Author + information
- Received January 20, 2000
- Revision received June 1, 2000
- Accepted July 14, 2000
- Published online December 1, 2000.
- Nitin K Ghaisas, MD, MRCPI∗,
- J.Brendan Foley, MD, MRCP∗,* (, )
- D.Sean O’Briain, MRCPath†,
- Peter Crean, FRCPI∗,
- Dermot Kelleher, MD, MRCPI‡ and
- Michael Walsh, MD, FRCPI, FACC∗
- ↵*Reprint requests and correspondence: Dr. Brendan Foley, Department of Cardiology, CREST Directorate, St. James’s Hospital, James’s Street, Dublin 8, Republic of Ireland
We studied the expression of intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule 1 (VCAM-1) and endothelial selectin (E-selectin) on aortic valve endothelium in patients undergoing valve replacement. We also assessed the relation between serum levels and endothelial expression and also the changes in serum levels following surgery.
Nonrheumatic aortic valve disease is believed to be a degenerative condition. Increased tissue and soluble adhesion molecule levels are described in inflammatory conditions.
Aortic valves from 22 surgical (16 bicuspid, 6 tricuspid) and 6 autopsy (4 normal, 2 thickened) cases were studied by immunohistochemistry. Soluble adhesion molecules were measured in peripheral blood preoperatively, and at 6 and 18 months postoperatively, and compared with controls.
The majority of the surgically removed tricuspid and bicuspid valves expressed adhesion molecules (E-selectin, 75% and 100%; ICAM-1, 75% and 80%; VCAM-1, 69% and 60%, respectively). The normal postmortem valves did not express these, while the diseased ones did. Endothelial expression of E-selectin correlated strongly with serum levels (r = 0.695, p = 0.004). Soluble E-selectin levels were significantly higher at baseline compared with controls (p = 0.017) and fell significantly at 18 months postoperatively (p = 0.005).
Adhesion molecule expression on diseased valves supports an inflammatory component in “degenerative” aortic valve disease. The diseased valves may be the main source of elevated soluble E-selectin in this condition as blood levels correlate with endothelial expression and blood levels fall at 18 months postoperatively.
Nonrheumatic stenosis of bicuspid and tricuspid aortic valves is the most common form of valvular heart disease in adults in Western societies (1). It is generally believed to be a degenerative condition resulting from mechanical stress (2,3). We have previously reported elevated levels of soluble cellular adhesion molecules (CAMs) in nonrheumatic aortic stenosis (AS) of tricuspid valves, suggesting an inflammatory response in this condition (4).
Adhesion between endothelial cells and circulating leukocytes is a key initial event in targeting leukocytes to sites of inflammation. Cellular adhesion molecules mediate the adhesion of cells with other cells or components of the extracellular matrix. They are expressed on vascular endothelium and on immune and inflammatory cells and are critical in targeting circulating leukocytes to sites of inflammation, transmigration of leukocytes across vascular endothelium and in immune effector functions. The adhesion molecules intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule 1 (VCAM-1) and endothelial selectin (E-selectin) are expressed on vascular endothelium and serve as ligands for counter-receptors on circulating inflammatory cells. The extracellular or soluble portions of CAMs are detectable at low levels in serum of healthy people, and raised levels have been described in various inflammatory conditions (5).
This study examines the expression of ICAM-1, VCAM-1 and E-selectin on aortic valve (AV) endothelium in patients with “degenerative” AV disease, assesses the relation between soluble CAM levels and their endothelial expression and also the effect of valve replacement surgery on soluble CAM levels.
Twenty-two patients undergoing AV replacement for critical AV disease were included. Exclusion criteria were evidence of rheumatic disease, endocarditis and other coexisting valvular disease. All patients had a detailed history and general medical examination. Echocardiography and cardiac catheterization data were reviewed. The mean age of the patients was 65 ± 12.43 years; there were 15 men and 6 women. There were 16 tricuspid and 6 bileaflet valves. Seventeen patients had AS and five had AS and AR. Four patients (18%) had hypertension and five (22%) were current smokers. Valves were examined grossly immediately after surgical removal, and segments of the valve were snap frozen in liquid nitrogen and stored at −80°C for immunohistochemistry.
