1Groupe de Neurobiologie,
Benquet, Pascal,
Janine Le Guen,
Govindan Dayanithi,
Yves Pichon, and
François Tiaho.
In the CNS, voltage-dependent
Ca2+ channels (VDCC) are involved in
neurotransmission, regulation of cell excitability, and gene transcription. Multiple types of Ca2+ current
(ICa) have been distinguished and
named T-, L-, N-, P-, Q-, and R-type according to biophysical and
pharmacological criteria (for review see Catterall et al.
1995 In the insect CNS, LVA or M-LVA (Baines and Bate 1998 Embryonic cockroach brain neurons in primary culture have been shown to
express voltage-dependent calcium currents in their soma
(Christensen et al. 1988 Cell culture
The culture technique was derived from that of Chen and
Levi-Montalcini (1970) Electrophysiology
Ca2+ currents of the neurons were studied
using the whole cell configuration of the patch-clamp technique
(Hamill et al. 1981 Because the neurons studied (3-15 days in culture) had processes, the
quality of the whole cell voltage clamp depended on the location of the
calcium channels. Study of the transient current elicited by a
hyperpolarizing voltage step (from Data analysis
The pClamp 5.5 program (Axon Instruments) was used for
stimulation, data acquisition, and analysis. Furthermore, data were analyzed off-line using different software packages: Excel (Microsoft), Freelance Graphics (Lotus), and Sigmaplot (Jandel Scientific). Student's t-test was used for statistical analysis. The
peak current-to-voltage relationship (I-V curve) was fitted
with the following Boltzmann equation:
ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
-AgaIVA-Sensitive (P/Q-type) and -Resistant (R-type)
High-Voltage-Activated Ba2+ Currents in Embryonic
Cockroach Brain Neurons.
J. Neurophysiol. 82: 2284-2293, 1999.
By means of the whole cell patch-clamp
technique, the biophysical and pharmacological properties of
voltage-dependent Ba2+ currents
(IBa) were characterized in embryonic
cockroach brain neurons in primary culture.
IBa was characterized by a threshold of
approximately
30 mV, a maximum at ~0 mV, and a reversal potential near +40 mV. Varying the holding potential from
100 to
40 mV did
not modify these properties. The steady-state, voltage-dependent activation and inactivation properties of the current were determined by fitting the corresponding curves with the Boltzmann equation and
yielded V0.5 of
10 ± 2 (SE) mV and
30 ± 1 mV, respectively. IBa was
insensitive to the dihydropyridine (DHP) agonist BayK8644 (1 µM) and
antagonist isradipine (10 µM) but was efficiently and reversibly
blocked by the phenylalkylamine verapamil in a dose-dependent manner
(IC50 = 170 µM). The toxin
-CgTxGVIA (1 µM) had no significant effect on IBa.
Micromolar doses of
-CmTxMVIIC were needed to reduce the current
amplitude significantly, and the effect was slow. At 1 µM, 38% of
the peak current was blocked after 1 h. In contrast,
IBa was potently and irreversibly blocked by nanomolar concentrations of
-AgaTxIVA in ~81% of the neurons. Approximately 20% of the current was unaffected after treatment of the neurons with
high concentrations of the toxin (0.4-1 µM). The steady-state dose-response relationship was fitted with a Hill equation and yielded
an IC50 of 17 nM and a Hill coefficient
(n) of 0.6. A better fit was obtained with a combination
of two Hill equations corresponding to specific
(IC50 = 9 nM; n = 1) and
nonspecific (IC50 = 900 nM; n
= 1)
-AgaTxIVA-sensitive components. In the remaining 19%
of the neurons, concentrations
100 nM
-AgaTxIVA had no visible
effect on IBa. On the basis of these results, it is
concluded that embryonic cockroach brain neurons in primary culture
express at least two types of voltage-dependent,
high-voltage-activated (HVA) calcium channels: a specific
-AgaTxIVA-sensitive component and DHP-,
-CgTxGVIA-, and
-AgaTxIVA-resistant component related respectively to the P/Q- and
R-type voltage-dependent calcium channels.
INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
; Dunlap et al. 1995
; Olivera et al.
1994
; Reuter 1996
; Tsien et al.
1995
). These currents have been subdivided in
low-voltage-activated channels (LVA) for the T-type (Nowycky et
al. 1985
), mid-low-voltage-activated channel (M-LVA) for the
R-type (Dunlap et al. 1995
; Ellinor et al.
1993
; Soong et al. 1993
), and
high-voltage-activated (HVA) for the L-, N-, P-, and Q-types
(Llinas et al. 1989
; Nowycky et al. 1985
; Sather et al. 1993
).
