Rab8 Interacts with Distinct Motifs in α2B- and β2-Adrenergic Receptors and Differentially Modulates Their Transport*

The molecular mechanism underlying the post-Golgi transport of G protein-coupled receptors (GPCRs) remains poorly understood. Here we determine the role of Rab8 GTPase, which modulates vesicular protein transport between the trans-Golgi network (TGN) and the plasma membrane, in the cell surface targeting of α2B- and β2-adrenergic receptors (AR). Transient expression of GDP- and GTP-bound Rab8 mutants and short hairpin RNA-mediated knockdown of Rab8 more potently inhibited the cell surface expression of α2B-AR than β2-AR. The GDP-bound Rab8(T22N) mutant attenuated ERK1/2 activation by α2B-AR, but not β2-AR, and arrested α2B-AR in the TGN compartment. Co-immunoprecipitation revealed that both α2B-AR and β2-AR physically interacted with Rab8 and glutathione S-transferase fusion protein pulldown assays demonstrated that Rab8 interacted with the C termini of both receptors. Interestingly, mutation of the highly conserved membrane-proximal C terminus dileucine motif selectively blocked β2-AR interaction with Rab8, whereas mutation of residues Val431-Phe432-Asn433-Gln434, Pro447-Trp448, Gln450-Thr451, and Trp453 in the C terminus impaired α2B-AR interaction with Rab8. Furthermore, transport inhibition by Rab8(T22N) of a chimeric β2-AR carrying the α2B-AR C terminus was similar to α2B-AR. These data provide strong evidence indicating that Rab8 GTPase interacts with distinct motifs in the C termini of α2B-AR and β2-AR and differentially modulates their traffic from the TGN to the cell surface.

G protein-coupled receptors (GPCRs) 3 are the largest superfamily of cell surface receptors that regulate a variety of cell functions by response to a myriad of stimuli. The magnitude of ligand-elicited cellular response is at least in part dictated by the level of receptor expressed at the plasma membrane available for binding to the ligand and subsequently activating downstream effectors (1)(2)(3). The cell surface expression of GPCRs at a given time is determined by a number of elaborately regulated trafficking processes, including the export of newly synthesized receptors from the endoplasmic reticulum (ER) through the Golgi to the cell surface, the endocytosis of the cell surface receptors to endosomes upon agonist stimulation, the recycling of the internalized receptor from endosomes to the cell surface, and the transport to lysosome for degradation (4). However, compared with the extensive studies performed on the events of endocytosis, recycling, and degradation (1,(5)(6)(7)(8)(9)(10), the molecular mechanisms governing GPCR cell surface transport and its role in regulating receptor function are relatively less well understood.
A number of studies have demonstrated that GPCR export from the ER, the first step in intracellular trafficking of GPCRs, is modulated by specific, conserved motifs located in the membrane-proximal C termini, multiple regulatory proteins that interact with GPCRs, and constitutive dimerization in the ER (4,(11)(12)(13)(14)(15). Several studies have also demonstrated that post-Golgi transport of GPCRs is a regulated process. For example, G protein-coupled olfactory and chemokine receptors are released from the ER, but accumulated in the Golgi (16,17). The opsin mutant E150K is also accumulated in the cis and medial Golgi (18). We have recently identified the YS motif in the N termini, which regulates the exit of ␣ 2 -adrenergic receptors (␣ 2 -ARs) from the Golgi (19).
Rab GTPases are the largest branch of the Ras-related GTPase superfamily, consisting of more than 60 members in mammals and 11 in yeast, and are involved in almost every step of vesicle-mediated protein transport. Each Rab GTPase has a distinct subcellular localization pattern that correlates with the compartments between which it coordinates transport (20). The function of Rab GTPases in coordinating vesicular transport is mediated through their GTP/GDP exchange cycle, which superimposes with Rab protein association with, and dissociation from, subcellular organelle membranes. The inactive, GDP-bound conformation of Rab GTPases is maintained in the cytosol through an association with GDP dissociation inhibitors, which function as chaperones and mediate Rab translocation from cytosol to membrane. Membrane-associated GDPbound Rab-GDP dissociation inhibitor complexes undergo GDP exchange for GTP. GTP-bound Rab GTPases recruit their effectors, which mediate the migration, docking, and fusion of transport vesicles to acceptor membrane. GTP-bound Rab is then hydrolyzed to GDP-bound Rab for recycling back to donor membrane (20,21). Therefore, GDP-bound Rab mutants, which are unable to undergo exchange for GTP, and GTP-bound Rab mutants, which prevent the recycling of Rab to be reused, function as negative regulators in protein transport.
Most studies on the roles of Rab GTPases in the intracellular trafficking of GPCRs have been focused on the events involved in the internalization, degradation, and transport between the ER and the Golgi of the receptors (22)(23)(24). In contrast, much less is known about the involvement of Rab GTPases in GPCR export from the Golgi to the plasma membrane. Our previous studies have demonstrated the roles of several Rab GTPases in regulating GPCR transport along the early secretory pathway (25)(26)(27)(28)(29). In this article, we expand these studies to determine the role of Rab8 GTPase in the cell surface targeting of ␣ 2B -AR and ␤ 2 -AR and the possible mechanism involved. Rab8 GTPase has been demonstrated to regulate protein transport from the trans-Golgi network (TGN) to the apical/basolateral plasma membrane (30 -37). Our data demonstrate for the first time that Rab8 differentially modulates the transport of two adrenergic receptors, which is likely determined by direct interactions of Rab8 with distinct motifs in the C termini of the receptors.
