Regulation of Angiotensin II Type 1A Receptor Intracellular Retention, Degradation, and Recycling by Rab5, Rab7, and Rab11 GTPases*

Previous studies have demonstrated that the interaction of the angiotensin II type 1A receptor (AT1AR) carboxyl-terminal tail with Rab5a may modulate Rab5a activity, leading to the homotypic fusion of endocytic vesicles. Therefore, we have investigated whether AT1AR/Rab5a interactions mediate the retention of AT1AR·β-arrestin complexes in early endosomes and whether the overexpression of Rab7 and Rab11 GTPases influences AT1AR lysosomal degradation and plasma membrane recycling. We found that internalized AT1AR was retained in Rab5a-positive early endosomes and was neither targeted to lysosomes nor recycled back to the cell surface, whereas a mutant defective in Rab5a binding, AT1AR-(1-349), was targeted to lysosomes for degradation. However, the loss of Rab5a binding to the AT1AR carboxyl-terminal tail did not promote AT1AR recycling. Rather, it was the stable binding of β-arrestin to the AT1AR that prevented, at least in part, AT1AR recycling. The overexpression of wild-type Rab7 and Rab7-Q67L resulted in both increased AT1AR degradation and AT1AR targeting to lysosomes. The Rab7 expression-dependent transition of “putative” AT1AR·β-arrestin complexes to late endosomes was blocked by the expression of dominant-negative Rab5a-S34N. Rab11 overexpression established AT1AR recycling and promoted the redistribution of AT1AR·β-arrestin complexes from early to recycling endosomes. Taken together, our data suggest that Rab5, Rab7, and Rab11 work in concert with one another to regulate the intracellular trafficking patterns of the AT1AR.

family of integral membrane receptor proteins. The AT 1A R is coupled via G q to the stimulation of phospholipase C␤, leading to increases in intracellular diacylglycerol and inositol 1,4,5triphosphate and the release of calcium from intracellular stores (1). Agonist activation of the AT 1A R also leads to the desensitization of AT 1A R second messenger responses and the removal of cell-surface AT 1A R into the intracellular compartment of the cell (2)(3)(4). The agonist-stimulated desensitization and endocytosis of many GPCRs is initiated by GPCR kinasemediated phosphorylation, followed by ␤-arrestin binding (5). Both ␤-arrestin-dependent and ␤-arrestin-independent mechanisms of AT 1A R endocytosis have been reported; and as such, the precise mechanism(s) regulating AT 1A R internalization remain unclear (6,7).
The AT 1A R is a member of a class of GPCRs that remain associated with ␤-arrestins following their endocytosis (4,8). Internalized AT 1A R is targeted to enlarged hollow core vesicular structures, but is neither dephosphorylated nor efficiently recycled back to the cell surface (4). The inability of the vasopressin V2 receptor to recycle back to the cell surface appears to be regulated, at least in part, by the stable formation of a complex between ␤-arrestin and the G protein receptor kinasephosphorylated carboxyl-terminal tail of the receptor (9). This has led to the hypothesis that the formation of stable GPCR⅐␤arrestin complexes in endosomes may prevent the GPCR recycling and resensitization (10). However, exceptions do exist because other GPCRs, such as the neurokinin-1 receptor, internalize in a complex with ␤-arrestin and are still efficiently recycled back to the plasma membrane (11,12).
The internalization and trafficking of GPCRs between intracellular membrane compartments play a crucial role in regulating the overall balance of GPCR activity by governing whether GPCRs are recycled back to the cell surface or degraded in lysosomes. Although the sorting of GPCRs between distinct intracellular membrane organelles, including early, recycling, and late endosomes, may be regulated in part by receptor/␤-arrestin interactions, there is evidence that components of the endocytic machinery may directly influence the GPCR trafficking between these intracellular compartments. For example, AT 1A Rs preferentially traffic to Rab5-positive endosomal structures as a consequence of the agonist-dependent formation of AT 1A R⅐Rab5a protein complexes and AT 1A R-stimulated Rab5a GTP binding (13). However, the final intracellular destination of the AT 1A R internalized to the Rab5-positive early endosomal compartment remains less clear.
