Activation of the orphan receptor GPR55 by lysophosphatidylinositol promotes metastasis in triple-negative breast cancer

Abstract

The orphan G protein-coupled receptor GPR55 has been directly or indirectly related to basic alterations that drive malignant growth: uncontrolled cancer cell proliferation, sustained angiogenesis, and cancer cell adhesion and migration. However, little is known about the involvement of this receptor in metastasis. Here, we show that elevated GPR55 expression in human tumors is associated with the aggressive basal/triple-negative breast cancer population, higher probability to develop metastases, and therefore poor patient prognosis. Activation of GPR55 by its proposed endogenous ligand lysophosphatidylinositol confers pro-invasive features on breast cancer cells both in vitro and in vivo. Specifically, this effect is elicited by coupling to Gq/11 heterotrimeric proteins and the subsequent activation, through ERK, of the transcription factor ETV4/PEA3. Together, these data show that GPR55 promotes breast cancer metastasis, and supports the notion that this orphan receptor may constitute a new therapeutic target and potential biomarker in the highly aggressive triple-negative subtype.

Keywords: G protein-coupled receptor; GPR55; cannabinoids; metastasis; triple-negative breast cancer.

Figure 1. GPR55 expression correlates with triple-negative tumors and poor patient prognosis

(A–C) Kaplan-Meier curves for disease-free survival (A), overall survival (B), and metastasis-free survival (C). Data plotted in A correspond to the human breast tumor tissues with complete clinical information contained in the 483-sample tissue microarray generated at the University Hospital of Kiel and described in [19]. Cases were scored as 0 (no staining), 1 (weak staining), 2 (moderate staining), or 3 (high staining), for GPR55 expression. A representative image of each category is shown in (D) upper panel. Scale bar, 0.25 mm. Samples scoring 0 were grouped as “low GPR55 expression”, and cases scoring 1-3 as “high GPR55 expression”. Data plotted in (B and C) were obtained from the microarray data sets published by the TCGA network in [20] (B), and from the microarray data set published in [21] (C). In these two panels, samples were ranked by GPR55 mRNA expression, and the best cutoff was manually selected. Survival curves were statistically compared by the log-rank test. (D) Association between GPR55 expression (as determined by staining scoring) and the molecular features of the breast tumor samples included in the TMAs described in (A). The Pearson's chi-squared test was used for statistical analysis. (E–G) Relative GPR55 mRNA expression in the indicated breast cancer subtypes (E and F) or human breast cancer cell lines (G). Data in (E and F) were obtained from the same database as in (B) (for E), and from database in [22] (in F). **p < 0.01 vs. the rest of the subgroups [except Normal-like in (E)], and vs. MDA-MB-231 cells in (G).

Figure 2. GPR55 confers pro-metastatic advantages on breast cancer cells in vitro and in vivo

(A–F) Cell invasion assays were performed with four different cell lines: MDA-MB-231 cells stably expressing a shRNA selectively targeting GPR55 (shGPR55) or a non-targeted shRNA (shC) (A–C), or shGPR55 cells stably expressing a GPR55 overexpression plasmid (shGPR55 + GPR55) or the corresponding empty vector (shGPR55 + pcDNA3) (D–F). Results are expressed as invasion fold increase towards 10% FBS or 0.5 μM LPI vs. shC (A and B) or shGPR55 + pcDNA3 (D and E) vehicle-treated cells, set at 1. (C and F) Relative GPR55 mRNA expression in GPR55 knocked-down (C) and GPR55 overexpressing cells (F). (G) Angular histograms of shC (left panel) and shGPR55 (right panel) cells migrating in response to a LPI gradient. (H) Representative images of the lung metastases generated by injection of the indicated luciferase expressing cell lines. (I) Quantification of the lung bioluminescence signal. *p < 0.05; **p < 0.01 vs. shC or shGPR55 + pcDNA3 vehicle-treated cells/animals. #p < 0.05; ##p < 0.01 vs. shC or shGPR55 + pcDNA3 FBS- or LPI-treated cells/animals.

Figure 3. GPR55-driven pro-invasive responses involve coupling to G_q/11 and activation of RhoA

(A) Activation of G_q/11 or G_12/13 proteins by LPI (10⁻⁵ M) as determined by Antibody-capture [³⁵S]GTPγS scintillation proximity assay (SPA). Results are expressed as percentage of [³⁵S]GTPγS basal binding (BB, binding obtained in the absence of the agonist, set as 100% for each cell line, and represented as a single bar in the graphs) to the indicated subunit. (B) Invasion of shGPR55 + GPR55 cells after incubation with 0.5 μM LPI or the corresponding vehicle (PBS) for 24 h, in the presence of a construct blocking G_q/11 (GRK2) or G_12/13 signaling (RGS), or the corresponding empty vector (GFP). (C) Activation of the indicated small GTPases after a 3 min incubation with 0.5 μM LPI. Results are expressed as fold increase activation over the corresponding vehicle-treated cells, set at 1. *p < 0.05; **p < 0.01 vs. vehicle-treated cells. ##p < 0.01 vs. GFP LPI-treated cells. #p < 0.01 vs. LPI-treated cells.

Figure 4. GPR55 activates the transcription factor ETV4/PEA3 through coupling to G_q/11 and stimulation of ERK

Relative ETV4 mRNA expression in the indicated cell lines (A and B), in the metastases derived from those cell lines (C) (see Figure 2G–2H for details on metastasis generation and treatment), and in shC cells transfected with constructs blocking blocking G_q/11 (GRK2) or G_12/13 signaling (RGS), or the corresponding empty vector (GFP) (E). In A, ETV4 expression was determined by the RT2 Profiler PCR Array of Human Tumor Metastases (see Supplementary Table S1 legend), and in (B), (C and E) by real-time quantitative PCR. In (B), siGPR55 1 and 2 represent two different GPR55 siRNAs. (D) Invasion of the indicated cell lines upon transient ETV4 knockdown. (F and G) Western blot analysis of phospho-ERK in shC and shGPR55 cells after treatment with 0.5 μM LPI or the corresponding vehicle (PBS) for 3 min (F) and shC cells after incubation with LPI and expressing the indicated G protein signaling blocking constructs (G). Representative luminograms are shown. Numbers on top of the images correspond to the densitometric analysis of pERK levels and are expressed as fold increase vs. the corresponding vehicle-treated cells, set at 1 (n = 3). (H) Relative ETV4 mRNA expression, as determined by real-time quantitative PCR, in shC cells challenged with 0.5 μM LPI and the MEK inhibitor U0126 (1μM) for 3 min. **p < 0.01 vs. siC or shC vehicle-treated cells. ##p < 0.01 vs. siC or shC LPI-treated cells (C), GFP LPI-treated cells (E) or LPI-treated cells (H).