Abstract
Mutant 5-hydroxytryptamine (5-HT)3A receptors, in which changes were made at Ile294, position 16′, of the second transmembrane domain, were assessed for alterations in macroscopic response kinetics and modulation by alcohols. Function of heterologously expressed receptors was measured in Xenopus oocytes in the two-electrode voltage clamp configuration and in human embryonic kidney 293 cells using whole cell patch-clamp electrophysiological recordings with rapid drug application. Compared with the wild-type receptor, a decrease in the 5-HT EC50 value in the Ile294Thr mutant was observed, whereas an increase in the 5-HT EC50 value in the Ile294Leu mutant was measured. Ile294Thr receptors showed a marked reduction in the extent of desensitization. Ethanol and 2,2,2-trichloroethanol (TCEt) enhanced 5-HT-mediated currents in wild-type and Ile294Leu receptors, but inhibited or had little stimulatory effect in the Ile294Thr mutant. Kinetic analysis revealed that in the presence of TCEt, the slope of activation was unchanged in the Ile294Thr mutant and increased in the wild-type receptor. Alcohol cutoff was altered with wild-type = heptanol and Ile294Leu = hexanol. Kinetic changes in the Ile294Thr mutant that favor the open channel state, as well as reduction in the rate of channel activation in the presence of TCEt, likely underlie this mutant's altered response to n-chain alcohols.
The 5-hydroxytryptamine3 (5-HT3) receptor is a member of the superfamily of ligand-gated ion channels, of which the nicotinic acetylcholine (nACh) receptor is the prototype (Derkach et al., 1989; Peters et al., 1994). To date, two subunits of the 5-HT3 receptor, 5-HT3A (Maricq et al., 1991; Belelli et al., 1995; Miyake et al., 1995; Lankiewicz et al., 1998) and 5-HT3B (Davies et al., 1999; Dubin et al., 1999) have been cloned. Recent reverse transcriptase and immunohistochemical studies demonstrated that the B subunit is largely expressed in the peripheral nervous system, suggesting that the predominant form in the brain is the A homomer (Morales and Wang, 2002). The most well documented actions of 5-HT3 receptors are to alter gastrointestinal motility and to regulate the vomiting reflex (Aapro, 1991). Moreover, the 5-HT3 receptor has been implicated in altering the voluntary intake of ethanol in humans (Johnson et al., 1993) and rodents (Knapp and Pohorecky, 1992; Hodge et al., 1993).
The 5-HT3 receptor is modulated by pharmacologically relevant concentrations of n-chain alcohols and anesthetics. Alcohols and volatile anesthetics, in the presence of low 5-HT concentrations, potentiate native and heterologously expressed 5-HT3 receptors (Lovinger and White, 1991; Machu and Harris, 1994). The mechanism for enhancement of receptor function by alcohols has been assessed in NCB-20 cells, which express the 5-HT3 receptor endogenously. Ethanol and 2,2,2-trichloroethanol (TCEt) enhanced peak currents evoked by a maximally effective concentration of dopamine, which is a weak partial agonist at the 5-HT3 receptor (Lovinger et al., 2000). That alcohols would enhance receptor function under conditions of full agonist occupancy suggests that they increase probability of channel opening independent of any effect on agonist affinity. Whole cell patch-clamp electrophysiological recordings conducted with rapid drug superfusion suggest that alcohols increase the activation rate, decrease the intrinsic desensitization rate, and decrease the deactivation rate of the 5-HT3 receptor channel. These effects act in concert to favor and stabilize the open channel state (Zhou et al., 1998).
The binding domain in the 5-HT3 receptor with which alcohols and/or anesthetics interact remains unknown. On the basis of bidirectional modulation of the nACh α7 receptor and 5-HT3A receptors by ethanol, a chimera of the two was used to define a candidate region for ethanol sensitivity in these two receptors (Yu et al., 1996). The chimera was composed of the N terminus of the nACh α7 receptor, and the balance 5-HT3A receptor. Inhibition by ethanol was observed, consistent with the hypothesis that the N terminus confers sensitivity. In contrast, amino acids at position 15′ of the second transmembrane domain (TM2) and at position 16′ of the third transmembrane domain (TM3) of other ligand-gated ion channels, the γ-aminobutyric acid (GABA) type A and rho1 receptors, as well as the glycine receptor, determine the direction of alcohol and anesthetic modulation (Belelli et al., 1997; Mihic et al., 1997; Amin, 1999). Alcohol cutoff, which is the n-chain alcohol at which functional effects are lost altogether or are reduced to that of the n-1 alcohol, was altered in receptors containing mutations at these critical residues. Thus, mutant glycine receptors had the cutoff of GABA rho1 receptors, and vice versa (Wick et al., 1998). These results, plus correlation of molecular volume of these critical residues with direction or degree of modulation (Ye et al., 1998; Koltchine et al., 1999; Yamakura et al., 1999) has led to the speculation that an alcohol/anesthetic binding pocket may reside in a cavity between these amino acids in TM2 and TM3. However, the possibility that these drugs bind elsewhere in the receptor, with the mutations altering the transduction of the alcohol and anesthetic signal, cannot be ruled out.
