ACE-inhibition induces NO-formation in cultured bovine endothelial cells and protects isolated ischemic rat hearts

,J h,lol

Cell Cardiol

24. 909-919

( 1992)

ACE-inhibition Induces NO-formation Endothelial Cells and Protects Isolated Wolfgang

Linz,*

Gabriele

Wiemer

Hoechst AG, D-6230

in Cultured Bovine Ischemic Rat Hearts

and Bernward

Frankfurt/Main,

A. Schijlkens

Germay

(Received 3 December 1991, accepted in revised form

16 March

1992)

W. LINZ, G. WIEMER AND B. A. SCH~LKENS. ACE-inhibition Induces NO-Formation in Cultured Bo\inc Endothelial Cells and Protects Isolated Ischemic Rat Hearts. Journal ofMolecularandCe/lular Cardiology ( 1992, 24, 909-919. The role of NO-formation induced by accumulated endogenous bradykinin (BKJ via local ACE;inhibition with ramiprilat (RT) or by adding BK rxogenously was evaluated in cultured bovine aortic endothelial cells (BAEC) and in isolated rat hearts with post-ischaemic reperfusion injuries. Furthermore WC used the n-octyl-ester of ramipril (RA-octil) which was shown to have no ACE-inhibitory action. In BAEC, or addition of BK ( 1 x 10 ‘-1 x 10 “mol/l) stimulated the ACE-inhibition by RI‘ (1 x IO-“-l x 10 b moljl) formation of NO and prostacyclin (PGI,) as assessed by endothelial cyclic GMPand 6-keto-PGF,, formation. Cyclic GMP and PGI, synthesis was completely suppressed by the NO synthase inhibitor N”-nitro-L-argininr {L-NINA. 1 x 10 ’ mol/l) and by the B, kinin receptor antagonist HOE 140 i I x 10~ ’ mot/l). RA-octil (1 x IO ” I x 10 ‘mot/l) did not affect endothelial cyclic GMP production in BAEC. In isolated working rat hearts subjected to local ischemia with reperfusion both RT (1 X 10 ” moljl) and BK [ 1 X lo-” moljl) reduced the incidrncc and duration of ventricular fibrillation. In parallel myocardial function (left ventricular pressure, coronary flow) and metabolism (high energy rich phosphates) were improved showing a comparable fingerprint for RT and BK. Addition of L-NNA (1 x 10mb mol/l) or HOE 140 (1 X 10 ‘mol/l) abolished these protvcri\c effects of RT and BK. As in the BAEC studies RA-ortil was without beneficial effects on the isolated ischaemic rat heart. The findings on BAEC show that inhibition of ACE localized on the luminal side of the vascular endothelium results in increased synthesis of NO and prostacyclin by local accumulation of endotheliurnderived BK. Similar mechanisms may occur in the ischaemic rat heart leading to rardioprotcction. KEY WORDS: Nitric Bovine endothelial

oxide cells:

formation; Cardioprotection;

ACE-inhibition; Ramiprilat

Introduction It has been shown that endothelial cells are able to synthesize and release potent vasoconstrictor peptides such as angiotensins (Kifor and Dzau, 1987) and endothelin (Yanagisawa et al., 1988). Furthermore, there is some evidence that the potent stimulators of endothelial nitric oxide (NO) formation acetylcholine, ATP, and substance P are released from endothelial cells themselves (Milner et al., 1990; Kawashima et al., 1990) in this way establishing an effective paracrine dilator system. Stimulation of endothelial B,-kinin receptors by bradykinin (BK) increases the cytosolic Ca” concentration, which promotes synthesis of prostacyclin (PGI,) as well as NO as assessed by 6-keto-prostaglandin F,, (6-keto*Pleas? address 800320, U-6230 0022-2828/92/080909+

all correspondence Frankfurt am Main II

$08.00/O

to: Wolfgang 80, Germany.

Linz,

Bradykinin;

Isolated

rat heart:

Reperfusion

arrhythmias;

PGF,,) and endothelial cyclic GMP content, respectively, and is associated with relaxation of vascular smooth muscle (Wiemer tt al.. 1991; Liickhoff et al., 1988a). ACE-inhibition potentiates the vasodilation by BK because ACE, also named kininase II, is synthesized by endothelial cells and localized on the luminal surface, and degradates this peptide (Vanhoutte et al., 1989). However, the few studies on plasma kinin levels in hypertensive patients after ACE-inhibitor treatment present conflicting data on the contribution of kinin-mediated vasodilations in the hypotensive effects of ACE-inhibitors (Imura and Shimamoto, 1989; Biinner, 1988;. Thus, it has been proposed that locally-generated kinins in the vascular wall are responsiblr c/o Hoechst

AG.

