TOXICOLOGY
AND
APPLIED
PHARMACOLOGY
115,
156- 160 ( 1991)
Oxidative DNA Damage Levels in Rats Fed Low-Fat, High-Fat, or Calorie-Restricted Diets Z. DJURIC,’ Division
M. H. Lu,* S. M. LEWIS,* D. A. LUONGO, X. W. CHEN,* L. K. HEILBRUN, B. A. READING, P. H. DUFFY,* AND R. W. HART*
y/‘Hernatology
and *National
Ot~olog~, C’enter,for
Received
P.O. Box 02158, U hyne To.uicological Re.tearch.
December
State Univer.tir~~. Detroit. Michigatz .Jeffi’r.~or~. .-lrkan.sa.r 72079
13. 1991; accepted
March
48201;
and
30. 1992
been suggestedto modulate breast cancer risk in humans (Klurfeld and Kritchevsky, 1990; Schatskin et al., 1990a). Fat in the diet is thought to act as a promotional agent during the development of mammary tumors, and this was first described by Tannenbaum (1942). One mechanism of tumor promotion by dietary fat may involve the metabolism Increasedfat and caloric content of the diet hasheenassociated of fats. Fatty acids can be oxidized to form reactive oxygen with increasedmammary tumor incidence. The dietary modu- speciesthat can damage cellular constituents, such as DNA lation of cellular redox state may be one mechanismbehind this (Ames, 1983). In our laboratory. we have shown that a deassociation.We have examined the effects of changesin dietary creasein dietary fat can decreasethe levels of oxidative DNA fat and caloric intake on the levels of 5hydroxymethyluracil in damage, specifically 5-hydroxymethyluracil, in peripheral DNA from rat liver and mammary gland. Female Fischer 344 nucleated blood cells of women at high risk for breast cancer rats, 40 daysold, were maintainedon 3% (low-fat), 5% (control), or 20% (high-fat) corn oil diets for 2 weeks.A fourth group of (Djuric et ~1.. 1991a). In addition to dietary fat, the level of caloric intake also rats had the samedaily fat intake as the control group, but total caloric intake wasrestricted by 40%. As a measureof oxidative can modulate tumorigenesis.Mammary tumor incidence has DNA damage,5-hydroxymethyluracil levels were measuredin been shown to be decreaseddramatically by a 40% reduction the DNA extracted from liver and mammary gland by gaschro- in caloric intake (Klurfeld et al., 1989). Caloric restriction matography-massspectrometry. 5-Hydroxymethyluracil levels can extend the lifespan of rats and has beneficial effects on in the liver DNA of the low-fat, high-fat, and calorie-restricted the prevention of numerous diseases.One mechanism behind groups were decreasedrelative to that of control, but the only these observations may be the decreasedproduction of oxsignificantdecreasewasin the calorie-restrictedgroup (p < 0.01). In the mammary gland DNA, statistically significant decreases ygen-free radicals during caloric restriction (Weindruch and in damagewere found in each group relative to control (p < Walford, 1988). This could result from increased metabolic 0.05). The relationship between fat in the diet and oxidative efficiency or from increased detoxification of free radicals. stressis thus complex. Theseresultsshowthat changesin dietary Both of these pathways have been shown to be increased intake of both fat and calories can modulate oxidative DNA during caloric restriction of animals (Koizumi et al., 1987: damagelevels, and the effect of diet wasmore clearly evident in Feuers et al.. 1989) and this could decreasethe endogenous the DNA from mammary gland than in DNA from liver. c, 1992 levels of oxidative DNA damage. Another factor that can Academic Press, Inc. decreaseDNA damagelevels is DNA repair, and DNA repair ability has been shown to be increased by caloric restriction (Weraarchakul et al.. 1989). Two dietary factors which have been shown to modulate The levels of oxidative DNA damage therefore may be mammary tumor incidence in mice and rats are fat and cal- influenced by both fat and caloric intake. Oxidative DNA ories. The relative importance of these two factors is still a damage has been implicated in the process of tumor promatter of controversy, but a review of 100 studies in rats motion (Cerutti. 1987; Frenkel et al., 1986: Reid and Loeb. and mice has indicated that decreasedfat and caloric content 1992; McBride et ul.. 1991). When DNA is exposed to oxiof the diet each independently reduce mammary tumor in- dants, various types of modified DNA basescan be produced. cidence (Freedman et al., 1990). These two factors also have Of these products, relatively high levels of 5-hydroxymethyluracil have been found in human urine (reviewed in Mullaart et al.. 1990). 5-Hydroxymethyluracil is potentially a muta’ To whom correspondence should be addressed. Oxidative DNA Damage Levels in Rats Fed Low-Fat, HighFat, or Calorie-Restricted Diets. DJURIC, Z., Lu, M. H., LEWIS, S. M., LUONGO, D. A., CHEN, X. W., HEILBRUN, L. K., READING, B. A., DUFFY, P. H., AND HART, R. W. (1992). Toxicol. Appf. Pharmacol. 115, 156-l 60.
