Ropivacaine inhibits CSC-like properties of breast cancer cells

To determine the effects of ropivacaine on the CSC-like properties of breast cancer cells, MCF-7 and MDA-MB-231 cells were treated with ropivacaine and subjected to functional characterization. A significant decreased CD44+/CD24− CSC subpopulations (Fig. 1A) and mammosphere formation (Fig. 1B) were observed in breast cancer cells upon ropivacaine treatment. We also observed ropivacaine significantly inhibited migration and invasion of breast cancer cells (Fig. 1C and D). Consistently, ropivacaine treatment resulted in reduced expression levels of stem cell markers, including CD133, OCT4 and SOX2 (Fig. 1E). Given that CSCs are associated with chemoresistance [31], whether ropivacaine could resensitize doxorubicin-resistant MCF-7 cells (MCF-7/ADR) to doxorubicin was examined. As anticipated, ropivacaine treatment decreased cell viability (Fig. 1F) and increased apoptosis (Fig. 1G) in MCF-7/ADR cells treated with doxorubicin. Taken together, our results suggest ropivacaine inhibits the stemness of breast cancer cells.

Fig. 1

Effects of ropivacaine on CSC-like phenotypes of breast cancer cells in vitro. A CD44+/CD24.− subpopulation in MCF-7 and MDA-MB-231 treated with ropivacaine (10 μM) or negative control for 48 h was measured by FACS analysis. B Mammosphere formation of MCF-7 and MDA-MB-231 treated with ropivacaine (10 μM) or negative control for 48 h (scale bar = 100 μm). C-D Migration and invasion of MCF-7 and MDA-MB-231 treated with ropivacaine (10 μM) or negative control for 48 h were examined by transwell assays (scale bar = 100 μm). E Stem cell markers (CD133, OCT4, SOX2) in MCF-7 and MDA-MB-231 treated with ropivacaine (10 μM) or negative control for 48 h were analyzed by western blot. F Cell viability of ropivacaine (10 μM)- or negative control-treated MCF-7/ADR cells exposed to the indicated concentration of ADR. G Cell apoptosis of ropivacaine (10 μM)- or negative control-treated MCF-7/ADR cells exposed to ADR (10 μM) were assessed by FACS analysis. Results are shown are shown as mean ± S.D from at least three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001 (Two-way ANOVA test in F, others unpaired two-tailed Student’s t test)

Ropivacaine suppresses metastasis and tumorigenic capacity

To assess the effects of ropivacaine on the CSCs in vivo, a tail-vein injection model was employed to establish lung metastasis by using the MDA-MB-231 cells. Bioluminescence imaging revealed a significant reduction in pulmonary metastasis in mice treated with ropivacaine (Fig. 2A and B), which was confirmed by H&E staining of the lung sections from each group (Fig. 2C). Additionally, ropivacaine treatment led to decreased cell proliferation in the lung metastatic lesions (Fig. 2D and E), indicating its inhibitory effects on the colonization of metastatic cells. A higher survival rate for the tumor-bearing mice in ropivacaine-treated group was also observed as a result (Fig. 2F).

Fig. 2

Effects of ropivacaine on CSC-like phenotypes of breast cancer cells in vivo. A Bioluminescence images of the metastatic burden of ropivacaine (40 μmol/kg, intraperitoneal injection, once every other day)- or negative control-treated mice at the indicated days after injection with MDA-MB-231 cells intravenously (n = 4). B-C Bioluminescence images (B) and H&E staining (C) of lungs from ropivacaine- or negative control-treated mice intravenously injected with MDA-MB-231 cells (scale bar = 100 μm) (n = 4). D-E Ki-67 staining of lung sections from ropivacaine- or negative control-treated mice intravenously injected with MDA-MB-231 cells (scale bar = 100 μm) (n = 4). F Survival curve of mice intravenously injected with MDA-MB-231 cells and treated with ropivacaine or negative control (n = 12). G-H The incidence of tumors in mice injected with different numbers of MDA-MB-231 cells and treated with ropivacaine or negative control (n = 6). I The tumor-free survival curves of the mice that were inoculated with different numbers of MDA-MB-231 cells with and without ropivacaine treatment (n = 6). *p < 0.05; **p < 0.01; ***p < 0.001 (log-rank test in F and I, others unpaired two-tailed Student’s t test)

