tumors, the expression of ERINA was found to be high in ER+ breast cancer cell lines including T47D, MCF7, MB361, BT474, and ZR-75–1 cells, but not in a panel of ER cell lines as shown by qRT-PCR (Fig. 1G). Moreover, the high expression of ERINA was significantly associated with worse overall survival in ER+ patients (Fig. 1H) but not in ER patients (Fig. 1I), and similar associations were observed in an independent patient cohort GSE6532 (Supplementary Fig. S1A). The expression of ERINA was positively associated with the expression of cell proliferation marker genes PCNA and Ki67 in ER+ patients, particularly, in the LumA subtype (ρ = 0.25, P = 2.13e-08 for PCNA; ρ = 1.16, P = 0.001 for Ki67; Supplementary Fig. S1B), which may have contributed to the poor survival of ERINA high expression patients.

ERINA is an intergenic nuclear lncRNA

ERINA is an intergenic lncRNA located on chromosome 5q23.1. Detailed annotation of the ERINA gene locus was accomplished by using MiTranscriptome, a comprehensive lncRNA catalog (12). The ERINA gene contains three exons, and the transcript of ERINA was predicted to be 3262 nt in length (Supplementary Fig. S1C). We went on to clone ERINA by rapid amplification of cDNA ends (RACE) and the sequences were verified by DNA sequencing. Because the subcellular localization of lncRNAs is important for their functions (33), we separated the nuclear and cytoplasmic fractions of T47D and MCF7 cells, and measured the expression of ERINA in different fractions. As shown in Supplementary Fig. S1D, 75% to 78% of ERINA was enriched in the nucleus, suggesting ERINA may have important nuclear functions.

ERINA drives tumor cell growth in vitro and in vivo

To investigate the function of ERINA, we generated ERINA knockdown T47D and BT474 stable cell lines by using two independent lentiviral shRNAs, and the efficiencies of knockdown were verified by qRT-PCR (Supplementary Fig. S2A). T47D cells were chosen because they had the highest endogenous expression of ERINA, and BT474 cells were chosen because they had a relatively lower endogenous expression of ERINA (Fig. 1G). Knockdown of ERINA in both T47D and BT474 cells decreased the cell proliferation and colony formation as shown by MTT assay (Fig. 2A and C) and soft agar assay (Fig. 2B and D), respectively. Also, the morphologies of both T47D (Supplementary Fig. S2B) and BT474 (Supplementary Fig. S2C) cells were obviously changed after ERINA knockdown in that the knockdown cells became smaller and abnormal in shape. In vivo, knockdown of ERINA markedly inhibited the growth of xenograft tumors derived from T47D and BT474 shERINA cells compared with their shCtrl counterparts (Fig. 3E–J). The inhibition of tumor cell proliferation in the BT474 xenografts was also confirmed by Ki-67 immunostaining (Fig. 3K). In the gain-of-function model, overexpression of ERINA in ZR-75–1 cells promoted cell growth as shown by MTT (Fig. 3L) and soft agar assay (Fig. 3M). ZR-75–1 cells were chosen for overexpression because they had a very low expression of the endogenous ERINA (Fig. 1G), and the overexpression was verified by qRT-PCR (Supplementary Fig. S2D). These results demonstrated that ERINA is an oncogenic lncRNA.

ERINA promotes cell-cycle progression by targeting the E2F pathway

In understanding the oncogenic mechanism of ERINA, we found shRNA knockdown of ERINA in T47D cells led to cell-cycle arrest in the G0–G1 phase, which was accompanied by less cells in the S and G2–M phases (Fig. 3A). In contrast, overexpression of ERINA in ZR-75–1 cells led to promotion of cell-cycle progression with more cells in the S and G2–M phases (Fig. 3B).

