Mechanisms of endothelin receptor subtype-specific targeting to distinct intracellular trafficking pathways.

We recently reported that the endothelin (ET) receptor subtypes ET(A) and ET(B) are targeted to distinct intracellular destinations upon agonist stimulation (Bremnes, T., Paasche, J. D., Mehlum, A., Sandberg, C., Bremnes, B., and Attramadal, H. (2000) J. Biol. Chem. 275, 17596-17604). The ET(A) receptor was shown to follow the recycling route of transferrin, whereas ET(B) is targeted to lysosomes for degradation. In the present study we have investigated the mechanisms of ET receptor subtype-specific targeting to distinct intracellular trafficking pathways. Truncation mutants of the ET(A) and ET(B) receptors with deletions of the cytoplasmic carboxyl-terminal tail distal to the palmitoylation site were found to mediate inositol phosphate accumulation and to internalize upon agonist stimulation, although internalization occurred at a slower rate as compared with the wild-type receptors. However, the truncated ET(A) receptor was no longer able to undergo recycling. Rather, both truncation mutants were recognized by beta-arrestin for recruitment to endocytosis and were sorted to lysosomes by a dynamin-dependent internalization pathway. Furthermore, studies of chimeric ET(A) and ET(B) receptors where the cytoplasmic tail of ET(A) was swapped with the corresponding domain of ET(B), and vice versa, revealed that the cytoplasmic tail of ET(B) is required for efficient lysosomal sorting and that signals for targeting to recycling reside in the cytoplasmic tail of the ET(A) receptor.

The multiple physiological effects of the vasoactive peptide endothelin (ET) 1 (1) are mediated by the G protein-coupled receptors (GPCRs) ET A and ET B (2). In the vasculature, ET A receptors residing on the smooth muscle cells mediate prolonged vasoconstriction (3), whereas ET B receptors, which are on the plasma membrane of endothelial cells, are primarily considered to cause NO-mediated vasodilatation (4). In addition, considerable evidence now also supports a role for ET B receptors in the clearance of plasma ET-1 from the circulation (5)(6)(7)(8). In order to elucidate the molecular mechanisms of these distinct physiological responses, we recently characterized the intracellular trafficking pathways of the ET A and ET B receptors (9). Upon agonist stimulation both receptor subtypes are rapidly internalized by mechanisms that depend on G proteincoupled receptor kinase, arrestin, clathrin, and dynamin. Interestingly, the internalized ET A and ET B receptors initially appear to share a common path into Rab5-positive early endosomes. However, the two receptor subtypes are subsequently targeted to different intracellular fates. Whereas the ET A receptor follows the recycling pathway through the pericentriolar recycling compartment and reappears at the plasma membrane, the ET B receptor is directed to lysosomes for degradation. In terms of physiological effects, rapid recycling of the ET A receptor may provide the basis for reestablishment of the signaling response, and thus for the sustained vasoconstriction mediated through this receptor. Conversely, lysosomal targeting of the ET B receptor is consistent with a role for this receptor subtype in the clearance of ET-1 from the circulation. In this respect, it was recently also demonstrated that ET-1 is cotransported with ET B receptors to lysosomes (10). The mechanisms that allow for sorting of GPCRs to recycling versus degradation in lysosomes are unknown. However, recent evidence obtained with the substance P receptor and the protease-activated receptor-1 (PAR-1) indicates that the signals for sorting may reside in the cytoplasmic carboxyl-terminal tail (11). In the latter report analysis of chimeric substance P and PAR-1 receptors demonstrated that the cytoplasmic carboxyl-terminal domains of these receptors contained information critical for receptor-specific targeting to recycling or lysosomal degradation, respectively.
