Plasma Cholesterol Regulates Soluble Cell Adhesion Molecule Expression in Familial Hypercholesterolemia
Background Hypercholesterolemia is associated with endothelial dysfunction. On the basis of the hypothesis that high plasma cholesterol per se may be a sufficient stimulus to upregulate endothelial adhesiveness and that this phenomenon might be reversible, soluble endothelial leukocyte adhesion molecules (sELAMs) were studied in patients with familial hypercholesterolemia undergoing LDL apheresis.
Methods and Results Selective LDL absorption by dextran sulfate columns was used to treat plasma volumes of 6.5 to 9.2 L; after LDL apheresis, total cholesterol, LDL cholesterol, apolipoprotein B, triglycerides, and lipoprotein(a) levels were reduced by 74%, 82%, 79%, 56%, and 86%, respectively. Soluble intercellular adhesion molecule-1 (sICAM-1) and sELAM-1 were measured before, immediately after, and 2 and 6 days after LDL apheresis. Basal sICAM-1 and sELAM-1 values were higher than in healthy control subjects. After LDL apheresis, they decreased (P<.0001 and P<.0004, respectively); their removal by extracorporeal circulation components was excluded. Individual pretreatment and posttreatment values of sICAM-1 and sELAM-1 were positively correlated (P<.0001 and P<.001, respectively) with total cholesterol; their rebound curves showed patterns similar to the total cholesterol rebound curve but not to the triglyceride and lipoprotein(a) curves.
Conclusions In the absence of changes in clinical chemical parameters, tumor necrosis factor-α, interleukin-6, and acute-phase reactant proteins, these results confirm in a clinical setting the upregulation of endothelial adhesiveness observed in experimental hypercholesterolemia and suggest a direct role for cholesterol in regulating this phenomenon, at least in familial hypercholesterolemia.
Intensive clinical end-point trials have demonstrated that interventions aimed at lowering lipid levels have beneficial effects in terms of reduction in the incidence of cardiac events and in mortality from coronary artery disease.1 The positive clinical results are superior to angiographic evidences.2 To account for these large clinical benefits that have been observed even with relatively short periods of intervention,3 several hypotheses such as plaque progression halting and stabilizing of rupture-prone lesions have been suggested4 ; however, the mechanisms responsible for these observations remain poorly outlined. The restoration of endothelium-mediated vasomotion induced by antilipidemic interventions, recently demonstrated in the coronary3 and peripheral5 circulations, indicates that these therapies may act on the paracrine functions of the endothelium. Answers are still lacking to the questions not only of how these functional endothelial improvements are generated and which stimulus or stimuli is or are shut off but also of whether lipid-lowering therapies influence early and critical atherogenic events such as the adhesion and transendothelial migration of circulating leukocytes, an essential target for the “regression quartet.”6 Whereas (1) the adherence of circulating blood monocytes to the intact intimal surface is one of the earliest events in experimental atherogenesis,7 (2) this event is mediated by inducible ELAMs,8 and (3) higher levels of sICAM-1 have been found in human atherosclerosis,9 we addressed the specific question of whether this hallmark of molecule expression would be influenced by cholesterol levels and whether cholesterol-lowering therapy would reverse the phenomenon independently of any preexisting alteration. We studied the effect of an acute, selective, and massive removal of cholesterol, obtained by LDL apheresis, on soluble ELAMs as well as on TNF-α, the most representative among cell adhesion molecules regulating cytokines, in patients affected only by FH.
Eight subjects affected by FH were recruited: one 12-year-old homozygous boy and seven heterozygotes, two of whom were women, aged between 30 and 50 years. Diagnosis of heterozygous FH was based on the presence of primary hypercholesterolemia, tendon xanthomas, and family history of hypercholesterolemia. The homozygous patient is a carrier of a novel point mutation of the LDL receptor gene.10 The heterozygotes were treated by LDL apheresis because of limited efficacy of diet and combination drug therapy. All patients manifested various degrees of cardiovascular disease and were free from any other form of organ or systemic, chronic, or recurrent disease and additional risk factors for atherosclerosis; furthermore, when investigated for adhesion molecules and cytokines, they all had phogosis indexes and acute-phase reactant protein (transferrin, α-1 acid glycoprotein, haptoglobin, C-reactive protein complement fractions C3 and C4) levels within normal ranges and undetectable levels of IL-6.
