Previous studies have suggested that G protein coupling, phospholipase C activation, phosphoinositide hydrolysis, and protein kinase C activation may be required for α1B-adrenergic receptor regulation, particularly for their endocytosis into intracellular vesicles. Accordingly, the internalization and down-regulation properties of mutated receptors with defects in G protein coupling and second messenger generation were investigated. The Δ12 and Δ5 receptors, previously shown to be defective in G protein coupling, exhibited greater agonist-induced losses of cell surface accessibility assessed by radioligand binding to intact cells on ice than for the wild-type receptor; however, these receptors were completely defective in endocytosis into intracellular vesicles assessed by sucrose density gradient centrifugation. These receptors also did not undergo down-regulation with long-term agonist exposure as did the wild-type receptor; instead, a prominent up-regulation was observed. The Y348A receptor, previously shown to be defective in phosphoinositide hydrolysis and endocytosis was also defective in down-regulation but did not exhibit significant up-regulation. In contrast, a receptor construct with amino acid residues 246 to 261 deleted (Δ[246–261]) was also defective in stimulation of phosphoinositide hydrolysis but exhibited internalization and down-regulation properties essentially identical to those for the wild-type receptor. Together, these results suggest that stimulation of phosphoinositide hydrolysis by α1B-adrenergic receptors is not required for their endocytosis or down-regulation but that similar and overlapping receptor structural domains are involved in mediating these processes.
The effects of G protein-coupled receptors (GPCRs) are mediated by activation of specific G proteins and their downstream effector enzymes and second messenger pathways. Epinephrine binding to α1B-adrenergic receptors (α1BARs) activates members of the Gq/11 family of G proteins, which stimulate phospholipase C (PLC) and phosphoinositide (PI) hydrolysis, eliciting cellular responses by activation of protein kinase C (PKC) and regulation of intracellular Ca2+ (Graham et al., 1996; Zhong and Minneman, 1999). Agonist binding to GPCRs also initiates desensitization, a series of adaptive changes that decrease responsiveness of the receptor to subsequent or continued agonist exposure (Perkins et al., 1991). Receptor-specific changes occurring during desensitization include a rapid uncoupling of the receptor from G protein activation; a rapid internalization of receptors into compartments no longer accessible to ligands at the cell surface, due to receptor “sequestration” within the plasma membrane and/or “endocytosis” into intracellular vesicles; and a slower decrease in the total number of receptor binding sites, referred to as down-regulation and generally thought to result from receptor degradation (Ferguson et al., 1996; Bohm et al., 1997; Krupnick and Benovic, 1998). Although the cellular and molecular mechanisms involved in desensitization are best characterized for the Gs-coupled β2-adrenergic receptor (β2AR), similar changes have been shown to occur for Gq-coupled α1BARs (Cotecchia and Mhaouty-Kodja, 1999;Garcı́a-Sáinz et al., 2000; Toews, 2000).
The extent to which G protein coupling, effector enzyme activation, and second messenger generation and action contribute to GPCR internalization remains a question of interest. In the case of the β2AR, it has been known for many years that agonist-induced internalization can occur independently of G protein coupling, cAMP formation, and activation of the cAMP-dependent protein kinase, because β2AR internalization occurs normally in S49 lymphoma cell variants that lack Gs or functional coupling between β2ARs, Gs, and adenylyl cyclase (Clark et al., 1985; Mahan et al., 1985). Instead, considerable evidence indicates that internalization of β2ARs is mediated by second messenger-independent phosphorylation of agonist-occupied receptors by members of the GPCR kinase (GRK) family and their subsequent translocation to clathrin-coated pits mediated by binding of the adaptor protein β-arrestin to the phosphorylated receptor (Ferguson et al., 1996; Krupnick and Benovic, 1998). Recent studies have shown that GRKs are also able to phosphorylate and desensitize α1BARs and that β-arrestin can modulate α1BAR desensitzation and internalization (Diviani et al., 1996; Mhaouty-Kodja et al., 1999). However, we and others have shown that direct activation of PKC with the second messenger analog phorbol-12-myristate-13-acetate or related compounds can induce or promote α1BAR internalization, at least under some conditions (Cowlen and Toews, 1988; Fonseca et al., 1995; Zhu et al., 1996). Furthermore, both PLC inhibitors and PKC inhibitors have been reported to inhibit agonist-induced internalization of α1BARs (Fonseca et al., 1995; Zhu et al., 1996; Awaji et al., 1998), suggesting that G protein coupling, effector enzyme activation, and second messenger-regulated kinases may play a more important role in internalization for α1BARs than for β2ARs.
