Direct interaction of the resistance to inhibitors of cholinesterase (RIC-3) protein with the serotonin receptor type 3A (5-HT_3A) intracellular domain

Abstract

Pentameric ligand-gated ion channels (pLGIC) are expressed in both excitable and non-excitable cells that are targeted by numerous clinically used drugs. Assembly from five identical or homologous subunits yields homo- or heteromeric pentamers, respectively. The protein known as Resistance to Inhibitors of Cholinesterase (RIC-3) was identified to interfere with assembly and functional maturation of pLGICs. We have shown previously for serotonin type 3A homopentamers (5-HT_3A) that the interaction with RIC-3 requires the intracellular domain (ICD) of this pLGIC. After expression in Xenopus laevis oocytes RIC-3 attenuated serotonin-induced currents in 5-HT_3A wild-type channels, but not in functional 5-HT_3AglvM3M4 channels that have the 115-amino acid ICD replaced by a heptapeptide. In complementary experiments we have shown that engineering the Gloeobacter violaceus ligand-gated ion channel (GLIC) to contain the 5-HT_3A-ICD confers sensitivity to RIC-3 in oocytes to otherwise insensitive GLIC. In the present study, we identify endogenous RIC-3 protein expression in X. laevis oocytes. We purified RIC-3 to homogeneity after expression in Echericia coli. By using heterologously overexpressed and purified RIC-3 and the chimera consisting of the 5-HT_3A-ICD and the extracellular and transmembrane domains of GLIC in pull-down experiments, we demonstrate a direct and specific interaction between the two proteins. This result further underlines that the domain within 5-HT_3AR that mediates the interaction with RIC-3 is the ICD. Importantly, this is the first experimental evidence that the interaction between 5-HT_3AR-ICD and RIC-3 does not require other proteins. Additionally, we demonstrate that the pentameric assembly of the GLIC-5-HT_3A-ICD chimera interacts with RIC-3.

The Cys-loop family of pentameric ligand-gated ion channels (pLGICs) includes cation-conducting channels such as nicotinic acetylcholine (nACh), and serotonin type 3 (5-HT₃), as well as anion-conducting members like γ-aminobutyric acid (GABA) and glycine (Gly) receptors. The protein Resistance to Inhibitors of Cholinesterase (RIC-3) is a crucial mediator for functional maturation of members of the eukaryotic pLGIC superfamily, also known as Cys-loop receptors, based on a disulfide-linked loop in the extracellular domain of these receptors¹. Mutations in ric-3 were originally identified in a screen in Caenorhabditis elegans that was geared towards identifying suppressors of a dominant mutation in the acetylcholine receptor subunit DEG-3 acetylcholine receptor². Subsequently, it was found that RIC-3 modulates functional maturation of non-C. elegans acetylcholine receptors and also other pLGIC superfamily members, e.g. nAChRa7, nAChRα4β2, nAChRα3β4, 5-HT_3A^2–3. RIC-3 is highly charged. Specifically, human RIC-3 protein (Q7Z5B4-1) has a high charge density with 17% amino acids with carboxylic acid sidechain (5% aspartic acid and 12% glutamic acid), and with 12 % basic amino acids (5% arginine and 7% lysine). Human RIC-3 is predicted to be a single-pass membrane protein mainly residing in endoplasmic reticulum (ER) membranes with the N-terminus in the lumen and the C-terminal coiled-coil domain in the cytoplasm^{2, 4}. However, an alternative topology with two transmembrane segments and the N- and C-termini both in the cytoplasm is predicted by other algorithms that identify a transmembrane segment instead of a signal sequence^4–6. Surface biotinylation studies additionally identified RIC-3 located on the cell surface after expression in HEK293 cells⁷. It was previously shown by co-immunoprecipitation studies, that RIC-3 interacts with nAChR α3, α4, α7 as well as β2 and β4 subunits^7–8. In several cell types including Xenopus laevis oocytes and HEK cells, it was shown that RIC-3 co-expression is required for functional expression of homomeric nAChR α7. In contrast, for heteromeric nAChR α3β4 and α4β2 as well as homomeric 5-HT_3AR RIC-3 co-expression leads to inhibition of functional maturation in X. laevis oocytes^3–4. In HEK cells however, RIC-3 enhanced 5-HT_3A functional expression⁹. Additionally, a study investigating the surface delivery of α-bungarotoxin-binding sites (BgTRs) contributed by α7 subunits found that low levels of RIC-3 promote bungarotoxin binding, and high levels suppress binding¹⁰. It was postulated that low levels of RIC-3 lead to short-lived interactions with α7 and promote BgtR assembly and ER release, whereas at high RIC-3 levels interactions are longer-lived and result in ER retention. At present, it is not clear why RIC-3 modulation has opposing effects on different pLGIC, and in which manner expression levels alter its impact on pLGIC maturation. In a first step to elucidate the nature of modulation of pLGIC by RIC-3, it is essential to probe whether these proteins interact directly with each other, or whether other protein factors are required.

