Role of β3 subunit of the GABA type A receptor in triple negative breast cancer proliferation, migration, and cell cycle progression

ABSTRACT

Triple negative breast cancer (TNBC) is known for its heterogeneous nature and aggressive onset. The unresponsiveness to hormone therapies and immunotherapy and the toxicity of chemotherapeutics account for the limited treatment options for TNBC. Ion channels have emerged as possible therapeutic candidates for cancer therapy, but little is known about how ligand gated ion channels, specifically, GABA type A ligand-gated ion channel receptors (GABA_AR), affect cancer pathogenesis. Our results show that the GABA_A β3 subunit is expressed at higher levels in TNBC cell lines than non-tumorigenic cells, therefore contributing to the idea that limiting the GABA_AR via knockdown of the GABA_A β3 subunit is a potential strategy for decreasing the proliferation and migration of TNBC cells. We employed pharmacological and genetic approaches to investigate the role of the GABA_A β3 subunit in TNBC proliferation, migration, and cell cycle progression. The results suggest that pharmacological antagonism or genetic knockdown of GABA_A β3 subunit decreases TNBC proliferation and migration. In addition, GABA_A β3 subunit knockdown causes cell cycle arrest in TNBC cell lines via decreased cyclin D1 and increased p21 expression. Our findings suggest that membrane bound GABA_A receptors containing the β3 subunit can be further developed as a potential novel target for the treatment of TNBC.

Introduction

In 2023, an estimated 300,590 people will be diagnosed with invasive breast cancer [1]. Breast cancer can be divided into four subtypes with immunohistochemical analyzes: luminal A and luminal B, human epidermal growth factor receptor 2 (HER 2) positive, and basal-like [2]. The basal-like subtype, known as triple negative breast cancer (TNBC), lacks estrogen receptor (ER), progesterone receptor (PR), and HER2 expression [3]. Surgery, radiotherapy, chemotherapy, and immunotherapy are the treatment options for TNBC patients. However, TNBC treatment options are limited due to its unresponsiveness to hormone therapy and anti HER2 therapy. In addition, chemotherapy options can cause severe side effects including cardiotoxicity, neuropathy, and neutropenia [4]. TNBC is recognized for its heterogeneous nature and extremely aggressive onset, highlighting an urgent need to develop targeted therapies.

Ion channels have emerged as a possible therapeutic candidate in cancer therapy [5]. Recent studies have shown that voltage gated ion channels are upregulated in breast, cervical, and prostate cancers, and play a role in tumor progression [6,7]. Specifically, voltage gated chloride channels play a role in cell cycle regulation, cell proliferation, migration, and apoptosis in cancer cells [8,9]. Although ion channels have been identified as a potential therapeutic target in cancer therapy, little is known about how ligand gated chloride channels affect cancer pathogenesis. As compared to voltage gated ion channels, ligand gated ion channels are activated by specific neurotransmitters or ligands allowing for ligand-triggered modulation as opposed to voltage gated channels. While voltage gated channels are typically widespread throughout the body, certain subtypes of ligand gated channels are more enriched in some areas and/or overexpressed in cancers [10,11]. Moreover, several neurotransmitters have been shown to regulate breast cancer biology via effects on cell cycle, epithelial mesenchymal transition and the tumor microenvironment [12]. One such neurotransmitter is γ-amino butyric acid (GABA), a non-proteinogenic major inhibitory neurotransmitter in the central nervous system (CNS), which plays a tumor-supporting role in some solid cancers outside the brain, such as pancreatic, lung and colorectal cancers [13]. GABA signals via two mechanisms: GABA type A ligand-gated ion channel receptors (GABA_AR) and GABA type B G-protein coupled receptors (GABA_BR). While GABA signaling via metabotropic GABA_BR is implicated in lung and colon cancers, the role of membrane bound inhibitory GABA_AR in breast cancer, specifically TNBC is relatively understudied [14]. GABA_AR are ion channel coupled pentameric receptors that mediate chloride (Cl⁻) influx and neuronal inhibition in the adult brain, with 19 cloned subunits (α1–6, β1–3, γ1–3, δ, ε, θ, π, ρ1–3) [15]. There is consensus that most GABA_AR in the brain are assembled from two α, two β and one γ (or one δ) subunits [16]. The most abundant GABA_AR subtype in the CNS is α1β2/3γ2, which is a target for several clinically used anxiolytic, anesthetic and antiepileptic drugs. GABA binds in the extracellular domain at the α-β interface [12]. Importantly, a β subunit is required for the assembly of functional native GABA_ARs in the brain, and the expression of the β3 subunit is particularly crucial for proper inhibitory neurotransmission [17].

With respect to cancer, recent in vitro and in vivo studies have revealed that GABA_AR signaling via certain receptor subtypes is upregulated in prostate and pancreatic cancer as well as certain breast cancer subtypes [18–20]. In terms of TNBC, GABA_A π subunit has been shown to stimulate cell migration via the activation of extracellular regulated kinase 1/2 (ERK1/2) [21]. Recently, the GABA_A δ subunit has been implicated in the glutamic-pyruvate transaminase (GPT2)/GABA-induced metastasis via calcium-induced CREB signaling [22]. Employing Kaplan–Meier analysis of publicly available data, we found that higher gene expression of the GABA_A β3 and α1 subunits correlated with a significant drop in relapse-free survival rate in TNBC patients. Of the four major GABA_A subunits examined, higher GABA_A β3 subunit gene expression showed the most significant decrease in relapse-free survival in TNBC patients, but not in the cohort representing all breast cancer samples. Such a correlation was not observed with GABA_A δ and γ2 subunits in TNBC patients. Our western blotting analyses showed that GABA_A β3 subunit was expressed at higher levels in TNBC cell lines as compared to non-tumorigenic MCF 10A cells. Based on these observations and on the established role of GABA_A β3 subunit in inhibitory neurotransmission, we hypothesized that the β3 subunit containing GABA_ARs may play a role in TNBC growth. In this study, we employed pharmacological and genetic approaches to investigate the role of the GABA_A β3 subunit in proliferation, migration, and cell cycle progression in two TNBC cell lines. Our results suggest that pharmacological antagonism or genetic knockdown of GABA_A β3 subunit decreases TNBC proliferation and migration. Furthermore, GABA_A β3 subunit knockdown causes cell cycle arrest in TNBC cell lines via decreased cyclin D1 expression and increased p21 expression. These results suggest that GABA_A receptors containing the β3 subunit represent a novel cell surface target in the treatment of TNBC.

