Folate Profile in Tissues from Infant Primates Fed Folic Acid

Abstract

Background:

Diet-derived folic acid (FA) is found in circulation in a high percentage of United States subjects, but knowledge of tissue folate forms and concentrations in humans is limited. The infant rhesus macaque is an excellent model for the human infant and, like the human, has lower FA reductive activity than rodents.

Objectives:

This study explored tissue folate concentrations in FA-fed infant rhesus macaques.

Methods:

Archived tissue samples from 6-mo-old infant macaques who were fed breast milk and later some maternal diet (BMD, n = 8) or were fed a milk replacer (MR, n = 15) were analyzed for tetrahydrofolate (THF), 5-methyltetrahydrofolate (5-MTHF), and unmetabolized folic acid (UMFA); other forms were not measured. Folates were analyzed in plasma, primary organs, and multiple brain regions by liquid chromatography–tandem mass spectrometry.

Results:

All 3 folates were detected in each tissue with the exception that THF was undetectable in plasma. ΣFolate (THF+5-MTHF+UMFA) concentrations were highest in the liver and kidney (P < 0.05), then heart (P < 0.05) and brain, lung, and plasma (all P < 0.05 compared with kidney) with a 100-fold difference from highest to lowest. Tissue THF and 5-MTHF followed similar but not identical patterns. UMFA concentration was highest in the kidneys and lungs (P < 0.05). Prefrontal and temporal cortices and cerebellum were higher in Σfolate than motor cortex, occipital cortex, and striatum (P < 0.05). THF was the most abundant folate measured in each brain region. Tissue and brain concentrations were similar between BMD and MR with only a few sporadic differences.

Conclusions:

Differential tissue folate metabolism, uptake, and/or retention are suggested, and further research is necessary to explore the mechanisms involved.

Introduction

Folates are required for the biosynthesis of DNA and RNA, methionine recovery, methylation processes, 1-carbon metabolism, and redox chemistry needed for normal cellular growth and division [1]. Insufficient folate intake is associated with altered 1-carbon metabolism and megaloblastic anemia in adults and older children [2], as well as the manifestation of neural tube defects in neonates [3,4]. To reduce the incidence of neural tube defects [5] and low folate intake by United States adults [6], the Food and Drug Administration introduced folic acid (FA) fortification of staple grains in 1998. Following this United States–led initiative, >90 countries have implemented similar fortification programs. As a result, many people now consume folates from foods and supplements that are fortified with FA, in addition to naturally occurring folates in a variety of foods.

Folate is the umbrella term used for both natural and synthetic folates [7]. FA is a synthetic oxidized folate used in supplements and food fortification due to its low cost and stability in food matrices [8]. The tetrahydrofolates, 5-methyltetrahydrofolate (5-MTHF) and tetrahydrofolate (THF), are the most abundant natural folates in foods. Due to the action of intestinal and hepatic dihydrofolate reductase (DHFR) and hepatic methylene tetrahydrofolate reductase (MTHFR) [9], along with intestinal and hepatic 1-carbon metabolism, food folates generally enter the circulation in the form of 5-MTHF. Indeed, 5-MTHF is the predominant folate found in human circulation [10]. However, human DHFR capacity is limited when compared to that of rodents and pigs [9]. As a result, >95% of 2707 United States subjects had detectable unmetabolized FA (UMFA) in their plasma, independent of supplement use and age [11]. In addition, several reports have demonstrated that elevated FA intake raises plasma UMFA concentration in humans [12-16].