The aims of the present study were to investigate the mechanisms of ET receptor subtype-specific targeting to distinct intracellular trafficking pathways, i.e. recycling in the case of the ET A receptor and lysosomal sorting in the case of the ET B receptor. The kinetics of ET receptor internalization and recycling were investigated by monitoring 125 I-ET-1 trafficking in the absence and presence of Rab protein mutants known to interfere with the recycling process. The data provided in the present study corroborate the evidence that ET A receptors follow the recycling pathway through the pericentriolar recycling compartment via a transport route controlled by Rab5 and Rab11. To elucidate the role of the carboxyl-terminal domains of the ET receptors in intracellular trafficking, truncation mutants of the ET A and ET B receptors with deletions of the cytoplasmic tail were constructed and analyzed in transfected Chinese hamster ovary (CHO) cells. Furthermore, chimeric ET receptors where the carboxyl-terminal domains of ET A and ET B were interchanged were subjected to studies of agonist-* This work was supported by the MSD-Medinnova Cardiovascular Research Fund and grants from the National Research Council. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ Both authors contributed equally to this work. § To whom correspondence should be addressed: MSD-Cardiovascular Research Center, Institute of Surgical Research, A3.1013, The National Hospital, N-0027 Oslo, Norway. Tel.: 47-23071396; Fax: 47-23073530, E-mail: havard.attramadal@klinmed.uio.no. 1 The abbreviations used are: ET, endothelin; GPCR, G protein-coupled receptor; PAR-1, protease-activated receptor-1; CHO, Chinese hamster ovary; GFP, green fluorescent protein; Tf, transferrin; DiI-LDL, dioctadecyl-tetramethylindocarbocyanine-low density lipoprotein. dependent internalization and intracellular trafficking. Strikingly, the carboxyl-terminally truncated ET A receptor lost the ability to undergo recycling. Moreover, the chimeric ET B receptor with the carboxyl-terminal tail of ET A was capable of recycling, whereas the chimeric ET A receptor with the carboxylterminal domain of ET B was rapidly targeted to lysosomes. Both ET receptor subtypes with deletions of the cytoplasmic tail were targeted to lysosomal compartments in an agonist-dependent manner, although internalization occurred at a slower rate as compared with the wild-type receptors. However, overexpression of ␤-arrestin was able to rescue the reduced rate of internalization. Thus, although the cytoplasmic tail of ET B augments the rate of internalization, lysosomal trafficking appears to be a default pathway. Sorting to recycling, on the other hand, appears to depend on a signal residing in carboxylterminal domain of the ET A receptor.

EXPERIMENTAL PROCEDURES
Plasmid Constructs-Subcloning of the cDNAs of the human endothelin receptors ET A and ET B into pcDNA1 has been described previously (9). The cDNAs of hemagglutinin epitope-tagged wild-type dynamin and the dynamin K44A mutant (12) in pRK5 were gifts from Dr. C. Van Koppen (University of Essen, Germany). Rat ␤-arrestin-1 and ␤-arrestin-2 were in the expression vector pCMV5 (13). Their respective mutants V53D and V54D in pcDNA1 were from Dr. M. G. Caron (Duke University, Durham, NC). The Rab11 S25N and Rab5 Q79L cDNA constructs in pcDNA1 were from Dr. H. Stenmark (The Norwegian Radium Hospital, Oslo, Norway). Subcloning of the human transferrin receptor cDNA into pcDNA1 has been described previously (9).

Construction of ET Receptor Truncation Mutants and ET Receptor Chimera-Truncation mutants of ET A (ET A T) and ET B (ET B T) with
deletion of the cytoplasmic carboxyl-terminal tail were constructed in pcDNA1 by polymerase chain reaction-directed mutagenesis replacing Gln-390 and Gln-406, respectively, by stop codons and introduction of a downstream XbaI restriction site. The primers for ET A T were 5Ј-ccctcttcatttaagccg-3Ј and 5Ј-gctctagagcctagtaacagcagcagcagaggcatgactgg-3Ј (XbaI site underlined), and the primers for ET B T were 5Ј-gatcccaatagatgtgaac-3Ј and 5Ј-gctctagagcctagcaccagcagcataagcatgacttaaag-3Ј. The EcoRI-XbaI fragments of the wild-type receptors in pcDNA1 were replaced by the EcoRI/XbaI-digested polymerase chain reaction products. The cDNA constructs were verified by DNA sequence analysis using the dideoxy chain termination method. Construction of ET A -GFP and ET B -GFP in pEGFP-N1 (CLONTECH) has been described previously (9). ET A T-GFP and ET B T-GFP in pEGFP-N1 were made by introducing a BamHI site at the carboxyl-terminal truncation site for in frame cloning of the ET A T and ET B T sequences into pEGFP-N1. The primers for construction of ET A T-GFP were 5Ј-ccctcttcatttaagccg-3Ј and 5Ј-cgcggatcccggtaacagcagcagcagaggcatgactgg-3Ј (BamHI site underlined), and the primers for ET B T-GFP were 5Ј-gatcccaatagatgtgaac-3Ј and 5Ј-cgcggatcccggcaccagcagcataagcatgacttaaag-3Ј. The polymerase chain reaction products were digested with EcoRI and BamHI for insertion into EcoRI/BamHI-digested ET A -GFP and ET B -GFP. The chimeric receptors ET A/B and ET B/A were generated by interchanging the carboxyl-terminal cytoplasmic domains of the ET A and ET B receptors at the conserved EcoRI site located in the 7th transmembrane domain of the receptor cDNAs. Briefly, the EcoRI-XbaI fragment of pcDNA1-ET B was exchanged with the corresponding fragment of pcDNA1-ET A and vice versa. The ET A/B -GFP and ET B/A -GFP constructs were made by similarly interchanging the cytoplasmic tails of ET A -GFP and ET B -GFP in pEGFP-N1.
