Identification of a Polar Region in Transmembrane Domain 6 That Regulates the Function of the G Protein-Coupled α-Factor Receptor

Abstract

The α-factor pheromone receptor (Ste2p) of the yeast Saccharomyces cerevisiae belongs to the family of G protein-coupled receptors that contain seven transmembrane domains (TMDs). Because polar residues can influence receptor structure by forming intramolecular contacts between TMDs, we tested the role of the five polar amino acids in TMD6 of the α-factor receptor by mutating these residues to nonpolar leucine. Interestingly, a subset of these mutants showed increased affinity for ligand and constitutive receptor activity. The mutation of the most polar residue, Q253L, resulted in 25-fold increased affinity and a 5-fold-higher basal level of signaling that was equal to about 19% of the α-factor induced maximum signal. Mutation of the adjacent residue, S254L, caused weaker constitutive activity and a 5-fold increase in affinity. Comparison of nine different mutations affecting Ser²⁵⁴ showed that an S254F mutation caused higher constitutive activity, suggesting that a large hydrophobic amino acid residue at position 254 alters transmembrane helix packing. Thus, these studies indicate that Gln²⁵³ and Ser²⁵⁴ are likely to be involved in intramolecular interactions with other TMDs. Furthermore, Gln²⁵³ and Ser²⁵⁴ fall on one side of the transmembrane helix that is on the opposite side from residues that do not cause constitutive activity when mutated. These results suggest that Gln²⁵³ and Ser²⁵⁴ face inward toward the other TMDs and thus provide the first experimental evidence to suggest the orientation of a TMD in this receptor. Consistent with this, we identified two residues in TMD7 (Ser²⁸⁸ and Ser²⁹²) that are potential contact residues for Gln²⁵³ because mutations affecting these residues also cause constitutive activity. Altogether, these results identify a new domain of the α-factor receptor that regulates its ability to enter the activated conformation.

The α-factor mating pheromone receptor that promotes the conjugation of the yeast Saccharomyces cerevisiae belongs to the large family of G protein-coupled receptors (GPCRs) (13). Receptors in this family transmit the signals for a wide range of stimuli, including light, hormones, and neurotransmitters (63). GPCRs function in a similar manner because they transduce their signal across the plasma membrane by activating a heterotrimeric guanine nucleotide binding protein (G protein) (6). In the case of the pheromone pathway, identification of the components that transduce the signal has been greatly facilitated by the genetic accessibility of yeast (15, 36, 37). Interestingly, analysis of these components has demonstrated that they are remarkably similar to the signaling components in mammalian cells including the receptors, G protein subunits, an RGS protein that regulates G protein signaling, protein kinases that form a mitogen-activated protein kinase (MAP kinase) cascade, and transcription factors (13, 22, 37).

Although the α-factor receptor and other GPCRs respond to a wide range of diverse stimuli, it is interesting that they share a common structural organization in that they are all composed of seven transmembrane domains (TMDs) that are connected by intracellular and extracellular loops (6). The functional domains of the α-factor receptor are also similar to other GPCRs. For example, the core region of the α-factor receptor encompassing the seven TMDs carries out ligand binding and G protein activation (32, 47, 53). In addition, the third intracellular loop functions in G protein coupling (11, 49, 58), and the cytoplasmic C terminus of the α-factor receptor is a target for negative regulation by phosphorylation (10, 16, 24). Unfortunately, GPCRs do not contain significant sequence similarity across the whole receptor family to help identify the functionally important residues. In spite of this lack of sequence similarity, many mammalian GPCRs can activate the pheromone signal pathway if they are expressed in yeast (28, 43, 44). The ability of GPCRs to activate G protein signaling in such heterologous cell types indicates that there are underlying similarities in the mechanisms of GPCR activation that can be explored with the help of the genetic approaches possible in yeast.

The mechanisms of GPCR activation are poorly understood, but several different approaches indicate that the TMDs play an important role in signaling. Mutagenesis studies aimed at identifying the functional domains of GPCRs have shown that the TMDs play a key role in transducing the signal across the plasma membrane (6, 13). Biochemical studies indicate that ligand binding induces specific conformational changes in the TMDs that are associated with the activated state of the receptors (14, 17, 25). In the case of the α-factor receptor, a ligand-induced conformational change was identified by studies which showed that the binding of α-factor to the receptor increased the accessibility of the third intracellular loop to protease digestion (7). Analysis of constitutively active GPCR mutants that signal in a ligand-independent manner suggests that the effect of ligand binding may be to relieve constraints on receptor structure and allow for isomerization to the activated conformation (30, 38, 48, 50). Altogether, these studies indicate that receptor activation results from a change in the organization and packing of TMDs (6). However, the hydrophobic nature of the TMDs has made it difficult to analyze specific structural changes, and the key changes that promote G protein activation are not known.

Polar residues in TMDs are given special emphasis in models for receptor function because of their potential to form intramolecular contacts that can determine receptor conformation (2). Polar residues are not usually found in hydrophobic membrane spanning segments, but the polar residues in GPCRs can be shielded from the nonpolar membrane environment by facing inward toward the other TMDs, since the seven TMDs of GPCRs are thought to be clustered in a bundle (46, 61). This arrangement allows some of the polar residues to form intramolecular contacts between TMDs that can play a key role in receptor structure. As an example, a salt bridge between residues in TMD3 and TMD7 of rhodopsin helps to maintain photoreceptors in the inactive state (48). A polar region at the base of TMD3, termed the DRY motif, has also been implicated in the function of some GPCRs (42, 50). However, these domains are not conserved in all receptors and are not present in the α-factor receptor. In general, there are only limited data on the function of polar residues in TMDs, and their function in the α-factor receptor has not been analyzed. Therefore, we investigated the effects of mutating polar residues in TMD6 of the α-factor receptor. The polar residues in TMD6 were specifically targeted for analysis, because a previous study showed that a P258L mutation in TMD6 caused constitutive receptor activity, indicating that this region of the α-factor receptor is important for receptor function (33). Furthermore, these polar residues are mainly located in the cytoplasmic half of TMD6 and are adjacent to the third intracellular loop of the receptor that functions in G protein activation. The analysis of the effects of these mutations on ligand binding and G protein signaling in this study identified a novel polar region of TMD6 that plays an important role in receptor signaling.

MATERIALS AND METHODS

Strains and media.