;
Grolleau and Lapied 1996
; Wicher and Penzlin
1997
) and HVA calcium channels have been found similar to those
in vertebrates. However, HVA calcium channels that were sensitive to
phenylalkylamines (PAA) but insensitive to dihydropyridines (DHPs),
unlike vertebrate L-type calcium channels, have been identified in
several insect neurons, suggesting some difference between vertebrate
and invertebrate calcium channels (Bickmeyer et al.
1994a
; Pauron et al. 1987
; Pearson et al.
1993
; Pelzer et al. 1989
; Wicher and
Penzlin 1994
). With the discovery of toxins that are specific
blockers of vertebrate non-L-type HVA calcium channel blockers
(Llinas et al. 1989
; McCleskey et al.
1987
; Mintz et al. 1992b
), it has been shown
that micromolar concentrations of
-CgTxGVIA selectively block HVA
Ca2+ currents in adult cockroach abdominal ganglion dorsal
unpaired median (DUM) neurons (Wicher and Penzlin 1994
),
and that nanomolar concentrations of
-AgaTxIVA partially block HVA
Ca2+ currents in adult locust brain neurosecretory neurons
(Bickmeyer et al. 1994b
) and M-LVA Ca2+
currents in adult abdominal ganglion cockroach DUM neurons
(Wicher and Penzlin 1997
). Unfortunately, in all these
experiments, the potencies and efficacies of these blockers were not
determined by using dose-response curves. These parameters are
necessary to assess the specificity of these blockers for VDCCs in
insect neurons and to estimate the contribution of each type of channel to the macroscopic current.
). The pharmacology of the
corresponding channels was not studied in detail, and consequently, no
information concerning their possible diversity was available. As a
first step in understanding their physiological role in developing
neurons, we have analyzed the contribution of each type of VDCC to the macroscopic current, by using specific blockers and toxins of vertebrate VDCCs. The results show that embryonic cockroach brain neurons in primary culture expressed VDCC that were specifically and
potently blocked by
-AgaTxIVA, similar to P/Q-like VDCCs in
vertebrates. They also expressed DHP-,
-CgTxGVIA-, and
-AgaTxIVA-resistant VDCC, reminiscent of the vertebrate R-type. The
physiological significance of the currents flowing through these two
types of channels is discussed.
METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
, as described by Beadle and Hicks
(1985)
and recently modified (Amar et al. 1991
;
Van Eyseren et al. 1998
). Briefly, cells were isolated
from 21-23-day-old egg case embryos that were stored in an incubator
at 28-29°C in a humid atmosphere. The brains were then removed from
the head capsules and transferred into a glass tube, where they were
dissociated by gentle mechanical trituration with a Pasteur pipette (no
enzyme treatment was necessary) in a defined volume of culture medium.
In general, 30 brains were dissociated in 1 ml culture medium and
yielded a density of 3 × 104
neurons/cm2 adhering to the bottom of the dish.
The cultures were initiated in a medium (5 + 4) containing 5 parts of
Schneider's revised Drosophila medium and 4 parts of
Eagle's basal medium containing 100 IU/ml penicillin and 100 µg/ml
streptomycin complemented with 6 mg/ml
L-glutamine and 2.5 µg/ml fungizone. After 5 days, the first (5 + 4) culture medium was replaced by a second (L + G) medium made up of equal parts of Leibovitz's L-15 medium and Yunker's modified Grace medium containing penicillin, streptomycin, glutamine, and fungizone, supplemented with 10% fetal calf serum. This second medium was renewed every week. All culture media were obtained from
Gibco (Cergy Pontoise, France).
). Before the experiments, the
culture medium was replaced by a solution containing (in mM) 100 TEACl,
70 Tris-HCl, 10 4-AP, 10 BaCl2, 4 MgCl2, and 10 HEPES buffer adjusted to pH 7.2 with TEAOH. This experimental extracellular solution was designed to
eliminate any contaminating voltage-dependent sodium current
(INa+) and to markedly reduce the
voltage-dependent potassium current (IK+).
In addition, Ba2+ ions were used instead of
Ca2+ in this solution to avoid the run-down of
the current (ICa) through voltage-activated calcium channels (see also Wicher et Penzlin 1997
). The patch electrodes were made of borosilicate 1.5-mm
glass (Clark Electromedical) with a Flaming-Brown micropipette puller (Sutter Instruments). They were filled with a solution containing (in
mM) 120 CsF, 25 CsOH, 2 MgCl2, 10 EGTA, 3 ATP-Mg2+, 0.5 guanosine 5'-triphosphate-Tris, and
10 HEPES buffer adjusted to pH 7.3, using CsOH, and their resistance
ranged from 2 to 5 M
. Voltage-clamp experiments were performed with
the patch-clamp amplifier RK300 (Biologic Science Instruments, Claix,
France) at a holding potential (HP) of
70 mV, and all experiments
were performed at room temperature (20-27°C). In some experiments, a
ramp protocol was used to quickly assess the I-V
relationship of IBa. In this protocol,
the membrane potential was transiently and linearly varied from
100
to +50 mV for 500 ms.
