Available online at www.sciencedirect.

com

Journal of Nutritional Biochemistry 97 (2021) 108809

RESEARCH PAPER

Postprandial plasma lipidome responses to a high-fat meal among healthy women

Marcos Yukio Yoshinaga a ,1,∗, Bruna Jardim Quintanilha b,c ,1, Adriano Britto Chaves-Filho a ,1,
Sayuri Miyamoto a, Geni Rodrigues Sampaio b, Marcelo Macedo Rogero b,c,∗∗
a
Laboratory of Modified Lipids, Department of Biochemistry, Institute of Chemistry, University of São Paulo, São Paulo, Brazil
b
Nutritional Genomics and Inflammation Laboratory, Department of Nutrition, School of Public Health, University of São Paulo, São Paulo, Brazil
c
Food Research Center (FoRC), CEPID-FAPESP, Research Innovation and Dissemination Centers São Paulo Research Foundation, São Paulo, Brazil

Received 12 November 2020; received in revised form 27 April 2021; accepted 8 June 2021

Abstract

Postprandial lipemia consists of changes in concentrations and composition of plasma lipids after food intake, commonly presented as increased levels
of triglyceride-rich lipoproteins. Postprandial hypertriglyceridemia may also affect high-density lipoprotein (HDL) structure and function, resulting in a net
decrease in HDL concentrations. Elevated triglycerides (TG) and reduced HDL levels have been positively associated with risk of cardiovascular diseases
development. Here, we investigated the plasma lipidome composition of 12 clinically healthy, nonobese and young women in response to an acute high-
caloric (1135 kcal) and high-fat (64 g) breakfast meal. For this purpose, we employed a detailed untargeted mass spectrometry-based lipidomic approach and
data was obtained at four sampling points: fasting and 1, 3 and 5 h postprandial. Analysis of variance revealed 73 significantly altered lipid species between
all sampling points. Nonetheless, two divergent subgroups have emerged at 5 h postprandial as a function of differential plasma lipidome responses, and
were thereby designated slow and fast TG metabolizers. Late responses by slow TG metabolizers were associated with increased concentrations of several
species of TG and phosphatidylinositol (PI). Lipidomic analysis of lipoprotein fractions at 5 h postprandial revealed higher TG and PI concentrations in HDL
from slow relative to fast TG metabolizers, but not in apoB-containing fraction. These data indicate that modulations in HDL lipidome during prolonged
postprandial lipemia may potentially impact HDL functions. A comprehensive characterization of plasma lipidome responses to acute metabolic challenges
may contribute to a better understanding of diet/lifestyle regulation in the metabolism of lipid and glucose.
© 2021 Elsevier Inc. All rights reserved.

Keywords: high-fat meal; postprandial; plasma lipidome; triglycerides; phosphatidylinositol; lipidomics; high-density lipoprotein.

1. Introduction pecially TG, may lead to accelerated atherosclerosis and cardiovas-

cular diseases (CVD) [4].
Postprandial lipemia is defined as modulations in plasma The importance of postprandial TG in predicting CVD resides
lipid concentrations and composition occurring after a meal. in the fact that humans spend most of the day in the postpran-
Triglyceride-rich lipoproteins are the main constituents of these dial state. Both case-control [5–8] and prospective epidemiological
changes and they are derived from intestinal chylomicrons and studies [9–12] have provided strong evidence for a positive cor-
hepatic very low-density lipoprotein (VLDL) [1,2]. Although a close relation of postprandial TG and risk of CVD, even more so than
association exists between fasting and postprandial triglycerides fasting TG concentrations [9,10]. In addition, the postprandial state
(TG) concentrations, intra-individual differences in plasma TG lev- induces an intense lipid remodeling of high-density lipoprotein
els between fasting and postprandial conditions may vary substan- (HDL). This remodeling is mainly characterized by the exchange
tially, partially due to insulin sensitivity [3]. Postprandial lipemia of cholesteryl esters and TG between HDL and triglyceride-rich
has gained attention because elevated levels of remnant lipids, es- lipoproteins (mediated by cholesteryl ester transfer protein), re-
sulting in a TG enrichment of HDL [11–13]. The high catabolic rates
of large, TG-rich HDL are deemed a major driving force for reduc-
∗
Corresponding author at: Marcos Yukio Yoshinaga - Av. Lineu Prestes tion of plasma HDL cholesterol (HDL-C) concentrations, which in
748 (São Paulo, SP) - 05508-0 0 0, Brazil
∗∗
turn is a strong and independent cardiovascular risk factor [14–
Corresponding author at: Marcelo Macedo Rogero - Av. Dr. Arnaldo 715
16]. Nonetheless, circulating HDL exists as a spectrum of particles
(São Paulo, SP) - 01246-904, Brazil
sizes from which lipid composition is shaped by a complex inter-
E-mail addresses: [email protected] (M.Y. Yoshinaga),
[email protected] (M.M. Rogero).
play of factors regulating HDL’s synthesis, intravascular remodel-
1
These authors contributed equally to this manuscript. ing and catabolism [17]. Thus, the dynamic remodeling of rem-

https://doi.org/10.1016/j.jnutbio.2021.108809
0955-2863/© 2021 Elsevier Inc. All rights reserved.
2 M.Y. Yoshinaga, B.J. Quintanilha, A.B. Chaves-Filho et al. / Journal of Nutritional Biochemistry 97 (2021) 108809

