![[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}
|
|
|
|
|
||
The Journal of Nutrition Vol. 128 No. 7 July 1998,
pp. 1204-1212
, and
Western Human Nutrition Research Center, U.S. Department of Agriculture, Agricultural Research Service, Presidio of San Francisco, CA 94129; * National Center for Toxicological Research, FDA, Division of Nutritional Toxicology, Jefferson, AR 72079; and School of Public Health, University of California, Los Angeles, CA 90024
 |
ABSTRACT |
|---|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
To determine the human folate requirement on the basis of changes in biochemical pathways, we studied the effect of controlled folate intakes on plasma homocysteine and lymphocyte DNA methylation and deoxynucleotide content in healthy postmenopausal women. Eight women (49-63 y of age) were housed in a metabolic unit and fed a low folate diet containing 56 µg/d of folate for 91 d. Folate intake was varied by supplementing 55-460 µg/d of folic acid (pteroylglutamic acid) to the diet to provide total folate intake periods of 5 wk at 56 µg/d, 4 wk at 111 µg/d and 3 wk at 286-516 µg/d. A subclinical folate deficiency with decreased plasma folate was created during the first two periods. This resulted in significantly elevated plasma homocysteine and urinary malondialdehyde, and lymphocyte DNA hypomethylation. The folate depletion also resulted in an increased ratio of dUTP/dTTP in mitogen-stimulated lymphocyte DNA and decreased lymphocyte NAD, changes suggesting misincorporation of uracil into DNA and increased DNA repair activity. The DNA hypomethylation was reversed with 286-516 µg/d of folate repletion, whereas the elevated homocysteine decreased with 516 but not 286 µg/d of folate. The results indicate that marginal folate deficiency may alter DNA composition and that the current RDA of 180 µg/d may not be sufficient to maintain low plasma homocysteine concentrations of some postmenopausal women.
KEY WORDS: folic acid · homocysteine · methylation · DNA · humans
Anemia and other hematologic measures have traditionally been the criteria for gauging the adequacy of folate nutriture. Recent data, however, suggest that the elevation of plasma homocysteine (Hcy)6 and disturbances of DNA composition may provide more sensitive functional measures of folate deficiency (Blount et al. 1997 Recent studies have linked moderately elevated plasma Hcy (>14 µmol/L) to increased risk of coronary, cerebral and peripheral vascular diseases (Boushey et al. 1995 Studies with cells, as well as in vivo studies with animal models, have shown that folate deficiency disrupts the normal biosynthesis of deoxyribonucleotides, which are utilized for DNA replication and repair. Spleen cells from rats fed diets low in folate, choline and/or methionine had decreased amounts of dTMP and dTTP, consistent with impaired folate-dependent conversion of uridylate to thymidylate (James et al. 1992). Mitogen-stimulated rat lymphocytes cultured in a low folate medium showed significant decreases in dTMP, dGTP, dATP and dCTP (James et al. 1993 These new data suggest that the human requirement for folate should be re-evaluated by assessing molecular as well as clinical endpoints of folate deficiency, including plasma Hcy and DNA methylation and deoxynucleotide content. The objectives of this study were to investigate these variables in healthy postmenopausal women during controlled folate intakes and to determine the amount of dietary folate needed to maintain their integrity.
Protocol and subjects.
Ten healthy non-smoking postmenopausal volunteer women, ages 49-63 y, were admitted to the metabolic unit of the USDA Western Human Nutrition Research Center (WHNRC) after medical and psychological screening. Prestudy screening included medical histories, physical and dental examinations, hematologic and clinical chemistry tests, psychological testing, resting electrocardiogram, and tests for hepatitis, syphilis, tuberculosis, human papilloma virus (PAP smear) and HIV-antibody tests. Tests for plasma folate and vitamin B-12, alcohol, tobacco and drug use were also performed. All subjects were within 90-130% of desirable weight (Metropolitan Life Insurance Company, New York, NY), except for one subject who was 145% of desirable weight. Other specific criteria for subject selection included the following: non-smokers, non-users of vitamin or health food supplements containing folic acid, hemoglobin >115 g/L and hematocrit >0.34. Three subjects, who were receiving estrogen replacement therapy when they entered the study, continued the same regimen throughout.
