Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Jun 1.
Published in final edited form as: Neuropharmacology. 2008 Mar 27;54(8):1153–1165. doi: 10.1016/j.neuropharm.2008.03.009

The L293 Residue in Transmembrane Domain 2 of the 5-HT3A Receptor is a Molecular Determinant of Allosteric Modulation by 5-Hydroxyindole

Xiang-Qun Hu 1, David M Lovinger 1
PMCID: PMC2515613  NIHMSID: NIHMS55238  PMID: 18436267

Abstract

Allosteric modulation of ligand-gated ion channels can play important roles in shaping synaptic transmission. The function of the 5-hydroxytryptamine (serotonin) type 3 (5-HT3) receptor, a member of the Cys-loop ligand-gated ion channel superfamily, is modulated by a variety of compounds such as alcohols, anesthetics and 5-hydroxyindole (5-HI). In the present study, the molecular determinants of allosteric modulation by 5-HI were explored in N1E-115 neuroblastoma cells expressing the native 5-HT3 receptor and HEK 293 cells transfected with the recombinant 5-HT3A receptor using molecular biology and whole-cell patch-clamp techniques. 5-HI potentiated 5-HT-activated currents in both N1E-115 cells and HEK 293 cells, and significantly decreased current desensitization and deactivation. Substitution of Leu293 (L293, L15′) in the second transmembrane domain (TM2) with cysteine (L293C) or serine (L293S) abolished 5-HI modulation. Other mutations in the TM2 domain, such as D298A and T284F, failed to alter 5-HI modulation. The L293S mutation enhanced dopamine efficacy and converted 5-HI into a partial agonist at the mutant receptor. These data suggest that 5-HI stabilizes the 5-HT3A receptor in the open state by decreasing both desensitization and 5-HT unbinding/channel closing; and L293 is a common site for both channel gating and allosteric modulation by 5-HI. Our observations also indicate existence of a second 5-HI recognition site on the 5-HT3A receptor which may overlap with the 5-HT binding site and is not involved in the positive modulation by 5-HI. These findings support the idea that there are two discrete sites for 5-HI allosteric modulation and direct activation in the 5-HT3A receptor.

Keywords: 5-HT3A receptor, allosteric modulation, gating, desensitization, structure-function, electrophysiology

1. Introduction

Ligand-gated ion channels constitute a group of membrane-bound proteins that control the flux of ions across the cell membrane. The 5-hydroxytryptamine (serotonin) type 3 (5-HT3), nicotinic acetylcholine (nACh), γ-aminobutyric acid type A (GABAA), and glycine receptors belong to the Cys-loop ligand-gated ion channel superfamily, which mediate synaptic transmission and modulate neurotransmitter release (Ortells & Lunt, 1995;Reeves & Lummis, 2002). Members of the superfamily are pentamers formed by assembly of one (homomeric) or different (heteromeric) subunits. The subunits that form these receptors share similar topology, with each subunit possessing an extracellular N-terminal domain, four transmembrane domains (TMs) and loops connecting those TMs. The interfaces between subunits formed by the extracellular N-terminal domains comprise the agonist binding sites. The TM2 domains form the pore which is surrounded by an outer protein shell comprised of the other TMs.

The 5-HT3 receptor participates in a variety of physiological functions, such as cognitive processing, sensory transmission, regulation of autonomic function, integration of the vomiting reflex, pain processing and control of anxiety (Barnes & Sharp, 1999). Thus, agents that alter 5-HT3 receptor function, including allosteric modulators, are of interest because altering receptor function could be a useful therapeutic strategy for diseases associated with the receptor. The 5-HT3 receptor is not only the target of the natural agonist 5-HT but also a variety of allosteric modulators including the 5-HT analog 5-hydroxyindole (5-HI). Both 5-HT and 5-HI are metabolites of tryptophan (Mannaioni et al., 2003). 5-HI has been shown to increase release of neurotransmitter in the cerebellum and hippocampus (Zwart et al., 2002;Mannaioni et al., 2003) and triggers convulsion in rats (Mannaioni et al., 2003). 5-HI is a positive modulator of the 5-HT3 receptor (Kooyman et al., 1993;van Hooft et al., 1997;Gunthorpe & Lummis, 1999) and α7 nicotinic acetylcholine receptor (Zwart et al., 2002;Selina Mok & Kew, 2006). However, the molecular mechanisms by which 5-HI affects the 5-HT3 receptor are still poorly understood. Locating sites involved in the binding and allosteric actions of 5-HI has proven elusive due to the complex effects on 5-HT3A receptor-mediated responses and possible existence of more than one binding site. Previous studies suggested that 5-HI exhibits a combination of competitive and non-competitive actions at the 5-HT3 receptor (Kooyman et al., 1994), suggesting that there are at least two binding sites for 5-HI. One site likely overlaps with the orthosteric agonist binding site, whereas the second site may be in another region of the protein.

It has been established that amino acid residues in the transmembrane (TM) 2 domain and TM2-TM3 loop play critical roles in 5-HT3 receptor gating (Filatov & White, 1995;Grosman et al., 2000;Grosman & Auerbach, 2000) and allosteric modulation (Boileau & Czajkowski, 1999;Hu & Lovinger, 2005;Jones-Davis et al., 2005) of the Cys-loop ligand-gated ion channels. To identify amino acid residue(s) that participate in 5-HI modulation, we examined the roles of selected residues from TM2 domain of the 5-HT3A receptor by site-directed mutagenesis and whole-cell patch-clamp recording.

2. Materials and methods

2.1. Mutagenesis

Point mutation of the mouse 5-HT3A receptor (gift from Dr. D. Julius, San Francisco, CA) was accomplished using a QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). The successful incorporation of mutations was verified by sequencing the clones using an ABI Prism 377 automated DNA sequencer (Applied Biosystems, Foster City, CA). The cDNAs were then subcloned into the vector pCDNA3.1 (Invitrogen) for expression in human embryonic kidney (HEK) 293 cells (passage ≤ 12).

2.2. Cell Culture and Transient Receptor Expression

N1E-115 cells (American Type Culture Collection, Manassas, VA) were grown in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen) supplemented with 10% fetal bovine serum (FBS), and maintained in a humidified incubator at 37ºC in 5% CO2. HEK 293 cells (ATCC) were grown in minimum essential medium (MEM, Invitrogen) supplemented with 10% horse serum, and maintained in a humidified incubator at 37ºC in 5% CO2. HEK 293 cells were transiently transfected with the wild-type or mutant 5-HT3A receptor cDNA using the Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer’s instructions. Green fluorescent protein (pGreen Lantern, Invitrogen) was co-expressed with the 5-HT3A receptor subunit to permit selection of transfected cells under fluorescence optics. Each 35 mm dish was transfected with 3 μg of cDNA encoding the wild-type or mutant receptors along with 1 μg green fluorescent protein cDNA.

2.3. Whole-Cell Patch-Clamp Recording

Whole-cell patch-clamp recordings were performed in cultured cortical neurons 1 week after plating, N1E-115 cells 1 day after cell passage, or HEK 293 cells 1–3 day after transfection. HEK 293 cells were re-plated on the day of the experiment to ensure that recordings were only made from single, isolated cells. Cells were continuously superfused with a solution containing 140 mM NaCl, 5 mM KCl, 1.8 mM CaCl2, 1.2 mM MgCl2, 5 mM glucose, and 10 mM HEPES (pH was adjusted to 7.4 with NaOH and osmolarity adjusted to ~340 mosmol l−1 with sucrose). Pipettes were pulled from borosilicate glass (TW-150F, World Precision Instruments, Sarasota, FL) using a two-stage puller (Flaming-Brown P-97; Sutter Instruments, Novato, CA) and had resistances of 2–5 μM when filled with pipette solution containing 140 mM CsCl, 2 mM MgCl2, 10 mM EGTA, 10 mM HEPES (pH was adjusted to 7.2 with CsOH, and osmolarity adjusted to ~315 mosmol l−1 with sucrose). Membrane current was recorded in the whole-cell configuration using an Axopatch 200B amplifier (Axon Instruments, Foster City, CA) at 20–22ºC. Cells were held at −60 mV unless otherwise indicated. Data were acquired using pClamp9.0 software (Axon). Currents were filtered at 2 kHz and digitized at 5–10 kHz.

