Constitutive and Protein Kinase C-Induced Internalization of the Dopamine Transporter is Mediated by a Clathrin-Dependent Mechanism

Authors


Alexander Sorkin, alexander.sorkin@uchsc.edu

Abstract

The amount of dopamine transporter (DAT) present at the cell surface is rapidly regulated by the rates of DAT internalization to endosomes and DAT recycling back to the plasma membrane. The re-distribution of the transporter from the cell surface to endosomes was induced by phorbol ester activation of protein kinase C in porcine aortic endothelial cells stably expressing the human DAT. Inhibition of DAT recycling with the carboxylic ionophore monensin also caused significant accumulation of DAT in early endosomes and a concomitant loss of DAT from the cell surface, due to protein kinase C-independent constitutive internalization of DAT in the absence of recycling. Such monensin-induced relocation of DAT to endosomes was therefore utilized as a measure of the constitutive internalization of DAT. Knock-down of clathrin heavy chain or dynamin II by small interfering RNAs dramatically inhibited both constitutive and protein kinase C-mediated internalization of DAT. In contrast, neither monensin-dependent nor phorbol ester-induced re-distribution of DAT were affected by inhibitors of endocytosis through cholesterol-rich membrane microdomains. Mutational analysis revealed the potential importance of amino acid residues 587–597 in DAT internalization. Altogether, the data suggest that both constitutive and protein kinase C-mediated internalization of DAT utilize the clathrin-dependent endocytic pathway, but likely involve unconventional mechanisms.

The plasma membrane dopamine transporter (DAT) belongs to the family of plasma membrane Na+/Cl dependent neurotransmitter transporters, which includes transporters for norepinephrine, serotonin, glycine and GABA (1). The re-uptake of dopamine by DAT is the primary mechanism of termination of dopaminergic signaling in the brain. DAT is also a target of psychostimulants. In the mammalian central nervous system, DAT is expressed exclusively in a small subset of neurons, called dopaminergic neurons, the most prominent of which arise from the substantia nigra and ventral tegmental area and project to the striatum and cerebral cortex. Electron microscopy studies revealed DAT localization in the plasma membrane near the active zone in synapses located along the axonal processes, as well as in various intracellular membrane compartments in the soma and dendrites (2–4).

Several lines of evidence, largely generated in heterologous expression cell systems, suggest that DAT expression at the cell surface can be rapidly regulated by endocytosis. For instance, protein kinase C (PKC)-dependent endocytosis of DAT has been observed in various mammalian cell lines (5–9). Phorbol esters also reduce dopamine uptake capacity, DAT transport-associated currents and capacitance measurements in Xenopus oocytes expressing the human DAT (10). Direct evidence for regulated DAT trafficking in neuronal cells is limited to data obtained in vitro using rat brain synaptosomes (11). Down-regulation of surface DAT and the accumulation of DAT in endosomes can be also triggered by amphetamine and other DAT substrates in non-neuronal cells and synaptosome preparations (9,11–14). In all cases, it is unclear whether regulated changes of DAT distribution in the cell are due to acceleration of DAT internalization or slowing down of DAT recycling from endosomes. The first attempt to distinguish between these two possibilities suggested that PKC activity affects both steps of endocytic trafficking of DAT expressed in PC12 cells (15).

Many integral membrane proteins are constitutively internalized and recycled. Such membrane recycling is an essential component of the general process of membrane homeostasis. The observation of DAT localization in endosomes of dopaminergic neurons is an indirect indication of constitutive DAT cycling in neurons (16). Different amounts of DAT are found in endosomes under steady-state growth conditions in various non-neuronal expression systems, suggestive of different rates of constitutive endocytic cycling of DAT in these systems (5,6). The model is emerging whereby DAT is constitutively internalized and recycled (15). Thus activation of PKC or other signaling processes may shift DAT distribution towards endosomes or the plasma membrane by accelerating or slowing down DAT internalization or recycling, or by affecting both of these processes.

The molecular mechanisms of DAT internalization are unknown. Overexpression of dominant-negative mutants of dynamin blocked phorbol ester- and amphetamine-induced DAT down-regulation in MDCK and HEK293 cells, suggesting that regulated internalization of DAT involves clathrin-coated pits (5,12). However, the same mutants of dynamin effectively inhibit caveolae- and lipid raft-mediated, clathrin-independent endocytosis (reviewed in (17)). Moreover, the PKC-induced endocytosis of nor-epinephrine transporter, the closest sequence homolog of DAT, was recently shown in rat placental trophoblasts to be mediated by lipid rafts (18). Serotonin transporter has also been found to be associated with lipid rafts, and this association has been implicated in the regulation of serotonin transporter trafficking (19). Thus, mechanisms of PKC-induced endocytosis of DAT must be analyzed using more specific experimental approaches other than overexpression of dynamin mutants.

Importantly, whereas certain clues are available regarding possible mechanisms of stimulus-induced endocytosis of DAT, the pathways and mechanisms of constitutive DAT internalization are not known. Constitutive internalization of other plasma membrane neurotransmitter transporters has been investigated (20,21). In the case of the GABA transporter 1 (GAT1), constitutive internalization is inhibited by the K44A dynamin mutant (21).

Hence, we examined the internalization pathways of human DAT stably expressed in porcine aortic endothelial (PAE) cells using small interfering RNAs (siRNAs) targeted to clathrin-coated pit proteins. The siRNA analysis demonstrated the essential role of the clathrin-mediated pathway in both PKC-induced internalization and PKC-independent, constitutive internalization of DAT. Mutagenesis revealed a potential role of carboxyl- but not amino-terminal portions of DAT molecule in internalization.

