Activation, internalization, and recycling of the serotonin 2A receptor ...

18 downloads 0 Views 1MB Size Report
Oct 10, 2006 - for the 5-HT2A receptor, acts as a partial-efficacy agonist. At ... Dopamine and serotonin (5-HT) have been implicated in a number of ...
Activation, internalization, and recycling of the serotonin 2A receptor by dopamine Samarjit Bhattacharyya*†, Ishier Raote*, Aditi Bhattacharya*, Ricardo Miledi‡§, and Mitradas M. Panicker*§ *National Centre for Biological Sciences, Tata Institute of Fundamental Research, University of Agricultural Sciences–Gandhi Krishi Vignana Kendra Campus, Bellary Road, Bangalore-560065, Karnataka, India; and ‡Department of Neurobiology and Behavior, University of California, Irvine, CA 92697 Contributed by Ricardo Miledi, August 1, 2006

Serotonergic and dopaminergic systems, and their functional interactions, have been implicated in the pathophysiology of various CNS disorders. Here, we use recombinant serotonin (5-HT) 2A (5-HT2A) receptors to further investigate direct interactions between dopamine and 5-HT receptors. Previous studies in Xenopus oocytes showed that dopamine, although not the cognate ligand for the 5-HT2A receptor, acts as a partial-efficacy agonist. At micromolar concentrations, dopamine also acts as a partial-efficacy agonist on 5-HT2A receptors in HEK293 cells. Like 5-HT, dopamine also induces receptor-internalization in these cells, although at significantly higher concentrations than 5-HT. Interestingly, if the receptors are first sensitized or ‘‘primed’’ by subthreshold concentrations of 5-HT, then dopamine-induced internalization occurs at concentrations ⬇10-fold lower than when dopamine is used alone. Furthermore, unlike 5-HT-mediated internalization, dopamine-mediated receptor internalization, alone, or after sensitization by 5-HT, does not depend on PKC. Dopamine-internalized receptors recycle to the surface at rates similar to those of 5-HT-internalized receptors. Our results suggest a previously uncharacterized role for dopamine in the direct activation and internalization of 5-HT2A receptors that may have clinical relevance to the function of serotonergic systems in anxiety, depression, and schizophrenia and also to the treatment of these disorders. receptor priming 兩 receptor recycling

D

opamine and serotonin (5-HT) have been implicated in a number of pathological psychiatric disorders, including depression, anxiety, bipolar disorder, schizophrenia, and drug abuse (1–6). Many antipsychotic drugs bind to both dopamine and 5-HT receptors (4, 7), and several studies have indicated that 5-HT, acting through its receptors, can modulate dopamine function (8–14). In particular, the 5-HT2A receptor has been shown to be present on dopaminergic neurons in various regions of the brain; e.g., the ventral tegmental area in rats and humans (15, 16). These findings suggest that dopamine and 5-HT, interacting through their respective receptors and signal transduction pathways, may modulate each other’s response. Although dopamine is not the cognate ligand for 5-HT receptors, it has been shown to directly activate 5-HT1A, 5-HT2C, and 5-HT3 receptors (17, 18). Dopamine acts as a partial-efficacy agonist at these receptors, i.e., it activates these receptors to a lesser extent than does 5-HT. Although dopamine activates 5-HT receptors directly, the potency and efficacy of its binding vary significantly among subtypes. Early work showed that dopamine acts as a partial-efficacy agonist at the rat 5-HT2A receptor expressed in Xenopus oocytes (17), although micromolar concentrations of dopamine are required to activate the 5-HT2A receptor. More recent studies on the interactions of dopamine with 5-HT1A, 5-HT2C, and 5-HT3 receptors indicate that the affinity of the ligand to the receptor does not correlate with the efficacy of the ligand for activation (18). In addition, a number of dopamine-receptor antagonists (atypical antipsychotic drugs) were also found to internalize the 5-HT2A receptor, some at nanomolar concentrations. 15248 –15253 兩 PNAS 兩 October 10, 2006 兩 vol. 103 兩 no. 41

To date, there have been no detailed studies on the effects of dopamine on 5-HT2A receptors, studies which are important to more completely understand the potential for direct cross-talk between serotonergic and dopaminergic systems. Because 5-HT2A receptors localize to some dopaminergic neurons, and a number of clinically used drugs bind to both 5-HT and dopamine receptors, modulation of 5-HT2A-receptor responses by dopamine could play an important role in the CNS. The current study uses mammalian cell lines to characterize receptor activation and trafficking using a full-length rat 5-HT2A receptor tagged at the C terminus with EGFP (SR2-GFP receptor), stably expressed in HEK293 cells. An earlier study using this cell line showed that the tagged receptor is functional, easily visualized, and can be used to study trafficking and activation of the receptor (19). Results Dopamine Activates the Rat 5-HT2A Receptor in HEK293 Cells. In