60 to
70 mV) provided
information relevant to the quality of the voltage clamp (see
Byerly and Leung 1988
). In general, these transients
varied progressively from a single exponential time constant
[
1 = 0.3 ± 0.1 (SE) ms] for neurons
that had been <5 days in culture) to a complicated time course that
could be approximated by two exponential components
(
1 = 0.3 ± 0.1 ms;
2 = 2 ± 1 ms) for older neurons. In the
present experiments, we selected the neurons in which the amplitude of
the slow exponential component, when present, was <25% of that of the
fast component. Under these conditions, the control of the membrane
potential was fast enough to enable an adequate recording of the
currents. These neurons had an average input resistance of 2.2 ± 0.8 G
, and the estimated voltage error due to uncompensated series
resistance was <5 mV in all the neurons used for the analysis (see
also Tiaho et al. 1991
).
where Gmax is the maximal
conductance of the global calcium channels,
EBa is the reversal potential of
IBa estimated by the curve-fitting
program, V0.5 is the potential for
half-maximal steady-state activation of the barium current, and
K is a voltage-dependent slope factor. Steady-state
activation curves were fitted with the following Boltzmann equation:
(1)
where I is the peak amplitude of
IBa tail current on repolarization to
(2)
100 mV, Imax is the maximum
IBa tail current, and V0.5 and K have the same
meaning as in Eq. 1.
Steady-state inactivation curves were fitted with the following
Boltzmann equation:
![]() |
(3) |
To quantify the effect of calcium channel blockers, dose-response curve
were fitted with a Hill equation:
![]() |
(4) |
Drugs
Verapamil, amiloride, isradipine, nifedipine, and Bay K 8644 were purchased from Sigma (L'Isle d'Abeau Chesnes, France). Verapamil solutions were prepared immediately before use. Amiloride, isradipine, nifedipine, and Bay K 8644, were dissolved in DMSO to obtain a stock
solution that was further diluted. The final test concentrations contained 0.1% DMSO. At this concentration the solvent had no significant effect on IBa.
-CgTxGVIA,
-CmTxMVIIC, and
-AgaTxIVA (Neurex) were dissolved
in distilled water at
10
4-10
3 M and stored
at
70°C.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Voltage-dependent activation of Ba2+ currents
We used extracellular and intracellular media defined to minimize
the contribution of contaminating K+ outward and
Na+ inward currents (see METHODS).
Voltage-clamp pulses were applied from an HP of 70 mV to various
depolarized levels in the whole cell configuration of the patch-clamp
technique. As a result, net inward IBa
at potentials between
30 mV and +40 mV was observed. After 2 days in
culture, all neurons produced a detectable inward current. Typical
recordings of IBa are shown in Fig.
1A. In this neuron,
IBa was activated at
30 mV. Larger
depolarizing pulses of
0 mV progressively increased
IBa peak amplitude, and then further
depolarizations beyond this potential decreased
IBa peak amplitude. The peak current
reversed at membrane potential more positive than +50 mV. These inward
currents were completely blocked by millimolar concentrations of
cadmium, nickel, and cobalt (data not shown) and had characteristics
similar to those previously described by Christensen et al.
(1988)
in the same neurons. After a complete block of
IBa by cadmium, nickel, and cobalt, a
tiny time-independent, outwardly rectifying current reversing at
10 mV was generally observed (data not shown). Because this reversal potential does not correspond to that of any permeating ion, it was
considered nonspecific. This current component can be seen in Fig.
1A for the depolarization to +50 mV, in which the current is
inward at the peak and outward at the end of the pulse. A contamination of IBa with this nonspecific current
could be detected at potentials more positive than +10 mV, but in
general its average amplitude was small (<30% of
IBa peak amplitude at +50 mV). The
experiments presented here were performed at potential values at which
this nonspecific current was negligible, the current traces were
therefore not corrected for this contaminating component.
|
The average peak amplitude of IBa for 66 neurons is plotted against the membrane potential in Fig. 1B. The I-V curve was fitted with the Boltzmann Eq. 1 (see METHODS), assuming that the time-dependent inactivation of the current was negligible at the peak and that the current was flowing through a homogeneous population of calcium channels with the same steady-state voltage-dependent activation properties.
HVA and LVA calcium currents have been distinguished according to their
voltage-dependent activation properties both in vertebrate and
invertebrate neurons (Nowycky et al. 1985, and for
review see Bean 1989
; Hess 1990
). They
also differ by their inactivation properties: depolarized HP
preferentially inactivate LVA currents, and their time-dependent
inactivation kinetics are faster than those of HVA currents.
