![[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. 59, Issue 1, 54-61, January 2001
The Wolfson Institute for Biomedical Research, University College London, The Cruciform Building, Gower Street, London, United Kingdom
 |
Abstract |
|---|
 Top
 Abstract
 Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
Nitric oxide (NO) functions as a diffusible messenger in the central nervous system and elsewhere, exerting many of it physiological effects by activating soluble guanylyl cyclase, so increasing cellular cGMP levels. Hydrolysis of cyclic nucleotides is achieved by phosphodiesterases (PDEs) but the enzyme isoforms responsible for degrading cGMP in most cells have not been identified. We have devised a method for quantitatively monitoring the rate of breakdown of cGMP within intact cells and have applied it to rat cerebellar cell suspensions previously stimulated with NO. In contrast to previous findings in cultured cerebellar cells, there was no evidence from the use of selective inhibitors that PDE 1 participated importantly in cGMP hydrolysis. Moreover, procedures expected to increase PDE 1 activity by raising cytosolic Ca2+ concentrations (neurotransmitter agonists, Ca2+ ionophore) failed to influence cGMP breakdown. Instead, through the use of inhibitors selective for different PDE families, two isoforms were implicated: a "cGMP-specific" PDE (PDE 5), inhibited by sildenafil and zaprinast, and a "cAMP-specific" PDE (PDE 4), inhibited by low concentrations of rolipram and Ro-20-1724 and by milrinone. An explanation is offered for a participation of PDE 4 based on the high estimated intracellular cGMP concentration (~800 µM) and the low affinity of the enzyme for cGMP. In accordance with predictions, recombinant PDE 4 was shown to hydrolyze high cGMP concentrations in a rolipram-sensitive manner. The widespread use of rolipram to test for a specific involvement of cAMP in cellular phenomena must therefore be questioned.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
Nitric
oxide (NO), formed from L-arginine, serves in many
different tissues as an intercellular signaling molecule (Moncada et
al., 1991
; Garthwaite and Boulton, 1995). Under physiological conditions, the principal transduction pathway used by NO is activation of soluble guanylyl cyclase (sGC), leading to accumulation of cGMP
(Hobbs, 1997). The cellular levels of cGMP are regulated not only by
the rate of synthesis but also by the rate of breakdown by
phosphodiesterase (PDE) enzymes. At least 11 different families of PDE,
having varying affinities for cGMP and cAMP and differing pharmacological properties, are known to exist (Beavo, 1995). However,
with the exception of platelets (a homogenous cell population) the
isoforms hydrolyzing cGMP in cells stimulated by NO are uncertain: most
studies have been conducted on tissue homogenates in which PDEs
normally located in different cell types or cell compartments are mixed
together. Knowledge of the participating isoforms is important for
understanding the dynamics of the signal transduction pathways and for
providing pharmacological tools to probe those pathways. There is also
the potential for the identification of new medicines, as exemplified
by sildenafil for erectile dysfunction (Corbin and Francis, 1999) and
rolipram for inflammatory disorders (Teixeira et al., 1997).
The central nervous system, and the cerebellum in particular, expresses
abundantly the neuronal isoform of NO synthase (nNOS), which is
activated by Ca2+/calmodulin (Abu-Soud et al.,
1994). A Ca2+/calmodulin-dependent PDE, PDE 1, is
also abundant (Yan et al., 1994) and a current hypothesis is that a
rise of intracellular Ca2+ (as a result of
glutamate receptor activation, for example) simultaneously stimulates
nNOS activity and PDE 1 activity. In this way, cGMP accumulation in the
stimulated cells is inhibited but neighboring cells with low cytosolic
Ca2+ are free to respond (Mayer et al., 1992).
This could act as a device for ensuring that NO acts in a paracrine,
rather than autocrine, manner. In principle, such a mechanism would
also permit modulation of the NO-cGMP pathway by other signaling
systems. These could act on target cells to change cGMP PDE activity by
altering intracellular Ca2+, by modifying the
levels of the competing PDE 1 substrate (cAMP) or by changing the
phosphorylation state of the enzyme (Beavo, 1995).
