Integrated Cell-Based Platform to Study EGFR Activation and Transactivation Marie-Elaine Caruso, Paule Cle´ ment, Ste´ phane Parent, Vincent Dupriez, Roger Bosse´ , and Nathalie Rouleau PerkinElmer BioSignal, Inc., Montreal, Canada.

ABSTRACT The epidermal growth factor receptor (EGFR) pathway is one of the most deregulated molecular pathways in human epithelial cancers. Many approved drugs were optimized to directly target EGFR but yielded only modest clinical improvement in cancer patients due to low efficacy and drug resistance. Transactivation of EGFR by other cell surface receptors such as G-protein-coupled receptors (GPCRs) was proposed to explain this lack of efficacy. Even if direct EGFR activation and transactivation by GPCR contribute to the activation of the same signaling pathways, they are often studied as independent events resulting in partial investigation of a drug’s mechanism of action. We present a novel highthroughput approach that integrates interrogation of direct activation of EGFR and its transactivation via GPCR activation. Using distinct technology platforms, three readouts were used to measure (1) direct activation of GPCR via cyclic adenosine monophosphate (cAMP) detection, (2) direct activation of EGFR through the release of intracellular Ca2+, and (3) EGFR transactivation by GPCR using the detection of p-extracellular-signal-regulated kinases 1/2 (p-ERK1/2). In addition to being simple, quick, and homogenous, our methods were shown to be more sensitive than those in current use. These enabling tools should improve the knowledge pertaining to GPCRs and receptor tyrosine kinases trans-regulation and facilitate the design of more potent and better targeted new therapeutic strategies.

INTRODUCTION

T

he epidermal growth factor receptor (EGFR) is the prototype of the ErbB cell surface tyrosine kinase receptor family, and it plays a key role in cellular homeostasis. EGFR activation occurs when agonists, including epidermal growth factor (EGF), transforming growth factor alpha (TGFa), or amphiregulin (AR), bind to the receptor N-terminal extremity, leading to homo- or hetero-dimerization followed by trans-autophosphorylation at multiple tyrosine residue sites in the receptor’s C-terminus part.1 The

resulting phosphotyrosine residues create docking sites for several enzymes or adaptor proteins containing SH2 domains leading to downstream signaling. The four major EGFR signaling pathways include (1) the mitogen-activated protein kinase cascade (MAPKs), (2) the p85/p110 phorphatidylinositol-3 kinase (PI3-K) cascade, (3) members of the STAT (signal transducers and activator of transcription) family, and (4) the phospholipase C gamma (PLCg) cascades. Their differential activation ultimately culminates in the modulation of various transcription factor target genes, leading to changes in cellular proliferation, differentiation, migration, survival, metabolism, and cycle control.1 Recent investigations demonstrated that G-protein-coupled receptors (GPCRs) and other members of the receptor tyrosine kinase (RTK) family can also activate the EGFR signaling pathway through a mechanism called receptor transactivation. A first evidence for this mechanism was proposed by Daub et al., who showed that on stimulation of Rat-1 cells with GPCR agonists such as endothelin-1, lysophosphatic acid, and thrombin, EGFR became rapidly phosphorylated.2 Later, EGFR transactivation was shown to occur through activation of various GPCRs, including classical neurotransmitter receptors such as b-adrenergic receptors (b2AR)3,4 or dopamine receptors,5 by peptide receptors such as angiotensin II receptor,6 or by estrogen receptor GPR30.7,8 Moreover, transactivation was shown to occur in various cell types, including HEK, CHO, and HeLa, and in tissues.9 The molecular mechanism by which EGFR is transactivated involves the cleavage of transmembrane pro-ligands, such as TGF-a, AR, and heparin binding EGF, by members of the ADAM family of matrix metalloproteases (MMPs)—ADAM 10, 12 and 17.10 The processed pro-ligands can then bind and activate EGFR, which, in turn, transduces signals to its prominent downstream signaling pathways. Interestingly, MMP inhibitors were shown to abrogate GPCRmediated activation of EGFR.11 GPCRs can also transactivate EGFR through intracellular tyrosine kinase domain activation. This ligand-independent mechanism acts via Src phosphorylation and was validated using intracellular Src-like tyrosine kinases inhibitors (PP1 and PP2) that block EGFR transactivation.12 Finally, in certain cell types, RTKs were also demonstrated to be transactivated through inactivation of protein tyrosine phosphatase by

ABBREVIATIONS: AR, amphiregulin; cAMP, cyclic adenosine monophosphate; EGF, epidermal growth factor; EGFR, epidermal growth factor receptor; ERK1/2, extracellular-signal–regulated kinases 1/2; FGFR, fibroblast growth factor receptor; GPCRs, G protein-coupled receptors; HNSCC, head and neck squamous cell carcinoma; MAPK, mitogen-activated protein kinase cascade; MMP, matrix Metalloproteases; NSCLC, nonsmall cell lung cancer; PI3-K, p85/p110 phorphatidylinositol-3 kinase; PLCg, phospholipase C gamma; RTK, receptor tyrosine kinase; S/B, signal-to-background; STAT, signal transducers and activator of transcription; TGFa, transforming growth factor alpha; b2AR, b2-adrenergic receptor.

DOI: 10.1089/adt.2013.518

ª MARY ANN LIEBERT, INC.  VOL. 11 NO. 7  SEPTEMBER 2013 ASSAY and Drug Development Technologies

423

CARUSO ET AL.

