Potential state-selective hydrogen bond formation can modulate activation and desensitization of the α7 nicotinic acetylcholine receptor

Abstract

A series of arylidene anabaseines were synthesized to probe the functional impact of hydrogen bonding on human α7 nicotinic acetylcholine receptor (nAChR) activation and desensitization. The aryl groups were either hydrogen bond acceptors (furans), donors (pyrroles), or neither (thiophenes). These compounds were tested against a series of point mutants of the ligand-binding domain residue Gln-57, a residue hypothesized to be proximate to the aryl group of the bound agonist and a putative hydrogen bonding partner. Q57K, Q57D, Q57E, and Q57L were chosen to remove the dual hydrogen bonding donor/acceptor ability of Gln-57 and replace it with hydrogen bond donating, hydrogen bond accepting, or nonhydrogen bonding ability. Activation of the receptor was compromised with hydrogen bonding mismatches, for example, pairing a pyrrole with Q57K or Q57L, or a furan anabaseine with Q57D or Q57E. Ligand co-applications with the positive allosteric modulator PNU-120596 produced significantly enhanced currents whose degree of enhancement was greater for 2-furans or -pyrroles than for their 3-substituted isomers, whereas the nonhydrogen bonding thiophenes failed to show this correlation. Interestingly, the PNU-120596 agonist co-application data revealed that for wild-type α7 nAChR, the 3-furan desensitized state was relatively stabilized compared with that of 2-furan, a reversal of the relationship observed with respect to the barrier for entry into the desensitized state. These data highlight the importance of hydrogen bonding on the receptor-ligand state, and suggest that it may be possible to fine-tune features of agonists that mediate state selection in the nAChR.

FIGURE 1.

The α7 nAChR ligand binding domain model and structures of partial agonists. A, 4OH-GTS-21 in the α7 nAChR ligand binding domain of a homology model. The key elements of the α7 receptor were modeled using the 2WN9 (PDB ID) template. Residues within 5 Å of the phenyl ring of 4OH-GTS-21 are displayed except Trp-55, which is located behind the ligand and is hidden to clarify this presentation. B, the structures of 4OHGTS-21 and the six arylidene anabaseines used in this study are displayed and annotated with their corresponding hydrogen bonding properties.

FIGURE 2.

Functional tests of wild-type α7 Gln-57 mutants. A, comparison of the wild-type and Gln-57 mutant expression levels as estimated by the net charge response to ACh. All of the receptor types were tested 2 days after injection of RNA. The net charge responses of the 300 μm acetylcholine are normalized to that of the wild-type. **, p < 0.01; n ≥ 4. B, concentration-response curves of acetylcholine presented in net charge. C, representative traces of the acetylcholine activation on WT and Q57K mutant receptors at different concentrations. D and E, comparison of the PNU-120596-stimulated response when applied with ACh. Each oocyte received two initial 300 μm ACh controls followed by co-application of 300 μm acetylcholine and 300 μm PNU-120596. Both the peak response and the net charge response of the co-application were normalized to the average of the two initial controls. The peak responses of the co-application were displayed in panel D, and the ratio of the net charge response to the peak response of the co-application was displayed in panel E. F, traces for co-application of ACh and PNU-120596 to WT, Q57K, and Q57D receptors, with the accompanying ACh-only reference traces.

FIGURE 3.

Concentration-response curves of the six arylidene anabaseines with WT and Gln-57 mutant α7 receptors. 3FAB-evoked responses are not shown because most responses were marginally detectable at 300 μm concentration. All responses are the averages (±S.E.) of data from at least 4 oocytes.

FIGURE 4.

Representative traces of the 300 μm arylidene anabaseine-evoked currents and their inhibition of the subsequent ACh-evoked responses of the wild-type α7 nAChR and Gln-57 mutants. Each oocyte received two 300 μm ACh controls (only the second one is displayed), a 300 μm arylidene anabaseine treatment, followed by another 300 μm ACh control. All of the traces are scaled to the pre-drug ACh control responses, the absolute peak values of which varied from 1.9 to 11 μA.

FIGURE 5.

A summary of the residual desensitization profile of 2FAB with WT and Gln-57 mutant α7 receptors probed by PNU-120596. A, representative traces of 300 μm PNU-120596-evoked responses following the application 300 μm 2FAB to induce RD. Also shown are 300 μm ACh control responses before the sequential applications of 2FAB and PNU-120596. All responses are scaled relative to the initial ACh control responses. B, scatter plot of the PNU-120596-evoked response in comparison with the inhibitory effect of 300 μm 2FAB on a subsequent ACh application (as shown in Fig. 4). The values of the inhibition effect are obtained from the recovery values displayed in Table 2. C, recovery tests of different α7 receptor types after 2FAB application. Cells were washed with Ringers solution after application of 2FAB, and repeatedly stimulated with 300 μm ACh to evaluate the rates of recovery. All responses plotted in panels B and C are the averages (±S.E.) of at least 4 oocytes.

FIGURE 6.

Comparison of arylidene anabaseine-evoked current and currents evoked with the co-application of PNU-120596. A, representative traces for the application of 300 μm 3FAB (upper two traces) or 300 μm 3PyroAB (lower two traces) applied alone or co-applied with 300 μm PNU-120596. Each oocyte received two 300 μm ACh controls (only the first one was displayed in the representative traces), then was treated with one of the arylidene anabaseines at 300 μm, either with or without 300 μm PNU-120596. Each single trace displayed represents a single experiment, but all traces are normalized to its own ACh control, the absolute peak value of which varied from 0.6 to 10.2 μA. These control responses were scaled to the same level and omitted for clear presentation. B, the peak responses for co-applications of agonist and PNU-120596 relative to the average of the two initial ACh controls. C and D, the peak response (C) and net charge-to-peak response ratio (D) of the co-application responses relative to the average of the two initial controls. The ratios were calculated for each single cell. The mean ± S.E. were based on data obtained from at least four oocytes.

FIGURE 7.

Energy landscapes for ACh, FAB, and PyroAB agonists with different α7 receptors at high ligand occupancy level. Under equilibrium conditions, the distribution of receptors into the resting closed (C), open (O), PNU-120596-sensitive desensitized (D_s), and PNU-120596-insensitive desensitized (D_i) states are determined by the relative free energy of the states (represented by vertical displacements). The graphs were constructed as described in the text. The experimental support for transitions shown in bold are discussed in the text.