Agonist-Induced CXCR4 and CB2 Heterodimerization Inhibits Gα13

Published OnlineFirst January 12, 2018; DOI: 10.1158/1541-7786.MCR-16-0481

Signal Transduction Molecular Cancer Research Agonist-induced CXCR4 and CB2 Heterodimerization Inhibits Ga13/ RhoA-mediated Migration Kisha A. Scarlett1, El-Shaddai Z. White1,2, Christopher J. Coke1,2, Jada R. Carter1,2, Latoya K. Bryant1, and Cimona V. Hinton1,2

Abstract

G-protein–coupled receptor (GPCR) heterodimerization has is required for directional cell migration and the metastatic emerged as a means by which alternative signaling entities can potential of cancer cells. These studies in prostate cancer cells be created; yet, how receptor heterodimers affect receptor demonstrate decreased protein expression levels of Ga13 and pharmacology remains unknown. Previous observations sug- RhoA upon simultaneous CXCR4/CB2 agonist stimulation. gested a biochemical antagonism between GPCRs, CXCR4 and Furthermore, the agonist-induced heterodimer abrogated CB2 (CNR2), where agonist-bound CXCR4 and agonist-bound RhoA-mediated cytoskeletal rearrangement resulting in the CB2 formed a physiologically nonfunctional heterodimer attenuation of cell migration and invasion of an endothelial on the membrane of cancer cells, inhibiting their metastatic cell barrier. Finally, a reduction was observed in the expression potential in vitro. However, the reduced signaling entities of integrin a5 (ITGA5) upon heterodimerization, supported by responsible for the observed functional outputs remain elusive. decreased cell adhesion to extracellular matrices in vitro.Taken This study now delineates the signaling mechanism whereby together, the data identify a novel pharmacologic mechanism heterodimeric association between CXCR4 and CB2, induced for the modulation of tumor cell migration and invasion in the by simultaneous agonist treatment, results in decreased context of metastatic disease. CXCR4-mediated cell migration, invasion, and adhesion through inhibition of the Ga13/RhoA signaling axis. Activation Implications: This study investigates a signaling mechanism of CXCR4 by its cognate ligand, CXCL12, stimulates Ga13 by which GPCR heterodimerization inhibits cancer cell migra- (GNA13), and subsequently, the small GTPase RhoA, which tion. Mol Cancer Res; 1–12. 2018 AACR.

Introduction GPCR signaling has been that GPCRs can exist and function as monomers, homodimers, or as larger oligomeric complexes, a G-protein–coupled receptors (GPCR) represent a large fam- number of GPCRs have been identiﬁed to form functionally ily of cell surface signaling proteins that are involved in mul- relevant heterodimers after translation (4–7). GPCR heterodi- tiple physiologic functions and diseases (1, 2). In classical merization has slowly been accepted as a means by which new GPCR signaling, the binding of an agonist to a GPCR induces signaling entities can be created to amplify or desensitize a conformational change of the receptor that catalyzes the signaling mechanisms that would otherwise result from each exchange of guanosine diphosphate (GDP) from the Ga sub- individual receptor leading to diverse pharmacologic outcomes unit for guanosine triphosphate (GTP), and the dissociation of (4). Therefore, GPCR heterodimers have emerged as novel Ga from Gb subunits (2). Both Ga and Gb subunit com- targets for therapeutic development (3, 4, 7, 8). plexes stimulate several downstream effectors leading to diverse CXCR4 is a chemokine receptor, subclass of GPCR, which is physiologic outcomes (2, 3). While the conventional theory of deﬁned by its ability to induce directional migration of cells toward a chemotactic gradient (chemotaxis; refs. 9–11). The role of CXCR4, and its natural ligand CXCL12, has expanded 1Center for Cancer Research and Therapeutic Development, Clark Atlanta beyond leukocyte recruitment to include critical processes such University, Atlanta, Georgia. 2Department of Biological Sciences, Clark Atlanta as tissue remodeling, angiogenesis, hematopoiesis, cell prolif- University, Atlanta, Georgia. eration, and most notably, cell migration (10, 12). Its ability to Note: Supplementary data for this article are available at Molecular Cancer regulate multiple functions over the lifespan of neoplastic and Research Online (http://mcr.aacrjournals.org/). malignant cells makes CXCR4 a critical driver of tumorigenesis, Corresponding Author: Cimona V. Hinton, Center for Cancer Research and the progression of cancer, and metastasis (9, 10, 12, 13). Therapeutic Development, Department of Biological Sciences, Clark Atlanta Clinically, overexpression of CXCR4 correlates with increased University, 4017B Thomas Cole Science Research Center, 223 James P. Brawley tumor growth, invasion, angiogenesis, metastasis, and thera- Drive SW, Atlanta, GA 30314-4385. Phone: 404-880-8134; Fax: 404-880-6756; peutic resistance while CXCR4 antagonism has been shown to E-mail: [email protected] sensitize cancer cells to cytotoxic drugs, reduce tumor growth, doi: 10.1158/1541-7786.MCR-16-0481 and metastatic burden (10, 14), thus making CXCR4 an attrac- 2018 American Association for Cancer Research. tive target for early- and late-stage disease management.

www.aacrjournals.org OF1

Scarlett et al.

