| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Biophys J, April 1999, p. 1988-2003, Vol. 76, No. 4
*Department of Physiology and Biophysics, Cornell University Medical College, New York, New York 10021, and #Department of Physiology and Biophysics, The Chicago Medical School, North Chicago, Illinois 60064 USA
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
SUMMARY
APPENDIX
REFERENCES
|
|---|
The gating characteristics of two ion channels in the
inward-rectifier K+ channel superfamily were compared at
the single-channel level. The strong inward rectifier IRK1 (Kir 2.1)
opened and closed with kinetics that were slow relative to those of the
weakly rectifying ROMK2 (Kir 1.1b). At a membrane potential of 
60 mV,
both IRK and ROMK had single-exponential open-time distributions, with mean open times of 279 ± 58 ms (n = 4) for
IRK1 and 23 ± 1 ms (n = 7) for ROMK. At 60
mV (and no EDTA) ROMK2 had two closed times: 1.3 ± 0.1 and
36 ± 3 ms (n = 7). Under the same conditions, IRK1 exhibited four discrete closed states with mean closed times of
0.8 ± 0.1 ms, 14 ± 0.6 ms, 99 ± 19 ms, and 2744 ± 640 ms (n = 4). Both the open and the three
shortest closed-time constants of IRK1 decreased monotonically with
membrane hyperpolarization. IRK1 open probability
(Po) decreased sharply with
hyperpolarization due to an increase in the frequency of long closed
events that were attributable to divalent-cation blockade. Chelation of
divalent cations with EDTA eliminated the slowest closed-time
distribution of IRK1 and blunted the hyperpolarization-dependent fall
in open probability. In contrast, ROMK2 had shorter open and closed
times and only two closed states, and its Po
was less affected by hyperpolarization. Chimeric channels were
constructed to address the question of which parts of the molecules
were responsible for the differences in kinetics. The property of
multiple closed states was conferred by the second membrane-spanning
domain (M2) of IRK. The long-lived open and closed states, including
the higher sensitivity to extracellular divalent cations, correlated
with the extracellular loop of IRK, including the "P-region."
Channel kinetics were essentially unaffected by the N- and C-termini.
The data of the present study are consistent with the idea that the
locus of gating is near the outer mouth of the pore.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
SUMMARY
APPENDIX
REFERENCES
|
|---|
Members of the inward-rectifier (Kir) family of
K+ channels share a number of basic characteristics
including high selectivity for K+ over Na+,
high Po at the normal resting potential of the
cell, and properties indicative of multi-ion occupancy of the pore
(Latorre and Miller, 1983
; Eisenman and Dani, 1987; Hille, 1992).
However, within the inward rectifier family, there are significant
differences in the single-channel conductance and in the degree of
rectification (Jan and Jan, 1997; Nichols and Lopatin, 1997). The
channels also exhibit a wide range of gating characteristics. In this
paper we focus on the differences in the opening and closing rates of two prototypical inward rectifier K+ channels. These are
ROMK2 (Kir1.1b), a weakly rectifying channel found mainly in the kidney
(Zhou et al., 1994), and IRK1 (Kir2.1), a strongly rectifying channel
cloned from J774, a macrophage cell line (Kubo et al., 1993).
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
SUMMARY
APPENDIX
REFERENCES
|
|---|
Construction of chimeras
Chimeras were constructed using the splicing by overlap
extension method (Horton et al., 1989). Overall schemes for each
chimera are summarized in Table 1.
Briefly, after the synthesis of fragment 12 and fragment 34 using
polymerase chain reaction (PCR), a second PCR was conducted with primer
1, primer 4, and two fragments. The PCR product was cut with the
restriction enzymes 1 and 2 to replace the corresponding parts of ROMK2
(GenBank accession No. L29403) or IRK1 (GenBank accession No. X73052).
Primers (Table 2) were synthesized by
Operon Technologies, Inc. (Ala-meda, CA). Nucleotide sequences
between two restriction enzyme sites were checked using an ABI 377 automated DNA sequencer at The Rockefeller University DNA Technology
Center (New York, NY).
|
|
Expression of channels
Plasmids were linearized with Not I restriction enzyme and
transcribed in vitro with T7 RNA polymerase in the presence of the
GpppG cap using the mMESSAGE mMACHINE kit (Ambion, Austin, TX).
