European Journal of Pharmacology, 182 (1990) 171-174
171
Elsevier EJP 20645
Short communication
Different time course of atrial natriuretic factor-induced relaxation in rabbit aorta and portal vein H6ctor De Le6n, Gilberto Castafleda-Hernfindez and Enrique H o n g Secci6n de Terapdutica Experimental, Departamento de Farmacologla y Toxicologla, Centro de Investigaci6n y de Estudios Avanzados del I.P.N., Apartado Postal 22026, 14000 M~xico D.F., M~xico Received 9 March 1990, accepted 24 March 1990
Atrial natriuretic factor (ANF) and nitroglycerine were tested in superfused rabbit aorta and portal vein strips precontracted with norepinephrine. The ANF dose-response curves were similar in both tissues, however, relaxation was fast and transient in portal veins whereas it was sustained and of slow onset in aortas. Nitroglycerine was more potent in portal veins than in aortas but relaxation was fast and transient in both tissues. The results suggest that the time course of ANF-induced relaxation in different vessels may be of relevance to the overall hemodynamic profile of this peptide in intact animals. ANF (atrial natriuretic factor); Atrial peptides; Nitroglycerine; Vascular sensitivity
1. Introduction
It has been reported that the vasorelaxant profile of atrial natriuretic factor (ANF) is similar to that of nitrovasodilators, suggesting that these agents may have a c o m m o n mechanism of action (Winquist et al., 1984). Actually, it is known that both A N F and nitrates increase intracellular c G M P levels in vascular smooth muscle, but the atrial peptide stimulates the particulate form of guanylate cyclase whereas nitrovasodilators act on the soluble form of this enzyme (Waldman and Murad, 1987). Moreover, bolus injections of both types of compounds produce a dose-dependent hypotensive effect in whole animal preparations. However, this effect is reversed very rapidly (1-3 min) after administration of nitrovasodilators
Correspondence to: H. De Le6n, Clinical Research Institute of Montreal, Laboratory of Experimental Hypertension and Vasoactive Peptides, 110, Pine Avenue West, Montreal, Quebec, Canada H2W 1R7.
(Ignarro et al., 1981), whereas with the atrial peptide, the blood pressure remains under basal levels for longer periods (23 min) (Garcia et al., 1985). In this study we examined the potency and time course of A N F and nitroglycerine ( N T G ) to relax superfused rabbit aorta and portal vein strips arranged in a cascade fashion.
2. Materials and methods
Rabbit aorta helical strips and portal vein longitudinal strips were placed in a tissue chamber system arranged in a cascade fashion and superfused with Krebs solution at 5 m l / m i n , as described previously (De Le6n et al., 1988). The initial tension was 4 g for the aorta strips and 0.75 g for the portal vein strips. Strips were allowed to equilibrate for 1 h and then they were precontracted by a continuous infusion of 100 nM norepinephrine (NE) (Sigma, St. Louis, MO, USA). Once the contraction had reached a plateau,
0014-2999/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
172 A N F (Synthetic A N F corresponding to rat A N F 8-33, Merck Sharp and D o h m e Research Labs, Rahway, N J, USA) or N T G (Lilly, Indianapolis, IN, USA) was applied as a bolus dose dissolved in 0.5 ml of saline. The time course of the relaxant effects was studied with a dose of 10 -6 M. Data are presented as mean values + S.E.M. Effective dosess0 (EDs0) and 95% confidence limits were calculated and c o m p a r e d with Student's ttest, as described by Tallarida and M u r r a y (1981).
PORTAL
,25-
VEIN
AORTA
,"4:/u0¢~4,
I00 -
-o-
t,
#
z 0
75
b< c~
i.
50
© o 25
o0
10
15
30
TIME
3. ResMts
The potency and efficacy of A N F were similar in rabbit aorta and portal vein strips (fig. 1, left); the difference between the EDs0 values was not statistically significant. Furthermore, an A N F concentration of 10 -5 M inhibited contractions induced by N E by about 100% in the two vascular preparations. Unlike A N F , N T G was more potent in the portal vein than in the aorta (fig. 1, right). The difference between ED50 values was statistically significant ( P < 0 . 0 5 ) ; 10 -5 M N T G , the highest N T G dose assayed, inhibited N E induced contractions by more than 100% in portal vein strips but by only 83% in aorta strips. A N F and N T G (both at a dose of 10 -6 M) both produced a rapid and transient relaxation in the portal vein (fig. 2, left). The m a x i m u m relaxation was 98% with N T G and 84% with A N F , and 125•
AORTA
s PORTAL
J
VEiN
IO0 -
75.