An extracorporeal venous-venous circulation provided a blood flow of 90 to 120 mL/min; the initial heparin bolus was 1500 IU followed by continuous infusion of 1000 IU/h. Plasma was separated by a polysulfone hollow fiber filter (Sulflux FS-05). Two columns, each containing 150 mL of dextran sulfate cellulose, a specific sorbent of apo B–containing lipoproteins (LiposorberLA-15), were alternately flushed with plasma and regenerated with 0.7 mol/L saline solution followed by rinsing with Ringer’s solution under control of an automatic adsorption-desorption apparatus (MA-01; Kaneka Co). During a 3.5- to 4-hour period, a total plasma volume of 6.5 to 9.2 L was treated, corresponding to 2.5 to 3 times that of each patient’s plasma volume.
Laboratory Measurements and Study Design
Serum cholesterol, TG, and HDL-C levels were assessed enzymatically by automated procedures; HDLs were isolated with heparin and manganese. LDL-C values were calculated according to Friedewald et al.11 Apolipoproteins A-I (apoA-I) and apo B were assayed by rate immunonephelometry (Beckman BN 100). IRMA was used to measure Lp(a) (Pharmacia), IL-6, and TNF-α (Medgenix Diagnostic). Intra-assay and interassay coefficients of variation were 5.4% to 5.0% and 3.8% to 5.2%, respectively. ELISA tests were used for sICAM-1 (T Cell Diagnostics, Inc) and sELAM-1 (Bender Med System), with intra-assay and interassay coefficients of variation of 2.4% to 3.8% and 3.7% to 4%, respectively, and sensitivities of 3.3 and 1.6 ng/mL.
All samples for IRMAs and ELISAs, stored at −80°C, were analyzed in one batch. Routine chemical clinical analyses, including measurement of acute-phase reactant protein, determined by standard methods under strict quality control, were performed on samples drawn immediately before each procedure; postapheresis samples were taken immediately before saline infusion to wash out the extracorporeal circulation components so as to avoid the influence of hemodilution. For each patient, a rebound curve was constructed by assaying lipids, lipoproteins, and adhesion molecules in blood samples obtained before, immediately after, and 2 and 6 days after LDL apheresis. To verify possible absorption by the extracorporeal circulation components, sELAM-1 and sICAM-1 levels were monitored in samples taken simultaneously from the inlets and outlets of the plasma separator and the dextran sulfate column at different treated plasma volumes (400 and 9100 mL).
Routine laboratory tests and flogosis indexes were within normal ranges and did not change throughout the study. The mean lipid values before apheresis were as follows: total cholesterol, 319.9±27 mg/dL; LDL-C, 228.4±16 mg/dL; Apo B, 168.8±18 mg/dL; HDL-C, 43±22 mg/dL; TG, 110.6±35 mg/dL; and Lp(a), 35 mg/dL median (range, 17 to 114 mg/dL). After apheresis, mean percent reductions of total cholesterol, LDL-C, apo B, HDL-C, TG, and Lp(a) were 74%, 82%, 79%, 7%, 56%, and 86%, respectively. For the sake of brevity, we chose to refer and relate all other results and interpretations only to total cholesterol (Fig 1⇓).