Methods

Kaplan–Meier survival plots

Kaplan–Meier survival plots were generated online with the Kaplan–Meier plotter database (https://kmplot.com/analysis/). Affymetrix gene chip data from the Kaplan–Meier plotter database were analyzed. The entire transcriptome of breast cancer patient samples for the two populations, all breast cancer patients and TNBC-only patients, was collected. Redundant samples between the two populations were removed. The number of patients at risk, hazard ratios (HR) with 95% confidence intervals and log-rank p values were calculated to analyze relapse-free survival probability in patients with high versus low expression of GABA_A β3, α1, δ and γ2 subunits. Log rank p < 0.05 indicates a statistically significant difference.

Cell Culture

All cell lines were obtained from ATCC. MCF 10A cells were cultured in mammary epithelial basal medium (MEBM) (Lonza, MD) supplemented with 5% horse serum (Invitrogen), 20 ng/mL epidermal growth factor (Lonza), 0.5 mg/mL hydrocortisone (Lonza), 10 μg/mL insulin (Sigma-Aldrich, MO), 100 ng/mL cholera toxin (Sigma-Aldrich), and 1% penicillin/streptomycin (ThermoFisher Scientific, MA). MDA MB 231 cells were cultured in Dulbecco’s modified eagle medium (DMEM) (ThermoFisher Scientific) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (ThermoFisher Scientific), 1% sodium pyruvate (ThermoFisher Scientific), and 1% penicillin/streptomycin (ThermoFisher Scientific). BT 549 cells were cultured in RPMI-1640 growth medium (ATCC, VA) supplemented with 10% FBS, 0.023 IU/mL bovine insulin, and 1% penicillin/streptomycin. HCC 1806 cells were cultured in RPMI-1640 growth medium supplemented with 10% FBS and 1% penicillin/streptomycin. All cell lines were incubated at 37°C in a 5% CO₂ incubator.

Western Blotting

Samples from cell lines were lysed in radioimmunoprecipitation assay (RIPA) lysis buffer supplemented with complete mini protease and phosphatase inhibitor cocktail (ThermoFisher Scientific). Lysed samples were sonicated and centrifuged at 10,000 × rpm for 10 min (min) at 4°C. Supernatants were collected for measurement of protein concentration using a bicinchoninic acid (BCA) kit (ThermoFisher Scientific). Protein samples (20 μg) were separated by SDS-PAGE and transferred to a PVDF membrane for probing. Membranes were blocked in TBS-Tween supplemented with 5% nonfat dry milk for 1 h (h) at room temperature (RT). The membranes were incubated with a primary monoclonal antibody against various GABA_AR subunits. The following antibodies were employed: GABA_A β3 (1:1000, #73–149, RRID: AB_2109585, Antibodies Inc, CA), GABA_A α1 (1:1000, #75–136, RRID: AB_2108811, Antibodies Inc), GABA_A δ (1:1000, #75–171, (RRID: AB_2107254, Antibodies Inc), GABA_A y₂ (1:1000, #75–483, RRID: AB_2725814, Antibodies Inc). All GABA_AR monoclonal antibodies were purchased from the UC Davis/NIH NeuroMab Facility, sold by Antibodies Inc. Other antibodies used include cyclin D1 monoclonal antibody (1:800, #2978, RRID: AB_2259616, Cell Signaling, MA), monoclonal p21 antibody (1:1000, #2947, RRID: AB_823586, Cell Signaling) and Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:20000, #10 R–2932, RRID: AB_11199818, Fitzgerald, MA) as the loading control. Each blot was incubated with the respective dilution of primary antibody overnight at RT. IRDye 680RD secondary antibodies (1:10,000, LI-COR BioSciences, NE) were used to visualize bound primary antibodies. The Odyssey CLx Imaging System (LI-COR BioSciences) was utilized for near-infrared fluorescent detection of proteins. Image Studio software on the Odyssey CLx was used to carry out densitometry analysis of western blot data (LI-COR BioSciences).

Immunofluorescence

2.5 × 10⁵ cells were seeded on Poly-L-Lysine (ThermoFisher Scientific) coated coverslips and fixed in 2% paraformaldehyde (ThermoFisher Scientific) for 15 min at RT. Cells were blocked in a blocking buffer (DPBS 1×, 5% normal goat serum) for 1 hat RT. Cells were incubated with the same GABA_AR β3 subunit antibody employed for western blotting analysis (#73–149, RRID: AB_2109585, Antibodies Inc, diluted 1:500 in DPBS, 5% normal goat serum) for 1 h at 37°C. After 3X washes, cells were incubated with an Alexa Fluor dye conjugated goat anti-mouse secondary antibody (ThermoFisher Scientific) for 1 h at RT. After 3X washes, coverslips were mounted with Prolong Gold Antifade with 4 ’6-diamidino-2- phenylindole (DAPI) counterstain (ThermoFisher Scientific). The mouse IgG isotype control antibody (02–6502, ThermoFisher Scientific) was used as a negative control to rule out any nonspecific labeling. Using the EVOS FL microscope (ThermoFisher Scientific), images were acquired at a 40× objective using the appropriate light cube. All images were acquired and scored by an observer that was blind to the antibody treatment groups. Scoring was carried out by counting total number of DAPI-stained cells and the number of cells positive for GABA_AR β3 (Alexa fluor signal) in six fields per “n” for all cell lines.

