Background
Although the overall survival and prognosis of breast cancer patients have been improved in recent years [
32], metastasis is still the leading cause of mortality in breast cancer patients. Patients with metastases have a 5-year survival rate of only 26%, compared to that of 90% in overall breast cancer patients [
36]. Thus, it is important to identify underlying molecular mechanisms of breast cancer metastasis and develop new therapeutic targets. Cancer cells, which displayed high metastatic potential, invade locally and migrate to the secondary organ sites to form metastatic niches. Regional lymph nodes (LNs) are usually the primary sites of early metastasis and axillary lymph node metastasis is one of the most important prognostic factors for breast cancer patients [
30]. Thus, novel metastasis-promoting genes in breast cancer may be identified through combing the genetic changes between LN metastasis tissues and respective primary tumors.
In human genome, there is only up to 2% of protein-coding genes that are stably transcribed, whereas the vast majority are non-coding RNAs (ncRNAs) [
35]. Emerging evidence has shown that ncRNAs, in particular long non-coding RNAs (lncRNAs) and microRNAs (miRNAs), play crucial roles in various types of cancer through regulating coding gene expression and epigenetic signatures [
16,
27]. lncRNAs possess diverse biological functions, such as regulation of protein and RNA stability, transcriptional modulation, nuclear scaffold formation, or guiding protein-DNA interaction [
38]. Although previous studies have linked several lncRNAs to cancer metastasis [
38], regulatory lncRNAs that initiate and promote the transition from primary cancer to metastatic tumors remain elusive.
Here, we identified that lncRNA LINC02273 was significantly upregulated in metastatic LNs compared with primary tumors and could predict poor recurrence free survival in breast cancer patients. Knockout of LINC02273 could curb breast cancer metastasis in vitro and in vivo. Mechanistically, RNA-binding protein hnRNPL, which was also increased in metastatic tissues, bound with CA-repeats in the 3′ region of LINC02273 through its RRM1&2 motif and increased the stability of LINC02273. We further revealed that the hnRNPL-LINC02273 RNP complex could activate oncogene AGR2 transcription through enhancing H3K4me3 and H3K27ac levels around its promoter region. We also verified the potential clinical use of ASOs targeting LINC02273 to inhibit breast cancer metastasis. Altogether, our results demonstrated that LINC02273 played a pivotal role in enhancing breast cancer metastatic potential and could serve as a novel therapeutic target for mitigating breast cancer metastasis.
Methods
Tissue samples
Metastatic lymph node samples and paired primary tumors were collected at Fudan University Shanghai Cancer Center. All lymph node metastatic loci and primary tumors were confirmed by H&E staining. Lymph node metastatic loci and primary tumor sites were dissected and subjected to RNA extraction. A series of fresh frozen tissue and clinical data from breast cancer patients was collected at Fudan University Shanghai Cancer Center. Tumor tissue was obtained from patients undergoing surgery and immediately stored at − 80 °C. All the protocols were reviewed and approved by an independent institutional review board at Fudan University, and all patients gave their written informed consent before inclusion in this study. Post-operative chemotherapy and radiation therapy were suggested according to breast cancer guidelines of Fudan University Shanghai Cancer Center.
bc-GenExMiner 3.0 database was used to explore the predictive role of LINC02273 for metastatic events in breast cancer [
14]. Five GEO datasets were pooled (GSE6532, GSE9195, GSE19615, GSE17907, E_MTAB_365). Three hundred fifty-one patients were included, and 81 metastatic events were observed.