The present study was undertaken to examine alcohol action on 5-HT3A receptors containing mutations at Ile294 of TM2, a residue that has recently been shown to face the channel pore (Reeves et al., 2001; Panicker et al., 2002). Ile294 is at position 16′ of TM2, neighboring the amino acid that is critical for n-chain alcohol and anesthetic modulation of the GABAA, GABA rho1, and glycine receptors at position 15′ (Belelli et al., 1997; Mihic et al., 1997; Amin, 1999). We tested the hypothesis that mutation of Ileu294 in TM2 changes the response to n-chain alcohol sensitivity in the absence of any actions on an alcohol binding site. Ethanol and TCEt potentiation of receptor function is largely lost and change in n-chain alcohol cutoff occurs with mutation, suggesting that kinetics of channel gating have been altered rather than the dimension and/or physicochemical characteristics of an alcohol/anesthetic binding pocket.
Materials and Methods
Site-Directed Mutagenesis. The cDNA encoding the NCB-20 5-HT3A receptor (a gift from Dr. D. Julius, San Francisco, CA) was subcloned in pBluescript KS– (Stratagene, La Jolla, CA). Numbering of the amino acids began with the initiating methionine. Mutant cDNAs were obtained with the unique site elimination method (U.S.E. mutagenesis kit; Amersham Biosciences Inc., Piscataway, NJ). In each mutagenesis reaction, selection pressure was created by deleting a unique restriction site in the plasmid DNA. Mutations were confirmed by dideoxynucleotide sequencing of double-stranded DNA at the Texas Tech University Biotechnology Core Facility (Lubbock, TX).
Expression of 5-HT3A Receptors in Xenopus laevis Oocytes. Ovarian lobes were obtained from X. laevis frogs and placed in modified Barth's solution (MBS) containing 88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO3, 10 mM HEPES, 0.82 mM MgSO4, 0.33 mM Ca(NO3)2, and 0.91 mM CaCl2 (pH 7.5). Stage V and VI oocytes were dissected with fine surgical forceps in a hypertonic isolation medium containing 108 mM NaCl, 2 mM KCl, 1 mM EDTA, and 10 mM HEPES (pH 7.5) and placed in MBS. Dissected oocytes were incubated for 10 min in buffer containing 0.5 mg/ml collagenase type IA and 83 mM NaCl, 2 mM KCl, 1 mM MgCl2, and 10 mM HEPES (pH 7.5) to remove the follicular cell layer. After several rinses with MBS, oocytes were placed in incubation medium composed of ND96, containing 96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, and 5 mM HEPES (pH 7.5), or MBS, plus 10 mg/l streptomycin, 50 mg/l gentamicin, 10,000 units/l penicillin, 0.5 mM theophylline, and 2 mM sodium pyruvate.
Wild-type and mutant cDNAs were linearized with NotI, extracted with phenol-chloroform, precipitated with sodium acetate and ethanol, and resuspended in diethyl pyrocarbonate-treated water. The cDNAs were then transcribed with T3 mMESSAGE mMACHINE (Ambion, Austin, TX). An aliquot of cRNA was centrifuged at 15,000g and the ethanol was removed. The pellet was resuspended in RNase-free water, and 5 to 30 ng of cRNA were injected per oocyte. Oocytes were stored in incubation medium, and were recorded from days 2 through 7 after injection.
Two-Electrode Voltage-Clamp Electrophysiological Recordings. Oocytes were perfused (2 ml/min) in a 100-μl volume chamber with MBS via a roller pump (Cole-Parmer Instrument, Co., Vernon Hills, IL). Two glass electrodes (1.2 mm outside diameter and 1–10 MΩ resistance) filled with 3 M KCl were used to impale oocytes. A Warner Instruments model OC-725B or OC-725C oocyte clamp (Hamden, CT) was used to voltage clamp oocytes to –70 mV. Clamping currents were plotted on a strip chart recorder (Cole-Parmer Instrument, Co.). Serotonin (Sigma-Aldrich, St. Louis, MO), ethanol (AAPER Alcohol and Chemical Co., Shelbyville, KY), and other n-chain alcohols (Sigma-Aldrich), and TCEt (>99% pure, Sigma-Aldrich) were dissolved in MBS buffer. Serotonin was applied for 30 s. n-Chain alcohols were perfused for 1 min before the application of 5-HT + n-chain alcohol for 30 s; preapplication was used to ensure that the relatively low potency n-chain alcohols were equilibrated with their putative alcohol binding domains before exposure to 5-HT + n-chain alcohol. Equipotent concentrations of 5-HT (∼EC1) were used for each receptor construct.