NADPH-mediated release of reactive oxygen species, a mechanism that is still poorly understood.13 The EGFR pathway is one of the most deregulated molecular pathways in human epithelial cancers, including colon, lung, breast, head, and neck variants. Many drugs were developed and approved to treat cancer patients showing abrogated EGFR expression or presenting genetic EGFR alterations.14,15 However, these treatments resulted in only modest clinical improvement except for patients affected with nonsmall cell lung cancer (NSCLC) carrying EGFR activation mutations. It is worth mentioning that most frequent mutations found in NSCLC are attributable to the deletion of exon 19 and the single-point substitution mutation L858R in exon 21.16 However, this latter population represents only 10% of all NSCLC-affected patients.17 Growing evidence suggests that EGFR transactivation by GPCRs may explain the low treatment success rate. As a matter of fact, some GPCRs were shown to be overexpressed in some EGFRrelated cancers, and a study analyzing more than 60 human carcinoma cell lines underlined the importance of the GPCR-induced EGFR activation in carcinomas.18,19 Moreover, ADAM17 overexpression was shown to mediate GPCR ligand-dependent transactivation of EGFR in colon cancer.20 Finally, preclinical studies suggested that combining inhibition of GPCR and EGFR may result in additive or synergistic inhibition of head and neck cancers (HNSCC) and NSCLC.21,22 Therefore, therapeutic strategies to inhibit both GPCR and EGFR pathways may help increasing cancer treatment effectiveness. Despite the fact that GPCR and EGFR pathways are closely interconnected in a complex cellular crosstalk system, these pathways are often studied independently. Moreover, most of the existing biochemical assays developed to study receptor transactivation rely on Western blotting techniques that are inappropriate to carry out highthroughput drug screening. Last but not the least, there is a crying need for tools enabling the integrated study of signaling pathways that are specific to particular cell phenotypes and to find specific drugs eliciting responses from those cells. This work presents the development of new screening tools that are capable of studying RTK cellular functions through the integration of the multiple signaling pathways linked to GPCR-EGFR crosstalk. During our study, we used the EGFR-b2AR pair, as it is one of the most representative examples of GPCR-induced EGFR transactivation mechanism. Three distinct cell-based homogenous and high-throughput assay technologies were then chosen to discriminate between (1) direct activation of b2AR (using the LANCE Ultra cyclic adenosine monophosphate [cAMP] technology), (2) direct activation of the EGFR (using the Aequorin assay to measure mobilized Ca ++ ), and (3) EGFR transactivation by b2AR (using the AlphaScreen SureFire phospho–extracellular-signal-regulated kinases 1/2 [ERK1/2] platform). This novel approach should allow the identification of new compounds that are capable of selective modulation of the responses listed earlier. These compounds could be used to further delineate the mechanisms linking GPCR and RTK signaling pathways and may also facilitate the design of new therapeutic strategies for patients with EGFR-related cancers.

424 ASSAY and Drug Development Technologies SEPTEMBER 2013

MATERIALS AND METHODS Reagents and Plasmids Ham’s F12 culture media (Cat# SH30026.02) and trypsin (Cat# SV30031.01) were from Hyclone/Thermo Fisher Scientific, Inc. Fetal bovine serum (FBS; Cat# 80150) and G418 (Cat# 400-130) were from Multicell Technologies, Inc. Selection antibiotics Zeocin (Cat# R250-01) and Blasticidin (Cat# R210-01) as well as DMEM/F12 with HEPES and without phenol red media (Cat# 11039), HEPES 1M (Cat# 15630), and HBSS (Cat# 14025) were from Invitrogen/Life Technologies Corporation. FuGENE 6 (Cat# E2691) and coelenterazine h (Cat# S2011) were from Promega Corporation. Digitonin (Cat# D5628), ATP (Cat# A7699), Tyrphostin AG 1478 (Cat# T4182), Salbutamol (Cat# S5013), TGFa (Cat# T7924), ICI 118, 551 hydrochloride (Cat# I127), BSA (Cat# A9418), and IBMX (Cat# I7018) were from Sigma-Aldrich Co. LLC. EGF (Cat# 236-EG-200) and AR (Cat# 262-AR-100) were from R&D Systems, Inc., and PP2 (Cat# 1407) and Salmeterol (Cat# 1660) were from Tocris Bioscience. Both LANCE Ultra cAMP (Cat# TRF0262) and AlphaScreen SureFire p-ERK1/2 Thr202/tyr204 (Cat# TGRES10K) assay kits, AlphaLISA protein A Acceptor beads (Cat# AL101), assay plates (Black 384-well ViewPlate [Cat# 6007460], and White 384-well Optiplate [Cat# 6007290]), and TopSeal-A 384 (Cat# 6005250) microplates were from PerkinElmer, Inc. EGFR1, FGFR1, and b2AR genes were synthesized at GeneArt/Life Technologies Corporation and subcloned in the pEFIN5 vector containing a resistance gene for G418.

Cell Culture and Transfection Three selection antibiotics were required for the cell line development. Zeocin was used for the selection of the Aequorin gene; G418, for the selection of RTK; and, finally, Blasticidin for the selection of b2AR. Therefore, the culture media used for each cell line expansion was as follows: CHO-EGFR-b2AR: Ham’s F12, 10% FBS, 250 mg/mL Zeocin, 400 mg/mL G418, and 5 mg/mL Blasticidin; CHOEGFR: Ham’s F12, 10% FBS, 250 mg/mL Zeocin, and 400 mg/mL G418; CHO-b2AR: Ham’s F12, 10% FBS, 250 mg/mL Zeocin, and 5 mg/mL Blasticidin; and the wild-type CHO-K1 Aequorin parental cell line was grown in Ham’s F12, 10% FBS, and 250 mg/mL Zeocin. pEFIN-EGFR vector was transfected in CHO-K1 parental cells following Promega recommendations. Briefly, CHO-K1 AequoScreen parental cells (PerkinElmer ES-000-A12) were plated at 0.6 · 106 cells per well of a six-well plate in culture media (Ham’s F12 + 10% FBS + 250 mg/mL Zeocin) and grown overnight. At 50% confluence, cells were transfected with 2 mg of pEFIN-EGFR plasmid and 5 mL of the FuGENE6 reagent following the manufacturer’s recommendations. About 48 h post-transfection, the culture medium was exchanged to medium supplemented with 800 mg/mL G418 for plasmid selection. Selection proceeded for 10 days while changing media every 3 days. When colonies clearly formed, cells were trypsinized and transferred to a new flask containing G418 supplemented medium. Cells were then isolated following limiting dilution protocol to obtain a homogenous culture. Clones were tested for Aequorin response (see Aequorin Assay for EGFR1-Directed Activation). Those showing a response to 500 ng/mL EGF higher than 70% of the digitonin-positive

CELL-BASED ASSAYS TO STUDY EGFR

control treatment were selected. After the clone’s selection, cells were propagated as for untransfected cells, but in selective medium (Ham’s F12 + 10% FBS + 250 mg/mL Zeocin + 400 mg/mL G418). For b2AR, the same transfection procedures were used except that a mix of 5 mg of pEFIN5-b2AR and 0.5 mg of pEFIB3 were transfected into CHO cell to link b2AR expression to Blasticidin selection marker. Clones were selected using 10 mg/mL Blasticidin. Isolated clones were tested for b2AR expression using the LANCE Ultra cAMP assay as described next. Cells were propagated in Ham’s F12 + 10% FBS + 250 mg/mL Zeocin + 5 mg/mL Blasticidin.