;/?' T~
50-
25-
5
/
9" t
I
I
o
5
/
iO
I
i5
/~ 3o
(rmn)
Fig. 2. Time course of the relaxant effect of a 0.5 ml bolus injection of either ANF or nitroglycerine (NTG) (both 10 6 M) in superfused strips of rabbit portal vein (left) or aorta (fight) contracted by a continuous infusion of norepinephrine 10 7 M. Data correspond to means_+S.E.M, of six vascular preparations. in both cases it was achieved in 30 s. Thereafter, the portal vein strips gradually recovered their contractile tone. At 15 min, preparations showed a recovery ranging from 65 to 90% of the initial contraction level, and there was no statistically significant difference between A N F - and N T G treated strips. However, recovery from A N F - i n duced relaxation was slower than from N T G - i n duced relaxation. In the aorta (fig. 2, fight), the time courses of the relaxant effects of A N F and N T G were clearly different. The N T G effect was rapid and transient, as it was in the portal vein. The m a x i m u m relaxation was 71% of the N E - i n d u c e d contraction and was achieved in 30 s; Thereafter the aorta strips recovered, reaching the initial contraction level in 5-10 min. A N F induced a slow relaxation, reaching a m a x i m u m effect (inhibition of 75% of the N E - i n d u c e d contraction) in 10 min; however, aorta strips remained relaxed for the whole observation period, despite continuous washout and N E infusion.
41""
0 i
9 Log
I
I
I
-'8 7 6 -5 [SYNTHETIC ANF] (M)
Log [NITROGLYCERIN]
(M)
Fig. 1. Relaxant effect of ANF (left) and nitroglycerine (right) in aorta and portal vein superfused strips contracted by a continuous infusion of norepinephrine 10 7 M. Data correspond to means ___S.EM. of six vascular preparations.
4. Discussion
U n d e r our conditions of continuous superfusion of vascular strips, the relaxant profiles of A N F and N T G were dissimilar. The A N F dose-
173
response curves (constructed with the m a x i m u m effect observed after each dose) were similar in the aorta and portal vein, whereas N T G showed greater potency in the portal vein. This is consistent with the observations of Edwards and coworkers (1984) that nitrovasodilators are more potent in veins than in arteries. The N T G effect was rapid and transient in both tissues. The A N F effect was rapid and transient in the portal vein but in the aorta it was sustained and of slow onset, despite continuous washout and infusion of NE. This was also observed with histamine precontracted aorta strips (data now shown). These results confirm previous reports that ANF-induced relaxation is not abolished by washout in certain vascular preparations (Winquist et al., 1985). With this experimental model, portal veins did not develop tolerance to either A N F or N T G . When portal vein strips returned to pre-drug contraction levels, a subsequent challenge produced a relaxation comparable to the one previously observed. Similar results were obtained with N T G in the aorta. It was not possible to determine whether aorta strips developed tolerance to ANF, because of the long duration of the relaxation response. The lack of development of tolerance to N T G is probably because of the short exposure time to the compound. Both preparations were placed in a tissue chamber system arranged in a cascade fashion. In other words, the tissues were not incubated with a given dose of N T G , but the drug was applied as a bolus to the top of the superfusion system. A recent report (Kowaluk et al., 1989) demonstrates that a reduced time of exposure to N T G decreases the degree of in vitro tolerance in aortic rings. Moreover, vascular tolerance to N T G can be reversed completely if a long enough NTG-free washout period is allowed. It is known that A N F is relatively ineffective is antagonizing vascular contractile responses associated with cell m e m b r a n e depolarization, i.e. KC1induced contraction in the aorta and spontaneous contractions in the portal vein (Winquist et al., 1984). Considering these findings, the responses of aortas and portal veins observed can be explained by the ability of A N F to inhibit agonist-induced contractions in both tissues and by its small effect
on the intrinsic contractions of the portal vein. The present results extend previous observations on the transient nature of the ANF-induced relaxation in the pulmonary artery compared to the prolonged effect in the trachea (O'Donell et al., 1985). A N F and nitrovasodilators increase intracellular levels of c G M P in smooth muscle, with a time course that correlates with their relaxant effects (Harris et al., 1985). However, A N F stimulates the particulate form of guanylate cyclase, while nitrates act on the soluble form of this enzyme (Waldman and Murad, 1987). Thus, differences observed between the time course of the relaxant effect of A N F and N T G could be due to different activation kinetics of the two forms of guanylate cyclase. Recent studies have suggested that there may be at least two populations of A N F receptors in several tissues (Vandlen et al., 1986), one population being coupled to guanylate cyclase while the other seems to be coupled to a different second messenger system (Waldman and Murad, 1987). It is possible that differences in the distribution of A N F receptor populations in vascular tissues play a role in the time course of the relaxant responses to the atrial peptide. The results of the present study indicate that it is important to consider not only the potency of A N F , but also the time course of the relaxant effect on different blood vessels when characterizing the modulation of vascular tone induced by this peptide. It appears that differences between the duration of the hypotensive effects elicited by A N F and nitrates in whole animal preparations are probably due, at least in part, to the dissimilar time courses of the relaxation responses in the various vascular beds. Undoubtedly, further studies are required to completely understand the mechanisms of the overall hemodynamic effect of ANF.
