Combinatorial effects of cannabinoid receptor 1 and 2 agonists on characteristics and proteomic alteration in MDA-MB-231 breast cancer cells

Cell culture

In this study, MDA-MB-231 (RRID CVCL_0062; HTB-26, ATCC, VA, USA) was used as a human triple-negative breast cancer cell model; while, MCF-7 (RRID CVCL_0031; HTB-26, ATCC) was used as a control less aggressive breast cancer model. The cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) (12800017, Gibco, Texas, USA) supplemented by 10% fetal bovine serum (F7524, Sigma-Aldrich, MO, USA) and 1% penicillin-streptomycin (15140122, Gibco, Texas, USA). Moreover, 4T1 (RRID CVCL_0125; CRL-2539, ATCC) was used as a mouse triple-negative breast cancer cells, which were maintained in RPMI 1640 medium (Gibco) supplemented 10% fetal bovine serum (Sigma-Aldrich), 1% Sodium pyruvate (11360070, Gibco) and 1% penicillin-streptomycin (Gibco). The cells were maintained at 37°C with 5% CO₂ in a humidified incubator and routinely tested for mycoplasma contamination by using the MycoStrip™ (Mycoplasma detection kit; rep-mys-50, InvivoGen, CA, USA).

Chemicals

CB receptor agonists ACEA (A9719, Sigma-Aldrich) and GW405833 (G1421, Sigma-Aldrich) were used in this study. Since ACEA has K_{i (CB1)}/ K_{i (CB2)} = 1,428-fold and GW405833 has K_{i (CB2)}/ K_{i (CB1)} = 1,217-fold [20, 21]; therefore, ACEA and GW405833 represented the highly selective CB receptor agonists to CB1 and CB2 in our study, respectively. Since both agents were dissolved in DMSO, DMSO was used as vehicle control for all experiments. The concentration of CB agonists used in this study included 7.5 μM ACEA, 15 μM ACEA, 30 μM ACEA, 15 μM GW405833, 1:1 combination (15 μM ACEA: 15 μM GW405833), 1:2 combination (7.5 μM ACEA: 15 μM GW405833) and 2:1 combination (30 μM: 15 μM GW405833).

Cell viability assay

The cytotoxic effect of CB agonists was carried out by cell viability assay as previously described [10]. Briefly, MDA-MB-231, MCF-7 and 4T1 cells were seeded in 96-well tissue culture plate (Jet Bio-Fil, Guangzhou, China) at a density of 1.0×10⁴ cells/well. After 24 hours of incubation, cells were treated with CB agonists, ACEA and/or GW405833, for 48 hours. The cell viability was investigated by using MTT assay. The experiment was carried out in three independent biological replicates with three technical replicates each.

Clonogenic assay

MDA-MB-231 and MCF-7 cells were seeded in 6-well tissue culture plate (Jet Bio-Fil) at a density of 3.3×10⁵ cells/well. After incubation for 24 hours, the cells were treated with CB agonists for 48 hours. Subsequently, cells were harvested by trypsinization (Trypsin, 25200072, Gibco, Texas, USA) and seeded into the new 6-well plate at a density of 800 cells/well. Then, the cells were maintained in complete culture medium, and the medium was replaced every two days. After 12 days of incubation, all cell culture medium was discarded followed by two PBS washes. The cells were fixed by 100% methanol (106009, Sigma-Aldrich, PA, USA) for 30 minutes. Colonies were stained with a staining solution, 0.5% crystal violet (V5265, Sigma-Aldrich) in 0.25% methanol. After staining, scanned images were processed using ImageJ software (National Institutes of Health (NIH), MD, USA). To distinguish separated colony area, all images were analyzed through colony area plugin for ImageJ developed by Guzmán et al. [22]. The processed images were then adjusted to separate nearby colonies by using the watershed binary. To count the numbers of colonies, colony numbers were measured by using analyzed particles tool in ImageJ software. The clonogenic assay was performed in at least three independent biological replicates with two internal technical replicates each.