Biotinylation

The protocol followed was based on a cell surface protein biotinylation protocol employed in a study [23]. 1.5 × 10⁶ cells were seeded and incubated for 48 h. Cells were incubated for 30 min at 4°C, with gentle rocking with 400 μl per plate of membrane-impermeable biotin solution (2.5 mg/ml EZ-link Sulfo-NHS-LC-LC-biotin biotin reagent (ThermoFisher Scientific, cat# 21338) in DPBS). After biotinylation, each plate was washed with cold 100 mM Glycine in DPBS for 5 min at 4°C, with gentle rocking. Then, each plate was washed with cold 20 mM Glycine in DPBS for 5 min, with gentle rocking. Cells were lysed by adding 200 μl of lysis buffer LB3 into each plate and using cell scrapers to detach the cells. Cell lysates were collected in 1.5 ml tubes and placed in a rotating wheel at slow speed for 1 h at 4°C. While cells were lysing, immobilized NeutrAvidin Ultralink beads (ThermoFisher Scientific, cat#53150) were prepared by taking 40 μl of NeutrAvidin beads at 50% slurry and washing the beads with 0.5 mL DPBS, then with 0.5 mL lysis buffer 2 (LB2), and resuspended in μl of lysis buffer LB3. After the lysing incubation, the lysates were centrifuged at 16,000 × g for 15 min, and the supernatants were transferred to two 1.5 mL tubes. One tube was stored at −80°C and served as the “input” sample. The other tube (“pull-down” sample) was incubated with the 40 μl of NeutrAvidin beads overnight in a rotating wheel at 4°C. Then, the pull-down sample was centrifuged 16,000 × g for 30 s at 4°C. The beads were washed with LB3 buffer, LB2 buffer, saline washing solution (SWS) buffer, and LB1 buffer, in this order. Precipitated proteins were resuspended in a Laemmli sample buffer (BioRad, CA, cat#1610737) and analyzed via western blotting.

Cell Proliferation Assay

2.5 × 10⁴ cells were plated on 96 well plates and incubated overnight at 37°C, 5% CO₂. Cells were treated with GABA_AR agonist (muscimol) or antagonist (bicuculline methiodide (BCM)) for 48 h to assess the effects of pharmacological modulation on GABA_AR. We employed a dosage range of 10–100 μM for muscimol and 1–20 μM for bicuculline, based on previous in vitro and in vivo studies [24–26]. 20 μL of CellTiter 96® AQueous One Solution Reagent containing a tetrazolium compound[3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2 H-tetrazolium, inner salt; MTS] (Promega, WI) was added to each well, incubated at 37°C, 5% CO₂. Absorbance was read on a plate reader at 490 nm.

Cell Number

1 × 10⁶ cells were seeded in 100 mm culture dishes. Cells were either treated with 20 μM BCM or 100 μM muscimol for 24, 48, and 72 h or exposed to shRNA lentiviral particles to knock down the GABA_A β3 subunit. Dosing for BCM and muscimol was chosen based on our cell viability data showing a significant change in cell viability at 100 μM muscimol and 20 μM BCM in all TNBC cell lines. Floating cells were collected, and adherent cells were washed with DPBS (1X). The DPBS wash was collected. The cells were trypsinized with 0.05% trypsin-EDTA (ThermoFisher Scientific), collected, and combined with floating cells and DPBS wash to represent total cell number. The suspension was centrifuged, reconstituted in a complete cell medium, and counted using MUSE® Cell Count Reagent (Cytek Bio, MD) in the Guava® MUSE® Cell Analyzer (Cytek Bio).

Lentiviral Mediated Knockdown of GABA_AR subunit in TNBC Epithelial Cells

To knockdown the GABA_A β3 subunit, TNBC cells (HCC 1806 and BT 549) were cultured in appropriate complete medium until cells were 50% confluent. The medium was replaced with polybrene (5 μg/ml) containing medium to increase the efficiency of transduction. Cells were infected with transduction-ready scramble control shRNA lentiviral particles (#TR30021V, Origene, MD) or human GABA_AR shRNA lentiviral particles (four unique constructs targeting the GABA_A β3 gene, #TL304428V, Origene) for 24 h at multiplicity of infection (MOI) of 5. Plasmid vectors also harbored a puromycin resistant gene for selection and a green fluorescent protein (GFP) reporter as an indicator of transduction efficiency. Stably transduced cells were selected with puromycin (1–2.5 μg/mL).

Scratch wound healing assay

1 × 10⁶ cells were seeded in 100 mm culture dishes and left to grow until 100% confluent. Subsequently, using a sterile pipette tip, a scratch was induced on the monolayer of cells. Non-adherent cells were removed by washing twice with DPBS. Complete medium was added to the cells. For pharmacological experiments, 20 μM BCM or 100 μM muscimol was added to the medium. Dosing for BCM and muscimol was chosen based on the cell viability data showing a significant change in cell viability at 100 μM muscimol and 20 μM BCM in all TNBC cell lines. The images were captured at 0 h and respective time points using the transmitted light setting on the EVOS FL microscope at 4× objective. The distance of the scratch gap was quantified by measurement using Image J software (NIH) to calculate the rate of migration (Rm).

Transwell migration assay

Cells were resuspended in a complete growth medium containing a reduced serum concentration of 0.5% FBS. 2.5 × 10⁴ cells were seeded in transwell migration chambers with 8 μm pore membranes. Migration wells (VWR, PA) were placed in 24 well plates, each well containing 750 μL of complete growth medium at a normal serum concentration of 10% FBS. Chambers were incubated at 37°C, 5% CO₂ for 16 h to allow cells to migrate through the membrane. After incubation, the inner membranes of the chambers were rinsed with DPBS. The inner membranes were gently scraped with cotton swabs to remove any non-migrated cells. Migrated cells located on the outer membrane were fixed with 4% paraformaldehyde at RT and then rinsed with DPBS. Migrated cells on each membrane were stained with Prolong Gold Antifade with 4 ’6-diamidino-2-phenylindole (DAPI) (ThermoFisher Scientific) and captured at 20× objective using the EVOS FL microscope.