Xenopus oocytes, rat 5-HT2A receptors stimulate the phospholipase C-IP3 pathway upon activation by dopamine (17). To determine whether dopamine activates rat 5-HT2A receptors expressed in mammalian cells, intracellular Ca2⫹ levels were monitored by using the Ca2⫹-sensitive Rhod2-AM dye in SB1 cells (i.e., HEK293 cells stably expressing SR2-GFP receptors) after exposure to dopamine. Application of 10 ␮M dopamine elicited an increase in intracellular Ca2⫹ levels in SB1 cells (Fig. 1). This increase in Ca2⫹ levels was not seen in untransfected HEK293 cells, indicating that the 5-HT2A receptor mediates the response. Concentrations of dopamine ⬍5 ␮M did not cause a detectable increase in intracellular Ca2⫹ (data not shown). The increase in the levels of Ca2⫹-dependent intracellular fluorescence upon application of 10 ␮M dopamine was significantly smaller than that produced by 10 ␮M 5-HT (Fig. 1C). This result is consistent with the reduced efficacy of activation of the 5-HT2A receptor by dopamine observed earlier in Xenopus oocytes (17). These experiments suggest that, at micromolar concentrations, dopamine is a partial-efficacy agonist also at rat 5-HT2A receptors in HEK293. Dopamine-Activated Rat 5-HT2A Receptors Internalize in HEK293 Cells.

Because dopamine activated the rat 5-HT2A receptor, we examined whether it would also induce receptor internalization, as does serotonin (17). Cycloheximide-treated SB1 cells were stimulated with different concentrations of dopamine, and the cells were imaged to check for the presence of an internalized pool of receptors. These receptors were observed to internalize at dopamine concentrations ⱖ5 ␮M (Fig. 2), whereas 5-HT proAuthor contributions: S.B., R.M., and M.M.P. designed research; S.B., I.R., and A.B. performed research; S.B., I.R., A.B., and M.M.P. analyzed data; and S.B., I.R., A.B., R.M., and M.M.P. wrote the paper. The authors declare no conflict of interest. Abbrreviation: 5-HT, serotonin. †Present

address: Department of Psychiatry and Behavioral Sciences, Stanford University, Stanford, CA 94305.

§To

whom correspondence may be addressed. E-mail: [email protected] or [email protected].

© 2006 by The National Academy of Sciences of the USA

www.pnas.org兾cgi兾doi兾10.1073兾pnas.0606578103

Fig. 3. SR2-GFP receptors recycle to the cell surface after dopaminemediated internalization in SB1 cells. Before dopamine (Dop) application, control cells have the SR2-GFP receptors at the cell surface (A), and the receptors internalized upon application of 10 ␮M dopamine for 10 min (B), after which, cells were washed free of dopamine and incubated at 37°C in the absence of dopamine and in the presence of cycloheximide for 1 h (C), 1.5 h (D), 2 h (E), and 2.5 h (F). Almost all receptors recycled to the cell surface in 2.5 h (F). (Scale bar, 50 ␮m.)

Fig. 1. Dopamine (Dop) activates the SR2-GFP receptor in HEK293 cells. (A and B) Untransfected HEK293 cells or SB1 cells were loaded with the Rhod2-AM, a Ca2⫹-sensitive fluorescent dye. Ten micromolar dopamine was then applied to the cells. Untransfected cells showed no change in intracellular fluorescence before (Ai) and after (Aii) the addition of dopamine. (Bi) The basal level of fluorescence before the addition of dopamine in SB-1 cells. (Bii) SB1 cells showed an increase in the intracellular Ca2⫹ levels after the addition of 10 ␮M dopamine. (C) The relative increases in intracellular fluorescence after the application of 5-HT or dopamine normalized to basal levels of fluorescence seen in SB1 cells before application of the ligand. (Scale bar, 50 ␮m.)

duced internalization at 100 nM (Fig. 7, which is published as supporting information on the PNAS web site). Interestingly, only 70–75% of the cells showed receptor internalization upon

application of dopamine, as compared with ⬇100% for 5-HT. The reasons why 25–30% of the cells seem to be refractory to dopamine-mediated internalization remain to be determined. Dopamine-Internalized 5-HT2A Receptors Recycle to the Cell Surface with the Same Kinetics as 5-HT-Internalized 5-HT2A Receptors. To

determine whether dopamine-internalized 5-HT2A receptors recycle to the cell surface, 10 ␮M dopamine was applied for 10 min to SB1 cells after the initial cycloheximide treatment. Cells were then washed of dopamine and followed for various times in the continued presence of cycloheximide (Fig. 3). Receptors that internalized upon application of dopamine were seen to localize to the perinuclear region in 10 min, i.e., the recycling endosome (Fig. 8, which is published as supporting information on the PNAS web site). After a 1.5-h wash, receptors were seen to redistribute within the cytoplasm, and, at 2.5 h, receptors returned to the cell surface, as evidenced by the disappearance of the internal fluorescence and a coincident reappearance of surface fluorescence. Because there was no new protein synthesis during the experimental time period, the dynamics of receptor localization in the above experiment suggest that, after dopamine-mediated internalization, receptors recycle to the cell surface in 2.5 h in SB1 cells, similar to 5-HT-internalized receptors (19).

