Shotgun proteomic analysis reveals proteome alterations in HDL of patients with cholesteryl ester transfer protein deficiency

Journal of Clinical Lipidology (2019) -, -–-

Original Research

Shotgun proteomic analysis reveals proteome alterations in HDL of patients with cholesteryl ester transfer protein deficiency Takeshi Okada, MD, PhD, Tohru Ohama, MD, PhD, Kazuaki Takafuji, PhD, Kotaro Kanno, MD, Hibiki Matsuda, MD, Masami Sairyo, MD, PhD, Yinghong Zhu, MD, Ayami Saga, MS, Takuya Kobayashi, MS, Daisaku Masuda, MD, PhD, Masahiro Koseki, MD, PhD, Makoto Nishida, MD, PhD, Yasushi Sakata, MD, PhD, Shizuya Yamashita, MD, PhD* Department of Cardiovascular Medicine, Osaka University Graduate School of Medicine, Osaka, Japan (Drs Okada, Ohama, Kanno, Matsuda, Sairyo, Zhu, Saga, Kobayashi, Masuda, Koseki, Nishida, Sakata, and Yamashita); Department of Dental Anesthesiology, Osaka University Graduate School of Dentistry, Osaka, Japan (Dr Ohama); Department of Bio– System Pharmacology, Osaka University Graduate School Graduate, School of Medicine, Osaka, Japan (Dr Takafuji); Health Care Division, Health and Counseling Center, Osaka University, Osaka, Japan (Dr Nishida); Department of Community Medicine, Osaka University Graduate School of Medicine, Osaka, Japan (Dr Yamashita); and Department of Cardiology, Rinku General Medical Center, Osaka, Japan (Dr Yamashita) KEYWORDS: Cholesteryl ester transfer protein deficiency; High density lipoprotein; Shotgun proteomic analysis; Ultracentrifugation; LC-MS/MS analysis

BACKGROUND: We previously reported that the patients with cholesteryl ester transfer protein (CETP) deficiency (CETP-D) show marked changes in the size and lipid compositions of highdensity lipoprotein (HDL) and that they are not protected from atherosclerotic cardiovascular diseases, despite increased serum HDL-cholesterol (HDL-C) levels. HDL particles carry a variety of proteins, some of which are known to have antiatherogenic functions. OBJECTIVE: This study aimed to investigate the protein composition of HDL particles in patients with CETP-D. METHODS: Eight patients with complete deficiency of CETP and 8 normolipidemic healthy subjects were enrolled. We performed shotgun proteomic analysis to investigate the proteome of ultracentrifugally isolated HDL. RESULTS: We identified 79 HDL-associated proteins involved in lipid metabolism, protease inhibition, complement regulation, and acute-phase response, including 5 potential newly identified HDLassociated proteins such as angiopoietin-like3 (ANGPTL3). Spectral counts of apolipoprotein (apo) E were increased in patients with CETP-D compared with controls (60.3 6 6.9 vs 43.7 6 2.5, P , .001), which is concordant with our previous report. Complement regulatory proteins such as C3, C4a, C4b, and C9 were also significantly enriched in HDL from patients with CETP-D. Furthermore, apoC-III and ANGPTL3, both of which are now known to associate with increased

* Corresponding author. Department of Community Medicine, Osaka University Graduate School of Medicine, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan.

1933-2874/Ó 2019 National Lipid Association. All rights reserved. https://doi.org/10.1016/j.jacl.2019.01.002

E-mail address: [email protected] Submitted July 18, 2018. Accepted for publication January 7, 2019.

Journal of Clinical Lipidology, Vol -, No -, - 2019

2

atherosclerotic cardiovascular diseases, were enriched in patients with CETP-D compared with normolipidemic subjects (35.9 6 5.3 vs 27.1 6 3.7, 2.3 6 1.1 vs 0.4 6 1.1, respectively; P , .01). CONCLUSION: We have characterized HDL-associated proteins in patients with CETP-D. We identified a significant increase in the amount of apoE, apoC-III, ANGPTL3, and complement regulatory proteins. These proteomic changes might be partly responsible for the enhanced atherogenicity of patients with CETP-D. Ó 2019 National Lipid Association. All rights reserved.

