Download Stimulation of Soluble Guanylate Cyclase by an Acetylcholine

531
Stimulation of Soluble Guanylate Cyclase by an
Acetylcholine-Induced Endothelium-Derived Factor
from Rabbit and Canine Arteries
Ulrich Forstermann, Alexander Miilsch, Eycke Bohme, and Rudi Busse
From the Department of Pharmacology, University of Freiburg, Freiburg, F.R.C., the Department of Pharmacology, University of
Heidelberg, Heidelberg, F.R.C., and the Department of Applied Physiology, University of Freiburg, Freiburg, F.R.G.
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
SUMMARY. The present study was designed to investigate the hypothesis that, during acetylcholine-induced endothelium-dependent relaxation, a factors) is released from endothelial cells
which directly activates soluble guanylate cyclase. We attempted to determine what similarities
or differences existed between this factor and endothelium-derived relaxing factor. The study was
performed on segments of rabbit aorta and canine femoral artery. Purified soluble guanylate
cyclase was injected into the lumen of these vascular segments, together with its substrate, for
intraluminal incubation of the enzyme. In endothelium-intact vascular segments, the activity of
guanylate cyclase was enhanced over control values obtained by incubation in test tubes. The
stimulation was further increased by acetylcholine in concentrations which caused relaxation of
the vascular segments. The stimulating principle could not be transferred from the vessel lumen
to an external solution of guanylate cyclase, indicating a short life-time. Removal of the endothelium prevented formation and release of the guanylate cyclase stimulating factors). Atropine,
mepacrine, or nordihydroguaiaretic acid, which inhibit acetylcholine-induced endothelium-dependent relaxations, also inhibited acetylcholine-induced endothelium-mediated activation of
guanylate cyclase. The results support the hypothesis that acetylcholine-induced endotheliumderived relaxing factor increases cyclic guanosine monophosphate levels of vascular smooth
muscle by a stimulation of soluble guanylate cyclase. {Circ Res 58: 531-538, 1986)
RELAXATION of several isolated blood vessel preparations by acetylcholine (ACh) and numerous other
agents is dependent upon the presence of an intact
endothelium (Furchgott and Zawadzki, 1980; Chand
and Alrura, 1981; Furchgott et al., 1981; DeMey et
al., 1982; Furchgott, 1983, 1984; Furchgott et al.,
1984). Such agents induce the formation and release
of an unknown factor (or factors) which relaxes the
subjacent smooth muscle. It has been shown that
the ACh-induced endothelium-derived relaxing factor (EDRF) is a humoral agent with a short half life
in the range of seconds (Griffith et al., 1984; Forstermann et al., 1984). The chemical structure of
EDRF and its mechanism of action at the molecular
level are still unknown.
In blood vessel preparations with intact endothelium, relaxations elicited by muscarinic agonists, and
some other agents were found to be associated with
increased levels of cyclic guanosine monophosphate
(cGMP) in smooth muscle cells. In addition, in vessels with the endothelium removed or damaged, this
increase in cGMP levels was reduced or abolished
(Holzmann, 1982; Diamond and Chu, 1983; Furchgott and Jothianandan, 1983; Rapoport and Murad,
1983a, 1983b; Furchgott et al., 1984; Ignarro et al.,
1984; Miller et al., 1984; Rapoport et al., 1984). It
has been proposed that the increase of smooth muscle cGMP levels by endothelium-dependent vasodilators is mediated through activation of guanylate
cyclase (GC) (Furchgott et al., 1981; Holzmann,
1982; Rapoport and Murad, 1983a; Ignarro et al.,
1984).
Relaxation of a variety of blood vessel preparations by nitric oxide (NO) containing vasodilators
like organic nitrates and sodium nitroprusside (SNP)
is also associated with an increase of smooth muscle
cell cGMP levels (for review, see Rapoport and
Murad, 1983b; Ignarro and Kadowitz, 1985).
These vasodilators, or their metabolites, activate
soluble GC by interacting directly with the enzyme
(Mittal and Murad, 1982; Craven and DeRubertis,
1983; Bohme et al., 1984; Ignarro and Kadowitz,
1985).
These findings raise the possibility that EDRF may
also be a direct activator of soluble GC. We report
here that endothelial cells generate a factors) that
activates soluble guanylate cyclase, probably in a
direct manner. The formation and/or release of this
factor(s) is stimulated by ACh. ACh-induced soluble
GC-stimulating factors) and EDRF share several
common properties with respect to stability, formation, and/or release.
Methods
Preparation and Perfusion of Arterial Segments
Rabbits (2.5-3.5 kg) were killed by decapitation. The
descending thoracic aorta and a femoral artery were dis-
532
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
sected out. Mongrel dogs (25-35 kg) were anesthetized
with sodium pentobarbital, and the femoral artery and
larger side branches were excised. Arterial segments of a
length of 3 cm (rabbit aorta and canine femoral artery) or
1 cm (rabbit femoral arteries and side branches of canine
femoral arteries) were prepared free of connective tissue.
Any damage of the endothelium was avoided during the
preparation of the vascular segments. In some experiments, the endothelium was removed by gently rubbing
the intimal surface with a rough steel cannula or by
inverting the segments and carefully removing the endothelium with a razor blade. Both procedures proved to be
equieffective as indicated by the complete failure of ACh
to induce relaxation and confirmed by light microscopy
after silver staining (Poole et al., 1958). Contractile responses to norepinephrine (NE) were not affected by
either rubbing procedure.