Cell Lines and Transfection-CHO-K1 cells (ATCC number CRL-61) were maintained in Ham's F-12 Kaighn's modified medium supplemented with 10% fetal bovine serum (Life Technologies, Inc.) and 50 g/ml garamycin in a humidified atmosphere containing 5% CO 2 at 37°C. Cells were grown to 40 -60% confluency before transient transfection using LipofectAMINE reagent (Life Technologies, Inc.) according to the manufacturer's instructions. For fluorescence microscopy experiments, cells were plated onto glass coverslips before Lipo-fectAMINE transfection.
Kinetics of 125 I-ET-1 Internalizaton-Internalization of ET A and ET B receptors was determined using 125 I-ET-1 (Amersham Pharmacia Biotech) as we have described previously (9). Briefly, 24 h after transfection the cells were placed on ice, washed, and preincubated in binding buffer (Hanks' balanced salt solution with 20 mM Hepes, pH 7.4, 0.2% bovine serum albumin, and 0.1% glucose) before binding of 50 pM 125 I-ET-1 in binding buffer for 3 h at 4°C to obtain equilibrium binding. The cells were washed extensively in ice-cold phosphate-buffered saline to remove unbound radioligand before incubation in prewarmed binding buffer at 37°C for various time points. The cells were subsequently treated two times for 10 min at 4°C with 0.5 M acetic acid, 0.15 M NaCl, pH 2.5 (ET A ) or 50 mM glycine-HCl, 0.5 M NaCl, pH 2.2 (ET B ), to strip off surface-bound radioligand. Acid-stripped cells were then lysed for 10 min in 1 M NaOH, before the fraction of surface-bound ET-1 in the acid wash and the fraction of internalized ET-1 in the lysate were determined by ␥-spectrometry. Internalized receptor was calculated as the fraction of radioactivity that remained associated with the cells after acid wash.
Western Blot Analysis-To verify the expression of wild-type and mutant arrestins, Western blot analysis of samples from transiently transfected CHO cells was performed. The immunoblots were processed as described previously (9), using antisera against rat ␤-arrestin-1 and ␤-arrestin-2 (13), and subsequently horseradish peroxidase-conjugated donkey anti-rabbit IgG, and the ECL system (Amersham Pharmacia Biotech) for detection of immunoreactivity.
Phosphoinositide Hydrolysis-Determination of ET-stimulated inositol phosphate accumulation was performed after transfection of CHO cells as described previously (9). Briefly, the cells were metabolically labeled with myo-[ 3 H]inositol for 24 h and subsequently treated with 20 mM LiCl in cell culture medium with or without 0.1 M ET-1 with six parallels at each time point. Total inositol phosphates were separated by anion exchange chromatography on Dowex AG1-8x columns, and 3 H was determined by liquid scintillation counting.
Radioligand Binding-Radioligand binding analysis of membrane preparations of transiently transfected CHO cells was performed according to Elshourbagy et al. (14) with minor modifications (9). Briefly, equilibrium binding conditions were obtained after 1 h at 37°C, and saturation binding analysis was performed using 2-1000 pM 125 I-ET-1 in the absence or presence of a 100-fold excess of unlabeled ET-1. The samples were filtered through Whatman GF/C filters presoaked with 0.05% bovine serum albumin to remove unbound ligand. After three washes in 0.5 M NaCl, the filters were subjected to ␥-spectrometric analysis.
Analysis of Intracellular Trafficking of Receptor-GFP Fusion Proteins by Fluorescence Microscopy-For analysis of ET receptor trafficking, transfected CHO cells on glass coverslips were placed on ice and washed in ice-cold phosphate-buffered saline before binding of ET-1 (200 nM, 1 h at 4°C) in binding buffer (Hank's balanced salt solution with 20 mM Hepes, 0.2% bovine serum albumin, and 0.1% glucose). The cells were subsequently incubated for various time points at 37°C for internalization of the receptor-ET-1 complexes. The cells were fixed in 4% paraformaldehyde before mounting onto object slides using ProLong (Molecular Probes, Inc., Eugene, OR). The specimens were investigated using a Zeiss Axiovert-100 fluorescence microscope with a 100 ϫ 1.3 oil immersion objective. Images were processed and overlaid using Photoshop 5.5 and the Metamorph Imaging System.
