Pisupati A , Mickolajczyk KJ , Horton W , van Rossum DB , Anishkin A , Chintapalli SV , Li X , Chu-Luo J , Busey G , Hancock WO , Jegla T .

F99 CA223018 NCI NIH HHS , R01 GM076476 NIGMS NIH HHS , R01 NS069842 NINDS NIH HHS

             voltage-gated cation channel activity

	Figure 1. Tetrameric Shaker-like Kv channels have two major intersubunit interfaces. (A) Schematic cartoon depicting subunit domain arrangement in Shaker-like Kv channels. Two diagonally opposed subunits of the tetrameric channel are shown. The PDs (magenta) from each subunit surround a central ion-conducting pore, while the VSDs (blue) are physically isolated at the periphery. The conserved N-terminal cytoplasmic assembly domains (T1, black) form a ring beneath the pore. (B) A more detailed cartoon of the tetrameric ion conducting pore from an extracellular perspective, with adjacent subunits differentially shaded. Transmembrane helices S5 and S6 are depicted with cylinders, and a selectivity filter helix is shown as a black ribbon. The pore-lining S6 helix forms a major intersubunit interface (arrows), and the intracellular side comprises the activation gate. (C) Cartoon of the cytoplasmic T1 ring with adjacent subunits differentially shaded. There are major intersubunit contacts between neighboring, but not diagonally opposed, T1s (arrows). Helices are depicted with cylinders, and β-sheets are depicted with rectangles. (D) Amino acid alignment of mouse Kv2.1 and Kv6.4 with residue numbers given at the right margin. Identical residues are shaded black, and conservative substitutions are shaded gray. The T1 domain is underlined in magenta, transmembrane domains S1–S6 are underlined in dark blue, and the PD is underlined in light blue. The alignment was produced with the CLUSTALW algorithm as implemented in MEGA7 (Kumar et al., 2016), and un-conserved N and C termini have been trimmed. (E) T1 self-incompatibility in regulatory subunits (blue) theoretically allows formation of heterotetramers with a single (3:1R) or two diagonally opposed (2:2R) regulatory subunits because four compatible T1 contacts (+ signs) remain in each case. However, T1 self-incompatibility rules out formation of 2:2R with adjacent regulatory subunits, 1:3R and 4R tetramers, all of which have at least one incompatible intersubunit contact (minus signs).
	Figure 2. The S6 activation gate sequence differs from the Shaker-like Kv family consensus in evolutionarily independent regulatory subunit groups. (A) Cartoon showing the PD of two diagonally opposed subunits of a closed Shaker-like Kv channel tetramer with the location of a six–amino acid sequence (PVPVIV) comprising the S6 gating hinge and activation gate overlaid to show its approximate location at the intracellular side of the conduction pathway. (B) S6 consensus sequence logo constructed from all 27 mouse and sea anemone (N. vectensis) Shaker-like Kv family subunits that can form functional homotetrameric channels. The hinge and activation gate are boxed with a red outline, and positions 1–6 as used in the paper are indicated. A highly conserved section of S6 upstream of the gate is also shown in the logo. Amino acid frequency is encoded in letter height, and colors are used to depict amino acid class (blue, hydrophobic; green, hydrophilic; magenta, proline; black, aromatic; orange, acidic; purple, basic). Note the conservative substitution V2I and V4I are found in mouse Kv2.1 and Kv2.2 and are typical for the Kv2 subfamily. (C) A similar sequence logo constructed from 43 mouse and sea anemone Shaker-like Kv regulatory subunits shows degeneration of the gate PVPVIV consensus sequence; only P1 is found in a majority of regulatory subunits. (D) Sequence logo for nine mouse Kv2 subfamily regulatory subunits with a common evolutionary origin (Kv6.1-6.4, Kv8.1-8.2, Kv9.1-9.3). Asterisks mark positions with unusual substitutions in at least some members of the group. Note the loss of P3 (hydroxyl typical) and V6 (aromatic typical) across the group. (E–K) Sequence logos broken out for seven additional evolutionarily independent groups of mouse and sea anemone regulatory subunits. Each separate group of regulatory subunits has a distinct pattern of unusual gate substitutions, marked by asterisks. Accession numbers and sequences for channels used to make the sequence logos in B–E are given in Supplemental Table 1, and amino acid frequencies at each gate position are given in Supplemental Table 2.