Drugs were applied with a piezoelectric device (PZ-150M; EXFO Burleigh Products Group Inc., Victor, NY) through two-barrel theta glass tubing (TGC150, Warner Instruments, Hamden, CT) that had been pulled to a tip diameter of ~200 μm. The cell was placed in front of the stream of control solution. The piezoelectric device was driven by TTL pulses from the pClamp9.0 software. Voltage applied to the piezoelectric device produced a rapid lateral displacement (~ 50 μm) of the theta tubing to move the interface between control and drug solutions. Solution exchange rates for open pipette and whole-cell recording were estimated using the potential change induced by switching from the control solution to a 140 mM N-methyl-D-glucamine (NMDG) test solution at 0 mV in the absence of agonist; and the current rising phase was fit using an exponential function. The solution exchange time constants were ~0.3 ms for an open pipette tip and ~1.6 ms for whole-cell recording.

2.4. Data Analysis

Data analysis and curve fitting were performed with Origin7.0 (Microcal Software, Northampton, MA), pClamp9.0 (Axon), or GraphPad InStat3.0 (GraphPad Software Inc., San Diego, CA) software. Concentration-response data for dopamine (DA) and 5-HI were fit using the Hill equation, I/Imax 5-HT = 1/[1 + (EC50/[Agonist])nH], where I is the current amplitude activated by a given concentration of agonist ([Agonist]), Imax 5-HT is the current amplitude produced by a maximally efficacious concentration of 5-HT (30 μM), nH is the Hill coefficient and EC50 is the concentration eliciting a half-maximal response.

Parameters of channel deactivation and desensitization were estimated by fitting appropriate current components using exponential functions of the general form ΣAne(−t/τn) + As, where An is the relative amplitude of the respective component, As is the steady-state current, n is the optimal number of exponential components, t is time and τn is the respective time constant. Curve fitting was achieved in Clampfit9.0 using the Levenberg-Marquardt algorithm. Additional components were accepted only if they significantly improved the fit, as determined by an F test performed using the analysis software.

Desensitization rates were derived from exponential fits to the current decay starting just after the current peak and extending to the end of agonist application, whereas deactivation rates were derived from exponential fits to the current decay after the removal of agonist following a 2-ms application. To facilitate direct comparison of desensitization or deactivation with different components, a weighted summation of time constants (Σanτn) was used, where an is the fractional contribution of the respective component, τn is the respective time constant, and n is the optimal number of exponential components. The extent of desensitization was measured as the percentage of current loss relative to peak current using the equation [(Ipeak-I10 s)/Ipeak] × 100, where Ipeak is the peak current amplitude, I10 s is the current amplitude at the end of 10 s application of 5-HT alone or 5-HT plus 5-HI.

In some experiments voltage ramps were applied to measure reversal potential. A ramp that changed the membrane potential between −80 and +60 mV with a slew rate of 0.5 mV/ms was applied during the peak of current activated by 3 μM 5-HT, 5 mM 5-HI, 5 mM 5-methoxyindole and 3-(2-hydroxyethyl)indole. Current activated by a voltage ramp in the absence of agonist was subtracted from the ramp-activated current in the presence of agonist prior to plotting and analyzing these data.

To facilitate comparison between different Cys-loop ligand-gated ion channels, a common TM2 domain numbering system is used (Miller, 1989). In this system, the amino acid residue at the putative cytoplasmic end of TM2 domain is assigned as position 1′ (Fig. 3A).

Data are presented as mean ± S.E.M. Statistical significance was determined with the Student’s t test or one-way analysis of variance (ANOVA). Differences were considered significant at p < 0.05.

2.5. Kinetic Simulations

Kinetic simulations of current measured in HEK 293 cells were generated using Berkeley-Madonna V8.3 software (Macey and Oster, Berkeley, CA, USA) and the Runge-Kutta 4 integration algorithm. Channel states and most rate constant estimates were as in the allosteric model developed by Solt et al. (2007) except where noted. For simulation of continuous 5-HT application, 10 sec of simulated current was generated in the presence of 30 μM 5-HT with all channels initially set in the unliganded, closed state (R). For simulation of “deactivation”, 20 sec of simulated current was generated with all channels initially set in the triple-liganded, open state (A3O) and agonist concentration set to 0. The figure showing simulated current was generated in Prism software using exported simulated traces.

3. Results

3.1. Positive modulation of 5-HT3 receptor-mediated current by 5-HI

To study the effects of 5-HI, we examined whole-cell currents activated by 5-HT in N1E-115 neuroblastoma cells expressing the native mouse 5-HT3 receptor (Neijt et al., 1988;Hope et al., 1993), and subsequently in HEK 293 cells transiently expressing the recombinant mouse 5-HT3A receptor. Inward current with a rapid onset was observed upon application of a maximally efficacious concentration of 5-HT (30 μM) for 10 s to either N1E-115 cells or transfected HEK 293 cells (Fig. 1A, left). Co-application of 5 mM 5-HI with 5-HT enhanced the current mediated by the 5-HT3A receptor in both cell lines (Fig. 1A, right). The average peak current amplitude activated by 5-HT in the presence of 5-HI was ~120% of that in the absence of 5-HI (N1E-115: 125 ± 6%; HEK293: 119 ± 8 %), which is comparable to previous findings in N1E-115 cells (Kooyman et al., 1993).

Figure 1. Effects of 5-HI on 5-HT3 receptor-mediated currents.

Figure 1

Open in a new tab

A, traces show current activated by 30 μM 5-HT from individual N1E-115 cells (upper) and HEK 293 cells transiently expressing the wild-type receptor (WT, lower) in the absence and presence of 5 mM 5-HI. Bar indicates the time of agonist application. Cells were first exposed to 30 μM 5-HT for 10 s followed by complete washout of the agonist, followed by exposure to the same concentration of agonist in the presence of 5-HI. B, traces show alteration of desensitization by 5 mM 5-HI in N1E-115 cells (upper) and HEK 293 cells transfected with the WT receptor (lower). Current amplitudes were normalized to peak current and superimposed for comparison. Bar indicates the time of agonist application. Traces in the absence of 5-HI are in black, whereas those in the presence of 5-HI are in grey. Data are summarized in Table 1.

The 5-HT3 receptor-mediated current decayed toward baseline in both cell lines during prolonged agonist exposure (Fig. 1B). Desensitization time course was best fit using a bi-exponential function in N1E-115 cells, whereas it could be fit adequately with a mono-exponential function in HEK 293 cells transfected with the wild-type receptor. Desensitization (measured as weighted desensitization time constant) in N1E-115 cells was considerably faster than in the HEK 293 cells (Table 1, p < 0.01). Co-application of 5-HI decreased the rate of current decay in the presence of agonist in both cell lines. This finding is consistent with the decrease in desensitization rate previously observed during 5-HI application (Kooyman et al., 1993;Gunthorpe & Lummis, 1999). 5-HI did not alter the number of components of desensitization in either cell line. However, the proportion of the fast component in N1E-115 cells was reduced by 5-HI from ~70% to ~20%. In addition, 5-HI significantly increased the time constants of both fast and slow components (Table 1). 5-HI also drastically increased desensitization time constant in HEK 293 cells expressing the wild-type receptor (Table 1). We also examined the impact of 5-HI co-application on the extent of desensitization of 5-HT3 receptor-mediated current by normalizing the loss of current amplitude at the end of 10 s application of 5-HT over the peak current amplitude (see Materials and Methods). Although the extent of desensitization (measured as percent current decay) in the absence of 5-HI was greater in N1E-115 cells than in HEK 293 cells (N1E-115: 95.8 ± 0.8%; HEK 293: 81.8 ± 1.6%, p < 0.01), 5-HI significantly decreased this parameter in both cell lines (N1E-115: 25.8 ± 0.8%, p < 0.01; HEK 293: 24.2 ± 1.1%, p < 0.01) to a similar extent (p = 0.3 %).

Table 1.