The present study was aimed at identifying the PDE activities responsible for degrading cellular cGMP after an NO stimulus in the cerebellum. To do so, we have developed a method for measuring the rate of cGMP breakdown in living cells and have applied it to freshly isolated cell suspensions from the developing cerebellum. One of the advantages of this preparation for the present purposes is that, despite the cellular heterogeneity, the cGMP response to NO is apparently confined to a single cell type, the astrocytes. The results indicate that PDE 1 is not involved but rather that PDE 5 and, more surprisingly, a "cAMP-specific" PDE (PDE 4) are responsible for cGMP hydrolysis.
| |
Experimental Procedures |
|---|
|
Top
Abstract
Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
Measurement of cGMP and cAMP in Cerebellar Cell Suspensions.
Cell suspensions were prepared from 8-day-old rat cerebella as
described previously (Garthwaite and Garthwaite, 1987). Pooled cells
from 12 to 14 cerebella were divided into populations of 8 million
cells (0.4 ml) and incubated at 37°C in a solution containing 130 mM
NaCl, 3 mM KCl, 1.2 mM MgSO4, 1.2 mM
Na2HPO4, 15 mM Tris·HCl, 1.5 mM CaCl2, and 11 mM glucose, pH 7.4). To
avoid possible complications arising from endogenous NO production, the
NOS inhibitor L-nitroarginine (100 µM) was added at least
1 h before the experiments were begun.
). Subsequently, cGMP levels
were followed over time by withdrawing aliquots of cells and
inactivating them in boiling hypotonic buffer (50 mM Tris, 4 mM EDTA,
pH 7.4). cGMP levels were measured by radioimmunoassay. Extracellular
cGMP accounts for only 1 to 2% of the total generated and was not
deducted from the totals. cAMP was measured using a commercial kit
(Amersham Pharmacia Biotech, Bucks, UK).
The PDE inhibitors were dissolved in DMSO and added at a 1:100 dilution
to give the required concentration, 10 min before addition of DEA/NO.
The solvent itself was without effect on the cGMP response to DEA/NO
(not shown). Where neurotransmitter receptor agonists were used, they
were dissolved in DMSO or equimolar NaOH (as appropriate), and added
immediately before the addition of Hb. Results are given as the mean
cyclic nucleotide levels from independently treated cell
populations ± S.E.M.
Quantification of the Rate of cGMP Degradation.
The decay of
cGMP after application of Hb was described by the integrated
Michaelis-Menten equation: Vpt = Kpln
(Po/Pt) + (Po 
Pt),
where Po is concentration of cGMP
immediately before addition of Hb and Pt
is concentration of cGMP at time t. Vp, Kp, and Po
were found by iteration (Fernley, 1974) using Origin v.4.1 software
(MicroCal Software, Northampton, MA).
cGMP Degradation by Recombinant PDE 4B.
Extracts from BHK
cells transfected with human PDE 4B (Bardelle et al., 1999) were a
generous gift from Dr. J. Staddon (Eisai London Research Laboratories
Ltd., University College, London, UK). Cytosolic extracts (50 µl
containing 0.25 mg of protein in 25 mM HEPES, 5 mM EDTA, 1 mM
dithiothreitol, 10 µl/ml leupeptin, 50 µg/ml aprotinin, 200 µg/ml
benzamidine, 10 µg/ml soybean trypsin inhibitor, and 1 mM
phenylmethylsulfonyl fluoride) were incubated at 37°C with a solution
(5 µl) containing cGMP (0.72-20 mM), MgSO4 (0.5 M), EGTA (1 mM), 1% DMSO, ± rolipram (0.03-100 µM). After various intervals, aliquots were removed and inactivated (as above). Activity was assessed by loss of cGMP.
Materials. 1-H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-l-one (ODQ) and (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid [(1S,3R)-ACPD] were supplied by Tocris Cookson (Bristol, UK). Sildenafil was supplied by the Chemistry Division, Wolfson Institute for Biomedical Research. All other special chemicals were from Sigma-Aldrich (Poole, Dorset, UK).
| |
Results |
|---|
|
Top
Abstract
Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
Measuring PDE Activity in Intact Cells.
The technique has been
described briefly in a previous study designed to extract the kinetics
of sGC in cells (Bellamy et al., 2000). Hence, we begin by providing
fuller details of the procedure before using it to characterize the
relevant PDE activity.
