Construction of a novel detection system for protein–protein interactions using yeast G-protein signaling Nobuo Fukuda 1 , Jun Ishii 2 , Tsutomu Tanaka 2 , Hideki Fukuda 2 and Akihiko Kondo 1 1 Department of Chemical Science and Engineering, Graduate School of Engineering, Kobe University, Japan 2 Organization of Advanced Science and Technology, Kobe University, Japan Protein–protein interactions are essential for normal cellular function, and numerous studies have provided important insight into the molecular mechanisms underlying these interactions. In particular, develop- ment and use of the yeast two-hybrid (Y2H) system has greatly facilitated the study of protein–protein interactions. In order to exhaustively identify protein interaction pairs, including membrane-associated pro- teins, the SOS and Ras recruitment systems (SRS or RRS) using the Ras-signaling pathway in yeast cells as the readout have proven to be successful [1,2]. Mem- brane-associated proteins, which constitute approxi- mately 40% of the total cellular proteins, include many important drug receptors, channels and enzymes [3]. In the SRS and RRS systems, temperature-sensi- tive mutant strains are required for detection of pro- tein–protein interactions. If the proteins physically interact, the Ras-signaling pathway is activated, allow- Keywords G-protein signaling; membrane localization of Gc subunit; protein–protein interaction; yeast two-hybrid system Correspondence A. Kondo, Department of Chemical Science and Engineering, Graduate School of Engineering, Kobe University, 1-1 Rokkodaicho, Nada-ku, Kobe 657-8501, Japan Fax ⁄ Tel: +81 78 803 6196 E-mail: akondo@kobe-u.ac.jp (Received 18 December 2008, revised 18 February 2009, accepted 3 March 2009) doi:10.1111/j.1742-4658.2009.06991.x In the current study, we report the construction of a novel system for the detection of protein–protein interactions using yeast G-protein signaling. It is well established that the G-protein c subunit (Gc) is anchored to the inner leaflet of the plasma membrane via lipid modification in the C-termi- nus, and that this localization of Gc is required for signal transduction. In our system, mutated Gc (Gc cyto ) lacking membrane localization ability was genetically prepared by deletion of the lipid modification site. Complete disappearance of G-protein signal was observed when Gc cyto was expressed in the cytoplasm of yeast cells from which the endogenous Gc gene had been deleted. In order to demonstrate the potential use of our system, we utilized the Staphylococcus aureus ZZ domain and the Fc portion of human immunoglobulin G (IgG) as a model interaction pair. To design our detec- tion system for protein–protein interaction, the ZZ domain was altered so that it associates with the inner leaflet of the plasma membrane, and the Fc part was then fused to Gc cyto . The Fc–Gc cyto fusion protein migrated towards the membrane via the ZZ–Fc interaction, and signal transduction was therefore restored. This signal was successfully detected by assessing growth inhibition and transcription in response to G-protein signaling. Finally, several Z variants displaying affinity constants ranging from 8.0 · 10 3 to 6.8 · 10 8 m )1 were prepared, and it was demonstrated that our system was able to discriminate subtle differences in affinity. In conclusion, our system appears to be a reliable and versatile technique for detection of protein–protein interactions, and may prove useful in future protein inter- action studies. Abbreviations EGFP, enhanced green fluorescent protein; RRS, Ras recruitment system; SRS, Sos recruitment system; Z I31A , single-site mutant of the Z domain by altering isoleucine at position 31 to alanine; Z K35A , single-site mutant of the Z domain by altering the lysine at position 35 to alanine; Z WT , wild-type Z domain derived from the B domain of Staphylococcus aureus protein A. 2636 FEBS Journal 276 (2009) 2636–2644 ª 2009 The Authors Journal compilation ª 2009 FEBS ing yeast cells to grow at 37 °C. Although this system is advantageous for analysis of membrane-associated proteins, the yeast growth rate using this system is slow given that the optimal temperature for yeast repli- cation is 30 °C. To establish a system that allows rapid identification of protein–protein interactions, we focused on the yeast G-protein signaling pathway. The yeast G-pro- tein signaling pathway is a well characterized pathway that is activated via pheromone stimulation. Phero- mone stimulation leads to activation of heterotrimeric G-proteins comprising Gpa1 (Ga), Ste4 (Gb) and Ste18 (Gc) through the G-protein-coupled receptor (Fig. 1A). The activated G-protein subsequently disso- ciates into Ga and Gbc complex