Mutation Research, 250 (1991) 17-24 © 1991 Elsevier Science Publishers B.V. All rights reserved 002%5107/91/$03.50 ADONIS 0027510791001592
17
MUT 02519
Dietary modulation of DNA damage in human M i c h a e l G. Simic a n d D a v i d S. B e r g t o i d National Institute of Standards and Technology, Gaithersburg, MD 20899 (U.S.A.) (Accepted 5 April 1991)
Keywords: Aging; Antioxidants; Biomarkers; Caloric restriction; Diet; DNA damage; Hydroxyl radical; Oxidative processes; Prevention of degenerative diseases; Reactive oxygen species
Summary Manipulation of human diet can modulate urinary biomarkers of oxidative DNA base damage (UBODBD), reflecting changes in levels of DNA damage. When dietary composition is maintained but caloric intake is decreased (caloric restriction), UBODBD excretion is suppressed. At isocaloric dietary intake the level of damage depends on diet composition. For diets consisting of foods containing carbohydrates, proteins, and fats but lacking fruits and vegetables, the level of damage is higher than for diets including fruits and vegetables, which are rich in natural antioxidants. Assay of urinary biomarkers is suggested as a potential test for quantitative assessment of the carcinogenic or anticarcinogenic properties of foods, food components, and diets and for individual responses to nutritional regimens.
1. Introduction The basic conceptual foundation (including trans-species and high-to-low-dose extrapolation) of animal studies, mutation tests in bacteria, and other current practices for assessing the human carcinogenic potential of various agents are under scrutiny (Ames and Gold, 1990; Cohen and Ellwein, 1990). With the proliferation of dietary recommendations and speculative dietary regimens, the limitations of these methodologies for assessing the effects of foods, food components, and diets on cancer, degenerative diseases, aging, and life span are particularly evident. Similar
Correspondence: Dr. Michael G. Simic, National Institute of Standards and Technology, Gaithersburg, MD 20899 (U.S.A.).
problems are encountered in toxicological studies of exogenous agents. Despite dramatic achievements in extending average life expectancy, reaching as much as 80 years of age in some countries such as Japan and Iceland, maximum life span (MLS) apparently has not changed since ancient times (Weindruch and Walford, 1988). In the 1930s and 1940s McCay and coworkers (McCay et al., 1935, 1943), in a series of pioneering nutritional experiments, succeeded in extending MLS of mice and rats by about one third. More recently other investigators (Masoro et al., 1982; Weindruch and Walford, 1988) have confirmed and extended the original findings, leading to renewed interest in the phenomenon. These achievements were accomplished by giving the short-lived animals calorically restricted but 'nutritionally adequate'
18 diets. A dramatic decrease in cancer and degenerative diseases paralleled the success in extending the life span of the animals. Similar studies of cancer reduction or MLS extension in human subjects, however, would translate into experiments lasting from several decades to more than a century. For obvious reasons such studies are neither practical nor feasible.
2. Biomarkers of DNA damage Replacing long-term end points of animal studies by short-term, direct measurement of relevant biomarkers in humans, when possible, would greatly reduce the need for interspecies extrapolation, with its inherent uncertainties.
2.1. Rate of damage vs. cumulatit,e end points To bypass the constraints of classical bioassays and tests, we addressed the role of diet in extension of human maximum life span (MLS) and its corollary, carcinogenesis, through indirect measurement of genetic damage. The levels of urinary biomarkers of oxidative DNA base damage ( U B O D B D ) in humans were correlated with dietary caloric intake and type of food, and were found to decrease with reduced caloric intake and with increased consumption of fruits and vegetables. Our approach is based on short-term measurement of intensity factors (rate of DNA damage) rather than observing long-term capacity factors such as biological end points, e.g., MLS and onset of cancer. Our aim was to avoid present practices for studying cumulative effects, which are not only long term but also costly, paralyzing our ability to quantitatively assess foods and diets. 2.2. Principles of biomarker selection Urinary biomarkers of oxidative DNA-base damage were introduced recently by Ames and coworkers (Cathcart et al., 1984) as a novel approach to noninvasive assessment of genetic damage. Of all the possible products of oxidative damage, we have concentrated our investigations on thymidine glycol (dRTg) and 8-hydroxydeoxyguanosine (8-dRG-OH) for the following rea-
sons. First, these products are generated in cells in the reaction of OH radicals with DNA, i.e., nuclear thymine (T) and guanine (G), respectively, and the mechanisms are well understood (von Sonntag, 1987). Second, a direct quantitative correlation to radiation-generated free radicals in model systems and in vivo has been demonstrated (Bergtold and Simic, 1991; Simic eI al., 1989). Third, dRTg and 8-dRG-OH, in contrast to Tg and 8-G-OH, are not absorbed through the digestive tract (Cathcart ct al., 1984), thereby eliminating the confounding variables associated with direct dietary sources of the compounds or the biochemical reactions of the digestive system. Fourth, no enzymatic processes (except repair enzymes) are known that generate dRTg or 8dRG-OH, making these products specific for monitoring OH radicals in viw).