	Figure 3. Comparison of gating properties of Kv2.1 homomers and Kv2.1:Kv6.4 heteromers. (A) Example currents recorded under TEVC from oocytes expressing Kv2.1 (left) or Kv2.1:Kv6.4 (right) in response to 400-ms depolarizations ranging from −60 to +40 mV in 20-mV increments from a holding potential of −100 mV. Tail currents were recorded at −50 mV. Scale bars are included for current amplitude and time and the voltage protocol is shown below the currents. The step to 0 mV is highlighted in magenta to emphasize the left-shift in voltage activation in Kv2.1:Kv6.4. For Kv2.1:6.4, a 1:10 cRNA ratio was used for the experiments in this figure to produce a predominantly heteromeric current (Fig. 8). (B) Normalized GV relationships determined from isochronal tail currents recorded at −50 mV after 400-ms steps to the indicated voltages are shown for Kv2.1 (black, n = 7) and Kv2.1:Kv6.4 (magenta, n = 10). Data points show mean, error bars indicate standard error, and smooth curves represent a single Boltzmann fit of the data. V50 and slope values are given in Table 1. The V50 values were significantly different (t test, P < 0.001). (C) Example currents recorded from oocytes for Kv2.1 and Kv2.1:Kv6.4 during a protocol used to determine SSI. Oocytes were stepped from a holding potential of −100 mV to voltages ranging from −110 to +30 mV in 10-mV increments for 4 s before a 500-ms step to +40 ms to measure current size. For display purposes, here we show currents recorded in response to prepulses from −40 to +20 mV in 20 mV increments for Kv2.1, and from −100 to −40 mV in 20-mV increments for Kv2.1:Kv6.4. The −40 mV sweep is highlighted in magenta for comparison and represents a voltage at which SSI is virtually absent in Kv2.1 and near maximal in Kv2.1:Kv6.4. Thus, a comparison of currents recorded at +40 mV after prepulses to −100 mV and −40 mV can be used to distinguish homomeric and heteromeric current fractions. Scale bars are given for time and current amplitude, and the voltage protocol is shown below the currents. (D) Normalized SSI curves for Kv2.1 (black, n = 7) and Kv2.1:Kv6.4 (magenta, n = 10). Data show mean ± SEM, and curves show a single Boltzmann fit of the data. V50 and slope values are given in Table 1. Note SSI is incomplete for both Kv2.1 and Kv2.1:Kv6.4; residual pedestal current fractions as determined by Boltzmann fitting are given in Table 1.
	Figure 4. Determination of Kv2.1:Kv6.4 heteromer stoichiometry by TIRF photobleaching assay. (A) Example TIRF images from an oocyte expressing GFP-Kv2.1 before (left) and after photobleaching (middle). Spots that bleached to background during the photobleaching period (blue circles) were analyzed for number of bleaching steps. Bleach-resistant spots (yellow circles) or spots that bleached but did not have a stable baseline (magenta) were not included in the analysis. Example fluorescence traces for spots in the movie bleaching in one (I), two (II), three (III) and four (IV) steps are shown at the right margin. Fluorescent spots (circled) bleachable in discrete steps were observed only in oocytes expressing GFP-tagged channels. (B) Example TIRF images and fluorescence traces before and after photobleaching for Kv2.1 coexpressed with GFP-Kv6.4, labeled as in A. Only spots with one or two bleaching steps were observed. (C) Frequency distribution of bleaching steps for channel spots in oocytes expressing GFP-Kv2.1 (Kv2.1 homomers). Only spots bleaching in one to four steps were observed. Assuming all channels are tetramers with four GFPs, a binomial fit of the distribution (squares) estimates the probability of detecting GFP fluorescence and bleaching at 69% in our experimental setup. (D) Frequency distribution of bleaching steps for channel spots in oocytes coexpressing GFP-Kv2.1 and Kv6.4 compared with GFP-Kv2.1 alone. The addition of Kv6.4 RNA significantly reduces the frequency of four steps from 22% to 11% and reduces the frequency of three steps from 42% to 32%. There is a corresponding increase in the frequency of two steps from 25% to 39% and an increase in the frequency of one steps from 12% to 17% (P < 0.05, Fisher’s exact test). This shift in step distribution confirms detection of Kv2.1:Kv6.4 heteromer formation. (E) Frequency distribution of bleaching steps for channel spots in oocytes injected with Kv2.1:GFP-Kv6.4 in an ∼1:50 ratio to maximize heteromer formation. Only spots bleaching in one to two steps were observed, and 81% bleach in one step. (F) Relative frequency of 3:1R and 2:2R Kv2.1:GFP-Kv6.4 heteromers calculated from the bleaching step distribution using the 69% GFP detection probability determined for Kv2.1 homomers as described in C.
