1Department of Cell Biology and Anatomy, New York Medical College, Valhalla, New York 10595; and 2Department of Neurology and Neurological Sciences, Stanford University, Stanford, California 94305
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Kang, Jian,
John R. Huguenard, and
David A. Prince.
Voltage-Gated Potassium Channels Activated During Action
Potentials in Layer V Neocortical Pyramidal Neurons.
J. Neurophysiol. 83: 70-80, 2000.
To investigate
voltage-gated potassium channels underlying action potentials (APs), we
simultaneously recorded neuronal APs and single K+ channel
activities, using dual patch-clamp recordings (1 whole cell and 1 cell-attached patch) in single-layer V neocortical pyramidal neurons of
rat brain slices. A fast voltage-gated K+ channel with a
conductance of 37 pS (Kf) opened briefly during AP
repolarization. Activation of Kf channels also was
triggered by patch depolarization and did not require Ca2+
influx. Activation threshold was about 20 mV and inactivation was
voltage dependent. Mean duration of channel activities after single APs
was 6.1 ± 0.6 ms (mean ± SD) at resting membrane
potential (
64 mV), 6.7 ± 0.7 ms at
54 mV, and 62 ± 15 ms at
24 mV. The activation and inactivation properties suggest that
Kf channels function mainly in AP repolarization but not in
regulation of firing. Kf channels were sensitive to a low
concentration of tetraethylammonium (TEA, 1 mM) but not to
charybdotoxin (ChTX, 100 nM). Activities of A-type channels
(KA) also were observed during AP repolarization. KA channels were activated by depolarization with a
threshold near
45 mV, suggesting that KA channels
function in both repolarization and timing of APs. Inactivation was
voltage dependent with decay time constants of 32 ± 6 ms at
64
mV (rest), 112 ± 28 ms at
54 mV, and 367 ± 34 ms at
24
mV. KA channels were localized in clusters and were
characterized by steady-state inactivation, multiple subconductance
states (36 and 19 pS), and inhibition by 5 mM 4-aminopyridine (4-AP)
but not by 1 mM TEA. A delayed rectifier K+ channel
(Kdr) with a unique conductance of 17 pS was recorded from
cell-attached patches with TEA/4-AP-filled pipettes. Kdr channels were activated by depolarization with a threshold near
25 mV
and showed delayed long-lasting activation. Kdr channels were not activated by single action potentials. Large conductance Ca2+-activated K+ (BK) channels were not
triggered by neuronal action potentials in normal slices and only
opened as neuronal responses deteriorated (e.g., smaller or absent
spikes) and in a spike-independent manner. This study provides direct
evidence for different roles of various K+ channels during
action potentials in layer V neocortical pyramidal neurons.
Kf and KA channels contribute to AP
repolarization, while KA channels also regulate repetitive
firing. Kdr channels also may function in regulating
repetitive firing, whereas BK channels appear to be activated only in
pathological conditions.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Voltage-gated potassium channels play a major role
in neuronal action potential (AP) repolarization and repetitive firing (Connor and Stevens 1971a; Hille 1992
;
Hodgkin and Huxley 1952
; Kolb 1990
;
Rudy 1988
, Schwindt et al. 1988
). A
delayed rectifier K+ current
(IK) was described first by
Hodgkin and Huxley (1952)
and found to be responsible
for AP repolarization in squid giant axons. However, the slow
activation and inactivation properties of this current raised questions
about its function in repolarization of more rapid mammalian neuronal
APs. Subsequently, it was proposed that a transient
K+ current (IA)
contributed to neuronal repolarization and repetitive firing
(Connor and Stevens 1971a
,b
; Hagiwara et al.
1961
; Neher 1971
; Schwindt et al.
1988
; Storm 1987
). The channels responsible for
IA (KA channel)
begin to activate below the threshold for AP generation and inactivate
rapidly. Since these initial descriptions, a variety of transient
K+ currents with a wide range of
voltage-dependent activation and inactivation have been recorded in a
variety of neurons (Albert and Nerbonne 1995
;
Gestrelius and Grampp 1983
; Kasai et al.
1986
; Quandt 1988
; Penner et al.
1986
; Rudy 1988
; Solaro et al.
