Ginsenosides protect pulmonary vascular endothelium against free radical — induced injury

Vol. 189, No. 2, 1992 December 15, 1992

GINSENOSIDES

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMM;NICATIONS Pages 670-676

PROTECT PULMONARY VASCULAR ENDOTHELIUM FREE RADICAL - INDUCED INJURY 1

AGAlNST

Hyeyoung Kim (Jun) 2, Xiu Chen 3 and C. Norman Gillis 4 Departments of Anesthesiology and Pharmacology, Yale University School of Medicine, New Haven, CT 06510 Received

October

22,

1992

SUMMARY. We studied the actions of saponin (ginsenosides) from Panax ginseng on free radical-induced pulmonary endothelial injury which is manifest as reversal of the normal vasodilator response to acetylcholine in perfused, vasoconstricted lungs. 50 or 200 pg/ml ginsenosides prevented this injury response and also reduced the pulmonary edema which follows free radical injury but did not alter the normal ACh - induced vasodilation in intact lungs. In control perfused lungs preconstricted with U46619, the ginsenoside mixture or purified ginsenosides Rbl and Rgl caused vasodilatation. This effect was eliminated by 100 pM nitrol-arginine, an inhibitor of nitric oxide synthase. In cultured bovine aortic endothelial cells, ginsenosides (10 pg/ml) stimulated the conversion of [14C]-L-arginine to [14C]-L-citrulline. These data indicate that GS may cause vasorelaxation and prevent manifestations of oxygen free radical injury by promoting release of nitric oxide. e 1992 Academic press, hc.

We reported earlier that GS (a mixture of saponins from Panax ginseng) reversed NE or PGF2,induced contraction of rabbit pulmonary and intrapulmonary arteries (1). This effect qualitatively which the cholinergic

resembled that of ACh in perfused, pre-constricted agonist acts by generating

rabbit lungs in

NO from the vascular endothelium

(2,3).

‘Supported by grants from the National Heart, Lung and Blood Institute (HL-13315 and HL-40863) and from the Miles Institute, West Haven, Connecticut. ‘Current address: Laboratory of Biochemical Pharmacology, Tobacco Research tnstitute, Taejon 305345, South Korea. 3Current address: Department of Pharmacology, Changsa, People’s Republic of China. 4 To whom correspondence

Korea Ginseng

and

Hunan Medical University,

should be addressed.

Abbreviations A = aorta; ACh = Acetylcholine HCI; BAEC = bovine aortic endothelial cells; DMEM = Dulbeco’s modified Eagle medium; EDRF = endothelium-derived relaxing factor; GS = Ginsenoside; NE = norepinephrine; NO = nitric oxide; N-arg = N-o-Nitro-L-arginine; PBS = phosphate buffered saline; PA = pulmonary artery; SOD = superoxide dismutase; U46619 = 9,11 -dideoxyB,,l I,-methyl-anoepoxy prostaglandin F2m 0006-291X/92 Copyright All rights

$4.00

0 I992 by Academic Press, of reproduction in any form

Inc. reserved.

670

Vol.

189, No. 2, 1992

We therefore similar

BIOCHEMICAL

determined

actions

AND BIOPHYSICAL

whether GS (and the individual ginsenosides

in the perfused rabbit lung and in endothelial

Our experiments

RESEARCH COMMUNICATIONS

Rbl and Rgl) had

cells in culture.

with a perfused rabbit lung model indicated that GS, like ACh, could

dilate pre-constricted lungs. Prior studies in this laboratory (3) showed that the vasodilator response to ACh is converted to vasoconstriction following brief electrolysis of the inflowing perfusion

medium.

reflects oxygen free radical-mediated release of NO by ACh. serotonin protect

uptake, against

Consistent

an established myocardial

reducing the damaging

which disrupts the normal

with this proposal, electrolytic reflection

injury following several antioxidants

injury (4) we also determined properties (56)

electrolysis-induced

injury to the endothelium, of pulmonary

injury decreased

endothelial

ischemia/reperfusion

effects of free radicals and lipid peroxidation

this reason and because antioxidant

We proposed (3,4) that such reversal of ACh action

injury (2).