subunits, and the Gbc complex induces activation of the mitogen-acti- vated protein kinase cascade. The amplified signal results in various cellular responses, including global changes in transcription, growth arrest in the G 1 phase, and polarized morphogenesis for mating. One significant advantage of using G-protein signaling for the detection of protein–protein interactions is that the assays are undertaken at 30 °C and thus yeast cells are able to rapidly grow at their preferred temperature. Several methods of detection of protein–protein interactions have been developed using G-protein sig- naling. Medici et al. established the Gpa1–Gas chime- ric system in which a receptor is fused to protein X and Gpa1–Gas is fused to protein Y, restoring G-pro- tein signaling in response to the protein X–protein Y interaction [4]. Subsequently, Ehrhard et al. reported the use of a Gbc interfering system, in which the inter- action between protein X fused to Gc with the integral membrane protein Y disturbed contact of Gb with its effectors and thus inhibited G-protein signaling [3]. These assays can be undertaken at 30 °C and are therefore suitable for rapid yeast growth, unlike the SRS and RRS methods, which required temperature- sensitive mutant strains. Furthermore, these systems can be applied to the study of biologically important membrane-bound proteins. Unfortunately, however, previously reported systems using G-protein signaling resulted in high background signals, making it difficult to distinguish between subtle differences in affinity, and have therefore been considered unfavorable for extensive screening processes. In the current study, we established a technique for the successful identification of protein–protein interac- tions using yeast G-protein signaling. We previously utilized the G-protein-coupled receptor assay system, which involves growth inhibition following G 1 arrest and transcription of the enhanced green fluorescent protein (EGFP) reporter gene to detect protein–protein interactions [5]. Signal transduction defects resulting from dissociation of the Gc subunit from the mem- brane require localization of the Gbc complex to the plasma membrane through the lipidated Gc subunit [6], and our method to detect protein–protein interac- tions is based on this finding. The sequence encoding target protein ‘binder X’ is genetically fused to a Gc gene from which the lipidation sites (Gc cyto ) have been deleted. The gene encoding the binder X–Gc cyto fusion protein replaces the STE18 gene, which encodes intact Gc. Then the lipidation motif is genetically introduced to ‘binder Y’ and co-expressed with the binder X–Gc- cyto protein. Binder X–Gc cyto protein is expressed in the cytosol and the lipidated binder Y protein is local- ized to the plasma membrane. As a result, signal trans- duction did not occur. When binder X and binder Y interact with each other, the binder X–Gc cyto fusion protein becomes localized to the plasma membrane and thus activates G-protein signaling (Fig. 1C). In this study, we selected the Fc portion of human IgG and the ZZ domain derived from Staphylococ- cus aureus protein A as the model interaction pair (Fig. 2), and demonstrated protein–protein interactions using growth inhibition and transcription assays. Use C Pheromone A Receptor B EGFP gene transcription Signal transduction Signal transduction Growth arrest (growth inhibition assay) (transcription assay) Fig. 1. Schematic outline of the experimental design. (A) The wild- type Gc subunit induces pheromone-stimulated signaling. (B) Engi- neered Gc lacking membrane-localization ability (Gc cyto ) leads to a significant defect in G-protein signaling. As a result, the Gb and Gc cyto (Gbc cyto ) complex is released into the cytosol following dissociation from Ga due to ablation of plasma membrane associa- tions. (C) Protein–protein interaction re-establishes pheromone- stimulated signaling. Interaction between protein X fused to Gc cyto and protein Y anchored to the plasma membrane results in migra- tion of Gbc cyto to the inner leaflet of the plasma membrane. In our system, a transcription assay using the EGFP reporter gene fused to the pheromone-inducible FIG1 gene allows positive selection. A growth inhibition assay based on cell-cycle arrest permits negative selection. The conditions used for this system are suitable for yeast cell growth (30 °C). N. Fukuda et al. Detection system for protein–protein interactions FEBS Journal 276 (2009) 2636–2644 ª 2009 The Authors Journal compilation ª 2009 FEBS 2637 of this assay system also resulted in very low back- ground signal. Results and Discussion General strategy The aim of this study was to establish a rapid and reli- able