After ligation of the side branches, each segment was
cannulated with two L-shaped stainless steel cannulas and
fixed with ligatures. The segment was stretched to its
approximate in situ length, and the vertical parts of the
cannulas were fixed in a plastic frame at this distance.
This device containing the segment was then placed in a
special organ bath so that the top ends of the cannulas
protruded beyond the surface (cf. Busse et al., 1983,1984).
The bath contained oxygenated Tyrode's solution (Poj >
460 mm Hg, pH 7.4, 37°C) as the extraluminal medium.
The proximal end of the inflow cannula was connected to
a perfusion pump and the segment was perfused (0.66
ml/min) with modified Tyrode's solution (Na+, 144.0 ITIM;
Ca++, 1.6 mM; Mg++, 3.0 mM; Cl", 145.0 ITIM; HCO3", 11.9
ITIM; H2PO«~, 0.35 ITIM; CaNa2EDTA, 0.025 nw; glucose,
11.2 mM; glutathione, 0.5 ITIM; bovine 7-globulin, 0.1 mg/
ml; P02 = 140 mm Hg). The pH of the solution was
adjusted to 7.4 at 37°C by gassing with 5% CO^ 20% O2,
and 75% N2. Glutathione and bovine 7-globulin are essential for the determination of the soluble GC activity
described below. Glutathione protects soluble GC from
oxidative inactivation. Bovine 7-globulin prevents adsorption of the purified GC to different surfaces, e.g., test
tubes.
To assess mechanical responsiveness, vascular segments
were perfused with modified Tyrode's solution containing
NE (0.1 ^M) until a contraction plateau had been reached.
Then ACh (0.01-2 MM) was added to induce relaxation.
To test the effects of drugs in modified Tyrode's solution,
vascular segments were perfused for 60 minutes with
mepacrine (30 ^M), nordihydroguaiaretic acid (20 HM), or
indomethacin (20 JZM), respectively, before exposure to NE
and ACh. In some experiments sodium nitroprusside
(SNP, 1 nM to 1 J/M) was added instead of ACh.
Measurement of Smooth Muscle Tone
The experiments were performed with endotheliumintact rabbit aortas and canine femoral arteries as the
'donor' vessels for EDRF. For bioassay of EDRF released
into its lumen, the 'donor' vessel was perfused in series
with endothelium-denuded 'detector' vessel (rabbit femoral artery or side branch of canine femoral artery) as
described by Forstermann et al. (1984). The tone of the
vascular segments was measured by continuously recording the external diameters at their midpoints using photoelectric devices (Busse et al., 1983, 1984). For bioassay of
EDRF, the mechanical responses of 'donor' and 'detector'
segments were recorded simultaneously.
Circulation Research/Vol. 58, No. 4, April 1986
Purification of Soluble Guanylate Cyclase from
Bovine Lung
Soluble, heme-containing GC was purified from bovine
lung to apparent homogeneity according to the method
described by Gerzer et al. (1981), but with some modifications. We used Sepharose CL-6B containing 1 jtmol
Cibacron blue F3G-A/ml swollen gel prepared according
to Heyns and De Moor (1974). It allows elution of GC
under low ionic strength conditions with a yield of about
50%. The properties of the isolated enzyme with an approximative molecular weight of 150,000 daltons were
described briefly (Mulsch and Bohme, 1984).
Determination of the Activity of Soluble Guanylate
Cyclase
The incubation procedure to determine the activity of
purified soluble GC was carried out in test tubes as well
as in the lumen of vascular segments. Purified soluble GC
(10-100 nM), with and without drugs added, was preincubated in test tubes for 3 minutes at 37°C and pH 7.4 in
modified Tyrode's solution under 5% CO2/ 20% O2 and
75% N2 or in a triethanolamine/HCl-buffer (50 HIM) containing glutathione (0.5 ITIM), MgCl2 (3 mM), and bovine
7-globulin (0.1 mg/ml). All concentrations given are final
concentrations at the respective step. The incubations were
started by addition of [a-32P)]GTP (0.1 mM, 0.2 jtCi). If the
effect of SNP on the activity of GC was to be evaluated
SNP was added together with [a-32P]GTP. After 3 minutes
the reactions were stopped by addition of zinc acetate (50
ITIM) and subsequently sodium carbonate (55 mM). Isolation of cGMP and calculation of guanylate cyclase activity
were performed as described (Schultz and Bohme, 1984).
Control experiments for the determination of non-enzymatically formed cGMP were performed in the absence of
guanylate cyclase. For the determination of recovery of
cGMP during the incubation procedure, cyclic [3H]GMP
(0.05 mM, 20 nCi) was used.
Incubation in Test Tubes
Initially, we investigated whether the effluent of endothelium-intact rabbit aortic segments stimulates purified
soluble GC in a test tube. Modified Tyrode's solution
containing NE (0.1 /IM) with and without ACh (2 ^M) was
injected into the lumen of rabbit aorta segments with
intact endothelium, preincubated for 0.5 and 3 minutes,
and transferred within 5 or 60 seconds to test rubes
containing purified soluble GC. Subsequently, the incubations were started by addition of [a"P]GTP (0.1 mM,
0.5 iiCi). The reactions were stopped after 25 or 180
seconds, as described above. In control experiments preincubations of NE with and without ACh were performed
in test tubes instead of vascular segments.