The yeast strains used are described in Table 1. Cells were grown in media described by Sherman (56). Plasmid-containing cells were grown in synthetic medium containing adenine and amino acid additives but lacking uracil to select for plasmid maintenance. Yeast transformations were performed by the lithium acetate method (52). Deletion of the α-factor genes in strain JKY127-36-1 was accomplished by introducing a mfα1::LEU2 mutation (35) and a mfα2::his5⁺ mutation. In order to create the mfα2::his5⁺ allele, primers that contained 50 bp of homology to the sequences flanking the M F α2 gene followed by 17 bp flanking the Schizosaccharomyces pombe his5⁺ gene in plasmid pME3 (kindly provided by N. Dean) were used in a PCR with pME3 as the template. The resulting mfα2::his5⁺ DNA was used to transform his3⁻ S. cerevisiae cells, and the replacement of MFα2 with his5⁺ sequences was selected for by plating cells on medium lacking histidine, since the S. pombe his5⁺ gene will complement a his3 mutation in S. cerevisiae. Deletions were confirmed by PCR analysis of the genomic DNA and also by a backcross of this strain which showed that the combination of the mfα1::LEU2 and the mfα2::his5⁺ mutations blocked the ability of MATα cells to produce α-factor.

TABLE 1.

Yeast strains used

Strain	Genotype
JKY78	MATa far1 bar1::hisG ste2::LEU2 lys2::FUS1-lacZ arg4 his3 leu2 lys2o trp1 ura3
JKY79	MATa bar1::hisG ste2::LEU2 ste5-3ts lys2::FUS1-lacZ ade2-1o ade3 cry1 his4-580a leu2 lys2o trp1a tyr1o ura3 SUP4-3ts
JKY127-36-1	MATa bar1::hisG far1 sst2-1 ste2Δ mfα1::LEU2 mfα2::his5+ade2 his3 leu2 ura3
lys1α	MATα lys1

Mutagenesis of the α-factor receptor gene.

Mutations were introduced into the α-factor receptor gene (STE2) by PCR. Taq DNA polymerase and all other PCR reagents were purchased from Boehringer Mannheim. Plasmid pDB02 (STE2 CEN3 URA3 ARS1) (33) was used as template for all PCRs. Plasmid pDB02 was constructed by using PCR to add an SphI site 5′ of the STE2 coding region (minus 837 bp) and a SacI site 3′ of the coding region (695 bp downstream) and then inserting the modified STE2 fragment into pJK67, which was derived from YCplac33 (18) by destroying the endogenous AatII site by deleting the 3′ overhang. The PCR primers were complementary to the STE2 sequence, except for the noted codon change required to introduce the mutation. After PCR, the 276-bp AatII-ClaI fragment containing the desired change was subcloned into pDB02 to create ste2-S251L (TCA→CTA), ste2-C252L (TGT→CTT), ste2-Q253L (CAA→CTA), and ste2-S254L (TCT→CTT). ste2-S259L was created by first subcloning the 605-bp AatII-PstI fragment of STE2 into the AatII-PstI sites of pBR322 to create pPD03. A 283-bp AatII-SspI PCR fragment containing the S259L (TCG→CTG) change was subcloned into pPD03, and then the 605-bp AatII-PstI fragment of pPD03 was subcloned into pDB02 to create ste2-S259L. ste2-S254F was isolated by genetic screening based on the ability of this mutant to activate a FUS1-lacZ reporter gene in the absence of α-factor, and then this mutant receptor gene was sequenced to identify the mutation. The methods used in this genetic screen were similar to those used previously to isolate the constitutive mutant STE2-P258L (33). To verify that the S254F mutation accounted for the constitutive activity, an AatII-ClaI fragment carrying this mutation was subcloned into pDB02. Receptor genes containing other mutations at position 254 were constructed with a heterogeneous PCR primer that randomly introduced all four bases at the three positions of codon 254. After PCR, the 276-bp AatII-ClaI fragment was subcloned into pDB02. This pool of mutant plasmids was introduced into yeast strain JKY78, and then the cells were tested for mating ability. DNA sequence analysis of receptor plasmids recovered from 18 mating-competent strains that showed constitutive activity identified nine different substitution mutations at position 254. Mutations that resulted in the substitution of Ser²⁸⁸, Ser²⁹², and Ser²⁹³ with alanine were constructed with the SURE CHANGE mutagenesis kit (Stratagene). Plasmid pDB02 was used as the template for single mutants, and a pSTE2-Q253L plasmid was used as the template for the double mutants. Mutagenic oligonucleotides were designed according to manufacturer’s instructions and were complementary to the STE2 sequence, except for the indicated changes required to change the appropriate serine codon to alanine: STE2-S288A (TCT→GCT), STE2-S292A (TCA→GCA), and STE2-S293A (TCA→GCA). Plasmids with STE2 under the control of the galactose-inducible GAL1 promoter were created by subcloning the 605-bp AatII-PstI fragment from each of the mutants into pJK57. pJK57 was constructed by using PCR to add a BamHI site to the 5′ end of the STE2 open reading frame and a SacI site to the 3′ end so that it could be cloned into pCTG2 downstream of the GAL1 promoter. pCTG2 was a gift from Phil James and contains a 690-bp EcoRI-BamHI fragment containing the GAL1 promoter inserted into pRS314 (57). All of the mutations constructed in this study were confirmed by dideoxy sequencing of the double-stranded DNA with Sequenase (U.S. Biochemical).

α-Factor receptor analysis.

Western immunoblots were carried out essentially as described elsewhere (32). Mid-logarithmic-phase cells (2.5 × 10⁸) were harvested and lysed by agitation with glass beads in 250 μl of lysis buffer (2% sodium dodecyl sulfate [SDS], 100 mM Tris [pH 7.5], 8 M urea). A 100-μg amount of protein extract, as determined by the bicinchoninic acid protein assay kit (Pierce), was separated by electrophoresis on an 10% SDS-polyacrylamide gel, electrophoretically transferred to nitrocellulose, and probed with rabbit anti-Ste2p antibodies (32). Immunoreactive proteins were detected by chemiluminescence with an ECL kit (Amersham). α-Factor binding assays were conducted essentially as described elsewhere (49). Logarithmic-phase cells were collected by centrifugation, washed twice with ice-cold inhibitor medium (IM; YEPD medium containing 10 mM KF and 10 mM NaN₃), and resuspended at a density of 10⁹ cells/ml. A 50-μl volume of cells was mixed with 50 μl of ³⁵S-α-factor and incubated for 30 min, and then the cells were collected on a Whatman GF/C filter and the unbound α-factor was removed by washing. Nonspecific binding was determined by performing reactions in the presence of a 100-fold excess of cold α-factor. Scatchard plots represent the averages of three to six independent assays done in duplicate. Similar results were obtained when the data were plotted by the method of Klotz (31). ³⁵S-labeled α-factor was purified from the supernatant of MATα cells labeled with [³⁵S]SO₄ by chromatography on a Bio-Rex 70 column as described previously (49).

