GRK6 upregulation is extensively found in TNBC compared to non-TNBC and normal tissues and significantly correlates with a higher risk for distant metastasis in TNBC patients
First of all, we dissected the transcriptional profiling of GRK subfamily members (GRK1-7) in primary tumors and normal mammary epithelial tissues from TCGA breast cancer database. We found that GRK6 gene expression in primary tumors derived TNBC patients is significantly (p < 0.001) higher than that of normal tissues and primary tumors derived non-TNBC patients in TCGA breast cancer database (Fig. 1A). Moreover, in the paired normal adjacent tissues and primary tumors, the transcriptional levels of GRK6 was detected to be relatively higher in primary tumors compared to adjacent normal tissues (Fig. 1B). Accordingly, RT-PCR results showed that GRK6 upregulation are commonly found in primary tumors compared to adjacent normal tissues derived from 11 breast cancer patients (Fig. 1C). Immunohistochemistry staining results derived from The Human Protein Atlas database revealed that GRK6 protein expression in breast cancer tissues is higher than that in normal mammary tissues and mainly distributed in cytoplasm and cell membrane (Fig. 1D). Kaplan-Meier analyses demonstrated that a higher level of GRK6 transcript is closely associated with a poor distant metastasis-free survival probability in TNBC compared to other breast cancer subtypes (Fig. 1E). These findings implicate that GRK6 may act a pivotal role in the molecular mechanism underlying the metastatic progression of TNBC.
GRK6 expression causally associates with the metastatic potentials of TNBC cells in vitro and in vivo
To ascertain if GRK6 expression is associated with metastatic potentials of TNBC, we next performed a cellular migration assay using a trans-well cell culture against TNBC cell lines HCC38, HCC1806, HCC1937 and MDA-MB231. Whereas HCC1806 and HCC1937 cells harboring a lower expression of endogenous GRK6 expression exhibited a poorer migration ability, HCC38 and MDA-MB231 cells with a higher endogenous GRK6 levels showed a relatively stronger migration ability in a trans-well culture for 16 hours (Fig. 2A, B). The knockdown of GRK6 gene using its 2 independent shRNA clones in MDA-MB231 cells predominantly repressed the endogenous levels of GRK6 (Fig. 2C) and significantly (p < 0.001) suppressed the cellular migration ability (Fig. 2D). Robustly, GRK6 knockdown mitigated the lung colony-forming ability of MDA-MB231 cells in comparison with the non-silencing control cells in the tumor-bearing mice (Fig. 2E). Conversely, the enforced expression of exogenous GRK6A gene (Fig. 2F) dramatically enhanced the migration (Fig. 2G) and lung colony-forming (Fig. 2H) abilities of HCC1937 cells.
GRK6 membrane association activity is required for promoting the process of epithelial-mesenchymal transition and cellular migration ability in TNBC cells
Since the C-terminal palmotoylation of GRK6 is a critical process for its membrane translocation and subsequent interaction with an agonist-activated receptor [20], we generated HCC1937 cells overexpressing 2 exogenous GRK6 gene variants GRK6A with the palmitoylation sites at Cys-561/562/565 and GRK6B lacking the palmitoylation sites at C-terminal region (Fig. 3A). Thereafter, we defined GRK6A gene as wild-type and GRK6B as mutant form in this study. Western blot analyses indicated that the overexpression of wild-type, not mutant, GRK6 gene trigger the membrane localization of GRK6 in HCC1937 cells (Fig. 3B). Besides, the pharmaceutical inhibitor of GRK6 kinase activity by GRK-IN-2 dose-dependently suppressed the membrane localization of GRK6 in highly metastatic MDA-MB231 cells (Fig. 3C). Furthermore, our data showed that wild-type, not mutant, GRK6 overexpression triggers the progression of epithelial-mesenchymal transition (EMT) as judged by the decreased level of E-type marker E-cadherin and increased levels of M-type markers N-cadherin, Fibronectin, Vimentin and Slug (Fig. 3D). Accordingly, the lack of palmotoylation sites in the C-terminal region of GRK6 protein failed to force the cellular migration ability in HCC1937 cells (Fig. 3E). On the other hand, the treatment with GRK6 kinase inhibitor GRK6-IN-2 dose-dependently suppressed the EMT progression (Fig. 3F) and cellular migration ability (Fig. 3G) in the highly metastatic MDA-MB231 cells.