We are not aware of published reports investigating the impact of limiting DHFR activity combined with dietary FA on tissue folate profiles using DHFR knockout mice. However, MTHFR knockout mice have been studied in conjunction with dietary FA intake. MTHFR ± knockout mice fed a FA concentration well above the estimated requirement for mice had plasma ratios of 5-MTHF and UMFA [17-19] like those reported in humans [11]. However, the mice had markedly higher plasma THF and total folate concentrataions than humans. This likely reflects the higher DHFR activity persisting in MTHFR knockout mice than the reduced DHFR activity observed in humans. As a result, tissue folate profiles may offer another point of difference between mice and humans. Although both 5-MTHF and THF constitute a large portion of cellular folates in both species, concentrations of THF are generally higher than those of 5-MTHF in rodent tissue [17,20-23]. In contrast, limited data from human tissues reveal higher 5-MTHF than THF, although concentrations have only been reported for colon mucosa [24] and skin [10], which are both rapidly dividing tissues. Given the paucity of data from human tissues, especially human infants, we studied tissue folate profiles in infant rhesus macaques. Rhesus macaques have a close phylogenetic relationship with humans and are an excellent model of human infant brain development [25], physiology, and metabolism [26]. Hepatic DHFR activity in adult rhesus macaques is ~4 times higher than in humans [27], whereas, in rats, it is 16 to 50 times higher [28]. Young rats and humans had significantly lower hepatic DHFR activity than the corresponding adults [27], suggesting that DHFR activity might be lower in young rhesus macaques than in adults. This led us to hypothesize that the infant rhesus macaque may provide useful information on tissue folate profile as a model of the human being. The objective of this study, which is a secondary analysis of tissue samples previously collected from 6-mo-old infant rhesus macaques [29], was to investigate tissue folate profiles in a wild type animal model with the potential to have circulating UMFA.

Methods

Animals and diet

The complete details of the original study [29], along with other studies from the same cohort, have been previously published [30]. As noted earlier, from birth to 6 mo of age rhesus macaque infants were nursed by their dams and gradually also ate the mother’s standard monkey diet or were exclusively fed a well-defined milk replacer (MR) than contained supplemental FA. All procedures were approved by the Institutional Animal Care and Use Committee of Oregon Health & Science University and carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Two groups were fed MRs differing in carotenoid concentration and source of α-tocopherol but with the same concentration of FA; they were combined into one group in the present analysis, hereon referred to as MR-fed group (n = 15). The group of infant rhesus macaques that were not only breastfed but also had access to the maternal diet is hereon referred to as breastfed, mixed diet-fed (BMD) group (n = 8). The meam FA concentrations in the MR, breast milk, and the maternal diet are provided in Table 1. The MRs contained only FA and concentrations of the other nutrients and compounds, including DHA and arachidonic acid, were consistent as the MRs came from the same batch [30]. Breast milk from the dams contained THF, 5-MTHF, and UMFA. The manufacturer (Lab Diets Inc.) of the maternal diet provided values for FA fortification but not inherent THF and 5-MTHF concentrations likely to result from some ingredients used. Mothers received a mixture of a small amount fruits and vegetables daily, which was not quantified. It is likely that the amount of THF and/or 5-MTHF was substantially lower than that of FA.

TABLE 1.

Folate profiles of maternal and infant rhesus macaque diets¹

Folate species	Maternal diet (pmol/g)	Breast milk (pmol/mL)	Milk replacer (pmol/mL)
THF	—	17.1 ± 4.28	—
5-MTHF	—	19.9 ± 8.5	—
FA	21,700	164 ± 67.3	482 ± 58.1

Abbreviations: FA, folic acid; MTHF, methyltetrahydrofolate; THF, tetrahydrofolate.

¹Values are means ± SDs.

Tissue collection and folate measurement

Breast milk from dams and tissue samples were obtained from male and female infant primates (age = 6 mo) as previously described [29,30]. The animals were fasted prior to harvest. The concentration of UMFA, THF, and 5-MTHF were analyzed in breast milk, plasma, primary organs, and multiple brain areas by stable isotope dilution liquid chromatography–tandem mass spectrometry by an in-house method adapted from Szpylka et al. [31]. Briefly, tissues were homogenized with extraction buffer in a bead mill, treated with rat serum (Sigma Aldrich) to completely deconjugate the polyglutamate forms of folates, and cleaned up by solid-phase extraction before analysis. Plasma was diluted in extraction buffer and cleaned up by solid-phase extraction. We decided not to measure formyl, methylene, and methenyl folates due to their instability and interconversions under our analytical conditions [32,33].

Statistical analysis

Data were analyzed using GraphPad Prism version 9.3.1 for Windows (GraphPad Software; www.graphpad.com). The D’Agostino and Pearson normality test revealed that some of the data were not normally distributed. Data were not normalized by arithmetic transformation. Therefore, the Kruskal-Wallis test was used to make nonparametric comparisons between tissue folates. Dunn multiple comparisons test was used to correct for multiple comparisons. Plasma folates (expressed as pmol/mL) were compared with tissue folates (expressed as pmol/g tissue) based on the assumption of the weight of 1 mL of plasma (density = 1.025 g/mL). Folate form concentrations for comparisons between the BMD and MR feeding groups were done using two-tailed Mann-Whitney tests. Statistical significance was set at P ≤ 0.05. The sum of the concentrations of the folates (∑folate) was calculated by summing the concentrations of THF, 5-MTHF, and UMFA. Subsequently, the relative proportions of THF, 5-MTHF, and UMFA presented as pie charts by tissue were determined by dividing their individual concentrations by ∑folate for all animals.

