Am J Physiol Gastrointest Liver Physiol 288: G1179 –G1189, 2005.
First published December 9, 2004; doi:10.1152/ajpgi.00411.2004.
Hepatic triglyceride contents are genetically determined in mice:
results of a strain survey
Xiaobo Lin, Pin Yue, Zhouji Chen, and Gustav Schonfeld
Department of Medicine, Washington University School of Medicine, St. Louis, Missouri
Submitted 13 September 2004; accepted in final form 7 December 2004
fatty liver; genetics of hepatic steatosis; lipogenesis; fatty acid oxidation; very low-density lipoprotein
NONALCOHOLIC FATTY LIVER DISEASE (NAFLD) is highly prevalent (9, 20). Although it resembles alcoholic fatty liver histopathologically, NAFLD, by definition, develops in the absence of alcohol abuse (9). The majority of patients with
NAFLD are affected with various aspects of the metabolic
syndrome, i.e., obesity, hypertension, insulin resistance, Type
2 diabetes mellitus, and hyperlipidemia (4). In the setting of
insulin resistance, hepatic lipogenesis may be stimulated sufficiently to overcome the adaptive capacities of the liver to
secrete more VLDL and result in the accumulation of triglycerides (TG) in the liver (33). On the other hand, insulin
resistance may be the consequence of chronic lipid accumulation in insulin-sensitive tissues. For example, muscle- and
liver-specific overexpression of lipoprotein lipase causes TG
accumulation in muscle and liver as well as impairment of
insulin signaling and action in the respective tissues (17). A
Address for reprint requests and other correspondence: G. Schonfeld, 660 S.
Euclid, Campus Box 8046, St. Louis, MO 63110 (E-mail: [email protected]).
http://www.ajpgi.org
strong relationship between fatty liver and insulin resistance
also is established in patients (26, 27).
NAFLD poses a major health problem because it may lead
to steatohepatitis, cirrhosis, and liver failure. The pathophysiological mechanisms by which steatosis may progress to steatohepatitis, fibrosis, and cirrhosis remain unclear, but steatosis
appears to be an essential precursor (10). Thus controlling
hepatic TG levels may be important in reducing the risk of
developing chronic liver disease.
Fatty liver is due to an imbalance between the availability of
hepatic TG for export and the disposal system of the liver (1,
20, 36). Both genetic and environmental factors are likely to be
important. The major sources of hepatic fatty acids (FA) for
TG assembly are those synthesized de novo in the liver and
free FAs (FFA) derived from plasma. Genes involved in de
novo lipogenesis are controlled by sterol regulatory elementbinding protein (SREBP)-1c (40), a mediator of both nutritional stimulus and insulin action in hepatic lipogenesis (16).
Mice overexpressing SREBP-1a develop severe fatty liver due
to overproduction of cholesterol and FAs (32). Elevated levels
of FFA in plasma, such as those in fasting, increase delivery of
FFA to the liver, which may cause excessive hepatic TG
accumulation despite an accompanying increased FA oxidation (15).
The major disposal routes for hepatic FAs and TG are
composed of the export of hepatic TG via VLDL (39) and the
oxidation of component FAs in mitochondria, peroxisomes,
and microsomes (41). Here too, both genetic and environmental factors are important. For example, the impaired capacity
for FA ␤-oxidation on fasting in peroxisome proliferatoragonist receptor (PPAR)-␣ null mice results in the accumulation of hepatic TG (18). A defective VLDL transport system
that impairs hepatic lipid-exporting capacity also may result in
fatty liver. This is seen in abetalipoproteinemia, caused by
mutations in microsomal triglyceride transfer protein that is
essential in the assembly of chylomicrons and VLDL particles
(37), and in familial hypobetalipoproteinemia due to truncation-specifying mutations in apolipoprotein (apo)B, the indispensable structural protein in the formation and secretion of
VLDL (6, 31, 35).
Thus the major genes for developing fatty liver, identified to
date, are selected genes in the FA synthetic and oxidative
pathways and in the VLDL export system. However, modifier
genes (22, 23) as yet not identified may be contributing to the
development of steatosis. This is suggested by the finding that
engineered apoB gene mutation-bearing mice (e.g., apoB⫹/38.9
and apoB⫹/27.6 heterozygotes) produced on mixed C57BL/
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
0193-1857/05 $8.00 Copyright © 2005 the American Physiological Society
G1179
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
Lin, Xiaobo, Pin Yue, Zhouji Chen, and Gustav Schonfeld.
Hepatic triglyceride contents are genetically determined in mice:
results of a strain survey. Am J Physiol Gastrointest Liver Physiol
288: G1179 –G1189, 2005. First published December 9, 2004;
doi:10.1152/ajpgi.00411.2004.—To assess whether genetic factor(s)
determine liver triglyceride (TG) levels, a 10-mouse strain survey of
liver TG contents was performed. Hepatic TG contents were highest
in BALB/cByJ, medium in C57BL/6J, and lowest in SWR/J in both
genders. Ninety and seventy-six percent of variance in hepatic TG in
males and females, respectively, was due to strain (genetic) effects. To
understand the physiological/biochemical basis for differences in
hepatic TG among the three strains, studies were performed in males
of the BALB/cByJ, C57BL/6J, and SWR/J strains. In vivo hepatic
fatty acid (FA) synthesis rates and hepatic TG secretion rates ranked
BALB/cByJ ⬇ C57BL/6J ⬎ SWR/J. Hepatic 1-14C-labeled palmitate
oxidation rates and plasma ␤-hydroxybutyrate concentrations ranked
in reverse order: SWR/J ⬎ BALB/cByJ ⬇ C57BL/6J. After 14 h of
fasting, plasma-free FA and hepatic TG contents rose most in BALB/
cByJ and least in SWR/J. ␤-Hydroxybutyrate concentrations rose least
in BALB/cByJ and most in SWR/J. Adaptation to fasting was most
effective in SWR/J and least in BALB/cByJ, perhaps because BALB/
cByJ are known to be deficient in SCAD, a short-chain FA oxidizing
enzyme. To assess the role of insulin action, glucose tolerance test
(GTT) was performed. GTT-glucose levels ranked C57BL/6J ⬎
BALB/cByJ ⬇ SWR/J. Thus strain-dependent (genetic) factors play a
major role in setting hepatic TG levels in mice. Processes such as FA
production and hepatic export in VLDL on the one hand and FA
oxidation on the other, explain some of the strain-related differences
in hepatic TG contents. Additional factor(s) in the development of
fatty liver in BALB/cByJ remain to be demonstrated.
G1180
GENETICS OF LIVER TRIGLYCERIDE LEVEL IN MICE
MATERIALS AND METHODS
Mice and Diet
Ten inbred strains of mice (6 wk old), AKR/J (AKR), BALB/c,
C3H/HeJ (C3H), C57BL, C57BL/6ByJ (C57BLBY), DBA/1LacJ
(DBA), 129/SVJ, NZB/B1NJ (NZB), PL/J (PL), and SWR, of both
genders were purchased from Jackson Laboratory (Bar Harbor, ME).
AKR males were not studied because only AKR females were
available from the supplier. Mice were housed in a room maintained
at 24°C with a 12:12-h light-dark cycle (6:00 AM to 6:00 PM). All
mice were given Purina mouse chow 5053 containing 4.5% fat, 20.0%
protein, and 54.8% carbohydrate (LabDiet). Food was removed in the
beginning of the light cycle, and mice were fasted for 4 – 6 h, except
in the fasting experiment where times are indicated. Mice were 11–12
wk of age when killed. All animal procedures were performed in
accordance with guidelines of Washington University’s Animal Studies Committee and approved by the IACUC of Washington University.