Introduction Epidemiological studies have revealed that plasma HDL-cholesterol (HDL-C) levels showed a negative correlation with the incidence of atherosclerotic cardiovascular diseases (ASCVD).1 However, recent studies have elucidated that the association between HDL-C and all-cause mortality was U-shaped in general population, with extremely high HDL-C levels being associated with high mortality in general population.2,3 We previously reported 2 cases of cholesteryl ester transfer protein (CETP) deficiency with corneal opacity and premature coronary heart disease.4 Moreover, we also reported that in the Omagari district, Akita Prefecture of Japan, where subjects with a marked hyperalphalipoproteinemia (HALP) due to CETP deficiency (CETP-D) caused by the intron 14 splice donor site mutation are very frequent, U-shaped relationship was observed between serum HDL-C levels and the incidence of ischemic ECG changes.5 Furthermore, the prevalence of both marked HALP and the intron 14 splice donor site mutation was significantly lower in subjects aged .80 years than in the younger generation.5 Thus, a marked HALP caused by intron 14 splice donor site CETP gene mutation does not represent a longevity syndrome, suggesting the importance of re-evaluation of the clinical significance and pathophysiology of a marked HALP. However, the development of CETP inhibitors has been performed as a strategy to raise HDL-C levels and lower LDLcholesterol (LDL-C). The development of CETP inhibitors such as torcetrapib,6 dalcetrapib,7 and evacetrapib8 has been terminated because of the absence of significant clinical effects or even increased ASCVD events. Anacetrapib9 has been shown to give a modest reduction of coronary events, but it may be due to LDL-C reduction and not HDL-C increase. The development of anacetrapib has been discontinued recently. The major component of HDL, apolipoprotein (apo)A-I, which is synthesized by the liver and intestine, takes up cholesterol from lipid-laden macrophages through ABCA1. HDL particles also take up cholesterol from foam cells through ABCG1, growing up to mature HDL particles. The cholesterol in mature HDL is transferred by CETP to apoBcontaining lipoproteins, and then these lipoproteins are taken up by the liver through LDL receptor. This pathway is called ‘‘reverse cholesterol transport (RCT),’’ by which HDL exerts anti-atherosclerotic functions. Cholesterol efflux capacity of HDL has a strong inverse association

with ASCVD.10,11 Furthermore, in addition to cholesterol efflux capacity, HDL has a variety of antiatherogenic functions such as antioxidant, anti-inflammatory, antithrombotic, and endothelial function-improving effects. The protein cargo of HDL is considered to contribute to these functions.12 Recent proteomic analysis of HDL has identified more than 85 HDL-associated proteins, which may exert these antiatherogenic functions.13 Based on the atherogenic properties of patients with CETP-D, we hypothesized that CETP deficiency may demonstrate altered protein composition of HDL particles. The aim of the present study was to investigate the HDL proteome in patients with CETP-D using shotgun proteomic analysis.

Materials and methods Subjects We enrolled 8 patients with complete deficiency of CETP, whose serum CETP mass was ,0.1 mg/mL (reference range in controls [mean 6 SD]: 2.4 6 0.6 mg/mL),14 and age- and sex-matched 8 healthy normolipidemic subjects in the present study. Among 8 patients with CETP-D, 5 patients were homozygous for a G-to-A change at the 5’-splice donor site of the intron 14. One patient was heterozygous for the intron 14 mutation but totally lacked CETP.15 Another one was a compound heterozygous patient with the intron 14 mutation and a G-to-A change at codon 181 of exon 6 mutation (G181X),16 and the other one’s mutation was not determined. All subjects provided written informed consent, and the research protocol was approved by the Ethics Committee of Osaka University Hospital. Venous blood was drawn after overnight fasting (over 12 h). Serum was separated by lowspeed centrifugation (3000 rpm, 15 min, at 4 C), and aliquots were frozen at 280 C until use.

Analysis of serum lipids and lipoproteins Serum total cholesterol (TC), triglyceride (TG), HDL-C, and LDL-C levels were determined by enzymatic methods (Sekisui Medical Co, Tokyo, Japan). Serum concentrations of apoA-I, apoA-II, apoB, apoC-II, apoC-III, and apoE were measured by an immunoturbidimetric method (Sekisui Medical Co, Tokyo, Japan). Serum CETP mass levels were determined by an enzyme immunoassay (BML, INC, Tokyo, Japan).

Okada et al

HDL proteomic analysis in CETP-D

3

Isolation of HDL

Relative quantification based on spectral counts

HDL was isolated by sequential ultracentrifugation (d 5 1.063–1.21 g/mL), using solid sodium bromide for density adjustment, as described previously.17 The protein concentration of HDL was measured by modified BCA method.18

Normalized spectral counts were calculated using Scaffold version 4 (Proteome Software Inc).21 Database search files generated by Mascot were imported into Scaffold. Peptide and protein identifications were accepted if established at .95.0% probability as specified by the Peptide Prophet22 and Protein Prophet23 algorithms, respectively. Each identified protein required at least 2 unique peptides to be part of the data set. To correct for sampling differences between individuals, normalized spectral counts were calculated as follows. First, we calculated the total number of spectra in each subject. Next, we calculated the average of total number of spectra across all subjects. Finally, we multiplied the subject’s spectrum count of each HDL-associated protein by the average count over the subject’s total spectral count.24 To increase stringency of identification, proteins were required to be present in at least 5 of the 8 groups analyzed. The spectral index was used for the relative quantification of each protein of interest.25 The spectral index was calculated as: [(SCETP-D/(SCETP-D 1 SHealthy))x(NpCETP-D/NtCETP-D)]– [(SHealthy/(SCETP-D 1 SHealthy))!(NpHealthy/NtHealthy)], where SCETP-D and SHealthy represent normalized spectral count in HDLCETP-D and HDLHealthy, respectively; NpCETP-D and NpHealthy represent the number of samples where peptides were found for a protein in HDLCETP-D and HDLHealthy, respectively; and NtCETP-D and NtHealthy represent total number of samples of HDLCETP-D and HDLHealthy, respectively.

