Hypertension. 1996;27:1346-1352
(Hypertension. 1996;27:1346-1352.)
© 1996 American Heart Association, Inc.
Carbenoxolone Damages Endothelium and Enhances Vasoconstrictor Action in Aortic Rings
Michael E. Ullian;
Debra J. Hazen-Martin;
Lyle G. Walsh;
Rajesh K. Davda;
Brent M. Egan
From the Departments of Medicine (M.E.U., L.G.W., R.K.D., B.M.E.),
Pathology (D.J.H.-M.), and Pharmacology (B.M.E.), Medical University of South
Carolina and Ralph H. Johnson Veterans Administration Hospital, Charleston.
 |
Abstract
|
|---|
Abstract Carbenoxolone causes hypertension indirectly by
inhibition
of 11ß-hydroxysteroid dehydrogenase and consequent
elevation
of intracellular glucocorticoid levels and enhancement of
vasoconstrictor
action. We performed the present study to determine
whether
carbenoxolone also enhances vascular tone directly by
mechanisms
independent of glucocorticoids and other systemic
influences.
Exposure of rat aortic rings to 10 to 100 µmol/L
carbenoxolone
in aerated Krebs-Henseleit buffer for 24 hours resulted
in concentration-dependent
increases in angiotensin II
(Ang II) (100 nmol/L)stimulated
contractions and significant shifting
of the phenylephrine cumulative
contraction curve to the
left but not increases in KCl (120
mmol/L)stimulated contractions.
Maximal enhancement of
Ang II contraction was 39%. In contrast, brief
(15-minute) exposure
to 100 µmol/L carbenoxolone did not alter Ang II
contractions.
Mechanical denudation of the endothelium
obviated enhancement
of Ang II contractions by carbenoxolone,
suggesting interaction
of carbenoxolone with the
endothelium. Endothelium-dependent
relaxation
of precontracted rings to acetylcholine or ATP was reduced
by
more than 90% by 24-hour pretreatment with 100 µmol/L
carbenoxolone
but not with 100 µmol/L deoxycorticosterone acetate (a
mineralocorticoid)
or 100 µmol/L glycyrrhizic acid (a natural
11ß-hydroxysteroid
dehydrogenase inhibitor). Vascular
smooth muscle relaxation
with sodium nitroprusside was not inhibited by
carbenoxolone.
Incubation of cultured endothelial cells
with 100 µmol/L
carbenoxolone for 24 hours did not inhibit nitric
oxide synthase
activity, as measured by conversion of
[
3H]
L-arginine to
[
3H]
L-citrulline.
Electron micrography
demonstrated that endothelial cell ultrastructure
but
not vascular smooth muscle cell ultrastructure was abnormal
after
incubation of rings for 24 hours with 100 µmol/L
carbenoxolone. These
studies suggest that carbenoxolone concentrations
higher than 10
µmol/L enhance vasoconstrictor action via
selective toxicity to the
endothelium and elimination of
endothelium-dependent
relaxation.
Key Words: carbenoxolone endothelium muscle, smooth, vascular vasoconstriction receptors, adrenergic angiotensin II
 |
Introduction
|
|---|
Compounds that inhibit
the enzyme 11ß-hydroxysteroid dehydrogenase
cause
hypertension. These include glycyrrhizic acid and glycyrrhetinic
acid,
natural compounds found in licorice and tobacco, and
carbenoxolone,
the semisynthetic hemisuccinate derivative of
glycyrrhetinic
acid. Carbenoxolone has been used as a treatment for
peptic
ulcer disease. It has been hypothesized that
11ß-hydroxysteroid
dehydrogenase allows certain tissues to be
mineralocorticoid
targets.
1 Although type 1
adrenocorticosteroid receptors have
equal binding
affinity for the glucocorticoid cortisol and the
mineralocorticoid
aldosterone and although circulating glucocorticoid
concentrations
exceed circulating mineralocorticoid concentrations by
1000-fold,
glucocorticoids (but not mineralocorticoids) are metabolized
to
less active compounds by 11ß-hydroxysteroid dehydrogenase
in
certain tissues such as the kidney. Consequently, mineralocorticoids
can
gain access to type 1 receptors. In the setting of
11ß-hydroxysteroid
dehydrogenase inhibition, however,
glucocorticoids may flood
type 1 receptors and result in excess sodium
reabsorption, expansion
of extracellular fluid volume, and
hypertension.