In this study, we examine 1) whether the association of Rab5a with the AT 1A R carboxyl-terminal tail prevents the lysosomal degradation and/or plasma membrane recycling of the AT 1A R and 2) whether the overexpression of Rab11 and Rab7 GTPases promotes AT 1A R recycling and trafficking to late endosomes and lysosomes. We found that Rab5a binding to the * This work was supported in part by Grant T-4987 from the Heart and Stroke Foundation of Ontario (to S. S. G. F.). 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  carboxyl-terminal tail protected the AT 1A R from lysosomal degradation, but did contribute to the regulation of AT 1A R recycling. Moreover, Rab7 and Rab11 overexpression increased AT 1A R targeting to lysosomes and recycling to the plasma membrane, suggesting that the relative level of Rab GTPase expression may influence the intracellular trafficking patterns and fate of GPCRs.

EXPERIMENTAL PROCEDURES
Materials-Human embryonic kidney (HEK) 293 cells were provided by American Type Culture Collection. The tissue culture medium was from Invitrogen. Bovine serum albumin was obtained from Bioshop Canada Inc. Mouse anti-hemagglutinin (HA) monoclonal antibody 12CA5 was purchased from Roche Applied Science. Horseradish peroxidase-conjugated anti-mouse IgG secondary antibody was from Amersham Biosciences. LysoTracker Red was purchased from Molecular Probes, Inc. Fluorescein isothiocyanate-conjugated goat anti-mouse secondary antibody, mouse anti-FLAG monoclonal antibody M2, angiotensin II (AngII), isoproterenol, and all other biochemical reagents were purchased from Sigma.
Cell Culture-HEK 293 cells were grown in Eagle's minimal essential medium with Earle's salt (Invitrogen) supplemented with 8% (v/v) heatinactivated fetal bovine serum (Hyclone Laboratories) and gentamycin (100 g/ml). The cells were seeded at a density of 2.5 ϫ 10 6 /100-mm dish (Falcon) and were transiently transfected by a modified calcium phosphate method (14,15) with the cDNAs described in the figure legends. Following transfection (ϳ18 h), the cells were pooled and reseeded on 35-mm glass-bottomed culture dishes (Mattek Corp.) for confocal live cell imaging studies, on 12-well dishes (Falcon) for receptor recycling studies, or in 96-well dishes (Falcon) for receptor degradation assays. Receptor expression was 500 -1000 fmol/mg of whole cell protein.
Receptor Recycling Assays-Cells expressing HA epitope-tagged receptors were treated with and without agonist for 30 min at 37°C, washed three times with serum-free medium, and either kept on ice or allowed to recover at 37°C for 60 min. The cells were antibody-stained, and the cell-surface receptor density was determined by flow cytometry as described previously (15). Receptor recycling is defined as the recovery of cell-surface receptors accessible to antibodies outside the cell following the removal of agonist compared with the cell-surface expression of receptors in matched controls that were not exposed to agonist.
Qualitative Receptor Degradation Assays-HEK 293 cells expressing FLAG epitope-tagged AT 1A R were pretreated with 10 g/ml cycloheximide in Hanks' balanced salt solution (HBSS) for 2 h at 37°C. Cells were incubated for an additional 4 h at 37°C in HBSS containing 10 g/ml cycloheximide along with 100 nM AngII for 0, 2, or 4 h. The cells were subsequently washed on ice with cold phosphate-buffered saline and solubilized with lysis buffer containing protease inhibitors (25 nM HEPES, pH 7.5, 300 nM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.1% Triton X-100, aprotinin, leupeptin, and phenylmethylsulfonyl fluoride). The protein content of the cell lysates was determined using a Bio-Rad D c protein assay kit. FLAG-AT 1A R was immunoprecipitated with rabbit anti-FLAG polyclonal antibody using protein G-Sepharose beads from cell lysates containing 500 g of protein. The immunoprecipitated proteins were subjected to SDS-PAGE, followed by electroblotting onto nitrocellulose membranes. The membranes were blocked with 10% milk in wash buffer (Tris-buffered saline/Tween 20) and then incubated with anti-FLAG monoclonal antibody diluted 1:1000 in wash buffer containing 3% milk. The membranes were rinsed and then incubated with horseradish peroxidase-conjugated donkey anti-mouse IgG secondary antibody diluted 1:2500 in wash buffer containing 3% milk. The membranes were rinsed with wash buffer, incubated with ECL Western blot detection reagents, and then exposed to film.