Cell lines, transfection and treatments
The HEK293T, MDA-MB-231 and BT549 cell lines were obtained from American Type Culture Collection (ATCC) (Manassas, VA). LM2 was a subline of MDA-MB-231 with high lung metastasis tendency was described previously [
33]. All cell lines were genotyped (Genewiz) and tested routinely for Mycoplasma contamination (Vazyme). All cells are cultured in DMEM containing 10% FBS (Life Technologies), 100 U/ml penicillin and 100 μg/ml streptomycin (Invitrogen) at 37 °C in a humidified incubator with 5% CO2. siRNA and 2-O-methyl RNA/DNA antisense oligonucleotides (ASOs), which were modified by changing the five nucleotides at the 5′ and 3′ ends into 2′-O-methyl were synthesized (RiboBio). siRNA and ASOs transfection were performed using Lipofectamine RNAiMAX (Life Technologies), and plasmid were transfected with Lipofectamine 3000 (Life Technologies). In vivo use GapmeR ASOs with same sequences were chemically modified by Ribobio. The modification of GapmeR were described previously [
41]. GapmeR ASOs contains a central block of DNA with a wing region of artificial nucleotides exhibiting high affinity for its target RNA. GapmeR ASOs were dissolved and diluted in PBS to obtain the final concentration of 100 mM. For ASOs free uptake assay, 0.5 × 10
6 cells were seeded in 6-well plate. After 24 h, GapmeR ASOs were added directly to cell cultures. RNA was extracted 48 h after transfection. For actinomycin D RNA stability assay, 1 × 10
6 cells were seeded in 6-well plate. After 24 h, 5 μg/ml actinomycin D (Apexbio) were added. Cells were treated at the time point as indicated. Sequences of siRNAs and ASOs were listed in Additional file
1: Table S13.
RNA isolation, RT-PCR and RT-qPCR
Total RNAs from cultured cells were extracted with TRIzol (Life technologies) according to the manufacturer’s protocol. cDNAs were reverse transcribed with Hiscript III Reverse Transcriptase (Vazyme) with oligo (dT) and random hexamers followed by qRT-PCR analysis and applied for PCR/qPCR analysis. Real time quantitative PCR was performed with ChamQ SYBR qPCR Master Mix (Vazyme) and 7900HT Fast Real-Time PCR System or (Applied Biosystems). The relative expression of different sets of genes was quantified to GAPDH mRNA. Primer sequences for RT-qPCR and RT-PCR used were listed in Additional file
1: Table S12.
Immunoblotting analysis
Cells (1 × 106) were counted and plated in 6-well plate. After 24 to 48 h, cells were rinsed with PBS. Then, 80 μl 1 × SDS Protein Loading Buffer containing protease and phosphatase inhibitors (Bimake) were added to each well of the plate. Cell lysates were scraped down and sonicated with Bioruptor Plus for 3 × 30s with low power. After denaturation at 100 °C for 10 min, the samples were resolved by SDS-PAGE, transferred to PVDF membranes (Millipore). The membranes were blocked with 5% (w/v) skimmed milk in TBS-T for 1 h, then probing with antibodies at 4 °C overnight. Antibodies used in immunoblot includes hnRNPL (sc-32,317, Santa Cruz Biotechnology); AGR2 (66768–1-Ig, Proteintech); GAPDH (HRP-60004). HRP-conjugated secondary antibodies were used for detection.
Cell proliferation, wound healing and Transwell assays
For proliferation assay, cells were seeded in 96-well flat-bottomed plates with each well containing 3000 cells in 200 μl of culture medium and cultured in ambient environment described above. Plates were imaged by using IncuCyte ZOOM System (Essen Bioscience) at 12-h interval. The growth rate was measured according to confluence change analyzed by IncuCyte software. For wound healing assay, 30,000 cells were seeded in 96-well IncuCyte® ImageLock Plates (Essen Bioscience). After 24 h, a wound was made in each well with WoundMaker™. Plates were imaged as described above with 6-h interval. Wound width was measured and analyzed. For transwell migration assay, a total of 1 × 105 of MDA-MB-231, 5 × 104 BT549, or 1 × 105 LM2 cells were suspended in 200 μl of DMEM without FBS and seeded on the top chamber of 24-well plate-sized Transwell inserts (Corning Falcon). The lower chambers contained DMEM with 20% FBS. After incubation for 8 h, the inserts were fixed and stained with crystal violet. Cells in the upper chamber were removed with cotton swabs. The average confluence of migrated cells was analyzed by ImageJ according to three random fields captured by 100× microscope. Each experiment was conducted in triplicate. Matrigel invasion assays were performed using Matrigel-coated Transwell inserts with the procedure as described above.
