Penfluridol: an antipsychotic agent suppresses metastatic tumor growth in triple negative breast cancer by inhibiting integrin signaling axis

Abstract

Metastasis of breast cancer, especially to the brain, is the major cause of mortality. The inability of anti-cancer agents to cross the blood-brain-barrier represents a critical challenge for successful treatment. In the current study, we investigated the anti-metastatic potential of penfluridol (PF), an antipsychotic drug frequently prescribed for schizophrenia with anti-cancer activity. We show that PF induced apoptosis and reduced the survival of several metastatic triple negative breast cancer (TNBC) cell lines. Additionally, PF treatment significantly reduced the expression of integrinα6, integrin β4, Fak, Paxillin, Rac1/2/3, and ROCK1 in vitro. We further evaluated the efficacy of PF in three different in vivo tumor models. We demonstrate that PF administration to an orthotopic model of breast cancer suppressed tumor growth by 49%. On the other hand, PF treatment inhibited the growth of metastatic brain tumors introduced by intracardiac or intracranial injection of breast cancer cells by 90% and 72%, respectively. PF-treated tumors from all three models exhibited reduced integrinβ4 and increased apoptosis. Moreover, chronic administration of PF failed to elicit significant toxic or behavioral side effects in mice. Taken together, our result indicate that PF effectively reduces the growth of primary TNBC tumors and especially metastatic growth in the brain by inhibiting integrin signaling, and prompt further preclinical investigation into repurposing PF for the treatment of metastatic TNBC.

Introduction

Breast cancer is the most diagnosed cancer and the second leading cause of mortality (1–4). Triple negative breast cancer (TNBC) is characterized by the absence of receptors for estrogen (ER) and progesterone (PR) as well as lack of HER-2 gene (5). TNBC is considered to have basal characteristics, which are associated with more aggressive cancer phenotype (6). TNBC is more prevalent in younger females and accounts for about 10–20% of breast cancer incidence with poor prognosis (7). Metastatic TNBC is difficult to treat thus leading to extremely poor survival. About 25–46% patients with TNBC are at higher risk of brain metastasis (8). Hence it is crucial to search new treatment options for patients with metastatic TNBC.

Integrins are heterodimeric adhesion family receptors, which facilitate communication of the cells with extracellular matrix and promote cell survival (9). Accumulating evidence suggests a major role of integrins in cancer. Integrinα6β4 have been implicated in breast cancer progression and metastasis, making it an attractive target for breast cancer therapy (10–13).

Published studies suggest an overall reduced risk of cancer in schizophrenic patients using neuroleptic agents (14,15). Interestingly, few agents like chlorpromazine and thioridazine have shown anti-cancer activity (16–19). Penfluridol (PF) is an oral anti-psychotic drug available since 1970 for the treatment of schizophrenia (20). Limited studies indicate that PF exhibits anti-cancer activity however, mechanism of the anti-cancer effect of PF is not known (21).

In the present study, we investigated the anti-metastatic effect of PF in TNBC. We observed significant growth suppression of breast cancer cells by PF treatment through inhibition of integrin signaling. Reduced cell migration and invasion of metastatic TNBC cells was also observed by PF treatment. Oral administration of PF suppressed the tumor growth and metastasis of breast cancer cells to brain by inhibiting integrin α6β4 signaling. To the best of our knowledge, this the first report on the anti-metastatic effects of PF targeting integrin α6β4 signaling axis in TNBC.

Materials and methods

Ethics Statement

Investigation has been conducted in accordance with the ethical standards and according to approved protocol by Institutional Animal Care and Use Committee (IACUC).

Cell culture

Human triple negative breast carcinoma cell lines MDA-MB-231 and 4T1 were purchased from ATCC, Manassas, VA and Perkin Elmer, Waltham, MA respectively. HCC1806 cells were kindly provided by Dr. Sophia Ran, Southern Illinois University-School of Medicine. The cell lines were maintained in DMEM supplemented with 10% FBS and 5% PSN. All the cells used in this study were within twenty passages after receipt or resuscitation.

Cytotoxicity Studies

Cells were plated at a density of about 3000 cells/well in 96 well plates and incubated overnight. The cells were then treated with different concentrations of PF (Sigma-Aldrich, St. Louis, MO). After desired duration of treatment (24, 48 and 72h), cells were fixed using ice cold 10% trichloroacetic acid followed by washing and staining with Sulforhodamine B (SRB) dye. Plates were washed with 1% solution of acetic acid and the optical density was measured in 10 mM Tris-base solution, using plate reader (BioTek Instruments, VT) as described by us previously (22,23).

Wound Healing Assay

Wound healing assay was performed as described by us earlier (24,25). Briefly, 4T1 cells were incubated to form a monolayer in 6 well dishes. Wound was created by scratching the monolayer with a 1ml sterile tip. The cells were washed with sterile phosphate saline buffer (PBS) thrice to remove the floating cells and 4µM PF was added to the cell culture media. The cells were fixed using 10% tricholoroacetic acid at desired time points. The wound was imaged using bright field microscope (Olympus Inc, PA) after staining with SRB dye as described above. The wound widths were quantitated using Image J software.