Synthetic cRNA was dissolved in water and stored at 70°C before
use. Stage V-VI oocytes were obtained by partial ovariectomy of female
Xenopus laevis (NASCO, Ft. Atkinson, WI), anesthetized with
tricaine methanesulfonate (1.5 g/l, adjusted to pH 7.0). Oocytes were
defolliculated by incubation in OR2 solution (82.5 mM NaCl, 2 mM KCl, 1 mM MgCl2, and 5 mM HEPES, adjusted to pH 7.5 with NaOH)
containing 2 mg/ml collagenase type II and 2 mg/ml hyaluronidase type
II (Sigma Chemical, St. Louis, MO) for 90 min and (if necessary)
another 90 min in a fresh enzyme solution at 23°C. Oocytes were
injected with 0.5-1 ng of cRNA and incubated at 19°C in 2× diluted
Leibovitz medium (Life Technologies, Grand Island, NY) for 1 to 4 days
before measurements were made. For patch-clamp experiments, oocytes
were subjected to a hypertonic shrinking solution containing 200 mM
sucrose, thereby allowing the vitelline membrane to be easily removed.
Electrophysiology
Oocytes were bathed in a solution containing (in mM): KCl (110),
CaCl2 (2), MgCl2 (1), and HEPES (5) at pH 7.4. Patch-clamp pipettes were pulled from #7052 borosilicate glass
(Richland Glass Co., Richland, NJ) using a three-stage process and were
coated with Sylgard. They were filled with solutions containing (in
mM): KCl (110) and HEPES (5) at pH 7.4. In some cases the solution also
contained 5 mM EDTA. In one set of experiments KCl was replaced by 73 mM K2SO4. Pipette resistances ranged from 1 to
3 M
. Currents were recorded with either a List EPC-7 or a Dagan 8900 patch-clamp amplifier and stored, unfiltered, on videotape. For
off-line analysis, current records were replayed from videotape,
filtered at 1 kHz, and sampled at 5 kHz, using an Atari-based data
acquisition system (Instrutech, Mineola, NY). Construction of open- and
closed-time histograms and fitting with exponential distributions were
carried out using the data analysis TAC program (Sigworth and Sine,
1987). When selected records were sampled at 25 kHz rather than 5 kHz, no significant difference was found in the locations of peaks on the
closed-time histograms. Hence, data were routinely sampled at 5 kHz to
conserve disk space.
Data analysis
The voltage dependence of the open probability
(Po) was fit by the Boltzmann function:
|
(1) |
The smooth curves in Figs. 2, 3, 7, 8, and 9 were drawn using the equations and fitted parameters enumerated in the Appendix.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
SUMMARY
APPENDIX
REFERENCES
|
|---|
Kinetic properties of IRK1
The basic gating characteristics of IRK1 at three different
membrane potentials are shown in Fig. 1.
Both openings and closures tended to be long-lived, as previously
reported (Kubo et al., 1993). At a membrane potential of 100 mV
(oocyte negative) the duration histogram of Fig. 1 B
indicates that the open time is well described by a single-exponential
distribution, whereas the closed times required at least four separate
exponentials to fit the distribution well (Figs. 1 C and 6
A). Maximum likelihood analysis indicated that the fit with
four exponentials was significantly better than the fit obtained using
only three exponentials.
|
The voltage dependence of the kinetic parameters is illustrated in Fig.
2. The open probability
(Po) of IRK1 declined steeply as the voltage was
changed from 60 mV to 150 mV (Fig. 2 A). This is in
marked contrast to ROMK2, which exhibits little voltage dependence of
open probability (dashed line, Fig. 2 A). Over
the same voltage range, the mean open time of IRK1 decreased
exponentially, declining e-fold for a hyperpolarization of 64 mV
(solid line, Fig. 2 B). The closed-time constants
of IRK1 also declined with hyperpolarization, although the longest
closed time was much less sensitive to voltage than were the shorter
closed times. The curves in Fig. 2 C were fit to the data
according to Eqs. A2 and A3, given in the Appendix. Although the
duration of the longest closures was not strongly affected by voltage,
the number of such closures increased with hyperpolarization. This
largely accounted for the substantial voltage dependence of IRK1
Po.
|
Because native inward rectifiers are known to be sensitive to
extracellular divalent cations (Hagiwara et al., 1978; Sakmann and
Trube, 1984; Ho et al., 1993; Kubo et al., 1993; Elam and Lansman,
1995; Robertson et al., 1996), and because the kinetics of the cloned
ROMK2 channel were altered by chelation of divalent cations (Choe et
al., 1998), we examined the effects of adding 5 mM EDTA to the pipette
solution on the gating of IRK1. As shown in Fig.
