FEBS 15221 FEBS Letters 362 (1995) 15-18
o)-Agatoxin IVA blocks nicotinic receptor channels in
bovine chromaffin cells
Ricardo Granja, Jos6 M. Fern/mdez-Fernfindez, Victor Izaguirre, Carmen Gonz/dez-Garcia,
Departamento de Farmacologia and Instituto de Neurociencias, Universidad de Alicante, Apdo. Correos 374, E-03080 Alicante, Spain
Received 25 November 1994; revised version received 2 February 1995
Abstract We have studied the contribution of P-type voltage-
dependent Ca 2+ channels to both catacholamine (CA) and ATP
secretion from bovine chromaffin cells induced by high K ÷ or
nicotine using ¢o-agatoxin IVA, a selective blocker of P-type
voltage-dependent Ca 2÷ channels. We found that high K ÷ (75
raM) induced the release of about 13% of norepinephrine, 5%
epinephrine and 11% ATP, and that ¢o-agatoxin (100 nM) did not
affect this secretion. However, both nicotine-induced CA and
ATP secretion were significantly blocked (about 50%) by
¢o-agatoxin IVA (100 nM). In addition, this toxin also reversibly
blocked (about 70%) the inward current induced by nicotine in
bovine chromaffin cells. The results suggest that, besides its
known action of blocking P-type voltage-dependent channels,
¢o-agatoxin is a potent and reversible blocker of the nicotinic
receptor channel in chromaffin cells, and that this action would
explain the blockade of nicotine-induced secretion.
Key words: Calcium channel; P-type; Nicotinic receptor;
(o-Agatoxin IVA; Secretion; Chromaffin cell
Adrenomedullary chromaffin cells secrete catecholamines in
response to nicotinic cholinergic agonists [1-3]. Following acti-
vation of the nicotinic receptor, a non-selective cation channel,
associated to the receptor, opens [4,5], and the membrane of
chromaffin cells is depolarized [6-8]. This depolarization leads
to the opening of voltage-dependent Na + and Ca 2+ channels
located on the plasma membrane that elevate [Ca2+]i, which
induces catecholamine (CA) secretion. Using electrophysiologi-
cal techniques, the presence of L- [9,10], N-  and P-  type
voltage-dependent Ca 2+ channels located on the plasma mem-
brane of chromaffin cells has been demonstrated.
P-Type voltage-dependent Ca 2+ channels have been de-
scribed to carry a significant amount of voltage-activated Ca 2+
current in different neural cells [13-15], and have also been
implicated in neurotransmitter secretion in different systems
[16,17]. In adrenergic tissues, P-type voltage-dependent Ca 2+
channels seem to contribute significantly to Ca 2+ currents in
response to depolarization . However, its participation in
the secretory process is less clear, although a fraction of the
funnel-web spider poison (FTX) can partially block both the
increase in intracellular Ca 2+ levels and CA secretion in re-
sponse to K+-induced depolarization .
*Corresponding author. Fax: (34) (6) 565 5218.
In the present work, we have attempted to characterize the
contribution of P-type voltage-dependent Ca 2+ channels to CA
and ATP secretion from bovine chromaffin cells induced by
either depolarization with high extracellular K + or by nicotinic
receptor activation, by using the toxin o)-agatoxin IVA (o)-
AgaTx), obtained from a peptide fraction of the funnel-web
spider, Agenelopsis aperta, that is considered to be a specific
blocker of P-type voltage-dependent Ca 2+ channels .
2. Materials and methods
2.1. Cell culture
Chromaffin cells were isolated as previously described  with
minor modifications. Briefly, glands were perfused through the adre-
nolumbar vein with CaZ÷-free Locke's solution of the following compo-
sition (in mmol/l): NaC1, 154; KCI, 5.6; MgC12, 1; NaH2PO4, 3; HEPES,
10; and glucose, 10. The pH of the solution was adjusted to 7.4 with
NaOH. The glands were then digested with Locke's solution containing
0.2% collagenase (Boehringer-Mannheim, Indianapolis, IN) and 0.5%
bovine serum albumin (Calbiochem, La Jolla, CA) and incubated for
45 min at 37°C. Next, the medulla was separated from the cortex,
minced with scissors and further digested in the presence of collagenase
at 37°C for an additional 30 min. After filtering through a nylon mesh,
chromaffin cells were separated from erythrocytes using a Percoll gra-
dient. The yield varied between 20 and 30 x 106 cells per gland and the
viability was greater than 95% as defined by exclusion of the biological
dye Trypan blue. Cells were cultured in Dulbecco's modified Eagle's
medium (DMEM) supplemented with 10% fetal calf serum (FCS) under
an atmosphere of 5% CO2.
