Proteomic analysis of plasma-purified VLDL, LDL, and HDL fractions from atherosclerotic patients undergoing carotid endarterectomy: identification of serum amyloid A as a potential marker.

Apolipoproteins are very heterogeneous protein family, implicated in plasma lipoprotein structural stabilization, lipid metabolism, inflammation, or immunity. Obtaining detailed information on apolipoprotein composition and structure may contribute to elucidating lipoprotein roles in atherogenesis and to developing new therapeutic strategies for the treatment of lipoprotein-associated disorders. This study aimed at developing a comprehensive method for characterizing the apolipoprotein component of plasma VLDL, LDL, and HDL fractions from patients undergoing carotid endarterectomy, by means of two-dimensional electrophoresis (2-DE) coupled with Mass Spectrometry analysis, useful for identifying potential markers of plaque presence and vulnerability. The adopted method allowed obtaining reproducible 2-DE maps of exchangeable apolipoproteins from VLDL, LDL, and HDL. Twenty-three protein isoforms were identified by peptide mass fingerprinting analysis. Differential proteomic analysis allowed for identifying increased levels of acute-phase serum amyloid A protein (AP SAA) in all lipoprotein fractions, especially in LDL from atherosclerotic patients. Results have been confirmed by western blotting analysis on each lipoprotein fraction using apo AI levels for data normalization. The higher levels of AP SAA found in patients suggest a role of LDL as AP SAA carrier into the subendothelial space of artery wall, where AP SAA accumulates and may exert noxious effects.


Introduction
Cardiovascular diseases are the leading cause of death and illness in developed countries, with atherosclerosis being the most important contributor. Atherosclerosis is a chronic inflammatory condition that could turn into an acute clinical event due to plaque rupture and thrombosis [1]. Indeed, vascular inflammation not only plays a major role in the development of atherosclerosis but also contributes to the acute onset of thrombotic complications [2]. The selective retention of circulating apolipoprotein B100 containing lipoproteins in the subendothelial space, by means of specific interactions with artery wall proteoglycans, is currently thought to be the leading event in atherogenesis [3,4].
Lipoproteins are supramolecular complexes that deliver insoluble lipids from the tissues where they are synthesized to those that metabolize or store them. They consist of hydrophobic molecules (core), particularly triacylglycerol and cholesteryl esters, stabilized by a coat of amphipathic compounds, namely, phospholipids, unesterified cholesterol, and proteins, with the latter referred to as apolipoproteins (apo) [5]. Apolipoproteins are a very heterogeneous protein family implicated in lipoprotein stabilization, lipid metabolism, inflammation, or immunity [6]. Except for apo B100, the main structural apolipoprotein of VLDL and LDL, they may be exchanged among circulating lipoprotein classes during their physiological metabolism or in pathological conditions. For a long time now, lipoproteins have attracted a great deal of interest because of their implication in atherogenesis. Although it is well known that high LDLcholesterol and low HDL-cholesterol levels are associated with increased risk for the development of cardiovascular disease, clinical studies suggest that levels of apo B100 and apo AI may be better predictors [7]. Since the protein component of these particles is largely responsible for carrying out their various functions, detailed information about the apolipoprotein composition and structure may contribute to revealing their role in atherogenesis and to developing new therapeutic strategies for the treatment of lipoproteinassociated disorders. Applying proteomics to the study of lipoproteins, including gel-based or gel-free technologies, may significantly contribute to the achievement of this goal. Indeed, recent proteomic studies have revealed that lipoproteins carry an array of proteins previously unexpected [8].
As far as we know, about thirty proteomic studies on VLDL, LDL, and HDL have been published up to now, while no proteomic studies on chylomicrons are reported in the literature [8]. Recently, the proteome of lipoprotein (a) has been investigated too [9]. The majority of these studies focused on characterizing the apolipoprotein cargo of the different lipoprotein fractions in healthy subjects. Conversely, only few studies dealt with differential apolipoprotein expression in relation to cardiovascular diseases. In this respect, HDL fraction has been the most studied, in association with coronary artery disease (CAD) [10][11][12][13], high cardiovascular risk and acute myocardial infarction [14], hyperlipidemia [15], or low levels of HDL cholesterol [16,17]. One study on LDL fraction from atherosclerotic patients with the metabolic syndrome and diabetes [18] and one on VLDL fraction from hyperlipidemic subjects [19] were conducted. Several apolipoprotein isoforms involved in lipid metabolism, inflammation or immunity were detected as differentially expressed. Regrettably, these studies merely focused on one lipoprotein class at a time and were based on a low sample size.
Recently, by applying proteomics to the study of carotid plaque vulnerability, we identified a panel of proteins differentially expressed/oxidized in stable and unstable lesions [20,21]. In the present study we set up a method for characterizing the exchangeable apolipoprotein component of plasma VLDL, LDL, and HDL fractions from patients undergoing carotid endarterectomy, useful for identifying differentially expressed proteins with respect to healthy normolipidemic subjects. By means of two-dimensional electrophoresis (2-DE) and MALDI-TOF MS analysis we identified acute-phase serum amyloid A protein (AP SAA) as overexpressed in the three lipoprotein classes from patients. Results validation was performed by western blotting analysis. In our knowledge, this is the first study providing information on exchangeable apolipoprotein profiles in VLDL, LDL, and HDL fractions from atherosclerotic patients undergoing carotid endarterectomy.

