Cell surface expression of 5-hydroxytryptamine type 3 receptors is promoted by RIC-3.

RIC-3 has been identified as a molecule essential for the recruitment of functional nicotinic acetylcholine receptors composed of alpha7, but it exhibits inhibitory effects on alpha4beta2 or alpha3beta4 receptors. In this study, we investigated the role of RIC-3 in the recruitment of 5-hydroxytryptamine type 3A (5-HT(3A)) receptors to the cell surface. Although RIC-3 is not essential for the surface transport of 5-HT(3A) receptors, we found that its presence enhances both receptor transport and function in a concentration-dependent manner. RIC-3 is localized to the endoplasmic reticulum, as evidenced by co-localization with the chaperone molecule, binding protein (BiP). RIC-3 is not detected at significant levels on the cell surface when expressed alone or in the presence of 5-HT(3A). RIC-3 and 5-HT(3A) show a low level interaction that is transient (<4 h). That RIC-3 can interact with an endoplasmic reticulum-retained 5-HT(3A) construct, combined with the transient interaction observed and lack of significant surface-expressed RIC-3, suggests that RIC-3 may play a role in 5-HT(3A) receptor folding, assembly, or transport to the cell surface.

The most remarkable morphological feature of the brain is not the highly complex, interconnected neuronal pathways but the enormous number (ϳ10 15 ) of potentially distinct synaptic connections involved in information transfer between neurons. Moreover, synapses are not static and passive translators of information but can change their efficiency of synaptic transmission (1). This process is termed "synaptic plasticity" and is thought to lie at the heart of the capacity of the brain for learning and memory. Synaptic plasticity may reside pre-or postsynaptically (or both), modulating neurotransmitter release or neurotransmitter receptor responses, respectively.
A fundamental question in neurobiology is how receptor biogenesis is orchestrated. A vast array of receptor-interacting proteins have been identified as participating in ligand-gated ion channel trafficking and localization (1) and have lead to dramatic advances in our knowledge of synaptic plasticity.
The 5-HT 3 1 receptors belong to the Cys loop superfamily of ligand-gated ion channels that includes the nicotinic acetylcho-line, GABA A , and glycine receptors. The structural relationship (2)(3)(4)(5) of the members of this group suggests that their folding and assembly may involve similar posttranslational chaperone-mediated events (6 -8).
The forward transport of these receptors requires the appropriate assembly of specific subunits and release from the endoplasmic reticulum (ER) (6,7). Specific assembly signals have been identified in receptors for GABA A (9), glycine (10), and acetylcholine (11). The export of receptors from the ER represents a critical checkpoint for surface expression, with quality control within the lumen of the ER being performed by the resident chaperone proteins (9,12). In addition, cytoplasmically exposed ER retention signals within the receptors have been identified as elements that control protein export from the ER (13). The failure to pass these quality control checks results in ER-associated degradation by the proteasome (12,14,15).
Remarkably, despite the size of this growing list and apart from an implication for cyclophilin (based on modulation of receptor function) (16), no molecule has been implicated in the recruitment of 5-HT 3 receptors. As RIC-3 has been proposed to inhibit 5-HT 3 receptor expression (19), we addressed the role of RIC-3 on 5-HT 3A transport to the cell surface. RIC-3 was identified in a genetic screen for molecules that are required to maintain AChR function (18). Characterization of RIC-3 confirmed that it is required for the production of functional AChRs (18,19). The predominant localization of RIC-3 to the ER and not at synapses, combined with the observation that AChRs failed to exit the neuronal cell body in the absence of RIC-3 (19), is highly suggestive of a role in AChR transport from the ER to the surface. However, a contradictory study (34) reports that RIC-3 is not involved in the transport of AChR ␣7 receptors to the cell surface. Instead, the function of RIC-3 appears to be at the level of protein folding and is essential for the generation of the ␣7 ligand (␣-bungarotoxin) binding site. Furthermore, RIC-3 was found at the cell surface associated with ␣7 (34), raising the question regarding the location of RIC-3 chaperone activity.
We investigated the role of RIC-3 on the trafficking of 5-HT 3A receptors and discovered a significant role in the enhancement of receptor recruitment to the cell surface. The interaction of RIC-3 with 5-HT 3A occurs within the ER and is transient, and RIC-3 is not detected at the cell surface. We propose that RIC-3 plays a role in the transport of 5-HT 3A receptors, possibly by enhancing protein folding and/or stabilization.