Transwell cell invasion assay

Cell invasion was performed using Transwell Boyden’s chamber with 8.0µm pore size membrane (BD Biosciences, San Jose, CA). The assay was performed according to manufacturer’s instructions and as described by us previously (24,25). Briefly, serum starved cells were seeded in the upper well of Boyden’s chamber. The lower chamber was filled with cell culture medium containing 10% FBS and VEGF as chemo-attractants. After 2h of incubation, 4µM PF was added to upper chamber of the well. After 24h of PF treatment, cells from the upper side of membrane were removed whereas cells that migrated to the lower side of the membrane were fixed with 10% TCA and stained with 0.4% (w/v) SRB solution. The SRB dye was solubilized in 10 mM Tris buffer and the absorbance was taken using a microplate reader (BioTek Instruments, VT).

Integrin β4 and integrin α6 silencing

MDA-MB-231 cells were transfected with integrin β4 shRNA (Addgene, Cambridge, MA) or integrin α6 siRNA (Cell Signaling Technologies, Danvers, MA) using Xfect (Takara Clonetech, Mountain View, CA) or siPORT (Ambion Inc, Austin, TX) transfection reagent as per manufacturer’s protocol. Briefly, 2.5µg shRNA plasmid was suspended in dilution buffer and Xfect polymer was added accordingly. The complexes were added to the cells after 10 mins of incubation. In another experiment, cells were transfected with 100nM integrin α6 siRNA or scrambled siRNA using siPORT reagent and after 24h post transfection, cells were treated for additional 24h with 6µM PF. The cells were collected after treatment and processed for western blot analysis.

Laminin and TGFβ treatment

Laminin (Life technologies, Carlsbad, CA) was diluted in PBS to attain a final concentration of 50µg/ml and added to the wells (1ml/well in a 6 well plate) or coverslips (500µl/well in a 24 well plate) to cover the surface and incubated overnight at 4°C. Next day after removing laminin solution, wells were washed with PBS followed by a wash with pre-warmed cell culture medium and the cells were plated. The cells were processed for western blot analysis or microscopy after 24h treatment with PF.

For TGFβ (Peprotech, Rocky Hill, NJ) treatment, cells were serum starved overnight. Following day, media was replaced with normal cell culture media containing 10% serum and 20ng/ml TGFβ 1h prior to PF treatment. After 24h treatment with 6µM PF, cells were collected and processed for western blot analysis.

Western Blot Analysis

The whole cell lysates were prepared using 4% (w/v) CHAPS in urea-tris buffer. Proteins from whole cell lysates were subjected to SDS-PAGE and the resolved proteins were transferred to PVDF membrane. The membranes were probed for primary antibodies against integrin α6, integrin β4, integrin α4, integrin αv, integrin β1, integrin β3, p-FAK (Y397), FAK, p-Paxillin (Y118), Paxillin, Rac1/2/3, p-Rac1 (S71), ROCK1 and cleaved caspase 3. All primary antibodies were purchased from Cell Signaling Technologies (Danvers, MA) except integrinβ4 and FAK (Santa Cruz Biotechnology, Inc., Dallas, TX). The membranes were developed as described by us previously (22,23,26,27).

Immuno-precipitation

Immuno-precipitation was performed as described by us previously (10,28–30). Briefly, 0.75×10⁶ MDA-MB-231 cells were plated in 100mm tissue culture dishes and treated with 6µM PF. After 24h PF treatment, whole cell lysates were prepared using RIPA buffer and immuno-precipitated with integrin β4 antibody (Santa Cruz Biotechnology Inc, Dallas, TX). Immune complexes were resolved on SDS-PAGE and immuno-blotted for integrinα6.

PCR

Total RNA was extracted from control and treated cells using TRIzol reagent (Life Technologies, Inc., Carlsbad, CA) according to manufacturer instructions and the cDNA synthesis was carried out as described by us earlier (29). About 2 µg of template cDNA was used for cycling in a thermal cycler (Thermo Fisher, Pittsburgh, PA). The PCR products were separated on a 1.5% agarose gel, stained with 0.5 mg/mL ethidium bromide, and visualized under UV light using BioRad Versa Doc imager.

Immunofluorescence analysis

Immunofluorescence analysis was performed as described by us earlier (26). MDA-MB-231 cells were plated in a 24-well plate on a coverslip (uncoated or laminin coated) at a density of 0.1×10⁶ cells/well and allowed to attach overnight followed by treatment with 4µM PF for additional 24h. The cells were fixed with formalin and permeabilized using Triton-X100 solution. After blocking with goat serum, cells were incubated overnight with primary antibody for integrin β4 (1:150). Next day, cells were washed and incubated with AlexaFluor 488 secondary antibody (Invitrogen, Carlsbad, CA), followed by DAPI staining for nuclei. After washing, the coverslips were mounted on slides and images were taken using fluorescence microscope (Olympus, Center Valley, PA).