SBU

Cardiovascular

(‘. 1992

Agents

Academic

H 821,

Press

P.O.B.

Limitrd

910

W. Linz

for dilator action of ACE-inhibitors rather than circulating kinins (Banner et al., 1990; Scicli et al., 1991). On the other hand, increased BK levels during ischaemia were found in coronary sinus blood after coronary occlusion in dogs (Hashimoto et al., 1977). In humans, kinin levels in peripheral blood were found to increase soon after myocardial infarction, which led Hashimoto et al. (1978) to suggest that kinins released in patients with infarction may have a compensatory cardioprotective effect. Th e potential sources of intravascular kinins have not yet been identified. The present experiments with endothelial cells and isolated rat hearts with post-ischaemic reperfusion injuries were performed to further elucidate these mechanisms considering a possible interaction of ACE-inhibition and endogenous BK metabolism.

Methods Endothelial

cell culture

Bovine aortic endothelial cells (BAEC) were isolated by digestion with dispase and cultured as previously described (Luckhoff et al., 1988b). The cells were seeded on 6-well plates (Nunc Intermed@, Wiesbaden, Germany) and grown to confluence. The medium used was Dulbecco’s modified Eagle’s/Ham’s F-l 2 medium (1: 1) containing 20% fetal calf serum. The culture medium was supplemented with penicillin ( 10 U/ml), streptomycin ( 10 pg/ml), L-glutamine ( 1 mmol/l), glutathione and L( +) ascorbic acid (each 5 mg/ml; Biotect@ protection medium). Measurement

of cyclic GMP

and 6-keto PGF,,

Primary cultures of endothelial cells were used. Following removal of the culture medium by aspiration, the monolayer was washed twice with 2 ml of HEPES/Tyrode’s (37°C). Thereafter solution the cells were preincubated for 15 min at 37°C with 3-isobutyl- 1-methyl-xanthine (IBMX, 1 x 1O-4 mol/l). After this time, drugs or solvents and superoxide dismutase (SOD, 3 x lo-’ mol/l) were added to the cells at the concentrations and times indicated in the results. At the appropriate time, the incubation

et al. medium was quickly removed from the monolayers and frozen at -20°C until it was assayed for its content of 6-keto-PGF,,, the stable hydrolysis product of PGI,, by a specific radioimmunoassay (Amersham Buchler, Braunschweig, Germany). Simultaneously, remaining cells were immediately extracted with 0.8 ml of 1 N ice-cold formic acid/acetone (v/v 15:85) and scraped off with a rubber policeman. The cell suspension was then sonicated for 10 s before being centrifuged for 10 min at 1000 g. Supernatants were lyophilized and resuspended in 0.25 ml sodium acetate buffer (0.05 mol/l, pH 6.2) for determination of cyclic GMP by radioimmunoassay (New England Nuclear, Dreieich, Germany). Cyclic GMP and 6-keto-PGF,, content were expressed as pmol and ng per mg protein respectively. The NO synthase inhibitor LNNA (Mulsch and Busse, 1990) and the B, kinin receptor antagonist HOE 140 (Wirth et al., 1991) were preincubated for 5 min in inhibitory concentrations of 1 x 10-s mol/l and 1 x IO-’ mol/l respectively. Isolated working

rat heart preparation

Isolated working heart preparations from Wistar rats of either sex weighing 280-300 g were used in all experiments, as described previously (Linz et al., 1986). Isolated working hearts were perfused via the aorta at a constant perfusion pressure of 65 mmHg, with a modified Krebs-Henseleit buffer of the following composition? (mM): NaCl, 113.8; NaHCO,, 22; KCl, 4.7; KHsPO,, 1.2; MgSO,, 1.1; CaCl,, 2.5; glucose, 11; Na-pyruvate, 2. The medium was gassed with 95% 0, plus 5% CO, at 38°C and pH 7.4. The perfusate did not recirculate and the hearts were not stimulated. After an equilibration period of 20 min (preischaemic period) acute regional myocardial ischaemia was produced by clamping the left coronary artery close to its origin for 15 min (ischaemic period). The clip was then reopened, and changes during reperfusion were monitored for 30 min (reperfusion period). In the left ventricle a silicone balloon closely fitting the ventricular cavity was placed and connected to an artificial “systemic” circulation (Linz et al., 1986). Left ventricular pressure was measured via a

