(Received for publication, October 22, 1996)
From the Laboratoire de Physiologie Cellulaire et
Pharmacologie Moléculaire, CNRS ESA 5017, Université de
Bordeaux II, 146 rue Léo Saignat, 33076 Bordeaux Cedex, France
and ¶ Institut für Pharmakologie, Freie Universität
Berlin,
Thielallee 69/71, D-14195 Berlin, Federal Republic of Germany
In this study, we identified the subunit
composition of Gq and G11 proteins coupling
1-adrenoreceptors to increase in cytoplasmic Ca2+ concentration ([Ca2+]i) in rat
portal vein myocytes maintained in short-term primary culture. We used
intranuclear antisense oligonucleotide injection to inhibit selectively
the expression of subunits of G protein. Increases in
[Ca2+]i were measured in response to activation
of
1-adrenoreceptors, angiotensin AT1
receptors, and caffeine. Antisense oligonucleotides directed against
the mRNAs coding for
q,
11,
1,
3,
2, and
3 subunits selectively inhibited the increase in
[Ca2+]i activated by
1-adrenoreceptors. A corresponding reduction of the
expression of these G protein subunits was immunochemically confirmed.
In experiments performed in Ca2+-free solution only cells
injected with anti-
q antisense oligonucleotides displayed a reduction of the
1-adrenoreceptor-induced
Ca2+ release. In contrast, in Ca2+-containing
solution, injection of anti-
11 antisense
oligonucleotides suppressed the
1-adrenoreceptor-induced
stimulation of the store-operated Ca2+ influx. Agents that
specifically bound G
subunits (anti-
com antibody
and overexpression of a
-adrenergic receptor kinase carboxyl-terminal fragment) had no effect on the
1-adrenoreceptor-induced signal transduction. Taken
together, these results suggest that
1-adrenoreceptors
utilize two different G
subunits to increase [Ca2+]i. G
q may activate
phosphatidylinositol 4,5-bisphosphate hydrolysis and induce release of
Ca2+ from intracellular stores. G
11 may
enhance the Ca2+-activated Ca2+ influx that
replenishes intracellular Ca2+ stores.
In vascular smooth muscle, activation of
1-adrenoreceptors stimulates phospholipase C-
which
hydrolyzes phosphatidylinositol-4,5-bisphosphate to yield
diacylglycerol and inositol 1,4,5-trisphosphate. In portal vein
myocytes, the
1A-adrenoreceptors are coupled to
phospholipase C-
through G proteins which have been identified to be
Gq and/or G11, on the basis of intracellular
applications of an anti-G
q/
11 antibody.
Inositol 1,4,5-trisphosphate subsequently releases Ca2+
from the intracellular store. Diacylglycerol in concert with cellular
Ca2+ activates protein kinase C which, in turn, stimulates
Ca2+ influx through voltage-dependent
Ca2+ channels (1-2). In addition, depletion of the
intracellular store by norepinephrine promotes a sustained
Ca2+ entry through dihydropyridine-resistant
Ca2+ channels by an unknown mechanism (3). Both
norepinephrine-induced Ca2+ release and Ca2+
entry lead to a biphasic rise of the cytoplasmic Ca2+
concentration
([Ca2+]i).1 Although
the G protein subtypes are currently defined by their
subunits, of
which 23 (including splice variants) are known, a functionally active
heterotrimeric G protein includes an
,
, and
subunit. Up to
now, 5 different
and 11 different
subunits have been identified
(4). Thus, a great number of heterotrimers composed of specific
,
, and
subunits may exist and be involved in signal transduction
pathways. In many cases, the coupling between receptor and G protein
may appear unselective since one receptor may activate more than one G
protein and thus initiate more than one signal-transduction pathway.
However, there are many examples showing that different receptors
activate the same heterotrimeric G protein to regulate the same
effector system (5-6). The question remains whether in portal vein
myocytes
1-adrenoreceptors recognize a single
heterotrimeric G protein (Gq or G11) to induce a rise of [Ca2+]i or whether different
heterotrimers varying in the composition of
,
, and
subunits
are required for this coupling.
Antisense oligonucleotides can be used for selective and transient
knockout of cellular proteins (7). So far, microinjection is the only
method available that allows for controlled intranuclear application of
antisense oligonucleotides. Studies with this method in GH3
cells have revealed that the M4 muscarinic receptor in GH3 cells couples to the G protein trimer consisting of
o1
3
4, the somatostatin
receptor to the trimer
o2
1
3, and the galanin receptor to the trimers
o1
2
2 and
o1
3
4 to inhibit
voltage-dependent Ca2+ channels (8-11). In
RBL-2H3-hm1 cells, G proteins composed of G
q/
11·
1/
4·
4
are required for effective coupling between the stably expressed human
muscarinic m1 receptor and cellular increase in
[Ca2+]i (12).
