Endothelin Peptide and Receptors in Human Atherosclerotic Coronary Artery and Aorta

Impairment of endothelial function leading to an imbalance in the production of vasoconstrictor mitogenic factors such as ET and the antimitogenic vasodilator NO is believed to be an initiating factor in atherosclerosis.¹ Endothelial cells synthesize the vasoactive peptide ET, which when released appears to act on neighboring vascular receptors in an autocrine or paracrine manner.² ET_B receptors located on endothelial cells may mediate vasodilatation via the release of NO.³ In contrast, stimulation of contractile ET_A or ET_B receptors situated on VSMCs results in potent vasoconstriction.⁴⁵ ET_A receptors predominate in the medial smooth muscle layer of human coronary artery and aorta, both examples of arteries prone to atherosclerosis.⁶⁷⁸ Only small and inconsistent constrictions to ET_B receptor agonists have been observed in these vessels in vitro.⁹

As the most potent vasoconstrictor described, ET may be a contributing factor in the vasospasm associated with the atherosclerotic arterial wall.¹⁰ Furthermore, at subthreshold concentrations, lower than required for ET-induced contraction, it has been reported to potentiate the response to other vasoconstrictors such as serotonin.¹¹¹²

A key event in the development of an atherosclerotic plaque is the formation of a neointima due to proliferation of VSMCs. In common with other vasoconstrictors such as angiotensin II,¹³ ET has the capacity to promote the proliferation of vascular smooth muscle. In the presence of mitogens such as PDGF, ET is comitogenic for cultured rat VSMCs.¹⁴ A potentiating action of ET on neointimal formation after rat carotid artery balloon angioplasty has also been demonstrated.¹⁵

The release of ET from endothelial cells is known to be stimulated by a number of factors, including TGFβ, thrombin, and interleukin-1.¹⁶ Invading macrophages may be another potential source of ET in diseased arteries. In culture, these cells secrete ET,¹⁷ and release is enhanced in the presence of oxidized low-density lipoprotein,¹⁸ as it is in endothelial cells.¹⁹ Additionally, VSMCs in culture release the peptide when stimulated by factors such as angiotensin II, TGFβ, and PDGF.²⁰

A significant increase in the concentration of circulating ET in patients with advanced atherosclerosis has been reported previously.²¹ Likewise, risk factors for atherosclerosis, such as diabetes, hypercholesterolemia, and cigarette smoking, may be associated with elevated plasma ET.²²²³²⁴ It is unknown whether this is due to an increase in the production and release of ET or results solely from endothelial cell damage. Semiquantitative studies suggest that the local ET concentration within atherosclerotic tissue is actually increased.²⁵ Interestingly, more ET-1 mRNA has been detected in arteries containing atherosclerotic lesions obtained at carotid endarterectomy than in normal human aorta.²⁶

Given the possible importance of ET in the pathogenesis of atherosclerosis, the purpose of the present study was to qualitatively and quantitatively examine ET peptide and its receptors in human diseased arteries. The aim was to determine whether atherosclerosis is associated with an alteration in the expression of ET or the receptors that mediate its actions. Quantitative techniques including RIA and quantitative autoradiography were used to determine whether upregulation or downregulation of these proteins occurs in atherosclerotic coronary artery and aorta. Visualization by immunohistochemistry and autoradiography also allowed the distinction between VSMCs found in the neointima and those present in the vasoactive part of the vessel, the media. A spectrum of lesions at different stages of development was studied, ranging from preatherosclerotic to fully developed atherosclerotic plaques.

Materials and Methods

Materials

Amprep C2 minicolumns, Amerlex-M reagent, Hyperfilm β-max, ¹²⁵I microscales, and all radioligands ([¹²⁵I]ET-1, [¹²⁵I]PD151242, and [¹²⁵I]BQ3020 [2000 Ci/mmol]) were from Amersham International. Unlabeled ET-1 was from Novabiochem, BQ3020 was synthesized by solid-phase t-boc chemistry, and PD151242 was synthesized by Dr A.M. Doherty (Parke-Davis Pharmaceutical Research Division). Kodak NTB2 nuclear emulsion, D19 developer, Dektol developer, and fixer were from Eastman-Kodak. Bio-Rad DC reagents were obtained from Bio-Rad Laboratories. The following antibodies were purchased from Dako Corp: swine anti-rabbit immunoglobulins, rabbit PAP, rabbit anti-human von Willebrand factor, and monoclonal mouse anti-human macrophage CD68 PG-M1. Monoclonal anti–α smooth muscle actin was from Sigma Immuno Chemicals. The Vectastain ABC kit was from Vector Laboratories.