Acknowledgements The authors thank Mr. J. Sfinchez and Mr. J.J. Lrpez for technical assistance and Dr. Raul Garcia for valuable comments. This work was supported by COSNET-SEP, Grant PEP 04.88.
174
References De Le6n, H., G. Casta?teda-Hernb.ndez and E. Hong, 1987, Decreased ANF atrial content and vascular reactivity to ANF in spontaneous and renal hypertensive rats, Life Sci. 41, 341. Edwards, J.C., L.J. Ignarro, A.L. Hyman and P.J, Kadowitz, 1984, Relaxation of intrapulmonary artery and vein by nitrogen oxyde-containing vasodilators and cycfic GMP, J. Pharmacol. Exp. Ther. 228, 33. Garcia, R., G. Thibault, P. Hamet, J. Gutkowska, M. Cantin and J. Genest, 1985, Effect of atrial natriuretic factor [ANF (Arg 101-tyr 126)] on kallikreid and cyclic GMP in the renovascular hypertensive rat, Clin. Exp. Theory Pract. A7 (ll), 1597. Harris, D.W., C.A. Baker, H.H. Saneii and G.A. Johnson, 1985, Stimulation of cyclic GMP formation in smooth muscle cells by atriopeptin II, Life Sci. 37, 591. Ignarro, L.J., H. Lippton, J.C. Edwards, W.H. Baricos, A.L. Hyman, P.J. Kadowitz and C.A. Gruetter, 1981, Mechanism of vascular smooth muscle relaxation by organic nitrates, nitrites, nitroprusside and nitric oxide: evidence for the involvement of S-nitrosothiols as active intermediates, J. Pharmacol. Exp. Ther. 218, 739.
Kowaluk. E., K. Hough and H.L. Fung, 1989, Effect of intermittent exposure and drug-free intervals on the in vitro vascular tolerance to nitroglycerin, Life Sci. 44, 1157. O'Donnel, M., R. Garippa and F. Welton, 1985, Relaxant activity of atriopeptins in isolated guinea pig airways and vascular smooth muscle, Peptides 6, 597. Tallarida, R.J. and R.B. Murray, 1981, Manual of Pharmacologic Calculations with Computer Programs (SpringerVerlag, New York) p. 21. Vandlen, R.L., K.W. Arcuri, L. Phope, M.E. Keegan and M.A. Napier, 1986, Molecular characteristics of receptors for atrial natriuretic factor, Fed. Proc. 45, 2366. Waldman, S.A. and F. Murad, 1987, Cyclic GMP synthesis and function, Pharmacol. Rev. 39, 163. Winquist, R.J., E.P. Faison and R.F. Nutt, 1984, Vasodilator profile of synthetic atrial natriuretic factor, European J. Pharmacol. 102, 169. Winquist, R.J., M.A. Napier, R.L. Vandlen, K.E. Arcuri, M.E. Keegan, E.P. Faison and E.P. Basquin, 1985, Pharmacology and receptor binding of atrial natriuretic factor in vascular smooth muscle, Clin. Exp. Hypert. A7, 869.