Breast tumor orthotopic model

Female Balb/c mice (4–6 weeks old) were obtained from Harlan Laboratories (Livermore, CA). The experiments were conducted in strict compliance with the regulations of Institutional Animal Care and Use Committee (IACUC), Texas Tech University Health Sciences Center. Exponentially growing 4T-1 cells were harvested, washed twice with PBS and re-suspended in PBS at a density of 0.7×10⁶ cells per ml. A suspension of 0.1mL containing 0.07×10⁶ cells was injected in the inguinal mammary fat pads of each recipient mouse. Tumor volumes were measured three times a week as described by us previously (10,31) and calculated using the formula (length × (width)²/2) (32,33). Two days after tumor cells injection, mice were randomly segregated into two groups with seven mice in each group. Test group of mice received 10mg/kg PF by oral gavage every day till day 27, whereas control mice received vehicle alone. PF stock was made in DMSO which was further diluted in water/PEG300/ethanol/2% acetic acid in 8:3:3:1 v/v (34). Experiment was terminated at day 27 by euthanizing mice with CO₂ overdose. The tumors were removed aseptically from each mouse, weighed and snap frozen in liquid-nitrogen for western blot analysis. A part of tumor was fixed in formalin for immunohistochemical analysis.

Intracardiac Brain Metastasis Model

Female Balb/c mice (4–6 weeks old) were obtained from Charles River (Wilmington, MA) and maintained as per the IACUC guidelines. For the metastatic breast cancer model, we used a method first described by Conley et. al. in 1979 (35) and since then it is widely used to study metastasis to bone and brain (36,37). Briefly, 4T1 cells expressing luciferase (PerkinElmer, Waltham, Massachusetts), were harvested, washed and re-suspended in sterile phosphate buffered saline (PBS) at a density of 2.5×10⁴ cells/50µl. Mice were injected with 50µl of the cell suspension into the heart’s left ventricle using stereotaxic apparatus. The tumor growth in mice was monitored by using non-invasive imaging technique (IVIS, Perkin Elmer, Waltham, Massachusetts). After three weeks, mice were sacrificed and brain was collected in a sterile environment and minced aseptically to collect 4T1-luc cells. The 4T1-luc cells were selected using 60µM 6-thioguanine and cultured further and named them brain seeking 4T1-luc cells. These cells (2.5×10⁴ cells/50µl) were re-injected into the left ventricle of the heart of a new set of mice under isofluorane anesthesia. Animals were monitored till they became conscious. The mice injected with 4T1-luc cells were randomized and divided into two groups with 5 mice in each group. The 4T1 breast tumor cells usually reach the brain within 5 minutes following intracardiac injection as observed by imaging. The treatment group received 10mg/kg PF by oral gavage starting the same day after cell injection whereas control group received vehicle only. The mice were humanely sacrificed at day 12 as the control mice started showing signs of sickness due to metastatic tumor burden in brain. The brains were collected, imaged for luminescence and fixed in formalin overnight and processed for immunohistochemical analysis.

Intracranial tumor model

Female Balb/c mice (4–6 weeks old) from Harlan Laboratories (Livermore, CA) were used for intracranial injection and the experiments were conducted in strict compliance with the regulations of Institutional Animal Care and Use Committee (IACUC), Texas Tech University Health Sciences Center. Exponentially growing 4T1-luc cells were harvested, washed twice with sterile PBS and re-suspended in PBS at a density of 5×10⁶ cells per ml. A suspension of 5µl containing 0.025×10⁶ cells were injected by intracranial route at a flowrate of 1µl/min in each recipient mouse using stereotaxic apparatus. Following this, mice were randomly divided into two groups with 6 mice in each group. 10mg/kg PF by oral gavage was administered to mice a day after tumor cells injection and then every day after that. The tumor growth was monitored by non-invasive in vivo live animal imaging as described before (24). Experiment was terminated by humanely euthanizing the mice with CO₂ overdose and mice brain were carefully dissected out, weighed, imaged for luminescence and processed for immunohistochemical staining.

Immunostaining of brain sections

The immunohistochemistry (IHC) was performed as previously described by us (10,38). Briefly, fixed brains were dehydrated, embedded in paraffin and sectioned into 5µm thick sections using microtome (Leica Microsystems Inc., Buffalo Grove, IL). The sections were deparaffinized and rehydrated using xylene, ethanol and double-distilled water washes. Antigens were unmasked by boiling the sections in 10 mM sodium citrate buffer (pH 6.0) and the sections were washed and incubated in 3% hydrogen peroxide solution. The sections were blocked with 5% goat serum and incubated with primary antibodies for integrinβ4 (1:100) and cleaved caspase 3 (1:100) overnight at 4°C. Next day the slides were stained using Ultravision ONE HRP polymer kit (Thermofisher scientific, Fremont, CA) as per the manufacturer’s instructions. The sections were counterstained with Mayer’s hematoxylin and dehydrated. The slides were mounted using Permount (Fisher scientific, Fair Lawn, NJ) and imaged using Olympus microscope (Olympus America Inc, Center Valley, PA).