Results

Recycling inhibitor monensin causes accumulation of DAT in endosomes

We have previously characterized PAE cell lines stably transfected with wild-type human DAT, cyan (CFP)- or yellow fluorescent protein (YFP)-tagged DAT (9). These cell lines were used here to study the mechanisms of constitutive and regulated endocytosis of DAT. Under steady-state growth conditions in serum-containing medium, heterologously expressed DAT is mainly located in the plasma membrane of these cells, but a small pool of DAT is present in small intracellular vesicles and in the Golgi area ((9), see Figures 1 and 2A). Based on this subcellular distribution of DAT, the constitutive internalization of DAT in PAE cells must be substantially slower than the constitutive recycling of internalized DAT back to the plasma membrane.

Figure 1.

Figure 1.

    Biotinylation of live cells causes accumulation of DAT in early endosomes. PAE cells expressing CFP-DAT grown on coverslips were biotinylated using 1 mg/mL sulfo-NHS-SS-biotin at 4 °C as described in Materials and Methods, or incubated under the same conditions in PBS in the absence of sulfo-NHS-SS-biotin. The cells were then incubated for 30 min at 37 °C in the medium containing Tfn-TR and fixed. A z-stack of optical sections was acquired through CFP (green) and CY3 (red) filter channels. Single optical sections through the middle of the cells are shown. ‘Yellow’ in the merged images signifies colocalization of CFP (DAT) and TexasRed (Tfn-TR) fluorescence. Bars, 10 μm.

    Figure 2.

    Figure 2.

      Effects of monensin and PMA on the subcellular distribution of DAT. PAE/CFP-DAT cells grown on coverslips, 12-well or 24-well dishes were incubated with vehicle DMSO (0.01%), monensin (Mon, 20 μm), PMA (1 μm) or monensin (20 μm) and PMA (1 μm) together for 0–30 min at 37 °C. A) Cells on coverslips were fixed and the images were acquired through the CFP filter channel. Bars, 10 μm. B) Cells in 24-well dishes were subjected to surface biotinylation as described in Materials and Methods. CFP-DAT was detected in NeutrAvidin precipitates by Western blotting using rat anti-DAT. Each lane corresponds to the same number of cells and the same amount of total CFP-DAT as tested by blotting aliquots of the cell lysates equal to those taken for NeutrAvidin precipitation (not shown). C) Time course of the amount of CFP-DAT or YFP-DAT present at the cell surface in cells treated with Vehicle (DMSO), monensin (Mon), PMA or both monensin and PMA. The amount of surface (biotinylated) DAT was normalized to the total amount of DAT in the corresponding aliquot of cell lysate, and expressed as the percent of the amount of surface DAT at time point zero. Each point represents an averaged value (± SEM) obtained from five experiments described in B with PAE/CFP-DAT and PAE/YFP-DAT cells. D) Specific [3H]dopamine uptake was measured in cells grown in 12-well dishes and incubated for 30 min at 37 °C with DMSO, monensin and PMA. The data represent averaged values (± SEM) from three independent experiments.

      Recently, to measure the rates of constitutive internalization of DAT in PC12 cells, Loder & Melikian (15) used a reversible biotinylation assay, in which surface biotinylation of living cells was followed by the incubation of cells at 37 °C and stripping of biotin from noninternalized plasma membrane proteins, including DAT. However, we unexpectedly found that this type of biotinylation procedure (Procedure #2, see Materials and Methods) caused significant accumulation of DAT in endosomes labeled with transferrin conjugated with Texas Red (Tfn-TR), a marker of early and recycling endosomal compartments of PAE cells (Figure 1). Therefore, this assay could not be used to study DAT internalization.

      Another way to monitor constitutive endocytosis is to block constitutive recycling, which should lead to accumulation of internalized cargo in endosomes. To this end, we used an inhibitor of endosomal recycling, the carboxylic ionophore monensin (22–24). Monensin mediates proton movement across membranes and thereby raises the pH in acidic organelles, including endosomes, preventing transport of cargo from and through acidic endosomes (25). Figure 2A shows that incubation of PAE/CFP-DAT cells with monensin caused accumulation of CFP-DAT in the vesicular structures, consistent with constitutive endocytosis of CFP-DAT. This effect of monensin reached a maximum at concentrations of 15–20 μm, whereas general toxicity of monensin was observed at concentrations above 50 μm (data not shown). The accumulation of CFP-DAT in endosomes in monensin-treated cells was comparable to the accumulation of CFP-DAT in endosomes upon activation of PKC by phorbol 12-myristate 13-acetate (PMA) (Figure 2A). Simultaneous incubation of cells with both PMA and monensin resulted in the fastest accumulation of CFP-DAT in endosomes (Figure 2A). Essentially the same results were obtained in PAE cells expressing untagged DAT and YFP-DAT (data not shown).

      Surface biotinylation experiments (Procedure #1, Materials and Methods) revealed about 40% down-regulation of plasma membrane DAT after a 30-min incubation of PAE cells stably expressing YFP-DAT or CFP-DAT with monensin (Figures 2B, C). Similar monensin-induced loss of wild-type untagged DAT was observed (data not shown). PMA consistently caused greater DAT down-regulation than monensin, and treatment of cells with PMA and monensin together produced the largest reduction in the levels of plasma membrane DAT (∼ 75% down-regulation after a 30-min incubation) (Figures 2B,C).

