2.1 Reagents and antibodies

Cell culture reagents were purchased from Gibco Laboratories (Grand Island, NY). The following reagents and antibodies used in this study were purchased from different sources: Cannabidiol (Sigma Aldrich); human/murine EGF (Peprotech); GAPDH, AKT, p-EGFR/EGFR, Arginase-I, CD31, and p-ERK/ERK (Santa Cruz); and F4/80 (Abd Serotec) and Ki67 from (NeoMarkers); and p-AKT from (Cell Signaling). For flow-cytometery studies all antibodies were purchased from (Biolegend). For lung fixation Bouin's reagent was purchased from (Sigma Aldrich).

2.2 Cell culture

Human TNBC cell line SUM159 (Grigoriadis et al., 2012) was kindly provided by (Dr. Sarmila Majumder, The Ohio State University) and 4T1.2, a subclone of murine TNBC cell line 4T1 cells (Jin et al., 2014a; Lelekakis et al., 1999) was kindly provided Dr Robin Anderson (Eckhardt et al., 2005). SCP2, a subclone of MDA-MB-231 cells, was kindly provided by Dr. Joan Massagué (Minn et al., 2005). MVT-1 cell line was obtained from Dr. Johnson (Pei et al., 2004). MDA-MB-231 and RAW 264.7 cell lines were purchased from American Type Culture Collection (ATCC). The identity of these cells was verified regularly on the basis of cell morphology. Cells were cultured in DMEM containing 10% fetal bovine serum (FBS), 5 units/mL penicillin, and 5 mg/mL streptomycin (Corning Cellgro) and grown in 5% CO₂ incubator at 37 °C.

2.3 Cell proliferation assay

Cells were seeded in 96 well plates and treated with different concentrations of CBD with or without EGF (100 ng/ml) for 48 h in SFM and subjected to MTT assay (Roche) according to manufacturer's protocol.

2.4 Chemotactic assays

Chemotactic assays were performed using transwell chambers (Corning-Costar). Briefly, cells were pretreated with CBD or vehicle. Top chambers were loaded with (1.5 × 10⁵ cells for migration assay) or (2 × 10⁵ cells for invasion assay) in serum-free medium (SFM) and bottom chambers contained SFM in presence or absence of (EGF) or cancer cells conditioned media. Cells that migrated were stained and counted (Wani et al., 2014).

2.5 Immunofluorescence

Immunofluorescence was performed as described earlier (Ravi et al., 2014).

2.6 Flow-cytometry

Single cell suspensions were prepared from tumors (Nasser et al., 2012). Cells were incubated with anti-CD11b APC, anti-F4/80 PE, and anti-CD206 (Alexa Flour-488) for 1 h then fixed with 2% paraformaldehyde and acquired on a BD FACS caliber, then analyzed using Flowjo software.

2.7 Gelatin zymography

Gelatin zymography for collected conditioned media was performed as described earlier (Ravi et al., 2014).

2.8 Luciferase reporter assay

We used NF-kB luciferase reporter assay (Promega) to determine NF-kB activity as described previously (Ravi et al., 2014). Briefly, NF-kB luciferase constructs containing either wild type or NF-kB vector were transfected in the pretreated cells using lipofectamine 2000 (Invitrogen). For internal control, we co-transfected cells with Renilla luciferase vector. 24 h after transfection, EGF (100 ng/ml) was added and then incubated for another 24 h. Cells were lysed and luciferase assay was performed according to manufacturer's protocol.

2.9 Colony forming assay

One thousand cells were seeded in 60 mm² plates in DMEM supplied with 10% FBS. The next day the media were changed to DMEM with 3% FBS and incubated for 6 days with or without (EGF 100 ng/ml). Cells were fixed, stained and clones were counted (Ravi et al., 2014).