Table 1. Aequorin Protocol to Measure Direct Epidermal Growth Factor Receptor Activation and Inhibition Step

Parameter

Value

Description

Preparation of cells 1

Cell solution

2

Coelenterazine

2.5 · 106 cells/mL

In Aequorin assay buffer

5 mM final

h addition 3

Incubation time

4 h or overnight

Ambient temperature on vertical rotation wheel

Frozen Cell Bank Production Frozen cell banks were prepared from freshly cultured cells. Cells were detached using versene, counted, and resuspended at 4 · 106 cells/mL in culture media containing 10% FBS but no antibiotics. Cell suspension was diluted 1/2 in culture freezing media (culture media, 10% FBS, and 20% DMSO) for a final DMSO concentration of 10%. Cell suspension was distributed in 1 mL aliquot into frozen tubes at 2 · 106 cells/mL. Tubes were placed at -80C for 24 h before being transferred into liquid nitrogen for longer storage.

Aequorin Assay for EGFR1-Directed Activation For the Aequorin assay, cells were thawed and transferred in a tube containing 10 mL of complete media, counted using the Cedex cell counter (Roche Diagnostics Corporation), centrifuged at 110 g for 10 min, and resuspended in Aequorin assay buffer (DMEM/F12 with HEPES and without phenol red media, 0.1% BSA) at a concentration of 2.5 · 106 cells/mL. Cœlenterazine h (5 mM) was added to the cell suspension and incubated for at least 4 h at room temperature on a vertical rotation wheel. Cells were then diluted 10 times with Aequorin assay buffer at room temperature, and kept in a black bottle with magnetic stirring for 60 min with agitation for cell equilibration. For agonist assays, 20 mL of each agonist concentration was distributed in triplicate in black 384-well ViewPlate microplates. Using the Microbeta LumiJet (PerkinElmer) microplate counter, 20 mL of cell suspension was then injected in the plate containing ligands, and the signal was monitored for 30 s. For antagonist assays, 20 mL of cells were injected on black 384-well ViewPlate microplates (PerkinElmer) containing 20 mL of each concentration of antagonist and incubated 15 min. Then, 20 mL of agonist was injected at a concentration corresponding to its EC80, and plates were read for 30 s. EC80 was calculated using the following formula: EC80 = (80/100 - 80)1/hill slope · EC50. Digitonin (50 mM final concentration) was used as a positive control and utilized to normalize the signal between various cell suspension preparations (Table 1).

LANCE Ultra cAMP Assay for GPCR-Directed Activation To monitor GPCR direct activation, we used the LANCE Ultra cAMP detection kit. This assay measures cyclic AMP production by whole cells treated with GPCR agonists or antagonists. This assay was used to characterize the b2AR activity within the new cell lines developed. Briefly, cells were thawed and transferred in a tube containing 10 mL of complete media, counted using the Cedex cell

ª MARY ANN LIEBERT, INC.  VOL. 11

4

Cell dilution

5

Incubation time

1/10 dilution 60 min

In Aequorin assay buffer In black bottle with magnetic stirring

a: Agonist assay 6a

Dispense agonists

20 mL

1 mM to 1 pM dilution series

7a

Cell injection

20 mL

8a

Assay readout

30 s

Microbeta LumiJet

1 mM to 1 pM dilution series

b: Antagonist assay 6b

Antagonists

20 mL

7b

Cell injection

20 mL

8b

Incubation time

15 min

9b

Agonist injection

20 mL

At concentration corresponding to EC80

10b

Assay readout

30 s

Microbeta LumiJet

Step Notes 1. Cells are thawed from a frozen cell bank, counted, and resuspended to the desired concentration in Aequorin assay buffer. 4. This dilution is used to reduce the amount of coelenterazine used. It is facultative. If removed, cells should be resuspended at 0.25 · 106 cells/mL in Step 1. 6a,b. In a black 384-well Viewplate microplate; three wells per conditions; also add 50 mM Digitonin as a positive control. 7a,b. Use the MicrobetaJet injector 8b. During this incubation time, calculate agonist EC80 and prepare agonist dilution EC80 = (80/100 - 80)1/hill slope · EC50 9b. Use the MicrobetaJet injector

counter (Roche Diagnostics Corporation), centrifuged at 1,000 rpm for 10 min, and resuspended in stimulation buffer (1· HBSS, 5 mM HEPES pH 7.4, 0.5 mM IBMX, and 0.1% BSA) at a concentration of 0.5 · 106 cells/mL. For agonist testing, serial dilutions of b2AR agonists (Salmeterol, Salbutamol, Isoproterenol, and Norepinephrin) were prepared as a 2· concentrated solution in LANCE stimulation buffer. Five microliters of cell suspension (2,500 cells/well) were added in triplicate in a

NO. 7  SEPTEMBER 2013

ASSAY and Drug Development Technologies 425

CARUSO ET AL.

fold in dilution buffer) was then added (3.5 mL) under low light conwhite 384-well Optiplate microplate followed by 5 mL of each agonist ditions; the plate was incubated for 1 h at 23C and read using the dilutions. Plates were sealed using a TopSeal-A 384 sealing tape Envision Multilabel reader (Cat# 2104-0010; PerkinElmer) using (PerkinElmer) and incubated for 25 min at 37C. For detection, 5 mL of a standard AlphaScreen protocol (Table 2). 4· preparation of Eu-cAMP tracer diluted in LANCE detection buffer For the Z0 determination, cells were prepared as for transactivation and 5 mL of a 4· solution of ULight-anti-cAMP diluted in LANCE detection buffer were added, respectively, in the plate. Plates were sealed assays and were treated with buffer, 1 mM Isoproterenol, or 0.1 mM and incubated for 1 h before reading on the Envision Multilabel reader EGF for 10 min. Z0 values were calculated using the following (PerkinElmer) using the LANCE standard protocol. Note that the Topformula: 1 - ([3 · SDagonist + 3 · SDbuffer]/ABS [AVERAGEagonist Seal tape should be removed before performing LANCE assay readout. AVERAGEbuffer]). The detection step was identical as described for Antagonist (ICI 118,551 hydrochloride) was diluted from 1 pM to transactivation assays. 100 mM as a 4· solution in LANCE stimulation buffer. Isoproterenol was diluted as a 4· solution at a concentraTable 2. Epidermal Growth Factor Receptor Transactivation Protocol tion corresponding to EC90 in LANCE stimulation buffer. In Using AlphaScreen SureFire pERK1/2 Detection Kit a white 384-well Optiplate, 5 mL of cell suspension (2,500 cells/well) was added in triplicate followed by 2.5 mL of each Step Parameter Value Description antagonist and agonist dilutions. Plates were sealed using Cell culture and treatment steps TopSeal-A 384 tape and incubated for 25 min at 37C. Detection was performed as for agonist assays. 1 CHO cell resuspension 5 · 106 cells/mL In HBSS + 0.1% BSA

