Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Silencing of ANKRD12 circRNA induces molecular and functional changes associated with invasive phenotypes

Silencing of ANKRD12 circRNA induces molecular and functional changes associated with invasive... Background: Circular RNAs (circRNAs) that form through non-canonical backsplicing events of pre-mRNA transcripts are evolutionarily conserved and abundantly expressed across species. However, the functional relevance of circRNAs remains a topic of debate. Methods: We identified one of the highly expressed circRNA (circANKRD12) in cancer cell lines and characterized it validated it by Sanger sequencing, Real-Time PCR. siRNA mediated silencing of the circular junction of circANKRD12 was followed by RNA Seq analysis of circANKRD12 silenced cells and control cells to identify the differentially regulated genes. A series of cell biology and molecular biology techniques (MTS assay, Migration analysis, 3D organotypic models, Real-Time PCR, Cell cycle analysis, Western blot analysis, and Seahorse Oxygen Consumption Rate analysis) were performed to elucidate the function, and underlying mechanisms involved in circANKRD12 silenced breast and ovarian cancer cells. Results: In this study, we identified and characterized a circular RNA derived from Exon 2 and Exon 8 of the ANKRD12 gene, termed here as circANKRD12. We show that this circRNA is abundantly expressed in breast and ovarian cancers. The circANKRD12 is RNase R resistant and predominantly localized in the cytoplasm in contrast to its source mRNA. We confirmed the expression of this circRNA across a variety of cancer cell lines and provided evidence for its functional relevance through downstream regulation of several tumor invasion genes. Silencing of circANKRD12 induces a strong phenotypic change by significantly regulating cell cycle, increasing invasion and migration and altering the metabolism in cancer cells. These results reveal the functional significance of circANKRD12 and provide evidence of a regulatory role for this circRNA in cancer progression. Conclusions: Our study demonstrates the functional relevance of circANKRD12 in various cancer cell types and, based on its expression pattern, has the potential to become a new clinical biomarker. Keywords: Circular RNA, RNA-seq, Breast cancer, siRNA, Cancer invasion, OCR, OXPHOS * Correspondence: jom2042@qatar-med.cornell.edu Thasni Karedath and Ikhlak Ahmed contributed equally to this work. Department of Genetic Medicine, Weill Cornell Medicine-Qatar, Education City, Qatar Foundation, P.O. Box No, 24144 Doha, Qatar Genomics Core, Weill Cornell Medicine-Qatar, Education City, Qatar Foundation, Doha, Qatar Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Karedath et al. BMC Cancer (2019) 19:565 Page 2 of 17 Background regulation pathways in cancer. We confirmed the expres- Exonic circular RNAs (circRNAs) are a class of RNA in bio- sion of this circRNA across a variety of cancer cell lines logical systems for which function is not well understood. and provided evidence for its functional relevance through They possess distinct properties compared to linear RNAs downstream regulation of several tumor invasion genes. and arise from direct backsplicing events that covalently link the 3′ endofanexonwiththe 5′ end of either the Methods same exon or any other further upstream exon [1]. Cell lines and treatment circRNAs were initially considered as molecular ar- Ovarian cancer cell lines PA-1 (ATCC® CRL-1572), SKO- tifacts of aberrant RNA splicing [2]. This hypothesis V3(ATCC® HTB-77™),Caov3(ATCC® HTB-75™), NIH: was challenged by the observation that circRNAs are OVCAR-3 (ATCC® HTB-161™), breast cancer cell lines detected in various cell types in an evolutionarily con- MDA-MB-231(ATCC® HTB-26™), MCF7(ATCC® HTB- served manner [3]. The copy number of circRNAs 22™),T-47D (ATCC® HTB-133™) and breast normal cell can be up to 10 times greater than that of associated line MCF 10A(ATCC® CRL-10317™), Lung cancer cell linear RNAs, suggesting that these circRNAs may lines, HCC2935(ATCC® CRL-2869™) and NCI-H226 possess biological functions [4]. Studies have shown (ATCC® CRL-5826™) and Lung Normal Fibroblast cell that some circRNAs harbor multiple binding sites for line LL 24(ATCC® CCL-151™) (all from American type microRNAs, thereby “sponging” microRNAs and serv- Cell Collection, Manassas, VA), APOCC (ovarian pri- ing as competitive inhibitors for microRNA functions mary cell line derived from ascites fluid) (pers. commu- [5]. To serve as a sponge, however, a circular RNA nication Dr. Arash Tabrizi), A27809 (93112519-1VL, must contain multiple microRNA-binding sites and be Sigma), A2780 CIS (93112519-1VL,Sigma) were used for expressed at sufficiently high levels in the cytoplasm the current study. Cells were cultured in DMEM (Life [1]. The majority of circRNAs may not fit this Technologies, NY, USA) supplemented with 10% fetal category, and their functions remain uncertain. Some bovine serum (Life Technologies, USA). Low Passage circRNAs have been shown to interact with RNA number cells were used for all the experiments. Cell binding proteins to form RNA protein complexes culture were routinely checked for mycoplasma contam- thereby regulating canonical linear splicing of the ination using MycoAlert Mycoplasma detection kit gene [6]. These findings suggest that circRNAs hold a (Lonza, Basel, Switzerland). dynamic and distinct role in gene regulation. Emerging evidence suggests that regulation of circRNAs RNA preparation and qRT–PCR is closely associated with different diseases, particularly The nuclear and cytoplasmic RNA was extracted using cancer with aberrant expression pattern [7–10]. Thus, cir- The SurePrep™ Nuclear or Cytoplasmic RNA Purifica- cRNAs represent a new class of diagnostic biomarkers tion Kit (Fisher Grand Island, NY, USA). Total RNA with potential therapeutic significance [11, 12]. The longer from whole-cell lysates were isolated by using RNAe- half-lives compared to their linear counterparts makes cir- say mini kit (Qiagen Valencia, CA USA). For RNase R cRNAs long-acting regulators of cellular behavior and treatment, 2 μg of total RNA was incubated 60 min at − 1 robust biomarkers [13, 14]. A growing body of evidence 40 °C with or without 3 U μg of RNase R (Epicentre has implicated the functional involvement of circRNAs in Technologies, Madison, WI), and the resulting RNA was regulating cancer progression and proliferation [14–17]. subsequently purified using an RNeasy MinElute cleaning In the present study, we investigated the functional Kit (Qiagen, Valencia, CA USA). cDNA synthesis was role of a high abundance circRNA from Ankyrin Repeat carried out using First strand synthesis kit (AMV) from Domain 12 (ANKRD12) gene in cancer progression. Roche Biosciences or Biorad select cDNA synthesis kit ANKRD12 is a paralog of ANKRD11, a putative tumor using random primer for circRNA experiments. Fast Start suppressor gene with multiple functions including as a Universal SYBR Green Master mix (Roche, Clovis, CA) p53 co-activator [18, 19]. Low ANKRD12 level is an was used to amplify the specific gene using cDNA primes independent prognostic predictor of colorectal carcin- obtained from Primer bank (https://pga.mgh.harvard.edu/ oma patients [20]. Circular isoforms of ANKRD12 have primerbank/ (Additional file 4: Table S1). Each Real-Time been identified in cancer cells and patient samples in assay was done in triplicate on Step One Plus Real-time many recent studies including ours [21, 22]. In this PCR machine (Life Technologies, CA, USA). study, we validated one of the most predominant circu- lar isoforms of ANKRD12 gene that includes the back- Transfection splice junction of exons 8 and 2 (circANKRD12). We siRNA transfection was carried out using custom- report that circANKRD12 regulates the invasion, migra- designed siRNAs for both ANKRD12 circular and linear tion proliferation and cellular bioenergetics of cancer cells transcripts (Fig. 1 and Additional file 4: Table S1). The by modulating cell signaling, metabolic and cell cycle SKOV3, MDA-MB-231, OVCAR3, NCI-H226 cells were Karedath et al. BMC Cancer (2019) 19:565 Page 3 of 17 bc Fig. 1 siRNA mediated silencing of circANKRD12 in cancer cells a Two circANKRD12 siRNAs spanning the back-splice junction b qRT-PCR analysis for knockdown efficiency of circANKRD12 siRNA in 4 different cell lines. c qRT-PCR analysis for knockdown efficiency of two different circANKRD12 siRNA constructs in SKOV3 cells. d qRT-PCR analysis for silencing efficiency of ANKRD12 linear siRNA (exon9) in SKOV3 cells. (Data in b–d are the means with error bars indicating standard error of the mean (SEM) of three experiments. **P < 0.01 (Student’s t-test) grown in 6 well plates for transfection. The cells were Cell proliferation assay transfected at 24 h with 30 pmol concentration of siRNA Cells were cultured at a density of 5 × 10 cells per (VWR, Radnor, PA, USA) or scrambled control (Mission well in flat-bottomed 96-well plates in DMEM sup- siRNA universal negative control, Sigma, St.Louis, USA) plemented with 10% FBS with antimycotic anti- using Lipofectamine RNAi max (Invitrogen MA USA) biotic. The experiment was done at different time according to manufacturer’s protocol. These experiments points (24 h and 48 h or based on cell doubling were conducted in three different biological triplicates for time). CellTiter 96® Aq One Solution Reagent ueous subsequent RNA-sequencing. (Promega,Madison,WI) wasadded and theexperi- ment was conducted according to the manufac- RNAseq analysis turer’sinstructions. SKOV3, OVCAR3, NCI-H226, MDA-MB-231 cells CellTiter-Glo assay (Promega Madison USA) for transfected with either circANKRD12 or a universal determining cell proliferation was conducted according scrambled control were used for RNAseq analysis. RNA- to the manufacturer’s instructions. Briefly, CellTiter- seq library preparation and In silico detection of cir- Glo reagent was added directly to the wells of 96 well cRNA candidates from paired-end RNA-seq data was plate and luminescence was measured on an Envision conducted as described earlier [21]. reader (PerkinElmer). Karedath et al. BMC Cancer (2019) 19:565 Page 4 of 17 Scratch assay- cell migration assay Cell cycle analysis Scratch assay was conducted as described earlier [23]. Cell cycle analysis was done on cells fixed with Cells were plated into a 6-well plate with complete FxCycle™ PI/RNase Staining Solution using BD LSRFor- medium and grown to 80% confluence. Cells were trans- tessa™ cell analyzer. fected with respective siRNA in OPTIMEM medium, and the medium was replaced with serum-free DMEM Western blot analysis after transfection. After cells were grown to 100% Cellular protein was extracted after 48 h of transfection. confluence, a wound was created by scraping the conflu- The cells were lysed in 100ul of RIPA buffer with prote- ent monolayer cells with a p200 pipette tip. Cells were ase inhibitor cocktail. Then 40 micrograms of protein then grown either in serum-free medium or medium were resolved in SDS PAGE gel and transferred to a containing 3 mM Thymidine. The distance between the nitrocellulose membrane. The primary antibodies used two sides of the cell-free area was photographed using were anti-Cyclin D1, Anti- Cyclin B1, Anti CyclinD2, 10x objective in AXION Zeiss epifluorescence micro- anti-Cyclin Phospho B1 and β-actin (Cell Signaling, scope. The distance is measured using Zeiss Zen USA). The blots were visualized by ECL detection software (Carl Zeiss Carpenteria, CA, USA). (Amersham, N J, United States). The Western blot experiments and analysis were done as described earlier [25]. Briefly, cell lysate protein (25 μg) was separated on Trans-well migration and invasion assay an SDS-Polyacrylamide gel and transferred onto nitro- Cellular migration and invasion were determined cellulose membranes. Membranes were then blocked using a Transwell Boyden chamber assay as described with 5% (w/v) non-fat dry milk tris-buffered saline (TBS) previously [24]. containing 0.1% (v/v) Tween20 and incubated at 4 °C with the primary antibody (1:1000 dilution). The excess 3D organotypic spheroid model experiments primary antibody was washed off in TBS-T wash buffer, 3D anchorage-independent spheroids were developed and the membranes were incubated with HRP-linked in SKOV3 cancer cell lines; initially, cells were seeded secondary antibodies (1:2500) for one h at room on ultra-low attachment plate (Corning, NY, USA) for temperature. The excess secondary antibody was then three days. Transfection of the spheroids with cir- washed off in TBS-T, and the protein levels were cANKRD12 siRNA was conducted using the reverse detected using enhanced chemiluminescence reagent transfection method. (Sigma-Aldrich, Inc., MO, USA) and imaged on a Geliance P600 gel documentation system (PerkinElmer, Inc. MA, USA). β-Actin was used as the loading control. Spheroid area measurement Cells were seeded at a density of 300,000 cells/well into circANKRD12 and ANKRD12 siRNA longevity assay ultralow attachment plates. After 72 h, cells were trans- siRNA longevity was calculated based on cell doubling fected either with scrambled siRNA or circANKRD12 time. The cell doubling time of SKOV3 cells was calcu- siRNA. After 48 h the diameters of at least 50 spheroids lated based on the mathematical equation. were measured and the spheroid area was measured Zeiss Zen Software. Doubling Time ¼ duration  logðÞ 2 = logðÞ Final Concentration Cell proliferation assay in 3D organotypic models Ten thousand cells were seeded on each well of ultra- − logðÞ Initial Concentration : low attachment 96 well plate with OPTIMEM medium. After three days, once spheroids were formed, cells Cells were seeded on 60 mm petri plates and transfec- were transfected with either circANKRD12 siRNA or tion was done with siRNA of circANKRD12, or scrambled control. After 24 h and 48 h of transfection, ANKRD12 mRNA. Cells were harvested on an interval of MTS reagent was added to the medium. Measurements 48 h (doubling Time). The experiment was extended till were according to the manufacturer’sinstructions. 