Identifying mutation hotspots reveals pathogenetic mechanisms of KCNQ2 epileptic encephalopathy

Figure 1. Epilepsy mutations cluster at the S4, the pore, and the S6 domains of Kv7.2. (a) Pie charts showing the relative proportions of different types of epilepsy variants in human Kv7.2 (left) and of single amino acid Kv7.2 mutations that cause epilepsy with different clinical severity including “mild or BFNE”, “uncertain severity”, and “severe or epileptic encephalopathy (EE)” (right). All mutations identified up to December 31, 2017 are shown in Supplementary Table S1. (b) Single amino acid mutations except for 2 missense mutations in the primary start ATG codon (green markers) and silent mutations (purple markers) are mapped to Kv7.2 primary structure (NP_742105.1). The MHF statistical algorithm identified S4, the pore loop, and S6 as hotspots for all pathogenic epilepsy mutations (brackets, ***p < 0.005, Supplementary Table S2). (c) The MHF algorithm revealed S4, the pore loop, and S6 as hotspots for “severe or EE” mutations (brackets, **p < 0.01, ***p < 0.005, Supplementary Table S3). (d) The S1-S6 domains of tetrameric human Kv7.2 were modeled based on the cryoEM structure of Xenopus Kv7.1 (Protein Data Bank: 5VMS)28. At left, two opposing subunits are viewed from the plane of…

Figure 2. Epilepsy mutations cluster at CaM-binding helix B and the helix B-C linker of Kv7.2. (a) Tetrameric human Kv7.2 (ribbons) in complex with four CaM subunits (transparent green surfaces) was modeled based on the structure of Xenopus Kv7.1 (Protein Data Bank: 5VMS)28. Sites of pathogenic mutations in Kv7.2 C-terminal tail are highlighted with colored spheres on the C-alpha atoms of one subunit: mild or BFNE (blue), uncertain severity (purple), severe or EE (red). In addition to epilepsy mutation hotspots (S4, the pore loop, and S6), “severe or EE” mutations cluster near the inner leaflet of the plasma membrane and the C-terminus of S6 at the base of the gate. (b,c) The MHF statistical algorithm on the intracellular C-terminal tail of Kv7.2 identified helix B and the helix B-C linker as hotspots for all pathogenic epilepsy mutations (brackets, **p < 0.01, ***p < 0.005, Supplementary Table S2) (b) and EE mutations (brackets, *p < 0.05, ***p < 0.005, Supplementary Table S3) (c). (d) Location of amino acids mutated in mild or BFNE (blue), uncertain severity (purple), or severe or EE (red) in helix A containing a consensus IQ motif for binding CaM (underlined), helix B, and the helix B-C…

Figure 3. All selected EE mutations variably alter voltage-dependent activation of homomeric Kv7.2 channels and disrupt their current enhancement upon diC8-PIP2 inclusion. (a) Sites of selected EE mutations (L203P, L268F, K552T, and R553L) characterized in this study. These mutations are highlighted with red spheres on the C-alpha atoms of one subunit on the modeled tetrameric human Kv7.2 structure (ribbons) in complex with four CaM subunits (transparent green surfaces). (b) Localization of selected EE mutations are shown in red in the amino acid sequence of Kv7.2 (NP_742105.1). The EE mutations are shown in bold. The critical residues in S4 and the selectivity filter in the pore are underlined. (c–h) Whole cell voltage clamp recordings of macroscopic K+ currents in CHO hm1 cells transfected with GFP and Kv7.2 WT or EE mutants. Cells were held at -80 mV. Currents were evoked by depolarization for 1.5 s from −100 mV to +20 mV in 10 mV increments, followed by a step to 0 mV for 300 ms. To examine PIP2 sensitivity of Kv7.2 channels, the recording was repeated with internal patch pipette solution containing diC8-PIP2 (100 μM) which also contained EGTA to sequester free Ca2+. The raw current…

Figure 4. Modeled Kv7.2 structure and molecular dynamics simulation suggest that selected EE mutations reside in PIP2 binding regions of Kv7.2. (a) Homology structure of Kv7.2 from Fig. 3a and its surface charge representation were compared side-by-side with the cartoon representation of the TRPV1 structure in nanodiscs (PDB: 5irz)43. The cross-section view of Kv7.2 with electrostatic surface showing basic surface in blue, acidic surface in red and nonpolar surface in white. The surface charge representation reveals the nonpolar cavities between the voltage-sensor and the pore domain and indicates that parts of the voltage-sensor, S6, pre-helix A, helix B and helix B-C linker form two basic pockets. (b) The open-state conformation of Kv7.2 channel was embedded in the lipid bilayer. An R553 residue from the helix B of each subunit is represented as a purple sphere. Each subunit is labeled in green, blue, light grey or orange. (c) Sample trajectory of PIP2 in the modeled Kv7.2 structure viewed from the bottom (cytoplasmic) side in Fig. 4b. The initial locations of PIP2 at the beginning of the molecular dynamics (MD) simulations are indicated in red, and the final positions of PIP2 are…

