![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 303, Issue 1, 356-363, October 2002
Department of Pharmacology and Program in Neuroscience, University of Colorado Health Sciences Center, Denver, Colorado
 |
Abstract |
|---|
 Top
 Abstract
 Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Excitatory glutamatergic synapses in the hippocampal CA1 region of rats
are potently inhibited by purines, including adenosine, ATP, and ATP
analogs. Adenosine A1 receptors are known to mediate at
least part of the response to adenine nucleotides, either because adenine nucleotides activate A1 receptors directly, or
activate them secondarily upon the nucleotides' conversion to
adenosine. In the present studies, the inhibitory effects of adenosine,
ATP, the purportedly stable ATP analog
adenosine-5'-O-(3-thio)triphosphate (ATP
S), and
cyclic AMP were examined in mice with a null mutation in the adenosine
A1 receptor gene. ATPS displaced the binding of
A1-selective ligands to intact brain sections and brain
homogenates from adenosine A1 receptor wild-type animals.
In homogenates, but not in intact brain sections, this displacement was
abolished by adenosine deaminase. In hippocampal slices from wild-type
mice, purines abolished synaptic responses, but slices from mice
lacking functional A1 receptors showed no synaptic
modulation by adenosine, ATP, cAMP, or ATPS. In slices from
heterozygous mice the dose-response curve for both adenosine and ATP
was shifted to the right. In all cases, inhibition of synaptic
responses by purines could be blocked by prior treatment with the
competitive adenosine A1 receptor antagonist
8-cyclopentyltheophylline. Taken together, these results show that even
supposedly stable adenine nucleotides are rapidly converted to
adenosine at sites close to the A1 receptor, and that
inhibition of synaptic transmission by purine nucleotides is mediated
exclusively by A1 receptors.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
ATP
released during nerve activity can regulate transmitter release (for
review, see Cunha and Ribeiro, 2000
). Both excitatory and inhibitory
effects have been suggested in peripheral nerves, in ganglia, and in
the brain (von Kugelgen et al., 1993, 1994; Boehm, 1999; O'Kane and
Stone, 2000; Smith et al., 2001). For example, hippocampal synaptic
transmission is inhibited not only by adenosine but also by ATP, cAMP,
and numerous cAMP and ATP analogs (Dunwiddie and Hoffer, 1980; Lee et
al., 1981). The inhibitory effects of adenosine are mediated via the
adenosine A1 receptor (A1R)
(Johansson et al., 2001), but the identity of the receptor(s) by which
nucleotides exert their effects on transmission is more difficult to
ascertain. It is clear that ATP and other adenine nucleotides are very
rapidly broken down in hippocampus, and that adenosine is one of the
products formed (Dunwiddie et al., 1997). Hence, a major question is
whether ATP and other adenine nucleotides act at receptors other than
presynaptic A1Rs to inhibit transmitter release.
Several lines of evidence suggest that nucleotides per se can inhibit
transmission and that conversion to adenosine is not required. First,
many ATP analogs, generally considered metabolically stable, are at
least as potent as ATP in terms of inhibiting synaptic potentials (Lee
et al., 1981; von Kugelgen et al., 1992; Cunha et al., 1998;
Mendoza-Fernandez et al., 2000; Smith et al., 2001). Second, incubation
of brain slices with adenosine deaminase (ADA), which is expected to
convert adenosine formed from the metabolic conversion of nucleotides
into inosine (relatively inactive at A1Rs)
(Fredholm et al., 2001), does not completely inhibit responses to ATP
and ATP analogs, suggesting that there may be a component of the ATP
response that is not mediated via adenosine (Cunha et al., 1998;
Mendoza-Fernandez et al., 2000). Similarly, inhibitors of
5'-nucleotidase, the extracellular enzyme required to convert 5'-AMP to
adenosine, have limited effectiveness in blocking nucleotide responses.
Because of these observations and additional experimental evidence, the
hypothesis has been advanced that there may be a nucleotide receptor
that inhibits neurotransmitter release, is activated directly by ATP
and ATP analogs, and is insensitive to most ATP receptor (P2)
antagonists, but is sensitive to purportedly selective
A1R antagonists such as
8-cyclopentyl-theophylline (CPT) (Shinozuka et al., 1988; Smith et al.,
1997; Mendoza-Fernandez et al., 2000).