Quantitative Receptor Degradation Assays-Cell-surface FLAG epitope-tagged receptors were labeled for 1 h on ice with mouse anti-FLAG antibody diluted in HBSS. Cells were then washed and allowed to warm to 37°C prior to treatment with and without 100 nM AngII for 90 min at 37°C. Cells were subsequently fixed and permeabilized in 2% formaldehyde in HBSS and 0.1% saponin, washed, and incubated with horseradish peroxidase-conjugated anti-mouse IgG secondary antibody to measure to the total cellular complement of FLAG epitope-tagged receptors. The total cellular complement of the remaining internalized cell-surface receptors was determined following a 20-min incubation of solubilized cells with the horseradish peroxidase substrate tetramethylbenzidine (Sigma). The reaction was stopped with 1 N HCl, and the absorbance of the supernatant at 450 nm was measured using a microplate luminescence reader. Receptor degradation is defined as the loss of horseradish peroxidase activity following agonist stimulation. Experiments were done in quadruplicate for each condition and were repeated at least six times.
Rab7 Immunoblots-HA-Rab7 expression was confirmed by immunoblotting. In brief, 50 g of protein from each of the cell lysates used for immunoprecipitation were subjected to SDS-PAGE, transferred to nitrocellulose membrane, and subsequently immunoblotted for HA-Rab7 expression using anti-HA monoclonal antibody (1:1000 dilution).
Live Cell Imaging with Confocal Microscopy-Confocal microscopy was performed using a Zeiss LSM-510 META laser-scanning microscope with a Zeiss 100 ϫ 1.4 NA oil immersion lens. The Zeiss LSM-510 META system produces cross-talk-free images of fluorescent proteins (GFP and YFP) with closely overlapping emission spectra. This spectral analytical separation of GFP and YFP emission spectra functions to eliminate channel bleed-through. 2 HEK 293 cells expressing either FLAG-or HA-tagged AT 1A R and AT 1A R-(1-349) together with GFPand YFP-tagged Rab and ␤-arrestin constructs as described in the figure legends were plated on 35-mm glass-bottomed culture dishes, and live cell images were taken in the absence and presence of treatment with 100 nM AngII. Cells were loaded with 50 nM LysoTracker Red dye for 30 min at 37°C following agonist stimulation to assess the co-localization of receptors in lysosomes. Co-localization studies of ␤-arrestin-2-GFP and LysoTracker Red-labeled lysosomes were performed using dual excitation (488 and 543 nm) and emission (505-530 nm, GFP; and 590 -610 nm, LysoTracker Red) filter sets. The specificity of labeling and the absence of signal crossover were established by examination of single-labeled samples. Images showing scans of YFP-and GFP-expressing cells were unmixed using Zeiss LSM-510 META image processing software. 2 For live cell receptor labeling, anti-FLAG monoclonal antibody was conjugated to either Alexa Fluor 488 or Alexa Fluor 555 using a Zenon mouse IgG labeling kit (Molecular Probes, Inc.) following the manufacturer's directions. Cell-surface FLAG-AT 1A R was labeled for 10 min at room temperature with Alexa Fluor-tagged anti-FLAG antibody diluted 1:500 in HBSS; and subsequently, the cells were washed with HBSS to remove any unbound antibody.
Data Analysis-The means Ϯ S.E. are shown for values obtained for the number of independent experiments indicated in the figure legends. Data were analyzed for statistical significance using GraphPAD Prism software. Statistical significance was determined by an unpaired twotailed t test.