While previous studies indicated that RIC-3 modulates the functional expression of these receptors either by inhibiting or promoting their functional expression, it is also apparent, that other host cell factors contribute to the activity of RIC-3^{7–9, 11–12}. We therefore questioned whether RIC-3 interacts with pLGICs in a direct manner or whether other protein factors mediate the interaction. Given our extensive previous work probing the interaction of RIC-3 with 5-HT_3A, we wanted to investigate this question using 5-HT_3A-chimeras^13–15. Our hypothesis was that the interaction of RIC-3 and 5-HT_3A is directly between these two proteins, as opposed to mediated by other proteins. The results of the present investigations will be crucial in identifying which segments of the two proteins mediate the interaction, and also for further studies probing the mechanism of interaction. Previously, by using a chimera approach and expression in X. laevis oocytes, we have demonstrated, that for 5-HT_3A receptors the intracellular domain (ICD) is required for modulation by RIC-3. This was shown with two sets of complementary chimeras. 5-HT_3A receptors were inhibited by co-expression of RIC-3 in X. laevis oocytes, whereas the removal of the ICD yielded functional channels for which the functional expression on the plasma membrane was not modulated by RIC-3¹³. Additionally, when the 5-HT_3A-ICD was added to the Gloeobacter violaceus ligand-gated ion channel (GLIC), that naturally lacks an ICD and only contains a short loop between the transmembrane segments M3 and M4, this chimera conferred RIC-3 sensitivity to otherwise insensitive GLIC^14–15. In the present study, we utilized a functional chimera GLIC-5-HT_3A-ICD between the extracellular and transmembrane domains of GLIC and the ICD of 5-HT_3AR (Figure 1). The rationale for using this construct is two-fold. First, the GLIC-5-HT_3A-ICD chimera established that the ICD mediates the interaction with RIC-3. Second, GLIC chimeras are amenable to high-level overexpression and purification from Echericia coli (E. coli), a cell system that naturally lacks the expressed proteins and additionally other proteins that are known to interact with the proteins of interest. Human RIC-3 and the GLIC-5-HT_3A-ICD chimera were expressed individually in E. coli and purified to homogeneity. These purified proteins were then subjected to pull-down assays indicating that the interaction between these two proteins is direct and not mediated by other proteins.

Figure 1. Domain architecture of pentameric ligand-gated ion channels and their chimeras used in this study.

Both 5-HT_3A (A) and GLIC (B) contain a transmembrane (TMD) and extracellular domain (ECD). The latter contains a short α-helical segment only in 5-HT_3A. The transmembrane domain consists out of four α-helical segments (M1-M4). 5-HT_3A receptors contain a 115-amino acid intracellular domain (ICD) between M3 and M4, whereas GLIC only has a short 4-amino acid linker (ESQP). We obtained chimeras by exchanging the ICDs between these two channels. 5-HT_3AglvM3M4 (C) contains the ECD and TMD of the 5-HT_3A receptor and a M3M4 linker from GLIC, SQPARAA. GLIC-5-HT_3A-ICD (D) contains the ECD and TMD of GLIC and the 5-HT_3A-ICD of 115-amino acids.

Methods

Constructs

A chimera between GLIC and the 5-HT_3A-ICD was obtained by replacing 10 amino acids in the M3-M4 loop region of GLIC (KVESQPARAA, α-helical positions in M3 and M4 underlined) with the entire 5-HT_3A-ICD consisting of 115 amino acids (QDLQRPVP…RDWLRVGY). The chimera was generated similarly in both the pXOON vector for Xenopus laevis oocyte expression and the pET26b vector for E. coli expression, as described previously¹⁴. In the pET26b vector, the chimera follows a pelB signal sequence, polyhistidine (His₁₀) tag, maltose binding protein (MBP) tag and human rhino virus (HRV)-3C cleavage site¹⁷. The identity of all constructs was verified by DNA sequencing (Genewiz, South Plainfield, NJ) of the complete coding region. Human RIC-3 (NM_001206671.2) in pGEMH19 expression vector was a generous gift Dr. Millet Treinin. RIC-3 was cloned into the prokaryotic expression vector pMAL-c2x (New England Biolabs) as a fusion construct with N-terminal MBP and C-terminal His₆ tags⁵⁵. This construct was then transferred back into pGEMHE vector as MBP-RIC-3 for expression in Xenopus laevis oocytes.