10th doubling time (20 days). The silencing efficacy was measured by gene expression analysis (qRT-PCR). Collagen invasion assay in 3D organotypic models 3D organotypic models of either circANKRD12 trans- Assessment of mitochondrial function by seahorse fected or scrambled control in SKOV3 cells expressing extracellular flux analyzer GFP were placed on the top of jellified collagen matrix The mitochondrial oxygen consumption rate (OCR) in (Rat tail collagen1, 1 mg/ml). The invasiveness analyzed the MDA-MB-231 and SKOV3 cells was assessed by using after 24 h to 10 days. a Seahorse Bioscience XFe96 analyzer (Massachusetts, Karedath et al. BMC Cancer (2019) 19:565 Page 5 of 17 USA). Seahorse Bioscience XF Cell Mito Stress Test band of expected product size in PCR assay indicating assay kit was used for the study. In this assay, subse- the presence of the circular junction. The backsplice quent additions of the ATP synthase inhibitor oligo- junctional sequence was confirmed by Sanger sequen- mycin, the mitochondrial uncoupler carbonyl cyanide cing. The divergent primers, with respect to the genomic 4-(trifluoromethoxy) phenylhydrazone (FCCP) and the sequence, only amplified when cDNA was used as a complex I + II inhibitors rotenone + antimycin A were template. The same primers did not produce a product injected into the assay medium containing cells with from genomic DNA (gDNA). Conversely, PCR using different treatments as per manufactures protocol. 10, convergent primers with respect to genomic sequence 000 cells were seeded on assay plate and the assay could amplify both cDNA and gDNA templates from was completed in 48 h after the transfection. ANKRD12 gene (Fig. 2b). This strongly indicates that a head-to-tail backsplicing junction in circANKRD12 only Statistical analysis exists in the RNA form. Statistical analysis in all the experiments is based on at To examine whether circANKRD12 is resistant to least three biological replicates and the error bars are exonucleases, we treated the total RNA extracted from drawn with the standard error of means (SEM). The p the SKOV3 cell line with RNase R. This exoribonuclease value is calculated by using 1-tailed student’s T test. enzyme digests all linear RNA forms with a 3′ single- stranded region of greater than 7 nucleotides [26]. As Data access circRNAs are devoid of any 3′ single strand overhangs, The RNA-seq data for cell-lines has been submitted to they are expected to show resistance to digestion by RNase Sequence Read Archive (NCBI SRA) under Bio project R. Indeed, circANKRD12 was resistant to RNase R diges- accession number PRJNA526399. tion compared to linear forms of ANKRD12, HIF1 alpha, SRA records will be accessible with the following link: and GAPDH. Resistance to digestion with RNase R https://www.ncbi.nlm.nih.gov/sra/PRJNA526399 exonuclease confirmed that the circANKRD12 is a stable circularized transcript (Fig. 2c). Further, cDNA created by Results priming with oligo (dT) primers failed to produce any Characterization of circANKRD12 in cancer cells amplification products for circANKRD12 in the PCR assay We characterized the circular RNA expression using confirming the circular form does not have a typical polyA RNA sequencing (RNA-seq) of total RNA in ovarian tail as do linear mRNAs. On the other hand, priming with cancer cell lines (SKOV3 and CAOV3) undergoing random hexamers – which can also amplify non-poly- Epithelial to Mesenchymal Transition. Using our cir- adenylated RNAs – resulted in distinct PCR bands for the cRNA detection pipeline [21], we identified a total of ~ backsplice junction (Fig. 2d). The PCR and qRT-PCR 18,700 circRNAs in these cell lines. The circRNA from analysis of nuclear and cytoplasmic fractions of RNA dem- the ANKRD12 gene having a backsplice junction onstrated that ANKRD12 circRNA is predominantly local- between exons 8 and 2 (circANKRD12) was one of the ized in the cytoplasm (Additional file 1: Figure S1:Fig. 2a, highly abundant circular RNAs in both SKOV3 and b). The purity of cytoplasmic and nuclear fractions was CAOV3 cell lines. The circANKRD12 originates from confirmed by amplifying the fractions using nuclear and 18p11.22 chromosome locus with the backsplice junc- cytoplasmic specific markers (Additional file 1:FigureS1: tion forming between exons located ~ 39.3 kb apart (Fig. Fig a, b, Additional file 3:S2:8). 2a). By carefully designing two sets of divergent primer We performed Real-time PCR analysis on 12 different pairs in Exon2 and Exon8 and Sanger sequencing of the cell lines from breast, ovarian and lung cancer and amplified products, we were able to identify two circular normal breast and lung to assess the cell-type specific isoforms of 286 bp and 925 bp lengths sharing the same expression of circANKRD12 (Additional file 1: Figure backsplice junction (Additional file 3: S2:1–3), Add- S1:Fig. 2c). Most of these cell lines show a high abun- itional file 4: Table S2). The first isoform of 286 bp is dance of both ANKRD12 circRNA and mRNA. Ovarian comprised of only two exons (Exon2 and Exon8), while cancer cell lines show a higher abundance of ANKRD12 the 925 bp isoform is comprised of 6 exons (Exons 2,3,5, circRNA compared to breast and lung cancer cell lines. 6,7 and 8). All subsequent functional studies do not dis- tinguish between these two circular isoforms as they tar- siRNA-mediated knockdown of circANKRD12 is highly geted the junction sequence common to both circular specific forms but distinct vis-à-vis linear forms. To investigate the role of circANKRD12, we designed Multiple validation experiments were used to confirm siRNAs to target the backsplice junction (Fig. 1a) and circANKRD12 expression (Fig. 2b-d). Divergent primers transfected multiple cancer cell lines to induce siRNA- were designed to amplify the backsplice exon junction. mediated knockdown of the circle while leaving the lin- As expected, each primer pair produced a single distinct ear RNA unaffected. The circRNA specific siRNA was Karedath et al. BMC Cancer (2019) 19:565 Page 6 of 17 cd Fig. 2 Characterization of circANKRD12 in human cancer cells. a Transcript structure of the ANKRD12 gene. Two predominant circular isoforms (286 bp and 925 bp) were detected sharing the same backsplice junction between Exon8 and Exon2 as represented by the dashed line. The backsplice junction was sequenced for validation. b Divergent primers designed to detect the circANKRD12 backsplice junction amplified cDNA but not genomic DNA (4 pairs of different primers were used to amplify the circular junction). Linear ANKRD12 amplicon was detected in both cDNA and gDNA. M represents 100 bp ladder. c qRT-PCR for relative abundance of circANKRD12 and linear ANKRD12, GAPDH and HIF1 alpha in SKOV3 cells treated with RNase R d Backsplice junction was detected in cDNA synthesized using random hexamer but not oligoDT primed RNA. The data in c is means with error bars representing standard error of the mean (SEM) from three experiments; **P < 0.01 (Student’s t-test) designed against the backsplice junction spanning exons in four cell-line transfections (Fig. 1b). Using two 2 and 8 of the gene. We observed high knockdown effi- siRNA constructs against the circANKRD12, we ciency of greater than 90% of the circular junction when confirmed that knockdown of the circular RNA is using the siRNA versus the control (scrambled siRNA) specific and has no significant effect on the linear Karedath et al. BMC Cancer (2019) 19:565 Page 7 of 17 mRNA expression (Fig. 1c). The two different siRNAs estrogen mediated S phase entry through downregulat- had similar results reducing the likelihood that the ing cyclin D1 and cyclin A. Activation of IL8 signaling observed effects are due to off-target knockdown. We and inflamasome pathways, closely associated with im- also designed siRNAs targeting exon9 and exon7 of mune modulation is also predicted. IL8 signaling path- ANKRD12 gene to knockdown the linear mRNA. In- way upregulates IL8, IRAK, ICAM-1, COX-2, and VEGF deed, the siRNA designed against exon 9, which lies genes there by upregulating angiogenesis, inflammation, outside the circANKRD12 locus was successful at and inhibits cell proliferation by downregulating cyclin knocking down the mRNA and exhibits no effect on D1. A heatmap of differentially expressed genes from circANKRD12 levels (Fig. 1d).However,the siRNA RNAseq analysis shows consistent gene expression designed on Exon7 which is shared between mRNA changes in circANKRD12 silenced cells in cancer cell and one of the circANKRD12 isoforms shows a re- lines (Fig. 4a). The differentially regulated genes cyclin markable reduction in mRNA levels as well a minimal D1 (CCND1), CBX5, STAU1, and AK4 were found to be but significant reduction in circANKRD12 levels. We consistently downregulated in the circANKRD12 knock- investigated the specificity of siRNA constructs de- down across multiple cell lines (Fig. 4a). Using qRT- signed against circANKRD12, Exon7, and Exon9 PCR, we validated the differential expression of these through a series of knockdown experiments explained in selected genes in SKOV3 cells (Fig. 4c). Cyclin D1 Additional file 4: Supplementary file S2:4–7). (CCND1) was among the genes that showed consistent downregulation in the circANKRD12 knockdown across Silencing of circANKRD12 changes molecular phenotypes multiple cell lines (Fig. 4a). To ensure that the effect on of ovarian, breast and lung cancer cells CCND1 is related to circANKRD12 knockdown rather RNA-sequencing was performed in triplicate for two ovar- than off-target effects we compared the knockdown ian cancer cell lines (SKOV3, OVCAR3), breast (MDA- effects in SKOV3 and LL24 cells (Fig. 4b). Unlike SKOV3 MB-231) and a lung cancer cell line (NCI-H226) for both cells, the circANKRD12 level is low in LL24 cell line, at- scrambled control and siRNA targeting circANKRD12. tempts at knocking down circANKRD12 should not affect Table 1 gives the number of differentially expressed genes the cyclin D1 level in the cell line. Indeed, transfection of in circANKRD12 knockdown samples (Additional file 5: the siRNA against circANKRD12 in LL24 showed a reduc- Supplementary file S1) at a false discovery rate of less than tion of the already low levels of circANKRD12 but no ef- 5% with at least 1.5-fold changes in expression. fect on cyclin D1 expression, suggesting the siRNA effect The canonical pathways analysis by Ingenuity’sIPA on CCND1 is likely mediated through circANKRD12. Be- toolkit (IPA®, QIAGEN Redwood City, (https://www.qia- cause cell cycle regulation seems to be one of the most genbioinformatics.com/products/ingenuity-pathway-ana- deregulated pathways in circANKRD12 silenced cells, we lysis/) revealed an enrichment of differentially regulated decided to follow it up with further functional screening genes involved in cell cycle regulation, invasion, migra- using cell based phenotypic assays. tion, and interferon signaling pathways (Fig. 3, Additional file 2: Figure S2, Additional file 4: Table S3). siRNA mediated silencing of circANKRD12 increases We observed an upregulation of inflammatory pathways migration and invasion in ovarian cancer cells such as tumor necrosis pathway, NFkB pathway, inter- Wound healing assays show that silencing of cir- feron signaling and down regulation of the cell cycle ANKRD12 increased cell migration (Fig. 5a, b, pathway (Additional file 3: S2:15). Silencing of cir- Additional file 3: S2:10), in SKOV3 cells after 24 h and cANKRD12 affects cellular bioenergetics by downregula- 48 h. compared to scrambled control. This is confirmed tion of key genes involved in oxidative phosphorylation, again by cell migration assays in synchronized (Thymi- AMPK pathway and cell metabolism (Additional file 4: dine incorporation) cells, which show an increased mi- Table S4). IPA predicts a significant activation of inter- gration rate for circANKRD12 silenced SKOV3 cells feron signaling pathway through upregulation of STAT1, (Fig. 5c). Matrigel invasion (inserts coated with matrigel) STAT2, IFT1, MX1genes, and downregulation of and migration analysis using Boyden chamber shows cir- cANKRD12 silenced cells undergo significant increase in the invasion and migration compared to scrambled con- Table 1 Number differentially up and downregulated genes for trol (Additional file 3: S2:10). four cancer cell lines in siRNA-mediated knockdown of circANKRD12 Silencing of circANKRD12 decreases cell proliferation Si-circANKRD12 vs. Scrambled Knockdown of circANKRD12 and ANKRD12 mRNA SKOV3231 OVCAR3 MDA-MB-231 NCI-H226 significantly reduces cell proliferation (Fig. 5d, Additional Upregulated genes 25 104 237 168 file 3: S2:9). The results of MTS and ATP assays show a Downregulated genes 9 71 179 3 significant reduction in cell proliferation in circANKRD12 Karedath et al. BMC Cancer (2019) 19:565 Page 8 of 17 Fig. 3 Gene network analysis of RNA-Seq data from MDA-MB-231 and OVCAR3 cells. a Gene network analysis using ingenuity pathway analysis in MDA-MB-231 cells silenced for circANKRD12 shows upregulation of Interferon, JAK/STAT, IL-12 signaling pathways. b Gene network analysis using ingenuity pathway analysis in OVCAR3 cells silenced with circANKRD12 shows downregulation of Cyclins and Cell Cycle regulation pathways silenced cells compared to scrambled control (Fig. 5d, e. SKOV3 cells (Fig. 5f). These results indicate that knock- The Trypan blue exclusion assays show the silencing of downs of both circular and linear RNA forms of circANKRD12 is not affecting the cell viability signifi- ANKRD12 gene are capable of inducing strong pheno- cantly. However, silencing of ANKRD12 mRNA signifi- typic changes and modulate the growth or survival of can- cantly reduces both proliferation and cell viability in cer cells. Karedath et al. BMC Cancer (2019) 19:565 Page 9 of 17 Fig. 