Food Research International 107 (2018) 346–352

Contents lists available at ScienceDirect

Food Research International

journal homepage: www.elsevier.com/locate/foodres

Orange juice affects acylcarnitine metabolism in healthy volunteers as T

revealed by a mass-spectrometry based metabolomics approach
⁎
Vanessa Moreiraa,1, Elisa Brasilib,c, ,1, Jarlei Fiamoncinid, Federico Marinie,f, Alfredo Micchelie,
Hannelore Danield, Jennifer Ji Hye Leeb, Neuza Mariko Aymoto Hassimottob,c,
Franco Maria Lajolob,c
a
Department of Pharmacology, Paulista School of Medicine, Federal University of São Paulo, São Paulo, Brazil
b
Department of Food Science and Experimental Nutrition, School of Pharmaceutical Science, 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
d
ZIEL Institute for Food and Health, Technische Universiät München, Munich, Germany
e
Department of Chemistry, University of Rome "La Sapienza", Rome, Italy
f
Department of Food Science, Stellenbosch University, Private Bag X1, Matieland 7602, South Africa

A R T I C L E I N F O A B S T R A C T

Keywords: Citrus juices, especially orange juice, constitute rich sources of bioactive compounds with a wide range of health-
Orange juice promoting activities. Data from epidemiological and in vitro studies suggest that orange juice (OJ) may have a
Dried blood spots positive impact on lipid metabolism. However, the effect of orange juice intake on blood lipid profile is still
Targeted metabolic profile poorly understood.
Acylcarnitines
We have used two different blood samples, Dried Blood Spots (DBS) and plasma, to assess the effect of two-
Phosphatidycholines
LC-MS
week orange juice consumption in healthy volunteers by a mass-spectrometry based metabolomics approach.
GC–MS DBS were analysed by liquid chromatography mass spectrometry (LC-MS) and plasma samples were analysed by
the gas chromatography mass spectrometry (GC–MS).
One hundred sixty-nine lipids including acylcarnitines (AC), lysophosphatidylcholines (LysoPC), (diacyl- and
acyl-alkyl-) phosphatidylcholines (PC aa and PC ae) and sphingomyelins (SM) were identified and quantified in
DBS. Eighteen fatty acids were identified and quantified in plasma. Multivariate analysis allowed to identify an
increase in C3:1, C5-DC(C6-OH), C5-M-DC, C5:1-DC, C8, C12-DC, lysoPC18:3, myristic acid, pentadecanoic acid,
palmitoleic and palmitic acid and a decrease in nervonic acid, C0, C2, C10, C10:1, C16:1, C16-OH, C16:1-OH,
C18-OH, PC aa C40:4, PC ae C38:4, PC ae C42:3, PC ae C42:4 and cholesterol levels after orange juice intake.
A two-week period of orange juice intake could affect fatty acids β-oxidation through mitochondrial and
peroxisomal pathways, leading to an increase of short-chain acylcarnitines and a decrease of medium and long-
chain acylcarnitines. This is the first report analyzing the effect of orange juice intake in healthy volunteers using
a dried blood spot-based metabolomics approach.

1. Introduction
Citrus juices, especially orange juice (OJ), constitute rich sources of
vitamin C and bioactive compounds such as flavanones hesperidin and
naringenin with a wide range of health-promoting biological activities
(Li & Schluesener, 2017).
Several studies have shown that consumption of orange juice re-
duces oxidative stress and inflammation helping in preventing athero-
sclerosis, heart failure, Alzheimer's disease and other immunological
disorders (Azzini et al., 2017; Rampersaud & Valim, 2017; Rangel-
Huerta et al., 2017). These health effects were mainly associated to a
lowering of blood lipids and lipid peroxidation, decreased oxidized LDL
and maintenance of low levels of IL-6 and C-reactive protein (Escudero-
lópez et al., 2015; Zheng et al., 2017). On the other hand, new food
biomarkers are needed to evaluate the effect of diet on health and to
check adherence to dietary recommendations and healthy eating pat-
terns. In recent years, metabolomics has emerged as a key tool in search
for novel biomarkers of citrus consumption providing new insight in
nutritional science. In particular, metabolic profiling strategies allowed
the identification of proline betaine as urinary biomarker of citrus fruit

⁎
Corresponding author.
E-mail address: elisa.brasili@uniroma1.it (E. Brasili).
1
These authors contributed equally to this work.