	Figure 5. The Kv6.4 activation gate but not T1 limits the formation of Kv2.1:Kv6.4 2:2R heteromers. (A) Schematic diagram of the chimeric constructs used to test 2:2R formation by TIRF microscopy. The N-terminal T1 domain is labeled, and transmembrane domains S1–S6 are shown as rectangles, with a star next to the S6 activation gate. Sequences derived from Kv6.4 are shown in black, and sequences inserted from Kv2.1 are shown in blue. Chimera names are given below each diagram. (B) Percentages of 3:1R and 2:2R heteromers calculated from TIRF photobleaching assays for Kv2.1 coexpressed with GFP-tagged versions of WT Kv6.4 (magenta), Kv6.4-Kv2.1T1 (black), and Kv6.4-PIPIIV-Kv2.1CT (blue). The stoichiometry distribution for Kv6.4-Kv2.1T1 was unchanged, but Kv6.4-PIPIIV-Kv2.1CT had significantly fewer 3:1R heteromers and significantly more 2:2R heteromers (P < 0.05, χ2 test). Kv2.1 was expressed with Kv6.4-Kv2.1T1 and Kv6.4-PIPIIV-Kv2.1CT in a 1:80 and a 1:60 cRNA ratio, respectively, to eliminate Kv2.1 homomers and bias 2:2R heteromer formation. (C) Expression ratio dependence of 2:2R heteromer formation is shown for WT Kv6.4 and Kv6.4-PIPIIV-Kv2.1CT. 2:2R formation was significantly higher for both constructs at the highest concentration tested relative to lower concentrations (*, P < 0.05, χ2 test). Note 2:2R formation is higher for Kv6.4-PIPIIV-Kv2.1CT across a broad range of expression ratios. (D) Kv2.1:Kv6.4 2:2R heteromer percentages detected for all Kv6.4 chimeras tested in TIRF photobleaching assays at a highly Kv6.4-biased expression ratio. 2:2R formation was significantly increased relative to WT for the Kv6.4-PIPIIV-Kv2.1CT, Kv6.4-PIPIIV, and Kv6.4-Kv2.1T1-PIPIIV-Kv2.1CT constructs (*, P < 0.05, χ2 test). The P3T and V6F mutations in the Kv6.4-PIPIIV-Kv2.1CT background eliminate the increase in 2:2R formation. Bleaching step counts and expression ratios for the data in D were as follows (construct name, ratio, no. of one-steps, no. of two-steps): Kv6.4 WT, 1:200, 98, 23; Kv6.4-Kv2.1CT, 1:142, 87, 21; Kv6.4-PITIIF-Kv2.1CT, 1:154, 85, 13; Kv6.4-PITIIV-Kv2.1CT, 1:128, 64, 16; Kv6.4-PIPIIF-Kv2.1CT, 1:334, 67, 15; Kv6.4-PIPIIV-Kv2.1CT, 1:60, 72, 39; Kv6.4-PIPIIV, 1:185, 80, 39; and Kv6.4-Kv2.1T1-PIPIIV-Kv2.1CT, 1:78, 63, 32. We did not detect spots bleaching in three or four steps for any GFP-Kv6.4 chimera tested in this study.
	Figure 6. Biophysical properties of heteromeric currents from oocytes coexpressing Kv2.1:Kv6.4-PIPIIV-Kv2.1CT. (A) Example currents recorded from an oocyte expressing Kv2.1:Kv6.4-PIPIIV-Kv2.1CT in a 1:10 ratio in response to 1-s voltage steps ranging from −80 to +40 mV in 20-mV increments from a holding potential of −100 mV. The voltage protocol is shown below the currents, the 0-mV trace is highlighted in magenta, and scale bars are given for time and current amplitude. (B) Normalized GV relationship for Kv2.1+Kv6.4-PIPIIV-Kv2.1CT determined from isochronal tail currents recorded at −50 mV after 1-s steps to the indicated voltages. Data points show mean ± SEM (n = 10), and the solid blue curve represents a single Boltzmann distribution fit of the data (V50 and slope are included in Table 1). Dashed magenta and black curves show the Boltzmann fits for Kv2.1 homomers and Kv2.1:Kv6.4 WT, respectively. (C) Example current traces for the SSI protocol for an oocyte expressing Kv2.1:Kv6.4-PIPIIV-Kv2.1CT in a 1:10 cRNA ratio. The voltage protocol is indicated below the current, the oocyte was held at −100 mV and current traces recorded in response to 4-s prepulses ranging from −120 to −40 mV in 20-mV increments are shown, and the −40 mV trace is highlighted in magenta. Pre-pulses were followed by a 500-ms step to +40 mV to show current availability. (D) Normalized SSI relationship determined for Kv2.1:Kv6.4-PIPIIV-Kv2.1CT from peak current amplitudes recorded at +40 mV following 4-s prepulses to the indicated voltages. Data show mean ± SEM (n = 10), the blue curve represents a single Boltzmann fit (parameters in Table 1), and the dashed magenta and black curves show Boltzmann fits for Kv2.1 and Kv2.1:Kv6.4 WT, respectively.
	Figure 7. Predicted stoichiometry and abundance of surface-expressed channels for Kv2.1:Kv6.4 and Kv6.4-PIPIIV-Kv2.1CT at various expression ratios. Surface-expressed channel tetramers are depicted as open circles and colored-in quarters according to subunit composition (Kv2.1, black; Kv6.4, magenta; Kv6.4-PIPIIV-Kv2.1CT, blue). Monomer abundance is shown below the channels; Kv2.1 abundance remains constant while Kv6.4 abundance is increased from left to right. Relative abundance for each stoichiometry of surface-expressed channel and each type of monomer is represented by the size of the respective icon. (A) For Kv2.1:Kv6.4, the number of 3:1R heteromers initially increases as Kv6.4 expression is increased, but then drops as availability of Kv2.1 homodimers, which are required for 3:1R formation, becomes limiting at the highest Kv6.4 expression levels. TIRF results show that 2:2R formation is inefficient and never exceeds 30% of the total heteromer population, even at the highest Kv6.4 expression levels. Therefore, the total number of surface-expressed channels will reduce and become very small as Kv6.4 expression is increased against a constant level of Kv2.1 expression. (B) For Kv2.1:Kv6.4-PIPIIV-Kv2.1CT, channel abundance is predicted to decrease less precipitously as Kv6.4-PIPIIV-Kv2.1CT expression level increases because 2:2R formation is more efficient, reaching ∼60% of all heteromers at the most Kv6.4-PIPIIV-Kv2.1CT–biased expression ratios. Assuming similar levels of 3:1R formation, this represents an approximately fourfold increase in the number of 2:2R heteromers formed compared with Kv6.4.