To investigate the mechanism by which ERINA promotes cell-cycle progression, we performed RNA-seq analysis on T47D cells transfected with either control shRNA or two independent shRNAs targeting ERINA. Differential gene expression analysis identified 120 and 275 genes commonly upregulated (Supplementary Fig. S3A) and downregulated (Supplementary Fig. S3B) in both shERINA cells. Gene ontology (GO) analysis revealed that the commonly downregulated genes were highly enriched in cell cycle, mitotic prometaphase, DNA replication, and mitotic M_G1 phase pathways (Fig. 3C). The gene expression profile was consistent with the phenotype of inhibition of cell proliferation and tumor growth of the ERINA knockdown cells. To further explore the role of ERINA in clinical breast cancers, we used GSEA to identify genes whose expressions were correlated with ERINA expression across 559 TCGA breast tumors. Then, we overlapped the ERINA correlated genes in patients with those in T47D ERINA knockdown cells (Fig. 3D), and found the common signaling pathways affected by ERINA overexpression in breast tumors and ERINA knockdown in T47D cells included the E2F Target pathway (Fig. 3E). We went on to verify that ERINA knockdown T47D cells indeed showed decreased mRNA (Fig. 3F) and protein (Fig. 3G) expression of a panel of cell cycle–related genes, such as CCNA2, CDC20, CDC45, and CDK6. In contrast, ERINA overexpressing ZR-75–1 cells showed increased mRNA (Fig. 3H) and protein (Fig. 3I) expression of cell cycle–related genes.

Among genes downregulated by ERINA knockdown, CDC20, CDK2, CCNA2, CCNB2, CCNE1, and MCM2 are known E2F target genes, consistent with the RNA-seq results that the E2F pathway was downregulated in the ERINA knockdown cells (Fig. 3E). The E2F family of transcription factors promote cell-cycle progression by regulating the expression of genes involved in DNA replication and cell proliferation (34). These results led to our hypothesis that the E2F pathway might be the target of ERINA. To directly test whether ERINA can enhance the transcriptional activity of E2F, we generated the E2F-responsive 3X-E2F-Luc reporter gene by inserting three copies of the canonical E2F-binding motifs upstream of a luciferase reporter gene. This E2F binding motif 5′-TTTC[CG]CGC-3′ was found in the promoter regions of several E2F target genes involved in cell-cycle regulation (22). As shown in Fig. 3J, the E2F reporter activity was significantly decreased in T47D cells cotransfected with shERINA compared with cells transfected with shCtrl. Knockdown of ERINA also decreased the activity of a reporter gene that contained the promoter of the E2F target gene CCNA2 (Fig. 3K). In contrast, overexpressing ERINA in ZR-75–1 cells enhanced E2F1 responsive E2F reporter activity (Fig. 3L). Together, our results suggested that ERINA may promote cell-cycle progression by targeting the transcriptional factor E2F1.

ERINA interacts with E2F1 within the RB-binding domain

Knowing E2F is a potential target of ERINA and nuclear lncRNAs like ERINA often affect gene expression by interacting with gene regulatory proteins (35–37), we hypothesized that ERINA may interact with E2F and thus enhance the cell cycle–promoting effect of E2F. To test this hypothesis, we performed a biotin-labeled ERINA pull-down assay and found that E2F1 can indeed bind to the sense transcript, but not the antisense transcript of ERINA (Fig. 4A). The E2F1–ERINA interaction was further confirmed by the RIP assay showing that an E2F1 antibody can precipitate the endogenous ERINA (Fig. 4B). To map the E2F1-binding region within ERINA, we generated exon-specific deletion mutants of ERINA as outlined in Fig. 4C. In vitro biotin-labeled RNA pull-down assay using these deletion mutants showed the exon 1 of ERINA was required for its binding to E2F1, because the exon 1 deletion mutant failed to bind to E2F1 (Fig. 4D). We then performed further fine mapping by generating 8 deletion mutants within the exon 1. Our results showed the E2F1-binding regions within the exon 1 were mainly localized in nt 61–240 and nt 301–420, because deletions of these two regions resulted in a significant decrease of binding to E2F1 (Fig. 4E). At the functional level, exon 1 deletion mutant ERINA can no longer enhance the activity of E2F1 in the luciferase reporter gene assay (Fig. 4F). Our structural predictions of ERINA suggested that exon 1 was required for the formation of a secondary structure (Fig. 4G), which is known to be essential for the binding of lncRNAs to their target proteins (38). Consistent with the fine mapping results, deletions of nt 61–240 or nt 301–420 were predicted to be sufficient to disrupt the formation of the secondary structure (Fig. 4G).

To map the ERINA-binding domain of E2F1, 293T cells were transfected with FLAG-tagged WT or a panel of deletion mutants E2F1s (39) as outlined in Fig. 4H, followed by a pull-down assay using in vitro transcribed WT ERINA. Our mapping results revealed the region between amino acids 382–437 of E2F1 was required for binding to ERINA (Fig. 4I).