Sample Collection.
Analyses were performed on 4 pooled plasma samples from 57 healthy normolipidemic volunteers and 4 pools obtained from 79 patients undergoing carotid endarterectomy. Carotid atherosclerosis was assessed by ultrasonography using a Mylab 70 X vision echocolor Doppler equipped with an LA332 AppleProbe 11-3 MHz (Esaote). All patients selected for surgery had either a high-grade stenosis (>70%) or an ulcerated lesion of a medium grade based on echo-Doppler analysis. All patients with hypertension, dyslipidemia, and/or diabetes were under pharmacological treatment. The main clinical parameters of patients and controls are summarized in Table 1. Informed consent was obtained before enrolment. Institutional Review Board approval was obtained. The study was conducted in accordance with the ethical principles of the current Declaration of Helsinki. Fasting blood samples were collected into Vacutainer tubes containing EDTA and immediately centrifuged at 2000 ×g for 10 minutes at 4 ∘ C. Plasma was added with 100 M APMSF (4-amidinophenylmethane-sulfonylfluoride), 2 g/ mL Kallikrein Inactivator (cyclohexylacetyl-Phe-Arg-Ser-Val-Gln amide), and 50 M leupeptin (Acetyl-Leu-Leu-Arg-Hydrochloride) and stored at −80 ∘ C until analysis. Before lipoprotein purification, equal amounts of plasma samples were randomly combined to yield 4 homogeneous pooled VLDL, LDL, and HDL fractions from both patients and controls.

Lipoproteins
Purification. Lipoproteins were isolated by isopycnic ultracentrifugation according to the methods described by Himber et al. [22] and McDowell et al. [23] with slight modifications. Briefly, 3.9 mL of pooled plasma samples was added with NaBr to = 1.3 g/mL (472.2 mg NaBr/mL plasma) in centrifuge tubes (Beckman Coulter, Thin Wall Ultra-Clear, 14 mL, 14 × 95 mm), overlaid with 8.1 mL of a 0.6% NaCl solution ( = 1.006 g/mL), and centrifuged at 285,000 ×g for 48 h at 4 ∘ C in an Optima L90 series ultracentrifuge equipped with a SW40 Ti rotor (Beckman Coulter). Afterwards, VLDL, LDL, and HDL fractions were collected and further purified by a second centrifugation step, performed at 285,000 ×g for 24 h in saline solutions at density 1.006, 1.063, and 1.21 g/mL, respectively. Lipoprotein fractions were then collected, desalted (final salt concentration < 5 mM), and concentrated using Amicon Ultra-0.5 mL centrifugal filter units (10 KDa MWCO, Millipore). Protein concentration was determined using DC Protein Assay Kit (Bio-Rad), according to the manufacturer's instructions, using bovine serum albumin as standard. Aliquots of 500 g (as protein) of LDL and 300 g (as protein) of VLDL were delipidated by adding ice-cold tri-nbutylphosphate : acetone : methanol (1 : 12 : 1) as reported by Mastro and Hall [24]. Delipidated fractions were resolubilized with repeated boiling and sonication passages in 10 mM Tris buffer containing 4% CHAPS (w/v) and 1% DTT (w/v). After cooling to room temperature, samples were diluted with a solution containing 8 M urea, 4% CHAPS, 1% DTT, and 0.4% carrier ampholytes and subjected to further sonication passages followed by 2-DE. 50 g (as protein) of native HDL fraction were solubilized for 2-DE analysis as reported above for LDL and VLDL delipidated fractions.

