Chemotherapy-induced exosomal circBACH1 promotes breast cancer resistance and stemness via miR-217/G3BP2 signaling pathway

The human BC cell lines MDA-MB-231, MDA-MB-468, MCF-7 were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). The non-cancerous mammary epithelial cell line MCF-10A the human umbilical vein endothelial cells (HUVECs) were purchased from ATCC (Manassas, VA, USA). The MCF-7, HUVECs, MDA-MB-468 and MDA-MB-231 cells were cultured in RPMI 1640 (Gibco) containing 10% FBS at 37 °C under a humidified chamber supplemented with 5% CO₂. The MCF-10A cells were cultured in DMEM/F12 (invitrogen, Carlsbad, CA, USA) containing 10% FBS under a humidified atmosphere with 5% CO₂ at 37 °C.

A total of 9 BC patients were collected from 2021 to 2022 in Zhejiang Provincial People’s Hospital. The Blood of BC patients were collected before and after PTX treatment. The BC tissues and precancerous tissues were collected from the BC patients, then stored at − 80 °C for further analysis. The written informed consent forms were signed by all the BC patients. All the clinical experiments were approved by the Ethics Committee of Zhejiang Provincial People’s Hospital (No. 2022268).

Cells were treated with 100 nmol/l PTX, or PBS as control for 48 h. Then the supernatants of the medium were collected, which were centrifuged at 500g for 5 min, 2000g for 15 min, and 10,000g for 20 min at 4 °C to remove cells, debris, and large vesicles. Exosomes were obtained by ultracentrifugation of the supernatants at 110,000×g for 70 min. Exosomes were resuspended in PBS for further analysis. which were validated by transmission electron microscopy and Western blotting. ExoQuick Exosome Precipitation Solution (SBI System Biosciences, Mountain View, CA, USA) was used to isolate exosomes from plasma.

Total RNA extracted from exosomes derived from PTX-treated normal mammary epithelial cells (n = 4) and exosomes derived from PTX-treated MCF-7 cells (n = 6) RNA digestion was performed using RNase R (Epicentre, Madison, WI, USA) to remove linear RNA. According to the manufacturer’s protocol of the cRNA Amplification and Labeling Kit (KangChen, Shanghai, China), random primers were used to amplify the enriched circRNA and were subsequently transcribed into fluorescent circRNA. The amplified circRNA was purified using the RNeasy Mini kit (Qiagen, Düsseldorf, Germany). The labeled RNA was hybridized to the human circRNA Array (KangChen, Shanghai, China). The circRNA array was designed to have four identical arrays per slide, and each array contained probes to interrogate approximately 13,524 human circRNAs. Each circRNA was detected simultaneously by one long probe and one short probe.

The total RNA extraction was isolated from BC tissues and BC cells using the RNeasy Mini Kit (Qiagen, USA). Then the RNAs were converted into cDNA with the reverse Transcription Kit (Qiagen). The miScript SYBR Green PCR Kit (Qiagen) was used for amplification of cDNA and the RT-qPCR was performed on ABI PRISM 7500 real-time PCR System. The amplification program was set as follows: 50 °C for 2 min, 95 °C for 10 min followed by 50 cycles of 95 °C for 15 s, 72 °C for 30 s, 72 °C for 5 min. The forward primers of circBACH1: 5′-ACTGTGGATCAGCTCCCTTTG-3′, reverse primers of circBACH1: 5′-CACAGCCAATGTGTGAACCT-3′. The forward primers of miR-217: 5′-TACT GCATCAGGAACTGATTGGA-3′, reverse primers of miR-217: 5′-TCCAATCAGTTCCTGATG-3′. The forward primers of G3BP2: 5′-CCGGCAGAACCTGTTTCTCT-3′, reverse primers of G3BP2: 5′-GACCTCGATCATTGCGCCTA-3′. The relative RNA expression was calculated by the standard 2^−ΔΔCt method and the experiments were repeated three times.