Endothelial

NO-formation

Contributes

transducer (Statham P 23 Db), pressure which on differentiation yielded left ventricular dP/dt,,, and heart rate. Coronary flow was determined by an electromagnetic Bow probe in the aortic cannula. An epicardial electrocardiogram recording was obtained via two silver electrodes attached to the heart. All parameters were recorded (Brush 2600). For the determination of lactate release, lactate dehydrogenase and creatine kinase activities in the perfusate, samples were taken from the coronary effluent and analyzed spectrophotometrically (Linz et al., 1986). After the experiments, glycogen, lactate, ATP and creatine phosphate in the myocardial tissue were measured (Linz et al., 1986). The effects of ramiprilat (RT), the active moiety of ramipril [ 1 X lo-* mol/l), RA-octil, a derivative without inhibitory action on ACE I 1 X 10eh and 1 x 10m7mol/l), BK (1 X 10eq and 1 x IO-’ mol/I), B, kinin antagonist HOE 140 (1 x IO-” and 1 X IO-‘“mo!/l), and NCnitro-L-arginine (L-NNA, 1 x lo-” mol/l) alone and in combination were tested (n = 610 per group). Controls were perfused with Krebs-Henseleit medium (n = 10).

.-lgen ts 3-Isobutyl-I-methyl-xanthine (IBMX), LNNA and superoxide dismutase (bovine erythrocytes, specific activity 3300 Ujmg) were purchased from Serva (Heidelberg, Germany). RT, RA-octil, and HOE 140 were prepared in the Pharma Synthesis (Hoechst AG) and dissolved in saline. BK was obtained from Sigma (Deisenhofen, Germany). Salts, enzymes, coenzymes and adenine nucleotides were purchased from Boehringer Mannheim i Germany).

Statistical anatysis Unless indicated otherwise the data are reported as means & S.E.M. Statistical evaluation was performed with Student’s t test for unpaired data and with Dunnett’s test when more than two groups were compared. For data not normally distributed, Wilcoxon’s sign-rank test for paired data was applied. A probability of PC 0.05 was considered statistically significant.

to Cardioprotection

911

Results

E$ects of R T and BK on cyclic GMP formation in endothelial cells Incubation of endothelial monolayers with the ACE-inhibitor RT increased cyclic GMP production in a time- and concentrationdependent manner (Fig. 1). The increases in cyclic GMP developed slowly, reaching a plateau level after IO min and remained stable for at least 15min [Fig. l(a)]. Maximal increases were obtained between 1 x IO ’ 1 X lo-” mol/l RT at threshold concentrations of about 3 X lo-” mol/l, whereas high concentrations of RT( > 1 x lo-” mol/l) produced much less of an increase in cyclic GMP content (Fig. 1 (b)]. Preincubation (5 min) of the cells with either HOE 140 (1 x IO-’ mol/l:~ or LNNA (1 x lo-” mol/l:j [Fig. l(b)] abolished completely the increases in cyclic GMP. BK also induced a time- and concentrationdependent increase in the cyclic GMP content in BAEC. After addition of BK, the maximum increases were reached by 1 min and returned after a short plateau to baseline by 10 min [Fig. 2(a)]. Similarly to the RT effects BK resulted in a bell-shaped concentrationresponse curve [Fig. 2(b)]. Preincubation (5 min) of rndothelial cells with HOE 140 (1 x 10.’ mol/l) abolished completely the BKinduced increases of the cyclic nut-lrotide [Fig. 2(b) ]. A similar inhibition of BKinduced cyclic GMP formation was observed in monolayers that had been preincuhatrd for 5 min with L-NNA ( I x IU ’ nrol,‘lj [Fig. 2(b)]. RA-octil did not aIfect endothelial cyclic GMP production in BAEC (Fig. :i,. EJects of R’T and Bk‘ on PCI, bio.lvnthr.cir in endothelial cel1.r The RT-induced PGI, release was delayed and only significant after 10 min [Fig. 4(a)]. Incubation of BAEC with RT in the concentration range of 1 x 10eH and 1 x 1V” mol/l induced an increase in PGI, release that was significantly inhibited after preincubation with HOE 140 (I x IO-? mol/l) [Fig. 4(b)]. BK stimulated the release of PGI, in the endothelial cells faster than RT [Fig. 5(a) 1. Threshold concentrations were in the range of about 10-s mol/l, and maximal increases were seen between 1 x 1V7 and I x 10-j molil BK.

912

W. Linz

et al.

