Dual Role of the RIC-3 Protein in Trafficking of Serotonin and Nicotinic Acetylcholine Receptors*

,

The ric-3 gene is required for maturation of nicotinic acetylcholine receptors in Caenorhabditis elegans.The human homolog of RIC-3, hRIC-3, enhances expression of ␣7 nicotinic receptors in Xenopus laevis oocytes, whereas it totally abolishes expression of ␣4␤2 nicotinic and 5-HT 3 serotonergic receptors.Both the N-terminal region of hRIC-3, which contains two transmembrane segments, and the C-terminal region are needed for these differential effects.hRIC-3 inhibits receptor expression by hindering export of mature receptors to the cell membrane.By using chimeric proteins made of ␣7 and 5-HT 3 receptors, we have shown that the presence of an extracellular isoleucine close to the first transmembrane receptor fragment is responsible for the transport arrest induced by hRIC-3.Enhancement of ␣7 receptor expression occurs, at least, at two levels: by increasing the number of mature receptors and facilitating its transport to the membrane.Certain amino acids of a putative amphipathic helix present at the large cytoplasmic region of the ␣7 subunit are required for these actions.Therefore, hRIC-3 can act as a specific regulator of receptor expression at different levels.
Neurotransmitter-gated receptors constitute a gene superfamily of membrane proteins that mediate fast transmission in the nervous system.Receptors for acetylcholine (ACh) 1 (nicotinic type), glycine, ␥-aminobutyric acid (GABA A type), and serotonin (5-HT 3 type) are included within this family (1).They share the same oligomeric structure, being composed of five homologous subunits surrounding the ion pore (2).Receptors are assembled in the endoplasmic reticulum (ER) and transported to the cell surface through a process that may be of central importance in regulating the efficacy of synaptic transmission (3,4).This process is relatively inefficient (5,6), probably as a consequence of tight quality controls that guarantee the functional competence of the final product.For this purpose, specific proteins involved in assembly and trafficking of receptors may be required (7)(8)(9).The hRIC-3 protein could be one of them, as mutations in the ric-3 gene affect maturation of nicotinic acetylcholine receptors (nAChRs) in Caenorhabditis elegans (10).Moreover, the human homolog, hRIC-3, shows differential effects when co-expressed with several ligandgated receptors (11).Thus, hRIC-3 enhances surface expression and currents of ␣7 nAChRs, whereas it inhibits currents produced by other nAChRs subtypes (␣4␤2 and ␣3␤4) and by 5-HT 3 serotonin receptors (5-HT 3 Rs).These opposite effects suggested that the hRIC-3 protein might play a key role in the biogenesis of some ligand-gated receptors and prompted us to investigate how and where it performs its action.Here we show that both the N-terminal region of hRIC-3, which contains two transmembrane segments, and the C-terminal region are needed for its action.The hRIC-3 protein acts as a barrier for some receptors such as ␣4␤2 nAChRs and 5-HT 3 Rs, stopping the traffic of mature receptors to the membrane.By contrast, the inefficient transport of ␣7 nAChRs is enhanced by hRIC-3 in a process in which certain amino acids at the amphipathic helix located at the C-terminal region of the large cytoplasmic domain are involved.Thus, hRIC-3 plays a dual role in modulating trafficking of some ligand-gated ion channels.

MATERIALS AND METHODS
Plasmid Constructions-Truncated constructs of hRIC-3 were obtained by deletions produced by restriction digests of the human hRIC-3 cDNA (11).The C-terminal fusion of hRIC-3, ␣7, and ␤2 nAChR subunits to green fluorescent protein (GFP) were obtained by inserting the corresponding cDNAs into the pEGFP-N1 vector (Clontech).Maintenance of the appropriate reading frame and suppression of stop codons were achieved by ligating suitable double-stranded oligonucleotides.The C-terminal fusion of hRIC-3 to red fluorescent protein (RFP) was obtained by carrying the hRIC-3 cDNA directly from pEGFP-N1 to pDsRed1-N1 (Clontech), because both vectors have identical polylinker regions and reading frames.Nicotinic-serotonergic chimeric proteins were obtained by PCR, using as starting construct the V201 chimera, kindly provided by Dr. K. T. Dineley.Mutants of the amphipathic helix of the rat ␣7 nAChR subunit were generated by using annealed singlestranded oligonucleotides with the desired sequences and proper single strand ends that could be ligated easily to the ends generated by restriction enzymes present in previous constructs (12).