      The loss of functional CFP-DAT from the cell surface of monensin- and PMA-treated cells was confirmed by measurements of [3H]dopamine uptake. In these experiments, specific [3H]dopamine uptake, defined as the difference between [3H]dopamine uptake in the absence and presence of 1 mm cocaine, was significantly reduced by monensin or PMA treatment (Figure 2D). Interestingly, monensin treatment resulted in a slightly larger reduction of the uptake than would have been expected from the microscopy and biotinylation experiments (Figure 2A–C). It is possible that monensin-induced changes in the pH gradient have a direct inhibitory effect on DAT function in addition to the effect of monensin on DAT recycling (26,27).

      Analysis of the steady-state localization of CFP-DAT revealed that a small amount of CFP-DAT was located in endosomes that contained Tfn-TR (Figure 3). This indirectly confirms the constitutive internalization of DAT. In monensin-treated cells, a significant pool of CFP-DAT was colocalized with Tfn-TR, suggesting that CFP-DAT was accumulated in the early and/or recycling endosomes when DAT recycling was inhibited (Figure 3). CFP-DAT was colocalized to a similar extent with Tfn-TR in monensin- and PMA-treated cells (Figure 3). Essentially the same colocalization with Tfn-TR was observed in PAE/YFP-DAT cells (data not shown).

      Figure 3.

      Figure 3.

        Co-localization of CFP-DAT and Tfn-TR in early and recycling endosomes of cells treated with monensin or PMA. PAE/CFP-DAT cells were incubated with Tfn-TR (5 μg/mL) and with either DMSO (0.01%), monensin (20 μm) or PMA (1 μm) for 30 min at 37 °C. The cells were then fixed and a z-stack of optical sections was acquired through CFP (green) and CY3 (red) filter channels. Single optical sections through the middle of the cells are shown. ‘Yellow’ in the merged images signifies colocalization of CFP (DAT) and TexasRed (Tfn-TR) fluorescence. Insets represent higher magnification images of the regions marked by the white rectangles. Bars, 10 μm.

        Interestingly, while PMA-dependent YFP-DAT endocytosis was completely blocked by pretreatment with the general PKC inhibitor bisindolylmaleimide I (BIM, 1 μm), monensin-induced accumulation of YFP-DAT in early endosomes, identified by the early endosomal antigen 1 protein (EEA.1), was PKC-independent (Figure 4A). Similar results were obtained in PAE/DAT and PAE/CFP-DAT cells (data not shown). Furthermore, subsequent incubation of PMA-treated cells with BIM caused the disappearance of endosomal and re-appearance of plasma membrane YFP-DAT in about 50% of cells in the population (Figure 4B). Presumably, upon inhibition of PKC by BIM, DAT internalization and recycling processes return back to constitutive kinetics. The explanation why BIM has an effect in only a limited number of cells in these experiments is unknown. Perhaps, in a certain population of cells treated with PMA, DAT is trapped in the endosomal compartments or subcompartments of sorting endosomes from which recycling is not efficient. Consistent with the partial effects of BIM on DAT localization in PMA-treated cells, surface biotinylation assays detected only a relatively small (20%) increase in the surface DAT levels after BIM treatment of PMA-activated cells (data not shown). Nevertheless, monensin completely abolished the BIM-dependent re-distribution of endosomal DAT to the plasma membrane in all PMA-treated cells as detected by fluorescence microscopy (Figure 4B). Since monensin-induced accumulation of DAT in endosomes was not affected by BIM (Figure 4A), the simplest interpretation of these experiments is that monensin does indeed inhibit recycling of internalized DAT.

        Figure 4.

        Figure 4.

          Effect of the PKC inhibitor BIM and monensin on DAT endocytosis and recycling.A) PAE/YFP-DAT cells were incubated without or with 1 μm BIM for 15 min at 37 °C and then in the same media with monensin (Mon, 20 μm) or PMA (1 μm) for 30 min at 37 °C. The cells were then fixed, permeabilized and stained with EEA.1 antibody as described in ‘Methods’. A z-stack of optical sections was acquired through FITC (YFP-DAT, green) and Cy3 (EEA.1, red) filter channels. Single optical sections through the middle of the cell are shown. ‘Yellow’ in the merged images signifies colocalization of CFP (DAT) and Cy3 (EEA.1) fluorescence. Insets represent higher magnification images of the regions marked by the white rectangles. Bar, 10 μm. B) PAE/YFP-DAT cell were treated with PMA (1 μm) or monensin (20 μm) (top row) for 30 min at 37 °C. The cells treated with PMA were then washed and incubated with BIM (1 μm) for 120 min in the absence or presence of monensin (20 μm) (bottom row). Images were obtained through the YFP filter channel. Bars, 10 μm.

          The data in Figures 2–4 suggest that treatment of cells with monensin results in accumulation of DAT in early and recycling endosomes, as judged by colocalization of DAT with Tfn-TR and EEA.1, due to shifting of the equilibrium between internalization and recycling towards the internalization process. Thus, DAT down-regulation from the cell surface and accumulation in endosomes in the presence of monensin are the result of constitutive internalization of DAT.

          Effect of depletion of coated pit proteins by siRNA on DAT endocytosis

          Establishing the assays to follow constitutive, as well as PKC-induced, endocytosis of DAT allowed examination of the mechanism of this endocytosis. In particular, we tested the role of the major internalization pathway, clathrin-mediated endocytosis. Recently, we developed a number of siRNAs that can efficiently down-regulate proteins involved in clathrin-mediated endocytosis in human cells (28). siRNAs targeted to the clathrin heavy chain (CHC) and dynamin II were the most effective inhibitors of clathrin-mediated endocytosis of Tfn and the epidermal growth factor (EGF) receptors (28). These siRNAs were therefore employed to examine the role of the clathrin internalization pathway in DAT endocytosis in PAE cells.

