Document JJjdLQoKnXaa0Ed7NG29LMDeK

ELSEVIER Mutation Research 417 (1998) 101-114 Genetic Toxicology and Environmental Mutagenesis Micronuclei in lymphocytes and exfoliated buccal cells of postmenopausal women with dietary changes in folate Nina Titenko-Holland a *, Robert A. Jacob b, Nong Shang a, Anita Balaraman a, Martyn T. Smith a a Division ofEnvironmental Health Sciences, 217 Warren Hall, School ofPublic Health, Univ ersity of California, Berkeley, CA 94720-7360, USA b USDA, Western Human Nutrition Research Center, POB 29997, Presidio ofSan Francisco, San Francisco, CA 94129, USA Received 16 July 1997; revised 29 June 1998; accepted 7 July 1998 Abstract Folate deficiency is associated with anemia, birth defects, cancer and neuropsychiatric disorders. The purpose of this study was to determine if a moderate folate deficiency during controlled changes in folate intake would affect chromosomal damage in lymphocytes and buccal cells. A study of nine healthy postmenopausal women volunteers (age 49-63 years) was carried out in a metabolic unit (baseline week with folate intake of 195 mg/day, five-week depletion at 56 mg/day, and gradual repletion including four weeks at 111 mg/day, 11 days at 286 mg/day and 9 days at 516 mg/day). Plasma folate, vitamin B-12, and homocysteine were measured weekly. Cytogenetic damage was assessed by scoring micronucleus (MN) frequency in lymphocytes and buccal cells three times: (1) at the beginning of the study, (2) at the end of depletion, and (3) after repletion. The MN frequency increased in binucleated lymphocytes, as well as in all lymphocytes, after depletion (p = 0.037), and later decreased following repletion (p = 0.028). Both kinetochore-positive and kinetochore-negative MN were increased after depletion (p = 0.015 and 0.028), but after repletion only the change in kinetochore-positive MN was statistically significant (p = 0.048). The main variables affecting MN were: (1) vitamin B-12 level, (2) plasma folate level, and (3) baseline frequency of MN. The MN frequency in exfoliated buccal cells was decreased after dietary supplementation of 516 mg/day folate (p = 0.010). Thus, low folate, without clinical symptoms of anemia, results in higher levels of cytogenetic damage in both the blood and oral cavity of postmenopausal women. 1998 Elsevier Science B.V. All rights reserved. Keywords: Dietary folate; Micronucleus frequency; Postmenopausal woman; Lymphocyte; Buccal cell 1. Introduction Folate deficiency has been associated with can cers of epithelial origin [1,19,21,25], as well as * Corresponding author. Tel.: + 1-510-642-8781; Fax: + 1-510 642-0427; E-mail: ninah@uclink4.berkeley.edu neural tube defects and other birth abnormalities [22,35,48]. Folate deficiency may also contribute to myocardial infarction and neuropsychiatric disorders, especially in the elderly [6,7]. Folate requirements in different health groups have been the subject of extensive discussion [3,26]. More than 10% of the general population, an estimated 50% percent of 1383-5718/98/S19.00 1998 Elsevier Science B.V. All rights reserved. PII: S1383-5718(98)00104-1 102 N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 some low-income minority populations, and at least 25% of elderly people have a folic acid deficiency [3]. Biochemical studies have suggested that distur bances in the nucleotide pool, DNA synthesis, and cell growth are responsible for the adverse health effects associated with folate deficiency [18,47]. A recent study [4] showed that in folate deficiency, synthesis of thymidylic acid from dUMP, its precur sor, was inhibited, causing dUMP to accumulate in the cell. This condition increases the chance of misincorporation of uracil into DNA in place of thymine, and when the DNA repair system removes the uracil it creates breaks in both DNA and chromosomes. In the group of folate deficient volunteers, massive incorporation of uracil into DNA and an increased frequency of micronucleated erythrocytes, presum ably resulting from chromosome breaks, were ob served [4]. A possible link between chromosome damage and folate deficiency was suggested in early experimen tal studies with hereditary fragile sites in human chromosomes induced as a result of cultivation under low-folate conditions [41]. Increased micronucleus (MN) frequency and chromosome breakage in hu man lymphocytes, erythrocytes and bone marrow cells were observed in individuals low in folate (reviewed in Ref. [23]). One large population study in Australia showed plasma folate and vitamin B-12 levels in young individuals were more important than vitamin C and E levels in minimizing chromo some damage in binucleated lymphocytes [12,13]. The highest levels of MN erythrocytes were ob served in splenectomized individuals with exception ally low values of plasma folate, red cell folate, or plasma B-12 [24]. All these data suggest that low folate status is associated with increased cytogenetic damage. The frequency of chromosome aberrations and MN also increases with age [46], although there is limited data available on the relationship between low folate and cytogenetic damage in older people. Neither of these studies of folate deficiency in hu man populations used antikinetochore-antibody stain ing to analyze the mechanism of MN formation. Very few data are available on effect of folate status and MN frequency in exfoliated cells [33]. Metabolism of folic acid is intimately connected with vitamin B-12 and homocysteine. Several studies have shown a significant association between low plasma folate and B-12 levels, on one hand, and increased MN frequency on the other [9,13,24]. Moderate folate deficiency has been linked to in creased plasma homocysteine in both men and women [17,32]. A unique opportunity to study whether folate deficiency caused by controlled di etary folate depletion will result in cytogenetic changes, and whether vitamin B-12 and homo cysteine were associated with these changes, was presented in a project carried out at the metabolic unit of the Western Human Nutrition Research Cen ter, San Francisco [16]. While a group of women volunteers lived at the Center for 13 weeks, their folate intake was controlled to create a mild folate depletion (56 mg/day) followed by gradual folate repletion reaching 3 times the RDA-1998 of 180 mg/day (~ 516 mg/day). Postmenopausal women were chosen for this study to avoid the potential reproductive risk associated with low folate status. Multiple parameters of diet and energy consumption, as well as folate, vitamin B-12, and homocysteine status were monitored weekly. Results of this study showing that moderate folate depletion increased plasma homocysteine and decreased lymphocyte DNA methylation have been described elsewhere [16]. Here, we report our findings with the MN assay in this group, where MN levels in both lymphocytes and buccal epithelial cells were determined. 