Effects of selected mutations in TM2 and TM2-TM3 loop on modulation of desensitization kinetics by 5-HI

5-HT alone
5-HT + 5-HI
n
τFAST (ms) % Fast τSLOW (ms) % Slow τWEIGHTED (ms) τFAST (ms) % Fast τSLOW (ms) % Slow τWEIGHTED (ms)
N1E-115 295 ± 36 73 ± 3 2899 ± 457 27 ± 3 718 ± 118 613 ± 71** 20 ± 3 11119 ± 819** 80 ± 3 9092 ± 817 6
WT -- -- 3700 ± 353 100 3700 ± 353 -- -- 11153 ± 867** 100 11153 ± 867** 9
L293C 382 ± 87 68 ± 5 2646 ± 77 32 ± 5 1103 ± 162 313 ± 79 70 ± 5 2498 ± 153 30 ± 5 1093 ± 177 6
L293S 508 ± 46 66 ± 2 3013 ± 300 34 ± 2 1375 ± 159 392 ± 35 65 ± 3 2716 ± 193 35 ± 3 1212 ± 110 9
T284F 269 ± 21 73 ± 2 2155 ± 287 27 ± 2 783 ± 80 _ _ 3297 ± 206 100 3297 ± 206** 15
D298A 200 ± 26 81 ± 2 1353 ± 133 19 ± 2 425 ± 33 602 ± 48** 49 ± 1** 3174 ± 187** 52 ± 1** 1931 ± 148** 5
Open in a new tab

Desensitization time constants and percent contributions were derived from exponential fits to the current decay. To facilitate comparisons, the desensitization data were also transformed to weighted desensitization (see Materials and Methods). Data shown represent mean ± S.E.M. from n cells.

**

p < 0.01.

3.2. Role of L293 in the modulatory action of 5-HI

The sequence alignment of the amino acid residues around TM2 across members of Cys-loop ligand-gated ion channels is displayed in Fig. 2A. Residue L293 is in the same 15′ position as a residue in the GABAA receptor that has important roles in channel gating (Findlay et al., 2001;Scheller & Forman, 2002) and modulation by alcohols and general anesthetics (Mihic et al., 1997;Wick et al., 1998). This amino acid residue is thought to reside on the non-pore-facing side of TM2 domain (Panicker et al., 2002;Reeves et al., 2001). We recently demonstrated that alcohol enhancement of 5-HT3A receptor-mediated current was abolished with L293C and L293S mutations (Hu et al., 2006). One effect of alcohols is to decrease desensitization, and this alcohol action was also lost in receptors with mutations at L293. The actions of 5-HI on the 5-HT3A receptor are similar to those observed for alcohols. It is possible that the allosteric interaction described for alcohol/anesthetics might involve mechanisms similar to those involved in modulation by 5-HI. We therefore examined the action of 5-HI in the L293C and L293S mutant receptors.

Figure 2. Effects of selected mutations in TM2 domain and the TM2-TM3 loop on 5-HI modulation.

Figure 2

Open in a new tab

A, sequence alignment of the TM2 domains of selected Cys-loop ligand-gated ion channels. Asterisks denote putative channel-lining amino acid residues. The open box indicates the residues in the GABAA and glycine receptors that are thought to be part of anesthetic binding site. B, traces show modulation of current amplitude by 5-HI in the L293C and L293S receptors. C, traces show alteration of desensitization by 5 mM 5-HI in HEK 293 cells expressing the L293C and L293S receptors. D, traces show modulation of current amplitude by 5-HI in the T284F and D298A receptors. E, traces show alteration of desensitization by 5 mM 5-HI in HEK 293 cells expressing the T284F and D298A receptors. Traces in the absence of 5-HI are in black, whereas those in the presence of 5-HI are in grey. The bar indicates the time of agonist application. Currents were normalized to peak current and superimposed for comparison in C and E. Data are summarized in Table 1.

Interestingly, currents activated by 30 μM 5-HT in the L293C and L293S receptor were slightly inhibited by co-application of 5 mM 5-HI (Fig. 2B). The average current amplitudes produced by co-application of 5-HT and 5-HI in the L293C and L293S receptors were 80.5 ± 5.6% and 73.0 ± 3.4%, respectively, of the current amplitude activated by 5-HT alone. In both the L293C and L293S receptors there was a rebound current upon termination of co-application of 5-HT and 5-HI. Desensitization of 5-HT-activated current was faster in the L293C and L293S receptors than in the wild-type receptor (Fig. 2C, Table 1, ANOVA, p < 0.01). Unlike the wild-type receptor, a bi-exponential function was required to adequately fit the current decay in both the L293C and L293S receptors. The newly introduced fast component accounted for ~65% of the total current decay for both mutant receptors. Neither the time constant nor proportion of the fast and slow components of desensitization was significantly altered by 5-HI in the L293C and L293S receptors. 5-hydroxyindole did not alter the weighted desensitization time constant in either L293C (p = 0.9) or L293S (p = 0.4) receptors (Table 1). In addition, the extent of desensitization remained unchanged in the presence of 5-HI in both L293C (91.7 ± 1.1% → 91.3 ± 0.9%, p = 0.4) and L293S (88.8 ± 1.4% → 89.4 ± 1.1, p = 0.7) receptors.

A recent study in the α7 nACh receptor suggested that 5-HI enhancement of ACh-activated current involves T266, the 6′ residue of the TM2 domain (Placzek et al., 2004). In addition, we have recently demonstrated that D298, an amino acid in the TM2-TM3 loop of the 5-HT3A receptor participating in forming the outer mouth of the channel pore (although this residue may be part of the extended TM2 α-helic structure), plays a critical role in gating and desensitization as well as Ca2+ modulation (Hu & Lovinger, 2005). Therefore we tested whether the T284 (6′ residue of the TM2 domain in the 5-HT3A receptor) and D298 residues are involved in 5-HI action with the T284F and D298A mutations. The average current amplitudes in response to co-application of 5-HT and 5-HI in the T284F and D298A receptors were 123.1 ± 1.5% and 126.9 ± 9.3% of the current amplitude in the presence of 5-HT alone, respectively (Fig. 3D). Desensitization of T284F or D298A receptor-mediated current was faster than in the wild-type receptor (Fig. 3E, Table 1, ANOVA, p < 0.01). A bi-exponential function was required to adequately fit desensitization in these two mutant receptors (Table 1). The fast component of desensitization introduced by the mutations accounted for 70–80% of the total current amplitude. Co-application of 5-HI with 5-HT abolished the fast component of desensitization in the T284F receptor and reduced the fast component from ~ 80% to ~50% of the total current amplitude in the D298A receptor. The time constants for both fast and slow components of desensitization were significantly increased in the D298A receptor by 5-HI (p < 0.01; Table 1). The weighted desensitization time constants for both T284F and D298A receptor were also increased by 5-HI (p < 0.01; Table 1). In addition, the extent of desensitization was significantly reduced in both T284F (99.0 ± 0.4% → 78.8 ± 1.2%, p < 0.01) and D298A (99.5 ± 0.3% → 91.5 ± 0.4%, p < 0.01) receptors, although the effects were not as large as those observed in the wild-type receptor (ANOVA, p < 0.0 1).

Figure 3. Effects of L293S mutation on modulation of 5-HT3A receptor desensitization across a range of 5-HI concentrations.

Figure 3

Open in a new tab

A, traces show alterations of desensitization by various concentrations of 5-HI in HEK 293 cells expressing the WT (upper) and L293S (lower) receptors. The bar indicates the time of agonist application. Currents were normalized to peak current and superimposed for comparison. Time constants were estimated by fitting with either a mono-exponential (WT) or a bi-exponential (L293S) function. B, averaged data show effects of 5-HI on desensitization of currents activated by 30μM 5-HT in the WT and L293S receptors. Each bar represents mean ± S.E.M. from 5–10 cells. ** p < 0.01. The traces obtained in the absence of 5-HI are in black, whereas those obtained in the presence of 5-HI are in grey.