8 7- (a)

g’-

Eo 5 z ,a a

-I

1” 0

I

2

4

3

h

4

P-

8-

0’

a

5

3

14

IO II

f

12

4

13

t

14

3

15

16

(b)

Control

10-e

10-6

Io-7

Ramiprilot

FIGURE 1. Effects aortic endothelial cells. (b) Effects as a function ---0: RT+L-NNA). batches. HOE 140 (1 x the addition of RT. All *PiO.O5 us unstimulated

I”

6 7 8 9 Time (min)

concentration

10-5

10-4

(mol/l)

bovine of ramiprilat (RT; I X lO~“mol/l) on the accumulation of cyclic GMP in cultured (a) Effects as a function of time at a given concentration (0 ~ -0: controls; M: RT.) of concentration (10 min incubation with RT) (e-p+: RT; A---A: RT+HOE 140; Results are expressed as the meanfs.E.M. of 6-8 experiments performed on 4 different cell lO~‘mol/l) was added 5 min and N”-nitro-L-arginine (L-NNA; 1 x 10e5 mol/l) 5 min before experiments were performed in the presence of IBMX (0.1 mM) and SOD (3 x 10-‘mol/l). cells.

As in the RT experiments, HOE 140inhibited the BK-induced PGI, release [Fig. 5(b)]. Isolated rat hearts with post-ischaemic reperfusion arrhythmias

In isolated rat hearts subjected to regional myocardial ischaemia followed by reperfusion, RT (1 x lo-smol/l) and BK (1 x lo-’ and 1 x IO- lomol/l) reduced the incidence and duration of post-ischaemic ventricular fibrillation (Fig. 6). In parallel, myocardial functions such as left ventricular pressure,left ventricular dP/dt,,, and coronary flow increased in rat hearts perfused with RT or BK compared to vehicle-perfused rat hearts. During the ischaemic and reperfusion period the cytosolic enzymes of lactate dehydrogenaseand creatine kinase as well as lactate output were diminished in BK and RT perfused hearts.

At the end of the experiment after ischaemia and reperfusion, myocardial metabolism was improved via increased high energy rich phosphates (ATP and creatine phosphate) and glycogen stores as well as depressedlactate levels (Fig. 7). Both RT and BK showed a comparable “fingerprint” with respect to myocardial function and metabolism (Fig. 7). When LNNA (1 x lO-‘j mol/l) or HOE 140 (1 x lo-’ mol/l) were added to the perfusion medium the protective effects of RT or BK were abolished (Figs 6 and 8). Both the NO-synthase inhibitor (Fig. 9) or the B, kinin receptor antagonist alone reduced coronary Ilow compared to the vehicle perfused group and blocked the flow increase by RT or BK perfusion RA-octil (1 x 1O-6 and 1 x IO-’ mol/l) did not show protective effects on isolated ischaemic rat hearts (not shown).

Endothelial

NO-formation

0

Control

Contributes

10-e

10-T

Bradykmm

to Cardioprotection

10-G

concentration

10-s

913

10-G

(mol/l)

FIGURE 2. Effects of bradykinin (BK; I x lo- ’ mol/l) on the accumulation of cyclic GMP in cultured bovine aortic endothelial cells. (a) Effects as a function oftime at a given concentration (0 0: control; t-a: BK). (b) Effects as a function of concentration (1 min incubation with BK) (e-P l : BK; AmA:BK+HOE140; n - mm--H:BK+LNNA). HOE 140 (1 X IO-’ mol/l) was added 5 min and NC nitro-L-arginine (L-NNA; 1 x 10 5 mol/l) 5 min before the addition of BK. Results are expressed as the mean f S.E.M. of 4-6 experiments performed on 4 different cell batches. All experiments were performed in the presence of IBMX (0.1 mM) and SOD (3 x 10-j mol/l). *P
2--

l__il! O-

I Control

I 1x10-@

I 1X10-’

Romipril-octal

FIGURE conditions

3. Effects of ramipril-octil in the control experiments

on cyclic were similar

I IX 10-6 concentrotlon

I I x 10-5

I Ix 10-Q

(mol/L)

GMP formation in cultured bovine aortic endothelial to those in the control experiments of Figure I ibl.

culls. The

914

W. Linz

2 2 0a

40

et al.

J'

_----

I--/---

20

\r OT -I 0I II 2’ 3I 4I 5’ 6I 7/ 8I 9’ Time lmm)

.EE

Ii< :: A &

,

IO II

12 13 14 15 IE

70r

- ib)

6050 40-

2ot

I::

Control

10-s

Ramlprtlot

10-T

concentration

10-c

10-s

10-Q

(mol/l)