Results

We estimated the daily folate intake for each dietary group at 6 mo of age based on caloric intake calculated from week 24 formula intake data previously reported for these monkeys [29, 30,34] (Table 2). Mean daily caloric intake in the MR group was 294 kcal/d during the last week of the study. Breast milk intake values were not available for the BMD infants and from 2 mo of age BMD infants consumed increasing amounts of pelleted monkey diet fed to their mother, culminating in an estimated intake of ~235 kcal/d from that source. As previously reported [30], we assumed that BMD infants had the same daily energy intake as the MR infants since there were no differences between the dietary groups for either growth velocity or final body weight. Based on this assumption, the pelleted monkey diet provided most of the energy for the BMD group at 6 mo of age accompanied by a modest intake of breast milk of ~69 mL/d. The estimated folate intake in the BMD group was ca. >5000-fold higher than in the MR group driven by consumption of FA from the pelleted monkey diet.

TABLE 2.

Estimated food and folate intake at 6 mo of age in infant rhesus macaques that were fed a MR supplemented with folic acid or a mixture of mother’s milk and mother’s food containing FA^a.

Intake amount	Dietary group
	MR	BMD
MR intake (kcal/d)	2941	NA
Estimated breast milk intake (kcal/d)	—	294
Monkey diet intake
Estimated grams per day	—	822
Estimated kilocalories per day	—	2352
Breast milk intake
Estimated grams per day	—	593
Estimated kilocalories per day	—	693
Folate intake
MR (pmol/d)	2024	N/A
Monkey diet (pmol/d)	—	1,779,4004
Breast milk
UMFA (pmol/d)	—	10.74
THF (pmol/d)	—	1.14
5-MTHF (pmol/d)	—	1.34
Total folate intake (pmol/d)	202	1,779,413

Abbreviations: BMD, breastfed, mixed diet; FA, folic acid; MTHF, methyltetrahydrofolate; MR, milk replacer; THF, tetrahydrofolate; UMFA, unmetabolized FA.

¹From the study by Jeon et al. [29], intake per day was calculated from caloric density of the MRs as 0.712 kcal/mL. Calculated daily calorie intake for the MR groups combined was 294 kcal/d at 6 mo of age. Assumed BMD infants consumed similar calories per day as the MR infants since they grew at the same rate and finished at the same body weight [29].

²Estimated daily intake of monkey diet by the BMD infants at 6 mo of age. Monkey diet (LabDiets) had a caloric density of 2.87 kcal/g.

³Estimated caloric intake from breast milk by subtracting monkey diet caloric intake from the estimated daily calorie intake. Rhesus macaque milk has a caloric density of ~0.854 kcal/mL [34].

⁴Milk replacer contained 482 pmol FA/mL. Monkey diet contained 21,700 pmol FA/g diet, and breast milk contained 164 pmol FA/mL, 17.1 pmol THF/mL, and 19.9 pmol 5-MTHF/mL. Breast milk values were calculated by determining a mean for each dam, then calculating a mean for the dams together.

^aThe MR group received a MR supplemented with FA. The BMD group received their dam’s breast milk, followed by mother’s monkey diet.

Given the difference in estimated folate intake between the BMD and MR, we compared tissue folates between BMD and MR (Table 3). Liver Σfolate was 30% higher in BMD than that in MR infants (P < 0.05). There were no other statistical differences in ∑folate between the BMD and MR groups. Tissue THF, 5-MTHF, and UMFA amounts were generally similar between the BMD and MR groups but there were a few exceptions. Liver THF was ~40% higher in the BMD than in the MR group (P < 0.05), but kidney THF was 33% lower in BMD than in MR (P < 0.05). Plasma 5-MTHF was 75% higher in the BMD group than in the MR group (P < 0.05). UMFA was 83% lower in the BMD group for plasma (P < 0.05), 44% lower in the heart (P < 0.05), and 74% lower in the kidneys (P < 0.05) than those in the MR group.

TABLE 3.