The cannabinoids system consists of lipophilic cannabi- and functionally migrating through an endothelial cell layer in noids (endocannabinoids and exocannabinoids), cannabi- a RhoA-dependent manner. Altogether, our results further noid receptors (CB1 or CB2) and their enzymes for synthesis support agonist-induced CXCR4/CB2 heterodimerization and degradation (15–18). Delta-9-tetrahydrocannabinol as an emerging approach to inhibiting the metastatic spread (D9-THC), the major psychoactive and immunomodulatory of cancer. component of marijuana, is the most notable cannabinoid (18, 19), and clinically, is used for its antipalliative effects. Materials and Methods However, emergent data demonstrates that cannabinoids possess antiproliferative capabilities and regulate key cell Cell lines and reagents signaling pathways that promote cell survival, invasion, Human metastatic prostate cancer cell line (PC3), human angiogenesis, and metastasis in various cancer models, pro- metastatic breast cancer cell line (MDA-MB-231), and human pelling cannabinoids to the forefront of cancer therapeutics embryonic kidney cells (HEK 293T) were purchased from ATCC. research (3, 20). PC3 and MDA-MB-231 cells were grown in RPMI1640 and Allosteric interactions within chemokine receptor heterodi- HEK239T cells were grown in DMEM-F-12; all lines were supple- mers inhibit receptor activation in some instances, while other mented with 10% FBS, 1% nonessential amino acids, and 1% fi receptor pairings cause enhanced signaling and trigger entirely antibiotic/antimycotic in a humidi ed incubator (5% CO2)at new responses downstream of receptor activation (3, 21, 22). For 37 C. Human umbilical vein endothelial cells (HUVEC) were example, Decaillot and colleagues demonstrated that CXC che- purchased from ATCC and maintained in endothelial growth mokine receptors CXCR4 and CXCR7 form heterodimers that medium-2 [Lonza; 2% FBS, human endothelial growth factor enhance cell migration in response to CXCL12 through recruit- (hEGF), hydrocortisone, GA-1000 (gentamicin, amphotericin-B), ment of b-arrestin to the heterodimeric complex. In addition, human fibroblast growth factor-B (hFGF-B), R3-IGF-1, ascorbic fi extensive evidence points to heterodimerization between chemo- acid, and VEGF)] in a humidi ed incubator (5% CO2)at37C. kine receptors with nonchemokine GPCRs or non-GPCR recep- Cell lines were authenticated using a short tandem repeat (STR) tors (21, 22). In cancer, we (3) and our colleagues (5, 6) dem- DNA profiling by ATTC at time of purchase, and every 6 months onstrated that CXCR4-mediated functions were abrogated by thereafter. Mycoplasma was monitored using the MycoAlert simultaneous, agonist-induced heterodimerization of CXCR4 Detection Kit (Lonza). Human CXCR4 agonist, CXCL12 with GPCRs, such as we have demonstrated with CB2 (3), where- (100ng/mL working concentration), was purchased from Pepro- by heterodimers reduced cellular migration and the metastatic Tech, Inc. Human AM1241 ligand (CB2 agonist; 1 mmol/L) was potential of tumor cells. However, the underlying signaling purchased from Cayman Chemicals. The CXCR4 antagonist, mechanisms governing the observed reduction in cellular migra- AMD3100 (1 mg/mL), and Rho-associated kinase inhibitor tion were not elucidated. (ROCK), Y-27631 (10 mmol/L) were purchased from Sigma- The Rho family of GTPases are small GTP-binding proteins Aldrich. Human CXCR4-siRNA and control-siRNA (60 nmol/L) that function as molecular switches controlling a wide spec- was from Santa Cruz Biotechnology. Prior to treatment with trum of signal transduction pathways in eukaryotic cells agonists and antagonists, cells were serum-starved for 24 hours (23–26). They regulate actin cytoskeleton dynamics, and mod- in media only (0% FBS, 0% nonessential amino acids, and 0% ulate cell polarity, microtubule dynamics, membrane transport antibiotic/antimycotic) for 24 hours in 5% CO2 at 37 C. Samples pathways, and transcriptional activity (8, 21, 23–25). Expres- denoted as "control" or "untreated" received fresh media supple- sion of the GTPase, RhoA, is implicated in increased tumor mented with corresponding vehicles (e.g., 0.1%PBS/BSA, DMSO, aggressiveness in multiple cancer types, and loss of RhoA etc.) for each agonist or antagonist. prevents proliferation and cellular migration both in vitro and metastasis in vivo (23). Furthermore, the CXCL12/CXCR4 sig- RhoA activity assay naling axis stimulates RhoA activation through Ga13, leading Activated RhoA was detected using a RhoA Activity Assay Kit to phosphorylation of myosin light chain (MLC) by Rho- (Cell Biolabs, Inc.) as per the manufacturer's instructions. Briefly, 6 associated protein kinase (ROCK), which is required for direc- 1.5 10 cells were cultured to 80% confluence then serum- tional cell migration, the promotion of invasive protrusions, starved for 24 hours. Cells were pretreated with antagonists for and remodeling of the extracellular matrix (26–28). 1 hour prior to stimulation with various agonists for 1 minute We tested whether a physical heterodimeric association 5% CO2 at 37 C. Cells were harvested as described by the kit, and between CXCR4 and CB2 results in decreased CXCR4-medi- 1 mg of protein from each sample was immunoprecipitated ated cell migration via reduction of the Ga13–RhoA signaling with Rhotekin-RBD beads at 4 C overnight with gentle agitation. axis. First, we analyzed RhoA activation upon heterodimeriza- The beads were pelleted, washed thrice, and boiled in Laemmeli tion, and expression of downstream targets of RhoA signaling. sample buffer for 5 minutes. Samples were separated by 15% Second, we investigated alterations to cellular invasion, adhe- SDS-PAGE, transferred to a polyvinylidene difluoride (PVDF) sion, and morphology as a result of reduced RhoA expression membrane, and blocked in 3% BSA in 1 TBST. Rho-A was upon CXCR4/CB2 heterodimerization. Finally, to correlate the detected by anti-RhoA antibody (1:100) in 3% BSA/TBST over- relationship between the effects of CXCR4 heterodimerization, night at 4 C and harvested for Western blot analysis. RhoA activity reduced RhoA expression, and the migratory ability of cancer was normalized to total RhoA protein expression via Western cells, we probed for expression of metastatic markers. We blot analysis. observed that CXCR4/CB2 heterodimerization antagonized expression of Ga13, followed by reduced activity of RhoA and siRNA transfection further downstream targets. The resultant signaling blockade Transient transfection of CXCR4-specific human siRNA (Santa interfered with cancer cells developing a metastatic phenotypic Cruz Biotechnology; sc-35421) was performed on PC3 cells using