(7)

GS

reportedly

by

in the heart.

For

prevent electrolysis-induced

whether the ginsenoside

lung

endothelial

Rbl, which is reported

might also protect the pulmonary vasculature

to have

against

injury.

MATERIALS

and METHODS

Chemicals and Druas: Drugs and chemicals were from the following sources: U46619, Cayman Chemicals (Ann Arbor, MI); [14C]-L-arginine (57.8 mCi/mmol) and [14C]-Lcitrulline (53.7 mCi/mmol), NENlDupont (Willimington, DE); Dowex AG 50 WX-8, Bio-Rad Laboratories (Richmond,CA); Nw-nitro-L-arginine (N-arg) and all other chemicals, Sigma Chemical Co. (St.Louis, MO). Ginsenosides (GS) were extracted from Korean red ginseng (Panax ginseng C.A.Meyer) as described by Shibata et al (8). Ginsenosides Rbl and Rgl were purified from GS by the method of Ando et al (6). Molecular weights of Rbl and Rgl are 1109 and 801 respectively. Luno perfusion: Lungs were perfused as described earlier (2,4). After an initial period of 30 minutes of non-recirculating (SP) perfusion to remove blood (Fig. 1) recirculation was started and vascular reactivity was determined by 1) addition of U46619 (5-30 nM) to the recirculating reservoir to cause an increase in perfusion pressure to 30 - 35 mm Hg., and 2) cumulative addition of ACh (50-200 nM final concentration) to produce vasodilation. Drugs were then flushed from the vascular space by 10 min of SP perfusion. After this first “challenge”, lungs were again constricted with U46619 and the direct vasodilator effect of GS, Rbl or Rgl on preconstricted lungs was assessed by adding each (50-200 PM) instead of ACh. Effects of ginsenosides pretreatment on ACh-vasodilatation in control experiments or after ES were studied by adding the saponins (20-200 PM) to the recirculation medium for 10 minutes (Fig. 1). Following this pretreatment, ES was applied (2,4) by passing a 20 mA current for 2 minutes between platinum electrodes sealed through the perfusion tubing proximal to the lung. Next, recirculating perfusion was again started with fresh Krebs’ medium and challenge with U46619 and ACh was repeated as above. In some experiments, N-Arg (100 PM) was recirculated through the lungs for 5 min to block the release of NO and challenge with U46619 and ACh was repeated. After each experiment, lungs were weighed before and after drying to constant weight in an oven at 550 C. Wet to dry weight ratio is an index of edema production. 671

Vol.

189, No. 2, 1992

RECIRC SP : (Single PassI: 30’

I

BIOCHEMICAL

:

RECIRC SP

;

10’

I

:

:

RECIRC

I

RESEARCH COMMUNICATIONS

:

ISP :

: 10’

AND BIOPHYSICAL

: 10’

:2’:

10’

:

02

I

ELECTROLYSIS

0

5

10

15

U46619

20

25

30

(nM1

Fiaure

1.

Proctocol used to evaluate the pulmonary vascular response to ACh before and after electrolysis and the effects of ginsenosides (GS). U4 refers to infusion of U46619 (see “Methods and Materials” for details).

Fiaure

2.

Effect of GS on U46619 ug/ml), vasoconstrictor reduced (P < 0.01).

