Portal de Periódicos da CAPES

Efficient signal transduction by a chimeric yeast-mammalian G protein alpha subunit Gpa1-Gsalpha covalently fused to the yeast receptor Ste2

1997; Springer Nature; Volume: 16; Issue: 24 Linguagem: Inglês

10.1093/emboj/16.24.7241

1460-2075

R. Medici,

Cellular transport and secretion

Article15 December 1997free access Efficient signal transduction by a chimeric yeast–mammalian G protein α subunit Gpa1–Gsα covalently fused to the yeast receptor Ste2 Roberto Medici Roberto Medici Istituto G.Donegani, EniChem, 00016 Monterotondo (Rome), Italy Search for more papers by this author Emma Bianchi Emma Bianchi Istituto G.Donegani, EniChem, 00016 Monterotondo (Rome), Italy Search for more papers by this author Gianfranco Di Segni Corresponding Author Gianfranco Di Segni Istituto G.Donegani, EniChem, 00016 Monterotondo (Rome), Italy Search for more papers by this author Glauco P. Tocchini-Valentini Glauco P. Tocchini-Valentini Institute of Cell Biology, CNR, Department of Biochemistry and Molecular Biology, University of Chicago, 00137 Rome, Italy, Chicago, IL 60637 USA Search for more papers by this author Roberto Medici Roberto Medici Istituto G.Donegani, EniChem, 00016 Monterotondo (Rome), Italy Search for more papers by this author Emma Bianchi Emma Bianchi Istituto G.Donegani, EniChem, 00016 Monterotondo (Rome), Italy Search for more papers by this author Gianfranco Di Segni Corresponding Author Gianfranco Di Segni Istituto G.Donegani, EniChem, 00016 Monterotondo (Rome), Italy Search for more papers by this author Glauco P. Tocchini-Valentini Glauco P. Tocchini-Valentini Institute of Cell Biology, CNR, Department of Biochemistry and Molecular Biology, University of Chicago, 00137 Rome, Italy, Chicago, IL 60637 USA Search for more papers by this author Author Information Roberto Medici1, Emma Bianchi1, Gianfranco Di Segni 1 and Glauco P. Tocchini-Valentini2 1Istituto G.Donegani, EniChem, 00016 Monterotondo (Rome), Italy 2Institute of Cell Biology, CNR, Department of Biochemistry and Molecular Biology, University of Chicago, 00137 Rome, Italy, Chicago, IL 60637 USA *Corresponding author. E-mail: [email protected] The EMBO Journal (1997)16:7241-7249https://doi.org/10.1093/emboj/16.24.7241 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Saccharomyces cerevisiae uses G protein-coupled receptors for signal transduction. We show that a fusion protein between the α-factor receptor (Ste2) and the Gα subunit (Gpa1) transduces the signal efficiently in yeast cells devoid of the endogeneous STE2 and GPA1 genes. To evaluate the function of different domains of Gα, a chimera between the N-terminal region of yeast Gpa1 and the C-terminal region of rat Gsα has been constructed. This chimeric Gpa1–Gsα is capable of restoring viability to haploid gpa1Δ cells, but signal transduction is prevented. This is consistent with evidence showing that the C-terminus of the homologous Gα is required for receptor–G protein recognition. Surprisingly, a fusion protein between Ste2 and Gpa1–Gsα is able to transduce the signal efficiently. It appears, therefore, that the C-terminus of Gα is mainly responsible for bringing the G protein into the close proximity of the receptor's intracellular domains, thus ensuring efficient coupling, rather than having a particular role in transmitting the signal. To confirm this conclusion, we show that two proteins interacting with each other (such as Snf1 and Snf4, or Ras and Raf), each of them fused either to the receptor or to the chimeric Gα, allow efficient signal transduction. Introduction A major class of signal transducing proteins in eukaryotic cells, from yeasts to mammals, is composed of seven-transmembrane domain receptors coupled to GTP-binding proteins (G proteins). This class includes several hundred members which can respond to a variety of agents, such as hormones, neurotransmitters, odorants and light signals (reviewed by Baldwin, 1994; Strader et al., 1994). Upon ligand binding, the receptors activate αβγ heterotrimeric G proteins, promoting the exchange of bound GDP with GTP on Gα and the subsequent dissociation of Gα from Gβγ. Both the activated, GTP-bound Gα and/or the free Gβγ dimer regulate specific target effectors, stimulating