	Figure 8. Kv2.1:Kv6.4 expression ratio titrations predict that 2:2R heteromers have negligible conductance for Kv6.4 WT and are conducting for Kv6.4PIPIIV-Kv2.1CT. (A) Example currents from oocytes expressing Kv2.1 alone (left, black) or Kv2.1:Kv6.4 WT (magenta) at 9:1, 1:1, and 1:20 ratios. Kv2.1 cRNA level was kept constant, and currents were recorded at +40 mV following a 4-s prepulse to −120 mV to relieve SSI. Note the dramatic decrease in current size as Kv6.4 cRNA level is increased. Scale bars indicate current amplitude and time. (B) Example currents from a similar titration of Kv2.1 versus Kv6.4-PIPIIV-KV2.1CT; Kv2.1 control is shown in black, and cRNA ratios shown are 3:1, 1:1, and 1:20. More current is readily apparent at the 1:20 ratio for Kv6.4-PIPIIV-KV2.1CT than Kv6.4 WT. (C) Current amplitude versus expression ratio titrations are shown for Kv2.1:Kv6.4 WT (magenta) and Kv2.1:Kv6.4-PIPIIV-KV2.1CT (blue). Expression ratio is given as −log(Kv2.1fraction); 1.0 = a 1:10 Kv2.1:Kv6.4 cRNA ratio. The Kv2.1 cRNA amount remained constant while the Kv6.4 cRNA species amount was varied to achieve the given expression ratio. Data were normalized to the amplitude of control Kv2.1 homomeric currents. Data points show mean ± SEM (n = 10–18 eggs per ratio), and asterisks indicate a significant difference (P < 0.05, t test) between the indicated pairs of data points. (D) Normalized current size ratio for Kv2.1:Kv6.4-PIPIIV-Kv2.1CT versus Kv2.1:Kv6.4 WT increases approximately fourfold as the Kv6.4 species cRNA is increased. Data show mean ± SD and are derived from the data point pairs statistically compared in C. Kv2.1:Kv6.4 ratios were similar but not identical between the titrations, and data are plotted using the average of the Kv2.1:Kv6.4 ratio of the data pair. SD for the ratios was calculated by propagation of error from the original measurements in C. (E) Example single-channel current traces from an on-cell patch from an oocyte expressing Kv2.1 alone. Traces were recorded at −20 mV, 0 mV, and +20 mV; the dashed blue line indicates the average closed baseline, and the dashed magenta lines indicate 1× and 2× the average open-channel amplitude for the trace. (F) Traces labeled as in D for an on-cell patch containing two channels from an oocyte coexpressing Kv2.1 with Kv6.4. In this example, the dashed magenta lines indicate 1× and 2× the average open-channel amplitude of the largest channel recorded in the trace. Arrows point to examples of single- or double-channel openings with smaller than expected amplitudes, indicating a second channel with a lower conductance. (G) Plots of single-channel current amplitude versus voltage are given for Kv2.1 homomers expressed in isolation and for presumed homomers and heteromers from oocytes expressing Kv2.1 with Kv6.4. Single-channel conductance values shown on the graph were calculated from linear fits; n was 6 for the two Kv2.1 homomer measurements and 5 for the Kv2.1:Kv6.4 heteromers. The Kv2.1:Kv6.4 heteromer conductance was significantly smaller (P < 0.05, t test). Note there is a small offset in reversal potential between the experiments conducted on oocytes expressing Kv2.1 alone versus Kv2.1 + Kv6.4 that alters single-channel amplitude but does not interfere with conductance measurements; these experiments were conducted at different times with distinct solution batches. (H) Normalized total current versus heteromeric current for the Kv2.1:Kv6.4 WT coexpression titration shown in C. To determine the amplitude of the heteromeric current, we selectively inactivated heteromers with a 4-s prepulse to −40 mV before measuring current amplitude at +40 mV. The amplitude of the heteromeric current was determined by adjusting the current fraction removed by the −40 mV prepulse by the fractional inactivation determined from SSI analysis (Figure 3 D and Table 1). In this case, the inactivating fraction was multiplied by 1.38 to account for the observation that 27.8% of heteromeric channels are not expected to inactivate during the prepulse. (I) Normalized heteromeric versus total current amplitude for Kv2.1:Kv6.4-PIPIIV-Kv2.1CT; heteromeric fraction was calculated as in H except using a multiplication factor of 1.44 to match the 30.4% pedestal observed for Kv2.1:Kv6.4-PIPIIV-Kv2.1CT in SSI analysis. Note in both titrations the current approaches 100% heteromer at a 1:10 Kv2.1:Kv6.4 species expression ratio (−logKv2.1 fraction = 1).