RESULTS
The AT 1A R Is Not Targeted to Lysosomes following Prolonged Agonist Treatment-Previous studies have demonstrated that the AT 1A R does not efficiently recycle to the cell surface and is internalized as a complex with ␤-arrestin to enlarged Rab5positive endocytic structures (4,8,13). Therefore, we sought to determine whether the AT 1A R is either retained in early endosomes or is eventually targeted to lysosomes for degradation. To examine AT 1A R targeting to lysosomes in live HEK 293 cells, we utilized the acidotropic lysosomal compartment probe LysoTracker Red (16) in combination with ␤-arrestin-2-GFP to follow receptor trafficking (4,8,13). In the absence of agonist treatment, ␤-arrestin-2-GFP was diffusely localized throughout the cytoplasm, whereas LysoTracker Red stained punctu-ate vesicular structures within the cytosol of the cell (Fig. 1A). In response to a 3-h treatment of the same HEK 293 cell with 100 nM AngII, AT 1A R⅐␤-arrestin-2-GFP complexes internalized to large hollow core endosomal structures (Fig. 1B). Consistent with a previous study (13), AT 1A R⅐␤-arrestin-2-YFP complexes extensively co-localized with GFP-Rab5a (Fig. 1C). However, even following 3 h of agonist stimulation, the AT 1A R⅐␤-arrestin complexes rarely co-localized with the lysosomal marker Lyso-Tracker Red (Fig. 1B). Taken together, these observations indicate that internalized AT 1A R is retained in a Rab5-positive early endosomal compartment and does not target to lysosomes for degradation.
Rab5 Binding to the AT 1A R Carboxyl-terminal Tail Prevents AT 1A R Degradation-Truncation of the AT 1A R carboxyl-terminal tail to create the AT 1A R-(1-349) mutant lacking the distal 10 amino acid residues prevents Rab5 binding, but does not alter the stability of AT 1A R⅐␤-arrestin complexes (13). Furthermore, the AT 1A R-(1-349) mutant only partially co-localizes with Rab5a in endocytic vesicles (13). Therefore, we tested the possibility that lost AT 1A R-(1-349) binding to Rab5 may result in the lysosomal targeting of the mutant receptor. Following a 90-min exposure to 100 nM AngII, we observed substantial co-localization of AT 1A R⅐␤-arrestin-2-GFP with LysoTracker Red, whereas we did not observe co-localization between GFP-Rab5a and LysoTracker Red (Fig. 2, A and B). Agonist treatment of AT 1A R-expressing HEK 293 cells for 2 and 4 h did not promote the degradation of the AT 1A R (Fig. 3A). Moreover, only 14 Ϯ 7% of the wild-type AT 1A R was degraded following a 90-min exposure to 100 nM AngII, whereas 33 Ϯ 8% of the total cellular complement of AT 1A R-(1-349) was lost following a 90-min exposure to agonist (Fig. 3, B and C). Thus, the truncation of the AT 1A R tail, which impairs Rab5 interactions, also results in the targeting of the AT 1A R to lysosomes for degradation.
Rab7-dependent Regulation of AT 1A R Targeting to and Degradation in Lysosomes-Because Rab7 may regulate intracellular trafficking of vesicular cargo to late endosomes and lysosomes (16), we tested whether Rab7 overexpression might increase AT 1A R degradation. Following Rab7 overexpression, agonist treatment of AT 1A R-expressing HEK 293 cells for 2 and 4 h promoted significant AT 1A R degradation (Fig. 3A). We found that the overexpression of either wild-type Rab7 or Rab7-Q67L increased wild-type AT 1A R degradation to 29 Ϯ 5 and 37 Ϯ 7%, respectively (Fig. 3B). However, increased AT 1A R degradation in response to 100 nM AngII treatment for 90 min was not observed following the overexpression of the dominantnegative Rab7-N125I mutant (Fig. 3B). In contrast, the extent of AT 1A R-(1-349) mutant degradation following 90 min of agonist treatment was unaffected by Rab7 and Rab7-Q67L expression, but dominant-negative Rab7-N125I reduced AT 1A R-(1-349) degradation to 11 Ϯ 7% of the total cellular complement of the receptor (Fig. 3C). The overexpression of either Rab7 or Rab7-Q67L resulted in the co-localization of "putative" wild-type AT 1A R⅐␤-arrestin complexes with the lysosomal marker dye LysoTracker Red (Fig. 4, A and B). Although the overexpression of Rab7-N125I had no effect on the localization of the wild-type AT 1A R⅐␤-arrestin complexes, it prevented the redistribution of the AT 1A R-(1-349) mutant⅐␤-arrestin complexes to LysoTracker Red-positive lysosomes (Fig. 4, C and D). We have demonstrated that the AT 1A R remains associated with ␤-arrestin following the internalization of the receptor (8), but this does not rule out the possibility that only ␤-arrestin is targeted to lysosomes following Rab7 expression. Furthermore, Bhatnagar et al. (17) have reported previously that ␤-arrestin dissociates from the 5-hydroxytryptamine type 2c receptor in the endosomal compartment. Therefore, we tested whether the AT 1A R may be targeted to lysosomes following the prolonged treatment of cells with agonist. In the absence of agonist, AT 1A R labeled with Alexa Fluor 488-conjugated anti-FLAG antibody was localized to the cell surface (Fig. 5A). Agonist treatment for 2 h resulted in the internalization of anti-FLAG antibody-labeled AT 1A R to large hollow core vesicles, but the receptor did not co-localize with LysoTracker Red-positive lysosomes (Fig. 5B). However, identical to what was observed for AT 1A R⅐␤-arrestin complexes, the expression of both wild-type Rab7 and Rab7-Q79L resulted in the co-localization of the wild-type AT 1A R with LysoTracker Red (Fig. 5, C and D). Thus, it is unlikely that the receptor and ␤-arrestin traffic to distinct compartments following Rab7 overexpression. Taken together, the data suggest that, although the AT 1A R is normally retained within a Rab5-positive early endosomal compartment, increased Rab7 activity promotes the transition of the AT 1A R from the early endosomal compartment to lysosomes.