or inhibiting their function. Once GTP is hydrolyzed to GDP, the α and βγ subunits reassociate and the G protein returns to the receptor in its resting state (reviewed by Conklin and Bourne, 1993; Neer, 1995; Hamm and Gilchrist, 1996). The yeast Saccharomyces cerevisiae is a useful organism for studying signal transduction pathways regulated by G protein-coupled receptors. The two haploid cell types (a and α) each secrete a peptide pheromone (a-factor and α-factor, respectively), that acts on the other cell type to promote conjugation, resulting in the formation of a/α diploid cells. The a and α mating factors bind to cell surface receptors of the seven-transmembrane domain type (encoded by the STE3 and STE2 genes, respectively). In response to pheromone binding to the receptors, the G protein α subunit (Gpa1) replaces bound GDP with GTP and dissociates from the G protein βγ complex, which in turn triggers a cascade of events leading to transcriptional induction of specific genes, stimulation of morphological changes and inhibition of growth (reviewed by Kurjan, 1992, 1993; Sprague and Thorner, 1992). The deletion of the gene encoding the yeast Gα subunit (GPA1) results in lethality in haploid cells, because the free Gβγ complex constitutively activates the pathway which leads to growth inhibition. Mammalian Gα subunits (such as Gsα, Goα or Giα) or chimeric yeast–mammalian Gα subunits are able to bind to the yeast Gβγ complex, and thereby restore viability to haploid cells with the deleted GPA1 gene (gpa1Δ) (Dietzel and Kurjan, 1987; Kang et al., 1990). However, mammalian or even chimeric yeast–mammalian Gα subunits do not allow signal transduction in yeast: presumably the mammalian Gα cannot interact with the yeast receptor. This is consistent with evidence showing that the C-terminus of the homologous Gα is required for receptor–G protein recognition. To investigate specific interactions between receptors, G proteins and effectors in intact cells, Bertin et al. (1994) engineered a receptor–transducer fusion protein, by covalently linking the β2-adrenergic receptor to the Gsα subunit. They found that this fusion protein was able to restore efficient cellular signaling in mutant animal cells which, although expressing endogenous β2-adrenergic receptors, were devoid of endogenous Gsα subunits. Very recently, a fusion protein between the α2A-adrenergic receptor and Gi1α was shown to be functional, as judged by the induced stimulation of the G protein's GTPase activity (Wise et al., 1997). We decided to undertake a similar approach in yeast cells, where it is possible to obtain mutants with deletions of both the GPA1 gene and the STE2 gene. We show here that not only is a fusion protein between the Ste2 receptor and the Gpa1 protein functional in signal transduction, but surprisingly, even a fusion protein between the yeast Ste2 receptor and a chimeric yeast–mammalian Gα subunit is able to transduce the signal efficiently. This result is in apparent contrast with the above-mentioned inability of the chimeric Gα to interact with the yeast receptor. In other words, when the Ste2 receptor and the chimeric yeast–mammalian Gα subunit are separate, there is no response at all to α-factor, but when these two proteins are fused together, there is efficient signal transduction. We explain this finding by suggesting that the C-terminus of the Gα subunit is mainly responsible for bringing the G protein into the close proximity of the receptor's intracellular domains, thus ensuring efficient coupling. This function becomes dispensable when the two proteins are covalently fused. Results The Ste2–Gpa1 fusion protein complements the deletion of the endogenous GPA1 gene A fusion protein between the S.cerevisiae α-factor receptor (Ste2) and G protein α subunit (Gpa1) was obtained, as described in Materials and methods, by joining the entire STE2 gene, except for the bases encoding the last 62 amino acids of the long cytoplasmic tail, to the 5′-end of the complete GPA1 gene. This construct was inserted into either the multicopy episomal plasmid YEp24 or into the single-copy integrating plasmid YIp5, both of which carry the URA3-selectable marker. Two separate questions could be addressed concerning the functionality