Introduction

Top Summary Introduction Experimental Procedures Results Discussion References

The 5-hydroxytryptamine type 3 (5-HT₃) receptor is a transmitter-gated, cation-selective ion channel containing five transmembrane-spanning glycoprotein subunits (Derkach et al., 1989; Boess et al., 1995). A 5-HT₃ subunit, termed 5-HT_3A, has been identified by expression cloning from a murine hydridoma cell line cDNA library (Maricq et al., 1991). Subsequent homology screening has allowed the isolation of splice variants and 5-HT_3A subunit orthologs from human (Belelli et al., 1995; Miyake et al., 1995), rat (Isenberg et al., 1993; Miyake et al., 1995), and guinea pig (Lankiewicz et al., 1998) sources. Most recently, a novel 5-HT_3B subunit, capable of forming hetero-oligomeric complexes with the 5-HT_3A subunit, has been identified (Davies et al., 1999).

Despite their highly conserved structure, 5-HT_3A subunit orthologs display distinctive pharmacological profiles that closely reflect the interspecies variation in ligand binding described for neuronal 5-HT₃ receptors (Peters et al., 1997). A striking example is provided by the nicotinic acetylcholine receptor (AChR) antagonist (+)-tubocurarine [(+)-Tc], which binds to native (pK_i = 6.7-7.3; Bonhaus et al., 1993) or recombinant receptors (pK_i = 6.7-7.1; Bonhaus et al., 1995; Yan et al., 1999) of mouse origin (m5-HT_3A) with an affinity comparable with that found for the high- affinity site formed at the interfaces of the / subunits of the Torpedo nicotinic AChR (Pedersen and Cohen, 1990) and the / and / subunits of the mammalian skeletal muscle nicotinic AChR (Sine, 1993; Papineni and Pedersen, 1997; Bren and Sine, 1997). In contrast, (+)-Tc demonstrates much lower affinity at the human recombinant (h5-HT_3A; pK_i = 4.4; Hope et al., 1996) or native (pK_i = 4.8; Bufton et al., 1993) receptor isoforms.

Knowledge of the structural determinants of ligand binding at 5-HT₃ receptors is limited. That the extracellular N-terminal domain of the receptor imparts ligand binding specificity is evident from studies on a chimeric construct of the N-terminal domain of the nicotinic AChR 7 subunit and the residual sequence of the m5-HT_3A subunit (Eiselé et al., 1993). The chimera displayed nicotinic receptor pharmacology, but the ion channel properties of the m5-HT_3A receptor. Oxidation of tryptophan residues within the 5-HT₃ receptor reduces the binding of the selective antagonist [³H]zacopride in a manner that can be prevented by preincubation with some, but not all, 5-HT₃ receptor ligands (Miquel et al., 1991). For the m5-HT_3A subunit, the replacement of individual tryptophan residues within the N-terminal domain by either Tyr and/or Ser causes, with the exception of Trp-67, loss of ligand recognition and function (Spier et al., 1997). Replacement of Trp-67 by Tyr or Phe decreased the affinity of (+)-Tc, but also other ligands, such as granisetron (Spier et al., 1997; Yan et al., 1999). Interestingly, Trp occupies an homologous location in nicotinic AChR 7, , and  subunits (i.e., 7Trp-54, Trp-55, and Trp-57) and participates in the binding of (+)-Tc (O'Leary et al., 1994; Corringer et al., 1995). However, Trp-67 and adjacent residues are conserved in all 5-HT_3A subunit orthologs, indicating that additional residues must be responsible for the differential potency of (+)-Tc. Similarly, although mutation of Glu-106 of the m5-HT_3A subunit produces differing effects upon the binding of several ligands (Boess et al., 1997), conservation of this residue and flanking sequences argues against involvement in the differential affinity of (+)-Tc.