171
Elsevier EJP 20645
Short communication
Different time course of atrial natriuretic factor-induced relaxation in rabbit aorta and portal vein H6ctor De Le6n, Gilberto Castafleda-Hernfindez and Enrique H o n g Secci6n de Terapdutica Experimental, Departamento de Farmacologla y Toxicologla, Centro de Investigaci6n y de Estudios Avanzados del I.P.N., Apartado Postal 22026, 14000 M~xico D.F., M~xico Received 9 March 1990, accepted 24 March 1990
Atrial natriuretic factor (ANF) and nitroglycerine were tested in superfused rabbit aorta and portal vein strips precontracted with norepinephrine. The ANF dose-response curves were similar in both tissues, however, relaxation was fast and transient in portal veins whereas it was sustained and of slow onset in aortas. Nitroglycerine was more potent in portal veins than in aortas but relaxation was fast and transient in both tissues. The results suggest that the time course of ANF-induced relaxation in different vessels may be of relevance to the overall hemodynamic profile of this peptide in intact animals. ANF (atrial natriuretic factor); Atrial peptides; Nitroglycerine; Vascular sensitivity
1. Introduction
It has been reported that the vasorelaxant profile of atrial natriuretic factor (ANF) is similar to that of nitrovasodilators, suggesting that these agents may have a c o m m o n mechanism of action (Winquist et al., 1984). Actually, it is known that both A N F and nitrates increase intracellular c G M P levels in vascular smooth muscle, but the atrial peptide stimulates the particulate form of guanylate cyclase whereas nitrovasodilators act on the soluble form of this enzyme (Waldman and Murad, 1987). Moreover, bolus injections of both types of compounds produce a dose-dependent hypotensive effect in whole animal preparations. However, this effect is reversed very rapidly (1-3 min) after administration of nitrovasodilators
Correspondence to: H. De Le6n, Clinical Research Institute of Montreal, Laboratory of Experimental Hypertension and Vasoactive Peptides, 110, Pine Avenue West, Montreal, Quebec, Canada H2W 1R7.
(Ignarro et al., 1981), whereas with the atrial peptide, the blood pressure remains under basal levels for longer periods (23 min) (Garcia et al., 1985). In this study we examined the potency and time course of A N F and nitroglycerine ( N T G ) to relax superfused rabbit aorta and portal vein strips arranged in a cascade fashion.
2. Materials and methods
Rabbit aorta helical strips and portal vein longitudinal strips were placed in a tissue chamber system arranged in a cascade fashion and superfused with Krebs solution at 5 m l / m i n , as described previously (De Le6n et al., 1988). The initial tension was 4 g for the aorta strips and 0.75 g for the portal vein strips. Strips were allowed to equilibrate for 1 h and then they were precontracted by a continuous infusion of 100 nM norepinephrine (NE) (Sigma, St. Louis, MO, USA). Once the contraction had reached a plateau,
0014-2999/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
172 A N F (Synthetic A N F corresponding to rat A N F 8-33, Merck Sharp and D o h m e Research Labs, Rahway, N J, USA) or N T G (Lilly, Indianapolis, IN, USA) was applied as a bolus dose dissolved in 0.5 ml of saline. The time course of the relaxant effects was studied with a dose of 10 -6 M. Data are presented as mean values + S.E.M. Effective dosess0 (EDs0) and 95% confidence limits were calculated and c o m p a r e d with Student's ttest, as described by Tallarida and M u r r a y (1981).
PORTAL
,25-
VEIN
AORTA
,"4:/u0¢~4,
I00 -
-o-
t,
#
z 0
75
b< c~
i.
50
© o 25
o0
10
15
30
TIME
3. ResMts
The potency and efficacy of A N F were similar in rabbit aorta and portal vein strips (fig. 1, left); the difference between the EDs0 values was not statistically significant. Furthermore, an A N F concentration of 10 -5 M inhibited contractions induced by N E by about 100% in the two vascular preparations. Unlike A N F , N T G was more potent in the portal vein than in the aorta (fig. 1, right). The difference between ED50 values was statistically significant ( P < 0 . 0 5 ) ; 10 -5 M N T G , the highest N T G dose assayed, inhibited N E induced contractions by more than 100% in portal vein strips but by only 83% in aorta strips. A N F and N T G (both at a dose of 10 -6 M) both produced a rapid and transient relaxation in the portal vein (fig. 2, left). The m a x i m u m relaxation was 98% with N T G and 84% with A N F , and 125•
AORTA
s PORTAL
J
VEiN
IO0 -
75.