We have examined the actions of adenosine and adenine nucleotides in
mice with a null mutation in the A1R gene, which
results in a complete loss of functional A1Rs in
the central nervous system (Johansson et al., 2001). Our experiments
demonstrate that in mice lacking the A1R, the
inhibitory effects of adenosine, cAMP, ATP, and
adenosine-5'-O-(3-thio)triphosphate (ATPS) on synaptic transmission are completely absent, suggesting that all of these actions are mediated via conventional A1Rs.
| |
Materials and Methods |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Null Mutant Mice.
Mice with a null mutation in the
A1R were derived as described previously
(Johansson et al., 2001). Mice that were heterozygous for the null
mutation were bred, and littermates with the
A1R+/+,
A1R
/, and
A1R/ genotypes were
used for the following experiments. Mice were genotyped either by
Southern blotting (Johansson et al., 2001) or by a polymerase chain
reaction-based assay that amplified products of different sizes
from the wild-type and null mutant genes. Phenotypes were confirmed by
ligand binding assays conducted on cortical tissue obtained from the
mice used in the electrophysiological studies.
Membrane Receptor Binding Studies.
The assay method using
[3H]N6-cyclohexyladenosine
([3H]CHA; PerkinElmer Life Sciences,
Boston, MA) to measure A1Rs was similar to that
published previously (Mayfield et al., 1996). Crude membrane preparations were preincubated in 50 mM Tris buffer, pH 7.4, and ADA (2 U/ml; Roche Applied Science, Indianapolis, IN) for 30 min at 37°C.
After washing, membranes were incubated in 50 mM Tris buffer containing
10 mM MgCl2 and [3H]CHA
(0.14-25 nM for saturation curves, 30 nM for single point saturation
analysis, and 10 nM for competition curves) for 2 h at 25°C.
Nonspecific binding was measured in the presence of 2-chloroadenosine (2-CADO; 20 µM; Sigma/RBI, Natick, MA). Competition curves with 2-CADO and ATPS (Sigma-Aldrich, St. Louis, MO) were generated in the
absence and presence of ADA (2 U/ml) during the incubation. Incubations
were terminated by dilution with ice-cold buffer and rapid filtration
through glass fiber filters. Radioactivity was quantitated by liquid
scintillation counting. Saturation and competition data were analyzed
by nonlinear curve fitting using the one-site hyperbola and sigmoid
dose-response (variable slope) equations, respectively (GraphPad
Software, San Diego, CA). Protein was measured using the method of
Bradford (1976).
Receptor Autoradiography. Sagittal sections of brains from A1R+/+ mice (14 µm in thickness) were cut 1.2 to 1.8 mm from the midline and put on collagen-coated microscope slides. [3H]1,3-Dipropyl-8-cyclopentyl xanthine (DPCPX; 103 Ci/mmol; PerkinElmer Life Sciences) was used as the radioligand. Slides were preincubated in 170 mM Tris-HCl, pH 7.4, containing 1 mM EDTA and 4 U/ml ADA at room temperature for 30 min. Thereafter, the slides were prewashed twice for 10 min in Tris-HCl buffer containing 1 mM MgCl2 at room temperature. Then they were incubated in the same buffer with 0.5 nM [3H]DPCPX and 4 U/ml ADA and varying concentrations of ATP analogs for 2 h at room temperature. Nonspecific binding was studied in the presence of 20 µM R-N6-phenylisopropyl adenosine (R-PIA; Sigma/RBI). The incubation was stopped by washing twice in ice-cold Tris-HCl buffer for 5 min followed by three quick dips in ice-cold distilled water. After the incubation all slides were dried with a stream of air in a refrigerator overnight, and together with calibrated standards, apposed to film for the indicated times. Exposure time was 2 to 3 weeks. The autoradiograms were analyzed with an M5 Imaging Device (Imaging Research, St. Catharine's, ON, Canada). Optical densities were converted to binding density (femtomoles per milligram of gray matter) using the plastic standards and the specific activity of the radioligands. Results and dose-response curves were analyzed with GraphPad Prism.
Purine Metabolism.
