点击显示 收起
Baker Heart Research Institute, Melbourne, Australia (D.T.D., H.Q., W.G.T.)
Western Australian Institute for Medical Research, Centre for Medical Research, Sir Charles Gairdner Hospital, University of Western Australia, Perth, Australia (R.S., E.L., K.P., K.A.E.)
Abstract
-arrestins bind to phosphorylated, seven-transmembrane-spanning, G protein-coupled receptors (GPCRs), including the type 1 angiotensin II receptor (AT1R), to promote receptor desensitization and internalization. The AT1 R is a class B GPCR that recruits both -arrestin1 and -arrestin2, forming stable complexes that cotraffic to deep-core endocytic vesicles. -Arrestins contain one amphipathic and potentially amphitropic (membrane-targeting) -helix (helix I) that may promote translocation to the membrane or influence receptor internalization or trafficking. Here, we investigated the trafficking and function of -arrestin1 and -arrestin2 mutants bearing substitutions in both the hydrophobic and positively charged faces of helix I. The level of expression of these mutants and their cytoplasmic localization (in the absence of receptor activation) was similar to wild-type -arrestins. After angiotensin II stimulation, both wild-type and -arrestin mutants translocated to the cell membrane, although recruitment was weaker for mutants of the hydrophobic face of helix I. For all -arrestin mutants, the formation of deep-core vesicles was less observed compared with wild-type -arrestins. Furthermore, helix I conjugated to green fluorescent protein is not membrane-localized, suggesting that helix I, in isolation, is not amphitropic. Bioluminescence resonance energy transfer analysis revealed that both wild-type and -arrestin mutants retained a capacity to interact with the AT1R, although the interaction with the mutants was less stable. Finally, wild-type and mutant -arrestins fully supported receptor internalization in human embryonic kidney cells and mouse embryonic fibroblasts deficient in -arrestin1 and -2. Thus, helix I is implicated in postmembrane trafficking but is not strongly amphitropic.
The nonvisual arrestins, -arrestin1 (arr1) and -arrestin2 (arr2), are ubiquitously expressed and regulate the activity of hundreds of GPCRs, including the type 1 angiotensin II (AngII) receptor (AT1R). -Arrestins bind to activated and phosphorylated GPCRs, promoting receptor internalization and preventing further interaction of receptors with G proteins, thereby attenuating initial signaling. They can also function as scaffolds to recruit additional signaling/regulatory molecules to the receptor (Luttrell et al., 1999; DeFea et al., 2000; Hall and Lefkowitz, 2002). The recruitment and trafficking of -arrestins can be visualized using confocal microscopy of green fluorescent protein (GFP)-labeled arr1 and -2. Derived from preferential trafficking of arr1 and -2, Oakley et al. (2000) proposed the classification of GPCRs as either class A or class B—the AT1R is a class B GPCR that recruits both arr1 and -2 with equal affinity and forms stable complexes that internalize via clathrin-coated pits into deep-core endocytic vesicles. In contrast, class A GPCRs, such as the 2-adrenergic receptor (2AR), interact transiently and preferentially with arr2 and dissociate from it soon after receptor internalization.
arr1 (418 amino acids) shares 78% sequence homology with arr2 (410 amino acids). The crystal structure of arrestin in its resting state (Han et al., 2001) reveals N and C domains flanking a polar core. Elements within both domains have been shown to bind phosphorylated and activated GPCRs (Gurevich and Gurevich, 2004; Vishnivetskiy et al., 2004). Although the structural changes that underpin arrestin activation are poorly understood, it is predicted that this process probably involves disruption of a key salt bridge in the polar core of arrestin by phosphorylated amino acids in the receptor and subsequent destabilization of the arrestin structure to then accommodate the cytoplasmic face of the receptor (Han et al., 2001; Gurevich and Gurevich, 2004). A novel feature of this activation is the proposed release of an -helix in the N domain (helix I), which is highly conserved in arr1 (T98RLQERLIKKL108) and arr2 (T99RLQDRLLKKL109). In the basal state of -arrestin, the hydrophobic face of helix I is completely buried in the hydrophobic cage formed by the -strands of the N domain and the carboxyl terminus. Activation of -arrestin is believed to destabilize the constraining hydrophobic pocket, allowing helix I to swing from its resting position and insert within the membrane bilayer and strengthen the receptor-arrestin complex (Han et al., 2001). Helix I is amphipathic and potentially amphitropic (membrane-targeting), with hydrophobic residues (Leu100, Leu104, and Leu108, arr1 numbering) aligned on one side of the helix and positive charges (Arg99, Arg103, Lys106, and Lys107) aligned on the opposite side (Han et al., 2001). This suggests a possible role in membrane targeting and facilitation of receptor internalization and trafficking. However, there is currently no evidence to support this directly.
