Profound Inhibition of Myogenic Tone in Rat Cardiac Allografts Is Due to eNOS- and iNOS-Based Nitric Oxide and an Intrinsic Defect in Vascular Smooth Muscle Contraction
Background—The physiological consequences of inducible NO synthase (iNOS) expression were studied in allograft coronary arteries by pressure myography.
Methods and Results—Septal coronary arteries (diameter, 200.6±3.3 μm) were harvested from allograft and isograft hearts, and their myogenic properties were measured before and after iNOS and nonselective NOS inhibition with aminoguanidine (AG, 100 μmol/L) and NG-nitro-l-arginine methyl ester (L-NAME) (200 μmol/L). Fura 2 fluorescence microscopy was used to measure [Ca2+]i in isolated endothelial cells. Monoclonal anti-iNOS immunostains demonstrated iNOS protein in day 2, 7, 14, and 28 allograft vessels, but only in day 2 isograft vessels. Myogenic tone was profoundly inhibited in allograft vessels from day 4 onward. In day 4 allograft vessels, these differences were abolished by L-NAME but not AG, suggesting greater basal release of eNOS-based NO from allograft endothelium. Fluorescence measurements confirmed elevation of [Ca2+]i in day 4 allograft endothelium, providing a mechanism for enhanced eNOS activity. For days 7 to 28, AG potentiated myogenic tone in allograft but not isograft vessels, indicating that vasoactive iNOS-based NO was present. In mature vessels, constriction via agonist- and depolarization-mediated mechanisms showed parallel inhibition, suggesting an intrinsic defect in vascular smooth muscle cell contraction.
Conclusions—Our data indicate that the profound inhibition of myogenic tone in allograft arteries involves direct vasodilation by eNOS- and iNOS-based NO, as well as an intrinsic defect in vascular smooth muscle contraction. The hemodynamic profile resulting from these changes in allograft resistance vessel function would favor movement of extracellular fluid from the intravascular space into the myocardial interstitium, resulting in edema, increased ventricular stiffness, and poor ventricular performance.
Nitric oxide (NO) has recently been proposed to play an important role in transplant physiology, with several investigators having identified the inducible isoform of NO synthase (iNOS) in experimental and human cardiac allografts but not isografts.1 2 3 Unlike the constitutively expressed endothelial NOS (eNOS), which normally produces low basal levels of NO in a manner very tightly regulated by endothelial intracellular Ca2+, iNOS activity is Ca2+-independent and is regulated largely by the amount of enzyme present; thus, iNOS is capable of producing large quantities of NO for sustained periods of time.4 The coronary vasculature could be a target of iNOS-based NO, leading to sustained vasodilation of resistance vessels and increased coronary flow. During biopsy-proven cellular rejection in human cardiac allografts, basal work-corrected coronary flow is elevated, with a corresponding decrease in flow reserve, indicating enhanced dilation of the coronary circulation under immune assault.5 On the basis of this evidence, we hypothesized that the expression of iNOS in cardiac allografts would result in the production of NO and enhanced vasodilation. Our results indicate that the basal myogenic tone of isolated allograft arteries is profoundly inhibited, occurring through NO-dependent and NO-independent mechanisms.
iNOS and eNOS expression
iNOS expression was clearly evident in day 2, 7, 14, and 28 allograft arteries (Figure 1⇓). In contrast, isograft arteries showed clear iNOS expression only at day 2 posttransplantation, with infrequent staining for subsequent time points. Notably, day 4 lacked iNOS expression in both groups. Thus, allospecific iNOS expression is first identified at day 7 and persists at day 28, whereas transient nonspecific iNOS expression (allografts and isografts) occurs at day 2. eNOS protein was identified in allograft and isograft arteries at all time points.
Vascular Functional Studies
Vessel Size and Structure
The average diameter for all vessels was 200.6±3.3 μm at 10 mm Hg in Ca2+-free physiological salt solution (PSS) (n=85). For all graft and time-point groups, pressure-diameter curves in Ca2+-free PSS were similar, indicating identical passive characteristics.
Day 2 Posttransplantation.
Isograft and allograft vessels showed similar myogenic profiles at day 2; myogenic tone developed in a graded fashion as transmural pressure was increased in the physiological range. In both groups, there was a nonsignificant trend toward greater tone after aminoguanidine (AG), indicating the presence of iNOS-based vasoactive NO (Figure 2A⇓). NG-Nitro-l-arginine methyl ester (L-NAME) resulted in similar potentiation in both groups, indicating equal underlying myogenic tone (Figure 2B⇓).