In the present study, we used the method of intranuclear microinjection
of antisense oligonucleotides directed against individual G protein
subunits and determined the composition of Gq and
G11 proteins mediating the
1-adrenoreceptor-induced increase in
[Ca2+]i in short-term primary cultured rat portal
vein myocytes. We show that
1-adrenoreceptors utilize G
proteins composed of
q,
11,
1,
3,
2, and
3 subunits to increase [Ca2+]i and
that the effector coupling is mediated by the
subunits.
G
q subunit may activate release of Ca2+ from
intracellular stores and G
11 subunit may modulate
intracellular store-dependent Ca2+ entry.
Isolated myocytes from
rat portal vein were obtained by enzymatic dispersion, as described
previously (1). Cells were seeded at a density of about 103
cells per mm2 on glass slides imprinted with squares for
localization of injected cells and maintained in short-term primary
culture in medium M199 containing 2% fetal calf serum, 2 mM glutamine, 1 mM pyruvate, 20 units/ml
penicillin, and 20 µg/ml streptomycin; they were kept in an incubator
gassed with 95% air, 5% CO2 at 37 °C. The sequences of
the oligonucleotides used in this study were determined by sequence
comparison and multiple alignment using Mac Molly Tetra software (Soft
Gene, Berlin, Germany). Oligonucleotides were from MWG-Biotech
(Ebersberg, Germany) or synthesized in a DNA synthesizer (Milligen,
model 8600); for synthesis of phosphorothioate oligonucleotides, the
method described by Iyer et al. (13) was used. Injection of
oligonucleotides was performed into the nucleus of myocytes by a manual
injection system (Eppendorf, Hamburg, Germany). The injection solution
contained 10 µM oligonucleotides in water; approximately
10 fl were injected with commercially available microcapillaries
(Femtotips, Eppendorf) with an outlet diameter of 0.5 µm. In some
control experiments, myocytes were injected only with water and tested
in comparison with non-injected cells and cells injected with sense,
scrambled, and antisense oligonucleotides. The myocytes were cultured
for 3-4 days in culture medium, and the glass slides were transferred
into a perfusion chamber for intracellular Ca2+
measurements. The sequences of anti-ocom,
anti-
1, anti-
2, anti-
3,
anti-
4, anti-
1, anti-
4,
and anti-
5 antisense oligonucleotides have been
previously published (11). The sequences of the anti-
q, anti-
11, and anti-
14 have been published
(12). The sequence of anti-
q/11com is ATGGACTCCAGAGT and
that of sense
q/11com is ACTCTGGAGTCCAT corresponding to
nt 4-17 of
q cDNA (14), of scrambled
anti-
q/11com is TACGGTCCAGAGTA corresponding to a
scrambled sequence of nt 4-17 of
q cDNA, of
anti-
12 is CTCCGGCCTCGGCCGGCAGCAAGC corresponding to nt
32-55 of
12 cDNA (15), of anti-
5 is
TGCCATCTTCGTCCGGATGCAGCC corresponding to nt (
18)-(+6) of
5 cDNA (16), of anti-
2 is TTCCTTGGCATGCGCTTCAC corresponding to nt 122-141 of
2
cDNA (17), of anti-
3 is GTTCTCCGAAGTGGGCACAGGGGT
corresponding to nt 165-188 of
3 cDNA (18), of
anti-
7 is CTGGGCGACGTTGTTAGTACCTGA corresponding to nt
7-30 of rat
7 cDNA (19), of anti-
8
is GCGGGCCTCAGCGAT CTTGGCCAT corresponding to nt 13-36 of
8 cDNA (20).
cDNAs encoding -adrenergic receptor
kinase carboxyl-terminal fragment and the S65T green fluorescent
protein were cloned into cytomegalovirus expression plasmids
pRK5 and pcDNA3, respectively (Clontech,
Palo Alto, CA). Plasmids were injected directly into the nucleus of
vascular myocytes, as described for oligonucleotides. Briefly,
cDNAs were diluted with water from stock solutions (0.5 µg/µl)
to final concentrations of 0.1 µg/µl. The S65T green fluorescent protein was included to facilitate later identification of myocytes receiving a successful nuclear injection. Fluorescence produced by the
S65T green fluorescent protein was observed 3 days after injection with
a confocal microscope (Bio-Rad MRC 1000, Paris, France). The percentage
of successful nuclear injection was estimated to be 20%
(n = 185).
