Background
Breast cancer is the most common malignancy in women, with a rapidly increasing incidence [
1]. Triple-negative breast cancer (TNBC) accounts for approximately 15% of all breast cancers, with the highest recurrence, metastasis, and mortality rates and a lack of effective targeted treatment options [
1]. Characterized by the absence of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) expression, TNBCs are insensitive to endocrine therapies, molecular targeted therapies and current chemotherapeutics, thus increasing the difficulty of clinical treatment [
2,
3]. Although many reagents targeting signaling, such as phosphoinositide 3-kinase (PI3K), cyclin-dependent kinases (CDKs) and receptor tyrosine kinases (RTKs), have been approved by the Food and Drug Administration (FDA), their therapeutic efficacy is still disappointing [
4], and there is an urgent need to identify novel molecular markers for early diagnosis or targeted therapies for TNBCs.
Recently, several newly identified biomarkers, such as tumor-associated macrophages (TAMs) [
5], microRNAs (miRNAs) [
6,
7], and long noncoding RNAs (lncRNAs) [
8], have proven to have important prognostic value in TNBCs. Specifically, circular RNAs (circRNAs) have attracted our attention due to their key roles in human cancers, including TNBC. CircRNAs are closed single-stranded circular transcripts with no 5′ caps or 3′ poly(A) tails [
9]. CircRNAs have been reported to be distributed widely across species and to play crucial roles, such as miRNA sponges, RNA binding proteins (RBPs) and protein coding templates, in tumorigenesis and cancer progression. Given their unique biological structure and vital functions, looking for circRNAs with high specificity and sensitivity will provide a new opportunity for the early diagnosis, clinical treatment, and prognosis monitoring of TNBC [
10,
11]. To date, several TNBC-related circRNAs have been reported. CircSEPT9, induced by E2F1 and EIF4A3, promotes TNBC tumorigenesis and serves as a progression marker [
12]. Another circRNA, circHER2, encodes a polypeptide and resensitizes TNBC to pertuzumab-based treatment [
13]. Nevertheless, the roles of circRNAs in TNBC remain largely illusive.
In this study, we aimed to identify new TNBC biomarkers that predict the prognosis for clinical patient management and novel targets for precise therapy. We determined that circCD44 promotes TNBC progression via miR-502–5p/KRAS and IGF2BP2/C-Myc signaling and suggested that circCD44 is a potential therapeutic target for TNBCs.
Methods
Human samples
All samples were obtained from the Department of Breast and Thyroid Surgery of the First Affiliated Hospital of Sun Yat-sen University. Written and informed consent was obtained. The research was approved by the Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University.
RNA sequencing
RNA sequencing was applied as previously described [
13]. Briefly, total RNA was extracted, purified, and reverse transcribed. CircRNA junction data were downloaded from circBase (
http://www.circbase.org/). CircRNAs with a fold change > 2,
p < 0.01, FDR < 0.05 were identified as significantly differentially expressed circRNAs. The RNA seq data access number is NCBI SRP266211, totally, 1200 differentially expressed circRNAs were identified and the consistent up- or down- regulated circRNAs in 5 TNBCs were listed in Supplementary Table
4.
Cell lines
The human normal breast epithelium MCF-10A and human mammary cancer cell lines Hs578T, BT-549, MDA-MB-468 and MDA-MB-231 were obtained from the American Type Culture Collection (ATCC, Manassas, VA). Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM), minimum essential medium (MEM) and RPMI-1640 and supplied with 10% fetal bovine serum (Life Technologies, Grand Island, NY, USA), 1% penicillin G, and streptomycin.
Plasmid and stable cell line construction
The circCD44 plasmid was generated by chemical synthesis of the complete sequence of circCD44, and additional circularization promoter ALU sequences were added upstream and downstream [
14],the sequence was detailed as below: 5’TCTGACAACTGAACTGCTCTCGCCTTGAACCTGTTTTGGCACTAA AATAAAATCTGTTCAATTAACGAATTCTGAAATATGCTATCTTACAG----GTGAATATATTTTTTCTTGAGGATCCACTAATTTGGGATGATAACGCCAAAACAGGTTCAAGGCGAGAGCAGTTCAGTTGTCAGAA3’,the sequence of circCD44 was cloned between AG and GT in the blank. Mutant circCD44 was established by changing the interaction residue CAAGAA to GUUCUU, as shown in Fig.
