Abstract
Background/Aim: Breast cancer (BC) is the most common cancer and second leading cause of death in women worldwide. Triple-negative breast cancer (TNBC) is the most aggressive type of BC, while the treatment option is limited and has long been considered as a major unmet need. Meta-analysis indicated the anti-tumor potential of anti-depressants, especially selective serotonin-reuptake inhibitors (SSRIs). The SSRI fluoxetine has been shown to suppress BC and ovarian cancer cell growth; however, whether it suppresses tumor progression in vivo is unclear. Materials and Methods: We established an 4T1 bearing animal model, an orthotopic TNBC model, to identify the mechanism and therapeutic efficacy of fluoxetine. Results: Tumor growth evaluated by caliper and computed tomography scan demonstrated the inhibition effect by fluoxetine treatment. Immunohistochemistry showed that the expression of STAT3-mediated epithelial-to-mesenchymal transition (EMT) proteins and apoptosis-related proteins was decreased. Conclusion: Fluoxetine may induce an anti-TNBC effect via inactivating STAT3 signaling transduction and triggering the caspase-mediated apoptotic pathway.
Triple negative breast cancer (TNBC) is an aggressive subtype of breast cancer characterized by the absence of estrogen receptor (ER), progesterone receptor (ER), and human epidermal growth factor receptor 2 (HER2) expression. TNBC has poor outcome due to the high recurrence and metastasis rate (1, 2). The standard neoadjuvant chemotherapy regimen comprising adriamycin, cyclophosphamide, and taxane is used for the treatment of TNBC. For further improvement of pathological complete response of TNBC patients, potential drug combinations have been developing. New therapeutic strategies such as the combination of poly ADP-ribose polymerase 1 (PARP-1) inhibitor and/or immune check point inhibitor has been shown to increase anti-TNBC efficacy of chemotherapy (3, 4).
Signal transducer and activator of transcription 3 (STAT3) is an oncogenic transcription factor that participates in tumor growth, angiogenesis, metastasis, and evasion of apoptosis by inducing the expression of downstream target genes (5). Constitutively activated STAT3 signaling is an essential mediator of TNBC. The inhibition of STAT3 signaling has been found to mediate TNBC suppression but also up-regulate the sensitivity of TNBC cells to therapeutic agents including olaparib (PARP-1 inhibitor), doxorubicin, and docetaxel. Therefore, STAT3 has been recognized as a potential therapeutic target for the treatment of TNBC (2, 6-8).
Efficient induction of apoptosis is required for therapeutic agents-elicited tumor regression. Caspase-mediated apoptosis is activated through extrinsic and intrinsic pathways (9). Several antidepressants have been associated with anti-cancer effects against TNBC (10, 11). For instance, imipramine, a tricyclic antidepressant, has been reported to induce tumor regression through the induction of apoptosis, cell cycle arrest, and disruption of DNA repair activity in TNBC both in vitro and in vivo (10). Fluoxetine, a selective serotonin reuptake inhibitor (SSRI), has been demonstrated to trigger intrinsic apoptosis, resulting in inhibition of TNBC cell growth (11). Previous studies demonstrated that fluoxetine, as a complementary agent, reduced tumor progression through attenuation of nuclear factor-kappaB (NF-B) signaling or the induction of apoptosis in non-small cell lung cancer (NSCLC), hepatocellular carcinoma (HCC), and gastric adenocarcinoma (12-15). However, the anti-TNBC effect of fluoxetine has not yet been understood. The main purpose of the present study was to investigate the anticancer efficacy and mechanism of fluoxetine in TNBC in vivo.
Materials and Methods
Cell culture. 4T1 cells were purchased from the American Type Culture Collection (ATCC). Cells were cultured in DMEM, containing 10% fetal bovine serum, 1% penicillin and streptomycin, and 2 mM L-glutamine (Hyclone Laboratories LLC, Logan, UT, USA) in a humidified incubator at 37°C containing 5% CO2.