All other reagents were from Fisons and Sigma.

Tissue Samples

The present study was conducted with local ethical approval. Samples of human cardiac arteries were obtained from the recipient hearts of patients undergoing cardiac transplantation for coronary artery disease, cardiomyopathy, or congenital heart disease, with histologically normal coronary arteries, at Papworth Hospital, UK. Coronary artery samples were taken from 23 different individuals for which the average age was 46±16 years (mean±SEM). Drug therapy included Ca²⁺ antagonists, vasodilators, angiotensin-converting enzyme inhibitors, diuretics, digoxin, and anticoagulants. Aortic samples were also obtained from patients undergoing pulmonary transplantation or from tissue not subsequently used for transplantation.

The adventitia was removed from arteries used for RIA or saturation binding studies. All arteries were immediately snap-frozen in liquid nitrogen and stored at −70°C. Cryostat sections cut for immunohistochemistry, saturation binding assays, and autoradiography studies were thaw-mounted on slides coated with either gelatin or 3-aminopropyltriethoxysilane.

RIA for Mature ET and Big ET-1

RIA was used to compare the amount of immunoreactive ET and the precursor big ET-1 in diseased and nondiseased aorta. These large arteries were selected for RIA experiments because of the greater mass of tissue available for peptide extraction. Fifteen samples of aorta from 14 patients were classified as to whether they were histologically normal, possessed fatty streaks, or raised atherosclerotic plaques. With the adventitia removed, each sample was ground to a powder in a freezer-mill (Glen Creston Ltd), added to 10 vol of 0.5 mol/L acetic acid/0.01% Triton X-100, and heated in a boiling water bath for 15 minutes. This was followed by centrifugation at 48 000g for 20 minutes at 4°C, and resulting supernatants were applied to activated Amprep C2 minicolumns. The columns were washed with 0.1% trifluoroacetic acid before the elution of bound material with 80% methanol/0.1% trifluoroacetic acid. Resulting extracts were evaporated to dryness before reconstitution in RIA buffer (50 mmol/L sodium phosphate, 0.25% bovine serum albumin, 0.01% Tween-20, and 0.05% sodium azide, pH 7.4).

Antisera raised against the common C-terminus of the mature ETs (ET-1[15-21]) at a dilution of 1:10 000 were used to assay the reconstituted extracts for mature ET. An assay for big ET-1 used antisera against the C-terminus of big ET-1 (big ET-1[31-38]) diluted to 1:20 000 as previously described.²⁷

Tissue extracts were incubated with diluted antisera overnight at 4°C before incubating again with [¹²⁵I]ET-1 or [¹²⁵I]big ET-1 (≈10 000 cpm per tube) under the same conditions. Amerlex-M reagent was used to separate out bound radioligand, which was then counted. Standard curves constructed for unlabeled ET-1 or big ET-1 over the range 0.5 to 1000 fmol per tube (5×10⁻¹² to 10⁻⁸ mol/L) were used to determine the immunoreactivity in each tissue extract.

The sensitivity of detection of this assay has been calculated to be <1.25 fmol per tube, with ED₅₀ values between 20 and 25 fmol per tube. Intra-assay and interassay coefficients of variation for the ET RIA were, respectively, as follows: at 6 fmol per tube, 7.8% and 10.8%; at 30 fmol per tube, 9.7% and 12.9%; and at 350 fmol per tube, 13.3% and 19%. For the big ET-1 RIA, the respective values were as follows: at 6 fmol per tube, 8.5% and 10.7%; at 30 fmol per tube, 4.2% and 6.7%; and at 350 fmol per tube, 15.1% and 14.3% (n=25 and 6, respectively). The ET RIA cross-reacts equally with ET-1, ET-2, and ET-3, with <0.02% cross-reactivity with big ET peptides. The big ET-1 RIA only cross-reacts with big ET-1 and the C-terminal fragment of big ET-1.²⁷

There was no evidence of nonnormality using the Shapiro Francia test; data were analyzed by ANOVA, followed by Student's t test (P<.01).