https://doi.org/10.1016/j.foodres.2018.02.046
Received 30 November 2017; Received in revised form 8 February 2018; Accepted 17 February 2018
Available online 21 February 2018
0963-9969/ © 2018 Elsevier Ltd. All rights reserved.
V. Moreira et al.
intake (Heinzmann et al., 2010; Lang et al., 2017; Lloyd, Beckmann,
Favé, Mathers, & Draper, 2011). Yet, an untargeted serum metabo-
lomics-driven approach was also used to predict orange juice con-
sumption. Rangel-Huerta et al. (2017) observed a decrease in serum
levels of hydroxyoctadecadienoic acid (9-HODE+13-HODE) and dihy-
droxyoctadecanoic acid (12,13-DiHOME and 9,10-DiHOME) and an
increase in 12-hydroxyeicosatetraenoic acid (12-HETE) levels after
high-polyphenol orange juice supplementation.
Recently, the analysis of dried blood spot (DBS) matrix, commonly
used in newborn screening for the diagnosis of inborn errors of meta-
bolism, has been extended and became a praticable tool for many dif-
ferent applications ranging from therapeutic drug monitoring (Li & Tse,
2010), genomics (Tarini & Goldenberg, 2012), proteomics (Chambers,
Percy, Yang, & Borchers, 2015) and metabolomics (Michopoulos, Lai,
Gika, Theodoridis, & Wilson, 2009; Wilson, 2011). Using this metho-
dology blood samples are typically obtained from heel or finger pricks
and spotted onto filter paper for analysis. DBS offers some essential
advantages over whole blood, plasma or serum samples, including a
minimally invasive and easy to perform sampling, as well as a sig-
nificantly lower volume of required sample compared to the venous
collection and a low cost of sample transport (Gao et al., 2017; Zukunft,
Sorgenfrei, Prehn, & Adamski, 2013). In addition, DBS analysis could
lead to the detection of whole blood markers that might not be present
in plasma or serum samples.
In the present study we used two different blood samples, DBS and
plasma, to assess the effect of two-week orange juice consumption on
lipid metabolism in healthy volunteers by a mass-spectrometry based
metabolomics approach. DBS were analysed by liquid chromatography
mass spectrometry (LC-MS) and plasma samples were analysed by the
gas chromatography mass spectrometry (GC–MS). Multivariate statis-
tical models were constructed to explore the relationship between the
lipids in dried blood spot and plasma and the orange juice intake.
2. Materials and methods
2.1. Chemical composition of orange juice
Orange juice obtained from local producers was used. Flavonoid
identification and quantification were performed as described in Brasili
et al. (2017). Briefly, a total of 5 mL of orange juice was centrifuged for
10 min at 4 °C and 7000 g. An aliquot of the supernatant (1 mL) was
filtered and injected (10 μL) into a 1260 Infinity Quaternary LC System
(Agilent Technologies, USA) with an autosampler and a quaternary
pump, coupled to a DAD. Elution of the analytes was achieved on a
column Prodigy 5 μm ODS3 reversed-phase silica
(250 mm × 4.60161 mm) (Phenomenex Ltd., UK) with a flow rate of
1 mL/min at 25 °C. The mobile phase consisted of 0.5% formic acid in
water (solvent A) and 0.5% formic acid in acetonitrile (solvent B). The
eluates were monitored at 270 and 525 nm. Peak identification was
carried out by the combined information provided by comparison of
retention times, diode array spectral characteristics and mass spectra,
measured by LC- ESI-MS/MS, with the internal standards and the data
available in the literature. The equipment of LC-ESI-MS/MS was a
Prominence liquid chromatograph (Shimadzu, Japan) linked to an ion
trap mass analyzer (Esquire, Bruker Daltonics, Billerica, MA, USA) with
an electrospray ionization (ESI) interface in negative mode for flavo-
noids. The mass spectrometer operating conditions were as follows:
collision energy was 4000 V and capillary temperature was 275 °C. The
analysis was carried out using a full scan from m/z 100 to 1500.
Quantification was done by calibration curves of internal standards
(Extrasynthese, Genay, France).
The content of vitamin C was evaluated through the reduction of
ascorbic acid. This compound was extracted with metaphosphoric acid
(0.3% vv−1) and quantified by reversed-phase HPLC, in a Hewlett-
Packard 1100 system, with an autosampler and a quaternary pump,
coupled to a diode array detector (DAD). The column used was μ-
347
Food Research International 107 (2018) 346–352
Bondapack (300 mm × 3.9 mm i.d., Waters, Milford, MA, USA) column
and elution (flow rate of 1.5 mL/min) was carried out in isocratic
conditions with 0.2 M sodium acetate/acetic acid buffer (pH 4.2),
monitored at 280 nm. Total ascorbic acid was estimated after reduction
of dehydroascorbic acid (DHA) with 10 mM dithiothreitol, with a ca-
libration curve of ascorbic acid (0–600 mg/L).
2.2. Antioxidant activity
The antioxidant capacity of orange juice was assessed by free ra-
dical-scavenging activity according to Brand-Williams, Cuvelier, and
Berset (1995) method and by an oxygen radical absorbance capacity
assay according to Prior et al. (2003) method. To determine the radical
scavenging activity, a 0.1 mM solution of 1,1-diphenyl-2-picrylhydrazyl
(DPPH) in methanol was prepared. An aliquot of 50 μL of deproteinated
sample was added to 150 μL of this solution. The decrease in absor-
bance was determined at 517 nm using a microplate spectrophotometer
(Benchmark Plus, Bio-Rad, Hercules, CA, USA) when the reaction
reached a plateau (after 20 min). The area under the curve was calcu-
lated. The results were expressed as millimole Trolox equivalents. To
evaluate the oxygen radical absorbance capacity, 25 μL of diluted
sample (1:300) was added to 150 μL of sodium fluorescein (40 nM) and
incubated for 15 min at 37 °C. Subsequently, 0.6 mL of AAPH (4 mM) as
a peroxyl radical generator and Trolox as a control standard were
added, and the fluorescence was read with an excitation wavelength of
485 nm and an emission wavelength of 520 nm, every 5 min for
120 min. Final results were calculated from a standard curve using
different concentrations of Trolox (12.5–100 μM) and expressed in
millimole Trolox equivalents. Measurements were recorded on a Sy-
nergy TM HT-multimode microplate reader (Biotek Instruments, Wi-
nooski, VT, USA).
2.3. Subjects and study design
Fifteen healthy volunteers between 20 and 45 years old (7 males
and 8 females) with a body mass index (BMI) of 22.5 ± 3.5 kg/m2
(mean ± SD) participated in a longitudinal study. Informed consent
was obtained from all individual participants included in the study.
Exclusion criteria included diabetes, gastrointestinal and cardiovascular
diseases, liver and kidney dysfunctions, intake of supplements or anti-
biotics, current smokers, suspected or definite history of alcohol or drug
abuse history, being pregnant. The study protocol was approved by the
Ethics Committee of the Faculty of Pharmacy at University of São Paulo
(ref: 10207012.6.0000.0067). During the study period, participants
were forbidden to consume citrus fruits other than test orange juice.
After a 7 day washout period (citrus-free diet), subjects were asked to
consume a 250 mL of orange juice, twice daily for 15 consecutive days.
A total of 8 mL of blood by venous puncture was collected in hepar-
inized tubes at the start, corresponding to 7 day after washout period
(T0) and at the end of the study (T1), period related to 15 days after
orange juice intake. Aliquots of whole blood were spotted onto Guthrie
card filter papers (Whatman no. 903 Protein Saver TM cards, formerly
Schleicher & Schuell, Keene, USA) for the DBS analysis. The Guthrie
card filter papers were left to dry for at least 4 h at room temperature
and were stored at −20 °C in a foil bag with a desiccant package
pending further analysis. The blood tubes were then centrifuged and