). When sGC activity was
arrested by addition of Hb, cGMP declined progressively (Fig. 1a). At
its simplest, the rate of cGMP degradation
(vd) at a given substrate concentration
(P) should follow Michaelis-Menten-type kinetics according
to the equation: vd = VpP / (Kp + P). The parameters Vp and
Kp are empirical constants describing,
respectively, the apparent maximum PDE activity and the
apparent Michaelis constant. The integrated form of the expression (see Experimental Procedures) should then
describe the fall in cellular cGMP levels over time. As shown in Fig.
1a, the integrated equation provided a good fit to the experimental data. Mean values for the parameters Vp
and Kp derived from such fits were
0.81 ± 0.32 pmol/106 cells/s and 107 ± 50 pmol/106 cells, respectively
(n = 9). The rather large standard errors reflect the
fact that the absolute values can interchange markedly without having
much impact on the net rate of degradation at a given substrate
concentration. More revealing is the ratio
Vp/Kp, which
gives an index of catalytic efficiency, because this ratio cannot vary
much without the rate of degradation being significantly affected.
Accordingly, there was much less interexperiment variation in the
estimations of
Vp/Kp, which
averaged 0.011 ± 0.001 s
1
(n = 9).
|
Pharmacological Identification of Active PDE Isoforms.
Inhibitors selective for different classes of PDE were initially
examined using a screening procedure in which cGMP levels were measured
5 min after removal of NO (by addition of Hb). By this time, cGMP in
control cells had decayed almost to basal levels. An effective
inhibitor of cGMP breakdown, therefore, will cause cGMP to remain
elevated relative to that in control cells. The PDE 1-selective
inhibitor, nicardipine (300 µM; Agullo and Garcia, 1997), the PDE
2-selective inhibitor, erythro-9-(2-hydroxy-3-nonyl)adenine (EHNA; 300 µM; Michie et al., 1996), and the PDE 3-selective
inhibitor cilostamide (100 µM; Manganiello et al., 1995), all reduced
degradation by only a small extent (Table
1). Another PDE 3-selective inhibitor, milrinone (Manganiello et al., 1995; Schudt et al. 1996), was more
effective, causing a graded increase in residual cGMP at concentrations
of 30 µM and above (Fig. 2). The PDE
4-selective inhibitor rolipram (Schudt et al. 1996) also caused a
concentration-dependent increase in residual cGMP levels (Fig. 2). The
maximum effect was similar to that found with milrinone at 300 µM,
but the potency of rolipram was much higher (EC50
~30 nM). A second PDE 4-selective inhibitor, Ro-20-1724 (Schudt et
al. 1996), gave results very similar to those of rolipram, although its
potency was somewhat lower (EC50 ~100 nM; Fig.
2). No additional effects of either of the two drugs were observed at
concentrations as high as 100 µM (not illustrated).
|
|
), was the
single most effective agent tested. As the sildenafil concentration was
raised over the range 3 to 300 µM, degradation was progressively
blocked. At the highest concentration tested (300 µM), the residual
cGMP was approximately 10-fold higher than in control cells (Fig. 2).
The EC50 for sildenafil was about 50 µM. A
second PDE 5-selective inhibitor, zaprinast, also increased residual
cGMP but the effect was apparently submaximal at the highest
concentration tested (300 µM). Judging by the degree of rightward
shift of the concentration-response curve, zaprinast was about 30-fold
less potent than sildenafil (Fig. 2).
As a control, the compounds sildenafil, rolipram, and zaprinast were
examined for possible effects on the level of cGMP in the absence of
DEA/NO. After a 10-min incubation, none caused a significant change.
The values (pmol cGMP/106 cells ± S.E.M.)
were: control cells, 0.41 ± 0.08; sildenafil (300 µM),
0.29 ± 0.03; rolipram (1 µM), 0.35 ± 0.07; and zaprinast (300 µM), 0.60 ± 0.04 (n = 3-6).