1995
; Spain et al. 1991
; Stansfeld et al.
1986
).
In addition to IA, a fast
Ca2+-activated K+ current
(IC) also was reported to contribute to AP
repolarization and the fast spike afterhyperpolarization (AHP) in
peripheral and some central neurons (Adams et al. 1982;
Lancaster and Nicoll 1987
; Lattorre et al. 1989
; MacDermott and Weight 1982
;
Pennefather et al. 1985
; Storm 1987
).
IC was dependent on both membrane potential
and intracellular Ca2+ levels and was very sensitive to TEA
(0.2-1 mM) and charybdotoxin (ChTX) (Castle et al.
1989
; Lancaster and Nicoll 1987
; Storm
1987
). Large-conductance K+(BK) channels may
contribute to IC and AP repolarization in
muscle cells (Pallota et al. 1981
; Walsh and
Singer 1983
), chromaffin cells (Marty 1981
),
neurons of a GH3 anterior pituitary cell line (Lang
and Ritchie 1987
), sympathetic neurons (Pennefather et
al. 1985
), Helix neurons (Crest and Gola 1993
;
Gola et al. 1990
), and hippocampal neurons
(Lancaster et al. 1991
; Storm 1987
;
Yoshida et al. 1991
). Observations in large neocortical
pyramidal neurons from layer V of the cat sensorimotor cortex suggested
that two transient K+ currents contribute to AP
repolarization, a fast-transient current sensitive to a 1 mM TEA that
differed from IC in terms of its Ca2+ dependence and a slow-transient K+ current
that inactivated slowly and was attenuated by 4-aminopyridine (4-AP)
(Spain et al. 1991
). A recent study showed that AP
repolarization was not Ca2+ dependent in rat neocortical
pyramidal neurons (Pineda et al. 1998
). Several
other groups also indicated that voltage-gated K+ currents
play more important roles in AP repolarization than Ca2+-activated K+ currents (Albert and
Nerbonne 1995
; Foehring and Surmeier 1993
; Locke and Nerbonne 1997
). The channels responsible for
the fast- and slow-transient K+ currents in pyramidal
neurons have not yet been identified.
The application of cDNA cloning methods has identified a large number
of voltage-gated K+ channel genes and transcripts,
demonstrating a molecular diversity that parallels the functional
diversity. Two groups of mammalian neuronal voltage-gated
K+ channel genes have been discovered. The first group
consists of homologs of Drosophila Shaker gene (Sh gene
family), Shaker, Shab, Shaw (Kv3.x), and Shal (Jan and Jan
1990; Perney and Kaczmarek 1991
; Rudy et
al. 1992
). The other of homologs of Drosophila
eag gene (Warmke and Ganetzky 1994
;
Warmke et al. 1991
). At least 20 Sh genes and 25 Sh
transcripts have been identified. In addition, subunits of the same Sh
subfamily can combine with each other to form heteromultimeric
channels, and thereby further increase the diversity of K+
channels (Covarrubias et al. 1991
; Christie et
al. 1990
; Isacoff et al. 1990
; Luneau et
al. 1991
; McCormack et al. 1990
;
Ruppersberg et al. 1990
). Shal-related mRNAs (Kv4.1,
Kv4.2, and Kv4.3) were reported to be responsible for A-type K channels
in the CNS (Serodio et al. 1994
), whereas Shab-related
K+ channels (Kv2.1) were reported to contribute to the
delayed rectifier K+ current (Murakoshi and Trimmer
1999
). Channels, expressed from the Shaw-related gene subfamily
(Kv3.3, Kv3.4, and their heteromultimers) in oocytes, are sensitive to
1 mM TEA (Luneau et al. 1991
; McCormack et al.
1990
; Vega-Saenz de Miera et al. 1992
).
Wang et al. (1998)
reported that Kv3.1 in mouse auditory
neurons was also sensitive to 1 mM TEA. Shaw-related K+
channels were found to be broadly expressed in rat CNS neurons (Martina et al. 1998
; Weiser et al. 1994
,
1995
). Identification of the subunit structure of particular
native K+ channels in specific types of CNS neurons is a
challenging task. Kinetic and pharmacological properties of native
K+ channels in brain slices may provide useful data to
clarify the relationship between native K+ channels and
cDNA-expressed K+ channels.