FIGURE 4. Effects of ramiprilat (RT) on the accumulation of 6-keto-PGF,, in the supernatant of bovine aortic endothelial cells (BAEC). (a) Effects as a function of time (-: control; - - - : RT (1 X 10-6mol/l) ). (b) Effects as a function ofconcentration (10min incubation with RT) (t-o: RT; A---A: RT+HOE 140). HOE 140 (1 x lo’ mol/l) was added to the bovine cells 5 min before the addition of RT. Results are expressed as the mean ?ZS.E.M. of 6 experiments. All experiments were performed in the presence of IBMX (0.1 mM) and SOD (3 x 10-l mol/l). *P
Discussion ACE-inhibition by ramiprilat in, or the addition of BK to, BAEC led to increased NOformation as assessed by endothelial cyclic GMP and PGI, synthesis. In contrast captopril was not active in BAEC (de Nucci et al., 1988). A possible explanation for this discrepancy might be a loss of ACE activity during cell preparation and culturing. On the other hand, it was reported that captopril induced endothelium-dependent vasodilation. This effect was related to its sulfhydryl group and the ability of the latter to scavenge O,-, thereby protecting EDRF (Goldschmidt and Tallarida, 199 1; Chopra et al., 1989). The finding that endothelial cells grown in dishes under non-flow conditions are able to produce BK in the presence of the ACE-inhibitor RT fits very well with the observations seen in isolated rat hearts.

In our isolated rat hearts with post-ischaemic reperfusion the ACE-inhibitor RT as well as BK evoked a cardioprotective effect as indicated by a shortening of reperfusion arrhythmias and an improvement of cardiodynamics via increased coronary flow, left ventricular pressure and left ventricular dP/ dt,,,. Furthermore, cell membranes were stabilized, as indicated by a reduced outflow of cytosolic enzymes, and metabolism was improved via increased high energy rich phosphates and glycogen stores. Both RT and BK produced comparable changes indicating a common mechanism of action. It has been speculated that ACE-inhibitors promote vasodilation by increasing the level of endogenous BK in subthreshold concentrations in the vascular wall by a local kallikreinkinin system (Scicli et al., 1991). The presence of high molecular weight kininogen has been

Endothelial

NO-formation

400-

Contributes ___-

_____

(a)

300. ,'

,'

,'

,'

,’ T

J,',,'

2OO2' IOO,' 0 -I

91 5

to Cardioprotection

I 0

T'

I' J"

,'

I I

/'

I 2

IS 3 4

5

I 6

I 7 Time

10-E

Control

Bradyklmn

6

I 9

I II I / IO II 12 13 14 15

(mln)

10-T

10-C

concentration

10-5

10-q

(mot/l)

in the supernatant of hoviw aortic FIGURE 5. Effects of hradykinin (BK) on the accumulation of 6-keto-PGF,x cndothelial cells. (a) Effects as a function of time at a given concentration (~: control: - - - : BE;, 1 x 10 h mol/l). (bj l : BK; A A: BK + HOE 140). HOE 140 Effects as a function of concentration (1 min incubation with BK) (em : I x 10 ’ moljl) was added to the bovine cells 5 min before the addition of bradykinin. Results are exprrssed as the mean+ S.E.M. of six experiments. All experiments were performed in thr prtvnrr of IBMX 10.1 my and SOD 3 x 10 i mol/l) *PC 0.05 rs unstimulated cells.

DBK

+ I-NNA RT + I-NNA

Concentration FIGURE fibrillation

6. Incidence (e.g. ten out of ten (VF) (min) in isolated rat hearts. *PC 0.05 /I\ control hrarts. Mean f S.E.M.

hearts (BK:

(mol/O fibrillated bradykinin;

in

the RT:

control group) and ramiprilat; L-NNA:

duration of ventricular S”-llitrt,-I,-arqininr~.

916

W. Linz

Cardladynamic

Venous

et al.

effluent

Myocard~ol

tissue

FIGURE 7. Effects of ramiprilat (1 x 10~Rmol/l) (a) and bradykinin (I x 10 qmol/l) (b) on myocardial function (cardiodynamic) and myocardial metabolism (venous effluent, myocardial tissue) in isolated working rat hearts with post-ischaemic reperfusion arrhythmias. (LVP: left ventricular pressure; HR: heart rate; CF: coronary flow; LDH: lactate dehydrogenase; CK: creatine kinase; ATP: adenosine triphosphatr: CP: rreatinc phosphate. *P
demonstrated in cultured human endothelial cells (van Iwaarden et al., 1988), but it is not yet established from where the BK precursors come and to what type they belong. In our BAEC we observed a slow increase in cyclic GMP content in response to RT, but BK elicited a fast transient increase in this second messenger. Obviously endogenously formed kinins by RT (1 x IO-* to 1 x 10e6 mol/l) do not lead to receptor desensitization, whereas exposure to high concentrations of RT or BK develop a homologous desensitization of B, kinin receptors in endothelial cells (Luckhoff et al., 1988b).