Fig. 2. Dopamine induces SR2-GFP-receptor internalization in SB1 cells. (A–D) Most 5-HT2A receptors were localized to the plasma membrane of SB1 cells after cycloheximide treatment. Cells were then incubated in various concentrations of dopamine for 10 min. (E and F) Internalization of the receptor is seen at 5 ␮M dopamine (E) and increased on application of 10 ␮M dopamine (F). Less than 5 ␮M dopamine, i.e., 1 ␮M (B), 3 ␮M (C), and 4 ␮M (D) dopamine, did not induce any observable internalization of the receptor. (Scale bar, 50 ␮m.)

Bhattacharyya et al.

receptors to serotonin would induce internalization by lower concentrations of dopamine (⬍5 ␮M), SB1 cells were incubated with 50 nM 5-HT, a subthreshold concentration for internalization, for 10 min before the application of dopamine. As expected, 50 nM 5-HT, by itself, did not induce internalization of the SR2-GFP receptors (Fig. 4B). Subsequently, various concentrations of dopamine were applied in the continued presence of 5-HT for an additional 10 min, and receptors were observed to internalize from concentrations as low as 500 nM dopamine (Fig. 4C). Thus, sensitization or ‘‘priming’’ of SR2-GFP receptors with concentrations of 5-HT subthreshold for internalization decreased the threshold concentration of dopamine required to induce internalization by ⬇10-fold. As seen in internalization experiments with dopamine alone, SR2-GFP receptors internalPNAS 兩 October 10, 2006 兩 vol. 103 兩 no. 41 兩 15249

NEUROSCIENCE

5-HT2A Receptors Primed by 5-HT Require Lower Concentrations of Dopamine to Internalize. To test whether prior exposure of 5-HT2A

Fig. 4. Prior application of subthreshold concentration of 5-HT leads to the SR2-GFP receptor internalization at lesser concentration of dopamine (Dop) in SB1 cells. (A and B) Control cells have SR2-GFP receptors at the cell surface (A), and application of 50 nM 5-HT for 10 min did not induce any observable internalization of the receptor (B). (C and D) Receptors were observed to internalize at lower concentration, i.e., 500 nM (C) and 1 ␮M (D) dopamine, if cells were incubated in 50 nM 5-HT for 10 min before the application of dopamine. (Scale bar, 50 ␮m.)

ized in only 70–75% of the cells upon application of dopamine, even after prior sensitization by 5-HT. This priming was further investigated in terms of both the sequence and duration of application of the ligands. No internalization was seen when SB1 cells were treated with subthreshold levels of dopamine (i.e., 1 ␮M dopamine), followed by 50 nM 5-HT (data not shown). Internalization was also not observed when two separate pulses of 50 nM 5-HT were applied to SB1 cells (data not shown), suggesting that the priming phenomenon is not the same as mere sensitization, i.e., increased sensitivity, of the receptor to an agonist. Increased sensitivity seems to be specifically confined to dopamine, the interaction of both 5-HT and dopamine with the receptor seems essential, and the interaction has to occur in a defined temporal sequence. We then went on to test the duration for which the receptor remains primed. SB1 cells were incubated with 50 nM 5-HT for 10 min, after which the 5-HT was washed off, and cells were incubated for varying periods of time before the application of 500 nM dopamine. If the interval between the two applications exceeded 15 min, no internalization took place. This result indicated that the primed receptors retained increased sensitivity to dopamine for 15 min after the wash-out of 5-HT (Fig. 9, which is published as supporting information on the PNAS web site). 5-HT2A Receptors Sensitized by 5-HT Are also Activated by Lower Concentrations of Dopamine. Because priming of the receptor by

5-HT facilitated dopamine-mediated internalization, we next asked whether priming also activated the receptor or caused internalization without activation. To determine this, SB1 cells were loaded with the Ca2⫹-sensitive Rhod2-AM dye, and 50 nM 5-HT was applied. No increase in the Ca2⫹-dependent intracellular fluorescence could be measured, suggesting that 50 nM 5-HT did not result in an observable activation of SR2-GFP 15250 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0606578103

Fig. 5. Lesser concentrations of dopamine (Dop) can activate SR2-GFP receptors if a subthreshold concentration of 5-HT is applied before dopamine. (A) SB1 cells were loaded with the Rhod2-AM Ca2⫹-sensitive dye. Cells did not show a change in the intracellular fluorescence before (Ai) and after (Aii) the addition of 50 nM 5-HT. (B) Five hundred nanomolar dopamine could activate the SR2-GFP receptor, as seen by the increase in the intracellular Ca2⫹ levels, if receptors were exposed to 50 nM 5-HT before the application of dopamine. (Scale bar, 50 ␮m.)

