Neuropharmacology 42 (2002) 950–957
www.elsevier.com/locate/neuropharm
Adenovirus-mediated Gαq-protein antisense transfer in neurons
replicates Gαq gene knockout strategies
F.C. Abogadie a,∗, R. Bron b, S.J. Marsh c, L.J. Drew d, J.E. Haley a, N.J. Buckley e,
D.A. Brown c, P. Delmas a,f
a
Wellcome Laboratory for Molecular Pharmacology, University College London, Gower Street, London WC1E 6BT, UK
b
MRC Centre for Developmental Neurobiology, New Hunt’s House, King’s College London, London SE1 1UL, UK
c
Department of Pharmacology, University College London, Gower Street, London WC1E 6BT, UK
d
Department of Biology, University College London, Gower Street, London WC1E 6BT, UK
e
School of Biochemistry and Molecular Biology, University of Leeds, Leeds LS2 9JT, UK
f
ITIS-CNRS, 31 Chemin Joseph Aiguier, 13402 Marseille cedex 20, France
Received 12 November 2001; received in revised form 7 March 2002; accepted 18 March 2002
Abstract
Antisense approaches are increasingly used to dissect signaling pathways linking cell surface receptors to intracellular effectors.
Here we used a recombinant adenovirus to deliver G-protein αq antisense into rat superior cervical ganglion (SCG) neurons and
neuronal cell lines to dissect Gαq-mediated signaling pathways in these cells. This approach was compared with other Gαq gene
knockdown strategies, namely, antisense plasmid and knockout mice. Infection with adenovirus expressing Gαq antisense (GαqAS
AdV) selectively decreased immunoreactivity for the Gαq protein. Expression of other Gα protein subunits, such as GαoA/B, was
unaltered. Consistent with this, modulation of Ca2+ currents by the Gαq-coupled M1 muscarinic receptor was severely impaired in
neurons infected with GαqAS AdV whereas modulation via the GαoA-coupled M4 muscarinic receptor was unchanged. In agreement,
activation of phospholipase C and consequent mobilization of intracellular Ca2+ by UTP receptors was lost in NG108-15 cells
infected with GαqAS AdV but not in cells infected with the control GFP-expressing adenovirus. Results obtained with this recombinant AdV strategy qualitatively and quantitatively replicated results obtained using SCG neurons microinjected with Gαq antisense
plasmids or SCG neurons from Gαq knockout mice. This combined antisense/recombinant adenoviral approach can therefore be
useful for dissecting signal transduction mechanisms in SCG and other neurons.  2002 Elsevier Science Ltd. All rights reserved.
Keywords: Adenovirus; G-protein; Calcium current; Antisense; Superior cervical ganglion
1. Introduction
Modulation of gene expression using antisense strategies is an area of intense study, both from therapeutic
and research perspectives (Crooke, 2000; Pachori et al.,
2001; Su et al., 2001). Antisense approaches are increasingly used in the dissection of signaling pathways due
to their potential to specifically target individual members of homologous gene families. We have previously
used antisense cDNA expression vectors and antisense
oligonucleotides to specifically alter expression of indi-
Corresponding author. Tel.: +44-20-7679-2999; fax: +44-207679-7245.
E-mail address: f.abogadie@ucl.ac.uk (F.C. Abogadie).
∗
vidual G-protein subunits (Abogadie et al., 1997; Buckley et al., 2000). Using these approaches, we have identified the G-protein subunits involved in mediating the
inhibition of potassium M-channels and N-type Ca2+
channels in rat superior cervical ganglion (SCG) neurons
(Haley et al., 1998; Delmas et al., 1998a, 1999).
Although these studies have yielded a great deal of information on specific receptor/G-protein/effector interactions, both approaches suffer from limitations. We found
that the effects of antisense oligonucleotides can be
inconsistent (Abogadie et al., 1997). On the other hand,
cDNA expression vectors can only be delivered by
intranuclear injection or by transfection into a subpopulation of mitotic cells. We needed a method of gene delivery that is more suitable to large populations of cells and
to more refractory smaller neurons of the central nervous
0028-3908/02/$ - see front matter  2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 2 8 - 3 9 0 8 ( 0 2 ) 0 0 0 4 4 - 8
F.C. Abogadie et al. / Neuropharmacology 42 (2002) 950–957
system. Therefore, we explored the possibility of an
adenovirus-mediated gene transfer to deliver G-protein
antisense constructs to neurons.