For generation of the construct ␤ 2 -AR-ct in which the C-terminal 86 amino acid residues (Arg 328 -Leu 413 ) was truncated from ␤ 2 -AR, the full-length ␤ 2 -AR was amplified by PCR (forward primer, 5Ј-GATCAAGCTTATGGGGCAACCCGGGA-ACGGCAGC-3Ј and reverse primer, 5Ј-GATCGTCGACCA-GCAGTAGATAAGGGGATTG-3Ј) in which the ␤ 2 -AR-ct was in-frame with GFP, restricted with HindIII and SalI, and ligated into the pEGFP-N1 vector. For generation of chimeric receptor ␤ 2 -AR␣ 2B -ct in which the C terminus of ␤ 2 -AR was substituted with that of ␣ 2B -AR, two complementary oligonucleotides coding the ␣ 2B -AR C terminus and carrying the sticky ends of SalI and BamHI were annealed and ligated into the ␤ 2 -AR-ct-GFP in the pEGFP-N1 vector, which was cleaved with SalI and BamHI.
Rab8 and its GDP-bound mutant T22N tagged with the FLAG epitope at their N termini were kindly provided by Dr. Terrance E. Hebert (Department of Pharmacology, McGill University, Canada). Rab8 was also conjugated with GFP at its N terminus as described previously for Rab2 (29). The ␣ 2B -AR and ␤ 2 -AR gene segments encoding the C termini of the receptors were amplified by PCR and subcloned into the BamHI and XhoI restriction sites of pGEX-4T-1 as described previously (7). Receptor mutants were generated using QuikChange site-directed mutagenesis (Agilent Technologies, La Jolla, CA). The structure of each construct used in the present study was verified by restriction mapping and nucleotide sequence analysis.
Primary neuronal cultures were prepared from the thalamus of embryonic day 18 rat pups and grown on glass coverslips precoated with poly-L-lysine. After 10 days in vitro, GFP-Rab8 constructs were transfected into the neurons with the calcium phosphate precipitation method as described previously (42). Two days after transfection, the neurons were stained with antibodies recognizing ␣ 2B -AR at a dilution of 1:50 for 1 h at room temperature followed by incubation with Alexa 568-conjugated secondary antibodies for 1 h. Images were acquired using a confocal microscope (Zeiss 510 Meta) as described (43). Data were analyzed by NIH Image J software. Specifically, the ␣ 2B -AR images were thresholded to remove background. The amounts of the ␣ 2B -AR signal pixels in the dendrites (processes 3 m away from cell body) were calculated and normalized over the same dendrite area containing the Rab8-GFP signal.
Radioligand Binding-Cell surface expression of ␣ 2B -AR and ␤ 2 -AR was measured by ligand binding of intact live cells using [ 3 H]RX821002 and [ 3 H]CGP12177, respectively, as described previously (19,29). Briefly, cells were cultured in 6-well dishes and transfected with 0.3 g of the ␣ 2B -AR or ␤ 2 -AR plasmid together with 1.8 g of Rab8 or its mutant or empty vectors. After 6 h the cells were split into 12-well plates at a density of 4 ϫ 10 5 cells/well and cultured for an additional 24 h. For measurement of ␣ 2B -AR expression at the cell surface, the cells were incubated with DMEM plus [ 3 H]RX821002 at a concentration of 20 nM for 90 min at room temperature. The nonspecific binding was determined in the presence of rauwolscine (10 M). shRNA-mediated Knockdown of Endogenous Rab8-Human Rab8-specific and scrambled short hairpin RNAs (shRNA) in the psiSTRIKE hMGFP vector were generously provided by Dr. Johan Peranen (Institute of Biotechnology, University of Helsinki, Finland) as described (44). HEK293 and MCF-7 cells were cultured on 6-well dishes and transfected with 2 g of control or Rab8 shRNA with or without co-transfection with HA-tagged receptors. After transfection for 72 h, cell surface expression of the receptors was measured by intact cell ligand binding.
Measurement of ERK1/2 Activation-HEK293 cells were cultured in 6-well dishes and transfected with 0.5 g of ␣ 2B -AR or ␤ 2 -AR with or without co-transfection with 1.5 g of Rab8(T22N). At 6 -8 h after transfection, the cells were split into 6-well dishes and cultured for an additional 36 h. The cells were starved for at least 3 h and then stimulated with 1 M UK14304 or ISO for 5 min. For measurement of ERK1/2 activation by endogenous ␣ 2B -AR in NG108 cells, the cells were cultured in 6-well plates and transfected with 2 g of Rab8(T22N) with FuGENE 6 reagent in serum-free medium for 24 h. The cells were then cultured in fresh serum-free medium for 3 h and stimulated with 1 M UK14304 for 15 min. Stimulation was terminated by addition of 1ϫ SDS gel loading buffer. After solubilizing the cells, 20 l of total cell lysates was separated by 12% SDS-PAGE. ERK1/2 activation was determined by measuring the levels of phosphorylation of ERK1/2 with phospho-specific ERK1/2 antibodies by immunoblotting (25).
