Ingrid Remy and Stephen

Two strategies are described for detecting constitutive or induced protein–protein interactions in intact mammalian cells; these strategies are based on oligomerization domain-assisted complementation of rationally designed fragments of the murine enzyme dihydrofolate reductase (DHFR; EC 1.5.1.3). We describe a dominant clonal-selection assay of stably transfected cells expressing partner proteins FKBP (FK506 binding protein) and FRAP (FKBP–rapamycin binding protein) fused to DHFR fragments and show a rapamycin dose-dependent survival of clones that requires '25 molecules of reconstituted DHFR per cell. A f luorescence assay also is described, based on stoichiometric binding of f luorescein-methotrexate to reconstituted DHFR in vivo. Formation of the FKBP–rapamycin–FRAP complex is detected in stably and transiently transfected cells. Quantitative rapamycin dose-dependence of this complex is shown to be consistent with in vitro binding and distinguishable from a known constitutive interaction of FKBP and FRAP. We also show that this strategy can be applied to study membrane protein receptors, demonstrating dose-dependent activation of the erythropoietin receptor by ligands. The combination of these clonal-selection and fluorescence assays in intact mammalian cells makes possible selection by simple survival, f low cytometry, or both. High-throughput drug screening and quantitative analysis of induction or disruption of protein–protein interactions are also made possible. Many processes in biology are mediated by noncovalently associated multienzyme complexes (1, 2). Examples include cellular machineries for transcription, translation, and metabolic or signal-transduction pathways. Much of modern biological research is concerned with identifying proteins involved in these cellular processes and with determining their functions and how, when, and where they interact with other proteins involved in specific biochemical pathways. The yeast twohybrid system is a robust method to study protein–protein interactions in a specific cellular context (3). However, quantitative detection of protein–protein interactions in intact cells remains a significant experimental challenge. We have developed a strategy to study protein–protein interaction in vivo based on protein-fragment complementation assays (PCAs; refs. 4–6). The basic principles of this approach were pioneered by Johnsson and Varshavsky (7). Similar in vivo protein–protein interaction assays also have been described (8, 9). Here, we report PCAs, based on the murine enzyme dihydrofolate reductase (DHFR), that allow for quantitative characterization of protein–protein interactions in mammalian cells. Our goal has been to develop a strategy that allows detection of elemental protein–protein interactions in vivo and in appropriate biological contexts, such as specific cell types or within particular cellular compartments. We show how the DHFR PCA can be used simultaneously in a strategy for dominant clonal selection and as a universal in vivo f luorescence assay for quantitative pharmacological analysis of protein and protein–small molecule binding. MATERIALS AND METHODS DNA Constructs. Complementary oligonucleotides containing restriction sites and coding for a 10-aa flexible polypeptide linker consisting of (GlyzGlyzGlyzGlyzSer)2 were inserted into the eukaryotic expression vector pMT3 (10). Fragments of DHFR (F[1,2] and F[3]) were amplified by PCR from the constructs Z-F[1,2:Phe31Ser] and Z-F[3] (6) and introduced at the 39 end of the flexible linker. F[1,2] corresponds to amino acids 1–105, and F[3] corresponds to amino acids 106–186 of murine DHFR. FKBP (the FK506 binding protein) and FRB (FKBP–rapamycin binding domain of FRAP; FRAP is the FKBP–rapamycin binding protein; ref. 11] were amplified by PCR with appropriate eukaryotic initiation regions (12) from pNH1 (13) and pMRS315 (14), respectively. They were subcloned at the 59 end of the flexible linker, resulting in the following constructs: FRB–F[1,2] and FKBP–F[3]. As negative controls for interactions, we generated constructs ZIP–F[1,2] and ZIP–F[3] by using PCR fragments corresponding to residues 235–281 of the GCN4 leucine zipper. For the erythropoietin receptor (EpoR) fusion constructs, a fragment comprising the extracellular and transmembrane-domain of the receptor was generated by PCR amplification from murine EpoR and fused to DHFR F[1,2] and F[3] via a 5-aa flexible linker, resulting in EpoR(1–270)–F[1,2] and EpoR(1– 270)–F[3]. Creation of Stable Cell Lines (Survival Selection). CHO DUKX-B11 (15) cells were split 24 h before transfection at 1 3 105 in 12-well plates in a-MEM (Life Technologies; Grand Island, NY), which was enriched with dialyzed FBS (HyClone) and supplemented with 10 mgyml of adenosine, deoxyadenosine, and thymidine (Sigma). Cells were transfected with the different constructs by using Lipofectamine reagent (Life Technologies) according to the manufacturer’s instructions. At 48 h after the beginning of the transfection, cells were split at '5 3 104 in 6-well plates in selective medium consisting of a-MEM enriched with dialyzed FBS but without addition of nucleotides. Rapamycin (Calbiochem), for FRByFKBP transfection, or Epo (R. W. Johnson Pharmaceutical Research Institute, Raritan, NJ), for EpoRyEpoR transfection, was added to the cells at a final concentration of 10 nM and 2 nM, respectively. After 5 days, a dozen surviving colonies were The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked ‘‘advertisement’’ in accordance with 18 U.S.C. §1734 solely to indicate this fact. PNAS is available online at www.pnas.org. This paper was submitted directly (Track II) to the Proceedings office. Abbreviations: DHFR, dihydrofolate reductase; PCA, proteinfragment complementation assay; FKBP, FK506 binding protein; FRAP, FKBP–rapamycin binding protein; FRB, FKBP–rapamycin binding domain of FRAP; Epo, erythropoietin; EpoR, Epo receptor; fMTX, fluorescein-conjugated methotrexate; FACS, fluorescenceactivated cell sorter. *To whom reprint requests should be addressed. e-mail: stephen. [email protected].