Replication-deficient recombinant adenoviruses are
widely used for gene delivery to post-mitotic cells, both
in experimental systems and increasingly in the area of
gene therapy. Adenoviruses offer many advantages
including relatively low pathogenicity in humans, wide
host range and high replication efficiency (Le Gal La
Salle et al., 1993; Russell, 2000). In this study we show
that adenoviral-based vectors can be successfully used
to deliver Gαq antisense constructs to sympathetic neurons as well as neuronal cell lines, that they specifically
attenuate Gαq-mediated cellular functions, and that this
approach replicates our results obtained using other strategies such as the use of antisense plasmids and knockout animals. This combined strategy of using antisense
and adenovirus should enable a wide variety of studies
investigating receptor/G-protein/effector signaling pathways in neurons.
2. Methods
2.1. Cell cultures
SCG neurons were dissociated from young rats (15–
19 days old) and cultured on glass coverslips as
described previously (Delmas et al., 1998b). Rats were
sacrificed by exposure to carbon dioxide and decapitation according to the UK Animals (Scientific
Procedures) Act, 1986. Methods for culturing NG10815 mouse neuroblastoma rat glioma cells and mouse
sympathetic neurons were described previously
(Filippov and Brown, 1996; Haley et al., 2000).
Organotypic hippocampal cultures were prepared using
the interface method (Stoppini et al., 1991). Briefly, 350
µm thick hippocampal slices were taken from 8-day old
Sprague–Dawley rats and cultured on small pieces of
Millicell-CM membrane (Millipore) inside Millicell-CM
inserts. Cultures were maintained in 50% MEM, 25%
horse serum and 25% HBSS at 37 °C for 5 days and
then subsequently at 33 °C.
2.2. Generation of recombinant adenoviruses and
infection
The antisense insert corresponding to the 3⬘ untranslated region nucleotides 29–129 of rat Gαq was subcloned from clone C23-16 (Haley et al., 1998) to the
EcoRV/HindIII site of the shuttle vector pAT-CMV (He
et al., 1998). The resultant plasmid, pAT-CMV-GαqAS,
as well as control pAT-CMV plasmid without insert,
were linearized with PmeI and each co-transformed with
the supercoiled adenoviral plasmid pAdEasy-1 into electrocompetent BJ5183 cells. In each case, a recombinant
951
adenoviral plasmid was generated using homologous
recombination in bacteria (He et al., 1998). Plasmids
were screened by restriction digests and verified by
sequencing. The chosen recombinant clones were linearized and used to transfect host HEK 293A cells
(Quantum Biotechnologies Inc., Montreal, Canada) to
generate infectious virus particles. The viral production
process was monitored using GFP fluorescence. High
titer viral stocks (2 × 1010pfu / ml) were produced after
four rounds of amplification. Final virus harvests were
purified using CsCl gradient centrifugation and dialyzed
against Tris/HEPES buffer (200 mM Tris, 50 mM
HEPES at pH 8). Viruses were then stored in aliquots
at ⫺80 °C in the same buffer containing 10% glycerol.
Standard plaque assays using HEK 293A cells were done
to determine the titer of the viral stocks. Cultured cells
(SCG neurons in primary cultures and NG108-15 cells)
were infected with 1 × 107–1 × 108pfu / ml of either control GFP AdV or GαqAS AdV. Cultures were exposed
to the inoculum overnight before changing to fresh
medium the next day. Infection of CA3 hippocampal
neurons in organotypic slice cultures (21 days in vitro)
was achieved using local microinjection of viral solutions (1 × 108pfu / ml). Slices were fixed 2 days postinfection and GFP expression was visualized using laser
scanning confocal microscopy.
2.3. Immunodetection and image analysis
Immunostaining was performed 3 days post-infection
largely as described previously (Haley et al., 1998;
Delmas et al., 1998b) using anti-Gαq polyclonal antibody (IQB2, 1/1000) and anti-Gα0 monoclonal antibody
(MAB3073, 1/1000, Chemicon). Secondary anti-rabbit
and anti-mouse IgGs were from Dako (A/S, Denmark)
and Molecular Probes (Leiden, The Netherlands),
respectively. Digital images were collected with a Nikon
Eclipse 800 fluorescence microscope, connected to a PC
via a Hamamatsu CCD camera. The fluorescence intensity of the staining was analyzed using an ImagePro Plus
software package (Media Cybernetics, Silver Spring,
MD). Background fluorescence was excluded in Fig.