	Figure 9. Cd2+ block mediated by adjacent Kv2.1 I383C subunits predicts that 2:2R heteromers contribute to whole cell currents for Kv2.1:Kv6.4-PIPIIV-Kv2.1CT but not for Kv2.1:Kv6.4 WT. (A) Channels predicted to form for Kv2.1 I383C (black circle with C) + Kv6.4 WT (left, magenta) or Kv2.1 I383C + Kv6.4 PIPIIV-Kv2.1CT (right, blue) at Kv2.1-biased or Kv6.4-biased expression ratios. Cd2+-coordination sites between adjacent cysteine-containing subunits are marked with green circles; all channels with Cd2+-coordination sites are predicted to be sensitive to Cd2+ block. Cd2+-resistant 2:2R heteromers (magenta box) are predicted to contribute to currents in Kv6.4-PIPIIV-Kv2.1CT-biased mixes. (B) Examples current traces before and after addition of 500 µM Cd2+ for Kv2.1, Kv2.1:Kv6.4 and Kv2.1:Kv6.4-PIPIIV-Kv2.1CT recorded in response to a 500-ms +40 mV step following a prepulse of 4 s to −100 mV. Oocytes were held at −70 mV. Experiments were performed in the presence of 200 µM dithiotheitol (DTT), and a 1:10 ratio was used for coexpression experiments. Scale bars indicate current amplitude and time. (C) Fractional current remaining after 500 µM Cd2+ addition determined by measuring the peak current during the +40 mV test pulse using the protocol described in B. Cd2+ block was significantly greater of both heteromer species compared with Kv2.1 homomers (*, P < 0.05, ANOVA + Tukey post hoc, n = 10 for each, bars show SEM). (D) Examples traces before and after application of 500 µM Cd2+ for Kv2.1 I383C, Kv2.1 I383C:Kv6.4 WT, and Kv2.1 I383C:Kv6.4-PIPIIV-Kv2.1CT. Currents were recorded as in B, and a 1:10 Kv2.1:Kv6.4 species expression ratio was used. (E) 500 µM Cd2+-resistant current fraction (peak current during a 2-s test pulse to +40 mV) is shown for Kv2.1 I383C:Kv6.4 WT (magenta) and Kv2.1 I383C:Kv6.4-PIPIIV-Kv2.1CT (blue) at the indicated cRNA expression ratios. The solid gray and dashed gray lines mark the mean ± SEM of the Cd2+-resistant current fraction for Kv2.1I383C homomers. Cd2+ block was significantly reduced at high expression ratios for Kv6.4-PIPIIV-Kv2.1CT compared with the lowest expression ratio used (*, P < 0.05, ANOVA, n = 8–11 for each ratio, bars show SEM). (F) Kv2.1 I383C homomers (n = 10), Kv2.1 I383C:Kv6.4 M422C heteromers (1:10 ratio, n = 12) and Kv2.1:Kv6.4 M422C-PIPIIV-Kv2.1CT (1:11 ratio, n = 9) heteromers, which are all predicted to have 4 Cd2+-coordination sites are highly sensitive to block by 500 µM Cd2+ (**, P < 0.01, ANOVA; *, P < 0.05, ANOVA, bars show SEM).
	Figure 10. Structural homology models of Kv2.1 homomers and Kv2.1:6.4 heteromers suggest the Kv6.4 activation gate might perturb pore stability and function. (A) Sequence alignment of the Kv1.2, Kv2.1, and Kv6.4 PDs. Transmembrane domains S5 and S6 and the K+ selectivity filter are underlined. Residues identical or conservatively substituted across all three sequences are shaded magenta and black, respectively. The six–amino acid activation gate is boxed with positions 1–6 labeled. Note there are identical signpost residues throughout the PDs allowing precise alignment between Kv1.2 (determined structure, Long et al., 2005a), Kv2.1, and Kv6.4 (structural models based on Kv1.2 presented here). (B–E) Snapshots of the closed conduction pathway viewed from the extracellular side for structural models of a Kv2.1 homomer (B), a Kv2.1:Kv6.4 3:1R heteromer (C), a Kv2.1:Kv6.4 2:2R heteromer (D), and a Kv2.1:Kv6.4 2:2R heteromer with the Kv2.1 activation gate (PIPIIV) substituted for the Kv6.4 activation gate (PATSIF; E). The protein backbone (thin tubes) is colored white for Kv2.1 and light purple for Kv6.4. Side chains (van der Waals representation) are shown for positions 2, 3, and 6 of the activation gate and colored according to hydrophobicity index in the Kyte and Doolittle scale with their values ranging from 1.8 for alanine (green) to 4.5 for isoleucine (white; scale provided below panels). Residues that line the conduction pathway (I2 and V6 in Kv2.1; A2 and F6 in Kv6.4) are labeled in C. Yellow ribbons in all panels highlight the position of the gate backbone. In Kv2.1, there is an expected intersubunit hydrophobic vapor lock at the activation gate. While insertion of a single Kv6.4 subunit (C) is well tolerated, insertion of two diagonally opposed Kv6.4 subunits (D) simultaneously introduces a hydrophilic cleft that bisects