Vitamins
Relative Folate Bioavailability from Diets Containing Human, Bovine and Goat Milk12 ! NANCY SWIATLO, DEBORAH L O'CONNOR,3 JENNIFER ANDREWS AND MARY FRANCES PICCIANO* School of Human Resources and Family Studies and Diuision of Nutritional Sciences, College of Agriculture, University of Illinois at Urbana-Champaign, Urbana, IL 61801
amount of folate ingested from milk but not with the total amount of folate in the diet. This strong correlation between blood folate levels and milk folate intake sug gests that folate present in milk, whether from human milk, bovine milk or bovine milk-based formula, is more available to the infant than is folate present in other foods. Review of the current literature reveals little infor mation on the constituents of milk that may influence the absorption of folate from milk or diets containing milk. Milk folate binding proteins (MFBP) appear to influence absorption of bound folate by a process dis tinct from absorption of unbound folate in very young neonatal animals (5-10). In view of the importance of milk in folate nutriture of the infant, the present study was designed to assess the relative bioavailability of folate from diets containing human, bovine, and goat milk. This study was also designed to evaluate the relative predictive value of plasma, RBC, liver and kid ney folate concentrations as determinants of folate bioavailability.
ABSTRACT The present study was designed to de termine the relative folate bioavailability from diets containing human, bovine or goat milk and the relative sensitivity of various response criteria used in assess ing folate bioavailability. Following a 12-wk depletion period, 16 groups of male rats (n = 5/group) were fed experimental diets with or without 20% milk solids and graded levels of folie acid for 4 wk. Total foiates were measured in plasma, erythrocytes, liver and kidney. Bioavailability of dietary folate was determined using slope-ratio statistics. Plasma response was found to be the most sensitive indicator of folate bioavailability based on steepness of slope, goodness of fit (r = 0.96, P< 0.01) and linearity of response over the entire range of folate intakes. Kidney folate concentration also showed a significant linear relationship to total folate intake (r = 0.69, P < 0.01). Liver and erythrocyte folate concentrations were not correlated with folate intake (r = 0.33 and r = 0.22, respectively). Using plasma folate as the response criterion, dietary incorporation of human milk significantly enhanced folate bioavailability by 75% (P < O.O1). With kidney as the response tissue, folate bioavailability from diets containing human and bovine milk was significantly enhanced over milk-free diets. These results show that incorporation of human or bovine milk into diets en hances folate bioavailability and that plasma and kid ney folate concentrations are sensitive and specific indicators of folate bioavailability. J. Nutr. 120:172177, 1990.
MATERIALS AND METHODS Animal care. One hundred male weanling SpragueDawley rats (HarÃ-anIndustries, Indianapolis, IN) weigh ing approximately 100 g were housed individually in
INDEXING KEY WORDS:
•folate bioauailability
•milk •infancy
'Supported in part by U.S.D.A. Competitive
Grant A84-CRCR-1-
1493 and the Illinois Agriculture Experiment Station. Presented in part at the 72nd Annual Meeting of the Federation of the American Societies for Experimental Biology, Las Vegas, NV, May 1988 [SWIATLO, N. L. & PICCIANO,M. F. (1988) Relative folate bioavailability from human, bovine and goat milk containing diets. FASEB I. 2 (5):A1087]. 3Present address: Division of Applied Human Nutrition, Univer sity of Guelph, Guelph, Ontario, Canada NIG 2W1. 4To whom reprint requests should be sent. Present address: Depart
Provision of an adequate supply of folate to the infant is essential to facilitate the rapid rate of growth that occurs during this stage of the life cycle. The high dietary requirement for folate among infants is evident by the frequent occurrence of megaloblastic anemia among infants of both low and normal birth weights (1^3). Smith et al. (4) reported that infant folate status as determined by serum and red blood cell (RBC) folate concentrations is significantly correlated with the total
ment of Nutrition, 126 Henderson Building South, The Pennsylvania State University, University Park, PA 16802. 0022-3166/90 $3.00 ©1990 American Institute of Nutrition. Received 30 fune 1989. Accepted 25 September 1989. 172
Downloaded from https://academic.oup.com/jn/article-abstract/120/2/172/4743956 by University of Leeds Library user on 18 June 2018
173
FOLATE BIOAVAILABILITY
stainless steel wire cages with controlled temperature (20-22°C)and lighting (12-h light cycle). Rats were fed a folate-deficient, semipurified diet (66 ug/kg) for 12 wk to deplete body stores of folate. Concurrently, a separate group of rats was fed a semipurified diet containing adequate folie acid (1400 |ig/kg) to provide the NRC (11) requirement for folate and to assess the outcome of the depletion diet. Semipurified diets and demineralized water (Nanopure, Barnstead, Boston, MA) were supplied ad libitum. After 12 wk, four animals from each of the two groups were killed for determination of baseline plasma, RBC and liver folate concentrations. Remaining folate-depleted animals were then randomly allocated to 1 of 16 experimental diets. Four groups (n = 5 rats/ group) received milk-free (MF) diets with graded levels of folate; the remaining groups (12) received one of three milk-containing (MC) diets with graded levels of folate. Diets. All diets were formulated according to the AIN-76A diet (12) (Table 1). The human, bovine and goat milk incorporated into rat diets was obtained locally. Human milk (38.5 L) was collected from 19 apparently healthy, lactating women in the Champaign-Urbana, IL area at several nursings per day and at various stages during lactation. After each nursing, human milk sam ples were frozen. Immediately prior to freeze-drying, milk samples from all mothers were pooled. Bovine milk (39 L) was collected and pooled from five cows (University Dairy Herd, Department of Animal Sci ences, University of Illinois) after morning and evening milkings. Goat milk (32.3 L) was collected and pooled from seven goats (Agan's Caprine Farm, Sidney, IL) after morning and evening milkings. To protect labile folates, all milk was collected in opaque containers and stored at -70°Cuntil it was freeze-dried. Following proximate analysis of milk solids (Table 1), dietary constituents providing the protein (casein), carbohydrate (cornstarch/ sucrose) and fat (corn oil) fraction of MC diets were adjusted to ensure that all experimental diets were isocaloric and isonitrogenous. The analyzed mean folate concentration of human, bovine and goat milk solids averaged 354.4, 400.5 and 51.5 ng/g, respectively. Folate as pteroylglutamic acid was added to MF and MC diets to obtain 0, 200, 400 or 600 ug folie acid/kg diet. Tissue and blood collection. After the depletion and experimental periods, a portion of the whole blood (100 HL) collected by cardiac puncture was diluted 10 times with 0.1 M phosphate buffer (pH 7.0, 2% ascorbate). The remainder was centrifuged (800 x g) and the plasma was separated and stored with 1% ascorbate. Liver and kid neys were removed, rinsed with deionized water, blot ted dry and weighed. Tissues and blood samples were immediately frozen and stored at -70°Cuntil analyzed. folate analyses. Liver, kidney, plasma, RBC, lyophilized milk and dietary folate concentrations were de termined using the test microorganism Lactobacillus casei(l3}. Liver (2 g), kidney (-1 g), dietfl g) and milk (1 g) samples were diluted in 0.1 M phosphate buffer (pH Downloaded from https://academic.oup.com/jn/article-abstract/120/2/172/4743956 by University of Leeds Library user on 18 June 2018