2. Materials and methods 2.1. Sample collection Blood and exfoliated buccal cells were collected from a group of ten postmenopausal women. Health prestudy screening, study protocol and parameters of diet during folate depletion and repletion have been described elsewhere [16]. Briefly, healthy volunteer nonsmoking women, ages 49-63 years (57.9 + 5.5 years, Table 1), were admitted to the metabolic unit of the USDA, ARS Western Human Nutrition Re search Center (WHNRC) after medical and physio logical screening. Tests for plasma folate and vita min B-12, alcohol, tobacco, and drug use were also performed. Signed informed consent was obtained N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 Table 1 Demographics and vitamin levels in postmenopausal women during dietary study with controlled intake of folate Group description No. of participants Age (years) Weight (kg) Smoking Alcohol Hormonal supplements 9 57.9 + 5.5a 70.6 + 11.9 No No 2 Yes/7 No Folate intake (mg/day) Folate in plasma (ng/ml) B-12 in plasma (pg/ml) Plasma homocysteine (mmol/l) Sample collection d1-5 Baseline 196 9.6+ 5.8 462 + 184 10.0+ 1.2 d6-41 Depletion 56 4.4+ 2.3c 412+ 151b 12.0+ 2.7c d42-91 Repletion 228 7.4+ 1.9d 391+128 13.2+ 4.3 aMean + standard deviation for all data in the table. bp -- 0.05, cp -- 0.01, dp -- 0.001 statistically significant vs. previous collection (by-pair comparison, Student's t-test). 103 from each volunteer. For the duration of the 91 day study, the subjects lived and ate all meals in the WHNRC metabolic unit. Subjects consumed the same experimental low folate diet in a four day menu rotation for the entire period. The diet provided 56 m g/day of folate and was supplemented with vary ing amounts of synthetic folic acid to provide deple tion (56 mg/day, no folate supplement) and reple tion (up to 516 mg/day) of folate. Baseline screen ing (day 1-5) with folate intake of 195 mg/day was followed by the five-week depletion period with 56 mg/day, and repletion period consisting of 4 weeks of 111 mg/day, 11 days of 286 mg/day and 9 days of 516 mg/day of dietary folate. The five week folate depletion period (day 6-41) was designed to reduce body folate stores and produce moderate but not severe folate deficiency, analogous to that of free-living women with chronically low folate intake and body status but without macrocytosis or anemia. The folic acid supplement solutions were mixed into applesauce served at each breakfast and dinner dur ing the repletion period (day 42-91, an average folate intake of 228 mg/day). The diet was supple mented daily with three different vitamin/mineral tablets to provide at least 80% of the RDA-1989 [28] for each essential micronutrient except folate (the diet plus supplements provided 124% of the RDA for vitamin B-12). Individual energy intake and body weights were estimated and adjustments were made to energy intake if the subjects deviated beyond + 5% of the baseline weight (70.6 + 11.9 kg, Table 1 ). Specimens of blood and exfoliated buccal cells were collected from each individual three times: (1) at the beginning of the study (background), (2) at the end of depletion (day 41), and (3) at the end of repletion (days 89, 90 and 92). Blood for serum and red cell folate, plasma B-12 and homocysteine deter minations was collected by venipuncture using EDTA (7.5%) as anticoagulant and processed immediately. Methods for determining folate and B-12 levels, using a competitive protein binding radioassay kit (Quantaphase II B-12/Folate Radioassay, BioRad, Hercules, CA), were described previously [16]. Total plasma homocysteine was analyzed by HPLC fluo rescence detection after derivatization with a fluores cent sulfonic acid reagent [2]. One of the individuals with an abnormally high level of homocysteine (more than three standard deviations (sd) from the mean) was excluded from further analysis. 2.2. Isolated human lymphocyte culture Fasting blood was collected in heparin-vacutainers and lymphocytes were isolated using FicollPaque (Pharmacia, Piscataway, NJ) density gradients 104 N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 and cultured as described previously [45]. Briefly, blood was diluted 1:1 with phosphate buffered saline (PBS) and layered onto Ficoll-Paque with a ratio of blood + PBS: Ficoll maintained at 4:3. The blood was centrifuged at 2000 rpm for 35 min at room temperature. The lymphocyte layer was removed and washed twice in PBS at 1200 rpm for 10 min each, and then washed with RPMI 1640 media. Cell den sity was counted with a hemocytometer. Typically, each culture consisted of an initial density of 1 X 106 cells in 2 ml of culture medium. Culture medium consisted of RPMI 1640 supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), 2 mM L-glutamine, 100 units/ml penicillin, 100 ug/ml streptomycin (Gibco, Grand Island, NY) and 1.5% phytohemagglutinin (PHA, HA15, Burroughs-Wellcome, Greenville, NC). The lymphocyte cultures were grown in a humidified incubator with 5% CO2 at 37C in 15 ml conical polystyrene centrifuge tubes. Cytochalasin B (Sigma, St. Louis, MO) (5 m g/ml) was added to the cultures at 44 h post-ini tiation as described in Ref. [11]. Cytochalasin B prevents the cells from completing cytokinesis result ing in the formation of multinucleated cells. At 72 h, lymphocyte cultures were spun directly (48 g, 10 min) onto glass slides using a cytocentrifuge (Shandon, Sewickley, PA). Slides were allowed to air dry at room temperature for 15 min before methanol fixation. Slides were stored at -- 20C in a sealed box, desiccated under a N2 atmosphere. 2.3. Immunofluorescent staining Detailed procedures for performing the antikinetochore antibody staining method have been described elsewhere [8]. Briefly, methanol fixed slides were incubated for 5 min in PBS containing 0.1% Tween 20. Excess fluid was drained from the slide and 40-50 ml of the antikinetochore antibody (Chemicon, Temecula, CA) diluted 1:1 with PBS contain ing 0.2% Tween 20 is applied. The slide was then coverslipped and placed in a humidified chamber at 37C for 1 h. Following two washes in PBS contain ing 0.1% Tween 20 for 5 min each, excess fluid was again drained and the slides are covered with a 1:50 dilution of fluorescent goat anti-human IgG (Chemicon, Temecula, CA) and incubated again for 1 h. Because the fluorescent-labeled antibodies fade upon exposure to light, this and all subsequent steps were conducted under yellow light. The slides were rinsed twice in buffer plus 0.1% Tween 20 and counterstained with DNA-dye 4' 6-diamidino-2-phenylindole (DAPI) (2 mg/ml) in an antifade solution [20]. Slides were stored refrigerated for up to a week prior to analysis by microscopy. 2.4. Cell division kinetics and cell viability Cell viabilities were determined at 72 h using Trypan blue (0.16%). Replicative index (RI), a mea sure of cell division kinetics, was calculated by scoring at least 400 cells per dose or sample, by counting the percent of cells containing 1, 2, 3 or more nuclei per individual. RI was calculated as follows: RI = [(1 X % mononuclear cells) + (2 X % binuclear cells) + (3 X % tri) + (4 X % tetra)]/100. 