|
Figure 1. Two classes of KATP channel ATP inhibition gating mechanism make different kinetic predictions for the ATP dependence of open state duration. (A) ATP concentration decreases open-state duration and thereby burst duration. The simple ATP-dependent burst scheme makes specific quantifiable predictions about how decremental ATP destabilization of the open state necessarily will account for dramatic decreases in burst durations. (B) ATP concentration is excluded from determining open state or burst durations. ATP binding to the open state is so energetically unfavorable that there is essentially no open-state occupancy at any ATP concentration. The hypothesis we tested here is that ATP binds the open state of the KATP channel, which destabilizes it relative to the inactive interburst state, thus providing the mechanism by which increasing ATP speeds the rate of burst exit. The ATP dependence of the open times in the top model exactly depends not only on rate constant \documentclass[10pt]{article}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{pmc}
\usepackage[Euler]{upgreek}
\pagestyle{empty}
\oddsidemargin -1.0in
\begin{document}
\begin{equation*}{\mathrm{k}}_{{\mathrm{OC}}_{1}}\end{equation*}\end{document}[ATP], but also on rate constant \documentclass[10pt]{article}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{pmc}
\usepackage[Euler]{upgreek}
\pagestyle{empty}
\oddsidemargin -1.0in
\begin{document}
\begin{equation*}{\mathrm{k}}_{{\mathrm{OC}}_{{\mathrm{f}}}}\end{equation*}\end{document}. C, closed state; O, open state. Subscripts: f, fast; 0, 0 ATP ligands bound; and 1, 1 ATP ligand bound. Cf is the short-lived or fast intraburst closed state. C0 and C1 are long-lived or slow interburst closed states that differ by whether ATP is bound and thereby their mean duration.
|
|
Figure 2. ATP dependence of single KATP channel gating kinetics at 0, 0.2, and 0.6 mM ATP. (A) Effect of ATP on burst durations. (B) 10-fold expanded time scale to reveal effects of ATP on open durations. Single-channel currents from a representative inside-out patch within 30 s after patch excision continuously perfused with 0 ATP for 15 s. The patch was then exposed to 0.2 mM MgATP for 90 s, and finally 0.6 mM MgATP for 10 min. In between these [MgATP], the channel is exposed briefly to 0 MgATP (to check for rundown in activity) and to 0.6 mM MgATP to help maintain the initial high PO activity typical of the KATP channel in 0 MgATP. For each [ATP], single-channel currents are shown by using a slow time scale to emphasize burst durations reduction by increasing ATP (left) and by using a 10-fold faster time scale a segment is expanded to emphasize open durations reduction by ATP (immediately to the right).
|
|
Figure 3. Burst and open duration histograms. (A) Distribution of burst durations as a function of ATP concentration. The log-scale abscissa of the Sine-Sigworth plots results in the peak of these single-exponential distributions positioned at the value of the time constant. The dashed line is placed at the higher time constant (0 ATP) to facilitate comparison with the intermediate and lower time constants at other ATP concentrations. (B) Distribution of open times as a function of ATP concentration. Note the decrement in open time with increasing ATP concentration correlates with a dramatic decrease in the corresponding burst durations.
|
|
Figure 4. ATP has no effect on the intraburst shut times at concentrations where open times are reduced by greater than twofold. In general, the experiments showed if anything a slight increase in closed time durations with increasing ATP.
|
|
Figure 5. Evidence for ligand-independent gating transitions from the open state O to the inactive interburst C0. (A) In the absence of ATP, the Kir6.2::G334D/SUR1 mutant channel, which virtually eliminates ATP-dependent inhibition gating, exhibits gating transitions from the active burst to the inactive interburst, indicating little or no effect on ligand-gating to the inactive interburst, slow time scale. (B) Same as above except 10-fold faster time scale, emphasizing the open durations within these long bursts. (C) The distribution of burst durations for the G334D channel is single-exponential with mean duration of 39.8 ms. The reciprocal of the mean burst duration of the ATP-refractory Kir6.2::G334D/SUR1 is the first order rate constant \documentclass[10pt]{article}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{pmc}
\usepackage[Euler]{upgreek}
\pagestyle{empty}
\oddsidemargin -1.0in
\begin{document}
\begin{equation*}{\mathrm{k}}_{{\mathrm{OC}}_{0}}\end{equation*}\end{document} equal to 25.1 s−1, which is comparable to the \documentclass[10pt]{article}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{pmc}
\usepackage[Euler]{upgreek}
\pagestyle{empty}
\oddsidemargin -1.0in
\begin{document}
\begin{equation*}{\mathrm{k}}_{{\mathrm{OC}}_{0}}\end{equation*}\end{document} value 27 s−1 determined for the ATP-sensitive wild-type KATP channel. (D) The distribution of open durations for the G334D channel is single-exponential with mean duration of 1.78 ms, which is similar to the 1.68-ms value determined for the ATP-sensitive KATP channel. The G334D burst and open duration results provide strong support for the ligand-independent gating transition O to C0, and quantitatively exclude the possibility that the transition to C0 may be explained by ligand-dependent transitions driven by nominal ATP in the patch.