0041-008x/92
$5.00
Copyright rB 1992 by Academic Press. Inc. All rights of reproductmn in any form reserved.
156
DNA DAMAGE
genie lesion (Shirname-More et al., 1987). Unlike some base oxidation products, 5-hydroxymethyluracil appears to be relatively stable (Frenkel et al., 1986). In this study, we have investigated the effects of dietary fat and caloric restriction on the levels of 5-hydroxymethyluracil in the DNA from liver and mammary gland of Fischer 344 rats. MATERIALS
AND METHODS
Diets. The diets consisted of ingredients, including the vitamin and mineral mixes, that were obtained from Teklad Premier Laboratory Diets (a division of Harlan Sprague-Dawley, Inc., Madison, WI). The diets prepared were: ( I) 3% corn oil fed ad lib (low fat), (2) 5% corn oil fed ad lib (control), (3) 20%’ corn oil fed ad lib (high fat). and (4) 5% corn oil fed at 40% caloric restriction relative to the uu’ lib control diet (CR). The formulation of these diets is shown in Table I. All diets, with the exception of the high-fat diet. were prepared by blending the vitamin and mineral mixes with the dextrose in a Patterson-Kelley Vblender (HARSCO Corp.. East Stroudsburg, PA) for I5 min. The corn oil was hand-blended into the dry ingredients to obtain a thorough blend. The remaining ingredients were then added and the entire mixture was reblended in the V-blender for I5 min with the addition of an intensifier bar apparatus, This intensifier bar, with a peripheral speed of 1700 fpm. breaks soft agglomeratcs that can be found in raw materials or are formed during blending. The high-fat diet was prepared by dividing the total dextrose, corn oil, vitamin. and mineral mixes into three equal amounts, and each fraction was blended with the V-blender and by hand. as described above for the other three diets. By decreasing the amount of material being blended, a uniform blend with the required 20% corn oil was obtained. The fractions were subsequently mixed together and reblended for I5 min with the intensilier bar apparatus. The diets were stored refrigerated. Tests of homogeneity were performed on each diet by analyzing for zinc and calcium in samples of the blended diets. As the diets were removed from the V-blender, two 60-g samples were removed from three locations: the top. middle. and bottom layer of the V-blender. The total energy values given as kcal/g diet were calculated using the standard Atwater value estimates for physiological fuel values (Harper, 1979). Carbohydrates, protein, and cellulose were calculated at 4 kcal/g and fat at 9 kcal/g. Cellulose is often considered to be an inert filler: however, it can be fermented in the hindgut to a limited extent, ranging from I2 to 20% (VanSoest. 1982). In our calculation of the energy content of the diets. 12% digestibility of cellulose was used. Animals. Female. Fischer 344 rats were weaned at 2 I days of age and maintained on NIH-3 I diets. At 38 days of age. the rats were placed on the control diet (see Table I). Food intake was monitored such that the amount of food to be fed to the calorically restricted rats could be determined. At an age of 45 days. the animals were weighed and divided arbitrarily into the four diet groups with 12 animals per group. The animals were maintained on the diets for 2 weeks with daily food changes. The animals were housed singly in hanging, mesh-bottom cages fitted with trays to facilitate the collection of spilled food for the determination of total intake. Food intake, body weight, and the general health of the animals were monitored on a daily basis. The diets were presented to the animals in jars, and food intake was calculated from the amount of food left in the jar the next day plus the amount of food spilled. The lights in the animal room were kept on a schedule of 12 hr on and 12 hr off. The food was presented to the animals immediately before the lights were turned off. All animals thus consumed their diets during the same time period. After 2 weeks, the animals were euthanized by carbon dioxide inhalation and decapitation. The mammary gland tissue and livers were removed. Whole livers were frozen. The mammary gland epithelium was isolated from fresh
157
AND DIET
TABLE 1 Formulation of Diets Diet group Ingredient”
Control
High fat
Low fat
CRh
Corn oil Casein Dextrose Cellulose Vitamin mix (AIN76) Mineral mix (AlN76) DL-Methionine Choline bitartrate kcal/OO g diet
5 30 65 5 I
20 20 50 5 I
3 ‘0 67 5 I
x.33 33.3 41.8 x.33 1.67 6.16
3.7 0.3
3.7 0.3
3.7 0.3
0.2
0.2
0.2
387
462
377
0.5
0.2 379
’ Ingredients are given as percentage, by weight, in the diet. ’ The CR (calorically restricted) diet is formulated as indicated and fed at 40% restriction relative to the ad lib intake of the control diet.
