The Ethical Committee of Taihe Hospital, Hubei University of Medicine approved the present study (NO.2024KS03). Cancer and para-cancer tissue specimens of patients undergoing breast cancer surgery in Taihe Hospital, Hubei University of Medicine from July 2017 to August 2019 were collected. Permission for use of breast tissues in this research was signed by all the participating patients. A total of 142 female patients, aged between 35–55 years old, with an average age of 43.26 $\pm$ 4.31 years old, were included. All the patients were followed up successfully and the follow-up data were recorded and analyzed.

Inclusion criteria: (1) pathological diagnosis of invasive breast cancer; (2) complete clinical medical records; (3) no radiotherapy, chemotherapy or endocrine therapy administered prior to surgery. Exclusion criteria: (1) patients with other primary malignant tumors; (2) defecits in clinical data and incomplete treatment. All patients signed informed consent forms. All sample tissues were stored at –80 ℃.

The MDA-MB-231 and MCF-7 cell lines were obtained from Academia Sinica Cell Bank (Shanghai, China), MCF-10A, MDA-MB-468, T47D and BT20 were bought from Shanghai Institute of Cell Biology, Chinese Academy of Sciences (Shanghai, China) and were identified with the short tandem repeat (STR) method. The cell lines used have been tested for Mycoplasma. Cells were cultured under the condition of 37 °C with 5% carbon dioxide. Dulbecco Modified Eagle Medium (DMEM) (11965092, Thermo Fisher Scientific, Waltham, MA, USA) was added with 10% fetal bovine serum (FBS, Life Technologies, Carlsbad, CA, USA) as the culture medium, with antibiotics being streptomycin (100 µg/mL) (Beyotime, Shanghai, China) and penicillin (100 U/mL) (Beyotime, Shanghai, China).

The shRNA that specifically targets for circPVT1 (si-circPVT1) (20 nM) and a negative control that targets nothing (si-NC) were chemical synthesized by Shengong Biotechnology (Shanghan, China). The sequences of siRNA were si-circPVT1#1: 5^′-GAGCTTCGTTCAAGTATTT-3^′, si-circPVT1#2: 5^′-GAAATGTCCTCTCGCCTGC-3^′ and si-NC: 5^′-UUCUCCGAACGUGUCACGUTT-3^′. Lipofectamine 2000 (Invitrogen, Shanghai, China) was added for the transfer of MDA-MB-231 and MCF-7. After 48 h transfection, the silenced cells were harvested and applied for the following experiments.

The breast cancer cell lines MCF-7 and MDA-MB-231 were cultured without serum in 6-well plates. They were divided into si-circPVT1#1, si-circPVT1#2, and siRNA NC with a final concentration of 100 nM. After transfection for 24 h, 1000 cells were inoculated in 96-well plates with a volume of 200 µL per well. The light absorption values of cells were detected at culture time points 1, 2, 3 and 4 days. According to the description of CCK-8 proliferation detection reagent, each well had added 10 µL proliferation detection reagent, incubated for 2 h, and the optical density (OD) value of each well was measured at 450 nm wavelength. The growth curves of different groups of breast cancer cells were drawn.

The TRIzol reagent (Invitrogen, Shanghai, China) was adopted to extract tissue or cell RNA. The NanoDrop ND-1000 (Thermo Scientific, Waltham, MA, USA) was utilized to measure the RNA concentration and quality. The total RNA was reverse transcribed (Takara, Tokyo, Japan). Quantitative real-time PCR was completed with SYBR Green PCR Master Mix reagents (Takara, Tokyo, Japan). The primer sequences were as follows: circPVT1 forward, 5^′-CTTCCTGGTGAAGCATCTGAT-3^′ and reverse, 5^′-TTCAGCCTCCACTTAAAGTACC-3^′; and GAPDH forward, 5^′-CATGAGAAGTATGACAACAGCCT-3^′ and reverse, 5^′-AGTCCTTCCACGATACCAAAGT-3^′. The expression level of circPVT1 mRNA was relatively normalized to the GAPDH. The 2^{-Δ⁢Δ⁢Ct} method was utilized to compare the quantification difference.

A wound-healing assay was adopted to observe cell migration. The cells were seeded with a concentration of 5 $\times$ 10⁴ cells per well in six-well plates and were cultured for 24 hours in serum-free medium. Cells were individually scratched by a sterile 10 µL micropipette tip. Photos were taken at 0 and 24 hours after the scratching with the microscope (Leica, Malvern, PA, USA). Wound closure was measured and compared.