The isolated exosomes were fixed using 4% paraformaldehyde. Then the fixed exosomes were adsorbed onto a copper grids. The copper grids with adsorbed exosomes were stained with 2% uranyl acetate for 30 s. At last, the images were captured by the transmission electron microscopy (TEM, JEOL,USA).

The lysis buffer (100 mM Tris–HCl, 2% SDS, 1 mM Mercaptoethanol, 25% Glycerol) was used to extract the proteins. The BCA Protein Assay Kit (Beyotime, China) was applied to quantify the proteins. The proteins were separated by 10% polyacrylamide gel electrophoresis (SDS-PAGE), then they were transferred to nitrocellulose (NC) membrane. The NC membranes were blocked with 5% BSA for 1 h at room temperature. After washed with PBS, they were incubated with the primary antibody against CD81 (1:1000, Abcam, Cambridge, UK), CD63 (1:1000, Abcam, Cambridge, UK), TSG101 (1:1000, Abcam, Cambridge, UK), CD44 (1:1000, Abcam, Cambridge, UK), CD133 (1:1000, Abcam, Cambridge, UK), Oct-4 (1:1000, Abcam, Cambridge, UK), G3BP2 (1:1000, Cell Signaling Technology, Danvers, MA, USA), β-actin (1:1000, Cell Signaling Technology, Danvers, MA, USA) at 4 °C overnight. Next, the secondary antibody (1:7500, Abcam, Cambridge, UK) was incubated with the NC membranes at room temperature for 2 h. Finally, the chemiluminescence detection system was performed to visualize the protein bands, which was analyzed by the Image J software (NIH).

MCF-7 cells were collected and used for nuclear and cytoplasmic RNA detection. For nuclear and cytoplasmic RNA separation, the cells were collected and extracted using a PARIS kit (Life Technologies, Carlsbad, CA, USA).

The circBACH1 probes labeled by Cy3 were designed and synthesized by GenePharma (Shanghai, China). The circBACH1 probe was: 5′-GTCACCAGCTTCTCAATTTTT-3′. The probes were incubated in water at 75 °C for 5 min and immediately at 0 °C for 10 min for annealing. The cells were fixed with solution containing 4% paraformaldehyde and 0.5% Triton X-100 for 20 min. After dehydration with 70%, 95% and 100% ethanol, the cells were incubated with the annealed probes on the slides at 37 °C overnight for hybridization. The slides were incubated in washing buffer at 48 °C for 20 min after washed with hybridization buffer. The 4,6- diamidino- 2- phenylindole (DAPI, Beyotime) was added to stain the cyteblasts before sealed with Nail Polish. The results were observed for four fields using a fluorescence microscope (Leica, Wetzlar, Germany) and analyzed by ImageJ software (NIH, Bethesda, MD, USA).

Northern blot was performed as previously described (29). Briefly, 10 µl RNA was electrophoresed and transferred onto a positively charged nylon membrane (GE Healthcare, Shanghai, China). Hybridization was performed overnight with biotin-labeled circBACH1 probe. GAPDH was used as control. The membranes were incubated with HRP-linked streptavidin for 15 min, then incubated with chemiluminescent substrate and the blots were detected using the chemiluminescent detection system.

After treatment with PTX-EXO and/or si-circBACH1 for 24 h, cells were seeded at a density of 5 × 10⁴, and cultured in DMEM/F12 supplemented with B27 (Invitrogen), 20 ng/ml EGF (Sigma-Aldrich), 10 ng/ml bFGF (Sigma-Aldrich), and 20 ng/ml insulin-like growth factor I (Invitrogen) for 7 days. Mammosphere was calculated by larger than 50 μm in diameter under a microscope (Olympus, Tokyo, Japan).