4A. In Fig.
6F, circCD44-mut was generated by changing the interacting residues mentioned in Fig.
6E: U to A, A to U, C to G and G to C. The IGF2BP2 mutant allele was established by changing the interaction residues to IMKDNKHSDCDRQDT. The circCD44 shRNA-1 sequence was GAAGGATGGTCCAGGCAACTC, and the shRNA-2 sequence was AGAAGGATGGTCCAGGCAACT. The plasmids were transfected with Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions.
Antibodies
The antibodies used in this paper are detailed below:
N-cadherin (CST, cat# 13116), E-cadherin (CST, cat# 14472), PCNA (CST, cat# 13110), Slug (CST, cat# 9585), Snail (CST, cat# 3879), vimentin (CST, cat# 5741), β-actin (CST, cat# 3700), KRAS (CST, cat# 67648), p-AKT (473) (CST, cat# 4060), p-AKT (CST, cat# 9275), IGF2BP2 (CST, cat# 14672), HA (CST, cat# 5017), CD44 (BD Pharmingen; IM7) and C-myc (CST, cat# 18583).
Cell proliferation assay
A Cell Counting Kit-8 (CCK-8, Dojindo, Tabaru, Japan) was used to measure cell proliferation. Cells were seeded into 96-well plates, collected at different time points, and absorbance at 450 nm was detected. The relative level was then obtained by normalization from at least three independent experiments.
To investigate anchorage-independent effects on proliferation, a specified number of cells were seeded in a six-well plate. After incubation for 14 days, the colonies were stained with crystal violet (Sigma–Aldrich, St. Louis, MO, USA), and the colonies were counted.
Apoptosis assay
Cells with the indicated modifications were subjected to apoptosis assays following the assay kit (Invitrogen™ V35113) manufacturer’s instructions. This flow cytometry product detects the externalization of phosphatidylserine in apoptotic cells using recombinant annexin V conjugated to red laser-excited allophycocyanin, and dead cells using Green nucleic acid stain. Briefly, Apoptotic cells are detected by annexin V binding to externalized phosphatidylserine, and late apoptotic and necrotic cells have compromised membranes that permit green stain access to cellular nucleic acids. Cells with indicated modifications were subjected to apoptosis assay and live cells show little or no fluorescence, apoptotic cells show red fluorescence and very little green fluorescence, while late apoptotic cells show a higher level of red and orange fluorescence. Cells were distinguished using a flow cytometer that has both 488 nm and 633 nm excitation sources.
EdU assay
For the EdU assay, cells were seeded into 6-well plates and cocultured with an EdU-labeling reagent (Beyotime, C0071S) after incubation for 2 h. The cells were fixed and stained following the manufacturer’s protocol. Five fields of view were evaluated for each cell line.
Dual luciferase activity reporter system
The Renilla luciferase (Rluc) and firefly luciferase (Luc) sequences were amplified from the psiCheck 2 vector (Promega, USA). Rluc was placed in the upstream position, and Luc was placed in the downstream position. The KRAS sequence along with its 3′ UTR was amplified and inserted between Rluc and Luc. Relative activity was determined by normalization to Rluc and control.
LC–MS analysis
Proteins were separated via 12% SDS–PAGE and subjected to digestion with sequencing-grade trypsin (Promega, Madison, WI, USA). The fragments were analyzed using the National Center for Biotechnology Information nonredundant protein database with Mascot (Matrix Science, Boston, MA, USA) to identify specific peptides. The top 50 potential interacting proteins identified by LC/MS are shown in Supplementary Table
3.
Flow cytometry
Whole blood samples were collected from mice 5 weeks after establishing pulmonary metastatic models. Peripheral blood mononuclear cells (PBMCs) were enriched from the blood as a control. GFP-tagged tumor cells were selected using the selected GFP gate. Flow cytometric analysis was performed on a flow cytometer FACSCalibur (BD Biosciences, San Jose, CA).
Animal studies
Four-week-old female BALB/c nude mice were purchased from the Laboratory Animal Center of Sun Yat-sen University. All animals were treated in accordance with the guidelines of the Committee on Animals of Sun Yat-sen University. A total of 5 × 106 cells were inoculated into the mammary fat pad to establish subcutaneous xenografts. A total of 5× 103 cells were injected through the caudal vein. After 6 weeks, the samples were harvested and subjected to HE staining.