Analytical Procedures
Body fat (%) was assessed on anesthetized living mice by dualenergy X-ray absorptiometry with the use of a small animal densitometer (Lunar) (3). Plasma samples were assayed for triglycerides,
total cholesterol, and free cholesterol as described (7, 19). Lipids were
extracted from liver and assayed for TG (7). Hepatic TG levels were
somehow lower in the three strains compared with their counterparts
during the survey carried out at 4 – 6 h of fasting, probably due to the
use of a different TG analysis kit from a supplier, which systematically did yield lower hepatic TG levels when used on the surveyed
livers (data not shown). The same kit was used for all of the fasting
experiment. Cellular protein contents were determined as described
(7). Hepatic TG contents were expressed as milligrams of lipid per
gram of protein. Plasma FFA were analyzed by an enzymatic method
(Wako). Blood glucose was determined using a B-glucose analyzer
(Hemocue, Ängelholm, Sweden). The Core Laboratory of the General
Clinical Research Center at Washington University analyzed plasma
␤-hydroxybutyrate (BHB) concentration. Mice were weighed and
killed. Blood was obtained from the inferior vena cava. Livers were
excised, washed in cold PBS, weighed, cut into pieces that were then
frozen in liquid nitrogen, and stored at ⫺80°C until further analysis.
Glucose Tolerance Tests
After mice were fasted for 5 h, mice received an intraperitoneal
injection of 10% D-glucose (0.75 g/kg body wt) for glucose tolerance.
Blood (⬃10 ␮l) was drawn from the tail vein at 0, 30, 60, and 120 min
and assayed for glucose.
In Vivo Measurement of FA Synthesis
Mice were injected with [1-14C]acetate (0.25 Ci/ml) at 25 ␮Ci/30
g body wt and killed after 1.0 h. Liver (300 mg) was excised and
digested in potassium hydroxide. After extraction of the nonsaponifiable lipids with petroleum ether, the sample solution containing the
saponified lipids was acidified with sulfuric acid and FAs were
extracted with petroleum ether as described (11). The radioactivity in
total FAs was determined by scintillation counting. The FA synthesis
rates are reported as disintegrations per hour per milligram of liver.
Determination of In Vivo Hepatic TG Secretion Rates
Hepatic production of VLDL-TG was measured after injection
(intravenous) of Triton WR1339 (500 mg/kg body wt) on mice fasted
for 4 h (7). Tail vein blood samples were taken at the specified times
after injection for TG measurement and measured as described above.
Determination of FA Oxidation in Primary Cultures of Hepatocytes
Primary mouse hepatocytes were isolated as described (7). Cells
(1.0 ⫻ 106) were cultured in a flask for 2 h in 4 ml DMEM containing
10% FBS, and medium was then changed to DMEM containing 5%
FBS. After an overnight culture, cells were washed three times with
PBS. To determine FA oxidation, 14C-labeled palmitate (0.25 ␮Ci/ml)
was added to the medium. At the end of 2-h incubation, a portion of
the cell medium (250 ␮l) was removed to determine acid-soluble
products (ASP) as described (13). A septum and center well were then
fitted into the flask. Sodium hydroxide (2 N, 0.25 ml) was injected
onto the filter paper placed in the center well to capture CO2.
Hydrochloric acid (6 N, 2 ml) was injected into the medium to release
14
CO2. Both ASP and 14CO2 were counted. Two separate flasks
treated in the same way without the 14C-labeled palmitate were used
for assay of cellular protein. Total FA oxidation activity was obtained
by adding the counts of 14CO2 and ASP and expressed as DPM per
hour per microgram of cellular protein.
Short-chain Acyl-CoA Dehydrogenase Mutation Assay
A PCR assay was used to distinguish between the wild-type and
mutant alleles for the structural gene of short-chain acyl-CoA dehydrogenase (SCAD), which employs oligonucleotide primers that flank
a 278-bp deletion in the mutant allele to produce PCR products of 870
and 592 bp for the wild-type and mutant alleles for SCAD, respectively (38).
Responses of Liver TG Levels to Fasting
To determine whether there was a differential susceptibility of the
fasting raising effect on liver TG levels, male mice of BALB/c,
C57BL, and SWR were killed at 0, 6, and 14 h of fasting. Liver TG
contents, plasma FFA, and BHB concentrations were determined as
described in Analytical Procedures. A separate kit (Thermo Electron,
Melbourne, Australia) was used to determine hepatic TG levels,
because Wako discontinued its TG kit used in this study. The latter
kit, although yielding precise results, deviated systematically downward from the Wako kit, yielding somewhat lower values for the
fasting study (see RESULTS).
mRNA Expression Profiling
Dual-channel microarray analysis was performed on total liver
RNA pooled from each strain (male, n ⫽ 5 for each strain). Extracted
RNA was further purified using RNeasy spin columns (Qiagen,
AJP-Gastrointest Liver Physiol • VOL
288 • JUNE 2005 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
6JX129/SVJ genetic backgrounds manifest significantly elevated hepatic TG on average, but interindividual variation is
high (7, 8, 31), probably reflecting the heterogeneity of the
genetic backgrounds. Indeed, although groups of these mice on
average contain 50% C57BL/6J and 50% 129/SVJ genes, the
genetic complements of individual animals may deviate materially from the 50/50 mean (P. Yue, X. Lin, and G. Schonfeld,
unpublished observation). Interindividual variation of hepatic
TG levels should be significantly smaller in congenic strains
bearing apoB truncation-producing mutations than in the apoB
mutation-bearing mice of 50/50% C57BL/6J and 129/SVJ
mixed background.
To identify potentially useful parental strains for the production of congenic mice, we performed a mouse-strain survey
of liver TG levels in 10 inbred strains, the results of which
form the bulk of this presentation. Significant strain-related
differences in hepatic TG contents were found. To ascertain
which physiological, biochemical processes may play roles in
determining the differences in liver TG contents among the
selected strains, experiments, including hepatic mRNA expression profiling, were performed in the strains with the highest
[BALB/cByJ (BALB/c)], middle [C57BL/6J (C57BL)], and
lowest [SWR/J (SWR)] hepatic TG contents.
GENETICS OF LIVER TRIGLYCERIDE LEVEL IN MICE
G1181
Valencia, CA) following the manufacturer’s protocol. Purified RNA
was quantitated by ultraviolet absorbance at 260 and 280 nm and
assessed qualitatively using Bioanalyzer 2100 (Agilent, Palo Alto,
CA). Three comparison pairs (BALB/c vs. C57BL, BALB/c vs. SWR,
C57BL vs. SWR) were set up comparing against each pair of two
strains of mice. In each pair, 5 ␮g of purified total RNA were
converted to cDNA that was labeled either with Cy3 or Cy5, which
were hybridized to mouse Oligo Array (Sigma 65-mer Probe Set)
containing 21,676 transcripts. In “dye-flip” experiments with the same
pair of comparison, the dye labeling was reversed. Oligonucleotide
array analysis was performed by the Genomic Core Facility of
Digestive Diseases Research Core Center at Washington University in
St. Louis, MO. All protocols were performed as recommended by
Genisphere (Hatfield, PA).
Array images were scanned on an Axon scanner. Array data were
extracted and analyzed using GenePix Pro 4.1 software from Axon
Instruments. Further analysis was performed by using BRB-Array
Tools (version 3.1, http://linus.nci.nih.gov/BRB-ArrayTools.html) according to the instructions provided. Differential gene expression
between two of the three strains (BALB/c, SWR, and C57BL) was
done by using class comparison expression analysis.
sequence-detection system using SYBR Green dye binding to the
PCR product (5). Primer Express 2.0 software was used to design
amplimers for both genes of interest and GAPDH, the latter of which
was used for normalization in RT-PCR.
Quantitative Real-Time RT-PCR
Liver TG levels, averaged on both genders, varied over a
sixfold range among the strains, from 49.7 ⫾ 5.5 mg/g protein
(SWR) to 316.5 ⫾ 25.8 (BALB/c; Fig. 1, A and B). Liver TG
contents showed similar strain-related differences across both
genders. The presence of the SCAD mutation was confirmed in
All statistic analyses were conducted using SAS (version 9, SAS
Institute, Cary, NC). Data are expressed as means ⫾ SD. ANOVA
(PROC GLM) followed by either Duncan’s multiple tests or t-tests
were performed for comparisons between treatments as appropriate
with an overall ␣-level of 0.01 or 0.05 as indicated. Pearson’s
correlation was performed using SAS Proc CORR using data from all
10 strains. The total variance (Vtotal) in the general linear model was
partitioned into variance due to “environment” (Venv) and variance
due to strain or genetic effects (Vstrain), where Vstrain/Vtotal gave an
estimate of Vtotal explained by strain effect, an indication of heritability (12).