The two different maxima exhibited by the PDE 4-selective inhibitors
and sildenafil suggested that (at least) two discrete isoforms of PDE
were operating. This was investigated further by coapplying selected
inhibitors with a supramaximal concentration of rolipram (1 µM). With
a combination of rolipram and Ro-20-1724 (3 µM) or milrinone (300 µM), the residual cGMP was no different from that found with any of
the compounds individually. In contrast, zaprinast (300 µM) and
sildenafil (300 µM) both produced additional inhibition of cGMP
degradation (Fig. 3). Furthermore, the
combination of sildenafil plus rolipram was more effective than
sildenafil alone (Fig. 3), such that 5 min after the Hb addition, cGMP
remained at about 70% of its starting value.
|
1 to
0.006 s1. Similar experiments were carried out
using sildenafil (100 µM) plus rolipram (1 µM). The overall effects
were similar to, but greater than, those seen with zaprinast plus
rolipram (see Bellamy et al., 2000). Analysis of the decay phase
indicated that
Kp/Vp was
reduced 4-fold, from 0.011 to 0.003 s1. Neither
the combination of zaprinast plus rolipram (Fig. 1a) nor sildenafil
plus rolipram (Bellamy et al., 2000) influenced cGMP levels in the
absence of DEA/NO.
Further Tests for Involvement of PDE 1.
The above
pharmacological results indicated that PDE 1 does not participate
noticeably in cGMP hydrolysis in the cerebellar cells, a finding that
conflicts with conclusions based on measurements of PDE activity in
homogenates of cultured cerebellar cells (Agullo and Garcia, 1997). One
explanation could be that, because
Ca2+/calmodulin enhances the activity of PDE 1, it may be necessary to raise Ca2+ to observe the
enzyme in action. This was tested in two ways. First, neurotransmitter
receptor agonists previously shown to raise the cytosolic
Ca2+ concentration in cerebellar astrocytes
(Murphy and Pearce, 1987) were examined for their ability to accelerate
the decline in cGMP. For this purpose, the agonists were added
immediately before the Hb and cGMP was measured 2 min later, when the
level had normally fallen by 50% (Table 1). Neither glutamate itself
nor agonists activating selectively the AMPA/kainate or metabotropic
glutamate receptors [kainate and (1S,3R)-ACPD]
significantly affected the level of residual cGMP. Noradrenaline was
similarly without effect. Moreover, a 1-min pretreatment of the cells
with these receptor agonists had no effect on the subsequent
time-courses of DEA/NO-stimulated cGMP accumulation (data not shown).
). Furthermore, the concentrations of intracellular Ca2+ required
to cause this direct inhibition of sGC are severalfold greater than
those required for maximal activation of PDE 1 (Agullo and Garcia,
1997), indicating that the magnitude of the Ca2+
influx would have been sufficient to reveal any PDE 1-mediated degradation of cGMP. However, conclusions based on measurements at a
single time point could be wrong if, in the presence of increased cellular Ca2+, the kinetics of PDE becomes
aberrant. Hence, to examine the effect of A23187 more rigorously, the
time course of cGMP hydrolysis was followed over 3 min. There was no
alteration in the rate of decay compared with control cells throughout
the period (Fig. 1c).
Possible Role of cAMP.
An explanation for the ability of
inhibitors of PDE 4, normally considered a "cAMP-specific" isoform,
to attenuate cGMP hydrolysis is that, by inhibiting cAMP breakdown,
they might raise the level of cAMP that could, in turn, competitively
inhibit cGMP hydrolysis by a PDE able to act on both cyclic nucleotides
(i.e., PDEs 1, 2, 3, or 10; Beavo, 1995; Soderling et al., 1999). When
measured in the same cells, however, rolipram inhibited cGMP breakdown without altering cAMP levels (Fig. 4). To
examine the possibility further, cells were pretreated for 10 min with
forskolin (an activator of adenylyl cyclase) before being exposed to
DEA/NO and then Hb, as usual. Five minutes later, cAMP was at high
levels but there was little or no effect on cGMP breakdown (Fig. 4).
|
Degradation of cGMP by Recombinant PDE 4B.
To examine directly
the ability of PDE 4 to degrade cGMP, extracts of BHK cells
over-expressing recombinant human PDE 4B were used. cGMP was added at a
concentration estimated to be attained in the cerebellar cells exposed
to NO (~800 µM; see Discussion). The cGMP concentration
declined progressively to the extent that about 85% was degraded after
1 h and this hydrolysis was almost eliminated by 10 µM rolipram
(Fig. 5a). A Lineweaver-Burk analysis (Fig. 5a inset) indicated that the Km value
for cGMP was 4.8 mM and that the apparent
Vmax was 24 nmol cGMP/mg of protein/min. For comparison, the same cell extracts hydrolyze cAMP with a
Km value of 5.5 µM (Bardelle et al. 1999)
and an apparent Vmax (calculated from the
data of Bardelle et al.) of about 20 nmol/mg of protein/min at 20°C
or, assuming a temperature coefficient (Q10) of
2, 68 nmol/mg of protein/min at 37°C. A 3-fold greater maximal
activity of PDE 4 against cAMP compared with cGMP accords with an
earlier study (Beavo, 1988). The IC50 value of
rolipram for inhibition of degradation of 800 µM cGMP over 1 h
was 0.9 µM (Fig. 5b).