|
|
Figure 6. “Ligand-insensitive bursts” in the presence of 5 mM MgATP. Typically (92/97 patches) in the absence of creatine phosphate/kinase to scavenge any MgADP generated in the patch, tens and hundreds of one or two opening bursts would be accompanied by one to a few long bursts, as above. Note the occasional long bursts each with 10–100 openings with long mean durations, compared with the more frequent short bursts each with 1–2 openings with short mean durations. Thus, one long burst can have approximately the same number of openings as in 50 or more short bursts.
|
|
Figure 7. At 5 mM MgATP, long mean open times are kinetically comparable to mean open times in either 0 MgATP or 150 μM MgADP. (A) Long and short mean open duration components are clearly detectable when the KATP channel is exposed to 5 mM MgATP. The value of the long time constant is indistinguishable from the single open time component observed for the KATP channel in 0 ATP in the absence of rundown. (B) Only short open duration component is observed for the KATP channel when the creatine phosphate/kinase system is used to scavenge any MgADP generated in the patch. (C) Only long open duration component is observed for the KATP channel when only 150 μM MgADP is added. Thus, in the absence of creatine kinase and creatine phosphate, at ≪5mM MgATP, the ATP-dependent open durations become kinetically indistinguishable from the ATP-refractory, constant long open durations, which should not but might lead to confusion over the ATP dependence of KATP channel open times.
|
|
Figure 8. ATP dependence of single truncated Kir6.2ΔC26 channels without SUR at 0, 1, and 5 mM ATP. (A) Effect of ATP on open durations. Single-channel currents from a representative inside-out patch within 30 s after patch excision continuously perfused with 0 ATP for 1 min. The patch was then exposed to 1 mM MgATP for 3 min, and finally 5 mM MgATP for 15 min. In between these [MgATP], the channel is exposed briefly to 0 MgATP (to check for rundown in activity) and to 0.6 mM MgATP to help maintain the initial PO activity in 0 MgATP.
|
|
Figure 9. Open duration histograms of truncated Kir6.2ΔC26 channels without SUR at 0, 1, and 5 mM ATP. The log-scale abscissa of the Sine-Sigworth plots results in the peak of these single-exponential distributions positioned at the value of the time constant. The dashed line is placed at the higher time constant (0 ATP) to facilitate comparison with the intermediate and lower time constants at other ATP concentrations. Because the burst durations of the truncated ΔC26 channel without SUR even in 0 ATP are already very brief, the ATP dependence of burst times is not shown.
|
|
Figure 10. ATP-dependent open state mechanism for speeding burst exit, supported by single-channel kinetic data reported here. (A) Mean open time (red) or burst time (black) as a function of increasing [ATP]. The measured mean durations from individual patch experiments are shown by the position of individual symbols. The smooth curves through these symbols were generated as indicated in the text. Note that by using first order rate constants, together only with the decrementing mean open durations at increasing [ATP], we solved for the second order, ATP-dependent rate constant, \documentclass[10pt]{article}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{pmc}
\usepackage[Euler]{upgreek}
\pagestyle{empty}
\oddsidemargin -1.0in
\begin{document}
\begin{equation*}{\mathrm{k}}_{{\mathrm{OC}}_{1}}\end{equation*}\end{document}. This second order rate constant then was used to calculate the ATP dependence of not only the open, but also the burst durations shown by the smooth curves. The main plot shows that the rates determined from the decrementing open times quantitatively account for the mean open and burst duration data. The insetted plot shows the mean open duration ATP dependence with the y-axis expanded to emphasize that the mean open durations decrease with increasing ATP, as predicted by the determined rate constants. (B) The burst gating model supported by the results includes a second order, ATP-dependent rate constant, \documentclass[10pt]{article}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{pmc}
\usepackage[Euler]{upgreek}
\pagestyle{empty}
\oddsidemargin -1.0in
\begin{document}
\begin{equation*}{\mathrm{k}}_{{\mathrm{OC}}_{1}}\end{equation*}\end{document}[ATP], from the open state (O) to the interburst state bound by one ATP ligand, C1. Cf is the short-lived intraburst closed state, and C0 is the long-lived inactive interburst closed state, each with no ATP bound. The relative values of the rate constants determined indicate that about twofold decrease in mean open duration requires ∼500 μM ATP, or tens of Ki units for ATP inhibition. Note that in the new gating mechanism, the inactive interburst state C0 is not an obligatory state to which the open channel must transit before ATP is permitted to bind. The open channel either first transits to the inactive interburst (e.g., in low ATP) or first binds ATP, which speeds its transition to C1. Although beyond the scope of this study, C0 and C1 can interconvert by ATP binding and dissociation reactions where bound ATP would stabilize the closed allosteric gating conformation(s) without transitions from the single-channel closed current level. For both mean open durations and mean burst durations, differences between values at 0, 0.2, and 0.6 mM ATP were each statistically significant (P < 0.001). (C) The burst gating model for the KATP channel in the presence of 150 μM MgADP. The gating transitions are effectively directed by bound MgADP to those within the burst by reducing to little or nothing the two burst exit pathways (transition and rate constants in aqua). When MgADP is bound to the channel, presumably at SUR1, the open state is precluded from binding inhibitory ATP, and gating transitions are constrained to gating transitions within the active burst. Gating transitions to the long-lived interburst by ligand-independent gating or ATP-dependent gating are permitted only upon unbinding of the MgADP, as above.
|