2.10 Mouse models

Female Balb/C and FVB mice were purchased from (Charles River Laboratories Inc.). Tumors were formed by orthotopically injecting 4T1.2 (1 × 10⁵) or MVT-1 (5 × 10⁵) cells into the 4th mammary glands. When tumors became palpable, mice were randomized and injected peri-tumorally with CBD (10 mg/kg) or vehicle on alternate days for 3 weeks. Tumors were measured every week with external calipers and tumor volume was calculated according to the formula tumor volume = (length × width² × 0.52) (Nasser et al., 2012, 2011).

2.11 Western blot (WB), real time PCR (RT-PCR) and immunohistochemistry (IHC)

WB, RT-PCR and IHC were performed as described earlier (Nasser et al., 2012; Qamri et al., 2009).

2.12 Wound healing assay

Wound healing experiment has been performed as described earlier (Preet et al., 2011). Briefly, Cells were treated with CBD or vehicle for 24 h. Monolayers were scratched with a sterile 200 μL micropipette tip, washed, and incubated in media supplemented with 0.1% FBS in the presence or absence of CBD or vehicle and EGF (100 ng/ml). After another 24 h, cells were fixed and photographed.

2.13 Statistical analysis

Results were represented as mean ± SD. Student's t test was used to compare vehicle and CBD-treated groups. P < 0.05 was considered to be statistically significant. For all graphs, ^∗ indicates P < 0.05, ^∗∗ indicates P < 0.01 and ^∗∗∗ indicates P < 0.001.

3.1 CBD inhibits breast cancer cell proliferation, migration and invasion

Cell proliferation is an important characteristic of cancer cells to survive and form a tumor mass (Ravi et al., 2014). As shown, TNBC cells express EGFR (Figure S4), Furthermore, EGF/EGFR axis plays an important role in proliferation of breast cancer cells (Wheeler et al., 2010); therefore, we performed MTT assay to evaluate CBD's effect on cancer cell viability. We found that CBD induces a strong dose-dependent decrease in proliferation of SUM159 and SCP2 human TNBC as well as 4T1.2 murine cell lines after 48 h incubation (Figure 1 (A–C)). We chose lower concentrations (CBD, 3, 6 and 9 μM) that have lower direct effect on proliferation for further studies. We tested the anti-proliferative ability of low doses of CBD in presence of EGF. As shown in Figure 2-A and B, CBD significantly inhibits EGF-induced proliferation in SUM159 and 4T1.2 cells respectively. In order to test the ability of CBD to inhibit cancer cell survival and proliferation to form single cell clones, we performed colony forming assay. We found that CBD significantly inhibited the number of cancer colony forming cells (SUM159 and 4T1.2) respectively after EGF stimulation (Figure 2 C–D).

Migration and invasion are two important characteristics of cancer cells to metastasize and form secondary tumors (Yamaguchi et al., 2005). When there is a gradient of extracellular stimuli like EGF, cells will undergo cytoskeletal rearrangement associated with an increase in cell adhesion, resulting in directional migration (Xie et al., 1998). Therefore, we evaluated the ability of CBD to inhibit EGF-induced migration of 4T1.2 and SUM159. CBD (6 μM) significantly inhibited EGF-induced migration in these cell lines (Figure 2 E–F). To confirm, we analyzed CBD effect in wound healing assay. As expected, CBD 6 μM inhibits EGF-induced wound closure in MDA-MB231 and SUM159 cell lines (Supp. Figure 3 A–B).

The extracellular matrix (ECM) and the basement membrane are barriers of cancer cells that hinder migration and invasion, so we were interested to examine CBD's modulatory effect on cancer cell invasion through ECM. We found that CBD (6 μM) inhibited EGF-induced invasion of SUM159 cells and 4T1.2 cells through ECM and basal membrane coated filters (Figure 2 G–H).

These results suggest that CBD -even at low concentrations- has the ability to inhibit the proliferative, migratory and invasive ability of TNBC cells induced by EGF.