nant lipoprotein particles in the postprandial state opens the pos- centrifugation at 1200 g for 15 min at room temperature. Aliquots of plasma sam-
sibility for alternative markers of postprandial lipemia that may ples were stored at -80°C until analyses. The experimental protocol was approved
in December 2013 by the ethics research committee of the School of Public Health
complement results by existing methods to improve the prediction
at the University of São Paulo and each subject signed an informed consent (CAAE:
of disease states, including insulin resistance and CVD. 21959413.1.0 0 0 0.542).
Intervention studies on postprandial plasma lipid responses
have been focused on the evaluation of routine clinical mea-
2.2. Lipidomic analysis
surements (total TG, total cholesterol, low- and high-density
lipoproteins) and partly also apolipoprotein (apo) composition, Plasma lipidome extraction was performed as previously described [28]. In
particularly apoB48 and apoB100 from chylomicrons and VLDL, brief, 20 µL of plasma samples or 20 uL of water (extraction blanks) were spiked
with 50 μL of an internal standard mixture (Supplementary Table S2) and the vol-
respectively (reviewed in [18]). Given the worldwide rise in pa-
ume was adjusted to 450 μL using ice-cold methanol. To this solution, 1 mL of
tients with insulin resistance and type 2 diabetes mellitus and methyl tert-butyl ether (MTBE) was added and the mixture thoroughly vortexed for
their association with CVD risk, there exists a current burgeon- 15 s and stirred for 1h at 20 ºC. Simultaneously, pooled samples from each sampling
ing interest in mass spectrometry-based lipidomics for precision point and an external reference sample (National Institute of Standards Technology,
medicine and clinical studies [19,20]. This interest in the appli- NIST, human plasma 1950) were extracted. The pooled samples (n=4) represented
by an aliquot of each individual at a given sampling point (i.e., fasting, 1, 3 and 5
cation of lipidomics relies on the ability to monitor hundreds
h postprandial) were used for lipid identification by MS/MS. The NIST plasma 1950
of lipid molecular species in plasma, which may provide critical was used as quality control sample for analytical reproducibility tests. After cen-
information when examined in parallel to routine clinical mea- trifugation at 10,0 0 0 x g for 10 min at 4°C, the supernatant was transferred to a
surements. Despite the clinical relevance of postprandial lipemia, glass vial and dried under N2 gas. The dried lipid extracts were dissolved in 100 μL
of isopropanol. Blanks of injection (isopropanol) and quality control samples (NIST
very few studies have reported data on postprandial plasma
plasma 1950) were measured at the beginning and end of every batch composed of
lipidome responses [21–25]. Of note, two of these studies have 8 experimental samples, totalizing 6 measurements.
shown significant changes concerning the composition of specific Lipid extracts were analyzed by electrospray ionization time-of-flight mass
fatty acids linked to TG in the postprandial state [21,24], and more spectrometry (ESI-Q-TOFMS, Triple TOF® 6600, Sciex, Concord, US) interfaced with
recently an HDL enrichment in phosphatidylcholine was observed an ultra high-performance liquid chromatography (UHPLC Nexera, Shimadzu, Kyoto,
Japan). Details on the chromatographic conditions and mass spectrometry of this
in response to a high-fat meal [25].
untargeted lipidomic approach were described in a previous study by our group
In a previous investigation, the postprandial effects of a high-fat [29]. For reverse-phase LC, the injection volume was set to 2 µL and samples were
meal on plasma microRNA expression and inflammatory biomark- loaded into a CORTECS® (UPLC® C18 column, 1.6 µm, 2.1 mm i.d. x 100 mm) with a
ers levels were investigated in healthy women [26]. Here, our goal flow rate of 0.2 mL/min and the oven temperature maintained at 35 °C. The mobile
phase A consisted of water and/or acetonitrile (60:40) and mobile phase B of iso-
was to contribute to the understanding of postprandial lipemia
propanol/acetonitrile/water (88:10:2). Mobile phases A and B contained ammonium
elicited by an acute high-fat meal challenge in individuals with- acetate or ammonium formate (at a final concentration of 10 mM) for experiments
out metabolic syndrome by describing changes in plasma lipidome, performed in negative or positive ionization mode, respectively. Separation of lipids
particularly regarding the composition of remnant lipids. The ra- was achieved by a 20 min linear gradient as follows: from 40 to 100% B over the
tionale for investigating healthy individuals was to attenuate the first 10 min., hold at 100% B from 10 to 12 min., decreased from 100 to 40% B dur-
ing 12–13 min., and hold at 40% B until 20 min.
effects of insulin resistance, particularly exacerbated hypertriglyc-
The MS operated in both positive and negative ionization modes, with a scan-
eridemia, in postprandial lipemia. For this purpose, we applied a ning range of 20 0–20 0 0 Da. Data acquisition was performed with a cycle time pe-
comprehensive untargeted lipidomic approach using liquid chro- riod of 1.05 s with 100 ms for MS1 scan and 25 ms acquisition time to obtain the
matography coupled to mass spectrometry (MS), with the iden- top 36 precursor ions, using Analyst® 1.7.1. Ion spray voltages of −4.5 kV and 5.5 kV
were applied for negative and positive modes, respectively, and the cone voltage
tification of lipid species based exclusively on their MS/MS frag-
set at −/+80 V. The curtain gas was set to 25 psi, nebulizer and heater gases to
mentation. This approach allows discovery of a large number of 45 psi and interface heater at 450 °C. Lipid molecular species were manually iden-
molecules and is thus highly suited for plasma lipidome assess- tified exclusively based on their exact masses coupled to specific MS/MS fragments
ment of the postprandial state, which is characterized by an in- and/or neutral losses obtained by Information Dependent Acquisition (IDA) as out-
tense synthesis of diverse lipid species that are markedly distinct lined in Han (2016) [30]. We inspected ca. 300 ions in each ionization mode per
pooled sample (see above) and the quantification of each identified lipid was based
as compared to fasting conditions.
on their peak area using MS1 data. Quantification of lipid molecular species was
determined by comparison of chromatographic peak areas to those of the corre-
2. Methods
sponding internal standard (Supplementary Table S2), using a maximum of 5 mDa
as limit for attribution of the precursor ion. The amount of lipid molecular species
2.1. Study design and blood plasma sampling
was expressed in mg/dL of plasma to facilitate comparison with routine clinical
measurements, and data are presented as average ± standard error of the mean.
Twelve young, nonoverweight and clinically healthy women were enrolled in
Note that standards of some lipid classes (e.g., mono- and di-hexosyl ceramides
this study, with exclusion criteria being athletes, pregnancy, diabetes, intake of food
and acylcarnitines) were unavailable in the laboratory and thus lack absolute con-
supplements, prescribed anti-inflammatories, smokers, dysregulated lipid profiles
centrations. Thus, their concentrations are comparable among samples, but not with
and blood pressure, chronic gastrointestinal dysfunctions, among others. Detailed
other compounds. Data reproducibility analysis was performed using quality control
information on the study design and laboratory analyses (sugar, protein and lipid
sample measurements in both negative and positive ionization modes (the integral
contents in diets) is depicted elsewhere [26]. Anthropometric measures (waist cir-
lipidomics data are shown in Supplementary Table S3).
cumference, height and weight), blood pressure, routine biochemical (lipid profiles,
glycemia, C-reactive protein, insulin) and inflammatory biomarkers measurements,
including lipopolysaccharides (LPS), were described in a previous study [26] and 2.3. Statistical analyses
are displayed in Table 1 and Supplementary Table S1. The healthy status of par-
ticipants was based on the three out of five criteria for qualification of individu- Statistical analyses of data obtained from untargeted lipidomics and accom-
als with metabolic syndrome determined for Latin American women populations panying biochemical measurements were performed with Metaboanalyst (www.
[27]: waist circumference (WC ≥ 80 cm); TG ≥ 150mg/dL; high-density lipoprotein metaboanalyst.ca), following protocols by Xia and Wishart (2016) [31]. In brief, the
cholesterol (HDL-C ≤ 50 mg/dL); systolic blood pressure (SBP ≥ 130 mm Hg) and coefficient of variance (CV) was calculated for each quantified lipid in quality con-
diastolic blood pressure (DBP ≥ 85 mm Hg); glucose ≥ 100 mg/dL. According to trol samples, and lipid species displaying CV values above 20% were excluded from
Table 1, the participants lack traits of metabolic syndrome. further analyses. Given the often right-skewed and non-Gaussian distribution of the
Briefly, the participants received an identical standardized balanced dinner the lipidomic data [32], log-transformation was used to normalize the data prior to sta-
day before the test (before 8 PM), abstained from alcohol intake for 7 d prior, did tistical analyses. A general comparison of all samples was performed by one-way
not exercised 48 h and fasted for 12 h before blood sampling. The subjects ingested ANOVA, whereas pairwise comparisons were conducted by unpaired t-test. False
a high-caloric (1135 kcal – 47.3% lipids, 42.4% carbohydrates and 10.3% proteins) discovery rates were used throughout the analyses of variance or otherwise indi-
and high-fat (64 g fat, 69% saturated fatty acids) breakfast meal (Supplementary cated and significance was determined at P<.05. Total area under the curve values
Fig. S1). Blood sampling was conducted at fasting, 1 h, 3 h and 5 h after the in- were calculated using GraphPad Prism 8 as well as correlation analysis performed
take of the meal. Blood samples were obtained in EDTA tubes, kept on ice before by nonparametric Spearman correlation, after checking for data distribution. Please
 M.Y. Yoshinaga, B.J. Quintanilha, A.B. Chaves-Filho et al. / Journal of Nutritional Biochemistry 97 (2021) 108809 3

Table 1
Anthropometric and biochemical characteristics of participants at baseline (displayed as range
of values and average ± standard deviation (SD)) and number of individuals displaying traits of
metabolic syndrome (MS)

Clinical Parameter Range Average ± SD MS traits∗

Age (years) 21–36 27.2 ± 4.6 –
Weight (kg) 47.6–68.0 56.5 ± 5.9 –
Height (m) 1.56–1.69 1.62 ± 0.05 –
BMI (kg/m2) 18.7–24.8 21.5 ± 2.1 –
WC (cm) 65–86 72.0 ± 5.2 1
SBP (mmHg) 90–121 103.1 ± 9.5 0
DBP (mm Hg) 50–77 62.8 ± 8.3 0
Fasting TG (mg/dL)† 47–144 79.7 ± 30.8 0
Fasting HDL-C (mg/dL)† 31–81 58.5 ± 15.9 2
Fasting Glucose (mg/dL)† 86–99 89.2 ± 6.3 0
∗According to criteria by [27]. Please see Main Text for details on participants’ selection.
†
For biochemical measurements n=11, otherwise n=12. BMI, body mass index; WC, waist cir-
cumference; SBP and DBP, systolic and diastolic blood pressure, respectively; TG, triglycerides;
HDL-C, high-density lipoprotein cholesterol.