Cell cycle analysis

The percentage of cells in each cell cycle phase was determined using modified method from Lertsuwan et. al., [23]. Briefly, MDA-MB-231 cells were seeded in 6-well tissue culture plate (Jet Bio-Fil) at a density of 3.3×10⁵ cells/well for 24 hours. Then, they were treated with ACEA, GW405833 or the combination of ACEA and GW405833 in different ratios for 48 hours. After that, the cells were collected by trypsinization and fixed in cold 70% ethanol (A6531-04, Fulltime, China) at –20 ˚C. Afterwards, the samples were centrifuged for 5 minutes at 2,500 rpm and treated with 20 μg/mL propidium iodide (PI) DNA staining solution (P1304MP, Life Technologies, CA, USA) and 200 μg/mL PureLink™ RNase A (12091021, Life Technologies) for 30 minutes at room temperature in the dark. Cell cycle distribution was analyzed using a Flow Cytometer (Accuri™ C6 Plus; BD Biosciences, CA, USA). The experiment was performed in four independent biological replicates.

Sample preparation for proteomic analysis

MDA-MB-231 cells were plated at 3.3×10⁵ cells/well in 6-well tissue culture plate (Jet Bio-Fil) and allowed to attach overnight. The cells were then treated with ACEA, GW405833 or the 2:1 combination of both agents for 48 hours. At the end of the experiment, cell pellets were collected and were processed using a modified protocol based on a previously published proteome-profiling pipeline [24, 25]. Briefly, cell pellets were washed twice with PBS supplemented with protease inhibitor cocktail (04693159001, Roche, Mannheim, Germany). Afterwards, total proteins were extracted using modified radioimmunoassay precipitation (RIPA) buffer containing the protease inhibitor cocktail for 60 minutes at 4°C. Then, samples were centrifuged at 12,000 rpm for 30 minutes, and protein-containing supernatants were collected. Protein concentration of samples was measured by using BCA protein assay kit (23227, Thermo Scientific). In this study, protein samples from 3 independent experiments were used for further analysis.

To investigate protein expression profile in MDA-MB-231 cells exposed to CB agonists, extracted proteins were precipitated using ice-cold acetone. After precipitation, the protein pellet was reconstituted and adjusted to 1 μg/μL in reconstitution buffer, 0.2% RapiGest SF (186001860, Waters, UK) in 10 mM ammonium bicarbonate. To carry out sulfhydryl alkylation, fifty microliters of the protein solution (50 μg of proteins) were mixed with 5 mM DTT in 10 mM ammonium bicarbonate for 20 minutes at 90°C followed by incubation at room temperature for 40 minutes in the dark. After that, the mixture was placed in a desalting column (89883, Zebra-spin, Thermo Scientific) with trypsin at a ratio of 1:50 (enzyme: protein) to digest the peptide at 37°C. After 6 hours of incubation, 0.2% formic acid was added at a 1:1 (v/v) ratio to terminate the digestion process. The peptides were reconstituted in 0.1% formic acid and transferred to TruView LCMS vials (186005666CV, Waters, UK) for further analysis.

LC-MS/MS setting

The tryptic peptides were analyzed using LC-MS instrument consisting of an EASY-nLC 1200 system with a Q Exactive HF Orbitrap (Thermo Scientific). One microgram of peptides was separated on EASY-SPRAY PEPMAP RSLC C18 column (25 cm length x 75 μm inner diameter, 2 μm particle size (ES802, Thermo Scientific)). The flow rate was set at 300 nL/min with mobile phase A (0.1% formic acid in LC-MS grade water) and mobile phase B (0.1% in LC-MS grade acetonitrile). The column was maintained at 60°C. The linear gradient was changed from 2% to 45% of mobile phase B in 75 minutes and held for 15 minutes. The gradients were then increased to 95% of mobile phase B for 5 minutes followed by re-equilibration to 2% mobile phase B for additional 20 minutes. For MS1, the scanning range was set at 400–1,650 m/z with a resolution of 120,000, 3×10⁶ automated grain control (AGC) and 50 ms injection time (IT). For MS2, the resolution was set at 15,000 with 1×10⁵ AGC and 25 ms IT. The data was normalized to the higher energy collisional dissociation (HCD) fragmentation of 28 and processed by Proteome Discoverer 2.4 against the UniProt protein database for Homo sapiens. For protein and peptide identification, the parameters were set as follows: 1% of false discovery rates (FDR); peptide tolerance = 20 ppm; fragment tolerance = 0.05 Da; minimum fragment ion matches per peptide = 3; digest enzyme = trypsin; fixed modification = cysteine carbamidomethylation; variable modification = methionine oxidation. The normalization of relative protein abundance ratio was performed using the total peptide amount of each LC-runs (across all proteomic samples n = 24; 8 conditions with biological triplicate each). The proteomics data including raw files and search results have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD040666.