2.5. Exfoliated cell collection Buccal cells were collected and analyzed accord ing to a procedure described earlier [43]. Oral mu cosa was swabbed with a moistened wooden tongue depressor. We prepared slides both by direct smear ing of buccal cells or by dropping the washed cell suspension with determined density onto the slides. Washes were performed twice in a buffer solution containing 0.01 M Tris-HCl, 0.1 M EDTA and 0.02 M NaCl at pH 7.0, and cells (50-100 ml of 1.5 X 106-2 X 106/ml suspension) were dropped onto prewarmed slides (37C). Washes with this buffer helped to inactivate endogenous DNases present in the oral cavity and to remove bacteria and cell debris that would complicate scoring. Dropping was prefer able to smearing because it allowed for the even distribution of at least 3000-5000 cells per slide at an optimal density without overlap and eliminated debris and bacteria. After dropping, the cells were allowed to air-dry and fixed in methanol (80% v/v) at 0C for 20 min. After all specimens were col lected, they were mixed, coded and stained with propidium iodide (1 mg/ml, Sigma) in antifade solu tion [20]. N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 105 2.6. Scoring procedure and criteria Lymphocytes were analyzed under a total magni fication of X 1000 using a Nikon microscope with a filter for fluorescein (excitation at 470 nm, dichroic at 510 nm, emission at 520-560 nm) and quinacrine (excitation at 400-440 nm, dichroic at 450 nm, barrier at 470 nm). At least 1000 binucleated lym phocytes (those that had undergone one mitotic divi sion) were scored for the number of MN. All mono-, tri- and tetranucleated cells available at the same slide area where these binucleated cells were ana lyzed were also scored for the presence of MN. The number of these cells was dependent on proliferation (replicative index) and was defined by default (~ 850, 20 and 50 per individual, respectively). When a MN was located using the quinacrine filter, the presence or absence of kinetochore staining was determined by switching to the fluorescent filter. Exfoliated buccal cells were analyzed under a total magnification of X 1000 using a Nikon micro scope with a filter for propidium iodide (excitation at 470 nm, dichroic at 510 nm, emission at 520-560 nm). Two thousand cells were scored for each indi vidual and the presence of MN was recorded sepa rately for each thousand cells. Only cells that were not smeared, clumped or overlapped and that con tained intact nuclei were included in the analysis. Cells undergoing degenerative processes such as karyorhexis, karyolysis, and/or fragmentation of the nucleus [43] were recorded separately and were not included in MN analysis. The frequency of MN was estimated based on the number of normal exfoliated cells. The following criteria for MN analysis were used in lymphocytes and exfoliated cells. MN must have: (a) been less than 1/3 diameter of the main nucleus, (b) been on the same plane of focus, (c) had the same color, texture and refraction as the main nu cleus, (d) had smooth oval or round shape; and (e) been clearly separated from the main nucleus. Ques tionable MN were disregarded. 2.7. Statistical analysis Statistical analysis of data was defined by the design of the study with a relatively small number of subjects who had a high initial variability in both MN and plasma folate level. Longitudinal design allowed by-pair comparison of the data from each individual who served as her own control. This approach eliminated a significant part of the variabil ity by excluding (1) inter-individual variability, and (2) a part of intra-individual variability unrelated to dietary folate changes. A controlled environment of the metabolic unit with standardized diet and lifestyle allowed the changes introduced by folate to become a major variable. Differences in plasma folate, vita min B-12, and homocysteine between baseline, de pletion and repletion were assessed by paired t-tests of values at the end of each period. Non-parametric Wilcoxon's sign-rank test was used to assess the significance of the change in the frequency of MN and other parameters due to folate depletion or reple tion. All data were expressed as an average + sd throughout the text and in the tables and graphs since sd is an index of the inter-individual differences. The change in the MN frequency may depend on the baseline and change in plasma folate, homocysteine and vitamin B-12 levels. Age, body weight, and hormonal supplements were also considered as pos sible covariates. These associations were measured by non-parametric Spearman rank correlation coeffi cients. Linear correlation coefficients were also as sessed. p-Values were calculated based on the Spearman rank test. In all cases differences were considered significant where p was < 0.05. 3. Results 3.1. Micronuclei and replicative index in lymphocytes Controlled folate intake during the depletion phase of the study resulted in a significant decrease of plasma folate (9.6 + 5.8 ng/ml baseline level vs. 4.4 + 2.3 ng/ml at day 42, p - 0.01, Table 1) fol lowed by an increase after repletion (7.4 + 1.9, p -- 0.001). Initially, the nine study subjects differed appreciably in their plasma folate (2.9-19.5 ng/ml) and B-12 (251-831 pg/ml), but this variability de creased at the end of the study due to the controlled diet and environment (4.6-8.7 ng/ml and 314-698 pg/ml for folate and B-12, respectively) [16]. Vita min B-12 was decreased from the baseline level of 462 + 184 pg/ml to 412 + 151 pg/ml at the end of 106 N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 depletion period ( p < 0.05). A slightly lower level was observed after the repletion period (391 + 128 pg/ml) but this difference was not significant com pared to the day 42 level. At the end of the folate depletion period, five of the nine subjects had plasma folate concentrations below or near the lower limit of normal (3-20 ng/ml, [42]), while all vitamin B-12 concentrations remained in the normal range of 200 835 pg/ml [42]. Individual changes in the MN fre quency during folate depletion and repletion are shown in Table 2. Though the average levels of total MN, micronucleated cells, and kinetochore-positive and kinetochore-negative MN reflect the changes in the study group, more precise statistical analysis was performed using by-pair individual comparison of these parameters after depletion and repletion. The decrease in folate intake was accompanied by a significant increase in total MN frequency in binu- cleated cells ( p = 0.038, Table 3). Less than 20% of binucleated cells with MN had more than one MN, and there was no significant difference in the fre quency of multiple MN between different cell collec tions. No significant change in the replicative index was observed due to folate depletion and repletion (Table 2). To assess whether MN changes occur in subpopu lations of lymphocytes other than binucleated cells, we also analyzed mononucleated, trinucleated and tetranucleated cells. This approach allowed us to obtain more information regarding the effect of fo late depletion on cytogenetic damage in human lym phocytes in vivo. The same pattern of increased MN frequency after folate depletion followed by de creased