3.1 LAE inhibits ER− breast cancer cells migration independent of nuciferine
To investigate LAE effects on cell migration, metastatic SK-BR-3, MDA-MB-231 and HCC1806 ER− breast cancer cell lines were used in the study. Previous studies have shown that cell proliferation interferes with cell migration. To address this issue, the proliferation of breast cancer cells was examined by CCK8 assay. The viability of SK-BR-3, MDA-MB-231 and HCC1806 cells was similarly unchanged between with low dose (50 and 100 µg/ml) of LAE supplement and without LAE (Fig. 1a). All of SK-BR-3, MDA-MB-231 and HCC1806 cells exhibited no cytotoxicity in morphological observation with 100 µg/ml or less LAE supplement (Fig. 1b). However, cell number was significantly reduced by high doss (250 µg/ml) of LAE supplement for 48 hours (Fig. 1a). These results indicated that low dose LAE did not alter breast cancer cell proliferation.
Next, we examined the LAE role in cell migration by wound healing assays with 50 and 250 µg/ml LAE supplement for 36 hours. Previous studies found that FBS promote cell proliferation to alter cell migration. To rule out the effects of FBS, all experiments were performed in an FBS-free medium. Cell wound healing distance analysis showed that LAE supplement significantly reduced the wound healing distance in three subtypes cells, suggesting cell migration was suppressed (Fig. 1c). Interestingly, 50 or 250 µg/ml LAE supplement showed similarly inhibitory effect on SK-BR-3 cells migration, while LAE supplement inhibited cell migration of MDA-MB-231 and HCC1806 cells in a dose dependent manner.
Nuciferine, a major bioactive component of lotus, inhibits the growth of cancer cells and breast cancer-associated bone loss [24–25]. To verify if nuciferine is involved in the inhibition of cell migration, SK-BR-3, MDA-MB-231 and HCC1806 cells were treated with nuciferine supplement. Cell viability (Supplementary Fig. 1a) and migration (Supplementary Fig. 1b) were unchanged in the presence of high dose of nuciferine, indicating that nuciferine was not associated with LAE-inhibited migration. Together, these results indicate that LAE supplement suppresses ER− breast cancer cells migration independent of nuciferine.
3.2 LAE-inhibited cell migration is independent of autophagy and Wnt signaling
Autophagic process is recognized to suppress cancer metastasis through decreasing EMT [26]. To determine if LAE-suppressed cell migration was regulated by autophagy and EMT-related Wnt/β-catenin signaling pathways directly, we examined β-catenin, snail, p62 and LC3 proteins levels by immunoblotting. Elevated LC3 protein level was detected in cells with high dose (100, 250 µg/ml) of LAE supplement. However, p62 protein level was similarly in cells with low and high dose of LAE supplement (Supplementary Fig. 1a). To address the controversial results, we used autophagy/lysosome inhibitor chloroquine (CQ) to inhibit autophagy flux. Inhibition of autophagy flux did not alter cell migration, revealed by wound healing distance (Supplementary Fig. 1b). Unchanged β-catenin and snail protein levels were detected in three subtypes of breast cancer cells with low or high dose of LAE supplement (Supplementary Fig. 1c). To further determine the role of Wnt signaling on the effect of LAE-inhibited cell migration, LiCl was used to activate the Wnt signaling. Wound healing assays revealed that the activation of Wnt signaling did not reverse the cell migration inhibited by LAE (Supplementary Fig. 1d), suggesting LAE-suppressed cell migration was independent of Wnt signaling. Thus, LAE-suppressed cell migration is independent of autophagy and Wnt signaling.
3.3 TGF-β1 signaling is associated with LAE-suppressed cell migration
Considerable literatures showed that TGF-β1 is recognized as a major regulator of EMT to regulate tumor metastasis [27–29]. Then we tested the protein level of TGF-β1 after treating with LAE. The results showed that TGF-β1 was obviously reduced in SK-BR-3 and HCC1806 cells at both low (even 10 µg/ml) and high concentrations (Fig. 2a). To further determine whether TGF-β1 signal was required for LAE-inhibited cell migration, 10 ng/ml TGF-β1 growth factor with 100 µg/ml LAE was used to treat SK-BR-3, MDA-MB-231 and HCC1806 cells. LAE-inhibited cell migration was significantly restored by the addition of TGF-β1 growth factor, as revealed by wound healing assays (Fig. 2b). Consistent with literature, TGF-β1 growth factor changed cells morphology from epithelial to mesenchymal transition (EMT), which was suppressed by LAE supplement (Fig. 2c). Thus, these observations demonstrate that TGF-β1 signal is involved in LAE-inhibited cell migration via inhibition of EMT.
