CECA-70.

QXD

8/3/01 4:16 PM

Page 181

Research

Cell Calcium (2001) 30(3), 181190

2001 Harcourt Publishers Ltd
doi:10.1054/ceca.2001.0224, available online at http://www.idealibrary.com on

Stimulation by thimerosal of
histamine-induced Ca2
release
in intact HeLa cells seen with
aequorin targeted to the
endoplasmic reticulum
M. Montero, M. J. Barrero, F. Torrecilla, C. D. Lobatn, A. Moreno, J. Alvarez
Instituto de Biologa y Gentica Molecular (IBGM), Departamento de Bioqumica y Biologa Molecular y Fisiologa, Facultad de Medicina, Universidad de
Valladolid and CSIC, Valladolid, Spain

Summary The oxidizing thiol reagent, thimerosal, has been shown to activate reversibly the inositol 1,4,5-trisphosphate
(InsP3) receptor in several cell types. We have studied here the effects of thimerosal by monitoring the [Ca2
] inside the
endoplasmic reticulum (ER) of intact HeLa cells with targeted aequorin. We show that thimerosal produced little effects
on the ER-Ca2
-pump and only slightly increased the ER-Ca2
-leak in intact cells. Instead, thimerosal increased the
sensitivity to histamine of ER-Ca2
-release by about two orders of magnitude, made the response much more
prolonged at saturating histamine concentrations and enhanced both cytosolic and mitochondrial [Ca2
] responses to
histamine. Moreover, inhibition of ER-Ca2
release by cytosolic [Ca2
] microdomains was fully preserved and sensitive
to BAPTA-loading, and histamine-induced Ca2
release remained quantal in the presence of both thimerosal and
intracellular BAPTA. The effects of thimerosal were reversible in the presence of dithiotreitol, suggesting the possible
presence of a physiological redox regulatory mechanism. However, in permeabilized cells thimerosal potentiated InsP3induced Ca2
release but oxidized glutathione had no effect. In addition, thimerosal increased the [Ca2
]ER steady-state
level in permeabilized cells. Thimerosal partially inhibited also plasma membrane Ca2
extrusion and increased Ca2

(Mn2
) entry through the plasma membrane, both phenomena contributing to increase the steady-state cytosolic
[Ca2
]. Thimerosal-induced Ca2
entry was additive to that induced by emptying of the ER, suggesting that storeoperated Ca2
channels may not be involved. These results provide new insights on the mechanisms of activation and
inactivation of InsP3 receptors. 2001 Harcourt Publishers Ltd

INTRODUCTION
Ca2
-release from intracellular stores is mediated by
Ca2
-channels belonging to two main families, inositol
1,4,5-trisphosphate receptors (InsP3R) and ryanodine
receptors [1]. Activation of phospholipase C by extracellular agonists produces InsP3, which is the main activator of
Received 24 January 2001
Revised 28 April 2001
Accepted 4 May 2001
Published online 13 July 2001
Correspondence to: Javier Alvarez, Departamento de Bioqumica y Biol.
Mol. y Fisiologa, Facultad de Medicina, Ramn y Cajal, 7, E-47005
Valladolid, Spain. Tel.:
34 983 423 085; Fax:
34 983 423 588;
e-mail: jalvarez@ibgm.uva.es

InsP3R and triggers Ca2
-release through these channels.

However, the activity of InsP3R is not only dependent
on the concentration of InsP3, but can be modulated by
several mechanisms [2,3]. One of the most important
ones is the [Ca2
], particularly in the cytosolic side
([Ca2
]c) but also in the lumenal side [4,5]. Type I InsP3R
has a bell-shaped dependence on [Ca2
]c, so that concentrations above 300 nM stimulate Ca2
-release, but further
increase to above 12 M inhibits Ca2
-release [6,7]. This
biphasic mechanism appears to be very important to start
and maintain regenerative Ca2
-release phenomena such
as Ca2
-waves and Ca2
-oscillations [8]. In HeLa cells, we
have shown that this mechanism controls Ca2
release
induced by histamine in intact cells [9,10]. On the other
181

CECA-70.QXD

8/3/01 4:16 PM

Page 182

182 M Montero, MJ Barrero, F Torrecilla, CD Lobatn, A Moreno, J Alvarez

hand, the lumenal [Ca2
] has also been proposed to stimulate Ca2
release through InsP3R [4,5]. In fact, InsP3R
activate spontaneously in the presence of resting levels of
InsP3 when the Ca2
-stores are overloaded [11,12].
Activation of InsP3R is potentiated by some thiolreactive oxidizing agents such as oxidized glutathione or
thimerosal [1116]. Thimerosal has been shown to activate InsP3R by several mechanisms, including increase of
the affinity for InsP3 [12,13,16,17], and increase of the
mean open time of the channel and the conductance,
even at saturating InsP3 concentrations [15]. In many
cases, this activation has been shown to be reversible in
the presence of thiol-reducing agents such as dithiotreitol
[13,16], suggesting that the mechanism of activation may
involve reversible alkylation of critical sulfhydril groups.
The reversibility of this effect suggests that it could represent a physiological mechanism of redox modulation of
InsP3R. In fact, oxidized glutathione has been shown to
activate InsP3-induced Ca2
release in permeabilized
hepatocytes [11,12] and endothelial cells [14].
The investigation of the mechanisms of the effects of
thimerosal at different points of the Ca2
-homeostatic
machinery may provide important clues to understand
their modulation under physiological conditions. It is
obvious that, being a general thiol-oxidizing agent, thimerosal is probably acting on many cysteine residues, thus
producing many different effects. For example, regarding
to Ca2
-homeostasis, thimerosal has been shown to
inhibit the ER Ca2
-ATPase [13,18], although the effective concentration required was highly variable. We have
used here a new methodology that allows measuring
[Ca2
] inside the ER ([Ca2
]ER) of intact cells using ER-targeted aequorin [9,10]. This method is ideal to follow
specifically the effect of thimerosal on InsP3-induced
Ca2
-release, independently of any Ca2
fluxes occurring
at the plasma membrane. Moreover, it allows studying
also directly in intact cells the rate of Ca2
-pumping into
the ER, the steady-state [Ca2
]ER and the rate of Ca2
leak from the ER in the presence of thimerosal.

