![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
,
Department of Internal Medicine, * Department of Pathology, Division of Nutritional Sciences, University of Illinois College of Medicine, Urbana, IL 61801; Department of Medical Nutrition Research and Development, Ross Products Division, Abbott Laboratories, Columbus, OH 43216; and ** The Sapient's Institute, Molecular Genetics in Nutrition Program, Dallas, TX 75209
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ACKNOWLEDGMENT
FOOTNOTES
LITERATURE CITED8
ABSTRACT 
Chronic diseases develop in susceptible individuals following exposure to environmental conditions including high fat diets. Inbred strains of mice differing in susceptibility to atherosclerosis, diabetes, obesity and certain cancers are models for understanding the genetic basis and molecular mechanisms whereby diet influences these polygenic and multifactorial disorders. Expression sequence tags (EST) and disease quantitative trait loci (QTL) are also being identified with these strains. Reported here are comparisons of food intake, growth, nonfasting serum lipids and expression of mRNA for hepatic apolipoprotein E (ApoE), hepatic stearoyl CoA desaturase (Scd1) and heart lipoprotein lipase (Lpl) in a 2 × 2 × 2 design with C57BL/6J and BALB/cByJ mice fed semipurified diets with 4 or 20% saturated (coconut) or unsaturated (corn) oils for 4 mo. Histological studies of aortas and coronary arteries are also reported for these animals. After 4 mo, BALB/cByJ mice were significantly heavier and had significantly higher total serum cholesterol, HDL cholesterol and triglyceride concentrations in the fed state than C57BL/6J mice. Efficiency of utilizing dietary energy did not differ consistently between strains. Oil level affected serum total cholesterol, triglycerides and HDL cholesterol, which were significantly greater in mice fed high fat diets. Lpl and ApoE mRNA expression levels were not significantly affected by mouse strain, oil source or oil level. Scd1 mRNA expression, however, was significantly higher in C57BL/6J than in BALB/cByJ mice and was lower in all mice fed 20% compared with those fed 4% fat diets. Genes regulated differently by diet among strains with distinct susceptibility to diet-influenced disease may be associated with molecular pathways contributing to incidence or severity.
Key words: diet, inbred mouse strains, serum lipids, stearoyl CoA desaturase, disease genes.INTRODUCTION
Epidemiological and laboratory animal studies indicate that diets high in fat increase the incidence and severity of atherosclerosis, diabetes, obesity and cancer in susceptible individuals (reviewed in NRC 1989). Because diet changes disease phenotype, certain dietary components must regulate expression of a subset of genes whose involvement in disease development (Berg 1989
, Kaput et al. 1994, Kirk et al. 1995) appears to be regulated differently in genetically distinct individuals. Except for familial and dominant mutations, chronic diseases are outcomes of contributions from many genes interacting with environmental factors (e.g., Berg 1989, Grundy 1995, Hegele 1992, Kirk et al. 1995, NRC 1989).
Certain inbred strains of mice, established by at least 20 generations of brother × sister matings, show higher susceptibility than other strains to experimentally induced disease. They have been used experimentally to identify quantitative and qualitative chromosomal loci (reviewed in Frankel 1995) and to define molecular events responsible for disease development. Changes in abundances of mRNA-encoding genes participating in lipid metabolism have also been compared within or between inbred mouse strains fed normal diets (Kirk et al. 1995, Srivastava 1996, Srivastava et al. 1991 and 1992) or atherosclerosis-inducing diets (Kirchgessner et al. 1989, LeBeouf et al. 1994, Liao et al. 1993, Qiao et al. 1993, Uelmen et al. 1995, Warden et al. 1989). Based upon the evidence, some diet-regulated or other genes are likely to participate in disease induction or severity. These have been referred to as level and variability genes (Berg 1989, Kirk et al. 1995).