Experimental design and diet.
Subjects consumed the same experimental low folate diet in a 4-d menu rotation for the entire 91 d. The diet was also designed to limit choline intake as an exogenous methyl group source. The diet provided an average of 56 µg/d of folate; varying amounts of synthetic folic acid (pteroylglutamic acid) were added to the diet to provide folate intake periods of 56 to 516 µg/d as shown in Table 1, and Figures 1 and 2. Over the first 5 d, a total of 195 µg/d of folate (56 µg/d from the diet plus 139 µg/d of added folic acid) were provided to maintain adequate folate status while baseline measurements were taken. The succeeding 5-wk folate depletion period (d 6-41) was designed to reduce body folate stores and produce a moderate but not severe folate deficiency, analogous to free-living women with chronically low intake and body stores of folate but without macrocytosis or anemia.
Specimen collections and analytical methods.
A fasting blood sample was taken at 0700-0800 h by venipuncture 13 times for study-related biochemical determinations, and for complete blood count and clinical chemistry panels to monitor the subjects health (not all tests were performed 13 times). The blood was collected in evacuated glass tubes and immediately processed for serum, EDTA, heparin and citrate plasma. Plasma folate and vitamin B-12, and erythrocyte folate (from EDTA anticoagulated blood) were determined by competitive protein binding radioassay kits (Quantaphase II B-12/Folate Radioassay, BioRad, Hercules CA). The samples were prepared according to kit directions and stored at Peripheral blood mononuclear cell isolation and analyses.
The isolation of mononuclear cells (primarily T and B lymphocytes) was based on the method of Boyum (1968) Statistics.
Results for the plasma, lymphocyte and urine variables were analyzed for differences due to dietary folate intake. Descriptive statistics were computed for the end of each folate intake period and are shown as means ± SEM in Table 1. Correlations of study variables with folate intake or plasma folate within individuals were examined by calculation of Pearson product-moment coefficients, or Spearman coefficients if data were not normally distributed. An ANOVA model with repeated measures was performed to test for any effects of dietary folate intake on plasma folate, homocysteine and DNA methyl acceptance. If significant relationships were found, paired t statistics or the Wilcoxon signed-rank statistic (data not normally distributed) was used to determine differences between baseline, folate repletion and depletion periods (Glantz 1992 Of the total group of 10 subjects, data from two subjects were excluded from the analyses because of abnormal values related to the Hcy pathway. One subject began the study with an abnormally high plasma Hcy value, 19 µmol/L at d 5 compared with 8-11 µmol/L for the rest of the group (2.6 SD above the mean of the entire group of 10). The second excluded subject developed a mild anemia of unknown origin at d 56 and was then supplemented with 325 mg/d of ferrous sulfate daily for the remainder of the study. After this, the subject developed greatly elevated plasma Hcy levels, i.e., 31 µmol/L at d 84 vs. 10-15 µmol/L for the rest of the group.
Blood folate and vitamin B-12.
Repeated-measures ANOVA using values at the end of folate intake periods showed a strong relation between plasma folate concentrations and dietary folate intakes (P < 0.001). Plasma folate decreased significantly during the 5-wk depletion period when subjects consumed the low folate diet (56 µg/d) with no folic acid supplementation (Fig. 1 and Table 1). After the 5-wk folate depletion period, a total folate intake of 111 µg/d was not sufficient to raise plasma folate above the low levels reached during depletion (at d 63, five of the eight subjects had plasma folate values below normal, <6.8 nmol/L). Subsequent intake of 286 µg/d was sufficient to raise plasma folate concentrations in all subjects. Plasma folate continued to rise with folate intake of 516 µg/d, and the group mean was not significantly different than baseline at the end of the study, d 92.
Plasma homocysteine.
All subjects started in the normal range and increased above baseline by d 56; at that point, five of the eight subjects were above the upper limit of normal for this population, 12 µmol/L (Rasmussen et al. 1996 Hematologic measures.