          Fluorescence microscopy analysis of PAE/CFP-DAT cells transfected with the siRNA targeted to CHC showed that greater than 95% of the cells had a very low amount of, if any, detectable CHC. No significant accumulation of CFP-DAT in endosomes was observed in CHC-depleted cells treated with either PMA or monensin, indicative of the efficient inhibition of DAT endocytosis by the CHC siRNA (Figure 5). Although the dynamin II antibody did not recognize porcine dynamin II in immunofluorescence staining in PAE cells, inspection of the large number of PAE/CFP-DAT cell cultures transfected with dynamin II siRNA revealed that almost all cells lacked endosomal DAT (Figure 5). These data demonstrate that internalization of CFP-DAT in the presence of PMA or monensin requires major clathrin-coated pit proteins.

          Figure 5.

          Figure 5.

            Effects of CHC and dynamin II siRNAs on DAT endocytosis. PAE/CFP-DAT cells were mock-transfected or transfected with siRNAs to CHC or dynamin II (Dyn II) as described in Materials and Methods. The cells were incubated with DMSO, monensin (Mon) or PMA for 30 min at 37 °C as in Figure 2. The cells were then fixed, permeabilized and stained with the CHC antibodies as described in Materials and Methods. A z-stack of optical sections was acquired through CFP (green) and Cy3 (red) filter channels. Single optical sections through the middle of the cell are shown. Small images below the images of mock-transfected cells rep- resent higher magnification images of the regions marked by the white rectangles demonstrating examples of colocalization of CFP-DAT and CHC. Bars, 10 μm.

            In mock-transfected cells a number of CFP-DAT dots were colocalized with CHC-containing dots, presumably clathrin-coated pits or vesicles (Figure 5). Co-localization of CFP-DAT with another marker of plasma membrane coated pits, a YFP-tagged β2-subunit of clathrin adaptor complex AP-2, was observed in living cells (data not shown). The presence of DAT in coated pit/vesicles is consistent with the requirement of clathrin for DAT internalization revealed in the siRNA experiments. However, only a small pool of CFP-DAT was present in clathrin-positive structures in cells under all conditions due to the short residence time of internalizing cargo in the pits.

            A surface biotinylation assay was used to examine the effect of CHC and dynamin II siRNAs on DAT levels in the plasma membrane under conditions of PMA or monensin treatment. Figure 6 shows that while a 30-min incubation of PAE/CFP-DAT cells with PMA or monensin resulted in, correspondingly, 45% or 40% reduction in the amount of surface CFP-DAT, no down-regulation of plasma membrane CFP-DAT was observed in cells transfected with the CHC siRNA. Similarly, transfection of cells with the dynamin II siRNA blocked the loss of cell surface DAT induced by PMA or monensin. The data from Figures 5 and 6 demonstrated that the pathway of endocytosis through clathrin-coated pits plays an essential role during both constitutive and PKC-dependent internalization of DAT.

            Figure 6.

            Figure 6.

              Effects of CHC and dynamin II siRNAs on the surface expression of DAT. PAE/CFP-DAT cells grown on 24-well dishes were mock-transfected or transfected with siRNAs to CHC or dynamin II as described in Materials and Methods. The cells were incubated with DMSO, monensin (Mon) or PMA for 30 min at 37 °C as in Figure 2 and subjected to surface biotinylation as described in Materials and Methods. CFP-DAT was detected in NeutrAvidin precipitates by Western blotting using rat anti-DAT. Each lane in A corresponds to the same number of cells and the same amount of total CFP-DAT as determined by immunoprecipitation of CFP-DAT from aliquots of cell lysates equal to those taken for NeutrAvidin precipitation (not shown). The cell lysates were blotted with antibodies to CHC or dynamin II, as well as to the unrelated protein PP1, to demonstrate the depletion of CHC and dynamin II in cells transfected with corresponding siRNAs (B). C) Amounts of CFP-DAT present at the cell surface in cells treated with DMSO, monensin or PMA (± SEM) averaged from three experiments described in A. The amount of surface (biotinylated) CFP-DAT was normalized to the total amount of CFP-DAT in a corresponding aliquot of the cell lysate, and expressed as percent to the DMSO control.

              Effect of cholesterol inhibitors on DAT endocytosis

              Recently, it has been shown that PKC-dependent internalization of norepinephrine transporter can be blocked by cholesterol-disrupting compounds, indicative of the role of lipid raft-mediated endocytosis (18). To test whether DAT endocytosis also requires cholesterol-rich membrane microdomains, PAE/CFP-DAT cells were preincubated with the cholesterol-disrupting compounds filipin and nystatin, known to inhibit lipid raft- and caveolae-dependent endocytosis (17,29–31). The cells were then incubated with PMA or monensin for 30 min in the presence of inhibitors. As shown in Figure 7, neither filipin nor nystatin, at concentrations known to inhibit raft-mediated endocytosis (18), affected endocytosis of CFP-DAT. Under the same conditions, the endocytosis of Tfn-TR was not affected (data not shown). Concentrations of filipin and nystatin higher than 5 μg/mL and 15 μg/mL, respectively, resulted in general toxic effects and cell detachment. These data indicate that DAT does not use lipid raft-dependent pathways of endocytosis in PAE cells.

              Figure 7.

              Figure 7.

