Protamine Induces Vasorelaxation of Human Internal Thoracic Artery by Endothelial NO-Synthase Pathway Dmitry Pevni, MD, Jacob Gurevich, MD, Inna Frolkis, MD, PhD, Gad Keren, MD, Izhak Shapira, MD, Josef Paz, MD, Amir Kramer, MD, Chaim Locker, MD, and Rephael Mohr, MD Department of Thoracic and Cardiovascular Surgery, Tel Aviv Sourasky Medical Center and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
Background. Protamine is commonly used in cardiac surgery to reverse the anticoagulant effects of heparin. We investigated the role of different nitric oxide synthase pathways in the response of the human internal thoracic artery to protamine and evaluated whether heparin could prevent this effect. Methods. A tension-recording method was used to obtain baseline measurements of contractions of human internal thoracic artery rings achieved with norepinephrine. Isolated internal thoracic artery rings were suspended in two organ chambers. One contained KrebsHenseleit solution and served as control. The other contained a heparin or N-Nitro-L-arginine (L-NAM, an inhibitor of both endothelial and inducible nitric oxide synthase) or a specific inhibitor of inducible nitric oxide synthase, aminoguanidine. Increasing doses of prota-
mine were added to both chambers and dose-response curves were obtained. Results. Protamine was found to relax contracted internal thoracic arteries 56% ⴞ 4.7% of baseline measurements in a concentration-dependent manner. When LNAM was added, protamine caused only a slight decrease of tension. There were no differences in the relaxing effect of protamine in the presence of aminoguanidine or heparin. Conclusions. Protamine induces nitric oxidedependent relaxation of the internal thoracic artery by activation of endothelial nitric oxide synthase pathway. Heparin could not prevent this relaxing effect of protamine. (Ann Thorac Surg 2000;70:2050 –3) © 2000 by The Society of Thoracic Surgeons
P
endothelium and decreased adenosine triphosphate in the vascular wall [9, 10]. Recent animal laboratory studies reported that protamine caused endothelium-dependent relaxation by stimulating the production of endothelium derived relaxing factor [11, 12]. Arginine represents 67% of the amino acid composition of protamine [13]. The amino acid L-arginine is the physiologic precursor of nitric oxide (NO), the active component of endothelium derived relaxing factor [14, 15]. Despite several laboratory animal investigations, the mechanism of endothelium derived relaxing factor formation after exposure of vascular endothelium to protamine is not completely understood. The influence of heparin on the vascular wall and on protamineassociated vasorelaxation, which was demonstrated in a number of studies, is controversial. These studies demonstrated that anion-charged heparin prevented the anticipated effect of positive-charged protamine on the vascular wall by formation of nonactive complex [16]. However, another investigation found that the heparin itself caused increased NO production by endothelium [17], but neither influenced vascular tone [12, 17] nor prevented protamine-induced vasorelaxation [18]. This is the report of an investigation of the mechanism of protamine action on human internal thoracic arteries
rotamine sulfate (protamine) is widely used to reverse heparin anticoagulation in the practice of cardiovascular surgery. However, the use of protamine has been associated with clinically significant side effects, including systemic hypotension, bradycardia, bronchoconstriction, pulmonary edema, and left ventricular dysfunction [1–5]. Despite its extensive clinical application, the mechanism of protamine-induced hemodynamic changes is unclear: the three major hypotheses which have been proposed for protamine-induced vasodilatation are an immunologic reaction, a direct toxic action, and endothelium-dependent relaxation. Many clinical investigations have demonstrated the presence of some immunologic base for the undesirable protamine reactions. These include hypersensitivity reactions in patients treated with neutral protamine Hagedorn insulin or reactions in patients with fish allergy. Those patients had an elevation of IgE and IgG antibodies specific to protamine as well as complement system activation [6 – 8]. Some authors have reported that protamine causes direct cardiac depression, toxic influence on vascular
Accepted for publication March 28, 2000. Address reprint requests to Dr Mohr, Department of Thoracic and Cardiovascular Surgery, Tel-Aviv Sourasky Medical Center, 6 Weizmann St, Tel-Aviv, Israel 64239; e-mail:
[email protected].
© 2000 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
0003-4975/00/$20.00 PII S0003-4975(00)01678-7
Ann Thorac Surg 2000;70:2050 –3
PEVNI ET AL NO SYNTHASE IN PROTAMINE-INDUCED VASORELAXATION
2051
(ITA). We also tested the effect of heparin on the vascular wall and on protamine-induced vasorelaxation.