As a group, hematologic indices of the subjects did not change significantly throughout the study. From beginning to end, hemoglobin decreased slightly, from 129 ± 3 to 123 ± 2 g/L [normal = 117-160 g/L (Tietz et al. 1995)], hematocrit from 0.37 ± 0.01 to 0.36 ± 0.01 volume fraction (0.35-0.47 volume fraction), and MCV was unchanged at 89 ± 2 fL (81-101 fL). Individual values for these hematologic measures remained within or close to the normal range throughout. The nuclear lobe average of neutrophils increased significantly during folate depletion from d 6 to 42 (2.13 ± 0.09 to 2.37 ± 0.09 lobes/cell) and remained unchanged to the end of the study (2.44 ± 0.08 lobes/cell). All individual lobe averages were < 3, well below the value of 3.5 indicating folate deficiency (Herbert and Das 1994 Lymphocyte DNA methylation and deoxynucleotide concentrations.
The group means for the four time points at which the lymphocyte DNA methyl acceptance assay was performed are shown in Figure 1 and Table 1. The DNA methyl acceptance shown is inversely related to the degree of DNA methylation. Genome-wide DNA hypomethylation was inversely related to dietary folate (ANOVA, P < 0.001) and also inversely related to plasma folate as seen in Figure 1. All subjects except one (whose DNA methyl uptake did not change throughout) showed DNA methyl acceptance peak values at d 69 after a folate intake of 111 µg/d, and subsequent decreases with folate repletion of 286 to 516 µg/d.
MDA and creatinine measures.
Group means for urinary MDA determined in urine pools collected for 7 d are given in Table 1. Values for all subjects increased at d 67 (midpoint of 7-d pool) compared with baseline and decreased from d 67 to 88 after folate repletion. All subjects' values peaked at d 67, similar to the DNA methyl acceptance pattern.
Folate status and requirement to maintain plasma folate concentrations.
During 5 wk of folate intake at 56 µg/d followed by 4 wk at 111 µg/d, the subjects attained a state of moderate folate depletion characterized by low plasma folate and elevated Hcy. However, because red cell folate concentrations, hematologic indices and neutrophil segmentation remained normal throughout, no appreciable tissue folate depletion or folate-deficient hematopoiesis occurred. Although plasma folate and Hcy respond to changes in folate intake within weeks, tissue folate pools would not have reached steady state during the relatively short dietary treatment periods of this study (Stites et al. 1997 Folate requirement to maintain plasma homocysteine concentrations.
The observed inverse relation between plasma Hcy and folate intake in healthy postmenopausal women is consistent with similar relationships reported for healthy premenopausal women, adult men and the elderly (Jacob et al. 1994 DNA methylation and deoxynucleotide content.
The increased DNA methyl acceptance seen at d 69 (Fig. 1, Table 1), a point of low plasma folate and elevated Hcy, indicates that folate depletion resulted in genome-wide DNA hypomethylation, which was reversible within 3 wk by folate repletion at 286-516 µg/d. Despite the well-known role of folate as a supplier of one-carbon units for in vivo methylation reactions, no previous studies have shown that folate intake affects DNA methylation in humans. Our finding of DNA hypomethylation is similar to that of Pogribny et al. (1995) Folate status and lipid peroxidation.
The observed inverse relation of urinary MDA equivalents, a lipid peroxidation measure, to folate intake is very similar to the pattern of plasma Hcy and lymphocyte DNA methyl uptake (Table 1). Increased lipid peroxidation during folate deficiency may be a consequence of increased circulating Hcy, which has been postulated to act as a pro-oxidant. Olszewski and McCully (1993) suggested that high levels of Hcy may promote oxidative damage because the sulfhydryl group of Hcy acts catalytically with ferric or cupric ions to generate hydrogen peroxide, oxygen radicals and homocysteinyl radicals. Sparrow and Olszewski (1993)
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
, Jacob et al. 1994, James et al. 1997, O'Keefe et al. 1995, Pogribny et al. 1995). These new data raise the question of whether the current Recommended Dietary Allowance (RDA) for folate is adequate to satisfy pathways of Hcy metabolism and deoxynucleotide synthesis. Folate is needed to provide one-carbon units for over one hundred biochemical processes, including the methylation of Hcy to form methionine and the biosynthesis of deoxynucleotides dTMP, dAMP and dGMP needed for DNA replication (Selhub and Rosenberg 1996). Increased Hcy has been associated with premature vascular disease (Boushey et al. 1995), and disturbed deoxynucleotide synthesis has been associated with aberrant DNA synthesis and cell proliferation (James et al. 1993 and 1994b).