The activation of the current was therefore studied for three
different HPs: 100 mV,
70 mV, and
40 mV. As illustrated in Fig.
2A, the currents were the same
for
100- and
70-mV HPs. Furthermore, as illustrated in the
left panel, no detectable current was seen for a
depolarization from
100 mV or
70 mV to
40 mV, ruling out the
existence of a low-voltage-activated current component that may have
been hidden at
70 mV. For an HP of
40 mV, the current was
significantly inactivated, and the amount varied from neuron to neuron
by about one third to two thirds of the maximum peak current. The
average peak amplitude of IBa for the
three HPs versus the command potential for at least eight neurons is illustrated in Fig. 2B. Apart from the scaling factor, the
I-V curves could also be fitted with the described Boltzmann
activation equation with the same parameters (see legend of Fig. 2).
|
For suprathreshold potentials, IBa
activation was fast (that is, in the ms range) and the time to peak
decreased with increasing depolarizations. For the cell illustrated in
Fig. 3A the times to peak were
6.8 ms at 20 mV, 5.7 ms at 0 mV, 4.4 ms at +20 mV, and 3.2 ms at +30 mV, indicating that activation kinetics were voltage dependent.
|
The steady-state activation properties of
IBa were studied using short-duration
pulses that were adequate to fully activate IBa and minimize time-dependent
inactivation (Fig. 3A). Tail currents observed on
hyperpolarizing the membrane to 100 mV (Fig. 3A) were used
to estimate the steady-state conductance of the calcium channels (Fig.
3C). The mean potential of half-maximum activation of the
conductance (
10 ± 2mV) was similar to that found for the peak
conductance (
10 ± 1 mV, see legend to Fig. 1B). The
small discrepancy found in the slope factor (K = 9 ± 1 for the steady-state activation curve against K = 7 ± 1 for the I-V curve) is within the limit of the
values reported by different investigators (for review see
Pelzer et al. 1990
).
Voltage-dependent inactivation of IBa
We studied the steady-state, voltage-dependent inactivation
properties of IBa using 7-s prepulses
as illustrated in Fig. 3B. This duration was chosen because
in most neurons complete time-dependent inactivation of
IBa needed >1 second (see Fig.
1A). A typical illustration of the steady-state inactivation
of IBa is illustrated in Fig.
3B. After conditioning prepulses of increasing positivity, the peak amplitude of the current evoked by a pulse to +10
mV progressively decreased and was almost completely suppressed for a
10 mV prepulse potential. The average peak amplitude was plotted against the membrane prepulse potential (Fig. 3C). The fit
of the data with the Boltzmann Eq. 3 yielded a
V0.5 of
30 ± 1 mV and a slope
factor of K = 15 ± 1.
Pharmacological properties of IBa
LVA ICa (or
IBa) that was sensitive to amiloride
have been recorded in cockroach abdominal ganglionic neurosecretory DUM
neurons (Grolleau and Lapied 1996),
Drosophila CNS embryonic neurons (Baines and Bate
1998
), and Drosophila larval muscle (Gielow
et al. 1995
). To determine whether an equivalent LVA
IBa was absent from our preparation,
we tested the effect of bath superfusion of the neurons with 1 mM
amiloride for
2 min. The current was not significantly reduced
(90 ± 17%, n = 6 neurons, of the control peak
current; data not shown). In some preparations there appeared to be a
gradual decrease in the peak amplitude of
IBa. This, however, turned out to be
statistically insignificant. The absence of effect of amiloride could
not be attributed to the slow perfusion rate, because other pharmacological agents acted rapidly (for example, see effects of
verapamil described later). We therefore concluded from the voltage-dependent activation and pharmacological properties that IBa was an HVA calcium channel current.
To identify the type of channel carrying this HVA IBa, we used agents that selectively block individual classes of both vertebrate and invertebrate VDCC.
HVA L-type currents are characterized by their selective sensitivity to
DHPs and PAAs (for review see Bean 1989; Hess
1990
). After a
2-min superfusion of the neurons with the DHP
agonist Bay K 8644 (1 µM, n = 7 neurons) and the
antagonist isradipine (10 µM, n = 7 neurons), the peak
amplitude of IBa was not significantly changed. In the presence of the DHPs the peak amplitudes of
IBa were 88 ± 6% (n = 7 neurons) and 91 ± 15% (n = 7 neurons) of the control
values with Bay K 8644 and isradipine, respectively (data not shown).
However, larger DHP concentrations (100 µM nifedipine) were found to
partially block IBa (data not shown).