We also examined the effect of 5-HI at lower concentrations on 5-HT3A receptor-mediated current in both the wild-type and L293S receptors (Fig. 3A). As shown in Fig. 3B, 5-HI at 30 μM decreased desensitization of 5-HT-activated current in the wild-type receptor (p < 0.05). The increase in desensitization time constant was 5-HI concentration-dependent in the wild-type receptor, reaching a maximum at ~300 μM 5-HI. In addition, 5-HI also produced a concentration-dependent reduction in the extent of desensitization in the wild-type receptor (30 μM: 70.7 ± 1.6%; 100 μM: 45.9 ± 3.0%; 300 μM: 39.4 ± 2.2%; and 1 mM: 29.3 ± 0.4%). However, the desensitization rate and relative contribution of the fast and slow components in the L293S receptor were not altered by either 100 μM or 1 mM 5-HI (Fig. 3B). The weighted desensitization time constant remained unchanged when either concentration of 5-HI was co-applied with 5-HT in the L293S receptor (100 μM: p = 0.8; 1 mM: p = 0.7). Furthermore, the extent of desensitization remained unchanged in the presence of 100 μM (86.6 ± 2.2%, p = 0.7) and 1 mM (88.2 ± 2.9%, p = 0.9) 5-HI.

5-HI was without effect on the peak current amplitude in the wild-type receptor at 30 μM (99.8 ± 1.3% of control). Concentration-dependent enhancement of 5-HT-activated current was seen with 5-HI concentrations ≥ 100 μM in the wild-type receptor (108.6 ± 1.7%, 116.4 ± 2.0% and 118.1 ± 2.9% of the control for 100 μM, 300 μM and 1 mM 5-HI, respectively). 5-HI at 100 μM caused no change in the peak current amplitude in the L293S receptor (97.3 ± 2.0% of control). However, 5-HI at 1 mM slightly inhibited 5-HT-activated current in the L293S receptor; and the current amplitude in the presence of 1 mM 5-HI was 90.1 ± 1.7% of control.

The impact of 5-HI on responses to a lower concentrations of 5-HT was also examined. Fig. 4 depicts the current activated by 1 μM and 3 μM 5-HT in the absence and presence of 5 mM 5-HI in the wild-type, L293C and L293S receptors. In the wild-type receptor 5-HT at 1 μM activated a slowly rising current that showed minimal decay during 10 s of agonist application (Fig. 4A, left). In the presence 5-HI, the current activated by 5-HT developed gradually and never attained a true steady-state value during agonist application (Fig. 4A, right). The initial phase of current activation appeared similar to that observed with 5-HT alone, but current in the presence of 5-HI continued to increase at time points where current plateaued with agonist alone (Fig. 4A, inset). The average current amplitude at the end of a 10 s agonist application in the presence of 5-HI was 329.6 ± 34.5% of current amplitude in response to agonist alone. However, in both the L293C and L293S receptors the initial rising phase of the current activated by 1 μM 5-HT was faster than that observed in the wild-type receptor; and the current amplitude was inhibited by 5-HI (Fig. 4B). The average peak current amplitude in the presence of 5-HI was 68.5 ± 3.4% and 71.0 ± 2.9% of control (agonist alone) for the L293C and L293S receptors, respectively. 5-HT at 3 μM activated a relative fast rising current in the wild-type receptor (Fig. 4C, left). The response activated by 5-HT was enhanced by co-application of 5-HI (134.7 ± 10.3% of the control; Fig. 4C, right). The current decayed after reaching peak in the absence of 5-HI and plateaued in the presence of 5-HI (Fig. 4C, inset). 5-HI inhibited 3 μM 5-HT-activated current in the L293S receptor (77.1 ± 5.6% of the control; Fig. 4D). A rebound current appeared at the end of co-application of 5-HT and 5-HI in the L293C and L293S receptors.

Figure 4. Effects of 5-HI on current activated by low concentrations of agonist.

Figure 4

Open in a new tab

A, traces show current activated by 1μM 5-HT in the WT receptor in the absence (left) and presence (right) of 5 mM 5-HI. Inset depicts the superimposed traces in the absence and presence of 5-HI on an expanded time scale to emphasize the initial current activation phase. B, traces show current activated by 1μM 5-HT in the L293C (left) and L293S (right) receptors in the absence and presence of 5 mM 5-HI. Similar responses were obtained from 6–8 cells. C, traces show current activated by 3μM 5-HT in the WT receptor in the absence (left) and presence (right) of 5 mM 5-HI. Inset depicts the superimposed traces in the absence and presence of 5-HI on an expanded time scale to emphasize the initial current activation phase. D, traces show current activated by 3μM 5-HT in the L293S receptor in the absence and presence of 5 mM 5-HI. Similar responses were obtained from 7–8 cells. The traces obtained in the absence of 5-HI are in black, whereas those obtained in the presence of 5-HI are in grey.

To determine whether 5-HI alters deactivation of 5-HT-activated current, a short pulse (2 ms) of 1 mM 5-HT was followed by washout with either control solution or 5 mM 5-HI-containing solution. The current activated by a brief application of 5-HT decayed gradually back to baseline levels in both the wild-type and L293S receptors (Fig. 5A). However, deactivation in the L293S receptor appeared to be faster than in the wild-type receptor. As shown in Fig. 5A, 5-HI slowed deactivation in the wild-type receptor but not in the L293S receptor. The current decay after removal of 5-HT in the absence or presence of 5-HI followed a mono-exponential time-course in the wild-type receptor, whereas it was best described by a bi-exponential function in the L293S receptor. 5-HI significantly increased the weighted deactivation time constant in the wild-type receptor (p < 0.01), whereas it did not alter this value in the L293S receptor (p = 0.5, Fig. 5B).

Figure 5. Effects of L293S mutation on 5-HI modulation of 5-HT3A receptor deactivation.

Figure 5

Open in a new tab

A, traces show alterations of deactivation by 5 mM 5-HI in HEK 293 cells expressing the WT (upper) and L293S (lower) receptors. Arrow indicates the brief (2 ms) application of 5-HT. Currents were normalized to peak current and superimposed for comparison of deactivation. Time constants were estimated by fitting with either a mono-exponential (WT) or a bi-exponential (L293S) function. The traces obtained in the absence of 5-HI are in black, whereas those obtained in the presence of 5-HI are in grey. B, Averaged data show effects of 5-HI on deactivation of currents activated by 30μM 5-HT in the WT and mutant receptors. Each bar represents mean ± S.E.M. from 6–8 cells. ** p < 0.01.

3.3. Conversion of 5-HI into an agonist by the L293S mutation

Dopamine is a partial agonist at the 5-HT3 receptor (van Hooft & Vijverberg, 1996), and application of a maximally efficacious concentration of dopamine (3 mM) activated a small current in the wild-type receptor expressed in HEK 293 cells. On the other hand, 5 mM 5-HI did not induce any detectable current when applied in the absence of 5-HT in the wild-type receptor (Fig. 6A, middle). In the L293S receptor 3 mM dopamine activated a current of larger amplitude relative to 30 μM 5-HT than that observed in the wild-type receptor. Surprisingly, 5 mM 5-HI showed agonist activity when applied in the absence of 5-HT to the L293S receptor, activating a relatively small amplitude current (Fig. 6A, right). The currents activated by various concentrations of 5-HI are presented in Fig. 6B. A rebound current occurred upon removal of high concentrations (≥ 5 mM) of 5-HI. Concentration-response curves for agonists were constructed by normalizing the amplitude of current activated by a range of dopamine and 5-HI concentrations to that produced by a maximally-effective 5-HT concentration (I/I30 μM 5-HT) (Fig. 6C). The amplitude of current activated by dopamine was concentration-dependent in both the wild-type and L293S receptors. The concentration-response curve for dopamine was shifted to the left and maximal dopamine efficacy was increased in the L293S receptor. The L293S receptor exhibited a 3.5-fold increase in dopamine potency (EC50) (L293S: 0.04 ± 0.01 mM; wild-type: 0.14 ± 0.05 mM; p < 0.01). The L293S mutation also decreased the Hill slope of the concentration-response curve for dopamine (wild-type: 2.10 ± 0.16; L293S: 1.17 ± 0.25; p < 0.05). The maximal response to dopamine was also greatly increased by the L293S mutation, from ~30 to ~65% of the current activated by 30 μM 5-HT (p < 0.01). In the wild-type receptor 5-HI alone failed to activate a response at concentrations up to 10 mM. However, a detectable current was observed at 300μM 5-HI in the L293S receptor (Figs. 6B and C). The amplitude of current activated by 5-HI was also concentration-dependent in the L293S receptor and the maximal response to 5-HI was ~25% of the current activated by 30 μM 5-HT (Fig. 6C). Fitting the pooled data with the logistic equation revealed an EC50 value of 1.05 ± 0.02 mM and a Hill slope of 2.31 ± 0.19 for 5-HI activation of the L293S receptor.