Material and Methods Preparation of Vessels Human ITA segments were taken from patients undergoing a coronary artery bypass procedure using ITA grafts. After the grafts were harvested as skeletonized vessels, and before a papaverine bath, the required length was carefully measured. The distal segments, about 6 mm proximal to bifurcation, were immediately collected, placed in a container with physiologic KrebsHenseleit (KH) solution and transferred to the laboratory. The vessels were cleaned of surrounding tissue and cut into 3-mm rings, which were suspended on wire hooks in a 100-ml jacket glass chamber maintained at 37°C. The upper hook was connected to a force transducer (Instrument & Control Ltd, Medagri Ltd, Limassol, Cyprus), and the lower hook was fixed. The force used for stretching the ring was recorded on a Beckman Dynograph Recorder R611 (Beckman Coulter, Fullerton, CA).The modified KH buffer solution was composed of the following: NaCL, 118mmole; KCL, 4.7 mmole; CaCL2, 2.5 mmole; MgSO4 ⫻ 7H2O, 1.2 mmole; KH2PO4, 1.2 mmole; glucose, 11.1 mmole; NaHCO3, 25 mmole. The perfusate was bubbled continuously with 95% O2 and 5% CO2. Two chambers arrangements were run simultaneously.
Protocol We used the commonly accepted procedure described above to stretch human ITA rings [19]. Rings were placed at the optimal point of length-tension relationship by progressively stretching them to 90% of their internal circumference. After the normalizing procedure, the rings were left resting for 30 minutes. A steady level of active contraction was then established by adding norepinephrine (NE) 10⫺6 mole/L. This concentration gave 60% to 80% of maximum contraction for this agent in ITA rings. In our pilot study there were no changes in ITA ring contraction during 30 minutes of NE exposure. In all experiments, the presence of functional endothelium was confirmed by determining the response to acetylcholine (3 ⫻ 10-6 mole/L) in each ring contracted with NE. The cumulative concentration-response curves for protamine were obtained by adding protamine (50 to 800 g/ml) to the organ chamber (group A). To study the possible role of the NO pathway in protamine-induced vasorelaxation, we used two NO synthase (NOS) inhibitors: N -Nitro-L-arginine (LNAM,10-5 mole/L, group B) which is a general inhibitor of NOS activity [20], and aminoguanidine (5 mmole/L, group C) which is a selective inhibitor of the inducible isoform of NO (iNOS) [21]. The inhibitors were added 30 minutes before NE exposure. To study a possible effect of heparin (80 U/mL, group D) on protamine-induced relaxation, heparin was added 5 minutes before protamine action was started. The mean curves were obtained from at least 10 different ITA rings for each group from
Fig 1. Original traces showing the effect of protamine on precontracted human internal thoracic artery rings. The protamine caused dose-dependent relaxation of the ring (curve A). Protamine caused a slight relaxing effect (curve B) when the NOS inhibitor L-NAM was added. Protamine concentration: 1 to 50 g/ml, 2 to 100 g/ml, 3 to 200 g/ml, and so on, to 9 to 800 g/ml.
different patients. Only one concentration-response curve was obtained for each ring.
Drug and Chemicals Protamine sulfate and heparin sodium was purchased from Kamada (Rehovot, Israel); aminoguanidine and other chemicals were purchased from Sigma Chemical Co (St Louis, MO).
Statistical Analysis Data are presented as the mean ⫾ standard deviation. Statistical comparisons were performed with two-way analysis of variance (ANOVA). A p value of 0.05 was considered statistically significant.
Results Concentration-Dependent Relaxation Progressive addition of protamine (from 50 g/mL to 800 g/mL, control group A) produced a sustained, dose-dependent relaxation of the ITA rings that were contracted with NE (Fig 1, curve A). The maximal relaxation was 56% ⫾ 4.7% (Fig 2).
Endothelium-Dependent Relaxation: Endothelial NOS (eNOS) and iNOS Pathways The pilot study did not show any changes in ITA rings contraction during 30 minutes of NE exposure. The addition of L-NAM, the competitive inhibitor of eNOS and iNOS, did not induce any changes in tension of the arterial segments. However, pretreatment of arterial rings with L-NAM almost completely blocked protamine-induced relaxation (Fig 1, curve B; Fig 2; p ⬍ 0.001). The NO-independent relaxation was slight (11.7% ⫾
2052
PEVNI ET AL NO SYNTHASE IN PROTAMINE-INDUCED VASORELAXATION
Ann Thorac Surg 2000;70:2050 –3
Fig 2. Relaxation induced by protamine sulfate in precontacted human internal thoracic artery rings with and without the NOS inhibitor, L-NAM.
Fig 4. Relaxation induced by protamine sulfate in precontracted human internal thoracic artery rings with and without heparin.