Oocytes

Oocytes were obtained from EcoCyte Bioscience, Austin, USA, or harvested in house as described previously, following procedures for maintenance and surgery of the frogs that were approved by the local animal welfare committee (Institutional Animal Care and Use Committee, IACUC #: 08014, PHS Assurance # A 3056-01).

Oocyte Expression

GLIC-5-HT_3A-ICD pXOON chimera plasmid was linearized with XbaI and RIC-3 pGEMH19 and MBP-RIC-3 pGEMHE plasmids were linearized with NheI for in vitro transcription using T7 RNA polymerase (mMESSAGE mMACHINE kit, Applied Biosystems/Ambion, Austin, TX). Mouse 5-HT_3A and 5-HT_3AglvM3M4, a construct with the 115-amino acid ICD replaced by SQPARAA, in pGEMHE were linearized with XbaI for cRNA preparation. Capped cRNA was purified with the MEGAclear kit, from Applied Biosystems/Ambion and then precipitated with ammonium acetate. cRNA was dissolved in nuclease-free water and stored at −80 °C. The integrity of the cRNA was tested by agarose gel electrophoresis. X. laevis oocytes were harvested and defolliculated as described previously¹⁴. Unless otherwise noted, oocytes were injected 24 h after isolation with 10 ng of ion channel-coding cRNA (0.2 µg/µl) and kept in standard oocyte saline medium (SOS, 100 mM NaCl, 2mM KCl, 1mM MgCl₂, 1.8mM CaCl₂, 5 mM HEPES, pH7.5) supplemented with 100 IU/ml penicillin, 100 µg/ml streptomycin, and 250 ng/ml amphotericin B (Invitrogen) and 5% horse serum (Sigma-Aldrich) for 2–7 days at 16 °C. For co-expression experiments ion channel-coding cRNA (5 ng at 0.2 µg/µl) and RIC-3 cRNA (2.5 ng at 0.2 µg/µl), or a molar equivalent of MBP-RIC-3 cRNA (5 ng at 0.2 µg/µl) was co-injected. Recordings were performed 48 hours or later after injection.

Oocyte two-electrode voltage-clamp (TEVC) experiments

TEVC experiments were conducted 2–5 days after cRNA injection at a holding potential of −60 mV and at room temperature. A ~250 µl chamber harboring a single oocyte was continuously perfused at a rate of 5–6 ml/min with the appropriate buffer: either GLIC oocyte recording buffer (GORB, 100 mM NaCl, 20 mM NaOH, 2.5 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 5 mM HEPES, 5 mM Citric Acid, pH adjusted to 7.5 or as indicated with HCl), or oocyte recording buffer 2 (OR2, 115 mM NaCl, 2.5 mM KCl, 1.8 mM MgCl₂, 10 mM HEPES, pH 7.5 with NaOH) containing serotonin. The ground electrode was connected to the bath by a 3 M KCl/agar bridge. The glass microelectrode resistance was < 2 MΩ when filled with 3 M KCl. Data were acquired and analyzed using a TEV-200 amplifier (Dagan Instruments, Minneapolis, MN), a Digidata 1440A data interface (Molecular Devices, Sunnyvale, CA), and pClamp 10.4 software (Molecular Devices). All experiments were performed on at least 3 oocytes from two different batches of oocytes.

RIC-3 expression in oocytes

Uninjected oocytes and RIC-3 injected oocytes were homogenized in lysis buffer (0.3 M sucrose, 10 mM sodium phosphate buffer pH 7.5, protease inhibitor cocktail) on ice. After homogenization samples were centrifuged at 3,000 × g for 10 min at 4 °C. The clear middle layer was mixed with Laemmli buffer for separation on 4–15 % TGX gels and electrophoresed and blotted to PVDF membranes. Blotting and detection was conducted as described below.

GLIC wild-type and GLIC-5HT_3A-ICD chimera expression in E. coli and purification

GLIC and GLIC-5-HT_3A-ICD chimera were expressed and purified essentially as described previously for GLIC crystallization^{15, 17}. In brief, constructs were overexpressed in E. coli C41 cells after induction with IPTG overnight at 18°C. All subsequent steps were performed at 4 °C. Cells were harvested and lysed using a microfluidizer (Microfluidics, Inc.). Membranes were isolated by ultracentrifugation, solubilized in a 2% n-dodecyl-β-D-maltoside (DDM) (Anatrace) buffer, purified by affinity chromatography (nickel or amylose) (New England Biolabs), and the maltose binding protein tag was cleaved with HRV-3C protease (Sino Biological Inc.). Finally, GLIC or chimera were subjected to gel filtration on a Superdex 200 10/300 column (GE Healthcare)⁵⁶. Peak protein samples corresponding to pentameric GLIC or GLIC-5-HT_3A-ICD were pooled and concentrated.