Baseline sICAM-1 and sELAM-1 levels were significantly higher in FH patients than in healthy control subjects (n=13) matched for age and sex (358.7±61.4 versus 304±52 ng/mL and 43.5±23.7 versus 31.6±12.8 ng/mL, respectively; P<.01). They were significantly (P<.0001 and P<.004, respectively) reduced after each LDL apheresis (Fig 1⇑); the postapheresis mean value for sICAM-1 was 266.4±62.1 ng/mL, and the mean value for sELAM-1 was 32.1±16.5 ng/mL. After treatment of 400 mL of plasma, mean concentrations of sICAM-1 measured in the inlets and outlets of the plasma separators and the dextran sulfate column were 358.5±62, 337.4±63, 350.1±60, and 340.3±62 ng/mL, respectively; sELAM-1 levels were 43.5±21, 42.1±24, 43.0±21, and 42.8±23 ng/mL, respectively; after 9100 mL, sICAM-1 and sELAM-1 levels were 269.5±61, 270.4±60, 270.5±63, and 270.5±63 ng/mL and 32.8±16, 31.9±15, 31.6±14, and 30.9±17 ng/mL, respectively. Whatever the treated volumes, no significant differences were found between sICAM-1 and sELAM-1 levels in inlets versus outlets of the plasma separator or in inlets versus outlets of the dextran sulfate column by ANOVA. Individual pretreatment and posttreatment values of both sICAM-1 and sELAM-1 were positively and significantly (P<.0001 and P<.001, respectively) correlated with total cholesterol and with each other (Fig 2⇓). Values of sICAM-1 and sELAM-1 determined during this time course sampling correlated with total cholesterol values (P<.02). At 48 hours and 6 days after apheresis, cholesterol increased to 158.3±32 and 210.3±51.9 mg/dL, respectively (mean±SD); sICAM-1 increased to 253.3±41.3 and 267±74.2 ng/mL, respectively (mean±SD); and sELAM-1 increased to 36.8±14 and 38±6.9 ng/mL, respectively (mean±SD). Rebound curves after apheresis of sICAM-1 and sELAM-1 showed a pattern very similar to that shown by total cholesterol (Fig 3⇓). TG and Lp(a) levels regressed to baseline values in 48 hours. TNF-α plasma values in FH patients were significantly (P<.01 by unpaired Student’s t test) higher than in control subjects (21.23±12 versus 12.93±3.1 pg/mL [mean±SD]) and were unmodified by the treatment.
FH patients undergoing therapy with selective LDL apheresis12 may represent an intriguing experimental model because it is possible to achieve a safe, acute, selective, and massive lowering of cholesterol levels and to perform multiple observations in the same subject in well-circumstanced conditions; moreover, the conditions of these patients naturally return to baseline conditions within a fixed period of time.
We exploited this opportunity to investigate the possible role of hypercholesterolemia in inducing endothelial adhesiveness by studying the expression of molecules that mediate the adhesion of leukocytes to vascular endothelium. The possibility of measuring cell adhesion molecule levels in the blood13 has offered the opportunity to study ex vivo their role in a number of pathological conditions (diabetes, immunologic disorders, chronic inflammation, and cancer). The issue of whether and to what extent cell adhesion molecule levels reflect the pathological status of the endothelium is still a matter of intensive and inconclusive debate,14 but as far as the specific point of endothelial adhesiveness is concerned, at the moment it seems likely that ELAM-1 (otherwise known as E-selectin) “is found only on activated endothelium. The demonstration of soluble E-Selectin in the blood would therefore be taken as conclusive evidence of endothelial activation.”14 15
In the present study, we have shown for the first time that FH is associated with elevated levels of sICAM-1 and sELAM-1; in our patients, high levels of soluble adhesion molecules seem not to be determined by inflammation events16 because the markers of inflammatory status, including IL-6,17 were within normal limits and did not change over time. The sizeable reduction of sICAM-1 and sELAM-1 by an acute and massive lowering of cholesterol levels (the mean mass decrease was ≈414 000 and 49 680 ng, respectively) cannot be attributed to a specific adsorption into the extracorporeal circulation components; the sensitivity of the assays used would have revealed differences between samples taken from the inlets and outlets of the extracorporeal circulation components inasmuch as the mean sICAM-1 and sELAM-1 concentration disappearances were 36.5 and 4.4 ng/mL, respectively, throughout the apheresis. By this analysis, the data would indicate that the decrease depends on a synthesis/release reduction and/or clearance increase that appeared to already be activated during the 4 hours of apheresis, but additional studies are necessary to elucidate which of the above-cited mechanisms are implicated.