Antibodies-Anti-HA and anti-Myc monoclonal antibodies were used directly as supernatant (20 g/ml) or purified on immobilized protein A. Antiserum to RIC-3 was raised in sheep using either an extracellular/ luminal (depending on whether RIC-3 is expressed on the cell surface) epitope (RIC-3b, SDGQTPGARFQRSHL) or an intracellular epitope (RIC-3a, KAYTGSMLRKRNP) as the antigen. RIC-3b was used for immunofluorescence/ELISA, and RIC-3a was used for immunoprecipitations. Neither RIC-3 antibody produced a significant signal by immunofluorescence on rat neurons (hippocampal). The secondary antibodies, goat anti-mouse Alexa Fluor 488/568 and donkey anti-sheep Alexa Fluor 488, were purchased from Molecular probes and used at 1 g/ml. The secondary antibody, sheep anti-mouse horseradish peroxidase (HRP), was purchased from Amersham Biosciences and used at 1/1000.
Immunofluorescence-COS7 cells were fixed in 3% paraformaldehyde (in PBS), washed twice in 50 mM NH 4 Cl (in PBS), and blocked (10% FBS, 0.5% bovine serum albumin in PBS) for 30 min. Subsequent washes and antibody dilutions were performed in PBS containing 10% FBS and 0.5% bovine serum albumin. Following surface immunofluorescence, cells were permeabilized by the addition of 0.5% Triton X-100 (10 min), and the immunofluorescence protocol was repeated from the NH 4 Cl step. Where applicable, the fluorophores, Alexa 488 and Alexa 568, were used to detect surface or total receptor populations, respectively. Cells were examined using a wide-field imaging system (Improvision).
Quantification of Cell Surface Expression-HEK293 cells were plated into 96-well dishes. Eight transfections (2 g of 5-HT 3A -HA and 0 -8 g of RIC-3, total DNA maintained at 10 g using GABA A receptor ␣-1 cDNA) were used in each dish (with each condition in sextuplet). Cells were fixed in 3% paraformaldehyde (in PBS). Cell surface detection was performed in the absence of detergent, and total expression levels were determined following Triton X-100 treatment (0.5% for 15 min). Cells were washed twice in 50 mM NH 4 Cl (in PBS) and blocked (10% FBS, 0.5% bovine serum albumin in PBS) for 1 h. Subsequent washes were performed in block. Receptor expression was determined using an HRP-conjugated secondary antibody and assayed using 3,3Ј,5,5Ј-tetramethylbenzidine (Sigma) as the substrate, with detection at 450 nm after 30 min following the addition of 0.5 M H 2 SO 4 . The reaction rate was determined to remain linear for up to 1 h.
Membrane Potential Assay-HEK293 cells were plated into 96-well dishes (Molecular Devices). Eight transfections (as above) were used for each dish. After 24 h, cells were incubated in Red Membrane Potential® assay solution diluted 1:10 (Molecular Devices) for 1 h at room temperature. A range of concentrations of 5-HT (in Hanks' solution) was added automatically by the Flexstation II® apparatus and responses recorded over 90 s using SoftMax Pro 4.6 to analyze the results and plot doseresponse curves according to the manufacturer's instructions. Each point is the average of at least 6 wells monitored over a total area of ϳ6 mm 2 .
Cell Surface Biotinylation-Cells were washed three times in cold PBS, and cell surface proteins were biotinylated using 0.5 mg/ml sulfo-NHS-LC-biotin (Pierce) for 30 min on ice followed by quenching in cold PBS containing 100 mM glycine. Cells were lysed in PBS containing 1% Triton X-100, and biotinylated proteins were isolated overnight using NeutrAvidin-agarose (Pierce) at 4°C. Beads were washed five times and bound protein eluted in SDS-PAGE sample buffer.

RESULTS
To examine the subcellular distribution of 5-HT 3A receptors and RIC-3, heterologous expression in COS-7 cells was utilized. COS-7 cells were chosen because of their flattened phenotype, offering higher morphological resolutions of intracellular structures. To facilitate biochemical and morphological analyses, the 5-HT 3A subunit was tagged using the epitopes of HA or Myc. These epitope tags were added to the amino terminus of 5-HT 3A between amino acids 5 and 6 of the mature polypeptide (downstream from the predicted signal sequence cleavage site) to create 5-HT 3A -HA or 5-HT 3A -Myc. No functional effects of these tags were evident (8).