We also evaluated the effect of ropivacaine on the tumor-initiating capability of breast cancer cells by using limited dilution assay. 1 × 106 MDA-MB-231 cells formed tumor xenografts with 100% efficiency in the control group, but the tumor formation efficiency of the ropivacaine-treated group decreased to 60%. When 1 × 105 and 1 × 104 cells per site were implanted, the tumor formation efficiency in ropivacaine-treated group decreased to 20% and 0%, whereas the control group retained 60% and 40%, respectively (Fig. 2G and H). Not surprisingly, the onset of tumor growth in the ropivacaine-treated group was delayed compared with its control counterparts (Fig. 2I). Collectively, our study demonstrates that ropivacaine exhibits anti-CSC effects in vivo.

Ropivacaine represses GGT1 expression

We subsequently conducted RNA-seq analysis of MCF-7 cells treated with or without ropivacaine for the mechanistic insight. The scatter plot, volcano plot and heat map displayed dysregulated genes in MCF-7 cells in response to ropivacaine treatment (Fig. 3A-C). The upregulated genes in mammosphere cells compared with their parental MCF-7 and MDA-MB-231 cells were screened out in silico (GSE182532 and GSE136190). Upregulated genes in breast cancer tissues compared to their normal counterparts from TCGA dataset were selected. By combining the data from all these four cohorts, we focused on three candidate genes: GGT1, TAPAN1 and ANO7 (Fig. 3D). Among the identified genes, GGT1 emerged as the most down-regulated target responding to ropivacaine treatment (Fig. 3E). We further validated the suppressive effects of ropivacaine on GGT1 at the RNA and protein levels in MCF-7 and MDA-MB-231 cells using qRT-PCR and Western blot analysis (Fig. 3F and G). These results suggest GGT1 is a downstream target gene of ropivacaine.

Fig. 3

Identification of GGT1 as a ropivacaine-responding gene in breast cancer cells. A-B RNA-seq data of ropivacaine- or negative control-treated MCF-7 cells. The scatter plot (A) and volcano map (B) showing differentially-expressed genes. C Heatmap representing the top 50 up-regulated and down-regulated genes in MCF-7 cells treated with ropivacaine or negative control. D The intersection among four databases. E RNA levels of GGT1, TSPAN1 and ANO7 in MCF-7 cells treated with ropivacaine or negative control from RNA-seq data. F-G RNA (F) and protein (G) level of GGT1 in MCF-7 and MDA-MB-231 treated with ropivacaine (10 μM) or negative control for 48 h were was determined by qRT-PCR and western blot, respectively. Results are shown are shown as mean ± S.D from at least three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001 (unpaired two-tailed Student’s t test)

NF-κB is an upstream transcriptional factor of GGT1

For mechanistic insight of ropivacaine repressed GGT1 expression, we employed the PROMO (https://alggen.lsi.upc.es/cgi-bin/promo_v3/promo/promoinit.cgi?dirDB=TF_8.3) online prediction tool to identify potential upstream transcriptional factors of GGT1. NF-κB was predicted to potentially bind to the GGT1 promoter region. This prediction was supported as p65 overexpression significantly increased GGT1 expression (Fig. 4A and B), while the NF-κB inhibitor, BAY 11–7802, reduced GGT1 expression (Fig. 4C and D), at both protein and RNA levels in MCF-7 and MDA-MB-231 cells. Further, luciferase reporter assay was employed by inserting GGT1 promoter region containing NF-κB-binding sites into pGL3-basic reporter vector (Fig. 4E). BAY 11–7802 remarkably inhibited GGT1 promoter elicited luciferase activity (Fig. 4F and G). ChIP-PCR assays was preformed to have confirmed the reduced enrichment of p65 at the GGT1 promoter region upon NF-κB inhibition (Fig. 4H and I). These results collectively demonstrate that NF-κB is a bona fide upstream transcriptional factor of GGT1 in breast cancer cells.