We have presented evidence that sequestration of receptors within the plasma membrane (defined as a loss of cell surface accessibility of radioligand binding sites) and endocytosis within intracellular vesicles (defined as a shift of receptors from the plasma membrane fraction to a light vesicle fraction with sucrose density gradient centrifugation) are two distinct steps or components of receptor internalization for α1BARs (Cowlen and Toews, 1988; Wang et al., 1997; Toews, 2000). Two specific aspects of those studies prompted us to further investigate the possible requirement of G protein coupling and downstream signaling events specifically in the endocytosis step of α1BAR internalization. First, we showed that agonist plus the PKC activator phorbol-12-myristate-13-acetate could induce endocytosis of α1BARs in DDT1 MF-2 cells, a cell line in which treatment with agonist alone induced sequestration without endocytosis (Cowlen and Toews, 1988). Second, in a subsequent study aimed at investigating the role of the NPIIY sequence in the seventh transmembrane domain of the α1BAR as an internalization signal, we found that the Y348A-mutated α1BAR was essentially completely defective in stimulation of PI hydrolysis and in agonist-induced endocytosis, although agonist-induced sequestration of the mutated receptors within the plasma membrane was similar to that for the wild-type receptor (Wang et al., 1997).
To investigate whether the defect in G protein coupling and PI hydrolysis of the Y348A receptor was responsible for its failure to endocytose and to further investigate the requirement of G protein coupling for other aspects of receptor function and regulation, we have now investigated the binding, functional, and regulatory properties of three additional mutated α1BARs that are defective in stimulation of PI hydrolysis. The results suggest that G protein coupling is not required for α1BAR endocytosis or down-regulation but that the receptor determinants involved in G protein coupling are very similar to, yet distinct from, those required for endocytosis and down-regulation.
Cell culture medium, serum, trypsin, G418, and LipofectAMINE reagent were from Invitrogen (Carlsbad, CA). The Muta-Gene In vitro Mutagenesis kit was obtained from Bio-Rad (Hercules, CA); other enzymes were from New England BioLabs (Beverly, MA). [3H]Prazosin was from PerkinElmer Life Sciences (Boston, MA) and [3H]inositol was from Amersham Pharmacia Biotech (Piscataway, NJ). Epinephrine, phentolamine, sucrose, and other biochemicals were from Sigma Chemical (St. Louis, MO).
Mutagenesis, Transfections, and Cell Culture.
The preparation of the stably transfected CHO cells expressing the wild-type and Y348A α1BARs used in these studies has been described in detail previously (Wang et al., 1997). To generate the Δ[246–261] receptor construct, the cDNA encoding the α1BAR inserted between theHindIII and XbaI sites of phage M13 mp18 was used as the substrate for oligonucleotide-directed mutagenesis with the Bio-Rad Muta-Gene M13 kit as in our previous studies (Wang et al., 1997, 2000). The mutagenic primer for the Δ[246–261] receptor construct was as follows, where the carat represents the site where the deletion was introduced: GTCATGAAGGAGATG ∧ GAGGACACCCTCAGC, deleting residues 246 through 261 (Fig. 1). After confirmation of the mutation by DNA sequencing, the mutated α1BAR cDNA was cut from M13 mp18 by usingHindIII/XbaI and subcloned into the expression vector pRC/CMV, followed by DNA sequencing to reconfirm the mutation. The mutated α1BAR plasmid was then stably transfected into CHO-K1 Chinese hamster ovary cells by using LipofectAMINE, and clones resistant to 400 μg/ml G418 were isolated and screened for α1BAR expression, as in our previous studies (Wang et al., 1997, 2000).
