CircWHSC1 Promotes Breast Cancer Progression by Regulating the FASN/AMPK/mTOR Axis Through Sponging miR-195-5p

Numerous studies reveal that circular RNAs (circRNAs) affect cancer progression. CircWHSC1 is a novel circRNA that accelerates ovarian cancer progression. Nevertheless, the function of circWHSC1 in regulating breast cancer (BC) is elusive. Here, quantitative reverse transcription-polymerase chain reaction (qRT-PCR) was carried out to detect the profiles of circWHSC1 and miR-195-5p in BC tissues and corresponding non-tumor tissues. Gain- and loss-of-function assays were implemented both in vivo and ex vivo to verify the significance of circWHSC1 in BC development. BC cell proliferation was estimated by the cell counting kit-8 (CCK-8) and BrdU assays. Transwell assay was implemented to test BC cell migration and invasion. The protein levels of FASN, AMPK and mTOR were determined by Western blot. Moreover, immunohistochemistry was performed to examine Ki67 and FASN expression. As shown by the result, circWHSC1 was up-regulated in BC tissues versus adjacent non-tumor tissues. circWHSC1 overexpression was correlated with higher tumor stages, lymphatic metastasis and worse survival of BC patients. Functionally, overexpressing circWHSC1 amplified proliferation, migration and invasion of BC cell lines and boosted xenograft tumor growth in nude mice. Bioinformatics uncovered that circWHSC1 functioned as a competitive endogenous RNA by sponging miR-195-5p, which was further corroborated by the dual-luciferase reporter assay and RNA immunoprecipitation. miR-195-5p delayed BC progression, which was dampened by circWHSC1 up-regulation. Fatty acid synthase (FASN) was affirmed as a direct target of miR-195-5p. miR-195-5p overexpression curbed FASN expression and activated its downstream AMPK pathway. Inhibition of FASN or activation of the AMPK pathway reversed circWHSC1-mediated oncogenic effects. Collectively, CircWHSC1 acted as an oncogene to expedite BC evolvement by modulating the miR-195-5p/FASN/AMPK/mTOR pathway.


INTRODUCTION
Breast cancer (BC) is a familiar malignancy harming women's health worldwide (1). Hormone therapy is only appropriate for HR-positive cancer patients. Surgery is usually suitable for early BC patients or those who have received neoadjuvant treatment instead of metastatic patients. Thus, the overall morbidity of BC patients is still growing due to limited methods in treating metastatic BC or triple-negative breast cancer (TNBC) (2,3). Hence, elucidating the molecular mechanism of BC progression is vital for improving the efficacy and prognosis of BC patients.
Circular RNAs (circRNAs) are long noncoding endogenous RNAs with limited protein-coding abilities and are involved in various transcriptional and post-transcriptional genomic regulations (4). Accumulating reports have stated that circRNAs participate in tumorigenesis and tumor evolvement (5). For example, circ-0001742 regulates the miR-431-5p/ATF3 axis by functioning as a competitive endogenous RNA (ceRNA), thereby aggravating tongue squamous cell carcinoma (TSCC) progression (6). circ-0000267 promotes the proliferation, migration, invasion and epithelial-mesenchymal transition (EMT) of gastric cancer cells by up-regulating the miR-503-5p/ HMGA2 axis (7). In contrast, circCDYL represses colon cancer cells' growth and migration by inhibiting miR-150-5p (8). CircWHSC1 is a newly discovered circRNA. Overexpressing circWHSC1 inhibits miR-145 and miR-1182, and aggravates the growth and metastasis of ovarian cancer cells (9). Yet, there are few reports on the function of circWHSC1 in BC.
MicroRNAs (miRNAs) are noncoding RNAs with 20-24 nucleotides in length, which modulate gene expression through diverse mechanisms (10). miRNAs modulate cell growth and differentiation, act as promising tumor markers and contribute to cancer development. For example, miR-424-5p is a contributor of the proliferation, migration, invasion and adhesion of laryngeal squamous cell carcinoma (LSCC) cells by up-regulating CADM1 (11). miR-627-5p decreases the proliferation and metastasis of hepatoma cells by suppressing BCl3 (12). As an important miRNA, miR-195-5p is located at 17p13.1 and 87bp in length, which has a role in controlling cell proliferation and migration. For example, INC-ROR facilitates the proliferation and migration of aortic smooth muscle cells by regulating the miR-195-5p/FGF2 pathway (13). The mechanism by which miR-195-5p modulates BC progression is, however, unclear and this requires further study.
Fatty acid synthase (FASN) is located at 17q25.3 and is 8464bp in length. Functionally, FASN catalyzes the synthesis of acetyl CoA and malonyl CoA and palmitic acid into long saturated fatty acids in the presence of NADPH (14). Interestingly, FASN is abnormally expressed in tumor tissues and mediates tumor evolvement (15). For instance, FASN enhances EMT in ovarian cancer through transcriptional regulation of E-cadherin and N-cadherin. Meanwhile, knocking down FASN dampens ovarian cancer invasion and migration both in vivo and in vitro (16). In addition, FASN gets involved in the regulation of cancer-associated metabolic reprogramming, and targeting FASN has been regarded as an effective method in treating BC (17).
Here, we found that circWHSC1 was up-regulated in BC, and higher expression of circWHSC1 predicted worse survival of BC patients. Functionally, circWHSC1 promoted BC cells' proliferation, migration and invasion. Moreover, we verified a targeted regulatory correlation among circWHSC1, miR-195-5p and FASN through biological information analysis. By detecting circWHSC1, miR-195-5p and FASN expressions in BC tissues and cells and probing the association between these molecules, we confirmed that circWHSC1 targets miR-195-5p by acting as a ceRNA and up-regulates FASN. In summary, this research sheds light on the molecular mechanisms in BC evolvement and presents novel references for BC therapy.