RESULTS
Hepatic TG Levels
Fig. 1. Strain survey. Hepatic triglyceride (TG) levels of male
(A) and female (B) mice of 10 different strains. Values are
means ⫾ SD, n ⫽ 5 for each gender and strain. The multiple
comparisons of means were obtained using Duncan’s multiple
range test with an overall ␣ of 0.01. Mean values with different
lowercase letters differ from each other (P ⬍ 0.01).
AJP-Gastrointest Liver Physiol • VOL
288 • JUNE 2005 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
Total RNA from male BALB/c, C57BL, and SWR each was pooled
and digested with RNase-free DNase, followed by purification with
RNeasy Mini Kit (Qiagen). mRNAs of interest were quantified by
fluorescence RT-PCR with an Applied Biosystems GeneAmp 5700
Statistics
G1182
GENETICS OF LIVER TRIGLYCERIDE LEVEL IN MICE
Table 1. Strain survey: body and liver weights and body fat
BW, g
Strain
BALB/c
C57BL
C57BLBY
129/SVJ
AKR
NZB
PL
DBA
C3H
SWR
Body Fat, %
Male
Female
a
25.3⫾2.7
27.4⫾0.7a
27.6⫾3.2a
25.0⫾1.8a
NA
26.5⫾3.3a
24.1⫾2.0ab
24.6⫾1.1ab
26.0⫾0.7a
20.8⫾2.8b
Male
bc
Female
c
20.9⫾1.7
18.4⫾0.8cde
19.0⫾2.2cd
20.3⫾2.0bcd
26.8⫾1.7a
22.3⫾0.4b
15.7⫾0.6f
18.3⫾0.8de
19.7⫾1.0cd
16.5⫾0.4ef
LW, g
12.2⫾1.4
13.5⫾2.2c
13.7⫾0.6c
13.7⫾0.8c
NA
15.3⫾2.1bc
20.9⫾3.1a
18.5⫾4.5ab
14.3⫾1.0c
12.2⫾1.9c
LB, %
Male
abc
Female
a
16.4⫾3.8
15.4⫾2.6bcd
13.0⫾0.9cd
14.5⫾1.7bcd
19.5⫾2.9a
12.3⫾1.1d
18.1⫾1.8ab
12.1⫾1.4d
12.5⫾1.6cd
12.8⫾1.4cd
1.49⫾0.05
1.20⫾0.06bc
1.31⫾0.20ab
0.98⫾0.10cd
NA
1.03⫾0.20cd
0.95⫾0.07d
1.07⫾0.04cd
1.17⫾0.06bcd
1.13⫾0.07bcd
Male
b
1.08⫾0.08
0.80⫾0.08cd
0.94⫾0.12bc
0.75⫾0.15d
1.32⫾0.13a
1.00⫾0.03b
0.67⫾0.08d
0.77⫾0.06d
0.94⫾0.11bc
0.64⫾0.03d
Female
a
5.3⫾0.2
4.4⫾0.1b
4.7⫾0.2b
3.9⫾0.2c
NA
3.8⫾0.3c
3.9⫾0.2c
4.4⫾0.1b
4.5⫾0.2b
4.7⫾0.2b
5.2⫾0.2a
4.3⫾0.3cd
5.0⫾0.2ab
3.6⫾0.4e
4.9⫾0.3ab
4.5⫾0.2bc
4.3⫾0.5cd
4.2⫾0.4cd
4.8⫾0.4abc
3.9⫾0.1de
both BALB/c males and females (data not shown), but it was
not detected in any of the other strains (data not shown).
ranging from 9.3 ⫾ 1.0 mg/dl for BALB/c to 23.8 ⫾ 2.0 mg/dl
for NZB; Table 2).
Body Weight, Body Fat, Liver Weight, and Liver-to-Body
Weight Ratio
Plasma BHB and Glucose Concentrations
Body weights ranged from 25.3 ⫾ 2.7 g in BALB/c males to
20.8 ⫾ 2.8 g in SWR males. Among females, AKR were
heaviest (26.8 ⫾ 1.7 g), and PL were the lightest (24.1 ⫾ 2.0 g;
Table 1). Body fat contents were highest in AKR females
(males were not available), followed by PL of both genders
(Table 1). BALB/c and SWR males and NZB females were
among the lowest. With respect to liver-to-body weight ratio,
BALB/c males and females ranked at the top, and 129/SVJ
ranked at the bottom (Table 1).
Plasma Lipids
Plasma TG concentrations varied approximately threefold,
ranging from 59.2 ⫾ 16.6 (C57BLBY) to 173.8 ⫾ 12.4 mg/dl
(BALB/c) for males and from 61.1 ⫾ 17.5 (C57BL) to 203.8 ⫾
22.7 mg/dl (PL) in females (Table 2). Plasma TC ranged
⬃2.5-fold, from 70.6 ⫾ 5.0 (C57BLBY) to 153.1 ⫾ 4.7 mg/dl
(NZB) for males and from 63.3 ⫾ 7.7 (C57BLBY) to 164.8 ⫾
10.1 mg/dl (C3H) in females (Table 2). The range of concentrations for plasma-free cholesterol was greater in males (4.3fold difference ranging from 6.1 ⫾ 1.0 mg/dl for BALB/c to
26.2 ⫾ 1.1 mg/dl for NZB) than in females (2.6-fold difference
Plasma BHB concentrations were alike in males (although
male SWR had the highest level numerically) but differed
among strains in females (Table 3). SWR had the highest
concentration, whereas PL had the lowest. Significant differences in glucose concentration existed among the strains of
both genders. The C57BL strain had the highest plasma glucose concentrations for both males and females (Table 3).
Correlation and Variance Component Analysis
Many of the parameters reported above have been shown to
correlate with liver fat in humans (31). Therefore, a series of
correlations was sought in an attempt to identify correlates of
hepatic fat in mice. Correlations between body weight and
hepatic TG were significant only in males (data not shown).
Correlations between body fat and hepatic TG were significant
only in females. Hepatic TG level was significantly and positively correlated with liver-to-body weight ratio for both males
and females (Table 4). Hepatic TG was positively correlated
with plasma TG only in males. No significant correlations
existed between hepatic TG and TC (Table 4). However,
hepatic TG level was negatively correlated with plasma-free
cholesterol level for both genders (Table 4). No significant
Table 2. Strain survey: plasma lipid concentrations
Plasma TG, mg/dl
Plasma TC, mg/dl
Plasma FC, mg/dl
Strain
Male
Female
Male
Female
Male
Female
BALB/c
C57BL
C57BLBY
129/SVJ
AKR
NZB
PL
DBA
C3H
SWR
173.8⫾12.4a
88.1⫾4.2de
59.2⫾16.6e
153.9⫾16.3ab
NA
113.8⫾8.4cd
71.6⫾11.5e
125.5⫾29.9bcd
139.0⫾32.7abc
95.9⫾13.6de
171.3⫾33.9ab
61.1⫾17.5d
125.2⫾20.0bc
152.3⫾59.8abc
178.2⫾18.8ab
69.0⫾16.3d
203.8⫾22.7a
113.8⫾19.4cd
177.2⫾48.0ab
174.3⫾14.6ab
88.4⫾10.0c
84.7⫾4.3c
70.6⫾5.0d
108.3⫾9.8b
NA
153.1⫾4.7a
81.3⫾2.9cd
76.4⫾4.2cd
114.2⫾4.3b
83.5⫾6.5cd
86.5⫾17.3cd
73.1⫾7.0de
63.3⫾7.7e
99.2⫾9.8c
69.5⫾0.7de
124.2⫾16.7b
96.8⫾4.5c
66.6⫾4.5e
164.8⫾10.1a
73.5⫾9.1de
6.1⫾1.0d
8.0⫾0.8d
13.5⫾0.3c
11.2⫾0.8c
NA
26.2⫾1.1a
14.2⫾0.5c
12.6⫾0.4c
18.2⫾0.6b
11.6⫾0.6c
9.3⫾1.0c
9.4⫾1.0c
9.7⫾0.7c
13.0⫾0.7bc
10.3⫾1.7c
23.8⫾2.0a
17.0⫾0.9b
13.0⫾0.7bc
22.1⫾1.9a
15.6⫾0.8b
Values are means ⫾ SD (n ⫽ 5 males and 5 females for each strain). Duncan’s multiple tests were used to compare the differences of means with an overall
␣ ⫽ 0.01. Values in the same column with different superscripts are different (P ⬍ 0.01). TG, triglycerides; TC, total cholesterol; FC, free cholesterol.