Figure 6. L293S mutation alters gating of the 5-HT3A receptor.

Figure 6

Open in a new tab

A, chemical structure of 5-HT, dopamine (DA) and 5-HI (left) and their agonist activity in the WT (middle) and L293S (right) receptors. Cells expressing the WT and L293S receptors were first exposed to 30 μM 5-HT. After complete washout of the agonist, the cells were challenged with either 3 mM DA or 5 mM 5-HI. The arrow indicates the application of 5-HT for 250 ms. Bars indicate the application times for DA and 5-HI. B, traces show current activated by 30 μM 5-HT and various concentrations of 5-HI in the L293S receptor. Cells expressing the L293S receptor were first exposed to 30 μM 5-HT. After complete washout of the agonist, the cells were challenged with 5-HI concentrations ranging from 0.1 to 10 mM. The trace activated by 5-HT is in grey, whereas those activated by 5-HI are in black. Inset depicts currents activated by 5-HI with expanded current and time scales. C, concentration-response curves for dopamine and 5-HI in the WT and mutant receptors. The responses elicited by dopamine or 5-HI were normalized to peak current activated by 30 μM 5-HT for each cell expressing either the wild-type or L293S receptor. Each data point represents mean ± S.E.M. from 5–9 cells.

3.4. Properties of 5-HI-activated currents

Both 8-(diethylamino)octyl-3,4,5-trimethoxybenzoate (TMB-8) and the 5-HT3 receptor competitive antagonist MDL 72222 have been used to detect spontaneous channel activity in the 5-HT3A mutant receptors (Zhang et al., 2002;Bhattacharya et al., 2004). The application of 100 μM TMB-8 or 300 nM MDL 72222 failed to alter the holding current in cells expressing the L293S receptor (Fig. 7A), suggesting that the L293S mutation did not produce spontaneous channel activity. The current activated by 5 mM 5-HI alone in the L293S receptor was completely abolished by 300 nM MDL 72222 (Fig. 7B). Fig. 7C shows the current-voltage (I-V) relationship obtained with a voltage-ramp from −80 to +60 mV at the peak of current activated by 3 μM 5-HT in the wild-type receptor and by 5 mM 5-HI in the L293S receptor. The shape of the I-V relationship was similar for both 5-HI and 5-HT; and the current reversed at 2.8 ± 0.9 mV for 5-HI in the L293S receptor and 3.2 ± 1.0 mV for 5-HT in the wild-type receptor (p = 0.8). We next tested whether the membrane potential influences the rebound current generated upon the termination of 5-HI application in the L293S receptor. The rebound current was preserved at holding potentials ranging from −80 to +40 mV (Fig. 7D).

Figure 7. Properties of 5-HI-activated current in the L293S receptor.

Figure 7

Open in a new tab

A, traces show the inability of TMB-8 (middle) and MDL 72222 (right) to alter the holding current in a cell expressing the L293S receptor. Cells expressing the L293S receptor were first exposed to 30 μM 5-HT. After complete washout of the agonist, the cells were challenged with 100 μM TMB-8 or 300 nM MDL 72222 for 10 s. Similar responses were observed in 5 cells. B, traces show current activated by 5 mM 5-HI in the absence (left) and presence of 300 nM MDL 72222 (right). The bar indicates the application time for 5-HI (black) and MDL 72222 (grey). Similar responses were observed in 6 cells. C, voltage ramps obtained at the peak of current activated by 3 μM 5-HT in the WT receptor and 5 mM 5-HI in the L293 receptor. Currents activated by the voltage protocol in the absence of 5-HT and 5-HI have been subtracted. Data are normalized to the current amplitude obtained at −80 mV. Similar results were obtained from 6–8 cells. D, trace shows the effects of holding potential on the rebound current observed upon removal of 5-HI. Similar responses were observed in 3 cells.

In the present study, 5-HI was able to function as an agonist in the L293S receptor. Analogs of 5-HI, such as tryptamine, are weak partial agonists at the wild-type 5-HT3 receptor (van Hooft & Vijverberg, 1996). These findings suggest a possible interaction of 5-HI analogs with the 5-HT recognition site on the 5-HT3A receptor. Therefore we wished to find out whether other closely-related analogs of 5-HI (Fig. 8A, left) could activate the wild-type 5-HT3A receptor. We examined the agonist activities of 5-methoxyindole (5-MI) and 3-(2-hydroxyethyl)indole (3-2-HEI) in HEK 293 cells transfected with the wild-type receptor. Both 5-MI and 3-2-HEI were unable to elicit a current in untransfected HEK 293 cells (data not shown). However, we observed that application of either compound alone at 5 mM directly activated inward current in HEK 293 cells transfected with the wild-type receptor (Fig. 8A, right). The onset of current was relatively slow in response to an application of either compound when compared to 5-HT-activated current. A rebound current upon agonist removal was observed following cessation of application of 5-MI, but not 3-2-HEI.

Figure 8. Agonist activity of 5-HI analogs in the WT receptor.

Figure 8

Open in a new tab

A, chemical structure of 5-HT, 5-methoxyindole (5-MI) and 3-(2-hydroxyethyl)indole (3-2-HEI) (left) and traces showing the agonist activity of 5-MI and 3-2-HEI (right). Cells expressing the WT receptor were first exposed to 30 μM 5-HT. After complete washout of the agonist, the cells were challenged with either 5 mM 5-MI or 3-2-HEI. B, averaged data show the efficacy of 5-HI analogs in the WT receptor. The relative efficacy of the partial agonists was determined by normalizing the current amplitude activated by partial agonists as a percentage of that activated by 30 μM 5-HT. Each bar represents mean ± S.E.M. from 7–10 cells. C, voltage ramps obtained at the peak of current activated by 5 mM 5-MI or 3-2-HEI in the WT receptor. Currents activated by the voltage protocol in the absence of agonists have been subtracted. Data are normalized to the current amplitude obtained at −80 mV. Similar results were obtained from 5–9 cells.

Averaged data indicate that 5-MI and 3-2-HEI were weak agonists at the wild-type receptor; the maximal responses to 5 mM 5-MI and 3-2-HEI were ~30% and ~20% of the current activated by 30 μM 5-HT, respectively (Fig. 8B). The I-V relationship obtained with a voltage-ramp from −80 to +60 mV at the peak of current activated by the agonists, as shown in Fig. 8C, revealed similar rectification and reversal potential for current activated by 5-MI (3.0 ± 0.5 mV) and 3-2-HEI (2.8 ± 0.4 mV), and these current characteristics were comparable to that for 5-HT in the wild-type receptor.

A sequential application protocol was used to examine the effect of pre-incubation of 5-HI on 5-HT-activated current in the wild-type and L293 receptors (Fig. 9). Responses to 30 μM 5-HT applications (250 ms and 10 s) in the absence of 5-HI immediately after pre-incubation with 5 mM 5-HI for 20 s were slightly reduced in the wild-type receptor (Fig. 9A and C, upper). In addition, 5-HI pre-incubation also reduced desensitization (Fig. 9C). As seen before, pre-incubation with 5 mM 5-HI activated an inward current in the L293S receptor which gradually decayed in the continued presence of 5-HI. Responses to 250-ms or 10-s applications of 5-HT immediate following 5-HI pre-incubation were almost abolished (Fig. 9A & C, lower). The average responses after 5-HI pre-incubation were ~90% and 5% for the wild-type and L293S receptors, respectively (Fig. 9B and D).

Figure 9. Effects of pre-incubation with 5-HI on 5-HT-activated currents in the WT and L293S receptors.

