Enhanced ubiquitylation and accelerated degradation of the dopamine transporter mediated by protein kinase C.

Dopamine transporter (DAT) localization in dopaminergic neurons plays an important role in regulating dopamine signaling. However, the mechanisms of DAT trafficking that control DAT localization are still poorly understood. To gain insight into these mechanisms, human DAT was purified in large amounts using a two-step affinity chromatography procedure from untreated HeLa cells or cells treated with phorbol 12-myristate 13-acetate (PMA). Mass spectrometric analysis of purified DAT complexes revealed the presence of several proteins, among which ubiquitin was particularly abundant in the PMA-treated sample. Western blotting of highly purified DAT protein confirmed constitutive ubiquitylation of DAT and a dramatic increase in DAT ubiquitylation in cells treated with PMA. This increase was blocked by pretreatment with the protein kinase C (PKC) inhibitor bis-indolylmaleimide. DAT ubiquitylation by ectopically expressed ubiquitin was demonstrated in cells transiently transfected with yellow fluorescent protein-tagged ubiquitin. In addition, fluorescence resonance energy transfer was detected between cyan fluorescent protein-tagged DAT and yellow fluorescent protein-tagged ubiquitin, indicative of DAT-ubiquitin conjugation. Interestingly, the largest fluorescence resonance energy transfer signals were observed in endosomes. Ubiquitylated DAT was detected in the plasma membrane using cell surface biotinylation as well as in intracellular compartments, suggesting that ubiquitylation begins at the plasma membrane and is maintained in endosomes. In both porcine aortic endothelial and HeLa cells, where PKC-dependent DAT ubiquitylation was observed, PKC activation resulted in rapid degradation of DAT (t12 = 1-2 h). Altogether, these data suggest that PKC-induced DAT ubiquitylation may target DAT to lysosomal degradation.

The dopamine transporter (DAT) 2 is a member of the family of Na ϩ / Cl Ϫ -dependent plasma membrane transporters (SLC6 gene family) that are responsible for rapid clearance of neurotransmitters from the extracellular space. This family also includes norepinephrine, serotonin, ␥-aminobutyric acid, and glycine transporters (1). Members of the SLC6 family share similar predicted topology of a single polypeptide that contains twelve transmembrane segments, a large second extracel-lular glycosylated loop, and cytoplasmic amino-and carboxyl-terminal tails (2,3). DAT and several other transporters of this family have been found to be constitutively oligomerized in vitro and in cells (4 -8).
The amount of DAT at the plasma membrane and, therefore, dopamine uptake capacity are determined by trafficking of the DAT protein, which is regulated by several signaling cascades including signaling through protein kinase C (PKC). Activation of PKC by 4␣-phorbol 12-myristate 13-acetate (PMA) leads to a reduction in the V max of dopamine transport without a change in the substrate affinity (K m ) as well as in down-regulation of surface DAT protein (9 -15). This downregulation of DAT activity and levels has been shown to be due to accelerated internalization of the transporter through clathrin-coated pits and possibly reduced recycling of DAT (16,17). Similar effects of phorbol esters on the serotonin transporter, the norepinephrine transporter, ␥-aminobutyric acid transporter 1, and glycine transporter 1 have also been reported (reviewed in Ref. 18). In addition to downregulation of the plasma membrane pool of DAT, PKC activation caused degradation of DAT in Madin-Darby canine kidney cells (14). In contrast, it has been suggested that internalized DAT is mostly sorted to the recycling pathway in PC12 cells stimulated with PMA (15).
The mechanisms of PKC effects on DAT endocytosis and dopamine uptake are not fully understood. Phosphorylation of the amino-terminal tail of DAT by PKC has been observed in rat striatal synaptosomal preparations and transfected mammalian cell lines (10,19,20). However, truncations or point mutations of the phosphorylation sites in DAT did not inhibit PKC-induced down-regulation of DAT, suggesting that this PKC effect is not mediated by DAT phosphorylation (17,21,22). On the other hand, amino-terminal phosphorylation has been shown to be important for reversed dopamine transport by DAT (23). In contrast, PKC-dependent phosphorylation of the second intracellular loop of glycine transporter 1 has been implicated in the phorbol estermediated down-regulation of this transporter (24), thus illustrating differences in the regulation of endocytosis of different members of this transporter family (25).