the gate and increases the distance between diagonally opposed subunits. Both changes favor disruption of the intersubunit vapor lock and could hypothetically reduce tetramer stability at the gate intersubunit interface. (F–M) The bottom two rows are the view from the cytoplasmic side—one row for the closed state (F–I) and another for the open (J–M). Side chains at the gate constriction are shown in space fill and colored by the residue type (Kv2.1: I2, white; P3, cyan; V6 gray; Kv6.4: A2, gray; T3, brown; F6, pink). In the closed models, the hydroxyl group of T3 in the Kv6.4 gate faces the neighboring subunit (G and H) and thus contributes to the hydrophilic cleft observed in the extracellular view of the 2:2R tetramer (D). In the open conformations, a dehydrated K+ ion (CHARMM36 radius, 1.76 Å) is depicted with a blue circle in the pore opening at the narrowest point of the conduction pathway in the activation gate region. A magenta dashed line surrounds the portion of the opening accessible to the center of the dehydrated K+ ion (magenta dot) as determined by rolling the ion against the side chains lining the pore. Blue mesh covers the inaccessible region of the pore. The black circle around the ion roughly approximates the radius of the first hydration shell (i.e., K+ radius plus the diameter of the TIP3P water molecule in CHARMM36). While not an explicit simulation in an all-atom setting, it is nevertheless obvious that F6 in the Kv6.4 activation gate narrows the gate opening proportional to the number of Kv6.4 gates. In the 2:2R conformation, almost complete dehydration of a K+ ion would be needed for passage. Alternatively, more drastic rearrangements of the backbone than can be observed in our short vacuum simulations might reduce the constriction, but still block conduction by disrupting gating as observed for the F6 substitution in Shaker (Kitaguchi et al., 2004). Note that substitution of the Kv2.1 activation gate into Kv6.4 in the 2:2R simulation restores both the hydrophobic vapor lock of the closed state (E and I) and a wide conduction pathway to the open state (M).
	Figure 11. Model for the evolution of Shaker-like Kv regulatory subunits adopting a 3:1R functional stoichiometry. (A) A precondition for the evolution of the regulatory phenotype in Shaker-like Kv family channels is coexpression of two subunits (A and B) from the same subfamily (Kv1, Kv2, Kv3, or Kv4) in at least a subset of cells. Each subunit forms homomeric channels, and because they have cross-compatibility, they can also form heterotetramers in four possible stoichiometries (3A:1B, 2A:2B adjacent, 2A:2B diagonal, and 1A:3B). All channels formed are functional, indicated by the presence of a K+ ion (black circle) in the pore. (B) Model 1, our favored model, provides a two-step path to evolution of the regulatory phenotype (in subunit B in this example) functioning in a 3:1R stoichiometry. In step 1, mutation(s) in subunit B establish self-incompatibility (shown here as mutations in T1, star), but do not eliminate cross-compatibility with subunit A. This restricts channel assembly to three possible stoichiometries (4A, 3A:1B, 2A:2B, diagonal); all assembled channels are functional. Expression of subunit B increases the number of channels formed (relative to expression of subunit A alone) and is therefore critical for maintaining currents at or near their premutation starting levels. Step 1 effectively established the regulatory phenotype for subunit B. In step 2 of the model, subunit B accumulates gate mutation(s) tolerated in only a single subunit (star), disrupting 2:2R assembly, conduction, or gating. This establishes 3:1R as the single functional heteromer stoichiometry. Loss of function in 2:2R channels is depicted by replacement of the K+ ion with a black X. Dominant-negative suppression of current will occur at highly subunit B–biased expression ratios as few A-A contacts will form during assembly, but reasonable currents will remain with balanced expression. (C) In the alternative model 2, subunit B gate incompatibility (star) evolves first, while self-compatibility remains. This results in strong dominant negative suppression because most subunits are tied up in nonfunctional channels. Model 2 is likely an evolutionary dead end because the gate mutation will engage negative selection to preserve current levels.