Heterologous Expression in Xenopus Oocytes-DNAs of human neuronal nAChR subunits (␣7, ␣4, and ␤2) (13), rat ␣7 nAChR (14) and its mutants, mouse 5-HT 3 R (15), ␣7-5-HT 3 R chimeras, and hRIC-3 and its truncated forms were inserted into the pSP64T vector (16) or a derivative thereof.Capped mRNA was synthesized in vitro using SP6 RNA polymerase and the mMESSAGEmMACHINE kit (Ambion).Defoliculated Xenopus laevis oocytes were injected with 10 ng of total subunit cRNA and 5 ng of hRIC-3 cRNA in 50 nl of sterile water.The final concentrations of subunits cRNAs were kept constant between injections with or without hRIC-3 for each experiment.All experiments were performed within 3 days after cRNA injection.
Specific surface expression of 125 I-␣-bungarotoxin (␣-Bgt) binding sites was tested with 5 nM 125 I-␣-Bgt as described (17).Briefly, oocytes were incubated for 2 h at 18 °C.At the end of the incubation, unbound I-␣-Bgt was removed, oocytes were washed, and bound radioactivity was counted.Binding to internal and external sites was obtained by carrying out the incubations in the presence of 0.1% saponin.Nonspecific binding was determined using noninoculated oocytes in the same conditions.Electrophysiological recordings were done as described previously (17).Functional expression of each construct was estimated as the peak ionic current evoked by a 1-s application of 1 mM ACh at Ϫ80 mV, and no correction for desensitization was made.Errors are standard error of means.
For Western blots, oocytes were lysed and proteins separated by 10% SDS-polyacrylamide gel electrophoresis.After the transfer, nitrocellulose membranes were blocked for 30 min at 20 °C with 5% dry milk in phosphate-buffered saline, 0.2% Tween 20 and incubated with an anti-GFP antibody (Clontech) diluted 1:100 in phosphate-buffered saline, 0.2% Tween 20 and 1% dry milk for 1 h.After incubation with the secondary antibody at room temperature for 1 h, the bands were visualized by a chromogenic reaction (Sigma Fast, nitro blue tetrazolium, Sigma).
Cell Culture and Transfection-COS cells were grown in 90% Dulbecco's modified Eagle's medium and 10% fetal calf serum.Plasmids used in the transfections were purified by Concert columns (Invitrogen).Cells (4 ϫ 10 4 ) on 3.5-cm-diameter plates were transfected with Lipofectamine (Invitrogen) as indicated by the manufacturer, using a total of 1.5 g of DNA (ratio of hRIC-3 to receptor was 1:1).The density of ␣-Bgt binding sites was determined as described previously (18).
Confocal Microscopy Studies of the Cellular Distribution of hRIC-3 and nAChRs-Direct observation of proteins fused to GFP or RFP was performed with an Olympus Fluoview FV300 confocal laser system mounted in a BX-50 WI upright microscope incorporating a 100ϫ LUMPlan FI water immersion objective.This system allows for z axis reconstruction with a theoretical z-slice ϳ0.5 m thick and sequential mode studies in double labeling experiments.Analysis of fluorescent images was performed using the public domain program ImageJ with Plugins for region of interest measurement, image average, comparison of multiple channel images, and co-localization studies.Staining of the ER and the Golgi apparatus was carried out with 100 nM 3,3Ј-dipentyloxacarbocyanine iodide (DiOC 5 (3)) and Bodipy TR ceramide (BTC) (1 g/ml) (Molecular Probes), respectively.Cells transfected with hRIC-3 were treated for 30 min at 37 °C with any of the two markers before visualization.