Rapid amplification of cloned cDNA ends (RACE)
The 3′ and 5′ RACE (Rapid amplification of cloned cDNA ends) was performed using the SMARTer RACE kit (Clontech) following the manufacturer’s instruction. RNA was extracted from MDA-MB-231 cells. Primers used for 3′ and 5′ RACE were listed in Additional file
1: Table S12.
Ectopic expression and gene knockdown by shRNA
LINC02273 1653 nt full length cDNA were synthesized (Genewiz) according to 5′ and 3′ RACE result and cloned into pCDH-CMV-Puro and pcDNA3.1-myc vector. The ORF of hnRNPL, AGR2, ETS1, and STAT3 were cloned from MDA-MB-231 cDNA and inserted into pCDH-CMV-Puro with C-terminus 3 × FLAG tag. Serial mutations of hnRNPL were subcloned and constructed with C-terminus 3 × FLAG tag. The shRNA oligos were annealed and cloned into pLKO.1-Puro vector. Lentivirus was produced by co-transfecting HEK293T cells with desired plasmids together with psPAX2 and pMD2.G. After 72 h, virus was harvested by passing through a 0.45 mm filter. Collected lentivirus was used directly to infect cells with the addition of 8 mg/ml polybrene (Sigma-Aldrich) or stored in − 80 °C. Infected cells were selected with puromycin (Invivogen) at 1 μg/ml.
Generation of LINC02273 knockout cell lines by CRISPR-cas9
The gRNA oligos were annealed and cloned into the lentiGuide-Puro (Addgene, 52,963). To delete LINC02273, MDA-MB-231 and BT549 cells were transiently transfected with lentiGuide-Puro plasmid carrying two gRNAs and pLentiCas9-Blast (Addgene, 52,963) and selected with 1 μg/ml puromycin and 2 μg/ml blasticidin for 5 days. Cells were then seeded in 96 well plates in low density (1 cell per well) to form single colonies. Cell clones were genotyped by PCR and sequencing to detect the deletions. PCR primers were listed in Additional file
1: Table S12.
RNA fish
LINC02273 RNA FISH was performed using Stellaris FISH probes as listed in Additional file
1: Table S11 (Biosearch Technologies) according to the manufacturer’s protocol. In brief, MDA-MB-231 cells were grown on coverslips in a 24-well culture plate. Cells were fixed with 4% (w/v) paraformaldehyde in 1 × PBS for 10 min. Fixed cells were permeabilized with 70% ethanol for 1 h at 4 °C. The coverslips were washed one time with Buffer A (Biosearch Technologies) for 5 min at room temperature and incubated with Stellaris RNA FISH Hybridization Buffer (Biosearch Technologies) containing 10% formamide and 250 nM of the LINC02273-specific Stellaris probe labeled with CAL Fluor Red 590 in the dark at 37 °C overnight. The coverslips were then washed once with Wash Buffer A and once with Wash Buffer A containing 0.1 mg/ml DAPI, both washes in the dark at 37 °C for 30 min and mounted onto microscope slides using VECTASHIELD Antifade Mounting Medium (Vector Laboratories). Images were acquired with Leica confocal microscope.
1 × 107 cells were collected and used for nuclear and cytoplasmic protein extraction using the NE-PER Nuclear and Cytoplasmic Extraction Reagents (Life Technologies). For nuclear and cytoplasmic RNA separation, 1 × 106 cells were collected and extracted using PARIS™ kit (Life Technologies).
RIP assay
1 × 10
7 cells were harvested, resuspended in 1 mL lysis buffer followed by 3 × 10 s sonication with high power with an interval of 30 s. After centrifuging at 12,000 rpm for 15 min at 4 °C, the supernatant was pre-cleared with 30 μl Dynabeads Protein G (Invitrogen). The precleared lysates were incubated with 4 μg anti-hnRNPL antibody (Santa Cruz Biotechnology) for 3 h at 4 °C. Then 40 μl Dynabeads Protein G beads (blocked with 1% BSA and 20 mg/ml yeast tRNA for 1 h at 4 °C) were added to the mixture and incubated for another 1 h at 4 °C followed by washing with wash buffer. The RNA-protein complex was eluted with 200 μl TRIzol at room temperature for 10 min. RNA was purified with Direct-zol RNA Microprep (Zymo Research). For qRT-PCR, reverse transcription was performed with Hiscript III Reverse Transcriptase (Vazyme) with oligo (dT) and random hexamers followed by qRT-PCR analysis. Primers were listed in Additional file
1: Table S12.