Sixteen of these 22 patients had blood taken for soluble CAM measurements preoperatively and at 6 and 18 months postoperatively. The remaining six patients were excluded, five because of angiographically demonstrated coronary artery disease and one because of history of treated polymyalgia rheumatica. Sixteen healthy volunteers age matched to within five years and gender matched, with normal physical examination, electrocardiogram and echocardiogram were included as controls. Blood samples were centrifuged after clotting at 3,500 rpm for 10 min, and serum was separated and stored at −70°C for analysis.
For comparison purposes, six AVs from autopsies (mean age, 63.7 ± 9.1 years) were studied (four normal, two thickened but not stenosed). These were collected within 48 h of death, previous studies having shown good preservation of cell antigens up to three days (6).
Soluble ICAM-1, VCAM-1 and E-selectin were measured by enzyme-linked immunosorbent assay (ELISA), using commercially available kits (R & D Systems Ltd; Abingdon, United Kingdom). The assay uses a sandwich ELISA that involves the simultaneous reaction of the antigen in the sample to two antibodies directed against different epitopes on the antigen molecule. All samples were analyzed in duplicate, and the mean of two readings was used to calculate the results. The correlation between the two sets of readings was excellent (r = 0.985).
Cryostat sections 10 μm thick were cut onto silane-coated slides. The primary antibodies used were monoclonal mouse antibodies against human ICAM-1 (Novocastra 1:200), human VCAM-1 (Novocastra, 1:50) and human E-selectin (Novocastra, 1:40). Endothelium was confirmed by antibodies to human CD34 (Dako, 1:20)/CD31 (Dako, 1:40). The secondary antibody used was rabbit anti-mouse biotin-labeled antibody (Dako, 1:300). The avidin-biotin complex kit used was Vectastain Elite ABC (Vector Laboratories).
The slides were stained immunohistochemically by the indirect avidin-biotin complex immunoperoxidase technique. Sections were fixed in acetone-methanol for 10 min, the slides were blocked with 3% hydrogen peroxide, washed with trishydroxymethylaminomethane (TRIS) saline, incubated with the primary antibody at 4°C for 24 h and then washed again with TRIS saline. The biotin-labeled secondary antibody was applied for 30 min at room temperature, followed by the avidin-biotin conjugate for 30 min. Diaminobenzidine tetrahydrochloride was used as the chromagen, and the sections were counterstained with hematoxylin. Reactive tonsils were used as positive controls, and negative controls were by substitution of primary antibody with TRIS buffer to abolish staining. Staining was graded semiquantitatively from 0 to 4+ on the endothelium. Immunohistochemical stains were assessed independently by an experienced observer who was blinded to patient details. Staining of intensity 2+ or higher was considered positive.
Quantitative data for soluble CAM levels are presented as median and range. The Fisher exact test was used to compare differences between independent proportions. In investigating the associations of the intensity of staining for the three adhesion molecules, pairwise nonparametric correlation (Spearman’s) was employed. A coefficient of linear correlation was calculated to measure the strength of association between quantitative variables. The nonparametric paired signed rank test was used to compare the levels of soluble adhesion molecules for controls, preoperatively and postoperatively with a Bonferroni-type correction for multiple comparisons (p < 0.025 considered significant).
Endothelial adhesion molecule expression
Most of the surgically removed valves stained positive for the three adhesion molecules. One of the 22 valves did not have adequate endothelial preservation. The results of staining for the remaining 21 valves are summarized in Table 1. Examples of staining of the valves are shown in Figures 1 to 3. ⇓⇓⇓ In the surgical cases, there was no significant difference in the proportions of valves staining positive for any of the three molecules between the bicuspid and tricuspid groups (Fisher exact test—ICAM-1, p = 1.0; VCAM-1, p = 1.0; E-selectin: p = 0.532).
There was good pairwise association between endothelial expression of all three adhesion molecules (Spearman’s correlation coefficient)—ICAM-1 and VCAM-1, r = 0.68, p = 0.001; ICAM-1 and E-selectin, r = 0.69, p = 0.001; E-selectin and VCAM-1, r = 0.64, p = 0.002. The four normal autopsy valves were negative for all three adhesion molecules, while the two thickened valves were positive for the three molecules.