2(B).
2.4. Whole-cell recording
Calcium currents were recorded using the whole-cell
configuration of the patch clamp technique as described
in Delmas et al. (2000). Briefly, pipettes had resistances
of 2–3.5 M⍀ when filled with an internal solution consisting of (in mM): 120 NMDG, 14 CsCl, 10 HEPES, 5
EGTA, 0.5–1 CaCl2, 10 phosphocreatine, 4 MgATP and
0.2 Na2GTP (pH 7.34). A solution containing 20 mM
BAPTA and 0.1 mM CaCl2 was used to record the M4
muscarinic inhibition in relative isolation (Beech et al.,
1991). The extracellular solution consisted of (in mM):
952
F.C. Abogadie et al. / Neuropharmacology 42 (2002) 950–957
130 NaCl, 3 KCl, 1 MgCl2, 10 Hepes, 0.0005 TTX, 2
CaCl2, 11 glucose (pH 7.3). Currents were measured
with an Axopatch 200A amplifier (Axon Instruments),
filtered at 2 kHz and leak subtracted. All experiments
were performed at 30–32 °C.
2.5. Measurement of intracellular calcium
Intracellular free Ca2+ was measured using dual excitation fluorescence of the Ca2+-sensitive probe Fura-2.
Fura-2 was loaded into NG108-15 cells using the acetoxymethyl-ester form of the probe (Fura-AM, 5 M, for
30 min). Images were recorded using an 8-bit gray scale
camera controlled using OpenLab acquisition software.
Images were obtained at excitation wavelengths of 345
and 380 nm and emitted light captured at 520 nm.
Exposure time was 100 ms. Images were ratioed after
background subtraction and free Ca2+ was estimated
from within 25 µm2 areas of the cytoplasm. Calibration
curves were constructed using whole-cell estimates of
Ca2+ concentration (Trouslard et al., 1993).
3. Results
Fig. 1. Generation and expression of recombinant GαqAS adenovirus. (A) Nucleotide sequence of target rat Gαq 3⬘ untranslated region.
(B) Structure of the recombinant plasmid pAdV GFP GαqAS. The
plasmid contains the following genetic elements: LITR, left internal
terminal repeat; PA, polyadenylation site; GFP, green fluorescent protein; CMV, cytomegalovirus promoter; Kan, kanamycin resistance
gene; GαqAS, a 101-bp long gene fragment antisense to a segment of
rat Gαq 3⬘ untranslated region. (C) Shown are CA3 hippocampal neurons within an organotypic hippocampal culture (23 days in vitro) 3
days post-infection. Cell bodies labeled red (arrows) indicate dying
cells that have absorbed propidium iodide, a dye that is taken up by
cells with damaged membranes and binds to nucleic acids (Noraberg
et al., 1999).
The Gαq antisense construct used here corresponds to
nucleotides 29–129 of the 3⬘ untranslated region of the
rat G-protein αq subunit (Fig. 1(A)) and has previously
been shown to be specific for G-protein αq subunit versus other G-protein α subunits (Haley et al., 1998;
Delmas et al., 1998a). The antisense insert was subcloned into the shuttle vector pAT-CMV giving the plasmid pAT-CMV-GαqAS which contains a GFP reporter
gene and the Gαq antisense construct (Fig. 1(B)). Replication-deficient recombinant adenoviruses were then
generated by bacterial recombination (see Section 2) (He
et al., 1998).