Although it contains less than the kidney, the vasculature also
contains 11ß-hydroxysteroid dehydrogenase.2 3 Since
glucocorticoids potentiate vascular constrictor responses to a number
of pressor hormones such as norepinephrine and
angiotensin II (Ang II),4 5 6 7 it is possible
that inhibition of 11ß-hydroxysteroid dehydrogenase results in
increased glucocorticoid levels in vascular smooth muscle and
further potentiation of vasoconstrictor action. In fact, oral treatment
with carbenoxolone or skin treatment with glycyrrhetinic acid
resulted in increased dermal constriction to
endogenous pressors, and carbenoxolone treatment resulted
in increased blood pressure responses and heightened forearm blood flow
reductions to exogenous norepinephrine.8 9
Recent studies have suggested that carbenoxolone may alter vascular
tone by mechanisms independent of 11ß-hydroxysteroid
dehydrogenase or glucocorticoid levels. Short-term exposure of
aortic rings to carbenoxolone resulted in alterations in
endothelium-dependent relaxation,10
and carbenoxolone treatment of aortic rings from adrenalectomized
animals (which lack circulating corticosteroids)
resulted in potentiation of norepinephrine-stimulated
vasoconstriction.11 Therefore, we hypothesized that
carbenoxolone, in addition to its indirect enhancement of vascular
smooth muscle tone via alterations in glucocorticoid
metabolism, directly enhances vascular responses to pressor
hormones via attenuation of endothelium-dependent
relaxation. We performed contraction studies after exposure of
endothelium-denuded and
endothelium-intact rings to carbenoxolone ex vivo.
We used this ex vivo experimental system to avoid secondary systemic
effects of carbenoxolone.
 |
Methods
|
|---|
Aortic Ring Contractions
Aortic ring segments were obtained from Sprague-Dawley
rats
by procedures approved by the Institutional Animal Care and
Utilization
Committee of the Medical University of South Carolina.
Thoracic
aortas were cleaned of fat and adventitia and cut into rings
5
mm in length. Rings were treated with the desired agents for
the
desired time periods in Krebs-Henseleit bicarbonate buffer
bubbled with
95% O
2/5% CO
2 to pH 7.4 at room
temperature in
an Erlenmeyer flask. Rings were then attached to an
isometric
force-displacement transducer under 2 g of tension
and equilibrated
for 1 hour in the same buffer at 37°C. Effectors and
inhibitors
in 10-µL volumes were added
simultaneously to control
and experimental rings in
side-by-side 10-mL organ chambers
to achieve the desired final
concentrations. Cumulative contractions
to phenylephrine
were achieved by administration of increasing
concentrations of
phenylephrine when the contraction from the
previous
concentration had plateaued. Contraction intensity
was expressed as
grams tension per milligram dry weight of aorta.
Maintenance of
vascular segments ex vivo for prolonged periods
of time has been
described previously.
12 13
Nitric Oxide Synthase Activity
Confluent endothelial cells in triplicate in 12-
or 24-well plates were washed twice with phosphate-buffered saline
and then exposed to 1 mL of [3H]L-arginine (3
µCi/mL) for 20 minutes. Reactions were terminated after various
stimulations by washing cells with ice-cold calcium-free buffer
containing 5 mmol/L EDTA and adding 1 mL of 0.3 mol/L perchloric acid.
Nitric oxide (NO) synthase activity was measured as the conversion of
[3H]L-arginine to
[3H]L-citrulline after separation of these
amino acids by ion-exchange chromatography.
Vascular Ultrastructure
Aortic rings were fixed in 2.5% glutaraldehyde
in phosphate-buffered saline for 1.5 hours after the desired
treatment. After fixation, rings were rinsed in phosphate-buffered
saline and post-fixed in 2% aqueous osmium tetroxide. The rings
were then rinsed in distilled water, dehydrated in a graded series of
ethanols, cleared in propylene oxide, and infiltrated and embedded in
Embed 812. Half-micron-thick sections were obtained and stained
in toluidine blue for evaluation. Representative areas
were used to obtain 70-nm-thin sections that were stained with
uranyl acetate and lead citrate. Micrographs of thin section profiles
were obtained with an electron microscope (JEOL 100S).
Materials
All chemicals and reagents were obtained from Sigma Chemical Co
except the following: [3H]L-arginine (New
England Nuclear), rats (Harlan Sprague Dawley, Indianapolis, Ind),
ion-exchange resin (Bio-Rad), Ang II (Peninsula Laboratories),
Embed 812 (EM Sciences), and pulmonary artery
endothelial cells (American Type Culture Collection).
Aortic endothelial cells were kindly provided by Dr
W.C. O'Neill, Emory University School of Medicine, Atlanta, Ga.
 |
Results
|
|---|
Effects on Vasoconstrictor Action
Ring contraction studies were performed after rings had been
exposed
to carbenoxolone ex vivo. As demonstrated in Fig 1

, exposure
of rings to 100 µmol/L carbenoxolone for
24 hours resulted
in significant enhancement of contractions to 100
nmol/L Ang
II (39% greater than control) but no enhancement of KCl
contractions.
Incubation with 50 µmol/L carbenoxolone also enhanced
Ang
II contractions (0.34±0.13 [control] versus 0.73±0.09
[carbenoxolone]
g tension/mg dry aortic weight, n=4,
P<.01), but 10 µmol/L
carbenoxolone was ineffective
(0.58±0.12 [control] versus
0.53±0.08 [carbenoxolone], n=4,
P=NS). Fifteen-minute
exposure of rings to 100 µmol/L
carbenoxolone did not
increase contractile responses to 100 nmol/L Ang
II (0.39±0.11
[control] versus 0.25±0.08 [carbenoxolone], n=4).