3.2 CBD modulates EGF/EGFR signaling

EGFR signaling pathways are known to activate many cellular targets that are important for cancer cell survival, migration, and invasion. Both AKT and ERK are downstream targets of the EGF/EGFR pathway and they are crucial for cell growth and metastasis (Seshacharyulu et al., 2012). NF-kB activation is well known to promote the migration and invasion of TNBC cells (Helbig et al., 2003; Singel et al., 2014). Furthermore, there is a correlation between high levels of NF-kB activation and EGFR overexpression in breast cancer (Biswas and Iglehart, 2006). Therefore, we examined the ability of CBD to inhibit NF-kB signaling. As shown in Figure 3-A, CBD 6 μM inhibited EGF-induced translocation of NF-kB into the nucleus in SUM159 cells compared to vehicle-treated cells. To further understand the molecular mechanism by which CBD modulates the EGF/EGFR pathway, we treated SUM159 and 4T1.2 with CBD 6 μM for 48 h and then we stimulated the cells with EGF (100 ng/ml) for 15 and 30 min. As shown, CBD (6 μM) inhibited EGF-induced phosphorylation of EGFR, AKT and ERK in SUM159 (Figure 3-B). We also showed inhibition of phosphorylation of AKT and ERK by CBD after EGF stimulation of 4T1.2 cells (Supp. Figure 1-A).

EGF has also been reported to induce tumor cell invasion through induction of matrix metalloproteinases (Kim et al., 2009; Xu et al., 2011). In order to understand whether CBD-mediated inhibition of invasion is due to CBD effect on MMPs, we treated cells with CBD and analyzed MMPs activity in cancer cell-conditioned media by Zymography. As shown, CBD has reduced MMP2 and MMP9 activities in SUM159 and 4T1.2 respectively (Supp. Figure 1 B–C).

Cancer cells make protrusions to help them in adhesion and migration, and these protrusions are rich in actin filaments and adhesion molecules (Yamaguchi and Condeelis, 2007). We examined the effect of CBD on EGF-induced actin stress fiber and focal adhesion formation. By immunofluorescence, we showed that CBD 6 μM inhibits EGF-induced actin stress fiber formation as shown by changes of Phalloidin expression and focal adhesion formation as detected by changes in Vinculin expression in SUM159 cells (Figure 3-C).

These results indicate that CBD can inhibit EGF-induced tumorigenic properties of cancer cells through inhibition of the activation of EGFR, AKT, ERK and NF-kB signaling; in addition it blocks MMPs secretion, and suppresses phalloidin expression and actin stress fiber formation induced by EGF.

3.3 CBD inhibits tumor growth in breast cancer mouse models

To confirm the in vitro data, we evaluated CBD ability to inhibit breast cancer growth in vivo. We found that CBD inhibited tumor growth, in 4T1.2 mouse model, as shown by reduction in tumor volume (Figure 4-A) and weight (Figure 4-C) compared to control group. By IHC staining, we observed a dramatic decrease in the tumor cell proliferation, tumor vascularization and p-EGFR expression in CBD-treated group (Figure 4-G). To further investigate CBD's mechanism of tumor suppression, we used the tumor lysates to examine CBD effect on signaling pathways in vivo. CBD treated tumors showed decreased phosphorylation of AKT and ERK proteins (Figure 4-I).

We confirmed in vivo results using another highly aggressive mouse cell line (MVT-1). CBD treatment (10 mg/kg) significantly reduced tumor volume and weight in MVT1-1 model (Figure 4-B and D, respectively). We also analyzed Ki67, CD31 and p-EGFR levels by IHC staining. CBD-treated tumors had decreased proliferative activity, vessel formation and p-EGFR expression (Figure 4-H). Furthermore, CBD-treated group also showed less activation of AKT, and ERK proteins in MVT-1 tumor lysates compared to the control group (Figure 4-J).

These results suggest that CBD has the potential to inhibit tumor growth through suppression of tumor cell proliferation, angiogenic potential and inhibition of the activation of EGFR, AKT and ERK proteins.

