Metabolomic assessment of the effect of dietary cholesterol in the progressive development of fatty liver disease.

Abstract

Nonalcoholic fatty liver disease is considered to be the hepatic manifestation of metabolic syndrome and is usually related to high-fat, high-cholesterol diets. With the rationale that the identification and quantification of metabolites in different metabolic pathways may facilitate the discovery of clinically accessible biomarkers, we report the use of (1)H NMR metabolomics for quantitative profiling of liver extracts from LDLr(-/-) mice, a well-documented mouse model of fatty liver disease. A total of 55 metabolites were identified, and multivariate analyses in a diet- and time-comparative strategy were performed. Dietary cholesterol increased the hepatic concentrations of cholesterol, triglycerides, and oleic acid but also decreased the [PUFA/MUFA] ratio as well as the relative amount of long-chain polyunsaturated fatty acids in the liver. This was also accompanied by variations of the hepatic concentration of taurine, glutathione, methionine, and carnitine. Heat-map correlation analyses demonstrated that hepatic inflammation and development of steatosis correlated with cholesterol and triglyceride NMR derived signals, respectively. We conclude that dietary cholesterol is a causal factor in the development of both liver steatosis and hepatic inflammation.

Full text

Page 1

Metabolomic Assessment of the Effect of Dietary Cholesterol in the
Progressive Development of Fatty Liver Disease
Maria Vinaixa,*,†Miguel A´ngel Rodrı ´guez,†Anna Rull,‡Rau ´l Beltra ´n,‡Cinta Blade ´,†
Jesu ´s Brezmes,†Nicolau Can ˜ellas,†Jorge Joven,‡and Xavier Correig†
Metabolomics Platform, CIBER de Diabetes y Enfermedades Metabo ´licas Asociadas (CIBERDEM), IISPV, Universitat
Rovira i Virgili, Avda. Paı ¨sos Catalans 26, 43007 Tarragona, Spain, and Centre de Recerca Biome `dica, Hospital
Universitari Sant Joan de Reus, IISPV, Universitat Rovira i Virgili, c/Sant Joan s/n, 43201 Reus, Spain
Received December 28, 2009
Nonalcoholic fatty liver disease is considered to be the hepatic manifestation of metabolic syndrome
and is usually related to high-fat, high-cholesterol diets. With the rationale that the identification and
quantification of metabolites in different metabolic pathways may facilitate the discovery of clinically
accessible biomarkers, we report the use of1H NMR metabolomics for quantitative profiling of liver
extracts from LDLr-/-mice, a well-documented mouse model of fatty liver disease. A total of 55
metabolites were identified, and multivariate analyses in a diet- and time-comparative strategy were
performed. Dietary cholesterol increased the hepatic concentrations of cholesterol, triglycerides, and
oleic acid but also decreased the [PUFA/MUFA] ratio as well as the relative amount of long-chain
polyunsaturated fatty acids in the liver. This was also accompanied by variations of the hepatic
concentration of taurine, glutathione, methionine, and carnitine. Heat-map correlation analyses
demonstrated that hepatic inflammation and development of steatosis correlated with cholesterol and
triglyceride NMR derived signals, respectively. We conclude that dietary cholesterol is a causal factor
in the development of both liver steatosis and hepatic inflammation.
Keywords: dietary cholesterol •1H NMR spectroscopy • liver steatosis • LDLr-deficient mice • metabolomics
quantitative profiling • NASH
1. Introduction
The liver is the major metabolic organ that performs a variety
of biochemical functions necessary for whole-body metabolic
homeostasis. It is sensitive to many pathological insults leading
to an array of different clinical signs and symptoms. In
particular, nonalcoholic fatty liver disease (NAFLD) is the most
common cause of liver dysfunction characterized by fatty
infiltration of the liver in the absence of alcohol abuse. NAFLD
is currently considered the hepatic manifestation of metabolic
syndrome and other related metabolic derangements such as
obesity, insulin resistance, hypertension, and dyslipidemia.1-3
NAFLD ranges from simple steatosis to nonalcoholic steato-
hepatitis (NASH), in which inflammation is present.4-7Al-
though steatosis is considered a relatively benign and reversible
condition, the progression from steatosis toward NASH rep-
resents a critical step in the progression toward more harmful
conditions including fibrosis, cirrhosis, or liver cancer.1,8
The actual risk factors that drive hepatic inflammation during
the progression from steatosis to NASH remain unknown,
although inflammatory, oxidative, and infectious cellular insults
have been previously implicated.6-9Experimental studies in
several animal models of hepatic steatosis demonstrate a direct
relationship between dyslipidemia and dietary cholesterol with
the development of liver injury.10,11It has been generally
accepted that hepatic steatosis is the critical first step and
prerequisite for development of hepatic inflammation.6How-
ever, recent evidence have raised doubts about this hypothesis.
Dietary cholesterol, rather than hepatic steatosis, may be a risk
factor for NASH development in hyperlipidemic mouse mod-
els.12Therefore, the characterization of the metabolic events
in the etiology of progression from steatosis to NASH might
provide a better knowledge-base to understand the exact
mechanisms underlying such progression and to predict as well
as prevent further complications.
In this regard, metabolomics is a highly valuable tool because
it can provide a holistic evaluation of low molecular weight
compounds present in hepatic tissue of animal models indicat-
ing a particular biochemical phenotype.13The utility of the so-
called fingerprinting NMR-based metabolomics approach in
the analysis of liver tissue extracts has already been demons-
trated.13-18However, most of these works report their findings
upon a statistical comparison of samples based on spectral
patterns or statistically relevant NMR spectral signatures.
Currently, there is an emerging and growing preference for
using quantitative metabolic profiling as it offers a number of
important advantages over existing fingerprinting methods.14
Profiling studies can be either holistic or targeted and quantita-
* To whom correspondence should be addressed. Maria Vinaixa. E-mail:
maria.vinaixa@urv.cat. Tf.: +34-977256587. Fax: +34-977559605. Website:
www.metabolomicsplatform.com.
†CIBER de Diabetes y Enfermedades Metabo ´licas Asociadas.
‡Hospital Universitari Sant Joan de Reus.
10.1021/pr901203w
 2010 American Chemical Society
Journal of Proteome Research 2010, 9, 2527–2538 2527
Published on Web 04/19/2010