In-solution digestion For tryptic digestion, 20 mg of HDL protein was solubilized in 50 mM Tris-HCl (pH 9.0), containing 6 M urea and 5% sodium deoxycholate and reduced with 10 mM dithiothreitol for 60 minutes at 37 C and alkylated with 55 mM iodoacetamide for 30 minutes in the dark at 25 C. The reduced and alkylated samples were diluted 10-fold with 50 mM Tris-HCl (pH 9.0) and digested with trypsin at 37 C for 16 hours (trypsin-to-protein ratio of 1:20 (w/w)). An equal volume of ethyl acetate was added to each sample solution, and the mixtures were acidified with the final concentration of 0.5% trifluoroacetic acid. The mixtures were shaken for 1 minute and centrifuged at 15,700 ! g for 2 minutes. Then, the aqueous phase was collected. Digested samples were desalted with C18-StageTips.19,20

LC-MS/MS analysis LC-MS/MS analysis was performed by UltiMate 3000 Nano LC systems (Thermo Fisher Scientific) coupled to QExactive hybrid quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific) with a nanoelectrospray ionization source. Digested samples (5 mg protein for each sample) were injected by an autosampler and enriched on a C18 reverse phase trap column (100 mm I.D. ! 5 mm length, Thermo Fisher Scientific) at a flow rate of 4 mL/min. The sample was subsequently separated by a C18 reverse phase column (75 mm I.D. ! 150 mm length, Nikkyo Technos Co Ltd, Tokyo, Japan) at a flow rate of 300 nL/min with a linear gradient from 2% to 35% mobile phase B. Mobile phase B consisted of 95% acetonitrile with 0.1% formic acid, whereas mobile phase A consisted of 2% acetonitrile with 0.1% formic acid. The peptides were ionized using nanoelectrospray ionization in positive ion mode.

Data processing The raw data files were analyzed by Mascot Distiller v2.3 (Matrix Science, London) to create peak lists on the basis of the recorded fragmentation spectra. Peptides and proteins were identified by Mascot v2.3 (Matrix Science, London) against UniProt database with a precursor mass tolerance of 10 ppm, a fragment ion mass tolerance of 0.01 Da, and strict trypsin specificity allowing for up to 1 missed cleavage. The carbamidomethylation of cysteine and the oxidation of methionine were allowed as variable modification.

Immunoblot analysis HDL proteins were applied to an SDS PAGE gel (Tefco, Tokyo, Japan), separated using SDS running buffer and subsequently transferred onto polyvinylidene difluoride membranes. The membranes were incubated for 30 minutes at room temperature in Blocking One blocking buffer (Nacalai Tesque, Kyoto, Japan). Then, they were probed with rabbit polyclonal anti-apoA-I antibody, goat polyclonal anti-apoC-III antibody, rabbit polyclonal antiapoE antibody, or goat polyclonal anti-complement C3 antibody (1:1000, Santa Cruz Biotechnology, Dallas, TX) in TBS with 0.1%Tween-20 (TBS-T) overnight at 4 C. After corresponding anti-rabbit or anti-goat secondary antibodies were exposed to membranes for 30 minutes, ECL Prime Western Blotting Substrate (Thermo Scientific, Rockford, IL) was reacted. Images were analyzed by ImageQuant LAS 4000 mini (GE Healthcare, Uppsala, Sweden).

Statistical analysis Data were analyzed by the Graphpad Prism Ver.7.01 software program (GraphPad Software, CA) and were expressed as the mean 6 SD. Two-group comparison was performed with Student’s t-test or Mann-Whitney U test, as

Journal of Clinical Lipidology, Vol -, No -, - 2019

4 appropriate. Statistical significance was established at a P value of ,.05.

haptoglobin (HP), which were already listed as HDLassociated proteins.13

Results

Quantitative analysis of HDL protein abundance between normolipidemic subjects and patients with CETP-D

Lipid profiles of patients with CETP-D The lipid profile of healthy subjects and patients with CETP-D is shown in Table 1. As we reported previously,26 serum HDL-C and apoA-I levels were markedly higher in patients with CETP-D than in healthy subjects. Furthermore, serum apoC-III and apoE levels were also significantly higher, whereas serum LDL-C and apoB levels were significantly lower in CETP-D than in healthy subjects.