2-DE Analysis.
2-DE was performed as previously described [20]. Briefly, samples were applied to 70 mm IPG strips (pH 4-7, Bio-Rad), by overnight rehydration loading at 20 ∘ C, and subsequently focused at 50 A/IPG strip for 22 kVh at 20 ∘ C. After focusing, proteins were in-gel reduced by incubating IPG strips with a 50 mM Tris buffer containing 6 M urea, 30% glycerol (v/v), 3% SDS (w/v) and 1% DTT (w/v), followed by in-gel alkylation, using the same solution containing 2.5% iodoacetamide (w/v) in place of DTT. Each step was performed keeping strips under continuous shaking for 15 minutes. Then, IPG strips were sealed, with 0.5% low melting point agarose (w/v) in SDS running buffer, at the top of second dimension gels (8 cm × 7 cm × 0.1 cm). SDS-PAGE was carried out on 15% T/3% C polyacrylamide gels at 50 V for 15 minutes and subsequently at 150 V for about 90 minutes. Then, gels were fixed in 30% ethanol (v/v), 2% phosphoric acid (v/v) for 1 h, washed twice in 2% phosphoric acid (v/v) solution for 10 minutes, equilibrated in 18% ethanol (v/v) and 2% phosphoric acid (v/v), and 15% ammonium sulfate (w/v) solution for 30 minutes, and stained in the same solution added with 0.02% Coomassie Brilliant Blue G-250 (w/v) for 48 h. Gel images were acquired by using GS-800 calibrated densitometer (Bio-Rad) at 36.3 m resolution. Image analyses were performed using PD-Quest 2-D analysis software V 8.0.1 (Bio-Rad). Each spot was assigned a relative value corresponding to the single spot volume compared to the volume of all spots in the gel, following background subtraction and normalization between gels. In particular, in the adopted normalization method, the raw quantity of each spot in a member gel was divided by the total quantity of all the spots in that gel that have been included in the Master.

In-Gel Digestion and MALDI-TOF MS Analysis.
Spots of interest were excised with sterile pipette tips, destained with a solution containing 2.5 mM NH 4 HCO 3 and 50% acetonitrile (v/v), dehydrated with 100% acetonitrile, and dried at room temperature before proteolytic treatment. Tryptic digestion was performed by incubating dried spots in 5 mM NH 4 HCO 3 buffer containing 10 ng/ L trypsin overnight at 37 ∘ C. The resulting peptides were mixed with an equal volume of -cyano-4-hydroxycinnamic acid saturated solution (40% acetonitrile (v/v) and 0.1% trifluoroacetic acid (v/v)) and applied as a microcrystalline thin film onto a stainless steel 96-spot MALDI target. Mass analyses were performed using a MALDI micro-MX-mass-spectrometer (Waters), according to the tuning procedures suggested by the manufacturer. Peak lists were generated using Protein Lynx Global Server v.2.2.5 (Waters) and searched against the Swiss-Prot human database (version 57.4) using Mascot (http://www.matrixscience.com). Research parameters included taxa (Homo sapiens), trypsin digest, monoisotopic peptide masses, iodoacetamide modifications, one missed cleavage by trypsin, and a mass deviation of 50 ppm. Only protein identifications with significant Mascot scores ( < 0.05) were taken into consideration.

Western Blotting Analysis.
Western blotting analyses on proteins resolved by SDS-PAGE were performed as previously described [20]. Briefly, resolved proteins were electroblotted onto Hybond-P PVDF membranes (GE Healthcare) at 250 mA for 1.5 hours. Afterwards, membranes were incubated with blocking solution (PBS, 0.1% Tween-20, 3% nonfat dry milk) for 1 h at room temperature followed by overnight incubation at 4 ∘ C with a monoclonal anti-SAA antibody (Abcam, Ab81483) diluted 1 : 1000 or a polyclonal anti-apo AI antibody (Millipore, AB740) diluted 1 : 8000 with blocking solution. Then, after 30 minutes washing (PBS, 0.1% Tween-20), membranes were incubated with HRPconjugated secondary antibody solution for 1 h at room temperature. Following further membrane washing, proteins were revealed by enhanced chemiluminescence using ChemiDoc XRS System (Bio-Rad). Densitometric analysis was performed using Quantity One 4.6.3 software (Bio-Rad).

Statistical Analysis.
Student's -test for unpaired samples was performed to compare exchangeable apolipoproteins expression in each purified lipoprotein fraction among atherosclerotic patients and healthy controls, using the software package Sigma Stat 3 (Systat Software). Significance was set at < 0.05.  Exchangeable apolipoproteins from purified lipoprotein fractions were resolved by isoelectric focusing on pH 4-7 IPG strips followed by SDS-PAGE on 15% T and 3% C gels (a-c). Plasma proteins were resolved by isoelectric focusing on pH 3-10 IPG strip followed by SDS-PAGE on 10% T and 3% C gel (d). Protein spots identified by MALDI-TOF MS analysis are marked.