frequency after repletion, described above for binucleated lymphocytes, was also observed for mononucleated cells (Fig. 1). The frequency of MN Table 2 Cytogenetic effects of dietary folate changes on micronuclei in binucleated human lymphocytes Micronuclei/1000 cells d1-5 Baseline d6-41 Depletion Individuals (ID no.) 1 2 3 4 5 7b 8 9 10 5.9a 22.7 6.9 25.1 11.6 17.0 8.2 22.0 14.0 6.6 19.0 8.7 23.0 33.5 20.5 14.9 23.7 25.2 d42-91 Repletion 3.9 24.7 12.7 16.7 9.6 26.4 8.9 14.8 8.9 Group total MN cells/1000 cells MN(total)/1000 cells Kinetochore-positive MN + per 1000 cells Kinetochore-negative MN -- per 1000 cells Replicative indexh 14.8+7.3c 17.2+9.2 9.4+4.8 7.4 + 5.5 1.6 + 0.2 19.5+ 8.4d 23.3 + 11.1f 12.5+ 6.4g 10.8+ 7.6 1.7+ 0.1 14.1 + 7.5e 16.2+ 9.0 8.7+ 4.6 7.4+ 5.0 1.7+ 0.2 aMicronucleated cells per 1000 binucleate lymphocytes (between 980 and 1055 cells were scored from two parallel cultures for each individual). bSubject 6 was excluded as an outlier based on an unusually high level of plasma homocysteine. cMean + sd. dDifference between the levels of MN at the first and second collection was calculated using by-pair comparison, Wilcoxon's sign rank test, p = 0.066. eDifference between the levels of MN at the second (depletion) and third (repletion) collections, by Wilcoxon's one-tail test, p = 0.082. fp = 0.038. gp= 0.029. hReplicative index = [(% mononuclear cells) + (2 X % binuclear cells) + (3 X % trinuclear cells) + (4 X %tetra or > tetranuclear cells)]/100. N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 107 Table 3 Micronucleus frequency in subpopulations of lymphocytes with different nuclear number Lymphocytes Mononucleated Binucleated Trinucleated Tetranucleated Cell number 823.6 + 419.2a 1014.6+ 36.4 MN per 1000 cells 1st collection 3.4 + 3.1 17.2+ 9.2 2nd collection 5.1 + 4.7 23.3+ 11.1 3rd collection 3.4 + 2.1 16.2+ 9.0 p-Values for changes between 1st and 2nd collection MN cells 0.103d 0.066 Total MN 0.103 0.038 MN + 0.066 0.029 MN -- 0.285 0.126 p-Values for changes between 2nd and 3rd collection MN cells 0.150 0.082 Total MN 0.125 0.082 MN + 0.049 0.180 MN -- 0.410 0.213 17.9+ 15.1 32.2+ 43.6 128.9+ 116.1 167.7 + 119.5 0.038 0.038 0.038 0.455 0.066 0.212 0.325 0.455 51.5+ 33.0 238.5+ 301.2 195.1+ 111.4 192.2+ 174.8 0.367 0.455 0.103 0.455 0.249 0.249 0.064 0.367 Total nuclei 3112.2+ 355.9b 8.9+ 3.3c 13.0+ 5.7 9.7+ 5.2 0.037 0.037 0.082 0.028 0.020 0.248 aAverage number of cells (mean + sd) scored for each of three cell collections for each of nine study subjects. bTo compare the MN frequency for lymphocytes with different nuclear number, total nucleus number for all scored cells was determined as a sum of mononucleated cells + number of binucleated cell multiplying by two, + trinucleated cells X 3, and + tetranucleated cells X 4. cWeighted MN frequency per 1000 nuclei of mono-, bi-, tri- and tetranucleated cells normalized for the nucleus number (mean + sd). dp-Values for the MN changes by Wilcoxon's one-tail rank test. in mononucleated cells varied between 3.4 MN (1st and 3rd collections) and 5.1 MN (2nd collection) per 1000 cells (Table 3). The frequency was significantly higher for binucleated cells (16.2-23.3 MN per 1000 cells, p < 0.01). An estimation for tri- and tetranucleated cells was less stable since comparatively small numbers of these cells were available for analysis ( ~ 20 and 50 for each individual in each of the three cell collections). However, they had the highest frequency of MN per cell (Table 3). The MN frequency for trinucleated cells (32.2-167.7 MN per 1000 cells) was statistically different from that in the tetranucleated cells (192.2 and 238.5 MN per 1000 cells, p< 0.05). After MN frequencies for all cells were normalized per nucleus number to account for the difference in mono-, bi-, tri- and tetranucleated cells, the significance of the differences in the MN frequency per 1000 nuclei between these subpopula tions was sustained ( p < 0.05, Fig. 2). The effect of both depletion and repletion on total MN frequency adjusted for nucleus number was significant (p = 0.037 and 0.028, respectively, Table 3). The mechanism of MN formation was examined using antikinetochore-antibody staining to distin guish between kinetochore-positive (MN + ) and kinetochore-negative (MN -- ) micronuclei. A slightly larger fraction (54-56%) of lymphocyte MN at all collection times were MN + (Table 3), reflect ing aneugenic mechanism of cytogenetic damage. Both MN + and MN -- were similarly affected by dietary changes in folate intake, but only changes in the MN + frequency were significant after depletion ( p = 0.029) and repletion ( p = 0.027) for all ana lyzed lymphocytes (Table 3). An increase in MN -- after depletion approached statistical significance ( p = 0.066). Similar results were obtained when only mononucleated and binucleated cells were included in the analysis (Fig. 3). Total MN were increased after depletion (p = 0.014) and decreased after re pletion ( p = 0.066) as were MN + ( p = 0.015 and p= 0.048 for depletion and repletion, respectively). MN -- were elevated after depletion (p = 0.028), while a small decrease following repletion was not statistically significant (p = 0.41). These data sug gest that both chromosome breakage and aneuploidy were induced by folate depletion. 3.2. Micronuclei in exfoliated buccal cells Between 0 and 6 MN per thousand buccal cells were observed during dietary folate study of post- 108 N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 1st collection Ei 2nd collection H 3rd collection 12 Number of nuclei per cell Fig. 1. Micronuclei in mono and binucleated lymphocytes during dietary folate changes. The average number of mononucleated cells scored for each of the three cell collections was 824 + 419, and the number of binucleated was 1015 + 36. The MN frequency was normalized for the number of nuclei per cell. The difference in the MN frequency of three cell subpopulations was statistically significant ( p - 0.01). both Spearman rank correlation and linear regres sion: (1) age, (2) body weight, (3) hormonal supple ments, (4) individual folate level and change in plasma folate, (5) vitamin B-12 baseline and change in plasma level, (6) homocysteine level, and (7) MN, MN + , MN -- baseline and changes after depletion and repletion. Variables that were significant or ap proaching significance are shown in Table 5. Data shown were based on the analysis of all nuclei of mono-, bi, tri- and tetranucleated lymphocytes. We also performed a similar analysis for binucleated lymphocytes only and for binucleated and mononucleated cells together. Since these results generally supported our findings based on the analysis of all lymphocytes, we present the latter as more compre hensive. In lymphocytes during the depletion phase of the study, the most noticeable negative association was observed between MN frequency and plasma vitamin B-12 level (p = 0.02 for total MN, and menopausal women (Table 4). An average baseline frequency of 2.56 + 1.78 MN cells was not apprecia bly affected by folate depletion but was significantly decreased after repletion (1.06 + 0.73, p = 0.11, Table 4). Very few cells had more than one MN; thus, the frequency of MN cells and MN per 1000 cells were almost identical. Degenerated cells result ing from karyorhesis, karyolysis, and fragmentation were decreased by the end of the depletion period (2.9 + 1.3 vs. 5.4 + 3.6 baseline, p = 0.014), though there was no further change after repletion. 