To
determine whether the effect of carbenoxolone on vascular reactivity
was
specific for Ang II, we repeated studies with the
1-adrenergic
agonist phenylephrine. Unlike
Ang II, whose contractions desensitize
on repeated
exposure,
14 phenylephrine can be repetitively
administered
to generate cumulative contraction curves. Exposure of
rings
ex vivo to 0 or 100 µmol/L carbenoxolone for 24 hours
resulted
in maximal contractions (to 10 µmol/L phenylephrine)
that
were not different (1.52±0.30 [control] versus 1.71±0.21
[carbenoxolone],
P=NS). However, carbenoxolone
significantly potentiated phenylephrine
contractions (ie,
shifted the cumulative contraction curve to
the left) (Fig 2

). In summary, these studies demonstrate that
prolonged
incubations with carbenoxolone concentrations higher
than 10 µmol/L
result in enhanced vasoconstriction to
at least two distinct pressor
hormones.

View larger version (21K):
[in this window]
[in a new window]
|
Figure 1. Effects of 24-hour carbenoxolone (Carb) exposure on
KCl- and Ang IIstimulated contractions in aortic rings. Aortic rings
were isolated and aerated for 24 hours in the presence of 0 (water) or
100 µmol/L carbenoxolone. After stretching and equilibration (see
"Methods"), rings were stimulated with 120 mmol/L KCl. Once
maximal contraction was achieved, rings were washed and original
baseline tension was reestablished. Then, stimulation with 100 nmol/L
Ang II was performed. Data are expressed as increases over basal
tension in response to effector. Lines connect data from rings from the
same animal. Comparisons were by paired t test. Con
indicates control.
|
|

View larger version (18K):
[in this window]
[in a new window]
|
Figure 2. Effects of 24-hour carbenoxolone (Carb) exposure on
phenylephrine-stimulated contractions in aortic rings.
Aortic rings were exposed to 0 or 100 µmol/L carbenoxolone for 24
hours, and cumulative contractions to phenylephrine were
performed. Comparisons were between control (Con) and experimental at
each phenylephrine concentration by unpaired t
test.
|
|
Role of Endothelium
Since a previous study suggested that carbenoxolone affected
endothelial cell function,10 we performed
studies to determine whether carbenoxolone would enhance Ang II
contractions in the absence of endothelium. After rings
were exposed to 0 or 100 µmol/L carbenoxolone for 24 hours,
endothelium was removed from all rings with a
cotton-tipped applicator. Absence of endothelium
was confirmed by assessing the degree of relaxation to 1 µmol/L
acetylcholine (an agent that causes release of NO from
endothelial cells) after precontraction with 1 µmol/L
phenylephrine. In both control and
carbenoxolone-treated rings, acetylcholine-mediated relaxation
was less than 10% after physical manipulation of
endothelium. In contrast to the results of experiments
with intact endothelium, carbenoxolone did not enhance
Ang II contractions in the absence of endothelium
(0.60±0.07 [control] versus 0.56±0.05 [carbenoxolone] g
tension/mg dry aortic weight, n=7, P=NS). It should be noted
that Ang II contractions tended to be greater in the absence of
endothelium than in its presence.
Since enhancement of Ang IIstimulated contractions by carbenoxolone
was not observed after endothelium was removed, the
effects of carbenoxolone on endothelium-dependent
relaxation were investigated. Fig 3
shows a
representative tracing of contractions and relaxations
of rings from a single aorta treated with 0 or 100 µmol/L
carbenoxolone for 24 hours ex vivo. Contractions to 100 nmol/L Ang II
but not to 120 mmol/L KCl were greater in carbenoxolone-treated
rings than in control rings, as demonstrated in Fig 1
.
Endothelium-dependent relaxation (to 1 µmol/L
acetylcholine and to 100 µmol/L ATP) from
phenylephrine-mediated precontraction was greater than
80% in the control rings but less than 10% in the treated rings. Fig 4
summarizes these data from rings from a number of
rats. Fig 5
demonstrates that loss of
endothelium-dependent relaxation was concentration
dependent from 10 to 100 µmol/L carbenoxolone. For determination of
whether carbenoxolone affected the ability of vascular smooth muscle to
relax, rings treated with 0 or 100 µmol/L carbenoxolone for 24 hours
were precontracted with 1 µmol/L phenylephrine and then
exposed to 100 µmol/L sodium nitroprusside, an agent that elaborates
NO in vascular smooth muscle. Relaxation to sodium nitroprusside was
greater than 90% in both groups of rings (data not shown).

View larger version (15K):
[in this window]
[in a new window]
|
Figure 3. Representative tracings of
contractile and relaxation responses of aortic rings treated with
carbenoxolone for 24 hours. Aortic rings were exposed to 0 or 100
µmol/L carbenoxolone for 24 hours and then in succession to 120
mmol/L KCl, 100 nmol/L Ang II, 1 µmol/L phenylephrine
(PE), 1 µmol/L acetylcholine (ACh), 1 µmol/L
phenylephrine, and 100 µmol/L ATP. Washing with buffer
(Wash) was performed, and original baseline tension was reestablished
as indicated. These tracings are representative of five
similar experiments.
|
|

View larger version (19K):
[in this window]
[in a new window]
|
Figure 4. Effects of 24-hour carbenoxolone (Carb) exposure on
endothelium-dependent relaxation in precontracted
aortic rings. Data were derived from experiments like the one shown in
Fig 3 . Relaxation is defined as percent reversal of
phenylephrine contraction by acetylcholine (ACh) or ATP.