MATERIAL AND METHODS

[Ca2
]ER measurements
HeLa cell clones EM26 and EM56, producing ER-targeted
low-Ca2
-affinity mutated aequorin [19] were grown in
Dulbeccos modified Eagles medium supplemented with
10% foetal calf serum and 0.2 mg/ml G418. Cell clones
were plated onto 13 mm round coverslips. Before reconstituting aequorin, [Ca2
]ER was reduced by incubating
the cells for 510 min at 37C with the sarcoplasmic and
endoplasmic reticulum Ca2
-ATPase (SERCA) inhibitor
2,5-di-tert-buthyl-benzohydroquinone (BHQ) 10 M in
standard external medium containing (in mM): NaCl,
Cell Calcium (2001) 30(3), 181190

145; KCl, 5; MgCl2, 1; glucose, 10; HEPES, 10, pH 7.4, supplemented with 3 mM EGTA. Cells were then incubated
for 1 h at room temperature in standard medium containing 0.5 mM EGTA, 10 M BHQ and 1 M coelenterazine n.
The coverslip was then placed in the perfusion chamber
of a purpose-built thermostatized luminometer, and standard medium containing 1mM Ca2
was perfused to refill
the ER with Ca2
. For experiments with permeabilized
cells, the coverslip was placed in the perfusion chamber
and treated for 1 min with 100 M digitonin suspended in
intracellular medium containing (in mM): KCl, 130; NaCl,
10; MgCl2, 1; K3PO4, 1; EGTA, 0,2; ATP-Mg, 1; Hepes, 20,
pH 7. Then, intracellular medium containing a Ca2
EGTA buffer providing a [Ca2
] of 100 nM was perfused
to refill the ER. In some experiments, the 100 nM Ca2

buffer was included also in the permeabilization medium.
Measurements were performed at 22C and [Ca2
]ER values were calculated from the luminescence records using
a computer algorithm [20] which follows the calibration
curve reported before [10].
Mitochondrial [Ca2
] ([Ca2
]M) measurements
HeLa cells were grown in Dulbeccos modified Eagles
medium supplemented with 10% foetal calf serum. Cells
were plated onto 13 mm round coverslips and transfected
transiently using Fugene (Gibco) with a pCDNA 3.1 plasmid encoding mitochondrially targeted wild-type aequorin.
After 1824 h, aequorin was reconstituted by incubation
with 1 M coelenterazine for 1 h prior to the measurements. Experiments were carried out at 22C and [Ca2
]M
values were calculated as described above.
[Ca2
]c measurements
Cells from the same EM26 and EM56 HeLa cell clones
were plated on coverslips and loaded with Fura-2 by
incubation for 1 h with 4 M Fura-2AM. Measurements
were performed in cell monolayers using a Cairn spectrophotometer equipped with a six-filter rotating wheel
as described previously [21]. [Ca2
] values were calculated from the ratio between the fluorescence obtained at
340 and 380 nm excitation wavelength. Cells used for
[Ca2
]c measurements were always depleted of Ca2
in
the same way as those used for [Ca2
]ER measurements,
in order to allow comparison of both types of data in the
same conditions. Ca2
entry was studied by following the
rate of Fura-2 quenching induced by Mn2
entry, used as
a Ca2
surrogate. Cells were loaded with Fura-2 as above
and the fluorescence excited at 360 nm (insensitive to
Ca2
) was monitored. Data were then normalized as
%F360. Coelenterazine n, Fura2-AM and BAPTA-AM were
obtained from Molecular Probes. Other reagents were
from Sigma, Madrid or Merck, Darmstadt.
2001 Harcourt Publishers Ltd

CECA-70.QXD

8/3/01 4:16 PM

Page 183

Stimulation by thimerosal of Ca2
release 183

RESULTS
Reconstitution of ER-targeted aequorin with the semisynthetic prosthetic group, coelenterazine n, requires previous depletion of Ca2
of the ER to prevent aequorin
consumption during reconstitution [9]. Once aequorin
has been reconstituted, the ER is refilled again with Ca2

by perfusing the cells with extracellular medium containing 1 mM Ca2
. This leads to an increase in [Ca2
]ER, that
reaches a steady-state of around 600 M within 35 min
in HeLa cells [9,10]. Figure 1 shows a comparison of the
effects of histamine on both [Ca2
]ER (upper panels) and
[Ca2
]c (lower panels), either in the presence or in the
absence of thimerosal or dithiotreitol. Comparing panels
a and b, we can see that the rate of refilling of the ER was
not significantly modified by thimerosal. In fact, Figure 1
shows that even the refilling of the ER after histamine
addition, i.e. after 1520 min in the presence of thimerosal
100 M, proceeded at a similar rate. In 15 similar experiments, the rate of refilling of the ER in cells treated for at
least 7 min with thimerosal 100 M was 1.80.2 M/s
(meanSD). This value is identical to that obtained under
control conditions (1.80.5 M/s, see [10]). The rate in
each experiment was obtained by averaging the rate of
refilling during 100 s within the period of maximum rate
of ER refilling. Another parameter that should be affected
if the ER-Ca2
-pump was inhibited by thimerosal is the