mammary gland tissue by the method of Moon et al. (1969). Briefly. the tissue was minced and incubated with collagenase and the fat was separated from the epithelial cells by centrifugation. The epithelial cells were frozen until DNA extractions could be performed. DNA extraction. The mammary gland epithelium and liver tissue were homogenized in 1% sodium dodecyl sulfate and I mM EDTA and the DNA was extracted as previously described (Djuric E( a/.. 1988). The procedure involved incubation of the homogenates with protease followed by successive organic extractions with phenol. chloroform/isoamyl alcohol/phenol (24: 1:25) and chloroform/isoamyl alcohol (24: 1). treatment of the semipurified nucleic acid with RNases, and another series of the same organic extractions. The DNA was recovered by precipitation and quantified spectrophotometrically. Analysis spectrometry
of S-hydroxymethyluracil
levels by gas chromatography-mass
The purified DNA samples were hydrolyzed to (X-MS). nucleosides enzymatically. and trimethylsilyl derivatives were prepared for analysis by GC-MS as previously described (Djuric e( al.. 199 lb). The GCMS analyses were carried out with a Hewlett-Packard 597 I mass spectrometer operated in the EI mode. and positive ions were monitored. The HewlettPackard 5890 CC was operated in the splitless injection mode and injections were made with a Hewlett-Packard 7673 autosampler. Statistical methods. Dependent variables were analyzed to determine the effectof the four experimental diets on mean DNA damage level, nutrient intake, and animal weight. Nutrient intakes were calculated based on a simple mean food intahe over the I4 days of study. DNA damage level was assessed at Day I4 only. Due to the heavily nonnormal distribution of DNA damage level and nutrient intake, nonparametric methods of statistical analysis were used. For each dependent variable. separate analyses were performed using the k-sample Kruskal-Wallis rank-sum test to compare mean levels by diet group (Hollander and Wolfe, 1973). To perform the multiple comparisons of interest (all pairwise comparisons of diet groups), a nonparametric rank-sum method was used. based on a large sample approximation for balanced designs (Miller, 1981). This allowed us to maintain the experimentwise Type I error rate at 0.05 while performing all pairwise comparisons of diet groups (Miller. 1981). There were I3 rats for each diet group. yielding a completely balanced design. and there were no missing data. Rats were consecutively and arbitrarily assigned to diet groups. The data were viewed as resulting from separate ( 1 X 4) complete one-way layout experimental designs.
1.58
DJURIC ET AL
RESULTS
Diets and Food Intake The prepared diets were homogeneous as determined by the analyses of calcium and zinc in samples of the diets removed from the top, middle, and bottom of the V-blender. The values for percentage of calcium (by weight) in the control, high-fat, low-fat, and CR diets were 0.147 f 0.010, 0.137 * 0.004, 0.146 + 0.004, and 0.196 * 0.002, respectively, given as means + standard deviation. The values for zinc, in mean mg zinc/kg diet + standard deviation, were 44.5 ? 3.10, 40.2 t- 2.63, 40.2 -+_2.21, and 69.5 -+ 1.29. The maximum standard deviation in any individual diet preparation after blending was only 7% of the mean. The caloric content of each diet was calculated as described under Materials and Methods. In terms of kcal/g diet, all the diets were similar with the exception of the high-fat diet, which had a considerably higher energy content (Table 1). Mean food intake was significantly different by diet group (p < 0.0001) because the mean for the CR rats was significantly different from that of all other diets, especially the low-fat and control diets (Table 2). Food consumption in the high-fat group was, however, 12% lower than that in the control group. With regard to mean daily caloric intake, the significant difference between diet groups was due to the decreased caloric intake of the CR group.