Comparisons were by unpaired t test. Con indicates
control.
|
|

View larger version (35K):
[in this window]
[in a new window]
|
Figure 5. Concentration-dependent effects of 24-hour
carbenoxolone exposure on acetylcholine-mediated relaxation in
precontracted aortic rings. Studies were performed as described in Fig 3 , except that three different carbenoxolone concentrations (10, 50, or
100 µmol/L) were used during the 24-hour exposure. Each carbenoxolone
concentration was administered to different sets of rings. Comparisons
were by unpaired t test.
|
|
To determine whether the loss of
endothelium-dependent relaxation after incubation
with carbenoxolone could be related to depletion of the substrate for
NO synthase, we repeated some of the studies described above in the
presence of the NO precursor L-arginine. Rings were
incubated with 0 or 10 mmol/L L-arginine for 1 hour before
the addition of 100 µmol/L carbenoxolone for an additional 24 hours.
Fig 6
demonstrates that preincubation with
L-arginine prevented a portion of
carbenoxolone-mediated loss of
endothelium-dependent relaxation.

View larger version (25K):
[in this window]
[in a new window]
|
Figure 6. Effect of L-arginine on
carbenoxolone-mediated loss of
endothelium-dependent relaxation. Aortic rings were
exposed to 0 or 10 mmol/L L-arginine for 1 hour ex vivo and
then 0 or 100 µmol/L carbenoxolone for an additional 24 hours in the
continued presence (or absence) of L-arginine. Contractions
to 1 µmol/L phenylephrine were performed, and relaxation
was effected with 1 µmol/L acetylcholine. Data were analyzed
by one-way ANOVA and means compared post hoc with the
Student-Newman-Keuls test.
|
|
Since the triterpenoid carbenoxolone is similar in structure to steroid
hormones, we performed studies to determine whether other triterpenoids
or adrenocorticosteroid hormones also inhibit
endothelium-dependent relaxation. We were unable to
dissolve glycyrrhetinic acid in buffer at the desired concentrations
and therefore were unable to evaluate its effects. Incubation with
either the mineralocorticoid deoxycorticosterone acetate or the
naturally occurring triterpenoid glycyrrhizic acid at 100 µmol/L for
24 hours did not significantly inhibit relaxation of precontracted
rings to 1 µmol/L acetylcholine (71±5% [control] versus 70±6%
[glycyrrhizic acid] versus 53±7% [deoxycorticosterone acetate],
n=5, P=NS).
We performed an additional study to determine whether inhibition of
endothelium-dependent relaxation by an agent
different from carbenoxolone would also enhance Ang II contractions.
Rings were incubated for 15 minutes with 0 or 100 µmol/L of the NO
synthase inhibitor
NG-nitro-L-arginine
methyl ester (L-NAME), and contractions to 100 nmol/L Ang II were
measured. L-NAME treatment, like carbenoxolone treatment, enhanced Ang
II contractions (0.45±0.11 [control] versus 0.87±0.11 [L-NAME] g
tension/mg dry aortic weight, P<.01 by paired t
test).
Effects on NO Synthase Activity
The results described above are consistent with the
possibility that endothelial cell function but not
vascular smooth muscle cell function is impaired by carbenoxolone.
Therefore, we performed studies to determine the effect of
carbenoxolone on NO synthase activity. Confluent bovine
pulmonary artery endothelial cells and bovine
aortic endothelial cells were treated with 0 or 100
µmol/L carbenoxolone for 24 hours. Then, conversion of
[3H]L-arginine to
[3H]L-citrulline as a measure of NO synthase
activity was determined in the basal state and in response to ionomycin
(maximal) and ATP (receptor-mediated). After 24-hour treatment with
carbenoxolone, endothelial cells in culture were not
grossly altered by carbenoxolone, as assessed by visual inspection. NO
synthase activity was stimulated by ATP and ionomycin, but this
stimulation was similar in control and carbenoxolone-treated cells
(Fig 7
).

View larger version (34K):
[in this window]
[in a new window]
|
Figure 7. Effects of 24-hour carbenoxolone (Carb) exposure on
nitric oxide (NO) synthase activity in cultured
endothelial cells. Confluent bovine pulmonary
artery endothelial cells and bovine aortic
endothelial cells were treated with 0 or 100 µmol/L
carbenoxolone for 24 hours, and then NO synthase activity was measured
(see "Methods"). Basal activity was measured after exposure to
water for 20 minutes. Hormone-stimulated and maximal NO synthase
activities were measured after exposure for 20 minutes to 100 µmol/L
ATP and 2 µmol/L ionomycin (Iono), respectively. Stimulated NO
synthase activity is presented as percentage of basal activity.