steady-state [Ca2
]ER. In a series of experiments performed in parallel in control cells or in cells treated for at
least 5 min with 100 M thimerosal, the steady-state
[Ca2
]ER was 66080 (meanSD, n30) in control cells
and 71090 (meanSD, n36) in cells treated with
thimerosal. These data suggest that thimerosal does not
modify significantly the activity of the ER-Ca2
-ATPase
(SERCA, from sarcoplasmic and endoplasmic reticulum
Ca2
-ATPase) in intact HeLa cells.
A very distinct effect of thimerosal was observed
instead on histamine-induced [Ca2
]c peaks and [Ca2
]ER
decreases. Panels a and d of Figure 1 show the effect of
histamine on both [Ca2
]ER and [Ca2
]c in control cells.
As we have reported previously [10], histamine produces
a small and fast [Ca2
]ER decrease (panel a), that translates in the cytosol into the [Ca2
]c peak shown in panel
d. If the cells were pretreated with thimerosal for 57 min,
the effects of histamine on both [Ca2
]ER and [Ca2
]c
were much stronger ( panels b & e) and had a very peculiar kinetics. Histamine induced a fast drop in [Ca2
]ER
coincident with the [Ca2
]c peak. This was followed by a
short period of rapid refilling and then by a second and
much slower phase of [Ca2
]ER decrease. This second
phase was coincident with a persistent increase in
[Ca2
]c. A magnification of these experiments, shown in
the inset, reveals interesting kinetic details. The initial

Fig. 1 Effects of thimerosal and dithiotreitol on histamine-induced [Ca2
]ER decrease and [Ca2
]c increase in HeLa cells. Cells were
reconstituted with coelenterazine n (panels ac) or loaded with Fura-2 (panels df). Then, medium containing 1 mM Ca2
(Ca2
), 100 M
histamine (His), 100 M thimerosal (Thim) or 1 mM dithiotreitol (DTT) was perfused as indicated. The inset shows a magnified superimposition
of the histamine addition in panels b and e. These experiments are representative of from five to 13 similar ones of each kind.

2001 Harcourt Publishers Ltd

Cell Calcium (2001) 30(3), 181190

CECA-70.QXD

8/3/01 4:16 PM

Page 184

184 M Montero, MJ Barrero, F Torrecilla, CD Lobatn, A Moreno, J Alvarez

fast [Ca2
]ER decrease coincides with the [Ca2
]c peak.

Then, Ca2
release stops suddenly and [Ca2
]ER increases
for a few seconds while [Ca2
]c decreases down to near
background levels. These phenomena suggest that inhibition of Ca2
release by local [Ca2
]c microdomains
leads to rapid back-pumping of Ca2
into the ER. Finally,
[Ca2
]ER turns to decrease again slowly while [Ca2
]c
increases and stabilizes around 400 nM. Once histamine
was washed away, [Ca2
]c suddenly dropped to resting
levels and the ER started to refill again with Ca2
normally. A subsequent addition of histamine still in the
presence of thimerosal produced similar effects. We
should remark that thimerosal induces a slow but persistent increase in the resting [Ca2
]c before histamine addition. However, once histamine was washed, [Ca2
]c
rapidly dropped to resting levels for a couple of minutes
before starting to increase again. That period closely
matches the time required to refill the ER with Ca2
(compare panels b and e), suggesting that SERCAs are able to
reduce [Ca2
]c to resting levels while they are refilling the
ER. Once the ER is full of Ca2
, the pump-leak equilibrium
at the plasma membrane appears unable to keep low the
resting [Ca2
]c levels in the presence of thimerosal. The
effects of thimerosal on resting [Ca2
]c and histamineinduced [Ca2
]c peaks were not reversible by washing
(not shown). However, they could be reversed by the
reducing agent dithiotreitol (Fig. 1, panels c & f). The first
histamine was added after 7 min preincubation with
thimerosal, and produced the same effects described
before. Then, refilling of the ER took place in the presence
of dithiotreitol. After that, the second addition of histamine produced a much smaller effect. Actually, the height
of the histamine-induced [Ca2
]c peak was smaller in the
presence of dithiotreitol (386% of the control peak,
meanSEM, n6, compare Figs. 1d & f ). Consistently, the
step of decrease in [Ca2
]ER induced by histamine in the
presence of dithiotreitol was only of 386 M
(meanSEM, n5) compared to 1026 M (meanSEM,
n10) in the controls (compare Figs. 1a & c).
The effects of thimerosal were dependent on the concentration. Thimerosal was almost inactive a 1 M, and
10 M thimerosal produced a clear activation of histamine-induced Ca2
-release, though smaller than with
100 M thimerosal (data not shown). Regarding the activation by thimerosal of histamine-induced Ca2
-release,
it has been reported that thimerosal does not increase
InsP3 production [13], but increases the sensitivity of the
InsP3R to InsP3 (see Introduction). We would, therefore,
expect that the effects of thimerosal should be made
more evident by using a submaximal concentration of
histamine. Figure 2 shows that this is the case. The left
panel shows that 2.5 M histamine produced no
detectable effect on [Ca2
]ER, and a small [Ca2
]c transient in control HeLa cells. In the presence of 100 M
Cell Calcium (2001) 30(3), 181190

thimerosal, the same concentration of histamine induced

a strong decrease in [Ca2
]ER with the typical biphasic
kinetics. The [Ca2
]c peak was also bigger and followed
by a persistent increase in [Ca2
]c.
The effect of histamine on [Ca2
]ER in HeLa cells is
strongly potenciated by loading the cells with a Ca2
chelator, BAPTA [9,10]. We have proposed that this effect
may be due to the inhibition of Ca2
release by local
[Ca2
]c microdomains in the absence of BAPTA. Figure 3
shows that this effect was additive to that of thimerosal.
The left panel shows the effect of histamine on [Ca2
]ER
both in control and in BAPTA-loaded cells. Histamine produced a very fast decrease of [Ca2
]ER (about 80%) in

Fig. 2 Effect of thimerosal on the [Ca2
]ER decrease and [Ca2
]c

increase induced by 2.5 M histamine. Cells were reconstituted with
coelenterazine n (lower panels) or loaded with Fura-2 (upper
panels). Then, medium containing 1 mM Ca2
(Ca2
), 2.5 M
histamine (His 2.5 M) or 100 M thimerosal (Thim) was perfused
as indicated. These experiments are representative of from five to
six similar ones of each kind.