                Effects of cholesterol-disrupting drugs on DAT endocytosis. PAE/CFP-DAT cells were incubated without or with nystatin (Nys, 15 μm) or filipin (Fil, 5 μm) for 45 min at 37 °C and then in the same media with DMSO, monensin (Mon) or PMA for 30 min at 37 °C as in Figure 2. Cells were then fixed and the images were acquired through the CFP filter channel. Bars, 10 μm.

                Role of amino- and carboxyl-termini of DAT in DAT internalization

                Since DAT is internalized via clathrin-coated pits, there must be amino acid sequences in the DAT molecule that mediate its interaction with coated pit proteins. Such sequences are typically recognized by adaptor proteins in linear conformation and positioned at some distance from the transmembrane domains of the cargo. Therefore, internalization signals are likely to be situated in the cytoplasmic amino- and carboxyl-terminal tails, rather than in small intracellular loops of DAT.

                To test for the role of the amino-terminus in internalization, a DAT mutant lacking the entire amino-terminus of DAT (first 65 residues) (YFP-ΔN-DAT) was stably expressed in PAE cells. Truncation of the amino-terminus results in the retention of DAT in the endoplasmic reticulum (ER) (32). However, if YFP is attached to this truncation mutant (as in YFP-ΔN-DAT), the mutant normally exits ER and is efficiently delivered to the plasma membrane (9). Treatment of PAE/YFP-ΔN-DAT cells with PMA or monensin led to dramatic accumulation of this truncated transporter in endosomes (Figure 8). Both PMA and monensin caused down-regulation of surface YFP-ΔN-DAT as measured with the surface biotinylation assay (Figure 8). These data indicate that the amino-terminus of DAT is not essential for internalization.

                Figure 8.

                Figure 8.

                  Amino-terminus of DAT is not essential for internalization. PAE cells stably expressing YFP-ΔN-DAT were treated with DMSO, monensin or PMA as described in Figure 3. Cells grown on coverslips were analyzed by fluorescence microscopy. The images were acquired through the YFP filter channel. Bars, 10 μm. The cells grown on 12-well dishes were surface-biotinylated as described in Figure 2. YFP-ΔN-DAT was detected in NeutrAvidin precipitates by Western blotting using anti-GFP. Each lane corresponds to the same number of cells and the same amount of total YFP-ΔN-DAT as tested by blotting aliquots of the cell lysates equal to those taken for NeutrAvidin precipitation (not shown).

                  Similar experiments could not, however, be performed with DATs lacking the carboxyl-terminal tail because large deletions in this portion of the molecule produce DAT mutants that are retained in the ER (33). To circumvent this problem we generated a chimeric protein consisting of the extracellular and transmembrane domain of the EGF receptor linked to the carboxyl-terminal tail of DAT (residues 578–620) (Figure 9A). Such an N-EGFR-DAT-C chimera has been previously shown to be well expressed at the cell surface in a manner independent of the ER export signals in the carboxyl-terminal tail of DAT (33). In experiments where cells were incubated with EGF receptor antibody at 37 °C or immunostained after cell permeabilization, a large pool of the chimera was also located in intracellular vesicles that also contained YFP-EEA.1 (Figure 9B). In contrast, the chimeric protein lacking the DAT carboxyl-terminus (N-EGFR) was exclusively localized at the cell surface (Figure 9B). The above suggests that the carboxyl-terminus of DAT alone is sufficient to support constitutive endocytosis of the chimera.

                  Figure 9.

                  Figure 9.

                    EGF receptor extracellular domain – DAT carboxyl-terminus chimeras demonstrate the potential importance of the carboxyl-terminus for DAT endocytosis.A) The schematic structure of N-EGFR-DAT-C chimera. Numbering of residues is according to human DAT. B) PAE cells were transiently transfected with YFP-EEA.1 and either N-EGFR-DAT-C, N-EGFR (lacking an entire DAT sequence 578–620), N-EGFR-DAT-CΔ598-617 (deletion of residues 598–617 of DAT) or N-EGFR-DAT-CΔ587-617 (deletion of residues 587–617 of DAT). The cells were incubated with antibody 528 to EGF receptor for 30 min at 37 °C, fixed, permeabilized and stained with Cy3-labeled secondary antibody. Images were taken through Cy3 and YFP filter channels and deconvoluted as in Figure 3. Merge represents overlap of YFP and Cy3 image. Bar, 10 μm. C) PAE cells were transfected with CHC siRNA or mock-transfected. The cells were additionally transiently transfected with N-EGFR-DAT-C. The cells were incubated for 30 min with antibody 528 and 5 μg/mL Tfn-TR at 37 °C then, fixed, permeabilized and stained with secondary anti-mouse antibody labeled with fluorescein. The cells were imaged through Cy3 (Tfn-TR) and FITC (Ab528) filter channels as described in B. Bars, 10 μm.

                    To test whether endocytosis of the N-EGFR-DAT-C chimera is clathrin-dependent, cells transiently expressing this chimera were transfected with clathrin siRNA. In addition, cells were incubated with Tfn-TR internalized through coated pits to control for the inhibitory effects of siRNA on clathrin-mediated endocytosis. As shown in Figure 9C, clathrin siRNA dramatically inhibited Tfn-TR endocytosis. In the same cells the N-EGFR-DAT-C chimera was mostly localized at the plasma membrane, indicative of the block of constitutive endocytosis of the chimera (Figure 9C). We concluded from these experiments that N-EGFR-DAT-C is internalized through the clathrin-dependent pathway.