Cerebral cortical and hippocampal tissue
from A1R+/+ and
A1R/ mice was
homogenized in 5 ml of Tris-buffered sucrose. After removal of nuclei
and debris by centrifugation, ATP was added to a final concentration of
30 µM. Using high-performance liquid chromatography, the levels of
ATP, ADP, AMP, adenosine, and inosine were determined at 5-min
intervals during a 30-min incubation at 30°C.
Electrophysiology.
Hippocampal slices (400 µm) were
prepared from 4- to 24-week-old mice and studied using conventional
extracellular and whole-cell recording techniques (Dunwiddie and Diao,
1994; Poelchen et al., 2000). Brain slices were maintained at a
constant temperature of 31-33°C in a submersion chamber and
constantly superfused (2 ml/min) with gassed (95%
O2, 5% CO2) artificial
cerebrospinal fluid containing 126 mM NaCl, 3 mM KCl, 1.5 mM
MgCl2, 2.4 mM CaCl2, 1.2 mM
NaH2PO4, 11 mM glucose, and
26 mM NaHCO3 (Sigma-Aldrich). For recordings of
field excitatory postsynaptic potential (fEPSP) responses, the
recording electrode was placed in stratum radiatum of the CA1 region
and the stimulation electrode in stratum radiatum near the border of
the CA1 and CA2 regions; stimuli were delivered at 15-s intervals.
EPSPs were recorded from CA1 neurons using the whole-cell patch-clamp
technique. Patch recording electrodes were filled with a solution
containing 125 mM potassium-gluconate, 5 mM KCl, 10 mM HEPES, 0.1 mM
CaCl2, 1 mM potassium-EGTA, 2 mM MgCl2, 2 mM magnesium-ATP, and 0.2 mM Tris-GTP
(Sigma-Aldrich). Series resistances ranged from 10 to 41 M
(average
30 ± 1.5 M).
S (Sigma-Aldrich), or 8-cyclopentyltheophylline (CPT;
Sigma/RBI) were prepared at 50 to 200 times the desired final
concentration of drug, and a calibrated syringe pump (Razel, Stamford,
CT) was used to add the drugs directly to the superfusion system.
Concentration-response curves were made by superfusing slices with
incrementally higher concentrations of drugs until a near-maximal
inhibition of the EPSP was observed, because in no case was there any
evidence of desensitization of the response. EC50
values and Hill slopes for the concentration-response curves were
determined using the InPlot program (GraphPad Software), with the
Emin constrained to 0 and the
Emax to 100%.
| |
Results |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
ATP and ATP Analogs Displace Binding of Selective A1R
Ligands.
In the present experiments, agonist
([3H]CHA) binding was measured in the cerebral
cortices of the same mice used in the electrophysiological experiments.
As with an antagonist radioligand ([3H]DPCPX;
Johansson et al., 2001), the Bmax in
the A1R/ mice was half
of that in the A1R+/+ mice
(52%), and there was virtually no specific binding in the A1R/ animals (4.5% of
A1R+/+ mice; Table
1). Thus, these experiments confirm that
the null mutation results in a complete loss of
A1Rs and that neither
[3H]DPCPX nor [3H]CHA
bind with high-affinity binding to other
non-A1R-related binding sites in brain.
|
S, was able to displace their binding. In
sections of hippocampus, cerebellum, and cortex from
A1R+/+ mice, ATPS was a
potent inhibitor (IC50 values of 5 ± 1.4, 3 ± 1.5, and 2 ± 1.4 µM, respectively), even in the
presence of 4 U/ml ADA (Fig. 1).
Analogous data were obtained using 5'-adenylimidodiphosphate (data not
shown). These results indicate that these ATP analogs can interact with
A1Rs and that this is not due to the ATP analogs being contaminated by adenosine.
|
S to displace [3H]CHA from high-affinity binding sites.
Because of limited tissue availability, these experiments were
conducted in membranes prepared from rat cortex; previous studies have
shown that the rat and mouse receptors have very similar
pharmacological properties (Maemoto et al., 1997). When
A1R+/+ mouse and
Sprague-Dawley rat cortical membranes were assayed with
[3H]CHA in parallel, they showed nearly
identical Kd (1.12 ± 0.29 versus
0.97 ± 0.12 nM) and Bmax (724 ± 28.5 versus 726 ± 7.99 fmol/mg protein) values in mice and in
rats, respectively. All membranes were pretreated with ADA to eliminate
endogenous adenosine, and subsequently the competition curves were
generated either in the presence or absence of ADA. The ability of
2-CADO to displace [3H]CHA binding was
unaffected by the presence or absence of ADA in the assay (Fig.
