The Journal of Nutrition Vol. 128 No. 11 November 1998, pp. 1956-1960

Properties of Food Folates Determined by Stability and Susceptibility to Intestinal Pteroylpolyglutamate Hydrolase Action^1,2,3

Elias Seyoum and Jacob Selhub⁴

Jean Mayer USDA Human Nutrition Research Center on Aging at Tufts University, Boston, MA 02111

	ABSTRACT

Abstract Introduction Methods Results Discussion References

The intestinal absorption of folate occurs at the monoglutamyl level, and an important measure of food folate bioavailability is how much folate from the food reaches the intestinal sites in forms that can readily be absorbed. In the absence of protecting agents, e.g., vitamin C and reduced thiols, many labile folates may be lost during cooking and during residence in the acid-peptic milieu of the stomach. On the other hand, the presence of polyglutamyl folate necessitates the action of intestinal hydrolases, which could be affected by food constituents. In this study, we developed an in vitro assay for the determination of an index of food folate availability. The index of folate availability in this study was defined as that proportion of folate that has been identified as monoglutamyl derivatives after tests for stability and susceptibility to an enzymatic hydrolysis. The index of folate availability varied widely among foods. The highest index was for egg yolk (72.2%), followed by cow`s livers (55.7%), orange juice (21.3%), cabbage (6.0%), lima beans (4.5%) and lettuce (2.9%). Yeast folate had the lowest index (0.3%). The availability indices generated by this study correlate with the indices of the bioavailability of the corresponding food folate observed in earlier studies, R² = 0.529 (P = 0.068). Additional information is required on the bioavailability of other food products to test the usefulness of this in vitro approach for assessing food folate availability.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Food folate is a mixture of many derivatives; they differ by the state of oxidation, one-carbon substitution of the pteridine ring and by the number of glutamate residues. These differences are associated with different physicochemical properties, which, together with certain food constitutents, could influence folate bioavailabilty. One important factor of folate bioavailability is folate stability. With the exception of 5-formyltetrahydropteroylglutamates, reduced folates are unstable; those without any carbon substitution, i.e., dihydropteroylglutamates and tetrahydropteroylglutamates are the most unstable, followed by 5-methyltetrahydropteroylglutamates, which exhibit intermediate stability (Cooper et al. 1978, Ghitis 1969, Gregory 1989, Hawks and Villota 1989, O`Broin et al. 1975). Exposure to air oxidation, heat and most importantly, the acid-peptic milieu of the stomach increases folate instability. The presence in foods of antioxidants such as ascorbic acid and reduced thiols protects against this instability (Chen and Cooper 1979, Gregory 1989, Hawks and Villota 1989, Malin 1977).

Another factor that influences folate bioavailability is the presence of polyglutamyl folates. The available information indicates that bioavailability of monoglutamyl folate is higher than the bioavailability of polyglutamyl folate (Godwin and Rosenberg 1975, Gregory 1989, Keagy et al. 1988, Perry and Chanarin 1968). Polyglutamyl folates must be hydrolyzed to the respective monoglutamyl derivatives before they are absorbed by the intestine. This conversion is catalyzed by intestinal hydrolases (Halsted et al. 1975, 1981 and 1986, Halsted 1989). The activity of these enzymes, however, is susceptible to inhibition by the constituents found in some foods. (Bhandari and Gregory 1990, Butterworth et al. 1974, Rosenberg and Godwin 1971). In addition to the inhibitory effect caused by the constituents found in the foods, the activity of these enzymes could also be influenced by the glutamate chain length. However, little is known regarding the activity of this enzyme as far as the pteridine ring structure is concerned.

In this study, we have tested a number of foods for stability of their folate to heat and/or the acid-peptic milieu that exists in the stomach. These same foods were also tested for the susceptibility of the polyglutamated folates to the action of intestinal pteroylpolyglutamate hydrolases. The data obtained were used to create an index of folate availability, which is defined as that proportion of folate that has been identified as monoglutamyl derivatives after tests for stability and susceptibility to enzymatic hydrolysis.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Materials.  Sodium ascorbate, pepsin, zinc acetate, dimethyl glutaric acid and bis(2-hydroxyethyl)imino-tris(hydroxymethyl)methane (Bis-Tris) were obtained from Sigma Chemical (St. Louis, MO). Frozen cow's liver, frozen concentrated orange juice (Minute Maid, Houston, TX), fresh green cabbage and fresh romaine lettuce, active dry baker's yeast (Fleishmann brand, San Francisco, CA), large lima beans (Goya brand, Secaucus, NJ) and medium-sized hen eggs were obtained from local supermarkets.