Data analysis, bioinformatic analysis and meta-analysis

Data analysis of proteomic profile from MDA-MB-231 exposed to CB agonists was performed by using Reactome (reactome.org) [26]. Then, the proteins with statistical significance (p < 0.01) were processed to gene ontology enrichment analysis by using Panther (pantherdb.org) [27]. Differential expression of proteins from each treatment as compared to control was reported as log₂ fold change normalized to control.

After that, protein-protein interaction (PPI) analysis was performed through STRING (string-db.org) [28] to observe the protein relationship and possible affected pathway. Furthermore, all PPI networks were constructed and visualized by using Cytoscape 3.9.1 software (cytoscape.org) [29].

For the meta-analysis study of gene expression, the analyses were acquired by using GEPIA2 (gepia2.cancer-pku.cn) [30]. In addition, to analyze the expression profile of breast cancer, the 179 datasets were obtained from the Cancer Genome Atlas Program (TCGA) and the Genotype-Tissue Expression (GTEx) databases for each breast cancer subtype in comparison to normal tissues. According to breast cancer subtypes, they are categorized based on their mutation and expression profiles into 4 molecular subtypes including luminal A (ER+, PR+ ≥ 20%, HER2− and Ki67+ < 20%), luminal B (ER+, PR+ < 20%, HER2− or HER2+, Ki67+ ≥ 20%), non-luminal (ER-, PR-, overexpressed HER2) and basal-like, namely triple-negative breast cancer (TNBC), (ER−, PR−, HER2−) [31, 32]. The statistical analysis was carried out by using analysis of variance (ANOVA).

Transwell invasion assay

Before cell seeding, the membrane of upper chamber was coated with diluted Matrigel (356234, Corning, USA) at a ratio of 1:5 with serum-free culture medium. MDA-MB-231 cells were seeded in serum-free culture media at 5×10⁴ cells per well on an 8 μm pore size Transwell insert (Falcon, Arizona, USA). In the lower chamber, DMEM culture medium supplemented with 20% FBS was added. Cells were treated with either 30 μM ACEA, 15 μM GW405833 or their 2:1 combination. Then, cells were cultured at 37°C with 5% CO₂ for 12 hours before staining with 0.5% crystal violet at room temperature. Pictures of invasive cells on the underside of the membrane were taken under a light microscope (Model MBL3200, A. KRÜSS Optronic GmbH, Hamburg, Germany). After that, the number of invasive cells was quantified using ImageJ software (NIH). The quantification of invasive cells was carried out following the previous published protocol [33]. All images were converted into 8-bit format, and the nearby invasive cells were separated by watershed binary tool. The numbers of invasive cells were counted by analyzed particles tool. Invasion assay was carried out for four independent biological replicates.