the resulting plasma was stored at −20 °C until further analysis.
2.4. Biochemical parameters
Concentrations of glucose, triglycerides, total cholesterol (TC), HDL
and LDL levels in plasma were determined by spectrophotometry using
commercial kits (Labtest, Brazil) adapted to a biochemical analyzer
equipment (LabMax 240, Labtest, Santa Clara, Brazil). The determina-
tion of LDL-cholesterol was calculated using the Friedwald formula.
V. Moreira et al.
2.5. Metabolic profiling
2.5.1. Metabolic profiling of DBS by LC-MS
Metabolic profiling was carried out in DBS on filter paper.
Acylcarnitines, lysophosphatidylcholines, phosphatidylcholines and
sphingomyelins were quantified in DBS using flow injection analysis
(FIA) coupled to mass spectrometry. Briefly, the metabolites were ex-
tracted from a 3 mm DBS punch (assumed to contain around 3.1 μL of
total blood) with LC-MS grade methanol containing 5 mM NH4Ac and
deuterated internal standards of acylcarnitines (Chromsystems –
Gräfelfing, Germany), glycerolipids and sphingomyelins (Avanti Polar
Lipids, USA). The punches were incubated with the solvent for 30 min
at room temperature under agitation, followed by centrifugation at
12800 g and recovery of the supernatant, which was then evaporated to
dryness and re-suspended in 100 μL of 5 mM NH4Ac methanol prior to
the injection in the mass spectrometer. This procedure allows the ex-
traction of the metabolites and precipitation of the proteins. The me-
thanolic extracts were injected (20 μL) at 35 μL/min flow into a triple
quadrupole mass spectrometer (QTRAP5500 – Sciex, Framingham, MA,
USA), coupled to a HPLC system (Agilent, Santa Clara, CA, USA) and
the mass spectra acquired in positive mode (Han, Yang, & Gross, 2012).
The mobile phase consisted of LC-MS grade methanol:H2O (96.7:3.3)
containing 5 mM NH4Ac. Peak integration was performed with the
Analyst 1.5 software (Sciex) and metabolite concentrations calculated
multiplying the ratio of metabolite: internal standard peak area by the
concentration of deuterated internal standards (Supplementary Mate-
rial). Results were expressed in μM, assuming that each punch of the
DBS contained 3.1 μL of total blood (Hall, Flores, & De Jesús, 2015).
2.5.2. Preparation and analysis of FAMEs from plasma by GC/MS
Fatty acids were extracted from plasma using Folch method (Folch,
Lees, & Stanley, 1957). Plasma was mixed with a chloroform:methanol
(2:1 v/v) solution at a final volume equivalent to 20-fold the sample
volume. After the homogenization for 15 min, samples were centrifuged
at 14000 rpm, for 10 min. The upper phase was separated and a volume
of milli-Q water (0.2%) was added. Sample was homogenized for 1 min
and centrifuged at 2000 rpm. The upper phase was collected for the
derivatization of fatty acids to methyl esters (FAMEs) according to the
protocol described by Ichihara and Fukubayashi (2010) with some
modifications. Briefly, samples were dried under a gentle flux of ni-
trogen, dissolved in 0.20 mL of toluene and transferred to a screw-
capped glass test tube. A volume of 1.5 mL of methanol and 0.30 mL of
the 8% HCl were added. The tubes were vortexed and then incubated at
100 °C for 1 h. After cooling to room temperature, 1 mL of hexane and
1 mL of milli-Q water were added. The tubes were vortexed for 1 min
and kept at room temperature for a complete separation of the two
phases. The organic phase was collected, dried under N2 stream and
dissolved in 50 μL of hexane and finally analysed by GC–MS.
The samples were injected (1 μL) in split mode (5:1) at 250 °C in the
GC/MS system (Agilent 6890 coupled to an Agilent 5973 Mass Selective
Detector). The column was a DB-5MS capillary column
(30 m × 0.25 mm inner diameter, 0.25 μm film thickness; Agilent).
Helium was used as gas carrier at the rate of 1 mL/min. The oven
temperature was programmed from 70 °C (isothermal for 0.5 min), with
an increase of 5 °C/min to 175 °C, then 2 °C/min to 280 °C for 0.5 min.
The FAMEs were identified on the basis of retention time and mass
spectrum and by comparing the retention time with their respective
standards as proposed by Sumner et al. (2007). A mass spectral survey
was performed using the NIST Mass Spectral Library (2008). The ana-
lyses were performed in triplicate for each sample and each FAME was
quantified based on curves made with authentic standards (Supelco,
Inc).
2.6. Multivariate analysis
All analyses were performed using in-house routines running under
348
Food Research International 107 (2018) 346–352
Matlab R2015b environment (The MathWorks, Natick, MA, USA).
Normality of data was checked by Shapiro-Wilk test. Baseline clinical
characteristics were assessed via Student t-test. Differences in lipids and
fatty acids compounds were evaluated via the Mann-Whitney U test. A
p-value < 0.05 was considered statistically significant.
The relationship between acylcarnitines, lysophosphatidylcholines,
phosphatidylcholines, sphingomyelins and fatty acids was explored by
constructing and validating a predictive classification model. Given the
multilevel structure of the data, since multiple points were recorded on
the same subjects, a preliminary analysis was conducted to characterize
the differences among subjects (inter-individual variation) and separate
it from the treatment-related (intra-individual variation). In order to
investigate the effect of orange juice intake in participants (intra-in-
dividual variation), a Partial least squares-discriminant analysis (PLS-
DA) classification model was constructed and validated using post-in-
tervention-baseline values. Double cross-validation was used to obtain
an unbiased estimate of the prediction error. To further validate the
classification model, the results of the cross-validation were compared
with the distributions of specific figures of merit under the null hy-
pothesis, estimated by means of permutation tests. Three figures of
merit were considered: (1) the number of misclassifications (NMC), (2)
the area under the receiver operating characteristic (ROC) curve
(AUROC), and (3) the value of the discriminant Q2 (DQ2) (Szymanska,
Saccenti, Smilde, & Westerhuis, 2012). NMC is the most intuitive of all
diagnostic statistics as it simply indicates the number of samples (or
participants, as in the present investigation), which are wrongly clas-
sified by the model. AUROC is a figure of merit borrowed from signal
processing and is particularly useful to characterize binary classifiers.
Its values range between 1 (perfect classification) and 0 (no dis-
crimination). DQ2 was introduced by Van Velzen et al. (2008) as a
modification of the standard Q2 (R2 in cross-validation) to cope with the
peculiarities of classification problems addressed by regression
methods. Like its regression analog, DQ2 assumes its highest values in
the case of a perfect discrimination between classes. Differently from
standard Q2, DQ2 can also be negative (i.e., it is not bound to the 0–1
range of values). Once the model was calculated, information about the
experimental variables more important in the discrimination was ob-
tained by inspecting the rank products (Smit et al., 2007). Metabolites
with a low rank product are typically the most important metabolites,
while metabolites with a high rank product are typically in the lower
regions of the sorted list, indicating that often other metabolites were
more important.
3. Results
3.1. Orange juice bioactive compounds and antioxidant activity
The chemical composition of orange juice was investigated through
a targeted LC-ESI-MS/MS approach. Hesperidin and narirutin were the
major flavonoid glycosides found in the selected orange juice with a
concentration equal to 44.4 ± 4.44 and 12.0 ± 0.8 mg/200 mL, re-
spectively. The levels of vicenin-2 and didymin were 6.92 ± 0.45 and