ERINA competes with RB for its binding to E2F1, and the RB/E2F1 axis is required for the effect of ERINA on cell cycle

Interestingly, the ERINA-interacting region of E2F1 overlaps with its retinoblastoma-associated protein (RB) binding domain, which was mapped to residues 409–426 (40). The RB protein (pRB), encoded by the RB transcriptional corepressor 1 (RB1) gene, is a tumor suppressor. RB proteins negatively regulate the cell cycle by binding to and inhibiting the transcriptional activity of E2F1 (41). Our RNA-seq analysis also showed ERINA played a crucial role in RB1-involved pathway, as shown by GO analysis (Supplementary Fig. S4A) and GSEA analysis (Supplementary Fig. S4B).

The E2F1-binding interface shared by both ERINA and RB suggested that ERINA may have enhanced the transcriptional activity of E2F1 by the sequestration of RB. Indeed, our Co-IP results showed that the E2F1–pRB interaction was increased upon ERINA knockdown in both T47D and BT474 cells (Fig. 5A). At the functional level, ERINA knockdown in T47D cells reduced the recruitment of E2F1 onto the promoter of its target genes, including TIMENESS, UBE2T. CYCLINE, CDK1, CDC6, and CCNA2, as shown by ChIP analysis (Fig. 5B).

To determine whether the RB/E2F1 axis is required for the effect of ERINA on the cell cycle in breast cancer cells, we knocked down RB in T47D shCtrl and shERINA cells, and the knockdown efficiency was verified by Western blotting (Fig. 5C). The cell-cycle analysis showed that the effect of ERINA knockdown on the inhibition of cell-cycle progression was attenuated upon RB knockdown (Fig. 5C), suggesting the effect of ERINA knockdown on cell cycle was RB dependent. The knockdown of E2F1 was performed in ZR-75–1 ERINA-overexpressing cells, and the knockdown efficiency was also confirmed by Western blotting (Fig. 5D). As shown in Fig. 5D, the cell cycle–promoting effect of ERINA overexpression in ZR-75–1 cells was abolished upon E2F1 knockdown, suggesting the effect of ERINA overexpression on cell cycle was E2F1 dependent. In determining the effect of RB knockdown on ERINA-responsive E2F1 target gene expression, we found RB knockdown attenuated or abolished the downregulation of several E2F1 target genes in ERINA knockdown cells, such as CDC20, CDK2, CDK6, CCNA2, and CCNE1 (Fig. 5E). Given the link between the expression of ERINA and ER status, we segregated patients with TCGA breast cancer based on their differential expression of 50 genes (PAM50) profile, and observed an overall higher expression of E2F target genes in luminal A and B subtypes (Fig. 5F and G).

ERINA is an estrogen-responsive gene

Having shown ERINA was an oncogenic lncRNA highly expressed in patients with ER+ breast cancer and cell lines, and knowing the expression of ERINA was positively correlated with poor survival of patients with ER+ breast cancer, we speculated that ERINA is an estrogen-responsive gene. Indeed, treatment with estradiol (E2) induced the expression of ERINA along with the known ER target gene TFF1 in the ER-positive T47D and MCF7 cells, but not in the ER-negative BT20 cells (Fig. 6A). In contrast, treatment of T47D and MCF7 cells with the ER antagonist fulvestrant (ICI 182780) or siRNA knockdown of ERα decreased the basal expression of ERINA and TFF1 (Fig. 6B). The efficiency of ERα siRNA knockdown in T47D and MCF7 cells was verified by Western blotting (Fig. 6C). These results were also supported by our bioinformatic analysis of the NCBI GEO database showing that treatment with E2 (Fig. 6D) and siRNA knockdown of ERα (Fig. 6E) increased and decreased the expression of ERINA, respectively.

To understand the mechanism by which ER regulates the expression of ERINA, we analyzed the publicly available ER ChIP-seq data on MCF7 and BT474 cells. No prominent ER binding peak was found within the promoter region of ERINA. However, a marked ER binding peak was found in the intron region in both MCF7 and BT474 cells, and this peak only appeared in MCF7 cells when treated with E2 (Fig. 6F). Enrichment of H3K27Ac is a hallmark of active enhancers (42). Analysis of publicly available MCF7 cell H3K27Ac and RNA polymerase II ChIP-seq results revealed an enrichment of H3K27Ac binding signal around the ER binding peak (Fig. 6F), as well as a strong enrichment of RNA polymerase II binding signal in the promoter region of ERINA (Supplementary Fig. S5A). These results strongly suggested that this ER-binding peak represented an active ER-responsive enhancer.