of the Ste2–Gpa1 fusion protein: is the fusion protein able to function as a Gα subunit, and is it also able to function as a receptor? To answer the first question, the recombinant plasmids encoding the fusion protein were transformed into the yeast diploid strain RM7, where both alleles of STE2 and a single allele of GPA1 were deleted, and the diploids were sporulated and dissected. In haploid cells, the deletion of the GPA1 gene would result in constitutive activation of the signal transduction pathway and, therefore, inhibition of growth. Based upon this, we expected the diploid strain RM7 without a Gpa1-encoding plasmid to give rise to two viable spores (those that received the wild-type chromosomal GPA1 allele) and two spores unable to grow (those that received the gpa1Δ mutant allele). However, in the presence of a plasmid encoding a functional Gα subunit, four viable spores could arise, provided that the two gpa1Δ spores received the plasmid. Figure 1 shows that the Ste2–Gpa1 fusion protein is able to complement the deletion of GPA1, yielding tetrads with more than two viable spores. Because the fusion gene is under the MATa-specific promoter of STE2, only those spores that are MATa will express the fusion protein and thus be viable. This fact explains why some tetrads do not contain four viable spores, but only three or two. The ability of the Ste2–Gpa1 fusion protein to complement the deletion of GPA1 is seen when either a multicopy plasmid or a single-copy plasmid, which was integrated at the STE2 chromosomal locus, are used (Figure 1C and D). Figure 1.Complementation of gpa1Δ growth defect by the fusion protein Ste2–Gpa1. Yeast diploid strains were sporulated and dissected. Homozygous wild-type GPA1. GPA1 diploid strain GDS94 gives rise to four viable spores (A); heterozygous mutant gpa1Δ/GPA1 strain RM7, with the deletion of one GPA1 allele, gives rise to only two viable spores (B); a multicopy plasmid (YEp24) or an integrating plasmid (YIp5) encoding a Ste2–Gpa1 fusion protein complement the gpa1Δ mutation, giving rise to three or four viable spores (C and D). Download figure Download PowerPoint To investigate further the functionality of the Ste2–Gpa1 fusion protein, cells not containing the wild-type Gpa1 protein, which could interfere with the analysis, were needed. Therefore, to discriminate the spores that received the wild-type chromosomal GPA1 allele from those that inherited the gpa1Δ mutation, the cells of the tetrads were examined microscopically. A pattern was observed where two colonies of each tetrad contained well-growing cells while the other one or two colonies contained 2–5% morphologically aberrant cells, most likely due to the occasional loss of the complementing plasmid and subsequent activation of the inhibiting pathway. We therefore assigned the wild-type phenotype to the GPA1 cells, and the aberrant one to the gpa1Δ mutant. Confirmation of this line of reasoning was provided by the fact that this pattern was observed only when cells carried the episomal recombinant plasmid, and not when they contained the integrated one, in which case plasmid loss is a rare event. Because no aberrant phenotype was visible with the integrating plasmid, we utilized another method to distinguish between the gpa1Δ mutants and the wild-type GPA1 colonies. We streaked the colonies arising from the tetrads onto plates containing 5-fluoroorotic acid (5-FOA), which counterselect for the presence of the URA3 marker: rare events of popping-out of the URA3 integrating plasmid could give rise to Ura− viable cells only in wild-type GPA1 segregants, whereas in gpa1Δ mutants the presence of the URA3 plasmid containing the STE2–GPA1 fusion gene is absolutely necessary for growth, and no pop-out events could be selected for. Indeed, two colonies from each tetrad were growing on 5-FOA plates, while the other one or two were not. The latter were considered to be gpa1Δ. Consistent with these observations, when the cells containing the episomal URA3 plasmid YEp24-STE2/GPA1 were grown in a non-selective liquid medium, those assumed to bear the chromosomal GPA1 gene allowed the loss of the plasmid and became Ura−, whereas those assumed to be gpa1Δ cells did not, indicating that the retention