RESULTS
The hRIC-3 Protein Is Present at the ER and the Golgi Apparatus-The first transmembrane segment of hRIC-3 is adjacent to the N terminus and could play the role of a leader peptide.A prediction program (19) was used to analyze the first 50 amino acids of the hRIC-3 protein.For the sake of comparison we also analyzed the N-terminal regions of the rat ␣7 nAChR and the mouse 5-HT 3 R, which have well defined leader peptides.Results were similar in the three cases (see Fig. 1A, where only the first 30 amino acids of each protein are shown as the rest have a score close to zero), suggesting that hRIC-3 is synthesized in the ER and might be addressed to the ER and/or some other organelle along the secretory pathway, thus facilitating the encounter with ion channels subunits, which transit the same route.In fact, hRIC-3 exhibited an intracellular localization when transfected into different cell lines.We studied its localization in more detail by using confocal microscopy of a transfected hRIC-3 fused to either RFP or GFP together with DiOC 5 (3) or BTC, which are specific markers for the ER (20) and the Golgi apparatus (21), respectively.When hRIC-3-RFP (Fig. 1B) was used in combination with DiOC 5 (3) (Fig. 1C), partial co-localization of both markers at the ER was observed (Fig. 1D).DiOC 5 (3) labeling was more extensive, whereas hRIC-3-RFP expression was restricted to only a frac-tion of the areas labeled by DiOC 5 (3).The presence of hRIC-3-GFP at the Golgi apparatus (Fig. 1E) was evidenced by its co-localization (Fig. 1G) with BTC (Fig. 1F).The detection of the hRIC-3 protein at intracellular membranes is in accordance with previous results obtained in C. elegans that located it within the cell bodies (10) and suggests that this protein may be involved in the membrane trafficking of nAChRs.
Regions of hRIC-3 That Are Required for Its Function-To determine which regions of hRIC-3 are required for its action, we co-expressed several versions of this protein together with different ligand-gated receptors and compared their effects on receptor expression (Fig. 2).As shown previously (11), the whole hRIC-3 protein (construct 1-369) totally inhibited the expression of mouse 5-HT 3 Rs and human ␣4␤2 nAChRs.Deletion of 166 amino acids from the C-terminal region of hRIC-3 (construct 1-203) did not significantly affect the observed inhibition.By contrast, the additional deletion of 78 amino acids (construct 1-125) abolished the inhibition, and receptor levels were similar to the ones observed in the absence of hRIC-3.Given the apparent inhibitory effect of the deleted domain, we tested another protein in which only the C-terminal region was present (construct 134 -369).Again, no inhibition of receptor expression was observed, suggesting that the N-terminal region that contains the transmembrane segments of hRIC-3 and the loop between them might be required for hRIC-3 action.However, deletion of this loop (construct ⌬37-90) did not abolish the inhibition observed with constructs 1-369 and 1-203, and the same happened when a coiled-coil domain present at the C-terminal region was also deleted (construct ⌬137-163).Therefore, these hRIC-3 regions are not needed for the inhibition of receptor expression.
We have shown previously that the presence of hRIC-3 enhances the expression of human ␣7 nAChRs, as judged by the currents evoked by ACh and the binding sites for ␣-Bgt detected at the oocyte membrane (11).The hRIC-3 constructs (Fig. 2) were also co-expressed with human ␣7 nAChRs.The truncated construct 1-203 increased ␣7 nAChR expression, although to a lesser extent than the whole hRIC-3 protein.However, no increase was observed with the constructs that include either the C-terminal (construct 134 -369) or the Nterminal (construct 1-125) regions.Interestingly, when the loop between the transmembrane fragments was deleted (⌬37-90) no activation was observed, suggesting that the loop is necessary for activation.By contrast, activation was still observed when the coiled-coil domain was eliminated (⌬137-163).
An initial approach to determining which region of the receptor subunits might be relevant for the hRIC-3 action consisted of co-injecting the hRIC-3 constructs mentioned above with a chimeric protein made of the extracellular N-terminal region of the rat ␣7 nAChR until Val 201 and the C-terminal region of the mouse 5-HT 3 R, which includes the four transmembrane segments and the large cytoplasmic loop (22).As shown previously by others (22), this so-called V201 chimera produced receptors that were activated by ACh and able to bind ␣-Bgt.The behavior of V201 with the hRIC-3 constructs was similar to that observed with the 5-HT 3 R (Fig. 2); thus, the whole hRIC-3 protein, its truncated version 1-203, and the deletions ⌬37-90 and ⌬136 -163 inhibited receptor expression, whereas the N-terminal and C-terminal regions had no effect when acting individually (constructs 1-125 and 134 -369, respectively).In no case was receptor expression increased, as happened with ␣7 nAChRs, so it can be concluded that the observed inhibitory effects of hRIC-3 were due to the presence in V201 receptors of the C-terminal moiety contributed by the 5-HT 3 R.