Anishkin, Hydration properties of mechanosensitive channel pores define the energetics of gating. 2010, Pubmed

Anishkin, Hydration properties of mechanosensitive channel pores define the energetics of gating. 2010, Pubmed
Beckstein, Liquid-vapor oscillations of water in hydrophobic nanopores. 2003, Pubmed
Bocksteins, The subfamily-specific interaction between Kv2.1 and Kv6.4 subunits is determined by interactions between the N- and C-termini. 2014, Pubmed
Bocksteins, Independent movement of the voltage sensors in KV2.1/KV6.4 heterotetramers. 2017, Pubmed
Carrasquillo, IA channels: diverse regulatory mechanisms. 2014, Pubmed
Chen, Molecular counting by photobleaching in protein complexes with many subunits: best practices and application to the cellulose synthesis complex. 2014, Pubmed
Clancy, KCNE1 and KCNE3 beta-subunits regulate membrane surface expression of Kv12.2 K(+) channels in vitro and form a tripartite complex in vivo. 2009, Pubmed , Xenbase
Coste, Piezo proteins are pore-forming subunits of mechanically activated channels. 2012, Pubmed
Covarrubias, Shaker, Shal, Shab, and Shaw express independent K+ current systems. 1991, Pubmed , Xenbase
Crooks, WebLogo: a sequence logo generator. 2004, Pubmed
del Camino, Status of the intracellular gate in the activated-not-open state of shaker K+ channels. 2005, Pubmed , Xenbase
del Camino, Tight steric closure at the intracellular activation gate of a voltage-gated K(+) channel. 2001, Pubmed
Dodson, Two heteromeric Kv1 potassium channels differentially regulate action potential firing. 2002, Pubmed
Du, Frequency-dependent regulation of rat hippocampal somato-dendritic excitability by the K+ channel subunit Kv2.1. 2000, Pubmed
Eswar, Comparative protein structure modeling using MODELLER. 2007, Pubmed
Hackos, Scanning the intracellular S6 activation gate in the shaker K+ channel. 2002, Pubmed , Xenbase
Humphrey, VMD: visual molecular dynamics. 1996, Pubmed
Jegla, A novel subunit for shal K+ channels radically alters activation and inactivation. 1997, Pubmed
Jegla, Expanded functional diversity of shaker K(+) channels in cnidarians is driven by gene expansion. 2012, Pubmed
Jegla, Multiple Shaker potassium channels in a primitive metazoan. 1995, Pubmed , Xenbase
Jegla, Molecular evolution of K+ channels in primitive eukaryotes. 1994, Pubmed
Jiang, X-ray structure of a voltage-dependent K+ channel. 2003, Pubmed
Kerschensteiner, Fluorescence measurements reveal stoichiometry of K+ channels formed by modulatory and delayed rectifier alpha-subunits. 2005, Pubmed
King, A unique ion channel clustering domain on the axon initial segment of mammalian neurons. 2014, Pubmed
Kitaguchi, Stabilizing the closed S6 gate in the Shaker Kv channel through modification of a hydrophobic seal. 2004, Pubmed , Xenbase
Kramer, Modulation of potassium channel gating by coexpression of Kv2.1 with regulatory Kv5.1 or Kv6.1 alpha-subunits. 1998, Pubmed , Xenbase
Kreusch, Crystal structure of the tetramerization domain of the Shaker potassium channel. 1998, Pubmed
Krovetz, Atomic distance estimates from disulfides and high-affinity metal-binding sites in a K+ channel pore. 1997, Pubmed
Kumar, MEGA7: Molecular Evolutionary Genetics Analysis Version 7.0 for Bigger Datasets. 2016, Pubmed
Lau, Impaired fast-spiking, suppressed cortical inhibition, and increased susceptibility to seizures in mice lacking Kv3.2 K+ channel proteins. 2000, Pubmed
Li, Major diversification of voltage-gated K+ channels occurred in ancestral parahoxozoans. 2015, Pubmed , Xenbase
Li, Ether-à-go-go family voltage-gated K+ channels evolved in an ancestral metazoan and functionally diversified in a cnidarian-bilaterian ancestor. 2015, Pubmed , Xenbase
Li, Bimodal regulation of an Elk subfamily K+ channel by phosphatidylinositol 4,5-bisphosphate. 2015, Pubmed
Lien, Kv3 potassium conductance is necessary and kinetically optimized for high-frequency action potential generation in hippocampal interneurons. 2003, Pubmed
Liu, Gated access to the pore of a voltage-dependent K+ channel. 1997, Pubmed
Long, Crystal structure of a mammalian voltage-dependent Shaker family K+ channel. 2005, Pubmed
Long, Voltage sensor of Kv1.2: structural basis of electromechanical coupling. 2005, Pubmed
MacKerell, All-atom empirical potential for molecular modeling and dynamics studies of proteins. 1998, Pubmed
MacKinnon, Determination of the subunit stoichiometry of a voltage-activated potassium channel. 1991, Pubmed , Xenbase
Malin, Delayed rectifier K+ currents, IK, are encoded by Kv2 alpha-subunits and regulate tonic firing in mammalian sympathetic neurons. 2002, Pubmed
Malin, Elimination of the fast transient in superior cervical ganglion neurons with expression of KV4.2W362F: molecular dissection of IA. 2000, Pubmed
Malin, Molecular heterogeneity of the voltage-gated fast transient outward K+ current, I(Af), in mammalian neurons. 2001, Pubmed
Martinson, Functional evolution of Erg potassium channel gating reveals an ancient origin for IKr. 2014, Pubmed , Xenbase