α-Factor-induced responses.

To examine the ability of cells to mate, patches of yeast strain JKY78 carrying the indicated STE2 allele on a plasmid pJK67 were replica plated to YPD plates containing a lawn of MATα cells (lys1α). The cells were incubated at 30°C for 12 h to allow mating and were then replica plated to minimal plates lacking amino acids, incubated at 30°C for 48 h to select for the growth of diploids, and then photographed. Quantitative mating assays were performed by mixing 3 × 10⁶ lys1α cells with various dilutions of MATa cells containing the indicated mutant receptor plasmids on a synthetic medium plate lacking lysine. The plates were incubated at 30°C for 72 h, and then efficiency of mating was determined by counting the number of diploid colonies that were formed at each dilution of cells. To assay induction of FUS1-lacZ in far1⁻ cells, cultures were grown overnight to logarithmic phase in selective medium, diluted to 4 × 10⁶ cells/ml and incubated with the indicated concentrations of synthetic α-factor (Bachem) for 2 h. Inductions were stopped by adding cycloheximide (final concentration, 10 μg/ml), and then β-galactosidase assays were performed by using the colorimetric substrate O-nitrophenyl-β-d-galactopyranoside (ONPG) as described elsewhere (41). Basal levels of FUS1-lacZ expression were determined as described above, except that the cells were incubated in the absence of α-factor. To assay the effects of galactose-induced expression of receptor mutants, the cells were first grown overnight in synthetic medium containing raffinose and diluted to 4 × 10⁶ cells/ml, galactose was added to the medium at a final concentration of 2%, and then after a 5-h incubation the cells were assayed for FUS1-lacZ induction as described above. Induction of FUS1-lacZ was assayed in ste5-3^ts cells (JKY79) that were grown overnight to logarithmic phase in selective medium at 34°C which was adjusted to 2.5 × 10⁶ cells/ml and then shifted to 23°C for 8 h. α-Factor was added to a final concentration of 10⁻⁶M to induce signaling in the wild-type cells. Inductions were stopped by adding cycloheximide (final concentration, 10 μg/ml), and then β-galactosidase assays were performed in duplicate with the colorimetric substrate chlorophenyl red-β-d-galactopyranoside (CPRG). The results of at least two independent assays, each done in duplicate, are reported.

RESULTS

Mutation of polar residues in TMD6.

Our previous studies implicated TMD6 in α-factor receptor activation because a mutation of Pro²⁵⁸ to Leu (P258L) in TMD6 caused constitutive receptor activity in the absence of α-factor (33). To analyze the mechanisms of receptor activation further, we screened for additional constitutive mutations and identified another mutation in TMD6, S254F, that also caused constitutive signaling (see Materials and Methods). Interestingly, Ser²⁵⁴ is one of five polar residues found in TMD6 that may interact with the other TMDs to influence receptor structure. Therefore, the function of the polar amino acids in TMD6 of the α-factor receptor was examined by mutating the corresponding codons in the receptor gene (STE2) to leucine codons by site-directed mutagenesis (see Materials and Methods). As shown in Fig. 1, the most polar amino acids in TMD6 include Ser²⁵¹, Cys²⁵², Gln²⁵³, Ser²⁵⁴, and Ser²⁵⁹. In addition, we also studied the ste2-S254F mutant that was isolated by genetic screening for constitutive α-factor receptor mutants. To confirm that the mutant receptor proteins were produced, the plasmids carrying the mutant receptor genes were introduced into a yeast strain (JKY78) lacking the wild-type STE2 gene, and then cell extracts were analyzed by Western blotting (Fig. 2). As expected, the wild-type receptor protein (Ste2p) was detected as multiple bands on Western blots due to the N-linked glycosylation of the receptor (4). The mutant receptor proteins displayed a similar heterogeneous pattern, indicating that they entered the secretory pathway efficiently. All of the mutants appeared to produce similar levels of receptor protein, which was slightly less than the wild-type level.

FIG. 1.

Polar residues in TMD6 of the α-factor receptor. The predicted two-dimensional topology of the receptor in the plasma membrane, with the extracellular domain at the top, is shown. The expanded view of TMD6 details the relative positions of amino acid residues. The polar amino acid residues investigated in this study are in boldface.

FIG. 2.

Immunoblot analysis of mutant receptor proteins. Extracts of yeast strain JKY78 carrying the indicated wild-type or mutant STE2 allele on plasmid vector pJK67 were separated by electrophoresis on an SDS-polyacrylamide gel, transferred to nitrocellulose, and probed with affinity-purified rabbit anti-Ste2p antibodies (32). Immunoreactive proteins were detected by chemiluminescence. The relative molecular weights of prestained molecular weight markers (BioRad) are noted to the left.

Ligand responsiveness of mutant receptors.

The overall ability of the mutant receptors to function was assessed by examining the mating ability of JKY78 cells carrying the mutant STE2 genes on plasmids. As shown in Fig. 3, MATa cells carrying the wild-type STE2 gene mated with MATα cells, as evidenced by formation of diploid cells that grew on selective medium. Similarly, all of the cells carrying mutant receptor genes mated. Some of the mutant strains appeared to mate poorly compared to wild type; thus, quantitative mating assays were carried out (Fig. 3). Of the group, the ste2-S259L mutant mated the most poorly, with an efficiency of 1.4% of that of the wild type. The ste2-S251L and -S254F mutants mated with slightly better efficiency (7.5 and 5.4%, respectively), and the other mutants (ste2-C252L, -Q253L, and -S254L) mated essentially as well as the wild type. Thus, all of the mutants are at least partially functional to carry out the full range of receptor activities.

FIG. 3.

Mating efficiencies of receptor mutants. Mating abilities were tested by replica plating patches of MATa strain JKY78 carrying the indicated wild-type or mutant STE2 allele on plasmid vector pJK67 to a lawn of MATα cells (lys1α). After the cells were allowed to mate, they were replica plated to selective plates for 2 days at 30°C, and then the growth of diploid cells that were formed by mating was photographed. Mating efficiencies are expressed as a percentages of wild type as determined by quantitative mating analysis.