Role of Rab7 and Rab5 in Regulating the Targeting of the AT 1A R to Late Endosomes-To determine how Rab7 overexpression might facilitate the lysosomal targeting of the AT 1A R in HEK 293 cells, we examined whether the subcellular distribution of Rab7 overlaps with either LysoTracker Red in lysosomes or Rab5 in early endosomes. We found that, although YFP-Rab7 exhibited extensive co-localization with Lyso-Tracker Red-positive lysosomes (Fig. 6A, yellow), YFP-Rab7 also labeled an endosomal compartment that was not Lyso-Tracker Red-positive (green). In addition, we observed Lyso-Tracker Red-positive lysosomes that did not contain YFP-Rab7 (Fig. 6A, red). Although GFP-Rab5a and YFP-Rab7 primarily labeled distinct membrane compartments, there appears to be some overlap in GFP-Rab5a and YFP-Rab7 in HEK 293 cells (Fig. 6B, arrows). Consistent with a role for Rab7 in regulating vesicular trafficking from early to late endosomes, internalized AT 1A R⅐␤-arrestin-2-GFP complexes became extensively co-localized with overexpressed YFP-Rab7 (Fig. 7A, arrows). Similarly, Alexa Fluor 555-conjugated anti-FLAG antibody-labeled AT 1A R was also targeted to YFP-Rab7-positive vesicles (Fig.  7B). The overexpression of the dominant-negative Rab5a-S34N mutant prevented both the co-localization of AT 1A R⅐␤-arrestin-2-GFP complexes with YFP-Rab7 and the targeting of the AT 1A R-(1-349) mutant to lysosomes (Fig. 7, C and D). These results support previous observations that both Rab5 and Rab7 regulate the trafficking of cargo proteins from early endosomes to late endosomes and lysosomes (16,18). Thus, we propose that increased Rab7 activity promotes the exit of AT 1A Rs from early endosomes, ultimately allowing AT 1A R degradation in lysosomes.
AT 1A R Plasma Membrane Recycling-The AT 1A R is not efficiently recycled back to the cell surface following agonist removal (4). Therefore, we investigated the possibility that loss of Rab5 binding to the AT 1A R carboxyl-terminal tail might pro-mote AT 1A R recycling. However, we found that neither the AT 1A R-(1-349) mutant nor the wild-type AT 1A R was effectively recycled (16 Ϯ 5 and 20 Ϯ 5% cell-surface recovery, respectively), whereas 77 Ϯ 7% of the ␤ 2 -adrenergic receptor (␤ 2 AR) was efficiently recycled back to the cell surface 1 h following agonist removal (Fig. 8, A-C). In contrast, the AT 1A R-AALAA mutant (10), which lacks serine and threonine residues required for stable ␤-arrestin binding, recycled more efficiently than the wild-type AT 1A R, but less efficiently than the ␤ 2 AR (Fig. 8D). Thus, the plasma membrane recycling of the AT 1A R is prevented, at least in part, by the formation of stable recep-tor⅐␤-arrestin complexes in endosomes.