AJP-Gastrointest Liver Physiol • VOL
288 • JUNE 2005 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
Values are means ⫾ SD (n ⫽ 5 males and 5 females for each strain). Duncan’s multiple tests were used to compare the differences of means with an overall
␣ ⫽ 0.01. Values in the same column with different superscripts are different (P ⬍ 0.01). BW, body wt; LW, liver wt; LB, liver-to-body wt ratio; NA, not
applicable.
G1183
GENETICS OF LIVER TRIGLYCERIDE LEVEL IN MICE
Table 3. Strain survey: plasma BHB and
glucose concentrations
Blood BHB, mM
Strain
BALB/c
C57BL
C57BLBY
129/SVJ
AKR
NZB
PL
DBA
C3H
SWR
Male
0.87⫾0.13
0.88⫾0.19
0.94⫾0.16
1.03⫾0.31
NA
0.77⫾0.25
0.87⫾0.18
0.69⫾0.71
0.69⫾0.18
1.12⫾0.35
Glucose, mg/dl
Female
Male
abc
0.86⫾0.21
1.08⫾0.20ab
0.84⫾0.10abc
0.86⫾0.29abc
0.71⫾0.12bc
0.80⫾0.19bc
0.63⫾0.18c
0.83⫾0.21abc
0.82⫾0.12abc
1.20⫾0.31a
Female
d
132.2⫾17.6
198.0⫾13.6a
173.8⫾9.9abc
154.2⫾22.2cd
NA
182.9⫾12.9ab
186.0⫾22.2ab
169.9⫾6.2abc
160.4⫾9.4bc
163.7⫾16.3bc
129.4⫾8.6ef
188.3⫾34.2a
184.6⫾9.5ab
142.5⫾12.5def
151.7⫾5.6cde
150.1⫾8.8cde
176.1⫾5.5abc
159.2⫾10.5bcd
142.7⫾8.8def
119.8⫾10.4f
correlations existed between hepatic TG levels and plasma
BHB concentrations (Table 4). Thus univariate analysis did not
identify any consistent covariates of hepatic TG except for the
liver-to-body weight ratio.
We next examined how much the differences among strains
were due to “environmental” effects versus strain (genetic)
effects, using variance component analysis. On two-way
ANOVA to assess the effects of strains and gender, there were
significant differences in plasma lipids and plasma BHB and
glucose concentrations among 10 strains (Tables 1–3). An
overall gender effect was observed for plasma TG levels in the
mouse strains. When the total liver TG variance was partitioned into Venv and Vstrain, significant strain effect was observed in both genders [P ⬍ 0.001]. Interstrain variation
explained ⬃90% of total hepatic TG variation in males and
⬃76% of total variation in females.
Physiological/Biochemical Studies on Male BALB/c, C57BL,
and SWR
Hepatic lipogenesis, TG secretion, and FA oxidation. In
vivo rates of hepatic lipogenesis, measured by [1-14C]-acetate
incorporation were similar in BALB/c and C57BL, but smaller
in SWR (Fig. 2A). Rates of liver TG secretion, quantified by
Triton WR-1339 injection, were similar in BALB/c and
C57BL and lower in SWR (Fig. 2B). FA ␤-oxidation rates,
determined in primary hepatic cell cultures, were similar in
BALB/c and C57BL and higher in SWR (Fig. 2C). These data
suggest that SWR livers synthesize FAs at a lesser rate and
oxidize them at a faster rate than the other two strains, leaving
less TG for hepatic accumulation. Clearly, low levels of hepatic TG in SWR are not due to enhanced export from the liver.
Responses to fasting: liver TG levels, plasma FFA, and
BHB concentrations. Mean hepatic TG contents in the fed state
were BALB/c⬎C57BL⬎SWR (same rank order as noted during the 10-strain survey). Hepatic TG contents rose significantly with fasting in BALB/c and C57BL at 6 and 14 h, but in
SWR, only at 14 h (Table 5). Mean fasting-induced rises
(⌬14 – 0 h) were ⬃80 mg/g in SWR, ⬃200 mg/g in C57BL,
and ⬃230 mg/g in BALB/c (P ⬍ 0.01 for the overall trend). Rises
also were largest for BALB/c, at 6 h, i.e., the livers of BALB/c
and C57 BL were less able to “adapt” to fasting than SWR.
Table 4. Strain survey: correlation between hepatic
TG and metabolic parameters
Hepatic TG
BF
P
LB
P
BHB
P
PTG
P
PTC
P
PFC
P
Male
Female
⫺0.268
0.0791
0.568
⬍0.0001
0.096
0.5455
0.418
0.0065
⫺0.158
0.3242
⫺0.459
0.0025
0.246
0.0501
0.469
⬍0.0001
0.121
0.3403
⫺0.075
0.5534
⫺0.196
0.1196
⫺0.502
⬍0.0001
BF, body fat; PTG, plasma total triglycerides; PTC, plasma total cholesterol;
PFC, plasma-free cholesterol. Pearson’s correlation was performed using SAS
Proc CORR on all mice (n ⫽ 5 males and 5 females for each strain).
AJP-Gastrointest Liver Physiol • VOL
288 • JUNE 2005 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
Values are means ⫾ SD (n ⫽ 5 males and 5 females for each strain).
Duncan’s multiple tests were used to compare the differences of means with an
overall ␣ ⫽ 0.01. Values in the same column with different superscripts are
different (P ⬍ 0.01). BHB, ␤-hydroxybutyrate.
In the fed state, FFA concentrations were highest in SWR
and rose in all strains with fasting, reaching the highest level in
BALB/c at 14 h. The rise in FFA levels reflects the stimulatory
effect of fasting on lipolytic rates in adipose tissue, balanced by
rates of uptake by liver, muscle, and other tissues (25). It is not
clear whether the largest fasting-induced rise in BALB/c is due
to more rapid lipolysis, slower rate of removal from plasma, or
slower rate of hepatic oxidation perhaps due to the deficiency
of SCAD.
In the fed state, BHB levels were highest in SWR. With
fasting, levels rose in all strains, but rises were greatest in SWR
and least in BALB/c. These data are compatible with a
greater capacity by livers of SWR to adapt to fastinginduced FFA mobilization by most effectively enhancing
FA oxidation among the three strains, thus limiting hepatic TG
accumulation.
Glucose tolerance. Responses to the glucose tolerance test
were similar between male and female; thus results were
combined for both genders. Glucose levels were highest in
C57BL in response to glucose loading (Fig. 3). Levels were
similar in BALB/c and SWR.
Hepatic gene expression profiling. A 2.5-fold or greater
difference in gene expression was considered a notable change.
Expressed transcripts were analyzed in pairwise comparison
between two strains of mice. Eight transcripts were overexpressed, and 14 were underexpressed in C57BL relative to
BALB/c (Table 6). Ten were overexpressed, and 13 were
underexpressed in C57BL relative to SWR (Table 7). Fifteen
genes were overexpressed, and 12 were underexpressed in
SWR relative to BALB/c (Table 8). When one strain was
compared with two others, three transcripts [metallothionein 1
(MT1), MHC class II H2-IA-␣ gene (q haplotype), and hemoglobin-␣, adult chain 1] were upregulated, and four [selenium
binding protein 1 (SELENBP1), hemolytic complement,
G0/G1 switch gene 2, and adult male kidney cDNA, RIKEN
full-length enriched] were downregulated in SWR compared
with both C57BL and BALB/c. Three transcripts [stearoylcoenzyme A desaturase 1 (SCD1), serum amyloid A1, and
MHC class I H-2K1-k pseudogene] were more expressed, and
one [RIKEN cDNA 4933406L18 gene (4933406L18Rik)] was
less expressed in C57BL compared with both SWR and
G1184
GENETICS OF LIVER TRIGLYCERIDE LEVEL IN MICE
BALB/c. Four transcripts (SELENBP1, phosphatidylethanolamine binding protein, amino levulinate synthase, and 18-day
embryo cDNA, RIKEN full-length enriched) were overexpressed in BALB/c relative to both SWR and C57BL, and three
[0-day neonate head cDNA, RIKEN full-length enriched, clone
MGC:19436 IMAGE:3495833, and growth arrest specific 5]
were underexpressed. It is worth noting that SCD-1 was 3.2- to
3.6-fold overexpressed in C57BL relative to both BALB/c and
SWR. This is compatible with the relative glucose intolerance
exhibited by this strain (see Fig. 3) that may lead to increased
synthesis of FAs.