Mapping of Exchangeable
According to echo-Doppler analysis, patients selected for surgery had either a high-grade carotid stenosis (>70%) or an ulcerated lesion. Since apo B100 is the most abundant apolipoprotein of both LDL and VLDL, representing up to 95% of their protein moiety, large amounts of LDL and VLDL were processed in order to allow the detection and identification also of less abundant apolipoproteins. Furthermore, due to the high lipid/protein ratio in both lipoprotein classes, a delipidation step was mandatory.
Reproducible 2-DE maps of exchangeable apolipoproteins from VLDL, LDL, and HDL fractions were obtained (Figure 1). The adopted fractionation procedure, which included an isopycnic gradient ultracentrifugation followed by an additional washing step, allowed high purification yields of all lipoprotein fractions. As a matter of fact, no albumin or additional plasma proteins were detectable in 2-DE gels after Coomassie Brilliant Blue G-250 staining. Moreover, western blotting analysis of the HDL fraction did not reveal any apolipoprotein B-100 contamination (data not shown).
Almost all the resolved protein spots were identified by peptide mass fingerprinting (Table 2). In more detail, we identified 2 isoforms of apo J, 2 isoforms of apo AIV, 5 isoforms of apo E, 6 isoforms of apo AI, 3 isoforms of apo D, 2 isoforms of acute phase serum amyloid A protein, 1 isoform of apo CII and 2 isoforms of apo CIII, in both VLDL and LDL fractions. Due to its high molecular mass (about 500 kDa) and hydrophobicity, apo B100 was not detectable in any VLDL or LDL 2-DE maps. Regarding to HDL, 3 isoforms of apo E, 6 of apo AI, 3 of apo D, 2 of AP SAA, 1 of apo CII, and 2 of apo CIII were identified.

Differential Apolipoproteins Expression Analysis.
Differential expression analysis was performed on VLDL, LDL, and HDL fractions from 4 pooled plasma samples of 79 patients having severe atherosclerosis and 4 pooled plasma samples of 57 healthy volunteers revealing no differences in the relative abundance of all identified apolipoproteins (Figure 2) with the exception of AP SAA that was more abundant (2.3-, 14.0-, and 1.5-fold in VLDL, LDL, and HDL fractions, resp.) in patients ( = 0.001, = 0.003, and = 0.045, for VLDL, LDL, and HDL fractions, resp.) (Figure 3). Differential AP SAA expression was confirmed by western blotting analysis   Because of its high percentage in respect to the other apolipoproteins, apo AI was omitted in the panels corresponding to LDL (b) and HDL (c) fractions, being 72.9 ± 9.8% and 91.5 ± 3.7%, respectively. Relative amount of AP SAA is reported in Figure 3. Data are expressed as mean ± SD. on each lipoprotein fraction using apo AI levels for data normalization ( = 0.03, = 0.03, and = 0.05, for VLDL, LDL, and HDL fractions, resp.) (Figure 4). To further corroborate these preliminary observations, the same western blotting assay was performed on a randomly chosen sample from each group (Figure 4).