Measurement of cAMP Production-cAMP production in response to stimulation with ISO was measured by using the cAMP enzyme immunoassay system (Biotrak, Amersham Biosciences, Piscataway, NJ) as described previously (45). HEK293 cells were cultured in 100-mm dishes and transfected with 3 g of ␤ 2 -AR. After 6 h, the cells were split into 12-well plates and cultured for 12 h. The cells were then starved for 24 h before stimulation with ISO at a concentration of 10 M for 10 min. The reactions were stopped by aspirating the medium and then the cells were lysed using 200 l of dodecyltrimethylammonium (2.5%). One hundred l of cell lysate was transferred into microtiter plates and incubated with anti-cAMP antiserum, followed by incubation with cAMP peroxidase. After washing and addition of substrate, peroxidase activity was measured by spectrometry. cAMP concentrations were calculated based on the competition of cAMP in samples with a fixed quantity of peroxidase-labeled cAMP.
Immunofluorescence Microscopy-For fluorescence microscopic analysis of receptor subcellular localization, HEK293 cells were grown on coverslips pre-coated with poly-L-lysine in 6-well plates and transfected with 50 ng of GFP-tagged receptors with or without co-transfection together with 400 ng of Rab8(T22N). The cells were fixed with 4% paraformaldehyde, 4% sucrose mixture in PBS for 15 min. For co-localization studies involving immunostaining, HEK293 cells were permeabilized with PBS containing 0.2% Triton X-100 for 5 min, and blocked with 5% normal donkey serum for 1 h. The cells were then incubated with p230 antibodies for 1 h. After washing with PBS (3 ϫ 5 min), the cells were incubated with Alexa Fluor 594-labeled secondary antibody (1:2000 dilution) for 1 h at room temperature. The coverslips were mounted, and fluorescence was detected with a Leica DMRA2 epifluorescent microscope as described previously (25). Images were deconvolved using SlideBook software and the nearest neighbors deconvolution algorithm (Intelligent Imaging Innovations, Denver, CO).

Co-immunoprecipitation of Receptors and Rab8
GTPase-HEK293 cells cultured on 100-mm dishes were transfected with 2 g of HA-tagged receptor together with 2 g of the pEGFP-C1 vector or GFP-tagged Rab8 GTPase in the pEGFP-C1 vector for 28 h. The cells were washed twice with PBS, harvested, and lysed with 500 l of lysis buffer containing 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS and Complete Mini protease inhibitor mixture. After gentle rotation for 1 h, samples were centrifuged for 15 min at 14,000 ϫ g and the supernatant was incubated with 50 l of protein G-Sepharose for 1 h at 4°C to remove nonspecific bound proteins. Samples were then incubated with 3 g of anti-HA antibodies overnight at 4°C with gentle rotation followed by incubation with 50 l of protein G-Sepharose 4B beads for 5 h. Resin was collected by centrifugation and washed three times each with 500 l of lysis buffer without SDS. Immunoprecipitated receptors were eluted with 100 l of 1ϫ SDS gel loading buffer. Thirty l from each sample was then separated by SDS-PAGE to probe for GFP-Rab8 in the immunoprecipitates by immunoblotting using GFP antibodies. In parallel, each sample was further diluted 5 times with 1ϫ SDS gel loading buffer, separated by SDS-PAGE, and probed with anti-HA antibodies to determine the amount of the receptor in the immunoprecipitates (19).
Protein Interaction Assay-The GST fusion proteins were expressed in bacteria and purified using a glutathione affinity matrix as described previously (5, 7). Immobilized fusion pro-teins were either used immediately or stored at 4°C for no longer than 3 days. Each batch of fusion protein used in experiments was first analyzed by Coomassie Blue staining following SDS-PAGE. Total HEK293 cell homogenates were incubated with ϳ5 g of GST fusion proteins tethered to the glutathione resin in 250 l of binding buffer containing 20 mM Tris-HCl, pH 7.5, 70 mM NaCl, and 2% Nonidet P-40 for 2.5 h at 4°C. The resin was washed three times with 0.5 ml of binding buffer, and the retained proteins were solubilized in SDS gel loading buffer and separated by SDS-PAGE. Rab8 GTPase bound to the GST fusion proteins was detected by immunoblotting using anti-Rab8 antibodies.
Statistical Analysis-Differences were evaluated using Student's t test, and p Ͻ 0.05 was considered as statistically significant. Data are expressed as the mean Ϯ S.E.