MBP-RIC-His6 expression in E. coli and purification

The MBP-RIC-3-His₆ fusion construct was overexpressed in E. coli BL21-CodonPlus® (DE3)-RIPL competent cells using Terrific Broth (TB) medium supplemented with ampicillin (100 µg/mL) and chloramphenicol (34 µg/mL). Cultures were grown at 37 °C to an optical density (OD₆₀₀) of 0.6. Expression was induced with 0.2 mM IPTG for eight hours at 18 °C. Cells were harvested by centrifugation at 6,500 rpm for 15 min at 4 °C and the resulting cell pellet was resuspended in lysis buffer (20 mM Tris-HCl, pH 7.4 and 0.2 M sodium chloride) containing the following protease inhibitors: 1 mM PMSF, 10 µg/mL leupeptin, and 7 µg/mL pepstatin. Lysozyme and DNaseI were also added to a final concentration of 1 mg/mL and 20 µg/mL, respectively. The resuspended cells were then further lysed using an EmulsiFlex-C3 French press-style high-pressure homogenizer (AVESTIN, Inc., Ottawa, ON, Canada). Unbroken cells and debris were removed from the cell lysate by centrifugation (40 min at 12,500×g, 4 °C); and the membranes were isolated by ultracentrifugation (60 min at 100,000×g, 4°C). These initial small-scale purifications involved harvesting a 50–100 mL culture to yield a total of up to 100 µg of MBP-RIC-3-His₆.

For investigating the extraction and recovery properties of different detergents, standard Tris lysis buffer or a phosphate buffer (50 mM potassium phosphate, pH 7.4 and 150 mM sodium chloride) were modified by addition of detergents as follows: in phosphate buffer, 0.5% DDM, 0.05% Triton X-100, 0.8% CHAPS, 0.1% C12E9, and 0.8% sodium cholate; in Tris buffer, 0.5% DDM. Membrane pellet aliquots were treated with the respective detergent buffer for 2 hours at 4 °C. The solubilized fraction was separated from insolubilized material by ultracentrifugation (60 min at 100,000g, 4°C).

For large-scale preparation and for use in pull-down experiments, membranes from 2–4 L of cultures were solubilized with 0.5% DDM in Tris lysis buffer and protease inhibitors under agitation at 4°C for two hours, and the solubilized fraction was cleared by ultracentrifugation (60 min at 100,000g, 4°C). Solubilized proteins were purified by affinity chromatography using an amylose-resin column and subsequently a His-Pur Cobalt or TALON superflow resin column. The purified fusion protein was concentrated using Amicon® Ultra – 15 centrifugal filters with a 50 kDa molecular weight cutoff. The concentrated protein was filtered through a 0.22 µm poresize low protein binding Durapore® membrane, and stored at 4°C with 2× protease inhibitor cocktail. For the last purification step protein was subjected to SEC on a Superdex 200 10/300 column. The yield was 0.5 mg for 2 L of culture. For control experiments MBP was purified from pMAL-c2x similarly, except that the resuspended cells were lysed using a sonicator (QSONICA, LLC, Newtown, CT) and a single affinity purification step of the soluble fraction with amylose resin was used in Tris buffer.

Pull-down assay between 5-HT_3A-ICD and RIC-3

MBP-RIC-3-His₆ (2.5 µg) was immobilized on amylose resin by incubation for 30 min on ice, with occasional flicking. After repeated washes to remove unbound protein, purified GLIC WT and GLIC-5-HT_3A-ICD (10 µg each) were added to separate tubes and allowed to incubate with amylose resin-immobilized MBP-RIC-3 for 30 min on ice, with occasional flicking. After repeated washes to remove unbound protein, the proteins were eluted from the resin by adding 2× Laemmli buffer and boiling at 95 °C for 5 min. MBP without RIC-3 was used as a control. Samples were separated by sodium dodecylsulfate polyacrylamide electrophoresis (SDS-PAGE) using 4 – 15% precast gradient Mini PROTEAN TGX Stain-Free gels (Bio-Rad). Resolved protein bands were visualized by stain-free enabled imager (Gel Doc_TM EZ Imager, Bio-Rad) or Coomassie Blue Staining. The pull-down assay was also conducted using differing ratios between MBP-RIC-3 (2.5 to 15 µg) and GLIC-5-HT_3A-ICD (10 µg).