L-Arginine uptake, estimated as the sum of
[3H]L-arginine and
[3H]L-citrulline radioactivity (250 000 to
300 000 cpm/mg protein), basal NO synthase activity, and protein
content of each well were not different between untreated and
carbenoxolone-treated cells. For each cell line, data were
analyzed by one-way ANOVA.
|
|
Effects on Endothelial Ultrastructure
Although endothelial cell structure and NO
synthase activity were normal in culture, endothelial
integrity in rings exposed to carbenoxolone ex vivo may have been
compromised. Aortic rings were fixed after various treatments and time
periods ex vivo, and vascular ultrastructure was examined with electron
microscopy. Rings fixed immediately after isolation (Fig 8A
) revealed intact endothelial lining
separated from smooth muscle cells and connective tissues by a
continuous basal lamina. Endothelial cells were
characterized by numerous micropinocytotic vesicles at the
plasmalemma surfaces and intact intercellular junctional
complexes. Rings treated with 100 µmol/L carbenoxolone for 24 hours
ex vivo in aerated buffer (Fig 8B
) exhibited focal sites of
endothelial damage, whereas the smooth muscle
components of the rings appeared unaffected.
Endothelial cells frequently revealed damaged
mitochondrial profiles, multivesicular bodies, and secondary
lysosomes. To explore the possibility that the prolonged
aeration, independent of carbenoxolone, might have compromised
endothelial structural integrity, we examined rings
after 24 hours of ex vivo aeration in the absence of carbenoxolone (Fig 8C
). These rings, like those acutely fixed (Fig 8A
), revealed intact
endothelial cell lining with normal intercellular
complexes and numerous pinocytotic vesicles. Smooth muscle components
were also normal.

View larger version (166K):
[in this window]
[in a new window]
|
Figure 8. Effects of 24-hour carbenoxolone exposure on aortic
ring ultrastructure. Rings fixed immediately after isolation (A)
exhibit a continuous layer of endothelial cells
separated from normal smooth muscle cells by basal lamina (left).
Endothelial cells (right) contain numerous
plasmalemmal vesicles (arrowhead) and intercellular
junctional complexes (arrow). Rings treated with 100 µmol/L
carbenoxolone in aerated buffer for 24 hours (B) exhibit focal sites of
endothelial damage (left). Endothelial
cells (right) contain multivesicular bodies (arrow) and damaged
mitochondrial profiles (arrowhead). Rings in aerated buffer for 24
hours without carbenoxolone (C) display normal
endothelial and vascular smooth muscle morphology
(left). Endothelial cells (right) reveal the normal
complement of plasmalemmal vesicles (arrowhead) and
intercellular junctions (arrow). Bar=1 µm; bl indicates basal lamina;
e, endothelium; and sm, smooth muscle. This study was
performed two times.
|
|
 |
Discussion
|
|---|
These studies demonstrate that prolonged exposure of blood vessels
to
carbenoxolone at concentrations higher than 10 µmol/L
causes
selective endothelial damage. Enhanced responses to
vasoconstrictors,
maintained relaxation to sodium nitroprusside, and
normal vascular
smooth muscle ultrastructure suggest that vascular
smooth muscle
integrity and action were not affected by 24-hour
incubation
with carbenoxolone. Similarly, there are clinical
hypertensive
states in which endothelium is selectively
damaged but underlying
vascular smooth muscle is not, eg, eclampsia and
preeclampsia.
It is unclear why 100 µmol/L carbenoxolone was toxic to
endothelial
cells in rings but not to
endothelial cells in culture, but
this fortuitous set
of circumstances contributed to our eliminating
NO synthase as the
target of carbenoxolone action. Structural
or functional changes in
endothelial cells resulting from multiple
passage in
culture or different attachment surfaces for
endothelial
cells in culture and for
endothelium in vivo may explain why
cultured cells were
more resistant to the deleterious effects
of carbenoxolone.
Maintained endothelium-dependent relaxation
and
normal endothelial ultrastructural appearance in rings
aerated
ex vivo for 24 hours in the absence of carbenoxolone imply that
the
ex vivo conditions themselves were not toxic to the
endothelium.
A number of reports describe toxicities of
carbenoxolone,
15 and a repeated theme with this drug is
loss of cell membrane
integrity as evidenced by acute renal tubular
necrosis in patients
on oral carbenoxolone therapy
16 17
and membrane leakage after
ex vivo exposure of
leukocytes,
18 hepatic lysosomes,
19
and
artificial lecithin:cholesterol
liposomes
20 to 10 to 100 µmol/L
carbenoxolone. It should
be noted that the concentration range
of carbenoxolone (10 to 100
µmol/L) that caused membrane
leakage in those studies is identical to
the carbenoxolone concentration
range that eliminated
endothelium-dependent relaxation in the
present
study.
The mechanisms of endothelial damage from carbenoxolone
are unknown. A single study has suggested that short-term exposure
to 100 to 300 µmol/L carbenoxolone causes release of NO from the
endothelium.10 It is possible that this
release was the earliest manifestation of carbenoxolone-mediated
endothelium toxicity, with leak of NO to the underlying
vascular smooth muscle. More prolonged carbenoxolone exposure, as in
the present study, might have resulted in enough
endothelial damage to deplete existing NO and eliminate
the ability of the endothelium to form additional NO.
