Enzymatically Modified Nonoxidized Low-Density Lipoprotein Induces Interleukin-8 in Human Endothelial Cells
Role of Free Fatty Acids
Background— Treatment of low-density lipoprotein (LDL) with a protease and cholesterolesterase transforms the lipoprotein to an entity that resembles lipoprotein particles in atherosclerotic lesions, which have a high content of free cholesterol, reflecting extensive de-esterification in the intima. Because de-esterification would occur beneath the endothelium, we examined the effects of enzymatically modified LDL (E-LDL) on cultured endothelial cells.
Methods and Results— Incubation of endothelial cells with E-LDL provoked selective accumulation of interleukin (IL)-8 mRNA and production of the cytokine. Chemical analyses and depletion experiments indicated that the effect was caused by the presence of free fatty acids in the altered lipoprotein. Reconstitution studies demonstrated that the oleic and linoleic acids associated with E-LDL are particularly effective IL-8 inducers. The effects of E-LDL on endothelial cells could be abrogated with albumin.
Conclusion— IL-8 is required for rolling monocytes to adhere firmly to the endothelium; thus, the findings reveal a link between subendothelial entrapment of LDL, cleavage of cholesterol esters, and monocyte recruitment into the lesion.
Received July 2, 2002; revision received August 23, 2002; accepted September 3, 2002.
Although upregulation of adhesion molecules and chemotaxis are obviously fundamental to monocyte recruitment to the atherosclerotic lesion, identification of the cytokines and mediators involved is still ongoing.1 Macrophage-derived monocyte chemotactic protein (MCP)-1 seems to be of prime importance, but the question still remains of how recruitment is initiated. One possibility is that the lipoprotein itself triggers this event, whereby 2 distinct mechanisms can be envisaged. (1) Low-density lipoprotein (LDL) may become altered in the intima so that it acquires immune-activating capacity, thus leading to the generation of chemotactic molecules. (2) LDL might be modified so that it can directly induce production of a suitable chemokine in the overlying endothelial cells. In fact, oxidized LDL (ox-LDL) stimulates endothelial cells to secrete interleukin (IL)-8.2 The potential relevance of this finding has been highlighted by 2 recent discoveries. (1) IL-8 is crucial for firm adherence and transmigration of circulating monocytes into the intima,3 and (2) elimination of the IL-8 receptor analogue reduces atherogenesis in apolipoprotein-E knockout mice.4 It is known that IL-8 is present in human atherosclerotic lesions,5 and all the data thus implicate IL-8 as a key player in atherogenesis.
Assuming that IL-8 is initially produced by endothelial cells, what is the nature of the stimulus involved? Ox-LDL is naturally a first candidate. However, we are pursuing a new hypothesis on atherogenesis that requires us to consider another possibility. Our hypothesis envisages tissue-stranded LDL as subject to enzymatic, nonoxidative modification through the combined action of proteinases and cholesterolesterase.6,7 Treatment of LDL with these enzymes in vitro generates lipoprotein particles that are similar to lesion-derived LDL in structure, biological properties, and composition. Trypsin has routinely been used as the proteolytic enzyme, but other proteases, including cathepsin H (P. Suriyaphol, unpublished data, 2002), have been found to exert equivalent effects. In contrast to ox-LDL, enzymatically modified LDL (E-LDL) binds C-reactive protein and activates complement; this could provide a first signal for monocyte recruitment.8 E-LDL is recognized and taken up by macrophages and stimulates MCP-1 production in these cells.9 E-LDL can also directly induce adhesion and transmigration of monocytes through endothelial cell monolayers.10 These in vitro findings have their in vivo correlates: deposits of E-LDL are detectable in early atherosclerotic lesions in colocalization with C-reactive protein and complement.11 Like in vitro–generated E-LDL, lesioned LDL has a high content of free cholesterol, rendering it apparent that extensive de-esterification of cholesterol esters must indeed occur in the lesions.12,13 In addition, like E-LDL, lesioned LDL does not contain large amounts of oxidized lipids.14–16
In the present article, we report that free fatty acids (FFAs) present in E-LDL selectively stimulate IL-8 secretion in endothelial cells. A direct link emerges between subendothelial entrapment of LDL and monocyte recruitment to the lesion.