2.2. Catecholamine secretion
After 3-7 days of culture, chromaffin cells were washed for 60 rain
(12 x 5 min) with 800/.d Krebs-HEPES (K-H) of the following compo-
sition (in mmol/1): NaCI, 140; KC1, 5.9; MgSO4, 1.2; CaC12, 2.5;
HEPES, 15; and glucose, 11; pH was adjusted to a value between 7.3
and 7.4. After washing, a 5 min sample was taken to determine basal
CA release. The solution in the well was changed, for 5 min, to one
containing the desired secretagogue, either high K ÷ or nicotine. A final
5 min sample was taken following the 5 min period of stimulation.
Samples were kept on ice-cold water and quickly acidified with per-
chloric acid (PCA) to a final concentration of 0.05 N. To determine
total CA cell content at the end of the experiment, 800 ¢tl of PCA
(0.05 N) was added to each well. High K ÷ solution was prepared by
isosmotically substituting NaC1 by KCI. o)-Agatoxin IVA, when pres-
ent, was added 5 min before, and maintained through, the stimulus.
Fractional release was calculated by dividing the total amount of either
norepinephrine (NE) or epinephrine (E) released during a 5 min period
by the total amount of that amine present at the beginning of that
particular period of time. Individual net fractional release was calcu-
lated for each period either during the secretory stimulus or 5 min after
the stimulus, by subtracting the release during the basal 5 min period.
Total net fractional release was calculated by adding individual net
fractional releases during and 5 rain after K ÷ stimulation.
CA release was determined using a HPLC system with electrochem-
ical detection, provided with two Gilson pumps (model 303), two
Gilson automatic injectors, two amperometric detectors (Model LC4B,
Bioanalytical Systems, West Lafayette, IN and model 656, Metrohm,
0014-5793/95/$9.50 © 1995 Federation of European Biochemical Societies. All rights reserved.
R. Granja et al./FEBS Letters 362 (1995) 15-18
Switzerland with voltages set at 0.65 V and 0.9 V, respectively). Signals
recorded from the detectors were integrated, the area under the peaks
measured and the CA concentration calculated from internal standards
using a Gilson software (model 714). The precolumn and the analytical
columns were identical in both systems and consisted of a replaceable
cartridge reverse-phase C]s precolumn (5 ,urn particle size, 30 × 4 mm
i.d.; Macherey-Nagel, Duren, Germany) and an analytical reverse-
phase C]8 replaceable cartridge column (5/.tm particle size, 100 × 4 mm
i.d.; Machinery-Nagel, Duren, Germany). Flow was set at 1 ml/min and
100 pl samples were injected. The mobile phase was similar to the one
previously described  and had the following composition (in
mmol/1): NaHzPO4, 80; EDTA, 0.13; heptane sulfonic acid, 5; methanol
10%, the pH was adjusted to 3 with orthophosphoric acid. Retention
times were 5.1 rain for NE and 7.9 min for E.
2.3. Electrophysiological recording
Recording of chromaffin cell nicotinic receptor currents under volt-
age-clamp was done as previously described  with some modifica-
tions. Chromaffin cells were bathed in a solution with the following
ionic composition (in mmol/1): NaCl, 140; KCI, 5.9; CaCl2, 2.5; MgCh_,
1; HEPES, 10; glucose, 11 (pH 7.4). The ionic composition of the
pipette solution was the following (in mmol/1): KCI, 40; K2SO4, 50;
MgSO4, 7; HEPES, 10 and nystatin, 250/lg/ml (pH 7.2). A high resis-
tance seal was established in the cell-attached configuration (open tip
resistance of the pipette ranged between 3 and 5 Mr2). Capacity tran-
sients were canceled using the built-in circuitry of the patch-clamp
amplifier (EPC-7; List, Darmstadt, Germany) and the pipette potential
set at -55 mV. After 2 or 3 min a small capacity transient indicating
electrical continuity between the cell interior and pipette solution was
observed. The size of the transient stabilized about 5 min after forma-
tion of the gigaseal. Nicotine (10pM) was applied for 3.5 s using a fast
perfusion system (DAD-12; Adams & List, NY) 10 min after gigaseal
formation. Nicotine (10 pM) application was repeated 3 times (S~, $2,
$3) with an interval of 15 min. The average capacitance of the cells was
11.1 + 0.99 pF and the series resistance (Rs) 9.8 + 0.9 M/2 (n = 12). No
compensation for Rs was used. When used, co-AgaTx or D-tubocurarine
were perfused to the cell 5 s before and during $2.