of the plasmid was necessary for growth. We concluded from this analysis that the Gα subunit Gpa1, even when fused to the Ste2 receptor, is indeed able to complement efficiently the deletion of the endogenous GPA1 gene. The Ste2–Gpa1 fusion protein also complements the deletion of STE2 The presence of three or four viable spores shown in Figure 1 is an indication that the fusion protein Ste2–Gpa1 is able to bind the G protein βγ complex, and thereby keeps the signal transduction pathway in the resting state. However, it does not reveal whether or not the fusion protein is able to transmit the signal upon binding of the α-factor to its receptor moiety. This question was addressed by several approaches. The halo assay (Figure 2A) is used to measure the inhibition of growth caused by the α-factor binding to the receptor and subsequent onset of signal transduction. It can be seen that cells devoid of the endogenous Ste2 receptor and Gpa1 subunit, but carrying a plasmid encoding the fusion protein Ste2–Gpa1, have a response to the α-factor similar to wild-type cells. The size of the halo is approximately equivalent in cells harboring both types of recombinant plasmids, either multicopy or integrating. Moreover, the presence of the chromosomal wild-type GPA1 gene, together with the fusion gene, does not increase the efficiency of the response significantly. Figure 2.Response of the Ste2–Gpa1 fusion protein to α-factor. Signal transduction by the Ste2–Gpa1 fusion protein was analyzed by its ability to arrest the cell cycle (A) and to induce gene expression (B) in response to α-factor in MATa segregants of strain RM7, with (GPA1 wt) or without (gpa1Δ) the chromosomal GPA1 gene, carrying an episomal, multicopy plasmid (YEp24), or an integrating plasmid (YIp5), encoding the Ste2–Gpa1 fusion protein. The controls were wild-type GPA1 RM7 segregants expressing the wild-type or the truncated (lacking the last 62 amino acids of the cytoplasmic tail) Ste2 receptor. (A) Growth inhibition induced by α-factor was analyzed by the halo assay. Different doses (left, 0.4 μg; right, 4 μg) of α-factor were spotted on filter disks on a lawn of cells. Plates were photographed after 48 h of incubation at 30°C. (B) Pheromone-induced expression of FUS1–lacZ was checked by the β-galactosidase assay. Cells were incubated with (+) or without (−) α-factor (2.5 μg/ml) for 6 h. The activation of the signal transduction pathway was measured by assaying the β-galactosidase activity in permeabilized cells. The data represent averages of three experiments; error bars indicate 1 SD. Download figure Download PowerPoint The fusion protein lacks the last 62 amino acids of the Ste2 receptor, which are part of the long cytoplasmic tail, a region known to be involved in the desensitization process of the signal transduction pathway. Its loss causes a supersensitive phenotype (Konopka et al., 1988; Reneke et al., 1988), and therefore it could be argued that the efficiency of signal transduction observed with the Ste2–Gpa1 fusion protein is due to its inability to undergo desensitization. To rule out this possibility, we tested the response of a truncated Ste2 receptor, lacking those 62 amino acids, to α-factor in GPA1 cells. We found no significant increase in the signal transduction by truncated Ste2 as compared with wild-type Ste2 (Figure 2A). Induction of FUS1–lacZ expression by α-factor binding to the Ste2–Gpa1 fusion protein A second, more sensitive and quantitative approach to analyze signal transduction in yeast involves assaying the activity of a transfected Escherichia coli β-galactosidase gene, placed under the control of the yeast FUS1 gene promoter. FUS1 is one of several yeast genes activated by the signal transduction pathway when α-factor binds to the Ste2 receptor. The β-galactosidase activity is therefore a measure of the induction level of the signal transduction pathway. As shown in Figure 2B, incubating α-factor with haploid gpa1Δ cells harboring the fusion protein Ste2–Gpa1 causes a strong induction of FUS1–lacZ expression, comparable with expression levels obtained in wild-type cells. Cells that had the chromosomal GPA1 gene deleted, but carried the fusion gene STE2–GPA1 on the episomal plasmid YEp24, had a basal