Page 2

tive or semiquantitative. A targeted metabolic profiling focuses
on a group or category of metabolites of interest defined a
priori. It involves the quantification or semi- quantification of
such group of metabolites using an analytical methodology
specifically developed for such purposes. When no a priori
assumptions on metabolites describing the biological phenom-
ena under study are made, a holistic or global profiling
methodology can be attempted. Then, data are acquired by
high-throughput generic analytical methods under certain
analytical conditions. When referring to NMR spectroscopy,
holistic quantitative profiling is based on the deconvolution
of resonances in each NMR spectrum into a list of metabolites
and their associated concentrations obtained by properly
matching and fitting the reference peaks to the sample
peaks.14,19,20
In the current study, we have performed a NMR-based
metabolomics study to investigate the metabolic effects of the
dietary cholesterol in the etiology of progression from hepatic
steatosis to NASH. A comparative study based on a nutritional
intervention in male LDLr-/-mice on a C57BL/6J background
was designed according to previous data.11,13,21At 10 weeks of
age, animals were assigned to three dietary groups, chow diet,
high-fat diet and high-fat, high-cholesterol diet. At baseline (10
weeks of age), 16- and 32-week time point variables were
analyzed and the metabolic rearrangements in the diet-and
time-dependent progression from steatosis to NASH were
assessed in liver tissue extracts using1H NMR spectroscopy.13
We have attempted a holistic metabolomics quantitative profil-
ing study performing a quantification of individual metabolites
present in aqueous and lipidic liver extracts analyzed by1H
NMR spectroscopy. In addition, we have submitted this
quantification to unsupervised multivariate analysis obtaining
a combination of biochemical markers that characterize the
metabolome changes in the progressive development of fatty
liver disease induced by the addition of dietary cholesterol. We
reasoned that the identification and quantification of metabo-
lites might facilitate downstream pathways and network analy-
ses and, therefore, may lead to drawing specific conclusions
and formulating testable hypothesis.
2. Materials and Methods
2.1. Experimental Animals and Laboratory Procedures. All
animal studies were carried out under appropriate guidelines
according to the code of conduct of Animal Care Committee
of the Universitat Rovira i Virgili and previously published
results.21Mice were housed at constant room temperature (22
°C), air humidity (55%), and a light/dark cycle of 12 h. Water
and food (14% protein rodent maintenance diet, Harlan,
Barcelona, Spain) were given ad libitum.
A comparative study was designed with male LDLr-/-mice
on a C57BL/6J background, which were the progeny of mice
purchased from the Jackson Laboratory (Bar Harbor, ME). The
animals were allocated to experimental groups by computer-
generated randomization schedules and investigators respon-
sible for the assessment of outcomes had no knowledge of the
experimental group to which the animals belonged. Mice were
maintained under a normal chow diet for 10 weeks. At 10 weeks
of age, four animals were sacrificed, and plasma and liver tissue
samples were obtained. The remaining littermates were as-
signed to three dietary groups (n ) 8 each) and maintained
under the same regime condition until 16 and 32 weeks of age
where plasma and liver tissue samples were taken in four
animals from each group. Chow diet (considered the control
diet) is the same as the maintenance diet (3% fat by weight,
0.03% cholesterol). The second diet, which was high-fat, was
commercially prepared with the addition of palm oil to diet 1,
providing a source of fat (20% w/w) that provides equal
amounts of saturated (42.5%) and monounsaturated (43.4%)
fat (PF 1973, Mucedola, Harlan, Barcelona, Spain). Finally, a
third diet was prepared with the addition of cholesterol (0.25%
w/w) to diet 2 to resemble some of the characteristics of the
Western-type diet (PF 1973/A, Mucedola, Harlan, Barcelona,
Spain). Diets were prepared and labeled by an independent
investigator according to the randomization schedule to ensure
allocation concealment. There were no animals excluded from
analysis.
Body weight was monitored weekly. Plasma cholesterol and
triglyceride concentration were determined by standard labora-
tory procedures.10The liver-tissue content of cholesterol and
triglycerides was determined after lipid extraction with isopro-
pyl alcohol-hexane method.22Liver histological examination
was performed to measure the lipid droplet content (percentage
of lipid droplet area/cell) using the AnaliSYS system (Soft
Imaging System, Mu ¨nster, Germany). The amount of inflam-
mation and the degree of steatosis were estimated as previously
described.22,23Oil Red O stains were prepared using reagents
from Sigma-Aldrich (St. Louis, MO). The hepatic proportion
of macrophages (anti-F4/80, Serotec, Oxford, UK) was deter-
mined by immunohistochemistry. For each individual mouse,
30 fields from each of three different sections were analyzed.21
For the assessment of atherosclerosis, hearts and regions 2 mm
below the ascending aorta were removed, cut transversely,
embedded in OCT, and immediately flash-frozen. Aortic valve
leaflets were used as an anatomic reference point. Pathological
evaluation and quantitative measurements were performed as
previously described24using the image analysis software
AnaliSYS.
2.2. Liver Extraction Procedures for
Metabolomics Assays. Liver extraction was performed following
the procedure described by Shi et al.25with slight modifica-
tions.13A portion of hepatic tissue (∼50 mg) was removed,
flash-frozen, and mechanically homogenized using the Pre-
cellys24 instrument in 1 mL of H2O/CH3CN (1/1). The homo-
genates were centrifuged at 5000× g for 15 min at 4 °C.
Supernatants (hydrophilic metabolites) and pellets (lipophilic
metabolites) were separately lyophilized overnight to remove
water for NMR experiments and stored at -80 °C until further
analysis. For NMR measurements, the hydrophilic extracts were
reconstituted in 600 µL D2O containing 0.67 mM trisilylpropi-
onic acid (TSP). The lipophilic extracts were subsequently
extracted in 700 µL of a solution CDCl3/CD3OD (2:1) containing
1.18 mM tetramethylsilane (TMS) and then vortexed, homog-
enized for 20 min, centrifuged for 15 min at 6000× g at room
temperature, and transferred into 5-mm NMR glass tubes.
2.3.
NMR spectra were measured at a 600.20 MHz frequency using
an Avance III-600 Bruker spectrometer equipped with an
inverse TCI 5 mm cryoprobe. For 1D aqueous extract spectra,
a one-dimensional (1D) nuclear Overhauser effect spectroscopy
with a spoil gradient (noesygppr1d) was used. Solvent presatu-
ration with low irradiation power (10 Hz) was applied during
recycling delay and mixing time (tm ) 100 ms) to suppress
residual water. A total of 256 transients of 12 kHz of spectral
width were collected at 300 K into 64 k data points, and
exponential line broadening of 0.3 Hz was applied before
Fourier transformation. A recycling delay time of 8 s was
1H NMR Based
1H NMR Measurements. One- and two-dimensional1H
research articles
Vinaixa et al.
2528Journal of Proteome Research • Vol. 9, No. 5, 2010