A putative ERE (5′-AGGTTGActcTGACCT-3′) was found in the peak region. This ERE was functional, because treatment of MCF7 cells with E2 increased the recruitment of ERα to this ERE as shown by ChIP analysis (Fig. 6G). The recruitment of ERα to the TFF1 gene promoter was included as a positive control. Moreover, the tk-(ERE)x3-Luc reporter gene that contains three copies of ERINA ERE, but not its mutant variant (5′-AGTTTCActcTTATCT-3′ with the mutated nucleotides underlined), was activated in 293T cells cotransfected with ERα and treated with E2 (Fig. 6H). Taken together, our results revealed ERINA as an estrogen-responsive gene and the intronic ERE may have functioned as an enhancer to mediate the induction of ERINA.

ERINA affects responses of ER+ breast cancer cells to tamoxifen and CDK inhibitors

Because ERINA is an estrogen-responsive lncRNA in ER+ breast cancers, we went on to determine whether ERINA expression affected the therapeutic response to tamoxifen, the first-line endocrine therapy for patients with ER+ breast cancer. We first selected a total of 225 ER+ patients who have been treated with tamoxifen from the TCGA patient cohort, and found patients with high expression of ERINA was associated with poor prognosis (Fig. 7A, left). The same pattern of association was independently validated in two addition datasets GSE6532 (n = 263) and GSE9195 (n = 77; Fig. 7A, middle and right). These results suggested that although the treatment of tamoxifen may have the tendency to suppress the expression of ERINA, the expression level of ERINA continues playing a dominating role in determining the clinical outcome. To determine whether ERINA can affect the estrogen–ER signaling, we treated ERINA knockdown cells with E2 or tamoxifen. As expected, treatment with E2 induced the expression of ERINA and the known ER target gene TFF1 (Fig. 7B), whereas tamoxifen suppressed the expression of ERINA and TFF1 (Fig. 7C). The E2 and tamoxifen-responsive regulation of TFF1 was intact in ERINA knockdown cells (Fig. 7B and C), suggesting that although ERINA is a downstream target of ER, it had little effect on the estrogen–ER signaling.

CDK4/6 inhibitors such as palbociclib suppress breast cancer by inhibiting CDK-mediated RB phosphorylation and preventing cell-cycle progression. Accordingly, deficiencies in the RB–E2F1 axis has been reported as a major mechanism for ER+ breast cancer patients' resistance to palbociclib (43). Because we have shown that ERINA competes with RB for its binding to E2F1 and promotes cell cycle, we went on to determine whether the expression of ERINA affected breast cancer cell response to CDK4/6 inhibitors. Analysis of published RNA-seq dataset GSE130437 showed the expression of ERINA was significantly elevated in palbociclib-resistant ER+ MCF-7 cells compared with the parental palbociclib-sensitive MCF-7 cells (Fig. 7D). The expression of ERINA was barely detectable in ER- MDA-MB-231 cells regardless of their sensitivity to palbociclib. These results suggested that high expression of ERINA may have contributed to the resistance to palbociclib in ER+ breast cancer cells.

To experimentally demonstrate the effect of ERINA on the therapeutic response to palbociclib, we found that knockdown of ERINA sensitized T47D cells to increasing concentrations of palbociclib (Fig. 7E). The combined ERINA knockdown and palbociclib treatment further reduced cell growth than palbociclib alone, as shown by cell proliferation assay (Fig. 7F) and colony formation assay (Fig. 7G), respectively. Collectively, our results showed the expression of ERINA had a major effect on the therapeutic response to CDK inhibitors, consistent with the notion that RB and E2F1 are common downstream targets of ERINA and CDK inhibitors.

Cancer treatment has advanced a great deal with continual development of therapeutic drugs since the 1940s. However, resistance to cancer therapeutic drugs remains a major clinical challenge. Intrinsic factors attained within the tumor can cause the induction of drug resistance. In the current study, we have discovered and functionally characterized a novel lncRNA ERINA whose expression level can predict sensitivity to CDK inhibitors in breast cancer.