hRIC-3 Protein Inhibits Receptor Expression by Preventing Export of Mature Receptors to the Membrane-To determine at which step hRIC-3 acts when inhibiting receptor expression, we used the V201 chimera, because its capacity to bind 125 I-␣-Bgt facilitates follow-up of the receptor fate.Thus, we asked whether the inhibitory effect of hRIC-3 takes place on receptors that have been synthesized and are mature enough for export to the membrane.In this case they should be able to bind ␣-Bgt and show the typical pentameric form.To test for this possibility, we performed binding in the absence or presence of 0.1% saponin and compared the amount of receptors present at the oocyte surface and at internal sites.In addition, we analyzed the oligomeric state of the receptors.As shown in Fig. 3A, in the absence of hRIC-3 most of the V201 receptors are exported to the external membrane.By contrast, only a small fraction of V201 reached the membrane when hRIC-3 was present.Notice that in the presence of hRIC-3 the total amount of V201 was almost as high as in its absence, suggesting that a large population of receptors remained stable inside the oocytes without suffering substantial degradation, despite being unable to reach the external membrane.Sucrose gradient sedimentation experiments revealed that surface V201 receptors expressed in the absence of hRIC-3 migrated as the typical 9 S pentamer (Fig. 3B, V201 External) and their sedimentation profile was indistinguishable when 0.1% saponin was present (Fig. 3B cated intracellularly and also show the same sedimentation behavior (Fig. 3B, V201ϩhRIC-3 Total), suggesting that hRIC-3 arrests the transport of mature receptors to the membrane.
hRIC-3 Protein Enhances Receptor Expression by Acting at Two Different Steps of Receptor Biogenesis-␣-Bgt binding experiments in the presence of saponin were also performed with rat ␣7 nAChRs, in which expression is strongly enhanced by hRIC-3.When hRIC-3 was absent, only a small percentage of the total mature receptors (12.6%) was able to reach the oocyte membrane (Fig. 4).By contrast, in the presence of hRIC-3 almost half of the total number of receptors was transported to the external membrane.Moreover, in this case the total number of receptors increased about 5-fold, suggesting that hRIC-3 acts on ␣7 nAChRs at two different steps: one related to facil-itation of receptor synthesis (or inhibition of receptor degradation) and the other to expedite receptor transport.A further experiment, independent of ␣-Bgt binding, validated the previous results.For this purpose a Western blot of ␣7 nAChRs fused to GFP and expressed in oocytes was probed with an anti-GFP antiserum to show that the total amount of ␣7 subunits did also increase in the presence of hRIC-3 (see Fig. 4, inset).
Another experimental approach based on confocal microscopy observations of COS cells co-transfected with hRIC-3 and nAChRs confirmed the latter results (Fig. 5).Localization of these proteins was possible by attaching RFP to the C terminus of hRIC-3 and GFP to the C termini of ␤2 and ␣7 nAChR subunits.The fusion of GFP did not significantly affect the function of the resultant ␣4␤2 and ␣7 nAChRs when they were expressed in oocytes (results not shown).Cells transfected with only ␣4 and ␤2-GFP subunits showed a broad distribution of fluorescence (Fig. 5A), suggestive of receptors being synthesized and transported to the membrane.However, in the presence of hRIC-3, receptors were present as aggregates inside the cells (Fig. 5B), and the same was true for hRIC-3 (Fig. 5C), so that both proteins co-localized (Fig. 5D).These results would explain why ␣4␤2 receptors are not detected at the oocyte membrane in the presence of hRIC-3 (Fig. 2), confirming what was observed with the V201 chimera (Fig. 3A), that transport of receptors is prevented by hRIC-3.Moreover, the co-localization of both proteins suggests that this impairing action is the consequence of a lasting and close association of hRIC-3 and ␣4␤2 nAChRs, which could be direct or indirect through some other protein.