Our laboratories have collaborated in developing a multistep procedure (Elliott et al. 1993, Kaput et al. 1994, Paisley et al. 1996, Swartz et al. 1996) for identifying diet-regulated genes and for testing the hypothesis that such genes participate in disease development (Kaput et al. 1994 and this report). The first step in our protocol involves the isolation of genes or expression sequence tags (EST)7 regulated by diet (Elliott et al. 1993) in tissues of disease-free virgin female mice fed semipurified diets. When desirable, the precision of the model can be refined by the feeding of chemically purified diets. Step 2 analyzes mRNA abundance between inbred strains differing in disease susceptibility before signs of the disease are evident. Our working hypothesis is that genes regulated differently between strains by the same diet may be among the subset involved in producing differences in disease phenotypes between strains. The third step compares the chromosomal map position of the differently regulated genes to independently derived quantitative trait loci (QTL). Others have proposed mapping disease-specific expressed sequence tags (EST) with independently derived QTL (Berry et al. 1995) for identifying candidate disease genes, an example of association analyses (Risch and Merikangas 1996). EST regulated by the same nutritional factors that produce the disease and overlapping disease QTL maps have been defined as candidate disease genes (Risch and Merikangas 1996) for analyses in humans or animals showing the disease, the last step of our proposed protocol.
We previously isolated hepatic apolipoprotein E (ApoE) and stearoyl CoA desaturase (Scd1) (Elliott et al. 1993) using the above described experimental model in screens for diet-regulated genes. We report here the analyses of hepatic ApoE, hepatic Scd1, and heart Lpl mRNA abundance in C57BL/6J and BALB/cByJ mice fed semipurified diets with 4 or 20% corn or coconut oils for 4 mo. C57BL/6 are more susceptible to diet-induced atherosclerosis (Paigen et al. 1987 and 1990), type II diabetes (Seldin et al. 1994, Surwit et al. 1995), and express certain genes involved in lipid metabolism differently than BALB/c mice (Kirk et al. 1995 and this report). Food intake, growth, nonfasting serum lipids, histological sections of aortas and coronary arteries were also analyzed to assess their influence or correlation with gene regulation. This report is an example of the second step of our protocol.
MATERIALS AND METHODS
). The diets (Table 1), formulated according to modified AIN-76 guidelines (NRC 1995, Reeves et al. 1993), were pelleted and color coded by Research Diets (New Brunswick, NJ). Mice were caged and fed individually with free access to food and distilled water in temperature-controlled rooms maintained at 23 ± 1°C with a 12-h light:dark cycle. Animal care met University of Illinois and National Institutes of Health guidelines. Food spillage was also monitored throughout the course of the experiment. Efficiencies of energy utilization were calculated from the recorded weekly weight gain/calculated weekly energy intake. At 16 wk, all mice were deprived of food for 12 h and offered a preweighed 3-g pellet of their assigned diet. After 2 h, the uneaten food was removed; 2 h later, all mice were injected intramuscularly with 0.02 mL/g body weight of ketaset/xylazine mixture (Ketaset, Fort Dodge Laboratories, Ft. Dodge, IA) for collection of blood via cardiac puncture. Immediately thereafter, they were killed by cervical dislocation and their livers and hearts were removed, individually frozen in liquid nitrogen, and stored at 
80°C for mRNA isolation.
|
Table 1. Diet composition |
-actin and stearoyl CoA desaturase (Scd1, Genbank/EMBL: J04190, M21285) were as described previously (Paisley et al. 1996, Swartz et al. 1996). A DNA fragment encoding nucleotides 681 to 806 of apolipoprotein E mRNA (Genbank/EMBL: D00466, M12414) was isolated in a second screen for diet-regulated genes. cDNA were excised from the vectors, isolated (Ausubel et al. 1987) and labeled by random priming or polymerase chain reaction (PCR) amplification in the presence of 
-32P-deoxycytidine-5-triphosphate (dCTP, 111 TBq/mmol, ICN, Irvine, CA) as described by Paisley at al. (1996) and Swartz et al. (1996).
Northern hybridization analyses.