2.3. Measurement of biomarkers * Biomarkers of oxidative DNA-base damage were measured using gas chromatography-mass s p e c t r o m e t r y with s e l e c t - i o n m o n i t o r i n g ( G S / M S - S I M ) after pre-purification and derivatization of samples as described previously (Dizdaroglu and Bergtold, 1986; Bergtold et al., 1988). Each value represents the average of four injections. Quantitation of products was achieved using deuterated standards (Bergtold et al., 1990). Daily dietary caloric intake (DCI) was calculated from the weight of the foods eaten and their caloric values (USDA Handbook 8). The measurements in caloric restriction experiments (high and low DCI) were made from urine samples taken after 10 days of constant DCI. In the experiments in which the type of food was altered, the measurements were taken after 7 days of equilibration.
* Certain commercial equipment, instruments, and materials are identified in this work in order to specifyadequately lhe experimental procedure. Such identification does not imply recommendation nor endorsement by the National Institute of Standards and Technology, nor does it imply that the material or equipment identified is necessarily the best available for these purposes.
19
3. Effect of diet on oxidative biomarkers
Demonstrating an effect of dietary caloric intake or food composition on oxidative damage of DNA (i.e., the corresponding biomarkers) would implicate the basic source of biochemical energy as a major factor in causation of degenerative diseases and would elucidate mechanisms for their prevention.
3.1. Caloric restriction The effects of dietary caloric intake on the levels of UBODBD are shown in Table 1. The measurements were conducted from one male individual in his 50s, subject A, over a period of one year under strictly controlled conditions (level of exercise, type of diet, general health), with four H i / L o (high calorie/low calorie) dietary cycles. As a consequence, surprising reproducibility of UBODBD levels was achieved within Hi and Lo caloric periods of different cycles. Invariably in all low-calorie intake periods, lower levels of UBODBD were observed. These observations were confirmed with female subject B, in her 30s. Here again Hi caloric intake resulted in higher levels of UBODBD than Lo caloric intake, although the absolute values at Hi caloric diets were higher than those of subject A (Table 1). The difference in UBODBD levels between individuals may be attributed to different diets and to individual responses.
3.2. Type of food To test whether energy input is the sole determinant of DNA damage, isocaloric feeding experiments were conducted in which subject A ingested different types of food (Fig. 1). On highcalorie diets consisting of foods completely lacking fruits and vegetables, i.e., plant antioxidants (Hi(-)), the UBODBD levels were about three times higher than on a diet rich in fruits and vegetables (Hi(+), Fig. 1). The average levels of UBODBD (Table 1) for Lo and Hi caloric diets with an average fruit and vegetable content are also included in Fig. 1 for comparison. Similar observations were made for female subject B (Simic and Bergtold, 1991).
3.3. Individual vs. population responses There are many advantages of using individuals who serve as their own controls for the study of dietary effects, provided diverse variables are carefully controlled. Physiological and genetic differences among individuals are eliminated. In a heterogeneous population (n = 20), the 'normal' levels of UBODBD may vary greatly; the levels of dRTg, for example, vary from 0.1 to 0.7 nmol kg-~ day-~. Because inter-individual variation in UBODBD levels is great, averaging the values for many individuals may obscure significant trends in UBODBD levels observed for each individual, yielding meaningless results. It would be interesting to compare individual values in different subpopulations and average values between the sub-
TABLE 1 URINARY BIOMARKERS OF OXIDATIVE DNA BASE DAMAGE (UBODBD) IN (nmoles kg- t day- l) FOR SUBJECT A (MALE, 50s) AS A FUNCTION OF LOW AND HIGH DIETARY CALORIC INTAKE. MEASUREMENTS WERE TAKEN AFTER 10 DAYS OF CONSTANT DAILY CALORIC INTAKE Date (1989/90)
May 5
July 18
July 27
Aug. 19
Aug. 30
Jan. 8
Jan. 18
Caloric intake (kcal/day)
2,000
2,100
1,200
2,000
1,200
2,200
1,100
180
180
175
180
173
181
173
Body weight (Ib) dRTg
0.26
0.27
0.11
0.19
0.10
0.28
0.11
8-d RG -OH
0.35
0.40
0.08
0.30
0.16
0.33
0.09
20 P
.
. . . . . . . . . . . . . .
.
.
.
-10.6
3000~g
"i O
O
2000
~0.4 E
: ¸''I¸¸
J
,,,."
ot,,,, 1
r~ ..