Mutations of the amino acid motifs that correspond to consensus sequences for conventional internalization and sorting signals, such as tyrosine-and dileucine-based signals, did not establish any role for these motifs in PKC-induced DAT down-regulation (17,21). Recently, monoubiquitylation has been proposed to serve as a sorting signal for internalization and lysosomal targeting of membrane proteins (26). Unlike polyubiquitylation, which typically targets proteins for proteasomal degradation, monoubiquitin moieties can be recognized by the components of plasma membrane internalization and endosomal sorting machinery. Hence, we performed mass spectrometry and immunoblot analyses of DAT and demonstrated constitutive ubiquitylation of DAT and a dramatic increase of DAT ubiquitylation upon PKC activation. Because PMA caused a substantial decrease in the DAT half-life in our model expression systems, our data imply that DAT ubiquitylation may be responsible for the accelerated degradation of DAT caused by PKC activation.

EXPERIMENTAL PROCEDURES
Chemicals and Antibodies-PMA, N-ethylmaleimide, cycloheximide, and anti-FLAG M2 affinity gel were purchased from Sigma. Bisindolylmaleimide I (BIM) was from Calbiochem. Polyclonal rabbit antibody ab290 to green fluorescent protein (GFP) was purchased from Abcam Ltd. (Cambridge, UK); monoclonal mouse antibody to GFP was from Zymed Laboratories Inc. (South San Francisco, CA); monoclonal rat antibody against the amino terminus of DAT was from Chemicon, Inc. (Temecula, CA); and monoclonal mouse antibody P4D1 to ubiquitin was from Santa Cruz Biotechnology (Santa Cruz, CA).
Plasmid Construction-The full-length human DAT in pcDNA3.1 was kindly provided by Dr. G. W. Miller (Emory University). To generate the FLAG epitope and His-tagged DAT, a set of forward primers containing the NheI site and sequences encoding the FLAG and His tags (6 or 10 histidines) and reverse primers carrying an EcoRI site after the stop codon were used to amplify the human DAT sequence by PCR using Pfu polymerase (Stratagene, La Jolla, CA). The DNA fragment was cloned into pcDNA3.1(ϩ) (Invitrogen). To generate wild-type and mutant YFP-ubiquitin (YFP-Ub and YFP-UbAA, respectively), a fragment of BsrgI-HindIII containing the ubiquitin sequence was transferred from the plasmid GFP 2 -UbiWT or GFP 2 -UbiAA kindly donated by Dr. M. Bouvier (University of Montreal, Quebec, Canada) (27) into the pEYFP-C1 vector (Clontech, Palo Alto, CA). All constructs and point mutations were verified by automatic dideoxynucleotide sequencing. The plasmid CFP-DAT was described previously (6). YFP-Hrs (hepatocyte growth factor receptor substrate) was provided by Dr. H. Stenmark (Radium Institute, Oslo, Norway).
Cell Culture and Transfections-Human cervical carcinoma HeLa cells were grown in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and antibiotics. Porcine aortic endothelial (PAE) cells were grown in F12 medium containing 10% fetal bovine serum and antibiotics. Cells were grown to 50 -80% confluence and transfected with appropriate plasmids using Effectene (Qiagen, Hilden, Germany). The constructs FLAG/His 6 -DAT or FLAG/His 10 -DAT (FH-DAT) were transfected into HeLa and PAE cells using Effectene, and the cells stably expressing DAT were selected by growing them in the presence of G418 (400 g/ml). For DAT purification, the cells were plated into 20-cmsquare dishes for large scale purification or 35-mm dishes for small scale purification and used at near 100% confluence. For microscopy, the cells were split 1 day after transfection onto glass coverslips and used for experiments on the second or third day.
