STEATOHEPATITIS/METABOLIC LIVER DISEASE
Dysfunctional Very-Low-Density Lipoprotein Synthesis
and Release Is a Key Factor in Nonalcoholic
Steatohepatitis Pathogenesis
Koji Fujita,1 Yuichi Nozaki,1 Koichiro Wada,2 Masato Yoneda,1 Yoko Fujimoto,3 Mihoyo Fujitake,3 Hiroki Endo,1
Hirokazu Takahashi,1 Masahiko Inamori,1 Noritoshi Kobayashi,1 Hiroyuki Kirikoshi,1 Kensuke Kubota,1 Satoru Saito,1
and Atsushi Nakajima1
The specific mechanisms of nonalcoholic fatty liver (NAFL) and nonalcoholic steatohepatitis
(NASH) pathogenesis remain unknown. In the present study we investigated the differences
between NAFL and NASH in terms of liver lipid metabolites and serum lipoprotein. In all, 104
Japanese subjects (50 men and 54 postmenopausal women) with histologically verified NAFL
disease (NAFLD) (51 with NAFL, 53 with NASH) were evaluated; all diagnoses were based on
liver biopsy findings and the proposed diagnostic criteria. To investigate the differences between
NAFL and NASH in humans, we carefully examined (1) lipid inflow in the liver, (2) lipid outflow
from the liver, (3) very-low-density lipoprotein (VLDL) synthesis in the liver, (4) triglyceride
(TG) metabolites in the liver, and (5) lipid changes and oxidative DNA damage. Most of the
hepatic lipid metabolite profiles were similar in the NAFL and NASH groups. However, VLDL
synthesis and lipid outflow from the liver were impaired, and surplus TGs might have been
produced as a result of lipid oxidation and oxidative DNA damage in the NASH group. Conclusion: A growing body of literature suggests that a deterioration in fatty acid oxidation and VLDL
secretion from the liver, caused by the impediment of VLDL synthesis, might induce serious lipid
oxidation and DNA oxidative damage, impacting the degree of liver injury and thereby contributing to the progression of NASH. Therefore, dysfunctional VLDL synthesis and release may be
a key factor in progression to NASH. (HEPATOLOGY 2009;50:772-780.)
A
lmost one-quarter of adults in many industrialized
countries have excessive hepatic fat accumulation.1 Nonalcoholic fatty liver disease (NAFLD)
represents a wide spectrum of conditions ranging from
nonalcoholic fatty liver (NAFL), which generally follows
a benign, nonprogressive clinical course, to nonalcoholic
steatohepatitis (NASH), a more serious form of NAFL
that may progress to cirrhosis and endstage liver disease.2
Although NASH represents steatosis with ballooning
and/or fibrosis (Type 3-4 NAFLD), NAFL is characterized by simple steatosis or steatosis with only inflammation (Type 1-2 NAFLD),3 indicating that NAFL
might represent the first stage of NASH pathogenesis.4
The current model of NASH pathogenesis suggests two
stages of progression. First, insulin resistance causes
lipid accumulation in hepatocytes; and second, cellular
Abbreviations: ADRP, adipose differentiation-relative protein; ALT, alanine aminotransferase; apoB100, apolipoproteinB100; BMI, body mass index; CDAA,
choline-deficient L-amino acid-defined; CM, chylomicron; DGAT, diacylglycerol acyltransferase; FFA, free fatty acid; fCh, free choline; HDL, high-density lipoprotein;
HF/HC, high-fat/high-calorie; HOMA-IR, homeostasis assessment insulin resistance; HPLC, high-performance liquid chromatography; LDL, low-density lipoprotein;
MDA, malondialdehyde; MTTP, microsomal triglyceride transfer protein; NAFL, nonalcoholic fatty liver; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic
steatohepatitis; PC, phosphatidylcholine; PET, positron emission tomography; PL, phospholipid; PPAR-␣, peroxisome proliferator-activated receptor alpha; SFA, subcutaneous fat area; SREBP, sterol regulatory element binding protein; TC, total cholesterol; TG, triglycerides; TNF-␣, tumor necrosis factor alpha; VFA, visceral fat area;
VLDL, very-low-density lipoprotein.
From the 1Division of Gastroenterology, Yokohama City University Graduate School of Medicine, Yokohama, Japan; 2Department of Pharmacology, Graduate School
of Dentistry, Osaka University, Osaka, Japan; 3Osaka University of Pharmaceutical Sciences, Osaka, Japan.
Received February 16, 2009; accepted April 26, 2009.
Supported in part by a Grant-in-Aid for research on the Third Term Comprehensive Control Research for Cancer from the Ministry on Health, Labor and Welfare, Japan
(to A.N.); a grant from the National Institute of Biomedical Innovation (NBIO) (to A.N.); a grant from the Ministry of Education, Culture, Sports, Science and
Technology, Japan (KIBAN-B) (to A.N.); a grant program “Collaborative Development of Innovative Seeds” from the Japan Science and Technology Agency (JST); and
a grant from the Ministry of Education, Culture, Sports, Science and Technology, Japan Wakate-kenkyuu start up (grant no. 20890185) (to K.F.). This study was not
supported by private funding.