Immunoblotting

Proteins were separated using 4 – 15% precast gradient Mini PROTEAN TGX Stain-Free gels (Bio-Rad), blotted onto PVDF membranes, blocked overnight at 4°C with 5% milk in Tris-buffered saline containing Tween 20, probed with a primary murine polyclonal antibody raised against human RIC-3 (Novus Biologicals, Littleton, CO) at 1:500 or 1:1,000 dilution for four hours at room temperature, and then with a goat anti-mouse HRP conjugated secondary antibody at 1:100,000 dilution for two hours at room temperature. Blots were developed with SuperSignal West Femto Substrates using a digital imaging system (ImageQuant™ LAS 4000, GE Healthcare).

Statistics

Statistical significance was determined with one-way ANOVA and Dunnett’s multiple comparisons test (**** denotes p < 0.0001) of RIC-3 or MBP-RIC-3 co-injected oocytes vs pLGIC alone (Prism 6, GraphPad Software, Inc.).

Results

Maltose binding protein (MBP) tagged RIC-3 and RIC-3 both inhibit agonist-induced current amplitudes of 5-HT_3A-ICD containing receptors

We introduced a MBP-tag N-terminal to RIC-3 to facilitate its use as a bait protein in pull-down assays. We showed previously that co-expression of human RIC-3 with mouse 5HT_3A significantly attenuated 5-HT-induced currents in X. laevis oocytes¹³. This effect was abolished upon replacement of the ICD with a short 7-amino acid linker (SQPARAA) that is found in the M3-M4 linker region of GLIC in the 5-HT_3AglvM3M4 construct. In the opposite chimera, when the 5-HT_3A-ICD was inserted into GLIC to obtain GLIC-5HT_3A-ICD chimeras, RIC-3 co-expression significantly inhibited proton-induced currents, but not when it was co-expressed with wild-type GLIC¹⁵. To probe if MBP-RIC-3 was similarly able to interfere with functional 5HT_3AR maturation, we compared co-expression of either MBP-RIC-3 or RIC-3 together with 5-HT_3AR in X. laevis oocytes (Figure 2). Three groups of oocytes were injected, the first only with cRNA for wild-type 5-HT_3A (5 ng), the second with cRNA for wild-type 5-HT_3A (5 ng) together with RIC-3 cRNA (2.5 ng), and the third with cRNA for wild-type 5-HT_3A (5 ng) together with MBP-RIC-3 cRNA (5 ng, equimolar to 2.5 ng RIC-3). Both co-injection of RIC-3 or MBP-RIC-3 decreased the currents induced by application of the agonist 5-HT significantly, as compared to injection of 5-HT_3A alone, thereby confirming that the presence MBP-RIC-3 modulates 5-HT_3A functional maturation similar to RIC-3, and can thus be used as a bait protein (Figure 2). Neither the construct MBP-RIC-3 or RIC-3, when co-expressed with 5-HT_3AglvM3M4, a construct that lacks the 5-HT_3A-ICD, had any effect on agonist-induced current amplitudes. Similarly, proton-induced current amplitudes in GLIC were not altered by RIC-3 or MBP-RIC-3 co-expression, whereas the current amplitudes were significantly attenuated in the GLIC-5-HT_3A-ICD chimera. These results indicate, that the 5-HT_3A-ICD is required for the modulation by both RIC-3 and MBP-RIC-3.

Figure 2. RIC-3 and MBP-RIC-3 modulate functional maturation of 5-HT_3A-ICD-containing receptors.

Agonist induced current amplitudes (nA) were measured 48 hours or later after injection of cRNA into X. laevis oocytes in two-electrode voltage-clamp experiments. Currents recorded in oocytes injected with pLGIC cRNA (black bars), after co-expression of pLGIC together with MBP-RIC-3 (grey bars), and after co-expression of pLGIC with RIC-3 (white bars) are given for each pLGIC. Statistical significance was determined with one-way ANOVA and Dunnett’s multiple comparisons test (**** denotes p < 0.0001). Note that the currents were almost completely abolished for MBP-RIC-3 and RIC-3 co-expression with GLIC-5-HT_3A-ICD and bars are therefore not visible.

RIC-3 is endogenously expressed in X. laevis oocytes

RIC-3 mRNA homologues have been identified in X. tropicalis as well as laevis¹⁶. Specifically, RIC-3 mRNA was identified in brain and spinal cord samples. Here we show that a murine polyclonal antibody developed against full-length human RIC-3 detects a protein corresponding to a size of 56 kDa (Figure 3) in X. laevis oocyte lysates. Upon injection of RIC-3 cRNA, this band intensified. The running behavior corresponds to previous reports⁷. The predicted weight of both X. as well as Homo sapiens RIC-3 that have 49 % sequence identity (BLAST) is 41 kDa. This result indicates that RIC-3 is indeed natively expressed in X. laevis oocytes.