Data Analysis. The values in the 5-HT concentration response curves for wild-type and mutant 5-HT3A receptors were expressed as a percentage of the respective maximal 5-HT (10 μM) responses. Unless otherwise noted, in all other experiments, data were expressed as percent change from the control, baseline response. GraphPad Prism (GraphPad Software Inc., San Diego, CA) was used to calculate EC50 values, Hill coefficients, and two-way analysis of variance (ANOVA). GraphPad InStat was used to perform Student's t tests.
Expression of 5-HT3A Receptors in HEK 293 Cells. Wild-type and mutant 5-HT3A cDNAs were subcloned into pcDNA3.1 (Invitrogen, Carlsbad, CA). HEK 293 cells grown in standard 35-mm-diameter culture dishes were transfected by calcium phosphate precipitation as previously described (Lovinger and Zhou, 1994). Cells were maintained in a 95% O2/5% CO2 incubator, and recombinant receptor activity was measured beginning ∼48 h after transfection.
Whole Cell Patch-Clamp Electrophysiological Recordings with Rapid Drug Application. In all experiments transfected cells were replated as single cells onto suspension dishes before initiating the recording. Cells were resuspended in a phosphate-buffered saline solution containing 3 mM EDTA. The cell suspension was centrifuged, and cells were resuspended with light trituration in the external medium used for electrophysiological recording (see below). Cells were then plated onto 35-mm-diameter suspension dishes and were allowed to settle on the dish for at least 5 min before recording was initiated.
Whole cell patch-clamp recordings were performed on transfected HEK 293 cells bathed in external solution containing (in mM): 150 NaCl, 2.5 KCl, 2.5 CaCl2, 10 HEPES, 10 d-glucose (pH adjusted to 7.4 with NaOH and osmolality adjusted to 340 mOsmol/kg with sucrose) as described previously (Zhou and Lovinger, 1996; Zhou et al., 1998) with an Axopatch 200 amplifier (Axon Instruments, Union City, CA). The solution constantly superfused cells at a rate of 2 to 3 ml/min. Patch pipettes had resistances of 1 to 2 MΩ when filled with (in mM): 140 CsCl, 10 cesium-EGTA, and 10 HEPES (pH adjusted to 7.4 with CsOH and osmolality adjusted to 313–316 mOsmol kg–1 with sucrose). Cells were voltage-clamped with the membrane potential held at –70 mV. After establishing a recording, the cell was lifted clear of the bottom of the suspension dish and was placed into a solution stream coming from the drug application micropipette. In selected experiments, alcohols were applied alone for 30 s before application of agonist + alcohol.
Alcohol and other pharmacological agents were dissolved directly into external solution, and applied to transfected HEK cells using two different techniques. In the majority of experiments, agonist was applied with a θ-tube whose lateral movement was controlled by a piezoelectric manipulator, as previously described (Zhou et al., 1998). The solution exchange time using this technique is 20 to 40 ms in the whole cell recording mode. In a few experiments drug application was accomplished with a stepper-motor mounted on a micromanipulator (Warner Instruments). In this system solution was applied from glass tubes mounted on the manipulator, and solution exchange time was 22 ± 5 ms. At the beginning of each experiment the cell was placed in front of a solution stream of standard extracellular solution. Agonist application was initiated under computer control, resulting in lateral displacement of the application pipette by 100 to 200 μm such that the cell was superfused by a solution containing agonist ± alcohol. Cessation of the computer-generated pulse reversed the pipette displacement, returning the cell to the standard external solution stream.
Solution applications ranged in duration from 1 to 30 s. Data were filtered at 5 kHz with a 3-pole Bessel filter and digitized at 10 kHz. In the majority of experiments, data were acquired using an Axopatch 200 amplifier and digitized using an ITC-16 analog/digital interface linked to a MacIntosh IIfx computer by means of IGOR software (Wavemetrics Inc., Lake Oswego, OR). In a few experiments data acquisition was accomplished using a TL-1-125 interface (Scientific Solutions, Mentor, OH) linked to a Pentium microcomputer running pClamp v5.5 software (Axon Instruments). Data were analyzed offline using either IGOR v3.0 or pClamp v8.0 software. Current amplitudes were determined using a cursor-based system to measure current at a particular time point, and these amplitudes were then subtracted from the baseline, preagonist current. Peak current was the maximal agonist-induced current, and was defined as the difference between the maximal current and the baseline preagonist holding (trough) current. Steady-state current was the stable current observed during prolonged (30-s) agonist application. Steady-state currents were measured at 30 s. It must be noted that at 2 μM 5-HT (see Figs. 2B and 5A) the Ile294Thr mutant receptor reached an apparent steady state at 30 s. Very slight decrements in current were observed between 25 and 30 s of agonist exposure. Nonlinear curve-fitting and estimates of the time course of exponential decay were obtained using IGOR software. The initial current slope was calculated for data points between 10 and 20% of the maximal current on the downward-sloping initial phase of current. Measurement of 10% of the peak response eliminates the contributions of nonlinear components of current at the bottom of a presumed sigmoidal-shaped response. Measurement at 20% of the peak current, rather than at higher percentages of peak current, ensures that slope can be measured accurately, with the least possible contamination from any other current component. All current amplitudes were normalized to whole cell capacitance and are reported as pA pF–1. All averaged measurements are expressed as mean ± S.E.M. values.