The window of selectivity between the binding of (+)-Tc at high (i.e., / or /) and low (i.e., /) affinity subunit interfaces of the muscle nicotinic AChR has, via the construction of chimeras and site-directed mutations, revealed residues that participate in (+)-Tc binding (Pedersen and Cohen, 1990; Sine, 1993; Papineni and Pedersen, 1997; Bren and Sine, 1997). Here, we exploit the differential affinity of (+)-Tc at m5-HT_3A and h5-HT_3A receptors to identify residues that might influence binding. We elected to quantify antagonism by (+)-Tc using an electrophysiological assay of wild-type and mutant 5-HT_3A subunits expressed in Xenopus laevis oocytes because of the robust difference (i.e., ~1,850-fold) in the potency of the antagonist at human and mouse 5-HT_3A receptors in this system (Hope et al., 1993; Belelli et al., 1995; Peters et al., 1997). Subunit chimeras wherein the N-terminal domains of human and mouse 5-HT_3A receptor subunits were exchanged reciprocally were constructed to verify that differences in the potency of (+)-Tc are due wholly to binding in the extracellular N-terminal region. By the construction of mutant receptors, we demonstrate the reciprocal exchange of small numbers of residues between human and mouse receptors to produce opposite effects upon the potency of (+)-Tc.

	Experimental Procedures

Top Summary Introduction Experimental Procedures Results Discussion References

Materials. (+)-Tubocurarine hydrochloride and 5-HT creatine sulfate complex were purchased from Sigma Chemical Co. (Poole, Dorset, UK). Oligonucleotides were synthesized by Cruachem Ltd. (Glasgow, Strathclyde, UK).

Construction of Chimeric 5-HT₃ Receptor Subunits and Mutagenesis. Two chimeric 5-HT₃ receptor subunits containing human and mouse 5-HT_3A sequences were constructed using a polymerase chain reaction (PCR)-based approach (Yon and Fried, 1989). The chimeric receptor subunit h218m 5-HT_3A, where the number denotes the position of the residue immediately amino terminal to the chimeric junction, comprised the extracellular N-terminal domain of the h5-HT_3A subunit and the remaining sequence of the m5-HT_3A ortholog. The 5' and 3' ends of the h218m 5-HT_3A chimera were defined by the oligonucleotides H1 (CCGGAATTCCGGGGCCACGAGAGGCAG) and M1 (CCGCTCGAGAAGATATCATAGCATTTTTATT) containing EcoRI and XhoI restriction sites (underlined), respectively. The chimeric junction was defined by a single large oligonucleotide C1 (ATGTGGTCATCCGCCGGCGGCCTTTATTCTATGCAGTCAG). Conversely, the amino- and carboxy-terminal components of m223h 5-HT_3A were derived from the corresponding regions of m5-HT_3A and h5-HT_3A, respectively. The 5' and 3' ends of m223h 5-HT_3A were defined by the oligonucleotides M2 (CCGGCTCGAGACATCTGGGAAGCTTGCCAT) and H2 (CCGGAATTCCAAAGTCCC) embodying an XhoI or EcoRI site. The chimeric junction was defined by oligonucleotide C2 (ACGTGATCATCCGCCGGAGGCCCCTCTTCTATGTGGTCAG). Both chimeric 5-HT_3A subunits were amplified using a 30-cycle PCR (denaturing, 95°C, 30 s; annealing, 65°C, 30 s; and extension, 75°C, 4 min). Each PCR contained 10 ng of m5-HT_3A and h5-HT_3A cDNAs, 10 mM KCl 10, 10 mM (NH₄)₂SO₄, 20 mM Tris-Cl (pH 8.8), 2 mM MgSO₄, 0.1% Triton X-100, 100 µg ml¹ BSA, 200 µM each dNTP and 2.5 U of plaque-forming unit polymerase. In both cases, the outer (5' and 3') primers were present at a concentration of 1 µM, while the central oligonucleotide (C1 or C2) was present at 0.01 µM. h218m 5-HT_3A and m223h 5-HT_3A were cloned into pcDNA1amp (In Vitrogen BV, NV Leek, The Netherlands) and Bluescript SK+ (Stratagene Ltd., Cambridge, UK) respectively, before expression in Xenopus laevis oocytes.