3.12 ± 0.27 mg/200 mL, respectively. The ascorbic acid content was
equal to 0.489 ± 0.01 mg/mL. The antioxidant capacity of orange
juice was evaluated using ORAC and DPPH assays. The DPPH values
were of 202.4 ± 22.9 (mM eq Trolox/200 mL) while ORAC values
were 925.4 ± 23.5 (mM eq Trolox/200 mL).
3.2. Metabolomic profile of DBS and plasma
The clinical characteristics of the volunteers are summarized in
Table 1. No statistical significance was observed in glucose, HDLeC,
LDL-C, TG and TC levels after orange juice intake.
Metabolomic profile of DBS was generated using a targeted LC-MS
approach. In total, 169 lipids were identified and quantified including
40 acylcarnitines, 18 lysophosphatidylcholines, 38 diacyl
V. Moreira et al.
Table 1
Clinical characteristics of the volunteers before and after orange juice intake.
Variable Before OJ intake After OJ intake
(Mean ± SD) (Mean ± SD)
Glucose 85.8 ± 6.9 89.9 ± 7.9
Triglycerides 76.7 ± 30.0 81.8 ± 30.0
HDL-Cholesterol 50.1 ± 14.1 48.4 ± 14.1
LDL-Cholesterol 102.1 ± 23.9 95.1 ± 25.1
Total Cholesterol 178.2 ± 25.2 173.8 ± 26.9
Mean and standard deviation (SD) of the clinical characteristics before and after orange
juice intake are shown. Value are expressed in mg/dL.
phosphatidylcholines, 38 alkyl: ethyl phosphatidylcholines and 15
sphingomyelins. Table S1 contains a list of all metabolites profiled as
well as the transmissions and deuterated internal standards used for the
quantitation.
The targeted GC–MS approach was able to identify 18 fatty acids.
The complete list of all quantified metabolites is reported in Table S2.
The LC-MS and GC–MS data were combined for the statistical analysis.
The optimal PLS-DA model was built using three LVs that accounted
for > 75% of the variance originally present in the X block. As in-
dicated by the cross-validation procedure, the model allowed to cor-
rectly predicting the effect of orange juice intake in 100.0% of parti-
cipants in the calibration phase and 99.1 ± 2.3% of participants in
double cross-validation (outer loop). The classification capacity of the
model is also evident by inspecting the projection of participants into
the space spanned by the first three LVs of the PLS-DA model (Fig. 1),
which shows a clear separation between subjects at baseline time and
after orange juice consumption. To further validate the classification
model, the results of the double cross-validation were compared with
the distributions of specific figures of merit under the null hypothesis.
The distribution of NMC, AUROC, and DQ2 under their null hypothesis,
as estimated by the permutation tests, is shown in Fig. 2. The corre-
sponding values obtained by the PLS-DA model on unpermuted data,
are also shown. The results of the PLS-DA classification model were
statistically significant, as for all of the three figures of merit p-values
equal to 0 were obtained.
In order to identify the metabolites which were mostly involved in
discriminating between baseline and treatment categories, the values of
rank products were inspected. The metabolites that contributed the
most to the separation between T0 and T1 are in Table 2. Twenty-five
variables were found to contribute significantly to the discrimination
Fig. 1. Projection of participants on to the space spanned by the first three latent vari-
ables (LVs) of the PLS-DA model. Red circles correspond to subjects at baseline, blue
circles identify subjects after two-week period orange juice intake. (For interpretation of
the references to colour in this figure legend, the reader is referred to the web version of
this article.)
349
Food Research International 107 (2018) 346–352
model, including short- and long-chain acylcarnitines, phosphati-
dylcholines, lysophosphatidylcholines and fatty acids. The sign of the
regression coefficients of the PLS-DA model is reported to indicate how
the concentration of each metabolite varies after orange juice intake.
All metabolites with positive regression coefficient had a higher con-
centration after orange juice ingestion and vice versa. Accordingly, the
levels of C3:1, C5-DC(C6-OH), C5-M-DC, C5:1-DC, C8, C12-DC, ly-
soPC18:3, myristic acid, pentadecanoic and palmitoleic acid (9-Hex-
adecenoic acid) and palmitic acid increased after orange juice ingestion
whereas the levels of nervonic acid, C0, C2, C10, C10:1, C16:1, C16-
OH, C16:1-OH, C18-OH, PC aa C40:4, PC ae C38:4, PC ae C42:3, PC ae
C42:4 and cholesterol decreased.
4. Discussion
4.1. Orange juice bioactive compounds and antioxidant activity
The chemical composition of orange juice in terms of flavonoid
compounds including hesperidin, narirutin, vicenin-2 and didymin, was
in agreement with those of the other authors (Brasili et al., 2017;
Escudero-lópez et al., 2013). Also the ascorbic acid content was within
the range of average vitamin C concentration in orange juice reported
in several studies (Escudero-lópez et al., 2016). The ORAC and DPPH
methods, based on different chemical mechanisms, are able to take into
account the wide variety and range of action of antioxidant compounds
present in orange juice. The obtained results showed an antioxidant
activity in agreement with other studies on orange juice (Klimczak,
Małecka, Szlachta, & Gliszczynśka-Sẃigło, 2007; Pellegrini et al., 2003;
Seeram et al., 2008).
4.2. Effect of orange juice intake on DBS and plasma metabolic profile
The most important metabolite highlighted by PLS-DA analysis was
nervonic acid (C24:1). This monounsatured fatty acid belongs to very
long chain fatty acids (VLCFA) and is beta-oxidized preferentially in
peroxisomes. It was found in amphipathic membrane lipids, most
abundantly in sphingolipids, but its role in maintaining health remains
poorly defined. It is unclear whether nervonic acid exerts biophysical or
biochemical properties associated with anti-inflammatory or protective
actions. Among long-chain fatty acids, myristic acid, pentadecanoic and
palmitoleic acids have been recently identified as biomarkers of dietary
fiber intake (Weitkunat et al., 2017). Previous studies have largely
demonstrated that higher concentrations of these fatty acids are asso-
ciated with a lower risk in diabetes and its underlying disorders
(Santaren et al., 2014; Weitkunat et al., 2017).
Our results suggest that a two-week period of orange juice intake
affects the peroxisomal and mitochondrial fatty acid β-oxidation.
Peroxisomal and mitochondrial β-oxidations are the main processes by
which fatty acids are oxidized by means of a sequential removal of
two‑carbon units from the acyl chain to generate acyl coenzyme A
(acyl-CoA) (Poirier, Antonenkov, Glumoff, & Hiltunen, 2006). Once in
the peroxisome, acyl-CoAs are initially oxidized until the carbon chain
is shortened to C8– or C6-CoA. After their conversion to acylcarnitines,
by carnitine octanoyltransferase (CrOT) or peroxisomal carnitine acet-
yltransferase (CrAT), these intermediates are exported from the per-
oxisome by a still unknown mechanism for further oxidation in mi-
tochondria (Violante et al., 2013). The carnitine shuttle transports long-
chain acyl-CoAs into mitochondria as their corresponding carnitine
ester (Ramsay, Gandour, & van der Leij, 2001). Here, long-chain acyl-
CoAs are converted to acylcarnitines by carnitine palmitoyltransferase
1 (CPT1), which exchanges the CoA moiety for carnitine. After pro-
duction of acylcarnitines by CPT1, the mitochondrial inner membrane
transporter carnitine acylcarnitine translocase (CACT, or SLC25A20)
transports the acylcarnitines into the mitochondrial matrix. Finally, the
enzyme CPT2 reconverts acylcarnitines back into free carnitine and
long-chain acyl-CoAs, which can then be oxidized (Schooneman, Vaz,
V. Moreira et al. Food Research International 107 (2018) 346–352