In this study, we used dual patch-clamp techniques, with one pipette in
the cell-attached patch configuration and one to obtain whole cell
currents (Hamill et al. 1981) to allow simultaneous recordings of single K+ channel activities and action
potentials in individual large pyramidal neurons from layer V of the
rat sensorimotor cortex. We were able to identify three distinct
AP-activated K channels that presumable have different functions in
spike generation.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Brain slices
Brain slices were prepared using previously described techniques
(Kang et al. 1996). Eight- to 15-day-old (P8-P15)
Sprague-Dawley rats of either sex (Simonsen Breeders) were anesthetized
with pentobarbital sodium (55 mg/kg) and decapitated. The brain was removed rapidly, blocked in the coronal plane, and glued to the stage
of a vibratome (TPI, St. Louis, MO) with the posterior surface down.
The block was covered with ice-cold cutting solution (contents in
Solutions and drugs), and 300-µm coronal slices were
cut. Slices containing sensorimotor cortex (FL and Par1) (Zilles
1985
) were incubated for 1-8 h in standard slice solution (see
Solutions and drugs) gassed with 5%
CO2-95% O2 at room temperature (23-25°C) before being transferred to the recording chamber for patch-clamp recordings.
Patch-clamp recordings
Neocortical slices were placed in a small chamber that had
a volume of 1.5 ml and were superfused with standard slice solution gassed with 5% CO2-95% O2
at 31°C. Cells in slices were visualized with a ×63 water immersion
objective and differential inference contrast (DIC) optics (Axioskop,
Zeiss, Germany; Fig. 1A). Two electrically controlled mechanical manipulators (Newport) were mounted
on the microscope stage. Recording electrodes with resistances of 3-7
M were pulled from KG-33 glass capillaries (1.0 mm ID, 1.5 mm OD,
Garner Glass) using a PP-83 electrode puller (Narishige, Japan). Dual
patch-clamp recordings were performed on single pyramidal neurons,
using an Axoclamp 2B for whole cell current-clamp and an Axopatch 200A
for cell-attached patch configurations (Axon Instruments, Burlingame,
CA). Neurons with seal resistances <1 G
were rejected. Single
channel recordings were acquired using the cell-attached patch
configuration (Hamill et al. 1981
) and filtered through
an 8-pole Bessel low-pass filter with a 1-kHz cutoff frequency.
Neuronal spikes were recorded in the whole cell current-clamp
configuration with 3-kHz bandwidth. The resting membrane potentials
were measured in the whole cell recording. Patch membrane potentials
were calculated from the equation: Vp = RMP
Vd, where
Vd is the patch depolarization step
and RMP is the resting membrane potential obtained with whole cell
recordings. The reported voltages were corrected for the liquid
junction potential. In data from cell-attached patches with APs evoked
by extracellular stimulation, an average RMP of
64 mV was used.
Signals were acquired with Pclamp6.0-Clampex and stored on a video-tape
recorder via a NEURO-CORDER converter (Cygnus Technology, Delaware
Water Gap, PA). Extracellular stimuli (0.1 Hz) were applied through a
bipolar tungsten electrode (impedance, 10 M
; FHC, Brunswick, ME)
placed in neocortical layer II/III.
|
Channels permeable to K+ ions were distinguished
according to concentrations of ions across patch membrane and reversal
potentials. Because the patch pipette solution contained neither
Na+ nor Ca2+ and
intracellular levels of Na+ and
Ca2+ are relative low, unitary currents in
cell-attached patch recordings should be carried by
K+ or Cl. Cell-attached
patches contained 140 mM KCl, similar to the expected intracellular
[K+], thus the reversal potential for
K+ was close to 0 mV. The reversal potential for
[Cl
] was
40 to
60 mV in these neurons, as
estimated from recordings of GABAA single channel
currents obtained with GABA-filled cell-attached patch pipettes
(unpublished data).