The most interesting finding of ours is that ERINA promotes drug resistance by promoting cell cycle progress through targeting the E2F1/RB1 pathway. An often-used mechanism by which lncRNAs affect gene expression is recruiting or blocking transcriptional factors' binding to DNA or its coregulators (44). In this study, we have mechanistically demonstrated that ERINA can directly bind to E2F1 through the first exon. Further mapping results showed the ERINA-binding region within E2F1 overlapped the Rb family binding domain of E2F1. These results led to our hypothesis that ERINA enhances the transcriptional activity of E2F1 by sequestration of pRB, a well-established tumor suppressor that interacts with and inactivates E2F1. When ERINA is overexpressed and RB is sequestered, E2F1 is freed from the RB inhibition and subsequently induces the expression of its target genes, especially the cell cycle–related genes. We have also provided direct evidence that the RB/E2F1 axis is required for the effect of ERINA on cell cycle.

The notion that the RB–E2F1 axis is required for the oncogenic activity of ERINA was supported by our observation that the expression of ERINA was positively correlated with resistance to and affected sensitivity to CDK inhibitors, a new class of anti–breast cancer drugs that target RB to prevent cell-cycle progression. E2F1 has been previously reported to play an important role in the development of cancer drug resistance. For example, E2F1 induces the expression of genes involved in DNA repair that can protect tumor cells from the lethal effect of DNA-damaging agents, such as doxorubicin and docetaxel (45). Besides promoting cell cycles, E2F1 has also been reported to promote chemoresistance by inducing multidrug resistance related genes, such as the ABC transporters and drug metabolizing enzymes (46). It will be interesting to determine whether and how ERINA may affect the expression of multidrug resistance genes.

Another interesting finding is that ERINA is an estrogen-inducible lncRNA. This explains why the expression of ERINA is specifically elevated in ER+ breast cancer and positively associated with poor prognosis in patients with ER+ breast cancer. The poor survival of ERINA-overexpressing and patients with ER+ breast cancer may have been accounted for by its activities in oncogenesis and drug resistance. Mechanistically, we found a strong binding between ERα and the intron region of ERINA. An ER-binding element was identified within the intronic region and the functionality of this ER-binding element was confirmed by luciferase reporter assay and ChIP analysis. The coenrichment of H3K27Ac strongly suggested that this intronic ERα-binding region functions as an enhancer to mediate the estrogen-responsive induction of ERINA. The intronic regulation of ERINA was reminiscent of the reported regulation of lncRNA PCAT19 by an intronic enhancer (25).

Besides ERINA, several other lncRNA species have been implicated in cancer drug resistance via various mechanisms. We recently reported that EPIC1, the top lncRNA that can predict resistance to the bromodomain and extraterminal motif (BET) inhibitors, strongly promoted resistance to iBET762 and JQ-1 by increasing the transcriptional activity of the MYC oncogene (19). Our subsequent characterization showed that EPIC1 promotes cell-cycle progression by its direct interaction with MYC. EPIC1 knockdown reduces the occupancy of MYC to its target genes (23). The oncogenic activity of EPIC1 in vitro and in vivo was abolished when MYC was depleted, reminiscent of the dependence of the RB/E2F1 axis in the cell-cycle effect of ERINA observed in our current study. Among other examples, LncRNA Activated in RCC with Sunitinib Resistance (lncARSR) was reported to promote doxorubicin resistance in hepatocellular carcinoma by activating the PTEN-PI3K/Akt pathway (47). LncRNA FAM84B-AS promotes resistance of gastric cancer to platinum drugs through the inhibition of FAM84B expression (48). Knockdown of LncRNA-H19 decreases resistance of human glioma cells to temozolomide by suppressing the epithelial–mesenchymal transition via the Wnt/β-catenin pathway (49). In the current study, besides ERINA, we have identified more than 30 other lncRNAs whose expression was associated with cancer drug resistance (Fig. 1B). Future studies are necessary to determine whether and how these additional lncRNAs may play a role in cancer drug resistance.

In summary, we have discovered and established ERINA as an estrogen-responsive oncogenic lncRNA that drives breast cancer by targeting the E2F1/RB1 pathway. The overexpression of ERINA may have contributed to drug resistance and poor survival of patients with ER+ breast cancer. We propose ERINA as a novel biomarker and potential therapeutic target of breast cancer.

No potential conflicts of interest were disclosed.

Z. Fang: Conceptualization, investigation, methodology, writing-original draft, writing-review and editing. Y. Wang: Conceptualization, investigation, writing-original draft, writing-review and editing. Z. Wang: Conceptualization, investigation, methodology, writing-review and editing. M. Xu: Investigation. S. Ren: Investigation. D. Yang: Conceptualization, supervision, writing-original draft, writing-review and editing. M. Hong: Conceptualization, supervision, writing-original draft, writing-review and editing. W. Xie: Conceptualization, supervision, funding acquisition, writing-original draft, project administration, writing-review and editing.

