Atherosclerosis (AS) is a challenge for public health, and its major clinical sequelae, including coronary heart disease (CHD) and cerebrovascular disease, account for approximately half of all deaths occurring annually.¹ Multiple risk factors such as serum lipid concentrations, smoking, diabetes mellitus, and hypertension contribute to the development of AS.² Recent studies in humans established a low level of high-density lipoprotein-cholesterol (HDL-C) in plasma as an independent risk factor of AS.^{[3], [4] and} 5 T. Gordon, W.P. Castelli, M.C. Hjortland, W.B. Kannel and T.R. Dawber, High density lipoprotein as a protective factor against coronary heart disease. The Framingham Study, Am J Med 62 (1977), pp. 707–714. Abstract | PDF (855 K) | View Record in Scopus | Cited By in Scopus (1178)^[5] Endothelial lipase (EL), which is a newly identified member of the lipase family, catalyzes the hydrolysis of HDL phospholipids and facilitates the clearance of HDL from the circulation.^{[6] and [7]} In recent years, EL has been a topic of intense research involved in metabolic syndrome,⁸ inflammation,⁹ hypertension,¹⁰ and AS.¹¹ This article is a review of the association between EL and several key characteristics of AS, namely low HDL-C, inflammation, and atherosclerotic plaque.

Molecular Structure and Function

EL was discovered in 1999 by 2 independent research groups,^{[12] and [13]} and it was established as a new member of the triglyceride lipase gene family, which also includes pancreatic lipase, lipoprotein lipase (LPL), and hepatic lipase (HL). EL exhibits a considerable molecular homology with LPL (45%) and HL (40%).^{[14] and [15]} It contains 482 amino acids and has a molecular mass of 55 kDa, with conservation of the catalytic triad as well as potential heparin and lipoprotein binding sites. Although alignment of EL with the human LPL and HL amino acid sequences revealed conservation of the catalytic triads, EL has a minimal homology in the lid domain,^{[12] and [13]} which is known to be critical in determining substrate specificity. So the substrate specificity of EL might be different from those of LPL and HL.

Unlike LPL or HL, EL did not hydrolyze triolein in the presence or absence of apolipoprotein C-II (apoC-II) by using medium from COS cells transfected with the human EL complementary DNA (cDNA).¹² However, a subsequent investigation indicated that EL has detectable triglyceridase activity, which does not require apoC-II.¹⁵ It is most important that EL has a significant activity of phospholipase A1, which cleaves fatty acids from the sn-1 position of phosphatidylcholine.^{[12] and [13]} The ratio of triglyceridase to phospholipase activity of EL was 0.65 compared with ratios of 24.1 for HL and 139.9 for LPL. This was confirmed by in vivo studies in which mice overexpressing EL seem to have a substantially higher phospholipase A1 activity in post-heparin plasma as compared with wild-type or knockout mice.⁶ Recently, Griffon et al¹⁶ documented that the active form of EL had a molecular weight higher than a simple monomer but less than a dimer.

EL Regulates HDL-C

Recently, increasing evidence has indicated that EL is important to serum HDL-C levels regulated in part by the lipase family. EL has mostly phospholipase and little triglyceridase activity, and HDL phospholipids represent a preferred substrate for this enzyme in in vitro assays.¹⁵

As a pattern of physiologic relevance, the EL-mediated hydrolysis of HDL particles determines the plasma levels of HDL-C, which was established in experimental animals using both overexpression as well as loss-of-function models. Transgenic expression of human EL in mice resulted in a 19% reduction in plasma HDL-C levels compared with wild-type control animals.⁶ Consistent with this finding, adenovirus-mediated over-expression of human EL in LDL receptor-deficient mice lead to more significantly decreased plasma concentrations of HDL-C than those of very LDL and LDL cholesterol.¹³ Furthermore, plasma post-heparin phospholipase activity was paralleled by an increase in the fractional catabolic rate of HDL in dose-dependent pattern when the recombinant adenoviral vector encoding human EL cDNA was injected into mice with different doses.¹⁷ Accordingly, similar results were also observed in EL overexpressed model of the C57 B6 mice.¹⁸ However, Jin et al⁷ observed that significantly increased (>40%) HDL-C and phospholipid levels were associated with increased HDL particle size and reduced HDL phospholipid turnover after EL was inhibited by administering a polyclonal antimurine EL antibody in wild-type, apoA-I-transgenic, and HL-deficient mice. In addition, HDL-C concentration is highly heritable and modifiable by physical activity, and its response to exercise varies among individuals. This variability might be associated with genetic polymorphisms of EL.¹⁹ In conclusion, the existing studies indicate EL takes part in the regulation of HDL-C metabolism. Respecting these results, EL might be a potential target to intervention of HDL-C, thus exerting effects on AS.