Clinical Specimen Collection and Handling
The tumor tissues and adjacent non-tumor tissues from 50 primary BC patients who received surgery in Minhang Hospital of Fudan University from October 2014 to October 2015 were collected. No chemotherapy, radiotherapy, or other neoadjuvant treatments were performed on those patients before operation. No cancer cells were observed in the adjacent non-tumor tissues by postoperative pathological examination. All samples were isolated and promptly preserved in liquid nitrogen at -196°C for RNA extraction. The research ethics committee of Minhang Hospital of Fudan University authorized this study, and informed agreement was obtained from all participating patients.

Cell Transfection
MCF7 and MDA-MB-231 cells in the logarithmic growth phase received trypsinization by 0.25% trypsin. The cells were collected and seeded in 6-well plates with 5×10 6 cells per well. 24 hours later, the two cell lines were transfected with circWHSC1 overexpressing plasmids, miR-195-5p mimics and their negative controls by applying the Lipofectamine ™ 3000 Transfection Reagent (Thermo, Shanghai, China) as per the manufacturer's instructions. Afterward, the cells were kept at 37°C with 5% CO 2 . After 24 hours of transfection, the medium was exchanged with a new compete medium and the cells were kept culturing for another 24 hours. After that, the total RNA was extracted from the cells, and quantitative reverse transcription-polymerase chain reaction (qRT-PCR) was performed to detect the expressions of circWHSC1 and miR-195-5p.

qRT-PCR
The total RNA was extracted out of cells by using the TRIzol reagent. After RNA concentration and purity measurement, the total RNA was reversely transcribed into cDNA with the PrimeScript ™ RT Reagent kit (Invitrogen, Shanghai, China). The procedures were conducted according to the manufacturer's instructions. Subsequently, the Bio-Rad CFX96 quantitative PCR system and SYBR Green Real-Time PCR Master Mixes (Thermo Fisher Scientific, Shanghai, China) were utilized to conduct qRT-PCR. The reaction conditions included: pre-denaturation (95°C, 5 min), denaturation (95°C, 15s), and annealing (60, 30s). GAPDH acted as a housekeeping gene for calculating the relative circWHSC1 and FASN expression, and U6 was utilized as an internal control for calculating the relative miR-195-5p expression. The 2 (-DDCt) method was employed for relative gene expression analysis. The primers were as follows: miR-195-5p Forward

Cell Counting Kit-8 (CCK-8) Assay
MCF7 and MDA-MB-231 cells in the logarithmic growth phase were trypsinized, resuspended (2×10 3 /mL) and inoculated in 96well plates with 100 mL cell suspension per well. Three replicate wells were established per group. The plates were then maintained in an incubator for further culture. After 24 hours, 10 mL of CCK-8 solution (Hubei Biossci Biotechnology Co., Ltd.) was added to each well, and the cells were cultured for 1 hour. After the culture was completed, the optical density (OD) value of each well at 450 nm was measured at the 24 th , 48 th , 72 th , and 96 th hours.

