Lysophosphatidylcholine Inhibits Endothelium-Dependent Hyperpolarization and Nω-Nitro-l-Arginine/IndomethacinResistant Endothelium-Dependent Relaxation in the Porcine Coronary Artery
Background Oxidized LDL and lysophosphatidylcholine (LPC) have been reported to inhibit the endothelium-dependent relaxation (EDR) mediated by nitric oxide. Recently, a new vasorelaxing factor, endothelium-derived hyperpolarizing factor (EDHF), which hyperpolarizes and relaxes the porcine coronary artery in the presence of Nω-nitro-l-arginine (NNA) and indomethacin (IM), has been reported. We examined whether LPC also inhibits both the EDHF-mediated relaxation and membrane hyperpolarization of the porcine coronary artery.
Methods and Results EDHF was evaluated as the bradykinin- or A23187-induced relaxation of the porcine coronary artery contracted by prostaglandin F2α in the presence of NNA and IM. We also directly measured the membrane potential of the porcine coronary artery. The effects of LPC on both relaxation and membrane hyperpolarization were investigated. At concentrations of 0 to 20 μmol/L, LPC dose-dependently inhibited the NNA/IM-resistant EDR induced by bradykinin and A23187, and the relaxation was reversible after the absorption of LPC with albumin. LPC also inhibited the bradykinin- and A23187-induced hyperpolarization of the porcine coronary artery.
Conclusions In the present study, LPC was found to inhibit not only nitric oxide–mediated but also EDHF-mediated relaxation of the porcine coronary artery. Our findings suggest a new regulatory mechanism in the atherosclerotic coronary artery.
One of the most important factors in atherogenesis is LDL, which is oxidated in vivo, is incorporated into macrophages via scavenger receptors, and causes foam cell formation in the atherosclerotic lesions.1 In the oxidation of LDL, lecithin is degraded to lysolecithin (LPC),2 which is thought to be a major atherogenic component. The EDR mediated by nitric oxide has been reported to be inhibited by oxidized LDL and LPC,3 4 5 6 7 8 9 10 11 and the suppression of EDR in the atherosclerotic lesions was thought to be attributable to this mechanism.
A new vasorelaxing factor, which is produced by endothelial cells and hyperpolarizes vascular smooth muscle cells, has been designated as an EDHF.12 13 EDHF is thought to be a humoral factor released in response to several agents (acetylcholine, bradykinin, A23187, and thrombin) and is considered to hyperpolarize smooth muscle cells by opening potassium channels.14 However, EDHF has not yet been completely characterized or chemically identified. Some EDRs are known to be resistant to inhibitors of the nitric oxide pathway such as oxyhemoglobin, methylene blue, and NNA.15 Nagao and Vanhoutte16 and Pacicca et al17 reported that EDR in the presence of IM and NNA is mediated by EDHF and plays an essential part in the regulation of vascular tone in the porcine coronary artery. They proposed that EDHF can be detected by isometric tension recording of the EDR resistant to both NNA and IM.
To understand the pathological regulation of vascular tone in atherosclerosis, it is necessary to clarify the effects of LPC on not only nitric oxide– but also EDHF-mediated EDR. The present study was designed to examine the effects of LPC on the EDHF-mediated relaxation in the porcine coronary artery.
NNA, IM, bradykinin acetate, l-α-LPC from egg yolk, glibenclamide, PMA, staurosporine, and BSA were obtained from Sigma Chemical Co. TEA, TBA, and BaCl2 were from Nacalai Tesque Inc. A23187 was from Calbiochem. PGF2α was a gift from Ono Pharmaceutical Co. Nicorandil was a gift from Chugai Pharmaceutical Co.
The modified Krebs-Henseleit solution had the following composition (mmol/L): NaCl 121.9, KCl 4.7, NaHCO3 15.5, CaCl2 2.5, MgCl2 1.2, KH2PO4 1.2, and glucose 11.5, pH 7.4. LPC was suspended in Krebs-Henseleit solution and sonicated for 30 seconds twice.4 6 8 9
PMA, staurosporine, glibenclamide, and A23187 were dissolved in DMSO. The final concentration of DMSO in the organ bath was <0.1%, at which EDR was not affected.