HVA calcium currents of several insect neurons exhibit a component that
is blocked by PAA but is insensitive to DHP (Bickmeyer et al.
1994a; Pearson et al. 1993
; Pelzer et al.
1989
; Wicher and Penzlin 1997
). Superfusion of
neurons with the PAA verapamil in the bath solution at concentrations
<10 µM had no significant effect on the peak amplitude of
IBa. A block was seen at
concentrations starting at 10 µM (Fig.
4B). For concentrations that
were just suprathreshold (50 µM), the reduction of the current at the
end of an 800-ms test pulse was often larger than that observed at the
peak amplitude of the current (Fig. 4, A and C;
see also Wicher and Penzlin 1997
: Fig.
5A). This effect is
reminiscent of the open-channel-blocking properties of the PAAs (for
review see Hondeghem and Katzung 1984
). The effect of
verapamil on IBa peak amplitude was
dose-dependent (Fig. 4B) and could be fitted using the Hill equation (see METHODS). This fit yielded an
IC50 of 170 µM and a Hill
coefficient of 0.96. A near complete block was achieved at millimolar
concentrations. At all tested concentrations, a steady-state block was
achieved within 40 s, and the effect of verapamil was at least
partially reversed by washing. At our usual HP (
70 mV) the blocking
effect of verapamil was independent of the test membrane potential
(Fig. 4D). That IBa was
insensitive to DHPs and weakly sensitive to PAA suggests that it could
be different from an L-type IBa.
|
|
In vertebrate neurons the toxins -CgTxGVIA,
-CmTxMVIIC, and
-AgaTxIVA are specific blockers of N-, Q-, and P- type currents, respectively. These toxins were tested using two protocols.
Bath perfusion (superfusion protocol) of 1 µM -CgTxGVIA or 1 µM
-CmTxMVIIC for
5 min left the peak amplitude of
IBa virtually unchanged (not shown).
This was confirmed by coapplication to the same neuron of 1 µM
-CgTxGVIA and 1 µM
-CmTxMVIIC (n = 8 neurons; Fig.
5, A and B). However, subsequent
superfusion of the same neurons with 200 nM
-AgaTxIVA resulted in a significant reduction of
IBa peak amplitude after 1 min (Fig.
5, A and B). For a population of 26 neurons, we
found that concentration of
-AgaTxIVA
100 nM, had no detectable
effect on 2 neurons, a hardly detectable effect on 3, an effect of
between 25 and 80% on 8, and an effect >80% on 13. This result
illustrates some heterogeneity of the neurons regarding their
sensitivity to the toxin with a clear continuum between neurons that
are barely (19%, n = 5 neurons) and neurons that were
highly (50%, n = 13 neurons) sensitive to the toxin.
Considering the neurons with clear sensitivity to the toxin, the
average steady-state reduction of
IBa peak amplitude for concentrations
of
-AgaTxIVA
100 nM was 74 ± 4% (n
= 21 neurons). The effect of
-AgaTxIVA was not reversed after a
5-min washing (data not shown) and was not reversed by depolarizing
prepulses (Fig. 5B) in contrast with the observation of
Mintz et al. (1992a)
in rat neurons. The persistence of
a residual current in high concentrations of
-AgaTxIVA (
100 nM)
suggests the presence in these neurons of an
-AgaTxIVA-resistant
current component.
To reduce the amount of toxin used for the experiment and avoid the
run-down of IBa, we used a second
protocol. In this (incubation) protocol, the neurons were incubated
with the toxins before any recording for 10 min. At that time we
assumed that the effect of the toxins had reached a steady state, at
least in large concentrations. The peak amplitude of
IBa was measured for 5-10 neurons per
concentration and per culture dish. This protocol had the advantage of
reducing the number of culture dishes needed per experiment but had the minor drawback of increasing the values of the SE (Fig. 5, C
and D). The effect of the toxins was estimated from a
comparison of maximum peak amplitudes of
IBa (obtained from the I-V
curves) of treated and untreated (control) neurons of the same age
originating from the same culture. For the experiments illustrated in
Fig. 5C, the average maximum amplitude of the current in 1 µM
-CgTxGVIA was
195 ± 22 pA (n = 6 neurons)
compared with
214 ± 52 pA (n = 6) in control
neurons, the average maximum amplitude of the current in 1 µM
-CmTxMVIIC was
192 ± 30 pA (n = 19 neurons) compared with
158 ± 17 pA (n = 24 neurons) in
control neurons, and the average maximum amplitude of the current in
400 nM
-AgaTxIVA was
15 ± 8 pA (n = 7 neurons)
compared with
149 ± 40 pA (n = 5 neurons) in control
neurons. In the first two cases, the difference between the treated and
untreated neurons was not statistically significant (P = 0.4 for 1 µM
-CgTxGVIA and P = 0.08 for 1 µM
-CmTxMVIIC). On the contrary, the blocking effect of 400 nM
-Aga-IVA (81 ± 5%, n = 7 neurons) was highly
significant (P = 0.002).