Various studies have shown that many intracellular signaling and regulatory proteins (e.g., the Ras family of GT-Pases and Src tyrosine kinases) use amphitropism to reversibly attach to the cell membrane during activation (Johnson and Cornell, 1999). Recent studies have revealed that components of the GPCR activation/deactivation cycle also seem to use amphitropism to modulate function. For example, helix VIII in the proximal carboxyl terminus of rhodopsin, angiotensin, and oxytocin receptors is a membrane-based sensor (Krishna et al., 2002; Mozsolits et al., 2002; Zhong et al., 2004). The basically charged, amphipathic -helices in the regulators of G protein signaling (RGS) proteins and GPCR kinases (GRKs) also promote reversible membrane association (Chen et al., 1999; Bernstein et al., 2000; Thiyagarajan et al., 2004). The hydrophobic face of the -helix is believed to intercalate into the lipid bilayer, whereas the positively charged amino acids on the opposing face interact with the head groups of anionic phospholipids.
In this study, we investigated the role of helix I in -arrestin trafficking, receptor interaction, and function. We observed that mutations to helix I that negate the hydrophobic and charged aspects do not abrogate membrane recruitment of arrestins in response to receptor activation. It is interesting that these mutants retain a capacity to interact with activated receptors and to fully support receptor internalization, yet they do not traffic efficiently to deep-core endocytic vesicles.
Materials and Methods
Materials. Anti-GFP polyclonal antibody was purchased from BD Biosciences Clontech (Palo Alto, CA), and AngII was from Auspep (Melbourne, Australia). The SuperSignal West Pico Chemiluminescent was purchased from Pierce Chemical (Rockford, IL). Radiolabeled [125I]AngII (specific activity >2000 Ci/mmol) was provided by ProSearch (Melbourne, Australia). All other chemicals were from Sigma-Aldrich (St. Louis, MO) or BDH Laboratory Supplies (Poole, Dorset, UK).
Plasmids. The K-ras-GFP construct (referred to as tK-GFP) was provided by J. Hancock (Department of Pathology, University of Queensland, Brisbane, Australia). 2AR plasmid was provided by R. Summers (Department of Pharmacology, Monash University, Melbourne, Australia). The construction of hemagglutinin epitope-tagged wild-type AT1AR, an AT1R-EGFP, and the dominant-negative arr1 (arr1318eC419) have been reported previously (Thomas et al., 1998; Qian et al., 2001; Holloway et al., 2002). A thyrotropin-releasing hormone receptor (TRHR1)-Renilla reniformis luciferase (Rluc) construct was described previously (Hanyaloglu et al., 2002) and the AT1R-Rluc construct was generated by amplifying the hemagglutinin epitope AT1R using PCR (sense primer, T7 primer; antisense primer, 5'-GAAGCGGCCGCTCCACCTCAAAACAAGACGCAGG) with an extra NotI site (boldface italic) and subsequently subcloning into pcDNA3/Rluc.
GFP and HcRed versions of the -arrestin mutants were constructed using PCR-based site-directed mutagenesis (ExSite; Stratagene, La Jolla, CA) from arr1-GFP, arr2-GFP, and arr2-HcRed (kindly provided by M. G. Caron, Duke University Medical Center, Durham, NC). LLL/A mutant versions of arr1-GFP and arr2-GFP contained three substitutions of leucine residues with alanine (arr1: Leu99, Leu103, and Leu107; arr2: Leu100, Leu104, and Leu108), whereas RRKK/Q mutant versions of arr1-GFP and arr2-GFP are characterized by quadruple substitutions of arginine and lysine residues with glutamine (arr1: Arg98, Arg102, Lys105, and Lys106; arr2: Arg99, Arg103, Lys106, and Lys107). 5'-Oligonucleotides used for mutagenesis were (5'-3'): arr1LLL/A, CGGGCCATCAAGAAGGCGGGCGAGCATGCCTACCCC (sense) and CTCTTGTGCCCGAGTCAGTGGCTTCTTGTC (antisense); arr2LLL/A, GCCCTGAAGAAGGCGGGCCAGCATGCCCACCCC (sense) and CCGGTCCTGTGCGCGGGTGGGGGGCCGAGG (antisense); arr1RRKK/Q, CTGATCCAGCAGCTGGGCGAGCATGCCTACCCC (sense) and CTGCTCTTGTAGCTGAGTCAGTGGCTTCTTGTC (antisense); and arr2RRKK/Q, CTGCTGCAGCAGTTGGGCCAGCATGCCCACCCC (sense) and CTGGTCCTGTAGCTGGGTGGGGGGCCGAGGTGG (antisense).