Purification of FH-DAT by Ni-NTA-Agarose and FLAG M2 Affinity Chromatography-To purify FH-DAT, HeLa cells stably expressing FH-DAT were grown to near 100% confluence and treated with vehicle (Me 2 SO) or PMA in the absence or presence of BIM. The cells were placed on ice and washed three times with Ca 2ϩ -and Mg 2ϩ -free cold phosphate-buffered saline (PBS), and the proteins were solubilized in lysis buffer (25 mM HEPES, pH 7.6, 10% glycerol, 100 mM NaCl, 10 mM sodium fluoride, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 10 g/ml aprotinin, 1% Triton X-100, 10 mM N-ethylmaleimide, 15 mM imidazole) for 20 min at 4°C. The lysate was then centrifuged at 100,000 ϫ g for 45 min to remove insoluble material. After centrifugation, cleared lysate was incubated with Ni-NTA, previously equilibrated with the same lysis buffer, for 1 h at 4°C on the nutator. The mixture was then transferred to a plastic column (Poly-Prep chromatography columns, Bio-Rad), the flow-through was discarded, and the column was washed with lysis buffer. FH-DAT was then eluted with 250 mM imidazole in lysis buffer. The eluate was diluted 10 times with the FLAG binding buffer (50 mM Tris, 150 mM NaCl, 10% glycerol, 1% Triton) and incubated for 2 h with FLAG M2 affinity gel. The mixture was washed five times with 1 ml of FLAG binding buffer, and FH-DAT was eluted with 0.1 M glycine (pH 3.5). The eluted fraction was quickly mixed with an equal volume of 1 M Tris (pH 7.6). All procedures were performed at 4°C.
Mass Spectrometry and Data Processing-Affinity-purified human DAT samples were analyzed using mass spectrometry. Samples were either first resolved by gel electrophoresis followed by gel excision, digestion, and extraction of both tryptic and non-tryptic (elastase digests) peptides (28,29) or directly digested in solution to produce both tryptic and non-tryptic peptides (30). Peptides prepared using both methods were analyzed by multidimensional protein identification technology (MudPIT) (31) with a Finnigan LTQ linear ion trap (Ther-moElectron, San Jose, CA) using an in-house constructed microspray ion source. Briefly, peptide samples were desalted and loaded onto a biphasic chromatography column constructed from 100-m inner diameter, 365-m outer diameter fused silica capillary tubing and packed with 8 cm of 5-m Aqua C18 reverse phase packing material (Phenomenex) followed by 4 cm of 5-m Partisphere strong cation exchanger (Whatman). Peptide digests were loaded using a pressure bomb, and an automated MudPIT analysis was performed with six salt elution steps as described previously (30). All tandem mass spectrometry spectra were searched against an NCBI human-mouse-rat data base concatenated onto a randomized "shuffled" data base using the Sequest algorithm (32). Ubiquitylations were identified by searching for mass shifts localized to lysine residues (ϩ114 and ϩ383) from the samples digested with trypsin (33). Positive protein identifications required normalized cross-correlation scores (X corr ) Ն 0.3 for any charge state (29) and delta correlation scores (⌬C n ) Ն 0.1. False discovery rates were Յ5%, as determined by hits to the randomized shuffled data base. Partial tryptic status was required for ubuitylation identifications. Polyubiquitin sites were identified on ubiquitin lysine residues (34,35).
Surface Biotinylation-Cells expressing FH-DAT proteins were grown in 35-mm dishes and treated with vehicle (Me 2 SO) or PMA for the indicated times. Cell surface biotinylation was performed as described previously (36). Briefly, the cells were washed with cold PBS containing 0.1 mM CaCl 2 and 1 mM MgCl 2 and incubated for 20 min on ice with 1 mg/ml sulfo-N-hydroxysuccinimidobiotin (EZ-Link TM , Pierce) in PBS followed by a second incubation with fresh sulfo-Nhydroxysuccinimidobiotin. After biotinylation, the cells were washed twice with cold PBS, incubated on ice with 0.1 M glycine in PBS, and washed with PBS again. The cells were then solubilized at 4°C in lysis buffer (described above) supplemented with 10 mM Tris-HCl (pH 7.6). The lysates were cleared by centrifugation for 10 min at 16,000 ϫ g, and the FH-DAT was precipitated with 50 l of Ni-NTA, as described above. After elution of FH-DAT from the Ni-NTA-agarose, the eluted fraction was diluted with lysis buffer, and the biotinylated proteins were precipitated with NeutrAvidin TM beads (Pierce), washed five times with lysis buffer, and denatured by heating the beads in sample buffer at 95°C for 5 min. The supernatant from the NeutrAvidin TM bead precipitation was incubated with 50 l of FLAG M2 affinity gel, and non-biotinylated DAT was purified as described above. Western blotting was performed with mouse or rat monoclonal antibodies to ubiquitin or DAT, respectively, followed by secondary antibodies conjugated with horseradish peroxidase. Detection was accomplished using enhanced chemiluminescence (Pierce).