3, long-lived closures that were
prevalent under control conditions were virtually abolished by EDTA.
The Po-V curves associated with these
experiments are shown in Fig. 3 B. The results of this figure strongly suggest that the reduction in Po
is a consequence of block by divalent cations, even though none were
added to the control pipette solution.
|
To investigate the possibility that the voltage-dependent reduction in
Po arose specifically from block by trace
amounts of Ba2+, we repeated the measurements using 73 mM
K2SO4 in the pipette solution. Since the
solubility product of BaSO4 is 1.05 × 10
10 (Weast and Selby, 1966), the ionized
Ba2+ concentration would be <2 nM under these conditions.
As shown in Fig. 3 B, complexation of Ba2+
reduced, but did not eliminate, the fall in Po
at hyperpolarizing voltages. This is consistent with reports that trace
amounts of Ba+2 affect the gating of calcium-activated
K+ channels (Diaz et al., 1996) and ROMK2 channels (Choe et
al., 1998).
Comparison of Figs. 3 C and 1 B illustrate that
addition of EDTA has a relatively small effect on the mean open time of
IRK1. With 5 mM EDTA in the pipette solution, the mean open time was 270 ± 26 ms (n = 5) at 100 mV (compared to the
value of 131 ms for the experiment of Fig. 1). However, chelation
of divalents virtually eliminated the longest closed time (compare
Figs. 3 D and 1 C). In some experiments, the
shortest closed time also disappeared (or became too short to measure)
under these conditions.
Comparison with ROMK gating
Although IRK1 and ROMK2 show many structural similarities, their
gating characteristics differ significantly (Chepilko et al., 1995;
Choe et al., 1998). Compared to ROMK, IRK1 channels have much longer
mean open times (300 ms vs. 20 ms), more resolvable closed states (four
versus two under control conditions) and a stronger voltage dependence
(see Fig. 2 A).
A major goal of this study was to investigate which aspects of the ROMK and IRK structures are responsible for these differences. To address this question we divided the primary amino acid sequence of both channels into seven different sections based on hydropathy analysis and sequence homology (Fig. 4). These are 1) the cytoplasmic N-terminal; 2) the first predicted transmembrane domain (M1); 3) the extracellular region from M1 to the P-region (MP); 4) the P-region itself containing the GYG K+-channel signature sequence; 5) the remainder of the extracellular region from the P-region to the second predicted membrane spanning domain (PM); 6) the second predicted membrane spanning domain itself (M2); and 7) the cytoplasmic C-terminal.
|
The alignment of several of these regions in IRK1 and ROMK2 is shown in
Fig. 4 B. Clearly the strongest homology is in the P-region,
where 15 of 17 amino acids are identical. Significant homology also
exists in M1 (7/22 identical), PM (3/8 identical) and M2 (11/22
identical). In contrast, the MP regions are quite different, as are the
N- and C-termini. In Fig. 4 B the shaded residues comprise
the putative pore helix according to the crystal structure of KcsA
(Doyle et al., 1998). Chimeric molecules were constructed by combining
regions from IRK1 and ROMK2, as described in the Methods section.
N- and C-termini do not determine gating characteristics
To examine the contribution of the cytoplasmic parts of the
proteins to channel gating, we attached the N- and C-termini of IRK1
separately and together to the core component (M1, MP, P, PM, and M2)
of ROMK. As indicated in Fig. 5
A, the kinetics of the channels changed little with
substitution of these segments. In all cases, the closed-time
histograms were well described by two exponentials (see below) and the
open-time histograms by one exponential, similar to the case of ROMK2
channels (Chepilko et al., 1995; Choe et al., 1998). A quantitative
comparison of the kinetics is shown in Fig. 5 B. None of the
kinetic parameters were altered significantly by changes in the
cytoplasmic domains, although single-channel conductance was higher
when the C-terminal region was derived from ROMK.