Immunohistochemistry for Mature ETs

Immunohistochemistry was carried out as previously described²⁸ to visualize the distribution of the mature ET peptides in sections of healthy and diseased vessels. The antisera were raised against the common C-terminus of mature ETs and have been shown by enzyme-linked immunosorbent assays to cross-react with ET-1, ET-2, and ET-3 but have no detectable cross-reactivity with big ET precursors.²⁸ Serial 30-μm slide-mounted vessel sections were fixed in 4% paraformaldehyde (4°C, 30 minutes). After they were washed in 0.1 mol/L PBS and permeabilized with 0.1% Triton X-100, the sections were incubated with 10% swine serum (22°C, 30 minutes) as a blocking step. Excess serum was then removed and replaced with the primary rabbit anti-ET antibodies, initially tested at dilutions of 1/100 to 1/1000 and subsequently used at 1/300. In all cases, an adjacent section was treated with preimmune rabbit serum at the same dilution as a negative control. Sections were incubated with primary antibody at 4°C overnight.

A 10-minute wash period preceded the addition of secondary swine anti-rabbit antibody (1/200, 22°C, 60 minutes). The sections were again washed before incubation with rabbit PAP (1/400, 22°C, 60 minutes). A solution of diaminobenzidine was applied to the washed sections; the appearance of the brown reaction product indicated regions of positive immunostaining. Deionized water was used to stop the reaction, and the sections were dehydrated by passage through an alcohol series and cleared in xylene before mounting.

Positive specificity controls included the demonstration that ET antibodies stained the cytoplasm of endothelial cells from human umbilical vein and artery. Specificity was also tested by preabsorbing ET antibodies with ET-1, ET-2, and ET-3.²⁸ In addition to substituting immune serum with preimmune serum, negative controls included the absence of staining when the antiserum was absorbed with its antigen before incubation with tissue sections. Staining was unaffected when antibodies were absorbed with a range of peptides unrelated to ET, at a concentration of 1 μmol/L, including calcitonin gene-related peptide, atrial natriuretic peptide, bradykinin, neuropeptide Y, human vasoactive intestinal peptide, cholestokinin-4, pentagastrin, somatostatin, substance P, eledoisin, dynorphin A (fragment 1-6), and [Met]enkephalin.

[¹²⁵I]ET-1 Saturation Binding Studies

The binding of [¹²⁵I]ET-1 in sections of diseased coronary arteries from three patients with ischemic heart disease was compared with that of three histologically normal coronary arteries. Endothelial cells were removed by gentle rubbing, and consecutive 10-μm-thick cryostat sections were cut longitudinally through the media of the arteries. Slide-mounted sections were incubated with HEPES buffer (pH 7.4) containing increasing concentrations of [¹²⁵I]ET-1 for 2 hours at 22°C as previously described.²⁹ Nonspecific binding was determined at each concentration by inclusion of unlabeled ET-1 (1 μmol/L). After washing, bound radioactivity was determined by gamma counting. Protein concentrations were calculated using the Bio-Rad DC 96-well microtiter method. Binding data were analyzed using the nonlinear iterative curve-fitting program LIGAND.³⁰

Quantitative Autoradiography

Sections were cut to a 10-μm thickness at different orientations through a range of lesions varying from those with predominantly intracellular lipid to fully established atherosclerotic plaques with necrotic cores. ET_A receptors were labeled with the ET_A-selective radioligand [¹²⁵I]PD151242³¹ ; ET_B receptors were labeled with the ET_B-selective radioligand [¹²⁵I]BQ3020.³² Sections were incubated in the same conditions as the binding assays with [¹²⁵I]ET-1 (100 pmol/L), the ET_A-selective radioligand [¹²⁵I]PD151242 (100 pmol/L), or the ET_B-selective radioligand [¹²⁵I]BQ3020 (300 pmol/L). The concentrations used were calculated to label ≈30% of receptors according to the dissociation constant of each radioligand.⁷ Nonspecific binding was defined by incubation of adjacent sections with the corresponding unlabeled ligand (1 μmol/L). After they were washed and dried, the sections were apposed, along with [¹²⁵I]microscale standards, to Hyperfilm βmax radiation-sensitive film for 7 days at 22°C.