). Even moderate folate deficiency has been linked to increased plasma Hcy concentrations in both women and men. Healthy non-pregnant women given a diet containing 200 µg/d of folate (slightly more than the current RDA of 180 µg/d) had significantly higher plasma Hcy (12.6 ± 3.7 µmol/L) than women given the same diet with 400 µg/d of folate intake (7.7 ± 1.6 µmol/L) (O'Keefe et al. 1995). In healthy adult men, 84% of the current RDA did not normalize plasma Hcy concentrations elevated by experimental folate depletion, suggesting that the current folate RDA for adult men may not provide the expected margin of protection (Jacob et al. 1994).
). Additionally, Pogribny et al. (1995) reported that folate/methyl deficiency in rats induced DNA strand breaks and hypomethylation within the hepatic p53 tumor suppressor gene. Similar results were reported by Kim et al. (1997) with isolated folate deficiency in rats. Consistent with the above results, recent studies in humans indicated that folate-deficient individuals had elevated incorporation of uracil into DNA, accompanied by an increased frequency of cellular micronuclei, a measure of DNA and chromosome damage (Blount et al. 1997, MacGregor et al. 1997).
SUBJECTS AND METHODS
Abstract
Introduction
Methods
Results
Discussion
References
View this table:
Table 1.
Plasma folate and biochemical indices in postmenopausal women with various dietary folate intakes1
View larger version (29K):
[in a new window]
Fig 1.
Plasma folate and homocysteine (lower panel), and lymphocyte DNA methyl acceptance (upper panel) for eight postmenopausal women receiving various dietary intakes of folate shown at bottom in µg/d (folate intake values are 96 µg/d higher if results from the tri-enzyme food folate assay procedure are used, Tamura et al. 1997 ). Plotted values are means ± SEM. Means at end of folate intake periods are different: *from d 5 or 
from d 84 (d 69 for methyl acceptance test), P < 0.05 by paired t test. Lower limit of normal range for plasma folate is shown at bottom as horizontal dashed line at 6.8 nmol/L.
View larger version (23K):
[in a new window]
Fig 2.
Red cell folate concentrations of total, young and old red cell populations of eight postmenopausal women receiving various dietary intakes of folate shown at bottom in µg/d (folate intake values are 96 µg/d higher if results from the tri-enzyme food folate assay procedure are used, Tamura et al. 1997 ). Plotted values are means + SEM. Lower limit of normal range for red cell folate is shown as horizontal dashed line at 317 nmol/L. All means for young red cells are greater than old, P < 0.05. Means at end of folate intake periods are different: *from d 6 or from d 42, P < 0.05 by paired t test.
View this table:
Table 2.
Low folate diet menus1
). Concentrations of folic acid determined in this way were within 6% of the values calculated gravimetrically. Weekly spectrophotometric checks of the refrigerated folic acid supplement solutions showed no deterioration of folic acid content over 5 wk.
) showed that this procedure reduced their folate content by ~50%. Approximately 5 g/d of a nondigestible fiber supplement (Alphacel, ICN Biomedicals, Aurora OH) was added to the diet by mixing with the applesauce at 2 g/100 g.
70°C until thawed for determination of total folate by treatment with folate conjugase and microbiological assay with L. casei as the assay organism (Tamura et al. 1997). The mean folate content of the four daily menus was 56 µg/d (range 39-71). This compared with a mean value of 31 µg/d (range 28-34) calculated from food composition tables (USDA 1991). The calculated value may be lower because folate values for 13 of the 40 food items were missing from the nutrient database including values for the low protein and low carbohydrate products, spices and the fiber supplement. Although the treatment with folate conjugase alone is the traditional technique for food folate analysis, use of the newer tri-enzyme technique (treatment of the homogenate with 
-amylase, protease and folate conjugase) resulted in a mean daily intake value of 152 µg/d, nearly triple that when folate conjugase alone was used (Tamura et al. 1997). To interpret the results from this study, we used the value of 56 µg/d derived from the traditional assay method with the use of folate conjugase alone because this allows comparability to previous studies (Jacob et al. 1994, O'Keefe et al. 1995, Sauberlich et al. 1987) and data upon which the current RDA is based. Use of the higher tri-enzyme value of 152 µg/d would increase the folate intake of each period by 96 µg/d, as noted in footnote 2 of Table 1 and the figure legends.