|
Multiple closed states are conferred by the M2 region
One of the striking differences between ROMK and IRK gating is
that four time constants are required to fit the closed-time histograms
of IRK1 (Figs. 1 C and 6 A). Only two time
constants are necessary to describe ROMK kinetics over the same
frequency bandwidth (Choe et al., 1998). The M2 region of IRK is
essential for this difference. Substitution of only the M2 region of
ROMK into the IRK channel produced a channel with gating patterns that resembled those of ROMK2 (Fig. 6
B). A similar result was obtained if both M1 and M2 regions
of ROMK were substituted into IRK (Fig. 6 D). The chimera
with just the M1 region of ROMK substituted into IRK also displayed
altered kinetics, indicative of three discrete closed states (Fig. 6,
C and E). In both chimeras Chm-108 (Fig. 6
C) and Chm-4 (Fig. 6 E), best fits to the data
were obtained with three time constants (corresponding to three peaks
in the distribution. Although the middle peak in Fig. 6 E is
difficult to discern, the fitting routine gave significantly better
results with three exponentials compared to two (see inset).
In Fig. 6, C and E, the shortest closures were
not observed, possibly because they were too short to be detected at
the bandwidth used.
|
Every chimera tested in which the M2 domain was from IRK1 exhibited
three or four closed states, although the values of the time constants
depended on both the M1 domain and the extracellular loop (Fig. 6,
A, C, and E). Conversely, all chimeras
tested in which the M2 domain was from ROMK2 had only two closed time
constants (Fig. 6, B and D). Unfortunately,
chimeras with an M2 region from IRK and C-terminal region from ROMK2
were nonfunctional (Taglialatela et al., 1994; Doi et al., 1996).
Hence, it was not possible to test whether having the transmembrane
domains from IRK is sufficient for multiple closed states.
Closed times are affected by the extracellular region
We analyzed the closed times of those chimeras for which the
closed-time distributions could be described by two exponentials, corresponding to well-defined "long" and "short" closures. All of these chimeras contained the M2 region of ROMK2, as described above.
In all cases the mean durations of both short and long closures were
biphasic functions of voltage, as previously described for ROMK2 itself
(Choe et al., 1998).
The extracellular loop (ECL) has a major role in determining the long
closed times. In Fig. 7, chimeras Chm-107
and Chm-34 were formed by respectively replacing the ROMK-ECL regions
of ROMK2 and Chm-25 with the corresponding region of IRK (shaded rectangles). This alteration in the ECL regions increased maximum long closed times by about sixfold (Fig. 7). This was true regardless of whether the N- and C-termini were from ROMK2 (Chm-107 vs. ROMK2) or
from IRK1 (Chm-34 vs. Chm-25). The closed times of all these constructs
showed the same qualitative, biphasic voltage dependence, where the
maximum closed time was in the range of 120 to 150 mV.
|
The effect of the ECL on the shorter closed times was qualitatively
similar (Fig. 8). Namely, replacement of
the ROMK-ECL region by its IRK counterpart increased maximum short
closed times by three to sixfold, regardless of whether the cytoplasmic
parts of the channel were from IRK or from ROMK. As with the long
closures, the biphasic voltage dependence of the short closures was
preserved. In one case, where the transmembrane domains were from ROMK
but the N- and C-termini were from IRK, there was an apparent shift in
the maximum short closed time from 100 to 160 mV (Chm-34 vs.
Chm-25) when the ECL was exchanged.
|
The ECL also determined the sensitivity of the
Po to hyperpolarization and divalent cations. As
indicated in Fig. 9, chimeras having the
ECL from ROMK2 had a mild dependence of Po on
voltage, similar to the ROMK2 channel itself. In contrast, when the ECL was from IRK1, there was a steep fall in Po as
the membrane was hyperpolarized from 50 to 150 mV. There was some
dependence of this effect on the transmembrane domains. For example,
replacement of M1 and M2 of IRK1 with the corresponding regions of
ROMK2 shifted the voltage dependence by ~25 mV, but the steepness of
the voltage effect was primarily a function of the ECL. As discussed
above, this effect of hyperpolarization reflects interactions of
divalent blocking ions with the channel. These ions probably interact
directly with the extracellular portion of the channel protein.
|
We attempted to localize the determinants of the closed times more precisely by dividing the ECL into three regions: MP, P, and PM. As shown in Figs. 10 and 11, the main effects of the ECL on both closed times occurred with exchange of the P and PM regions. The MP region, which accounts for ~50% of the ECL, had relatively little influence on the gating parameters.
|
|
The determinants of the long and short closed times were somewhat different. For the long closures, exchanging only the P region increased the mean closed time. Exchanging the PM region had little effect by itself, but this exchange augmented the effects of the P region. In the case of the short closures, the converse pattern was seen. Exchanging only the PM region increased the duration of the short closures. Swapping only the P region had little effect, but this augmented the influence of the PM region.