receptors under these conditions. Subsequently, various concentrations of dopamine were applied in the continued presence of 5-HT. Increases in the intracellular Rhod2-AM fluorescence, i.e., increases in Ca2⫹ levels were observed at dopamine concentrations starting at 500 nM, indicating that SR2-GFP receptors were activated by concentrations of 500 nM dopamine if subthreshold concentrations of 5-HT were applied before dopamine (Fig. 5). Dopamine at lower levels than 500 nM did not result in an observable increase in intracellular Ca2⫹ levels. To study further similarities兾differences between the two forms (dopamine alone and 5-HT-primed) of dopaminemediated internalization of the rat 5-HT2A receptor, the time course of recycling of 50 nM 5-HT-primed 500 nM dopamineexposed receptors was examined (Fig. 10, which is published as supporting information on the PNAS web site). It was found that 5-HT-primed dopamine-internalized receptors take 2.5 h to recycle to the surface, similar to the time taken by receptors activated by 10 ␮M 5-HT or dopamine alone. Dopamine-Mediated Internalization of the Rat 5-HT2A Receptor Is Both Ca2ⴙ- and PKC-Independent. The 5-HT-mediated internalization of

5-HT2A receptors depends on PKC (19). To determine the role of PKC in dopamine-mediated internalization of 5-HT2A receptors, experiments were repeated after prior incubation in 50 ␮M sphingosine for 10 min. This concentration of sphingosine completely inhibits 10 ␮M 5-HT-mediated internalization of the receptor, thus serving as a positive control (19). Subsequently, cells were either treated with varying concentrations of dopamine alone or primed with 50 nM 5-HT, followed by varying concentrations of dopamine. Inhibition of PKC did not block Bhattacharyya et al.

internalization of the receptor by 10 ␮M dopamine alone or internalization of 5-HT-primed receptors stimulated with 500 nM dopamine (Fig. 6). Lower concentrations (10 ␮M) of sphingosine also did not inhibit the sensitization of the receptor by 5-HT. These results suggest that PKC is not involved in the sensitization and dopamine-mediated internalization of 5-HT2A receptors. Because 5-HT-primed receptors are activated and internalized by 500 nM dopamine, it was possible that Ca2⫹ surges resulting from receptor activation by dopamine could be responsible for internalization. To test this hypothesis, SB1 cells were loaded with 25 ␮M BAPTA-AM to chelate cytosolic Ca2⫹. When these cells were stimulated with dopamine after being primed by 5-HT, internalization was not inhibited (Fig. 11, which is published as supporting information on the PNAS web site). This result was consistent with the observation that no internalization of receptors took place when SB1 cells were treated with the Ca2⫹ ionophore A23187, alone or followed by 500 nM dopamine (data not shown). These results indicate that, although there is a Ca2⫹ surge upon activation of the receptor by 50 nM 5-HT followed by 500 nM dopamine, 5-HT, or dopamine alone, the Ca2⫹ rise makes little or no contribution to the internalization of the receptor under these conditions. Discussion The major findings of this study are twofold: (i) Dopamine, at micromolar concentrations, is a partial-efficacy agonist at recombinant rat 5-HT2A receptors expressed in HEK293 cells and can induce receptor internalization and subsequent recycling; and (ii) interactions between dopamine and 5-HT2A receptors could be physiologically relevant when the receptor is previously primed by subthreshold concentrations of 5-HT, resulting in Bhattacharyya et al.

PNAS 兩 October 10, 2006 兩 vol. 103 兩 no. 41 兩 15251

NEUROSCIENCE

Fig. 6. Inhibition of PKC does not inhibit the dopamine-mediated internalization or the sensitization of the SR2-GFP receptor by 5-HT. (A) Control cells showed receptors localized to the cell surface after the initial cycloheximide treatment. (B) The receptor internalized upon application of 10 ␮M dopamine (Dop). (C) Internalization of SR2-GFP receptors by 10 ␮M dopamine was not inhibited by 50 ␮M sphingosine. (D) SR2-GFP receptors sensitized by prior application of 50 nM of 5-HT internalized upon application of 500 nM dopamine in the sphingosine-treated SB1 cells. (Scale bar, 50 ␮m.)