To examine the function of the mutant receptors in more detail, dose-response assays were carried out to quantify their ability to induce a reporter gene in response to α-factor. This analysis was also carried out in the JKY78 yeast strain, because it carries a pheromone-responsive FUS1-lacZ reporter gene (60) that provides a quantitative assay for the induction of signaling. As shown in Fig. 4, the ste2-S251L, -C252L, -Q253L, and -S254L mutants could be induced to approximately the same maximum level as wild-type cells when treated with a dose of α-factor that is essentially saturating for wild type (10⁻⁷M). The exceptions were the ste2-S254F mutant, which was induced only to about 80% of the maximum wild-type level, and the ste2-S259L mutant, which was induced only to about 20% of the maximum. Although some mutants were partially defective, these results indicate that none of the polar residues in TMD6 were absolutely required for the receptors to respond to α-factor. Interestingly, the ste2-Q253L and -S254F mutants displayed an increased level of basal signaling that demonstrates a role for these residues in preventing spontaneous activation of the receptor in the absence of α-factor (Fig. 4A). The constitutive activity of the mutant receptors will be described in detail below.

FIG. 4.

Ligand-dependent activation of a FUS1-lacZ reporter gene by mutant receptors; dose-response assays for yeast strain JKY78 carrying the indicated wild-type (WT) or mutant STE2 allele on plasmid vector pJK67. Logarithmic-phase cells were incubated with the indicated concentrations of α-factor for 2 h and then assayed for β-galactosidase activity to measure the induction of the pheromone-responsive FUS1-lacZ reporter gene. The results represent two to six independent assays, each done in duplicate. The standard deviation was less than 8% for all data points. (A) Results for the ste2-Q253L and -S254F mutants; (B) results for the ste2-S251L, -C252L, -S254L, and -S259L mutants.

Ligand-binding properties of mutant receptors.

Radioligand-binding assays were performed to determine whether mutations affecting the polar residues altered the ligand-binding properties. Ligand binding is a sensitive measure of the structure of mutant receptors, because the α-factor binding pocket is formed by the region of the receptor that contains the seven TMDs. Equilibrium-binding assays were carried out on whole cells with ³⁵S-labeled α-factor, and then the data were analyzed by the Scatchard method. Scatchard plots for the wild type and the mutants that had the strongest effect on binding (ste2-Q253L and -S254F) are shown in Fig. 5. Binding data for the wild type and for all of the mutants are summarized in Table 2. The K_d for wild-type receptors was 10 nM, which is in agreement with previous studies (27, 33). Interestingly, the mutants all bound α-factor with an affinity equal to or greater than that of the wild-type receptors. The mutation affecting the most polar residue in TMD6, Q253L, had the strongest effect on binding. The K_d for the binding of α-factor to ste2-Q253L cells was 25-fold lower than that for wild-type cells, indicating that these cells have significantly increased affinity for α-factor. The mutations affecting Ser²⁵⁴ also increased binding affinity: the ste2-S254F and the ste2-S254L mutants showed 11- and 5-fold increased affinities, respectively. The ste2-S259L mutant displayed 3.7-fold increased affinity, and the ste2-S251L and -C252L mutants bound α-factor with about the same affinity as that of the wild type. Thus, none of the polar residues were essential for ligand binding, and, in fact, binding affinity was improved in some mutants. Increased binding affinity for the mutant receptors is interesting, because an increase in binding affinity is often associated with the activated conformation of GPCRs (5, 38, 50).

FIG. 5.

Radioligand-binding studies. Scatchard plot analysis of wild-type (A), Q253L (B), and S254F (C) receptors. Equilibrium binding studies were carried out by measuring the binding of ³⁵S-α-factor to strain JKY78 carrying the indicated wild-type or mutant STE2 allele on plasmid vector pJK67. The small upper panels are plots of the bound versus the free α-factor under the conditions used. The lower panels show the results plotted by the Scatchard method. Lines, best fits of the data determined by the least-squares method; data points, averages of three to eight independent assays, each done in duplicate.

TABLE 2.

Pharmacological properties of mutant receptors

Receptor	Kd (nM)a	No. of receptors per cella	EC50 (nM)b	Maximum α-factor induced activity (%)b
Wild type	10.1 ± 0.7	11,000 ± 300	3.5	100
S251L	15.0 ± 6.1	810 ± 60	4.0	110
C252L	5.2 ± 1.7	2,400 ± 630	2.2	106
Q253L	0.4 ± 0.2	170 ± 8	0.5	110
S254L	1.9 ± 1.0	2,290 ± 168	2.0	93
S254F	0.9 ± 0.6	90 ± 13	1.2	80
S259L	2.8 ± 0.7	160 ± 3	4.0	20

^aLigand-binding properties of yeast strain JKY78 carrying the indicated receptor genes on a plasmid were analyzed in equilibrium binding assays with ³⁵S-labeled α-factor. The K_ds and the numbers of receptors per cell were determined by Scatchard plot analysis. 

^bEC₅₀s and maximum α-factor-induced values were extrapolated from dose-response curves for FUS1-lacZ induction. Results are the averages of two to six independent assays, each done in duplicate (± standard deviations). The α-factor-induced activity of wild-type receptor cells was set to 100%. 

To assess the efficiency of receptor function further, the K_d for binding was compared to the abilities of the mutant receptors to be stimulated by α-factor. To facilitate this comparison, the dose-response curves shown in Fig. 4 were analyzed to determine the ligand concentration that gave the half-maximal induction (EC₅₀) of the reporter gene (Table 2). The EC₅₀ for the wild type was 3.5 nM, which corresponds well with the dissociation constant for α-factor binding (K_d = 10 nM). The ste2-Q253L and -S254F mutants showed EC₅₀s of 0.5 and 1.2 nM, respectively, indicating a greater sensitivity to α-factor for these mutants. This increased sensitivity correlates well with the increased affinity of these mutants for α-factor described above. The EC₅₀s for the other mutants were within twofold of the wild type value. Even the S254F and S259L mutants that were not able to give the same maximum level of induction as the wild type showed EC₅₀s that were less than or equal to that of the wild type. These results demonstrate that mutation of a subset of the polar residues in TMD6 causes a coordinate increase in the binding and signaling of the α-factor receptor.

Radioligand binding studies were also used to determine the number of cell surface receptors (Table 2). The maximum binding value obtained by Scatchard analysis indicated that wild-type cells contained approximately 11,000 receptors, as expected (27, 33). In contrast, all of the mutants contained fewer cell surface-binding sites. The mutation affecting the most polar residue in TMD6 again had a strong effect, since the number of receptors was reduced by about 65-fold in the ste2-Q253L mutant. The effects of mutating the adjacent Ser²⁵⁴ codon varied with the substituted amino acid, since the ste2-S254F mutant showed more than 100-fold fewer cell surface binding sites and the ste2-S254L mutant was decreased only by about 5-fold. The magnitude of the decrease in receptor number did not correlate with the altered affinity of the mutant receptors, because the ste2-S259L mutant, which showed a relatively minor increase in affinity, contained about 70-fold fewer surface receptors. Furthermore, the ste2-S251L and -C252L mutants did not display a significant change in affinity, and yet they contained 14- and 5-fold decreased receptor numbers, respectively. The decreased cell surface expression may be a result of less efficient transport to the cell surface, as has been observed for other mutant receptors (26, 59), or of decreased stability of receptors at the cell surface, since some of the mutants may mimic the ability of ligand binding to promote rapid endocytosis of α-factor receptors (23, 49). The observation that the decreased surface receptor number did not prevent the detection of signaling by the mutant receptors is consistent with previous studies indicating that the number of receptors is not limiting for signaling in wild-type cells (32, 54).