After development in D19 developer, the film images were quantified using computer-assisted image analysis (Quantimet 920, Cambridge Instruments) as previously described.³³

Microautoradiography

For higher resolution autoradiography, labeled sections were immersed in Kodak NTB2 nuclear emulsion in complete darkness and allowed to dry. The sections were exposed at 4°C for 7 days before development with Kodak Dektol developer and fixing.

Histology

In all cases, adjacent sections were stained with hematoxylin and eosin to enable histological examination. For further clarification, sections were also stained for the presence of endothelial cells, human smooth muscle α-actin, or macrophages.

Endothelial cell staining was achieved using polyclonal rabbit anti–human von Willebrand factor antibodies (1/1000 dilution) on acetone-fixed sections. The PAP detection system was used as described above for the immunohistochemistry for ET.

A monoclonal mouse anti–human smooth muscle α-actin antibody and a monoclonal mouse anti–human macrophage antibody (CD68 PG-M1) were used at dilutions of 1/1000 and 1/50, respectively. Staining with these antibodies was achieved by an avidin biotin system using the Vectastain ABC kit with diaminobenzidine as a substrate.

Results

RIA

Samples of aorta containing raised atherosclerotic plaques or fatty streaks were obtained from five different individuals each (total, 10 samples). Histologically normal aorta was taken from another four individuals and from a different unaffected region of one of the arteries providing raised plaques.

Levels of mature ET and big ET-1 detectable in the aorta samples are shown in Fig 1. There was no significant difference between the amounts detectable in nondiseased aorta and tissue with fatty streaks. However, there was a highly significant increase in both big ET-1 and mature ET in aorta containing fully developed raised atherosclerotic plaques (ANOVA followed by Student's t test, P<.01). In all three categories of aorta, the quantity of immunoreactive big ET-1 was similar to that of the mature peptide.

Immunohistochemistry

ET-like immunoreactivity was demonstrated in the luminal endothelial cells of histologically normal vessels and in perivascular microvessels, with little seen in the media. In arteries with atherosclerotic disease, ET staining was seen in the endothelial cells lining the remains of the lumen. Positive staining was also observed in the endothelial cells in newly formed microvessels, including neovascularization extending through the media and recanalization at sites of previous thrombosis (Fig 2). Although weak diffuse staining occurred across sections, no strong positive signal was obtainable in the cytoplasm of neointimal smooth muscle cells in any of the arteries studied. Discrete patches of immunoreactivity, visualized in the neointima of some vessels, corresponded to regions of macrophage staining.

Saturation Binding Assays

Saturation binding assays demonstrated that [¹²⁵I]ET-1 binds with high affinity and that Hill coefficients are close to unity in both diseased and histologically normal coronary arteries (Table). The dissociation constants were in the same subnanomolar range, indicating that the affinity of the arterial receptors is unchanged in the atherosclerotic artery. The B_max values calculated were also comparable in ischemic and nonischemic arteries.

Autoradiography

Dense binding of [¹²⁵I]ET-1 was observed in the vascular smooth muscle of the media of all vessels studied. In complete contrast, the smooth muscle of the neointima was virtually devoid of ET receptors (Fig 3). This was apparent in sections cut in a number of different orientations through vessels. Receptors were lacking in the thickened intima of macroscopically normal arteries, as well as in fatty streaks and fully developed atherosclerotic plaques.

In agreement with the [¹²⁵I]ET-1 images, ET_A receptors, although present on the thinned media underlying atherosclerotic plaques, were sparse or undetectable in the neointima. Specific [¹²⁵I]PD151242 binding demonstrated the presence of ET_A receptors in the contractile medial smooth muscle, with comparatively little ET_B binding in this region. ET_B receptors were completely absent in neointimal smooth muscle. In regions of organized thrombus, microautoradiography revealed ET_A receptors on the smooth muscle of recanalization vessels (Fig 4).

Quantification of autoradiograms was used to determine whether the number of receptors or the ratio of ET_A to ET_B in the arterial media correlates with atherosclerotic disease. Similar levels of binding were observed, and there was no statistically significant difference in the proportion of ET_B receptors detectable in human coronary arteries with and without atherosclerotic disease (Fig 5).