Differential Regulation of the Cell Surface Expression of ␣ 2B -AR and ␤ 2 -AR by Rab8
GTPase-We first determined the effect of manipulating the function of Rab8 by transient expression of wild-type Rab8 and its GDP-bound Rab8(T22N) and GTP-bound Rab8(Q67L) mutants on cell surface expression of ␣ 2B -AR and ␤ 2 -AR. Both receptors were tagged with either GFP at their C termini or three HA at their N termini. Rab8 was tagged with FLAG at its N terminus and expressed together with individual receptors in HEK293 cells. Rab8 expression was measured by immunoblotting using anti-FLAG and anti-Rab8 antibodies (Fig. 1A) and the cell surface expression of ␣ 2B -AR and ␤ 2 -AR was quantified by ligand binding in intact live cells using the membrane-impermeable ligands [ 3 H]RX821002 and [ 3 H]CGP12177, respectively. Transient expression of wild-type Rab8 did not significantly alter the cell surface numbers of either receptor (Fig. 1B). In contrast, expression of Rab8(T22N) and Rab8(Q67L) inhibited the cell surface expression of ␣ 2B -AR and ␤ 2 -AR, but at different magnitudes. Rab8(T22N) and Rab8(Q67L) inhibited the cell surface transport of ␣ 2B -AR by 45%, whereas the cell surface expression of ␤ 2 -AR was moderately attenuated by 20% (Fig. 1B). Similar results were obtained with GFP-and HA-tagged receptors. The overall expression of these receptors was not significantly altered by Rab8 and its mutants as determined by measuring the total GFP signal (data not shown). As the total cell surface numbers of ␣ 2B -AR and ␤ 2 -AR were similar (about 350 pmol/well), their differential inhibition by the Rab8 mutants was not due to the different cell surface expression levels of the receptors.
To further define if Rab8 GTPase could differentially regulate the cell surface expression of ␣ 2B -AR and ␤ 2 -AR, both receptors were co-expressed in the same cell populations with or without co-expression with Rab8(T22N) or Rab1(S25N) and the cell surface expression of ␣ 2B -AR and ␤ 2 -AR was measured by intact cell ligand binding. We have previously demonstrated that ␤ 2 -AR transport to the cell surface is mediated through a Rab1-dependent pathway, whereas cell surface transport of ␣ 2B -AR uses a non-conventional, Rab1-independent pathway (25). Consistent with our previous report, Rab1(S25N) selectively attenuated the cell surface number of ␤ 2 -AR, but not ␣ 2B -AR. In contrast, Rab8(T22N) inhibited the cell surface transport of both receptors. The inhibitory effect on ␣ 2B -AR was significantly higher than that on ␤ 2 -AR (Fig. 1C).
We then determined the effect of transient expression of Rab8 and Rab8(T22N) ( Fig. 2A) on cell surface expression of endogenous receptors. NG108 and MCF-7 cells express endogenous ␣ 2B -AR, whereas HEK293 and HL-1 cells express endogenous ␤-AR. Similar to the results obtained in HEK293 cells transfected with exogenous receptors, Rab8(T22N) significantly inhibited the cell surface expression of endogenous ␣ 2 -AR in NG108 and MCF-7 cells by about 50%, whereas the cell surface expression of endogenous ␤-AR in HEK293 and

Rab8 Regulation of GPCR Post-Golgi Traffic
HL-1 cells was inhibited by about 20% (Fig. 2B). These data further suggest that Rab8 differentially modulates the export trafficking of ␣ 2B -AR and ␤ 2 -AR.
Our preceding data suggest that Rab8 more potently influences the cell surface targeting of ␣ 2B -AR than ␤ 2 -AR in cell lines. We then asked the question if Rab8 could play a role in the transport of ␣ 2B -AR in the primary cultures of neurons isolated from the thalamus, which has been demonstrated to highly express endogenous ␣ 2B -AR (46). In the primary neurons, tran-sient expression of Rab8(T22N) for 2 days significantly reduced ␣ 2B -AR expression in the dendrites by 44% as compared with that in neurons transfected with wild-type Rab8 (Fig. 2C).
Effect of shRNA-mediated Knockdown of Rab8 on the Cell Surface Expression of ␣ 2B -AR and ␤ 2 -AR-To further define the role of Rab8 in regulating the cell surface transport of ␣ 2B -AR and ␤ 2 -AR, we determined the effect of shRNA-mediated knockdown of endogenous Rab8. Transient expression of shRNA targeting onto human Rab8 in HEK293 and MCF-7 cells markedly reduced the expression level of endogenous Rab8 as compared with cells transfected with control shRNA (Fig. 3, A and B). We then determined the effect of Rab8 shRNA on the cell surface expression of exogenously transfected HA-␣ 2B -AR and HA-␤ 2 -AR and endogenous ␣ 2 -AR and ␤-AR in HEK293 and MCF-7 cells. Similar to transient expression of the dominant-negative Rab8 mutants, down-regulation of endogenous Rab8 by shRNA transfection resulted in significant decreases in the cell surface expression of HA-␣ 2B -AR and HA-␤ 2 -AR by 58 and 22%, respectively (Fig. 3C). The cell surface expression of endogenous ␣ 2 -AR in MCF-7 cells and endogenous ␤-AR in HEK293 cells were attenuated by 40 and 19%, respectively, by Rab8 shRNA (Fig. 3D). These data further  suggest that the normal function of Rab8 GTPase is required for the cell surface transport of both ␣ 2B -AR and ␤ 2 -AR.