Results
Receptors containing mutations at Ile294 of the mouse 5-HT3A receptor were initially compared with the wild-type receptor for their sensitivity to 5-HT (Fig. 1, A and B). In oocytes expressing the receptors, mutations at Ile294 resulted in shifts of the 5-HT concentration response curves (Fig. 1A), n = 4. Two-way ANOVA demonstrated that mutation had a significant effect on the 5-HT concentration response curves (F(2,54) = 3.88, p = 0.03). The EC50 value (in μM) of the Ile294Thr 5-HT3A receptor (0.65 ± 0.015) was significantly less than that of the wild-type receptor (0.93 ± 0.023), Student's t test, p < 0.0001. The EC50 value of the Ile294Leu mutant (1.15 ± 0.14) was significantly greater than that of the wild-type receptor, Student's t test, p = 0.0003. Hill coefficients were 3.2 ± 0.22 (wild-type), 4.1 ± 0.36 (Ile294Thr), and 3.1 ± 0.14 (Ile294Leu).
In HEK 293 cells expressing wild-type and Ile294Thr 5-HT3A receptors, 5-HT concentration response curves were generated (Fig. 1B). Two-way ANOVA revealed that the curves were not significantly different (F(1,26) = 3.36, p = 0.08). EC50 values (in μM) and Hill coefficients were 3.82 ± 0.11 and 2.2 ± 0.13 (wild-type) and 2.73 ± 0.18 and 1.7 ± 0.2 (Ile294Thr). It should be noted that the difference in 5-HT potency in oocytes compared with HEK 293 cells is due to the relatively slow agonist application in oocytes where current is measured after many of the receptors have undergone desensitization, in which case apparent agonist affinity and potency are increased. In addition, the observation in the oocyte expression system that the Hill coefficient of the Ile294Thr mutant was greater than that of the wild-type receptor may reflect an underestimation of the peak current, and hence a steeper curve.
Currents generated in the presence of 5-HT in cells expressing wild-type or mutant receptors are presented in Fig. 2. In Fig. 2A, representative currents obtained from oocytes expressing wild-type, Ile294Thr, and Ile294Leu receptors are shown. Similar desensitization of the responses to 10 μM 5-HT were observed between wild-type and Ile294Leu receptors. In contrast, the most striking aspect of the currents generated in the Ile294Thr mutant expressed in oocytes is the slow decay toward baseline of current during the prolonged application of 10 μM 5-HT (Fig. 2A). Rapid current decay was observed in the wild-type receptor; however, recovery of currents in the Ile294Thr mutant was only ∼33% at the end of the 30-s application of 5-HT (10 μM). Generally, a minimum of 4 min was required for the holding current to reach the baseline, pre-5-HT application levels. In Fig. 2B, current traces obtained in wild-type and Ile294Thr 5-HT3A receptors expressed in HEK 293 cells are shown. The most pronounced effect is the reduction in the amount of desensitization upon prolonged exposure to 10 μM 5-HT (30 s) in the Ile294Thr mutant. Furthermore, 5-HT-mediated current activated by a low concentration of agonist (2 μM) exhibited a sustained current in the Ile294Thr mutant, in contrast to the rapid peak and desensitization of currents produced by the same agonist concentrations in the wild-type receptor (Fig. 2B). The insets in Fig. 2B are presented with an expanded time scale and highlight the differences in current onset and desensitization of the wild-type and mutant receptors.
With the rapid drug application protocol we measured several of the parameters of whole cell current in receptors expressed in HEK 293 cells that provide information about macroscopic kinetics and modulation of ion channel function by alcohols, as shown previously (Zhou et al., 1998). These include the initial slope of current during the activation phase, the 20 to 80% rise-time, the steady state/peak (SS/Peak) current ratio, and the rate of current decay during prolonged agonist application. The Ile294Thr mutation did not produce a significant increase in initial current slope (101.2 ± 35 pA pF–1 s–1, n = 5) relative to the wild-type receptor (89.3 ± 62.7 pA pF–1 s–1, n = 3), (Student's t test, ⋆p = 0.86). The 20 to 80% rise-time of the Ile294Thr mutant (1.7 ± 0.46 s, n = 5) was not significantly greater than that of the wild-type receptor (0.37 ± 0.22 s, n = 3), (Student's t test, ⋆p = 0.09), given the large standard error in the measurements. However, a longer rise-time would indicate that during this phase of activation the mutant is slower than the wild-type receptor. In the example shown in Fig. 2B in the inset at the extreme right of the figure, wild-type and Ile294Thr receptor current traces in response to 2 μM 5-HT are overlaid. The slower rise time in the mutant receptor is apparent. A greater than twofold increase in SS/Peak current ratio was observed in the Ile294Thr mutant relative to the wild-type receptor upon application of 5-HT (10 μM) (Fig. 2C) (Student's t test, ⋆p = 0.0006). This increase reflects a decrease in the extent of desensitization; i.e., steady-state levels were achieved after receptors had desensitized ∼60% upon prolonged exposure to 5-HT (10 μM). In contrast, in the wild-type receptor, currents achieved steady state after ∼95% desensitization. Furthermore, the rate of desensitization (1/τ) was reduced in the presence of the Ile294Thr mutation, as measured by the time constant for current decay (τ) following the peak response in the continued presence of agonist (Fig. 2D) (Student's t test, ⋆p = 0.0006). The mutation decreases the rate and extent of desensitization of the 5-HT3A receptor. One probable outcome of the changes in desensitization is the open state of the mutant receptor channel would likely be stabilized relative to the wild-type receptor.