Table 2
Area under the curve values (average ± standard error of the mean) for significantly altered parameters in slow
versus fast TG metabolizers (t-test, P<.05 as shown in Fig. 2) and baseline anthropometric and biochemical
characteristics (average ± standard deviation) of participants relevant to metabolic syndrome [27]. Pairwise
comparisons between slow and fast TG metabolizers were performed by t-test (P<.05 are shown in italics). For
biochemical measurements n=11, otherwise n=12.

Parameters Slow Fast P value

AUC values
Insulin-clinical (µL U h mL−1 ) 240.0 ± 45.4 132.5 ± 9.7 .0495
TG-clinical (mg h dL−1 ) 681.6 ± 66.7 428.4 ± 18.1 .0064
LPS-clinical (EU h mL−1 ) 1.35 ± 0.08 1.01 ± 0.07 .0138
1H-Cer (mg h dL−1 ) 0.101 ± 0.005 0.081 ± 0.007 .0375
PC (mg h dL−1 ) 158.6 ± 4.6 135.2 ± 7.4 .0176
PI (mg h dL−1 ) 8.1 ± 0.7 4.5 ± 0.6 .0043
TG (mg h dL−1 ) 2197.7 ± 458.4 857.9 ± 224.7 .0437
Baseline characteristics
WC (cm) 69.9 ± 3.1 75.0 ± 6.3 .0886
SBP (mm Hg) 104.1 ± 10.4 101.8 ± 9.1 .7037
DBP (mm Hg) 62.3 ± 9.7 63.6 ± 6.8 .8011
Fasting TG (mg/dL) 90.8 ± 37.4 66.4 ± 14.9 .2059
Fasting HDL-C (mg/dL) 61.8 ± 8.9 54.6 ± 22.3 .4811
Fasting Glucose (mg/dL) 90.2 ± 5.3 88.0 ± 7.7 .5966

note that due to the availability of data, the number of samples used in statisti- h, and 5 h postprandial revealed that postprandial TG reflect to a
cal analyses for lipidomics (n=12) is distinct from other biochemical measurements large extent the composition of dietary fatty acids (Supplementary
(n=11).
Fig. S1).
Image deleted. Please check.Image deleted. Please check.Image
3. Results
deleted. Please check.Image deleted. Please check.Image deleted.
Please check.Image deleted. Please check.
In this study, a total of 12 young and nonobese women with
Analysis of variance revealed 73 significantly altered lipid
healthy clinical scores, according to routine biochemical measure-
species between all sampling points (one-way ANOVA; P<.05).
ments (Table 1; Supplementary Table S1), were enrolled [26]. The
These differentially modulated species were plotted in a heatmap
subjects were examined at four-time points: fasting and after a
as the fold-change from the mean for each individual sample (as
high-fat meal intake at 1 h, 3 h, and 5 h postprandial. By using an
shown in Fig. 1B). Baseline plasma lipidome was similar among
untargeted lipidomics approach, we identified and quantified 274
participants and clustering of fasting samples on the horizontal
individual species of lipids in plasma distributed into 18 classes
axis of the heatmap was link to low plasma triglycerides (TG) and
and/or subclasses as displayed in Figure 1A. We applied quality
high free fatty acids (FFA) concentrations. Clustering together with
control samples throughout the LC-MS routines and according to
fasting samples, six individuals at 1 h and two individuals at 3 h
the coefficient of variance (CV) for each lipid, 15 molecular species
displayed a delayed postprandial TG enrichment in plasma. More
were excluded from statistical analyses given their CV values above
importantly, two divergent subgroups (indicated by arrowheads in
20%. Qualitative analysis of fatty acids composition in TG species
Fig. 1B) have emerged as a function of plasma lipidome profiles
that displayed concentrations five fold higher than fasting at 1 h, 3
4 M.Y. Yoshinaga, B.J. Quintanilha, A.B. Chaves-Filho et al. / Journal of Nutritional Biochemistry 97 (2021) 108809

Fig. 1. Plasma lipidome composition (A) and postprandial alterations (B) after a high-fat meal. In A, The numerical abundance of lipid molecular species are given for TG,
coenzyme ubiquinone 10 (Q10), free cholesterol (FC), cholesteryl esters (CE), acylcarnitines (AC), sphingomyelin (SM), plasmenyl phosphatidyl-ethanolamine and -choline (pPE
and pPC), phosphatidyl-inositol, -ethanolamine and -choline (PI, PE and PC), plasmanyl phosphatidyl-ethanolamine and -choline (oPE and oPC), lyso-phosphatidylcholine (LPC),
free fatty acids (FFA), ceramides (Cer), di- and mono-hexosyl ceramides (2H-Cer and 1H-Cer). In B, the heatmap displays values of fold-change differences from the mean
(from 3 to -3) for the 73 significantly altered lipid species (one-way ANOVA; P<.05, false discovery rate or FDR<0.05 and Tukey post-hoc test detailed in Supplementary Table
S4A). The heatmap was constructed using Euclidean distance and Ward clustering (see methods section). Clustering of lipid species on the vertical axis of the heatmap was
simplified to the total number of species per lipid classes within the two major clusters. Samples on the horizontal axis are split into fasting, 1 h, 3 h, and 5 h postprandial.
The blue and orange arrowheads inside the heatmap indicate, respectively, fast and slow TG metabolizers, based on differential plasma lipidome profiles at 5h postprandial.

at 5 h postprandial. These groups were designated fast and slow tabolizers at 5 h postprandial exhibited not only higher overall TG
TG metabolizers, the former represented by individuals that re- relative to fast TG metabolizers, but also increased concentrations
turned to the cluster containing all fasting samples 5 h after meal of 9 PI and other 5 species (4 of them represented by cholesteryl
intake (i.e., relatively low plasma TG concentrations). We then esters) exceptionally modulated at this sampling point (Fig. 3C).
tested through pairwise comparisons whether temporal variations These differences are remarkable when considering that plasma
in clinical parameters (Supplementary Table S1) and the sum of concentrations of 82 of 91 total TG and 10 of 11 total PI species
lipid classes (measured by lipidomics) could differentiate these were increased in slow compared to fast TG metabolizers at 5 h
groups based on total area under curve (AUC) values. The re- postprandial. It is also worth noting that fold-change differences in
sults revealed significant differences in AUC values for insulin, plasma TG concentrations between fast and slow TG metabolizers
total TG and LPS as well as for lipid classes, including TG and were maximized in magnitude at 5 h compared to 1 h postprandial
phosphatidylinositol (PI), between fast and slow TG metabolizers (Fig. 3B and C).
(Fig. 2; Table 2). Importantly, differences in baseline charac- We next evaluated to what extent the plasma lipidome in fast
teristics of fast and slow TG metabolizers were not detectable and slow TG metabolizers returned to baseline conditions 5 h af-
(Table 2). Moreover, our data further indicated positive associa- ter the high-fat meal intake (Fig. 4A). As expected, the results re-
tions in total AUC values between clinical measurements and the vealed a higher number of lipid species modulated in slow than
lipidomic data, particularly for PI (Table 3). fast TG metabolizers at 5 h compared to fasting (Fig. 4B and C). In-
Pairwise comparison of plasma lipidome of fast versus slow TG terestingly, 23 lipid species were commonly modulated in plasma
metabolizers revealed minor differences at fasting and 3 h post- from fast and slow TG metabolizers at 5 h postprandial relative
prandial, while major differences appeared at 1 h and 5 h post- to fasting, including lower concentrations of 4 FFA and 2 acyl-
prandial (Fig. 3A). Whereas significant differences at 1 h postpran- carnitines as well as increased concentrations of 16 TG. Qualita-
dial were mostly linked to TG species (Fig. 3B), the slow TG me- tively, the fatty acid chain composition of the latter commonly
 M.Y. Yoshinaga, B.J. Quintanilha, A.B. Chaves-Filho et al. / Journal of Nutritional Biochemistry 97 (2021) 108809 5