Dose tolerance and mice behavioral analysis

Female CD1 mice (4–6 weeks old) were obtained from Charles River (Wilmington, MA, USA). The use of CD1 mice and their treatment was approved by the Institutional Animal Care and Use Committee (IACUC), Texas Tech University Health Sciences Center. Mice were randomly divided into two groups with 5 mice per group. Mice were administered 10mg/kg PF by oral gavage every day for 55 days. Control mice received the vehicle only. Mice weights were monitored once a week and mice were observed for general signs of toxicity. After 55 days of treatment, behavioral activity of mice was assessed using Versamax (AccuScan Instruments Inc., Columbus, OH, USA). Versamax is a ventilated chamber equipped with infrared sensors along the side wall to monitor mice activity. Each mouse was acclimatized in chamber for 15min before taking the reading. The readings were taken for control and PF treatment groups. At the end of experiment mice were humanely euthanized and plasma was carefully collected for analysis of liver transaminases. The enzymatic activities of AST and ALT were determined using a commercially available kit (Pointe Scientific, Inc, Canton, MI, USA), according to manufacturer’s instructions and as described previously (10). Mice organs were also collected and weighed for comparison between control and PF treatment group.

Statistical Analysis

Prism 6.0 software was used for all the statistical analysis (GraphPad software Inc., San Diego, CA). Results are represented as means ± standard deviation (SD) or standard error of means (SEM). Statistical significance was analyzed using Student’s t-test or Mann Whitney test and outcomes were considered statistically significant at p<0.05.

Results

PF suppresses proliferation of TNBC breast cancer cells

To evaluate the growth suppressive effects of PF, we first performed the cytotoxicity assay in MDA-MB-231, HCC1806 and 4T1 TNBC cells. The cells were treated with varying concentrations of PF for 24, 48 and 72h. Our results showed that treatment with increasing concentrations of PF significantly suppressed the growth of all the three breast cancer cell lines in a concentration and time-dependent manner. The IC₅₀ of PF after 24h treatment ranged 6–8µM in all the three cell lines (Fig.1A). The IC₅₀ values were further reduced to about 4–5µM and 2–4µM after 48 and 72h treatment respectively in all the cell lines (Fig. 1A). These results suggest potential cytotoxic effects of PF in TNBC cells.

Figure 1. Penfluridol suppresses cell survival, cells migration and invasion.

(A) MDA-MB-231, HCC1806 and 4T1 cells were treated with different concentrations of penfluridol at 24, 48 and 72 hours. Cell survival was measured by sulforhodamine B assay to estimate IC₅₀ values. The experiments were repeated three times with 8 replicates in each experiment. (B) 4T1 cells were grown in 6 well plates to form monolayer. Wounds were created using 1ml sterile tip. Cell migration capacity was measured by the time taken to heal the wound in control and penfluridol (4µM) treated cells. Statistically different at p<0.05 when compared to control. (C) 4T1 cells were starved in serum free medium overnight before plating cells in transwell. Invading potential of the cells was estimated by staining the cells invaded to lower side of the membrane using sulphorhodamine B dye. Invading potential of the cells in the treatment was compared with the control. Values were plotted as means ± SD. Experiment was repeated three times and considered statistically significant with control at p<0.05.

PF inhibits migration and invasion of breast cancer cells

To determine the effects of PF on metastatic potential of cells, cell migration using wound healing assay was evaluated. Our results indicated that the migration of PF treated 4T1 cells was significantly delayed as compared to control cells. PF treatment inhibited the migration of 4T1 cells by 61% and 76% at 18h and 36h respectively (Fig 1B). Furthermore, the effect of PF on cell invasion was confirmed by transwell invasion assay. Our results showed that invasion of PF treated cells was only 60% as compared to 100% in control cells (Fig. 1C). These observations indicated that PF treatment inhibits cell migration and invasion of breast cancer cells suggesting anti-metastatic potential.

PF inhibits integrin signaling

To elucidate the molecular mechanism of the growth suppressive effects of PF, we performed western blot analysis of whole cell lysates of MDA-MB-231, HCC1806 and 4T1 cells treated with 0, 2.5, 5 and 7.5µM PF for 24h. Our results showed that expression of integrinα6 and integrinβ4 were significantly reduced in a concentration-dependent manner by PF treatment in MDA-MB-231 and HCC1806, human breast cancer cell lines (Fig. 2A–B). Surprisingly, in 4T1, a murine breast cancer cells, we did not observe much change in integrinα6 but integrinβ4 expression was significantly reduced in a concentration dependent manner by PF treatment (Fig. 2C). PF treatment also reduced the expression of integrin α4, integrin αv, integrin β1 and integrin β3 in concentration-dependent manner in MDA-MB-231, HCC-1806 and 4T1 cells (Supplemental Fig. 1). In addition, we observed a notable inhibition of the downstream effector molecules of integrin signaling such as FAK, p-Paxillin (Y118) and Paxillin by PF treatment. Integrin signaling modulates Rac and ROCK1 proteins to activate cell migration and invasion (Fig. 2A–B). Interestingly, PF treatment significantly inhibited p-Rac1 (S71), Rac1/2/3 and ROCK1 expression (Fig. 2A–B). These results indicate that PF suppresses cell survival and motility by inhibiting integrin signaling.

Figure 2. Suppression of integrin α6β4 and associated signaling by penfluridol treatment.