Tension Recordings of Coronary Arterial Strips
Porcine hearts were obtained from a nearby slaughterhouse. The proximal half of the left anterior descending coronary artery was carefully dissected out and placed in Krebs-Henseleit solution. The coronary artery was cut into 3-mm rings, opened, and mounted isometrically at a resting tension of 1.0 g in a 3-mL organ bath containing oxygenated (95% O2/5% CO2) Krebs-Henseleit solution at 37°C. Isometric tension was recorded as previously described.3 Before the experiments were started, the strips were contracted by 60 mmol/L KCl two or three times until the contractions stabilized.
The membrane potential was measured by the previously reported method.12 15 16 The coronary artery, opened by a longitudinal cut, was mounted on a silicon rubber plate fixed in the experimental chamber with the intimal side up. The tissue was superfused with modified Krebs-Henseleit solution at a constant flow rate of 2 mL/min at 37°C. After 90 minutes of incubation, a glass capillary microelectrode filled with 3 mol/L KCl (tip resistance, 40 to 60 MΩ) was inserted into the smooth muscle cells from the intimal side of the vessel. The membrane potential was recorded on a pen recorder. The following criteria were used to assess the validity of a successful impalement: (1) a sudden negative shift in voltage followed by (2) a stable negative voltage for more than 1 minute and (3) an instantaneous return to the previous voltage level on dislodgment of the microelectrode.
Inhibition of NNA/IM-Resistant EDR by LPC
To block prostanoids and nitric oxide formation, the strips were preincubated with 10 μmol/L IM and 30 μmol/L NNA for at least 30 minutes, then contracted by 3 μmol/L PGF2α. As Fig 1⇓ shows, the blockade of nitric oxide synthesis by 30 μmol/L NNA was complete. The relaxation by EDHF was inhibited by either 10 mmol/L TBA (Fig 1⇓, I) or 60 mmol/L KCl (Fig 1⇓, II).16 17 The addition of 30 μmol/L NNA in the presence of TBA or KCl eliminated the bradykinin-induced relaxation. Fig 2⇓ shows that NNA completely abolished the relaxation to serotonin (a nonhyperpolarizing endothelium-dependent agonist) over a wide range of concentrations (10−13 to 10−5 mol/L). Therefore, nitric oxide production was completely blocked by NNA.
When the contraction reached a plateau, 10−7 mol/L bradykinin was added and the relaxation was recorded (first relaxation). The bath was washed with Krebs-Henseleit solution several times until tension decreased to a minimum. The strips were then preincubated with LPC in addition to IM and NNA for 10 minutes, contracted by PGF2α, and relaxed by bradykinin again as above (second relaxation). Percentage relaxation was defined as the relaxation divided by maximal contraction (B/A×100 in Fig 3⇓, bottom). The inhibitory effects of LPC on NNA/IM-resistant EDR were evaluated as a ratio of percentage of the second relaxation divided by that of the first relaxation. Because LPC may damage the endothelial cells, the LPC-treated strips were washed three times with 0.1% albumin to absorb LPC,8 and relaxations were recorded again as above (third relaxation). The reversibility was evaluated as a ratio of percentage relaxation of the third divided by that of the first relaxation. Fig 4⇓ shows a representative tracing. The force generation following the preconstriction with PGF2α increased slightly after the administration of LPC. The increase of the force generation after preconstriction with PGF2α did not vary with respect to LPC concentration (data not shown). When the forces before the administration of PGF2α were compared among the control, LPC, and albumin experiments, there was a statistically significant difference (control, 0.36±0.03 g [n=33]; LPC, 0.44±0.04 g [n=33]; and albumin, 0.41±0.05 g [n=23], P<.05 for control versus LPC and control versus albumin).
In another experiment, the calcium ionophore A23187 at a concentration of 10−7 mol/L instead of bradykinin was added to the PGF2α-contracted strips.
Characterization of NNA/IM-Resistant EDR
To characterize the NNA/IM-resistant EDRs, relaxation induced by bradykinin was examined with a denuded endothelium or in the absence of NNA. To examine the effects of potassium channel blockers, 10 mmol/L TBA, 10 mmol/L TEA, 100 μmol/L BaCl2, or 1 μmol/L glibenclamide was added after the first relaxation.
To determine whether the NNA/IM-resistant EDR is regulated by PKC, 100 nmol/L PMA or 100 nmol/L staurosporine was added after the first bradykinin-induced relaxation, and the second relaxation was recorded. Ten minutes before incubation with 20 μmol/L LPC, 100 nmol/L staurosporine or 100 nmol/L PMA was also added to examine the influence, if any, on the effect of LPC.
