The voltage-gated sodium channel EF-hands form an interaction with the III-IV linker that is disturbed by disease-causing mutations

Figures

• 

  Figure 1. Disease mutations in the C-terminus of Nav1.5. (a) Schematic depiction of the structural entities of the channel. (b) Disease-associated mutations plotted on Nav1.5 C-terminal domain (red, light brown) interacting with Mg2+-loaded CaM (grey; PDB ID 4OVN). This region is immediately downstream of the last transmembrane segment and contains the EF-hand domain (beige), an additional helix (orange), and an IQ motif (red) that binds CaM. The disease-causing mutations are linked to LQT3 (black), BrS (blue), or both (brown). (c) Thermal denaturation curves for wild type and disease mutant EF-hand domain constructs using a Thermofluor assay. The displayed curves are an average of at least three measurements. The melting temperatures, standard errors, and number of repeats are shown in Table S1. R1860Gfs*12 is a frameshift mutation, which leads to a premature stop of translation after 11 amino acids of non-WT sequence.

• 

  Figure 2. Inactivation-voltage relationship for wild-type and mutant Nav1.5 channels determined by two-electrode voltage clamp. (a) Representative raw traces of the test pulses for wild-type and disease mutants. The trace with −80 mV pre-pulse potential is shown in red. (b) Pulse protocol used to obtain steady-state inactivation curves. (c) Steady-state inactivation curve of wild-type (•),L1825P (□), R1826H (Δ) and Y1795insD (∇). The lines represent fits of the data with a Boltzmann equation. V1/2 values for inactivation are presented in Table S2.

• 

  Figure 3. Isothermal titration calorimetry data of the EF-hand domain interaction with the III-IV linker. (a) EF-hand domain (160 μM) was titrated with aliquots of 2 mM III-IV linker. Average of three repeats yielded a Kd of 87 μM ± 6 μM. (b) III-IV linker (175 μM) was titrated with 2 mM EF-hand domain in buffer containing 4 mM CaCl2. One set of sites model fit in Origin yielded a Kd of 87 μM ± 12 μM. (c) III-IV linker (175 μM) premixed with 285 μM of CaM titrated with EF-hand domain in buffer containing 4 mM CaCl2. These show that the interaction now is exothermic instead of endothermic. The integrated heats are too small to allow a reliable fit of the data.

• 

  Figure 4. NMR spectroscopic characterization of the EF-hand domain interaction with III-IV linker peptide. (a) Assigned 15N−HSQC spectrum of the 15N/13C-labelled EF-hand domain. (b) Superimposed 15N-HSQC spectra (boxed region in A) showing chemical shift perturbations of the 15N-isotopically labelled EF-hand domain before and after addition of the III-IV linker peptide. Spectra are coloured according to the molar ratio of EF-hand domain: III-IV linker, 1:0 blue, 1:0.25 cyan, 1:0.5 green, 1:1 yellow, 1:2 orange, 1:4 red. Annotated peaks show weighted amide 1HN-15N shift changes of more than 0.03 ppm at the final titration point. (c) Weighted chemical shift perturbations mapped on the EF-hand domain structure 2KBI. The ribbon representation is annotated with residues corresponding to peaks found to shift. The bottom panel shows a surface representation.

• 

  Figure 5. Isothermal titration calorimetry data demonstrating that the EF-hand domain does not interact with III-IV linker mutant K1504E/K1505E. (a) Purified EF-hand domain sample at 175 µM was titrated with 2 mM K1504E/K1505E peptide. Compared with Fig. 3, the mutation clearly affects binding. The binding isotherm is too shallow to allow a reliable fit. (b) Theoretical electrostatic surface potential of the interaction partners plotted on models of III-IV linker (PDB ID 1BYY, extended for missing Nav1.5 sequence via Modeller61) and EF-hand domain (PDB ID 2KBI, shown are only residues included in our constructs). Potentials calculated with the APBS tool in UCSF Chimera, red indicates negative, blue positive charge.

• 

  Figure 6. Two-electrode voltage clamp measurement of steady-state inactivation shows differences in voltage-dependent inactivation for designed inactivation gate mutant. (a) Representative raw traces of the test pulse for wild-type and K1504E K1505E NaV1.5. The −80 mV pre-pulse potential trace is shown in red. (b) Steady state inactivation data of wild-type (•) and mutant channels (ο) fitted with a Boltzmann function. The holding potential was −110 mV. The mutant shows a pronounced difference in inactivation and late sodium current.

• 

  Figure 7. Sequence alignment of mammalian voltage-gated sodium channel isoforms with NavPaS, EeNav1.4 and human CaV1.1. (a) III-IV linker region and (b) C-terminal domain region aligned with Clustal Omega. Fully conserved residues are shaded. Known disease associated residues in Nav1.5 are marked by boxes coloured according to their disease phenotype LQT3 (black), BrS (blue), or both (brown). The phosphorylation site in the III-IV linker is marked by an asterix51. Secondary structure has been annotated according to the DSSP calculations of Nav1.2 (PDB ID 1BYY) and Nav1.5 (PDB ID 4OVN) structures in the protein structure database.

• 

  Figure 8. Model for the functional importance of the EF-hand domain: III-IV linker interaction. (a) Structure of NaVPaS (PDB ID 5X0M) showing the III-IV linker (blue) in close contact with the EF-hand domain (beige). (b) In the structure of EeNav1.4 (PDB ID 5XSY) the EF-hand like domain is not resolved and the III-IV linker (blue) adopts a conformation where the IFM equivalent is interacting with a hydrophobic pocket, possibly representing an inactivated state. The EF-hand domain is not present in the structural model indicating possible mobility. Subunits other than the pore forming α-subunit have been omitted for clarity. In both structures the equivalents of the IFM motif are depicted as black sticks. The arrows indicate the position of the engineered mutation site (K1504E/K1505E in NaV1.5). Voltage sensors of domain II and IV are labelled; domain III is oriented towards the viewer. For ease of orientation the end of helix 6 transmembrane domain IV is coloured orange. (c) The EF hand domain interacts with the III-IV linker, thus imposing an energetic barrier for channel inactivation at a given membrane potential. Abolishing the interaction (bottom), either through disease mutations in…