Methods.  Preparation of jejunal brush border membrane from pig intestine. We used brush border membranes from pig intestine as the source of pteroylpolyglutamate hydrolase. This brush border membrane enzyme, first identified by Reisenauer et al. (1985) in human intestine is unique because no other species except pig intestine was found to contain this enzyme. Subsequent studies by Gregory et al. (1987) demonstrated a large number of similarities between the enzymes from human and pig intestine. Membranes from pig intestine were prepared by the method of Selhub and Rosenberg (1981). Mucosal scrapings of pig jejunum were homogenized in a Waring blender for 2 min in 5 vol of a homogenization buffer (300 mmol/L Mannitol, 10 mmol/L Tris and 16 mmol/L HEPES, pH 7.5); the suspension was filtered through cheesecloth. The filtrate was centrifuged for 10 min at 10,000 × g and the supernatant passed through glasswool. After removing the filtrate, the pellet was subjected to homogenization for 2 min with 5 vol of the homogenization buffer and then combined with original filtrate. MgCl₂ (1 mol/L) was added to this filtrate to a final concentration of 10 mmol/L. After standing in an ice bath for 15 min, the suspension was centrifuged for 15 min at 3000 × g. The pellet was discarded and the supernatant was subjected to a second centrifugation for 15 min at 37,000 × g. The resulting pellet was resuspended in 5 vol of homogenization buffer containing 10 mmol/L MgCl₂. After standing 15 min in an ice-water bath, the two centrifugation steps were repeated. The final pellet was reconstituted in a homogenization buffer (0.125 mL/g mucosa), divided into aliquots and stored at 70°C. Protein was determined by the Lowry method (Lowry et al. 1951). Pteroylpolyglutamate hydrolase activity was estimated by using PteGlu₃ as a substrate for the enzyme (Reisenauer et al. 1985).

Sample preparation. Foods were prepared as follows. Whole egg was boiled for 15 min and the whites discarded. Frozen cow's liver was thawed and minced. Dry lima beans were softened by soaking them in water for 2 h at room temperature and then draining the liquid. Bakers' yeast was used as purchased. The outer leaves of lettuce and cabbage were discarded and the remaining portions were finely chopped. Frozen and concentrated orange juice was first thawed and diluted with 3 vol of water. Each food was divided into two equal portions as follows: 1) the control portion in which folate was extracted under protective (reducing) conditions and analyzed as such without further treatment; and 2) the experimental portion in which extraction and subsequent treatments were performed under simulated conditions.

Control group.  Folate in egg yolk, cow's liver, lima beans, baker's yeast, lettuce and cabbage was extracted by autoclaving for 30 min at 120°C and 103 kPa in 10 vol of 20 g/L sodium ascorbate and 100 mmol/L Bis-Tris buffer, pH 7.8 (Seyoum and Selhub 1993). After cooling in an ice bath, the extract was centrifuged at 40,000 × g for 20 min at 4°C and the supernatant fraction then stored at 70°C. Diluted orange juice solution was neutralized with 1 mol/L NaOH and stored at 70°C.

Folate content and distribution in each extract and orange juice was determined by the affinity/HPLC method as described by Seyoum and Selhub (1993). This method is capable of identification and quantitative estimations of the individual components in any given folate mixture or extract.

Experimental group.  The food portions allocated to this group underwent a series of treatments.

Cooking. Cow`s liver, lima beans, yeast and cabbage were boiled for 10 min in 10 vol of water and then subjected to 30 s homogenization in a Waring blender. Egg yolk and raw lettuce were suspended in 10 vol of room temperature water and homogenized for 30 s in a Waring blender.