- induced response

vasoconstriction. In the presence of GS (20 to 10 - 30 nM U46619 was significantly

Bovine aortic endothelial cell culture: BAEC, isolated from segments of bovine aortae (9) were grown in Dulbeco’s modified Eagle medium (DMEM) containing 0.48 mM Larginine, 2 mM L-glutamine, 1 mM sodium pyruvate, 1 mM Hepes and 10 % (vol/vol) fetal calf serum. Endothelial cells were identified by their cobblestone - like appearance when seeded onto 6 well plates and by positive immunostaining for angiotensin converting enzyme and factor VII I. Assav of NO svnthase: NO synthase activity was measured by monitoring the conversion of [14C]-L-arginine to [14C]-L-citrulline (9) in confluent BAEC which were arginine deprived for 30 min before addition of [14C]-L-arginine (final concentration 850 nM, 0.05 $J/ml). 1 min later, GS, Rbl or Rgl were added (final concentration 10 PM) and incubation continued for 15 min at 370 C. Cells were then washed with PBS, harvested in ice-cold Hepes-EDTA buffer (20 mM12 mM, pH 6.0), sonicated (10 sec., three times) and centrifuged (10,000 x g, for 15 min at 4 oC ) and the supernatant applied to a 1 -ml Dowex 50 cation exchange column. [14C]-L-citrulline in the effluent and a subsequent 1 ml water wash was quantified by liquid scintillation spectrometry. The sole radioactive component was verified as [14C]-L-citrulline by thin layer chromatography (data not shown). Percent conversion was calculated as dpm of [14C]-L-citrullineldpm [14C]-Larginine x 100. Statistics: determined considered

All values represent mean + S.E.M. by one - way ANOVA with Newman significantly different if P c 0.05.

Differences among - Keuls test (11).

groups Values

were were

RESULTS Effect of GS on basal vascular tone, U46619-induced vasoconstriction and ACh vasodilatation: Recirculation of 20 pg/ml GS did not alter basal vascular tone (i.e. perfusion pressure), although we consistently noted that more U46619 was required to 672

Vol.

189, No. 2, 1992

BIOCHEMICAL

produce a 30 - 35 mm Hg change observations

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

in perfusion pressure (Fig. 2) thus confirming

(1) which pointed to inhibition of smooth muscle contractility

prior

by GS.

We confirmed our earlier Effect of GS on ACh-induced vasoconstriction after ES: observations (4) that after ES, ACh consistently produced vasoconstriction (Fig. 3). Treatment

of lungs with 50 pg/ml or 200 pg/ml GS before and during ES prevented

induced vasoconstriction, less

protective

vasoconstriction

which normally followed the injury. Although 20 pg/ml GS had

effect

than

higher

concentrations,

it did

eliminate

ACh-induced

after ES (Fig. 3).

Effect of GS on ES-induced

pulmonarv edema:

Control wet/dry weight ratios for lungs in

this series of experiments

were 7.5+0.1; these values are somewhat

data in other experiments

carried out in this laboratory

ES had significantly experimental

ACh-

(2).

higher than control

However lungs exposed to

higher ratios (1 l.O+l.O; P
controls in the presence of 50 pg/ml (7.5+0.3) or 200 pg/ml GS (7.620.4);

20 pglml GS did not protect against edema (wet/dry weights were 8.721.4).

GS in the

absence of ES did not cause pulmonary edema. Vasodilator

actions

significantly

dilated lungs preconstricted

of ACh and GS before

and after

N-arq:

by U46619 (PcO.05).

GS, Rgl

and Rbl

200 vg/ml of GS or 200 p

M of either Rgl or Rbl caused 39 %, 22 % or 15 % reduction,

respectively,

in the peak

U44619-induced pressure change (Fig. 4). After blocking the release of NO by N-arg (100 PM), a relatively lower concentration (10 nM ~30 nM) of U46619 was required to

ES ONLY

%

Q)

t 0)

8 2

146

120

1

(N = 3 per Group

)

z P

105 -

z c s ‘3 ‘=3 ,?

03

Rbl

ES+GS 20pg/ml

f z

Ml 63-

ES+GS 50pgh

7 20 0

/ 50

100

150

GS

ES+GS 200pg/ml CONTROL I 200 250

ACh (PM)

0

U46619

4

CT

0

50 100 Ginsenoride

150 (pM1

200

U46619

Fiaure3.