To confirm that the large-scale analysis had correctly identified differentially expressed transcripts, four transcripts were
selected for validation by RT-PCR. Results were similar to
those obtained from gene-expression profiling (Table 9).
Primer pairs used also are listed in Table 9.
DISCUSSION
Although, as mentioned above, genetic manipulation of
genes involving the FA biosynthetic, oxidation, and VLDL-TG
exporting pathways has produced fatty livers (18, 30, 32), here
we provide evidence of a significant genetic contribution to
hepatic TG contents in inbred mice. The 10 mouse strains were
of similar age and were maintained under currently accepted
identical environmental conditions and diet. The highest liver
TG levels were obtained in both male and female BALB/c,
whereas the lowest ones were in SWR, showing six- to sevenfold difference between the two extremes (Fig. 1). ANOVA
analysis indicated that strain-related effects per se accounted
for the majority of the interstrain variation of hepatic TG level.
The genetic bases for the interstrain differences in hepatic TG
are not clear, with the possible exception of BALB/c, which
were SCAD deficient (see below). However, significant strainrelated effects on hepatic TG were seen even when BALB/c
were left out of the ANOVA calculation.
In an attempt to identify metabolic covariates of liver TG,
we performed a series of correlation analyses. Only liver-tobody weight ratios were significantly and positively correlated
with liver TG in both genders. The other parameters of body
weight, body fatness and plasma metabolites such as glucose,
BHB, plasma TG were not consistently correlated across
strains and genders (Table 4). Thus the parameters of adiposity
or insulin action that correlate with liver fat in humans (31, 34)
appear not to be correlated in mice, for reasons that are not
clear. The negative correlation of hepatic TG with plasma-free
cholesterol remains unexplained.
Table 5. Effects of fasting on hepatic TG, plasma FFA, and BHB levels in 3 strains
Hepatic TG, mg/g protein
FFA, mM
BHB, mM
Strain
0h
6h
14 h
0h
6h
14 h
0h
6h
14 h
BALB/c
C57BL
SWR
159.3⫾20.1cA
53.2⫾5.7cB
32.6⫾4.6bC
247.3⫾33.5bA
116.0⫾32.0bB
56.7⫾13.8bC
391.5⫾36.1aA
247.1⫾11.1aB
113.6⫾21.2aC
0.24⫾0.08cB
0.25⫾0.04bAB
0.34⫾0.06bA
0.80⫾0.12bA
0.43⫾0.14bB
0.36⫾0.05bB
1.39⫾0.31aA
0.95⫾0.18aB
0.80⫾0.09aB
0.63⫾0.10bAB
0.50⫾0.10bB
0.78⫾0.23bA
0.75⫾0.19bAB
0.56⫾0.03bB
0.80⫾0.04bA
0.97⫾0.12aB
1.13⫾0.17aB
1.70⫾0.18aA
Values are means ⫾ SD. Male mice were killed after being fasted for 0 (BALB/c, n ⫽ 6; C57BL, n ⫽ 6; and SWR, n ⫽ 3), 6 (BALB/c, n ⫽ 3; C57BL, n
⫽ 3; and SWR, n ⫽ 3), or 14 h (BALB/c, n ⫽ 3; C57BL, n ⫽ 3; and SWR, n ⫽ 3). FFA, free fatty acids, Duncan’s multiple tests were used to compare the
differences of means with an overall ␣ ⫽ 0.05. Lower-case superscripts refer to differences across the rows, i.e. related to increasing duration of fasting within
the same strain. Upper case superscripts refer to columns, i.e. same duration of fasting across the 3 strains. Different superscript letters indicate significant
differences (P ⬍ 0.05).
AJP-Gastrointest Liver Physiol • VOL
288 • JUNE 2005 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
Fig. 2. Three-strain physiological/biochemical studies. Rates of hepatic fatty acid synthesis (A), TG secretion (B), and oxidation (C) were studied in males of
the BALB/c, C57BL, and SWR strains. Values are means ⫾ SD. For rates of hepatic fatty acid synthesis n ⫽ 7 for each strain, for TG secretion, n ⫽ 8 for each
strain, and for fatty acid oxidation, n ⫽ 3 for each strain. The multiple comparisons of means were obtained using Duncan’s multiple-range test with an overall
␣ of 0.01. Means differ from each other among strains with different lowercase letters (P ⬍ 0.01).
G1185
GENETICS OF LIVER TRIGLYCERIDE LEVEL IN MICE
Table 6. Differential gene expression: C57BL vs. BALB/c
Parametric P Value
Fold Difference
GB acc
Description
Transcripts upregulated in C57BL vs. BALB/c
Immune response
0.0004651
0.001183
Acute phase response
0.0004773
Lipid metabolism
0.0014558
cDNA or unknown
0.004716
0.001354
0.0038124
2.61e-05
9.98
2.66
X57330
M27134
mRNA for Qa-2 antigen.
MHC class I H-2K1-k pseudogene
3.60
NM_009117
serum amyloid A 1 (SAA1)
3.62
NM_009127
stearoyl-Coenzyme A desaturase 1 (SCD1)
10.42
10.00
8.53
7.85
BC003986
AK014743
BC011511
NM_013525
clone IMAGE:3491638
0 day neonate head cDNA, RIKEN full-length enriched
clone MGC:19436 IMAGE:3495833
growth arrest specific 5 (GAS5)
Transcripts downregulated in C57BL vs. BALB/c
Acute phase response
0.0017946
0.001338
0.0018503
Immune response
0.0008274
Lipid metabolism
0.0012252
0.0036857
Transcriptional factors
0.0012464
Heme synthesis
0.0010088
Steroid binding
0.0034141
cDNA or unknown
0.0001068
0.0001021
0.0001142
0.0022098
0.0015055
0.32
0.30
0.39
NM_009150
NM_019414
NM_010361
selenium binding protein 1 (SELENBP1)
selenium binding protein 2 (SELEBP2)
glutathione S-transferase, theta 2 (GSTT2)
0.34
X16203
Q5 class I MHC gene
0.28
0.04
NM_018858
NM_021304
phosphatidylethanolamine binding protein (PBP)
alpha/beta hydrolase-1 (LOC57742)
0.28
NM_009883
CCAAT/enhancer binding protein (C/EBP), beta
0.29
M63245
amino levulinate synthase (ALAS-H)
0.33
NM_007618
corticosteroid binding globulin (CBG)
0.20
0.18
0.21
0.31
0.38
AK013108
AK003567
NM_028740
AK007895
AK012224
10, 11 days embryo cDNA, RIKEN full-length enriched
18 days embryo cDNA, RIKEN full-length enriched
RIKEN cDNA 4933406L 18 gene (4933406L 18Rik)
10-day-old male pancreas cDNA, RIKEN full-length
11 days embryo cDNA, RIKEN full-length enriched
Hepatic gene expression profiling and analysis were performed from pooled total liver RNA (n ⫽ 5 each strain, male), as described in MATERIALS AND METHODS.
A 2.5-fold difference or greater was considered a notable change. GB acc, GenBank accession number.
AJP-Gastrointest Liver Physiol • VOL
288 • JUNE 2005 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
Fig. 3. Three-strain glucose tolerance tests (GTT). Both genders of the
BALB/c, C57BL, and SWR strains were used. Because the response to the
glucose load was similar between genders, results were combined within each
strain. Values are means ⫾ SD, n ⫽ 5 for each strain and gender. The multiple
comparisons of means were obtained using Duncan’s multiple-range test with
an overall ␣ of 0.01. Means differ from each other among strains with different
lowercase letters (P ⬍ 0.01).