Fig. 2. Temporal variation in plasma concentrations of clinical parameters (A; n=11) and lipid classes (B; n=12) in fast and slow TG metabolizers. These variables were selected
based on significant differences in area under the curve (AUC) values between the two groups (Table 2). Concentrations are given as average ± standard error of the mean and
star symbols indicate P<.05 (t-test). Correlation analyses between AUC values of lipids classes and clinical parameters are show in Table 3.

Table 3
Spearman rank (rs) correlation analysis of AUC values of parameters significantly altered in slow versus fast TG
metabolizers. ∗ TG-c = TG-clinic

Parameters Insulin TG-c∗ LPS TG PC PI 1H-Cer

Insulin 1
TG-c∗ 0.70∗ 1
LPS 0.63 0.77∗ 1
TG 0.61 0.93‡ 0.83† 1
PC 0.61 0.50 0.78† 0.48 1
PI 0.86† 0.79† 0.84† 0.83† 0.62 1
1H-Cer 0.19 0.33 0.30 0.24 0.42 0.11 1
∗ P values: <.05,
† P values: <.01,
‡ P values: <.001.

modulated TG species was mainly represented by saturated fatty eases (CVD) than fasting TG concentrations [8–10,18,33,34]. This
acids (mostly prominently 14:0, 16:0 and 18:0; Supplementary study investigated postprandial changes in plasma lipidome of
Fig. S2A), which were the main saturated fatty acids in the diet otherwise clinically healthy, nonoverweight and young women
(Supplementary Fig. S1A). A remarkable number of 63 TG were (Table 1) up to 5 h after an acute high-caloric high-fat break-
uniquely found in higher concentrations in plasma of slow TG me- fast meal intake. As a classical result of the regulatory role of in-
tabolizers at 5 h postprandial relative to fasting, and they were sulin after food intake, plasma FFA concentrations dropped signif-
qualitatively represented by molecular species containing mono- icantly within the first hour postprandial relative to fasting condi-
and polyunsaturated fatty acids, predominantly oleic and linoleic tions (Fig. 1B) as well as TG achieved their maximum concentra-
acids (18:1 and 18:2, respectively; Supplementary Fig. S2B). More- tions 3 h after the meal (Fig. 2). Although these changes in plasma
over, slow TG metabolizers also displayed higher plasma concen- lipids are well within the expected for such an acute metabolic
trations of 7 PI 5 h after the meal intake relative to fasting plasma, challenge, our study provides evidence for differential postpran-
among other minor changes (Fig. 4C). dial responses among currently healthy individuals based solely
on plasma lipidome alterations. This variability in postprandial re-
4. Discussion sponses among apparently healthy individuals is not novel, and has
been already reported as a function of incremental AUC values for
Postprandial lipemia is putatively characterized by hypertriglyc- TG [13]. Remarkably, however, our data reveal that prolonged post-
eridemia and possibly a stronger risk factor for cardiovascular dis- prandial lipemia in slow TG metabolizers is associated not only
6 M.Y. Yoshinaga, B.J. Quintanilha, A.B. Chaves-Filho et al. / Journal of Nutritional Biochemistry 97 (2021) 108809

Fig. 3. Pairwise comparisons of plasma lipidome of fast vs. slow TG metabolizers. Fig. 4. Pairwise comparisons of plasma lipidome of 5 h postprandial vs. fasting
(A) Venn diagram analysis of altered plasma lipids in fast versus slow TG metaboliz- state in fast and slow TG metabolizers. (A) Venn diagram analysis of significantly
ers at fasting, 1 h, 3 h, and 5 h postprandial. Significant differences in lipid species altered plasma lipids (t-test; P<0.05) at 5 h postprandial versus fasting in fast and
(t-test; P<.05) were most pronounced at 1 h and 5 h postprandial. (B and C) Fold slow TG metabolizers. (B and C) Fold change (FC, scaled to log2) plots of signifi-
change (FC, scaled to log2) plots of significantly altered lipid species according to cantly altered lipid species according to their respective classes at 5h postprandial
their respective classes in fast versus slow TG metabolizers at 1 h (B) and 5 h (C) versus fasting in fast (B) and slow (C) TG metabolizers. Filled and empty circles
postprandial. Filled and empty circles represent altered species (t-test; P<.05) with represent altered species with and without FDR<0.05, respectively. The statistical
and without FDR<0.05, respectively. The statistical analyses displayed in this figure analyses displayed in this figure are detailed in Supplementary Table S4C.
are detailed in Supplementary Table S4B.
 M.Y. Yoshinaga, B.J. Quintanilha, A.B. Chaves-Filho et al. / Journal of Nutritional Biochemistry 97 (2021) 108809 7