Expression of Chimeras, Mutant Receptors, and Electrophysiological Analysis. Xenopus laevis ooctyes were isolated and enzymatically defolliculated as previously described (Hope et al., 1993). cDNA (20 nl; 5-250 ng µl¹) encoding chimeric or mutant 5-HT_3A receptor subunits was injected into the nucleus of Stage V-VI oocytes, which were subsequently stored individually at 19-20°C for 2 to 10 days in 96-well microtiter plates in 200 µl of Barth's solution (composition 88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO₃, 1 mM MgSO₄, 0.5 mM CaCl₂, 0.5 mM Ca(NO₃)₂, and 15 mM HEPES, pH 7.5) supplemented with gentamicin (100 µg ml¹).

1

	Results

Top Summary Introduction Experimental Procedures Results Discussion References

Mouse and Human 5-HT_3A Subunit Chimeras. Previous studies examining the antagonist potency of (+)-Tc at h5-HT_3A and m5-HT_3A receptors expressed in Xenopus laevis oocytes yielded IC₅₀ values of 2.6 µM and 1.4 nM, respectively (Hope et al., 1993; Belelli et al., 1995). To confirm that this large differential is due exclusively to differences in primary amino acid sequence residing within the extracellular N-terminal domain preceding the first transmembrane span, the chimeras m223h 5-HT_3A and h218m 5-HT_3A were constructed and examined for sensitivity to block by (+)-Tc. At a holding potential of 60 mV, both chimeric constructs mediated large inward current responses to 5-HT applied at a half-maximally effective concentration (EC₅₀; Fig. 1). As anticipated from the close similarity in the EC₅₀ values for 5-HT at wild-type m5-HT_3A and h5-HT_3A subunits (Hope et al., 1993; Belelli et al., 1995), the apparent affinity of 5-HT at either chimera was unaltered (Table 1). In contrast, the concentration of (+)-Tc required to reduce the control response to 5-HT by 50% (i.e., IC₅₀) at chimera m233h 5-HT_3A was reduced by over 2200-fold relative to the wild-type h5-HT_3A subunit (Table 1), such that the concentration-inhibition curve shifted leftward to superimpose upon that obtained for (+)-Tc acting at the wild-type m5-HT_3A subunit (Fig. 1). The converse result was obtained for the h218m 5-HT_3A subunit chimera, where the IC₅₀ for (+)-Tc was increased by over 1300-fold in comparison to the m5-HT_3A subunit to yield an inhibition curve that approximated closely to that found for (+)-Tc at the wild-type h5-HT_3A subunit. These results indicate that the structural determinants of the discriminatory potency of (+)-Tc are located entirely within the N-terminal domain.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Antagonist potency of (+)-Tc at wild-type human and mouse 5-HT3A receptor subunits in comparison to m223h 5-HT3A and h218m 5-HT3A subunit chimeras. A, traces illustrating potency of (+)-Tc in inhibiting currents evoked by equieffective (i.e., EC50) concentrations of 5-HT at wild-type human and mouse 5-HT3A receptor subunits and m223h 5-HT3A and h218m 5-HT3A subunit chimeras. (+)-Tc (1 nM) inhibits ~50% of current evoked by 3 µM 5-HT both at wild-type m5HT3A receptor subunit and m223h 5-HT3A subunit chimera. In contrast, 3 µM (+)-Tc is required to produce ~50% inhibition of 5-HT (3 µM)-evoked current both at h5HT3A receptor subunit and h218m 5-HT3A subunit chimera. B, graph illustrates relationship between concentration of (+)-Tc in medium (abscissa, log scale) and peak amplitude of inward current response evoked by 5-HT at EC50 as a percentage of control response to 5-HT (ordinate, linear scale) for wild-type m5-HT3A (), wild-type h5-HT3A (), m223 h 5-HT3A (), and h218m 5-HT3A () subunits. Data points were fitted with equation described in Experimental Procedures to yield the following parameters: m5-HT3A, IC50 = 1.4 nM, n = 0.9; h5-HT3A, IC50 = 2.55 µM, n = 1.00; m223h 5-HT3A, IC50 = 1.16 nM, n = 1.06; h218m 5-HT3A, IC50 = 1.87 µM, n = 1.28. Each point represents mean ± S. E. of experiments conducted upon a minimum of three oocytes. Data for m5-HT3A and h5-HT3A were taken from Hope et al. (1993) and Belelli et al. (1995), respectively.