Immunofluorescence Staining-The cells grown on glass coverslips were washed with Ca 2ϩ -, Mg 2ϩ -free PBS, fixed with freshly prepared 4% paraformaldehyde (Electron Microscopy Sciences, Ft. Washington, PA) for 15 min at room temperature, and mildly permeabilized using a 3-min incubation in Ca 2ϩ -, Mg 2ϩ -free PBS containing 0.1% Triton X-100 and 0.5% bovine serum albumin at room temperature. Immunostaining with DAT antibodies and image acquisition using epifluorescence Mariannas TM work station (Intelligent Imaging Innovation, Denver, CO) were performed essentially as described previously (36).
Live Cell Microscopy and Fluorescence Energy Transfer (FRET) Measurements-PAE stably expressing CFP-DAT cells, PAE or HeLa cells transiently expressing CFP-DAT, and YFP-tagged ubiquitin or YFP-Hrs were grown on 25-mm glass coverslips. After treatments, the coverslips were mounted in a microscope chamber and placed on a microscope stage. In experiments with YFP-Hrs, a z-stack of 30 images was acquired through CFP and YFP channels and deconvoluted using a nearest neighbor algorithm.
The method of sensitized FRET measurements used to examine FRET between CFP and YFP has been described previously (37,38). Briefly, images through YFP, CFP, and FRET filter channels were acquired using a Mariannas TM fluorescence imaging work station consisting of a Zeiss inverted microscope equipped with a cooled charged couple device CoolSnap HQ (Roper, CA), dual filter wheels, and a Xenon 175-watt light source, all controlled by SlideBook software. Images were background-subtracted, and the corrected FRET was calculated. Corrected FRET images were presented in a quantitative pseudocolor. The apparent FRET efficiency (E d ) was calculated as described previously (38). All calculations were performed using the FRET statistic module of SlideBook 4.0 or 4.1.

RESULTS
To isolate DAT in amounts sufficient for mass spectrometry analysis, a two-step affinity purification procedure was developed. Two sequences corresponding to the FLAG epitope and His 6 or His 10 were introduced in tandem at the amino terminus of DAT. Although we initially used a hexahistidine tag, His 10 allowed tighter binding of tagged DAT to Ni-NTA-agarose and therefore more stringent washing conditions, thus eliminating a significant amount of nonspecific interactions. Therefore, in most of the experiments FLAG-His 10 -DAT (FH-DAT) was used (Fig. 1A). FH-DAT was stably expressed in HeLa and PAE cells. FH-DAT migrated on SDS-PAGE as a smeared band of 85-90 kDa (Fig. 1). Localization of FH-DAT at steady-state growth conditions or in cells stimulated with 1 M PMA (condition of increased DAT endocytosis, Fig. 1A) was essentially similar to that observed in numerous studies of untagged DAT or various DAT fusion proteins (Fig. 1B) (6,14,21). It has been demonstrated in several studies that attachment of small epitope tags or GFP to the DAT amino terminus does not affect DAT function and trafficking (6,14,21,39).
Purification of FH-DAT from HeLa cells by consecutive binding and elution from Ni-NTA-agarose and FLAG M2 gel typically yielded several micrograms of FH-DAT readily detectable as 85-90 kDa (monomer) and ϳ150 -170 kDa (SDS-resistant dimer) bands by Coomassie staining (Fig. 1C). As shown in Fig. 1B, the second purification step dramatically enriched DAT and eliminated most of the contaminating proteins from the preparation. Several proteins that are specifically or nonspecifically associated with DAT were also detected in the final samples (Fig. 1B). Surprisingly, liquid chromatography-tandem mass spectrometry analysis of the 85-kDa FH-DAT band or total eluates from the FLAG M2 gel revealed the presence of the large number of peptides corresponding to human ubiquitin. Furthermore, Lys 11 -, Lys 48 -, and Lys 63 -linked ubiquitin chains were detected in PMA-treated and untreated FH-DAT (TABLE ONE). Moreover, the number of ubiquitylated DAT peptide hits was ϳ3-fold higher by spectral count in DAT purified from HeLa cells treated with PMA than from untreated cells. Lysines 19,35, and 599 of DAT were found to be ubiquitylated (TABLE  ONE), although additional experiments are necessary to complete the mapping of DAT ubiquitylation sites.