Digestion with acid/peptic mixture. Homogenate and orange juices were acidified to pH 2 with 1 mol/L HCL and then thoroughly mixed with freeze-dried pepsin (0.5 g/100 g food). The mixtures were incubated for 2 h at 37°C in a shaking water bath. After incubation, the pH of the mixture was adjusted to 6.5 with 1 mol/L NaOH.

To extract the folates, the acid/peptic digest was boiled at 120^oC for 30 min in equal volumes of extraction buffer (20 g/L sodium ascorbate and 100 mmol/L Bis-Tris buffer, pH 7.8.). After cooling in an ice bath, the extract was centrifuged at 40,000 × g for 20 min at 4°C and the supernatant fraction was divided into two equal portions to determine susceptibility of folates to pteroylpolyglutamate hydrolase action before and after purification by affinity chromatography.

Incubation before purification: One portion of each supernatant containing ~3 nmol folate was incubated as such at 37°C for 4 h in a shaking water bath with pig intestinal brush border membrane preparation (5 mg protein), 33 mmol/L dimethylglutarate buffer (pH 6.5), 0.25 mmol/L sodium chloride, 0.1 mmol/L zinc acetate and 2.5 g/L sodium ascorbate solution (Reisenauer et al. 1985). After incubation, the distribution of folate was determined by the affinity/HPLC method as described by Seyoum and Selhub (1993).

Incubation after purification. The second portion of the supernatant was first subjected to a purification on the affinity column and then incubated with the brush border membrane preparation exactly as described above. After incubation, the distribution of folate was then determined by affinity/HPLC as described by Seyoum and Selhub (1993). Throughout these studies, the activity of pteroylpolyglutamate hydrolase was monitored independently using pteroylhexaglutamate as the substrate.

Data calculation.  Stability of folate. Stability was calculated on the basis of a comparison of total folate concentration in the experimental group with that of a control group as follows: where T2 is the total folate concentration in the experimental group and T1 is the total folate concentration in the control group.

Susceptibility of folate polyglutamate to hydrolysis. Susceptibility to hydrolysis was estimated by comparing the monoglutamyl folate concentration with the total folate concentration within the experimental group as follows: hydrolysis = (M1/T2) × 100 where M1 is the monoglutamyl folate concentration after the treatments and T2 is the total folate concentration after the treatments.

Index of folate availability. This was assessed by comparing the concentration of the monoglutamyl folate in the experimental group to the total folate concentration in the control group as follows: where M1 is the monoglutamyl folate concentration after treatment and T1 is the total folate concentration in the control group.

Data analysis.  Two to three tests were performed per food product. Differences in the release of monoglutamyl folate before and after purification were compared by Student`s paired t test statistics. The relation between the availability index and bioavailability was examined by the Pearson correlation test. The level of significance in both instances was set at P-values <0.05.

	RESULTS

Abstract Introduction Methods Results Discussion References

Folate content and distribution in foods.  The content and distribution of folate in the various foods as obtained under conditions of maximal folate protective measures are shown in Table 1. The food products differed substantially with respect to pteridine ring structure and glutamic acid chain length distributions. Both egg yolk and thawed frozen cow's liver contained monoglutamyl derivatives; 5-methylH₄PteGlu was the only form in egg yolk, whereas in cow's liver only 37% was 5-methylH₄PteGlu (M1), with the rest comprised of H₂PteGlu (D1), and H₄PteGlu (T1). Cabbage, lettuce and orange juice contained methylated tetrahydrofolate with various glutamate chain length and at various proportions. The folate in these foods was 15-36% 5-methylH₄PteGlu (M1). Folate in lima beans was comprised mainly of pentaglutamyl derivatives of which 61.4% were formylated H₄PteGlu₅ (L5/F5); the remainder was comprised of 5-methylH₄PteGlu₅ (30.3%) and the unsubstituted H₄PteGlu₅ (T5). Yeast folates were a mixture of various polyglutamylated derivatives.