). The choline content of the four daily menus averaged 147 mg/d (range 132-171 mg/d). The diet was supplemented daily with three different vitamin/mineral tablets to provide at least 80% of the RDA for each essential micronutrient (except 75% of vitamin D RDA). The subjects received one folate-free vitamin/mineral tablet daily with breakfast (Fosfree, Mission Pharmacal, San Antonio TX), one multimineral tablet at each of the three daily meals (Chelated Solamins, Solgar Vitamins, Lynbrook NY) and one calcium/magnesium tablet at each dinner (Bronson Calcium Complex & Magnesium, Bronson, St. Louis MO). This regimen provided the following amounts of total daily B vitamins (diet plus supplements), in % of RDA (NRC 1989): 234% of B-6, 124% of B-12, 200% of riboflavin, 556% of thiamin and 214% of niacin (21.1 niacin equivalents from preformed niacin and 8.6 niacin equivalents from 516 mg of tryptophan).
70°C until analysis.
, which is based on the lower density of young compared with older red cells. After hematocrit was determined and the red cell fractions separated, 100 µL aliquots from total, young and old red cells plus plasma were added to 1 mL of 22.7 mmol/L ascorbic acid, mixed well and frozen at 70°C for folate determination. Upon thawing, the whole-blood hemolysates, which included the added back plasma, were incubated for 90 min before radioassay to hydrolyze polyglutamyl folate in red cells. Erythrocyte folate was calculated as the difference between plasma and whole-blood folate using the hematocrit to correct for red cell volume differences. Determination of folate in six plasma and whole blood samples on three consecutive days gave a mean day-to-day variance (CV%) of 4.0 and 6.7% for plasma and red cell folate assays, respectively.
). Malondialdehyde (MDA) was determined in EDTA plasma by a modified procedure from Chirico (1994) involving HPLC separation and fluorescence detection of the MDA-thiobarbituric acid (TBA) adduct, MDA[TBA]2. The results were calculated from a calibration curve based on MDA generated in vitro by hydrolysis of 1,1,3,3-tetramethoxypropane (TMP) and were expressed as micromolar MDA equivalents. Day-to-day variation of frozen aliquots of a plasma control over 10 runs was 15% (0.242 ± 0.036 µmol/L MDA equivalents).
).
, which employs TBA derivatization and fluorometric detection at 525 nm excitation/547 nm emission. Results were calculated from the average slope of calibration curves prepared from hydrolysis of TMP over 26 assay runs. Day-to-day variation of frozen urine aliquots at low and high concentrations over 26 assays was 10.9 and 5.1%, respectively (within-run variations were 3% or less).
). Complete blood counts including differential white cell counts were taken weekly with an automated cell counter, which calculates the hematocrit from the red cell count and mean corpuscular volume (MCV), and determines hemoglobin by the cyanmethemoglobin reaction (System 9000 Diff, Serono Baker Diagnostics, Allentown PA).
. Histopaque-1077 (12 mL) (Sigma Chemical, St. Louis MO) was layered into the bottom of tubes containing 8 mL of heparinized blood and 20 mL PBS. After centrifugation, removal of plasma and washing with PBS, the isolated lymphocytes were then processed for DNA methylation, strand breaks, deoxynucleotides and tritiated deoxyuridine uptake tests. For DNA methylation and strand break tests, the PBS supernatant was removed after centrifugation and the cell pellets stored at 70°C until DNA extraction (Ausebel et al. 1989).
70°C for later determination of deoxynucleotides.