To understand some of the physiological/biochemical processes underlying the strain-related differences in hepatic TG,
we studied mice representing the highest (BALB/c), middle
range (C57BL), and the lowest (SWR) levels of hepatic TG.
Males were used because they were more readily available, and
we saw similar strain differences in hepatic TG in both genders. Glucose tolerance, reflecting insulin action, was studied
in the 4- to 6-h fasted state. Processes contributing to hepatic
TG levels such as hepatic FA synthesis and oxidation and
hepatic TG transport were studied (7, 9, 15, 20) in the fed state.
Finally, responses of hepatic TG to 6 and 14 h of fasting were
compared with the fed state. We attempt to explain the differences among the three strains in terms of the relative rates of
the above processes we studied.
SWR showed significantly lower rates of hepatic lipogenesis
(Fig. 2A) and TG secretion (Fig. 2B) and higher rates of FA
oxidation (Fig. 2C) than both the BALB/c and C57BL strains,
whereas these parameters were similar in BALB/c and C57BL.
These data suggest that SWR may maintain lower hepatic TG
levels than the other two strains by presenting smaller TG loads
to the VLDL export system. Clearly, because hepatic TG
export was low, liver TG levels were not low because export
was increased. SWR livers also have more efficient capacities
to resist the hepatic TG-raising effects of fasting (Table 5).
G1186
GENETICS OF LIVER TRIGLYCERIDE LEVEL IN MICE
Table 7. Differential gene expression: C57BL vs. SWR
Parametric P Value
Fold Difference
GB acc
Description
Transcripts upregulated in C57BL vs. SWR
3.26
NM_009127
stearoyl-Coenzyme A desaturase 1 (SCD1)
2.57
2.63
3.20
3.27
4.05
M23894
NM_013819
M27134
NM_010406
NM_008059
PYS-2
histocompatibility 2, M region locus 3 (H2-M3),
MHC class I H-2K1-k pseudogene
hemolytic complement (HC)
G0/G1 switch gene 2 (G0S2)
3.27
4.36
NM_009117
NM_009150
serum amyloid A 1 (SAA1)
selenium binding protein 1 (SELENBP1)
2.55
2.82
AK003407
AK002480
18 days embryo cDNA, RIKEN full-length enriched
adult male kidney cDNA, RIKEN full-length enriched
Transcripts downregulated in C57BL vs. SWR
Signal transduction
2.68e-05
Transcriptional factors
0.0028646
Nucleic acid binding
0.0009634
Immune response
0.0040761
0.0001713
0.000482
Lipid metabolism
0.0019276
0.0018488
0.0005654
cDNA
0.0012349
0.0014501
0.0043909
0.0030628
0.13
NM_013602
metallothionein 1 (MT 1)
0.28
NM_020255
peroxisome proliferative activated receptor, gamm
0.32
NM_007447
angiogenin (ANG)
0.30
0.25
0.28
K01925
AF071431
NM_008218
MHC class II H2-IA-alpha gene (q haplotype)
Beta globin
hemoglobin alpha, adult chain 1 (HBA-A1)
0.32
0.35
0.31
AF047725
NM_010004
NM_007815
CYP2C38
cytochrome P450, 2c40 (CYP2C40)
cytochrome P450, 2c29 (CYP2C29)
0.29
0.37
0.39
0.39
AK004319
NM_028740
AK004939
AK009784
18 days embryo cDNA, RIKEN full-length enriched
RIKEN cDNA 4933406L 18 gene (4933406L 18Rik)
adult male liver cDNA, RIKEN full-length enriched
adult male tongue cDNA, RIKEN full-length enriched
Hepatic gene expression profiling and analysis were performed from pooled total liver RNA (n ⫽ 5 each strain, male), as described in MATERIALS AND METHODS.
A 2.5-fold difference or greater was considered a notable change.
They appear to accomplish this by keeping plasma FFA levels
low, which may indicate relatively low rates of lipolysis in
adipose tissue and less delivery of FFA substrate to hepatic TG
production, and BHB levels high, which may indicate efficient
rates of ketogenesis. Both processes would tend to reduce the
amounts of hepatic TG loads available for export via the VLDL
system during fasting, compared with BALB/c and C57BL
strains.
Glucose levels were highest in C57BL in the survey (Table
3) and during the glucose tolerance test (Fig. 3). The relatively
high expression level of the SCD-1 gene in C57BL (Tables 6
and7) is compatible with the relatively high rates of hepatic FA
synthesis and the relative glucose intolerance of C57BL. These
data suggest that although livers of C57BL may produce higher
amounts of hepatic TG (certainly relative to SWR), the presence of adequate VLDL-TG export rates and FA oxidation
rates was able to restrain hepatic TG contents. It is possible that
the accumulation of hepatic TG over time, in C57BL, may
further contribute to the increasing glucose intolerance with
age (4, 26), but high levels of hepatic TG do not explain the
absence of glucose intolerance in BALB/c. It is possible that
not all causes of fatty liver produce insulin resistance.
SCAD is a mitochondrial enzyme that catalyzes the first
reaction in the ␤-oxidation of short-chain FAs. BALB/c mice
bear a spontaneous deletion in SCAD gene and develop fatty
liver after 18 h of fasting (28, 29). However, it is worth noting
that in the fed state, whereas BALB/c had considerably higher
hepatic TG levels than C57BL, the two strains had similar rates
of hepatic FA synthesis, hepatic TG secretion, and similar
concentrations of plasma FFA and BHB. Similar rates of
long-chain FA ␤-oxidation, as measured by using 14C-labeled
palmitate as the substrate, were also found between C57BL and
BALB/c. This raises questions as to whether the SCAD mutation is sufficient to explain the higher hepatic TG contents in
BALB/c in the fed state.
It is possible that static measurements of plasma concentrations do not adequately reflect the deficiency in the action of
SCAD. Plasma levels of FFA are determined by relative rates
of lipolysis in adipose tissue (input) and rates of uptake by the
liver, muscle, and other organs (output) (25). Plasma BHB
concentrations are determined by relative rates of ketogenesis
in liver (input) and ketone body uptake by brain and other
organs (output) (21). It is impossible from the present experiments to determine whether the similar plasma levels of the
two metabolites in fed BALB/c and C57BL reflect similar rates
in input or output. It is conceivable that the rates of input and
output differ in the two strains, but the balance between the two
rates resulted in similar plasma levels. For example, if rates of
AJP-Gastrointest Liver Physiol • VOL
288 • JUNE 2005 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
Lipid metabolism
0.0004591
Immune response
0.0033534
0.0019176
0.0008779
0.0003951
0.0008195
Acute phase response
0.0004732
0.0007113
cDNA
0.0012738
0.0022096
G1187
GENETICS OF LIVER TRIGLYCERIDE LEVEL IN MICE
Table 8. Differential gene expression: SWR vs. BALB/c
Parametric P Value
Fold Difference
GB acc
Description
Transcripts upregulated in SWR vs. BALB/c
2.654
NM_010634
Fatty acid binding protein 5, epidermal (FABP5),
3.04
2.535
2.539
K01925
M19689
NM_008218
MHC class II H2-IA-alpha gene (q haplotype)
MHC class I H-2 (q-haplotype) classic transplantation
hemoglobin alpha, adult chain 1 (HBA-A1)
19.496
NM_013602
Metallothionein 1 (MT1)
3.518
J03953
glutathione transferase GT9.3
3.358
AF171080
histone macroH2A1.2 variant mRNA
2.712
NM_010324
glutamate oxaloacetate transaminase 1, soluble
4.103
15.253
7.764
3.505
2.944
4.522
3.036
AK014743
BC011511
AK004796
NM_024223
AJ278735
NM_013525
AJ279951
0 day neonate head cDNA, RIKEN full-length enriched
Clone MGC:19436 IMAGE:3495833
adult male lung cDNA, RIKEN full-length enriched
RIKEN cDNA 0610010I23 gene (0610010I23Rik)
hypothetical protein (ORF1), 1975 BP
growth arrest specific 5 (GAS5)
Hypothetical protein, clone mv10
Transcripts downregulated in SWR vs. BALB/c
Xenobiotic metabolism
0.0015764
0.0006006
Lipid metabolism
0.0012188
Immune response
2.13e-05
0.0003417
Acute phase response
0.0002434
Heme biosynthesis
0.0019222
cDNA or unknown
0.0005279
0.0016523
0.0005856
0.0020696
0.0015822
0.385
0.315
NM_032541
NM_009676
hepcidin antimicrobial peptide (HAMP)
aldehyde oxidase 1 (AOX1)
0.312
NM_018858
phosphatidylethanolamine binding protein (PBP)
0.145
0.316
NM_008059
NM_010406
G0/G1 switch gene 2 (G0S2)
hemolytic complement (HC)
0.148
NM_009150
selenium binding protein 1 (SELENBP1)
0.158
M63245
amino levulinate synthase (ALAS-H)
0.215
0.324
0.229
0.385
0.359
AK020099
AK003665
AK003567
AK002480
AB054000
adult male olfactory brain cDNA, RIKEN full-leng
18 days embryo cDNA, RIKEN full-length enriched
18 days embryo cDNA, RIKEN full-length enriched
adult male kidney cDNA, RIKEN full-length enriched
Cg10671-like
Hepatic gene expression profiling and analysis were performed from pooled total liver RNA (n ⫽ 5 each strain, male), as described in MATERIALS AND METHODS.