Cell cycle analysis

1.5 × 10⁶ cells were seeded and harvested after 96 h. Adherent and non-adherent cells were collected, counted using a Muse® Cell Count & Viability Kit (Cytek Bio), and analyzed using a Muse® Cell Cycle Kit (Cytek Bio). 1.0 × 10⁶ were collected from the respective cell suspension and centrifuged at 1,500 RPM for 5 min at 4°C. The cell pellets were washed with 100 μL DPBS, and 70% ethanol was added for fixation and stored in −20°C overnight. Following, 200 μL of the fixed cell samples were collected and centrifuged and washed with 100 μL DPBS (1×). 200 μL of Muse® Cell Cycle Kit solution was added to the samples. Samples were incubated in the dark at RT for 30 min and analyzed on the Guava® MUSE® Cell Analyzer (Cytek Bio).

Statistical analysis

t-test and one-way ANOVA followed by a Tukey post-hoc test in GraphPad Prism 8.0 (Boston, MA) were employed as needed. p < 0.05 was considered significant.

Results

TNBC patients with high expression of GABA_A β3 and α1 subunits have lower relapse-free survival probability Kaplan–Meier plots show the relapse-free survival rate for patients with breast cancer.

Patients that have higher mRNA expression of GABA_A β3 and α1 subunits showed a significant decrease in relapse-free survival rate, specifically in the TNBC cohort and not in the overall breast cancer cohort (Figure 1). GABA_A β3 expression showed the highest statistical significance (log p = 0.0014) in the TNBC cohort, wherein higher GABA_A β3 gene expression correlated with lower relapse-free survival probability. Higher GABA_A α1 mRNA expression was also associated with lower relapse-free survival probability in TNBC patients (log p = 0.047). However, patients with higher expression of the δ and γ2 GABA_A subunits did not show a significant change in relapse-free survival rate in TNBC patients.

Figure 1.

TNBC patients with high expression of GABA_A β3 and α1 subunits have lower relapse free survival probability. Kaplan-Meier survival plots of the entire transcriptome of all breast cancer patient samples for each GABA_A receptor gene. (b) Kaplan-Meier survival plots of the entire transcriptome of TNBC samples for each GABA_A receptor gene. The gene probe used for each subunit is shown in parentheses next to each GABA_A receptor gene. Each Kaplan-Meier survival plot shows low (black) and high (red) expression of each gene, number at risk, hazard ratio (HR), 95% confidence intervals (CI) and log-rank P-values of relapse-free survival from the Kaplan-Meier plotter database (https://kmplot.com/analysis/). Log rank p < 0.05 indicates a statistically significant difference.

Human TNBC cell lines show higher protein expression of GABA_A β3 and α1 subunits as compared to non-tumorigenic MCF 10A cells.

Western blot analyzes showed significantly higher expression of the GABA_A β3 and α1 subunits in TNBC cell lines (MDA MB 231, HCC 1806, and BT 549) as compared to the non-tumorigenic MCF 10A cells (Figure 2(b,c)). Within the three TNBC cell lines, HCC 1806 and BT 549 showed significantly higher GABA_A β3 expression as compared to MDA MB 231 cells. Only the HCC 1806 cell line showed a higher expression of the GABA_A δ subunit (Figure 2(d)). There was no significant difference in GABA_A γ2 subunit protein expression between TNBC and MCF 10A cells (Figure 2(e)). Additionally, we also found that GAD 65/67 (GABA synthesizing enzyme) expression is significantly increased as compared to the non-tumorigenic cells (Figure 2(f)). In support of the western blotting results, immunofluorescence analyses showed higher GABA_A β3 expression in non-permeabilized HCC 1806 and BT 549 cells as compared to MDA MB 231 and MCF 10A cells (Figure 2(g,h)). Further, biotinylation experiments confirmed the localization of the GABA_A β3 subunit on the cell surface. These results also showed the same trend of higher GABA_A β3 subunit expression in HCC 1806 and BT 549 cell lines, as observed in western blotting and immunofluorescence studies (Figure 2(i,j)).

Figure 2.

Human TNBC cell lines show higher protein expression of GABA_A β3 and α1 subunits as compared to non-tumorigenic MCF 10A cells. a) Representative images of cell morphology of each cell line. Scale bar = 200 µm. b) Representative western blot image of GABA_A β3 subunit and corresponding densitometry analysis in nontumorigenic MCF 10A cells versus three TNBC cell lines (MDA MB 231, HCC 1806 and BT 549), n = 3. Representative western blot images of (c) GABA_Aα1 subunit and corresponding densitometry analysis in nontumorigenic MCF 10A cells versus TNBC cell lines, (d) GABA_Aδ subunit and corresponding densitometry analysis in MCF 10A cells versus TNBC cell lines and (e) GABA_Aγ2 subunit and corresponding densitometry analysis in MCF 10A cells versus TNBC cell lines, n = 3. (f) GAD 65/67 expression and corresponding densitometry analysis in MCF 10A cells versus TNBC cell lines, n = 3. (g) Representative immunofluorescence images of MCF 10A, MDA MB 231, HCC 1806, and BT 549 cells stained with DAPI (blue) and probed with GABA_A β3 subunit antibody (red). Scale bar = 75 µm. (h) GABA_A β3 subunit expression scored as arbitrary units (AU) in all TNBC cell lines versus MCF 10A cells. n = 3. (i) GABA_A β3 subunit protein expression in whole cell samples (input) versus biotinylated samples (pulldown) in TNBC cells and MCF10A cells showing that GABA_AR β3 subunits are localized on the cell surface, n = 3. GAPDH is a cytosolic marker shown as loading control. ABCB1 protein expression was used as a positive control to confirm the detection of a membrane-bound protein in the samples. (j) Corresponding densitometry of the ratio of pulldown to input samples normalized to MCF 10As. * represent significance compared to MCF10A. # represent significance compared to MDA MB 231. Data are presented as mean ± SE, *p < 0.05, **p < 0.01, ****p < 0.0001, #<0.05, ####p < 0.0001 (ANOVA).