We next examined the modulatory actions of ethanol on wild-type and mutant 5-HT3A receptors expressed in Xenopus oocytes and HEK 293 cells (Fig. 3). The function of the wild-type and Ile294Leu 5-HT3A receptors expressed in oocytes was similarly enhanced by ethanol over the range of concentrations tested (Fig. 3A). Stimulation increased from an average of 15% at ethanol (50 mM) to ∼55% at ethanol (200 mM). In contrast, the Ile294Thr 5-HT3A receptor was inhibited by ethanol, but the concentration-response curve for inhibition was relatively flat, ranging from ∼15 to 27%. Two-way ANOVA demonstrated that mutation had a significant effect (F(3,54) = 228, p < 0.0001). The Ile294Thr 5-HT3A receptor was significantly different from the wild-type receptor at all ethanol concentrations tested (Tukey's post hoc test, ⋆p < 0.05). In HEK 293 cells (Fig. 3B), 5-HT-mediated currents were measured in the absence and presence of ethanol in the wild-type and Ile294Thr receptors. Stimulation of ∼17 and 26% was observed in the wild-type receptor with 50 and 100 mM ethanol, respectively. No effects on receptor function were obtained in the Ile294Thr 5-HT3A receptor with ethanol (10–100 mM). The percentage changes in response to ethanol were not significantly different between wild-type and Ile294Thr receptors, two-way ANOVA, (F(1,22) = 3.47, p = 0.08). Ethanol applied alone did not elicit any detectable current in the wild-type or Ile294Thr 5-HT3A receptors expressed in oocytes (n = 5) or HEK 293 cells (n = 3–4) (data not shown).
In previous studies we have demonstrated that wild-type 5-HT3A receptor function is more robustly enhanced by TCEt than by ethanol (Lovinger and Zhou, 1996; Zhou et al., 1998), and we have also performed extensive kinetic analysis of TCEt's actions at this receptor (Zhou et al., 1998). Therefore, it was of interest to examine the actions of TCEt on Ile294Thr receptors and to compare the receptor-channel macroscopic kinetic parameters of channel function in the presence of TCEt in wild-type and Ile294Thr receptors. We first examined the effect of TCEt on wild-type and Ile294Thr 5-HT3A receptors expressed in Xenopus oocytes; the Ile294Leu mutant was not studied because of the lack of change in modulation by ethanol due to this mutation (Fig. 4). A concentration-dependent enhancement of 5-HT-evoked currents was seen in wild-type receptors, ranging from ∼55 to 2646% with 0.25 to 10 mM TCEt. A biphasic response was observed in oocytes expressing the Ile294Thr mutant. Receptor function was inhibited by ∼58 and 34% at 0.25 and 5 mM TCEt, respectively. No effects were observed at 1 and 2 mM TCEt, whereas stimulation occurred at 5 and 10 mM TCEt. Two-way ANOVA revealed that the curves were significantly different (F(1,78) = 106, p < 0.0001). TCEt applied alone did not elicit any detectable current in the wild-type or Ile294Thr 5-HT3A receptors expressed in oocytes (n = 4–6) (data not shown).
In studies with rapid drug application, we focused on measuring kinetic parameters of wild-type and Ile294Thr 5-HT3A receptors expressed in HEK 293 cells in the absence and presence of 3 mM TCEt. No detectable currents were observed in wild-type or Ile294Thr 5-HT3A receptors when TCEt was applied alone (n = 3–4) (data not shown). Representative tracings of 5-HT evoked currents generated in wild-type and mutant receptors are shown (Fig. 5A). As demonstrated previously (Zhou et al., 1998), TCEt enhanced wild-type receptor function (panel a). However, Ile294Thr receptor currents were slightly inhibited by TCEt, with an apparent decrease in the slope of activation (panel b). The insets (panels a and b) in Fig. 5A are presented using an expanded time scale and highlight the onset of current in the wild-type and mutant receptors. In Fig. 5A, panels c and d, wild-type and Ile294Thr receptor current traces in response to 5-HT or 5-HT + TCEt are overlaid; these are the same traces as presented in panels a and b, but are overlaid to illustrate the differences between the constructs. Activation is slower in the mutant (panel c, inset), but is even slower in the mutant in the presence of TCEt (panel d, inset). Both peak and isochronal (4.5 s) measurements were made in each receptor construct (Fig. 5B), and data were expressed as ((ITCEt/Icontrol) × 100), where ITCEt is the current obtained in the presence of TCEt and Icontrol is the current obtained in its absence. Both peak and isochronal measurements were performed to take into account the fact that peak currents occur at different time points and that desensitization is also altered in the two constructs. Clearly, in the wild-type receptor both peak (Student's paired t test, *p = 0.04) and isochronal currents (Student's paired t test, *p = 0.04) were increased, whereas in the mutant receptor an insignificant decrease was observed in peak (Student's paired t test, p = 0.37) and isochronal currents (Student's paired t test, p = 0.17). The wild-type and Ile294Thr receptor peak (Student's t test, +p = 0.03) and isochronal currents (Student's t test, +p = 0.01) were different from each other.