Figure 3. Identification of endogenous RIC-3 protein in X. laevis oocytes.

Cleared whole cell lysates of uninjected (uninj.) or RIC-3 cRNA-injected oocytes were separated on TGX gels, and probed with RIC-3 antibody (Novus, 1:500) and HRP-conjugated secondary antibody (1:100,000). Uninjected oocytes show a faint band of endogenous RIC-3 detected by the antibody that intensifies upon injection with RIC-3 cRNA.

GLIC and GLIC-5-HT_3A-ICD are purified from E. coli plasma membrane fractions

GLIC and GLIC-5-HT_3A-ICD were purified following previously published protocols (Figure 4)¹⁷. Both proteins localized to the membrane fraction in E. coli. GLIC and GLIC-5-HT_3A-ICD chimera were solubilized with n-dodecyl-β-D-maltoside (DDM). Molecular sizing using size exclusion chromatography (SEC) on a Superdex 200 10/300 column indicates a pentameric assembly of both proteins. The pentameric assembly was further confirmed using dynamic light scattering. Additionally, GLIC-5-HT_3A-ICD when injected into oocytes as a purified protein, leads to proton-induced currents upon exposure to acidic pH buffer¹⁵. Consequently, the GLIC-5-HT_3A-ICD chimera used for subsequent pull-down experiments represents a functional pentameric assembly, as opposed to monomers or assembly intermediates.

Figure 4. Purification of GLIC and GLIC-5HT_3A-ICD chimera.

Gels shown are Coomassie-stained TGX SDS PAGE gels. (A) MBP-GLIC (calculated molecular mass 83 kDa) was purified by amylose resin chromatography of DDM-solubilized membranes. Stars, *, indicate two minor impurities after this first purification step. HRV protease was used to cleave MBP-GLIC into MBP (calculated molecular mass 46 kDa) and GLIC (calculated molecular mass 37 kDa). MBP carries an N-terminal pelB signal sequence and His₁₀. Size-exclusion chromatography (SEC) subsequently provided pure GLIC. Major proteins are indicated to the right of the Coomassie-stained gel. SEC chromatogram shown to the right. The peak corresponding to GLIC is indicated. (B) GLIC-5-HT_3A-ICD chimera was purified similarly. Coomassie-stained SDS PAGE gel compares MBP-GLIC and MBP-GLIC-5-HT_3A-ICD chimera (calculated molecular mass 95 kDa) fractions obtained by amylose resin chromatography. Cleavage by HRV yielded MBP and GLIC-5-HT_3A-ICD (calculated weight 49 kDa) that run as one band under the conditions shown. (C) GLIC-5-HT_3A-ICD chimera SEC chromatogram indicates homogeneous pentameric assemblies (calculated molecular mass 245 kDa) that were readily separated from MBP by SEC.

RIC-3 is purified from E. coli plasma membrane fractions

Human RIC-3 protein is predicted to contain a signal peptide and a single transmembrane span. We purified MBP-RIC-3-His₆ using the E. coli membrane fraction (Figure 5). To extract MBP-RIC-3 from the membrane fraction, we intended to use DDM, since this is the detergent that has been extensively used for studies of GLIC¹⁷. Additionally, we have succeeded in purifying the GLIC-5-HT_3A-ICD chimera as a stable pentamer using this non-ionic detergent^14–15. In initial screens, we briefly tested five detergents for solubilization of RIC-3, DDM, 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate hydrate (CHAPS), Triton X-100, polyoxyethylene (9) dodecyl ether (C₁₂E₉), as well as sodium cholate (Figure 5). While the ionic detergents CHAPS and sodium cholate extracted approximately two-fold more RIC-3 protein, the milder detergent DDM solubilized RIC-3 in significant amounts and does not carry the risk of denaturing membrane proteins¹⁸. Since GLIC and GLIC chimera were purified using DDM, we decided to use DDM for further studies of RIC-3 as well, so that a single detergent was used for purifying all membrane proteins. Purified MBP-RIC-3-His₆ was recognized by an antibody directed against human RIC-3 (Figure 5). Size exclusion chromatography (SEC) on a Superdex 200 10/300 column produced highly-purified protein as represented by a symmetric peak in the SEC profile (Figure 6).