Cells transfected with only ␣7-GFP nAChRs showed a general distribution of fluorescence (Fig. 5E), probably at internal sites, because ␣7 receptors were not detected at the plasma membrane by ␣-Bgt binding experiments.In the presence of hRIC-3 this pattern was not strongly modified (Fig. 5F), although some aggregates were observed, and in general, labeling was stronger.hRIC-3 labeling (Fig. 5G) only coincided with ␣7 receptor labeling in some places (Fig. 5H) and did not show the exact match observed with ␣4␤2 receptors.It is possible that, in contrast to ␣4␤2 receptors, ␣7 receptors interact transiently with hRIC-3 receptors, so that their transport to the cell membrane is facilitated.In fact, in the presence of hRIC-3 a significant number of binding sites for ␣-Bgt were detected at FIG. 3. The hRIC-3 protein arrests transport of mature V201 receptors to the cell membrane.A, binding of ␣-Bgt to oocytes injected with V201 Ϯ hRIC-3 cRNAs was performed with or without saponin so that the total (open boxes) or only the external (closed boxes) receptors, respectively, were detected.Results are expressed as percentage of the total amount of receptors obtained in the absence of hRIC-3 and in the same conditions as shown in Fig. 2. In the absence of hRIC-3 most of the receptors are able to reach the membrane, whereas in its presence the total amount of receptor remains similar, but only a small fraction is able to reach the external oocyte membrane.B, sedimentation profile on a 5-20% sucrose gradient of 125 I-␣-Bgt-labeled V201 receptors expressed in oocytes in the absence or presence of hRIC-3.Labeling was performed in the same conditions as in A. Fractions of 300 l were collected, and their radioactivity and refraction index were determined.

FIG. 4. The hRIC-3 protein facilitates ␣7 nAChR biogenesis.
Binding of ␣-Bgt to oocytes injected with ␣7 Ϯ hRIC-3 cRNAs was performed with or without saponin in the same conditions described for Fig. 3. Results are expressed as percentage of the total amount of ␣7 receptors obtained in the absence of hRIC-3 and in the same conditions as in Fig. 2. In the absence of hRIC-3 a large proportion of the receptors are unable to reach the membrane, whereas in its presence the total amount of receptors increases about 5-fold, with about half of them transported to the external oocyte membrane.The inset shows a Western blot of a lysate of oocytes expressing ␣7-GFP subunits with (ϩ) or without (Ϫ) hRIC-3 and probed with an anti-GFP antibody.
A Single Amino Acid at the Extracellular Side Close to M1 Is Responsible for the Inhibition Produced by hRIC-3-The fact that V201 is expressed at relatively high levels in mammalian cells gave us the opportunity to study the effect of hRIC-3 in an expression system different from the oocyte system.Moreover, we tried to approach the question of which receptor domain would be involved in the hRIC-3-mediated inhibition.Although ␣4␤2 nAChRs are also inhibited by hRIC-3 (Fig. 2), the use of V201 was more advantageous because its homomeric nature would simplify the production and analysis of mutants and also because receptor expression could be followed easily by ␣-Bgt binding.Expression of V201 receptors at the membrane of transfected COS cells was inhibited in the presence of hRIC-3 (Fig. 6) as observed previously in oocytes.We then tried a construction called N213, which extended the ␣7 sequence 12 more amino acids to the C terminus.No inhibition of hRIC-3 was observed in this case (Fig. 6), and the same happened with two other constructions (T207 and T202), so that we concluded that the presence of a threonine at position 202 instead of isoleucine present in V201 is sufficient to abolish hRIC-3 inhibition.To confirm this result in a wild-type receptor, the same substitution of isoleucine by threonine was made in 5-HT 3 Rs.However, no expression of this mutant was detected, thus precluding study of the possible effect of hRIC-3.