TABLE1 Composition of depletion and experimental diets Depletion diet
Ingredient
Milk-free diet
Milkcontaining diet
g/kg solids'Casein2D,L-Methionine3Cornstarch4Sucrose5Fiber6Corn Milk
oil7AIN-76A mix(without vitamin acid)8AIN-76A folie mix9Addedsalt pteroylglutamic 'Incorporation
acid10—2005300300501001035——2005300300501001035V
of milk solids at 200 g/kg of diet provided the
following: human milk: 2.6 g total N, 129.7 g carbohydrate, 43.4 g fat, 70.9 ug total folate; bovine milk: 7.2 g total N, 90.7 g carbohydrate, 42.4 g fat, 80.1 ug total folate; and goat milk: 7.6 total N, 61.3 carbohydrate, 74.6 g fat, 10.4 ug total folate. 2Vitamin-free test casein, Teklad, Madison, WI. 'Sigma Chemical, St. Louis, MO. 4Staley Manufacturing, Decatur, IL. 5Amstar, New York. "Alphacel, ICN Biochemicals, Cleveland, OH 'Anderson Clayton Foods, Dallas, TX. "No. V14901 (Research Diets, New Brunswick, NJ). Composition of vitamin mix (g/kg of vitamin mix) was as follows: menadione sodium bisulfite, 0.08; biotin 1.0%, 2.0; cyanocobalamin 0.1%, 1.0; niacin, 3.0; calcium pantothenate, 1.6; pyridoxine-HCl, 0.7; riboflavin, 0.6; thiamin-HCl, 0.6; sucrose, 979-37, Supplied per kilogram of vita min mix: vitamin A palmitate, 400,000 iu; vitamin D-3, 100,000 iu; vitamin E acetate, 5,000 ru. 9No. S10001 (Research Diets, New Brunswick, NJ). Composition of mineral mix (g/kg of mineral mix) was as follows: CaHPO4, dibasic, 500.0; MgO, 24.0; potassium citrate, monohydrate, 220.0; K2SO4, 52.0; NaCl, 74.0; Crk(SO4)2, 0.55; cupric carbonate, 0.3; KIO3, 0.01; ferric citrate, 6.0; MnCO3, 3.5; Na2SeO-5H2O, 0.01; ZnCO3, 1.6; sucrose, 118.03. '"Pteroylglutamic acid (Sigma Chemical, St. Louis, MO) was added to experimental diets at either 0, 200, 400 or 600 ng/kg. A control diet was made by adding 1.40 mg pteroylglutamic acid per kg depletion diet.
7.0, 2% ascorbate). Diet, liver and kidney samples were homogenized (Brinkman Instruments, Westbury, NY), boiled and immediately cooled on ice. Milk samples were autoclaved (121°C),cooled and centrifuged (2,000 x g) at 4°C.Fat was subsequently removed by aspiration. Folate concentrations of the resultant supernatant were determined following incubation with conjugase. Par tially purified conjugase from desiccated chicken pan creas was prepared according to a method previously described (14). Folate concentrations in plasma and RBC were deter mined as described by Scott et al. (15) and Cooperman (16) with modifications. Plasma folate concentrations
174
SWIATLO ET AL.
„ TABLE 2
Food intake, weight gain, blood folate concentration and liver folate concentration of rats fed either depletion or control diet during a 12-wk period'-2 Measure g/wkWeight Food intake, g/wkFinal gain, gPlasma weight, nmoI/LRed folate, nmol/LLiver blood cell folate, folate, nmol/g113.71
Depletion diet3
Control diet4
30-
z
20
i
10
2.0-23.1
3.822.7 ± 2.6378.2 ± 2.5379.0 ± 6.310.2 ± 12.982.0+ ± 2.0'379.1129.2-12.9 ± 5.4988.7+31.020.4 ± 1.6125.2
± 5.7
'Values are means ±SEMfor 3-4 rats. 2'Denotes statistically significant differences between dietary treatment groups at the end of the 12-wk period (P < 0.05]. '66 ng Folate/kg diet. "1.4 mg Folate/kg diet.
were determined following incubation of samples at room temperature for 15 min. To ensure lysis of RBC, whole blood was diluted in 0.005 Mphosphate buffer (pH 6.4,0.15% ascorbate) and subjected to ultrasound (Sonifier, Heat Systems-Ultrasonics, Plainview, NY). Samples were then autoclaved, cooled and centrifuged (1000 x g). The resultant supernatant was then assayed for folate concentration. Statistical analyses. Unavailability of folate from experimental diets was determined using slope-ratio statistics (17). Unavailability estimates were deter mined by the multiple linear regression model when dose-response curves shared a common y-intercept. Otherwise, the simple linear regression model was uti lized. The regression equation consisted of an indepen dent variable (x); total folate intake,- and one of four dependent variables (y), either plasma, RBC, liver or kidney folate concentrations. Differences between slopes of the regression lines and calculated folate bioavailabilities were determined by Student's i-test. Cor relation coefficients were assessed for significance using Pearson's test (18).
RESULTS Status of rats following a 12-wk depletion period. Food consumption was significantly lower in rats fed the folate-deficient diet than in rats fed the control diet (Table 2). The difference in food intake was found to have no effect on final weight of folate-deficient rats compared to control animals. Feeding the folate-defi cient diet for 12 wk resulted in significant decreases in both mean plasma and RBC folate concentrations. Mean plasma and RBC folate concentrations of folate-depleted rats were 12% and 38% of those in control rats at the end of the depletion period, respectively (P < 0.05). A similar but statistically insignificant reduction in mean Downloaded from https://academic.oup.com/jn/article-abstract/120/2/172/4743956 by University of Leeds Library user on 18 June 2018
• 100
200
300
y «3.3 + O.OOSx
5-
n=0.69
0
100
200
300
Total Folate Intake (|ig) for a 4 Week Period
FIGURE 1 Dose-response curves for plasma and kidney folate concentrations of rats as a function of dietary folate intake from milk-free diets with graded levels of dietary folate. Correlation coefficients are significant at P < 0.01.
liver folate concentration was noted for folate-deficient compared to control animals at the end of the depletion period. Status of rats following a 4-wk experimental period. Dietary folate concentration and the presence of milk in diets had no effect on either weekly food consump tion or weight gain. Mean (±SEM)body weights (g) and food intakes (g/wk) of all rats were 383.3 ±2.2 and 121.2 ±0.6, respectively, at the end of the depletion period and were 418.6 ±2.6 and 135.0 ±4.8 at the end of the experimental period. Mean liver and kidney weights were also similar among treatment groups throughout the study. Effectiveness of blood and tissue folate concentra tions as indicators of folate bioavailability. The con centrations of folate in plasma, RBC, liver and kidney were regressed against total folate consumption of rats fed milk-free diets with graded levels of folie acid for 4 wk. Relative effectiveness of the above indices as mea sures of folate bioavailability were determined by lin earity over the range of folate intakes, steepness of the slope, and goodness of fit. Plasma folate concentrations were significantly correlated with total folate intake from milk-free diets (r = 0.96, P < 0.01; Fig. 1). Kidney folate concentrations showed a less sensitive but linear relationship to total folate intake (r = 0.69, P < 0.01; Fig. 1). In contrast, neither liver (r = 0.33) nor RBC (r = 0.22) folate concentrations were significantly related to total
175
FOLATE BIOAVAILABILITY
TABLE 3
y «7.48 + 0.161 r=0.67
60 -\
y =7.48 + 0.11x r=0.82 y = 7.48+ 0.09x
Relative bioavailability of folate from milk-free and milk-containing diets1 y . 7.48 + 0.07»
Milk-containing
1=0.66
diets
W*1.0"Humanmilk1.75C3.0CBovinemilk1.25a2.0bGoatmilk0.75b1.5ab PlasmaKidneyMilk-freedietl. 'Relative bioavailabilities were calculated using the slopes from the regression lines of the milk-free diet and the milk-containing (human, bovine, goat| diets. Values in each row with unlike super scripts are significantly different at P < 0.01.