Fig. 3 Effects of thimerosal and intracellular BAPTA on

histamine-induced [Ca2
]ER decrease. Cells were reconstituted
with coelenterazine n either in the presence (traces labeled
BAPTA) or in the absence (traces labeled Control) of 10 M
BAPTA-AM. Then, medium containing 1 mM Ca2
(Ca2
),
100 M histamine (His) or 100 M thimerosal (Thim) was
perfused as indicated. These experiments are representative
of from four to 13 similar ones of each kind.

2001 Harcourt Publishers Ltd

CECA-70.QXD

8/3/01 4:16 PM

Page 185

Stimulation by thimerosal of Ca2
release 185

BAPTA-loaded cells. This was followed by a partial refilling

still in the presence of histamine and then by a rapid and
complete refilling when the agonist was washed away.
This last refilling period usually led to an overshoot of
[Ca2
]ER, reaching levels 268% (meanSEM, n8)
higher than those prior to stimulation. In terms of
absolute [Ca2
]ER values, the pre-stimulation level was
57060 nM (meanSD, n8) and increased to 710
100 nM (meanSD, n8) after recovering of stimulation.
In the same series of experiments, mean steady-state
[Ca2
]ER values obtained in cells not loaded with BAPTA
was 66080 nM (meanSD, see above). Therefore, the
starting [Ca2
]ER value in BAPTA-loaded cells was somewhat lower than in control cells, probably because the
buffering capacity of BAPTA decreases the rate of refilling
of the ER [22]. Agonist-induced Ca2
release may then
stimulate Ca2
entry and increase slightly the steadystate [Ca2
]c [9], stimulating SERCA to accumulate Ca2

into the ER. The right panel shows the same experiments
but carried out in cells pretreated with thimerosal. In
BAPTA-loaded cells, histamine induced a complete and
persistent emptying of the stores, which was fully and
rapidly reversible after washing out the agonist.
It has been reported recently [23] that mitochondrial
depolarization may modify (first enhance, then inhibit)
InsP3-induced Ca2
release in HeLa cells. Thimerosal is
an oxydizing agent, and a series of oxidants have been
shown to promote opening of the mitochondrial permeability transition pore [24], a mechanism that would
rapidly lead to mitochondrial depolarization. It could be
argued, therefore, that the effects of thimerosal on Ca2

release from the ER could be indirect, mediated by its
effects on mitochondria. However, if that were the case,
mitochondrial Ca2
uptake induced by histamine addition should be strongly inhibited in the presence of
thimerosal. Figure 4 shows instead that treatment with
thimerosal enhanced the peak of [Ca2
]M induced by histamine. In several similar experiments, the [Ca2
]M peak
induced by 100 M histamine increased from 2.40.8 M
(meanSD, n5) to 4.70.6 (meanSD, n6) after 57
min of incubation with 100 M thimerosal. These data
are consistent with a larger Ca2
release from the ER in
the presence of thimerosal, even though the real [Ca2
]M
values in each case may be underestimated due to both
subcellular [Ca2
]M heterogeneity and preferential consumption of aequorin in high [Ca2
]M compartments
[25]. The control [Ca2
]M peak observed here corresponds to the nearly full consumption of aequorin in a
mitochondrial subcompartment containing about 5%
(5.31.3%, meanSEM, n5) of aequorin. In the presence of thimerosal, the size of this subcompartment
increased to 202% (meanSEM, n6), and this produced the increase in the calibrated signal. Regarding the
size of these compartments, we should note that our
2001 Harcourt Publishers Ltd

Fig. 4 Effect of thimerosal on the [Ca2
]M peak induced by

histamine. HeLa cells expressing mitochondrially targeted aequorin
were reconstituted with wild type colenterazine. Then 100 M
histamine was added either to control cells or to cells treated for
5 min with 100 mM thimerosal, as indicated. Thimerosal was also
present during and after histamine addition. These experiments
are representative of from five to six experiments of each kind.

experiments were performed at 22C to match the temperature used in the [Ca2
]ER experiments. At 37C, the
histamine-induced [Ca2
]M peak consumed 292%
(meanSEM, n8) of aequorin (see also [25]) and
thimerosal increased that percentage to 442%
(meanSEM, n7 ).
We have shown previously that Ca2
-release induced
by histamine in HeLa cells is quantal in the presence of
BAPTA, when the inhibition by [Ca2
]c has been
removed [9]. This means that submaximal concentrations
of histamine produce a rapid but incomplete Ca2

release, leading to a new steady-state at a level of [Ca2
]ER
that depends on the histamine concentration. The mechanism of this quantal effect is not clear and may be multiple. Inactivation of the InsP3-gated channels, regulation
of Ca2
release by [Ca2
]ER or inhibition of Ca2
release
by local microdomains of [Ca2
]c are among the proposed mechanisms [26]. In BAPTA-loaded cells, however,
this quantal effect cannot be attributed to local
microdomains of high [Ca2
]c. Regarding the inactivation
of InsP3-gated Ca2
channels, we have shown here that
thimerosal facilitates a full and persistent activation of
these channels in BAPTA-loaded cells. We have, therefore, investigated if Ca2
release induced by histamine
was still quantal under these conditions. Figure 5, upper
panel, shows that the sensitivity to histamine was dramatically increased by thimerosal. In BAPTA-loaded cells,
2.5 M histamine produced near half-maximal Ca2

release from the ER [10], and no significant effect was
obtained at concentrations at or below 1 M (Fig. 5, panel
a). In the presence of thimerosal, instead, 1 M histamine
produced a stronger and more persistent Ca2
release
than that induced by 100 M histamine in the absence of
Cell Calcium (2001) 30(3), 181190

CECA-70.QXD

8/3/01 4:16 PM

Page 186

186 M Montero, MJ Barrero, F Torrecilla, CD Lobatn, A Moreno, J Alvarez

Fig. 5 Effect of thimerosal on the sensitivity to histamine of Ca2

release. Cells were reconstituted with coelenterazine n in the
presence of 10 M BAPTA-AM. Then, medium containing 1 M
Ca2
(Ca2
), 100 M thimerosal (Thim), and either 10 nM, 1 M or
100 M histamine was perfused as indicated. These experiments
are representative of 1214 similar ones of each kind.