Stability and susceptibility of food folate to pteroylpolyglutamate hydrolase action.  Cooking and exposure to an acid/peptic milieu were associated with a number of changes in food folates. The most noticeable change was the loss of folate activity; the degree to which this occurred was different in the various foods (Table 2). The highest loss was in lettuce in which the activity was reduced to ~30% of the original level. Folates in orange juice and egg yolk were the most stable, whereas those in the other foods exhibited intermediate stability. Other changes were more qualitative in nature and are characterized by the following: 1) selective destruction of labile folates, i.e., H₂PteGlu (D1),and H₄PteGlu (T1), in liver extracts (data are not shown), and 2) conversion of reduced folates to oxidized forms particularly to a form with spectral characteristics that correspond to 10-formyl folic acids (see Fig. 1). These oxidized forms were found in cabbage and lima beans, but not in orange juice.

View larger version (11K): [in this window] [in a new window]   Fig 1. The diode array absorption spectrum of the unknown peak from extracts of cabbage and lima beans after the heat and the acid/peptic treatment. The spectrum shows maximal absorption at 272 nm and resembles the spectrum of 10-formyl folic acid provided by Rabinowitz (1960).

The results of the treatment of the acid/peptic digest with the pteroylpolyglutamate hydrolase-containing brush border membrane from pig intestine are presented in Table 3. The amount of monoglutamyl folates released after such a treatment was quite small in incubations containing the crude digest (the high amount of monoglutamyl folates in orange juice represents original monoglutamyl folates that withstood the acid peptic digestion). After purification of the folates by affinity chromatography, incubation with the hydrolase preparation resulted in significant releases of monoglutamyl folate derivatives amounting to 13.4% (vs. 6.1% in the crude extract) of the total folate in cabbage to as high as 56.9% (vs. 7.1% in the crude extract) of the folate in lima beans.

The index of folate availability for the various foods used in this study is given in Table 4. The highest folate availability index was in egg yolk (72.2%) and the lowest in baker's yeast (0.3%).

Availability vs. bioavailability.  Presented in Figure 2 is the comparison between our estimates of the availability indices of food folate and the bioavailability data for folate in the same foods reported by Tamura and Stokstad (1973) and Babu and Sirikantia (1976). Relation between these measurements was described by the following equation: where Y represents the bioavailability of food folate and X the availability index. The two values correlate with an R² value of 0.529 and a P-value that was equal to 0.068. 

View larger version (10K): [in this window] [in a new window]   Fig 2. Relationship between the availability indices and the reported bioavailability for folate from several foods in which R2 = 0.529 and P = 0.068 (Babu and Sirikantia 1976, Tamura and Stokstad 1973).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

In this study, we determined the properties of folates in a number of food products with emphasis on several factors that could influence their bioavailability. As expected, the various foods differed substantially in folate content and distribution. Egg yolk and thawed cow's liver contained only monoglutamyl derivatives, whereas other foods contained mixtures of mono- and polyglutamyl folate derivatives. With the exception of lima beans, which contained formylated tetrahydrofolates, much of the reduced folates in the other foods was comprised of methylated tetrahydrofolates with various chain lengths. Exposure to an acid/peptic milieu followed by incubation with a pteroylpolyglutamate hydrolase preparation revealed even greater differences among the various foods. The most profound difference was seen in folate stability, which ranged from a high of 86% in orange juice to a low of 27% in lettuce. This difference in stability is due to a small extent to differences in the pteridine structure of the folate molecule. The labile folates (dihydro- and tetrahydro- derivatives) found in liver were mainly destroyed upon exposure to the acid/peptic milieu. Other factors, however, appeared to be even more important. Folate distribution in lettuce, cabbage and orange juice is quite similar; the great majority of folates in these foods are methylated tetrahydofolates with a different number of glutamate residues (Table 1). Yet the difference in stability is remarkable. This difference is probably attributable to differences in the antioxidant activities of these foods. For example, egg yolk and liver are abundant in cysteine, and orange juice and cabbage in vitamin C. In contrast, lima beans, lettuce and yeast are deficient in cysteine and/or in vitamin C (Adams 1988, Pennington 1989).