AlphaScreen SureFire p-ERK1/2 Assay for EGFR1 Transactivation AlphaScreen SureFire phospho/ERK1/2 is a unique platform that is used for measuring phosphorylated ERK1/ 2 in cell lysates. The principle of the assay relies on the use of the Donor and Acceptor beads that emit light when brought into close proximity. For the p-ERK1/2 assay, protein A acceptor beads bind to an anti-phosphorylated ERK1/2 antibody; while on the other hand, the streptavidine Donor beads recognize a biotinylated anti-ERK1/2 antibody. Briefly, cells were thawed and transferred in a tube containing 10 mL of complete media without FBS; cells were counted using the Cedex cell counter (Roche Diagnostics Corporation), centrifuged at 110 g for 10 min, and re-suspended in SureFire assay buffer (1· HBSS, 0.1% BSA) at a concentration of 5 · 106 cells/ mL. Four microliters (20,000 cells/well) of cell suspension was added in triplicate to a white 384-well Optiplate. Antagonists were diluted as 3· solutions (AG1478 or ICI118, 551, and 1 mM final) in SureFire assay buffer and added to the plate. When cells were treated with agonists only, SureFire assay buffer containing the corresponding amount of DMSO was used. Plates were sealed and incubated for 2 h at 37C. Cells were then stimulated with 2 mL of each agonist (EGF 0.1 mM, isoproterenol 1 mM final concentration in 8 mL) or assay buffer for control and incubated for 10 min at 23C. Cells were lysed using 2 mL of 5 · SureFire lysis buffer, and the plate was agitated for 10 min at 23C. During this incubation, the acceptor mix was prepared as follows: In reaction buffer, the activation buffer was diluted 7-fold and AlphaLISA protein A Acceptor beads were diluted 70-fold. Then, 8.5 mL of this mix was added in the plate and incubated for 2 h at 23C. Donor mix (Alpha Donor beads diluted 20-

426 ASSAY and Drug Development Technologies

SEPTEMBER 2013

2

Plate cells

4 mL

20,000 cells/well

3

Antagonists or controls

2 mL

AG1478 or ICI-118,551 1 lM final

4

Incubation time

2h

37C, 5% CO2

5

Agonists

2 mL

EGF or ISO 1 lM to 1 pM dilution series

6

Incubation time

7

Lysis buffer

8

Incubation time

10 min 2 mL 10 min

23C 5 · solution 23C

Detection steps 9

Acceptor mix

8.5 mL

Dilute activation buffer 7-fold and AlphaLISA protein A Acceptor beads 70-fold in reaction buffer

10

Incubation time

11

Donor beads

2h 3.5 mL

23C Dilute Donor beads 20-fold in dilution buffer; work under low light conditions

12

Incubation time

13

Assay readout

1h 520–620 nm

23C Envision, Alpha mode

Step Notes 1. CHO cells are thawed in culture media without fetal bovine serum, centrifuged, and counted 2. White 384-well Optiplate—triplicate wells per set of assay conditions 3. Use HBSS + 0.1% BSA + DMSO as a negative control 4. Plate are sealed with plastic TopSeal 5. Use HBSS + 0.1% BSA + DMSO as a negative control 6. Plate are sealed with plastic TopSeal 7. Positive and negative lysates provided in the assay kit should be added here as detection controls The major steps required for optimization of EGFR transactivation are presented in boldface. ERK1/2, extracellular-signal-regulated kinases 1/2.

CELL-BASED ASSAYS TO STUDY EGFR

Data Analysis All data generated in this project were analyzed using GraphPad Prism software. Graphs were drawn using nonlinear regressions curve fitting. EC50 and IC50 represented the agonist or antagonist concentration that gives a response half way between the bottom and top plateaus of the dose-response curve.

RESULTS AND DISCUSSION With the goal of providing powerful tools to study RTK-GPCR crosstalk, we developed and optimized three homogenous and highthroughput assays to measure (1) b2AR direct activation using the LANCE Ultra cAMP assay, (2) EGFR1 direct activation using either

Aequorin Ca2+ assay or the AlphaScreen SureFire phospho-ERK1/2 assay, and (3) EGFR1 transactivation by b2AR using AlphaScreen SureFire phospho-ERK1/2 assay (Fig. 1). These assays provide three different and orthogonal readouts that can be used to profile compounds during lead optimization in order to increase our understanding of compound mechanism of action. As a proof of concept, we selected the EGFR1 and the b2AR, as both receptors were well characterized and the transactivation of the EGFR1 by b2AR was previously demonstrated.22 Indeed, Maudsley et al. were the first to demonstrate that the EGFR1 is trans-activated through b2AR stimulation, leading to ERK1/2 phosphorylation in COS-7 recombinant cells.22 Moreover, EGFR1 represents an important target for cancer

Fig. 1. Technological platform developed to study RTK direct activation and transactivation by GPCRs. This platform allows independent measurements of (1) GPCR activation using LANCE Ultra cAMP assay, (2) direct RTK activation using Aequorin assay, and (3) RTK transactivation by GPCR using the AlphaScreen SureFire p-ERK1/2 (Thr202/Tyr204) assay. All technologies are homogenous and allow high-throughput studies. Principles of each technology are described in the Material and Methods section. cAMP, cyclic adenosine monophosphate; ERK1/2, extracellular-signal-regulated kinases 1/2; GPCRs, G protein-coupled receptors; RTK, receptor tyrosine kinase.

