Molecular recognition of the neurotransmitter acetylcholine by an acetylcholine binding protein reveals determinants of binding to nicotinic acetylcholine receptors

Abstract

Despite extensive studies on nicotinic acetylcholine receptors (nAChRs) and homologues, details of acetylcholine binding are not completely resolved. Here, we report the crystal structure of acetylcholine bound to the receptor homologue acetylcholine binding protein from Lymnaea stagnalis. This is the first structure of acetylcholine in a binding pocket containing all five aromatic residues conserved in all mammalian nAChRs. The ligand-protein interactions are characterized by contacts to the aromatic box formed primarily by residues on the principal side of the intersubunit binding interface (residues Tyr89, Trp143 and Tyr185). Besides these interactions on the principal side, we observe a cation-π interaction between acetylcholine and Trp53 on the complementary side and a water-mediated hydrogen bond from acetylcholine to backbone atoms of Leu102 and Met114, both of importance for anchoring acetylcholine to the complementary side. To further study the role of Trp53, we mutated the corresponding tryptophan in the two different acetylcholine-binding interfaces of the widespread α4β2 nAChR, i.e. the interfaces α4(+)β2(-) and α4(+)α4(-). Mutation to alanine (W82A on the β2 subunit or W88A on the α4 subunit) significantly altered the response to acetylcholine measured by oocyte voltage-clamp electrophysiology in both interfaces. This shows that the conserved tryptophan residue is important for the effects of ACh at α4β2 nAChRs, as also indicated by the crystal structure. The results add important details to the understanding of how this neurotransmitter exerts its action and improves the foundation for rational drug design targeting these receptors.

Figure 1. Structure of ACh bound to Ls-AChBP.

(A) The structure of acetylcholine (ACh). (B) Displacement of tritium-labeled epibatidine (³H-Epi) bound to Ls-AChBP by ACh was used to determine the IC₅₀ value of ACh. The data points shown are from one determination of the IC₅₀ value. The average of three such experiments were converted to the K_i value of ACh by the Cheng-Prusoff equation. (C) Top-view of a cartoon representation of the structure of one Ls-AChBP pentamer with an ACh molecule bound in each interface. (D) Side-view of a cartoon representation of the Ls-AChBP with ACh shown in green stick representation. The ACh molecule is located between two colored subunits: the green subunit forms the principal side of the binding pocket, (+) interface, while the orange subunit forms the complementary side, (−) interface.

Figure 2. Close-up view on the ACh-binding site and comparison to other structures.

Structures are shown with ligands as green sticks and side chains within 5 Å of ligands as lines. (A) ACh bound to Ls-AChBP, as reported here, with principal side-chain carbon atoms in green and complementary side-chain carbon atoms in orange. The mesh around ACh corresponds to a 2mFo-DFc omit map calculated in PHENIX, shown at a level of 1ó and carved at 2.2 Å around the ligand. (B) ACh bound to a MMTS-modified Y53C mutant of Ac-AChBP , with principal side residues in cyan and complementary side residues in magenta. (C) Carbamoylcholine bound to Ls-AChBP with coloring as in (A). (D) Nicotine bound to Ls-AChBP with coloring as in (A). (E) ACh bound to ELIC , with principal side residues in blue and complementary side residues in salmon.

Figure 3. Two conformations of Trp53, Leu112 and Met114.

(A) On the complementary side of the interface, three residues near ACh adopt two distinct sets of conformations (shown in purple and light-blue sticks, respectively), at some interfaces occurring separately and at other interfaces with both conformations occurring as shown here. (B) and (C) Trp53 is shown in stick representation in the two different orientations observed, in interfaces where distinct orientations are seen. Principle side carbon atoms are colored green, while complementary side carbon atoms are orange. A mesh is shown in each case, corresponding to a partial omit map shown at 1ó and carved at 2 Å around Trp53. The partial omit map was generated using PHENIX by refining the structure after changing all Trp53 residues to alanine, thus alleviating side-chain orientation bias for this residue. (B) In one possible orientation, the Trp53 side-chain nitrogen atom is pointing “away” from Met114 with Trp53 and Trp143 aligned for T-type ππstacking. (C) In the other conformation, which is favored when a PEG400 molecule is present nearby, the Trp53 side-chain nitrogen atom is pointing towards Met114 and can form a hydrogen bond to the backbone carbonyl oxygen atom of this residue.

Figure 4. Concentration-response relationships at mutated nAChRs measured by two-electrode voltage clamp on X. laevis oocytes.

(A) α4β2 nAChRs with 3α:2β stoichiometry have three binding sites for ACh (black arrows): two with high sensitivity (HS) and one with low sensitivity (LS). Point-mutation of a central tryptophan residue in the α4 subunit will change the complementary (−) side of the LS site (red circle and arrow). Point-mutation of the corresponding tryptophan in the β2 subunit will change the complementary side of both HS binding sites (blue circles and arrows). (B) Concentration-response relationships (CRRs) of ACh at α4^(W88A)β2 and α4β2^(W82A) receptors. The black curve is drawn from previously published data for the ACh CRR at type α4β2 nAChRs with 3α:2β stoichiometry , with EC₅₀ and fraction values listed below the figure. ‘Fraction’ describes the fraction of the maximum response that is elicited by the high-sensitivity phase. Numbers in parenthesis refer to 95% confidence intervals. ‘n’ is the range of the number of measurements that were made of each point on a curve. An F test was carried out in GraphPad Prism 4 against the null hypothesis of a monophasic fit, which was rejected for α4^(W88A)β2 (F = 120, DFnN = 2, DFnD = 288) and accepted for α4β2^(W82A) (F = 0.73, DFnN = 2, DFnD = 72), where DFnN and DFnD are the degrees of freedom of the numerator and denominator in the F test, respectively.