Amino Acids Involved in Activation of nAChR ␣7 Expression-Although constructs T202, T207, and N213 were not inhibited by hRIC-3 (Fig. 6), they did not have the same behavior as the ␣7 subunit, in which expression is strongly increased by hRIC-3, suggesting that this effect is localized in a different receptor domain that is present in ␣7 but not in these constructs, i.e. in a region between Asn 213 and the C terminus.We hypothesized whether a direct or indirect interaction between hRIC-3 and a putative amphipathic helix (24,25) within the C-terminal region of the large cytoplasmic loop of the receptor would be possible, because these motifs are often involved in both intra-and intermolecular protein-protein interactions (26,27).Moreover, within this helix we observed the sequence 424 RFR 426 , which is reminiscent of the ER retention/retrieval motifs RKR of ATP-sensitive K ϩ channels (28) and RXR(R) of GABA B receptors (29).If this hypothesis were true, one would expect that mutants of the ␣7 subunit amphipathic helix might have a different behavior than the wild-type ␣7 subunit in regard to the enhancement of expression induced by hRIC-3.Moreover, this behavior might depend on mutations location.Thus, we mutated to alanines several amino acids that were located at the hydrophobic side of the helix (see Fig. 7, left), Leu 411 , Val 418 , and Phe 425 , and others, Lys 413 and Asn 423 ,  located at the opposite side, which is more polar.In addition, Arg 424 and Arg 426 , which might form the RXR motif, were also mutated.In contrast with the expression increase produced by hRIC-3 on wild-type rat ␣7 receptors (about 14-fold), mutant L411A showed similar expression levels in the presence or in the absence of hRIC-3 (Fig. 7, right); therefore, a single amino acid change is able to abolish the action of hRIC-3.Mutants V418A and F425A, which are located at the same side as L411A, have their expression slightly activated by the presence of hRIC-3 (about 1.7-and 2.2-fold over the control), but this activation did not reach the level observed with the wild-type ␣7.By contrast, mutant K413A, at the other side of the helix, showed strong activation with hRIC-3; the same happened with N423A, although to a lower extent in this case.Therefore, to observe the activating effect of hRIC-3 on ␣7 nAChRs, certain amino acids appear to be required at the hydrophobic side of the amphipathic helix.However, not only these hydrophobic amino acids are important, because mutants R424A and R426A also showed low levels of activation, suggesting that also the motif RXR might be involved in the hRIC-3 activation.

DISCUSSION
The precise mechanisms by which the hRIC-3 protein exerts its effects on nicotinic and serotonergic receptors are not known.In this work we have tried to elucidate at which steps of receptor biosynthesis hRIC-3 acts and which domains of both hRIC-3 and the receptors are involved.
Inhibitory Effect of hRIC-3-Our data show that hRIC-3 hinders the trafficking of 5-HT 3 Rs and ␣4␤2 nAChRs to the plasma membrane.By using the chimera ␣7-5-HT 3 V201 we also observed that the amount of receptors retained is close to the total amount of receptors expressed in the absence of hRIC-3 (Fig. 3).Therefore, it seems that hRIC-3 acts as a barrier without affecting other steps of receptor biogenesis.Retention of the components of multimeric protein complexes might be related to avoiding the surface delivery of improperly assembled proteins, but in this case the retained receptors appear to be fully functional, because they bind ␣-Bgt, they form pentamers, and most of them reach the membrane when hRIC-3 is absent.Therefore, in the case of ␣4␤2 nAChRs and 5-HT 3 Rs, hRIC-3 would act as a full negative regulator at the posttranslational level.Alternatively, we could consider that hRIC-3 is associated to a larger protein complex in which its inhibitory action might be attenuated or restricted by other(s) protein(s) in the complex, so that receptor retention might only be transient.Such a modulation would not take place in the heterologous expression systems that we used if the whole complex were not present or if hRIC-3 outnumbered the other components of the complex.
Regarding the receptor domains that are the target of the hRIC-3 inhibition, our data show that just the presence of an isoleucine instead of a threonine at position 202 is enough to induce the retention of V201 receptors (Fig. 6).The nAChR ␣4 and ␤2 subunits have a valine and an isoleucine at this position, respectively, and make up receptors in which transport is completely inhibited by hRIC-3 (Fig. 2).The same happens with 5-HT 3 Rs (Fig. 2), which also have an isoleucine.By contrast, ␣3␤4 nAChR expression is only partially inhibited (about 50%) by hRIC-3 (11).In this case ␤4 does have an isoleucine, but the ␣3 subunit has a polar tyrosine.Therefore, the presence of a hydrophobic amino acid at this position appears to be required for hRIC-3 inhibition.Recently, a transmembrane motif, PLYFXXN, has been found to control surface trafficking of muscle nAChRs (30).This motif starts at a position equivalent to Pro 207 in V201 and, therefore, is located at a 5-amino acid distance from Ile 202 downward from the C terminus.Moreover, we have previously shown that an arginine (Arg 205 according to the numbering system used here) located between Ile 202 and the latter motif and conserved in all ligand-gated channel subunits is required for transport of assembled ␣7 receptors to the cell surface (31).Also in this region certain domains have been found that affect surface expression of ␣7-5HT 3 chimeric receptors (32).All these data suggest that this region, adjacent to the first transmembrane segment, constitutes a "hot spot" with potential retention/retrieval signals involved in receptor trafficking.Probably, interactions at different places of this region might influence the whole molecular interactions with the transport machinery.