Page 3

applied between scans to ensure correct quantification. In the
case of lipophilic extracts, a 90° pulse with presaturation
sequence (zgpr) was used. We performed measurements at 287
K, shifting the residual water signal to 4.65 ppm to allow the
quantification of the characteristic glycerol-backbone signals.
Besides, residual water was presaturated during recycling delay
(RD ) 8 s) using a low irradiation power (10 Hz). A total of 256
FIDs of 12 kHz of spectral width were collected into 64 k data
points and exponential line broadening of 0.3 Hz was applied
before Fourier transformation. The frequency spectra were
phased, baseline corrected, and then calibrated (TMS or TSP,
0.0 ppm) using TopSpin software (version 2.1, Bruker).
2.4. Metabolite Identification and Quantification. Reso-
nance assignments were done on the basis of literature
values26-33and different database search engines (BBioref
AMIX database, Bruker and HMDB). Chemical shifts were
identified as described elsewhere.27,29,30,34Both a two-dimen-
sional (2D)-1H,13C- HSQC (heteronuclear single quantum cor-
relation) and a two-dimensional (2D)-1H-13H COSY (correlation
spectroscopy) were used for structural confirmation. The one-
dimensional S-TOCSY (Statistical Total Correlation Spectros-
copy) approach35was also used to elucidate some of these
assignments.
After baseline correction, selected peaks in the 1D-NMR
spectra were integrated using the AMIX 3.8 software package
(Bruker, GmBH). The absolute concentration of single metabo-
lites either in water-soluble or lipid extracts was assessed
according to the methodology described by Serkova et al.36The
integral at 0.87 corresponding to ω-CH3 was used as a reference
for total fatty acid chains to estimate the molar percentage of
fatty acid signals.
A list of metabolites identified in
aqueous liver extracts that have been integrated and further
used to assess metabolic rearrangements produced by progres-
sive development of fatty liver disease is summarized in Table
1.
2.5. Data Processing and Multivariate Analysis. Absolute
concentrations derived from both lipophilic and hydrophilic
extracts were arranged together in one single data matrix, which
was used as the input matrix for the PCA multivariate model.
Previously, data were scaled to unit variance to give all the
identified metabolites the same opportunity to enter to the
model. Data (pre-) processing, data analysis, and statistical
calculations were performed with Matlab (Matlab version 6.5.1,
Release 13, The Mathworks, 2003 and the PLS Toolbox, version
4.2). For1H NMR correlation heat-maps and S-TOCSY calcula-
tion, a Matlab in-house script based on Cloarec et al.35was
used.
1H NMR lipidic and
3. Results
3.1. Course of Liver Steatosis in the Animal Model. A
summary of relevant changes in this model under different
dietary conditions is depicted in Figure 1. Mean body weight
increased in an age- and diet-dependent manner, and the
maximal differences were observed from the high-fat, high-
cholesterol diet at the end of dietary treatment (Figure 1A).
These changes were paralleled by significant increases in
plasma cholesterol and triglyceride concentrations (Figure 1B,
C), significantly apparent in mice on the high-fat, high-
cholesterol diet at the 32-week time point. For descriptive
purposes, this model develops significant atherosclerosis (Fig-
ure 1D) that is closely linked to hyperlipidemia and other
metabolic disturbances. At 16 weeks of age, the atherosclerotic
lesion size was not detectable in animals fed the chow diet and
was minimal in mice fed high-fat diet, high-cholesterol diets.
At 32 weeks of age, the plaque progression increased in mice
fed high-fat diets with respect to controls, although the
differences in lesion size were only significant in animals fed
on the high-fat, high-cholesterol diet. In the liver, the intake
of a high-fat diet produced inflammation and steatosis (Figure
1E, F), an effect clearly evident when there was a supplementa-
tion of cholesterol to the diet. However, when the infiltration
of inflammatory cells in the liver was measured as the hepatic
presence of F4/80 cells, there were minimal differences among
dietary treatments (Figure 1G). Dietary cholesterol significantly
increased the hepatic cholesterol concentration at both the 16-
and 32-week time point (Figure 1H). This trend was also
observed in the hepatic triglyceride content, although it did
not reach statistical significance (Figure 1I). Hepatic neutral
lipid storage was also confirmed by examining the amount of
Oil red O stained cells (Figure 1J-L).
3.2. Analysis of1H NMR Spectroscopic Profiles of Liver
Extracts. Aqueous extracts1H NMR profiles of liver from mice
underwent a chow diet were markedly different to those in a
high-fat, high-cholesterol diet (Supplementary Figure S1, Sup-
porting Information).1H NMR spectrum of aqueous soluble
liver tissue extracts shows resonances mainly associated with
low molecular weight metabolites such as amino acids and
related compounds, glucose, lactate, nucleotides, intermediate
metabolites, and soluble membrane components, such as
choline.1H NMR spectra of the lipophilic extracts are composed
of several dominating regions with major peaks attributable
to double bonds mainly from protons belonging to di- or tri-
acylated glycerols, the phospholipids polar head groups, me-
thylene and methyl groups of the fatty acyl chains, and methyl
cholesterol. The spectra allowed a detailed assignment of
unsaturated fatty acyl moieties from components such as
phospholipids and triglycerides (Supplementary Figure S2,
Supporting Information).
1H NMR resonance assignments with chemical shifts, mul-
tiplicity, and J-coupling constants of the signals elucidated in
1H NMR spectra of both the water and lipid-soluble mice liver
extracts are shown in Table 1.
3.3. Diet- and Time-Dependent Metabolic Changes As-
sociated with Gradual Development of Liver Steatosis and
Inflammation. Multivariate PCA analysis in a strategy designed
to compare diets were initially performed on the data derived
from metabolites quantification at either 16 or 32 weeks of age.
Supplementary Figure S3 (A, B, Supporting Information) shows
the scatter scores and loading plots for these two PCA models.
The PCA scores plot at 16 weeks of age revealed a distinct mice
group clustering trend according to their dietary manipulation
along PC1, which accounts for a 37% of total variation within
the data matrix (Figure 2A). Along PC1, samples appeared
gradually arranged according to the presence of hepatic lesions.
This was minimal in mice fed on the chow diet at 10 weeks of
age, which had positive PC1 values and were considered as
baseline. Those mice fed with the cholesterol supplemented
diet characteristically showed negative values of PC1 and
presented significant hepatic lesions. Thus, moving along PC1
from positive to negative values, liver damage produced by
different dietary conditions appears progressive. At the 16-week
time point, hepatic inflammation was evident in all groups,
especially in mice fed on cholesterol supplemented diets, and
steatosis was considered incipient for mice fed on high-fat,
high-cholesterol diets. Conversely, at 32 weeks of age, steatosis
Effect of Dietary Cholesterol on Fatty Liver Disease
research articles
Journal of Proteome Research • Vol. 9, No. 5, 2010
2529