(A) MDA-MB-231, (B) HCC1806, (C) 4T1 cells were treated with different concentrations of penfluridol for 24 hours. Representative blots showing concentration-dependent effect of penfluridol on integrin α6, integrin β4, p-FAK (Y397), FAK, p-Paxillin (Y118), Paxillin, Rac1/2/3, p-Rac (S71), ROCK1 and cleaved caspase 3. Actin was used as loading control. Figure shown is the representative blot of at least three independent experiments. Blots were quantitated and normalized with actin.

Silencing integrins 6 and 4 enhances the effects of PF

To confirm the role of integrins in PF attributed effects, we genetically silenced integrinvα6 and β4 in MDA-MB-231 cells using siRNA and shRNA respectively. Using siRNA, silencing 32% integrin 6 caused suppression of p-FAK (Y397), p-Rac-1, ROCK1 (Fig. 3A) and augmented the effects of PF treatment on integrin signaling as well as cleavage of caspase3 (Fig. 3A). Similarly, using shRNA, we were able to inhibit about 43% integrinβ4 expression (Fig. 3B). Our results further showed that silencing of integrin4 significantly suppressed the levels of p-FAK (Y397), Rac1/2/3 as well as ROCK1 (Fig. 3B). The effect of PF in suppressing the level of these molecules was further enhanced in MDA-MB-231 cells with silenced integrin β4. A 17 fold increased cleavage of caspase 3 was observed in PF treated cells with silenced integrin β4 expression as compared to only 6 fold in the cells with basal expression of integrin 4 (Fig. 3B). These observations prove that PF treatment inhibits integrin 64 signaling to suppress breast cancer cell growth as well as migration and invasion potential.

Figure 3. Penfluridol induces apoptosis through suppression of integrin α6β4 signaling cascade.

MDA-MB-231 cells were treated with penfluridol (6µM) for 24 hours after (A) transfecting the cells with integrin α6 siRNA (B) transfecting the cells with integrin β4 shRNA (C) Cells were plated in laminin coated plate followed by treatment with 6µM penfluridol for 24h (D) Cells were pretreated with 20ng/ml TGFβ for one hour followed by treatment with penfluridol as described above. Levels of p-FAK, p-Paxillin, p-Rac1, Rac1/2/3, ROCK1, cleaved caspase 3 were evaluated by western blotting. Figure shown is the representative blot of at least three independent experiments. Blots were quantitated and normalized with actin. (E) Integrin β4 was immunoprecipitated from control and 6µM penfluridol treated MDA-MB-231 cells and probed for integrin α6. (F) mRNA level of integrinα6 was evaluated after treating MDA-MB-231 cells with 6µM penfluridol for 24h by RT-PCR (G) Nuclear localization of integrin β4 by fluorescence microscopy in MDA-MB-231 cells treated with 4µM penfluridol for 24h. Green fluorescence represents integrin β4 whereas blue represents DAPI.

Activation of integrins 6 and 4 signaling suppresses efficacy of PF

We further used laminin, the ligand of integrin 64 and TGF to activate integrin signaling and then evaluated the effects of PF in breast cancer cells. Our results showed that the cells cultured on laminin coated surface increased p-FAK(Y397) and p-Paxillin (Y118) by 1.70 and 1.67 fold respectively however, levels of these phosphorylated protein was significantly reduced by PF treatment (Fig. 3C). In addition, increased integrin 64 signaling by laminin reduced cleavage of caspase 3 induced by PF treatment in MDA-MB-231 cells (Fig. 3C). Similarly, TGF treatment also activated integrin signaling, as exhibited by increased phosphorylation of FAK (Y397) and p-Paxillin (Y118) (Fig. 3D). In line with these observations, activation of integrin64 signaling by TGF blocked the cleavage of caspase 3 mediated by PF in MDA-MB-231 cells (Fig. 3D). Taken together these results confirmed integrin 64 as targets of PF in suppressing the growth of breast cancer cells.

PF treatment disrupts integrin 64 heterodimer

To delineate the mechanism of inhibition of integrins, effects of PF on heterodimerization of integrin 6 with integrin 4 was evaluated by immunoprecipitation studies. Immunoprecipitated integrin β4 from PF treated MDA-MB-231 cells showed reduced association of integrin 6 with integrin 4 (Fig. 3E). We further evaluated the effect of PF on integrin 6 at the transcriptional level by performing PCR analysis. Our results showed that PF treatment inhibited the mRNA levels of integrin α6 by 76% in MDA-MB-231 cells (Figure 3F). In addition, the immunofluorescence analysis also confirmed inhibition of integrin 4 by PF treatment (Fig. 3G). We also induced integrin β4 signaling by using laminin and then treated the cells with PF. The green staining for integrin β4 was reduced significantly in the cells treated with PF or in cells treated with laminin and PF combination. Taken together, these results demonstrate that integrin α6β4 are the targets of PF in breast cancer cells.