4 Silencing of circANKRD12 affects gene expression changes in cancer cells. a Heat map of differentially expressed genes from RNAseq analysis of SKOV3, NCI-H226, MDAMB 231 and OVCAR3 cells silenced with si-circANKRD12 compared to scrambled control. b qRT-PCR validation of expression level of cyclin D1 in circANKRD12 silenced SKOV3 and LL24 cell lines. In SKOV3 cells, circANKRD12 knockdown leads to a significant downregulation of Cyclin D1 (P = 7.8E-04), while as in LL24 Cyclin D1 shows a minimal non-significant change in expression (P = 0.18). c qRT-PCR validation of some selected genes differentially expressed in SKOV3 cells silenced with circANKRD12. Data in b–c are the means with error bars indicating standard error of the mean of three experiments. **P < 0.01 (Student’s t-test) Silencing of cirANKRD12 in 3D tumor models induces a [27], we investigated the effects of circANKRD12 knock- phenotypic switch from highly proliferative to an invasive down in 3D culture experiments. The 3D anchorage in- phenotype dependent growth of siRNA transfected SKOV3 cells Since 3D culture is known to have better cell-to-cell showed smaller, dense aggregates associated with an interactions and more closely mimics tumors in vivo invasive phenotype (Additional file 3:S2:12). The3D Karedath et al. BMC Cancer (2019) 19:565 Page 10 of 17 ab cd ef Fig. 5 Silencing of circANKRD12 results in an invasive phenotype with low proliferation rate. a Representative Wound healing migration assay image of circANKRD12 silenced SKOV3 cells and its scrambled control at 0 h and 24 h. b Bar diagram depicting the reduction in wound width after migration in scrambled and circANKRD12 silenced cells (average of 15 wounds from three different biological replicates). c Migration of circANKRD12 silenced SKOV3 cells after cell cycle synchronization with thymidine. d The proliferation of SKOV3 cells transfected with siRNA against circANKRD12 and assessed using MTS cell proliferation assay kit at the indicated hours and data shown as % proliferation compared to control. e Cell titer glow assay shows relative luminescence rate of ATP production in circANKRD12 silenced SKOV3 cells compared to scrambled control. f Cell viability is measured by using trypan blue exclusion assay. In circANKRD12 knockdown, the viability is reduced by roughly 2% while it is reduced by ~ 34% in ANKRD12-mRNA knockdown. (Data in b–f are the means with error bars indicating standard error of the mean of three experiments. **P < 0.01 (Student’s t-test) Karedath et al. BMC Cancer (2019) 19:565 Page 11 of 17 organotypic models were efficiently silenced with cir- Table S4). In order to determine whether the altered cANKRD12 siRNA with 74% knockdown efficacy (Fig. 6a). gene expression of these pathways translates to changes This reductioninspheroidsize in circANKRD12-silenced in basic metabolism, we analyzed metabolic phenotypes SKOV3 organotypic models resultsinanalterationof of circANKRD12 silenced MDA-MB-231 and SKOV3 phenotype from less invasive to a more invasive one (Fig. cells using Seahorse extracellular flux analyzer. The oxy- 6b). Cell proliferation MTS assay shows a significant reduc- gen consumption ratio (OCR) and ATP linked OCR ana- tion in cell growth within a time period of 24 h and 48 h lysis shows that knockdown of circANKRD12 decreases after knockdown of ANKRD12 circRNA (Fig. 6c). Collagen oxidative phosphorylation (OXPHOS) of MDAMB231 invasion of the circANKRD12 silenced spheroids shows in- cells and SKOV3 cells (Fig. 8c-e). Previous reports have creased invasion through the collagen gel with highly mo- suggested that high invasive potential of cancer cells is tile cells after 10 days of transfection (Fig. 6d). These results negatively correlated to high energetic cancer phenotype indicate the phenotypic switching from a highly prolifera- [29, 30]. These results thus indicate that circANKRD12 tive phenotype to a less proliferative phenotype with high silencing can induce phenotypic switching between invasion potential. highly proliferative cells to highly invasive cells through cyclin D1 deregulation and shifting the oncobioener- Cyclin D1 is down-regulated in circANKRD12 silenced cells getics to a low energy phenotype to facilitate invasion. and involved in phenotypic switching by facilitating G1 arrest circANKRD12 silencing results in cyclin D1 down- Discussion regulation and subsequent invasion and reduction in prolif- Circular RNAs are attracting greater attention in RNA eration in ovarian cancer SKOV3 cells. Both real-time PCR biology as there is growing evidence for their role in and RNA-sequencing showed down-regulation of cyclin gene expression regulation. Even though a large reper- D1 in circANKRD12 silenced cells. The qRT-PCR analysis toire of circRNAs has been identified in different organ- shows cyclin D1 expression is downregulated at least until isms, tissues, diseases, and developmental conditions, 48 h after transfection with the siRNA (Fig. 7a). This is only a few have been evaluated for their role in cellular also supported by the Western blot analysis, which shows functions [31–34]. Functional screening using siRNA a reduction in cyclin D1 protein level (Fig. 7b). Only cyclin targeting is difficult for circRNAs compared to other D1 shows differential regulation at the protein level in cir- RNA types as the choice of the target region is limited cANKRD12 silenced cells compared to other cyclin mem- to a few base pairs with specificity only at the backsplice bers including cyclin E1, pB1, and D2 (Additional file 3: junction. This constraint severely restricts the scale of S2:13). To estimate the duration of the RNAi effect initi- circRNAs suitable for further functional studies using ated by transfecting siRNA, we performed a longevity siRNA-mediated knockdown approaches. assay of circANKRD12 siRNA (si-circANKRD12). The In this study, we identified and characterized circular longevity, initially stable for six days (Additional file 3:S2: RNA isoforms from ANKRD12 gene (circANKRD12) 14) was checked on a cell doubling time basis estimated that are abundantly expressed in ovarian and breast to be 48 h (see Methods). The si-circANKRD12 exhibits cancer cells. A recent study identified stable levels of cir- an extended longevity period in SKOV3 cells, approaching cANKRD12 (exon 2–8) in young and old erythrocyte nine doublings (18 days). The reduction in the level of cyc- cells [35]. We show that circANKRD12 is a stable circu- lin D1 is on par with the observed silencing efficiency and larized transcript, resistant to RNase R digestion and longevity of siRNA (Fig. 7c). After the 10th doubling, the devoid of a polyA tail. In contrast to the linear mRNA effect on cell proliferation induced by silencing of cir- form, which is predominantly nuclear, circANKRD12 is cANKRD12 becomes insignificant (Fig. 7d), suggesting localized in the cytoplasm. that the knockdown efficiency is lost due to multiple pas- Differential gene expression analysis by RNA-seq of sages of cells. As cyclin D1 is involved in cell cycle pro- circANKRD12 silenced cells revealed that silencing may gression [28], cell cycle analysis of circANKRD12 accelerate cancer cell invasion, migration, and cellular knockdowns was conducted. Cell cycle analysis using movement by regulating a cascade of genes involved in FACS shows there is a significant G0/G1 cell cycle ar- these processes (Additional file 4: Table S3, 5). Inter- rest in SKOV3 cells silenced with circANKRD12 com- feron signaling and cell cycle checkpoint regulation by pared to scrambled control (Fig. 8a, b). G1/S transition are the top networks deregulated in circANKRD12-silenced cells. IPA analysis shows an in- Silencing of circANKRD12 affects oxygen consumption crease in pathways related to the invasion of cancer cells ratio (OCR) in SKOV3 and MDAMB231 cells with significant p-value and z score (Additional file 2: As AMPK signaling and other metabolic pathways are Figure S2: Fig. 2a,b,c) in MDA-MB-231 cells. The key affected in circANKRD12 knockdown (Additional file 4: genes upregulated in circANKRD12 silenced cells are Karedath et al. BMC Cancer (2019) 19:565 Page 12 of 17 ab Fig. 6 Silencing of circANKRD12 in 3D spheroid models of SKOV3 cells shows an invasive phenotype. a Knockdown efficacy of circANKRD12 in 3D cultures. siRNA directed against circANKRD12 downregulates circular RNA but shows no downregulation on mRNA of ANKRD12 gene in SKOV3 organotypic models (b) spheroid area is measured in anchorage-independent 3D spheroids of SKOV3 cell lines, average of 100 spheroids in 3 independent experiments were measured. c MTS assay shows cell viability in 3D organotypic models is reduced at 24 and 48 h of transfection. d Collagen invasion assay of spheroids 24,48 h and 10 days. circANKRD12 silenced cells were able to invade through the thick collagen matrix and circANKRD12 silenced cells shows an invasive phenotype. Data in a–c are the means with error bars representing standard error of the mean of three experiments. **P < 0.01 (Student’s t-test) Karedath et al. BMC Cancer (2019) 19:565 Page 13 of 17 ab Fig. 7 Longevity of circANKRD12 silencing in SKOV3 cells a qRT-PCR analysis of circANKRD12 knockdown efficiency at two different time points (24 h, 48 h). b Western blot analysis of cyclin D1 expression in circANKRD12 and ANKRD12 silenced SKOV3 cells at 48 h. c Longevity analysis of circANKRD12 silenced SKOV3 cells. qRT-PCR analysis for the longevity of siRNA transfection based on knockdown efficiency of circANKRD12 in each doubling time (48 h). The figure shows the Knockdown efficiency of circANKRD12 on cyclin D1 expression. qRT-PCR of the expression pattern of cyclin D1 from 1st to 9th doubling time d MTS cell proliferation assay of circANKRD12 knockdown cells after 10th doubling. Data in a is the means with error bars representing standard error of the mean. **P < 0.01 (Student’s t-test) Karedath et al. BMC Cancer (2019) 19:565 Page 14 of 17 ab cd Fig. 8 circANKRD12 silenced cells have an arrested G1stage and low OXPHOS compared to scrambled control. a Cell cycle analysis of SKOV3 cells silenced with circANKRD12 at 24 h. b Bar chart representation of % of cells in G0/G1 stage. Error bars represent standard error of the mean from three experimental replicates. c,d Oxygen consumption ratio and ATP production analyzed by seahorse extracellular flux analyzer in MDAMB231 cells silenced with si-circANKRD12, si-linear ANKRD12 and its control. e Oxygen consumption ratio analyzed by seahorse by extracellular flux analysis in SKOV3 cells. Data in c, d, and e are normalized with total protein(ug/ul) STAT-1, MX1, NFkB, MUC4, and SMAD3. Cell based reducing cell proliferation and increasing invasion. Pre- assays live cell migration, invasion and wound healing vious studies have reported a reduction in cyclin D1 assays also shows increased invasion of circANKRD12 levels, and G0/G1 cell cycle stage arrest leads to an in- silenced cells in SKOV3 cell line. crease in migratory activities of MDA-MB-231 breast Cell proliferation was significantly arrested without cancer cells [36]. Consistent with these findings, our re- any change in cell viability in circANKRD12 silenced sults of cell cycle analysis also show a substantial G0/G1 SKOV3 cells. Cyclin D1, a consistently affected gene by arrest with increased migration and invasion in cir- circANKRD12 silencing, is down regulated thereby cANKRD12 knockdown cells. The migration and Karedath et al. BMC Cancer (2019) 19:565 Page 15 of 17 invasion assays significantly correlate with gene expres- abundant circRNA in human ovarian and breast cancer sion analysis and indicate an augmented cell migration cells. Our results suggest that circANKRD12 could be and invasion rate. The 3D anchorage independent orga- involved in a diverse set of functions ranging from cell notypic tumor models of SKOV3 show similar patterns cycle arrest, tumor invasion to immune modulation. Ma- of phenotypic characterization where invasion through nipulating the levels of circANKRD12 can regulate mo- collagen is increased upon silencing of circANKRD12. lecular functions by altering different signaling pathways There is a strong phenotypic alteration after silencing and modifies the phenotype of the cells. The distinctive circANKRD12 in the ovarian cancer cells in both 2D change from a proliferative to a more invasive phenotype and 3D culture conditions suggesting that circANKRD12 by altering circANKRD12 levels could lead to a future is important in regulating proliferation, invasion, and circRNA based therapeutic intervention in cancer. migration. Reduction in cell proliferation and regulation of Cyclin D1 expression is confirmed by silencing cir- Additional files cANKRD12 by another construct of siRNA in SKOV3 Additional file 1: Figure S1. (a) The abundance of circANKRD12 and cells (Additional file 3: S2:16a, b). circRNAs are involved ANKRD12 linear RNA. Semi qRT-PCR data indicating the abundance of in regulating cellular movement. Indeed, circRNA (F-cir- circANKRD12 and ANKRD12 mRNA in the cytoplasmic and nuclear fractions cEA) produced from the EML4-ALK fusion gene which of SKOV3 cells, nuclear and cytoplasmic purity markers were also assessed using nuclear specific marker 7SK and cytoplasmic specific marker CYTB. The is independent of the EML4-ALK linear transcript and gel picture represents PCR amplified products of circular RNAs with three the fusion protein, can promote migration and invasion, different sets of primers. (b) Real-time data shows the abundance of thus contributing to tumor metastasis [37, 38]. circANKRD12 and ANKRD12 mRNA in the cytoplasm and nucleus. (c) The abundance of circANKRD12 and ANKRD12 mRNA in a panel of cancer cells. The Ankyrin Repeat Domain family of genes can act (PPTX 2728 kb) (PPTX 2750 kb) as putative tumor suppressors via p53 mediated feedback Additional file 2: Figure S2. Figure a, b, c Shows functional predicted or through recruiting histone deacetylases (HDACs) to networks