with elevated AUC values for TG and insulin, but also with a con- ester and phospholipid transfer proteins [45–48], accounting, re-
current plasma enrichment in several species of TG and PI (Fig. 2 spectively, for the enrichment in TG and PI content of HDL dur-
and 4; Table 2). ing prolonged postprandial lipemia. By comparing postprandial
Regarding the composition of remnant TG, fast and slow TG responses to a high-fat meal in patients with metabolic syn-
metabolizers at 5h postprandial shared an increase in TG species drome and healthy controls, incremental AUC values of TG, but
linked to saturated fatty acids relative to fasting, likely reflecting not metabolic syndrome status or fasting TG levels, were found
the high-fat meal composition. However, in slow TG metabolizers, to induce a shift towards large HDL particles, reflecting the in-
the rise in neutral lipids at 5 h postprandial relative to fasting was crease in TG content of HDL [13]. The high catabolic rates of TG-
associated with significant contribution of TG species containing rich HDL have been long suggested to contribute quantitatively
mono- and polyunsaturated fatty acids, primarily oleic and linoleic to reductions in the number of HDL particles [44,49,50], which
acids, but also arachidonic and docosahexaenoic acids (Supplemen- is a major quantitative lipoprotein abnormality in insulin resis-
tary Fig. S3). While chylomicron remnants were thought to rep- tance [15,51,52]. Nonetheless, whether PI enrichment of TG-rich
resent major contributors to postprandial hyperlipidemia [4], the HDL (elicited by postprandial lipemia as in slow TG metabolizers)
liver-derived very low-density lipoprotein (VLDL) remnant parti- exerts significant effects on distinct and diverse functions of HDL
cles were more recently found to account quantitatively for the is currently unknown.
increase in number of postprandial lipoproteins after food intake Regarding the role of negatively charged phospholipids in HDL
[35]. Here, we did not evaluate the abundance of apolipoproteins subpopulations, much attention has been given to phosphatidylser-
to ascertain the chylomicron or VLDL origin of remnant TG-rich ine, in particular its enrichment in small HDL associated with
lipoproteins. Nonetheless, it appears that remnant lipoproteins in high cholesterol efflux capacity [53,54]. However, given its preva-
slow TG metabolizers (i.e., 5 h after the meal) undergo an intense lence among negatively charged plasma phospholipids [36,37], PI
remodeling (towards an enrichment of TG with mono- and polyun- is likely to have a major impact on the electrostatic interactions
saturated fatty acids) relative to those TG commonly increased in of HDL with lipases, lipid transfer proteins, and cell surface pro-
all participants, which reflected the dietary fat composition (Sup- teins. In fact, in vitro and in vivo studies indicate that PI enrich-
plementary Fig. S1 and S2). ment in HDL inhibits its interactions with hepatic lipase (HL) and
While lipidomics studies of postprandial alterations are scarce lecithin:cholesterol acyltransferase (LCAT) [40,55,56]. These data in
in the literature [21–25], there is no report of PI enrichment among concert with the high clearance rates of PI-enriched HDL [39–
plasma phospholipids following a high-fat meal challenge. The 41] suggest a possible causal link between PI and low concentra-
close association of PI with the plasma lipidome of slow TG metab- tions of HDL particles. Low HDL concentrations as a consequence
olizers at 5 h postprandial suggests a role of this phospholipid in of familial apoA-I deficiency are characterized by PI enrichment
the metabolism of remnant TG-rich lipoproteins. Although present in HDL2 and HDL3 from heterozygotes relative to control subjects
as a minor phospholipid class, PI is the most abundant negatively [57]. Importantly, these authors suggest that this HDL lipidome sig-
charged lipid in the human plasma [36,37] and also known to be nature of apoA-I deficiency is associated with impaired cholesterol
exclusively bond to lipoproteins rather than nonlipoprotein frac- efflux capacity and anti-oxidative activity.
tions [38]. In normolipidemic subjects, PI is quantitatively present Lipidome profiling of HDL subpopulations from subjects with
in large high-density lipoprotein 2 (HDL2), at least three times metabolic syndrome revealed that HDL2 and HDL3 are enriched in
more concentrated than any other lipoprotein fraction such as low- both PI and TG relative to control individuals [58]. In epidemio-
density lipoprotein (LDL) or chylomicron [38]. Interestingly, PI ad- logical lipidomic studies, plasma concentrations of PI were found
ministration (enriched in chylomicrons or by venous injection) to positively associated with type 2 diabetes mellitus and prediabetes
animal models results in a rapid association of PI with HDL, which [59]. More recently, an independent positive correlation has been
in turn stimulates HDL catabolism [39–41]. The intense remodel- found between fasting lyso-PI concentrations and insulin secretion
ing of TG-rich lipoproteins during prolonged postprandial lipemia in healthy overweight subjects [60]. In support of the casual asso-
may also induce functional changes in HDL [42,43]. Particularly ciation of plasma PI and metabolic diseases evidenced by the above
well studied, an enrichment of TG content in HDL is stimulated in studies, our findings indicate a strong relationship between modu-
postprandial hypertriglyceridemia [11–13,44]. In contrast, qualita- lations in TG and PI content of HDL during prolonged postprandial
tive postprandial changes in HDL phospholipids composition have lipemia in apparently healthy subjects.
been only reported for phosphatidylcholine species [25], which in While routine clinical measurements of lipids or even
our study were not as strongly modulated as PI (Fig. 2). apolipoproteins, are largely applied to assess postprandial lipemia
To verify a potential link between plasma PI levels and HDL at 5 and its associated risk factors for CVD (reviewed in [18]), HDL
h postprandial, we have performed apoB precipitation for lipopro- lipidome profiling appear as an alternative tool, with potential use
tein fractionation followed by targeted lipidomic analysis of PI and in large clinical cohorts. Thus far, studies on HDL lipidome com-
TG species in both HDL and apoB fractions (please see Supplemen- bining structure and function analyses have been focused on static
tary Methods). For this analysis both slow and fast TG metaboliz- concentrations with no assessment of the dynamic processes regu-
ers were compared, but here the focus was given to individuals lating HDL-specific phospholipid composition [61]. A better under-
displaying the most discrepant AUC values for TG and insulin to standing of insulin actions in the HDL lipidome remodeling during
maximize differences between the groups. Our data indicate that postprandial lipemia may contribute to ascertain its functional or
PI and TG concentrations are significantly higher in HDL fraction of defective role in cardiovascular and metabolic diseases.
slow than fast TG metabolizers, but this difference was not appar- Limitations of this study include validation of results in a sec-
ent in the apoB fraction (Fig. 5A and 5B). Moreover, a great overlap ond and/or larger cohort. Although investigating a small number of
of altered lipid species in plasma and lipoprotein fractions was ob- subjects, we have focused on clinically healthy, nonoverweight and
served in the comparisons slow versus fast TG metabolizers at 5 young women to minimize putative variations in plasma lipidome
h postprandial (Fig. 5C). Thus, enhanced PI and TG concentrations composition according to metabolic conditions, gender and age
in HDL appear as major lipidome signatures of HDL remodeling in [62]. Therefore, our results may not be generalizable to other popu-
slow TG metabolizers 5 h after the meal. lations. In a recent large-scale study with healthy adults (n>1,0 0 0),
From a physiological standpoint, acute hyperinsulinemia and postprandial TG responses were considerably variable between in-
hypertriglyceridemia may lead to increased activities of cholesteryl dividuals, linked to factors like time of the day for a meal and
8 M.Y. Yoshinaga, B.J. Quintanilha, A.B. Chaves-Filho et al. / Journal of Nutritional Biochemistry 97 (2021) 108809

Fig. 5. Targeted lipidomic analysis of PI and TG species in HDL and apoB-containing fractions at 5 h postprandial. Total concentrations of PI and TG (A and B, respectively) in
HDL and apoB fractions from slow and fast TG metabolizers (blue and orange bars, respectively, n=4 each). Note that in B, distinct axes of TG concentrations are displayed for
HDL and apoB. Statistical analysis was performed by t-test (star symbols represent significant differences [P<.05], ns=not significant). Concentrations are given in area ratio,
that is, area of the lipid divided by the area of its corresponding internal standard. Please see Supplementary Methods for analytical details and selection of individuals for this
comparison. (C) Venn diagram analysis displaying significantly altered lipid species (t-test; P<.05) from the comparison fast versus slow TG metabolizers at 5 h postprandial
in plasma, HDL and apoB fractions (see statistical analyses in Supplementary Table S4D).