Beside the increase of BK by local inhibition of ACE in endothelial cells, it has been suggested that the flow itself by generating shear stress on the surface of the endothelium may enhance the release of kinins and thereby nitric oxide (Buga et al., 1991; Miller et al., 1990). By enhancing coronary flow BK can increase myocardial contractility via stimulation of cardiac afferent sympathetic nerves (Munch and Longhurst, 1991). This would explain the increase in left ventricular pressure and contractility after RT and BK. Additional to the increase in coronary flow, BK has favorable metabolic effects, optimizing

Endothelial

NO-formation

Contributes

91 7

to Cardioprotection

616 * 28

c

mBK-ant HOE 140

IO/IO *

(b)

35

t BKontagonkst

DRT

t BKontagomst

m

Control

+10-‘0

+lo-q LL >

DIRT

28

8K m

BK-ant HOE 140

iBK

t BKantagonist + BKantagonist

08K

+10-g Concentration

+10-q

(mol/l)

FIGURE 8. Abolition of the protective effects of (a) ramiprilat (RT) and (b) bradykinin isolated working rat hearts with post-ischaemic rcperfusion arrhythmias. (VF: ventricular numbers indicate the incidence of the hearts to fibrillatr). *P
(BK) by fibrillation

HOE tin

140 in mint:

20 m

Control BK

IEiIillRT FSZZL-NNA OBK+L-NNA RTtL-NNA

10-g

10-a

Concentration FIGURE xmiprilat;

9. BK:

Changes bradykinin;

of basal coronary flow (CF) L-NNA: NC-nitro-L-arginine).

10-G

10-g +10-G

10-e +10-e

(mot/l) in the preischaemic prriod *P
in isolated working hearts. Mcan+s.E.M.

rat

hrarts.

/ RT:

918

W. Linz

nutritional flow across the capillary wall which in turn leads to an elevated glucose uptake in isolated rat hearts (Rosen et al., 1983). Hearts perfused with glucose exhibited fewer reperfusion arrhythmias than hearts perfused with acetate or palmitate, suggesting that the substrates used before reperfusion may affect the incidence of arrhythmias during reperfusion (Bricknell and Opie, 1978; Bernier & Hearse 1988). The BK-induced rise in endothelial cyclic GMP may improve the energy state which results in high energy rich phosphates in isolated rat hearts after reperfusion (Vuorinen et al., 1984). BK also increased PGI, release from our endothelial cells. In the isolated ischaemic rat heart BK may act synergistically with the cyclic AMP-dependent effects of prostacyclins. Perfusion of isolated ischaemic hearts with PGI, reduced the incidence and duration of ventricular fibrillation (VF), increased coronary flow and decreased cytosolic enzyme release. In contrast, inhibition of PG biosynthesis with indomethacin prolonged the duration of VF, reduced coronary flow and increased cytosolic enzyme release and attenuated the cardioprotective effects of RT or BK (Linz et al., 1989). Thus, arachidonic acid derivatives like prostaglandins, PGI, and thromboxane and lipoxygenase products may also be involved in early ischaemic and reperfusion arrhythmias. The PGI,-thromboxane balance seems to be of great importance for the induction or prevention of reperfusion arrhythmias (Parrat et al., 1987). Blockade of B, kinin receptors with the specific B, kinin receptor antagonist HOE 140 suppressed cyclic GMP formation and PGI, synthesis induced by RT and BK. The NO synthase inhibitor L-NNA specifically suppressed only the cyclic GMP increase in

et al.

BAEC. The specificity of BK-mediated effects by local ACE-inhibition with RT is also underlined by the fact that RA-octil without any ACE inhibitory activity did not affect endothelial cyclic GMP production in BAEC nor have any protective effects on isolated ischaemic rat hearts. In isolated rat hearts addition of both LNNA or HOE 140 abolished the protective effects of RT or BK suggesting a mechanism in which the intact endothelium plays a dominant role. Reversal of protective effects of RT, BK or PGI, in isolated ischaemic rat hearts was reported earlier with another B, kinin receptor antagonist (Linz et al., 1990). Similar results were obtained in vascular tissue PGI,-synthesis and its inhibition with a B, kinin receptor antagonist (Beierwaltes and Carretero, 1989). Furthermore, in anesthetized dogs with coronary artery constriction and reperfusion, BK infused intracoronarily in low doses without changing local or systemic blood flow protected against reperfusion arrythmias (Linz et al., 1990). In anesthetized dogs with myocardial infarction both the ACE-inhibitor RT and BK reduced significantly myocardial infarct size. This protective effect could be blocked by the B, kinin receptor antagonist HOE 140 (Martorana et al., 1990). Conclusion Inhibition of ACE ,localized on the luminal site of the vascular endothelium results in increased synthesis of NO and PGI, by accumulation of endogenous BK. This mechanism might contribute to cardioprotection in isolated rat hearts with post-ischaemic reperfusion injuries. Thus, improvement of endothelial NO synthesis by ACE-inhibition might be of therapeutic importance for myocardial ischaemia and hypertension.