Enumeration of HDL-associated proteins Although we applied stringent criteria for the identification of HDL-associated proteins to increase the specificity of our study, we identified 79 HDL-associated proteins in the present study, including 5 potential newly identified HDL-associated proteins such as ACTB, KRTDAP, ANGPTL3, SELL, and IGF2 (Table 2). These 5 proteins were not documented by the HDL Proteome Watch (initiated by the Davidson Lab, Cincinnati, OH; database version August 14, 2015). It should be carefully discussed whether these proteins were novel HDLassociated proteins or just artifacts. Of these, KRTDAP, which was present in 7 of 8 subjects in both groups, may be a keratin contaminant during the sample handling. On the other hand, it is a notable finding that ANGPTL3, which is an important modulator of lipid metabolism, might be a potential novel HDL-associated protein. Because of the stringent criteria, we could not identify several proteins such as serotransferrin (TF), hemopexin (HPX), and

Table 1

To assess the relative protein abundance of HDL proteins of normolipidemic subjects and patients with CETP-D, we applied the same amount of HDL protein (5 mg for each sample) to LC-MS/MS, then evaluated spectral counting and spectral index. As shown in Figure 1, we identified 79 HDL-associated proteins, and spectral index shows that HDL proteome in CETP-D was quite different from that in controls. A positive spectral index suggests an enrichment of peptides derived from the protein of interest in HDLCETP-D, whereas a negative spectral index suggests reduction of peptides derived from the protein of interest in HDLCETP-D. The smallest value protein was CETP, and the largest value protein was ANGPTL3.

HDL-associated proteins involved in lipid metabolism, protease inhibition, complement regulation, and acute-phase response As shown in Table 2, supplementary Table 1, and supplementary Figure 1, when comparing with the same HDL protein weight, apoA-I was slightly, and apoA-II was moderately less in the HDL of patients with CETPD. CETP was only seen in normolipidemic subjects but not detected in patients with CETP-D. In contrast, apoE was significantly increased in patients with CETP-D, which is concordant with our previous report.27 ApoC-III is known to inhibit lipoprotein lipase (LPL) thereby attenuating the catabolism of triglyceride-rich lipoproteins.28 It is notable that apoC-III was also significantly increased in patients with CETP-D. Recent report has demonstrated

Clinical characteristics and lipid profiles of normolipidemic healthy subjects and patients with CETP-D Healthy subjects

Age (y) Sex (m/f) TC (mg/dL) TG (mg/dL) HDL-C (mg/dL) LDL-C (mg/dL) Apolipoprotein A-I (mg/dL) Apolipoprotein A-II (mg/dL) Apolipoprotein B (mg/dL) Apolipoprotein C-II (mg/dL) Apolipoprotein C-III (mg/dL) Apolipoprotein E (mg/dL) CETP mass (mg/mL)

56.6 3/5 211.8 102.8 60.9 116.6 154.0 30.1 96.3 4.6 10.2 4.6 3.4

6 15.5 6 6 6 6 6 6 6 6 6 6 6

24.9 26.3 8.2 19.5 20.6 2.9 14.4 1.0 2.2 1.1 1.0

Patients with CETP-D 56.0 3/5 320.1 113.1 167.6 49.7 250.0 40.1 73.7 7.8 29.2 13.9 ,0.1

P value

6 16.4

n.s.

6 6 6 6 6 6 6 6 6 6

.017 n.s. ,.001 ,.001 ,.001 .018 .006 .013 .015 ,.001 ,.001

110.4 89.3 44.7 11.4 61.7 10.1 13.7 3.0 19.1 6.2

HDL-C, high-density lipoprotein-cholesterol; LDL-C, low-density lipoprotein-cholesterol; TC, total cholesterol; TG, triglyceride. Data are shown as the mean 6 SD, and statistical significance was calculated by unpaired Student’s t test.

Okada et al Table 2

HDL proteomic analysis in CETP-D

5

HDL-associated proteins with significant difference between healthy subjects and patients with CETP-D HDL-derived peptide

Protein name Apolipoprotein A-I Apolipoprotein B-100 Apolipoprotein A-II Apolipoprotein E Apolipoprotein C-III Complement C3 Complement C4-B Vitamin D-binding protein Prenylcysteine oxidase 1 BPI fold-containing family B member 1 Vitronectin Cholesteryl ester transfer protein Actin, cytoplasmic 1* Alpha-2-antiplasmin Hemoglobin subunit beta Retinol-binding protein 4 Ig alpha-1 chain C region Integrin alpha-Iib Apolipoprotein A-V Ig mu chain C region Insulin-like growth factor-binding protein complex acid labile subunit Filamin-A Angiopoietin-related protein 3* Thrombospondin-1 Hemoglobin subunit alpha Prosaposin Mean total spectral counts/subject

Protein short name

Healthy subjects (n 5 8)

Patients with CETP-D (n 5 8)

P value

APOA1 APOB APOA2 APOE APOC3 C3 C4B GC PCYOX1 BPIFB1 VTN CETP ACTB SERPINF2 HBB RBP4 IGHA1 ITGA2B APOA5 IGHM IGFALS

938.8 168.0 158.2 43.7 27.1 16.9 15.7 14.5 11.1 5.6 5.2 11.2 6.9 3.7 6.6 5.2 2.0 4.4 1.4 0.5 0.9

872.1 243.3 102.7 60.3 35.9 34.4 30.9 12.0 14.9 10.6 6.1 0.0 2.5 5.9 2.6 2.9 5.8 1.6 3.3 2.7 3.0

.038 ,.001 ,.001 ,.001 .002 ,.001 ,.001 .028 .007 .010 .038 ,.001 .019 .038 .010 .028 .003 .011 .020 .049 .030