Fig. 4

NF-κB binds to the promoter region of GGT1 to activate its transcription. A Protein levels of GGT1 and p65 in MCF-7 and MDA-MB-231 cells transfected with p65 overexpression plasmid (P65-OE) or empty vector were examined by western blot. B RNA level of GGT1 in MCF-7 and MDA-MB-231 cells transfected with p65 overexpression plasmid (P65-OE) or empty vector was examined by qRT-PCR. C Protein levels of GGT1 and p65 in MCF-7 and MDA-MB-231 cells treated with BAY 11–7802 (1 μM) or DMSO were examined by western blot. D RNA level of GGT1 in MCF-7 and MDA-MB-231 cells treated with BAY 11–7802 (1 μM) or DMSO was examined by qRT-PCR. E Schematic of predicted binding sites between GGT1 promoter and p65. F-G Construction of luciferase reporter vectors comprising p65 binding sites in the DNA promoter region of GGT1 (2 kb). Dual-luciferase reporter assays were performed by transfecting the promoter region of GGT1 (GGT1 Pro) or pGL3-basic plasmid in MCF-7 and MDA-MB-231 cells treated with BAY 11–7802 (1 μM) or DMSO. H-I The binding of p65 to the GGT1 promoter region in MCF-7 and MDA-MB-231 cells treated with BAY 11–7802 (1 μM) or DMSO was examined by ChIP assay. Results are shown are shown as mean ± S.D from at least three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001 (unpaired two-tailed Student’s t test)

Ropivacaine interacts with AKT1 directly to repress NF-κB signaling

To identify the direct target of ropivacaine, SwissTargetPrediction (http://www.swisstargetprediction.ch/) with further bioinformatics analysis were performed to have suggested AKT1 as a potential direct binding partner. It is intriguing to note that AKT1 is widely known to activate NF-κB signaling pathway via catalyzing phosphorylation of iκB and releasing its inhibitory effect on NF-κB [32,33,34]. To confirm the direct interactions between ropivacaine and AKT1, a docking analysis of ropivacaine (the 2D and 3D structures are shown in Fig. 5A) and human AKT1 was performed by using Auto Dock Vina software. The binding poses and interactions of ropivacaine hydrochloride with AKT1 were obtained with Autodock Vina v.1.2.2 and binding energy for interaction was generated (Fig. 5A). Results showed that ropivacaine hydrochloride bound to domain area of AKT1 kinase through a hydrogen bond (THR291) and six hydrophobic interactions (LEU156, VAL164, ALA177, TYR229, THR291, ASP292) (Fig. S1). Moreover, the predicted ropivacaine-binding region partially coincided with the ATP-binding region (from LEU156 to VAL164), suggesting that ropivacaine as a potential ATP-competitive inhibitor of AKT1. Ropivacaine hydrochloride and AKT1 exhibited low binding energy value of -6.567 kcal/mol, suggestinga highly stable binding. Further, CETSA showed that ropivacaine treatment effectively protected the AKT1 protein from temperature-dependent degradation (Fig. 5B and C). Consistently, ropivacaine significantly reduced AKT1 kinase activity, as determined by the ADP-Glo™ kinase assay (Fig. 5D) and led to a notably decrease in the capacity of AKT1 to bind to ATP (Fig. 5E). As expected, the interaction between ropivacaine and AKT1 resulted in markedly decreased expression of p-AKT1, p-iκBa and p-NF-κB (Fig. 5F). Taken together, these data indicate ropivacaine inhibits NF-κB activity and GGT1 expression via its direct binding to AKT1.

Fig. 5

Ropivacaine interacts with AKT1 directly to repress NF-κB signaling. A The 2D and 3D structure of ropivacaine hydrochloride and the predicted binding modes of ropivacaine hydrochloride with AKT1 kinase domain. B-C The stabilizing effect of ropivacaine on AKT1 protein was assessed by western blot in cellular thermal shift assay using the indicated concentrations of ropivacaine-treated MCF-7 and MDA-MB-231 cells. D The effect of ropivacaine on the activity of AKT1 kinase has been estimated through kinase activity assay. E The competitive binding relationship between ropivacaine and ATP was confirmed by using a pull-down assay. F Protein levels of p-AKT1, t-AKT1, p-iκBa, iκBa, p-NF-κB and NF-κB in the indicated concentrations of ropivacaine-treated MCF-7 and MDA-MB-231 cells were determined by western blot. Results are shown are shown as mean ± S.D from at least three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001 (unpaired two-tailed Student’s t test)