Present knowledge about the metabolism of adhesion molecules does not provide an explanation of the acute plasma variations in sICAM (≈30% ) and sELAM (≈26%) or why after 6 days sELAM levels returned close to their pretreatment values whereas sICAM-1 levels remained reduced by ≈26%. Nevertheless, the data reported herein may lead to the elucidation of a regulatory role of plasma cholesterol levels on endothelial adhesiveness as described by soluble forms of endothelial adhesion molecules; the decrease in the level of these molecules was indeed concurrent with the removal of cholesterol during LDL apheresis, and their rebound curves in plasma paralleled that of cholesterol for 6 days, showing a positive correlation. TNF-α is a modulator of the inflammatory response that occurs once the endothelium has been exposed to injurious agents, and it regulates endothelial adhesiveness18 ; its plasma concentration in FH patients was significantly higher than in healthy control subjects. However, the fact that it was not affected by cholesterol reduction should rule out the possibility that the short-term cholesterol-lowering effect on endothelial adhesiveness was mediated by TNF-α in FH patients.
Since our study was completed, Hackman and colleagues19 have reported that severe hyperlipidemia is associated with increased levels of soluble cell adhesion molecules, but “aggressive” lipid-lowering treatment had only limited effects on their levels, suggesting that increased levels of soluble cell adhesion molecules may be related to underlying atherosclerosis. We agree with such an association; however, our findings of a correlation between lipid-lowering therapy and soluble cell adhesion molecules may be explained by different experimental characteristics, ie, the quality of the study population and/or drug versus selective LDL-apheresis therapy and/or the differences in cholesterol levels obtained; again, a cholesterol level <200 mg% seems to be critical, at least in a short therapeutic course. Whereas nothing can be inferred regarding other possible mechanisms by which cholesterol plasma levels may regulate endothelial adhesiveness, the decreased adhesiveness observed may be explained by the physical and biochemical shift of the milieu in which the endothelium finds itself (from a cholesterol concentration of ≈300 mg/dL to 70 mg/dL).
Additional clinical trials are necessary to determine whether there is a threshold plasma cholesterol level that must be achieved and maintained to prevent atherosclerotic vascular lesions and to stabilize atheromatous plaque and recover endothelial functions. Our data suggest that soluble cell adhesion molecules may be a possible marker of efficacy in lipid-lowering trials and that this model may be useful to establish, from a biological point of view, the plasma cholesterol levels desirable for a “healthy” (ie, nondysfunctional) endothelium.
Selected Abbreviations and Acronyms
|ELAM||=||endothelial leukocyte adhesion molecule|
|ELISA||=||enzyme-linked immunosorbent assay|
|sELAM-1||=||soluble endothelial leukocyte adhesion molecule-1|
|sICAM-1||=||soluble intercellular adhesion molecule-1|
|TNF-α||=||tumor necrosis factor-α|
- Received May 14, 1997.
- Revision received June 19, 1997.
- Accepted June 24, 1997.
- Copyright © 1997 by American Heart Association
Thompson GR. Angiographic trials of lipid-lowering therapy: end of an era? Br Heart J. 1995;74:143-147.
Schwartz CJ, Valente AJ, Sprague EA, Kelley JL, Cayatte AJ, Mowery J. Atherosclerosis, potential targets for stabilization and regression. Circulation. 1992;86(suppl III):III-117-III-123.
Cybulsky MI, Gimbrone MA. Endothelial expression of a mononuclear leukocyte adhesion molecule during atherogenesis. Science. 1991;251:788-791.
Lelli N, Garuti M, Ghisellini M, Tiozzo R, Rolleri M, Animale V, Ginocchio E, Naselli A, Bertolini S, Calandra S. Occurrence of multiple aberrantly spliced mRNAs of the LDL-receptor gene upon a donor splice site mutation that causes familial hypercholesterolemia. J Lipid Res. 1995;36:1315-1324.
Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low density lipoprotein cholesterol in plasma without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499-502.
Gordon BR, Kelsey SF, Bilheimer DW, Brown DC, Dau PC, Gotto AM, Illingworth DR, Jones PH, Leitman SF, Prihoda JS, Stein EA, Stern TN, Zavoral JH, Zwiener RJ. Treatment of refractory familial hypercholesterolemia by low-density lipoprotein apheresis using an automated dextran sulfate cellulose adsorption system. Am J Cardiol. 1992;70:1010-1016.
Hackman A, Abe Y, Insull W, Pownall H, Smith L, Dunn K, Gotto AM, Ballantyne CM. Levels of soluble cell adhesion molecules in patients with dyslipidemia. Circulation. 1996;93:1334-1338.