It has been reported previously (8) that 5-HT 3A -HA can access the cell surface and function as a homomeric 5-HT 3 receptor. In support of this assertion, surface immunofluorescence of the HA epitope can be detected in the absence of detergent, where it is evenly distributed in a punctate pattern (Fig. 1B). Intracellular staining is also evident (Fig. 1A). In higher expressing cells (not shown), receptors often exhibited localization to filopodial structures, as observed elsewhere (32). Upon permeabilization, the majority of intracellular receptors are observed in an ER-like pattern (8). In addition, small punctate spots are observed, indicative of their presence in transport vesicles (36). Upon the co-expression of 5-HT 3A -HA and RIC-3, a more robust cell surface expression of 5-HT 3A -HA is observed (Fig. 1D). Moreover, a striking localization to filopodia is observed, similar to that reported previously (32). The more dramatic localization to filopodia upon co-expression is most likely a result of increased expression rather than a distinct targeting mechanism. In addition to increased surface staining, a significant increase in immunofluorescence is observed in permeabilized cells (Fig. 1C).
Given the apparent increase in cell surface levels of 5-HT 3A , we endeavored to explore the subcellular localization of recombinantly expressed RIC-3 using the antibody raised against the extracellular/luminal (depending on localization to the surface/ intracellular compartments, respectively) region. It should be noted that although earlier analysis programs predicted two transmembrane domains (18), more recent data bases such as those for SignalP 2.0 (34) and SignalP 3.0 suggest that the first transmembrane region is a signal sequence. Regardless, the epitope used in this study is predicted to be extracellular/ luminal in either model. In the presence of detergent, a high level of RIC-3 is detected, showing a highly clustered distribu-tion as well as an ER-like pattern (Fig. 1E). No peripheral staining, which would indicate surface expression, was observed. In support of this observation, only very low levels of surface staining, similar to background levels in untransfected cells, were evident (Fig. 1F). To confirm RIC-3 localization to the ER, co-staining for BiP, an ER marker (7), was performed. Both RIC-3 and BiP staining show similar but not identical patterns (Fig. 1, G and H). It has been reported that RIC-3 is significantly localized to the plasma membrane in a cell line expressing the nicotinic acetylcholine receptors (nAChR) ␣7 subunit (34). We investigated, therefore, whether RIC-3 was similarly localized when 5-HT 3A was present. Upon the coexpression of RIC-3 with 5-HT 3A -HA, significant surface staining (not shown) or the presence of surface biotinylated protein (see Fig. 4D) for RIC-3 was still not evident, suggesting that RIC-3 may be functioning within the ER in the folding, assembly, or transport of 5-HT 3A receptors rather than as a plasma membrane clustering/anchoring molecule. However, we cannot rule out a transient presence of RIC-3 at the cell surface.
To quantify the apparent increased cell surface expression observed by immunofluorescence, we used the whole cell ELISA approach to quantify cell surface and total receptor expression. This approach is similar to immunofluorescence, except that an HRP-conjugated antibody is used to provide quantitative results. It is not possible to equate these values with an actual percentage distribution of a receptor, as the estimation of total receptors is consistently underestimated in the presence of detergent. In HEK293 cells expressing 5-HT 3A -HA, positive signals are apparent for cell surface receptors (mock-transfected cell values have been subtracted) (Fig. 2). Increasing amounts of RIC-3 cDNA (2-8 g) were co-transfected with 2 g of 5-HT 3A -HA. In support of a role in the recruitment of 5-HT 3A receptors to the cell surface by RIC-3, cell surface levels of 5-HT 3A -HA were enhanced as RIC-3 was increased. The most significant enhancement occurred at equimolar or 1:2 ratios, with diminishing additional effects at higher ratios. An examination of the ratio between surface and total receptor levels supported this observation and suggests that the transport pathway may become limiting for some other factor. Interestingly, a small increase in total receptor expression was observed, becoming significant when higher amounts of RIC-3 were co-transfected. This result suggests that RIC-3 may be exerting a stabilizing effect on 5-HT 3A .
To assess the impact of RIC-3 on 5-HT 3A receptor activation, we utilized a membrane potential reagent and analysis using a Flexstation ("mini FLIPR") apparatus (Molecular Devices). 5-HT 3A was co-transfected with increasing amounts of RIC-3 (0 -8 g). Maximal responses to a range of 5-HT concentrations were plotted against log 5-HT concentration to yield a doseresponse curve for each ratio of 5-HT 3A :RIC-3. In support of the ELISA results, increasing maximal responses were observed with increasing ratios of co-transfected RIC-3 (Fig. 3). However, the EC 50 for 5-HT responses was unaltered for the different ratios, ranging from 1.85 to 2.4 M, as reported previously for 5-HT 3A (35). Thus, these results are consistent with an increase in cell surface receptor number and a lack of change in the apparent potency of 5-HT.