Western blotting of an aliquot of the eluate from the FLAG-agarose confirmed the presence of ubiquitin immunoreactivity in FH-DAT recovered from untreated and PMA-treated cells (Fig. 1C). To ensure that ubiquitin is conjugated to DAT rather than to a DAT-associated protein, FH-DAT was purified in the presence of SDS to minimize protein-protein interactions. Staining of the specific eluate demonstrated substantially cleaner preparations than those obtained in the absence of SDS (Fig. 1C). Mass spectrometry as well as Western blotting of these DAT preparations detected, respectively, ubiquitin peptides and a char- A, schematic representation of DAT tagged with the FLAG epitope and His 10 sequences at the amino terminus (FH-DAT). HeLa cells stably expressing FH-DAT were incubated with vehicle (DMSO, dimethyl sulfoxide) or 1 M PMA for 30 min at 37°C, and the cells were immunostained using anti-DAT followed by secondary anti-rat labeled with CY3. Bar, 10 m. B, HeLa/FH-DAT cells grown in 20-cm trays were treated with vehicle or PMA as in A and then solubilized in 20-ml lysis buffer/tray. The cleared lysate (ϳ30 mg of protein) was affinity-purified using Ni-NTA-agarose and FLAG M2 gel as described under "Experimental Procedures." Coomassie-stained gels of the eluates from the Ni-NTA (left) and FLAG M2 (right) are shown. dDAT, DAT dimers; mDAT, mature DAT monomer; ngDAT, non-glycosylated DAT. C, FH-DAT was purified as described in B with the exception that 0.1% SDS was included throughout the purification procedure. A 10% aliquot of purified FH-DAT preparation was used for Western blotting (WB) with ubiquitin and DAT antibodies.
acteristic smear of the ubiquitylated signal with the maximum ubiquitin immunoreactivity at ϳ125-130 kDa.
DAT ubiquitylation in cells treated with 1 M PMA reached a maximum at ϳ15-30 min at 37°C and then decreased with continued incubation of cells with PMA (data not shown). PMA-induced ubiquitylation was blocked by the inhibitor of conventional PKCs, BIM (1 M), confirming that this PMA effect is PKC-dependent (Fig. 2). To test whether DAT can be ubiquitylated by ectopically expressed ubiquitin, HeLa/FH-DAT cells were transiently transfected with YFP-ubiquitin, and FH-DAT was then affinity-purified on Ni-NTA-agarose. Blotting with GFP antibodies revealed the presence of high molecular weight GFP immunoreactivity in purified FH-DAT, indicative of the conjugation of YFP-ubiquitin to FH-DAT (Fig. 3). There appear to be discrete YFP-ubiquitin bands, suggesting that these bands may correspond to mono-and diubiquitylated FH-DAT (Fig. 3). The specificity of ubiquitylation was confirmed by the absence of incorporation of a YFP-ubiquitin mutant incapable of conjugation to lysines (YFP-UbAA). The amount of YFP-ubiquitin conjugated to DAT was relatively small, probably because of the very high concentration of endogenous ubiquitin that competed with YFP-ubiquitin for conjugation to DAT. Importantly, although there was a significant increase of YFP-ubiquitin incorporation induced by PMA, the overall incorporation of YFP-ubiquitin in cellular proteins was not dependent on PKC activity (Fig. 3, supernatants). This suggests that the PKC effect on DAT is specific.   Fig. 1, and FH-DAT was pulled down using Ni-NTA. YFP-containing proteins were then immunoprecipitated (IP) with GFP antibodies. FH-DAT pull-downs were probed with GFP, ubiquitin, and DAT antibodies. Asterisks indicate nonspecific bands detected by GFP antibodies. To examine where in the cell FH-DAT is ubiquitylated, CFP-DAT was co-expressed with YFP-ubiquitin. Fluorescence microscopy revealed that a bulk of YFP-ubiquitin was distributed in the cytosol and nucleus (Fig. 4A). In addition, YFP-ubiquitin was accumulated in endosomes in untreated cells and cells stimulated with PMA, although CFP-DAT could be detected in YFP-ubiquitin-decorated endosomes only in cells treated with PMA. Co-localization of CFP-DAT and YFP-ubiquitin in the plasma membrane ruffles and cell edges was also detected, although this co-localization was much less pronounced. The YFP-UbAA mutant was diffusely distributed throughout the cytosol and nucleus, was not detected in endosomes or the plasma membrane, and was not co-localized with CFP-DAT (Fig. 4A). These data suggested that the plasma membrane and endosomal YFP-ubiquitin represent a pool of YFP-ubiquitin conjugated to membrane-associated proteins, including DAT.