100
200
300
400
y = 3.85 + 0.014l i»0.80 y = 3.85+ 0.009X
MM4
folate intake during the 4-wk experimental period. Relative bioavailability of folate from milk-free and milk-containing diets using slope-ratio statistics. The results of slope-ratio statistical analyses using plasma and kidney folate concentrations are presented in Table 3. Using either plasma or kidney folate concentration as the response variable (Fig.2), no statistically significant differences existed between the y-intercepts of the doseresponse curves. Therefore, bioavailability estimates were determined using the multiple linear regression model. Incorporation of human milk into the diet en hanced folate bioavailability. The relative bioavailabil ity of folate from human MC diets was 40% and 133% greater than that from bovine and goat MC diets, respec tively (P < 0.01). The folate bioavailability from bovine MC diets was 67% greater than the bioavailability of folate from goat MC diets (P < 0.01). Relative bioavaila bility of folate from bovine and goat MC diets did not differ significantly from that of milk-free diets. Using kidney folate concentration as the criterion of response (Fig. 2), folate bioavailability from diets con taining human milk was significantly enhanced (200%) compared to diets without milk (P < 0.01). The bio availability of folate from bovine MC diets was also elevated (100%) compared with milk-free diets (P for a 4 Week Period
FIGURE 2 Dose-response curves for plasma and kidney folate concentrations of rats from experimental diets with graded levels of dietary folate and containing either no milk solids (MF) or solids from human (HM), bovine (BM), and goat milk (GM).
availability of folate from human and bovine milk-con taining diets was significantly greater than that from milk-free diets. Folate bioavailability from human MC diets was also found to be superior to that from bovine and goat MC diets using either plasma or kidney folate concentration as the measure of bioavailability. Neither the bioavailability of folate from various milks nor the impact of various milks on the bioavaila bility of dietary folate has been widely investigated. In human infants, blood folates were linearly related to folate from milk but not to total dietary folate intake (4). The data from studies with humans suggest that folate present in various milks is more bioavailable than that in other infant foods, whereas the data from the present study indicate that human or bovine milks in the diet are capable of increasing total folate bioavaila bility under certain circumstances. Collectively, these data stress the importance of milk to folate nutrition of infants. The component or components of milk that have an impact on folate bioavailability are not well established. Considerable evidence exists to indicate that milk folate
176
SWIATLO ET AL.
binding proteins (MFBP) contribute to the absorption and/or retention of dietary folate in newborns (5-10). To determine whether concentration of MFBP in milk sol ids might relate to differences in the bioavailability of folate from milk-containing diets, the folate binding capacity of lyophilized human, bovine and goat milk was measured (19, 20). Endogenous folates were disso ciated from MFBP by acid before measurement of folate binding activity, so values reported herein reflect the total, not the unsaturated, folate binding capacity of human, bovine and goat milk solids. Mean total folate binding capacity of human milk solids (1.85 ±0.16 umol folie acid bound/kg milk solids) and bovine milk solids (2.28 ±1.35 nmol/kg) did not differ significantly. In contrast, the mean folate binding capacity of goat milk solids (4.48 ±1.98 umol/kg) was significantly greater than that of either human or bovine milk solids. Suffi cient binding protein was present in human, bovine and goat milk solids to bind all folate endogenous to milk and an additional 92.3, 121.5 and 384.6 ng folate per gram of diet, respectively. Therefore, in diets containing human, bovine and goat milk formulated to contain 600 ug/kg folate, for example, sufficient folate binding pro tein was present to bind approximately 15, 21 and 64% of total dietary folate, respectively. The role of MFBP in the absorption and retention of dietary folate appears beneficial. Ford (6) postulated that MFBP indirectly promote folate absorption by making dietary folate unavailable to folate-requiring bacteria. In the presence of goat colostrum, uptake of radiolabeled folie acid by folate-requiring strains of bacteria was considerably reduced compared to uptake of radiola beled folie acid in the absence of colostrum. Tani and Iwai (8) also have shown that uptake of folie acid by folate-requiring intestinal microorganisms was consid erably reduced when it was bound to bovine MFBP. Colman et al. (7) suggested a more direct role for MFBP in folate absorption. The uptake of folate by isolated rat intestinal cells was significantly greater when it was bound to the protein in goat and human milk than when it was unbound. In contrast, Tani and Iwai (8) and Said et al. (9) reported that folate bound to either bovine or human MFBP is absorbed more slowly than unbound folate. Tani and Iwai demonstrated that despite a slower rate of absorption the total amount of folate absorbed remained constant. They also showed that urinary fo late excretion is elevated when unbound folate versus bound folate is administered to rats via oral intubation, suggesting that MFBP enhance body retention and bioavailability of folate. Mason and Selhub (10) have dem onstrated differences in the absorption process of folie acid bound to rat MFBP and that of free folie acid in suckling rats. They suggested that these separate modes of transport for folate may be advantageous to suckling rats. It is appropriate to note here that the mechanism by which folylpolyglutamates (80% of dietary folate) are absorbed differs between human beings and rats (21). Downloaded from https://academic.oup.com/jn/article-abstract/120/2/172/4743956 by University of Leeds Library user on 18 June 2018
Thus, extending our findings in rats to humans is tenu ous, particularly under circumstances in which folate conjugase activity is the rate-limiting step in folate absorption. Assuming that MFBP do enhance the bioavailability of dietary folate, the bioavailability of folate from diets containing goat milk would be expected to be greater than that from diets containing human and bovine milk. In this study, however, the presence of goat milk solids did not enhance bioavailability of folate, suggesting a more complex relationship between folate and MFBP. In contrast, human and bovine milk solids in the diet did enhance bioavailability of folate. An intrinsic factor other than MFBP may be present in human and bovine milk that accounts for the increased bioavailability of folate from diets containing these milks. There is disagreement in the literature concerning the appropriate response tissue for measuring folate bioavailability (22-25). In this investigation, plasma fo late concentration was the most sensitive and specific measure of folate bioavailability as indicated by linear ity of response, steepness of slope and goodness of fit. Kidney folate concentration was likewise a good mea sure of available dietary folate. However, RBC and liver folate concentrations were not linearly related to total folate intake during the repletion period. Liver is often used as the response variable in the folate bioassay because it is believed to be the storage tissue for this nutrient and to reflect the cumulative effect of dietary folate intake over time (21). Abad and Gregory (23) previously reported that hepatic folate con centrations are inconsistent as a response criterion in the folate bioassay, but they found that fasting levels of plasma folate were reliable indicators of dietary folate concentration. Under seemingly similar experimental conditions, others reported (22, 24) that liver folate concentration is a good response variable, as indicated by a significant and positive correlation between hepatic folate and dietary folate concentrations. As previously discussed by Abad and Gregory (23) the concentration of folate in depletion diets does not ac count for these apparent discrepancies. The concentra tion of folate in depletion diets used in the present study, 66 ug/kg diet, was sightly lower than that used by Hoppner and Lampi (24), 75 ug/kg diet. Variation in the folate status of rats at the end of the depletion period in the present study may account for the inconsistent response of hepatic folate concentration to total folate ingestion. After 5 wk of folate depletion, Keagy and Oace (22) found mean liver folate concentrations in rats to be approximately half that of rats in the present investiga tion, which were depleted for 12 wk. Keagy and Oace (22) indicated depletion diets when rats weighed 43-73 g. In our investigation, the mean rat weight at the beginning of the depletion period was 100 g. By 8 wk of depletion, large weekly weight gains ceased, indicating that growth (accretion of lean body mass) had neared
FOLATE BIOAVAILABILITY
completion. Because liver folate concentration is dra matically affected by rate of growth, it is conceivable that the greater liver folate concentrations at the end of the depletion period in this study reflected a decreased requirement for folate among adult rats. Kidney folate concentrations are less frequently mea sured in bioavailability studies. Tigner and Roe (25) found a significant correlation between kidney folate concentrations and dietary intake of folate. There is ample evidence that the kidney is involved in folate homeostasis by acting as a storage site (26), and the tissue may be a sensitive and reliable indicator for measuring folate bioavailability. In summary, human and bovine milk incorporated into rat diets enhanced total folate bioavailability. The constituent in milk that enhances folate bioavailability is unknown. Though previous investigations have indi cated that there are beneficial effects of MFBP on folate absorption and retention, the concentrations of MFBP present in human, bovine and goat milk are not reflec tive of their ability to enhance dietary folate bioavaila bility in the rat. In this study, plasma and kidney were found to be sensitive indicators of dietary folate bio availability. Because there is disagreement among inves tigators concerning which tissue is the most reliable indicator of folate bioavailability in the rat bioassay, it is advisable to use more than one tissue in future stud ies. The experimental circumstances accounting for these reported differences in tissue responses to dietary folate need to be defined.