Figure 9

Open in a new tab

A and C, cells expressing the WT or L293S receptors were first exposed to 30 μM 5-HT for 250 ms. After complete washout of the agonist, the cells were exposed to 5 mM 5-HI for 20 s before challenging with 30 μM 5-HT alone for 250 ms (A) or 10 s (C). B and D, averaged data show the effect of 5-HI pre-incubation on 5-HT-activated responses. The current amplitude activated by 5-HT immediately followed by 5-HI pre-incubation was normalized as a percentage of that activated by 30 μM 5-HT. Each bar represents mean ± S.E.M. from 7–10 cells.

4. Discussion

Previous studies indicated that 5-HI has mixed effects on 5-HT3 receptor-mediated currents, producing a competitive inhibition and an allosteric potentiation of receptor function; and imply existence of two distinct sites of 5-HI action (Kooyman et al., 1994). In the present study, we investigated the mechanism underlying the actions of 5-HI on the 5-HT3A receptor. Our findings provide strong support for this two-site model. Mutations at the L293 residue in the TM2 domain eliminate the positive allosteric modulation of the receptor by 5-HI, and reveal a partial agonism by this compound. The latter finding provides the most direct evidence to date of an interaction of 5-HI with the agonist binding site, which persists in the absence of the positive allosteric modulatory effect.

5-HI shifts the 5-HT concentration-response relationship leftward (van Hooft et al., 1997). The apparent increase in agonist potency produced by 5-HI might result from increased agonist affinity or enhanced channel gating or both (Colquhoun, 1998). The observation that 5-HI slowed deactivation in the wild-type receptor indicates that the rate of channel closure and 5-HT unbinding after agonist removal are decreased by 5HI. Therefore an enhanced agonist affinity might contribute to 5-HI potentiation in the wild-type receptor. However, the observation of 5-HI potentiation of the response activated by a maximally efficacious agonist concentration indicates that a process other than increased agonist affinity is also involved since an increase in affinity alone should not enhance maximal agonist efficacy (Christopoulos & Kenakin, 2002). An increase in channel conductance could also produce an increase in peak current amplitude. However, single-channel measurements indicate that 5-HI does not alter 5-HT3 receptor conductance (van Hooft et al., 1997). The observations that 5-HI decreases the rate and extent of desensitization provide evidence supporting the idea that this agent stabilizes the open state of the receptor, which in turn likely contributes to the increase in agonist potency by slowing the rate of agonist unbinding (Chang & Weiss, 1999).

We also observed that 5-HI acts as a partial agonist in the L293S construct that lacks 5-HI-induced potentiation. Examination of this mutant receptor also revealed that the direct activation by 5-HI occurs at drug concentrations higher than those that elicit potentiation in the wild-type receptor. This observation is generally consistent with the previous finding that the 5-HI potentiating effect is non-competitive with respect to the agonist binding site, and that the potency of 5-HI for potentiating receptor function is greater than its potency for altering agonist binding (Kooyman et al., 1994). The observation that 5-HI effects in the presence of a low concentration of agonist are similar to those observed at high agonist concentrations in both the wild-type and L293 mutant receptors supports the idea that the potentiating effect is allosteric.

It is possible that slowing of desensitization by 5-HI could account for most or all of the potentiation produced by the drug. A recent study by Solt et al. (2007) indicates that desensitization produced by 5-HT dominates the channel closure process even following agonist removal (i.e. our “deactivation” paradigm). We performed kinetic simulations of receptor function designed to look at current decay in the continuous presence of 30 μM 5-HT and after agonist removal (Figure 10) to determine if our findings with 5-HI could be completely explained by reduced desensitization rate. We used the allosteric model developed by Solt et al. (2007) and the kinetic parameters shown in Table 2. Indeed, we could produce reasonable simulations of the 5-HI-induced prolongation of current decay under either continuous or transient agonist application conditions by decreasing the forward desensitization rate constant (Kd+). Changes in unbinding steps (K2 and K3) had no additional effect on simulated current, indicating that our experimental results could be due to changes in desensitization alone. Similar results were obtained with a linear kinetic model that we have previously used to describe 5-HT3 function (Zhou et. al., 1998).

Figure 10. Slowing of desensitization alone recapitulates effects of 5HI in simulations of current.

Figure 10

Open in a new tab

A Allosteric kinetic model (from Solt et al. 2007) used to simulate 5-HT activation of the receptor and potential 5-HI actions. B Simulated current produced using the allosteric model assuming continuous application of 30 μM 5-HT for 10 sec with kd+ set at an estimated normal level for wild-type 5-HT3A receptor (black trace) and reduced by ~1 order or magnitude to simulate the effect of 5-HI (Table 2). Note the similarity in current decay profile to the data presented in Figure 1B lower traces. C Simulated current produced using the allosteric model in a “deactivation” protocol (20 sec duration) in which agonist is set at 0 and all channels are assumed to be in the A3O state at the start of the trace. Black trace shows current with estimated WT kd+, grey trace with reduced kd+ as in A and Table 2. Note the similarity in time course to actual current traces in Figure 5A (top trace).

Table 2.

Kinetic parameters for current simulations

5-HT alone 5-HT + 5-HI
5-HT* 30 μM 30 μM
k1 1 × 107 M−1s−1 1 × 107 M−1s−1
k2 10 s−1 10 s−1
k3 0.01 s−1 0.01 s−1
β 400 s−1 400 s−1
α 1 s−1 1 s−1
kd+ 0.5 s−1 0.04 s−1
kd− 0.01 s−1 0.01 s−1
kDR 0.7 s−1 0.7 s−1
L0 3 × 106 3 × 106
Open in a new tab

5-HT concentration and rate constants used to generate simulated 5-HT3A receptor-mediated currents (Figure 10). Note that 5-HT concentration applies only to current simulation of response to continuous agonist application (Figure 10B), while current in simulated “deactivation” protocol run in agonist-free condition (Figure 10C).

The blockade of the partial agonism of 5-HI at the L293S receptor by a competitive 5-HT3 receptor antagonist suggests that 5-HI is activating the channel via actions at a site that largely overlaps with the orthosteric agonist/competitive antagonist binding site on the receptor. It is perhaps not surprising that 5-HI can bind to the 5-HT site on the receptor given the structural similarity of the two compounds (Fig. 7A, left). A number of other 5-HT analogs act as 5-HT3 partial agonists, including tryptamine (van Hooft & Vijverberg, 1996), and the previously untested indole derivatives 5-methoxyindole and 3-(2-hydroxyethyl)indole (Fig. 9). These findings indicate that the orthosteric site of the 5-HT3A receptor may recognize a variety of substituted indole compounds. The previous findings (Kooyman et al., 1994) together with the observations made at present suggest that 5-HI has a low affinity at the orthosteric site.

The binding of agonist to the orthosteric site on a Cys-loop ligand-gated ion channel induces a conformational change in the extracellular N-terminal domain, which is relayed to the TM2-TM3 loop (Lummis et al., 2005). One model of channel activation is that the movement of the TM2-TM3 loop then causes the TM2 domain to rotate leading to channel opening (Miyazawa et al., 2003). Structurally, L293 is in a position to promote communication between the TM2-TM3 loop and the TM2 domain. Certain L293 mutations result in greater agonist potency and efficacy (Hu et al., 2006). The increased dopamine efficacy in the L293S mutation suggests that L293 is critical for the coupling of agonist binding to channel gating in the 5-HT3A receptor. The equivalent position (S270) in the GABAA receptor has also been found to be an important gating site (Scheller & Forman, 2002). It is likely that the 15′ amino acid residue in the TM2 domain of the Cys-loop ligand-gated ion channels constitutes a constraint for receptor activation. Binding of agonist at the orthosteric site releases this intrinsic structural constraint in the receptor protein. 5-HI has no intrinsic agonist activity in the wild-type receptor, suggesting that binding of 5-HI is not sufficient to overcome energy barriers that prevent the channel from opening. The L293S mutation could release the molecular constraints and lower these energy barriers such that 5-HI binding to the mutant receptor is able to produce a sufficient conformational change to open the channel. A similar finding was made with respect to the R222A mutation which converts the antagonist apomorphine into a potent and efficacious agonist by facilitating the conversion between different states of the 5-HT3A receptor (Hu et al., 2003).