thimerosal. The lower panel shows that even histamine

concentrations as low as 10 nM were able to produce a
sizeable Ca2
release in the presence of thimerosal and
BAPTA. The kinetics of this response was still typically
quantal.
The possible role of GSSG as a physiological counterpart of thimerosal was studied by measuring the release
of Ca2
of the ER induced by InsP3 in permeabilized cells.
Figure 6a shows that InsP3 released Ca2
from the ER in
a quantal manner, and its potency was not modified by
previous incubation with GSSG. The percentage of Ca2

released by 100 nM InsP3 with respect to that released by
InsP3 2 M was 3217% (meanSD, n10) in control
conditions and 2914% (meanSD, n6) in the presence of 4 mM GSSG. In permeabilized cells, instead, 10 M
thimerosal strongly potentiated the effect of 100 nM InsP3
(Fig. 6b), but it also increased considerably the steady-state
level of [Ca2
]ER. In parallel experiments, the steady-state
[Ca2
]ER was 83060 M (meanSEM, n6) in the controls and 145070 M (meanSEM, n9) in the presence of 10 M thimerosal. Higher concentrations of
thimerosal (100 M) increased still further the refilling of
the ER, leading to a very fast consumption of aequorin,
that precluded studying the effect of InsP3 (not shown).
We have investigated also the effect of thimerosal on a
series of systems related to cell Ca2
homeostasis,
Cell Calcium (2001) 30(3), 181190

Fig. 6 Effect of oxidized glutathione and thimerosal on

InsP3-induced Ca2
release in permeabilized HeLa cells. Cells
reconstituted with coelenterazine n were permeabilized and then
perfused with an ATP-containing 100 nM [Ca2
] buffer to refill the
ER (see Experimental). Panel a: medium containing 100 nM or
2 M InsP3 (IP3) was perfused as indicated. In the right panel, 4 mM
oxidized glutathione (GSSG) was included in the perfusion medium
just after permeabilization (7 min before the first InsP3 addition).
These experiments are representative of 610 similar ones of each
kind. Panel b: cell permeabilization is started by perfusion of
100 M digitonin (dig) in medium containing 100 nM [Ca2
] buffer.
Then, InsP3 100nM or thimerosal 10 M were perfused as
indicated. These experiments are representative of from five to nine
similar ones of each kind carried out in parallel.

namely ER Ca2
leak, plasma membrane Ca2
pump and

Ca2
entry. The rate of Ca2
-leak from the ER can be easily estimated from the rate of [Ca2
]ER decrease after the
addition of a maximal concentration of a SERCA inhibitor
such as 2,5-di-tert-buthyl-benzohydroquinone (BHQ ).
Figure 7a shows that BHQ induces a rapid decrease of
[Ca2
]ER, leading to complete emptying of Ca2
of the ER
within 510 min (trace B). Thimerosal produced a small
but significant increase in the rate of Ca2
-leak from the
ER (trace C). In five similar experiments, the half time for
emptying of the ER was 1303 s in control conditions and
11113 s in cells treated with thimerosal (meansSD,
significantly different at P:0.05, independent t-test). The
trace labeled A corresponds to cells treated with thimerosal
but not with BHQ. Figure 7b shows the behaviour of
[Ca2
]c in a similar series of experiments. Trace A shows
that thimerosal alone induces a slow increase in [Ca2
]c
after a delay that was variable between 5 and 10 min.
Addition of BHQ induced a transient increase in [Ca2
]c
that returned to resting levels within 10 min in control
cells (trace B). Instead, in cells treated with thimerosal,
BHQ produced a persistent increase in [Ca2
]c (trace C).
The persistent elevation of [Ca2
]c observed in cells
treated with thimerosal (Figs 1 & 7) suggests that, in these
cells, the pump-leak equilibrium at the plasma membrane is unable to keep the resting [Ca2
]c levels. This
2001 Harcourt Publishers Ltd

CECA-70.QXD

8/3/01 4:16 PM

Page 187

Stimulation by thimerosal of Ca2
release 187

Fig. 7 Effects of thimerosal and BHQ on [Ca2
]ER and [Ca2
]c. Cells

were reconstituted with coelenterazine n (panel a) or loaded with
Fura-2 (panel b). In the curves labeled A and C, cells were treated
with 100 M thimerosal for 7 min before the arrow. The arrow indicates
the addition of 10 M BHQ to curves B and C. These experiments are
representative of four to five similar ones of each kind.

Fig. 8 Effects of thimerosal on the plasma membrane Ca2
pump

and on Ca2
(Mn2
) entry. Cells were loaded with Fura-2. Then,
medium containing 0.5 mM EGTA, 10 M BHQ or 1 mM Mn2
was
perfused as indicated. In the curves labeled Thimerosal or BHQ,
100 M thimerosal or 10 M BHQ were perfused for 7 min before
the additions. These experiments are representative of from three
to six similar ones of each kind.

means that thimerosal must produce either an increase of

the rate of Ca2
-entry through the plasma membrane or
an inhibition of the plasma membrane Ca2
-pump. To
investigate the effects of thimerosal on the plasma membrane Ca2
-pump, we treated the cells with BHQ in Ca2
free medium, as shown in Figure 8a. Under these
conditions, release of Ca2
from the ER increases [Ca2
]c
levels. Subsequent return to resting conditions depends
only of Ca2
-pumping mediated by the plasma membrane Ca2
-pump. Treatment with thimerosal modified
little the kinetics of return of [Ca2
]c to resting levels. The
rate of [Ca2
]c decrease, measured at a [Ca2
]c of
300350 nM, was 1.70.6 nM/s (meanSD, n5) in the
absence of thimerosal and 1.40.5 nM/s (meanSD,
n6) in the presence of thimerosal. The difference was
not significant (P 9 0.1, independent t-test), indicating
that the activity of the plasma membrane Ca2
-pump at
that [Ca2
] was scarcely affected by thimerosal. However,
the final steady-state [Ca2
]c reached was higher in the
presence of thimerosal (18941 nM vs 11124 nM in
control cells, meanSD, n6). The difference here was
highly significant (P : 0.005, independent t-test), suggesting that thimerosal may reduce the activity of the plasma
membrane Ca2
pump particularly at low [Ca2
]c.
To measure the effect of thimerosal on Ca2
-entry
through the plasma membrane, we have used Mn2
as a