Cells were loaded by incubation in physiological solution containing 1 µM fura-2-acetoxymethyl ester for 30 min at room temperature. These cells were washed and allowed to cleave the dye to the active fura-2 compound for at least 1 h. Fura-2 loading was usually uniform over the cytoplasm, and compartmentalization of the dye was never observed. Measurement of cytosolic Ca2+ concentration was carried out by the dual-wavelength fluorescence method, as described previously (1). Briefly, fura-2-loaded cells were mounted in a perfusion chamber and placed on the stage of an inverted microscope (Nikon Diaphot, Tokyo, Japan). Single cells were alternately excited with UV light at 340 and 380 nm through a 10 × oil immersion objective, and emitted fluorescent light from the Ca2+-sensitive dye was collected through a 510-nm-long pass filter with a charge-coupled device camera (Hamamatsu Photonics, Hamamatsu City, Japan). The signal was processed (Hamamatsu DVS 3000) by correcting each fluorescence image for background fluorescence and calculating 340/380 nm fluorescence ratios on a pixel-to-pixel basis. Averaged frames were usually collected at each wavelength every 0.5 s. In some experiments, cells were loaded through a patch-clamp pipette filled with a solution containing (in mM): 140 CsCl, 10 HEPES, 0.06 Fura-2, pH 7.3, as described previously (1). [Ca2+]i was calculated from mean ratios using a calibration for fura-2 determined in loaded cells. All measurements were made at 25 ± 1 °C.
The normal physiological solution contained (in mM): 130 NaCl, 5.6 KCl, 1 MgCl2, 2 CaCl2, 11 glucose, 10 HEPES, pH 7.4, with NaOH. Substances were applied to the cells by pressure ejection from a glass pipette for the period indicated on the records. Before each experiment, a fast application of physiological solution was tested, and cells with movement artifacts were excluded.
Results are expressed as means ± S.E. Significance was tested by means of Student's t test. p values of < 0.05 were considered as significant.
ImmunocytochemistryThree days after injection, venous myocytes were washed with phosphate-buffered saline solution (PBS), fixed with 3% formaldehyde (v/v) for 30 min at room temperature, and permeabilized in PBS containing 3% fetal calf serum and 0.01% (w/v) saponin for 30 min. Cells were incubated with the same buffer containing 5% fetal calf serum, 0.01% (w/v) saponin, and the anti-G protein antibody at 1:100 or 1:1000 dilution overnight at 4 °C. Then, cells were washed in PBS containing 3% fetal calf serum and 0.01 (w/v) saponin (4 × 10 min) and incubated with goat anti-rabbit IgG conjugated to fluorescein isothiocyanate (diluted 1:200) in the same solution for 8 h at 4 °C. Thereafter, cells were washed (4 × 10 min) in PBS and mounted in Moviol (Hoechst, Frankfurt, Germany). Images of the stained cells were obtained with a confocal microscope (Bio-Rad MRC 1000). Only cells on the same glass slide were compared with each other by keeping acquisition parameters (gray values, exposure time, aperture, etc.) constant. Immunostaining fluorescence was estimated by gray level analysis using the MPL software (Bio-Rad).
Chemicals and DrugsM199 medium was from Flow Laboratories
(Puteaux, France). Fetal calf serum was from Flobio (Courbevoie,
France). Streptomycin, penicillin, glutamate, and pyruvate were from
Life Technologies, Inc. (Paisley, UK). Fura-2, Fura-2/AM, and
anti-q (CN 371752) antibody were from Calbiochem
(Meudon, France). Norepinephrine, rauwolscine, and propranolol were
from Sigma (St. Quentin Fallavier, France).
Angiotensin II, CGP42112A
(N-
-nicotinoyl-Tyr-Lys[N-
-CBZ-Arg]-His-Pro-Ile-OH) was from Neosystem Laboratories (Strasbourg, France). Caffeine was from
Merck (Nogent sur Marne, France). Anti-
11 (SC 394), anti-
com (SC 378), anti-
1 (SC 379), and
anti-
3 (SC 375) were from Santa Cruz Biotechnology
(Santa Cruz, CA). Fluorescein isothiocyanate-conjugated goat
anti-rabbit IgG was from Immunotech (Marseille, France). Green
fluorescent protein S65T expression construct was from Clontech (Palo
Alto, CA). Oxodipine was a gift from Dr. Galiano (IQB, Madrid, Spain).