2A). In the absence of ADA, ATPS
displaced [3H]CHA significantly at
concentrations as low as 10 µM, but in the presence of ADA its
apparent potency was markedly reduced (Fig. 2B). Interestingly, the
potency of ATPS to interact with A1 receptors
in the brain section in the presence of ADA was as high as that in the
homogenate in the absence of ADA. Together, these results suggest that
even this "stable" adenine nucleotide is converted to adenosine,
which then interacts with the A1Rs. However, in a
preparation with a more intact cellular structure, such as a brain
section or brain slice, ADA has a limited capacity to eliminate
adenosine formed locally by ATP hydrolysis, perhaps due to limited
access, and the effect of an adenine nucleotide is profound even in the
presence of ADA. These results thus agree with initial observations by
Williams and Braunwalder (1986) and the more recent suggestion by Cunha
et al. (1998) that in an intact hippocampal slice, adenine nucleotides
can be hydrolyzed to adenosine in a local compartment that is not in
equilibrium with the incubation medium. This is particularly relevant,
because such preparations are used routinely in electrophysiological
studies of presynaptic regulation and synaptic function.
|
Electrophysiological Actions of Adenosine, ATP, and ATPS Are
Eliminated in A1R Knockout Mice.
The inhibitory
effects of adenosine on synaptically evoked fEPSP responses were
examined in A1R+/+,
A1R/, and
A1R/ mice. As we have
reported previously (Johansson et al., 2001), adenosine had an
inhibitory effect on fEPSPs in hippocampal slices from
A1R+/+ and
A1R/ mice but had no
effect on responses in
A1R/ slices. The time
course of inhibition was similar in
A1R+/+ and
A1R/ slices (Fig.
3A). The EC50 for
inhibition of the fEPSP response by adenosine was significantly greater
in the A1R/ slices
(72 ± 3.0 µM) than in the
A1R+/+ slices (40 ± 1.5 µM, p < 0.0001; Table
2). There was no difference in the
Emax values between the
A1R+/+ and
A1R/ slices, which
suggests that there are spare receptors for this inhibitory response in
the wild-type mice. As reported for rat hippocampal slices, whenever
adenosine depressed the fEPSP response there was a reduction in both
the peak amplitude and the slope of the falling phase of the fEPSP, but
there was no change in the presynaptic fiber volley.
|
|
/ slices, even
though the maximum depression required slightly more time in the
A1R/ slices (Fig. 3B).
Similar to adenosine, the EC50 value for ATP was
significantly higher in
A1R/ than
A1R+/+ slices (Table 2). In
addition, the EC50 value for ATP was higher than
that for adenosine (Table 2). ATP had no effect on synaptic transmission in A1R/
slices (Fig. 3B).
Because ectonucleotidases are known to convert ATP to adenosine in
brain (Zimmermann, 1996), and rapid conversion has been demonstrated in
the rat hippocampus (Dunwiddie et al., 1997), the lack of ATP effects
in the A1R/ slices
might be attributable to rapid breakdown of ATP. For this reason, many
studies of nucleotide receptors focus on metabolically stable ATP
analogs such as ATPS, which has been shown to be much more stable
than ATP when coincubated with hippocampal slices or synaptosomes
(Cunha et al., 1998). Therefore, we characterized the actions of
ATPS on fEPSP responses, to determine whether there were nucleotide
actions that were not observed with application of ATP because of its
rapid conversion to adenosine. When slices from
A1R+/+ mice were superfused
with 100 µM ATPS, the fEPSP response was inhibited in a manner
identical to that observed with ATP or adenosine (Fig. 3C). In the
A1R/ slices, similar to
ATP and adenosine, the response to ATPS was reduced. Most
significantly, like ATP and adenosine, 100 µM ATPS had no
detectable effect on fEPSP responses in slices from
A1R/ mice (Fig. 3C).
The comparable time course seen with all three compounds (Fig. 3)
suggests that if the purine nucleotides are being converted to
adenosine within the slice, this must be occurring relatively rapidly
with respect to the time required for the adenosine/ATP to diffuse into
the slice.