GRK6 recruits β-Arrestin 2 to promote the metastatic progression in TNBC cells
Because the interaction of GRK6 with β-Arrestin was found to regulate several cellular functions, we next examined the requirement of β-Arrestin activity for the GRK6-promoted metastatic progression in TNBC. Western blot analyses demonstrated that the enforced expression of wild-type, not mutant, GRK6 gene induces the membrane translocation and phosphorylation of β-Arrestin in HCC1937 cells (Fig. 4A). To delineate which β-Arrestin subtype is responsible for GRK6 membrane localization and activation, we next performed the membrane/cytosolic protein extraction. Another Western blot analysis revealed that β-Arrestin 2, not β-Arrestin 1, is activated and recruited to cell membrane in the GRK6A, not GRK6B, -overexpressing HCC1937 cells (Fig. 4A). The pharmaceutical inhibition of GRK6 kinase activity by GRK6-IN-2 predominantly suppressed phosphorylation of β-Arrestin and the membrane translocation β-Arrestin 2, not β-Arrestin 1, in MDA-MB231 cells in a dose-dependent manner (Fig. 4B). To delineate if the activity of β-Arrestin is exactly needed for the GRK6-triggered metastatic progression in TNBC, we next employed a selective β-Arrestin inhibitor Barbadin which is capable of blocking the agonist-promoted endocytosis of its associated receptors [21]. The treatment with Barbadin dramatically inhibited the GRK6-enhanced β-Arrestin phosphorylation/β-Arrestin 2 membrane translocation (Fig. 4C), EMT progression (Fig. 4D) and cellular migration ability (Fig. 4E) in the GRK6A-overexpressing HCC1937 cells.
GRK6/β-Arrestin 2-enhanced metastatic potentials of TNBC cells is mediated by the MAPK-NF-κB signalling axis
The interaction of GRK/β-Arrestin was found to foster cell migration via regulating the mitogen-activated protein kinase (MAPK) pathway [22]. Based on this finding, we next investigate the requirement of MAPK for the GRK6-enhanced metastatic potentials of TNBC. We found that the enforced expression of wild-type, not mutant, GRK6 induces the protein levels of phosphorylated MAPK members p38 and Erk1/2 in HCC1937 cells (Fig. 5A). The pharmaceutical inhibition of GRK6 kinase activity by GRK6-IN-2 was found to dose-dependently reduced the protein levels of phosphorylated p38 and Erk1/2 in MDA-MB231 cells (Fig. 5B). Accordingly, the block of β-Arrestin activity by Barbadin dose-dependently suppressed the GRK6-elevated protein phosphorylation of p38 and Erk1/2 in HCC1937 cells (Fig. 5C). The inhibition of p38 and Erk1/2 kinase activity by SB203580 and U0126, respectively, abrogated the GRK6-forced EMT progression (Fig. 5D) and cellular migration ability (Fig. 5E) in HCC1937 cells.
Since transcriptional activity of NF-κB is tightly regulated by MAPK and required for the expression of EMT-related genes [23], we further determined its roles in the GRK6-promoted metastatic progression in TNBC. We found that wild-type, not mutant, GRK6 overexpression robustly induces the protein phosphorylation and DNA-binding activity of NF-κB in HCC1937 cells (Fig. 6A). Similarly, the treatment with GRK6 kinase inhibitor GRK6-IN-2 dose-dependently suppressed the protein phosphorylation and DNA-binding activity of NF-κB in MDA-MB231 cells (Fig. 6B). Moreover, the inclusion of β-Arrestin, p38 and Erk1/2 inhibitors Barbadin, SB203580 and U0126 respectively reduced the GRK6-enhanced protein phosphorylation and DNA-binding activity of NF-κB in HCC1937 cells in a dose-dependent manner (Fig. 6C, D). Finally, the treatment with NF-κB inhibitors BAY-11-7082 and SN50 markedly suppressed the GRK6-enhanced EMT progression (Fig. 6E) and cellular migration ability (Fig. 6F) in HCC1937 cells