|
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
Despite extensive reports in the literature concerning
phosphodiesterases, quantitative analysis of their activity in intact cells does not seem to have been attempted previously. Monitoring the
behavior of the enzymes within their normal environment, however, is
essential for understanding their physiological functioning and the
possible modulation of their activity by cellular signaling pathways.
The cerebellum, used here, has the advantage of low PDE activity, such
that cGMP decay can be followed over several minutes. The fact that PDE
inhibitors did not reduce the PDE catalytic efficiency index
(Vp/Kp) very
dramatically (maximally 4-fold) reflects the fact that the index is so
low in the first place (0.01 s1). For
comparison, it can be estimated that in human platelets, the
uninhibited
Vp/Kp is
about 100-fold larger (1.4 s1) and that the
appropriate inhibitors (sildenafil plus EHNA) reduce the index 200-fold
(Bellamy et al., 2000).
PDEs Degrading cGMP after NO Stimulation of sGC.
The
inhibitors tested in this study have been well characterized in
homogenate assays and against purified PDE subtypes (Table 2). Their use in the cerebellar cells
indicates that two PDEs are at work: one inhibited by rolipram,
Ro-20-1724, and milrinone and the other inhibited by sildenafil and
zaprinast. These properties are those of PDE 4 and PDE 5, respectively.
Consistent with this conclusion are the relative inhibitory potencies:
rolipram>Ro-20-1724>milrinone is expected for PDE 4, and
sildenafil>zaprinast is expected for PDE 5. The PDE families 6,7, and
8 could not easily be examined because of the absence of selective
inhibitors. However, the functional and pharmacological properties of
these enzymes make them unlikely candidates: PDE 6 activity appears
confined to the retina and both PDE 7 and 8 are "cAMP-specific"
PDEs that are insensitive to rolipram or Ro-20-1724. Even in the
presence of maximal concentrations of sildenafil and rolipram, some
cGMP degradation persisted (30% during 5 min). Although this could
reflect the activity of another PDE isoform, it may simply be a
consequence of the mechanism of action of sildenafil and rolipram: the
efficacy of competitive inhibitors such as these will be inherently
limited by the associated rise in cGMP concentration.
|
; De Vente et
al., 1990; Bellamy et al., 2000), which represent about 6% of the
total cell number (Cohen et al., 1979) and are approximately spherical
with a radius of 5 µm (Garthwaite and Balazs, 1981). Thus, it
can be calculated that the 25 pmol of cGMP/106
cells formed in response to DEA/NO in the suspension as a whole (Fig.
1a), corresponds to a concentration of about 800 µM in the astrocytes. The potency of any competitive inhibitor (expressed as
IC50) depends on the substrate concentration
(S) according to the relationship:
IC50 = Ki (1+
S/Km), where
Ki is the inhibitory constant and
Km is the Michaelis constant (Cheng and
Prusoff, 1973). Unless S 
Km,
IC50 values will reflect the potency of an
inhibitor only under particular conditions; the values commonly cited
in homogenate studies are derived using cGMP concentrations of 0.5 to
10 µM. At higher substrate concentrations, the
IC50 value will be higher. From parameters in the
literature (Table 2), the predicted IC50 of
sildenafil for PDE 5 in cells containing 800 µM cGMP is 0.6 to 24 µM, the latter being similar to our estimate of 50 µM. For
zaprinast, the predicted IC50 is 0.14 to 2.3 mM, which is also consistent with its potency in our experiments.
Although PDE 4 is designated "cAMP specific," the isoform can
also metabolize cGMP but with a Km value
well above the concentrations normally used in homogenate assays.