This work was supported in part by NIH grants DK117370 and ES030429 (to W. Xie), CA222274 and the Shear Family Foundation (to D. Yang). W. Xie is supported in part by the Joseph Koslow Endowed Professorship from the University of Pittsburgh School of Pharmacy. Z. Fang is a visiting student from South China Agricultural University supported by a Visiting Student Scholarship from the Government of China's China Scholarship Council (File no. 201708440280).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Tao
Z
,
Shi
A
,
Lu
C
,
Song
T
,
Zhang
Z
,
Zhao
J
. 
Breast cancer: epidemiology and etiology
.
Cell Biochem Biophys
2015
;
72
:
333
8
.
2.
Salamanna
F
,
Borsari
V
,
Contartese
D
,
Costa
V
,
Giavaresi
G
,
Fini
M
. 
What is the role of interleukins in breast cancer bone metastases? A systematic review of preclinical and clinical evidence
.
Cancers
2019
;
11
:
2018
.
3.
Mills
CC
,
Kolb
EA
,
Sampson
VB
. 
Development of chemotherapy with cell-cycle inhibitors for adult and pediatric cancer therapy
.
Cancer Res
2018
;
78
:
320
5
.
4.
Nadji
M
,
Gomez-Fernandez
C
,
Ganjei-Azar
P
,
Morales
AR
. 
Immunohistochemistry of estrogen and progesterone receptors reconsidered: experience with 5,993 breast cancers
.
Am J Clin Pathol
2005
;
123
:
21
7
.
5.
Lichtenfels
M
,
Dornelles
ADS
,
Petry
FDS
,
Blank
M
,
de Farias
CB
,
Roesler
R
, et al
The anticancer estrogen receptor antagonist tamoxifen impairs consolidation of inhibitory avoidance memory through estrogen receptor alpha
.
J Neural Transm
2017
;
124
:
1331
9
.
6.
Sanchez-Martinez
C
,
Lallena
MJ
,
Sanfeliciano
SG
,
de Dios
A
. 
Cyclin dependent kinase (CDK) inhibitors as anticancer drugs: recent advances (2015–2019)
.
Bioorg Med Chem Lett
2019
;
29
:
126637
.
7.
Roskoski
R
 Jr
. 
Cyclin-dependent protein serine/threonine kinase inhibitors as anticancer drugs
.
Pharmacol Res
2019
;
139
:
471
88
.
8.
Chun
KH
,
Park
JH
,
Fan
S
. 
Predicting and overcoming chemotherapeutic resistance in breast cancer
.
Adv Exp Med Biol
2017
;
1026
:
59
104
.
9.
Gonzalez-Angulo
AM
,
Morales-Vasquez
F
,
Hortobagyi
GN
. 
Overview of resistance to systemic therapy in patients with breast cancer
.
Adv Exp Med Biol
2007
;
608
:
1
22
.
10.
Pan
JJ
,
Xie
XJ
,
Li
X
,
Chen
W
. 
Long non-coding rnas and drug resistance
.
Asian Pac J Cancer Prev
2015
;
16
:
8067
73
.
11.
Verma
A
,
Jiang
Y
,
Du
W
,
Fairchild
L
,
Melnick
A
,
Elemento
O
. 
Transcriptome sequencing reveals thousands of novel long non-coding RNAs in B cell lymphoma
.
Genome Med
2015
;
7
:
110
.
12.
Iyer
MK
,
Niknafs
YS
,
Malik
R
,
Singhal
U
,
Sahu
A
,
Hosono
Y
, et al
The landscape of long noncoding RNAs in the human transcriptome
.
Nat Genet
2015
;
47
:
199
208
.
13.
Fang
Y
,
Fullwood
MJ
. 
Roles, functions, and mechanisms of long non-coding rnas in cancer
.
Genomics Proteomics Bioinformatics
2016
;
14
:
42
54
.
14.
Ulitsky
I
,
Bartel
DP
. 
lincRNAs: genomics, evolution, and mechanisms
.
Cell
2013
;
154
:
26
46
.
15.
Arun
G
,
Diermeier
SD
,
Spector
DL
. 
Therapeutic targeting of long non-coding rnas in cancer