Effect of Rab8 on ␣ 2B -AR Signaling-To further confirm that Rab8 GTPase modulated the transport of ␣ 2B -AR and ␤ 2 -AR to the cell surface as measured by intact cell ligand binding and to define if Rab8 was capable of altering receptor signaling, we measured the effect of transient expression of Rab8(T22N) on ERK1/2 activation in response to stimulation with receptor agonists. ERK1/2 activation by UK14304 was reduced by 60% in HEK293 cells expressing ␣ 2B -AR and Rab8(T22N) compared with cells transfected receptor alone (Fig. 4, A and D). Furthermore, Rab8(T22N) expression also significantly attenuated ERK1/2 activation by endogenous ␣ 2B -AR in response to stimulation with UK14304 in NG108 cells (Fig. 4, C and D). In contrast, Rab8(T22N) did not significantly influence ERK1/2 activation by ISO in cells expressing ␤ 2 -AR (Fig. 4, B and D). These data indicate that inhibition of Rab8 function by expressing its dominant-negative mutants is able to differentially modulate the cell surface expression and signaling of ␣ 2B -AR and ␤ 2 -AR.
Rab8(T22N) Modulates Receptor Transport at the TGN-To define the intracellular compartment in which Rab8 regulates receptor transport, GFP-tagged ␣ 2B -AR and ␤ 2 -AR was expressed together with Rab8(T22N) and the subcellular distribution of the receptors was visualized by co-localizing with dif-ferent intracellular marker proteins. ␣ 2B -AR was strongly co-localized with the TGN marker p230 (Fig. 5A), but not with the ER marker calregulin (data not shown). In contrast, the majority of ␤ 2 -AR was expressed at the cell surface in cells expressing Rab8(T22N) (Fig. 5B). These results are consistent with the differential influences of Rab8 on the cell surface expression of ␣ 2B -AR and ␤ 2 -AR as measured by intact cell ligand binding. These data also suggest that Rab8(T22N) arrested ␣ 2B -AR in the TGN compartment, which is consistent with the Rab8 function in the TGN to plasma membrane traffic (31,32).
Association of ␣ 2B -AR and ␤ 2 -AR with Rab8 GTPase-To elucidate the molecular mechanism underlying the regulation of AR export trafficking by Rab8 GTPase, we determined if both ␣ 2B -AR and ␤ 2 -AR were capable of associating with Rab8 GTPase. HA-tagged ␣ 2B -AR or ␤ 2 -AR were transiently expressed together with GFP-tagged Rab8, Rab8(T22N), or Rab8(Q67L) in HEK293 cells and the receptors were immunoprecipitated using anti-HA antibodies. Rab8 in the immunoprecipitates was determined by Western blotting using anti-GFP antibodies. Rab8 was found in the immunoprecipitates from cells expressing ␣ 2B -AR or ␤ 2 -AR (Fig. 6A). Interestingly, the amounts of Rab8 in both receptor immunoprecipitates were higher in cells expressing Rab8(T22N) than in cells

Rab8 Regulation of GPCR Post-Golgi Traffic
expressing Rab8 or Rab8(Q67L) (Fig. 6, A and B). These data suggest that both ␣ 2B -AR and ␤ 2 -AR are able to interact with Rab8 GTPase, preferentially in its inactive, GDP-bound form.
Interaction of the C Termini of ␣ 2B -AR and ␤ 2 -AR with Rab8 GTPase-To identify the motif in ␣ 2B -AR that mediates receptor interaction with Rab8, the intracellular domains of ␣ 2B -AR including the first (ICL1, 10 residues), second (ICL2, 15 residues), and third (ICL3, 165 residues) intracellular loops and the C terminus (CT, 24 residues) (Fig. 7A) were generated as GST fusion proteins and their abilities to interact with Rab8 were determined in a GST fusion protein pulldown assay. The GST fusion proteins containing the C terminus, but not ICL1, ICL2, and ICL3 fusion proteins, were capable of interacting with Rab8 (Fig. 7B). The interaction of the GST-C-terminal fusion protein with GDP-bound Rab8(T22N) was stronger than Rab8 and Rab8(Q67L) (Fig. 7C), consistent with the co-immunoprecipitation results using the intact receptors.
We then determined if the ␤ 2 -AR C terminus was also able to interact with Rab8 GTPase. Similar to the ␣ 2B -AR C terminus, the ␤ 2 -AR C terminus consisting of 87 residues interacted with Rab8 in the GST fusion protein pulldown assay (Fig. 7D). In contrast to Rab8, Rab1 GTPase did not interact with the ␤ 2 -AR C terminus fusion protein (Fig. 7D). These data indicate that the ␣ 2B -AR and ␤ 2 -AR C termini are able to interact with Rab8 GTPase.
We next sought to determine relative binding affinities of the C termini of ␣ 2B -AR and ␤ 2 -AR for Rab8. GST fusion proteins containing the ␣ 2B -AR or ␤ 2 -AR C terminus were incubated with increasing concentrations of total cell homogenates prepared from cells transfected with GFP-Rab8 (from 0 to 2 mg). At lower concentrations of Rab8, both fusion proteins did not clearly bind Rab8. At higher concentrations of Rab8, the ␤ 2 -AR C terminus consistently bound more Rab8 than the ␣ 2B -AR C terminus (Fig. 7E).
Identification of the Rab8-binding Sites in the C Termini of ␣ 2B -AR and ␤ 2 -AR-To identify specific residues in the C termini interacting with Rab8, we first focused on the ␤ 2 -AR C

Rab8 Regulation of GPCR Post-Golgi Traffic
terminus and determined if the membrane-proximal ␣-helix region (Cys 327 -Cys 341 ) is essential for interaction with Rab8. Rab8 interaction with the ␤ 2 -AR C-terminal fusion protein containing the region Leu 342 -Leu 413 (lacking the region Cys 327 -Cys 341 ) was markedly reduced as compared with the whole C-terminal fusion protein (Fig. 8A).