method for the detection of protein–protein inter- actions using yeast G-protein signaling. In our system, protein–protein interactions were detected utilizing the knowledge that the Gc subunit localizes to the inner leaflet of the plasma membrane and that this localiza- tion is required for G-protein signaling (Fig. 1). For- mation of Gc mutants by deletion of their lipidation sites completely interrupts G-protein signaling [6], thus we expected that our system would permit a more accurate determination of protein–protein interactions. We chose the Fc portion of human IgG and the ZZ domain derived from protein A [7–9], and an STE18 gene encoding the yeast Gc subunit in which the lipi- dation sites had been mutated (Gc cyto ), as the key components of our system. The Fc was genetically fused to the C-terminus of Gc cyto (Gc cyto –Fc) and the lipidation motif was genetically added to the C-termi- nus of the ZZ domain (ZZ mem ). Gc cyto –Fc and ZZ mem were then co-expressed a yeast strain lacking endoge- nous STE18. Interaction between Fc and ZZ mem would then result in localization of Gc cyto to the plasma membrane, and signal transduction in response to pheromone stimulation will occur (Fig. 1C). Construction of a yeast strain lacking endogenous Gc In order to prepare a host strain that would accept the mutated Gc (Gc cyto ), which would in turn result in its altered localization to the membrane and therefore cause a strong defective signal (Fig. 1B), we con- structed an endogenous Gc-defective strain termed BWG2118 by deletion of the STE18 gene. BWG2118 was derived from the MC-F1 yeast strain that induces expression of the EGFP reporter gene in response to G-protein signaling (Table 1). To confirm STE18 gene deletion in BWG2118, pheromone-dependent growth inhibition (halo) and transcription assays were carried out. For growth inhibition assays, cells were plated and then exposed to synthetic pheromone spotted onto filter disks. MC-F1, in which endogenous Gc is intact, produced a clear halo in response to G-protein signal- ing, but BWG2118, in which endogenous Gc is defec- tive, did not exhibit a clear halo due to loss of signaling ability (Fig. 3A). For transcription assays, expression of the EGFP reporter gene under the con- trol of the pheromone-inducible FIG1 promoter was analyzed by flow cytometry. MC-F1 exhibited high flu- orescence as a result of signaling, but BWG2118 did not show EGFP reporter fluorescence even after the addition of pheromone (Fig. 3B). The fluorescence intensity for BWG2118 appeared similar to that for strain BY4741, which was the original BWG2118 strain and does not encode the EGFP reporter gene (data not shown). These results demonstrate that G-protein signaling was interrupted due to the absence of the STE18 gene, and that a yeast strain BWG2118, lacking endogenous Gc, had been successfully constructed. Co-expression of ZZ mem and Gc cyto –Fc proteins in an endogenous Gc-defective yeast strain To demonstrate detection of protein–protein interac- tions using mutated Gc, we used the ZZ domain and the Fc portion as the model pair in this system. The lipidation-defective Gc mutant (Gc cyto ) was con- structed by deleting five amino acids from the C-termi- nus, and then fusing Gc cyto with Fc (Fig. 2A). Alternatively, the ZZ domain, which demonstrates high specific affinity for Fc, was genetically fused to the lipidation motif sequence of yeast Gc at the Table 1. Yeast strains used in this study. Strain Genotype Reference BY4741 MATa his3D1 ura3D0 leu2D0 met15D0 [11] MC-F1 BY4741 P fig1 -FIG1-EGFP Ishii et al., (unpublished results) BWG2118 MC-F1 ste18D::kanMX4 Present study BZG2118 MC-F1 ste18D::kanMX4-P PGK -ZZ mem Present study BFG2118 BWG2118 his3D::URA3-P ste18 -Gc cyto -Fc Present study BZFG2118 MC-F1 ste18D::kanMX4-P PGK -ZZ mem his3D::URA3-P ste18 -Gc cyto -Fc Present study BFG2Z18-WT MC-F1 ste18D::kanMX4-P PGK -Z WT, mem his3D::URA3-P ste18 -Gc cyto -Fc Present study BFG2Z18-K35A MC-F1 ste18D::kanMX4-P PGK -Z K35A, mem his3D::URA3-P ste18 -Gc cyto -Fc Present study BFG2Z18-I31A MC-F1 ste18D::kanMX4-P PGK -Z I31A, mem his3D::URA3-P ste18 -Gc cyto -Fc Present study Detection system for protein–protein interactions N. Fukuda et al. 2638 FEBS Journal 276 (2009) 2636–2644 ª 2009 The Authors Journal compilation ª 2009 FEBS C-terminus (ZZ mem ; ZZ-SNSVCCTLM-COOH; Fig. 2A) [6]. The strains in which ZZ mem and the Gc cyto –Fc fusion genes were integrated into BWG2118 were termed BZG2118 (ZZ mem ), BFG2118 (Gc cyto –Fc) and BZFG2118 (ZZ mem ⁄ Gc cyto –Fc) (Table 1). Expres- sion of ZZ mem proteins in BZG2118 (lane 2) and BZFG2118 (lane 4), or of Gc cyto –Fc fusion proteins in BFG2118 (lane 3) and BZFG2118 (lane 4) was