Elevated GGT1 is associated with poor prognosis of breast cancer

Analysis of GGT1 RNA levels in breast cancer tissue samples from the TCGA datasets revealed higher GGT1 expression compared to normal tissue (Fig. 6A). In consistence, IHC analysis from our own 70 patients' cohort also validated GGT1 overexpression compared with peritumor specimens (Fig. 6B). As showed in Supplementary Table 3, further analysis of correlation between GGT1 expression and clinicopathological features of these enrolled breast cancer patients indicated GGT1 expression was positively correlated with tumor stage (p = 0.0258) and lymph node metastasis (p = 0.0266). We also showed that GGT1 expression was up-regulated in the tumor group relative to the matched adjacent samples (Fig. 6C). Compared to HMEC-hTERT cells, GGT1 expression levels were higher in 6 breast cancer cell lines (Fig. 6D). Furthermore, GGT1 was up-regulated in mammospheres as compared with their parental cells (Fig. 6E). Kaplan–Meier survival analysis displayed that breast cancer patients with high level of GGT1 exhibited a shorter overall survival, compared with patients with lower levels of GGT1 (Fig. 6F). Collectively, these results suggest elevated GGT1 predicts poor prognosis of breast cancer patients.

Fig. 6

Higher GGT1 correlates to shorter overall survival of breast cancer patients. A mRNA level of GGT1 in 1113 breast cancer tissues and 113 benign normal tissues from TCGA database. B IHC analysis of GGT1 expression in breast cancer (Tumor) and peritumor tissues (Normal) from our cohort (n = 70), (scale bar = 100 μm). C Protein level of GGT1 in 4 breast cancer tissues and their adjacent normal tissues was determined by western blot. D Protein level of GGT1 in six breast cancer cell lines (MCF-7, MDA-MB-231, SKBR-3, T47D, BT-549 and SUM149) and a non-transformed mammary epithelial cell line (HMEC-hTERT) was determined by western blot. E Protein level of GGT1 in mammosphere derived from MCF-7 and MDA-MB-231 cells and their parental cells was evaluated by western blot. F Kaplan–Meier analysis of the relationship between GGT1 expression levels and overall survival of breast cancer patients from TCGA database. *p < 0.05; **p < 0.01; ***p < 0.001 (log-rank test in F, others unpaired two-tailed Student’s t test)

GGT1 depletion impairs stemness of breast cancer cells

To investigate the functional role of GGT1 in maintaining stemness in breast cancer cells, MCF-7 and MDA-MB-231 cells were stably transfected with shRNAs against GGT1 (Fig. 7A and B, Fig. S2A and B). GGT1 depletion led to significantly reduced expression of stem cell markers (Fig. 7C and Fig. S2C), decreased percentage of CD44+/CD24− CSC subpopulations (Fig. 7D and Fig. S2D), impaired mammosphere formation (Fig. 7E and Fig. S2E), as well as reduced migratory and invasive capability (Fig. 7F and Fig. S2F) in MCF-7 and MDA-MB-231 cells. Endogenous GGT1 in MCF-7/ADR cells was obviously silenced by transfection with specific siRNAs targetting GGT1 (Fig. S3A and B). An attenuated cell viability (Fig. 7G) while increased cell apoptosis (Fig. 7H) were observed in GGT1-depleted MCF-7/ADR cells exposed to doxorubicin, suggesting that ablation of GGT1 restored the sensitivity of MCF-7/ADR cells to doxorubicin.

Fig. 7

Effects of GGT1 on CSC-like phenotypes of MCF-7 cells in vitro. A-B RNA (A) and protein (B) level of GGT1 in MCF-7 cells stably transfected with shRNAs against GGT1 (shGGT1-1 and shGGT1-2) or negative control (shCtrl) were determined by qRT-PCR and western blot, respectively. C Stem cell markers (CD133, OCT4, SOX2) in GGT1-depleting MCF-7 cells were analyzed by western blot. D CD44+/CD24.− subpopulation in GGT1-depleting MCF-7 cells was measured by FACS analysis. E Mammosphere formation of GGT1-depleting MCF-7 cells (scale bar = 100 μm). F Migration and invasion of GGT1-depleting MCF-7 cells were examined by transwell assays (scale bar = 100 μm). G Cell viability of GGT1-depleting MCF-7/ADR cells exposed to the indicated concentration of ADR. H Cell apoptosis of GGT1-depleting MCF-7/ADR cells exposed to ADR (10 μM) were assessed by FACS analysis. Results are shown are shown as mean ± S.D from at least three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001 (Two-way ANOVA test in G, others unpaired two-tailed Student’s t test)