3.3. Factors affecting MN variability Dietary folate intake changes were accompanied by changes in levels of vitamin B-12 and homo cysteine, which are involved in folate metabolism. Since we observed a high individual variability in the MN frequency in both lymphocytes and exfoli ated buccal cells, we performed a correlation analy sis to reveal the most significant covariates. Associa tions with the following variables were analyzed by Fig. 2. Micronuclei in the lymphocytes with different nucleus number. For each cell collection ~ 1000 binucleated cells were scored, and the number of mono ( ~ 850), tri ( ~ 20) and tetranu cleated ( ; 50) cells was defined by the replicative index. Total of > 3000 nuclei for each individual was analyzed for each of the three cell collections during folate depletion and repletion. The difference between the MN frequencies in mononucleated, binucleated, tri and tetranucleated lymphocytes was statistically signif icant ( p -- 0.05 by Wilcoxon's one-tail rank test). N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 109 the MN frequency after repletion (from -- 0.73 to -- 0.55, p = 0.01-0.06). Buccal cell MN were characterized by association with changes of vitamin B-12 in both the depletion and repletion phase (0.77-0.52, p = 0.009-0.074). A positive correlation with age (p = 0.02) and a negative correlation with MN level at the second collection (p = 0.02) were also observed. In summary, based on the p-values estimated by Spearman rank coefficients, the main factors affect ing MN frequency in lymphocytes were: (1) plasma vitamin B-12 and change in B-12 level, (2) MN and MN + baseline and change due to folate intake, and (3) plasma folate level and change in plasma folate due to depletion or repletion. Similar associations were observed for the buccal cell MN, though there Baseline Depletion Repletion Cell collection Fig. 3. Kinetochore-positive and kinetochore-negative micronuclei in human mono- and binucleated lymphocytes during dietary folate changes. Total number of nuclei analyzed per individual for each of the three cell collections was 2857.8 + 420.7. Kineto chore-positive and kinetochore-negative micronuclei were differ entiated by the presence of bright yellow-green signal after antikinetochore-antibody staining. Cells were counterstained with DAPI. Total MN were increased after depletion ( p = 0.014) and decreased after repletion ( p = 0.066) as were MN + ( p = 0.015 and p = 0.048 for depletion and repletion, respectively). MN -- were elevated after depletion ( p = 0.028), while a small decrease following repletion was not statistically significant ( p = 0.41). p = 0.009 for MN +). A weaker negative associa tion was also observed for the baseline plasma folate level ( -- 0.55, p = 0.06). However, Spearman corre lation coefficients between MN and folate changes were positive (0.53-0.57) with borderline statistical significance (p = 0.06-0.07). Association with age was observed only for MN + (p = 0.03). Repletion phase was mostly characterized by positive associa tion with vitamin B-12 level (0.62-0.67, p = 0.03 0.04) while the changes in MN + and B-12 were negatively correlated (p = 0.038). At this stage no direct association was observed between MN fre quency and folate level or change in plasma folate. Nevertheless, such a link was found between the increase in MN due to depletion and the decrease in Table 4 Cytogenetic effects of dietary folate changes on micronuclei in human exfoliated buccal mucosa cells Micronuclei/1000 cells Baseline Depletion Repletion Individuals (ID no.) 1 2 3 4 5 7b 8 9 10 1.5a 1.5 2.0 1.0 1.5 1.5 5.8 3.0 5.0 2.5 2.5 2.0 1.0 1.0 2.5 3.0 4.0 4.5 3.5 1.5 0.5 0 1.5 0.5 1.5 1.0 1.0 Group total MN cells/1000 cellsc MN per 1000 cells Degenerated cells (%) 2.56 + 1.78 2.56 + 1.78 5.4+ 3.6 2.44 + 1.26 2.50+1.27 2.9 + 1.3f 1.06 +0.73d 1.22 + 1.00e 3.6+1.4 aMicronucleated cells per 1000 buccal cells (2000 cells were scored from two slides for each individual except no. 8 with 2028 cells). bSubject 6 was excluded as an outlayer based on an unusually high level of plasma homocysteine. c Mean + sd. d Difference between the levels of MN in exfoliated buccal cells at the second (depletion) and third (repletion) collections by Wilcoxon's one-tail test, p = 0.010. e Difference between the levels of MN in exfoliated buccal cells at the second (depletion) and third (repletion) collections by Wilcoxon's one-tail test, p = 0.020. fFirst vs. second collection by Wilcoxon's one-tail test, p = 0.014. 110 N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 Table 5 Significant variables associated with micronucleus frequency in lymphocytes and exfoliated buccal cells Endpoint Variable L-associationa S-associationb p-Value Lymphocytesc Depletion MN1d-2 F MN1 2 b1 MN1-2 MN + 1-2 F1-2 F1 mn+1-2 B1 MN +1- 2 F1-2 mn+1-2 Age mn-1-2 F1-2 -- 0.65 --0.68 0.70 -- 0.56 -- 0.74 0.60 -- 0.73 0.58 -- 0.53 -- 0.68 0.57 -- 0.55 -- 0.77 0.53 -- 0.64 0.57 0.067 0.021 0.059 0.060 0.009 0.073 0.033 0.059 Repletion MN2 - 3 MN2-3 MN2 - 3 MN + 2-3 MN + 2 3 MN + 2-3 MN --2-3 MN1 2 B1 B2 MN +1 2 B2 B2-3 MN --1-2 -- 0.63 0.37 0.35 -- 0.58 0.40 -- 0.34 -- 0.82 -- 0.57 0.62 0.55 -- 0.55 0.67 -- 0.62 -- 0.73 0.054 0.040 0.064 0.060 0.028 0.038 0.013 Buccal cells Depletion MN1 - 2 B1-2 0.39 0.77 0.009 Repletion MN2-3 MN2 - 3 MN2-3 MN2 Age B2-3 -- 0.76 0.58 -- 0.27 -- 0.69 0.68 -- 0.52 0.021 0.024 0.074 aLinear regression coefficient. bSpearman rank correlation coefficient. cAll lymphocytes (mono-, bi-, tri- and tetranucleated) were in cluded into analysis, the frequency of MN was normalized by the number of nuclei per cell, e.g., the number of MN per nucleus was used. dMNi_2 was used to describe the change in MN between first and second collection (2-3 between second and third). F is plasma folate level at the first collection, while Fi-2 stands for the change of folate level between first and second collection (2-3 are used respectively for the change between second and third collections). B is plasma vitamin B-12 level at the first collection. was no significant correlation with folate. In two cases, one for lymphocytes and one for buccal cells, MN were associated with age. Thus, MN frequency in lymphocytes and exfoliated cells appeared to be affected by multiple factors both reflecting con trolled dietary intake and also individual genetic differences. 