receptor internalization driven by extrasynaptic concentrations of dopamine. We found that, although dopamine alone can activate and internalize rat 5-HT2A receptors expressed in HEK293 cells, it does so only at micromolar concentrations high enough to argue against having major physiological relevance. Both 5-HT and dopamine-mediated internalization cause receptors to enter a recycling compartment and, apparently, recycle to the surface in 2.5 h (19). Thus, 5-HT- and dopamine-mediated recycling appear to share common features. Moreover, because both ligands activate the phospholipase C-IP3 pathway, it is likely that components of the internalization pathways are also common to both processes. PKC activation is known to be sufficient and necessary for 5-HT-mediated internalization of the majority of the 5-HT2A receptors (19). Interestingly, however, the present study shows that dopamine-mediated internalization of the receptor does not depend on PKC. This result indicates that, although some aspects of the internalization and recycling may be similar, there are also fundamental differences in the mechanisms of 5-HT- and dopamine-mediated internalization of 5-HT2A receptors. The 5-HT2A receptor has been reported to also activate other signaling pathways (20–24), e.g., ERK MAPK in many cell types. In aortic myocytes, this activation is brought about by the activation of Src kinase, and, in PC12 cells, the 5-HT2A receptor has been shown to activate tyrosine kinases in a Ca2⫹兾calmodulin-dependent manner. Differential activation of one signaling pathway versus another, independent of ligand affinity (functional selectivity), has been observed with 5-HT2 receptors, and it is possible that dopamine differentially activates alternate effector pathways (25, 26) whose role in dopamine-mediated internalization remains to be explored. The subthreshold (50 nM) concentrations of 5-HT used in our experiments did not initiate detectable internalization or show an increase in intracellular Ca2⫹. Still, it is possible that Ca2⫹ levels may have risen, but the sensitivity of the assay was insufficient to detect such an increase (27). Whether signal transduction pathways were activated or not, subthreshold concentrations of 5-HT clearly had a priming action at the 5-HT2A receptors, such that they require 10-fold-lesser amounts of dopamine to cause receptor internalization and activation. This priming is not a mere sensitization of the receptor to an agonist; an interaction of both 5-HT and dopamine with the receptor is essential, and it has to occur in a defined temporal sequence. Such concentrations of dopamine as are required to internalize the 5-HT-primed receptor are within the range of concentrations of dopamine present at synapses in the brain. For example, extracellular concentrations of dopamine in conscious rats has been reported to be ⬇1 ␮M in rat striatum (28). Potential direct interactions of dopamine on 5-HT2A receptors would be most likely in areas of the brain where serotonergic and dopaminergic neurons coexist, e.g., the ventral tegmental area (VTA) (29). Because 5-HT receptors are most likely presynaptic to serotonergic neurons, it is likely that prevailing concentrations of 5-HT are sufficient to sensitize the 5-HT2A receptors before activation by dopamine, and this could result in an altered and increased sensitivity to dopamine. Interestingly, 5-HT2A receptors are present presynaptically on dopaminergic neurons of the VTA, and the interaction of these receptors with some antipsychotic drugs regulates the release of dopamine (30). This dopaminemediated partial activation and internalization could function in vivo, as a partial desensitization of synaptic 5-HT2A receptors. In addition, this process could act as a molecular AND gate, exhibiting activity only in the presence of both 5-HT and dopamine in a defined temporal sequence. The priming of 5-HT2A receptors by 5-HT shows some important features. After internalization, the primed receptor recycles to the cell surface in the same time as taken for 5-HTor dopamine-mediated recycling. Furthermore, our experiments

indicate that dopamine-mediated internalization of primed receptors is independent of PKC activation and independent also of the Ca2⫹ transients produced in the cells. Nonetheless, primed receptors recycle to the cell surface in the same time as that taken for receptors internalized by 5-HT or dopamine alone. These results suggest that 5-HT2A receptors internalized by dopamine, with or without priming by 5-HT, are internalized by a different mechanism from that used by 5-HT. The results also suggest that, when 5-HT occupies the receptor at subthreshold concentrations, pathways other than a PKC-dependent pathway may be activated. The activation and internalization of 5-HT2A receptors may have an important physiological function. For example, this process would be different from interactions of a receptor with an antagonist or inverse agonist, where the second-messenger system is not activated or the basal activity is reduced. The order of application of the agonist (5-HT) and partial-efficacy agonist (dopamine), having very different effects, would also result in a system with the capability of responding to a specific sequence of events, i.e., temporal coding, which is evident from the sequence of neurotransmitter application required for priming to occur. One mechanism by which priming could occur is through 5-HT2A-receptor multimers. Like other 5-HT receptors and many other seven-transmembrane receptors, 5-HT2A receptors may form multimers (31–33). It has been shown that, in some receptor homodimers and heterodimers, binding of ligand to one receptor site changes the conformation of the binding pocket of the other receptor partner (34, 35). One could, therefore, propose a model where, at subthreshold concentrations of 5-HT, one receptor subunit is occupied by 5-HT, and the other subunit then acquires a conformation having higher affinity for dopamine. This situation could result in dopamine binding at lower concentrations. Alternatively, partially activated or modified receptor states may exist at concentrations of 5-HT that we empirically define as subthreshold under our assay conditions. Because all of the experiments used a GFP-tagged receptor, it is important to stress that the C-terminal addition of EGFP did not appear to be responsible for introducing artifacts in our observations, because the internalization of an N-terminaltagged myc-5-HT2A receptor was similar to what was observed with GFP-tagged receptors (data not shown). We therefore conclude that the EGFP tag does not significantly affect any of our results with dopamine, although it has been reported that the dendritic localization of this receptor may depend partially on a C-terminal PDZ-binding domain (36). In summary, there is extensive literature describing interactions between dopaminergic and serotonergic systems in mammalian brain (37, 38). The present study suggests that some of these results could be explained by direct interactions of dopamine with 5-HT receptors. The differential regulation of the 5-HT2A receptors by an agonist (5-HT) or a partial-efficacy agonist (dopamine) adds another dimension to the control of dopamine release and function. Such direct interactions could have implications for our understanding of the effects of a broad range of neuropsychiatric therapeutics on 5-HT and dopamine receptors as well as on normal brain function. Materials and Methods Materials. The rat 5-HT2A-receptor full-length cDNA construct was