;/?' T~
50-
25-
5
/
9" t
I
I
o
5
/
iO
I
i5
/~ 3o
(rmn)
Fig. 2. Time course of the relaxant effect of a 0.5 ml bolus injection of either ANF or nitroglycerine (NTG) (both 10 6 M) in superfused strips of rabbit portal vein (left) or aorta (fight) contracted by a continuous infusion of norepinephrine 10 7 M. Data correspond to means_+S.E.M, of six vascular preparations. in both cases it was achieved in 30 s. Thereafter, the portal vein strips gradually recovered their contractile tone. At 15 min, preparations showed a recovery ranging from 65 to 90% of the initial contraction level, and there was no statistically significant difference between A N F - and N T G treated strips. However, recovery from A N F - i n duced relaxation was slower than from N T G - i n duced relaxation. In the aorta (fig. 2, fight), the time courses of the relaxant effects of A N F and N T G were clearly different. The N T G effect was rapid and transient, as it was in the portal vein. The m a x i m u m relaxation was 71% of the N E - i n d u c e d contraction and was achieved in 30 s; Thereafter the aorta strips recovered, reaching the initial contraction level in 5-10 min. A N F induced a slow relaxation, reaching a m a x i m u m effect (inhibition of 75% of the N E - i n d u c e d contraction) in 10 min; however, aorta strips remained relaxed for the whole observation period, despite continuous washout and N E infusion.
41""
0 i
9 Log
I
I
I
-'8 7 6 -5 [SYNTHETIC ANF] (M)
Log [NITROGLYCERIN]
(M)
Fig. 1. Relaxant effect of ANF (left) and nitroglycerine (right) in aorta and portal vein superfused strips contracted by a continuous infusion of norepinephrine 10 7 M. Data correspond to means ___S.EM. of six vascular preparations.
4. Discussion
U n d e r our conditions of continuous superfusion of vascular strips, the relaxant profiles of A N F and N T G were dissimilar. The A N F dose-
173
response curves (constructed with the m a x i m u m effect observed after each dose) were similar in the aorta and portal vein, whereas N T G showed greater potency in the portal vein. This is consistent with the observations of Edwards and coworkers (1984) that nitrovasodilators are more potent in veins than in arteries. The N T G effect was rapid and transient in both tissues. The A N F effect was rapid and transient in the portal vein but in the aorta it was sustained and of slow onset, despite continuous washout and infusion of NE. This was also observed with histamine precontracted aorta strips (data now shown). These results confirm previous reports that ANF-induced relaxation is not abolished by washout in certain vascular preparations (Winquist et al., 1985). With this experimental model, portal veins did not develop tolerance to either A N F or N T G . When portal vein strips returned to pre-drug contraction levels, a subsequent challenge produced a relaxation comparable to the one previously observed. Similar results were obtained with N T G in the aorta. It was not possible to determine whether aorta strips developed tolerance to ANF, because of the long duration of the relaxation response. The lack of development of tolerance to N T G is probably because of the short exposure time to the compound. Both preparations were placed in a tissue chamber system arranged in a cascade fashion. In other words, the tissues were not incubated with a given dose of N T G , but the drug was applied as a bolus to the top of the superfusion system. A recent report (Kowaluk et al., 1989) demonstrates that a reduced time of exposure to N T G decreases the degree of in vitro tolerance in aortic rings. Moreover, vascular tolerance to N T G can be reversed completely if a long enough NTG-free washout period is allowed. It is known that A N F is relatively ineffective is antagonizing vascular contractile responses associated with cell m e m b r a n e depolarization, i.e. KC1induced contraction in the aorta and spontaneous contractions in the portal vein (Winquist et al., 1984). Considering these findings, the responses of aortas and portal veins observed can be explained by the ability of A N F to inhibit agonist-induced contractions in both tissues and by its small effect
on the intrinsic contractions of the portal vein. The present results extend previous observations on the transient nature of the ANF-induced relaxation in the pulmonary artery compared to the prolonged effect in the trachea (O'Donell et al., 1985). A N F and nitrovasodilators increase intracellular levels of c G M P in smooth muscle, with a time course that correlates with their relaxant effects (Harris et al., 1985). However, A N F stimulates the particulate form of guanylate cyclase, while nitrates act on the soluble form of this enzyme (Waldman and Murad, 1987). Thus, differences observed between the time course of the relaxant effect of A N F and N T G could be due to different activation kinetics of the two forms of guanylate cyclase. Recent studies have suggested that there may be at least two populations of A N F receptors in several tissues (Vandlen et al., 1986), one population being coupled to guanylate cyclase while the other seems to be coupled to a different second messenger system (Waldman and Murad, 1987). It is possible that differences in the distribution of A N F receptor populations in vascular tissues play a role in the time course of the relaxant responses to the atrial peptide. The results of the present study indicate that it is important to consider not only the potency of A N F , but also the time course of the relaxant effect on different blood vessels when characterizing the modulation of vascular tone induced by this peptide. It appears that differences between the duration of the hypotensive effects elicited by A N F and nitrates in whole animal preparations are probably due, at least in part, to the dissimilar time courses of the relaxation responses in the various vascular beds. Undoubtedly, further studies are required to completely understand the mechanisms of the overall hemodynamic effect of ANF.
