Am J Physiol Endocrinol Metab 282: E703–E713, 2002. First published October 2, 2001; 10.1152/ajpendo.00072.2001.
Diet-induced obesity and hepatic gene expression alterations in C57BL/6J and ICAM-1-deficient mice FRANCINE M. GREGOIRE,* QIN ZHANG,* STEVEN J. SMITH, CARMEN TONG, DAVID ROSS, HENRY LOPEZ, AND DAVID B. WEST Pfizer Global R&D, Alameda, California 94502 Received 28 February 2001; accepted in final form 26 October 2001
Gregoire, Francine M., Qin Zhang, Steven J. Smith, Carmen Tong, David Ross, Henry Lopez, and David B. West. Diet-induced obesity and hepatic gene expression alterations in C57BL/6J and ICAM-1-deficient mice. Am J Physiol Endocrinol Metab 282: E703–E713, 2002. First published October 2, 2001; 10.1152/ajpendo.00072.2001.—The effects of high-fat feeding on the development of obesity were evaluated in intercellular adhesion molecule-1 (ICAM-1) knockout and C57BL/6J (B6) male mice fed a high-fat diet for ⱕ50 days. Serum and tissues were collected at baseline and after 1, 11, and 50 days on the diet. After 11 days on the diet, ICAM-1-deficient, but not B6, mice developed fatty livers and showed a significant increase in inguinal fat pad weight. At day 50, ICAM-1-deficient mice weighed less, and their adiposity index and circulating leptin levels were significantly lower than those of B6 controls. To better understand the early differential response to the diet, liver gene expression was analyzed at three time points by use of Affymetrix GeneChips. In both strains, a similar pattern of gene expression was detected in response to the high-fat diet. However, sterol regulatory element-binding protein-1, apolipoprotein A4, and adipsin mRNAs were significantly induced in ICAM1-deficient livers, suggesting that these genes and their associated pathways may be involved in the acute diet response observed in the knockout mice.
THE PHYSIOLOGICAL EFFECTS of murine diet composition have been the focus of much study, particularly regarding obesity, diabetes, and atherosclerosis (11, 26, 29, 31, 37, 41, 44). The studies of these human pathologies have been greatly furthered by the use of the C57BL/6J inbred strain because of its susceptibility to these disorders, thus serving as a model for the human diseases. Liver steatosis, or fatty liver, is characterized by excessive lipid accumulation in the hepatocytes and is a related pathology associated with obesity and type 2 diabetes (23, 36). The excess liver lipid accumulation may be comprised of cholesterol or triglycerides or both. Although hepatic lipid deposition may result from a number of causes, few genes have been positively correlated with this pathology (12, 19, 23, 33, 36).
Intercellular adhesion molecule-1 (ICAM-1) is a cytokine-responsive integral membrane receptor having five immunoglobulin (Ig)-like extracellular domains, a transmembrane domain, and a short cytoplasmic tail. ICAM-1-deficient mice were previously reported to spontaneously develop maturity-onset obesity and fatty livers without an increase in food intake, and they had an increased susceptibility to obesity induced by a high-fat diet (6). Adhesion molecules such as ICAM-1 are known to play an important role in the development of atherosclerosis (1, 40) and may also contribute to the atherosclerosis pathology in diabetic individuals (3, 14). There has been no reported characterization of the molecular mechanism involved in ICAM-1-deficient obesity, and the role of leukocytes in the regulation of both lipid metabolism and energy expenditure remains obscure. A possible role for the peroxisome proliferator-activating receptor-␣ (PPAR␣) transcription factor pathway was suggested (6), but this hypothesis has not been validated. Although it is not known whether repression of leukocyte adhesion receptors is associated with obesity in humans, ICAM-1 (⫺/⫺) mice represent an interesting animal model for studying potentially novel mechanisms that regulate adiposity without significant appetite modification. In this report, we have characterized the response of C57BL/6J (B6) and ICAM-1 (⫺/⫺) mice to high-fat feeding at both the physiological and molecular levels. With short-term high-fat feeding (11 days), the ICAM-1 (⫺/⫺) mice accumulated significantly more adipose and liver lipid than B6 controls. However, after 50 days of high-fat feeding, the B6 mice weighed more than ICAM-1 (⫺/⫺) mice, and the liver lipid accumulation was similar in the two strains. By assessing the expression levels of thousands of genes by use of an Affymetrix GeneChip array, we sought to gain insights into the molecular events that are altered in the livers during the development of dietary obesity. Microarray technology has been successfully used to identify differences in gene expression in various altered metabolic conditions, including aging and caloric restriction, genetic obesity, diabetes, and cancer (5, 9, 21, 25, 38). However, this
* These authors contributed equally. Address for reprint requests and other correspondence: F. M. Gregoire, Metabolex, Inc., 3876 Bay Center Place, Hayward, CA 94545 (E-mail: [email protected]
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.