Partial restoration of endothelium-dependent
relaxation with an excess of L-arginine (Fig 6
) suggests
that L-arginine depletion is playing a role in the loss of
endothelium-dependent relaxation upon carbenoxolone
exposure and that endothelial damage is not total.
In a study similar to the present study, vascular segments were
exposed to carbenoxolone ex vivo for assessment of the direct effects
of carbenoxolone on vascular reactivity.11 However,
findings similar to ours were not reported because of the following
methodological differences: rings were mechanically
deendothelialized before carbenoxolone
exposure; carbenoxolone exposure was only 2 to 5 hours; and the
carbenoxolone concentration used was 10 µmol/L, which is at the
threshold for the effects observed in the present study.
The ability of carbenoxolone but not related compounds to disrupt
endothelial function is of interest. It has been
suggested that carbenoxolone and other triterpenoids act as
corticosteroids.21 22 However, in our
studies, neither the mineralocorticoid deoxycorticosterone acetate nor
the triterpenoid glycyrrhizic acid significantly reduced
endothelium-dependent relaxation as carbenoxolone
did. Although these results appear to imply a degree of specificity for
carbenoxolone among the steroids in inhibiting
endothelium-dependent relaxation, cautions in
interpretation are necessary. Just as glycyrrhetinic acid is a far
stronger inhibitor of 11ß-hydroxysteroid
dehydrogenase than glycyrrhizic acid,23 glycyrrhetinic
acid may have been as efficacious as carbenoxolone in inhibiting
endothelium-dependent relaxation, whereas
glycyrrhizic acid was not. We could not test glycyrrhetinic acid in the
present study because of lack of solubility. In addition, we did
not examine endothelial ultrastructure after
deoxycorticosterone acetate or glycyrrhizic acid exposure as it was
after carbenoxolone exposure.
A secondary result of the present study is that contractions to Ang
II are enhanced by reductions in NO production, whether this
reduction is mediated by the NO synthase inhibitor L-NAME
or by toxicity to endothelium. Similarly, Ang II
contractions were heightened in rabbit aortic segments after physical
denudation of endothelium or after treatment with NO
synthase inhibitors,24 and Ang II did not
cause contraction of isolated rabbit afferent arterioles unless an NO
synthase inhibitor was present.25 These
results suggest either that there is chronic vasodilator tone mediated
by the spontaneous elaboration of NO in the vasculature or that Ang II
stimulates the formation and/or release of NO. The latter possibility
is strengthened by the fact that endothelial NO
synthase is calcium dependent and that Ang II causes increases in
intracellular calcium concentration.26 Other studies have
demonstrated that NO synthase inhibition in intact animals causes
hypertension that is Ang II dependent.27 28
The role of endothelial toxicity and loss of NO
elaboration in carbenoxolone-mediated hypertension is unknown at
present. Carbenoxolone may foster hypertension by multiple
mechanisms. It is possible that carbenoxolone, over a wide range of
concentrations, inhibits 11ß-hydroxysteroid dehydrogenase in
vascular smooth muscle, increases intracellular glucocorticoid
concentrations, and accentuates vasoconstrictor action. These
corticosteroids may be formed locally in the blood
vessel29 30 31 or may reach the vascular smooth muscle after
synthesis in the adrenal cortex. At concentrations of carbenoxolone
higher than 10 µmol/L, however, loss of endothelial
cell integrity and endothelium-dependent relaxation
may also contribute to increased vascular tone. If incubation of rings
with glucocorticoid (eg, corticosterone) and carbenoxolone together
were to enhance vasoconstrictor action more than incubation with
carbenoxolone alone, support for both endothelial and
smooth muscle loci of carbenoxolone action would be forthcoming.
Although cultured cells and isolated rings are extremely useful assay
systems because systemic influences can be eliminated, differences in
these isolated preparations from blood vessels in the in vivo setting
necessitate repetition of the present study after treatment of
intact animals with carbenoxolone. We will perform these studies in the
near future. Carbenoxolone concentrations that result in loss of
endothelium-dependent relaxation (Fig 5
) are
relevant to blood concentrations of carbenoxolone attained in vivo. One
to 2 hours after oral ingestion of 100 mg carbenoxolone in humans, a
blood level of 20 to 30 µmol/L is attained.32 Greater
and more sustained blood levels might result from ingestion of a higher
dose or during clinical states in which protein binding or
metabolism of carbenoxolone are reduced (hypoproteinemia,
hepatic dysfunction).
 |
Acknowledgments
|
|---|
Support for this research was obtained from a Grant-in-Aid
from
the American Heart Association, South Carolina Affiliate, and
a
Biomedical Research Grant from the Medical University of South
Carolina.
Ultrastructural data were obtained with the Medical
University
of South Carolina Electron Microscopy University Shared
Instrument
Resource. The authors thank Jana J. Fine and Gregory D.
Miller
for expert technical assistance.
 |
Footnotes
|
|---|
Reprint requests to Michael E. Ullian, Medical University of
South Carolina, Division of Nephrology, Clinical Sciences Building
829, 171 Ashley Ave, Charleston, SC 29425. E-mail michael_ ullian@smtpgw-musc.edu.
