Role of the Differentially Spliced Carboxyl Terminus in Thromboxane A2 Receptor Trafficking

The thromboxane A2 receptor (TP) is a G protein-coupled receptor that is expressed as two alternatively spliced isoforms, α (343 residues) and β (407 residues) that share the first 328 residues. We have previously shown that TPβ, but not TPα, undergoes agonist-induced internalization in a dynamin-, GRK-, and arrestin-dependent manner. In the present report, we demonstrate that TPβ, but not TPα, also undergoes tonic internalization. Tonic internalization of TPβ was temperature- and dynamin-dependent and was inhibited by sucrose and NH4Cl treatment but unaffected by wild-type or dominant-negative GRKs or arrestins. Truncation and site-directed mutagenesis revealed that a YX 3φ motif (whereX is any residue and φ is a bulky hydrophobic residue) found in the proximal portion of the carboxyl-terminal tail of TPβ was critical for tonic internalization but had no role in agonist-induced internalization. Interestingly, introduction of either a YX 2φ or YX 3φ motif in the carboxyl-terminal tail of TPα induced tonic internalization of this receptor. Additional analysis revealed that tonically internalized TPβ undergoes recycling back to the cell surface suggesting that tonic internalization may play a role in maintaining an intracellular pool of TPβ. Our data demonstrate the presence of distinct signals for tonic and agonist-induced internalization of TPβ and represent the first report of a YX 3φ motif involved in tonic internalization of a cell surface receptor.

tropin, M 2 muscarinic, and thrombin receptors also undergo tonic internalization (2)(3)(4)(5). Although no particular motif responsible for tonic internalization of GPCRs has been identified, tyrosine-containing (YXX and NPXY) and dileucine motifs have been shown to be determinants for a number of other receptor types (1). Various studies have demonstrated direct interaction between YXX motifs and the chain of the clathrin-associated proteins AP-1, AP-2 (Ref. 6 and references therein), and AP-3 (7,8), allowing the efficient targeting of transmembrane proteins containing these motifs to clathrin-coated vesicles.
Thromboxane has been implicated in a number of cardiovascular, bronchial, and kidney diseases (9,10). It is produced by the sequential metabolism of arachidonic acid by cyclooxygenase and thromboxane synthase following activation of a variety of cell types including platelets, macrophages, and vascular smooth muscle cells (11). Thromboxane is a strong activator of platelet aggregation and smooth muscle cell proliferation and mediates its effects via interaction with a specific GPCR. The thromboxane A 2 receptor (TP) is encoded by a single gene that is alternatively spliced in the carboxyl terminus resulting in two variants, TP␣ (343 residues) and TP␤ (407 residues) that share the first 328 amino acids (12)(13)(14).
In a previous study, we demonstrated that TP␤, but not TP␣, undergoes agonist-induced internalization in a variety of cell types (15). Internalization of TP␤ was dynamin-, GRK-, and arrestin-dependent in HEK293 cells, suggesting the involvement of receptor phosphorylation and clathrin-coated pits in this process. Additional characterization of the role of arrestins in this process revealed that arrestin-3 coexpression promoted agonist-induced internalization of both TP␣ and TP␤ but not of a mutant truncated after residue 328. Analysis of various carboxyl-terminal deletion mutants revealed that a region between residues 355 and 366 in TP␤ was essential for agonistpromoted internalization. During the course of these studies, we observed that TP␤, but not TP␣, also undergoes tonic internalization. In the present study, we characterize the mechanisms involved in tonic internalization of TP␤. These studies reveal that a YX 3 motif found in the proximal portion of the carboxyl-terminal tail of TP␤ is responsible for tonic internalization.

EXPERIMENTAL PROCEDURES
Cell Culture and Expression Systems-Human embryonic kidney cells (HEK293) were maintained in Dulbecco's modified Eagle's Me-dium (DMEM, Life Technologies, Inc.) supplemented with 10% fetal bovine serum. P388D1, A431, Mia Paca (a human pancreatic carcinoma cell line of epithelial morphology), COS-1 and CHO cells were obtained from ATCC and grown as recommended by the supplier. Cells were kept in a humidified atmosphere of 95% air, 5% CO 2 at 37°C. Transfections were done with Fugene-6 (Roche Molecular Biochemicals) following the manufacturer's recommendations.