intercellular adhesion molecule-1 knockout mice; obesity; dyslipidemia; steatosis; microarray
0193-1849/02 $5.00 Copyright © 2002 the American Physiological Society
is the first report to describe global changes in liver gene expression upon high-fat feeding. MATERIALS AND METHODS
Animals. ICAM-1-deficient mice [ICAM-1 (⫺/⫺); stock no. 002867] and C57BL/6J mice (B6; stock no. 000664) were obtained from the Jackson Laboratory (Bar Harbor, ME). The ICAM-1 (⫺/⫺) mice were originally generated in a mixed 129-B6-DBA background (43) and then backcrossed to the C57BL/6J background for 8 generations. The homozygous knockouts are viable and fertile, allowing ICAM-1 (⫺/⫺) breeding pairs to be intercrossed to maintain the line. The experimental colony described here were bred and maintained on a mouse chow diet (Laboratory Rodent Diet 5001, PMI Feeds, St. Louis, MO). Male mice were individually housed at 6 wk of age. At 7 wk of age, these mice were switched to the Western type of high-fat diet that contained 42% of calories as fat (Adjusted Calories Diet no. 88137, Harlan Teklad, Madison, WI). Body weights and food intakes were measured twice a week from the start of high-fat feeding. Mice were maintained on a 12:12-h light-dark cycle, with water ad libitum. All reported investigations were carried out in accordance with the Guidelines of the National Institutes of Health and Medical Research (NIH publication No. 87– 848, 1987). To generate ICAM-1 (⫺/⫺) and control mice of the proper background strain, ICAM-1 (⫺/⫺) males were mated to C57BL/6J females. The progeny of this backcross were intercrossed to produce F2-generation littermates of the three genotypes (⫺/⫺, ⫹/⫺, and ⫹/⫹). Genotypes of the wild type and knockout ICAM-1 alleles of all of the experimental mice were confirmed by PCR. The wild-type allele was amplified by PCR primer pairs IC1–1 (5⬘ TGCCTCTGAAGCTCGGATATACC-3⬘) and IC1–2 (5⬘ CTGTAGACTGTTAAGGTCCTCTGCG-3⬘). The knockout allele was amplified with primers IC1–1 and PGK-1 (5⬘-TGAGCCCAGAAAGCGAAGGAA-3⬘). PCR was performed in a 15-l volume containing 0.8 M of pairs of either primer and 25 ng of genomic DNA. During amplification, a 5-min denaturing step at 95°C was followed by 30 cycles of 30 s at 94°C, 30 s at 57°C, and 2 min at 72°C. Serum chemistry and tissue collection. Mice from all experimental groups were killed at baseline (day 0) or after 1, 11, or 50 days on the diet. The animals were fasted for 4–5 h before blood and tissue collection. After a drop of blood was obtained from the tail vein for glucose measurements, each mouse was anesthetized with isoflurane and weighed. Blood was collected by cardiac puncture and subsequently assayed for insulin, leptin, and lipid levels. Circulating insulin and leptin levels were determined by radioimmunoassays (Linco Research, St. Charles, MO). Serum triglyceride levels were determined by a colorimetric assay kit (Wako Chemicals USA, Richmond, VA). Serum lipoproteins were separated by HPLC (Varian, Walnut Creek, CA) by use of a Superose 6 column (Amersham Pharmacia Biotech, Piscataway, NJ), and the cholesterol content in each lipoprotein fraction was determined by use of a cholesterol esterase kit (Roche Diagnostic Systems, Somerville, NJ). Livers and individual fat pads were dissected, weighed, and snap-frozen in liquid nitrogen. The adiposity index for individual animals was calculated by dividing the sum of the weights of the dissected fat depots by the body weight (minus the fat depot weight). Although this method is not as accurate as carcass analysis, it gives a more rapid and reproducible evaluation of body fat content in mice (40). Liver histology was evaluated on either paraffin or frozen sections. For paraffin sections, the livers from three B6 and five ICAM-1 (⫺/⫺) males at day 11 of AJP-Endocrinol Metab • VOL
high-fat feeding were dissected, fixed in 10% formalin, and embedded in paraffin. Sections (5 m) were stained with hematoxylin and eosin. For frozen sections, livers from five male ICAM-1 (⫺/⫺) and B6 mice fed the high-fat diet for 11 days were snap-frozen. Sections of 10–15 m were cut, fixed in formalin, and stained with Oil Red O. Expression profiling. Total RNA was isolated from the livers of seven mice in each group at days 0 and 11 and of four ICAM-1 (⫺/⫺) and six B6 livers at day 11 with Trizol reagent (Life Technologies, Gaithersburg, MD). To prepare samples for Affymetrix GeneChip analysis, cDNA was generated from 15 g of pooled total RNA (50 g RNA per liver per time point and genotype, as indicated above) by use of a modified oligo-dT primer and a 5⬘ T7 RNA polymerase promoter oligo primer with the Superscript Choice System for cDNA Synthesis (Life Technologies). After phenol-chloroform extraction and ethanol precipitation, one-half of the cDNA reaction (0.5–1.0 g) was used as a template for an in vitro transcription reaction (BioArray High Yield kit, Enzo Biochem, Farmingdale, NY) by following the manufacturer’s protocol. The resulting cRNA was purified on an affinity resin column (RNeasy, Qiagen, Valencia, CA) and quantified by ultraviolet (UV) absorbance. For each reaction, 15 g of biotinylated cRNA were randomly fragmented to an average size of 50 nucleotides by incubating them at 94°C for 35 min in 40 mM Tris-acetate, pH 8.1, 1,000 mM potassium acetate, and 30 mM magnesium acetate. The fragmented cRNA was divided into two aliquots that were each used for hybridization to an MU6500 Affymetrix GeneChip according to the manufacturer’s protocol (Affymetrix, Santa Clara, CA), with a duplicate data set generated for all samples. Data were analyzed with the Affymetrix software algorithm to generate “average difference” and/or degree of difference. When the degree of difference was calculated, each average difference ⬍10 was brought up to 10 to get meaningful fold difference value. Variation between duplicate samples did not exceed 30%. Therefore, the mean of the average difference values obtained for the duplicates was used to generate Table 3. For Northern blot analyses, duplicate samples of 20 g of pooled total liver RNA were separated on 1% agarose gels containing 6.7% formaldehyde in 1⫻ MOPS buffer and blotted onto Duralon-UV membranes (Stratagene, La Jolla, CA). 