We previously showed that in portal vein myocytes,
activation of 1A-adrenoreceptors mediates both release
of Ca2+ from intracellular stores and stimulation of
voltage-dependent Ca2+ channels through a
Gq/G11 protein that activates phospholipase C-
(1). In order to identify the heterotrimeric G proteins involved
in the
1-adrenoreceptor-induced increase in
[Ca2+]i, we injected phosphorothioate-modified
antisense oligonucleotides directed against
,
, and
subunits
into the nucleus of vascular myocytes. By measuring the
norepinephrine-induced increases in [Ca2+]i after
injection of an antisense oligonucleotide directed against both
q and
11 subunits
(anti-
q/11com), the highest inhibition (76 ± 12%,
n = 7) was obtained 3 days after injection (data not
shown). Therefore, all further measurements were performed 3 days after
injection. We measured the increase in [Ca2+]i
induced by successive applications of 10 µM
norepinephrine (in the presence of 10 nM rauwolscine and 1 µM propranolol to inhibit both
2- and
-adrenoreceptors (2)), 10 mM caffeine and 10 nM angiotensin II (in the presence of 1 µM
CGP42112A to inhibit angiotensin AT2 receptors (21)) on the
same cells (Fig. 1). For each experiment, we compared
the Ca2+ responses of antisense oligonucleotide-injected
cells located within a marked area of the glass slide to sense or
scrambled oligonucleotide-injected cells or non-injected cells outside
this marked area. This procedure guaranteed that antisense
oligonucleotide-injected cells were always compared with control cells
that were otherwise grown, treated, and analyzed under identical
conditions, i.e. culture, incubation, microinjection, and
loading with fura-2AM. The increase in [Ca2+]i
was measured for each cell, and mean values were calculated from all
cells of each experiment. Myocytes injected with 10 µM antisense oligonucleotides directed against the mRNAs encoding for
q subunit (anti-
q) showed strongly
reduced (75%)
1-adrenoreceptor-induced Ca2+
responses, as compared with non-injected cells (Figs. 1B and 2A). Myocytes injected with 10 µM antisense
oligonucleotides directed against the
11
(anti-
11) subunit showed reduced (40%)
1-adrenoreceptor-induced Ca2+ responses as
well (Figs. 1C and 2A). Interestingly, injection of both anti-
q and anti-
11
oligonucleotides (anti-
q+11) did not induce a larger
decrease of the
1-adrenoreceptor-induced Ca2+ response than that evoked by anti-
q
oligonucleotides alone (Fig. 2A). Myocytes
injected with 10 µM antisense oligonucleotides directed against
o1 and
o2
(anti-
ocom),
12 (anti-
12),
and
14 (anti-
14) subunits were comparable
with non-injected cells. None of them showed a significant reduction in
[Ca2+]i responses evoked by activation of
1-adrenoreceptors (Figs. 1D and
2A). Furthermore, we used sense
q/11com and
scrambled anti-
q/11com oligonucleotides which do not
efficiently anneal to the target sequence of G
q/11
subunits. Ca2+ responses evoked by activation of
1-adrenoreceptors were not significantly affected by
injection of these oligonucleotides (non-injected cells = 418 ± 53 nM, n = 12; sense
q/11-injected cells = 379 ± 42 nM, n = 9; and scrambled
anti-
q/11com-injected cells = 390 ± 32 nM, n = 13).
In order to identify the subunits involved in the
1-adrenoreceptor-induced Ca2+ response, we
used antisense oligonucleotides directed against the mRNAs coding
for
1,
2,
3,
4, and
5 subunits (Fig.
3A). Injection of 10 µM
oligonucleotides directed against the
1
(anti-
1) and
3 (anti-
3)
subunits significantly reduced the
1-adrenoreceptor-induced Ca2+ responses.
Inhibition of the
1-adrenoreceptor-induced
Ca2+ responses evoked by the oligonucleotides directed
against the
1 and
3 subunits (80 and
40%, respectively) was quantitatively similar to those induced by the
oligonucleotides directed against the
q and
11 subunits. No significant reduction of the
1-adrenoreceptor-induced Ca2+ response was
seen in myocytes injected with oligonucleotides directed against
2 (anti-
2),
4
(anti-
4), and
5 (anti-
5) subunits. Injection of 10 µM antisense oligonucleotides
against different
subunits showed that anti-
1,
-
4, -
5, -
7, and
-
8 oligonucleotides had no significant effect on the
1-adrenoreceptor-induced Ca2+ responses
(Fig. 3B). In contrast, injection of anti-
3
and -
2 oligonucleotides resulted in significant
reduction of the
1-adrenoreceptor-induced Ca2+ responses (73 and 40%, respectively). Amplification
of cDNA fragments revealed that in portal vein smooth muscle five
(
1-
5) and six
(
2-
8) subunits were expressed (data not
shown). These results indicate that
1-adrenoreceptors
utilize G proteins composed of
q,
11,
1,
3,
2, and
3 subunits to increase
[Ca2+]i.