Recently, several human genome-wide association studies have suggested that a common variation near the EL gene (LIPG) locus is associated with HDL-C in humans.^{[20], [21], [22], [23], [24] and [25]} deLemos et al²⁶ first reported EL gene variants and confirmed 4 were derived from amino acid substitutions (Gly26Ser, Thr111Ile, Thr298Ser, and Asn396Ser), and 2 were located in the promoter region of LIPG (-410C/G and -303A/C). In addition, 3 rare alleles of EL (Gly26Ser, Thr298Ser, and -303A/C) were different between white individuals with normal (50 mg/dL) and high HDL-C (89 mg/dL), which suggests a potential association with higher HDL-C. These rare variants occurred in the black control cohort, with allele frequencies of 5.7%, 2.3%, and 1.8%, respectively. These variants were completely absent in the white control population, which suggests that these variants could conceivably contribute to the higher HDL-C levels in blacks compared with whites. Only 1 nonsynonymous variant Thr111Ile was found to be common (minor allele frequency >0.05) without significant differences in HDL-C levels in carriers of this variant and without significant difference in the allele frequency between the white high HDL-C cohort and control.

After that, several other studies have attempted to determine whether there is any impact of the Thr111Ile variant on a variety of cardiovascular index. An examination of the Thr111Ile variation in 372 individuals from the Lipoprotein and Coronary Atherosclerosis Study (LCAS) revealed a significant association between a SNP 584C/T in the LIPG and mean plasma levels of HDL-C. Patients with the TT allele have a 14% higher mean plasma HDL-C level compared with those with the CC allele. No significant genotype-by-treatment interactions were observed between the LIPG 584C/T SNP and response of HDL-C to fluvastatin treatment in the 2.5-year prospective LCAS.²⁷ Paradis et al²⁸ also found an association of Thr111Ile with increased levels of the HDL₃ subfraction in 281 females. Other investigators showed a weak association of Thr111Ile with increased HDL-C among 541 Japanese Americans.²⁹ In addition, the similar result was also shown by Tang et al³⁰ in 265 Chinese coronary artery disease cases and controls. However, these studies have been underpowered and lack functional evidence to support their conclusions.

In contrast, several studies failed to find association between LIPG variants Thr111Ile and HDL-C levels. Yamakawa-Kobayashi et al³¹ failed to detect an association of Thr111Ile with HDL-C levels in 340 Japanese children; Jensen et al³² presented the same findings among healthy Caucasian men and women from 3 independent studies. Shimizu et al³³ did not find an association of Thr111Ile with HDL-C in 107 Japanese acute myocardial infarction cases and controls. Mank-Seymour et al³⁴ confirmed the lack of significant association of Thr111Ile when individuals with normal and high HDL-C were compared (P = 0.36) in an intermediate HDL-C population (between 35 and 60 mg/dL HDL-C, n = 1458). More recently, Edmondson et al³⁵ definitively established that Thr111Ile was not associated with HDL-C from a combined sample of 3845 participants, and in vitro studies showed that Thr111Ile has normal lipolytic activity.

The underlying reasons for these discrepant findings might be derived from the differences in the subjects of different ethnicities, pathophysiologic circumstances, and lifestyle characteristics. In general, the more powerful and larger studies mentioned previously indicated that Thr111Ile was not significantly associated with HDL-C level.