Measurements of initial slope and rise-time (20–80% of the response), which represent activation of the receptor, in HEK 293 cells were performed in the presence of TCEt and expressed as a percentage of that obtained in its absence (Fig. 5C). Activation rate of the wild-type receptor was enhanced ∼70% in the presence of TCEt. The slope of 61.4 ± 18 pA pF–1 s–1 in the absence of TCEt was significantly different from the slope of 87.8 ± 20.6 pA pF–1 s–1 in the presence of TCEt (Student's paired t test, *p = 0.03). No effect was observed in the mutant receptor (Student's t test, p = 0.42). The slopes were significantly different between the wild-type and Ile294Thr receptors (Student's t test, +p = 0.004). Rise-time was unaffected by the presence of TCEt in either construct (wild-type: Student's paired t test, p = 0.26; mutant: Student's paired t test, p = 0.81). The rise-times of the wild-type and mutant receptors were not significantly different from one another (Student's t test, p = 0.66). Finally, SS/Peak current ratios were compared in the absence and presence of TCEt in the wild-type and mutant receptors (Fig. 5D). As seen in Fig. 5A, wild-type receptors desensitize rapidly in the presence of 2 μM 5-HT and thus the SS/Peak current ratios are less than 0.1. Addition of TCEt increases both peak and steady-state currents in the wild-type receptor, and the increase in the SS/Peak current ratio in the presence of TCEt is significant (Student's paired t test, p = 0.03). Minimal desensitization takes place in the Ile294Thr 5-HT3A receptor (Fig. 5A), which is reflected by an SS/Peak current ratio that is greater than 0.8. The addition of TCEt does not further increase the SS/Peak current ratio in the mutant (Fig. 5D), (Student's paired t test, p = 0.13). The SS/Peak current ratios are different between wild-type and Ile294Thr 5-HT3A receptors in the presence of 2 μM 5-HT (Student's t test, +p < 0.0001) or 2 μM 5-HT + 3 mM TCEt (Student's t test, + p < 0.0001).
Alcohol cutoffs were determined in wild-type and mutant 5-HT3A receptors expressed in Xenopus oocytes. The alcohol cutoff for a protein is the n-chain alcohol at which functional effects are lost altogether or are reduced to that of the n-1 alcohol. Jenkins et al. (1996) previously showed that hexanol and lower chain alcohols stimulate 5-HT3A receptor evoked currents but that octanol and higher chain alcohols produce inhibition. A series of n-chain alcohols were examined. In Fig. 6A, pentanol (0.25–4 mM) enhanced the function of both wild-type and Ile294Leu 5-HT3A receptors, although the mutant receptor was less sensitive as determined by two-way ANOVA (F(1,39) = 53, p < 0.0001). The actions of hexanol on wild-type and mutant receptors were examined. In B and C, hexanol (0.25–10 mM) enhanced the function of wild-type receptors, but inhibited or had no effect on Ile294Leu and Ile294Thr receptors. Where responses to hexanol (0.1–1 mM) were measured, two-way ANOVA yielded significant effect of mutation (F(6,52) = 198, p < 0.0001), and two-way ANOVA also revealed a significant effect of mutation for hexanol at the 5 and 10 mM concentrations (F(2,25) = 54.96, p < 0.0001) (Tukey's post hoc test, *p < 0.05, compared with wild-type). These results suggest that the alcohol cutoff for the Ile294Leu 5-HT3A receptor is hexanol. Actions of heptanol (0.1–1 mM) and octanol (0.03125–1 mM) were evaluated in wild-type and Ile294Thr receptors (Fig. 6D). Minimal potentiation of 5-HT-mediated currents (<27%) was observed in wild-type receptors over the range of heptanol concentrations tested, suggesting that the alcohol cutoff for the wild-type receptor is heptanol. Robust inhibition of the Ile294Thr mutant was produced with heptanol, and the wild-type and Ile294Thr receptors' concentration response curves were significantly different, two-way ANOVA (F(1,24) = 112.5, p < 0.0001). The Ile294Thr mutant was modulated by octanol to an extent similar to the wild-type receptor, two-way ANOVA (F(1,25) = 2.45, p = 0.1302). Inhibition that is observed with higher concentrations of octanol in both of the receptors has been previously reported for the wild-type receptor and is believed to represent a distinct alcohol site(s) from the potentiating site(s) (Jenkins et al., 1996). Given that the Ile294Thr 5-HT3A receptor is inhibited by all n-chain alcohols, it is difficult to ascertain alcohol cutoff for the inhibition observed in the mutated receptor.