We then determined the effect of mutating the dileucine (Leu 339 -Leu 340 ) motif to alanines on the interaction with Rab8. Similar to the deletion of the LL-containing region Cys 327 -Cys 341 , mutation of the LL motif in the ␤ 2 -AR C terminus dramatically reduced the interaction with Rab8 (Fig. 8B). Surprisingly, mutation of the Ile 443 -Leu 444 motif in the ␣ 2B -AR C terminus did not influence Rab8 interaction (Fig. 8B). These data demonstrate that the LL motif specifically mediates the interaction of ␤ 2 -AR, but not ␣ 2B -AR, with Rab8 GTPase.
To further search for the Rab8 binding motif in ␣ 2B -AR, each residue in the ␣ 2B -AR C terminus was mutated to alanine individually or in combination and the effect on the interaction with Rab8 was tested by GST fusion protein pulldown assays. Mutation of Thr 430 , Asp 435 , Phe 436 /Ile 443 /Leu 444 , five Arg (Arg 437 /Arg 438 /Arg 441 /Arg 442 /Arg 446 ), and Thr 449 did not influence Rab8 binding. In contrast, mutation of Val 431 , Phe 432 , Gln 434 , Trp 448 , Gln 450 , Thr 451 , and Trp 453 clearly inhibited and mutation of Asn 433 and Pro 447 almost abolished Rab8 binding (Fig. 8C). These data suggest that multiple Rab8 binding sites exist in the membrane proximal region and distal end of the ␣ 2B -AR C terminus (Fig. 8D).
Effect of the ␣ 2B -AR C Terminus on the ␤ 2 -AR Responsiveness to Rab8(T22N)-Our preceding data have demonstrated that Rab8 differentially modulates the cell surface expression of ␣ 2B -AR and ␤ 2 -AR and interacts with distinct motifs in the C termini of the receptors. To test if the differential regulation of ␣ 2B -AR and ␤ 2 -AR by Rab8 is determined by the C termini of the receptors, we measured the effect of substituting the ␤ 2 -AR C terminus with the ␣ 2B -AR C terminus on the ␤ 2 -AR response to Rab8(T22N). We generated a chimeric receptor, ␤ 2 -AR␣ 2B -ct in which the ␤ 2 -AR C terminus was replaced by the ␣ 2B -AR C terminus (Fig. 9A). Truncation of the C terminus from ␤ 2 -AR (␤ 2 -AR-ct) dramatically blocked receptor transport to the cell surface as measured by intact cell ligand binding (Fig. 9B) and subcellular localization (Fig. 9C). In contrast to ␤ 2 -AR-ct, the chimeric receptor ␤ 2 -AR␣ 2B -ct retained its abilities to transport to the cell surface (Fig. 9, B and C), which are comparable with its wild-type counterpart. Consistently, cAMP production in response to ISO stimulation was almost completely inhibited in cells expressing ␤ 2 -AR-ct and was about normal in cells expressing ␤ 2 -AR␣ 2B -ct as compared with cells expressing ␤ 2 -AR (Fig. 9D).
We then determined the effect of Rab8(T22N) on the cell surface expression of the chimeric receptor. Measurement of the cell surface receptor expression revealed that, in contrast to ␤ 2 -AR, ␤ 2 -AR␣ 2B -ct transport to the cell surface was inhibited by 45% (Fig. 10A). Such an inhibitory effect of Rab8(T22N) on the transport of ␤ 2 -AR␣ 2B -ct was very much similar to that on ␣ 2B -AR (Fig. 10A). The chimeric receptor ␤ 2 -AR␣ 2B -ct was partially co-localized with p230 in cells expressing Rab8(T22N) (Fig. 10B). These data suggest that the C termini of ␣ 2B -AR and ␤ 2 -AR play an important role in defining the sensitivity of the receptors in response to the inhibition of Rab8 function.

DISCUSSION
The molecular mechanism underlying the transport of nascent GPCRs from the Golgi to the functional destination, the plasma membrane, remains poorly understood. This issue is addressed in this article by determining the functional role of Rab8 GTPase, which has been demonstrated to modulate vesicular protein transport from the TGN to the apical/basolateral plasma membrane, in the cell surface targeting of ␣ 2B -AR and ␤ 2 -AR. We first demonstrated that expression of Rab8 mutants and Rab8 shRNA differentially modulated the cell surface expression of ␣ 2B -AR and ␤ 2 -AR. The Rab8 mutants and Rab8 B, effect of mutating the LL motif on the interaction of the ␣ 2B -AR and ␤ 2 -AR C termini with Rab8. Bottom panel, quantitative data presented as the mean Ϯ S.E. of three experiments. *, p Ͻ 0.05 versus ␤ 2 -AR CT. C, site-directed alanine scanning mutagenesis to identify Rab8 binding sites in the C terminus of ␣ 2B -AR. The ␤ 2 -AR C terminus and its truncated form lacking the region Cys 327 -Cys 341 (A), the ␣ 2B -AR and ␤ 2 -AR termini and their mutants in which the LL motif were mutated to alanines (B), and the ␣ 2B -AR C-terminal mutants in which each residue was individually or in combination mutated to alanines (C) were generated as GST fusion proteins and their abilities to interact with GFP-Rab8 GTPase were determined as described in the legend of Fig. 7. Similar results were obtained in at least three independent experiments. In A-C, upper panels, Coomassie Blue staining of GST fusion proteins; lower panels, GFP-Rab8 bound to the GST fusion proteins revealed by immunoblotting using anti-GFP antibodies. D, a summary of the Rab8-binding sites in the C terminus of ␣ 2B -AR as demonstrated in C.