Fig. 2. Distribution of number of misclassifications (NMC), area under the ROC curve (AUROC), and discriminant Q2 (DQ2) values under their respective null hypothesis as estimated by
Permutation tests with 1,000 randomization (blue histograms) and corresponding values obtained by the PLS-DA model on unpermuted data (red lines). (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article.)

Houten, & Soeters, 2013). chain acylcarnitines such as C10, C10:1, C16-OH, C16:1-OH, C18-OH
We have observed a significant decrease of free carnitine (C0) as- after orange juice intake.
sociated to an increase of short-chain acylcarnitines such as C3:1, C5-M- To confirm these results we have calculated the C0/(C16 + C18)
DC, C5-DC(C6-OH), C5:1-DC, C8 and to a decrease of medium and long- ratio and a decrease from 96.3 to 80.3 was observed after two-week

Table 2
Significant metabolites in dried blood spots and plasma after orange juice intake.

Metabolite Rank product Sign Before OJ intake (Mean ± SD) After OJ intake (Mean ± SD) p-Value

C12-DC 2.85355 1 0.24 ± 0.03 0.25 ± 0.03 0.150

C5-DC(C6-OH) 3.60022 1 0.10 ± 0.09 0.13 ± 0.10 0.022
C5-M-DC 3.72336 1 0.16 ± 0.14 0.21 ± 0.16 0.009
Nervonic acid (C24:1) 6.27522 −1 2.48 ± 0.70 2.28 ± 0.59 0.080
C5:1-DC 6.75999 1 0.11 ± 0.10 0.14 ± 0.12 0.024
Cholesterol 7.00546 −1 0.25 ± 0.10 0.24 ± 0.11 0.404
Myristic acid 9.06343 1 3.39 ± 1.21 4.08 ± 1.67 0.038
PC aa C40:4 9.70732 −1 5.18 ± 0.90 4.22 ± 0.95 0.000
C0 9.94848 −1 32.81 ± 11.9 24.38 ± 5.86 0.008
9-Hexadecenoic acid (C16:1) 9.95045 1 7.59 ± 3.15 9.04 ± 6.13 0.068
PC ae C42:3 13.5638 −1 3.88 ± 0.60 3.22 ± 0.71 0.000
C16:1 15.6495 −1 0.13 ± 0.02 0.12 ± 0.03 0.006
C3:1 16.789 1 0.08 ± 0.02 0.08 ± 0.02 0.439
Pentadecanoic acid (C15:0) 18.3526 1 3.85 ± 0.72 4.13 ± 0.72 0.057
C8 21.2364 1 0.64 ± 0.16 0.65 ± 0.20 0.455
C10 21.9752 −1 2.44 ± 3.96 2.24 ± 3.45 0.193
C18-OH 24.4734 −1 0.11 ± 0.07 0.10 ± 0.05 0.025
C16-OH 25.4928 −1 0.10 ± 0.04 0.09 ± 0.03 0.014
PC ae C42:4 25.9587 −1 2.13 ± 0.28 1.76 ± 0.39 0.000
PC ae C38:4 26.0636 −1 8.95 ± 1.52 7.30 ± 1.40 0.000
Palmitic acid 26.5175 1 152.96 ± 30.0 170.87 ± 66.7 0.105
C10:1 34.4434 −1 0.38 ± 0.08 0.35 ± 0.07 0.010
C16:1-OH 35.1083 −1 0.13 ± 0.06 0.12 ± 0.04 0.063
LysoPC a C18:3 37.0876 1 0.33 ± 0.11 0.34 ± 0.17 0.386
C2 41.201 −1 8.03 ± 1.57 6.82 ± 1.61 0.005

Metabolites were ranked in descendant order of the rank products that was determined by cross validation. Mean and standard deviation (SD) of the metabolites before and after orange
juice intake are shown. For clarity, the p-value are bold where they are also significant at Mann-Whitney U test. Acylcarnitines, phosphatidylcholines and cholesterol are expressed in μM,
fatty acids are expressed in mM.