a gift from David Julius (University of California, San Francisco, CA). pEGFP-N1 and pCruz-Myc vectors were purchased from Clontech (Mountain View, CA) and Santa Cruz Biotechnology (Santa Cruz, CA), respectively. Tissue-culture reagents were purchased from Life Technologies (Carlsbad, CA) and Sigma (St. Louis, MO). Rhod2-AM dye, BAPTA-AM, and pluronic were from Molecular Probes (Eugene, OR). Poly (DL-ornithine), 5-HT, dopamine, cycloheximide, sphingosine, calphostin C, and 1,4diazabicyclo[2.2.2]octane (DABCO) were purchased from Sigma. 15252 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0606578103

Tissue-culture plastic wares were obtained from Nalge Nunc International (Roskilde, Denmark). Coverslips (Gold seal cover glass, No. 1) were purchased from Clay Adams (Franklin Lakes, NJ). Functional and Internalization Studies. The functional SR2-GFP

receptor was constructed by tagging EGFP at the C terminus of the full-length rat 5-HT2A receptor, and a stable line, SB1, expressing fluorescent SR2-GFP receptors, was established in the HEK293 cell line, as described earlier (19). The effects of dopamine and 5-HT on the activation of SR2-GFP receptors were studied in SB1 cells by monitoring increases in intracellular Ca2⫹ levels as described (19). Briefly, SB1 cells were grown to 65–70% confluence on 35-mm coverslip dishes coated with 20 ␮g兾ml poly (DLornithine), at 37°C and loaded with Rhod2-AM, a Ca2⫹-sensitive fluorescent dye. Fluorescence was monitored before and after the addition of the ligand to determine functional activation of the SR2-GFP receptors. Untransfected HEK293 cells served as the control. To study the effect of dopamine on receptor activation when receptors were preexposed to subthreshold concentration (for activation) of 5-HT, SB1 cells were loaded with Rhod2-AM dye and incubated with 50 nM 5-HT. Subsequently, various concentrations of dopamine were applied to the cells, and fluorescence changes were monitored before and after the addition of dopamine. In all experiments, total fluorescence in cells under various experimental conditions was determined by using a confocal microscope and quantified by using ImageJ software (National Institutes of Health, Bethesda, MD). Internalization studies were done as described (19). Briefly, SB1 cells were treated with 100 ␮g兾ml cycloheximide for 5–6 h to clear internal fluorescence. At the end of this treatment, receptor fluorescence was entirely confined to the cell surface, with no intracellular fluorescence. Various concentrations of dopamine (1, 3, 4, 5, and 10 ␮M) were applied to the cells for 10 min at 37°C. Cells were fixed in 4% paraformaldehyde in phosphate buffer for 30 min, washed with PBS, and imaged by using a confocal microscope, as detailed below. The method of fixation used was identical for all experimental procedures. To study dopamine-mediated internalization of the receptor when receptors were exposed to subthreshold concentration (of internalization) of 5-HT before the application of dopamine, SB1 cells were incubated with 50 nM 5-HT for 10 min. Subsequently, dopamine (500 nM and 1 ␮M) were applied for 10 min. Cells were fixed and imaged. To determine whether two pulses of 50 nM 5-HT, one after the other, were sufficient to cause internalization of the 5-HT2A receptor, SB1 cells were treated with cycloheximide for 6 h. This procedure was followed by a 10-min pulse of 50 nM 5-HT. Subsequently, the cells were washed twice to remove any trace of 5-HT. Then, another pulse of 50 nM 5-HT was provided for 10 min. Cells were fixed and imaged. To determine the time for which the receptor stayed primed after being exposed to 50 nM 5-HT, SB1 cells were treated with cycloheximide for 5 h, after which they were exposed to 50 nM 5-HT in DMEM for 10 min. The cells were then washed twice with DMEM and kept in plain DMEM (with cycloheximide) for varying amounts of time (0–20 min). Dopamine (500 nM) was applied for 10 min. Cells were fixed and imaged. To determine the role of PKC in dopamine-mediated receptor internalization, SB1 cells were preincubated for 10 min in 10 ␮M and 50 ␮M sphingosine (an inhibitor of PKC), after which 10 ␮M 5-HT or 10 ␮M dopamine was applied for an additional 10 min in the continued presence of sphingosine. Cells were then fixed and imaged. Similar experiments were done by using calphostin C, another specific inhibitor of PKC. To determine whether inhibition of PKC inhibits the sensitization of the receptor by 5-HT, SB1 cells were treated with 10 ␮M and 50 ␮M sphingosine for 10 min after the initial cycloheximide treatment. Subsequently, 50 nM 5-HT was applied for 10 min. After that, 500 nM Bhattacharyya et al.