772
HEPATOLOGY, Vol. 50, No. 3, 2009
insults such as oxidative stress, lipid oxidation, direct
lipid toxicity, mitochondrial dysfunction, and/or bacterial endotoxins from the gut cause hepatic inflammation, resulting in NASH.5 In most cases, however,
NAFL does not progress to more severe pathological
conditions, i.e., 10%-20% of patients with NAFLD
eventually exhibit signs of fibrosis, necrosis, and inflammation, indicating the presence of NASH.6 The
factors involved in the progression of NAFL to NASH
are not fully understood. Moreover, the real percentage
of NAFL that progresses to NASH is unknown, and
some researchers have suggested that NAFL and NASH
are two distinct entities. Therefore, it is critically important to clarify this issue.
In general, two experimental animal models are often used to study NAFLD: a high-fat/high-calorie
(HF/HC) diet as a generalized NAFL model,7 and a
choline-deficient/1-amino acid-defined (CDAA) diet
as a liver-specific NASH model.8 Although both are
well established animal models of NAFLD, the former
involves only fatty liver, not steatohepatitis, whereas
the latter includes both steatohepatitis and liver cirrhosis.9 The mechanisms of fatty liver pathogenesis also
differ in that lipids accumulate in the liver because of
increased fatty acid (FA) inflow in the former model
and because of the blockage of very-low-density lipoprotein (VLDL) secretion in the latter.7,8
The specific mechanism by which lipids accumulate
in the liver in patients with NAFLD remains unknown,
and identifying this mechanism would advance efforts
toward the prevention and reversal of this condition.
Based on the findings of the above-mentioned animal
models, we hypothesized that blocking VLDL secretion might be a key factor in NASH pathogenesis in
humans. The hepatic profiles of lipid metabolites,
especially FA inflow and VLDL secretion, have not
been previously compared in human patients with
NAFL or NASH. In the present study, we investigated
the differences between NAFL and NASH in terms of
liver lipid metabolites and serum lipoprotein levels. We
compared the expression patterns, activities, and functional roles of hepatic lipid metabolites in patients with
NAFL and those with NASH.
FUJITA ET AL.
773
Patients and Methods
Subjects. The study protocol was reviewed and approved by an institutional ethical review committee.
Written informed consent was obtained from all the subjects before examination. In all, 104 consecutive subjects,
who were restricted to men and postmenopausal women
over 20 years of age, were analyzed in the present study.
Subjects with a history of excessive alcohol consumption
(weekly consumption: men, ⬎210 g; women, ⬎140 g),
other liver diseases, use of drugs associated with fatty liver,
weight reduction, etc., were excluded. A liver biopsy was
performed and three specimens obtained from the same
lobe and on the same occasion were obtained from each
subject using a 16-gauge needle biopsy kit according to a
standard protocol; three specimens were obtained to enable a sufficient sample size for analysis and to reduce
histological errors. All the subjects were histologically
confirmed as having NAFLD based on the liver biopsy
findings, and the diagnoses were further confirmed in
accordance with the categories defined by Matteoni et al.3
As a control group, 88 healthy subjects with a mean age
and sex ratio comparable to that of the NAFL and NASH
groups were also analyzed. All control subjects were confirmed to have normal liver enzyme levels, a liver/spleen
ratio of greater than 1.0 according to a computed tomography (CT) examination, and no other liver disease or
metabolic disorder.
Measurement of Serum Biochemical Markers. Venous blood samples were obtained after the subjects had
fasted overnight (12 hours) and were subsequently used to
measure the serum alanine aminotransferase (ALT), glucose, insulin, total cholesterol (TC), low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL)
cholesterol, triglyceride (TG), hyaluronic acid, free fatty
acid (FFA), free choline (fCh), chylomicron (CM)-TG,
and VLDL-TG levels. The serum TC, TG, FFA, and
plasma lipoprotein levels were analyzed using an online
dual-enzymatic method for the simultaneous quantification of cholesterol and triglycerides using high-performance liquid chromatography (HPLC) at Skylight
Biotech (Akita, Japan), according to the procedure described by Usui et al.10 CM (80-1,000 nm), VLDL1
(50-75 nm), VLDL2 (35-50 nm), large buoyant LDL
Address reprint requests to: Atsushi Nakajima, Division of Gastroenterology, Yokohama City University Graduate School of Medicine, 3-9 Fuku-ura, Kanazawa-ku,
Yokohama 236-0004, Japan. E-mail: [email protected]; fax: ⫹81-45-787-8988.
Copyright © 2009 by the American Association for the Study of Liver Diseases.
Published online in Wiley InterScience (www.interscience.wiley.com).
DOI 10.1002/hep.23094
Potential conflict of interest: Nothing to report.
774
FUJITA ET AL.
(25.5-30 nm), small dense LDL (17-25.5 nm), and HDL
(7-15 nm) were defined according to particle size.