2.4. ATP determination
ATP release in response to high K + or nicotine was determined using
the luciferin luciferase method as previously described . Total ATP
content in the cell was determined by solubilizing the cells using Ex-
tralight (Analytical Luminescence, San Diego, CA).
Culture reagents were obtained from Gibco. Nicotine was purchased
¢ m 6
NE E ATP
Fig. 1. Effect of to-agatoxin IVA on high K+-induced NE, E and ATP
secretion. Secretion of norepinephrine (NE), epinephrine (E) and ATP
after exposure of bovine chromaffin cells to high extracellular K + (75
mM) for 5 min in the absence (open bars) and in the presence (filled
bars) of to-agatoxin (100 nM). Data represent mean _+ S.E.M. of
7 experiments for NE and E secretion and 28 for ATE
NE E ATP *"
Fig. 2. Effect of m-agatoxin IVA on nicotine-induced NE, E and ATP
secretion. Secretion of norepinephrine (NE), epinephrine (E) and ATP
after exposure of bovine chromaffin cells to nicotine (10/~M) for 5 rain
in the absence (open bars) and in the presence (filled bars) ofto-agatoxin
IVA (100 nM). Data represent mean + S.E.M. of 8 experiments.
*P < 0.01 as compared to control.
from Sigma and co-Agatoxin IVA was obtained from RBI (Natick,
MA). All other reagents were obtained from commercial sources and
were of the highest quality available.
3. Results and discussion
Two different populations of chromaffin cells preferentially
secreting NE and E have been described in bovine adrenal
medulla [21,22]. Exposure of bovine chromaffin cells in culture
to high extracellular K ÷ (75 mM) induced the release of about
13% of the total NE and 5% of total E content in the cells
(n = 7) (Fig. 1). High K + (75 mM) induced the release of 12%
of total ATP present in chromaffin cells (n = 28). There is a
good correlation between CA and ATP secretion, as would be
expected since the more likely source of released ATP in re-
sponse to depolarization is the chromaffin granule where it is
co-stored with CA [23,24]. In the presence of 100 nM of m-
AgaTx, a dose high enough to completely block P-type voltage-
dependent Ca 2+ channels , CA or ATP secretion induced by
high K ÷ (75 mM) were not affected (Fig. 1). This suggests that
although about 50% of the Ca 2+ entering bovine chromaffin
cells following a depolarizing pulse is sensitive to co-AgaTx ,
P-type Ca 2+ channels are not involved in either CA or ATP
secretion in response to depolarization.
The physiological stimulus for CA and ATP secretion in the
adrenal medulla is acetylcholine that activates cholinergic nico-
tinic receptors . Exposure of bovine chromaffin cells to the
nicotinic cholinergic agonist nicotine (10 pM) induced the se-
cretion of 30 + 1.8% of total NE and 17.9 + 1.6% of total E
present in chromaffin cells at the beginning of the stimulation
(n -- 8) (Fig. 2). Nicotine treatment also induced the release of
8.4 + 0.6% of total ATP present in the cells (n = 8) (Fig. 2).
m-AgaTx (100 nM) decreased by about 40% nicotine-induced
NE and E release (Fig. 2). In addition, nicotine-induced ATP
secretion was significantly blocked (about 60%) by m-AgaTx
(100 nM) (Fig. 2). These results suggest that voltage-dependent
P-type Ca 2+ channels are selectively coupled, in bovine chro-
R. Granja et al./FEBS Letters 362 (1995) 15-18
_, J , J
Fig. 3. Effect of co-agatoxin IVA on nicotine-induced currents. Inward current induced by 3 consecutive applications of nicotine (10/IM) with an
interval of 15 min as described in section 2. The second application of nicotine was done in the presence of to-agatoxin IVA (100 nM). The figure
represents a typical experiment that was repeated 4 times with similar results.