from RNAseq data of circANKRD12 silenced MDA-MB-231 cells the p160 coactivator to repress transcriptional activities induces inflammatory immune responses and cancer cell invasion with activation Z score and p values. (PPTX 6928 kb) (PPTX 6949 kb) [18, 19, 39]. A clinical study of gene expression of Additional file 3: S2. File represents additional figures for validating the ANKRD12 in colorectal cancer revealed that low circANKRD12 in cell lines. (PPTX 17563 kb) (PPTX 17572 kb) ANKRD12 expression is correlated with overall poor Additional file 4: Tables S1-S5. Tables representing primers used for survival and liver metastasis of CRC patients [20]. different gene expression studies, siRNAs and pathways involved in We observed that the silencing of ANKRD12 mRNA circANKRD12 gene knockdown condition. (PPTX 3253 kb) (PPTX 3268 kb) reduces cell proliferation, induces cell death and down Additional file 5: Supplementary file S1. List of genes differentially expressed in circANKRD12 silenced cells compared to control in different regulates cyclin D1. On the contrary, silencing of cell lines. (XLSX 160 kb) circANKRD12 arrests the cell cycle progression and Additional file 6: Table S6. List of microRNAs that can target increases tumor invasion without significantly affecting circANKRD12 and CyclinD1. (XLSX 9 kb) cell viability. The knockdown of circANKRD12 can change the oncobioenegetics as its shift from a higher Abbreviations OXPHOS to a lower OXPHOPS phenotype which is 3D culture: Three-Dimensional cell culture; circRNA: Circular RNA; FDR: False discovery rate; OCR: Oxygen Consumption Ratio; OXPHOS: Oxidative highly invasive. circANKRD12 may act as competing Phosphorylation; qRT-PCR: Quantitative Real Time PCR; RNAseq: RNA endogenous RNA (ceRNA) to regulate a circRNA- sequencing; siRNA: Small (or short) interfering RNA miRNA-mRNA network. Our insilco analysis shows that Acknowledgments cyclin D1 and circANKRD12 have shared binding sites for The work was supported by grants from Basic Medical Research Program several different microRNAs (Additional file 6: Table S6). (BMRP) grant from Qatar Foundation to WCM-Q. Thus, circANKRD12 could act as a microRNA sponge to We thank Shameem Yunuskunju form Genomics Core at WCM-Q for RNA -Seq data handling. We thank Dr. Anna Halama from Dr. Karsten Suhre’s Bio- regulate cyclin D1 levels. Our preliminary analysis shows informatics Core lab at WCM-Q for providing the LL24, NCI-H226, T47D, overexpression of hsa-miR-4768-5p reduces the level of HCC2935 cells. We also thank Ms. Aleksandra M. Liberska from Microscopy Cyclin D1 by 20% (hsa-miR-4768-5p has common binding Core at WCM-Q for helping with Flow Cytometry. sites for circANKRD12 and Cyclin D1) (Data not shown). Funding Thus, circANKRD12 could act as a microRNA sponge to This study was supported by Weill Cornell Medicine and Qatar Foundation regulate cyclin D1 levels. We also observed that (BMRP 1 - Malek Pilot FY17). The funders had no role in the design of the study,data analysis, interpretation of data and writing the manuscript. circANKRD12 contains some putative open reading frames (ORFs) and therefore its translation through in- Availability of data and materials ternal ribosomal entry sites cannot be ruled out [40, 41]. The datasets supporting the conclusions of this article are included in this article and the Supplementary Data. Conclusion Authors’ contributions In conclusion, our study provides the molecular, pheno- JM, TK, and IA designed the research; TK, WAA, FAD, SA, and SS performed typic and metabolic characterization of one of the most cellular experiments. TK and IA contributed equally to this work. IAA and Karedath et al. BMC Cancer (2019) 19:565 Page 16 of 17 YAM performed RNA-seq and IA and JM performed bioinformatic analysis; 15. Xia W, Qiu M, Chen R, Wang S, Leng X, Wang J, et al. Circular RNA has_circ_ AR provided the cell lines and helped with experimental design; TK and IA 0067934 is upregulated in esophageal squamous cell carcinoma and promoted wrote the manuscript. All authors read and approved the final manuscript. proliferation. Sci Rep. 2016;6:35576. https://doi.org/10.1038/srep35576. 16. Li F, Zhang L, Li W, Deng J, Zheng J, An M, et al. Circular RNA ITCH has Ethics approval and consent to participate inhibitory effect on ESCC by suppressing the Wnt/$β$-catenin pathway. Not applicable for the present study. All experiments were conducted in Oncotarget. 2015;6:6001–13. https://doi.org/10.18632/oncotarget.3469. cancer cell lines, no human or animal subject is involved in the study. 17. Dong Y, He D, Peng Z, Peng W, Shi W, Wang J, et al. Circular RNAs in cancer: an emerging key player. J Hematol Oncol. 2017. https://doi.org/10. Consent for publication 1186/s13045-016-0370-2. Not applicable. 18. Li C-W, Dinh GK, Zhang A, Chen JD. Ankyrin repeats-containing cofactors interact with ADA3 and modulate its co-activator function. Biochem J. 2008. Competing interests https://doi.org/10.1042/BJ20071484. The authors declare that they have no competing interests. 19. Neilsen PM, Cheney KM, Li C-W, Chen JD, Cawrse JE, Schulz RB, et al. Identification of ANKRD11 as a p53 coactivator. J Cell Sci. 2008;121 https:// www.ncbi.nlm.nih.gov/pubmed/?term=Identification+of+ANKRD11+as+a+ Publisher’sNote p53+coactivator.+J+Cell+Sci.+2008. Accessed 11 Apr 2017. Springer Nature remains neutral with regard to jurisdictional claims in 20. Bai R, Li D, Shi Z, Fang X, Ge W, Zheng S. Clinical significance of Ankyrin published maps and institutional affiliations. repeat domain 12 expression in colorectal cancer. J Exp Clin Cancer Res. https://doi.org/10.1186/1756-9966-32-35. Author details 21. Ahmed I, Karedath T, Andrews SS, Al-Azwani IK, Ali Mohamoud Y, Querleu Department of Genetic Medicine, Weill Cornell Medicine-Qatar, Education D, et al. Altered expression pattern of circular RNAs in primary and City, Qatar Foundation, P.O. Box No, 24144 Doha, Qatar. Genomics Core, metastatic sites of epithelial ovarian carcinoma. Oncotarget. 2016. https:// Weill Cornell Medicine-Qatar, Education City, Qatar Foundation, Doha, Qatar. doi.org/10.18632/oncotarget.8917. Department of Pharmacology, Weill Cornell Medicine-Qatar, Education City, 22. Zheng Q, Bao C, Guo W, Li S, Chen J, Chen B, et al. Circular RNA profiling Qatar Foundation, Doha, Qatar. Stem Cell and Microenvironment reveals an abundant circHIPK3 that regulates cell growth by sponging multiple Laboratory, Weill Cornell Medicine-Qatar, Education City, Qatar Foundation, miRNAs. Nat Commun. 2016;7:11215. https://doi.org/10.1038/ncomms11215. Doha, Qatar. 23. Liang C-C, Park AY, Guan J-L. In vitro scratch assay: a convenient and inexpensive method for analysis of cell migration in vitro. Nat Protoc. 2007; Received: 25 July 2018 Accepted: 23 April 2019 2:329–33. https://doi.org/10.1038/nprot.2007.30. 24. Luo J, Wang X, Xia Z, Yang L, Ding Z, Chen S, et al. Transcriptional factor specificity protein 1 (SP1) promotes the proliferation of glioma cells by up- References regulating midkine (MDK). Mol Biol Cell. 2015;26:430–9. https://doi.org/10. 1. Jeck WR, Sharpless NE. Detecting and characterizing circular RNAs. Nat 1091/mbc.E14-10-1443. Biotechnol. 2014;32. https://doi.org/10.1038/nbt.2890. 25. Samuel SM, Ghosh S, Majeed Y, Arunachalam G, Emara MM, Ding H, et al. 2. Cocquerelle C, Mascrez B, Hétuin D, Bailleul B. Mis-splicing yields circular Metformin represses glucose starvation induced autophagic response in RNA molecules. FASEB J. 1993;7:155–60 https://www.fasebj.org/doi/abs/10. microvascular endothelial cells and promotes cell death. Biochem 1096/fasebj.7.1.7678559. Accessed 1 May 2017. Pharmacol. 2017;132:118–32. https://doi.org/10.1016/J.BCP.2017.03.001. 3. Wang PL, Bao Y, Yee M-C, Barrett SP, Hogan GJ, Olsen MN, et al. Circular 26. Vincent HA, Deutscher MP. Substrate recognition and catalysis by the RNA is expressed across the eukaryotic tree of life. PLoS One. 2014;9:e90859. exoribonuclease RNase R. J Biol Chem. 2006;281:29769–75. https://doi.org/ https://doi.org/10.1371/journal.pone.0090859. 10.1074/jbc.M606744200. 4. Dang Y, Yan L, Hu B, Fan X, Ren Y, Li R, et al. Tracing the expression of 27. Pampaloni F, Reynaud EG, Stelzer EHK. The third dimension bridges the gap circular RNAs in human pre-implantation embryos. Genome Biol. 2016. between cell culture and live tissue. Nat Rev Mol Cell Biol. 2007;8:839–45. https://doi.org/10.1186/s13059-016-0991-3. https://doi.org/10.1038/nrm2236. 5. Hansen TB, Jensen TI, Clausen BH. Natural RNA circles function as efficient 28. Baldin V, Lukas J, Marcote MJ, Pagano M, Draetta G. Cyclin D1 is a nuclear microRNA sponges. Nature. 2013;495. https://doi.org/10.1038/nature11993. protein required for cell cycle progression in G1. Genes Dev. 1993;7:812–21. 6. Ashwal-Fluss R, Meyer M, Pamudurti NR, Ivanov A, Bartok O, Hanan M, et al. https://doi.org/10.1101/GAD.7.5.812. circRNA biogenesis competes with pre-mRNA splicing. Mol Cell. 2014;56. 29. Yang L, Moss T, Mangala LS, Marini J, Zhao H, Wahlig S, et al. Metabolic https://doi.org/10.1016/j.molcel.2014.08.019. shifts toward glutamine regulate tumor growth, invasion and bioenergetics 7. Li J, Yang J, Zhou P, et al. Circular RNAs in cancer: novel insights into origins, in ovarian cancer. Mol Syst Biol. 2014;10. https://doi.org/10.1002/msb. properties, functions and implications. Am J Cancer Res. 2015;5(2):472–80. 30. Vayalil PK, Landar A. Mitochondrial oncobioenergetic index: a potential 8. He J, Xie Q, Xu H, Li J, Li Y. Circular RNAs and cancer. Cancer Lett. 2017;396: biomarker to predict progression from indolent to aggressive prostate cancer. 138–44. https://doi.org/10.1016/j.canlet.2017.03.027. Oncotarget. 2015;6:43065–80. https://doi.org/10.18632/oncotarget.5487. 9. Qu S, Liu Z, Yang X, Zhou J, Yu H, Zhang R, et al. The emerging functions 31. Guarnerio J, Bezzi M, Jeong JC, Paffenholz SV, Berry K, Naldini MM, et al. and roles of circular RNAs in cancer. Cancer Lett. 2018;414:301–9. https://doi. Oncogenic role of fusion-circRNAs derived from Cancer-associated chromosomal org/10.1016/J.CANLET.2017.11.022. translocations. Cell. 2016;165:289–302. https://doi.org/10.1016/j.cell.2016.03.020. 10. Li Z, Yanfang W, Li J, Jiang P, Peng T, Chen K, et al. Tumor-released 32. Du WW, Yang W, Chen Y, Wu Z-K, Foster FS, Yang Z, et al. Foxo3 circular exosomal circular RNA PDE8A promotes invasive growth via the miR-338/ RNA promotes cardiac senescence by modulating multiple factors MACC1/MET pathway in pancreatic cancer. Cancer Lett. 2018. https://doi. associated with stress and senescence responses. Eur Heart J. 2016:ehw001. org/10.1016/j.canlet.2018.04.035. https://doi.org/10.1093/eurheartj/ehw001. 11. Chen B, Huang S. Circular RNA: An emerging non-coding RNA as a 33. Huang G, Zhu H, Shi Y, Wu W, Cai H, Chen X, et al. Cir-ITCH plays an inhibitory regulator and biomarker in cancer. Cancer Lett. 2018;418:41–50. https://doi. role in colorectal Cancer by regulating the Wnt/$β$-catenin pathway. PLoS org/10.1016/j.canlet.2018.01.011. One. 2015;10:e0131225. https://doi.org/10.1371/journal.pone.0131225. 12. Meng X, Li X, Zhang P, Wang J, Zhou Y, Chen M. Circular RNA: an emerging 34. Dong W, Bi J, Liu H, Yan D, He Q, Zhou Q, et al. Circular RNA ACVR2A key player in RNA world. Brief Bioinform. 2016:bbw045. https://doi.org/10. suppresses bladder cancer cells proliferation and metastasis through miR- 1093/bib/bbw045. 626/EYA4 axis. https://doi.org/10.1186/s12943-019-1025-z. 13. Memczak S, Papavasileiou P, Peters O, Rajewsky N. Identification and characterization of circular RNAs as a new class of putative biomarkers in 35. Nicolet BP, Engels S, Aglialoro F, van den Akker E, von Lindern M, Wolkers human blood. PLoS One. 2015;10:e0141214. https://doi.org/10.1371/journal. MC. Circular RNA expression in human hematopoietic cells is widespread pone.0141214. and cell-type specific. Nucleic Acids Res. 2018;46:8168–80. https://doi.org/10. 14. Hang D, Zhou J, Qin N, Zhou W, Ma H, Jin G, et al. A novel plasma circular 1093/nar/gky721. RNA circFARSA is a potential biomarker for non-small cell lung cancer. 36. Lehn S, Tobin NP, Berglund P, Nilsson K, Sims AH, Jirström K, et al. Down- Cancer Med. 2018. https://doi.org/10.1002/cam4.1514. regulation of the oncogene cyclin D1 increases migratory capacity in breast Karedath et al. BMC Cancer (2019) 19:565 Page 17 of 17 cancer and is linked to unfavorable prognostic features. Am J Pathol. 2010; 177:2886–97. https://doi.org/10.2353/ajpath.2010.100303. 37. Tan S, Gou Q, Pu W, Guo C, Yang Y, Wu K, et al. Circular RNA F-circEA produced from EML4-ALK fusion gene as a novel liquid biopsy biomarker for non-small cell lung cancer. Cell Res. 2018;28:693–5. https://doi.org/10. 1038/s41422-018-0033-7. 38. Tan S, Sun D, Pu W, Gou Q, Guo C, Gong Y, et al. Circular RNA F-circEA-2a derived from EML4-ALK fusion gene promotes cell migration and invasion in non-small cell lung cancer. Mol Cancer. 2018;17:138. https://doi.org/10. 1186/s12943-018-0887-9. 39. Zhang A, Li C-W, Chen JD. Characterization of transcriptional regulatory domains of ankyrin repeat cofactor-1. Biochem Biophys Res Commun. 2007; 358:1034–40. https://doi.org/10.1016/j.bbrc.2007.05.017. 40. Legnini I, Di Timoteo G, Rossi F, Morlando M, Briganti F, Sthandier O, et al. Circ- ZNF609 is a circular RNA that can be translated and functions in Myogenesis. Mol Cell. 2017;66:22–37.e9. https://doi.org/10.1016/j.molcel.2017.02.017. 41. Pamudurti NR, Bartok O, Jens M, Ashwal-Fluss R, Stottmeister C, Ruhe L, et al. Translation of CircRNAs. Mol Cell. 2017;66:9–21.e7. https://doi.org/10. 1016/j.molcel.2017.02.021. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Cancer Springer Journals

Silencing of ANKRD12 circRNA induces molecular and functional changes associated with invasive phenotypes

Download PDF
17 pages

Loading next page...
 
/lp/springer-journals/silencing-of-ankrd12-circrna-induces-molecular-and-functional-changes-42T9noqaHF

References (43)

Publisher
Springer Journals
Copyright
Copyright © 2019 by The Author(s).