Chromatin isolation by RNA purification
Each probe was designed using LGC Biosearch’s web-based designer, biotin-TEG-labeled at its 3′ terminus and divided into odd or even groups. Probes targeted LacZ were used as negative control. For each sample, 4 × 107 MDA-MB-231 cells were used and crosslinked in 1% glutaraldehyde at room temperature for 10 min, and quenched with 125 mM glycine for 5 min. The cells were lysed by Bioruptor plus in ChIRP lysis buffer and sheared to 100–500 bp fragments at 4 °C. Cleared lysates were hybridized with probes in 37 °C in a hybridization oven for 4 h. After washes and elution, RNA and DNA were purified accordingly. qPCR and RT-qPCR were performed as described above. DNA for library construction were performed as described in the library preparation section. The probes used in the ChIRP assay are listed in Additional file
1: Table S11.
RNA pull-down assay
PCR products were verified by gel electrophoresis and purified by phenol: chloroform extraction. Biotin-labelled RNAs was in vitro transcribed with HiScribe™ T7 Quick High Yield RNA Synthesis Kit (NEB) with Biotin-16-UTP (Roche), treated with RNase-free DNase I on column during RNA purification with RNA Clean & Concentrator-25 (Zymo Research). For each sample, 5 μg RNA was mixed with 1 × 107 cell extract and incubated at 4 °C for 1 h, followed by incubating with Dynabeads M-280 Streptavidin (Invitrogen) at 4 °C overnight. After washes, the pull-down complexes were eluted by denaturation in 1× protein loading buffer for 10 min at 100 °C. The samples were detected by Western blot or proceed to mass spectrometry.
Mass spectrometry after RNA pull-down
After RNA pull-down, equal amounts of samples pulled down by sense and anti-sense LINC02273 were loaded on SDS-PAGE gel. Then the gel was stained with Fast Silver Stain Kit (Beyotime) according to the manufacturer’s instructions. Specific bands were cut and analyzed by LC-MS/MS (Shanghai Applied Protein Technology, Shanghai, China). Protein identification was retrieved in the human RefSeq protein database (National Center for Biotechnology Information), using Mascot version 2.4.01 (Matrix Science, London, UK).
Cells were trypsinized, and 3 × 104 cells per well in 2.5 ml of completed MammoCult™ Medium (Stemcell) in 6-well ultralow attachment plates (Corning). After 5–10 days, mammospheres larger than 100 μm were imaged and counted under microscope.
Chromatin Immunoprecipitation (ChIP)
MDA-MB-231 cell lines (2 × 106 cells per assay) were cross-linked with 1% formaldehyde at RT for 10 min and quenched in 125 mM glycine for 5 min. The cross-linked chromatin was sonicated to generate DNA fragments averaging 200–500 bp in length by Bioruptor plus. Chromatin fragments were immunoprecipitated with antibodies against mouse or rabbit normal IgG (4 μg, Santa cruz), hnRNPL (4 μg, Santa Cruz), H3K27ac (1:200 CST), H3K4me3 (1:200 CST). The precipitated DNA were purified using the ChIP DNA Clean & Concentrator Kits (Zymo Research).
Luciferase assay
LINC02273 promoter (chr14:24160663–24,161,805, hg38), AGR2 promoter region, and serial truncations were cloned into pGL3-basic vector. Full length LINC02273 and LINC02273 without CA-repeat were cloned into pMIRGLO vector at the 3′ UTR of luciferase ORF. 1 × 105 cells were plated in each well of 24-well plates. On the next day, 250 ng reporter vector, 250 ng ectopic expression vector and 10 ng pRL-SV40 vector were transfected with Lipofectamine 3000 (Invitrogen). For LINC02273 stability assay, 500 ng reporter vector and 20 pmol siRNA were co-transfected with Lipofectamine 3000 (Invitrogen). After 48 h, cells were lysed in 100 μl of passive lysis buffer. The firefly luciferase activity and the Renilla activity were determined by Dual-Luciferase® Reporter Assay System. For each experiment, the firefly luciferase activity was normalized to Renilla activity. The results were expressed as fold induction compared to negative control indicated.