con- firmed by Western blot analysis using anti-protein A or anti-human IgG (Fig. 4A). Migration of mutated Gc to the plasma membrane by protein–protein interaction restores signal transduction in an endogenous Gc-defective yeast strain To test our hypothesis that the mutated Gc (Gc cyto ) migrates to the plasma membrane and restores signal transduction via protein–protein interaction, we inves- tigated whether the endogenous Gc-defective yeast strain expressing the ZZ mem protein or the Gc cyto –Fc fusion protein induced signal transduction in growth inhibition assays. In order to achieve this, cells were plated and then exposed to synthetic pheromone spot- ted onto filter disks (Fig. 4B). The endogenous Gc-defective yeast strain (BWG2118) and the cells expressing ZZ mem or Gc cyto –Fc (BZG2118 or BFG2118) did not show halo formation even after pheromone stimulation (Fig. 4B, panels 1–3). How- ever, the yeast strain BZFG2118, which expresses both ZZ mem and Gc cyto –Fc, did show a clear halo in response to pheromone induction, demonstrating that co-expression of ZZ mem and Gc cyto –Fc was able to restore signal transduction (Fig. 4B, panel 4). We also prepared a yeast strain expressing ZZ without the lipidation motif in place of ZZ mem (termed BFG2118 ⁄ ZZ), which co-expressed Gc cyto –Fc and ZZ without the lipidation motif, as a negative control strain. As ZZ and Fc are known to interact in the cytosol [9], the BFG2118 ⁄ ZZ strain did not exhibit cell-cycle arrest in the halo assay (data not shown). In addition, as a positive control, yeast cells express- ing the lipidation motif attached to Gc cyto –Fc (Gc cyto –Fc mem ) formed a clear halo in response to pheromone induction, as expected (data not shown). These results demonstrate that the mutated Gc (Gc cyto ) strain utilized in this study had completely lost its membrane associations; however, recruitment of Gc cyto to the membrane following interaction with ZZ and Fc recovered G-protein signaling. These results suggest that interactions between membrane proteins or cytoplasmic proteins modified to contain membrane lipidation motifs and cytoplasmic proteins may be detected using our system. As the growth inhibition assay based on cell-cycle arrest allowed for negative selection, our system may also be success- fully used in high-throughput screening of signal- defective mutants to determine the specific amino acids required for protein–protein interactions. Evaluation of the affinity constant via transcription assays using the EGFP fluorescence reporter gene To corroborate the results of the growth inhibition assay, we performed reporter transcription assays. As shown in Fig. 4C, co-expression of ZZ mem and Gc cyto – Fc (BZFG2118) resulted in remarkably high fluores- cence following transcriptional activation of the EGFP reporter gene. The fluorescence intensity was equiva- lent to that of the MC-F1 positive control strain shown in Fig. 3B. In contrast, BFG2118, which expressed Gc cyto –Fc without ZZ mem , did not show reporter expression, and the fluorescence intensity was equivalent to that of the negative control strain BWG2118 (Fig. 3B). These results demonstrate that our system resulted in very low background signal and therefore confers a significantly high signal-to-noise (S ⁄ N) ratio in the detection of protein–protein inter- actions. Detection of interactions in the absence of A BC Fig. 2. Schematic outline of gene construction. (A) Structural fea- tures of the yeast endogenous Gc gene (STE18), and design of the Gc–Fc fusion gene excluding the lipidation motif (Gc cyto –Fc) and the lipidation motif attached to the ZZ gene (ZZ mem ). (B) Plasmid map for integration of the ZZ mem gene into the yeast chromosome. (C) Plasmid map for integration of the Gc cyto –Fc gene into the yeast chromosome. N. Fukuda et al. Detection system for protein–protein interactions FEBS Journal 276 (2009) 2636–2644 ª 2009 The Authors Journal compilation ª 2009 FEBS 2639 background signal generation was also shown for the growth inhibition assay (Fig. 4B). Transcription assays using the EGFP reporter allow the quantitative assess- ment of changes in G-protein signaling and high- throughput selection of positive interaction pairs by flow cytometric screening [5]. Numerous previous studies have reported detection of protein–protein interactions; however, few methods have allowed evaluation of affinity constant. To assess the correlation between the affinity constant and the fluorescence intensity, we prepared several partners for Fc (Z WT , 5.9 · 10 7 m )1 ;Z K35A , 4.6 · 10 6 m )1 ;Z I31A , 8.0 · 10 3 m )1 ) instead of ZZ (6.8 · 10 8 m )1 ) [10] (where Z WT is the wild-type Z domain derived from the B domain of