Subject
Biomedicine; Cancer Research; Oncology; Surgical Oncology; Health Promotion and Disease Prevention; Biomedicine, general; Medicine/Public Health, general
eISSN
1471-2407
DOI
10.1186/s12885-019-5723-0
Publisher site
See Article on Publisher Site

Abstract

Background: Circular RNAs (circRNAs) that form through non-canonical backsplicing events of pre-mRNA transcripts are evolutionarily conserved and abundantly expressed across species. However, the functional relevance of circRNAs remains a topic of debate. Methods: We identified one of the highly expressed circRNA (circANKRD12) in cancer cell lines and characterized it validated it by Sanger sequencing, Real-Time PCR. siRNA mediated silencing of the circular junction of circANKRD12 was followed by RNA Seq analysis of circANKRD12 silenced cells and control cells to identify the differentially regulated genes. A series of cell biology and molecular biology techniques (MTS assay, Migration analysis, 3D organotypic models, Real-Time PCR, Cell cycle analysis, Western blot analysis, and Seahorse Oxygen Consumption Rate analysis) were performed to elucidate the function, and underlying mechanisms involved in circANKRD12 silenced breast and ovarian cancer cells. Results: In this study, we identified and characterized a circular RNA derived from Exon 2 and Exon 8 of the ANKRD12 gene, termed here as circANKRD12. We show that this circRNA is abundantly expressed in breast and ovarian cancers. The circANKRD12 is RNase R resistant and predominantly localized in the cytoplasm in contrast to its source mRNA. We confirmed the expression of this circRNA across a variety of cancer cell lines and provided evidence for its functional relevance through downstream regulation of several tumor invasion genes. Silencing of circANKRD12 induces a strong phenotypic change by significantly regulating cell cycle, increasing invasion and migration and altering the metabolism in cancer cells. These results reveal the functional significance of circANKRD12 and provide evidence of a regulatory role for this circRNA in cancer progression. Conclusions: Our study demonstrates the functional relevance of circANKRD12 in various cancer cell types and, based on its expression pattern, has the potential to become a new clinical biomarker. Keywords: Circular RNA, RNA-seq, Breast cancer, siRNA, Cancer invasion, OCR, OXPHOS * Correspondence: jom2042@qatar-med.cornell.edu Thasni Karedath and Ikhlak Ahmed contributed equally to this work. Department of Genetic Medicine, Weill Cornell Medicine-Qatar, Education City, Qatar Foundation, P.O. Box No, 24144 Doha, Qatar Genomics Core, Weill Cornell Medicine-Qatar, Education City, Qatar Foundation, Doha, Qatar Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Karedath et al. BMC Cancer (2019) 19:565 Page 2 of 17 Background regulation pathways in cancer. We confirmed the expres- Exonic circular RNAs (circRNAs) are a class of RNA in bio- sion of this circRNA across a variety of cancer cell lines logical systems for which function is not well understood. and provided evidence for its functional relevance through They possess distinct properties compared to linear RNAs downstream regulation of several tumor invasion genes. and arise from direct backsplicing events that covalently link the 3′ endofanexonwiththe 5′ end of either the Methods same exon or any other further upstream exon [1]. Cell lines and treatment circRNAs were initially considered as molecular ar- Ovarian cancer cell lines PA-1 (ATCC® CRL-1572), SKO- tifacts of aberrant RNA splicing [2]. This hypothesis V3(ATCC® HTB-77™),Caov3(ATCC® HTB-75™), NIH: was challenged by the observation that circRNAs are OVCAR-3 (ATCC® HTB-161™), breast cancer cell lines detected in various cell types in an evolutionarily con- MDA-MB-231(ATCC® HTB-26™), MCF7(ATCC® HTB- served manner [3]. The copy number of circRNAs 22™),T-47D (ATCC® HTB-133™) and breast normal cell can be up to 10 times greater than that of associated line MCF 10A(ATCC® CRL-10317™), Lung cancer cell linear RNAs, suggesting that these circRNAs may lines, HCC2935(ATCC® CRL-2869™) and NCI-H226 possess biological functions [4]. Studies have shown (ATCC® CRL-5826™) and Lung Normal Fibroblast cell that some circRNAs harbor multiple binding sites for line LL 24(ATCC® CCL-151™) (all from American type microRNAs, thereby “sponging” microRNAs and serv- Cell Collection, Manassas, VA), APOCC (ovarian pri- ing as competitive inhibitors for microRNA functions mary cell line derived from ascites fluid) (pers. commu- [5]. To serve as a sponge, however, a circular RNA nication Dr. Arash Tabrizi), A27809 (93112519-1VL, must contain multiple microRNA-binding sites and be Sigma), A2780 CIS (93112519-1VL,Sigma) were used for expressed at sufficiently high levels in the cytoplasm the current study. Cells were cultured in DMEM (Life [1]. The majority of circRNAs may not fit this Technologies, NY, USA) supplemented with 10% fetal category, and their functions remain uncertain. Some bovine serum (Life Technologies, USA). Low Passage circRNAs have been shown to interact with RNA number cells were used for all the experiments. Cell binding proteins to form RNA protein complexes culture were routinely checked for mycoplasma contam- thereby regulating canonical linear splicing of the ination using MycoAlert Mycoplasma detection kit gene [6]. These findings suggest that circRNAs hold a (Lonza, Basel, Switzerland). dynamic and distinct role in gene regulation. Emerging evidence suggests that regulation of circRNAs RNA preparation and qRT–PCR is closely associated with different diseases, particularly The nuclear and cytoplasmic RNA was extracted using cancer with aberrant expression pattern [7–10]. Thus, cir- The SurePrep™ Nuclear or Cytoplasmic RNA Purifica- cRNAs represent a new class of diagnostic biomarkers tion Kit (Fisher Grand Island, NY, USA). Total RNA with potential therapeutic significance [11, 12]. The longer from whole-cell lysates were isolated by using RNAe- half-lives compared to their linear counterparts makes cir- say mini kit (Qiagen Valencia, CA USA). For RNase R cRNAs long-acting regulators of cellular behavior and treatment, 2 μg of total RNA was incubated 60 min at − 1 robust biomarkers [13, 14]. A growing body of evidence 40 °C with or without 3 U μg of RNase R (Epicentre has implicated the functional involvement of circRNAs in Technologies, Madison, WI), and the resulting RNA was regulating cancer progression and proliferation [14–17]. subsequently purified using an RNeasy MinElute cleaning In the present study, we investigated the functional Kit (Qiagen, Valencia, CA USA). cDNA synthesis was role of a high abundance circRNA from Ankyrin Repeat carried out using First strand synthesis kit (AMV) from Domain 12 (ANKRD12) gene in cancer progression. Roche Biosciences or Biorad select cDNA synthesis kit ANKRD12 is a paralog of ANKRD11, a putative tumor using random primer for circRNA experiments. Fast Start suppressor gene with multiple functions including as a Universal SYBR Green Master mix (Roche, Clovis, CA) p53 co-activator [18, 19]. Low ANKRD12 level is an was used to amplify the specific gene using cDNA primes independent prognostic predictor of colorectal carcin- obtained from Primer bank (https://pga.mgh.harvard.edu/ oma patients [20]. Circular isoforms of ANKRD12 have primerbank/ (Additional file 4: Table S1). Each Real-Time been identified in cancer cells and patient samples in assay was done in triplicate on Step One Plus Real-time many recent studies including ours [21, 22]. In this PCR machine (Life Technologies, CA, USA). study, we validated one of the most predominant circu- lar isoforms of ANKRD12 gene that includes the back- Transfection splice junction of exons 8 and 2 (circANKRD12). We siRNA transfection was carried out using custom- report that circANKRD12 regulates the invasion, migra- designed siRNAs for both ANKRD12 circular and linear tion proliferation and cellular bioenergetics of cancer cells transcripts (Fig. 1 and Additional file 4: Table S1). The by modulating cell signaling, metabolic and cell cycle SKOV3, MDA-MB-231, OVCAR3, NCI-H226 cells were Karedath et al. BMC Cancer (2019) 19:565 Page 3 of 17 bc Fig. 1 siRNA mediated silencing of circANKRD12 in cancer cells a Two circANKRD12 siRNAs spanning the back-splice junction b qRT-PCR analysis for knockdown efficiency of circANKRD12 siRNA in 4 different cell lines. c qRT-PCR analysis for knockdown efficiency of two different circANKRD12 siRNA constructs in SKOV3 cells. d qRT-PCR analysis for silencing efficiency of ANKRD12 linear siRNA (exon9) in SKOV3 cells. (Data in b–d are the means with error bars indicating standard error of the mean (SEM) of three experiments. **P < 0.01 (Student’s t-test) grown in 6 well plates for transfection. The cells were Cell proliferation assay transfected at 24 h with 30 pmol concentration of siRNA Cells were cultured at a density of 5 × 10 cells per (VWR, Radnor, PA, USA) or scrambled control (Mission well in flat-bottomed 96-well plates in DMEM sup- siRNA universal negative control, Sigma, St.Louis, USA) plemented with 10% FBS with antimycotic anti- using Lipofectamine RNAi max (Invitrogen MA USA) biotic. The experiment was done at different time according to manufacturer’s protocol. These experiments points (24 h and 48 h or based on cell doubling were conducted in three different biological triplicates for time). CellTiter 96® Aq One Solution Reagent ueous subsequent RNA-sequencing. (Promega,Madison,WI) wasadded and theexperi- ment was conducted according to the manufac- RNAseq analysis turer’sinstructions. SKOV3, OVCAR3, NCI-H226, MDA-MB-231 cells CellTiter-Glo assay (Promega Madison USA) for transfected with either circANKRD12 or a universal determining cell proliferation was conducted according scrambled control were used for RNAseq analysis. RNA- to the manufacturer’s instructions. Briefly, CellTiter- seq library preparation and In silico detection of cir- Glo reagent was added directly to the wells of 96 well cRNA candidates from paired-end RNA-seq data was plate and luminescence was measured on an Envision conducted as described earlier [21]. reader (PerkinElmer). Karedath et al. BMC Cancer (2019) 19:565 Page 4 of 17 Scratch assay- cell migration assay Cell cycle analysis Scratch assay was conducted as described earlier [23]. Cell cycle analysis was done on cells fixed with Cells were plated into a 6-well plate with complete FxCycle™ PI/RNase Staining Solution using BD LSRFor- medium and grown to 80% confluence. Cells were trans- tessa™ cell analyzer. fected with respective siRNA in OPTIMEM medium, and the medium was replaced with serum-free DMEM Western blot analysis after transfection. After cells were grown to 100% Cellular protein was extracted after 48 h of transfection. confluence, a wound was created by scraping the conflu- The cells were lysed in 100ul of RIPA buffer with prote- ent monolayer cells with a p200 pipette tip. Cells were ase inhibitor cocktail. Then 40 micrograms of protein then grown either in serum-free medium or medium were resolved in SDS PAGE gel and transferred to a containing 3 mM Thymidine. The distance between the nitrocellulose membrane. The primary antibodies used two sides of the cell-free area was photographed using were anti-Cyclin D1, Anti- Cyclin B1, Anti CyclinD2, 10x objective in AXION Zeiss epifluorescence micro- anti-Cyclin Phospho B1 and β-actin (Cell Signaling, scope. The distance is measured using Zeiss Zen USA). The blots were visualized by ECL detection software (Carl Zeiss Carpenteria, CA, USA). (Amersham, N J, United States). The Western blot experiments and analysis were done as described earlier [25]. Briefly, cell lysate protein (25 μg) was separated on Trans-well migration and invasion assay an SDS-Polyacrylamide gel and transferred onto nitro- Cellular migration and invasion were determined cellulose membranes. Membranes were then blocked using a Transwell Boyden chamber assay as described with 5% (w/v) non-fat dry milk tris-buffered saline (TBS) previously [24]. containing 0.1% (v/v) Tween20 and incubated at 4 °C with the primary antibody (1:1000 dilution). The excess 3D organotypic spheroid model experiments primary antibody was washed off in TBS-T wash buffer, 3D anchorage-independent spheroids were developed and the membranes were incubated with HRP-linked in SKOV3 cancer cell lines; initially, cells were seeded secondary antibodies (1:2500) for one h at room on ultra-low attachment plate (Corning, NY, USA) for temperature. The excess secondary antibody was then three days. Transfection of the spheroids with cir- washed off in TBS-T, and the protein levels were cANKRD12 siRNA was conducted using the reverse detected using enhanced chemiluminescence reagent transfection method. (Sigma-Aldrich, Inc., MO, USA) and imaged on a Geliance P600 gel documentation system (PerkinElmer, Inc. MA, USA). β-Actin was used as the loading control. Spheroid area measurement Cells were seeded at a density of 300,000 cells/well into circANKRD12 and ANKRD12 siRNA longevity assay ultralow attachment plates. After 72 h, cells were trans- siRNA longevity was calculated based on cell doubling fected either with scrambled siRNA or circANKRD12 time. The cell doubling time of SKOV3 cells was calcu- siRNA. After 48 h the diameters of at least 50 spheroids lated based on the mathematical equation. were measured and the spheroid area was measured Zeiss Zen Software. Doubling Time ¼ duration  logðÞ 2 = logðÞ Final Concentration Cell proliferation assay in 3D organotypic models Ten thousand cells were seeded on each well of ultra- − logðÞ Initial Concentration : low attachment 96 well plate with OPTIMEM medium. After three days, once spheroids were formed, cells Cells were seeded on 60 mm petri plates and transfec- were transfected with either circANKRD12 siRNA or tion was done with siRNA of circANKRD12, or scrambled control. After 24 h and 48 h of transfection, ANKRD12 mRNA. Cells were harvested on an interval of MTS reagent was added to the medium. Measurements 48 h (doubling Time). The experiment was extended till were according to the manufacturer’sinstructions. 