amount of sleep, but largely independent of a genetic compo- evaluation of postprandial plasma or lipoprotein-specific lipidome
nent [63]. Interestingly, these authors also demonstrated that a responses to a high-fat meal challenge may potentially help to de-
person’s postprandial TG response to two distinct meals (high-fat tect, predict and monitor metabolic diseases. Plasma lipidome pro-
and high-carbohydrate) of the same macronutrient content is con- filing may also contribute to a better understanding of diet and/or
sistently comparable despite the substantial inter-individual dif- lifestyle regulation in the metabolism of lipid and glucose.
ferences. While TG responses are observed for longer periods af-
ter food intake (up to 8 h) [64,65], our study was designed for CRediT authorship contribution statement
5 h postprandial, which is at the lower end of the postprandial
hyperlipidemic period [66]. Nonetheless, we employed an acute Marcos Yukio Yoshinaga: Conceptualization, Methodology, For-
metabolic challenge using 4 sampling points, with a detectable mal analysis, Investigation, Writing – original draft, Supervision.
peak in plasma TG concentrations after 3 h in both fast and slow Bruna Jardim Quintanilha: Conceptualization, Methodology, In-
TG metabolizers, as observed in many investigations [66]. It is im- vestigation, Resources, Writing – original draft. Adriano Britto
portant to mention that while plasma TG concentrations exhibited Chaves-Filho: Conceptualization, Methodology, Formal analy-
a trend of returning to baseline values in slow TG metabolizers, sis, Investigation, Writing – original draft. Sayuri Miyamoto:
concentrations of PI remained relatively high at 5 h postprandial Methodology, Writing – review & editing, Funding acquisition.
(Fig. 2). In contrast, both TG and PI plasma concentrations in fast Geni Rodrigues Sampaio: Methodology, Resources, Writing – re-
TG metabolizers returned to baseline conditions after 5 h post- view & editing. Marcelo Macedo Rogero: Conceptualization, Re-
prandial. Finally, consistent with previous investigations, no major sources, Writing – original draft, Supervision, Funding acquisition.
postprandial changes in plasma cholesterol concentrations (either
free or esterified as cholesteryl esters) were noticed (Fig. 1B and 4) Acknowledgments
[13,67].
Our findings provide new insights into differential postpran- The authors are thankful to Dr. Luiz Sérgio de Carvalho and Dr.
dial plasma lipidome responses to a high-fat meal among clin- Andrei Sposito for insightful discussions on the role of HDL in post-
ically healthy women. Inter-individual variability in postprandial prandial lipemia. We are also grateful to two anonymous reviewers
plasma lipidome composition (i.e., slow and fast TG metaboliz- for the thoughtful comments that significantly improved this work.
ers) was mainly related to TG and PI plasma concentrations, in
Declaration of competing interest
particular their distribution in HDL. Our data suggest that pro-
longed postprandial lipemia induces not only a quantitative trans-
The authors declare no conflict of interests.
fer of TG, but also PI to HDL. Whereas large, TG-rich HDL2 gen-
erated by postprandial lipemia or insulin resistant states displays Funding
high clearance rates that are implicated in overall reductions in
numbers of HDL particles [15,44,51,52], the role of PI in HDL is This work was supported by the São Paulo Research Founda-
still unknown. Given the quantitative relevance of phospholipids tion (FAPESP) (CEPID-Redoxoma grant 2013/07937–8 and CEPID-
in HDL (ca. 40–60% of its lipid content; [68]), the prevalence of Food Research Center (FoRC) grant 2013/07914-8). Postdoc fellow-
PI in HDL2 [38] and its importance as the most abundant neg- ships were received from the Brazilian Federal Agency for Support
atively charged phospholipid in the human plasma [36,37], it is and Evaluation of Graduate Education within the Ministry of Edu-
likely that PI plays a major role in the electrostatic interactions cation of Brazil (CAPES) to MYY, FAPESP to ABC-F.
during intravascular HDL metabolism, with possible consequences
for functions. Thus, establishing a direct link between PI enrich- Supplementary materials
ment in HDL and insulin resistance may provide a better under-
standing of how specific lipids impact HDL functions associated Supplementary material associated with this article can be
with elevated cardiovascular risk. If these premises are validated, found, in the online version, at doi:10.1016/j.jnutbio.2021.108809.
 M.Y. Yoshinaga, B.J. Quintanilha, A.B. Chaves-Filho et al. / Journal of Nutritional Biochemistry 97 (2021) 108809 9