PF inhibits the growth of 4T1 orthotopic tumors

To test the efficacy of PF in vivo, we implanted aggressive 4T1 murine breast tumor cells, representing stage IV breast cancer, orthotopically in the mammary fat pads of Balb/c mice and the mice were treated with 10mg/kg PF by oral gavage every day. Our results showed significant suppression of tumor growth in PF treated mice as compared to control mice. At the end of experiment, average tumor volume of PF treated group was about 342mm³ as compared to 668mm³ of control group, indicating 49% reduction in tumor volume (Fig. 4A). Tumors were collected and weighed after humanely euthanizing the mice. Average weight of PF treated tumors was about 42% less as compared to control group (Fig. 4B & C). Tumor lysates were subjected to western blot analysis. In Fig. 4D, each band represents lysate from a separate tumor. In agreement with our in vitro observation with 4T1 cells, no significant change in the expression of integrinα6 was observed in PF treated tumors as compared to control tumors (Fig. 4D & E). However, a remarkable suppression of FAK, ROCK1, Rac1/2/3 and enhanced cleavage of caspase 3 was observed in PF treated tumor lysate (Fig. 4D & E). These observations were also confirmed by immunohistochemical staining of tumors from control and PF treated mice for integrin4 and cleaved caspase 3. Our results demonstrated reduced expression of integrin 4 as well as increased cleavage of caspase 3 in tumor samples from PF treated mice (Fig. 4F). These results indicated that breast tumor growth suppression by PF was associated with inhibition of integrin signaling and induction of apoptosis.

Figure 4. Penfluridol suppresses the growth of breast tumor and integrin signaling in vivo.

(A) About 0.07×10⁶ 4T1 breast cancer cells were injected orthotopically in the mammary fat pads of 4–6 weeks old Balb/c female mice. Treatment with 10 mg/kg penfluridol by oral gavage everyday started 2 days after tumor cells injection, till day 27. Values were plotted as means ± SEM (B) Tumors were weighed once isolated from control and penfluridol treated mice. Values were plotted as means ± SEM (C) Representative tumor pictures from control and penfluridol treated mice. Orthotopically implanted tumors were removed aseptically after terminating the experiments. Tumors were homogenized, lysed and analyzed for integrin α6, integrin β4, FAK, ROCK1, Rac1/2/3 and cleaved caspase 3. Actin was used as loading control. (D) Each lane of blot represents tumor from individual mouse. (E) Blots were quantitated, normalized with actin and represented as bars. Values were plotted as means ± SEM (F) Tumors were sectioned and immunostained for integrin β4 and cleaved caspase 3, as described in method section.

PF inhibits in vivo brain metastasis of 4T1 cells

The in vivo efficacy of PF was further validated in an in situ metastatic breast cancer model. The 4T1-luc cells were injected into the left ventricle of mouse heart so that the cells lodge in the brain. Luminescence was detected in the brain of mice within minutes of intra-cardiac injection of the tumor cells (data not shown). Mice in experimental group received 10 mg/kg PF by oral gavage every day. Control mice bearing metastatic breast tumor in brain starts succumbing after 12 days so the experiment was terminated at day 12. Our results showed an increase in brain luminescence starting day 6 after injection (Fig. 5A). At the end of the experiment, there was a massive increase in luminescence signal in control group as compared to PF treated mice (Fig. 5A). Based on luminescence, our results showed about 90% inhibition in tumor growth by PF treatment (Fig. 5A). At the end of the experiment, brain from control and PF treated mice were removed and imaged. Average brain luminescence from PF treated group was also suppressed by 90% as compared to control group (Fig. 5B). The brain from control and PF treated mice were analyzed by immunohistochemistry. Consistent with our previous observations, tumors from PF treated mice exhibited reduced expression of integrin β4 and increased staining for cleaved caspase 3 (Fig. 5C). These results clearly indicated that PF treatment suppressed metastatic growth of triple negative breast cancer cells by inhibiting integrins and inducing apoptosis.

Figure 5. Penfluridol inhibits the growth of metastatic breast tumors in the brain in intracardiac model.

About 2.5×10⁴ 4T1-luc brain seeking aggressive breast cancer cells were injected in left ventricle of 4–6 weeks old female balb/c mouse heart. Mice were treated with 10mg/kg penfluridol by oral gavage everyday till day 12. (A) Brain luminescence (photons/sec) of each mouse was normalized with initial average luminescence of all the control and treated group mice. Relative increase in brain luminescence of control and treated group was plotted against days till day 12. (B) After terminating the experiment, brain of each mouse was imaged and luminescence in respective group was plotted. Values were plotted as means ± SEM Representative brains from control and penfluridol treated mice. (C) Brains were processed, sectioned and immunostained for integrin β4 and cleaved caspase 3.