Rab11-dependent Regulation of AT 1A R Plasma Membrane Recycling-Because Rab7 overexpression influenced the targeting of the AT 1A R to lysosomes, we sought to determine whether wild-type, constitutively active (Q70L), and dominantnegative (S25N) Rab11 protein overexpression might alter the plasma membrane recycling of the ␤ 2 AR and AT 1A R. We found that the plasma membrane recycling of the ␤ 2 AR was unaffected by either wild-type Rab11 or Rab11-Q70L overexpression, whereas Rab11-S25N reduced ␤ 2 AR recycling by 24 Ϯ 4% (Fig. 8A). However, the overexpression of both the wild-type Rab11 and Rab11-Q70L proteins significantly increased AT 1A R recycling to 61 Ϯ 9 and 61 Ϯ 10%, respectively (Fig. 8B). However, only Rab11-Q70L expression increased the plasma membrane recycling of the AT 1A R-(1-349) and AT 1A R-AALAA mutants (Fig. 8, C and D). Rab11-S25N did not alter the extent of membrane recycling for any of the AT 1A R constructs (Fig. 8,  B-D). Thus, similar to what we observed for AT 1A R degradation following Rab7 overexpression, Rab11 overexpression facilitated AT 1A R recycling.
Effect of Rab11 on the Subcellular Distribution of AT 1A R⅐␤-Arrestin Complexes-Consistent with previous observations that Rab5 and Rab11 are each localized to overlapping, yet distinct early endosomal domains (19), we observed the punctuate co-localization of GFP-Rab11 at the rim of endosomes containing AT 1A R⅐␤-arrestin-2-YFP complexes (Fig. 9A, arrows). In contrast, we found that GFP-Rab11-Q70L was extensively co-localized with AT 1A R⅐␤-arrestin-2-YFP complexes in enlarged endosomes (Fig. 9B, arrows). Because Rab7-Q70L overexpression increased the plasma membrane recycling of the AT 1A R, AT 1A R-(1-349), and AT 1A R-AALAA, we examined the effect of Rab11-Q70L overexpression on the co-localization of internalized ␤-arrestin-2-YFP and Alexa Fluor 555-conjugated anti-FLAG antibody-labeled AT 1A R with GFP-Rab5a. To our surprise, we found that Rab11-Q70L overexpression resulted in the loss of both ␤-arrestin-2-YFP and anti-FLAG antibody-labeled AT 1A R co-localization with GFP-Rab5a (Fig.  9, C and D, arrows). Taken together, our observations suggest that the overexpression of both Rab7 and Rab11 GTPases not only results in the redistribution of the AT 1A R to late and recycling endosomes, but also appears to dynamically regulate AT 1A R/Rab5 interactions. DISCUSSION In this study, we have investigated whether the association of Rab5 with the AT 1A R carboxyl-terminal tail results in the retention of the receptor in early endosomes, thereby preventing AT 1A R recycling and degradation. We found that the wildtype AT 1A R endocytosed to Rab5-positive endosomes was neither targeted to lysosomes nor recycled back to the cell surface, whereas the AT 1A R-(1-349) mutant, which did not bind Rab5, was targeted to lysosomes for degradation. However, the loss of Rab5 binding to the AT 1A R carboxyl-terminal tail did not reestablish AT 1A R recycling. Rather, a loss of stable ␤-arrestin binding to the AT 1A R, at least in part, allowed the AT 1A R to recycle. Thus, Rab5 and ␤-arrestin binding to the AT 1A R carboxyl-terminal tail appear to function together to mediate the retention of the AT 1A R in early endosomes. We also observed that the transition of the wild-type AT 1A R from early endosomes to late and recycling endosomal compartments was facilitated by the overexpression of Rab7 and Rab11 proteins, respectively. The targeting of the AT 1A R to late endosomes and lysosomes was blocked by the expression of the dominantnegative Rab5a-S34N mutant. Taken together, our data suggest that Rab5, Rab7, and Rab11 have the potential to work in concert with one another to regulate the intracellular trafficking patterns of the AT 1A R.