Agonist-induced Internalization Assays-For quantitation of receptor internalization, ELISA assays were performed as described (15). Cells were plated at 6 ϫ 10 5 cells per 60-mm dish, transfected with 6 g of DNA and split after 24 h into 6 wells of a 24-well tissue culture dish coated with 0.1 mg/ml poly(L-lysine) (Sigma). After another 24 h, the cells were washed once with phosphate buffered saline (PBS) and incubated in DMEM at 37°C for several minutes. The TP agonist U46619 was added at a concentration of 100 nM in prewarmed DMEM to the wells, the cells were incubated for 2 h at 37°C, and reactions were stopped by removing the medium and fixing the cells in 3.7% formaldehyde in Tris-buffered saline (TBS) for 5 min at 22°C. The cells were washed three times with TBS and nonspecific binding was blocked with TBS containing 1% bovine serum albumin (BSA) for 45 min at room temperature. The first antibody (monoclonal HA 101R from Babco or FLAG M1 from Sigma) was added at a dilution of 1:1000 in TBS/BSA for 1 h at 22°C. The cells were washed three times with TBS and then reblocked for 15 min at 22°C. Incubation with goat anti-mouse-conjugated alkaline phosphatase (Bio-Rad) diluted 1:1000 in TBS/BSA was carried out for 1 h at 22°C. The cells were washed three times with TBS, and a colorimetric alkaline phosphatase substrate was added. When adequate color change was reached, 100-l samples were taken for colorimetric readings. Cells transfected with pcDNA3 alone were studied concurrently to determine background, and all experiments were done in triplicate.
Immunofluorescence Microscopy-Cells (HEK293, P388D1, Mia Paca, A431, and CHO) were grown in 35-mm dishes on coverslips and transfected as described above with 2 g of DNA/well. After 48 h, cells were incubated with FLAG M1 antibody (1:500 dilution) for 1 h at 4°C in DMEM supplemented with 1% BSA and 1 mM CaCl 2 . Cells were washed twice with PBS containing 1 mM CaCl 2 and then treated with 100 nM U46619 for 1 h at 37°C in DMEM with 20 mM HEPES, pH 7.4, 0.5% BSA, 1 mM CaCl 2 . The cells were then fixed with 3.7% formaldehyde/PBS for 15 min at 22°C, washed with PBS/CaCl 2 , and permeabilized with 0.05% Triton X-100/PBS/CaCl 2 for 10 min at 22°C. Nonspecific binding was blocked with 0.05% Triton X-100/PBS/CaCl 2 containing 5% nonfat dry milk for 30 min at 37°C. Goat anti-mouse fluorescein isothiocyanate (FITC)-conjugated secondary antibody (Molecular Probes) was then added at a dilution of 1:150 for 1 h at 37°C. The cells were washed six times with permeabilization buffer with the last wash left at 37°C for 30 min. Finally, the cells were fixed with 3.7% formaldehyde as described. Coverslips were mounted using Slow-Fade mounting medium (Molecular Probes) and examined by microscopy on a Nikon Eclipse E800 fluorescence microscope using a Plan Fluor 60ϫ objective under oil immersion. Cells expressing the lowest levels of transfected proteins, but clearly above those of nonexpressing cells, were chosen for view. Images were collected using the QED Camera software and processed with Adobe Photoshop v. 3.0. The images were then imported into the NIH Image 1.62b27f software and black and white binary images were used for quantitation of fluorescence. Measurements were made separately of areas corresponding to the intracellular content (whole cell minus the plasma membrane) and to the whole cell. The percentage of internalized receptors was then calculated as the intracellular to whole cell fluorescence ratio ϫ 100. Results shown represent the mean Ϯ S.E. of three independent experiments where immunofluorescence of at least ten cells was evaluated in each experiment.
Recycling Studies-The recovery of receptors at the cell surface was evaluated using a modified version of the immunofluorescence microscopy procedure described above. Following internalization, cells were stripped of M1 antibody with two quick washes of cold PBS, 1 mM EDTA. Cells were then washed with PBS to remove residual EDTA and then reincubated for 1 h at 37°C in DMEM with 20 mM HEPES, pH 7.4, 0.5% BSA, 1 mM CaCl 2 to allow receptors to recycle back to the cell surface. The cells were fixed with 3.7% formaldehyde/PBS and treated as above for immunofluorescence analysis.

RESULTS AND DISCUSSION
A schematic representation of the carboxyl terminus of the two isoforms of the human thromboxane A 2 receptor is shown in Fig. 1. To investigate tonic internalization of TP␣ and TP␤, epitope-tagged receptors were transiently expressed in HEK293 cells. Previous studies have shown that the agonist binding affinities of TP␣ and TP␤ are similar (14)(15)(16)(17). Moreover, addition of a FLAG epitope at the amino terminus of the receptors did not alter ligand affinities, nor did it affect the activation characteristics of the receptors as determined by their respective EC 50 values for agonist-promoted generation of inositol phosphate (data not shown).