32Plabeled cDNA probes were hybridized to the Northern blots by use of the ExpressHyb solution (Clontech, Palo Alto, CA), and radioactive signals were analyzed by the Storm System phosphoimager with ImageQuant 5.0 software (Molecular Dynamics, Sunnyvale, CA). The 36B4 message was used as a control for gene expression levels (20). The cDNA probes were generated either by RT-PCR by use of gene-specific primers and liver total RNA as template (adipsin probe only, forward primer: 5⬘-TCCGCCCCTGAACCCTACAAG-3⬘; reverse primer: 5⬘-CTTTTTGCCATTGCCACAGACG-3⬘; product size 447bps) or by PCR of IMAGE clones obtained from Research Genetics (Huntsville, AL) as described below. The IMAGE clone numbers are for ICAM-1, 315622; malic enzyme (ME), 876123; sterol regulatory element-binding protein (SREBP)-1, 747669; fatty acid synthase (FAS), 851908; apolipoprotein A (apoA)-4, 481110; keratinocyte fatty acid binding protein (mal 1), 1196423; squalene synthase, 1347569; PPAR␣, 677502; PPAR␥, 317536; carnitine o-palmitoyltransferase 1 (CPT-1), 335258; and 36B4, 331627. Esherichia coli harboring the desired IMAGE clones were grown in a 96-well plate overnight in LB medium containing the appropriate antibiotic for plasmid selection. To amplify the plasmid inserts, PCR was performed on 1 l of liquid culture in a 50-l reaction using T3 and T7 primers (category no. 302001, Stratagene) and the following PCR conditions: denaturation at 96°C for 3 min
282 • MARCH 2002 •
followed by 35 cycles of 30 s at 94°C, 60 s at 55°C, and 90 s at 72°C, finishing with 5 min at 72°C for a final elongation. All amplified inserts generated by PCR were sequence-verified, as previously described (10). Data and statistical analysis. The data were analyzed by two-way ANOVA (strain ⫻ diet) with SAS (Morrisville, NC) version 8 and the ANALYST program set. Pair-wise statistical significance was established using a post hoc Student t-test. In some cases, statistical outliers were removed and the analysis was revisited; however, no effect was found on the overall significance levels. Data are presented as means ⫾ SE. Statistical signifiance is defined as P ⬍ 0.05. Affymetrix GeneChip studies considered those gene expression changes ⬎2-fold and those ⬍0.5-fold to be significantly changed, as recommended by Affymetrix. RESULTS
ICAM-1 (⫺/⫺) and C57BL/6J (B6) mice were fed a high-fat diet, containing 42% of calories as fat, beginning at 7 wk of age. Body weights and food consumption were measured twice a week beginning at day 0 on the diet and ending at day 50. Contrary to a published report (6), ICAM-1-deficient male mice did not gain
more weight than control mice upon high fat feeding. Although both strains were diet responsive, B6 male mice were on average 4 g heavier than ICAM-1 (⫺/⫺) animals by day 50 (Table 1). Moreover, no significant differences in body weight and adiposity were found between high fat-fed B6 and ICAM-1 (⫺/⫺) females, and for both sexes, maturity-onset obesity was not observed in animals on the chow diet (not shown). In all cases, no significant difference in food intake was detected between strains (not shown). Overall, independent of the gender, no obvious obesity phenotype was detected in ICAM-deficient mice. Therefore, a detailed phenotyping analysis was performed on male mice only, and the following parameters were measured: body mass index (BMI); liver weight; fat pad weights; adiposity index (Table 1); leptin, glucose, and insulin levels (Table 1); and blood lipid levels (Table 2). Individual fat pad weights from ICAM-1 (⫺/⫺) male mice were compared with those from B6 male mice. Before high-fat diet feeding, there were no significant differences in the relative weights of the examined fat
Table 1. Comparison of phenotypes between C57BL/6J and ICAM-1 (⫺/⫺) mice
Before high fat diet Body wt, g Body mass index, g/cm2 Liver wt, g Inguinal fat pad wt, g Epididymal fat pad wt, g Retroperitoneal fat pad wt, g Mesenteric fat pad wt, g Brown fat pad wt, g Adiposity index, % Serum leptin, ng/ml Blood glucose, mg/dl Serum insulin, ng/ml 11 Days on high fat diet Body wt, g Body mass index, g/cm2 Liver wt, g Inguinal fat pad wt, g Epididymal fat pad wt, g Retroperitoneal fat pad wt, g Mesenteric fat pad wt, g Brown fat pad wt, g Adiposity index, % Serum leptin, ng/ml Blood glucose, mg/dl Serum insulin, ng/ml 50 Days on high fat diet Body wt, g Body mass index, g/cm2 Liver wt, g Inguinal fat pad wt, g Epididymal fat pad wt, g Retroperitoneal fat pad wt, g Mesenteric fat pad wt, g Brown fat pad wt, g Adiposity index, % Serum leptin, ng/ml Blood glucose, mg/dl Serum insulin, ng/ml
ICAM-1 (⫺/⫺) vs. C57BL/6J, %
20.771 ⫾ 0.408 0.233 ⫾ 0.003 1.093 ⫾ 0.027 0.244 ⫾ 0.015 0.282 ⫾ 0.018 0.056 ⫾ 0.005 0.079 ⫾ 0.007 0.050 ⫾ 0.002 3.533 ⫾ 0.173 1.436 ⫾ 0.059 145.357 ⫾ 4.554 0.316 ⫾ 0.045
20.769 ⫾ 0.460 0.237 ⫾ 0.003 1.097 ⫾ 0.037 0.274 ⫾ 0.016 0.181 ⫾ 0.009 0.053 ⫾ 0.005 0.072 ⫾ 0.007 0.057 ⫾ 0.003 3.171 ⫾ 0.160 1.740 ⫾ 0.280 154.563 ⫾ 6.629 0.793 ⫾ 0.172
100 102 100 112 64 96 90 115 90 121 106 251
0.997 0.320 0.925 0.175 ⬍0.001 0.731 0.432 0.064 0.122 0.311 0.260 0.015
23.894 ⫾ 0.440 0.259 ⫾ 0.003 1.116 ⫾ 0.033 0.473 ⫾ 0.021 0.642 ⫾ 0.031 0.145 ⫾ 0.011 0.210 ⫾ 0.019 0.092 ⫾ 0.005 6.969 ⫾ 0.308 4.515 ⫾ 1.023 143.944 ⫾ 5.135 0.708 ⫾ 0.126
24.406 ⫾ 0.456 0.267 ⫾ 0.004 1.428 ⫾ 0.054 0.689 ⫾ 0.028 0.418 ⫾ 0.002 0.166 ⫾ 0.007 0.289 ⫾ 0.022 0.142 ⫾ 0.009 7.481 ⫾ 0.260 8.467 ⫾ 0.914 156.000 ⫾ 4.022 1.381 ⫾ 0.191
102 103 128 146 65 114 138 154 107 188 108 195
0.412 0.110 ⬍0.001 ⬍0.001 ⬍0.001 0.127 0.008 ⬍0.001 0.203 0.006 0.068 0.004
34.657 ⫾ 0.690 0.349 ⫾ 0.007 1.820 ⫾ 0.122 1.186 ⫾ 0.048 1.972 ⫾ 0.081 0.456 ⫾ 0.020 0.508 ⫾ 0.034 0.198 ⫾ 0.012 14.200 ⫾ 0.393 23.675 ⫾ 1.713 166.929 ⫾ 3.848 1.551 ⫾ 0.238
30.577 ⫾ 0.761 0.314 ⫾ 0.005 1.847 ⫾ 0.107 1.183 ⫾ 0.055 0.878 ⫾ 0.034 0.338 ⫾ 0.016 0.391 ⫾ 0.031 0.186 ⫾ 0.012 10.278 ⫾ 0.272 14.965 ⫾ 1.203 157.923 ⫾ 9.954 1.865 ⫾ 0.265
88 90 101 100 45 74 77 94 76 63 95 120
⬍0.001 ⬍0.001 0.872 0.973 ⬍0.001 ⬍0.001 0.021 0.528 ⬍0.001 ⬍0.001 0.375 0.368
Values are means ⫾ SE. AJP-Endocrinol Metab • VOL
282 • MARCH 2002 •
Table 2. Comparison of blood lipids between C57BL/6J and ICAM-1 (⫺/⫺) mice
11 Days on high fat diet Total cholesterol, mg/dl VLDL cholesterol, mg/dl LDL cholesterol, mg/dl HDL cholesterol, mg/dl Triglyceride, mg/dl 11 Days on chow Total cholesterol, mg/dl VLDL cholesterol, mg/dl LDL cholesterol, mg/dl HDL cholesterol, mg/dl Triglycerides, mg/dl
ICAM-1 (⫺/⫺) vs. C57BL/6J, %
169.698 ⫾ 5.301 10.202 ⫾ 1.201 28.501 ⫾ 1.704 129.897 ⫾ 4.999 72.587 ⫾ 8.295
196.402 ⫾ 14.300 14.896 ⫾ 1.794 27.006 ⫾ 1.687 153.704 ⫾ 14.001 124.202 ⫾ 11.785
116 146 95 118 171
0.248 0.072 0.285 0.295 0.003
64.495 ⫾ 5.101 3.803 ⫾ 0.801 8.298 ⫾ 1.101 52.104 ⫾ 4.386 52.578 ⫾ 7.201
72.198 ⫾ 4.897 7.789 ⫾ 0.687 7.501 ⫾ 0.589 56.496 ⫾ 4.001 151.200 ⫾ 11.101
112 205 90 108 293
0.271 0.001 0.509 0.446 ⬍0.001
Values are means ⫾ SE. VLDL, LDL, and HDL, very low density, low-density, and high-density lipoprotein, respectively.
