Figure 2. State-dependent accessibility of residues in S2 and S4 segments of HCN channel.

(A) Accessibility of the S2 charge transfer center (Phe residue) in representative molecular models of the activated and resting states Kv1.2/2.1 VSDs and the resting and activated states HCN1 VSDs extracted from molecular simulations. The charge transfer center is only accessible from the intracellular medium in the activated (Down) state of HCN1. A cutaway through the VSD is represented as a slightly transparent light brown surface; the charge transfer center (Phe residue) is represented as red spheres. (B) Predicted accessibility to internally applied polar cysteine modifying reagents. Accessibility to S4 cysteines will increase in the Down state compared to the Up state for both models as has been shown previously. The S2 cysteine at or proximal to the charge transfer center is expected to show very little modification in both Up and Down states in the canonical model. However, if the helix breaks and opens a large cavity in the Down state, the cysteine at the charge transfer center will become accessible in a state-dependent fashion. (C) Solvent accessibility estimates of S4 (left) and S2 (right) residues for both up and down state models of HCN1. The accessibility of each residue is estimated for the six voltage sensors that underwent activation and shown as one symbol each (closed blue circles for the resting Up state, open orange circles for the activated Down state). (D) State-dependent accessibility of the charge transfer center of hyperpolarization-activated ion channel. Wild type spHCN channel does not react with MTSET under hyperpolarized or depolarized condition (top two panels). Substitution of the charge transfer center, F186C, on the other hand, shows state-dependent reactivity with MTSET (bottom two panels). The rate of reactivity in the activated Down state is 440 ± 30 M⁻¹s⁻¹ versus 7.5 ± 0.3 in the resting Up state.

Figure 2—figure supplement 1. Activated state model explains state-dependent accessibility of S4 residues.

(A) Comparison of solvent accessibility data reported by Bell et al. (2004); Vemana et al. (2004) (left plot) and calculated from the MD simulations (right plot). Solvent accessibility is estimated experimentally as a rate of modification by MTSET and in the simulations as the surface area accessible to MTSET (SASA). For the simulations, the data was collected from the six voltage sensors that underwent activation; each data point corresponds to the average SASA estimated for one voltage sensor. The values for the resting (Up) state are shown as filled cyan circles, and those for the activated (Down) state as open red circles. (B) Models of the voltage sensor domains in the resting (Up) and activated (Down) conformations with residues in panel A highlighted. Residues accessible in both states are colored in cyan, and residues accessible mostly in the activated state are colored in red. (C) A cartoon highlighting the key features of the two possible models for hyperpolarization-dependent gating. In the canonical helical screw motion, the S4 translates across the membrane in a helical screw motion, past the charge transfer center within the gating scaffold. Our MD simulations of HCN1 channel show that the S4 helix moves down and breaks into two parts with the lower helix becoming almost parallel to the membrane.