The 8 µm pore transwell (BD Biosciences, San Jose, CA, USA) was utilized to observe cell invasion. Briefly, 1 $\times$ 10⁵ cells in 200 µL serum-free culture medium were placed in the upper transwell chamber with Matrigel (Corning, New York, NY, USA), and 10% FBS medium was added to the transwell bottom chamber. Incubation time was 48 hours. The non-invading cells were moved away by a cotton swab from the upper chamber. The invaded cells on the transwell membrane were fixed with 4% paraformaldehyde (NO.P1110, Maichen Gene Technology, Beijing, China) and then dyed with 1% crystal violet (Sigma-Aldrich Co., St Louis, MO, USA). The invaded cells were randomly calculated in five different fields with a microscope (Leica, Malvern, PA, USA).

Sixty micrograms of total protein were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) electrophoresis and electrophoretically transferred onto PVDF (poly vinylidene fluoride) membranes. Five percent milk was adopted to block the membrane. The membrane was incubated with astrocyte elevated gene 1 (AEG-1) or GAPDH (1:1000) antibodies overnight at 4 °C. AEG-1 (1:1000) was purchased from Abcam (No. ab229128, Cambridge, MA, USA) and GAPDH was from Cell Signaling Technology, Inc. (No. #5174, Beverley, MA, USA). The proteins were observed by enhanced chemiluminescence (Millipore, Beijing, China) following the manufacturer’s procedure. Image J 1.53 (https://imagej.net/ij/) to compare western blot (WB) gray value. AEG-1 gray value is normalized to GAPDH.

The luciferase assay kit was purchased from Vigorous Biotech (Beijing, China) and the TOP/FOP Flash luciferase reporter plasmid was bought from Biovector NTCC Ltd (Beijing, China). Luciferase assays were completed following by the manufacturer’s protocol. The miRNA-30b-5p binding site sequence of circPVT1 or AEG-1’s 3^′untranslated region (3^′-UTR) was cloned into pGL3 vector (wild type), respectively. The 3^′-UTR wild type (WT) and mutatin (MUT) luciferase reporter plasmids containing circPVT1 or AEG-1 were synthesized and transfected into MDA-MB-231 or MCF-7 cells cultured in 24-well plates. Simultaneously transfected 100 ng wild type or mutate type plasmid, 100 nm invalid nucleotide or miRNA-30b-5p in the MDA-MB-231 or MCF-7 cells were placed in 24-well plates. The luciferase assay was done after 24 hours transfection. Firefly luciferase was utilized as a reference, and Renilla luciferase was utilized as the internal control.

The 6-week-old female Balb/c mice were obtained from Charles River (Beijing, China). MDA-MB-231 cells transfected with si-circPVT1 or si-NC and were separately injected into the dorsum of the mice (n = 7 per group), 5 000 000 cells/200 µL phosphate buffered saline (PBS) were inoculated. Tumor sizes were carefully measured weekly and the formula L $\times$ S/2 (L: tumor’s long diameter; S: tumor’s short diameter) was applied to calculate tumor volume. At 25 days post-injection, all mice were euthanasia executed by decapitation and tumor tissues were removed. Quantitative real-time polymerase chain reaction (q-PCR) was used to examine the relative circPVT1 expression in tumor tissues. Animal studies had obtained the agreement of the Animal Care and Use Committee of Taihe Hospital (NO.2024KS03).

Tumor tissue samples were collected, embedded in paraffin with 10% formaldehyde solution. Antigen retrieval was enhanced by microwaving in sodium citrate solution antigen repair liquid antigen of 98 °C for 15 minutes to repair, rabbit anti human AEG-1 (Abcam, No. ab229128, Cambridge, MA, USA) or PCNA monoclonal antibody (Abcam, No. ab92552, Cambridge, MA, USA) (1:100) 4 °C for the night, horseradish peroxidase labelled goat anti-rabbit second antibody (Shenggong Bioengineering Company, NO. E670020, Shanghai, China) incubation for 1 hour, colour reaction with 3,3-diaminobenzidine (DAB) (NO.DAB-0031, MaiXin Bioengineering Company, Fuzhou, Fujian, China), and then all the sections were lightly counterstained with hematoxylin and examined by light microscopy. The results were finally evaluated by the staining intensity and staining area score. The staining intensity was negative to strong (0–3), negative (0 point), weak positive (1 point), moderate positive (2 points), and strong positive (3 points). The staining positive area score was 0 (10%); 1 (10–25%); 2 points (26–50%); 3 (51–75%); 4 ($>$76%). Staining area score $\times$ staining intensity score was the final score, with a score $\ge$4 to judge positive and a score 4 to judge negative.