Complete concentration-response curves were obtained for adenosine and
ATP. In both the A1R+/+ and
A1R/ mouse hippocampal
slices, high concentrations of both adenosine (Fig.
4A) and ATP (Fig. 4B) were able to
completely inhibit the fEPSP response. However, in the
A1R/ slices, neither
adenosine nor ATP had a significant effect on the response waveform at
concentrations up to 200 µM (Fig. 4).
|
;
Lee et al., 1981; Dunwiddie et al., 1997). These effects have generally
been ascribed to metabolism of these nucleotides to adenosine by
ectophosphodiesterases and ectonucleotidases, because these actions are
antagonized by adenosine receptor antagonists (Dunwiddie et al., 1997).
However, there is considerable variation between species with respect
to some of the nucleotidases, and the mouse in particular shows
considerably less 5'-nucleotidase activity in the CA1 region of the
hippocampus than does the rat (Lee et al., 1986). Therefore, the effect
of cAMP was also determined in hippocampal slices from all three types
of mice. As with ATP and ATPS, superfusion of slices with cAMP
inhibited the fEPSP responses. Complete concentration-response curves
were not obtained for cAMP, but the magnitude of the inhibition produced by 50 µM cAMP was quite comparable with the inhibition induced by 50 µM adenosine, ATP, or ATPS in both the
A1R+/+ and
A1R/ hippocampal slices
(Fig. 5A). As with adenosine and the
other adenine nucleotides, cAMP had no detectable effect on responses evoked in the A1R/
slices (Fig. 5A).
|
/ cortex) or the
rate of AMP accumulation (data not shown). Additionally, there was no
difference between the genotypes in the rate of adenosine elimination
(adenosine levels decreased from 3.7 ± 0.2 to 2.3 ± 0.4 µM in A1R+/+ cortex and
from 3.5 ± 0.5 to 2.2 ± 0.3 µM in
A1R/ cortex during 15 min) or inosine formation (data not shown). Furthermore, results were
similar in hippocampus and in cortex (data not shown).
Finally, to compare between species, we measured the effects of 100 µM ATPS on the fEPSP recorded from wild-type mouse versus rat
hippocampal slices. There was no difference in the amount of inhibition
observed in slices obtained from the two species (mouse, 
79.4 ± 6.0%; rat, 89.1 ± 1.4%; n = 6 for each; N.S.)
The results obtained with the
A1R/ slices suggest
that the responses to ATPS are mediated by
A1Rs. This was confirmed pharmacologically by
determining the sensitivity of these responses to CPT, a selective A1R antagonist. As has been reported previously
in rat, responses in both the
A1R+/+ and
A1R/ slices to 50 µM
adenosine, ATP, ATPS (Mendoza-Fernandez et al., 2000), and cAMP were
completely antagonized by prior superfusion with 100 nM CPT (Fig. 5B).
A previous report suggesting that nucleotide effects on hippocampal
synaptic transmission might be mediated via other purine receptors
(Mendoza-Fernandez et al., 2000) used intracellular recording of EPSPs
to characterize modulation of transmission, rather than measurements of
synaptic field potentials. Although fEPSPs and intracellularly recorded
EPSPs are generally affected in the same ways by pharmacological
agents, this is not always the case (Dunwiddie et al., 1980; Robinson
and Deadwyler, 1981; Valentino et al., 1982), so parallel experiments
were conducted using whole-cell patch recordings from mouse hippocampal
slices to confirm the previous observations. As was observed with
extracellular field potentials, in these experiments both ATP and
ATPS inhibited the intracellularly recorded EPSPs, although the
intracellularly recorded responses seemed to be slightly more sensitive
in this regard (Fig. 6). Moreover, as
seen in the field potentials, the effects of both nucleotides are
essentially lost in
A1R/ slices, clearly
implicating the involvement of A1Rs in this
response.
|
),
and we have previously reported that adenosine induces an outward
(i.e., hyperpolarizing) conductance in pyramidal neurons from
A1R+/+ mice that is absent
in A1R/ mice (Johansson
et al., 2001). In the present studies, we have observed that ATP
induced a small but statistically significant hyperpolarizing effect on
the resting membrane potential of CA1 neurons in
A1R+/+ mice (hyperpolarized
by 4.2 ± 1.2 mV; p < 0.05, n = 4) that was not observed in
A1R/ mice. In contrast,
ATPS depolarized A1R+/+
CA1 neurons (depolarized by 7.5 ± 1.5 mV; p < 0.05, n = 3), and this effect seemed to be present in
the A1R/ mice
(depolarized by 8.9 ± 4.8 mV; N.S., n = 3).