pads, except for the relatively smaller epididymal fat in ICAM-1 (⫺/⫺) (Table 1). At day 11, inguinal and brown fat pad weights were significantly greater in ICAM-1 (⫺/⫺) mice than in B6 mice. However, the difference in weight of these fat pads disappeared by day 50 (Table 1). The intra-abdominal mesenteric, retroperitoneal, and epididymal fat depots showed a slightly different profile but were all significantly heavier in B6 than in ICAM-1 (⫺/⫺) mice by day 50 (Table 1). As a result, the adiposity indexes for the two strains were the same at days 0 and 11 but were significantly higher for B6 mice at day 50 (Table 1). Serum leptin levels were the same at day 0, high at day 11, and low at day 50 in ICAM-1 (⫺/⫺) males compared with the B6 males (Table 1). Eleven days after initiation of the high-fat feeding, a number of differences were apparent in livers taken from ICAM-1 (⫺/⫺) male mice. Upon gross observation, ICAM-1-deficient livers were paler, distended, and heavier than B6-derived livers, suggesting hepatic steatosis (not shown and Table 1). The histological analyses revealed major hepatic abnormalities in ICAM-1-deficient mice that corresponded to excessive lipid accumulation, as assessed by the lipid-specific Oil Red O-positive staining of frozen liver sections (Fig. 1). After 50 days of high-fat feeding, there were no more significant differences between B6 and ICAM-1 (⫺/⫺) liver appearance, with both strains exhibiting equally fatty livers (not shown). To characterize the effects of high-fat feeding on glucose metabolism, blood glucose and insulin levels were measured in both strains. Serum insulin levels were higher in ICAM-1-deficient mice than in B6 mice before and also after 11 days of high-fat feeding (Table 1). The difference in blood insulin levels became statistically insignificant between the two strains after 50 days of high-fat feeding. Despite elevated insulin levels in the knockouts at days 0 and 11, the blood glucose levels were not significantly different at any time point in the two strains (Table 1). Glucose tolerance and insulin sensitivity were also evaluated after 6 wk of high-fat diet feeding but did not differ between strains (not shown). The analysis of blood lipid profiles of B6 and ICAM-1 (⫺/⫺) mice showed that the knockout mice had higher AJP-Endocrinol Metab • VOL
levels of serum triglycerides and very low density lipoprotein (VLDL) cholesterol than the B6 mice when fed a high-fat diet for 11 days (Table 2). Similar results were obtained from mice fed a chow diet, suggesting that the observed differences are not dependent on the dietary fat quantity or composition. In contrast, no significant differences were found in low-density lipoprotein (LDL), high-density lipoprotein (HDL), and total cholesterol levels between these two strains (Table 2).
Fig. 1. Frozen liver sections derived from C57BL/6J (A) and intercellular adhesion molecule-1 (ICAM-1) (⫺/⫺) (B) mice, which had been fed a high-fat diet for 11 days, were fixed in formalin and stained with Oil Red O.
282 • MARCH 2002 •
To determine whether the discrepancies in our results compared with those of Dong et al. (6) were due to a strain background effect rather than ICAM-1 deficiency, ICAM-1 (⫺/⫺) males were backcrossed to C57BL/6J females. The F1 progeny was intercrossed to obtain F2-generation animals of all genotypes at the ICAM-1 locus. Male knockout or wild-type F2 animals were individually housed and fed the high-fat diet, as described previously. At days 0, 11, and 50, animals were killed and fat pads were dissected and weighed as described. Figure 2 depicts body weight gain, liver, inguinal fat, and epididymal fat weights over time for
all genotypes. Figure 2A shows the weight gain over time in B6 male mice and ICAM-1 knockout males (left) and the weight gain over time of F2-generation ICAM-1 (⫹/⫹) male mice compared with that of their ICAM-1 (⫺/⫺) male littermates (right). Body weight gain over time was significantly higher for B6 control mice compared with ICAM-1 (⫺/⫺) mice (Fig. 2A, left). In contrast, no significant difference in weight gain was found between ICAM-1 (⫹/⫹) and ICAM-1 (⫺/⫺) F2 littermates (Fig. 2A, right). In neither case did ICAM-1-deficient mice gain weight as rapidly as wildtype animals. After 11 days of high-fat feeding, the
Fig. 2. Body and tissue weights of mice fed the high-fat diet. A: body wts of C57BL/6J (n ⫽ 14) and ICAM-1 (⫺/⫺) (n ⫽ 13) mice (left) or ICAM-1 (⫹/⫹) (n ⫽ 11) and ICAM-1 (⫺/⫺) (n ⫽ 9) F2-generation littermates (right) fed the high-fat diet for ⱕ50 days. B: weights of liver, inguinal fat pad (ING), and epididymal fat pad (EPI) from C57BL/6J (n ⫽ 14–18) and ICAM-1 (⫺/⫺) (n ⫽ 13–17) (left) or from ICAM-1 (⫹/⫹) (n ⫽ 7–13) and ICAM-1 (⫺/⫺) (n ⫽ 6–11) F2-generation littermates (right) fed the high-fat diet for 50 days. Error bars denote SE; *P ⬍ 0.05, statistically significant differences.