10th doubling time (20 days). The silencing efficacy was measured by gene expression analysis (qRT-PCR). Collagen invasion assay in 3D organotypic models 3D organotypic models of either circANKRD12 trans- Assessment of mitochondrial function by seahorse fected or scrambled control in SKOV3 cells expressing extracellular flux analyzer GFP were placed on the top of jellified collagen matrix The mitochondrial oxygen consumption rate (OCR) in (Rat tail collagen1, 1 mg/ml). The invasiveness analyzed the MDA-MB-231 and SKOV3 cells was assessed by using after 24 h to 10 days. a Seahorse Bioscience XFe96 analyzer (Massachusetts, Karedath et al. BMC Cancer (2019) 19:565 Page 5 of 17 USA). Seahorse Bioscience XF Cell Mito Stress Test band of expected product size in PCR assay indicating assay kit was used for the study. In this assay, subse- the presence of the circular junction. The backsplice quent additions of the ATP synthase inhibitor oligo- junctional sequence was confirmed by Sanger sequen- mycin, the mitochondrial uncoupler carbonyl cyanide cing. The divergent primers, with respect to the genomic 4-(trifluoromethoxy) phenylhydrazone (FCCP) and the sequence, only amplified when cDNA was used as a complex I + II inhibitors rotenone + antimycin A were template. The same primers did not produce a product injected into the assay medium containing cells with from genomic DNA (gDNA). Conversely, PCR using different treatments as per manufactures protocol. 10, convergent primers with respect to genomic sequence 000 cells were seeded on assay plate and the assay could amplify both cDNA and gDNA templates from was completed in 48 h after the transfection. ANKRD12 gene (Fig. 2b). This strongly indicates that a head-to-tail backsplicing junction in circANKRD12 only Statistical analysis exists in the RNA form. Statistical analysis in all the experiments is based on at To examine whether circANKRD12 is resistant to least three biological replicates and the error bars are exonucleases, we treated the total RNA extracted from drawn with the standard error of means (SEM). The p the SKOV3 cell line with RNase R. This exoribonuclease value is calculated by using 1-tailed student’s T test. enzyme digests all linear RNA forms with a 3′ single- stranded region of greater than 7 nucleotides [26]. As Data access circRNAs are devoid of any 3′ single strand overhangs, The RNA-seq data for cell-lines has been submitted to they are expected to show resistance to digestion by RNase Sequence Read Archive (NCBI SRA) under Bio project R. Indeed, circANKRD12 was resistant to RNase R diges- accession number PRJNA526399. tion compared to linear forms of ANKRD12, HIF1 alpha, SRA records will be accessible with the following link: and GAPDH. Resistance to digestion with RNase R https://www.ncbi.nlm.nih.gov/sra/PRJNA526399 exonuclease confirmed that the circANKRD12 is a stable circularized transcript (Fig. 2c). Further, cDNA created by Results priming with oligo (dT) primers failed to produce any Characterization of circANKRD12 in cancer cells amplification products for circANKRD12 in the PCR assay We characterized the circular RNA expression using confirming the circular form does not have a typical polyA RNA sequencing (RNA-seq) of total RNA in ovarian tail as do linear mRNAs. On the other hand, priming with cancer cell lines (SKOV3 and CAOV3) undergoing random hexamers – which can also amplify non-poly- Epithelial to Mesenchymal Transition. Using our cir- adenylated RNAs – resulted in distinct PCR bands for the cRNA detection pipeline [21], we identified a total of ~ backsplice junction (Fig. 2d). The PCR and qRT-PCR 18,700 circRNAs in these cell lines. The circRNA from analysis of nuclear and cytoplasmic fractions of RNA dem- the ANKRD12 gene having a backsplice junction onstrated that ANKRD12 circRNA is predominantly local- between exons 8 and 2 (circANKRD12) was one of the ized in the cytoplasm (Additional file 1: Figure S1:Fig. 2a, highly abundant circular RNAs in both SKOV3 and b). The purity of cytoplasmic and nuclear fractions was CAOV3 cell lines. The circANKRD12 originates from confirmed by amplifying the fractions using nuclear and 18p11.22 chromosome locus with the backsplice junc- cytoplasmic specific markers (Additional file 1:FigureS1: tion forming between exons located ~ 39.3 kb apart (Fig. Fig a, b, Additional file 3:S2:8). 2a). By carefully designing two sets of divergent primer We performed Real-time PCR analysis on 12 different pairs in Exon2 and Exon8 and Sanger sequencing of the cell lines from breast, ovarian and lung cancer and amplified products, we were able to identify two circular normal breast and lung to assess the cell-type specific isoforms of 286 bp and 925 bp lengths sharing the same expression of circANKRD12 (Additional file 1: Figure backsplice junction (Additional file 3: S2:1–3), Add- S1:Fig. 2c). Most of these cell lines show a high abun- itional file 4: Table S2). The first isoform of 286 bp is dance of both ANKRD12 circRNA and mRNA. Ovarian comprised of only two exons (Exon2 and Exon8), while cancer cell lines show a higher abundance of ANKRD12 the 925 bp isoform is comprised of 6 exons (Exons 2,3,5, circRNA compared to breast and lung cancer cell lines. 6,7 and 8). All subsequent functional studies do not dis- tinguish between these two circular isoforms as they tar- siRNA-mediated knockdown of circANKRD12 is highly geted the junction sequence common to both circular specific forms but distinct vis-à-vis linear forms. To investigate the role of circANKRD12, we designed Multiple validation experiments were used to confirm siRNAs to target the backsplice junction (Fig. 1a) and circANKRD12 expression (Fig. 2b-d). Divergent primers transfected multiple cancer cell lines to induce siRNA- were designed to amplify the backsplice exon junction. mediated knockdown of the circle while leaving the lin- As expected, each primer pair produced a single distinct ear RNA unaffected. The circRNA specific siRNA was Karedath et al. BMC Cancer (2019) 19:565 Page 6 of 17 cd Fig. 2 Characterization of circANKRD12 in human cancer cells. a Transcript structure of the ANKRD12 gene. Two predominant circular isoforms (286 bp and 925 bp) were detected sharing the same backsplice junction between Exon8 and Exon2 as represented by the dashed line. The backsplice junction was sequenced for validation. b Divergent primers designed to detect the circANKRD12 backsplice junction amplified cDNA but not genomic DNA (4 pairs of different primers were used to amplify the circular junction). Linear ANKRD12 amplicon was detected in both cDNA and gDNA. M represents 100 bp ladder. c qRT-PCR for relative abundance of circANKRD12 and linear ANKRD12, GAPDH and HIF1 alpha in SKOV3 cells treated with RNase R d Backsplice junction was detected in cDNA synthesized using random hexamer but not oligoDT primed RNA. The data in c is means with error bars representing standard error of the mean (SEM) from three experiments; **P < 0.01 (Student’s t-test) designed against the backsplice junction spanning exons in four cell-line transfections (Fig. 1b). Using two 2 and 8 of the gene. We observed high knockdown effi- siRNA constructs against the circANKRD12, we ciency of greater than 90% of the circular junction when confirmed that knockdown of the circular RNA is using the siRNA versus the control (scrambled siRNA) specific and has no significant effect on the linear Karedath et al. BMC Cancer (2019) 19:565 Page 7 of 17 mRNA expression (Fig. 1c). The two different siRNAs estrogen mediated S phase entry through downregulat- had similar results reducing the likelihood that the ing cyclin D1 and cyclin A. Activation of IL8 signaling observed effects are due to off-target knockdown. We and inflamasome pathways, closely associated with im- also designed siRNAs targeting exon9 and exon7 of mune modulation is also predicted. IL8 signaling path- ANKRD12 gene to knockdown the linear mRNA. In- way upregulates IL8, IRAK, ICAM-1, COX-2, and VEGF deed, the siRNA designed against exon 9, which lies genes there by upregulating angiogenesis, inflammation, outside the circANKRD12 locus was successful at and inhibits cell proliferation by downregulating cyclin knocking down the mRNA and exhibits no effect on D1. A heatmap of differentially expressed genes from circANKRD12 levels (Fig. 1d).However,the siRNA RNAseq analysis shows consistent gene expression designed on Exon7 which is shared between mRNA changes in circANKRD12 silenced cells in cancer cell and one of the circANKRD12 isoforms shows a re- lines (Fig. 4a). The differentially regulated genes cyclin markable reduction in mRNA levels as well a minimal D1 (CCND1), CBX5, STAU1, and AK4 were found to be but significant reduction in circANKRD12 levels. We consistently downregulated in the circANKRD12 knock- investigated the specificity of siRNA constructs de- down across multiple cell lines (Fig. 4a). Using qRT- signed against circANKRD12, Exon7, and Exon9 PCR, we validated the differential expression of these through a series of knockdown experiments explained in selected genes in SKOV3 cells (Fig. 4c). Cyclin D1 Additional file 4: Supplementary file S2:4–7). (CCND1) was among the genes that showed consistent downregulation in the circANKRD12 knockdown across Silencing of circANKRD12 changes molecular phenotypes multiple cell lines (Fig. 4a). To ensure that the effect on of ovarian, breast and lung cancer cells CCND1 is related to circANKRD12 knockdown rather RNA-sequencing was performed in triplicate for two ovar- than off-target effects we compared the knockdown ian cancer cell lines (SKOV3, OVCAR3), breast (MDA- effects in SKOV3 and LL24 cells (Fig. 4b). Unlike SKOV3 MB-231) and a lung cancer cell line (NCI-H226) for both cells, the circANKRD12 level is low in LL24 cell line, at- scrambled control and siRNA targeting circANKRD12. tempts at knocking down circANKRD12 should not affect Table 1 gives the number of differentially expressed genes the cyclin D1 level in the cell line. Indeed, transfection of in circANKRD12 knockdown samples (Additional file 5: the siRNA against circANKRD12 in LL24 showed a reduc- Supplementary file S1) at a false discovery rate of less than tion of the already low levels of circANKRD12 but no ef- 5% with at least 1.5-fold changes in expression. fect on cyclin D1 expression, suggesting the siRNA effect The canonical pathways analysis by Ingenuity’sIPA on CCND1 is likely mediated through circANKRD12. Be- toolkit (IPA®, QIAGEN Redwood City, (https://www.qia- cause cell cycle regulation seems to be one of the most genbioinformatics.com/products/ingenuity-pathway-ana- deregulated pathways in circANKRD12 silenced cells, we lysis/) revealed an enrichment of differentially regulated decided to follow it up with further functional screening genes involved in cell cycle regulation, invasion, migra- using cell based phenotypic assays. tion, and interferon signaling pathways (Fig. 3, Additional file 2: Figure S2, Additional file 4: Table S3). siRNA mediated silencing of circANKRD12 increases We observed an upregulation of inflammatory pathways migration and invasion in ovarian cancer cells such as tumor necrosis pathway, NFkB pathway, inter- Wound healing assays show that silencing of cir- feron signaling and down regulation of the cell cycle ANKRD12 increased cell migration (Fig. 5a, b, pathway (Additional file 3: S2:15). Silencing of cir- Additional file 3: S2:10), in SKOV3 cells after 24 h and cANKRD12 affects cellular bioenergetics by downregula- 48 h. compared to scrambled control. This is confirmed tion of key genes involved in oxidative phosphorylation, again by cell migration assays in synchronized (Thymi- AMPK pathway and cell metabolism (Additional file 4: dine incorporation) cells, which show an increased mi- Table S4). IPA predicts a significant activation of inter- gration rate for circANKRD12 silenced SKOV3 cells feron signaling pathway through upregulation of STAT1, (Fig. 5c). Matrigel invasion (inserts coated with matrigel) STAT2, IFT1, MX1genes, and downregulation of and migration analysis using Boyden chamber shows cir- cANKRD12 silenced cells undergo significant increase in the invasion and migration compared to scrambled con- Table 1 Number differentially up and downregulated genes for trol (Additional file 3: S2:10). four cancer cell lines in siRNA-mediated knockdown of circANKRD12 Silencing of circANKRD12 decreases cell proliferation Si-circANKRD12 vs. Scrambled Knockdown of circANKRD12 and ANKRD12 mRNA SKOV3231 OVCAR3 MDA-MB-231 NCI-H226 significantly reduces cell proliferation (Fig. 5d, Additional Upregulated genes 25 104 237 168 file 3: S2:9). The results of MTS and ATP assays show a Downregulated genes 9 71 179 3 significant reduction in cell proliferation in circANKRD12 Karedath et al. BMC Cancer (2019) 19:565 Page 8 of 17 Fig. 3 Gene network analysis of RNA-Seq data from MDA-MB-231 and OVCAR3 cells. a Gene network analysis using ingenuity pathway analysis in MDA-MB-231 cells silenced for circANKRD12 shows upregulation of Interferon, JAK/STAT, IL-12 signaling pathways. b Gene network analysis using ingenuity pathway analysis in OVCAR3 cells silenced with circANKRD12 shows downregulation of Cyclins and Cell Cycle regulation pathways silenced cells compared to scrambled control (Fig. 5d, e. SKOV3 cells (Fig. 5f). These results indicate that knock- The Trypan blue exclusion assays show the silencing of downs of both circular and linear RNA forms of circANKRD12 is not affecting the cell viability signifi- ANKRD12 gene are capable of inducing strong pheno- cantly. However, silencing of ANKRD12 mRNA signifi- typic changes and