Comparison of folate concentrations in plasma and various tissues from infant rhesus macaques fed either a mixed diet or a MR¹

Dietary group	Folate species	Plasma	Liver	Kidney	Heart	Brain	Lung
BMD	∑Folates	20.8 ± 2.64	15039 ± 1350*	5346 ± 402.5	734 ± 72.1	267 ± 24.1	107 ± 9.81
	THF	ND	5552 ± 747.5*	3104 ± 252.4*	117 ± 10.6	221 ± 20.1	52.1 ± 9.88
	5-MTHF	17.4 ± 2.99*	9467 ± 937.5	2104 ± 301.7	607 ± 70.3	44.8 ± 3.64	9.23 ± 1.13
	UMFA	2.09 ± 0.861*	20.0 ± 1.97	139 ± 34.6*	10.1 ± 1.55*	12.9 ± 0.40	45.9 ± 3.57
MR	∑Folates	14.9 ± 0.952	11346 ± 647.3*	6443 ± 399.6	589 ± 39.8	307 ± 28.6	103 ± 4.63
	THF	ND	3912 ± 345.5*	4136 ± 188.3*	122 ± 17.6	238 ± 21.8	51.1 ± 4.62
	5-MTHF	9.92 ± 0.837*	7409 ± 485.0	2064 ± 293.8	453 ± 40.6	56.6 ± 7.79	8.35 ± 0.824
	UMFA	3.84 ± 0.499*	24.5 ± 2.93	243 ± 26.9*	14.5 ± 1.11*	12.7 ± 0.99	43.5 ± 2.77

Abbreviations: BMD, breastfed, mixed diet; FA, folic acid; MTHF, methyltetrahydrofolate; MR, milk replacer; ND, not detected; THF, tetrahydrofolate; UMFA, unmetabolized FA.

¹Values are expressed as pmol/mL for plasma and pmol/g for all other tissues. Values are mean ± SEM. BMD group, received a diet consisting of dam’s milk and the mother’s monkey diet, n = 8; MR group, received a MR, n = 15. ∑Folates, THF + 5-MTHF + UMFA. The lower limit of detection for THF was 0.11 pmol/g.

^*P < 0.05 BMD vs. MR for same folate, 2-tailed Mann-Whitney test.

As tissue folate concentrations were generally similar between the feeding groups and lacked marked quantitative differences, the groups were combined to better compare the folates between different tissues (Table 4). The 3 folates were detected in all samples except in plasma in which THF was undetectable, but concentrations varied among tissues. ∑Folate was highest in liver and kidneys, followed by heart (P < 0.05 compared with the liver), brain (P < 0.05 compared with the kidneys and liver), lung (P < 0.05 compared with the heart, kidneys, and liver), and plasma (P < 0.05 compared with all tissues except the lungs) (Table 4). THF and 5-MTHF were the most abundant folates measured in different tissues. THF was higher in the liver and kidneys than in the brain and heart (P < 0.05 compared with the liver and kidney) and lungs (P < 0.05 compared with the liver, kidneys, and brain) with plasma being undetectable. Comparing the tissues for 5-MTHF concentration revealed that the liver and kidneys were highest, followed by the heart (P < 0.05 compared with the liver), brain (P < 0.05 compared with the liver and kidneys), plasma (P < 0.05 compared with the liver, kidney, and heart), and lungs (P < 0.05 compared with all tissues except plasma). Notably, 5-MTHF was the most abundant folate measured in the liver, heart, and plasma, despite liver also being the highest in THF. UMFA was found in each tissue tested with the highest concentration in the kidneys and lung, followed by the liver, brain, heart, and plasma (all P < 0.05 compared with the kidneys). Notably, the concentration of UMFA in lung was 5-fold higher than that of 5-MTHF. The folate profile for each tissue is presented in Supplementary Figure 1 expressed as a percentage of the sum of the 3 folates measured. Note that these percentages do not consider formyl, methenyl, and methylene folates.

Table 4.

Folate concentrations in plasma and various tissues from infant rhesus macaques^a

Folate species	Plasma	Liver	Kidney	Heart	Brain	Lung
∑Folates	16.9 ± 1.23	12,630 ± 718.51	6062 ± 310.81,2	639 ± 38.12,3	293 ± 20.53,4	105 ± 4.514,5
THF	ND	4483 ± 372.91	3777 ± 181.01	120 ± 11.92,3	232 ± 15.61	51.5 ± 4.493,4
5-MTHF	12.5 ± 1.37	8125 ± 488.51	2078 ± 214.31,2	507 ± 38.42,3	52.5 ± 5.303,4	8.67 ± 0.6565
UMFA	3.23 ± 0.466	22.9 ± 2.052,3	207 ± 23.41	12.9 ± 0.9911	12.8 ± 0.653	44.4 ± 2.151,2

Abbreviations: FA, folic acid; MTHF, methyltetrahydrofolate; ND, not detected; THF, tetrahydrofolate; UMFA, unmetabolized FA.