AJP-Endocrinol Metab • VOL
282 • MARCH 2002 •
livers and the inguinal fat pads derived from ICAM-1 (⫺/⫺) males weighed more, as a percentage of body weight, than those of B6 mice (Fig. 2B, left) or than those of their ICAM-1 (⫹/⫹) littermates (Fig. 2B, right). In contrast, at every time point analyzed, a distinct picture emerged for the epididymal fat pad weight in the F2 generation. The significant differences that were detected between B6 controls and ICAMdeficient mice (Fig. 2B, left) were abolished between (⫹/⫹) and (⫺/⫺) F2 littermates (Fig. 2B, right). Because ICAM-1 (⫺/⫺) mice developed fattier livers, heavier fat pads, and hyperlipidemia by day 11 of high-fat feeding, specific mRNA levels were compared from livers of ICAM-1 (⫺/⫺) and B6 male mice at this time point by use of Affymetrix MU6500 GeneChips. RNAs harvested from B6 livers obtained at day 0 were used as baseline controls. To address potential immediate responses to high-fat feeding, day 1 liver total RNA samples were also obtained for expression analysis. To generate pair-wise comparisons, the gene expression data were analyzed with the Affymetrix software. First, pair-wise comparisons were performed in a time-dependent fashion: for each strain, gene expression levels at day 1 and day 11 were compared with the day 0 levels. Table 3, which is available as Supplementary Material at the AJP-Endocrinology and Metabolism web site, lists the genes that were differentially regulated at day 1 and/or day 11 in B6 controls and/or ICAM-1-deficient livers. In addition, pair-wise comparisons were also performed in a strain-dependent way: for each time point, the gene expression levels were compared between B6 control and ICAM-deficient mice. Overall, there were fewer expression changes greater than twofold between the two strains at any of the time points examined than there were within a strain at the three time points of high-fat feeding (data not shown). Pair-wise comparisons of mRNA levels between days 1 or 11 and day 0 indicated that several genes involved in lipogenesis and lipid transport were similarly induced at day 1 but returned to baseline levels by day 11 in both ICAM-1 (⫺/⫺) and B6 male mice (Table 3). In both strains, genes involved in cholesterol synthesis, including hydroxymethyl glutaryl-coenzyme A (HMGCoA) reductase, squalene synthase, and squalene epoxidase, were repressed at day 11. Interestingly, SREBP-1, an important liver transcription factor controlling a large number of genes involved in the metabolism of cholesterol and other lipids, was induced at day 1 and day 11 in ICAM-1-deficient mice only. Among the secreted proteins, the pattern of expression of adipsin markedly differed between the two strains: a dramatic induction of the adipsin message was observed in ICAM-deficient livers (Table 3 and Fig. 3). In addition, significant differences in apoA-4 mRNA levels were also detected between strains. Although the difference in apoA-4 expression was minimal when the pair-wise comparisons were performed in a time-dependent fashion for each strain (day 1 vs. 0 and day 11 vs. day 11, Supplementary Table 3), this was not the case when the pair-wise comparisons were performed AJP-Endocrinol Metab • VOL
Fig. 3. Northern blot showing gene expression in livers of ICAM-1 (⫺/⫺) (I) and C57BL/6J (C) mice fed the high-fat diet for 0, 1, and 11 days. ME, malic enzyme; FAS, fatty acid synthase; mal 1, keratinocyte lipid-binding protein; Sq. Synth., squalene synthase; SREBP-1, sterol regulatory element-binding protein 1; ApoA4, apolipoprotein A4. The 36B4 message was used as a control for gene expression levels. Total RNA samples (20 g) were prepared and probed as described in MATERIALS AND METHODS.
in a strain-dependent fashion (ICAM-deficient vs. B6 control) for a chosen time point. At day 11, the average signal was about fivefold higher in ICAM-deficient livers compared with B6 control livers (average signal 2164 for ICAM-deficient vs. 439 for B6 control). To validate the GeneChip results, a few genes involved in lipid metabolism were arbitrarily selected, and Northern blot analysis was used to confirm differential expression. Lack of ICAM-1 message in the knockout mice was also verified by this method (Fig. 3). As shown in Fig. 3, the transcriptional alterations observed by GeneChip analysis for ME, FAS, mal 1, SREBP-1, apoA-4, and adipsin were confirmed. A comparison of the degree of differences in hybridization intensity over time from the Northern blot analysis and Affymetrix data is shown in Fig. 4. Overall, this graphic comparison of the data shows that the results obtained by the two methods are very similar. The only exception is squalene synthase, for which the message levels were method dependent, both in degree of change and in overall trend (Fig. 4). The PPAR␣ transcription factor pathway has been suggested to be involved in ICAM-1 (⫺/⫺) obesity (6). As shown in Fig. 5, no difference in message levels of either PPAR␣ or CPT-1, a gene transcriptionally controlled by PPAR␣, was found between ICAM-1-deficient and B6 control livers at any of the time points examined. Figure 5A shows the Northern blot, and Fig. 5B shows the relative expression level, expressed as degrees of difference over time, for both Northern blot
282 • MARCH 2002 •
Fig. 4. Comparison of liver gene expression changes by Northern blot and the Affymetrix GeneChip after 0, 1, or 11 days on the high-fat diet for C57BL/6J and ICAM-1 (⫺/⫺) mice. Degree of change in gene expression is based on the day 0 baseline expression level of C57BL/6J. Affymetrix data presented for SREBP-1 reflect averages of the 4 Affymetrix probe sets.
and GeneChip analysis. Although PPAR␣ and CPT-1 mRNA levels did not differ between strains, the Northern blot results revealed that both genes were induced by approximately twofold after 11 days of high-fat feeding (Fig. 5A). In contrast, PPAR␥ mRNA levels were slightly elevated only at day 11 and only in the knockout mice (Fig. 5, A and B). DISCUSSION
ICAM-1-deficient mice were previously reported to develop maturity-onset obesity spontaneously without an increase in food intake and also to demonstrate increased susceptibility to obesity when fed a Western type of high-fat diet (6). These authors suggested that leukocytes can suppress excess triglyceride deposition and may be involved in modulating lipid transport and storage. We sought to understand the underlying molecular mechanism responsible for these reported differences in fat deposition in ICAM-1 (⫺/⫺) animals compared with control animals. Interestingly, our findings did not confirm those of Dong et al. (6). We did not observe spontaneous obesity in either gender, even when ICAM-1 (⫺/⫺) animals AJP-Endocrinol Metab • VOL
were fed chow diet for over a year (data not shown). Moreover, after 50 days of high-fat feeding, no significant increase in body weight, adiposity index, or fat pad weights was observed in male ICAM-1 (⫺/⫺) mice over that of the B6 controls (Table 1). Although we cannot exclude a role of environmental differences such as batch of diet, source of water, type of cage, ambient temperature, and/or the immune environment, we believe that discrepancies in results between the two studies are due to differences in strain background. The animals that Dong et al. (6) used in their work were originally established on a mixed 129/ B6/DBA2 background and were backcrossed four times (N4) to the B6 background. The animals used in this study were similar except that they were backcrossed either eight (N8) or nine (N9 for the F2 generation) times to B6 mice. Therefore, our animals had a significantly higher percentage of B6 genome than the mice used in the study conducted by Dong et al. The contribution of 129/DBA2 genomes at the N4 is statistically expected to be ⬃4.7%, whereas it should be ⬍0.3% at N8-N9. A genome-wide scan for allelic contribution would be necessary to assess the fundamental differ-
282 • MARCH 2002 •
Fig. 5. Comparison of expression changes of peroxisome proliferator-activated receptor-␣ (PPAR␣)-related liver genes by Northern blot and Affymetrix GeneChip methods after 0, 1, or 11 days on the high-fat diet for ICAM-1 (⫺/⫺) (I) and C57BL/6J (C) mice. CPT-1, carnitine palmitoyltransferase 1. The 36B4 message was used as a control for gene expression levels. Total RNA samples (20 g) were prepared and probed as described in MATERIALS AND METHODS. Degree of change in gene expression is based on the day 0 baseline expression level of C57BL/6J.