We first determined the optimal virus dose that would
give the desired infection rate and a convenient window
of time during which electrophysiological and immunological experiments could be carried out. When using
purified GαqAS AdV at 1 × 107–1 × 108pfu / ml, about
40–60% of SCG neurons in primary culture were GFPpositive 2–3 days after exposure to the virus. The same
titer of virus was able to infect ⬎60% of NG108-15 cells
and many hippocampal neurons in organotypic cultures
(data not shown but see Fig. 1(C)). We then tested for
the development of cytopathic effects by incubating
infected cells with propidium dye, a dye that is taken up
by cells with damaged membranes and binds to nucleic
acids (Noraberg et al., 1999). Using a range of titers
from 1 × 107 to 1 × 108pfu / ml, we did not notice any
cell death among the infected neurons up to 5 days postinfection. A representative example of hippocampal neurons maintained in organotypic culture 3 days post-infection is shown in Fig. 1(C). This titer range,
F.C. Abogadie et al. / Neuropharmacology 42 (2002) 950–957
953
Fig. 2. GαqAS adenovirus specifically reduces Gαq expression in infected SCG neurons. (A) Representative immunostaining for Gαq and Gαo
subunits in neurons unexposed to the virus. (B) Corresponding images of neurons infected with either control GFP AdV or GαqAS AdV (left
panels) and stained for Gαq or Gαo subunits (right panels). (C) Population-based image analysis of Gαq and Gαo immunoreactivities in unexposed
neurons and infected neurons. Data were fitted to Gaussian distributions (0.945⬍R2⬍0.985) giving luminance values (mean ± SD, in arbitrary
units) for Gαq immunofluorescence of: 739 ± 135 (unexposed), 764 ± 107 and 598 ± 118 (control GFP AdV) and 376 ± 183 (GαqAS AdV) and
for Gαo immunofluorescence: 763 ± 172 (unexposed), 707 ± 115 (control GFP AdV) and 657 ± 154 (GαqAS AdV). n ⬎ 100 for each group except
for cells treated with the control GFP AdV and stained for Gαo, n ⫽ 48.
1 × 107–1 × 108pfu / ml, was therefore used in all subsequent experiments.
Next, we studied the effects of GαqAS AdV on the
level of expression of Gαq subunits in SCG neurons 3
days after infection. For controls, we looked at the
effects of GαqAS AdV on the expression of Gαo subunits and at the immunostaining levels of Gαq in neurons
infected with the control GFP AdV (Fig. 2(A) and (B)).
In most SCG neurons infected with GαqAS AdV, a clear
decrease in levels of Gαq immunofluorescence could be
seen when compared with unexposed SCG neurons or
neurons infected with the control GFP AdV (titer
1 × 108pfu / ml). Gαo immunofluorescence in cells
infected with GαqAS AdV was not appreciably altered.
We tried to assess the effects of GαqAS AdV on the
expression levels of Gα11, a closely related sequence to
Gαq, but were precluded by the low levels of Gα11
expression in these cells (Caulfield et al., 1994). Hence,
although the antisense was specifically directed against
Gαq, we cannot exclude non-specific changes in Gα11
expression in these experiments.
To account for cell-to-cell variability in staining and
infection we performed a population-based image analysis of the level of Gα immunofluorescence in individual
neurons for each treatment (Fig. 2(C)). This analysis of
over 100 cells for each group revealed that the control
GFP AdV by itself produced a slight reduction in
expression of both Gαo and Gαq subunits (7–19%
decrease in mean immunoreactivity). This effect is consistent with previous reports showing a decrease in total
protein synthesis in various cell types including SCG
neurons infected with replication-deficient adenovirus
(Paquet et al., 1996; Easton et al., 1998). Expression of
GαqAS AdV caused a further reduction of mean Gαq
immunofluorescence to 51% of its control value but had
no significant effect on Gαo staining (87% of the control
value). As shown by the Gaussian distribution in Fig.
2(C) reduction of Gαq staining was very variable, ranging from strong attenuation to no attenuation at all.
We then tested the effects of GαqAS AdV on the M1
muscarinic inhibition of Ca2+ currents (ICa) in response
to 10 µM oxotremorine-M (Oxo-M). This pathway
mainly requires Gαq (Delmas et al., 1998a; Haley et al.,
2000). Neurons were recorded 48 h after infection and
the M4 muscarinic inhibition of ICa (a pathway mediated
via GaoA, Delmas et al., 1998a) was taken as control.