). For the tritiated deoxyuridine uptake test, the cells were added at a concentration of 105/well in six replicate wells containing 100 µL media. The media contained folic acid (1 g/L) for the control wells and folate positive wells. To the folate positive and negative, but not the control wells, PHA-P was added to a final concentration of 10 µg/mL for mitogen stimulation. The cells were incubated at 37°C for 48 h; then 50 µL [6-3H]tritiated deoxyuridine [1.48 GBq (40 mCi)/L] (Dupont NEN Research Products, Boston MA) was added to each well, and cells were returned to the incubator for another 24 h. The cells were harvested onto glass fiber filters and the radioactivity measured using a Matrix 9600 direct beta counter (Packard Instrument, Downers Grove, IL).
was used to assess the methylation status of lymphocyte DNA. This assay reflects the capacity of genomic DNA to accept 3H-labeled methyl groups, and the DNA incorporated radioactivity is inversely proportional to the level of DNA methylation. Genomic DNA (0.5 µg) extracted from lymphocytes was incubated with 3.0 µmol/L [3H-methyl] S-adenosyl L-methionine (Dupont NEN) containing 74 kBq and 3 units Sss I CpG methylase (New England Biolabs, Beverly, MA) in 1X Sss I buffer at 30°C for 1 h. The samples were subsequently applied to Whatman DE-81 ion exchange filters and washed in 0.5 mol/L sodium phosphate buffer followed by 70% ethanol to remove unincorporated precursor. The air-dried filters were placed in scintillation vials and radioactivity was quantified in a 1900 TR Packard counter using Ultima Gold Scintillant (Packard, Meriden, CT). The results are expressed as cpm of 3H-methyl incorporation/0.5 µg DNA to compare relative changes in methylation status.
), 3'OH DNA fragments in the high molecular weight DNA are initially separated into single strands by heat denaturation. After reassociation, these fragments serve as primers, and the excess of high molecular weight DNA serves as a template in a reaction with DNA polymerase. As a result, incorporation of [32P]dCTP (Dupont NEN) initiated by the Klenow fragment of DNA polymerase I is proportional to the number of 3'OH breaks present. Because a DNA denaturation step is included in the assay, both single- and double-strand DNA breaks are detectable.
).
). The statistical analysis was conducted using SAS version 6.12 (SAS Institute, Cary NC) and SigmaStat version 1.02 (Jandel Scientific Software, San Rafael CA) statistical software programs. Differences were considered significant at P < 0.05.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
) except for one subject who started and remained low (239-253 nmol/L). Folate concentrations of the young red cells were always significantly higher than those of the old red cell fraction; however, they were no more reflective of changes in dietary folate intake than were the total or old red cell fractions.
) throughout.
), with a mean increase of 3.3 µmol/L. The ANOVA showed a significant inverse relation between Hcy and folate intake (P = 0.001), and the Hcy increases from baseline were significant beginning at d 27 (P < 0.02 by paired t test).
).
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
). Hence, changes (or lack of change) in some folate-dependent measures taken in this study might have been different with longer dietary treatment periods.
that 200-250 µg/d of food folate restored low plasma folate concentrations in folate-depleted premenopausal women.
, O'Keefe et al. 1995, Selhub et al. 1993). Although all subjects followed the general inverse relation, the magnitude of the Hcy response to folate depletion was different among individuals. The Hcy values for three subjects changed only within the normal range, whereas five subjects showed mild elevations, between 12 and 18 µmol/L. These results are similar to those from a previous study of controlled folate intake in healthy men, which also showed clear inter-individual differences of Hcy response to folate depletion (Jacob et al. 1994). This indicates that even under precisely controlled nutriture, a variety of genetic and environmental factors may affect circulating Hcy concentrations.