The pCMV plasmids encoding the Δ12 and Δ5 receptors, whose preparation has been previously described (Wu et al., 1995), were kindly provided by Dr. Dianqing Wu (University of Connecticut, Farmington, CT). These plasmids were transfected into CHO-K1 cells together with the pMAMneo plasmid encoding neomycin resistance by using LipofectAMINE, and clones resistant to G418 (400 μg/ml) were isolated and screened for α1BAR expression. Sequencing of Δ12 confirmed the correct deletion of the YIV sequence at residues 227 to 229. However, sequencing of the Δ5 plasmid indicated the presence of additional changes beyond those originally described (Wu et al., 1995), as described under Results and in Fig. 1.
Cells were maintained in monolayer culture in Ham's F12 medium supplemented with 10% fetal bovine serum and 200 μg/ml G418 at 37°C in a humidified incubator with a 5% CO2atmosphere. Cells from confluent flasks were trypsinized and plated in culture dishes at 3000 to 5000 cells/cm2. Cells were typically used for experiments on the 4th day of culture.
Binding and Functional Assays.
Binding and functional assays were conducted as described in greater detail for previous studies (Wang et al., 1997, 2000), with minor modifications. For membrane binding assays, cells grown on 100- or 150-mm dishes were lysed by scraping in hypotonic buffer and membranes were isolated by centrifugation. The membrane pellets were resuspended in binding buffer (20 mM Tris, pH 7.4, 2 mM MgCl2, 140 mM NaCl) and aliquots were incubated with [3H]prazosin in binding buffer for 60 min at 37°C; for these studies 100 μM phentolamine was used to define nonspecific binding. The reactions were stopped by filtration over Whatman GF/B glass fiber filters (Whatman, Maidstone, UK) followed by washing and quantitation by liquid scintillation counting. For saturation assays, six to seven different concentrations of [3H]prazosin were used. For competition binding assays, [3H]prazosin was used at approximately 300 pM and the concentration of epinephrine as competing ligand was varied; this high concentration of [3H]prazosin was used to avoid binding more than 10% of the radioligand for those clones with high levels of receptor expression.
For PI hydrolysis assays, cells grown on 35-mm dishes were prelabeled overnight with 2 μCi/ml [3H]inositol and then stimulated for 20 min with various concentrations of epinephrine in medium containing 10 mM LiCl. Labeled compounds were then extracted from the cells and the aqueous and organic phases were separated. Inositol phosphates in the aqueous phase were isolated by chromatography on Dowex 1-X8 (formate form) columns and quantitated by liquid scintillation counting. Data are presented as the percentage of conversion of [3H]inositol phospholipids to [3H]inositol phosphates.
Internalization and Down-Regulation Assays.
Assays of receptor internalization and down-regulation were all conducted essentially identically to our previous studies (Wang et al., 1997,2000). Assays of radioligand binding to intact cells on ice, referred to as “ice binding assays”, were used to assess decreases in cell surface accessibility of receptors, referred to as “loss of ice binding”. Cells grown on 35-mm dishes were incubated for 30 min in the absence or presence of 10 μM epinephrine to induce receptor internalization. Cells were quickly washed and then incubated on ice for 4 h with 1.8 nM [3H]prazosin in serum-free medium; nonspecific binding was defined as that occurring in the presence of 10 μM phentolamine. Cells were then rapidly washed in medium containing 10 μM phentolamine and dissolved in 1 ml of 0.2 N NaOH. Radioactivity associated with the dissolved cells was assessed by liquid scintillation counting.
For sucrose density gradient centrifugation assays of receptor endocytosis, cells grown on 100-mm dishes were incubated for 30 min in the absence or presence of 10 μM epinephrine to induce internalization. Cells were washed and then lysed by scraping in ice-cold hypotonic buffer. The lysate was layered on top of a discontinuous sucrose density gradient consisting of 1.7 ml 15% sucrose (w/v), 5.0 ml of 30% sucrose, and 2.5 ml of 60% sucrose. Samples were centrifuged at 28,000 rpm for 60 min at 4°C in an SW41 rotor in a Beckman L8–70 refrigerated ultracentrifuge. Fractions were then collected from the top of the tubes, and binding of [3H]prazosin (1.4 nM) to the membranes in each fraction was determined essentially as described above.
For down-regulation assays, cells grown on 100-mm dishes were incubated in the absence or presence of 10 μM epinephrine for 24 h and then washed and lysed. Membranes were isolated by centrifugation, and binding of [3H]prazosin (3.7 nM) to the isolated membranes was then determined as described above.