The impact of reducing caloric intake was evident in the animal body weights. Initially the mean animal weight did not differ significantly by diet group. After 2 weeks on the diets, mean animal body weight did differ significantly by diet group, and this was entirely due to the CR mean being significantly lower than the other three means (Table 2). Daily fat intake differed significantly by diet group and in the expected fashion: the means for the high-fat and low-fat rats each differed significantly from that of the other three diet groups (Table 2). Mean fat intake of the high-fat group was three- to sixfold higher than that in any of the other diet groups. Conversely, the daily fat intake of the low-fat group was lower than that in any of the other groups. The daily fat intake of the CR and control groups was virtually the same. DNA Damage Levels As shown in Table 3, the mean liver DNA damage level differed significantly across the four diet groups (p = 0.002). This difference was entirely due to only two pairs of diet groups, as revealed by the rank-sum multiple-comparisons analysis. Mean DNA damage levels in the liver of CR rats differed significantly from that of control rats (p < 0.0 1) and that of rats in the high-fat group (p < 0.05). Mean 5-hydroxymethyluracil levels in the liver DNA of rats fed highfat and low-fat diets were decreased relative to control, but this was not significant.
TABLE 2 Mean Dietary Intakes and Body Weights by Diet Group Diet group” Variable Food intake (g/day)
Control 12.10 f 1.07
High fat
Low fat
10.67 f 0.79
11.84 f 0.54
d Caloric intake (kcal/day)
46.81 !I 4.13
49.27 k 3.64
44.62 f 2.04
d Fat intake (g/day)
0.61 + 0.05
2.13 + 0.16
0.36 kO.02 d
d e Initial body weight (g/animal) Final body weight (g/animal)
105.3 f 7.7 137.7 + 9.7
111.3 k5.9 146.9 k 9.2 c
p Values/
CR 7.14 + 0.07 l’ d 26.77 f
0.27
AND
APPLIED
PHARMACOLOGY
115,
156- 160 ( 1991)
Oxidative DNA Damage Levels in Rats Fed Low-Fat, High-Fat, or Calorie-Restricted Diets Z. DJURIC,’ Division
M. H. Lu,* S. M. LEWIS,* D. A. LUONGO, X. W. CHEN,* L. K. HEILBRUN, B. A. READING, P. H. DUFFY,* AND R. W. HART*
y/‘Hernatology
and *National
Ot~olog~, C’enter,for
Received
P.O. Box 02158, U hyne To.uicological Re.tearch.
December
State Univer.tir~~. Detroit. Michigatz .Jeffi’r.~or~. .-lrkan.sa.r 72079
13. 1991; accepted
March
48201;
and
30. 1992
been suggestedto modulate breast cancer risk in humans (Klurfeld and Kritchevsky, 1990; Schatskin et al., 1990a). Fat in the diet is thought to act as a promotional agent during the development of mammary tumors, and this was first described by Tannenbaum (1942). One mechanism of tumor promotion by dietary fat may involve the metabolism Increasedfat and caloric content of the diet hasheenassociated of fats. Fatty acids can be oxidized to form reactive oxygen with increasedmammary tumor incidence. The dietary modu- speciesthat can damage cellular constituents, such as DNA lation of cellular redox state may be one mechanismbehind this (Ames, 1983). In our laboratory. we have shown that a deassociation.We have examined the effects of changesin dietary creasein dietary fat can decreasethe levels of oxidative DNA fat and caloric intake on the levels of 5hydroxymethyluracil in damage, specifically 5-hydroxymethyluracil, in peripheral DNA from rat liver and mammary gland. Female Fischer 344 nucleated blood cells of women at high risk for breast cancer rats, 40 daysold, were maintainedon 3% (low-fat), 5% (control), or 20% (high-fat) corn oil diets for 2 weeks.A fourth group of (Djuric et ~1.. 1991a). In addition to dietary fat, the level of caloric intake also rats had the samedaily fat intake as the control group, but total caloric intake wasrestricted by 40%. As a measureof oxidative can modulate tumorigenesis.Mammary tumor incidence has DNA damage,5-hydroxymethyluracil levels were measuredin been shown to be decreaseddramatically by a 40% reduction the DNA extracted from liver and mammary gland by gaschro- in caloric intake (Klurfeld et al., 1989). Caloric restriction matography-massspectrometry. 