shRNA more potently inhibited the transport of exogenously expressed ␣ 2B -AR than ␤ 2 -AR in HEK293 cells. ␣ 2B -AR was strongly colocalized with the TGN marker p230 in cells expressing the GDP-bound mutant Rab8(T22N), suggesting that Rab8 GTPase likely controls GPCR movement at the TGN level, which is consistent with the Rab8 function in vesiclemediated transport from the TGN (30 -36). Expression of Rab8(T22N) also more strongly inhibited the cell surface expression of endogenous ␣ 2B -AR in NG108 and MCF-7 cells than endogenous ␤-AR in HL-1 and HEK293 cells. Moreover, expression of the Rab8(T22N) mutant clearly attenuated the transport of ␣ 2B -AR to the dendrites in the primary cultures of neurons. In addition, expression of Rab8(T22N) significantly attenuated ERK1/2 activation by ␣ 2B -AR, but not ␤ 2 -AR, paralleling the effects of Rab8 on the transport of these receptors to the cell surface. Therefore, it is likely that the influences of Rab8(T22N) on receptor-mediated ERK1/2 activation were due at least in part to its effects on the receptor transport to the cell surface. However, we cannot exclude that altering Rab8 function may also modulate the intracellular trafficking of other signaling molecules involved in receptor signaling systems, which may also contribute to disruption of the normal signaling of the receptors. Nevertheless, our data demonstrated that the cell surface targeting of different GPCRs may have different sensitivities in response to the inhibition of Rab8 function, which will ultimately influence receptor signal propagation.

Rab8 Regulation of GPCR Post-Golgi Traffic
We have previously demonstrated that inhibition of Rab1 function similarly blocks the transport of ␣ 1A -AR, ␣ 1B -AR, ␤ 1 -AR, ␤ 2 -AR, and angiotensin II type 1 receptor (AT1R), without influencing the transport of ␣ 2B -AR (25)(26)(27). Similar to Rab1, inhibition of Rab6 function, which modulates retrograde protein transport from the Golgi to the ER, blocks the transport of ␤ 2 -AR and AT1R, but not ␣ 2B -AR (29), demonstrating that ␣ 2B -AR transport along the early secretory pathway uses an unconventional pathway independent of Rab1 and Rab6. Here we have demonstrated that Rab8 differentially modulates the post-Golgi transport of distinct GPCRs. These data further indicate that different Rab GTPases may selectively coordinate GPCR transport at different intracellular compartments.
To explore the molecular mechanism underlying the function of Rab8 GTPase in regulating GPCR post-Golgi transport, we determined if ␣ 2B -AR and ␤ 2 -AR could directly associate with Rab8 by coimmunoprecipitation. It has been shown the transport machinery of Rab GTPases may directly associate with cargo proteins. For example, Rab4 and Rab11 GTPases interact with the C termini of AT1R, ␤ 2 -AR, and the thromboxane A 2 receptor, which is involved in the recycling of internalized receptors from endosomes back to the plasma membrane (47)(48)(49). Rab3 modulates intracellular localization of polymeric immunoglobulin receptor via directly interacting with the receptor (50). Our data demonstrate that both ␣ 2B -AR and ␤ 2 -AR are able to form a complex with Rab8 GTPase, suggesting that the cargo GPCRs may directly interact with Rab8 GTPase to coordinate their transport from the TGN to the plasma membrane. Interestingly, similar to the interaction of ␤ 2 -AR and thromboxane A 2 receptor with Rab11 (47,49), ␣ 2B -AR and ␤ 2 -AR preferentially associate with the inactive, GDP-bound form of Rab8. Therefore, the receptors unlikely function as the downstream effectors of Rab8 GTPase, as it has been well demonstrated that the downstream effectors strongly interact with the GTP-bound Rab GTPase mutants. It is possible that the receptors may function as anchoring proteins for Rab8 GTPase localization to the TGN by providing docking sites for inactive, GDP-bound Rab8. The receptors may also function as guanine nucleotide exchange factors to facilitate the exchange of GDP for GTP and promote activation of Rab8 GTPase. This possibility is supported by the fact that some cargo proteins can activate transport machinery to modulate their transport. For example, AT1R is able to interact with and activate Rab5 GTPase, which is involved in the regulation of receptor internalization (48).