ACKNOWLEDGMENTS The authors wish to express their extreme grateful ness to the lactating women who collected the large quantities of human milk required in this study. Appre ciation also is expressed to Jimmy Clark from the De partment of Animal Sciences for providing the bovine milk and to Don Angus for the goat milk. The assistance of the Department of Food Science for lyophilizing milk samples and of Don Wade for animal care is acknowl edged.
LITERATURE CITED 1. GRAY,O. P. & BUTLER,E. B. (1965) Megaloblastic anaemia in premature infants. Arch. Dis. Child. 40: 53-58. 2. STRELLING, M. K., BLACKLEDGE, G. D., GOODALL,H. B. & WALKER, C. H. M. (1966) Megaloblastic anaemia and whole blood folate levels in premature infants. Lancet 1: 898-900. 3. BECROFT,D. M. O. & HOLLAND,J. T. (1966) Goat's milk and megaloblastic anemia in infancy. N. Z. Med. f. 65: 303-307.
Downloaded from https://academic.oup.com/jn/article-abstract/120/2/172/4743956 by University of Leeds Library user on 18 June 2018
177
4. SMITH,A. M., PICCIANO,M. F. &. DEERING,R. H. (1985| Folate intake and blood concentrations of term infants. Am. J. Clin. Nutr. 41: 590-598. 5. FORD,J. E., KNAGGS,G. S., SALTER,D. N. & SCOTT,K. J. (1972) Folate nutrition in the kid. Bi. J. Nutr. 27: 571-583. 6. FORD,J.E. (1974) Some observations on the possible nutritional significance of vitamin B-12 and folate-binding proteins in milk. Br.f. Nutr. 31:243-257. 7. COLMAN,N., HETTIARACHY, N. & HERBERT,V. (1981) Detection of a milk factor that facilitates folate uptake by intestinal cells. Science 211: 1427-1429. 8. TANT,M. & IWAI,K. (1984) Some nutritional effects of folatebinding protein in bovine milk on the bioavailability of folate to rats./. Nutr. 114: 778-785. 9. SAID,H. M., HÖRNE, D. W. & WAGNER,C. (1986) Effect of human folate binding protein on folate intestinal transport. Archiv. Biochem. Biophys. 251: 114-120. 10. MASON,J. B. & SELHUB,J. (1988) Folate-binding protein and the absorption of folie acid in the small intestine. Am. ]. Clin. Nutr. 48: 620-625. 11. NATIONALRESEARCH COUNCIL (1978) Nutrient requirements of laboratory animals, No. 10, 3rd rev. ed., National Academy of Sciences, Washington, DC. 12. AMERICANINSTITUTEOFNUTRITION(1980) Second report of the Ad Hoc Committee on standards for nutritional studies. /. Nutr. 110: 1726. 13. KOCHANOWSKI, B. A., SMITH,A. M., PICCIANO,M. F. & SHERMAN, A. R. (1983) Folate depletion secondary to iron deficiency in the neonatal rat. J. Nutr. 113: 2471-2478. 14. KEAGY,P. M. (1984) Folacin-microbiological and animal assays. In: Methods of Vitamin Assay (Augustin, J., Klein, B. P., Becker, D. & Venugopal, P. B., eds.), 4th éd., pp. 445^171, Wiley-Interscience, New York. 15. SCOTT, J. M., CHANTA,V. & HERBERT,V. (1974) Trouble-free microbiologie serum and red cell folate assays. Am. /. Med. Techno!. 40: 125-134. 16. COOPERMAN,J. M. (1971) Microbiological assay of folie acid activity in serum and whole blood. Methods Enzymol. 15: 629642. 17. FINNEY,D. J. (1978) Statistical Method in Biological Assay, 3rd ed., Macmillan, New York. 18. SAS INSTITUTE(1985) SAS User's Guide: Statistics. 5th éd. SAS (Statistical Analysis System) Institute, Cary, NC. 19. BROWN,C. M. (1985) The forms of human milk folacin and variation patterns and relative bioavailability of folacin added to diets containing cow, goat and human milk to rats. MS thesis, University of Illinois, Urbana-Champaign. 20. SELHUB,J., ARNOLD,R., SMITH,A. M. & PICCIANO,M. F. (1984) Milk folate protein (FBP):A secretory protein for folate. Nutr. Res. 4: 181-187. 21. WANG, T. T Y, REISENAUER, A. M. & HALSTED,C. H. (1985) Comparison of folate conjugase activities in human, pig, rat and monkey intestine. /. Nutr. 115: 814-819. 22. KEAGY,P. M. & OACE,S. M. (1982) Development of a folacin bioassay in rats. /. Nutr. 112: 87-91. 23. ABAD,A. &. GREGORY,J. F. (1987) Determination of folate bio availability with a rat bioassay. /. Nutr. 117: 866-873. 24. HOPPNER,K. & LAMPI,B. (1985) Liver folate as a response param eter for folate bioassay in rats. Nutr. Rep. Intern. 31: 203-211. 25. TIGNER,]. & ROE, D. A. (1979) Tissue folacin stores in rats measured by radioassay. Proc. Soc. Exp. Biol. Med. 160: 445-448. 26. STEINBERG, S. E. (1984) Mechanisms of folate homeostasis. Am. /. Physiol. G319-324.