In orange juice, lettuce and lima beans, the release of monoglutamyl folates after incubation with the pig intestinal brush border membrane preparation was significantly lower with the crude extracts than with the affinity purified extracts. These differences in monoglutamyl folate release are probably due to the presence in these extracts of enzyme inhibitors from a number of sources. Inhibitors of pteroylpolyglutamate hydrolases have been reported to be present in lettuce (Santini et al. 1962), in a wide variety of legume extracts including lima beans (Butterworth et al. 1973) and in orange juice (Bhandari and Gregory 1990). Most recently, Wei and Gregory (1998) have shown that organic acids in selected foods inhibit intestinal brush border pteroylpolyglutamate hydrolase activity in vitro. They proposed that this inhibition was one of the mechanisms that affected dietary polyglutamate folate. However, as seen in Table 3, the presence of inhibitors is not the only explanation because even after affinity purification, the releases of monglutamyl folate from both yeast and cabbage extracts after incubation with the intestinal hydrolase were marginal, suggesting perhaps that the purified preparations are not free of inhibitors. The "folate availability indices" reflect the properties of folates in the various foods as they pertain to distribution, stability and susceptibility of pteroylpolyglutamates to intestinal hydrolase. We assessed the usefulness of these indices by comparing them with the indices of bioavailability for the same foods as reported by Tamura and Stokstad (1973) and Babu and Sirikantia (1976). These two quoted studies represent the best available in vivo data on folate bioavailablity. We found that the two set of indices are marginally correlated (R² = 0.529, P = 0.068). These results are encouraging. However, additional information on folate bioavailability of other food products is required to test the usefulness of this in vitro approach in assessing the bioavailability of food folate. A promising approach for studying bioavailability of food folates was introduced by Wei et al. (1996) who mixed the food with isotopically labeled folate polyglutamates. However, even this approach requires verification. There is no assurance that the exogenously added folate undergoes complete mixing with endogenous folate.

