The development of small molecule and antibody inhibitors targeting the interaction of receptor tyrosine kinases (RTKs), such as epidermal growth factor receptor (EGFR), is of high pharmacological and biological interest. Unfortunately, conventional biochemical techniques using cell or tissue lysates and co-immunoprecipitation experiments to investigate EGFR dimerization are not always conclusive. Here we describe a series of technical and biological validation experiments demonstrating the utility of a proximity ligation assay (PLA)–based methodology for in situ visualization and quantification of ligand-dependent EGFR receptor dimerization in intact cells. Using the PLA approach combined with a universally applicable epitope tagging strategy, we detected EGFR dimers in cells transiently co-expressing FLAG-tagged and MYC-tagged human EGFRs. Our data strongly suggest that PLA can be used to detect ligand-dependent EGFR dimerization and this signal is generated in a protein interaction–based manner, rather than solely due to proximity of target proteins. This application represents a generalized RTK expression strategy for protein-interaction analysis in a transient expression system where antibody epitopes are not known or not unique enough to discriminate between interaction partners. This assay also holds promise as a general RTK dimerization screening tool in tissue specimens to identify potential dimerization inhibitors with clinical relevance.

Introduction

Conventional biochemical approaches to assess receptor tyrosine kinase (RTK) dimerization in vitro rely on immunoprecipitation (IP) and chemical crosslinking techniques. However, these strategies are not particularly sensitive, do not preserve physiological or cellular compartmentalization, and are not easily amenable to high-throughput screening. Moreover, the lack of suitable antibodies toward the most common RTK deletion and point mutants precludes detailed dimerization analysis by traditional biochemical methods.

The epidermal growth factor receptor (EGFR) is a prototypical member of the ErbB family of RTKs and is aberrantly overexpressed or activated in numerous human cancers (1). Receptor activation is mediated by ligand binding to the extracellular domain, leading to intramolecular conformational rearrangements that facilitate homo- or heterodimerization with ErbB family members. Dimerization promotes intrinsic kinase domain activation, resulting in phosphorylation on specific tyrosine residues within the cytoplasmic tail and subsequent stimulation of intracellular signaling pathways (2,3). Tumors harboring overexpressed EGFR and ErbB members have the potential for oncogenic signaling and deregulated cellular behavior. In recent years, therapeutic focus has shifted to abrogating aberrant EGFR function at the critical event of dimerization. Clearly, there exists an urgent need for adequate methodologies to evaluate EGFR dimerization from a biological and therapeutic perspective.

In the present report, we describe our work developing and standardizing an application of proximity ligation strategy (4,5) to allow EGFR homo- and heterodimer detection and quantification in a transient expression system. Furthermore, we utilized this system to confirm that dimerization signal is generated in a protein interaction–based manner, rather than solely due to proximity of target proteins. Additionally, this is to our knowledge the first application of in situ PLA in a transient expression system to detect protein interactions.

Methods and materials

Expression constructs

Human wild-type EGFR-encoding cDNA (lacking stop codon) was PCR-amplified from the pGEM7zf(-):EGFRwt (Promega, Madison, WI, USA) shuttle vector (primer sequences available in Supplementary Table S1) and cloned into pCR2.1-TOPO using the TOPO-TA system (Invitrogen, Burlington, ON, Canada) for sequence verification. EGFRwt-pCR2.1-TOPO was SacII-digested and blunt-end–cloned into the HindIII site of pLPCX expression vector (Clontech, Mountain View, CA, USA) to generate EGFR-pLPCX. To prepare FLAG and MYC epitope tags, oligonucleotides (Integrated DNA Technologies, Coralville, IA, USA) FLAG sense/antisense and MYC sense/antisense (Supplementary Table S1), respectively, were annealed, phosphorylated, and blunt-end–cloned into blunted EcoRI site of EGFR-pLPCX to generate EGFR-FLAG-pLPCX and EGFR-MYC-pLPCX. Untagged EGFR was expressed from a pcDNA3.1/Zeo (+) construct (Invitrogen). Mammalian expression constructs encoding MYC tag translational fusions of fibroblast growth factor receptor-1 (FGFR1) and ephrin receptor B1 (EPHB1) were generated from hOR Feome v5.1 cDNA shuttled into pcDNA3.2/myc-DEST using Gateway (Invitrogen) cloning technology (SIDNET, The Hospital for Sick Children). Y246D point mutant construct was generated using EGFRwt-FLAG-pLPCX as template and mutagenized with a site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA). Nucleotide sequencing was performed to confirm the mutagenesis.

Cell culture and transfection

Chinese hamster ovary (CHOK1) cells were cultured in 1× F-12K nutrient medium containing 10% FCS, 100 U/mL penicillin, 100 µg/mL streptomycin, and 250 ng/mL amphotericin B (Wisent, St-Bruno, Quebec, Canada). For immunofluorescence experiments, cells were grown to 70% confluence on 18 × 18-mm glass coverslips (Fisher Scientific, Ottawa, ON, Canada) placed in six-well plates (BD Biosciences, Mississauga, ON, Canada) and transfected with indicated expression constructs using Fugene HD (Roche, Mississauga, ON, Canada), according to the manufacturer's instructions. After 6 h, the transfection medium was changed to serum-free medium. Cells were incubated another 18 h and subsequently washed twice in ice-cold PBS and fixed for 10 min with 4% paraformaldehyde in PBS at room temperature. For immunoblotting experiments, cells were grown to 70% confluence in six-well plates and transfected as in immunofluorescence experiments, and serum-starved cells were stimulated with 100 ng/mL EGF (Chemicon; Billerica, MA, USA).