These results are in perfect agreement with those obtained by
superfusion: IBa is insensitive to
micromolar concentrations of -CgTxGVIA and
-CmTxMVIIC, and part
of the current is highly sensitive to
-AgaTxIVA. Furthermore, these
experiments showed that most neurons exhibited two current components
of IBa: an
-Aga-IVA-sensitive
component that represented ~80% of the macroscopic current and a
DHP-,
-CgTxGVIA-, and
-AgaTxIVA-resistant current component
representing ~20% of the total current.
The incubation protocol was used to study the dose-response
relationship for -AgaTxIVA. As illustrated in Fig. 5D,
the effect of this toxin was dose-dependent. The parameters of the fit
of this curve with the Hill equation were, respectively, 17 nM for the
IC50 and 0.6 for the Hill coefficient
n (Fig. 5D, dashed line). This low value of the
Hill coefficient in addition to the fact that the block was incomplete
at high toxin concentrations suggests that the overall response was
made of the combination of at least two populations of calcium channels
with clearly different affinities for the toxin. This hypothesis was
tested, and it was found that the data points could be equally well
fitted with the combination of two Hill equations with a Hill
coefficient of 1 and an IC50 of,
respectively, 9 nM and 900 nM, corresponding to high-affinity and
low-affinity components (Fig. 5D, straight line). The
low-affinity
-AgaTxIVA component obtained from the fit represented
20% of IBa and could be suggestive of
its nonspecific effect on the
-AgaTxIVA-resistant component at
higher concentrations.
In the next series of experiments, we used the incubation protocol to
compare the time course and the voltage-dependent activation properties
of the barium current before and after a 10-min exposure to 400 nM
-Aga-IVA, which completely blocks P/Q-type calcium currents in
vertebrate neurons (Mintz et al.1992a
; Zhang et
al. 1993
). We found that the kinetics of the current for
test-potential values ranging from
40 to +40 mV were
apparently not different between treated and untreated neurons (data
not shown) and that the current-voltage relationship of the residual
peak IBa was not altered by the toxin
(Fig. 6C). Similar results
were obtained using the superfusion protocol in conditions under which
the residual currents were larger and could therefore be analyzed with
more precision. The kinetics were apparently not altered, as
illustrated in Fig. 6A, for a depolarization to
+5 mV. The
-AgaTxIVA-sensitive component had the same
inactivation kinetics as the current recorded before the superfusion.
The I-V curve was not significantly modified by
-AgaTxIVA
superfusion, as illustrated in Fig. 6B, and the percentage
of block was found to be insensitive to membrane potential between
30
and +20 mV (data not shown). We concluded from these experiments that the IBa of embryonic
cockroach brain neurons possesses
-AgaTxIVA-sensitive and
-resistant components bearing similar voltage-dependent properties.
The distribution of the two components varied from neuron to neuron.
|
It is well known that, in vertebrate CNS neurons, the
-AgaTxIVA-sensitive current components are also sensitive to
-CmTxMVIIC. The effect of this latter toxin could need an incubation
time of
30 min to reach steady state (Sabatier et al.
1997), and therefore the absence of effect seen with 1 µM of
-CmTxMVIIC could reflect an insufficient incubation time. To verify
this possibility, we decided to probe the effects of the application of
the usual 1-µM concentrations of
-CmTxMVIIC but for a much longer
period (1-h incubation). Under these conditions, the mean peak
amplitude of the current was reduced from 204 ± 37 pA (n
= 6 neurons) for untreated neurons to 126 ± 19 pA (n
= 8 neurons), a 38% reduction (data not shown). The blocking
effect of
-CmTxMVIIC was not accompanied by a significant
modification of the activation and inactivation kinetics or of the peak
current-voltage relationship (data not shown).
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Our patch-clamp study of the barium currents provide new and interesting results concerning the biophysical and pharmacological properties of the voltage-dependent calcium channels of embryonic cockroach brain neurons. The main findings can be summarized as follows.
The voltage-dependent activation and inactivation properties of IBa in these neurons were typical of HVA calcium currents seen in other preparations. Their pharmacological properties, which were studied quantitatively using both fast-bath superfusion and incubation techniques, were original in several respects. The current was not significantly affected by DHP agonists or antagonists but was efficiently and reversibly reduced by verapamil, a PAA.
