A novel role of ADGRF1 (GPR110) in promoting cellular quiescence and chemoresistance in human epidermal growth factor receptor 2-positive breast cancer

Abstract

While G protein-coupled receptors (GPCRs) are known to be excellent drug targets, the second largest family of adhesion-GPCRs is less explored for their role in health and disease. ADGRF1 (GPR110) is an adhesion-GPCR and has an important function in neurodevelopment and cancer. Despite serving as a poor predictor of survival, ADGRF1's coupling to G proteins and downstream pathways remain unknown in cancer. We evaluated the effects of ADGRF1 overexpression on tumorigenesis and signaling pathways using two human epidermal growth factor receptor-2-positive (HER2+) breast cancer (BC) cell-line models. We also interrogated publicly available clinical datasets to determine the expression of ADGRF1 in various BC subtypes and its impact on BC-specific survival (BCSS) and overall survival (OS) in patients. ADGRF1 overexpression in HER2+ BC cells increased secondary mammosphere formation, soft agar colony formation, and % of Aldefluor-positive tumorigenic population in vitro and promoted tumor growth in vivo. ADGRF1 co-immunoprecipitated with both Gαs and Gαq proteins and increased cAMP and IP1 when overexpressed. However, inhibition of only the Gαs pathway by SQ22536 reversed the pro-tumorigenic effects of ADGRF1 overexpression. RNA-sequencing and RPPA analysis revealed inhibition of cell cycle pathways with ADGRF1 overexpression, suggesting cellular quiescence, as also evidenced by cell cycle arrest at the G0/1 phase and resistance to chemotherapy in HER2+ BC. ADGRF1 was significantly overexpressed in the HER2-enriched BC compared to luminal A and B subtypes and predicted worse BCSS and OS in these patients. Therefore, ADGRF1 represents a novel drug target in HER2+ BC, warranting discovery of novel ADGRF1 antagonists.

Keywords: ADGRF1; GPR110; HER2; breast cancer; chemoresistance; quiescence; tumorigenesis.

FIGURE 1

ADGRF1 overexpression using pHAGE lentiviral-mediated infection of BT474 and SKBR3 parental cells. A, The normalized values of ADGRF1 mRNA expression and copy number in panel of HER2+ cell-lines from publicly available CCLE dataset. Arrows indicate two cell-lines with low ADGRF1 mRNA expression and copy number (BT474 and SKBR3), which were used to generate stable doxycycline (Dox)- inducible ADGRF1 overexpressing clones. B, immunoblotting analysis to detect the expression of full length ADGRF1 using anti-HA antibodies in clones 1 and 5 of BT474 cells and clones 1 and 2 of SKBR3 cells in absence (−) or presence (+) of Dox. Representative blot is shown from immunoblotting performed every 2 months during ongoing experiments. C and D, qRT-PCR showing the expression of full-length ADGRF1 in clones 1 and 5 of BT474 cells and clones 1 and 2 of SKBR3 cells grown in −/+ Dox. *indicates statistically significant difference compared to −Dox; P < .05 by unpaired t test (N = 5)

FIGURE 2

ADGRF1 overexpression increased anchorage-independent cell growth, mammosphere formation, and Aldefluor positivity of HER2+ BC cells in vitro and increased the rate of tumor growth in vivo. BT474 clones 1 and 5 and SKBR3 clones 1 and 2 were grown in absence (−) or presence (+) of doxycycline (Dox). A,B, Anchorage-independent cell growth by soft agar assay. C, Mammosphere formation assay. D, Aldefluor assay using FACS analysis. E, For in vivo studies, 1 million BT474 clone 1 cells grown in +Dox or −Dox for 72 hours, confirmed for ADGRF1 overexpression in +Dox versus −Dox using anti-ADGRF1 antibody, were injected subcutaneously in athymic nu/nu mice (n = 10 per group). Tumor growth was measured twice weekly. The rate of tumor growth was significantly higher in +Dox group versus −Dox group. *indicates statistically significant difference compared to −Dox; P < .05 by unpaired t test (N = 3–4)

FIGURE 3

ADGRF1 overexpression had no effects on total and phosphorylated HER1 and HER2 expression by immunoblotting. A, BT474 clones 1 and 5 and SKBR3 clones 1 and 2 were grown in absence (−) or presence (+) of doxycycline (Dox). After 72 hours of Dox treatment, protein was extracted, and the expression of phosphorylated HER1 and HER2 and total HER1 and HER2 was analyzed using immunoblotting. A representative immunoblot images of three individual replicates is shown. B, Densitometric quantitation of the relative intensity of phosphorylated (p)-HER1 and HER2 over total (t)-HER1 and HER 2 bands from three independent experiments (mean ± SEM). GAPDH was used as a loading control for visual assessment