We previously developed a simple and reliable method
for the quantitative detection of serum fCh using HPLC,
with a slight modification.10 Briefly, 4 ␮L of sample was
automatically injected and carried over the column’s surface using an elution buffer that had first been delivered
by way of a pump at a flow rate of 0.24 mL/min; the
column effluent was then mixed with an fCh reagent. The
enzymatic reagent was pumped at a flow rate of 0.12
mL/min, and the mixture was passed through a 15M
reaction-coil in a postcolumn reactor at 37°C. Finally, the
absorption was monitored using an ultraviolet-visible
(UV-VIS) detector at 550 nm, and data were collected
every 300 ms. The fCh concentrations were calculated by
comparing the results with the fCh area under the chromatographic curves of a calibration material (choline
chloride; TOYOBO) of known concentration. The degree of insulin resistance (IR) was calculated using the
homeostasis model for the assessment of insulin resistance
(HOMA-IR) according to the following formula: [fasting
serum insulin (␮U/mL) ⫻ fasting plasma glucose (mg/
dL)/405]. However, the HOMA-IR assessment was only
conducted in the 166 subjects with a fasting glucose level
of less than 140 mg/dL, because HOMA-IR has been
reported to be a suitable method for evaluating insulin
resistance only in such patients.11
Anthropometry and Abdominal Fat Distribution.
Abdominal fat distribution was determined using a CT
scanning method with the subjects in a supine position, in
accordance with a described procedure.12 The subcutaneous fat area (SFA) and the intraabdominal visceral fat area
(VFA) were measured at the level of the umbilicus and
determined using a standardized method based on the CT
number. The liver/spleen (L/S) ratio was also calculated.
A histogram representing the fat tissue was computed
based on the mean attenuation ⫹2 SD.
[11C]Choline Dynamic Positron Emission Tomography (PET) for Assessing Hepatic Phospholipid Metabolites. [11C]Choline was synthesized according to the
loop methylation method of Wilson et al.13 [11C]Choline
PET was performed as described.14 In brief, [11C]choline
PET was performed after 8 hours of fasting using an integrated PET scanner following the intravenous injection
of 370 MBq of [11C]choline. After a transmission scan,
[11C]choline was injected and the first emission scan was
started at 30 seconds after injection. The scans were performed continuously every 30 seconds for a total of 1,500
seconds. Once the scans were completed, the PET images
were reconstructed and the time-activity count was calculated for the appropriate liver sections.
Measurement of Hepatic TG Content. One and
HEPATOLOGY, September 2009
one-half liver biopsy samples per subject (25 mg of liver
tissue) were used to measure the hepatic triglyceride content. The liver biopsy samples were homogenized in 50
mM Tris/HCl buffer (pH 7.4) containing 150 mM
NaCl, 1 mM ethylene diamine tetraacetic acid (EDTA),
and 1 ␮M phenylmethylsulfonyl fluoride (PMSF). TGs
were analyzed enzymatically using a diagnostic kit (Infinity; Thermo DMA, Arlington, TX) and measured using
spectrophotometry (Beckman Coulter, Fullerton, CA).15
Histopathologic and Immunohistochemical Evaluations. Liver biopsy samples were excised and embedded
in Tissue-Tek OCT compound (Sakura Finetek, Torrance, CA) and paraffin for histological analysis. Fivemicrometer-thick sections were processed routinely using
hematoxylin-eosin staining. The presence of collagen, as
an index of fibrosis in the lesions, was examined in Masson’s trichrome-stained preparations. OCT-embedded
samples were serially sectioned at 4 ␮m. To evaluate fat
deposition and to determine the fat droplet size, the liver
sections were stained with Oil Red O and the diameters of
the fat droplets were measured using a microscope at a
40⫻ magnification using a grid of 0.0625 mm2 with 100
points. All histopathological findings were scored by the
same two pathologists (Y.N. and S.M.), who were unaware of the other biochemical and anthropometric parameters.
RNA Isolation, Reverse Transcription, and Quantification of Gene Expression Using Real-time ReverseTranscription Polymerase Chain Reaction (RT-PCR).
One-half of the liver biopsy samples per subject (10 mg of
liver tissue) was used to measure gene expression. Total
RNA was isolated from the samples using an RNeasy
Mini Kit (Qiagen, Hilden, Germany; Cat. No. 74126).
Reverse transcription to produce cDNA was performed
using a TaqMan Gold RT-PCR Kit (Applied Biosystems,
Foster City, CA), according to the manufacturer’s instructions. The reaction mixtures (100 ␮L) contained 2.5
␮g of total RNA, and the reaction was carried out for 50
minutes at 48°C; reverse transcriptase was then inactivated by heating the samples to 95°C for 5 minutes.16 The
messenger RNA (mRNA) levels of liver microsomal triglyceride transfer protein (MTTP), apolipoproteinB100
(apoB100), peroxisome proliferator activated receptor alpha
(PPAR-␣), sterol regulatory element binding protein-1c
(SREBP-1c), SREBP-2, diacylglycerol acyltransferase 1
(DGAT1), diacylglycerol acyltransferase 2 (DGAT2), perilipin, and adipose differentiation-relative protein (ADRP) as
well as a housekeeping gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), were determined in the liver tissue
using fluorescence-based RT-PCR and an ABI PRISM 7700
Sequence Detection System (Applied Biosystems). RT-PCR
was performed using the TaqMan Universal PCR Master
HEPATOLOGY, Vol. 50, No. 3, 2009
FUJITA ET AL.