ª MARY ANN LIEBERT, INC.  VOL. 11

NO. 7  SEPTEMBER 2013

ASSAY and Drug Development Technologies 427

CARUSO ET AL.

therapy, and, therefore, these assays could ultimately enable the development of new therapeutic approaches targeting EGFR-GPCR crosstalk. Such a strategy has successfully been investigated for patients with HNSCC. In HNSCC tumors, the high expression level of EGFR predicts decreased survival, and EGFR inhibitor monotherapy has shown limited antitumor effect due to drug resistance.4,21,23 However, when a gastrin-releasing peptide receptor inhibitor (PD176252) was used in conjunction with an EGFR inhibitor (erlotinib), enhanced antitumor effect was observed compared with EGFR monotherapy.21

CHO-Aeq-EGFR1 Cell Line Development and Characterization Using Aequorin Technology Aequorin is a Ca2 + -binding photoprotein that, in the presence of oxygen, spontaneously interacts with its cofactor, coelenterazine. On Ca2 + binding, the protein undergoes a conformational change and behaves as an oxygenase that converts coelenterazine into excited coelenteramide and carbon dioxide. As the excited coelenterazide relaxes to its ground state, blue light is emitted and can be easily measured at a wavelength of 469 nm (Fig. 1).24 Aequorin cell lines are useful sensors to study receptors that induce mobilization of intracellular calcium release. Over the last decade, this technology has been used to study many GPCRs and to screen modulators of their activities.24–26 In this work, the Aequorin technology was adapted to the RTK family (Fig. 2A). It is well established that activation of certain members of the RTK family, including EGFR, fibroblast growth factor receptor (FGFR), platelet-derived growth factor receptor (PDGFR), and vascular endothelial growth factor receptor (VEGFR), triggers a series of intracellular transduction cascade, one of them being the activation of PLCg, leading to the generation of inositol phosphate and release of intracellular Ca2 + .1 Using CHO-K1 Aequorin parental cells, which express a mitochondrial form of the Aequorin,24 EGFR1 was transfected, and clones were isolated to obtain a homogenous cell population expressing both the Aequorin and the EGFR1. The CHO-Aeq-EGFR1 cell line was then tested for its Aequorin response to EGFR1 stimulation. Figure 2B shows that on CHO-Aeq-EGFR1 cell line treatment with EGFR agonists (EGF, TGFa, and AR), an Aequorin signal was obtained with a signal-to-background (S/B) ratio ranging from 60 to 75. EC50 values of 3.6, 1.4, and 24.8 nM were obtained for EGF, TGFa, and AR, respectively, with rank order of potency as expected from the literature.24,27,28 Antagonist assays were also performed using the Tyrphostin AG 1478, an inhibitor of the EGFR kinase domain (Fig. 2C). Again, CHO-AeqEGFR1 cells responded well to AG 1478 antagonist with an expected IC50 value of 57.5 nM.28 Taken together, these results demonstrate that the CHO-Aeq-EGFR1 cell line is responding as expected to EGFR1 ligands and is a suitable tool to study EGFR1 activation. The same procedure was followed to develop a CHO-Aequorin cell line for the FGFR1 (Supplementary Fig. S1; Supplementary Data are available online at www.liebertpub.com/ adt), supporting the adaptability of the Aequorin technology for studying RTKs.

428 ASSAY and Drug Development Technologies SEPTEMBER 2013

In order to provide tools to measure EGFR transactivation by GPCR, we then coexpressed the b2AR and the EGFR in the CHO cell line. The b2AR was thus transfected in the CHO-Aeq-EGFR1 cell lines. The functionality of the EGFR in this double-transfected cell line was tested using the Aequorin assay and normalized to the digitonin signal. Digitonin is a nonionic detergent that is used to permeabilize cell membranes, inducing release of the intracellular Ca2 + from the endoplasmic reticulum. This digitonin response represents the highest signal expected from the Aequorin assay. It is used as a positive control and also serves to normalize the signal between different cell line suspensions. An EGF dose-response curve was performed in parallel on the CHO-Aeq-EGFR and CHO-Aeq-EGFRb2AR cell lines. The resulting Aequorin signal and EC50 values obtained with both cell lines were similar (Fig. 1D), proving that the expression of b2AR did not affect the expression or the activity of the EGFR1. Moreover, no response was obtained in the control cell lines CHO and CHO-b2AR, demonstrating that CHO-K1 cells do not have intrinsic expression of EGFR1.

b2AR Direct Activation Measurement Using LANCE Ultra cAMP The CHO-Aeq-EGFR-b2AR cell line was then tested for the functionality of the b2AR receptor using the LANCE Ultra cAMP assay kit that measures levels of intracellular cAMP. On activation of the b2AR by its agonists, cAMP is produced, competing with the Eu-tagged cAMP probe binding to a specific antibody. A signal decrease is, therefore, observed as an interaction between the probe Eu-cAMP, and anti-cAMP-Ulight tagged antibody is reduced (Fig. 3A). As shown in Figure 3B, all b2AR agonists tested were able to activate the GPCR, and IC50 values obtained are in the same order of magnitude as reported by others.* The basal level obtained with salmeterol is higher compared with isoproterenol, because this agonist is weaker and does not induce, at the cell concentration selected, enough cAMP to compete totally with the Eu-cAMP probe. Isoproterenol provides the highest assay window and a strong potency toward the b2AR and was, therefore, selected for the following experiments. As expected, no LANCE cAMP response was obtained when CHO-Aeq-EGFR-B2AR cells were treated with EGF (data not shown) or when Aequorin parental cells CHO-Aeq or CHO-EGFR1 cells were treated with isoproterenol (Fig. 3D). These last results demonstrate that CHO cells do not have an intrinsic b2AR response. An ICI-118, 551 (specific b2AR antagonist) dose-response curve was then performed using CHO-EGFRb2AR cells. As expected when an increased concentration of ICI118, 551 was used, less cAMP was produced and an increase in the LANCE signal was observed. An EC50 value of 20 nM was obtained (Fig. 3C) compared with the reported 70 nM.29,30 These results confirmed that the new cell lines developed respond as expected to

*Brescia P, Larson B, Banks P: Live-cell assay to interrogate GPCRs by monitoring cAMP levels using a bioluminescent readout. Application notes from BioTek; 2010 (unpublished).