Discussion
In the present study we examined the effects of two mutations at position 16′ of TM2 of the 5-HT3A receptor on kinetics of channel gating, allosteric modulation by n-chain alcohols, and n-chain alcohol cutoff. The two mutations, Ile294Leu and Ile294Thr, were conservative and nonconservative, respectively. Ile and Leu are very similar, with Ile having slightly greater hydrophobicity. Thr has less bulk and hydrophobicity, but greater polarity than Ile. Ile294 faces the channel pore and therefore is a poor candidate as a residue that forms an alcohol-binding domain, given that n-chain alcohols enhance 5-HT3A receptor function. We tested the hypothesis that mutation of Ile294 may alter the receptor's response to n-chain alcohols in the absence of any change in a putative alcohol-binding domain. In the Ile294Thr mutant, the rate and extent of desensitization were reduced relative to those observed in the wild-type receptor. We have demonstrated previously that TCEt similarly enhances activation rate and decreases the intrinsic desensitization rate in the wild-type receptor (Zhou et al., 1998). Thus, we propose that the loss of enhancing actions of TCEt in the Ile294Thr receptor is not due to a change in an alcohol-binding domain, but instead occurs as a consequence of kinetic changes that favor the open channel state and likely occlude the effects of alcohol.
Each of the members of the superfamily of ligand-gated ion channels has been mutated in the TM2 domain, but the difficulty in studying such receptors lies with the fact that the pore-lining domains are critical in participating in the conformational changes that occur when the channels open. This gating process may be fundamentally disturbed in a variety of ways (reviewed in Colquhoun, 1998). Changes in desensitization are demonstrated in the present study, where the Ile294Thr 5-HT3A receptor shows a marked reduction in extent and rate of desensitization. Other mutation-induced changes that have been reported are changes in affinity, shift from antagonism to agonism (Bertrand et al., 1992), loss of allosteric modulation (Findlay et al., 2001), and production of tonically open channels (Findlay et al., 2001). Collectively, these data suggest that mutations in TM2 alter the state-dependence of these channels by shifting the probability of closed → open → desensitized transitions. Such changes would have a high likelihood of interfering with allosteric changes in channel gating such as those produced by alcohols.
Mutation at Ile294 produced a number of changes in gating and allosteric modulation in the 5-HT3A receptor. The EC50 value for the Ile294Thr 5-HT3A receptor was significantly reduced relative to the wild-type receptor when oocytes were used as an expression system. Although not significantly lower, the EC50 value for the Ile294Thr 5-HT3A receptor was reduced in HEK 293 cells. In contrast, the Ile294Leu mutant, when expressed in oocytes, had a greater EC50 value than the wild-type receptor. The Ile294Thr and not the Ile294Leu mutant lost positive allosteric modulation by ethanol. These results are consistent with the findings of others that demonstrate that a TM2 mutation-induced increase in agonist potency correlates with a decrease in alcohol or anesthetic modulatory actions (Yamakura et al., 1999). A mutant 5-HT3A receptor with greater agonist potency than the wild-type receptor is more likely to open at any given 5-HT concentration. From a mechanistic standpoint, the resting conformation of such a mutant is more thermodynamically favored to make a transition from a closed to an open state. Positive allosteric modulation is typically lost when probability of opening is increased, as exemplified by loss of ethanol modulatory actions at high concentration of 5-HT in the wild-type receptor (Machu and Harris, 1994).
The most striking change produced by the Ile294Thr mutation was the reduction in the rate and extent of desensitization compared with wild-type receptors. Changes in desensitization in the mutant receptor suggest that once activated by 5-HT, it is more likely than the wild-type receptor to remain in the open channel state. The most parsimonius explanation for the reduction in enhancing effect of ethanol and TCEt on the Ile294Thr 5-HT3A receptor that we have observed is that the open channel state is already favored in the absence of alcohols and that the mutation actually mimics the actions of alcohols, albeit to a greater extent, on desensitization. Thus, the Ile294Thr mutation may actually mask the stimulatory effects of alcohols. Our data support this conclusion in that TCEt significantly increases steady-state to peak current ratios in the wild-type receptor, but not the Ile294Thr5-HT3A receptor. Further supporting evidence comes from work in the GABAA receptor. Mutations in either TM2 or in the N terminus have been described that produce tonically open channels, an extreme case of increased probability of channel opening wherein a certain fraction of receptors are constantly open. These receptors have very reduced or no sensitivity to allosteric modulation by alcohols and anesthetics (Thompson et al., 1999; Findlay et al., 2001; Zhang et al., 2002).