Discussion
By means of two-dimensional electrophoresis (2-DE) coupled with Mass Spectrometry (MS) analysis on plasmapurified VLDL, LDL, and HDL fractions we were able to identify acute phase serum amyloid A protein (AP SAA) as a potential marker of advanced carotid atherosclerosis.
The protein moiety of the different lipoprotein species is largely responsible for their functions, being involved in several metabolic events, such as recognition by specific receptors on cell surfaces or regulation of the activity of various enzymes [6]. Obtaining detailed information about apolipoprotein composition may contribute to revealing their role in atherogenesis. To date, only few studies have dealt with lipoproteomics in relation to atherosclerosis or atherosclerosis-related diseases, wih most of them being performed on HDL from coronary artery disease (CAD) patients [10][11][12][13]. This is the first report focusing on differential apolipoprotein expression in atherosclerotic patients undergoing carotid endarterectomy compared to healthy controls. As the adopted method is very labour intensive, we decided to set it up on randomly combined pooled plasma samples. In our opinion, this is the major limitation of the study, since each data reported represents an average value from 15 to 20 samples and a measure of how much variation or dispersion from the average exists is missing. On the other hand, this approach allowed for comparing levels of the exchangeable apolipoproteins in each lipoprotein fraction from a noteworthy number of patients and controls (79 patients and 57 controls). Indeed, the preliminary results obtained from differential protein expression analysis may represent a useful starting point for further, more specific analyses that could be carried out on each sample separately (e.g., ELISA on plasma samples and plaque extracts, and immunohistochemistry on endarterectomy specimens).
The apolipoprotein profile of VLDL, LDL, and HDL fractions, obtained from patients with advanced atherosclerosis was analysed, identifying 23 different protein isoforms by peptide mass fingerprinting. Moreover, we compared the obtained profiles with those from healthy volunteers. The adopted purification procedure provided highly purified lipoprotein fractions as confirmed by the absence of  8). Following data normalization for apo AI levels, AP SAA was confirmed as differentially expressed in all purified fractions from patients compared to controls ( = 0.03, = 0.03, and = 0.05, for VLDL, LDL, and HDL fractions, resp.). Results obtained on a randomly chosen sample from controls group (lane 9) and from patients group (lane 10) are shown. 10 g and 1 g of proteins were loaded for AP SAA and apo AI immunodetection, respectively. 9 plasma proteins as well as of apo B-100 in HDL (assessed by western blotting). Furthermore, due to the high number of resolved protein isoforms and the low limit of detection, 2-DE has proven to have high resolution and sensitivity.
Image analysis allowed for identifying AP SAA as differentially expressed in atherosclerotic patients. Its levels were higher in all lipoprotein fractions, increasing up to fourteenfold in atherosclerotic patients LDL fraction. Such evidence was confirmed by western blotting analysis on each lipoprotein purified fraction.
Serum amyloid A protein belongs to a family of acutephase proteins, synthesized primarily in the liver, that circulate mainly associated to HDL. During the acute-phase reaction, plasma AP SAA concentration may increase up to 1000-fold due to an upregulation of its synthesis by inflammatory cytokines or peptide hormone signals [25]. Besides some evidence of an antiatherogenic role in promoting cholesterol efflux from cells [26][27][28][29][30], a plethora of studies have suggested proatherogenic effects for SAA, such as promotion of monocytes and neutrophils chemotaxis [31][32][33], macrophage foam cells formation [34], stimulation of proinflammatory cytokines secretion by monocytesmacrophages and lymphocytes [33,35,36], and induction of endothelial dysfunction [33,37,38]. Furthermore, SAA may potentiate prothrombotic events by inducing tissue factor expression [39]. A role for SAA in lipoprotein subendothelial retention has been postulated, since it is known that it could stimulate the synthesis of vascular proteoglycans with an increased LDL-binding affinity [40] and a proteoglycan binding domain on its carboxy-terminal region has been identified [41]. Furthermore, a study on in vivo murine model demonstrated that SAA deposited in atherosclerotic areas at all stages of lesion development and highly correlated with lesion area, apo AI area, apoB-100 area, and perlecan area [42]. Nevertheless, both protein localization [43] and mRNA expression [44] in human atherosclerotic plaques were demonstrated. SAA was also shown to increase in patients with atherosclerosis of the coronary and peripheral arteries [45][46][47][48][49][50]. Interestingly, Ogasawara et al. evidenced that LDL-associated SAA represents a marker of intravascular inflammation in patients with stable CAD, more sensitive than C-reactive protein, or free SAA [51]. Finally, a strong independent relationship between SAA and future cardiovascular events was demonstrated [52,53]. All this evidence underlines not only the potential of this protein as a marker of advanced vascular disease but also the mechanisms by which SAA plays a direct role in atherosclerosis development.
Our results further support findings on proatherogenic effects of AP SAA and suggest a role of LDL as AP SAA carrier into the artery wall, where it could exert its noxious effects.

Conclusions
Our aim was to evaluate the apolipoprotein expression profiles in VLDL, LDL, and HDL fractions from patients undergoing carotid endarterectomy and healthy normolipidemic subjects by means of 2-DE coupled to Mass Spectrometry. By this approach, for the first time, AP SAA was identified as differentially expressed. Its levels were higher in all lipoprotein fractions, rising up to fourteen fold in LDL from atherosclerotic patients, indicating a potential association between AP SAA and the presence of advanced carotid lesions. Although preliminary, our findings are in accordance with a number of in vitro and in vivo evidence sources for a contributory role of this protein in atherogenesis. The adopted methodological approach was proven to be a powerful tool for identifying lipoprotein-associated markers in cardiovascular research. Further studies are in progress to evaluate the potential of LDL-associated AP SAA as a marker of carotid atherosclerosis as well as the effects of its increase in inducing inflammatory mechanisms underlying plaque development and progression/vulnerability.