References W. H., CARRETERO 0. A. (1989) Kinin antagonist reverses converting enzyme inhibitor-stimulated prostaglandin I, synthesis. Hypertension 13: 754-758. BERNIER M., HEARSE D. J. (1988) Reperfusion-induced arrhythmias: mechanisms of protection by glucose and mannitol. Am J Physiol 254: H862-H870. BANNER G. (1988) Haben die Kinine eine Bedeutung ftir die antihypertensive Wirkung der ACE-Hemmer? Z Kardiol 77 (Suppl3): 23-27. BANNER G., PREIS S., SCHUNK U., TOUSSAINT C., KAUFMANN W. (1990) Hemodynamic effects of bradykinin on sS;;;mic and pulmonary circulation in health and hypertensive humans. J Cardiovasc Pharmacol 15 (Suppl6): S4> BEIERWALTES

vascular

Endothelial

NO-formation

Contributes

to Cardioprotection

919

BRICKNEI.L 0. K., OPIE L. H. (1978) Effects of substrates on tissue metabolic changes in the isolated rat heart during underperfusion and on release of lactate dehydrogenase and arrhythmias during reperfusion. Circ Res 43: 102-I 15. BUGA G. M., GOLD M. E., FUKUTO J. M., IGNARRO L. J. (1991) Shear stress-induced release of nitric oxide from rndothelial cells grown on beads. Hypertension 17: 187-193. CHOPRA M., SCOTT N., MCMURRAY J., MCFAY J., BRIDGES A., SMITH W. E.. BELCH J. .J. F. ( 1989) Captopril: A frer radical scavenger. Br J Clin Pharmacol 27: 396-399. GOLDSCHUIDT J. E., TALLARIDA R. J. (1991) Pharmacological evidence that captopril possesses an endotheliummediated component of vasodilation: Effect of sulfhydryl groups on rndothelium-derived relaxing factor. J Pharmacol Exp Ther 257: 1136-l 145. HASHIMOTO K., HIROSE M., FURUKAWA S., HAYAKAWA H., KIMURA E. I 1977) Changes in hemodynamics and bradykinin concentration in coronary sinus blood in experimental coronary occlusion. Jpn Heart J 18: 679-689. HASHIMOTO K., HAMAMOTO H., HONDA Y., HIROSE M., FURUKAWA S., KIMURA E. (1978) Changes in components ol kinin system and hemodynamics in acute myocardial infarction. Am Heart J 95: 619426. I.MC.RA O., SHIMAMOTO K. (1989) Role of kallikrein-kinin system in the hypotensive mechanisms of converting-enzymr inhibitors in essential hypertension. J Cardiovasc Pharmacol 13 (Suppl3): S63pS66. “AI\ IWAARDEN F., DE GROOT P. G., SIXMA J. J,, BARRETTINI M., BOUMA BN (1988) High-molecular weight kininogen is present in cultured human endothelial cells: localization, isolation, and characterization. Blood 71: 1268--1276. KAWASHIMA K., WATANABE N., OOHATA H.. FUJIMOTO K., SUZUKI T., ISHIZAKI Y., MORITA I., MUROTA S. I. / 19901 Synthesis and release of acetylcholine by cultured bovine arterial endothelial cells. Neurosci Lrtt 119: 156. 158. KIFOR I., DZAU V. J. (1987) Endothelial renin-angiotensin pathway: Evidence for intracellular synthesis and wcretion of angiotensins. Circ Res 60: 422-428. LINZ W.. MARTORANA P. A., SCH~LKENS B. A. (1990) Local inhibition of hradykinin degradation in ischemic hearts. J Cardiovasc Pharmacol 15 (Suppl6): S99-SlO9. LINZ W’.. SCH~LKENS B. A., HAN Y. F. (1986) Beneficial effects of the converting enzyme inhibitor, ramipril, in ischcmic rat hearts. J Cardiovasc Pharmacol 8 (Suppl 10): S91-S99. LINZ W.. SCH~LKENS B. A., KAISER J., JUST M., BEI-YIN Q, ALBUS C., PETRY P. 1,1989) Cardiac arrhythmias arc ameliorated by local inhibition of angiotensin formation and hradykinin degradation with the convrrtinq-cnzymc inhibitor ramipril. Cardiovasc Drug Ther 3: 