Incubations of Guanylate Cyclase Injected into the
Lumen of Vascular Segments
To circumvent technical difficulties arising from the
short half life of ACh-induced EDRF, we determined the
activity of purified soluble GC after injecting it into the
lumen of vascular segments. The segment was at first
perfused and tested for mechanical responsiveness to
ACh, as described above. Then the perfusion was stopped
Forstermann et a/./Endothelial Relaxing Factor and Guanylate Cydase
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
and the perfusion system was disconnected from the top
ends of the L-shaped steel cannulas while the arterial
segment remained in place in the organ bath. The remaining inrraluminal medium was cleared from the segment
by quickly removing the device containing the arterial
segment from the organ bath, turning it sideways, and
passing a gentle stream of air through the lumen. The
segment was promptly returned into the organ bath. Then
a preincubated (3 minutes, 37°C) purified soluble GC (34170 nM) in modified Tyrode's solution was mixed with
[a-32P]GTP (0.1 ITIM, 0.5 nCi) and cyclic [3H]GMP (0.05
ITIM, 20 nCi) and injected within 20 seconds into the lumen
of the vascular segments. The mixture contained cyclic
[3H]GMP to determine the recovery of cGMP. The ends
of both cannulas were closed with plugs and the incubation was allowed to proceed for a specified time. The
cannulas were then reopened, and the intraluminal incubation mixture was rapidly transferred from the vascular
lumen into the zinc acetate solution by gently blowing air
through the segment as described above. Following this,
the vascular segment was reconnected to the perfusion
system. The isolation and determination of the cyclic
[ PJGMP formed and calculation of GC activity was performed as described (Schultz and Bohme, 1984). Control
experiments were performed in test tubes instead of vascular segments under comparable conditions. Blank values
were obtained by omitting purified soluble GC from the
reaction mixture and were the same in test tubes and
isolated vascular segments.
After the first intraluminal determination of soluble GC
activity, the vascular segment was perfused for at least 16
minutes to wash out residual GC-containing solution prior
to the next incubation procedure. A maximum of five runs
was performed on each segment. In the experiments in
which the effects of drugs on EDRF formation and/or
release were to be evaluated, the activity of GC injected
into the lumen of vascular segments was first determined
in the presence of NE (0.1 ^M) and then in the presence
of both NE (0.1 /tM) and ACh (2 MM). Subsequently, the
segments were perfused for 60 minutes with modified
Tyrode's solution containing the drug under study. Then,
the determinations of GC activity were repeated in the
presence or absence of the drug.
GC activities are expressed as percent of test tube
controls (mean ± SE). Differences were tested for significance by Student's Ntest. Significance was accepted at the
0.05 level of probability.
Materials
Norepinephrine-HCl, acetylcholine-HCl, mepacrineHC1, nordihydroguaiaretic acid, indomethacin, and glutathione were purchased from Sigma. Atropine sulfate
was obtained from Drobena, and bovine 7-globulin was
from Serva. Other materials for the purification of soluble
GC and determination of enzyme activity were obtained
as previously described (Gerzer et al., 1981). Stock solutions of drugs were prepared immediately before use in
different solvents, and were then diluted with modified
Tyrode's solution or triethanolamine/HCl buffer. Nordihydroguaiaretic acid was dissolved in modified Tyrode's
solution (37°C) and indomethacin in 25% (vol/vol)
ethanol, 0.675% (wt/vol) NaCl, and 0.75% (wt/vol)
NaHCO3; all other drugs were dissolved in twice-distilled
water.
533
Results
Relaxation of Vascular Segments by
Acetylcholine and Sodium Nitroprusside
ACh (0.01-2 HM) induced relaxations of precontracted (NE, 0.1 HM), endothelium-intact rabbit aortas, and canine femoral arteries in modified Tyrode's
solution. Similarly, the ACh-induced release of
EDRF into the lumen of endothelium-intact 'donor'
rabbit aortic segments and the transfer of this EDRF
to endothelium-denuded 'detector* rabbit femoral
segments could be demonstrated in modified Tyrode's solution (Fig. 1). GTP (0.1 ITIM) and cGMP
(0.05 mM) had no effect on ACh-induced relaxation
and intraluminal release of EDRF (data not shown).
Furthermore, ACh-induced relaxations of 'donor" as
well as 'detector' vascular segments were abolished
in the presence of atropine (2 /XM), after pretreatment
of the 'donor' segments for 60 minutes with mepacrine (30 ^M) or nordihydroguaiaretic acid (20 JIM),
and after removal of the endothelium from the
'donor' segments. Thus, the arteries showed identical mechanical response in modified as in unmodified Tyrode's solution (cf. Forstermann et al., 1984).
0.1 uM Norepinephrine
[0.05
|2uM ACh I
3860 -,
1480-1
FIGURE 1. Representative diameter recordings (D) of an endotheliumintact rabbit aortic segment ("donor" segment, upper panel) and of an
endothelium-denuded rabbit femoral segment ("detector" segment,
lower panel) perfused in series with modified Tyrode's solution. The
effluent from the endothelium-intact vascular segment was subsequently perfused through the endothelium-denuded segment as described previously (Forstermann et al, 1984). Both vascular segments
were precontracted by perfusion with norepinephrine (NE, 0.1 pu).