The toxin -CgTxGVIA had no significant effect, micromolar
concentrations of
-CmTxMVIIC partly reduced the peak amplitude of
the current after a 1-h incubation, and nanomolar concentrations of
-AgaTxIVA irreversibly blocked the major part of the current in the
majority (81%) of the tested neurons (in which 400 nM
-AgaTxIVA reduced the current by 80%). A small proportion of the neurons (19%)
were significantly less sensitive to that toxin (their barium current
was unaffected by 100 nM toxin).
Biophysical properties: HVA versus LVA currents
Our experiments clearly demonstrate that, in embryonic cockroach
neurons, the macroscopic barium current activates at potential values
more positive than 50 mV, in agreement with previous observations on
the same preparation (Christensen et al. 1988
) and on
adult cockroach brain neurons (Amar and Pichon 1992
).
Changes in the HP between
100 and
60 mV failed to modify this
threshold or the time course of the current and did not shift the
I-V relationship. Furthermore, amiloride which has been
shown to inhibit LVA currents in other insect preparations
(Baines and Bate 1998
; Gielow et al.
1995
; Grolleau and Lapied 1996
) did not modify
the current. Conversely, agents that are known to modify HVA currents
in other preparations, namely
-AgaTxIVA and verapamil, blocked the
current in embryonic cockroach brain neurons. Altogether, these results strongly suggest that embryonic cockroach brain neurons in primary culture express only HVA calcium channels. They differ in that respect
from the neurosecretory DUM neurons of this same species which also
exhibit one or two LVA/M-LVA-activated components (Grolleau and
Lapied 1996
; Wicher and Penzlin 1997
) that are
likely to play an important role in the generation of their pacemaker
activity. They also differ from developing neurons from
Drosophila embryos (Baines and Bate 1998
).
The origin of this discrepancy remains speculative. It is tempting,
however, to suggest that it may be because our cultures were grown in
high-potassium media (see METHODS) in which the
cells were likely to have a low resting potential (between
40 and
20 mV, Lees et al. 1985
) and that at these potential levels, which are known to inactivate LVA currents (Baines and Bate 1998
; Grolleau and Lapied 1996
;
Wicher and Penzlin 1997
), the expression of LVA currents
is repressed.
Sensitivity to DHPs, PAAs, and -CgTxGVIA
Our observation that the barium current was not significantly
affected by micromolar concentrations of the DHP isradipine (10 µM)
or Bay K 8644 (1 µM) but was blocked by the PAA verapamil is
consistent with earlier observations on insect neurons
(Bickmeyer et al. 1994a; Pearson et al.
1993
; Pelzer et al. 1989
; Wicher and
Penzlin 1997
). In this respect, insect HVA calcium channels differ from their vertebrate L-type counterparts, which are sensitive to both DHP and PAA. As previously shown by Mills and Pitman
(1997)
, larger DHP concentrations (100 µM) block
IBa, but this effect is considered
nonspecific. That a relatively large concentration of verapamil was
needed (IC50 of 170 µM) to block the
current is consistent with the observations of Byerly and Leung
(1988)
on Drosophila embryonic neurons and suggests
that the very-high-affinity PAA-binding sites detected by Pauron
et al. (1987)
in Drosophila head membranes
(Kd = 4.7 pM for [3H]-verapamil) may not correspond to the channel-blocking sites in native membranes (see also
Pelzer et al. 1989
). Low-affinity blocking sites have
also been observed in some vertebrate preparations in which non-L-type
HVA calcium currents have been blocked by supramicromolar
concentrations of verapamil (Diochot et al. 1995
;
Ishibashi et al. 1995
).
The insensitivity of the embryonic cultured neurons to -CgTxGVIA, a
selective blocker of the neuronal vertebrate N-type calcium channels
(McCleskey et al. 1987
) that also blocks the HVA
component of the calcium current in insect adult DUM neurons
(Grolleau and Lapied 1996
; Wicher and Penzlin
1997
), is an interesting feature of our preparation. These
differences among various neuronal populations in the same species
could result from cell-specific differential expression of the
channels, which ought to be studied in more detail. It could also
reflect the growth conditions in vitro or a difference between
embryonic and adult neurons.
Sensitivity to -AgaTxIVA: evidence for two populations of HVA
calcium channels
The experiments clearly indicate that most neurons (81%) were
sensitive to -AgaTxIVA but that a fraction of the current (varying between 10 and 75%, mean = 25%) in these neurons was
resistant to the toxin. Furthermore, a significant proportion of the
neurons (19%) was much less sensitive to the toxin. These results
strongly suggest that the total current was associated with the opening of two different populations of HVA calcium channels, the proportion of
which varied from neuron to neuron. The first (
-AgaTxIVA-sensitive) would be of the P/Q type (i.e., resistant to DHP and
-CgTxGVIA), the
second of the R type (i.e., resistant to all). Interestingly, the
dose-response curve of the
-AgaTxIVA effect could be fitted with a
combination of two Hill equations assuming a low
(IC50 = 9 nM) and high affinity
(IC50 = 900 nM) for the toxin. The
low-affinity binding site would correspond to the nonspecific effect of
-AgaTxIVA on the R-type calcium channels when higher concentrations
of the toxin were used (
100 nM).