Constitutive receptor activity.

The dose-response assays shown in Fig. 4 indicated that at least two of the mutants, ste2-Q253L and -S254F, displayed elevated levels of basal signaling. To facilitate the analysis of this ligand-independent signaling activity, plasmids containing mutant receptor genes were transformed into a yeast strain (JKY79) that carries a temperature-sensitive mutation in a postreceptor component of the pheromone signaling pathway, ste5-3^ts (21). This strain also carries a pheromone-responsive FUS1-lacZ reporter gene and lacks the chromosomal copy of the receptor gene to prevent interference from the wild-type receptors. The advantage of incorporating the ste5-3^ts mutation into the strain is that it allows for propagation of the cells without activation of the pheromone pathway by growing the cells at the restrictive temperature (34°C), at which there is essentially no basal activation of the FUS1-lacZ reporter gene. The pheromone signal pathway can then be made competent by shifting the cells to the permissive temperature (23°C). When the ligand-independent activation of the FUS1-lacZ reporter gene was determined at 23°C, two of the mutants, ste2-Q253L and -S254F, displayed significantly elevated basal levels of signaling (Table 3). The basal activity of ste2-Q253L was elevated by about 5-fold, and that of ste2-S254F was elevated 2.4-fold compared to the wild type.

TABLE 3.

Constitutive activities of α-factor receptor mutants

Receptor	β-Galactosidase activity (U)a	Fold increase in basal activityb	Basal activity relative to maximum α-factor-induced activityc
Wild-type STE2	2.7 ± 0.2		3.8
S251L	1.5 ± 0.1	0.5	2.0
C252L	1.4 ± 0.1	0.5	1.9
Q253L	13.7 ± 0.6	4.9	19.0
S254L	2.1 ± 0.1	0.7	2.9
S254F	6.6 ± 0.7	2.4	9.2
S259L	2.5 ± 0.1	0.9	3.4

^aBasal signaling activities in yeast strain JKY79 carrying the indicated receptor genes on a plasmid were assayed. The activity of a FUS1-lacZ reporter gene was determined by quantifying β-galactosidase activity. Results are the averages of two to six independent assays, each done in duplicate (± standard deviations). 

^bRatio of mutant basal activity to wild-type basal activity. 

^cComparison of the ligand-independent basal activities of the mutants to the maximum ligand-induced activities of wild-type cells treated with 10⁻⁶ M α-factor (72.1 ± 1.7 U). 

The ste5-3^ts strain was then used to compare the ligand-independent constitutive signaling of the mutant receptors to the ligand-induced activity of wild-type receptors. For this comparison, cells carrying the wild-type receptor were shifted to the permissive temperature in the presence of a saturating dose of α-factor (10⁻⁶ M) to determine the maximum level of FUS1-lacZ activation. This level of activity was then compared to the basal signaling activity in the receptor mutant strains in the absence of α-factor. As shown in Table 3, the basal activity of ste2-Q253L was equivalent to about 19% of the maximum ligand-induced level, and the basal level of ste2-S254F was equivalent to about 10% of the maximum. This level of signaling activity is comparable to those of the previously identified constitutive mutants, STE2-P258L and STE2-P258L-S259L, whose basal signaling activities were equivalent to about 12 and 45%, respectively, of the ligand-induced maximum (33). These results indicate that the Q253L and S254F mutations affect the structure of the receptor so as to partially mimic the activated conformation of the receptor. Thus, a subset of the polar residues in TMD6 is required to prevent receptors from entering the activated conformation in the absence of ligand.

To examine whether the number of receptors was limiting for detection of constitutive signaling, we tested the effects of overproducing the mutant receptors. For this analysis, we introduced the mutant receptor genes into cells on multicopy plasmid vectors, and in separate studies we also placed the mutant receptor genes under control of a strong galactose-inducible GAL1 promoter that results in approximately 30-fold overproduction of receptors. The results showed that overproduction did not help to reveal constitutive activity for the other receptor mutants and had only small positive effects on the basal levels of signaling for the ste2-Q253L and -S254F mutants (data not shown).

Basal levels of signaling in an sst2⁻ adaptation-defective strain.

In order to gain increased sensitivity for the analysis of the basal signaling by the mutant receptors, we analyzed the effects of the mutants in an adaptation-defective sst2-1 strain. These cells are more sensitive to α-factor because of the inability of Sst2p to regulate the pheromone-responsive G_α subunit (1, 8). To help maintain a low basal level of signaling in these cells, the genes encoding α-factor were deleted. This prevents the rare cells that switch mating type from producing α-factor that could otherwise stimulate this supersensitive strain. The cells also carry a mutation in the FAR1 gene to prevent pheromone-induced cell division arrest (9). Analysis of the ste2-Q253L and -S254F mutants in this strain showed that they both caused greater than a 10-fold increase in the basal level of signaling (Fig. 6). These results demonstrate that the elevated basal level of signaling by these mutants is not due to autocrine signaling, because the α-factor genes are deleted from this strain. Furthermore, the detection of elevated basal signaling by the ste2 mutants in an sst2⁻ strain shows that the constitutive receptor activity is independent of SST2 and is not a consequence of a defect in SST2-mediated adaptation. Interestingly, the ste2-S254L mutant showed a readily detectable sixfold increase in basal signaling in this sst2⁻ strain that was not obvious in the SST2⁺ strain. The elevated level of basal signaling detected for the ste2-S254L mutant in this strain demonstrates that the sst2-1 mutation had the predicted effect of increasing the sensitivity of the assay. In spite of this, the other receptor mutants (ste2-S251L, -C252L, and -S259L) did not show a significant increase in basal activity. These results demonstrate that in comparison to the other polar residues in TMD6, Gln²⁵³ and Ser²⁵⁴ play a special role in receptor function.

FIG. 6.

Analysis of the basal signaling levels by receptor mutants in an sst2⁻ adaptation-defective strain. Supersensitive sst2-1 strain JKY127-36-1 carrying the indicated wild-type or mutant STE2 allele on plasmid vector pJK67 was assayed for β-galactosidase activity to measure the basal level of the pheromone-responsive FUS1-lacZ reporter gene in the absence of added α-factor. The results are averages of two independent assays done in duplicate (± standard deviations).