Microarray analysis
Five pairs of primary tumor and LN metastatic loci were dissected and verified by H&E staining, tumor percentage > 70%. Followed by RNA extraction with TRIzol reagent, samples were analyzed by Shanghai Qiming Biotechnology Co. Ltd. with Affymetrix HTA 2.0.
Library preparation and deep sequencing
The total RNA samples (1 μg) of MDA-MB-231 WT and LINC02273 KO cells were treated with VAHTS mRNA Capture Beads (Vazyme) to enrich polyA+ RNA before constructing the RNA-seq libraries. RNA-seq libraries were prepared using VAHTS mRNA-seq v2 Library Prep Kit for Illumina (Vazyme) following the manufacturer’s instructions. Briefly, polyA+ RNA samples were fragmented and then used for first- and second-strand cDNA synthesis with random hexamer primers. The cDNA fragments were treated with DNA End Repair Kit to repair the ends, then modified with Klenow to add an A at the 3′ end of the DNA fragments, and finally ligated to adapters. Purified dsDNA was subjected to 12 cycles of PCR amplification, and the libraries were sequenced by Illumina sequencing platform on a 150 bp paired-end run. Sequencing reads from RNA-seq data were aligned using the spliced read aligner HISAT2, which was supplied with the Ensembl human genome assembly (Genome Reference Consortium GRCh38) as the reference genome. Gene expression levels were calculated by the FPKM (fragments per kilobase of transcript per million mapped reads). Gene Set Enrichment Analysis (GSEA) pre-ranked was run on the ranked list using the Molecular Signatures Database (MSigDB) as the gene sets.
Total RNA was isolated with TRIzol reagent. Next, strand-specific RNA-seq libraries were prepared using the NEBNext Ultra Directional RNA Library Prep kit (New England Biolabs, Beverly, MA, USA), according to the manufacturer’s instructions. For the data processing, the raw sequencing reads were aligned to human reference genome (hg38) using the splice-aware aligner HISAT246. Read counts for each gene were normalized into FPKM (Fragments per Kilobase of transcript per Million mapped reads) values. The cutoff of differential gene expression was FDR < 0.05, normalized by the wild type cells. The FPKM fold change of the overlap genes of the two subgroups independent KO cells were used for further analysis.
ChIP and ChIRP DNA libraries were prepared using the NEBNext DNA Library Prep Kit for Illumina (New England Biolabs) according to the manufacturer’s instructions. The raw reads were aligned by Bowtie2 (v2.2.9) with a human reference genome (hg38). The uniquely mapped reads were subjected to the peak calling algorithm, MACS (v1.4.2) with default parameters.
In vivo assays
Four-week-old female NOD/SCID mice or female athymic BALB/c nude mice were purchased from Shanghai Slack Laboratory animal Co., LTD and housed under SPF conditions at the animal care facility of the Experimental Animal Center of Shanghai University of Traditional Chinese Medicine.
For xenograft models, 1 × 106 LM2 cells were orthotopically injected directly into the inguinal mammary fat pads of mice in 50 μl of sterile PBS (n = 8 in each group).
For ASO in vivo treatment experiment, 1 × 106 LM2 cells were orthotopically injected directly into the mammary fat pads of mice in 100 μl of sterile PBS. Two weeks after injection, mice were randomized into three groups (n = 8 in each group). ASOs were delivered according to groups, through tail vein injection, 10 nmol (100 μl, 100 mM ASO in PBS without transfection reagent) each mouse, and twice a week. Tumor were measured by caliper twice a week. The xenografts were fixed in polyformaldehyde for paraffin embedding or frozen for RNA. For tail vein injection model, 1 × 105 LM2 cells were diluted in 100 μl of sterile PBS and injected through tail vein.
To investigate tumor metastasis, 8 weeks after injection, each mouse was intraperitoneally injected 200 mg/kg D-Luciferin. Bioluminescence were analyzed by IVIS system. All experiments were performed in accordance with relevant institutional and national guidelines and regulations of Shanghai Medical Experimental Animal Care Commission.