Measurements of FRET between CFP-DAT and YFP-ubiquitin revealed specific FRET signals in endosomes of cells treated with PMA, indicative of the close proximity of CFP-DAT and YFP-ubiquitin (Fig.  4B). This result is consistent with conjugation of YFP-ubiquitin to CFP-DAT. The relatively low apparent efficiency of energy transfer (E d ) could be due to a very low extent of DAT ubiquitylation by YFP-ubiquitin as evident from immunoprecipitation experiments in Fig. 3. Altogether, the experiments presented in Figs. 3 and 4 suggest that endosomes represent the major localization site of ubiquitylated DAT.
Surface biotinylation of cells incubated with PMA revealed the presence of ubiquitylated DAT in the extracellularly exposed pool of proteins, suggesting that DAT ubiquitylation may begin at the cell surface (Fig. 5). The amount of ubiquitylated DAT was, however, higher in the non-biotinylated fraction of cells (Fig. 5). This suggests that a large pool of mature DAT protein, presumably endosomes, is ubiquitylated in intracellular membranes.
Because ubiquitylation of integral membrane proteins has been implicated in lysosomal targeting of these proteins, we examined the rate of degradation of DAT in cells treated with PMA. In these experiments protein synthesis was inhibited by cycloheximide to eliminate the contribution of newly synthesized DAT in the total pool of cellular DAT. In both HeLa (Fig. 6A) and PAE cells (data not shown), PMA treatment resulted in dramatic down-regulation of the FH-DAT protein with a half-life time of ϳ1-2 h. These experiments suggest that DAT ubiquitylation correlates with PKC-dependent degradation and may, therefore, underlie the PKC-dependent acceleration of DAT turnover.
To test whether internalized DAT is endocytosed in the endosomes containing ubiquitin recognition sorting machinery, the localization of CFP-DAT was compared with the localization of overexpressed YFP-  Hrs. The Hrs-signal-transducing adaptor molecule complex is directly involved in the recognition of ubiquitylated cargo in sorting endosomes (40). Live cell three-dimensional imaging and deconvolution revealed significant accumulation of DAT in large endosomes decorated by YFP-Hrs (Fig. 6B). Overexpression of Hrs is known to result in dramatic enlargement of endosomes and blockade of trafficking through multivesicular endosomes (MVBs) of various cargos, in particular, ubiquitylated membrane proteins destined for the lysosomal degradation pathway (41,42). Fig. 6B shows that overexpression of YFP-Hrs did result in accumulation of intracellular CFP-DAT. CFP-DAT was partially colocalized with YFP-Hrs on the limiting membranes of endosomes. In addition, CFP-DAT could be seen in the lumen of many large endosomes. This pool of CFP-DAT probably represented transporters that were incorporated into internal vesicles of MVBs.

DISCUSSION
Ubiquitylation of transmembrane proteins, such as various receptors and channels, has been implicated in the regulation of their endocytosis, post-endocytic trafficking, and turnover (26,43). In our study we detected constitutive and PKC-dependent ubiquitylation of DAT by endogenous ubiquitin in HeLa and PAE cells. We provide several lines of evidence that ubiquitin is conjugated to the DAT molecule rather than to DAT-associated proteins. First, endogenous ubiquitin and polyubiquitin were detected in highly purified DAT preparations by mass spectrometry. Second, several ubiquitylated peptides of DAT were identified. Third, a high molecular weight band was detected by immunoblotting with the ubiquitin antibody in highly purified DAT preparations. Fourth, PKC-induced ubiquitylation of DAT by heterologously expressed YFP-ubiquitin was demonstrated.