Effects of different substitution mutations at residue 254.

The observation that substitution of Ser²⁵⁴ with Phe caused greater constitutive activity than substitution with Leu suggested that the type of amino acid substituted at position 254 could influence the degree of constitutive activity. To examine this in more detail, site-directed mutagenesis was used to introduce all possible codon combinations at this position, and then the pool of mutant receptor plasmids was introduced into yeast cells for analysis (see Materials and Methods). Yeast colonies that contained functional receptors were identified by testing for mating ability, and then nine different constitutive mutations were identified after sequencing 18 independent isolates. The basal level of signaling was at least slightly increased in all of the mutant strains (Fig. 7). Substitution of Ser²⁵⁴ with Ala, Asp, Val, Gln, or Tyr all caused about a 2-fold increase in the basal level of signaling, indicating that the loss of serine at position 254 promoted higher basal levels of signaling. Higher levels of basal signaling (3- to 7.5-fold) were observed when Ser²⁵⁴ was substituted with Gly, Leu, Phe, or Trp, indicating that the insertion of these residues caused an additional effect on constitutive signaling. All of the mutants with nonaromatic amino acids at residue 254 (Gly, Ser, Ala, Asp, Val, Gln, and Leu) displayed the ability to maximally induce a reporter gene (Fig. 4 and data not shown), indicating that these receptors were not obviously impaired in signaling. However, substitution of Ser²⁵⁴ with aromatic amino acids (Phe, Tyr, or Trp) resulted in a partial defect in α-factor-induced signaling (Fig. 4 and data not shown), indicating that the substitution of Ser²⁵⁴ with these larger amino acids altered receptor structure in a way that also adversely affected G protein activation. Thus, the basal signaling activity of this latter group of mutants may be slightly lower than might otherwise be expected based on the correlation with the type of amino acid substituted at position 254 (see Discussion). Altogether, the differential abilities of the various substitution mutants to promote ligand-independent activation of the receptor indicate that Ser²⁵⁴ is located within a region of the receptor that can alter the structure of the TMDs in a way that mimics the activated state of the receptor.

FIG. 7.

Constitutive signaling activity of Ser²⁵⁴ substitution mutants. Strain JKY127-36-1 carrying the wild-type STE2 allele or the indicated substitution mutations affecting Ser²⁵⁴ on plasmid vector pJK67 was assayed for β-galactosidase activity in the absence of α-factor to determine the basal levels of the pheromone-responsive FUS1-lacZ reporter gene. The different mutants are listed in order of increasing size of the residue substituted at position 254. Results are fold increases in basal activity over that of the wild type and are the averages of two independent assays, each done in duplicate (± standard deviations).

Analysis of potential contact residues.

The phenotypes caused by mutations affecting Gln²⁵³ and Ser²⁵⁴ suggest that these residues may form important contacts with the other TMDs. Inspection of the α-factor receptor sequence indicates that there are more than 40 residues in the predicted TMDs that may be capable of forming hydrogen bond interactions. As an approach to identifying potential contact residues for Gln²⁵³ and Ser²⁵⁴, we reasoned that mutations affecting the contact residues should cause constitutive activity similar to the effects of mutating Gln²⁵³ and Ser²⁵⁴. To focus our efforts on identifying residues that may contact TMD6, we analyzed the effects of mutating the polar residues in TMD5 and TMD7, since they are likely to be adjacent to TMD6 in the receptor (2). In addition, we further narrowed our analysis to examining the effects of mutating the residues in the cytoplasmic halves of TMD5 and -7 that are likely to be in a position to interact with either Gln²⁵³ or Ser²⁵⁴ in TMD6.

Analysis of the three polar residues in the cytoplasmic half of TMD7 (Ser²⁸⁸, Ser²⁹², and Ser²⁹³) was carried out by mutating the corresponding codons to alanine (see Materials and Methods). The only polar residue in the cytoplasmic half of TMD5 (Lys²²⁵) was previously mutated to Cys as part of a separate study in our lab and found not to display significant constitutive activity (unpublished data), although it should be noted that these studies were carried out with different plasmid vectors and yeast strains. In contrast, the ste2-S288A mutant and the ste2-S292A mutants displayed about 3- and 4.5-fold increased basal signaling activities, respectively, in the sst2⁻ strain (JKY127-36-1) (Fig. 8A). The constitutive activity was specific to these mutants, since the ste2-S293A mutant did not display elevated basal levels of signaling. The mutations affecting the three serine residues in TMD7 did not appear to cause deleterious effects on receptor signaling, because the mutant strains could induce a FUS1-lacZ reporter gene to maximum levels when treated with α-factor (Fig. 8B). Furthermore, these receptor mutants also showed EC₅₀s similar to that for the wild type for induction of the reporter gene (not shown). These results demonstrate further that a specific subset of the polar residues in the α-factor receptor is required to prevent ligand-independent receptor activation.

FIG. 8.

Mutational analysis of polar residues in TMD7. Strain JKY127-36-1 carrying the indicated wild-type or mutant STE2 allele on plasmid vector pJK67 was assayed for β-galactosidase activity to determine the relative levels of the pheromone-responsive FUS1-lacZ reporter gene. Cells were assayed in the absence of α-factor (A) and in the presence of 10⁻⁶ M α-factor (B) to determine the basal and pheromone-induced levels, respectively. The results are the averages of two independent assays, each done in duplicate (± standard deviations).

Since we predict that Gln²⁵³ may face in the direction of TMD7 (see Discussion), it is possible that the Ser²⁸⁸ and Ser²⁹² residues may contact Gln²⁵³. This contact could be stabilized by hydrogen bond formation with the hydroxyl groups on the serine side chain acting as hydrogen bond donors and the oxygen on the glutamine side chain acting as an acceptor. Therefore, we carried out double mutant analysis to help determine the relationship between the mutations affecting Gln²⁵³ and the three serine residues in TMD7. The rationale for this approach was that the effects of two different constitutive receptor mutations should not be additive if they affect the same receptor contact and should have additive effects if they affect different intramolecular contacts. The results of this analysis showed that none of the double mutant combinations displayed significantly greater basal signaling activity than the ste2-Q253L mutant alone (Fig. 8A). Two of the double mutants, ste2-Q253L-S292A and ste2-Q253L-S293A, showed basal activity that was slightly lower than that of the ste2-Q253L single mutant. This lower basal level of signaling did not appear to be due to a defect in receptor signaling, because all of the double mutants could be induced efficiently with α-factor (Fig. 8B and data not shown). Thus, the failure of the S288A and S292A mutations to further increase the constitutive activity caused by the Q253L mutation identifies Ser²⁸⁸ and Ser²⁹² as potential contact residues for Gln²⁵³.