                    Because the extracellular domain of the EGF receptor anchored to the membrane is normally transported to the plasma membrane, the structure-function analysis of the DAT carboxyl-terminus was performed using the N-EGFR-DAT-C chimera. Truncation of the last 10 residues in N-EGFR-DAT-C did not affect constitutive endocytosis of the chimera (data not shown). Deletion of residues 598–617 also did not result in inhibition of the accumulation of N-EGFR-DAT-C in endosomes (Figure 9B). However, deletion of residues 587–617 produced a mutant chimera that was localized at the cell surface (Figure 9B). Residues 587–597 would therefore seem to be important for internalization of the N-EGFR-DAT-C chimera. The importance of these residues in the context of wild-type DAT could not, however, be tested because similar deletions and truncations of the carboxyl terminus produce ER-export-deficient DAT mutants (33). Single/multiple alanine substitutions of residues 587–597 and surrounding residues in the wild-type DAT either did not affect DAT internalization or led to the ER retention of the mutants (see Table in the Supplemental Materials, http://www.traffic.dk/supp_material.asp). It is possible that residues 587–597 are important for the overall folding of the carboxyl-terminus necessary for interaction with endocytic machinery or represent a part of an extended internalization signal motif in this part of the DAT molecule.

                    Discussion

                    In the present study we analyzed endocytic trafficking of DAT in PAE cells. Because we observed an effect of biotinylation on DAT localization in living cells (Figure 1), an internalization assay based on biotinylation of living cells (Procedure #2, see Materials and Methods) could not be used to study constitutive endocytosis of DAT. This finding suggests that extracellular lysine residues in DAT may be involved in regulation of DAT trafficking; for example, by interfering with DAT interactions at the plasma membrane or with DAT recycling from endosomes. These data also suggest that results obtained using biotinylation assays with subsequent chase incubations must be interpreted with caution. Hence, to directly visualize constitutive internalization of DAT, the cells were here treated with monensin, a carboxylic ionophore known to block recycling of many types of receptors from endosomes to the plasma membrane. Monensin increases the pH in endosomes and other acidic compartments in the cell, leading to a ‘traffic jam’ in these compartments. Obviously, monensin may affect processes other than endosomal sorting, as well. However, within the time frame of our internalization assays (15–30 min), it is important to note that no data indicated that monensin can up-regulate the internalization step of endocytic trafficking. Therefore, the parallel accumulation of cargo in endosomes and decrease in the plasma membrane pool of cargo in monensin-treated cells is due to unaltered, continuous internalization of cargo in the absence of recycling.

                    In our experiments, inhibition of recycling by monensin caused an accumulation of DAT in endosomes that was clearly visible in the majority of cells in the population. Similar extents of accumulation of endosomal DAT were observed for YFP-DAT, CFP-DAT and untagged DAT expressed in PAE, immortalized neuronal 1RB3AN27 and other cell types (data not shown). A 30-min monensin treatment also caused significant (by 40%) down-regulation of surface DAT (Figure 2). Analysis of the time-courses of DAT loss from the cell surface in the presence of monensin yields a rate for constitutive internalization of DAT in the range of 1.3–1.5%/min. This rate may, however, be an underestimation as recycling may be incompletely blocked by monensin. In particular, monensin can be ineffective in mildly acidic early endosomes. Although similar rates for the loss of DAT from the cell surface of PC12 cells were observed after incubating the cells at 18 °C, which slows down recycling (15), results obtained at 18 °C can be difficult to interpret. This temperature dramatically inhibits the slow recycling pathway through the recycling compartment but does not significantly affect the rapid recycling pathway from early endosomes (34). In addition, internalization rates are a magnitude slower at 18 °C than at 37 °C. In fact, no accumulation of CFP-DAT in endosomes was detected after incubation of PAE cells, with or without PMA, at 20 °C for 30 min (data not shown).

                    PMA caused a more pronounced accumulation of endosomal DAT and a faster down-regulation of surface DAT (2.5–3.0%/min), as compared with the effects of the monensin treatment. Monensin treatment further increased the rate of DAT loss from the surface of cells treated with PMA (about 60% loss in 15 min equals ∼ 4%/min). These data are consistent with the model in which PMA accelerates internalization of DAT and may also partially inhibit recycling of DAT. When monensin is added together with PMA, the acceleration of the internalization is combined with a block of recycling by monensin, thus leading to a larger loss of DAT from the cell surface than in cells treated with PMA alone.

                    The experiments described in Figures 2–4 validated the usefulness of the monensin treatment protocol as an approach to analyze the mechanisms of constitutive internalization of DAT. The mechanisms of DAT internalization have only recently begun to be elucidated. Originally, Daniels & Amara (5) demonstrated that PMA-induced internalization of DAT in MDCK cells was blocked by overexpression of the K44E mutant of dynamin. Likewise, it has been demonstrated that constitutive internalization of GAT1 is dynamin-dependent (21). K44E, K44A and other mutants of dynamin have dominant-negative effects on clathrin-mediated endocytosis by interfering with the complete constriction of coated pits and fission of coated vesicles (35). More recently, however, it has been demonstrated that dynamin mutants also block caveolae-mediated endocytosis (see for example (36–38)). Moreover, it has been found that overexpression of dynamin mutants inhibits activation of extracellular response kinase ERK1/2 by opioid receptors in a manner independent of receptor endocytosis (39). Therefore, given the pleiotropic effects of overexpression of dominant-negative dynamin mutants, interpretation of the results of these experiments is not unequivocal.