Figure 5. Purification of MBP-RIC-3-His₆.

(A) MBP-RIC-3-His₆ solubilized with DDM was purified first by amylose affinity chromatography (lanes 1–4) and then by metal-affinity chromatography using either HisPur cobalt (lanes 5–8) or nickel (lanes 10–14) resin. The last lane for each resin corresponds to final purified fraction from that purification step. Note that after amylose purification there are several other protein bands visible, most notably one corresponding to MBP. (B) Detergent screen with various detergents in phosphate buffer as indicated above each lane, or using DDM in Tris buffer. (C) Purified MBP-RIC-3-His₆ protein (theoretical weight 84 kDa) is detected by an antibody directed against human RIC-3.

Figure 6. Purification of MBP-RIC-3-His₆ to homogeneity by size exclusion chromatography (SEC).

After a two-step affinity chromatography (amylose and metal affinity), MBP-RIC-3-His₆ was passed through a SEC Superdex 200 10/300 column which produced highly-purified protein as represented by a symmetric peak on the SEC profile (A), and as visualized on an analytical SDS-PAGE gel (B).

MBP-RIC-3 pulls down GLIC-5-HT_3A-ICD

We expressed GLIC and GLIC-5-HT_3A-ICD in E. coli and purified their pentameric assemblies to homogeneity (Figure 4). Additionally, we expressed in E. coli and purified to homogeneity MBP-RIC-3-His₆ (Figure 5 and Figure 6). MBP-RIC-3-His₆ was used in pull-down experiments with wild-type GLIC and GLIC-5-HT_3A-ICD chimera. MBP-RIC-3-His₆ bound to GLIC-5-HT_3A-ICD chimera, but not to GLIC (Figure 7). This indicates a direct interaction between RIC-3 and GLIC-5-HT_3A-ICD chimera, as opposed to an interaction mediated by or requiring additional proteins/chaperones, since the only proteins present were the purified starting materials, contrary to the complex interaction networks possible when co-expressing in X. laevis oocytes or mammalian cell lines. Since the 5-HT_3A-ICD is only present in the chimera and not GLIC WT, we infer that the interaction occurs between the ICD and RIC-3.

Figure 7. Direct interaction between GLIC-5-HT_3A-ICD and MBP-RIC-3.

Coomassie-blue stained gels or stain-free (Biorad) gels showing the results of the pull-down assay for MBP-RIC-His₆ (RIC-3) (84 kDa) and MBP (44 kDa) with GLIC-5-HT_3A-ICD (49 kDa) and GLIC (37 kDa). (A) Input lanes and samples after elution from amylose resin shown. Note that RIC-3 pulls down GLIC-5-HT_3A-ICD, but not GLIC. (B) Using successively more RIC-3 (2.5 through 15 µg) and a constant amount of GLIC-5-HT_3A-ICD (10 µg) and separation by amylose beads leads to a proportional increase in pulled down GLIC-5-HT_3A-ICD.

Discussion

Recently published X-ray structures of both prokaryotic and eukaryotic pLGICs have shown that both the agonist-binding ECD as well as the TMD are structurally well conserved^{17, 19–26}. Only eukaryotic pLGICs, however, contain an ICD between the third (M3) and fourth (M4) transmembrane segments of the TMD, whereas these two segments are linked by a short linker of up to 15 amino acids in prokaryotes²⁷. Intriguingly, the ICD is also the most diverse domain in eukaryotic pLGICs, exemplified by lengths ranging from 50 to 280 amino acids. Inspired by the demonstration that GLIC was indeed a functional ligand-gated ion channel²⁸ we developed chimeras in which the short M3-M4 linker of GLIC was exchanged for the large ICD in eukaryotic pLGICs. We have shown that functional channels are obtained upon replacement of the ICD in both anion- (GABAρ1) and cation-conducting (5-HT_3A) pLGIC by a short heptapeptide linker (SQPARAA)¹³. Subsequently, several other studies investigated comparable chimeras or deletions of the long M3-M4 linker, essentially confirming that the ICD can be removed while maintaining the overall ability to fold, assemble into pentamers, traffick to the plasma-membrane, and function as ion channels^29–32. The same heptapeptide linker yielded functional Glyα1 chimeras³³. Chimeras between the ECD of nAChR α7 and the TMD and ICD GluClβ functioned with the ICD removed³⁴. The ICD of eukaryotic Cys-loop receptors has been shown to play a crucial role in many functions such as assembly, trafficking, sorting, and anchoring of the receptors, as well as post-translational modifications^{2, 4, 35–42}. Interactions of GlyR with G-protein coupled receptors⁴³, which can change GlyR potentiation by ethanol, are also mediated by the ICD⁴⁴. The ICD has been implicated in associating with a number of different cytosolic proteins such as 14-3-3η⁴⁵, rapsyn^46–47, gephyrin⁴⁸, BiP⁴⁹, calnexin⁵⁰, Src kinases⁴³, and RIC-3⁵¹. Specific sites within the ICD can have very specific functional effects. For example, mutations of the α1 subunit in GABA α1β2γ2s receptors lead to altered agonist apparent affinities, or changes in desensitization kinetics. Another study showed that the sensitivity of nAChR α4β2 towards nicotine may be abolished by inhibiting PKC sites while preserving normal function of these channels⁵². Among all of these chaperones, RIC-3 is one of the few that has been studied extensively in multiple expression systems and with different receptor subtypes. It has already been shown to be implicated strongly in post-translational control of nAChRs and 5-HT_3ARs and has been shown to be mandatory for expression of nACh α7 receptors in different expression systems⁵¹. We have shown previously, that chimeras combining the ECD and TMD of GLIC with the ICD of 5-HT_3A are modulated by RIC-3 co-expression in X. laevis oocytes^14–15. Such a modulation is not observed for GLIC co-expressed with RIC-3, indicating that the RIC-3 modulation requires the 5-HT_3A-ICD. Similarly, we have shown that RIC-3 modulation of 5-HT_3A is abolished when the ICD is replaced by a heptapeptide linker¹³.