775
Table 1. Clinical and Biochemical Characteristics of NAFL, NASH, and Control Groups
Number (n)
Age (years)
Gender (male; female)
BMI (kg/m2)
Visceral fat area (cm2)
Subcutaneous fat area (cm2)
Liver/spleen ratio
Fasting blood sugar (mg/dL)
Insulin (␮g/L)
HOMA-IR
LDL-cholesterol (mg/dL)
HDL-cholesterol (mg/dL)
Triacylglycerol (mg/dL)
ALT (U/mL)
Hyaluronic acid (ng/dL)
Fibrosis stage F0/F1/F2/F3/F4
Obesity (BMI ⬎ 25) (%)
Diabetes mellitus (%)
Hyperlipidemia (%)
Hypertension (%)
IR (HOMA-IR ⬎ 2.5) (%)
Metabolic syndrome (%)
Control
NAFL
NASH
88
49.2 ⫾ 13.3
41;47
21.1 ⫾ 2.4
76.9 ⫾ 14.2
134.8 ⫾ 42.4
1.08 ⫾ 0.08
82.3 ⫾ 20.1
4.22 ⫾ 1.83
0.83 ⫾ 0.29
75.6 ⫾ 20.7
64.8 ⫾ 14.8
82.4 ⫾ 38.1
17.2 ⫾ 8.1
6.0 ⫾ 3.9
—
0
0
0
0
0
0
51
47.1 ⫾ 13.6
24;27
28.6 ⫾ 6.1
121.6 ⫾ 38.2*
217.3 ⫾ 63.8*
0.70 ⫾ 0.26*
108.5 ⫾ 33.9*
9.87 ⫾ 3.51*
2.88 ⫾ 1.14*
133.1 ⫾ 37.7*
51.2 ⫾ 12.4*
159.5 ⫾ 71.1*
49.2 ⫾ 33.8*
13.2 ⫾ 8.4
51/0/0/0/0
60.8*
31.4*
37.3*
25.5*
53*
33.3*
53
53.6 ⫾ 12.8
26;27
28.2 ⫾ 5.7
119.4 ⫾ 35.3*
207.3 ⫾ 61.4*
0.71 ⫾ 0.29*
120.8 ⫾ 28.1*†
11.21 ⫾ 3.89*†
3.32 ⫾ 1.53*†
143.3 ⫾ 32.9*†
47.3 ⫾ 11.4*
173.3 ⫾ 62.6*
58.1 ⫾ 42.2*
46.5 ⫾ 31.2*†
0/32/15/6/0†
60.4*
39.6*
37.7*
30.2*
64.2*†
37.7*
Several parameters were measured and are expressed as the mean ⫾ SD. *P ⬍0.05, compared with control group. †P ⬍0.05, compared with NAFL group.
Mix reagent (Applied Biosystems).16,17 The values in all the
samples were normalized to the expression level of the endogenous control, GAPDH.16,18
Statistical Analysis. All results are expressed as the
mean ⫾ SD. A one-way Student’s t test was performed for
comparisons between two groups, and an analysis of variance (ANOVA) was performed to compare the means of
multiple groups, followed by an unpaired Student’s t test
for comparisons between two groups. All analyses were
two-sided, and P-values ⬍ 0.05 were considered statistically significant.
Results
Subject Characteristics. In all, 192 subjects (104
subjects with NAFLD, 88 control subjects) were analyzed
in this study. The main features of the NAFL, NASH, and
control groups are shown in Table 1. Compared with the
control group, both the NAFL and NASH groups had
high body mass index (BMI), VFA, SFA, fasting blood
sugar, insulin, HOMA-IR, ALT, TG, LDL-cholesterol,
hyaluronic acid scores, percentage of diabetes mellitus,
hyperlipidemia, hypertension, obesity, and IR values and
low L/S ratios and HDL-cholesterol scores (all P ⬍ 0.05)
(Table 1). Most of the parameters, such as BMI, VFA,
SFA, L/S ratios, HDL-cholesterol, TG, ALT, percent of
obesity, diabetes mellitus, hyperlipidemia, and hypertension, showed no significant difference between the NAFL
and NASH groups. Compared with the NAFL group,
however, the NASH group had a higher fasting blood
sugar, insulin, HOMA-IR, LDL-cholesterol, hyaluronic
acid score, fibrosis stage, and percentage of IR (all P ⬍
0.05) (Table 1).