GABA _A R specific agonist muscimol increases cell viability, cell number, and migration, while GABA _A R specific antagonist bicuculline methiodide decreases cell viability, cell number, and migration in TNBC cell lines.

As shown in Figure 3(a,b) there was a significant increase in cell viability and cell number in all TNBC cell lines compared to the non-tumorigenic MCF 10A cells when exposed to 100 μM GABA_AR specific agonist muscimol. An increase in cell viability suggests that cells may be undergoing a proliferative effect when exposed to muscimol. In addition, BT 549 cells showed a significant increase in cell viability at 10 μM muscimol, indicating that this cell line may be more sensitive to muscimol.

Figure 3.

GABA_AR specific agonist, muscimol, increases cell viability and cell number, while GABA_AR specific antagonist, bicuculline methiodide, decreases cell viability and cell number in TNBC cell lines. (a) Cell viability of MCF 10A cells and TNBC (MDA MB 231, HCC 1806, and BT 549) cell lines exposed to muscimol for 48 h at various concentrations, n = 3. (b) Cell number of MCF 10A cells and TNBC cell lines exposed to 0 µM and 100 µM muscimol, n = 3. (c) Cell viability of MCF 10A cells and TNBC cell lines exposed to bicuculline methiodide for 48 h at various concentrations, n = 3. (d) Cell number of MCF 10A cells and TNBC cell lines exposed to 0 µM and 20 µM Bicuculline methiodide, n = 3. Data are presented as mean ± SE, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, (ANOVA).

In contrast to results with muscimol, GABA_AR antagonist BCM showed an opposite effect. BCM was chosen as the GABA_AR antagonist since it is a competitive inhibitor of GABA_AR and blood–brain barrier impermeable. 20 μM BCM decreased cell viability in all TNBC cell lines as compared to MCF 10A cells (Figure 3(c)). However, BT 549 cells and HCC 1806 cells were more sensitive to BCM as compared to MDA MB 231 cells since they responded to lower concentrations of BCM. 1 μM BCM decreased cell viability in BT 549 cells and 15 μM BCM decreased cell viability in HCC 1806 cells, MDA MB 231 cells showed a statistically significant decrease in viability only at 20 μM. Regarding cell number, there was a significant decrease in cell number in all TNBC cell lines compared to MCF 10A cells at 24, 48, and 72 h upon incubation with 20 μM BCM (Figure 3(d)). Muscimol increases migration, while bicuculline methiodide decreases migration in TNBC cells. As demonstrated in a wound healing assay, there was a significant increase in the rate of migration of HCC 1806 and BT 549 cells after exposure to 100 μM muscimol for 24 and 48 h (Figure 4). MDA MB 231 cells only show a significant change in the rate of migration after 48 h exposure to muscimol. As further demonstrated in transwell assays, TNBC cell lines showed an increase in the number of cells that migrated when exposed to muscimol (Figure 4). Conversely, all three TNBC cell lines showed a significant decrease in the rate of migration when exposed to 10 and 20 μM BCM for 24 and 48 h (Figure 5). In addition, transwell assays similarly demonstrated that the number of cells that migrated showed a significant decrease in all three TNBC cell lines when exposed to 20 μM BCM (Figure 5).

Figure 4.

Muscimol increases the rate of migration in TNBC cells in wound healing and transwell assays. (a) Representative images of scratch assay of MDA MB 231, HCC 1806, and BT 549 cells at 0,24, and 48 hours. Scale bar represents 650 µm. (b) Rate of migration of MDA MB 231, HCC 1806, and BT 549 cells exposed to Muscimol at various concentrations and time points, n = 3. (c) Representative images of transwell migration assay in TNBC cells exposed to muscimol. DAPI-stained cells (blue) represent cells migrated through the membrane. Scale bar represents 275 µm. (d) Number of cells migrated in transwell migration assay, n = 3. * Represent significance compared to 0 µM control at respective time points. Data are presented as mean ± SE, *p < 0.05, ****p < 0.0001, (ANOVA).

Figure 5.

GABA_AR antagonist bicuculline methiodide decreases the rate of migration in TNBC cells in wound healing and transwell assays. (a) Representative images of scratch assay of MDA MB 231, HCC 1806, and BT 549 cells at 0, 24, and 48 h. Scale bar represents 650 µm. (b) Rate of migration of MDA MB 231, HCC 1806, and BT 549 cells exposed to bicuculline methiodide at various concentrations and time points, n = 3. (c) Representative images of transwell migration assay in TNBC cells exposed to bicuculline methiodide. DAPI-stained cells (blue) represent cells migrated through the membrane. Scale bar represents 275 µm. (d) Number of cells migrated in transwell migration assay, n = 3. * represent significance compared to 0 µM control at respective time points. Data are presented as mean ± SE, *p < 0.05, ***p < 0.01, ****p < 0.0001, (ANOVA).

GABA_A β3 lentiviral shRNA knockdown decreases cell viability and cell number in HCC 1806 and BT 549 cells

HCC 1806 and BT 549 cell lines were chosen for lentiviral knockdown studies since they showed the highest expression of GABA_A β3 and α1 subunits as well as the highest sensitivity to modulation by both GABA_AR agonist and antagonist in cell migration analyses. For this study, we focused on the knockdown of the GABA_A β3 subunit. Each cell line was exposed to lentiviral particles harboring four different shRNA constructs of GABA_A β3 knockdown and a scramble control construct to optimize the efficiency of the knockdown (Figure 6). Western blot confirmed statistically significant GABA_A β3 lentiviral knockdown with all constructs in HCC 1806 cells, but only constructs 2 and 3 caused significant knockdown in BT 549 cells (Figure 6(c,d)). Therefore, lentiviral constructs 2 and 3 were chosen for further characterization in both HCC 1806 and BT 549 cells. Notably, constructs 2 and 3 also caused a concomitant significant decrease in GABA_A α1 subunit expression in BT 549 cells (Figure 6(e)). While in HCC 1806 cells, construct 3 was the only construct that caused a statistically significant decrease in GABA_A α1 subunit expression (Figure 6(f)). GABA_A β3 lentiviral knockdown caused a significant decrease in cell viability as compared to scramble control in both HCC 1806 and BT 549 cells (Figure 7). It is important to note that HCC 1806 and BT 549 cells showed maximal decrease in cell viability at 96 h. Regarding cell number, HCC 1806 and BT 549 cells exposed to GABA_A β3 lentiviral knockdown showed a decrease in cell number at 72 h (Figure 7). Additionally, construct 3 was more effective than construct 2 in decreasing cell viability and cell number in both BT 549 and HCC 1806 cells.