GGT1 depletion dampens lung metastasis and tumor-initiating ability

To further determine the roles of GGT1 in regulating CSC-like properties in vivo, 3 × 104 MDA-MB-231 cells stably transfected with shRNAs against GGT1 or negative control were injected into Balb/c nude mice via tail vein. Bioluminescence imaging displayed that GGT1 depletion significantly inhibited lung metastasis derived from the MDA-MB-231 cells (Fig. 8A and B). H&E staining revealed smaller metastatic lung lesions and less incidence of lung metastasis in mice injected with GGT1-silenced MDA-MB-231 cells (Fig. 8C). Consistently, decreased cell proliferation in the lung metastatic lesions of mice derived from GGT1-depleted MDA-MB-231 cells was observed (Fig. 8D and E), indicating the role of GGT1 in promoting the colonization of metastatic cells. Moreover, mice injected with GGT1-depleted MDA-MB-231 cells exhibited a higher survival rate (Fig. 8F).

Fig. 8

Effects of GGT1 on CSC-like phenotypes of breast cancer cells in vivo. A Bioluminescence images of the metastatic burden of mice at the indicated days after injection with MDA-MB-231-shCtrl or -shGGT1 cells intravenously (n = 4). B-C Bioluminescence images (B) and H&E staining (C) of lungs from mice intravenously injected with MDA-MB-231-shCtrl or -shGGT1 cells (scale bar = 100 μm) (n = 4). D-E Ki-67 staining of lung sections from mice intravenously injected with MDA-MB-231-shCtrl or -shGGT1 cells (scale bar = 100 μm) (n = 4). F Survival curve of mice intravenously injected with MDA-MB-231-shCtrl or -shGGT1 cells (n = 12). G-H The incidence of tumors in mice injected with different numbers of MDA-MB-231-shCtrl or -shGGT1 cells (n = 5). I The tumor-free survival curves of the mice that were inoculated with different numbers of MDA-MB-231-shCtrl or -shGGT1 cells (n = 5). J-K IF examination of GGT1 and CD44 expression in tumor sections from both lung tissues (J) and primary tumor tissues (K) derived from MDA-MB-231-shCtrl or -shGGT1 cells (scale bar = 100 μm) (n = 4).*p < 0.05; **p < 0.01; ***p < 0.001 (log-rank test in F and I, others unpaired two-tailed Student’s t test)

Limited dilution analysis was also conducted to determine whether GGT1 mediated tumorigenesis of breast cancer cells. 1 × 106 MDA-MB-231/shCtrl cells formed tumor xenografts with 100% efficiency, whereas the tumor formation efficiency of MDA-MB-231/shGGT1 cells decreased to 40%. When cells were implanted at a density of 1 × 105 and 1 × 104 cells per site, the tumor formation efficiency of MDA-MB-231/shGGT1 cells decreased to 20% and 0%, but the MDA-MB-231/shCtrl cells retained 60% and 40% (Fig. 8G and H). Furthermore, the onset of tumor growth of MDA-MB-231/shGGT1 cells was much delayed compared with its control counterparts (Fig. 8I). Additionally, IF examination of tumor sections from both lung tissues and primary tumor tissues indicated GGT1 depletion accompanied by low level of CD44 (Fig. 8J and K). Collectively, these results highlight the role of GGT1 in maintaining stem-like phenotypes of breast cancer.

GGT1 activates NF-κB signaling

To elicit the signaling pathways affected by GGT1 in the regulation of breast cancer stem cell-like phenotypes, we conducted RNA-seq analysis in GGT1-depleted MCF-7 cells (shGGT1) and control cells. The volcano plot displayed differentially expressed genes (Fig. 9A). KEGG analysis revealed that GGT1 depletion impinged on several signaling pathways, with the NF-κB signaling pathway scoring on the top (Fig. 9B and C). Consistently, GSEA analysis indicated NF-κB signaling pathway was inactivated in MCF-7 cells with GGT1 depletion (Fig. 9D). We subsequently showed that GGT1 depletion impaired NF-κB activity (Fig. 9E), while GGT1 overexpression enhanced NF-κB activity (Fig. 9F). Further, knockdown of GGT1 reduced expression of p-iκBα and p-NF-κB, whereas forced expression of GGT1 exhibited opposite effects (Fig. 9G and H). Based on the essential role of NF-κB signaling pathway in maintaining stemness of breast cancer cells [35,36,37], we thereby propose GGT1 activates NF-κB signaling pathway in breast cancer cells.