Intense [¹²⁵I]BQ3020 binding was visible on perivascular ET_B receptors localized to adventitial lymphatics, nerves, and periadventitial lymphoids. Discrete clusters of ET_B receptors were apparent in longitudinal and transverse sections of the media and neointima of diseased vessels but absent in nondiseased arteries. Endothelial cell staining of adjacent sections produced a superimposable pattern of positivity suggesting that ET_B receptors are present on neovascularization.

In some atherosclerotic lesions studied, distinct areas of ET_B receptors that were not matched by human smooth muscle α-actin or von Willebrand factor staining were demonstrated. These cells with ET_B receptors were subsequently identified as macrophages by microautoradiography with hematoxylin and eosin staining and macrophage staining (Fig 6).

Discussion

The results of the present study demonstrate for the first time in a quantitative manner that there is a significant increase in the amount of ET present in human arteries with fully developed atherosclerosis. The ratio of ET to big ET-1 was similar in diseased and normal aorta, implying that ET upregulation in diseased tissue occurs at the level of gene expression rather than at the ET-converting enzyme. The antisera used to measure the amount of mature ET in the RIA²⁷ and to localize the peptide by immunohistochemistry²⁸ cross-react with ET-1, ET-2, and ET-3, and the precise isoform expressed in human arteries has not been determined. Visualization of the actual distribution of peptide by immunohistochemistry suggests that a significant proportion of the increase in mature ET is due to heightened endothelial production. Evidence is emerging that ET-1 and its precursor are the predominant isoforms in human endothelial cells. For example, ET-1 and big ET-1 were detected in human endothelial cells from the endocardium and umbilical vein after separation by high-performance liquid chromatography and RIA of the resulting fractions. However, ET-2 and ET-3 could not be detected.²⁷

Intense staining for ET was seen in the endothelium of newly formed microvessels associated with atherosclerosis, including neovascularization and recanalization of organized thrombus. It has been suggested that the rich microvasculature of neovascularization may promote localized vasospasm and intramural hemorrhage by providing increased tissue concentrations of vasoconstrictors.³⁴

Increased ET levels could also be traced to the presence of invading macrophages in the intimal layer of some arteries. ET immunoreactivity has also been localized to invading inflammatory cells in coronary artery lesions of patients with crescendo angina.²⁵ Activated leukocytes are known to release a stable vasoconstrictor factor.³⁵ Furthermore, circulating lymphocytes and monocytes from patients with chronic heart failure have been shown to spontaneously release ET; this release was not observed in normal subjects.³⁶

In contrast, little immunoreactive ET was detectable in VSMCs, both in the media and neointima, regions where the cells would be expected to be of the contractile or proliferative phenotypes, respectively. A lack of ET in the media of coronary artery has been previously reported.²⁸³⁷ Conversely, other studies have described ET immunoreactivity localized to vascular smooth muscle of diseased human arteries.²¹²⁵ In addition, cultured VSMCs have the ability to synthesize and secrete ET.²⁰ However, the results of the present study indicate that VSMC-derived ET does not contribute greatly to the increase in immunoreactivity observed here in atherosclerotic aorta.

No significant difference in the amount of ET detectable in normal aorta and in tissue with fatty streaks implies that the concentration of ET is unchanged within the fatty streak. However, if increases are focal in nature, local changes in concentration may have remained undetected when the entire piece of artery was assayed. Therefore, a role for ET in the development of the fatty streak cannot be eliminated on the basis of the present study.

Radiolabeled ET-1 bound with similar affinity to receptors in the media of diseased and nondiseased coronary artery. The difference in B_max values in these tissues was small, ruling out a large upregulation or downregulation of ET receptors associated with ischemic heart disease. Quantification of autoradiograms also indicated similar densities of receptors in arteries with and without atherosclerosis. Functional studies have indicated that healthy arteries are more sensitive than diseased arteries to low concentrations of ET-1, although the maximal contractile effect was similar.¹²

Contractile smooth muscle, in the arterial media and in the walls of new vessels in regions of recanalization, contained predominantly ET_A receptors. This is consistent with a number of functional studies with human arteries.⁶⁹ An example of species differences is provided by porcine coronary artery, where a significant non-ET_A contractile component has been demonstrated.⁴

