Dynamic rearrangement of the intrinsic ligand regulates KCNH potassium channels

Figures

• 

  Figure 1.. Mutating and deleting the intrinsic ligand altered the gating of zELK channels. (A) Ribbon diagram showing the interaction between the eag domain (consisting of the PAS domain and PAS cap) and CNBHD for KCNH channels, based on the cryo-EM structure of rEAG1 channel (Protein Data Bank accession no. 5K7L). Residues of the intrinsic ligand are highlighted in yellow. The position of cAMP in the corresponding structure of mHCN2-CNBD (Protein Data Bank accession no. 1Q5O) is shown in purple. (B) Representative I-V current traces of zELK wild-type channels in the cell-attached configuration using the voltage protocol on the left. (C) I-V current traces of zELK wild-type channels in the inside-out configuration (15 min after excision, same patch as B). (D) Normalized G-V curves for zELK wild-type and Δ740–742 channels in the cell-attached configuration. (E) Normalized G-V curves for zELK wild-type and Y740A channels in the inside-out configuration. The dashed line is the G-V curve for zELK wild-type in the cell-attached configuration. (F) Summary of the V1/2 obtained from experiments related to D. (G) Summary of the V1/2 obtained from experiments related to E. Data are shown as mean ±…

• 

  Figure 2.. VDP was attenuated by mutating the intrinsic ligand. (A) Representative I-V current traces for zELK wild-type and Δ740–742 channel activation in the cell-attached configuration using the voltage protocol on the left. The red trace is the double-exponential fit to the current elicited by the +120-mV voltage pulse (τ1 = 3 ms and τ2 = 130 ms). (B) Normalized G-V curve of zELK wild-type and Δ740–742 channel activation with and without a 500-ms, +60-mV prepulse. The purple dashed curves indicate the G-V curves of zELK wild-type and Δ740–742 channels after run-up caused by patch excision, highlighting the similar degree of hyperpolarizing shift compared with that caused by the +60-mV prepulse. (C) Current traces of zELK wild-type and Δ740–742 channels elicited by a voltage protocol with increasing durations of a +60-mV pulse (left). For the wild-type channels, τ of the tail current (red traces) increased from 8 to 17 ms. For zELKΔ740–742 channels, τ of the tail current (red traces) changed from 10 to 13 ms. (D) Summary of the changes in V1/2 for different zELK channels with and without the +60-mV prepulse using the protocol in B. n.s., not statistically significant. (E) Summary of the…

• 

  Figure 3.. Intrinsic-ligand peptide bound to the CNBHD of zELK channels and modulated the channel gating. (A and B) G-V curves of zELKΔ740–742 channels before and after applying the intrinsic ligand peptide in the inside-out patch configuration in the presence (A) or absence (B) of MgATP. The dashed lines are G-V curves for zELK wild-type in the presence (A) or absence (B) of MgATP as references. (C) Coomassie-stained gel showing purified C-linker/CNBHD and C-linker/CNBHDΔ740–742 with MBP fused to the amino-terminal end. (D) Fluorescence anisotropy of the mEAG1 and the hERG1 intrinsic-ligand peptides with an increasing concentration of wild-type or Δ740–742 CNBHD protein (mean ± SEM, n = 4–5).

• 

  Figure 4.. VCF revealed voltage-dependent changes of Anap fluorescence during VDP of zELK channels. (A) Ribbon diagram highlighting the residues within or flanking the intrinsic ligand of zELK that were mutated to L-Anap for VCF experiments. (B–E) Simultaneous Anap fluorescence measurements and two-electrode voltage-clamp recordings showing the kinetic change of Anap fluorescence during the depolarizing pulses for zELK-L738Anap (B), zELK-E744Anap (C), zELK-N741Anap (D), and zELK-N741Anap, ΔPAS cap (E) channels. The fluorescence traces are shown as the mean of 6–13 repetitive measurements using the same oocyte. (F) Summary of the percent increase in Anap fluorescence induced by depolarization for constructs in B–E (increase in fluorescence at the end of the +60-mV pulse versus that at −100 mV); mean ± SEM (n = 3–5). *, P < 0.05.

• 

  Figure 5.. PCF/tmFRET revealed the movement of the intrinsic ligand relative to CNBHD. (A) Ribbon diagram showing the incorporation of L-Anap (red) within the intrinsic ligand and Co2+ coordinated by a dihistidine pair in the β-roll region of the CNBHD of zELK channels. (B) Relationship of Anap fluorescence versus YFP fluorescence measured by PCF for zELK-N741Anap,L682H,G684H channels. The illustration at the top shows the positions of Anap and YFP in the sequence. A.U., arbitrary units. (C) Apparent tmFRET efficiency measured at different concentrations of Co2+ using PCF at −100 mV and in the presence of MgATP for zELK-N741Anap,L682H,G684H channels. The smooth curve is the fit of Langmuir isotherm, Apparent FRETeff. = FRETeff [Co2+]/(K1/2 + [Co2+]), with the following parameters: FRETeff = 0.64, K1/2 = 30.3 µM. (D) Representative heatmaps of PCF images of zELK-N741Anap,L682H,G684H channels in the presence of 1 mM Co2+ and 2 mM MgATP at −100-mV or +120-mV voltages. The brightfield view of the same patch is shown on the left. The Anap fluorescence intensity of the patch membrane at +120 mV decreased 34% compared with that at −100 mV. (E) Spectra of L-Anap emission at different voltages from…

• 

  Figure 6.. Kinetic tmFRET measurement demonstrated the movement of the intrinsic ligand during VDP. (A–C) Simultaneous tmFRET and current measurements using PCF for zELK-N741Anap,L682H,G684H channels with 1 mM Co2+ in the presence (A) or absence (B) of 2 mM MgATP or after deleting the PAS cap (C). The green trace in A is the single-exponential fit with a τ of 219 ms. (D) Ribbon diagram showing the full-length KCNH channel (rEAG1; Protein Data Bank accession no. 5K7L), highlighting the intrinsic ligand (in yellow) and the intersubunit interaction between the eag domain (in blue) and CNBHD (in green). Only two subunits are shown for simplicity except that the C-linker region (in magenta) is displayed in a tetrameric assembly. Ribbon diagram is the amplified view of the structure in the dashed-line rectangle. The red arrow indicates the eag domain–CNBHD rearrangement associated with the VDP.