Prince-Carter A , Pfaffinger PJ .

P01NS37444 NINDS NIH HHS , R01NS31583 NINDS NIH HHS , R01 GM090029-01 NIGMS NIH HHS , R01 GM090029-02 NIGMS NIH HHS , R01 GM090029-03 NIGMS NIH HHS , R01 GM090029 NIGMS NIH HHS , P30 HD024064 NICHD NIH HHS , P01 NS037444 NINDS NIH HHS , R01 NS031583 NINDS NIH HHS

	Figure 1. N-type inactivation gating models. Stable terminal states boxed: green, hyperpolarized; blue, depolarized. Rate-limiting transitions in blue with key rate-limiting directional reactions in red. (A) General gating cycle where depolarization gates the formation of the opened state from which the free N terminus can bind to the channel to block it. The number of steps between these two states is unknown, but at some point in the chain the rate-limiting transition occurs for the ON and OFF reactions. At hyperpolarized potentials, the channel gates from the open blocked state back to the closed free state. The rate-limiting step for recovery occurs somewhere along this chain and may or may not be identical to the rate-liming OFF step at depolarized potentials. (B) Single-step inactivation model. Model incorporates a single block ON and recovery transition after channel voltage–dependent gating. Voltage dependence for inactivation is due to the voltage-dependent formation of the open state. (C) Preinactivation model. Model is similar to the single-step inactivation model, except an intermediate preinactivated state is proposed to occur between open and inactivated. Model proposes that the formation and breakdown of this preinactivated state are rate limiting for both inactivation and recovery.
	Figure 2. Homology relationships between different Shaker channel N-type inactivation domains. (A) Alignments suggest the partitioning of N terminus into three distinct regions: a hydrophobic latch region, a glycine-rich flex region, and a charged polar region. Specific sequence conservation between N termini is only seen in the latch and flex regions (see Consensus; h, a hydrophobic residue; p, a polar residue). The number below the consensus indicates the number of the six sequences that matches the consensus. Percent identity with consensus ranges from 100% for AKv1 to 47% for Kvβ1.1 (%). Structural prediction for the consensus is given: C, coil; E, extended; t, β turn. Although polar region sequence conservation is low, there is a general charge trend, where this region is predominantly positively charged close to the N terminus but becomes more negative near the T1 domain at the C terminus. Positive charges are not typically found in the latch and flex regions, which typically have a net negative charge. Net charge of N termini varies widely from +5 to −2; however, the general pattern of a more positive N-terminal part of the polar region remains. (B) Structural predictions for selected N termini. General trend for a latch region extended structure and a flex region β turn is present in all N termini but (7–15)Ala, which predicts an α helix at the beginning of the flex region. Turn type code: 1, type I; 2, type II; 3, type VIII; 4, type I' ; 5, type II'; 6, type IV (Fuchs and Alix, 2005).
	Figure 3. Scanning mutagenesis identifies regions of the N terminus that are important during different phases of N-type inactivation. (A) Wild-type AKv1 N terminus sequences and sequences of scanning mutants made. (B) Polar region scanning mutations slow inactivation with the largest effect produced by (16–25)Ala. (C) Recovery of polar region mutants from inactivation is only slightly different from wild-type, typically showing a slight acceleration. (D) Latch region mutant EVA(2–4)ATT inactivates rapidly but incompletely, resulting in a large sustained current that is not seen in the wild-type channel. (E) Representative experiment showing that recovery from inactivation for EVA(2–4)ATT is accelerated by ∼10 times compared with wild type, as expected for a destabilized binding to the channel pore. (F) Flex region mutation (7–15)Ala produces a pronounced two-exponential inactivation not seen with other mutations. Inset shows inactivation of (7–15)Ala during a 5-s depolarization to +50 mV, emphasizing the slower inactivation kinetic. (G) Recovery of (7–15)Ala from the N-type–inactivated state is dramatically slowed with a time constant ∼14 times slower than wild type.
