Ng CA et al. (2013), C-terminal β9-strand of the cyclic nucleotide-b...

XB-ART-49878

PLoS One 2013 Oct 04;810:e77032. doi: 10.1371/journal.pone.0077032.

Show Gene links Show Anatomy links

C-terminal β9-strand of the cyclic nucleotide-binding homology domain stabilizes activated states of Kv11.1 channels.

Ng CA , Ke Y , Perry MD , Tan PS , Hill AP , Vandenberg JI .

???displayArticle.abstract???
Kv11.1 potassium channels are important for regulation of the normal rhythm of the heartbeat. Reduced activity of Kv11.1 channels causes long QT syndrome type 2, a disorder that increases the risk of cardiac arrhythmias and sudden cardiac arrest. Kv11.1 channels are members of the KCNH subfamily of voltage-gated K(+) channels. However, they also share many similarities with the cyclic nucleotide gated ion channel family, including having a cyclic nucleotide-binding homology (cNBH) domain. Kv11.1 channels, however, are not directly regulated by cyclic nucleotides. Recently, crystal structures of the cNBH domain from mEAG and zELK channels, both members of the KCNH family of voltage-gated potassium channels, revealed that a C-terminal β9-strand in the cNBH domain occupied the putative cyclic nucleotide-binding site thereby precluding binding of cyclic nucleotides. Here we show that mutations to residues in the β9-strand affect the stability of the open state relative to the closed state of Kv11.1 channels. We also show that disrupting the structure of the β9-strand reduces the stability of the inactivated state relative to the open state. Clinical mutations located in this β9-strand result in reduced trafficking efficiency, which suggests that binding of the C-terminal β9-strand to the putative cyclic nucleotide-binding pocket is also important for assembly and trafficking of Kv11.1 channels.

???displayArticle.pubmedLink??? 24204727
???displayArticle.pmcLink??? PMC3808384
???displayArticle.link??? PLoS One
???displayArticle.grants??? [+]