Lamellipodia formation assay

MDA-MB-231 cells were seeded into 8-well glass bottom chamber (LAB-TEK, New York, USA) at a density of 1×10⁴ cells per chamber. After 24 hours of seeding, 30 μM ACEA, 15 μM GW405833 or their 2:1 combination were applied to MDA-MB-231 cells for 12 hours. Then, cells were fixed with 4% formaldehyde (20909.330, VWR international, Rontenay-Sous-Bois, France) before staining with Alexa Fluor™ 546 Phalloidin (A22283, Thermo Scientific, MA, USA) for F-actin staining and Hoechst (62249, Thermo Scientific) for nuclei staining. Filopodia and lamellipodia formation was observed using confocal laser scanning microscope (FV1000, Olympus, Tokyo, Japan). Image analysis was performed using ImageJ software (NIH). To perform the measurement, all images were converted into 8-bit format. The quantification was carried out by using modified protocol from Weaver et. al. [34]. The polygon selection tool was used to select the edge of individual cells and to measure the whole cell spread in μm² according to scale bar. Whole cell spread was the average surface area of individual cells. Experiments were performed in six independent biological replicates.

Quantitative reverse transcription polymerase chain reaction (qRT-PCR)

To validate the proteomic results, the expression of representative genes in each figure was evaluated by qRT-PCR. MDA-MB-231 cells were seeded at a density of 3.3×10⁵ cells/well into 6-well plates (Jet Bio-Fil). Treatments including vehicle control (DMSO), 30 μM ACEA, 15 μM GW405833, and the 2:1 combination (30 μM ACEA and 15 μM GW405833) were applied to MDA-MB-231 cells after 24 hours of seeding. MDA-MB-231 cells were collected 48 hours post-treatment, and their RNA was extracted using RNeasy Mini Kit (74104, Qiagen, Hilden, Germany). One microgram of RNA was converted into cDNA by using iScript™ (1708841, Bio-Rad, CA, USA). Thereafter, 100 nanograms of cDNA were combined with primers specified in the S1 Table and Luna® Universal qPCR Master Mix (M3003S, Biolabs, Herts, UK). After that, the mixtures were added to qPCR microstrips (TLS0801, Bio-Rad), and qRT-PCR was performed utilizing the Bio-rad CFX connect™ Real-Time PCR Detection System. The relative expression was quantified from Ct value normalized to β-actin of their own samples. Experiments were performed in three independent biological replicates.

Statistical analysis

Statistical analysis of all results was performed by using GraphPad Prism 9 (GraphPad Software Inc., USA). The comparison between 2 groups was performed using a T-test. For the multiple comparisons among treatment groups, the results were analyzed by using one-way ANOVA. Then, Tukey post-test was carried out to determine the statistical difference between pairs of means. For all statistical analysis, p-value ≤ 0.05 was considered as statistical significance.

Pre-exposure of specific combination of ACEA and GW405833 suppressed MDA‑MB‑231 colony formation

According to our previous study, ACEA, GW405833 and their combination at different ratios suppressed MDA-MB-231 cell viability by apoptosis induction, especially in the 2:1 combination of ACEA and GW405833 [10]. In this study, cytotoxicity of CB agonist treatment was also investigated in the less aggressive human breast cancer cells (MCF-7) and mouse triple-negative breast cancer cells (4T1) as shown in S1 Fig. Our results showed that MCF-7 was notably less sensitive to CB agonists as compared to MDA-MB-231. A specific 2:1 combination ratio also showed a significant cytotoxic effect on both human and mouse triple-negative breast cancer cell lines but not for MCF-7 (S1I Fig). However, whether the growth inhibitory effects would still persist in the surviving cells after the treatment withdrawal is not known. CB agonist concentrations were determined by cell viability test as reported in our previously published data [10]. Accordingly, the same concentrations were used in this study. The long-term suppression was then examined by clonogenic assay in this study. After MDA-MB-231 cells were pre-exposed to ACEA and GW405833 at different concentrations and combination ratios including vehicle control, 7.5μM ACEA, 15 μM ACEA, 30μM ACEA, 15μM GW405833, 1:1 combination (15 μM ACEA and 15 μM GW405833), 1:2 combination (7.5 μM ACEA and 15 μM GW405833) and 2:1 combination (30 μM ACEA and 15 μM GW405833), surviving cells were seeded into the new plate for colony formation assay. The pre-exposure of 100 μg/mL (768.76 μM) of 5-Fluorouracil (5-FU) was used as positive control, which completely inhibited MDA-MB-231 colony formation even after treatment withdrawal for 12 days. Our results showed that MDA-MB-231 cells were more sensitive to GW405833 than ACEA as shown in Fig 1. At the same concentration of 15 μM, MDA-MB-231 colony formation was significantly lowered in GW405833 treated cells than ACEA treated group. As none of individual treatments significantly decreased colony number as compared to control, our results showed that only the combination at 2:1 ratio (30 μM ACEA and 15 μM GW405833) significantly reduced the number of MDA-MB-231 colony formation by 23.14% as compared to control (Fig 1A and 1B). Nevertheless, the 2:1 combination of ACEA and GW405833 did not affect the colony formation in MCF-7 cells (S2A and S2B Fig). The findings indicated that the 2:1 combination of ACEA and GW405833 promoted long-term suppression of MDA-MB-231 colony formation, while individual treatments and other combinations did not.