Multiple Point Mutant m5-HT_3A and h5-HT_3A Subunits. Sequence alignment of 5-HT_3A subunit orthologs (Fig. 2) reveals a cluster of nonconserved residues contained within a domain homologous to ligand binding "loop" 3 (also termed "C") of certain nicotinic AChR (Galzi and Changeux, 1995), -aminobutyric acid type A (Smith and Olsen, 1995), and strychnine-sensitive glycine receptor subunits (Rajendra et al., 1997). In view of the fact that (+)-Tc is known to photoaffinity label aromatic residues within loop 3 of nicotinic AChR -subunits, which stabilize the binding of the antagonist (Sine et al., 1994; Chiara and Cohen, 1997), we focused upon the homologous region of m5-HT_3A and h5-HT_3A in subsequent mutagenesis experiments.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Sequence alignment of species homologs of 5-HT3A subunit and nicotinic acetylcholine receptor  subunit of mouse and Torpedo over domain corresponding to loop 3 of nicotinic receptor. Residues that are not conserved between human and mouse 5-HT3A subunit orthologs are highlighted in bold. Residues corresponding to Tyr- 190, Cys-192, Cys-193, and Tyr-198 of nicotinic 1 sequence are indicated (^).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Antagonist potency of (+)-Tc at the wild-type m5-HT3A receptor subunit in comparison with m5-HT3A subunits bearing triple, quintuple, or septuple amino acid mutations. Graph depicts relationship between concentration of (+)-Tc in medium (abscissa, log scale) and peak amplitude of inward current response evoked by 5-HT at EC50 as a percentage of control response to 5-HT (ordinate, linear scale) for wild-type m5-HT3A (), mutant 1 (i.e., m5-HT3A I205M/D206E/I207S; ), mutant 3 (i.e., m5-HT3A Q199Y/K201R/I205M/D206E/I207S; ) and mutant 5 (i.e., m5-HT3A Q199Y/K201R/I205M/D206E/I207S/S210Y/I219V; ) receptor subunits. Data points were fitted with equation described in Experimental Procedures to yield the following parameters: m5-HT3A, IC50 = 1.4 nM, n = 0.9; mutant 1, IC50 = 27.5 nM, n = 1.53; mutant 3, IC50 = 76 nM, n = 1.39; and mutant 5, IC50 = 228 nM, n = 1.15. Each point represents mean ± S.E. of experiments conducted on a minimum of three oocytes. Data for m5-HT3A were taken from Hope et al. (1993).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Antagonist potency of (+)-Tc at the wild-type h5-HT3A receptor subunit in comparison with h5-HT3A subunits bearing triple, quintuple, or septuple amino acid mutations. Graph depicts relationship between concentration of (+)-Tc in medium (abscissa, log scale) and peak amplitude of inward current response evoked by 5-HT at EC50 as a percentage of control response to 5-HT (ordinate, linear scale) for wild-type h5-HT3A (), mutant 2 (i.e., h5-HT3A M200I/E201D/S202I; ), mutant 4 (i.e., h5-HT3A Y194Q/R196K/M200I/E201D/S202I; ) and mutant 6 (i.e., m5-HT3A Y194Q/R196K/M200I/E201D/S202I/Y205S/V214I; ) receptor subunits. Data points were fitted with equation described in Experimental Procedures to yield the following parameters: h5-HT3A, IC50 = 2.55 µM, n = 1.0; mutant 2, IC50 = 486 nM, n = 1.53; mutant 4, IC50 = 99.5 nM, n = 1.17; and mutant 6, IC50 = 49.4 nM, n = 1.37. Each point represents mean ± S.E. of experiments conducted on a minimum of three oocytes. Data for h5-HT3A were taken from Belelli et al. (1995).