Received October 10, 1995;
first decision November 14, 1995;
accepted February 12, 1996.
 |
References
|
|---|
-
Funder JW, Pearce PT, Smith R, Smith AI.
Mineralocorticoid action: target tissue specificity is enzyme, not
receptor, mediated. Science. 1988;242:583-585. [Abstract/Free Full Text]
-
Funder JW, Pearce PT, Smith R, Campbell J.
Vascular type I aldosterone binding sites are
physiological mineralocorticoid receptors.
Endocrinology. 1989;125:2224-2226. [Abstract]
-
Walker BR, Yau JL, Brett LP, Seckl JR, Monde C,
Williams BC, Edwards CRW. 11ß-Hydroxysteroid dehydrogenase in
vascular smooth muscle and heart: implications for
cardiovascular responses to glucocorticoids.
Endocrinology. 1991;129:3305-3312. [Abstract]
-
Handa M, Kondo K, Suzuki H, Saruta T.
Dexamethasone hypertension in rats: role of
prostaglandins and pressor sensitivity to
norepinephrine. Hypertension. 1984;6:236-241. [Abstract/Free Full Text]
-
Grunfeld JP, Eloy L. Glucocorticoids modulate
vascular reactivity in the rat.
Hypertension. 1987;10:608-618. [Abstract/Free Full Text]
-
Yagil Y, Krakoff LR. The differential effect of
aldosterone and dexamethasone on pressor
responses in adrenalectomized rats.
Hypertension. 1988;11:174-178. [Abstract/Free Full Text]
-
Pirpiris M, Sudhir K, Yeung S, Jennings G, Whitworth
JA. Pressor responsiveness in
corticosteroid-induced hypertension in
humans. Hypertension. 1992;19:567-574. [Abstract/Free Full Text]
-
Teelucksing S, Mackie ADR, Burt D, McIntyre MA, Brett
L, Edwards CRW. Potentiation of hydrocortisone activity in skin
by glycyrrhetinic. Lancet. 1990;335:1060-1063. [Medline]
[Order article via Infotrieve]
-
Walker BR, Connacher AA, Webb, DJ, Edwards CRW.
Glucocorticoids and blood pressure: a role for the cortisol/cortisone
shuttle in the control of vascular tone in man. Clin
Sci. 1992;83:171-178. [Medline]
[Order article via Infotrieve]
-
Dembinska-Kiec A, Pallapies D, Simmet T, Peskar BM,
Peskar BA. Effect of carbenoxolone on the biological activity of
nitric oxide: relation to gastroprotection. Br J
Pharmacol. 1991;104:811-816. [Medline]
[Order article via Infotrieve]
-
Walker BR, Sang KS, Williams BC, Edwards CRW.
Direct and indirect effects of carbenoxolone on responses to
glucocorticoids and noradrenaline in rat aorta.
J Hypertens. 1994;12:33-39. [Medline]
[Order article via Infotrieve]
-
Todd ME, Friedman SM. The rat-tail artery
maintained in culture: an experimental model. In
Vitro. 1978;14:757-770. [Medline]
[Order article via Infotrieve]
-
Merrilees MJ, Scott L. Organ culture of rat
carotid artery: maintenance of morphological characteristics
and of pattern of matrix synthesis. In Vitro. 1982;18:900-910. [Medline]
[Order article via Infotrieve]
-
Lin L, Nasjletti A. Role of
endothelium-derived prostanoid in
angiotensin-induced vasoconstriction.
Hypertension. 1991;18:158-164. [Abstract/Free Full Text]
-
Monder C. Corticosteroids,
kidneys, sweet roots, and dirty drugs. Mol Cell
Endocrinol. 1991;78:C95-C98. [Medline]
[Order article via Infotrieve]
-
Descamps C, Vandenbroucke JM, van Ypersele de Strihou
C. Rhabdomyolysis and acute tubular necrosis associated with
carbenoxolone and diuretic treatment. Br Med
J. 1977;1:277.
-
Hurley B. Acute renal failure associated with
carbenoxolone treatment. Br Med J. 1977;1:1472.
-
Baron DN, Green RJ. Action of compounds with
effective in vivo mineralocorticoid activity on ion transport in
leucocytes. Br J Clin Pharmacol. 1986;21:27-34. [Medline]
[Order article via Infotrieve]
-
Symons AM, Johnston B, Parke D. The effect of
sodium carbenoxolone on lysosomal enzyme release. Biochem
Pharmacol. 1978;27:2461-2463. [Medline]
[Order article via Infotrieve]
-
Symons AM. The effect of sodium carbenoxolone on
the permeability of phosphatidylcholine and
phosphatidylcholine:cholesterol liposomes.
Biochem Pharmacol. 1976;25:1545-1547. [Medline]
[Order article via Infotrieve]
-
Porter GA, Rhodes C, Sacra P. Comparative
studies on the mineralocorticoid action of aldosterone and
carbenoxolone sodium in the adrenalectomized rat.