Figure 6.

GABA_A β3 shRNA lentiviral knockdown shows a significant decrease in GABA_A β3 subunit expression in HCC 1806 and BT 549 cells as well as a decreasing trend in GABA_A α1 subunit expression in both cell lines. a) Representative images of HCC 1806 (a) knockdown (KD) with scramble control, shRNA construct 2 and shRNA construct 3, respectively. Scale bar represents 125 µm. (c) Representative western blot images of GABA_A β3 expression and densitometry analysis in (c) HCC 1806 cells after GABA_A β3 shRNA lentiviral knockdown and (d) in BT 549 cells after GABA_A β3 shRNA lentiviral knockdown, n = 3. Representative western blot images of GABA_A α1 expression and densitometry analysis in (E) HCC 1806 cells after GABA_A β3 shRNA lentiviral knockdown and (f) in BT 549 cells after GABA_A β3 shRNA lentiviral knockdown, n = 5. * represents significance compared to scramble control. Data are presented as mean ± SE. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (ANOVA).

Figure 7.

GABA_A β3 lentiviral shRNA knockdown decreases cell viability and cell number in HCC 1806 and BT 549 cells. (a) Representative images of cell number of HCC 1806 cells after GABA_A β3 lentiviral shRNA knockdown (KD). Scale bar represents 200 µm. (b) Cell viability of HCC1806 cells after GABA_A β3 lentiviral shRNA KD, n = 3. (c) Cell number of HCC1806 cells after GABA_A β3 lentiviral shRNA KD, n = 3. (d) Representative images of cell number of BT 549 cells after GABA_A β3 lentiviral KD. Scale bar represents 200 µm. (e) Cell viability of BT 549 cells after GABA_A β3 lentiviral shRNA KD, n = 3. (f) Cell number of BT 549 cells after GABA_A β3 lentiviral shRNA KD, n = 3. * represent significance compared to scramble control. # represent significance compared to construct 2. $ represent significance compared to 24 h time point within each treatment, respectively. Data are presented as mean ± SE, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, #p < 0.05, ##p < 0.01, ####p < 0.0001,$$<0.01, $$<0.001 (ANOVA).

GABA_A β3 lentiviral shRNA knockdown decreased migration in HCC 1806 and BT 549 cells. As measured in a wound healing assay, HCC 1806 and BT 549 cell lines both showed a decrease in rate of migration after GABA_A β3 knockdown with construct 2 and construct 3 compared to scramble control (Figure 8(a-d)). Similarly, in a transwell migration assay, we observed a decrease in the number of migrated HCC 1806 and BT 549 cells after GABA_A β3 knockdown (Figure 8(e-h)). In agreement with trends seen in cell viability and cell number experiments, construct 3 showed the most significant decrease in both cell lines. In BT 549 cells, construct 3 was more effective than construct 2 in decreasing the number of migrated cells (Figure 8).

Figure 8.

GABA_Aβ3 lentiviral shRNA knockdown decreased migration in HCC 1806 and BT 549 cells. (a) Representative images of scratch migration assay using HCC 1806 cells after GABA_A β3 lentiviral shRNA KD. Scale bar represents 275 µm. (b) Rate of migration of HCC 1806 cells after GABA_A β3 lentiviral shRNA knockdown (KD), n = 3. (c) Representative images of scratch migration assay using BT 549 cells after GABA_A β3 lentiviral shRNA KD. Scale bar represents 275 µm. (d) Rate of migration of BT 549 cells after GABA_A β3 lentiviral shRNA knockdown (KD), n = 3. e) Representative images of transwell migration assay on HCC 1806 cells. DAPI-stained cells (blue) represent cells migrated through the membrane. Scale bar represents 275 µm. f) Representative images of transwell migration assay on BT 549 cells. g) Number of migrated HCC1806 cells after GABA_A β3 lentiviral shRNA KD n = 3. h) Number of migrated BT 549 cells after GABA_A β3 lentiviral shRNA KD n = 3. * represent significance compared to scramble control. # represents significance compared to construct 2. Data are presented as mean ± SE, **p < 0.01, ***p < 0.001, ****p < 0.0001, ##p < 0.01, ####p < 0.0001 (ANOVA).

GABA_A β3 lentiviral shRNA knockdown induces cell cycle arrest in HCC 1806 and BT 549 cells, decreases cyclin D1 expression, and increases p21 expression

After GABA_A β3 knockdown, TNBC cells showed cell cycle arrest in G0/G1 as well as a decrease in the S phase of the cell cycle (Figure 9) as measured by flow cytometry. Additionally, HCC 1806 cells showed a decrease in the M phase of the cell cycle. Further western blotting analyses revealed that cyclin D1 was downregulated after GABA_A β3 knockdown compared to scramble control in HCC 1806 and BT 549 cells (Figure 10(a,b)). Moreover, p21 expression was increased after GABA_A β3 knockdown in HCC 1806 and BT 549 cells as compared to scramble control (Figure 10(c,d)).

Figure 9.