Colocalization Studies. To determine whether 5-HT2A receptors are targeted to the recycling compartment upon dopaminemediated internalization, transferrin coupled to Alexa Fluor 568 was used as a marker of recycling endosomes (39–41). SB1 cells were treated with cycloheximide as described before. Dopamine (10 ␮M) and 10 ␮g兾ml transferrin (conjugated to Alexa Fluor 568) were applied to the cells for 15 min. Subsequently, the cells were kept on ice, washed twice with cold DMEM, and treated with 200 ␮l of ascorbate buffer (pH 4.5) (140 mM sodium ascorbate兾65 mM ascorbic acid兾1 mM CaCl2兾1 mM MgCl2) for 10 min to remove membrane-bound noninternalized transferrin. After 10 min, the cells were fixed with 4% ice-cold paraformaldehyde for 30 min, washed with PBS, and stored in DABCO at 4°C until they were imaged.

To study recycling after priming, followed by dopaminemediated internalization, SB1 cells were treated with cycloheximide for 6 h to clear internal fluorescence. A 10-min pulse of 50 nM 5-HT was applied to the cells, after which the cells were incubated with 500 nM dopamine for 10 min. Cells were washed free of dopamine and incubated at 37°C for various times (0.5–2.5 h) in the continued presence of cycloheximide, fixed, and imaged. Confocal Microscopy. A laser-scanning confocal microscope

(Model MRC1024; Bio-Rad) attached to an inverted microscope (Eclipse TE300; Nikon, East Rutherford, NJ), and Lasersharp acquisition software was used for imaging. A ⫻60 oil-immersion objective (N.A. ⫽ 1. 4) was used with laser power at 30% in all studies. GFP excitation兾emission was achieved with a filter set (488 nm兾510 nm) designed for fluorescein detection. Images were processed with Photoshop (Adobe Systems, Mountain View, CA) by using identical values for contrast and brightness. For Figs. 9–11, confocal images were acquired on the Olympus (Melville, NY) FV 1000 confocal setup attached to an Olympus inverted microscope (IX81). FV10-ASW 1.3 acquisition software was used for imaging. A ⫻60 oil-immersion objective (N.A. ⫽ 1. 4) was used with laser power at 7% of the 488-nm line of the argon ion laser. GFP excitation was achieved with a primary dichroic mirror 405兾488. Fluorescent signals were acquired by using a spectral detector whose bandwidth was automatically set by using the on-line dye database option for FITC. Images were processed by using Photoshop with identical values for contrast and brightness.

internalization, a 10-min pulse of 10 ␮M dopamine was applied to the SB1 cells after 100 ␮g兾ml cycloheximide treatment for 5–6 h. Cells were then washed free of dopamine and incubated at 37°C for various times (1, 1.5, 2, and 2.5 h) in the presence of cycloheximide, fixed, and imaged.

We thank members of our laboratory, particularly Saptarshi Mandal, for valuable discussions and support; Gaiti Hasan and Satyajit Mayor for related discussions; the National Centre for Biological Sciences Confocal Facility and Dr. H. Krishnamurthy for imaging-related discussions; and Rahul Chadda and Satyajit Mayor (National Centre for Biological Sciences, Tata Institute of Fundamental Research) for the generous gift of Alexa Fluor 568-labeled transferrin. This work was supported by the National Centre for Biological Sciences, Tata Institute of Fundamental Research. S.B. and A.B. were recipients of the Kanwal Rekhi Career Development Fellowship from the Tata Institute for Fundamental Research endowment fund.