References Rogero MM. Circulating plasma microRNAs dysregulation and metabolic endo-
toxemia induced by a high-fat high-saturated diet. Clin Nutr 2020;39:554–62.
[1] Masuda D, Yamashita S. Postprandial hyperlipidemia and remnant lipoproteins. doi:10.1016/j.clnu.2019.02.042.
J Atheroscler Thromb 2017;24:95–109. doi:10.5551/jat.RV16003. [27] Alberti KGMM, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI, Donato KA,
[2] Nakajima K, Tokita Y, Tanaka A, Takahashi S. The VLDL receptor plays a key et al. Harmonizing the metabolic syndrome: a joint interim statement of
role in the metabolism of postprandial remnant lipoproteins. Clin Chim Acta the International Diabetes Federation Task Force on Epidemiology and Pre-
2019;495:382–93. doi:10.1016/j.cca.2019.05.004. vention; National Heart, Lung, and Blood Institute; American Heart Associ-
[3] Delawi D, Meijssen S, Castro Cabezas M. Intra-individual variations of fasting ation; World Heart Federation; International Atherosclerosis Society; and In-
plasma lipids, apolipoproteins and postprandial lipemia in familial combined ternational Association for the Study of Obesity. Circulation 2009;120:1640–5.
hyperlipidemia compared to controls. Clin Chim Acta 2003;328:139–45. doi:10. doi:10.1161/CIRCULATIONAHA.109.192644.
1016/S0 0 09-8981(02)0 0420-5. [28] Matyash V, Liebisch G, Kurzchalia TV, Shevchenko A, Schwudke D. Lipid ex-
[4] Zilversmit DB. Atherogenesis: A postprandial phenomenon. Circulation traction by methyl- tert -butyl ether for high-throughput lipidomics. J Lipid
1979;60:473–85. doi:10.1161/01.CIR.60.3.473. Res 2008;49:1137–46. doi:10.1194/jlr.d70 0 041-jlr20 0.
[5] Groot PHE, Van Stiphout WAHJ, Krauss XH, Jansen H, Van Tol A, Van [29] Chaves-Filho AB, Pinto IFD, Dantas LS, Xavier AM, Inague A, Faria RL, et al.
Ramshorst E, et al. Postprandial lipoprotein metabolism in normolipi- Alterations in lipid metabolism of spinal cord linked to amyotrophic lateral
demic men with and without coronary artery disease. Arterioscler. Thromb. sclerosis. Sci Rep 2019;9:1–14. doi:10.1038/s41598- 019- 48059- 7.
1991;11:653–62. doi:10.1161/01.atv.11.3.653. [30] Han X. Lipidomics: comprehensive mass spectrometry of lipids. Hoboken, NJ,
[6] Karpe F, Boquist S, Tang R, Bond GM, De Faire U, Hamsten A. Remnant lipopro- USA: John Wiley & Sons, Inc; 2016. doi:101002/9781119085263.
teins are related to intima-media thickness of the carotid artery independently [31] Xia J, Wishart DS. Using metaboanalyst 3.0 for comprehensive
of LDL cholesterol and plasma triglycerides. J Lipid Res 2001;42:17–21. metabolomics data analysis. Curr Protoc Bioinforma 2016;2016 14.10.1-
[7] Lopez-Miranda J, Williams C, Larion D. Dietary, physiological, genetic 14.10.91https://doi.org/. doi:10.1002/cpbi.11.
and pathological influences on postprandial lipid metabolism. Br J Nutr [32] Mundra PA, Shaw JE, Meikle PJ. Lipidomic analyses in epidemiology. Int J Epi-
2007;98:458–73. doi:10.1017/S000711450774268X. demiol 2016;45:1329–38. doi:10.1093/ije/dyw112.
[8] Karpe F. Postprandial lipoprotein metabolism and atherosclerosis. J Intern Med [33] Bozzetto L, Della Pepa G, Vetrani C, Rivellese AA. Dietary impact on postpran-
1999;246:341–55. doi:10.1046/j.1365-2796.1999.00548.x. dial lipemia. Front Endocrinol (Lausanne) 2020;11. doi:10.3389/fendo.2020.
[9] Bansal S, Buring JE, Rifai N, Mora S, Sacks FM, Ridker PM. Fasting compared 00337.
with nonfasting triglycerides and risk of cardiovascular events in women. J Am [34] DKolovou G, P Mikhailidis D, Kovar J, Lairon D, G Nordestgaard B, Chye
Med Assoc 2007;298:309–16. doi:10.1001/jama.298.3.309. Ooi T, et al. Assessment and clinical relevance of non-fasting and postprandial
[10] Nordestgaard BG, Benn M, Schnohr P, Tybjærg-Hansen A. Nonfasting triglyc- triglycerides: an expert panel statement. Curr Vasc Pharmacol 2011;9:258–70.
erides and risk of myocardial infarction, ischemic heart disease, and death in doi:10.2174/157016111795495549.
men and women. J Am Med Assoc 2007;298:299–308. doi:10.1001/jama.298.3. [35] Nakajima K, Nakano T, Tokita Y, Nagamine T, Inazu A, Kobayashi J, et al.
299. Postprandial lipoprotein metabolism: VLDL vs chylomicrons. Clin Chim Acta
[11] Patsch JR, Prasad S, Gotto AM Jr, Bengtsson-Olivecrona G. Postprandial lipemia. 2011;412:1306–18. doi:10.1016/j.cca.2011.04.018.
A key for the conversion of high density lipoprotein2 into high density lipopro- [36] Quehenberger O, Dennis EA. The human plasma lipidome. N Engl J Med
tein3 by hepatic lipase. J Clin Invest 1984;74:2017–23. doi:10.1172/JCI111624. 2011;365:1812–23. doi:10.1056/NEJMra1104901.
[12] Lewis GF, O’Meara NM, Soltys PA, Blackman JD, Iverius PH, Druetzler AF, et al. [37] Burla B, Bendt AK, Wenk MR, Cazenave-Gassiot A, Lin MK, Torta F, et al. MS-
Postprandial lipoprotein metabolism in normal and obese subjects: compari- based lipidomics of human blood plasma: A community-initiated position pa-
son after the vitamin A fat-loading test. J Clin Endocrinol Metab 1990;71:1041– per to develop accepted guidelines. J Lipid Res 2018;59:2001–17. doi:10.1194/
50. doi:10.1210/jcem- 71- 4- 1041. jlr.S087163.
[13] Quintanilla-Cantú A, Peña-de-la-Sancha P, Flores-Castillo C, Mejía- [38] Dashti M, Kulik W, Hoek F, Veerman EC, Peppelenbosch MP, Rezaee F. A
Domínguez AM, Posadas-Sánchez R, Pérez-Hernández N, et al. Small HDL phospholipidomic analysis of all defined human plasma lipoproteins. Sci Rep
subclasses become cholesterol-poor during postprandial period after a fat 2011;1. doi:10.1038/srep00139.
diet intake in subjects with high triglyceridemia increases. Clin Chim Acta [39] Nilsson A, Chen Q, Dahlman E. Metabolism of chylomicron phosphatidylinosi-
2017;464:98–105. doi:10.1016/j.cca.2016.11.018. tol in the rat: Fate in vivo and hydrolysis with lipoprotein lipase and hepatic
[14] Gotto AM Jr, Brinton EA. Assessing low levels of high-density lipoprotein lipase in vitro. J Lipid Res 1994;35:2151–60.
cholesterol as a risk factor in coronary heart disease: a working group report [40] Stamler CJ, Breznan D, Neville TAM, Viau FJ, Camlioglu E, Sparks DL.
and update. J Am Coll Cardiol 2004;43:717–24. doi:10.1016/j.jacc.2003.08.061. Phosphatidylinositol promotes cholesterol transport in vivo. J Lipid Res
[15] Lewis GF, Rader DJ. New insights into the regulation of HDL metabolism and 20 0 0;41:1214–21.
reverse cholesterol transport. Circ Res 2005;96:1221–32. doi:10.1161/01.RES. [41] Burgess JW, Boucher J, Neville TAM, Rouillard P, Stamler C, Zachariah S, et al.
0 0 0 0170946.56981.5c. Phosphatidylinositol promotes cholesterol transport and excretion. J Lipid Res
[16] Collaboration Emerging Risk Factors, E Di Angelantonio, Sarwar N, Perry P, 20 03;44:1355–63. doi:10.1194/jlr.M30 0 062-JLR20 0.
Kaptoge S, Ray KK, et al. Major lipids, apolipoproteins, and risk of vascular [42] Guerin M, Egger P, Soudant C, Le Goff W, van Tol A, Dupuis R, et al. Cholesteryl
disease. JAMA 2009;302:1993–2000. doi:10.1001/jama.2009.1619. ester flux from HDL to VLDL-1 is preferen- tially enhanced in type IIB hyper-
[17] Karathanasis SK, Freeman LA, Gordon SM, Remaley AT. The changing face of lipidemia in the postprandial state. J Lipid Res 2002;43:1652–60. doi:10.1194/
HDL and the best way to measure it. Clin Chem 2017;63:196–210. doi:10.1373/ jlr.M20 0135-JLR20 0.
clinchem.2016.257725. [43] Julia Z, Duchene E, Fournier N, Bellanger N, Chapman MJ, Le Goff W, et al.
[18] Borén J, Matikainen N, Adiels M, Taskinen MR. Postprandial hypertriglyc- Postprandial lipemia enhances the capacity of large HDL2 particles to mediate
eridemia as a coronary risk factor. Clin Chim Acta 2014;431:131–42. doi:10. free cholesterol efflux via SR-BI and ABCG1 pathways in type IIB hyperlipi-
1016/j.cca.2014.01.015. demia. J Lipid Res 2010;51:3350–8. doi:10.1194/jlr.P009746.
[19] Hyötyläinen T, Ahonen L, Pöhö P, Orešič M. Lipidomics in biomedical [44] Patsch JR, Prasad S, Gotto AM Jr, Patsch W. High density lipoprotein2. Relation-
research-practical considerations. Biochim Biophys Acta - Mol Cell Biol Lipids ship of the plasma levels of this lipoprotein species to its composition, to the
2017;1862:800–3. doi:10.1016/j.bbalip.2017.04.002. magnitude of postprandial lipemia, and to the activities of lipoprotein lipase
[20] O’Donnell VB, Ekroos K, Liebisch G, Wakelam M. Lipidomics: current state of and hepatic lipase. Clin Invest 1987;80:341–7. doi:10.1172/JCI113078.
the art in a fast moving field. Wiley Interdiscip Rev Syst Biol Med 2020;12. [45] Riemens S, van Tol A, Sluiter W, Dullaart R. Elevated plasma cholesteryl ester
doi:10.1002/wsbm.1466. transfer in NIDDM: relationships with apolipoprotein B-containing lipoproteins
[21] Bonham MP, Linderborg KM, Dordevic A, Larsen AE, Nguo K, Weir JM, et al. and phospholipid transfer protein. Atherosclerosis 1998;140:71–9. doi:10.1016/
Lipidomic profiling of chylomicron triacylglycerols in response to high fat s0 021-9150(98)0 0111-7.
meals. Lipids 2013;48:39–50. doi:10.1007/s11745- 012- 3735- 5. [46] Rye KA, Clay MA, Barter PJ. Remodelling of high density lipoproteins by plasma
[22] Meikle PJ, Barlow CK, Mellett NA, Mundra PA, Bonham MP, Larsen A, et al. factors. Atherosclerosis 1999;145:227–38. doi:10.1016/s0021-9150(99)00150-1.
Postprandial plasma phospholipids in men are influenced by the source of di- [47] Huuskonen J, Olkkonen VM, Jauhiainen M, Ehnholm C. The impact of
etary fat. J Nutr 2015;145:2012–18. doi:10.3945/jn.115.210104. phospholipid transfer protein (PLTP) on HDL metabolism. Atherosclerosis
[23] Grace MS, Dempsey PC, Sethi P, Mundra PA, Mellett NA, Weir JM, et al. 2001;155:269–81. doi:10.1016/s0021-9150(01)00447-6.
Breaking up prolonged sitting alters the postprandial plasma lipidomic pro- [48] Borggreve SE, de Vries R, Dullaart RPF. Alterations in high-density lipoprotein
file of adults with type 2 diabetes. J Clin Endocrinol Metab 2017;102:1991–9. metabolism and reverse cholesterol transport in insulin resistance and type 2
doi:10.1210/jc.2016-3926. diabetes mellitus: role of lipolytic enzymes, lecithin:cholesterol acyltransferase
[24] Kessler K, Gerl MJ, Hornemann S, Damm M, Klose C, Petzke KJ, et al. Shotgun and lipid transfer proteins. Eur J Clin Invest 2003;33:1051–69. doi:10.1111/j.
lipidomics discovered diurnal regulation of lipid metabolism linked to insulin 1365-2362.2003.01263.x.
sensitivity in nondiabetic men. J Clin Endocrinol Metab 2020;105. doi:10.1210/ [49] Lamarche B, Uffelman KD, Carpentier A, Cohn JS, Steiner G, Barrett PHR,
clinem/dgz176. et al. Triglyceride-enrichment of HDL enhances the in vivo metabolic clear-
[25] Averill M, Rubinow KB, Cain K, Wimberger J, Babenko I, Becker JO, et al. Post- ance of HDL-apo A-1 in healthy men. J Clin Invest 1999;103:1191–9. doi:10.
prandial remodeling of high-density lipoprotein following high saturated fat 1172/JCI5286.
and high carbohydrate meals. J Clin Lipidol 2020;14:66–76. doi:10.1016/j.jacl. [50] Pont F, Duvillard L, Florentin E, Gambert P, Vergès B. High-density lipopro-
2019.11.002. tein apolipoprotein A-I kinetics in obese insulin resistant patients. An in
[26] Quintanilha BJ, Pinto Ferreira LR, Ferreira FM, Neto EC, Sampaio GR, vivo stable isotope study. Int J Obes Relat Metab Disord 2002;26:1151–8
https://doi.org/10.1038=sj.ijo.0802070.
10 M.Y. Yoshinaga, B.J. Quintanilha, A.B. Chaves-Filho et al. / Journal of Nutritional Biochemistry 97 (2021) 108809