Nonlinear regression analyses of saturation and competition binding assay and dose-response curve data were performed with GraphPad Prism (GraphPad Software, San Diego, CA). Values for all parameters for all mutations are the averages from multiple experiments, with duplicate or triplicate determinations in each experiment, including assays with at least two different clones and performed on at least two different days. Data are presented as the means ± S.E.M. (n = x, y), where x indicates the total number of determinations and y represents the number of different clones tested.
Mutated Receptor Constructs.
The G protein-coupling properties of two mutated α1BAR constructs with deletions at the amino-terminal end of the third intracellular loop (Fig. 1) were reported previously (Wu et al., 1995). The mutation referred to as Δ12 (deleting three amino acids, residues 227–229) was shown to be defective in coupling to all members of the Gq/11family of G proteins (Gq, G11, G14, and G16), whereas the mutation referred to as Δ5 (deleting 35 amino acids, residues 227–261, including the residues deleted in Δ12) was defective in coupling to Gq, G11, and G14 but retained a low level of coupling to G16. We confirmed the deletion of three amino acids (residues 227–229) in the Δ12 plasmid cDNA by sequencing. Our sequencing of the Δ5 plasmid indicated that 35 amino acids (residues 227–261) were deleted, as originally described; however, our sequencing indicated the insertion of an Asn and a Ser residue at the carboxyl-terminal end of this deletion, as indicated in Fig. 1. Whether these two residues were present in the receptors in the previous study is unknown. We generated an additional related α1BAR mutant, referred to as Δ[246–261], in which only the carboxyl-terminal 16 amino acids (residues 246–261) of the originally reported Δ5 deletion were deleted. Each of these constructs was stably transfected in CHO cells, and the binding, functional and regulatory properties of multiple clones of each construct were investigated and compared with those for the wild-type α1BAR.
Functional Properties of Mutated Receptors.
Epinephrine stimulated PI hydrolysis in cells expressing the wild-type receptor by approximately 4-fold, with half-maximal stimulation at approximately 200 nM (Fig. 2). The EC50 value for the wild-type receptor from this set of experiments is higher than the values of 31 and 43 nM reported in our two previous studies (Wang et al., 1997, 2000). The reason for this lower potency compared with our earlier studies is not completely clear; however, it is likely to result from differences in “coupling efficiency” or “receptor reserve” between the cells used in the earlier studies and those in the current studies, because the affinity of the receptor for epinephrine in competition binding assays has remained essentially constant in all of these studies (Table1). The Δ12 and Δ5 receptors expressed in CHO cells appeared to be completely defective in epinephrine stimulation of PI hydrolysis, similar to the previously reported results with these receptors in transiently transfected COS-7 cells (Wu et al., 1995). The newly generated Δ[246–261] receptor was similarly defective in stimulation of PI hydrolysis. The averages ± S.E.M. of the fold stimulation values with 10 μM epinephrine from the individual PI hydrolysis experiments were 5.08 ± 2.01 for wild-type, 0.85 ± 0.12 for Δ12, 1.14 ± 0.19 for Δ5, and 1.20 ± 0.13 for Δ[246–261]; these values were significantly different from 1.0 (p < 0.05, analysis of variance followed by Dunnett's multiple comparison test) only for the wild-type receptor.
Binding Properties of Mutated Receptors.
Clones with high-level expression were obtained for all of the mutated receptors. Receptor expression levels and antagonist binding properties were tested in saturation binding assays with the antagonist radioligand [3H]prazosin (Table 1).B max values ranged from 0.1 to nearly 10 pmol/mg of membrane protein in the clones analyzed, similar to the range of expression levels obtained for the wild-type receptor. The affinities of all three of the mutated receptors for [3H]prazosin were 2- to 3-fold lower than that of the wild-type receptor. These results are similar to the decrease in antagonist radioligand affinity observed for the F303G and F303N α1BARs that are uncoupled from G protein activation due to mutations in transmembrane domain VI (Chen et al., 2000); however, they are in contrast to our previously characterized Y348A-mutated α1BAR, which had [3H]prazosin binding affinity identical to the wild-type receptor (Wang et al., 1997).