The results are expressed as mean±SEM. The SAS system was used for the statistical analysis described in this article. Statistical evaluation of the data was performed by Dunnett’s T test except as shown in Fig 7⇓, where ANOVA was used. A two-tailed value of P<.05 was considered to be significant.
Characteristics of NNA/IM-Resistant EDR
The EDRs induced both by bradykinin and by A23187 were maximal at concentrations of 10−7 mol/L in the presence of 30 μmol/L NNA and 10 μmol/L IM (Fig 5⇓). Thus, in the following experiments, the concentrations of bradykinin and A23187 were fixed at 10−7 mol/L.
No bradykinin-induced relaxation was observed when the endothelium was denuded (Fig 3⇑, top). EDR was sustained in the absence of NNA (Fig 3⇑, middle) but was relatively transient in the presence of NNA (Fig 3⇑, bottom).
NNA/IM-resistant EDR was completely abolished by 10 mmol/L TBA but was not inhibited by 10 mmol/L TEA, 0.1 mmol/L BaCl2, or 1 μmol/L glibenclamide. The ratios of the percentage relaxations were 0 (P<.01), 0.94±0.09 (P=NS), 1.07±0.15 (P=NS), and 0.88±0.10 (P=NS), respectively, versus the control value of 0.95±0.04 (n=7).
Inhibitory Effects of LPC on NNA/IM-Resistant EDR
NNA/IM-resistant EDR induced by 10−7 mol/L bradykinin was dose-dependently inhibited by LPC (Fig 6⇓). The EDR was almost completely abolished at a concentration of 20 μmol/L. The relaxation response evoked by 100 μmol/L nicorandil, a potassium channel opener, and 100 nmol/L sodium nitroprusside was preserved after LPC treatment (data not shown). The relaxation of arterial smooth muscle per se was not inhibited by LPC at these concentrations. The inhibitory effects were almost completely reversed by 0.1% albumin, an absorber of LPC (Fig 6⇓, top). At LPC concentrations of 15 to 20 μmol/L, the ratios of percentage relaxation were significantly improved. Hence, the inhibitory effects of LPC were reversible, and treatment with LPC at these concentrations for 10 minutes did not cause endothelial cell damage. Treatment with LPC for 30 minutes also showed the same results (data not shown).
The effects of LPC on NNA/IM-resistant EDR induced by the calcium ionophore A23187 were also examined. LPC dose-dependently inhibited the EDR similarly to bradykinin, and the inhibition was significant at concentrations >10 μmol/L (Fig 6⇑, bottom). Albumin treatment also reversed the inhibitory effects of LPC. The effects of LPC concentrations on the ratio of percent relaxation in the absence of NNA were the same as those in the presence of NNA (Fig 7⇓).
The experiment using 10−6 mol/L ADP as an additional EDHF-dependent dilator was performed. The IC50 value for ADP was 19.3±2.1 μmol/L (IC50 values for bradykinin and A23187 were 25.1±1.4 μmol/L and 20±1.3 μmol/L, respectively).
Effects of LPC on Bradykinin- or A23187-Induced Hyperpolarization
To directly confirm the inhibition of membrane hyperpolarization by LPC, we measured the membrane potential by the microelectrode. Fig 8⇓ (top) shows a representative presentation of bradykinin-induced hyperpolarization of the endothelium-intact porcine coronary artery. After bradykinin-induced hyperpolarization was confirmed in the presence of NNA and IM, LPC was administered. LPC inhibited the hyperpolarization. Table 1⇓ summarizes the results. LPC inhibited the membrane hyperpolarization. A23187 showed the same results as bradykinin. TBA (5 mmol/L) reduced the membrane potential. TEA (10 mmol/L), BaCl2 (0.1 mmol/L), glibenclamide (1 μmol/L), and NNA (30 μmol/L) had no effect (data not shown).
Effects of PMA and Staurosporine on NNA/IM-Resistant EDR
PMA at a concentration of 100 nmol/L partially inhibited the NNA/IM-resistant EDR induced by bradykinin, whereas 100 nmol/L staurosporine had no effect (Table 2⇓).
The ratio of percentage relaxation did not increase after the addition of 100 nmol/L PMA or staurosporine before 20 μmol/L LPC. The inhibitory effects of LPC were not reversed by PMA or staurosporine.