Relative Folate Bioavailability from Diets Containing Human, Bovine and Goat Milk12 ! NANCY SWIATLO, DEBORAH L O'CONNOR,3 JENNIFER ANDREWS AND MARY FRANCES PICCIANO* School of Human Resources and Family Studies and Diuision of Nutritional Sciences, College of Agriculture, University of Illinois at Urbana-Champaign, Urbana, IL 61801
amount of folate ingested from milk but not with the total amount of folate in the diet. This strong correlation between blood folate levels and milk folate intake sug gests that folate present in milk, whether from human milk, bovine milk or bovine milk-based formula, is more available to the infant than is folate present in other foods. Review of the current literature reveals little infor mation on the constituents of milk that may influence the absorption of folate from milk or diets containing milk. Milk folate binding proteins (MFBP) appear to influence absorption of bound folate by a process dis tinct from absorption of unbound folate in very young neonatal animals (5-10). In view of the importance of milk in folate nutriture of the infant, the present study was designed to assess the relative bioavailability of folate from diets containing human, bovine, and goat milk. This study was also designed to evaluate the relative predictive value of plasma, RBC, liver and kid ney folate concentrations as determinants of folate bioavailability.
ABSTRACT The present study was designed to de termine the relative folate bioavailability from diets containing human, bovine or goat milk and the relative sensitivity of various response criteria used in assess ing folate bioavailability. Following a 12-wk depletion period, 16 groups of male rats (n = 5/group) were fed experimental diets with or without 20% milk solids and graded levels of folie acid for 4 wk. Total foiates were measured in plasma, erythrocytes, liver and kidney. Bioavailability of dietary folate was determined using slope-ratio statistics. Plasma response was found to be the most sensitive indicator of folate bioavailability based on steepness of slope, goodness of fit (r = 0.96, P< 0.01) and linearity of response over the entire range of folate intakes. Kidney folate concentration also showed a significant linear relationship to total folate intake (r = 0.69, P < 0.01). Liver and erythrocyte folate concentrations were not correlated with folate intake (r = 0.33 and r = 0.22, respectively). Using plasma folate as the response criterion, dietary incorporation of human milk significantly enhanced folate bioavailability by 75% (P < O.O1). With kidney as the response tissue, folate bioavailability from diets containing human and bovine milk was significantly enhanced over milk-free diets. These results show that incorporation of human or bovine milk into diets en hances folate bioavailability and that plasma and kid ney folate concentrations are sensitive and specific indicators of folate bioavailability. J. Nutr. 120:172177, 1990.
MATERIALS AND METHODS Animal care. One hundred male weanling SpragueDawley rats (HarÃ-anIndustries, Indianapolis, IN) weigh ing approximately 100 g were housed individually in
INDEXING KEY WORDS:
•folate bioauailability
•milk •infancy
'Supported in part by U.S.D.A. Competitive
Grant A84-CRCR-1-
1493 and the Illinois Agriculture Experiment Station. Presented in part at the 72nd Annual Meeting of the Federation of the American Societies for Experimental Biology, Las Vegas, NV, May 1988 [SWIATLO, N. L. & PICCIANO,M. F. (1988) Relative folate bioavailability from human, bovine and goat milk containing diets. FASEB I. 2 (5):A1087]. 3Present address: Division of Applied Human Nutrition, Univer sity of Guelph, Guelph, Ontario, Canada NIG 2W1. 4To whom reprint requests should be sent. Present address: Depart
Provision of an adequate supply of folate to the infant is essential to facilitate the rapid rate of growth that occurs during this stage of the life cycle. The high dietary requirement for folate among infants is evident by the frequent occurrence of megaloblastic anemia among infants of both low and normal birth weights (1^3). Smith et al. (4) reported that infant folate status as determined by serum and red blood cell (RBC) folate concentrations is significantly correlated with the total
ment of Nutrition, 126 Henderson Building South, The Pennsylvania State University, University Park, PA 16802. 0022-3166/90 $3.00 ©1990 American Institute of Nutrition. Received 30 fune 1989. Accepted 25 September 1989. 172
Downloaded from https://academic.oup.com/jn/article-abstract/120/2/172/4743956 by University of Leeds Library user on 18 June 2018
173
FOLATE BIOAVAILABILITY
stainless steel wire cages with controlled temperature (20-22°C)and lighting (12-h light cycle). Rats were fed a folate-deficient, semipurified diet (66 ug/kg) for 12 wk to deplete body stores of folate. Concurrently, a separate group of rats was fed a semipurified diet containing adequate folie acid (1400 |ig/kg) to provide the NRC (11) requirement for folate and to assess the outcome of the depletion diet. Semipurified diets and demineralized water (Nanopure, Barnstead, Boston, MA) were supplied ad libitum. After 12 wk, four animals from each of the two groups were killed for determination of baseline plasma, RBC and liver folate concentrations. Remaining folate-depleted animals were then randomly allocated to 1 of 16 experimental diets. Four groups (n = 5 rats/ group) received milk-free (MF) diets with graded levels of folate; the remaining groups (12) received one of three milk-containing (MC) diets with graded levels of folate. Diets. All diets were formulated according to the AIN-76A diet (12) (Table 1). The human, bovine and goat milk incorporated into rat diets was obtained locally. Human milk (38.5 L) was collected from 19 apparently healthy, lactating women in the Champaign-Urbana, IL area at several nursings per day and at various stages during lactation. After each nursing, human milk sam ples were frozen. Immediately prior to freeze-drying, milk samples from all mothers were pooled. Bovine milk (39 L) was collected and pooled from five cows (University Dairy Herd, Department of Animal Sci ences, University of Illinois) after morning and evening milkings. Goat milk (32.3 L) was collected and pooled from seven goats (Agan's Caprine Farm, Sidney, IL) after morning and evening milkings. To protect labile folates, all milk was collected in opaque containers and stored at -70°Cuntil it was freeze-dried. Following proximate analysis of milk solids (Table 1), dietary constituents providing the protein (casein), carbohydrate (cornstarch/ sucrose) and fat (corn oil) fraction of MC diets were adjusted to ensure that all experimental diets were isocaloric and isonitrogenous. The analyzed mean folate concentration of human, bovine and goat milk solids averaged 354.4, 400.5 and 51.5 ng/g, respectively. Folate as pteroylglutamic acid was added to MF and MC diets to obtain 0, 200, 400 or 600 ug folie acid/kg diet. Tissue and blood collection. After the depletion and experimental periods, a portion of the whole blood (100 HL) collected by cardiac puncture was diluted 10 times with 0.1 M phosphate buffer (pH 7.0, 2% ascorbate). The remainder was centrifuged (800 x g) and the plasma was separated and stored with 1% ascorbate. Liver and kid neys were removed, rinsed with deionized water, blot ted dry and weighed. Tissues and blood samples were immediately frozen and stored at -70°Cuntil analyzed. folate analyses. Liver, kidney, plasma, RBC, lyophilized milk and dietary folate concentrations were de termined using the test microorganism Lactobacillus casei(l3}. Liver (2 g), kidney (-1 g), dietfl g) and milk (1 g) samples were diluted in 0.1 M phosphate buffer (pH Downloaded from https://academic.oup.com/jn/article-abstract/120/2/172/4743956 by University of Leeds Library user on 18 June 2018