OF2 Mol Cancer Res; 2018 Molecular Cancer Research

Inhibition of Ga13/RhoA-mediated Migration

Lipofectamine 2000 (Invitrogen). Cells (1 106) were plated in Filamentous (F)-actin immunocytochemistry 10% FBS/RPMI in 100-mm culture dishes and then transfected Cells were seeded on glass coverslips at a density of 2 105 with 60 mmol/L CXCR4-siRNA or scramble/control-siRNA (Santa cells/well and incubated for 72 hours at 5% CO2 at 37 C Cruz Biotechnology) in Opti-MEM at 37 C/5% CO2 for 12 hours. until they were approximately 80% confluent. Cells were then Cells recovered in 10% FBS/RPMI for 24 hours, starved for serum starved for 24 hours, then pretreated with 1 mg/mL of 24 hours, followed by treatment with 100 ng/mL CXCL12 and/ AMD3100 or 10 mmol/L of Y-27632 for 1 hour prior to or 1 mmol/L AM1241, and then were harvested for immunopre- treatment with 100 ng/mL CXCL12 or 1 mmol/L AM1241 each cipitation and/or Western blot analysis. alone, or combined CXCL12/AM1241 at 5% CO2 and 37 C in serum-free media for 90 minutes. Coverslips were rinsed Immunoprecipitation twice with ice-cold 1 PBS, fixed with 4% paraformaldehyde A total of 1.5 106 cells per sample were serum starved for for 10 minutes, and washed again with 1 PBS. Cells were 24 hours prior to pretreatment with 1 mg/mL AMD3100 or blocked with 1% BSA/TBST and incubated with FITC-conju- 10 mmol/L Y-27632, followed by treatment with 100 ng/mL gated phalloidin (100 nmol/L in 1 PBS, Enzo Life Science) CXCL12, 1 mmol/L AM1241, or CXCL12 and AM1241 simulta- to detect F-actin for 1 hour at room temperature with gentle neously for 1 (PC3 and HEK 293) or 10 (MDA-MB-231) minutes. agitation. Cells were then washed thrice with 1 PBS and Cells were lysed in Rho A lysis buffer from above, sonicated, then counterstained with DAPI for nuclear visualization. Slides were followed by incubation on ice for 30 minutes. Lysates were visualized and images were captured using a Zeiss LSM 700 centrifuged at maximum speed for 10 minutes at 4 C, then confocal microscope. Scale bar ¼ 50 mm. 500–750 mg of protein from each sample was incubated with Ga13 antibody (1:100; Abcam) overnight at 4C on a rocking Cell migration platform. Protein A/G PLUS Agarose Beads (1:10, Santa Cruz MDA-MB-231 cells (1 106 per sample) were plated in p-100 Biotechnology) were added to each sample for 2 hours at 4C tissue culture plates and serum-starved for 24 hours. Cells were with gentle agitation. Cell lysates were washed twice in 1 PBS detached with 1 Citric Saline, washed in 1 PBS, pelleted by (maximum speed for 5 minutes at 4C), boiled in centrifugation at 3,000 g for 5 minutes, and then counted to Laemelli sample buffer for 5 minutes, followed by 12% prepare 50,000 cells per sample in 150 mL of media. Transwell SDS-PAGE. Protein was transferred to a PVDF membrane, blocked inserts were prepared by precoating the outer membrane with in 5% milk/TBST, immunoblotted in 5% BSA/TBST with an 50 mL rat tail collagen (50 mg/mL in 0.02N acetic acid; BD anti-RhoA (1:1,000; Cell Signaling Technology). Primary Biosciences) for cell adherence to the transwell insert, and incu- antibodies were detected with corresponding horseradish bating at 4C overnight. The next day, the inner membrane was peroxidase (HRP)-conjugated secondary antibodies (1:1,000; coated as above, and incubated for 1 hour at room temperature. Jackson) and enhanced chemiluminescence (Luminata; Thermo Inserts were washed thrice in 1 PBS prior to use for assay. Fisher Scientific). Treatment (100 ng/mL CXCL12 and 1 mmol/L AM1241 each or in combination, 1 mg/mL AMD3100 or 10 mmol/L Y-27632) were Antibodies and Western blotting calculated for a final concentration of 500 mL each, but prepared PDZ-RhoGEF and LKB1 antibodies were purchased from in 350 mL per sample, and placed into the bottom well of each Santa Cruz Biotechnology. F-actin was purchased from Abcam. chamber. Fifty-thousand cells per sample were resuspended in Phosphorylated myosin light chain (pMLC), integrin aV, and 150 mL of media and were seeded into the transwell insert of integrin b3 were purchased from Cell Signaling Technology. A 24-well transmigration chambers (8-mm pore; BD Falcon), bring- total of 2.0 106 cells per sample were serum starved for ing the total volume of each chamber to 500 mL. To allow for 24 hours prior to pretreatment with 1 mg/mL AMD3100 or migration, plates were incubated in 5% CO2 and 37 C for 6 hours, 10 mmol/L Y-27632 followed by treatment with 100 ng/mL and the inner-top chambers were cleaned with a cotton swab. CXCL12, 1 mmol/L AM1241, or CXCL12 and AM1241 simul- Cells attached to the outer-top chambers were fixed with 10% taneously for various time points. Cells were lysed and soni- formalin for 15 minutes at room temperature, washed thrice in cated in 1 Cell Signaling Technology lysis buffer prior to 1 PBS, stained with 0.05% crystal violet for 15 minutes, then incubation on ice for 30 minutes. Lysates were centrifuged at washed 6 times in water. Five representative fields of each insert maximum speed for 10 minutes at 4C, and then equal were counted under a light microscope, and the migration index amounts of protein per sample were separated by SDS-PAGE was calculated and graphed as the x-fold change in migration and transferred to PVDF membrane. Protein-bound mem- observed over control cells or CXCL12-treated cells. Experiments branes were blocked in 3% BSA/1 TBST, (1 TBST for pMLC), were performed at least thrice, and significant significance was and subsequently incubated with primary antibodies: RhoA analyzed via GraphPad Prism ( , P < 0.05; , P < 0.001). (1:500, Cell Signaling Technology); F-actin (1:1,000, Abcam); pMLC (1:500, Cell Signaling Technology); LKB1 (1:100, Santa Transendothelial migration assay Cruz Biotechnology); integrin a5 (1:1,000, Cell Signaling Migration though an endothelial cell barrier was performed Technology); integrin b3 (1:1,000, Cell Signaling Technology); via a Transwell Migration Assay Kit (Cell Biolabs, Inc.) per the total AKT (1:1,000, Cell Signaling Technology); or total ERK1/2 manufacturer's instructions. Briefly, 1 105 HUVECS per well (1:1,000, Cell Signaling Technology) overnight at 4Cin3% were seeded inside of Transwell inserts in 500-mLcomplete BSA/TBST; (1 TBST alone for pMLC). Primary antibodies were EGM-2 media at 5% CO2 and 37 C for 72 hours to allow detected by HRP-conjugated secondary antibodies at 1:5000– formation of a cell monolayer. In the interim, 1 106 of PC3, 1:10,000 diluted in 3% BSA/1 TBST (1 TBST alone for MDA-MB-231, or HEK293T cells were preincubated with pMLC) and enhanced chemiluminescence (SuperSignal West 1 CytoTracker (kit provided) allowing each cell to become Pico Chemiluminescent Substrate or Lumniglo). fluorescently labeled, then were pelleted and washed 3 in

www.aacrjournals.org Mol Cancer Res; 2018 OF3

Scarlett et al.

serum-free media and resuspended in 500 mLofserum-free Statistical analysis media. EGM-2 media was removed from inside each HUVEC Student t test, two-way ANOVA, or Bonferroni multiple transwell and replaced with 300 mLeachofPC3,MDA-MB- comparison test were performed to determine statistically signi- 231, or HEK293T serum-free RPMI cell suspension. Transwell ficant differences between groups. Statistical analysis was per- inserts were transferred to fresh plates containing serum-free formed via GraphPad Prism. media supplemented with combinations of 100 ng/mL CXCL12, 1 mmol/L of AM1241, 1 mg/mL of AMD3100, and/or Results 10 mmol/L Y-27632 and the plate was incubated in 5% CO2 at 37C for 24 hours. Fluorescently labeled cells that transmi- Simultaneous treatment with CXCL12 and AM1241 inhibits grated were lysed in 1 Cell Lysis Buffer (Cell Biolabs, Inc.) and Ga13/PRG–mediated activation of RhoA analyzed using a plate reader at OD 480 nm/520 nm. Exper- We have shown formation of a physical heterodimer iment was repeated thrice, and statistical significance was between CXCR4 and CB2 upon simultaneous (combined) analyzed via GraphPad Prism (, P < 0.05; , P < 0.001). treatment with agonists CXCL12 and AM1241, respectively (3). This heterodimer between CXCR4 and CB2 impeded cellular migration; however, a mechanism was not defined. Wound healing assay Independent studies have demonstrated that CXCR4-mediated A total of 1 106 cells per sample were seeded in a 6-well migration occurs via the RhoA–PRG–Ga13 pathway (27, 28). culture plate overnight, and then serum-starved for 24 hours. Therefore, we first investigated whether heterodimerization Cells were pretreated with 1 mg/mL of AMD3100 or 10 mmol/L of CXCR4/CB2 decreased RhoA activation in the malignant of Y-27632 for 1 hour, removed, and then plates were stored on human prostate cancer cells (PC3). Isolating active forms ice. One milliliter of 1 PBSwasaddedtoeachwell,avertical (GTP-bound) of RhoA with Rhotekin-glutathione beads wound was made across each well using a micropipette tip, and demonstrated that CXCL12 caused an increase in expression nonadherent cells were removed by washing with 1 PBS. of RhoA-GTP compared with control (untreated) cells (Fig. 1A). Serum-free media with supplements of 100 ng/mL CXCL12, However, simultaneous treatment with CXCL12 and AM1241, and 1mmol/L of AM1241 each or in combination, along with which induced the heterodimer (3), resulted in decreased m m 1 g/mL AMD3100, or 10 mol/L Y-27632. Wound closure was protein expression of RhoA-GTP, indicating a decrease in acti- allowed for up to 6 hours at 5% CO and 37C, with pictures 2 vation of RhoA compared with CXCL12-stimulated cells captured at 0, 2, 4, and 6 hours. The wound area of each sample (Fig. 1A). The decrease in activated Rho-A expression levels was quantified using ImageJ software. Experiments were repeat- was consistent in cells pretreated with AMD3100 (CXCR4 fi ed thrice; statistical signi cance was analyzed via GraphPad antagonist) or Y-27632 (RhoA/ROCK inhibitor) prior to treat- Software (, P < 0.05; , P < 0.001). ment with CXCL12 (Fig. 1A). Tan and colleagues demonstrated that CXCR4-mediated acti- Extracellular matrix adhesion assay vation of RhoA, and resulting directional cellular migration, was The adhesion of cells to components of the extracellular specifically mediated by the small G protein Ga13 (27). Consis- matrix (ECM) was quantified using the CytoSelect Adhesion tent with the pattern for RhoA activity, we observed that CXCL12 Assay (ECM Array, Colorimetric Format; Cell Biolabs, Inc.). A led to increased expression Ga13, as determined by expression of total of 1.5 106 cells per sample were seeded and grown to RhoA via Ga13/RhoA immune complexes, compared with approximately 80% confluence. Cells were serum-starved for 24 untreated cells in PC3 and the human breast cancer cell line, hours and then harvested by 0.25% trypsin/EDTA for 5 min- MDA-MB-231 (Fig. 1B and C). Conversely, simultaneous stimu- utes. Trypsin was removed after centrifugation (2,000 g for 5 lation of PC3 or MDA-MB-231 cells with CXCL12 and AM1241 minutes) and cells were resuspended in serum-free RPMI at a resulted in decreased expression of Ga13, which was consistent concentration of 1.0 106 cells/mL. The cell suspension with expression levels of Ga13 in cells treated with AMD3100 or was preincubated with 1 mg/mL AMD3100 or 10 mmol/L Y-27632 (Fig. 1B and C). Y-27632 in serum-free media for 1 hour in 5% CO2 at 37 C. To support our premise that a dimer between CXCR4 and Thereafter, additional treatments of 100 ng/mL of CXCL12 CB2 inhibits the action of CXCR4 to activate Ga13-RhoA, we alone, 1 mmol/L of AM1241 each, or combination was added used the human embryonic kidney cell line, HEK 293T, as we to the cell suspension. A total of 150 mLofeachcellsuspension (3) and others have published is CXCR4-null. Pretreatment was plated onto a 48-well culture plate precoated with various with the RhoA/ROCK inhibitor Y-27632 visibly reduced ECM components (fibronectin, collagen I, and laminin I), and Ga13/RhoA immune complexes. However, RhoA protein levels then incubated for 2 hours in 5% CO2 at 37 Ctoallow displayed consistent expression among all other treatment attachment. Cells were washed carefully with 1 PBS to remove groups, with no visible induction by CXCL12 (Fig. 1D); the unbound cells, and adherent cells were incubated in 200 mLof basal level of Ga13–RhoA immune complexes could have been CellStainSolutionfor10minutesonanorbitalshaker(Cell mediated by any other GPCR. Likewise, we used siRNA target- Biolabs, Inc). Thereafter, each well was gently washed 4 with ing CXCR4 to further show that the dimer between CXCR4 and 500 mL of double deionized water, followed by incubation in CB2 has similar antagonistic effects as would a knockdown of 200 mL Extraction Solution (Cell Biolabs, Inc.) for 10 minutes the CXCR4 receptor (Fig. 1E). In PC3 cells, we observed on an orbital shaker. Following extraction, 150 mLofeach reduction of Ga13/RhoA immune complexes in CXCL12-trea- sample was transferred to a 96-well plate and measured using ted cells transfected with CXCR4-siRNA compared with cells a spectrophotometer at OD 560 nm. Each experiment was without siRNA. However, the reduction in Ga13/Rho A repeated thrice and statistical significance via GraphPad Prism immune complexes in cells was more robust in cells treated (, P < 0.05; , P < 0.001). CXCL12 and AM1241 (both with and without siRNA),