Acknowledgements The authors thank Mr. J. Sfinchez and Mr. J.J. Lrpez for technical assistance and Dr. Raul Garcia for valuable comments. This work was supported by COSNET-SEP, Grant PEP 04.88.
174
References De Le6n, H., G. Casta?teda-Hernb.ndez and E. Hong, 1987, Decreased ANF atrial content and vascular reactivity to ANF in spontaneous and renal hypertensive rats, Life Sci. 41, 341. Edwards, J.C., L.J. Ignarro, A.L. Hyman and P.J, Kadowitz, 1984, Relaxation of intrapulmonary artery and vein by nitrogen oxyde-containing vasodilators and cycfic GMP, J. Pharmacol. Exp. Ther. 228, 33. Garcia, R., G. Thibault, P. Hamet, J. Gutkowska, M. Cantin and J. Genest, 1985, Effect of atrial natriuretic factor [ANF (Arg 101-tyr 126)] on kallikreid and cyclic GMP in the renovascular hypertensive rat, Clin. Exp. Theory Pract. A7 (ll), 1597. Harris, D.W., C.A. Baker, H.H. Saneii and G.A. Johnson, 1985, Stimulation of cyclic GMP formation in smooth muscle cells by atriopeptin II, Life Sci. 37, 591. Ignarro, L.J., H. Lippton, J.C. Edwards, W.H. Baricos, A.L. Hyman, P.J. Kadowitz and C.A. Gruetter, 1981, Mechanism of vascular smooth muscle relaxation by organic nitrates, nitrites, nitroprusside and nitric oxide: evidence for the involvement of S-nitrosothiols as active intermediates, J. Pharmacol. Exp. Ther. 218, 739.
Kowaluk. E., K. Hough and H.L. Fung, 1989, Effect of intermittent exposure and drug-free intervals on the in vitro vascular tolerance to nitroglycerin, Life Sci. 44, 1157. O'Donnel, M., R. Garippa and F. Welton, 1985, Relaxant activity of atriopeptins in isolated guinea pig airways and vascular smooth muscle, Peptides 6, 597. Tallarida, R.J. and R.B. Murray, 1981, Manual of Pharmacologic Calculations with Computer Programs (SpringerVerlag, New York) p. 21. Vandlen, R.L., K.W. Arcuri, L. Phope, M.E. Keegan and M.A. Napier, 1986, Molecular characteristics of receptors for atrial natriuretic factor, Fed. Proc. 45, 2366. Waldman, S.A. and F. Murad, 1987, Cyclic GMP synthesis and function, Pharmacol. Rev. 39, 163. Winquist, R.J., E.P. Faison and R.F. Nutt, 1984, Vasodilator profile of synthetic atrial natriuretic factor, European J. Pharmacol. 102, 169. Winquist, R.J., M.A. Napier, R.L. Vandlen, K.E. Arcuri, M.E. Keegan, E.P. Faison and E.P. Basquin, 1985, Pharmacology and receptor binding of atrial natriuretic factor in vascular smooth muscle, Clin. Exp. Hypert. A7, 869.