Isolation and Modification of LDL
LDL was isolated as described previously9 and stored in the presence of 0.5 mmol/l EDTA for a maximum of 3 weeks. The preparations contained no detectable amounts of thiobarbiturate-reactive substances. To prepare ox-LDL, EDTA was removed by dialysis. LDL samples were oxidized at 37°C in the presence of 25 μmol/L CuSO4 for up to 18 hours and then dialyzed against 3 changes of PBS. For enzymatic modification, 1-mL samples of LDL containing 3 mg/mL cholesterol were sequentially incubated with 6 μg/mL trypsin (Sigma) for 5 hours and then 10 μg/mL trypsin inhibitor (Sigma) and 40 μg/mL cholesterolesterase (Roche) for 10 hours at 37°C.
To deplete lipoproteins of FFAs, FFA-free human serum albumin (HSA) at a final concentration of either 4% or 10% (wt/vol) was added to E-LDL or ox-LDL, and the samples were incubated for 1 hour at 37°C. Thereafter, 0.6 g of sucrose was dissolved in each sample, which was overlayered with 3 mL of 25% (wt/vol) sucrose and 0.4 mL of PBS. Lipoproteins were floated by centrifugation for 10 hours at 45 000 rpm at 10°C in a Beckman ultracentrifuge with swing-out rotor SW50.1, and the top fractions were collected. E-LDL preparations depleted of FFAs with 4% or 10% albumin are henceforth designated as E-LDL(d4) and E-LDL(d10), respectively. In reconstitution experiments, E-LDL(d10) was incubated with the following fatty acids for 1 hour at 37°C: linoleic acid (Sigma), oleic acid (Sigma), stearic acid (Sigma) and arachidonic acid (Calbiochem). Lipoprotein concentrations refer to the cholesterol concentrations throughout.
Labeling of Lipoproteins
Lipoproteins were labeled with 1,1′-dioctadecyl-3,3,3′,3′-tetramethyl-indocarbocyanine perchlorate (DiI; Molecular Probes), as described previously.17 One milligram of lipoprotein was admixed with 2 mL of lipoprotein-deficient serum and incubated for 8 hours at 37°C with 50 μL of DiI (3 mg/mL). The DiI-labeled lipoproteins were reisolated by floatation, as described above.
Cholesterol and FFA Analysis
Total cholesterol and free cholesterol were determined by cholesterol measurement assays from Roche and Wako, respectively. Lipoprotein preparations were analyzed for their FFA content by gas-liquid chromatography-mass spectrometry (GLC-MS) as fatty acid methyl esters; 100-μL aliquots were diluted with 750 μL of water, and 60 μg of heptadecanoid acid (17:0; Sigma) was added as an internal standard for quantification. The mixture was extracted 3 times with 500 μL of chloroform, and the FFAs in the combined organic layer were transferred to their fatty acid methyl ester derivatives by treatment with ethereal diazomethane (500 μL). After drying in a stream of nitrogen, samples were dissolved in 50 μL of chloroform and analyzed by GLC-MS.
GLC-MS was performed with a Hewlett-Packard model 5989 instrument equipped with a HP-5 MS capillary column (30 m, 0.25 mm ID, 0.25 μm film thickness) using a temperature gradient of 150°C (3 minutes) to 320°C at 5°C per minute.
Human endothelial cells (Ea.hy.926)18 were cultured in Dulbecco’s modified eagle medium containing 10% fetal calf serum at 37°C in 5% CO2. Cultures were passaged every third day. Primary endothelial cells from human aorta were purchased from Promocell and cultured in media supplied with the cells. Cells were seeded 1 day before treatment with the various agents for 24 hours.