Hi
Lo
'
17
MUT 02519
Dietary modulation of DNA damage in human M i c h a e l G. Simic a n d D a v i d S. B e r g t o i d National Institute of Standards and Technology, Gaithersburg, MD 20899 (U.S.A.) (Accepted 5 April 1991)
Keywords: Aging; Antioxidants; Biomarkers; Caloric restriction; Diet; DNA damage; Hydroxyl radical; Oxidative processes; Prevention of degenerative diseases; Reactive oxygen species
Summary Manipulation of human diet can modulate urinary biomarkers of oxidative DNA base damage (UBODBD), reflecting changes in levels of DNA damage. When dietary composition is maintained but caloric intake is decreased (caloric restriction), UBODBD excretion is suppressed. At isocaloric dietary intake the level of damage depends on diet composition. For diets consisting of foods containing carbohydrates, proteins, and fats but lacking fruits and vegetables, the level of damage is higher than for diets including fruits and vegetables, which are rich in natural antioxidants. Assay of urinary biomarkers is suggested as a potential test for quantitative assessment of the carcinogenic or anticarcinogenic properties of foods, food components, and diets and for individual responses to nutritional regimens.
1. Introduction The basic conceptual foundation (including trans-species and high-to-low-dose extrapolation) of animal studies, mutation tests in bacteria, and other current practices for assessing the human carcinogenic potential of various agents are under scrutiny (Ames and Gold, 1990; Cohen and Ellwein, 1990). With the proliferation of dietary recommendations and speculative dietary regimens, the limitations of these methodologies for assessing the effects of foods, food components, and diets on cancer, degenerative diseases, aging, and life span are particularly evident. Similar
Correspondence: Dr. Michael G. Simic, National Institute of Standards and Technology, Gaithersburg, MD 20899 (U.S.A.).
problems are encountered in toxicological studies of exogenous agents. Despite dramatic achievements in extending average life expectancy, reaching as much as 80 years of age in some countries such as Japan and Iceland, maximum life span (MLS) apparently has not changed since ancient times (Weindruch and Walford, 1988). In the 1930s and 1940s McCay and coworkers (McCay et al., 1935, 1943), in a series of pioneering nutritional experiments, succeeded in extending MLS of mice and rats by about one third. More recently other investigators (Masoro et al., 1982; Weindruch and Walford, 1988) have confirmed and extended the original findings, leading to renewed interest in the phenomenon. These achievements were accomplished by giving the short-lived animals calorically restricted but 'nutritionally adequate'
18 diets. A dramatic decrease in cancer and degenerative diseases paralleled the success in extending the life span of the animals. Similar studies of cancer reduction or MLS extension in human subjects, however, would translate into experiments lasting from several decades to more than a century. For obvious reasons such studies are neither practical nor feasible.
2. Biomarkers of DNA damage Replacing long-term end points of animal studies by short-term, direct measurement of relevant biomarkers in humans, when possible, would greatly reduce the need for interspecies extrapolation, with its inherent uncertainties.
2.1. Rate of damage vs. cumulatit,e end points To bypass the constraints of classical bioassays and tests, we addressed the role of diet in extension of human maximum life span (MLS) and its corollary, carcinogenesis, through indirect measurement of genetic damage. The levels of urinary biomarkers of oxidative DNA base damage ( U B O D B D ) in humans were correlated with dietary caloric intake and type of food, and were found to decrease with reduced caloric intake and with increased consumption of fruits and vegetables. Our approach is based on short-term measurement of intensity factors (rate of DNA damage) rather than observing long-term capacity factors such as biological end points, e.g., MLS and onset of cancer. Our aim was to avoid present practices for studying cumulative effects, which are not only long term but also costly, paralyzing our ability to quantitatively assess foods and diets. 2.2. Principles of biomarker selection Urinary biomarkers of oxidative DNA-base damage were introduced recently by Ames and coworkers (Cathcart et al., 1984) as a novel approach to noninvasive assessment of genetic damage. Of all the possible products of oxidative damage, we have concentrated our investigations on thymidine glycol (dRTg) and 8-hydroxydeoxyguanosine (8-dRG-OH) for the following rea-
sons. First, these products are generated in cells in the reaction of OH radicals with DNA, i.e., nuclear thymine (T) and guanine (G), respectively, and the mechanisms are well understood (von Sonntag, 1987). Second, a direct quantitative correlation to radiation-generated free radicals in model systems and in vivo has been demonstrated (Bergtold and Simic, 1991; Simic eI al., 1989). Third, dRTg and 8-dRG-OH, in contrast to Tg and 8-G-OH, are not absorbed through the digestive tract (Cathcart ct al., 1984), thereby eliminating the confounding variables associated with direct dietary sources of the compounds or the biochemical reactions of the digestive system. Fourth, no enzymatic processes (except repair enzymes) are known that generate dRTg or 8dRG-OH, making these products specific for monitoring OH radicals in viw).