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Using brain slices and tissue homogenates from mice bred to be
+/+, /, or / for the A1R, we clarify
several issues that have remained difficult to resolve with
pharmacologically based experiments. We show that the two commonly used
radioligands, the A1R agonist
[3H]CHA, and the A1R
antagonist [3H]DPCPX, are selective for
A1Rs in concentrations commonly used in binding
studies. Hence, previous suggestions that there are binding sites for
[3H]DPCPX at some as yet poorly defined
nucleotide receptor (Smith et al., 1997) were not supported. In
addition, we confirmed that ATP and some purportedly stable ATP analogs
do compete for A1R ligands at their binding sites
and have similar electrophysiological effects as
A1R agonists on hippocampal slices from all three
types of mice.
Less is known about the effects of adenosine in the mouse compared with
the rat. Nevertheless, presynaptic as well as postsynaptic effects of
adenosine observed in the rat are recapitulated in the mouse (Luscher
et al., 1997; Greif et al., 2000; Jarolimek et al., 2000). Herein, we
report that fEPSP responses evoked by Schaffer collateral/commissural
stimulation are potently inhibited by adenosine, although the
EC50 for this response was higher than we
observed previously in the rat (Dunwiddie et al., 1986; Brundege et
al., 1997). Given the nearly identical
Bmax and
Kd for A1Rs between the two species (Maemoto et al., 1997), this difference is most
likely due to different coupling between the receptor and modulation of
transmitter release, although differences in adenosine uptake could
also be involved. In
A1R/ mice, which
express approximately 50% of the A1Rs seen in
the wild-type mice, there was a significant increase in the
EC50 for the inhibition of the fEPSP by
adenosine, although there was no change in the maximal inhibition.
Similar results (i.e., a change in EC50 but not
in Emax) have been reported in the rat
when the number of A1Rs is reduced (Sebastiao et
al., 2000). These data are consistent with the conclusion that there
are spare receptors for the inhibition of synaptic potentials and that
the reduction in the number of receptors in the
A1R/ animals results in
a synapse where the same maximal inhibition can be achieved but
requires occupation of a greater fraction of the receptors in the
A1R/ versus the
A1R+/+ animals. The ligand
binding studies clearly suggest no difference in receptor affinity for
agonist between the A1R+/+
and the A1R/ animals,
so the changes observed are most likely related to changes in the
number of receptors.
Insofar as the responses to ATP and ATPS are concerned, there are
three basic possibilities: 1) nucleotides bind to and directly activate
A1Rs; 2) ectonucleotidases metabolize
extracellular nucleotides to adenosine, which then activates
A1Rs; or 3) nucleotides activate novel ATP
receptors antagonized by classical A1R
antagonists such as CPT. Regarding the first possibility, the present
experiments show ATPS to be a very weak displacer of
[3H]CHA binding. This agrees with results from
other groups for both the A1R (Schwabe and Trost,
1980; Ragazzi et al., 1991) as well as the A2A
receptor (Pirotton and Boeynaems, 1993). In contrast, the second
possibility, metabolism of nucleotides by ectonucleotidases, is
supported both by our studies as well as those of others. For example,
in brain membranes sufficient adenosine can be formed from
stable ATP analogs to interfere with the binding of
high-affinity ligands (Figs. 1 and 2; Pirotton and Boeynaems, 1993). In
homogenates of rat brain the displacement by ATPS was virtually
eliminated by incubation with the enzyme adenosine deaminase. This
suggests that these ATP analogs interact with A1
receptors because they are converted to adenosine. It may be argued
that the addition of exogenous ATP could in some way alter the
breakdown of endogenous adenosine and that this would explain our
results. However, this would not explain the complete absence of any
response in the A1R/.
This cannot be addressed readily in the intact slice, but when using
brain homogenates the rate of adenosine elimination in the presence of
ATP was not different between wild-type and knockout tissues. The rate
of ATP hydrolysis was also indistinguishable.