Day 4 Posttransplantation.
In contrast to day 2, significant differences in myogenic tone were evident between isograft and allograft vessels at day 4. AG did not potentiate tone in either group, indicating an absence of iNOS-based NO (Figure 3A⇓). However, nonselective NOS inhibition with L-NAME abolished the tone differences. This observation suggests greater basal release of eNOS-based NO in allograft vessels (Figure 3B⇓).
Because eNOS activity is [Ca2+]i-dependent, we reasoned that enhanced Ca2+ availability in allograft endothelial cells might underlie the enhanced eNOS activity in day 4 allografts. We used fura 2 fluorescence imaging to study [Ca2+]i homeostasis in freshly isolated aortic valvular endothelial cells from day 4 allograft and isograft hearts. A representative tracing, along with cumulative data, is shown in Figure 4⇓. Calculated basal [Ca2+]i was significantly elevated in allograft endothelium (78.3±4.9 versus 42.8±5.7 nmol/L; 35.6±5.7 nmol/L [control], P<0.05 allograft versus isograft and control). Inhibition of the endoplasmic reticulum Ca2+-ATPase with cyclopiazonic acid (CPA) in the presence of extracellular Ca2+ caused a large increase in 340/380 ratio in allograft endothelium. In isografts and controls, the increase was small and transient. The elevated 340/380 ratio in allografts could not be rapidly reversed by tetraethylammonium (TEA) (which depolarizes the endothelial cells, thus decreasing the electrical driving force for Ca2+ entry); even removal of CPA caused a very slow decrease in 340/380 ratio. When Mn2+ quenching of the 360-nm signal was used as a reflection of Ca2+ influx, no differences between allograft and isograft cells were observed, although both were elevated compared with control cells.
Days 7 to 28 posttransplantation.
Inhibition of allograft myogenic tone persisted for days 7 to 28. The tracings in Figure 5⇓ demonstrate the profound vasodilation of an allograft artery compared with a matched isograft artery. For days 7 to 28, incubation with AG abolished the statistical differences between isograft and allograft vessels, indicating that iNOS-based NO is present and is vasoactive (Figure 6A⇓ through 6C). However, a trend toward less tone in allograft vessels persists after AG, suggesting either incomplete iNOS blockade, greater eNOS activity, or another mechanism of tone inhibition. In these vessels, L-NAME unmasked a time-dependent deterioration of underlying myogenic tone in allograft groups (Figure 7A⇓). Neither indomethacin 1 μmol/L nor endothelium removal could reverse the profound inhibition of tone in L-NAME–treated day 28 allograft arteries (not shown).
Agonist- and Potassium-Induced Tone
Allograft and isograft vessels (with AG) were constricted by agonist and depolarization (potassium) mechanisms for comparison with myogenic tone. As shown in Figure 7B⇑, there is a parallel inhibition of these 3 mechanisms of constriction in mature allograft vessels (day 28), whereas all 3 are preserved in early allografts and matched isografts. This pattern of multimodal inhibition suggests a decrease in the number of viable or functional smooth muscle cells,6 7 or possibly a signaling defect in the distal common pathway of constriction.
Graft Weight and Wet/Dry Ratios
Dry weight was greater in allografts than in matched isografts from day 4 onward. Wet/dry ratio was significantly elevated in day 4, 7, and 28 allograft hearts, indicating greater myocardial free water content (Table⇓).
Drugs and Concentrations
Previous work has shown inhibition of iNOS in vascular preparations by 100 to 300 μmol/L AG.8 9 10 11 12 Figures 3A⇑, 4A⇑, and 7A⇑ through 7C show potentiation of tone only in vessels expressing iNOS protein. We observed preservation of acetylcholine-induced dilation in native arteries in the presence of 100 μmol/L AG; 200 μmol/L L-NAME abolished acetylcholine-induced dilation, indicating complete blockade of eNOS (not shown).