modulate the growth or survival of can- cantly reduces both proliferation and cell viability in cer cells. Karedath et al. BMC Cancer (2019) 19:565 Page 9 of 17 Fig. 4 Silencing of circANKRD12 affects gene expression changes in cancer cells. a Heat map of differentially expressed genes from RNAseq analysis of SKOV3, NCI-H226, MDAMB 231 and OVCAR3 cells silenced with si-circANKRD12 compared to scrambled control. b qRT-PCR validation of expression level of cyclin D1 in circANKRD12 silenced SKOV3 and LL24 cell lines. In SKOV3 cells, circANKRD12 knockdown leads to a significant downregulation of Cyclin D1 (P = 7.8E-04), while as in LL24 Cyclin D1 shows a minimal non-significant change in expression (P = 0.18). c qRT-PCR validation of some selected genes differentially expressed in SKOV3 cells silenced with circANKRD12. Data in b–c are the means with error bars indicating standard error of the mean of three experiments. **P < 0.01 (Student’s t-test) Silencing of cirANKRD12 in 3D tumor models induces a [27], we investigated the effects of circANKRD12 knock- phenotypic switch from highly proliferative to an invasive down in 3D culture experiments. The 3D anchorage in- phenotype dependent growth of siRNA transfected SKOV3 cells Since 3D culture is known to have better cell-to-cell showed smaller, dense aggregates associated with an interactions and more closely mimics tumors in vivo invasive phenotype (Additional file 3:S2:12). The3D Karedath et al. BMC Cancer (2019) 19:565 Page 10 of 17 ab cd ef Fig. 5 Silencing of circANKRD12 results in an invasive phenotype with low proliferation rate. a Representative Wound healing migration assay image of circANKRD12 silenced SKOV3 cells and its scrambled control at 0 h and 24 h. b Bar diagram depicting the reduction in wound width after migration in scrambled and circANKRD12 silenced cells (average of 15 wounds from three different biological replicates). c Migration of circANKRD12 silenced SKOV3 cells after cell cycle synchronization with thymidine. d The proliferation of SKOV3 cells transfected with siRNA against circANKRD12 and assessed using MTS cell proliferation assay kit at the indicated hours and data shown as % proliferation compared to control. e Cell titer glow assay shows relative luminescence rate of ATP production in circANKRD12 silenced SKOV3 cells compared to scrambled control. f Cell viability is measured by using trypan blue exclusion assay. In circANKRD12 knockdown, the viability is reduced by roughly 2% while it is reduced by ~ 34% in ANKRD12-mRNA knockdown. (Data in b–f are the means with error bars indicating standard error of the mean of three experiments. **P < 0.01 (Student’s t-test) Karedath et al. BMC Cancer (2019) 19:565 Page 11 of 17 organotypic models were efficiently silenced with cir- Table S4). In order to determine whether the altered cANKRD12 siRNA with 74% knockdown efficacy (Fig. 6a). gene expression of these pathways translates to changes This reductioninspheroidsize in circANKRD12-silenced in basic metabolism, we analyzed metabolic phenotypes SKOV3 organotypic models resultsinanalterationof of circANKRD12 silenced MDA-MB-231 and SKOV3 phenotype from less invasive to a more invasive one (Fig. cells using Seahorse extracellular flux analyzer. The oxy- 6b). Cell proliferation MTS assay shows a significant reduc- gen consumption ratio (OCR) and ATP linked OCR ana- tion in cell growth within a time period of 24 h and 48 h lysis shows that knockdown of circANKRD12 decreases after knockdown of ANKRD12 circRNA (Fig. 6c). Collagen oxidative phosphorylation (OXPHOS) of MDAMB231 invasion of the circANKRD12 silenced spheroids shows in- cells and SKOV3 cells (Fig. 8c-e). Previous reports have creased invasion through the collagen gel with highly mo- suggested that high invasive potential of cancer cells is tile cells after 10 days of transfection (Fig. 6d). These results negatively correlated to high energetic cancer phenotype indicate the phenotypic switching from a highly prolifera- [29, 30]. These results thus indicate that circANKRD12 tive phenotype to a less proliferative phenotype with high silencing can induce phenotypic switching between invasion potential. highly proliferative cells to highly invasive cells through cyclin D1 deregulation and shifting the oncobioener- Cyclin D1 is down-regulated in circANKRD12 silenced cells getics to a low energy phenotype to facilitate invasion. and involved in phenotypic switching by facilitating G1 arrest circANKRD12 silencing results in cyclin D1 down- Discussion regulation and subsequent invasion and reduction in prolif- Circular RNAs are attracting greater attention in RNA eration in ovarian cancer SKOV3 cells. Both real-time PCR biology as there is growing evidence for their role in and RNA-sequencing showed down-regulation of cyclin gene expression regulation. Even though a large reper- D1 in circANKRD12 silenced cells. The qRT-PCR analysis toire of circRNAs has been identified in different organ- shows cyclin D1 expression is downregulated at least until isms, tissues, diseases, and developmental conditions, 48 h after transfection with the siRNA (Fig. 7a). This is only a few have been evaluated for their role in cellular also supported by the Western blot analysis, which shows functions [31–34]. Functional screening using siRNA a reduction in cyclin D1 protein level (Fig. 7b). Only cyclin targeting is difficult for circRNAs compared to other D1 shows differential regulation at the protein level in cir- RNA types as the choice of the target region is limited cANKRD12 silenced cells compared to other cyclin mem- to a few base pairs with specificity only at the backsplice bers including cyclin E1, pB1, and D2 (Additional file 3: junction. This constraint severely restricts the scale of S2:13). To estimate the duration of the RNAi effect initi- circRNAs suitable for further functional studies using ated by transfecting siRNA, we performed a longevity siRNA-mediated knockdown approaches. assay of circANKRD12 siRNA (si-circANKRD12). The In this study, we identified and characterized circular longevity, initially stable for six days (Additional file 3:S2: RNA isoforms from ANKRD12 gene (circANKRD12) 14) was checked on a cell doubling time basis estimated that are abundantly expressed in ovarian and breast to be 48 h (see Methods). The si-circANKRD12 exhibits cancer cells. A recent study identified stable levels of cir- an extended longevity period in SKOV3 cells, approaching cANKRD12 (exon 2–8) in young and old erythrocyte nine doublings (18 days). The reduction in the level of cyc- cells [35]. We show that circANKRD12 is a stable circu- lin D1 is on par with the observed silencing efficiency and larized transcript, resistant to RNase R digestion and longevity of siRNA (Fig. 7c). After the 10th doubling, the devoid of a polyA tail. In contrast to the linear mRNA effect on cell proliferation induced by silencing of cir- form, which is predominantly nuclear, circANKRD12 is cANKRD12 becomes insignificant (Fig. 7d), suggesting localized in the cytoplasm. that the knockdown efficiency is lost due to multiple pas- Differential gene expression analysis by RNA-seq of sages of cells. As cyclin D1 is involved in cell cycle pro- circANKRD12 silenced cells revealed that silencing may gression [28], cell cycle analysis of circANKRD12 accelerate cancer cell invasion, migration, and cellular knockdowns was conducted. Cell cycle analysis using movement by regulating a cascade of genes involved in FACS shows there is a significant G0/G1 cell cycle ar- these processes (Additional file 4: Table S3, 5). Inter- rest in SKOV3 cells silenced with circANKRD12 com- feron signaling and cell cycle checkpoint regulation by pared to scrambled control (Fig. 8a, b). G1/S transition are the top networks deregulated in circANKRD12-silenced cells. IPA analysis shows an in- Silencing of circANKRD12 affects oxygen consumption crease in pathways related to the invasion of cancer cells ratio (OCR) in SKOV3 and MDAMB231 cells with significant p-value and z score (Additional file 2: As AMPK signaling and other metabolic pathways are Figure S2: Fig. 2a,b,c) in MDA-MB-231 cells. The key affected in circANKRD12 knockdown (Additional file 4: genes upregulated in circANKRD12 silenced cells are Karedath et al. BMC Cancer (2019) 19:565 Page 12 of 17 ab Fig. 6 Silencing of circANKRD12 in 3D spheroid models of SKOV3 cells shows an invasive phenotype. a Knockdown efficacy of circANKRD12 in 3D cultures. siRNA directed against circANKRD12 downregulates circular RNA but shows no downregulation on mRNA of ANKRD12 gene in SKOV3 organotypic models (b) spheroid area is measured in anchorage-independent 3D spheroids of SKOV3 cell lines, average of 100 spheroids in 3 independent experiments were measured. c MTS assay shows cell viability in 3D organotypic models is reduced at 24 and 48 h of transfection. d Collagen invasion assay of spheroids 24,48 h and 10 days. circANKRD12 silenced cells were able to invade through the thick collagen matrix and circANKRD12 silenced cells shows an invasive phenotype. Data in a–c are the means with error bars representing standard error of the mean of three experiments. **P < 0.01 (Student’s t-test) Karedath et al. BMC Cancer (2019) 19:565 Page 13 of 17 ab Fig. 7 Longevity of circANKRD12 silencing in SKOV3 cells a qRT-PCR analysis of circANKRD12 knockdown efficiency at two different time points (24 h, 48 h). b Western blot analysis of cyclin D1 expression in circANKRD12 and ANKRD12 silenced SKOV3 cells at 48 h. c Longevity analysis of circANKRD12 silenced SKOV3 cells. qRT-PCR analysis for the longevity of siRNA transfection based on knockdown efficiency of circANKRD12 in each doubling time (48 h). The figure shows the Knockdown efficiency of circANKRD12 on cyclin D1 expression. qRT-PCR of the expression pattern of cyclin D1 from 1st to 9th doubling time d MTS cell proliferation assay of circANKRD12 knockdown cells after 10th doubling. Data in a is the means with error bars representing standard error of the mean. **P < 0.01 (Student’s t-test) Karedath et al. BMC Cancer (2019) 19:565 Page 14 of 17 ab cd Fig. 8 circANKRD12 silenced cells have an arrested G1stage and low OXPHOS compared to scrambled control. a Cell cycle analysis of SKOV3 cells silenced with circANKRD12 at 24 h. b Bar chart representation of % of cells in G0/G1 stage. Error bars represent standard error of the mean from three experimental replicates. c,d Oxygen consumption ratio and ATP production analyzed by seahorse extracellular flux analyzer in MDAMB231 cells silenced with si-circANKRD12, si-linear ANKRD12 and its control. e Oxygen consumption ratio analyzed by seahorse by extracellular flux analysis in SKOV3 cells. Data in c, d, and e are normalized with total protein(ug/ul) STAT-1, MX1, NFkB, MUC4, and SMAD3. Cell based reducing cell proliferation and increasing invasion. Pre- assays live cell migration, invasion and wound healing vious studies have reported a reduction in cyclin D1 assays also shows increased invasion of circANKRD12 levels, and G0/G1 cell cycle stage arrest leads to an in- silenced cells in SKOV3 cell line. crease in migratory activities of MDA-MB-231 breast Cell proliferation was significantly arrested without cancer cells [36]. Consistent with these findings, our re- any change in cell viability in circANKRD12 silenced sults of cell cycle analysis also show a substantial G0/G1 SKOV3 cells. Cyclin D1, a consistently affected gene by arrest with increased migration and invasion in cir- circANKRD12 silencing, is down regulated thereby cANKRD12 knockdown cells. The migration and Karedath et al. BMC Cancer (2019) 19:565 Page 15 of 17 invasion assays significantly correlate with gene expres- abundant circRNA in human ovarian and breast cancer sion analysis and indicate an augmented cell migration cells. Our results suggest that circANKRD12 could be and invasion rate. The 3D anchorage independent orga- involved in a diverse set of functions ranging from cell notypic tumor models of SKOV3 show similar patterns cycle arrest, tumor invasion to immune modulation. Ma- of phenotypic characterization where invasion through nipulating the levels of circANKRD12 can regulate mo- collagen is increased upon silencing of circANKRD12. lecular functions by altering different signaling pathways There is a strong phenotypic alteration after silencing and modifies the phenotype of the cells. The distinctive circANKRD12 in the ovarian cancer cells in both 2D change from a proliferative to a more invasive phenotype and 3D culture conditions suggesting that circANKRD12 by altering circANKRD12 levels could lead to a future is important in regulating proliferation, invasion, and circRNA based therapeutic intervention in cancer. migration. Reduction in cell proliferation and regulation of Cyclin D1 expression is confirmed by silencing cir- Additional files cANKRD12 by another construct of siRNA in SKOV3 Additional file 1: Figure S1. (a) The abundance of circANKRD12 and cells (Additional file 3: S2:16a, b). circRNAs are involved ANKRD12 linear RNA. Semi qRT-PCR data indicating the abundance of in regulating cellular movement. Indeed, circRNA (F-cir- circANKRD12 and ANKRD12 mRNA in the cytoplasmic and nuclear fractions cEA) produced from the EML4-ALK fusion gene which of SKOV3 cells, nuclear and cytoplasmic purity markers were also assessed using nuclear specific marker 7SK and cytoplasmic specific marker CYTB. The is independent of the EML4-ALK linear transcript and gel picture represents PCR amplified products of circular RNAs with three the fusion protein, can promote migration and invasion, different sets of primers. (b) Real-time data shows the abundance of thus contributing to tumor metastasis [37, 38]. circANKRD12 and ANKRD12 mRNA in the cytoplasm and nucleus. (c) The abundance of circANKRD12 and ANKRD12 mRNA in a panel of cancer cells. The Ankyrin Repeat Domain family