SPSS 21.0 (SPSS Inc., Chicago, IL, USA) was used to analyse the data. The measurement data were analyzed with t test, and the Kaplan–Meier method was used to evaluate disease free survival (DFS) and overall survival (OS). The Chi-square Test was adopted to compare the difference between circPVT1 expression in breast cancer patients and clinicopathological characteristics. Values of p 0.05 were defined as being statistically significant.

CircPVT1 was expressed in 142 cases of breast cancer and corresponding paracancer tissues. The results indicated that the relative expression level of circPVT1 in breast cancer tissues was 4.21 $\pm$ 0.17, while that in corresponding paracancer tissues was 2.57 $\pm$ 0.12. Statistical analysis using matched data showed a significant difference (t = 8.613, p 0.001), indicating that circPVT1 expression was increased in cancer tissues (Fig. 1A). q-PCR was used to detect the circPVT1 expression in breast cancer cells. Compared with normal breast cancer epithelial cells, the expression of circPVT1 in breast cancer cell line was increased (Fig. 1B). The difference was statistically significant (p 0.05), and the highest expression level was found in MDA-MB-231 cells (p 0.01). MCF-7 is a widely used cell line in breast cancer research. The Kaplan-Meier results indicated that high circPVT1 expression correlated with a worse prognosis as compared to those with low circPVT1 expression in DFS ($\chi$² = 7.174, p = 0.007) and OS ($\chi$² = 3.946, p = 0.047) (Fig. 1C,D).

Fig. 1.

Increasing expression of circPVT1 in breast cancer tissues and cells. (A) CircPVT1 expression was elevated in breast cancer tissues compared to the corresponding paracancer tissues, n = 142. GAPDH was adopted as the internal control. Data come from triplicate experiments. (B) q-PCR to detect circPVT1 RNA levels in different breast cancer cell lines. n = 3. MCF-10A is a normal mammary cell. (C) The Kaplan-Meier results indicated that higher circPVT1 expression correlated with poorer disease free survival. (D) The Kaplan-Meier results indicated that higher circPVT1 expression correlated with poorer overall survival. circPVT1, circular RNA PVT1; q-PCR, quantitative real-time polymerase chain reaction. *p 0.05 vs. the control; **p 0.01 vs. the control; ***p 0.001 vs. the control.

The mean expression level of circPVT1 in breast cancer tissues was 4.21. High expression of circPVT1 was related to axillary lymph node metastasis ($\chi$² = 4.108, p = 0.043) and differentiation degree ($\chi$² = 4.823, p = 0.028). However, there was no correlation with patient age, Ki67 index, HER2 (human epidermal growth receptor 2), ER (estrogen receptor), PR (progesterone receptor), tumor diameter or tumor stage (p $>$ 0.05) (Table 1).

In order to further study the role of circPVT1 in breast cancer, cell lines with circPVT1 knockout were established in breast cancer cells. The cell lines with circPVT1 knockout were successfully established by q-PCR (Fig. 2A). Further cell count assay showed that the elimination of circPVT1 significantly inhibited the proliferation ability of MCF-7 and MDA-MB-231 cells (Fig. 2B,C) (p 0.05).

Fig. 2.

Knockout of circRNA PVT1 in breast cancer cells. (A) Knockout of circRNA PVT1 expression compared to negative control (NC) group. (B,C) Cell proliferation assay results suggested that knockout of circRNA PVT1 inhibited the breast cancer cell proliferation. (D) Scratch assay suggested that knockout of circRNA PVT1 inhibited the migration of MCF-7 and MDA-MB-231 breast cancer cells, scale bar = 200 µm. (E) Cell invasion assay suggested that knockout of circRNA PVT1 effectively inhibited the breast cancer cell invasion, scale bar = 100 µm. ***p 0.001 vs. the control. n = 3.