[51] Rashid S, Watanabe T, Sakaue T, Lewis GF. Mechanisms of HDL lowering [60] Mousa A, Naderpoor N, Mellett N, Wilson K, Plebanski M, Meikle PJ, et al.
in insulin resistant, hypertriglyceridemic states: the combined effect of HDL Lipidomic profiling reveals early-stage metabolic dysfunction in overweight or
triglyceride enrichment and elevated hepatic lipase activity. Clin Biochem obese humans. Biochim Biophys Acta - Mol Cell Biol Lipids 2019;1864:335–43.
2003;36:421–9. doi:10.1016/s0009- 9120(03)00078- x. doi:10.1016/j.bbalip.2018.12.014.
[52] Vergès B. Pathophysiology of diabetic dyslipidaemia: where are we? Diabetolo- [61] Rosenson RS, Brewer HB Jr, Ansell B, Barter P, Chapman MJ, Heinecke JW, et al.
gia 2015;58:886–99. doi:10.10 07/s0 0125-015- 3525- 8. Translation of high-density lipoprotein function into clinical practice: current
[53] Camont L, Lhomme M, Rached F, Le Goff W, Nègre-Salvayre A, Salvayre R, et al. prospects and future challenges. Circulation 2013;128:1256–67. doi:10.1161/
Small, dense high-density lipoprotein-3 particles are enriched in negatively CIRCULATIONAHA.113.0 0 0962.
charged phospholipids: relevance to cellular cholesterol efflux, antioxidative, [62] Sales S, Graessler J, Ciucci S, Al-Atrib R, Vihervaara T, Schuhmann K, et al. Gen-
antithrombotic, anti- inflammatory, and antiapoptotic functionalities. Arte- der, Contraceptives and Individual Metabolic Predisposition Shape a Healthy
rioscler Thromb Vasc Biol 2013;33:2715–23. doi:10.1161/ATVBAHA.113.301468. Plasma Lipidome. Sci Rep 2016;6. https://doi.org/ 10.1038/srep27710.
[54] Darabi M, Kontush A. Phosphatidylserine in atherosclerosis. Curr Opin Lipidol [63] Berry SE, Valdes AM, Drew DA, Asnicar F, Mazidi M, Wolf J, et al. Human
2016;27:414–20. doi:10.1097/MOL.0 0 0 0 0 0 0 0 0 0 0 0 0298. postprandial responses to food and potential for precision nutrition. Nat Med
[55] Sparks DL, Frank PG, Neville TA. Effect of the surface lipid composition of 2020;26:964–73. doi:10.1038/s41591- 020- 0934- 0.
reconstituted LPA-I on apolipoprotein A-I structure and lecithin: cholesterol [64] Patsch JR, Miesenböck G, Hopferwieser T, Mühlberger V, Knapp E, Dunn JK,
acyltransferase activity. Biochim Biophys Acta 1998;1390:160–72. doi:10.1016/ et al. Relation of triglyceride metabolism and coronary artery disease - Stud-
s0 0 05-2760(97)0 0172-0. ies in the postprandial state. Arterioscler Thromb Vasc Biol 1992;12:1336–45.
[56] Boucher JG, Nguyen T, Sparks DL. Lipoprotein electrostatic properties regulate doi:10.1161/01.ATV.12.11.1336.
hepatic lipase association and activity. Biochem Cell Biol 2007;85:696–708. [65] Williams CM, Moore F, Morgan L, Wright J. Effects of n -3 fatty acids on post-
doi:10.1139/O07-137. prandial triacylglycerol and hormone concentrations in normal subjects. Br J
[57] Rached F, dos Santos R, Camont L, Miname MH, Lhomme M, Dauteuille C, Nutr 1992;68:655–66. doi:10.1079/bjn19920123.
et al. Defective functionality of HDL particles in familial apolipoprotein A-I de- [66] Lairon D, Lopez-Miranda J, Williams C. Methodology for studying postpran-
ficiency: relevance of alterations in HDL lipidome and proteome. J Lipid Res dial lipid metabolism. Eur J Clin Nutr 2007;61:1145–61. doi:10.1038/sj.ejcn.
2014;55:2509–20. doi:10.1194/jlr.M051631. 1602749.
[58] Denimal D, Nguyen A, Barros J-PP, Bouillet B, Petit J-M, Vergès B, et al. [67] Nordestgaard BG, Langsted A, Freiberg JJ. Nonfasting hyperlipidemia and
Major changes in the sphingophospholipidome of HDL in non-diabetic pa- cardio- vascular disease. Curr Drug Targets 2009;10:328–35. doi:10.2174/
tients with metabolic syndrome. Atherosclerosis 2016;246:106–14. doi:10. 138945009787846434.
1016/j.atherosclerosis.2015.12.042. [68] Kontush A, Lhomme M, Chapman MJ. Unraveling the complexities of the HDL
[59] Meikle PJ, Wong G, Barlow CK, Weir JM, Greeve MA, MacIntosh GL, et al. lipidome. J Lipid Res 2013;54:2950–63. doi:10.1194/jlr.R036095.
Plasma lipid profiling shows similar associations with prediabetes and type 2
diabetes. PLoS One 2013;8:e74341. doi:10.1371/journal.pone.0074341.