Fig. 9

GGT1 enhances activation of NF-κB signaling pathway. A RNA-seq data of MCF-7-shCtrl or -shGGT1 cells. The volcano map showing differentially-expressed genes. B KEGG analysis of the top 10 enriched pathways in MCF-7-shCtrl or -shGGT1 cells from RNA-seq data. C Heatmap showing differentially expressed genes in NF-κB signaling pathway in MCF-7-shCtrl or -shGGT1 cells. D GSEA analysis of the enrichment of NF-κB signaling pathway in MCF-7-shCtrl or -shGGT1 cells from RNA-seq data. E NF-κB transcriptional activity in MCF-7 and MDA-MB-231 cells stably transfected with shRNAs against GGT1 (shGGT1-1 and shGGT1-2) or negative control (shCtrl) was determined by NF-κB activation reporter assay. F NF-κB transcriptional activity in MCF-7 and MDA-MB-231 cells transfected with GGT1 overexpression plasmid (GGT1-OE) or empty plasmid (Vector) was determined by NF-κB activation reporter assay. G Protein levels of GGT1, p-iκBa, iκBa, p-NF-κB and NF-κB in GGT1-depleting and GGT1-overexpressing MCF-7 and MDA-MB-231 cells were determined by western blot. Results are shown are shown as mean ± S.D from at least three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001 (unpaired two-tailed Student’s t test)

NF-κB/GGT1 feedback loop mediates inhibitory effects of ropivacaine on breast CSC-like phenotypes

To determine whether ropivacaine-repressed stem cell-like traits is mediated by GGT1/NF-κB signaling pathway. MCF-7 and MDA-MB-231 cells were transfected with GGT1 ovexpression plasmid before ropivacaine treatment. Ropivacaine-induced suppression of GGT1, p-iκBα and p-NF-κB was significantly rescued by GGT1 overexpression (Fig. 10A and B, Fig. S4A and B). In line with GGT1/NF-κB signaling pathway, GGT1 overexpression reverted ropivacaine-repressed expression of stem cell markers (Fig. 10C and Fig. S4C), CD44+/CD24− CSCs subpopulation (Fig. 10D and Fig. S4D), mamosphere formation (Fig. 10E and Fig. S4E), migration and invasion (Fig. 10F and Fig. S4F). Additionally, forcing expression of GGT1 dramatically restored ropivacaine-induced decreased cell viability (Fig. 10G) and increased cell apoptosis (Fig. 10H) in MCF-7/ADR cells exposed to doxorubicin. Taken together, these findings highlight GGT1/NF-κB feedback loop was involved in mediation of the anti-CSCs effects of ropivacaine.

Fig. 10

The anti-CSC effects of ropivacaine on MCF-7 cells is dependent on NF-κB/GGT1 feedback loop. MCF-7 cells was transfected with GGT1 overexpression plasmid (GGT1-OE) or empty plasmid (Vector) and treated with ropivacaine (10 μM) or negative control for 48 h. A RNA level of GGT1 in the indicated treated-MCF-7 cells was determined by qRT-PCR. B Protein levels of GGT1, p-iκBa, iκBa, p-NF-κB and NF-κB in the indicated treated-MCF-7 cells were examined by western blot. C Stem cell markers (CD133, OCT4, SOX2) in the indicated treated-MCF-7 cells were analyzed by western blot. D CD44+/CD24.− subpopulation in the indicated treated-MCF-7 cells was measured by FACS analysis. E Mammosphere formation of the indicated treated-MCF-7 cells (scale bar = 100 μm). F Migration and invasion of the indicated treated-MCF-7 cells were examined by transwell assays (scale bar = 100 μm). G Cell viability of MCF-7/ADR cells transfected with GGT1 overexpression plasmid (GGT1-OE) or empty plasmid (Vector) and treated with ropivacaine (10 μM) or negative control for 48 h as indicated, exposure to the indicated concentration of ADR. H Cell apoptosis of the indicated treated-MCF-7/ADR cells exposed to ADR (10 μM) were assessed by FACS analysis. Results are shown are shown as mean ± S.D from at least three independent experiments.. *p < 0.05; **p < 0.01; ***p < 0.001 (Two-way ANOVA test in G, others unpaired two-tailed Student’s t test)