OF4 Mol Cancer Res; 2018 Molecular Cancer Research

Inhibition of Ga13/RhoA-mediated Migration

A PC3 B PC3 C MDA-MB-231 CXCL12 -+-+++ CXCL12 -+-+++ CXCL12 - -++ ++

AM1241 --++-- AM1241 --++-- AM1241 - + + -- -

Y-27632 - - - - + - Y-27632 - - - - + - Y-27632 ----+ -

AMD3100 - - - - - + AMD3100 - - - - - + AMD3100 -----+ RhoA-GTP (Eluent) (Eluent) Gα13/RhoA IP Gα13/RhoA IP

Total RhoA Total Gα13 prior to IP Total Gα13 prior to IP

D E F HEK 293T PC3 PC3

CXCL12 - + - + + + CXCL12 - + +

AM1241 - - + + - - AM1241 --+ CXCL12 - + - + + +

Gα13/RhoA IP AM1241 Y-27632 - - - - + - (Eluent) CXCR4 knockdown - - + + - -

Total ERK1/2 Control siRNA Y-27632 AMD3100 - - - - - + prior to IP - + - - - - - + -

(Eluent) CXCR4 siRNA AMD3100 Gα13/RhoA IP --+ - - - - - + +CXCR4 siRNA CXCR4 ~45Kd Total Gα13 prior to IP PDZ CXCL12 - + + RhoGEF Total AKT α-Tubulin AM1241 --+ Gα13/RhoA IP (Eluent)

Total ERK1/2 prior to IP

Figure 1. CXCR4/CB2 heterodimerization attenuates Ga13/PRG–mediated RhoA activation in PC3 cells. A, Serum-starved PC3 cells were pretreated for 1 hour with Y-27632 or AMD3100 followed by treatment with CXCL12 for 1 minute. Alternatively, PC3 cells were treated for 1 minute with CXCL12 or AM1241 alone, or CXCL12/AM1241 simultaneously. Lysates were incubated with Rhotekin-RBD beads prior to harvesting for immunoblotting for RhoA-GTP protein expression; total RhoA represents lysate prior to pull-down with Rhotekin-RBD beads. PC3 cells (B), MDA-MB-231 cells (C), and HEK 293T cells (D) were treated as described in Materials and Methods, and then the lysates were incubated with anti-Ga13 antibody overnight at 4C followed by standard immunoprecipitation techniques. Isolated protein lysates were immunoblotted for RhoA protein expression; input represents lysate priorto immunoprecipitation. E, PC3 were transfected with 60 mmol/L human siRNA targeting CXCR4 or nonspecific control siRNA for 12 hours, followed by recovery in 10% FBS/RPMI for 24 hours. Cells were then harvested for Ga13/RhoA immunoprecipitation or Western blotting. Total ERK1/2 and Total AKT were used as: (i) loading input controls to demonstrate that siRNA was specific; and (ii) to demonstrate that equal amounts of protein were used for assay. F, Serum-starved PC3 cells were treated as described above, and whole protein lysates were immunoblotted for PRG protein expression; a-tubulin was used as a loading control. Each experiment was performed at least twice. , P < 0.05; , P < 0.001. suggesting that the heterodimer is an efficient way to manage Agonist-induced CXCR4/CB2 heterodimerization inhibits CXCR4 signaling (Fig. 1E). RhoA-mediated cytoskeleton reorganization and attenuates In breast cancer cells stimulated with CXCL12, Struckhoff RhoA-dependent cell migration and colleagues demonstrated that PDZ-RhoGEF (PRG) RhoA plays a critical role in cell migration through its guanine nucleotide exchange factor was required for RhoA- intimate relationship with the actin cytoskeleton, which has dependent cell migration (28). We, too, observed that CXCL12 long been implicated in generating the contractile forces increased PRG protein expression levels in PC3 cells (Fig. 1C). needed for membrane blebbing, lamellae formation, and Contrastingly, when PC3 cells were treated with CXCL12 membrane ruffling at the leading edge and retraction of the and AM1241 simultaneously, expression levels of PRG pro- trailing edge of migratory cells (29–32). To evaluate whether tein expression decreased compared with that of CXCL12 our induced heterodimer could inhibit RhoA-mediated treatments, further supporting that agonist-induced heterodi- changes in the actin cytoskeleton that are necessary for migra- merization is an effective mechanism to antagonize CXCR4- tion, we first examined phenotypic changes in filamentous mediated signaling that leads to cell migration. As a control, actin (F-actin) in response to independent (CXCL12) and PC3 cells treated with AMD3100 or Y-27632 virtually abol- combinatory (CXCL12/AM1241) agonist stimulation in PC3 ished PRG protein expression in the presence of CXCL12 and MDA-MB-231 cell lines. CXCL12 produced expected pro- (Fig. 1F), indicating that the observed results were specifically migratory alterations in cell morphology such as membrane mediated by CXCR4–RhoA signaling. Collectively, coinciding ruffling, lamellae formation, and membrane protrusions in with our previously observed decrease in cellular migration cancer cell lines compared with untreated cells that remained (3), an induced GPCR heterodimer between CXCR4 and CB2 characteristically smooth at the cell membrane (Fig. 2 and 3). antagonizes CXCR4-mediated cell migration via the RhoA– However, cells simultaneously stimulated with CXCL12 Ga13–PRG pathway. and AM1241 also exhibited a smooth cell membrane, similar

www.aacrjournals.org Mol Cancer Res; 2018 OF5

Scarlett et al.