^1–5Values in row with unlike superscript differed as assessed by the Kruskal-Wallis test with Dunn multiple comparisons correction, P < 0.05. Plasma values were not compared with those of tissues, given the differences in units of measurement.

^aInfant rhesus macaques were fed a milk replacer (MR, n = 15) or a mixed diet consisting of dam’s milk and the mother’s monkey diet (BMD, n = 8). Values are expressed as pmol/mL for plasma and pmol/g for all other tissues. Values are mean + SEM, n = 23. ∑Folates, THF + 5-MTHF + UMFA.

Brain folate profile was analyzed in the prefrontal cortex (PFC), occipital cortex (OC), temporal cortex (TC), striatum (ST), motor cortex (MC), and the cerebellum (CB). Folate profiles for these brain regions were similar between the 2 feeding groups with 2 exceptions (Table 5). UMFA concentrations were higher in PFC (P < 0.01) and ST (P < 0.01) from the BMD group than those from the MR group (P < 0.05). It is notable that the ST had the lowest concentrations of both THF and 5-MTHF and had a relatively high concentration of UMFA than the other brain regions. CB, however, had similar concentrations of UMFA as ST but did not have similarly low concentrations of THF and 5-MTHF. In contrast, the PFC and TC had higher concentrations of THF and 5-MTHF than the other brain regions and had relatively lower concentrations of UMFA.

TABLE 5.

Comparison of folate concentrations in various brain regions from infant rhesus macaques fed either a mixed diet or a MR¹

Dietary group	Folate species	PFC	OC	TC	ST	MC	CB
BMD	∑Folates	464 ± 40.6	167 ± 40.0	343 ± 61.5	148 ± 38.5	264 ± 46.0	291 ± 37.6
	THF	394 ± 36.7	131 ± 38.6	269 ± 56.2	87.1 ± 37.6	210 ± 43.8	233 ± 39.7
	5-MTHF	60.6 ± 6.35	29.1 ± 2.62	64.5 ± 9.02	30.8 ± 6.13	47.5 ± 5.99	36.0 ± 4.35
	UMFA	9.08 ± 0.554*	7.28 ± 0.607	9.64 ± 0.624	20.2 ± 1.19*	7.12 ± 0.831	22.3 ± 1.87
MR	∑Folates	419 ± 43.5	214 ± 22.2	455 ± 50.9	110 ± 23.0	258 ± 42.9	388 ± 58.6
	THF	339 ± 36.3	171 ± 23.8	353 ± 44.4	72.2 ± 21.6	195 ± 33.7	293 ± 43.1
	5-MTHF	73.2 ± 11.7	36.2 ± 6.52	94.3 ± 10.9	17.9 ± 2.47	57.6 ± 11.5	60.4 ± 17.0
	UMFA	7.01 ± 0.570*	7.12 ± 0.866	7.95 ± 0.945	14.7 ± 1.22*	5.40 ± 0.338	34.6 ± 5.17

Abbreviations: BMD, breastfed, mixed diet; CB, cerebellum; FA, folic acid; MTHF, methyltetrahydrofolate; MC, motor cortex; MR, milk replacer; OC, occipital cortex; PFC, prefrontal cortex; ST, striatum; TC, temporal cortex; THF, tetrahydrofolate; UMFA, unmetabolized FA.

¹Values are expressed as pmol/g tissue and means ± SEM. The BMD group received a diet consisting of dam’s milk and the mother’s monkey diet, n = 8; The MR group received a MR, n = 15.

^*P < 0.05 BMD vs. MR for same folate, 2-tailed Mann-Whitney test.