level of FUS1–lacZ expression higher than cells possessing the wild-type chromosomal GPA1 gene. This basal activity could be due to an impairment of the interaction between the Gα moiety of the fusion protein and the Gβγ complex. More probably, however, it is caused by the occasional loss of the episomal plasmid encoding the fusion protein in some of the cells of the culture, which would render the Gβγ complex free to activate the pathway. In fact, when the fusion gene was inserted into the integrating plasmid YIp5, there was no substantial basal activity (Figure 2B). Consistently, when a wild-type GPA1 gene is placed on an episomal plasmid, it produces a basal activation of the signal transduction pathway in gpa1Δ cells (data not shown). A fusion protein between the Ste2 receptor and a chimeric yeast–mammalian Gα subunit is able to transduce the signal efficiently It has been shown previously that a mammalian G protein α subunit is able to complement the deletion of the GPA1 gene in S.cerevisiae (Dietzel and Kurjan, 1987). The efficiency of complementation is increased if a chimera between the N-terminal region of the yeast Gα subunit and the C-terminal region of a mammalian Gα is utilized (Kang et al., 1990). However, the complementation by mammalian Gα or by the yeast–mammalian chimeric Gα is only limited. These Gα subunits are able to interact with the yeast Gβγ complex, but not with the yeast pheromone receptor: therefore, a yeast gpa1Δ strain containing either the mammalian or the yeast–mammalian chimeric Gα subunit is viable but sterile, and does not respond to pheromones (Kang et al., 1990). This is consistent with the generally accepted model according to which the C-terminus of Gα is involved in receptor recognition (see Discussion). We wondered, therefore, whether a fusion protein between Ste2 and a chimeric yeast–mammalian Gα subunit would be active in signal transduction. First, we constructed a chimera between yeast Gpa1 and rat Gsα, as described in Materials and methods. It contains the N-terminal 362 amino acids of Gpa1 and the C-terminal 128 amino acids of rat Gsα. The junction site is within a highly conserved sequence, and therefore it is likely that this chimera would retain the G protein's normal structure. The GPA1–Gsα chimeric gene was placed under the control of the GAL1 promoter into the very high copy number plasmid pGAL (derived from pEMBLyex2), which would compensate the possible low efficiency of the chimeric Gα in sequestering the Gβγ complex. This plasmid was introduced into the GPA1. gpa1Δ diploid strain RM7, and the diploids were sporulated and dissected. Figure 3A shows that the chimera Gpa1–Gsα is able to complement the deletion of the GPA1 gene efficiently, producing tetrads with four viable spores. As expected, when the tetrads were grown on glucose instead of galactose, only two viable spores arose (Figure 3B). Figure 3.Complementation of gpa1Δ growth defect by a chimeric yeast–mammalian Gα subunit fused to the Ste2 receptor. Yeast diploid strain RM7 (gpa1Δ/GPA1) was sporulated and dissected. A pGAL plasmid encoding a yeast–mammalian chimeric Gα subunit Gpa1–Gsα, or a fusion protein Ste2–Gpa1–Gsα, complements the gpa1Δ mutation, giving rise to four viable spores when cells are grown on a galactose-containing medium (A and C), and to only two viable spores on a glucose medium (B and D). Download figure Download PowerPoint We then constructed a fusion protein between the Ste2 receptor and the yeast–mammalian chimera Gpa1–Gsα, as described in Materials and methods. Immunoblot analysis verified the proper presence of the fusion protein Ste2–Gpa1–Gsα with the expected molecular weight (Figure 4). This fusion protein was found to complement the lethality of the GPA1 deletion in haploid cells, as shown by dissection of tetrads of the GPA1. gpa1Δ diploid strain RM7 (Figure 3C). Figure 4.Visualization of the Ste2–Gpa1–Gsα fusion protein using anti–Gsα antibodies. The immunoblot shows proteins extracted from membranes (lanes 1 and 2) or from cytoplasm (lanes 3 and 4) of gpa1Δ segregants of strain RM7, containing the fusion protein between the Ste2 receptor and the Gpa1–Gsα chimera (lanes 2 and 3) or the Gpa1–Gsα chimera alone (lanes 1 and 4). The antibodies were