However, Asn396Ser, which is a less common mutation in LIPG, was found to be significantly associated with increased HDL-C. The allele frequency of the Asn396Ser variant was higher in the high HDL-C group (2.4%) compared with the controls (1.2%) in white participants.²⁶ Asn396Ser is a nonconservative substitution, and it occurs in an area of sequence homology to LPL and HL where functional mutations have been reported. Asn396Ser was also significantly associated with increased HDL₂ and HDL₃ subfractions, increased HDL particle size, increased large HDL particles, and increased apoA-I levels.³⁵ In addition, a meta-analysis across 5 cohorts demonstrated that the low-frequency Asn396Ser variant is significantly associated with increased HDL-C. A subsequent functional analysis confirmed that the Asn396Ser variant decreased lipase activity significantly both in vitro and in vivo.³⁵

Several other LIPG gene variants were also reported to be associated with HDL-C. Mank-Seymour et al³⁴ found the intronic variants C+42T/In5 and T+2864C/In8 (P = 0.007 and 0.004, respectively) were significantly associated with HDL-C. Yamakawa-Kobayashi et al³¹ reported the promoter variant -384A/C and a variant in the 3' untranslated region of exon 10 (2237G/A) were significantly associated with increased HDL-C levels among healthy, school-aged Japanese children. Brown et al³⁶ demonstrated that Gly26Ser was significantly more common in African-American individuals with elevated HDL-C levels, and carriers of the Gly26Ser variant had significantly reduced plasma levels of EL protein. Although these rare variant were demonstrated to be associated with HDL-C levels, large cohorts are needed to detect a phenotypic effect on atherosclerotic diseases.

EL Interacts With Inflammation

It is well known that inflammation plays an important role in atherogenesis; thus, several groups have studied the effects of inflammatory cytokines on EL expression in endothelial cells in vitro. Hirata et al³⁷ examined the regulation of EL expression in cultured human umbilical vein and coronary artery endothelial cells. They found that EL mRNA levels were up-regulated in both cell types by inflammatory cytokines implicated in vascular disease etiology, including tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β). In a subsequent investigation, Jin et al³⁸ confirmed these results and extended them by demonstrating that secretion and activity of EL protein were also upregulated in endothelial cells by TNF-α and IL-1β in a dose- and time-dependent pattern. In vitro data have suggested that EL might also promote monocyte adhesion to the vascular endothelium through an interaction with heparin sulfate proteoglycans.³⁹ A relatively strong association was observed between plasma concentrations of proinflammatory cytokines, such as C-reactive protein (CRP) and IL-6, and postheparin plasma EL concentrations in overweight sedentary men.⁴⁰ Badellino et al⁹ found some inflammatory markers, such as soluble TNF receptor II, soluble intercellular adhesion molecule 1, and leptin, besides CRP and IL-6, were correlated directly with plasma EL concentrations, and experimental endotoxemia significantly increased plasma EL concentrations, which proved that EL is upregulated by inflammation in humans.⁹ As we know, low-grade systemic inflammation occurs in obesity, insulin resistance, type 2 diabetes mellitus, hypertension, hyperlipidemia, and CHD.⁴¹ Increased plasma EL concentration has been reported in subjects with features of the metabolic syndrome and with obesity.⁸ Recently, Shiu et al⁴² found serum EL concentration increased in type 2 diabetic patients and was associated with the degree of subclinical inflammation, and exogenous insulin therapy lowered serum EL concentration.⁴²

On the other hand, the expression of cytokines was also affected by EL. Qiu et al⁴³ showed inhibiting EL expression within THP-1 macrophages decreased induction of pro-inflammatory genes, growth factors, and anti-apoptotic genes and altered lipid composition with reduced percentages of cholesterol, triglycerides, and lysophosphatidylcholine. The expression of EL messenger RNA (mRNA) was associated positively with CD68 mRNA in advanced human atherosclerotic lesions, and it increased markedly when monocytes differentiated into macrophages.⁴⁴ These data provided potential pathologic roles of EL in human macrophages within atherosclerotic lesions. Qiu et al⁴⁵ also found that atorvastatin and simvastatin decreased EL expression dose dependently, as well as Rho, liver X receptor-α, and nuclear factor-kB activation in THP-1 macrophages, and it was independent of reductase inhibition, which likely contributed to the ability of statins treatment to mitigate lipid accumulation in macrophages.