Pharmacology. 1974;12:224-229. [Medline]
[Order article via Infotrieve]
-
Takeda R, Miyamori I, Soma R, Matsubara T, Ikeda
M. Glycyrrhizic acid and its hydrolysate as mineralocorticoid
agonist. J Steroid Biochem. 1987;27:845-849. [Medline]
[Order article via Infotrieve]
-
Monder C, Steware PM, Lakshmi V, Valentino R, Burt D,
Edwards CRW. Licorice inhibits corticosteroid
11ß-dehydrogenase of rat kidney and liver: in vivo and in vitro
studies. Endocrinology. 1989;125:1046-1053. [Abstract]
-
Zhang J, van Meel JCA, Pfaffendorf M, Zhang J, van
Zwieten PA. Endothelium-dependent, nitric
oxide-mediated inhibition of angiotensin II-induced
contractions in rabbit aorta. Eur J Pharmacol. 1994;262:247-253. [Medline]
[Order article via Infotrieve]
-
Yoshida H, Tamaki T, Aki Y, Kimura S, Takenaka I, Abe
Y. Effects of angiotensin II on isolated rabbit
afferent arterioles. Jpn J Pharmacol. 1994;66:457-464. [Medline]
[Order article via Infotrieve]
-
Emori T, Hirata Y, Ohta K, Kanno K, Eguchi S, Imai T,
Shichiri M, Marumo F. Cellular mechanism of endothelin-1 release
by angiotensin and vasopressin.
Hypertension. 1991;18:165-170. [Abstract/Free Full Text]
-
Ribeiro MO, Antunes E, de Nucci G, Lovisolo SM, Zatz
R. Chronic inhibition of nitric oxide synthesis: a new model of
arterial hypertension.
Hypertension. 1992;20:298-303. [Abstract/Free Full Text]
-
Pollack DM, Polakowski JS, Divish BJ, Opgenorth
TJ. Angiotensin blockade reverses hypertension
during long-term nitric oxide synthase inhibition.
Hypertension. 1993;21:660-666. [Abstract/Free Full Text]
-
Hatakeyama H, Miyamori I, Fujit T, Takeda Y, Takeda R,
Yamamoto H. Vascular aldosterone: biosynthesis and a
link to angiotensin II-induced hypertrophy of
vascular smooth muscle cells. J Biol Chem. 1994;269:24316-24320. [Abstract/Free Full Text]
-
Takeda Y, Miyamori I, Yoneda T, Iki K, Hatakeyama H,
Blair IA, Hsieh FY, Takeda R. Production of
aldosterone in isolated rat blood vessels.
Hypertension. 1994;25:170-173. [Abstract/Free Full Text]
-
Takeda Y, Miyamori I, Yoneda T, Iki K, Hatakeyama H,
Blair IA, Hsieh FY, Takeda R. Synthesis of corticosterone in the
vascular wall. Endocrinology. 1994;135:2283-2286. [Abstract]
-
Parke DV. The pharmacology and biochemistry of
carbenoxolone. In: Beck IT, ed. North American Symposium
on Carbenoxolone. New York, NY: American Elsevier Publishing Co,
Inc, Excerpta Medical; 1976.
This article has been cited by other articles:

|
 |

|
 |
 
V. M. Vehaskari and L. L. Woods
Prenatal Programming of Hypertension: Lessons from Experimental Models
J. Am. Soc. Nephrol.,
September 1, 2005;
16(9):
2545 - 2556.
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
J. W. Tomlinson, E. A. Walker, I. J. Bujalska, N. Draper, G. G. Lavery, M. S. Cooper, M. Hewison, and P. M. Stewart
11{beta}-Hydroxysteroid Dehydrogenase Type 1: A Tissue-Specific Regulator of Glucocorticoid Response
Endocr. Rev.,
October 1, 2004;
25(5):
831 - 866.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
J. R. de Groot, T. Veenstra, A. O. Verkerk, R. Wilders, J. P.P. Smits, F. J.G. Wilms-Schopman, R. F. Wiegerinck, J. Bourier, C. N.W. Belterman, R. Coronel, et al.
Conduction slowing by the gap junctional uncoupler carbenoxolone
Cardiovasc Res,
November 1, 2003;
60(2):
288 - 297.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
M. DODIC, V. HANTZIS, J. DUNCAN, S. REES, I. KOUKOULAS, K. JOHNSON, E. M. WINTOUR, and K. MORITZ
Programming effects of short prenatal exposure to cortisol
FASEB J,
July 1, 2002;
16(9):
1017 - 1026.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
P. W.F. Hadoke, C. Christy, Y. V. Kotelevtsev, B. C. Williams, C. J. Kenyon, J. R. Seckl, J. J. Mullins, and B. R. Walker
Endothelial Cell Dysfunction in Mice After Transgenic Knockout of Type 2, but Not Type 1, 11{beta}-Hydroxysteroid Dehydrogenase
Circulation,
December 4, 2001;
104(23):
2832 - 2837.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
E. P. Gomez-Sanchez and C. E. Gomez-Sanchez
Maternal Hypertension and Progeny Blood Pressure : Role of Aldosterone and 11ß-HSD
Hypertension,
June 1, 1999;
33(6):
1369 - 1373.
[Abstract]
[Full Text]
[PDF]
|
 |
|