350
V. Moreira et al.
orange juice intake. An elevated ratio between free carnitine (C0) and
the sum of palmitoylcarnitine and stearoylcarnitine (C16 + C18) in
DBS samples was previously proposed as diagnostic criteria to identify
CPT − 1 deficiency (De Sain-van der Velden et al., 2013). In effect,
acylcarnitines have a long history in the diagnosis and neonatal
screening of fatty acid oxidation defects and other inborn errors of
metabolism (Rinaldo, Cowan, & Matern, 2008). In addition, the acyl-
carnitines accumulation has been positively associated to insulin re-
sistance (Schooneman et al., 2013).
Our results suggest that OJ intake affects liver fatty acids β-oxida-
tion, both mitochondrial and peroxisomal β-oxidation, decreasing
medium and long-chain acylcarnitines and increasing short-chain
acylcarnitines blood levels.
Although the mitochondria have the same enzyme machinery to
metabolize all the double bonds of polyunsaturated fatty acids, ex-
periments with isolated organelles have demonstrated that poly-
unsaturated fatty acids are well oxidized in peroxisomes and slowly
oxidized in mitochondria (Hiltunen, Kärki, Hassinen, & Osmundsen,
1986).
In addition, two weeks of orange supplementation lead to a decrease
in levels of choline linked to alkyl-acyl groups or polyunsaturated
phosphatidyl choline (Choline-EPL) such as PC ae C38:4, PC ae C42:3,
PC ae C42:4, suggesting that the effect of the orange juice intake could
be more complex, involving the regulation system of the liver perox-
isomal metabolism. The first two enzymes of the EPL biosynthetic
pathway, dihydroxyacetone phosphate acyltransferase (DHAPAT) and
alkyl-dihydroxyacetone phosphate synthase (alkylDHAP), are mainly
localized in peroxisomes (Wanders & Brites, 2010). The increased
synthesis of medium-chain carnitines associated with the decrease in
long-chain carnitines and EPLs supports the hypothesis of an effect
induced by OJ intake on the regulation of EPL synthesis and fatty acid
metabolism in peroxisomes, even though a contribution of branched
chain amino acid catabolism in the production of C5-acylcarnitines
cannot be excluded.
Indeed, previous studies demonstrated that a diet characterized by
higher intake of vegetables and fruits was positively associated with
higher concentrations of C5-DC/C6-OH (Bouchard-Mercier,
Rudkowska, Lemieux, Couture, & Vohl, 2013).
More evidences suggested that dietary supplementation of Citrus
extracts influenced fatty acid oxidation via upregulation of phosphor-
ylation levels of AMP-activated protein kinase (AMPK), acetyl-CoA
carboxylase (ACC) and via peroxisome proliferator-activated receptors
(PPARs) pathway (Huong, Takahashi, & Ide, 2006). It is well known
that PPARs are nuclear hormone receptors that mediate the carbohy-
drate, fatty acid and lipid metabolism of tissues and cells at the tran-
scriptional level (Kota, Huang, & Roufogalis, 2005; Varga, Czimmerer,
& Nagy, 2011). In particular, the genes encoding the β-oxidation
pathway are transcriptionally regulated by PPARs. Several studies have
suggested that the inhibitory effect of citrus flavonoids on adipogenesis
and adiposity can be partially attributed to regulation of the PPAR
expression levels both in cell and animal models. In this context,
Fukuchi et al. (2008) have demonstrated that supplementation with
lemon polyphenols suppressed body weight gain and body fat accu-
mulation, by increasing peroxisomal β-oxidation through up-regulation
of the mRNA level of the acyl-CoA oxidase enzyme that is involved in
peroxisomal fatty acid oxidation in the liver and white adipose tissue.
This process was likely mediated via up-regulation of the mRNA levels
of PPARα.
One of the major components of orange juice is hesperidin (3,5,7-
trihydroxyflavanone 7- rhamnoglucoside), an abundant flavanone gly-
coside in Citrus fruits. It has been demonstrated that hesperidin might
be effective in treating a great variety of diseases, including cardio-
vascular diseases and neurological disorders, due to its antiin-
flammatory, antioxidant, lipid-lowering and insulin-sensitizing prop-
erties (Li & Schluesener, 2017).
Some studies evidenced that hesperidin targets peroxisome
351
Food Research International 107 (2018) 346–352
proliferator-activated receptor-gamma (PPAR-γ) to exert biological ac-
tions (Agrawal et al., 2014; Salam et al., 2008). PPAR-γ is the nuclear
receptor for 15D–prostaglandin J2, which derives from PGH2, a cy-
clooxygenase 2-dependent metabolite of arachidonic acid.
It is important to highlight that an inverse correlation between or-
ange juice intake and concentrations of plasmatic prostaglandins has
been also observed by Sánchez-Moreno et al., 2003. Certainly, the
measurement of prostaglandins and other eicosanoids in these experi-
mental conditions could help to support these evidences.
5. Conclusions
In this study, for the first time, two different blood samples, DBS and
plasma, were used to assess the effect of two-week orange juice con-
sumption in healthy volunteers by a mass-spectrometry based meta-
bolomics approach. The observed metabolic variations suggest that
orange juice stimulates and regulates the mitochondrial and perox-
isomal fatty acid β-oxidation at liver compartment, as shown by of the
changes in short-, medium- and long-chain acylcarnitines. This effect of
orange juice could have an important impact on some metabolic dis-
eases, such as obesity and liver steatosis.
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.foodres.2018.02.046.
Acknowledgements
This work was supported by the São Paulo Research Foundation
(FAPESP) (Grant 2013/07914-8), the National Council of Technological
and Scientific Development (CNPq Grants 150905/2012-2 and 150033/
2015-0) and Coordination for the Improvement of Higher Education
Personnel (CAPES Grant 88887.091278/2014-00) for their financial
support.
Compliance with ethical standardConflict of interest