4. Discussion The MN assay in binucleated human lymphocytes and cultured cell lines has been shown to be an effective tool in measuring cytogenetic damage of agents with different mechanisms of genotoxicity in vitro and in vivo [8,11,29]. The MN assay has been used to analyze the effect of low folate status in several cohorts using binucleated lymphocytes [12,13] and erythrocytes in splenectomized patients [4,24]. A principal conclusion was that low folate is generally associated with increased cytogenetic dam age. We reached the same conclusion after monitoring a group of postmenopausal women living in a highly controlled environment in the metabolic unit. Though initially these women had quite different MN, folate, B-12 and homocysteine levels, the difference in mi cronutrients was noticeably decreased during the study, while MN variability did not change apprecia bly. Moreover, although we observed a statistically significant increase in lymphocyte MN after folate depletion and decrease as a result of repletion, there was no correlation between individual folate mea surements and MN frequencies at any collection time. Our findings corroborate those in Ref. [13], where in a large study of 126 vegetarians and 138 non-vegetarians of different ages (including a small group of individuals older than 60 years), where such a correlation, except in young females, was not seen either. The lack of association in older individuals may be explained by insufficient statistical power and high inter-individual variability in epidemiologi cal studies. It is also possible that such an association could be disguised by the effect of other nutrients closely related to folate. In fact, Fenech and Rinaldi [13] observed a negative correlation between MN frequency and the level of B-12 both in young males and females, and in males vitamin C was also signif icant in defining the level of MN. A study of 122 splenectomized healthy individu als [24] also found no correlation between individual levels of folate and MN. However, seven individuals with the highest observed frequencies of micronucleated erythrocytes from this group together with 64 individuals studied by Everson et al. [9] and Smith et al. [39] all had exceptionally low values of plasma folate, red cell folate, and plasma vitamin B-12. N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 111 Moreover, such an association was observed in an individual with Crohn's disease, a condition which results in the inhibition of dietary folate absorption and transport. These data suggest that folate is one of the major determinants of the types of damage that lead to spontaneous MN formation in erythrocytic cells. Another important observation made in Refs. [13,24], and our study is that vitamin B-12 status also appears to have a major influence on the fre quency of micronucleated cells. Plasma B-12 values in our group of postmenopausal women indicated no apparent vitamin B-12 deficiency throughout the study. Nevertheless, it is possible that a negative B-12 balance may have occurred since plasma B-12 values declined from baseline. According to Ref. [15], a negative vitamin B-12 balance begins when vitamin B-12 absorption falls low enough to deplete the amount of vitamin B-12 on its primary delivery protein, TCII, resulting in a low holo-TCII level, even though total vitamin B-12 levels remain within normal limits. Fenech [10] found the same connec tion between MN and vitamin B-12 in a group of 50to 70-year-old men whose diet was supplemented with folate above RDA either through tablets or cereal. A significant improvement in folate status was not accompanied by a change in MN frequency in binucleated lymphocytes, while an inverse correla tion observed between plasma B-12 and MN fre quency was statistically significant. Both folate and B-12 are intimately involved in metabolic pathways responsible for maintenance of the DNA precursor pool. Blount et al. [4] made a further step in establish ing the mechanism by which folate deficiency may cause chromosome damage. They supplemented the diet of 19 splenectomized individuals with folic acid (5 mg pteroylglutamic acid per day) and measured plasma and erythrocyte folate, DNA uracil levels, and MN frequency. The study concluded that DNA uracil levels are closely associated with frequencies of MN, and that both are minimized by supplementa tion with folic acid. Uracil misincorporation is asso ciated with an increased frequency of abasic sites and subsequent strand incisions [4], suggesting that folate deficiency can result in chromosome breakage. Also supporting this hypothesis is data indicating a high frequency of fragile sites in low folate cultures [41]. However, no analysis of the mechanism of MN formation was performed in any previous MN study of lymphocytes or erythrocytes. In the present study, we applied antikinetochoreantibody staining, which allows differentiation of MN formed due to either chromosome lagging or chromosome breakage. Both types of MN were in creased due to folate depletion ( p = 0.015 and p = 0.028 for MN + and MN -- , respectively). These results provide the first evidence that not only chro mosome breakage, but the integrity of chromosome segregation and the cell division mechanism as well, could be affected by folate deficiency. Advances in the MN assay of human lymphocytes (cytokinesis block, antikinetochore antibody staining or FISH with centromeric probe) have led to its increased use and its recommendation as a `routine' index of cytogenetic damage [14]. Binucleated lym phocytes have been used for MN analysis in the majority of studies. Although reports of the MN assay as a biomarker of environmental exposures continue to accumulate, many questions exist regard ing its limitations and interpretation of the data from population studies. These important issues include intra- and inter-individual variability [34], the effect of cytochalasin B [30], differential sex effect favor ing a greater MN frequency in women [5,34], age, smoking, and other sources of variability [46]. The modification of the MN assay introduced in Ref. [11] allows differentiation of subpopulations of lympho cytes which have undergone a different number of divisions in culture by arrest of the cytoplasm divi sion (using cytochalasin B) while nuclei divide nor mally. In our study, we initially performed analysis of binucleated lymphocytes separately to obtain data comparable to other published results. A mild re sponse to dietary changes in folate was observed, but only the increase in total MN and MN + due to depletion was significant ( p = 0.038 and p = 0.029, respectively). We also collected data on the MN frequency in mono-, tri- and tetranucleated cells. Since the number of mononucleated and binucleated cells was similar, as was their response to folate changes, we combined MN data for all nuclei. This approach provided more statistical power by increas ing the total cell number, and incorporating all cells possibly affected by folate deficiency in vivo. As a result, a significant change in MN frequency due to 112 N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 depletion and repletion (p = 0.037 and 0.028, re spectively) was revealed. These findings highlight the importance of proper selection of cell and target tissue for folate defi ciency analysis. In all previous publications, only binucleated lymphocytes and erythrocytes were used