Solutions and drugs
The slice cutting solution contained (in mM) 2.5 KCl, 1.25 NaH2PO4, 10 MgSO4, 5 CaCl2, 10 glucose, 26 NaHCO3, and 230 sucrose. The standard slice solution contained (in mM) 126 NaCl, 2.5 KCl, 1.25 NaH2PO4, 2 MgCl2, 2 CaCl2, 10 glucose, and 26 NaHCO3 (pH = 7.4 when gassed with 95% O2-5% CO2). The intracellular solution for filling whole cell electrodes contained (in mM) 117 KMeSO4, 13 KCl, 0.1 EGTA, 2 MgCl2, 10 HEPES, 1 ATP, 0.2 GTP, and 4 dextrose (pH adjusted to 7.2 with KOH; osmolarity 280). The solution for filling cell-attached patch electrodes contained (in mM) 130 KCl, 5 EGTA, 2 MgCl2, 10 HEPES, and 4 dextrose (pH adjusted to 7.2 with KOH). Estimated final [K+] was 140 mM after adding KOH to adjust pH. Dipotassium ATP (ATP), N-2-hydroxyethylpiperazine-N'-2-ethane sulfonic acid (HEPES), tetra-ethyl ammonium chloride monohydrate (TEA), ethylene glycol-bis(b-aminoethyl ether)N, N,N',N'-tetraacetic acid (EGTA), poly-L-lysine, 1,2-bis(2-aminophenoxy)ethane-N, N,N',N'-tetraacetic acid (BAPTA), and 4-AP were purchased from Sigma. Other chemicals were purchased from Mallinckrodt Specialty Chemicals (Paris, Kentucky).
Data analysis
PCLAMP 6.0 was used to plot recording traces and analyze amplitude histograms. Single channel currents were sampled every 50 µs with Fetchex and analyzed with Fetchan programs. A 50% threshold criterion was used to determine the duration of open and closed events. The collected open and closed intervals were binned with Pstat, and logarithmic distributions of the open and closed duration were exponentially fitted using the maximal likelihood method. The binwidth was determined automatically with Pstat. Amplitude histograms were analyzed via Fetchan and Pstat programs. Unitary currents for I-V curves were obtained from amplitude histograms. The time course of channel activities that followed single APs, "postspike duration," was measured from the positive AP peak to the end of channel activities. The mean value for postspike duration was obtained from 10 individual traces for each channel studied. Statistical data are presented as means ± SE.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Channel activities during action potentials
Large pyramidal neurons in layer V of the rat neocortex were
visualized with DIC optics and identified by their oblong somata and
single long apical dendrite extending toward to pia (Fig. 1A, inset). Dual patch-clamp recordings were
performed on somata of these neurons to simultaneously acquire single
channel activities (cell-attached patch) and neuronal action potentials
(whole cell current clamp). The mean value of resting membrane
potentials measured by whole cell electrodes was 64.0 ± 0.3 mV
(n = 67 cells). When a neuronal AP was evoked with a
depolarizing current pulse, openings of two types of
K+ channels were observed in cell attached
patches. A fast-opening channel (Kf) was
characterized by very brief openings (Fig. 1Ab,
),
whereas a second channel (KA) was characterized
by longer openings and clustering (Fig. 1Bb,
).
Twenty-three of 65 patches contained both Kf and
KA channels, 12 patches contained
Kf channels without KA
channels, and 21 patches contained only KA
channels. A single Kf channel often was observed
(7 of 12) in a patch (Fig. 1Ab,
), whereas multiple
KA channels always were recorded (21 of 21) from
one patch (Fig. 1Bb), indicating that
KA channels were localized in clusters.
Electrophysiological properties of Kf channels
To examine voltage-dependent properties of
Kf channels, we obtained cell-attached patch
recordings while evoking APs with extracellular stimulation.
Cell-attached patches were conditioned by delivering voltage steps to
the patch electrode (Fig. 2A,
bottom), and an extracellular stimulus (200 µs) was
delivered to layer II/III (Fig. 2A, arrow, Stim). Stimulus
intensity was adjusted to reliably elicit a single spike with a delay
of 7-50 ms (Fig. 2A, AP). Because only the patch membrane
was depolarized, we could test voltage-dependent properties of channels
without changing RMP or neuronal excitability. In addition,
AP-activated channel activities could be measured at different patch
membrane potentials to examine voltage-dependent inactivation of the
channel. Data from seven patches, which contained only
Kf channels without KA channels, were analyzed. When cell-attached patches were depolarized to
a potential below 14 mV, the Kf channel was not
activated by depolarization steps but opened by APs (Fig.