To determine the existence of direct protein interactions between 5-HT 3A -HA and RIC-3, we performed co-immunopre- amounts of RIC-3 are shown. GABA A receptor ␣-1 cDNA was included to ensure that equal amounts of cDNA (10 g) were used in all transfections. Surface receptor levels were detected using antibodies to Myc followed by secondary antibodies conjugated to HRP in the absence of detergent. Total receptor levels were determined as above, in the presence of detergent. Surface/total ratios were derived from this data set. Each value represents the mean Ϯ S.E. of at least six determinants in at least three independent experiments. *, significant difference from the control without RIC-3 (p Ͻ 0.001, t test). cipitation experiments on [ 35 S]methionine-labeled HEK293 cells. In cells transfected with 5-HT 3A -HA and immunoprecipitated with HA antibodies, characteristic (8) 5-HT 3A bands (48 -54 kDa) were observed (Fig. 4A). Upon the co-expression of RIC-3, an additional band of ϳ60 kDa was observed (34). This band co-migrated with the band identified when the RIC-3 antibody was used to immunoprecipitate from RIC-3 transfected cells. However, this band does not appear to be present in stoichiometric amounts despite the RIC-3 being transfected in a 2:1 ratio to 5-HT 3A -HA. Plus, both molecules possess nine methionines (so should be labeled equivalently), suggesting that the interaction might be transient. No co-immunoprecipitation is observed when the reciprocal experiment is performed using the RIC-3 antibody.
Because 5-HT 3 receptor assembly occurs within the ER, and because this is the predominant location for RIC-3, it seems likely that their interaction would at least initially occur within this compartment. To confirm that these two molecules are able to interact within the ER, we repeated the experiment using a 5-HT 3A construct with an engineered ER retention signal, 5-HT 3A ЈCRARЈ, that has been shown to be retained in the ER (13). The results of this experiment (Fig. 4B) are identical to those for the wild-type 5-HT 3A -HA, confirming that the interaction can occur within the ER. However, it is possible that the interaction is maintained throughout its transport to, and residence at, the plasma membrane.
To address whether the 5-HT 3A -HA and RIC-3 interaction is transient, we radiolabeled HEK293 cells expressing 5-HT 3A -HA and RIC-3 with [ 35 S]methionine, and then radiolabel was removed and cells were chased in the presence of excess cold methionine (DMEM/FBS) for 0 -4 h (Fig. 4C). Over this period of time, we found that the 5-HT 3A -HA expression level was stable. However, the amount of RIC-3 co-immunoprecipitating was reduced significantly, suggesting that the interaction between 5-HT 3A and RIC-3 is transient, lasting just hours.
Given the recent report of RIC-3 at the cell surface when co-expressed with the nicotinic acetylcholine ␣7 subunit in A7-3 cells (34), we addressed whether the same might be true for 5-HT 3A and RIC-3. The cell surface of HEK293 cells expressing RIC-3 or RIC-3 and 5-HT 3A -HA were biotinylated. The biotinylated surface proteins were purified using NeutrAvidinagarose and probed for the existence of RIC-3 by Western blotting. These results show that RIC-3 was not present at the cell surface (Fig. 4D), either when expressed alone (lane 2) or in combination with 5-HT 3A (lane 3). Blots of total lysate from cell surface biotinylated cells were probed with streptavidin-HRP to confirm that the biotinylation reaction had been successful (not shown). DISCUSSION

RIC-3 was identified in a genetic screen in
Caenorhabditis elegans for suppression of a dominant mutation in the acetylcholine receptor subunit DEG-3 (18). The DEG-3 mutation leads to the production of a non-desensitizing channel that is responsible for causing necrotic death to neurons. In this screen, mutations in RIC-3 were found to suppress this toxicity, implying an essential role for RIC-3 in acetylcholine receptor function. This hypothesis was confirmed by electrophysiological analysis, in which responses to acetylcholine, nicotine, and levamisole, but not GABA, were eliminated in ric-3 mutant cells (18). Furthermore, RIC-3 was localized to neuronal cell bodies, as were DEG-3 receptors in ric-3 mutants, suggesting an essential role for RIC-3 in the ER on the formation of nAChR.
Later analysis extended this role of RIC-3 to human nACh ␣7 receptors (19). Remarkably, an inhibitory effect of RIC-3 was observed on the function of nACh ␣4␤2 and ␣3␤4 and 5-HT 3 receptors, with no effect on glycine receptors. In contrast to the apparent requirement for RIC-3 on DEG-3 transport from the cell body (18), Williams et al. (34) identified no role for RIC-3 in the transport of nAChR ␣7 receptors to the cell surface. Instead these authors determined a role in the generation of the ligand (␣-bungarotoxin) binding site. Although this role might occur within the ER and be a prerequisite for export of nACh ␣7 receptors to the cell surface, the findings of this study suggests that surface nACh ␣7 occurs in the absence of exogenous RIC-3 and that the nACh ␣7-RIC-3 interactions were present on the cell surface.