Previous reports have indicated that, in atherosclerotic lesions, the content of oxidized LDL and LPC is increased18 and the EDRs are impaired.19 Several studies have demonstrated the inhibition of EDRs ex vivo by oxidized LDL and LPC,3 4 5 6 7 8 9 10 11 but in these reports the NNA-sensitive (mediated by nitric oxide) EDR and the NNA-resistant EDR were not clearly distinguished. Experiments using rabbit aorta4 6 11 have shown that oxidized LDL and LPC (10 to 20 μmol/L) inhibit the EDR induced by acetylcholine or serotonin but not that evoked by A23187. Thus, the inhibition was considered to be due to interference with intramembrane regulatory pathways.4 Inoue et al8 used the cascade method, which consists of bovine aortic endothelial cells and endothelium-denuded rabbit aorta, and reported that the EDR evoked by bradykinin was abolished by LPC (20 μmol/L). Experiments with the porcine coronary artery7 9 10 have demonstrated that serotonin-, thrombin-, and UK14304 (an α2-adrenergic agonist)–induced EDRs were inhibited by oxidized LDL or LPC (10 to 20 μmol/L), while neither bradykinin- nor A23187-induced EDRs showed such an inhibition. Thus, the inhibitory effects of LPC on EDR seem to be dependent on agonists, vessels, and species. One explanation for this is differences in the signal transduction system,10 but the EDR may be mediated by several factors,7 the ratios of which differ in response to different agonists or in different vessels. In our preliminary study, no NNA/IM-resistant bradykinin-induced EDR was observed in the rabbit aorta (data not shown). In the rabbit aorta, the EDR seemed to be mediated mainly by nitric oxide. However, in the porcine coronary artery, EDR was observed to be mediated not only by nitric oxide but also by an NNA/IM-resistant relaxing factor (Fig 3⇑, bottom) that is considered to be identical to EDHF.16 17 It is important to distinguish between nitric oxide–mediated and EDHF-mediated EDRs when the effects of LPC on EDR are examined.
Previous studies have indicated that EDHF is a potassium channel opener.12 14 In the present study, the NNA/IM-resistant EDR was completely abolished by TBA but was not inhibited by TEA, BaCl2, or glibenclamide. EDHF has been demonstrated to be inhibited by TBA, a nonselective potassium channel blocker,16 but not by the selective blockers BaCl2 and glibenclamide.20 The effects of TEA on EDHF are at present controversial; EDHF induced by acetylcholine was inhibited by TEA,21 whereas the NNA/IM-resistant EDR induced by bradykinin was not.20 22 This might be a result of the antimuscarinic effect of TEA, which inhibits only acetylcholine receptor–mediated pathways.22
The effects of LPC on the NNA/IM-resistant EDR were dose dependent. At a concentration of 5 μmol/L, LPC did not significantly inhibit the EDRs induced by bradykinin or A23187 (Fig 7⇑). Although in previous reports, the EDRs induced by acetylcholine or serotonin in rabbit aorta were inhibited at these concentrations,4 6 11 these EDRs are thought to be mediated mainly by nitric oxide, and LPC appears to inhibit the nitric oxide– but not the EDHF-mediated EDRs. At concentrations of 15 to 20 μmol/L LPC, the NNA/IM-resistant EDR was significantly inhibited.
This inhibition was not due to the decreased reactivity of arterial smooth muscle cells, since the relaxation in response to nicorandil and sodium nitroprusside was not inhibited after LPC treatment.
We directly measured the membrane potentials of the porcine coronary artery smooth muscle. LPC was found to inhibit bradykinin- and A23187-induced membrane hyperpolarization. These findings are compatible with those obtained by the relaxation of the porcine coronary artery.
Because the critical micelle concentration of LPC in a Krebs-Henseleit solution at pH 7.4 and 37°C has been reported to be 40 to 50 μmol/L,23 this inhibitory effect might be due to cell lysis by the detergent actions of LPC micelles on cell membranes. In the present study, the inhibitory effects were reversed by albumin, a scavenger of LPC, and were not due to endothelial cell lysis (Fig 6⇑). The endothelial cell function may be reversibly altered by LPC.