Residues in the TM2 domain of Cys-loop ligand-gated channels participate in forming the channel gate, controlling ionic selectivity, and binding agents of physiological and pharmacological importance (Karlin, 2002;Lummis, 2004). Position 15′ of the 5-HT3A receptors TM2 domain is a non-pore-facing amino acid residue (Reeves et al., 2001;Panicker et al., 2002). In addition to its role in channel gating, the amino acid residue at 15′ also appears to have important roles in channel modulation by alcohol and anesthetics (Mihic et al., 1997;Wick et al., 1998), and binding of these drugs to a hydrophobic pocket involving this residue alters receptor function. In a separate study, we have demonstrated that L293 is critical for alcohol actions on the 5-HT3A receptor; and L293C and L293S mutations abolish alcohol modulation (Hu et al., 2006). Alcohols and 5-HI have similar actions on the 5-HT3A receptor, and mutations at L293 prevent the effects of these agents. Thus, L293 may be part of a common structural element necessary for the modulation by alcohol/anesthetics and 5-HI, although these compounds differ substantially in chemical structure.

The L293 residue could be part of a binding site for 5-HI involved in allosteric potentiation of the receptor. However, testing this hypothesis will have to await the development of technologies to more directly assess drug interactions with this residue. Furthermore, our previous studies suggest that there is little correlation between the physicochemical properties of the amino acid residues substituted at L293 (15′) and alcohol action (Hu et al., 2006). It is believed that allosteric agents act by modulating receptor activity through conformational changes in the receptor transmitted from the allosteric to the orthosteric site and/or to the coupling sites (Christopoulos & Kenakin, 2002). It is possible that L293 is a key residue in regions linking the allosteric site of 5-HI binding to agonist binding and/or channel gating. In this regard, it is interesting to note that amino acid residues in the TM2 domain and TM2-TM3 loop of the GABAA receptor are the structural determinants for coupling benzodiazepine binding to increased channel gating (Boileau & Czajkowski, 1999;Jones-Davis et al., 2005).

It is possible that the impairment of 5-HI modulation in the L293 mutant receptors is an indirect consequence of the conversion of 5-HI into a partial agonist in the L293S receptor. The decreased 5-HT occupancy at the receptor in the presence of 5-HI in this construct might preclude potentiation. However, this does not appear to be the case. A relatively low concentration of 5-HI (100 μM) that did not directly activate the L293S receptor produced no potentiation of the function of this construct, despite the fact that this concentration of 5-HI decreases desensitization in the wild-type receptor.

Allosteric modulation of ligand-gated ion channels is a well-known phenomenon. Positive allosteric modulation provides an attractive mechanism for the facilitation of physiologically appropriate receptor activation. Similarly, greater selectivity, potency and/or safety can often be achieved using allosteric modulators, as opposed to orthosteric ligands. Benzodiazepines are widely used central nervous system depressants which allosterically enhance the activity of GABAA receptor (Sigel & Buhr, 1997). Therefore, the allosteric modulation of the 5-HT3A receptor may hold promise for the development of novel therapeutics in disorders involving this receptor.

In summary, using N1E-115 cells expressing the native 5-HT3A receptor and HEK 293 cells transfected with the recombinant 5-HT3A receptor we have confirmed the positive allosteric action of 5-HI. This positive modulation is accompanied by a decrease in rates of channel desensitization and deactivation. Site-directed mutagenesis reveals that L293 constitutes an important gating site and is crucial for 5-HI positive allosteric modulation. L293 may constitute a common transduction pathway for both agonists and allosteric modulators. In addition, our data support the existence of a second 5-HI binding site at the 5-HT3A receptor that is not involved in the positive modulation, but likely overlaps with the orthosteric site.

Acknowledgments

This work was supported by the NIAAA Division of Intramural Clinical and Basic Research.