CELL-BASED ASSAYS TO STUDY EGFR

Fig. 2. Development of the CHO-Aeq-EGFR1 cell line and characterization using the Aequorin technology. (A) Schematic representation of the Aequorin assay. In response to EGFR1 stimulation with specific agonists, intracellular Ca2 + is released, activating the apo-Aequorin, which then converts the cœlenterazine h into excited cœlenteramide. As the excited cœlenteramide relaxes to its ground state, a 469 nm flash light signal is emitted and measured using the MicroBeta LumiJET instrument. (B) The Aequorin CHO-EGFR cells were treated with increasing concentrations of EGFR agonists (EGF [C], TGFa [-], or AR [A]). Dose-response curves were obtained and EC50 values were determined. (C) For the antagonist assay, CHO-EGFR cells were first treated with an increasing concentration of antagonist (AG 1478 [C]) followed by stimulation with EGF at a concentration corresponding to the EC80 value. Dose-response curves were obtained, and IC50 value was determined. (D) Aequorin signal generated by EGF in Aequorin wild-type CHO cells (:), CHO-b2AR (A), CHO-EGFR (C), and CHO-EGFR- b2AR (-) cell lines. Percent of digitonin response is used to normalize the Aequorin signal obtained with each cell suspension. These experiments were repeated at least twice in triplicate. EGF, epidermal growth factor; TGFa, transforming growth factor alpha; AR, amphiregulin; b2AR, b2-adrenergic receptor.

ª MARY ANN LIEBERT, INC.  VOL. 11

NO. 7  SEPTEMBER 2013

ASSAY and Drug Development Technologies 429

CARUSO ET AL.

Fig. 3. Development of the CHO-EGFR1-b2AR cell line expression and characterization of the b2AR activity using the LANCE Ultra cAMP detection kit. (A) Schematic representation of the LANCE Ultra cAMP assay. On stimulation of CHO-EGFR-b2AR cells by b2AR agonists, cAMP is produced and competes with the Eu-cAMP tracer, leading to a decreased TR-FRET signal. (B) CHO-EGFR-b2AR cells were treated with increasing concentrations of b2AR agonists (salmeterol [salm; C], salbutanol [salb; -], isoproterenol [iso; :], and norepinephrin [nor; [;]). Dose-response curves were obtained and IC50 values were determined for each agonist. (C) In contrast, when cells were treated with increasing concentrations of a b2AR antagonist (ICI-118,551; C)before activation with an agonist (1 mM isoproterenol), intracellular levels of cAMP decreased proportionally to the antagonist’s concentration, leading to a TR-FRET signal increase. (D) LANCE Ultra signal obtained for CHO-EGFR (C) and CHO (-) cell lines. These experiments were repeated at least twice in triplicate.

430 ASSAY and Drug Development Technologies SEPTEMBER 2013

Fig. 4. EGFR transactivation detection using the AlphaScreen SureFire phospho-ERK1/2 assay. (A) On stimulation of b2AR, EGFR is transactivated and phosphorylates ERK1/2. Phospho-ERK1/2 is then detected using an anti-phospho-ERK1/2 antibody recognized by the AlphaScreen protein A beads and with a biotinylated anti-ERK1/2 captured by the Alpha streptavidin donor beads. ICI-118,551, a selective b2AR antagonist, and AG 1478, an EGFR kinase domain inhibitor, were used as controls. (B, C) EGF and isoproterenol dose-response curves on CHO-EGFR-b2AR pretreated with EGFR inhibitor AG1478 (-) or with buffer (C) as control. (D) Transactivation characterization using different receptor inhibitors. CHO-EGFR ( ), CHO-b2AR (-), and CHO-EGFR-b2AR ( ) cell lines were first treated for 2 h with buffer, AG 1478 or ICI-118, 551 and then with buffer (control), EGF (EGFR direct activation), or isoproterenol (b2AR direct activation and EGFR1 transactivation) for 10 min. S/B is the ratio of the AlphaScreen signal obtained for cells stimulated with EGF or isoproterenol to the value obtained for unstimulated cells (buffer). These experiments were repeated at least twice in triplicate. (E) Z0 determination using CHO-EGFRb2AR cells treated with 0.1 mM EGF (-), 1 mM isoproterenol (:), or buffer (C) as control. S/B, signal-to-background.

ª MARY ANN LIEBERT, INC.  VOL. 11

NO. 7  SEPTEMBER 2013

ASSAY and Drug Development Technologies 431

CARUSO ET AL.

EGFR and b2AR ligands and can, therefore, be used to study the receptor’s transactivation.

EGFR Transactivation Assay Development Using AlphaScreen SureFire pERK1/2 AlphaScreen SureFire is a platform for measuring endogenous proteins from cell lysates, and the principle of this technology is illustrated in Table 2 and Figure 4A. The AlphaScreen SureFire pERK1/2 assay kit was chosen to measure EGFR transactivation, because this reagent is very sensitive and EGFR transactivation or its direct activation leads to ERK1/2 phosphorylation.22 Receptor transactivation is influenced by numerous factors relative to cell culture and treatment and, therefore, assay parameters were optimized. Table 2 describes the EGFR transactivation protocol, and the major steps required for its optimization are shown in boldface. First, selecting the appropriate cell line was crucial for a highly performing assay. If recombinant cell lines are used to study receptor-specific effects, a cell line that does not endogenously express the receptors of interest should be selected. The CHO-K1 cell line is often used to express recombinant protein, because these cells grow quickly, are easy to transfect, and its genome is completely sequenced.31 For the EGFR-b2AR crosstalk study, wild-type CHO-K1 cells were tested for both receptor activities and selected, because they did not present any EGFR1 or b2AR activities (Figs. 2D and 3D, respectively). The use of cancer cell lines expressing the receptors of interest naturally is another alternative. Indeed, the human epithelial carcinoma cell line A431 was tested for EGFR1 transactivation by b2AR, but a higher background and large dayto-day variability was observed compared with CHO-EGFR1-b2AR recombinant cell line (data not shown). Second, we observed that cell confluence has a major impact on the transactivation results. Indeed, we tested the transactivation protocol using cells harvested at confluences ranging from 10% to 100%. Interestingly, only cells with a confluence ranging from 70% to 85% generated a reproducible transactivation response when measured using the AlphaScreen SureFire p-ERK1/2 optimized protocol (data not shown). Therefore, a frozen cell bank prepared from an 80% confluence culture was produced and compared with fresh 80% confluent cell culture using the transactivation protocol. Both conditions yielded similar results, but the assay reproducibility was much higher and the assay time considerably reduced with the frozen cells preparation. The frozen cell bank was consequently used for all subsequent transactivation assays. Once the cell culture conditions were established, optimal agonists, antagonists concentration, and stimulation time were determined for each receptor before assessing the pathway transactivation. Using the CHO-Aeq-EGFR-b2AR cell line, dose-response curves using EGF and isoproterenol were performed. EGF treatment directly activated EGFR1 (Fig. 4B), whereas a transactivation of EGFR could be observed using the b2AR agonist isoproterenol (Fig. 4C). EC50 values of 2.6 and 1.0 nM and S/B ratio of 10-fold and 5-fold were obtained for EGF and isoproterenol, respectively. Moreover, a 10 min stimulation time was determined as optimal for both EGF and