                    As new approaches for the direct analysis of the importance of endogenous proteins, such as protein knock-down using RNAi, became available, we developed a number of siRNAs that effectively block clathrin-dependent endocytosis of EGF and Tfn receptors (28). The most effective siRNA in blocking endocytosis of the Tfn receptor in our experiments and other studies in human cells was the siRNA to human CHC (28,40). Transfection of porcine (PAE) cells with this siRNA also led to dramatic inhibition of Tfn-TR endocytosis (Figure 9C), although the extent of CHC depletion was slightly less than in human cells (Figure 6). We therefore have extended the application of the RNAi approach to the analysis of another type of endocytic cargo, namely DAT. The data presented in Figures 5 and 6 demonstrate that knock-down of CHC effectively blocks both constitutive (PKC-independent) and PMA-induced (PKC-dependent) endocytosis of DAT. Similar inhibition of DAT endocytosis was observed when dynamin II was knocked-down in PAE cells. On the other hand, we did not observe any effects of the cholesterol-disrupting drugs filipin and nystatin, which are known to inhibit caveolae and lipid-raft-mediated endocytosis (Figure 7). Altogether the data suggest that the clathrin-dependent pathway is the main route of both constitutive and PMA-induced internalization of DAT in PAE cells.

                    The molecular mechanisms of DAT internalization through coated pits remain unknown. For instance, internalization sequences have not been mapped in DAT or in other plasma membrane neurotransmitter transporters. Mutations of Tyr335 and Tyr575 or Leu440–441 that can potentially serve as internalization signals of the YxxΘ or di-leucine type, respectively, in the DAT molecule did not affect PKC-dependent endocytosis of DAT (8). Likewise, we performed systematic mutagenesis of tyrosine and leucine residues, as well as lysine residues that can be monoubiquitinylated and thus serve as internalization signals. However, this mutagenesis did not reveal any such internalization signals (Table, Supplemental Materials). In the search for internalization sequences, we initially focused on the amino- and carboxyl-termini of DAT molecules. While deletion of the amino-terminus did not inhibit either constitutive or PMA-induced DAT endocytosis, analysis of the carboxyl-terminus revealed the potential role of residues 587–597 (FREKLAYAIAP) in DAT internalization. These experiments utilized EGFR-DAT chimeric proteins (Figure 9) because the corresponding mutations in the full-length DAT inhibited transport of newly synthesized DAT to the plasma membrane. However, single or multiple site-point mutations within this region either produced no endocytic phenotype or resulted in ER export-deficient DATs (Table, Supplemental Materials). Therefore, although it is possible that the 587–597 residues represent a part of the internalization signal, it is also possible that the deletion of the last 34 residues of the EGFR-DAT chimera led to conformational changes in the DAT carboxyl-terminus that resulted in reduced internalization of the chimera. Our data suggest that the recruitment of DAT into coated pits may be controlled by unconventional and unique mechanisms, and/or may involve an additional adaptor mediating DAT interactions with coated pit proteins.

                    Materials and Methods

                    Antibodies

                    Monoclonal rat antibody against the amino terminus of human DAT was purchased from Chemicon International, Inc. (Temecula, CA); polyclonal rabbit antibody to green fluorescent protein (GFP) was from Abcam Ltd. (Cambridge, UK); monoclonal EEA.1 and PP1 antibodies were from BD Transduction Laboratories (Los Angeles, CA); mouse monoclonal antibody X.22 to Clathrin heavy chain from American Type Cell Culture Collection (Manassas, VA); monoclonal antibody to GFP from Zymed; dynamin II polyclonal antibody was from Affinity Bioreagents, Inc. (Golden, CO). Monoclonal antibody 528 to the EGF receptor was from American Type Cell Culture Collection.

                    Plasmids

                    A construct consisting of the extracellular and transmembrane domain of the EGF receptor fused to the carboxyl-terminus of DAT (N-EGFR-DAT-C) was described previously (33). All mutations of DAT, as well as truncation and deletions of N-EGFR-DAT-C were made using Stratagene quick-change mutagenesis kit as described (33). YFP fusion of EEA.1 fragment (YFP-EEA.1) has been described previously (41). YFP-ΔN-DAT was described previously (9).

                    Cell culture

                    PAE cells stably expressing wild-type human DAT, CFP-DAT or YFP-DAT were described previously (9). The cells were grown in F12 medium containing 10% fetal bovine serum (FBS) and antibiotics. DNA transfections were performed using Effectine (Qiagen, Hilden, Germany).

                    siRNA Transfection

                    CHC and dynamin II siRNAs were described previously (28). siRNA duplexes were resuspended in 1X siRNA Universal buffer (Dharmacon, Inc., Lafayette, CO) to 20 μm prior to transfection. PAE cells in 12-well plates (50–60% confluent; 1 mL F12/FBS per well) were transfected with 4 μL siRNA duplex and 2 μL Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA) in 95 μL Opti-MEM medium according to the manufacture's recommendations. Cells were incubated in OptiMEM/siRNA overnight and then placed into normal culture medium. The experiments were performed 3–4 days after transfection.