5-Hydroxymethyluracil levels can extend the lifespan of rats and has beneficial effects on in the liver DNA of the low-fat, high-fat, and calorie-restricted the prevention of numerous diseases.One mechanism behind groups were decreasedrelative to that of control, but the only these observations may be the decreasedproduction of oxsignificantdecreasewasin the calorie-restrictedgroup (p < 0.01). In the mammary gland DNA, statistically significant decreases ygen-free radicals during caloric restriction (Weindruch and in damagewere found in each group relative to control (p < Walford, 1988). This could result from increased metabolic 0.05). The relationship between fat in the diet and oxidative efficiency or from increased detoxification of free radicals. stressis thus complex. Theseresultsshowthat changesin dietary Both of these pathways have been shown to be increased intake of both fat and calories can modulate oxidative DNA during caloric restriction of animals (Koizumi et al., 1987: damagelevels, and the effect of diet wasmore clearly evident in Feuers et al.. 1989) and this could decreasethe endogenous the DNA from mammary gland than in DNA from liver. c, 1992 levels of oxidative DNA damage. Another factor that can Academic Press, Inc. decreaseDNA damagelevels is DNA repair, and DNA repair ability has been shown to be increased by caloric restriction (Weraarchakul et al.. 1989). Two dietary factors which have been shown to modulate The levels of oxidative DNA damage therefore may be mammary tumor incidence in mice and rats are fat and cal- influenced by both fat and caloric intake. Oxidative DNA ories. The relative importance of these two factors is still a damage has been implicated in the process of tumor promatter of controversy, but a review of 100 studies in rats motion (Cerutti. 1987; Frenkel et al., 1986: Reid and Loeb. and mice has indicated that decreasedfat and caloric content 1992; McBride et ul.. 1991). When DNA is exposed to oxiof the diet each independently reduce mammary tumor in- dants, various types of modified DNA basescan be produced. cidence (Freedman et al., 1990). These two factors also have Of these products, relatively high levels of 5-hydroxymethyluracil have been found in human urine (reviewed in Mullaart et al.. 1990). 5-Hydroxymethyluracil is potentially a muta’ To whom correspondence should be addressed. Oxidative DNA Damage Levels in Rats Fed Low-Fat, HighFat, or Calorie-Restricted Diets. DJURIC, Z., Lu, M. H., LEWIS, S. M., LUONGO, D. A., CHEN, X. W., HEILBRUN, L. K., READING, B. A., DUFFY, P. H., AND HART, R. W. (1992). Toxicol. Appf. Pharmacol. 115, 156-l 60.
0041-008x/92
$5.00
Copyright rB 1992 by Academic Press. Inc. All rights of reproductmn in any form reserved.
156
DNA DAMAGE
genie lesion (Shirname-More et al., 1987). Unlike some base oxidation products, 5-hydroxymethyluracil appears to be relatively stable (Frenkel et al., 1986). In this study, we have investigated the effects of dietary fat and caloric restriction on the levels of 5-hydroxymethyluracil in the DNA from liver and mammary gland of Fischer 344 rats. MATERIALS
AND METHODS
Diets. The diets consisted of ingredients, including the vitamin and mineral mixes, that were obtained from Teklad Premier Laboratory Diets (a division of Harlan Sprague-Dawley, Inc., Madison, WI). The diets prepared were: ( I) 3% corn oil fed ad lib (low fat), (2) 5% corn oil fed ad lib (control), (3) 20%’ corn oil fed ad lib (high fat). and (4) 5% corn oil fed at 40% caloric restriction relative to the uu’ lib control diet (CR). The formulation of these diets is shown in Table I. All diets, with the exception of the high-fat diet. were prepared by blending the vitamin and mineral mixes with the dextrose in a Patterson-Kelley Vblender (HARSCO Corp.. East Stroudsburg, PA) for I5 min. The corn oil was hand-blended into the dry ingredients to obtain a thorough blend. The remaining ingredients were then added and the entire mixture was reblended in the V-blender for I5 min with the addition of an intensifier bar apparatus, This intensifier bar, with a peripheral speed of 1700 fpm. breaks soft agglomeratcs that can be found in raw materials or are formed during blending. The high-fat diet was prepared by dividing the total dextrose, corn oil, vitamin. and mineral mixes into three equal amounts, and each fraction was blended with the V-blender and by hand. as described above for the other three diets. By decreasing the amount of material being blended, a uniform blend with the required 20% corn oil was obtained. The fractions were subsequently mixed together and reblended for I5 min with the intensilier bar apparatus. The diets were stored refrigerated. Tests of homogeneity were performed on each diet by analyzing for zinc and calcium in samples of the blended diets. As the diets were removed from the V-blender, two 60-g samples were removed from three locations: the top. middle. and bottom layer of the V-blender. The total energy values given as kcal/g diet were calculated using the standard Atwater value estimates for physiological fuel values (Harper, 1979). Carbohydrates, protein, and cellulose were calculated at 4 kcal/g and fat at 9 kcal/g. Cellulose is often considered to be an inert filler: however, it can be fermented in the hindgut to a limited extent, ranging from I2 to 20% (VanSoest. 