XA-VP16 (A), ADRB2-Cub-LexA-VP16 (B), or HTR1A-Cub-LexAVP16 (C). The control prey plasmid 15826876 was pPR3-C mock vector. Error bars represent the standard deviations (n = 3). doi:10.1371/journal.pone.0066793.gScreening of Human GPCR HeterodimerFigure 8. Detection for dimerization and ligand-induced conformational changes of human GPCRs. (A) Quantitative b-galactosidase assays for heterodimerization of AGTR1 in NMY63 strain. NMY63 yeast strain was transformed with GPCR-NubG indicated at the bottom and AGTR1Cub-LexA-VP16. (B ) Ligand assays for detection of conformational changes in GPCR dimerizations. (B) AT1/AT2 (AGTR1/AGTR2) heterodimers. Incubation time, 18 h. Castanospermine web Angiotensin II conc., 0 or 10 mM. (C) MT1/MT2 (MTNR1A/MTNR1B) heterodimers. Incubation time, 18 h. Melatonin conc., 0 or 10 mM. (D) GABAB1a/GABAB2 (GABBR1a/GABBR2) heterodimers. Incubation time, 18 h. GABA conc., 0 or 100 mM. The control prey plasmid was pPR3-C mock vector. Error bars represent the standard deviations (n = 3). (*P,0.05). doi:10.1371/journal.pone.0066793.gallowed successful detection of the conformational change of MTNR1A/MTNR1B heterodimers in the presence of melatonin (Fig. 8C and Fig. S4F), suggesting that our system can detect ligand-mediated conformational changes as well as the heterodimer formations. Moreover, we also tested the detection of the conformational change of GABAB2/GABAB1a heterodimers in the presence of GABA (Fig. 8D). While the addition of 100 mM of GABA did not affect the states of the GABAB2 homodimers, the bgalactosidase assay exposed the conformational change of GABAB2/GABAB1a heterodimers in consistency with previous reports [28] (Fig. 8D). Because a positive result in the assay potentially could come about through indirect association of receptors via a third protein or close co-localization, it is important to re-evaluate the irrefragability with other methods or in human cells. Additionally, there is no guarantee that association is actually physiologically relevant. However, it is also true that this system could narrow down the new candidates of GPCR heterodimers with a bit of effort. If one accepts these uncertainties, the assay provides a nice way of monitoring changes that does not depend on effective downstream signaling through the GPCR pathway. In summary, we have developed a specialized method to screen candidate heterodimer partners for target GPCRs based on the split-ubiquitin membrane yeast two-hybrid method. This modified system permitted the rapid and facile detection of not only the heterodimer formation of target human GPCRs, but also theligand-mediated conformational changes in living yeast cells. Since budding yeast Saccharomyces cerevisiae can functionally express human GPCRs [29,30], construction of a large prey library would be beneficial for the identification of heterodimer candidates as the partners of target human GPCRs. Our system will be a useful tool to assist in the intermolecular mapping of interactions among GPCRs and uncover potential targets for the development of new therapeutic agents.Materials and Methods MediaSynthetic dextrose (SD) media contained 6.7 g/l yeast nitrogen base without amino acids (YNB) (BD Human parathyroid hormone-(1-34) Diagnostic Systems, Sparks, MD, USA) and 20 g/l glucose. YPDA media contained 10 g/l yeast extract, 20 g/l peptone, 20 g/l glucose and 55 mg/l adenine. Amino acids and nucleotides (60 mg/l leucine, 40 mg/l tryptophan, 40 mg/l adenine, 20 mg/l histidine or 20 mg/l uracil) were supplemented into SD media t.XA-VP16 (A), ADRB2-Cub-LexA-VP16 (B), or HTR1A-Cub-LexAVP16 (C). The control prey plasmid 15826876 was pPR3-C mock vector. Error bars represent the standard deviations (n = 3). doi:10.1371/journal.pone.0066793.gScreening of Human GPCR HeterodimerFigure 8. Detection for dimerization and ligand-induced conformational changes of human GPCRs. (A) Quantitative b-galactosidase assays for heterodimerization of AGTR1 in NMY63 strain. NMY63 yeast strain was transformed with GPCR-NubG indicated at the bottom and AGTR1Cub-LexA-VP16. (B ) Ligand assays for detection of conformational changes in GPCR dimerizations. (B) AT1/AT2 (AGTR1/AGTR2) heterodimers. Incubation time, 18 h. Angiotensin II conc., 0 or 10 mM. (C) MT1/MT2 (MTNR1A/MTNR1B) heterodimers. Incubation time, 18 h. Melatonin conc., 0 or 10 mM. (D) GABAB1a/GABAB2 (GABBR1a/GABBR2) heterodimers. Incubation time, 18 h. GABA conc., 0 or 100 mM. The control prey plasmid was pPR3-C mock vector. Error bars represent the standard deviations (n = 3). (*P,0.05). doi:10.1371/journal.pone.0066793.gallowed successful detection of the conformational change of MTNR1A/MTNR1B heterodimers in the presence of melatonin (Fig. 8C and Fig. S4F), suggesting that our system can detect ligand-mediated conformational changes as well as the heterodimer formations. Moreover, we also tested the detection of the conformational change of GABAB2/GABAB1a heterodimers in the presence of GABA (Fig. 8D). While the addition of 100 mM of GABA did not affect the states of the GABAB2 homodimers, the bgalactosidase assay exposed the conformational change of GABAB2/GABAB1a heterodimers in consistency with previous reports [28] (Fig. 8D). Because a positive result in the assay potentially could come about through indirect association of receptors via a third protein or close co-localization, it is important to re-evaluate the irrefragability with other methods or in human cells. Additionally, there is no guarantee that association is actually physiologically relevant. However, it is also true that this system could narrow down the new candidates of GPCR heterodimers with a bit of effort. If one accepts these uncertainties, the assay provides a nice way of monitoring changes that does not depend on effective downstream signaling through the GPCR pathway. In summary, we have developed a specialized method to screen candidate heterodimer partners for target GPCRs based on the split-ubiquitin membrane yeast two-hybrid method. This modified system permitted the rapid and facile detection of not only the heterodimer formation of target human GPCRs, but also theligand-mediated conformational changes in living yeast cells. Since budding yeast Saccharomyces cerevisiae can functionally express human GPCRs [29,30], construction of a large prey library would be beneficial for the identification of heterodimer candidates as the partners of target human GPCRs. Our system will be a useful tool to assist in the intermolecular mapping of interactions among GPCRs and uncover potential targets for the development of new therapeutic agents.Materials and Methods MediaSynthetic dextrose (SD) media contained 6.7 g/l yeast nitrogen base without amino acids (YNB) (BD Diagnostic Systems, Sparks, MD, USA) and 20 g/l glucose. YPDA media contained 10 g/l yeast extract, 20 g/l peptone, 20 g/l glucose and 55 mg/l adenine. Amino acids and nucleotides (60 mg/l leucine, 40 mg/l tryptophan, 40 mg/l adenine, 20 mg/l histidine or 20 mg/l uracil) were supplemented into SD media t.