Lipid Inflow (FA Inflow and Synthesis) in the
Liver. We analyzed the serum levels of CM-TG, glucose, and FFA and the hepatic expressions of SREBP1c
and SREBP2 mRNA as markers of lipid inflow in the
liver. The serum levels of CM-TG, glucose, and FFA
were significantly higher in both the NAFL and NASH
groups than in the control group, in spite of the serum
samples having been obtained after fasting (P ⬍ 0.05)
(Fig. 1A-C). We next compared the serum levels of
CM-TG, glucose, and FFA between the NAFL and
NASH groups, but no significant differences were observed (Fig. 1A-C). In support of these data, the hepatic expressions of SREBP1c and SREBP2 mRNA,
which are markers of FA inflow and synthesis, were
similar in the NAFL and NASH groups (Fig. 1D,E).
Moreover, the liver steatosis levels (L/S ratio, an index
of the hepatic fat content) and the TG content were
almost equal in the NAFL and NASH groups (Table 1;
Fig. 1F),
Lipid Outflow (VLDL Secretion and FA Oxidation) from the Liver. We analyzed the serum levels of
VLDL-TG, which is an index of lipid outflow from the
liver, and the hepatic expression of PPAR-␣ mRNA,
which stimulates FA oxidation in the liver. The serum
VLDL-TG levels were significantly higher in both the
NAFL and NASH groups than in the control group (P ⬍
776
FUJITA ET AL.
HEPATOLOGY, September 2009
Fig. 1. Lipid inflow in the liver. Subjects with NAFLD (NAFL ⫹ NASH) had significantly higher serum levels of CM-TG (A), glucose (B), and FFA (C)
than the control group (both P ⬍ 0.05), although no differences were noted between the NAFL and NASH groups. The hepatic mRNA expressions
of SREBP1c (D) and SREBP2 (E) were similar in the NAFL and NASH groups. The liver TG contents (F) in the NAFL and NASH groups were almost
equal. All data are expressed as the mean ⫾ SD.
0.05, Fig. 2A). Interestingly, the NASH patients had significantly lower serum VLDL-TG levels (P ⫽ 0.035)
(Fig. 2A), a lower expression of liver PPAR-␣ mRNA
(P ⫽ 0.039) (Fig. 2B) than the NAFL group.
Fig. 2. Lipid outflow from the liver. (A) Subjects with NAFLD had a
significantly higher serum level of VLDL-TG than the control group (P ⬍
0.05). However, the NASH group had a significantly lower serum level of
VLDL-TG than the NAFL group (P ⫽ 0.035). (B) The hepatic expression
of PPAR-␣ mRNA was significantly lower in the NASH group than in the
NAFL group (P ⬍ 0.001). All data are expressed as the mean ⫾ SD.
VLDL Synthesis in the Liver. We analyzed the hepatic expression of apoB100 mRNA, which is a structural protein of VLDL and controls the synthesis of
VLDL in the liver and the release of lipoprotein from
the liver. The NASH patients had a significantly lower
expression of liver apoB100 mRNA (P ⫽ 0.026) (Fig.
3A). In addition, we analyzed the scores for serum
fCh, which is required for lipoprotein synthesis. Contrary to our expectations, the serum fCh level was significantly higher in both the NAFL and NASH groups
than in the control group (P ⬍ 0.05); moreover, the
serum fCh level was significantly higher in the NASH
group than in the NAFL group (P ⫽ 0.022, Fig. 3B).
To investigate the levels of choline-phospholipid (PL)
metabolites in the liver, we measured [11C]choline uptake in the liver using [11C]choline dynamic PET imaging. As shown in Fig. 3C,D, [11C]choline
accumulation in the liver was significantly lower in the
NASH group (at all timepoints) than in the NAFL or
control groups (both P ⬍ 0.05). We also analyzed TG
transportation by way of MTTP mRNA expression,
HEPATOLOGY, Vol. 50, No. 3, 2009
FUJITA ET AL.
777
Fig. 3. VLDL synthesis in the liver. The hepatic mRNA expressions of apoB100 (A) and MTTP (E) were significantly lower in the NASH group than in the
NAFL group (P ⫽ 0.026, P ⬍ 0.001, respectively). (B) The serum fCh levels increased incrementally from the control group to the NAFL group and finally
to the NASH group (all P ⬍ 0.05). (C) Photographic images of each stage of pathogenesis in the control, NAFL, and NASH groups (early and late phases).
Depending on the amount of [11C]choline accumulation in the liver, the color signal changed from blue to red. (D) [11C]choline accumulation was significantly
lower in the NASH group (both timepoints) than in the NAFL or control groups (P ⬍ 0.05). The horizontal axis signifies the passage of time from the beginning
of [11C]choline injection. The vertical axis signifies the amount of [11C]choline accumulation in the liver. All data are expressed as the mean ⫾ SD.
because MTTP is a key factor for TG transport in the
liver. The hepatic expression of MTTP mRNA was
significantly lower in the NASH group than in the
NAFL group (P ⬍ 0.001) (Fig. 3E). The hepatic expression level of MTTP mRNA for differentiating between
NAFL and NASH, as determined using a receiver operating
characteristic (ROC) curve, was 0.775 (area under the curve,
0.910). This hepatic MTTP mRNA expression level had a
sensitivity of 91.3%, a specificity of 82.3%, a positive predictive value of 87.5%, and a negative predictive value of
87.5% for the differentiation of NAFL and NASH.