DISCUSSION

In order to gain insight into the mechanisms of GPCR activation, we investigated the effects of mutating the polar residues in TMD6 of α-factor receptor. The function of polar residues was examined because they have the ability to form contacts between TMDs that can play a key role in receptor structure. We focused on the polar residues in TMD6 because of the previous identification of constitutive receptor mutations in this domain. Also, TMD6 is adjacent to the third intracellular loop that plays a key role in G protein activation. The mutant receptor proteins were produced at levels similar to that for the wild type and were glycosylated, indicating that they enter the secretory pathway efficiently. However, the mutant cells contained fewer surface receptors for α-factor than wild-type cells. This decrease in cell surface receptors suggests that the mutations affecting the polar residues altered receptor structure in a way that was recognized by the mechanisms that control the trafficking of receptors to or from the plasma membrane (23, 26, 59). The decrease in receptor number was not so high as to prevent detection of receptor signaling, because all of the mutant receptors showed at least partial ability to respond to α-factor. Evolutionarily conserved polar residues in TMDs are often given special emphasis in molecular models for receptor function (2, 3), so it is worth noting that two of the polar residues in TMD6 (Ser²⁵¹ and Gln²⁵³) are conserved in the pheromone receptors from Saccharomyces kluyveri (40) and Schizosaccharomyces pombe (29). Mutations affecting Gln²⁵³ had significant effects on ligand binding and signaling. In contrast, mutations affecting Ser²⁵¹ did not appear to alter receptor function aside from lowering the number of cell surface binding sites. This indicates that some conserved polar residues, such as Ser²⁵¹, may be more important for proper receptor production than for signal transduction.

Effects on ligand binding.

Previous studies have shown that α-factor binding is mediated by the central region of the α-factor receptor that contains the TMDs (32, 47). Studies of chimeric receptors formed between the α-factor receptors from S. cerevisiae and S. kluyveri further indicate that ligand binding is carried out by noncontiguous domains within this central region (39, 53). It was, therefore, very interesting that mutating the most polar residue in TMD6 to Leu (Q253L) resulted in 25-fold increased binding affinity. A mutation affecting the adjacent Ser²⁵⁴ residue (S254L) also caused a significant increase in affinity (5-fold). The mutated residues are unlikely to be directly involved in ligand binding, because two different substitutions at position 254 (S254L and S254F) both increased binding affinity. Instead, these results suggest that the mutated residues influence binding affinity indirectly by altering the packing of the TMDs. Increased binding affinity may result from the abilities of these mutants to mimic the activated state of the receptor, because they also displayed constitutive receptor activity. Many GPCRs display increased affinity for ligand when they are coupled to a G protein; thus, it has been suggested that the increased affinity of constitutive mutants is due to their ability to mimic the activated state of the receptor (38). Altogether, these results suggest that the polar residues in TMD6, particularly Gln²⁵³ and Ser²⁵⁴, influence the packing of the TMDs in a way that affects the ligand binding pocket.

Effects on constitutive signaling.

Mutations affecting a subset of the polar residues in TMD6 also altered receptor structure in a way that caused higher basal levels of signaling in the absence of α-factor. In particular, mutation of the Ser²⁵⁴ and Gln²⁵³ residues increased basal signaling. To examine the strength of this constitutive receptor activity, we made use of an ste5^ts strain (Table 3) so that we could compare the levels of constitutive signaling by the mutants with the maximum levels of α-factor-induced signaling by wild-type receptors. In these studies, we found that, in the absence of α-factor, ste2-Q253L signaled at 19% and ste2-S254F signaled at about 9% of the ligand-induced maximum. The ste2-S254L mutant displayed weaker constitutive activity that was not readily detected in the ste5^ts strain, but in a supersensitive sst2⁻ strain the ste2-S254L receptors increased basal signaling by about fivefold above that of background. In contrast, the basal signaling levels were not significantly increased by the ste2-S251L, -C252L, or -S259L mutant receptors, even in a supersensitive strain. These results indicate that Gln²⁵³ and Ser²⁵⁴ in TMD6 play a special role in stabilizing receptors in the inactive conformation and in preventing constitutive activity in the absence of ligand.

The observation that substituting Ser²⁵⁴ with Phe caused stronger constitutive activity than substitution with Leu suggested that the character of the residue at position 254 had significant effects on receptor structure and function. Therefore, we examined a set of nine different substitution mutations at residue 254. All of the mutants showed at least slightly elevated basal signaling. The simple loss of the Ser²⁵⁴ hydroxyl group was not sufficient to cause strong constitutive activity, because substitution with Ala caused only about a twofold increase in basal signaling. Substitution of a more hydrophobic amino acid at residue 254 was also not sufficient, because Val had much weaker effects than Leu. The residues that resulted in the most significant degree of constitutive activity when substituted at position 254 were Gly, Leu, Phe, and Trp. It was interesting that three of these residues are larger than ser (Leu, Phe, and Trp). This suggested that there might be some similarities between the activation of the α-factor receptor and rhodopsin, because studies of rhodopsin indicate that TMD3 and -6 move apart during activation (14). Furthermore, certain substitution mutations that introduce larger amino acid residues into TMD3 of rhodopsin affect the positioning of the retinal chromophore in a way that causes higher basal levels of signaling in the dark (20). However, substitution of Ser²⁵⁴ with Gly, the smallest amino acid, also gave high levels of constitutive activity for the α-factor receptor. This indicates either that the size of the residue at position 254 is not related to constitutive activity, or that Gly at position 254 alters transmembrane packing in a distinct way to cause constitutive activity. Nonetheless, the differential effects on basal signaling caused by the different substitution mutants demonstrate that Ser²⁵⁴ is in a conformationally sensitive region of the receptor and that the type of residue substituted at 254 strongly influences the structure of the receptor.

Orientation of TMD6.