The transport of proteins between distinct intracellular organelles is a highly regulated process that involves multiple vesicular membrane budding and fusion events between donor and acceptor membranes. Rab GTPases are key players that regulate vesicular trafficking between these endosomal compartments (20 -22). This study focused on the specific role of three Rab GTPases, Rab5, Rab7, and Rab11, in regulating the intracellular trafficking of the AT 1A R. We have found that endocytosed AT 1A R⅐␤-arrestin complexes are normally retained in enlarged endosomal structures (4,13), but that the overexpression of Rab7 and Rab11 GTPases promotes the redistribution of these putative complexes to late and recycling endosomal membrane compartments. The retention of the AT 1A R in early endosomes is likely the result of AT 1A R/Rab5 interactions, leading to increased Rab5 activity, and increased Rab7 and Rab11 activity may be sufficient to alter the trafficking of the receptor between membrane compartments. Thus, despite the fact that the AT 1A R may regulate Rab5 activity and early endosomal fusion (13), Rab7 and Rab11 overexpression may overcome this activity to substantially increase AT 1A R transit to late and recycling endosomal compartments. This conclusion is consistent with previous observations that the expression of the constitutively active Rab5-Q79L mutant increases the rate of transferrin uptake and results in the accumulation of the ␤ 2 AR in enlarged Rab5-positive endosomes (15,23). Consequently, despite the fact that Rab5, Rab7, and Rab11 GTPases are ubiquitously expressed, cell type and tissue differences in Rab GTPase protein expression levels may have profound effects on AT 1A R trafficking between intracellular membrane compartments and the plasma membrane surface. Thus, the AT 1A R may exhibit diverse desensitization and resensitization profiles in different cell types.
Many GPCRs, including the ␤ 2 AR, ␦-opioid receptor, -opioid receptor, endothelin B receptor, protease-activated receptor-1, CXC chemokine receptor-2, and CXC chemokine receptor-4, are degraded in response to prolonged agonist stimulation (27)(28)(29). Consequently, the mechanisms contributing to the targeting of GPCRs for degradation in both proteasomes and lysosomes have become the subject of intensive investigation. The degradation of the ␤ 2 AR and -opioid receptor is reported to be mediated by both lysosomes and proteasomes (24,26,30), whereas ubiquitin-dependent lysosomal degradation has been reported for CXC chemokine receptor-4 (28). ␤-Arrestins have also been implicated in the ubiquitin-mediated down-regulation of the ␤ 2 AR via association with Mdm2 (31). However, Mdm2 does not appear to be the same ubiquitin ligase that mediates ␤ 2 AR ubiquitination (31). The targeting of proteaseactivated receptor-1 for lysosomal degradation is mediated by the association of sorting nexin-1 with the carboxyl-terminal tail of the receptor (32). Due to the localization of Rab7 to late endosomes and lysosomes, Rab7 is proposed to regulate vesicular trafficking from early endosomes to late endosomes and from late endosomes to lysosomes (16,18). Thus, it is likely that, no matter which mechanism and/or molecular intermediate is required for the targeting of GPCRs to lysosomes, the receptors may be mobilized to lysosomes via Rab7-positive late endosomes. Consistent with this hypothesis, identical to what we observed for the AT 1A R-(1-349) mutant, the overexpression of dominant-negative Rab7 mutants prevents the down-regulation of the -opioid receptor, CXC chemokine receptor-2, low density lipoprotein receptor, and epidermal growth factor receptor (16, 23-26, 29, 33). However, unlike other GPCRs that are normally targeted to lysosomes for degradation, the AT 1A R does not become localized in late endosomes and lysosomes unless Rab7 is overexpressed. Moreover, the movement of the AT 1A R from the early to late endosomal compartment is dependent upon both Rab5 and Rab7 activity, suggesting that the trafficking of proteins between early and late endosomes requires the coordination of Rab5 and Rab7 function.