1.7%) but attained statistical significance (p ⬍ 0.001). The addition of aminoguanidine, a selective inhibitor of iNOS, neither changed the tension of ITA rings nor had any effect on the extent of protamine-induced relaxation (47.2% ⫾ 1.7%, Fig 3). There was no significant different in relaxation between groups A and C (p ⬍ 0.247).
in predisposing individuals to protamine reactions [2, 6 – 8]. Kimmel and associates [7] reported three risk factors, all of which were independently associated with systemic hypotension following protamine administration: history of protamine insulin use, fish allergy, and nonprotamine drug-induced allergy. Immunologic factors probably play an important role in protamine-induced hypotension. However, most patients who develop protamine associated cardiopulmonary collapse do not have the above risk factors. In the present study, we found that protamine caused concentration-dependent vasorelaxation, which could explain the association between rapid protamine administration and systemic hypotension. Thus it would appear that slow protamine administration from a peripheral line might prevent circulatory collapse. Recent laboratory studies with animal vessels showed that protamine caused endothelium-dependent relaxation of the vascular wall by stimulating the production of endothelium-derived relaxing factor [11, 12]. We hypothesized that protamine, which is rich in L-arginine, a physiologic precursor of endothelium-derived relaxing factor [14, 15], augmented endothelium-derived relaxing factor formation by supplying exogenous L-arginine as a metabolic substrate for NO production. Nitric oxide is an active part of endothelium-derived relaxing factor and is synthesized by different isoforms of NO synthase by oxidative deamination of the amino acid L-arginine [22]. A constitutive isoform of NO synthase, eNOS, which is calcium/calmodulin regulated and was originally described in endothelial cells, causes continuous NO production and plays an important role in vascular tone regulation [23]. The second type of NO synthase is calcium independent–inducible NOS, which can be activated by cytokines, leading to NO overproduction and inappropriate vasodilatation by activation of the enzyme guanylate cyclase [24, 25]. The present study confirmed that protamine induces endothelium-dependent vasorelaxation of human ITA. This vasorelaxation was inhibited by L-NAM (a general
Effect of Heparin on Protamine-Induced Vasorelaxation Heparin (80 U/mL, group D) did not cause any change in the tension of precontracted arterial rings. However, even with heparin, the addition of protamine caused concentration-dependent relaxation in precontracted ITA rings (Fig 4). This response was comparable to the response to protamine in the absence of heparin.
Comment The administration of protamine to reverse the anticoagulant effects of heparin may lead to systemic hypotension. The exact mechanism by which protamine produces this adverse reaction is not completely understood. Some authors suggest that immunologic factors are important
Fig 3. Relaxation induced by protamine sulfate in precontracted human internal thoracic artery rings with and without aminoguanidine.
Ann Thorac Surg 2000;70:2050 –3
inhibitor of both iNOS and eNOS); however; it was not inhibited by aminoguanidine (the selective iNOS inhibitor). These results provide evidence that the protamineinduced vasorelaxation is eNOS dependent. We also found that protamine induces endotheliumindependent relaxation. This minimal relaxation was seen after the blocking of NOS function by L-NAM (group B). This part of the vasorelaxing effect of protamine was slight but statistically significant. The direct toxic effect of protamine on ITA smooth muscles and decrease in the amount of ATP [9] available for contraction might explain this minimal relaxation effect. One of the aims of our study was to investigate the effect of heparin on vascular tone and on the vasorelaxing effect of protamine. Some authors reported that heparin increased NO production in the vascular wall [17]. This heparin-induced NO production can influence vascular tone and aggravate protamine-induced circulatory collapse. The influence of heparin on the vasodilatatory effect of protamine is controversial. Akata and colleagues [18] reported that heparin can prevent the side effects of protamine, but Ordonez and colleagues [19] found no influence of heparin on protamine-induced vasodilatation. In the current study, we found no changes in vascular tone after the addition of heparin. The presence of heparin had no influence on ITA ring relaxation after the addition of protamine. Our results show that protamine sulfate causes dosedependent relaxation of human ITA rings. The relaxation is endothelium-dependent and caused by stimulation of eNOS. Heparin had no influence on vascular tone nor did it prevent vasorelaxation following protamine administration.
PEVNI ET AL NO SYNTHASE IN PROTAMINE-INDUCED VASORELAXATION
8.
9. 10.
11. 12. 13.
14. 15.
16. 17.
18.
19.