Agonist binding properties were assessed in assays of epinephrine competition for [3H]prazosin binding (Table 1). The Δ12 and Δ5 receptors both exhibited markedly higher affinities for epinephrine than did the wild-type receptor, similar to the results with the Y348A-mutated α1BAR (Wang et al., 1997) and with other α1BARs with defects in G protein coupling and PI hydrolysis due to mutations in transmembrane domain VI (Chen et al., 2000) or in the DRY motif in the second intracellular loop (Scheer et al., 1996). In the previous study with transient transfection in COS-7 cells, the Δ5 receptor exhibited higher affinity than the wild-type receptor for the agonist norepinephrine, but the affinity of the Δ12 receptor for norepinephrine was the same as for the wild-type receptor (Wu et al., 1995). In contrast to the results with the other mutated receptors, the Δ[246–261] receptor exhibited an affinity for epinephrine that was slightly lower than that of the wild-type receptor, suggesting that the nature of the coupling defect for this receptor may be different from that for the other mutated receptors. Together, these binding and functional assays indicate that the failure of these mutated receptors to stimulate PI hydrolysis is not due to a failure of the receptors to be expressed or an inability to bind agonist.
Assays of Agonist-Induced Changes in Cell Surface Accessibility of Receptor Binding Sites.
Assays of [3H]prazosin binding to intact cells on ice were used to assess agonist-induced changes in the cell surface accessibility of the various receptor constructs (Fig.3, top). Pretreatment of cells expressing the wild-type receptor for 30 min with 10 μM epinephrine led to a 26 ± 2% decrease in [3H]prazosin binding. Agonist pretreatment of cells expressing the Δ12 and Δ5 receptors led to 80 ± 3% and 53 ± 6% decreases in [3H]prazosin binding to intact cells on ice, respectively, decreases that were greater than that seen for the wild-type receptor assayed under the same conditions. In contrast to the results with the Δ12 and Δ5 receptors, the agonist-induced decrease in [3H]prazosin binding to intact cells on ice for cells expressing the Δ[246–261] receptor was 24 ± 5%, essentially identical to that for the wild-type receptor.
The loss of ice binding was further investigated in saturation binding assays with intact cells on ice for the Δ12, Δ[246–261], and wild-type receptors (Table 2). Agonist pretreatment induced both a decrease inB max and a decrease in affinity for the Δ12 receptor, similar to our previous results with the Y348A receptor (Wang et al., 1997). However, in contrast to the previous results with the Y348A receptor, the decrease inB max observed in the saturation assays with the Δ12 receptor was about twice as great as for the wild-type receptor, similar to the data in Fig. 3. The values for the loss of ice binding observed in the assays with a single concentration of [3H]prazosin that are shown in Fig. 3 are more similar to the decreases in B maxobserved in the saturation assays in Table 2 for the current experiments than in the previous study, because a much higher and nearly saturating concentration of [3H]prazosin was used in the single concentration assays in the current studies. In contrast to the results with the Δ12 receptor, the loss of ice binding for the Δ[246–261] receptor was due to a decrease inB max with no change in affinity, and the decrease in B max for the Δ[246–261] receptor was similar to that for the wild-type receptor (Table 2).
One possibility for the apparent decreases in affinity seen for the Δ12 receptor in this study and for the Y348A receptor in our previous study (Wang et al., 1997) is incomplete removal of the pretreatment agonist during the washes before the binding assay. Low concentrations of retained agonist from the pretreatment would be more likely to interfere with the subsequent binding assay for these receptors because they have significantly higher affinity for epinephrine compared with the wild-type receptor and the Δ[246–261] receptor, and this is in fact the pattern of effects that is observed. In experiments with the Δ12 receptor pretreated with 0.1 μM rather than 10 μM epinephrine to lessen the possible contribution of retained agonist, the decrease in affinity after epinephrine pretreatment was smaller but still present; thus, retained agonist likely contributes to the observed decrease in affinity, but additional changes in receptor binding properties induced by epinephrine pretreatment may also be involved.
Sucrose Density Gradient Centrifugation Assays of Agonist-Induced Endocytosis.