The authors declare that they have no conflict of interest.Ethical
approval
All procedures performed in study were in accordance with the
ethical standards of the Faculty of Pharmacy at University of São Paulo
committee and with the 1964 Helsinki declaration and its later
amendments or comparable ethical standards.
References
Agrawal, Y. O., Sharma, P. K., Shrivastava, B., Ojha, S., Upadhya, H. M., Arya, D. S., &
Goyal, S. N. (2014). Hesperidin produces cardioprotective activity via PPAR-γ
pathway in ischemic heart diseasemodel in diabetic rats. PLoS One, 9, e111212.
Azzini, E., Venneria, E., Ciarapica, D., Foddai, M. S., Intorre, F., Zaccaria, M., ... Polito, A.
(2017). Effect of red orange juice consumption on body composition and nutritional
status in overweight/obese female: A pilot study. Oxidative Medicine and Cellular
Longevity, 2017, 1672567.
Bouchard-Mercier, A., Rudkowska, I., Lemieux, S., Couture, P., & Vohl, M. C. (2013). The
metabolic signature associated with the western dietary pattern: A cross-sectional
study. Nutrition Journal, 12, 158.
Brand-Williams, W., Cuvelier, M. E., & Berset, C. (1995). Use of a free radical method to
evaluate antioxidant activity. LWT - Food Science and Technology, 28, 25−30.
Brasili, E., Seixas Chaves, D. F., Xavier, A. A., Mercadante, A. Z., Hassimotto, N. M. A., &
Lajolo, F. M. (2017). Effect of pasteurization on flavonoids and carotenoids in Citrus
sinensis (L.) Osbeck cv. Cara Cara and Bahia juices. Journal of Agricultural and Food
Chemistry, 65, 1371–1377.
Chambers, A. G., Percy, A. J., Yang, J., & Borchers, C. H. (2015). Multiple reaction
monitoring enables precise quantification of 97 proteins in dried blood spots.
Molecular & Cellular Proteomics, 14, 3094–3104.
De Sain-van der Velden, M. G., Diekman, E. F., Jans, J. J., van der Ham, M., Prinsen, B. H.,
Visser, G., & Verhoeven-Duif, N. M. (2013). Differences between acylcarnitine pro-
files in plasma and bloodspots. Molecular Genetics and Metabolism, 110, 116–121.
Escudero-lópez, B., Berná, G., Ortega, Á., Herrero-martín, G., Cerrillo, I., Martín, F., &
Fernández-pachón, M. (2015). Consumption of orange fermented beverage reduces
cardiovascular risk factors in healthy mice. Food and Chemical Toxicology, 78, 78–85.
Escudero-lópez, B., Cerrillo, I., Gil-izquierdo, Á., Herrero-martín, G., Berná, G., Medina,
S., ... Mart, F. (2016). Effect of thermal processing on the profile of bioactive com-
pounds and antioxidant capacity of fermented orange juice. International Journal of
V. Moreira et al.
Food Sciences and Nutrition, 67, 779–788.
Escudero-lópez, B., Cerrillo, I., Herrero-martín, G., Hornero-me, D., Gil-izquierdo, A.,
Medina, S., ... Martín, F. (2013). Fermented orange juice: Source of higher carotenoid
and flavanone contents. Journal of Agricultural and Food Chemistry, 61, 8773–8782.
Folch, J., Lees, M., & Stanley, G. H. S. (1957). A simple method for the isolation and
purification of total lipides from animal tissues. The Journal of Biological Chemistry,
226, 497–509.
Fukuchi, Y., Hiramitsu, M., Okada, M., Hayashi, S., Nabeno, Y., Osawa, T., & Naito, M.
(2008). Lemon polyphenols suppress diet-induced obesity by up-regulation of mRNA
levels of the enzymes involved in beta-oxidation in mouse white adipose tissue.
Journal of Clinical Biochemistry and Nutrition, 43, 201–209.
Gao, F., McDaniel, J., Chen, E. Y., Rockwell, H. E., Drolet, J., Vishnudas, V. K., ... Kiebish,
M. A. (2017). Dynamic and temporal assessment of human dried blood spot MS/
MSALL shotgun lipidomics analysis. Nutrition & Metabolism (London), 14, 28.
Hall, E. M., Flores, S. R., & De Jesús, V. R. (2015). Influence of hematocrit and total-spot
volume on performance characteristics of dried blood spots for newborn screening.
International Journal of Neonatal Screening, 1, 69–78.
Han, X., Yang, K., & Gross, R. W. (2012). Multi-dimensional mass spectrometry-based
shotgun lipidomics and novel strategies for lipidomic analyses. Mass Spectrometry
Reviews, 31, 134–178.
Heinzmann, S. S., Brown, I. J., Chan, Q., Bictash, M., Dumas, M., Kochhar, S., ...
Nicholson, J. K. (2010). Metabolic profiling strategy for discovery of nutritional
biomarkers: Proline betaine as a marker of citrus consumption. The American Journal
of Clinical Nutrition, 92, 436–443.
Hiltunen, J. K., Kärki, T., Hassinen, I. E., & Osmundsen, H. (1986). Beta-oxidation of
polyunsaturated fatty acids by rat liver peroxisomes. A role for 2,4-dienoyl-coenzyme
a reductase in peroxisomal beta-oxidation. The Journal of Biological Chemistry, 261,
16484–16493.
Huong, D. T., Takahashi, Y., & Ide, T. (2006). Activity and mRNA levels of enzymes
involved in hepatic fatty acid oxidation in mice fed citrus flavonoids. Nutrition, 22,
546–552.
Ichihara, K., & Fukubayashi, Y. (2010). Preparation of fatty acid methyl esters for gas-
liquid chromatography. Journal of Lipid Research, 51, 635–640.
Klimczak, I., Małecka, M., Szlachta, M., & Gliszczynśka-Sẃigło, A. (2007). Effect of sto-
rage on the content of polyphenols, vitamin C and the antioxidant activity of orange
juices. Journal of Food Composition and Analysis, 20, 313–322.
Kota, B. P., Huang, T. H., & Roufogalis, B. D. (2005). An overview on biological me-
chanisms of PPARs. Pharmacological Research, 51, 85–94.
Lang, R., Lang, T., Bader, M., Beusch, A., Schlagbauer, V., & Hofmann, T. (2017). High-
throughput quantitation of proline betaine in foods and suitability as a valid bio-
marker for citrus consumption. Journal of Agricultural and Food Chemistry, 65,
1613–1619.
Li, C., & Schluesener, H. (2017). Health-promoting effects of the citrus flavanone. Critical
Reviews in Food Science and Nutrition, 57, 613–631.
Li, W., & Tse, F. L. (2010). Dried blood spot sampling in combination with LC-MS/MS for
quantitative analysis of small molecules. Biomedical Chromatography, 249, 49–65.
Lloyd, A. J., Beckmann, M., Favé, G., Mathers, J. C., & Draper, J. (2011). Proline betaine