Twenty micrograms of total RNA, isolated from livers by the Ultra-spec II RNA isolation system (Biotecx, Houston, TX), was resolved by electrophoresis in gels containing 1.2% agarose and 2.2 mol/L formaldehyde for 2.5 h at 110 V. Sizes of mRNA were estimated by comparisons to a synthetic RNA ladder electrophoresed simultaneously. Gels were subsequently stained with 0.5 mg/L ethidium bromide and photographed to verify rRNA quality. RNA was blotted onto nylon membranes (U.S. Biochemical, Cleveland, OH) following standard protocols (Ausubel et al. 1987). Blots were prehybridized at 42°C with 50% formamide, 5X standard sodium citrate (SSC), 1X Denhardt's solution, 0.02 mol/L sodium phosphate (pH 6.8), 0.1 g/L denatured salmon sperm DNA, 1 mmol/L EDTA, 0.2% SDS and 10% dextran sulfate (Ausubel et al. 1987). After overnight hybridization at 42°C, the membranes were washed twice at room temperature with 1X SSC containing 0.5% SDS for 20 min. Two 15-min washes with 0.5X SSC/0.5% SDS and one 15-min wash in 0.25X SSC/0.5% SDS at 65°C followed. Membranes were placed next to preflashed (Laskey 1980) Kodak XAR film at 80°C, and electrophoretic patterns were analyzed in the linear range of the film (Laskey 1980). Membranes of liver RNAs were rehybridized with a -actin to control for the total RNA per well. Glyceraldehyde-3-phosophate dehydrogenase (Paisley et al. 1996) served as the control for RNA in hearts, which express two actin mRNAs. -Actin and GAPDH are believed to be unaffected by changes in metabolism and are used as controls for mRNA abundance measurements (e.g., Ausubel et al. 1987). Because ethidium bromide staining of nucleic acids is a relatively insensitive measure of RNA loading with 10-20 µg loaded per lane, mRNA abundance is reported as the ratio of specific transcripts to control transcripts. RNA from livers of 80 mice (10 from each dietary treatment) and hearts from 40 mice (5 from each dietary treatment) were analyzed. Hybridization signals were quantified with a Molecular Dynamics 425S phosphorimager (Molecular Dynamics, Sunnyvale, CA). Abundances of specific mRNA were normalized to the amount of -actin or GAPDH mRNA within the same electrophoretic lane.
|
Table 2. Serum total cholesterol, HDL cholesterol and triglyceride concentrations in BALB/cByJ and C57BL/6J mice fed 4 or 20% corn oil or coconut oil for 15 wk1,2 |
Serum lipids, histology and statistical analyses. Serum lipid profiles were determined for total cholesterol (Sigma Kit 352-20), triglycerides (Sigma 339-10) and HDL cholesterol (Sigma 352-3) using kit protocols (Sigma Chemical, St. Louis, MO). Samples were analyzed in duplicate and reaction mixtures were adjusted to 200 µL for the well volume of 96 well plates. Hearts and aortas from 40 mice (5 from each dietary treatment) were fixed in Karnovski's fixative and embedded in JB4 media (Polysciences, Warrington, PA) for sectioning, staining (Aparicio and Marsden 1969
) and microscopic examination at magnifications of 40X and 400X. Identity of all sections was coded and unknown to the examiner. The Statistical Analyses System-General Linear Model (SAS GLM) program (SAS 1987) was employed for statistical computing with post-hoc analyses of statistical significance by Tukey's Multiple Range Comparison (HSD) (Steele and Torrie 1980). Results are reported as means ± SEM.
RESULTS
Serum lipid concentrations. Fed BALB/cByJ mice had higher serum triglycerides (P < 0.01), HDL cholesterol (P < 0.0001) and total cholesterol (P < 0.0001) than did C57BL/6J mice (Table 2). The higher concentration of dietary oil also resulted in greater serum HDL cholesterol (P < 0.01), serum triglycerides (P < 0.001) and total cholesterol (P < 0.001) (Table 2). Serum triglyceride concentrations showed significant interactions between strain and oil source (P < 0.001) and between strain and dietary oil level and source (P < 0.01) (Fig. 2), but no interactions were significant for cholesterol (Fig. 3). Interactions tended to be significant between strain and dietary oil concentration (P < 0.06) and between strain and source (P < 0.08) and source and dietary oil level (P < 0.08) for serum total cholesterol concentrations (Fig. 4). Significant correlations between total cholesterol and HDL cholesterol (r = 0.631, P < 0.001) and triglycerides and total cholesterol were found (Table 3). There were no significant correlations between these serum lipid concentrations and body weight, total energy intake or energy eaten in the last meal (not shown).