Accordingly, we have found that recombinant human PDE 4B is able to
degrade cGMP at concentrations calculated to exist in NO-stimulated
cells (~800 µM). The initial rate of cGMP hydrolysis by PDE 4B in
the BHK cell extracts was about 45 pmol/mg of protein/s (Fig. 5a),
which compares with a calculated initial rate of rolipram-sensitive
cGMP hydrolysis in the cerebellar
astrocytes1 of 28 pmol/mg of protein/s. Although crude, this comparison supports the
feasibility of PDE 4 functioning to degrade cGMP in cells.
The estimated Km value for PDE 4B of 4.8 mM
(Fig. 5) is high compared with another published value for cGMP
degradation by PDE 4 (Km = 310 µM; Beavo,
1988) or with the cGMP concentrations that inhibit cAMP degradation by
PDE 4 (approximate Ki = 360 µM; Herman et
al., 2000), possibly reflecting differences between PDE 4 isoforms. Collectively, it seems that that the affinity of PDE 4 isoforms for cGMP is in the high micromolar-millimolar range. Hence, at
an intracellular cGMP concentration of 800 µM, PDE 4 would be
expected to hydrolyze cGMP at about half-maximal rate. Addition of PDE
4-selective inhibitors with Ki values
substantially less than the Km value should
then potently inhibit cGMP degradation. For the recombinant PDE 4B, we
found an IC50 value for rolipram of 0.9 µM
(versus 800 µM cGMP). The lower IC50 value in
the cerebellar cells (30 nM) may reflect accumulation of the drug
intracellularly, a different PDE 4 isoform, or intracellular regulatory
mechanisms, such as phosphorylation (McPhee et al., 1999).
The PDE 3-selective inhibitor milrinone can also inhibit PDE 4 (Ki = 11 µM) so inhibition of cGMP
hydrolysis by PDE 4 could clearly be achieved at the concentrations
used in this study. Indeed, the predicted IC50
value of 39 µM for milrinone accords well with our experimental data
(Fig. 2). Thus, this "PDE 3-selective" drug can inhibit breakdown
of cGMP by "cAMP-specific" PDE 4!
Despite initial appearances, therefore, the pharmacological evidence
supports the conclusion that the principal cGMP-degrading enzyme in
cerebellar astrocytes is PDE 5, with an additional contribution by PDE
4. Other evidence suggests that both these isoforms are expressed in
the rat cerebellum, but their precise cellular locations are unclear
(Iwahashi et al., 1996; Kotera et al., 1997). In contrast, PDE 1 seems
to be the major cGMP-degrading enzyme in homogenates of cultured
cerebellar astrocytes (Agullo and Garcia, 1997). Despite exhaustive
tests, we could find no obvious role for this enzyme. The discrepancy
may be attributable to methodological differences or to differences in
PDE expression between cultured and freshly isolated astrocytes. Our
findings indicate that, in these target cells, the downstream effector
mechanism for NO signaling is unlikely to undergo modulation by
Ca2+-mediated changes in cGMP PDE activity.
Whether this applies to other NO targets in the brain remains to be addressed.
From a pharmacological perspective, this study reveals potential
pitfalls associated with applying data on PDE activity in homogenates
to the regulation of cGMP metabolism within intact cells. In
particular, although designated "cAMP specific," PDE 4 could have a
direct role to play in regulating cGMP metabolism in cells that
accumulate the second messenger to high concentrations. In this
respect, it could be argued that the high cGMP concentrations accumulating in the cells over a 2-min exposure to DEA/NO are unphysiological and, hence, that this conclusion is only of academic interest. Examination of Fig. 1a, however, shows that with an exposure
to NO of only 4 s, a not unreasonable period from a physiological perspective (Wood and Garthwaite, 1994), cGMP levels are about 25% of
maximum. This equates to a cGMP concentration in the astrocytes of
around 200 µM, which means (assuming a high micromolar-millimolar Km range for PDE 4) that significant cGMP
hydrolysis would occur through PDE 4. Thus, the phenomenon is of
potential physiological significance. Obviously, cGMP concentrations of
this magnitude are very high compared with those needed to activate
cGMP-dependent protein kinases but cGMP does have other targets. In
particular, increasing evidence indicates that cyclic nucleotide-gated
ion channels are expressed widely in the brain and other organs (Biel et al., 1994; Kingston et al., 1999; Strijbos et al., 1999; Yao et al.,
1999) and, depending on the channel type, cGMP concentrations in the 10 or 100 µM ranges are needed for maximal activity. It is interesting
to speculate that these channels represent targets for cGMP, generated
in response to NO, in cerebellar astrocytes.