GABA_A β3 lentiviral shRNA knockdown induces cell cycle arrest in HCC 1806 and BT 549 cells. (a) Flow cytometry DNA contentdistribution of HCC 1806 cells after GABA_A β3 lentiviral shRNA knockdown (KD), n = 3. (b) Cell cycle distribution of HCC 1806 cells after GABA_A β3 lentiviral shRNA knockdown (KD), n = 3. (c) Flow cytometry DNA contentdistribution of BT 549 cells after GABA_A β3 lentiviral shRNA knockdown (KD), n = 3. (d) Cell cycle distribution of BT 549 cells after GABA_A β3 lentiviral shRNA knockdown (KD), n = 3. * represents significance compared to scramble control. # represents significance compared to construct 2. Data are presented as mean ± SE, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, #p < 0.05 (ANOVA).

Figure 10.

GABA_Aβ3 lentiviral shRNA knockdown in TNBC cells causes a decrease in cyclin D1 and an increase in p21 expression. Representative images of western blot showing cyclin D1 expression and corresponding densitometry analysis in a) HCC 1806 cells after GABA_A β3 lentiviral shRNA KD and in (b) BT 549 cells after GABA_A β3 lentiviral shRNA KD, n = 3. Representative images of western blot showing p21 expression and corresponding densitometry analysis in (c) HCC 1806 cells after GABA_Aβ3 lentiviral shRNA KD and in (d) BT 549 cells after GABA_A β3 lentiviral shRNA KD, n = 4. * represents significance compared to scramble control. # represents significance compared to construct 2. Data are presented as mean ± SE, **p < 0.01, ***p < 0.001, ****p < 0.0001, #p < 0.05 (ANOVA).

Discussion

It is well established that γ-aminobutyric acid (GABA) mediated signaling via GABA_A ion channel receptors mediates inhibitory neurotransmission. Emerging evidence suggests that GABA plays a role in supporting cancer cell proliferation and metastasis, as well as the antitumor immune response [14]. However, mechanisms by which GABA_AR ion channel receptors affect TNBC growth remain unclear. Our results reported here demonstrate that the GABA_A β3 subunit, which is critical for binding of GABA to the α1-β3 interface as well as central inhibitory neurotransmission, also plays a critical role in TNBC proliferation, migration, and cell cycle progression. In parallel, Kaplan–Meier analyses of publicly available human gene expression data show that higher GABA_A β3 gene expression is associated with decreased relapse-free survival probability specifically in TNBC patients. To represent the heterogeneous nature of TNBC, we employed three TNBC cell lines: MDA MB 231 (mesenchymal stem-like), HCC 1806 (basal-like), and BT 549 (mesenchymal-like) [27]. Additionally, these cell lines represent African American (HCC 1806) and Caucasian (MDA MB 231 and BT 549) populations. MCF 10A cells were chosen as the cell line to represent normal human breast epithelial cells, as they are derived from mammary gland benign epithelial tissue and are a widely used model [28,29]. Muscimol and BCM were chosen for pharmacological modulation of the GABA_AR. Muscimol is a GABA_AR specific agonist that mimics GABA and binds to the α-β subunit interface but does not activate metabotropic GABA_BR [30]. BCM was chosen as the pharmacological antagonist as it is a competitive inhibitor of the GABA_AR [31]. As compared to the parent compound bicuculline, BCM has a quaternary ammonium charge that renders it blood-brain barrier impermeable. Thus, BCM would also be more suitable for future in vivo studies where this compound would spare GABA_AR in the brain and lead to fewer CNS adverse effects such as convulsions. It is also important to note that the IC₅₀ of BCM for GABA_AR is five times higher than its parent drug bicuculline [32,33]. Therefore, it is possible that the parent compound, bicuculline, may be more potent than BCM in decreasing TNBC cell proliferation and migration. We found that BT 549 cells were more sensitive to lower concentrations of muscimol and BCM as compared to the other TNBC cell lines in cell viability experiments. Additionally, BT 549 cells showed a significant increase at an earlier time point of 24 h when exposed to muscimol as compared to other TNBC cell lines. HCC 1806 cells and MDA MB 231 cells were sensitive to higher doses of muscimol and BCM in cell viability experiments. However, there was a more significant increase in migration in HCC 1806 and BT 549 than MDA MB 231 cells. Along with these trends, we observed that both HCC 1806 and BT 549 cells showed the highest expression of GABA_A β3 and α1 subunits. Thus, we postulate that higher expression of GABA_AR containing β3 and α1 subunits accounts for the increased pharmacological sensitivity of these two TNBC cell lines as compared to MDA MB 231 cell line. Similarly, GABA_A β3 knockdown caused a decrease in cell viability and migration in both HCC 1806 and BT 549 cell lines. However, BT 549 cells show a larger decrease in cell proliferation than HCC 1806 cells after GABA_A β3 knockdown. We reason that this larger decrease in BT 549 cells occurred due to better GABA_A β3 subunit knockdown efficiency in BT 549 cells (~60% protein knockdown) than in HCC 1806 cells (~40% protein knockdown).

A previous report has shown that GABA_A π subunit increases breast cancer metastasis through increasing migration through ERK 1/2 phosphorylation [34]. We found that the antiproliferative effects induced by GABA_A β3 knockdown were associated with G0/G1 cell cycle arrest and a reduction in cyclin D1 levels. Cyclin D1 is overexpressed in many cancers, including breast cancer, and is associated with G1 to S phase cell cycle progression [35]. Cyclin D1 often correlates with poor prognosis and increased metastasis [36]. Additionally, it has been shown that the expression of cyclin D1 led to increased TNBC migration in vitro [37]