Page 4

Table 1.
Water and Lipid-Soluble Extractsa
1H NMR Resonance Assignments with Chemical Shifts, Multiplicity, and J-Coupling Constants for Signals Identified in
code metabolite
δ (H1shift)
ppm# protons moieties assignmentmultiplicity
Water-soluble metabolites
5.23 + 4.641(C-R) + 1(C-?)
5.491
7.961
2.403
1.203
4.111
6.521
3.652)1(C1) + 1(C3)
8.251
8.341
7.871
5.801
4.581
3.209
1.913
4.501
0.966
1.043
1.483
1.013
1.722
2.452
2.352
2.642
7.385
4.251
6.902
7.061
3.262
2.554
3.043
3.053
1
2
3
4
5
6
7
8
9
Glucose-6-Phosphate (q)
Glucose-1-Phophate-(Glycogen) (q)
UDPG (q)
Pyruvate (q)
3-hidroxybutyrate (q)
Lactate (q)
Fumarate (q)
Free glycerol (q)
NAD/NADP/NADPH (q)
ATP/ADP/AMP (q)
UTP/UDP/UMP (q)
Uracil (q)
Carnitine (q)
Cholines (q)
Acetates (q)
Ascorbic acid (q)
Leucine (q)
Valine (q)
Alanine (q)
Isoleucine (q)
Lysine (q)
Glutamine (q)
Glutamate (q)
Methionine (q)
Phenylalanine (q)
Threonine (q)
Tyrosine (q)
Histidine (q)
Taurine (q)
Glutathione (oxidized) (q)
Creatine (q)
Creatinine (q)
HC-R + HC-?
C1-H
H6-ring
CH3-CO-
CH3-CHO-
RCH
CH)
CH2-OH
H4-ring nicotinamide
H2-ring adenine
H6-ring
ring C5-H
?(CH)-OH
(CH3)3-N
CH3-CO
C4-H
γ(CH3) + γ(CH3)
γ(CH3)
?(C H3)
γ(CH3)
δ(CH2)
γ(CH2)
γ(CH2)
γ(CH2)
H2-H6 ring
? (CH)
H3, H5 ring
H4-ring
CH2-N
γ(CH2) Glu
N-CH3
N-CH3
d + d
m
d (8.2)
s
d (6.20)
q (7.0)
s
dd (7.4, 4.4)
m
s
d (8.10)
d (7.70)
s
s
s
d (2.0)
t (6.40, 6.40)
d (4.40)
d (7.20)
d (7.0)
m
m (broad)
m (broad)
t (7.3, 7.3)
m (broad)
m
d (8.50)
s
t (6.80, 6.80)
m
s
s
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Lipid-soluble metabolites
3
3
3
3 (+3)
1
2
3
4
Total Cholesterol (q)
Free Cholesterol (q)
Esterified Cholesterol (q)
Total Cholesterol
0.69
1.02
1.04
0.86
C18-H3
C19-H3
C19-H3
C26/27-H3(one of the doublets is
embedded in 0.88 triplet signal
arising from ω-CH3)
C21-H3
Glycerol (C1-Hu) and (C3-Hu)
Glycerol (C1-Hd) and (C3-Hd)
Glycerol (C2-H)
Glycerol (C3-H2). C-3 position is
not acylated, then C-3 has fast free
rotation leading to a doublet.
Glycerol (C1-Hd) (The other proton
(Hu) attached to C1 remains
embedded inside 4.34 dd). Only a
doublet of dd is visible
Glycerol (C2-H)
FA RH -CH2-CO-O-C2
Glycerol (C1,3-H2) acylated in pos
C-2
Glycerol (C2-H) acylated in pos C-2
Glycerol (C3-H2)
s
s
s
2 × d
5
6
7
8
9
Total Cholesterol
Triglycerides
Triglycerides
Triglycerides (q)
Diglycerides
0.92
4.17
4.34
5.29
3.69
3
2
2
1
2
d (6.6)
dd (11.8,6.1)
dd (12.1,4.1)
q (5.1)
d (5.2)
10Diglycerides4.381/2 (+1/2)
d (5.3)
11
12
13
Diglycerides (q)
Monoglycerides
Monoglycerides
5.10
2.29
3.72
1
2
4, 2 + 2
m
t (7.5)
t(5.2)
14
15
Monoglycerides (q)
Total phospholipids
4.89
4.01
1
2
q (5.1)
m (broad) two
overlapping
pairs of
quartets
m
16Total phospholipids (except
lysophosphatidylcholine)
4.421 (+1) Glycerol (C1-H2) (The other proton
attached to C1 remains embedded
inside 4.34 position)
Glycerol (C2-H)
17Total phospholipids (except
lysophosphatidylcholine) (q)
5.241quartet of
doublets
research articles
Vinaixa et al.
2530 Journal of Proteome Research • Vol. 9, No. 5, 2010

Page 5

clearly increased, and the PCA model did not show such a
gradual separation on the three different diets (Supplementary
Figure S3B, Supporting Information). PC1 gathering the highest
variance is rather discriminating among baseline and 32 weeks
of age mice regardless of the dietary manipulation intervention.
However, mice fed the cholesterol-supplemented diet were
clearly differentiated along PC2, which explained 22% of the
total variation (Figure 2B). To assess the metabolites putatively
implicated in the variations observed in inflammation and
steatosis through the progressive lesions induced by dietary
modification, we have studied the loadings bar plots of PC1 at
16 weeks of age (Figure 2A) and PC2 at 32 weeks of age (Figure
2B). Metabolites accounting for higher absolutes value in the
loadings bar plot exert higher influence in the PCA model. Data
are also color-coded according to significance determined by
a Kruskal-Wallis test (p < 0.05). Gray bars indicate that their
corresponding metabolites are significantly different among the
dietary groups entering the PCA model. Thus, at the 16-week
time point, when steatosis is not yet fully developed, livers from
mice fed the cholesterol supplemented diet presented signifi-
cantly raised levels of free and esterified cholesterol as well as
triglycerides and oleic acid in comparison to their counterparts
on cholesterol free diets. However, mice undergoing lower
levels of hepatic lesions, inflammation and steatosis score
(baseline and chow diet mice), which are characterized for
positive values of PC1, presented depleted levels of triglycerides
and total cholesterol but increased hepatic levels of PUFA (ω-3
fatty acyls, docosahexanoic 22:6(n-3) (DHA), arachidonic 20:
4(n-6) (ARA) + eicosapentaenoic 20:5(n-3) (EPA), and PUFA/
MUFA and PC/PE ratios. These changes were paralleled by
increased levels of metabolites such as free glycerol, lactate,
G-6-P, carnitine, taurine, glutathione, and methionine, among
others. At the 32-week time point, in the presence of severe
steatosis and clear evidence of hepatic lesions, mice fed the
cholesterol supplemented diet showed significantly higher
levels of triglycerides, free and esterified cholesterol, oleic acid,
and some amino acids, such as leucine, valine, and lysine. Also,
at this end time point there was a significant depletion of PUFA
(ARA + EPA, DHA, and linoleic) and carnitine levels, among
others.
3.4. Characterization of Differential Metabolite Patterns
Associated with Liver Steatosis and Inflammation. The actual
concentration of the 55 metabolites identified and segregated
by dietary condition and time of observation is shown in
Supplementary Table S1 (Supporting Information). For com-
parison, baseline values are also included and significant
changes between chow and high-fat or high-fat supplemented
cholesterol diets (Mann-Whitney U test, p < 0.05) are indicated.
Of note, dietary cholesterol significantly raises the levels of total
hepatic cholesterol and triglycerides at both the 16- and 32-
week time points (Figure 3A, B). Additionally, the levels of
MUFA were increased significantly at the expense of PUFA,
resulting in a significantly decreased PUFA/MUFA ratio (Figure
3C) that correlated with hepatic lesions.
Table 1 Continued
code metabolite
δ (H1shift)
ppm # protons moieties assignment multiplicity
18
19
Phosphatidylcholine
Phosphatidylcholine (q)
3.22
3.60
9
2
-CH2-N-(CH3)3
Alkyl-PC (Phosphoether)
CH2-N-(CH3)3
Alkyl-PE (Phosphoether)
R-PO-CH-CH2-N+H3
Inositol cycle 6′ -CH-OH
NH3+-CH-COO
C-3′ CH2-OH
C-3 CH2-OP
C-1 CH2-OC
-CH2-N-(CH3)3
CH3(CH2)n-CbH)CaH-C1HOH-
C2HN-C3H2-O--P
CH3(CH2)n-CbHdCaH-C1H2OR-
C2HOCO-C3H2O-P
FA chain CH3(CH2)n
ω-3 CH3-CH2-CdC
FA chain -(CH2)n-
?H2R-CH2CH2CO-OR
?H2
-CHdCH-CH2-CH2-CH2CO-OR
-CHdCH-CH2-
γH2
-CHdCH-CH2-CH2-CH2CO-OR
RH2-CH2-CO-OR
RH2and ?H2
-CHdCH-CH2-CH2-CO-OR
-CHdCH-CH2-(CHdCH-CH2-)n,
n ) 1
-CHdCH-CH2-(CHdCH-CH2-)n,
n g 2
-CH)CH-
s
m
20Phosphatidylethanolamine (q)
3.112m
21
22
23
24
25
26
27
Phosphatidylinositol (q)
Phosphatidylserine (q)
Phosphatidylglycerol
Lysophosphatidylcholine
Lysophosphatidylcholine (q)
Sphingomyelin
Sphingomyelin (q)
3.79
6.67
3.59
3.87
4.12
3.21
5.68
1
3
2
2
2
9
1
t (9.5)
m
dd
m
m
s
m
28 Plasmalogen (q)
5.921d (6.7)
29
30
31
32
33
FA, ω-CH3
ω-3, (DHA+ EPA+ linolenic) (q)
FA, (Total Fatty acyl chains)
FA, ?H2
FA, ARA+EPA
0.88
0.98
1.30
1.62
1.70
3
3
2
t (6.90)
t (7.50)
m
m
m
2
34
35
FA, OLEIC (q)
FA, ARA+EPA (q)
2.02
2.12
4
2
m
m
36
37
FA, RH2
FA, DHA (q)
2.35
2.41
2
4
m
m
38FA, LINOLEIC (q)
2.784 t (6.4)
39FA, PUFA (q)
2.852m
40FA, MUFA (with PUFA) (q)
5.362m
aIn water-soluble extracts, only quantitative signals are indicated. Subscript (q) indicates signals used for quantitative purposes. ARA, arachidonic acid;
DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; SM, sphingomyelin; FA: fatty acid chain.
s ) singlet; d ) doublet; dd)double doublet; t ) triplet; q ) quartet; m ) multiplet.
Effect of Dietary Cholesterol on Fatty Liver Disease
research articles
Journal of Proteome Research • Vol. 9, No. 5, 2010
2531