Uptake of Red Fluorescent Transferrin and LDL-Loading of the pericentriolar recycling compartment with 25 g/ml tetramethylrhodamine-conjugated transferrin (Tf) (Molecular Probes) was performed in CHO cells cotransfected with ET receptor-GFP cDNA and cDNA for the human transferrin receptor as described previously (9). For labeling of lysosomes, LDL-DiI (Molecular Probes) was bound to the cells together with ET-1, and internalization was performed for 60 min allowing receptor-bound LDL to enter lysosomal compartments. The cells were fixed in 4% paraformaldehyde before mounting of the samples and investigation by fluorescence microscopy. Fig. 1A is shown the characteristic kinetics of ET A and ET B receptor internalization as determined by 125 I-ET-1 uptake at 37°C in transiently transfected CHO cells. As shown, the time course of 125 I-ET-1 internalization is principally different in ET A receptor-expressing cells as compared with ET B receptor-expressing cells. Whereas internalization of the ET B receptor is characterized by a monophasic time course that asymptotically increases toward a maximum level, the ET A receptor displays a biphasic internalization curve. Internalization of the ET A receptor peaks after 10 min of agonist stimulation and subsequently declines from 10 to 60 min. As shown in Fig. 1, B and C, the sum of surface-bound and inter-nalized 125 I-ET-1 was constant during the time course of the assays for both the ET A and the ET B receptors. Thus, a decline in the fraction of internalized ET A receptors in the time span from 10 to 60 min reflects recycling of ligand-bound ET A back to the cell surface. To substantiate the kinetic evidence of ET A receptor recycling and to provide further insights into the mechanisms of ET A receptor recycling, we employed Rab protein mutants known to interfere with the recycling process. Rab11 S25N, a dominant negative form of Rab11 retained in the GDP-bound state, has been shown to block transport from early endosomes to the recycling compartment (15), and Rab5 Q79L, a GTPase defective mutant of Rab5, has been shown to slow down the recycling process by increasing homotypic fusion of early endosomes (16). In Fig. 2 is shown how these Rab mutants affect the kinetics of ET A ( Fig. 2A) and ET B receptor (Fig. 2B) internalization in CHO cells. Strikingly, cotransfection of Rab11 S25N altered internalization of the ET A receptor from a biphasic event consistent with recycling to a monophasic process similar to that characteristic of the ET B receptor. A less dramatic, albeit similar alteration in the kinetics of ET A receptor internalization was observed with cotransfection of Rab5 Q79L. The alterations in the kinetics of ET A receptor recycling caused by Rab11 S25N and Rab5 Q79L can be visualized even more clearly by transforming the data obtained from the time course of ET A receptor internalization in Fig. 2A. Thus, in Fig. 2C changes in the rate of ET A receptor externalization during the time span from 10 to 30 min after initiation of agonistinduced internalization were calculated. As shown, in CHO cells transfected with ET A receptor alone, the level of surface receptors increased, consistent with net externalization during this time. In cells cotransfected with Rab5 Q79L or Rab11 S25N externalization of ET A receptor in the same time span was substantially blunted or even reversed. Conversely, neither Rab5 Q79L nor Rab11 S25N altered the rate of ET B receptor internalization during the same time span. However, the initial rate of ET B receptor internalization appeared to be lower in the presence of Rab11 S25N. Taken together, these data show that when the recycling pathway is inhibited by mutant Rab proteins, the typical biphasic internalization curve of recycling ET A receptors shifts to a monophasic process characteristic of terminally internalized ET B receptors. Hence, the distinct intracellular pathways of the ET A and the ET B receptors appear to reflect on the internalization kinetics of these receptors.

Effects of Rab Mutants on ET Receptor Internalization and Recycling-In
Internalization and Intracellular Trafficking of ET Receptors with Deletion of the Cytoplasmic Carboxyl-terminal Tail-To investigate the role of the cytoplasmic carboxyl-terminal tail of the ET receptors in internalization and intracellular trafficking, we constructed deletion mutants of ET A (ET A T) and ET B (ET B T) by introducing stop codons at positions Gln-390 and Gln-406, respectively, as illustrated in Table I. Thus, both mutants retained the clusters of cysteines representing the putative sites of palmitoylation (17,18). As shown in Table II, the ligand binding characteristics of these truncated receptors did not differ significantly from those of wild-type ET A and ET B receptors. Furthermore, deletion of the cytoplasmic carboxylterminal tail did not appear to alter the capacity of the ET receptors to activate phospholipase C. As shown in Fig. 3, C and D, ET-1-stimulated accumulation of inositol phosphates was nearly identical for the truncated receptors and their wildtype counterparts. However, agonist-stimulated internaliza- tion of ET A T and ET B T was substantially impaired as compared with wild-type ET A and ET B , respectively (Fig. 3, A and  B). The biphasic internalization curve characteristic for the recycling ET A wild-type receptor was abolished in agoniststimulated internalization of ET A T. Thus, the monophasic internalization of ET A T provides evidence that deletion of the carboxyl-terminal tail of ET A prevents recycling of this receptor subtype. Deletion of the cytoplasmic tail of ET B reduced the rate and extent of agonist-stimulated internalization of this receptor subtype, indicating that the mechanism for fast sorting of ET B may reside in its carboxyl-terminal tail.