The third possibility, nucleotides activate novel ATP receptors, is
disproved by the finding that the inhibition of fEPSP responses by
adenosine, ATP, ATPS, and cAMP was reduced in the A1R/ and completely
lost in the A1R/
slices. The most straightforward conclusion based upon the present results is that in the hippocampus these nucleotides inhibit excitatory neurotransmission via A1Rs, and most likely
following their conversion to adenosine by ectonucleotidases.
Previously, Cunha et al. (1998) suggested that nucleotide hydrolysis
may occur at a site quite close to the adenosine receptors themselves.
Our data showing that ADA does not prevent adenine nucleotides from
interacting with A1 receptors in brain sections (even though ADA prevents binding in homogenates) is entirely compatible with localized generation of adenosine from adenine nucleotides. The observation that both ectonucleotidases and
A1Rs are localized to caveolae in non-neural
tissues (Kittel and Bacsy, 1994; Lasley et al., 2000) suggests there
may be mechanisms by which these proteins associate in similar domains
in cellular membranes. Thus, using the terminology of Cunha et al.
(1998), adenosine that is produced by the nucleotidases is
"channeled" to the adenosine receptors. Such localized adenosine
production accounts for the fact that significant adenosine-mediated
inhibition can be achieved even when there is very limited conversion
of stable nucleotides to adenosine, as determined using biochemical methods (Cunha et al., 1998). Thus, an active nucleotidase might generate substantial local concentrations of adenosine from an adenine
nucleotide, whereas converting relatively little of the nucleotide to
adenosine in a quantitative sense.
Even though ATP and ATPS had very similar effects on presynaptic
release of transmitter, this was not the case at postsynaptic A1Rs where ATP hyperpolarized and ATPS
depolarized the membrane potential. One explanation for this
discrepancy between the presynaptic and postsynaptic
A1Rs could be that ectonucleotidases are located in proximity to the presynaptic but not postsynaptic receptors, and
thus the local concentration of adenosine at the postsynaptic site is
insufficient to activate the A1Rs. Alternatively,
the depolarizing response to ATPS may simply have occluded a
hyperpolarizing response to this nucleotide.
The findings of Mendoza-Fernandez et al. (2000) are somewhat difficult
to reconcile with the conclusions of the present study. In particular,
their observation that ATP responses were substantially more sensitive
to pertussis toxin than adenosine responses suggests there may be
additional complexities involved in the actions of ATP. Colocalization
and coimmunoprecipitation of A1R with
P2Y1 receptors after their coexpression in human
embryonic kidney cells was recently reported (Yoshioka et al., 2001).
Further experiments are needed to determine whether functional
P2Y/A1 receptors exist in hippocampus and whether
they are peculiarly sensitive to pertussis toxin. Another possibility
is that the nucleotidase is in some way dependent on a G protein
pathway to be fully competent, and additional more or less far-fetched
explanations can be devised. We favor the view that adenosine in the
extracellular buffer reaches a somewhat different population of
A1Rs than adenosine channeled to the receptors
from ATP or its analogs. Given that there are many spare receptors,
equivalent responses may be produced via slightly different entities
that may differ in their ability to be reached by pertussis toxin.
The apparent Hill slopes for the adenosine and ATP
concentration-response curves were significantly different. Although
this might be taken as evidence that different receptors mediate the two responses, it is perhaps not surprising because the concentrations quantified are bath, not tissue concentrations of drug. Previous work
from our laboratory has demonstrated that upon superfusion with an
EC50 concentration of adenosine (approximately 25 µM in rat), approximately 97% of the adenosine in the extracellular buffer is taken up by cells or metabolized before it reaches receptor sites (Dunwiddie and Diao, 1994). If ATP must be converted to adenosine
to be effective at A1Rs then this concentration
additionally depends on its conversion by nucleotidases. Thus, one
would not necessarily expect the concentration of adenosine or ATP in
the vicinity of the receptors to be linearly related to the bath
concentration, because both metabolism and transport are saturable processes.