Note that none of the chimeras had closures as long as those seen with IRK1 (>1 s). It therefore appears that both the transmembrane domains (at least M2) as well as the P-region and the flanking extracellular regions are involved to some extent in determining the stability of this closed state which, as suggested above, may involve the binding of divalent cations within the pore.
Open times are affected by several regions
Another significant difference between IRK1 and ROMK2 is the mean
open time (Table 3). At 100 mV, the
mean open time of IRK1 is ~160 ms, 10 times longer than that of
ROMK2. Replacement of almost any segment of IRK from M1 to M2 with the
corresponding part of ROMK2 reduced the mean open time (Table 3). The
sole exception is the P region; exchange of this segment with that from
ROMK2 increased the open time from 162 to 458 ms. The converse of this
paradoxical effect was also observed; replacement of the P region of
ROMK2 with that of IRK1 decreased the mean open time significantly
(Table 4). The mean open time was most
sensitive to replacement of either the PM or M2 segments of IRK by
their corresponding ROMK segments (Table 3). However, the short (16 ms) open times of ROMK2 were only achieved when the entire
M1-MP-P-PM-M2 region was derived from ROMK2 (Table 3).
|
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
SUMMARY
APPENDIX
REFERENCES
|
|---|
The main conclusions of this paper are: 1) The gating properties of IRK1 differ markedly from those of ROMK2. The gating of IRK1 is slower, with longer mean open and closed times and four discernible closed states versus two for ROMK2. The Po of IRK1 decreases markedly with hyperpolarization, an effect which required only trace amounts of extracellular divalent cations. 2) Channels in which the extracellular segment (MP, P, and PM in Fig. 4) was derived from IRK1 have longer open and closed times and greater sensitivity to divalent cations. 3) The number of observable closed states depends on the second membrane-spanning domain (M2), which may interact with the extracellular loop to stabilize some of the closed states.
These conclusions are discussed in more detail below.
Kinetics of IRK1
The data in Figs. 1-3 constitute, to our knowledge, the first
detailed report of the kinetics of the cloned IRK1 channel. Our results
are consistent with the general properties of inward rectifiers that
have been previously reported (Kubo et al., 1993; Aleksandrov et al.,
1996). Our results on long open times and multiple closed states are
also similar to those reported for native inward rectifier channels in
skeletal muscle (Matsuda and Stanfield, 1989) and aortic endothelial
cells (Elam and Lansman, 1995). However, hyperpolarization decreased
the closed-state lifetime for IRK1 that was expressed in oocytes,
whereas it had little effect on the closed times of native inward
rectifiers. We do not know the basis for this discrepancy.
The most striking features of IRK kinetics are the long open and
closed times compared to ROMK. The IRK1 open times were distributed exponentially, consistent with domination of this distribution by a
single open state of the channel. Similar results were obtained for
native channels (Matsuda and Stanfield, 1989). The mean open time of
IRK was a weak function of voltage, decreasing e-fold for a 200 mV hyperpolarization.
The closed-time distribution for IRK implied the existence of
multiple closed states since at least four exponentials were required
to adequately fit the data. This is clearly a minimum estimate because
the kinetics of IRK1 are so slow that it was difficult to obtain
recordings with enough events to resolve the histograms in greater
detail than that of Fig. 6 A. Previous reports of native
inward-rectifier channels also indicated multiple closed states
(Sakmann and Trube, 1984; Matsuda and Stanfield, 1989). It has not been
deter mined which of these states interact with each other or with the
open state.
Similar to what was found with ROMK2 (Choe et al., 1998), the longest
closed state of IRK1 probably represents block by extracellular divalent cations. Although we did not add divalents to the pipette solution, chelation of divalent cations with EDTA abolished the closures longer than 1 s (Fig. 3). Reduction in Ba2+
concentration by precipitation with SO42
produced a result intermediate between control and EDTA. These results
imply either that there is a very high affinity site for Ba2+ block or that other divalent cations contribute to
this phenomenon.