Specificity of the Antisense Oligonucleotides
In order to
verify that injection of antisense oligonucleotides directed against
specific G protein subunits suppressed involvement of these subunits in
the 1-adrenoreceptor-activated transduction couplings,
we performed two types of control experiments. First, we showed that
injection of a specific antisense oligonucleotide inhibited only the
immunofluorescence signal of the corresponding G protein subunit and
did not affect the expression of other subunits. Cells were stained
with either anti-
q or anti-
11 specific
antibodies, and the immunofluorescence was quantified by using the MPL
software of the confocal microscope (Fig.
4A). Cells injected with either of the two
different antisense oligonucleotides and non-injected cells located on
the same glass slide were compared with each other, so that the
staining procedure 3 days after injection of oligonucleotides was
identical for the different cells. In cells injected with
anti-
q oligonucleotides the immunofluorescence signal
for the G
q subunit was reduced by 76%
(n = 9), whereas that for the G
11
subunit was only slightly affected (10%, n = 8).
Similarly, in cells injected with anti-
11 antisense
oligonucleotides, the immunofluorescence signal for the
G
11 subunit was reduced by 70% (n = 7),
whereas that for the G
q subunit was only slightly affected (12%, n = 12). Then we tested the effects of
injection of anti-
1, anti-
3,
anti-
2, and anti-
3 antisense
oligonucleotides on the expression of
G
q/
11 subunits by staining with
anti-
q/
11 antibody (Fig. 4B).
Although the immunofluorescence signal appeared to be slightly reduced
in cells injected with
and
antisense oligonucleotides (between
15 and 20%, n = 21), only the cells injected with the
q/
11com antisense oligonucleotides showed a considerable inhibition of the immunofluorescence signal (85%, n = 7). Finally, we verified that in cells stained with
an anti-
1 antibody, the immunofluorescence signal was
inhibited in cells injected with anti-
1 antisense
oligonucleotides (77%, n = 13), whereas it was
slightly affected in cells injected with either anti-
3
and anti-
2 antisense oligonucleotides (15%,
n = 12). In cells stained with an anti-
3
antibody, the immunofluorescence signal was inhibited in cells injected
with anti-
3 antisense oligonucleotides (81%,
n = 12), whereas it was slightly affected in cells
injected with either anti-
1 or anti-
3
antisense oligonucleotides (22%, n = 12). In cells
stained with an anti-
3 antibody, the immunofluorescence
signal was inhibited in cells injected with anti-
3
antisense oligonucleotides (75%, n = 8), whereas it
was slightly affected in cells injected with either
anti-
3 or anti-
2 antisense
oligonucleotides (18%, n = 8). Taken together, these results indicate that each antisense oligonucleotide is selective for
inhibiting the expression of the corresponding G protein subunit.
Second, we compared the effects of norepinephrine to those of
angiotensin II (activating AT1 receptors, 22) and caffeine
(releasing Ca2+ from the intracellular stores) in each cell
studied (Fig. 1). We recently showed that activation of angiotensin
AT1 receptors releases intracellularly stored
Ca2+ without involving inositol 1,4,5-trisphosphate but
through a Ca2+ release mechanism activated by
Ca2+ influx through L-type Ca2+ channels
(22-23). In the same cells injected with anti-q/11com, -
q, and -
11 antisense oligonucleotides,
angiotensin II (in the presence of 1 µM CGP42112A) and
caffeine evoked large Ca2+ responses, whereas
1-adrenoreceptor-induced Ca2+ responses were
inhibited (Fig. 2, A-C). We noted unspecific effects of
phosphorothioate-modified antisense oligonucleotides only when oligonucleotides were injected at concentrations of 50 µM, i.e. 5 times higher than the concentration
used in these experiments (n = 15). Taken together,
these data indicate that suppression of
1-adrenoreceptor-activated effects by antisense
oligonucleotides does not interfere with other signaling pathways
(e.g. that of angiotensin II) and with the intracellular
Ca2+ stores of vascular myocytes.
We previously showed that norepinephrine
activates Ca2+ entry even if the intracellular
Ca2+ store is not completely emptied (3), possibly by
involving a mechanism independent of Ca2+ store depletion.
Therefore, experiments were performed in external Ca2+-free
solution (containing 0.5 mM EGTA) on myocytes injected with anti-q or anti-
11 antisense
oligonucleotides. As illustrated in Fig. 5, the
1-adrenoreceptor-induced Ca2+ release was
inhibited in cells injected with anti-
q oligonucleotides but was not affected in cells injected with anti-
11
oligonucleotides. These results suggest different tasks for
Gq and G11 proteins, i.e. induction
of Ca2+ release from intracellular stores and induction of
Ca2+ influx from extracellular medium, respectively.
The sarcoplasmic reticulum Ca2+-ATPase inhibitor
thapsigargin is commonly used to study the Ca2+ entry
pathway activated after Ca2+ store depletion (24).