PTX-EXO and/or si-circBACH1 treated cells in the presence of PTX were dual-labeled with a APC-conjugated anti-CD44 antibody and an FITC-conjugated anti-CD24 antibody (BD Pharmingen, San Diego). 1 × 10⁵ cells were labeled and incubated with FITC-conjugated anti-CD24 and APC-conjugated anti-CD44 antibody at 4 °C in the dark for 1 h. The cells were washed and then analyzed by flow cytometer (BD FACS Aria II).

The BC cells were harvested to detect the cell viability by CCK-8 (Dojindo, Kumamoto, Japan) according to the manufacturer’s instructions. Briefly, 100 μl Cell suspension containing 5 × 10³ cells were seeded onto 96-well culture plates with corresponding treatment. After incubation with PTX (100 nmol/l), PBS-EXO, PTX-EXO, si-circBACH1 treatment for 0 h, 24 h, 48 h, 72 h, 96 h, 10 μl CCK-8 solution was added and incubated for 2 h at 37 °C respectively. Then, the absorbance at 450 nm was determined with a microplate reader (ELX800, BioTeK, USA).

The miR-217 mimics, inhibitors, si-circBACH1, and the negative control (NC) were constructed and synthesized by Sangon Biotech (Shanghai, China). The circRNA overexpression vector (#GS0108) was obtained from Geneseed (Guangzhou, China). The BC cells were transfected with the circBACH1 vectors, miR-217 inhibitors or si-circBACH1 using the Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) with the manufacturer’s protocol.

The Matrigel-coated transwell cell culture chambers (BD Matrigel Invasion Chamber, BD Biosciences, USA) was applied to determine the migration capability of the BC cells with corresponding treatment. About 1 × 10⁴ BC cells were suspended in 200 μl medium without serum and seeded into the upper chambers of each transwell (8 μm pore size) coated with Matrigel (BD Biosciences, USA). Medium containing 10% FBS was added to the bottom chamber. After incubation with PTX, PBS-EXO, PTX-EXO, si-circBACH1 treatment for 24 h, cells in the upper chamber were removed with cotton swabs and the cells on the bottom chamber were fixed by 4% formaldehyde for 15 min, stained for 20 min using 0.1% crystal violet. An inverted light microscope was applied to count the migration cells (Olympus, Tokyo, Japan).

HUVECs were cultured in DMEM without serum for 24 h, then transferred into 24-well plates pre-coated with Matrigel (Invitrogen) and stimulated for 12 h with MCF-7 medium post PTX, PBS-EXO, PTX-EXO or si-circBACH1 treatment. The relative length of tubes in HUVECs were measured by Image J (NIH).

The wild type (WT) or the mutant type (Mut) of circBACH1 and G3BP2 gene fragments were respectively cloned into psiCHECK-2 vector (Promega, Madison, WI). For circBACH1-miR-217 luciferase assay, 0, 1, 5, 10 nm of miR-217 was co-transfected with WT or Mut circBACH1 luciferase reporter into 4 × 10⁵ MCF-7 cells, respectively. For G3BP2-miR-217 luciferase assay, 0, 1, 5, 10 nm of miR-217 was cotransfected with WT or Mut G3BP2 luciferase reporter into MCF-7 cells, respectively. After incubation for 48 h, dual Luciferase Reporter Assay System (Promega) was applied to determine the luciferase activity of the MCF-7 cells.

MCF-7 cell lysates from different groups were incubated with RIP buffer containing magnetic beads conjugated with anti-human argonaute 2 (Ago2) or IgG antibodies (Millipore, Billerica, MA, USA) overnight under 4 °C. Then streptavidin-coated magnetic beads were added for incubation for 2 h. The isolated and purified RNAs was subjected to qRT-PCR measurement. The immunoprecipitated RNAs were then extracted and detected by qRT-PCR to confirm the enrichment of binding targets and the products were then subjected to agrose gel electrophoresis. The primers used for detecting circBACH1 or miR-217 were listed in RT-qPCR.