ences between the strains used in both studies. Our results reemphasize the often-overlooked power of gene interactions and their influence on the phenotype of knockout animals. In our study, the comparison between either B6 and ICAM-1-deficient mice or wild-type and ICAM-1 (⫺/⫺) F2 littermates revealed that significant differences were present at day 11 for inguinal fat and liver tissue weights, expressed as a percentage of body weight. Although these differences were smaller or had vanished by day 50, they may indicate that ICAM deficiency alters the early metabolic response to high-fat feeding. Therefore, we used DNA microarrays to assess the changes in liver gene expression associated with high-fat feeding and hepatic lipid deposition in B6 vs. ICAM-1-deficient mice. In addition to comparing the two mouse strains, we also analyzed different time points to better dissect the metabolic response to highfat feeding. A total of 83 genes were differentially regulated either between strains or within a strain in a time-dependent fashion. These genes were distributed in nine different classes, including cell growth, arrest, and death, transcription factors, and carbohydrate and lipid metabolism enzymes. Initially we were struck by AJP-Endocrinol Metab • VOL
the low number of expression differences between strains at the three time points analyzed. However, upon examination of the data, we found that most of the expression changes were in a similar direction in both strains, but the magnitude of the change differed. As a result, few of the expression changes were greater than twofold when the two strains were compared with one another at the specific time points. The expression differences between strains are a measure of the compensatory changes resulting from the ICAM-1 (⫺/⫺) mice and also of the differences in strain background. Because we were primarily interested in the fatty liver phenotype, several genes falling into the “lipid metabolism” category were chosen for the follow-up studies. Most of these genes were acutely regulated, because they were markedly induced after 1 day and repressed after 11 days of high-fat feeding. This reemphasized the necessity of including multiple time points in gene expression profiling studies, because a single time point allows only the detection of too much (day 1) or too little (day 11) change in mRNA levels. Previous research on hepatic steatosis in mice has examined the control of metabolic pathways involved in this disease, but so far only a few genes have been
282 • MARCH 2002 •
positively correlated with this pathology (23, 36). SREBP-1 protein and FAS mRNA levels were found to be elevated in the fatty livers of ob/ob mice, implicating increased cholesterol and fatty acid biosynthesis as potential causes of the lipid deposition (36). SREBP-1 is an important liver transcription factor controlling a large number of genes involved in the metabolism of cholesterol and other lipids. There are two splice variants of the SREBP-1 protein, known as SREBP-1a and SREBP-1c. Elevated SREBP-1 gene expression is observed at both day 1 and day 11 in ICAM-deficient livers, suggesting that the SREBP-1 pathway is involved in the rapid hepatic lipid accumulation observed in the knockout animals. However, neither the GeneChip nor the Northern blot analysis can differentiate between the splice variants, impairing accurate identification of the regulated isoform. SREBP-1c is a likely candidate, because circulating insulin levels are elevated in ICAM-deficient mice at day 11. Insulin was indeed shown to selectively increase SREBP-1c mRNA in both cultured hepatocytes and diabetic rat liver (8, 35, 36). In contrast, the expression of several SREBP-1 downstream genes, including FAS, ME, ATP-citrate lyase , HMG-CoA reductase, and squalene synthase (34), did not follow a similar pattern and was generally similar in both strains. Transcriptional activation is a very complex phenomenon, and these genes are most likely coregulated by other transcription factors. These additional players may not be differentially regulated by high-fat feeding or may not differ between the two mouse strains. In addition, it should be kept in mind that the mRNA level may not appropriately reflect the protein level and/or the transcriptional activity. Among other genes that were differentially expressed but not selected for confirmatory studies, SPOT14 would deserve further examination. Although SPOT14 expression was repressed after 11 days of high-fat feeding in both strains, its mRNA level was still significantly elevated in ICAM-1 knockout mice compared with B6 control mice. Interestingly, SREBP-1 was shown to regulate SPOT14 gene transcription positively (12, 24). Therefore, the elevated SREBP-1 expression detected in ICAM-deficient mice at day 11 may help to sustain SPOT14 mRNA. Although the function of SPOT14 is still poorly understood, it was reported to play a role in lipid synthesis (17), indicating that it may play a key role in the rapid development of the fatty liver phenotype observed in ICAM-deficient mice. At all time points examined that were independent of the diet, apoA-4 message levels were higher in ICAM-1 (⫺/⫺) animals than in B6 mice, and the differences in message levels were particularly pronounced after 11 days of high-fat feeding. In the knockout mice, both SREBP-1 and PPAR-␣ were induced at day 11, suggesting that perhaps these two transcription factors may act in a coordinate fashion to increase apoA-4 gene expression. Other apolipoproteins, such as apoA-2, are indeed known to be transcriptionally regulated by both SREBP-1 and PPAR␣ (27, 28), and a similar mechanism may be taking place for apoA-4 in AJP-Endocrinol Metab • VOL