Oligonucleotides were 5'-phosphorylated using T4 polynucleotide kinase. Silent restriction sites (BsrBI, DraII, or PvuII) were introduced to facilitate the screening of mutated clones (boldface italic) formed after ligation of PCR product created by arr1LLL/A, arr2LLL/A, arr1RRKK/Q, or arr2RRKK/Q primers, respectively.
Expression plasmids for GFP-helix I were constructed by ligating HindIII/BamHI cDNA fragments of arr1 and -2, corresponding to helix I, into the cloning site of pEGFP-C1 (BD Biosciences Clontech). GFP-arr1helix ended with TRLQERLIKKL, and GFP-arr2helix terminated with TRLQDRLLKKL. All constructs were verified by sequencing.
Cell Culture and Transfection. Mouse embryonic fibroblasts (MEFs; provided by R. J. Lefkowitz, Duke University Medical Center, Durham, NC) from a arr1/2 knockout (arr1/2KO) (Kohout et al., 2001) and HEK-293 (American Type Culture Collection, Manassas, VA) cells were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 2 mM glutamine, and 0.1 mg/ml streptomycin/penicillin (Invitrogen, Melbourne, Australia). For receptor expression and internalization studies, cells were grown to 60 to 80% confluence on 12-well culture plates. Cells were transiently cotransfected with 0.3 e DNA/well wild-type AT1R with or without 0.3 e DNA/well of either arr1-GFP, arr2-GFP, or mutants and various amount of pRc/CMV to a total of 0.6 e DNA/well using LipofectAMINE (Invitrogen). For confocal experiments, cells were plated on collagen-coated 35-mm glass-bottom dishes (MatTek, Ashland, MA) and transiently cotransfected with 30 ng DNA of either wild-type arr1-GFP, arr2-GFP, arr2-HcRed, or mutants with either 70 ng of DNA of AT1R or 2AR (or 10 ng of AT1R-EGFP) and various amount of pRc/CMV to a total of 0.3 e DNA/dish, using LipofectAMINE 2000 (Invitrogen). Cells were assayed 24 to 48 h after transfection.
Confocal Microscopy. HEK-293 cells were changed into serum-free Dulbecco's modified Eagle's medium at least 2 h before commencing confocal experiments. Localization and trafficking of eCarrestins in response to AngII (100 nM) or isoproterenol (10 e) stimulation was viewed with a 63 x 1.2 numerical aperture water-immersion objective on a heated stage at 37°C. Images were collected using a Zeiss LSM 510 META confocal microscope (Carl Zeiss Inc., Thornwood, NY). To compare trafficking between wild-type and mutant -arrestins after 60 min of AngII stimulation, we randomly chose five fields (10 cells per field) and scored whether -arrestins had trafficked into vesicles or were retained at the cell surface. Values are expressed as a percentage of the total number of cells examined.
Bioluminescence Resonance Energy Transfer Assay. -arrestin-AT1R interaction in living cells was examined using bioluminescence resonance energy transfer (BRET) as described previously (Kroeger et al., 2001; Hanyaloglu et al., 2002). In brief, AT1R-Rluc and either wild-type arr1-GFP, arr2-GFP, or mutants were coexpressed in COS-7 cells (American Type Culture Collection). The transfected cells were stimulated with AngII, and readings were measured at 0 and 10 min in the presence of the substrate of R. reniformis luciferase, coelenterazine (Molecular Probes, Eugene, OR). In addition, a time-course study was performed in which cells were incubated in the presence of a long-acting substrate, EnduRen (Promega, Madison, WI). Thereafter, cells were stimulated with AngII, and consecutive readings were collected for 60 min. The binding of arr1-GFP and arr2-GFP to the AT1R-Rluc was measured as changes in the BRET ratio calculated using the equation 515 nm/475 nm)eC(515 nm/475 nm for Rluc alone). Measurements were performed at the wavelengths of 475 and 515 nm. An increased ratio indicates -arrestin-AT1R interaction.