Paraffin-embedded slides of BC tissue were dewaxed in xylol for 20 min and rinsed with 100% ethanol for the staining process. Each slide was incubated with G3BP2 (1:200, Cell Signaling Technology, Danvers, MA, USA) for 24 h at 4 °C. Immunostaining was visualized with the DAB staining system (Beyotime). These slices were Images were visualized using a microscope system (Leica, Wetzlar, Germany).

All experiments were repeated for 3 times. Statistics were presented as mean ± SD. Student’s t-test was used to assess the statistical significance for comparisons between two groups. One way ANOVA was used for the comparison of multiple groups (> 2). GraphPad Prism (v9.0, GraphPad Software Inc., San Diego, CA, USA) was applied for statistical analyses. P < 0.05 was considered as a statistically significant difference.

To identify the exosomes derived from BC cells, the morphology and the surface markers of the exosomes were measured by TEM and western blotting, respectively. The representative micrograph showed that the size of exosomes is nearly 50-150 nm diameter (Fig. 1A). Moreover, the exosome markers, including CD81, CD63, and TSG101, were positively expressed pre- or post-PTX treatment (Fig. 1B). CircRNA profiles were characterized by circRNA microarray analysis between four gorups of exosomes derived from PTX-treated normal mammary epithelial (MCF-10A) cells (C1, C2, C3, and C4) and six gorups of exosomes derived from PTX-treated MCF-7 cells (M1, M2, M3, M4, M5 and M6). In total, 13,524 circRNAs were detected, and 10,411 circRNAs were identified as being dysregulated in BC cells. Among them, 5,015 were up-regulated and 5,396 were down-regulated in M group compared with C group. A volcano plot was drawn using the two factors of P value and FC value for the significantly differently expressed circRNAs obtained from the above analysis (Fig. 1C). The hierarchical clustering analysis of the differentially expressed circRNAs clearly differentiated M group from C group (Fig. 1D). As a result, circBACH1 was upregulated in the exosomes derived from PTX-treated BC cells.

CircBACH1 was reported to promoted tumor progress in various cancer types [25, 29‐31]. The expression of circBACH1 was significantly increased in the exosomes derived from PTX treated-BC cells (PTX-EXO) compared with control (Fig. 1E). To explore whether exosomes could deliver circBACH1 to the BC cells, we measureded the circBACH1 expression in MDA-MB-231 and MCF-7 post PTX-EXO treatment. The circBACH1 expression was significantly increased in the PTX-EXO-treated MCF-7, MDA-MB-231 and MDA-MB-468 cells compared with control (Fig. 1F). Interestingly, we also found that the expression of circBACH1 was significantly increased in the exosomes derived from PTX-treated BC patients compared with pre-PTX treatment (Fig. 1G). FISH assay showed that circBACH1 was mainly located in the cytoplasm of MCF-7 cells (Fig. 2A and B). Compared with control, the relative circBACH1 expression was significantly increased in the PTX-treated MDA-MB-231 cell lysates, while it was not significantly changed in the PTX-treated MCF-7 cell lysates (Fig. 2C). Consistently, the circBACH1 levels were apparently increased in the BC tissues compared with the adjacent tissues (Fig. 2D). We knocked down circBACH1 by transfection with si-circBACH1. The efficiency of si-circBACH1 transfection was characterized by northern blot (Fig. 2E). The si-circBACH1 #1 was used in the further investigation. To determine the effect of circBACH1 on the function of BC cells, the MCF-7 and MDA-MB-231 cells were treated with PTX and PTX-EXO or si-circBACH1. We observed PTX-EXO enhanced mammosphere formation in PTX treated MCF-7 and MDA-MB-231 cells, whereas loss of circBACH1 significantly attenuated mammosphere formation in PTX-EXO treated MCF-7 and MDA-MB-231 cells in the presence of PTX (Fig. 2F and H). In addition, we found that the proportion of CD44⁺CD24⁻ cells were enhanced in PTX + PTX-EXO group, whereas downregulation of circBACH1 significantly decreased the proportion of CD44⁺CD24⁻ cells in PTX-EXO treated MCF-7 cells in the presence of PTX (Fig. 2G and I).