Co-activation of CB receptors induced cell cycle arrest in MDA‑MB‑231

As shown in Fig 1, a specific combination of CB agonists suppressed colony formation ability of MDA-MB-231 cells. To further study whether cell cycle alteration involved in CB-induced growth inhibition in breast cancer cells (MDA-MB-231) seen in previous and current studies, flow cytometry was carried out to investigate changes in cell cycle progression as shown in representative histograms in Fig 2A. Our results showed that there was no significant difference of cells in Sub-G0/G1, G0/G1 and G2/M phases between treatments as compared to control (Fig 2B, 2C and 2E). However, the proportion of cells in S-phase was markedly increased in MDA-MB-231 exposed to 30 μM ACEA and the 2:1 (ACEA: GW405833) combination by 33.51% and by 41.18% as compared to control, respectively (Fig 2D). Moreover, this combination also significantly augmented the percentage of cells in S-phase than the individual treatment of 15 μM GW405833. In addition, the 2:1 combination was shown to have the lowest proportion of MDA-MB-231 cells in the G2/M phase as compared to other combination treatments (Fig 2E). These results suggested that colony formation suppression in 2:1 combination corresponded to S-phase cell cycle arrest.

Proteomic profile alteration of proteins related to cell proliferation and survival of MDA-MB-231

To identify the proteomic signatures that were altered in breast cancer cells exposed to CB agonists, protein expression profile of MDA-MB-231 cells exposed to vehicle control, CB agonists alone or their different combination was determined using LC-MS/MS. The schematic workflow of proteomic study in this research was shown in Fig 3A. Firstly, the differential protein expression analysis was performed between each treatment group against vehicle control using Reactome (reactome.org), and the proteins with p-values lower than 0.01 and more than 1.5-fold-change were shown in volcano plots in Fig 3B–3H. The numbers of significantly downregulated and upregulated proteins as compared to vehicle control were shown in Table 1. Among those, the top 5 downregulated and upregulated proteins from each treatment as compared to control were shown in Table 2, and the numbers of unique proteins that were significantly altered in each individual and combination treatment were shown in Fig 3I (in red). Since the 2:1 (ACEA:GW405833) combination was shown to have the most significant effects on MDA-MB-231 cell growth in previous and current studies, proteomic signature in this group was further analyzed. The results showed that there were 26 unique proteins altered in the 2:1 combination treatment as compared to other treatments (S2 Table and Fig 3I). We next performed an enrichment analysis of gene ontology on these unique proteins using Panther (pantherdb.org). Our results indicated that 15 unique proteins in 2:1 combination treatment were involved in cellular processes (Fig 3J). Within these 15 unique changed proteins, one protein related to cell population proliferation process (Zinc Finger Protein; ZPR1) and one protein involved in cell cycle (Anaphase Promoting Complex Subunit 1; ANAPC1) were identified (Fig 3K). For proliferation-related protein, we found that ZPR1 was significantly decreased by 1.54-fold in MDA-MB-231 treated with the 2:1 combination of CB agonists as compared to a vehicle control (Fig 4A). Considering all significantly changed proteins in MDA-MB-231 exposed to the 2:1 combination as compared to control, other cell survival and proliferation-related proteins were also detected. SHC-transforming protein 1 (SHC1), mitogen-activated protein kinase 15 (MAPK15) and tumor protein P53 (TP53) were downregulated by 1.71-, 1.57- and 2.49-fold upon the exposure of this combination, respectively (Fig 4B–4D). Gene expression validation of ZPR1, SHC1, MAPK15 and TP53 in MDA-MB-231 cells exposed to CB agonists was also shown by qRT-PCR (S3A–S3D Fig). On the other hand, our results showed that the exposure of the 2:1 combination induced the expression of signal transducer and activator of transcription 1 (STAT1), mucin 1 (MUC1) and caveolin 1 (CAV1) by 2.75-, 2.42- and 1.71-fold, respectively (Fig 4E–4G).