BrdU Assay
MCF7 and MDA-MB-231 cells were seeded into 24-well plates at 1×10 5 cells/well. 24 hours later, the BrdU reagent (Sigma, Beijing, China) was added into each group, and the plates were incubated at 37°C with 5% CO 2 . After the cells were cultured for 48 hours, immunofluorescence staining was performed according to the anti-BrdU antibody (Abcam) operation guidelines. DAPI (Beyotime, Shanghai, China) was used for labeling the nucleus. The BrdU-positive cells and DAPI-positive cells were counted in 3 random areas with the fluorescence microscope (Olympus, Japan). Cell proliferation rate=the number of BrdU-positive cells/DAPI-positive cells×100%.

Transwell Assay
MCF7 and MDA-MB-231 were seeded in 24-well culture plates with a transwell chamber (Corning, Beijing, China). The chambers were coated with Matrigel (Cat.No. 356234, BD company, Germany) in the invasion test, while they were not pre-coated with Matrigel in the migration experiment. The upper chambers were supplemented with 5×10 4 cells. In parallel, 10% FBS medium and 400 mL of RPMI-1640 was added to the lower chamber. After an incubation at 37°C for 24 hours, non-migrated or noninvasive cells were wiped off. 4% paraformaldehyde was used for fixing the cells for 10 minutes at room temperature, and then 0.5% crystal violet was adopted for cell staining. After rinsing with running water, an inverted microscope was used for cell counting. All tests were done in triplicate.

RNA Immunoprecipitation (RIP)
RIP was carried out with the Magna RIP RNA binding protein immunoprecipitation kit (Millipore, Bedford, MA, USA). BC cells transfected with miR-195-5p mimics or miR-NC at 80% confluence were collected, then lysed using RIP lysis buffer. Then, the anti-Ago2 antibody (Abcam, Shanghai, China) was applied for Ago2 immunoprecipitation, and the immunoglobulin G (IgG) antibody served as a negative control. Immunoprecipitated RNA was separated and the enrichment of CircWHSC1 and FASN in the lysates was assessed by qRT-PCR.

The Xenograft Tumor in Nude Mice
MCF7 and MDA-MB-231 cells transfected with circWHSC1 overexpression plasmids or negative controls were collected by 0.25% trypsin. Single-cell suspensions (3×10 7 /mL) were made using the serum-free medium. 6-week-old BALB/c-nu nude mice were used for establishing xenograft tumor model. 40 nude mice were divided into 4 groups, and 0.1 mL of cell suspension was administered subcutaneously into each mouse's left forelimb. After the injection, the tumor sizes were calculated weekly for five weeks, with the mass of the tumor tissues derived from mice measured. The lung tissues of all nude mice were collected and stained by H&E to evaluate pulmonary metastases of BC cells.

Immunohistochemistry (IHC)
Following paraffin embedding and sectioning (4 mM), the tumor tissue sections were dewaxed with xylene, hydrated with gradient alcohol, and treated with 3% H 2 O 2 for 10 minutes. Microwave repair was made by applying 0.01 mol/L sodium citrate buffer (pH=6.0, 15 min), and the sections were blocked with 5% bovine serum albumin (BSA) for 20 minutes, followed by the incubation with anti-FASN (ab22759, Abcam, MA, USA) and Ki67 (ab15580, Abcam, MA, USA) overnight at 4°C. The following day, the goat-anti-rabbit IgG was added and kept at 37°C for 60 minutes. After rinsing with PBS, DAB was used for color development. After hematoxylin re-staining, the sections underwent dehydration, permeation, sealing, and examination under a microscope (Olympus, Japan).
Data Analysis SPSS 20.0 (SPSS Inc., Chicago, IL, USA) was utilized for data analysis. Mean ± standard deviation (`x ± s) was used for data measurement. The multivariate comparison was conducted with one-way analysis of variance, and pairwise comparison was made by t test. P <0.05 was considered as statistical significance.

The Clinical Importance of circWHSC1 in BC Tissues
To verify the function of circWHSC1 in BC, we performed qRT-PCR. The data indicated that circWHSC1 was notably overexpressed in BC tissues compared with that in the adjacent non-tumor tissues (P <0.05, Figure 1A). Besides, circWHSC1 was distinctly up-regulated in BC cell lines (MCF7, BT474, SKBr-3, ZR-75-30, MDA-MB-453 and MDA-MB-231) by contrast with that in human normal breast epithelial cell MCF-10A (P <0.05, Figure 1B). Next, we evaluated the function of circWHSC1 in predicting the survival of BC patients. As a result, a higher level of circWHSC1 was markedly linked to worse overall survival of BC patients ( Figure 1C). Meanwhile, BC patients with a higher expression of circWHSC1 had a more advanced tumor stage, more lymphatic metastasis and distant metastasis and higher Ki67 levels ( Table 1). Therefore, circWHSC1 was a potential oncogene in BC.