3.4 SMAD3 and Erk1/2 phosphorylation are inhibited
To explore the signaling pathway involved in LAE-inhibited cell migration, we examined the canonical TGF-β1/SMAD3 signaling pathway and the non-canonical TGF-β1/PTEN/AKT and MAPKs signaling pathways [30–31]. Immunoblotting analysis showed that AKT phosphorylation and PTEN levels were unchanged in three subtypes of breast cancer cells with LAE supplement (Fig. 3a). Similar p38 and JNK phosphorylation levels were detected in cells with and without LAE supplement (Fig. 3a). However, the phosphorylated Erk1/2 level was significantly decreased in SK-BR-3 and HCC1806 cells even at a low LAE supplement (10 µg/ml), and mildly reduced in MDA-MB-231 cells (Fig. 3b). Interestingly, with LAE supplement, the phosphorylated SMAD3 level was significantly reduced in SK-BR-3 and MDA-MB-231 cells, but not in HCC1806 cells (Fig. 3c). Moreover, TGF-β1 supplement significantly restored Erk1/2 and SMAD3 phosphorylation levels reduced by LAE supplement in three subtypes of breast cancer cells (Fig. 3d).
Since mutant p53 mediates the TGF-β1 signaling pathway via Erk1/2 as well as SMAD3 signals, and the p53 in all of the three subtypes of breast cancer cells are mutant [32–35]. To address this issue, we examined the mutant p53 protein level in three subtypes of breast cancer cells with LAE supplement. Although p53 level was unchanged in SK-BR-3 and HCC1806 cells with LAE supplement compared to cells without LAE supplement, immunoblotting analysis showed that p53 level was significantly decreased in MDA-MB-231 cells with LAE supplement in a dose dependent manner (Fig. 3e). It provides another explanation for the effect of LAE on the inhibition of the metastasis in MDA-MB-231 cells via Erk1/2 as well as SMAD3 signals. Above all, these results demonstrate that LAE inhibited the cell migration of ER- breast cancer cells via suppressing the phosphorylation of TGF-β1/Erk1/2 as well as TGF-β1/SMAD3 signals.
3.5 LAE inhibits TGF-β1-related cell migration via downregulating hydrogen peroxide
Reactive oxygen species (ROS), especially hydrogen peroxide (H2O2) are recognized as associated with tumor metastasis [36–38]. TGF-β-related signaling promotes hydrogen peroxide production in several types of cells [39–41]. Simultaneously, H2O2 enhances TGF-β-mediated EMT via SMAD and MEK/ERK signaling [42]. To determine if intracellular H2O2 participate in the TGF-β1-related inhibition of cell migration, we first measured H2O2 level in cells with LAE supplement for 48 hours. With LAE supplement, intracellular H2O2 level was significantly downregulated in MDA-MB-231 and HCC1806 cells, but not in SK-BR-3 (Fig. 4a). Consistently, H2O2 scavenger catalase protein level was significantly increased even at a low concentration LAE supplement (10 µg/ml) (Fig. 4b). However, the proteins levels of SOD1, NOX2 and NOX4, responsible for the production of large amounts of ROS, were similarly in cells with or without LAE supplement. Thus, LAE supplement reduced H2O2 production in MDA-MB-231 and HCC1806 cells, likely via enhancing the catalase level.
To determine if downregulated H2O2 level caused LAE-induced inhibition of breast cancer cell migration, we used 25 nM H2O2 with 100 µg/ml LAE supplement to treat SK-BR-3, MDA-MB-231 and HCC1806 cells. Wound healing distance analysis showed that addition of H2O2 significantly restored LAE-inhibited cell migration in MDA-MB-231 and HCC1806 cells (Fig. 4c). However, SK-BR-3 cells displayed similarly cell migration with or without H2O2 supplement, suggesting other signals involved in LAE-inhibited cell migration. Interestingly, elevated intracellular H2O2 level was significantly detected in MDA-MB-231 and HCC1806 cells with 10 ng/ml TGF-β1 and 100 µg/ml LAE compared to cells with single LAE supplement, but not in SK-BR-3 cells (Fig. 4d). These findings indicate that LAE inhibits TGF-β1-related cell migration via downregulating hydrogen peroxide in MDA-MB-231 and HCC1806.