2.3. Measurement of biomarkers * Biomarkers of oxidative DNA-base damage were measured using gas chromatography-mass s p e c t r o m e t r y with s e l e c t - i o n m o n i t o r i n g ( G S / M S - S I M ) after pre-purification and derivatization of samples as described previously (Dizdaroglu and Bergtold, 1986; Bergtold et al., 1988). Each value represents the average of four injections. Quantitation of products was achieved using deuterated standards (Bergtold et al., 1990). Daily dietary caloric intake (DCI) was calculated from the weight of the foods eaten and their caloric values (USDA Handbook 8). The measurements in caloric restriction experiments (high and low DCI) were made from urine samples taken after 10 days of constant DCI. In the experiments in which the type of food was altered, the measurements were taken after 7 days of equilibration.
* Certain commercial equipment, instruments, and materials are identified in this work in order to specifyadequately lhe experimental procedure. Such identification does not imply recommendation nor endorsement by the National Institute of Standards and Technology, nor does it imply that the material or equipment identified is necessarily the best available for these purposes.
19
3. Effect of diet on oxidative biomarkers
Demonstrating an effect of dietary caloric intake or food composition on oxidative damage of DNA (i.e., the corresponding biomarkers) would implicate the basic source of biochemical energy as a major factor in causation of degenerative diseases and would elucidate mechanisms for their prevention.
3.1. Caloric restriction The effects of dietary caloric intake on the levels of UBODBD are shown in Table 1. The measurements were conducted from one male individual in his 50s, subject A, over a period of one year under strictly controlled conditions (level of exercise, type of diet, general health), with four H i / L o (high calorie/low calorie) dietary cycles. As a consequence, surprising reproducibility of UBODBD levels was achieved within Hi and Lo caloric periods of different cycles. Invariably in all low-calorie intake periods, lower levels of UBODBD were observed. These observations were confirmed with female subject B, in her 30s. Here again Hi caloric intake resulted in higher levels of UBODBD than Lo caloric intake, although the absolute values at Hi caloric diets were higher than those of subject A (Table 1). The difference in UBODBD levels between individuals may be attributed to different diets and to individual responses.
3.2. Type of food To test whether energy input is the sole determinant of DNA damage, isocaloric feeding experiments were conducted in which subject A ingested different types of food (Fig. 1). On highcalorie diets consisting of foods completely lacking fruits and vegetables, i.e., plant antioxidants (Hi(-)), the UBODBD levels were about three times higher than on a diet rich in fruits and vegetables (Hi(+), Fig. 1). The average levels of UBODBD (Table 1) for Lo and Hi caloric diets with an average fruit and vegetable content are also included in Fig. 1 for comparison. Similar observations were made for female subject B (Simic and Bergtold, 1991).
3.3. Individual vs. population responses There are many advantages of using individuals who serve as their own controls for the study of dietary effects, provided diverse variables are carefully controlled. Physiological and genetic differences among individuals are eliminated. In a heterogeneous population (n = 20), the 'normal' levels of UBODBD may vary greatly; the levels of dRTg, for example, vary from 0.1 to 0.7 nmol kg-~ day-~. Because inter-individual variation in UBODBD levels is great, averaging the values for many individuals may obscure significant trends in UBODBD levels observed for each individual, yielding meaningless results. It would be interesting to compare individual values in different subpopulations and average values between the sub-
TABLE 1 URINARY BIOMARKERS OF OXIDATIVE DNA BASE DAMAGE (UBODBD) IN (nmoles kg- t day- l) FOR SUBJECT A (MALE, 50s) AS A FUNCTION OF LOW AND HIGH DIETARY CALORIC INTAKE. MEASUREMENTS WERE TAKEN AFTER 10 DAYS OF CONSTANT DAILY CALORIC INTAKE Date (1989/90)
May 5
July 18
July 27
Aug. 19
Aug. 30
Jan. 8
Jan. 18
Caloric intake (kcal/day)
2,000
2,100
1,200
2,000
1,200
2,200
1,100
180
180
175
180
173
181
173
Body weight (Ib) dRTg
0.26
0.27
0.11
0.19
0.10
0.28
0.11
8-d RG -OH
0.35
0.40
0.08
0.30
0.16
0.33
0.09
20 P
.
. . . . . . . . . . . . . .
.
.
.
-10.6
3000~g
"i O
O
2000
~0.4 E
: ¸''I¸¸
J
,,,."
ot,,,, 1
r~ ..
Hi
Lo
'