The source of the Ba2+ responsible for differences in
kinetics between control and EDTA solutions was never positively
identified. However, this effect of trace amounts of barium on inward
rectifier kinetics was previously discussed in detail (Choe et al.,
1998); where it was shown that the long closed state of ROMK2 could be attributed to the presence of 0.1 µM Ba2+, an amount that
might easily have come from the reagent grade KCl used to prepare the solutions.
Although much work has been done on the structural regions responsible
for rectification in the Kir family (Lu and MacKinnon, 1994; Taglialatela et al., 1994, 1995; Wible et al., 1994; Pessia et
al., 1995; Yang et al., 1995) relatively little attention has been paid
to the structural basis underlying differences in the kinetic
properties of these channels. In one study using chimeras of IRK1 and
GIRK1, the fast kinetics of GIRK1 could largely be attributed to its
hydrophobic core (Slesinger et al., 1995). This would correspond to the
M1, MP, P, PM, and M2 regions of ROMK2 or IRK1 in Fig. 4. These
results are consistent with the role of the ECL region as described in
the present study (Figs. 5 A, 7, 8, 9, and 12).
Role of the N- and C-termini
Exchanging the N- and C-termini of ROMK2 and IRK1 had negligible
effect on channel kinetics under the conditions of our experiments. Since these regions share relatively little homology, one might expect
to see changes in kinetics if these regions were involved in the gating
process. On this basis we conclude that the cytoplasmic N- and
C-termini do not participate in channel opening and closing, at least
at voltages near the resting potential. However, the C-terminus is
clearly involved in the Mg2+-dependent closures of the
channel observed at positive membrane potentials (Taglialatela et al.,
1995; Yang et al., 1995).
Role of the transmembrane domains
The transmembrane domains of IRK and ROMK determine the number of closed states as well as the rates of transitions into the closed states, i.e., the mean open time. Replacement of the M2 region of IRK1 with the corresponding region of ROMK2 was sufficient to produce the basic ROMK2 kinetic pattern with two easily resolved closed states (Fig. 6). All chimeras with M2 from ROMK had two closed states, whereas all chimeras having an IRK1-M2 region had more than two closed states.
In addition, replacement of the IRK1-M2 segment by its ROMK counterpart decreased mean open time by an order of magnitude (from 162 ms to 16 ms). This may be related to the associated reduction in the number of closed states. When there are rapid transitions from the open state to a single short-lived closed state, this state will dominate the closed-time histograms, making the others more difficult to resolve.
Role of the extracellular loop
IRK1 mean open time was also affected by the ECL since any chimera with a ROMK2-ECL had short open times that were characteristic of ROMK. Thus, the long duration of the IRK1 open state appears to be determined by interactions between the M2 and ECL regions. In the absence of these interactions mean open times are short (<50 ms), and transitions to the short-lived closed state were frequent.
The ECL also determines the duration of closed times. For chimeras that
exhibit ROMK2-like gating, with two distinct closed states, both the
long and the short closed times are increased when the ECL is
derived from IRK1. We previously proposed that these closed states
represent block of the channel by permeant divalent and monovalent
cations, respectively (Choe et al., 1998). In this hypothesis,
increases in mean closed times reflect tighter binding of the blocking
ions. The finding that both the long and the short closed times are
affected by the same regions of the protein support the view that these
two states are closely related. The results are consistent with the
idea that channel closures are the result of interactions of
extracellular ions with the outer part of the conducting pore that is
formed by the ECL (and stabilized by interactions with the M2
transmembrane segment).
Role of the P-region
The P-region is thought to be an important site of interaction of
ions with the channel and a determinant of channel conductance and
selectivity. Results of the present study implicate the P region in
channel gating as well. The P regions of IRK1 and ROMK2 differ by only
two amino acids: a leucine replaces an isoleucine and a valine replaces
a threonine. We have previously shown that these substitutions alter
the affinity of the pore for block by Ba2+ and by
Cs+ (Zhou et al., 1996). The substitution also increased
the frequency of the "spontaneous" long closed state of ROMK,
consistent with the view that this state results from block by
contaminant divalent cations, probably Ba2+.