Ca2
-surrogate. Figure 8b compares the rates of Mn2

entry obtained in control cells and in cells treated with
thimerosal, both in the presence and in the absence of
BHQ. Mn2
entry was faster in cells treated with BHQ
with respect to the controls (2.20.4 fold, meanSD,
n4) due to the activation of the store-operated
Ca2
-channels. In the presence of thimerosal, the rate of
Mn2
entry increased similarly both in controls and in
BHQ-treated cells (1.50.1 and 1.40.3 fold, respectively; meanSD, n3). This suggests that thimerosal
may activate Ca2
-entry through a pathway independent
of the store-operated channels.

2001 Harcourt Publishers Ltd

DISCUSSION
We have investigated the effects of thimerosal on Ca2

homeostasis in HeLa cells by monitoring [Ca2
]ER, [Ca2
]c
and [Ca2
]M. Our results show that thimerosal dramatically increases the sensitivity and the maximum rate of
InsP3-induced Ca2
-release. In addition, it produced also
smaller modifications in other parameters related to Ca2

homeostasis in intact cells. Thimerosal activated Ca2
entry through the plasma membrane and decreased the
activity of the plasma membrane Ca2
pump under resting [Ca2
]c conditions. Instead, the activity of the ERCa2
-pump in intact cells was not affected and the rate of
Cell Calcium (2001) 30(3), 181190

CECA-70.QXD

8/3/01 4:16 PM

Page 188

188 M Montero, MJ Barrero, F Torrecilla, CD Lobatn, A Moreno, J Alvarez

the ER-Ca2
-leak was only slightly increased. In permeabilized cells, however, the steady-state [Ca2
]ER level
was increased by thimerosal. The effects of thimerosal
were not mediated by mitochondrial depolarization,
because the ability of mitochondria to take up Ca2
was
fully preserved in the presence of thimerosal.
The effects of thimerosal increasing the sensitivity to
InsP3 of the InsP3R have been reported in many cell
types, including HeLa cells [13]. In these cells, concentrations of thimerosal as high as 100 M have been shown
to produce still activation of Ca2
-release, contrarily to
other cell types where this concentration produced a
strong inhibition [16,17,26,27]. By looking at [Ca2
]ER, we
show here that 100 M thimerosal strongly activates histamine-induced Ca2
-release. Activation involved several
mechanisms. On the first place, thimerosal increased by
about two orders of magnitude the sensitivity to histamine of Ca2
release (Fig. 5). This effect can only be interpreted as an increase in the sensitivity to InsP3 of the
InsP3R, because thimerosal does not increase InsP3 production in these cells [13]. On the second place,
thimerosal increased the magnitude of histamineinduced Ca2
release, even in the presence of maximum
concentrations of histamine. This is evident both in the
absence (Fig. 1) and in the presence (Fig. 5a) of BAPTA.
This effect is consistent with the reported ability of
thimerosal to increase the mean open times of the InsP3R
channels and to shift them to higher subconductance
states [15]. On the third place, thimerosal increased the
duration of histamine-induced Ca2
release, leading to
complete and persistent Ca2
-depletion of the ER in
BAPTA-loaded cells. In control cells, instead, depletion
was not complete and the ER started to refill with Ca2

still in the presence of histamine (Fig. 3). This suggests
that InsP3R channels undergo a slow developing inactivation even in BAPTA-loaded cells, when the generation of
local [Ca2
]c microdomains is prevented. This inactivation is abolished by thimerosal.
The interplay among the effects of thimerosal and the
inhibition of InsP3R by local [Ca2
]c microdomains
provided a very peculiar kinetics to histamine-induced
Ca2
-release, revealing interesting kinetic details about
that process of inhibition. The effect of histamine was
biphasic, composed of a fast initial release (:10 s) followed
by a short period (20 s) of refilling and then again by
release, but at a much slower rate. These [Ca2
]ER phenomena closely correlate with the biphasic [Ca2
]c dynamics (see Figs 1df and inset) induced by histamine,
originally described by Bootman et al. [28]. This pattern
of response can be easily explained by the inhibition of
InsP3R by microdomains of high [Ca2
]c. After the initial
Ca2
release, InsP3R are rapidly blocked by the accumulation of Ca2
in microdomains around the channels. This
stops Ca2
release and leads to the rapid dissipation of the
Cell Calcium (2001) 30(3), 181190

[Ca2
]c microdomains. However, the inhibition remains

longer that the microdomains, leading to rapid refilling of
the ER through the SERCAs, that are strongly stimulated
because [Ca2
]c is then at the peak. After about 20 s, the
InsP3R channels reactivate partially and we observe a second phase of slow release. This phase coincides with a
persistent increase of [Ca2
]c, which can be attributed in
part to Ca2
release but also to other effects of thimerosal
on Ca2
entry and extrusion at the plasma membrane
(see below). In the absence of thimerosal, the response to
histamine at the [Ca2
]c level follows a similar, although
attenuated, pattern (Fig. 1d). Instead, only a single step in
[Ca2
]ER is observed (Fig. 1a). This apparent discrepancy
may be explained by the generation, in the last sustained
phase of the [Ca2
]c transient, of a new pump-leak equilibrium in which increased Ca2
release is exactly balanced by stimulated Ca2
pumping due to the high
[Ca2
]c. Then, as soon as histamine is washed, Ca2