432 ASSAY and Drug Development Technologies SEPTEMBER 2013

isoproterenol treatment (data not shown). Using these data, we then confirmed that the observed signal resulted from a transactivation phenomenon using three cell lines: CHO-b2AR, CHO-EGFR, and CHO-EGFR-b2AR. As expected, only cells expressing the EGFR (CHO-Aeq-EGFR and CHO-Aeq-EGFR-b2AR cell lines) showed a 30to 40-fold increase in ERK1/2 phosphorylation when treated with EGF compared with untreated cells (Fig. 4D). When cells were treated with isoproterenol, a fourfold increase in the phosphorylation state of ERK1/2 was observed only in cells expressing both EGFR and b2AR. No response was measured in cells expressing only b2AR, suggesting that b2AR activation alone does not lead to ERK1/2 phosphorylation in CHO-K1 cells. Two inhibitors (AG 1478 and ICI-118, 551) were also evaluated to confirm that ERK1/2 phosphorylation was specific to EGFR1 transactivation. As expected, preincubating CHO-EGFR and CHO-EGFR-b2AR cell lines with EGFR1 kinase inhibitor (AG 1478) completely abolished ERK1/2 phosphorylation when treated with either EGF or isoproterenol, supporting the notion that b2AR activation requires EGFR to phosphorylate ERK1/2. The b2AR antagonist, the ICI-118,551, does not affect the direct activation of EGFR by EGF; however, inhibiting b2AR with ICI-118,551 abolished ERK1/2 phosphorylation when cells are treated with isoproterenol (Fig. 4D). Since EGFR transactivation is known to occur through activation of MMPs and Src kinase, EDTA and PP2 were used to inhibit MMPs and Src, respectively. Both inhibitors led to a signal decrease in CHO-Aeq-EGFR-b2AR cells treated with isoproterenol but not in cells treated with EGF (Supplementary Fig. S2). Taken together, these results prove that the assay conditions developed allow for the specific measurement of EGFR1 transactivation by b2AR. This assay can be used for high-throughput screening as confirmed by a Z0 value of 0.54 for cells treated with isoproterenol (Fig. 4E). Moreover, to enable high-throughput screening, both acceptor and donor beads can be added in a single step followed by a 2 h incubation step. This would simply the assay but result in minor loss of assay sensitivity based on past experience. Interestingly, when the transactivation protocol was tested using the Aequorin assay, no signal was obtained in any of the cell lines treated with isoproterenol (Supplementary Fig. S3). This suggests that EGFR1 transactivation by b2AR does not activate the EGFR1PLCg signaling pathway. This hypothesis is supported by a study which showed that EGFR docking sites pY992 and pY1173, required for PLCg activation, are not phosphorylated by the b2AR-mediated transactivation of the EGFR.32 We can postulate that transactivation by b2AR only activates a subset of EGFR signaling pathways, including ERK but not PLCg, and that different phosphorylated residue could be at play during transactivation versus direct activation of the receptor. Further studies would be required to identify these events. In conclusion, our work describes a new methodology to decipher the biochemical pathways leading to RTK activation. This new method combines three different readouts (cAMP, p-ERK1/2, and PLCg) with each optimized to measure EGFR activation and transactivation by b2AR. It also demonstrates a novel use of the Aequorin

CELL-BASED ASSAYS TO STUDY EGFR

technology for studying RTK activation. These nonwash 384-well format assays are very sensitive, because they only require as little as 10,000 cells/well for both RTK systems tested (EGFR1 and FGFR1). We also believe that this technology could be applied to other members of the RTK family signaling through PLCg and thus affecting intracellular level of Ca2 + . These Aequorin assays represent a novel screening tool which is used to identify new RTK inhibitors that are suitable for cancer treatment. Since the direct activation of EGFR and its transactivation by GPCR are closely interconnected, it is primordial to measure all these events simultaneously, especially when searching for new drugs with less resistance and side effects. To overcome drug resistance, the development of new molecules targeting specific intracellular pathways such as the transactivation of RTKs was already proposed.19,21 Our new platform could be used to identify molecules specifically inhibiting the EGFR transactivation without affecting other GPCR signaling pathways, thereby reducing side effects. The knowledge of potential off-target effects and improved understanding of the mechanism of action at early stages of drug discovery should lead to developing compounds with better efficacy and specificity in the development pipeline.

ACKNOWLEDGMENTS The authors thank G. Consentino, P. Roby, M. Loignon, L. Rihakova, and N. MacDonald (PerkinElmer BioSignal, Inc.) for critical discussions and for technical help. This work was funded by the Natural Sciences and Engineering Research Council of Canada.

DISCLOSURE STATEMENT No competing financial interests exist.

REFERENCES 1. Lemmon MA, Schlessinger J: Cell signaling by receptor tyrosine kinases. Cell 2010;141:1117–1134. 2. Daub H, Weiss FU, Wallasch C, Ullrich A: Role of transactivation of the EGF receptor in signaling by G-protein-coupled receptors. Nature 1996;379:557– 560. 3. Chen LF, Cohen EE, Grandis JR: New strategies in head and neck cancer: understanding resistance to epidermal growth factor receptor inhibitors. Clin Cancer Res 2010;16:2489–2495. 4. Kim IM, Tilley DG, Chen J, et al.: Beta-blockers alprenolol and carvedilol stimulate beta-arrestin-mediated EGFR transactivation. Proc Natl Acad Sci USA 2008;105:14555–14560. 5. Wang C, Buck DC, Yang R, Macey TA, Neve KA: Dopamine D2 receptor stimulation of mitogen-activated protein kinases mediated by cell typedependent transactivation of receptor tyrosine kinases. J Neurochem 2005;93:899–909. 6. Eguchi S, Numaguchi K, Iwasaki H, et al.: Calcium-dependent epidermal growth factor receptor transactivation mediates the angiotensin II-induced mitogenactivated protein kinase activation in vascular smooth muscle cells. J Biol Chem 1998;273:8890–8896. 7. Filardo EJ: Epidermal growth factor receptor (EGFR) transactivation by estrogen via the G-protein-coupled receptor, GPR30: a novel signaling pathway with potential significance for breast cancer. J Steroid Biochem Mol Biol 2002;80:231–238.