raised against the C-terminus of Gsα. The molecular weights of markers are shown in kDa (M). Download figure Download PowerPoint Surprisingly, when we checked the ability of gpa1Δ RM7 segregants containing the fusion protein Ste2–Gpa1–Gsα to respond to α-factor, a strong response was obtained, both in the halo assay (Figure 5A) and in the FUS1–lacZ induction assay (Figure 5B). On the other hand, as expected, gpa1Δ RM7 segregants expressing, separately, the chimera Gpa1–Gsα and the Ste2 receptor were found to be completely defective in α-factor response, in the halo assay (Figure 5A) as well as in the β-gal assay (Figure 5B). The elevated basal activity in gpa1Δ cells could be caused by the occasional loss of the plasmid encoding the chimeric Gα (even though this is a very high copy number plasmid), but it is also likely that the Gpa1–Gsα chimera, as opposed to wild-type Gpa1, is relatively inefficient in sequestering the Gβγ complex, thereby causing a certain amount of activation of the pathway. Figure 5.Efficient signal transduction by a chimeric yeast–mammalian Gα subunit covalently fused to the Ste2 receptor, analyzed by its ability to arrest the cell cycle and to induce gene expression in response to α-factor. (A) MATa segregants of strain RM7, with (GPA1 wt) or without (gpa1Δ) the chromosomal GPA1 gene, expressing the Ste2–Gpa1–Gsα fusion protein, were analyzed for growth inhibition by the halo assay. The controls were wild-type GPA1 RM7 segregants expressing the wild-type Ste2 receptor, and gpa1Δ RM7 segregants expressing, separately, the Ste2 receptor and the yeast–mammalian chimera Gpa1–Gsα. Synthetic α-factor (4 μg) was spotted on filter disks on a lawn of cells. Plates were photographed after 48 h of incubation at 30°C. (B) Pheromone-induced expression of FUS1–lacZ was checked by the β-galactosidase assay. MATa segregants of strain RM7, with (GPA1 wt) or without (gpa1Δ) the chromosomal GPA1 gene, carrying a pGAL plasmid encoding the Ste2–Gpa1–Gsα fusion protein or carrying two separate plasmids encoding the Ste2 receptor and the chimeric Gpa1–Gsα, respectively, were incubated with (+) or without (−) α-factor (2.5 μg/ml) for 6 h. The control was a wild-type GPA1 strain carrying a YEp24-STE2 plasmid. The activation of the signal transduction pathway was measured by assaying the β–galactosidase activity in permeabilized cells. The data represent averages of three experiments; error bars indicate 1 SD. Download figure Download PowerPoint To rule out the possibility that gene conversion or other recombination processes occurred during meiotic division between the chromosomal GPA1 gene and the chimeric GPA1–Gsα gene carried by the plasmid, we recovered the plasmid from several gpa1Δ RM7 segregants, and checked its identity. We could verify, by restriction mapping and sequencing of a dozen independently rescued plasmids, that these indeed contained the chimeric gene. We conclude, therefore, that when the Ste2 receptor (both the wild-type and the 62 amino acid-less truncated form) and the chimeric yeast–mammalian Gα subunit are separate, there is no response at all to α-factor. On the other hand, when the two proteins are covalently fused together, there is a very efficient signal transduction. Ste2/Snf1 couples to Gpa1–Gsα/Snf4 and Ste2/Raf couples to Gpa1–Gsα/Ras to transmit the signal The results obtained with the Ste2–Gpa1–Gsα fusion protein suggest that the specific interaction of the receptor with the C-terminus of Gα is mainly necessary to bring the two proteins into close proximity, rather than having a particular role in transmitting the signal. The fusion between the receptor and the Gα subunit overcomes the requirement for this specific interaction. If this hypothesis is correct, we reasoned that it should be possible to reach the same goal by utilizing two other proteins (say X and Y), interacting with each other, each of them fused to a separate component of the signal transduction system, either the receptor or the Gα subunit. It will be the interaction of the two separate X and Y proteins which will bring the receptor and Gα into close proximity, thus ensuring the transmission of the signal between the latter two proteins (Figure 6). Figure 6.Coupling of the receptor to