During our immunofluorescence analysis of agonist-induced internalization of thromboxane A 2 receptors, we noted that TP␤ could undergo tonic internalization (15). To further characterize this phenomenon, we performed a series of immunofluorescence studies on cells transiently expressing FLAGtagged TP receptors. Cells were initially incubated with the FLAG antibody at 4°C for 1 h, washed, and then incubated at different temperatures so that tonic receptor internalization could be followed. As shown in Fig. 2A, there is significant redistribution of TP␤ to intracellular compartments following incubation at 37°C whereas minimal internalization of TP␣ is observed. Quantitation of intracellular and total cell fluorescence revealed that ϳ60% of TP␤ was tonically redistributed to a subcellular compartment following a 1-h incubation at 37°C. These observations confirm that TP␤, but not TP␣, undergoes significant tonic internalization. In contrast, when the cells are incubated for 1 h at 4°C or 16°C, TP␣ and TP␤ remain entirely FIG. 1. Schematic representation of the amino acid sequence of the TP␣ and TP␤ carboxyl terminus. Vertical lines illustrate the sites where truncation mutations were made. The indicated residues were individually mutated to alanines (bold, underlined) whereas additional mutations that were introduced into TP␣ are represented with arrows.
at the cell surface. Interestingly, the inability of TP␤ to undergo tonic internalization at 16°C is a property shared with agonist-induced internalization of TP␤ (data not shown) and the ␤ 2 AR (18), but quite distinct from tonic internalization of the transferrin receptor which still occurs at 16°C (data not shown) (18).
To verify that the internalization of TP␤ was not triggered by endogenous production of thromboxane or by trace amounts of agonist in the media, cells were pretreated with the TP antagonist SQ29548 (10 M) or the cyclooxygenase inhibitor indomethacin (10 M) for 15 min at 4°C prior to incubation at 37°C. These treatments had no effect on internalization of TP␤ confirming that the receptor is undergoing tonic internalization (Fig. 2B). The same trafficking properties for TP␣ and TP␤ were observed in P388D1 and A431 cells, which express endogenous TP receptors, and Mia Paca, COS-1 and CHO cells, which lack endogenous TP receptors, as well as in HEK293 cells stably expressing low levels of throm-boxane receptors (15) (data not shown). These results demonstrate that the differential tonic internalization of TP␣ and TP␤ is a property of the receptor.
We previously demonstrated that agonist-induced internalization of TP␤ is dynamin-, GRK2-, and arrestin-dependent (15). Thus, we next determined the role of these proteins in tonic internalization of TP␤. Immunofluorescence analysis of TP␤ redistribution in the presence of dominant-negative mutants of dynamin (19), GRK2 (20), and arrestin-3 (21) was performed. Coexpression of dynamin-K44A inhibited tonic internalization (Fig. 2B), whereas GRK2-K220R and arrestin-3 (201-409) had no effect (data not shown). Interestingly, coexpression of dynamin-K44A, but not GRK2-K220R or arrestin-3 (201-409), also resulted in an ϳ2-fold higher cell surface expression of TP␤ as assessed by ELISA (data not shown). In contrast, dynamin-K44A had no effect on cell surface expression of TP␣ (data not shown). When cells were preincubated with inhibitors of clathrin-coated pit formation such as sucrose and NH 4 Cl, tonic internalization of TP␤ was also suppressed (Fig. 2B). Whereas these results demonstrate that tonic and agonist-induced internalization of TP␤ are both dynamin-dependent, tonically internalized TP␤ is targeted to clathrin-coated pits via a mechanism independent of GRKs and arrestins.