Therefore, the decrease in cyclin D1 seen after GABA_A β3 knockdown may be contributing to decreased migration. All these observations support the use of cyclin-dependent kinase (CDK) inhibitors for breast cancer treatment; without CDKs, cyclin D1 is unable to form the necessary complex for G1 to S phase cell cycle progression [38]. However, CDK inhibitors are typically considered more effective in hormone-positive breast cancers due to the positive feedback loop between estrogen and cyclin D1 resulting in CDK 4/6- cyclin D1 complex-dependent cell cycle progression [39]. Since TNBC lacks hormone receptors, unraveling alternative mechanisms of CDK- cyclin D1 complex inhibition to stall cell cycle progression could be useful as a treatment for TNBC. Since GABA_A β3 knockdown led to cyclin D1 downregulation, studying if pharmacological inhibition of GABA_AR can be employed as a novel way to target cell cycle progression in TNBC is essential. To this end, there is a wealth of GABA_AR antagonists and negative allosteric modulators that could be repurposed to target migration and cell cycle progression in TNBC [40]. Along with cyclin D1 downregulation, we also found that p21 expression was increased after GABA_A β3 knockdown in HCC 1806 and BT 549 cells as compared to scramble control. p21 (also known as wildtype activating factor-1/cyclin dependent kinase inhibitory protein-1 or WAF1/CIP1), is a well-established CDK inhibitor of the G1 phase of the cell cycle that plays an important role in controlling cell cycle progression [41]. Supporting the tumor suppressor role of p21 in breast cancer, a recent study found that a combined delivery of p21 and p53 plasmid loaded nanoparticles significantly decreased cell growth of MDA MB 231 cells [42]. Our findings suggest that GABA_A β3 knockdown increases p21 expression which likely plays a role in decreased cell proliferation and cell cycle arrest that were observed. Previous research has shown that p21 plays a complex role in various cancers and can act via p53 dependent and independent pathways [41]. Therefore, our future research will focus on unraveling the mechanisms by which GABA_A β3 knockdown causes an increase in p21 as well as a decrease in cyclin D1 levels.

The relationship between TNBC and GABA_AR expression needs to be studied further to understand how GABA_AR overexpression leads to increased cell proliferation and migration. One aspect of this novel relationship includes the ion channel function of the GABA_AR. When GABA_AR in the mature CNS are activated, chloride (Cl⁻) ion influx and hyperpolarization occurs [16]. On the other hand, in brain cancer such as medulloblastoma, activation of alpha5 containing GABA_AR is associated with Cl⁻ efflux [43]. However, to our knowledge, the ion channel function of GABA_AR overexpressed in cancers outside of the brain is largely unknown. Our biotinylation data strongly suggest that β3 containing GABA_AR are expressed at the cell surface at higher levels in HCC 1806 and BT 549 cells. We also observed that α1 expression is reduced after β3 knockdown in HCC 1806 and BT 549 cells, suggesting that β3 KD causes a decrease in α1-β3 interfaces and likely a decrease in activation of GABA_AR. Therefore, our future studies will evaluate the role of GABA_AR ion channel function in TNBC. Interestingly, previous reports on gastric cancer suggest a strong link between intracellular Cl⁻ concentrations ([Cl⁻]_i) and cell cycle arrest. Specifically, gastric cancer cells cultured in low Cl⁻ had decreased [Cl⁻]_i, inhibiting cell growth via cell cycle arrest at the G0/G1 phase caused by reduction of CDK2 and phosphorylated Rb [44]. Moreover, a similar role for [Cl⁻]_i was also noted in prostatic cancer [45] and mouse osteoblast cells [46]. We hypothesize that GABA_A β3 knockdown results in a decrease in functional GABA_AR, leading to a similar decrease in [Cl⁻]_i which causes G0/G1 cell cycle arrest and decrease in cyclin D1. Incidentally, a decrease in cyclin D1 also affects cell adhesion, which could also account for the decrease in migration seen after GABA_A β3 knockdown [47].

Furthermore, gene editing studies have shown that the expression of GABA_A β3 subunit alone, can rescue inhibitory neurotransmission in a β1–3 subunit knockout mouse. Also, the inclusion of the β3 subunit is crucial for GABA_AR function in the CNS [17]. In our study, GABA_A β3 subunit knockdown did not cause a compensatory increase in α1 subunit, which forms the GABA binding site with the β3 subunit. On the contrary, we saw that the β3 subunit knockdown (construct 3) caused a decrease in GABA_A α1 subunit expression in both BT 549 and HCC 1806 cells. Notably, BT 549 cells showed a larger reduction than HCC 1806 cells in GABA_A α1 subunit expression. We suspect that the decrease in GABA_A α1 subunit expression in HCC 1806 cells was more modest due to lower GABA_A β3 subunit knockdown efficiency. We speculate that higher GABA_A β3 subunit knockdown efficiency (and resultant GABA_A α1 subunit reduction) in BT 549 cells may account for the greater effects seen with this cell line in cell viability, wound healing, migration, cell cycle, p21 and cyclin D1 protein expression experiments. Alpha and beta subunits are critical for assembly of most functional GABA_AR and are critical for GABA binding. Therefore, disrupting the assembly of these two subunits can dramatically decrease GABA_AR function. Overall, the subunit composition of upregulated GABA_AR in TNBC and the signaling mechanism(s) that lead to cell cycle arrest, decreased proliferation, decreased migration, and reduction of cyclin D1 levels are areas that warrant further investigation.

In further experiments, increased levels of glutamic acid decarboxylase (GAD 65/67) were detected in all TNBC cell lines as compared to MCF 10A cells, indicating that the TNBC cell lines examined have the enzymatic machinery required to synthesize GABA from glutamate (Figure 2(f)). A recent study showed that significant levels of intracellular and extracellular GABA (~6 mM intracellular and ~40 µM extracellular GABA) were detected in BT 549 TNBC cell lines [22]. Based on these findings, we did not add exogenous GABA to cell culture media in our experiments. Moreover, evidence from CNS literature shows that GABA_AR show significant basal (constitutive) GABA-independent activity which can be blocked by antagonists such as bicuculline and picrotoxin [26,48]. Whether the GABA_AR expressed in TNBC cell lines is basally active remains to be investigated.

In conclusion, the results presented here point to a novel role of β3 subunit containing GABA_AR in TNBC proliferation, migration, and cell cycle progression. Further studies are needed to fully elucidate the signaling mechanisms of GABA_AR in TNBC, so that therapeutics targeting this membrane-bound receptor can be developed as potential options for the treatment of TNBC.

Funding Statement

The work was supported by the Pennsylvania Department of Health.

Disclosure statement

No potential conflict of interest was reported by the author(s).