FLNA ANGPTL3 THBS1 HBA1 PSAP

3.5 (5/8) 0.4 (1/8) 2.0 (6/8) 2.5 (6/8) 1.8 (6/8) 2054.8 6 19.8

(8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (7/8) (7/8) (8/8) (8/8) (5/8) (7/8) (5/8) (2/8) (3/8)

(8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (8/8) (0/8) (4/8) (8/8) (7/8) (7/8) (8/8) (5/8) (8/8) (6/8) (6/8)

0.2 (1/8) 2.3 (7/8) 0.6 (2/8) 0.0 (0/8) 0.0 (0/8) 2056.7 6 11.2

.026 .008 .049 .007 .007

HDL was isolated from 8 control subjects and 8 patients with CETP-D by ultracentrifugation. The HDL proteome was analyzed by LC-MS/MS, and data were analyzed by searching the Uniprot database with Mascot 2.3 (MatrixScience). Values represent the mean spectral counts of peptides. The number of subjects with the protein identified is noted in parentheses. Statistical significance was calculated with the Mann-Whitney U test. *Newly identified HDL-associated proteins in this study.

that the HDL of patients with coronary artery disease (CAD) is enriched with apoC-III, which leads to stimulate potential endothelial proapoptotic pathways.21 Furthermore, ANGPTL3, which has an inhibitory effect on LPL activity and recently regarded as a potent therapeutic target as well as apoC-III, was also increased in the HDL of patients with CETP-D. There were no significant differences in apoA-IV, apoM, apoC-II, apoL-I, PLTP, and LCAT. We also identified HDL-associated proteins involved in protease inhibition such as AHSG, AGT, SERPINA1, SERPINF1, and SERPINF2. Among these proteins, SERPINF2, which inactivates plasmin, the important enzyme that participates in fibrinolysis, was significantly increased in patients with CETP-D. Regarding proteins involved in complement pathway and regulation, we have identified 4 HDL-associated proteins including C3, C4a, C4b, and C9. Especially, C3 and C4b were significantly increased in patients with CETP-D. C4a and C9 also tended to be increased in patients with CETP-D, although the difference was not significant. We also identified proteins involved in acute-phase response such as SAA1, SAA2,

SAA4, and F2. However, there were no significant differences in these proteins between normolipidemic subjects and patients with CETP-D. Regarding as the other proteins, GC, ACTB, HBB, RBP4, ITGA2B, FLNA, THBS1, HBA1, and PSAP were significantly decreased, whereas PCYOX1, BPIFB1, IGHA1, APOA-V, IGHM, and IGFALS were significantly increased in patients with CETP-D. To further validate the proteomic results, we performed immunoblot analysis of major apolipoproteins (apoA-I, apoC-III, apoE) and complement C3. Protein enrichment could be confirmed by appearance of specific bands for those proteins, and then we also performed densitometric scanning for quantification (Fig. 2A). When we compared in the same weight of HDL protein, apoA-I was slightly but significantly decreased, whereas apoE, apoC-III, and complement C3 were significantly increased in patients with CETP-D. These results were concordant with the data obtained from proteomic analysis (Table 2 and Fig. 2B), suggesting that these proteomic results are reliable.

6

Journal of Clinical Lipidology, Vol -, No -, - 2019

Figure 1 Relative abundance of proteins isolated from HDL of controls and patients with CETP-D, shown by the spectral index as described in methods. A positive spectral index suggests enrichment of peptides derived from the protein of interest in CETP-D, whereas a negative spectral index suggests reduction of peptides derived from the protein of interest in CETP-D.

Discussion We previously reported that a U-shaped relationship was observed between serum HDL-C levels and the incidence of ischemic ECG changes in the Omagari area (Akita Prefecture, Japan), where patients with HALP with the intron 14 splicing defect in the CETP gene are very frequent.5 Also, some epidemiological studies have demonstrated that lower plasma CETP activity was associated with greater CVD risk.29–32 Taken together, patients with HALP due to CETP deficiency were not protected against ASCVD, despite increased HDL-C levels. Furthermore, patients with HALP with SR-BI mutation were shown to have an increased risk of ASCVD.33 Recent evidence suggests that patients with a marked HALP is associated with an

increased mortality and cardiovascular mortality in European and Japanese populations.2,3 In the present study, we have focused on the proteome of HDL in patients with CETP-D rather than HDL-C levels. We compared the protein composition of HDL in normolipidemic subjects and patients with CETP-D and demonstrated that the composition was quite different between the 2 groups. When we compared in the same HDL protein weight, the spectral counts of apoA-I in CETP-D were lower than those of controls, in spite of higher serum apoAI levels. This is not only because particle numbers of larger HDLs in CETP-D are markedly increased34 but also because HDL of CETP-D is relatively enriched with several major apolipoproteins such as apoE and apoC-III, compared with controls. In the present study, apoE was

Okada et al

HDL proteomic analysis in CETP-D

7

Figure 2 Validation of apoA-_, apoE, apoC-___, and C3 levels in HDL isolated from controls and patients with CETP-D by western blots. Quantification by western blot followed by densitometric scanning (A) and normalized spectrum counts (B) of apoA-_, apoE, apoC-___, and C3.*P , .005 **P , .01 ***P , .001 by two-tailed Student’s t-test.