2A). Brief openings of the channel indicate fast
inactivation (Fig. 2, A and B). The mean
postspike duration of Kf channel activities was
6.1 ± 0.6 ms at
64 mV, 6.7 ± 0.7 ms at
54 mV, and
62 ± 15 ms at
24 mV (Fig. 2B, n = 7 patches). The results show that inactivation of Kf channels is voltage dependent. The
current-voltage relationship (I-V) corresponded to a slope
conductance of 37 ± 1.5 pS (Fig. 2C, n = 7 patches). Single channel currents were reversed at +1.3 ± 2.5 mV (n = 7 patches), which was close to estimated
K+ equilibrium potential for cell-attached
patches (EK
0 mV). Kf channels were activated by strong patch
depolarization (Fig. 2A, +36 and +56 mV), indicating that
they are voltage-gated channels. To further test voltage-dependent
activation of Kf channels, cell-attached patches
were depolarized without extracellular stimulation. Patch depolarization induced openings of Kf channels
(Fig. 3A) with a threshold of
22 ± 2.5 mV (n = 7 patches). Because only a
small piece of membrane in the patch pipette was depolarized, no AP or
AP-induced Ca2+ influx occurred. Therefore
activation of Kf channels was independent of
AP-induced Ca2+ influx. The amplitude histogram
for Kf channel openings had a single peak (Fig.
3B, o), reflecting a single conductance state. Kinetic
analysis demonstrated a mean open time (
o) of
6.8 ± 0.7 ms at membrane potential of +56 mV (n = 5 patches), which was close to the postspike duration of
Kf channel activities at
64 mV (6.1 ± 0.6 ms), suggesting that the channel might open once, or very few times,
during AP repolarization. A double exponential with a fast time
constant (
c1) of 1.1 ± 0.1 ms and slow
time constant (
c2) of 19 ± 6 ms was
needed to best fit the closed time distribution (Fig. 3C,
n = 5 patches).
|
|
Pharmacological properties of Kf channels
In rat neocortical pyramidal neurons, both AP repolarization and
the fast-transient K+ current were inhibited by
low concentrations of TEA (1 mM, data not shown), similar to
observations in cat neocortical pyramidal neurons (Spain et al.
1991). To further determine whether Kf
channels contribute to the TEA-sensitive component of AP
repolarization, we attempted to block Kf channels
by including 1 mM TEA in the patch pipette solution.
Kf channel activities were not detected in any of
the 35 TEA-containing cell-attached patches (Fig.
4A), indicating that
Kf channels were TEA sensitive. Because
Ca2+-activated BK channels are very sensitive to
both TEA (1 mM) and charybdotoxin (ChTX, 100 nM), we tested sensitivity
of Kf channels to ChTX. Kf
channel activities were not blocked in five cell-attached patch
recordings obtained with pipettes containing 100 nM ChTX (Fig.
4B), demonstrating that Kf channels
are pharmacologically distinct from BK channels (Kang et al.
1996
).
|
Electrophysiological properties of KA channels
To distinguish KA channels from
Ca2+-activated K+ channels,
we dissolved 10 mM BAPTA, a calcium chelator (Tsien
1980), in the whole cell pipette solution. Before the whole
cell patch membrane was ruptured, a depolarizing current pulse induced
a large voltage jump (>200 mV, Fig.
5Aa), due to the high-input
resistance (>1 G
) of the intact of patch membrane, together with
a train of AP transient currents that showed frequency
adaptation in cell-attached patch-clamp recordings (Fig.
5Ab). Action potentials were followed by activities of
multiple KA channels (Fig. 5Ab). Once
the patch membrane was ruptured, spike frequency adaptation was reduced significantly (Fig. 5B, a and b), presumably due
to inactivation of Ca2+-dependent
channel activities by chelation of intracellular
Ca2+. However,
KA channel activities were not inhibited (Fig.
5Bb,
), demonstrating that KA
channels are not Ca2+ dependent.
|
KA channels showed multiple subconductances, a
finding similar to that reported in tissue cultures (Forsythe et
al. 1992). Because each patch often contained multiple
KA channels, it was difficult to distinguish all
subconductance states from simultaneous openings of multiple channels.