In some animal models, an upregulation of ET_B receptors has been identified, which is related to high cholesterol diet³⁸ and vascular disease.³⁹ Additionally, a phenotypic change toward ET_B receptors has been reported in late-passage rat cultured VSMCs.⁴⁰ To examine whether such a shift occurs in human vessels, quantitative autoradiography was used to compare the proportions of ET_A and ET_B receptors in the media of normal and diseased coronary arteries. The results demonstrate that there is no significant change in the proportion of ET_B receptors present on medial smooth muscle of the human artery associated with atherosclerotic disease.

The absence of ET receptors in the intimal layer of coronary arteries has been previously reported.⁸⁴¹ In the present study, this has been observed in the muscular elastic layer of normal arteries and in the neointima of atherosclerotic lesions at all stages of development. These ranged from the early (including areas of intimal thickening with predominantly intracellular lipid) to the advanced (with necrotic lipid core) atherosclerotic plaque. Such findings are consistent with the lack of expression of ET_A receptor mRNA detectable in the thickened arterial intima of hypertensive patients.⁴² It remains to be seen whether ET receptors are actually downregulated in the proliferative VSMC phenotype or whether their absence reflects an undifferentiated state of the cells in the intima. In human cultured VSMCs, the mitogenic effect of ET-1, which correlated with ET receptor density, decreased with increasing population doubling of cells.⁴³ Interestingly, this effect on receptor number and responsiveness did not correlate with the rate of ET secretion by the cells. Cultured rabbit VSMCs have also been demonstrated to downregulate ET receptors when switching to the proliferative phenotype.⁴⁴ In the study by Kanse et al,⁴³ the ET_A antagonist BQ123 completely inhibited [¹²⁵I]ET-1 binding, and ET-1 induced mitogenesis of cultured human VSMCs. The order of potency of the ET isopeptides (ET-1=ET-2>ET-3) also indicated that the mitogenic effect of ET-1 is mediated by ET_A receptors in these cells. Therefore, it is possible that ET_A receptors are involved in an initial proliferative response in human VSMCs, but the progression of cells through successive divisions appears to be accompanied by downregulation of the receptors. Perhaps this downregulation accounts for the deficiency of ET_A receptors in the intimal layer of human arteries.

Interestingly, ET_B receptors were detected on macrophages within atherosclerotic plaques and on aggregates of lymphocytes. Previously, ET_B receptors also have been identified on mouse peritoneal macrophages.⁴⁵ ET has chemoattractant activity for macrophages and stimulates the release of factors such as interleukin-6.⁴⁶⁴⁷ Furthermore, monocytic cells have been reported to increase the vascular contractile potency of ET-1, possibly via the stimulation of monocyte ET receptors.⁴⁸ Given the importance of these cells in the pathogenesis of atherosclerosis, the significance of ET in their function is intriguing.

ET_B receptors were also localized by microautoradiography to endothelial cells of neovascularization extending through the media of diseased arteries. The presence of ET_B receptors on these new microvessels may explain the sites reported to be resistant to ET_A receptor antagonists in atherosclerotic arteries.⁴⁹

In previous binding studies with sections of the media of histologically normal aorta, ET_B receptors were undetectable.⁷ However, in other studies with aortic media from patients with ischemic heart disease, a small population of ≈10% ET_B receptors was detected.⁵⁰ The results of the present study suggest that this small difference is due to endothelial cells of neovascularization and possibly the presence of inflammatory cells in diseased arteries. The endogenous concentration of the potent vasoconstrictor ET is increased in human atherosclerotic arteries. When released from damaged endothelium, it may contribute to the development of atherosclerosis by promoting proliferation of VSMCs. Additionally, by interacting with the ET_A receptors, which predominate on human vascular smooth muscle, it could contribute to vasospasm and the events accompanying plaque rupture. Pharmacological intervention in the form of ET receptor antagonists is currently being considered for the treatment of vascular disease.⁵¹ A better understanding of the implications of these therapeutic strategies is likely to result from identification of cell types expressing ET and characterization of ET receptors within atherosclerotic plaques.

Selected Abbreviations and Acronyms

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Footnotes