Because the carboxyl terminus of TP␤ appears critical in tonic internalization of the receptor, we next determined whether this function could be ascribed to any particular residues. Progressive deletion mutants were first used to address this question (Fig. 1). All constructs were transiently transfected in HEK293 cells using transfection conditions that yielded comparable levels of receptor expression (ϳ1 pmol/mg protein). The removal of up to 63 residues from the carboxyl terminus (S344Stop) appeared to have no effect on tonic internalization (Fig. 3), whereas agonistinduced internalization was completely blocked (15). However, truncation of an additional 7 amino acids (L337Stop) completely abolished tonic internalization (Fig. 3). Thus, the region found between residues 338 and 344 seems to play a critical role in tonic internalization of TP␤. Comparison of this region of TP␤ (EYSG-TIS) with the corresponding region of TP␣ (TQRSGLQ) suggests that Tyr-339 in TP␤ might be an important component of a tonic internalization motif. A role for tyrosine-based internalization motifs in tonic endocytosis of a variety of receptors has been demonstated (1,6,22). To test this hypothesis, we generated a Y339A mutant TP␤ and characterized tonic and agonist-induced internalization. Indeed, TP␤-Y339A did not undergo any tonic internalization (Fig. 4), suggesting a critical role for Tyr-339 in this process. Additional amino acids between residues 338 and 344 in TP␤ were then individually mutated to alanine in an attempt to identify a motif for tonic internalization. The E338A, S340A, G341A, T342A, and S344A mutants of TP␤ appeared to undergo normal tonic internalization, whereas I343A was completely inhibited (Fig. 4). None of these mutations affected agonist-induced internalization of TP␤ (Fig. 4C), suggesting that the motif for tonic internalization is distinct from the region required for agonist-induced trafficking of TP␤ (15). Thus, our data demonstrate that the YXXXI motif plays a critical role in tonic internalization of TP␤. This sequence is closely related to the YXX motif identified as playing an important role in tonic internalization of the transferrin receptor, T-cell receptor (CD3), lgp-A/ lamp-1, lgp-B/lamp-2, lysosomal acid phosphatase, CI-mannose-6-phosphate receptor, polymeric Ig receptor, and TGN38 receptor (23).
To further clarify the importance of Tyr-339 in tonic internalization, we generated Q338Y and R339Y mutants in the carboxyl terminus of TP␣. Interestingly, introduction of a Tyr at position 339 in TP␣, creating a YXX motif, induced tonic

FIG. 2. Immunofluorescence analysis of TP␣ and TP␤ distribution in HEK293 cells.
A, FLAG-tagged receptors were transiently transfected in HEK293 cells. Cells were incubated with the FLAG antibody at 4°C prior to any other treatment to detect receptors that were present initially at the cell surface. Immunofluorescence detection was performed as described under "Experimental Procedures." Top panel, receptor distribution in cells expressing TP␣ (left) and TP␤ (right) when incubated at 4°C. Middle panel, after a 1 h incubation at 16°C. Bottom panel, after a 1 h incubation at 37°C. B, HEK293 cells, transiently transfected with TP␤, were labeled with FLAG antibody at 4°C prior to incubation at 37°C for 1 h. Tonic internalization of TP␤ in the presence or absence of the TP receptor antagonist SQ29548, the inhibitors of clathrin-coated pit formation NH 4 Cl and sucrose, or dynamin-K44A was analyzed by immunofluorescence and quantitated. Data represent the percentage of intracellular immunofluorescence relative to total immunofluorescence of individual cells. Immunofluorescence was measured using the NIH Image 1.62b7f software. Results shown represent the mean Ϯ S.E. of three independent experiments, where immunofluorescence of at least ten cells was evaluated for each experiment. Refer to "Experimental Procedures" for details. internalization of TP␣ (Fig. 5). Mutation of Leu-342 to Ala in the R339Y mutant inhibited tonic internalization, demonstrating the importance of the hydrophobic residue in this process (Fig. 5). In an effort to determine the importance of the spacing between the Tyr and hydrophobic residues, we introduced a Thr between Gly-341 and Leu-342 in the R339Y mutant (R339Y-T342) to create a YXXX motif similar to that found in TP␤. This latter addition did not affect the internalization induced by the Tyr residue in R339Y, verifying that both YX 2 and YX 3 can function as efficient internalization motifs. Interestingly, a Q338Y mutant (also creating a YX 3 motif but one residue closer to the plasma membrane than in TP␤ and TP␣ (R339Y-T342)) did not induce tonic internalization of TP␣. This suggests that the position of the Tyr in the receptor carboxyl tail is also an important determinant in this process. As expected, none of these mutations conferred agonist-induced internalization of TP␣ (data not shown). Our data suggest that both the distance between the Tyr and the hydrophobic residue and the position of the YX 2-3 motif in the receptor carboxyl tail are important determinants in tonic internalization of the thromboxane receptor. It is interesting to note that a YLGI peptide sequence found in the second intracellular loop of both TP receptor isoforms is evidently not sufficient to induce tonic internalization because this is not observed for TP␣. Moreover, mutation of residues within this motif in TP␤ did not affect tonic internalization (data not shown).