The effects of mutating Gln²⁵³ and Ser²⁵⁴ described above, as well as the previously described effects of mutating Pro²⁵⁸ (33), indicate that these residues in TMD6 play a special role in the structure of the α-factor receptor. To examine the potential significance of these residues further, the amino acids of TMD6 were displayed on a helical wheel projection to predict their orientation in an α-helix, as shown in Fig. 9. TMD6 of the α-factor receptor is assumed to form an α-helix similar to those of the TMDs in rhodopsin and other GPCRs (2, 61). Consistent with this, a synthetic peptide corresponding to the sequence of TMD6 exhibited helical character (45). Interestingly, the residues in TMD6 that cause constitutive activity when mutated (Gln²⁵³, Ser²⁵⁴, and Pro²⁵⁸) are predicted to reside on one side of the transmembrane helix. In contrast, the polar residues that did not cause constitutive activity when mutated to leucine (Ser²⁵¹, Cys²⁵², and Ser²⁵⁹) all lie on the opposite side of TMD6. This alignment indicates that the side of TMD6 containing Gln²⁵³ and Ser²⁵⁴ is likely to be involved in interactions with other TMDs. Furthermore, Gln²⁵³ is likely to be facing inward toward the other TMDs, since it is the most polar residue in TMD6. Thus, the polar nature of Gln²⁵³ and Ser²⁵⁴ and the phenotypes caused by mutations that affect these residues indicate that they are likely to face inward toward the other TMDs and to be in a position to have cooperative effects on the packing arrangement of the other TMDs.

FIG. 9.

Helical wheel diagram of TMD6. The residues of TMD6 are plotted on a helical wheel diagram to display their relative orientations in an α-helix as viewed from the cytoplasmic side of the plasma membrane. Polar residues that were mutated as part of this study are boxed. Pro²⁵⁸ is also boxed, because a ste2-P258L mutation was identified previously as causing constitutive activity (33). The black boxes with white letters identify the residues that caused constitutive activation of the receptor when they were mutated to leucine. Interestingly, these residues were all located on one side of the helical wheel as shown. The white boxes with black letters on the opposite side of the wheel identify the residues that did not cause constitutive activity when they were mutated to leucine. Although proline residues can perturb the structure of an α-helix, Pro²⁵⁸ is not expected to cause a significant change in the relative orientations of the residues in this model, because proline residues are well tolerated in the transmembrane helices of membrane proteins (19, 62).

This analysis provides the first experimental evidence to suggest the orientation of a transmembrane helix in the α-factor receptor; thus, it was of interest to compare this arrangement with the predicted structure of the other members of the GPCR family. Based on the comparison of the sequences of a large set of GPCRs, Baldwin predicted that the TMDs are arranged in a clockwise manner as viewed from the intracellular side, as has been determined for rhodopsin (2, 61). Since >90% of GPCRs contain a proline residue in the middle region of TMD6, this residue forms an important landmark in the structure of TMD6. In the case of the rhodopsin-adrenergic subfamily of GPCRs, Baldwin predicted that the conserved proline faces TMD7 (2). In contrast, when the TMDs of the α-factor receptor are arranged according to Baldwin’s model, the orientation of TMD6 predicted above places Pro²⁵⁸ facing generally toward TMD5 in the α-factor receptor. This suggests that although a proline in TMD6 may be important for a wide range of GPCRs, the relative position of the proline residue may not be conserved. Perhaps this difference contributes to the observation that mutating Pro²⁵⁸ causes strong constitutive activity in the α-factor receptor but not in other GPCRs (33, 59).

According to the orientation of TMD6 described above, Gln²⁵³ is predicted to face in the direction of TMD7. It seems likely that significant intramolecular contacts occur between TMD6 and TMD7 because they are predicted to be adjacent in the Baldwin model (2). TMD6 and TMD7 are also expected to be in close proximity because they are connected by a very short extracellular loop (nine amino acids). Therefore, we investigated the possibility that Gln²⁵³ contacts any of the three polar residues in the cytoplasmic half of TMD7 (Ser²⁸⁸, Ser²⁹², or Ser²⁹³) by mutating each of the corresponding codons to encode alanine. Interestingly, the ste2-S288A and -S292A mutants displayed constitutive activity, indicating that these residues play an important role in receptor structure, possibly by forming an important contact with Gln²⁵³. Double mutant analysis was also consistent with Gln²⁵³ interacting with both Ser²⁸⁸ and Ser²⁹², since the corresponding double mutants did not display additive effects on constitutive activity. Contact between these residues could be stabilized by hydrogen bond interaction, with the serines acting as hydrogen bond donors and glutamine acting as an acceptor. Furthermore, Gln²⁵³ may be able to interact with both Ser²⁸⁸ and Ser²⁹², since the oxygen in the glutamine side chain can act as an acceptor to form two hydrogen bonds (12). The serine residues may also be close enough to simultaneously interact with Gln²⁵³, because Ser²⁸⁸ and Ser²⁹² are predicted to be in close proximity on the same side of TMD7, i.e., a little more than one turn apart in an α-helix. The fact that Gln²⁵³, Ser²⁸⁸, and Ser²⁹² are all conserved in the homologous pheromone receptors from S. kluyveri (40) and S. pombe (29) also supports the idea that they may form important intramolecular contacts. We are currently developing alternative approaches to confirm the identities of the key contact residues in the receptor to complement the genetic methods used in this study.

Mechanism of receptor activation.

Polar residues have been implicated in the activation of GPCRs because of their ability to form intramolecular contacts. In the case of rhodopsin, a salt bridge between TMD3 and TMD7 is thought to keep rhodopsin in the inactive conformation until photoactivation of the retinal chromophore leads to disruption of the salt bridge and subsequent receptor activation. This mechanism, which is specific for the photoreceptors, was confirmed in part by showing that mutations affecting the residues that form the salt bridge caused constitutive signaling in the dark (48). Another polar region, which is known as the DRY motif, is thought to play a key role in activation of a broader range of GPCRs, including the members of the rhodopsin-adrenergic receptor family (42, 50). The DRY motif, which is found at the cytoplasmic end of TMD3, is proposed to form a polar pocket in conjunction with polar residues from other TMDs until protonation of the Asp (or Glu) residue in the DRY motif causes it to move out of the polar pocket and to permit receptor activation. Consistent with this, certain mutations in this sequence cause constitutive receptor activity and increased affinity for ligand (51). Although this motif is found in many GPCRs, its absence from the pheromone receptors and other receptors indicates that additional mechanisms must exist to regulate receptor activation. In this study, we identified a novel polar region in TMD6 of the α-factor receptor that influences ligand binding and G protein activation. Since this domain occurs in the cytoplasmic half of TMD6 adjacent to the third intracellular loop and since the third intracellular loop is involved in receptor activation, it is expected that conformational changes in this part of TMD6 can play a significant role in receptor activation. This aspect of receptor regulation may be conserved in some of the other GPCRs, because constitutively active mutations in the luteinizing hormone (55) and the thyrotropin receptors (34) that are caused by mutation of polar residues in TMD6 have been discovered. The constitutive activity caused by these mutations has been implicated in their ability to cause precocious puberty and hyperfunctioning thyroid adenomas in humans (34, 55). Thus, analogous polar pockets in TMD6 of other GPCRs may function in regulation of signaling and consequently may also be important sites for mutations that cause human diseases.