TG Metabolites in the Liver. We analyzed the hepatic
expressions of PAT family proteins (perilipin and ADRP),
which control lipid droplet accumulation and decomposition. The NASH patients had significantly higher expressions of hepatic perilipin1 and perilipin2 mRNA (P ⬍
0.001, P ⬍ 0.001, respectively) (Fig. 4A,B), whereas the
hepatic expressions of ADRP mRNA were similar in the
NAFL and NASH groups (Fig. 4C). Next we analyzed the
hepatic expressions of DGAT1 and DGAT2 mRNA, which
control TG synthesis in the liver. No significant differences
were observed in the NAFL and NASH groups (Fig. 4D,E).
Lipid Changes and Oxidative DNA Damage. Based
on the results mentioned above, we measured the
changes in the liver lipid droplet sizes and numbers, the
serum TG rich VLDL (VLDL1) ratio and small dense
LDL, and the liver expression of markers of oxidative
stress-induced DNA damage (malondialdehyde
[MDA], 8-hydroxydeoxyguanine [8-OHdG]). The
liver droplet sizes were significantly higher in the
NASH group than in the NAFL group (P ⫽ 0.043, Fig.
5A), whereas the liver droplet numbers tended to be
lower in the NASH group than in the NAFL group
(P ⫽ 0.062, Fig. 5B). Both the serum VLDL1 ratio and
small dense LDL, a serum marker of oxidation, were significantly higher in the NASH group than in the NAFL
group (P ⫽ 0.019, P ⫽ 0.043, Fig. 5C,D). Moreover, the
NASH group had significantly higher expressions of liver
MDA levels (P ⬍ 0.001) (Fig. 5E) and liver 8-OHdG
levels (P ⫽ 0.014) (Fig. 5F) than the NAFL group.
778
FUJITA ET AL.
HEPATOLOGY, September 2009
Fig. 4. TG metabolism in the liver. The hepatic mRNA expressions of perilipin1 (A) and perilipin2 (B) were significantly higher in the NASH group
than in the NAFL group (both P ⬍ 0.001). The hepatic mRNA expressions of ADRP (C), DGAT1 (D), and DGAT2 (E) were similar in the NAFL and
NASH groups. All data are expressed as the mean ⫾ SD.
Discussion
The key findings of this study are that NAFL and
NASH have distinct hepatic lipid metabolite profiles, and
different patterns of lipogenesis are seen in NAFL and
NASH. A scheme illustrating the differences in hepatic
lipid metabolism in NAFL and NASH is shown in Fig. 6.
The accumulation of fat, mainly TG, within the hepatocytes is a prerequisite for the development of NAFLD.19
The primary metabolic abnormalities leading to lipid accumulation are not well understood. TG is delivered to
adipose tissue by way of intestinal CM and hepatic
VLDL, and significant amounts of TGs are secreted by
the liver.20 These TGs are mainly delivered to the muscle,
heart, and adipose tissue.21 Fatty liver reportedly occurs
when the hepatic production of TG is not matched by its
secretion as VLDL or its degradation by ␤-oxidation.22 In
the present study, patients with NAFL and NASH exhibited highly similar tendencies and degrees of TG pooling
in the liver, but the manner of lipogenesis differed.
Namely, the capacity for lipid outflow from the liver was
deteriorated to a greater extent in NAFL than in NASH,
even though lipid inflow remained the same. These results
suggest that the blockage of VLDL-TG secretion and the
reduction in FA oxidation might be key factors in the
pathogenesis of NASH.
The blockage of VLDL-TG secretion, which can be
caused by a dietary choline deficiency, reportedly caused
steatohepatitis in an animal model.8,9 Choline deficiency
in animal models may be responsible for a lack of PLs,
more specifically of phosphatidylcholine (PC), which are
necessary for VLDL synthesis and secretion.23 However,
VLDL secretion from the liver was lower in the NASH
group than in the NAFL group in the present study, contrary to our expectations. The NASH group also had a
significantly higher serum fCh level than the NAFL
group. These phenomena seem to contradict each other,
because fCh is essential to VLDL synthesis by way of
phosphatidylcholine formation in the liver.24 To clarify
this discrepancy, we compared the ability to synthesize
VLDL in NAFL and NASH under the condition of a high
serum fCh level in NASH patients. In the present study,
[11C]choline accumulation in the liver was significantly
lower in the NASH group than in the NAFL and control
groups (all P ⬍ 0.05), in spite of the presence of abundant
fCh in the NASH group. Moreover, the hepatic expression of apoB100 mRNA and MTTP mRNA were impaired in the NASH group. A correlation between MTTP
activation and NASH has been suggested in some studies
using polymorphisms or MTTP knockout mice,25 but the
present finding is clear evidence that hepatic MTTP expression is significantly lower in patients with NASH than
in those with NAFL. Charlton et al.26 reported that
NASH is associated with a markedly altered hepatic synthesis of apoB100, compared with obese (BMI-matched)
controls without NASH. This report is in accordance
with our observation that the hepatic expression of
apoB100 mRNA was significantly lower in the NASH
group than in the NAFL group (P ⫽ 0.026). These results
suggest that VLDL synthesis is impaired in NASH, in
spite of the presence of abundant fCh in the serum, as a
result of some obstacle in choline utilization.