If PDE 4 can contribute to cGMP hydrolysis physiologically, another
implication is for the interpretation of results from the many studies
that have used rolipram (or other PDE 4 inhibitors) as probes for an
involvement of cAMP. As we have demonstrated, low concentrations of
rolipram and Ro-20-1724 can have pronounced effects on cGMP
degradation. Indeed, with equal concentrations of cGMP and cAMP, a
competitive PDE 4 inhibitor will preferentially influence the
degradation of cGMP by this enzyme. Furthermore, as exemplified by
milrinone, drugs with a weak effect on degradation of cAMP by PDE 4 can
have a significant effect on cGMP degradation by the enzyme because of
the disparity in affinity (Km) of PDE 4 for
the alternative cyclic nucleotides.
| |
Acknowledgments |
|---|
We thank the members of Eisai's Research Laboratories in Tsukuba (Japan) and London (UK) for the generous gift of recombinant human PDE 4B. We also thank Dr. D. Madge (Chemistry Division, Wolfson Institute for Biomedical Research) for supplying sildenafil.
| |
Footnotes |
|---|
Received March 28, 2000; Accepted September 27, 2000
1 This calculation assumes that, in the cerebellar cell suspension, the initial rate of cGMP degradation is 0.1 pmol/106 cells/s (see Fig. 1), that a third of this rate is rolipram-sensitive (see Fig. 3), and that astrocytes make up 6% of the cells (see text) and contain 20 µg cytosolic protein/106 cells (based on authors' unpublished observation that 106 cells in the cell suspension contain about 50 µg of total protein, of which 40% is cytosolic).
This work was supported by The Wellcome Trust and a Medical Research Council Studentship (to T.C.B.).
Send reprint requests to: Professor John Garthwaite, Wolfson Institute for Biomedical Research, University College London, The Cruciform Building, Gower Street, London, WC1E 6BT (E-mail: john.garthwaite{at}ucl.ac.uk)
| |
Abbreviations |
|---|
NO, nitric oxide; sGC, soluble guanylyl cyclase; PDE, phosphodiesterase; nNOS, neuronal nitric-oxide synthase; DEA, diethylamine; Hb, hemoglobin; ODQ, 1-H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-l-one; (1S,3R)-ACPD, (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid; EHNA, erythro-9-(2-hydroxy-3-nonyl)adenine.
| |
References |
|---|
|
Top
Abstract
Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  B. Roy and J. Garthwaite Nitric oxide activation of guanylyl cyclase in cells revisited PNAS, August 8, 2006; 103(32): 12185 - 12190. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Hausser, I. M. Raman, T. Otis, S. L. Smith, A. Nelson, S. du Lac, Y. Loewenstein, S. Mahon, C. Pennartz, I. Cohen, et al. The Beat Goes On: Spontaneous Firing in Mammalian Neuronal Microcircuits J. Neurosci., October 20, 2004; 24(42): 9215 - 9219. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Jurado, J. Sanchez-Prieto, and M. Torres Differential expression of NO-sensitive guanylyl cyclase subunits during the development of rat cerebellar granule cells: regulation via N-methyl-D-aspartate receptors J. Cell Sci., August 1, 2003; 116(15): 3165 - 3175. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. U. Suvarna and J. M. O'Donnell Hydrolysis of N-Methyl-D-aspartate Receptor-Stimulated cAMP and cGMP by PDE4 and PDE2 Phosphodiesterases in Primary Neuronal Cultures of Rat Cerebral Cortex and Hippocampus J. Pharmacol. Exp. Ther., July 1, 2002; 302(1): 249 - 256. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. C. Bellamy and J. Garthwaite Sub-second Kinetics of the Nitric Oxide Receptor, Soluble Guanylyl Cyclase, in Intact Cerebellar Cells J. Biol. Chem., February 2, 2001; 276(6): 4287 - 4292. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
,
) and the subsequent
degradation of cGMP after addition of 10 µM Hb (
,
) in the
presence (
). Data after addition of Hb are
fitted to the integrated Michaelis-Menten equation (see
Experimental Procedures). For control cells (
P < 0.05 compared with sildenafil alone.