1. Kahn RS, Davidson M (1993) Psychopharmacology (Berlin) 112:S1–S4. 2. McBride WJ, Murphy JM, Gatto GJ, Levy AD, Yoshimoto K, Lumeng L, Li TK (1993) Alcohol Suppl 2:463–467. 3. Kosten TR, Markou A, Koob GF (1998) J Nerv Ment Dis 186:737–745. 4. Baldessarini RJ, Tarazi FI (2001) in Goodman & Gilman’s The Pharmacological Basis of Therapeutics, eds Hardman JG, Limbird LE, Goodman Gilman A (McGraw–Hill, New York), pp 485–520. 5. Meltzer HY, Matsubara S, Lee JC (1989) J Pharmacol Exp Ther 251:238–246. 6. Yatham LN, Goldstein JM, Vieta E, Bowden CL, Grunze H, Post RM, Suppes T, Calabrese JR (2005) J Clin Psychiatry 66 (Suppl 5):40–48. 7. Meltzer HY, Matsubara S, Lee JC (1989) Psychopharmacol Bull 25:390–392. 8. Schmidt CJ, Fadayel GM (1995) Eur J Pharmacol 273:273–279. 9. Gobert A, Millan MJ (1999) Neuropharmacology 38:315–317. 10. Lucas G, Spampinato U (2000) J Neurochem 74:693–701. 11. Pehek EA, McFarlane HG, Maguschak K, Price B, Pluto CP (2001) Brain Res 888:51–59. 12. Pehek EA, Nocjar C, Roth BL, Byrd TA, Mabrouk OS (2006) Neuropsychopharmacology 31:265–277. 13. Iyer RN, Bradberry CW (1996) J Pharmacol Exp Ther 277:40–47. 14. Auclair A, Blanc G, Glowinski J, Tassin JP (2004) J Neurochem 91:318–326. 15. Doherty MD, Pickel VM (2000) Brain Res 864:176–185. 16. Ikemoto K, Nishimura A, Okado N, Mikuni M, Nishi K, Nagatsu I (2000) Brain Res 853:377–380. 17. Woodward RM, Panicker MM, Miledi R (1992) Proc Natl Acad Sci USA 89:4708–4712. 18. Oz M, Zhang L, Rotondo A, Sun H, Morales M (2003) Synapse 50:303–313. 19. Bhattacharyya S, Puri S, Miledi R, Panicker MM (2002) Proc Natl Acad Sci USA 99:14470–14475. 20. Banes A, Florian JA, Watts SW (1999) J Pharmacol Exp Ther 291:1179–1187. 21. Quinn JC, Johnson-Farley NN, Yoon J, Cowen DS (2002) J Pharmacol Exp Ther 303:746–752. 22. Turner JH, Raymond JR (2005) J Biol Chem 280:30741–30750. 23. Guillet-Deniau I, Burnol AF, Girard J (1997) J Biol Chem 272:14825–14829.

24. Kurrasch-Orbaugh DM, Parrish JC, Watts VJ, Nichols DE (2003) J Neurochem 86:980–991. 25. Urban JD, Clarke WP, von Zastrow M, Nichols DE, Kobilka BK, Weinstein H, Javitch JA, Roth BL, Christopoulos A, Sexton P, et al. (2006) J Pharmacol Exp Ther, in press. 26. Berg KA, Maayani S, Goldfarb J, Scaramellini C, Leff P, Clarke WP (1998) Mol Pharmacol 54:94–104. 27. Ohkura M, Tanaka N, Kobayashi H, Wada A, Nakai J, Yamamoto R (2005) Eur J Pharmacol 518:18–21. 28. Garris PA, Christensen JR, Rebec GV, Wightman RM (1997) J Neurochem 68:152–161. 29. Ikemoto K, Nishimura A, Okado N, Mikuni M, Nishi K, Nagatsu I (2000) Brain Res 853:377–380. 30. Matsui-Sakata A, Ohtani H, Sawada Y (2005) Drug Metab Pharmacokinet 20:187–199. 31. Bouvier M (2001) Nat Rev Neurosci 2:274–286. 32. Herrick-Davis K, Grinde E, Mazurkiewicz JE (2004) Biochemistry 43:13963– 13971. 33. Xie Z, Lee SP, O’Dowd BF, George SR (1999) FEBS Lett 456:63–67. 34. Rocheville M, Lange DC, Kumar U, Patel SC, Patel RC, Patel YC (2000) Science 288:154–157. 35. Herrick-Davis K, Grinde E, Harrigan TJ, Mazurkiewicz JE (2005) J Biol Chem 280:40144–40151. 36. Xia Z, Hufeisen SJ, Gray JA, Roth BL (2003) Neuroscience 122:907–920. 37. Bortolozzi A, Diaz-Mataix L, Scorza MC, Celada P, Artigas F (2005) J Neurochem 95:1597–1607. 38. Esposito E (2006) Curr Drug Targets 7:177–185. 39. Schlierf B, Fey GH, Hauber J, Hocke GM, Rosorius O (2000) Exp Cell Res 259:257–265. 40. Dautry-Varsat A, Ciechanover A, Lodish HF (1983) Proc Natl Acad Sci USA 80:2258–2262. 41. Lamb JE, Ray F, Ward JH, Kushner JP, Kaplan J (1983) J Biol Chem 258:8751–8758.

Recycling Studies. To study recycling after dopamine-mediated

Bhattacharyya et al.

PNAS 兩 October 10, 2006 兩 vol. 103 兩 no. 41 兩 15253

NEUROSCIENCE

and 1 ␮M dopamine was applied for an additional 10 min. Cells were fixed and imaged in the confocal microscope. To determine the role played by Ca2⫹, SB1 cells were first treated with cycloheximide for 5 h. They were then loaded with 25 ␮M BAPTA-AM for 45 min. Cells were washed free of BAPTA-AM and incubated at room temperature for 30 min, followed by 50 nM 5-HT for 10 min and, subsequently, 500 nM dopamine for 10 min. Cells were then fixed and imaged. Also, cells were preincubated with the Ca2⫹ ionophore A-23187 (20 ␮M) for 15 min, and then the SB1 cells were incubated with 50 nM 5-HT for 10 min or with 500 nM dopamine. Cells were then fixed and imaged. In all experiments, cycloheximide was present until the cells were fixed. Typically 100–150 cells were chosen randomly and imaged in any one experiment. All experiments were repeated at least three times.