Whole-cell Ca2+ currents recorded at this time showed
normal current density (25–45 pA/pF), activation kinetics and voltage dependence. The M4 muscarinic pathway was recorded in isolation by using pipette solution
954
F.C. Abogadie et al. / Neuropharmacology 42 (2002) 950–957
containing a high concentration of BAPTA, whereas the
M1 muscarinic pathway was recorded in cells treated
with Pertussis toxin (PTX) (Beech et al., 1991; Delmas
et al., 1998a). The voltage dependence of inhibition was
routinely tested to make sure of the adequate isolation
of the M4 (V-dependent) and M1 (V-independent) muscarinic pathways (Fig. 3(A)). Cells infected with GαqAS
AdV had M4 muscarinic inhibition of ICa not signifi-
cantly different from uninfected cells or cells infected
with control GFP AdV (Fig. 3(A) and (B)). Thus, stimulating M4 muscarinic receptors inhibited ICa by
52 ± 5% in uninfected cells and by 49 ± 6% and
54 ± 7% in cells infected with either control GFP AdV
or GαqAS AdV, respectively. In contrast, M1 muscarinic
inhibition was reduced by 70–75% in cells infected with
GαqAS AdV when compared with uninfected neurons
Fig. 3. GαqAS AdV specifically blocks M1 muscarinic pathway. (A) M4 (left panels) and M1 (right panels) muscarinic inhibition of Ca2+ currents
by 10 µM oxotremorine-M (Oxo-M) in SCG neurons infected with either control GFP AdV or GαqAS AdV. Currents were evoked using the threevoltage step protocol illustrated in inset. The time between the prepulse to +90 mV and the second test pulse to 0 mV was incremented by 5 ms
on successive recordings. M4-mediated inhibition was isolated by adding 20 mM BAPTA to the pipette solution, whereas M1-mediated inhibition
was isolated by pretreatment with Pertussis toxin (see Section 2). Inhibition by M4 receptor stimulation is transiently reversed by the depolarizing
prepulse (i.e. it is voltage dependent) whereas M1-mediated inhibition is not (see Delmas et al., 1998a). Note that M1 but not M4 muscarinic
inhibition was strongly attenuated by GαqAS AdV. (B) Summary of Ca2+ current inhibition (mean ± S.E.M) by M4 and M1 muscarinic receptors
in SCG neurons infected with either control GFP AdV or GαqAS AdV and for comparison in neurons microinjected with Gαq antisense-plasmids
(Delmas et al., 1998a) and in SCG neurons collected from Gαq knockout mice (Haley et al., 2000). Patch clamp recordings were made 48 h after
infection or injection. ∗∗, P ⬍ 0.001 (two-tailed Student t-test).
F.C. Abogadie et al. / Neuropharmacology 42 (2002) 950–957
(not shown) or neurons infected with control GFP AdV
(Fig. 3(A) and (B)). A similar amount of reduction was
obtained in neurons nuclearly microinjected with Gαq
antisense plasmids (400 µg/ml, Fig. 3(B); see also
Delmas et al., 1998a) or when the M1 muscarinic modulation was recorded in SCG neurons isolated from Gαq
knockout mice (Fig. 3(B); see also Haley et al., 2000).
Finally, we tested whether the GαqAS AdV would
prevent mobilization of intracellular Ca2+ by the phospholipase C-linked P2Y receptor in NG108-15 neuronal
hybrid cells (Erb et al., 1993; Lustig et al., 1993). Intracellular free Ca2+ was simultaneously monitored in noninfected cells and cells infected with GαqAS AdV using
the calcium probe Fura-2 (Fig. 4). The P2Y receptor
agonist UTP was quite effective in mobilizing Ca2+ in
uninfected
NG108-15
cells
(peak [Ca2+]I ⫽
590 ± 45nM, n ⫽ 6) as well as in cells infected with
control GFP AdV (495 ± 30nM, n ⫽ 7) but was largely
ineffective in cells infected with the GαqAS AdV
(45 ± 10nM, n ⫽ 7). The lack of response in these cells
was not due to prior Ca2+ depletion of intracellular stores
by GαqAS AdV since application of thapsigargin, an
inhibitor of endoplasmic Ca2+ ATPase, still released Ca2+
from intracellular stores (data not shown). Since Gα11AS
AdV was not tested in the present study, we cannot
exclude a contribution of Gα11 in the coupling of P2Y
receptors to PLC in NG108-15 cells.