However, we could clearly distinguish at least two single channel
current levels of 1.2 ± 0.03 pA and 2.3 ± 0.04 pA after
evoked APs (n = 15 patches), especially in late
openings (Fig. 6, 1 and 2). The slope
conductance of KA channels was not resolvable
because channels were localized in clusters and multiple channels were
activated when patches were depolarized (Fig. 7,
A and B). When
conductance was estimated by dividing the mean current amplitude by
64 mV (assumed mean resting membrane potential from whole cell
recordings), two major subconductances of 19 and 36 pS were obtained.
The different conductances are likely to represent two subconductances
of the KA channel, rather than two subtypes of
KA channels because these two conductances always
coexisted in patches containing KA channels;
transition between two states could be seen occasionally; they showed
similar activation and inactivation; 4-AP blocks both of them; and
similar subconductances of KA channels have been
reported previously in neuronal tissue cultures (Forsythe et al.
1992
).
|
|
KA channels are voltage-gated and were opened by
patch depolarization with a threshold of 44.8 ± 2.9 mV (Fig.
7A, n = 13 patches). Channel activities were
larger during tail currents than during the depolarization period
because of the larger driving force at
64 mV than at
44 mV (Fig.
7A, bottom). Stronger patch depolarizations
induced opening of more channels (Fig. 7A, top). To test the
voltage-dependence of KA channel inactivation,
cell-attached patches were step depolarized (Fig. 7B,
bottom) and APs were evoked by extracellular stimulation
(Fig. 7B, arrow, Stim). Action-potential-activated KA channel activities lasted much longer when the
patch membrane potential was more positive than resting potential (Fig.
7, B and C,
54,
44, and
34). Postspike
duration of KA channel activities was 32 ± 6 ms at
64 mV, 112 ± 28 ms at
54 mV, and 367 ± 34 ms at
24 mV (n = 5 patches). The results indicated that
inactivation from the opening state of KA
channels is voltage dependent. The time course of
KA channel activities would allow the channel to play roles in both AP repolarization and regulation of the repetitive firing rate. The steady-state inactivation property, previously described for IA (Connor and
Stevens 1971a
,b
), also was observed in KA
channel activities. When patch membrane was held at rest (
64 mV),
AP-activated KA channel activities were similar
to those in Fig. 7B (Fig.
8A). When the patch membrane
was held at
44 mV (5-15 s), KA channels were
steady-state inactivated (Fig. 8B). When the patch membrane
then was held at
84 mV (10-20 s), KA channels
recovered from inactivation, and even more channels were activated by
the AP than at rest (cf traces at Vh
64 mV, Fig. 8A and Vh
84 mV Fig. 8C), suggesting that some channels had been inactivated at rest. Steady-state inactivation occurs from the closed
state of KA channels and is different from
inactivation occurring from opening state of the channel (Fig. 7,
B and C). Steady-state inactivation is also
voltage dependent and occurs in less channels when the patch membrane
is more hyperpolarized (Fig. 8), whereas inactivation from opening
state of the channel is faster when the patch membrane is more
hyperpolarized (Fig. 7, B and C). Therefore
voltage dependence for steady-state inactivation is opposite in
direction to that for inactivation from the opening state of
KA channels.
|
Pharmacological properties of KA channels
To examine sensitivity of KA channels to
4-AP, dual patch-clamp recordings were performed in the absence or
presence of 5 mM 4-AP in cell-attached patch pipettes. BAPTA (10 mM)
and 1 mM TEA were added to the whole cell and cell-attached pipettes,
respectively, to eliminate activities of
Ca2+-activated K+ and
Kf channels. Six of 10 control patches showed
KA channel activities, and 4/10 patches obtained
with 4-AP-filled pipettes showed 4-AP-insensitive
KA channel activities (Fig.