	FOOTNOTES

Manuscript received 12 December 1997. Initial reviews completed 1 March 1998. Revision accepted 6 July 1998.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Adams, C. F. (1988) The Nutritive Value of American Foods in Common Units. Agriculture handbook No. 456. Agriculture Research Service, U.S. Department of Agriculture, Washington, DC.
• Babu S., Sirikantia S. G. Availability of folates from some foods. Am. J. Clin. Nutr. 1976; 29:376-379[Abstract/Free Full Text]
• Bhandari S. D., Gregory J. F. Inhibition by selected food components of human and porcine intestinal pteroylpolyglutamyl folate hydrolase activity. Am. J. Clin. Nutr. 1990; 51:87-94[Abstract/Free Full Text]
• Butterworth C. E., Newman A. J., Krumdieck C. L. Tropical sprue: a consideration of possible etiologic mechanisms with emphasis on pteroylglutamate metabolism. Trans. Am. Clin. Climatol. Assoc. 1974; 86:11-22
• Chen T. S., Cooper R. G. Thermal destruction of folacin: effect of ascorbic acid, oxygen and temperature. J. Food Sci. 1979; 44:713-716
• Cooper R. G., Chen T.-S., King M. A. Thermal destruction of folacin in microwave and conventional heating. J. Am. Diet. Assoc. 1978; 73:406-410[Medline]
• Ghitis J. The labile folate of milk. Am. J. Clin. Nutr. 1966; 18:452-455[Abstract]
• Godwin H. A., Rosenberg I. H. Comparative studies of the intestinal absorption of [³H]pteroylmonoglutamate and [³H]pteroylheptaglutamate in man. Gastroenterology. 1975; 69:364-373[Medline]
• Gregory J. F. Chemical and nutritional aspects of folate research: analytical procedures, methods of folate synthesis, stability, and bioavailability of dietary folates. Adv. Food Nutr. Res. 1989; 33:1-101[Medline]
• Gregory, J. F., Ink, S. L. & Cerda, J. J. (1987) Comparison of pteroylpolyglutamate hydrolase (folate conjugase) from porcine and human intestinal brush border membrane. Comp. Biochem. Physiol. 88B: 1135-1141.
• Halsted C. H. The intestinal absorption of dietary folates in health and disease. J. Am. Coll. Nutr. 1989; 8:650-658[Abstract]
• Halsted C. H., Baugh C. M., Butterworth C. E. Jejunal perfusion of simple and conjugated folates in man. Gastroenterology. 1975; 68:261-269[Medline]
• Halsted C. H., Beer W.H., Chandler C. J., Ross K., Wolfe B. M., Bailley L., Cerda J. J. Clinical studies of intestinal folate conjugases. J. Lab. Clin. Med 1986; 107:228-232[Medline]
• Halsted C. H., Gandhi G., Tamura T. Sulfasalazine inhibits the absorption of folates in ulcerative colitis. N. Engl. J. Med. 1981; 305:1513-1517[Medline]
• Hawks J. G., Villota R. Folates in foods: reactivity, stability during processing, and nutritional implications. Crit. Rev. Food Sci. Nutr. 1989; 28:439-538[Medline]
• Keagy P. M., Shane B., Oace S. Folate bioavailability in humans: effect of wheat bran and beans. Am. J. Clin. Nutr. 1988; 47:80-88[Abstract/Free Full Text]
• Lowry O. H., Rosebrough N. H., Farr A. L., Randall R. J. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 1951; 193:265-275[Free Full Text]
• Malin J. D. Total folate activity in Brussels sprouts: the effect of storage, processing, cooking and ascorbic acid content. J. Food. Technol. 1977; 12:623-632
• O'Broin J. D., Temperley I. J., Brown J. P., Scott J. M. Nutritional stability of various naturally occurring monoglutamyl folate derivatives of folic acid. Am. J. Clin. Nutr. 1975; 8:438-444
• Pennington, J.A.T. (1989) Bowes and Church`s Food Values of Portions Commonly Used (Cleary, P. & Stewart, L. J., eds.). J. B. Lippincott Company, Philadelphia, PA.
• Perry J., Chanarin I. Absorption and utilization of polyglutamyl forms of folate in man. Br. Med. J. 1968; 4:546-549
• Rabinowitz, J. C. (1960) Folic acid. In: The Enzymes (Boyer, P., Lardy, H. & Myrback, K., eds.), pp. 185-252. Academic Press, New York, NY.
• Reisenauer A. M., Halsted C. H., Jacobs L. R., Wolfe B. M. Human intestinal folate conjugase: adaptation after jejunoileal bypass. Am. J. Clin. Nutr. 1985; 42:660-665[Abstract/Free Full Text]
• Rosenberg I. H., Godwin H. A. Inhibition of intestinal gamma-glutamyl carboxypeptidase by yeast nucleic acid: an explanation of variability in utilization of dietary polyglutamyl folate. J. Clin. Investig. 1971; 50:78a
• Santini R., Berger F. M., Berdasco G., Sheehy T., Aviles J., Davila I. Folic acid activity in Puerto Rican foods. J. Am. Diet. Assoc. 1962; 41:562-567[Medline]
• Selhub J., Rosenberg I. H. Folate transport in isolated brush border membrane vesicles from rat intestine. J. Biol. Chem. 1981; 256:4489-4493[Abstract/Free Full Text]
• Seyoum E., Selhub J. Combined affinity and ion pair column chromatographies for the analysis of food folate. J. Nutr. Biochem. 1993; 4:488-494
• Swendseid M. E., Bird O. D., Brown R. A., Bethell F. H. Metabolic functions of folic acid and its hexaglutamyl conjugate. II. Urinary excretion studies on normal persons. Effect of conjugase inhibitor. J. Lab. Clin. Med. 1947; 32:23-27
• Tamura T., Stokstad E.L.R. The availability of food folate in man. Br. J. Haematol. 1973; 25:513-532[Medline]
• Wei M.-M., Bailey L. B., Toth J. P., Gregory J. F. Bioavailability for humans of deuterium-labeled monoglutamyl and polyglutamyl folates is affected by selected foods. J. Nutr. 1996; 126:3100-3108
• Wei M.-M., Gregory J. F. Organic acids in selected foods inhibit intestinal brush border pteroylpolyglutamate hydrolase in vitro: potential mechanism affecting the bioavailability of dietary polyglutamyl folate. J. Agric. Food Chem. 1998; 46:211-219[Medline]

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences