Background
Plasma high-density lipoprotein (HDL) is a well-established cardiovascular protective factor with multiple functions, including mediating reverse cholesterol transport, and exerting anti-oxidative, anti-inflammatory, anti-thrombotic and vasodilatory activities [
1]. Epidemiological and clinical studies have shown that HDL cholesterol (HDL-C) level is inversely correlated with coronary heart disease risk [
2]. However, nearly half of cardiovascular clinical events occurred in subjects with normal or even high levels of HDL-C [
3]; and clinical trials with drugs that raised HDL-C levels by cholesteryl ester transfer protein (CETP) inhibitors, had failed to reduce the risk of cardiovascular events [
4]. These controversial findings suggest that HDL-associated functional factors, instead the levels of HDL-C, are important for the antiatherogenic properties. Indeed, it has been known that HDL particles have substantial compositional heterogeneity in its lipids, apolipoproteins, lipid transfer proteins, enzymes and numerous other proteins under pathophysiological conditions because of their complex metabolic function. The evidence from transgenic mice indicates that protein composition determines the antiatherogenic properties of HDL. Alterations in protein composition involved in HDL metabolism could promote atherosclerosis, even if plasma HDL-C was normal or elevated [
5]. Recent studies come up with a new concept of “dysfunctional HDL”, which suggests that antiatherogenic HDL can be converted to proatherogenic HDL under dyslipidemic and inflammatory states in metabolic diseases, and thus may accelerate atherosclerosis [
6]. Therefore, the assessment of dysfunctional HDL by measuring the protein composition, and the identification of biomarkers of dysfunctional HDL in different experimental models, will provide important information for the evaluation of cardiovascular risk in patients and for the development of new antiatherogenic therapies [
7].
Scavenger receptor class B type I (SR-BI), a well-defined HDL receptor, is a prominent regulator of HDL metabolism, which controls the plasma HDL-C level. It mediates the selective uptake of cholesteryl esters from HDL by the liver and steroidogenic organs, and facilitates the efflux of cholesterol from peripheral tissues to HDL [
8]. It has been reported that several mutations in human SR-BI gene, which are associated with low SR-BI protein levels, caused elevated plasma HDL-C levels, a reduction in cholesterol efflux from macrophages, altered platelet function, and decreased adrenal steroidogenesis [
9‐
11]. The SR-BI deficient (SR-BI
−/−) mice showed a nearly twofold increase in plasma total cholesterol (TC) and HDL-C levels, accompanied with impaired anti-oxidative function, and lipid deposition in the aorta, contributing to the enhanced susceptibility to atherosclerosis [
12‐
14]. However, the protein composition of defective HDL in SR-BI
−/− mice has not been well studied.
Therefore, in this study we used shotgun proteomics, a powerful tandem mass spectrometry (MS/MS) approach following the protein separation by liquid chromatography (LC) [
15], to compare the protein composition of HDL particles isolated from SR-BI
−/− mice and wild type (WT) mice. Meanwhile, HDL-mediating cholesterol flux in macrophages, anti-oxidative and anti-inflammatory function of HDL were detected to verify the functional status of HDL. Considering that probucol as a cholesterol-lowering and antioxidant drug, could alter the abnormal HDL-C level in SR-BI
−/− mice [
16], probucol treatment was used to investigate whether the alteration of protein markers and HDL function could be affected in SR-BI
−/− mice. The goal of this study is to quantify the protein characteristics of SR-BI
−/− HDL, providing the valuable protein markers for diagnosis and therapeutics of dysfunctional HDL-related metabolic diseases.
Discussion
In current study, we demonstrate that the defective HDL functions in SR-BI−/− mice are associated with the altered protein composition by shotgun proteomic profiling and biochemical analyses, despite the higher HDL-C level in SR-BI−/− mice. We observed that a significantly lower abundance of lipid metabolism and antioxidant related proteins, such as apolipoproteins (apoAI, apoAII and apoCI) and HDL-associated PON1 in SR-BI−/− mice, whereas levels of proteins involved in inflammatory and immune response, proteinase inhibition and blood coagulation, including SAA, apoAIV, complement C3 and A1AT, were increased. These protein composition alterations result in impairment of HDL function, including impaired function in maintaining cholesterol homeostasis in macrophages, and the anti-oxidative and anti-inflammatory capacity. Furthermore, the lipid-lowering and antioxidant drug probucol can improve the HDL functions in SR-BI−/− mice by reversing the contents of some critical proteins, including increasing plasma apoAI and PON1 levels, and decreasing plasma SAA, apoAIV and A1AT levels.
There is accumulating evidence that alterations in HDL protein quality and composition result in HDL function impairment [
24,
25]. In this study, we employed shotgun proteomics methods to identify the quantitative and qualitative alterations in HDL-associated proteins that may render SR-BI
−/− HDL dysfunctional. The proteomic profiling of HDL particles shown that some of essential protective apolipoproteins and enzymes were decreased, and the oxidative and inflammatory related proteins identified in SR-BI
−/− HDL appeared or were increased, compared to SR-BI
+/+ HDL. We observed that relative levels of HDL-associated apolipoproteins involved in lipid metabolism, such as apoAI, apoAII, and apoCI, were significantly decreased in SR-BI
−/− HDL. As the key apolipoprotein on HDL, apoAI is a well-known anti-atherogenic molecule by mediating reverse cholesterol transport (RCT) and exerting anti-inflammatory effects [
26‐
28]. ApoAII, as the second most abundant protein on HDL, also acts as a primary acceptor and efficiently removes cholesterol from macrophages in vivo [
29]. ApoCI, as a component of HDL and VLDL, was considered to play a beneficial role in lipid metabolism, but recently study reported that apoCI could increase the LPS-induced inflammation in macrophages in vitro and in apoE
−/− mice [
30]. In the proteomic profiling, we found that the relative abundance of apoCI in SR-BI
−/− HDL was decreased. Therefore, the role of apoCI in SR-BI
−/− mice remains to be further studied. An interesting finding is that the percentage of apoAI is decreased in HDL by spectral counting, while the plasma apoAI content is not reduced in SR-BI
−/− mice. Meanwhile, we also identified by 2D gel electrophoresis that the SR-BI
−/− mice had an accumulation of large HDL migrating between pre-β and α, with a higher level of apoAI. These suggested that apoAI in abnormal SR-BI
−/− HDL might have underwent dramatic modification, and became loosely associated with the larger HDL, thus, part of apoAI was likely to lost from isolation of HDL by precipitation and ultracentrifugation. Considering that SR-BI
−/− mice had comparable plasma apoAI concentration as WT mice (108 versus 100 mg/dL), as reported by a previous study [
12], it is conceivable that the relative level of apoAI in SR-BI
−/− HDL was significantly reduced due to modification of apoAI and elevation of other proteins such as inflammatory proteins. Indeed, our results further confirmed that SR-BI
−/− HDL contributed to the accumulation of cholesterol in macrophages, indicating that SR-BI
−/− HDL not only stimulated FC influx towards the cells with gradient-regulated transport, but also impaired the ability of mediating cholesterol efflux because of modification of apoAI and other compositional alterations.
Besides involved in lipid metabolism, apoAI also exerts the anti-oxidative and anti-inflammatory functions of HDL. A recent study has demonstrated that impaired antioxidant HDL function is independently associated with the development of premature acute myocardial infarction [
31]. In addition to apoAI, several important enzymes such as PON1, PON3 and PAFAH in HDL play anti-oxidative role. As an essential HDL-associated enzyme, PON1 contributes substantially to the anti-inflammatory properties of HDL through preventing LDL oxidation by hydrolyzing oxidized lipids. Our previous study has reported that lower plasma PON1 activity is associated with increased atherosclerotic lesions in apoE
−/− mice and in coronary artery disease (CAD) patients [
21]. This present study showed that the abundance and activities of PON1 in HDL and plasma of SR-BI
−/− mice were markedly reduced, indicating that anti-oxidative activity of SR-BI
−/− HDL was hampered. Therefore, oxidative stress was increased in SR-BI
−/− mice, which is evidenced by increased plasma MPO activity and HDL-related oxidative property, consistent with a previous study [
14].
