An Allosteric Model for Electromechanical Coupling in Cardiac CNBD Channels

Abstract

Ion channels in the cyclic nucleotide-binding domain (CNBD) family, including hyperpolarization-activated cyclic nucleotide-gated (HCN) channels and human ether-à-go-go-related gene (hERG) channels, play pivotal roles in regulating cardiac action potentials. HCN channels are uniquely activated by hyperpolarization, rather than depolarization, a critical mechanism for controlling the involuntary pacemaker activity of the heart. In contrast, hERG channels are depolarization-activated and mediate K⁺ currents essential for action potential repolarization. Notably, certain hERG mutations, including those associated with long-QT syndrome, can induce biphasic activation by both hyperpolarization and depolarization. Despite the diverse voltage-dependent gating behaviors observed in CNBD channels, a unified mechanistic framework remains lacking. Here, we propose an allosteric model for their electromechanical coupling, featuring a single voltage-sensor transition coupled to two distinct conformational coupling modes between voltage-sensing and pore domains. With only three or four free parameters, this model recapitulates the biphasic U-shaped and bell-shaped conductance-voltage relationships commonly seen in CNBD channels. Fluorescence anisotropy-based homo-FRET experiments employing site-specifically incorporated noncanonical amino acids provide further support for the hypothesis, suggesting that the S5 helix movement plays a key role in hyperpolarization-dependent activation, while S4-S6 helix interactions are required for depolarization-dependent gating.

Figure 1.. An allosteric model for bidirectional voltage-dependent gating in CNBD channels.

(A) Schematic of the dual-coupling modular framework, where the voltage-sensing domain (VSD) bidirectionally regulates the pore domain (PD) via facilitation (D, positive coupling) or inhibition (H, negative coupling). (B) Cubic 8-state gating scheme integrating baseline pore opening, single-coupling (D or H), and dual-coupling (D and H) contributions. The model includes four closed states (C₀₀, C_D0, C_DH, C_0H), representing the channel’s closed, pre-open conformations under different coupling conditions: no coupling C₀₀, D-coupling alone (C_D0), both D- and H-coupling (C_DH), or H-coupling alone (C_0H). The corresponding open states (O₀₀, O_D0, O_DH, O_0H) mirror these configurations, reflecting pore opening under equivalent coupling conditions. The model assumes four identical VSDs acting cooperatively. (C-D) Simulated conductance-voltage (G-V) relationships for HCN1, HCN4, and hERG channels, fitted using the dual-coupling model (parameters in Table S1). Data points were sampled every 1 mV. HCN1 and HCN4 exhibit hyperpolarization-driven opening, while hERG shows hyperpolarization-driven closure. (E-H) Simulated biphasic G-V curves: bell-shaped (E, e.g., spHCN mutants) and U-shaped (F, e.g., hERG-D540K), using the same allosteric model. The model captures diverse gating polarities, including hyperpolarization-shifted (U-shape II) in G, and depolarization-shifted (U-shape III) profiles in H.

Figure 2.. Homo-FRET suggests S5 helix movement for channel opening, unique in HCN channels.

(A) The amber stop-codon suppression strategy for incorporating L-Anap. (B-C) Schematic of hHCN4 and hERG channels showing L-Anap labeling sites in S5 (V419/F551) and C-linker (Q536/L678). The structures shown are single subunits from either AlphaFold generated models of hHCN4 or hERG. (D-E) Fluorescence anisotropy traces during hyperpolarization (NMDG) and depolarization (KCl). In D, hHCN4-V419Anap (S5) shows increased anisotropy (reduced homo-FRET) during hyperpolarization, indicating S5 movement, while Q536Anap (C-linker) remains unchanged. In E, hERG-F551Anap (S5) exhibits no anisotropy shift, whereas L678Anap (C-linker) shows modest changes. Data suggest S5 movement drives HCN activation but is dispensable for hERG gating, supporting distinct coupling mechanisms. Error bars represent SEM, n = 4 for all conditions, two-sided t-test was used.

Figure 3.. Cartoons highlighting the two electromechanical coupling modes for CNBD channels.

(A) In HCN channels, downward movement and tilting of the S4 voltage sensor displaces the S5 helix, which then shifts the S6 helix to open the pore. S4 does not directly interact with S6. (B) In the S4-down conformation of hERG channels, S4 directly pushes against the S6 helix, forcing pore closure. S4 upward movement opens the pore. The cartoons were based on solved cryo-EM structures of HCN and hERG.