Statistics and reproducibility
Statistical analyses were performed using SPSS v.23.0 (SPSS) or Prism GraphPad 8.0. For most of the experiments, independent sample t-tests were used to calculate the P values. For bioluminescence, Mann-Whitney tests were used to calculate the P values. Spearman correlation analysis was performed to assess the relationship between different factors. Survival curves were plotted using the Kaplan–Meier method and compared using log-rank tests. All statistical analyses were performed using two-tailed P values. The experiments, with the representative data shown in the figures, were repeated three times independently with similar results obtained. Statistical details and methods used are indicated in the figure legends, text or methods.
Discussion
Through profiling transcriptome of metastatic LNs and matched primary cancer tissues, we identified LINC02273 as a lncRNA significantly enriched in LN metastatic lesions, compared to the primary tumors. We further showed that downregulation of LINC02273 impaired the migration and invasion ability of breast cancer cells while its overexpression significantly promoted breast cancer cell mobility. Moreover, LINC02273 knock-down remarkably reduced lung metastasis in breast cancer xenograft models. In breast cancer patients, increased LINC02273 was associated with high burden of LN metastasis. Patients with high LINC02273 level had shorter RFS and worse clinical outcome than those with low LINC02273 expression. Therefore, LINC02273 could serve as an independent prognostic factor in breast cancer.
Recent studies highlighted the critical roles that RNA binding proteins (RBP) may play in modulating cancer initiation and progression through binding to RNAs and regulating their processing, stability, localization, modification or translation [
3,
10]. Our mechanistic studies revealed that LINC02273 was stabilized by interaction with hnRNPL. hnRNPL is a RNA binding protein which predominantly binds to the consensus CA repeat motif within RNA [
49]. We confirmed that hnRNPL bound to the CA repeats of LINC02273 with its RRM1&2 motif, which is essential for hnRNPL-dependent stabilization of LINC02273. Previous reports indicated that hnRNPL played critical roles in cancer progression [
8,
17,
24]. For instance, hnRNPL could promote lymphatic metastasis of bladder cancer through binding with LNMAT1 [
4]. hnRNPL was also involved in prostate cancer progression through regulating the alternative splicing of a set of RNAs, including those encoding androgen receptor, the key lineage-specific oncogene of prostate cancer [
8]. Here we demonstrated that hnRNPL-promoted breast cancer metastasis was dependent on, at least in part, stabilizing LINC02273. Increased level of hnRNPL was found in metastatic LNs compared to that in matched primary tumors. These results revealed a novel mechanism by which hnRNPL promoted breast cancer progression via its RNA-binding activity.
The molecular mechanisms by which lncRNAs exert their biological functions are diverse and complex. Accumulating evidence suggests that lncRNAs can affect neighboring intra-chromosomal genes
in cis or genes on different chromosomes
in trans. Transcriptional regulation by lncRNAs may depend on their roles in modulating signaling transduction, regulating transcription factor recruitment and histone modification, as well as serving as a scaffold for assembly of multiple regulatory molecules at single locus. Since LINC02273 mainly localized in the nucleus, RNA-seq and ChIRP-seq were carried out to identify genes regulated by LINC02273 and respective DNA loci that LINC02273 bound. A number of LINC02273-regulated gene, such as CHRNA7, ERP27, GABRG1, FOXN4, ZNF831, AGR2, EBF1 and GUCY2C, have been reported to be involved in cancer progression and metastasis. CHRNA7 promoted pancreatic and lung cancer metastasis [
34,
39], while EBF1 could inhibit colorectal cancer cell proliferation and induce cell apoptosis through the p53/p21 pathway [
40]. As a member of protein disulfide isomerase family, AGR2 is involved in maturation of proteins in endoplasmic reticulum and participates in maintaining endoplasmic reticulum homeostasis. Recently studies have revealed varying oncogenic roles of AGR2 in cancer biology. It has been reported that AGR2 could regulate epithelial-mesenchymal transition (EMT) in cancer development [
5]. Overexpression of AGR2 exhibited enhanced cancer cell adhesion to a plastic substratum and extracellular AGR2 enhanced the rate of cancer cell attachment [
25]. In breast cancer, increased AGR2 had an unfavorable impact on survival [
45]. Upregulation of AGR2 could decrease cancer cell sensitivity to fulvestrant and epirubicin [
22,
23]. Furthermore, AGR2 inhibits p53 activation after DNA damage [
12]. Although the expression of AGR2 has been reported to be regulated by ERK, ER and HIF signaling pathways as well as by its 3’UTR [
5,
13,
23,
31], the molecular mechanisms controlling AGR2 transcription remains poorly understood. In our study, we found a novel mechanism that oncogene AGR2 could be modulated by LINC02273. LINC02273 binds to AGR2 promoter along with hnRNPL which changes its epigenetic landscape. Enrichment of LINC02273-hnRNPL complex at AGR2 promoter increased local H3K4me3 and H3K27ac levels, thereby promoting AGR2 transcription. It is plausible that hnRNPL-bound LINC02273 serves as a scaffold and a guide RNA to recruit additional histone writer proteins and/or transcriptional coactivators to AGR2 promoter region, which enhances AGR2 transcription. The LINC02273-recuited histone writers responsible for increased H3K4me3 and H3K27ac levels and other potential transcriptional activators warrants further investigation.