In summary, the results of the present experiments suggest that bath
superfusion with adenine nucleotides, including ATP and ATPS, leads
to the activation of hippocampal adenosine A1Rs. Ligand binding studies provide direct evidence that these nucleotides themselves have low affinity for A1Rs, so it is
most probable that they activate A1Rs after
conversion to adenosine by ectonucleotidases, although other
possibilities (e.g., activation after conversion to AMP) cannot be
ruled out. In general, ATP responses have been controversial and
difficult to study in the hippocampus, due to some inadequacies of
available pharmacological tools and potential contamination with
adenosine responses. The A1R null mutant mouse avoids a number of these issues, which are difficult to resolve in
wild-type animals, and serves as a useful model to study and clarify
purinergic receptor function.
| |
Acknowledgments |
|---|
During the final stages of preparation of this manuscript Tom Dunwiddie suffered a fatal climbing accident. This article is dedicated to his memory.
| |
Footnotes |
|---|
Accepted for publication July 8, 2002.
Received for publication April 4, 2002.
This study was supported by National Institutes of Health Grant 29173, the Department of Veterans Affairs Medical Center; a visiting professorship from the VW-foundation (Hannover), the Swedish Medical Research Council (project numbers 2553 and 12587), and the Tore Nilsson, Bergwall, Eriksson, Thuring and Wiberg Foundations.
DOI: 10.1124/jpet.102.036731
Address correspondence to: Susan A. Masino, Neuroscience Program B138, University of Colorado Health Sciences Center, 4200 E. 9th Ave., Denver, CO 80262. E-mail: susan.masino{at}uchsc.edu
| |
Abbreviations |
|---|
A1R, adenosine A1
receptor;
ADA, adenosine deaminase;
CPT, 8-cyclopentyl-theophylline;
ATPS, adenosine-5'-O-(3-thio)triphosphate;
CHA, N6-cyclohexyladenosine;
2-CADO, 2-chloroadenosine;
DPCPX, 1,3-dipropyl-8-cyclopentyl xanthine;
R-PIA, R-N6-phenylisopropyl
adenosine;
fEPSP, field excitatory postsynaptic potential;
EPSP, excitatory postsynaptic potential.
| |
References |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
(o).
J Neurophysiol
83:
1010-1018)-[3H]N6-phenylisopropyladenosine.
Naunyn-Schmiedeberg's Arch Pharmacol
313:
179-187[CrossRef][Medline].This article has been cited by other articles:
|
|
 |
|
  J. Jones, E. A. Stubblefield, T. A. Benke, and K. J. Staley Desynchronization of Glutamate Release Prolongs Synchronous CA3 Network Activity J Neurophysiol, May 1, 2007; 97(5): 3812 - 3818. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Burnstock Physiology and Pathophysiology of Purinergic Neurotransmission Physiol Rev, April 1, 2007; 87(2): 659 - 797. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. K. Filippov, R. C. Y. Choi, J. Simon, E. A. Barnard, and D. A. Brown Activation of P2Y1 Nucleotide Receptors Induces Inhibition of the M-Type K+ Current in Rat Hippocampal Pyramidal Neurons J. Neurosci., September 6, 2006; 26(36): 9340 - 9348. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. Abbracchio, G. Burnstock, J.-M. Boeynaems, E. A. Barnard, J. L. Boyer, C. Kennedy, G. E. Knight, M. Fumagalli, C. Gachet, K. A. Jacobson, et al. International Union of Pharmacology LVIII: Update on the P2Y G Protein-Coupled Nucleotide Receptors: From Molecular Mechanisms and Pathophysiology to Therapy Pharmacol. Rev., September 1, 2006; 58(3): 281 - 341. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Rodrigues, T. Almeida, P. J. Richardson, C. R. Oliveira, and R. A. Cunha Dual Presynaptic Control by ATP of Glutamate Release via Facilitatory P2X1, P2X2/3, and P2X3 and Inhibitory P2Y1, P2Y2, and/or P2Y4 Receptors in the Rat Hippocampus J. Neurosci., July 6, 2005; 25(27): 6286 - 6295. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kawamura, C. Gachet, K. Inoue, and F. Kato Direct Excitation of Inhibitory Interneurons by Extracellular ATP Mediated by P2Y1 Receptors in the Hippocampal Slice J. Neurosci., December 1, 2004; 24(48): 10835 - 10845. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Ortinau, B. Laube, and H. Zimmermann ATP Inhibits NMDA Receptors after Heterologous Expression and in Cultured Hippocampal Neurons and Attenuates NMDA-Mediated Neurotoxicity J. Neurosci., June 15, 2003; 23(12): 4996 - 5003. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