Regarding the domains of hRIC-3 required for this inhibitory effect, our data (Fig. 2) show that the transmembrane fragments, but not the region between them, are necessary but not sufficient.The same occurs with the C-terminal region in which a coiled-coil domain is present, although the latter is not needed for inhibition to be observed.The need for the transmembrane segments may be related to a proper localization of the hRIC-3 protein at the secretory pathway.However, a direct involvement of the transmembrane segments cannot be discarded and will be explored further.Regarding the C-terminal region, and given that the construct with the first 203 amino acids of hRIC-3 was able to inhibit expression, whereas the one with 125 amino acids was not, it could be deduced that the remaining 78 amino acids are responsible for the effect.Moreover, within this stretch we also showed that amino acids 137-163, which conform to the coiled-coil, can be deleted without affecting the inhibition, so that we could conclude that, to observe inhibition, only amino acids 164 -203 would be required.This may be an extreme simplification that should be tested, as other regions might also be needed though in an indirect way.
Activation of ␣7 nAChRs by hRIC-3-Previous studies performed with mutants of hRIC-3 in C. elegans (10) have shown that the DEG-3 neuronal nAChR subunit accumulates in cell bodies; therefore, it appears that hRIC-3 is needed for DEG-3 transport to the cell membrane.Our results show that ␣7 nAChRs are transported very inefficiently to the membrane of Xenopus oocytes, because only about 13% of mature receptors FIG. 7. Expression of mutants of the ␣7 subunit amphipathic helix when co-expressed with hRIC-3.An alanine-mutagenic analysis was performed on the amino acids enclosed by rectangles, which were chosen according to their position at the putative amphipathic helix present at the C-terminal region of the cytoplasmic loop of the rat ␣7 subunit.Thus, Leu 411 , Val 418 , and Phe 425 lay at the hydrophobic side of the helix, whereas Lys 413 and Asn 423 lay at its polar side (the helical wheel diagram looks down the axis of the ␣-helix from the N terminus starting at amino acid Asp-410 to the C terminus finishing at Cys 427 ).Expression was tested by ␣-Bgt binding to receptors present at the oocyte surface, and all data were normalized to the mean obtained in the same oocyte batch with the mutant in the absence of hRIC-3.The mean Ϯ S.E.from at least two experiments with 30 -40 oocytes each is shown for every mutant.ACh-induced currents produced by the mutants had properties similar to those obtained with wild-type ␣7, and their magnitude paralleled ␣-Bgt binding data; therefore, they have not been included.reach their final destination (Fig. 4).In other heterologous expression systems based on transfected cell lines, the situation is even worse, as no functional receptors are detected at the cell membrane (32)(33)(34).However, the presence of hRIC-3 facilitates the transport of ␣7 nAChRs so that 41% of total mature receptors reach the cell membrane.Moreover, the total amount of mature receptors as well as the overall expression levels of ␣7 protein also increased (Fig. 4).These results are consistent with a similar role of hRIC-3 in the expression of the C. elegans DEG-3/DES-2 receptor and the ␣7 nAChR.A recent report (23) pointed out that neither the expression levels of ␣7 protein nor the trafficking of ␣7 receptors to the cell surface was affected by hRIC-3 in a cell culture system.Although the discrepancy could be because of using a different expression system, it also might be related to the different way receptor expression is detected and/or defined.In our case we have considered receptors able to bind ␣-Bgt, in the typical pentameric state, i.e. with a high degree of maturity and functional competence.The total amount of this kind of receptors was clearly increased in the presence of hRIC-3 as was its transport to the oocyte cell membrane.We also observed the same in transfected COS cells (results not shown).