.
Trends Mol Med
2018
;
24
:
257
77
.
16.
Petryszak
R
,
Keays
M
,
Tang
YA
,
Fonseca
NA
,
Barrera
E
,
Burdett
T
, et al
Expression atlas update–an integrated database of gene and protein expression in humans, animals and plants
.
Nucleic Acids Res
2016
;
44
:
D746
52
.
17.
Iorio
F
,
Knijnenburg
TA
,
Vis
DJ
,
Bignell
GR
,
Menden
MP
,
Schubert
M
, et al
A landscape of pharmacogenomic interactions in cancer
.
Cell
2016
;
166
:
740
54
.
18.
Colaprico
A
,
Silva
TC
,
Olsen
C
,
Garofano
L
,
Cava
C
,
Garolini
D
, et al
TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data
.
Nucleic Acids Res
2016
;
44
:
e71
.
19.
Wang
Y
,
Wang
Z
,
Xu
J
,
Li
J
,
Li
S
,
Zhang
M
, et al
Systematic identification of non-coding pharmacogenomic landscape in cancer
.
Nat Commun
2018
;
9
:
3192
.
20.
Benskey
MJ
,
Manfredsson
FP
. 
Lentivirus production and purification
.
Methods Mol Biol
2016
;
1382
:
107
14
.
21.
Kumar
P
,
Nagarajan
A
,
Uchil
PD
. 
Analysis of cell viability by the MTT assay
.
Cold Spring Harb Protoc
2018
.
DOI: 10.1101/pdb.prot095505
.
22.
Iglesias-Ara
A
,
Osinalde
N
,
Zubiaga
AM
. 
Detection of e2f-induced transcriptional activity using a dual luciferase reporter assay
.
Methods Mol Biol
2018
;
1726
:
153
66
.
23.
Wang
Z
,
Yang
B
,
Zhang
M
,
Guo
W
,
Wu
Z
,
Wang
Y
, et al
LncRNA epigenetic landscape analysis identifies epic1 as an oncogenic lncrna that interacts with myc and promotes cell-cycle progression in cancer
.
Cancer Cell
2018
;
33
:
706
20
.
24.
Niu
Y
,
Xu
M
,
Slagle
BL
,
Huang
H
,
Li
S
,
Guo
GL
, et al
Farnesoid X receptor ablation sensitizes mice to hepatitis b virus X protein-induced hepatocarcinogenesis
.
Hepatology
2017
;
65
:
893
906
.
25.
Hua
JT
,
Ahmed
M
,
Guo
H
,
Zhang
Y
,
Chen
S
,
Soares
F
, et al
Risk SNP-mediated promoter-enhancer switching drives prostate cancer through lncrNA PCAT19
.
Cell
2018
;
174
:
564
75
.
26.
Tsai
MC
,
Manor
O
,
Wan
Y
,
Mosammaparast
N
,
Wang
JK
,
Lan
F
, et al
Long noncoding RNA as modular scaffold of histone modification complexes
.
Science
2010
;
329
:
689
93
.
27.
Nelson
JD
,
Denisenko
O
,
Bomsztyk
K
. 
Protocol for the fast chromatin immunoprecipitation (ChIP) method
.
Nat Protoc
2006
;
1
:
179
85
.
28.
Karimzadeh
M
,
Ernst
C
,
Kundaje
A
,
Hoffman
MM
. 
Umap and Bismap: quantifying genome and methylome mappability
.
Nucleic Acids Res
2018
;
46
:
e120
.
29.
Xue
W
,
Li
L
,
Tian
X
,
Fan
Z
,
Yue
Y
,
Zhang
C
, et al
Integrated analysis profiles of long non-coding RNAs reveal potential biomarkers of drug resistance in lung cancer
.
Oncotarget
2017
;
8
:
62868
79
.
30.
Tani
H
,
Torimura
M
. 
Identification of short-lived long non-coding RNAs as surrogate indicators for chemical stress response
.
Biochem Biophys Res Commun
2013
;
439
:
547
51
.
31.
Jiang
X
,
Guo
S
,
Zhang
Y
,
Zhao
Y
,
Li
X
,
Jia
Y
, et al
LncRNA NEAT1 promotes docetaxel resistance in prostate cancer by regulating ACSL4 via sponging miR-34a-5p and miR-204-5p
.
Cell Signal
2020
;
65
:
109422
.
DOI: 10.1016/j.cellsig.2019.109422
.
32.
Chen
S
,