Impacts of circWHSC1 on BC Cell Proliferation, Migration and Invasion
For the purpose of exploring how circWHSC1 affected BC evolvement, we constructed circWHSC1 overexpression cell models in MCF7 and MDA-MB-231 ( Figure 2A). CCK-8 and BrdU methods demonstrated that cell proliferation was markedly facilitated after circWHSC1 overexpression versus the NC group (P <0.05, Figures 2B-D). Further, transwell assay was conducted to study circWHSC1's impact on cell migration and invasion. As displayed in the figure, cell migration and invasion were dramatically enhanced after circWHSC1 overexpression (P <0.05, Figures 2E, F). WB results testified that circWHSC1 overexpression reduced BAX, c-Caspase3, and E-cadherin levels, whereas up-regulated bcl2, Ncadherin, and Vimentin ( Figures 2G, H). Those data indicated that circWHSC1 promotes the malignant phenotypes of BC cells and plays a cancerigenic role.

Overexpressing circWHSC1 Amplified BC Cells Growth In Vivo
We constructed circWHSC1 overexpression models in MCF7 and MDA-MB-231 to verify its role on cell growth in vivo. It was found that in the MCF7 cell lines, the tumor size and mass in the circWHSC1 overexpression group were distinctly increased versus the vector group (P <0.05, Figures 3A-D). Immunohistochemical outcomes corroborated that the ratio of Ki67-positive cells was increased after circWHSC1 overexpression, indicating a significant increase in tumor proliferation (P <0.05, Figure 3E). At the same time, lung metastases increased (P <0.05, Figure 3F). Similar outcomes were found in MDA-MB-231 cell lines with circWHSC1 overexpression (P <0.05, Figures 3G-L). Thus, overexpressing circWHSC1 enhanced BC cell growth.
circWHSC1 Targeted miR-195-5p and Suppressed miR-195-5p-Mediated Inhibitive Effects on BC Cells By detecting the expression of miR-195-5p in BC tissues, we concluded that the miR-195-5p level was significantly decreased in BC tissues versus non-tumor tissues (P <0.05, Figure 5A). Similar result was discovered by the data analysis on Starbase ( Figure 5B). The lower level of miR-195 was linked to the worse overall survival of BC patients, as evidenced by KM plotter analysis (https://kmplot.com/analysis/index.) ( Figure 5C).

FASN Was Targeted by miR-195-5p
We were curious about the downstream mechanism of miR-195-5p in BC. Through miRpath (http://snf-515788.vm.okeanos. grnet.gr/index.php?r=miRpath) analysis, we believed that miR-195-5p is effective in modulating the fatty acid-related pathways ( Figure 6A). Next, we analyzed the target genes regulated by miR-195-5p via Starbase and validated that FASN is a candidate (Figures 6B, C). To test the targeting association between miR-195-5p and FASN, we performed RIP assay. It turned out that enriched FSAN level was significantly higher in the Ago2 antibody group versus the IgG group after miR-195-5p transfection, indicating FASN bound to Ago2 via miR-195-5p (P <0.05, Figure 6D). We performed dual-luciferase reporter gene assay, and found that miR-195-5p greatly decreased the luciferase activity of FASN-WT-transfected cells, while it had little influence on that of FASN-MT (P> 0.05, Figure 6E). As shown by qRT-PCR result, FASN was markedly up-regulated versus that in adjacent non-tumor tissues (P <0.05, Figure 6F). Furthermore, Pearson linear regression analysis disclosed that FASN had a negative relationship with miR-195-5p in BC tissues (P <0.0001, R 2 = 0.5705, Figure 6G). In contrast, FASN and circWHSC1 levels were positively correlated in BC tissues (P <0.0001, R 2 = 0.4972, Figure 6H). We detected FASN levels at low or high levels of circWHSC1 and miR-195-5p. WB results showed that FASN had higher expression in BC tissues with high CircWHSC1 expression and low miR-195-5p expression ( Figures 6I, J). These data indicated that FASN acts as a downstream target of miR-195-5p, and there is a potential circWHSC1/miR-195-5p/FASN axis in BC.