Although the duration of the short closed states was not dramatically
changed by this substitution, the rate of occurrence of the short
closures was increased in those chimeras having an IRK P-region. In
terms of the previous model, this might represent either an increase in
the occupancy of a monovalent binding site by K+ or an
increase in the probability of induction of a deeper energy well to
trap permeant ions (Choe et al., 1998).
| |
SUMMARY |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
SUMMARY
APPENDIX
REFERENCES
|
|---|
Fig. 12 is a cartoon of the gating
process, based on data from the present study as well as earlier work
(Choe et al., 1998). The GYG selectivity site of the channel may be
stabilized in a relatively rigid configuration by hydrogen bonding and
van der Waals interactions (Doyle et al., 1998). However, the inner
region of the pore could be flexible enough to interact dynamically
with permeant ions (Fig. 12 B). In this model, long closures
are due to a high-affinity binding of permeant divalents such as
Ba2+, whereas short closures involve a conformational
change of the inner P region, correlated with the presence of a
permeant monovalent ion within the pore. This involves the coordinate
movements of the P, PM, and M2 regions, although the exact nature of
this interaction between protein and permeant ion is unknown. One
possibility is a movement of the pore helix, which sits at an angle to
the plane of the membrane (Doyle et al., 1998). Because an 
-helix
has an associated macro dipole (Hol, 1985) it will feel a torque in the presence of an electric field. Voltage-dependent changes in the position of this and associated parts of the protein could constrict the pore, transiently trapping a permeating ion and producing a channel
closure.
|
| |
APPENDIX |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
SUMMARY
APPENDIX
REFERENCES
|
|---|
The data of Figs. 2 A, 3 B, and 9 were fit
to Eq. 1 of the text:
|
(A1) |
:
|
(A2) |
|
(A3) |
|
(A4) |
|
|
|
|
|
|
|
| |
ACKNOWLEDGMENTS |
|---|
The IRK1 clone was a generous gift of Dr. William Thornhill (Mt. Sinai School of Medicine).
This work was supported by National Institutes of Health Grants DK46950 (to H.S.) and DK27847 (to L.G.P.).
| |
FOOTNOTES |
|---|
Received for publication 29 September 1998 and in final form 27 January 1999.
Address reprint requests to Dr. Henry Sackin, Department of Physiology and Biophysics, The Chicago Medical School, 3333 Green Bay Road, North Chicago, IL 60064. Tel.: 847-578-8329; Fax: 847-578-3265; E-mail: sackinh{at}mis.finchcms.edu.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
SUMMARY
APPENDIX
REFERENCES
|
|---|
-helix dipole in protein function and structure.
Prog. Biophys. Mol. Biol.
45:149-195[Medline].
Biophys J, April 1999, p. 1988-2003, Vol. 76, No. 4
© 1999 by the Biophysical Society 0006-3495/99/04/1988/16 $2.00
This article has been cited by other articles:
 |
|
 |
|
  S. M. Y. Makary, T. W. Claydon, K. M. Dibb, and M. R. Boyett Base of Pore Loop Is Important for Rectification, Activation, Permeation, and Block of Kir3.1/Kir3.4 Biophys. J., June 1, 2006; 90(11): 4018 - 4034. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Sackin, M. Nanazashvili, L. G. Palmer, M. Krambis, and D. E. Walters Structural Locus of the pH Gate in the Kir1.1 Inward Rectifier Channel Biophys. J., April 1, 2005; 88(4): 2597 - 2606. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-Y. Zhang, J. L. Robertson, D. A. Gray, and L. G. Palmer Carboxy-terminal Determinants of Conductance in Inward-rectifier K Channels J. Gen. Physiol., November 29, 2004; 124(6): 729 - 739. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L.-H. Xie, S. A John, B. Ribalet, and J. N Weiss Regulation of gating by negative charges in the cytoplasmic pore in the Kir2.1 channel J. Physiol., November 15, 2004; 561(1): 159 - 168. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. C. Hardie, Y. Gu, F. Martin, S. T. Sweeney, and P. Raghu In Vivo Light-induced and Basal Phospholipase C Activity in Drosophila Photoreceptors Measured with Genetically Targeted Phosphatidylinositol 4,5-Bisphosphate-sensitive Ion Channels (Kir2.1) J. Biol. Chem., November 12, 2004; 279(46): 47773 - 47782. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
|