While these previous experiments show that RIC-3 modulates 5-HT_3A functional expression, the co-expression studies in oocytes cannot distinguish between an interaction that involves direct binding of the two proteins, or a modulation that requires other proteins factors. Since oocytes contain endogenously expressed RIC-3, it is possible, that other proteins of such a more complex modulatory system are also present in oocytes. Here, using proteins expressed and purified from E. coli, we show, that GLIC-5-HT_3A-ICD chimera interacts with RIC-3. Importantly, our results indicate that this interaction is not mediated by other proteins that may be present during immunoprecipitation experiments investigating protein-protein interactions using proteins in complex cellular environments. Thus, we provide evidence for a direct interaction between the GLIC-5-HT_3A-ICD chimera and RIC-3. Future studies are required to determine the nature of the interaction between the ICD and RIC-3, since only the linear amino acid sequence but not the three-dimensional structure of either protein is known at present. The postulated mechanism of RIC-3 modulation of pLGIC plasma-membrane expression involves RIC-3 interacting with pLGIC monomers or assembly intermediates within the ER^{4, 6, 8, 53}. However, surface-biotinylation studies have also indicated the presence of RIC-3 on the plasma membrane⁷. We show, that pentameric assemblies of GLIC-5-HT_3A-ICD chimeras directly interact with RIC-3. This further substantiates the observation of RIC-3 not only in the ER, but also on the plasma membrane. Assembly of pLGIC subunits occurs in the ER where their interaction with RIC-3 may occur, however, further folding completes after the formation of pentamers⁵⁴. Experimental evidence indicates that this functional maturation occurs on the plasma membrane, and importantly, is mediated by RIC-3⁷.

Conclusions

In summary, we provide direct experimental evidence for a direct interaction between the chaperone protein RIC-3 and the 5-HT_3A receptor ICD. By using highly-purified proteins, we demonstrate that the interaction does not necessitate other protein factors, and that the 5-HT_3A-ICD in its pentameric assembly state interacts with RIC-3. In addition, to the best of our knowledge, this is the first study demonstrating the heterologous overexpression and purification to homogeneity of RIC-3. Our results provide the basis for further establishing the specific interacting domains as well as mechanisms of interaction of the 5-HT_3A-ICD and RIC-3. The ICD or also RIC-3 represent promising targets for neurological drugs that involve disturbances of pLGIC functional expression.

Abbreviations used

BgTRs

α-bungarotoxin-binding sites

CHAPS

3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate hydrate

DDM

n-dodecyl-β-D-maltoside

ECD

extracellular domain

GABA

γ-amino butyric acid

ER

endoplasmic reticulum

GLIC

Gloeobacter violaceus ligand-gated ion channel

GlyR

glycine receptor

GORB

GLIC oocyte recording buffer

5-HT₃

5-hydroxy tryptamine (serotonin) Type 3

ICD

intracellular domain

pLGIC

pentameric ligand-gated ion channel

MBP

maltose binding protein

nACh

nicotinic acetylchoine

RIC-3

resistance to inhibitors of cholinesterase type 3

SEC

size exclusion chromatography

TEVC

two-electrode voltage clamp

TMD

transmembrane domain