Addition of acetylcholine (ACh, 0.05 and 2 yiM) induced relaxation of
the endothelium-intact vascular segment and the release of EDRF
from the "donor" segment which relaxed the endothelium-denuded
"detector" segment. The constituents of modified Tyrode's solution,
glutathione (0.5 mM), MgCl2 (3 mM), and bovine -/-globulin (0.1 mg/
ml), as well as GTP (0 1 mM) and cGMP (0.05 mM), did not affect
ACh-induced formation and/or release of EDRF.
Circulation Research/Vol. 58, No. 4, April 1986
534
The NO containing vasodilator SNP in a concentration of 10 nM caused the same degree of relaxation
of precontracted (NE, 0.1 /XM) endothelium-free
rabbit femoral arteries as did the effluent of endothelium-intact arteries after stimulation with ACh
(2
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
Activity of Purified Soluble Guanylate Cyclase
The activity of soluble, heme-containing GC purified from bovine lung was significantly inhibited
with modified Tyrode's solution instead of the triethanolamine/HCl buffer described in methods (Table 1). Basal GC activity determined in modified
Tyrode's solution was increased about 2-fold by
omission of NaCl (134 mM)or CaCl2 (1.6 ITIM), respectively. In the absence of NaCl and CaCl2, the
enzyme activity was comparable to that determined
in triethanolamine/HCl buffer. Correspondingly,
GC activity determined in triethanolamine/HCl
buffer was reduced to about 30% by the addition of
NaCl or CaCl2. The enzyme activity in the presence
of NaCl and CaCl2 was comparable to that measured
in modified Tyrode's solution (Table 1).
Also, the stimulation of soluble GC by SNP was
reduced in modified Tyrode's solution. Maximal
stimulation achieved with SNP was about 60-fold
in triethanolamine/HCl buffer, but only about 20fold in modified Tyrode's solution, both at about 20
HM SNP. The SNP concentration for half-maximum
stimulation was about 2 HM, independent of the
buffer used (data not shown). At 1 HM SNP the
extent of stimulation was about 20-fold in triethanolamine/HCl buffer and about 8-fold in modified
Tyrode's solution (Table 1). Addition of NaCl (134
mM), CaCl2 (1.6 ITIM), or both to the triethanolamine/
HC1 buffer reduced enzyme activities measured in
the presence of SNP and reduced the extent of
TABLE 1
Inhibition of Purified Soluble Guanylate Cyclase by Some
Constituents of Tyrode's Solution
Guanylate cyclase activity
(nmol/mg protein per min)
Buffer system
Without SNP
With SNP
(1 MM) added
Triethanolamine/HCl
+ CaCl2
+ NaCl
+ CaCl2 + NaCl
67.0 ± 10.0
25 1 ±4.7
20 1 ±4.2
9.4 ± 24
1340 ± 136
377 ±51.2
261 ±42.0
65.8 ± 8 5
Tyrode's solution
- CaCl2
- NaCl
- CaClj - NaCl
10.0 ± 2 2
20 0 ±3.0
24.1 ± 2.8
68.3 ± 6.5
80.0 ±9.8
240 ± 43.1
313 ±49 4
1230 ± 125
The activity of soluble, heme-containing guanylate cyclase
purified from bovine lung was determined in the presence and
absence of sodium nitroprusside (SNP, 1 MM) Incubations were
carried out in triethanolamine/HCl (5 IT\M) buffer or modified
Tyrode's solution, as described in Methods. CaClj (1.6 ITIM) and
NaCl (134 mM) were added (+) to or omitted (—) from the
incubation medium Data given are mean ± SE of three experiments performed in triplicates.
stimulation. Correspondingly, enzyme activities and
the extent of stimulation were increased by omission
of NaCl, CaCl2, or both, from the modified Tyrode's
solution (Table 1).
Basal activity of GC in modified Tyrode's solution
was 10.0 ± 2.2 nmol/mg per min. NE, ACh, atropine, mepacrine, nordihydroguaiaretic acid, and indomethacin, in the concentrations used, did not
affect basal GC activity or stimulation of the enzyme
by SNP when determined in modified Tyrode's
solution. In contrast, nordihydroguaiaretic acid (20
HM) and indomethacin (20 ^M) caused inhibition
(about 20%) of GC when determined in triethanolamine/HCl buffer.
ACh-Induced Stimulation of Purified Soluble
Guanylate Cyclase
In our first attempts to demonstrate activation of
soluble GC by an ACh-induced stimulating factor,
we transferred the effluent from the lumen of endothelium-intact rabbit aortic segments into test
tubes containing purified soluble GC. There was no
significant effect on GC activity under the conditions
described in Methods.
Therefore, the purified soluble GC was injected
into the lumen of rabbit aortic segments, where the
incubations for the determination of the enzyme
activity were carried out. There was a significant
increase in GC activity (164 ± 12%; n = 12) in the
absence of ACh and NE, in comparison to control
experiments performed in test tubes (100%) (Fig.