Ubiquitylation of immunoprecipitated DAT with epitope-tagged ubiquitin has been reported recently in cells overexpressing Parkin, a RING-containing E3 ubiquitin ligase (44). The latter study implicated Parkin-induced DAT ubiquitylation in the quality control of newly synthesized DAT processing in the endoplasmic reticulum. Our data show that the mature DAT protein is constitutively ubiquitylated by endogenous ubiquitin at the plasma membrane and that ubiquitylation of the mature DAT is markedly enhanced by PKC activity (Figs. 1, 2, and 5). Furthermore, we have not detected ubiquitylation of immature nonglycosylated DAT. Therefore, our observations imply that DAT ubiquitylation may have a role in the endocytic trafficking of DAT. The effects of PKC on DAT endocytosis and post-endocytic trafficking have been observed in various experimental systems, but the molecular mechanisms underlying these effects are unknown. Recently, it has been demonstrated that the simultaneous replacement of the carboxylterminal residues 587-591 with alanines in the Tac-DAT chimera or the full-length human DAT inhibits PKC-dependent endocytosis of the chimera or the dopamine uptake capacity of DAT, respectively (45). Based on these data the latter report suggested that residues 587-591 constitute an internalization signal. However, the crystal structure of the homologous bacterial leucine transporter revealed that residues 587 and 588 represent a part of the 12-transmembrane ␣-helix (46). Therefore, multiple alanine substitutions of these two and neighboring residues may lead to misfolding of this part of the DAT molecule. In fact, our previous studies revealed that mutation of Lys 590 to alanine resulted in the retention of the newly synthesized DAT in the endoplasmic reticulum, presumably because of transporter misfolding (36). Thus, residues 587-591 may not represent an endocytosis signal per se but rather may be indirectly required for PKC-dependent endocytosis.
Demonstration of DAT ubiquitylation in response to PKC activation may shed light on some of the mechanisms of PKC-induced DAT trafficking. A growing literature supports the model by which ubiquitylation of various transmembrane endocytic cargos facilitates their internalization and subsequent degradation (26,43). Although the mechanisms of internalization of ubiquitylated cargo remain obscure, the involvement of ubiquitylation in the endosomal sorting step of trafficking and the mechanisms of this step involving recognition of ubiquitylated cargo by endosomes Hrs-signal-transducing adaptor molecule and endosomal sorting complex required for transport complexes are well established (47). Because PKC activation led to accelerated degradation of DAT in both HeLa and PAE cells in our experiments (Fig. 6A), such a PKC effect on DAT turnover is likely a common phenomenon in different cells. Thus, it is possible that DAT ubiquitylation could be responsible for DAT interactions with Hrs and ESCRT complexes at the limiting membrane of MVBs followed by DAT incorporation into the internal vesicles of MVBs. In fact, we have frequently observed DAT in the lumen of large endosomes containing Hrs at the limiting membrane (Fig. 6B).
Typically, degradation of ubiquitylated proteins is blocked by inhibitors of lysosomes suggesting that these cargos are degraded through lysosomal, rather than proteasomal, pathways. Although in several cases, degradation of transmembrane cargo is also blocked by proteasomal inhibitors, the mechanisms of these effects are not understood (48,49). Degradation of DAT induced by PMA in Madin-Darby canine kidney cells was shown to be blocked by lysosomal but not proteasomal inhibitors (14). Sorting of transmembrane cargo in endosomes is thought to be mediated by monoubiquitylation rather than polyubiquitylation, which is responsible for targeting of proteins to the proteasome (Lys 48 -linked ubiquitin chains) (26). However, recent reports suggested a role for polyubiquitylation in lysosomal targeting (48,50). Likewise, polyubiquitylation of DAT may be involved in PKC-induced DAT degradation. In particular, detection of Lys 63 -linked chains in DAT in our experiments is especially interesting because Lys 63 -mediated diubiquitylation has been suggested to serve as a trafficking signal (51).
The mechanisms by which PKC activation leads to DAT ubiquitylation are unknown. PKC-dependent ubiquitylation must be mediated by either direct or indirect interaction of DAT with E3 ubiquitin ligase(s). In dopaminergic neurons, the RING domain containing the E3 ligase Parkin and ␣-synuclein may mediate DAT ubiquitylation (44,52). However, these proteins are not expressed in HeLa and PAE cells. Yeast membrane transporters and several other proteins are ubiquitylated by the HECT domain-containing E3 ubiquitin ligase Rsp5p/Npi1 (26). Therefore, it is also possible that the mammalian homolog of Rsp5, NEDD4 E3 ligase, which has a Ca 2ϩ -and phospholipid-binding C2 domain, could mediate PKC effects on DAT ubiquitylation (53). Interestingly, activation of PKC by phorbol esters leads to polyubiquitylation and degradation of PKC itself (54). This may suggest that there is an E3 ubiquitin ligase that is associated with PKC and that can participate in PKC-dependent ubiquitylation of transmembrane proteins like DAT.