1982). In our calculation of the energy content of the diets. 12% digestibility of cellulose was used. Animals. Female. Fischer 344 rats were weaned at 2 I days of age and maintained on NIH-3 I diets. At 38 days of age. the rats were placed on the control diet (see Table I). Food intake was monitored such that the amount of food to be fed to the calorically restricted rats could be determined. At an age of 45 days. the animals were weighed and divided arbitrarily into the four diet groups with 12 animals per group. The animals were maintained on the diets for 2 weeks with daily food changes. The animals were housed singly in hanging, mesh-bottom cages fitted with trays to facilitate the collection of spilled food for the determination of total intake. Food intake, body weight, and the general health of the animals were monitored on a daily basis. The diets were presented to the animals in jars, and food intake was calculated from the amount of food left in the jar the next day plus the amount of food spilled. The lights in the animal room were kept on a schedule of 12 hr on and 12 hr off. The food was presented to the animals immediately before the lights were turned off. All animals thus consumed their diets during the same time period. After 2 weeks, the animals were euthanized by carbon dioxide inhalation and decapitation. The mammary gland tissue and livers were removed. Whole livers were frozen. The mammary gland epithelium was isolated from fresh
157
AND DIET
TABLE 1 Formulation of Diets Diet group Ingredient”
Control
High fat
Low fat
CRh
Corn oil Casein Dextrose Cellulose Vitamin mix (AIN76) Mineral mix (AlN76) DL-Methionine Choline bitartrate kcal/OO g diet
5 30 65 5 I
20 20 50 5 I
3 ‘0 67 5 I
x.33 33.3 41.8 x.33 1.67 6.16
3.7 0.3
3.7 0.3
3.7 0.3
0.2
0.2
0.2
387
462
377
0.5
0.2 379
’ Ingredients are given as percentage, by weight, in the diet. ’ The CR (calorically restricted) diet is formulated as indicated and fed at 40% restriction relative to the ad lib intake of the control diet.
mammary gland tissue by the method of Moon et al. (1969). Briefly. the tissue was minced and incubated with collagenase and the fat was separated from the epithelial cells by centrifugation. The epithelial cells were frozen until DNA extractions could be performed. DNA extraction. The mammary gland epithelium and liver tissue were homogenized in 1% sodium dodecyl sulfate and I mM EDTA and the DNA was extracted as previously described (Djuric E( a/.. 1988). The procedure involved incubation of the homogenates with protease followed by successive organic extractions with phenol. chloroform/isoamyl alcohol/phenol (24: 1:25) and chloroform/isoamyl alcohol (24: 1). treatment of the semipurified nucleic acid with RNases, and another series of the same organic extractions. The DNA was recovered by precipitation and quantified spectrophotometrically. Analysis spectrometry
of S-hydroxymethyluracil
levels by gas chromatography-mass
The purified DNA samples were hydrolyzed to (X-MS). nucleosides enzymatically. and trimethylsilyl derivatives were prepared for analysis by GC-MS as previously described (Djuric e( al.. 199 lb). The GCMS analyses were carried out with a Hewlett-Packard 597 I mass spectrometer operated in the EI mode. and positive ions were monitored. The HewlettPackard 5890 CC was operated in the splitless injection mode and injections were made with a Hewlett-Packard 7673 autosampler. Statistical methods. Dependent variables were analyzed to determine the effectof the four experimental diets on mean DNA damage level, nutrient intake, and animal weight. Nutrient intakes were calculated based on a simple mean food intahe over the I4 days of study. DNA damage level was assessed at Day I4 only. Due to the heavily nonnormal distribution of DNA damage level and nutrient intake, nonparametric methods of statistical analysis were used. For each dependent variable. separate analyses were performed using the k-sample Kruskal-Wallis rank-sum test to compare mean levels by diet group (Hollander and Wolfe, 1973). To perform the multiple comparisons of interest (all pairwise comparisons of diet groups), a nonparametric rank-sum method was used. based on a large sample approximation for balanced designs (Miller, 1981). This allowed us to maintain the experimentwise Type I error rate at 0.05 while performing all pairwise comparisons of diet groups (Miller. 1981). There were I3 rats for each diet group. yielding a completely balanced design. and there were no missing data. Rats were consecutively and arbitrarily assigned to diet groups. The data were viewed as resulting from separate ( 1 X 4) complete one-way layout experimental designs.