Here again, our results differ in several respects from those reported
for other insect preparations. Thus, whereas an -AgaTxVIA-sensitive component has also been observed in adult locust neurosecretory cells
(Bickmeyer et al. 1994b
), this component was present in only 30% of the neurons (compared with 81% in our experiments) and,
importantly, the blocking effects were reversed by a brief train of
strong depolarizations (as shown earlier by Mintz et al.
1992a
in rat central and peripheral neurons).
-AgaTxIVA also blocked part of the barium current in cockroach DUM neurons, but
-AgaTxIVA was less potent than
-CmTxMVIIC, its effects were reversed after washing out the toxins, and the fraction of the current
affected by
-AgaTxIVA had the activation and inactivation properties
of an M-LVA current (Wicher and Penzlin 1997
).
The pharmacological properties of the -AgaTxIVA-sensitive component
of the embryonic cockroach brain neurons resemble that of the
vertebrate P-type calcium currents (Llinas et al. 1989
; Mintz et al. 1992a
) rather than that of the Q-type
calcium channel (Sather et al. 1993
; Wheeler et
al. 1994
). However, this current component differs from the
vertebrate P-type calcium current in the nonreversibility of the
blocking effect of
-AgaTxIVA after application of conditioning
depolarizations. In contrast with data reported in vertebrates for the
P/Q-type and R-type, the two components of the barium current exhibited
indistinguishable voltage-dependent activation and inactivation
properties. It would be interesting to see whether this is also true at
the single-channel level.
Physiological significance of the P/Q and R-type currents
The precise role of the calcium current in the embryonic cockroach
brain neurons in primary culture remains speculative. Under normal
conditions (i.e., in cockroach saline), these neurons, similar to most
insect neurons, have been shown to be electrically silent, partly
because of the small size of the TTX-sensitive sodium current, and
partly because of its very large delayed rectifier current (see
Christensen et al. 1988).
In the culture medium, however, the external potassium concentration
was much higher: 16 mM in the first medium and 30 mM in the second
medium (see METHODS). Under these conditions, the neurons
were depolarized (estimated resting potentials of ~ 40 and
20 mV; see Lees et al. 1985
). The window potential of
the barium current that can be calculated from the steady-state
activation and inactivation curves (Fig. 2C) lay between
40 mV and +20 mV in 10 mM Ba2+. In the culture
dish, there was no barium, and the calcium concentration was ~2 mM.
Under these conditions, the activation and inactivation curves were
displaced by ~10 mV toward more hyperpolarized potentials, and the
window current was shifted (between
50 and +10 mV), a potential range that enabled calcium to enter the cell. One role for
the observed calcium current could be to enable survival and differentiation of the neurons in culture. This hypothesis is strengthened by the report that, in vertebrate cultures, a better survival after growth factor deprivation is obtained in
high-K+ solutions, which induces membrane
depolarization and thereby activates voltage-dependent calcium channels
(Collins et al. 1991
; Koike et al. 1989
;
for review see Finkbeiner and Greenberg 1998
). We
observed, in agreement with Beadle and Hicks (1985)
,
that our second culture medium (L + G), which contained more potassium, enabled better development of the neurons than the first medium (5 + 4). The recent findings that CNS neurons during the early embryonic
development of Drosophila express voltage-dependent calcium
channels (Baines and Bate 1998
) and that the mutation of
the genes encoding for these channels is lethal (Smith et al. 1996
), also suggest that calcium channels play a prominent role during embryogenesis. Interestingly, preliminary experiments showing that verapamil induces the death of all cultured neurons in 72 h
also suggest that calcium channels are important for survival and
growth of embryonic cockroach neurons. Because the R-type calcium
channels are present in all neurons, it is tempting to suggest that
they could be involved in neuronal survival and differentiation shared
by all neurons, whereas the P/Q type channels, which are expressed in
only a subset of neurons, are implicated in another cellular function.
![]() |
ACKNOWLEDGMENTS |
---|
We thank Drs. F. Grolleau, B. Lapied, and S. Richard for comments on the manuscript.
![]() |
FOOTNOTES |
---|
Address for reprint requests: F. Tiaho, Equipe Canaux et Récepteurs Membranaires, UPRES-A 6026, Campus de Beaulieu, Bât. 13, 35042 Rennes Cedex, France.
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 5 March 1999; accepted in final form 7 July 1999.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|