Species referenced: Xenopus laevis
Genes referenced: ddx59 gnl3 kcnh2

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

	Figure 2. Gating Phenotype of AAA mutant.(A) Family of current traces recorded during a 3 s isochronal activation protocol for (i) WT and (ii) AAA mutant channels. (iii) Isochronal activation curves for WT (filled black circle) and AAA mutant (filled grey circle). The mean V0.5 of isochronal activation for the AAA mutant and WT were −15.1±1.0 mV (n = 4) and −23.1±0.4 mV (n = 4), respectively; see Table S1 in File S1. (B) Families of −70 mV tail current traces recorded during a 3 s isochronal deactivation protocol for (i) WT and (ii) AAA mutant. The dotted box in the voltage protocol indicates the portion of the traces shown in the current recordings. (iii) Isochronal deactivation curves for WT (filled black circle) and AAA mutant (filled grey circle). Data are presented as mean ± SEM for n = 4 experiments. The mean V0.5 of isochronal deactivation for the AAA mutant and WT were −23.4±1.1 mV (n = 4) and −61.3±0.8 mV (n = 4), respectively. (C) Typical family of current traces recorded between −60 to −160 mV at 20 mV intervals, corresponding to the dotted box in the voltage protocol, used to measure rates of deactivation for (i) WT and (ii) AAA mutant. (iii) Summary of the rates of deactivation for AAA mutant (grey) and WT (black). Data shown as mean ± SEM (n = 4), error bars are within the symbols.
	Figure 3. Inactivation phenotype of AAA mutant.(A) Current traces correspond to dotted box in the voltage protocol used to measure the recovery of inactivation for (i) WT and (ii) AAA mutant. Current traces recorded at −80 mV are highlighted to show the faster recovery of inactivation for AAA mutant. (B) Current traces correspond to dotted box in the voltage protocol used to measure the onset of inactivation for (i) WT and (ii) AAA mutant. Current traces recorded at 0 mV are highlighted to show the slower onset of inactivation for the AAA mutant. (C) Summary of rates of recovery and onset of inactivation plotted against voltages between −130 and +50 mV. The data points for −80 and 0 mV are indicated by the arrows. The mid-point of steady-state inactivation for the AAA mutant (grey) is right-shifted by ∼33 mV from WT (black). (D) V0.5 of steady-state inactivation for WT (−51.7±1.9 mV, n = 7; filled black circle) and AAA mutant (−18.7±2.7 mV, n = 4; filled gray circle) (*indicates p<0.05 versus WT, ANOVA). Data are presented as mean ± SEM.
	Figure 4. Summary data for individual β9-strand mutants.(A) Scatter plot of the V0.5 values for the 3 s isochronal activation (open symbols) and 3 s isochronal deactivation (closed symbols) for WT (black), AAA (grey), F860A (red), N861A (magenta), L862A (blue), F860L (orange), F860Y (green) and F860R (cyan). (B) Scatter plot of the V0.5 values for the steady-state inactivation (open symbols) for WT, AAA, F860A, N861A, L862A, F860L, F860Y and F860R (same colour scheme as in panel A). In all panels, the mean and SEM are indicated by horizontal bars and asterisks indicate values that are statistically significantly different to WT (P<0.05, ANOVA). The dashed horizontal lines indicate mean values for WT. The values for all mutants are summarized in Table S1 in File S1.
	Figure 5. Secondary structure prediction and MD simulations of cNBH domain.Sequence prediction of the cNBH domain around the β9-strand for (A) WT (i) and AAA mutant (ii). (B) RMSD (i) and RMSF (ii) of WT (red) and AAA mutant (black) from the 60 ns of MD simulations. The blue box highlights the most significant difference between WT and AAA mutant in the backbone fluctuation. (C) The structures that have the lowest structural fluctuation to the centroid structure in the most populated cluster from the last 10 ns for WT (i) and AAA mutant (ii). Residues involved in hydrophobic interactions, defined by being within 4 Å of residues 860, 861 and 862 (cyan), are highlighted in magenta for WT (i) and AAA mutant (ii). There are reduced hydrophobic interactions in the AAA mutant. (D) Summary of residues that participate in hydrogen bonds with residues 860, 861 and 862 in WT (i) and the AAA mutant (ii) that are present for more than 5% of the 60 ns of MD simulation (see Table 1 for details).
	Figure 6. Trafficking assay of LQT2 mutants located within β9-strand.(A) Typical western blot of WT, N861I and N861H mutant channels. WT shows two bands at ∼155 kDa and ∼135 kDa. The ∼155 kDa band disappears following digestion of surface proteins with proteinase K. The N861H mutant shows only a single ∼135 kDa band. N861I contains both ∼155 kDa and ∼135 kDa bands. Arrow indicates degradation band after proteinase K digestion. (B) Normalized expression levels of N861H and N861I relative to WT for the fully glycosylated (∼155 kDa band) and core-glycosylated (∼135 kDa band) proteins. (C) The partially trafficking defective N861I can be rescued by incubation with cisparide whereas N861H was not rescued by cisapride. (D) Co-imunpreciptation of HA-tagged mutant subunits with Flag-tagged WT subunits. (E) Top panel: Summary of 3 s isochronal activation V0.5 (open symbols) and 3 s isochronal deactivation V0.5 (closed symbols) for WT (black), N861H (magenta) and N861I (blue). Asterisks indicate P<0.05 (ANOVA) compared to WT. Bottom panel: Summary of the V0.5 of steady-state inactivation for WT, N861H and N861I (same colours as in top panel). Mean data for all mutants are summarized in Table S1 in File S1.
	Figure 1. Topology of Kv11.1 channels and sequence analysis of cNBH domains.(A) Topology of Kv11.1 channel showing the intracellular N-terminal PAS domain (blue), transmembrane voltage sensing domain (green), pore domain (yellow) and intracellular C-terminal C-linker and cNBH domains (orange). Inset shows the homology model of the cNBH domain of Kv11.1 generated based on the mEAG crystal structure [13]. (B) Sequence alignment of mHCN2, zELK, mEAG and human Kv11.1 extracted from a Clustalw alignment of the entire family of KCNHx/HCNx/CNGx ion channels. Sequences shown correspond to the dotted box region shown in panel A. Sequence similarity to the Kv11.1 are marked by white text/red box (identical) and black text/yellow box (similar). Non-conserved sequences are in grey. Clear rods and arrows represent the consensus α-helices and β-strands while filled rods and arrows indicate the differences with orange, green and blue representing mHCN2, zELK and mEAG, respectively. The hydrogen bond between asparagine (arrow) and tyrosine (asterisk) in zELK is not observed in the others.

References [+] :

Akhavan, Identification of the cyclic-nucleotide-binding domain as a conserved determinant of ion-channel cell-surface localization. 2005, Pubmed