Page 6

Fatty acyl chain composition was analyzed and represented
in terms of molar percentages, in both time points (Figure 3D,
E). Dietary cholesterol seems to be responsible for a significant
increase in the proportion of oleic acid and a significant
clearance of long chain polyunsaturated fatty acids such as
ARA, EPA, and DHA.
The hepatic concentrations of taurine, glutathione, carnitine,
and methionine were also decreased with respect to baseline
values (Figure 4). At 16 weeks of age, the hepatic taurine,
glutathione, and carnitine levels (Figure 4A, B, C) were signifi-
cantly decreased for those mice fed a cholesterol enriched diet
when compared to their chow diet counterparts.
Finally, heat-map correlations of lipid-soluble
spectra with inflammation and steatosis (Figure 5) demonstrate
that the inflammation score correlates with dietary cholesterol
as this is observed in all resonances from different proton nuclei
belonging to cholesterol molecules (Figure 5A). Although
cholesterol also contributes to the steatosis score, this is mainly
correlated with the triglyceride NMR-derived signals (Figure
5B).
1H NMR
4. Discussion
Our data indicate that dietary cholesterol has a profound
impact in the composition of fatty acids in the liver. In
Figure 1. High-fat, high-cholesterol diets elicited metabolic changes in LDLr-/-mice. There was a significant increase in body weight
(A), hyperlipidemia (B, C), and arteriosclerosis progression (D). The influence of diet is also evident in the liver as indicated by the
measurement of inflammation (E), the degree of steatosis (F), and the infiltration of inflammatory cells (G). These effects were
accompanied by an increased accumulation of cholesterol and triglycerides in the liver (H, I), confirmed histologically by Oil Red O
staining (J-L) from which a representative microphotograph of each diet is shown at the 32-week time point. Values are mean ( SD
* p < 0.05 with respect to controls.
research articles
Vinaixa et al.
2532Journal of Proteome Research • Vol. 9, No. 5, 2010

Page 7

Figure 2. Principal component analyses (PCA) based on the metabolic patterns obtained from aqueous and lipophilic liver tissue extracts at 16 (A) and 32 weeks of age (B). The PCA
scores plot showed an important separation of clusters according to dietary intervention which was more evident at 16 weeks of age. A PCA loadings plot revealed the metabolites
implicated in the induced hepatic lesions. Gray bars indicate which metabolites are significantly different among the dietary groups based on a Kruskal-Wallis test (p < 0.05).
Effect of Dietary Cholesterol on Fatty Liver Disease
research articles
Journal of Proteome Research • Vol. 9, No. 5, 2010
2533

Page 8

particular, a decreased PUFA/MUFA ratio is observed at an
early stage, probably indicating excessive lipid peroxidation and
oxidative stress.31,37This is in accordance with previous find-
ingsobtainedinobesehumanswithsignificantliversteatosis.5,38-40
Moreover, the significant depletion of PUFA indicates a reduc-
tion in fatty acid oxidation and triglyceride release from the
liver with a consequent increase in triglyceride synthesis that
may contribute significantly to the development of triglyceride
accumulation in hepatocytes.41
Our dietary conditions, as shown by the relative enrichment
of several fatty acids, clearly indicate variations in metabolism
that suggest that 12/15 lipoxygenase (12/15LO) or related
molecules may play a role in exacerbating the situation. High-
fat diet-induced liver steatosis is an inflammatory condition
that involves the recruitment of macrophages. Among candi-
date genes that regulate inflammation in tissues, it has recently
been proposed that 12/15LO is involved. Cells overexpressing
12/15LO secret potent chemokines, mainly monocyte chemoat-
tractant protein-1, and 12/15LO knockout mice exhibit no high-
fat diet induced change in tissues.42The family of 12/15LO
enzymes has a relevant role in the metabolism of fatty acids.
For instance, 12/15LO catalyzes the insertion of molecular
oxygen in ARA resulting in a fatty acid hydroperoxide and also
oxygenates linoleic acid 18:2(n-6). Free unsaturated fatty acids
and fatty acids esterified in phospholipids and cholesteryl esters
are substrates for 12/15LO.43
We found decreased hepatic methionine concentrations
respect to baseline values regardless of the dietary treatment.
This is also accompanied by a significant decrease in glu-
tathione and taurine concentrations in animals fed a high-fat,
high-cholesterol diet. Both taurine and glutathione are down-
stream products of methionine metabolism via S-adenosyl-L-
Figure 3. Dietary cholesterol elicited metabolic changes in lipid-soluble metabolites associated with liver steatosis and inflammation.
A significant increase in hepatic cholesterol (A) and triglyceride (B) content was detected whereas the PUFA/MUFA ratio (C) was decreased
in mice fed high-fat diets. The effect of dietary cholesterol was also evident in the fatty acyl chain composition (D, E). There was a
significant increase in the hepatic proportion of oleic acid, although decreased ARA, DHA, EPA, and ω-3 content was detected.
research articles
Vinaixa et al.
2534Journal of Proteome Research • Vol. 9, No. 5, 2010