Figure 5. The L268F and K552T mutations decreased current expression of heteromeric Kv7 channels whereas all selected EE mutations disrupted their current potentiation by diC8-PIP2 inclusion. Whole cell patch clamp recordings were measured using the voltage clamp protocol described in Fig. 3 from GFP-positive CHOhm1 cells cotransfected with Kv7.3 and Kv7.2 WT (1:1 ratio) or Kv7.3, Kv7.2 WT, Kv7.2 mutant (2:1:1 ratio). The raw current traces and data are shown in Supplementary Figs. S6 and S7. (a) Representative leak-subtracted current traces. (b) Average leak-subtracted peak current densities at all voltage steps. *p < 0.05, ***p < 0.005 based on one-way ANOVA Fisher’s test. (c) Average leak-subtracted peak current densities at -20 mV (top) and at + 20 mV (bottom). p values are computed from one-way ANOVA Tukey test. (d) Normalized conductance (G/Gmax) at all voltage steps. (e) Activation time constant (τ) at + 20 mV. The number of GFP-cotransfected cells that were recorded without diC8-PIP2: Kv7.2 WT (n = 14), L203P (n = 15), L268F (n = 18), K552T (n = 15), or R553L (n = 15). The number of GFP-cotransfected cells that were recorded with diC8-PIP2: Kv7.2 WT (n = 12), L203P (n = 12), L268F…

Figure 6. The K552T and R553L mutations reduced CaM binding to Kv7.2 whereas none of the tested EE mutations affected Kv7.3 interaction with Kv7.2. (a,b) Co-immunoprecipitation of YFP-CaM with wild-type Kv7.2 (WT) or Kv7.2 containing EE mutations in the presence of EGTA. (a) Representative immunoblots of HEK293T cells expressing Kv7.2 and YFP-CaM. For clarity, cropped gel images are shown. Full-length gels can be found in Supplementary Fig. S9. (b) Quantification of immunoblots: untransfected/None (n = 6), YFP-CaM (n = 6), YFP-CaM cotransfection with Kv7.2 WT (n = 6), L203P (n = 3), L268F (n = 3), K552T (n = 3), or R553L (n = 3). (c,d) Co-immunoprecipitation of HA-Kv7.3 with wild-type Kv7.2 (WT) or Kv7.2 containing EE mutations in the presence of EGTA. (c) Representative immunoblots of HEK293T cells expressing Kv7.2 and Kv7.3. For clarity, cropped gel images are shown. Full-length gels can be found in Supplementary Figs. S9–10. (d) Quantification of immunoblots: untransfected cells (None: n = 3), or cells transfected with HA-Kv7.3 (n = 4), HA-Kv7.3 and Kv7.2 WT (n = 4), L203P (n = 3), L268F (n = 3), K552T (n = 4), or R553L (n = 3). GAPDH served as a loading control. Both monomeric Kv7.2…

Figure 7. L268F, K552T, and R553L mutations decreased enrichment of heteromeric Kv7 channels at the axonal surface in cultured hippocampal neurons. Immunostaining of surface Kv7.3 and total Kv7.2 containing an extracellular hemagglutinin epitope (HA-Kv7.3) in healthy hippocampal neurons cotransfected with Kv7.2 WT or Kv7.2 with EE mutations L203P, L268F (a–e), K552T, and R553L (f–j). (a,f) Representative images of surface HA-Kv7.3 (Upper) as pseudo-color that display differences in the surface HA intensities from high (red) to low (blue). Total Kv7.2 (Middle-inverted gray) and the AIS identified by antibodies for phospho IκBα-Ser32 (14D4) (Lower-fluorescence) are shown in same neurons. Arrows mark the AIS. Scale bars: 25 μm. (b,g) Normalized background-subtracted mean intensities of surface HA fluorescence from neurons expressing WT and EE mutant HA-Kv7.3/Kv7.2 channels. The number of transfected neurons that were analyzed in Fig. 7b: WT (n = 41), L203P (n = 39), L268F (n = 34), untransfected (UT) (n = 20). The number of transfected neurons that were analyzed in Fig. 7g: WT (n = 42), K552T (n = 34), R553L (n = 21), untransfected (UT) (n = 19). The raw data from 3 independent experiments…

Figure 8. A summary for pathogenetic mechanisms of epilepsy mutation hotspots in Kv7.2. Our functional characterization of EE mutations selected from the epilepsy mutation hotspots of Kv7.2 show that they not only disrupt the functions of key protein domains they reside in, but also reduce PIP2-dependent current potentiation and axonal surface expression of Kv7 channels. The missense EE mutations are highlighted with red spheres on the C-alpha atoms of one subunit on the modeled tetrameric human Kv7.2 structure (ribbons) in complex with four CaM subunits (transparent green surfaces). The list of these EE variants can be found in Supplementary Table S7.