Given the requirement for RIC-3 in the transport of DEG-3 (18) and the folding of ␣7 (34), the inhibitory function of RIC-3 on nACh ␣4␤2 and ␣3␤4 and 5-HT 3 receptors (19), and the lack of effect of RIC-3 on GABA A or glycine receptors (18,19), we decided to investigate the role of RIC-3 on GABA A and 5-HT 3A receptor transport to the cell surface. First, in agreement with earlier findings for GABA A receptor function (18), we found no modulatory role for RIC-3 on the surface transport of GABA A ␣1␤2 receptors (results not shown). In contrast to the inhibition of RIC-3 on 5-HT 3 receptor function, we observed a robust enhancement of 5-HT 3A receptor surface expression and function. It is not clear why our findings are contradictory, but they may reflect differences in the expression system (oocytes versus HEK293/COS-7 cells) or species differences of the 5-HT 3A cDNA (murine versus Homo sapien).
Regardless, we observed clear increases in cell surface immunofluorescence in the presence of RIC-3, leading to the dramatic localization of 5-HT 3A to filopodia. This apparent increase in filopodial association is most likely the result of increased surface expression rather than specific re-localization because of RIC-3, because a similar distribution in the same cell type has been reported previously when expression at higher levels was evident (32). The ER is a critical checkpoint in the transport of cell surface proteins. In keeping with a role for RIC-3 within this compartment, RIC-3 exhibits co-localization with the ER chaperone, BiP. Furthermore, we could not detect significant levels of RIC-3 at the cell surface by either immunofluorescence or cell surface biotinylation. This finding appears to contradict the results observed previously (34). However, an important difference is that in the previous study, cells co-expressed nACh ␣7 receptors. Thus RIC-3 may be stabilized at the cell surface by ␣7 receptors. However, this is not the case for 5-HT 3A receptors, in keeping with the low level, transient interaction that can be detected. Furthermore, significant levels of surface RIC-3 could not be determined when co-expressed with 5-HT 3A . The low level of detectable interactions between 5-HT 3A and RIC-3 is consistent with a transient or low affinity interaction. Further studies are required to confirm and map the interaction sites to address the mechanism of action of RIC-3 on 5-HT 3A receptors.
Using a cell-based ELISA method to quantify 5-HT 3A receptor recruitment, we observed that the increase in 5-HT 3A receptor surface expression related to the level of RIC-3, a result supported by our functional analysis. Interestingly, an increase in total receptor levels was consistently observed, particularly when RIC-3 was present in excess. This is unlikely to be because of a specific transcriptional regulation, because all flanking regions of the 5-HT 3A have been removed, and receptor expression is driven from the vector promoter. Furthermore, GABA A receptor expression from the same vector is unaffected by RIC-3 expression (results not shown). Rather, this would be consistent with a role for RIC-3 in stabilizing 5-HT 3A receptor subunits by a mechanism such as promoting receptor folding (as observed for nACh ␣7 receptors) (34) or permitting more receptors to avoid degradation and thus, traffic to the cell surface (in contrast to the mechanism observed previously for nACh ␣7 receptors) (34). When we examined the ratio of surface to total receptors, it appeared that the enhancement in cell surface levels became saturated at low levels (1:2 ratio for 3A to RIC-3) of co-transfected RIC-3, suggesting the limitation of some other factor(s) in receptor transport.
Given the effect of RIC-3 on both surface and total 5-HT 3A receptor levels, the ER localization of RIC-3 (and absence from the cell surface) and the transient, low level interaction with 5-HT 3A receptors suggest that RIC-3 promotes some posttranslational modification that is a prerequisite for protein structure/stability and transport to the cell surface. In contrast to DEG-3 in C. elegans, RIC-3 is not an absolute requirement for the surface expression of 5-HT 3A in COS-7/HEK293 cells, as it is not endogenously expressed (34). An attractive yet highly speculative possibility to explain the striking variability in the promotion or inhibition of receptor function (18,19,34) is that RIC-3 may play a role in defining receptor composition by promoting the folding of certain receptor subunits into a particular conformation that is either a prerequisite for or a barrier to assembly with another subunit. The nAChR ␣7 receptor may be able to traffic in either conformation (37,38).