2a). In the presence of ACh (2 ^M), enzyme activity
was about 4.4-fold higher than in control experiments (Fig. 2c). The effect was dependent on the
concentration of ACh (Fig. 2, a-c), but independent
of the presence of NE (Fig. 2, d-f). The effects
described were independent of enzyme concentrations used (10-100 nM, data not shown). SNP (1 ^M)
enhanced the activity of soluble GC about 5-fold in
the arterial lumen (Fig. 2g), and about 8-fold in test
tubes (Table 1). This difference probably results
from diffusive loss of SNP into vessel walls and the
surrounding medium during the incubation period.
The activity of GC injected into canine femoral
segments was increased to 189 ± 39% (n = 6) of
control experiments. The activity was further increased to 341 ± 44% (n = 6) in the presence of NE
(0.1 fiM) and ACh (2 HM).
Formation of cGMP by purified soluble GC was
approximately proportional to the incubation time
(at least up to 720 seconds) in test tubes and in
rabbit aortic segments (Fig. 3). Also, the AChinduced stimulation of GC remained nearly constant
between 160 and 720 seconds, as indicated by the
approximately constant relation between stimulated
and control values (Fig. 3).
Inhibition of ACh-Induced Stimulation of
Purified Soluble Guanylate Cyclase
The activity of purified soluble GC injected into
the lumen of endothelium-denuded rabbit aortic
Forstermann et a/./Endothelial Relaxing Factor and Guanylate Cyclase
500
535
300-i
NE * ACh
(aorta)
400
8 ?
—
250
300
N
JS
u
© O
I
200-
200
150-
c
1001OO0
J
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
NE [|jM] ACh [yM] SNP[uM] -
—
0.05
-
—
2
-
01
—
-
01
Q05
-
01
2
-
a
1
FIGURE 2. Stimulation of soluble guanylate cyclase injected into the
lumen of rabbit aortic segments. Purified soluble guanylate cyclase
(CQ in modified Tyrode's solution was injected, together with its
substrate, into the lumen of vascular segments disconnected from the
perfuswn. The enzyme activity was determined intraluminaily in the
absence and presence of ACh, NE, and ACh with NE, as indicated.
CC activities obtained by incubation of the enzyme in modified
Tyrode's solution in test tubes were taken as control activities (100%).
The control value for the enzyme activity in the presence of sodium
nitroprusside (SNP) was obtained by incubation of the enzyme in test
tubes without SNP. The incubation time was always 3 minutes.
Enzyme activities are given as mean ± SE of seven or more experiments. CC activity obtained by intralummal incubation without
additions was significantly greater than that obtained by incubation
of the enzyme m test tubes + = Significantly greater than column a
(P < 0.01). * = Significantly greater than column d (P < 0.01)
z
50-
2
4
6
8
X)
12
incubation time [min]
FIGURE 3. Time-dependent formation of cGMP by soluble guanylate
cyclase. Purified soluble GC was incubated in modified Tyrode's
solution in test tubes (
) and intraluminaily in rabbit aortic
segments (
) for various times. NE (0.1 ^M) alone (open symbols),
or NE (0.1 IIM) ACh (2 HM) (closed symbols) were added to the
incubation medium. Each point represents means ± SE of four to six
experiments. + = Significantly greater than in test tubes at the same
incubation time (+ = P < 0.05; ++ = P < 0.01). * = Significantly
greater than with NE alone in the vascular segment at the same
incubation time C = P < 0.05, " = P < 0.01).
stimulation of GC was not affected by these drugs
(Fig. 4, c-e).
segments was not significantly different from test
tube controls (Fig. 4a). In addition, the ACh-induced
stimulation of GC was abolished in these segments
(Fig. 4a). The addition of atropine (2 ^M) significantly
reduced ACh-induced stimulation of GC in endothelium-intact segments to 163 ± 20% (n = 8) of
control values, whereas basal (ACh-independent)
enzyme stimulation due to endothelial cells was not
significantly altered (Fig. 4b). Pretreatment of endothelium-intact rabbit aortic segments for 60 minutes with mepacrine (30 MM) or nordihydroguaiaretic
acid (20 MM) significantly reduced ACh-induced
stimulation of GC to 198 ± 17% (n = 6) or 221 ±
10% (n = 6) of control values, respectively (Fig. 4, c
and d). These effects are independent of the presence of mepacrine or nordihydroguaiaretic acid, during the inrraluminal incubation of soluble GC (data
not shown). Pretreatment of the vascular segments
with indomethacin (20 ^M) did not significantly alter
ACh-induced GC stimulation (Fig. 4e). The AChindependent and basal endothelium-dependent
Discussion
To demonstrate the proposed stimulatory effect
of an ACh-induced endothelial factor on soluble
GC, it was necessary to use a medium which does
not disturb the integrity and responsiveness of the
endothelial cell membrane with respect to the formation and release of endothelial relaxing factoids).