TABLE1 Composition of depletion and experimental diets Depletion diet
Ingredient
Milk-free diet
Milkcontaining diet
g/kg solids'Casein2D,L-Methionine3Cornstarch4Sucrose5Fiber6Corn Milk
oil7AIN-76A mix(without vitamin acid)8AIN-76A folie mix9Addedsalt pteroylglutamic 'Incorporation
acid10—2005300300501001035——2005300300501001035V
of milk solids at 200 g/kg of diet provided the
following: human milk: 2.6 g total N, 129.7 g carbohydrate, 43.4 g fat, 70.9 ug total folate; bovine milk: 7.2 g total N, 90.7 g carbohydrate, 42.4 g fat, 80.1 ug total folate; and goat milk: 7.6 total N, 61.3 carbohydrate, 74.6 g fat, 10.4 ug total folate. 2Vitamin-free test casein, Teklad, Madison, WI. 'Sigma Chemical, St. Louis, MO. 4Staley Manufacturing, Decatur, IL. 5Amstar, New York. "Alphacel, ICN Biochemicals, Cleveland, OH 'Anderson Clayton Foods, Dallas, TX. "No. V14901 (Research Diets, New Brunswick, NJ). Composition of vitamin mix (g/kg of vitamin mix) was as follows: menadione sodium bisulfite, 0.08; biotin 1.0%, 2.0; cyanocobalamin 0.1%, 1.0; niacin, 3.0; calcium pantothenate, 1.6; pyridoxine-HCl, 0.7; riboflavin, 0.6; thiamin-HCl, 0.6; sucrose, 979-37, Supplied per kilogram of vita min mix: vitamin A palmitate, 400,000 iu; vitamin D-3, 100,000 iu; vitamin E acetate, 5,000 ru. 9No. S10001 (Research Diets, New Brunswick, NJ). Composition of mineral mix (g/kg of mineral mix) was as follows: CaHPO4, dibasic, 500.0; MgO, 24.0; potassium citrate, monohydrate, 220.0; K2SO4, 52.0; NaCl, 74.0; Crk(SO4)2, 0.55; cupric carbonate, 0.3; KIO3, 0.01; ferric citrate, 6.0; MnCO3, 3.5; Na2SeO-5H2O, 0.01; ZnCO3, 1.6; sucrose, 118.03. '"Pteroylglutamic acid (Sigma Chemical, St. Louis, MO) was added to experimental diets at either 0, 200, 400 or 600 ng/kg. A control diet was made by adding 1.40 mg pteroylglutamic acid per kg depletion diet.
7.0, 2% ascorbate). Diet, liver and kidney samples were homogenized (Brinkman Instruments, Westbury, NY), boiled and immediately cooled on ice. Milk samples were autoclaved (121°C),cooled and centrifuged (2,000 x g) at 4°C.Fat was subsequently removed by aspiration. Folate concentrations of the resultant supernatant were determined following incubation with conjugase. Par tially purified conjugase from desiccated chicken pan creas was prepared according to a method previously described (14). Folate concentrations in plasma and RBC were deter mined as described by Scott et al. (15) and Cooperman (16) with modifications. Plasma folate concentrations
174
SWIATLO ET AL.
„ TABLE 2
Food intake, weight gain, blood folate concentration and liver folate concentration of rats fed either depletion or control diet during a 12-wk period'-2 Measure g/wkWeight Food intake, g/wkFinal gain, gPlasma weight, nmoI/LRed folate, nmol/LLiver blood cell folate, folate, nmol/g113.71
Depletion diet3
Control diet4
30-
z
20
i
10
2.0-23.1
3.822.7 ± 2.6378.2 ± 2.5379.0 ± 6.310.2 ± 12.982.0+ ± 2.0'379.1129.2-12.9 ± 5.4988.7+31.020.4 ± 1.6125.2
± 5.7
'Values are means ±SEMfor 3-4 rats. 2'Denotes statistically significant differences between dietary treatment groups at the end of the 12-wk period (P < 0.05]. '66 ng Folate/kg diet. "1.4 mg Folate/kg diet.
were determined following incubation of samples at room temperature for 15 min. To ensure lysis of RBC, whole blood was diluted in 0.005 Mphosphate buffer (pH 6.4,0.15% ascorbate) and subjected to ultrasound (Sonifier, Heat Systems-Ultrasonics, Plainview, NY). Samples were then autoclaved, cooled and centrifuged (1000 x g). The resultant supernatant was then assayed for folate concentration. Statistical analyses. Unavailability of folate from experimental diets was determined using slope-ratio statistics (17). Unavailability estimates were deter mined by the multiple linear regression model when dose-response curves shared a common y-intercept. Otherwise, the simple linear regression model was uti lized. The regression equation consisted of an indepen dent variable (x); total folate intake,- and one of four dependent variables (y), either plasma, RBC, liver or kidney folate concentrations. Differences between slopes of the regression lines and calculated folate bioavailabilities were determined by Student's i-test. Cor relation coefficients were assessed for significance using Pearson's test (18).
RESULTS Status of rats following a 12-wk depletion period. Food consumption was significantly lower in rats fed the folate-deficient diet than in rats fed the control diet (Table 2). The difference in food intake was found to have no effect on final weight of folate-deficient rats compared to control animals. Feeding the folate-defi cient diet for 12 wk resulted in significant decreases in both mean plasma and RBC folate concentrations. Mean plasma and RBC folate concentrations of folate-depleted rats were 12% and 38% of those in control rats at the end of the depletion period, respectively (P < 0.05). A similar but statistically insignificant reduction in mean Downloaded from https://academic.oup.com/jn/article-abstract/120/2/172/4743956 by University of Leeds Library user on 18 June 2018
• 100
200
300
y «3.3 + O.OOSx
5-
n=0.69
0
100
200
300
Total Folate Intake (|ig) for a 4 Week Period
FIGURE 1 Dose-response curves for plasma and kidney folate concentrations of rats as a function of dietary folate intake from milk-free diets with graded levels of dietary folate. Correlation coefficients are significant at P < 0.01.
liver folate concentration was noted for folate-deficient compared to control animals at the end of the depletion period. Status of rats following a 4-wk experimental period. Dietary folate concentration and the presence of milk in diets had no effect on either weekly food consump tion or weight gain. Mean (±SEM)body weights (g) and food intakes (g/wk) of all rats were 383.3 ±2.2 and 121.2 ±0.6, respectively, at the end of the depletion period and were 418.6 ±2.6 and 135.0 ±4.8 at the end of the experimental period. Mean liver and kidney weights were also similar among treatment groups throughout the study. Effectiveness of blood and tissue folate concentra tions as indicators of folate bioavailability. The con centrations of folate in plasma, RBC, liver and kidney were regressed against total folate consumption of rats fed milk-free diets with graded levels of folie acid for 4 wk. Relative effectiveness of the above indices as mea sures of folate bioavailability were determined by lin earity over the range of folate intakes, steepness of the slope, and goodness of fit. Plasma folate concentrations were significantly correlated with total folate intake from milk-free diets (r = 0.96, P < 0.01; Fig. 1). Kidney folate concentrations showed a less sensitive but linear relationship to total folate intake (r = 0.69, P < 0.01; Fig. 1). In contrast, neither liver (r = 0.33) nor RBC (r = 0.22) folate concentrations were significantly related to total
175
FOLATE BIOAVAILABILITY
TABLE 3
y «7.48 + 0.161 r=0.67
60 -\
y =7.48 + 0.11x r=0.82 y = 7.48+ 0.09x
Relative bioavailability of folate from milk-free and milk-containing diets1 y . 7.48 + 0.07»
Milk-containing
1=0.66
diets
W*1.0"Humanmilk1.75C3.0CBovinemilk1.25a2.0bGoatmilk0.75b1.5ab PlasmaKidneyMilk-freedietl. 'Relative bioavailabilities were calculated using the slopes from the regression lines of the milk-free diet and the milk-containing (human, bovine, goat| diets. Values in each row with unlike super scripts are significantly different at P < 0.01.
