Figure 1.
Modeling of the full-length cone channel in its resting state, on the basis of two separate templates.
(A) Side view of the TM region of the bacterial channel MlotiK1 in a closed state, PDB entry 3BEH [10]. This structure was the template for modeling the TM region of the cone channel. (B) Side view of the mouse HCN2 CNBDs in a resting state, PDB entry 1Q3E [15]. This structure served as a template for the cytosolic domain of the cone channel. (A, B) The regions of conflict between the templates are in red. (C) Side view of the resultant model structure of the human cone channel, shown in cartoon representation. CNGA3 subunits are colored cyan (light and dark); CNGB3 subunits are colored orange (light and dark).
Figure 2.
The evolutionary conservation profile supports the model structure.
The channel is colored by conservation grades according to the color-coding bar, with variable-through-conserved corresponding to turquoise-through-maroon. Overall, the variable residues are peripheral, whereas the conserved residues are in the structural core and channel pore. (A) Side view of the model structure in cartoon representation; the CNBDs are in their holo-state. (B) Extracellular view of the TM domain of the model structure. The proposed locations of the two CNGA3 and two CNGB3 subunits are marked [7].
Figure 3.
Evolutionary coupling (EC) calculations support the model.
Contact maps of top-ranked predicted ECs (red) overlaid on monomer (light grey) and intermonomer (dark grey) contacts from (A) the model structure of the TM domain of CNGA3; (B) the model structure of the cytosolic domain of CNGA3. The insets show the contact maps of alternative CNGA3 model structures; the orange arrows point to evolutionary couplings between amino acid pairs that are not in contact in the alternative models, but are in contact in our final model. Clearly, the overlay of the ECs with the contact map is far better for the chosen model structure than for the alternatives.
Figure 4.
The cross-correlations between the residues in each motion type are shown on the left panels (A, C, E); the channel is colored according to the correlation of the VSD of chain A with the other residues. The corresponding motions are presented on the right panels (B, D, F). Snapshots are colored according to the direction of motion, ranging from green to red. (A, C, E) In each panel, chain A represents an arbitrary CNGA3 chain (the channel comprises four identical chains). The magnitude of positive and negative correlations between the fluctuations of the residues is color-coded according to the blue-to-red scale at the bottom of the picture. Positive correlation indicates motion of two residues in the same direction, while negative correlation indicates motion in opposite directions. In panel A, the arrows indicate the location of the pivot points of the rotational motion at the termini of the S6 helices, i.e., the border between the dynamics units (See Discussion). The CNBDs in panel D and the TM domain in panel F are omitted for clarity.
Figure 5.
Motion I of CNGA3 appears to describe channel gating.
For clarity, only helices S5 and S6 (or corresponding KirBac helices) of two juxtaposed subunits are shown in each panel. (A, B) The similarity between the predicted conformations in panel B and the crystal structures in panel A is apparent, verifying the relation between these conformations and channel gating. (A) Side view of the KirBac3.1 crystal structures in open (pale green, PDB entry 3ZRS [43]) and closed (pale red, PDB entry 2WLJ [44]) states. The α-carbons of the two gate-residues, namely G120 and Y132, are shown as a space-filling model. (B–D) The edge conformations of KirBac3.1 (B), CNGA3 (C) and CNGA3 lacking the VSDs (D), as predicted by the elastic network models in the slowest mode of motion. The two edge conformations are shown in red and green. (C) The CNGA3 conformations resemble the conformations predicted for the KirBac channel (panel B), but the pore region is rigid. (D) The CNGA3 without VSD conformations are identical to the KirBac conformations (panel B).