ª MARY ANN LIEBERT, INC.  VOL. 11

8. Filardo EJ, Quinn JA, Frackelton AR Jr, Bland KI: Estrogen action via the G protein-coupled receptor, GPR30: stimulation of adenylyl cyclase and cAMPmediated attenuation of the epidermal growth factor receptor-to-MAPK signaling axis. Mol Endocrinol 2002;16:70–84. 9. Gschwind A, Zwick E, Prenzel N, Leserer M, Ullrich A: Cell communication networks: epidermal growth factor receptor transactivation as the paradigm for interreceptor signal transmission. Oncogene 2001;20:1594–1600. 10. Ohtsu H, Dempsey PJ, Eguchi S: ADAMs as mediators of EGF receptor transactivation by G protein-coupled receptors. Am J Physiol Cell Physiol 2006;291:C1–C10. 11. Prenzel N, Zwick E, Daub H, et al.: EGF receptor transactivation by G-proteincoupled receptors requires metalloproteinase cleavage of proHB-EGF. Nature 1999;402:884–888. 12. Bokemeyer D, Schmitz U, Kramer HJ: Angiotensin II-induced growth of vascular smooth muscle cells requires an Src-dependent activation of the epidermal growth factor receptor. Kidney Int 2000;58:549–558. 13. Fischer OM, Giordano S, Comoglio PM, Ullrich A: Reactive oxygen species mediate Met receptor transactivation by G protein-coupled receptors and the epidermal growth factor receptor in human carcinoma cells. J Biol Chem 2004;279:28970–28978. 14. Manash KP, Anup KM: Tyrosine kinase–Role and significance in cancer. Int J Med Sci 2004;1:101–115. 15. Han W, Lo HW: Landscape of EGFR signaling network in human cancers: biology and therapeutic response in relation to receptor subcellular locations. Cancer Lett 2012;318:124–134. 16. Gazdar AF: Activating and resistance mutations of EGFR in non-small-cell lung cancer: role in clinical response to EGFR tyrosine kinase inhibitors. Oncogene 2009;28 Suppl 1:S24–S31. 17. Kuzumaki N, Suzuki A, Narita M, et al.: Multiple analyses of G-protein coupled receptor (GPCR) expression in the development of gefitinib-resistance in transforming non-small-cell lung cancer. PLoS One 2012;7:e44368. 18. Fischer OM, Hart S, Gschwind A, Ullrich A: EGFR signal transactivation in cancer cells. Biochem Soc Trans 2003;31(Pt 6):1203–1208. 19. Bhola NE, Thomas SE, Freilino M, et al.: Targeting GPCR-mediated p70S6K activity may improve head and neck cancer response to cetuximab. Cancer Res 2011;17:4996–5004. 20. Lin P, Sun X, Feng T, et al.: ADAM17 regulates prostate cancer cell proliferation through mediating cell cycle progression by EGFR/PI3K/AKT pathway. Mol Cell Biochem 2012;359:235–243. 21. Zhang Q, Bhola NE, Lui VW, et al.: Antitumor mechanisms of combined gastrinreleasing peptide receptor and epidermal growth factor receptor targeting in head and neck cancer. Mol Cancer Ther 2007;6:1414–1424. 22. Maudsley S, Pierce KL, Zamah AM, et al.: The beta(2)-adrenergic receptor mediates extracellular signal-regulated kinase activation via assembly of a multi-receptor complex with the epidermal growth factor receptor. J Biol Chem 2000;275:9572–9580. 23. Thomas SM, Bhola NE, Zhang Q, et al.: Cross-talk between G protein-coupled receptor and epidermal growth factor receptor signaling pathways contributes to growth and invasion of head and neck squamous cell carcinoma. Cancer Res 2006;66:11831–11839. 24. Dupriez VJ, Maes K, Le Poul E, Burgeon E, Detheux M: Aequorin-based functional assays for G-protein-coupled receptors, ion channels, and tyrosine kinase receptors. Receptors Channels 2002;8:319–30. 25. Brough SJ, Shah P: Use of Aequorin for G protein-coupled receptor hit identification and compound profiling. Methods Mol Biol 2009;552:181–198. 26. Gilchrist MA 2nd, Cacace A, Harden DG: Characterization of the 5-HT2b receptor in evaluation of Aequorin detection of calcium mobilization for miniaturized GPCR high-throughput screening. J Biomol Screen 2008;13:486–493. 27. Nakamura K, Zhou CJ, Parente J, Chew CS: Parietal cell MAP kinases: multiple activation pathways. Am J Physiol 1996;271(4 Pt 1):G640–G649. 28. Li RK, Qiang O, Liao J, Wang CH: Effects of EGFR inhibitor AG1478 in combination with celecoxib on the growth of gastric cancer cells. Sichuan Da Xue Xue Bao Yi Xue Ban 2010;41:595–599.

NO. 7  SEPTEMBER 2013

ASSAY and Drug Development Technologies 433

CARUSO ET AL.

29. O’Donnell SR, Wanstall JC: Evidence that ICI 118, 551 is a potent, highly Beta 2selective adrenoceptor antagonist and can be used to characterize Betaadrenoceptor populations in tissues. Life Sci 1980;27:671–677. 30. Tang LQ, Hong PH, Siddiqui Y, et al.: Effect of beta-adrenergic agents on intracellular potential of rabbit ciliary epithelium. Curr Eye Res 1998;17:24–30. 31. Xu X, Nagarajan H, Lewis NE, et al.: The genomic sequence of the Chinese hamster ovary (CHO)-K1 cell line. Nat Biotechnol 2011;29:735–741. 32. Drube S, Stirnweisse J, Valkova C, Liebmann C: Ligand-independent and EGF receptor-supported transactivation: Lessons from B2-adrenergic receptor signaling. Cell Signal 2006;18:1633–1646.

434 ASSAY and Drug Development Technologies SEPTEMBER 2013

Address correspondence to: Roger Bosse´, PhD PerkinElmer BioSignal, Inc. 744 William St. Montreal Que´bec H3J1R4 Canada E-mail: [email protected]