	Figure 4. Linear energy analysis of N-terminal mutants predicts differential consolidation into inactivated structure. Table I values for energetic effects of mutations on τon and Keq are plotted. Linear energy plot shows that polar region mutations have their primary effect early in the inactivation cycle because they plot near the Φ =1 line (Black). (7–15)Ala channels inactivating with the fast kinetic plot near the Φ = 0 line (red), suggesting that a later step in the inactivation process is impacted by this mutation in these channels. (7–15)Ala channels inactivating with a slow kinetic likely reflect channels that do not reach transition normally because they plot outside the normal, grayed out region of the linear energy plot. There is almost no effect on Keq, suggesting a similar impact of this mutation on both ON and recovery reactions. EVA(2–4)ATT also plots outside the normal region of the linear energy plot but shifts to the right, indicating a greater impact on recovery, suggesting a later impact of this mutation in addition to its effect on the intermediate transition reaction.
	Figure 5. Functional effects of changing charges in the 16–25 segment of the polar region. Charged residues R18 and D19 were systematically mutated from (positive, zero, negative) charge to determine the magnitude of the electrostatic potential changes these residues experience during N-type inactivation. (A) Inactivation slows as position 18 is made more negative from its normal positive charge. (B) Recovery from inactivation shows little sensitivity to charge at position 18. (C) Inactivation accelerates slightly as position 19 is made more positive from its normal negative charge. (D) Recovery shows a greater sensitivity to charge at position 19, becoming slower as position 19 is made more positive. The results suggest that both positions 18 and 19 experience a negative potential change between the free and bound states, but the magnitude and phase vary between the two residues.
	Figure 6. Inactivation reaction energetics predict that a negative electrostatic potential change is experienced by residues 18 and 19 during N-type inactivation. Energy state diagrams are constructed based on the effects of charge-changing mutations at residues 18 and 19 on the ON and recovery reactions. Free-state energy level is assumed to be unaffected by mutations and is set to 0. (A) Progressively higher transition-state energy as position 18 is made more negative suggests a negative electric field experienced by this residue during inactivation. There is no significant further effect between transition and the bound state. (B) Change in energy plotted versus the charge at position 18. Energy change is linear as expected for an electrostatic effect, with the slope of the line giving the potential change experienced by position 18 from the free state to either the transition or bound states. (C) Residue 19 mutations show a greater impact on recovery than inactivation, unlike other polar region mutations. Energy levels increase for more negative charges at residue 19, again suggesting a negative potential change during inactivation. (D) The effects of charge-changing mutations at residue 19 are linear for both inactivation and recovery, suggesting that D19 is primarily interacting electrostatically with the channel core. The net potential change experienced by D19 during binding is less than half that experienced by R18, suggesting significant local heterogeneity in experienced potential changes. The late effect of residue 19 mutations, after the ON transition state, may indicate that this residue is pointed away from the channel polar region binding surface and thus is not directly involved in polar region binding to the channel, but rather primarily affects how the latch and flex regions enter and exist from the pore block site.
	Figure 7. Electrostatic interactions occur throughout the polar region. (A) Location of charged residue within the polar region of the AKv1 N terminus. (B) Neutralization of positive charges by mutation to alanine produces a slowing of inactivation that becomes progressively bigger as the mutations are made closer to the N terminus. (C) Neutralizing positive charges has little effect on recovery from inactivation. (D) Estimated potential change experienced by these residues upon binding. Largest change is an ∼30-mV change for position 18. Slow decay in potential (e-fold per 16.6 residues) is less than what would be predicted for a single site acting from a region near the pore on an unstructured chain with a Debye length of ∼9 à for frog Ringer. (E) No significant effect of position on potential change experienced during recovery suggests that all the potential change experienced during binding occurs after the transition state for recovery. (F) A series of mutations to normally negative residues in the polar region of the N terminus. Slope of line indicates the sign and strength of the electric field change experienced at this site between the free and bound states. Biggest positive potential change is experienced by residue 47, with only residue 19 experiencing a large negative potential change. (G) Summary plot showing the potential changes experienced by charges in the polar region during N-type inactivation. Results show a strong trend with more negative potentials experienced in regions dominated by positively charged residues. D19 experiences a strong negative potential, but this potential is in a local minimum compared with the adjacent R18 and R26.