PF inhibits the growth of 4T1 tumors in intracranial tumor model

The anti-tumor effect of PF was further validated in an intracranial tumor model. 4T1-luc tumor cells were injected directly into the brain through intracranial injection using stereotaxic apparatus and the mice were treated with 10mg/kg PF by oral gavage every day. Our results showed a steady increase in brain luminescence indicating fast tumor growth of 4T1-luc cells in the brain of control mice whereas not much increase of luminescence was observed in the brain of PF treated mice. At the end of the experiment (day 21), luminescence in control mice was 3.5 fold higher than the luminescence in PF treated mice (Fig. 6A). We also analyzed luminescence in the isolated brains from both groups after euthanizing the mice. Our results showed a clear suppression of tumor growth by PF treatment as indicated by significantly reduced luminescence in PF treated brains as compared to control brains (Fig. 6B). In addition, we also observed reduction in average brain weight of PF treated group, suggesting reduced tumor load (Fig. 6C). The isolated brains were also examined by immunohistochemistry. Consistently, we observed suppression of integrin β4 expression as well as increase in cleaved caspase 3 in the tumors from the brain of PF treated mice as compared to control mice (Fig. 6D). These results established the growth inhibitory effects of PF in metastatic tumors.

Figure 6. Penfluridol suppresses the growth of intracranially implanted breast tumors in brain.

About 0.025 × 10⁶ 4T1-luc breast cancer cells were injected in the brain of 4–6 week old female balb/c mice using stereotaxic apparatus, connected to motorized injector. 10mg/kg penfluridol by oral gavage was given to each mouse everyday till day 21. (A) Luminescence in brain (photons/sec) of each mouse was normalized to initial average luminescence of all the control and treated group mice. Relative increase in the luminescence from brain was plotted in control and penfluridol treated group till day 21. Values were plotted as means ± SEM (B) At day 21, experiment was terminated and brain from control and penfluridol treated mice were removed, imaged and luminescence of control and PF treated group was plotted. Values were plotted as means ± SEM. Also shown is the representative image of control and penfluridol treated brain after terminating the experiment. (C) Individual brains were weighed and results were plotted as means ± SEM (D) Brains were sectioned, processed and immunostained for integrin β4 and cleaved caspase 3.

PF does not exhibit any toxicity in vivo in a chronic toxicity model

Our in vitro and in vivo studies conclusively showed breast tumor growth suppression by PF in three different in vivo models. However, it was not clear whether long term treatment with PF would cause any unwanted side effects or toxicity. Hence we studied long term effects by treating the mice with 10mg/kg PF by oral gavage every day for 55 days. General signs of toxicity, such as, body weights, organ weights and plasma transaminases (ALT, AST) were evaluated. Our results showed no significant difference in the overall weights of PF treated mice as compared to control mice (Fig. 7A). PF treatment modestly increased ALT activity, whereas AST activity was reduced. The overall high basal level of AST and ALT in our study compared to reported normal range could be due to lysis of RBC while collection of plasma (Fig. 7B & C). We also weighed the brains, livers, kidneys and spleens of control and PF treated mice after the experiment was terminated. No difference in the average weights of any of these critical organs was observed in PF treated group as compared to control group of mice (Fig. 7D–G). These results suggested that chronic treatment with 10mg/kg PF was not associated with any toxicity or side effects.

Figure 7. Penfluridol does not induce any side effects or affect behavioral activity in mice after chronic administration.

Female CD1 mice were given 10 mg/kg penfluridol by oral gavage everyday till day 55 while control group received vehicle only. Weight of mice was taken once a week. After terminating the experiment at day 55, plasma from control and penfluridol treated group mice was collected and analyzed for liver transaminase. The weight of brains, livers, kidneys and spleens of mice was also taken and plotted. (A) Mice weights (B) Aspartate transaminase (AST) in the plasma (C) Alanine transaminase (ALT) and (D–E) Weight of brains, livers, kidney and spleen. Behavioral activity of mice was assessed using Versamax (Accuscan Instruments) after 55 days of 10 mg/kg penfluridol administration by oral gavage every day. (H) Clockwise revolution (I) Counter clockwise revolution (J) Total distance (K) Horizontal activity (L) Vertical activity. Values were plotted as means ± SD.

No apparent effects of PF on behavioral activity of mice

PF is a CNS acting drug; hence besides general signs of toxicity, we also monitored behavioral activity of mice after long term PF administration. The behavioral activity of mice was monitored after 55 days of PF treatment using Versamax (AccuScan Instruments Inc., Columbus, OH, USA). The readings were taken after administration of 10mg/kg PF every day for 55 days. Our results showed no significant changes in terms of clockwise or counter-clockwise movement, total distance covered, vertical and horizontal activity in PF treated mice as compared to control mice (Fig. 7H–L). These results suggested that chronic administration of PF does not has any significant effect on the behavioral activity of mice, indicating that perhaps it is relatively safe for long term use.

Discussion

Our current study provides the first in vitro and in vivo evidence for significant activity of an antipsychotic drug PF against metastatic TNBC, a type of breast cancer, which is currently untreatable. Our results indicated that PF treatment reduced the proliferation of TNBC by inhibiting integrin α6β4 signaling axis. Surprisingly, inhibition of integrin α6 was not observed by PF treatment in 4T1, a murine breast cancer cell line. Inhibitory activity of PF on integrin axis was also confirmed by silencing integrinα6 and integrinβ4, which further enhanced the effects of PF in TNBC cells. Furthermore, activation of integrin α6β4 by laminin or TGFβ, suppressed the inhibitory effect of PF, which was demonstrated by reduction in cleavage of caspase3. To determine the efficacy of PF in vivo, we used three different tumor models. Interestingly, PF suppressed the growth of breast tumors in brain as well as in the breast as evaluated in intracardiac, intracranial and orthotopic models respectively. Suppression of metastatic breast tumors in brain and primary breast tumors by PF was due to inhibition of integrin signaling as demonstrated by western blot and immunohistochemical staining of tumors and consistent with our in vitro observations. Moreover, anti-cancer dose of PF used in our experiment was safe to use without any side effect or behavioral changes as shown in our chronic toxicity study where 10mg/kg PF was administered by oral gavage in mice everyday till 55 days. Hence our study reveals a highly impressive activity of PF against TNBC.