Lipid Staining With Oil-Red-O
Cells were washed twice with PBS, fixed in 10% formaldehyde in PBS for 30 minutes at room temperature, and stained with a saturated solution of Oil-Red-O (Sigma) in 60% isopropanol in PBS for 3 hours.
Measurement of Lipoprotein Uptake
After incubation with DiI-labeled lipoproteins, endothelial cells were washed twice with PBS containing 0.4% BSA and 3 times with PBS alone; 1 mL of cell lysis reagent (0.1% wt/vol; SDS dissolved in 0.1 mol/L NaOH) was added, and fluorescence was measured in triplicate of 200 μL of lysate in a Titertek Fluoroskan II microplate reader (Flow Laboratories) with excitation and emission wavelengths set at 544 and 590 nm, respectively. Cellular protein was determined in duplicate 25-μL samples by the Lowry method. Fluorescence of DiI-lipoproteins diluted in the lysis reagent was measured to determine the specific fluorescence intensity of each preparation. Uptake measurements were expressed as nanograms of lipoprotein cholesterol per milligram of cellular protein.17
Immunoassay kits for the determination of chemokines and cytokines were purchased from R&D, and tests were performed according to the supplier’s instructions.
RNA Isolation and RNase Protection Assay
Total RNA was isolated using the RNAwiz kit (Ambion) and dissolved in diethylpyrocarbonate-treated water. RNase protection assays were performed as detailed in the manual accompanying the BD RiboQuant kit (Becton Dickinson). Chemokine mRNAs (LTN, RANTES, Ip-10, MIP-1β, MIP-1α, MCP-1, IL-8, and I309) and 2 control RNAs (L32, GAPDH) were simultaneously analyzed using the hCK-5 template set. Separation of protected probes was done on NOVEX QuickPoint gels.
Data were analyzed by ANOVA.
Combined Treatment With Trypsin and Cholesterolesterase Results in Cleavage of Cholesterol Esters in E-LDL
When LDL was treated with cholesterolesterase alone, no alteration in the ratio of cholesterol esters to free cholesterol was noted. However, when a combination of trypsin plus cholesterolesterase was used, the ratio of cholesterol esters to free cholesterol progressively decreased (Figure 1). These results confirmed the report that proteolytic nicking of apolipoprotein B renders cholesterol esters accessible to the action of cholesterolesterase.19
Identification and Partial Removal of FFAs From E-LDL
The amount of total FFAs in LDL (7.7 μg/mg cholesterol) increased by almost 2 orders of magnitude after enzymatic treatment (to 600 μg/mg in E-LDL). E-LDL preparations were analyzed for their content of various FFAs, and the results are given in the Table. Attempts to deplete the lipoprotein preparations of FFAs by dialysis failed, and a protocol was therefore developed that rendered their removal possible without the use of delipidating and denaturing steps. The method was based on the binding of FFAs by delipidated human serum albumin, followed by reisolation of E-LDL by floatation in sucrose. As shown in the Table, ≈50% and 90% of the FFAs could be removed with 4% and 10% HSA, respectively; thus differences in the extractability of individual FFAs were observed. In particular, linoleic and oleic acid were efficiently removed, but ≈50% of stearic acid remained associated with the lipoprotein, even when 10% albumin was used.
FFAs Render E-LDL Cytotoxic for Endothelial Cells
When freshly prepared E-LDL was applied to cultured endothelial cells, considerable cytotoxicity was observed, and incubation with 50 μg/mL of the lipoprotein led to the death of >95% of the cells within 4 hours, as evidenced by depletion of cellular ATP and by positive trypan blue staining (data not shown). Cytotoxicity was markedly diminished after partial removal of the FFAs with albumin. At this point, the cells tolerated 100 μg/mL E-LDL(d4) and up to 300 μg/mL E-LDL(d10).