The principal finding of this study is that myogenic tone in allograft coronary arteries is profoundly inhibited compared with site- and size-matched isograft arteries, resulting in a greatly enhanced arterial diameter. We have identified 3 mechanisms: (1) tone inhibition can occur through enhanced basal release of eNOS-based NO due to enhanced endothelial Ca2+ availability; (2) when iNOS is expressed, tone inhibition can occur through release of iNOS-based NO; and (3) in contrast to the NO-dependent vasodilation mechanisms mentioned above, the concerted deterioration of pressure-, agonist-, and depolarization-induced tone seen in mature allograft vessels is consistent with an intrinsic defect in vascular smooth muscle contraction, involving either common signal transduction events or a decrease in the number of viable smooth muscle cells. These findings, which precede the obliterative arteriopathy characteristic of chronic rejection, predict a pattern of coronary hemodynamics that would favor movement of extracellular fluid from the intravascular compartment into the myocardial interstitium, resulting in myocardial edema and ventricular stiffness.
Mechanisms of Myogenic Tone Inhibition
eNOS and Endothelial [Ca2+]i
Arteries from allografts at day 4 manifest less myogenic tone than matched isograft arteries. Selective iNOS inhibition with AG did not potentiate tone in either group, and no iNOS protein was identified immunohistochemically. The tone differences were abolished after nonselective NOS inhibition with L-NAME, indicating a greater basal release of eNOS-based NO from allograft endothelium. Resting 340/380 ratio (and calculated [Ca2+]i) was elevated in isolated allograft endothelial cells; because eNOS is very tightly regulated by [Ca2+]i, this observation provides a mechanism for enhanced basal release of eNOS-based NO.
iNOS-Based Vasoactive NO
From day 7 onward, the inhibition of myogenic tone in allograft arteries paralleled the expression of iNOS protein. In these vessels, selective inhibition of iNOS with AG potentiated tone; thus, iNOS-based NO is vasoactive and is an important mechanism of tone inhibition. To our knowledge, this is the first report to show immunohistochemical evidence of iNOS protein together with vascular hyporesponsiveness and iNOS inhibitor potentiation of resistance vessels. The general theme, however, has been addressed in other models.10 11
Smooth Muscle Contractile Defect
A trend toward less tone in day 7 to 28 allografts persists even after iNOS inhibition. These residual differences could be due to alloimmune alteration in endothelial [Ca2+]i as described above; in mature grafts, however, the residual differences in tone were not abolished by nonselective NOS inhibition. In fact, this approach unmasked a pattern of progressive, time-dependent deterioration of myogenic tone that is not solely due to NO vasodilation, arguing against eNOS as a preeminent factor. Because neither indomethacin nor endothelium removal altered this residual inhibition, prostaglandins and other endothelium-derived vasodilators were eliminated as candidate mechanisms. In these mature (day 28) allograft vessels, profound inhibition of myogenic, agonist, and depolarization-induced tone was observed, suggesting a defect in vascular smooth muscle contraction through a common signaling event. Potential events include those of the distal common pathway: Ca2+ influx, calmodulin, myosin light chain kinase, contractile filaments, and phosphatases.
It is also possible that the impairment of constriction in mature allograft vessels is due to a decrease in the number of viable smooth muscle cells. Apparent loss of medial cells has been observed by Dong et al13 in human coronary arteries, and recent evidence indicates that apoptosis may play a key role in cardiac allograft vasculopathy. Szabolcs et al14 used DNA laddering, terminal dUTP nick end-labeling (TUNEL), and in situ nick translation to identify apoptotic cells in Lewis-to–Wistar-Furth allografts. Apoptotic nuclei were identified in cardiac myocytes, endothelial cells, and infiltrating monocytes; iNOS protein was identified in the same cell types. Importantly, the temporal pattern of apoptosis paralleled that of iNOS expression, NOS activity, and nitrotyrosine staining, suggesting that apoptosis may be triggered by iNOS and peroxynitrite. With this in mind, it is possible that the effect of iNOS expression on allograft arteries is 2-fold and time-dependent: an early phase due to vasodilation by NO itself and a delayed phase due to smooth muscle apoptosis by NO and/or NO adducts.