9A). Measurements of the
integrated charge for channel currents over a period of 50 ms indicated
that 4-AP significantly inhibited KA channel
activities (Fig. 9B, P < 0.05, t-test, n = 5 and 4 for control and 4-AP
groups, respectively). KA channels that were
insensitive to 4-AP had properties similar to those of 4-AP-sensitive
KA channels, including subconductances, clustered
localization, voltage-dependent activation and inactivation, and
steady-state inactivation (data not shown). From these data, we cannot
determine whether there are two types of KA
channel that are differentially sensitive to 4-AP or one type the
sensitivity of which to 4-AP is relatively low, so that 5 mM 4-AP could
not block all channels. KA channel activities
were not blocked by 1 mM TEA, although the fast component of decay
component of postspike ensemble channel activity (1), which
reflected Kf channel activities, was eliminated
(Fig. 9Cb).
|
Activities of delayed rectifier K+ channels (Kdr)
Activities of a Kdr channel were observed in
5/10 patches recorded with TEA/4-AP-filled pipettes.
Kdr channels were not activated by single APs
when the patch membrane potential was more negative than 24 mV (Fig.
10A). Strong patch
depolarization induced delayed long-lasting openings of
Kdr channels (Fig. 10B). The threshold for Kdr channel openings was
23 ± 2.4 mV
(n = 5 patches). A unique slope conductance of 17 ± 0.7 pS for Kdr channels was determined from
I-V curves (Fig. 10C, n = 5 patches). The
channel could not be a subconductance state of KA
channels because Kdr channels never showed
multiple conductances; Kdr channels were not
distributed in clusters; activation and inactivation of
Kdr channels were slower than those of
KA channels; the opening threshold for
Kdr channels (
23 ± 2.4 mV) was higher
than that of KA channels (
44.8 ± 2.9 mV);
and insensitivity to 1 mM TEA or 5 mM 4-AP. The results suggest that
Kdr channels may not contribute to repolarization of a single AP because of their delayed activation but, with a slow
inactivation time course, may play a role in regulation of AP wave form
during repetitive firing.
|
Activities of BK channels
Although BK channels have been hypothesized to contribute to AP
repolarization and the fast AHP in hippocampal pyramidal neurons (Lancaster and Nicoll 1987; Storm 1987
;
Yoshida et al. 1991
), we did not observe BK channel
openings in any of the 65 "normal" cell-attached patches in rat
neocortical pyramidal neurons even under conditions where
Ca2+ (2 mM) was included in the patch pipette
(n = 5). Further, BK channels were not activated when
>20 APs were evoked by the depolarizing current (data not shown).
However, after patches were excised from neurons and changed into the
inside-out configuration (8/8 of patches), BK channel openings were
observed, indicating that patches contained BK channels, but they did
not open in intact neurons under our recording conditions. These
results are in accordance with previous observations by other groups
that no IC was found in whole cell
currents of cat neocortical pyramidal neurons (Spain et al.
1991
) and spike repolarization was Ca2+
independent (Pineda et al. 1998
; Schwindt et al.
1988
). In six experiments, BK channels were not activated when
neuronal APs appeared normal (Fig. 11,
20 min, a). However, BK channels started to open in a spike-independent
manner when neuronal responses deteriorated as evidenced by smaller or
absent spikes (Fig. 11, 40 and 42 min). Initiation of BK channel
activities often occurred over a short time (1-2 min) accompanied by a
change in cellular morphology observed under DIC optics, implying that
some untoward intracellular events might have occurred. The results
suggest that somatic BK channels do not contribute to AP repolarization and the fast AHP in these cells but may play roles under pathological conditions.
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In these experiments, we examined the properties of two
voltage-gated K+ channels activated by APs,
Kf and KA, using dual
patch-clamp techniques applied to large layer V pyramidal neurons in
neocortical slices. The properties of these two channels and a third,
Kdr, that was activated by depolarizing steps,
are summarized in Table 1. Fast and
transient opening of Kf channels with a threshold of 22 mV suggests that they contribute to AP repolarization but not
to sculpting interspike voltage trajectories. Openings were brief (Fig.