). The lack of decline in the elevated Hcy values at d 70-83 with folate intake of >286 µg/d (Fig. 1) suggests that >286 µg/d of folate was required to lower the elevated Hcy in these women. The 14-d period of 286+ µg/d folate intake during d 70-83 should have been enough time for plasma Hcy to at least begin to decline because the high values declined significantly within 8 d in the final period when 516 µg/d was fed, and within 9 d when men in a previous study were repleted with 440 µg/d (Jacob et al. 1994). The Hcy concentrations did not return to baseline, possibly because the 8-d final repletion period was not long enough to see additional decline. The question arises why 286+ µg/d did not lower elevated plasma Hcy when all women started with baseline Hcy levels <12 µmol/L, and their mean free-living folate intake could be estimated as 272 µg/d from NHANES II food records of white women age 51-65 y (Subar et al. 1989). The difference may be due to the low choline content of the experimental diet, because choline can methylate Hcy via betaine (Selhub and Rosenberg 1996). By comparison, O'Keefe et al. (1995) reported that elevated Hcy levels (>16 µmol/L) were observed in premenopausal women ingesting 200 µg/d but not 300 or 400 µg/d of folate. Taken together, data from these two studies suggest that, for some women, folate intake of 300 µg/d or more is required to maintain Hcy levels <12 µmol/L.
, who found that chronic severe folate/methyl deficiency in rats induced genome-wide and p53 gene-specific hypomethylation in preneoplastic liver, along with an increase in DNA strand breaks. In other studies of folate deficiency in rats, Balaghi and Wagner (1993) found genomic hypomethylation of hepatic DNA after 4 wk of folate deficiency, Kim and Christman (1995) found no hypomethylation of hepatic and colonic DNA during moderate folate deficiency and Kim et al. (1997) found hypomethylation of the p53 tumor suppressor gene (but not genome-wide hypomethylation) and DNA strand breaks due to isolated folate deficiency. Although genome-wide DNA hypomethylation was observed in human lymphocytes in this study, DNA strand breaks were found to be decreased, not increased, with short-term moderate folate deprivation.
). The different results may be due to an increased sensitivity of the DNA methyl acceptance assay compared with the metabolic methylation tests used in the previous study (measuring methylated niacin metabolites after an oral nicotinamide dose and urinary excretion of the methyl-rich compounds creatinine and carnitine), or to an increased sensitivity of the methylation pathways to folate deficiency in the postmenopausal women compared with the younger men. Although the methionine intake of the women was adequate (780 mg/d), the relatively low amount of choline in the diet (147 mg/d) may have contributed to the sensitivity of the observed inverse relationship between folate intake and DNA hypomethylation.
, Glynn and Albanes 1994, MacGregor et al. 1997). Recently, Fang et al. (1997) reported genomic DNA hypomethylation of gastric cancer compared with normal tissue, and an association between hypomethylation and low serum folate concentrations in the cancer patients.
, MacGregor et al. 1997). Our results are similar to the findings of Blount et al. (1997), who reported that elevated incorporation of uracil into DNA in folate-deficient individuals was reversed by supplementation with folate. However, the dUTP/dTTP ratio in our subjects did not decrease upon folate repletion. This may be because of the much shorter and smaller folate repletion of our study (20 d of 286-516 µg/d) compared with that of Blount et al. (6-8 wk of 5 mg/d).
found that lowered NAD and de novo thymidylate synthesis accompanied increased DNA strand breaks in spleen cells of folate and folate-methyl-donor-deficient rats. Henning et al. (1997) recently showed that in liver from male rats fed methyl- and folate-deficient diets with or without niacin, PARP activity was altered and NAD concentrations significantly decreased. The decline in lymphocyte NAD observed in this study occurred despite a niacin intake of 29.7 mg of niacin equivalents per day, 214% of the current RDA for adult women (NRC 1989).
) and also in liver of rats fed chronically a folate/methyl-deficient diet (James et al. 1997). In the latter study, increased cell apoptosis was accompanied by elevated liver PARP and a significant reduction in NAD concentrations. The decline from baseline in [3H]deoxyuridine uptake in the mitogen-stimulated lymphocytes (Table 1) indicates a decreased ability of the cells to proliferate in vitro upon ingestion of the experimental diet and follows the same pattern as NAD concentrations and DNA strand breaks, in that short-term folate repletion did not reverse the decline.
showed that LDL oxidation can occur by the interaction of sulfhydryl groups from cells with transition metal ions, and Jones et al. (1994) reported that lipid peroxidation accompanied the toxicity of Hcy to endothelial cells in culture. However, Dudman et al. (1993) found that HDL cholesterol ester hydroperoxides were not elevated in four patients with hyperhomocysteinemia and concluded that Hcy does not impose significant oxidant stress. Recently, folate deficiency in rats was reported to increase plasma Hcy and lipid peroxidation products along with a decrease in plasma and platelet fatty acid unsaturation (Durand et al. 1996). Although no relation of plasma MDA to folate intake was found, our urinary MDA results are consistent with previous work suggesting a lipid pro-oxidant effect of elevated Hcy concentrations.