Correlation between endothelial and soluble adhesion molecule levels
There was a strong association between the intensity of endothelial staining and soluble molecule levels for E-selectin but not for ICAM-1 or VCAM-1 (Table 2).
Soluble adhesion molecule levels
Soluble E-selectin levels were higher at baseline in patients compared with controls, were unchanged at six months postoperatively, but were significantly lower at 18 months postoperatively (Fig. 4). Soluble ICAM-1 did not differ significantly between patients and controls and showed a rise at six months and 18 months postoperatively. Soluble VCAM-1 did not differ significantly between patients preoperatively and controls and did not change significantly at 6 months and 18 months postoperatively.
This study demonstrates for the first time, expression of the endothelial adhesion molecules ICAM-1, VCAM-1 and E-selectin in nonrheumatic degenerative AV disease. Soluble E-selectin levels in peripheral blood correlated well with valvular endothelial expression. Soluble E-selectin levels were higher in patients compared with controls and fell significantly at 18 months postoperatively.
Mechanical stress has been proposed as a possible mechanism for the development of degenerative AV disease (2,3). However, all AVs are subject to long-term mechanical stress, and valvular calcification of tricuspid and even bicuspid AVs is not an inevitable consequence of aging (7). Thus, even if mechanical stress is an initiating factor, other factors may be important in initiation and subsequent disease progression.
Changes in soluble adhesion molecules levels after valve replacement
Elevated levels of the soluble ICAM-1, VCAM-1 and E-selectin have been reported in a variety of inflammatory disorders (5). Elevated soluble E-selectin particularly has been thought to reflect systemic inflammatory conditions. In our study looking at levels serially, E-selectin levels were elevated at baseline compared with controls, remained high at six months postoperatively and fell significantly 18 months after surgery. This may suggest that the postoperative inflammatory response does not settle down until at least six months after surgery. The subsequent fall in levels suggests that the diseased valve itself may have been either the main source of elevated levels or the source of cytokines that may activate endothelium and thereby result in elevated levels. We did not find any previous reports on long-term effects of either valve replacement or coronary bypass surgery on soluble adhesion molecule levels.
Correlation of soluble adhesion molecule levels with endothelial expression
Measurement of soluble CAMs in peripheral blood presents a potentially attractive means of assessing inflammatory activity and the state of endothelial activation in a variety of disease processes. An understanding of the relationship between expression of adhesion molecules in affected tissue and soluble CAM levels is therefore important in determining the status of soluble adhesion molecules as markers of tissue adhesion molecule expression. Soluble CAMs are known to be generated by at least two mechanisms: enzymatic cleavage from the cell surface and alternative splicing of messenger RNA (without being expressed at cell surface) (8). Endothelial-selectin is expressed only on activated endothelial cells, and soluble E-selectin is generated only by enzymatic cleavage from cell surface E-selectin (5). Therefore, on theoretical grounds, there should be good correlation between serum levels and tissue expression in nonrheumatic AS, in which the disease process is relatively localized. Our findings support the usefulness of soluble E-selectin levels to assess endothelial activation in “degenerative” AV disease.
There was poor correlation between soluble ICAM-1 levels and ICAM-1 on valvular endothelium. ICAM-1 is found on a variety of cell types, including leukocytes, endothelium, epithelium and smooth muscle cells (5), therefore sICAM-1 in serum has many potential sources of origin. Soluble ICAM-1 can be generated by both enzymatic cleavage and alternative splicing, and it is possible that the major source of sICAM-1 formation is alternative splicing. Studies in vascular beds in mice and in patients with inflammatory bowel disease did not demonstrate a relationship between soluble ICAM-1 and endothelial expression of ICAM-1. Our results support the lack of association between soluble ICAM-1 and endothelial expression of ICAM-1 (9,10). VCAM-1 is also expressed on multiple cell types (5), and therefore, soluble VCAM-1 levels may not reliably reflect endothelial expression of the molecule.
