Dai G et al. (2021), Electromechanical coupling mechanism for activa...

ECB-ART-49180

Nat Commun 2021 May 14;121:2802. doi: 10.1038/s41467-021-23062-7.

Show Gene links Show Anatomy links

Electromechanical coupling mechanism for activation and inactivation of an HCN channel.

Dai G , Aman TK , DiMaio F , Zagotta WN .

???displayArticle.abstract???
Pacemaker hyperpolarization-activated cyclic nucleotide-gated (HCN) ion channels exhibit a reversed voltage-dependent gating, activating by membrane hyperpolarization instead of depolarization. Sea urchin HCN (spHCN) channels also undergo inactivation with hyperpolarization which occurs only in the absence of cyclic nucleotide. Here we applied transition metal ion FRET, patch-clamp fluorometry and Rosetta modeling to measure differences in the structural rearrangements between activation and inactivation of spHCN channels. We found that removing cAMP produced a largely rigid-body rotation of the C-linker relative to the transmembrane domain, bringing the A' helix of the C-linker in close proximity to the voltage-sensing S4 helix. In addition, rotation of the C-linker was elicited by hyperpolarization in the absence but not the presence of cAMP. These results suggest that - in contrast to electromechanical coupling for channel activation - the A' helix serves to couple the S4-helix movement for channel inactivation, which is likely a conserved mechanism for CNBD-family channels.

???displayArticle.pubmedLink??? 33990563
???displayArticle.pmcLink??? PMC8121817
???displayArticle.link??? Nat Commun
???displayArticle.grants??? [+]

???attribute.lit??? ???displayArticles.show???

	Fig. 1. Structural features and properties of spHCN channels in the presence and absence of cyclic nucleotide.a Side view of a homology structure of spHCN channels based on the cryo-EM structure of human HCN1 (PDB 5U6P), highlighting the proximity of the S4 and S5 transmembrane helices relative to the C-linker domain of the adjacent subunit. Only two adjacent subunits are shown. b Top and bottom views of the same homology model of spHCN showing the arrangement of its quaternary structure. c Left: spHCN channel currents elicited by a hyperpolarizing voltage pulse from 0 to −100 mV, in the presence (red) or in the absence (black) of 1 mM cAMP. Middle: Channel inactivation without cyclic nucleotide, elicited by a series of hyperpolarizing voltages from 0 to −120 mV (with 10 mV steps), happened during the time frame indicated by the dashed box on the left panel. Note the blown up current and time scale. Right: Channel currents elicited by a series of hyperpolarizing voltages from 0 to −120 mV (with 10 mV steps), in the presence of 1 mM cAMP. d Left, cartoon showing the FRET between S346Anap in the S4 helix and Co2+ chelated by a di-histidine L182H and L186H in the HCN domain. e Fluorescence measurements for spHCN-S346Anap, L182H, and L186H channels with a −100 mV voltage pulse, before and after applying 1 mM Co2+. f Summary of the FRET efficiencies measured for spHCN-S346Anap, L182H, and L186H channels at 0 and −100 mV in the absence of cyclic nucleotide monophosphate (no cNMP; n = 4 independent patches), and the presence of cGMP (n=3 independent patches) or cAMP (n=3 independent patches). Data shown are mean ± s.e.m. Source data are provided in the Source Data file.
	Fig. 2. tmFRET between the C-terminal end of the S4 helix and A′ or B′ helix of the C-linker.a Cartoon showing tmFRET between W355Anap in the S4 helix and Cu2+-TETAC attached to L481C in the A′ helix. b Cartoon showing tmFRET between W355Anap and Cu2+-TETAC attached to E506C in the B′ helix. c Summary of the fraction of Anap fluorescence quenched by Cu2+-TETAC for the spHCN-W355Anap channels in panels a and b, without (left) and with (middle, right) the introduced cysteines L481C or E506C in four conditions: the presence and absence of cAMP, and at 0 and −100 mV; n = 3–5 independent patches. Data shown are mean ± s.e.m.; Source data in the forms of FCys, and Fno Cys are provided in the Source Data file. d FRET efficiencies of spHCN-W355Anap, L481C channels measured in four conditions: the presence and absence of cAMP, and at 0 and −100 mV; n = 3 independent patches for cAMP conditions and n = 5 independent patches for no cNMP conditions. p < 0.001 for the statistical comparisons. e FRET efficiencies of spHCN-W355Anap, E506C channels; n = 3 patches for cAMP and n = 4 patches for no cNMP conditions. p < 0.001 for the statistical comparisons. Data shown are mean ± s.e.m.; source data are provided in the Source Data file.
	Fig. 3. tmFRET between the N-terminal end of the S5 helix and A′ helix of the C-linker.a Cartoon showing tmFRET between C369Anap in the S5 helix and Cu2+-TETAC attached to Y477C in the A′ helix. b Cartoon showing tmFRET between W355Anap in the S4 helix and Cu2+-TETAC attached to Y477C in the A′ helix. c FRET efficiencies of spHCN-C369Anap, Y477C channels measured in four conditions: the presence and absence of cAMP, and at 0 and −100 mV; n = 7 independent patches for 0 mV conditions and n = 4 independent patches for −100 mV conditions. p < 0.001; p = 0.1 between no cNMP, 0 mV and no cNMP, −100 mV conditions. d FRET efficiencies of spHCN-W355Anap, Y477C channels measured in four conditions: the presence and absence of cAMP, and at 0 and −100 mV; n = 3 independent patches for cAMP and n = 4 independent patches for no cNMP. p < 0.001 and *p < 0.05. Data shown are mean ± s.e.m.; source data are provided in the Source Data file.
	Fig. 4. Distance determination using measured FRET efficiencies, guided by FCG and homologous structures.a Eight FRET pairs used to obtain the Förster Convolved with Gaussian (FCG) relation to determine the distance dependence of the FRET efficiency. The Cβ–Cβ distances were measured from the homology models based on human HCN1 (PDB:5U6P and PDB: 6UQF) assuming they correspond to the structures at 0 and −100 mV respectively (in the presence of cAMP). b The measured FRET efficiencies are plotted versus the distances measure in panel a (in R0 units) as 4 black (0 mV) and 4 purple (−100 mV) circles. Predicted distance dependencies of the Förster equation (black) and the FCG relation (green) are shown. The FCG relation used a Gaussian distribution with a standard deviation, σ, of 5 Å (inset).
	Fig. 5. Comparison of the measured distances from tmFRET and the corresponding distances from Rosetta models.a–d Comparison of the distances of FRET pairs measured from tmFRET (circle, mean ± s.e.m, n = 3–7) using the FCG relation in Fig. 4b and mean Cβ–Cβ distances of these FRET pairs in the 17 top scoring Rosetta models (square) based on the experimentally-determined distance constraints in four states: resting-cAMP (a), activated (b), resting-apo (c), and inactivated (d). The standard deviation of the Cβ–Cβ distances from the top 17 scoring Rosetta models for each state was less than 0.3 Å.
	Fig. 6. Rosetta models for activation and inactivation of spHCN channels and their regulation by cAMP.a–d Modeled structures (side view and bottom view) highlighting the rearrangement of the C-linker region of one subunit relative to the S4 and S5 helices of the adjacent subunit for four states: resting-cAMP (a), activated (b), resting-apo (c), and inactivated (d). The S4 helix moves downward and bends into two segments after hyperpolarization in both activated and inactivated states. The bottom view highlights the rigid-body rotational movements of the C-linker around the central axis of the pore in the absence of cyclic nucleotide. See also Supplementary Movies 1–4.

References [+] :

Aldrich, Fifty years of inactivation. 2001, Pubmed