). Using the latter figures, increased folic acid consumption estimated at ~100 µg/d as a result of pending fortification of enriched cereal-grain products (USDHHS 1996) would still leave about 15% of U.S. women consuming less than the current RDA. In 310 women, age 67-74 y, of the Framingham Heart Study, 20% had elevated plasma Hcy, which was strongly associated with low folate status in the overall population studied (Selhub et al. 1993). Therefore, the present findings may be relevant to an appreciable segment of postmenopausal U.S. women with low dietary folate intakes. These results underscore the need for further research on the molecular and functional consequences of inadequate folate nutriture.
| |
FOOTNOTES |
|---|
Manuscript received 6 January 1998. Initial reviews completed 3 March 1998. Revision accepted 27 March 1998.
| |
ACKNOWLEDGMENTS |
|---|
The authors thank Virginia Gildengorin for help with statistical analysis of study results, the Bionetics Corporation of Hampton, VA and the staff of the WHNRC Human Nutrition Suite for providing support in operating the Metabolic Unit and for providing nursing and dietary services. Thanks also to the staff of the WHNRC Bioanalytical Laboratory for specimen processing and various chemical analyses, and Mark Kutnink of the WHNRC for plasma MDA determinations. We also thank Russell Okoji of the UCLA School of Public Health and Donald Jenden and staff of the UCLA School of Medicine for processing and analysis of diet homogenates for choline. Lastly, the study subjects are acknowledged, for the dedication and generosity they showed in completing the study.
| |
LITERATURE CITED |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
use of CpG methylase.
Biochem. Biophys. Res. Commun.
1993;
193:1184-1190[Medline] relevance to development of arteriosclerosis.
Arterioscler. Thromb.
1993;
13:512-516-amylase, and folate conjugase treatments.
J. Agric. Food Chem.
1997;
45:135-139 This article has been cited by other articles:
|
|
 |
|
  L. S. Kristensen and A. Dobrovic Direct Genotyping of Single Nucleotide Polymorphisms in Methyl Metabolism Genes Using Probe-Free High-Resolution Melting Analysis Cancer Epidemiol. Biomarkers Prev., May 1, 2008; 17(5): 1240 - 1247. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. E. Kelemen, T. A. Sellers, J. M. Schildkraut, J. M. Cunningham, R. A. Vierkant, V. S. Pankratz, Z. S. Fredericksen, M. K. Gadre, D. N. Rider, M. Liebow, et al. Genetic Variation in the One-Carbon Transfer Pathway and Ovarian Cancer Risk Cancer Res., April 1, 2008; 68(7): 2498 - 2506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. O. Wright and A. Baccarelli Metals and Neurotoxicology J. Nutr., December 1, 2007; 137(12): 2809 - 2813. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-F. Chou, C.-C. Yu, and R.-F. S. Huang Changes in Mitochondrial DNA Deletion, Content, and Biogenesis in Folate-Deficient Tissues of Young Rats Depend on Mitochondrial Folate and Oxidative DNA Injuries J. Nutr., September 1, 2007; 137(9): 2036 - 2042. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Gemmati, A. Ongaro, S. Tognazzo, L. Catozzi, F. Federici, E. Mauro, M. Della Porta, D. Campioni, A. Bardi, G. Gilli, et al. Methylenetetrahydrofolate reductase C677T and A1298C gene variants in adult non-Hodgkin's lymphoma patients: association with toxicity and survival Haematologica, April 1, 2007; 92(4): 478 - 485. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. T. Heijmans, D. Kremer, E. W. Tobi, D. I. Boomsma, and P. E. Slagboom Heritable rather than age-related environmental and stochastic factors dominate variation in DN |