ICAM-1-deficient liver. The differential apoA-4 expression observed between ICAM-1 knockout and control mice may also be due to the genetic variation between these strains rather than to the lack of ICAM-1. Striking genetic variations in the levels of apoA-4 mRNA in the liver have previously been reported among inbred mouse strains (30), and, as stated above, the genetic background of ICAM-1-deficient and control mice used in this study significantly differed. The determination of apoA-4 expression levels in (⫺/⫺) and (⫹/⫹) F2 littermates would be necessary to clarify this issue. Overexpression of apoA-4 was shown to result in high plasma triglyceride, free fatty acid, total cholesterol, and HDL cholesterol levels when mice were fed an atherogenic diet (4). Therefore, the elevated apoA-4 expression may contribute to the elevated circulating triglyceride levels observed in ICAM-1 (⫺/⫺) males fed either the chow or high-fat diet. However, triglyceride and cholesterol levels did not correlate well with the pronounced elevation in apoA-4 message detected in high fat-fed knockout mice. Further investigations will be necessary to support this hypothesis. In rodents, the intestine accounts for the major proportion of circulating apoA-4 (15). Therefore, it will be of interest to determine whether the message levels of apoA-4 are also elevated in the intestine of ICAM-1 (⫺/⫺) mice, and to determine whether these mice have elevated levels of plasma apoA-4. The most unexpected and novel finding of this study concerns the dramatic induction of adipsin mRNA in high fat-fed ICAM-deficient livers. In humans and in rodents, the major site of synthesis of adipsin, also known as factor D, is adipose tissue (42), and this is the first study to report adipsin expression in rodent liver. Our data also demonstrate that liver adipsin mRNA levels are diet regulated and indicate that this regulation is tissue specific, because adipose tissue adipsin mRNA levels are not changed by overfeeding or highfat feeding (7, 32). It is very likely that adipsin synthesis occurs in the hepatocyte rather than in any other liver cell type, because adipsin/factor D was recently reported to be constitutively synthesized by normal cultured human hepatocytes (18). Adipsin/factor D, together with factor B and factor C3, are components of the alternative complement pathway circulating in plasma. Interaction among these three factors leads to production of C3a, which is almost immediately converted to C3adesarg, also known as acylation stimulation protein or ASP (2). ASP is a major determinant of the rate of triglyceride synthesis in the adipocyte and therefore may also play a role in hepatic lipid storage. However, assessing ASP production in liver and in adipose tissues would be necessary to determine whether the increase in adipsin mRNA levels correlates with increased level of ASP. Although adipsin gene induction was much more dramatic in ICAM-1deficient livers, adipsin message was also significantly increased in control mice after 11 days of high-fat feeding. Interestingly, both strains had equally fatty liver by day 50, supporting the hypothesis that elevation of adipsin message in liver may contribute to
282 • MARCH 2002 •
hepatic lipid accumulation. In light of our findings, it would be interesting to determine whether the plasma levels of adipsin and/or ASP are altered in ICAM-1 (⫺/⫺) and control mice. The PPAR␣ pathway is also involved in the development of fatty liver: PPAR␣ deficiency led to massive hepatic accumulation in response to short-term starvation or a high-fat diet (16, 22). In addition, the PPAR␣ pathway was suggested to be involved in the development of obesity in ICAM-1 (⫺/⫺) mice (6). However, neither our GeneChip nor our Northern blot data support this hypothesis. No significant difference in expression levels of PPAR␥ or CPT-1, a PPAR␣ downstream gene, was observed between the two strains at any of the three time points. Moreover, both PPAR␣ and CPT-1 mRNA levels were elevated in both strains at day 11, suggesting that fatty acid ␤-oxidation was increased in response to high-fat feeding. Overall, our results clearly indicate that ICAM-1 deficiency does not result in an obesity phenotype, and they question the function of this adhesion molecule in the regulation of body weight and adipose tissue mass. In contrast, the physiological analysis has revealed a differential response to feeding a high-fat diet, mostly in terms of lipid deposition in the liver and in specific fat pads. Additionally, the hepatic mRNA expression analysis has provided a number of tantalizing targets and possible novel regulatory pathways involved in liver lipid metabolism that deserve further study. Confirming these changes at the protein level would be the first necessary step in assessing their biological relevance. We thank Steve Madore and Tim Jatkoe for assistance with the Affymetrix studies, Judy Udove and Eric Kaldjian for histological preparations, and Rong Ni and Arnold Essenburg for excellent technical help. REFERENCES 1. Bourdillon MC, Poston RN, Covacho C, Chignier E, Bricca G, and McGregor JL. ICAM-1 deficiency reduces atherosclerotic lesions in double-knockout mice ApoE(⫺/⫺)/ICAM-1(⫺/⫺) fed a fat or a chow diet. Arterioscler Thromb Vasc Biol 20: 2630–2635, 2000. 2. Cianflone K, Maslowska M, and Sniderman AD. Acylation stimulating protein (ASP), an adipocyte autocrine: new directions. Semin Cell Dev Biol 10: 31–41, 1999. 3. Clausen P, Jacobsen P, Rossing K, Jensen JS, Parving HH, and Feldt-Rasmussen B. Plasma concentrations of VCAM-1 and ICAM-1 are elevated in patients with Type 1 diabetes mellitus with microalbuminuria and overt nephropathy. Diabet Med 17: 644–649, 2000. 4. Cohen RD, Castellani LW, Qiao JH, Van Lenten BJ, Lusis AJ, and Reue K. Reduced aortic lesions and elevated high density lipoprotein levels in transgenic mice overexpressing mouse apolipoprotein A-IV. J Clin Invest 99: 1906–1916, 1997. 5. Coller HA, Grandori C, Tamayo P, Colbert T, Lander ES, Eisenman RN, and Golub TR. Expression analysis with oligonucleotide microarrays reveals that MYC regulates genes involved in growth, cell cycle, signaling, and adhesion. Proc Natl Acad Sci USA 97: 3260–3265, 2000. 6. Dong ZM, Gutierrez-Ramos JC, Coxon A, Mayadas TN, and Wagner DD. A new class of obesity genes encodes leukocyte adhesion receptors. Proc Natl Acad Sci USA 94: 7526–7530, 1997. AJP-Endocrinol Metab • VOL
7. Flier JS, Cook KS, Usher P, and Spiegelman BM. Severely impaired adipsin expression in genetic and acquired obesity. Science 237: 405–408, 1987. 8. Foretz M, Guichard C, Ferre P, and Foufelle F. Sterol regulatory element binding protein-1c is a major mediator of insulin action on the hepatic expression of glucokinase and lipogenesis-related genes. Proc Natl Acad Sci USA 96: 12737– 12742, 1999. 9. Golub TR, Slonim DK, Tamayo P, Huard C, Gaasenbeek M, Mesirov JP, Coller H, Loh ML, Downing JR, Caligiuri MA, Bloomfield CD, and Lander ES. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 286: 531–537, 1999. 10. Gregoire FM, Chomiki N, Kachinskas D, and Warden CH. Cloning and developmental regulation of a novel member of the insulin-like gene family in Caenorhabditis elegans. Biochem Biophys Res Commun 249: 385–390, 1998. 11. Harte RA, Kirk EA, Rosenfeld ME, and LeBoeuf RC. Initiation of hyperinsulinemia and hyperleptinemia is diet dependent in C57BL/6 mice. Horm Metab Res 31: 570–575, 1999. 