Staphylococcus aureus protein A, Z K35A is a single-site mutant of the Z domain by alter- ing the lysine at position 35 to alanine, and Z I31A is a single-site mutant of the Z domain by altering isoleu- cine at position 31 to alanine), and introduced them into yeast chromosomes (BFG2Z18-WT, BFG2Z18- K35A and BFG2Z18-I31A, as shown in Table 1). Expression of ZZ mem and the ZZ mem variants was confirmed by Western blot analysis (Fig. 5A), and reporter transcription assays were performed for each strain (Fig. 5B). The fluorescence intensities of the strains were obviously altered according to the affinity constants of the Fc partners. It was notable that the relatively faint interaction between Fc and Z I31A , whose affinity constant was 8.0 · 10 3 m )1 , could be successfully detected. Furthermore, we identified a log- arithmic proportional relationship between fluores- cence intensity and affinity constant (Fig. 5C). Such accurate quantitative capability may be helpful for discrimination of doubtful interaction candidates using our system. (kDa) 14 1 A B C a b c 2 3 4 14.3 36.9 (ZZ mem ) (Gγ cyto -Fc) 42.0 (β-actin) 3 4 3 4 1 2 3 4 Fig. 4. Restoration of signal transduction following interaction between ZZ mem and Gc cyto –Fc. (A) Western blot analyses were per- formed using the following primary antibodies: (a) anti-protein A for the ZZ domain, (b) anti-IgG for Fc, and (c) anti-b-actin as the loading control. (B) Halo bioassays were performed with 10 ng of synthetic a-factor pheromone spotted onto filter disks. (C) Transcription assays were performed using flow cytometric EGFP fluorescence analysis. The histogram plots show the analytical data for 10 000 cells. ‘1’ indicates BWG2118 (negative control strain), ‘2’ indicates BZG2118 (the constructed strain expressing ZZ mem ), ‘3’ indicates BFG2118 (the constructed strain expressing Gc cyto –Fc), and ‘4’ indi- cates BZFG2118 (the constructed strain expressing both ZZ mem and Gc cyto –Fc). A B Fig. 3. Confirmation of signal response in the endogenous Gc-defective yeast strain. (A) Halo bioassays were performed with 10 ng of synthetic a-factor pheromone spotted onto filter disks. (B) Transcription assays were performed by flow cytometric EGFP fluo- rescence analysis. The histogram plots show the analytical data for 10 000 cells. ‘1’ indicates BWG2118 (the constructed ste18D strain), and ‘2’ indicates MC-F1 (the STE18-intact strain). Detection system for protein–protein interactions N. Fukuda et al. 2640 FEBS Journal 276 (2009) 2636–2644 ª 2009 The Authors Journal compilation ª 2009 FEBS In conclusion, we have established a novel detection system based on G-protein signaling for detection of protein–protein interactions, using a mutated Gc that lacks membrane-localization ability. Our assay can be performed under conditions suitable for maximal yeast cell growth, and the effects can be assessed in terms of transcription (positive selection) and growth inhibition assays (negative selection). In addition, our system is a reliable, quantitative technique that largely avoids background signals. As a result, we were able to evalu- ate a wide range of affinity constants from 8.0 · 10 3 to 6.8 · 10 8 m )1 . We suggest that our system can be uti- lized as a reliable and versatile system for detection of protein–protein interactions using G-protein signaling. Experimental procedures Strains and media Details of Saccharomyces cerevisiae BY4741 [11], MC-F1 (J. Ishii, M. Moriguchi, S. Matsumura, K. Tatematsu, S. Kuroda, T. Tanaka, T. Fujiwara, H. Fukuda & A. Kondo, unpublished results) and other constructed strains used in this study and their genotypes are given in Table 1. MC-F1 derived from BY4741 was engineered to express the EGFP fusion gene in response to a-factor pher- omone induction using the pheromone-inducible FIG1 gene. The yeast strains were grown in YPD medium containing 1% w ⁄ v yeast extract, 2% peptone and 2% glucose, or in SD medium without uracil (SD-Ura) containing 0.67% yeast nitrogen base without amino acids (Becton Dickinson and Company, Franklin Lakes, NJ, USA), 2% glucose, 20 mgÆL )1 histidine, 30 mgÆL )1 leucine and 30 mgÆL )1 methionine. Agar (2% w ⁄ v) was added to the media described above to produce YPD and SD-Ura agar. Construction of plasmids for yeast chromosome substitution Plasmids used for deletion of STE18 gene by substitution of the kanMX4 gene (G418 resistance gene) on the yeast chromosome were constructed by amplifying the fragment encoding the upstream region of STE18 (STE18p, STE18 promoter region) from MC-F1 genomic DNA using prim- ers 1 and 2 (Table 2). This fragment was then inserted into the XhoI site of pGK426 (J. Ishii, K. Izawa, S. Matsumura, K. Wakamura, T. Tanino, T. Tanaka, C. Ogino, H. Fukuda & A. Kondo, unpublished results), yielding plasmid pGK426-GP. The fragment encoding the down- stream region of STE18 (STE18t, STE18 terminator region) was amplified from MC-F1 genomic DNA using primers 3 and 4 (Table 2), and inserted into the BamHI– EcoRI sites of pGK426-GP yielding plasmid pGK426-GPT. The fragment containing kanMX4 was amplified from pUG6 (EUROSCARF, Frankfurt, Germany) [12] using primers 5 and 6 (Table 2), and inserted into the XhoI–SalI site of pGK426-GPT yielding plasmid pGK426-GPTK. The plasmid used for integration of the ZZ domain fused to the lipidation motif gene (ZZ mem ) at the STE18 locus of the yeast chromosome was constructed by amplifying the fragment encoding the ZZ domain from pMWIZ1 [13] Fluorescence intensity 1 2 3 4 5 1 10 100 1000 Fluorescence intensity 10 3 10 5 10 7 10 9 Affinity constant [ M –1 ] 2 3 4 5 (kDa) 14.3 8.1 36.9 42.0 (ZZ mem ) (β-actin) (Z mem ) (Gγ cyto -Fc) 1A B C a b c 2345 Fig. 5. Quantitative analysis of the signal responses and interaction strength. (A) Western blot analyses were performed using the fol- lowing primary antibodies: (a) anti-protein A for the ZZ domain, (b) anti-IgG for Fc, and (c) anti-b-actin as the loading control. (B) Flow cytometric EGFP fluorescence analysis. (C) Logarithmic plots of flu- orescence intensity against the affinity constants. ‘1’ indicates BFG2118 (negative control strain), ‘2’ indicates BFG2Z18-I31A (the constructed strain expressing Z I31A ), ‘3’ indicates BFG2Z18-K35A (the constructed strain expressing Z K35A ), ‘4’ indicates BFG2Z18- WT (the constructed strain expressing Z WT ), and ‘5’ indicates BZFG2118 (the strain expressing ZZ mem ). Standard errors of three independent experiments are presented. N. Fukuda et al. Detection system for protein–protein interactions FEBS Journal 276 (2009) 2636–2644 ª 2009 The Authors Journal compilation ª 2009 FEBS 2641 using primers 7 and 8 (Table 2), and inserting it into the SalI–BamHI site of pGK426, yielding plasmid pUMZZ. The fragment encoding the PGK1 promoter (PGK5¢), the ZZ mem gene and the PGK1 terminator (PGK3¢) was ampli- fied from pUMZZ using primers 9 and 10 (Table 2), and inserted into the XhoI site of pGK426-GPTK, yielding plas- mid pUMGPT-ZZK (Fig. 2). The plasmid used for integration of the Gc cyto –Fc gene at the HIS3 locus of the yeast chromosome was constructed by amplifying the fragment encoding STE18p and Gc delet- ing the lipidation sites (Gc cyto ) from MC-F1 genomic DNA using primers 11 and 12 (Table 2), and inserted into the XhoI–BamHI sites of pGK426, yielding plasmid pUMGP- GcM. The fragment encoding Fc was amplified from pUF318-Fc [14] using primers 13 and 14 (Table 2), and inserted into the BamHI–EcoRI site of pUMGP-GcM, yielding plasmid pUMGP-GcMFc. A fragment encoding the HIS3 terminator region (HIS3t) was amplified from MC-F1 genomic DNA using primers 15 and 16 (Table 2), and inserted into the NotI–SacI sites of pUMGP-GcMFc, yielding plasmid pUMGP-GcMFcH (Fig. 2). The plasmid used for integration of the Z mem gene at the STE18 locus of the yeast chromosome was constructed by amplifying the fragment encoding the Z domain from pMWIZ1 using primers 17 and 18 (Table 2), and inserted into the SalI–BamHI sites of pGK426 yielding plasmid pUMZ-WT. To prepare single amino acid-substituted Z variants, the following plasmids were constructed from pUMZ-WT using the Quick-Change method (Stratagene, La Jolla, CA, USA). For Z K35A and Z I31A , plasmids pUMZ-K35A and pUMZ-I31A were constructed using primers 19 and 20 and primers 21 and 22, respectively. The fragment encoding PGK5¢, the Z mem genes (wild-type, K35A and I31A) and PGK3¢ were amplified from pUMZ- WT, pUMZ-K35A and pUMZ-I31A, respectively, using primers 23 and 24 (Table 2), and inserted into the XhoI site of pGK426-GPTK, yielding plasmids pUMGPT-ZK-WT, pUMGPT-ZK-K35A and pUMGPT-ZK-I31A. Construction of yeast strains The strains used in this study are described in Table 1. The genes were introduced into yeast cells using the lithium acetate method [15]. Substitution of the STE18 gene by kanMX4 in the yeast chromosome was achieved by amplifying the DNA frag- ment containing STE18p–kanMX4–STE18t from pGK426- GPTK using primers 25 and 26 (Table 2). The amplified DNA fragment was then used to transform MC-F1, and the transformant was selected on YPD solid medium containing 500 ngÆ mL )1 G418 (geneteccin; Nacalai Tesque Inc., Kyoto, Japan) to yield the BWG2118 strain. Integration of the ZZ mem gene was achieved by amplifying a DNA fragment containing STE18p–PGK5¢– ZZmem–PGK3¢–kanMX4–STE18t from pUMGPT-ZZK using primers 25 and 26 (Table 2). The amplified DNA fragment was used to transform MC-F1, and the transfor- mant was