Xia
X
. 
Long noncoding RNA NEAT1 suppresses sorafenib sensitivity of hepatocellular carcinoma cells via regulating miR-335-c-Met
.
J Cell Physiol
2019
;
234
:
14999
15009
.
33.
Renganathan
A
,
Felley-Bosco
E.
. 
Long noncoding RNAs in cancer and therapeutic potential
.
Adv Exp Med Biol
2017
;
1008
:
199
222
.
34.
Sheldon
LA
. 
Inhibition of E2F1 activity and cell cycle progression by arsenic via retinoblastoma protein
.
Cell Cycle
2017
;
16
:
2058
72
.
35.
Chu
C
,
Zhang
QC
,
da Rocha
ST
,
Flynn
RA
,
Bharadwaj
M
,
Calabrese
JM
, et al
Systematic discovery of Xist RNA binding proteins
.
Cell
2015
;
161
:
404
16
.
36.
Ferre
F
,
Colantoni
A
,
Helmer-Citterich
M
. 
Revealing protein-lncRNA interaction
.
Brief Bioinform
2016
;
17
:
106
16
.
37.
Mohamadkhani
A
. 
Long noncoding rnas in interaction with RNA binding proteins in hepatocellular carcinoma
.
Hepat Mon
2014
;
14
:
e18794
.
38.
Mathews
DH
. 
Revolutions in RNA secondary structure prediction
.
J Mol Biol
2006
;
359
:
526
32
.
39.
Wang
C
,
Rauscher
FJ
 3rd
,
Cress
WD
,
Chen
J
. 
Regulation of E2F1 function by the nuclear corepressor KAP1
.
J Biol Chem
2007
;
282
:
29902
9
.
40.
Liban
TJ
,
Thwaites
MJ
,
Dick
FA
,
Rubin
SM
. 
Structural conservation and E2F binding specificity within the retinoblastoma pocket protein family
.
J Mol Biol
2016
;
428
:
3960
71
.
41.
Nevins
JR
,
Leone
G
,
DeGregori
J
,
Jakoi
L
. 
Role of the Rb/E2F pathway in cell growth control
.
J Cell Physiol
1997
;
173
:
233
6
.
42.
Creyghton
MP
,
Cheng
AW
,
Welstead
GG
,
Kooistra
T
,
Carey
BW
,
Steine
EJ
, et al
Histone H3K27ac separates active from poised enhancers and predicts developmental state
.
Proc Natl Acad Sci U S A
2010
;
107
:
21931
6
.
43.
Johnson
J
,
Thijssen
B
,
McDermott
U
,
Garnett
M
,
Wessels
LF
,
Bernards
R
. 
Targeting the RB-E2F pathway in breast cancer
.
Oncogene
2016
;
35
:
4829
35
.
44.
Dykes
IM
,
Emanueli
C
. 
Transcriptional and post-transcriptional gene regulation by long non-coding RNA
.
Genomics Proteomics Bioinformatics
2017
;
15
:
177
86
.
45.
Engelmann
D
,
Putzer
BM
. 
The dark side of E2F1: in transit beyond apoptosis
.
Cancer Res
2012
;
72
:
571
5
.
46.
Yan
LH
,
Wei
WY
,
Cao
WL
,
Zhang
XS
,
Xie
YB
,
Xiao
Q
. 
Overexpression of E2F1 in human gastric carcinoma is involved in anti-cancer drug resistance
.
BMC Cancer
2014
;
14
:
904
.
47.
Li
Y
,
Ye
Y
,
Feng
B
,
Qi
Y
. 
Long noncoding RNA lncARSR promotes doxorubicin resistance in hepatocellular carcinoma via modulating pten-pi3k/akt pathway
.
J Cell Biochem
2017
;
118
:
4498
507
.
48.
Zhang
Y
,
Li
Q
,
Yu
S
,
Zhu
C
,
Zhang
Z
,
Cao
H
, et al
Long non-coding RNA FAM84B-AS promotes resistance of gastric cancer to platinum drugs through inhibition of FAM84B expression
.
Biochem Biophys Res Commun
2019
;
509
:
753
62
.
49.
Jia
L
,
Tian
Y
,
Chen
Y
,
Zhang
G
. 
The silencing of LncRNA-H19 decreases chemoresistance of human glioma cells to temozolomide by suppressing epithelial-mesenchymal transition via the Wnt/beta-Catenin pathway
.
OncoTargets Ther
2018
;
11
:
313
21
.