significantly increased in the HDL of patients with CETP-D as well as in the serum levels. As we have reported recently, the particle number of larger size HDLs, which are enriched in cholesteryl ester and apoE, was markedly increased in patients with CETP-D.34 These apoE-rich large HDLs are not able to prevent macrophages from accumulating cholesterol.35 Furthermore, the concentration of apoE in HDL is an independent predictor of recurrent coronary events.36 We also demonstrated that apoC-III was increased in CETP-D not only at serum level but also on HDL. ApoC-III is known to inhibit LPL or hepatic lipase,37 which mediates the hydrolysis of triglyceride-rich lipoproteins such as chylomicrons and VLDL, or intermediate density lipoprotein (IDL) and HDL, respectively. Elevated levels of apoC-III in plasma lead to inhibition of lipolysis and thereby impaired clearance of triglyceride-rich lipoproteins, resulting in the accumulation of atherogenic VLDL and chylomicron remnants. Furthermore, recent genomewide association studies have shown that loss-of-function

mutations in apoC-III lower triglycerides and protect from CAD.38 The antisense oligonucleotide against apoCIII mRNA decreases both apoC-III production and triglyceride concentrations39 and is currently evaluated in phase 3 trials. Regarding the function of apoC-III on HDL, previous studies showed a high ratio of apoC-III to apoA-I in HDL predicts recurrent coronary events.36 Rewanto et al also showed that the HDL isolated from patients with CAD was increased in apoC-III, and such apoC-III–rich HDL could not activate endothelial antiapoptotic pathways, but rather stimulated potential endothelial proapoptotic pathways.21 This mechanism could be also applied to the present study; the HDL of patients with CETP-D, enriched in apoC-III, might be dysfunctional HDL. Another novel finding in the present study may be the increased complement components such as C3, C4a, C4b, and C9 in the HDL of patients with CETP-D. Complement system plays a pivotal role in innate immunity by promoting antibody binding and opsonization or by direct killing

8 of invading pathogens.40 On the other hand, complement proteins are also involved in chronic inflammatory disease such as ASCVD. Actually, the complement system was shown to be activated within atherosclerotic plaques.41 Vaiser et al demonstrated that complement proteins such as complement C3 or C4 are associated with HDL.12 Several reports have shown that significant difference in complement C3 were noted in the HDL of patients with stable CAD or acute coronary syndrome (ACS),12,42 which may indicate that these HDLs are shifted to an inflammatory profile. Interestingly, the serum levels of complement C3 and C4 were not changed between healthy subjects and patients with CETP-D in the present study. Our finding that complement C3 or C4b was enriched in the HDL of patients with CETP-D might indicate the remodeling of HDL toward proinflammatory profile. The other novel and interesting finding in the present study may be that ANGPTL3 was identified as one of the HDL-associated proteins. In the healthy normolipidemic subjects, only 1 of 8 subjects showed the presence of ANGPTL3 on HDL. In contrast, in the HDL of patients with CETP-D, ANGPTL3 was identified in 7 of 8 patients with CETP-D. Although the potential roles of ANGPTL3 on HDL particles are currently unknown, ANGPTL3 inhibits LPL activity43 and loss-of-function variants of ANGPTL3 gene were associated with a decreased risk of CAD.44 These data may suggest the ANGPTL3 on enlarged HDL particles in patients with CETP-D may lead to an impaired function of HDL. Our study has several limitations. First, the sample size of this study was small. It was not easy to enroll CETPcompletely-deficient patients because of low prevalence even in Japan. Despite this limitation, however, our results could demonstrate the big difference of HDL proteome between the 2 groups. Second, because spectral counting is only semiquantitative, the accuracy of this LC-MS/MS analysis was limited. To validate the proteomic results, further immunoblot analysis was performed in several proteins. These data were consistent, showing that our proteomic data are valid. Third, the proteomic data show the contamination of ApoB-100 in HDL samples. There are several methods for HDL isolation, and ultracentrifugation can most purely separate HDLs. Indeed, most studies of HDL proteomic analysis reported the presence of ApoB100 in HDL fraction.13 Finally, we analyzed entire HDL population collected by ultracentrifugation. Given the fact that HDL is highly heterogeneous in particle size or composition, it would be better to perform proteomic analysis of HDL fractionated in detailed subclass by highperformance liquid chromatography, for example, in the future study. In conclusion, we have characterized HDL-associated proteins in patients with CETP-D and identified a significant increase in the amount of apoE, apoC-III, ANGPTL3, and complement regulatory proteins. These proteomic changes might be partly responsible for the enhanced atherogenicity of patients with CETP-D.

Journal of Clinical Lipidology, Vol -, No -, - 2019

Acknowledgments The authors are grateful to Ms. Kyoko Ozawa and Masumi Asaji for their assistance. This study was selffunded through the Department of Cardiovascular Medicine, Osaka University Graduate School of Medicine. All authors are responsible for the work in this article and were involved in at least one of the following: conception, design of the study, acquisition of data, analysis and interpretation of data, drafting the article and/or revising the article for important intellectual content.