Animal experiments. All procedures followed the China Medical University IACUC guidelines (Approval number: CMUIACUC-2022-299). Six-week-old male BALB/cByJNarl mice were purchased from the National Laboratory Animal Center (Taipei, Taiwan). 1×105 4T1 cells were resuspended in PBS (containing 30% Matrigel) with the 29G needle tip turned upwards, to enter the skin subcutaneously at 10 mm from the upper inoculation nipple (16); treatment was started when the tumor size reached approximately 50 mm3 as measured by a digital caliper. Tumor volume was measured every three days and calculated using the following formula: Volume=Height×Weight2×0.523. Mice were randomly divided into three group: Control (0.1% DMSO in 100 μl distilled water/day), fluoxetine (5 mg/kg fluoxetine in 100 μl distilled water/day), and fluoxetine (10 mg/kg fluoxetine in 100 μl distilled water/day). All treatments were delivered through gavage. Mice were humanely sacrificed post treatment on day 22, and the tumor and organs were collected for further experiments.
In vivo computed tomography (CT) scanning. After twenty-one days of treatment, the tumor size of mice was assessed by CT scan (Mediso Ltd., Budapest, Hungary). 4T1 bearing mice were sedated by using 1-3% isoflurane. The scanning parameters were listed as follows: tube energy=55 kVp×145 μA; direction=360°; Voxel size=145×145×145 μm (17).
Immunohistochemistry (IHC) staining. After 22 days’ treatment, mice were sacrificed. Tumors were isolated and fixed with 10% neutral formalin solution (containing 4% paraformaldehyde) at 4°C. Paraffin-embedded tumor specimens were sliced into 5-μm thick specimens by Bio-Check Laboratories Ltd. (New Taipei City, Taiwan, ROC). IHC staining was used to evaluate protein expression according to the instructions of the IHC kit manufacturer (DAB500, Millipore, Burlington, VT, USA). Primary antibodies STAT3 (Tyr705) (#9145, Cell Signaling Technology, Danvers, MA, USA, Twist (#46702, Cell Signaling Technology), SNAIL (#3879, Cell Signaling Technology), SLUG (#9585, Cell Signaling Technology), E-cadherin (#3195, Cell Signaling Technology), N-cadherin (#13116, Cell Signaling Technology), ZEB1 (#70512, Cell Signaling Technology), ZEB2 (#97885, Cell Signaling Technology), cleaved caspase-3 (#9661, Cell Signaling Technology), cleaved caspase-8 (#9496, Cell Signaling Technology), cleaved caspase-9 (#9509, Cell Signaling Technology) and Ki-67(E-AB-63523, Elabscience) were used in the present experiment (18).
Hematoxylin & Eosin staining. Mice were sacrificed on day 22. The heart, lung, kidney, liver, spleen, and small intestine were fixed with 10% neutral formalin solution at 4°C. Paraffin-embedded tumor specimens were sliced into 5-μm thick specimens and H&E staining was performed by Bio-Check Laboratories (New Taipei City, Taiwan, ROC) (19).
Statistical analysis. Statistical analysis was performed by Microsoft Excel 2017 version (Redmond, WA, USA) and One-way ANOVA and student t-test was used to evaluate significant differences. A p-value <0.05 defined a statistically significant difference. Data are displayed as mean±standard.
Results
Fluoxetine inhibited tumor growth in a 4T1 orthotopic mice model. In order to investigate the anti-tumor effect of fluoxetine in vivo, a breast cancer model was established by orthotopic injection (4T1 cells) in the upper nipple (as indicated in methods). The design of the animal experiments is presented in Figure 1A. The tumor volume of mice treated with 10 mg/kg fluoxetine was significantly smaller than that of the CTRL group (p-value <0.01) after twelve days of treatment (Figure 1B). Tumor growth inhibition by fluoxetine was also dose-depended; 10 mg/kg fluoxetine showed greater tumor growth inhibition than 5 mg/kg fluoxetine. After humanely sacrificing the mice, tumors were isolated, imaged, and weighted (Figure 1C-E). As indicated in Figure 1C-E, the 10 mg/kg fluoxetine group showed superior tumor size and weight inhibition than the 5 mg/kg fluoxetine group (p-value <0.01). Furthermore, in order to provide a more accurate assessment of the 3D tumor growth pattern of each treatment group in vivo, we performed CT scanning on days 0 and 21. A lateral, frontal, and transverse view from one representative mouse of each group is displayed in Figure 1F. Notable tumor growth inhibition was observed in the 5 mg/kg and 10 mg/kg fluoxetine groups compared to the 0.1% CTRL group. Taken together, our results indicated that fluoxetine inhibited 4T1 tumor growth.