In the present study the hepatic TG content was found
to be similar between NAFL and NASH in spite of the
fact that VLDL synthesis and secretion was impaired to a
greater extent in NASH patients than in NAFL patients.
We next examined TG metabolism in the liver to elucidate how the surplus TGs were used. A drastic increase in
the mRNA expression of one member of the PAT family
of proteins, perilipin, might have led to the formation of
larger lipid droplets in NASH. Moreover, the surplus TGs
HEPATOLOGY, Vol. 50, No. 3, 2009
FUJITA ET AL.
779
Fig. 5. Lipid changes and oxidative DNA damage. The liver droplet size (A) was significantly greater in the NASH group than in the NAFL group
(P ⫽ 0.043), whereas the liver droplet number (B) tended to be smaller in the NASH group than in the NAFL group (P ⫽ 0.062). Both the serum
VLDL1 ratio (C) and the level of serum small dense LDL (D) were significantly higher in the NASH group than in the NAFL group (P ⫽ 0.019, P ⫽
0.043, respectively). The NASH group had significantly higher expression levels of liver MDA (E) (P ⬍ 0.001) and liver 8-OHdG (F) (P ⫽ 0.014) than
the NAFL group. All data are expressed as the mean ⫾ SD.
Fig. 6. Schema showing the changes in lipid flux through the liver in
NAFL and NASH. The blue arrows indicate no significant change between
NAFL and NASH. The red arrows indicate a significant increase or
decrease in NASH, compared with in NAFL. The red X shows the
dysfunctional point in NASH, compared with in NAFL.
would have been transported as a TG-rich lipoprotein
(VLDL1), leading to the formation of small dense LDL in
NASH. These alterations in lipid packaging might have
been caused by lipid oxidation and DNA oxidative damage, which are typically observed in patients with NASH
rather than patients with NAFL. The deposition of fat in
the liver, and, more specifically, the type of fat that is
deposited might in fact directly damage the liver and precipitate the development of NASH. Moreover, the accumulation of oxidative lipids in the liver is thought to play
a key role in the pathogenesis of NASH, as such increases
would activate hepatic nonparenchymal cells, including
Kupffer cells and hepatic stellate cells, leading to liver
injury during both the early stage of NASH and the late
fibrotic stage.27
Certainly, whether patients with NASH have had an
antecedent period of time in which they had NAFL or
whether these two populations diverge at the inception of
fatty liver remains unclear. A growing body of literature
suggests that deterioration in FA oxidation and VLDL
780
FUJITA ET AL.
secretion from the liver, caused by the impediment of
VLDL synthesis, might result in serious lipid oxidation
and DNA oxidative damage, impacting the degree of liver
injury and therefore contributing to the progress of
NASH. Therefore, dysfunctional VLDL synthesis and release might be a key factor in NASH pathogenesis.
Several potential drug therapies have been tested in
animal models of NASH.9,28 Although large studies (multicenter clinical trials, case-control studies, and family
studies) along with detailed analyses using animal models
are needed to define the precise mechanism involved in
the development and progression of NASH, the measurement of hepatic MTTP expression may be helpful for
diagnosis. Moreover, the development of drugs capable of
reactivating hepatic VLDL synthesis may provide a novel
treatment for NASH.
Acknowledgment: We thank Machiko Hiraga for
skillful technical assistance.
References
1. Lazo M, Clark JM. The epidemiology of nonalcoholic fatty liver disease: a
global perspective. Semin Liver Dis 2008;28:339-350.
2. Wieckowska A, McCullough AJ, Feldstein AE. Noninvasive diagnosis and
monitoring of nonalcoholic steatohepatitis: present and future. HEPATOLOGY 2007;46:582-589.
3. Matteoni CA, Younossi ZM, Gramlich T, Boparai N, Liu YC, McCullough AJ, et al. Nonalcoholic fatty liver disease: a spectrum of clinical and
pathological severity. Gastroenterology 1999;116:1413-1419
4. Farrell GC. Non-alcoholic steatohepatitis: what is it, and why is it important in the Asia-Pacific region? J Gastroenterol Hepatol 2003;18:124-138.
5. Matthews DR, Hosker JP, Rudenski AS. Homeostasis model assessment:
insulin resistance and beta-cell function from fasting plasma glucose and
insulin concentrations in man. Diabetalogia 1985;28:412-419.
6. Marchesini G, Bugianesi E, Forlani G, Cerrelli F, Lenzi M, Manini R, et al.
Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome.
HEPATOLOGY 2003;37:917-923.
7. Deng XQ, Chen LL, Li NX. The expression of SIRT1 in nonalcoholic fatty
liver disease induced by high-fat diet in rats. Liver Int 2007;27:708-715.