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Determinants of species-dependent antagonist potency of (+)-Tc at human (h) and mouse (m) 5-HT3A receptor subunits. Left, potency of (+)-Tc as an antagonist of inward current response to 5-HT at EC50 is expressed as log of ratio of the IC50 at chimeric and mutant receptors divided by IC50 at appropriate wild-type receptor. Vertical dashed lines indicate log IC50 ratios m5-HT3A/h5-HT3A and h5-HT3A/m5-HT3A. Right, schematic representation of chimeras h218m 5-HT3A and m223h 5-HT3A (mouse sequence shaded) and position of loop 3 domain. Loop 3 residues exchanged between h5-HT3A and m5-HT3A subunits are detailed in text, with sequence corresponding to m5-HT3A underlined. Bar diagram was constructed from data presented in Table 1.

Single Point Mutant m5-HT_3A Subunits. Table 2 summarizes the IC₅₀ values for (+)-Tc obtained when the seven nonidentical amino acids within the putative loop 3 of the m5-HT_3A subunit were individually mutated to the homologous residues of the h5-HT_3A species ortholog. The mutations m5-HT_3AK201R, I205 M, I207S, and I219V were associated with IC₅₀ values for (+)-Tc that were essentially indistinguishable from that found for the wild-type m5-HT_3A subunit. Very modest increases in IC₅₀ (~2.5-3.0-fold) were obtained for m5-HT_3AQ199Y and S210Y, whereas a substantial increment (~9-fold) was found for the D206E mutant.

	Discussion

Top Summary Introduction Experimental Procedures Results Discussion References

Domains contributing to the ligand binding sites of nicotinic AChRs were initially identified by affinity labeling using derivatives of agonist and antagonist compounds. Such studies provided a valuable framework for subsequent molecular biological approaches employing chimeric and point mutant receptor subunits (reviewed in Galzi and Changeux, 1995). In the absence of similar information, we exploited the species-dependent pharmacology of 5-HT_3A receptor subunits and sequence comparisons between subunits as an attempt to delineate components of the ligand binding site.

(+)-Tc is a structurally rigid antagonist that displays a large difference in potency across 5-HT_3A subunit orthologs (Peters et al., 1997). One potential advantage in evaluating this ligand is the existence of comparative information concerning specific amino acid residues that influence the binding of (+)-Tc to the interfaces formed between muscle nicotinic AChR subunits (Pedersen and Cohen, 1990; O'Leary et al., 1994; Sine, 1993; Bren and Sine, 1997; Chiara and Cohen, 1997). In the latter case, the binding of (+)-Tc involves specific domains provided both by the  and non- subunits, which have been termed the "principal" and "complementary" components of the site respectively (Bertrand and Changeux, 1995). The structure activity relationships for curariform antagonists at Torpedo and mouse nAChR / subunit interfaces and the m5-HT_3A subunit are broadly similar (Pedersen and Papineni, 1995; Papineni and Pedersen, 1997; Yan et al., 1998), suggesting that the binding of (+)-Tc at the 5-HT₃ receptor may also involve principal and complementary components provided, in this instance, by the opposite faces of adjacent, structurally identical, 5-HT_3A subunits. The results of the present study implicating loop 3 residues ("principal component") in the binding of (+)-Tc when combined with data demonstrating an influence of Trp-67 ("complementary component") upon antagonist binding (Yan et al., 1999) support such a scheme.