for MN analysis. Since folate deficiency was associ ated with cancers of epithelial origin [19], and a significant reduction in oral leukoplakias accompa nied by a lower MN frequency in buccal cells was observed in tobacco/betel quid chewers treated with vitamins [36,40], it was of interest to determine if cytogenetic damage in buccal cells would be associ ated with controlled changes in folate intake. The MN assay in exfoliated cells proved to be a sensitive biomarker of environmental exposure to smoking/tobacco chewing [36], aerial formaldehyde [44], and a high level of arsenic in drinking water [27]. However, in the study of smokers performed in Ref. [33] no association between folate status and MN frequency was found. The folate repletion protocol in the present study included the gradual increase in folate intake from 111 mg/day (day 42-69) to 286 mg/day (day 70 80), and finally 516 mg/day (day 81-91), with an average level during the repletion period of 228 mg/day. It is possible that the cytogenetic effect of the increased folate level was accumulating through out the entire repletion period. This scenario is even more plausible for lymphocytes, since they circulate in the blood stream for several weeks and even months [46]. Therefore, the most adequate measure of their exposure will be defined by an average level of folate intake (228 mg/day). On the other hand, the period in which the effect of folate repletion in exfoliated cells could be registered (by MN de crease) is the last 7-14 days before cell collection. Based on the dynamics of cells with MN induced by radiotherapy and chemotherapy, this is the typical lag period between an occurrence of MN in the basal layer of epithelium and the time cells migrate through the cellular strata and are sloughed off [37,38]. Thus, the MN that were observed in these exfoliated cells would reflect damage caused by factors in the oral cavity present 7-14 days earlier. In our study, the intake of 516 mg/day was maintained for the last 9 days before cell collection. This more than tripling of the RDA-1989 for folate resulted in a significant decrease in MN frequency in buccal cells. Lympho cytes appear to be more sensitive to dietary folate changes than exfoliated cells, since a statistically significant MN change in the former was observed after only a two-fold decrease/increase in folate intake, while in the latter a three-fold increase was necessary. Further studies are warranted to compare dose response of cytogenetic damage to dietary fo late changes in different cell types. We speculate that the cytogenetic response in both exfoliated cells and lymphocytes would be even stronger were we to study a group of individuals with clinical folate deficiency. Our data add strength to the recent suggestions that a higher RDA for elderly people, premenopausal and pregnant women, and children may better protect their health [31]. Further, our study also indicates that MN in exfoli ated cells are a sensitive biomarker of cytogenetic damage. In conclusion, cytogenetic monitoring of post menopausal women showed that MN frequency in lymphocytes increased after controlled dietary folate depletion, and later decreased following repletion. MN analysis in all lymphocytes was more statisti cally powerful than analysis of binucleated cells to assess cytogenetic damage in vivo. Both kinetochore-positive and kinetochore-negative MN demon strated similar changes, suggesting that anomalies in cell division and chromosome segregation, in addi tion to chromosome breakage, are associated with folate deficiency. The main covariables affecting MN changes during dietary depletion and repletion were plasma vitamin B-12, folate, and the individual baseline frequency of MN. MN frequency in buccal exfoliated cells did not respond to depletion, but was significantly decreased after dietary folate repletion up to 516 mg/day folate. Further study is needed to establish whether higher folate intake will lower cytogenetic damage in different epithelial tissues and thus decrease the risk of cancer, especially in older people. Acknowledgements The authors wish to acknowledge the staff of the WHNRC Human Nutrition Suite for providing sup port in operating the Metabolic Unit and help with N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 113 collection of samples. This research was partially supported by the NIH grants P42 ES04705 and P30 ES01896. We thank Ms. Sauda Parvatham for tech nical assistance with slide preparation and analysis. References [1] Anonymous, Folate, alcohol, methionine, and colon cancer risk: is there a unifying theme?, Nutrition Reviews 52 (1) (1996) 18-28. [2] A. Araki, Y. Sako, Determination of free and total homo cysteine in human plasma by high-performance liquid chro matography with fluorescence detection, J. Chromatogr. 442 (1987) 43-52. [3] L.B. Bailey, Folate nutrition in adolescents and adults, in: M.F. Picciano, E.L.R. Stockstad, J.F. Gregory (Eds.), Folic Acid Metabolism in Health and Disease, Wiley-Liss, New York, 1990, pp. 253-276. [4] B.C. Blount, M.M. Mack, C.M. Wehr, J.T. MacGregor, R.A. Hiatt, G. Wang, S.N. Wickramasinghe, R.B. Everson, B.N. Ames, Folate deficiency causes uracil misincorporation into human DNA and chromosome breakage: implications for cancer and neuronal damage, Proc. Natl. Acad. Sci. 94 (1997) 3290-3295. [5] J. Catalan, K. Autio, M. Wessman, C. Lindholm, S. Knuutila, M. Sorsa, H. Norppa, Age-associated micronuclei con taining centromeres and the X chromosome in lymphocytes of women, Cytogenet. Cell Genet. 68 (1995) 11-16. [6] L. Chasan-Taber, J. Selhub, I.H. Rosenberg, M.R. Malinow, P. Terry, P.V. Tishler, W. Willett, C.H. Hennekens, M.J. Stampfer, A prospective study of folate and vitamin B6 and risk of myocardial infarction in US physicians, J. of the American College of Nutrition 15 (2) (1996) 136-143. [7] R. Crellin, T. Bottiglieri, E.H. Reynolds, Folates and psychi atric disorders, Drugs 45 (5) (1993) 623-636. [8] D.A. Eastmond, J.D. Tucker, Identification of aneuploidy-inducing agents using cytokinesis-blocked human lymphocytes and an antikinetochore antibody, Environ. Mol. Mutagenesis 13 (1989) 34-43. [9] R.B. Everson, C.M. Wehr, G.L. Erexson, J.T. MacGregor, Association of marginal folate depletion with increased hu man chromosomal damage in vivo: demonstration by analy sis of micronucleated erythrocytes, J. Natl. Cancer Inst. 80 (1988) 525-529. [10] M. Fenech, Folate supplementation and genetic damage, Proceedings of 7th International Conference on Environmen tal Mutagens, Toulouse, France, 1997. [11] M. Fenech, A.A. Morley, Measurement of micronuclei in lymphocytes, Mutation Res. 147 (1985) 29-36. [12] M. Fenech, J. Rinaldi, The relationship between micronuclei in human lymphocytes and plasma levels of vitamin C, vitamin E, vitamin B12 and folic acid, Carcinogenesis 15 (7) (1994) 1405-1411. [13] M. Fenech, J. Rinaldi, A comparison of lymphocyte mi cronuclei and plasma micronutrients in vegetarians and non vegetarians, Carcinogenesis 16 (2) (1995) 223-230. [14] J.A. Heddle, M.C. Cimino, M. Hayashi, F. Romagna, M.D. Shelby, J.D. Tucker, P. Vanparys, J.T. MacGregor, Micronu clei as an index of cytogenetic damage: past, present and future, Environ. Mol. Mutagen. 