Fig. 4. Block of PLC-induced Ca2+ mobilization by GαqAS AdV in
NG108-15 cells. Upper panels: NG108-15 cells infected with GαqAS
AdV (GFP) and imaged 52 h later. The arrow shows an uninfected
cell. Intracellular free Ca2+ was monitored prior (1) and upon (2) application of UTP (50 µM). Calcium release from the intracellular stores
was strongly prevented in the infected cell.
955
4. Discussion
We set out to determine the feasibility of using recombinant adenovirus to deliver antisense transcripts to target neurons. Our results show that: (1) recombinant
adenovirus can be successfully used to deliver Gαq antisense to neurons in culture, (2) its delivery results in a
specific reduction of the target Gαq subunit which consequently attenuates Gαq-mediated cellular functions, and
(3) virally induced antisense transgene expression replicates data obtained using other Gαq knockout strategies.
At virus titers of at least 1 × 107pfu / ml nearly 60%
of the SCG neurons in our primary culture and most of
the NG108-15 cells were infected. This titer seems to be
optimal for obtaining a high level of transgene
expression without major cytopathic effect. This infection rate is a vast improvement over intranuclear injections, where we could inject only a few cells at a time.
Infection with GαqAS AdV inhibited M1 modulation
of ICa in sympathetic neurons and Ca2+ mobilization produced by PLC-coupled UTP receptors in neuroblastoma
cells. These inhibitory effects were not shared by the
control GFP AdV, indicating that the blockade of these
two well-known Gαq-dependent pathways resulted from
the depletion of Gαq subunits. The maximal amount of
inhibition of the M1/Gαq pathway obtained with viral
expression of Gαq antisense was very similar to those
obtained by using antisense plasmid and knockout strategies (see Fig. 3(B)). This suggests that the residual M1
modulation in Gαq-depleted cells may in fact result from
the participation of a G-protein other than Gαq. To this
extent, these findings confirm previous conclusions using
G-protein antibody injections, antisense depletion and
knockout animals on the principal role played by Gαq
in coupling M1 muscarinic receptor to Ca2+ channels in
sympathetic neurons (Delmas et al., 1998b; Haley et
al., 2000).
Gαq coupling to Ca2+ channels was virtually abolished
in most recorded cells, while decrease in Gαq protein
level, as detected by immunostaining, was variable. This
discrepancy could be because patch clamp experiments
were made on brighter cells, which would express more
antisense transcripts. Also, while electrophysiology
reveals levels of functional G-proteins in the plasma
membrane, immunostaining detects Gαq epitope in both
the cytosol and the membrane. In cells in which translation was prevented, the detected Gαq epitope may be
part of a degraded, non-functional, protein.
Previous studies have demonstrated the efficiency of
adenoviral vectors in transferring foreign genes into
post-mitotic neurons (Ehrengruber et al., 1998). Our
work points to the feasibility of using adenoviralmediated gene transfer in combination with an antisense
strategy to investigate the role of G-proteins in neuronal
signaling. At the titer we used, viral infection did not
noticeably alter cellular function, although we observed
956
F.C. Abogadie et al. / Neuropharmacology 42 (2002) 950–957
a general reduction of G-protein levels at early stages
after infection that was not related to the expression of
the inserted antisense transgene. These effects may result
from the expression of adenoviral genes, which leads to
biochemical and/or genetic perturbations within the
infected cell (Easton et al., 1998). These perturbations
complicated the interpretation of data regarding transgene antisense function and emphasized the need for us
to assess the effects of control GFP AdV on different
Gα subunit proteins and to define a time frame in which
cells were not affected by the viral infection. From our
analysis, no severe cellular alteration was observed
within the first 72 h, the time frame of our functional
experiments, suggesting that short-term analysis of the
antisense transgene may be done with confidence.
The present study sets the stage for the use of adenoviral-mediated antisense gene transfer to study G-protein
signaling in neurons and defines the limit of the usefulness of these vectors in antisense studies. Further
improvements of this delivery system rely on the development of adenoviral vectors that have high transduction
efficiency and high insert capacity, express minimal or
no adverse side effects and incorporate elements for
tissue-specific expression. Currently, under development
and looking promising in this respect are vectors devoid
of all but the essential viral coding sequences and vectors
that allow for regulated expression (Schiedner et al.,
1998; Burcin et al., 1999; Millecamps et al., 1999; Reynolds et al., 2001).