Application of 1 µM thapsigargin depleted the
intracellular Ca2+ stores (evidenced by the lack of
caffeine-induced Ca2+ response) and increased the
[Ca2+]i level to 118 ± 17 nM
(n = 29; Fig. 6A). In the
continuous presence of 10 µM oxodipine (a light-resistant
dihydropyridine) to block voltage-dependent
Ca2+ channels, activation of
1-adrenoreceptors during the thapsigargin-induced Ca2+ plateau evoked a further rise in
[Ca2+]i reaching 161 ± 15 nM
(n = 5; Fig. 6A). This
1-adrenoreceptor-induced Ca2+ response was
never observed in Ca2+-free, 0.5 mM
EGTA-containing solution (n = 15). In cells injected with anti-
11 oligonucleotides, the
1-adrenoreceptor-induced rise in
[Ca2+]i was not observed (n = 7;
Fig. 6B), although thapsigargin produced a progressive
increase in [Ca2+]i reaching 123 ± 13 nM (n = 7). In contrast, in cells injected
with anti-
q oligonucleotides, the amplitude of the
1-adrenoreceptor-induced increase in
[Ca2+]i (52 ± 12 mM,
n = 6) was comparable with that obtained in
non-injected control cells (48 ± 14 nM,
n = 11; Fig. 6B).
Depletion of caffeine-sensitive intracellular Ca2+ stores
induced a Ca2+ response similar to that evoked by
activation of 1-adrenoreceptors. Fig. 7
displays representative traces of these experiments. In the continuous
presence of 10 µM oxodipine, application of 10 mM caffeine for 50 s in the external solution (Fig.
7Aa) produced a large transient increase in
[Ca2+]i (375 ± 20 nM,
n = 16) and a sustained plateau of 70 ± 9 nM (n = 16). The rapid initial increase in
[Ca2+]i was reduced in Ca2+-free
solution (298 ± 25 nM, n = 10), and
the subsequent sustained phase was absent (Fig. 7Ab). This
indicates that in venous myocytes caffeine is able to induce a
transient increase in [Ca2+]i due to
Ca2+ release and a sustained phase representing
Ca2+ entry into the cell from the extracellular space. As
illustrated in Fig. 2B, the caffeine-induced
Ca2+ responses were not affected by inhibition of the
expression of any G
subunits, including G
q and
G
11 subunits. Activation of
1-adrenoreceptors (in Ca2+-containing
solution) during the second sustained phase of the caffeine-evoked
Ca2+ response resulted in a 2-fold increase in
[Ca2+]i which reached 134 ± 16 nM (n = 23; Fig. 7Ba). The
1-adrenoreceptor-induced enhancement of
[Ca2+]i during the second phase of the
caffeine-induced Ca2+ response (64 ± 6 nM, n = 23) was not observed at all when
norepinephrine was applied without Ca2+ and in the presence
of 0.5 mM EGTA (n = 12; Fig.
7Bb), indicating that it corresponded to a Ca2+
entry from the extracellular medium. In myocytes injected with the
anti-
q antisense oligonucleotides, the
1-adrenoreceptor-induced Ca2+ entry in the
continuous presence of caffeine (55 ± 9 nM,
n = 6) was similar to that obtained in non-injected
cells (61 ± 8 nM, n = 6; Fig.
7Bc). In contrast, in cells injected with
anti-
11 antisense oligonucleotides, no
1-adrenoreceptor-induced Ca2+ entry in the
continuous presence of caffeine was observed (n = 8;
Fig. 7Bd). These results further support the idea that
G
11 subunit is involved in the modulation of
store-operated Ca2+ entry by
1-adrenoreceptors.
Effector Coupling Is Dependent on
The
anti-q/
11 antibody and antisense
oligonucleotide block of the
1-adrenoreceptor-induced
Ca2+ response cannot distinguish whether
or
subunits are transducing the signal that activate Ca2+
release from the intracellular stores or Ca2+ entry. To
determine which G protein subunits were involved in the
1-adrenoreceptor-mediated effects, an
anti-
com antibody was dialyzed into the cell by the
patch pipette for 3 min. Anti-
com antibody (10 µg/ml
in pipette solution) had no effect on the
1-adrenoreceptor-induced Ca2+ response (Fig.
8A) since neither the transient peak
(control = 305 ± 30 nM; in the presence of
anti-
com antibody = 295 ± 35 nM;
n = 10) nor the sustained plateau (control = 55 ± 4 nM; in the presence of anti-
com
antibody = 53 ± 6 nM, n = 10)
were significantly affected. In the same cells, the
anti-
com antibody inhibited the sustained angiotensin
II-induced Ca2+ response in a
concentration-dependent manner, with a maximal inhibition
obtained at an antibody concentration of 10 µg/ml (n = 10).2 In a second set of experiments, we
overexpressed a carboxyl-terminal fragment of
ARK1 by
intranuclear microinjection of expression plasmids containing cDNA
inserts coding for
ARK1.