Disclosures Daisaku Masuda and Shizuya Yamashita received research funds from Merck. The other authors report no conflicts.

Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.jacl.2019.01.002.

References 1. Sharrett AR, Ballantyne CM, Coady SA, et al. Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: the atherosclerosis risk in communities (ARIC) study. Circulation. 2001;104(10):1108–1113. 2. Madsen CM, Varbo A, Nordestgaard BG. Extreme high high-density lipoprotein cholesterol is paradoxically associated with high mortality in men and women: two prospective cohort studies. Eur Heart J. 2017; 38:2478–2486. 3. Hirata A, Sugiyama D, Watanabe M, et al. Association of extremely high levels of high-density lipoprotein cholesterol with cardiovascular mortality in a pooled analysis of 9 cohort studies including 43,407 individuals: the EPOCH-JAPAN study. J Clin Lipidol. 2018;12: 674–684.e5. 4. Matsuzawa Y, Yamashita S, Funahashi T, Yamamoto A, Seiichiro T. Selective reduction of cholesterol in HDL2 fraction by probucol in familial hypercholesterolemia and hyper HDL2 cholesterolemia with abnormal cholesteryl ester transferTitle. Am J Cardiol. 1988;62(3): 66B–72B. 5. Hirano K, Yamashita S, Nakajima N, et al. Genetic cholesteryl ester transfer protein deficiency is extremely frequent in the Omagari area of Japan. Marked hyperalphalipoproteinemia caused by CETP gene mutation is not associated with longevity. Arterioscler Thromb Vasc Biol. 1997;17(6):1053–1059. 6. Nissen SE, Tardif J-C, Nicholls SJ, et al. Effect of torcetrapib on the progression of coronary atherosclerosis. N Engl J Med. 2007;356(13): 1304–1316. 7. Schwartz GG, Olsson AG, Abt M, et al. Effects of dalcetrapib in patients with a recent acute coronary syndrome. N Engl J Med. 2012; 367(22):2089–2099. 8. Lincoff AM, Nicholls SJ, Riesmeyer JS, et al. Evacetrapib and cardiovascular outcomes in high-risk vascular disease. N Engl J Med. 2017; 376(20):1933–1942. 9. HPS3/TIMI55–REVEAL Collaborative Group, Bowman L, Hopewell JC, et al. Effects of anacetrapib in patients with atherosclerotic vascular disease. N Engl J Med. 2017;377(13):1217–1227.

Okada et al

HDL proteomic analysis in CETP-D

10. Khera AV, Cuchel M, de la Llera-Moya M, et al. Cholesterol efflux capacity, high-density lipoprotein function, and atherosclerosis. N Engl J Med. 2011;364(2):127–135. 11. Rohatgi A, Khera A, Berry JD, et al. HDL cholesterol efflux capacity and incident cardiovascular events. N Engl J Med. 2014;371(25): 2383–2393. 12. Vaisar T, Pennathur S, Green PS, et al. Shotgun proteomics implicates protease inhibition and complement activation in the antiinflammatory properties of HDL. J Clin Invest. 2007;117(3):746–756. 13. Shah AS, Tan L, Long JL, Davidson WS. Proteomic diversity of high density lipoproteins: our emerging understanding of its importance in lipid transport and beyond. J Lipid Res. 2013;54(10):2575–2585. 14. Nagano M, Yamashita S, Hirano K, et al. Point mutation (-69 G–.A) in the promoter region of cholesteryl ester transfer protein gene in Japanese hyperalphalipoproteinemic subjects. Arterioscler Thromb Vasc Biol. 2001;21(6):985–990. 15. Yamashita S, Hui DY, Sprecher DL, et al. Total deficiency of plasma cholesteryl ester transfer protein in subjects homozygous and heterozygous for the intron 14 splicing defect. Biochem Biophys Res Commun. 1990;170(3):1346–1351. 16. Arai T, Yamashita S, Sakai N, et al. A novel nonsense mutation (G181X) in the human cholesteryl ester transfer protein gene in Japanese hyperalphalipoproteinemic subjects. J Lipid Res. 1996;37(10): 2145–2154. 17. Havel BRJ, Eder HA, Bragdon JH. The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest. 1955;34:1345–1353. 18. Morton RE, Evans TA. Modification of the bicinchoninic acid protein assay to eliminate lipid interference in determining lipoprotein protein content. Anal Biochem. 1992;204(2):332–334. 19. Rappsilber J, Mann M, Ishihama Y. Protocol for micro-purification, enrichment, pre-fractionation and storage of peptides for proteomics using StageTips. Nat Protoc. 2007;2(8):1896–1906. 20. Masuda T, Tomita M, Ishihama Y. Phase transfer surfactant-aided trypsin digestion for membrane proteome analysis. J Proteome Res. 2008;7(2):731–740. 21. Riwanto M, Rohrer L, Roschitzki B, et al. Altered activation of endothelial anti-and proapoptotic pathways by high-density lipoprotein from patients with coronary artery disease: role of highdensity lipoprotein-proteome remodeling. Circulation. 2013;127(8): 891–904. 22. Keller A, Nesvizhskii AI, Kolker E, Aebersold R. Empirical statistical model to estimate the accuracy of peptide identifications made by MS/MS and database search. Anal Chem. 2002;74(20):5383–5392. 23. Nesvizhskii AI, Keller A, Kolker E, Aebersold R. A statistical model for identifying proteins by tandem mass spectrometry. Anal Chem. 2003;75(17):4646–4658. 24. Freund DM, Prenni JE. Improved detection of quantitative differences using a combination of spectral counting and MS/MS total ion current. J Proteome Res. 2013;12(4):1996–2004. 25. Fu X, Gharib SA, Green PS, et al. Spectral index for assessment of differential protein expression in shotgun proteomics. J Proteome Res. 2008;7(3):845–854. 26. Yamashita S, Maruyama T, Hirano KI, Sakai N, Nakajima N, Matsuzawa Y. Molecular mechanisms, lipoprotein abnormalities and atherogenicity of hyperalphalipoproteinemia. Atherosclerosis. 2000; 152(2):271–285.