The agonist-induced endocytosis of the various receptor constructs was assessed by sucrose density gradient centrifugation to separate plasma membrane receptors from those in intracellular endocytotic vesicles (Fig. 3, middle). Pretreatment of cells expressing the wild-type receptor with the agonist epinephrine (10 μM) for 30 min induced a shift of 24% of the receptors initially in the plasma membrane fraction to the light vesicle fraction, indicating that for these receptors, endocytosis into intracellular vesicles is sufficient to account for essentially all of the loss of cell surface-accessible receptors detected in the ice binding assays. In contrast, pretreatment of cells expressing the Δ12 and Δ5 receptors with epinephrine did not induce a shift of receptors from the plasma membrane fraction to the light vesicle fraction. This lack of agonist-induced endocytosis is similar to the results obtained previously for the Y348A α1BAR (Wang et al., 1997). However, pretreatment of cells expressing the Δ[246–261] receptor with epinephrine induced a shift of plasma membrane receptors to the light vesicle fraction that was similar in magnitude to the shift for the wild-type receptor. Together with the results from the cell surface accessibility assays mentioned above, these results indicate that agonist exposure leads to sequestration of the Δ12 and Δ5 receptors within the plasma membrane but not to their endocytosis into intracellular vesicles, similar to the results reported previously for the Y348A α1BAR. In contrast, the Δ[246–261] receptor, in spite of its defect in coupling to PI hydrolysis, undergoes both sequestration and endocytosis similar to the wild-type receptor.
Agonist-Induced Changes in Receptor Expression.
Cells expressing each of the receptor constructs were pretreated with10 μM epinephrine for 24 h to assess their ability to undergo agonist-induced down-regulation (Fig. 3, bottom). Under these conditions, the wild-type receptor was down-regulated by 39 ± 2%. For cells expressing the Δ12 and Δ5 receptors, 24-h pretreatment with epinephrine did not induce down-regulation but led instead to a marked up-regulation of the total number of [3H]prazosin binding sites. For the Δ12 receptor, the value for epinephrine-pretreated cells was 245 ± 29% of the value for control cells (145% up-regulation) and for the Δ5 receptor the value for epinephrine-pretreated cells was 255 ± 16% of the value for control cells (155% up-regulation). In contrast, agonist pretreatment led to 46 ± 2% down-regulation of the Δ[246–261] receptor, similar to results with the wild-type receptor. The down-regulation properties of the Y348A α1BAR were also investigated, because these assays were not included in our original studies of this receptor (Wang et al., 1997). The Y348A receptor did not undergo either down-regulation or up-regulation with agonist pretreatment; the value for binding to membranes from Y348A cells pretreated with10 μM epinephrine for 24 h was 107 ± 6% of the value for control cells (n = 8, 8).
The goal of these studies was to investigate the hypothesis that G protein coupling and subsequent PKC activation are not required for α1BAR sequestration within the plasma membrane but are required for endocytosis into intracellular vesicles and for the subsequent down-regulation of these receptors. Three of the four mutated α1BARs that were defective in stimulation of PI hydrolysis were also defective in agonist-induced endocytosis and down-regulation (Δ12, Δ5, and Y348A), suggesting a relationship between PI hydrolysis and endocytosis and down-regulation. However, the ability of the fourth mutated receptor (Δ[246–261]) to undergo normal endocytosis and down-regulation, despite an apparently complete lack of PI hydrolysis, strongly suggests that PI hydrolysis is not required for either process. Recent studies with other Gq-coupled receptors have also provided evidence that Gq coupling and PI hydrolysis are not required for receptor internalization. These include studies with mutated receptors for AT1 angiotensin receptors (Hunyady et al., 1994), ETA endothelin receptors (Bhowmick et al., 1998), and pituitary adenylyl cyclase-activating protein type 1 receptors (Lyu et al., 2000) as well as studies demonstrating thyrotropin-releasing hormone receptor internalization in cells lacking Gq and G11(Yu and Hinkle, 1999). The data presented here appear to conflict with other evidence that the PKC inhibitor staurosporine is effective at inhibiting α1BAR internalization (Fonseca et al., 1995; Zhu et al., 1996; Awaji et al., 1998). One possible explanation for this discrepancy is that the effects of staurosporine on internalization are due to inhibition of a kinase other than PKC. Alternatively, it is possible that ongoing PKC activity is required but that further activation of PKC by the specific receptor being internalized is not required.