ARK has been used to bind
subunits and block activation of effectors (25-26). Overexpression of
ARK1 had no effect on the
1-adrenoreceptor-induced Ca2+ response (Fig.
8B) since neither the transient peak (control = 290 ± 25 nM; in the presence of
ARK1 = 285 ± 20 nM; n = 12) nor the
sustained plateau (control = 65 ± 5 nM; in the
presence of
ARK1 = 60 ± 8 nM;
n = 12) were significantly affected. In contrast, the
angiotensin II-induced Ca2+ response was inhibited when
ARK1 was overexpressed in the same cells
(n = 12).2 Taken together, our results
indicate that application of anti-
com antibody and
ARK1, both able to bind free
subunits, had no effects on both Ca2+ release and Ca2+ entry
induced by activation of
1-adrenoreceptors.
Here we show that the
1-adrenoreceptor-induced increase in
[Ca2+]i in rat portal vein myocytes involves both
Gq and G11 proteins. Using nuclear injection of
antisense oligonucleotides corresponding to the mRNA sequences
coding for G protein
,
, and
subunits, we identified G
protein heterotrimers composed of
q,
11,
1,
3,
2, and
3 involved in the coupling of
1-adrenoreceptors to Ca2+ release and
intracellular store-dependent Ca2+ entry.
We proved the specificity of the injected oligonucleotides by studying
the increase in [Ca2+]i induced by application of
two hormonal stimuli (norepinephrine and angiotensin II) and caffeine.
All three substances release Ca2+ from the same
intracellular store via different pathways. Caffeine is known to
release Ca2+ by acting on the ryanodine-sensitive
Ca2+ channels of the sarcoplasmic reticulum. In portal vein
myocytes, we recently showed that activation of angiotensin
AT1 receptors evoked an increase in
[Ca2+]i which depended on both activation of
L-type Ca2+ channels and opening of ryanodine-sensitive
Ca2+ channels of the sarcoplasmic reticulum, without
involving inositol 1,4,5-trisphosphate generation (22); this effect is
mediated by G proteins different from Gq/11 protein, as
shown by the absence of effect of intracellular application of
anti-q/
11 antibodies (27). Therefore, we
used caffeine as a control for the availability of the Ca2+
stores in control and oligonucleotide-injected cells, whereas the
angiotensin AT1 receptor-induced response was used as a
control for the specificity of antisense oligonucleotide effects on G protein subunits. Antisense oligonucleotides against G
subunits showed no effect on the caffeine-induced Ca2+ responses
excluding non-antisense effects on intracellular Ca2+
stores (see Fig. 2). Such non-antisense but sequence-specific effects
have been described for oligonucleotides directed against c-myb
and p53 (28-29). To have a second control for the specificity of
the antisense oligonucleotides, we compared injected cells to
non-injected control cells located on the same glass slide in each
experiment. We show that the value of the increase in [Ca2+]i induced by norepinephrine is not
significantly different comparing non-injected cells to cells injected
with
q/11 sense or scrambled antisense oligonucleotides
and antisense oligonucleotides directed against G protein subunits
which are not involved in the
1-adrenoreceptor-induced
increase in [Ca2+]i (see Figs. 1 and 2).
Furthermore, injection of antisense oligonucleotides directed against
G
subunits involved in
1-adrenoreceptor-mediated effects did not change the increase in [Ca2+]i
achieved by angiotensin II or caffeine (see Fig. 2, B-C). To demonstrate the extent of the antisense
knockout effects, we studied protein depletion by immunocytochemistry
of G
subunits. The results of these experiments revealed that the
time course by which anti-
q or anti-
11
antisense oligonucleotides were effective in suppressing functional
receptor-mediated effects paralleled suppression of the
G
q or G
11 protein level which decreased
maximally within 3 days after injection. Similar results were recently
obtained using
o-,
i-, and
q/
11-antisense oligonucleotides in
different cells (8, 12, 30). In addition, we demonstrated that
injection of antisense oligonucleotides directed against a given G
protein subunit did not modify significantly the expression of other G protein subunits (see Fig. 4). Therefore, the fact that
anti-
1, -
3, -
2, and
-
3 oligonucleotides inhibit the
1-adrenoreceptor-induced [Ca2+]i
increase cannot be related to an inhibition of the G
subunit
expression and means that these
and
subunits are necessary for
activation of the Gq and G11 proteins by
1-adrenoreceptors. Since anti-
q,
-
1, and -
3 oligonucleotides largely
inhibited the
1-adrenoreceptor-induced increase in
[Ca2+]i whereas anti-
11,
-
3 and -
2 oligonucleotides produced a
limited inhibition, one may speculate that the composition of G protein
heterotrimers required for the two phases of the
1-adrenoreceptor-induced Ca2+ response is
q/
1/
3 and
11/
3/
2.