Following their internalization, GPCRs may return to the cell surface by at least two distinct recycling pathways, either "rapidly" from sorting endosomes or "slowly" from recycling endosomes (34). Rab4 governs the rapid cell-surface recycling of proteins from early endosomes, whereas Rab11 controls the slow recycling route of proteins and nutrients from recycling endosomes back to the plasma membrane (34). Several studies have examined the role of Rab4 and Rab11 in regulating the plasma membrane recycling of GPCRs. For example, ␤ 2 AR recycling is blocked by the expression of dominant-negative Rab4 and Rab11 mutant proteins (Fig. 7) (15, 23). The recycling of many other GPCRs, including the somatostatin-3 receptor, vasopressin V2 receptor, neurokinin-1 receptor, chemokine CXC receptor-2, and m4 muscarinic acetylcholine receptor, is also regulated by Rab4 and Rab11 (29,(35)(36)(37)(38)(39). In HEK 293 cells, we observed discrete co-localization of the putative AT 1A R⅐␤-arrestin complexes with GFP-Rab11 and substantial co-localization of the putative AT 1A R⅐␤-arrestin complexes with GFP-Rab11-Q70L in enlarged endosomal structures (Fig. 9). The overexpression of either Rab11 or Rab11-Q70L increased the plasma membrane recycling of the AT 1A R. This observation is consistent with a previous report suggesting that the AT 1A R may recycle through the Rab11-mediated slow recycling route (40). However, only Rab11-Q70L stimulated the recycling of the AT 1A R-(1-349) mutant. It is not clear why wild-type Rab11 regulates only wild-type AT 1A R recycling; it is possible that, similar to what we have observed for Rab5 (13), the activity of overexpressed Rab7 may also be regulated by the AT 1A R and that ␤-arrestin binding may be required for this activity. This possibility warrants future investigation.
Recycling of the AT 1A R is increased following loss of ␤-arrestin binding, but not Rab5 interactions, supporting the idea that the formation of stable receptor⅐␤-arrestin complexes may retard receptor recycling (10). However, the neurokinin-1 receptor internalizes bound to ␤-arrestin, but is efficiently recycled back to the cell surface (36). The overexpression of Rab11-Q70L not only promotes the redistribution of putative AT 1A R⅐␤-ar-restin complexes out of the Rab5-positive endosomes, but increases the recycling of wild-type and mutant AT 1A Rs. Thus, it is possible that Rab protein expression may differentially regulate the trafficking of different GPCRs through distinct Rabregulated endosomal compartments. The effect of increased Rab11 activity may be to overcome the ␤-arrestin-dependent inhibition of GPCR recycling.
In conclusion, it is now clear that the targeting of GPCRs between intracellular membrane compartments involves a complex series of protein/protein interactions, such as the binding of sorting nexin-1 to protease-activated receptor-1 and Rab5 to the AT 1A R (13,32). In addition, the targeting of proteins between intracellular compartments requires the coordinated regulation of vesicular trafficking by Rab GTPases. In this study, we have demonstrated that, by virtue of their ability either to bind to the AT 1A R or to regulate vesicular trafficking to late and recycling endosomes, Rab GTPases interactively coordinate the intracellular trafficking fate of the AT 1A R. Because Ͼ60 Rab GTPases may exist (21), it is likely that other Rab GTPases may regulate the trafficking of GPCRs between additional membrane compartments and/or that it will be possible to identify increasingly discrete membrane compartments that regulate the activation, inactivation, and reactivation of GPCRs. Future challenges will be to further characterize the precise roles of other Rab GTPases in the intracellular trafficking of the AT 1A R and other GPCRs. showing the co-localization of GFP-Rab11 (green) with AT 1A R⅐␤-arrestin-2-YFP complexes (red) following treatment of live HEK 293 cells with 100 nM AngII for 60 min. B, representative laser-scanning confocal micrographs showing the co-localization of GFP-Rab11-Q70L (green) with AT 1A R⅐␤-arrestin-2-YFP complexes (red) following treatment of live HEK 293 cells with 100 nM AngII for 60 min. C, representative laser-scanning confocal micrographs showing the effect of Rab11-Q70L on the co-localization of GFP-Rab5a (green) with AT 1A R⅐␤-arrestin-2-YFP complexes (red) following treatment of live HEK 293 cells with 100 nM AngII for 60 min. D, representative laser-scanning confocal micrographs showing the effect of Rab11-Q70L on the co-localization of GFP-Rab5a (green) with Alexa Fluor 555-conjugated anti-FLAG monoclonal antibody-labeled FLAG-AT 1A R (red) following treatment of live HEK 293 cells with 100 nM AngII for 60 min. Yellow indicates co-localization. Data are representative images of multiple cells from three to five independent experiments. HEK 293 cells were transiently transfected with plasmid cDNAs encoding FLAG-AT 1A R (10 g), ␤-arrestin-2-YFP (5 g), GFP-Rab11 (3.5 g), GFP-Rab11-Q70L (3.5 g), GFP-Rab5a (5 g), and HA-Rab11-Q70L (5 g). Bars ϭ 10 m.