maffin cells, to nicotine-induced secretion. This indicates that
P channels are located in the vicinity of the nicotinic receptor
and so would be preferentially activated by nicotinic stimula-
tion. This suggests that P-type voltage-dependent Ca 2÷ chan-
nels may play a role in secretion in chromaffin cells similar to
other systems [26,27]. However, since co-AgaTx did not affect
high K+-induced secretion, an effect of the toxin on a target
different from voltage-dependent P-type Ca 2÷ channels should
Nicotine-induced secretion from chromaffin cells begins by
activation of nicotinic receptors that increase the permeability
to Na ÷ and K + ions and very little, under physiological concen-
trations of extracellular Na +, to Ca 2+ . This change in perme-
ability depolarizes chromaffin cells and produces opening of
voltage-dependent Ca 2+ channels. Although og-AgaTx is con-
sidered a highly selective blocker of P-type voltage-dependent
Ca 2+ channels , we decided to investigate the effect of this
toxin on the current induced by activation of nicotinic receptors
to exclude the possibility that the nicotinic receptor could be
a primary target for co-AgaTx. If this is the case it could explain
the selective inhibitory action of the toxin on nicotine-induced
secretion without affecting high K÷-induced secretion.
To explore this possibility we decide to use nystatin-perfo-
rated patch-clamp recordings to preserve physiological regula-
tion of the cell, and use a holding potential of -55 mV that is
Effect of agatoxin IVA and D-tubocurarine on nicotine-induced inward
currents and net charge influx in bovine chromaffin cells
Drug added together
with nicotine during $2 (S2/St ratio)
Peak current Total net charge
($2/S 1 ratio)
0.55 + 0.05
0.39 + 0.05*
0.15 _+ 0.06**
0.18 + 0.04**
0.09 + 0.04***
og-AgaTx 1 nM
to-AgaTx 10 nM
¢o-AgaTx 100 nM
D-Tubocurarine 10 ktM
Bovine chromaffin cells were exposed for 3.5 s to nicotine (10/.tM) as
stated in section 2. Nicotine stimulation was repeated 3 times (S~, $2,
$3) at 15 min interval, o-Tubocurarine or og-AgaTx IVA were perfused
to the cell 5 s before and during the second (Sz) nicotine stimulation.
The average value for inward peak current and total net charge influx
during S~ were -204 + 25 pA (n = 23) and 256 + 19 pC (n = 23). Data
represent mean + S.E.M. *P < 0.05; **P < 0.005; ***P < 0.001 as
compared to control.
0.81 _+ 0.12
0.48 _+ 0.06*
0.26 + 0.07**
0.34 _+ 0.04**
0.10 _+ 0.02***
the resting membrane potential measured under current-clamp
conditions (data not shown). Perfusion of bovine chromaffin
cells under voltage-clamp with nicotine (10/IM) for 3.5 s in-
duced an inward current that amounted to -204 + 25 pA
(n = 23) (Fig. 3). That current inactivated quickly with a time
constant of about 1 s. Peak current was very little affected by
a second exposure of the cell to nicotine (10/~M), with a SJS1
ratio of 0.81 + 0.12 (Table 1). However, in the presence of
o~-AgaTx, peak current was markedly decreased in a dose-
dependent manner ($2/$1 ratio of 0.34 + 0.04 in the presence of
co-AgaTx 100 nM; Fig. 3). In addition, nicotine-evoked inward
current was almost completely blocked by the nicotinic recep-
tor antagonist D-tubocurarine (Table 1). The time integral on
the current gives the total amount of charge entering the cell
during the first nicotine stimulation (S 0 as 256.5 _+ 19.2 pC
(n = 23). When nicotine perfusion was repeated 15 min later,
about 35% of inactivation was observed (SJS~ ratio for the total
charge of 0.55 + 0.05). However, when, after an initial exposure
to nicotine (10/.tM) for 3.5 s, a second perfusion with nicotine
was repeated 15 min later in the presence of og-AgaTx, a dose-
dependent blocking action of og-AgaTx could be observed
(Table 1). Consistent with its effect on peak current, the nico-
tinic receptor blocker D-tubocurarine blocked net charge influx
in response to nicotine. The blocking effect of both og-AgaTx
(Fig. 3) and D-tubocurarine (data not shown) could be partially
reverted by washing out the toxin or the drug.
The data indicate that besides its known action as a blocker
of P-type voltage-dependent Ca 2+ channels [14,25], og-AgaTx is
a very potent blocker of the neuronal cholinergic nicotinic
receptor present in bovine chromaffin cells and that its blocking
actions on nicotine-induced secretion might be explained, not
by a specific location of P-type voltage-dependent Ca 2+ chan-
nels close to nicotinic receptors, but rather by a specific revers-
ible blockade of agonist-induced current through the nicotinic
receptor channels. The blocking effect of og-AgaTx on nicotinic
receptor channels is shared by different compounds, including
somatostatin , methoxyverapamil  and atropine ,
suggesting that nicotinic receptor function might be regulated
by multiple pathways.