PC3 Untreated CXCL12 AM1241

Figure 2. Agonist-induced CXCR4/CB2 heterodimerization inhibits RhoA-mediated cytoskeleton reorganization. Serum-starved PC3 cells were pretreated for 1 hour with Membrane ruffling Y-27632 or AMD3100 prior to treatment with CXCL12 for 90 minutes. 50 μm 50 μm 50 μm Subsequently, cells were treated for CXCL12 + AM1241 Y-27632 + CXCL12 AMD3100 + CXCL12 90 minutes with CXCL12 or AM1241 alone, or CXCL12/AM1241 simultaneously. Samples were fixed and stained with FITC-phalloidin to visualize F-actin; nuclei were counterstained with DAPI. Arrows highlight areas of lamellae formation and membrane ruffling.

50 μm 50 μm 50 μm

to untreated cells, and indicated little-to-no changes in cyto- pretreatment with Y-27632 resulted in apoptotic phenotypes skeletal rearrangement; this same phenotype was observed for with an overall decrease in volume of cell cytoplasm and cells pretreated with AMD3100 (Fig. 2 and 3). Contrastingly, membrane blebbing in PC3 and MDA-MB-231 cells (Fig. 2

MDA-MB-231 CXCL12

Untreated

Membrane ruffling

50 μm

50 μm Membrane ruffling AM1241 CXCL12 + AM1241 Y-27632 + CXCL12 AMD3100 + CXCL12

50 μm 50 μm 50 μm 50 μm

Figure 3. Agonist-induced CXCR4/CB2 heterodimerization inhibits RhoA-mediated cytoskeleton reorganization. Serum-starved MDA-MB-231 cells were pretreated for1hourwithY-27632orAMD3100priortotreatmentwithCXCL12for90minutes. Subsequently, cells were treated for 90 minutes with CXCL12 or AM1241 alone, or CXCL12/AM1241 simultaneously. Samples were ﬁxed and stained with FITC-phalloidin to visualize F-actin; nuclei were counterstained with DAPI. Arrows highlight areas of lamellae formation and membrane rufﬂing.

OF6 Mol Cancer Res; 2018 Molecular Cancer Research

Inhibition of Ga13/RhoA-mediated Migration

HEK 293T A Untreated CXCL12 AM1241

50 μm 50 μm 50 μm

CXCL12 + AM1241 Y-27632 + CXCL12 AMD3100 + CXCL12

50 μm 50 μm 50 μm B C D --++++ --++++ CXCL12 CXCL12 CXCL12 --++++

AM1241 - -++-- AM1241 --++ -- AM1241 --++------+- ---- +- Y-27632 Y-27632 Y-27632 ---- +-

AMD3100 ------+ AMD3100 ------+ AMD3100 ------+ Phosphorylated F-Actin MLC LKB1

a-Tubulin a-Tubulin a-Tubulin

Figure 4. Agonist-induced CXCR4/CB2 heterodimerization inhibits RhoA-mediated cytoskeleton reorganization. A, Serum-starved HEK 293T cells were pretreated for1hourwithY-27632orAMD3100priortotreatmentwithCXCL12for90minutes. Subsequently, cells were treated for 90 minutes with CXCL12 or AM1241 alone, or CXCL12/AM1241 simultaneously. Samples were fixed and stained with FITC-phalloidin to visualize F-actin; nuclei were counterstained with DAPI. Arrows highlight areas of lamellae formation and membrane ruffling. B–D, Serum-starved PC3 cells were pretreated with Y-27632 or AMD3100 followed by treatment with CXCL12, AM1241, or agonists simultaneously for 90 minutes. Whole protein lysates were analyzed for F-actin (B), pMLC (C), or LKB1 (D). Each experiment was performed at least twice. and 3), which is confirmed in literature (33, 34). In HEK 293T Failed directional migration is usually a consequence of loss cells, as expected, we did not observe phenotypic changes of cell polarity as marked by decreased expression of human in response to CXCL12, AM1241, or the agonist combined tumorsuppressorandserine-threonine kinase LKB1 (29). (Fig. 4A) as they are CXCR4-null. Therefore, we harvested treated samples and immunoblotted We examined protein expression levels of F-actin to support the for LKB1, an established regulator of RhoA-dependent cell phenotypic changes observed above where expression of F-actin polarity (30, 31). Simultaneous stimulation of PC3 cells with in PC3 cells, simultaneously stimulated with CXCL12 and CXCL12 and AM1241 decreased protein expression levels of AM1241, was comparable with untreated cells, further supporting LKB1 compared with untreated control and CXCL12, as did that the CXCR4/CB2 heterodimer prevents the promigratory treatment with Y-27632 and AMD3100 (Fig. 4D), indicative of phenotype necessary for cellular migration (Fig. 4B). Likewise, reduced polarity for migration. AMD3100 and Y-27632 displayed markedly reduced F-actin To attribute the reduction in RhoA-mediated signaling, as a protein levels (Fig. 4B). RhoA regulates actomyosin contraction result of heterodimerization, to migratory behavior and events in part by stimulating phosphorylation of myosin light chain (p- (Fig. 5A and B), we observed that CXCL12 significantly stim- MLC; refs. 27, 28, 35) which is also required for phenotypic ulated wound closure of PC3 cells compared with untreated changes associated with cellular migration. Therefore, we deter- cells (Fig. 5B, P ¼ 0.0462). CXCL12 and AM1241 combined mined whether agonist-induced CXCR4/CB2 heterodimerization were deficient in stimulating migration across the wound area led to downstream attenuation of p-MLC expression in PC3 cells. when compared with CXCL12 alone (Fig. 5A and B; P ¼ 0.042). CXCL12 increased expression of p-MLC protein levels compared Likewise, pretreatment of cells with AMD3100 (P ¼ 0.0099) or with the untreated control; however, protein levels of p-MLC in Y-27632 decreased wound closure efficiency compared with simultaneously treated samples were reduced compared with CXCL12, lending more credence that GPCR heterodimerization both untreated control and CXCL12-treated cells (Fig. 4C). Taken attenuates RhoA signaling associated with cancer-progressing together, these results further support that an induced CXCR4/ activities. CB2 heterodimer effectively attenuates signaling pathways that Via simple Transwell migration, CXCL12-treated MDA-MB- leads to a migratory phenotype in cancer. 231 cells exhibited significantly increased migration compared

www.aacrjournals.org Mol Cancer Res; 2018 OF7

Scarlett et al.