Our conclusion that PI hydrolysis is not required for internalization is based on the properties of the Δ[246–261] receptor. However, we cannot rule out the possibility that an undetectably low level of PI hydrolysis occurs with the Δ[246–261] receptor and that this very low level of second messenger generation is sufficient to promote receptor internalization. A previous study using confocal microscopy to monitor internalization of green fluorescent protein-tagged α1BARs stably expressed in αT3 pituitary gonadotroph cells showed that receptor internalization was blocked by the PLC inhibitor U73122, suggesting that PLC activation is required for α1BAR internalization (Awaji et al., 1998). We attempted to use U73122 as a second approach to investigate the role of PI hydrolysis in α1BAR internalization. However, we discovered that U73122 inhibits ligand binding to the α1BAR at similar concentrations to those that inhibit PLC activation (unpublished data), preventing its straightforward use in these studies. The contribution of receptor blockade to the effects of U73122 will be particularly prominent when high concentrations of U73122 are used together with relatively low concentrations of α1BAR agonist, such as the combination of 10 μM U73122 with 100 nM norepinephrine used in the previous study (Awaji et al., 1998). Thus, it is possible that the inhibition of α1BAR internalization observed in their studies resulted from U73122 blockade of receptors rather than from PLC inhibition. More detailed studies of the multiple and complex effects of U73122 and related compounds on PLC activation, receptor binding, and receptor internalization assessed by various assays are in progress.
Although our results suggest that G protein coupling and PI hydrolysis are not required for endocytosis, the observation that three of the four mutated receptors that were unable to stimulate PI hydrolysis were also defective in endocytosis suggests that highly similar and overlapping receptor domains are required for G protein activation and for endocytosis. The Δ[246–261] mutation differentiates between these two responses, blocking PI hydrolysis but not endocytosis. Thus, amino acid residues 246 to 261 of the α1BAR are apparently critical for proper functional interaction of the receptor with G proteins but not for its interaction with the cellular machinery involved in receptor endocytosis. In contrast, residues 227 to 229 are important for both processes. Whether these residues are directly involved in G protein coupling and endocytosis or whether their mutation alters the structure of other regions of the receptor remains to be determined. Studies of additional mutations, including smaller deletions as well as substitution mutations, will be required to more precisely identify the roles of this region of the receptor and of individual amino acid residues in both G protein coupling and endocytosis. Several recent studies with other receptors have succeeded in separating the structural domains involved in internalization from those involved in G protein activation for various Gq-coupled receptors (Hunyady et al., 1994;Bhowmick et al., 1998; Lyu et al., 2000). Of greatest interest in relation to our data is a recent study of the regulatory properties of the constitutively active D142A-mutated α1BAR, which was able to activate Gq and PI hydrolysis but was defective in both agonist-induced receptor phosphorylation and internalization (Mhaouty-Kodja et al., 1999). This mutation also separates G protein activation from receptor internalization, but with the “opposite” phenotype from our Δ[246–261] mutation, which is defective in G protein activation but able to undergo normal internalization. Further studies based on this interesting pair of mutated α1BARs could yield important information on the distinct structural domains and conformations required for interaction with G proteins versus those required for interaction with the cellular endocytosis machinery.
The results presented here provide further evidence that the process that we call sequestration does not require coupling to PI hydrolysis, because it occurred for all three mutated receptors studied here to an extent equal to or greater than that for the wild-type receptor, similar to previous results with the Y348A receptor (Wang et al., 1997). Phosphorylation of GPCRs by GRKs requires only agonist occupancy of receptors, but not second messenger generation, and GRK-mediated phosphorylation and subsequent binding of β-arrestins have been implicated in the pathway for internalization of many GPCRs (Krupnick and Benovic, 1998; Ferguson, 2001), including α1BARs (Diviani et al., 1996, 1997). GRK-mediated phosphorylation of α1BARs followed by β-arrestin-mediated targeting of the phosphorylated receptors to an inaccessible subdomain within the plasma membrane is thus a likely mechanism to account for the process that we call sequestration, although this remains to be demonstrated. Whether this “compartment of sequestration” is an intermediate compartment on the pathway to endocytosis in clathrin-coated pits, vesicles, and endosomes, or whether it represents an alternate pathway to remove receptors from accessibility to ligands at the cell surface, remains to be determined. Identification of this compartment may also help to explain the greater extent of sequestration observed with the Δ12 and Δ5 receptors than with the wild-type and other mutated receptors. The Δ12, Δ5, and Y348A mutated α1BARs, which traffic to this compartment but do not undergo endocytosis, provide powerful tools for assessing the nature of this compartment and the molecular mechanisms involved in receptor transport to and from this compartment.
The three mutated receptors that were defective in endocytosis were also defective in down-regulation. The Y348 receptor did not down-regulate in response to agonist, whereas the Δ12 and Δ5 receptors showed a marked up-regulation in response to agonist. The Δ[246–261] receptor, which endocytosed normally, also down-regulated normally. If down-regulation is mediated by receptor protein degradation in lysosomes after receptor endocytosis, still the most widely accepted mechanism (Tsao et al., 2001), then it would be expected that defects in endocytosis would correlate with defects in down-regulation, as observed with these mutated receptors. However, our recent studies of α1BARs with carboxyl-terminal tail mutations showed that down-regulation can occur even for receptors that are completely defective in rapid receptor sequestration and endocytosis, raising questions about the lysosomal degradation hypothesis for down-regulation of these receptors (Wang et al., 2000). Down-regulation without endocytosis has been reported for other receptors as well (Tsao et al., 2001). Studies are in progress to further characterize the prominent up-regulation observed with the Δ12 and Δ5 receptors and to identify the mechanisms involved; potential mechanisms include the nuclear factor-κB-mediated mechanism recently proposed for up-regulation of serotonin receptors (Cowen et al., 1997) and the ligand-mediated stabilization of “inherent instability” that has been shown to occur for both constitutively active and inactive mutations of adrenergic and other receptors (Gether et al., 1997; Alewijnse et al., 2000; Wilson and Limbird, 2000).
The three mutated receptors that were defective in endocytosis also exhibited affinities for the agonist epinephrine that were 8- to 40-fold higher than that for the wild-type receptor (Table 1; Wang et al., 1997), similar to results with several other α1BAR mutations that inhibit coupling to PI hydrolysis (Scheer et al., 1996; Chen et al., 2000). In contrast, the Δ[246–261] receptor exhibited an affinity for epinephrine slightly lower than the wild-type receptor. Thus, although it does not stimulate PI hydrolysis, the Δ[246–261] receptor is otherwise more similar to the wild-type receptor than to the other three mutated receptors in terms of agonist binding affinity as well as internalization and down-regulation properties. These results suggest that the nature of the coupling defect for the Δ[246–261] receptor may be different from that for the other mutated receptors and that further studies of the Δ[246–261] receptor could provide new insights into G protein coupling for α1BARs.
In summary, we have generated a new mutated α1BAR that is defective in stimulation of PI hydrolysis but exhibits binding and regulatory properties different from those of three other α1BARs with defects in PI hydrolysis that have been described previously. Together, our studies of these mutated receptors suggest that at least four responses to agonist occupancy of α1BARs can occur without detectable stimulation of PI hydrolysis, namely, sequestration, endocytosis, down-regulation, and a novel up-regulation. Occupancy of the ligand binding site of the receptor by agonist and the resulting conformational changes alone may be sufficient to mediate these responses; alternatively, it is possible that one or more of these changes are mediated by as-yet-unidentified G protein-independent signaling pathways, several of which have been recently identified for other GPCRs (Heuss and Gerber, 2000). These mutated receptors provide a powerful set of tools for more detailed investigation of the specific receptor domains and conformations mediating both receptor signaling and the multiple adaptive responses that occur after agonist binding to these receptors.
We thank Dr. Dianqing Wu (University of Connecticut) for providing the Δ12 and Δ5 mutated receptor constructs, and Drs. David Bylund (University of Nebraska Medical Center) and Susanna Cottechia (Institut de Pharmacologie et de Toxicologie, Facultéde Médecine, Lausanne, Switzerland) for helpful discussions.
- G protein-coupled receptor
- α1B-adrenergic receptor
- phospholipase C
- protein kinase C
- β2-adrenergic receptor
- G protein-coupled receptor kinase
- Chinese hamster ovary
- Received July 17, 2001.
- Accepted September 21, 2001.
- The American Society for Pharmacology and Experimental Therapeutics