selected on YPD solid medium containing 500 ngÆmL )1 G418 to yield the BZG2118 strain. Integration of the GcM–Fc gene was achieved by amplifying a DNA fragment containing URA3–STE18p– GcM–Fc–PGK3¢–HIS3t from pUMGP-GcMFcH using primers with 50-nucleotide 5¢ segments that were homolo- gous to the region directly upstream of the HIS3 gene (primers 27 and 28; Table 2). The amplified DNA frag- ment was then used to transform BWG2118 and BZG2118, and the transformants were selected on SD-Ura solid medium, yielding the BFG2118 and BZFG2118 strains, respectively. Table 2. Primers used for construction of plasmids and yeast strains. Primer number Sequence (5¢-to3¢) 1 GCCCGTCGACATATTATATATATATATAGG 2 CCCGCTCGAGTCTTAGAATTATTGAGAACG 3 GCCCGGATCCTGATAGTAATAGAATCCAAA 4 CCCCGAATTCAAATTATAGAAAGCAGTAGA 5 AAGGCTCGAGAGATCTGTTTAGCTTGCCTC 6 AAAAGTCGACGAGCTCGTTTTCGACACTGG 7 TTTTGTCGACATGGCGCAACACGATGAAGC CGTAGACAAC 8 GGGGGGATCCTTACATAAGCGTACAACAAA CACTATTTGATTTCGGCGCCTGAGCATCA TTTAGCTTTTT 9 TTTTCTCGAGAAAGATGCCGATTTGGGCGC 10 GGGGCTCGAGGTTTTATATTTGTTGTAAAA 11 GCCCCTCGAGATATTATATATATATATAGG 12 TAAAGGATCCCTTGTCATCGTCATCCTTGT AGTCAACACTATTTGAGTTTGACATTTGGC 13 GAGAGAATTCGGGGGACCGTCAGTCTTCCT CTTCCCCC 14 TTCCGAATTCTCATTTACCCGGAGACAGGG 15 CCCCGCGGCCGCTGACACCGATTATTTAAA 16 TTTTGAGCTCGGAGCCATAATGACAGCAGT 17 TTTTGTCGACATGGCGCAACACGATGAAGC CGTAGACAAC 18 GGGGGGATCCTTACATAAGCGTACAACAAA CACTATTTGATTTCGGCGCCTGAGCATCA TTTAGCTTTTT 19 ATCCAAAGTTTAGCCGATGACCCAAGCCAA 20 TTGGCTTGGGTCATCGGCTAAACTTTGGAT 21 AAACGCCTTCGCCCAAAGTTTAAAAGATGA 22 TCATCTTTTAAACTTTGGGCGAAGGCGTTT 23 TTTTCTCGAGAAAGATGCCGATTTGGGCGC 24 GGGGCTCGAGGTTTTATATTTGTTGTAAAA 25 ATATTATATATATATATAGGGTCGTATATA 26 AAATTATAGAAAGCAGTAGA TAAAACAATG 27 CTTCGAAGAATATACTAAAAAATGAGCAGG CAAGATAAACGAAGGCAAAGTTCAATTCA TCATTTTTTTTTTATTCTTTT 28 GGAGCCATAATGACAGCAGT Detection system for protein–protein interactions N. Fukuda et al. 2642 FEBS Journal 276 (2009) 2636–2644 ª 2009 The Authors Journal compilation ª 2009 FEBS Construction of yeast strains containing Z mem genes rather than the ZZ mem gene were achieved using a process similar to construction of the BZG2118 strain, with the exception that the plasmids pUMGPT-ZK-WT, pUMGPT- ZK-K35A or pUMGPT-ZK-I31A were used instead of pUMGPT-ZZK. In addition, integration of the Gc cyto –Fc gene into these transformants was achieved as shown in Fig. 2C, yielding BFG2Z18-WT, BFG2Z18-K35A and BFG2Z18-I31A strains. Halo bioassay to test growth arrest via the pheromone response An agar diffusion bioassay (halo assay) was undertaken to measure the response to and recovery from pheromone- induced cell-cycle arrest as described previously [16]. The yeast strains were grown in YPD medium at 30 °C over- night. Sterilized paper filter disks (6 mm in diameter) were placed on the dishes, and 10 ng of a-factor pheromone was spotted onto the disks. The cells were then inoculated into fresh YPD medium containing 2% w ⁄ v agar (20 mL, main- tained at 60 °C), grown until they reached an absorbance at 600 nm (A 600 )of10 )3 , and the suspension was immedi- ately poured into a dish. The plates were then incubated at 30 °C for 24 h. Flow cytometric EGFP fluorescence analysis The fluorescence intensity of EGFP fusion proteins in yeast cells stimulated with 5 l m of a-factor in YPD medium for 6 h was measured using a FACSCalibur flow cytometer equipped with a 488 nm air-cooled argon laser (Becton Dickinson and Company), and the data were analyzed using cellquest software (Becton Dickinson and Com- pany). Parameters were as follows: the amplifiers were set in linear mode for forward scattering and in logarithmic mode for the green fluorescence detector (FL1, 530 ⁄ 30 nm bandpass filter) and the orange fluorescence detector (FL2, 585 ⁄ 21 nm bandpass filter). The amplifier gain was set at 1.00 for forward scattering; the detector voltage was set to E00 for forward scattering and 600 V for FL1, and the threshold for forward scattering was set at 52. The EGFP fluorescence signal was collected using a 530 ⁄ 30 nm band- pass filter (FL1), and the fluorescence intensity of 10 000 cells was defined as the FL1-height (FL1-H) geometric mean. Western blot analysis Yeast cells were cultured in YPD medium overnight. The cells were then harvested, washed in NaCl ⁄ Pi to remove culture media and resuspended in sample buffer for SDS ⁄ PAGE at an A 600 of 20. Fractionated cell lysates were prepared by glass bead vortex homogenization for 15 min. Protein extracts were separated by 15% SDS ⁄ PAGE, and Western blot analysis was performed using the primary antibodies goat anti-protein A (Rockland, Gilbertsville, PA, USA) for the ZZ or Z domain, and goat anti-human IgG (Fc) (Kirkegaard Perry Laboratories, Gaithersburg, MD, USA) for the Fc portion. Alkaline