Page 9

methionine (SAMe) and transsulfuration pathways. Thus, the
concomitant decrease of these metabolites together with
methionine depletion suggests that SAMe and transsulfuration
pathways fluxes are probably decreased because of dietary
conditions. Such a relationship is more evident with inflam-
mation (16 weeks of age) rather than with steatosis, suggesting
that dietary cholesterol may elicit a relative SAMe deficiency.
This may be supported by the fact that intact SAMe concentra-
tions are crucial for the critical balance of pro-inflammatory
and anti-inflammatory cytokines in the pathogenesis of liver
disease.44Our data may also be concordant with recent findings
indicating that the transition from fatty liver to a more intense
inflammatory condition is associated with SAMe depletion in
ob/ob mice fed a methionine and choline-deficient diet.45
Moreover, knockout mice deficient in hepatic SAMe synthesis
(MAT1A-/-)46developed steatohepatitis associated with a
significant downregulation of mitochondrial proteins (e.g.,
prohibitin 1, cytochrome c oxidase I and II, and ATPase
b-subunit), some of them directly regulated by SAMe. Ad-
ditionally, hepatic taurine and glutathione depletion might also
be related to other metabolic pathways. Taurine is linked to
the activity of the hepatic cholesterol-7R-hydroxylase (CYP7A1),
a key enzyme in the process of cholesterol excretion and bile
acid synthesis. As we previously described, a decrease in the
hepatic concentration of taurine in animals fed high-fat, high-
cholesterol diets could be due to increased excretion of taurine-
conjugated bile acids caused by an excess of cholesterol
accumulation in the liver.13Likewise, the possible metabolic
consequences related to glutathione depletion are also remark-
able. Glutathione, the major low-molecular-weight thiol in
animal cells, plays crucial roles in antioxidant defense, nutrient
metabolism, and the regulation of whole-body homeostasis.47
Depleted glutathione levels can be either attributed to increased
gamma-glutamyl cycle48or glutathione-S-transferase activi-
ties.49
Increased dietary cholesterol also resulted in a significant
decrease of hepatic carnitine concentrations. It is well docu-
mented that the essential role of carnitine consists of the
facilitation of mitochondrial import and oxidation of long chain
fatty acids but that it also functions as an acyl group acceptor
that facilitates mitochondrial export of excess carbons in the
form of acylcarnitine. It is therefore possible that diminished
carnitine reserves in the liver may be accompanied by marked
perturbations in mitochondrial fuel metabolism, including low
rates of complete fatty acid oxidation. Interestingly, it has been
recently demonstrated that carnitine insufficiency may be
caused by chronic overnutrition in mice and that this com-
promises not only mitochondrial performance but also meta-
bolic control.50
Taken together, our results suggest that dietary cholesterol
is a causal factor in the development of liver steatosis, probably
through an impact in the overall metabolism of fatty acids via
a mitochondrial impairment. Dietary cholesterol is also caus-
ative of hepatic inflammation and consequently of associated
metabolic abnormalities, indicating that dietary manipulation
is critical in the clinical care.
5. Conclusions
The present work offers a proof-of-concept, where we are
able to detect clear changes in metabolite levels driving hepatic
inflammation to NASH using a holistic, quantitative profiling
NMR-based metabolomics approach. NMR-derived quantita-
tive data have been used for further multivariate modeling to
ascertain the role of dietary cholesterol in the progression of
Figure 4. Representation of the most significant changes in water-soluble metabolites. Significant decreases in the hepatic taurine (A),
glutathione (B), and carnitine (C) levels were observed at 16 weeks of age in mice on high-fat, high-cholesterol diet. At 16 weeks of
age, hepatic methionine concentration is also decreased as compared to baseline levels in both chow diet and high-fat diets (D).
Effect of Dietary Cholesterol on Fatty Liver Disease
research articles
Journal of Proteome Research • Vol. 9, No. 5, 2010
2535

Page 10

fatty liver disease. This method may become a useful comple-
mentary tool in the in vivo mice experiments in which
metabolic phenotyping is required.
Our data established dietary cholesterol as a causative factor
in the development of both liver steatosis and hepatic inflam-
mation. Analysis of the affected metabolites suggests that the
mechanisms involved are related to impairment in mitochon-
drial function, particularly to variations in normal fatty acid
metabolism. This metabolomic approach may provide infor-
mation to indicate dietary modifications to modify the revers-
ible components of the associated metabolic derangements.
Acknowledgment. This work was supported by grants
PI051606 and PI08/1381 from the Fondo de Investigacio ´n
Sanitaria (FIS) and by CIBER de Diabetes y Enfermedades
Metabo ´licas (CIBERDEM). CIBERDEM is an initiative of
ISCIII (Instituto de Salud Carlos III), Madrid, Spain. Anna
Rull is the recipient of a fellowship from the Generalitat de
Catalunya (FI-G 0503).
Supporting Information Available: Figure S1 shows
a comparison between
aqueous extracts from LDLr-/-mice-fed chow diet and high-
fat +0.25% cholesterol at 16 weeks of age. Figure S2 shows the
identified metabolites in the1H NMR spectra (600 MHz) of liver
lipidic extract from mice under high-fat +0.25% cholesterol diet
at 32 weeks of age. Figure S3 shows the scores and loadings
scatter plot of PCA analysis performed either at 16 and 32 weeks
of ages. Table S1 summarizes the quantification of the 55
identified metabolites within the different dietary and time-
1H NMR spectra (600 MHz) of liver
Figure 5. Heat-map correlation of lipid-soluble1H NMR spectra demonstrated the crucial role of dietary cholesterol in liver steatosis
and inflammation. The hepatic inflammation score was correlated to all resonance related to cholesterol molecules (A) and the hepatic
steatosis score was mainly correlated with triglyceride NMR derived signals (B).
research articles
Vinaixa et al.
2536Journal of Proteome Research • Vol. 9, No. 5, 2010