8. Kawaguchi K, Sakaida I, Tsuchiya M, Omori K, Takami T, Okita K, et al.
Pioglitazone prevents hepatic steatosis, fibrosis, and enzyme-altered lesions
in rat liver cirrhosis induced by a choline-deficient L-amino acid-defined
diet. Biochem Biophys Res Commun 2004;27:187-195.
9. Fujita K, Nozaki Y, Yoneda M, Wada K, Saito S, Nakajima A. et al.
Effectiveness of antiplatelet drugs against experimental non-alcoholic fatty
liver disease. Gut 2008;57:1583-1591.
10. Usui S, Hara Y, Hosaki S, Okazaki M. A new on-line dual enzymatic
method for simultaneous quantification of cholesterol and triglycerides in
lipoproteins by HPLC. J Lipid Res 2002;43:805-814.
11. Turner RC, Holman RR, Matthews D, Hockaday TD, Peto J. Insulin
deficiency and insulin resistance interaction in diabetes: estimation of their
HEPATOLOGY, September 2009
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
relative contribution by feedback analysis from basal plasma insulin and
glucose concentrations. Metabolism 1979;28:1086-1096.
Yoshizumi T, Nakamura T, Yamane M, Islam AH, Menju M, Yamasaki K,
et al. Abdominal fat: standardized technique for measurement at CT.
Radiology 1999;211:283-286.
Wilson AA, Garcia G, Jin L, Houle S. Radiotracer synthesis from [11C]iodomethane: a remarkably simple captive solvent method. Nucl Med Biol
2000;27:529-532.
Reske SN, Blumstein NM, Neumaier B, Gottfield HW, Finsterbusch F,
Kocot D, et al. Imaging prostate cancer with 11C-choline PET/CT. J Nucl
Med 2006;47:1249-1254.
Sahai A, Malladi P, Pan X, Paul R, Melin-Aldana H, Whitington PF, et al.
Obese and diabetic db/db mice develop marked liver fibrosis in a model of
nonalcoholic steatohepatitis: role of short-form leptin receptors and osteopontin. Am J Physiol Gastrointest Liver Physiol 2004;287:1035-1043.
User Bulletin 2. ABI PRISM 7700 Sequence Detection System. Foster
City, CA: Applied Biosystems; December 11, 1997 (updated October
2001).
Marshall SE, Welsh KI. Tumor necrosis factor polymorphisms and acute
rejection: proven link or provisional association? Transplantation 2000;69:
1237-1239.
Zegarska J, Paczek L, Mroz A. Quantitative gene expression of TGF-␤1,
TNF-␣, IL-1␤ and IL-6 in the renal artery wall of chronically rejected
human renal allografts. Transpl Proc 2002;34:3176-3179.
Donnelly KL, Smith CI, Schwarzenberg SJ, Jessurun J, Boldt MD, Parks
EJ, et al. Sources of fatty acids stored in liver and secreted via lipoproteins
in patients with nonalcoholic fatty liver disease. J Clin Invest 2005;115:
1343-1351.
Lee RG, Shah R, Sawyer JK, Hamilton RL, Parks JS, Rudel LL, et al.
ACAT2 contributes cholesteryl esters to newly secreted VLDL, whereas
LCAT adds cholesteryl ester to LDL in mice. J Lipid Res 2005;46:12051212.
Havel RJ. Lipoproteins and lipid transport. Adv Exp Med Biol 1975;63:
37-59.
Hermier D, Salichon MR, Guy G, Peresson R. Differential channeling of
liver lipids in relation to susceptibility to hepatic steatosis in the goose.
Poultry Sci 1999;78:1398-1406.
Yao Z, Vance DE. The active synthesis of phosphatidylcholine is required
for very low density lipoprotein secretion from rat hepatocytes. J Biol
Chem 1988;263:2998-3004.
Li Z, Agellon LB, Vance DE. Phosphatidylcholine homeostasis and liver
failure. J Biol Chem 2005;280:37798-37802.
Gambino R, Cassader M, Pagano G, Durazzo M, Musso G. Polymorphism in microsomal triglyceride transfer protein: a link between liver
disease and atherogenic postprandial lipid profile in NASH. HEPATOLOGY
2007;45:1097-1107.
Charlton M, Sreekumar R, Rasmussen D, Lindor K, Nair S. Apolipoprotein synthesis in nonalcoholic steatohepatitis. HEPATOLOGY 2002;35:898904.
kura Y, Ohsawa M, Suekane T, Kawada N, Arakawa T, Ueda M, et al.
Localization of oxidized phosphatidylcholine in nonalcoholic fatty
liver disease: impact on disease progression. HEPATOLOGY 2006;43:
506-514.
Fujita K, Yoneda M, Wada K, Saito S, Terauchi Y, Nakajima A, et al.
Telmisartan, an angiotensin II type 1 receptor blocker, controls progress of
nonalcoholic steatohepatitis in rats. Dig Dis Sci 2007;52:3455-3464.