iNOS Protein Expression
Expression of iNOS mRNA and protein has been demonstrated in cardiac allografts from several heterotopic animal models, and our analysis is similar to these.1 2 In the Lewis-to-F344 model, reverse transcription–polymerase chain reaction on ventricular homogenate RNA identified iNOS transcript in allografts at days 7, 14, 28, and 75, with a small amount at day 3. Immunohistochemical stains with polyclonal rabbit antisera to macrophage NOS showed iNOS protein predominantly in mononuclear inflammatory cells within the interstitial and perivascular spaces of allograft hearts (days 7, 28, and 75); there was minimal staining in isografts. iNOS protein in smooth muscle and endothelial cells was identified in significant quantity only at later time points (days 75 and 120), although lesser staining was seen in all allografts.2 Our results with a monoclonal mouse anti-iNOS antibody show that allospecific expression of iNOS in day 7 to 28 allograft arteries occurs mainly in the intimal space. In this location, NO from functional iNOS would be expected to have access to the vascular smooth muscle with resistance vessel sequelae as we have shown. Our structure-function comparison demonstrates that the pattern of iNOS expression parallels the pattern of functional change in allograft vessels and confirms an important physiological consequence of iNOS expression in transplantation.
Our experimental observations provide a mechanistic basis for several clinical observations. In the absence of fixed distal disease, a pattern of enhanced arterial diameter due to myogenic tone inhibition would predict a hemodynamic profile of supranormal coronary flow and reduced flow reserve. In nonrejecting human cardiac grafts, resting coronary flow is indeed elevated, with a proportional decrease in flow reserve15 16 ; these findings have been attributed to increased cardiac work in recipients with systolic hypertension and tachycardia, and when corrected for this, coronary flow is appropriate. However, during biopsy-proven acute rejection, corrected coronary flow is significantly elevated (and coronary resistance depressed) compared with flow after successful recovery on immunosuppression,4 indicating that vasodilation accompanies uncontrolled graft rejection. Interestingly, corrected coronary flow remains significantly elevated after a single episode of rejection compared with patients without previous rejection episodes, consistent with a persistent defect in resistance vessel tone due to the rejection event. Because iNOS expression in cardiac allografts is inhibited by immunosuppression,17 it is possible that the residual elevation in resting corrected coronary flow in transplant patients after rejection is due to an iNOS-independent event such as enhanced endothelial [Ca2+]i and NO or a smooth muscle contractile defect, as we have shown. If so, these alterations in resistance vessel function may be irreversible.
Elevated coronary flow in the face of an unrestrained immune assault may at first consideration appear to be an appropriate and beneficial response. However, normal coronary myogenic behavior is necessary not only to regulate myocardial blood flow but also to provide graded vascular resistance. Appropriate vascular resistance protects the microvasculature from central arterial pressures and so preserves the important balance of hydrostatic and oncotic forces at the capillary and venular levels.18 A massively dilated coronary circulation, predicted by our results, would transmit abnormally high perfusion pressures to nutritive vessels, distorting the balance of intravascular oncotic and hydrostatic pressure and favoring a net movement of fluid into the myocardial interstitium. In support of this hypothetical pathogenesis, our results show a greater wet/dry ratio in allograft hearts, reflecting greater myocardial free water content. Because endothelial permeability to serum protein is enhanced in allograft rejection (also an iNOS-dependent event),19 these 2 mechanisms could act synergistically to cause significant myocardial edema, thereby compromising ventricular compliance and performance.
The association between iNOS expression and ventricular performance in human grafts was recently published. In support of our hypothesis that the altered allograft coronary physiology (in part due to iNOS-based NO) would favor myocardial edema, ventricular stiffness, and poor performance, Lewis et al3 reported that expression of iNOS correlates with cGMP levels and systolic and diastolic left ventricular contractile dysfunction; importantly, iNOS did not associate with International Society of Heart and Lung Transplantation (ISHLT) rejection grade. Together with our results, this key study indicates that current techniques of rejection assessment (ISHLT grade) may ignore coexistent physiological mechanisms of allograft dysfunction.
Summary and Conclusions
We have shown that myogenic tone is profoundly inhibited in cardiac allograft arteries, in part by excess vasoactive NO and in part by a defect in vascular smooth muscle contractility. Excess NO can be derived from eNOS and iNOS isoforms. These findings predict a hemodynamic pattern within the rejecting heart that would favor myocardial edema, ventricular stiffness, and poor myocardial performance.
Transplants, Immunostains, and Tissue Preparation
Heterotopic cardiac transplants were created with anastomosis of donor ascending aorta to recipient abdominal aorta and donor pulmonary artery to recipient inferior vena cava.20 Hearts were harvested at days 2, 4, 7, 14, and 28 for immunohistochemical and functional studies. This protocol was in accordance with Animal Care Guidelines set forth by the University of British Columbia.