release stops and both [Ca2
]ER and [Ca2
]c return to resting values. This pump-leak equilibrium phase was abolished by dithiotreitol (Fig. 1f ), suggesting that this
compound may not only revert the stimulation by
thimerosal but also inhibit Ca2
release with respect to
the control condition. In fact, both the [Ca2
]c peak and
the step of Ca2
release from the ER were reduced in the
presence of dithiotreitol. This effect could be interpreted
by assuming the presence of a certain resting level of oxidized/activated InsP3R, which could be increased by
thimerosal or decreased by dithiotreitol. In cells loaded
with BAPTA, the [Ca2
]c microdomains are not generated
and only the initial rapid phase of Ca2
release was
observed. Then, a slow inactivation of the channels
develops, Ca2
release stops and the ER starts refilling
still in the presence of the agonist.
Our results suggest, therefore, that InsP3-induced
Ca2
release is limited by two types of inactivation of the
channels, one dependent of the inhibition by [Ca2
]c
microdomains and sensitive to BAPTA, and the other sensitive to thimerosal. In the presence of BAPTA, only the
thimerosal-sensitive mechanism is operative, leading to a
delayed slow inactivation. In the presence of thimerosal,
only the [Ca2
]c-dependent mechanism is operative, leading to the biphasic kinetics described above. In the presence of both BAPTA and thimerosal, the channels do not
inactivate and the ER remains empty while the agonist is
present. [Ca2
]c-dependent inactivation was long-lasting
and persisted for at least 20 s in the absence of [Ca2
]c
microdomains. This suggests that the mechanism of
this inhibition may include phenomena such as Ca2
dependent phosphorylation of the channels or interaction with Ca2
-sensitive proteins [29]. Ca2
release
induced by histamine remained quantal even in the presence of both BAPTA and thimerosal, indicating that the
quantal nature of Ca2
release under these conditions
2001 Harcourt Publishers Ltd

CECA-70.QXD

8/3/01 4:16 PM

Page 189

Stimulation by thimerosal of Ca2
release 189

cannot be attributed to inactivation of the channels. The

most probable alternative explanation appears to be the
dependence of the rate of Ca2
release on [Ca2
]ER.
The lack of effect of oxidized glutathione on InsP3induced Ca2
release in permeabilized HeLa cells was
surprising. This compound has been shown to activate
InsP3-induced Ca2
release in hepatocytes [11,12], and
depletion of intracellular reduced glutathione sensitized
HeLa cells to the effect of thimerosal [13]. We have no
explanation for the discrepancy, but it may be possible
that a cytosolic factor required for this effect is lost after
permeabilization of HeLa cells. In permeabilized cells,
thimerosal was still able to stimulate InsP3-induced Ca2

release from the ER. However, the effect of thimerosal in
permeabilized cells was more complex, as it strongly
increased also Ca2
uptake by the ER and the steadystate [Ca2
]ER levels. This made more difficult comparing
the effects of InsP3 in control and thimerosal-treated cells,
because differences in the [Ca2
]ER level may also modify
the response to InsP3. The effects of thimerosal on the
[Ca2
]ER level are discrepant with those obtained by
Bootman et al. [13], that showed a substantial reduction in
the Ca2
content of the ER in permeabilized cells treated
with 10 M thimerosal. The reasons for the discrepancy
are unclear, and may rely on the different experimental
conditions, in particular the permeabilization procedure.
While we measure the rate of ER refilling immediately
after permeabilization (on line 1-min permeabilization,
see Fig. 6b), the protocol of Bootman et al. was much
longer, including a 10-min permeabilization at 37C, centrifugation and resuspension, treatment with mitochondrial inhibitors and storage on ice for up to 2 h. The
mechanism of the increased [Ca2
]ER level induced by
thimerosal is unknown. As far as we know, redox modulation of SERCA has not been described, and this phenomenon deserves further study.
Regarding to other components of the Ca2
homeostatic
machinery, we show here that thimerosal activates Ca2

entry through a pathway independent and additive to that
induced by emptying of the intracellular Ca2
stores. In
addition, thimerosal reduced the activity of the plasma
membrane Ca2
pump at resting [Ca2
]c but not at higher
[Ca2
]c. Both effects contribute to the slow-developing
increase in the mean steady-state [Ca2
]c levels induced
by thimerosal, as reported previously [13]. Thimerosal had
little effects on the ER Ca2
pump in intact cells and produced only a small, although significant, increase in the
rate of ER Ca2
leak. This effect, however, was unable to
modify the steady-state [Ca2
]ER level. In permeabilized
cells, instead, thimerosal strongly increased Ca2
uptake
by the ER and the steady-state [Ca2
]ER level.
In conclusion, we show here a direct view of the effects
of thimerosal on InsP3-induced Ca2
release. Activation
of Ca2
-release by thimerosal revealed interesting kinetic
2001 Harcourt Publishers Ltd

details on the mechanisms of inactivation of the InsP3R

channels and on the mechanism of quantal Ca2
release.
In addition, the effects of thimerosal were reversible in
the presence of dithiotreitol, suggesting that they could
have relevance from a physiological point of view.
ACKNOWLEDGEMENTS
Financial support from Direccin General de Enseanza
Superior (PM 98-0142) and Junta de Castilla y Len ( VA
19/99) are gratefully acknowledged. We thank Jess
Fernndez for technical assistance.
REFERENCES
1. Berridge, MJ. Elementary and global aspects of calcium
signalling. J Physiol 1997; 499.2: 291306.
2. Ehrlich, BE. Functional properties of intracellular calcium-release
channels. Curr Op Neurobiol 1995; 5: 304309.
3. Mikoshiba K. The InsP3 receptor and intracellular Ca2