For cell cycle-related proteins, ANAPC1, a regulator of metaphase-to-anaphase transition, was upregulated in MDA-MB-231 exposed to 2:1 combination of ACEA and GW405833 as compared to vehicle control (Fig 5A). The increase in ANAPC1 expression in MDA-MB-231 cells exposed to CB agonists was also confirmed by qRT-PCR (S3E Fig). On the other hand, cell division cycle protein 20 homolog B (CDC20B), dynactin subunit 3 (DCTN3), centrosomal protein 295 (CEP295), timeless circadian regulator (TIMELESS) and Ras-related nuclear protein (RAN) were significantly decreased by 1.64-, 1.94-, 1.81-, 1.71- and 1.67-fold upon the exposure of 2:1 CB agonists, respectively (Fig 5B–5F). As these proteins may also interact with other proliferation-related and cell cycle-related proteins presented in MDA-MB-231 treated with 2:1 combination treatment, we next investigated potential protein-protein interaction of all proteins with significant alterations (p < 0.01) in our samples through STRING analysis. The protein-protein interaction (PPI) diagram for cell proliferation-related proteins was shown in Fig 4H and cell cycle-related proteins were shown in Fig 5G. Together, these results demonstrated the potential regulatory proteins propensity for colony formation suppression and S-phase cell cycle arrest in MDA-MB-231 exposed to 2:1 (ACEA:GW405833) as compared to control.

CB receptor activation suppressed MDA‑MB‑231 cell invasion

Since the exposure of 2:1 (ACEA:GW405833) combination significantly reduced growth and promoted S-phase cell cycle arrest in MDA-MB-231, we aimed to test whether this condition also affected metastasis ability of MDA-MB-231 by conducting transwell invasion assay. Our results demonstrated that 30 μM ACEA significantly reduced number of invasive cells by 27.45% as compared to control (Fig 6A and 6B). Besides, MDA-MB-231 cell invasion was also markedly reduced in 15 μM GW405833 by 41.65% as compared to control. These results revealed the higher sensitivity of MDA-MB-231 cells to GW405833 than ACEA. Interestingly, combination between ACEA and GW405833 at ratio of 2:1 further suppressed MDA-MB-231 cell invasion by 59.80% as compared to control. Taken together, our results demonstrated that this specific combination was also shown to have significantly greater suppression of cell invasion as compared to individual treatments.

Whole cell spreading and lamellipodia formation were reduced in MDA-MB-231 after being exposed to CB agonists

According to Fig 6A, 6B, MDA-MB-231 cell invasion capability was decreased from CB agonist exposure. Since the formation of lamellipodia is known as one of the key characteristics for migrating cells [35, 36], we then performed fluorescent staining of F-actin in MDA-MB-231 exposed to CB agonists. Our results revealed that filopodia and lamellipodia formation was significantly decreased in MDA-MB-231 exposed to CB agonist treatments as shown in Fig 6C. Whole cell spread area of MDA-MB-231 cells was considerably reduced by 43.18% and 40.52% in 30 μM ACEA and 15 μM GW405833, respectively. The most dramatic reduction was found in the 2:1 combination with the decreased whole cell spread area by 54.57% as compared to control (Fig 6D). These results showed that reduction of cell invasion-related morphology correspond to decreased cell invasion in CB agonist-treated cells.