To investigate the intracellular trafficking pathways of the truncation mutants ET A T and ET B T, we constructed GFP fusion proteins of these receptor mutants for fluorescence microscopy studies. ET A T-GFP and ET B T-GFP were transiently transfected into CHO cells and subjected to analysis of agoniststimulated internalization. The GFP-tagged wild-type receptors ET A -GFP and ET B -GFP were analyzed in parallel for comparison. ET A -GFP colocalized with red fluorescent transferrin in the pericentriolar recycling compartment after 15 min of ET-induced internalization (Fig. 4, A-C) consistent with our previous report (9). ET-induced internalization of ET A T-GFP, however, was associated with spread endosome-like structures throughout the cytoplasm with only minor colocalization in the recycling compartment (Fig. 4, D-F). This finding indicates that the transport route of the carboxyl-terminally truncated ET A receptor differs from its wild-type counterpart. To investigate whether the spread vesicular distribution of internalized ET A T was consistent with transport to lysosomes, we labeled the lysosomal compartments of ET A -GFP-or ET A T-GFP-trans-fected cells with red fluorescent LDL (DiI-LDL). DiI-LDL was bound to the cells simultaneously with ET-1 at 4°C, and internalization was performed at 37°C as described under "Experimental Procedures." As shown in Fig. 5, A-C, ET A -GFP was able to recycle efficiently, and no colocalization with LDL in lysosomes was seen after 60 min of internalization. However, the spread endosome-like structures of internalized ET A T-GFP demonstrated extensive colocalization with red fluorescent LDL (Fig. 5, D-F). ET B -GFP was not found to enter the recycling compartment (Fig. 4, G-I), and the same was the case for the truncated ET B T-GFP (Fig. 4, J-L). Rather, both ET B -GFP (Fig. 5, G-I) and ET B T-GFP (Fig. 5, J-L) were found to accu-   mulate in LDL-positive lysosomes after 60 min of internalization. Thus, deletion of the cytoplasmic tail of ET B did not seem to alter the intracellular trafficking route of this subtype. Taken together, our data suggest that the cytoplasmic tail of ET A contains information for receptor recycling. B T were found to both internalize and follow a typical intracellular trafficking route to lysosomes, we pursued their mechanisms of internalization by employing wild-type and dominant negative mutants of the regulatory proteins arrestin and dynamin. 125 I-ET-1 was bound to the transfected CHO cells at 4°C, and agonist-stimulated internalization was subsequently analyzed at 37°C. As shown in Fig. 6, cotransfection with the dominant negative mutants ␤-arrestin-1 V53D or ␤-arrestin-2 V54D significantly reduced the rate of internalization of both the ET A receptor (Fig. 6A) and the ET B receptor (Fig. 6D), consistent with our previous findings (9). However, similar expression levels of the ␤-arrestin mutants did not alter the internalization kinetics of the truncated ET A T and ET B T receptors (Fig. 6,  B and E). To investigate to what extent increasing expression levels of the mutant ␤-arrestins might affect the internalization of the truncated ET receptors, we performed a dose-response study with increasing concentrations of ␤-arrestin-1 V53D or ␤-arrestin-2 V54D cDNA in the transfection of the CHO cells. Western blot analysis was performed to confirm that increasing DNA levels during transfection resulted in proportional elevations in protein levels (Fig. 7E). As shown in Fig. 7, A-D, cotransfection with increasing amounts of ␤-arrestin-1 V53D or ␤-arrestin-2 V54D resulted in a concentrationdependent decrease in the internalization of the wild-type receptors (Fig. 7, A and C), whereas less pronounced effects were seen for the carboxyl-terminally truncated receptors ET A T and ET B T (Fig. 7, B and D). Conversely, cotransfection of wild-type ␤-arrestin-1 or ␤-arrestin-2 significantly augmented the internalization rates of ET A T and ET B T (Fig. 6, C and F). Thus, the ␤-arrestins appear to be able to rescue the internalization defect imposed by deletion of the carboxyl-terminal cytoplasmic tail, indicating that the ␤-arrestins are able to bind and recruit the carboxyl-terminally truncated ET receptors to endocytosis, albeit with lower affinities than the wild-type receptors.

Role of Arrestin and Dynamin in Regulation of ET A and ET B Receptors with Carboxyl-terminal Truncations-Since the carboxyl-terminally truncated receptors ET A T and ET
In Fig. 8, A-D, are shown the effects of dynamin and the dominant negative mutant dynamin K44A on the internalization rate of wild-type and truncated ET receptors. The internalization rates of ET A and ET B are dramatically reduced upon overexpression of dynamin K44A. Dynamin K44A also reduced the rate of ET A T and ET B T internalization (Fig. 8, C-D), indicating that the residual capacity of these receptor mutants to enter endocytosis is through a dynamin-dependent mechanism similar to the wild-type ET A and ET B receptors.
Intracellular Trafficking of Wild-type and Chimeric ET Receptors-To investigate whether the characteristic properties of recycling and lysosomal sorting are controlled by the cytoplasmic tail of ET A and ET B , respectively, we constructed chi-   ligand binding characteristics of the chimeric receptors and the corresponding GFP-tagged variants did not differ significantly from those of the wild-type ET receptors (Table II). For characterization of the intracellular trafficking pathways of ET A/B -GFP and ET B/A -GFP after agonist-stimulated internalization, we performed simultaneous uptake of red fluorescent transferrin (Tf) or red fluorescent LDL as described under "Experimental Procedures." As shown in Fig. 9, A-C, ET A/B -GFP did not colocalize with Tf in the pericentriolar recycling compartment, as opposed to ET A -GFP. Strikingly, ET A/B was found to colocalize with red fluorescent LDL in lysosomes, as shown in Fig. 10, A-C. Conversely, ET B/A reflected the characteristics of the ET A receptor and was found to colocalize with Tf in the recycling compartment (Fig. 9, D-F). Although the carboxylterminal tail of ET A conferred capacity of ET B/A to be sorted to the pericentriolar recycling compartment, some degree of lysosomal sorting could also be detected in a minor fraction of the cells (Fig. 10, D-F). Taken together, however, the results obtained with the carboxyl-terminally swapped receptors demonstrate that the cytoplasmic tails of the ET A and ET B receptors play decisive roles in determining the intracellular trafficking pathways of these receptors.

. Effects of dominant negative ␤-arrestin mutants on the internalization kinetics of ET A T and ET B T compared with the wild-type ET receptors. A-F demonstrate the kinetics of agonist-induced internalization of ET A and ET A T (upper panels) and ET B and ET B T (lower panels) in CHO cells cotransfected with wild-type or dominant negative
Agonist-stimulated Internalization and Phosphoinositide Hydrolysis of the Chimeric ET Receptors-The assay of ligandinduced internalization demonstrated that ET A/B internalized more efficiently than wild-type ET A (Fig. 11A). Furthermore, the carboxyl-terminal tail of ET B conferred a monophasic time course of ET-1-stimulated internalization of ET A/B as opposed to the biphasic course of internalization of wild-type ET A . Conversely, the ET B/A chimera internalized at a lower rate and extent than ET B (Fig. 11B), and the carboxyl-terminal tail of ET A switched the internalization kinetics to a biphasic course similar to that of ET A . Interestingly, analysis of the capacity of the chimeric receptors ET A/B and ET B/A to mediate ET-1-induced inositol phosphate accumulation showed that the carboxyl-terminal tails conferred the characteristics of their respective wild-type counterparts. Thus, ET A/B , the ET A chimera bearing the cytoplasmic tail of ET B , was not able to stimulate accumulation of inositol phosphate to the same extent as the ET A wild-type receptor. Conversely, ET B/A , the ET B chimera bearing the cytoplasmic tail of ET A , was able to stimulate higher levels of inositol phosphate accumulation than the ET B wild-type receptor. These data demonstrate that ET-1-stimulated phospholipase C activity reflect on the intracellular trafficking pathways of the ET receptor subtypes. DISCUSSION In the present study we have investigated the mechanisms of ET receptor subtype-specific targeting to distinct intracellular trafficking pathways. Recycling is a characteristic feature of the ET A receptor and a plausible mechanism by which a long lasting signaling response can be achieved upon persistent agonist activation of this subtype. ET A receptors are transported to the pericentriolar recycling compartment upon stimulation with ET, and recycling back to the cell surface appears to occur within 1 h (9). A significant proportion of 125 I-ET-1 has been shown to remain associated with ET A receptors for up to 2 h after endocytosis (19), and ET-1 is neither degraded nor returned to the medium during the first 60 min after its binding to the ET A receptor (20). Thus, the biphasic internalization of 125 I-ET-1 in ET A receptor-transfected cells observed in the present study provides a means of quantitative assessment of the recycling process. As demonstrated, coexpression of ET A receptor and Rab5 Q79L altered the characteristic biphasic pattern of ET A receptor internalization, consistent with the ability of Rab5 Q79L to retard the recycling process (16) and to cause entrapment of ET A receptors in early endosomes (9). The dominant negative mutant Rab11 S25N is considered to be a more potent inhibitor of the recycling process, as transport from early endosomes to the pericentriolar recycling compartment is blocked upon its overexpression (15). As expected, coexpression of Rab11 S25N resulted in a more prominent change in the time course of ET A receptor internalization, altering the biphasic internalization curve to a monophasic event in which externalization of the receptor was blunted. Thus, the internalized receptor accumulated intracellularly and asymptotically reached maximal levels after 30 min of agonist-stimulated internalization. These data not only show that ET receptor trafficking is regulated by Rab proteins along the endocytic pathway but also provide quantitative evidence for recycling of the ET A receptors. As to the latter, the characteristic patterns of ET-1 internalization appeared to reflect the intracellular fate of the ET receptors. Interestingly, when correlating the internalization curves of 125 I-ET-1 in cells transfected with mutant ET receptors to the intracellular trafficking pathways of the corresponding GFP-tagged receptors, we found those with lysosomal sorting, i.e. ET B T, ET A T, and ET A/B , to display internalization curves characteristic of the ET B wildtype receptor. Conversely, with the ET B/A receptor chimera, a biphasic internalization event similar to that of the wild-type ET A receptor was observed, which correlates well with the recycling kinetics of these receptors as analyzed by fluorescence microscopy. It could be argued that inhibition of internalization and endocytic traffic of ET A by coexpression of the Rab mutants or by truncation of the carboxyl-terminal tail of ET A may shift the internalization from a biphasic to a monophasic event by slowing the initial rate of internalization. However, Rab11 S25N blocks recycling, and carboxyl-terminal truncation of ET A redirects the receptor to the lysosomal pathway. These alterations in intracellular transport of ET A are likely to be the primary causes of the shift of the internalization from a biphasic to a monophasic event. Furthermore, in the case of the ET receptor chimeras, the monophasic and biphasic internalization curves correlated with lysosomal transport and recycling, respectively, and not with the initial rates of receptor internalization. In this respect, lysosomal transport of ET A/B was associated with monophasic internalization at higher initial rates than that of the wild-type ET A receptor. Conversely, the recycling ET B/A chimera displayed a biphasic internalization curve despite lower initial rates of internalization as compared with ET B wild type.
In the present work we have demonstrated that the cytoplasmic tail of the ET A receptor is responsible for efficient targeting of this subtype to the recycling pathway. Thus, the carboxylterminal tail of ET A contains a signal, which we have also shown is able to redirect the ET B receptor to a recycling transport route. Although agonist-induced internalization of the carboxyl-terminally truncated ET A T and ET B T receptors was found to be slower than for the corresponding wild-type receptors, both ET receptor mutants were found to accumulate in lysosomal compartments within 60 min of internalization. These results suggest that lysosomal targeting is a default pathway occurring in the absence of a specific signal targeting the receptor to recycling. Our data also demonstrate that the cytoplasmic tail of ET B confers a higher rate of receptor internalization, which leads to efficient lysosomal transport of both ET B and the ET A/B chimera. However, the data do not indicate whether additional signals for lysosomal transport of the ET B receptor reside in the cytoplasmic tail of this subtype. The carboxyl terminus of ET B appears to contain more serine and threonine residues in contexts similar to those motifs described to be phosphorylated by G protein-coupled receptor kinases (21) than that of the ET A receptor. Thus, the cytoplasmic tail of ET B may confer higher affinities of ␤-arrestin binding. Although this remains to be demonstrated, Cramer et al. (22) have reported that the ET B receptor subtype is more efficiently phosphorylated than the ET A receptor subtype after agonist stimulation in transfected CHO cells. Furthermore, this finding was found to correlate with the rapid inactivation of the inositol phosphate response through the ET B receptors as compared with the more sustained activity observed through agonist-activated ET A receptors in these cells.
One other report demonstrates that signals for sorting to a specific pathway may reside in the cytoplasmic carboxyl-terminal tail of GPCRs. By analyzing chimeras of the substance P receptor and the PAR-1 receptor where the cytoplasmic carboxyl-terminal domains of the two receptors were interchanged, Trejo and Coughlin (11) demonstrated that the carboxyl-terminal tails of these GPCRs contain information critical for receptor-specific targeting to recycling or lysosomal degradation, respectively. Furthermore, maximal levels of agonist-stimulated phosphoinositide hydrolysis were higher for PAR-1 bearing the carboxyl terminus of the substance P receptor, and lower for the chimeric substance P receptor bearing the carboxyl terminus of PAR-1, as compared with their wild-type counterparts. Consistently, our analyses also showed higher levels of agonist-stimulated inositol phosphate accumulation for ET B/A than for ET B and lower levels for ET A/B than for ET A . Thus, the signaling properties appear to reflect on the characteristics of receptor desensitization as well as on the intracellular trafficking pathways of a particular receptor.
Our findings that the ␤-arrestin mutants V53D and V54D had no significant effects on the kinetics of ET A T and ET B T internalization suggest that these receptors have reduced affinities for arrestins. However, since wild-type ␤-arrestins were able to rescue the internalization defect of ET A T and ET B T, residual binding sites for ␤-arrestin appear to be present in these carboxyl-terminally truncated receptors. Results from previous studies of the ␤ 2 -adrenergic receptor and the M2 muscarinic cholinergic receptor have demonstrated that binding of ␤-arrestin engages two domains of the receptor interacting with two corresponding regions of the ␤-arrestin molecule (23). Whereas a large region within the amino-terminal half of ␤-arrestin interacts with the third intracellular loop of agonistactivated GPCRs (24), a smaller, positively charged domain termed the phosphorylation recognition domain appears to interact with G protein-coupled receptor kinase-phosphorylated residues in the carboxyl terminus of the receptor (25). The ability of ␤-arrestin to recognize the carboxyl-terminally truncated ET receptors supports previous data showing that carboxyl-terminal truncations of ET A and ET B are still able to undergo ligand-induced desensitization (26 -28). However, recent evidence obtained with the CXCR4 receptor shows that although ␤-arrestin is able to interact functionally with this GPCR both in the presence and absence of the carboxyl-terminal tail of the receptor, and to augment receptor internaliza-tion, truncation of the carboxyl terminus abolished the capacity of ␤-arrestin to interdict G protein coupling (29). The interaction between ␤-arrestin and carboxyl-terminally truncated receptors is probably relatively weak, and recognition of phosphorylated residues in the carboxyl-terminal tail might be crucial for efficient desensitization to occur. Thus, our data indicate that the ability of ET A T and ET B T to stimulate inositol phosphate accumulation to the same levels as their wild-type counterparts could be the sum of less efficient desensitization and a slower rate of internalization.
Differential regulation of GPCRs might also be due to the existence of specificity in the interaction between arrestin family members and distinct receptors. In a recent report, Oakley et al. (30) identified two classes of GPCRs as follows: class A receptors, including the ET A receptor and the ␤ 2 -adrenergic receptor; and class B receptors, including the V2 vasopressin receptor and the substance P receptor. Class A receptors displayed higher affinities for ␤-arrestin-2 than for ␤-arrestin-1 and visual arrestin, whereas class B receptors were found to bind with similar affinities to all three arrestins. Furthermore, class B receptors generally bound with higher affinities to arrestins due to the presence of a cluster of serines in their cytoplasmic tail, which is not present in class A receptors (31,32). Interestingly, the ET B receptor contains a cluster of serines in its extreme carboxyl terminus, which is not present in the ET A receptor. Although the affinities of arrestin binding to the ET receptors have not been determined, it is tempting to hypothesize that the ET B receptor may conform to the class B receptor characteristics of high affinity binding to ␤-arrestin. Thus, a more stable binding to ␤-arrestin caused by the phosphorylated carboxyl terminus of ET B could account for the more efficient desensitization as well as the faster internalization rates of ET B and ET A/B . Conversely, less stable association with ␤-arrestin due to the less phosphorylated ET A tail could serve to explain both the higher levels of inositol phosphate accumulation and the lower rates of internalization of ET A and ET B/A .
Based on the data of the present study, we conclude that the cytoplasmic tails of the ET receptors play decisive roles in the intracellular trafficking of these receptors. The future challenge will be to elucidate what specific signals residing in the carboxyl-terminal tail of the ET A receptor direct this receptor subtype to recycling. Arrestin evidently plays a key role in modulating signaling responses through GPCRs. Not only receptor desensitization and internalization but also the rate of receptor resensitization seems to be determined by arrestin, the latter correlating with the ability of arrestin to dissociate from the receptor (31). In addition, emerging evidence implicates ␤-arrestin as a scaffolding protein for GPCR-orchestrated stimulation of mitogen-activated protein kinase (33,34). However, the putative role of arrestin in targeting receptors to distinct intracellular trafficking pathways remains to be investigated. We believe our model system is of general interest in the field of GPCRs, as we deal with two receptors that are activated by the same ligand but that follow different intracellular routes and show distinct signaling properties.