Two contrasting effects on activation were observed in the Ile294Thr 5-HT3A receptor. Both the initial slope and the 20 to 80% rise-time were increased, although not significantly, in the Ile294Thr mutant relative to the wild-type receptor. An increase in activation rate would favor the open channel state in the mutant relative to the wild-type receptor. However, an increase in rise-time means that the time required for opening is longer in the mutant, suggesting that the wild-type receptor is more favored to open when this parameter is considered. It is important to note that the 20 to 80% rise-time is a less precise measure of activation, since it may be contaminated with other components of the whole cell current. We suggest that the desensitization of some of the mutant channels may be occurring while others are still opening. Although it is the summation of all factors that contribute to the stabilization of the open state, we suggest that the effects of the mutation on rate and extent of desensitization are likely to be the most important. That is, once the Ile294Thr receptor opens, it stays open, which prolongs the later stages of current onset, as is evidenced by the overlaid tracings in Fig. 5A, panel c.
TCEt had contrasting effects on 5-HT-mediated currents in the wild-type and Ile294Thr 5-HT3A receptors. In the wild-type receptor, TCEt increased the initial slope, but had no effect on 20 to 80% rise-time. In the mutant receptor, the initial slope of current is reduced, although nonsignificantly, by TCEt. Rise-time in the mutant receptor was unaffected by TCEt. Taken together, these data suggest that inhibition of activation of channel opening is the third mechanism through which the potentiating effects of alcohol are lost in the mutant receptor. In fact, inhibition of activation may in part account for blockade of mutant receptor function (oocyte expression system) with TCEt at lower concentrations and of mutant receptor function (HEK 293 cell expression system) at 3 mM TCEt. Thus, slower current activation, combined with reduced desensitization, likely contribute to the loss of alcohol effects. The biphasic response observed over the TCEt concentrations tested (0.125–10 mM), with potentiation occurring at 5 and 10 mM, may reflect TCEt binding at lower affinity sites whose allosteric modulation is less affected or unaffected by the mutation.
The kinetic data that we have shown provide us with a compelling argument that the lack of stimulatory action of alcohols on the Ile294Thr 5-HT3A receptor is not due to an alteration of a binding pocket for alcohols and anesthetics. Another supporting piece of evidence is the finding that Ile294 faces the pore of the channel (Reeves et al., 2001; Panicker et al., 2002). Given the fact that alcohols enhance 5-HT3A receptor function, it would be difficult to reconcile the notion of an alcohol binding domain that included amino acids in the pore. A pore-lining pocket would be more consistent with an inhibitory, channel-blocking action of alcohols. Furthermore, in the face of other evidence presented, our observation that the Ile294Leu mutant has a reduced cutoff relative to the wild-type receptor does not lend credence to the idea of a binding pocket at Ile294, and is likely a coincidental finding.
In summary, we have shown that the Ile294Thr 5-HT3A receptor has a markedly altered kinetic profile in comparison with the wild-type receptor. Changes in the extent and rate of desensitization and increased activation rate, as well as slowing of activation in the presence of alcohols, can explain the loss of stimulatory effects of ethanol and TCEt. These results, coupled with the finding that Ile294 faces the pore of the channel, cast doubt that Ile294 is part of an alcohol-binding domain. Our results suggest that loss of alcohol and anesthetic modulation of function and changes in alcohol cutoff should not be used as the sole criteria for identifying a putative alcohol-binding domain. Kinetics of channel function should be evaluated in a series of mutants at an apparently critical residue before an alcohol-binding site should be invoked.
Footnotes
-
Financial support was provided by grants from the Alcoholic Beverage Medical Research Foundation (T.K.M.) and National Institute of Alcohol Abuse and Alcoholism Grants AA12643 (T.K.M.) and AA08986 (D.M.L.).
-
DOI: 10.1124/jpet.103.050542.
-
ABBREVIATIONS: 5-HT, 5-hydroxytryptamine; TCEt, 2,2,2-trichloroethanol; nACh, nicotinic acetylcholine; TM2, transmembrane 2; MBS, modified Barth's solution; ANOVA, analysis of variance; HEK, human embryonic kidney.
-
↵1 Current address: Dept. of Pharmacology, Duke University Medical Center, Research Drive, Durham, NC 27710.
-
↵2 Current address: Laboratory of Integrative Neuroscience, National Institute of Alcohol Abuse and Alcoholism, Room 158H Park 5 Bldg., 12420 Parklawn Drive, Rockville, MD 20852.
- Received February 17, 2003.
- Accepted May 2, 2003.
- The American Society for Pharmacology and Experimental Therapeutics
References
JPET articles become freely available 12 months after publication, and remain freely available for 5 years.Non-open access articles that fall outside this five year window are available only to institutional subscribers and current ASPET members, or through the article purchase feature at the bottom of the page.
|