                    Surface biotinylation and immunoprecipitation

                    Two procedures of cell biotinylation were used. In Procedure #1 (surface DAT down-regulation), the cells expressing DAT proteins were grown in 12-well or 24-well plates, incubated in binding medium (F12, 0.1% bovine serum albumin (BSA)) under appropriate conditions and biotinylated as described previously (9). Briefly, the cells were washed with cold phosphate- buffered saline (PBS) containing 0.1 mm CaCl2 and 1 mm MgCl2 and incubated for 20 min on ice with 1 mg/mL sulfo-N-hydroxysuccinimidobiotin (EZ-LinkTM sulfo-NHS-biotin) (Pierce, Inc., Rockford, IL) in PBS, followed by a second incubation with fresh sulfo-NHS-biotin. After biotinylation, the cells were washed twice with cold PBS, incubated on ice with 0.1 m glycine in PBS for 20 min and washed with PBS again and then either incubated in binding medium (F12, 0.1% bovine serum albumin (BSA)) for 30 min at 37 °C or lysed. The cells were solubilized in lysis buffer (50 mm NaCl, 50 mm HEPES, pH 7.2, 10% glycerol, 1 mm EGTA, 1 mm EDTA, 1 mm phenylmethylsulfonyl fluoride, 10 μg/mL leupeptin, 10 μg/mL iodoacetamide, 10 μg/mL aprotinin, 1% Triton X-100, 0.1% SDS, 1% sodium deoxycholate, 10 mm Tris-HCl (pH 6.8) at 4 °C. The lysates were cleared by centrifugation for 10 min at 16 000 × g. Half of the total lysates were directly resolved by SDS-PAGE or subjected to immunoprecipitation with either rat anti-DAT antibodies or polyclonal anti-GFP antibodies to determine the total amount of DAT present in each well. Another half of the lysate was used for precipitation of the biotinylated proteins with NeutrAvidinTM beads (Pierce). The beads were washed five times with lysis buffer, and denatured by heating the beads in sample buffer at 95 °C for 5 min. In some experiments supernatants from the NeutrAvidin precipitation were further subjected to immunoprecipitation with polyclonal anti-GFP to determine the amount of nonbiotinylated DAT. The NeutrAvidin beads, aliquots of lysates and immunoprecipitates were subjected to electrophoresis on 7.5% SDS-PAGE, and the proteins were transferred to nitrocellulose membrane. Western blotting was performed with monoclonal rat or mouse antibodies, followed by corresponding secondary antibodies conjugated with horseradish peroxidase, and bands were detected using the enhanced chemiluminescence kit from Pierce. Several X-ray films were analyzed by densitometry to determine the linear range of the chemiluminescence signals, and the quantifications were performed using NIH Image software.

                    In Procedure #2 (postbiotinylation endocytosis of DAT), the cells expressing CFP-DAT and grown on glass coverslips were biotinylated using sulfo-N-hydroxysuccinimido-SS-biotin (Pierce) at 4 °C as described above in Procedure #1, washed twice with cold PBS, incubated with binding medium (F12 medium containing 0.1% BSA) at 4 °C for 15 min and then further incubated for 30 min at 37 °C in binding medium, washed with ice-cold CMF-PBS and fixed with 4% paraformaldehyde (Electron Microscopy Sciences, Ft. Washington, PA) for 15 min at room temperature.

                    Fluorescence microscopy

                    The cells grown on glass coverslips were incubated with DMSO, PMA, monensin and/or in some experiments with Tfn-TR (Molecular Probes, Inc., Eugene, OR), washed with Ca2+, Mg2+-free PBS (CMF-PBS), fixed with freshly prepared 4% paraformaldehyde for 15 min at room temperature and either taken for imaging directly or, for immunofluorescence staining, mildly permeabilized using a 3-min incubation in CMF-PBS containing 0.1% Triton X-100 and 0.5% bovine serum albumin at room temperature. Cells were then incubated in CMF-PBS containing 0.5% bovine serum albumin at room temperature for 1 h with primary antibodies, and subsequently incubated for 30 min with secondary antibodies labeled with FITC, CY3 or CY5 (Jackson Laboratories, West Grove, PA). Both primary and secondary antibody solutions were precleared by centrifugation at 100 000 ×g for 20 min. After staining, the coverslips were mounted in Mowiol (Calbiochem).

                    To obtain high resolution three-dimensional images of the cells, the fluorescence imaging workstation consisted of a Nikon inverted microscope equipped with a 100× oil immersion objective lens, cooled CCD SensiCam QE 16 MHz (Cooke, Germany), z-step motor, dual filter wheels and a Xenon 175 W light source, all controlled by slidebook 4 software (Intelligent Imaging Innovation, Denver, CO). Typically, 15–25 serial two-dimensional images were recorded at 100–200 nm intervals. A Z-stack of images obtained was deconvoluted using a modification of the constrained iteration method (Gaussian noise smoothing). Final arrangement of all images was performed using Adobe Photoshop.

                    Dopamine uptake assays

                    The cells were grown in 12-well plates for 2–3 days. The cells were rinsed at 4 °C and then assayed in Krebs-Ringer HEPES buffer (KRH; 120 mm NaCl, 4.7 mm KCl, 2.2 mm CaCl2, 1.2 mm Mg SO4, 1.2 mm KH2PO4, 10 mm glucose, 10 mm HEPES, pH 7.4) supplemented with 10 μm pargyline, 10 μm ascorbic acid, and 10 μm catechol. Assays (1 mL) included 50 nm[3H]dopamine (Perkin Elmers Life Sciences, Boston, MA). The specific activity was 39.3 Ci/mmol. Nonspecific [3H]dopamine accumulation was determined in the presence of 1 mm cocaine HCl (NIDA/RTI International, Research Triangle Park, NC). After 10 min of incubation at 22 °C, uptake was terminated by quickly washing the cells three times with 1 mL of ice-cold KRH. Cells were then solubilized in 0.5 mL of 3% trichloroacetic acid for 60 min with gentle shaking. Accumulated [3H]dopamine was determined by liquid scintillation counting.

                    Acknowledgments

                    This work was supported by the NIH grants DA014204 (T.S. and A.S.), DA016860 (B.R.H.) and DA015050 (N.R.Z.).

                    Ancillary