FIGURE 4

ADGRF1 overexpression did not alter lapatinib activity on anchorage–dependent and –independent cell growth. BT474 clones 1 and 5 and SKBR3 clones 1 and 2 were grown in absence (−) or presence (+) of doxycycline (Dox) and in absence and presence of various concentrations of lapatinib for (A-B) the MTT assay or 1 nM of lapatinib for (C-F) soft agar assay (N = 3–4). For the determination of IC₅₀ of lapatinib, and neratinib, the data was fitted using no-linear regression analysis and 3-parameter logistic equation: Y = Bottom + (Top-Bottom)/ (1 + 10^{^}(X-LogIC50)) using GraphPad Prism version 8.0c. *indicates statistically significant difference compared to −Dox and indicates statistically significant difference compared to vehicle in the same Dox group by Two-way ANOVA; P < .05

FIGURE 5

ADGRF1 overexpression activates Gαs/Gαq pathways. BT474 clone 1 and SKBR3 clone 2 were grown in absence (−) or presence (+) of doxycycline (Dox). A, Co-immunoprecipitation with anti-Gαs and immunoblotting with anti-HA indicate that ADGRF1 couples to Gαs in both BT474 and SKBR3. B, Basal levels of cAMP were significantly increased upon ADGRF1 overexpression with Dox in both cells. C, Co–immunoprecipitation with anti-Gαq and immunoblotting with anti-HA indicate that ADGRF1 couples also to Gαq in both cells. D, Basal levels of IP1 were significantly increased upon ADGRF1 overexpression with Dox in BT474 but not SKBR3 cells. *indicates statistically significant difference P < .05 by unpaired t test (N = 3–4)

FIGURE 6

ADGRF1 coupling to Gαs pathway is pro-tumorigenic. BT474 clone 1 and SKBR3 clone 2 were grown in absence (−) or presence (+) of doxycycline (Dox). A, Mammosphere formation assay showing that SQ22536 (100 μM) decreased secondary mammospheres compared to Vehicle (Veh) in +Dox cells only. Whereas BIM I (10 μM) increased the number of secondary mammospheres compared to vehicle (Veh) in both −Dox and +Dox cells, B, Mammosphere formation assay showing that Forskolin (Fsk, 10 μM) increased whereas m-3M3FBS (m3M, 50 μM) decreased the number of secondary mammospheres compared to vehicle (Veh) in both −Dox and +Dox cells. Synaptamide (Syn, 10 nM) showed no change in secondary mammospheres compared to Veh in both −Dox and +Dox cells. Synaptamide (Syn, 10 nM) did not alter basal levels of cAMP upon ADGRF1 overexpression in (C) BT474 Clone 1 or in (D) SKBR3 Clone 2. Synaptamide (Syn, 10 nM) did not alter basal levels of IP1 upon ADGRF1 overexpression in (E) BT474 Clone 1 or in (F) SKBR3 Clone 2. *indicates statistically significant difference compared to −Dox by unpaired t test, P < .05. ^#indicates statistically significant difference between various treatment groups; P < .05 by Two-way ANOVA, Sidak’s multiple comparisons test, (N = 3–4)

FIGURE 7

Bioinformatic analysis and validation studies identify cell cycle arrest and chemoresistance indicating promotion of quiescence with ADGRF1 overexpression in HER2+ BC. BT474 clones 1 and 5 were grown in absence (−) or presence (+) of doxycycline (Dox) for 72 hours. A, RNAseq analysis was performed (N = 2, each in triplicates) and Over-Representation Analysis (ORA) using 663 differentially expressed downregulated genes with a false discovery rate (FDR) q-value < .05 showed eight enriched pathways with FDR q-value < .1 and two pathways with FDR q-value < .05, and the enrichment score was plotted as enrichment ratio. B, RPPA analysis was performed (N = 3, each in triplicates) and the heatmap showing the common downregulated proteins related to cell cycle upon ADGRF1 overexpression with FDR q value < .05. C, Cell cycle analysis (N = 3) showing G0/1 arrest induced by ADGRF1 overexpression. D, A reduction in Ki67 expression upon ADGRF1 overexpression was confirmed using IHC in BT474 clones 1 and 5. E, MTT assay was performed for cells grown with various concentrations of docetaxel (N = 3). ADGRF1 overexpression led to about 10-fold reduction in docetaxel potency, suggesting chemoresistance

FIGURE 8

ADGRF1 is overexpressed and amplified in HER2-enriched and basal subtypes of breast cancer and predict BCSS and OS in HER2-enriched but not basal subtypes. Analysis of The Cancer Genome Atlas (TCGA) RNA-Seq and copy number dataset (http://gdac.broadinstitute.org/) showing: A, ADGRF1 RNA expression; and B, corresponding copy number alterations in different BC subtypes. At RNA level, ADGRF1 gene expression was significantly higher in HER2+ and basal subtypes compared to luminal A and B BCs (Wilcoxon test, P < .05). ADGRF1 gene was amplified in basal and HER2+ subtypes of BC (Wilcoxon test, P < .05). C-F, Survival curves using METABRIC database in patients with HER2-enriched and basal BC subtypes with high versus low ADGRF1 expression. Kaplan-Meier curves for BC-Specific Survival (left panel) and Overall Survival (right panel) of patients with; (C and D) HER2-enriched (n = 220); and (E and F) basal (n = 199) BC subtypes with high versus low expression of ADGRF1