18 (1991) 277-291. [15] V. Herbert, K. Das, Folic acid and vitamin B-12, in: M.E. Shils, J.A. Olson, M. Shike (Eds.), Modern Nutrition in Health and Disease, 8th edn., Vol. 1, Lea and Febiger, Philadelphia, PA, 1994, pp. 402-425. [16] R.A. Jacob, D.N. Gretz, P. Taylor, S.J. James, I.P. Pogribny, B.J. Miller, S.M. Henning, M.E. Swendseid, Moderate folate depletion increases plasma homocysteine and decreases lym phocyte DNA methylation in postmenopausal women, J. Nutr. 128 (1998) 1204-1212. [17] R.A. Jacob, M.M. Wu, S.M. Henning, M.E. Swendseid, Homocysteine increases as folate decreases in plasma of healthy men during short-term dietary folate and methyl group restriction, J. Nutr. 124 (1994) 1072-1080. [18] S.J. James, B.J. Miller, D.R. Cross, L.J. McGarrity, S.M. Morris, The essentiality of folate for the maintenance of deoxynucleotide precursor pools, DNA synthesis, and cell cycle progression in PHA-stimulated lymphocytes, Environ. Health Perspectives 101 (1993) 173-178. [19] E. Jennings, Folic acid as a cancer-preventing agent, Medical Hypotheses 45 (1995) 297-303. [20] G.D. Johnson, G.M. Norueira Araujo, A simple method of reducing the fading of immunofluorescence during mi croscopy, J. Immunol. Methods 43 (1981) 349-350. [21] B.A. Lashner, P.A. Heidenreich, G.L. Su, Effect of folate supplementation on the incidence of dysplasia and cancer in chronic ulcerative colitis, Gastroenterology 97 (1989) 225 229. [22] M.D. Lucock, J. Wild, C.J. Schorah, M.I. Levene, R. Hart ley, The methylfolate axis in neural tube defects: in vitro characterization and clinical investigation, Biochem. Med. Met. Biol. 52 (1994) 101-114. [23] J.T. MacGregor, Dietary factors affecting spontaneous chro mosome damage in man, Prog. Clin. Biol. Res. 347 (3) (1990) 139-153. [24] J.T. MacGregor, C.M. Wehr, R.A. Hiatt, B. Peters, J.D. Tucker, R.G. Langlois, R.A. Jacob, R.H. Jensen, J.W. Yager, M.K. Shigenaga, B. Frei, B.P. Eynon, B.N. Ames, `Sponta neous' genetic damage in man: evaluation of interindividual variability, relationship among markers of damage, and influ ence of nutritional status, Mutation Res. 377 (1997) 125-135. [25] J.B. Mason, Folate and colonic carcinogenesis: searching for a mechanistic understanding, J. Nutr. Biochem. 5 (1994) 170-175. [26] H. McNulty, Folate requirements for health in different population groups, British J. of Biomedical Science 52 (1995) 110-119. [27] L.E. Moore, A.H. Smith, C. Hopenhayn-Rich, M.L. Biggs, D.A. Kalman, M.T. Smith, Micronuclei in exfoliated bladder cells among individuals chronically exposed to arsenic in drinking water, Cancer Epi., Biomarkers and Prevention 6 (1997) 31-36. [28] National Research Council, Recommended Dietary Al lowances, 10th edn., National Academy Press, Washington, DC, 1989. 114 N. Titenko-Holland et al. / Mutation Research 417 (1998) 101-114 [29] H. Norppa, S. Luomahaara, H. Heikanen, S. Roth, M. Sorsa, L. Renzi, C. Lindholm, Micronucleus assay in lymphocytes as a tool to biomonitor human exposure to aneuploidogens and clastogens, Environ. Health Perspective 101 (1993) 139-- 143, Suppl. 3. [30] H. Norppa, L. Renzi, C. Lindholm, Detection of whole chromosomes in micronuclei of cytokinesis-blocked human lymphocytes by antikinetochore staining and in situ hy bridization, Mutagenesis 8 (1993) 519-525. [31] G.P. Oakley, M.J. Adams, C.M. Dickinson, More folic acid for everyone, now, J. Nutr. 126 (1996) 751-755. [32] C.A. O'Keefe, L.B. Bailey, E.A. Thomas, S.A. Hofler, B.A. Davis, J.J. Cerda, J.F. Gregory III, Controlled dietary folate affects folate status in nonpregnant women, J. Nutr. 125 (1995) 2717-2725. [33] C.J. Piyathilake, M. Macaluso, R.J. Hine et al., Cigarette smoking, intracellular vitamin deficiency, and occurrence of micronuclei in epithelial cells of the buccal mucosa, Cancer Epidem. Biomar. 4 (7) (1995) 751-758. [34] K.L. Radack, S.M. Pinney, G.K. Livingston, Sources of variability in the human lymphocyte micronucleus assay: a population-based study, Environ. Mol. Mutagenesis 26 (1) (1995) 26-36. [35] W.F. Rayburn, J.R. Stanley, M.E. Garrett, Periconceptional folate intake and neural tube defects, J. of the American College of Nutrition 15 (2) (1996) 121-125. [36] M.P. Rosin, Antigenotoxic activity of carotenoids in carcino gen-exposed populations, in: Antimutagenesis and Anticar cinogenesis Mechanisms II, Plenum, Vancouver, 1990, pp. 45-59. [37] M.P. Rosin, Genetic and proliferation markers in clinical studies of the premalignant process, Cancer Bull. (Houst.) 43 (1991) 507-514. [38] F. Sarto, R. Tomanin, L. Giacomelli, G. Iannini, A.R. Cupiraggi, The micronucleus assay in human exfoliated cells of the nose and mouth: application to occupational exposures to chromic acid and ethylene oxide, Mutation Res. 244 (1990) 345-351. [39] D. Smith, J. MacGregor, R. Hiatt, N. Hooper, C. Wehr, B. Peters, L. Goldman, L. Yuan, P. Smith, C. Becker, Micronucleated erythrocytes as an index of cytogenetic damage in humans: demographic and dietary factors associated with micronucleated erythrocytes in splenectomized subjects, Can cer Res. 50 (1990) 5049-5054. [40] H.F. Stich, M.P. Rosin, A.P. Hornby, B. Mathew, R. Sankaranarayanan, M.K. Nair, Remission of oral leuko plakias and micronuclei in tobacco betel quid chewers treated with beta-carotene and with beta-carotene plus vitamin-A, Int. J. of Cancer 42 (2) (1988) 195-199. [41] G.R. Sutherland, Heritable fragile sites on human chromo somes: I. Factors affecting expression in lymphocyte culture, Am. J. Hum. Genet. 31 (1979) 125-135. [42] N.W. Tietz (Ed.), Clinical Guide to Laboratory Tests, 3rd edn., Saunders, Philadelphia, PA, 1995. [43] N. Titenko-Holland, L. Moore, M.T. Smith, Measurement and characterization of micronuclei in exfoliated human cells by fluorescence in situ hybridization (FISH) with a cen tromeric probe, Mutation Res. 312 (1994) 39-50. [44] N. Titenko-Holland, A.J. Levine, M.T. Smith, P.J.E. Quin tana, M. Boeniger, R. Hayes, A. Suruda, P. Schulte, Quan tification of epithelial cell micronuclei by fluorescence in situ hybridization (FISH) in mortuary science students exposed to formaldehyde, Mutation Res. 371 (1996) 237-248. [45] N. Titenko-Holland, G. Windham, P. Kolachana, F. Reinisch, S. Parvatham, A.M. Osorio, M.T. Smith, Genotoxicity of malathion in human lymphocytes assessed using the mi cronucleus assay in vitro and in vivo: a study of malathionexposed workers, Mutation Res. 388 (1997) 85-95. [46] J.D. Tucker, R.J. Preston, Chromosome aberrations, mi cronuclei, aneuploidy, sister chromatid exchanges, and can cer risk assessment, Mutation Res. 365 (1996) 147-159. [47] M.B. van der Weyden, R.J. Hayman, I.S. Rose, J. Brumley, Folate-deficient human lymphoblasts: changes in deoxynucleotide metabolism and thymidylate cycle activities, Eur. J. Haematol. 47 (1991) 109-114. [48] N.J. Wald, A.K. Hackshaw, R. Stone, N. Sourial, Blood folic acid and vitamin B12 in relation to neural tube defects, British J. of Ob. and Gyn. 103 (1996) 319-324.