The most interesting data presented here are that ␣ 2B -AR and ␤ 2 -AR use distinct motifs in the C termini to interact with Rab8 GTPase and modulate their post-Golgi transport. We first demonstrated that the ␣ 2B -AR C terminus, but not three intracellular loops, interacted with Rab8 GTPase in GST fusion protein pulldown assays. Similar to the ␣ 2B -AR C terminus, the ␤ 2 -AR C terminus also interacted with Rab8. We further demonstrated that mutation of the LL motif located in the membrane-proximal C-terminal region markedly and specifically reduced Rab8 interaction with the C terminus of ␤ 2 -AR, but not ␣ 2B -AR, suggesting that the LL motif selectively mediates ␤ 2 -AR interaction with Rab8. It has been well documented that the LL motif functions as sorting signals at the TGN for basolateral cell surface transport and at the plasma membrane for endocytosis in clathrin-coated vesicles through interacting directly with the clathrin adaptor protein complex (51)(52)(53). The LL motif is highly conserved in the membrane-proximal C-terminal portion among the family A GPCRs (54) and we and others have demonstrated that the LL motif regulates export trafficking of a number of GPCRs (54,55). However, the molecular mechanism underlying the function of the LL motif in GPCR transport remains unknown. It has been suggested that the LL motif is involved in the folding of GPCRs, as mutation of the LL motif disrupts receptor ligand binding (55). Our data presented in this article have demonstrated a novel function of the LL motif to mediate receptor interaction with Rab8 GTPase, which is likely to modulate receptor transport at the TGN level. These data also suggest that a single LL motif may modulate export trafficking of newly synthesized GPCRs at multiple organelles. In addition to regulating ER export, the LL motif may also coordinate nascent GPCR exit from the Golgi.
In contrast to ␤ 2 -AR using the LL motif to interact with Rab8, ␣ 2B -AR uses multiple sites located in the membrane-proximal (VFNQ) and distal (PW and QTGW) C terminus to interact with Rab8 GTPase. In particular, residues Asn 433 and Pro 447 likely play a crucial role in mediating ␣ 2B -AR interaction with Rab8 GTPase as mutation of either one almost abolished Rab8 interaction in GST fusion protein pulldown assays. These data suggest that different GPCRs (i.e. ␣ 2B -AR and ␤ 2 -AR) may provide distinct docking sites for Rab8 GTPase to coordinate their export from the TGN.
To define if differential modulation of ␣ 2B -AR and ␤ 2 -AR transport by Rab8(T22N) is determined by the receptor C termini, we determined the effect of Rab8(T22N) on the transport of the chimeric receptor ␤ 2 -AR␣ 2B -ct in which the ␤ 2 -AR C terminus was substituted with the ␣ 2B -AR C terminus. The cell surface transport of ␤ 2 -AR␣ 2B -ct was inhibited at a level similar to ␣ 2B -AR, but greater than ␤ 2 -AR. These data suggest that the C termini may possess the structural determinants that modulate receptor response to Rab8(T22N). As we have shown that Rab8 GTPase interacts with distinct motifs in the C termini of ␣ 2B -AR and ␤ 2 -AR, different responsiveness of ␣ 2B -AR and ␤ 2 -AR to the inhibition of Rab8 function is likely determined by their differential interactions with Rab8 GTPase. Interestingly, ␤ 2 -AR and ␣ 2B -AR interaction with Rab8 and their responsiveness to Rab8 inhibition appear to be opposite. The simplest explanation could be that in cells that express Rab8 dominant mutants or shRNA and have less functional Rab8, ␤ 2 -AR, which has higher affinity for Rab8, is still able to bind Rab8 to maintain its transport to the cell surface. In contrast, under the same condition, ␣ 2B -AR, which has lower affinity for Rab8, might not be able to associate enough Rab8 required for its normal export. However, whether or not differential regulation of distinct receptors by Rab8 GTPase is indeed dictated by their differential interactions with Rab8 needs further investigation.
There are several possibilities regarding the differential regulation of ␣ 2B -AR and ␤ 2 -AR transport by Rab8 GTPase. It is possible that transport from the TGN to the cell surface of ␣ 2B -AR and ␤ 2 -AR is mediated through distinct pathways, which have different requirements for Rab8. Consistent with this possibility, we have demonstrated that ␣ 2B -AR and ␤ 2 -AR use different routes to move from the ER to the Golgi (25). Furthermore, a recent study has demonstrated that vesicular stomatitis virus glycoprotein and Na ϩ -K ϩ -ATPase follow different pathways to the basolateral cell surface in polarized cells (37). It is also possible that post-Golgi transport of ␣ 2B -AR and ␤ 2 -AR may be mediated through different transport vesicles that are differentially modulated by Rab8 GTPase. Consistent with this possibility, a number of proteins are involved in sorting of cargo proteins at the TGN to different destinations, such as the plasma membrane and the early and late endosomes. For example, adaptor proteins and GGAs (Golgi-localized ␥-earcontaining ADP-ribosylation factor-binding proteins) directly interact with different LL-based motifs in cargo proteins to direct the sorting at the TGN (53). The post-translational modification of ␣ 2B -AR and ␤ 2 -AR may also play a role in their differential regulation by Rab8 GTPase. Specifically, ␤ 2 -AR is a glycosylated receptor that has three putative N-linked glycosylation sites at positions 4, 176, and 188, whereas ␣ 2B -AR does not contain glycosylation signals. Glycosylation of the receptors occurs during their transport to the Golgi to achieve a fully maturated conformation competent for export from the TGN. Different glycosylation statuses of ␣ 2B -AR and ␤ 2 -AR may influence the characteristics of receptor export from the TGN and regulation by Rab8 GTPase.