Martí-Renom, Comparative protein structure modeling of genes and genomes. 2000, Pubmed
Misonou, Kv2.1: a voltage-gated k+ channel critical to dynamic control of neuronal excitability. 2005, Pubmed
Mohapatra, Interdomain cytoplasmic interactions govern the intracellular trafficking, gating, and modulation of the Kv2.1 channel. 2008, Pubmed
Murakoshi, Identification of the Kv2.1 K+ channel as a major component of the delayed rectifier K+ current in rat hippocampal neurons. 1999, Pubmed
Nakajo, Stoichiometry of the KCNQ1 - KCNE1 ion channel complex. 2010, Pubmed , Xenbase
Ogawa, Postsynaptic density-93 clusters Kv1 channels at axon initial segments independently of Caspr2. 2008, Pubmed
Ottschytsch, Domain analysis of Kv6.3, an electrically silent channel. 2005, Pubmed
Ottschytsch, Obligatory heterotetramerization of three previously uncharacterized Kv channel alpha-subunits identified in the human genome. 2002, Pubmed
Patel, Kv2.1/Kv9.3, a novel ATP-dependent delayed-rectifier K+ channel in oxygen-sensitive pulmonary artery myocytes. 1997, Pubmed
Pathak, Closing in on the resting state of the Shaker K(+) channel. 2007, Pubmed
Pau, Crystal structure of an inactivated mutant mammalian voltage-gated K+ channel. 2017, Pubmed
Pfaffinger, Shaker K+ channel T1 domain self-tetramerizes to a stable structure. 1995, Pubmed
Phillips, Scalable molecular dynamics with NAMD. 2005, Pubmed
Post, Kv2.1 and electrically silent Kv6.1 potassium channel subunits combine and express a novel current. 1996, Pubmed , Xenbase
Robinson, Coupled tertiary folding and oligomerization of the T1 domain of Kv channels. 2005, Pubmed , Xenbase
Rosenthal, Molecular identification of SqKv1A. A candidate for the delayed rectifier K channel in squid giant axon. 1996, Pubmed , Xenbase
Roth, Bubbles, gating, and anesthetics in ion channels. 2008, Pubmed
Rudy, Kv3 channels: voltage-gated K+ channels designed for high-frequency repetitive firing. 2001, Pubmed
Sali, Derivation of rules for comparative protein modeling from a database of protein structure alignments. 1994, Pubmed
Salinas, Modes of regulation of shab K+ channel activity by the Kv8.1 subunit. 1997, Pubmed , Xenbase
Salinas, New modulatory alpha subunits for mammalian Shab K+ channels. 1997, Pubmed
Sand, jShaw1, a low-threshold, fast-activating K(v)3 from the hydrozoan jellyfish Polyorchis penicillatus. 2011, Pubmed , Xenbase
Shen, Molecular recognition and assembly sequences involved in the subfamily-specific assembly of voltage-gated K+ channel subunit proteins. 1995, Pubmed
Sukhareva, Constitutive activation of the Shaker Kv channel. 2003, Pubmed , Xenbase
Trimmer, Subcellular localization of K+ channels in mammalian brain neurons: remarkable precision in the midst of extraordinary complexity. 2015, Pubmed
Tsunoda, Genetic analysis of Drosophila neurons: Shal, Shaw, and Shab encode most embryonic potassium currents. 1995, Pubmed
Tsunoda, The major delayed rectifier in both Drosophila neurons and muscle is encoded by Shab. 1995, Pubmed
Tu, Evidence for dimerization of dimers in K+ channel assembly. 1999, Pubmed , Xenbase
Ulbrich, Subunit counting in membrane-bound proteins. 2007, Pubmed , Xenbase
Upadhyay, Nicotinamide is an endogenous agonist for a C. elegans TRPV OSM-9 and OCR-4 channel. 2016, Pubmed , Xenbase
Wang, Heteromultimeric K+ channels in terminal and juxtaparanodal regions of neurons. 1993, Pubmed , Xenbase
Wang, Contribution of the Kv3.1 potassium channel to high-frequency firing in mouse auditory neurones. 1998, Pubmed
Wang, Two-stage PCR protocol allowing introduction of multiple mutations, deletions and insertions using QuikChange Site-Directed Mutagenesis. 1999, Pubmed
Waterhouse, Jalview Version 2--a multiple sequence alignment editor and analysis workbench. 2009, Pubmed
Webster, Intracellular gate opening in Shaker K+ channels defined by high-affinity metal bridges. 2004, Pubmed
Wei, K+ current diversity is produced by an extended gene family conserved in Drosophila and mouse. 1990, Pubmed , Xenbase
Weitz, Subunit stoichiometry of the CNG channel of rod photoreceptors. 2002, Pubmed
Xu, Assembly of voltage-gated potassium channels. Conserved hydrophilic motifs determine subfamily-specific interactions between the alpha-subunits. 1995, Pubmed
Xue, Dominant-negative suppression of HCN1- and HCN2-encoded pacemaker currents by an engineered HCN1 construct: insights into structure-function relationships and multimerization. 2002, Pubmed , Xenbase
Zhang, An enhanced green fluorescent protein allows sensitive detection of gene transfer in mammalian cells. 1996, Pubmed
Zheng, Rod cyclic nucleotide-gated channels have a stoichiometry of three CNGA1 subunits and one CNGB1 subunit. 2002, Pubmed , Xenbase
Zhong, The heteromeric cyclic nucleotide-gated channel adopts a 3A:1B stoichiometry. 2002, Pubmed
Zhu, Structural and functional characterization of Kv6.2 a new gamma-subunit of voltage-gated potassium channel. 1999, Pubmed