LPC was reported to stimulate PKC at low concentrations (<20 μmol/L) and show an inhibitory effect at high concentrations (>30 μmol/L).24 Since pretreatment with the PKC inhibitor staurosporine attenuated LPC (5 μmol/L)–induced impairment of the EDR in response to thrombin, LPC was thought to inhibit EDR by a pathway involving PKC activation.9 In the present study, PMA partially inhibited the NNA/IM-resistant EDR in response to bradykinin, but staurosporine had no effect. The inhibitory effect of LPC was not attenuated by staurosporine or PMA (Table 2⇑). Thus, the NNA/IM-resistant EDR may be partially regulated by PKC, but the inhibitory mechanism of LPC seems to be independent of the PKC pathway.
Flavahan10 demonstrated that LPC (10 μmol/L) selectively inhibited the EDRs coupled to the pertussis toxin–sensitive Gi protein and did not affect that evoked by bradykinin, which stimulates phospholipase C via the pertussis toxin–insensitive Gq protein. On the other hand, Inoue et al8 reported that LPC (10 to 40 μmol/L) dose-dependently inhibited both phosphoinositide hydrolysis and the increases in intracellular Ca2+ concentration evoked by bradykinin and that the bradykinin-induced EDR was abolished by LPC (20 μmol/L). This discrepancy may be due to the differences in concentration of LPC. In the latter study, the bioassay system for EDR was the cascade method using bovine aortic endothelial cells and endothelium-denuded rabbit aorta. Because the NNA/IM-resistant EDR induced by bradykinin has been reported to be mediated by a nontransferable factor,25 the observed EDR might be mediated only by nitric oxide. Nevertheless, LPC may also inhibit the bradykinin receptor–Gq protein–phospholipase C pathway at high concentrations.
The NNA/IM-resistant relaxation cannot be induced by serotonin (in the presence of ketanserin) (Fig 2⇑) and UK 14,304 (an α2-adrenergic agonist) (data not shown). The relaxation by serotonin and UK 14,304 in the absence of NNA is mediated by Gi protein.8 Bradykinin- and ADP-induced relaxation in the presence or absence of NNA was not inhibited or enhanced by pertussis toxin (100 ng/mL) or sodium fluoride (3.5 mmol/L) (data not shown). Gi is not considered to be involved in the EDHF- mediated relaxation. Thus, the inhibition of EDHF-induced relaxation by LPC is not related to Gi protein.
In the present study, LPC inhibited EDRs induced by bradykinin and A23187 almost equally. Bradykinin-induced EDR is receptor mediated, whereas A23187 is receptor independent and induces EDR by directly increasing the transmembrane Ca2+ influx and consequently intracellular Ca2+ concentration. This suggests that LPC directly inhibits the production or release of EDHF, not the receptor-mediated signal transduction system.
A large amount of LPC associated with lipoproteins or albumin in the circulating plasma is considered not to be transferable to the endothelial membranes. LPC in oxidized LDL in the arterial intima may come into contact with the endothelial membranes26 and may inhibit nitric oxide– and EDHF-induced relaxation. With regard to the mechanism involved in the LPC modification of the dilator responses, the possibility must also be considered that LPC puts the arterial wall under oxidative stress. LPC has been implicated in mediating endothelial responses that are redox sensitive, such as vascular cell adhesion molecule-1 gene induction.27 LPC might also affect EDHF in a redox-sensitive manner. In the porcine and human coronary arteries,28 EDHF appears to be one of the major components regulating vascular tone.
Selected Abbreviations and Acronyms
|EDHF||=||endothelium-derived hyperpolarizing factor|
|PKC||=||protein kinase C|
|PMA||=||phorbol 12-myristate 13-acetate|
This study was supported by a research grant from the Ministry of Education, Science, and Culture, Japan. We express our appreciation to Dr Chuichi Kawai for his pertinent support and advice.
Reprint requests to Yoshiki Yui, MD, Third Division, Department of Internal Medicine, Faculty of Medicine, Kyoto University, 54 Shogoin-Kawaracho, Sakyo-Ku, Kyoto 606-01, Japan.
- Received April 18, 1995.
- Revision received July 17, 1995.
- Accepted August 3, 1995.
- Copyright © 1995 by American Heart Association
Parthasarathy S, Steinberg UP, Barnett J, Witztum JL, Steinberg D. Essential role of phospholipase A2 activity in endothelial cell-induced modification of low density lipoprotein. Proc Natl Acad Sci U S A. 1985;82:3000-3004.