Recent studies have also identified anti-inflammatory properties of HDL that might cause lipase activity. The HDL-mediated repression of leukocyte adhesion to endothelial cells is decreased markedly after inhibition of lipase activity. Ahmed et al⁴⁶ showed EL limited the expression of vascular cell adhesion molecule-1 (VCAM-1) by activating peroxisome proliferator-activated receptor-α via HDL hydrolysis. They found that EL overexpression significantly decreased TNF-α–induced VCAM-1 expression and promoter activity in a manner dependent on HDL concentration and intact EL activity. The noncatalytic or bridging function of EL in macrophages mediates the interaction between lipoproteins and cell surfaces. Endogenously produced EL in transfected Chinese hamster ovary cells enhances the binding and cellular processing of plasma lipoproteins.⁴⁷ Moreover, EL can play a role as adhesion molecules through the interaction with heparin sulfate proteoglycans to facilitate the binding of monocytes onto the surface of endothelial cells.³⁸

All these data indicated the important role of EL in inflammatory conditions and its potential role in the modulation of lipoprotein metabolism during inflammatory conditions, including AS.

The Impacts of EL on Atherogenesis

Several studies showed EL expression associated with macrophages within human atherosclerotic lesions.^{[11] and [38]} Bartels et al⁴⁴ detected the expression of EL mRNA and protein in human atherosclerotic lesions in carotid endarterectomy specimens from patients with symptomatic cerebrovascular disease. They found that EL mRNA and/or protein were detected in areas between the necrotic core and the fibrotic cap. EL mRNA expression significantly increased when monocytes differentiated into macrophages and decreased when macrophages converted into foam cells. Azumi et al¹¹ demonstrated that EL was expressed in endothelial and medial smooth muscle cells in nonatherosclerotic coronary arteries, as well as in infiltrating cells and endothelial and smooth muscle cells, within atheromatous plaques in human coronary arterial specimens. Furthermore, EL immunoreactivity was also detected in neovasculature within atheromatous plaques in atherosclerotic coronary arteries. In a cohort of healthy subjects with a family history of premature CHD, increased human plasma EL concentrations were significantly associated with the metabolic syndrome and subclinical CHD.⁸ The results previously reported suggested EL was associated with AS development, but the precise mechanisms need additional investigation.

Qiu et al⁴⁸ used loss-of-function and gain-of-function approaches to demonstrate that THP-1 macrophage-derived EL promoted the binding and uptake of native LDL and oxidized-LDL, and the LDL receptor was partly responsible for the EL-enhanced uptake of native LDL. This group⁴⁹ then documented EL promoted apoAI-mediated cholesterol efflux through mechanisms related to its catalytic and noncatalytic functions. The impairment of lipid efflux in macrophages might aggravate lipid accumulation and accelerate the progression of AS.

Given the strong association between plasma HDL-C levels and CHD risk in humans, people would expect that alteration in plasma HDL levels by EL in experimental animals might affect the progression of AS. Recent studies dedicated to this hypothesis have suggested 2 different opinions. Ishida et al¹⁸ reported that the inactivation of EL retarded AS and was associated with an approximately 70% decrease in atherosclerotic disease area in apoE knockout mice compared with controls despite the presence of a proatherogenic lipid profile. They ascribed the protective effect mainly to the absence of EL proatherogenic functions during monocyte recruitment and cholesterol uptake in the vascular wall, as they found a major reduction in the macrophage content in the atherosclerotic lesion in apoE^-/- mice. Otherwise, Ko et al⁵⁰ found that although EL modulated the lipoprotein profile in mice, there was no effect of EL inactivation on AS development in 2 hyperlipidemic AS- prone mice models.

Both groups used apoE^−/− mice in the C57BL/6 background, and a targeting strategy led to the absence of functional EL expression, which demonstrated similar effects of EL inactivation on lipoprotein profiles and HDL-C levels. However, the reasons leading to the difference in AS development are unclear but might be related to environmental factors or the stage of development of the atherosclerotic lesions at the time of measurement and need further investigations.

Overall, whether EL participates in the development of AS, independent of its impacts on plasma lipids levels, deserves subsequent investigation.

Conclusion

The EL level in plasma has been associated with HDL-C concentration and inflammation, which are important in the initiation and development of AS. Although few data are available from human studies, it seems that plasma EL levels in individuals with AS might be higher than that in healthy individuals. Therefore, the subsequent characterization of EL might eventually help to produce more effective, targeted intervention strategies to elevate HDL levels and, thereby, to reduce risk of cardiovascular disease.

Ji Huang and Hai-Yan Qian contributed equally to this work.