signaling. Curr Op Neurobiol 1997; 7: 339-345.
4. Taylor CW. Inositol trisphosphate receptors: Ca2
-modulated
intracellular Ca2
channels. Biochim Biophys Acta 1998; 1436:
1933.
5. Missiaen L, De Smedt H, Droogmans G, Casteels R. Ca2
release
induced by inositol 1,4,5-trisphosphate is a steady-state
phenomenon controlled by luminal Ca2
in permeabilized cells.
Nature 1992; 357: 599602.
6. Bezprozvanny I, Watras J, Ehrlich BE. Bell-shaped calciumresponse curves of Ins(1,4,5)P3- and calcium-gated channels
from endoplasmic reticulum of cerebellum. Nature 1991; 351:
751754.
7. Kaftan EJ, Ehrlich BE, Watras J. Inositol 1,4,5-trisphosphate
(InsP3) and calcium interact to increase the dynamic range of
InsP3 receptor-dependent calcium signalling. J Gen Physiol 1997;
110: 529538.
8. Hagar, RE, Burgstahler AD, Nathanson MH, Ehrlich BE. Type III
InsP3 receptor channel stays open in the presence of increased
calcium. Nature 1998; 396: 8184.
9. Montero M, Barrero MJ, Alvarez J. [Ca2
] microdomains control
agonist-induced Ca2
release in intact HeLa cells. FASEB J 1997;
11: 881885.
10. Barrero MJ, Montero M, Alvarez J. Dynamics of [Ca2
] in the
endoplasmic reticulum and cytoplasm of intact HeLa cells.
J Biol Chem 1997; 272: 2769427699.
11. Missiaen L, Taylor CW, Berridge MJ. Spontaneous calcium
release from inositol trisphosphate-sensitive calcium stores.
Nature 1991; 352: 241244.
12. Missiaen L, Taylor CW, Berridge MJ. Luminal Ca2
promoting
spontaneous Ca2
release from inositol trisphosphate-sensitive
stores in rat hepatocytes. J Physiol 1992; 455: 623640.
13. Bootman MD, Taylor CW, Berridge MJ. The thiol reagent,
thimerosal, evokes Ca2
spikes in HeLa cells by sensitizing the
inositol 1,4,5-trisphosphate receptor. J Biol Chem 1992; 267:
2511325119.
14. Henschke PN, Elliott SJ. Oxidized glutathione decreases luminal
Ca2
content of the endothelial cell Ins(1,4,5)P3-sensitive Ca2

store. 1995. Biochem J 1995; 312: 485489.
15. Thrower EC, Duclohier H, Lea EJ, Molle G, Dawson AP. The
inositol 1,4,5,-trisphosphate-gated Ca2
channel: effect of the
protein thiol reagent thimerosal on channel activity. Biochem J
1996; 318: 6166.

Cell Calcium (2001) 30(3), 181190

CECA-70.QXD

8/3/01 4:16 PM

Page 190

190 M Montero, MJ Barrero, F Torrecilla, CD Lobatn, A Moreno, J Alvarez

16. Karhapaa L, Titievsky A, Kaila K, Tornquist K. Redox modulation

of calcium entry and release of intracellular calcium by
thimerosal in GH4C1 pituitary cells. Cell Calcium 1996; 20:
447457.
17. Missiaen L, De Smedt H, Parys JB, Sienaert I, Valingen S, Casteels
R. Threshold for inositol 1,4,5-trisphosphate action.
J Biol Chem 1996; 271: 1228712293.
18. Parys JB, Missiaen L, De Smedt H, Droogmans G, Casteels R. Bellshaped activation of inositol-1,4,5-trisphosphate-induced Ca2

release by thimerosal in permeabilized A7r5 smooth-muscle
cells. Pflugers Arch 1993; 424: 516522.
19. Montero M, Brini M, Marsault R, Alvarez J, Sitia R, Pozzan T,
Rizzuto R. Monitoring dynamic changes in free Ca2

concentration in the endoplasmic reticulum of intact cells.
EMBO J 1995; 14: 54675475.
20. Brini M, Marsault R, Bastianutto C, Alvarez J, Pozzan T, Rizzuto R.
Transfected aequorin in the measurement of cytosolic Ca2

concentration ([Ca2
]c). A critical evaluation. J Biol Chem 1995;
270: 98969903.
21. Villalobos C, Garca-Sancho J. Glutamate increases cytosolic
calcium in GH3 pituitary cells acting via a high affinity
glutamate transporter. FASEB J. 1995; 9: 815819.
22. Alonso MT, Barrero MJ, Michelena P, Carnicero E, Cuchillo I,
Garca AG, Garca-Sancho J, Alvarez J. Ca2
-induced Ca2

Cell Calcium (2001) 30(3), 181190

23.

24.
25.

26.

27.

28.

29.

release in chromaffin cells seen from inside the ER with targeted

aequorin. J. Cell Biol. 1999; 144: 241254.
Collins TJ, Lipp P, Berridge MJ, Li W, Bootman MD. Inositol 1,4,5trisphosphate-induced Ca2
release is inhibited by
mitochondrial depolarization. Biochem J 2000; 347: 593600.
Crompton M. The mitochondrial permeability transition pore
and its role in cell death. Biochem J 1999; 341: 233249.
Rizzuto R, Bastianutto C, Brini M, Murgia M, Pozzan T.
Mitochondrial Ca2
homeostasis in intact cells. J Cell Biol
1994; 126: 11831194.
Parys JB, Missiaen L, De Smedt H, Sienaert I, Casteels R.
Mechanisms responsible for quantal Ca2
release from inositol
trisphosphate-sensitive calcium stores. Pflugers Arch 1996;
432: 359367.
Mezna M, Michelangeli F. Effects of thimerosal on the transient
kinetics of inositol 1,4,5-trisphosphate-induced Ca2
release
from cerebellar microsomes. Biochem J 1997; 325: 177182.
Bootman MD. Berridge MJ, Taylor CW. All-or-nothing Ca2

mobilization from the intracellular stores of single histaminestimulated HeLa cells. J Physiol 1992; 450: 163178.
Mackrill JJ. Protein-protein interactions in intracellular Ca2

release channel function. Biochem J 1999; 337: 345361.

2001 Harcourt Publishers Ltd