The Modulation of circWHSC1 and miR-195-5p on FASN
For making certain the circWHSC1/miR-195-5p/FASN pathway in BC, the FASN mRNA and protein contents were evaluated by using qRT-PCR and WB, respectively. We found that circWHSC1 overexpression elevated FASN expression, while miR-195-5p upregulation inhibited FASN expression ( Figures 7A-D). In addition, the AMPK/mTOR pathway was detected. It was observed that overexpressing circWHSC1 markedly decreased p-AMPK level, and facilitated FASN and p-mTOR levels ( Figure 7C). By contrast, overexpressing miR- 195-5p up-regulated p-AMPK and down-regulated p-mTOR. However, up-regulating circWHSC1 in the miR-195-5p overexpressing group led to down-regulated p-AMPK and upregulated FASN and p-mTOR versus the miR-195-5p group (P <0.05, Figure 7D). IHC was conducted for evaluating FASN in the formed tumor tissues as shown in Figure 3. FASN level in the circWHSC1 overexpressing group was enhanced versus the vector group ( Figure 7E). These results displayed that circWHSC1 affected the FASN/AMPK/mTOR expression by regulating miR-195-5p.

Attenuation of FASN or Activation of AMPK Reversed circWHSC1-Mediated Pro-Carcinogenic Effects
To confirm the mechanism of FASN/AMPK pathway in circWHSC1-mediated carcinogenic effects, we treated MCF7 cells with the FASN inhibitor Cerulenin or the AMPK activator Metformin. The expression of FASN and the AMPK-mTOR pathway was checked by WB. Notably, in comparison to the circWHSC1 group, Cerulenin and Metformin treatment resulted in FASN and mTOR phosphorylation downregulation, and dramatic facilitation in AMPK phosphorylation level ( Figure 8A). Next, we conducted the CCK-8 assay and BrdU assay. Cerulenin and Metformin treatment caused a significant reduction in MCF7 viability and proliferation versus the circWHSC1 group ( Figures 8B, C). As evidenced by the Transwell assay, the application of Cerulenin and Metformin substantially attenuated the migratory and invasive ability of MCF7 cells versus the circWHSC1 group ( Figures 8D, E). WB was carried out to evaluate the apoptosis-associated proteins and EMT-associated proteins. The data revealed that Cerulenin and Metformin treatments up-regulated Bax, c-Caspase3, and Ecadherin, and reduced the levels of bcl2, N-cadherin and Vimentin (vs. circWHSC1 group, Figures 8F, G). Thus, circWHSC1 exerts carcinogenic effects mainly through regulating the FASN-AMPK axis (Figure 9).