Enhanced Uptake of E-LDL by Endothelial Cells
Cells were incubated with DiI-labeled LDL or E-LDL for 24 hours, and uptake of the lipoproteins was quantified by fluorescence measurements. Compared with native LDL, enhanced uptake of E-LDL was observed that was similar in extent to that previously reported for acetylated LDL20 (Figure 2a). When cells were stained with Oil-Red-O, cells that had been incubated with E-LDL, but not with LDL, assumed a discrete foam cell appearance (Figure 2b).
E-LDL Induces IL-8 Secretion in Endothelial Cells
In control pilot experiments, we found that incubation with non-delipidated HSA induced IL-8 in endothelial cells. The flotation protocol circumvented this problem because the albumin quantitatively remained at the bottom of the gradients and, consequently, there was no background stimulation. Cells were incubated with E-LDL, and IL-1β, IL-8, tumor necrosis factor (TNF)-α, and MCP-1 were quantified in the cell supernatants. The modified lipoprotein was found to induce secretion of IL-8 selectively. Thus, the dose-response curves obtained with E-LDL(d4) and E-LDL(d10) were different. As shown in Figures 3 and 4⇓, IL-8 induction peaked at 100 μg/mL E-LDL(d4), whereas 200 μg/mL E-LDL(d10) was required to produce the same peak effect. The drop in IL-8 response at high E-LDL(d4) concentrations correlated with cytotoxicity of the lipoprotein.
A comparison between the effects of E-LDL and ox-LDL is shown in Figure 5. In accord with published work,2 ox-LDL also induced IL-8, although to a weaker extent than observed with E-LDL(d4) (P<0.01). Incubation with HSA also abrogated the stimulating capacity of ox-LDL (P<0.01). There was no statistical difference between E-LDL(d10), ox-LDL(d4), and ox-LDL(d10).
RNase Protection Assays Confirm Selectivity of IL-8 Induction by E-LDL
We then screened for possible upregulation of 8 other cytokines using RNase protection assays. These experiments confirmed the remarkably specific effect of E-LDL on endothelial cells. As shown in Figure 6, E-LDL(d4) applied at 50 μg/mL provoked marked increases of IL-8 mRNA, but not of mRNA of other cytokines, including MCP-1. This contrasted conspicuously with the response to TNF-α, where accumulation of mRNA for both MCP-1 and IL-8 was observed, as was known from the literature.21 That our results might represent artifacts caused by lipopolysaccharide contamination of E-LDL preparations was virtually excluded by the finding that E-LDL(d10) was devoid of stimulating effects when applied at 50 μg/mL (Figure 6). When applied at 200 μg/mL, selective IL-8 mRNA accumulation was again noted (data not shown).
Reconstitution Experiments Identify FFAs in E-LDL as the Stimulus for IL-8 Secretion
The above findings led us to consider that the FFAs contained in E-LDL might be responsible for stimulating endothelial cells to secrete IL-8. To test this possibility, we used E-LDL(d10) at a concentration that was barely effective (50 μg/mL) and performed reconstitution experiments using arachidonic, linoleic, oleic, and stearic acid. As shown in Figure 7, reconstitution of E-LDL(d10) with either linoleic acid or oleic acid to concentrations equivalent to those found in E-LDL(d4) led to regaining IL-8 stimulatory capacity (P<0.01). This was not observed when stearic or arachidonic acid was used. Moreover, significant IL-8 stimulation did not occur when equivalent amounts of the FFAs were applied in cyclodextrin instead of in combination with E-LDL (P<0.01; Figure 7).