Fluoxetine inhibited the growth of 4T1 tumor. (A) Flow chart of orthotopic animal model is displayed. (B) Tumor growth, (C-D) isolated tumor photograph, and (E) tumor weight on day 22 are shown. (F) CT scanning results from one representative mouse of each group on days 0 and 21 are presented. (*p<0.05, **p<0.01, ***p<0.005 vs. CTRL; $p<0.05, $$p<0.01 vs. 5 mg/kg fluoxetine).
Fluoxetine induced expression of pro-apoptosis related proteins and inhibited expression of proliferation-related proteins in 4T1 orthotopic mice model. After 22 days of treatment, mice were sacrificed, and tumors were isolated and sections from them were analyzed by using IHC staining. The expression of pro-apoptotic proteins such as cleaved caspase-3, extrinsic signaling marker cleaved caspase-8, and intrinsic signaling marker cleaved caspase-9 were analyzed after treatment (Figure 2A and 2B). As indicated in Figure 2A and B, fluoxetine induced apoptosis of 4T1 tumor cells via activating both the extrinsic and intrinsic apoptosis signaling pathways. Additionally, expression of the tumor proliferation marker Ki-67 was examined in 4T1 tumor sections (Figure 2C and 2D). The expression levels of Ki-67 in the 4T1 tumor sections were significantly inhibited by fluoxetine treatment (p-value<0.01). These results indicated that fluoxetine may not only trigger apoptosis but may also reduce the proliferation capacity of 4T1 tumor cells.
Fluoxetine induced apoptosis and inhibited proliferation. (A, C) Protein expression of cleaved-caspase-3, -8, -9, and ki-67 in tumor sections from each group are displayed. (B, D) The quantified expression levels of cleaved-caspase-3, -8, -9, and ki-67 from each group is presented in a bar chart. (***p<0.005 vs. CTRL; $$$p<0.005 vs. 5 mg/kg fluoxetine).
Fluoxetine diminished STAT3 protein expression and its mediated epithelial mesenchymal transition (EMT)-related protein expression in a 4T1 orthotopic mice model. Next, we investigated whether STAT3 and its related EMT factors (18) were affected by fluoxetine treatment. As illustrated in Figure 3A and 3B, the phosphorylation of STAT3 (Tyr705) was suppressed by fluoxetine. Furthermore, E-cadherin, which suppresses EMT, was increased by fluoxetine (Figure 3C and 3D). In contrast, factors that may promote EMT, such as N-cadherin, Slug, Snail, ZEB1, and ZEB2 were decreased by fluoxetine (Figure 3C-F). Taken together, we suggest that fluoxetine-induced inhibition of EMT may associate with the reduction of the phosphorylation of STAT3 at the Tyr705 site (inactivation of STAT3).
Fluoxetine induced STAT3 and epithelial-mesenchymal transition-related proteins inhibition. (A, C, F) Protein expression pattern of STAT3 (Tyr705), E-cadherin, N-cadherin, Slug, Snail, ZEB1, and ZEB2 in tumor sections from each group are displayed. (B, D, E, G) The quantified expression levels of these proteins from each group are presented in a bar chart. (***p<0.005 vs. CTRL; $$p<0.01 vs. 5 mg/kg fluoxetine).
Mice body weight and normal organ morphology were not affected by fluoxetine treatment in the 4T1 orthotopic mice model. To investigate whether fluoxetine induces toxicity in mice, we weighted the mice every three days (Figure 4A). There was no statistically significant difference between the body weight of mice in the fluoxetine-treated mice and that of CTRL group (p-value> 0.05), which indicated no obvious acute toxicity induced by fluoxetine treatment. Furthermore, we examined the normal organ (heart, lung, liver, spleen, kidney, and small intestine) morphology using H&E staining, as shown in Figure 4B. Therefore, fluoxetine treatment is safe in the dosage used to inhibit tumor growth. No obvious general toxicity was observed in 4T1 orthotopic mice.