	Figure 8. Mutant cycle analysis between polar region residues and channel charges located near side windows. 161–3 is located at the periphery of the side windows, whereas 135 is located with the side windows at the axis of symmetry. We examined the impact of interactions between these channel residues and the polar region sites on the rate-limiting transitions for inactivation ON and recovery reactions. (A) The interaction between residues 18 and 161–3 during the ON process is the only one significantly different from all the others (indicated by *). Adjacent residue 19 has no significant interaction with 161–3, suggesting that the interaction with position 18 is specific and that residue 19 likely is pointed away from 161–3. (B) No residue shows a strong specific interaction with position 135. Coupling is greatest to D19 during the ON process, and it seems to fall off steadily with distance, suggestive of a diffuse electrostatic effect. Failure to see specific coupling suggests that the polar region does not reach far into the side windows during inactivation, but that D19 may be oriented toward the side window opening.
	Figure 9. Analysis of the importance of electrostatics for interactions between residue 18 and the side window residues 161–3 and 135. Charge-changing mutations were made in both the side window location and in residue 18, and the energetic effect on the rate-limiting inactivation ON and recovery reactions was determined. (A) Changing residue 18 charge alters the energetics of the inactivation on reaction in a linear manner. The slope changes depending on the charge present at residues 161–3. As 161–3 is made more positive, the energetic impact of changing residue 18 charge is lost, indicating a large electrostatic contribution of 161–3 on the potential change felt by residue 18. (B) There is little impact on recovery of charge changes to residue 18, and the slopes of these lines do not change much with changes in the charge at residues 161–3. (C) Charge at residue 135 has a similar effect on slope as residue 161–3, but the impact is less pronounced. Because residue135 is at the axis of symmetry, four net charges in the center of the side windows are changed with each residue 135 mutation; however, the number of these charges that are seen by residue 18 is unknown. (D) As expected, energetic coupling shows little charge dependence for residue 18 and 135 between the bound and transition state during recovery. (E) Summary diagram for coupling between residue 18 and 161–3. As expected, there is no significant change in electrostatic interaction between residue 18 and 161–3 during the recovery process, but there is a large potential change produced during inactivation as the polar region binds. The potential produced by EDE(161–3) is ∼−15 mV, about half the total potential change seen by residue 18. (F) Coupling between residue 18 and 135 shows a similar pattern with no change in interaction during recovery, but an estimated −8.5-mV change on binding (with V135D). Because curves are almost perfectly linear, it suggests that there is no significant non-electrostatic interaction between residue 18 and residue 135. Because residue 135 is normally uncharged, this interaction has no impact on inactivation in the wild-type channel.
	Figure 10. Linear energy plot for all mutations made to polar region–charged residues. In Tables II and IV, values for energetic effects of polar region mutations on τon and Keq are plotted. Plot shows the change in on and equilibrium energy compared with the base construct for every mutation we constructed in the polar region of the AKv1 N terminus. Mutations cluster tightly near the Φ =1 line with a slope of 0.91, indicating that as a group, polar region mutations primarily impact a step early in the inactivation process and do not alter the energetics for recovery. Residue 19 mutations may be plotting along a shallower line, indicating that they also affect a later step in the inactivation process.
	Figure 11. Charge at residue 161–3, but not at residue 18, affects voltage dependence for activation. (A) Activation curves for channels with different charges at residues 161–3 show a strong shift in activation with changes in charge at this site. Activation curves fit with fourth-power Boltzmann functions. (B) Plot of midpoint for subunit activation taken from fourth-power Boltzmann fits to the activation data for different charge substitutions at residue 161–3. Linear plot shows a clear electrostatic effect with an ∼−9-mV shift in activation for neutral to positive change at these residues. (C) Effect of changing residue 18 charge on inactivation produced by 161–3(KKK). Results show that the activation curves with different residue 18 charges stack on top of one another. (D) Plot of midpoint for subunit activation shows no effect of residue 18 charge on channel activation.
	Figure 12. Tail currents for EVA(2–4)ATT mutant show all N termini have progressed beyond the inactivation rate-limiting step. Currents and voltage commands for tail current recordings for AKv1(EVA(2–4)ATT) and AKv1(Δ2-57) mutants. Pulse potentials given in mV. Zero current potentials are −5 mV. (A) EVA(2–4)ATT mutant has surprisingly large tail currents given the relatively small amount of current decay. Entire tail current is well fit with a single-exponential function with a tau closely matching the time constant for recovery from inactivation. (B) Tail currents for AKv1(EVA(2–4)ATT) and AKv1(Δ2-57) are aligned and scaled. In the absence of an N-terminal inactivation domain, AKv1(Δ2-57) channel tails decay twice as fast as AKv1(EVA(2–4)ATT) tails. (C) AKv1(EVA(2–4)ATT) tail currents are driving force normalized to plot versus the amplitude of the current at +50 mV. Single-exponential component tails are similar in size to the sustained component at +50 mV.