|
Table 3. Pearson correlation analysis of strain, serum total cholesterol, HDL cholesterol and triglyceride concentrations in BALB/cByJ and C57BL/6J mice fed 4 or 20% corn oil or coconut oil for 15 wk1,2,3 |
-actin mRNA from individual BALB/cByJ (Fig. 5A) and C57BL/6J (Fig. 5B) mice fed one of the four diets with averages of Scd1/-actin mRNA ratios for all in each treatment shown in Figure 6A (BALB/cByJ) and Figure 6B (C57BL/6J). Relative Scd1 levels were significantly higher (P < 0.03) in mice fed diets containing 4% of either oil even though mice eating 20% oil diets consumed more energy during their last meal. Scd1/-actin mRNA ratios were significantly higher (P < 0.04) in C57BL/6J mice, but there were no significant differences in Scd1/-actin between corn oil- and coconut oil-fed mice. There were no significant interactions between strain and dietary source or concentration. Significant differences in hepatic ApoE and heart lipoprotein lipase mRNA expression due to strain, oil level or oil source were not found (data not shown).
-actin probes as described in Materials and Methods. Similar data from 8-10 mice from each dietary treatment are summarized in Figure 4.
-actin mRNA, from 8-10 BALB/cByJ (A) and 8-10 C57BL/6J (B) mice of each dietary treatment were analyzed. Results are means ± SEM. Scd1/-actin mRNA ratio was greater in C57BL/6J mice (P < 0.04) and higher in all mice fed 4% of either oil compared with those eating 20% of either oil.
0.237) between Scd1/-actin and HDL levels that approached significance (P < 0.08). Lpl/GAPDH mRNA ratios were significantly correlated (r = 0.517, P < 0.01 ) with cholesterol levels and there tended to be a negative correlation (r = 0.398) approaching significance between Scd1/-actin and Lpl/GAPDH mRNA ratios (P < 0.08).
DISCUSSION
. Our analyses were done in the absence of atherogenic chemicals added to a semipurified diet and without evidence of disease processes, which might influence gene expression or serum lipid concentrations in strains which differ in predisposition to diet-influenced diseases.
). The increases in total cholesterol, triglycerides and HDL cholesterol were qualitatively and in some cases quantitatively similar to those seen in these and other strains fed high fat diets with and without added cholesterol (Kirk et al. 1995, Srivastava 1996, Srivastava et al. 1991 and 1992) and atherogenic diets containing added cholic acid, cocoa butter and cholesterol (Hwa et al. 1992, LeBeouf et al. 1994, Nishina et al. 1993, Paigen et al. 1990, Warden et al. 1989). Saturated fat (coconut oil) significantly increased triglyceride levels in C57BL/6J but not in BALB/cByJ mice, and it did not alter total serum cholesterol or HDL cholesterol concentrations. Serum triglyceride concentrations were also affected by interactions among strain, oil level and oil source, demonstrating the importance of nutrient-genotype interactions. The variation in concentration in serum lipids for the same strains reported by different laboratories may be related to differences in diet composition because fatty acids and other ingredients vary among preparations of unpurified diets (e.g., Lardinos et al. 1989). Variations in serum lipid concentrations may also result from differences among studies in time between the last meal and the time of sampling (Kirk et al. 1995).
), Lpl mRNA levels are presumably not transcriptionally regulated by diet in hearts of mice (Kirchgessner et al. 1989) or rats (Erskine et al. 1993). Our Lpl data agree with this conclusion. Hepatic ApoE mRNA abundance was not regulated by level or source of oil. Others (Ishida et al. 1990) found strain (C57BL/6 and C3H/J)-specific differences in serum apolipoprotein E concentrations afterovernight food deprivation in mice fed unpurified vs. a highly atherogenic diet, suggesting that our results may be confounded because we analyzed expression 2 h postprandially. We previously found significant effects of fat level and time of eating on the regulation of Lpl (Paisley et al. 1996) and ApoE in 10-wk-old BALB/cHnn mice denied food for 12 h compared with those that had eaten. Some diet-regulated genes differ in their expression related to ime of eating (LeBeouf et al. 1994, Paisley et al. 1996, Swartz et al. 1996) or age (Rao et al. 1989), showing that experimental variables in addition to dietary constituents may alter mRNA concentrations.