Immunohistochemistry (IHC)
Tumor slices were cut to an 8 to 10-μM thickness. After being deparaffinized in xylene and rehydrated, the tissue sections were washed 3 times for 10 min with PBS and incubated for 1 h in goat serum dilution buffer at room temperature. Primary antibodies were applied overnight at 4 °C in a wet chamber. After washing three times for 10 min with wash buffer, the tumor sections were incubated with secondary antibodies for 60 min at room temperature. The tumor sections were subsequently washed three times with the above wash buffer. Diaminobenzidine (DAB) reagent was added to these tumor sections, which were then counterstained with hematoxylin to visualize the nuclei.
RNA fluorescence in situ hybridization (FISH)
Cells were incubated with 50% formamide, 2X SSC, 0.25 mg/mL Escherichia coli transfer RNA, 0.25 mg/mL salmon sperm DNA (Life Technologies), 2.5 mg/mL BSA, and fluorescently labeled junction probe for 12 h. The cells were then washed and incubated overnight at room temperature. Images were captured using confocal microscopy. The sequence of the detection probe was as follows: 5′ cy3-TAGGAGTTGCCTGGACCATCCTTCTTCCTG 3′.
Reverse transcription and real-time (RT) PCR
Total RNA was extracted with a PureLink RNA mini kit (Thermo Fisher Scientific). After reverse transcription, cDNAs were harvested, and the resulting cDNA was then subjected to real-time PCR analysis with SYBR Select Master Mix (Thermo Fisher Scientific) in a StepOne Plus real-time PCR system (Applied Biosystems). The results for each sample were normalized to β-actin mRNA. The key primers are listed in Supplementary Table
2.
Northern blotting
Twenty micrograms of total RNA was extracted and separated using 1.2% agarose gel electrophoresis. After nelon membrane permeabilization and fixation, specific probes were applied at 42 °C and washed with 0.1% SDS at 68 °C. The circCD44 detection probe was as follows: 5′ CTACTAGGAGTTGCCTGGACCATCCTTCTTCCTGCTTGA –DIG 3′.
RNA pulldown
The cells were harvested, lysed and incubated with beads at 4 °C overnight. The beads were coated with biotin tagged circRNA using Pierce™ RNA 3′ End Desthiobiotinylation Kit (Thermo 20,163). After washing with wash buffer, RNA-Protein complex was harvested, the RNA complex was purified with TRIzol reagent and then subjected to qRT–PCR analysis. The complex was incubated with loading buffer to separate the protein from the complex and subjected to immunoblot assay and LC/MS assay.
RNA immunoprecipitation
RNA immunoprecipitation (RIP) was performed using a Magna RIP Kit (Millipore) following the manufacturer’s protocol. Cells were lysed with lysis buffer and incubated with conjugated beads for 6 h at 4 °C. After treatment with proteinase K, the protein was removed. RNA was extracted and purified with TRIzol Reagent. The purified RNA was subjected to qRT–PCR for further analysis.
Western blotting
Equal amounts of protein were added to each well containing 12% SDS–PAGE gels. After separation and transfer to membranes, the bands were blocked with 5% FBS and incubated with the corresponding primary antibody at 4 °C overnight. After incubation with the secondary antibody, the bands were visualized with an ECL kit.
Molecular docking
The X-ray structure of IGF2BP2 was downloaded from the RCSB Protein Data Bank (PDB code: 6ROL). HDOCK3 was used for docking IGF2BP2 and RNA; RNA was selected as the receptor, and IGF2BP2 was selected as the ligand. The interaction of the ligand and receptor was analyzed in Molecular Operating Environment (MOE) v2014.094 and visualized with PyMOL (
www.pymol.org).
RNA stability assay
The cells were seeded into 6-well plates and grown to 50% confluence. They were then treated with 5 μg/ml actinomycin D and collected at the indicated points. RNA levels were detected using qRT–PCR, and the half-life of mRNA was evaluated according to a published paper [
15].
Statistical analysis
Statistical analysis was applied using GraphPad Prism software. The data are presented as the mean ± SD. As indicated, Student’s two tailed unpaired t test was used to determine the statistical significance of the in vitro experiments. The log-rank test or Gehan-Breslow-Wilcoxon test was used to determine significant differences in the survival data. A p value of less than 0.05 was considered statistically significant. For each experiment, data are representative of at least three replications with similar results.