1.58
DJURIC ET AL
RESULTS
Diets and Food Intake The prepared diets were homogeneous as determined by the analyses of calcium and zinc in samples of the diets removed from the top, middle, and bottom of the V-blender. The values for percentage of calcium (by weight) in the control, high-fat, low-fat, and CR diets were 0.147 f 0.010, 0.137 * 0.004, 0.146 + 0.004, and 0.196 * 0.002, respectively, given as means + standard deviation. The values for zinc, in mean mg zinc/kg diet + standard deviation, were 44.5 ? 3.10, 40.2 t- 2.63, 40.2 -+_2.21, and 69.5 -+ 1.29. The maximum standard deviation in any individual diet preparation after blending was only 7% of the mean. The caloric content of each diet was calculated as described under Materials and Methods. In terms of kcal/g diet, all the diets were similar with the exception of the high-fat diet, which had a considerably higher energy content (Table 1). Mean food intake was significantly different by diet group (p < 0.0001) because the mean for the CR rats was significantly different from that of all other diets, especially the low-fat and control diets (Table 2). Food consumption in the high-fat group was, however, 12% lower than that in the control group. With regard to mean daily caloric intake, the significant difference between diet groups was due to the decreased caloric intake of the CR group.
The impact of reducing caloric intake was evident in the animal body weights. Initially the mean animal weight did not differ significantly by diet group. After 2 weeks on the diets, mean animal body weight did differ significantly by diet group, and this was entirely due to the CR mean being significantly lower than the other three means (Table 2). Daily fat intake differed significantly by diet group and in the expected fashion: the means for the high-fat and low-fat rats each differed significantly from that of the other three diet groups (Table 2). Mean fat intake of the high-fat group was three- to sixfold higher than that in any of the other diet groups. Conversely, the daily fat intake of the low-fat group was lower than that in any of the other groups. The daily fat intake of the CR and control groups was virtually the same. DNA Damage Levels As shown in Table 3, the mean liver DNA damage level differed significantly across the four diet groups (p = 0.002). This difference was entirely due to only two pairs of diet groups, as revealed by the rank-sum multiple-comparisons analysis. Mean DNA damage levels in the liver of CR rats differed significantly from that of control rats (p < 0.0 1) and that of rats in the high-fat group (p < 0.05). Mean 5-hydroxymethyluracil levels in the liver DNA of rats fed highfat and low-fat diets were decreased relative to control, but this was not significant.
TABLE 2 Mean Dietary Intakes and Body Weights by Diet Group Diet group” Variable Food intake (g/day)
Control 12.10 f 1.07
High fat
Low fat
10.67 f 0.79
11.84 f 0.54
d Caloric intake (kcal/day)
46.81 !I 4.13
49.27 k 3.64
44.62 f 2.04
d Fat intake (g/day)
0.61 + 0.05
2.13 + 0.16
0.36 kO.02 d
d e Initial body weight (g/animal) Final body weight (g/animal)
105.3 f 7.7 137.7 + 9.7
111.3 k5.9 146.9 k 9.2 c
p Values/
CR 7.14 + 0.07 l’ d 26.77 f
0.27