The experiments presented in this work suggest that two
different heterotrimeric G proteins mediate Ca2+
release from intracellular stores and Ca2+ entry in
response to stimulation of 1-adrenoreceptors. Evidence supporting this proposal are the following. 1) Cells injected with a
mixture of anti-
q and anti-
11
oligonucleotides (anti-
q+11) showed no further reduction
of
1-adrenoreceptor-induced Ca2+ response
compared with cells injected with either anti-
q or anti-
q/
11 oligonucleotides. 2) In
Ca2+-free solution, anti-
q oligonucleotides
strongly reduced the
1-adrenoreceptor-induced
Ca2+ release, whereas anti-
11
oligonucleotides were without effect, suggesting that
G
11 protein-mediated modulation of oxodipine-resistant Ca2+ entry required a preceding Ca2+ release
mediated by G
q protein. 3) The oxodipine-resistant
Ca2+ entry evoked by activation of
1-adrenoreceptors in the presence of thapsigargin or
caffeine was selectively suppressed by anti-
11 oligonucleotides. Therefore, we propose that the G
11
subunit may enhance the
1-adrenoreceptor-induced
Ca2+ entry activated by a previous release of
Ca2+ from intracellular stores. Several types of
Ca2+ entry mechanisms have been described in various
cellular systems (31). In smooth muscle cells, activation of
Ca2+ entry by application of mediators (histamine,
endothelin, vasopressin) has been reported (32-34). We show that
caffeine also induces intracellular store-operated Ca2+
entry in rat portal vein myocytes (see Fig. 6), and we used caffeine and thapsigargin pretreatment to study the modulation of the
store-operated Ca2+ entry by activation of
1-adrenoreceptors. This dihydropyridine-resistant Ca2+ entry may be mediated by cation channels.
Interestingly, a nonselective cation channel, the Drosophila
trpl channel, has been shown to be stimulated in a membrane-confined
way by G
11 protein (35), and a similar nonspecific
cation channel permeable for Ca2+ ions has been previously
identified in portal vein myocytes (36). Furthermore, experiments
performed in an epithelial cell line have shown that overexpression of
G
q protein increases Ca2+ release-activated
Ca2+ influx (37). Thus, the ubiquitously expressed
G
q family may have a general role in modulating
Ca2+ entry through Ca2+-permeable nonselective
cation channels which may be controlled by both the filling state of
Ca2+ stores (38) and, as shown here, more directly by G
protein. Recently, the composition of G proteins coupling the stably
expressed human muscarinic m1 receptor in the rat
basophilic leukemia cell line (RBL-2H3-hm1) to increase in
[Ca2+]i has been determined by using the same
method and the same antisense oligonucleotides (12). In these cells,
the authors have proposed that a complex of G protein subunits,
i.e.
G
q/
11·
1/
4·
4, is activated by m1 receptors. As
[Ca2+]i measurements were performed in
Ca2+-containing solution, a differential coupling of
G
q and G
11 subunits to Ca2+
release and Ca2+ entry, respectively, could not have been
detected. Finally, transient expression of a carboxyl-terminal fragment
of
ARK1 that scavenged G
subunits after their
dissociation from the receptor-activated heterotrimers had no effect on
1-adrenoreceptor-induced Ca2+ responses
suggesting that
subunits did not display direct interactions
with the effectors, i.e. phospholipase C and
Ca2+-permeable nonselective cation channels. This
conclusion is supported by the results showing that intracellular
application of anti-
com antibody did not modify
significantly the
1-adrenoreceptor-mediated Ca2+ release and Ca2+ entry. The possibility
that G
subunits that are dissociated from both G
q
and G
11 subunits after activation of
1-adrenoreceptors may activate other cellular effectors
remains to be investigated.
In conclusion, we show that in rat venous myocytes, the Gq
proteins may couple by their subunits endogenous
1-adrenoreceptors to phospholipase C, whereas the
G
11 proteins, activated at the same time by the same
receptors, may couple to Ca2+ entry. These results point
out distinct functions of Gq and G11 in
receptor-activated [Ca2+]i increase.
We thank Dr. R. J. Lefkowitz for generously
donating ARK1 minigene; Drs. K. Spicher and B. Nürnberg for anti-
q/11, anti-
3, and
anti-
2 antibodies; Dr. A. Lückhoff for helpful
discussions, and N. Biendon for secretarial assistance.