Preservation of the ACh vasodilator response by (50 pg/ml or 200 pglml Notice that 20 pglml did not preserve the ACh vasodilatation.

Fioure 4.

Direct vasodilator actions of ginsenosides. Notice that GS (50-200 pg/ml) as well as ginsenosides Rgl and Rbl (50-200 PM), significantly vasodilated lungs preconstricted by U46619. GS, Rgl or Rbl reduced peak perfusion pressure during U46619 infusion by 39 %, 26 % or 15 % respectively (P c 0.05). 673

GS.

250

Vol.

189, No. 2, 1992

BIOCHEMICAL

AND BIOPHYSICAL

it 600 El W 0

50 25

ACh (0.05f.M) ACh (0.1j.M) ACh (O.ZpMl

g G

__@

“0 400 T 4 300

-

o--

--

-25

- llv

-50

-

--we

1 N=4 ACh

I GS

CONTROL

[

N=3

ACh

)

AFTER

I

Rbl

Rgl

200 100

2c

GS

vs CONTROL vs CONTROL

500

g ‘2 .z

-75

C5 >

or GS (50 pg/ml) or GS (100 pg/ml) or GS (200 ~g/mlJ

P
# P-co.01

0 CONTROL

-

TREATMENT

6 0 actions

N-ARG

RESEARCH COMMUNICATIONS

Fiaure 5.

Changes in vasodilator of ACh and GS before and after N-Arg. Vasodilator responses to ACh (n = 4) and GS (n = 3) were abolished by 100 TV y N-Arg. Notice that after N-Arg, ACh caused slight vasoconstriction.

Fiaure 6.

GS (IO pg/ml) and Rgl (IOpM) significantly (P c 0.01) increased both uptake of [14C]-L-arginine and conversion to [14C]-L-citrulline in citrulline while Rbl lacked this effect (n = 12).

raise perfusion

pressure to 30-35 mm Hg . In addition, the vasodilator

ACh and GS were completely caused slight vasoconstriction muscarinic

receptors

vasodilator

action of GS.

citrulline

production

significantly

enhanced

eliminated (Fig. 5). Notice (Fig. 5) that 200 nM ACh after N-arg, possibly because it acted directly on

in the vascular

Effect of GS on Arainine

smooth muscle (2,4).

to citrulline conversion

from

[14C]-L-arginine, In addition,

as an

index

of NO

the

production,

was

(20.3% or 22 % vs 14.9 % in

Rgl also significantly

arginine transport into cells. This enhanced the presence of 0.1 pM bradykinin

N-arg also eliminated

in BAEC: As shown in Fig. 6, [14C]-L-

by 10 pg/ml GS or 10 ~.IM Rgl

controls) but not by Rbl.

effects of both

increased total

[14C]-L-

uptake and conversion was also observed

in

(data not shown).

DISCUSSION The GS we used represents produce

a complex mixture of individual

a range of effects on the cardiovascular

Each of the ginsenosides

and central

ginsenosides,

nervous

of Panax Ginseng may have qualitatively

systems

which (12).

and quantitatively

different effects on the cardiovascular system producing a complicated range of actions (13) described as “adaptogenic” effects, which are reported to provide non-specific resistance to circulatory stress (12). The complex nature of the mixture probably explains the fact that, in our experiments, GS relaxed pulmonary vessels but constricted 674

Vol.

189,

No. 2, 1992

others, including

BIOCHEMICAL

AND BIOPHYSICAL

renal vessels (1) and that

study significantly

increased

NO production

RESEARCH COMMUNICATIONS

Rgl and GS (but not Rbl)

in the present

as reflected

of arginine

by conversion

to

citrulline in BAEC (Fig. 6). Our experiments

establish

activity of the thromboxane tone per se.

that GS significantly

analog U46619 (Fig 2.)