, Paisley et al. 1996). Others found that Scd1 was regulated by saturation of dietary fat and by carbohydrates in livers of CD-1 male mice and rats (reviewed in Ntambi 1995), and more recently, by the peroxisome proliferator-inducing chemical, clofibrate (Diczfalusy et al. 1995). Hepatic Scd1/-actin mRNA ratios tended to be higher in coconut oil-fed mice than in corn oil-fed mice. Diet composition or time between the last meal and time of killing may have been responsible for the observed differences between our studies and others (Ntambi 1992 and 1995). Scd1 expression showed significant strain differences, with BALB/c mice having lower expression than C57BL/6 mice, a strain susceptible to diet-induced atherosclerosis (Paigen et al. 1990), obesity and type II diabetes (Seldin et al. 1994, Surwit et al. 1995).
), may be identified genetically if they map within disease or complex trait loci (e.g., QTL) and if they have some association with the disease process (Risch and Merikangas 1996). A small but growing number of disease QTL have been identified in humans and in laboratory animals (e.g., Berry et al. 1995, Risch et al. 1993, Seldin et al. 1994, and reviewed in Frankel 1995, Sim et al. 1995), but many possible loci remain unidentified because of the number of possible nutritional and genetic combinations yet to be analyzed (e.g., Patterson et al. 1991, Risch and Merikangas 1996). Consequently, genes not mapping to a QTL can not be excluded from participating in disease processes. ApoE maps to a region of mouse chromosome (Chr) 7 (Encylcopedia of the Mouse Genome 1996), which overlaps a Type I diabetic QTL (Risch et al. 1993), and its overexpression in transgenic mice reduced the hyperlipidemia associated with experimentally induced diabetes (Yamamoto et al. 1995). Neither Scd1 (Chr19) (Keller et al. 1994) nor Lpl (Chr8) (Encylcopedia of the Mouse Genome 1996) map to known disease QTL, although Scd1 can be misregulated by streptozocin in rats (Waters and Ntambi 1995) and experimentally induced hypertriglyceridemia in diabetic mice was decreased in transgenic mice overexpressing Lpl (Shimada et al. 1995). Additional experiments are required to test whether Scd1 or other diet-regulated genes are involved in the molecular mechanism of disease.
ACKNOWLEDGMENT
FOOTNOTES
Manuscript received 7 August 1996. Initial reviews completed 23 September 1996. Revision accepted 10 January 1997.
LITERATURE CITED8
This article has been cited by other articles:
|
|
 |
|
  J. Kaput Nutrient selection through nutrigenomic approaches Am J Physiol Regulatory Integrative Comp Physiol, January 1, 2007; 292(1): R204 - R206. [Full Text] [PDF]
|
|
|
|
 |
|
 J. KAPUT Decoding the Pyramid: A Systems-Biological Approach to Nutrigenomics Ann. N.Y. Acad. Sci., December 1, 2005; 1055(1): 64 - 79. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Mutch, W. Wahli, and G. Williamson Nutrigenomics and nutrigenetics: the emerging faces of nutrition FASEB J, October 1, 2005; 19(12): 1602 - 1616. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kaput, K. G. Klein, E. J. Reyes, W. A. Kibbe, C. A. Cooney, B. Jovanovic, W. J. Visek, and G. L. Wolff Identification of genes contributing to the obese yellow Avy phenotype: caloric restriction, genotype, diet x genotype interactions Physiol Genomics, August 11, 2004; 18(3): 316 - 324. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Kaput and R. L. Rodriguez Nutritional genomics: the next frontier in the postgenomic era Physiol Genomics, January 15, 2004; 16(2): 166 - 177. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Castro-Chavez, V. K. Yechoor, P. K. Saha, J. Martinez-Botas, E. C. Wooten, S. Sharma, P. O'Connell, H. Taegtmeyer, and L. Chan Coordinated Upregulation of Oxidative Pathways and Downregulation of Lipid Biosynthesis Underlie Obesity Resistance in Perilipin Knockout Mice: A Microarray Gene Expression Profile Diabetes, November 1, 2003; 52(11): 2666 - 2674. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