the Gα subunit through the interaction of two sticky proteins. (A) A seven-transmembrane domain receptor is unable to interact with a G protein α subunit bearing a heterologous C-terminus (darkened segment). (B) Two hybrid proteins are constructed: one between the receptor and protein X, and another between Gα and protein Y. The interaction between the two proteins X and Y allows coupling of the Gα subunit to the receptor (the regions of contact between the receptor and Gα are only indicative). (C) The presence of the ligand activates the receptor which triggers the GDP/GTP exchange on the Gα subunit and the release of the βγ complex. The latter in turn activates a cascade of reactions, leading to the induction of specific genes. Download figure Download PowerPoint To test this hypothesis, we constructed two pairs of interacting hybrid proteins: Ste2/Snf1 matching with Gpa1–Gsα/Snf4, and Ste2/Raf matching with Gpa1–Gsα/Ras. Snf1 and Snf4 are required for glucose derepression in S.cerevisiae (Johnston and Carlson, 1992) and previously were shown to interact in the two-hybrid system (Fields and Song, 1989) and also by other means (Celenza et al., 1989). Ras and Raf are oncoproteins which also were shown to interact with each other (Vojtek et al., 1993; Van Aelst et al., 1993). We have used the constitutively active form of Ras (valine at position 12), which presumably would interact with Ste2/Raf more efficiently. According to our hypothesis, the interaction between Snf1 and Snf4, or between Raf and Ras, should bring the Ste2 receptor and the Gpa1–Gsα protein (which by themselves are unable to interact, as we have shown above) close enough to each other so that they could couple and transmit the signal. In Figure 7 we show that this is indeed the case. Haploid yeast cells responded to α-factor only when both matching hybrid proteins (Ste2/Snf1 with Gpa1–Gsα/Snf4, or Ste2/Raf with Gpa1–Gsα/Ras) were present (Figure 7A and B, respectively). The level of FUS1–lacZ induction in these cells was considerably higher than the basal activity without α-factor or the level obtained in the controls containing only one hybrid protein. Figure 7.The interactions between Snf1 and Snf4 and between Raf and Ras allow receptor–Gα coupling. (A) MATa segregants of strain RM7, with (GPA1 wt) or without (gpa1Δ) the chromosomal GPA1 gene, carrying the plasmids YIp5-STE2/SNF1 and pGAL-GPA1-Gsα/SNF4, were incubated with (+) or without (−) α-factor (2.5 μg/ml). Controls are cells expressing only one hybrid protein, i.e. cells carrying the plasmids YIp5-STE2/SNF1 and pGAL-GPA1-Gsα, or YEp24-STE2 and pGAL-GPA1-Gsα/SNF4. (B) MATa, gpa1Δ strain RM20 (devoid of the plasmid pYX212-GPA1), carrying the plasmids pYX123-STE2/Raf and pYX242-GPA1-Gsα/Ras, were incubated with (+) or without (−) α-factor (2.5 μg/ml). Controls are cells expressing only one hybrid protein, i.e. cells carrying the plasmids pYX123-STE2/Raf and pYX242-GPA1-Gsα, or pYX123-STE2 and pYX242-GPA1-Gsα/Ras. Other controls are cells expressing non-matching hybrid proteins, i.e. cells carrying the plasmids pYX123-STE2/Raf and pGAL-GPA1-Gsα/SNF4, or pYX123-STE2/SNF1 and pYX242-GPA1-Gsα/Ras. The activation of the signal transduction pathway was measured by assaying the β-galactosidase activity in permeabilized cells. The data represent averages of three experiments; error bars indicate 1 SD. Download figure Download PowerPoint To ascertain the specificity of the interaction further, we show that non-matching hybrid proteins (i.e. Ste2/Snf1 with Gpa1–Gsα/Ras, and Ste2/Raf with Gpa1–Gsα/Snf4) are unable to transduce the signal (Figure 7B). Discussion The main finding of this study is that two proteins (in this case, a seven-transmembrane domain receptor and a G protein α subunit), which are unable to interact with each other because one of them bears a heterologous portion (the C-terminal region of the mammalian Gα subunit replacing the corresponding region in the yeast Gα), nevertheless can be functionally coupled if they are covalently linked together. In other words, when the yeast α-factor receptor Ste2 and the chimeric yeast–mammalian Gα subunit Gpa1–Gsα are separate, there is no response at all to α-factor, but when the two proteins are fused together, there is ef