Standard immunohistochemical techniques were used to demonstrate iNOS and eNOS protein in fresh cardiac tissues. Fresh rat cardiac tissues from isograft, allograft, and normal control animals were harvested on days 2, 4, 7, 14, and 28, embedded in OCT compound, quick-frozen on dry ice, and stored at −80°C until analysis. Mouse monoclonal anti-iNOS and anti-eNOS antibodies (both from Transduction Laboratories) with biotinylated horse anti-mouse secondary antibody amplification and streptavidin-biotin detection were used for protein determination. For both antibodies, negative controls included (1) omission of the primary antibody and (2) use of an irrelevant, isotype-matched primary antibody at an equivalent concentration.
Septal coronary arteries (inner diameter, ≈200 μm) were obtained from grafted (n=75) and control (n=10, unoperated Lewis) hearts and were used for all functional studies. Through a right ventriculotomy, the septal coronary artery was identified on the ventricular septum. A segment of the vessel 0.6 to 1.0 mm long was dissected free of surrounding myocardium at the level of the superior septal papillary muscle, excised, and transferred to the experimental chamber of a pressure myograph as described previously.21 The artery was mounted between 2 microcannulas and then gradually pressurized to allow development of myogenic tone. The transilluminated image was then used to determine arterial diameter response to changes in transmural pressure.
For fluorescence experiments, endothelial cells were freshly isolated from isograft and allograft aortic valve leaflets by a previously described enzyme digestion method.22 The isolated cells were seeded onto glass coverslips precoated with poly-d-lysine and maintained at 37°C until transfer to the experimental perfusion chamber. The final preparation consisted of small clusters of 3 to 15 cells that maintained their typical tile-like morphology. [Ca2+]i of the isolated endothelial cells was evaluated by use of a fura 2fluorescence imaging system and the method of Grynkiewicz et al.23 We recognize the uncertainties of [Ca2+]i calculation by this fura 2 calibration method.24 Therefore, we elected to compare fluorescence intensity ratios between groups rather than their calculated [Ca2+]i. A Mn2+-quenching technique was used to study the rate of divalent cation influx into the isolated endothelial cells.
Myogenic, Agonist-Induced, and Depolarization-Induced Tone
After equilibration at 80 mm Hg, transmural pressure was decreased to 10 mm Hg. Vessels were subjected to stepwise increases in transmural pressure, from 10 to 120 mm Hg, to determine the degree of myogenic tone at each pressure. This protocol was repeated in the presence of inhibitors of NO synthases, and so 3 trials were completed: (1) physiological salt solution (PSS) alone, (2) iNOS inhibition with aminoguanidine (AG) 100 μmol/L, and (3) nonselective NOS inhibition with NG-nitro-l-arginine methyl ester (L-NAME) 200 μmol/L. The vessel was then cycled through the same pressure steps in Ca2+-free PSS to determine the passive diameter at each pressure. Constrictions in response to the thromboxane analogue U46619 10 μmol/L and potassium 80 mmol/L were measured in a separate group of arteries pressurized at 80 mm Hg.
Endothelial Ca2+ Signaling
To study the level and determinants of basal [Ca2+]i in isolated endothelial cells, 340/380 fluorescence intensity ratios were measured sequentially at rest, after endoplasmic reticulumCa2+-ATPase inhibition with cyclopiazonic acid (CPA, 10 μmol/L), after K+-channel blockade with tetraethylammonium (TEA, 5 mmol/L), and after washing with PSS. The rate of Mn2+ quenching was subsequently determined in the same cells.
Myogenic tone, expressed as percent constriction, was calculated as (DCa2+-free, P−DP)/DCa2+-free, P, where DCa2+-free, P is the diameter in Ca2+-free PSS at pressure P, and DP is the diameter in PSS with calcium at pressure P. Constrictions to agonist and potassium at 80 mm Hg were calculated similarly.
The ionic composition of PSS (in mmol/L) was NaCl 118, NaHCO3 24.9, KH2PO4 4.7, MgSO4 1.17, CaCl2 1.6, glucose 11.1, and EDTA 0.026. Potassium PSS was prepared by proportional mixing with 80 mmol/L KCl-PSS. Ca2+-free PSS contained 2.0 EGTA and no CaCl2.
Results are presented as mean±SEM. Data were analyzed with Student’s t test or 1-way/1-way repeated-measures ANOVA where appropriate and Student-Newman-Keuls or Dunn’s test for significant differences. The probability of a type I (α) error was set at P<0.05.