A 0 hours 6 hours 0 hours 6 hours B

CXCL12 + Untreated AM1241

AMD3100 + CXCL12 CXCL12

AM1241 Y-27632 + CXCL12

CDE Untreated CXCL12 AM1241

400

300 100,000

200 80,000

CXCL12 + AM1241 Y-27632 + CXCL12 AMD3100 + CXCL12 60,000 100 40,000 0 Average number of migrated cells 20,000

0 Fluoresence 480 nm/520 nm (RFU)

Treatment

Figure 5. CXCR4/CB2 heterodimerization attenuates RhoA-dependent cell migration and wound closure. A and B, Serum-starved PC3 cells were pretreated for 1 hour with Y-27632 or AMD3100 followed by treatment with CXCL12 and/or AM1241 for up to 6 hours as described in Materials and Methods. ImageJ was used to measure the width of each wound at 0 and 6 hours. Experiments were performed thrice. ,P< 0.05; , P < 0.001. C and D, 5 104 cells each were seeded into the top Transwell chamber and allowed to migrate toward combinations of agonists and chemical inhibitors in the bottom chamber for 6 hours at 37 C, 5% CO2.Fivefields of each Transwell insert were randomly selected and counted for migrated cells at 10 magnification using a Zeiss Axiovert 200M light microscope, and results were analyzed via GraphPad Prism. Experiments were repeated thrice, and data represent the average of three independent experiments. D, A graphical representation of total migrated cells. Data represents the average of three independent experiments; , P < 0.05; , P < 0.001. E, HUVECs were seeded in the top chamber of Transwell inserts allowing for the formation of a confluent monolayer. Fluorescently labeled, serum-starved PC3 cells were added on top of HUVECs in the top chamber, and then, the combinations of agonists and chemical inhibitors were added to the bottom chambers followed by a 24-hour incubation. Fluorescent invasive cells were quantified at OD 480 nm/520 nm. Experiments were performed thrice (, P < 0.05; , P < 0.001).

with untreated migrating cells (P ¼ 0.043; Fig. 5C and D). both Y-27632 (P ¼ 0.0054) and AMD3100 (P ¼ 0.0187) Likewise, CXCL12 and AM1241 combined exhibited signifi- significantly reduced transmigration when compared with cells cantly decreased migratory capacity compared with CXCL12 migrating toward CXCL12 alone (Fig. 5E). alone (P < 0.0001; Fig. 5C and D). In addition, both Y-27632 (P < 0.0001) and AMD3100 (P < 0.0001) significantly Inhibition of CXCR4-mediated signaling via reducedtransmigrationwhencomparedwithcellsmigrating heterodimerization results in decreased integrin expression toward CXCL12 alone (Fig. 5C and D). To effectively metas- and RhoA-mediated adhesion to the extracellular matrix tasize from a primary tumor site to a distal organs, aggressive Rho GTPase molecules regulate cell adhesion (36, 37) further cancer cells must demolish and migrate through an endothelial implicating its role in tumor cell invasion and migration cell barrier, which is characteristic of intravasation and extrav- (26). Multiple studies report that CXCR4 signaling leads to asation in vivo (25, 32), and CXCL12-driven transendothelial enhanced expression of a5 integrin and b3 integrin, which migration of metastatic tumor cells has been shown to be Rho- trigger persistent adhesion of tumor cells to immobilized dependent (20, 33, 34). As such, we observed that CXCL12 laminin, collagen, and fibronectin (38). In PC3 cells, CXCL12 resulted in significantly increased transmigration compared alone stimulated increased expression levels of a5integrin with untreated transmigrating cells (P ¼ 0.0296; Fig. 5E). compared with the untreated control, while AM1241 in com- Likewise, CXCL12 and AM1241 combined exhibited signifi- bination with CXCL12 reduced expression (Fig. 6A). Although cantly decreased transendothelial migratory capacity com- we observed an increase in b3expressionuponCXCL12 pared with CXCL12 alone (P ¼ 0.0476; Fig. 5E). In addition, treatment, which aligns with observations in literature,

OF8 Mol Cancer Res; 2018 Molecular Cancer Research

Inhibition of Ga13/RhoA-mediated Migration

Figure 6. Inhibition of CXCR4-mediated signaling via heterodimerization results in decreased integrin expression and RhoA-mediated adhesion to the extracellular matrix. A, Serum-starved PC3 cells were pretreated for 1 hour with Y-27632 or AMD3100 followed by treatment with CXCL12 or AM1241 alone, or CXCL12/AM1241 simultaneously for 6 hours. Cell lysates were harvested for integrin a5 and integrin b3 protein expression; a-tubulin was used as a loading control. Experiment was performed at least twice. B–D, Cells were serum-starved and pretreated Y-27632 and AMD3100 for 1 hour prior to the addition of CXCL12 or AM1241 alone, or CXCL12/AM1241 simultaneously on matrix-coated plates for 2 hours, stained with Cell Stain Solution (Cell Biolabs, Inc.), and then analyzed at OD 560 nm. Each assay was performed at least thrice. , P < 0.05; ,P< 0.001. heterodimerization of CXCR4 with CB2 caused no significantly 0.0317) compared with CXCL12 alone; likewise, Y-27632 observable reduction of b3 protein levels when compared with (fibronectin; P ¼ 0.0154), (collagen I; P ¼ 0.0031), and CXCL12; chemical inhibition with AMD3100 and Y-27632 (laminin I; P ¼ 0.005), and AMD3100 (fibronectin; P ¼ resulted in reduced levels of both a5 integrin and b3 integrin 0.0322), (collagen I; P ¼ 0.0086), and (laminin I; P ¼ (Fig. 6A). 0.0264), both yielded reduced adhesion of cells to ECM com- Finally, we confirmed whether CXCR4/CB2 heterodimeriza- ponents (Fig. 6B–D). These results suggest that a physical tion regulates adherence to components of the extracellular CXCR4/CB2 heterodimer is sufficient to decrease tumor cell matrix (ECM). Using in vitro models of ECM components, adhesion, likely due to decreased expression of integrin alpha we observed that CXCL12 led to a significant increase in and subsequent inhibition of RhoA activation. adhesion to fibronectin (P ¼ 0.0484), collagen I (P ¼ 0.0287) and laminin I (P < 0.0001) in PC3 cells when compared with untreated cells (Fig. 6B–D). CXCL12 and AM1241 Discussion combined significantly reduced cellular adhesion to fibronectin Our results provide mechanistic insight into our previous (P ¼ 0.0171), collagen I (P ¼ 0.0354), and laminin I (P ¼ observation that a physical CXCR4/CB2 heterodimer reduces

www.aacrjournals.org Mol Cancer Res; 2018 OF9

Scarlett et al.

Figure 7. Simultaneous agonist-induced CXCR4/CB2 heterodimerization inhibits RhoA-mediated cell migration and adhesion. A, The binding of CXCL12 to its cognate receptor, CXCR4, initiates the activation of Ga13 and its dissociation from Gbg subunits. Ga13 directly activates the DBl family RhoGEF, PDZ-RhoGEF (PRG), which subsequently activates the small GTPase, RhoA. The downstream target of RhoA, Rho-associated protein kinase (ROCK), phosphorylates myosin light chain (MLC) leading to directional cell migration, the formation of invasive protrusions including membrane ruffling, and increased adhesion to the extracellular matrix. Integrin signaling through an alternate pathway can modulate RhoA activity through activation of focal adhesion kinase1(FAK1) that activates RhoA through Ras protein–specific guanine nucleotide-releasing factor 1 (RASGRF1). B, Agonist-induced heterodimerization of CXCR4 and CB2 inhibits Ga13 activation and dissociation from Gbg subunits. As a result, PRG ineffectively activates RhoA, which in turn, fails to mediate phosphorylation of myosin light chain leading to abrogation of cell migration, invasion, and adhesion. Signaling through alternate pathways, including integrin-mediated signaling, is insufficient to overcome inhibition caused by heterodimerization.