This is qualitatively

diminished

the vasoconstrictor

but did not directly alter vascular

similar to effects on

pulmonary

arterial segments

yi& described previously (1). Also, GS and Rgl significantly vasodilated lungs. Both effects could be due to release of an endogenous GS.

The fact that both actions were inhibited

synthase

vasodilator

pre-constricted substance

by N-arg (Fig. 5), an inhibitor

(IO), suggests that NO itself could be the substance

involved.

h by

of NO

Experiments

with BAEC in culture support this possibility; we found (Fig. 6) that GS and Rgl (but not Rbl) significantly

enhanced

conversion

of arginine to citrulline, which is a measure

of

concomitant NO production. These observations are also consistent with the fact (Fig. 4) that Rbl had less inhibitory action on U46619-vasoconstriction than either GS or Rgl. If some ginsenosides cholinergic

agonists.

vasoconstriction, vasodilatation.

enhance NO production,

However,

this seems unlikely

while ACh in the presence We therefore

they might do so by direct action as because,

of 50 or 200 pg

after ES, ACh caused of GS per ml caused

propose that this protective effect is due to the production

of endogenous NO at a site distal to the cholinergic receptor. We reported earlier (3) that ES injury could be prevented by antioxidants including SOD, catalase or salicylate (a It is thus of interest that NO acts as an antioxidant and also hydroxyl radical scavenger). decreases

lipid peroxidation

in vitro (14). Perhaps, therefore,

as a direct antioxidant

action of the ginsenosides

against lung endothelial

injury caused by electrolysis.

(5)

NO released by GS as well

might contribute

to protection

ACKNOWLEDGMENT We gratefully acknowledge the supply of GS and purified Rbl and Rgl Hoon Park, Korea Ginseng and Tobacco Research Institute.

by Dr.

REFERENCES

1 2. 3. 4. 5 6. 7.

Chen,X., Gillis,C.N. and Moalli,R. (1984) Br. J. Pharmacol. 82,485-491. Gillis,C.N., Chen,X. and Merker,M.M. (1992) J. Pharmacol. Exp. Ther., 262, 212-216. Cherry, P.D. and Gillis, C.N. (1987) J. Pharmacol. Exp. Ther., 241, 5165205. Chen,X. and Gillis,C.N. (1991) J. Appl. Physiol., 71,821-825. Zhang,J.T., Qu,Z.W., Liu,Y. and Deng,H.L. (1990) Chinese Med. J., 103, 932-938. Ando,T., Tanaka,O. and Shibata,S. (1971) Syoyakugaku Zasshi. 25, 28-34. Yang, B.C. and Chen, X. (1989) Asia Pacific J. Pharmacol. 4, 265-272. 675

Vol.

8. 9. IO. 11. 12. 13. 14.

189, No. 2, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

Shibata,S., Tanaka,O., Soma,K., Ilda,Y., Ando,T. and Namamura,H. (1969) Tetrahed. Letts., 3,207-213. Boulanger, C., Schini, V.B., Moncada, S. and Vanhoutte, D.M. (1990) Br. J. Pharmacol., 101,152-l 56. Bogle,R.G., Coade,S.B., Moncada,S., Pearson,J.D. and Mann,G.E. (1991) Biochem. Biophys. Res. Comm., 180,926-932. Zar,J.H. (1984) Biostatistical Analysis, 2nd Edition, Prentice-Hall, Englewood Cliffs, N.J. Brekhmaql.1. and Dardymov,I.V. (1969) Ann. Rev. Pharmacol., 9, 419-430. Kaku,T., Miyata,T., Usuno,T., Sako,l. and Kinoshita,A. (1975) Arzneim Forsch., 25, 539- 547 Kanner, J. (1990) Free Radical Biol. Med. 9 (Suppl I), 15.

676