9 27. Yamashita S, Sprecher DL, Sakai N, Matsuzawa Y, Tarui S, Hui DY. Accumulation of apolipoprotein E-rich high-density lipoproteins in hyperalphalipoproteinemic human-subjects with plasma cholesteryl ester transfer protein-deficiency. J Clin Invest. 1990;86(3):688–695. 28. Huff MW, Hegele RA. Apolipoprotein C-III: going back to the future for a lipid drug target. Circ Res. 2013;112(11):1405–1408. 29. Vasan RS, Pencina MJ, Robins SJ, et al. Association of circulating cholesteryl ester transfer protein activity with incidence of cardiovascular disease in the community. Circulation. 2009;120(24): 2414–2420. 30. Robins SJ, Lyass A, Brocia RW, Massaro JM, Vasan RS. Plasma lipid transfer proteins and cardiovascular disease. The framingham heart study. Atherosclerosis. 2013;228(1):230–236. 31. Ritsch A, Scharnagl H, Eller P, et al. Cholesteryl ester transfer protein and mortality in patients undergoing coronary angiography: the ludwigshafen risk and cardiovascular health study. Circulation. 2010; 121(3):366–374. 32. Duwensee K, Breitling LP, Tancevski I, et al. Cholesteryl ester transfer protein in patients with coronary heart disease. Eur J Clin Invest. 2010;40(7):616–622. 33. Vergeer M, Korporaal SJA, Franssen R, et al. Genetic variant of the scavenger receptor BI in humans. N Engl J Med. 2011;364(2): 136–145. 34. Okada T, Ohama T, Okazaki M, et al. Particle number analysis of lipoprotein subclasses by gel permeation HPLC in patients with cholesteryl ester transfer protein deficiency. PLoS One. 2018;13(1): e0190875. 35. Ishigami M, Yamashita S, Sakai N, et al. Large and cholesteryl esterrich high-density lipoproteins in cholesteryl ester transfer protein (CETP) deficiency can not protect macrophages from cholesterol accumulation induced by acetylated low-density lipoproteins. J Biochem. 1994;116(2):257–262. 36. Sacks M, Alaupovic P, Moye LA, et al. VLDL, apolipoproteins B, CIII, and E, and risk of recurrent coronary events in the cholesterol and recurrent events (CARE) trial. Circulation. 2000;102:1886–1892. 37. Ooi EMM, Barrett PHR, Chan DC, Watts GF. Apolipoprotein C-III: understanding an emerging cardiovascular risk factor. Clin Sci (Lond). 2008;114(10):611–624. 38. The TG and HDL Working Group of the Exome Sequencing Project, National Heart, Lung, and Blood Institute. Loss-of-function mutations in APOC3, triglycerides, and coronary disease. N Engl J Med. 2014; 371:22–31. 39. Gaudet D, Alexander VJ, Baker BF, et al. Antisense inhibition of apolipoprotein C-III in patients with hypertriglyceridemia. N Engl J Med. 2015;373(5):438–447. 40. Sarma JV, Ward PA. The complement system. Cell Tissue Res. 2011; 343(1):227–235. 41. Speidl WS, Kastl SP, Huber K, Wojta J. Complement in atherosclerosis: friend or foe? J Thromb Haemost. 2011;9(3):428–440. 42. Alwaili K, Bailey D, Awan Z, et al. The HDL proteome in acute coronary syndromes shifts to an inflammatory profile. Biochim Biophys Acta. 2012;1821(3):405–415. 43. Tikka A, Jauhiainen M. The role of ANGPTL3 in controlling lipoprotein metabolism. Endocrine. 2016;52(2):187–193. 44. Dewey FE, Gusarova V, Dunbar RL, et al. Genetic and pharmacologic inactivation of ANGPTL3 and cardiovascular disease. N Engl J Med. 2017;377(3):211–221.