A 2.5-fold difference or greater was considered a notable change.
lipolysis and tissue uptake are both higher in BALB/c, FFA levels
in plasma could remain similar to C57BL levels, but the higher
hepatic uptake of FFA in BALB/c could result in higher levels of
hepatic TG, not explainable by the SCAD defect alone.
The fasting experiment appeared to support an FA ␤-oxidation defect in BALB/c mice (the least rise in plasma BHB
levels in response to fasting). However, this reflected
more of a defect in the fasting situation, consistent with
Table 9. mRNA levels by real-time PCR analysis
Accession Number
Primers (5⬘ to 3⬘)
BALB/c
C57BL
SWR
Stearoyl-CoA desaturase 1: SCD1
Genes
NM_009127
1.0
5.2
2.4
Aldehyde oxidase 1: AOX1
NM_009676
1.0
0.9
0.2
Metallothionein 1: MT1
NM_013602
1.0
2.5
13.2
Cytochrome P450, 2c29: CYP2C29
NM_007815
1.0
0.1
1.1
GAPDH
M32599
Forward primer, GTTGGCTCCATCCATTGC;
Backward primer, AACCATGGGAAGCCAAGTTT
Forward primer, CACGCTGTAGGCTGGTTT;
Backward primer, CCTCCAGGTAGAACTTGAAAA
Forward primer, CGTGCTGTGCCTGATGTG;
Backward primer, GGAAGACGCTGGGTTGGT
Forward primer, AAGAACATCAGCCAATCC;
Backward primer, TTCACCGCTTCATACCC
Forward primer, TGGCCTTCCGTGTTCCT;
Backward primer, AGGCGGCACGTCAGATC
NA
NA
NA
Liver total RNA was isolated and pooled from 5 male mice from each strain. Quantitative real-time PCR was performed from the pooled total RNA as
described in MATERIALS AND METHODS.
AJP-Gastrointest Liver Physiol • VOL
288 • JUNE 2005 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
Lipid metabolism
0.0009032
Immune response
0.0055601
0.0020473
0.0019972
Signal transduction
2.1e-06
Acute phase response
0.0002479
DNA binding
0.001726
Amino acid metabolism
0.0006943
cDNA or unknown
0.0001433
0.0055816
2.33e-05
0.0009285
0.0010505
7.89e-05
0.0010782
G1188
GENETICS OF LIVER TRIGLYCERIDE LEVEL IN MICE
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
GRANTS
20.
This work was supported by grants from The Alan and Edith Wolf
charitable fund, and National Institutes of Health (NIH) Grants R37-HL424460 and RO1-HL-59515. Support from the following sources also is
greatly appreciated: the General Clinical Research Center (NIH Grant
5MO1RR00036), the Diabetes Research and Training Center (5P60DK20579), the Clinical Nutrition Research Unit (DK-56341), the Digestive
Diseases Research Care Center (1P30DK-52574), and ROI HL-52139 (to Z.
Chen).
21.
22.
23.
24.
REFERENCES
1. Angulo P. Nonalcoholic fatty liver disease. N Engl J Med 346: 1221–
1231, 2002.
2. Beattie JH, Wood AM, Newman AM, Bremner I, Choo KH, Michalska AE, Duncan JS, and Trayhurn P. Obesity and hyperleptinemia in
metallothionein (-I and -II) null mice. Proc Natl Acad Sci USA 95:
358 –363, 1998.
3. Bernal-Mizrachi C, Weng S, Li B, Nolte LA, Feng C, Coleman T,
Holloszy JO, and Semenkovich CF. Respiratory uncoupling lowers
25.
26.
27.
AJP-Gastrointest Liver Physiol • VOL
blood pressure through a leptin-dependent mechanism in genetically obese
mice. Arterioscler Thromb Vasc Biol 22: 961–968, 2002.
Bugianesi E, Zannoni C, Vanni E, Marzocchi R, and Marchesini G.
Non-alcoholic fatty liver and insulin resistance: a cause-effect relationship? Dig Liver Dis 36: 165–173, 2004.
Bustin SA. Absolute quantification of mRNA using real-time reverse
transcription polymerase chain reaction assays. J Mol Endocrinol 25:
169 –193, 2000.
Castellano G, Garfia C, Gomez-Coronado D, Arenas J, Manzanares J,
Colina F, and Solis-Herruzo JA. Diffuse fatty liver in familial heterozygous hypobetalipoproteinemia. J Clin Gastroenterol 25: 379 –382, 1997.
Chen Z, Fitzgerald RL, Averna MR, and Schonfeld G. A targeted
apolipoprotein B-38.9-producing mutation causes fatty livers in mice due
to the reduced ability of apolipoprotein B-389 to transport triglycerides.
J Biol Chem 275: 32807–32815, 2000.
Chen Z, Fitzgerald RL, and Schonfeld G. Hypobetalipoproteinemic
mice with a targeted apolipoprotein (Apo) B-27.6-specifying mutation: in
vivo evidence for an important role of amino acids 1254 –1744 of ApoB
in lipid transport and metabolism of the apoB-containing lipoprotein.
J Biol Chem 277: 14135–14145, 2002.
Day CP. Pathogenesis of steatohepatitis. Best Pract Res Clin Gastroenterol 16: 663– 678, 2002.
Day CP and James OF. Steatohepatitis: a tale of two “hits”? Gastroenterology 114: 842– 845, 1998.
Entenman C. General procedures for separating lipid components of
tissue. Methods Enzymol 3: 299 –317, 1957.
Falconer DS and MacKay TFC. Introduction to Quantitative Genetics.
Essex, England: Longman, 1996.
Froyland L, Madsen L, Vaagenes H, Totland GK, Auwerx J, Kryvi H,
Staels B, and Berge RK. Mitochondrion is the principal target for
nutritional and pharmacological control of triglyceride metabolism. J
Lipid Res 38: 1851–1858, 1997.
Giometti CS, Liang X, Tollaksen SL, Wall DB, Lubman DM, Subbarao V, and Rao MS. Mouse liver selenium-binding protein decreased
in abundance by peroxisome proliferators. Electrophoresis 21: 2162–
2169, 2000.
Kersten S, Seydoux J, Peters JM, Gonzalez FJ, Desvergne B, and
Wahli W. Peroxisome proliferator-activated receptor alpha mediates the
adaptive response to fasting. J Clin Invest 103: 1489 –1498, 1999.
Kim JB, Sarraf P, Wright M, Yao KM, Mueller E, Solanes G, Lowell
BB, and Spiegelman BM. Nutritional and insulin regulation of fatty acid
synthetase and leptin gene expression through ADD1/SREBP1. J Clin
Invest 101: 1–9, 1998.
Kim JK, Fillmore JJ, Chen Y, Yu C, Moore IK, Pypaert M, Lutz EP,
Kako Y, Velez-Carrasco W, Goldberg IJ, Breslow JL, and Shulman
GI. Tissue-specific overexpression of lipoprotein lipase causes tissuespecific insulin resistance. Proc Natl Acad Sci USA 98: 7522–7527, 2001.