Abbreviations

5-HT3

5-hydroxytryptamine (serotonin) type 3

nACh

nicotinic acetylcholine

TM

transmembrane domain

5-HI

5-hydroxyindole

HEK cells

human embryonic kidney cells

WT

wild-type

DA

dopamine

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Barnes NM, Sharp T. A review of central 5-HT receptors and their function. Neuropharmacology. 1999;38:1083–1152. doi: 10.1016/s0028-3908(99)00010-6. [DOI] [PubMed] [Google Scholar]
  2. Bhattacharya A, Dang H, Zhu QM, Schnegelsberg B, Rozengurt N, Cain G, Prantil R, Vorp DA, Guy N, Julius D, Ford AP, Lester HA, Cockayne DA. Uropathic observations in mice expressing a constitutively active point mutation in the 5-HT3A receptor subunit. Journal of Neuroscience. 2004;24:5537–5548. doi: 10.1523/JNEUROSCI.5658-03.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Boileau AJ, Czajkowski C. Identification of transduction elements for benzodiazepine modulation of the GABAA receptor: three residues are required for allosteric coupling. Journal of Neuroscience. 1999;19:10213–10220. doi: 10.1523/JNEUROSCI.19-23-10213.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chang Y, Weiss DS. Channel opening locks agonist onto the GABAC receptor. Nature Neuroscience. 1999;2:219–225. doi: 10.1038/6313. [DOI] [PubMed] [Google Scholar]
  5. Christopoulos A, Kenakin T. G protein-coupled receptor allosterism and complexing. Pharmacological Reviews. 2002;54:323–374. doi: 10.1124/pr.54.2.323. [DOI] [PubMed] [Google Scholar]
  6. Colquhoun D. Binding, gating, affinity and efficacy: the interpretation of structure-activity relationships for agonists and of the effects of mutating receptors. British Journal of Pharmacology. 1998;125:924–947. doi: 10.1038/sj.bjp.0702164. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Filatov GN, White MM. The role of conserved leucines in the M2 domain of the acetylcholine receptor in channel gating. Molecular Pharmacology. 1995;48:379–384. [PubMed] [Google Scholar]
  8. Findlay GS, Ueno S, Harrison NL, Harris RA. Allosteric modulation in spontaneously active mutant γ-aminobutyric acidA receptors. Neuroscience Letters. 2001;305:77–80. doi: 10.1016/s0304-3940(01)01646-9. [DOI] [PubMed] [Google Scholar]
  9. Grosman C, Auerbach A. Asymmetric and independent contribution of the second transmembrane segment 12′ residues to diliganded gating of acetylcholine receptor channels: a single-channel study with choline as the agonist. Journal of General Physiology. 2000;115:637–651. doi: 10.1085/jgp.115.5.637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Grosman C, Salamone FN, Sine SM, Auerbach A. The extracellular linker of muscle acetylcholine receptor channels is a gating control element. Journal of General Physiology. 2000;116:327–340. doi: 10.1085/jgp.116.3.327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gunthorpe MJ, Lummis SC. Diltiazem causes open channel block of recombinant 5-HT3 receptors. Journal of Physiology. 1999;519:713–722. doi: 10.1111/j.1469-7793.1999.0713n.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hope AG, Downie DL, Sutherland L, Lambert JJ, Peters JA, Burchell B. Cloning and functional expression of an apparent splice variant of the murine 5-HT3 receptor A subunit. European Journal of Pharmacology. 1993;245:187–192. doi: 10.1016/0922-4106(93)90128-v. [DOI] [PubMed] [Google Scholar]
  13. Hu XQ, Hayrapetyan V, Gadhiya JJ, Rhubottom HE, Lovinger DM, Machu TK. Mutations of L293 in transmembrane two of the mouse 5-hydroxytryptamine3A receptor alter gating and alcohol modulatory actions. British Journal of Pharmacology. 2006;148:88–101. doi: 10.1038/sj.bjp.0706685. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Hu XQ, Lovinger DM. Role of aspartate 298 in mouse 5-HT3A receptor gating and modulation by extracellular Ca2+ Journal of Physiology. 2005;568:381–396. doi: 10.1113/jphysiol.2005.092866. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hu XQ, Zhang L, Stewart RR, Weight FF. Arginine 222 in the pre-transmembrane domain 1 of 5-HT3A receptors links agonist binding to channel gating. Journal of Biological Chemistry. 2003;278:46583–46589. doi: 10.1074/jbc.M308974200. [DOI] [PubMed] [Google Scholar]
  16. Jones-Davis DM, Song L, Gallagher MJ, Macdonald RL. Structural determinants of benzodiazepine allosteric regulation of GABAA receptor currents. Journal of Neuroscience. 2005;25:8056–8065. doi: 10.1523/JNEUROSCI.0348-05.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Karlin A. Emerging structure of the nicotinic acetylcholine receptors. Nature reviews Neuroscience. 2002;3:102–114. doi: 10.1038/nrn731. [DOI] [PubMed] [Google Scholar]
  18. Kooyman AR, van Hooft JA, Vanderheijden PM, Vijverberg HP. Competitive and non-competitive effects of 5-hydroxyindole on 5-HT3 receptors in N1E-115 neuroblastoma cells. British Journal of Pharmacology. 1994;112:541–546. doi: 10.1111/j.1476-5381.1994.tb13107.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kooyman AR, van Hooft JA, Vijverberg HP. 5-Hydroxyindole slows desensitization of the 5-HT3 receptor-mediated ion current in N1E-115 neuroblastoma cells. British Journal of Pharmacology. 1993;108:287–289. doi: 10.1111/j.1476-5381.1993.tb12795.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lummis SC. The transmembrane domain of the 5-HT3 receptor: its role in selectivity and gating. Biochemical Society Transactions. 2004;32:535–539. doi: 10.1042/BST0320535. [DOI] [PubMed] [Google Scholar]
  21. Lummis SC, Beene DL, Lee LW, Lester HA, Broadhurst RW, Dougherty DA. Cis-trans isomerization at a proline opens the pore of a neurotransmitter-gated ion channel. Nature. 2005;438:248–252. doi: 10.1038/nature04130. [DOI] [PubMed] [Google Scholar]
  22. Mannaioni G, Carpenedo R, Moroni F. 5-hydroxyindole causes convulsions and increases transmitter release in the CA1 region of the rat hippocampus. British Journal of Pharmacology. 2003;138:245–253. doi: 10.1038/sj.bjp.0705007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mihic SJ, Ye Q, Wick MJ, Koltchine VV, Krasowski MD, Finn SE, Mascia MP, Valenzuela CF, Hanson KK, Greenblatt EP, Harris RA, Harrison NL. Sites of alcohol and volatile anaesthetic action on GABAA and glycine receptors. Nature. 1997;389:385–389. doi: 10.1038/38738. [DOI] [PubMed] [Google Scholar]
  24. Miller C. Genetic manipulation of ion channels: a new approach to structure and mechanism. Neuron. 1989;2:1195–1205. doi: 10.1016/0896-6273(89)90304-8. [DOI] [PubMed] [Google Scholar]
  25. Miyazawa A, Fujiyoshi Y, Unwin N. Structure and gating mechanism of the acetylcholine receptor pore. Nature. 2003;423:949–955. doi: 10.1038/nature01748. [DOI] [PubMed] [Google Scholar]
  26. Neijt HC, te Duits I, Vijverberg HP. Pharmacological characterization of serotonin 5-HT3 receptor-mediated electrical response in cultured mouse neuroblastoma cells. Neuropharmacology. 1988;27:301–307. doi: 10.1016/0028-3908(88)90048-2. [DOI] [PubMed] [Google Scholar]
  27. Ortells MO, Lunt GG. Evolutionary history of the ligand-gated ion-channel superfamily of receptors. Trends in Neurosciences. 1995;18:121–127. doi: 10.1016/0166-2236(95)93887-4. [DOI] [PubMed] [Google Scholar]
  28. Panicker S, Cruz H, Arrabit C, Slesinger PA. Evidence for a centrally located gate in the pore of a serotonin-gated ion channel. Journal of Neuroscience. 2002;22:1629–1639. doi: 10.1523/JNEUROSCI.22-05-01629.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Placzek AN, Grassi F, Papke T, Meyer EM, Papke RL. A single point mutation confers properties of the muscle-type nicotinic acetylcholine receptor to homomeric α7 receptors. Molecular Pharmacology. 2004;66:169–177. doi: 10.1124/mol.66.1.169. [DOI] [PubMed] [Google Scholar]
  30. Reeves DC, Goren EN, Akabas MH, Lummis SC. Structural and electrostatic properties of the 5-HT3 receptor pore revealed by substituted cysteine accessibility mutagenesis. Journal of Biological Chemistry. 2001;276:42035–42042. doi: 10.1074/jbc.M106066200. [DOI] [PubMed] [Google Scholar]
  31. Reeves DC, Lummis SC. The molecular basis of the structure and function of the 5-HT3 receptor: a model ligand-gated ion channel. Molecular membrane biology. 2002;19:11–26. doi: 10.1080/09687680110110048. [DOI] [PubMed] [Google Scholar]
  32. Scheller M, Forman SA. Coupled and uncoupled gating and desensitization effects by pore domain mutations in GABAA receptors. Journal of Neuroscience. 2002;22:8411–8421. doi: 10.1523/JNEUROSCI.22-19-08411.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Selina Mok MH, Kew JN. Excitation of rat hippocampal interneurons via modulation of endogenous agonist activity at the α7 nicotinic ACh receptor. Journal of Physiology. 2006;574:699–710. doi: 10.1113/jphysiol.2006.104794. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Sigel E, Buhr A. The benzodiazepine binding site of GABAA receptors. Trends in pharmacological sciences. 1997;18:425–429. doi: 10.1016/s0165-6147(97)01118-8. [DOI] [PubMed] [Google Scholar]
  35. Solt K, Ruesch D, Forman SA, Davies PA, Raines DE. Differential effects of serotonin and dopamine on human 5-HT3A receptor kinetics: interpretation within an allosteric kinetic model. Journal of Neuroscience. 2007;27(48):13151–13160. doi: 10.1523/JNEUROSCI.3772-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. van Hooft JA, van der Haar E, Vijverberg HP. Allosteric potentiation of the 5-HT3 receptor-mediated ion current in N1E-115 neuroblastoma cells by 5-hydroxyindole and analogues. Neuropharmacology. 1997;36:649–653. doi: 10.1016/s0028-3908(97)00045-2. [DOI] [PubMed] [Google Scholar]
  37. van Hooft JA, Vijverberg HP. Selection of distinct conformational states of the 5-HT3 receptor by full and partial agonists. British Journal of Pharmacology. 1996;117:839–846. doi: 10.1111/j.1476-5381.1996.tb15269.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wick MJ, Mihic SJ, Ueno S, Mascia MP, Trudell JR, Brozowski SJ, Ye Q, Harrison NL, Harris RA. Mutations of gamma-aminobutyric acid and glycine receptors change alcohol cutoff: evidence for an alcohol receptor? Proceedings of the National Academy of Sciences of the United States of America. 1998;95:6504–6509. doi: 10.1073/pnas.95.11.6504. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zhang L, Hosoi M, Fukuzawa M, Sun H, Rawlings RR, Weight FF. Distinct molecular basis for differential sensitivity of the serotonin type 3A receptor to ethanol in the absence and presence of agonist. Journal of Biological Chemistry. 2002;277:46256–46264. doi: 10.1074/jbc.M207683200. [DOI] [PubMed] [Google Scholar]
  40. Zhou Q, Verdoorn TA, Lovinger DM. Alcohols potentiate the function of 5-HT3 receptor-channels on NCB-20 neuroblastoma cells by favouring and stabilizing the open channel state. Journal of Physiology (London) 1998;507:335–352. doi: 10.1111/j.1469-7793.1998.335bt.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zwart R, De Filippi G, Broad LM, McPhie GI, Pearson KH, Baldwinson T, Sher E. 5-Hydroxyindole potentiates human α7 nicotinic receptor-mediated responses and enhances acetylcholine-induced glutamate release in cerebellar slices. Neuropharmacology. 2002;43:374–384. doi: 10.1016/s0028-3908(02)00094-1. [DOI] [PubMed] [Google Scholar]

RESOURCES