	Figure 13. Summary of the polar region binding energetics. Sequence of AKv1 N-terminal polar region shown with charged residues highlighted. Arrows point toward the estimated equilibrium binding energy provided by the charged residue, in kT. Total estimated electrostatic equilibrium binding energy for all charged residues, ΔEo, is around −4 kT. Poly-Ala scanning regions are shown by blue bars, with estimated binding energies for the probed regions provided below the bar, in kT. For the 26–45 region, binding energy is estimated from the charged residue contributions. The total binding energy contributed by the polar region, ΔGo, is ∼−6.5 kT. Most of the missing −2.5 kT comes from non-electrostatic interactions occurring in the 16–25 region.
	Figure 14. Structural models of the AKv1 channel identifying a potential polar region binding site on the channel surface. (A) Poisson-Boltzmann calculations of electric fields (isopotential surfaces: red, −1 kT/z; blue, +1 kT/z) emanating from the surface of the AKv1 protein (left) compared with an EM reconstruction of a Kv1 channel on the right (Sokolova et al., 2001) showing the side windows that provide access to the pore block site. The grayed-out region is assumed to be inaccessible to the N terminus due to the lipid bilayer. Star indicates the N-terminal end of the T1 domain where the N-type inactivation domain attaches to the channel. White arrow traces a possible path to the side window that shows a strong increase in negative potential as the arrow approaches the side window. (B) ±1-kT/z electrostatic surfaces for the AKv1 N terminus based on the Kv1.4 N terminus model (Wissmann et al., 2003). Flex and latch regions are not explicitly modeled due to flexibility in the Kv1.4 N-terminal model in this region. Bottom surface of the model (Underneath View) shows the presumed polar region interaction surface. The model shows how the polar region might expose a progression of positive charges to one side of an α-helical structure with D19 pointed in the opposite direction. (C) N terminus–T1 domain interaction model constructed by overlaying R18 and EDE(161–3) along the track proposed in A. Electric Surface: ±1-kT/z electrostatic surfaces for the channel Poisson-Boltzmann model focusing on the T1 domain. Peptide Footprint: Underneath N-terminal model in B is flipped horizontally to show how the exposed polar region electrostatic surface might interact with the T1 domain. Binding Site: Outline of the Peptide Footprint showing the T1 domain electrostatic potentials that underlie the proposed polar region binding site. Although the precise binding site is not known, this analysis indicates that a region of negative charge on the T1 domain surface complements the positively charged side of the polar region model. In addition, D19 points away from 161–3 toward a region of locally positive potential.
	Figure 15. A multistate N-type inactivation model. Model shows gating cycle driven by depolarization and hyperpolarization. Blue, rate-controlling reactions; red, unidirectional reactions; black, reactions in rapid equilibrium. Stable state at depolarized potential is boxed in blue, and at hyperpolarized potentials is boxed in green. Depolarization triggers gating charge movement, which opens the channel (α(+V)) and exposes the polar region binding site. After polar region binding (PB) to the open state, the channel is able to overcome the rate-limiting step to inactivate. Formation of the PB intermediate is a key step in inactivation. At least two additional states exist after the rate limiting on transition is completed, a pre-block (PreBl) and a pore block (Bl) state. Both pre-block and pore block recover as if they are inactivated, but only the pore block state fails to conduct. For the recovery reaction, after membrane hyperpolarization, the pre-block and pore block states likely recover by a rate-limiting transition from the pre-block state. The exact order of steps to recover from pre-block are not clear, but it is likely that polar region unbinding occurs very rapidly once the channel closes (β(−V)) due to the low affinity binding.