DISCUSSION
Emerging circRNAs are found to play a role in BC (18). Therefore, studying the biological function of circRNAs and their potential mechanisms makes sense in the clinical treatment and diagnosis of BC. Here, we found that circWHSC1 was upregulated in BC tissues and acted as a potential prognostic biomarker for BC patients. CircRNAs are a special type of noncoding RNAs, which modulate tumor-related gene expression and participate in tumorigenesis as a ceRNA (19,20). For example, circACAP2 is up-regulated in BC tissues, which elevates cell growth and metastasis by controlling the miR-29A/B-3P/COL5A1 axis (21). Zhang Wei et al. reported that circHOXC-AS3 upregulates PPP1R1A by sponging miR-3922-5p, thus promoting BC cell growth and metastasis (22). Besides, circRNAs modulate the chemotherapy sensitivity of BC cells. As an example, circ_0025202 hampers BC cells' proliferation, colony formation, and migration and enhances their apoptosis and sensitivity to tamoxifen (23). circWHSC1 functions as an oncogene in multiple tumors, including ovarian cancer (9), endometrial cancer (24), and hepatocellular carcinoma (25). In terms of its functions, circWHSC1 increases tumor cell proliferation, migration and invasion, and suppresses cell apoptosis. Here, we concluded that circWHSC1 functions as an unfavorable biomarker in BC patients, and overexpressing circWHSC1 remarkably accelerates BC cells' proliferation, reduces apoptosis and promotes metastasis, indicating that circWHSC1 is an oncogene in BC.
miRNAs have been a research hotspot in the field of oncology, including BC (26,27). miRNAs are dysregulated during BC progression, and several miRNAs serve as survival and prognosis biomarkers (28,29). Remarkedly, miRNAs modulate the proliferation, metastasis, chemosensitivity, apoptosis, and other malignant phenotypes (30). For example, miR-532-5p is overexpressed in BC tissues. It reduces RERG expression and activates the MAPK/ERK signaling to facilitate BC cell growth and migration (31). Besides, miR-302b is obviously downregulated in BC tissues and cell lines, and overexpressing miR-302b dramatically attenuates the growth and metastasis of BC cells (BT549 and MCF-7) (32). Moreover, miR-195-5p modulates the evolvement of BC. For instance, miR-195-5p is down-regulated in 40 BC specimens versus adjacent normal breast tissues. Overexpressing miR-195-5p notably represses BC cell proliferation, migration, and invasion by dampening CCNE1 expression (33). Consistent with earlier reports, miR-195-5p is found down-regulated in BC in this study Forced miR-195-5p up-regulation significantly declines cell proliferation and metastasis, confirming the tumor-repressive role of miR-195-5p in BC.
As reported, circRNAs function by sponging their target miRNAs. When more circRNAs are expressed in tumor cells, their targeted miRNAs become inactivated or degraded, resulting in lower miRNA levels in the cytoplasm (34,35). For example, Yang Rui et al. claimed that circAGFG1 accelerates BC cell proliferation and migration by sponging miR-195-5p (36). Presently, we found that circWHSC1 level was reversely related to miR-195-5p level in BC tissues. By conducting RIP and dual-luciferase reporter assay, we confirmed that circWHSC1 bound to miR-195-5p and inhibited its expression. Functionally, overexpressing circWHSC1 reversed the anti-tumor effects caused by miR-195-5p, which partly explained how circWHSC1 exerted its promotive effects on the development of BC.
Fatty acid metabolism dysregulation has been regarded as a component of malignant transformation in BC (37). FASN is a multifunctional enzyme responsible for endogenous fatty acids synthesis, which is up-regulated in various human malignancies and related to poor prognosis (38). SPIN1 triggers FASN signals through SREBP1c to regulate abnormal liver lipid metabolism and promote liver cancer cell growth (39). More impressively, FASN inhibition enhances the sensitiveness of tumor cells to chemotherapy (40) and radiotherapy (41). Those studies indicate that FASN acts as an oncogene in diversified tumors. In addition, FASN, an important functional target of miR-127, also makes a great contribution to BC. miR-127 targets FASN to reduce the viability and motility of triple-negative BC cells (42). In this study, the bioinformatic analysis showed that miR-195-5p had binding sites with FASN. Further experiments exhibited that FASN was obviously overexpressed in BC and was reversely related to miR-195-5p. circWHSC1 was positively linked to FASN expression, and circWHSC1 overexpression significantly increased FASN expression in BC. Hence, circWHSC1 could promote BC evolvement by modulating the miR-195-5p/FASN axis. Protein kinase AMP-activated catalytic subunit alpha 1 (AMPK) is a cellular energy transducer that is conserved in all eukaryotic cells. It controls the activities of various key metabolic enzymes via phosphorylation modification (43). On the other hand, the mechanistic target of rapamycin kinase (mTOR) belongs to the phosphatidylinositol kinase-related kinase family. These kinases coordinate cellular responses to stresses such as DNA injury and nutrient insufficiency (44). During cancer development, AMPK/ mTOR pathway becomes dysregulated (45). What is more, FASN enhances CRC cells' proliferation and metastasis by attenuating the AMPK/mTOR pathway activation (46). Here, we observed that overexpressing miR-195-5p increased AMPK phosphorylation and dampened mTOR phosphorylation. Overexpressing circWHSC1 had the opposite effects. Meanwhile, the influence of miR-195-5p on the AMPK/mTOR axis activation was mostly altered by circWHSC1 up-regulation. We treated BC cells with the FASN inhibitor Cerulenin (47) or the AMPK agonist Metformin (48). The proliferation, invasion and migration of BC cells with circWHSC1 overexpression were notably inhibited. Considering that circWHSC1/miR-195-5p had obvious effects on modulating FASN expression, we believed that circWHSC1 regulated the FASN/AMPK/mTOR axis by modulating miR-195-5p expression. Overall, our research confirms that circWHSC1 is a good prognostic biomarker for BC. In addition, circWHSC1 promotes BC cell proliferation and metastasis as a therapeutic target for BC. Therefore, it is believed that there is a novel circWHSC1-miR-195-5p-FASN/AMPK/mTOR axis in BC ( Figure 9). This study offers a better insight into the molecular mechanism of BC, which is helpful for BC's early diagnosis and treatment.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Ethics Review Board of Minhang Hospital, Fudan University. The patients/participants provided their written informed consent to participate in this study. The animal study was reviewed and approved by The Ethics Review Board of Minhang Hospital, Fudan University. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.