In addition, several studies have reported the role of hnRNPL in regulating gene transcription. For instance, hnRNPL plays an important role in gene regulation in epithelial cells [
8]. Recent studies have also demonstrated that hnRNPL might be involved in transcriptional regulation through histone modifications [
21]. We found hnRNPL enhanced AGR2 transcription in breast cancer cells. Our ChIP analysis revealed that hnRNPL was enriched in the same AGR2 promoter region where LINC02273 bound. Knock-down of hnRNPL decreased the histone H3K4me3 and H3K27ac around AGR2 promoter region. Furthermore, overexpressing hnRNPL failed to rescue the decreased H3K4me3 and H3K27ac at AGR2 promoter region, and AGR2 transcription in LINC02273-knockout cells, indicating that the transactivating effect of hnRNPL on AGR2 promoter was dependent on LINC02273. In reciprocal, increased LINC02273 barely restored the decreased H3K4me3 and H3K27ac modifications in hnRNPL-depleted cells, supporting an essential role of hnRNPL-LINC02273 complex in enhancing transactivating histone modification at AGR2 promoter region. Nevertheless, we noticed that AGR2 transcription were increased by LINC02273 overexpression, although at a reduced level, in hnRNPL-depleted cells. Therefore, beside hnRNPL, LINC02273 may bind additional transcription regulators to enhance AGR2 expression, which remain to be identified.
Consistent with our in vitro findings, a positive correlation between LINC02273 and AGR2 was confirmed in both primary tumors and metastatic LNs from breast cancer patients. In addition, shorter RFS was significantly associated with high expression of both AGR2 and LINC02273, suggesting combined expression levels of AGR2 and LINC02273 may serve as promising biomarkers to determine prognosis in breast cancer.
Overall, hnRNPL mRNA expression was relative stable in breast cancer samples (RPKM mean = 138, mean fold change = 4.7, coefficient of variation = 20.8% )[
28]. Indeed, we also found hnRNPL expression displayed relative low variance in our cohort, although increased expression of hnRNPL in metastatic LNs, compared to the matched primary tumors, was observed. On the contrary, we found LINC02273 was overexpressed in breast tumor samples with high variance. In addition, knockout or knockdown of LINC02273 could significantly curb cell migration and invasion ability, which indicates its potential as a promising therapeutic target for antagonizing breast cancer metastasis. Recently, antisense oligos (ASO) and locked nucleic acid (LNA) has been used to target mRNAs in vivo and several ASOs have been investigated in the clinical trials [ClinicalTrials Identifier: NCT01839604, NCT02423590 ][
2,
11,
15,
37,
50]. We explored whether LINC02273 expression could be inhibited by ASO both in vitro and in vivo
. Our data indicated that ASO could effectively suppress LINC02273 in cells and xenograft tumors. Moreover, LINC02273-targeting ASOs substantially reduced primary tumor growth and lung metastasis in our xenograft model. The observation strongly supports that suppressing LINC02273 with specific ASOs may serve as an effective therapeutic approach to mitigate LINC02273-promoted breast cancer progression.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.