As brain folate concentrations were similar between the feeding groups, the groups were combined to better compare the folates between different brain regions (Table 6). As in the peripheral tissues, Σfolate varied markedly among brain regions with PFC, TC, and CB being higher (P < 0.05) than MC, OC, and ST (P < 0.05 compared with MC). THF was the most abundant folate measured in each region and generally followed the same relative order as Σfolate with PFC, TC, and CB being higher (P < 0.05) than the other brain regions. MC and CB had higher THF than ST (P < 0.05). Similar to THF, PFC and TC had higher 5-MTHF than OC (P < 0.05) and ST (P < 0.05) but were not different from MC and CB. ST and OC had the lowest 5-MTHF concentration and ST was lower than all other regions (P < 0.05). The concentration of 5-MTHF was substantially lower than that of THF in each region, but it was at least 4-fold higher than UMFA in each brain region, except ST and CB. Brain region concentrations of UMFA followed a different profile with CB and ST having the highest concentrations (P < 0.05), with TC, PFC, OC, and MC having similar and lower concentrations. The folate profile for each tissue is presented in Supplementary Figure 2 expressed as a percentage of the sum of the 3 folates measured. Note that these percentages did not consider the formyl, methenyl, and methylene folates.

TABLE 6.

Folate concentrations in various brain regions from infant rhesus macaques^a.

Folate species	PFC	OC	TC	ST	MC	CB
∑Folates	435 ± 31.41	198 ± 22.62,3	416 ± 40.01,4	122 ± 19.73	260 ± 32.72,4	354 ± 40.91,4
THF	358 ± 27.01	157 ± 20.42	323 ± 35.21,4	80.6 ± 18.73	200 ± 26.62,4	272 ± 31.41,2
5-MTHF	68.8 ± 7.941	33.7 ± 4.352,3	84.0 ± 8.221	22.4 ± 2.902	54.7 ± 8.361,3	51.9 ± 11.31,3
UMFA	7.73 ± 0.4612	7.18 ± 0.5842	8.54 ± 0.6652	16.6 ± 1.041	5.89 ± 0.3692	30.3 ± 3.611

Abbreviations: BMD, breastfed, mixed diet; CB, cerebellum; FA, folic acid; MTHF, methyltetrahydrofolate; MC, motor cortex; MR, milk replacer; OC, occipital cortex; PFC, prefrontal cortex; ST, striatum; TC, temporal cortex; THF, tetrahydrofolate; UMFA, unmetabolized FA.

^1–4Values in row with unlike superscript differed as assessed by the Kruskal-Wallis test with Dunn multiple comparisons correction, P < 0.05.

^aInfant rhesus macaques were fed a milk replacer (MR, n = 15) or a mixed diet consisting of dam’s milk and the mother’s monkey diet (BMD, n = 8). Values are expressed as pmol/g tissue and means ± SEM, n = 23. PFC, prefrontal cortex; OC, occipital cortex; TC, temporal cortex; ST, striatum; MC, motor cortex; CB, cerebellum.

Discussion

Tissue folate profiles were retrospectively explored in rhesus macaques to provide information on profiles in an animal that has less tissue DHFR than common rodent models [27,28]. Human beings also have less DHFR than common rodent models, and there is limited information on tissue folate profiles. Rhesus macaques are an excellent model for humans due to their close phylogenetic relationship with humans [25,26]. For these reasons, we used preserved tissues from a previously completed trial in infant rhesus macaques to provide insight into tissue folate profiles in response to a diet rich in FA. The original study divided the rhesus macaque infants into 2 feeding groups with respect to folate intake: MR and BMD. Infants in the BMD group received milk from their dam and consumed some of their mother’s food, which was a standard FA-supplemented chow--based monkey diet. Newborns rely exclusively on their dam’s milk but transition over the 6-mo trial period to less milk and more of the dam’s chow-based diet [29]. The second group of infants, MR, was fed a well-controlled MR with FA. Although the MR contained ~2 times as much folate as the dam’s milk, the mother’s food was highly concentrated in FA compared with the MR. As a result, we estimated that folate intake, predominantly as FA, was markedly higher in the BMD group than in the MR group. Thus, not only do the data provide insight into tissue folate profiles in macaque infants consuming FA, but they also provide insight into tissue profiles in animals consuming very different concentrations of FA.

It is striking that we observed only modest changes in the range of 30% to 80% in tissue folate concentrations in response to an estimated >5000-fold higher intake of FA in the BMD group than in the MR group. 5-MTHF was higher in the BMD group in plasma and the liver, with only a nonsignificant trend in ∑folate toward higher concentrations in both plasma and the liver. The BMD group also had higher UMFA in both PFC and ST. Interestingly, the only changes in other tissues were lower THF in kidney and lower UMFA in both the kidneys and the heart in the BMD group. This is a surprising finding given the dramatically higher FA intake in the BMD group and is consistent with systemic resistance to changes in ∑folate and tissue folate distributions. This research was not designed to shed light on the implications of the folate profile perturbations detected.