873-882. L~CKHOFF A., POHL U., M~~LSCH A., BUSSE R. (1988a) Differential role ofextraand intracellular calcium in the rtleast of EDRF and prostacyclin from cultured endothelial cells. Br J Pharmacol95: 189-196. LUCKHOFF A., ZEH R., BIXSE R. (1988b) Desensitization of the bradykinin induced rise in intracellular free c&ium in cultured endothelial cells. Pflugers Arch 412: 654-658. MARTORANA P. A., KETTENBACH B., BREIPOHL G., LINZ W., SCH~LKENS B. A. (1990) Reduction ofinfarct size by local angiotensin-converting enzyme inhibition is abolished by a bradykinin antagonist. Eur J Pharmacol 182: 395-396. \IILLER V. .%I. (1990) Does antihypertensive therapy improve the function of the vascular endothelium? Hypertension 16: 541~ 543. MILNER P., KIRKPATRICK K. A., RALEVIC V., TOOTHILL V., PEARSON J., BURNSTOCK G. (1990) Endothclial cells cultured from human umbilical vein release ATP, substance P and acetylcholinr in response to increased flow. Proc R Sot Land [Biol] 241: 245-248. M~LSCH A., BLUE R. ( 1990) N”-nitro-L-arginine (N5-[imino(nitroamino)methyl]-L-ornithine) impairs endothrliumdependent dilations by inhibiting cytosolic nitric oxide synthesis from L-argininc. Naunyn Schmiedebrrgs Arch Pharmacol 341: 143%147. ~I~NCH P. A., LONGHURST J. C. (1991) Bradykinin increases myocardial contractility: relation to the Gregg phrnomenon. Am J Physiol 260: RlO95-RllO3. DE N~u:I G.. WARNER ‘r., VANE J. R. (1988) Effect ofcaptopril on the bradykinin-induced release of prostacyrlin from guinea-pig lungs and bovine aortic endothelial cells. Br J Pharmacol 95: 783-788. PARRA.I J, R., COCKER S. J., WAINWRIGHT C. L. (1987) Eicosanoids and susceptibility to ventricular arrhythmias during myocardial ischemia and reperfusion. J Mel Cell Cardiol 19 (Suppl V): 55566. ROSEN l’., ECKEL J,. REINAUER H. (1983) Influence of bradykinin on glucose uptake and metabolism studies in isolated cardiac myorytes and isolated perfused rat hearts. Hoppe Seylers Z Physiol Chem 364: 431-438. SCICLI ;2. G.. FARHY R., SCICLI G., NOLLY H. (1991) The kallikrein-kinin system in heart and vascular tissue. In: “‘I‘hr rol? of Bradykinin in the Cardiovascular action of the converting enzyme inhibitor ramipril.” Eds: C. B6nner, B. A. Schblkrns and A. C. Scicli: 17-27. Published by Media Medica, 51 West Street, Chichester. Sussex PO19 1RP. rK. \.ANHOVTTE P. M., AUCHSCHWELK W., BIONDI M. L., LORENZ R. R., SC~NI V. B., VIDAL M. J. I 1989‘ Why ltr~ converting enzyme inhibitors vasodilators? Br J Clin Pharmacol 28 (Suppl2): 95-104. Vt.oRmEN P.. LAI.STIOLA K., METSA-KETELA T. (1984) The effects of cyclic AMP and cyclic GMP on rrdcrx htatt’ and energy state in hypoxic rat atria. Life Sci 35: 155-161. \~IEMER G.. SCHOLKENS B. A., BECKER R. H. A., BUSSE R. (1991) Ramiprilat enhances endothrlial autaroid t;)rmation IJ>- inhibiting breakdown of endothelium-derived bradykinin. Hypertension 18: 558-563. \~;IRTH Ii., HOCK F. J., ALBUS U., LINZ W., ALPERMANN H. G., ANAGNOSTOPOULOS H.. HENKB cr., BREIP~HL G., Kbh.10 W., KNOLLE J., SCH~LKENS B. A. (1991) HOE 140, a new potent and long acting bradykinin-antagonist: in &J studies. Br J Pharmacol 102: 774-777. YAKAGISAWA M., KURIHARA H., KIMURA S., TOMMOBE Y., KOBAYASHI X1., MITS~I Y., YAZAKI Y.. G(yro K., %IASAK ‘1’. 1988) A nobe potent vasoconstrictor peptide produced by vascular mothrlial cells. Nature 332: 41 1-41.5.