Studies on native and recombinant 5-HT₃ receptors indicate that (+)-Tc acts in a manner consistent with competitive antagonism (Higashi and Nishi, 1982; Maricq et al., 1991; Newberry et al., 1991; Hope et al., 1993; Yan et al., 1998). Indeed, as assessed by computational chemistry, there is considerable structural congruence between (+)-Tc, 5-HT, and several 5-HT₃ receptor-selective agonists (Aprison et al., 1996). However, high concentrations of (+)-Tc insurmountably antagonize electrical responses mediated by 5-HT₃ receptors endogenous to rabbit (Higashi and Nishi, 1982) and guinea pig (Newberry et al., 1991) neurons. By analogy to the muscle nAChR, such an action could potentially be due to open channel blockade by (+)-Tc (Colquhoun et al., 1979). Thus, it was important to confirm that the differential potency of (+)-Tc at mouse and human 5-HT_3A subunit orthologs is entirely due to differences in primary amino acid sequence within the extracellular N-terminal domain. That this is so is indicated by the results obtained with the chimeras m223h 5-HT_3A and h218m 5-HT_3A, where the reciprocal exchange of the N-terminal domain was shown to entirely account for the species dependent pharmacology of (+)-Tc.

We studied the domain homologous to loop 3 (or C) of nAChR  subunits for several reasons. First, sequence alignment of the four orthologs of the 5-HT_3A subunit currently isolated identify this region as containing a particularly high incidence of unconserved residues relative to the remainder of the extracellular N-terminal domain (Fig. 2). Secondy, a chimeric construct of the human and guinea pig 5-HT_3A subunit orthologs reveals this region to strongly contribute to a differential potency of 1-phenylbiguanide, a 5-HT₃ receptor-selective agonist, at the wild-type subunits. Third, a tyrosine residue (Tyr-198) within loop 3 of nAChR  subunits that is photoaffinity labeled by (+)-Tc (Chiara and Cohen, 1997) and that constitutes an important element of the principal binding component (Sine, 1993; O'Leary et al., 1994), is conserved in the 5-HT_3A subunit (Fig. 2 and see below).

The substitution of amino acid residues from m5-HT_3A into h5-HT_3A sequence and vice versa caused qualitatively opposite changes in the potency of (+)-Tc. By contrast, only very modest effects upon the agonist potency of 5-HT were observed, militating against a nonspecific effect of the mutations upon receptor structure or function. The progressively larger shift in the IC₅₀ of (+)-Tc associated with the triplet, quintuple, and septuple amino acid substitutions in loop 3 (Table 1) indicates that multiple nonconserved residues contribute to the differential potency of the antagonist at mouse and human 5-HT_3A receptor orthologs. Furthermore, it is clear that additional residues located elsewhere within the N-terminal domain must also be involved, because the exchange of all seven nonconserved loop 3 amino acids between the subunit orthologs is insufficient to completely convert the IC₅₀ for (+)-Tc to that of either the human (cf. mutant 5) or mouse (cf. mutant 6) 5-HT_3A subunit. Moreover, such unidentified residues appear to contribute unequally to the binding of (+)-Tc in the two subunit orthologs as evidenced by the consistently smaller impact upon the IC₅₀ of (+)-Tc when homologous residues from the mouse are grafted into the human subunit versus the converse exchange (compare mutants 1, 3, and 5 with mutants 2, 4, and 6). In this respect, domains of the 5-HT_3A subunit that are homologous to the complementary binding sites for (+)-Tc presented by vertebrate muscle nicotinic AChR / (high affinity) and  (low affinity) subunits are of interest, particularly because Trp-67, which is homologous to 7Trp-54, Trp-55, and Trp-57, participates in the binding of (+)-Tc (Spier et al., 1997; Yan et al., 1999). However, at loci corresponding to the critical residues identified in the nicotinic AChR subunits [i.e., fetal receptor: Ile-116/Val-118; Tyr-117/Thr-119 (Sine, 1993); Ser-161/Lys-163; adult receptor: Ile-58/His-60; and Asp-59/Ala-61 (Bren and Sine, 1997)] the primary sequence across 5-HT_3A subunit orthologs is invariant, or shows only a conservative substitution (i.e., Phe/Tyr) that, in any event, does not correlate with the apparent affinity of (+)-Tc.