, , Pubmed

, , Pubmed
, , Pubmed , Xenbase
Aldrich, Fifty years of inactivation. 2001, Pubmed
Antz, NMR structure of inactivation gates from mammalian voltage-dependent potassium channels. 1997, Pubmed
Baker, NMR-derived dynamic aspects of N-type inactivation of a Kv channel suggest a transient interaction with the T1 domain. 2006, Pubmed
Baukrowitz, Modulation of K+ current by frequency and external [K+]: a tale of two inactivation mechanisms. 1995, Pubmed
Borys, A diffusive model of the ball and chain inactivation. 2008, Pubmed
Choi, Tetraethylammonium blockade distinguishes two inactivation mechanisms in voltage-activated K+ channels. 1991, Pubmed
Cushman, Voltage dependent activation of potassium channels is coupled to T1 domain structure. 2000, Pubmed
Decher, Structural determinants of Kvbeta1.3-induced channel inactivation: a hairpin modulated by PIP2. 2008, Pubmed
Demo, The inactivation gate of the Shaker K+ channel behaves like an open-channel blocker. 1991, Pubmed
Fersht, The folding of an enzyme. I. Theory of protein engineering analysis of stability and pathway of protein folding. 1992, Pubmed
Fuchs, High accuracy prediction of beta-turns and their types using propensities and multiple alignments. 2005, Pubmed
Giese, Modulation of excitability as a learning and memory mechanism: a molecular genetic perspective. 2001, Pubmed
Gilboa, History-dependent multiple-time-scale dynamics in a single-neuron model. 2005, Pubmed
Guex, SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. 1997, Pubmed
Gulbis, Structure of the cytoplasmic beta subunit-T1 assembly of voltage-dependent K+ channels. 2000, Pubmed , Xenbase
Hoshi, Biophysical and molecular mechanisms of Shaker potassium channel inactivation. 1990, Pubmed , Xenbase
Humphrey, VMD: visual molecular dynamics. 1996, Pubmed
Jerng, Multiprotein assembly of Kv4.2, KChIP3 and DPP10 produces ternary channel complexes with ISA-like properties. 2005, Pubmed , Xenbase
Jerng, DPP10 splice variants are localized in distinct neuronal populations and act to differentially regulate the inactivation properties of Kv4-based ion channels. 2007, Pubmed , Xenbase
Jerng, Modulation of Kv4.2 channel expression and gating by dipeptidyl peptidase 10 (DPP10). 2004, Pubmed , Xenbase
Kurata, A structural interpretation of voltage-gated potassium channel inactivation. 2006, Pubmed
Lee, N-type inactivation in the mammalian Shaker K+ channel Kv1.4. 1996, Pubmed , Xenbase
Long, Crystal structure of a mammalian voltage-dependent Shaker family K+ channel. 2005, Pubmed
Long, Atomic structure of a voltage-dependent K+ channel in a lipid membrane-like environment. 2007, Pubmed
Molina, N-type inactivation of the potassium channel KcsA by the Shaker B "ball" peptide: mapping the inactivating peptide-binding epitope. 2008, Pubmed
Murrell-Lagnado, Energetics of Shaker K channels block by inactivation peptides. 1993, Pubmed , Xenbase
Murrell-Lagnado, Interactions of amino terminal domains of Shaker K channels with a pore blocking site studied with synthetic peptides. 1993, Pubmed , Xenbase
Ono, Subfamily-specific posttranscriptional mechanism underlies K(+) channel expression in a developing neuronal blastomere. 1999, Pubmed , Xenbase
Pfaffinger, Cloning and expression of an Aplysia K+ channel and comparison with native Aplysia K+ currents. 1991, Pubmed , Xenbase
Phillips, Scalable molecular dynamics with NAMD. 2005, Pubmed
Rettig, Inactivation properties of voltage-gated K+ channels altered by presence of beta-subunit. 1994, Pubmed , Xenbase
Roberds, Cloning and tissue-specific expression of five voltage-gated potassium channel cDNAs expressed in rat heart. 1991, Pubmed
Sokolova, Three-dimensional structure of a voltage-gated potassium channel at 2.5 nm resolution. 2001, Pubmed
Wang, Genetically encoding unnatural amino acids for cellular and neuronal studies. 2007, Pubmed
Wissmann, Solution structure and function of the "tandem inactivation domain" of the neuronal A-type potassium channel Kv1.4. 2003, Pubmed
Zagotta, Restoration of inactivation in mutants of Shaker potassium channels by a peptide derived from ShB. 1990, Pubmed , Xenbase
Zhou, Potassium channel receptor site for the inactivation gate and quaternary amine inhibitors. 2001, Pubmed , Xenbase
Zhou, Phi-value analysis of a linear, sequential reaction mechanism: theory and application to ion channel gating. 2005, Pubmed