cell migration via antagonism of the RhoA pathway (Fig. 7A RhoGTPases are key regulators of cytoskeletal dynamics that and B). CXCL12 binds to CXCR4 and activates RhoA specif- lead to cell adhesion, polarity, and directional movement. ically through activation of the small G protein, Ga13 (27), Across several studies, CXCR4 was shown to induce cell migra- which leads to the subsequent activation of PRG, and is tion via signaling through Rho GTPases, which therefore sug- required for directional cell migration (ref. 35; Fig. 7A). Once gests this group of proteins as a treatment target for tumors that activated, RhoA induces changes in the actin cytoskeleton express CXCR4 (27, 42). During cell movement, Rho is impor- through phosphorylation of myosin light chain (26, 27), tant for regulating the formation of contractile actin–myosin increased F-actin protein levels, and changes in integrin filaments and stress fibers, and maintaining focal adhesions at expression (39–41). Molecular signaling events initiated by the rear of the migrating cells, whereas Rac is involved in the CXCL12/CXCR4 cascade enhances cellular functionality forming actin-rich membrane ruffles, or lamellipodia forma- by increasing cell migration and adhesion, ultimately contrib- tion, at the leading edge of the migrating cells, which is uting to tumor progression (Fig. 7A). We demonstrated that required to determine directional movement (42). Therefore, the physical heterodimer between CXCR4/CB2 reduced Ga13 the collective events resulting in Rho GTPases represent a key recruitment of PRG and subsequent RhoA activation (Fig. 7B). regulatory event for the intentional chemotaxis of cancer cells We also demonstrated that reduced migration, adhesion, and (27). Our findings, that heterodimerization reduced CXCR4- other coordinating events that encourage tumor cell migration mediated RhoA signaling and phenotypic changes associated was specifically attributed to the RhoA pathway as reduction with cell migration, may have broad implications as the upon heterodimerization was parallel to treatment with RhoA/ CXCL12/CXCR4 signaling axis is critical to a successful meta- ROCK inhibitor, Y-27632. static tumor.

OF10 Mol Cancer Res; 2018 Molecular Cancer Research

Inhibition of Ga13/RhoA-mediated Migration

Receptor dimerization is becoming a key paradigm in GPCR present this specific heterodimer as a target for metastasis biology and cancer therapeutics. It has been reported for most intervention. aspects of GPCR function: trafficking, internalization, and pharmacologic antagonism and signal transduction (3, 36, Disclosure of Potential Conflicts of Interest – 43 45). Although the functional consequences of receptor No potential conflicts of interest were disclosed. heterodimerization are still emerging, physical heterodimers amplify or antagonize signals that would otherwise result from Authors' Contributions fi each receptor individually, and change the signaling pro le Conception and design: K.A. Scarlett, C.V. Hinton or direction to favor one downstream pathway over another Development of methodology: K.A. Scarlett, E.Z. White, C.V. Hinton (43, 44). When CXCR4/CB2 heterodimerize, the receptor pair Acquisition of data (provided animals, acquired and managed patients, reduces signals that would come from CXCR4 alone, strongly provided facilities, etc.): K.A. Scarlett, E.Z. White, J.R. Carter, L.K. Bryant, supporting our global results that inhibiting the action of C.V. Hinton, C.J. Coke Analysis and interpretation of data (e.g., statistical analysis, biostatistics, CXCR4 whether via: (i) induced receptor heterodimerization; computational analysis): K.A. Scarlett, E.Z. White, J.R. Carter, L.K. Bryant, (ii) chemical antagonism of CXCR4 with AMD3100; or (iii) C.V. Hinton, C.J. Coke reduced translation of CXCR4 protein via siRNA will reduce Writing, review, and/or revision of the manuscript: K.A. Scarlett, E.Z. White, the propensity of CXCR4-mediated signaling to activate cell C.V. Hinton migration. Administrative, technical, or material support (i.e., reporting or organiz- GPCR heterodimers also alter ligand selectivity where ing data, constructing databases): K.A. Scarlett, C.V. Hinton Study supervision: K.A. Scarlett, E.Z. White, C.V. Hinton physical heterodimerization resulted in negative binding cooperativity—only one chemokine ligand binds with high affinity to the receptor dimer pair (36–38). Considering the Acknowledgments overwhelming clinical and social support for medical canna- This work was supported by grant funding to C.V. Hinton from NIH/ NIGMS (R01GM106020). Training support for students was provided, in binoids in cancer treatment, our current and previous (3) part, by NIH/NIGMS 2R25GM060414 and NIH/NIHMD 4P20MD002285; studies mechanistically demonstrate cannabinoid applica- confocal analyses was funded by Research Centers in Minority Institutions tions, and opens avenues for considering antagonizing CXCR4 Program (RCMI)/NIMHD (5G12MD007590). via induced CXCR4/CB2 heterodimers to slow the metastatic process. What's more, we use agonists, instead of antagonists, The costs of publication of this article were defrayed in part by the which currently result in severe immune dysfunction due to payment of page charges. This article must therefore be hereby marked inhibition of CXCR4 (46). The established role of CXCR4 in advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. metastasis and the broad role for cannabinoids in cancer metastasis treatment, pain management, palliative care, Received January 3, 2017; revised July 12, 2017; accepted December 20, 2017; bioavailability, and noninvasive administration (47, 48) published OnlineFirst January 12, 2018.

References 1. Lappano R, Maggiolini M. G protein-coupled receptors: novel targets for 12. Li S, Huang S, Peng S. Overexpression of G protein-coupled receptors drug discovery in cancer. Nat Rev Drug Discov 2011;10:47–60. in cancer cells: involvement in tumor progression. Int J Oncol 2005;27: 2. Dorsam RT, Gutkind JS. G-protein-coupled receptors and cancer. Nat Rev 1329–39. Cancer 2007;7:79–94. 13. Teicher BA, Fricker SP. CXCL12 (SDF-1)/CXCR4 pathway in cancer. 3. Coke C, Scarlett K, Chetram M, Jones K, Sandifer B, Davis A, et al. Clin Cancer Res 2010;16:2927–31. Simultaneous activation of induced heterodimerization between CXCR4 14. Duda DG, Kozin SV, Kirkpatrick ND, Xu L, Fukumura D, Jain RK. CXCL12 chemokine receptor and cannabinoid receptor 2 (CB2) reveals a mecha- (SDF1alpha)-CXCR4/CXCR7 pathway inhibition: an emerging sensitizer nism for regulation of tumor progression. J Biol Chem 2016;291:9991– for anticancer therapies? Clin Cancer Res 2011;17:2074–80. 10005. 15. Mackie K. Cannabinoid receptors: where they are and what they do. 4. Prinster SC, Hague C, Hall RA. Heterodimerization of g protein-coupled J Neuroendocrinol 2008;20:10–4. receptors: specificity and functional significance. Pharmacol Rev 2005;57: 16. Pacher P, Batkai S, Kunos G. The endocannabinoid system as an emerging 289–98. target of pharmacotherapy. Pharmacol Rev 2006;58:389–462. 5. Pello OM, Martinez-Munoz L, Parrillas V, Serrano A, Rodriguez-Frade JM, 17. Howlett AC, Barth F, Bonner TI, Cabral G, Casellas P, Devane WA, et al. Jose Toro M, et al. Ligand stabilization of CXCR4/delta-opioid receptor International union of pharmacology. XXVII. Classification of cannabi- heterodimers reveals a mechanism for immune response regulation. Eur J noid receptors. Pharmacol Rev 2002;54:161–202. Immunol 2008;38:537–449. 18. Cabral GA, Griffin-Thomas L. Emerging role of the cannabinoid receptor 6. Nasser M, Qamri Z, Deol Y, Smith D, Shilo K, Zou X, et al. Crosstalk CB2 in immune regulation: therapeutic prospects for neuroinflammation. between chemokine receptor CXCR4 and cannabinoid receptor CB2 in Expert Rev Mol Med 2009;11:e3. modulating breast cancer growth and invasion. PLoS One 2011;6:e23901. 19. Chakravarti B, Ravi J, Ganju RK. Cannabinoids as therapeutic agents 7. Satake H, Matsubara S, Aoyama M, Kawada T, Sakai T. Functional Regu- in cancer: current status and future implications. Oncotarget 2014;5: lation and Implication for Biodiversity. Front Endocrinol 2013;4:100. 5852–72. 8. Garland SL. Are GPCRs still a source of new targets? J Biomol Screening 20. Balkwill F. The significance of cancer cell expression of the chemokine 2013;18:947–66. receptor CXCR4. Semin Cancer Biol 2004;14:171–9. 9. Balkwill F. Cancer and the chemokine network. Nat Rev Cancer 2004;4: 21. Munoz~ LM, Lucas P, Holgado BL, Barroso R, Vega B, Rodríguez-Frade JM, 540–50. et al. Receptor oligomerization: a pivotal mechanism for regulating che- 10. Chatterjee S, Behnam Azad B, Nimmagadda S. The intricate role of mokine function. Pharmacol Thera 2011;131:351–8. CXCR4 in cancer. Adv Cancer Res 2014;124:31–82. 22. Stephens B, Handel TM. Chemokine receptor oligomerization and allo- 11. Feng Y, Broder CC, Kennedy PE, Berger EA. HIV-1 entry cofactor: functional stery. Prog Mol Biol Translat Sci 2013;115:375–420. cDNA cloning of a seven-transmembrane, G protein-coupled receptor. 23. Sahai E, Marshall CJ. RHO-GTPases and cancer. Nat Rev Cancer Science 1996;272:872–7. 2002;2:133–42.