On the other hand, this medium must allow the
determination of the activity of a purified soluble
GC. To meet the first demand, we had to use an
extracellular medium like Tyrode's solution; however, additions such as glutathione, 7-gIobulin,
GTP, cGMP, and additional MgC^ were necessary
to achieve a sufficient GC activity and recovery of
the cGMP formed. The constituents of Tyrode's
solution NaCl and CaCl2 reduced significantly the
activity of the purified soluble, heme-containing
GC, and, to a greater degree, the enzyme stimulation
by SNP. The effect of NaCl is probably caused by
the increased ionic strength of the buffer. Effects of
Ca++ on soluble GC have been described by several
Circulation Research/Vol. 58, No. 4, April 1986
536
400 -i
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
-Endo
Atrop
Mepa
NDGA
Indo
a
b
C
d
e
FIGURE 4. Effect of different expenmental conditions on the stimulation of soluble guanylate cyclase injected into the lumen of rabbit aortic
segments. The vascular segments were disconnected from the perfusion system; purified soluble GC was injected together with its substrate into
the lumen. The enzyme activity was determined in the presence of NE (0.1 \IM); incubation time was 3 minutes. After a wash-out period
(perfusion), the enzyme activity was determined again in the presence of NE (0.1 fiM) and ACh (2 ^M). The determinations of enzyme activities
were repeated after the following treatments of the vascular segments: (a) removal of the endothelium (-Endo), (b) addition of atropine (Atrop,
2 HM), (C) perfusion of the segments for 60 minutes with mepacrine (Mepa, 30 HM), (d) with nordihydroguaiaretic acid (NDGA, 20 fiin), (e) with
indomethacin (Indo, 20 fiM). Each column represents the mean ± SE of at least four experiments. GC activities determined in the lumen of the
vascular segments are given as percent of the enzyme activities obtained in parallel test tube incubations with the respective additions. Plusses
= significantly greater or smaller than the first column of the same group, i.e., NE alone (+ = P < 0.05; ++ = P < 0.005). Asterisks =
significantly smaller than the second column of the same group, i.e., NE and ACh (* = P < 0.05; ** = P < 0.005). n.s. = not significantly different.
authors. So far, these studies show conflicting results, probably because different enzyme preparations were used. Inhibitory effects of Ca on the
stimulation of soluble GC by NO-containing compounds but not on the basal enzyme activity have
been described in more detail, e.g., by Gruetter et
al. (1980). The effect of Ca++ as well as ionic strength
on the purified, soluble, heme-containing guanylate
cyclase certainly deserves further investigation.
If purified soluble GC was injected, together with
its substrate, into the lumen of endothelium-intact
rabbit aortic or canine femoral arterial segments, a
significant stimulation of the enzyme was observed
in the absence of ACh. Since removal of endothelium reduced the enzyme activities to values obtained in the test tubes, this points to a basal release
of a GC-stimulating factors) from endothelial cells.
Similarly, a basal release of ERDF has recently been
demonstrated by Griffith et al. (1984). Furthermore,
it has been shown that removal of endothelium
decreases cGMP levels in the subjacent smooth muscle cells (Holzmann, 1982; Diamond and Chu, 1983;
Rapoport and Murad, 1983a, 1983b; Furchgott et
al., 1984; Miller et al., 1984).
The activity of the purified soluble GC injected
into the lumen of vascular segments was further
stimulated by ACh in a concentration-dependent
manner. This stimulation was prevented by removal
of endothelium or by the addition of atropine. These
results indicate that ACh-induced stimulation of
soluble GC is endothelium dependent and is mediated by muscarinic receptors. The same has been
shown for ACh-induced formation and/or release
of EDRF (Furchgott and Zawadzki, 1980).
The extent of the ACh-induced stimulation of
soluble GC was independent of the incubation time.
The stimulatory factor(s) obviously reached a steady
state level, suggesting a short life time as described
for the ACh-induced EDRF. Similarly, we were unable to demonstrate any significant stimulation of
soluble GC in test tubes by the effluent of endothelium-intact rabbit aortic segments treated with ACh.
This further indicates that the stimulating activity
does not accumulate in the vascular lumen, and is
obviously chemically unstable, as has been reported
for EDRF (Griffith et al., 1984; Forstermann et al.,
1984).
ACh-induced stimulation of soluble GC was inhibited after pretreatment of endothelium-intact
rabbit aortic segments with mepacrine or nordihydroguaiaretic acid at concentrations that also inhibited the ACh-induced formation and/or release of
EDRF (Furchgott and Zawadzki, 1980; Chand and
Alrura, 1981; Furchgott, 1983, 1984; Forstermann et
al., 1984; Forstermann and Neufang, 1984). The
blockade of the ACh-induced GC stimulation was
Forstermann et a/./Endothelial Relaxing Factor and Guanylate Cyclase
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
evident, even though these drugs were absent during the actual incubation. Thus, inhibition takes
place at the level of the endothelium, and is not
caused by an interaction of these drugs with soluble
GC itself. The ineffectiveness of indomethadn to
inhibit ACh-induced stimulation of soluble GC implies that the formation and/or release of the stimulating principle is insensitive to indomethacin, at
least in the concentration used, as shown for EDRF
(Furchgott and Zawadzki, 1980; Furchgott, 1983,
1984; Forstermann and Neufang, 1984).
It has previously been reported that platelet membranes can release fatty acids that modulate GC
activity (Gerzer et al., 1983). Arachidonic acid or
linoleic acid are known to stimulate basal activity of
soluble GC (Bohme et al., 1983; Gerzer et al., 1983;
Wolin and Ignarro, 1983). However, it is unlikely
that these fatty acids, or any of their relatively stable
metabolites, are responsible for the activation of GC
observed here, since the GC stimulating factors)
obviously has a short life time.