Page 11

point conditions. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
(1) Tiniakos, D. G.; Vos, M. B.; Brunt, E. M. Nonalcoholic fatty liver
disease: Pathology and pathogenesis. Annu. Rev. Pathol.: Mech.
Dis. 2010, 5, 145–171.
(2) Fabbrini, E.; Sullivan, S.; Klein, S. Obesity and nonalcoholic fatty
liver disease: biochemical, metabolic, and clinical implications.
Hepatology 2010, 51, 679–689.
(3) Malaguerna, M.; Di Rosa, M.; Nicoletti, F.; Malaguerna, L. Molec-
ular mechanisms involved in NAFLD progression. J. Mol. Med.
2009, 87, 679–695.
(4) Neuschwander-Tetri, B.; Caldwell, S. Nonalcoholic steatohepatitis:
summary of an AASLD single topic conference. Hepatology 2003,
37, 1202–1219.
(5) Johnson, N. A.; Walton, D. W.; Sachinwalla, T.; Thompson, C. H.;
Smith, K.; Ruell, P. A.; Stannard, S. R.; George, J. Noninvasive
assessment of hepatic lipid composition: Advancing understanding
and management of fatty liver disorders. Hepatology 2008, 47 (5),
1513–1523.
(6) Brunt, E. M. Nonalcoholic steatohepatitis. Semin. Liver Dis. 2004,
24 (1), 3–20.
(7) Day, C. P.; James, O. F. W. Steatohepatitis: A tale of two “hits.
Gastroenterology 1998, 114 (4), 842–845.
(8) Browning, J. D.; Horton, J. D. Molecular mediators of hepatic
steatosis and liver injury. J. Clin. Invest. 2004, 114 (2), 147–152.
(9) Marı ´, M.; Caballero, F.; Colell, A.; Morales, A.; Caballeria, J.;
Fernandez, A.; Enrich, C.; Fernandez-Checa, J. C.; Garcı ´a-Ruiz, C.
Mitochondrial free cholesterol loading sensitizes to TNF- and Fas-
mediated steatohepatitis. Cell Metab. 2006, 4 (3), 185–198.
(10) Rull, A.; Escola `-Gil, J. C.; Julve, J.; Rotllan, N.; Calpe-Berdiel, L.;
Coll, B.; Aragones, G.; Marsillach, J.; Alonso-Villaverde, C.; Camps,
J.; Blanco-Vaca, F.; Joven, J. Deficiency in monocyte chemoattrac-
tant protein-1 modifies lipid and glucose metabolism. Exp. Mol.
Pathol. 2007, 83 (3), 361–366.
(11) Joven, J.; Rull, A.; Ferre ´, N.; Escola `-Gil, J. C.; Marsillach, J.; Coll,
B.; Alonso-Villaverde, C.; Aragones, G.; Claria, J.; Camps, J. The
results in rodent models of atherosclerosis are not interchangeable:
The influence of diet and strain. Atherosclerosis 2007, 195 (2), e85–
e92.
(12) Wouters, K.; Gorp, P. J. v.; Bieghs, V.; Gijbels, M. J.; Duimel, H.;
Lu ¨tjohann, D.; Kerksiek, A.; Kruchten, R. v.; Maeda, N.; Staels, B.;
Bilsen, M. v.; Shiri-Sverdlov, R.; Hofker, M. H. Dietary cholesterol,
rather than liver steatosis, leads to hepatic inflammation in
hyperlipidemic mouse models of nonalcoholic steatohepatitis.
Hepatology 2008, 48 (2), 474–486.
(13) Rull, A.; Vinaixa, M.; A´ngel Rodrı ´guez, M.; Beltra ´n, R.; Brezmes, J.;
Can ˜ellas, N.; Correig, X.; Joven, J. Metabolic phenotyping of
genetically modified mice: An NMR metabonomic approach.
Biochimie 2009, 91 (8), 1053–1057.
(14) Wishart, D. S. Quantitative metabolomics using NMR. TrAC, Trends
Anal. Chem. 2008, 27 (3), 228–237.
(15) Xu, E. Y.; Schaefer, W. H.; Xu, Q. Metabolomics in pharmaceutical
research and development: metabolites, mechanisms and path-
ways. Curr. Opin. Drug Discovery Dev. 2009, 12 (1), 40–52.
(16) Ekman, D.; Keun, H.; Eads, C.; Furnish, C.; Murrell, R.; Rockett, J.;
Dix, D. Metabolomic evaluation of rat liver and testis to character-
ize the toxicity of triazole fungicides. Metabolomics 2006, 2 (2),
63–73.
(17) Waters, N. J.; Holmes, E.; Williams, A.; Waterfield, C. J.; Farrant,
R. D.; Nicholson, J. K. NMR and Pattern Recognition Studies on
the Time-Related Metabolic Effects of R-Naphthylisothiocyanate
on Liver, Urine, and Plasma in the Rat: An Integrative Metabo-
nomic Approach. Chem. Res. Toxicol. 2001, 14 (10), 1401–1412.
(18) Atherton, H. J.; Bailey, N. J.; Zhang, W.; Taylor, J.; Major, H.;
Shockcor, J.; Clarke, K.; Griffin, J. L. A combined H-1-NMR
spectroscopy- and mass spectrometry-based metabolomic study
of the PPAR-alpha null mutant mouse defines profound systemic
changes in metabolism linked to the metabolic syndrome. Physiol.
Genomics 2006, 27 (2), 178–186.
(19) Serkova, N. J.; Zhang, Y.; Coatney, J. L.; Hunter, L.; Wachs, M. E.;
Niemann, C. U.; Mandell, M. S. Early detection of graft failure using
the blood metabolic profile of a liver recipient. Transplantation
2007, 83 (4), 517–521.
(20) Griffiths, W. J. Metabolomics, Metabonomics and Metabolite Profil-
ing; The Royal Society of Chemistry: Cambridge, U.K, 2008.
(21) Rull, A.; Rodrı ´guez, F.; Aragone `s, G.; Marsillach, J.; Beltra ´n, R.;
Alonso-Villaverde, C.; Camps, J.; Joven, J. Hepatic monocyte
chemoattractant protein-1 is upregulated by dietary cholesterol
and contributes to liver steatosis. Cytokine 2009, 48 (3), 273–279.
(22) Tous, M.; Ferre ´, N.; Camps, J.; Riu, F.; Joven, J. Feeding apolipo-
protein E-knockout mice with cholesterol and fat enriched diets
may be a model of non-alcoholic steatohepatitis. Mol. Cell.
Biochem. 2005, 268 (1-2), 53–58.
(23) Tous, M.; Ferre ´, N.; Rull, A.; Marsillach, J.; Coll, B.; Alonso-
Villaverde, C.; Camps, J.; Joven, J. Dietary cholesterol and dif-
ferential monocyte chemoattractant protein-1 gene expression in
aorta and liver of apo E-deficient mice. Biochem. Biophys. Res.
Commun. 2006, 340 (4), 1078–1084.
(24) Calleja, L.; Paris, M. A.; Paul, A.; Vilella, E.; Joven, J.; Jimenez, A.;
Beltran, G.; Uceda, M.; Maeda, N.; Osada, J. Low-cholesterol and
high-fat diets reduce atherosclerotic lesion development in ApoE-
knockout mice. Arterioscler. Thromb. Vasc. Biol. 1999, 19 (10),
2368–2375.
(25) Shi, C.; Wu, C. Q.; Cao, A. M.; Sheng, H. Z.; Yan, X. Z.; Liao, M. Y.
NMR-spectroscopy-based metabonomic approach to the analysis
of Bay41-4109, a novel anti-HBV compound, induced hepatotox-
icity in rats. Toxicol. Lett. 2007, 173 (3), 161–167.
(26) Coen, M.; Lenz, E. M.; Nicholson, J. K.; Wilson, I. D.; Pognan, F.;
Lindon, J. C. An integrated metabonomic investigation of acetami-
nophen toxicity in the mouse using NMR spectroscopy. Chem. Res.
Toxicol. 2003, 16 (3), 295–303.
(27) Casu, M.; Anderson, G. J.; Choi, G.; Gibbons, W. A. NMR Lipid
Profiles of Cells, Tissues and Body Fluids (1D and 2D Proton NMR
of Lipids from Rat Liver). Magn. Reson. Chem. 1991, 29, 594–602.
(28) Sparling, M. L. Analysis of Mixed Lipid Extracts Using H-1-Nmr
Spectra. Comput. Appl. Biosci. 1990, 6 (1), 29–42.
(29) Sparling, M. L.; Zidovetzki, R.; Muller, L.; Chan, S. I. Analysis of
Membrane-Lipids by 500 MHz H-1-Nmr. Anal. Biochem. 1989, 178
(1), 67–76.
(30) Adosraku, R. K.; Choi, G. T.; Constantinou-Kokotos, V.; Anderson,
M. M.; Gibbons, W. A. NMR lipid profiles of cells, tissues, and body
fluids: proton NMR analysis of human erythrocyte lipids. J. Lipid
Res. 1994, 35 (11), 1925–1931.
(31) Serkova, N. J.; Jackman, M.; Brown, J. L.; Liu, T.; Hirose, R.; Roberts,
J. P.; Maher, J. J.; Niemann, C. U. Metabolic profiling of livers and
blood from obese Zucker rats. J. Hepatol. 2006, 44 (5), 956–962.
(32) Pollesello, P.; Masutti, F.; Croce, L. S.; Toffanin, R.; Eriksson, O.;
Paoletti, S.; Hockerstedt, K.; Tiribelli, C. 1H-NMR Spectroscopic
Studies of Lipid Extracts from Human Fatty Liver. Biochem.
Biophys. Res. Commun. 1993, 192 (3), 1217–1222.
(33) Claus, S. P.; Tsang, T. M.; Wang, Y.; Cloarec, O.; Skordi, E.; Martin,
F.-P.; Rezzi, S.; Ross, A.; Kochhar, S.; Holmes, E.; Nicholson, J. K.
Systemic multicompartmental effects of the gut microbiome on
mouse metabolic phenotypes. Mol. Syst. Biol. 2008, 4.
(34) Papathanasiou, A.; Kostara, C.; Cung, M. T.; Seferiadis, K.; Elisaf,
M.; Bairaktari, E.; Goudevenos, I. A. Analysis of the composition
of plasma lipoproteins in patients with extensive coronary heart
disease using H-1 NMR spectroscopy. Hellenic J. Cardiol. 2008,
49 (2), 72–78.
(35) Cloarec, O.; Dumas, M.-E.; Craig, A.; Barton, R. H.; Trygg, J.;
Hudson, J.; Blancher, C.; Gauguier, D.; Lindon, J. C.; Holmes, E.;
Nicholson, J. Statistical Total Correlation Spectroscopy: An Ex-
ploratory Approach for Latent Biomarker Identification from
Metabolic 1H NMR Data Sets. Anal. Chem. 2005, 77 (5), 1282–
1289.
(36) Serkova, N.; Fuller, T. F.; Klawitter, J.; Freise, C. E.; Niemann, C. U.
H-1-NMR-based metabolic signatures of mild and severe is-
chemia/reperfusion injury in rat kidney transplants. Kidney Int.
2005, 67 (3), 1142–1151.
(37) Galli, F.; Varga, Z.; Balla, J.; Ferraro, B.; Canestrari, F.; Floridi, A.;
Vitamin, E. lipid profile, and peroxidation in hemodialysis patients.
Kidney Int. Suppl. 2001, 78, S148–S154.
(38) Araya, J.; Rodrigo, R.; Videla, L. A.; Thielemann, L.; Orellana, M.;
Pettinelli, P.; Poniachik, J. Increase in long-chain polyunsaturated
fatty acid n-6/n-3 ratio in relation to hepatic steatiosis in patients
with non-alcoholic fatty liver disease. Clin. Sci. 2004, 106 (6), 635–
643.
(39) Elizondo, A.; Araya, J.; Rodrigo, R.; Poniachik, J.; Csendes, A.;
Maluenda, F.; Diaz, J. C.; Signorini, C.; Sgherri, C.; Comporti, M.;
Videla, L. A. Polyunsaturated fatty acid pattern in liver and
erythrocyte phospholipids from obese patients. Obesity 2007, 15
(1), 24–31.
(40) Puri, P.; Baillie, R. A.; Wiest, M. M.; Mirshahi, F.; Choudhury, J.;
Cheung, O.; Sargeant, C.; Contos, M. J.; Sanyal, A. J. A lipidomic
analysis of nonalcoholic fatty liver disease. Hepatology 2007, 46
(4), 1081–1090.
(41) Videla, L. A.; Rodrigo, R.; Araya, J.; Poniachik, J. Oxidative stress
and depletion of hepatic long-chain polyunsaturated fatty acids
Effect of Dietary Cholesterol on Fatty Liver Disease
research articles
Journal of Proteome Research • Vol. 9, No. 5, 2010
2537