3.6 LAE inhibits tumor metastasis
To investigate the inhibitory effects of LAE on metastasis in vivo, we used subcutaneous xenograft nude mice model to examine tumor metastasis. The schedule of experiments was showed in Figure. 5a. To exclude the toxicity of LAE in vivo, xenograft nude female mice were fed with 0.5% LAE for 56 days and body weight of mice was monitored. Body weight in mice with and without LAE supplement was similar (Fig. 5b), indicating that LAE supplement did not cause significant systemic toxicity. Consistent with the above results, LAE supplement did not affect the growth of in-situ tumor compared to controls (Fig. 5c). However, the number and weight of new metastatic tumors derived from the primary tumor were significantly decreased in mice with LAE supplement compared to controls (Fig. 5d). Immunoblotting assays showed that LAE did not inhibit the phosphorylation of Erk1/2 and SMAD3 in liver. However, LAE supplement significantly downregulated the phosphorylation of Erk1/2 but not the phosphorylation of SMAD3 in lung (Fig. 5e). These data indicate that the effect of LAE on SMAD3 and Erk1/2 is dependent on tissue environment or extracellular signal stimulation.
Next, we used intravenous injection nude mice model to investigate the anti-metastatic role of LAE. The experiments schedule was showed in Fig. 6a. Strikingly, LAE supplement significantly suppressed mice body weight loss induced by tumor compared to control mice (Fig. 6b), which indicated that LAE may have the effect of improving the weight loss in patients with advanced breast cancer. And the number and weight of new formed metastatic tumors were significantly decreased in mice with LAE supplement compared to control animals (Fig. 6c). Interestingly, we found less lung congestion in mice with LAE supplement than that of control animals (Fig. 6d). The number and size of metastatic tumor nodules in lung and liver was decreased significantly. H&E staining revealed inflammatory infiltration and vascular invasion in lung of mice, while LAE supplement improved inflammatory infiltration and vascular invasion in lung of mice (Fig. 6e). In liver, LAE supplement significantly decreased the number of large metastatic tumors. Mice with LAE supplement showed significantly downregulated the phosphorylation of SMAD3 in lung and liver, but not the phosphorylation of Erk1/2 (Fig. 6f). Taken together, these data suggest that LAE inhibited tumor metastasis via suppressing the phosphorylation of SMAD3 and Erk1/2, which is dependent on different tissue environment and extracellular signal.
3.7 Extracellular matrix signaling are associated with LAE-suppressed cell migration
To further explore the role of LAE in inhibiting cell migration, we performed RNA-sequence to analysis genes change in SK-BR-3, HCC1806 and MDA-MB-231. After LAE treatment, all the differentially expressed genes were analyzed by (Gene Ontology) GO method in DAVID with the standard of P < 0.001 and FDR < 1. Results showed that extracellular matrix (ECM) organization, oxidation-reduction process, negative regulation of cell migration and cell adhesion were significantly enriched by LAE supplement in MDA-MB-231 (Supplementary Fig. 2a), and oxidation-reduction process, extracellular matrix organization, cell-cell adhesion and cell migration were significantly enhanced by LAE supplement in SK-BR-3 (Supplementary Fig. 2b), while extracellular region part and extracellular matrix part were significantly changed by LAE supplement in HCC1806 (Supplementary Fig. 2c). Furthermore, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis found that pathways involved in ECM-receptor interaction and focal adhesion as well as their related genes were significantly enriched (Supplementary Fig. 2d-f) in SK-BR-3, HCC1806 and MDA-MB-231. Notably, the ECM components and cell adhesion could be induced by the TGF-β1 and FAK signaling pathway [43–46]. Taken together, our results demonstrate that LAE-suppressed cell migration are associated with multiple different pathways in different cell lines, likely via extracellular matrix signaling.