12. Hashimoto T, Fujita T, Usuda N, Cook W, Qi C, Peters JM, Gonzalez FJ, Yeldandi AV, Rao MS, and Reddy JK. Peroxisomal and mitochondrial fatty acid beta-oxidation in mice nullizygous for both peroxisome proliferator-activated receptor alpha and peroxisomal fatty acyl-CoA oxidase. Genotype correlation with fatty liver phenotype. J Biol Chem 274: 19228– 19236, 1999. 13. Jump DB and Clarke SD. Regulation of gene expression by dietary fat. Annu Rev Nutr 19: 63–90, 1999. 14. Kado S and Nagata N. Circulating intercellular adhesion molecule-1, vascular cell adhesion molecule-1, and E-selectin in patients with type 2 diabetes mellitus. Diabetes Res Clin Pract 46: 143–148, 1999. 15. Kalogeris TJ, Rodriguez MD, and Tso P. Control of synthesis and secretion of intestinal apolipoprotein A-IV by lipid. J Nutr 127: 537S-543S, 1997. 16. 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. 17. Kinlaw WB, Church JL, Harmon J, and Mariash CN. Direct evidence for a role of the “spot 14” protein in the regulation of lipid synthesis. J Biol Chem 270: 16615–16618, 1995. 18. Kitano E and Kitamura H. Synthesis of factor D by normal human hepatocytes. Int Arch Allergy Immunol 122: 299–302, 2000. 19. Klingenspor M, Xu P, Cohen RD, Welch C, and Reue K. Altered gene expression pattern in the fatty liver dystrophy mouse reveals impaired insulin-mediated cytoskeleton dynamics. J Biol Chem 274: 23078–23084, 1999. 20. Laborda J. 36B4 cDNA used as an estradiol-independent mRNA control is the cDNA for human acidic ribosomal phosphoprotein PO. Nucleic Acids Res 19: 3998, 1991. 21. Lee CK, Klopp RG, Weindruch R, and Prolla TA. Gene expression profile of aging and its retardation by caloric restriction. Science 285: 1390–1393, 1999. 22. Leone TC, Weinheimer CJ, and Kelly DP. A critical role for the peroxisome proliferator-activated receptor alpha (PPARalpha) in the cellular fasting response: the PPARalpha-null mouse as a model of fatty acid oxidation disorders. Proc Natl Acad Sci USA 96: 7473–7478, 1999. 23. Lin HZ, Yang SQ, Chuckaree C, Kuhajda F, Ronnet G, and Diehl AM. Metformin reverses fatty liver disease in obese, leptin-deficient mice. Nat Med 6: 998–1003, 2000. 24. Mater MK, Thelen AP, Pan DA, and Jump DP. Sterol response element-binding protein 1c (SREBP-1c) is involved in the polyunsaturated fatty acid suppression of hepatic S14 gene transcription. J Biol Chem 274: 32725–32732, 1999. 25. Nadler ST, Stoehr JP, Schueler KL, Tanimoto G, Yandell BS, and Attie AD. The expression of adipogenic genes is decreased in obesity and diabetes mellitus. Proc Natl Acad Sci USA 97: 11371–11376, 2000. 26. Nishina PM, Lowe S, Verstuyft J, Naggert JK, Kuypers FA, and Paigen B. Effects of dietary fats from animal and plant
282 • MARCH 2002 •
sources on diet-induced fatty streak lesions in C57BL/6J mice. J Lipid Res 34: 1413–1422, 1993. Peters JM, Hennuyer N, Staels B, Fruchart JC, Fievet C, Gonzalez FJ, and Auwerx J. Alterations in lipoprotein metabolism in peroxisome proliferator-activated receptor alphadeficient mice. J Biol Chem 272: 27307–27312, 1997. Pissios P, Kan HY, Nagaoka S, and Zannis VI. SREBP-1 binds to multiple sites and transactivates the human ApoA-II promoter in vitro: SREBP-1 mutants defective in DNA binding or transcriptional activation repress ApoA-II promoter activity. Arterioscler Thromb Vasc Biol 19: 1456–1469, 1999. Plump AS, Smith JD, Hayek T, Aalto-Setala K, Walsh A, Verstuyft JG, Rubin EM, and Breslow JL. Severe hypercholesterolemia and atherosclerosis in apolipoprotein E-deficient mice created by homologous recombination in ES cells. Cell 71: 343–353, 1992. Reue K, Purcell-Huynh DA, Leete TH, Doolittle MH, Durstenfeld A, and Lusis AJ. Genetic variation in mouse apolipoprotein A-IV expression is determined pre- and posttranscriptionally. J Lipid Res 34: 893–903, 1993. Schreyer SA, Wilson DL, and LeBoeuf RC. C57BL/6 mice fed high fat diets as models for diabetes-accelerated atherosclerosis. Atherosclerosis 136: 17–24, 1998. Shillabeer G, Hornford J, Forden JM, Wong NC, Russell JC, and Lau DC. Fatty acid synthase and adipsin mRNA levels in obese and lean JCR:LA-cp rats: effect of diet. J Lipid Res 33: 31–39, 1992. 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. Shimomura I, Bashmakov Y, and Horton JD. Increased levels of nuclear SREBP-1c associated with fatty livers in two mouse models of diabetes mellitus. J Biol Chem 274: 30028– 30032, 1999. Shimomura I, Bashmakov Y, Ikemoto S, Horton JD, Brown MS, and Goldstein JL. Insulin selectively increases
AJP-Endocrinol Metab • VOL
37. 38. 39.
SREBP-1c mRNA in the livers of rats with streptozotocin-induced diabetes. Proc Natl Acad Sci USA 96: 13656–13661, 1999. Shimomura I, Shimano H, Korn BS, Bashmakov Y, and Horton JD. Nuclear sterol regulatory element-binding proteins activate genes responsible for the entire program of unsaturated fatty acid biosynthesis in transgenic mouse liver. J Biol Chem 273: 35299–35306, 1999. Smith BK, Andrews PK, and West DB. Macronutrient diet selection in thirteen mouse strains. Am J Physiol Regulatory Integrative Comp Physiol 278: R797–R805, 2000. Soukas A, Cohen P, Socci ND, and Friedman JM. Leptinspecific patterns of gene expression in white adipose tissue. Genes Dev 14: 963–980, 2000. Surwit RS, Edwards CL, Murthy S, and Petro AE. Transient effects of long-term leptin supplementation in the prevention of diet-induced obesity in mice. Diabetes 49: 1203–1208, 2000. Walker G, Langheinrich AC, Dennhauser E, Bohle RM, Dreyer T, Kreuzer J, Tillmanns H, Braun-Dullaeus RC, and Haberbosch W. 3-Deazaadenosine prevents adhesion molecule expression and atherosclerotic lesion formation in the aortas of C57BL/6J mice. Arterioscler Thromb Vasc Biol 19: 2673–2679, 1999. Wang DQ, Lammert F, Paigen B, and Carey MC. Phenotypic characterization of lith genes that determine susceptibility to cholesterol cholelithiasis in inbred mice. Pathophysiology Of biliary lipid secretion. J Lipid Res 40: 2066–2079, 1999. White RT, Damm D, Hancock N, Rosen BS, Lowell BB, Usher P, Flier JS, and Spiegelman BM. Human adipsin is identical to complement factor D and is expressed at high levels in adipose tissue. J Biol Chem 267: 9210–9213, 1992. Xu H, Gonzalo JA, St Pierre Y, Williams IR, Kupper TS, Cotran RS, Springer TA, and Gutierrez-Ramos JC. Leukocytosis and resistance to septic shock in intercellular adhesion molecule 1-deficient mice. J Exp Med 180: 95–109, 1994. York B, Lei K, and West DB. Sensitivity to dietary obesity linked to a locus on chromosome 15 in a CAST/Ei ⫻ C57BL/6J F2 intercross. Mamm Genome 7: 677–681, 1996.
282 • MARCH 2002 •