100
200
300
400
y = 3.85 + 0.014l i»0.80 y = 3.85+ 0.009X
MM4
folate intake during the 4-wk experimental period. Relative bioavailability of folate from milk-free and milk-containing diets using slope-ratio statistics. The results of slope-ratio statistical analyses using plasma and kidney folate concentrations are presented in Table 3. Using either plasma or kidney folate concentration as the response variable (Fig.2), no statistically significant differences existed between the y-intercepts of the doseresponse curves. Therefore, bioavailability estimates were determined using the multiple linear regression model. Incorporation of human milk into the diet en hanced folate bioavailability. The relative bioavailabil ity of folate from human MC diets was 40% and 133% greater than that from bovine and goat MC diets, respec tively (P < 0.01). The folate bioavailability from bovine MC diets was 67% greater than the bioavailability of folate from goat MC diets (P < 0.01). Relative bioavaila bility of folate from bovine and goat MC diets did not differ significantly from that of milk-free diets. Using kidney folate concentration as the criterion of response (Fig. 2), folate bioavailability from diets con taining human milk was significantly enhanced (200%) compared to diets without milk (P < 0.01). The bio availability of folate from bovine MC diets was also elevated (100%) compared with milk-free diets (P for a 4 Week Period
FIGURE 2 Dose-response curves for plasma and kidney folate concentrations of rats from experimental diets with graded levels of dietary folate and containing either no milk solids (MF) or solids from human (HM), bovine (BM), and goat milk (GM).
availability of folate from human and bovine milk-con taining diets was significantly greater than that from milk-free diets. Folate bioavailability from human MC diets was also found to be superior to that from bovine and goat MC diets using either plasma or kidney folate concentration as the measure of bioavailability. Neither the bioavailability of folate from various milks nor the impact of various milks on the bioavaila bility of dietary folate has been widely investigated. In human infants, blood folates were linearly related to folate from milk but not to total dietary folate intake (4). The data from studies with humans suggest that folate present in various milks is more bioavailable than that in other infant foods, whereas the data from the present study indicate that human or bovine milks in the diet are capable of increasing total folate bioavaila bility under certain circumstances. Collectively, these data stress the importance of milk to folate nutrition of infants. The component or components of milk that have an impact on folate bioavailability are not well established. Considerable evidence exists to indicate that milk folate
176
SWIATLO ET AL.
binding proteins (MFBP) contribute to the absorption and/or retention of dietary folate in newborns (5-10). To determine whether concentration of MFBP in milk sol ids might relate to differences in the bioavailability of folate from milk-containing diets, the folate binding capacity of lyophilized human, bovine and goat milk was measured (19, 20). Endogenous folates were disso ciated from MFBP by acid before measurement of folate binding activity, so values reported herein reflect the total, not the unsaturated, folate binding capacity of human, bovine and goat milk solids. Mean total folate binding capacity of human milk solids (1.85 ±0.16 umol folie acid bound/kg milk solids) and bovine milk solids (2.28 ±1.35 nmol/kg) did not differ significantly. In contrast, the mean folate binding capacity of goat milk solids (4.48 ±1.98 umol/kg) was significantly greater than that of either human or bovine milk solids. Suffi cient binding protein was present in human, bovine and goat milk solids to bind all folate endogenous to milk and an additional 92.3, 121.5 and 384.6 ng folate per gram of diet, respectively. Therefore, in diets containing human, bovine and goat milk formulated to contain 600 ug/kg folate, for example, sufficient folate binding pro tein was present to bind approximately 15, 21 and 64% of total dietary folate, respectively. The role of MFBP in the absorption and retention of dietary folate appears beneficial. Ford (6) postulated that MFBP indirectly promote folate absorption by making dietary folate unavailable to folate-requiring bacteria. In the presence of goat colostrum, uptake of radiolabeled folie acid by folate-requiring strains of bacteria was considerably reduced compared to uptake of radiola beled folie acid in the absence of colostrum. Tani and Iwai (8) also have shown that uptake of folie acid by folate-requiring intestinal microorganisms was consid erably reduced when it was bound to bovine MFBP. Colman et al. (7) suggested a more direct role for MFBP in folate absorption. The uptake of folate by isolated rat intestinal cells was significantly greater when it was bound to the protein in goat and human milk than when it was unbound. In contrast, Tani and Iwai (8) and Said et al. (9) reported that folate bound to either bovine or human MFBP is absorbed more slowly than unbound folate. Tani and Iwai demonstrated that despite a slower rate of absorption the total amount of folate absorbed remained constant. They also showed that urinary fo late excretion is elevated when unbound folate versus bound folate is administered to rats via oral intubation, suggesting that MFBP enhance body retention and bioavailability of folate. Mason and Selhub (10) have dem onstrated differences in the absorption process of folie acid bound to rat MFBP and that of free folie acid in suckling rats. They suggested that these separate modes of transport for folate may be advantageous to suckling rats. It is appropriate to note here that the mechanism by which folylpolyglutamates (80% of dietary folate) are absorbed differs between human beings and rats (21). Downloaded from https://academic.oup.com/jn/article-abstract/120/2/172/4743956 by University of Leeds Library user on 18 June 2018
Thus, extending our findings in rats to humans is tenu ous, particularly under circumstances in which folate conjugase activity is the rate-limiting step in folate absorption. Assuming that MFBP do enhance the bioavailability of dietary folate, the bioavailability of folate from diets containing goat milk would be expected to be greater than that from diets containing human and bovine milk. In this study, however, the presence of goat milk solids did not enhance bioavailability of folate, suggesting a more complex relationship between folate and MFBP. In contrast, human and bovine milk solids in the diet did enhance bioavailability of folate. An intrinsic factor other than MFBP may be present in human and bovine milk that accounts for the increased bioavailability of folate from diets containing these milks. There is disagreement in the literature concerning the appropriate response tissue for measuring folate bioavailability (22-25). In this investigation, plasma fo late concentration was the most sensitive and specific measure of folate bioavailability as indicated by linear ity of response, steepness of slope and goodness of fit. Kidney folate concentration was likewise a good mea sure of available dietary folate. However, RBC and liver folate concentrations were not linearly related to total folate intake during the repletion period. Liver is often used as the response variable in the folate bioassay because it is believed to be the storage tissue for this nutrient and to reflect the cumulative effect of dietary folate intake over time (21). Abad and Gregory (23) previously reported that hepatic folate con centrations are inconsistent as a response criterion in the folate bioassay, but they found that fasting levels of plasma folate were reliable indicators of dietary folate concentration. Under seemingly similar experimental conditions, others reported (22, 24) that liver folate concentration is a good response variable, as indicated by a significant and positive correlation between hepatic folate and dietary folate concentrations. As previously discussed by Abad and Gregory (23) the concentration of folate in depletion diets does not ac count for these apparent discrepancies. The concentra tion of folate in depletion diets used in the present study, 66 ug/kg diet, was sightly lower than that used by Hoppner and Lampi (24), 75 ug/kg diet. Variation in the folate status of rats at the end of the depletion period in the present study may account for the inconsistent response of hepatic folate concentration to total folate ingestion. After 5 wk of folate depletion, Keagy and Oace (22) found mean liver folate concentrations in rats to be approximately half that of rats in the present investiga tion, which were depleted for 12 wk. Keagy and Oace (22) indicated depletion diets when rats weighed 43-73 g. In our investigation, the mean rat weight at the beginning of the depletion period was 100 g. By 8 wk of depletion, large weekly weight gains ceased, indicating that growth (accretion of lean body mass) had neared
FOLATE BIOAVAILABILITY
completion. Because liver folate concentration is dra matically affected by rate of growth, it is conceivable that the greater liver folate concentrations at the end of the depletion period in this study reflected a decreased requirement for folate among adult rats. Kidney folate concentrations are less frequently mea sured in bioavailability studies. Tigner and Roe (25) found a significant correlation between kidney folate concentrations and dietary intake of folate. There is ample evidence that the kidney is involved in folate homeostasis by acting as a storage site (26), and the tissue may be a sensitive and reliable indicator for measuring folate bioavailability. In summary, human and bovine milk incorporated into rat diets enhanced total folate bioavailability. The constituent in milk that enhances folate bioavailability is unknown. Though previous investigations have indi cated that there are beneficial effects of MFBP on folate absorption and retention, the concentrations of MFBP present in human, bovine and goat milk are not reflec tive of their ability to enhance dietary folate bioavaila bility in the rat. In this study, plasma and kidney were found to be sensitive indicators of dietary folate bio availability. Because there is disagreement among inves tigators concerning which tissue is the most reliable indicator of folate bioavailability in the rat bioassay, it is advisable to use more than one tissue in future stud ies. The experimental circumstances accounting for these reported differences in tissue responses to dietary folate need to be defined.