The extent of ACh-induced increase in GC activity
shown here would not account for the increase of
cGMP levels in rabbit aorta (Diamond and Chu,
1983; Furchgott and Jothianandan, 1983; Furchgott
et al., 1984). However, the intracellular medium is
different from the intraluminal medium in which
the determination of soluble GC was performed.
The basal enzyme activity, and especially the stimulation by SNP, were reduced in modified TyTode's
solution. Therefore, it is likely that the effect of the
ACh-induced stimulatory factor(s) on GC is reduced
as well. Furthermore, the release of ACh-induced
EDRF seems to be greater in the abluminal than in
the intraluminal direction (Busse et al., 1985).
In conclusion, the results presented demonstrate
that endothelial cells, of at least two types of arteries
(rabbit aorta and canine femoral artery), produce a
labile factor(s) which stimulates soluble GC. The
formation and release of this factor is increased by
ACh. The factor may be identical with the AChinduced EDRF. This assumption is supported by
correlations between known properties of EDRF and
those found for the GC-stimulating factors). Thus,
stimulation of GC, at least of the soluble enzyme
form, could be responsible for the increase in cGMP
levels observed in vascular smooth muscle cells during endothelium-dependent relaxation.
We would like to acknowledge the excellent technical assistance of
Cornelia Kellner.
This work was presented in part at the Spring Meeting of the
German Pharmacological Society, March 1985 (Mulsch et al., 1985)
The work was supported by the Deutsche Forschungsgemeinschaft
(Grants Schu 2/14-5 and Bu 436/2-2).
Dr. Forstermann was affiliated with the Department of Pharmacology at the University of Freiburg, Drs Mulsch and Bohme with
the Department of Pharmacology at the University of Heidelberg, and
Dr. Busse with the Department of Applied Physiology at the University of Freiburg.
Dr. U Fdrstermann's present address is: Department of Clinical
537
Pharmacology, Hannover Medical School, Konstanty-Gutschow-Str.
8, D-3000 Hannover 61, F.R.G.
Address for reprints. Dr. Rudi Busst, Department of Applied
Physiology, University of Freiburg, Hermann-Herder-Str. 7, D-7800
Freiburg, F.R.G.
Received December 11, 1984; accepted for publication October 3,
3985.
References
Bohme E, Gerzer R, Grossmann G, Herz J, Mulsch A, Spies C,
Schultz G (1983) Regulation of soluble guanylate cyclase activity. In Hormones and Cell Regulation, vol 7, edited by JE
Dumont, J Nunez, RM Denton. Amsterdam, Elsevier Biomedical Press, pp 147-161
Bohme E, Grossmann G, Herz J, Mulsch A, Spies C, Schultz G
(1984) Regulation of cyclic GMP formation by soluble guanylate cyclase: Stimulation by NO-containing compounds. In
Advances in Cyclic Nucleotide and Protein Phosphorylation
Research, vol 17, edited by P Greengard, GA Robison, R
Paoletti, S Nicosia. New York, Raven Press, pp 259-266
Busse R, Pohl U, Kellner C, Klemm U (1983) Endothelial cells are
involved in the vasodilatory response to hypoxia. Pflugers Arch
397: 78-80
Busse R, Forstermann U, Matsuda H, Pohl U (1984) The role of
prostaglandins in the endothelium-mediated vasodilatory response to hypoxia. Pflugers Arch 401: 77-83
Busse R, Pohl U, Luckhoff A, Bassenge E (1985) Evidence for
abluminal release of endothelium derived relaxant factor
(absrr). Circulation 72 (suppl II): 81
Chand N, AJtura BM (1981) Acetylcholine and bradykinin relax
intrapulmonary arteries by acting on endothelial cells: Role in
lung vascular diseases Science 213: 1376-1379
Craven PA, DeRubertis FR (1983) Requirement for heme in the
activation of purified guanylate cyclase by nitnc oxide. Biochim
Biophys Acta 745: 310-321
DeMey JG, Claeys M, Vanhoutte PM (1982) Endothelium-dependent inhibitory effects of acetylcholine, adenosine triphosphate, thrombin and arachidonic acid in the canine femoral
artery. J Pharmacol Exp Ther 222: 166-173
Diamond J, Chu EB (1983) Possible role for cyclic GMP in
endothelium-dependent relaxation of rabbit aorta by acetylcholine. Comparison with nitroglycerin. Res Commun Chem Pathol Pharmacol 41: 369-381
Forstermann U, Neufang B (1984) The endothelium-dependent
vasodilator effect of acetylcholine: Characterization of the endothelial relaxing factor with inhibitors of arachidonic acid
metabolism. Eur J Pharmacol 103: 65-70
Forstermann U, Trogjsch G, Busse R (1984) Species-dependent
differences in the nature of endothelium-derived vascular relaxing factors. Eur J Pharmacol 106: 639-643
Furchgott RF (1983) Role of endothelium in responses of vascular
smooth muscle. Circ Res 53: 557-573
Furchgott RF (1984) The role of endothelium in the responses of
vascular smooth muscle to drugs. Annu Rev Pharmacol Toxicol