Le May C, Pineau T, Bigot K, Kohl C, Girard J, and Pegorier JP.
Reduced hepatic fatty acid oxidation in fasting PPARalpha null mice is
due to impaired mitochondrial hydroxymethylglutaryl-CoA synthase gene
expression. FEBS Lett 475: 163–166, 2000.
Lin X, Schonfeld G, Yue P, and Chen Z. Hepatic fatty acid synthesis is
suppressed in mice with fatty livers due to targeted apolipoprotein B38.9
mutation. Arterioscler Thromb Vasc Biol 22: 476 – 482, 2002.
Malnick SD, Beergabel M, and Knobler H. Non-alcoholic fatty liver: a
common manifestation of a metabolic disorder. QJM 96: 699 –709, 2003.
Mitchell GA, Kassovska-Bratinova S, Boukaftane Y, Robert MF,
Wang SP, Ashmarina L, Lambert M, Lapierre P, and Potier E.
Medical aspects of ketone body metabolism. Clin Invest Med 18: 193–216,
1995.
Nadeau JH. Modifier genes and protective alleles in humans and mice.
Curr Opin Genet Dev 13: 290 –295, 2003.
Nadeau JH. Modifier genes in mice and humans. Nat Rev Genet 2:
165–174, 2001.
Ntambi JM. The regulation of stearoyl-CoA desaturase (SCD). Prog
Lipid Res 34: 139 –150, 1995.
Owen OE, Reichard GA Jr, Patel MS, and Boden G. Energy metabolism in feasting and fasting. Adv Exp Med Biol 111: 169 –188, 1979.
Petersen KF, Befroy D, Dufour S, Dziura J, Ariyan C, Rothman DL,
DiPietro L, Cline GW, and Shulman GI. Mitochondrial dysfunction in
the elderly: possible role in insulin resistance. Science 300: 1140 –1142,
2003.
Petersen KF, Oral EA, Dufour S, Befroy D, Ariyan C, Yu C, Cline
GW, DePaoli AM, Taylor SI, Gorden P, and Shulman GI. Leptin
288 • JUNE 2005 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
the development of fatty liver on 18 h of fasting in this
strain (29).
One way to settle these issues would be to examine another
BALB/c strain with intact SCAD function. We have studied
another BALB/c strain available from Jackson Laboratories:
the BALB/cJ strain. Hepatic TG contents of male BALB/cJ in
fact are lower, resembling those of SWR. However, the two
BALB/c strains also differ in at least 43 of ⬃300 microsatellite
markers genotyped across the genome (http://www.cidr.jhml.
edu/mouse/mouse.html). Thus, at this time, it is not possible
unequivocally to ascribe the excess hepatic TG in the fed state
to the SCAD deficiency alone. It remains to be investigated
whether any other factor(s) besides SCAD mutation causes
fatty liver in this mouse strain.
We performed a gene expression survey expecting to find
additional explanations for the interstrain differences in hepatic
TG contents. Most of the differences in hepatic gene expression remain to be explained. However, as noted, the relative
overexpression of the SCD-1 transcript in C57BL was useful in
explaining the contribution of presumed increased palmitate
desaturation to oleate to the higher hepatic TG levels in C57BL
(Tables 6 and 7) (24). It is not clear whether high hepatic FA
synthesis with lower SCD1 mRNA level in BALB/c results
from a lower oxidation rate of newly synthesized FA. MT1 has
been linked to the regulation of energy balance in mice,
because mice deficient in both MT1 and MT2 were obese (2).
SWR had the highest level of expression in MT1 when compared with both C57BL and BALB/c, consistent with the
lowest body weight of SWR in the three strains. Mouse hepatic
SELENBP1 was decreased in response to peroxisome proliferators such as ciprofibrate (14). The lowest expression of
hepatic SELENBP1 in SWR relative to both C57BL and
BALB/c may be related to the highest hepatic FA oxidation in
SWR. Another consideration is that differential gene expression may be the consequence, rather than the cause of differences in liver TG levels. For example, hepatic TG accumulation may raise the level of reactive oxygen products in liver (1)
stimulating alterations in the expression levels of mRNAs of
acute phase reactant molecules.
In conclusion, results of these studies have demonstrated a
significant genetic contribution to hepatic TG contents in
inbred mice. They also helped identify potential parental
strains for construction of congenic strains bearing apoB truncations. Breeding is in progress.
GENETICS OF LIVER TRIGLYCERIDE LEVEL IN MICE
28.
29.
30.
31.
32.
34.
35. Tarugi P, Lonardo A, Ballarini G, Grisendi A, Pulvirenti M, Bagni A,
and Calandra S. Fatty liver in heterozygous hypobetalipoproteinemia
caused by a novel truncated form of apolipoprotein B. Gastroenterology
111: 1125–1133, 1996.
36. Teli MR, James OF, Burt AD, Bennett MK, and Day CP. The natural
history of nonalcoholic fatty liver: a follow-up study. Hepatology 22:
1714 –1719, 1995.
37. Wetterau JR, Aggerbeck LP, Bouma ME, Eisenberg C, Munck A,
Hermier M, Schmitz J, Gay G, Rader DJ, and Gregg RE. Absence of
microsomal triglyceride transfer protein in individuals with abetalipoproteinemia. Science 258: 999 –1001, 1992.
38. Wood P, Hinsdale ME, and Kelly CL. Molecular detection of the Bcd-1
null allele in BALB/cByJ mice by polymerase chain reaction: a simple
assay for genetic monitoring. Mouse Genome 91: 342–344, 1993.
39. Yao Z and McLeod RS. Synthesis and secretion of hepatic apolipoprotein
B-containing lipoproteins. Biochim Biophys Acta 1212: 152–166, 1994.
40. Yokoyama C, Wang X, Briggs MR, Admon A, Wu J, Hua X, Goldstein JL, and Brown MS. SREBP-1, a basic-helix-loop-helix-leucine
zipper protein that controls transcription of the low density lipoprotein
receptor gene. Cell 75: 187–197, 1993.
41. Yu S, Rao S, and Reddy JK. Peroxisome proliferator-activated receptors,
fatty acid oxidation, steatohepatitis and hepatocarcinogenesis. Curr Mol
Med 3: 561–572, 2003.
AJP-Gastrointest Liver Physiol • VOL
288 • JUNE 2005 •
www.ajpgi.org
Downloaded from http://ajpgi.physiology.org/ by 10.220.33.6 on June 18, 2017
33.
reverses insulin resistance and hepatic steatosis in patients with severe
lipodystrophy. J Clin Invest 109: 1345–1350, 2002.
Reue K and Cohen RD. Acads gene deletion in BALB/cByJ mouse strain
occurred after 1981 and is not present in BALB/cByJ-fld mutant mice.
Mamm Genome 7: 694 – 695, 1996.
Schiffer SP, Prochazka M, Jezyk PF, Roderick TH, Yudkoff M, and
Patterson DF. Organic aciduria and butyryl CoA dehydrogenase deficiency in BALB/cByJ mice. Biochem Genet 27: 47–58, 1989.
Schonfeld G. Familial hypobetalipoproteinemia: a review. J Lipid Res 44:
878 – 883, 2003.
Schonfeld G, Patterson BW, Yablonskiy DA, Tanoli TS, Averna M,
Elias N, Yue P, and Ackerman J. Fatty liver in familial hypobetalipoproteinemia: triglyceride assembly into VLDL particles is affected by the
extent of hepatic steatosis. J Lipid Res 44: 470 – 478, 2003.
Shimano H, Horton JD, Hammer RE, Shimomura I, Brown MS, and
Goldstein JL. Overproduction of cholesterol and fatty acids causes
massive liver enlargement in transgenic mice expressing truncated
SREBP-1a. J Clin Invest 98: 1575–1584, 1996.
Silverman JF, Pories WJ, and Caro JF. Liver pathology in diabetes
mellitus and morbid obesity. Clinical, pathological, and biochemical
considerations. Pathol Annu 24: 275–302, 1989.
Tanoli T, Yue P, Yablonskiy D, and Schonfeld G. Fatty liver in familial
hypobetalipoproteinemia: roles of the APOB defects, intra-abdominal
adipose tissue, and insulin sensitivity. J Lipid Res 45: 941–947, 2004.
G1189