In order to further study the functions of circPVT1 on breast cancer cell biology, the cell scratch test was applied to study the migration ability. The elimination of circPVT1 significantly inhibit the breast cancer cell migration ability (Fig. 2D). The transwell invasion assay further indicated that the invasion ability of knockout circPVT1 in MCF-7 and MDA-MB-231 cells was significantly suppressed compared to the control group (Fig. 2E).

Since circular RNA can act as a miRNA sponge, by searching the circinteractome database, we found that the circPVT1 sequence contained the binding site of microRNA-30b-5p (miR-30b-5p) (Fig. 3A), and further confirmed the relationship between circPVT1 and miR-30b-5p. The luciferase results indicated that overexpression of miR-30b-5p could inhibit luciferase activity in circPVT1 3^′UTR wild type group. When the circPVT1 3^′UTR binding site was mutated, the overexpression of miR-30b-5p could not effectively inhibit luciferase activity (Fig. 3A). To confirm the reliability of our results, q-PCR was applied to detect the expression level of miR-30b-5p in MCF-7 and MDA-MB-231 cells, miR-30b-5p expression was significantly increased after circPVT1 knockout (Fig. 3B, p 0.05). Therefore, it could be inferred that circPVT1 might be involved in breast cancer progression through sponging miR-30b-5p.

Fig. 3.

CircRNA PVT1 regulates miR-30b-5p/AEG-1 pathway. (A) The binding places between miR-30b-5p and circRNA PVT1. Luciferase assay results indicated that miR-30b-5p mimics could target circRNA PVT1 3^′UTR-WT (3^′-untranslated region-wild type). (B) CircRNA PVT1 knockout by the small interference RNA could up-regulate miR-30b-5p expression. (C) The binding sites between miR-30b-5p and AEG-1. Luciferase assay results indicated that miR-30b-5p could target AEG-1 3^′UTR-WT. (D) Small si-RNA knockout circRNA PVT1 in MCF-7 and MDA-MB-231 cells or miR-30b-5p mimics could downregulate AEG-1 protein expression. MiR-30b-5p inhibitor reversed si-circRNA PVT1 function. AEG-1, astrocyte elevated gene 1; MUT, mutatin. ***p 0.001; n = 3.

By searching the target scan database, we found the presence of miR-30b-5p binding site 3^′UTR in AEG-1. Moreover, the biding between AEG-1 and miR-30b-5p was confirmed by luciferase reporter assay and the results revealed that overexpression of miR-30b-5p could inhibit luciferase activity in AEG-1 3^′-UTR wild type group. The inhibitory effect disappeared in the mutation group of 3ʹ-UTR binding site of AEG-1 (Fig. 3C). In order to confirm the reliability of our results, we used Western blotting to detect the overexpression of miR-30b-5p in MCF-7 and MDA-MB-231 or the elimination of circPVT1 with si-circPVT1#1, both of which inhibited the expression of AEG-1. However, simultaneous inhibition of miR-30b-5p expression in the circPVT1 knockout group could reverse the inhibition of AEG-1 expression (Fig. 3D).

The expression level of MDA-MB-231 in mice was transfected with si-NC or si-circPVT1#1, and the results demonstrated that si-circPVT1 effectively inhibited the tumor growth (Fig. 4A) and the expression of circPVT1 (Fig. 4B), reduced the tumor volume and weight (Fig. 4C,D), and inhibited the expression levels of AEG-1 and proliferating cell nuclear antigen (PCNA) in nude mice (Fig. 4E).

Fig. 4.

CircPVT1 knockdown inhibited tumor growth in vivo. (A) The MDA-MB-231 cells were transfected with si-circPVT1 or si-NC respectively, and injected into the dorsum of Balb/c mice. Tumor tissues were collected after 5 weeks. (B) q-PCR was used to detect the relative circPVT1 expression in tumor tissues. (C) Tumor sizes were carefully measured weekly and the formula L $\times$ S/2 was adopted to calculate tumor volume. (D) The difference of tumor weight between the si-circPVT1 and si-NC groups. *p 0.05 vs. the control; ***p 0.001 vs. the control. n=5, there are five mice in si-circPVT1 and si-NC respectively. (E) CircPVT1 knockdown inhibit tumor AEG-1and proliferating cell nuclear antigen (PCNA) expression in nude mice. Immunohistochemical method was used to detect AEG-1and proliferating cell nuclear antigen (PCNA) expression in nude mice (representative images). Scale bar = 100 µm.