Acknowledgements
The authors wish to thank Dr T.C. He for the gift of
the pAdEasy-1 and pAT-CMV plasmids as well as the
BJ5183 cells, Professor G. Milligan for the Gαq antiserum, Dr A. Roopra for valuable discussions, and Ms
M. Dayrell for technical expertise. This work was supported by the Wellcome Trust and the UK Medical
Research Council.
References
Abogadie, F.C., Vallis, Y., Buckley, N.J., Caulfield, M.P., 1997. Use
of antisense-generating plasmids to probe the function of signal
transduction proteins in primary neurons. In: Chaliss, R.A.J. (Ed.),
Methods in Molecular Biology. Receptor Signal Transduction Protocols, 83. Humana Press, Totowa, NJ, pp. 217–225.
Beech, D.J., Bernheim, L., Mathie, A., Hille, B., 1991. Intracellular
Ca2+ buffers disrupt muscarinic suppression of Ca2+ current and
M current in rat sympathetic neurons. Proceedings of the National
Academy of Sciences 88, 652–656.
Buckley, N.J., Abogadie, F.C., Brown, D.A., Dayrell, M., Caulfield,
M.P., Delmas, P., Haley, J.E., 2000. Use of antisense expression
plasmids to attenuate G-protein expression in primary neurons.
Methods in Enzymology 314, 136–148.
Burcin, M.M., Schiedner, G., Kochanek, S., Tsai, S.Y., O’Malley,
B.W., 1999. Adenovirus-mediated regulable target gene expression
in vivo. Proceedings of the National Academy of Sciences 96,
355–360.
Caulfield, M.P., Jones, S., Vallis, Y., Buckley, N.J., Kim, G.D., Milligan, G., Brown, D.A., 1994. Muscarinic M-current inhibition via
G alpha q/11 and alpha-adrenoceptor inhibition of Ca2+ current via
G alpha o in rat sympathetic neurones. Journal of Physiology 477,
415–422.
Crooke, S.T., 2000. Progress in antisense technology: the end of the
beginning. Methods in Enzymology 313, 3–45.
Delmas, P., Abogadie, F.C., Dayrell, M., Haley, J.E., Milligan, G.,
Caulfield, M.P., Brown, D.A., Buckley, N.J., 1998a. G-proteins and
G-protein subunits mediating cholinergic inhibition of N-type calcium currents in sympathetic neurones. European Journal of Neuroscience 10, 1654–1666.
Delmas, P., Brown, D.A., Dayrell, M., Abogadie, F.C., Caulfield, M.P.,
Buckley, N.J., 1998b. On the role of the endogenous G-protein βγ
subunits in N-type Ca2+ current inhibition by neurotransmitters in
rat sympathetic neurones. Journal of Physiology (London) 506,
319–329.
Delmas, P., Abogadie, F.C., Milligan, G., Buckley, N.J., Brown, D.A.,
1999. βγ dimers derived from Go and Gi proteins contribute different components of adrenergic inhibition of Ca2+ channels in sympathetic neurons. Journal of Physiology (London) 518, 23–36.
Delmas, P., Abogadie, F.C., Buckley, N.J., Brown, D.A., 2000. Calcium channel gating and modulation by transmitters depend on
cellular compartmentalization. Nature Neuroscience 3, 670–678.
Easton, R.M., Johnson, E.M., Creedon, D.J., 1998. Analysis of events
leading to neuronal death after infection with E1-deficient adenoviral vectors. Molecular and Cellular Neuroscience 11, 334–347.
Ehrengruber, M.U., Lanzrein, M., Xu, Y., Jasek, M.C., Kantor, D.B.,
Schuman, E.M., Lester, H.A., Davidson, N., 1998. Recombinant
adenovirus-mediated expression in nervous system of genes coding
for ion channels and other molecules involved in synaptic function.
In: Conn, P.M. (Ed.), Ion Channels Part B. Methods in Enzymology, 293. Academic Press, New York, pp. 483–503.
Erb, L., Lustig, K.D., Sullivan, D.M., Turner, J.T., Weisman, G.A.,
1993. Functional expression and photoaffinity labelling of a cloned
P2U purinergic receptor. Proceedings of the National Academy of
Sciences 90, 10449–19453.