Discussion
CircRNAs are the most recently identified key players in TNBCs. circRNAs can act as competing endogenous RNAs to sponge miRNAs, to act as protein scaffolds, or to act as translational templates. These multiple functions suggested the critical roles of circRNAs during tumorigenesis and progression in TNBCs. Combination therapy was the most reliable for cancer therapy in TNBCs [
23,
24]. Generally, different reagents target different molecules in the same pathway to avoid drug resistance. However, TNBCs still develop acquired drug resistance when targeting one pathway, and the gene status involved in other pathways affects the efficiency of targeted therapy [
25]. In our study, we identified an uncharacterized circRNA - circCD44 - that was upregulated in TNBCs and promoted the progression of TNBCs by targeting both the miR-502-5p-KRAS axis and C-myc. Due to its dual targeting effects on KRAS and C-myc, targeting circCD44 in combination with other therapeutic approaches may not only achieve better therapeutic efficiency but also minimize resistance.
KRAS was reported to participate in the tumorigenesis and progression of breast cancers [
26], and it plays a central role in the activation of many pathways, such as AKT/MEK/ERK signaling. Researchers have shown that overactivation of the KRAS pathway occurs in TNBCs and causes chemoresistance [
27]. In recent decades, many efforts have been made to improve KRAS inhibitors. However, to date, only two compounds are currently in phase I clinical trials, highlighting the difficulties of this approach. Apart from KRAS inhibitors, in recent years, studies targeting the KRAS gene instead of the KRAS protein have attracted increasing attention. The G-quadruplex DNA sequence in the promoter region was reported to be essential for the transcription of KRAS. This region can be recognized by many transcription factors, such as MYC, which promote its downstream transcription [
28]. miRNAs targeting the 5′ untranslated region of KRAS were reported to reduce the level of KRAS and thus inhibit the progression of TNBCs [
29]. However, none of these miRNAs were reported to meet the clinical requirements, and more targets are still urgently needed. Our research indicated that circCD44 promotes KRAS expression by inhibiting miR-502–5p-induced KRAS degradation and that targeting circCD44 could be an attractive future approach to manipulate aberrant KRAS signaling in TNBCs.
C-Myc participates in physiological progression, including proliferation, differentiation and apoptosis [
30]. C-Myc also plays central roles in the tumorigenesis and progression of TNBCs [
31]. This centrality is due to its vital role in normal cells and cancer cells. Targeting C-Myc is a challenge but is still an urgent requirement [
32]. The C-Myc-targeting strategy can be summarized as follows: 1) Targeting C-Myc transcription: stabilization of G-quadruplexes with small molecules such as QN-1 [
33]. 2) Targeting C-Myc translation, such as targeting mTORC1 [
34] and eIF4A [
35] or destabilizing mRNA stability [
36]. 3) Targeting C-Myc stability [
37,
38]. 4) Targeting C-Myc downstream genes [
38]. To date, no molecules have been applied in clinical practice. Our research provides a new choice for targeting C-Myc by affecting mRNA stability by modulating IGF2BP2 activity. As an m6A reader, IGF2BP2 was reported to determine the fate of mRNAs and to be associated with the methylation of mRNA, thus influencing mRNA stability. IGF2BP2 was also reported to be involved in the progression of numerous cancers, and it was recently suggested as a potential biomarker predicting prognosis [
39]. The overexpression of IGF2BP2 may lead to the aberrant expression of many genes, including C-Myc [
15,
40]. IGF2BP2 was also reported to regulate the expression of glycolysis genes and thus switch cancer metabolism to adapt to environmental changes [
40]. In TNBCs, IMP2 and IMP3 were reported to cooperate to stabilize IGF2BP2 and thus contribute to metastasis [
41‐
43]. However, although the function of IGF2BP2 is vital and indispensable, the crosstalk of IGF2BP2 and circRNAs in TNBCs is still unknown. We showed that IGF2BP2 bound to circCD44 was functional and stabilized C-Myc mRNA, which further explained the complex role of IGF2BP2 in TNBCs and re-enforced the critical role of circCD44 in TNBCs.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.