Regarding the domains of the ␣7 subunit involved in the hRIC-3 effects, we have found that the amphipathic helix at the large cytoplasmic loop plays a prominent role.Thus, the effect of hRIC-3 was dramatically reduced when hydrophobic amino acids at one side of the helix were mutated (Fig. 7), suggesting the possibility that direct or indirect interactions with hRIC-3 through these amino acids are disrupted at the mutants.Moreover, mutation of any of the two arginines (Arg 424 and Arg 426 ) of the putative retention motif RXR (29) also reduced the activation produced by hRIC-3, perhaps because the latter shields the retention motif.In a recent work in which a refined structure of the muscle nAChR at 4 Å resolution has been proposed (35), the amphipathic helix of each subunit is shown as an extension of the M4 segment that interacts with the corresponding amphipathic helixes of the neighboring subunits, mainly through their N-terminal and central regions, where Leu 411 , Arg 424 , and Arg 426 would be located.Therefore, it is possible that in the ␣7 subunit the interaction between these helixes is modified by the hRIC-3 protein, and as a consequence, its transport is facilitated.Moreover, considering all of the subunits used in this study and in a previous one (11), the only subunit having the signal RXR in this domain is ␣7, as well as the DEG3DES2 receptor from C. elegans (Fig. 8).
However, an understanding of the action mechanism of hRIC-3 is complicated by the fact that this protein not only facilitates trafficking of ␣7 receptors but also seems to promote other receptor biosynthesis processes that result in an increase in the number of mature receptors (Fig. 4).These processes could be related to the folding, assembly, stabilization, and/or degradation of receptors and will have to be investigated in the future.
The biogenesis of nAChRs is a complex and inefficient process (8,9) in which several cellular factors have been found to be involved, such as BIP (36), calnexin (37,38), 14 -3-3 (39), lynx1 (40), VILIP-1 (41), and the ubiquitin-proteasome system (42).We have shown that the hRIC-3 protein plays opposite roles in this process, acting on different receptor subtypes.In this way, it can exert a powerful posttranslational control of receptor expression.

FIG. 1 .
FIG. 1.Localization of hRIC-3 at the endoplasmic reticulum and the Golgi apparatus.A, the S-score for the signal peptide prediction (19) is reported for every single amino acid position in the hRIC-3, ␣7 nAChR, and 5-HT 3 R N-terminal sequences.High scores indicate that the corresponding amino acid is part of a signal peptide, and low scores indicate that the amino acid is part of a mature protein.The three proteins show a similar pattern for the first 30 amino acids.B-D, depicted are two transfected COS cells expressing hRIC-3-RFP (B), whereas the same and other nontransfected cells around show their endoplasmic reticulum labeled by DiOC 5 (3) (C), so that partial colocalization of both labels can be observed in the merged image (D).E-G, two transfected COS cells express hRIC-3-GFP (E), and the same and other nontransfected cells around show their Golgi apparatus labeled by BTC (F).The merged image shows co-localization of both labels (G).The bar represents 10 m.
FIG. 2. Effects of different hRIC-3 constructs on nAChRs and 5-HT 3 receptors.Expression of receptors in the presence of the indicated hRIC-3 constructs was tested by measuring whole cell currents (open boxes) elicited by the application of 1 mM ACh in oocytes expressing ␣4␤2, ␣7, and V201 receptors or 100 M serotonin in oocytes expressing 5-HT 3 Rs.Expression was also measured by ␣-Bgt binding to oocytes expressing ␣7 and V201 receptors (closed boxes).Data represent the mean Ϯ S.E. of 20 -40 oocytes (n ϭ 2-4 frogs) and are expressed as percentage of controls obtained in the same conditions but without co-expression of the hRIC-3 constructs.The scheme on the left depicts each construct with the transmembrane segments and the coiled-coil domain shown as open and closed boxes, respectively.

FIG. 6 .
FIG. 6.A single amino acid appears to be responsible for the transport inhibition produced by hRIC-3.Different chimeric proteins made of the rat ␣7 N-terminal extracellular domain and the transmembrane segments and cytoplasmic loop of the mouse 5-HT 3 receptor were transfected into COS cells with or without hRIC-3.Expression of receptors at the cell membrane was tested by 125 I-␣-Bgt binding, and the results (mean Ϯ S.E. of at least two experiments by quadruplicate) are expressed as percentage of the binding shown by the same chimera in the absence of hRIC-3.A diagram at the left shows the amino acid sequence of each chimera at the region, where the ␣7 and 5-HT 3 receptor sequences are linked; the latter are italicized and underlined.Each chimera is named according to the amino acid and its numbering in the ␣7 sequence at which the chimeric fusion occurs.