Quantitative proteomic analysis revealed the reduction of lamellipodia-promoting proteins; while, the induction of cell invasion suppressor proteins was observed upon CB agonist exposure

Proteomic profile of MDA-MB-231 cells exposed to 30 μM ACEA, 15 μM GW405833 and their 2:1 combination was analyzed to compare the proteomic signature changes against vehicle control. Hence, proteins with statistical difference as compared to control were selected and analyzed. Among these altered proteins, we found that 421 proteins were commonly expressed in MDA-MB-231 cells exposed to these 3 conditions (Fig 7A). For proteins associated with cell locomotion, AXL receptor tyrosine kinase (AXL) was downregulated by 1.68-fold and 1.98-fold in MDA-MB-231 exposed to 30 μM ACEA and 15 μM GW405833, respectively. The 2:1 (ACEA:GW405833) combination ratio was found to further downregulate AXL protein by 2.79-fold as compared to control (Fig 7B). These CB agonist treatments significantly reduced the expression of cellular actin dynamic regulator, VAV2 guanine nucleotide exchange factor (VAV2), by 1.82, 2.13 and 2.08-fold, respectively (Fig 7C). Likewise, Rac family small GTPase 1 (RAC1) and Ras-related protein Rab-5A (Rab5A) were significantly downregulated in MDA-MB-231 cells after exposed to 2:1 (ACEA:GW405833) combination (Fig 7D and 7E). The decreased expression of AXL, VAV2 and RAC1 in MDA-MB-231 cells exposed to CB agonists was also confirmed by qRT-PCR as shown in S3F–S3H Fig. Additionally, proteins playing a role in cell-to-cell adhesion and actin polymerization, catenin alpha 2 (CTNNA2) and beta-actin-like protein 2 (ACTBL2) were downregulated in MDA-MB-231 exposed to CB agonists as shown in Fig 7D and 7G. On the other hand, engulfment cell motility 2 (ELMO2) was upregulated after CB agonist treatment (Fig 7H). Similar to Figs 4 and 5, cell locomotion-related proteins with significant change (p < 0.01) in 2:1 combination treatment were further analyzed for potential interaction by using STRING as shown as PPI diagram in Fig 7I. The PPI diagram illustrated the direct interaction among effected protein including VAV2, RAC1 and other lamellipodia-related proteins [37, 38]. Accordingly, our results suggested that CB agonists suppressed MDA-MB-231 cell invasion, and this effect was associated with lamellipodia formation suppression.

ALDOC was upregulated in MDA-MB-231 exposed to CB agonists

Even though only 2:1 (ACEA:GW405833) combination significantly inhibited MDA-MB-231 colony formation after treatment withdrawal (Fig 1), individual and combination treatments were shown to suppress MDA-MB-231 cell viability and motility in our previous study [10] and Fig 6. We then investigated the common altered proteins in MDA-MB-231 cells exposed to CB agonists. There were 187 commonly altered proteins as compared to control from all treatments. Among these, our results showed that aldolase C (ALDOC) increased significantly in all conditions at least 11.85-fold (Fig 8A) as compared to control. The upregulation of ALDOC expression in MDA-MB-231 cells exposed to CB agonists was also confirmed by qRT-PCR as shown in S3I Fig. With this significant alteration, meta-analysis of ALDOC was performed to examine its association with breast cancer. Interestingly, by performing meta-analysis, the expression of ALDOC was downregulated in all breast cancer subtypes as shown in Fig 8C. Even though the results were not statistically significant, the increasing trend of percent survival in high ALDOC group was observed (Fig 8B). Since ALDOC downregulation was shown in all breast cancer subtypes, our results demonstrated the potential negative effect of CB agonist exposure on breast cancer via ALDOC induction.