ACKNOWLEDGMENTS The authors wish to express their extreme grateful ness to the lactating women who collected the large quantities of human milk required in this study. Appre ciation also is expressed to Jimmy Clark from the De partment of Animal Sciences for providing the bovine milk and to Don Angus for the goat milk. The assistance of the Department of Food Science for lyophilizing milk samples and of Don Wade for animal care is acknowl edged.
LITERATURE CITED 1. GRAY,O. P. & BUTLER,E. B. (1965) Megaloblastic anaemia in premature infants. Arch. Dis. Child. 40: 53-58. 2. STRELLING, M. K., BLACKLEDGE, G. D., GOODALL,H. B. & WALKER, C. H. M. (1966) Megaloblastic anaemia and whole blood folate levels in premature infants. Lancet 1: 898-900. 3. BECROFT,D. M. O. & HOLLAND,J. T. (1966) Goat's milk and megaloblastic anemia in infancy. N. Z. Med. f. 65: 303-307.
Downloaded from https://academic.oup.com/jn/article-abstract/120/2/172/4743956 by University of Leeds Library user on 18 June 2018
177
4. SMITH,A. M., PICCIANO,M. F. &. DEERING,R. H. (1985| Folate intake and blood concentrations of term infants. Am. J. Clin. Nutr. 41: 590-598. 5. FORD,J. E., KNAGGS,G. S., SALTER,D. N. & SCOTT,K. J. (1972) Folate nutrition in the kid. Bi. J. Nutr. 27: 571-583. 6. FORD,J.E. (1974) Some observations on the possible nutritional significance of vitamin B-12 and folate-binding proteins in milk. Br.f. Nutr. 31:243-257. 7. COLMAN,N., HETTIARACHY, N. & HERBERT,V. (1981) Detection of a milk factor that facilitates folate uptake by intestinal cells. Science 211: 1427-1429. 8. TANT,M. & IWAI,K. (1984) Some nutritional effects of folatebinding protein in bovine milk on the bioavailability of folate to rats./. Nutr. 114: 778-785. 9. SAID,H. M., HÖRNE, D. W. & WAGNER,C. (1986) Effect of human folate binding protein on folate intestinal transport. Archiv. Biochem. Biophys. 251: 114-120. 10. MASON,J. B. & SELHUB,J. (1988) Folate-binding protein and the absorption of folie acid in the small intestine. Am. ]. Clin. Nutr. 48: 620-625. 11. NATIONALRESEARCH COUNCIL (1978) Nutrient requirements of laboratory animals, No. 10, 3rd rev. ed., National Academy of Sciences, Washington, DC. 12. AMERICANINSTITUTEOFNUTRITION(1980) Second report of the Ad Hoc Committee on standards for nutritional studies. /. Nutr. 110: 1726. 13. KOCHANOWSKI, B. A., SMITH,A. M., PICCIANO,M. F. & SHERMAN, A. R. (1983) Folate depletion secondary to iron deficiency in the neonatal rat. J. Nutr. 113: 2471-2478. 14. KEAGY,P. M. (1984) Folacin-microbiological and animal assays. In: Methods of Vitamin Assay (Augustin, J., Klein, B. P., Becker, D. & Venugopal, P. B., eds.), 4th éd., pp. 445^171, Wiley-Interscience, New York. 15. SCOTT, J. M., CHANTA,V. & HERBERT,V. (1974) Trouble-free microbiologie serum and red cell folate assays. Am. /. Med. Techno!. 40: 125-134. 16. COOPERMAN,J. M. (1971) Microbiological assay of folie acid activity in serum and whole blood. Methods Enzymol. 15: 629642. 17. FINNEY,D. J. (1978) Statistical Method in Biological Assay, 3rd ed., Macmillan, New York. 18. SAS INSTITUTE(1985) SAS User's Guide: Statistics. 5th éd. SAS (Statistical Analysis System) Institute, Cary, NC. 19. BROWN,C. M. (1985) The forms of human milk folacin and variation patterns and relative bioavailability of folacin added to diets containing cow, goat and human milk to rats. MS thesis, University of Illinois, Urbana-Champaign. 20. SELHUB,J., ARNOLD,R., SMITH,A. M. & PICCIANO,M. F. (1984) Milk folate protein (FBP):A secretory protein for folate. Nutr. Res. 4: 181-187. 21. WANG, T. T Y, REISENAUER, A. M. & HALSTED,C. H. (1985) Comparison of folate conjugase activities in human, pig, rat and monkey intestine. /. Nutr. 115: 814-819. 22. KEAGY,P. M. & OACE,S. M. (1982) Development of a folacin bioassay in rats. /. Nutr. 112: 87-91. 23. ABAD,A. &. GREGORY,J. F. (1987) Determination of folate bio availability with a rat bioassay. /. Nutr. 117: 866-873. 24. HOPPNER,K. & LAMPI,B. (1985) Liver folate as a response param eter for folate bioassay in rats. Nutr. Rep. Intern. 31: 203-211. 25. TIGNER,]. & ROE, D. A. (1979) Tissue folacin stores in rats measured by radioassay. Proc. Soc. Exp. Biol. Med. 160: 445-448. 26. STEINBERG, S. E. (1984) Mechanisms of folate homeostasis. Am. /. Physiol. G319-324.