Because tonic internalization has also been reported for other GPCRs (2)(3)(4)(5)24), it is important to consider the biological role of this process. It has been proposed that tonic internalization of the thrombin receptor generates an intracellular pool of receptors that is used to repopulate the cell surface with functional receptors (24). If a similar role can be attributed to tonic internalization of TP␤, we would expect that these receptors would recycle back to the cell surface following tonic internalization. To address this issue, cell surface receptors were labeled with M1 anti-FLAG antibody at 4°C and then allowed to undergo tonic internalization for 1 h at 37°C. The cells were washed briefly with PBS/EDTA to strip the cell surface antibody (which binds in a Ca 2ϩ -dependent manner), reincubated at 37°C, fixed, and then receptor distribution determined by immunofluorescence. TP␤ was only detected intracellularly after cell surface antibody was stripped with PBS/EDTA (Fig. 6,  panel B). However, following incubation at 37°C, there was extensive redistribution of the intracellular receptors to the cell surface (Fig. 6, panel C). This recycling was not caused by new protein synthesis or to transport of new receptors from intracellular stores because visualized receptors originated from the initial labeling of cell surface receptors with antibody. These data suggest that there is constant recycling of the tonically internalized TP␤ between the cell surface and an unidentified intracellular compartment, similar to what has been observed for the thrombin receptor (24). Thus, tonic internalization of TP␤ likely helps to maintain an intracellular pool of functional receptors that recycle to the cell surface to preserve agonist sensitivity.
FIG. 3. Tonic internalization of different carboxyl tail truncation mutants of TP␤. HEK293 cells were transfected with amounts of TP␤ DNA yielding receptor expression of ϳ1 pmol/mg protein for each construct. Cells were labeled with the FLAG antibody at 4°C prior to being incubated at 37°C for 1 h. A, tonic internalization of each receptor construct was analyzed by immunofluorescence. B, quantitation of tonic internalization of the different receptor constructs was performed as described in Fig. 2.
The YXX motif is one of the most extensively characterized motifs within cytosolic domains involved in the targeting of integral membrane proteins. Tyrosine-based sorting signals conforming to the YXX motif have been shown to interact directly with the 1, 2, and 3 subunits of the adaptor complexes AP-1, AP-2, and AP-3, respectively (reviewed in Ref. 7). The critical tyrosine does not need to be phosphorylated and, in fact, the interaction of YXX and may actually be reduced by phosphorylation (25,26). The AP-1 complex associates with the trans-Golgi network and directs the transport of lysosomal enzymes to endosomes, whereas the AP-2 complex associates with the plasma membrane and directs the trafficking of cell surface proteins via clathrin-coated pits. AP-3 is involved in the delivery of proteins to lysosomes and lysosome-related organelles (27). Recent studies also suggest that there is a fourth adaptor-related protein complex, AP-4, that is associated with nonclathrin-coated vesicles in the region of the trans-Golgi network (27,28). The 4 subunit of this complex specifically interacts with a tyrosine-based sorting signal, suggesting that AP-4 is also involved in the recognition and sorting of proteins with tyrosine-based motifs (27).
Ohno et al. (7) investigated the selectivity for interaction of tyrosine-based sorting signals with 1, 2, 3A, and 3B subunits via screening of a combinatorial XXXYXX library using the yeast two-hybrid system. Their results revealed that there was no absolute requirement for the presence of specific residues at any of the X or positions. This contrasted with the critical tyrosine residue that could not be substituted by any other residue without a dramatic decrease in sorting activity (6, 7, 29 -32) and binding affinity for subunits (7,25,26,33,FIG. 4. Identification of the residues involved in tonic internalization of TP␤. Amino acids between residues 338 and 344, inclusively, were individually mutated to alanines. A, immunofluorescence analysis of the tonic internalization of each receptor construct. B, quantitation of intracellular fluorescence, as described in Fig. 2. C, agonistinduced internalization of the receptor mutants. HEK293 cells were transiently transfected with FLAG epitope-tagged receptors and the percentage of receptors remaining at the cell surface after 2 h of stimulation with 100 nM U46619 was measured by ELISA analysis, as described under "Experimental Procedures." The results represent the mean Ϯ S.E. of three independent experiments, each done in triplicate. 