and its biotransformation products in fasting urine samples are potential biomarkers
of habitual citrus fruit consumption. The British Journal of Nutrition, 106, 812–824.
Michopoulos, F., Lai, L., Gika, H., Theodoridis, G., & Wilson, I. (2009). UPLC-MS-based
analysis of human plasma for metabonomics using solvent precipitation or solid
phase extraction. Journal of Proteome Research, 8, 2114–2121.
Pellegrini, N., Serafini, M., Colombi, B., Del Rio, D., Salvatore, S., Bianchi, M., &
Brighenti, F. (2003). Total antioxidant capacity of plant foods, beverages and oils
consumed in Italy assessed by three different in vitro assays. Journal of Nutrition, 133,
2812–2819.
Poirier, Y., Antonenkov, V. D., Glumoff, T., & Hiltunen, J. K. (2006). Peroxisomal beta-
oxidation—a metabolic pathway with multiple functions. Biochimica et Biophysica
Acta, 1763, 1413–1426.
Prior, R. L., Hoang, H., Gu, L., Wu, X., Bacchiocca, M., Howard, L., ... Jacob, R. (2003).
Assays for hydrophilic and lipophilic antioxidant capacity (oxygen radical absor-
bance capacity (ORACFL)) of plasma and other biological and food samples. Journal
of Agricultural and Food Chemistry, 51, 3273–3279.
Rampersaud, G. C., & Valim, M. F. (2017). 100% citrus juice: Nutritional contribution,
dietary benefits, and association with anthropometric measures. Critical Reviews in
352
Food Research International 107 (2018) 346–352
Food Science and Nutrition, 57, 129–140.
Ramsay, R. R., Gandour, R. D., & van der Leij, F. R. (2001). Molecular enzymology of
carnitine transfer and transport. Biochimica et Biophysica Acta, 1546, 21–43.
Rangel-Huerta, O. D., Aguilera, C. M., Perez-de-la-Cruz, A., Vallejo, F., Tomas-Barberan,
F., Gil, A., & Mesa, M. D. (2017). A serum metabolomics driven approach predicts
orange juice consumption and its impact on oxidative stress and inflammation in
subjects from the BIONAOS study. Molecular Nutrition & Food Research, 61. http://dx.
doi.org/10.1002/mnfr.201600120.
Rinaldo, P., Cowan, T. M., & Matern, D. (2008). Acylcarnitine profile analysis. Genetics in
Medicine, 10, 151–156.
Salam, N. K., Huang, T. H., Kota, B. P., Kim, M. S., Li, Y., & Hibbs, D. E. (2008). Novel
PPAR-gamma agonists identified from a natural product library: A virtual screening,
induced-fit docking and biological assay study. Chemical Biology & Drug Design, 71,
57–70.
Sánchez-Moreno, C., Cano, M. P., de Ancos, B., Plaza, L., Olmedilla, B., Granado, F., &
Martín, A. (2003). Effect of orange juice intake on vitamin C concentratios and
biomarkers of antioxidant status in human. American Journal of Clinical Nutrition, 78,
454–460.
Santaren, I. D., Watkins, S. M., Liese, A. D., Wagenknecht, L. E., Rewers, M. J., Haffner, S.
M., ... Hanley, A. J. (2014). Serum pentadecanoic acid (15:0), a short-term marker of
dairy food intake, is inversely associated with incident type 2 diabetes and its un-
derlying disorders. The American Journal of Clinical Nutrition, 100, 1532–1540.
Schooneman, M. G., Vaz, F. M., Houten, S. M., & Soeters, M. R. (2013). Acylcarnitines:
Reflecting or inflicting insulin resistance? Diabetes, 62(1), 1–8.
Seeram, N. P., Aviram, M., Zhang, Y., Henning, M., Feng, L., Dreher, M., & Heber, D.
(2008). Comparison of antioxidant potency of commonly consumed polyphenol-rich
beverages in the United States. Journal of Agricultural and Food Chemistry, 56,
1415−1422.
Smit, S., van Breemen, M. J., Hoefsloot, H. C. J., Smilde, A. K., Aerts, J. M. F. G., & de
Koster, C. G. (2007). Assessing the statistical validity of proteomics based biomarkers.
Analytica Chimica Acta, 592, 210–217.
Sumner, L. W., Amberg, A., Barrett, D., Beale, M. H., Beger, R., Daykin, C. A., ...
Hankemeier, T. (2007). Proposed minimum reporting standards for chemical ana-
lysis. Metabolomics, 3, 211–221.
Szymanska, E., Saccenti, E., Smilde, A. K., & Westerhuis, J. A. (2012). Double-check:
Validation of diagnostic statistics for PLS-DA models in metabolomics studies.
Metabolomics, 8, 3–16.
Tarini, B. A., & Goldenberg, A. J. (2012). Ethical issues with newborn screening in the
genomics era. Annual review of genomics and human genetics era. 13. Annual review of
genomics and human genetics era (pp. 381–393).
Van Velzen, E. J. J., Westerhuis, J. A., Van Duynhoven, J. P. M., Van Dorsten, F. A.,
Hoefsloot, H. C. J., Jacobs, D. M., ... Smilde, K, A. (2008). Multilevel data analysis of a
crossover designed human nutritional intervention study. Journal of Proteome
Research, 7, 4483–4491.
Varga, T., Czimmerer, Z., & Nagy, L. (2011). PPARs are a unique set of fatty acid regu-
lated transcription factors controlling both lipidmetabolism and inflammation.
Biochimica et Biophysica Acta, 1812, 1007–1022.
Violante, S., Ijlst, L., Ruiter, J., Koster, J., van Lenthe, H., Duran, M., ... Ventura, F. V.
(2013). Substrate specificity of human carnitine acetyltransferase: Implications for
fatty acid and branched-chain amino acid metabolism. Biochimica et Biophysica Acta,
1832, 773–779.
Wanders, R. J., & Brites, P. (2010). Biosynthesis of ether-phospholipids including plas-
malogens, peroxisomes and human disease: New insights into an old problem. Journal
of Clinical Lipidology, 5(3), 379–386.
Weitkunat, K., Schumann, S., Nickel, D., Hornemann, S., Petzke, K. J., Schulze, M. B., ...
Klaus, S. (2017). Odd-chain fatty acids as a biomarker for dietary fiber intake: A novel
pathway for endogenous production from propionate. The American Journal of Clinical
Nutrition, 105, 1544–1551.
Wilson, I. (2011). Global metabolic profiling (metabonomics/metabolomics) using dried
blood spots:Advantages and pitfalls. Bioanalysis, 3, 2255–2257.
Zheng, J., Zhou, Y., Li, S., Zhang, P., Zhou, T., Xu, D., & Li, H. (2017). Effects and me-
chanisms of fruit and vegetable juices on cardiovascular diseases. International
Journal of Molecular Sciences, 18 (pii, E555).
Zukunft, S., Sorgenfrei, M., Prehn, C. M. G., & Adamski, J. (2013). Targeted metabolomics
of dried blood spot extracts. Chromatographia, 76, 1295–1305.