We also found some other proteins which may contribute to the pro-inflammatory properties of SR-BI
−/− HDL, such as acute-phase protein SAA, apoAIV, several cell adhesion molecules, complement C3 and immune response proteins. SAA is demonstrated as a clinically useful marker for inflammation and is strongly associated with increased risk of cardiovascular events [
32]. A previous report showed that SAA could result in altered metabolic properties of HDL by displacing apoAI, rendering HDL proatherogenic [
33]. Human apoAIV has been reported with multiple functions related to lipid metabolism through activation of LCAT and modulation of lipoprotein lipase, and anti-oxidative activity [
34]; however, mouse apoAIV, as a positive acute phase protein, was increased during inflammation [
35]. Our study identified that SR-BI
−/− mice had higher HDL-associated apoAIV levels, supporting that SR-BI
−/− HDL in the inflammatory state. The previous study showed that SR-BI
−/− mice have a significant lower LCAT activity [
36], Our results suggested that decreased activity of LCAT maybe related with apoAI modification and higher apoAIV levels in SR-BI
−/− mice, although the relative LCAT level of SR-BI
−/− HDL was increased in proteomic profile. In addition, proteins involved in cell adhesion like intercellular adhesion molecule 1 (ICAM1), integrin β1, and vitronectin were found in SR-BI
−/− HDL, but not in WT HDL. Several proteins with serine proteinase inhibitor domains were detected with increased levels in SR-BI
−/− HDL, such as A1AT, antithrombin-III, inter-α-trypsin inhibitor, angiotensinogen, α-2-macroglobulin and murinoglobulin-1, compared to SR-BI
+/+ HDL. A recent study of HDL proteomics indicated that protease inhibition ability modulates the anti-inflammatory properties of HDL [
37]. Our proteomic data supported that accumulation of the damaged proteins in SR-BI
−/− HDL could initiate inflammation, coagulation, and complement activation responses. Furthermore, we observed that HDL isolated from SR-BI
−/− mice exhibited less capability of suppressing inflammatory cytokine (MCP-1, TNF-α) expression induced by ox-LDL in macrophages. Taken together, the present study demonstrated that SR-BI deficiency results in a significant alteration in the HDL proteome, which contributes to impair the critical HDL functions, may explain increased risk of cardiovascular disease despite higher HDL-C level in SR-BI
−/− mice.
Certainly, mouse does not express CETP, which transfers cholesteryl ester from HDL to VLDL/LDL for further uptake by the liver, it means that HDL metabolism and its related proteins of mice are not exactly the same as those of humans. For example, the changes of apoAIV and A1AT in defective HDL in SR-BI
−/− mice were different from those in patients with HDL-associated metabolic diseases [
35,
38]. However, the previous studies showed that dysfunctional HDL with increased plasma HDL-C levels in SR-BI
−/− mice mirrored that in those patients who harbor SR-BI mutations [
9‐
11]. Accordingly, our observation made in SR-BI
−/− mice studies also supported that the identification of significant proteins in dysfunctional HDL using SR-BI
−/− mice as a model, might provide a powerful clue for exploring the biomarkers in clinical diagnosis and therapeutics of dysfunctional HDL-related metabolic diseases. SR-BI
−/− mice could be considered as a good preclinical animal model for studying the structure–function relationship of HDL.
Probucol is lipid-lowering and antioxidant drug that has been used in clinic for the treatment of CAD. It could promote RCT, normalize the FC level of HDL [
39,
40], and reverse female infertility in SR-BI
−/− mice [
16]. Therefore, we explored the effects of probucol on several critical HDL protein components and HDL dysfunction in SR-BI
−/− mice. The current study indicated that accompanied with decreased plasma TC, FC and HDL-C levels, the alterations of some representative proteins of SR-BI
−/− HDL could be reversible, and anti-oxidative function was also improved by probucol treatment. These are in agreement with the other studies, that small, protein-enriched, cholesterol-depleted HDL particles are more effective in mediating cholesterol efflux, and with more potent antioxidant, anti-inflammatory, and anti-apoptotic capacity [
41,
42]. The derived protein biomarkers in dysfunctional HDL may eventually help to exploit diagnostics targets and therapeutic strategies for HDL-related metabolic diseases.