phosphatase-conjugated anti-goat IgG (Vector Laboratories, Burlingame, CA, USA) was used as the secondary antibody, and colorimetric detec- tion of alkaline phosphatase activity was performed using 5-bromo-4-chloro-3-indolyl-phosphate and nitroblue tetra- zolium (Promega Co., Madison, WI, USA). Acknowledgements This work was supported by a Research Fellowship for Young Scientists from the Japanese Society for the Promotion of Science, and in part by the Global COE Program ‘Global Center for Education and Research in Integrative Membrane Biology’ from the Ministry of Education, Culture, Sports, Science and Technology of Japan. References 1 Aronheim A, Engelberg D, Li N, ai-Alawi N, Schles- singer J & Karin M (1994) Membrane targeting of the nucleotide exchange factor Sos is sufficient for activat- ing the Ras signaling pathway. Cell 78 , 949–961. 2 Broder YC, Katz S & Aronheim A (1998) The ras recruitment system, a novel approach to the study of protein–protein interactions. Curr Biol 8 , 1121–1124. 3 Ehrhard KN, Jacoby JJ, Fu XY, Jahn R & Dohlman HG (2000) Use of G-protein fusions to monitor integral membrane protein–protein interactions in yeast. Nat Biotechnol 18, 1075–1079. 4 Medici R, Bianchi E, DiSegni G & Tocchini-Valentini GP (1997) Efficient signal transduction by a chimeric yeast-mammalian G protein a subunit Gpa1-Gsa cova- lently fused to the yeast receptor Ste2. EMBO J 24, 7241–7249. 5 Ishii J, Tanaka T, Matsumura S, Tatematsu K, Kuroda S, Ogino C, Fukuda H & Kondo A (2008) Yeast-based fluorescence reporter assay of G protein-coupled recep- tor signalling for flow cytometric screening: FAR1-dis- ruption recovers loss of episomal plasmid caused by signalling in yeast. J Biochem 143, 667–674. 6 Manahan CL, Patnana M, Blumer KJ & Linder ME (2000) Dual lipid modification motifs in Ga and Gc subunits are required for full activity of the pheromone response pathway in Saccharomyces cerevisiae. Mol Biol Cell 18, 957–968. 7 Kronvall G & Williams RC Jr (1969) Differences in anti-protein A activity among IgG subgroups. J Immu- nol 103, 828–833. N. Fukuda et al. Detection system for protein–protein interactions FEBS Journal 276 (2009) 2636–2644 ª 2009 The Authors Journal compilation ª 2009 FEBS 2643 8 Nilsson B, Moks T, Fansson B, Abrahmsen L, Elmbrad A, Holmgren E, Henrichson C, Jones TA & Uhlen MA (1987) Synthetic IgG-binding domain based on staphy- lococcal protein A. Protein Eng 1, 107–113. 9 Shibasaki S, Kuroda K, Nguyen HD, Mori T, Zou W & Ueda M (2006) Detection of protein–protein interac- tions by a combination of a novel cytoplasmic mem- brane targeting system of recombinant proteins and fluorescence resonance energy transfer. Appl Microbiol Biotechnol 70, 451–457. 10 Jendeberg L, Persson B, Anderssson R, Karlsson R, Uhlen M & Nilsson B (1995) Kinetic analysis of the interaction between protein A domain variants and human Fc using plasmon resonance detection. J Mol Recognit 8, 270–278. 11 Brachann CB, Davies A, Cost GJ, Caputo E, Li J, Hieter P & Boeke JD (1998) Designer deletion strains derived from Saccharomyces cerevisiae S288C: a useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast 14, 115–132. 12 Gu ¨ ldener U, Heck S, Fielder T, Beinhauer J & Hege- mann JH (1996) A new efficient gene disruption cassette for repeated use in budding yeast. Nucleic Acids Res 24, 2519–2524. 13 Nakamura Y, Shibasaki S, Ueda M, Tanaka A, Fuku- da H & Kondo A (2001) Development of novel whole- cell immunoadsorbents by yeast surface display of the IgG-binding domain. Appl Microbiol Biotechnol 57, 500–505. 14 Fukuda N, Ishii J, Shibasaki S, Ueda M, Fukuda H & Kondo A (2007) High-efficiency recovery of target cells using improved yeast display system for detection of protein–protein interactions. Appl Microbiol Biotechnol 76, 151–158. 15 Gietz D, St. Jean A, Woods RA & Schiestl RH (1992) Improved method for high efficiency transformation of intact yeast cells. Nucleic Acids Res 20, 1425. 16 Ishii J, Matsumura S, Kimura S, Tatematsu K, Kuroda S, Fukuda H & Kondo A (2006) Quantitative and dynamic analyses of G protein-coupled receptor signal- ing in yeast using Fus1, enhanced green fluorescence protein (EGFP), and His3 fusion protein. Biotechnol Prog 22, 954–960. Detection system for protein–protein interactions N. Fukuda et al. 2644 FEBS Journal 276 (2009) 2636–2644 ª 2009 The Authors Journal compilation ª 2009 FEBS . AAATTATAGAAAGCAGTAGA TAAAACAATG 27 CTTCGAAGAATATACTAAAAAATGAGCAGG CAAGATAAACGAAGGCAAAGTTCAATTCA TCATTTTTTTTTTATTCTTTT 28 GGAGCCATAATGACAGCAGT Detection system. GCCCGGATCCTGATAGTAATAGAATCCAAA 4 CCCCGAATTCAAATTATAGAAAGCAGTAGA 5 AAGGCTCGAGAGATCTGTTTAGCTTGCCTC 6 AAAAGTCGACGAGCTCGTTTTCGACACTGG 7 TTTTGTCGACATGGCGCAACACGATGAAGC CGTAGACAAC 8