www.aacrjournals.org Mol Cancer Res; 2018 OF11

Scarlett et al.

24. Spiering D, Hodgson L. Dynamics of the Rho-family small GTPases in actin 37. El-Asmar L, Springael JY, Ballet S, Andrieu EU, Vassart G, Parmentier M. regulation and motility. Cell Adh Migr 2011;5:170–80. Evidence for negative binding cooperativity within CCR5-CCR2b hetero- 25. Ridley AJ. Rho GTPase signalling in cell migration. Curr Opin Cell Biol dimers. Mol Pharmacol 2005;67:460–9. 2015;36:103–12. 38. Gillies K, Wertman J, Charette N, Dupre DJ. Anterograde trafﬁcking of 26. Vega FM, Ridley AJ. Rho GTPases in cancer cell biology. FEBS Lett 2008; CXCR4 and CCR2 receptors in a prostate cancer cell line. Cell Physiol 582:2093–101. Biochem 2013;32:74–85. 27. Tan W, Martin D, Gutkind JS. The Galpha13-Rho signaling axis is required 39. Lim SM, Trzeciakowski JP, Sreenivasappa H, Dangott LJ, Trache A. RhoA- for SDF-1-induced migration through CXCR4. J Biol Chem 2006;281: induced cytoskeletal tension controls adaptive cellular remodeling to 39542–9. mechanical signaling. Integr Biol 2012;4:615–27. 28. Struckhoff AP, Vitko JR, Rana MK, Davis CT, Foderingham KE, Liu CH, et al. 40. Marjoram RJ, Lessey EC, Burridge K. Regulation of RhoA activity by Dynamic regulation of ROCK in tumor cells controls CXCR4-driven adhesion molecules and mechanotransduction. Curr Mol Med 2014;14: adhesion events. J Cell Sci 2010;123:401–12. 199–208. 29. Chan KT, Asokan SB, King SJ, Bo T, Dubose ES, Liu W, et al. LKB1 loss in 41. Engl T, Relja B, Marian D, Blumenberg C, Muller I, Beecken WD, et al. melanoma disrupts directional migration toward extracellular matrix cues. CXCR4 chemokine receptor mediates prostate tumor cell adhesion J Cell Biol 2014;207:299–315. through alpha5 and beta3 integrins. Neoplasia 2006;8:290–301. 30. Nakano A, Takashima S. LKB1 and AMP-activated protein kinase: regula- 42. Azab AK, Azab F, Blotta S, Pitsillides CM, Thompson B, Runnels JM, et al. tors of cell polarity. Genes Cells 2012;17:737–47. RhoA and Rac1 GTPases play major and differential roles in stromal cell- 31. Xu X, Jin D, Durgan J, Hall A. LKB1 controls human bronchial epithelial derived factor-1-induced cell adhesion and chemotaxis in multiple mye- morphogenesis through p114RhoGEF-dependent RhoA activation. Mol loma. Blood 2009;114:619–29. Cell Biol 2013;33:2671–82. 43. Milligan G. G-protein-coupled receptor heterodimers: pharmacology, 32. Roh-Johnson M, Bravo-Cordero JJ, Patsialou A, Sharma VP, Guo P, Liu H, function and relevance to drug discovery. Drug Discov Today 2006;11: et al. Macrophage contact induces RhoA GTPase signaling to trigger tumor 541–9. cell intravasation. Oncogene 2014;33:4203–12. 44. Terrillon S, Bouvier M. Roles of G-protein-coupled receptor dimerization. 33. Yagi H, Tan W, Dillenburg-Pilla P, Armando S, Amornphimoltham P, EMBO Rep 2004;5:30–4. Simaan M, et al. A synthetic biology approach reveals a CXCR4-G13-Rho 45. Decaillot FM, Kazmi MA, Lin Y, Ray-Saha S, Sakmar TP, Sachdev P. CXCR7/ signaling axis driving transendothelial migration of metastatic breast CXCR4 heterodimer constitutively recruits beta-arrestin to enhance cell cancer cells. Sci Signal 2011;4:ra60. migration. J Biol Chem 2011;286:32188–97. 34. Reymond N, Riou P, Ridley AJ. Rho GTPases and cancer cell transendothe- 46. Nagarsheth N, Wicha MS, Zou W. Chemokines in the cancer microenvi- lial migration. Methods Mol Biol 2012;827:123–42. ronment and their relevance in cancer immunotherapy. Nat Rev Immunol 35.StruckhoffAP,RanaMK,KherSS,BurowME,HaganJL,DelValleL, 2017;17:559–72. et al. PDZ-RhoGEF is essential for CXCR4-driven breast tumor cell 47. Percherancier Y, Berchiche YA, Slight I, Volkmer-Engert R, Tamamura H, motility through spatial regulation of RhoA. J Cell Sci 2013;126: Fujii N, et al. Bioluminescence resonance energy transfer reveals ligand- 4514–26. induced conformational changes in CXCR4 homo- and heterodimers. 36. Mellado M, Rodriguez-Frade JM, Vila-Coro AJ, Fernandez S, Martin J Biol Chem 2005;280:9895–903. de Ana A, Jones DR, et al. Chemokine receptor homo- or hetero- 48. Ramos JA, Bianco FJ. The role of cannabinoids in prostate cancer: Basic dimerization activates distinct signaling pathways. EMBO J 2001;20: science perspective and potential clinical applications. Indian J Urol 2497–507. 2012;28:9–14.

OF12 Mol Cancer Res; 2018 Molecular Cancer Research

Agonist-induced CXCR4 and CB2 Heterodimerization Inhibits G α13/RhoA-mediated Migration

Kisha A. Scarlett, El-Shaddai Z. White, Christopher J. Coke, et al.

Mol Cancer Res Published OnlineFirst January 12, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-16-0481

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/early/2018/03/06/1541-7786.MCR-16-0481. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.