Filippov, A.K., Brown, D.A., 1996. Activation of nucleotide receptors
inhibits high-threshold calcium currents in NG108-15 neuronal
hybrid cells. European Journal of Neuroscience 8, 1149–1155.
Haley, J.E., Abogadie, F.C., Delmas, P., Dayrell, M., Vallis, Y., Milligan, G., Caulfield, M.P., Brown, D.A., Buckley, N.J., 1998. The α
subunit of Gq contributes to muscarinic inhibition of the M-type
potassium current in sympathetic neurons. Journal of Neuroscience
18, 4521–4531.
Haley, J.E., Delmas, P., Offermanns, S., Abogadie, F.C., Simon, M.I.,
Buckley, N.J., Brown, D.A., 2000. Muscarinic inhibition of calcium current and M current in Gαq-deficient mice. Journal of Neuroscience 20, 3973–3979.
He, T.C., Zhou, S., Da Costa, L.T., Yu, J., Kinzler, K.W., Vogelstein,
B., 1998. A simplified system for generating recombinant adenoviruses. Proceedings of the National Academy of Sciences 95,
2509–2514.
Le Gal La Salle, G.L.G., Robert, J.J., Berrard, R.S., Ridoux, V., Stratford-Perricaudet, L.D., Perricaudet, M., 1993. An adenovirus vector
for gene transfer into neurons and glia in the brain. Science 259,
988–990.
Lustig, K.D., Shiau, A.K., Brake, A.J., Julius, D., 1993. Expression
cloning of an ATP receptor from mouse neuroblastoma cells. Proceedings of the National Academy of Sciences 90, 5113–5117.
Millecamps, S., Kiefer, H., Navarro, V., Geoffroy, M.C., Robert, J.J.,
Finiels, F., Mallet, J., Barkats, M., 1999. Neuron-restrictive silencer
elements mediate neuron specificity of adenoviral gene expression.
Nature Biotechnology 17, 865–869.
F.C. Abogadie et al. / Neuropharmacology 42 (2002) 950–957
Noraberg, J., Kristensen, B.W., Zimmer, J., 1999. Markers for neural
degeneration in organotypic slice cultures. Brain Research Protocols 3, 278–290.
Pachori, A.S., Huentelman, M.J., Francis, S.C., Gelband, C.H., Katovich, M.J., Raizada, M.K., 2001. The future of hypertension therapy:
sense, antisense, or nonsense? Hypertension 37 (2), 357–364.
Paquet, L., Massie, B., Mains, R.E., 1996. Proneuropeptide Y processing in large dense-core vesicles: manipulation of prohormone
convertase expression in sympathetic neurons using adenoviruses.
Journal of Neuroscience 16, 964–973.
Reynolds, P.N., Nicklin, S.A., Kaliberova, L., Boatman, B.G., Grizzle,
W.E., Balyasnikova, I.V., Bakers, A.H., Danilov, S.M., Curiel,
D.T., 2001. Combined transductional and transcriptional targeting
improves the specificity of transgene expression in vivo. Nature
Biotechnology 19, 838–842.
Russell, W.C., 2000. Update on adenovirus and its vectors. Journal of
General Virology 81, 2573–2604.
957
Schiedner, G., Morral, N., Parks, R.J., Wu, Y., Koopmans, S.C., Langston, C., Graham, F.L., Beaudet, A.L., Kochanek, S., 1998. Genomic DNA transfer with a high-capacity adenovirus vector results
in improved in vivo gene expression and decreased toxicity. Nature
Genetics 18, 180–183.
Stoppini, L., Buchs, P.-A., Muller, D., 1991. A simple method for
organotypic cultures of nervous tissue. Journal of Neuroscience
Methods 37, 173–182.
Su, Z.Z., Lebedeva, I.V., Gopalkrishnan, R.V., Goldstein, N.I., Stein,
C.A., Reed, J.C., Dent, P., Fisher, P.B., 2001. A combinatorial
approach for selectively inducing programmed cell death in human
pancreatic cancer cells. Proceedings of the National Academy of
Sciences 98, 10332–10337.
Trouslard, J., Marsh, S.J., Brown, D.A., 1993. Calcium entry through
nicotinic receptor channels and calcium channels in cultured rat
superior cervical ganglion cells. Journal of Physiology 468, 53–71.