Takahashi M, Yui Y, Yasumoto H, Aoyama T, Morishita H, Hattori H, Kawai C. Lipoproteins are inhibitors of endothelium-dependent relaxation of rabbit aorta. Am J Physiol. 1990;258:H1-H8.
Simon BC, Cunningham LD, Cohen RA. Oxidized low density lipoproteins cause contraction and inhibit endothelium-dependent relaxation in the pig coronary artery. J Clin Invest. 1990;86:75-79.
Tanner FC, Noll G, Boulanger CM, Luscher TF. Oxidized low density lipoproteins inhibit relaxations of porcine coronary arteries: role of scavenger receptor and endothelium-derived nitric oxide. Circulation. 1991;83:2012-2020.
Inoue N, Hirata K, Yamada M, Hamamori Y, Matsuda Y, Akita H, Yokoyama M. Lysophosphatidylcholine inhibits bradykinin-induced phosphoinositide hydrolysis and calcium transients in cultured bovine aortic endothelial cells. Circ Res. 1992;71:1410-1421.
Kugiyama K, Ohgushi M, Sugiyama S, Murohara T, Fukunaga K, Miyamoto E, Yasue H. Lysophosphatidylcholine inhibits surface receptor-mediated intracellular signals in endothelial cells by a pathway involving protein kinase C activation. Circ Res. 1992;71:1422-1428.
Flavahan NA. Lysophosphatidylcholine modifies G protein-dependent signaling in porcine endothelial cells. Am J Physiol. 1993;264:H722-H727.
Mangin EJ, Kugiyama K, Nguy JH, Kerns SA, Henry PD. Effects of lysolipids and oxidatively modified low density lipoprotein on endothelium-dependent relaxation of rabbit aorta. Circ Res. 1993;72:161-166.
Olesen SP, Davies PF, Clapham DE. Muscarinic-activated K+ current in bovine aortic endothelial cells. Circ Res. 1988;62:1059-1064.
Ylä-Herttuala S, Palinski W, Rosenfeld ME, Parthasarathy S, Carew TE, Butler S, Witztum JL, Steinberg D. Evidence for the presence of oxidatively modified low density lipoprotein in atherosclerotic lesions of rabbit and man. J Clin Invest. 1989;84:1086-1095.
Bossaller C, Habib BG, Yamamoto H, Williams C, Wells S, Henry PD. Impaired muscarinic endothelium-dependent relaxation and cyclic guanosine 5′-monophosphate formation in atherosclerotic human coronary artery and rabbit aorta. J Clin Invest. 1987;79:170-174.
Cowan CL, Cohen RA. Different mechanisms of relaxation of pig coronary artery to bradykinin and cromakalim are distinguished by potassium channel blockers. J Pharmacol Exp Ther. 1992;260:248-253.
Chen G, Yamamoto Y, Miwa K, Suzuki H. Hyperpolarization of arterial smooth muscle induced by endothelial humoral substances. Am J Physiol. 1991;260:H1888-H1892.
Bergmann SR, Bruce Ferguson TJ, Sobel BE. Effects of amphiphiles on erythrocytes, coronary arteries, and perfused hearts. Am J Physiol. 1981;240:H229-H237.
Oishi K, Raynor RL, Charp PA, Kuo JF. Regulation of protein kinase C by lysophospholipids: potential role in signal transduction. J Biol Chem. 1988;263:6865-6871.
Kauser K, Rubanyi GM. Bradykinin-induced, Nω-nitro-l-arginine-insensitive endothelium-dependent relaxation of porcine coronary arteries is not mediated by bioassayable relaxing substances. J Cardiovasc Pharmacol. 1992;20(suppl 12):S101-S104.
Kume N, Cybulsky MI, Gimbrone MA Jr. Lysophosphatidylcholine, a component of atherogenic lipoproteins, induces mononuclear leukocyte adhesion molecules in cultured human and rabbit arterial endothelial cells. J Clin Invest. 1992;90:1138-1144.
Khan BV, Parthasarathy SS, Alexander RW, Medford RM. Modified low density lipoprotein and its constituents augment cytokine-activated vascular cell adhesion molecule-1 gene expression in human vascular endothelial cells. J Clin Invest. 1995;95:1262-1270.
Nakashima M, Mombouli J-V, Taylor AA, Vanhoutte PM. Endothelium-dependent hyperpolarization caused by bradykinin in human coronary arteries. J Clin Invest. 1993;92:2867-2871.