AT1R Internalization. Receptor internalization assays were determined as acid-insensitive [125I]AngII receptor binding as described previously (Thomas et al., 1995).
Results
Trafficking of arr1 and -2 Helix I Mutants Using Confocal Microscopy. To examine the role of helix I in arrestin function, we constructed various -arrestin mutants (Fig. 1) in which hydrophobic residues on one face of the helix were substituted to alanine (arr1LLL/A and arr2LLL/A) and in which positively charged amino acids were replaced with glutamine (arr1RRKK/Q and arr2RRKK/Q). Before investigating the function of these -arrestin mutants, we confirmed that all constructs expressed at equivalent levels in HEK-293 cells (Fig. 2).
The AT1R is a class B GPCR in that it recruits and binds both arr1 and arr2 and forms stable complexes that traffic to deep-core endocytic vesicles. As expected, wild-type arr1-GFP rapidly translocated from the cytoplasm to the cell surface (5 min) and then redistributed into deep-core endocytic vesicles (60 min) after AngII-induced AT1R stimulation (Fig. 3A). In addition, we randomly chose five fields (10 cells per field) and scored whether wild-type or -arrestin mutants trafficked into vesicles or were retained at the cell surface; for the wild-type arr1-GFP, after 60 min of AngII stimulation, all 50 cells examined trafficked into deep-core vesicles. It is interesting that the arr1LLL/A-GFP construct, bearing substitutions within the hydrophobic face, translocated to the cell membrane, and although some pit formation occurred with longer AngII stimulation, the majority was localized to the juxta-membrane region (74% of cells were recruited to the cell membrane versus 26% that were trafficked to vesicles). The arr1RRKK/Q mutant, bearing substitutions within the positively charged residues, also maintained the capacity to translocate to the cell membrane. After 1 h of AT1R activation, although some perimembrane pits and intracellular vesicles formed (in 40% of cells, arr1RRKK/Q remained at the cell surface, whereas in 60% of the cells, it trafficked to vesicles), the robust development of clustered, perinuclear deep-core vesicles was not as apparent compared with the wild-type arr1. Thus, the hydrophobicity of helix I seems to contribute more than the charged residues to the membrane translocation of arr1.
Figure 3B shows the trafficking of wild-type and mutant arr2-GFP after AT1R activation. Wild-type arr2-GFP displayed complete translocation to the membrane followed by subsequent robust redistribution to endocytic vesicles (observed in all 50 cells). In contrast, arr2LLL/A-GFP translocated to the cell surface and formed pits at the membrane (60 min) but did not traffic strongly into cytoplasmic deep-core vesicles (20% vesicular, 80% membrane-localized). Likewise, the arr2RRKK/Q mutant translocated to the cell periphery in response to AngII stimulation. After 1 h of receptor activation, most arr2RRKK/Q persisted at the membrane, with only some small pits/vesicles observed (26% vesicular, 74% membrane-localized). Together, these results indicate that mutation of helix I mostly affects postmembrane sorting of -arrestins after AngII stimulation. Both hydrophobic and positively charged facets of helix I seem to serve a vital role in postmembrane events for both -arrestin isoforms.
To examine the effect of mutating helix I of -arrestin on the trafficking pattern of the AT1R, we coexpressed an AT1R-EGFP receptor with either wild-type arr2-HcRed or arr2LLL/A-HcRed. As shown in Fig. 4, AngII stimulation promoted internalization of the AT1R into deep-core endocytic vesicles where it is colocalized with the wild-type -arrestin. When coexpressed with arr2LLL/A-HcRed, the AT1R-GFP moved from the cell surface to small pits/vesicles, which also contained arr2LLL/A-HcRed; however, these did not resemble coalesced, deep-core vesicles.