In the intima, one or several biochemical modifications transforms LDL into an entity that acquires the capacity to signal to monocytes and/or macrophages and ultimately to cause its own removal. Accumulating evidence leads us to depart from the conventional view that oxidative processes play the central role. Instead, we propose that nonoxidative, enzymatic modification of LDL with concomitant de-esterification of cholesterol esters renders the lipoprotein atherogenic. Lesioned LDL does have a high content of free cholesterol, which implies that de-esterification occurs in tissues.12,13 Enzymatic modification in vitro transforms LDL to lipidic droplets resembling those derived from atherosclerotic lesions,6,8,16 and extensive deposits of E-LDL can be detected by immunohistochemistry in early lesions.11
Therefore, we examined the effects of E-LDL on endothelial cells. The results are of interest because they show that (1) E-LDL itself can provoke IL-8 secretion, and (2) FFAs contained in an atherogenic LDL-derivative may activate endothelial cells independent of oxidation.
We previously observed that cytotoxicity of E-LDL toward monocytes and smooth muscle cells could be abrogated through the addition of 4% HSA.9,22 Similar findings were then made with endothelial cells, indicating that FFAs were responsible. Avid binding of FFAs to HSA23 was subsequently exploited to obtain FFA-depleted E-LDL preparations. We then observed that E-LDL preparations depleted of FFAs with 4% HSA [E-LDL(d4)] vigorously stimulated IL-8 secretion when applied at subcytotoxic concentrations (50 μg/mL), whereas E-LDL(d10) applied at the same concentration did not. However, IL-8 secretion occurred when the concentration of E-LDL(d10) was raised so that the content of FFAs approached that of stimulatory doses of E-LDL(d4). Similarly, IL-8 induction by ox-LDL was also abrogated after incubation with HSA. That FFAs might directly be the stimulating element was then confirmed by reconstitution experiments. When E-LDL(d10) was reconstituted with either linoleic acid or oleic acid, stimulatory activity reappeared. This was not observed when stearic acid was used. Because stimulatory effects were low when the fatty acids were applied in methyl-β-cyclodextrin, FFAs may need to be contained in their natural vehicle to exert their action optimally. In this context, the observation that endothelial cells take up E-LDL may be relevant: staining with Oil-Red-O revealed the accumulation of cholesteryl esters in the cells. Endothelial cells take up acetylated20,24 and ox-LDL,25 and ongoing experiments are addressing whether uptake of E-LDL follows overlapping pathways. Like recent findings made in macrophages,26 we could not inhibit E-LDL–mediated foam cell formation with conventional inhibitors (fucoidan and polyinosidic acid, inhibitors of scavenger receptor (SR)-A used at 200 μg/mL, and anti-CD36, inhibitor of SR-B type II used at 5 μg/mL).
Electronegative LDL from normolipemic subjects reportedly induced IL-8 and MCP-1 in human endothelial cells. Among the few identified differences between electronegative and nonelectronegative LDL, a 4-fold higher concentration of FFAs was noted to be present in the former.27 We have calculated that the amounts of FFAs applied by Castellarneau et al27 were of the same order of magnitude as those used in our experiments. The cause of the discrepancy with regard to MCP-1 stimulation remains to be clarified. Slight stimulation of IL-8 secretion in endothelial cells treated with linoleic acid (180 μmol/L) was described by Young et al,28 and we observed a comparable, minor effect on EA.hy.926 cells when linoleic acid was applied at 50 μmol/L in methyl-β-cyclodextrin. If our concept that enzymatic modification of LDL is a key event in atherogenesis is correct, FFAs may emerge as far more important mediators than hitherto suspected.
The mechanisms underlying the remarkably selective IL-8 stimulation await elucidation. Peroxisome proliferator-activated receptors (PPARs) are likely to be involved, because these nuclear receptors are activated by FFAs.29 Oxidized phospholipids induce IL-8 and MCP-1 in endothelial cells through the activation of PPARs, a process that may be dependent on phospholipase-A2 (ie, on the liberation of FFAs).30
Our hypothesis proposes that enzymatic modification of LDL is a meaningful process that triggers physiological events leading to the removal of stranded cholesterol, a process that occurs only when the normal transport system is overloaded. Enzymatic modification confers on LDL the capacity to bind C-reactive protein, activate complement, and induce macrophage foam cell formation.6–8 Because the concomitantly liberated fatty acids will stimulate IL-8 production, a novel link is now provided between intimal LDL retention, its enzymatic modification, and monocyte recruitment to the lesion.