Page 12

may contribute to nonalcoholic fatty liver disease. Free Radical
Biol. Med. 2004, 37 (9), 1499–1507.
(42) Sears, D. D.; Miles, P. D.; Chapman, J.; Ofrecio, J. M.; Almazan, F.;
Thapar, D.; Miller, Y. I. 12/15-Lipoxygenase Is Required for the
Early Onset of High Fat Diet-Induced Adipose Tissue Inflammation
and Insulin Resistance in Mice. PLoS ONE 2009, 4 (9), e7250.
(43) Zhao, L.; Funk, C. D. Lipoxygenase Pathways in Atherogenesis.
Trends Cardiovasc. Med. 2004, 14 (5), 191–195.
(44) McClain, C. J.; Hill, D. B.; Song, Z.; Chawla, R.; Watson, W. H.;
Chen, T.; Barve, S. S-Adenosylmethionine, cytokines, and alcoholic
liver disease. Alcohol 2002, 27 (3), 185–192.
(45) Wortham, M.; He, L.; Gyamfi, M.; Copple, B.; Wan, Y.-J. The
Transition from Fatty Liver to NASH Associates with SAMe
Depletion in db/db Mice Fed a Methionine Choline-Deficient Diet.
Dig. Dis. Sci. 2008, 53 (10), 2761–2774.
(46) Santamarı ´a, E.; Avila, M. A.; Latasa, M. U.; Rubio, A.; Martı ´n-Duce,
A.; Lu, S. C.; Mato, J. M.; Corrales, F. J. Functional proteomics of
nonalcoholic steatohepatitis: Mitochondrial proteins as targets of
S-adenosylmethionine. Proc. Natl. Acad. Sci. U.S.A. 2003, 100 (6),
3065–3070.
(47) Wu, G.; Fang, Y.-Z.; Yang, S.; Lupton, J. R.; Turner, N. D.
Glutathione metabolism and its implication for health. J. Nutrition
2004, 134, 489–492.
(48) Yuan, L.; Kaplowitz, N. Glutathione in liver diseases and hepato-
toxicity. Mol. Aspects Med. , 30 (1-2), 29–41.
(49) Jennifer, E. M.; Rodman, D. S.; John, E; V, K. Gamma-Glutamyl
Transferase: A Novel Cardiovascular Risk BioMarker. Prev. Cardiol.
, 13 (1), 36–41.
(50) Noland, R. C.; Koves, T. R.; Seiler, S. E.; Lum, H.; Lust, R. M.;
Ilkayeva, O.; Stevens, R. D.; Hegardt, F. G.; Muoio, D. M. Carnitine
Insufficiency Caused by Aging and Overnutrition Compromises
Mitochondrial Performance and Metabolic Control. J. Biol. Chem.
2009, 284 (34), 22840–22852.
PR901203W
research articles
Vinaixa et al.
2538 Journal of Proteome Research • Vol. 9, No. 5, 2010