No general toxicity was found after fluoxetine treatment in a 4T1 orthotopic mice model. (A) Body weight of mice on day 0 to 21. (B) H&E staining of the heart, lung, liver, spleen, kidney, and small intestine of each treatment group is shown.
Discussion
Extrinsic and intrinsic pathways mediate apoptosis through the activation of caspase-8 and caspase-9, respectively. Activated caspase-8 and -9 trigger expression of cleaved caspase-3, -6, and -7, which lead to cell death. Cleaved caspase-3 is a central player in apoptosis, specifically for the induction of apoptotic DNA fragmentation and the cleavage of poly [ADP-ribose] polymerase 1 (9, 21, 22). Bowie et al. found that the intrinsic pathway participated in fluoxetine-induced apoptosis in TNBC in vitro (11). Our data showed that the increased expression of cleaved caspase-3, -8, and -9 was significantly augmented by treatment with fluoxetine. According to these data, it is suggested that fluoxetine induces apoptosis through both the extrinsic and intrinsic pathways in TNBC in vivo (Figure 2A and B).
The EMT is the conversion of non-invasive epithelial cells to an invasive mesenchymal phenotype. EMT-related proteins such as N-cadherin, Slug, Snail, Twist, ZEB1, and ZEB2 facilitate tumor metastasis through up-regulation of EMT (23-25). E-cadherin, a crucial component of cell-cell adhesion junctions, restrains tumor invasion and metastasis by preventing tumor cell dissociation. Loss of E-cadherin has been recognized as an important hallmark of EMT (26, 27). The decreased expression of E-cadherin has been associated with unfavorable chemotherapy response and prognosis in TNBC (28). Our results indicated that fluoxetine not only effectively reduced tumor growth but also induced E-cadherin expression in TNBC in vivo (Figure 1B and Figure 3C).
Slug, Snail, Twist, ZEB1, and ZEB2 as E-cadherin repressors, inhibit E-cadherin expression to promote metastasis in various cancers (29-31). Constitutive STAT3 signaling is associated with a high expression of EMT-related proteins that correlate with metastasis and poor survival of breast cancers (32-36). The reduced STAT3 signaling down-regulates EMT and associates with a favorable TNBC survival (32, 37). Furthermore, inhibition of EMT-related proteins also contributes to a reduction in metastasis in TNBC (38-40). The results showed that the protein levels of STAT3(Try705), N-cadherin, slug, snail, Twist, ZEB1, and ZEB2 were significantly decreased by treatment with fluoxetine in TNBC in vivo (Figure 3C-G).
In conclusion, this study indicated that fluoxetine induces apoptosis through extrinsic/intrinsic pathways while suppresses tumor growth, STAT3 signaling, and expression of EMT-related proteins. We suggested that induction of apoptosis and STAT3 inactivation are associated with fluoxetine-induced inhibition of cell growth and EMT in TNBC.
Acknowledgements
The Authors thank the Medical Research Core Facilities Center, Office of Research & Development at China Medical University (Taichung, Taiwan, R.O.C) for the technical support.
Footnotes
Authors’ Contributions
PAL, PYC, ZLT performed the experiments, derived the models and analyzed the data. ZLT and FTH prepared the initial version of the paper. FTH, YCL, and HJW conceived of the presented idea, supervised the findings of this work, performed the literature review, and prepared the final versions of the paper.
Conflicts of Interest
The Authors declare no competing financial interests regarding this study.
Funding
This study was supported by Tainan Municipal Hospital, Tainan, Taiwan (RD-108-04) and Show Chwan Memorial Hospital, Changhua, Taiwan (ID: SRD-110033), respectively.
- Received June 6, 2022.
- Revision received June 24, 2022.
- Accepted June 25, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.