This work was supported by the Deutsche Forschungsgemeinschaft (grant Bh 2/3-1) and by a scholarship from the Mahidol-University, Bangkok, Thailand (to Dr Suriyaphol).
Bhakdi S, Dorweiler B, Kirchmann R, et al. On the pathogenesis of atherosclerosis: enzymatic transformation of human low density lipoprotein to an atherogenic moiety. J Exp Med. 1995; 182: 1959–1971.
Bhakdi S, Torzewski M, Klouche M, et al. Complement and atherogenesis: binding of CRP to degraded, nonoxidized LDL enhances complement activation. Arterioscler Thromb Vasc Biol. 1999; 19: 2348–2354.
Klouche M, Gottschling S, Gerl V, et al. Atherogenic properties of enzymatically degraded LDL: selective induction of MCP-1 and cytotoxic effects on human macrophages. Arterioscler Thromb Vasc Biol. 1998; 18: 1376–1385.
Klouche M, May AE, Hemmes M, et al. Enzymatically modified, nonoxidized LDL induces selective adhesion and transmigration of monocytes and T-lymphocytes through human endothelial cell monolayers. Arterioscler Thromb Vasc Biol. 1999; 19: 784–793.
Torzewski M, Klouche M, Hock J, et al. Immunohistochemical demonstration of enzymatically modified human LDL and its colocalization with the terminal complement complex in the early atherosclerotic lesion. Arterioscler Thromb Vasc Biol. 1998; 18: 369–378.
Steinbrecher UP, Lougheed M. Scavenger receptor-independent stimulation of cholesterol esterification in macrophages by low density lipoprotein extracted from human aortic intima. Arterioscler Thromb. 1992; 12: 608–625.
Seifert PS, Hugo F, Tranum-Jensen J, et al. Isolation and characterization of a complement-activating lipid extracted from human atherosclerotic lesions. J Exp Med. 1990; 172: 547–557.
Edgell CJ, McDonald CC, Graham JB. Permanent cell line expressing human factor VIII-related antigen established by hybridization. Proc Natl Acad Sci U S A. 1983; 80: 3734–3737.
Chao FF, Blanchette-Mackie EJ, Tertov VV, et al. Hydrolysis of cholesteryl ester in low density lipoprotein converts this lipoprotein to a liposome. J Biol Chem. 1992; 267: 4992–4998.
Lakshminarayanan V, Beno DW, Costa RH, et al. Differential regulation of interleukin-8 and intercellular adhesion molecule-1 by H2O2 and tumor necrosis factor-alpha in endothelial and epithelial cells. J Biol Chem. 1997; 272: 32910–32918.
Klouche M, Rose-John S, Schmiedt W, et al. Enzymatically degraded, nonoxidized LDL induces human vascular smooth muscle cell activation, foam cell transformation, and proliferation. Circulation. 2000; 101: 1799–1805.
Hamilton JA. Fatty acid transport: difficult or easy? J Lipid Res. 1998; 39: 467–481.
Voyta JC, Via DP, Butterfield CE, et al. Identification and isolation of endothelial cells based on their increased uptake of acetylated-low density lipoprotein. J Cell Biol. 1984; 99: 2034–2040.
Kruth HS, Zhang WY, Skarlatos SI, et al. Apolipoprotein B stimulates formation of monocyte-macrophage surface- connected compartments and mediates uptake of low density lipoprotein- derived liposomes into these compartments. J Biol Chem. 1999; 274: 7495–7500.
De Castellarnau C, Sanchez-Quesada JL, Benitez S, et al. Electronegative LDL from normolipemic subjects induces IL-8 and monocyte chemotactic protein secretion by human endothelial cells. Arterioscler Thromb Vasc Biol. 2000; 20: 2281–2287.