To determine whether PTX-EXO mediated the cell viability and stemness in the BC cells, the MCF-7 and MDA-MB-231 cells were treated with PTX and PBS-EXO or PTX-EXO. Then we analyzed the cell viability and stemness of the BC cells after the corresponding treatment to confirm the PTX resistance of the BC cells. The cell viability of MCF-7 and MDA-MB-231 cells was markedly increased in the PTX + PTX-EXO group compared with the PTX + PBS-EXO group (Fig. 3A and B). Moreover, we compared the IC₅₀ of PTX between breast cancer cells and PTX-EXO treated breast cancer cells. We found that the IC₅₀ for PTX in MDA-MB-231 and MCF-7 cells was nearly 10 nM and 5 nM, respectively, whereas the IC₅₀ for PTX was not achieved in PTX-EXO treated MDA-MB-231 or MCF-7 cells (Additional file 1: Fig. S1A and B). The stemness markers such as CD44, CD133, and Oct were increased in the PTX + PTX-EXO group in comparison to the PTX + PBS-EXO group in MCF-7 and MDA-MB-231 cells (Fig. 3C and D). To explore the effect of exosomal circBACH1 on PTX-resistance, circBACH1 was downregulated by si-circBACH1 in BC cells. Compared with the PTX + PTX-EXO group, the cell viability was significantly decreased in the PTX + PTX-EXO + si-circBACH1 group in MDA-MB-231 and MCF-7 cells (Fig. 3A and B). Interestingly, the expression levels of CD44, CD133, and Oct were reduced in the PTX + PTX-EXO + si-circBACH1 group in comparison to the PTX + PTX-EXO group in MDA-MB-231 and MCF-7 cells (Fig. 3C and D). Consistently, the expression of CD44, CD133, Oct were obviously increased by transfected with circBACH1 vectors compared with control, whereas decreased by transfected with si-circBACH1 compared with si-NC (Fig. 3E and F). The overexpression of circBACH1 was determined by northern blot (Additional file 1: Fig. S1C). Collectively, the results revealed that PTX-EXO promoted the PTX-resistance of BC cells through enhancing cell viability and stemness, which was reversed by the downregulation of circBACH1.

To further confirm that the PTX-EXO mediated the PTX-resistance in the BC cells, the migration was analyzed in MCF-7 and MDA-MB-231 cells. The migrated cell number was significantly increased in the PTX + PTX-EXO group compared with the PTX + PBS-EXO group in MCF-7, MDA-MB-231, and MDA-MB-468 cells (Fig. 4A–D, Additional file 1: Fig. S2A and B). However, the migrated cell number was significantly decreased in the PTX + PTX-EXO + si-circBACH1 group compared with PTX + PTX-EXO group in MCF-7, MDA-MB-231, and MDA-MB-468 cells (Fig. 4A–D, Additional file 1: Fig. S2A and B). To determine whether PTX-EXO affected the angiogenesis of HUVECs, we monitored the tube formation of HUVECs following stimulation with MCF-7 cell medium post corresponding treatment. Compared with the control group, the relative length of tubes in the PTX + PTX-EXO group was substantially increased (Fig. 4E and F). However, the relative length of tubes was significantly decreased in the PTX + PTX-EXO + si-circBACH1 group in comparison to the PTX + PTX-EXO group (Fig. 4E and F). The results demonstrated that PTX-EXO promoted the migration of BC cells and the tube formation of HUVECs, whereas the downregulation of circBACH1 reduced migration to improve PTX-sensitiveness in BC cells, as well as inhibited the PTX-EXO-induced angiogenesis of HUVECs.