References 1. Horrow J.C. Protamine. A review of its toxicity. Anest Analg 1985;64:348– 61. 2. Takenoshita M, Sugiama M, Okuno Y, Inagoki Y, Yoshiya I, Shimazaki Y. Anaphylactoid reaction to protamine confirmed by plasma tryptase in a diabetic patient during open-heart surgery. Anesthesiology 1996;84:233–5. 3. Kindler C, Bircher A, Stulz P. Protamine-induced fulminant noncardiac pulmonary edema following cardiopulmonary bypass. Eur J Cardiovasc Surg 1996;10:463– 6. 4. Katz NM, Kim YD, Siegelman R, Ved SA, Ahmed SW, Wallace RB. Hemodynamics of protamine administration. J Thorac Cardiovasc Surg 1987;94:881– 6. 5. Hird BR, Crawford FA, Mukherjee R, Zile MR, Spinale FG. Effect of protamine on myocyte contractile function and adrenergic responsiveness Ann Thorac Surg 1994;57: 1066–75. 6. Weiss ME, Nyhan D, Peng Z, Horrow JC, Lowenstain E, Hirshman C, et al. Association of protamine IgE and IgG antibodies with life-threatening reactions to intravenous protamine. N Engl J Med 1989;320:886–92. 7. Kimmel SE, Sekeres MA, Berlin JA, Ellison N, Disesa VJ, Strom BL. Risk factor for clinically important adverse events
20. 21.
22. 23. 24.
25.
2053
after protamine administration following cardiopulmonary bypass. Am J Cardiol 1998;32:1916–22. Stewart WJ, McSweeney SM, Kellett MA, Faxon DP, Ryan TJ. Increased risk of severe protamine reactions in NPH insulin-dependent diabetics undergoing cardiac catheterization. Circulation 1984;70:788–92. Wakefield TW, Hinshaw DB, Burger JM, Burkel WE, Stanley JC. Protamine-induced reductions of endothelial cell ATP. Surgery 1989;106:378– 85. Wakefield TW, Ucros I, Kresawik TF Hinshaw DB, Stanley JC. Decreased oxygen consumption as a toxic manifestation of protamine sulfate reversal of heparin anticoagulation. J Vasc Surg 1989;9:772–7. Akata T, Yoshitake J, Nakashima M, Itoh T. Effect of protamine on vascular smooth muscle of rabbit mesenteric artery. Anesthesiology 1991;75:833– 846. Pearson PJ, Evora PRB, Ayrancioglu K, Shaff HV. Protamine releases endothelium-derived relaxing factor from systemic arteries. Circulation 1992;86:289–94. Ando T, Yamasaki M, Suzuki K. Protamines. In. Kleinzeller A, Springer GF, Wittman HG, editors. Molecular biology, biochemistry and biophysics, vol. 12, Berlin, Springer-Verlag 1973:1–30. Palmer RMJ, Ferrige AG, Moncada S. Nitric oxide accounts for the biological activity of endothelium derived relaxifactor. Nature 1987;327:524– 6. Palmer RMJ, Rees DD, Ashton DS, Moncada S. L Arginine is the physiological precursor for the formation of nitric oxide in endothelium-dependent relaxation. Biochem Byophys Res Commun 1988;153:1251– 6. Akata T, Kodama K, Yoshitake J, Takahashi S. Heparin prevents the vasodilatating action of protamine on human small mesenteric arteries. Anesth Analg 1993;76:1213–1221. Li JM, Hajarizadeh JLH, La Rosa CA, Rohrer MJ, Vander Salm TJ, Cutler BC. Heparin and protamine sulfate stimulate the production of nitric oxide. J Cardiovasc Surg (Torino)1996;37:445–52. Ordonez FA, Hernandez FA, Borrego DJM, Garcia TP, Perez BS, Hinajosa R., et al. The systemic vasodilatory action of protamine. is it inhibited or mediated by heparin? Res Exp Med (Berl) 1998;197:337– 47. Cohen RA, Shepherd JT, Vanhoutte PM. Inhibitory role of the endothelium in the response of isolated coronary arteries to platelets. Science 1983;221:273– 4. Wolff DS, Lubeskie A. Inactivation of nitric oxide synthase isoforms by diaminoguanidine and Ng-amino-l-arginine. Arch Biochem Biophys 1996;325:227–34. Okuda Y, Sakoda S, Fujimura H, Yanagihara T. Aminoguanidine a selective inhibitor of the inducible nitric oxide synthase, has different effect on experimental allergic encephalomyelitis in induction and progressive phase. J Neuroimmunol 1998;81:201–10. Fan B, Wang J, Stuehr DS, Rousean DL. NO synthase isoenzymes have distinct substrate binding sites. Biochemistry 1997;21:12660 –5. Buckley BS, Mirza Z, Whorton AR. Regulation of Ca2⫹ dependent nitric oxide synthase in bovine aortic endothelial cells. Am J Physiol 1995;269:C757– 65. Sperelakis N, Xiong Z, Haddad G; Masuda H. Regulation of slow calcium channels of myocardial cells and vascular smooth muscle cells by cyclic nucleotides and phosphorylation. Mol Cell Biochem 1994;140:103–17. Joe EK, Schussheim AE, Longrois D, et al. Regulation of cardiac myocyte contractile function by inducible nitric oxide synthase (iNOS). mechanism of contractile depression by nitric oxide. J Mol Cell Cardiol 1998;30:303–15.