Adhesion molecule expression on valvular endothelium
Adhesion molecule expression on diseased valves may play a part in the initiation or progression of the disease process. Adhesion molecule expression can be upregulated in response to cytokines or because of hemodynamic forces. The cytokines tumor necrosis factor alpha and interleukin-1 have been shown to induce production of ICAM-1, VCAM-1 and E-selectin (11). In our study, there was a strong association between the endothelial expression of all three adhesion molecules. The biomechanical paradigm of endothelial cell activation has not been as clearly delineated as the humoral model (12). Human umbilical vein endothelial cells exposed to steady unidirectional laminar shear stress in vitro show a time-dependent increase in ICAM-1 expression, while VCAM-1 and E-selectin remain unchanged. In contrast, humoral (cytokine and bacterial endotoxin) activation resulted in induction of all three adhesion molecules in the same model (12–14). The effect that turbulent pulsatile flow, as seen in AV disease, would have on adhesion molecule expression is unknown. These considerations suggest that the increased expression of adhesion molecules on valvular endothelium that we observed is more likely to be cytokine induced rather than due to shear forces. This is supported by our finding of adhesion molecules on autopsy valves that were diseased but not stenotic and therefore not subject to hemodynamic disturbances. We observed adhesion molecules on tricuspid and bicuspid valves, suggesting a common mechanism of disease.
The nature of the inflammatory stimulus that may be involved is not clear. Both diabetes and hyperlipidemia have been reported as risk factors for AS (15). No reports on the possible role of micro-organisms have yet been published. Activated T lymphocytes and macrophages have been demonstrated in degenerative tricuspid AV disease (16), but this aspect has not been examined in bicuspid aortic valve disease.
Adhesion molecules were not demonstrated on all of the degenerative valves studied. This may be due to many factors. It is possible that degenerative AV disease may have heterogeneous etiologies where the molecular mechanisms may vary. Also temporal factors may be responsible, with expression varying at different stages of the disease process.
We studied segments of the surgically removed AVs. The expression of adhesion molecules may not be uniform across the valve, and we could have missed areas of the valve that had more intense expression of adhesion molecules. Changes in shear stress are known to cause changes in adhesion molecule expression (13), and it is possible that there might have been increased adhesion molecule expression on aortic endothelium beyond the valve that we did not sample. We did not study adhesion molecule expression on circulating leukocytes that are known to express ICAM-1 and VCAM-1 and could have been important sources of soluble ICAM and VCAM-1 formation in our study. The standard deviation for soluble E-selectin levels was large, particularly in the six month postoperative samples. However, there was no overlap in the 95% confidence intervals for the means between preoperative levels when compared with the controls or the 18-month postoperative levels.
Conclusions and possible clinical implications
This study has demonstrated for the first time the expression of the adhesion molecules E-selectin, ICAM-1 and VCAM-1 in degenerative AV disease. Soluble E-selectin levels were elevated preoperatively, correlated strongly with endothelial expression and fell significantly 18 months postoperatively. This suggests that soluble E-selectin levels in peripheral blood may be useful as a marker of endothelial activation. These findings support the possibility that inflammation may play a role in the pathogenesis of the disease. Further studies exploring the role of cytokines and transcription factors in the process may refine targets for preventing or slowing disease progression.
We are grateful to Dr. Alan Kelly, biostatistician, for his statistical advice and analyses.
☆ This research was supported by a grant from the Board of the Royal City of Dublin Hospital to Dr. Ghaisas.
- aortic regurgitation
- aortic stenosis
- aortic valve
- cellular adhesion molecule(s)
- enzyme-linked immunosorbent assay
- endothelial selectin
- intercellular adhesion molecule 1
- vascular cell adhesion molecule 1
- Received January 20, 2000.
- Revision received June 1, 2000.
- Accepted July 14, 2000.
- American College of Cardiology
- Schoen F.J
- Mehrabian M,
- Demer L.L,
- Lusis A.J
- Jones S.C,
- Banks R.E,
- Haidar A,
- et al.
- Olsson M,
- Dalsgaard C,
- Haegarstrand A,
- Rosenqvist M,
- Ryden L