Drs Skarsgard and Laher are supported by the Medical Research Council of Canada. Drs van Breemen and McManus are supported by the Heart and Stroke Foundation of British Columbia and the Yukon (both by grants-in-aid and a program project award), Biomax Technologies, and the St Paul’s Hospital Foundation. The authors acknowledge the technical assistance of Dean English, Jennifer Kenyon, and Janet Wilson-McManus.
The Methods section of this article can be found at http://www.circulationaha.org
- Received April 8, 1999.
- Revision received September 30, 1999.
- Accepted October 8, 1999.
- Copyright © 2000 by American Heart Association
Yang X, Chowdhury N, Bolin C, Brett J, Marboe C, Sciacca RR, Michler RE, Cannon PJ. Induction of myocardial nitric oxide synthase by cardiac allograft rejection. J Clin Invest. 1994;94:714–721.
Russell ME, Wallace AF, Wyner LR, Newell JB, Karnovsky MJ. Upregulation and modulation of inducible nitric oxide synthase in rat cardiac allografts with chronic rejection and transplant arteriosclerosis. Circulation. 1995;92:457–464.
Lewis NP, Tsao PS, Rickenbacher PR, Xue C, Johns RA, Haywood GA, von der Leyen H, Trindade PT, Cooke JP, Hunt SA, Billingham ME, Valantine HA, Fowler MB. Induction of nitric oxide synthase in the human cardiac allograft is associated with contractile dysfunction of the left ventricle. Circulation. 1996;93:720–729.
Chan SY, Kobashigawa J, Stevenson LW, Brownfield E, Brunken RC, Schelbert HR. Myocardial blood flow at rest and during pharmacological vasodilation in cardiac transplants during and after successful treatment of rejection. Circulation. 1994;90:204–212.
Meininger GA, Davis MJ. Cellular mechanisms involved in the vascular myogenic response. Am J Physiol. 1992;263:H647–H659.
Schrör K. The effect of prostaglandins and thromboxane A2 on coronary vessel tone: mechanisms of action and therapeutic implications. Eur Heart J. 1993;14(suppl I):34–41.
Corbett JA, Tilton RG, Chang K, Hasan KS, Ido Y, Wang JL, Sweetland MA, Lancaster JR Jr, Williamson JR, McDaniel ML. Aminoguanidine, a novel inhibitor of nitric oxide formation, prevents diabetic vascular dysfunction. Diabetes. 1992;41:552–556.
Scott JA, Machoun M, McCormack DG. Inducible nitric oxide synthase and vascular reactivity in the rat thoracic aorta: effect of aminoguanidine. J Appl Physiol. 1996;80:271–277.
Szalbolcs M, Michler RE, Yang X, Aji W, Roy D, Athan E, Sciacca RR, Minanov OP, Cannon PJ. Apoptosis of cardiac myocytes during cardiac allograft rejection: relation to induction of nitric oxide synthase. Circulation. 1996;94:1665–1673.
Kofoed KF, Czernin J, Johnson J, Kobashigawa J, Phelps ME, Laks H, Schelbert HR. Effects of cardiac allograft vasculopathy on myocardial blood flow, vasodilatory capacity, and coronary vasomotion. Circulation. 1997;95:600–606.
Chilian WM, Layne SM, Klausner EC, Eastham CL, Marcus ML. Redistribution of coronary microvascular resistance produced by dipyridamole. Am J Physiol. 1989;256:H383–H390.
Worrall NK, Chang K, Suau GM, Allison WS, Misko TP, Sullivan PM, Tilton RG, Williamson JR, Ferguson TB. Inhibition of inducible nitric oxide synthase prevents myocardial and systemic vascular barrier dysfunction during early cardiac allograft rejection. Circ Res. 1996;78:769–779.
Ono K, Lindsay ES. Improved technique of heart transplantation in rats. J Thorac Cardiovasc Surg.. 1969;7:225–229.
Skarsgard P, van Breemen C, Laher I. Estrogen regulates myogenic tone in pressurized cerebral arteries by enhanced basal release of nitric oxide. Am J Physiol. 1997;273:H2248–H2256.
Wang X, Lau F, Li L, Yoshikawa A, van Breemen C. Acetylcholine-sensitive intracellular Ca2+ store in fresh endothelial cells and evidence for ryanodine receptors. Circ Res. 1995;77:37–42.
Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem. 1985;260:3440–3450.