Acknowledgements: This work was supported, in part, by grants from
Fundaci6n Salud 2000, Direcci6n General de Investigaci6n Cientifica
y T6cnica, Grant PM92-0112, and the European Economic Commu-
nity, Grant SC1"-CT91-0709. R.G. was supported by a University of
Alicante fellowship, J.M.F. by a CAM fellowship, and V.I. by a Gener-
alitat Valenciana fellowship.
R. Granja et al./FEBS Letters 362 (1995) 15-18
 Douglas, W.W. and Rubin, R.P. (1961) J. Physiol. 159, 40 57.
 Leszczyszyn, D.J., Jankowski, J.A., Viveros, O.H., Diliberto, E.J.,
Near, J.A. and Wightman, R.M. (1990) J. Biol. Chem. 265, 1473(~
 Cefia, V. and Rojas, E. (1990) Biochim. Biophys. Acta 1023, 213
 Zhou, Z. and Neher, E. (1993) Pfliigers Arch. 425, 511 517.
 Wada, A., Takara, H., Izumi, F., Kobayashi, H. and Yanagihara,
N. (1985) Neuroscience 15, 283-292.
 Biales, B., Dichter, M. and Tischler, A. (1976) J. Physiol. 262,
 Douglas, W.W., Kanno, T. and Sampson, S.R. (1967) J. Physiol.
 Nassar-Gentina, V., Pollard, H.B. and Rojas, E. (1988) Am. J.
Physiol. 254, C675-C683.
 Fenwick, E.M., Marty, A. and Neher, E. (1982) J. Physiol. 331,
 Cefia, V., Stutzin, A. and Rojas, E. (1989) J. Membrane Biol. 112,
 Bossu, J.L., De Waard, M. and Feltz, A. (1991) J. Physiol. 437,
 Albillos, A., Garcia, A.G. and Gandia, L. (1993) FEBS Lett. 336,
 Mintz, I.M., Adams, M.E. and Bean, B.E (1992) Neuron 9, 85 95.
 Mintz, I.M., Venema, V.J., Swiderek, K.M., Lee, T.D., Bean, B.P.
and Aadams, M.E. (1992) Nature 355, 82%829.
 Regan, L.J., Sah, D.W.Y. and Bean, B.E (1991) Neuron 6, 269
 Uchitel, O.D., Protti, D.A., Sanchez, V., Cherksey, B.D.,
Sugimori, M. and Llinfis, R. (1992) Proc. Natl. Acad. Sci. USA 89,
 Luebke, J.I., Dunlap, K. and Turner, T.J. (1993) Neuron 11,895-
 Duarte, C.B., Rosario, L.M., Sena, C.M. and Carvalho, A.P.
(1993) J. Neurochem. 60, 908-913.
 Gonz~lez-Garcia, C., Cefia, V., Keiser, H.R. and Rojas, E. (1993)
Biochim. Biophys. Acta Mol. Cell Res. 1177, 99-105.
 Kempf, E. and Mandel, P. (1981) Anal. Biochem. 112, 223-231.
 Moro, M.A., L6pez, M.G., Gandia, L., Michelena, E and Garcia,
A.G. (1990) Anal. Biochem. 185, 243-248.
 Marley, ED. and Livett, B.G. (1987) Neurosci. Lett. 77, 81-
 Weber, A. and Winkler, H. (1981) Neuroscience, 226%2276.
 Inoue, S. (1981) J. Gen. Physiol. 78, 43-61.
 Mintz, I.M. and Bean, B.P. (1993) Neuropharmacology 32, 1161-
 Turner, T.J., Adams, M.E. and Dunlap, K. (1992) Science 258,
 Turner, T.J., Adams, M.E. and Dunlap, K. (1993) Proc. Natl.
Acad. Sci. USA 90, 9518-9522.
 Inoue, M. and Kuriyama, H. (1991) Biochem. Biophys. Res. Com-
mun. 174, 750-757.
 Boehm, S. and Huck, S. (1993) Eur. J. Neurosci. 5, 1280-1286.
 Inoue, M. and Kuriyama, H. (1991) Pfltigers Arch. 419, 13-20.