To determine whether mutations in helix I can also affect arrestin trafficking in class A GPCRs (which preferentially traffic arr2), we next compared wild-type and mutant arr2-GFP translocation after activation of the 2AR. As shown in Fig. 5, stimulation of 2AR, using the agonist isoproterenol, leads to the membrane targeting of wild-type arr2 and the formation of membrane-localized pits; deep-core vesicles did not develop. Compared with the wild type, the -arrestin mutant containing substitutions within the hydrophobic residues, arr2LLL/A, translocated weakly to the cell membrane and remained in membrane-proximal pits/vesicles. In contrast, the arr2RRKK/Q mutant that has mutations within the positively charged facet rapidly and more completely redistributed to the cell surface and remained there for up to 60 min after stimulation. These data support our findings that alterations in helix I do not prevent the translocation of -arrestins to the plasma membrane after receptor stimulation.
GFP Fusions of Helix I Are Not Localized to the Plasma Membrane. To further investigate whether helix I is potentially amphitropic, we engineered constructs in which the 11 amino acid helix I of arr1 and -2 was fused at the C terminus of GFP to generate GFP-arr1helixI and GFP-arr2helixI. Previous studies have shown that amphitropic segments can act in isolation to target reporter proteins to the cell membrane. A key example of this is the positively charged motif within the signaling molecule K-ras, which promotes strong plasma membrane localization when fused to GFP (tK-GFP) (Apolloni et al., 2000). We expressed the positive control, tK-GFP, and our GFP fusion constructs of helix I from arr1 and arr2 in HEK-293 cells and examined their cellular location using confocal microscopy (Fig. 6). In contrast to tK-GFP, which is confined exclusively to cell surface, both GFP-arr1helixI and GFP-arr2helixI showed a diffuse cytoplasmic localization. This indicates that helix I alone is not sufficient to target membrane anchoring.
Interaction of AT1R with Helix I Mutants. Confocal microscopy can visually detect -arrestin trafficking but does not provide information on the direct association of -arrestin with receptors. Given the altered trafficking of mutant helix I -arrestins in response to AT1R activation, we wondered whether these mutants maintained the capacity to interact with activated AT1R and affect function. Hence, we engineered the AT1R to contain R. reniformis luciferase (AT1R-Rluc) as a C-terminal fusion protein and measured protein-protein interactions with wild-type and mutant versions of arr-GFP via BRET in living cells. Upon AngII stimulation, cells containing AT1R and either wild-type arr1 or arr2 showed a similar increase in BRET signal (Fig. 7A). For comparison, BRET signaling for both arr1 and arr2 was confirmed using the TRHR1 as a positive control—this receptor is a class B receptor that has been reported previously to interact strongly with both arr1 and arr2 (Hanyaloglu et al., 2002). As shown in Fig. 7B, mutation in either the hydrophobic or positively-charged aspects of helix I does not severely impair the strength of the receptor-arrestin interaction measured after AngII stimulation.
We also used a long-acting luciferase substrate, EnduRen, to examine the kinetics and stability of the receptor-arrestin interactions over a 1-h time course. Immediately after AngII stimulation, arr1 (Fig. 7C) and arr2 (Fig. 7D), as well as their LLL/A and RRKK/Q mutants, rapidly associated with the AT1R receptor. The association of wild-type -arrestins was stable and maintained over 60 min, whereas the interaction with the mutants was less so.
Helix I Mutants Support AT1R Internalization. Given that -arrestin mutants translocate to the membrane but do not traffic effectively into deep-core vesicles, we investigated whether they could support AT1R internalization or indeed hinder associations with endogenous arrestins, thereby interfering with receptor internalization. The latter possibility is supported by our observation that helix I mutants still complex with the AT1R. As shown in Fig. 8, expression of wild-type and helix I mutants of both arr1 (Fig. 8A) and arr2 (Fig. 8B) in HEK-293 cells did not affect the rapid and robust internalization (80% after 20-min stimulation) of the AT1R. These data suggest that helix I mutants do not act in a dominant-negative manner. For comparison, in this assay, a well-established arrestin dominant-negative (arr1319eC418) that binds and sequesters clathrin causes a significant reduction of AT1R internalization (Qian et al., 2001).
We next assessed whether these mutants could rescue receptor internalization in a situation in which endogenous -arrestins were absent. For this we used an MEF cell line derived from arr1 and arr1/2KO mice (Kohout et al., 2001). It has been demonstrated previously that AT1R internalization is abrogated in this cell line and that ectopic expression of wild-type arr1 and arr2 can rescue receptor endocytosis (Kohout et al., 2001). As expected, arr1/2KO MEF lines transfected with AT1R alone exhibited a dramatic reduction in receptor internalization (Fig. 9, A and B). Coexpression of either arr1 or arr2 restored receptor internalization, confirming -arrestineCdependent endocytosis for this receptor. It is interesting that LLL/A and RRKK/Q mutants of both arr1 and -2 supported receptor internalization in the arr1/2KO line.