To further explore the downstream of circBACH1 regulating PTX-EXO-induced resistance, the interacted-miRNAs were predicted based on the target prediction program of circBank, NIH, and ENCORI. As a result, the overlapped miRNA (miR-217) was obtained from the Venn diagram of the three databases (Fig. 5A). The predicted target sites between circBACH1 and miR-217 were shown in Fig. 5B. To clarify the interaction between miR-217 and circBACH1, the vectors containing wild type or mutant circBACH1 were co-transfected with miR-217 into MCF-7. The luciferase activity diminished gradually with the increased-concentration of miR-217 mimics following co-transfected with the wild type circBACH1, whereas the luciferase activity barely changed with the miR-217 and the mutant circBACH1 co-transfection (Fig. 5C). RNA-IP results showed that the relative abundance of circBACH1 and miR-217 in the anti-Ago2 group significantly increased compared to the anti-IgG group, whereas in the miR-217 inhibitors group, the relative abundance of circBACH1 and miR-217 immunoprecipitated with Ago2 was lower than that in the control group (Fig. 5D and E). To further validate the interaction between circBACH1 and miR-217, miR-217 expression levels were detected following PTX-EXO treatment in BC cells. The results showed that the expression levels of miR-217 were significantly decreased in PTX-EXO-treated MDA-MB-231 and MCF-7 cells compared with the control (Fig. 5F). Consistently, the expression levels of miR-217 were also significantly reduced in PTX-treated BC tissue in comparison to the adjacent tissue (Fig. 5G). Moreover, the expression levels of miR-217 reduced significantly following the overexpression of circBACH1 compared with control, whereas miR-217 expression was obviously increased following transfection with si-circBACH1 compared with control in MCF-7 cells (Fig. 5H and I). In addition, we also found that the mammosphere formation was significantly reduced in miR-217 mimics-treated MCF-7 cells compared with control (Additional file 1: Fig. S2C and D).

miR-217 was involved in the suppression of the proliferation, invasion, and migration of cancer cells [32‐34]. To explore the downstream mechanism of miR-217, the target gene was predicted by TargetScan 7.2 program. G3BP2 was an oncogene involved in breast cancer progression and metastasis [35]. The predicted binding sites of miR-217 in G3BP2 mRNA 3'UTR were provided by the TargetScan 7.2 program (Fig. 6A). The luciferase reporter vectors containing wild type or mutant G3BP2 and miR-217 were co-transfected into MCF-7 cells. Interestingly, the luciferase activity of wild type G3BP2 was significantly reduced in the increased concentration of miR-217 compared with control, whereas the luciferase activity of mutant G3BP2 barely changed between miR-217 group and control group in MCF-7 cells (Fig. 6B). We observed that the expression levels of G3BP2 were significantly increased in PTX-EXO-treated MDA-MB-231 and MCF-7 cells compared with the control (Fig. 6C). Moreover, overexpression of circBACH1 increased the G3BP2 levels in MCF-7 cells compared with control (Fig. 6D). We determined the expression of the G3BP2 proteins in MCF-7 cells by western blot assay after cotransfected with circBACH1 and miR-217. The G3BP2 proteins were upregulated by circBACH1 transfection, while downregulated by miR-217 transfection compared with control in MCF-7 cells (Fig. 6E). However, circBACH1 and miR-217 cotransfection suppressed the expression of G3BP2 proteins compared with circBACH1 treatment in MCF-7 cells (Fig. 6E). To elucidate the effect of G3BP2 on BC progression, G3BP2 was downregulated in MCF-7 cells. The protein expression of G3BP2 was testified following downregulation of G3BP2 (Fig. 6F). To explore the effect of G3BP2 on BC cell function, cell migration and mammosphere formation were determined following downregulation of G3BP2. The migrated cell numbers were significantly reduced by treatment with si-G3BP2 #1 compared with control in MCF-7 cells (Fig. 6G and H). Downregulation of G3BP2 significantly attenuated mammosphere formation compared with control in MCF-7 cells (Fig. 6L and J). Consistent with in vitro results, the G3BP2 levels were increased in BC tissue compared with the adjacent tissue, which was confirmed by IHC assay and RT-PCR (Fig. 6K and L).