3.4 CBD inhibits metastasis of breast cancer cells to the lung

Breast cancer patients' survival is limited in part by the development of distant metastases (Jin et al., 2014b). To study the efficacy of CBD to inhibit metastasis, two highly aggressive breast cancer cell lines (4T1.2 and MVT-1) were employed. The 4T1.2 subclone is known for its propensity to metastasize with rates superior to that of its parental cell line (4T1) (Lelekakis et al., 1999). After 3 weeks of treatment, CBD-treated groups showed significantly less number of metastatic nodules and less total lung weight in 4T1.2 and MVT-1 mouse models than control groups (Figure 5 A–F) and (Supp. Figure 5 A–D).

Since tumor cells secrete MMPs which enhance their ability to invade and metastasize to distant organs (Stamenkovic, 2000), we analyzed MMP2 and MMP9 expressions in tumor lysates of CBD-treated and control groups. The CBD-treated group had significantly lower expression of MMP2 and MMP9 in both 4T1.2 and MVT-1 tumors (Figure 5 G–H), suggesting that the anti-metastatic effects of CBD are likely mediated through decreased MMP2 and MMP9 secretion by the tumor cells.

3.5 CBD inhibits tumor growth and metastasis through inhibition of macrophage recruitment to tumor sites

Since tumor-associated macrophages (TAMs) are known to modulate tumor angiogenesis, cancer cell proliferation, and invasion (Place et al., 2011), we investigated CBD's effect on macrophage populations within the breast tumor microenvironment. First, we examined breast cancer xenografts for total (F4/80-positive) and M2 (Arginase-I-positive) macrophages by IHC. CBD-treated tumors showed decreased F4/80 and Arginase-I-positive cells compared to control group (Figure 6-A and B). Next, we analyzed macrophage populations in the primary tumors by flow-cytometry. The CBD-treated tumors have significantly smaller percentages of total macrophage (CD11b⁺, F4/80⁺) in MVT-1 and 4T1.2 tumors (Figure 6-C and Supp. Figure 2, respectively). Decreased percentage of M2 (F4/80⁺, CD206⁺) macrophages are detected in CBD-treated tumors of MVT-1 tumor-bearing mice (Figure 6-D).

Macrophages facilitate tumor angiogenesis, ECM breakdown, and cancer cell motility, which are crucial components of the metastatic process (Pollard, 2004). Furthermore, macrophages are needed for metastatic cancer cell survival and formation of pre-metastatic niche (Gil-Bernabé et al., 2012). Therefore, to investigate CBD effect on macrophage populations within secondary lung metastases, we analyzed F4/80 and arginase-I expressions in the lungs of vehicle- and CBD-treated groups by IHC staining. As expected, CBD-treated group had less F4/80 and Arginase-I expression in the lung metastases compared to vehicle-treated groups (Figure 6 E–F).

In order to understand why there is a reduction of macrophage population within the tumor, we treated 4T1.2 cells with CBD 9 μM or vehicle for 48 h, and the collected conditioned media were used as a chemo-attractant to test migration of RAW 264.7 cells. We observed a significant reduction in the number of migrated RAW 264.7 cells towards CBD-conditioned medium compared to the control-conditioned medium (Figure 7-C). We hypothesized that CBD might modulate cytokine production from tumor cells, which in turn affects macrophage recruitment towards the cancer cells. To test this hypothesis, we performed cytokine array analysis (R&D Systems), using the previously mentioned conditioned media. Interestingly, we found that CBD-treated 4T1.2 cells secreted less CCL3, GM-CSF and MIP-2 proteins compared to vehicle-treated cells (Figure 7 A–B).

These results suggest that CBD modulates cytokine production from tumor cells which leads to less recruitment of total macrophages and M2 macrophages into the primary and secondary tumor sites. This partially explains the ability of CBD to modulate the breast tumor microenvironment which helps in inhibition of tumor progression and metastasis to distant organs.