With the groups combined, we found that UMFA comprised ~20% of ∑folate in plasma, whereas Pfeiffer et al. [11] reported that UMFA accounted for 2% to 4% of the total plasma folate (5-MTHF+UMFA) in humans in the United States. The formyl, methenyl, and methylene folate forms were not detected in plasma. The former finding provided the incentive to examine tissue folate profiles in these infant rhesus macaques that, like the human, have circulating UMFA concentrations resulting from intake of FA. THF, 5-MTHF, and UMFA were each found in each tissue and indeed in each individual sample analyzed. Despite this, the folate concentrations varied widely by tissue.

We found that liver had the highest ∑folate, which was 100fold higher than the lowest ∑folate, which was observed in lung. We observed a similar range in tissue THF and a >100-fold range for 5-MTHF. We also observed different folate profiles between tissues. Whereas 5-MTHF was the most abundant folate in liver, heart and plasma, THF was most abundant in the brain and kidneys among the folates measured. Notably, plasma had a significant proportion of UMFA as discussed earlier. In lung, the UMFA proportion was similar to that of THF and exceeded 5-MTHF by 5-fold. The significance of these different folate concentrations and patterns is not clear; however, as reported elsewhere organ weights and appearance assessed by a veterinarian were not different between the feeding groups, except the liver and adrenal glands [29]. Despite the differences in estimated folate intake between the feeding groups, we found only a few minor differences between the BMD and MR groups and those differences did not fit a clear pattern.

The tissue ∑folate concentrations reported in this study for infant macaque tissues are in line with both rodent folate organ concentrations and the general order of organ concentrations [20-23] despite the higher plasma folate concentrations found in those rodents. The higher plasma concentrations in the rodents [20-23] resulted from high diet concentrations of FA that delivered an estimated 360 nmol folate/d to the rats and ~70 nmol/d to the mice studied. For comparison, the reference daily intake (RDI) for human adults is ~1 μmol/d. Taken together, rodents receiving a high concentration of dietary FA, nonetheless, had tissue folate concentrations similar to those reported in this study for infant macaques. The ∑folate concentrations reported for human epidermis and dermis [10] fall within the range reported in this study with epidermis being similar to infant macaque heart and human dermis being similar to infant macaque lung. Likewise, the percentage of ∑folate found as 5-MTHF by Hasoun et al. [10] was similar to the estimated percentage of 5-MTHF we report in this study for the liver and kidneys, respectively.

Folate profiles differed among the brain regions studied. ∑Folate was highest in PFC, TC, CB, and the cerebellum, followed by MC, OC, and ST. This pattern was mostly driven by the concentration of THF, the predominant folate in each brain region. In each case, the concentration of 5-MTHF was lower than that of THF but higher than that of UMFA. Interestingly, UMFA concentrations were highest in the CB, followed by ST and the other brain regions. UMFA concentrations in CB and ST were significantly higher than in other areas of the brain. It is of note that the ranges in folate concentrations and profiles were not as variable between brain regions as noted earlier for peripheral tissues. Surprisingly, UMFA concentration in the striatum was roughly the same as the concentration of 5-MTHF, and in the cerebellum, UMFA concentration was ~75% that of 5-MTHF. There was no clear pattern indicating that tissue concentrations of UMFA impacted tissue concentrations of THF or 5-MTHF. However, given the small size of this study, additional research could further test these observations and could investigate the mechanism by which certain brain regions accumulate more UMFA than other regions. The detection of UMFA in these brain samples raises a number of interesting questions as described further:

Mechanism of UMFA uptake

The uptake of folates into brain regions primarily involves mediated transport across the epithelial cells of the blood-brain barrier and the brain-cerebrospinal fluid barrier, also termed the choroid plexus [35,36]. Three folate transporters have been identified that are involved in brain uptake of circulating folates: the folate receptor (FR)α, the reduced folate carrier (RFC) and the proton-coupled folate transporter. UMFA and 5-MTHF exhibit high-affinity binding to FRα, with the affinity of UMFA ~10-fold than that of 5-MTHF, whereas the affinity of 5-MTHF for the RFC is ~100-fold greater than that of UMFA [37]. Although the current study provides no evidence of the mechanism by which UMFA occurs in these infant monkey brains, transport via FRα may be the likely route. Presumably, this involves endocytosis of the FRα-bound UMFA, release, and subsequent transport by the proton-coupled folate transporter [36]. Consistent with this hypothesis, UMFA has been reported to inhibit transport of 5-MTHF across the blood-brain barrier [38]. Arguing against this hypothesis is the observation that UMFA binds with such high affinity to FRα that it may be considered to be irreversibly bound [39].