Discussion
The major outcome of the present study is that the amphipathic -helix I of -arrestin is not strongly amphitropic, contrary to previous conjecture (Han et al., 2001). The mutations in both hydrophobic and positively charged facets did not abrogate translocation to the membrane. Moreover, the isolated helix was unable to localize GFP to the plasma membrane, as could be demonstrated for another well-established amphitropic sequence (K-ras). The most obvious effect of mutating helix I was that -arrestins trafficked poorly beyond their initial translocation to the membrane, and few deep-core endocytic vesicles were observed. This occurred despite the capacity of -arrestin mutants to bind the AT1R and support its rapid internalization. We conclude from this that functionally, with respect to trafficking, -arrestins are only required at the membrane and that helix I is not strictly essential for receptor internalization. Whether these mutants are inhibited in respect to other -arrestin functions (e.g., scaffolding of signaling/regulatory molecules) (Luttrell et al., 1999; DeFea et al., 2000; Hall and Lefkowitz, 2002) has not been explored in this study but would be of interest.
Few studies have directly investigated the role of helix I in arrestin function. Structural data revealed that helix I of arrestin is normally constrained in a hydrophobic pocket within the N domain and that it is involved in holding arrestin in its basal state (Vishnivetskiy et al., 2000; Han et al., 2001; Gurevich and Gurevich, 2004). Han et al. (2001) predicted that helix I is displaced from the pocket after -arrestin activation, allowing it to serve as an additional membrane anchor or to enhance receptor binding. However, this was not tested experimentally. Using purified rhodopsin in in vitro arrestin binding assays, Vishnivetskiy et al. (2000) reported that mutation of the leucines (to alanines) in helix I caused a modest increase in the constitutive activity (i.e., receptor binding) of arrestin. This constitutive activity was not enough for us to observe as an appreciable basal translocation of any of the -arrestin mutants (either with substitutions in the hydrophobic or positively charged faces) because these mutants were found to distribute uniformly in the cytoplasm, like wild-type -arrestins. In addition, these mutants did not display increased basal binding to receptor, in our hands, using BRET assays, which are a direct measurement of receptor-arrestin interaction in living cells. After AngII stimulation, the arrestin mutants trafficked to the membrane. Furthermore, helix I in isolation (as a GFP fusion construct) was not strongly membrane-targeted (Fig. 6). It is clear that other parts of -arrestin must be more important in this regard. It is of interest that residues within -arrestin (K233Q, R237Q, and K251Q) have been shown to bind to phosphoinositides, and -arrestin mutants lacking the phosphoinositides binding sequence do not traffic into pits or support receptor internalization (Gaidarov et al., 1999) and prevent arrestin trafficking and light adaptation in Drosophila melanogaster (Lee et al., 2003). A similar phenotype is observed with mutants lacking a binding site (R394A and R396A) for the clathrin-adaptor molecule (Laporte et al., 2000). Thus, -arrestin binding to phosphoinositides or clathrin-adaptor molecule is an essential initial step in the endocytic pathway.
Rather than membrane targeting, the function of helix I in -arrestin activity seemed to correlate more closely with postendocytic routing. Unlike the wild-type proteins, mutant -arrestins were retained near the cell surface, presumably because helix I contributes to the processes that target arrestins into deep-core endocytic vesicles. Despite this increased accumulation at the membrane, expression of these mutant arrestins did not alter AT1R binding and internalization. All helix I mutants rapidly interacted with activated AT1Rs, as measured in BRET assays, and helix I mutants did not compete with endogenous -arrestins for receptor binding in HEK-293 cells. Thus, they do not behave in a dominant-negative manner. Moreover, these mutants all fully supported AT1R internalization in experiments using MEF cell lines deficient in both arr1 and -2. Consistent with other reports (Kohout et al., 2001), AT1R internalization was impaired in arr1/2KO MEF cells, whereas re-expression of either wild-type arr1 or -2 fully rescued receptor internalization, confirming -arrestineCdependent endocytosis for this receptor. Our data clearly demonstrate that, despite their modified trafficking, -arrestin mutants retained the capacity to efficiently drive AT1R internalization. Thus, the integrity of helix I is not paramount for arrestin binding to the receptor or for the promotion of receptor internalization. Instead, helix I was shown to be critical for normal trafficking. This may reflect a role for helix I in the long-term stability of receptor-arrestin complexes and trafficking to endocytic vesicles.
We were surprised that helix I was not amphitropic, especially considering recent evidence that many proteins involved in GPCR signaling contain amphitropic sequences that allow reversible recruitment to the receptor signaling complex. For example, RGS4 and RGS16 proteins, which are involved in enhancing the GTPase catalytic activity of G proteins, require an amphipathic -helix for plasma membrane association (Chen et al., 1999; Bernstein et al., 2000). Likewise, Thiyagarajan et al. (2004) demonstrated that GRK5, which interacts and phosphorylates GPCRs, also contains an amphipathic -helix, which helps to tether GRK5 to the cytoplasmic membrane. This helix in isolation (fused to GFP) mediated membrane localization, and mutations at this helix disrupted membrane targeting (Thiyagarajan et al., 2004). Moreover, ADP-ribosylation factor, a vesicular trafficking regulator, binds to the lipid bilayer with high affinity via a basic amphipathic -helix (Johnson and Cornell, 1999). On the other hand, alterations of the membrane binding motif eliminated membrane interaction (Antonny et al., 1997). Finally, we and others have observed a key role for helix VIII (a positively charged amphipathic helix) in the proximal carboxyl terminus of GPCRs in membrane tethering and receptor activation (Krishna et al., 2002; Mozsolits et al., 2002; Zhong et al., 2004).
In summary, helix I of -arrestin is not strongly amphitropic. In contrast, mutants bearing changes in this helix still traffic to the cell surface, although they seem to be blocked in their capacity to strongly target into deep-core vesicles. Given that these mutants fully support AT1R internalization, we predict that the major function of helix I is postmembrane endocytic targeting rather than amphitropism.
Acknowledgements
We thank R. J. Lefkowitz, M. G. Caron, J. Hancock, and R. Summers for the constructs and cell lines.
doi:10.1124/mol.104.004721.
References
Antonny B, Beraud-Dufour S, Chardin P, and Chabre M (1997) N-terminal hydrophobic residues of the G protein ADP-ribosylation factor-1 insert into membrane phospholipids upon GDP to GTP exchange. Biochemistry 36: 4675eC4684.
Apolloni A, Prior IA, Lindsay M, Parton RG, and Hancock JF (2000) H-ras but not K-ras traffics to the plasma membrane through the exocytic pathway. Mol Cell Biol 20: 2475eC2487.
Bernstein LS, Grillo AA, Loranger SS, and Linder ME (2000) RGS4 binds to membranes through an amphipathic -helix. J Biol Chem 275: 18520eC18526.
Chen C, Seow KT, Guo K, Yaw LP, and Lin SC (1999) The membrane association domain of RGS16 contains unique amphipathic features that are conserved in RGS4 and RGS5. J Biol Chem 274: 19799eC19806.
DeFea KA, Zalevsky J, Thoma MS, Dery O, Mullins RD, and Bunnett NW (2000) -Arrestin-dependent endocytosis of proteinase-activated receptor 2 is required for intracellular targeting of activated ERK1/2. J Cell Biol 148: 1267eC1281.
Gaidarov I, Krupnick JG, Falck JR, Benovic JL, and Keen JH (1999) Arrestin function in G protein-coupled receptor endocytosis requires phosphoinositide binding. EMBO (Eur Mol Biol Organ) J 18: 871eC881.
Gurevich VV and Gurevich EV (2004) The molecular acrobatics of arrestin activation. Trends Pharmacol Sci 25: 105eC111.
Hall RA and Lefkowitz RJ (2002) Regulation of G protein-coupled receptor signaling by scaffold proteins. Circ Res 91: 672eC680.
Han M, Gurevich VV, Vishnivetskiy SA, Sigler PB, and Schubert C (2001) Crystal structure of -arrestin at 1.9 A: possible mechanism of receptor binding and membrane translocation. Structure (Camb) 9: 869eC880.
Hanyaloglu AC, Seeber RM, Kohout TA, Lefkowitz RJ, and Eidne KA (2002) Homo- and hetero-oligomerization of thyrotropin-releasing hormone (TRH) receptor subtypes. Differential regulation of -arrestins 1 and 2. J Biol Chem 277: 50422eC50430.
Holloway AC, Qian H, Pipolo L, Ziogas J, Miura S, Karnik S, Southwell BR, Lew MJ, and Thomas WG (2002) Side-chain substitutions within angiotensin II reveal different requirements for signaling, internalization and phosphorylation of type 1A angiotensin receptors. Mol Pharmacol 61: 768eC777.
Johnson JE and Cornell RB (1999) Amphitropic proteins: regulation by reversible membrane interactions (review). Mol Membr Biol 16: 217eC235.
Kohout TA, Lin FS, Perry SJ, Conner DA, and Lefkowitz RJ (2001) -Arrestin 1 and 2 differentially regulate heptahelical receptor signaling and trafficking. Proc Natl Acad Sci USA 98: 1601eC1606.
Krishna AG, Menon ST, Terry TJ, and Sakmar TP (2002) Evidence that helix 8 of rhodopsin acts as a membrane-dependent conformational switch. Biochemistry 41: 8298eC8309.
Kroeger KM, Hanyaloglu AC, Seeber RM, Miles LE, and Eidne KA (2001) Constitutive and agonist-dependent homo-oligomerization of the thyrotropin-releasing hormone receptor. Detection in living cells using bioluminescence resonance energy transfer. J Biol Chem 276: 12736eC12743.
Laporte SA, Oakley RH, Holt JA, Barak LS, and Caron MG (2000) The interaction of -arrestin with the AP-2 adaptor is required for the clustering of 2-adrenergic receptor into clathrin-coated pits. J Biol Chem 275: 23120eC23126.
Lee S, Xu H, Kang L, Amzel LM, and Montell C (2003) Light adaptation through phosphoinositide-regulated translocation of Drosophila visual arrestin. Neuron 39: 121eC132.
Luttrell LM, Ferguson SS, Daaka Y, Miller WE, Maudsley S, Della Rocca GJ, Lin F, Kawakatsu H, Owada K, Luttrell DK, et al. (1999) -arrestin-dependent formation of 2-adrenergic receptor-Src protein kinase complexes. Science (Wash DC) 283: 655eC661.
Mozsolits H, Unabia S, Ahmad A, Morton CJ, Thomas WG, and Aguilar MI (2002) Electrostatic and hydrophobic forces tether the proximal region of the angiotensin II receptor (AT1A) carboxyl terminus to anionic lipids. Biochemistry 41: 7830eC7840.
Oakley RH, Laporte SA, Holt JA, Caron MG, and Barak LS (2000) Differential affinities of visual arrestin, -arrestin1 and -arrestin2 for G protein-coupled receptors delineate two major classes of receptors. J Biol Chem 275: 17201eC17210.
Qian H, Pipolo L, and Thomas WG (2001) Association of -arrestin 1 with the type 1A angiotensin II receptor involves phosphorylation of the receptor carboxyl terminus and correlates with receptor internalization. Mol Endocrinol 15: 1706eC1719.
Thiyagarajan MT, Stracquatanio RP, Pronin AN, Evanko DS, Benovic JL, and Wedegaertner PB (2004) A predicted amphipathic helix mediates plasma membrane localization of GRK5. J Biol Chem 279: 17989eC17995.
Thomas WG, Baker KM, Motel TJ, and Thekkumkara TJ (1995) Angiotensin II receptor endocytosis involves two distinct regions of the cytoplasmic tail. A role for residues on the hydrophobic face of a putative amphipathic helix. J Biol Chem 270: 22153eC22159.
Thomas WG, Motel TJ, Kule CE, Karoor V, and Baker KM (1998) Phosphorylation of the angiotensin II (AT1A) receptor carboxyl terminus: a role in receptor endocytosis. Mol Endocrinol 12: 1513eC1524.
Vishnivetskiy SA, Hosey MM, Benovic JL, and Gurevich VV (2004) Mapping the arrestin-receptor interface: structural elements responsible for receptor specificity of arrestin proteins. J Biol Chem 279: 1262eC1268.
Vishnivetskiy SA, Schubert C, Climaco GC, Gurevich YV, Velez MG, and Gurevich VV (2000) An additional phosphate-binding element in arrestin molecule. J Biol Chem 275: 41049eC41057.
Zhong M, Navratil AM, Clay C, and Sanborn BM (2004) Residues in the hydrophobic face of putative helix 8 of oxytocin receptor are important for receptor function. Biochemistry 43: 3490eC3498.