1Ab, postspike duration: 6.1 ± 0.6 ms), suggesting that they did not have a role in setting interspike intervals. The
fast-transient K+ current previously reported in
whole cell recordings (Spain et al. 1991
) likely results
from Kf channel activities because both events
are similar in terms of voltage-dependent activation and sensitivity to
1 mM TEA. Although Kf channels are very sensitive to TEA, they are distinct from BK channels in terms of their
conductance of 37 pS (Fig. 2C), voltage-gated activation
without requirement for AP-induced Ca2+ influx
(Fig. 3A) and insensitivity to ChTX.
|
KA channels also were activated by action
potentials. Action-potential-activated KA
channels might have been misinterpreted as
Ca2+-activated channels because
KA channel activities were attenuated by
perfusion of cadmium in previous studies (Alger et al.
1994; Kang et al. 1995
). However, results of
BAPTA experiments (Fig. 5) showed that KA
channels are not Ca2+ dependent.
Voltage-dependent activation of KA channels at a
low threshold of
45 mV suggests that these channels influence AP generation. Postspike duration of KA channel
activities was 32 ± 6 ms at
64 mV, indicating that these
channels contribute to both AP repolarization and regulation of
repetitive firing rate. When patches were depolarized, postspike
duration of channel activities was much longer than that at rest (Fig.
7, B and C), suggesting voltage-dependent
inactivation from the opening state of KA
channels. This property suggests that KA channels
play even larger roles when neurons are transiently depolarized, such
as during repetitive firing. However, if neurons are depolarized for a
prolonged time, KA channels will be inactivated
because of steady-state inactivation that occurs from closed state of
channels (Fig. 8B). Steady-state inactivation is a
well-known property of IA and explains
why spike frequency was higher when neurons were held at depolarized
potentials (data not shown).
Although spike frequency adaptation was suppressed by chelation
of intracellular Ca2+ in BAPTA diffusion
experiments (Fig. 5), when using pipettes filled with control solution,
we could not record activities of single
Ca2+-activated K+ channels
in cell-attached patches during the spike frequency adaptation or slow
AHPs after repetitive spikes. A possible explanation is that
Ca2+-activated K+ channels
responsible for the spike adaptation and the slow AHP have too small a
conductance (Sah and Isaacson 1995) to allow their
detection from the background noise of our recording system. Another
possibility is that SK channels are primarily dendritic in location.
Although BK channels have been hypothesized to contribute to AP
repolarization and the fast AHP in muscle (Pallota et al. 1981; Walsh and Singer 1983
), chromaffin cells
(Marty 1981
), GH3 anterior
pituitary cells (Lang and Ritchie 1987
), and hippocampal neurons (Lancaster and Nicoll 1987
; Storm
1987
; Yoshida et al. 1991
), direct evidence in
this study demonstrates that somatic BK channels do not open during AP
repolarization and the fast AHP in intact neocortical pyramidal
neurons. We did not observe any BK channel activities in cell-attached
patches from "healthy" neurons during normal neuronal firing. The
absence of BK channel activities did not result from dilution of
intracellular Ca2+ by whole cell electrodes
because the same results were observed in cell-attached patch
recordings combined with extracellular stimulation-induced APs, where
the neuronal membrane was intact. BK channels were not activated even
when neurons were firing at a high rate (>20 spikes per train). At
this spiking rate, intracellular Ca2+ should be
higher than the threshold concentration of Ca2+
necessary for activation of BK channels (Kang et al.
1996
; Ross 1993
). Absence of BK channel openings
was not due to lack of local Ca2+ entry because
BK channel activities were not observed during APs when the
cell-attached patch contained CaCl2. Under such
conditions, APs should have depolarized the patch membrane
intracellularly and induced Ca2+ influx through
Ca2+ channels in the patch. The results in the
present study are in accordance with the previous observations that
spike repolarization was not Ca2+ dependent in
neocortical pyramidal neurons (Lorenzon and Foehring 1995
; Pineda et al. 1998
; Schwindt et al.
1988
). BK channels were observed only as neurons deteriorated.
At the time BK channels started to open, neuronal spikes had always
become smaller or absent (Fig. 7, 40 min) and morphological alterations
were often seen. Therefore BK channels may function in pathological
conditions but do not contribute to AP repolarization and the fast AHP
in these cells.
![]() |
FOOTNOTES |
---|
Address for reprint requests: J. Kang, Dept. of Cell Biology and Anatomy, New York Medical College, BSB Rm 220, Valhalla, NY 10595.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 15 March 1999; accepted in final form 21 September 1999.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|