Is the UMFA detected in brain samples an artifact?

The oxidation of reduced folates, including THF and 5-MTHF, does not form FA [40,41]. THF degradation primarily involves oxidative cleavage to a pterin and p-aminobenzoylglutamate, and 5-MTHF involves several pathways of rearrangement. Thus, the UMFA detected in these samples should not be an artifact of folate degradation during frozen storage.

Are there any adverse effects of UMFA in brain?

As previously reported [29], high level measures of brain development were not different between the feeding groups. These assessments included daily veterinarian assessments of behavioral development, head circumference at regular intervals, and brain weight at 6 mo of age [29]. UMFA has no known metabolic function, but there is some evidence from cell-free studies that cellular UMFA may interact with some cellular enzymes. MTHFR isolated from rat brain homogenates was inhibited by both 5-MTHFR and FA [42], which is one aspect of MTHFR regulation. FA was shown to competitively inhibit rabbit brain purine nucleoside phosphorylase (PNP), an enzyme integral to the purine salvage pathway, in the presence of PNP’s natural substrate, guanosine [43]. The current research was not designed to address whether the tissue concentrations of UMFA reported in this study have any metabolic impact. Additional research is necessary to address that question.

A key limitation of this work is that this study was not designed to explore potential manifestations of UMFA in the tissues. In addition, the formyl, methenyl, and methylene folate forms were not measured due to their interconversion during analysis. This should be considered when comparing these results to other works that reported those folate forms. Additional limitations of this study include the number of animals studied, although this is generally typical of studies in nonhuman primates. Concentrations of UMFA in the dam’s milk (~164 pmol/mL) were substantially higher than those previously reported in human breast milk (~14–55 pmol/mL) [44-47]. It is not clear if this reflected the predominant intake of FA by the monkey dams (~1000 vs. 500 μg/d for a lactating woman) or a reduced ability to reduce FA compared with humans. Maternal intake of naturally occurring folates from supplemental lettuce and fruit, given to help provide fiber, was not quantified but we note their intake of FA, as noted earlier, was high. We note that comparisons across tissue types and brain regions were based on folate content per gram of tissue, and therefore, cell densities could differ for both sets of comparisons. Finally, DHFR and MTHFR activities were not determined in this study.

In conclusion, limited data are available describing tissue folate profiles for human tissues other than colonic mucosa, epidermis, and dermis. This study provides insight into folate profiles in a selection of tissues by leveraging a controlled feeding in species similar to humans. Tissue folate concentrations and profiles differed dramatically among the tissues studied. Similarly, although to a lesser extent, brain region concentrations and profiles also varied. Our findings are consistent with low capacity for FA reduction, as we detected UMFA in addition to THF and 5-MTHF in each sample. Notably, the lungs had 5-fold more UMFA than 5-MTHF, and both CB and ST had higher UMFA concentrations relative to other brain regions. These findings warrant further studies to examine the impact of UMFA on folate status and metabolism to understand the basis for the differences in folate profiles between tissues. In addition, it will be important to determine the concentration and distribution of folate species in plasma and tissues as a function of supplemental folate forms other than FA, such as D, L-folinic acid, levofolinate, and L-5-MTHF. To our knowledge, few such data are currently available, aside from a study in rats [48].

Abbreviations:

BMD

breastfed, mixed diet

CB

cerebellum

DHFR

dihydrofolate reductase

FA

folic acid

FR

folate receptor

MC

motor cortex

MR

milk replacer

MTHFR

methylene tetrahydrofolate reductase

OC

occipital cortex

PFC

prefrontal cortex

PNP

purine nucleoside phosphorylase

ST

striatum

TC

temporal cortex

THF

tetrahydrofolate

UMFA

unmetabolized folic acid

RFC

reduced folate carrier

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.tjnut.2025.101297.

Data availability

Data described in the manuscript will be made available upon request pending application and approval.