34). It was shown that each subunit exhibits a preference for certain XXXYXX signals; however, there was also considerable overlap in specificity (7). Although these studies did not characterize interaction with YXXX motifs, the YX 3 motif found in TP␤ displays a serine at position YϪ3 and a glutamic acid at YϪ1, analogous to one of the specific sequences that binds to 2 (SFEYQPL) (7). Similarly, the asialoglycoprotein receptor has a threonine and a glutamic acid at the YϪ3 and YϪ1 positions, respectively. A serine or threonine are also found at YϪ3 of the YXX motif of the CI mannose-6-phosphate receptor, EGF receptor, and CTLA-4. Similar to TP␤, the CI mannose-6-phosphate receptor, poly(Ig) receptor and HIV gp41 have a serine at position Yϩ1 whereas the CD mannose-6-phosphate receptor and furin have a glycine at position Yϩ2. On the other hand, the 3A subunit preference for position YϪ1 is also a glutamic acid (7). Amino acid differences in or around the signal may possibly confer preferences for targeting to particular cell compartments. A glycine preceding the Tyr in a YXX motif enhances targeting of lamp-1 and acid phosphatase to lysosomes (6 and references therein). The position of the motif within the cytoplasmic domain may also influence its activity (6). Indeed, displacement of the YXX motif in lamp-1 by a single residue with respect to the transmembrane domain was reported to disrupt lysosomal targeting (35), similar to our observations in the TP␣ mutants. Similar findings were also reported when a Tyr was inserted in the cytoplasmic terminus of influenza virus hemagglutinin to generate an artificial internalization signal. In these studies, internalization was dependent on the position of the Tyr relative to the cell membrane indicating that the structural environment of the Tyr was important (22).
Our results identified a distinct motif (YX 3 ) in the carboxyl terminus of a G protein-coupled receptor responsible for tonic internalization. Both the distance between the Tyr and the hydrophobic residue and the position of the YX 3 motif in the receptor carboxyl tail appear to be important determinants in this process. In addition, secondary signals that function in concert with the YX 3 motif may influence the destination of the receptor. Further analysis of sequence and contextual requirements for the function of the YX 3 signal will be necessary to fully understand its specificity. Importantly, two adjacent sequences in the carboxyl terminus of TP␤, the YX 3 motif and the region between residues 355 and 366, seem to distinguish between tonic and agonist-induced internalization, respectively. Molecular analysis of the interaction of these sequences with their recognition molecules will provide additional clues as to their role in receptor trafficking. For example, it will be interesting to determine whether YX 3 and AP-2 directly interact and, if so, which AP-2 subunit contributes to such interaction. Comparatively, the region between residues 355 and 366 may dictate, at least in part, the interaction of the receptor with arrestins (15). This would suggest FIG. 5. Redistribution of receptors in HEK293 cells after introduction of a tyrosine residue at position 339 in the carboxyl tail of TP␣. HEK293 cells were transfected with amounts of DNA yielding receptor expression of ϳ1 pmol/mg of protein for each TP␣ construct. A, redistribution of TP␣ mutants was assessed by immunofluorescence. B, intracellular fluorescence was evaluated as described in Fig. 2. FIG. 6. Tonically internalized TP␤ receptors recycle to the cell surface. FLAG epitope-tagged receptors were transiently transfected in HEK293 cells. A, cells were labeled with the FLAG antibody at 4°C and incubated at 37°C for 1 h to allow tonic internalization. B, the cell surface was then stripped of the remaining FLAG antibody with two quick washes with cold PBS/EDTA (1 mM). C, cells were then reincubated for 1 h at 37°C to permit recycling of receptors to the cell surface (see "Experimental Procedures" for details). Results shown are representative of at least five independent experiments. tight regulation of the factors involved. The YX 3 motif could be continuously available for interaction with proteins involved in tonic internalization whereas ligand occupancy of the receptor might expose the adjacent sequence to proteins involved in agonist-induced internalization.
In summary, our results demonstrate that the alternative splicing of the carboxyl terminus of the thromboxane A 2 receptor generates isoforms that show distinct trafficking characteristics. We had previously shown that TP␤, but not TP␣, could undergo agonist-induced internalization (15). In the present report, we demonstrate that only TP␤ is capable of undergoing tonic internalization. Tonic internalization was attributed to a YX 3 motif, which is distinct from the sequence required for agonist-promoted trafficking and is the first such motif identified for tonic internalization of a GPCR. These findings raise important questions concerning how trafficking differences between TP␣ and TP␤ might contribute to mechanistic differences in the desensitization, resensitization, and/or degradation of these receptors as well as in their overall cellular physiology. Similar studies on other GPCRs will help to further characterize the signals, proteins, and cellular compartments involved in these processes.