The analysis of the contribution of individual residues suggests that, in all but one instance, interactions with (+)-Tc are likely to be indirect. Thus, the mutations m5-HT_3A Q199Y, K201R, I205M, I207S, S210Y, or I219V produce, at most, a 3-fold change in the blocking potency of (+)-Tc. Several of these exchanges conserve gross physicochemical properties of the residue such as positive charge (K201R) and aliphatic (I219V) character, whereas others are associated with the incorporation of aromatic groups and a concomitant increase in side chain volume (Q199Y and S210Y). It is noteworthy that there is little conservation of these residues across the four orthologs of 5-HT_3A subunit thus far identified (Fig. 2). In contrast, in the case of the solitary mutation producing a substantial increase in the IC₅₀ of (+)-Tc (i.e., m5-HT_3A D206E) Glu is the aligned residue in nonmouse 5-HT_3A subunits, all of which demonstrate reduced affinity toward (+)-Tc. Interestingly, m5-HT_3A Asp-206 aligns with Cys-193 of the nicotinic AChR -subunit which, along with Cys-192, Tyr-190, and Tyr198, contributes to the loop 3 component of the principal nicotinic binding site (Galzi and Changeux, 1995; Bertrand and Changeux, 1995).

There is considerable evidence that both Tyr-190 and Tyr-198 of the muscle nicotinic AChR -subunit act to stabilize the binding of the curariform antagonist dimethyl-d-tubocurarine, principally through quaterary ammonium-aromatic interactions (O'Leary et al., 1994; Sine et al., 1994). An homologous tyrosine residue common to human, mouse, and rat 5-HT_3A subunits, or phenylalanine in the guinea pig sequence (Fig. 2), may play a similar role. However, a serine residue conserved across all 5-HT_3A subunit orthologs aligns with Tyr-190 (Fig. 2). Mutagenesis of Tyr-190 to Ser in the nicotinic AChR subunits is associated with a pronounced reduction in the affinity of dimethyl-d-tubocurarine and the virtual abolition of / versus / interface selectivity (Sine et al., 1994). It is conceivable that this substitution contributes to the low affinity of (+)-Tc for the human 5-HT_3A receptor and that the multiple amino acid differences between human and mouse sequences within loop 3 compensate for the presence of serine in the mouse subunit, perhaps by allowing the adjacent phenylalanine residue to assume an orientation that permits interaction with (+)-Tc. Overall, the present data are compatible with a scheme wherein loop 3 residues collectively determine the shape of an element of the ligand pocket for (+)-Tc with m5-HT_3A Asp-206, perhaps making direct contact with the ligand. The additional methylene group in the side chain of Glu versus Asp may place the negatively charged carboxylic acid group in a less favorable orientation for interaction with (+)-Tc.

A precedent for a diffuse influence of loop 3 residues upon ligand binding derives from a recent study of the chimeric 7-5-HT_3A subunit (Corringer et al., 1998). In the latter, the exchange of five residues from an agonist binding domain of the 4- nicotinic subunit into the 7-5-HT_3A chimera selectively enhanced the apparent affinity of ACh by 30-fold, abolishing the difference in potency between the latter and nicotine to confer a pharmacological phenotype typical of the 42-nicotinic receptor. In common with the results of the present study, the mutation of individual residues revealed only one that exerted a substantial (i.e., 7-fold) influence upon the apparent affinity of ACh. Interestingly, several of the residues exerting a subtle effect upon the apparent affinity of ACh are homologous to amino acids within mouse and human 5-HT_3A subunits that contribute to the differential potency of (+)-Tc.