Published studies indicate reduced cancer incidence with consumption of neuroleptic agents in patients with schizophrenia (14,15,39). PF is an approved antipsychotic drug for schizophrenia and upon administration shows good bioavailability in brain. Few studies have suggested that antipsychotic drugs such as haloperidol, perphenazine, fluphenazine have anti-cancer effects. (17,40). Consistent with those studies, our results demonstrated the anti-metastatic effects of PF in breast cancer model.

Wu et. al. recently have the shown the anti-cancer effects of PF by modulating cholesterol homeostasis (21). The study showed that IC₅₀ of PF was between 2–4µM after 48h treatment in B16/F10, LL/2, 4T1 and CT26 cells. Similar to Wu et. al., we observed an IC₅₀ of about 3.7µM in 4T1 cells at 48h. Our results showed an IC₅₀ ranging between 5–8µM and 3–5µM after 24h and 48h treatment respectively in all the breast cancer cell lines tested. In another recent study, PF was shown to induce apoptosis in pancreatic cancer cells by activating protein phosphatase 2A (PP2A) (41). Our results also showed suppression of cell migration and invasion by PF treatment, suggesting anti-metastatic effects in TNBC cells.

Integrins have recently gained attention as an important therapeutic target in various cancer types (42,43). The heterodimerization of integrin α6 with β4 plays significant role in breast tumor progression. In fact, overexpression of integrin α6β4 has been observed in breast tumors (10). Our results showed significant suppression of integrin α6β4 by PF treatment. Integrin signaling is known to be mediated by downstream activation of FAK, Paxillin, Rac, ROCK proteins (44). Interestingly, PF treatment also inhibited the expression as well as activation of these downstream proteins in all the breast cancer cell lines tested. We have previously shown that TGFβ treatment increased the expression of integrin α6, leading to reduced cleavage of caspase 3 induced by cucurbitacin B (10). In agreement, results from current study also showed increased integrinα6 expression by TGFβ treatment and reduced cleavage of caspase 3 by PF, /indicating integrin α6β4 as a target of PF in breast cancer cells. Integrin α6β4 has been shown to have role in cell migration by laminin (45). Hence laminin coated culture dishes were used to evaluate the anti-cancer effects of PF. Our results showed that laminin reduced PF-mediated suppression of integrin signaling. Interestingly, we also observed disruption of integrinα6β4 dimerization by PF treatment. Furthermore, our results showed significantly reduced mRNA level of integrin α6 by PF treatment indicating that the inhibition of integrin α6 was at transcriptional level. In addition to integrin α6β4, integrin αvβ3 heterodimer plays a critical role in breast cancer metastasis to bones. Our results showed that PF treatment down-regulated the expression of αv and β3 integrin as well. These results clearly indicate that PF specifically inhibits integrin signaling in triple negative breast cancer cells.

PF is an orally bioavailable antipsychotic drug with doses ranging from 20–250 mg/week. Wu et. al., did not observe any statistically significant 4T1 tumor growth suppression with an oral dose of 0.06–0.12 mg/week/mouse (3–6 mg/kg/week) PF in mice (21). Interestingly, our results showed significant tumor growth suppression in all the three in vivo models by 10mg/kg PF administration every day by oral gavage. The human equivalent dose of 10 mg/kg PF used in mice is about 0.83 mg/kg. Moreover, 10mg/kg everyday administration of PF in mice is equivalent to 1.4 mg/week. PF treatment substantially reduced the establishment of metastasized breast tumors in the brain of mice in the intracardiac model. PF treatment also significantly inhibited the growth of established breast tumors in the brain of mice in an intracranial model. However, the inhibitory effect of PF was not as robust in orthotopic tumor model as compared to intracardiac or intracranial model. The exact reason behind the difference in efficacy of PF in different models is not clear at this point. Consistent with our in vitro observations, the tumor growth suppressive effects of PF were associated with reduced expression of integrinβ4 and increase in cleavage of caspase 3, as evaluated by immunohistochemistry in all the three tumor models, suggesting similar mechanism. Mice did not showed any significant behavioral side effects or any sign of toxicity when treated with 10mg/kg PF by oral gavage for 55 days.

Overall, our study provides convincing results to establish strong anti-tumor and anti-metastatic effects of PF in TNBC. To the best of our knowledge, ours is the first study to demonstrate use of PF for metastatic TNBC. Taken together, the outcomes from our study are very encouraging as they lay a foundation for repurposing PF for TNBC, which currently lacks any specific treatment options.