24: 175-197
Furchgott RF, Jothianandan D (1983) Relation of cyclic GMP
levels to endothelium-dependent relaxation by acetylcholine in
rabbit aorta (absrr). Fed Proc 42: 619
Furchgott RF, Zawadzki JV (1980) The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by
acetylcholine. Nature 288: 373-376
Furchgott RF, Zawadzki JV, Cherry PD (1981) Role of endothelium in the vasodilator response to acetylcholine. In Vasodilation, edited by PM Vanhoutte, I Leusen. New York, Raven
Press, pp 49-66
Furchgott RF, Cherry PD, Zawadzki JV, Jothianandan D (1984)
Endothelial cells as mediators of vasodilation of arteries. J
Cardiovasc Pharmacol 6: S336-S343
Gerzer R, Hofmann F, Schultz G (1981) Purification of a soluble,
sodium-nitroprusside-stimulated guanylate cyclase from bovine lung. Eur J Biochem 116: 479-486
Gerzer R, Hamet P, Ross AH, Lawson JA, Hardman JG (1983)
Calcium-induced release from platelet membranes of fatty acids
538
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
that modulate soluble guanylate cyclase. J Pharmacol Exp Ther
226: 180-186
Griffith TM, Edwards DH, Lewis M], Newby AC, Henderson AH
(1984) The nature of endothelium-derived vascular relaxant
factor. Nature 308: 645-647
Gruetter DY, Gruetter CA, Barry BK, Bancos WH, Hyman AL,
Kadowitz PJ, Ignarro LJ (1980) Activation of coronary arterial
guanylate cyclase by nitric oxide, nitroprusside, and nirrosoguanidine: Inhibition by calcium, lanthanum, and other cations,
enhancement by thiols. Biochem Pharmacol 29: 2943-2950
Heyns W, De Moor P (1974) A 3(17)/9-hydroxysteroid dehydrogenase in rat erythrocytes. Conversion of 5a-dihydrotestosterone into 5«-androstane-3jS,170-diol and purification of the
enzyme by affinity chromatography. Biochim Biophys Acta
358: 1-13
Holzmann S (1982) Endothelium-induced relaxation by acetylcholine assooated with larger rises in cyclic GMP in coronary
arterial strips. J Cycl Nucl Res 8: 409-419
Ignarro LJ, Kadowitz PJ (1985) The pharmacological and physiological role of cyclic GMP in vascular smooth muscle relaxation. Annu Rev Pharmacol Toxicol 25: 171-191
Ignarro LJ, Burke TM, Wood KS, Wolin MS, Kadowitz PJ (1984)
Association between cyclic GMP accumulation and acetylcholine-elicited relaxation of bovine intrapulmonary artery. J Pharmacol Exp Ther 228: 682-690
Miller RC, Mony M, Schini V, Schoeffter P, Stoclet JC (1984)
Endothelial mediated inhibition of contraction and increase in
cyclic GMP levels evoked by the a-adrenoceptor agonist B-HT
920 in rat isolated aorta. Br J Pharmacol 83: 903-908
Mittal CK, Murad F (1982) Guanylate cyclase: Regulation of cyclic
GMP metabolism. In Handbook of Experimental Pharmacology, vol 52/1, edited by JA Nathanson, JW Kebabian. Berlin,
Circulation Research/Vol. 58, No. 4, April 1986
Springer-Verlag, pp 225-260
Mulsch A, Bohme E (1984) UV/VIS- and EPR-spectroscopic investigation of soluble guanylate cyclase activation by NO and
sydnonimines (abstr). Naunyn Schmiedebergs Arch Pharmacol
325 (suppl): R31
Mulsch A, Busse R, Bohme E (1985) Evidence for stimulation of
soluble guanylate cyclase by endothelium-derived relaxing factor (abstr). Naunyn Schmiedebergs Arch Pharmacol 329
(suppl): R36
Poole JCF, Sanders AG, Florey UW (1958) The regeneration of
aortic endothelium. J Pathol Bacteriol 75: 133-143
Rapoport RM, Murad F (1983a) Agonist-induced endotheliumdependent relaxation in rat thoracic aorta may be mediated
through cGMP. Circ Res 52: 352-357
Rapoport RM, Murad F (1983b) Endothelium-dependent and
nitrovasodilator-induced relaxation of vascular smooth muscle:
role of cyclic GMP. J Cyclic Nucleotide Protein Phosphor Res
9: 281-296
Rapoport RM, Draznin MB, Murad F (1984) Mechanisms of
adenosine tnphosphate-, thrombin-, and trypsin-induced relaxation of rat thoracic aorta. Circ Res 55: 468-479
Schultz G, Bohme E (1984) Guanylate Cyclase. In Methods of
Enzymatic Analysis, vol 3, edited by HU Bergmeyer. Weinheim,
Verlag Chemie, pp 379-389
Wolin MS, Ignarro LJ (1983) Activation of purified soluble guanylate cyclase by arachidonic acid and catalase: Evidence for
two separate mechanisms (abstr). Fed Proc 41 1853
INDEX TERMS: Endothelium-denved relaxing factor • Soluble
guanylate cyclase • Acetylcholine • Atropine • Mepacrine •
Nordihydroguaiaretic acid
Stimulation of soluble guanylate cyclase by an acetylcholine-induced endothelium-derived
factor from rabbit and canine arteries.
U Förstermann, A Mülsch, E Böhme and R Busse
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
Circ Res. 1986;58:531-538
doi: 10.1161/01.RES.58.4.531
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1986 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/58/4/531
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/