of genes can act (PPTX 2728 kb) (PPTX 2750 kb) as putative tumor suppressors via p53 mediated feedback Additional file 2: Figure S2. Figure a, b, c Shows functional predicted or through recruiting histone deacetylases (HDACs) to networks from RNAseq data of circANKRD12 silenced MDA-MB-231 cells the p160 coactivator to repress transcriptional activities induces inflammatory immune responses and cancer cell invasion with activation Z score and p values. (PPTX 6928 kb) (PPTX 6949 kb) [18, 19, 39]. A clinical study of gene expression of Additional file 3: S2. File represents additional figures for validating the ANKRD12 in colorectal cancer revealed that low circANKRD12 in cell lines. (PPTX 17563 kb) (PPTX 17572 kb) ANKRD12 expression is correlated with overall poor Additional file 4: Tables S1-S5. Tables representing primers used for survival and liver metastasis of CRC patients [20]. different gene expression studies, siRNAs and pathways involved in We observed that the silencing of ANKRD12 mRNA circANKRD12 gene knockdown condition. (PPTX 3253 kb) (PPTX 3268 kb) reduces cell proliferation, induces cell death and down Additional file 5: Supplementary file S1. List of genes differentially expressed in circANKRD12 silenced cells compared to control in different regulates cyclin D1. On the contrary, silencing of cell lines. (XLSX 160 kb) circANKRD12 arrests the cell cycle progression and Additional file 6: Table S6. List of microRNAs that can target increases tumor invasion without significantly affecting circANKRD12 and CyclinD1. (XLSX 9 kb) cell viability. The knockdown of circANKRD12 can change the oncobioenegetics as its shift from a higher Abbreviations OXPHOS to a lower OXPHOPS phenotype which is 3D culture: Three-Dimensional cell culture; circRNA: Circular RNA; FDR: False discovery rate; OCR: Oxygen Consumption Ratio; OXPHOS: Oxidative highly invasive. circANKRD12 may act as competing Phosphorylation; qRT-PCR: Quantitative Real Time PCR; RNAseq: RNA endogenous RNA (ceRNA) to regulate a circRNA- sequencing; siRNA: Small (or short) interfering RNA miRNA-mRNA network. Our insilco analysis shows that Acknowledgments cyclin D1 and circANKRD12 have shared binding sites for The work was supported by grants from Basic Medical Research Program several different microRNAs (Additional file 6: Table S6). (BMRP) grant from Qatar Foundation to WCM-Q. Thus, circANKRD12 could act as a microRNA sponge to We thank Shameem Yunuskunju form Genomics Core at WCM-Q for RNA -Seq data handling. We thank Dr. Anna Halama from Dr. Karsten Suhre’s Bio- regulate cyclin D1 levels. Our preliminary analysis shows informatics Core lab at WCM-Q for providing the LL24, NCI-H226, T47D, overexpression of hsa-miR-4768-5p reduces the level of HCC2935 cells. We also thank Ms. Aleksandra M. Liberska from Microscopy Cyclin D1 by 20% (hsa-miR-4768-5p has common binding Core at WCM-Q for helping with Flow Cytometry. sites for circANKRD12 and Cyclin D1) (Data not shown). Funding Thus, circANKRD12 could act as a microRNA sponge to This study was supported by Weill Cornell Medicine and Qatar Foundation regulate cyclin D1 levels. We also observed that (BMRP 1 - Malek Pilot FY17). The funders had no role in the design of the study,data analysis, interpretation of data and writing the manuscript. circANKRD12 contains some putative open reading frames (ORFs) and therefore its translation through in- Availability of data and materials ternal ribosomal entry sites cannot be ruled out [40, 41]. The datasets supporting the conclusions of this article are included in this article and the Supplementary Data. Conclusion Authors’ contributions In conclusion, our study provides the molecular, pheno- JM, TK, and IA designed the research; TK, WAA, FAD, SA, and SS performed typic and metabolic characterization of one of the most cellular experiments. TK and IA contributed equally to this work. IAA and Karedath et al. BMC Cancer (2019) 19:565 Page 16 of 17 YAM performed RNA-seq and IA and JM performed bioinformatic analysis; 15. Xia W, Qiu M, Chen R, Wang S, Leng X, Wang J, et al. Circular RNA has_circ_ AR provided the cell lines and helped with experimental design; TK and IA 0067934 is upregulated in esophageal squamous cell carcinoma and promoted wrote the manuscript. All authors read and approved the final manuscript. proliferation. Sci Rep. 2016;6:35576. https://doi.org/10.1038/srep35576. 16. Li F, Zhang L, Li W, Deng J, Zheng J, An M, et al. Circular RNA ITCH has Ethics approval and consent to participate inhibitory effect on ESCC by suppressing the Wnt/$β$-catenin pathway. Not applicable for the present study. All experiments were conducted in Oncotarget. 2015;6:6001–13. https://doi.org/10.18632/oncotarget.3469. cancer cell lines, no human or animal subject is involved in the study. 17. Dong Y, He D, Peng Z, Peng W, Shi W, Wang J, et al. Circular RNAs in cancer: an emerging key player. J Hematol Oncol. 2017. https://doi.org/10. Consent for publication 1186/s13045-016-0370-2. Not applicable. 18. Li C-W, Dinh GK, Zhang A, Chen JD. Ankyrin repeats-containing cofactors interact with ADA3 and modulate its co-activator function. Biochem J. 2008. Competing interests https://doi.org/10.1042/BJ20071484. The authors declare that they have no competing interests. 19. Neilsen PM, Cheney KM, Li C-W, Chen JD, Cawrse JE, Schulz RB, et al. Identification of ANKRD11 as a p53 coactivator. J Cell Sci. 2008;121 https:// www.ncbi.nlm.nih.gov/pubmed/?term=Identification+of+ANKRD11+as+a+ Publisher’sNote p53+coactivator.+J+Cell+Sci.+2008. Accessed 11 Apr 2017. Springer Nature remains neutral with regard to jurisdictional claims in 20. Bai R, Li D, Shi Z, Fang X, Ge W, Zheng S. Clinical significance of Ankyrin published maps and institutional affiliations. repeat domain 12 expression in colorectal cancer. J Exp Clin Cancer Res. https://doi.org/10.1186/1756-9966-32-35. Author details 21. Ahmed I, Karedath T, Andrews SS, Al-Azwani IK, Ali Mohamoud Y, Querleu Department of Genetic Medicine, Weill Cornell Medicine-Qatar, Education D, et al. Altered expression pattern of circular RNAs in primary and City, Qatar Foundation, P.O. Box No, 24144 Doha, Qatar. Genomics Core, metastatic sites of epithelial ovarian carcinoma. Oncotarget. 2016. https:// Weill Cornell Medicine-Qatar, Education City, Qatar Foundation, Doha, Qatar. doi.org/10.18632/oncotarget.8917. Department of Pharmacology, Weill Cornell Medicine-Qatar, Education City, 22. Zheng Q, Bao C, Guo W, Li S, Chen J, Chen B, et al. Circular RNA profiling Qatar Foundation, Doha, Qatar. Stem Cell and Microenvironment reveals an abundant circHIPK3 that regulates cell growth by sponging multiple Laboratory, Weill Cornell Medicine-Qatar, Education City, Qatar Foundation, miRNAs. Nat Commun. 2016;7:11215. https://doi.org/10.1038/ncomms11215. Doha, Qatar. 23. Liang C-C, Park AY, Guan J-L. In vitro scratch assay: a convenient and inexpensive method for analysis of cell migration in vitro. Nat Protoc. 2007; Received: 25 July 2018 Accepted: 23 April 2019 2:329–33. https://doi.org/10.1038/nprot.2007.30. 24. Luo J, Wang X, Xia Z, Yang L, Ding Z, Chen S, et al. Transcriptional factor specificity protein 1 (SP1) promotes the proliferation of glioma cells by up- References regulating midkine (MDK). Mol Biol Cell. 2015;26:430–9. https://doi.org/10. 1. Jeck WR, Sharpless NE. Detecting and characterizing circular RNAs. Nat 1091/mbc.E14-10-1443. Biotechnol. 2014;32. https://doi.org/10.1038/nbt.2890. 25. Samuel SM, Ghosh S, Majeed Y, Arunachalam G, Emara MM, Ding H, et al. 2. Cocquerelle C, Mascrez B, Hétuin D, Bailleul B. Mis-splicing yields circular Metformin represses glucose starvation induced autophagic response in RNA molecules. FASEB J. 1993;7:155–60 https://www.fasebj.org/doi/abs/10. microvascular endothelial cells and promotes cell death. Biochem 1096/fasebj.7.1.7678559. Accessed 1 May 2017. Pharmacol. 2017;132:118–32. https://doi.org/10.1016/J.BCP.2017.03.001. 3. Wang PL, Bao Y, Yee M-C, Barrett SP, Hogan GJ, Olsen MN, et al. Circular 26. Vincent HA, Deutscher MP. Substrate recognition and catalysis by the RNA is expressed across the eukaryotic tree of life. PLoS One. 2014;9:e90859. exoribonuclease RNase R. J Biol Chem. 2006;281:29769–75. https://doi.org/ https://doi.org/10.1371/journal.pone.0090859. 10.1074/jbc.M606744200. 4. Dang Y, Yan L, Hu B, Fan X, Ren Y, Li R, et al. Tracing the expression of 27. Pampaloni F, Reynaud EG, Stelzer EHK. The third dimension bridges the gap circular RNAs in human pre-implantation embryos. Genome Biol. 2016. between cell culture and live tissue. Nat Rev Mol Cell Biol. 2007;8:839–45. https://doi.org/10.1186/s13059-016-0991-3. https://doi.org/10.1038/nrm2236. 5. Hansen TB, Jensen TI, Clausen BH. Natural RNA circles function as efficient 28. Baldin V, Lukas J, Marcote MJ, Pagano M, Draetta G. Cyclin D1 is a nuclear microRNA sponges. Nature. 2013;495. https://doi.org/10.1038/nature11993. protein required for cell cycle progression in G1. Genes Dev. 1993;7:812–21. 6. Ashwal-Fluss R, Meyer M, Pamudurti NR, Ivanov A, Bartok O, Hanan M, et al. https://doi.org/10.1101/GAD.7.5.812. circRNA biogenesis competes with pre-mRNA splicing. Mol Cell. 2014;56. 29. Yang L, Moss T, Mangala LS, Marini J, Zhao H, Wahlig S, et al. Metabolic https://doi.org/10.1016/j.molcel.2014.08.019. shifts toward glutamine regulate tumor growth, invasion and bioenergetics 7. Li J, Yang J, Zhou P, et al. Circular RNAs in cancer: novel insights into origins, in ovarian cancer. Mol Syst Biol. 2014;10. https://doi.org/10.1002/msb. properties, functions and implications. Am J Cancer Res. 2015;5(2):472–80. 30. Vayalil PK, Landar A. Mitochondrial oncobioenergetic index: a potential 8. He J, Xie Q, Xu H, Li J, Li Y. Circular RNAs and cancer. Cancer Lett. 2017;396: biomarker to predict progression from indolent to aggressive prostate cancer. 138–44. https://doi.org/10.1016/j.canlet.2017.03.027. Oncotarget. 2015;6:43065–80. https://doi.org/10.18632/oncotarget.5487. 9. Qu S, Liu Z, Yang X, Zhou J, Yu H, Zhang R, et al. The emerging functions 31. Guarnerio J, Bezzi M, Jeong JC, Paffenholz SV, Berry K, Naldini MM, et al. and roles of circular RNAs in cancer. Cancer Lett. 2018;414:301–9. https://doi. Oncogenic role of fusion-circRNAs derived from Cancer-associated chromosomal org/10.1016/J.CANLET.2017.11.022. translocations. Cell. 2016;165:289–302. https://doi.org/10.1016/j.cell.2016.03.020. 10. Li Z, Yanfang W, Li J, Jiang P, Peng T, Chen K, et al. Tumor-released 32. Du WW, Yang W, Chen Y, Wu Z-K, Foster FS, Yang Z, et al. Foxo3 circular exosomal circular RNA PDE8A promotes invasive growth via the miR-338/ RNA promotes cardiac senescence by modulating multiple factors MACC1/MET pathway in pancreatic cancer. Cancer Lett. 2018. https://doi. associated with stress and senescence responses. Eur Heart J. 2016:ehw001. org/10.1016/j.canlet.2018.04.035. https://doi.org/10.1093/eurheartj/ehw001. 11. Chen B, Huang S. Circular RNA: An emerging non-coding RNA as a 33. Huang G, Zhu H, Shi Y, Wu W, Cai H, Chen X, et al. Cir-ITCH plays an inhibitory regulator and biomarker in cancer. Cancer Lett. 2018;418:41–50. https://doi. role in colorectal Cancer by regulating the Wnt/$β$-catenin pathway. PLoS org/10.1016/j.canlet.2018.01.011. One. 2015;10:e0131225. https://doi.org/10.1371/journal.pone.0131225. 12. Meng X, Li X, Zhang P, Wang J, Zhou Y, Chen M. Circular RNA: an emerging 34. Dong W, Bi J, Liu H, Yan D, He Q, Zhou Q, et al. Circular RNA ACVR2A key player in RNA world. Brief Bioinform. 2016:bbw045. https://doi.org/10. suppresses bladder cancer cells proliferation and metastasis through miR- 1093/bib/bbw045. 626/EYA4 axis. https://doi.org/10.1186/s12943-019-1025-z. 13. Memczak S, Papavasileiou P, Peters O, Rajewsky N. Identification and characterization of circular RNAs as a new class of putative biomarkers in 35. Nicolet BP, Engels S, Aglialoro F, van den Akker E, von Lindern M, Wolkers human blood. PLoS One. 2015;10:e0141214. https://doi.org/10.1371/journal. MC. Circular RNA expression in human hematopoietic cells is widespread pone.0141214. and cell-type specific. Nucleic Acids Res. 2018;46:8168–80. https://doi.org/10. 14. Hang D, Zhou J, Qin N, Zhou W, Ma H, Jin G, et al. A novel plasma circular 1093/nar/gky721. RNA circFARSA is a potential biomarker for non-small cell lung cancer. 36. Lehn S, Tobin NP, Berglund P, Nilsson K, Sims AH, Jirström K, et al. Down- Cancer Med. 2018. https://doi.org/10.1002/cam4.1514. regulation of the oncogene cyclin D1 increases migratory capacity in breast Karedath et al. BMC Cancer (2019) 19:565 Page 17 of 17 cancer and is linked to unfavorable prognostic features. Am J Pathol. 2010; 177:2886–97. https://doi.org/10.2353/ajpath.2010.100303. 37. Tan S, Gou Q, Pu W, Guo C, Yang Y, Wu K, et al. Circular RNA F-circEA produced from EML4-ALK fusion gene as a novel liquid biopsy biomarker for non-small cell lung cancer. Cell Res. 2018;28:693–5. https://doi.org/10. 1038/s41422-018-0033-7. 38. Tan S, Sun D, Pu W, Gou Q, Guo C, Gong Y, et al. Circular RNA F-circEA-2a derived from EML4-ALK fusion gene promotes cell migration and invasion in non-small cell lung cancer. Mol Cancer. 2018;17:138. https://doi.org/10. 1186/s12943-018-0887-9. 39. Zhang A, Li C-W, Chen JD. Characterization of transcriptional regulatory domains of ankyrin repeat cofactor-1. Biochem Biophys Res Commun. 2007; 358:1034–40. https://doi.org/10.1016/j.bbrc.2007.05.017. 40. Legnini I, Di Timoteo G, Rossi F, Morlando M, Briganti F, Sthandier O, et al. Circ- ZNF609 is a circular RNA that can be translated and functions in Myogenesis. Mol Cell. 2017;66:22–37.e9. https://doi.org/10.1016/j.molcel.2017.02.017. 41. Pamudurti NR, Bartok O, Jens M, Ashwal-Fluss R, Stottmeister C, Ruhe L, et al. Translation of CircRNAs. Mol Cell. 2017;66:9–21.e7. https://doi.org/10. 1016/j.molcel.2017.02.021.

Journal

BMC Cancer – Springer Journals

Published: Jun 11, 2019

There are no references for this article.